diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,45823 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.998968008255934,
+  "eval_steps": 500,
+  "global_step": 6540,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00045866299736268775,
+      "grad_norm": 11.983423233032227,
+      "learning_rate": 5.076142131979695e-06,
+      "loss": 3.3181,
+      "step": 1
+    },
+    {
+      "epoch": 0.0009173259947253755,
+      "grad_norm": 20.939693450927734,
+      "learning_rate": 1.015228426395939e-05,
+      "loss": 4.644,
+      "step": 2
+    },
+    {
+      "epoch": 0.0013759889920880633,
+      "grad_norm": 15.526450157165527,
+      "learning_rate": 1.5228426395939088e-05,
+      "loss": 3.3677,
+      "step": 3
+    },
+    {
+      "epoch": 0.001834651989450751,
+      "grad_norm": 15.404317855834961,
+      "learning_rate": 2.030456852791878e-05,
+      "loss": 4.1666,
+      "step": 4
+    },
+    {
+      "epoch": 0.002293314986813439,
+      "grad_norm": 7.165552616119385,
+      "learning_rate": 2.5380710659898476e-05,
+      "loss": 2.5469,
+      "step": 5
+    },
+    {
+      "epoch": 0.0027519779841761265,
+      "grad_norm": 2.2743093967437744,
+      "learning_rate": 3.0456852791878175e-05,
+      "loss": 1.4774,
+      "step": 6
+    },
+    {
+      "epoch": 0.0032106409815388145,
+      "grad_norm": 12.052650451660156,
+      "learning_rate": 3.553299492385787e-05,
+      "loss": 3.9692,
+      "step": 7
+    },
+    {
+      "epoch": 0.003669303978901502,
+      "grad_norm": 3.099673271179199,
+      "learning_rate": 4.060913705583756e-05,
+      "loss": 1.4465,
+      "step": 8
+    },
+    {
+      "epoch": 0.0041279669762641896,
+      "grad_norm": 5.82925271987915,
+      "learning_rate": 4.568527918781726e-05,
+      "loss": 3.3387,
+      "step": 9
+    },
+    {
+      "epoch": 0.004586629973626878,
+      "grad_norm": 4.784450531005859,
+      "learning_rate": 5.076142131979695e-05,
+      "loss": 2.6028,
+      "step": 10
+    },
+    {
+      "epoch": 0.0050452929709895655,
+      "grad_norm": 6.859081745147705,
+      "learning_rate": 5.583756345177665e-05,
+      "loss": 3.8062,
+      "step": 11
+    },
+    {
+      "epoch": 0.005503955968352253,
+      "grad_norm": 2.7776687145233154,
+      "learning_rate": 6.091370558375635e-05,
+      "loss": 2.3401,
+      "step": 12
+    },
+    {
+      "epoch": 0.0059626189657149406,
+      "grad_norm": 3.001207113265991,
+      "learning_rate": 6.598984771573605e-05,
+      "loss": 2.3286,
+      "step": 13
+    },
+    {
+      "epoch": 0.006421281963077629,
+      "grad_norm": 4.640460968017578,
+      "learning_rate": 7.106598984771574e-05,
+      "loss": 2.7322,
+      "step": 14
+    },
+    {
+      "epoch": 0.0068799449604403165,
+      "grad_norm": 3.0203094482421875,
+      "learning_rate": 7.614213197969543e-05,
+      "loss": 1.9691,
+      "step": 15
+    },
+    {
+      "epoch": 0.007338607957803004,
+      "grad_norm": 4.914385795593262,
+      "learning_rate": 8.121827411167512e-05,
+      "loss": 3.811,
+      "step": 16
+    },
+    {
+      "epoch": 0.007797270955165692,
+      "grad_norm": 1.8599748611450195,
+      "learning_rate": 8.629441624365482e-05,
+      "loss": 1.743,
+      "step": 17
+    },
+    {
+      "epoch": 0.008255933952528379,
+      "grad_norm": 4.672982215881348,
+      "learning_rate": 9.137055837563452e-05,
+      "loss": 2.9532,
+      "step": 18
+    },
+    {
+      "epoch": 0.008714596949891068,
+      "grad_norm": 2.165299654006958,
+      "learning_rate": 9.644670050761421e-05,
+      "loss": 2.6724,
+      "step": 19
+    },
+    {
+      "epoch": 0.009173259947253756,
+      "grad_norm": 2.6288881301879883,
+      "learning_rate": 0.0001015228426395939,
+      "loss": 3.1768,
+      "step": 20
+    },
+    {
+      "epoch": 0.009631922944616443,
+      "grad_norm": 3.7706544399261475,
+      "learning_rate": 0.00010659898477157361,
+      "loss": 2.6769,
+      "step": 21
+    },
+    {
+      "epoch": 0.010090585941979131,
+      "grad_norm": 1.690014362335205,
+      "learning_rate": 0.0001116751269035533,
+      "loss": 2.0408,
+      "step": 22
+    },
+    {
+      "epoch": 0.010549248939341819,
+      "grad_norm": 1.388680100440979,
+      "learning_rate": 0.000116751269035533,
+      "loss": 1.7454,
+      "step": 23
+    },
+    {
+      "epoch": 0.011007911936704506,
+      "grad_norm": 1.0925803184509277,
+      "learning_rate": 0.0001218274111675127,
+      "loss": 1.7225,
+      "step": 24
+    },
+    {
+      "epoch": 0.011466574934067194,
+      "grad_norm": 1.5495200157165527,
+      "learning_rate": 0.0001269035532994924,
+      "loss": 2.7841,
+      "step": 25
+    },
+    {
+      "epoch": 0.011925237931429881,
+      "grad_norm": 0.8375623822212219,
+      "learning_rate": 0.0001319796954314721,
+      "loss": 1.4273,
+      "step": 26
+    },
+    {
+      "epoch": 0.01238390092879257,
+      "grad_norm": 1.5786348581314087,
+      "learning_rate": 0.00013705583756345178,
+      "loss": 2.4528,
+      "step": 27
+    },
+    {
+      "epoch": 0.012842563926155258,
+      "grad_norm": 0.9938832521438599,
+      "learning_rate": 0.00014213197969543148,
+      "loss": 1.9099,
+      "step": 28
+    },
+    {
+      "epoch": 0.013301226923517945,
+      "grad_norm": 1.1450473070144653,
+      "learning_rate": 0.00014720812182741116,
+      "loss": 2.4707,
+      "step": 29
+    },
+    {
+      "epoch": 0.013759889920880633,
+      "grad_norm": 0.5717598795890808,
+      "learning_rate": 0.00015228426395939087,
+      "loss": 1.4649,
+      "step": 30
+    },
+    {
+      "epoch": 0.01421855291824332,
+      "grad_norm": 1.1892627477645874,
+      "learning_rate": 0.00015736040609137057,
+      "loss": 2.762,
+      "step": 31
+    },
+    {
+      "epoch": 0.014677215915606008,
+      "grad_norm": 0.8061103224754333,
+      "learning_rate": 0.00016243654822335025,
+      "loss": 2.5979,
+      "step": 32
+    },
+    {
+      "epoch": 0.015135878912968696,
+      "grad_norm": 1.37034010887146,
+      "learning_rate": 0.00016751269035532995,
+      "loss": 2.8462,
+      "step": 33
+    },
+    {
+      "epoch": 0.015594541910331383,
+      "grad_norm": 5.973114967346191,
+      "learning_rate": 0.00017258883248730963,
+      "loss": 1.8631,
+      "step": 34
+    },
+    {
+      "epoch": 0.01605320490769407,
+      "grad_norm": 1.1910300254821777,
+      "learning_rate": 0.00017766497461928934,
+      "loss": 2.1458,
+      "step": 35
+    },
+    {
+      "epoch": 0.016511867905056758,
+      "grad_norm": 0.8178476691246033,
+      "learning_rate": 0.00018274111675126904,
+      "loss": 2.6722,
+      "step": 36
+    },
+    {
+      "epoch": 0.016970530902419446,
+      "grad_norm": 0.6606747508049011,
+      "learning_rate": 0.00018781725888324875,
+      "loss": 2.0547,
+      "step": 37
+    },
+    {
+      "epoch": 0.017429193899782137,
+      "grad_norm": 0.5291973352432251,
+      "learning_rate": 0.00019289340101522843,
+      "loss": 1.2715,
+      "step": 38
+    },
+    {
+      "epoch": 0.017887856897144824,
+      "grad_norm": 1.221416711807251,
+      "learning_rate": 0.00019796954314720813,
+      "loss": 3.2524,
+      "step": 39
+    },
+    {
+      "epoch": 0.018346519894507512,
+      "grad_norm": 0.751660168170929,
+      "learning_rate": 0.0002030456852791878,
+      "loss": 2.2274,
+      "step": 40
+    },
+    {
+      "epoch": 0.0188051828918702,
+      "grad_norm": 0.6112833023071289,
+      "learning_rate": 0.00020812182741116754,
+      "loss": 1.6378,
+      "step": 41
+    },
+    {
+      "epoch": 0.019263845889232887,
+      "grad_norm": 0.40640926361083984,
+      "learning_rate": 0.00021319796954314722,
+      "loss": 1.395,
+      "step": 42
+    },
+    {
+      "epoch": 0.019722508886595574,
+      "grad_norm": 0.3838483989238739,
+      "learning_rate": 0.0002182741116751269,
+      "loss": 1.3842,
+      "step": 43
+    },
+    {
+      "epoch": 0.020181171883958262,
+      "grad_norm": 0.5843013525009155,
+      "learning_rate": 0.0002233502538071066,
+      "loss": 2.1636,
+      "step": 44
+    },
+    {
+      "epoch": 0.02063983488132095,
+      "grad_norm": 0.5792063474655151,
+      "learning_rate": 0.00022842639593908628,
+      "loss": 2.4495,
+      "step": 45
+    },
+    {
+      "epoch": 0.021098497878683637,
+      "grad_norm": 0.48895737528800964,
+      "learning_rate": 0.000233502538071066,
+      "loss": 2.0795,
+      "step": 46
+    },
+    {
+      "epoch": 0.021557160876046325,
+      "grad_norm": 0.6037505865097046,
+      "learning_rate": 0.0002385786802030457,
+      "loss": 2.6077,
+      "step": 47
+    },
+    {
+      "epoch": 0.022015823873409012,
+      "grad_norm": 0.11669456958770752,
+      "learning_rate": 0.0002436548223350254,
+      "loss": 0.535,
+      "step": 48
+    },
+    {
+      "epoch": 0.0224744868707717,
+      "grad_norm": 0.5221993327140808,
+      "learning_rate": 0.0002487309644670051,
+      "loss": 1.9393,
+      "step": 49
+    },
+    {
+      "epoch": 0.022933149868134387,
+      "grad_norm": 0.6841984391212463,
+      "learning_rate": 0.0002538071065989848,
+      "loss": 2.3887,
+      "step": 50
+    },
+    {
+      "epoch": 0.023391812865497075,
+      "grad_norm": 0.5550611019134521,
+      "learning_rate": 0.0002588832487309645,
+      "loss": 2.0587,
+      "step": 51
+    },
+    {
+      "epoch": 0.023850475862859762,
+      "grad_norm": 0.5314993858337402,
+      "learning_rate": 0.0002639593908629442,
+      "loss": 2.7202,
+      "step": 52
+    },
+    {
+      "epoch": 0.024309138860222453,
+      "grad_norm": 0.6494979858398438,
+      "learning_rate": 0.00026903553299492385,
+      "loss": 2.7383,
+      "step": 53
+    },
+    {
+      "epoch": 0.02476780185758514,
+      "grad_norm": 0.6311330795288086,
+      "learning_rate": 0.00027411167512690355,
+      "loss": 2.7266,
+      "step": 54
+    },
+    {
+      "epoch": 0.02522646485494783,
+      "grad_norm": 0.6023324728012085,
+      "learning_rate": 0.00027918781725888326,
+      "loss": 2.8926,
+      "step": 55
+    },
+    {
+      "epoch": 0.025685127852310516,
+      "grad_norm": 0.4306228458881378,
+      "learning_rate": 0.00028426395939086296,
+      "loss": 1.8117,
+      "step": 56
+    },
+    {
+      "epoch": 0.026143790849673203,
+      "grad_norm": 0.4600723385810852,
+      "learning_rate": 0.0002893401015228426,
+      "loss": 1.4924,
+      "step": 57
+    },
+    {
+      "epoch": 0.02660245384703589,
+      "grad_norm": 0.5145777463912964,
+      "learning_rate": 0.0002944162436548223,
+      "loss": 1.8133,
+      "step": 58
+    },
+    {
+      "epoch": 0.02706111684439858,
+      "grad_norm": 0.4451518654823303,
+      "learning_rate": 0.000299492385786802,
+      "loss": 1.453,
+      "step": 59
+    },
+    {
+      "epoch": 0.027519779841761266,
+      "grad_norm": 0.47992944717407227,
+      "learning_rate": 0.00030456852791878173,
+      "loss": 2.0383,
+      "step": 60
+    },
+    {
+      "epoch": 0.027978442839123954,
+      "grad_norm": 0.49620696902275085,
+      "learning_rate": 0.00030964467005076144,
+      "loss": 1.7496,
+      "step": 61
+    },
+    {
+      "epoch": 0.02843710583648664,
+      "grad_norm": 0.6371163725852966,
+      "learning_rate": 0.00031472081218274114,
+      "loss": 2.4441,
+      "step": 62
+    },
+    {
+      "epoch": 0.02889576883384933,
+      "grad_norm": 0.38818782567977905,
+      "learning_rate": 0.00031979695431472085,
+      "loss": 1.4037,
+      "step": 63
+    },
+    {
+      "epoch": 0.029354431831212016,
+      "grad_norm": 1.4600757360458374,
+      "learning_rate": 0.0003248730964467005,
+      "loss": 2.6814,
+      "step": 64
+    },
+    {
+      "epoch": 0.029813094828574704,
+      "grad_norm": 0.394088476896286,
+      "learning_rate": 0.0003299492385786802,
+      "loss": 1.3568,
+      "step": 65
+    },
+    {
+      "epoch": 0.03027175782593739,
+      "grad_norm": 1.3067290782928467,
+      "learning_rate": 0.0003350253807106599,
+      "loss": 2.6444,
+      "step": 66
+    },
+    {
+      "epoch": 0.03073042082330008,
+      "grad_norm": 1.0073161125183105,
+      "learning_rate": 0.0003401015228426396,
+      "loss": 2.074,
+      "step": 67
+    },
+    {
+      "epoch": 0.031189083820662766,
+      "grad_norm": 0.4958036243915558,
+      "learning_rate": 0.00034517766497461927,
+      "loss": 1.9275,
+      "step": 68
+    },
+    {
+      "epoch": 0.03164774681802546,
+      "grad_norm": 0.6386379599571228,
+      "learning_rate": 0.00035025380710659897,
+      "loss": 2.5089,
+      "step": 69
+    },
+    {
+      "epoch": 0.03210640981538814,
+      "grad_norm": 0.3506165146827698,
+      "learning_rate": 0.0003553299492385787,
+      "loss": 1.1556,
+      "step": 70
+    },
+    {
+      "epoch": 0.03256507281275083,
+      "grad_norm": 0.511293888092041,
+      "learning_rate": 0.0003604060913705584,
+      "loss": 1.7764,
+      "step": 71
+    },
+    {
+      "epoch": 0.033023735810113516,
+      "grad_norm": 0.5064987540245056,
+      "learning_rate": 0.0003654822335025381,
+      "loss": 2.0087,
+      "step": 72
+    },
+    {
+      "epoch": 0.03348239880747621,
+      "grad_norm": 0.6523762345314026,
+      "learning_rate": 0.0003705583756345178,
+      "loss": 2.1339,
+      "step": 73
+    },
+    {
+      "epoch": 0.03394106180483889,
+      "grad_norm": 0.7642560005187988,
+      "learning_rate": 0.0003756345177664975,
+      "loss": 2.7424,
+      "step": 74
+    },
+    {
+      "epoch": 0.03439972480220158,
+      "grad_norm": 0.8428514003753662,
+      "learning_rate": 0.00038071065989847715,
+      "loss": 3.062,
+      "step": 75
+    },
+    {
+      "epoch": 0.034858387799564274,
+      "grad_norm": 0.5961275696754456,
+      "learning_rate": 0.00038578680203045685,
+      "loss": 1.8469,
+      "step": 76
+    },
+    {
+      "epoch": 0.03531705079692696,
+      "grad_norm": 0.7156499624252319,
+      "learning_rate": 0.00039086294416243656,
+      "loss": 2.3454,
+      "step": 77
+    },
+    {
+      "epoch": 0.03577571379428965,
+      "grad_norm": 0.4211914837360382,
+      "learning_rate": 0.00039593908629441627,
+      "loss": 1.3773,
+      "step": 78
+    },
+    {
+      "epoch": 0.03623437679165233,
+      "grad_norm": 0.25154200196266174,
+      "learning_rate": 0.0004010152284263959,
+      "loss": 0.834,
+      "step": 79
+    },
+    {
+      "epoch": 0.036693039789015024,
+      "grad_norm": 0.7084411978721619,
+      "learning_rate": 0.0004060913705583756,
+      "loss": 2.5184,
+      "step": 80
+    },
+    {
+      "epoch": 0.03715170278637771,
+      "grad_norm": 0.18035106360912323,
+      "learning_rate": 0.00041116751269035533,
+      "loss": 0.7498,
+      "step": 81
+    },
+    {
+      "epoch": 0.0376103657837404,
+      "grad_norm": 0.6386528611183167,
+      "learning_rate": 0.0004162436548223351,
+      "loss": 2.762,
+      "step": 82
+    },
+    {
+      "epoch": 0.03806902878110308,
+      "grad_norm": 0.5148580074310303,
+      "learning_rate": 0.00042131979695431474,
+      "loss": 2.1558,
+      "step": 83
+    },
+    {
+      "epoch": 0.038527691778465774,
+      "grad_norm": 0.5644561052322388,
+      "learning_rate": 0.00042639593908629444,
+      "loss": 2.4407,
+      "step": 84
+    },
+    {
+      "epoch": 0.03898635477582846,
+      "grad_norm": 0.21473997831344604,
+      "learning_rate": 0.00043147208121827415,
+      "loss": 0.9734,
+      "step": 85
+    },
+    {
+      "epoch": 0.03944501777319115,
+      "grad_norm": 0.29117411375045776,
+      "learning_rate": 0.0004365482233502538,
+      "loss": 0.8773,
+      "step": 86
+    },
+    {
+      "epoch": 0.03990368077055383,
+      "grad_norm": 0.5848171710968018,
+      "learning_rate": 0.0004416243654822335,
+      "loss": 1.5979,
+      "step": 87
+    },
+    {
+      "epoch": 0.040362343767916524,
+      "grad_norm": 0.7745836973190308,
+      "learning_rate": 0.0004467005076142132,
+      "loss": 2.6626,
+      "step": 88
+    },
+    {
+      "epoch": 0.04082100676527921,
+      "grad_norm": 0.6489730477333069,
+      "learning_rate": 0.0004517766497461929,
+      "loss": 2.2454,
+      "step": 89
+    },
+    {
+      "epoch": 0.0412796697626419,
+      "grad_norm": 0.38306012749671936,
+      "learning_rate": 0.00045685279187817257,
+      "loss": 1.1562,
+      "step": 90
+    },
+    {
+      "epoch": 0.04173833276000459,
+      "grad_norm": 0.6314185261726379,
+      "learning_rate": 0.0004619289340101523,
+      "loss": 2.3352,
+      "step": 91
+    },
+    {
+      "epoch": 0.042196995757367274,
+      "grad_norm": 0.5400259494781494,
+      "learning_rate": 0.000467005076142132,
+      "loss": 2.261,
+      "step": 92
+    },
+    {
+      "epoch": 0.042655658754729965,
+      "grad_norm": 0.5282353758811951,
+      "learning_rate": 0.00047208121827411174,
+      "loss": 1.9561,
+      "step": 93
+    },
+    {
+      "epoch": 0.04311432175209265,
+      "grad_norm": 0.6686978936195374,
+      "learning_rate": 0.0004771573604060914,
+      "loss": 2.3158,
+      "step": 94
+    },
+    {
+      "epoch": 0.04357298474945534,
+      "grad_norm": 0.5369504690170288,
+      "learning_rate": 0.0004822335025380711,
+      "loss": 2.0291,
+      "step": 95
+    },
+    {
+      "epoch": 0.044031647746818024,
+      "grad_norm": 0.34599074721336365,
+      "learning_rate": 0.0004873096446700508,
+      "loss": 1.1915,
+      "step": 96
+    },
+    {
+      "epoch": 0.044490310744180715,
+      "grad_norm": 0.45278364419937134,
+      "learning_rate": 0.0004923857868020305,
+      "loss": 2.0,
+      "step": 97
+    },
+    {
+      "epoch": 0.0449489737415434,
+      "grad_norm": 0.7745117545127869,
+      "learning_rate": 0.0004974619289340102,
+      "loss": 2.5948,
+      "step": 98
+    },
+    {
+      "epoch": 0.04540763673890609,
+      "grad_norm": 0.3228628635406494,
+      "learning_rate": 0.0005025380710659899,
+      "loss": 1.236,
+      "step": 99
+    },
+    {
+      "epoch": 0.045866299736268774,
+      "grad_norm": 0.49457353353500366,
+      "learning_rate": 0.0005076142131979696,
+      "loss": 1.9651,
+      "step": 100
+    },
+    {
+      "epoch": 0.046324962733631465,
+      "grad_norm": 0.6033787727355957,
+      "learning_rate": 0.0005126903553299493,
+      "loss": 2.657,
+      "step": 101
+    },
+    {
+      "epoch": 0.04678362573099415,
+      "grad_norm": 0.529002845287323,
+      "learning_rate": 0.000517766497461929,
+      "loss": 2.249,
+      "step": 102
+    },
+    {
+      "epoch": 0.04724228872835684,
+      "grad_norm": 0.3869992792606354,
+      "learning_rate": 0.0005228426395939087,
+      "loss": 1.3741,
+      "step": 103
+    },
+    {
+      "epoch": 0.047700951725719525,
+      "grad_norm": 0.4783024787902832,
+      "learning_rate": 0.0005279187817258884,
+      "loss": 1.8258,
+      "step": 104
+    },
+    {
+      "epoch": 0.048159614723082216,
+      "grad_norm": 0.5782725214958191,
+      "learning_rate": 0.0005329949238578681,
+      "loss": 2.4186,
+      "step": 105
+    },
+    {
+      "epoch": 0.04861827772044491,
+      "grad_norm": 0.5999312400817871,
+      "learning_rate": 0.0005380710659898477,
+      "loss": 2.6179,
+      "step": 106
+    },
+    {
+      "epoch": 0.04907694071780759,
+      "grad_norm": 0.6067392826080322,
+      "learning_rate": 0.0005431472081218274,
+      "loss": 2.1979,
+      "step": 107
+    },
+    {
+      "epoch": 0.04953560371517028,
+      "grad_norm": 0.5408378839492798,
+      "learning_rate": 0.0005482233502538071,
+      "loss": 1.9679,
+      "step": 108
+    },
+    {
+      "epoch": 0.049994266712532966,
+      "grad_norm": 0.8069329857826233,
+      "learning_rate": 0.0005532994923857868,
+      "loss": 2.8796,
+      "step": 109
+    },
+    {
+      "epoch": 0.05045292970989566,
+      "grad_norm": 0.7999544143676758,
+      "learning_rate": 0.0005583756345177665,
+      "loss": 2.3101,
+      "step": 110
+    },
+    {
+      "epoch": 0.05091159270725834,
+      "grad_norm": 0.8346303701400757,
+      "learning_rate": 0.0005634517766497462,
+      "loss": 3.0054,
+      "step": 111
+    },
+    {
+      "epoch": 0.05137025570462103,
+      "grad_norm": 0.4418480396270752,
+      "learning_rate": 0.0005685279187817259,
+      "loss": 1.3883,
+      "step": 112
+    },
+    {
+      "epoch": 0.051828918701983716,
+      "grad_norm": 0.6058046221733093,
+      "learning_rate": 0.0005736040609137056,
+      "loss": 1.7998,
+      "step": 113
+    },
+    {
+      "epoch": 0.05228758169934641,
+      "grad_norm": 0.4952194392681122,
+      "learning_rate": 0.0005786802030456852,
+      "loss": 1.7445,
+      "step": 114
+    },
+    {
+      "epoch": 0.05274624469670909,
+      "grad_norm": 0.6057190895080566,
+      "learning_rate": 0.0005837563451776649,
+      "loss": 1.9283,
+      "step": 115
+    },
+    {
+      "epoch": 0.05320490769407178,
+      "grad_norm": 0.4411911964416504,
+      "learning_rate": 0.0005888324873096446,
+      "loss": 1.3921,
+      "step": 116
+    },
+    {
+      "epoch": 0.053663570691434466,
+      "grad_norm": 0.7016524076461792,
+      "learning_rate": 0.0005939086294416243,
+      "loss": 2.9248,
+      "step": 117
+    },
+    {
+      "epoch": 0.05412223368879716,
+      "grad_norm": 0.7117334008216858,
+      "learning_rate": 0.000598984771573604,
+      "loss": 2.7302,
+      "step": 118
+    },
+    {
+      "epoch": 0.05458089668615984,
+      "grad_norm": 0.49954333901405334,
+      "learning_rate": 0.0006040609137055838,
+      "loss": 1.8414,
+      "step": 119
+    },
+    {
+      "epoch": 0.05503955968352253,
+      "grad_norm": 0.6712334156036377,
+      "learning_rate": 0.0006091370558375635,
+      "loss": 2.4685,
+      "step": 120
+    },
+    {
+      "epoch": 0.05549822268088522,
+      "grad_norm": 0.37909215688705444,
+      "learning_rate": 0.0006142131979695432,
+      "loss": 1.1743,
+      "step": 121
+    },
+    {
+      "epoch": 0.05595688567824791,
+      "grad_norm": 0.42544159293174744,
+      "learning_rate": 0.0006192893401015229,
+      "loss": 1.4775,
+      "step": 122
+    },
+    {
+      "epoch": 0.0564155486756106,
+      "grad_norm": 0.5652533769607544,
+      "learning_rate": 0.0006243654822335026,
+      "loss": 2.1067,
+      "step": 123
+    },
+    {
+      "epoch": 0.05687421167297328,
+      "grad_norm": 0.4516564607620239,
+      "learning_rate": 0.0006294416243654823,
+      "loss": 1.7084,
+      "step": 124
+    },
+    {
+      "epoch": 0.05733287467033597,
+      "grad_norm": 0.6015462279319763,
+      "learning_rate": 0.000634517766497462,
+      "loss": 2.3374,
+      "step": 125
+    },
+    {
+      "epoch": 0.05779153766769866,
+      "grad_norm": 0.5467323660850525,
+      "learning_rate": 0.0006395939086294417,
+      "loss": 2.2559,
+      "step": 126
+    },
+    {
+      "epoch": 0.05825020066506135,
+      "grad_norm": 0.5923335552215576,
+      "learning_rate": 0.0006446700507614214,
+      "loss": 2.4232,
+      "step": 127
+    },
+    {
+      "epoch": 0.05870886366242403,
+      "grad_norm": 0.6827653050422668,
+      "learning_rate": 0.000649746192893401,
+      "loss": 2.7278,
+      "step": 128
+    },
+    {
+      "epoch": 0.05916752665978672,
+      "grad_norm": 0.30980947613716125,
+      "learning_rate": 0.0006548223350253807,
+      "loss": 1.019,
+      "step": 129
+    },
+    {
+      "epoch": 0.05962618965714941,
+      "grad_norm": 0.6105501651763916,
+      "learning_rate": 0.0006598984771573604,
+      "loss": 2.1824,
+      "step": 130
+    },
+    {
+      "epoch": 0.0600848526545121,
+      "grad_norm": 0.292199969291687,
+      "learning_rate": 0.0006649746192893401,
+      "loss": 1.0942,
+      "step": 131
+    },
+    {
+      "epoch": 0.06054351565187478,
+      "grad_norm": 0.6059619188308716,
+      "learning_rate": 0.0006700507614213198,
+      "loss": 2.2701,
+      "step": 132
+    },
+    {
+      "epoch": 0.06100217864923747,
+      "grad_norm": 0.27790480852127075,
+      "learning_rate": 0.0006751269035532995,
+      "loss": 1.1452,
+      "step": 133
+    },
+    {
+      "epoch": 0.06146084164660016,
+      "grad_norm": 0.47722020745277405,
+      "learning_rate": 0.0006802030456852792,
+      "loss": 2.2821,
+      "step": 134
+    },
+    {
+      "epoch": 0.06191950464396285,
+      "grad_norm": 0.6286266446113586,
+      "learning_rate": 0.0006852791878172588,
+      "loss": 2.3686,
+      "step": 135
+    },
+    {
+      "epoch": 0.06237816764132553,
+      "grad_norm": 0.6163983345031738,
+      "learning_rate": 0.0006903553299492385,
+      "loss": 2.3126,
+      "step": 136
+    },
+    {
+      "epoch": 0.06283683063868822,
+      "grad_norm": 0.5008450150489807,
+      "learning_rate": 0.0006954314720812182,
+      "loss": 1.8563,
+      "step": 137
+    },
+    {
+      "epoch": 0.06329549363605091,
+      "grad_norm": 0.6025516986846924,
+      "learning_rate": 0.0007005076142131979,
+      "loss": 2.11,
+      "step": 138
+    },
+    {
+      "epoch": 0.0637541566334136,
+      "grad_norm": 0.4481363892555237,
+      "learning_rate": 0.0007055837563451776,
+      "loss": 1.8222,
+      "step": 139
+    },
+    {
+      "epoch": 0.06421281963077628,
+      "grad_norm": 0.5932562351226807,
+      "learning_rate": 0.0007106598984771574,
+      "loss": 2.28,
+      "step": 140
+    },
+    {
+      "epoch": 0.06467148262813897,
+      "grad_norm": 0.3764328062534332,
+      "learning_rate": 0.0007157360406091371,
+      "loss": 1.6057,
+      "step": 141
+    },
+    {
+      "epoch": 0.06513014562550166,
+      "grad_norm": 0.5842316746711731,
+      "learning_rate": 0.0007208121827411168,
+      "loss": 2.1699,
+      "step": 142
+    },
+    {
+      "epoch": 0.06558880862286436,
+      "grad_norm": 0.3900845944881439,
+      "learning_rate": 0.0007258883248730965,
+      "loss": 1.5842,
+      "step": 143
+    },
+    {
+      "epoch": 0.06604747162022703,
+      "grad_norm": 0.19777247309684753,
+      "learning_rate": 0.0007309644670050762,
+      "loss": 0.7936,
+      "step": 144
+    },
+    {
+      "epoch": 0.06650613461758972,
+      "grad_norm": 0.1068076640367508,
+      "learning_rate": 0.0007360406091370559,
+      "loss": 0.4987,
+      "step": 145
+    },
+    {
+      "epoch": 0.06696479761495241,
+      "grad_norm": 0.5474305152893066,
+      "learning_rate": 0.0007411167512690356,
+      "loss": 2.032,
+      "step": 146
+    },
+    {
+      "epoch": 0.0674234606123151,
+      "grad_norm": 0.6003263592720032,
+      "learning_rate": 0.0007461928934010153,
+      "loss": 2.3611,
+      "step": 147
+    },
+    {
+      "epoch": 0.06788212360967778,
+      "grad_norm": 0.3899264633655548,
+      "learning_rate": 0.000751269035532995,
+      "loss": 1.4559,
+      "step": 148
+    },
+    {
+      "epoch": 0.06834078660704047,
+      "grad_norm": 0.4279090464115143,
+      "learning_rate": 0.0007563451776649747,
+      "loss": 1.479,
+      "step": 149
+    },
+    {
+      "epoch": 0.06879944960440317,
+      "grad_norm": 0.4912429749965668,
+      "learning_rate": 0.0007614213197969543,
+      "loss": 1.6797,
+      "step": 150
+    },
+    {
+      "epoch": 0.06925811260176586,
+      "grad_norm": 0.23800012469291687,
+      "learning_rate": 0.000766497461928934,
+      "loss": 0.8729,
+      "step": 151
+    },
+    {
+      "epoch": 0.06971677559912855,
+      "grad_norm": 0.6027427315711975,
+      "learning_rate": 0.0007715736040609137,
+      "loss": 1.8105,
+      "step": 152
+    },
+    {
+      "epoch": 0.07017543859649122,
+      "grad_norm": 0.6931683421134949,
+      "learning_rate": 0.0007766497461928934,
+      "loss": 2.2175,
+      "step": 153
+    },
+    {
+      "epoch": 0.07063410159385392,
+      "grad_norm": 0.4811568558216095,
+      "learning_rate": 0.0007817258883248731,
+      "loss": 1.689,
+      "step": 154
+    },
+    {
+      "epoch": 0.0710927645912166,
+      "grad_norm": 0.6901246905326843,
+      "learning_rate": 0.0007868020304568528,
+      "loss": 1.6627,
+      "step": 155
+    },
+    {
+      "epoch": 0.0715514275885793,
+      "grad_norm": 12.634415626525879,
+      "learning_rate": 0.0007918781725888325,
+      "loss": 2.9146,
+      "step": 156
+    },
+    {
+      "epoch": 0.07201009058594197,
+      "grad_norm": 0.3894835114479065,
+      "learning_rate": 0.0007969543147208121,
+      "loss": 1.461,
+      "step": 157
+    },
+    {
+      "epoch": 0.07246875358330467,
+      "grad_norm": 0.4583110511302948,
+      "learning_rate": 0.0008020304568527918,
+      "loss": 1.3597,
+      "step": 158
+    },
+    {
+      "epoch": 0.07292741658066736,
+      "grad_norm": 0.5084859728813171,
+      "learning_rate": 0.0008071065989847715,
+      "loss": 2.131,
+      "step": 159
+    },
+    {
+      "epoch": 0.07338607957803005,
+      "grad_norm": 0.6448908448219299,
+      "learning_rate": 0.0008121827411167512,
+      "loss": 2.3026,
+      "step": 160
+    },
+    {
+      "epoch": 0.07384474257539272,
+      "grad_norm": 0.4896964728832245,
+      "learning_rate": 0.000817258883248731,
+      "loss": 2.0397,
+      "step": 161
+    },
+    {
+      "epoch": 0.07430340557275542,
+      "grad_norm": 0.6146819591522217,
+      "learning_rate": 0.0008223350253807107,
+      "loss": 2.1047,
+      "step": 162
+    },
+    {
+      "epoch": 0.0747620685701181,
+      "grad_norm": 0.2950705587863922,
+      "learning_rate": 0.0008274111675126904,
+      "loss": 1.1177,
+      "step": 163
+    },
+    {
+      "epoch": 0.0752207315674808,
+      "grad_norm": 0.39602458477020264,
+      "learning_rate": 0.0008324873096446702,
+      "loss": 1.6326,
+      "step": 164
+    },
+    {
+      "epoch": 0.07567939456484347,
+      "grad_norm": 0.46519267559051514,
+      "learning_rate": 0.0008375634517766498,
+      "loss": 1.7606,
+      "step": 165
+    },
+    {
+      "epoch": 0.07613805756220617,
+      "grad_norm": 0.31276392936706543,
+      "learning_rate": 0.0008426395939086295,
+      "loss": 1.2241,
+      "step": 166
+    },
+    {
+      "epoch": 0.07659672055956886,
+      "grad_norm": 0.48499414324760437,
+      "learning_rate": 0.0008477157360406092,
+      "loss": 1.9176,
+      "step": 167
+    },
+    {
+      "epoch": 0.07705538355693155,
+      "grad_norm": 0.43687963485717773,
+      "learning_rate": 0.0008527918781725889,
+      "loss": 1.6424,
+      "step": 168
+    },
+    {
+      "epoch": 0.07751404655429424,
+      "grad_norm": 0.4287634491920471,
+      "learning_rate": 0.0008578680203045686,
+      "loss": 1.8438,
+      "step": 169
+    },
+    {
+      "epoch": 0.07797270955165692,
+      "grad_norm": 0.28460821509361267,
+      "learning_rate": 0.0008629441624365483,
+      "loss": 1.1942,
+      "step": 170
+    },
+    {
+      "epoch": 0.0784313725490196,
+      "grad_norm": 0.30606964230537415,
+      "learning_rate": 0.000868020304568528,
+      "loss": 1.1019,
+      "step": 171
+    },
+    {
+      "epoch": 0.0788900355463823,
+      "grad_norm": 0.38955485820770264,
+      "learning_rate": 0.0008730964467005076,
+      "loss": 1.7985,
+      "step": 172
+    },
+    {
+      "epoch": 0.07934869854374499,
+      "grad_norm": 0.41820240020751953,
+      "learning_rate": 0.0008781725888324873,
+      "loss": 1.6667,
+      "step": 173
+    },
+    {
+      "epoch": 0.07980736154110767,
+      "grad_norm": 0.47719693183898926,
+      "learning_rate": 0.000883248730964467,
+      "loss": 2.0426,
+      "step": 174
+    },
+    {
+      "epoch": 0.08026602453847036,
+      "grad_norm": 0.49073362350463867,
+      "learning_rate": 0.0008883248730964467,
+      "loss": 2.0519,
+      "step": 175
+    },
+    {
+      "epoch": 0.08072468753583305,
+      "grad_norm": 0.4831169843673706,
+      "learning_rate": 0.0008934010152284264,
+      "loss": 2.511,
+      "step": 176
+    },
+    {
+      "epoch": 0.08118335053319574,
+      "grad_norm": 0.549760103225708,
+      "learning_rate": 0.0008984771573604061,
+      "loss": 2.2359,
+      "step": 177
+    },
+    {
+      "epoch": 0.08164201353055842,
+      "grad_norm": 0.5303396582603455,
+      "learning_rate": 0.0009035532994923858,
+      "loss": 2.5146,
+      "step": 178
+    },
+    {
+      "epoch": 0.08210067652792111,
+      "grad_norm": 0.47503864765167236,
+      "learning_rate": 0.0009086294416243654,
+      "loss": 2.0219,
+      "step": 179
+    },
+    {
+      "epoch": 0.0825593395252838,
+      "grad_norm": 0.4175845682621002,
+      "learning_rate": 0.0009137055837563451,
+      "loss": 1.9651,
+      "step": 180
+    },
+    {
+      "epoch": 0.08301800252264649,
+      "grad_norm": 0.536845862865448,
+      "learning_rate": 0.0009187817258883248,
+      "loss": 2.4031,
+      "step": 181
+    },
+    {
+      "epoch": 0.08347666552000918,
+      "grad_norm": 0.6141538023948669,
+      "learning_rate": 0.0009238578680203045,
+      "loss": 2.6653,
+      "step": 182
+    },
+    {
+      "epoch": 0.08393532851737186,
+      "grad_norm": 0.31810981035232544,
+      "learning_rate": 0.0009289340101522843,
+      "loss": 1.1417,
+      "step": 183
+    },
+    {
+      "epoch": 0.08439399151473455,
+      "grad_norm": 0.4755139648914337,
+      "learning_rate": 0.000934010152284264,
+      "loss": 2.1119,
+      "step": 184
+    },
+    {
+      "epoch": 0.08485265451209724,
+      "grad_norm": 0.14865882694721222,
+      "learning_rate": 0.0009390862944162437,
+      "loss": 0.6605,
+      "step": 185
+    },
+    {
+      "epoch": 0.08531131750945993,
+      "grad_norm": 0.31906870007514954,
+      "learning_rate": 0.0009441624365482235,
+      "loss": 1.2437,
+      "step": 186
+    },
+    {
+      "epoch": 0.08576998050682261,
+      "grad_norm": 0.5705398917198181,
+      "learning_rate": 0.0009492385786802031,
+      "loss": 2.3323,
+      "step": 187
+    },
+    {
+      "epoch": 0.0862286435041853,
+      "grad_norm": 0.3601418137550354,
+      "learning_rate": 0.0009543147208121828,
+      "loss": 1.5806,
+      "step": 188
+    },
+    {
+      "epoch": 0.08668730650154799,
+      "grad_norm": 0.4881735146045685,
+      "learning_rate": 0.0009593908629441625,
+      "loss": 1.9485,
+      "step": 189
+    },
+    {
+      "epoch": 0.08714596949891068,
+      "grad_norm": 0.15371385216712952,
+      "learning_rate": 0.0009644670050761422,
+      "loss": 0.683,
+      "step": 190
+    },
+    {
+      "epoch": 0.08760463249627336,
+      "grad_norm": 0.41688668727874756,
+      "learning_rate": 0.0009695431472081219,
+      "loss": 1.6802,
+      "step": 191
+    },
+    {
+      "epoch": 0.08806329549363605,
+      "grad_norm": 0.429606556892395,
+      "learning_rate": 0.0009746192893401016,
+      "loss": 1.9552,
+      "step": 192
+    },
+    {
+      "epoch": 0.08852195849099874,
+      "grad_norm": 0.549892008304596,
+      "learning_rate": 0.0009796954314720812,
+      "loss": 2.1486,
+      "step": 193
+    },
+    {
+      "epoch": 0.08898062148836143,
+      "grad_norm": 0.3614405393600464,
+      "learning_rate": 0.000984771573604061,
+      "loss": 1.3867,
+      "step": 194
+    },
+    {
+      "epoch": 0.08943928448572411,
+      "grad_norm": 0.43819788098335266,
+      "learning_rate": 0.0009898477157360406,
+      "loss": 1.8317,
+      "step": 195
+    },
+    {
+      "epoch": 0.0898979474830868,
+      "grad_norm": 0.14601579308509827,
+      "learning_rate": 0.0009949238578680203,
+      "loss": 0.6363,
+      "step": 196
+    },
+    {
+      "epoch": 0.09035661048044949,
+      "grad_norm": 0.2820906937122345,
+      "learning_rate": 0.001,
+      "loss": 1.0806,
+      "step": 197
+    },
+    {
+      "epoch": 0.09081527347781218,
+      "grad_norm": 0.41677334904670715,
+      "learning_rate": 0.0009999999386731974,
+      "loss": 1.79,
+      "step": 198
+    },
+    {
+      "epoch": 0.09127393647517487,
+      "grad_norm": 0.3319721519947052,
+      "learning_rate": 0.0009999997546928047,
+      "loss": 1.4927,
+      "step": 199
+    },
+    {
+      "epoch": 0.09173259947253755,
+      "grad_norm": 0.37898340821266174,
+      "learning_rate": 0.000999999448058867,
+      "loss": 1.6826,
+      "step": 200
+    },
+    {
+      "epoch": 0.09219126246990024,
+      "grad_norm": 0.5117692351341248,
+      "learning_rate": 0.0009999990187714593,
+      "loss": 2.368,
+      "step": 201
+    },
+    {
+      "epoch": 0.09264992546726293,
+      "grad_norm": 0.44219571352005005,
+      "learning_rate": 0.0009999984668306874,
+      "loss": 2.3706,
+      "step": 202
+    },
+    {
+      "epoch": 0.09310858846462562,
+      "grad_norm": 0.38528987765312195,
+      "learning_rate": 0.0009999977922366863,
+      "loss": 1.6523,
+      "step": 203
+    },
+    {
+      "epoch": 0.0935672514619883,
+      "grad_norm": 0.5601181983947754,
+      "learning_rate": 0.0009999969949896215,
+      "loss": 2.0021,
+      "step": 204
+    },
+    {
+      "epoch": 0.09402591445935099,
+      "grad_norm": 0.34919580817222595,
+      "learning_rate": 0.0009999960750896888,
+      "loss": 1.5776,
+      "step": 205
+    },
+    {
+      "epoch": 0.09448457745671368,
+      "grad_norm": 0.5601129531860352,
+      "learning_rate": 0.0009999950325371137,
+      "loss": 2.25,
+      "step": 206
+    },
+    {
+      "epoch": 0.09494324045407637,
+      "grad_norm": 0.5060522556304932,
+      "learning_rate": 0.0009999938673321519,
+      "loss": 2.4441,
+      "step": 207
+    },
+    {
+      "epoch": 0.09540190345143905,
+      "grad_norm": 0.44873717427253723,
+      "learning_rate": 0.0009999925794750893,
+      "loss": 1.6166,
+      "step": 208
+    },
+    {
+      "epoch": 0.09586056644880174,
+      "grad_norm": 0.1810014545917511,
+      "learning_rate": 0.0009999911689662418,
+      "loss": 0.7684,
+      "step": 209
+    },
+    {
+      "epoch": 0.09631922944616443,
+      "grad_norm": 0.3902459144592285,
+      "learning_rate": 0.0009999896358059556,
+      "loss": 1.5992,
+      "step": 210
+    },
+    {
+      "epoch": 0.09677789244352712,
+      "grad_norm": 0.6398706436157227,
+      "learning_rate": 0.0009999879799946067,
+      "loss": 2.5659,
+      "step": 211
+    },
+    {
+      "epoch": 0.09723655544088981,
+      "grad_norm": 0.6179486513137817,
+      "learning_rate": 0.000999986201532601,
+      "loss": 2.6672,
+      "step": 212
+    },
+    {
+      "epoch": 0.09769521843825249,
+      "grad_norm": 0.46587255597114563,
+      "learning_rate": 0.0009999843004203748,
+      "loss": 1.9471,
+      "step": 213
+    },
+    {
+      "epoch": 0.09815388143561518,
+      "grad_norm": 0.5521425604820251,
+      "learning_rate": 0.0009999822766583947,
+      "loss": 2.51,
+      "step": 214
+    },
+    {
+      "epoch": 0.09861254443297787,
+      "grad_norm": 0.43424972891807556,
+      "learning_rate": 0.0009999801302471574,
+      "loss": 1.8136,
+      "step": 215
+    },
+    {
+      "epoch": 0.09907120743034056,
+      "grad_norm": 0.3763306140899658,
+      "learning_rate": 0.0009999778611871888,
+      "loss": 1.3743,
+      "step": 216
+    },
+    {
+      "epoch": 0.09952987042770324,
+      "grad_norm": 0.34321171045303345,
+      "learning_rate": 0.0009999754694790459,
+      "loss": 1.4966,
+      "step": 217
+    },
+    {
+      "epoch": 0.09998853342506593,
+      "grad_norm": 0.46348294615745544,
+      "learning_rate": 0.0009999729551233155,
+      "loss": 1.6602,
+      "step": 218
+    },
+    {
+      "epoch": 0.10044719642242862,
+      "grad_norm": 0.5280048847198486,
+      "learning_rate": 0.000999970318120614,
+      "loss": 2.2937,
+      "step": 219
+    },
+    {
+      "epoch": 0.10090585941979131,
+      "grad_norm": 0.4680062234401703,
+      "learning_rate": 0.0009999675584715887,
+      "loss": 1.999,
+      "step": 220
+    },
+    {
+      "epoch": 0.10136452241715399,
+      "grad_norm": 0.48422807455062866,
+      "learning_rate": 0.0009999646761769162,
+      "loss": 2.1548,
+      "step": 221
+    },
+    {
+      "epoch": 0.10182318541451668,
+      "grad_norm": 0.30261266231536865,
+      "learning_rate": 0.0009999616712373035,
+      "loss": 1.3619,
+      "step": 222
+    },
+    {
+      "epoch": 0.10228184841187937,
+      "grad_norm": 0.549519956111908,
+      "learning_rate": 0.0009999585436534883,
+      "loss": 2.4467,
+      "step": 223
+    },
+    {
+      "epoch": 0.10274051140924206,
+      "grad_norm": 0.4387172758579254,
+      "learning_rate": 0.0009999552934262374,
+      "loss": 2.0479,
+      "step": 224
+    },
+    {
+      "epoch": 0.10319917440660474,
+      "grad_norm": 0.49094218015670776,
+      "learning_rate": 0.0009999519205563482,
+      "loss": 2.4268,
+      "step": 225
+    },
+    {
+      "epoch": 0.10365783740396743,
+      "grad_norm": 0.5436589121818542,
+      "learning_rate": 0.0009999484250446478,
+      "loss": 2.5486,
+      "step": 226
+    },
+    {
+      "epoch": 0.10411650040133012,
+      "grad_norm": 0.5063910484313965,
+      "learning_rate": 0.000999944806891994,
+      "loss": 2.364,
+      "step": 227
+    },
+    {
+      "epoch": 0.10457516339869281,
+      "grad_norm": 0.5463952422142029,
+      "learning_rate": 0.0009999410660992743,
+      "loss": 2.6033,
+      "step": 228
+    },
+    {
+      "epoch": 0.1050338263960555,
+      "grad_norm": 0.507409393787384,
+      "learning_rate": 0.0009999372026674063,
+      "loss": 2.32,
+      "step": 229
+    },
+    {
+      "epoch": 0.10549248939341818,
+      "grad_norm": 0.5893042087554932,
+      "learning_rate": 0.0009999332165973379,
+      "loss": 2.7153,
+      "step": 230
+    },
+    {
+      "epoch": 0.10595115239078087,
+      "grad_norm": 0.284311980009079,
+      "learning_rate": 0.0009999291078900466,
+      "loss": 0.9774,
+      "step": 231
+    },
+    {
+      "epoch": 0.10640981538814356,
+      "grad_norm": 0.42511793971061707,
+      "learning_rate": 0.0009999248765465406,
+      "loss": 1.924,
+      "step": 232
+    },
+    {
+      "epoch": 0.10686847838550625,
+      "grad_norm": 0.6010558605194092,
+      "learning_rate": 0.0009999205225678575,
+      "loss": 2.7175,
+      "step": 233
+    },
+    {
+      "epoch": 0.10732714138286893,
+      "grad_norm": 0.3992350697517395,
+      "learning_rate": 0.000999916045955066,
+      "loss": 1.9003,
+      "step": 234
+    },
+    {
+      "epoch": 0.10778580438023162,
+      "grad_norm": 0.3539673388004303,
+      "learning_rate": 0.0009999114467092636,
+      "loss": 1.4302,
+      "step": 235
+    },
+    {
+      "epoch": 0.10824446737759431,
+      "grad_norm": 0.3749694228172302,
+      "learning_rate": 0.0009999067248315787,
+      "loss": 1.7714,
+      "step": 236
+    },
+    {
+      "epoch": 0.108703130374957,
+      "grad_norm": 0.47365519404411316,
+      "learning_rate": 0.0009999018803231697,
+      "loss": 1.9969,
+      "step": 237
+    },
+    {
+      "epoch": 0.10916179337231968,
+      "grad_norm": 0.3063698709011078,
+      "learning_rate": 0.000999896913185225,
+      "loss": 1.3087,
+      "step": 238
+    },
+    {
+      "epoch": 0.10962045636968237,
+      "grad_norm": 0.2695063352584839,
+      "learning_rate": 0.0009998918234189632,
+      "loss": 1.203,
+      "step": 239
+    },
+    {
+      "epoch": 0.11007911936704506,
+      "grad_norm": 0.3138861060142517,
+      "learning_rate": 0.0009998866110256326,
+      "loss": 1.1181,
+      "step": 240
+    },
+    {
+      "epoch": 0.11053778236440776,
+      "grad_norm": 0.4734935462474823,
+      "learning_rate": 0.000999881276006512,
+      "loss": 2.2582,
+      "step": 241
+    },
+    {
+      "epoch": 0.11099644536177045,
+      "grad_norm": 0.3687781095504761,
+      "learning_rate": 0.0009998758183629099,
+      "loss": 1.6593,
+      "step": 242
+    },
+    {
+      "epoch": 0.11145510835913312,
+      "grad_norm": 0.33596163988113403,
+      "learning_rate": 0.0009998702380961655,
+      "loss": 1.4227,
+      "step": 243
+    },
+    {
+      "epoch": 0.11191377135649581,
+      "grad_norm": 0.37901684641838074,
+      "learning_rate": 0.0009998645352076471,
+      "loss": 1.6273,
+      "step": 244
+    },
+    {
+      "epoch": 0.1123724343538585,
+      "grad_norm": 0.30980101227760315,
+      "learning_rate": 0.0009998587096987544,
+      "loss": 1.0549,
+      "step": 245
+    },
+    {
+      "epoch": 0.1128310973512212,
+      "grad_norm": 0.4566017687320709,
+      "learning_rate": 0.0009998527615709158,
+      "loss": 1.9674,
+      "step": 246
+    },
+    {
+      "epoch": 0.11328976034858387,
+      "grad_norm": 0.4420877695083618,
+      "learning_rate": 0.0009998466908255907,
+      "loss": 2.1819,
+      "step": 247
+    },
+    {
+      "epoch": 0.11374842334594656,
+      "grad_norm": 0.29177185893058777,
+      "learning_rate": 0.0009998404974642684,
+      "loss": 1.3174,
+      "step": 248
+    },
+    {
+      "epoch": 0.11420708634330926,
+      "grad_norm": 0.2630375623703003,
+      "learning_rate": 0.000999834181488468,
+      "loss": 1.0881,
+      "step": 249
+    },
+    {
+      "epoch": 0.11466574934067195,
+      "grad_norm": 0.36166948080062866,
+      "learning_rate": 0.0009998277428997387,
+      "loss": 1.725,
+      "step": 250
+    },
+    {
+      "epoch": 0.11512441233803462,
+      "grad_norm": 0.4986696243286133,
+      "learning_rate": 0.0009998211816996602,
+      "loss": 2.5789,
+      "step": 251
+    },
+    {
+      "epoch": 0.11558307533539731,
+      "grad_norm": 0.3712925612926483,
+      "learning_rate": 0.0009998144978898421,
+      "loss": 2.1168,
+      "step": 252
+    },
+    {
+      "epoch": 0.11604173833276,
+      "grad_norm": 0.29228702187538147,
+      "learning_rate": 0.0009998076914719237,
+      "loss": 1.5106,
+      "step": 253
+    },
+    {
+      "epoch": 0.1165004013301227,
+      "grad_norm": 0.41084229946136475,
+      "learning_rate": 0.0009998007624475746,
+      "loss": 1.99,
+      "step": 254
+    },
+    {
+      "epoch": 0.11695906432748537,
+      "grad_norm": 0.394399493932724,
+      "learning_rate": 0.0009997937108184951,
+      "loss": 2.3965,
+      "step": 255
+    },
+    {
+      "epoch": 0.11741772732484806,
+      "grad_norm": 0.27776357531547546,
+      "learning_rate": 0.0009997865365864143,
+      "loss": 1.3273,
+      "step": 256
+    },
+    {
+      "epoch": 0.11787639032221076,
+      "grad_norm": 0.23769012093544006,
+      "learning_rate": 0.0009997792397530925,
+      "loss": 1.1331,
+      "step": 257
+    },
+    {
+      "epoch": 0.11833505331957345,
+      "grad_norm": 0.3943609297275543,
+      "learning_rate": 0.0009997718203203197,
+      "loss": 2.2263,
+      "step": 258
+    },
+    {
+      "epoch": 0.11879371631693614,
+      "grad_norm": 0.3729076087474823,
+      "learning_rate": 0.0009997642782899158,
+      "loss": 1.9142,
+      "step": 259
+    },
+    {
+      "epoch": 0.11925237931429881,
+      "grad_norm": 0.4220196604728699,
+      "learning_rate": 0.000999756613663731,
+      "loss": 2.4495,
+      "step": 260
+    },
+    {
+      "epoch": 0.1197110423116615,
+      "grad_norm": 0.43063583970069885,
+      "learning_rate": 0.0009997488264436455,
+      "loss": 2.1042,
+      "step": 261
+    },
+    {
+      "epoch": 0.1201697053090242,
+      "grad_norm": 0.31665846705436707,
+      "learning_rate": 0.000999740916631569,
+      "loss": 1.3811,
+      "step": 262
+    },
+    {
+      "epoch": 0.12062836830638689,
+      "grad_norm": 0.5183250904083252,
+      "learning_rate": 0.0009997328842294428,
+      "loss": 2.5891,
+      "step": 263
+    },
+    {
+      "epoch": 0.12108703130374956,
+      "grad_norm": 0.5458334684371948,
+      "learning_rate": 0.000999724729239237,
+      "loss": 2.8789,
+      "step": 264
+    },
+    {
+      "epoch": 0.12154569430111226,
+      "grad_norm": 0.32633325457572937,
+      "learning_rate": 0.0009997164516629515,
+      "loss": 1.5408,
+      "step": 265
+    },
+    {
+      "epoch": 0.12200435729847495,
+      "grad_norm": 0.3069404661655426,
+      "learning_rate": 0.0009997080515026175,
+      "loss": 1.4044,
+      "step": 266
+    },
+    {
+      "epoch": 0.12246302029583764,
+      "grad_norm": 0.3677540719509125,
+      "learning_rate": 0.0009996995287602953,
+      "loss": 2.0308,
+      "step": 267
+    },
+    {
+      "epoch": 0.12292168329320032,
+      "grad_norm": 0.37254124879837036,
+      "learning_rate": 0.0009996908834380756,
+      "loss": 1.7371,
+      "step": 268
+    },
+    {
+      "epoch": 0.123380346290563,
+      "grad_norm": 0.351275771856308,
+      "learning_rate": 0.0009996821155380793,
+      "loss": 1.7554,
+      "step": 269
+    },
+    {
+      "epoch": 0.1238390092879257,
+      "grad_norm": 0.45772960782051086,
+      "learning_rate": 0.0009996732250624571,
+      "loss": 2.5547,
+      "step": 270
+    },
+    {
+      "epoch": 0.12429767228528839,
+      "grad_norm": 0.3280022144317627,
+      "learning_rate": 0.00099966421201339,
+      "loss": 1.5541,
+      "step": 271
+    },
+    {
+      "epoch": 0.12475633528265107,
+      "grad_norm": 0.5368868708610535,
+      "learning_rate": 0.0009996550763930892,
+      "loss": 2.1713,
+      "step": 272
+    },
+    {
+      "epoch": 0.12521499828001376,
+      "grad_norm": 0.2520690858364105,
+      "learning_rate": 0.000999645818203795,
+      "loss": 0.9838,
+      "step": 273
+    },
+    {
+      "epoch": 0.12567366127737645,
+      "grad_norm": 0.3174075186252594,
+      "learning_rate": 0.0009996364374477793,
+      "loss": 1.5059,
+      "step": 274
+    },
+    {
+      "epoch": 0.12613232427473914,
+      "grad_norm": 0.26668769121170044,
+      "learning_rate": 0.0009996269341273427,
+      "loss": 1.3316,
+      "step": 275
+    },
+    {
+      "epoch": 0.12659098727210183,
+      "grad_norm": 0.3275998532772064,
+      "learning_rate": 0.000999617308244817,
+      "loss": 1.717,
+      "step": 276
+    },
+    {
+      "epoch": 0.12704965026946452,
+      "grad_norm": 0.4276018738746643,
+      "learning_rate": 0.0009996075598025628,
+      "loss": 2.5447,
+      "step": 277
+    },
+    {
+      "epoch": 0.1275083132668272,
+      "grad_norm": 0.3532467782497406,
+      "learning_rate": 0.000999597688802972,
+      "loss": 1.939,
+      "step": 278
+    },
+    {
+      "epoch": 0.12796697626418987,
+      "grad_norm": 0.3706674575805664,
+      "learning_rate": 0.000999587695248466,
+      "loss": 1.9899,
+      "step": 279
+    },
+    {
+      "epoch": 0.12842563926155257,
+      "grad_norm": 0.38228681683540344,
+      "learning_rate": 0.0009995775791414958,
+      "loss": 2.1842,
+      "step": 280
+    },
+    {
+      "epoch": 0.12888430225891526,
+      "grad_norm": 0.4526541829109192,
+      "learning_rate": 0.0009995673404845434,
+      "loss": 2.1307,
+      "step": 281
+    },
+    {
+      "epoch": 0.12934296525627795,
+      "grad_norm": 0.44101178646087646,
+      "learning_rate": 0.0009995569792801205,
+      "loss": 2.179,
+      "step": 282
+    },
+    {
+      "epoch": 0.12980162825364064,
+      "grad_norm": 0.15456262230873108,
+      "learning_rate": 0.0009995464955307684,
+      "loss": 0.6752,
+      "step": 283
+    },
+    {
+      "epoch": 0.13026029125100333,
+      "grad_norm": 0.4431912302970886,
+      "learning_rate": 0.000999535889239059,
+      "loss": 2.4602,
+      "step": 284
+    },
+    {
+      "epoch": 0.13071895424836602,
+      "grad_norm": 0.4478715658187866,
+      "learning_rate": 0.0009995251604075943,
+      "loss": 2.3623,
+      "step": 285
+    },
+    {
+      "epoch": 0.1311776172457287,
+      "grad_norm": 0.332527756690979,
+      "learning_rate": 0.000999514309039006,
+      "loss": 1.4232,
+      "step": 286
+    },
+    {
+      "epoch": 0.13163628024309137,
+      "grad_norm": 0.3733260929584503,
+      "learning_rate": 0.0009995033351359558,
+      "loss": 1.7972,
+      "step": 287
+    },
+    {
+      "epoch": 0.13209494324045407,
+      "grad_norm": 0.4460364282131195,
+      "learning_rate": 0.000999492238701136,
+      "loss": 2.227,
+      "step": 288
+    },
+    {
+      "epoch": 0.13255360623781676,
+      "grad_norm": 0.35513222217559814,
+      "learning_rate": 0.0009994810197372684,
+      "loss": 2.0901,
+      "step": 289
+    },
+    {
+      "epoch": 0.13301226923517945,
+      "grad_norm": 0.3551701009273529,
+      "learning_rate": 0.0009994696782471054,
+      "loss": 1.7401,
+      "step": 290
+    },
+    {
+      "epoch": 0.13347093223254214,
+      "grad_norm": 0.37001582980155945,
+      "learning_rate": 0.000999458214233429,
+      "loss": 2.0237,
+      "step": 291
+    },
+    {
+      "epoch": 0.13392959522990483,
+      "grad_norm": 0.39629852771759033,
+      "learning_rate": 0.0009994466276990511,
+      "loss": 1.9617,
+      "step": 292
+    },
+    {
+      "epoch": 0.13438825822726752,
+      "grad_norm": 0.30655738711357117,
+      "learning_rate": 0.0009994349186468144,
+      "loss": 1.4343,
+      "step": 293
+    },
+    {
+      "epoch": 0.1348469212246302,
+      "grad_norm": 0.3813563585281372,
+      "learning_rate": 0.000999423087079591,
+      "loss": 2.0818,
+      "step": 294
+    },
+    {
+      "epoch": 0.1353055842219929,
+      "grad_norm": 0.4179360866546631,
+      "learning_rate": 0.0009994111330002835,
+      "loss": 2.1136,
+      "step": 295
+    },
+    {
+      "epoch": 0.13576424721935557,
+      "grad_norm": 0.341799259185791,
+      "learning_rate": 0.000999399056411824,
+      "loss": 1.9364,
+      "step": 296
+    },
+    {
+      "epoch": 0.13622291021671826,
+      "grad_norm": 0.3983851969242096,
+      "learning_rate": 0.0009993868573171748,
+      "loss": 1.4841,
+      "step": 297
+    },
+    {
+      "epoch": 0.13668157321408095,
+      "grad_norm": 0.32653865218162537,
+      "learning_rate": 0.0009993745357193293,
+      "loss": 1.5058,
+      "step": 298
+    },
+    {
+      "epoch": 0.13714023621144364,
+      "grad_norm": 0.5542972087860107,
+      "learning_rate": 0.0009993620916213092,
+      "loss": 2.5276,
+      "step": 299
+    },
+    {
+      "epoch": 0.13759889920880633,
+      "grad_norm": 0.4614558517932892,
+      "learning_rate": 0.0009993495250261676,
+      "loss": 2.0603,
+      "step": 300
+    },
+    {
+      "epoch": 0.13805756220616902,
+      "grad_norm": 0.1725156605243683,
+      "learning_rate": 0.0009993368359369867,
+      "loss": 0.7159,
+      "step": 301
+    },
+    {
+      "epoch": 0.1385162252035317,
+      "grad_norm": 0.466024786233902,
+      "learning_rate": 0.0009993240243568798,
+      "loss": 2.0669,
+      "step": 302
+    },
+    {
+      "epoch": 0.1389748882008944,
+      "grad_norm": 0.5493818521499634,
+      "learning_rate": 0.0009993110902889893,
+      "loss": 2.6599,
+      "step": 303
+    },
+    {
+      "epoch": 0.1394335511982571,
+      "grad_norm": 0.4970023036003113,
+      "learning_rate": 0.0009992980337364882,
+      "loss": 2.4265,
+      "step": 304
+    },
+    {
+      "epoch": 0.13989221419561976,
+      "grad_norm": 0.37026146054267883,
+      "learning_rate": 0.0009992848547025791,
+      "loss": 1.4705,
+      "step": 305
+    },
+    {
+      "epoch": 0.14035087719298245,
+      "grad_norm": 0.4260435402393341,
+      "learning_rate": 0.0009992715531904955,
+      "loss": 1.6935,
+      "step": 306
+    },
+    {
+      "epoch": 0.14080954019034514,
+      "grad_norm": 0.08674909919500351,
+      "learning_rate": 0.0009992581292034997,
+      "loss": 0.4348,
+      "step": 307
+    },
+    {
+      "epoch": 0.14126820318770783,
+      "grad_norm": 0.3797175884246826,
+      "learning_rate": 0.0009992445827448852,
+      "loss": 1.7039,
+      "step": 308
+    },
+    {
+      "epoch": 0.14172686618507052,
+      "grad_norm": 0.24065154790878296,
+      "learning_rate": 0.0009992309138179745,
+      "loss": 1.0594,
+      "step": 309
+    },
+    {
+      "epoch": 0.1421855291824332,
+      "grad_norm": 0.4619709849357605,
+      "learning_rate": 0.000999217122426121,
+      "loss": 2.0485,
+      "step": 310
+    },
+    {
+      "epoch": 0.1426441921797959,
+      "grad_norm": 0.4283906817436218,
+      "learning_rate": 0.000999203208572708,
+      "loss": 2.2198,
+      "step": 311
+    },
+    {
+      "epoch": 0.1431028551771586,
+      "grad_norm": 0.39942678809165955,
+      "learning_rate": 0.0009991891722611484,
+      "loss": 1.8766,
+      "step": 312
+    },
+    {
+      "epoch": 0.14356151817452126,
+      "grad_norm": 0.37499570846557617,
+      "learning_rate": 0.0009991750134948857,
+      "loss": 1.4081,
+      "step": 313
+    },
+    {
+      "epoch": 0.14402018117188395,
+      "grad_norm": 0.4207157790660858,
+      "learning_rate": 0.0009991607322773928,
+      "loss": 2.1097,
+      "step": 314
+    },
+    {
+      "epoch": 0.14447884416924664,
+      "grad_norm": 0.39866119623184204,
+      "learning_rate": 0.000999146328612173,
+      "loss": 1.8325,
+      "step": 315
+    },
+    {
+      "epoch": 0.14493750716660933,
+      "grad_norm": 0.507705807685852,
+      "learning_rate": 0.00099913180250276,
+      "loss": 2.0061,
+      "step": 316
+    },
+    {
+      "epoch": 0.14539617016397202,
+      "grad_norm": 0.4097268283367157,
+      "learning_rate": 0.0009991171539527168,
+      "loss": 1.9861,
+      "step": 317
+    },
+    {
+      "epoch": 0.1458548331613347,
+      "grad_norm": 0.36734098196029663,
+      "learning_rate": 0.000999102382965637,
+      "loss": 1.6204,
+      "step": 318
+    },
+    {
+      "epoch": 0.1463134961586974,
+      "grad_norm": 0.36929383873939514,
+      "learning_rate": 0.0009990874895451439,
+      "loss": 1.9098,
+      "step": 319
+    },
+    {
+      "epoch": 0.1467721591560601,
+      "grad_norm": 0.40100765228271484,
+      "learning_rate": 0.000999072473694891,
+      "loss": 1.9401,
+      "step": 320
+    },
+    {
+      "epoch": 0.14723082215342279,
+      "grad_norm": 0.38179731369018555,
+      "learning_rate": 0.0009990573354185617,
+      "loss": 2.0457,
+      "step": 321
+    },
+    {
+      "epoch": 0.14768948515078545,
+      "grad_norm": 0.4014952480792999,
+      "learning_rate": 0.0009990420747198697,
+      "loss": 1.5037,
+      "step": 322
+    },
+    {
+      "epoch": 0.14814814814814814,
+      "grad_norm": 0.2382311075925827,
+      "learning_rate": 0.0009990266916025585,
+      "loss": 1.1044,
+      "step": 323
+    },
+    {
+      "epoch": 0.14860681114551083,
+      "grad_norm": 0.40865546464920044,
+      "learning_rate": 0.0009990111860704019,
+      "loss": 1.8592,
+      "step": 324
+    },
+    {
+      "epoch": 0.14906547414287352,
+      "grad_norm": 0.41064581274986267,
+      "learning_rate": 0.000998995558127203,
+      "loss": 1.9371,
+      "step": 325
+    },
+    {
+      "epoch": 0.1495241371402362,
+      "grad_norm": 0.29274582862854004,
+      "learning_rate": 0.0009989798077767959,
+      "loss": 1.3743,
+      "step": 326
+    },
+    {
+      "epoch": 0.1499828001375989,
+      "grad_norm": 0.3344130218029022,
+      "learning_rate": 0.0009989639350230439,
+      "loss": 1.3143,
+      "step": 327
+    },
+    {
+      "epoch": 0.1504414631349616,
+      "grad_norm": 0.25171759724617004,
+      "learning_rate": 0.0009989479398698413,
+      "loss": 0.9927,
+      "step": 328
+    },
+    {
+      "epoch": 0.15090012613232429,
+      "grad_norm": 0.13870039582252502,
+      "learning_rate": 0.0009989318223211112,
+      "loss": 0.7593,
+      "step": 329
+    },
+    {
+      "epoch": 0.15135878912968695,
+      "grad_norm": 0.5177074074745178,
+      "learning_rate": 0.0009989155823808076,
+      "loss": 2.2173,
+      "step": 330
+    },
+    {
+      "epoch": 0.15181745212704964,
+      "grad_norm": 0.3356158435344696,
+      "learning_rate": 0.0009988992200529144,
+      "loss": 1.5492,
+      "step": 331
+    },
+    {
+      "epoch": 0.15227611512441233,
+      "grad_norm": 0.44678938388824463,
+      "learning_rate": 0.000998882735341445,
+      "loss": 1.7162,
+      "step": 332
+    },
+    {
+      "epoch": 0.15273477812177502,
+      "grad_norm": 0.31965455412864685,
+      "learning_rate": 0.0009988661282504438,
+      "loss": 1.2234,
+      "step": 333
+    },
+    {
+      "epoch": 0.1531934411191377,
+      "grad_norm": 0.4562229812145233,
+      "learning_rate": 0.0009988493987839841,
+      "loss": 2.0798,
+      "step": 334
+    },
+    {
+      "epoch": 0.1536521041165004,
+      "grad_norm": 0.36107146739959717,
+      "learning_rate": 0.00099883254694617,
+      "loss": 1.7864,
+      "step": 335
+    },
+    {
+      "epoch": 0.1541107671138631,
+      "grad_norm": 0.3504132330417633,
+      "learning_rate": 0.0009988155727411357,
+      "loss": 1.6267,
+      "step": 336
+    },
+    {
+      "epoch": 0.1545694301112258,
+      "grad_norm": 0.28946036100387573,
+      "learning_rate": 0.0009987984761730445,
+      "loss": 1.319,
+      "step": 337
+    },
+    {
+      "epoch": 0.15502809310858848,
+      "grad_norm": 0.49577978253364563,
+      "learning_rate": 0.0009987812572460905,
+      "loss": 1.9709,
+      "step": 338
+    },
+    {
+      "epoch": 0.15548675610595114,
+      "grad_norm": 0.49071890115737915,
+      "learning_rate": 0.000998763915964498,
+      "loss": 2.1024,
+      "step": 339
+    },
+    {
+      "epoch": 0.15594541910331383,
+      "grad_norm": 0.3110016882419586,
+      "learning_rate": 0.0009987464523325202,
+      "loss": 1.5035,
+      "step": 340
+    },
+    {
+      "epoch": 0.15640408210067652,
+      "grad_norm": 0.3991573452949524,
+      "learning_rate": 0.000998728866354442,
+      "loss": 1.731,
+      "step": 341
+    },
+    {
+      "epoch": 0.1568627450980392,
+      "grad_norm": 0.3533564805984497,
+      "learning_rate": 0.0009987111580345763,
+      "loss": 1.319,
+      "step": 342
+    },
+    {
+      "epoch": 0.1573214080954019,
+      "grad_norm": 0.4693927466869354,
+      "learning_rate": 0.000998693327377268,
+      "loss": 2.17,
+      "step": 343
+    },
+    {
+      "epoch": 0.1577800710927646,
+      "grad_norm": 0.4726630747318268,
+      "learning_rate": 0.0009986753743868905,
+      "loss": 2.1395,
+      "step": 344
+    },
+    {
+      "epoch": 0.1582387340901273,
+      "grad_norm": 0.27608227729797363,
+      "learning_rate": 0.0009986572990678482,
+      "loss": 1.3976,
+      "step": 345
+    },
+    {
+      "epoch": 0.15869739708748998,
+      "grad_norm": 0.34242314100265503,
+      "learning_rate": 0.000998639101424575,
+      "loss": 1.614,
+      "step": 346
+    },
+    {
+      "epoch": 0.15915606008485264,
+      "grad_norm": 0.41097334027290344,
+      "learning_rate": 0.0009986207814615348,
+      "loss": 2.111,
+      "step": 347
+    },
+    {
+      "epoch": 0.15961472308221533,
+      "grad_norm": 0.44249075651168823,
+      "learning_rate": 0.0009986023391832216,
+      "loss": 2.4585,
+      "step": 348
+    },
+    {
+      "epoch": 0.16007338607957802,
+      "grad_norm": 0.412911593914032,
+      "learning_rate": 0.0009985837745941596,
+      "loss": 1.9757,
+      "step": 349
+    },
+    {
+      "epoch": 0.1605320490769407,
+      "grad_norm": 0.22264568507671356,
+      "learning_rate": 0.0009985650876989025,
+      "loss": 0.9562,
+      "step": 350
+    },
+    {
+      "epoch": 0.1609907120743034,
+      "grad_norm": 0.449796199798584,
+      "learning_rate": 0.0009985462785020347,
+      "loss": 2.0573,
+      "step": 351
+    },
+    {
+      "epoch": 0.1614493750716661,
+      "grad_norm": 0.7880814671516418,
+      "learning_rate": 0.0009985273470081702,
+      "loss": 2.3027,
+      "step": 352
+    },
+    {
+      "epoch": 0.1619080380690288,
+      "grad_norm": 10.907687187194824,
+      "learning_rate": 0.0009985082932219527,
+      "loss": 2.3361,
+      "step": 353
+    },
+    {
+      "epoch": 0.16236670106639148,
+      "grad_norm": 2.09915828704834,
+      "learning_rate": 0.0009984891171480566,
+      "loss": 1.6033,
+      "step": 354
+    },
+    {
+      "epoch": 0.16282536406375417,
+      "grad_norm": 0.3674648106098175,
+      "learning_rate": 0.0009984698187911856,
+      "loss": 1.2239,
+      "step": 355
+    },
+    {
+      "epoch": 0.16328402706111683,
+      "grad_norm": 0.5814857482910156,
+      "learning_rate": 0.000998450398156074,
+      "loss": 1.4328,
+      "step": 356
+    },
+    {
+      "epoch": 0.16374269005847952,
+      "grad_norm": 0.46351417899131775,
+      "learning_rate": 0.0009984308552474857,
+      "loss": 2.0938,
+      "step": 357
+    },
+    {
+      "epoch": 0.16420135305584221,
+      "grad_norm": 0.5045230388641357,
+      "learning_rate": 0.0009984111900702146,
+      "loss": 2.2004,
+      "step": 358
+    },
+    {
+      "epoch": 0.1646600160532049,
+      "grad_norm": 0.311603844165802,
+      "learning_rate": 0.000998391402629085,
+      "loss": 1.5096,
+      "step": 359
+    },
+    {
+      "epoch": 0.1651186790505676,
+      "grad_norm": 0.40799856185913086,
+      "learning_rate": 0.0009983714929289508,
+      "loss": 2.1035,
+      "step": 360
+    },
+    {
+      "epoch": 0.1655773420479303,
+      "grad_norm": 0.1845995932817459,
+      "learning_rate": 0.0009983514609746959,
+      "loss": 0.8358,
+      "step": 361
+    },
+    {
+      "epoch": 0.16603600504529298,
+      "grad_norm": 0.36411669850349426,
+      "learning_rate": 0.0009983313067712344,
+      "loss": 1.5809,
+      "step": 362
+    },
+    {
+      "epoch": 0.16649466804265567,
+      "grad_norm": 0.3270757794380188,
+      "learning_rate": 0.00099831103032351,
+      "loss": 1.4844,
+      "step": 363
+    },
+    {
+      "epoch": 0.16695333104001836,
+      "grad_norm": 0.24170421063899994,
+      "learning_rate": 0.0009982906316364968,
+      "loss": 0.9695,
+      "step": 364
+    },
+    {
+      "epoch": 0.16741199403738102,
+      "grad_norm": 0.3997127413749695,
+      "learning_rate": 0.000998270110715199,
+      "loss": 2.1677,
+      "step": 365
+    },
+    {
+      "epoch": 0.16787065703474371,
+      "grad_norm": 0.251727432012558,
+      "learning_rate": 0.0009982494675646503,
+      "loss": 1.3295,
+      "step": 366
+    },
+    {
+      "epoch": 0.1683293200321064,
+      "grad_norm": 0.27435481548309326,
+      "learning_rate": 0.0009982287021899146,
+      "loss": 1.1655,
+      "step": 367
+    },
+    {
+      "epoch": 0.1687879830294691,
+      "grad_norm": 0.3682950437068939,
+      "learning_rate": 0.0009982078145960858,
+      "loss": 1.6586,
+      "step": 368
+    },
+    {
+      "epoch": 0.1692466460268318,
+      "grad_norm": 0.3245248794555664,
+      "learning_rate": 0.0009981868047882877,
+      "loss": 1.2151,
+      "step": 369
+    },
+    {
+      "epoch": 0.16970530902419448,
+      "grad_norm": 0.22152075171470642,
+      "learning_rate": 0.0009981656727716746,
+      "loss": 1.0169,
+      "step": 370
+    },
+    {
+      "epoch": 0.17016397202155717,
+      "grad_norm": 0.3736206293106079,
+      "learning_rate": 0.00099814441855143,
+      "loss": 1.5577,
+      "step": 371
+    },
+    {
+      "epoch": 0.17062263501891986,
+      "grad_norm": 0.3287566602230072,
+      "learning_rate": 0.0009981230421327674,
+      "loss": 1.4915,
+      "step": 372
+    },
+    {
+      "epoch": 0.17108129801628252,
+      "grad_norm": 0.26681220531463623,
+      "learning_rate": 0.000998101543520931,
+      "loss": 1.1304,
+      "step": 373
+    },
+    {
+      "epoch": 0.17153996101364521,
+      "grad_norm": 0.32426249980926514,
+      "learning_rate": 0.0009980799227211946,
+      "loss": 1.4148,
+      "step": 374
+    },
+    {
+      "epoch": 0.1719986240110079,
+      "grad_norm": 0.38737696409225464,
+      "learning_rate": 0.0009980581797388616,
+      "loss": 1.6913,
+      "step": 375
+    },
+    {
+      "epoch": 0.1724572870083706,
+      "grad_norm": 0.5705090761184692,
+      "learning_rate": 0.0009980363145792662,
+      "loss": 2.7029,
+      "step": 376
+    },
+    {
+      "epoch": 0.1729159500057333,
+      "grad_norm": 0.4684537649154663,
+      "learning_rate": 0.0009980143272477718,
+      "loss": 2.1586,
+      "step": 377
+    },
+    {
+      "epoch": 0.17337461300309598,
+      "grad_norm": 0.386222779750824,
+      "learning_rate": 0.0009979922177497718,
+      "loss": 2.2925,
+      "step": 378
+    },
+    {
+      "epoch": 0.17383327600045867,
+      "grad_norm": 0.2883818447589874,
+      "learning_rate": 0.0009979699860906902,
+      "loss": 1.1465,
+      "step": 379
+    },
+    {
+      "epoch": 0.17429193899782136,
+      "grad_norm": 0.30754733085632324,
+      "learning_rate": 0.0009979476322759806,
+      "loss": 0.9909,
+      "step": 380
+    },
+    {
+      "epoch": 0.17475060199518405,
+      "grad_norm": 0.42961201071739197,
+      "learning_rate": 0.0009979251563111261,
+      "loss": 2.0005,
+      "step": 381
+    },
+    {
+      "epoch": 0.17520926499254671,
+      "grad_norm": 0.41431599855422974,
+      "learning_rate": 0.0009979025582016409,
+      "loss": 1.6347,
+      "step": 382
+    },
+    {
+      "epoch": 0.1756679279899094,
+      "grad_norm": 0.39737799763679504,
+      "learning_rate": 0.000997879837953068,
+      "loss": 2.2396,
+      "step": 383
+    },
+    {
+      "epoch": 0.1761265909872721,
+      "grad_norm": 0.2799502909183502,
+      "learning_rate": 0.0009978569955709808,
+      "loss": 1.3289,
+      "step": 384
+    },
+    {
+      "epoch": 0.1765852539846348,
+      "grad_norm": 0.4221355617046356,
+      "learning_rate": 0.000997834031060983,
+      "loss": 2.1155,
+      "step": 385
+    },
+    {
+      "epoch": 0.17704391698199748,
+      "grad_norm": 0.48007211089134216,
+      "learning_rate": 0.0009978109444287078,
+      "loss": 2.677,
+      "step": 386
+    },
+    {
+      "epoch": 0.17750257997936017,
+      "grad_norm": 0.3006044626235962,
+      "learning_rate": 0.0009977877356798185,
+      "loss": 1.4993,
+      "step": 387
+    },
+    {
+      "epoch": 0.17796124297672286,
+      "grad_norm": 0.30362898111343384,
+      "learning_rate": 0.0009977644048200084,
+      "loss": 1.5726,
+      "step": 388
+    },
+    {
+      "epoch": 0.17841990597408555,
+      "grad_norm": 0.25762712955474854,
+      "learning_rate": 0.0009977409518550006,
+      "loss": 1.117,
+      "step": 389
+    },
+    {
+      "epoch": 0.17887856897144822,
+      "grad_norm": 0.2970077693462372,
+      "learning_rate": 0.0009977173767905486,
+      "loss": 1.5483,
+      "step": 390
+    },
+    {
+      "epoch": 0.1793372319688109,
+      "grad_norm": 0.40200671553611755,
+      "learning_rate": 0.0009976936796324352,
+      "loss": 2.1389,
+      "step": 391
+    },
+    {
+      "epoch": 0.1797958949661736,
+      "grad_norm": 0.4564474821090698,
+      "learning_rate": 0.0009976698603864737,
+      "loss": 2.4102,
+      "step": 392
+    },
+    {
+      "epoch": 0.1802545579635363,
+      "grad_norm": 0.5073770880699158,
+      "learning_rate": 0.0009976459190585072,
+      "loss": 2.1866,
+      "step": 393
+    },
+    {
+      "epoch": 0.18071322096089898,
+      "grad_norm": 0.24557775259017944,
+      "learning_rate": 0.0009976218556544084,
+      "loss": 1.0652,
+      "step": 394
+    },
+    {
+      "epoch": 0.18117188395826167,
+      "grad_norm": 0.3804593086242676,
+      "learning_rate": 0.0009975976701800803,
+      "loss": 1.7112,
+      "step": 395
+    },
+    {
+      "epoch": 0.18163054695562436,
+      "grad_norm": 0.414326012134552,
+      "learning_rate": 0.000997573362641456,
+      "loss": 1.8792,
+      "step": 396
+    },
+    {
+      "epoch": 0.18208920995298705,
+      "grad_norm": 0.34744778275489807,
+      "learning_rate": 0.000997548933044498,
+      "loss": 1.4276,
+      "step": 397
+    },
+    {
+      "epoch": 0.18254787295034974,
+      "grad_norm": 0.3834390640258789,
+      "learning_rate": 0.0009975243813951993,
+      "loss": 1.8641,
+      "step": 398
+    },
+    {
+      "epoch": 0.1830065359477124,
+      "grad_norm": 0.40305501222610474,
+      "learning_rate": 0.0009974997076995825,
+      "loss": 1.9985,
+      "step": 399
+    },
+    {
+      "epoch": 0.1834651989450751,
+      "grad_norm": 0.3667886555194855,
+      "learning_rate": 0.0009974749119637002,
+      "loss": 1.7085,
+      "step": 400
+    },
+    {
+      "epoch": 0.1839238619424378,
+      "grad_norm": 0.28065481781959534,
+      "learning_rate": 0.000997449994193635,
+      "loss": 1.42,
+      "step": 401
+    },
+    {
+      "epoch": 0.18438252493980048,
+      "grad_norm": 0.37087318301200867,
+      "learning_rate": 0.0009974249543954995,
+      "loss": 1.5901,
+      "step": 402
+    },
+    {
+      "epoch": 0.18484118793716317,
+      "grad_norm": 0.3824957013130188,
+      "learning_rate": 0.0009973997925754362,
+      "loss": 2.0566,
+      "step": 403
+    },
+    {
+      "epoch": 0.18529985093452586,
+      "grad_norm": 0.3663063049316406,
+      "learning_rate": 0.0009973745087396171,
+      "loss": 2.079,
+      "step": 404
+    },
+    {
+      "epoch": 0.18575851393188855,
+      "grad_norm": 0.30660781264305115,
+      "learning_rate": 0.0009973491028942448,
+      "loss": 1.5338,
+      "step": 405
+    },
+    {
+      "epoch": 0.18621717692925124,
+      "grad_norm": 0.25807487964630127,
+      "learning_rate": 0.0009973235750455516,
+      "loss": 1.2258,
+      "step": 406
+    },
+    {
+      "epoch": 0.1866758399266139,
+      "grad_norm": 0.288686066865921,
+      "learning_rate": 0.0009972979251997994,
+      "loss": 1.6471,
+      "step": 407
+    },
+    {
+      "epoch": 0.1871345029239766,
+      "grad_norm": 0.34755846858024597,
+      "learning_rate": 0.0009972721533632804,
+      "loss": 2.1007,
+      "step": 408
+    },
+    {
+      "epoch": 0.1875931659213393,
+      "grad_norm": 0.22822384536266327,
+      "learning_rate": 0.000997246259542317,
+      "loss": 1.1318,
+      "step": 409
+    },
+    {
+      "epoch": 0.18805182891870198,
+      "grad_norm": 0.12297138571739197,
+      "learning_rate": 0.0009972202437432604,
+      "loss": 0.6918,
+      "step": 410
+    },
+    {
+      "epoch": 0.18851049191606467,
+      "grad_norm": 0.3044014871120453,
+      "learning_rate": 0.000997194105972493,
+      "loss": 1.4567,
+      "step": 411
+    },
+    {
+      "epoch": 0.18896915491342736,
+      "grad_norm": 0.20039816200733185,
+      "learning_rate": 0.0009971678462364266,
+      "loss": 1.1349,
+      "step": 412
+    },
+    {
+      "epoch": 0.18942781791079005,
+      "grad_norm": 0.36312806606292725,
+      "learning_rate": 0.0009971414645415027,
+      "loss": 2.0817,
+      "step": 413
+    },
+    {
+      "epoch": 0.18988648090815274,
+      "grad_norm": 0.30179518461227417,
+      "learning_rate": 0.0009971149608941929,
+      "loss": 1.5486,
+      "step": 414
+    },
+    {
+      "epoch": 0.19034514390551543,
+      "grad_norm": 0.3225315809249878,
+      "learning_rate": 0.0009970883353009987,
+      "loss": 1.3529,
+      "step": 415
+    },
+    {
+      "epoch": 0.1908038069028781,
+      "grad_norm": 0.3565948009490967,
+      "learning_rate": 0.000997061587768452,
+      "loss": 1.9429,
+      "step": 416
+    },
+    {
+      "epoch": 0.1912624699002408,
+      "grad_norm": 0.6202376484870911,
+      "learning_rate": 0.0009970347183031133,
+      "loss": 1.3034,
+      "step": 417
+    },
+    {
+      "epoch": 0.19172113289760348,
+      "grad_norm": 0.31976190209388733,
+      "learning_rate": 0.0009970077269115748,
+      "loss": 1.833,
+      "step": 418
+    },
+    {
+      "epoch": 0.19217979589496617,
+      "grad_norm": 0.4025329649448395,
+      "learning_rate": 0.0009969806136004573,
+      "loss": 1.7521,
+      "step": 419
+    },
+    {
+      "epoch": 0.19263845889232886,
+      "grad_norm": 0.20742328464984894,
+      "learning_rate": 0.0009969533783764117,
+      "loss": 1.1387,
+      "step": 420
+    },
+    {
+      "epoch": 0.19309712188969155,
+      "grad_norm": 0.2832637429237366,
+      "learning_rate": 0.0009969260212461192,
+      "loss": 1.5901,
+      "step": 421
+    },
+    {
+      "epoch": 0.19355578488705424,
+      "grad_norm": 0.21540392935276031,
+      "learning_rate": 0.0009968985422162907,
+      "loss": 1.0676,
+      "step": 422
+    },
+    {
+      "epoch": 0.19401444788441694,
+      "grad_norm": 0.31164097785949707,
+      "learning_rate": 0.0009968709412936669,
+      "loss": 1.7977,
+      "step": 423
+    },
+    {
+      "epoch": 0.19447311088177963,
+      "grad_norm": 0.3007836937904358,
+      "learning_rate": 0.0009968432184850188,
+      "loss": 1.9207,
+      "step": 424
+    },
+    {
+      "epoch": 0.1949317738791423,
+      "grad_norm": 0.34054356813430786,
+      "learning_rate": 0.0009968153737971463,
+      "loss": 2.281,
+      "step": 425
+    },
+    {
+      "epoch": 0.19539043687650498,
+      "grad_norm": 0.3539496660232544,
+      "learning_rate": 0.0009967874072368808,
+      "loss": 2.0088,
+      "step": 426
+    },
+    {
+      "epoch": 0.19584909987386767,
+      "grad_norm": 0.4037904441356659,
+      "learning_rate": 0.0009967593188110822,
+      "loss": 2.0894,
+      "step": 427
+    },
+    {
+      "epoch": 0.19630776287123036,
+      "grad_norm": 0.36563950777053833,
+      "learning_rate": 0.0009967311085266408,
+      "loss": 1.7298,
+      "step": 428
+    },
+    {
+      "epoch": 0.19676642586859305,
+      "grad_norm": 0.4053013324737549,
+      "learning_rate": 0.000996702776390477,
+      "loss": 2.3652,
+      "step": 429
+    },
+    {
+      "epoch": 0.19722508886595574,
+      "grad_norm": 0.24620242416858673,
+      "learning_rate": 0.0009966743224095406,
+      "loss": 1.1821,
+      "step": 430
+    },
+    {
+      "epoch": 0.19768375186331844,
+      "grad_norm": 0.4127357006072998,
+      "learning_rate": 0.0009966457465908117,
+      "loss": 1.9592,
+      "step": 431
+    },
+    {
+      "epoch": 0.19814241486068113,
+      "grad_norm": 0.13345493376255035,
+      "learning_rate": 0.0009966170489413002,
+      "loss": 0.6845,
+      "step": 432
+    },
+    {
+      "epoch": 0.1986010778580438,
+      "grad_norm": 0.46981745958328247,
+      "learning_rate": 0.000996588229468046,
+      "loss": 2.3521,
+      "step": 433
+    },
+    {
+      "epoch": 0.19905974085540648,
+      "grad_norm": 0.414408802986145,
+      "learning_rate": 0.0009965592881781182,
+      "loss": 1.7859,
+      "step": 434
+    },
+    {
+      "epoch": 0.19951840385276917,
+      "grad_norm": 0.36975857615470886,
+      "learning_rate": 0.0009965302250786168,
+      "loss": 1.8119,
+      "step": 435
+    },
+    {
+      "epoch": 0.19997706685013186,
+      "grad_norm": 0.29966479539871216,
+      "learning_rate": 0.000996501040176671,
+      "loss": 1.6156,
+      "step": 436
+    },
+    {
+      "epoch": 0.20043572984749455,
+      "grad_norm": 0.3034079074859619,
+      "learning_rate": 0.00099647173347944,
+      "loss": 1.4443,
+      "step": 437
+    },
+    {
+      "epoch": 0.20089439284485724,
+      "grad_norm": 0.23707164824008942,
+      "learning_rate": 0.0009964423049941132,
+      "loss": 1.1766,
+      "step": 438
+    },
+    {
+      "epoch": 0.20135305584221994,
+      "grad_norm": 0.37824898958206177,
+      "learning_rate": 0.0009964127547279094,
+      "loss": 2.0259,
+      "step": 439
+    },
+    {
+      "epoch": 0.20181171883958263,
+      "grad_norm": 0.26142311096191406,
+      "learning_rate": 0.0009963830826880775,
+      "loss": 1.3159,
+      "step": 440
+    },
+    {
+      "epoch": 0.20227038183694532,
+      "grad_norm": 0.34730425477027893,
+      "learning_rate": 0.0009963532888818962,
+      "loss": 1.8903,
+      "step": 441
+    },
+    {
+      "epoch": 0.20272904483430798,
+      "grad_norm": 0.505499541759491,
+      "learning_rate": 0.0009963233733166744,
+      "loss": 2.4998,
+      "step": 442
+    },
+    {
+      "epoch": 0.20318770783167067,
+      "grad_norm": 0.38120755553245544,
+      "learning_rate": 0.0009962933359997505,
+      "loss": 1.8554,
+      "step": 443
+    },
+    {
+      "epoch": 0.20364637082903336,
+      "grad_norm": 0.2800237536430359,
+      "learning_rate": 0.0009962631769384928,
+      "loss": 1.4182,
+      "step": 444
+    },
+    {
+      "epoch": 0.20410503382639605,
+      "grad_norm": 0.35025089979171753,
+      "learning_rate": 0.0009962328961402992,
+      "loss": 1.7761,
+      "step": 445
+    },
+    {
+      "epoch": 0.20456369682375874,
+      "grad_norm": 0.446393221616745,
+      "learning_rate": 0.0009962024936125986,
+      "loss": 2.626,
+      "step": 446
+    },
+    {
+      "epoch": 0.20502235982112144,
+      "grad_norm": 0.26646509766578674,
+      "learning_rate": 0.0009961719693628484,
+      "loss": 1.2512,
+      "step": 447
+    },
+    {
+      "epoch": 0.20548102281848413,
+      "grad_norm": 0.4339228868484497,
+      "learning_rate": 0.0009961413233985363,
+      "loss": 2.1897,
+      "step": 448
+    },
+    {
+      "epoch": 0.20593968581584682,
+      "grad_norm": 0.31140047311782837,
+      "learning_rate": 0.0009961105557271803,
+      "loss": 1.3533,
+      "step": 449
+    },
+    {
+      "epoch": 0.20639834881320948,
+      "grad_norm": 0.34898841381073,
+      "learning_rate": 0.0009960796663563282,
+      "loss": 1.6303,
+      "step": 450
+    },
+    {
+      "epoch": 0.20685701181057217,
+      "grad_norm": 0.3299868106842041,
+      "learning_rate": 0.0009960486552935566,
+      "loss": 1.5021,
+      "step": 451
+    },
+    {
+      "epoch": 0.20731567480793486,
+      "grad_norm": 0.3481556177139282,
+      "learning_rate": 0.0009960175225464732,
+      "loss": 1.6241,
+      "step": 452
+    },
+    {
+      "epoch": 0.20777433780529755,
+      "grad_norm": 0.2556866407394409,
+      "learning_rate": 0.0009959862681227151,
+      "loss": 1.2629,
+      "step": 453
+    },
+    {
+      "epoch": 0.20823300080266025,
+      "grad_norm": 0.35603269934654236,
+      "learning_rate": 0.0009959548920299493,
+      "loss": 1.5148,
+      "step": 454
+    },
+    {
+      "epoch": 0.20869166380002294,
+      "grad_norm": 0.4437220096588135,
+      "learning_rate": 0.0009959233942758725,
+      "loss": 2.4865,
+      "step": 455
+    },
+    {
+      "epoch": 0.20915032679738563,
+      "grad_norm": 0.5016736388206482,
+      "learning_rate": 0.000995891774868211,
+      "loss": 2.6392,
+      "step": 456
+    },
+    {
+      "epoch": 0.20960898979474832,
+      "grad_norm": 0.32121479511260986,
+      "learning_rate": 0.000995860033814722,
+      "loss": 1.7887,
+      "step": 457
+    },
+    {
+      "epoch": 0.210067652792111,
+      "grad_norm": 0.316918283700943,
+      "learning_rate": 0.000995828171123191,
+      "loss": 1.5659,
+      "step": 458
+    },
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 0.2690139710903168,
+      "learning_rate": 0.0009957961868014347,
+      "loss": 1.4294,
+      "step": 459
+    },
+    {
+      "epoch": 0.21098497878683636,
+      "grad_norm": 0.3223515450954437,
+      "learning_rate": 0.0009957640808572988,
+      "loss": 1.7502,
+      "step": 460
+    },
+    {
+      "epoch": 0.21144364178419905,
+      "grad_norm": 0.267553448677063,
+      "learning_rate": 0.0009957318532986593,
+      "loss": 1.1748,
+      "step": 461
+    },
+    {
+      "epoch": 0.21190230478156175,
+      "grad_norm": 0.30810901522636414,
+      "learning_rate": 0.0009956995041334215,
+      "loss": 1.6794,
+      "step": 462
+    },
+    {
+      "epoch": 0.21236096777892444,
+      "grad_norm": 0.33142194151878357,
+      "learning_rate": 0.0009956670333695215,
+      "loss": 1.7014,
+      "step": 463
+    },
+    {
+      "epoch": 0.21281963077628713,
+      "grad_norm": 0.270020455121994,
+      "learning_rate": 0.0009956344410149242,
+      "loss": 1.2939,
+      "step": 464
+    },
+    {
+      "epoch": 0.21327829377364982,
+      "grad_norm": 0.39550352096557617,
+      "learning_rate": 0.0009956017270776247,
+      "loss": 1.9753,
+      "step": 465
+    },
+    {
+      "epoch": 0.2137369567710125,
+      "grad_norm": 0.3066462278366089,
+      "learning_rate": 0.0009955688915656482,
+      "loss": 1.4247,
+      "step": 466
+    },
+    {
+      "epoch": 0.21419561976837517,
+      "grad_norm": 0.3493700325489044,
+      "learning_rate": 0.0009955359344870495,
+      "loss": 1.6978,
+      "step": 467
+    },
+    {
+      "epoch": 0.21465428276573786,
+      "grad_norm": 0.2585456073284149,
+      "learning_rate": 0.0009955028558499125,
+      "loss": 1.2059,
+      "step": 468
+    },
+    {
+      "epoch": 0.21511294576310055,
+      "grad_norm": 0.31552475690841675,
+      "learning_rate": 0.0009954696556623529,
+      "loss": 1.6361,
+      "step": 469
+    },
+    {
+      "epoch": 0.21557160876046325,
+      "grad_norm": 0.31127098202705383,
+      "learning_rate": 0.0009954363339325137,
+      "loss": 1.5196,
+      "step": 470
+    },
+    {
+      "epoch": 0.21603027175782594,
+      "grad_norm": 0.4300401210784912,
+      "learning_rate": 0.0009954028906685698,
+      "loss": 2.5085,
+      "step": 471
+    },
+    {
+      "epoch": 0.21648893475518863,
+      "grad_norm": 0.08116165548563004,
+      "learning_rate": 0.0009953693258787247,
+      "loss": 0.5775,
+      "step": 472
+    },
+    {
+      "epoch": 0.21694759775255132,
+      "grad_norm": 0.3769618272781372,
+      "learning_rate": 0.0009953356395712121,
+      "loss": 2.0813,
+      "step": 473
+    },
+    {
+      "epoch": 0.217406260749914,
+      "grad_norm": 0.3906650245189667,
+      "learning_rate": 0.0009953018317542957,
+      "loss": 2.0854,
+      "step": 474
+    },
+    {
+      "epoch": 0.2178649237472767,
+      "grad_norm": 0.24557356536388397,
+      "learning_rate": 0.0009952679024362685,
+      "loss": 1.2442,
+      "step": 475
+    },
+    {
+      "epoch": 0.21832358674463936,
+      "grad_norm": 0.3962481915950775,
+      "learning_rate": 0.000995233851625454,
+      "loss": 2.356,
+      "step": 476
+    },
+    {
+      "epoch": 0.21878224974200206,
+      "grad_norm": 0.2204056829214096,
+      "learning_rate": 0.0009951996793302047,
+      "loss": 0.9941,
+      "step": 477
+    },
+    {
+      "epoch": 0.21924091273936475,
+      "grad_norm": 0.07944503426551819,
+      "learning_rate": 0.0009951653855589035,
+      "loss": 0.4608,
+      "step": 478
+    },
+    {
+      "epoch": 0.21969957573672744,
+      "grad_norm": 0.2788356840610504,
+      "learning_rate": 0.000995130970319963,
+      "loss": 1.2136,
+      "step": 479
+    },
+    {
+      "epoch": 0.22015823873409013,
+      "grad_norm": 0.10177599638700485,
+      "learning_rate": 0.000995096433621825,
+      "loss": 0.5138,
+      "step": 480
+    },
+    {
+      "epoch": 0.22061690173145282,
+      "grad_norm": 0.19846868515014648,
+      "learning_rate": 0.0009950617754729625,
+      "loss": 1.0419,
+      "step": 481
+    },
+    {
+      "epoch": 0.2210755647288155,
+      "grad_norm": 0.3579308092594147,
+      "learning_rate": 0.0009950269958818767,
+      "loss": 1.4526,
+      "step": 482
+    },
+    {
+      "epoch": 0.2215342277261782,
+      "grad_norm": 0.36856117844581604,
+      "learning_rate": 0.0009949920948570995,
+      "loss": 1.7526,
+      "step": 483
+    },
+    {
+      "epoch": 0.2219928907235409,
+      "grad_norm": 0.32889747619628906,
+      "learning_rate": 0.0009949570724071923,
+      "loss": 1.4898,
+      "step": 484
+    },
+    {
+      "epoch": 0.22245155372090356,
+      "grad_norm": 0.4301432967185974,
+      "learning_rate": 0.0009949219285407464,
+      "loss": 1.9407,
+      "step": 485
+    },
+    {
+      "epoch": 0.22291021671826625,
+      "grad_norm": 0.3047052323818207,
+      "learning_rate": 0.000994886663266383,
+      "loss": 1.3376,
+      "step": 486
+    },
+    {
+      "epoch": 0.22336887971562894,
+      "grad_norm": 0.27812737226486206,
+      "learning_rate": 0.0009948512765927525,
+      "loss": 1.2351,
+      "step": 487
+    },
+    {
+      "epoch": 0.22382754271299163,
+      "grad_norm": 0.34248921275138855,
+      "learning_rate": 0.0009948157685285362,
+      "loss": 1.8019,
+      "step": 488
+    },
+    {
+      "epoch": 0.22428620571035432,
+      "grad_norm": 0.295585036277771,
+      "learning_rate": 0.0009947801390824437,
+      "loss": 1.5117,
+      "step": 489
+    },
+    {
+      "epoch": 0.224744868707717,
+      "grad_norm": 0.3885442912578583,
+      "learning_rate": 0.000994744388263216,
+      "loss": 1.9371,
+      "step": 490
+    },
+    {
+      "epoch": 0.2252035317050797,
+      "grad_norm": 0.5128207206726074,
+      "learning_rate": 0.0009947085160796221,
+      "loss": 1.957,
+      "step": 491
+    },
+    {
+      "epoch": 0.2256621947024424,
+      "grad_norm": 0.3117857873439789,
+      "learning_rate": 0.0009946725225404623,
+      "loss": 1.6161,
+      "step": 492
+    },
+    {
+      "epoch": 0.22612085769980506,
+      "grad_norm": 0.3748393952846527,
+      "learning_rate": 0.000994636407654566,
+      "loss": 2.0451,
+      "step": 493
+    },
+    {
+      "epoch": 0.22657952069716775,
+      "grad_norm": 0.4382317364215851,
+      "learning_rate": 0.0009946001714307923,
+      "loss": 2.3245,
+      "step": 494
+    },
+    {
+      "epoch": 0.22703818369453044,
+      "grad_norm": 0.4299793243408203,
+      "learning_rate": 0.0009945638138780305,
+      "loss": 2.5256,
+      "step": 495
+    },
+    {
+      "epoch": 0.22749684669189313,
+      "grad_norm": 0.36116522550582886,
+      "learning_rate": 0.000994527335005199,
+      "loss": 1.8933,
+      "step": 496
+    },
+    {
+      "epoch": 0.22795550968925582,
+      "grad_norm": 0.39556559920310974,
+      "learning_rate": 0.0009944907348212464,
+      "loss": 1.9866,
+      "step": 497
+    },
+    {
+      "epoch": 0.2284141726866185,
+      "grad_norm": 0.2519306242465973,
+      "learning_rate": 0.0009944540133351512,
+      "loss": 1.0477,
+      "step": 498
+    },
+    {
+      "epoch": 0.2288728356839812,
+      "grad_norm": 0.23983240127563477,
+      "learning_rate": 0.0009944171705559214,
+      "loss": 0.9459,
+      "step": 499
+    },
+    {
+      "epoch": 0.2293314986813439,
+      "grad_norm": 0.2413322627544403,
+      "learning_rate": 0.0009943802064925947,
+      "loss": 0.9827,
+      "step": 500
+    },
+    {
+      "epoch": 0.22979016167870658,
+      "grad_norm": 0.48764920234680176,
+      "learning_rate": 0.0009943431211542387,
+      "loss": 2.16,
+      "step": 501
+    },
+    {
+      "epoch": 0.23024882467606925,
+      "grad_norm": 0.13628147542476654,
+      "learning_rate": 0.0009943059145499508,
+      "loss": 0.7079,
+      "step": 502
+    },
+    {
+      "epoch": 0.23070748767343194,
+      "grad_norm": 0.3403797447681427,
+      "learning_rate": 0.0009942685866888576,
+      "loss": 1.6746,
+      "step": 503
+    },
+    {
+      "epoch": 0.23116615067079463,
+      "grad_norm": 0.4511934518814087,
+      "learning_rate": 0.0009942311375801163,
+      "loss": 1.9286,
+      "step": 504
+    },
+    {
+      "epoch": 0.23162481366815732,
+      "grad_norm": 0.3697836399078369,
+      "learning_rate": 0.0009941935672329136,
+      "loss": 1.4048,
+      "step": 505
+    },
+    {
+      "epoch": 0.23208347666552,
+      "grad_norm": 0.30644330382347107,
+      "learning_rate": 0.0009941558756564653,
+      "loss": 1.3854,
+      "step": 506
+    },
+    {
+      "epoch": 0.2325421396628827,
+      "grad_norm": 0.2954138517379761,
+      "learning_rate": 0.0009941180628600178,
+      "loss": 1.3652,
+      "step": 507
+    },
+    {
+      "epoch": 0.2330008026602454,
+      "grad_norm": 0.30790993571281433,
+      "learning_rate": 0.0009940801288528466,
+      "loss": 1.447,
+      "step": 508
+    },
+    {
+      "epoch": 0.23345946565760808,
+      "grad_norm": 0.37096887826919556,
+      "learning_rate": 0.0009940420736442575,
+      "loss": 1.4563,
+      "step": 509
+    },
+    {
+      "epoch": 0.23391812865497075,
+      "grad_norm": 0.32588744163513184,
+      "learning_rate": 0.0009940038972435852,
+      "loss": 1.4545,
+      "step": 510
+    },
+    {
+      "epoch": 0.23437679165233344,
+      "grad_norm": 0.30430012941360474,
+      "learning_rate": 0.000993965599660195,
+      "loss": 1.509,
+      "step": 511
+    },
+    {
+      "epoch": 0.23483545464969613,
+      "grad_norm": 0.3035714626312256,
+      "learning_rate": 0.0009939271809034818,
+      "loss": 1.5743,
+      "step": 512
+    },
+    {
+      "epoch": 0.23529411764705882,
+      "grad_norm": 0.21552062034606934,
+      "learning_rate": 0.0009938886409828695,
+      "loss": 0.9872,
+      "step": 513
+    },
+    {
+      "epoch": 0.2357527806444215,
+      "grad_norm": 0.30967265367507935,
+      "learning_rate": 0.0009938499799078124,
+      "loss": 1.5457,
+      "step": 514
+    },
+    {
+      "epoch": 0.2362114436417842,
+      "grad_norm": 0.47051116824150085,
+      "learning_rate": 0.0009938111976877946,
+      "loss": 2.3921,
+      "step": 515
+    },
+    {
+      "epoch": 0.2366701066391469,
+      "grad_norm": 0.24845226109027863,
+      "learning_rate": 0.0009937722943323293,
+      "loss": 1.0681,
+      "step": 516
+    },
+    {
+      "epoch": 0.23712876963650958,
+      "grad_norm": 0.28985902667045593,
+      "learning_rate": 0.0009937332698509598,
+      "loss": 1.3932,
+      "step": 517
+    },
+    {
+      "epoch": 0.23758743263387228,
+      "grad_norm": 0.37183916568756104,
+      "learning_rate": 0.0009936941242532594,
+      "loss": 1.9424,
+      "step": 518
+    },
+    {
+      "epoch": 0.23804609563123494,
+      "grad_norm": 0.32926613092422485,
+      "learning_rate": 0.0009936548575488306,
+      "loss": 1.8109,
+      "step": 519
+    },
+    {
+      "epoch": 0.23850475862859763,
+      "grad_norm": 0.3398732542991638,
+      "learning_rate": 0.0009936154697473057,
+      "loss": 1.7718,
+      "step": 520
+    },
+    {
+      "epoch": 0.23896342162596032,
+      "grad_norm": 0.36996906995773315,
+      "learning_rate": 0.0009935759608583471,
+      "loss": 1.7703,
+      "step": 521
+    },
+    {
+      "epoch": 0.239422084623323,
+      "grad_norm": 0.10577350109815598,
+      "learning_rate": 0.0009935363308916463,
+      "loss": 0.5741,
+      "step": 522
+    },
+    {
+      "epoch": 0.2398807476206857,
+      "grad_norm": 0.4257866144180298,
+      "learning_rate": 0.0009934965798569248,
+      "loss": 2.071,
+      "step": 523
+    },
+    {
+      "epoch": 0.2403394106180484,
+      "grad_norm": 0.23754611611366272,
+      "learning_rate": 0.0009934567077639342,
+      "loss": 1.2965,
+      "step": 524
+    },
+    {
+      "epoch": 0.24079807361541108,
+      "grad_norm": 0.2694760859012604,
+      "learning_rate": 0.0009934167146224552,
+      "loss": 1.4501,
+      "step": 525
+    },
+    {
+      "epoch": 0.24125673661277378,
+      "grad_norm": 0.35148951411247253,
+      "learning_rate": 0.0009933766004422983,
+      "loss": 1.674,
+      "step": 526
+    },
+    {
+      "epoch": 0.24171539961013644,
+      "grad_norm": 0.3345791697502136,
+      "learning_rate": 0.000993336365233304,
+      "loss": 1.4195,
+      "step": 527
+    },
+    {
+      "epoch": 0.24217406260749913,
+      "grad_norm": 0.15900827944278717,
+      "learning_rate": 0.0009932960090053417,
+      "loss": 0.7744,
+      "step": 528
+    },
+    {
+      "epoch": 0.24263272560486182,
+      "grad_norm": 0.358396977186203,
+      "learning_rate": 0.000993255531768312,
+      "loss": 1.766,
+      "step": 529
+    },
+    {
+      "epoch": 0.2430913886022245,
+      "grad_norm": 0.35677099227905273,
+      "learning_rate": 0.0009932149335321438,
+      "loss": 1.7595,
+      "step": 530
+    },
+    {
+      "epoch": 0.2435500515995872,
+      "grad_norm": 0.19825054705142975,
+      "learning_rate": 0.000993174214306796,
+      "loss": 0.913,
+      "step": 531
+    },
+    {
+      "epoch": 0.2440087145969499,
+      "grad_norm": 0.2745889723300934,
+      "learning_rate": 0.0009931333741022574,
+      "loss": 1.4666,
+      "step": 532
+    },
+    {
+      "epoch": 0.24446737759431258,
+      "grad_norm": 0.2890368700027466,
+      "learning_rate": 0.0009930924129285465,
+      "loss": 1.6223,
+      "step": 533
+    },
+    {
+      "epoch": 0.24492604059167528,
+      "grad_norm": 0.21782201528549194,
+      "learning_rate": 0.0009930513307957114,
+      "loss": 1.2415,
+      "step": 534
+    },
+    {
+      "epoch": 0.24538470358903797,
+      "grad_norm": 0.3002353310585022,
+      "learning_rate": 0.0009930101277138299,
+      "loss": 1.6741,
+      "step": 535
+    },
+    {
+      "epoch": 0.24584336658640063,
+      "grad_norm": 0.21679870784282684,
+      "learning_rate": 0.000992968803693009,
+      "loss": 1.2902,
+      "step": 536
+    },
+    {
+      "epoch": 0.24630202958376332,
+      "grad_norm": 0.4101315140724182,
+      "learning_rate": 0.0009929273587433863,
+      "loss": 1.9695,
+      "step": 537
+    },
+    {
+      "epoch": 0.246760692581126,
+      "grad_norm": 0.355189710855484,
+      "learning_rate": 0.0009928857928751282,
+      "loss": 1.7978,
+      "step": 538
+    },
+    {
+      "epoch": 0.2472193555784887,
+      "grad_norm": 0.2238903045654297,
+      "learning_rate": 0.0009928441060984311,
+      "loss": 1.2032,
+      "step": 539
+    },
+    {
+      "epoch": 0.2476780185758514,
+      "grad_norm": 0.31160059571266174,
+      "learning_rate": 0.0009928022984235214,
+      "loss": 1.6759,
+      "step": 540
+    },
+    {
+      "epoch": 0.24813668157321409,
+      "grad_norm": 0.3302862346172333,
+      "learning_rate": 0.0009927603698606543,
+      "loss": 1.7676,
+      "step": 541
+    },
+    {
+      "epoch": 0.24859534457057678,
+      "grad_norm": 0.27699384093284607,
+      "learning_rate": 0.0009927183204201159,
+      "loss": 1.348,
+      "step": 542
+    },
+    {
+      "epoch": 0.24905400756793947,
+      "grad_norm": 0.33498769998550415,
+      "learning_rate": 0.0009926761501122205,
+      "loss": 1.6743,
+      "step": 543
+    },
+    {
+      "epoch": 0.24951267056530213,
+      "grad_norm": 0.2707799971103668,
+      "learning_rate": 0.0009926338589473134,
+      "loss": 1.3558,
+      "step": 544
+    },
+    {
+      "epoch": 0.24997133356266482,
+      "grad_norm": 0.2782471776008606,
+      "learning_rate": 0.0009925914469357685,
+      "loss": 1.2665,
+      "step": 545
+    },
+    {
+      "epoch": 0.2504299965600275,
+      "grad_norm": 0.4235142171382904,
+      "learning_rate": 0.0009925489140879898,
+      "loss": 2.0011,
+      "step": 546
+    },
+    {
+      "epoch": 0.2508886595573902,
+      "grad_norm": 0.3951006233692169,
+      "learning_rate": 0.000992506260414411,
+      "loss": 1.7933,
+      "step": 547
+    },
+    {
+      "epoch": 0.2513473225547529,
+      "grad_norm": 0.26279470324516296,
+      "learning_rate": 0.0009924634859254952,
+      "loss": 1.0981,
+      "step": 548
+    },
+    {
+      "epoch": 0.2518059855521156,
+      "grad_norm": 0.3145405650138855,
+      "learning_rate": 0.000992420590631736,
+      "loss": 1.5393,
+      "step": 549
+    },
+    {
+      "epoch": 0.2522646485494783,
+      "grad_norm": 0.2764291763305664,
+      "learning_rate": 0.0009923775745436549,
+      "loss": 1.2642,
+      "step": 550
+    },
+    {
+      "epoch": 0.25272331154684097,
+      "grad_norm": 0.3677400052547455,
+      "learning_rate": 0.0009923344376718047,
+      "loss": 1.1807,
+      "step": 551
+    },
+    {
+      "epoch": 0.25318197454420366,
+      "grad_norm": 0.4418051242828369,
+      "learning_rate": 0.000992291180026767,
+      "loss": 2.4766,
+      "step": 552
+    },
+    {
+      "epoch": 0.25364063754156635,
+      "grad_norm": 0.3161998391151428,
+      "learning_rate": 0.0009922478016191531,
+      "loss": 1.3802,
+      "step": 553
+    },
+    {
+      "epoch": 0.25409930053892904,
+      "grad_norm": 0.39167219400405884,
+      "learning_rate": 0.0009922043024596045,
+      "loss": 2.0244,
+      "step": 554
+    },
+    {
+      "epoch": 0.25455796353629173,
+      "grad_norm": 0.3040946125984192,
+      "learning_rate": 0.0009921606825587913,
+      "loss": 1.6199,
+      "step": 555
+    },
+    {
+      "epoch": 0.2550166265336544,
+      "grad_norm": 0.22090859711170197,
+      "learning_rate": 0.000992116941927414,
+      "loss": 1.0001,
+      "step": 556
+    },
+    {
+      "epoch": 0.25547528953101706,
+      "grad_norm": 0.29208970069885254,
+      "learning_rate": 0.0009920730805762026,
+      "loss": 1.3642,
+      "step": 557
+    },
+    {
+      "epoch": 0.25593395252837975,
+      "grad_norm": 0.3597604036331177,
+      "learning_rate": 0.0009920290985159165,
+      "loss": 1.6378,
+      "step": 558
+    },
+    {
+      "epoch": 0.25639261552574244,
+      "grad_norm": 0.36907076835632324,
+      "learning_rate": 0.0009919849957573449,
+      "loss": 2.0097,
+      "step": 559
+    },
+    {
+      "epoch": 0.25685127852310513,
+      "grad_norm": 0.5593513250350952,
+      "learning_rate": 0.0009919407723113062,
+      "loss": 2.4558,
+      "step": 560
+    },
+    {
+      "epoch": 0.2573099415204678,
+      "grad_norm": 0.38623642921447754,
+      "learning_rate": 0.0009918964281886492,
+      "loss": 1.9032,
+      "step": 561
+    },
+    {
+      "epoch": 0.2577686045178305,
+      "grad_norm": 0.3310829997062683,
+      "learning_rate": 0.0009918519634002515,
+      "loss": 1.403,
+      "step": 562
+    },
+    {
+      "epoch": 0.2582272675151932,
+      "grad_norm": 0.3338778018951416,
+      "learning_rate": 0.0009918073779570206,
+      "loss": 1.7481,
+      "step": 563
+    },
+    {
+      "epoch": 0.2586859305125559,
+      "grad_norm": 0.32507607340812683,
+      "learning_rate": 0.000991762671869894,
+      "loss": 1.4885,
+      "step": 564
+    },
+    {
+      "epoch": 0.2591445935099186,
+      "grad_norm": 0.13469289243221283,
+      "learning_rate": 0.0009917178451498382,
+      "loss": 0.7271,
+      "step": 565
+    },
+    {
+      "epoch": 0.2596032565072813,
+      "grad_norm": 0.28605028986930847,
+      "learning_rate": 0.0009916728978078494,
+      "loss": 1.3297,
+      "step": 566
+    },
+    {
+      "epoch": 0.26006191950464397,
+      "grad_norm": 0.4288422465324402,
+      "learning_rate": 0.0009916278298549538,
+      "loss": 2.4978,
+      "step": 567
+    },
+    {
+      "epoch": 0.26052058250200666,
+      "grad_norm": 0.2774169445037842,
+      "learning_rate": 0.0009915826413022064,
+      "loss": 1.5432,
+      "step": 568
+    },
+    {
+      "epoch": 0.26097924549936935,
+      "grad_norm": 0.23148323595523834,
+      "learning_rate": 0.0009915373321606928,
+      "loss": 1.2222,
+      "step": 569
+    },
+    {
+      "epoch": 0.26143790849673204,
+      "grad_norm": 0.4910981059074402,
+      "learning_rate": 0.0009914919024415275,
+      "loss": 2.5752,
+      "step": 570
+    },
+    {
+      "epoch": 0.26189657149409473,
+      "grad_norm": 0.2878377139568329,
+      "learning_rate": 0.0009914463521558546,
+      "loss": 1.4696,
+      "step": 571
+    },
+    {
+      "epoch": 0.2623552344914574,
+      "grad_norm": 0.24708285927772522,
+      "learning_rate": 0.000991400681314848,
+      "loss": 1.3876,
+      "step": 572
+    },
+    {
+      "epoch": 0.2628138974888201,
+      "grad_norm": 0.38844871520996094,
+      "learning_rate": 0.0009913548899297111,
+      "loss": 2.4133,
+      "step": 573
+    },
+    {
+      "epoch": 0.26327256048618275,
+      "grad_norm": 0.3721879720687866,
+      "learning_rate": 0.0009913089780116768,
+      "loss": 1.9523,
+      "step": 574
+    },
+    {
+      "epoch": 0.26373122348354544,
+      "grad_norm": 0.304868221282959,
+      "learning_rate": 0.000991262945572008,
+      "loss": 1.7794,
+      "step": 575
+    },
+    {
+      "epoch": 0.26418988648090813,
+      "grad_norm": 0.24861812591552734,
+      "learning_rate": 0.000991216792621996,
+      "loss": 1.2682,
+      "step": 576
+    },
+    {
+      "epoch": 0.2646485494782708,
+      "grad_norm": 0.44407734274864197,
+      "learning_rate": 0.0009911705191729633,
+      "loss": 2.2016,
+      "step": 577
+    },
+    {
+      "epoch": 0.2651072124756335,
+      "grad_norm": 0.33282753825187683,
+      "learning_rate": 0.0009911241252362603,
+      "loss": 1.6449,
+      "step": 578
+    },
+    {
+      "epoch": 0.2655658754729962,
+      "grad_norm": 0.3493805527687073,
+      "learning_rate": 0.0009910776108232686,
+      "loss": 1.6497,
+      "step": 579
+    },
+    {
+      "epoch": 0.2660245384703589,
+      "grad_norm": 0.36523211002349854,
+      "learning_rate": 0.000991030975945398,
+      "loss": 2.0684,
+      "step": 580
+    },
+    {
+      "epoch": 0.2664832014677216,
+      "grad_norm": 0.5898191928863525,
+      "learning_rate": 0.0009909842206140883,
+      "loss": 1.5042,
+      "step": 581
+    },
+    {
+      "epoch": 0.2669418644650843,
+      "grad_norm": 0.14400826394557953,
+      "learning_rate": 0.0009909373448408092,
+      "loss": 0.7091,
+      "step": 582
+    },
+    {
+      "epoch": 0.26740052746244697,
+      "grad_norm": 0.36130577325820923,
+      "learning_rate": 0.0009908903486370597,
+      "loss": 1.7273,
+      "step": 583
+    },
+    {
+      "epoch": 0.26785919045980966,
+      "grad_norm": 0.2697531282901764,
+      "learning_rate": 0.000990843232014368,
+      "loss": 1.4963,
+      "step": 584
+    },
+    {
+      "epoch": 0.26831785345717235,
+      "grad_norm": 0.34319204092025757,
+      "learning_rate": 0.0009907959949842925,
+      "loss": 1.804,
+      "step": 585
+    },
+    {
+      "epoch": 0.26877651645453504,
+      "grad_norm": 0.19325220584869385,
+      "learning_rate": 0.0009907486375584204,
+      "loss": 0.9886,
+      "step": 586
+    },
+    {
+      "epoch": 0.26923517945189773,
+      "grad_norm": 0.23447225987911224,
+      "learning_rate": 0.0009907011597483691,
+      "loss": 1.2031,
+      "step": 587
+    },
+    {
+      "epoch": 0.2696938424492604,
+      "grad_norm": 0.32570499181747437,
+      "learning_rate": 0.0009906535615657852,
+      "loss": 1.8141,
+      "step": 588
+    },
+    {
+      "epoch": 0.2701525054466231,
+      "grad_norm": 0.40676188468933105,
+      "learning_rate": 0.0009906058430223447,
+      "loss": 2.1973,
+      "step": 589
+    },
+    {
+      "epoch": 0.2706111684439858,
+      "grad_norm": 0.3603213429450989,
+      "learning_rate": 0.0009905580041297537,
+      "loss": 1.9151,
+      "step": 590
+    },
+    {
+      "epoch": 0.2710698314413485,
+      "grad_norm": 0.3043905794620514,
+      "learning_rate": 0.0009905100448997471,
+      "loss": 1.5438,
+      "step": 591
+    },
+    {
+      "epoch": 0.27152849443871113,
+      "grad_norm": 0.1927344799041748,
+      "learning_rate": 0.0009904619653440898,
+      "loss": 0.9269,
+      "step": 592
+    },
+    {
+      "epoch": 0.2719871574360738,
+      "grad_norm": 0.29657939076423645,
+      "learning_rate": 0.000990413765474576,
+      "loss": 1.4634,
+      "step": 593
+    },
+    {
+      "epoch": 0.2724458204334365,
+      "grad_norm": 0.1932910680770874,
+      "learning_rate": 0.0009903654453030293,
+      "loss": 1.029,
+      "step": 594
+    },
+    {
+      "epoch": 0.2729044834307992,
+      "grad_norm": 0.21558885276317596,
+      "learning_rate": 0.0009903170048413032,
+      "loss": 1.0732,
+      "step": 595
+    },
+    {
+      "epoch": 0.2733631464281619,
+      "grad_norm": 0.32209712266921997,
+      "learning_rate": 0.0009902684441012804,
+      "loss": 1.5956,
+      "step": 596
+    },
+    {
+      "epoch": 0.2738218094255246,
+      "grad_norm": 0.36815398931503296,
+      "learning_rate": 0.0009902197630948734,
+      "loss": 1.8898,
+      "step": 597
+    },
+    {
+      "epoch": 0.2742804724228873,
+      "grad_norm": 0.28202107548713684,
+      "learning_rate": 0.0009901709618340237,
+      "loss": 1.4518,
+      "step": 598
+    },
+    {
+      "epoch": 0.27473913542024997,
+      "grad_norm": 0.36555325984954834,
+      "learning_rate": 0.0009901220403307027,
+      "loss": 1.9685,
+      "step": 599
+    },
+    {
+      "epoch": 0.27519779841761266,
+      "grad_norm": 0.35351619124412537,
+      "learning_rate": 0.0009900729985969114,
+      "loss": 2.0413,
+      "step": 600
+    },
+    {
+      "epoch": 0.27565646141497535,
+      "grad_norm": 0.2880600690841675,
+      "learning_rate": 0.0009900238366446798,
+      "loss": 1.527,
+      "step": 601
+    },
+    {
+      "epoch": 0.27611512441233804,
+      "grad_norm": 0.24242261052131653,
+      "learning_rate": 0.000989974554486068,
+      "loss": 1.2521,
+      "step": 602
+    },
+    {
+      "epoch": 0.27657378740970073,
+      "grad_norm": 0.3457260727882385,
+      "learning_rate": 0.000989925152133165,
+      "loss": 1.9675,
+      "step": 603
+    },
+    {
+      "epoch": 0.2770324504070634,
+      "grad_norm": 0.337833046913147,
+      "learning_rate": 0.0009898756295980897,
+      "loss": 1.7602,
+      "step": 604
+    },
+    {
+      "epoch": 0.2774911134044261,
+      "grad_norm": 0.6493378281593323,
+      "learning_rate": 0.00098982598689299,
+      "loss": 1.5513,
+      "step": 605
+    },
+    {
+      "epoch": 0.2779497764017888,
+      "grad_norm": 0.34193724393844604,
+      "learning_rate": 0.0009897762240300442,
+      "loss": 1.8746,
+      "step": 606
+    },
+    {
+      "epoch": 0.2784084393991515,
+      "grad_norm": 0.33631783723831177,
+      "learning_rate": 0.000989726341021459,
+      "loss": 1.6999,
+      "step": 607
+    },
+    {
+      "epoch": 0.2788671023965142,
+      "grad_norm": 0.13958394527435303,
+      "learning_rate": 0.0009896763378794712,
+      "loss": 0.6808,
+      "step": 608
+    },
+    {
+      "epoch": 0.2793257653938768,
+      "grad_norm": 0.2809588313102722,
+      "learning_rate": 0.0009896262146163472,
+      "loss": 1.3181,
+      "step": 609
+    },
+    {
+      "epoch": 0.2797844283912395,
+      "grad_norm": 0.3490467965602875,
+      "learning_rate": 0.0009895759712443821,
+      "loss": 2.0132,
+      "step": 610
+    },
+    {
+      "epoch": 0.2802430913886022,
+      "grad_norm": 0.3802107274532318,
+      "learning_rate": 0.0009895256077759014,
+      "loss": 2.2368,
+      "step": 611
+    },
+    {
+      "epoch": 0.2807017543859649,
+      "grad_norm": 0.2700074017047882,
+      "learning_rate": 0.0009894751242232596,
+      "loss": 1.6661,
+      "step": 612
+    },
+    {
+      "epoch": 0.2811604173833276,
+      "grad_norm": 0.3280056416988373,
+      "learning_rate": 0.0009894245205988401,
+      "loss": 1.6721,
+      "step": 613
+    },
+    {
+      "epoch": 0.2816190803806903,
+      "grad_norm": 0.31935131549835205,
+      "learning_rate": 0.0009893737969150572,
+      "loss": 1.3081,
+      "step": 614
+    },
+    {
+      "epoch": 0.28207774337805297,
+      "grad_norm": 0.4210816025733948,
+      "learning_rate": 0.000989322953184353,
+      "loss": 2.1636,
+      "step": 615
+    },
+    {
+      "epoch": 0.28253640637541566,
+      "grad_norm": 0.38598984479904175,
+      "learning_rate": 0.0009892719894192003,
+      "loss": 2.2451,
+      "step": 616
+    },
+    {
+      "epoch": 0.28299506937277835,
+      "grad_norm": 0.28689202666282654,
+      "learning_rate": 0.0009892209056321009,
+      "loss": 1.0069,
+      "step": 617
+    },
+    {
+      "epoch": 0.28345373237014104,
+      "grad_norm": 0.23823553323745728,
+      "learning_rate": 0.0009891697018355858,
+      "loss": 1.0115,
+      "step": 618
+    },
+    {
+      "epoch": 0.28391239536750373,
+      "grad_norm": 0.4259023666381836,
+      "learning_rate": 0.0009891183780422158,
+      "loss": 1.9298,
+      "step": 619
+    },
+    {
+      "epoch": 0.2843710583648664,
+      "grad_norm": 0.3770276606082916,
+      "learning_rate": 0.0009890669342645807,
+      "loss": 1.7634,
+      "step": 620
+    },
+    {
+      "epoch": 0.2848297213622291,
+      "grad_norm": 0.42291995882987976,
+      "learning_rate": 0.0009890153705153005,
+      "loss": 1.9407,
+      "step": 621
+    },
+    {
+      "epoch": 0.2852883843595918,
+      "grad_norm": 0.35341787338256836,
+      "learning_rate": 0.0009889636868070238,
+      "loss": 1.7191,
+      "step": 622
+    },
+    {
+      "epoch": 0.2857470473569545,
+      "grad_norm": 0.3987595736980438,
+      "learning_rate": 0.000988911883152429,
+      "loss": 2.0328,
+      "step": 623
+    },
+    {
+      "epoch": 0.2862057103543172,
+      "grad_norm": 0.30864641070365906,
+      "learning_rate": 0.0009888599595642244,
+      "loss": 1.351,
+      "step": 624
+    },
+    {
+      "epoch": 0.2866643733516799,
+      "grad_norm": 0.25297942757606506,
+      "learning_rate": 0.0009888079160551466,
+      "loss": 1.2824,
+      "step": 625
+    },
+    {
+      "epoch": 0.2871230363490425,
+      "grad_norm": 0.4757532477378845,
+      "learning_rate": 0.0009887557526379626,
+      "loss": 2.8047,
+      "step": 626
+    },
+    {
+      "epoch": 0.2875816993464052,
+      "grad_norm": 0.3067730665206909,
+      "learning_rate": 0.0009887034693254683,
+      "loss": 1.437,
+      "step": 627
+    },
+    {
+      "epoch": 0.2880403623437679,
+      "grad_norm": 0.3403368294239044,
+      "learning_rate": 0.0009886510661304892,
+      "loss": 1.9881,
+      "step": 628
+    },
+    {
+      "epoch": 0.2884990253411306,
+      "grad_norm": 0.21943096816539764,
+      "learning_rate": 0.0009885985430658803,
+      "loss": 1.1086,
+      "step": 629
+    },
+    {
+      "epoch": 0.2889576883384933,
+      "grad_norm": 0.19929127395153046,
+      "learning_rate": 0.000988545900144526,
+      "loss": 1.0739,
+      "step": 630
+    },
+    {
+      "epoch": 0.28941635133585597,
+      "grad_norm": 0.30291882157325745,
+      "learning_rate": 0.0009884931373793395,
+      "loss": 1.7326,
+      "step": 631
+    },
+    {
+      "epoch": 0.28987501433321866,
+      "grad_norm": 0.23580683767795563,
+      "learning_rate": 0.0009884402547832643,
+      "loss": 1.3195,
+      "step": 632
+    },
+    {
+      "epoch": 0.29033367733058135,
+      "grad_norm": 0.24452835321426392,
+      "learning_rate": 0.0009883872523692727,
+      "loss": 1.2024,
+      "step": 633
+    },
+    {
+      "epoch": 0.29079234032794404,
+      "grad_norm": 0.3366321325302124,
+      "learning_rate": 0.0009883341301503666,
+      "loss": 1.7041,
+      "step": 634
+    },
+    {
+      "epoch": 0.29125100332530673,
+      "grad_norm": 0.23907805979251862,
+      "learning_rate": 0.0009882808881395773,
+      "loss": 1.2118,
+      "step": 635
+    },
+    {
+      "epoch": 0.2917096663226694,
+      "grad_norm": 0.28174489736557007,
+      "learning_rate": 0.0009882275263499655,
+      "loss": 1.494,
+      "step": 636
+    },
+    {
+      "epoch": 0.2921683293200321,
+      "grad_norm": 0.2500317394733429,
+      "learning_rate": 0.0009881740447946212,
+      "loss": 1.379,
+      "step": 637
+    },
+    {
+      "epoch": 0.2926269923173948,
+      "grad_norm": 0.3939465880393982,
+      "learning_rate": 0.000988120443486664,
+      "loss": 1.8527,
+      "step": 638
+    },
+    {
+      "epoch": 0.2930856553147575,
+      "grad_norm": 0.2760303318500519,
+      "learning_rate": 0.0009880667224392422,
+      "loss": 1.5243,
+      "step": 639
+    },
+    {
+      "epoch": 0.2935443183121202,
+      "grad_norm": 0.252908855676651,
+      "learning_rate": 0.0009880128816655343,
+      "loss": 1.0428,
+      "step": 640
+    },
+    {
+      "epoch": 0.2940029813094829,
+      "grad_norm": 0.5029557943344116,
+      "learning_rate": 0.0009879589211787478,
+      "loss": 2.1211,
+      "step": 641
+    },
+    {
+      "epoch": 0.29446164430684557,
+      "grad_norm": 0.3588217794895172,
+      "learning_rate": 0.0009879048409921196,
+      "loss": 1.4391,
+      "step": 642
+    },
+    {
+      "epoch": 0.2949203073042082,
+      "grad_norm": 0.28557729721069336,
+      "learning_rate": 0.000987850641118916,
+      "loss": 1.3001,
+      "step": 643
+    },
+    {
+      "epoch": 0.2953789703015709,
+      "grad_norm": 0.42318931221961975,
+      "learning_rate": 0.0009877963215724323,
+      "loss": 1.9521,
+      "step": 644
+    },
+    {
+      "epoch": 0.2958376332989336,
+      "grad_norm": 0.22379803657531738,
+      "learning_rate": 0.000987741882365994,
+      "loss": 1.101,
+      "step": 645
+    },
+    {
+      "epoch": 0.2962962962962963,
+      "grad_norm": 0.31854167580604553,
+      "learning_rate": 0.000987687323512955,
+      "loss": 1.4523,
+      "step": 646
+    },
+    {
+      "epoch": 0.29675495929365897,
+      "grad_norm": 0.2017553597688675,
+      "learning_rate": 0.0009876326450266992,
+      "loss": 0.9974,
+      "step": 647
+    },
+    {
+      "epoch": 0.29721362229102166,
+      "grad_norm": 0.19907866418361664,
+      "learning_rate": 0.0009875778469206396,
+      "loss": 0.9805,
+      "step": 648
+    },
+    {
+      "epoch": 0.29767228528838435,
+      "grad_norm": 0.36655089259147644,
+      "learning_rate": 0.0009875229292082184,
+      "loss": 2.0823,
+      "step": 649
+    },
+    {
+      "epoch": 0.29813094828574704,
+      "grad_norm": 0.30942440032958984,
+      "learning_rate": 0.0009874678919029073,
+      "loss": 1.5203,
+      "step": 650
+    },
+    {
+      "epoch": 0.29858961128310973,
+      "grad_norm": 0.3907017707824707,
+      "learning_rate": 0.0009874127350182078,
+      "loss": 2.2189,
+      "step": 651
+    },
+    {
+      "epoch": 0.2990482742804724,
+      "grad_norm": 0.3971099555492401,
+      "learning_rate": 0.00098735745856765,
+      "loss": 2.0348,
+      "step": 652
+    },
+    {
+      "epoch": 0.2995069372778351,
+      "grad_norm": 0.24980095028877258,
+      "learning_rate": 0.0009873020625647934,
+      "loss": 1.2255,
+      "step": 653
+    },
+    {
+      "epoch": 0.2999656002751978,
+      "grad_norm": 0.21592852473258972,
+      "learning_rate": 0.0009872465470232273,
+      "loss": 1.0114,
+      "step": 654
+    },
+    {
+      "epoch": 0.3004242632725605,
+      "grad_norm": 0.3222113847732544,
+      "learning_rate": 0.00098719091195657,
+      "loss": 1.4415,
+      "step": 655
+    },
+    {
+      "epoch": 0.3008829262699232,
+      "grad_norm": 0.3443264067173004,
+      "learning_rate": 0.000987135157378469,
+      "loss": 1.6269,
+      "step": 656
+    },
+    {
+      "epoch": 0.3013415892672859,
+      "grad_norm": 0.32391127943992615,
+      "learning_rate": 0.0009870792833026017,
+      "loss": 1.5997,
+      "step": 657
+    },
+    {
+      "epoch": 0.30180025226464857,
+      "grad_norm": 0.20386099815368652,
+      "learning_rate": 0.000987023289742674,
+      "loss": 1.1451,
+      "step": 658
+    },
+    {
+      "epoch": 0.30225891526201126,
+      "grad_norm": 0.24229376018047333,
+      "learning_rate": 0.0009869671767124217,
+      "loss": 1.2804,
+      "step": 659
+    },
+    {
+      "epoch": 0.3027175782593739,
+      "grad_norm": 0.3021294176578522,
+      "learning_rate": 0.00098691094422561,
+      "loss": 1.8578,
+      "step": 660
+    },
+    {
+      "epoch": 0.3031762412567366,
+      "grad_norm": 0.33602842688560486,
+      "learning_rate": 0.0009868545922960326,
+      "loss": 1.9222,
+      "step": 661
+    },
+    {
+      "epoch": 0.3036349042540993,
+      "grad_norm": 0.2946925759315491,
+      "learning_rate": 0.0009867981209375134,
+      "loss": 1.6532,
+      "step": 662
+    },
+    {
+      "epoch": 0.30409356725146197,
+      "grad_norm": 0.2836867570877075,
+      "learning_rate": 0.0009867415301639051,
+      "loss": 1.5413,
+      "step": 663
+    },
+    {
+      "epoch": 0.30455223024882466,
+      "grad_norm": 0.3318963050842285,
+      "learning_rate": 0.00098668481998909,
+      "loss": 1.9834,
+      "step": 664
+    },
+    {
+      "epoch": 0.30501089324618735,
+      "grad_norm": 0.2361811101436615,
+      "learning_rate": 0.0009866279904269793,
+      "loss": 1.1826,
+      "step": 665
+    },
+    {
+      "epoch": 0.30546955624355004,
+      "grad_norm": 0.20993195474147797,
+      "learning_rate": 0.0009865710414915137,
+      "loss": 1.0908,
+      "step": 666
+    },
+    {
+      "epoch": 0.30592821924091274,
+      "grad_norm": 0.11743535846471786,
+      "learning_rate": 0.0009865139731966633,
+      "loss": 0.6957,
+      "step": 667
+    },
+    {
+      "epoch": 0.3063868822382754,
+      "grad_norm": 0.4895835816860199,
+      "learning_rate": 0.0009864567855564273,
+      "loss": 2.6428,
+      "step": 668
+    },
+    {
+      "epoch": 0.3068455452356381,
+      "grad_norm": 0.3493427038192749,
+      "learning_rate": 0.0009863994785848343,
+      "loss": 1.5343,
+      "step": 669
+    },
+    {
+      "epoch": 0.3073042082330008,
+      "grad_norm": 0.36788055300712585,
+      "learning_rate": 0.0009863420522959422,
+      "loss": 1.7888,
+      "step": 670
+    },
+    {
+      "epoch": 0.3077628712303635,
+      "grad_norm": 0.3058752715587616,
+      "learning_rate": 0.0009862845067038378,
+      "loss": 1.4943,
+      "step": 671
+    },
+    {
+      "epoch": 0.3082215342277262,
+      "grad_norm": 0.29403895139694214,
+      "learning_rate": 0.0009862268418226376,
+      "loss": 1.4614,
+      "step": 672
+    },
+    {
+      "epoch": 0.3086801972250889,
+      "grad_norm": 0.3854977488517761,
+      "learning_rate": 0.0009861690576664873,
+      "loss": 1.4861,
+      "step": 673
+    },
+    {
+      "epoch": 0.3091388602224516,
+      "grad_norm": 0.404697984457016,
+      "learning_rate": 0.0009861111542495617,
+      "loss": 2.0234,
+      "step": 674
+    },
+    {
+      "epoch": 0.30959752321981426,
+      "grad_norm": 0.3911391496658325,
+      "learning_rate": 0.000986053131586065,
+      "loss": 2.011,
+      "step": 675
+    },
+    {
+      "epoch": 0.31005618621717695,
+      "grad_norm": 0.2827008068561554,
+      "learning_rate": 0.0009859949896902304,
+      "loss": 1.0028,
+      "step": 676
+    },
+    {
+      "epoch": 0.3105148492145396,
+      "grad_norm": 0.3550345003604889,
+      "learning_rate": 0.0009859367285763206,
+      "loss": 1.6724,
+      "step": 677
+    },
+    {
+      "epoch": 0.3109735122119023,
+      "grad_norm": 0.33875733613967896,
+      "learning_rate": 0.0009858783482586276,
+      "loss": 1.8332,
+      "step": 678
+    },
+    {
+      "epoch": 0.31143217520926497,
+      "grad_norm": 0.3210250437259674,
+      "learning_rate": 0.0009858198487514723,
+      "loss": 1.4062,
+      "step": 679
+    },
+    {
+      "epoch": 0.31189083820662766,
+      "grad_norm": 0.24080604314804077,
+      "learning_rate": 0.0009857612300692052,
+      "loss": 1.1792,
+      "step": 680
+    },
+    {
+      "epoch": 0.31234950120399035,
+      "grad_norm": 0.2961960434913635,
+      "learning_rate": 0.0009857024922262057,
+      "loss": 1.3523,
+      "step": 681
+    },
+    {
+      "epoch": 0.31280816420135305,
+      "grad_norm": 0.32249388098716736,
+      "learning_rate": 0.0009856436352368829,
+      "loss": 1.6275,
+      "step": 682
+    },
+    {
+      "epoch": 0.31326682719871574,
+      "grad_norm": 0.23962727189064026,
+      "learning_rate": 0.0009855846591156747,
+      "loss": 1.3343,
+      "step": 683
+    },
+    {
+      "epoch": 0.3137254901960784,
+      "grad_norm": 0.24461761116981506,
+      "learning_rate": 0.0009855255638770485,
+      "loss": 1.4974,
+      "step": 684
+    },
+    {
+      "epoch": 0.3141841531934411,
+      "grad_norm": 0.36309170722961426,
+      "learning_rate": 0.0009854663495355005,
+      "loss": 1.7183,
+      "step": 685
+    },
+    {
+      "epoch": 0.3146428161908038,
+      "grad_norm": 0.35972607135772705,
+      "learning_rate": 0.0009854070161055566,
+      "loss": 1.9509,
+      "step": 686
+    },
+    {
+      "epoch": 0.3151014791881665,
+      "grad_norm": 0.3379707932472229,
+      "learning_rate": 0.0009853475636017717,
+      "loss": 1.9371,
+      "step": 687
+    },
+    {
+      "epoch": 0.3155601421855292,
+      "grad_norm": 0.3258635401725769,
+      "learning_rate": 0.0009852879920387299,
+      "loss": 1.567,
+      "step": 688
+    },
+    {
+      "epoch": 0.3160188051828919,
+      "grad_norm": 0.36305832862854004,
+      "learning_rate": 0.0009852283014310445,
+      "loss": 2.1206,
+      "step": 689
+    },
+    {
+      "epoch": 0.3164774681802546,
+      "grad_norm": 0.3250496983528137,
+      "learning_rate": 0.000985168491793358,
+      "loss": 1.7597,
+      "step": 690
+    },
+    {
+      "epoch": 0.31693613117761726,
+      "grad_norm": 0.39209747314453125,
+      "learning_rate": 0.0009851085631403425,
+      "loss": 2.2836,
+      "step": 691
+    },
+    {
+      "epoch": 0.31739479417497996,
+      "grad_norm": 0.3178749978542328,
+      "learning_rate": 0.0009850485154866983,
+      "loss": 1.8535,
+      "step": 692
+    },
+    {
+      "epoch": 0.31785345717234265,
+      "grad_norm": 0.39518505334854126,
+      "learning_rate": 0.0009849883488471562,
+      "loss": 1.8896,
+      "step": 693
+    },
+    {
+      "epoch": 0.3183121201697053,
+      "grad_norm": 0.49922698736190796,
+      "learning_rate": 0.000984928063236475,
+      "loss": 1.8992,
+      "step": 694
+    },
+    {
+      "epoch": 0.318770783167068,
+      "grad_norm": 0.3633042871952057,
+      "learning_rate": 0.0009848676586694434,
+      "loss": 1.8554,
+      "step": 695
+    },
+    {
+      "epoch": 0.31922944616443066,
+      "grad_norm": 0.4339780807495117,
+      "learning_rate": 0.000984807135160879,
+      "loss": 2.4392,
+      "step": 696
+    },
+    {
+      "epoch": 0.31968810916179335,
+      "grad_norm": 0.2694472074508667,
+      "learning_rate": 0.0009847464927256288,
+      "loss": 1.1097,
+      "step": 697
+    },
+    {
+      "epoch": 0.32014677215915605,
+      "grad_norm": 0.3319615125656128,
+      "learning_rate": 0.000984685731378569,
+      "loss": 1.5832,
+      "step": 698
+    },
+    {
+      "epoch": 0.32060543515651874,
+      "grad_norm": 0.39388328790664673,
+      "learning_rate": 0.0009846248511346041,
+      "loss": 2.2849,
+      "step": 699
+    },
+    {
+      "epoch": 0.3210640981538814,
+      "grad_norm": 0.42792367935180664,
+      "learning_rate": 0.0009845638520086688,
+      "loss": 2.3538,
+      "step": 700
+    },
+    {
+      "epoch": 0.3215227611512441,
+      "grad_norm": 0.4416462182998657,
+      "learning_rate": 0.000984502734015727,
+      "loss": 2.5254,
+      "step": 701
+    },
+    {
+      "epoch": 0.3219814241486068,
+      "grad_norm": 0.3952227532863617,
+      "learning_rate": 0.000984441497170771,
+      "loss": 2.0428,
+      "step": 702
+    },
+    {
+      "epoch": 0.3224400871459695,
+      "grad_norm": 0.2395290583372116,
+      "learning_rate": 0.000984380141488823,
+      "loss": 0.9864,
+      "step": 703
+    },
+    {
+      "epoch": 0.3228987501433322,
+      "grad_norm": 0.4042597711086273,
+      "learning_rate": 0.0009843186669849333,
+      "loss": 1.9032,
+      "step": 704
+    },
+    {
+      "epoch": 0.3233574131406949,
+      "grad_norm": 0.3238537013530731,
+      "learning_rate": 0.0009842570736741827,
+      "loss": 1.6617,
+      "step": 705
+    },
+    {
+      "epoch": 0.3238160761380576,
+      "grad_norm": 0.3476428985595703,
+      "learning_rate": 0.0009841953615716804,
+      "loss": 1.761,
+      "step": 706
+    },
+    {
+      "epoch": 0.32427473913542026,
+      "grad_norm": 0.47932443022727966,
+      "learning_rate": 0.0009841335306925644,
+      "loss": 2.4644,
+      "step": 707
+    },
+    {
+      "epoch": 0.32473340213278296,
+      "grad_norm": 0.39882582426071167,
+      "learning_rate": 0.0009840715810520027,
+      "loss": 2.0681,
+      "step": 708
+    },
+    {
+      "epoch": 0.32519206513014565,
+      "grad_norm": 0.3567412793636322,
+      "learning_rate": 0.0009840095126651919,
+      "loss": 1.4626,
+      "step": 709
+    },
+    {
+      "epoch": 0.32565072812750834,
+      "grad_norm": 0.29540038108825684,
+      "learning_rate": 0.0009839473255473575,
+      "loss": 1.2916,
+      "step": 710
+    },
+    {
+      "epoch": 0.32610939112487103,
+      "grad_norm": 0.35967186093330383,
+      "learning_rate": 0.0009838850197137548,
+      "loss": 1.5812,
+      "step": 711
+    },
+    {
+      "epoch": 0.32656805412223366,
+      "grad_norm": 0.4076613485813141,
+      "learning_rate": 0.0009838225951796678,
+      "loss": 1.8511,
+      "step": 712
+    },
+    {
+      "epoch": 0.32702671711959636,
+      "grad_norm": 0.37241417169570923,
+      "learning_rate": 0.0009837600519604097,
+      "loss": 1.3398,
+      "step": 713
+    },
+    {
+      "epoch": 0.32748538011695905,
+      "grad_norm": 0.3827233612537384,
+      "learning_rate": 0.0009836973900713226,
+      "loss": 1.9355,
+      "step": 714
+    },
+    {
+      "epoch": 0.32794404311432174,
+      "grad_norm": 0.3503935635089874,
+      "learning_rate": 0.0009836346095277782,
+      "loss": 1.5865,
+      "step": 715
+    },
+    {
+      "epoch": 0.32840270611168443,
+      "grad_norm": 0.346055269241333,
+      "learning_rate": 0.000983571710345177,
+      "loss": 1.879,
+      "step": 716
+    },
+    {
+      "epoch": 0.3288613691090471,
+      "grad_norm": 0.3340051770210266,
+      "learning_rate": 0.0009835086925389484,
+      "loss": 1.531,
+      "step": 717
+    },
+    {
+      "epoch": 0.3293200321064098,
+      "grad_norm": 0.3591856360435486,
+      "learning_rate": 0.0009834455561245512,
+      "loss": 1.5611,
+      "step": 718
+    },
+    {
+      "epoch": 0.3297786951037725,
+      "grad_norm": 0.08197461068630219,
+      "learning_rate": 0.0009833823011174735,
+      "loss": 0.45,
+      "step": 719
+    },
+    {
+      "epoch": 0.3302373581011352,
+      "grad_norm": 0.35696059465408325,
+      "learning_rate": 0.0009833189275332318,
+      "loss": 2.2035,
+      "step": 720
+    },
+    {
+      "epoch": 0.3306960210984979,
+      "grad_norm": 0.21588194370269775,
+      "learning_rate": 0.0009832554353873721,
+      "loss": 1.1911,
+      "step": 721
+    },
+    {
+      "epoch": 0.3311546840958606,
+      "grad_norm": 0.313678503036499,
+      "learning_rate": 0.00098319182469547,
+      "loss": 1.956,
+      "step": 722
+    },
+    {
+      "epoch": 0.33161334709322327,
+      "grad_norm": 0.3138997554779053,
+      "learning_rate": 0.000983128095473129,
+      "loss": 1.5287,
+      "step": 723
+    },
+    {
+      "epoch": 0.33207201009058596,
+      "grad_norm": 0.38969993591308594,
+      "learning_rate": 0.0009830642477359828,
+      "loss": 2.4346,
+      "step": 724
+    },
+    {
+      "epoch": 0.33253067308794865,
+      "grad_norm": 0.21810193359851837,
+      "learning_rate": 0.0009830002814996935,
+      "loss": 1.0224,
+      "step": 725
+    },
+    {
+      "epoch": 0.33298933608531134,
+      "grad_norm": 0.11237790435552597,
+      "learning_rate": 0.0009829361967799526,
+      "loss": 0.6925,
+      "step": 726
+    },
+    {
+      "epoch": 0.33344799908267403,
+      "grad_norm": 0.31989771127700806,
+      "learning_rate": 0.0009828719935924804,
+      "loss": 1.4384,
+      "step": 727
+    },
+    {
+      "epoch": 0.3339066620800367,
+      "grad_norm": 0.4660812318325043,
+      "learning_rate": 0.0009828076719530265,
+      "loss": 2.5269,
+      "step": 728
+    },
+    {
+      "epoch": 0.33436532507739936,
+      "grad_norm": 0.37333112955093384,
+      "learning_rate": 0.0009827432318773694,
+      "loss": 2.13,
+      "step": 729
+    },
+    {
+      "epoch": 0.33482398807476205,
+      "grad_norm": 0.32906556129455566,
+      "learning_rate": 0.000982678673381317,
+      "loss": 1.5657,
+      "step": 730
+    },
+    {
+      "epoch": 0.33528265107212474,
+      "grad_norm": 0.3006758689880371,
+      "learning_rate": 0.0009826139964807055,
+      "loss": 1.4011,
+      "step": 731
+    },
+    {
+      "epoch": 0.33574131406948743,
+      "grad_norm": 0.3466986119747162,
+      "learning_rate": 0.0009825492011914009,
+      "loss": 2.008,
+      "step": 732
+    },
+    {
+      "epoch": 0.3361999770668501,
+      "grad_norm": 0.29313912987709045,
+      "learning_rate": 0.0009824842875292977,
+      "loss": 1.337,
+      "step": 733
+    },
+    {
+      "epoch": 0.3366586400642128,
+      "grad_norm": 0.3280053436756134,
+      "learning_rate": 0.0009824192555103202,
+      "loss": 1.6918,
+      "step": 734
+    },
+    {
+      "epoch": 0.3371173030615755,
+      "grad_norm": 0.22816585004329681,
+      "learning_rate": 0.0009823541051504207,
+      "loss": 1.1973,
+      "step": 735
+    },
+    {
+      "epoch": 0.3375759660589382,
+      "grad_norm": 0.3135545551776886,
+      "learning_rate": 0.0009822888364655813,
+      "loss": 1.6632,
+      "step": 736
+    },
+    {
+      "epoch": 0.3380346290563009,
+      "grad_norm": 0.30442047119140625,
+      "learning_rate": 0.0009822234494718128,
+      "loss": 1.3535,
+      "step": 737
+    },
+    {
+      "epoch": 0.3384932920536636,
+      "grad_norm": 0.2706790268421173,
+      "learning_rate": 0.0009821579441851552,
+      "loss": 1.0,
+      "step": 738
+    },
+    {
+      "epoch": 0.33895195505102627,
+      "grad_norm": 0.39678725600242615,
+      "learning_rate": 0.0009820923206216774,
+      "loss": 1.9822,
+      "step": 739
+    },
+    {
+      "epoch": 0.33941061804838896,
+      "grad_norm": 0.38112956285476685,
+      "learning_rate": 0.0009820265787974772,
+      "loss": 2.0559,
+      "step": 740
+    },
+    {
+      "epoch": 0.33986928104575165,
+      "grad_norm": 0.32253459095954895,
+      "learning_rate": 0.0009819607187286816,
+      "loss": 1.389,
+      "step": 741
+    },
+    {
+      "epoch": 0.34032794404311434,
+      "grad_norm": 0.34095335006713867,
+      "learning_rate": 0.0009818947404314465,
+      "loss": 1.7615,
+      "step": 742
+    },
+    {
+      "epoch": 0.34078660704047703,
+      "grad_norm": 0.36802440881729126,
+      "learning_rate": 0.0009818286439219571,
+      "loss": 1.8613,
+      "step": 743
+    },
+    {
+      "epoch": 0.3412452700378397,
+      "grad_norm": 0.415279746055603,
+      "learning_rate": 0.0009817624292164271,
+      "loss": 1.9945,
+      "step": 744
+    },
+    {
+      "epoch": 0.3417039330352024,
+      "grad_norm": 0.3048365116119385,
+      "learning_rate": 0.0009816960963310998,
+      "loss": 1.6786,
+      "step": 745
+    },
+    {
+      "epoch": 0.34216259603256505,
+      "grad_norm": 0.23415561020374298,
+      "learning_rate": 0.0009816296452822464,
+      "loss": 1.0875,
+      "step": 746
+    },
+    {
+      "epoch": 0.34262125902992774,
+      "grad_norm": 0.3987458348274231,
+      "learning_rate": 0.0009815630760861684,
+      "loss": 1.94,
+      "step": 747
+    },
+    {
+      "epoch": 0.34307992202729043,
+      "grad_norm": 0.31393012404441833,
+      "learning_rate": 0.000981496388759196,
+      "loss": 1.4587,
+      "step": 748
+    },
+    {
+      "epoch": 0.3435385850246531,
+      "grad_norm": 0.29491788148880005,
+      "learning_rate": 0.0009814295833176871,
+      "loss": 1.3318,
+      "step": 749
+    },
+    {
+      "epoch": 0.3439972480220158,
+      "grad_norm": 0.23884756863117218,
+      "learning_rate": 0.0009813626597780305,
+      "loss": 0.908,
+      "step": 750
+    },
+    {
+      "epoch": 0.3444559110193785,
+      "grad_norm": 0.3067830502986908,
+      "learning_rate": 0.0009812956181566425,
+      "loss": 1.4077,
+      "step": 751
+    },
+    {
+      "epoch": 0.3449145740167412,
+      "grad_norm": 0.42813050746917725,
+      "learning_rate": 0.000981228458469969,
+      "loss": 2.1209,
+      "step": 752
+    },
+    {
+      "epoch": 0.3453732370141039,
+      "grad_norm": 0.20336052775382996,
+      "learning_rate": 0.0009811611807344848,
+      "loss": 0.9812,
+      "step": 753
+    },
+    {
+      "epoch": 0.3458319000114666,
+      "grad_norm": 0.35509541630744934,
+      "learning_rate": 0.0009810937849666938,
+      "loss": 2.0477,
+      "step": 754
+    },
+    {
+      "epoch": 0.34629056300882927,
+      "grad_norm": 0.3570297360420227,
+      "learning_rate": 0.0009810262711831283,
+      "loss": 2.0165,
+      "step": 755
+    },
+    {
+      "epoch": 0.34674922600619196,
+      "grad_norm": 0.3037537932395935,
+      "learning_rate": 0.0009809586394003502,
+      "loss": 1.6942,
+      "step": 756
+    },
+    {
+      "epoch": 0.34720788900355465,
+      "grad_norm": 0.30765625834465027,
+      "learning_rate": 0.00098089088963495,
+      "loss": 1.7054,
+      "step": 757
+    },
+    {
+      "epoch": 0.34766655200091734,
+      "grad_norm": 0.3930741250514984,
+      "learning_rate": 0.0009808230219035471,
+      "loss": 2.2827,
+      "step": 758
+    },
+    {
+      "epoch": 0.34812521499828003,
+      "grad_norm": 0.34178081154823303,
+      "learning_rate": 0.00098075503622279,
+      "loss": 1.7407,
+      "step": 759
+    },
+    {
+      "epoch": 0.3485838779956427,
+      "grad_norm": 0.2623182237148285,
+      "learning_rate": 0.0009806869326093562,
+      "loss": 1.2565,
+      "step": 760
+    },
+    {
+      "epoch": 0.3490425409930054,
+      "grad_norm": 0.4342869520187378,
+      "learning_rate": 0.0009806187110799518,
+      "loss": 2.2317,
+      "step": 761
+    },
+    {
+      "epoch": 0.3495012039903681,
+      "grad_norm": 0.11686166375875473,
+      "learning_rate": 0.0009805503716513121,
+      "loss": 0.5233,
+      "step": 762
+    },
+    {
+      "epoch": 0.34995986698773074,
+      "grad_norm": 0.3906189799308777,
+      "learning_rate": 0.0009804819143402013,
+      "loss": 1.9533,
+      "step": 763
+    },
+    {
+      "epoch": 0.35041852998509343,
+      "grad_norm": 0.4378143548965454,
+      "learning_rate": 0.0009804133391634125,
+      "loss": 2.1091,
+      "step": 764
+    },
+    {
+      "epoch": 0.3508771929824561,
+      "grad_norm": 0.2528671622276306,
+      "learning_rate": 0.0009803446461377676,
+      "loss": 1.0801,
+      "step": 765
+    },
+    {
+      "epoch": 0.3513358559798188,
+      "grad_norm": 0.2157050520181656,
+      "learning_rate": 0.0009802758352801177,
+      "loss": 1.1054,
+      "step": 766
+    },
+    {
+      "epoch": 0.3517945189771815,
+      "grad_norm": 0.4617224335670471,
+      "learning_rate": 0.0009802069066073423,
+      "loss": 2.3428,
+      "step": 767
+    },
+    {
+      "epoch": 0.3522531819745442,
+      "grad_norm": 0.3335023820400238,
+      "learning_rate": 0.0009801378601363502,
+      "loss": 1.601,
+      "step": 768
+    },
+    {
+      "epoch": 0.3527118449719069,
+      "grad_norm": 0.3270067870616913,
+      "learning_rate": 0.000980068695884079,
+      "loss": 1.4522,
+      "step": 769
+    },
+    {
+      "epoch": 0.3531705079692696,
+      "grad_norm": 0.2701742947101593,
+      "learning_rate": 0.0009799994138674955,
+      "loss": 1.1839,
+      "step": 770
+    },
+    {
+      "epoch": 0.35362917096663227,
+      "grad_norm": 0.3580704629421234,
+      "learning_rate": 0.0009799300141035945,
+      "loss": 1.8822,
+      "step": 771
+    },
+    {
+      "epoch": 0.35408783396399496,
+      "grad_norm": 0.37125471234321594,
+      "learning_rate": 0.0009798604966094007,
+      "loss": 2.0035,
+      "step": 772
+    },
+    {
+      "epoch": 0.35454649696135765,
+      "grad_norm": 0.3791847229003906,
+      "learning_rate": 0.000979790861401967,
+      "loss": 1.8893,
+      "step": 773
+    },
+    {
+      "epoch": 0.35500515995872034,
+      "grad_norm": 0.3457506000995636,
+      "learning_rate": 0.0009797211084983757,
+      "loss": 1.3708,
+      "step": 774
+    },
+    {
+      "epoch": 0.35546382295608303,
+      "grad_norm": 0.4049646854400635,
+      "learning_rate": 0.0009796512379157375,
+      "loss": 1.923,
+      "step": 775
+    },
+    {
+      "epoch": 0.3559224859534457,
+      "grad_norm": 0.41086849570274353,
+      "learning_rate": 0.000979581249671192,
+      "loss": 1.848,
+      "step": 776
+    },
+    {
+      "epoch": 0.3563811489508084,
+      "grad_norm": 0.32881003618240356,
+      "learning_rate": 0.000979511143781908,
+      "loss": 1.7467,
+      "step": 777
+    },
+    {
+      "epoch": 0.3568398119481711,
+      "grad_norm": 0.3037266135215759,
+      "learning_rate": 0.0009794409202650831,
+      "loss": 1.6274,
+      "step": 778
+    },
+    {
+      "epoch": 0.3572984749455338,
+      "grad_norm": 0.34774550795555115,
+      "learning_rate": 0.0009793705791379436,
+      "loss": 1.6964,
+      "step": 779
+    },
+    {
+      "epoch": 0.35775713794289643,
+      "grad_norm": 0.19517025351524353,
+      "learning_rate": 0.0009793001204177444,
+      "loss": 1.1024,
+      "step": 780
+    },
+    {
+      "epoch": 0.3582158009402591,
+      "grad_norm": 0.25842177867889404,
+      "learning_rate": 0.0009792295441217699,
+      "loss": 1.3525,
+      "step": 781
+    },
+    {
+      "epoch": 0.3586744639376218,
+      "grad_norm": 0.2821122407913208,
+      "learning_rate": 0.0009791588502673326,
+      "loss": 1.2322,
+      "step": 782
+    },
+    {
+      "epoch": 0.3591331269349845,
+      "grad_norm": 0.40477707982063293,
+      "learning_rate": 0.0009790880388717746,
+      "loss": 1.7263,
+      "step": 783
+    },
+    {
+      "epoch": 0.3595917899323472,
+      "grad_norm": 0.39657363295555115,
+      "learning_rate": 0.0009790171099524662,
+      "loss": 2.3909,
+      "step": 784
+    },
+    {
+      "epoch": 0.3600504529297099,
+      "grad_norm": 0.31579139828681946,
+      "learning_rate": 0.0009789460635268067,
+      "loss": 1.5805,
+      "step": 785
+    },
+    {
+      "epoch": 0.3605091159270726,
+      "grad_norm": 0.4094010889530182,
+      "learning_rate": 0.0009788748996122246,
+      "loss": 2.2917,
+      "step": 786
+    },
+    {
+      "epoch": 0.36096777892443527,
+      "grad_norm": 0.31713518500328064,
+      "learning_rate": 0.0009788036182261767,
+      "loss": 1.483,
+      "step": 787
+    },
+    {
+      "epoch": 0.36142644192179796,
+      "grad_norm": 0.38277119398117065,
+      "learning_rate": 0.000978732219386149,
+      "loss": 2.1552,
+      "step": 788
+    },
+    {
+      "epoch": 0.36188510491916065,
+      "grad_norm": 0.2237308770418167,
+      "learning_rate": 0.000978660703109656,
+      "loss": 1.2989,
+      "step": 789
+    },
+    {
+      "epoch": 0.36234376791652334,
+      "grad_norm": 0.32253265380859375,
+      "learning_rate": 0.0009785890694142413,
+      "loss": 1.4193,
+      "step": 790
+    },
+    {
+      "epoch": 0.36280243091388603,
+      "grad_norm": 0.28654542565345764,
+      "learning_rate": 0.0009785173183174767,
+      "loss": 0.8873,
+      "step": 791
+    },
+    {
+      "epoch": 0.3632610939112487,
+      "grad_norm": 0.313298761844635,
+      "learning_rate": 0.000978445449836964,
+      "loss": 1.4503,
+      "step": 792
+    },
+    {
+      "epoch": 0.3637197569086114,
+      "grad_norm": 0.3323945701122284,
+      "learning_rate": 0.0009783734639903325,
+      "loss": 1.4465,
+      "step": 793
+    },
+    {
+      "epoch": 0.3641784199059741,
+      "grad_norm": 0.34506741166114807,
+      "learning_rate": 0.0009783013607952411,
+      "loss": 1.4567,
+      "step": 794
+    },
+    {
+      "epoch": 0.3646370829033368,
+      "grad_norm": 0.36164945363998413,
+      "learning_rate": 0.000978229140269377,
+      "loss": 1.8277,
+      "step": 795
+    },
+    {
+      "epoch": 0.3650957459006995,
+      "grad_norm": 0.40301600098609924,
+      "learning_rate": 0.0009781568024304567,
+      "loss": 1.7805,
+      "step": 796
+    },
+    {
+      "epoch": 0.3655544088980621,
+      "grad_norm": 0.2755573093891144,
+      "learning_rate": 0.0009780843472962251,
+      "loss": 1.461,
+      "step": 797
+    },
+    {
+      "epoch": 0.3660130718954248,
+      "grad_norm": 0.3744322657585144,
+      "learning_rate": 0.000978011774884456,
+      "loss": 2.009,
+      "step": 798
+    },
+    {
+      "epoch": 0.3664717348927875,
+      "grad_norm": 0.4854291081428528,
+      "learning_rate": 0.0009779390852129519,
+      "loss": 1.8751,
+      "step": 799
+    },
+    {
+      "epoch": 0.3669303978901502,
+      "grad_norm": 0.21938319504261017,
+      "learning_rate": 0.000977866278299544,
+      "loss": 0.996,
+      "step": 800
+    },
+    {
+      "epoch": 0.3673890608875129,
+      "grad_norm": 0.3286993205547333,
+      "learning_rate": 0.0009777933541620925,
+      "loss": 1.4633,
+      "step": 801
+    },
+    {
+      "epoch": 0.3678477238848756,
+      "grad_norm": 0.3668621778488159,
+      "learning_rate": 0.000977720312818486,
+      "loss": 1.9308,
+      "step": 802
+    },
+    {
+      "epoch": 0.36830638688223827,
+      "grad_norm": 0.36969929933547974,
+      "learning_rate": 0.0009776471542866424,
+      "loss": 1.9707,
+      "step": 803
+    },
+    {
+      "epoch": 0.36876504987960096,
+      "grad_norm": 0.22716723382472992,
+      "learning_rate": 0.0009775738785845077,
+      "loss": 0.9617,
+      "step": 804
+    },
+    {
+      "epoch": 0.36922371287696365,
+      "grad_norm": 0.1840977519750595,
+      "learning_rate": 0.0009775004857300572,
+      "loss": 1.0723,
+      "step": 805
+    },
+    {
+      "epoch": 0.36968237587432634,
+      "grad_norm": 0.2663559019565582,
+      "learning_rate": 0.0009774269757412944,
+      "loss": 1.6182,
+      "step": 806
+    },
+    {
+      "epoch": 0.37014103887168903,
+      "grad_norm": 0.22786211967468262,
+      "learning_rate": 0.0009773533486362523,
+      "loss": 1.11,
+      "step": 807
+    },
+    {
+      "epoch": 0.3705997018690517,
+      "grad_norm": 0.10840432345867157,
+      "learning_rate": 0.0009772796044329917,
+      "loss": 0.6749,
+      "step": 808
+    },
+    {
+      "epoch": 0.3710583648664144,
+      "grad_norm": 0.3396431803703308,
+      "learning_rate": 0.000977205743149603,
+      "loss": 2.0026,
+      "step": 809
+    },
+    {
+      "epoch": 0.3715170278637771,
+      "grad_norm": 0.34386134147644043,
+      "learning_rate": 0.0009771317648042043,
+      "loss": 2.0295,
+      "step": 810
+    },
+    {
+      "epoch": 0.3719756908611398,
+      "grad_norm": 0.22869761288166046,
+      "learning_rate": 0.0009770576694149435,
+      "loss": 1.061,
+      "step": 811
+    },
+    {
+      "epoch": 0.3724343538585025,
+      "grad_norm": 0.20139172673225403,
+      "learning_rate": 0.000976983456999997,
+      "loss": 0.9954,
+      "step": 812
+    },
+    {
+      "epoch": 0.3728930168558652,
+      "grad_norm": 0.21274548768997192,
+      "learning_rate": 0.0009769091275775687,
+      "loss": 1.0131,
+      "step": 813
+    },
+    {
+      "epoch": 0.3733516798532278,
+      "grad_norm": 0.4451286196708679,
+      "learning_rate": 0.0009768346811658932,
+      "loss": 2.5039,
+      "step": 814
+    },
+    {
+      "epoch": 0.3738103428505905,
+      "grad_norm": 0.2519133388996124,
+      "learning_rate": 0.0009767601177832319,
+      "loss": 1.1851,
+      "step": 815
+    },
+    {
+      "epoch": 0.3742690058479532,
+      "grad_norm": 0.36184462904930115,
+      "learning_rate": 0.0009766854374478765,
+      "loss": 1.9926,
+      "step": 816
+    },
+    {
+      "epoch": 0.3747276688453159,
+      "grad_norm": 0.2603318393230438,
+      "learning_rate": 0.0009766106401781459,
+      "loss": 1.3791,
+      "step": 817
+    },
+    {
+      "epoch": 0.3751863318426786,
+      "grad_norm": 0.26677289605140686,
+      "learning_rate": 0.0009765357259923885,
+      "loss": 1.2404,
+      "step": 818
+    },
+    {
+      "epoch": 0.37564499484004127,
+      "grad_norm": 0.1966245174407959,
+      "learning_rate": 0.000976460694908982,
+      "loss": 1.0012,
+      "step": 819
+    },
+    {
+      "epoch": 0.37610365783740396,
+      "grad_norm": 0.24759866297245026,
+      "learning_rate": 0.0009763855469463314,
+      "loss": 1.4005,
+      "step": 820
+    },
+    {
+      "epoch": 0.37656232083476665,
+      "grad_norm": 0.387797474861145,
+      "learning_rate": 0.000976310282122871,
+      "loss": 2.082,
+      "step": 821
+    },
+    {
+      "epoch": 0.37702098383212934,
+      "grad_norm": 0.3424549698829651,
+      "learning_rate": 0.0009762349004570644,
+      "loss": 1.84,
+      "step": 822
+    },
+    {
+      "epoch": 0.37747964682949203,
+      "grad_norm": 0.35073143243789673,
+      "learning_rate": 0.0009761594019674027,
+      "loss": 1.981,
+      "step": 823
+    },
+    {
+      "epoch": 0.3779383098268547,
+      "grad_norm": 0.11143548041582108,
+      "learning_rate": 0.0009760837866724063,
+      "loss": 0.6522,
+      "step": 824
+    },
+    {
+      "epoch": 0.3783969728242174,
+      "grad_norm": 0.2800116539001465,
+      "learning_rate": 0.0009760080545906244,
+      "loss": 1.4713,
+      "step": 825
+    },
+    {
+      "epoch": 0.3788556358215801,
+      "grad_norm": 0.365592896938324,
+      "learning_rate": 0.0009759322057406346,
+      "loss": 2.3906,
+      "step": 826
+    },
+    {
+      "epoch": 0.3793142988189428,
+      "grad_norm": 0.3204135000705719,
+      "learning_rate": 0.0009758562401410429,
+      "loss": 1.6213,
+      "step": 827
+    },
+    {
+      "epoch": 0.3797729618163055,
+      "grad_norm": 0.32533779740333557,
+      "learning_rate": 0.0009757801578104846,
+      "loss": 1.8047,
+      "step": 828
+    },
+    {
+      "epoch": 0.3802316248136682,
+      "grad_norm": 0.30973535776138306,
+      "learning_rate": 0.0009757039587676229,
+      "loss": 1.8927,
+      "step": 829
+    },
+    {
+      "epoch": 0.38069028781103087,
+      "grad_norm": 0.36675432324409485,
+      "learning_rate": 0.0009756276430311501,
+      "loss": 1.9817,
+      "step": 830
+    },
+    {
+      "epoch": 0.3811489508083935,
+      "grad_norm": 0.5217725038528442,
+      "learning_rate": 0.000975551210619787,
+      "loss": 1.6754,
+      "step": 831
+    },
+    {
+      "epoch": 0.3816076138057562,
+      "grad_norm": 0.22404442727565765,
+      "learning_rate": 0.0009754746615522832,
+      "loss": 1.0305,
+      "step": 832
+    },
+    {
+      "epoch": 0.3820662768031189,
+      "grad_norm": 1.0014019012451172,
+      "learning_rate": 0.0009753979958474164,
+      "loss": 1.6633,
+      "step": 833
+    },
+    {
+      "epoch": 0.3825249398004816,
+      "grad_norm": 6.931515216827393,
+      "learning_rate": 0.0009753212135239935,
+      "loss": 2.2435,
+      "step": 834
+    },
+    {
+      "epoch": 0.38298360279784427,
+      "grad_norm": 4.728778839111328,
+      "learning_rate": 0.0009752443146008496,
+      "loss": 2.3665,
+      "step": 835
+    },
+    {
+      "epoch": 0.38344226579520696,
+      "grad_norm": 0.5018776655197144,
+      "learning_rate": 0.0009751672990968486,
+      "loss": 1.4681,
+      "step": 836
+    },
+    {
+      "epoch": 0.38390092879256965,
+      "grad_norm": 1.0198359489440918,
+      "learning_rate": 0.0009750901670308831,
+      "loss": 2.5583,
+      "step": 837
+    },
+    {
+      "epoch": 0.38435959178993234,
+      "grad_norm": 0.3947281241416931,
+      "learning_rate": 0.0009750129184218741,
+      "loss": 1.788,
+      "step": 838
+    },
+    {
+      "epoch": 0.38481825478729503,
+      "grad_norm": 0.4941509962081909,
+      "learning_rate": 0.0009749355532887712,
+      "loss": 1.4786,
+      "step": 839
+    },
+    {
+      "epoch": 0.3852769177846577,
+      "grad_norm": 0.20619411766529083,
+      "learning_rate": 0.0009748580716505523,
+      "loss": 0.8842,
+      "step": 840
+    },
+    {
+      "epoch": 0.3857355807820204,
+      "grad_norm": 0.4519241750240326,
+      "learning_rate": 0.0009747804735262249,
+      "loss": 1.1354,
+      "step": 841
+    },
+    {
+      "epoch": 0.3861942437793831,
+      "grad_norm": 0.5304630398750305,
+      "learning_rate": 0.0009747027589348239,
+      "loss": 2.087,
+      "step": 842
+    },
+    {
+      "epoch": 0.3866529067767458,
+      "grad_norm": 0.4589567184448242,
+      "learning_rate": 0.0009746249278954134,
+      "loss": 1.955,
+      "step": 843
+    },
+    {
+      "epoch": 0.3871115697741085,
+      "grad_norm": 0.4906143248081207,
+      "learning_rate": 0.0009745469804270857,
+      "loss": 2.0344,
+      "step": 844
+    },
+    {
+      "epoch": 0.3875702327714712,
+      "grad_norm": 0.330793172121048,
+      "learning_rate": 0.0009744689165489621,
+      "loss": 1.4565,
+      "step": 845
+    },
+    {
+      "epoch": 0.38802889576883387,
+      "grad_norm": 0.3725048899650574,
+      "learning_rate": 0.0009743907362801923,
+      "loss": 1.9084,
+      "step": 846
+    },
+    {
+      "epoch": 0.38848755876619656,
+      "grad_norm": 0.5816589593887329,
+      "learning_rate": 0.0009743124396399541,
+      "loss": 2.4114,
+      "step": 847
+    },
+    {
+      "epoch": 0.38894622176355925,
+      "grad_norm": 0.2848202586174011,
+      "learning_rate": 0.0009742340266474547,
+      "loss": 1.4275,
+      "step": 848
+    },
+    {
+      "epoch": 0.3894048847609219,
+      "grad_norm": 0.28104984760284424,
+      "learning_rate": 0.0009741554973219291,
+      "loss": 1.0759,
+      "step": 849
+    },
+    {
+      "epoch": 0.3898635477582846,
+      "grad_norm": 0.39782270789146423,
+      "learning_rate": 0.000974076851682641,
+      "loss": 1.6635,
+      "step": 850
+    },
+    {
+      "epoch": 0.39032221075564727,
+      "grad_norm": 0.25713643431663513,
+      "learning_rate": 0.0009739980897488831,
+      "loss": 1.3452,
+      "step": 851
+    },
+    {
+      "epoch": 0.39078087375300996,
+      "grad_norm": 0.12562298774719238,
+      "learning_rate": 0.0009739192115399762,
+      "loss": 0.65,
+      "step": 852
+    },
+    {
+      "epoch": 0.39123953675037265,
+      "grad_norm": 0.22468183934688568,
+      "learning_rate": 0.0009738402170752693,
+      "loss": 1.1257,
+      "step": 853
+    },
+    {
+      "epoch": 0.39169819974773534,
+      "grad_norm": 0.34269779920578003,
+      "learning_rate": 0.0009737611063741407,
+      "loss": 1.5716,
+      "step": 854
+    },
+    {
+      "epoch": 0.39215686274509803,
+      "grad_norm": 0.3019089102745056,
+      "learning_rate": 0.0009736818794559967,
+      "loss": 1.5173,
+      "step": 855
+    },
+    {
+      "epoch": 0.3926155257424607,
+      "grad_norm": 0.19614103436470032,
+      "learning_rate": 0.0009736025363402723,
+      "loss": 0.6453,
+      "step": 856
+    },
+    {
+      "epoch": 0.3930741887398234,
+      "grad_norm": 0.21685221791267395,
+      "learning_rate": 0.000973523077046431,
+      "loss": 1.0789,
+      "step": 857
+    },
+    {
+      "epoch": 0.3935328517371861,
+      "grad_norm": 0.30863749980926514,
+      "learning_rate": 0.0009734435015939644,
+      "loss": 1.4852,
+      "step": 858
+    },
+    {
+      "epoch": 0.3939915147345488,
+      "grad_norm": 0.1502714902162552,
+      "learning_rate": 0.0009733638100023932,
+      "loss": 0.7736,
+      "step": 859
+    },
+    {
+      "epoch": 0.3944501777319115,
+      "grad_norm": 0.3218334913253784,
+      "learning_rate": 0.0009732840022912664,
+      "loss": 1.662,
+      "step": 860
+    },
+    {
+      "epoch": 0.3949088407292742,
+      "grad_norm": 0.23470452427864075,
+      "learning_rate": 0.0009732040784801611,
+      "loss": 1.1917,
+      "step": 861
+    },
+    {
+      "epoch": 0.39536750372663687,
+      "grad_norm": 0.3882734775543213,
+      "learning_rate": 0.0009731240385886835,
+      "loss": 1.7597,
+      "step": 862
+    },
+    {
+      "epoch": 0.39582616672399956,
+      "grad_norm": 0.41674375534057617,
+      "learning_rate": 0.0009730438826364679,
+      "loss": 1.887,
+      "step": 863
+    },
+    {
+      "epoch": 0.39628482972136225,
+      "grad_norm": 0.3275260329246521,
+      "learning_rate": 0.0009729636106431768,
+      "loss": 1.5059,
+      "step": 864
+    },
+    {
+      "epoch": 0.39674349271872494,
+      "grad_norm": 0.262455016374588,
+      "learning_rate": 0.000972883222628502,
+      "loss": 1.4503,
+      "step": 865
+    },
+    {
+      "epoch": 0.3972021557160876,
+      "grad_norm": 0.3371032774448395,
+      "learning_rate": 0.0009728027186121629,
+      "loss": 1.7706,
+      "step": 866
+    },
+    {
+      "epoch": 0.39766081871345027,
+      "grad_norm": 0.4705641269683838,
+      "learning_rate": 0.0009727220986139079,
+      "loss": 2.3818,
+      "step": 867
+    },
+    {
+      "epoch": 0.39811948171081296,
+      "grad_norm": 0.37057843804359436,
+      "learning_rate": 0.0009726413626535136,
+      "loss": 1.912,
+      "step": 868
+    },
+    {
+      "epoch": 0.39857814470817565,
+      "grad_norm": 0.19416803121566772,
+      "learning_rate": 0.0009725605107507851,
+      "loss": 0.9414,
+      "step": 869
+    },
+    {
+      "epoch": 0.39903680770553834,
+      "grad_norm": 0.26515352725982666,
+      "learning_rate": 0.0009724795429255559,
+      "loss": 1.5286,
+      "step": 870
+    },
+    {
+      "epoch": 0.39949547070290103,
+      "grad_norm": 0.33706986904144287,
+      "learning_rate": 0.0009723984591976882,
+      "loss": 1.4502,
+      "step": 871
+    },
+    {
+      "epoch": 0.3999541337002637,
+      "grad_norm": 0.0702093318104744,
+      "learning_rate": 0.0009723172595870724,
+      "loss": 0.4485,
+      "step": 872
+    },
+    {
+      "epoch": 0.4004127966976264,
+      "grad_norm": 0.376376211643219,
+      "learning_rate": 0.000972235944113627,
+      "loss": 2.1362,
+      "step": 873
+    },
+    {
+      "epoch": 0.4008714596949891,
+      "grad_norm": 0.3718342185020447,
+      "learning_rate": 0.0009721545127972998,
+      "loss": 2.3489,
+      "step": 874
+    },
+    {
+      "epoch": 0.4013301226923518,
+      "grad_norm": 0.41915109753608704,
+      "learning_rate": 0.0009720729656580658,
+      "loss": 1.9443,
+      "step": 875
+    },
+    {
+      "epoch": 0.4017887856897145,
+      "grad_norm": 0.2805352210998535,
+      "learning_rate": 0.0009719913027159298,
+      "loss": 1.397,
+      "step": 876
+    },
+    {
+      "epoch": 0.4022474486870772,
+      "grad_norm": 0.23363643884658813,
+      "learning_rate": 0.000971909523990924,
+      "loss": 0.9596,
+      "step": 877
+    },
+    {
+      "epoch": 0.40270611168443987,
+      "grad_norm": 0.41541317105293274,
+      "learning_rate": 0.0009718276295031091,
+      "loss": 1.8851,
+      "step": 878
+    },
+    {
+      "epoch": 0.40316477468180256,
+      "grad_norm": 0.37006086111068726,
+      "learning_rate": 0.0009717456192725747,
+      "loss": 1.9034,
+      "step": 879
+    },
+    {
+      "epoch": 0.40362343767916525,
+      "grad_norm": 0.334375262260437,
+      "learning_rate": 0.0009716634933194386,
+      "loss": 1.5997,
+      "step": 880
+    },
+    {
+      "epoch": 0.40408210067652794,
+      "grad_norm": 0.3790551424026489,
+      "learning_rate": 0.0009715812516638466,
+      "loss": 1.7657,
+      "step": 881
+    },
+    {
+      "epoch": 0.40454076367389064,
+      "grad_norm": 0.3914640247821808,
+      "learning_rate": 0.0009714988943259733,
+      "loss": 1.7908,
+      "step": 882
+    },
+    {
+      "epoch": 0.40499942667125327,
+      "grad_norm": 0.2521947920322418,
+      "learning_rate": 0.0009714164213260215,
+      "loss": 1.1801,
+      "step": 883
+    },
+    {
+      "epoch": 0.40545808966861596,
+      "grad_norm": 0.3940063714981079,
+      "learning_rate": 0.0009713338326842225,
+      "loss": 1.5258,
+      "step": 884
+    },
+    {
+      "epoch": 0.40591675266597865,
+      "grad_norm": 0.41144511103630066,
+      "learning_rate": 0.0009712511284208358,
+      "loss": 2.3174,
+      "step": 885
+    },
+    {
+      "epoch": 0.40637541566334134,
+      "grad_norm": 0.31135112047195435,
+      "learning_rate": 0.0009711683085561496,
+      "loss": 1.8175,
+      "step": 886
+    },
+    {
+      "epoch": 0.40683407866070403,
+      "grad_norm": 0.27680450677871704,
+      "learning_rate": 0.0009710853731104798,
+      "loss": 1.4197,
+      "step": 887
+    },
+    {
+      "epoch": 0.4072927416580667,
+      "grad_norm": 0.3753867745399475,
+      "learning_rate": 0.0009710023221041712,
+      "loss": 1.8083,
+      "step": 888
+    },
+    {
+      "epoch": 0.4077514046554294,
+      "grad_norm": 0.39393678307533264,
+      "learning_rate": 0.0009709191555575972,
+      "loss": 1.7822,
+      "step": 889
+    },
+    {
+      "epoch": 0.4082100676527921,
+      "grad_norm": 0.24905888736248016,
+      "learning_rate": 0.0009708358734911586,
+      "loss": 1.3235,
+      "step": 890
+    },
+    {
+      "epoch": 0.4086687306501548,
+      "grad_norm": 0.2505285143852234,
+      "learning_rate": 0.0009707524759252855,
+      "loss": 1.2241,
+      "step": 891
+    },
+    {
+      "epoch": 0.4091273936475175,
+      "grad_norm": 0.39112672209739685,
+      "learning_rate": 0.0009706689628804355,
+      "loss": 1.9813,
+      "step": 892
+    },
+    {
+      "epoch": 0.4095860566448802,
+      "grad_norm": 0.3590749502182007,
+      "learning_rate": 0.0009705853343770954,
+      "loss": 1.4428,
+      "step": 893
+    },
+    {
+      "epoch": 0.41004471964224287,
+      "grad_norm": 0.4543074369430542,
+      "learning_rate": 0.0009705015904357797,
+      "loss": 2.1425,
+      "step": 894
+    },
+    {
+      "epoch": 0.41050338263960556,
+      "grad_norm": 0.5381583571434021,
+      "learning_rate": 0.0009704177310770313,
+      "loss": 1.4238,
+      "step": 895
+    },
+    {
+      "epoch": 0.41096204563696825,
+      "grad_norm": 0.5682449340820312,
+      "learning_rate": 0.0009703337563214216,
+      "loss": 0.8868,
+      "step": 896
+    },
+    {
+      "epoch": 0.41142070863433094,
+      "grad_norm": 0.8263031244277954,
+      "learning_rate": 0.0009702496661895501,
+      "loss": 2.1425,
+      "step": 897
+    },
+    {
+      "epoch": 0.41187937163169364,
+      "grad_norm": 0.07518908381462097,
+      "learning_rate": 0.0009701654607020449,
+      "loss": 0.4335,
+      "step": 898
+    },
+    {
+      "epoch": 0.4123380346290563,
+      "grad_norm": 0.43035972118377686,
+      "learning_rate": 0.0009700811398795622,
+      "loss": 1.9609,
+      "step": 899
+    },
+    {
+      "epoch": 0.41279669762641896,
+      "grad_norm": 0.33483147621154785,
+      "learning_rate": 0.0009699967037427864,
+      "loss": 1.614,
+      "step": 900
+    },
+    {
+      "epoch": 0.41325536062378165,
+      "grad_norm": 0.20277926325798035,
+      "learning_rate": 0.0009699121523124301,
+      "loss": 0.9445,
+      "step": 901
+    },
+    {
+      "epoch": 0.41371402362114434,
+      "grad_norm": 0.20781390368938446,
+      "learning_rate": 0.0009698274856092348,
+      "loss": 1.0036,
+      "step": 902
+    },
+    {
+      "epoch": 0.41417268661850704,
+      "grad_norm": 0.3994620740413666,
+      "learning_rate": 0.0009697427036539696,
+      "loss": 2.8057,
+      "step": 903
+    },
+    {
+      "epoch": 0.4146313496158697,
+      "grad_norm": 0.20834676921367645,
+      "learning_rate": 0.0009696578064674322,
+      "loss": 1.2151,
+      "step": 904
+    },
+    {
+      "epoch": 0.4150900126132324,
+      "grad_norm": 0.21917149424552917,
+      "learning_rate": 0.0009695727940704484,
+      "loss": 1.2262,
+      "step": 905
+    },
+    {
+      "epoch": 0.4155486756105951,
+      "grad_norm": 0.21042834222316742,
+      "learning_rate": 0.0009694876664838725,
+      "loss": 1.0464,
+      "step": 906
+    },
+    {
+      "epoch": 0.4160073386079578,
+      "grad_norm": 0.2943893074989319,
+      "learning_rate": 0.0009694024237285868,
+      "loss": 1.5404,
+      "step": 907
+    },
+    {
+      "epoch": 0.4164660016053205,
+      "grad_norm": 0.24305228888988495,
+      "learning_rate": 0.000969317065825502,
+      "loss": 1.4753,
+      "step": 908
+    },
+    {
+      "epoch": 0.4169246646026832,
+      "grad_norm": 0.29043078422546387,
+      "learning_rate": 0.000969231592795557,
+      "loss": 1.8857,
+      "step": 909
+    },
+    {
+      "epoch": 0.4173833276000459,
+      "grad_norm": 0.2063802182674408,
+      "learning_rate": 0.000969146004659719,
+      "loss": 1.133,
+      "step": 910
+    },
+    {
+      "epoch": 0.41784199059740856,
+      "grad_norm": 0.38645628094673157,
+      "learning_rate": 0.0009690603014389831,
+      "loss": 2.2798,
+      "step": 911
+    },
+    {
+      "epoch": 0.41830065359477125,
+      "grad_norm": 0.2220807522535324,
+      "learning_rate": 0.0009689744831543734,
+      "loss": 0.8575,
+      "step": 912
+    },
+    {
+      "epoch": 0.41875931659213395,
+      "grad_norm": 0.4347355365753174,
+      "learning_rate": 0.0009688885498269416,
+      "loss": 2.3926,
+      "step": 913
+    },
+    {
+      "epoch": 0.41921797958949664,
+      "grad_norm": 0.45031097531318665,
+      "learning_rate": 0.0009688025014777672,
+      "loss": 2.4253,
+      "step": 914
+    },
+    {
+      "epoch": 0.4196766425868593,
+      "grad_norm": 0.19847296178340912,
+      "learning_rate": 0.0009687163381279593,
+      "loss": 0.8247,
+      "step": 915
+    },
+    {
+      "epoch": 0.420135305584222,
+      "grad_norm": 0.3677137792110443,
+      "learning_rate": 0.0009686300597986541,
+      "loss": 1.9451,
+      "step": 916
+    },
+    {
+      "epoch": 0.42059396858158465,
+      "grad_norm": 0.2694656550884247,
+      "learning_rate": 0.0009685436665110161,
+      "loss": 1.5455,
+      "step": 917
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 0.21316292881965637,
+      "learning_rate": 0.0009684571582862382,
+      "loss": 1.0842,
+      "step": 918
+    },
+    {
+      "epoch": 0.42151129457631004,
+      "grad_norm": 0.32941651344299316,
+      "learning_rate": 0.0009683705351455419,
+      "loss": 1.709,
+      "step": 919
+    },
+    {
+      "epoch": 0.4219699575736727,
+      "grad_norm": 0.3653143644332886,
+      "learning_rate": 0.0009682837971101762,
+      "loss": 1.8724,
+      "step": 920
+    },
+    {
+      "epoch": 0.4224286205710354,
+      "grad_norm": 0.37952494621276855,
+      "learning_rate": 0.0009681969442014183,
+      "loss": 1.7211,
+      "step": 921
+    },
+    {
+      "epoch": 0.4228872835683981,
+      "grad_norm": 0.3981945216655731,
+      "learning_rate": 0.0009681099764405743,
+      "loss": 2.2534,
+      "step": 922
+    },
+    {
+      "epoch": 0.4233459465657608,
+      "grad_norm": 0.0702991932630539,
+      "learning_rate": 0.0009680228938489777,
+      "loss": 0.4506,
+      "step": 923
+    },
+    {
+      "epoch": 0.4238046095631235,
+      "grad_norm": 0.2196217030286789,
+      "learning_rate": 0.0009679356964479908,
+      "loss": 1.088,
+      "step": 924
+    },
+    {
+      "epoch": 0.4242632725604862,
+      "grad_norm": 0.2972463071346283,
+      "learning_rate": 0.0009678483842590034,
+      "loss": 1.4828,
+      "step": 925
+    },
+    {
+      "epoch": 0.4247219355578489,
+      "grad_norm": 0.10213679820299149,
+      "learning_rate": 0.000967760957303434,
+      "loss": 0.628,
+      "step": 926
+    },
+    {
+      "epoch": 0.42518059855521156,
+      "grad_norm": 0.294698029756546,
+      "learning_rate": 0.0009676734156027292,
+      "loss": 1.504,
+      "step": 927
+    },
+    {
+      "epoch": 0.42563926155257426,
+      "grad_norm": 0.4071488380432129,
+      "learning_rate": 0.0009675857591783634,
+      "loss": 2.2822,
+      "step": 928
+    },
+    {
+      "epoch": 0.42609792454993695,
+      "grad_norm": 0.16480673849582672,
+      "learning_rate": 0.0009674979880518393,
+      "loss": 0.8937,
+      "step": 929
+    },
+    {
+      "epoch": 0.42655658754729964,
+      "grad_norm": 0.2734246253967285,
+      "learning_rate": 0.0009674101022446879,
+      "loss": 1.451,
+      "step": 930
+    },
+    {
+      "epoch": 0.42701525054466233,
+      "grad_norm": 0.3893333971500397,
+      "learning_rate": 0.0009673221017784683,
+      "loss": 1.9644,
+      "step": 931
+    },
+    {
+      "epoch": 0.427473913542025,
+      "grad_norm": 0.2807949483394623,
+      "learning_rate": 0.0009672339866747675,
+      "loss": 1.5402,
+      "step": 932
+    },
+    {
+      "epoch": 0.4279325765393877,
+      "grad_norm": 0.37752094864845276,
+      "learning_rate": 0.0009671457569552009,
+      "loss": 2.1098,
+      "step": 933
+    },
+    {
+      "epoch": 0.42839123953675035,
+      "grad_norm": 0.29235443472862244,
+      "learning_rate": 0.0009670574126414118,
+      "loss": 1.3341,
+      "step": 934
+    },
+    {
+      "epoch": 0.42884990253411304,
+      "grad_norm": 0.3632213771343231,
+      "learning_rate": 0.0009669689537550717,
+      "loss": 1.5638,
+      "step": 935
+    },
+    {
+      "epoch": 0.4293085655314757,
+      "grad_norm": 0.30011284351348877,
+      "learning_rate": 0.0009668803803178803,
+      "loss": 1.6048,
+      "step": 936
+    },
+    {
+      "epoch": 0.4297672285288384,
+      "grad_norm": 0.21687191724777222,
+      "learning_rate": 0.0009667916923515651,
+      "loss": 0.9312,
+      "step": 937
+    },
+    {
+      "epoch": 0.4302258915262011,
+      "grad_norm": 0.2912392020225525,
+      "learning_rate": 0.0009667028898778822,
+      "loss": 1.5125,
+      "step": 938
+    },
+    {
+      "epoch": 0.4306845545235638,
+      "grad_norm": 0.3935389816761017,
+      "learning_rate": 0.0009666139729186152,
+      "loss": 2.1857,
+      "step": 939
+    },
+    {
+      "epoch": 0.4311432175209265,
+      "grad_norm": 0.4185419976711273,
+      "learning_rate": 0.0009665249414955761,
+      "loss": 2.4136,
+      "step": 940
+    },
+    {
+      "epoch": 0.4316018805182892,
+      "grad_norm": 0.10364656150341034,
+      "learning_rate": 0.0009664357956306051,
+      "loss": 0.6298,
+      "step": 941
+    },
+    {
+      "epoch": 0.4320605435156519,
+      "grad_norm": 0.2796267867088318,
+      "learning_rate": 0.0009663465353455703,
+      "loss": 1.5578,
+      "step": 942
+    },
+    {
+      "epoch": 0.43251920651301456,
+      "grad_norm": 0.2670913338661194,
+      "learning_rate": 0.0009662571606623678,
+      "loss": 1.1976,
+      "step": 943
+    },
+    {
+      "epoch": 0.43297786951037726,
+      "grad_norm": 0.41438791155815125,
+      "learning_rate": 0.0009661676716029219,
+      "loss": 2.4133,
+      "step": 944
+    },
+    {
+      "epoch": 0.43343653250773995,
+      "grad_norm": 0.31299319863319397,
+      "learning_rate": 0.000966078068189185,
+      "loss": 1.5249,
+      "step": 945
+    },
+    {
+      "epoch": 0.43389519550510264,
+      "grad_norm": 0.3784157931804657,
+      "learning_rate": 0.0009659883504431373,
+      "loss": 1.8156,
+      "step": 946
+    },
+    {
+      "epoch": 0.43435385850246533,
+      "grad_norm": 0.3633187413215637,
+      "learning_rate": 0.0009658985183867872,
+      "loss": 2.0233,
+      "step": 947
+    },
+    {
+      "epoch": 0.434812521499828,
+      "grad_norm": 0.29936981201171875,
+      "learning_rate": 0.0009658085720421714,
+      "loss": 1.4108,
+      "step": 948
+    },
+    {
+      "epoch": 0.4352711844971907,
+      "grad_norm": 0.24031949043273926,
+      "learning_rate": 0.0009657185114313541,
+      "loss": 0.9012,
+      "step": 949
+    },
+    {
+      "epoch": 0.4357298474945534,
+      "grad_norm": 0.30812880396842957,
+      "learning_rate": 0.000965628336576428,
+      "loss": 1.739,
+      "step": 950
+    },
+    {
+      "epoch": 0.43618851049191604,
+      "grad_norm": 0.27589166164398193,
+      "learning_rate": 0.0009655380474995137,
+      "loss": 1.387,
+      "step": 951
+    },
+    {
+      "epoch": 0.43664717348927873,
+      "grad_norm": 0.3418666422367096,
+      "learning_rate": 0.0009654476442227595,
+      "loss": 1.8832,
+      "step": 952
+    },
+    {
+      "epoch": 0.4371058364866414,
+      "grad_norm": 0.18190939724445343,
+      "learning_rate": 0.000965357126768342,
+      "loss": 0.8432,
+      "step": 953
+    },
+    {
+      "epoch": 0.4375644994840041,
+      "grad_norm": 0.336883544921875,
+      "learning_rate": 0.0009652664951584662,
+      "loss": 1.9048,
+      "step": 954
+    },
+    {
+      "epoch": 0.4380231624813668,
+      "grad_norm": 0.2961789667606354,
+      "learning_rate": 0.0009651757494153642,
+      "loss": 1.473,
+      "step": 955
+    },
+    {
+      "epoch": 0.4384818254787295,
+      "grad_norm": 0.2900097370147705,
+      "learning_rate": 0.0009650848895612969,
+      "loss": 1.2979,
+      "step": 956
+    },
+    {
+      "epoch": 0.4389404884760922,
+      "grad_norm": 0.13419854640960693,
+      "learning_rate": 0.0009649939156185526,
+      "loss": 0.8149,
+      "step": 957
+    },
+    {
+      "epoch": 0.4393991514734549,
+      "grad_norm": 0.2546023428440094,
+      "learning_rate": 0.000964902827609448,
+      "loss": 1.2925,
+      "step": 958
+    },
+    {
+      "epoch": 0.43985781447081757,
+      "grad_norm": 0.28633105754852295,
+      "learning_rate": 0.0009648116255563279,
+      "loss": 1.4249,
+      "step": 959
+    },
+    {
+      "epoch": 0.44031647746818026,
+      "grad_norm": 0.34882158041000366,
+      "learning_rate": 0.0009647203094815644,
+      "loss": 1.8254,
+      "step": 960
+    },
+    {
+      "epoch": 0.44077514046554295,
+      "grad_norm": 0.2623085677623749,
+      "learning_rate": 0.0009646288794075582,
+      "loss": 1.2006,
+      "step": 961
+    },
+    {
+      "epoch": 0.44123380346290564,
+      "grad_norm": 0.3005825877189636,
+      "learning_rate": 0.0009645373353567377,
+      "loss": 1.4745,
+      "step": 962
+    },
+    {
+      "epoch": 0.44169246646026833,
+      "grad_norm": 0.4481685757637024,
+      "learning_rate": 0.0009644456773515595,
+      "loss": 1.6587,
+      "step": 963
+    },
+    {
+      "epoch": 0.442151129457631,
+      "grad_norm": 0.34620991349220276,
+      "learning_rate": 0.0009643539054145077,
+      "loss": 2.0671,
+      "step": 964
+    },
+    {
+      "epoch": 0.4426097924549937,
+      "grad_norm": 0.10259232670068741,
+      "learning_rate": 0.0009642620195680948,
+      "loss": 0.6523,
+      "step": 965
+    },
+    {
+      "epoch": 0.4430684554523564,
+      "grad_norm": 0.2266424596309662,
+      "learning_rate": 0.0009641700198348608,
+      "loss": 1.033,
+      "step": 966
+    },
+    {
+      "epoch": 0.4435271184497191,
+      "grad_norm": 0.2258816808462143,
+      "learning_rate": 0.0009640779062373743,
+      "loss": 1.3325,
+      "step": 967
+    },
+    {
+      "epoch": 0.4439857814470818,
+      "grad_norm": 0.3065774142742157,
+      "learning_rate": 0.0009639856787982313,
+      "loss": 1.7267,
+      "step": 968
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 0.2785046696662903,
+      "learning_rate": 0.0009638933375400555,
+      "loss": 1.4358,
+      "step": 969
+    },
+    {
+      "epoch": 0.4449031074418071,
+      "grad_norm": 0.40370169281959534,
+      "learning_rate": 0.0009638008824854995,
+      "loss": 2.2593,
+      "step": 970
+    },
+    {
+      "epoch": 0.4453617704391698,
+      "grad_norm": 0.2614261507987976,
+      "learning_rate": 0.0009637083136572426,
+      "loss": 1.4521,
+      "step": 971
+    },
+    {
+      "epoch": 0.4458204334365325,
+      "grad_norm": 0.3734627068042755,
+      "learning_rate": 0.0009636156310779928,
+      "loss": 2.2771,
+      "step": 972
+    },
+    {
+      "epoch": 0.4462790964338952,
+      "grad_norm": 0.3273562490940094,
+      "learning_rate": 0.0009635228347704861,
+      "loss": 1.7556,
+      "step": 973
+    },
+    {
+      "epoch": 0.4467377594312579,
+      "grad_norm": 0.28261798620224,
+      "learning_rate": 0.0009634299247574858,
+      "loss": 1.409,
+      "step": 974
+    },
+    {
+      "epoch": 0.44719642242862057,
+      "grad_norm": 0.409447580575943,
+      "learning_rate": 0.0009633369010617834,
+      "loss": 2.3015,
+      "step": 975
+    },
+    {
+      "epoch": 0.44765508542598326,
+      "grad_norm": 0.44366922974586487,
+      "learning_rate": 0.0009632437637061982,
+      "loss": 2.4729,
+      "step": 976
+    },
+    {
+      "epoch": 0.44811374842334595,
+      "grad_norm": 0.35576725006103516,
+      "learning_rate": 0.0009631505127135778,
+      "loss": 2.0571,
+      "step": 977
+    },
+    {
+      "epoch": 0.44857241142070864,
+      "grad_norm": 0.2606424391269684,
+      "learning_rate": 0.000963057148106797,
+      "loss": 1.4125,
+      "step": 978
+    },
+    {
+      "epoch": 0.44903107441807133,
+      "grad_norm": 0.3204779028892517,
+      "learning_rate": 0.0009629636699087591,
+      "loss": 1.6771,
+      "step": 979
+    },
+    {
+      "epoch": 0.449489737415434,
+      "grad_norm": 0.3472746014595032,
+      "learning_rate": 0.0009628700781423947,
+      "loss": 1.7863,
+      "step": 980
+    },
+    {
+      "epoch": 0.4499484004127967,
+      "grad_norm": 0.2507071793079376,
+      "learning_rate": 0.0009627763728306626,
+      "loss": 1.2164,
+      "step": 981
+    },
+    {
+      "epoch": 0.4504070634101594,
+      "grad_norm": 0.284404993057251,
+      "learning_rate": 0.0009626825539965497,
+      "loss": 1.3602,
+      "step": 982
+    },
+    {
+      "epoch": 0.4508657264075221,
+      "grad_norm": 0.23801104724407196,
+      "learning_rate": 0.00096258862166307,
+      "loss": 1.1913,
+      "step": 983
+    },
+    {
+      "epoch": 0.4513243894048848,
+      "grad_norm": 0.08279567956924438,
+      "learning_rate": 0.0009624945758532662,
+      "loss": 0.5228,
+      "step": 984
+    },
+    {
+      "epoch": 0.4517830524022475,
+      "grad_norm": 0.31264549493789673,
+      "learning_rate": 0.0009624004165902078,
+      "loss": 1.3771,
+      "step": 985
+    },
+    {
+      "epoch": 0.4522417153996101,
+      "grad_norm": 0.290018767118454,
+      "learning_rate": 0.0009623061438969934,
+      "loss": 1.4374,
+      "step": 986
+    },
+    {
+      "epoch": 0.4527003783969728,
+      "grad_norm": 0.30359646677970886,
+      "learning_rate": 0.0009622117577967486,
+      "loss": 1.7618,
+      "step": 987
+    },
+    {
+      "epoch": 0.4531590413943355,
+      "grad_norm": 0.3439119756221771,
+      "learning_rate": 0.0009621172583126267,
+      "loss": 1.7233,
+      "step": 988
+    },
+    {
+      "epoch": 0.4536177043916982,
+      "grad_norm": 0.3586854636669159,
+      "learning_rate": 0.0009620226454678093,
+      "loss": 2.0002,
+      "step": 989
+    },
+    {
+      "epoch": 0.4540763673890609,
+      "grad_norm": 0.3362513780593872,
+      "learning_rate": 0.0009619279192855056,
+      "loss": 2.1243,
+      "step": 990
+    },
+    {
+      "epoch": 0.45453503038642357,
+      "grad_norm": 0.2629103362560272,
+      "learning_rate": 0.0009618330797889527,
+      "loss": 1.3026,
+      "step": 991
+    },
+    {
+      "epoch": 0.45499369338378626,
+      "grad_norm": 0.3186989426612854,
+      "learning_rate": 0.0009617381270014154,
+      "loss": 1.8407,
+      "step": 992
+    },
+    {
+      "epoch": 0.45545235638114895,
+      "grad_norm": 0.40748247504234314,
+      "learning_rate": 0.0009616430609461861,
+      "loss": 1.9366,
+      "step": 993
+    },
+    {
+      "epoch": 0.45591101937851164,
+      "grad_norm": 0.37845298647880554,
+      "learning_rate": 0.0009615478816465854,
+      "loss": 1.8237,
+      "step": 994
+    },
+    {
+      "epoch": 0.45636968237587433,
+      "grad_norm": 0.3755371570587158,
+      "learning_rate": 0.0009614525891259613,
+      "loss": 1.8331,
+      "step": 995
+    },
+    {
+      "epoch": 0.456828345373237,
+      "grad_norm": 0.28780463337898254,
+      "learning_rate": 0.00096135718340769,
+      "loss": 1.2989,
+      "step": 996
+    },
+    {
+      "epoch": 0.4572870083705997,
+      "grad_norm": 0.3778688609600067,
+      "learning_rate": 0.000961261664515175,
+      "loss": 1.7416,
+      "step": 997
+    },
+    {
+      "epoch": 0.4577456713679624,
+      "grad_norm": 0.2470480501651764,
+      "learning_rate": 0.0009611660324718479,
+      "loss": 0.9676,
+      "step": 998
+    },
+    {
+      "epoch": 0.4582043343653251,
+      "grad_norm": 0.5491411685943604,
+      "learning_rate": 0.0009610702873011676,
+      "loss": 1.8815,
+      "step": 999
+    },
+    {
+      "epoch": 0.4586629973626878,
+      "grad_norm": 0.38404297828674316,
+      "learning_rate": 0.0009609744290266216,
+      "loss": 1.8572,
+      "step": 1000
+    },
+    {
+      "epoch": 0.4591216603600505,
+      "grad_norm": 0.4080640971660614,
+      "learning_rate": 0.0009608784576717242,
+      "loss": 2.2996,
+      "step": 1001
+    },
+    {
+      "epoch": 0.45958032335741317,
+      "grad_norm": 0.39164450764656067,
+      "learning_rate": 0.000960782373260018,
+      "loss": 1.6304,
+      "step": 1002
+    },
+    {
+      "epoch": 0.4600389863547758,
+      "grad_norm": 0.2890069782733917,
+      "learning_rate": 0.0009606861758150733,
+      "loss": 1.2357,
+      "step": 1003
+    },
+    {
+      "epoch": 0.4604976493521385,
+      "grad_norm": 0.08920864015817642,
+      "learning_rate": 0.0009605898653604881,
+      "loss": 0.5615,
+      "step": 1004
+    },
+    {
+      "epoch": 0.4609563123495012,
+      "grad_norm": 0.13560301065444946,
+      "learning_rate": 0.0009604934419198877,
+      "loss": 0.7203,
+      "step": 1005
+    },
+    {
+      "epoch": 0.4614149753468639,
+      "grad_norm": 0.27159520983695984,
+      "learning_rate": 0.0009603969055169258,
+      "loss": 1.6746,
+      "step": 1006
+    },
+    {
+      "epoch": 0.46187363834422657,
+      "grad_norm": 0.19911514222621918,
+      "learning_rate": 0.0009603002561752832,
+      "loss": 1.0421,
+      "step": 1007
+    },
+    {
+      "epoch": 0.46233230134158926,
+      "grad_norm": 0.3992009162902832,
+      "learning_rate": 0.0009602034939186691,
+      "loss": 2.3098,
+      "step": 1008
+    },
+    {
+      "epoch": 0.46279096433895195,
+      "grad_norm": 0.28715813159942627,
+      "learning_rate": 0.0009601066187708194,
+      "loss": 1.3643,
+      "step": 1009
+    },
+    {
+      "epoch": 0.46324962733631464,
+      "grad_norm": 0.3220798671245575,
+      "learning_rate": 0.0009600096307554987,
+      "loss": 1.6344,
+      "step": 1010
+    },
+    {
+      "epoch": 0.46370829033367733,
+      "grad_norm": 0.3547840416431427,
+      "learning_rate": 0.0009599125298964987,
+      "loss": 1.7168,
+      "step": 1011
+    },
+    {
+      "epoch": 0.46416695333104,
+      "grad_norm": 0.29928240180015564,
+      "learning_rate": 0.000959815316217639,
+      "loss": 1.4865,
+      "step": 1012
+    },
+    {
+      "epoch": 0.4646256163284027,
+      "grad_norm": 0.32575827836990356,
+      "learning_rate": 0.0009597179897427668,
+      "loss": 1.4995,
+      "step": 1013
+    },
+    {
+      "epoch": 0.4650842793257654,
+      "grad_norm": 0.36625438928604126,
+      "learning_rate": 0.000959620550495757,
+      "loss": 1.8375,
+      "step": 1014
+    },
+    {
+      "epoch": 0.4655429423231281,
+      "grad_norm": 0.3549942970275879,
+      "learning_rate": 0.0009595229985005121,
+      "loss": 1.7886,
+      "step": 1015
+    },
+    {
+      "epoch": 0.4660016053204908,
+      "grad_norm": 0.33019477128982544,
+      "learning_rate": 0.0009594253337809621,
+      "loss": 1.618,
+      "step": 1016
+    },
+    {
+      "epoch": 0.4664602683178535,
+      "grad_norm": 0.31643784046173096,
+      "learning_rate": 0.0009593275563610655,
+      "loss": 1.5107,
+      "step": 1017
+    },
+    {
+      "epoch": 0.46691893131521617,
+      "grad_norm": 0.33537599444389343,
+      "learning_rate": 0.0009592296662648072,
+      "loss": 1.6339,
+      "step": 1018
+    },
+    {
+      "epoch": 0.46737759431257886,
+      "grad_norm": 0.3272285759449005,
+      "learning_rate": 0.0009591316635162006,
+      "loss": 1.4943,
+      "step": 1019
+    },
+    {
+      "epoch": 0.4678362573099415,
+      "grad_norm": 0.2963853180408478,
+      "learning_rate": 0.0009590335481392863,
+      "loss": 1.5295,
+      "step": 1020
+    },
+    {
+      "epoch": 0.4682949203073042,
+      "grad_norm": 0.38083258271217346,
+      "learning_rate": 0.000958935320158133,
+      "loss": 2.0518,
+      "step": 1021
+    },
+    {
+      "epoch": 0.4687535833046669,
+      "grad_norm": 0.2800465524196625,
+      "learning_rate": 0.0009588369795968366,
+      "loss": 1.386,
+      "step": 1022
+    },
+    {
+      "epoch": 0.46921224630202957,
+      "grad_norm": 0.31138157844543457,
+      "learning_rate": 0.0009587385264795206,
+      "loss": 1.2996,
+      "step": 1023
+    },
+    {
+      "epoch": 0.46967090929939226,
+      "grad_norm": 0.27252325415611267,
+      "learning_rate": 0.0009586399608303364,
+      "loss": 1.4318,
+      "step": 1024
+    },
+    {
+      "epoch": 0.47012957229675495,
+      "grad_norm": 0.2721007764339447,
+      "learning_rate": 0.0009585412826734627,
+      "loss": 1.3777,
+      "step": 1025
+    },
+    {
+      "epoch": 0.47058823529411764,
+      "grad_norm": 0.28045183420181274,
+      "learning_rate": 0.0009584424920331063,
+      "loss": 1.5048,
+      "step": 1026
+    },
+    {
+      "epoch": 0.47104689829148033,
+      "grad_norm": 0.3029325604438782,
+      "learning_rate": 0.000958343588933501,
+      "loss": 1.6938,
+      "step": 1027
+    },
+    {
+      "epoch": 0.471505561288843,
+      "grad_norm": 0.2408834993839264,
+      "learning_rate": 0.0009582445733989086,
+      "loss": 1.4271,
+      "step": 1028
+    },
+    {
+      "epoch": 0.4719642242862057,
+      "grad_norm": 0.10467322915792465,
+      "learning_rate": 0.0009581454454536182,
+      "loss": 0.6104,
+      "step": 1029
+    },
+    {
+      "epoch": 0.4724228872835684,
+      "grad_norm": 0.3287009000778198,
+      "learning_rate": 0.0009580462051219465,
+      "loss": 1.9726,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4728815502809311,
+      "grad_norm": 0.37531885504722595,
+      "learning_rate": 0.0009579468524282381,
+      "loss": 2.3677,
+      "step": 1031
+    },
+    {
+      "epoch": 0.4733402132782938,
+      "grad_norm": 0.4152393937110901,
+      "learning_rate": 0.0009578473873968649,
+      "loss": 2.207,
+      "step": 1032
+    },
+    {
+      "epoch": 0.4737988762756565,
+      "grad_norm": 0.26491448283195496,
+      "learning_rate": 0.0009577478100522261,
+      "loss": 1.406,
+      "step": 1033
+    },
+    {
+      "epoch": 0.47425753927301917,
+      "grad_norm": 0.3497820198535919,
+      "learning_rate": 0.0009576481204187492,
+      "loss": 1.9156,
+      "step": 1034
+    },
+    {
+      "epoch": 0.47471620227038186,
+      "grad_norm": 0.3127298951148987,
+      "learning_rate": 0.0009575483185208884,
+      "loss": 1.3849,
+      "step": 1035
+    },
+    {
+      "epoch": 0.47517486526774455,
+      "grad_norm": 0.386515736579895,
+      "learning_rate": 0.000957448404383126,
+      "loss": 1.9648,
+      "step": 1036
+    },
+    {
+      "epoch": 0.4756335282651072,
+      "grad_norm": 0.30607542395591736,
+      "learning_rate": 0.0009573483780299717,
+      "loss": 1.5166,
+      "step": 1037
+    },
+    {
+      "epoch": 0.4760921912624699,
+      "grad_norm": 0.3027437627315521,
+      "learning_rate": 0.0009572482394859625,
+      "loss": 1.6056,
+      "step": 1038
+    },
+    {
+      "epoch": 0.47655085425983257,
+      "grad_norm": 0.3490167558193207,
+      "learning_rate": 0.0009571479887756633,
+      "loss": 1.7711,
+      "step": 1039
+    },
+    {
+      "epoch": 0.47700951725719526,
+      "grad_norm": 0.3557809293270111,
+      "learning_rate": 0.0009570476259236662,
+      "loss": 1.7281,
+      "step": 1040
+    },
+    {
+      "epoch": 0.47746818025455795,
+      "grad_norm": 0.3299933075904846,
+      "learning_rate": 0.000956947150954591,
+      "loss": 1.3677,
+      "step": 1041
+    },
+    {
+      "epoch": 0.47792684325192064,
+      "grad_norm": 0.4228198528289795,
+      "learning_rate": 0.000956846563893085,
+      "loss": 1.8817,
+      "step": 1042
+    },
+    {
+      "epoch": 0.47838550624928333,
+      "grad_norm": 0.40233686566352844,
+      "learning_rate": 0.0009567458647638228,
+      "loss": 2.2048,
+      "step": 1043
+    },
+    {
+      "epoch": 0.478844169246646,
+      "grad_norm": 0.3554650843143463,
+      "learning_rate": 0.0009566450535915066,
+      "loss": 1.8723,
+      "step": 1044
+    },
+    {
+      "epoch": 0.4793028322440087,
+      "grad_norm": 0.33305272459983826,
+      "learning_rate": 0.0009565441304008661,
+      "loss": 1.67,
+      "step": 1045
+    },
+    {
+      "epoch": 0.4797614952413714,
+      "grad_norm": 0.35247936844825745,
+      "learning_rate": 0.0009564430952166587,
+      "loss": 1.9382,
+      "step": 1046
+    },
+    {
+      "epoch": 0.4802201582387341,
+      "grad_norm": 0.23185671865940094,
+      "learning_rate": 0.0009563419480636689,
+      "loss": 0.9469,
+      "step": 1047
+    },
+    {
+      "epoch": 0.4806788212360968,
+      "grad_norm": 0.29733264446258545,
+      "learning_rate": 0.0009562406889667088,
+      "loss": 1.4415,
+      "step": 1048
+    },
+    {
+      "epoch": 0.4811374842334595,
+      "grad_norm": 0.2600105106830597,
+      "learning_rate": 0.0009561393179506181,
+      "loss": 1.2259,
+      "step": 1049
+    },
+    {
+      "epoch": 0.48159614723082217,
+      "grad_norm": 0.36176133155822754,
+      "learning_rate": 0.0009560378350402637,
+      "loss": 1.6957,
+      "step": 1050
+    },
+    {
+      "epoch": 0.48205481022818486,
+      "grad_norm": 0.29688432812690735,
+      "learning_rate": 0.0009559362402605403,
+      "loss": 1.3275,
+      "step": 1051
+    },
+    {
+      "epoch": 0.48251347322554755,
+      "grad_norm": 0.2176387459039688,
+      "learning_rate": 0.0009558345336363695,
+      "loss": 0.9416,
+      "step": 1052
+    },
+    {
+      "epoch": 0.48297213622291024,
+      "grad_norm": 0.35077396035194397,
+      "learning_rate": 0.0009557327151927009,
+      "loss": 1.6619,
+      "step": 1053
+    },
+    {
+      "epoch": 0.4834307992202729,
+      "grad_norm": 0.35128337144851685,
+      "learning_rate": 0.0009556307849545114,
+      "loss": 1.5695,
+      "step": 1054
+    },
+    {
+      "epoch": 0.48388946221763557,
+      "grad_norm": 0.37428727746009827,
+      "learning_rate": 0.000955528742946805,
+      "loss": 1.9629,
+      "step": 1055
+    },
+    {
+      "epoch": 0.48434812521499826,
+      "grad_norm": 0.28633585572242737,
+      "learning_rate": 0.0009554265891946133,
+      "loss": 1.4113,
+      "step": 1056
+    },
+    {
+      "epoch": 0.48480678821236095,
+      "grad_norm": 0.25998106598854065,
+      "learning_rate": 0.0009553243237229956,
+      "loss": 1.1547,
+      "step": 1057
+    },
+    {
+      "epoch": 0.48526545120972364,
+      "grad_norm": 0.4127240478992462,
+      "learning_rate": 0.0009552219465570382,
+      "loss": 2.1592,
+      "step": 1058
+    },
+    {
+      "epoch": 0.48572411420708633,
+      "grad_norm": 0.23034657537937164,
+      "learning_rate": 0.000955119457721855,
+      "loss": 1.0124,
+      "step": 1059
+    },
+    {
+      "epoch": 0.486182777204449,
+      "grad_norm": 0.4003170430660248,
+      "learning_rate": 0.0009550168572425873,
+      "loss": 1.8563,
+      "step": 1060
+    },
+    {
+      "epoch": 0.4866414402018117,
+      "grad_norm": 0.34863001108169556,
+      "learning_rate": 0.0009549141451444036,
+      "loss": 1.8716,
+      "step": 1061
+    },
+    {
+      "epoch": 0.4871001031991744,
+      "grad_norm": 0.34055593609809875,
+      "learning_rate": 0.0009548113214525,
+      "loss": 1.8115,
+      "step": 1062
+    },
+    {
+      "epoch": 0.4875587661965371,
+      "grad_norm": 0.2745518684387207,
+      "learning_rate": 0.0009547083861921,
+      "loss": 1.3137,
+      "step": 1063
+    },
+    {
+      "epoch": 0.4880174291938998,
+      "grad_norm": 0.19711460173130035,
+      "learning_rate": 0.0009546053393884542,
+      "loss": 0.9982,
+      "step": 1064
+    },
+    {
+      "epoch": 0.4884760921912625,
+      "grad_norm": 0.2679455280303955,
+      "learning_rate": 0.0009545021810668406,
+      "loss": 1.0912,
+      "step": 1065
+    },
+    {
+      "epoch": 0.48893475518862517,
+      "grad_norm": 0.2457709014415741,
+      "learning_rate": 0.000954398911252565,
+      "loss": 0.9668,
+      "step": 1066
+    },
+    {
+      "epoch": 0.48939341818598786,
+      "grad_norm": 0.22733516991138458,
+      "learning_rate": 0.0009542955299709601,
+      "loss": 0.9916,
+      "step": 1067
+    },
+    {
+      "epoch": 0.48985208118335055,
+      "grad_norm": 0.3566371500492096,
+      "learning_rate": 0.000954192037247386,
+      "loss": 1.8339,
+      "step": 1068
+    },
+    {
+      "epoch": 0.49031074418071324,
+      "grad_norm": 0.40455371141433716,
+      "learning_rate": 0.0009540884331072304,
+      "loss": 2.1667,
+      "step": 1069
+    },
+    {
+      "epoch": 0.49076940717807593,
+      "grad_norm": 0.43050408363342285,
+      "learning_rate": 0.0009539847175759077,
+      "loss": 2.3535,
+      "step": 1070
+    },
+    {
+      "epoch": 0.49122807017543857,
+      "grad_norm": 0.3837912380695343,
+      "learning_rate": 0.0009538808906788608,
+      "loss": 1.995,
+      "step": 1071
+    },
+    {
+      "epoch": 0.49168673317280126,
+      "grad_norm": 0.28267940878868103,
+      "learning_rate": 0.0009537769524415585,
+      "loss": 1.4414,
+      "step": 1072
+    },
+    {
+      "epoch": 0.49214539617016395,
+      "grad_norm": 0.43233123421669006,
+      "learning_rate": 0.0009536729028894979,
+      "loss": 2.0576,
+      "step": 1073
+    },
+    {
+      "epoch": 0.49260405916752664,
+      "grad_norm": 0.27310910820961,
+      "learning_rate": 0.0009535687420482031,
+      "loss": 1.0809,
+      "step": 1074
+    },
+    {
+      "epoch": 0.49306272216488933,
+      "grad_norm": 0.2695041596889496,
+      "learning_rate": 0.0009534644699432254,
+      "loss": 1.2148,
+      "step": 1075
+    },
+    {
+      "epoch": 0.493521385162252,
+      "grad_norm": 0.4794595241546631,
+      "learning_rate": 0.0009533600866001437,
+      "loss": 2.3083,
+      "step": 1076
+    },
+    {
+      "epoch": 0.4939800481596147,
+      "grad_norm": 0.2349144071340561,
+      "learning_rate": 0.0009532555920445638,
+      "loss": 0.9781,
+      "step": 1077
+    },
+    {
+      "epoch": 0.4944387111569774,
+      "grad_norm": 0.4289840757846832,
+      "learning_rate": 0.000953150986302119,
+      "loss": 1.9989,
+      "step": 1078
+    },
+    {
+      "epoch": 0.4948973741543401,
+      "grad_norm": 0.4208228588104248,
+      "learning_rate": 0.0009530462693984698,
+      "loss": 2.0288,
+      "step": 1079
+    },
+    {
+      "epoch": 0.4953560371517028,
+      "grad_norm": 0.34878844022750854,
+      "learning_rate": 0.0009529414413593043,
+      "loss": 1.4258,
+      "step": 1080
+    },
+    {
+      "epoch": 0.4958147001490655,
+      "grad_norm": 0.23603618144989014,
+      "learning_rate": 0.0009528365022103371,
+      "loss": 1.1298,
+      "step": 1081
+    },
+    {
+      "epoch": 0.49627336314642817,
+      "grad_norm": 0.27409201860427856,
+      "learning_rate": 0.000952731451977311,
+      "loss": 1.2806,
+      "step": 1082
+    },
+    {
+      "epoch": 0.49673202614379086,
+      "grad_norm": 0.37235966324806213,
+      "learning_rate": 0.0009526262906859953,
+      "loss": 1.9988,
+      "step": 1083
+    },
+    {
+      "epoch": 0.49719068914115355,
+      "grad_norm": 0.3669985830783844,
+      "learning_rate": 0.0009525210183621869,
+      "loss": 1.7487,
+      "step": 1084
+    },
+    {
+      "epoch": 0.49764935213851624,
+      "grad_norm": 0.3774386942386627,
+      "learning_rate": 0.0009524156350317099,
+      "loss": 1.0475,
+      "step": 1085
+    },
+    {
+      "epoch": 0.49810801513587893,
+      "grad_norm": 0.30981799960136414,
+      "learning_rate": 0.0009523101407204154,
+      "loss": 1.3907,
+      "step": 1086
+    },
+    {
+      "epoch": 0.4985666781332416,
+      "grad_norm": 0.3435545861721039,
+      "learning_rate": 0.0009522045354541822,
+      "loss": 1.6863,
+      "step": 1087
+    },
+    {
+      "epoch": 0.49902534113060426,
+      "grad_norm": 0.23086117208003998,
+      "learning_rate": 0.0009520988192589158,
+      "loss": 1.0635,
+      "step": 1088
+    },
+    {
+      "epoch": 0.49948400412796695,
+      "grad_norm": 0.21928969025611877,
+      "learning_rate": 0.0009519929921605493,
+      "loss": 0.9675,
+      "step": 1089
+    },
+    {
+      "epoch": 0.49994266712532964,
+      "grad_norm": 0.2767171263694763,
+      "learning_rate": 0.0009518870541850426,
+      "loss": 1.4333,
+      "step": 1090
+    },
+    {
+      "epoch": 0.5004013301226924,
+      "grad_norm": 0.288322776556015,
+      "learning_rate": 0.0009517810053583835,
+      "loss": 1.6043,
+      "step": 1091
+    },
+    {
+      "epoch": 0.500859993120055,
+      "grad_norm": 0.32782241702079773,
+      "learning_rate": 0.0009516748457065862,
+      "loss": 1.8718,
+      "step": 1092
+    },
+    {
+      "epoch": 0.5013186561174178,
+      "grad_norm": 0.2817911207675934,
+      "learning_rate": 0.0009515685752556924,
+      "loss": 1.5955,
+      "step": 1093
+    },
+    {
+      "epoch": 0.5017773191147804,
+      "grad_norm": 0.3410284221172333,
+      "learning_rate": 0.0009514621940317712,
+      "loss": 1.8585,
+      "step": 1094
+    },
+    {
+      "epoch": 0.5022359821121432,
+      "grad_norm": 0.24062463641166687,
+      "learning_rate": 0.0009513557020609185,
+      "loss": 1.2591,
+      "step": 1095
+    },
+    {
+      "epoch": 0.5026946451095058,
+      "grad_norm": 0.26802435517311096,
+      "learning_rate": 0.0009512490993692578,
+      "loss": 1.3988,
+      "step": 1096
+    },
+    {
+      "epoch": 0.5031533081068684,
+      "grad_norm": 0.3186005651950836,
+      "learning_rate": 0.0009511423859829392,
+      "loss": 1.3655,
+      "step": 1097
+    },
+    {
+      "epoch": 0.5036119711042312,
+      "grad_norm": 0.378814160823822,
+      "learning_rate": 0.0009510355619281403,
+      "loss": 1.9003,
+      "step": 1098
+    },
+    {
+      "epoch": 0.5040706341015938,
+      "grad_norm": 0.12915131449699402,
+      "learning_rate": 0.0009509286272310662,
+      "loss": 0.7532,
+      "step": 1099
+    },
+    {
+      "epoch": 0.5045292970989566,
+      "grad_norm": 0.3437194228172302,
+      "learning_rate": 0.0009508215819179484,
+      "loss": 1.8945,
+      "step": 1100
+    },
+    {
+      "epoch": 0.5049879600963192,
+      "grad_norm": 0.3381500542163849,
+      "learning_rate": 0.0009507144260150459,
+      "loss": 1.8967,
+      "step": 1101
+    },
+    {
+      "epoch": 0.5054466230936819,
+      "grad_norm": 0.31556692719459534,
+      "learning_rate": 0.0009506071595486449,
+      "loss": 1.6499,
+      "step": 1102
+    },
+    {
+      "epoch": 0.5059052860910446,
+      "grad_norm": 0.19522079825401306,
+      "learning_rate": 0.0009504997825450586,
+      "loss": 1.1095,
+      "step": 1103
+    },
+    {
+      "epoch": 0.5063639490884073,
+      "grad_norm": 0.2746993601322174,
+      "learning_rate": 0.0009503922950306274,
+      "loss": 1.3311,
+      "step": 1104
+    },
+    {
+      "epoch": 0.50682261208577,
+      "grad_norm": 0.191048264503479,
+      "learning_rate": 0.0009502846970317187,
+      "loss": 0.9343,
+      "step": 1105
+    },
+    {
+      "epoch": 0.5072812750831327,
+      "grad_norm": 0.24246980249881744,
+      "learning_rate": 0.0009501769885747271,
+      "loss": 0.9377,
+      "step": 1106
+    },
+    {
+      "epoch": 0.5077399380804953,
+      "grad_norm": 0.2772439122200012,
+      "learning_rate": 0.0009500691696860743,
+      "loss": 1.4328,
+      "step": 1107
+    },
+    {
+      "epoch": 0.5081986010778581,
+      "grad_norm": 0.11574332416057587,
+      "learning_rate": 0.0009499612403922091,
+      "loss": 0.7223,
+      "step": 1108
+    },
+    {
+      "epoch": 0.5086572640752207,
+      "grad_norm": 0.2554071247577667,
+      "learning_rate": 0.0009498532007196071,
+      "loss": 1.3056,
+      "step": 1109
+    },
+    {
+      "epoch": 0.5091159270725835,
+      "grad_norm": 0.2135448306798935,
+      "learning_rate": 0.0009497450506947714,
+      "loss": 1.161,
+      "step": 1110
+    },
+    {
+      "epoch": 0.5095745900699461,
+      "grad_norm": 0.1852402538061142,
+      "learning_rate": 0.000949636790344232,
+      "loss": 1.002,
+      "step": 1111
+    },
+    {
+      "epoch": 0.5100332530673088,
+      "grad_norm": 0.3562169373035431,
+      "learning_rate": 0.0009495284196945458,
+      "loss": 1.8675,
+      "step": 1112
+    },
+    {
+      "epoch": 0.5104919160646715,
+      "grad_norm": 0.336135596036911,
+      "learning_rate": 0.0009494199387722969,
+      "loss": 1.9324,
+      "step": 1113
+    },
+    {
+      "epoch": 0.5109505790620341,
+      "grad_norm": 0.38499322533607483,
+      "learning_rate": 0.0009493113476040966,
+      "loss": 2.3125,
+      "step": 1114
+    },
+    {
+      "epoch": 0.5114092420593969,
+      "grad_norm": 0.20560574531555176,
+      "learning_rate": 0.0009492026462165831,
+      "loss": 0.9785,
+      "step": 1115
+    },
+    {
+      "epoch": 0.5118679050567595,
+      "grad_norm": 0.36569714546203613,
+      "learning_rate": 0.0009490938346364215,
+      "loss": 1.9463,
+      "step": 1116
+    },
+    {
+      "epoch": 0.5123265680541222,
+      "grad_norm": 0.2573243975639343,
+      "learning_rate": 0.0009489849128903041,
+      "loss": 0.9407,
+      "step": 1117
+    },
+    {
+      "epoch": 0.5127852310514849,
+      "grad_norm": 0.3620086908340454,
+      "learning_rate": 0.0009488758810049503,
+      "loss": 2.1328,
+      "step": 1118
+    },
+    {
+      "epoch": 0.5132438940488476,
+      "grad_norm": 0.14623965322971344,
+      "learning_rate": 0.0009487667390071064,
+      "loss": 0.777,
+      "step": 1119
+    },
+    {
+      "epoch": 0.5137025570462103,
+      "grad_norm": 0.2728992998600006,
+      "learning_rate": 0.0009486574869235453,
+      "loss": 1.5247,
+      "step": 1120
+    },
+    {
+      "epoch": 0.514161220043573,
+      "grad_norm": 0.379833459854126,
+      "learning_rate": 0.0009485481247810681,
+      "loss": 1.8743,
+      "step": 1121
+    },
+    {
+      "epoch": 0.5146198830409356,
+      "grad_norm": 0.1932978332042694,
+      "learning_rate": 0.0009484386526065014,
+      "loss": 0.795,
+      "step": 1122
+    },
+    {
+      "epoch": 0.5150785460382984,
+      "grad_norm": 0.3890305459499359,
+      "learning_rate": 0.0009483290704266999,
+      "loss": 2.0994,
+      "step": 1123
+    },
+    {
+      "epoch": 0.515537209035661,
+      "grad_norm": 0.2669411897659302,
+      "learning_rate": 0.0009482193782685449,
+      "loss": 1.5108,
+      "step": 1124
+    },
+    {
+      "epoch": 0.5159958720330238,
+      "grad_norm": 0.2894323766231537,
+      "learning_rate": 0.0009481095761589445,
+      "loss": 1.3339,
+      "step": 1125
+    },
+    {
+      "epoch": 0.5164545350303864,
+      "grad_norm": 0.32780617475509644,
+      "learning_rate": 0.0009479996641248339,
+      "loss": 1.2261,
+      "step": 1126
+    },
+    {
+      "epoch": 0.5169131980277492,
+      "grad_norm": 0.2773579955101013,
+      "learning_rate": 0.0009478896421931755,
+      "loss": 1.3108,
+      "step": 1127
+    },
+    {
+      "epoch": 0.5173718610251118,
+      "grad_norm": 0.25270065665245056,
+      "learning_rate": 0.0009477795103909586,
+      "loss": 1.2581,
+      "step": 1128
+    },
+    {
+      "epoch": 0.5178305240224745,
+      "grad_norm": 0.34188172221183777,
+      "learning_rate": 0.000947669268745199,
+      "loss": 1.7618,
+      "step": 1129
+    },
+    {
+      "epoch": 0.5182891870198372,
+      "grad_norm": 0.30379557609558105,
+      "learning_rate": 0.00094755891728294,
+      "loss": 1.5002,
+      "step": 1130
+    },
+    {
+      "epoch": 0.5187478500171998,
+      "grad_norm": 0.29648905992507935,
+      "learning_rate": 0.0009474484560312514,
+      "loss": 1.1797,
+      "step": 1131
+    },
+    {
+      "epoch": 0.5192065130145626,
+      "grad_norm": 0.3193150758743286,
+      "learning_rate": 0.0009473378850172303,
+      "loss": 1.899,
+      "step": 1132
+    },
+    {
+      "epoch": 0.5196651760119252,
+      "grad_norm": 0.3248406946659088,
+      "learning_rate": 0.0009472272042680005,
+      "loss": 1.7289,
+      "step": 1133
+    },
+    {
+      "epoch": 0.5201238390092879,
+      "grad_norm": 0.26000529527664185,
+      "learning_rate": 0.000947116413810713,
+      "loss": 1.4107,
+      "step": 1134
+    },
+    {
+      "epoch": 0.5205825020066506,
+      "grad_norm": 0.27710065245628357,
+      "learning_rate": 0.0009470055136725451,
+      "loss": 1.3737,
+      "step": 1135
+    },
+    {
+      "epoch": 0.5210411650040133,
+      "grad_norm": 0.3785195052623749,
+      "learning_rate": 0.0009468945038807018,
+      "loss": 1.7585,
+      "step": 1136
+    },
+    {
+      "epoch": 0.521499828001376,
+      "grad_norm": 0.3857591450214386,
+      "learning_rate": 0.0009467833844624142,
+      "loss": 2.188,
+      "step": 1137
+    },
+    {
+      "epoch": 0.5219584909987387,
+      "grad_norm": 0.30060064792633057,
+      "learning_rate": 0.0009466721554449412,
+      "loss": 1.424,
+      "step": 1138
+    },
+    {
+      "epoch": 0.5224171539961013,
+      "grad_norm": 0.29208019375801086,
+      "learning_rate": 0.0009465608168555677,
+      "loss": 1.2755,
+      "step": 1139
+    },
+    {
+      "epoch": 0.5228758169934641,
+      "grad_norm": 5.316793441772461,
+      "learning_rate": 0.0009464493687216058,
+      "loss": 2.1242,
+      "step": 1140
+    },
+    {
+      "epoch": 0.5233344799908267,
+      "grad_norm": 0.2195115089416504,
+      "learning_rate": 0.0009463378110703949,
+      "loss": 1.0337,
+      "step": 1141
+    },
+    {
+      "epoch": 0.5237931429881895,
+      "grad_norm": 0.24998384714126587,
+      "learning_rate": 0.0009462261439293005,
+      "loss": 1.4111,
+      "step": 1142
+    },
+    {
+      "epoch": 0.5242518059855521,
+      "grad_norm": 0.3403361737728119,
+      "learning_rate": 0.0009461143673257156,
+      "loss": 1.9103,
+      "step": 1143
+    },
+    {
+      "epoch": 0.5247104689829148,
+      "grad_norm": 0.17883948981761932,
+      "learning_rate": 0.0009460024812870598,
+      "loss": 0.9974,
+      "step": 1144
+    },
+    {
+      "epoch": 0.5251691319802775,
+      "grad_norm": 0.24725860357284546,
+      "learning_rate": 0.0009458904858407794,
+      "loss": 1.4016,
+      "step": 1145
+    },
+    {
+      "epoch": 0.5256277949776402,
+      "grad_norm": 0.29898151755332947,
+      "learning_rate": 0.0009457783810143479,
+      "loss": 1.701,
+      "step": 1146
+    },
+    {
+      "epoch": 0.5260864579750029,
+      "grad_norm": 0.27674156427383423,
+      "learning_rate": 0.0009456661668352652,
+      "loss": 1.3168,
+      "step": 1147
+    },
+    {
+      "epoch": 0.5265451209723655,
+      "grad_norm": 0.2375446856021881,
+      "learning_rate": 0.0009455538433310584,
+      "loss": 1.0746,
+      "step": 1148
+    },
+    {
+      "epoch": 0.5270037839697282,
+      "grad_norm": 0.3496798574924469,
+      "learning_rate": 0.0009454414105292812,
+      "loss": 1.881,
+      "step": 1149
+    },
+    {
+      "epoch": 0.5274624469670909,
+      "grad_norm": 0.3575010895729065,
+      "learning_rate": 0.0009453288684575143,
+      "loss": 1.9427,
+      "step": 1150
+    },
+    {
+      "epoch": 0.5279211099644536,
+      "grad_norm": 0.3948631286621094,
+      "learning_rate": 0.0009452162171433648,
+      "loss": 2.2327,
+      "step": 1151
+    },
+    {
+      "epoch": 0.5283797729618163,
+      "grad_norm": 0.1527937948703766,
+      "learning_rate": 0.0009451034566144671,
+      "loss": 0.8336,
+      "step": 1152
+    },
+    {
+      "epoch": 0.528838435959179,
+      "grad_norm": 0.3415633738040924,
+      "learning_rate": 0.0009449905868984822,
+      "loss": 1.9399,
+      "step": 1153
+    },
+    {
+      "epoch": 0.5292970989565416,
+      "grad_norm": 0.2828967571258545,
+      "learning_rate": 0.0009448776080230979,
+      "loss": 1.3918,
+      "step": 1154
+    },
+    {
+      "epoch": 0.5297557619539044,
+      "grad_norm": 0.20446327328681946,
+      "learning_rate": 0.0009447645200160285,
+      "loss": 1.025,
+      "step": 1155
+    },
+    {
+      "epoch": 0.530214424951267,
+      "grad_norm": 0.3204951286315918,
+      "learning_rate": 0.0009446513229050154,
+      "loss": 1.8108,
+      "step": 1156
+    },
+    {
+      "epoch": 0.5306730879486298,
+      "grad_norm": 0.38850653171539307,
+      "learning_rate": 0.0009445380167178266,
+      "loss": 2.2024,
+      "step": 1157
+    },
+    {
+      "epoch": 0.5311317509459924,
+      "grad_norm": 0.22903253138065338,
+      "learning_rate": 0.0009444246014822571,
+      "loss": 1.0826,
+      "step": 1158
+    },
+    {
+      "epoch": 0.5315904139433552,
+      "grad_norm": 0.2147892266511917,
+      "learning_rate": 0.0009443110772261286,
+      "loss": 1.0546,
+      "step": 1159
+    },
+    {
+      "epoch": 0.5320490769407178,
+      "grad_norm": 0.34580257534980774,
+      "learning_rate": 0.0009441974439772889,
+      "loss": 1.7944,
+      "step": 1160
+    },
+    {
+      "epoch": 0.5325077399380805,
+      "grad_norm": 0.2733863592147827,
+      "learning_rate": 0.0009440837017636134,
+      "loss": 1.3638,
+      "step": 1161
+    },
+    {
+      "epoch": 0.5329664029354432,
+      "grad_norm": 0.3368569016456604,
+      "learning_rate": 0.000943969850613004,
+      "loss": 1.8119,
+      "step": 1162
+    },
+    {
+      "epoch": 0.5334250659328059,
+      "grad_norm": 0.25083136558532715,
+      "learning_rate": 0.0009438558905533889,
+      "loss": 1.414,
+      "step": 1163
+    },
+    {
+      "epoch": 0.5338837289301686,
+      "grad_norm": 0.3574562966823578,
+      "learning_rate": 0.0009437418216127236,
+      "loss": 2.1129,
+      "step": 1164
+    },
+    {
+      "epoch": 0.5343423919275313,
+      "grad_norm": 0.2785975933074951,
+      "learning_rate": 0.0009436276438189899,
+      "loss": 1.5375,
+      "step": 1165
+    },
+    {
+      "epoch": 0.5348010549248939,
+      "grad_norm": 0.34642040729522705,
+      "learning_rate": 0.0009435133572001965,
+      "loss": 1.8992,
+      "step": 1166
+    },
+    {
+      "epoch": 0.5352597179222566,
+      "grad_norm": 0.4057437479496002,
+      "learning_rate": 0.0009433989617843786,
+      "loss": 2.4216,
+      "step": 1167
+    },
+    {
+      "epoch": 0.5357183809196193,
+      "grad_norm": 0.2808201313018799,
+      "learning_rate": 0.0009432844575995983,
+      "loss": 1.3363,
+      "step": 1168
+    },
+    {
+      "epoch": 0.536177043916982,
+      "grad_norm": 0.29180237650871277,
+      "learning_rate": 0.0009431698446739443,
+      "loss": 1.7092,
+      "step": 1169
+    },
+    {
+      "epoch": 0.5366357069143447,
+      "grad_norm": 0.28107771277427673,
+      "learning_rate": 0.000943055123035532,
+      "loss": 1.2791,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5370943699117073,
+      "grad_norm": 0.17355158925056458,
+      "learning_rate": 0.0009429402927125035,
+      "loss": 0.8209,
+      "step": 1171
+    },
+    {
+      "epoch": 0.5375530329090701,
+      "grad_norm": 0.31418073177337646,
+      "learning_rate": 0.0009428253537330272,
+      "loss": 1.6716,
+      "step": 1172
+    },
+    {
+      "epoch": 0.5380116959064327,
+      "grad_norm": 0.3649182617664337,
+      "learning_rate": 0.0009427103061252989,
+      "loss": 1.8427,
+      "step": 1173
+    },
+    {
+      "epoch": 0.5384703589037955,
+      "grad_norm": 0.41031497716903687,
+      "learning_rate": 0.0009425951499175404,
+      "loss": 2.2205,
+      "step": 1174
+    },
+    {
+      "epoch": 0.5389290219011581,
+      "grad_norm": 0.2333560585975647,
+      "learning_rate": 0.0009424798851380003,
+      "loss": 0.9493,
+      "step": 1175
+    },
+    {
+      "epoch": 0.5393876848985208,
+      "grad_norm": 0.31085172295570374,
+      "learning_rate": 0.0009423645118149539,
+      "loss": 1.4224,
+      "step": 1176
+    },
+    {
+      "epoch": 0.5398463478958835,
+      "grad_norm": 0.41655832529067993,
+      "learning_rate": 0.0009422490299767032,
+      "loss": 2.1355,
+      "step": 1177
+    },
+    {
+      "epoch": 0.5403050108932462,
+      "grad_norm": 0.390559583902359,
+      "learning_rate": 0.0009421334396515766,
+      "loss": 1.7881,
+      "step": 1178
+    },
+    {
+      "epoch": 0.5407636738906089,
+      "grad_norm": 0.3651013970375061,
+      "learning_rate": 0.0009420177408679294,
+      "loss": 1.9978,
+      "step": 1179
+    },
+    {
+      "epoch": 0.5412223368879716,
+      "grad_norm": 0.2856077551841736,
+      "learning_rate": 0.0009419019336541431,
+      "loss": 1.3818,
+      "step": 1180
+    },
+    {
+      "epoch": 0.5416809998853342,
+      "grad_norm": 0.2839828431606293,
+      "learning_rate": 0.0009417860180386264,
+      "loss": 1.3959,
+      "step": 1181
+    },
+    {
+      "epoch": 0.542139662882697,
+      "grad_norm": 0.338052898645401,
+      "learning_rate": 0.0009416699940498139,
+      "loss": 1.8724,
+      "step": 1182
+    },
+    {
+      "epoch": 0.5425983258800596,
+      "grad_norm": 0.26623108983039856,
+      "learning_rate": 0.0009415538617161672,
+      "loss": 1.5506,
+      "step": 1183
+    },
+    {
+      "epoch": 0.5430569888774223,
+      "grad_norm": 0.3088790774345398,
+      "learning_rate": 0.0009414376210661746,
+      "loss": 1.7001,
+      "step": 1184
+    },
+    {
+      "epoch": 0.543515651874785,
+      "grad_norm": 0.2527974545955658,
+      "learning_rate": 0.0009413212721283505,
+      "loss": 1.5046,
+      "step": 1185
+    },
+    {
+      "epoch": 0.5439743148721476,
+      "grad_norm": 0.29992103576660156,
+      "learning_rate": 0.0009412048149312364,
+      "loss": 1.405,
+      "step": 1186
+    },
+    {
+      "epoch": 0.5444329778695104,
+      "grad_norm": 0.26170700788497925,
+      "learning_rate": 0.0009410882495033998,
+      "loss": 1.1761,
+      "step": 1187
+    },
+    {
+      "epoch": 0.544891640866873,
+      "grad_norm": 0.1813340187072754,
+      "learning_rate": 0.0009409715758734352,
+      "loss": 0.9035,
+      "step": 1188
+    },
+    {
+      "epoch": 0.5453503038642358,
+      "grad_norm": 0.3251456916332245,
+      "learning_rate": 0.0009408547940699634,
+      "loss": 1.5636,
+      "step": 1189
+    },
+    {
+      "epoch": 0.5458089668615984,
+      "grad_norm": 0.3176114857196808,
+      "learning_rate": 0.0009407379041216321,
+      "loss": 1.402,
+      "step": 1190
+    },
+    {
+      "epoch": 0.5462676298589612,
+      "grad_norm": 0.21962450444698334,
+      "learning_rate": 0.0009406209060571149,
+      "loss": 1.009,
+      "step": 1191
+    },
+    {
+      "epoch": 0.5467262928563238,
+      "grad_norm": 0.25622203946113586,
+      "learning_rate": 0.0009405037999051125,
+      "loss": 0.9907,
+      "step": 1192
+    },
+    {
+      "epoch": 0.5471849558536865,
+      "grad_norm": 0.4034391939640045,
+      "learning_rate": 0.0009403865856943516,
+      "loss": 2.1548,
+      "step": 1193
+    },
+    {
+      "epoch": 0.5476436188510492,
+      "grad_norm": 0.10982491821050644,
+      "learning_rate": 0.0009402692634535861,
+      "loss": 0.5598,
+      "step": 1194
+    },
+    {
+      "epoch": 0.5481022818484119,
+      "grad_norm": 0.21411536633968353,
+      "learning_rate": 0.0009401518332115957,
+      "loss": 0.9482,
+      "step": 1195
+    },
+    {
+      "epoch": 0.5485609448457746,
+      "grad_norm": 0.35385677218437195,
+      "learning_rate": 0.0009400342949971868,
+      "loss": 1.8314,
+      "step": 1196
+    },
+    {
+      "epoch": 0.5490196078431373,
+      "grad_norm": 0.3468371033668518,
+      "learning_rate": 0.0009399166488391927,
+      "loss": 2.0469,
+      "step": 1197
+    },
+    {
+      "epoch": 0.5494782708404999,
+      "grad_norm": 0.315140038728714,
+      "learning_rate": 0.0009397988947664727,
+      "loss": 1.8289,
+      "step": 1198
+    },
+    {
+      "epoch": 0.5499369338378627,
+      "grad_norm": 0.2434529811143875,
+      "learning_rate": 0.0009396810328079126,
+      "loss": 1.2443,
+      "step": 1199
+    },
+    {
+      "epoch": 0.5503955968352253,
+      "grad_norm": 0.24115651845932007,
+      "learning_rate": 0.0009395630629924248,
+      "loss": 1.2274,
+      "step": 1200
+    },
+    {
+      "epoch": 0.550854259832588,
+      "grad_norm": 0.12398785352706909,
+      "learning_rate": 0.0009394449853489484,
+      "loss": 0.6473,
+      "step": 1201
+    },
+    {
+      "epoch": 0.5513129228299507,
+      "grad_norm": 0.3112662732601166,
+      "learning_rate": 0.0009393267999064486,
+      "loss": 1.5162,
+      "step": 1202
+    },
+    {
+      "epoch": 0.5517715858273133,
+      "grad_norm": 0.36855632066726685,
+      "learning_rate": 0.0009392085066939169,
+      "loss": 1.9268,
+      "step": 1203
+    },
+    {
+      "epoch": 0.5522302488246761,
+      "grad_norm": 0.24338075518608093,
+      "learning_rate": 0.0009390901057403716,
+      "loss": 1.1956,
+      "step": 1204
+    },
+    {
+      "epoch": 0.5526889118220387,
+      "grad_norm": 0.3083001375198364,
+      "learning_rate": 0.0009389715970748575,
+      "loss": 1.5538,
+      "step": 1205
+    },
+    {
+      "epoch": 0.5531475748194015,
+      "grad_norm": 0.2836032807826996,
+      "learning_rate": 0.0009388529807264455,
+      "loss": 1.4889,
+      "step": 1206
+    },
+    {
+      "epoch": 0.5536062378167641,
+      "grad_norm": 0.2570768892765045,
+      "learning_rate": 0.000938734256724233,
+      "loss": 1.3693,
+      "step": 1207
+    },
+    {
+      "epoch": 0.5540649008141268,
+      "grad_norm": 0.3241545855998993,
+      "learning_rate": 0.0009386154250973438,
+      "loss": 1.6445,
+      "step": 1208
+    },
+    {
+      "epoch": 0.5545235638114895,
+      "grad_norm": 0.29821258783340454,
+      "learning_rate": 0.0009384964858749283,
+      "loss": 1.4775,
+      "step": 1209
+    },
+    {
+      "epoch": 0.5549822268088522,
+      "grad_norm": 0.22773027420043945,
+      "learning_rate": 0.000938377439086163,
+      "loss": 1.2509,
+      "step": 1210
+    },
+    {
+      "epoch": 0.5554408898062149,
+      "grad_norm": 0.31152674555778503,
+      "learning_rate": 0.0009382582847602512,
+      "loss": 1.6638,
+      "step": 1211
+    },
+    {
+      "epoch": 0.5558995528035776,
+      "grad_norm": 0.11729778349399567,
+      "learning_rate": 0.0009381390229264221,
+      "loss": 0.5536,
+      "step": 1212
+    },
+    {
+      "epoch": 0.5563582158009402,
+      "grad_norm": 0.1624806672334671,
+      "learning_rate": 0.0009380196536139315,
+      "loss": 0.7977,
+      "step": 1213
+    },
+    {
+      "epoch": 0.556816878798303,
+      "grad_norm": 0.32244980335235596,
+      "learning_rate": 0.0009379001768520615,
+      "loss": 1.5859,
+      "step": 1214
+    },
+    {
+      "epoch": 0.5572755417956656,
+      "grad_norm": 0.2572208344936371,
+      "learning_rate": 0.0009377805926701208,
+      "loss": 1.2651,
+      "step": 1215
+    },
+    {
+      "epoch": 0.5577342047930284,
+      "grad_norm": 0.34207504987716675,
+      "learning_rate": 0.0009376609010974442,
+      "loss": 1.817,
+      "step": 1216
+    },
+    {
+      "epoch": 0.558192867790391,
+      "grad_norm": 0.33612552285194397,
+      "learning_rate": 0.0009375411021633927,
+      "loss": 1.7496,
+      "step": 1217
+    },
+    {
+      "epoch": 0.5586515307877536,
+      "grad_norm": 0.291958212852478,
+      "learning_rate": 0.0009374211958973542,
+      "loss": 1.6475,
+      "step": 1218
+    },
+    {
+      "epoch": 0.5591101937851164,
+      "grad_norm": 0.28733986616134644,
+      "learning_rate": 0.0009373011823287422,
+      "loss": 1.434,
+      "step": 1219
+    },
+    {
+      "epoch": 0.559568856782479,
+      "grad_norm": 0.2283989042043686,
+      "learning_rate": 0.0009371810614869971,
+      "loss": 0.9546,
+      "step": 1220
+    },
+    {
+      "epoch": 0.5600275197798418,
+      "grad_norm": 0.330026239156723,
+      "learning_rate": 0.0009370608334015856,
+      "loss": 1.5756,
+      "step": 1221
+    },
+    {
+      "epoch": 0.5604861827772044,
+      "grad_norm": 0.37356528639793396,
+      "learning_rate": 0.000936940498102,
+      "loss": 2.1412,
+      "step": 1222
+    },
+    {
+      "epoch": 0.5609448457745672,
+      "grad_norm": 0.35292860865592957,
+      "learning_rate": 0.0009368200556177598,
+      "loss": 1.697,
+      "step": 1223
+    },
+    {
+      "epoch": 0.5614035087719298,
+      "grad_norm": 0.30706706643104553,
+      "learning_rate": 0.0009366995059784104,
+      "loss": 1.2438,
+      "step": 1224
+    },
+    {
+      "epoch": 0.5618621717692925,
+      "grad_norm": 0.3499191999435425,
+      "learning_rate": 0.0009365788492135235,
+      "loss": 1.5356,
+      "step": 1225
+    },
+    {
+      "epoch": 0.5623208347666552,
+      "grad_norm": 0.16763897240161896,
+      "learning_rate": 0.0009364580853526967,
+      "loss": 0.8041,
+      "step": 1226
+    },
+    {
+      "epoch": 0.5627794977640179,
+      "grad_norm": 0.2519925534725189,
+      "learning_rate": 0.0009363372144255548,
+      "loss": 1.1698,
+      "step": 1227
+    },
+    {
+      "epoch": 0.5632381607613806,
+      "grad_norm": 0.17550460994243622,
+      "learning_rate": 0.0009362162364617479,
+      "loss": 0.9668,
+      "step": 1228
+    },
+    {
+      "epoch": 0.5636968237587433,
+      "grad_norm": 0.31531476974487305,
+      "learning_rate": 0.0009360951514909529,
+      "loss": 1.7645,
+      "step": 1229
+    },
+    {
+      "epoch": 0.5641554867561059,
+      "grad_norm": 0.26998183131217957,
+      "learning_rate": 0.0009359739595428729,
+      "loss": 1.194,
+      "step": 1230
+    },
+    {
+      "epoch": 0.5646141497534687,
+      "grad_norm": 0.3305562734603882,
+      "learning_rate": 0.000935852660647237,
+      "loss": 1.6608,
+      "step": 1231
+    },
+    {
+      "epoch": 0.5650728127508313,
+      "grad_norm": 0.3273736536502838,
+      "learning_rate": 0.000935731254833801,
+      "loss": 1.4538,
+      "step": 1232
+    },
+    {
+      "epoch": 0.5655314757481941,
+      "grad_norm": 0.3271407186985016,
+      "learning_rate": 0.0009356097421323461,
+      "loss": 1.8257,
+      "step": 1233
+    },
+    {
+      "epoch": 0.5659901387455567,
+      "grad_norm": 0.1826079934835434,
+      "learning_rate": 0.0009354881225726808,
+      "loss": 0.8564,
+      "step": 1234
+    },
+    {
+      "epoch": 0.5664488017429193,
+      "grad_norm": 0.255459725856781,
+      "learning_rate": 0.0009353663961846389,
+      "loss": 1.4302,
+      "step": 1235
+    },
+    {
+      "epoch": 0.5669074647402821,
+      "grad_norm": 0.2576141357421875,
+      "learning_rate": 0.0009352445629980809,
+      "loss": 1.3641,
+      "step": 1236
+    },
+    {
+      "epoch": 0.5673661277376447,
+      "grad_norm": 0.31090545654296875,
+      "learning_rate": 0.0009351226230428934,
+      "loss": 1.6024,
+      "step": 1237
+    },
+    {
+      "epoch": 0.5678247907350075,
+      "grad_norm": 0.2738056778907776,
+      "learning_rate": 0.0009350005763489888,
+      "loss": 1.4413,
+      "step": 1238
+    },
+    {
+      "epoch": 0.5682834537323701,
+      "grad_norm": 0.34732604026794434,
+      "learning_rate": 0.0009348784229463065,
+      "loss": 2.0205,
+      "step": 1239
+    },
+    {
+      "epoch": 0.5687421167297328,
+      "grad_norm": 0.2433692067861557,
+      "learning_rate": 0.0009347561628648115,
+      "loss": 1.4734,
+      "step": 1240
+    },
+    {
+      "epoch": 0.5692007797270955,
+      "grad_norm": 0.2669577896595001,
+      "learning_rate": 0.0009346337961344948,
+      "loss": 1.3418,
+      "step": 1241
+    },
+    {
+      "epoch": 0.5696594427244582,
+      "grad_norm": 0.2639061510562897,
+      "learning_rate": 0.0009345113227853741,
+      "loss": 1.4639,
+      "step": 1242
+    },
+    {
+      "epoch": 0.5701181057218209,
+      "grad_norm": 0.29163965582847595,
+      "learning_rate": 0.000934388742847493,
+      "loss": 1.4735,
+      "step": 1243
+    },
+    {
+      "epoch": 0.5705767687191836,
+      "grad_norm": 0.1659363955259323,
+      "learning_rate": 0.0009342660563509211,
+      "loss": 0.8228,
+      "step": 1244
+    },
+    {
+      "epoch": 0.5710354317165462,
+      "grad_norm": 0.2903018295764923,
+      "learning_rate": 0.0009341432633257543,
+      "loss": 1.3949,
+      "step": 1245
+    },
+    {
+      "epoch": 0.571494094713909,
+      "grad_norm": 0.39374253153800964,
+      "learning_rate": 0.0009340203638021149,
+      "loss": 2.1997,
+      "step": 1246
+    },
+    {
+      "epoch": 0.5719527577112716,
+      "grad_norm": 0.31729447841644287,
+      "learning_rate": 0.0009338973578101506,
+      "loss": 1.4435,
+      "step": 1247
+    },
+    {
+      "epoch": 0.5724114207086344,
+      "grad_norm": 0.12577004730701447,
+      "learning_rate": 0.000933774245380036,
+      "loss": 0.6955,
+      "step": 1248
+    },
+    {
+      "epoch": 0.572870083705997,
+      "grad_norm": 0.3122677206993103,
+      "learning_rate": 0.0009336510265419712,
+      "loss": 1.5015,
+      "step": 1249
+    },
+    {
+      "epoch": 0.5733287467033598,
+      "grad_norm": 0.2251952886581421,
+      "learning_rate": 0.000933527701326183,
+      "loss": 0.9793,
+      "step": 1250
+    },
+    {
+      "epoch": 0.5737874097007224,
+      "grad_norm": 0.2624088227748871,
+      "learning_rate": 0.0009334042697629235,
+      "loss": 1.2029,
+      "step": 1251
+    },
+    {
+      "epoch": 0.574246072698085,
+      "grad_norm": 0.2972775399684906,
+      "learning_rate": 0.0009332807318824717,
+      "loss": 1.3044,
+      "step": 1252
+    },
+    {
+      "epoch": 0.5747047356954478,
+      "grad_norm": 0.6643463969230652,
+      "learning_rate": 0.0009331570877151324,
+      "loss": 1.9626,
+      "step": 1253
+    },
+    {
+      "epoch": 0.5751633986928104,
+      "grad_norm": 0.25192347168922424,
+      "learning_rate": 0.0009330333372912361,
+      "loss": 1.2851,
+      "step": 1254
+    },
+    {
+      "epoch": 0.5756220616901732,
+      "grad_norm": 0.28251874446868896,
+      "learning_rate": 0.0009329094806411401,
+      "loss": 1.4486,
+      "step": 1255
+    },
+    {
+      "epoch": 0.5760807246875358,
+      "grad_norm": 0.3544906675815582,
+      "learning_rate": 0.0009327855177952267,
+      "loss": 1.7761,
+      "step": 1256
+    },
+    {
+      "epoch": 0.5765393876848985,
+      "grad_norm": 0.18498443067073822,
+      "learning_rate": 0.0009326614487839053,
+      "loss": 0.8643,
+      "step": 1257
+    },
+    {
+      "epoch": 0.5769980506822612,
+      "grad_norm": 0.2767449617385864,
+      "learning_rate": 0.0009325372736376109,
+      "loss": 1.3204,
+      "step": 1258
+    },
+    {
+      "epoch": 0.5774567136796239,
+      "grad_norm": 0.32360634207725525,
+      "learning_rate": 0.0009324129923868048,
+      "loss": 1.555,
+      "step": 1259
+    },
+    {
+      "epoch": 0.5779153766769866,
+      "grad_norm": 0.09593556076288223,
+      "learning_rate": 0.0009322886050619735,
+      "loss": 0.5922,
+      "step": 1260
+    },
+    {
+      "epoch": 0.5783740396743493,
+      "grad_norm": 0.2054739147424698,
+      "learning_rate": 0.0009321641116936306,
+      "loss": 0.8165,
+      "step": 1261
+    },
+    {
+      "epoch": 0.5788327026717119,
+      "grad_norm": 0.6027858257293701,
+      "learning_rate": 0.0009320395123123149,
+      "loss": 1.8843,
+      "step": 1262
+    },
+    {
+      "epoch": 0.5792913656690747,
+      "grad_norm": 0.31743377447128296,
+      "learning_rate": 0.0009319148069485917,
+      "loss": 1.2451,
+      "step": 1263
+    },
+    {
+      "epoch": 0.5797500286664373,
+      "grad_norm": 0.2888292670249939,
+      "learning_rate": 0.0009317899956330522,
+      "loss": 1.4385,
+      "step": 1264
+    },
+    {
+      "epoch": 0.5802086916638001,
+      "grad_norm": 0.17408441007137299,
+      "learning_rate": 0.0009316650783963132,
+      "loss": 0.7488,
+      "step": 1265
+    },
+    {
+      "epoch": 0.5806673546611627,
+      "grad_norm": 0.3600836396217346,
+      "learning_rate": 0.0009315400552690181,
+      "loss": 1.5487,
+      "step": 1266
+    },
+    {
+      "epoch": 0.5811260176585255,
+      "grad_norm": 0.2011416256427765,
+      "learning_rate": 0.0009314149262818358,
+      "loss": 0.825,
+      "step": 1267
+    },
+    {
+      "epoch": 0.5815846806558881,
+      "grad_norm": 0.37469393014907837,
+      "learning_rate": 0.0009312896914654616,
+      "loss": 1.4067,
+      "step": 1268
+    },
+    {
+      "epoch": 0.5820433436532507,
+      "grad_norm": 0.2877853512763977,
+      "learning_rate": 0.0009311643508506162,
+      "loss": 1.4527,
+      "step": 1269
+    },
+    {
+      "epoch": 0.5825020066506135,
+      "grad_norm": 0.3054576814174652,
+      "learning_rate": 0.0009310389044680467,
+      "loss": 1.39,
+      "step": 1270
+    },
+    {
+      "epoch": 0.5829606696479761,
+      "grad_norm": 0.09884677827358246,
+      "learning_rate": 0.000930913352348526,
+      "loss": 0.6243,
+      "step": 1271
+    },
+    {
+      "epoch": 0.5834193326453389,
+      "grad_norm": 0.25772392749786377,
+      "learning_rate": 0.0009307876945228528,
+      "loss": 1.4088,
+      "step": 1272
+    },
+    {
+      "epoch": 0.5838779956427015,
+      "grad_norm": 0.3259493112564087,
+      "learning_rate": 0.0009306619310218521,
+      "loss": 1.6538,
+      "step": 1273
+    },
+    {
+      "epoch": 0.5843366586400642,
+      "grad_norm": 0.09127166122198105,
+      "learning_rate": 0.0009305360618763745,
+      "loss": 0.5361,
+      "step": 1274
+    },
+    {
+      "epoch": 0.5847953216374269,
+      "grad_norm": 0.18113906681537628,
+      "learning_rate": 0.0009304100871172967,
+      "loss": 0.9267,
+      "step": 1275
+    },
+    {
+      "epoch": 0.5852539846347896,
+      "grad_norm": 0.32655301690101624,
+      "learning_rate": 0.000930284006775521,
+      "loss": 1.6382,
+      "step": 1276
+    },
+    {
+      "epoch": 0.5857126476321523,
+      "grad_norm": 0.19575290381908417,
+      "learning_rate": 0.0009301578208819758,
+      "loss": 0.9375,
+      "step": 1277
+    },
+    {
+      "epoch": 0.586171310629515,
+      "grad_norm": 0.3941885828971863,
+      "learning_rate": 0.0009300315294676158,
+      "loss": 2.1287,
+      "step": 1278
+    },
+    {
+      "epoch": 0.5866299736268776,
+      "grad_norm": 0.28158482909202576,
+      "learning_rate": 0.0009299051325634208,
+      "loss": 1.3167,
+      "step": 1279
+    },
+    {
+      "epoch": 0.5870886366242404,
+      "grad_norm": 0.1906464397907257,
+      "learning_rate": 0.000929778630200397,
+      "loss": 0.8625,
+      "step": 1280
+    },
+    {
+      "epoch": 0.587547299621603,
+      "grad_norm": 0.1061529740691185,
+      "learning_rate": 0.0009296520224095764,
+      "loss": 0.6863,
+      "step": 1281
+    },
+    {
+      "epoch": 0.5880059626189658,
+      "grad_norm": 0.26948490738868713,
+      "learning_rate": 0.0009295253092220166,
+      "loss": 0.886,
+      "step": 1282
+    },
+    {
+      "epoch": 0.5884646256163284,
+      "grad_norm": 0.36432507634162903,
+      "learning_rate": 0.0009293984906688016,
+      "loss": 1.8413,
+      "step": 1283
+    },
+    {
+      "epoch": 0.5889232886136911,
+      "grad_norm": 0.5231337547302246,
+      "learning_rate": 0.0009292715667810406,
+      "loss": 1.7709,
+      "step": 1284
+    },
+    {
+      "epoch": 0.5893819516110538,
+      "grad_norm": 0.3480357825756073,
+      "learning_rate": 0.000929144537589869,
+      "loss": 1.9674,
+      "step": 1285
+    },
+    {
+      "epoch": 0.5898406146084164,
+      "grad_norm": 0.24957574903964996,
+      "learning_rate": 0.0009290174031264482,
+      "loss": 1.2784,
+      "step": 1286
+    },
+    {
+      "epoch": 0.5902992776057792,
+      "grad_norm": 0.33761247992515564,
+      "learning_rate": 0.000928890163421965,
+      "loss": 1.9197,
+      "step": 1287
+    },
+    {
+      "epoch": 0.5907579406031418,
+      "grad_norm": 0.35372909903526306,
+      "learning_rate": 0.0009287628185076322,
+      "loss": 2.0242,
+      "step": 1288
+    },
+    {
+      "epoch": 0.5912166036005045,
+      "grad_norm": 0.21915297210216522,
+      "learning_rate": 0.0009286353684146884,
+      "loss": 0.9164,
+      "step": 1289
+    },
+    {
+      "epoch": 0.5916752665978672,
+      "grad_norm": 0.3058471083641052,
+      "learning_rate": 0.0009285078131743982,
+      "loss": 1.5614,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5921339295952299,
+      "grad_norm": 0.1828153282403946,
+      "learning_rate": 0.0009283801528180517,
+      "loss": 0.8684,
+      "step": 1291
+    },
+    {
+      "epoch": 0.5925925925925926,
+      "grad_norm": 0.1839916706085205,
+      "learning_rate": 0.000928252387376965,
+      "loss": 0.8804,
+      "step": 1292
+    },
+    {
+      "epoch": 0.5930512555899553,
+      "grad_norm": 0.21669816970825195,
+      "learning_rate": 0.0009281245168824799,
+      "loss": 1.1169,
+      "step": 1293
+    },
+    {
+      "epoch": 0.5935099185873179,
+      "grad_norm": 0.2863485813140869,
+      "learning_rate": 0.0009279965413659637,
+      "loss": 1.5664,
+      "step": 1294
+    },
+    {
+      "epoch": 0.5939685815846807,
+      "grad_norm": 0.22065885365009308,
+      "learning_rate": 0.00092786846085881,
+      "loss": 1.0018,
+      "step": 1295
+    },
+    {
+      "epoch": 0.5944272445820433,
+      "grad_norm": 0.3145039975643158,
+      "learning_rate": 0.0009277402753924376,
+      "loss": 1.4487,
+      "step": 1296
+    },
+    {
+      "epoch": 0.5948859075794061,
+      "grad_norm": 0.2897392809391022,
+      "learning_rate": 0.0009276119849982917,
+      "loss": 1.384,
+      "step": 1297
+    },
+    {
+      "epoch": 0.5953445705767687,
+      "grad_norm": 0.3938884139060974,
+      "learning_rate": 0.0009274835897078425,
+      "loss": 2.1997,
+      "step": 1298
+    },
+    {
+      "epoch": 0.5958032335741315,
+      "grad_norm": 0.17570100724697113,
+      "learning_rate": 0.0009273550895525864,
+      "loss": 0.8196,
+      "step": 1299
+    },
+    {
+      "epoch": 0.5962618965714941,
+      "grad_norm": 0.32507872581481934,
+      "learning_rate": 0.0009272264845640455,
+      "loss": 1.4766,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5967205595688568,
+      "grad_norm": 0.19803722202777863,
+      "learning_rate": 0.0009270977747737675,
+      "loss": 0.9604,
+      "step": 1301
+    },
+    {
+      "epoch": 0.5971792225662195,
+      "grad_norm": 0.32639971375465393,
+      "learning_rate": 0.0009269689602133258,
+      "loss": 1.7936,
+      "step": 1302
+    },
+    {
+      "epoch": 0.5976378855635821,
+      "grad_norm": 0.35615068674087524,
+      "learning_rate": 0.0009268400409143195,
+      "loss": 1.7267,
+      "step": 1303
+    },
+    {
+      "epoch": 0.5980965485609449,
+      "grad_norm": 0.33645865321159363,
+      "learning_rate": 0.0009267110169083734,
+      "loss": 1.6238,
+      "step": 1304
+    },
+    {
+      "epoch": 0.5985552115583075,
+      "grad_norm": 0.29043132066726685,
+      "learning_rate": 0.0009265818882271384,
+      "loss": 1.2636,
+      "step": 1305
+    },
+    {
+      "epoch": 0.5990138745556702,
+      "grad_norm": 0.41903120279312134,
+      "learning_rate": 0.0009264526549022903,
+      "loss": 2.3948,
+      "step": 1306
+    },
+    {
+      "epoch": 0.5994725375530329,
+      "grad_norm": 0.2502308189868927,
+      "learning_rate": 0.0009263233169655309,
+      "loss": 1.1672,
+      "step": 1307
+    },
+    {
+      "epoch": 0.5999312005503956,
+      "grad_norm": 0.1772756576538086,
+      "learning_rate": 0.000926193874448588,
+      "loss": 0.8739,
+      "step": 1308
+    },
+    {
+      "epoch": 0.6003898635477583,
+      "grad_norm": 0.24992649257183075,
+      "learning_rate": 0.0009260643273832147,
+      "loss": 1.4184,
+      "step": 1309
+    },
+    {
+      "epoch": 0.600848526545121,
+      "grad_norm": 0.1002533808350563,
+      "learning_rate": 0.0009259346758011898,
+      "loss": 0.6456,
+      "step": 1310
+    },
+    {
+      "epoch": 0.6013071895424836,
+      "grad_norm": 0.3838384449481964,
+      "learning_rate": 0.0009258049197343177,
+      "loss": 2.1184,
+      "step": 1311
+    },
+    {
+      "epoch": 0.6017658525398464,
+      "grad_norm": 0.1766338050365448,
+      "learning_rate": 0.0009256750592144287,
+      "loss": 0.8759,
+      "step": 1312
+    },
+    {
+      "epoch": 0.602224515537209,
+      "grad_norm": 0.37328869104385376,
+      "learning_rate": 0.0009255450942733783,
+      "loss": 1.9109,
+      "step": 1313
+    },
+    {
+      "epoch": 0.6026831785345718,
+      "grad_norm": 0.19689656794071198,
+      "learning_rate": 0.0009254150249430479,
+      "loss": 0.9383,
+      "step": 1314
+    },
+    {
+      "epoch": 0.6031418415319344,
+      "grad_norm": 0.0794055312871933,
+      "learning_rate": 0.0009252848512553447,
+      "loss": 0.4766,
+      "step": 1315
+    },
+    {
+      "epoch": 0.6036005045292971,
+      "grad_norm": 0.277407705783844,
+      "learning_rate": 0.0009251545732422009,
+      "loss": 1.7058,
+      "step": 1316
+    },
+    {
+      "epoch": 0.6040591675266598,
+      "grad_norm": 0.28111398220062256,
+      "learning_rate": 0.0009250241909355746,
+      "loss": 1.8057,
+      "step": 1317
+    },
+    {
+      "epoch": 0.6045178305240225,
+      "grad_norm": 0.2518330514431,
+      "learning_rate": 0.0009248937043674499,
+      "loss": 1.3856,
+      "step": 1318
+    },
+    {
+      "epoch": 0.6049764935213852,
+      "grad_norm": 0.3877377510070801,
+      "learning_rate": 0.0009247631135698358,
+      "loss": 2.1638,
+      "step": 1319
+    },
+    {
+      "epoch": 0.6054351565187478,
+      "grad_norm": 0.2325863391160965,
+      "learning_rate": 0.0009246324185747672,
+      "loss": 1.3607,
+      "step": 1320
+    },
+    {
+      "epoch": 0.6058938195161105,
+      "grad_norm": 0.23192419111728668,
+      "learning_rate": 0.0009245016194143047,
+      "loss": 1.1667,
+      "step": 1321
+    },
+    {
+      "epoch": 0.6063524825134732,
+      "grad_norm": 0.0964299887418747,
+      "learning_rate": 0.000924370716120534,
+      "loss": 0.623,
+      "step": 1322
+    },
+    {
+      "epoch": 0.6068111455108359,
+      "grad_norm": 0.23401454091072083,
+      "learning_rate": 0.0009242397087255667,
+      "loss": 1.2682,
+      "step": 1323
+    },
+    {
+      "epoch": 0.6072698085081986,
+      "grad_norm": 0.30539819598197937,
+      "learning_rate": 0.0009241085972615401,
+      "loss": 1.9392,
+      "step": 1324
+    },
+    {
+      "epoch": 0.6077284715055613,
+      "grad_norm": 0.1904824823141098,
+      "learning_rate": 0.0009239773817606165,
+      "loss": 0.8026,
+      "step": 1325
+    },
+    {
+      "epoch": 0.6081871345029239,
+      "grad_norm": 0.31935641169548035,
+      "learning_rate": 0.0009238460622549842,
+      "loss": 1.7756,
+      "step": 1326
+    },
+    {
+      "epoch": 0.6086457975002867,
+      "grad_norm": 0.232622429728508,
+      "learning_rate": 0.0009237146387768567,
+      "loss": 1.2202,
+      "step": 1327
+    },
+    {
+      "epoch": 0.6091044604976493,
+      "grad_norm": 0.22386907041072845,
+      "learning_rate": 0.0009235831113584732,
+      "loss": 1.0802,
+      "step": 1328
+    },
+    {
+      "epoch": 0.6095631234950121,
+      "grad_norm": 0.2252710908651352,
+      "learning_rate": 0.0009234514800320983,
+      "loss": 0.9511,
+      "step": 1329
+    },
+    {
+      "epoch": 0.6100217864923747,
+      "grad_norm": 0.3520384430885315,
+      "learning_rate": 0.0009233197448300221,
+      "loss": 1.898,
+      "step": 1330
+    },
+    {
+      "epoch": 0.6104804494897375,
+      "grad_norm": 0.3444898724555969,
+      "learning_rate": 0.0009231879057845601,
+      "loss": 1.77,
+      "step": 1331
+    },
+    {
+      "epoch": 0.6109391124871001,
+      "grad_norm": 0.24436596035957336,
+      "learning_rate": 0.0009230559629280535,
+      "loss": 1.054,
+      "step": 1332
+    },
+    {
+      "epoch": 0.6113977754844628,
+      "grad_norm": 0.11772722005844116,
+      "learning_rate": 0.0009229239162928689,
+      "loss": 0.6621,
+      "step": 1333
+    },
+    {
+      "epoch": 0.6118564384818255,
+      "grad_norm": 0.2954722046852112,
+      "learning_rate": 0.0009227917659113982,
+      "loss": 1.5235,
+      "step": 1334
+    },
+    {
+      "epoch": 0.6123151014791882,
+      "grad_norm": 0.2937714755535126,
+      "learning_rate": 0.0009226595118160588,
+      "loss": 1.3464,
+      "step": 1335
+    },
+    {
+      "epoch": 0.6127737644765509,
+      "grad_norm": 0.27231964468955994,
+      "learning_rate": 0.0009225271540392934,
+      "loss": 1.2777,
+      "step": 1336
+    },
+    {
+      "epoch": 0.6132324274739135,
+      "grad_norm": 0.19506557285785675,
+      "learning_rate": 0.0009223946926135709,
+      "loss": 0.8657,
+      "step": 1337
+    },
+    {
+      "epoch": 0.6136910904712762,
+      "grad_norm": 0.39066511392593384,
+      "learning_rate": 0.0009222621275713844,
+      "loss": 1.8091,
+      "step": 1338
+    },
+    {
+      "epoch": 0.6141497534686389,
+      "grad_norm": 0.4800299108028412,
+      "learning_rate": 0.0009221294589452535,
+      "loss": 2.0739,
+      "step": 1339
+    },
+    {
+      "epoch": 0.6146084164660016,
+      "grad_norm": 0.3111729621887207,
+      "learning_rate": 0.0009219966867677226,
+      "loss": 1.5457,
+      "step": 1340
+    },
+    {
+      "epoch": 0.6150670794633643,
+      "grad_norm": 0.30846676230430603,
+      "learning_rate": 0.0009218638110713615,
+      "loss": 1.4479,
+      "step": 1341
+    },
+    {
+      "epoch": 0.615525742460727,
+      "grad_norm": 0.3237318992614746,
+      "learning_rate": 0.0009217308318887659,
+      "loss": 1.5574,
+      "step": 1342
+    },
+    {
+      "epoch": 0.6159844054580896,
+      "grad_norm": 0.2283388376235962,
+      "learning_rate": 0.0009215977492525565,
+      "loss": 1.1732,
+      "step": 1343
+    },
+    {
+      "epoch": 0.6164430684554524,
+      "grad_norm": 0.3451511263847351,
+      "learning_rate": 0.0009214645631953791,
+      "loss": 1.7261,
+      "step": 1344
+    },
+    {
+      "epoch": 0.616901731452815,
+      "grad_norm": 0.3029400110244751,
+      "learning_rate": 0.0009213312737499055,
+      "loss": 1.3227,
+      "step": 1345
+    },
+    {
+      "epoch": 0.6173603944501778,
+      "grad_norm": 0.17842943966388702,
+      "learning_rate": 0.0009211978809488327,
+      "loss": 0.5785,
+      "step": 1346
+    },
+    {
+      "epoch": 0.6178190574475404,
+      "grad_norm": 0.3857697546482086,
+      "learning_rate": 0.0009210643848248824,
+      "loss": 1.8008,
+      "step": 1347
+    },
+    {
+      "epoch": 0.6182777204449031,
+      "grad_norm": 0.23998984694480896,
+      "learning_rate": 0.0009209307854108026,
+      "loss": 1.1037,
+      "step": 1348
+    },
+    {
+      "epoch": 0.6187363834422658,
+      "grad_norm": 0.2030552476644516,
+      "learning_rate": 0.0009207970827393661,
+      "loss": 0.8847,
+      "step": 1349
+    },
+    {
+      "epoch": 0.6191950464396285,
+      "grad_norm": 0.17879889905452728,
+      "learning_rate": 0.0009206632768433711,
+      "loss": 0.938,
+      "step": 1350
+    },
+    {
+      "epoch": 0.6196537094369912,
+      "grad_norm": 0.30431312322616577,
+      "learning_rate": 0.0009205293677556413,
+      "loss": 1.2913,
+      "step": 1351
+    },
+    {
+      "epoch": 0.6201123724343539,
+      "grad_norm": 0.09835697710514069,
+      "learning_rate": 0.0009203953555090252,
+      "loss": 0.4719,
+      "step": 1352
+    },
+    {
+      "epoch": 0.6205710354317165,
+      "grad_norm": 0.3493775427341461,
+      "learning_rate": 0.0009202612401363972,
+      "loss": 1.7051,
+      "step": 1353
+    },
+    {
+      "epoch": 0.6210296984290792,
+      "grad_norm": 0.35240983963012695,
+      "learning_rate": 0.0009201270216706568,
+      "loss": 1.2707,
+      "step": 1354
+    },
+    {
+      "epoch": 0.6214883614264419,
+      "grad_norm": 0.08556913584470749,
+      "learning_rate": 0.0009199927001447287,
+      "loss": 0.4551,
+      "step": 1355
+    },
+    {
+      "epoch": 0.6219470244238046,
+      "grad_norm": 0.18923942744731903,
+      "learning_rate": 0.000919858275591563,
+      "loss": 0.9558,
+      "step": 1356
+    },
+    {
+      "epoch": 0.6224056874211673,
+      "grad_norm": 0.22347323596477509,
+      "learning_rate": 0.000919723748044135,
+      "loss": 1.0806,
+      "step": 1357
+    },
+    {
+      "epoch": 0.6228643504185299,
+      "grad_norm": 1.7624882459640503,
+      "learning_rate": 0.0009195891175354451,
+      "loss": 1.8876,
+      "step": 1358
+    },
+    {
+      "epoch": 0.6233230134158927,
+      "grad_norm": 0.20518071949481964,
+      "learning_rate": 0.0009194543840985193,
+      "loss": 0.9335,
+      "step": 1359
+    },
+    {
+      "epoch": 0.6237816764132553,
+      "grad_norm": 0.24788199365139008,
+      "learning_rate": 0.0009193195477664087,
+      "loss": 1.2626,
+      "step": 1360
+    },
+    {
+      "epoch": 0.6242403394106181,
+      "grad_norm": 0.1483655720949173,
+      "learning_rate": 0.0009191846085721896,
+      "loss": 0.8627,
+      "step": 1361
+    },
+    {
+      "epoch": 0.6246990024079807,
+      "grad_norm": 0.42687055468559265,
+      "learning_rate": 0.0009190495665489635,
+      "loss": 1.7659,
+      "step": 1362
+    },
+    {
+      "epoch": 0.6251576654053435,
+      "grad_norm": 0.23796336352825165,
+      "learning_rate": 0.0009189144217298571,
+      "loss": 0.9816,
+      "step": 1363
+    },
+    {
+      "epoch": 0.6256163284027061,
+      "grad_norm": 0.30603915452957153,
+      "learning_rate": 0.0009187791741480227,
+      "loss": 1.3668,
+      "step": 1364
+    },
+    {
+      "epoch": 0.6260749914000688,
+      "grad_norm": 0.2691842019557953,
+      "learning_rate": 0.0009186438238366373,
+      "loss": 1.4046,
+      "step": 1365
+    },
+    {
+      "epoch": 0.6265336543974315,
+      "grad_norm": 0.3278261125087738,
+      "learning_rate": 0.0009185083708289032,
+      "loss": 1.6694,
+      "step": 1366
+    },
+    {
+      "epoch": 0.6269923173947942,
+      "grad_norm": 0.319321870803833,
+      "learning_rate": 0.0009183728151580484,
+      "loss": 1.8997,
+      "step": 1367
+    },
+    {
+      "epoch": 0.6274509803921569,
+      "grad_norm": 0.2504161596298218,
+      "learning_rate": 0.0009182371568573252,
+      "loss": 1.1439,
+      "step": 1368
+    },
+    {
+      "epoch": 0.6279096433895196,
+      "grad_norm": 0.35240206122398376,
+      "learning_rate": 0.0009181013959600119,
+      "loss": 1.7751,
+      "step": 1369
+    },
+    {
+      "epoch": 0.6283683063868822,
+      "grad_norm": 0.30558404326438904,
+      "learning_rate": 0.0009179655324994114,
+      "loss": 1.4472,
+      "step": 1370
+    },
+    {
+      "epoch": 0.6288269693842449,
+      "grad_norm": 0.3027516305446625,
+      "learning_rate": 0.0009178295665088522,
+      "loss": 1.431,
+      "step": 1371
+    },
+    {
+      "epoch": 0.6292856323816076,
+      "grad_norm": 0.1590060144662857,
+      "learning_rate": 0.0009176934980216876,
+      "loss": 0.7231,
+      "step": 1372
+    },
+    {
+      "epoch": 0.6297442953789703,
+      "grad_norm": 0.280381441116333,
+      "learning_rate": 0.0009175573270712961,
+      "loss": 1.2524,
+      "step": 1373
+    },
+    {
+      "epoch": 0.630202958376333,
+      "grad_norm": 0.2787134051322937,
+      "learning_rate": 0.0009174210536910816,
+      "loss": 1.3318,
+      "step": 1374
+    },
+    {
+      "epoch": 0.6306616213736956,
+      "grad_norm": 0.24251703917980194,
+      "learning_rate": 0.0009172846779144729,
+      "loss": 1.301,
+      "step": 1375
+    },
+    {
+      "epoch": 0.6311202843710584,
+      "grad_norm": 0.20021305978298187,
+      "learning_rate": 0.0009171481997749239,
+      "loss": 1.0351,
+      "step": 1376
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 0.19602487981319427,
+      "learning_rate": 0.0009170116193059138,
+      "loss": 0.8918,
+      "step": 1377
+    },
+    {
+      "epoch": 0.6320376103657838,
+      "grad_norm": 0.2832145690917969,
+      "learning_rate": 0.0009168749365409466,
+      "loss": 1.3265,
+      "step": 1378
+    },
+    {
+      "epoch": 0.6324962733631464,
+      "grad_norm": 0.2144933044910431,
+      "learning_rate": 0.0009167381515135515,
+      "loss": 0.8716,
+      "step": 1379
+    },
+    {
+      "epoch": 0.6329549363605091,
+      "grad_norm": 0.17951545119285583,
+      "learning_rate": 0.0009166012642572832,
+      "loss": 0.7958,
+      "step": 1380
+    },
+    {
+      "epoch": 0.6334135993578718,
+      "grad_norm": 0.06861251592636108,
+      "learning_rate": 0.0009164642748057208,
+      "loss": 0.403,
+      "step": 1381
+    },
+    {
+      "epoch": 0.6338722623552345,
+      "grad_norm": 0.08268727362155914,
+      "learning_rate": 0.0009163271831924689,
+      "loss": 0.4848,
+      "step": 1382
+    },
+    {
+      "epoch": 0.6343309253525972,
+      "grad_norm": 0.26671648025512695,
+      "learning_rate": 0.0009161899894511572,
+      "loss": 1.2793,
+      "step": 1383
+    },
+    {
+      "epoch": 0.6347895883499599,
+      "grad_norm": 0.31374937295913696,
+      "learning_rate": 0.00091605269361544,
+      "loss": 1.6044,
+      "step": 1384
+    },
+    {
+      "epoch": 0.6352482513473225,
+      "grad_norm": 0.3432426452636719,
+      "learning_rate": 0.0009159152957189975,
+      "loss": 1.7935,
+      "step": 1385
+    },
+    {
+      "epoch": 0.6357069143446853,
+      "grad_norm": 0.3372509479522705,
+      "learning_rate": 0.0009157777957955337,
+      "loss": 1.8975,
+      "step": 1386
+    },
+    {
+      "epoch": 0.6361655773420479,
+      "grad_norm": 0.3484379053115845,
+      "learning_rate": 0.000915640193878779,
+      "loss": 1.6467,
+      "step": 1387
+    },
+    {
+      "epoch": 0.6366242403394106,
+      "grad_norm": 0.268187016248703,
+      "learning_rate": 0.0009155024900024877,
+      "loss": 1.3937,
+      "step": 1388
+    },
+    {
+      "epoch": 0.6370829033367733,
+      "grad_norm": 0.27969205379486084,
+      "learning_rate": 0.0009153646842004396,
+      "loss": 1.2155,
+      "step": 1389
+    },
+    {
+      "epoch": 0.637541566334136,
+      "grad_norm": 0.22392551600933075,
+      "learning_rate": 0.0009152267765064395,
+      "loss": 0.8904,
+      "step": 1390
+    },
+    {
+      "epoch": 0.6380002293314987,
+      "grad_norm": 0.3370068371295929,
+      "learning_rate": 0.0009150887669543173,
+      "loss": 1.496,
+      "step": 1391
+    },
+    {
+      "epoch": 0.6384588923288613,
+      "grad_norm": 0.4173247814178467,
+      "learning_rate": 0.0009149506555779277,
+      "loss": 2.1025,
+      "step": 1392
+    },
+    {
+      "epoch": 0.6389175553262241,
+      "grad_norm": 0.3257886469364166,
+      "learning_rate": 0.0009148124424111501,
+      "loss": 1.2627,
+      "step": 1393
+    },
+    {
+      "epoch": 0.6393762183235867,
+      "grad_norm": 0.34521573781967163,
+      "learning_rate": 0.0009146741274878896,
+      "loss": 1.6971,
+      "step": 1394
+    },
+    {
+      "epoch": 0.6398348813209495,
+      "grad_norm": 0.43050798773765564,
+      "learning_rate": 0.0009145357108420756,
+      "loss": 1.8064,
+      "step": 1395
+    },
+    {
+      "epoch": 0.6402935443183121,
+      "grad_norm": 0.31571662425994873,
+      "learning_rate": 0.0009143971925076629,
+      "loss": 1.3123,
+      "step": 1396
+    },
+    {
+      "epoch": 0.6407522073156748,
+      "grad_norm": 0.3214929401874542,
+      "learning_rate": 0.0009142585725186307,
+      "loss": 1.5576,
+      "step": 1397
+    },
+    {
+      "epoch": 0.6412108703130375,
+      "grad_norm": 0.3679821491241455,
+      "learning_rate": 0.0009141198509089838,
+      "loss": 1.677,
+      "step": 1398
+    },
+    {
+      "epoch": 0.6416695333104002,
+      "grad_norm": 0.3775405287742615,
+      "learning_rate": 0.0009139810277127516,
+      "loss": 1.8577,
+      "step": 1399
+    },
+    {
+      "epoch": 0.6421281963077629,
+      "grad_norm": 0.12690681219100952,
+      "learning_rate": 0.0009138421029639882,
+      "loss": 0.6524,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6425868593051256,
+      "grad_norm": 0.22937628626823425,
+      "learning_rate": 0.0009137030766967731,
+      "loss": 0.9138,
+      "step": 1401
+    },
+    {
+      "epoch": 0.6430455223024882,
+      "grad_norm": 0.29358240962028503,
+      "learning_rate": 0.0009135639489452103,
+      "loss": 1.4675,
+      "step": 1402
+    },
+    {
+      "epoch": 0.643504185299851,
+      "grad_norm": 0.30726784467697144,
+      "learning_rate": 0.0009134247197434288,
+      "loss": 1.5634,
+      "step": 1403
+    },
+    {
+      "epoch": 0.6439628482972136,
+      "grad_norm": 0.2912197411060333,
+      "learning_rate": 0.0009132853891255827,
+      "loss": 1.3088,
+      "step": 1404
+    },
+    {
+      "epoch": 0.6444215112945763,
+      "grad_norm": 0.2756345868110657,
+      "learning_rate": 0.0009131459571258507,
+      "loss": 1.5412,
+      "step": 1405
+    },
+    {
+      "epoch": 0.644880174291939,
+      "grad_norm": 0.41657257080078125,
+      "learning_rate": 0.0009130064237784364,
+      "loss": 1.8962,
+      "step": 1406
+    },
+    {
+      "epoch": 0.6453388372893016,
+      "grad_norm": 0.270457923412323,
+      "learning_rate": 0.0009128667891175685,
+      "loss": 1.441,
+      "step": 1407
+    },
+    {
+      "epoch": 0.6457975002866644,
+      "grad_norm": 0.3133867383003235,
+      "learning_rate": 0.0009127270531775003,
+      "loss": 1.7017,
+      "step": 1408
+    },
+    {
+      "epoch": 0.646256163284027,
+      "grad_norm": 0.44148820638656616,
+      "learning_rate": 0.00091258721599251,
+      "loss": 1.6197,
+      "step": 1409
+    },
+    {
+      "epoch": 0.6467148262813898,
+      "grad_norm": 0.33237019181251526,
+      "learning_rate": 0.0009124472775969006,
+      "loss": 1.6628,
+      "step": 1410
+    },
+    {
+      "epoch": 0.6471734892787524,
+      "grad_norm": 0.3485216796398163,
+      "learning_rate": 0.0009123072380250003,
+      "loss": 0.9962,
+      "step": 1411
+    },
+    {
+      "epoch": 0.6476321522761151,
+      "grad_norm": 0.25593242049217224,
+      "learning_rate": 0.0009121670973111616,
+      "loss": 1.3634,
+      "step": 1412
+    },
+    {
+      "epoch": 0.6480908152734778,
+      "grad_norm": 0.20914927124977112,
+      "learning_rate": 0.000912026855489762,
+      "loss": 0.9994,
+      "step": 1413
+    },
+    {
+      "epoch": 0.6485494782708405,
+      "grad_norm": 0.2085810899734497,
+      "learning_rate": 0.0009118865125952038,
+      "loss": 1.0074,
+      "step": 1414
+    },
+    {
+      "epoch": 0.6490081412682032,
+      "grad_norm": 0.1792818307876587,
+      "learning_rate": 0.0009117460686619143,
+      "loss": 0.774,
+      "step": 1415
+    },
+    {
+      "epoch": 0.6494668042655659,
+      "grad_norm": 0.24848847091197968,
+      "learning_rate": 0.0009116055237243454,
+      "loss": 1.1238,
+      "step": 1416
+    },
+    {
+      "epoch": 0.6499254672629285,
+      "grad_norm": 0.2472166121006012,
+      "learning_rate": 0.0009114648778169735,
+      "loss": 1.1925,
+      "step": 1417
+    },
+    {
+      "epoch": 0.6503841302602913,
+      "grad_norm": 0.30023932456970215,
+      "learning_rate": 0.0009113241309743003,
+      "loss": 1.6653,
+      "step": 1418
+    },
+    {
+      "epoch": 0.6508427932576539,
+      "grad_norm": 0.2693098783493042,
+      "learning_rate": 0.0009111832832308522,
+      "loss": 1.4279,
+      "step": 1419
+    },
+    {
+      "epoch": 0.6513014562550167,
+      "grad_norm": 0.2827019691467285,
+      "learning_rate": 0.0009110423346211797,
+      "loss": 1.5595,
+      "step": 1420
+    },
+    {
+      "epoch": 0.6517601192523793,
+      "grad_norm": 0.31124547123908997,
+      "learning_rate": 0.0009109012851798588,
+      "loss": 1.7781,
+      "step": 1421
+    },
+    {
+      "epoch": 0.6522187822497421,
+      "grad_norm": 0.23482073843479156,
+      "learning_rate": 0.0009107601349414899,
+      "loss": 1.1257,
+      "step": 1422
+    },
+    {
+      "epoch": 0.6526774452471047,
+      "grad_norm": 0.34504595398902893,
+      "learning_rate": 0.0009106188839406982,
+      "loss": 1.9086,
+      "step": 1423
+    },
+    {
+      "epoch": 0.6531361082444673,
+      "grad_norm": 0.1269492506980896,
+      "learning_rate": 0.0009104775322121334,
+      "loss": 0.7008,
+      "step": 1424
+    },
+    {
+      "epoch": 0.6535947712418301,
+      "grad_norm": 0.2349303960800171,
+      "learning_rate": 0.0009103360797904705,
+      "loss": 1.3405,
+      "step": 1425
+    },
+    {
+      "epoch": 0.6540534342391927,
+      "grad_norm": 0.27141550183296204,
+      "learning_rate": 0.0009101945267104084,
+      "loss": 1.4506,
+      "step": 1426
+    },
+    {
+      "epoch": 0.6545120972365555,
+      "grad_norm": 0.28015798330307007,
+      "learning_rate": 0.0009100528730066712,
+      "loss": 1.3356,
+      "step": 1427
+    },
+    {
+      "epoch": 0.6549707602339181,
+      "grad_norm": 0.07972334325313568,
+      "learning_rate": 0.0009099111187140078,
+      "loss": 0.5118,
+      "step": 1428
+    },
+    {
+      "epoch": 0.6554294232312808,
+      "grad_norm": 0.2610866129398346,
+      "learning_rate": 0.0009097692638671913,
+      "loss": 1.1424,
+      "step": 1429
+    },
+    {
+      "epoch": 0.6558880862286435,
+      "grad_norm": 0.46445217728614807,
+      "learning_rate": 0.0009096273085010197,
+      "loss": 2.2559,
+      "step": 1430
+    },
+    {
+      "epoch": 0.6563467492260062,
+      "grad_norm": 0.2951407730579376,
+      "learning_rate": 0.0009094852526503158,
+      "loss": 1.4451,
+      "step": 1431
+    },
+    {
+      "epoch": 0.6568054122233689,
+      "grad_norm": 0.11164001375436783,
+      "learning_rate": 0.0009093430963499269,
+      "loss": 0.6334,
+      "step": 1432
+    },
+    {
+      "epoch": 0.6572640752207316,
+      "grad_norm": 0.3511189818382263,
+      "learning_rate": 0.0009092008396347249,
+      "loss": 1.946,
+      "step": 1433
+    },
+    {
+      "epoch": 0.6577227382180942,
+      "grad_norm": 0.29423317313194275,
+      "learning_rate": 0.0009090584825396064,
+      "loss": 1.4645,
+      "step": 1434
+    },
+    {
+      "epoch": 0.658181401215457,
+      "grad_norm": 0.2262077033519745,
+      "learning_rate": 0.0009089160250994928,
+      "loss": 1.0862,
+      "step": 1435
+    },
+    {
+      "epoch": 0.6586400642128196,
+      "grad_norm": 0.2860635221004486,
+      "learning_rate": 0.0009087734673493298,
+      "loss": 1.3347,
+      "step": 1436
+    },
+    {
+      "epoch": 0.6590987272101824,
+      "grad_norm": 0.3139696717262268,
+      "learning_rate": 0.0009086308093240878,
+      "loss": 1.4883,
+      "step": 1437
+    },
+    {
+      "epoch": 0.659557390207545,
+      "grad_norm": 0.3125250041484833,
+      "learning_rate": 0.0009084880510587619,
+      "loss": 1.6454,
+      "step": 1438
+    },
+    {
+      "epoch": 0.6600160532049077,
+      "grad_norm": 0.2761354446411133,
+      "learning_rate": 0.0009083451925883716,
+      "loss": 1.4274,
+      "step": 1439
+    },
+    {
+      "epoch": 0.6604747162022704,
+      "grad_norm": 0.20550890266895294,
+      "learning_rate": 0.0009082022339479615,
+      "loss": 0.9324,
+      "step": 1440
+    },
+    {
+      "epoch": 0.660933379199633,
+      "grad_norm": 0.18471978604793549,
+      "learning_rate": 0.0009080591751726,
+      "loss": 0.9747,
+      "step": 1441
+    },
+    {
+      "epoch": 0.6613920421969958,
+      "grad_norm": 0.23342064023017883,
+      "learning_rate": 0.0009079160162973805,
+      "loss": 1.2667,
+      "step": 1442
+    },
+    {
+      "epoch": 0.6618507051943584,
+      "grad_norm": 0.4921042025089264,
+      "learning_rate": 0.0009077727573574211,
+      "loss": 2.2029,
+      "step": 1443
+    },
+    {
+      "epoch": 0.6623093681917211,
+      "grad_norm": 0.2813926935195923,
+      "learning_rate": 0.000907629398387864,
+      "loss": 1.328,
+      "step": 1444
+    },
+    {
+      "epoch": 0.6627680311890838,
+      "grad_norm": 0.18378609418869019,
+      "learning_rate": 0.0009074859394238763,
+      "loss": 0.8692,
+      "step": 1445
+    },
+    {
+      "epoch": 0.6632266941864465,
+      "grad_norm": 0.11556479334831238,
+      "learning_rate": 0.0009073423805006495,
+      "loss": 0.556,
+      "step": 1446
+    },
+    {
+      "epoch": 0.6636853571838092,
+      "grad_norm": 0.28978320956230164,
+      "learning_rate": 0.0009071987216533999,
+      "loss": 0.9043,
+      "step": 1447
+    },
+    {
+      "epoch": 0.6641440201811719,
+      "grad_norm": 0.41565844416618347,
+      "learning_rate": 0.0009070549629173677,
+      "loss": 1.9819,
+      "step": 1448
+    },
+    {
+      "epoch": 0.6646026831785345,
+      "grad_norm": 0.3662017285823822,
+      "learning_rate": 0.0009069111043278181,
+      "loss": 1.5942,
+      "step": 1449
+    },
+    {
+      "epoch": 0.6650613461758973,
+      "grad_norm": 0.31095772981643677,
+      "learning_rate": 0.0009067671459200406,
+      "loss": 1.3741,
+      "step": 1450
+    },
+    {
+      "epoch": 0.6655200091732599,
+      "grad_norm": 0.33298200368881226,
+      "learning_rate": 0.0009066230877293493,
+      "loss": 1.777,
+      "step": 1451
+    },
+    {
+      "epoch": 0.6659786721706227,
+      "grad_norm": 0.28431054949760437,
+      "learning_rate": 0.0009064789297910826,
+      "loss": 1.4286,
+      "step": 1452
+    },
+    {
+      "epoch": 0.6664373351679853,
+      "grad_norm": 0.32433438301086426,
+      "learning_rate": 0.0009063346721406037,
+      "loss": 1.8176,
+      "step": 1453
+    },
+    {
+      "epoch": 0.6668959981653481,
+      "grad_norm": 0.24182525277137756,
+      "learning_rate": 0.0009061903148132997,
+      "loss": 1.2091,
+      "step": 1454
+    },
+    {
+      "epoch": 0.6673546611627107,
+      "grad_norm": 0.21921652555465698,
+      "learning_rate": 0.0009060458578445829,
+      "loss": 1.0043,
+      "step": 1455
+    },
+    {
+      "epoch": 0.6678133241600734,
+      "grad_norm": 0.3465001881122589,
+      "learning_rate": 0.0009059013012698892,
+      "loss": 1.97,
+      "step": 1456
+    },
+    {
+      "epoch": 0.6682719871574361,
+      "grad_norm": 0.3193022310733795,
+      "learning_rate": 0.0009057566451246797,
+      "loss": 1.6008,
+      "step": 1457
+    },
+    {
+      "epoch": 0.6687306501547987,
+      "grad_norm": 0.261091947555542,
+      "learning_rate": 0.0009056118894444396,
+      "loss": 1.1853,
+      "step": 1458
+    },
+    {
+      "epoch": 0.6691893131521615,
+      "grad_norm": 0.21906504034996033,
+      "learning_rate": 0.0009054670342646782,
+      "loss": 0.846,
+      "step": 1459
+    },
+    {
+      "epoch": 0.6696479761495241,
+      "grad_norm": 0.37668684124946594,
+      "learning_rate": 0.0009053220796209298,
+      "loss": 1.3909,
+      "step": 1460
+    },
+    {
+      "epoch": 0.6701066391468868,
+      "grad_norm": 0.38679805397987366,
+      "learning_rate": 0.000905177025548753,
+      "loss": 1.8932,
+      "step": 1461
+    },
+    {
+      "epoch": 0.6705653021442495,
+      "grad_norm": 0.09460335969924927,
+      "learning_rate": 0.00090503187208373,
+      "loss": 0.5168,
+      "step": 1462
+    },
+    {
+      "epoch": 0.6710239651416122,
+      "grad_norm": 0.1873873472213745,
+      "learning_rate": 0.0009048866192614685,
+      "loss": 1.0623,
+      "step": 1463
+    },
+    {
+      "epoch": 0.6714826281389749,
+      "grad_norm": 0.34352460503578186,
+      "learning_rate": 0.0009047412671175999,
+      "loss": 1.9176,
+      "step": 1464
+    },
+    {
+      "epoch": 0.6719412911363376,
+      "grad_norm": 0.36623653769493103,
+      "learning_rate": 0.0009045958156877801,
+      "loss": 1.3452,
+      "step": 1465
+    },
+    {
+      "epoch": 0.6723999541337002,
+      "grad_norm": 0.2123580127954483,
+      "learning_rate": 0.0009044502650076895,
+      "loss": 0.9063,
+      "step": 1466
+    },
+    {
+      "epoch": 0.672858617131063,
+      "grad_norm": 0.3140578866004944,
+      "learning_rate": 0.0009043046151130326,
+      "loss": 1.7998,
+      "step": 1467
+    },
+    {
+      "epoch": 0.6733172801284256,
+      "grad_norm": 0.32382750511169434,
+      "learning_rate": 0.0009041588660395385,
+      "loss": 1.5197,
+      "step": 1468
+    },
+    {
+      "epoch": 0.6737759431257884,
+      "grad_norm": 0.3363771140575409,
+      "learning_rate": 0.0009040130178229604,
+      "loss": 2.1333,
+      "step": 1469
+    },
+    {
+      "epoch": 0.674234606123151,
+      "grad_norm": 0.3622763156890869,
+      "learning_rate": 0.0009038670704990759,
+      "loss": 1.6564,
+      "step": 1470
+    },
+    {
+      "epoch": 0.6746932691205138,
+      "grad_norm": 0.3316439092159271,
+      "learning_rate": 0.000903721024103687,
+      "loss": 1.965,
+      "step": 1471
+    },
+    {
+      "epoch": 0.6751519321178764,
+      "grad_norm": 0.28195494413375854,
+      "learning_rate": 0.0009035748786726199,
+      "loss": 1.2933,
+      "step": 1472
+    },
+    {
+      "epoch": 0.6756105951152391,
+      "grad_norm": 0.26023930311203003,
+      "learning_rate": 0.0009034286342417251,
+      "loss": 1.2734,
+      "step": 1473
+    },
+    {
+      "epoch": 0.6760692581126018,
+      "grad_norm": 0.39502447843551636,
+      "learning_rate": 0.0009032822908468775,
+      "loss": 2.2822,
+      "step": 1474
+    },
+    {
+      "epoch": 0.6765279211099644,
+      "grad_norm": 0.20976006984710693,
+      "learning_rate": 0.0009031358485239761,
+      "loss": 1.0251,
+      "step": 1475
+    },
+    {
+      "epoch": 0.6769865841073271,
+      "grad_norm": 0.3310341238975525,
+      "learning_rate": 0.0009029893073089443,
+      "loss": 1.9667,
+      "step": 1476
+    },
+    {
+      "epoch": 0.6774452471046898,
+      "grad_norm": 0.2633422911167145,
+      "learning_rate": 0.0009028426672377297,
+      "loss": 1.5052,
+      "step": 1477
+    },
+    {
+      "epoch": 0.6779039101020525,
+      "grad_norm": 0.20135562121868134,
+      "learning_rate": 0.0009026959283463044,
+      "loss": 1.1064,
+      "step": 1478
+    },
+    {
+      "epoch": 0.6783625730994152,
+      "grad_norm": 0.43766430020332336,
+      "learning_rate": 0.000902549090670664,
+      "loss": 2.6384,
+      "step": 1479
+    },
+    {
+      "epoch": 0.6788212360967779,
+      "grad_norm": 0.35967737436294556,
+      "learning_rate": 0.0009024021542468292,
+      "loss": 2.2778,
+      "step": 1480
+    },
+    {
+      "epoch": 0.6792798990941405,
+      "grad_norm": 0.2625736892223358,
+      "learning_rate": 0.0009022551191108446,
+      "loss": 1.4241,
+      "step": 1481
+    },
+    {
+      "epoch": 0.6797385620915033,
+      "grad_norm": 0.41545170545578003,
+      "learning_rate": 0.0009021079852987788,
+      "loss": 2.2012,
+      "step": 1482
+    },
+    {
+      "epoch": 0.6801972250888659,
+      "grad_norm": 0.2915220856666565,
+      "learning_rate": 0.0009019607528467249,
+      "loss": 1.3696,
+      "step": 1483
+    },
+    {
+      "epoch": 0.6806558880862287,
+      "grad_norm": 0.274554967880249,
+      "learning_rate": 0.0009018134217907999,
+      "loss": 1.3186,
+      "step": 1484
+    },
+    {
+      "epoch": 0.6811145510835913,
+      "grad_norm": 0.2678084671497345,
+      "learning_rate": 0.0009016659921671454,
+      "loss": 1.5616,
+      "step": 1485
+    },
+    {
+      "epoch": 0.6815732140809541,
+      "grad_norm": 0.22896720468997955,
+      "learning_rate": 0.000901518464011927,
+      "loss": 1.2181,
+      "step": 1486
+    },
+    {
+      "epoch": 0.6820318770783167,
+      "grad_norm": 0.3303893506526947,
+      "learning_rate": 0.0009013708373613341,
+      "loss": 1.9144,
+      "step": 1487
+    },
+    {
+      "epoch": 0.6824905400756794,
+      "grad_norm": 0.34536314010620117,
+      "learning_rate": 0.0009012231122515807,
+      "loss": 1.5248,
+      "step": 1488
+    },
+    {
+      "epoch": 0.6829492030730421,
+      "grad_norm": 0.2860594093799591,
+      "learning_rate": 0.0009010752887189051,
+      "loss": 1.5133,
+      "step": 1489
+    },
+    {
+      "epoch": 0.6834078660704048,
+      "grad_norm": 0.15731200575828552,
+      "learning_rate": 0.0009009273667995691,
+      "loss": 0.7388,
+      "step": 1490
+    },
+    {
+      "epoch": 0.6838665290677675,
+      "grad_norm": 0.2773778438568115,
+      "learning_rate": 0.0009007793465298593,
+      "loss": 1.5673,
+      "step": 1491
+    },
+    {
+      "epoch": 0.6843251920651301,
+      "grad_norm": 0.3588542342185974,
+      "learning_rate": 0.000900631227946086,
+      "loss": 1.9658,
+      "step": 1492
+    },
+    {
+      "epoch": 0.6847838550624928,
+      "grad_norm": 0.3078388273715973,
+      "learning_rate": 0.0009004830110845838,
+      "loss": 1.25,
+      "step": 1493
+    },
+    {
+      "epoch": 0.6852425180598555,
+      "grad_norm": 0.34084051847457886,
+      "learning_rate": 0.0009003346959817113,
+      "loss": 1.7067,
+      "step": 1494
+    },
+    {
+      "epoch": 0.6857011810572182,
+      "grad_norm": 0.34470683336257935,
+      "learning_rate": 0.0009001862826738514,
+      "loss": 1.8342,
+      "step": 1495
+    },
+    {
+      "epoch": 0.6861598440545809,
+      "grad_norm": 0.281382292509079,
+      "learning_rate": 0.0009000377711974109,
+      "loss": 1.2902,
+      "step": 1496
+    },
+    {
+      "epoch": 0.6866185070519436,
+      "grad_norm": 0.2875047028064728,
+      "learning_rate": 0.0008998891615888205,
+      "loss": 1.4417,
+      "step": 1497
+    },
+    {
+      "epoch": 0.6870771700493062,
+      "grad_norm": 0.37324872612953186,
+      "learning_rate": 0.0008997404538845355,
+      "loss": 1.9172,
+      "step": 1498
+    },
+    {
+      "epoch": 0.687535833046669,
+      "grad_norm": 0.35222485661506653,
+      "learning_rate": 0.0008995916481210349,
+      "loss": 1.7999,
+      "step": 1499
+    },
+    {
+      "epoch": 0.6879944960440316,
+      "grad_norm": 0.28432413935661316,
+      "learning_rate": 0.0008994427443348217,
+      "loss": 1.3091,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6884531590413944,
+      "grad_norm": 0.3234626054763794,
+      "learning_rate": 0.0008992937425624235,
+      "loss": 1.6951,
+      "step": 1501
+    },
+    {
+      "epoch": 0.688911822038757,
+      "grad_norm": 0.3846684396266937,
+      "learning_rate": 0.0008991446428403909,
+      "loss": 2.1714,
+      "step": 1502
+    },
+    {
+      "epoch": 0.6893704850361198,
+      "grad_norm": 0.3131919205188751,
+      "learning_rate": 0.0008989954452052995,
+      "loss": 1.3457,
+      "step": 1503
+    },
+    {
+      "epoch": 0.6898291480334824,
+      "grad_norm": 0.19373203814029694,
+      "learning_rate": 0.0008988461496937485,
+      "loss": 0.9649,
+      "step": 1504
+    },
+    {
+      "epoch": 0.6902878110308451,
+      "grad_norm": 0.41193410754203796,
+      "learning_rate": 0.0008986967563423612,
+      "loss": 1.9963,
+      "step": 1505
+    },
+    {
+      "epoch": 0.6907464740282078,
+      "grad_norm": 0.3604571223258972,
+      "learning_rate": 0.0008985472651877847,
+      "loss": 1.7517,
+      "step": 1506
+    },
+    {
+      "epoch": 0.6912051370255705,
+      "grad_norm": 0.3667939007282257,
+      "learning_rate": 0.0008983976762666905,
+      "loss": 1.6728,
+      "step": 1507
+    },
+    {
+      "epoch": 0.6916638000229332,
+      "grad_norm": 0.3759375214576721,
+      "learning_rate": 0.0008982479896157737,
+      "loss": 2.1001,
+      "step": 1508
+    },
+    {
+      "epoch": 0.6921224630202958,
+      "grad_norm": 0.37497055530548096,
+      "learning_rate": 0.0008980982052717534,
+      "loss": 1.7802,
+      "step": 1509
+    },
+    {
+      "epoch": 0.6925811260176585,
+      "grad_norm": 0.23320765793323517,
+      "learning_rate": 0.0008979483232713731,
+      "loss": 1.096,
+      "step": 1510
+    },
+    {
+      "epoch": 0.6930397890150212,
+      "grad_norm": 0.27966445684432983,
+      "learning_rate": 0.0008977983436513997,
+      "loss": 1.3906,
+      "step": 1511
+    },
+    {
+      "epoch": 0.6934984520123839,
+      "grad_norm": 0.2946244180202484,
+      "learning_rate": 0.0008976482664486241,
+      "loss": 1.5397,
+      "step": 1512
+    },
+    {
+      "epoch": 0.6939571150097466,
+      "grad_norm": 0.3360169231891632,
+      "learning_rate": 0.0008974980916998618,
+      "loss": 1.6011,
+      "step": 1513
+    },
+    {
+      "epoch": 0.6944157780071093,
+      "grad_norm": 0.28576546907424927,
+      "learning_rate": 0.0008973478194419515,
+      "loss": 1.4437,
+      "step": 1514
+    },
+    {
+      "epoch": 0.6948744410044719,
+      "grad_norm": 0.25533977150917053,
+      "learning_rate": 0.000897197449711756,
+      "loss": 1.0789,
+      "step": 1515
+    },
+    {
+      "epoch": 0.6953331040018347,
+      "grad_norm": 0.08754102140665054,
+      "learning_rate": 0.000897046982546162,
+      "loss": 0.4806,
+      "step": 1516
+    },
+    {
+      "epoch": 0.6957917669991973,
+      "grad_norm": 0.2931134104728699,
+      "learning_rate": 0.0008968964179820806,
+      "loss": 1.8093,
+      "step": 1517
+    },
+    {
+      "epoch": 0.6962504299965601,
+      "grad_norm": 0.10447783023118973,
+      "learning_rate": 0.0008967457560564459,
+      "loss": 0.5872,
+      "step": 1518
+    },
+    {
+      "epoch": 0.6967090929939227,
+      "grad_norm": 0.1723310351371765,
+      "learning_rate": 0.0008965949968062166,
+      "loss": 0.758,
+      "step": 1519
+    },
+    {
+      "epoch": 0.6971677559912854,
+      "grad_norm": 0.27590087056159973,
+      "learning_rate": 0.000896444140268375,
+      "loss": 1.2286,
+      "step": 1520
+    },
+    {
+      "epoch": 0.6976264189886481,
+      "grad_norm": 0.3477742671966553,
+      "learning_rate": 0.0008962931864799272,
+      "loss": 1.5517,
+      "step": 1521
+    },
+    {
+      "epoch": 0.6980850819860108,
+      "grad_norm": 0.34647393226623535,
+      "learning_rate": 0.0008961421354779036,
+      "loss": 1.5421,
+      "step": 1522
+    },
+    {
+      "epoch": 0.6985437449833735,
+      "grad_norm": 0.36403578519821167,
+      "learning_rate": 0.0008959909872993574,
+      "loss": 1.9447,
+      "step": 1523
+    },
+    {
+      "epoch": 0.6990024079807362,
+      "grad_norm": 0.33395373821258545,
+      "learning_rate": 0.0008958397419813671,
+      "loss": 1.6851,
+      "step": 1524
+    },
+    {
+      "epoch": 0.6994610709780988,
+      "grad_norm": 0.23508980870246887,
+      "learning_rate": 0.0008956883995610338,
+      "loss": 1.063,
+      "step": 1525
+    },
+    {
+      "epoch": 0.6999197339754615,
+      "grad_norm": 0.29683560132980347,
+      "learning_rate": 0.0008955369600754831,
+      "loss": 1.3033,
+      "step": 1526
+    },
+    {
+      "epoch": 0.7003783969728242,
+      "grad_norm": 0.238669291138649,
+      "learning_rate": 0.0008953854235618641,
+      "loss": 1.0421,
+      "step": 1527
+    },
+    {
+      "epoch": 0.7008370599701869,
+      "grad_norm": 0.31371814012527466,
+      "learning_rate": 0.00089523379005735,
+      "loss": 1.0029,
+      "step": 1528
+    },
+    {
+      "epoch": 0.7012957229675496,
+      "grad_norm": 0.2695467472076416,
+      "learning_rate": 0.0008950820595991371,
+      "loss": 1.3376,
+      "step": 1529
+    },
+    {
+      "epoch": 0.7017543859649122,
+      "grad_norm": 0.17910736799240112,
+      "learning_rate": 0.0008949302322244465,
+      "loss": 0.9264,
+      "step": 1530
+    },
+    {
+      "epoch": 0.702213048962275,
+      "grad_norm": 0.4682910144329071,
+      "learning_rate": 0.0008947783079705223,
+      "loss": 1.4216,
+      "step": 1531
+    },
+    {
+      "epoch": 0.7026717119596376,
+      "grad_norm": 0.37215444445610046,
+      "learning_rate": 0.0008946262868746327,
+      "loss": 2.3613,
+      "step": 1532
+    },
+    {
+      "epoch": 0.7031303749570004,
+      "grad_norm": 0.25320783257484436,
+      "learning_rate": 0.0008944741689740695,
+      "loss": 1.2519,
+      "step": 1533
+    },
+    {
+      "epoch": 0.703589037954363,
+      "grad_norm": 0.24361571669578552,
+      "learning_rate": 0.0008943219543061481,
+      "loss": 1.191,
+      "step": 1534
+    },
+    {
+      "epoch": 0.7040477009517258,
+      "grad_norm": 0.3469868004322052,
+      "learning_rate": 0.0008941696429082084,
+      "loss": 1.9496,
+      "step": 1535
+    },
+    {
+      "epoch": 0.7045063639490884,
+      "grad_norm": 0.3765385150909424,
+      "learning_rate": 0.0008940172348176132,
+      "loss": 2.0288,
+      "step": 1536
+    },
+    {
+      "epoch": 0.7049650269464511,
+      "grad_norm": 0.23416545987129211,
+      "learning_rate": 0.0008938647300717491,
+      "loss": 0.9575,
+      "step": 1537
+    },
+    {
+      "epoch": 0.7054236899438138,
+      "grad_norm": 0.3623865246772766,
+      "learning_rate": 0.0008937121287080268,
+      "loss": 1.8331,
+      "step": 1538
+    },
+    {
+      "epoch": 0.7058823529411765,
+      "grad_norm": 0.20962877571582794,
+      "learning_rate": 0.0008935594307638806,
+      "loss": 1.1107,
+      "step": 1539
+    },
+    {
+      "epoch": 0.7063410159385392,
+      "grad_norm": 0.24036376178264618,
+      "learning_rate": 0.0008934066362767684,
+      "loss": 1.1372,
+      "step": 1540
+    },
+    {
+      "epoch": 0.7067996789359019,
+      "grad_norm": 0.0797274112701416,
+      "learning_rate": 0.0008932537452841716,
+      "loss": 0.4638,
+      "step": 1541
+    },
+    {
+      "epoch": 0.7072583419332645,
+      "grad_norm": 0.32924991846084595,
+      "learning_rate": 0.0008931007578235957,
+      "loss": 1.4589,
+      "step": 1542
+    },
+    {
+      "epoch": 0.7077170049306272,
+      "grad_norm": 0.3090713322162628,
+      "learning_rate": 0.0008929476739325694,
+      "loss": 1.2155,
+      "step": 1543
+    },
+    {
+      "epoch": 0.7081756679279899,
+      "grad_norm": 0.32841405272483826,
+      "learning_rate": 0.0008927944936486454,
+      "loss": 1.3251,
+      "step": 1544
+    },
+    {
+      "epoch": 0.7086343309253526,
+      "grad_norm": 0.2854306697845459,
+      "learning_rate": 0.0008926412170093998,
+      "loss": 1.2567,
+      "step": 1545
+    },
+    {
+      "epoch": 0.7090929939227153,
+      "grad_norm": 0.3006875514984131,
+      "learning_rate": 0.0008924878440524326,
+      "loss": 1.7976,
+      "step": 1546
+    },
+    {
+      "epoch": 0.7095516569200779,
+      "grad_norm": 0.42627203464508057,
+      "learning_rate": 0.0008923343748153674,
+      "loss": 1.429,
+      "step": 1547
+    },
+    {
+      "epoch": 0.7100103199174407,
+      "grad_norm": 0.36591845750808716,
+      "learning_rate": 0.0008921808093358512,
+      "loss": 2.0798,
+      "step": 1548
+    },
+    {
+      "epoch": 0.7104689829148033,
+      "grad_norm": 0.2793489694595337,
+      "learning_rate": 0.0008920271476515547,
+      "loss": 1.7866,
+      "step": 1549
+    },
+    {
+      "epoch": 0.7109276459121661,
+      "grad_norm": 0.24239283800125122,
+      "learning_rate": 0.0008918733898001721,
+      "loss": 1.1033,
+      "step": 1550
+    },
+    {
+      "epoch": 0.7113863089095287,
+      "grad_norm": 0.42722949385643005,
+      "learning_rate": 0.0008917195358194214,
+      "loss": 1.3158,
+      "step": 1551
+    },
+    {
+      "epoch": 0.7118449719068914,
+      "grad_norm": 0.10515722632408142,
+      "learning_rate": 0.0008915655857470443,
+      "loss": 0.5974,
+      "step": 1552
+    },
+    {
+      "epoch": 0.7123036349042541,
+      "grad_norm": 0.3216477334499359,
+      "learning_rate": 0.0008914115396208056,
+      "loss": 1.8257,
+      "step": 1553
+    },
+    {
+      "epoch": 0.7127622979016168,
+      "grad_norm": 0.27405673265457153,
+      "learning_rate": 0.000891257397478494,
+      "loss": 1.3561,
+      "step": 1554
+    },
+    {
+      "epoch": 0.7132209608989795,
+      "grad_norm": 0.2392241358757019,
+      "learning_rate": 0.0008911031593579217,
+      "loss": 1.3283,
+      "step": 1555
+    },
+    {
+      "epoch": 0.7136796238963422,
+      "grad_norm": 0.3045576810836792,
+      "learning_rate": 0.0008909488252969244,
+      "loss": 1.3752,
+      "step": 1556
+    },
+    {
+      "epoch": 0.7141382868937048,
+      "grad_norm": 0.34698760509490967,
+      "learning_rate": 0.0008907943953333613,
+      "loss": 1.8883,
+      "step": 1557
+    },
+    {
+      "epoch": 0.7145969498910676,
+      "grad_norm": 0.3657462000846863,
+      "learning_rate": 0.0008906398695051153,
+      "loss": 2.2324,
+      "step": 1558
+    },
+    {
+      "epoch": 0.7150556128884302,
+      "grad_norm": 0.2741200029850006,
+      "learning_rate": 0.0008904852478500927,
+      "loss": 1.3587,
+      "step": 1559
+    },
+    {
+      "epoch": 0.7155142758857929,
+      "grad_norm": 0.3094598352909088,
+      "learning_rate": 0.0008903305304062232,
+      "loss": 1.7721,
+      "step": 1560
+    },
+    {
+      "epoch": 0.7159729388831556,
+      "grad_norm": 0.45374658703804016,
+      "learning_rate": 0.0008901757172114601,
+      "loss": 2.1138,
+      "step": 1561
+    },
+    {
+      "epoch": 0.7164316018805182,
+      "grad_norm": 0.6525793671607971,
+      "learning_rate": 0.0008900208083037804,
+      "loss": 1.7445,
+      "step": 1562
+    },
+    {
+      "epoch": 0.716890264877881,
+      "grad_norm": 0.39578354358673096,
+      "learning_rate": 0.0008898658037211842,
+      "loss": 1.2509,
+      "step": 1563
+    },
+    {
+      "epoch": 0.7173489278752436,
+      "grad_norm": 0.348197340965271,
+      "learning_rate": 0.0008897107035016952,
+      "loss": 1.8569,
+      "step": 1564
+    },
+    {
+      "epoch": 0.7178075908726064,
+      "grad_norm": 0.3598276972770691,
+      "learning_rate": 0.0008895555076833607,
+      "loss": 1.7867,
+      "step": 1565
+    },
+    {
+      "epoch": 0.718266253869969,
+      "grad_norm": 0.2803894579410553,
+      "learning_rate": 0.0008894002163042514,
+      "loss": 1.3177,
+      "step": 1566
+    },
+    {
+      "epoch": 0.7187249168673318,
+      "grad_norm": 0.3473854064941406,
+      "learning_rate": 0.0008892448294024612,
+      "loss": 1.8008,
+      "step": 1567
+    },
+    {
+      "epoch": 0.7191835798646944,
+      "grad_norm": 0.2351778894662857,
+      "learning_rate": 0.0008890893470161078,
+      "loss": 0.7619,
+      "step": 1568
+    },
+    {
+      "epoch": 0.7196422428620571,
+      "grad_norm": 0.3453526794910431,
+      "learning_rate": 0.0008889337691833321,
+      "loss": 1.9944,
+      "step": 1569
+    },
+    {
+      "epoch": 0.7201009058594198,
+      "grad_norm": 0.40439310669898987,
+      "learning_rate": 0.0008887780959422984,
+      "loss": 1.496,
+      "step": 1570
+    },
+    {
+      "epoch": 0.7205595688567825,
+      "grad_norm": 0.3145516514778137,
+      "learning_rate": 0.0008886223273311946,
+      "loss": 1.8382,
+      "step": 1571
+    },
+    {
+      "epoch": 0.7210182318541452,
+      "grad_norm": 0.33060380816459656,
+      "learning_rate": 0.0008884664633882317,
+      "loss": 1.4937,
+      "step": 1572
+    },
+    {
+      "epoch": 0.7214768948515079,
+      "grad_norm": 0.11799333989620209,
+      "learning_rate": 0.0008883105041516445,
+      "loss": 0.6943,
+      "step": 1573
+    },
+    {
+      "epoch": 0.7219355578488705,
+      "grad_norm": 0.3208516538143158,
+      "learning_rate": 0.0008881544496596907,
+      "loss": 1.3055,
+      "step": 1574
+    },
+    {
+      "epoch": 0.7223942208462333,
+      "grad_norm": 0.29803258180618286,
+      "learning_rate": 0.0008879982999506518,
+      "loss": 1.3234,
+      "step": 1575
+    },
+    {
+      "epoch": 0.7228528838435959,
+      "grad_norm": 0.2925770580768585,
+      "learning_rate": 0.000887842055062832,
+      "loss": 1.3806,
+      "step": 1576
+    },
+    {
+      "epoch": 0.7233115468409586,
+      "grad_norm": 0.3469628095626831,
+      "learning_rate": 0.0008876857150345598,
+      "loss": 1.9561,
+      "step": 1577
+    },
+    {
+      "epoch": 0.7237702098383213,
+      "grad_norm": 0.2605293393135071,
+      "learning_rate": 0.0008875292799041863,
+      "loss": 1.2745,
+      "step": 1578
+    },
+    {
+      "epoch": 0.7242288728356839,
+      "grad_norm": 0.16196803748607635,
+      "learning_rate": 0.0008873727497100862,
+      "loss": 0.758,
+      "step": 1579
+    },
+    {
+      "epoch": 0.7246875358330467,
+      "grad_norm": 0.35878610610961914,
+      "learning_rate": 0.0008872161244906576,
+      "loss": 1.5375,
+      "step": 1580
+    },
+    {
+      "epoch": 0.7251461988304093,
+      "grad_norm": 0.43217015266418457,
+      "learning_rate": 0.0008870594042843216,
+      "loss": 2.2036,
+      "step": 1581
+    },
+    {
+      "epoch": 0.7256048618277721,
+      "grad_norm": 0.42131683230400085,
+      "learning_rate": 0.0008869025891295228,
+      "loss": 0.9508,
+      "step": 1582
+    },
+    {
+      "epoch": 0.7260635248251347,
+      "grad_norm": 0.36816248297691345,
+      "learning_rate": 0.0008867456790647292,
+      "loss": 2.1968,
+      "step": 1583
+    },
+    {
+      "epoch": 0.7265221878224974,
+      "grad_norm": 0.30012619495391846,
+      "learning_rate": 0.0008865886741284321,
+      "loss": 1.7261,
+      "step": 1584
+    },
+    {
+      "epoch": 0.7269808508198601,
+      "grad_norm": 0.29164719581604004,
+      "learning_rate": 0.0008864315743591457,
+      "loss": 1.5405,
+      "step": 1585
+    },
+    {
+      "epoch": 0.7274395138172228,
+      "grad_norm": 0.3047489821910858,
+      "learning_rate": 0.0008862743797954078,
+      "loss": 1.7762,
+      "step": 1586
+    },
+    {
+      "epoch": 0.7278981768145855,
+      "grad_norm": 0.25507205724716187,
+      "learning_rate": 0.0008861170904757794,
+      "loss": 1.3858,
+      "step": 1587
+    },
+    {
+      "epoch": 0.7283568398119482,
+      "grad_norm": 0.21503114700317383,
+      "learning_rate": 0.0008859597064388445,
+      "loss": 1.0641,
+      "step": 1588
+    },
+    {
+      "epoch": 0.7288155028093108,
+      "grad_norm": 0.2100914567708969,
+      "learning_rate": 0.0008858022277232107,
+      "loss": 1.0745,
+      "step": 1589
+    },
+    {
+      "epoch": 0.7292741658066736,
+      "grad_norm": 0.20805799961090088,
+      "learning_rate": 0.0008856446543675088,
+      "loss": 1.007,
+      "step": 1590
+    },
+    {
+      "epoch": 0.7297328288040362,
+      "grad_norm": 0.34034040570259094,
+      "learning_rate": 0.0008854869864103925,
+      "loss": 1.625,
+      "step": 1591
+    },
+    {
+      "epoch": 0.730191491801399,
+      "grad_norm": 0.3298024535179138,
+      "learning_rate": 0.000885329223890539,
+      "loss": 1.6735,
+      "step": 1592
+    },
+    {
+      "epoch": 0.7306501547987616,
+      "grad_norm": 0.33766958117485046,
+      "learning_rate": 0.0008851713668466484,
+      "loss": 1.7477,
+      "step": 1593
+    },
+    {
+      "epoch": 0.7311088177961242,
+      "grad_norm": 0.27652397751808167,
+      "learning_rate": 0.0008850134153174443,
+      "loss": 0.8909,
+      "step": 1594
+    },
+    {
+      "epoch": 0.731567480793487,
+      "grad_norm": 0.40618276596069336,
+      "learning_rate": 0.0008848553693416734,
+      "loss": 1.8905,
+      "step": 1595
+    },
+    {
+      "epoch": 0.7320261437908496,
+      "grad_norm": 0.3916473090648651,
+      "learning_rate": 0.0008846972289581053,
+      "loss": 1.6833,
+      "step": 1596
+    },
+    {
+      "epoch": 0.7324848067882124,
+      "grad_norm": 0.3562237024307251,
+      "learning_rate": 0.0008845389942055333,
+      "loss": 1.2002,
+      "step": 1597
+    },
+    {
+      "epoch": 0.732943469785575,
+      "grad_norm": 0.2905832827091217,
+      "learning_rate": 0.0008843806651227733,
+      "loss": 1.2439,
+      "step": 1598
+    },
+    {
+      "epoch": 0.7334021327829378,
+      "grad_norm": 0.2873367369174957,
+      "learning_rate": 0.0008842222417486646,
+      "loss": 1.3354,
+      "step": 1599
+    },
+    {
+      "epoch": 0.7338607957803004,
+      "grad_norm": 0.26076123118400574,
+      "learning_rate": 0.0008840637241220696,
+      "loss": 1.2548,
+      "step": 1600
+    },
+    {
+      "epoch": 0.7343194587776631,
+      "grad_norm": 0.2180601954460144,
+      "learning_rate": 0.0008839051122818737,
+      "loss": 1.0288,
+      "step": 1601
+    },
+    {
+      "epoch": 0.7347781217750258,
+      "grad_norm": 0.3309728801250458,
+      "learning_rate": 0.000883746406266986,
+      "loss": 1.4196,
+      "step": 1602
+    },
+    {
+      "epoch": 0.7352367847723885,
+      "grad_norm": 0.403773695230484,
+      "learning_rate": 0.0008835876061163377,
+      "loss": 2.1072,
+      "step": 1603
+    },
+    {
+      "epoch": 0.7356954477697512,
+      "grad_norm": 0.2361578643321991,
+      "learning_rate": 0.0008834287118688837,
+      "loss": 1.1724,
+      "step": 1604
+    },
+    {
+      "epoch": 0.7361541107671139,
+      "grad_norm": 0.3139882981777191,
+      "learning_rate": 0.0008832697235636023,
+      "loss": 1.5528,
+      "step": 1605
+    },
+    {
+      "epoch": 0.7366127737644765,
+      "grad_norm": 0.32773545384407043,
+      "learning_rate": 0.0008831106412394938,
+      "loss": 1.7954,
+      "step": 1606
+    },
+    {
+      "epoch": 0.7370714367618393,
+      "grad_norm": 0.36932218074798584,
+      "learning_rate": 0.0008829514649355829,
+      "loss": 2.0931,
+      "step": 1607
+    },
+    {
+      "epoch": 0.7375300997592019,
+      "grad_norm": 0.3764024078845978,
+      "learning_rate": 0.0008827921946909164,
+      "loss": 1.7544,
+      "step": 1608
+    },
+    {
+      "epoch": 0.7379887627565647,
+      "grad_norm": 0.3229760527610779,
+      "learning_rate": 0.0008826328305445644,
+      "loss": 1.8318,
+      "step": 1609
+    },
+    {
+      "epoch": 0.7384474257539273,
+      "grad_norm": 0.26198244094848633,
+      "learning_rate": 0.0008824733725356202,
+      "loss": 1.2097,
+      "step": 1610
+    },
+    {
+      "epoch": 0.7389060887512899,
+      "grad_norm": 0.24832715094089508,
+      "learning_rate": 0.0008823138207031999,
+      "loss": 1.4565,
+      "step": 1611
+    },
+    {
+      "epoch": 0.7393647517486527,
+      "grad_norm": 0.29865747690200806,
+      "learning_rate": 0.0008821541750864428,
+      "loss": 1.2469,
+      "step": 1612
+    },
+    {
+      "epoch": 0.7398234147460153,
+      "grad_norm": 0.3033508062362671,
+      "learning_rate": 0.0008819944357245111,
+      "loss": 1.3326,
+      "step": 1613
+    },
+    {
+      "epoch": 0.7402820777433781,
+      "grad_norm": 0.2659357190132141,
+      "learning_rate": 0.0008818346026565897,
+      "loss": 1.2563,
+      "step": 1614
+    },
+    {
+      "epoch": 0.7407407407407407,
+      "grad_norm": 0.3529365360736847,
+      "learning_rate": 0.0008816746759218874,
+      "loss": 1.9199,
+      "step": 1615
+    },
+    {
+      "epoch": 0.7411994037381034,
+      "grad_norm": 0.20898482203483582,
+      "learning_rate": 0.0008815146555596351,
+      "loss": 0.8429,
+      "step": 1616
+    },
+    {
+      "epoch": 0.7416580667354661,
+      "grad_norm": 0.39400166273117065,
+      "learning_rate": 0.0008813545416090869,
+      "loss": 1.9886,
+      "step": 1617
+    },
+    {
+      "epoch": 0.7421167297328288,
+      "grad_norm": 0.2804252803325653,
+      "learning_rate": 0.0008811943341095199,
+      "loss": 1.4959,
+      "step": 1618
+    },
+    {
+      "epoch": 0.7425753927301915,
+      "grad_norm": 0.24878603219985962,
+      "learning_rate": 0.0008810340331002341,
+      "loss": 1.1033,
+      "step": 1619
+    },
+    {
+      "epoch": 0.7430340557275542,
+      "grad_norm": 0.3392964005470276,
+      "learning_rate": 0.0008808736386205527,
+      "loss": 1.5595,
+      "step": 1620
+    },
+    {
+      "epoch": 0.7434927187249168,
+      "grad_norm": 0.25439828634262085,
+      "learning_rate": 0.0008807131507098213,
+      "loss": 1.127,
+      "step": 1621
+    },
+    {
+      "epoch": 0.7439513817222796,
+      "grad_norm": 0.2508736252784729,
+      "learning_rate": 0.0008805525694074093,
+      "loss": 1.2366,
+      "step": 1622
+    },
+    {
+      "epoch": 0.7444100447196422,
+      "grad_norm": 0.21780788898468018,
+      "learning_rate": 0.0008803918947527079,
+      "loss": 1.1689,
+      "step": 1623
+    },
+    {
+      "epoch": 0.744868707717005,
+      "grad_norm": 0.22065167129039764,
+      "learning_rate": 0.000880231126785132,
+      "loss": 0.8861,
+      "step": 1624
+    },
+    {
+      "epoch": 0.7453273707143676,
+      "grad_norm": 0.3314412236213684,
+      "learning_rate": 0.000880070265544119,
+      "loss": 1.9618,
+      "step": 1625
+    },
+    {
+      "epoch": 0.7457860337117304,
+      "grad_norm": 0.2850133180618286,
+      "learning_rate": 0.0008799093110691294,
+      "loss": 1.5959,
+      "step": 1626
+    },
+    {
+      "epoch": 0.746244696709093,
+      "grad_norm": 0.3236753046512604,
+      "learning_rate": 0.0008797482633996466,
+      "loss": 1.6779,
+      "step": 1627
+    },
+    {
+      "epoch": 0.7467033597064556,
+      "grad_norm": 0.29706433415412903,
+      "learning_rate": 0.0008795871225751766,
+      "loss": 1.6093,
+      "step": 1628
+    },
+    {
+      "epoch": 0.7471620227038184,
+      "grad_norm": 0.3429460823535919,
+      "learning_rate": 0.0008794258886352485,
+      "loss": 1.5433,
+      "step": 1629
+    },
+    {
+      "epoch": 0.747620685701181,
+      "grad_norm": 0.32680225372314453,
+      "learning_rate": 0.0008792645616194141,
+      "loss": 1.8085,
+      "step": 1630
+    },
+    {
+      "epoch": 0.7480793486985438,
+      "grad_norm": 0.2640790641307831,
+      "learning_rate": 0.0008791031415672482,
+      "loss": 1.277,
+      "step": 1631
+    },
+    {
+      "epoch": 0.7485380116959064,
+      "grad_norm": 0.3999061584472656,
+      "learning_rate": 0.000878941628518348,
+      "loss": 1.9884,
+      "step": 1632
+    },
+    {
+      "epoch": 0.7489966746932691,
+      "grad_norm": 0.28930339217185974,
+      "learning_rate": 0.0008787800225123341,
+      "loss": 1.2634,
+      "step": 1633
+    },
+    {
+      "epoch": 0.7494553376906318,
+      "grad_norm": 0.25914284586906433,
+      "learning_rate": 0.0008786183235888497,
+      "loss": 1.2084,
+      "step": 1634
+    },
+    {
+      "epoch": 0.7499140006879945,
+      "grad_norm": 0.3261665403842926,
+      "learning_rate": 0.0008784565317875604,
+      "loss": 1.5663,
+      "step": 1635
+    },
+    {
+      "epoch": 0.7503726636853572,
+      "grad_norm": 0.23450294137001038,
+      "learning_rate": 0.000878294647148155,
+      "loss": 1.1509,
+      "step": 1636
+    },
+    {
+      "epoch": 0.7508313266827199,
+      "grad_norm": 0.2073303908109665,
+      "learning_rate": 0.000878132669710345,
+      "loss": 0.9553,
+      "step": 1637
+    },
+    {
+      "epoch": 0.7512899896800825,
+      "grad_norm": 0.18613280355930328,
+      "learning_rate": 0.0008779705995138647,
+      "loss": 0.8871,
+      "step": 1638
+    },
+    {
+      "epoch": 0.7517486526774453,
+      "grad_norm": 0.3788055181503296,
+      "learning_rate": 0.000877808436598471,
+      "loss": 1.5571,
+      "step": 1639
+    },
+    {
+      "epoch": 0.7522073156748079,
+      "grad_norm": 0.2849988639354706,
+      "learning_rate": 0.0008776461810039437,
+      "loss": 1.4867,
+      "step": 1640
+    },
+    {
+      "epoch": 0.7526659786721707,
+      "grad_norm": 0.3063529133796692,
+      "learning_rate": 0.0008774838327700852,
+      "loss": 1.3499,
+      "step": 1641
+    },
+    {
+      "epoch": 0.7531246416695333,
+      "grad_norm": 0.35424184799194336,
+      "learning_rate": 0.0008773213919367206,
+      "loss": 1.7198,
+      "step": 1642
+    },
+    {
+      "epoch": 0.753583304666896,
+      "grad_norm": 0.32375872135162354,
+      "learning_rate": 0.0008771588585436982,
+      "loss": 1.8474,
+      "step": 1643
+    },
+    {
+      "epoch": 0.7540419676642587,
+      "grad_norm": 0.3555268943309784,
+      "learning_rate": 0.0008769962326308882,
+      "loss": 1.7086,
+      "step": 1644
+    },
+    {
+      "epoch": 0.7545006306616213,
+      "grad_norm": 0.32062697410583496,
+      "learning_rate": 0.000876833514238184,
+      "loss": 1.4244,
+      "step": 1645
+    },
+    {
+      "epoch": 0.7549592936589841,
+      "grad_norm": 0.5851942896842957,
+      "learning_rate": 0.0008766707034055017,
+      "loss": 1.8285,
+      "step": 1646
+    },
+    {
+      "epoch": 0.7554179566563467,
+      "grad_norm": 0.39790624380111694,
+      "learning_rate": 0.0008765078001727799,
+      "loss": 1.5838,
+      "step": 1647
+    },
+    {
+      "epoch": 0.7558766196537094,
+      "grad_norm": 0.22544077038764954,
+      "learning_rate": 0.00087634480457998,
+      "loss": 1.0953,
+      "step": 1648
+    },
+    {
+      "epoch": 0.7563352826510721,
+      "grad_norm": 0.4409390985965729,
+      "learning_rate": 0.000876181716667086,
+      "loss": 2.1729,
+      "step": 1649
+    },
+    {
+      "epoch": 0.7567939456484348,
+      "grad_norm": 0.09505387395620346,
+      "learning_rate": 0.0008760185364741045,
+      "loss": 0.6027,
+      "step": 1650
+    },
+    {
+      "epoch": 0.7572526086457975,
+      "grad_norm": 0.18144452571868896,
+      "learning_rate": 0.0008758552640410647,
+      "loss": 0.7814,
+      "step": 1651
+    },
+    {
+      "epoch": 0.7577112716431602,
+      "grad_norm": 0.25609463453292847,
+      "learning_rate": 0.0008756918994080184,
+      "loss": 1.2465,
+      "step": 1652
+    },
+    {
+      "epoch": 0.7581699346405228,
+      "grad_norm": 0.3209281265735626,
+      "learning_rate": 0.0008755284426150405,
+      "loss": 1.327,
+      "step": 1653
+    },
+    {
+      "epoch": 0.7586285976378856,
+      "grad_norm": 0.28706008195877075,
+      "learning_rate": 0.0008753648937022278,
+      "loss": 1.5922,
+      "step": 1654
+    },
+    {
+      "epoch": 0.7590872606352482,
+      "grad_norm": 0.32143130898475647,
+      "learning_rate": 0.0008752012527097003,
+      "loss": 1.2776,
+      "step": 1655
+    },
+    {
+      "epoch": 0.759545923632611,
+      "grad_norm": 0.29344642162323,
+      "learning_rate": 0.0008750375196776002,
+      "loss": 1.5062,
+      "step": 1656
+    },
+    {
+      "epoch": 0.7600045866299736,
+      "grad_norm": 0.38162222504615784,
+      "learning_rate": 0.0008748736946460922,
+      "loss": 1.892,
+      "step": 1657
+    },
+    {
+      "epoch": 0.7604632496273364,
+      "grad_norm": 0.28159940242767334,
+      "learning_rate": 0.0008747097776553639,
+      "loss": 1.3286,
+      "step": 1658
+    },
+    {
+      "epoch": 0.760921912624699,
+      "grad_norm": 0.09235129505395889,
+      "learning_rate": 0.0008745457687456255,
+      "loss": 0.5844,
+      "step": 1659
+    },
+    {
+      "epoch": 0.7613805756220617,
+      "grad_norm": 0.3361958861351013,
+      "learning_rate": 0.0008743816679571094,
+      "loss": 1.7726,
+      "step": 1660
+    },
+    {
+      "epoch": 0.7618392386194244,
+      "grad_norm": 0.31122922897338867,
+      "learning_rate": 0.0008742174753300707,
+      "loss": 1.3808,
+      "step": 1661
+    },
+    {
+      "epoch": 0.762297901616787,
+      "grad_norm": 0.3450722396373749,
+      "learning_rate": 0.000874053190904787,
+      "loss": 1.8993,
+      "step": 1662
+    },
+    {
+      "epoch": 0.7627565646141498,
+      "grad_norm": 0.3466764986515045,
+      "learning_rate": 0.0008738888147215584,
+      "loss": 1.9905,
+      "step": 1663
+    },
+    {
+      "epoch": 0.7632152276115124,
+      "grad_norm": 0.13895724713802338,
+      "learning_rate": 0.0008737243468207079,
+      "loss": 0.6572,
+      "step": 1664
+    },
+    {
+      "epoch": 0.7636738906088751,
+      "grad_norm": 0.32592159509658813,
+      "learning_rate": 0.0008735597872425804,
+      "loss": 1.7026,
+      "step": 1665
+    },
+    {
+      "epoch": 0.7641325536062378,
+      "grad_norm": 0.2743968367576599,
+      "learning_rate": 0.0008733951360275434,
+      "loss": 1.1798,
+      "step": 1666
+    },
+    {
+      "epoch": 0.7645912166036005,
+      "grad_norm": 0.1800556182861328,
+      "learning_rate": 0.0008732303932159873,
+      "loss": 0.8318,
+      "step": 1667
+    },
+    {
+      "epoch": 0.7650498796009632,
+      "grad_norm": 0.35121503472328186,
+      "learning_rate": 0.0008730655588483247,
+      "loss": 1.9238,
+      "step": 1668
+    },
+    {
+      "epoch": 0.7655085425983259,
+      "grad_norm": 0.31581225991249084,
+      "learning_rate": 0.0008729006329649906,
+      "loss": 1.5878,
+      "step": 1669
+    },
+    {
+      "epoch": 0.7659672055956885,
+      "grad_norm": 0.34449562430381775,
+      "learning_rate": 0.0008727356156064424,
+      "loss": 1.348,
+      "step": 1670
+    },
+    {
+      "epoch": 0.7664258685930513,
+      "grad_norm": 0.2677781581878662,
+      "learning_rate": 0.0008725705068131599,
+      "loss": 1.2385,
+      "step": 1671
+    },
+    {
+      "epoch": 0.7668845315904139,
+      "grad_norm": 0.3324863314628601,
+      "learning_rate": 0.0008724053066256461,
+      "loss": 1.827,
+      "step": 1672
+    },
+    {
+      "epoch": 0.7673431945877767,
+      "grad_norm": 0.32999104261398315,
+      "learning_rate": 0.0008722400150844252,
+      "loss": 1.4756,
+      "step": 1673
+    },
+    {
+      "epoch": 0.7678018575851393,
+      "grad_norm": 0.15800195932388306,
+      "learning_rate": 0.0008720746322300447,
+      "loss": 0.7594,
+      "step": 1674
+    },
+    {
+      "epoch": 0.768260520582502,
+      "grad_norm": 0.3041391968727112,
+      "learning_rate": 0.0008719091581030741,
+      "loss": 1.4358,
+      "step": 1675
+    },
+    {
+      "epoch": 0.7687191835798647,
+      "grad_norm": 0.3658222258090973,
+      "learning_rate": 0.0008717435927441053,
+      "loss": 1.684,
+      "step": 1676
+    },
+    {
+      "epoch": 0.7691778465772274,
+      "grad_norm": 0.3466409146785736,
+      "learning_rate": 0.0008715779361937528,
+      "loss": 1.8845,
+      "step": 1677
+    },
+    {
+      "epoch": 0.7696365095745901,
+      "grad_norm": 0.32669052481651306,
+      "learning_rate": 0.0008714121884926536,
+      "loss": 1.6675,
+      "step": 1678
+    },
+    {
+      "epoch": 0.7700951725719528,
+      "grad_norm": 0.35387712717056274,
+      "learning_rate": 0.0008712463496814662,
+      "loss": 1.9104,
+      "step": 1679
+    },
+    {
+      "epoch": 0.7705538355693154,
+      "grad_norm": 0.3723506033420563,
+      "learning_rate": 0.0008710804198008727,
+      "loss": 1.8469,
+      "step": 1680
+    },
+    {
+      "epoch": 0.7710124985666781,
+      "grad_norm": 0.19466747343540192,
+      "learning_rate": 0.0008709143988915763,
+      "loss": 0.8387,
+      "step": 1681
+    },
+    {
+      "epoch": 0.7714711615640408,
+      "grad_norm": 0.33413955569267273,
+      "learning_rate": 0.0008707482869943035,
+      "loss": 1.6544,
+      "step": 1682
+    },
+    {
+      "epoch": 0.7719298245614035,
+      "grad_norm": 0.29037922620773315,
+      "learning_rate": 0.0008705820841498029,
+      "loss": 1.5114,
+      "step": 1683
+    },
+    {
+      "epoch": 0.7723884875587662,
+      "grad_norm": 0.3319530785083771,
+      "learning_rate": 0.0008704157903988448,
+      "loss": 1.61,
+      "step": 1684
+    },
+    {
+      "epoch": 0.7728471505561288,
+      "grad_norm": 0.37831956148147583,
+      "learning_rate": 0.0008702494057822224,
+      "loss": 1.8689,
+      "step": 1685
+    },
+    {
+      "epoch": 0.7733058135534916,
+      "grad_norm": 0.24735870957374573,
+      "learning_rate": 0.0008700829303407514,
+      "loss": 1.0308,
+      "step": 1686
+    },
+    {
+      "epoch": 0.7737644765508542,
+      "grad_norm": 0.09444022178649902,
+      "learning_rate": 0.000869916364115269,
+      "loss": 0.5718,
+      "step": 1687
+    },
+    {
+      "epoch": 0.774223139548217,
+      "grad_norm": 0.4078546166419983,
+      "learning_rate": 0.0008697497071466351,
+      "loss": 1.4986,
+      "step": 1688
+    },
+    {
+      "epoch": 0.7746818025455796,
+      "grad_norm": 0.3958517611026764,
+      "learning_rate": 0.0008695829594757323,
+      "loss": 1.8461,
+      "step": 1689
+    },
+    {
+      "epoch": 0.7751404655429424,
+      "grad_norm": 0.2555053234100342,
+      "learning_rate": 0.0008694161211434645,
+      "loss": 1.2473,
+      "step": 1690
+    },
+    {
+      "epoch": 0.775599128540305,
+      "grad_norm": 0.30251985788345337,
+      "learning_rate": 0.0008692491921907586,
+      "loss": 1.225,
+      "step": 1691
+    },
+    {
+      "epoch": 0.7760577915376677,
+      "grad_norm": 0.2819230258464813,
+      "learning_rate": 0.0008690821726585634,
+      "loss": 1.0078,
+      "step": 1692
+    },
+    {
+      "epoch": 0.7765164545350304,
+      "grad_norm": 0.32672372460365295,
+      "learning_rate": 0.0008689150625878501,
+      "loss": 1.4543,
+      "step": 1693
+    },
+    {
+      "epoch": 0.7769751175323931,
+      "grad_norm": 0.282443642616272,
+      "learning_rate": 0.000868747862019612,
+      "loss": 1.412,
+      "step": 1694
+    },
+    {
+      "epoch": 0.7774337805297558,
+      "grad_norm": 0.33812659978866577,
+      "learning_rate": 0.0008685805709948644,
+      "loss": 1.8073,
+      "step": 1695
+    },
+    {
+      "epoch": 0.7778924435271185,
+      "grad_norm": 0.29638344049453735,
+      "learning_rate": 0.0008684131895546453,
+      "loss": 1.5007,
+      "step": 1696
+    },
+    {
+      "epoch": 0.7783511065244811,
+      "grad_norm": 0.36412540078163147,
+      "learning_rate": 0.0008682457177400141,
+      "loss": 1.8148,
+      "step": 1697
+    },
+    {
+      "epoch": 0.7788097695218438,
+      "grad_norm": 0.1184324324131012,
+      "learning_rate": 0.0008680781555920533,
+      "loss": 0.6136,
+      "step": 1698
+    },
+    {
+      "epoch": 0.7792684325192065,
+      "grad_norm": 0.32119160890579224,
+      "learning_rate": 0.000867910503151867,
+      "loss": 1.8235,
+      "step": 1699
+    },
+    {
+      "epoch": 0.7797270955165692,
+      "grad_norm": 0.36955249309539795,
+      "learning_rate": 0.0008677427604605816,
+      "loss": 2.0277,
+      "step": 1700
+    },
+    {
+      "epoch": 0.7801857585139319,
+      "grad_norm": 0.3423328697681427,
+      "learning_rate": 0.0008675749275593454,
+      "loss": 1.6159,
+      "step": 1701
+    },
+    {
+      "epoch": 0.7806444215112945,
+      "grad_norm": 0.2848832905292511,
+      "learning_rate": 0.000867407004489329,
+      "loss": 1.5393,
+      "step": 1702
+    },
+    {
+      "epoch": 0.7811030845086573,
+      "grad_norm": 0.31208136677742004,
+      "learning_rate": 0.0008672389912917253,
+      "loss": 1.3461,
+      "step": 1703
+    },
+    {
+      "epoch": 0.7815617475060199,
+      "grad_norm": 0.35254040360450745,
+      "learning_rate": 0.0008670708880077492,
+      "loss": 1.6527,
+      "step": 1704
+    },
+    {
+      "epoch": 0.7820204105033827,
+      "grad_norm": 0.2813601493835449,
+      "learning_rate": 0.0008669026946786375,
+      "loss": 1.372,
+      "step": 1705
+    },
+    {
+      "epoch": 0.7824790735007453,
+      "grad_norm": 0.29800623655319214,
+      "learning_rate": 0.0008667344113456495,
+      "loss": 1.4514,
+      "step": 1706
+    },
+    {
+      "epoch": 0.782937736498108,
+      "grad_norm": 0.3098766505718231,
+      "learning_rate": 0.0008665660380500657,
+      "loss": 1.6478,
+      "step": 1707
+    },
+    {
+      "epoch": 0.7833963994954707,
+      "grad_norm": 0.4111390709877014,
+      "learning_rate": 0.00086639757483319,
+      "loss": 2.3545,
+      "step": 1708
+    },
+    {
+      "epoch": 0.7838550624928334,
+      "grad_norm": 0.2358768880367279,
+      "learning_rate": 0.0008662290217363474,
+      "loss": 0.8783,
+      "step": 1709
+    },
+    {
+      "epoch": 0.7843137254901961,
+      "grad_norm": 0.3664570152759552,
+      "learning_rate": 0.0008660603788008847,
+      "loss": 2.2107,
+      "step": 1710
+    },
+    {
+      "epoch": 0.7847723884875588,
+      "grad_norm": 0.22186599671840668,
+      "learning_rate": 0.0008658916460681721,
+      "loss": 1.1024,
+      "step": 1711
+    },
+    {
+      "epoch": 0.7852310514849214,
+      "grad_norm": 0.36577606201171875,
+      "learning_rate": 0.0008657228235796002,
+      "loss": 1.7358,
+      "step": 1712
+    },
+    {
+      "epoch": 0.7856897144822842,
+      "grad_norm": 0.2555186450481415,
+      "learning_rate": 0.0008655539113765828,
+      "loss": 0.8061,
+      "step": 1713
+    },
+    {
+      "epoch": 0.7861483774796468,
+      "grad_norm": 0.24842973053455353,
+      "learning_rate": 0.0008653849095005551,
+      "loss": 0.882,
+      "step": 1714
+    },
+    {
+      "epoch": 0.7866070404770095,
+      "grad_norm": 0.37208500504493713,
+      "learning_rate": 0.0008652158179929746,
+      "loss": 1.7593,
+      "step": 1715
+    },
+    {
+      "epoch": 0.7870657034743722,
+      "grad_norm": 0.3234862685203552,
+      "learning_rate": 0.0008650466368953206,
+      "loss": 1.5933,
+      "step": 1716
+    },
+    {
+      "epoch": 0.7875243664717348,
+      "grad_norm": 0.36344775557518005,
+      "learning_rate": 0.0008648773662490944,
+      "loss": 1.6957,
+      "step": 1717
+    },
+    {
+      "epoch": 0.7879830294690976,
+      "grad_norm": 0.29362058639526367,
+      "learning_rate": 0.0008647080060958194,
+      "loss": 1.246,
+      "step": 1718
+    },
+    {
+      "epoch": 0.7884416924664602,
+      "grad_norm": 0.48402926325798035,
+      "learning_rate": 0.0008645385564770409,
+      "loss": 1.8261,
+      "step": 1719
+    },
+    {
+      "epoch": 0.788900355463823,
+      "grad_norm": 0.36820700764656067,
+      "learning_rate": 0.0008643690174343258,
+      "loss": 1.7743,
+      "step": 1720
+    },
+    {
+      "epoch": 0.7893590184611856,
+      "grad_norm": 0.15446464717388153,
+      "learning_rate": 0.0008641993890092637,
+      "loss": 0.8386,
+      "step": 1721
+    },
+    {
+      "epoch": 0.7898176814585484,
+      "grad_norm": 0.3198365271091461,
+      "learning_rate": 0.0008640296712434654,
+      "loss": 1.6002,
+      "step": 1722
+    },
+    {
+      "epoch": 0.790276344455911,
+      "grad_norm": 0.15601693093776703,
+      "learning_rate": 0.000863859864178564,
+      "loss": 0.7734,
+      "step": 1723
+    },
+    {
+      "epoch": 0.7907350074532737,
+      "grad_norm": 0.38354170322418213,
+      "learning_rate": 0.0008636899678562143,
+      "loss": 1.7987,
+      "step": 1724
+    },
+    {
+      "epoch": 0.7911936704506364,
+      "grad_norm": 0.07455827295780182,
+      "learning_rate": 0.0008635199823180931,
+      "loss": 0.4332,
+      "step": 1725
+    },
+    {
+      "epoch": 0.7916523334479991,
+      "grad_norm": 0.3557904064655304,
+      "learning_rate": 0.0008633499076058992,
+      "loss": 1.8337,
+      "step": 1726
+    },
+    {
+      "epoch": 0.7921109964453618,
+      "grad_norm": 0.20361699163913727,
+      "learning_rate": 0.000863179743761353,
+      "loss": 0.9175,
+      "step": 1727
+    },
+    {
+      "epoch": 0.7925696594427245,
+      "grad_norm": 0.2946905493736267,
+      "learning_rate": 0.000863009490826197,
+      "loss": 1.3851,
+      "step": 1728
+    },
+    {
+      "epoch": 0.7930283224400871,
+      "grad_norm": 0.26087355613708496,
+      "learning_rate": 0.0008628391488421956,
+      "loss": 1.1733,
+      "step": 1729
+    },
+    {
+      "epoch": 0.7934869854374499,
+      "grad_norm": 0.33554574847221375,
+      "learning_rate": 0.0008626687178511346,
+      "loss": 1.6946,
+      "step": 1730
+    },
+    {
+      "epoch": 0.7939456484348125,
+      "grad_norm": 0.08757393062114716,
+      "learning_rate": 0.0008624981978948223,
+      "loss": 0.4806,
+      "step": 1731
+    },
+    {
+      "epoch": 0.7944043114321752,
+      "grad_norm": 0.3665269911289215,
+      "learning_rate": 0.0008623275890150882,
+      "loss": 1.4377,
+      "step": 1732
+    },
+    {
+      "epoch": 0.7948629744295379,
+      "grad_norm": 0.2906312346458435,
+      "learning_rate": 0.0008621568912537841,
+      "loss": 1.0789,
+      "step": 1733
+    },
+    {
+      "epoch": 0.7953216374269005,
+      "grad_norm": 0.28736838698387146,
+      "learning_rate": 0.0008619861046527832,
+      "loss": 0.922,
+      "step": 1734
+    },
+    {
+      "epoch": 0.7957803004242633,
+      "grad_norm": 0.37983497977256775,
+      "learning_rate": 0.0008618152292539807,
+      "loss": 1.8016,
+      "step": 1735
+    },
+    {
+      "epoch": 0.7962389634216259,
+      "grad_norm": 0.347023606300354,
+      "learning_rate": 0.0008616442650992937,
+      "loss": 1.9502,
+      "step": 1736
+    },
+    {
+      "epoch": 0.7966976264189887,
+      "grad_norm": 0.3550783097743988,
+      "learning_rate": 0.000861473212230661,
+      "loss": 1.7224,
+      "step": 1737
+    },
+    {
+      "epoch": 0.7971562894163513,
+      "grad_norm": 0.3077528774738312,
+      "learning_rate": 0.0008613020706900429,
+      "loss": 0.9033,
+      "step": 1738
+    },
+    {
+      "epoch": 0.797614952413714,
+      "grad_norm": 0.4047601819038391,
+      "learning_rate": 0.0008611308405194218,
+      "loss": 2.2925,
+      "step": 1739
+    },
+    {
+      "epoch": 0.7980736154110767,
+      "grad_norm": 0.4590601623058319,
+      "learning_rate": 0.0008609595217608015,
+      "loss": 1.0867,
+      "step": 1740
+    },
+    {
+      "epoch": 0.7985322784084394,
+      "grad_norm": 0.23011907935142517,
+      "learning_rate": 0.000860788114456208,
+      "loss": 0.8041,
+      "step": 1741
+    },
+    {
+      "epoch": 0.7989909414058021,
+      "grad_norm": 0.3282072842121124,
+      "learning_rate": 0.0008606166186476884,
+      "loss": 1.417,
+      "step": 1742
+    },
+    {
+      "epoch": 0.7994496044031648,
+      "grad_norm": 0.3150783181190491,
+      "learning_rate": 0.0008604450343773122,
+      "loss": 1.3964,
+      "step": 1743
+    },
+    {
+      "epoch": 0.7999082674005275,
+      "grad_norm": 0.268691748380661,
+      "learning_rate": 0.0008602733616871701,
+      "loss": 1.0648,
+      "step": 1744
+    },
+    {
+      "epoch": 0.8003669303978902,
+      "grad_norm": 0.31134241819381714,
+      "learning_rate": 0.0008601016006193747,
+      "loss": 0.8968,
+      "step": 1745
+    },
+    {
+      "epoch": 0.8008255933952528,
+      "grad_norm": 0.2897193729877472,
+      "learning_rate": 0.0008599297512160602,
+      "loss": 1.3438,
+      "step": 1746
+    },
+    {
+      "epoch": 0.8012842563926156,
+      "grad_norm": 0.3896466791629791,
+      "learning_rate": 0.0008597578135193826,
+      "loss": 1.8129,
+      "step": 1747
+    },
+    {
+      "epoch": 0.8017429193899782,
+      "grad_norm": 0.34135007858276367,
+      "learning_rate": 0.0008595857875715191,
+      "loss": 1.311,
+      "step": 1748
+    },
+    {
+      "epoch": 0.8022015823873409,
+      "grad_norm": 0.49784937500953674,
+      "learning_rate": 0.0008594136734146694,
+      "loss": 1.9587,
+      "step": 1749
+    },
+    {
+      "epoch": 0.8026602453847036,
+      "grad_norm": 0.3799792230129242,
+      "learning_rate": 0.0008592414710910541,
+      "loss": 1.6164,
+      "step": 1750
+    },
+    {
+      "epoch": 0.8031189083820662,
+      "grad_norm": 0.4176270067691803,
+      "learning_rate": 0.0008590691806429155,
+      "loss": 1.7939,
+      "step": 1751
+    },
+    {
+      "epoch": 0.803577571379429,
+      "grad_norm": 0.31819960474967957,
+      "learning_rate": 0.000858896802112518,
+      "loss": 1.2059,
+      "step": 1752
+    },
+    {
+      "epoch": 0.8040362343767916,
+      "grad_norm": 0.34135720133781433,
+      "learning_rate": 0.000858724335542147,
+      "loss": 1.4622,
+      "step": 1753
+    },
+    {
+      "epoch": 0.8044948973741544,
+      "grad_norm": 0.33655521273612976,
+      "learning_rate": 0.0008585517809741102,
+      "loss": 1.8534,
+      "step": 1754
+    },
+    {
+      "epoch": 0.804953560371517,
+      "grad_norm": 0.24837389588356018,
+      "learning_rate": 0.000858379138450736,
+      "loss": 1.2936,
+      "step": 1755
+    },
+    {
+      "epoch": 0.8054122233688797,
+      "grad_norm": 0.302501380443573,
+      "learning_rate": 0.0008582064080143752,
+      "loss": 1.5486,
+      "step": 1756
+    },
+    {
+      "epoch": 0.8058708863662424,
+      "grad_norm": 0.38520947098731995,
+      "learning_rate": 0.0008580335897073999,
+      "loss": 2.1995,
+      "step": 1757
+    },
+    {
+      "epoch": 0.8063295493636051,
+      "grad_norm": 0.05348968505859375,
+      "learning_rate": 0.0008578606835722032,
+      "loss": 0.3607,
+      "step": 1758
+    },
+    {
+      "epoch": 0.8067882123609678,
+      "grad_norm": 0.2687481939792633,
+      "learning_rate": 0.0008576876896512006,
+      "loss": 1.4349,
+      "step": 1759
+    },
+    {
+      "epoch": 0.8072468753583305,
+      "grad_norm": 0.2733592689037323,
+      "learning_rate": 0.0008575146079868287,
+      "loss": 1.4257,
+      "step": 1760
+    },
+    {
+      "epoch": 0.8077055383556931,
+      "grad_norm": 0.27336040139198303,
+      "learning_rate": 0.0008573414386215457,
+      "loss": 1.2039,
+      "step": 1761
+    },
+    {
+      "epoch": 0.8081642013530559,
+      "grad_norm": 0.27744030952453613,
+      "learning_rate": 0.000857168181597831,
+      "loss": 1.4286,
+      "step": 1762
+    },
+    {
+      "epoch": 0.8086228643504185,
+      "grad_norm": 0.2781740725040436,
+      "learning_rate": 0.0008569948369581864,
+      "loss": 1.3666,
+      "step": 1763
+    },
+    {
+      "epoch": 0.8090815273477813,
+      "grad_norm": 0.2565990686416626,
+      "learning_rate": 0.0008568214047451339,
+      "loss": 1.2767,
+      "step": 1764
+    },
+    {
+      "epoch": 0.8095401903451439,
+      "grad_norm": 0.34465324878692627,
+      "learning_rate": 0.000856647885001218,
+      "loss": 1.6884,
+      "step": 1765
+    },
+    {
+      "epoch": 0.8099988533425065,
+      "grad_norm": 0.3289906978607178,
+      "learning_rate": 0.0008564742777690047,
+      "loss": 1.5647,
+      "step": 1766
+    },
+    {
+      "epoch": 0.8104575163398693,
+      "grad_norm": 0.46371495723724365,
+      "learning_rate": 0.0008563005830910804,
+      "loss": 2.2827,
+      "step": 1767
+    },
+    {
+      "epoch": 0.8109161793372319,
+      "grad_norm": 0.3772294521331787,
+      "learning_rate": 0.0008561268010100541,
+      "loss": 1.7618,
+      "step": 1768
+    },
+    {
+      "epoch": 0.8113748423345947,
+      "grad_norm": 0.3338968753814697,
+      "learning_rate": 0.0008559529315685558,
+      "loss": 1.4556,
+      "step": 1769
+    },
+    {
+      "epoch": 0.8118335053319573,
+      "grad_norm": 0.39005744457244873,
+      "learning_rate": 0.0008557789748092369,
+      "loss": 1.8439,
+      "step": 1770
+    },
+    {
+      "epoch": 0.81229216832932,
+      "grad_norm": 0.2635294497013092,
+      "learning_rate": 0.00085560493077477,
+      "loss": 0.9517,
+      "step": 1771
+    },
+    {
+      "epoch": 0.8127508313266827,
+      "grad_norm": 0.3014441430568695,
+      "learning_rate": 0.0008554307995078495,
+      "loss": 1.2068,
+      "step": 1772
+    },
+    {
+      "epoch": 0.8132094943240454,
+      "grad_norm": 0.297055721282959,
+      "learning_rate": 0.0008552565810511912,
+      "loss": 1.3474,
+      "step": 1773
+    },
+    {
+      "epoch": 0.8136681573214081,
+      "grad_norm": 0.39429694414138794,
+      "learning_rate": 0.000855082275447532,
+      "loss": 1.3822,
+      "step": 1774
+    },
+    {
+      "epoch": 0.8141268203187708,
+      "grad_norm": 0.4196150302886963,
+      "learning_rate": 0.0008549078827396304,
+      "loss": 1.9164,
+      "step": 1775
+    },
+    {
+      "epoch": 0.8145854833161335,
+      "grad_norm": 0.33066174387931824,
+      "learning_rate": 0.0008547334029702661,
+      "loss": 1.3359,
+      "step": 1776
+    },
+    {
+      "epoch": 0.8150441463134962,
+      "grad_norm": 0.5004546642303467,
+      "learning_rate": 0.0008545588361822403,
+      "loss": 1.6743,
+      "step": 1777
+    },
+    {
+      "epoch": 0.8155028093108588,
+      "grad_norm": 0.3157084882259369,
+      "learning_rate": 0.0008543841824183756,
+      "loss": 1.2809,
+      "step": 1778
+    },
+    {
+      "epoch": 0.8159614723082216,
+      "grad_norm": 0.4495697021484375,
+      "learning_rate": 0.0008542094417215156,
+      "loss": 2.0178,
+      "step": 1779
+    },
+    {
+      "epoch": 0.8164201353055842,
+      "grad_norm": 0.40004342794418335,
+      "learning_rate": 0.0008540346141345254,
+      "loss": 1.8539,
+      "step": 1780
+    },
+    {
+      "epoch": 0.816878798302947,
+      "grad_norm": 0.3421812355518341,
+      "learning_rate": 0.0008538596997002918,
+      "loss": 1.2512,
+      "step": 1781
+    },
+    {
+      "epoch": 0.8173374613003096,
+      "grad_norm": 0.28738144040107727,
+      "learning_rate": 0.0008536846984617224,
+      "loss": 1.2953,
+      "step": 1782
+    },
+    {
+      "epoch": 0.8177961242976722,
+      "grad_norm": 0.40851840376853943,
+      "learning_rate": 0.0008535096104617464,
+      "loss": 2.1592,
+      "step": 1783
+    },
+    {
+      "epoch": 0.818254787295035,
+      "grad_norm": 0.2928701937198639,
+      "learning_rate": 0.0008533344357433138,
+      "loss": 1.3396,
+      "step": 1784
+    },
+    {
+      "epoch": 0.8187134502923976,
+      "grad_norm": 0.08765708655118942,
+      "learning_rate": 0.0008531591743493966,
+      "loss": 0.4814,
+      "step": 1785
+    },
+    {
+      "epoch": 0.8191721132897604,
+      "grad_norm": 0.3306613564491272,
+      "learning_rate": 0.0008529838263229874,
+      "loss": 1.3589,
+      "step": 1786
+    },
+    {
+      "epoch": 0.819630776287123,
+      "grad_norm": 0.29500478506088257,
+      "learning_rate": 0.0008528083917071006,
+      "loss": 1.3019,
+      "step": 1787
+    },
+    {
+      "epoch": 0.8200894392844857,
+      "grad_norm": 0.1884388029575348,
+      "learning_rate": 0.0008526328705447712,
+      "loss": 0.7717,
+      "step": 1788
+    },
+    {
+      "epoch": 0.8205481022818484,
+      "grad_norm": 0.31622594594955444,
+      "learning_rate": 0.0008524572628790562,
+      "loss": 1.669,
+      "step": 1789
+    },
+    {
+      "epoch": 0.8210067652792111,
+      "grad_norm": 0.4221099615097046,
+      "learning_rate": 0.0008522815687530333,
+      "loss": 1.6276,
+      "step": 1790
+    },
+    {
+      "epoch": 0.8214654282765738,
+      "grad_norm": 0.3094578981399536,
+      "learning_rate": 0.0008521057882098015,
+      "loss": 1.3534,
+      "step": 1791
+    },
+    {
+      "epoch": 0.8219240912739365,
+      "grad_norm": 0.2990223169326782,
+      "learning_rate": 0.000851929921292481,
+      "loss": 1.2487,
+      "step": 1792
+    },
+    {
+      "epoch": 0.8223827542712991,
+      "grad_norm": 1.4571868181228638,
+      "learning_rate": 0.0008517539680442133,
+      "loss": 1.2966,
+      "step": 1793
+    },
+    {
+      "epoch": 0.8228414172686619,
+      "grad_norm": 0.22885964810848236,
+      "learning_rate": 0.0008515779285081608,
+      "loss": 0.8865,
+      "step": 1794
+    },
+    {
+      "epoch": 0.8233000802660245,
+      "grad_norm": 0.3707837164402008,
+      "learning_rate": 0.0008514018027275074,
+      "loss": 1.4297,
+      "step": 1795
+    },
+    {
+      "epoch": 0.8237587432633873,
+      "grad_norm": 0.34915557503700256,
+      "learning_rate": 0.0008512255907454584,
+      "loss": 1.7339,
+      "step": 1796
+    },
+    {
+      "epoch": 0.8242174062607499,
+      "grad_norm": 0.2898045480251312,
+      "learning_rate": 0.0008510492926052393,
+      "loss": 1.4886,
+      "step": 1797
+    },
+    {
+      "epoch": 0.8246760692581127,
+      "grad_norm": 0.3200799822807312,
+      "learning_rate": 0.0008508729083500974,
+      "loss": 1.5057,
+      "step": 1798
+    },
+    {
+      "epoch": 0.8251347322554753,
+      "grad_norm": 0.23627309501171112,
+      "learning_rate": 0.0008506964380233014,
+      "loss": 1.0652,
+      "step": 1799
+    },
+    {
+      "epoch": 0.8255933952528379,
+      "grad_norm": 0.19365330040454865,
+      "learning_rate": 0.0008505198816681403,
+      "loss": 0.8976,
+      "step": 1800
+    },
+    {
+      "epoch": 0.8260520582502007,
+      "grad_norm": 0.30963438749313354,
+      "learning_rate": 0.0008503432393279251,
+      "loss": 1.4713,
+      "step": 1801
+    },
+    {
+      "epoch": 0.8265107212475633,
+      "grad_norm": 0.30070507526397705,
+      "learning_rate": 0.000850166511045987,
+      "loss": 1.3842,
+      "step": 1802
+    },
+    {
+      "epoch": 0.826969384244926,
+      "grad_norm": 0.23169678449630737,
+      "learning_rate": 0.0008499896968656789,
+      "loss": 1.0347,
+      "step": 1803
+    },
+    {
+      "epoch": 0.8274280472422887,
+      "grad_norm": 0.27118271589279175,
+      "learning_rate": 0.0008498127968303747,
+      "loss": 1.2278,
+      "step": 1804
+    },
+    {
+      "epoch": 0.8278867102396514,
+      "grad_norm": 0.28942152857780457,
+      "learning_rate": 0.0008496358109834691,
+      "loss": 1.2407,
+      "step": 1805
+    },
+    {
+      "epoch": 0.8283453732370141,
+      "grad_norm": 0.34731897711753845,
+      "learning_rate": 0.000849458739368378,
+      "loss": 1.7978,
+      "step": 1806
+    },
+    {
+      "epoch": 0.8288040362343768,
+      "grad_norm": 0.33974307775497437,
+      "learning_rate": 0.0008492815820285384,
+      "loss": 1.8225,
+      "step": 1807
+    },
+    {
+      "epoch": 0.8292626992317395,
+      "grad_norm": 0.36200281977653503,
+      "learning_rate": 0.0008491043390074085,
+      "loss": 1.8848,
+      "step": 1808
+    },
+    {
+      "epoch": 0.8297213622291022,
+      "grad_norm": 0.3786148130893707,
+      "learning_rate": 0.0008489270103484668,
+      "loss": 1.8229,
+      "step": 1809
+    },
+    {
+      "epoch": 0.8301800252264648,
+      "grad_norm": 0.3229668140411377,
+      "learning_rate": 0.0008487495960952139,
+      "loss": 1.4528,
+      "step": 1810
+    },
+    {
+      "epoch": 0.8306386882238276,
+      "grad_norm": 0.34114035964012146,
+      "learning_rate": 0.00084857209629117,
+      "loss": 1.3704,
+      "step": 1811
+    },
+    {
+      "epoch": 0.8310973512211902,
+      "grad_norm": 0.22234633564949036,
+      "learning_rate": 0.0008483945109798778,
+      "loss": 0.8441,
+      "step": 1812
+    },
+    {
+      "epoch": 0.831556014218553,
+      "grad_norm": 0.2830922305583954,
+      "learning_rate": 0.0008482168402049001,
+      "loss": 1.3244,
+      "step": 1813
+    },
+    {
+      "epoch": 0.8320146772159156,
+      "grad_norm": 0.30978924036026,
+      "learning_rate": 0.0008480390840098207,
+      "loss": 1.2952,
+      "step": 1814
+    },
+    {
+      "epoch": 0.8324733402132783,
+      "grad_norm": 0.3014751076698303,
+      "learning_rate": 0.0008478612424382444,
+      "loss": 1.5973,
+      "step": 1815
+    },
+    {
+      "epoch": 0.832932003210641,
+      "grad_norm": 0.1848728209733963,
+      "learning_rate": 0.0008476833155337971,
+      "loss": 0.7438,
+      "step": 1816
+    },
+    {
+      "epoch": 0.8333906662080036,
+      "grad_norm": 0.18665684759616852,
+      "learning_rate": 0.0008475053033401256,
+      "loss": 0.854,
+      "step": 1817
+    },
+    {
+      "epoch": 0.8338493292053664,
+      "grad_norm": 0.36908119916915894,
+      "learning_rate": 0.0008473272059008976,
+      "loss": 1.7524,
+      "step": 1818
+    },
+    {
+      "epoch": 0.834307992202729,
+      "grad_norm": 0.2789961099624634,
+      "learning_rate": 0.0008471490232598016,
+      "loss": 1.1132,
+      "step": 1819
+    },
+    {
+      "epoch": 0.8347666552000917,
+      "grad_norm": 0.3717726171016693,
+      "learning_rate": 0.000846970755460547,
+      "loss": 1.8761,
+      "step": 1820
+    },
+    {
+      "epoch": 0.8352253181974544,
+      "grad_norm": 0.38276535272598267,
+      "learning_rate": 0.0008467924025468645,
+      "loss": 2.0549,
+      "step": 1821
+    },
+    {
+      "epoch": 0.8356839811948171,
+      "grad_norm": 0.3366381824016571,
+      "learning_rate": 0.000846613964562505,
+      "loss": 1.5439,
+      "step": 1822
+    },
+    {
+      "epoch": 0.8361426441921798,
+      "grad_norm": 0.21815823018550873,
+      "learning_rate": 0.0008464354415512409,
+      "loss": 1.1877,
+      "step": 1823
+    },
+    {
+      "epoch": 0.8366013071895425,
+      "grad_norm": 0.3503132462501526,
+      "learning_rate": 0.000846256833556865,
+      "loss": 1.9175,
+      "step": 1824
+    },
+    {
+      "epoch": 0.8370599701869051,
+      "grad_norm": 0.2623533010482788,
+      "learning_rate": 0.0008460781406231913,
+      "loss": 1.3655,
+      "step": 1825
+    },
+    {
+      "epoch": 0.8375186331842679,
+      "grad_norm": 0.3180966377258301,
+      "learning_rate": 0.0008458993627940541,
+      "loss": 1.6582,
+      "step": 1826
+    },
+    {
+      "epoch": 0.8379772961816305,
+      "grad_norm": 0.19976475834846497,
+      "learning_rate": 0.0008457205001133093,
+      "loss": 0.8719,
+      "step": 1827
+    },
+    {
+      "epoch": 0.8384359591789933,
+      "grad_norm": 0.2790711224079132,
+      "learning_rate": 0.000845541552624833,
+      "loss": 1.736,
+      "step": 1828
+    },
+    {
+      "epoch": 0.8388946221763559,
+      "grad_norm": 0.25180596113204956,
+      "learning_rate": 0.0008453625203725224,
+      "loss": 1.3915,
+      "step": 1829
+    },
+    {
+      "epoch": 0.8393532851737187,
+      "grad_norm": 0.35062751173973083,
+      "learning_rate": 0.0008451834034002954,
+      "loss": 1.631,
+      "step": 1830
+    },
+    {
+      "epoch": 0.8398119481710813,
+      "grad_norm": 0.20606881380081177,
+      "learning_rate": 0.0008450042017520905,
+      "loss": 0.9485,
+      "step": 1831
+    },
+    {
+      "epoch": 0.840270611168444,
+      "grad_norm": 0.34081435203552246,
+      "learning_rate": 0.0008448249154718675,
+      "loss": 2.0344,
+      "step": 1832
+    },
+    {
+      "epoch": 0.8407292741658067,
+      "grad_norm": 0.2597459852695465,
+      "learning_rate": 0.0008446455446036063,
+      "loss": 1.5004,
+      "step": 1833
+    },
+    {
+      "epoch": 0.8411879371631693,
+      "grad_norm": 0.19320417940616608,
+      "learning_rate": 0.0008444660891913079,
+      "loss": 0.9119,
+      "step": 1834
+    },
+    {
+      "epoch": 0.841646600160532,
+      "grad_norm": 0.3378828763961792,
+      "learning_rate": 0.0008442865492789943,
+      "loss": 1.5369,
+      "step": 1835
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 0.2489069253206253,
+      "learning_rate": 0.0008441069249107076,
+      "loss": 1.3823,
+      "step": 1836
+    },
+    {
+      "epoch": 0.8425639261552574,
+      "grad_norm": 0.299431174993515,
+      "learning_rate": 0.000843927216130511,
+      "loss": 1.6409,
+      "step": 1837
+    },
+    {
+      "epoch": 0.8430225891526201,
+      "grad_norm": 0.17401325702667236,
+      "learning_rate": 0.0008437474229824886,
+      "loss": 0.735,
+      "step": 1838
+    },
+    {
+      "epoch": 0.8434812521499828,
+      "grad_norm": 0.267111212015152,
+      "learning_rate": 0.0008435675455107448,
+      "loss": 1.342,
+      "step": 1839
+    },
+    {
+      "epoch": 0.8439399151473455,
+      "grad_norm": 0.17649328708648682,
+      "learning_rate": 0.0008433875837594048,
+      "loss": 0.9951,
+      "step": 1840
+    },
+    {
+      "epoch": 0.8443985781447082,
+      "grad_norm": 0.24475933611392975,
+      "learning_rate": 0.0008432075377726144,
+      "loss": 1.2884,
+      "step": 1841
+    },
+    {
+      "epoch": 0.8448572411420708,
+      "grad_norm": 0.2003568857908249,
+      "learning_rate": 0.0008430274075945405,
+      "loss": 1.0649,
+      "step": 1842
+    },
+    {
+      "epoch": 0.8453159041394336,
+      "grad_norm": 0.26941803097724915,
+      "learning_rate": 0.0008428471932693701,
+      "loss": 1.4079,
+      "step": 1843
+    },
+    {
+      "epoch": 0.8457745671367962,
+      "grad_norm": 0.2373347133398056,
+      "learning_rate": 0.0008426668948413111,
+      "loss": 1.1529,
+      "step": 1844
+    },
+    {
+      "epoch": 0.846233230134159,
+      "grad_norm": 0.2616652846336365,
+      "learning_rate": 0.0008424865123545921,
+      "loss": 1.233,
+      "step": 1845
+    },
+    {
+      "epoch": 0.8466918931315216,
+      "grad_norm": 0.28413131833076477,
+      "learning_rate": 0.0008423060458534621,
+      "loss": 1.3795,
+      "step": 1846
+    },
+    {
+      "epoch": 0.8471505561288843,
+      "grad_norm": 0.2537064552307129,
+      "learning_rate": 0.000842125495382191,
+      "loss": 1.3186,
+      "step": 1847
+    },
+    {
+      "epoch": 0.847609219126247,
+      "grad_norm": 0.21368277072906494,
+      "learning_rate": 0.0008419448609850689,
+      "loss": 0.9202,
+      "step": 1848
+    },
+    {
+      "epoch": 0.8480678821236097,
+      "grad_norm": 0.2566043734550476,
+      "learning_rate": 0.000841764142706407,
+      "loss": 1.2372,
+      "step": 1849
+    },
+    {
+      "epoch": 0.8485265451209724,
+      "grad_norm": 0.34998470544815063,
+      "learning_rate": 0.0008415833405905366,
+      "loss": 1.5085,
+      "step": 1850
+    },
+    {
+      "epoch": 0.848985208118335,
+      "grad_norm": 0.13233597576618195,
+      "learning_rate": 0.0008414024546818098,
+      "loss": 0.6812,
+      "step": 1851
+    },
+    {
+      "epoch": 0.8494438711156977,
+      "grad_norm": 0.21131940186023712,
+      "learning_rate": 0.0008412214850245991,
+      "loss": 1.2487,
+      "step": 1852
+    },
+    {
+      "epoch": 0.8499025341130604,
+      "grad_norm": 0.35627928376197815,
+      "learning_rate": 0.000841040431663298,
+      "loss": 1.9961,
+      "step": 1853
+    },
+    {
+      "epoch": 0.8503611971104231,
+      "grad_norm": 0.37637192010879517,
+      "learning_rate": 0.0008408592946423198,
+      "loss": 2.4197,
+      "step": 1854
+    },
+    {
+      "epoch": 0.8508198601077858,
+      "grad_norm": 0.12372195720672607,
+      "learning_rate": 0.0008406780740060991,
+      "loss": 0.7192,
+      "step": 1855
+    },
+    {
+      "epoch": 0.8512785231051485,
+      "grad_norm": 0.24412862956523895,
+      "learning_rate": 0.0008404967697990902,
+      "loss": 1.1282,
+      "step": 1856
+    },
+    {
+      "epoch": 0.8517371861025111,
+      "grad_norm": 0.2743844985961914,
+      "learning_rate": 0.0008403153820657688,
+      "loss": 1.4214,
+      "step": 1857
+    },
+    {
+      "epoch": 0.8521958490998739,
+      "grad_norm": 0.3629717230796814,
+      "learning_rate": 0.0008401339108506302,
+      "loss": 1.7277,
+      "step": 1858
+    },
+    {
+      "epoch": 0.8526545120972365,
+      "grad_norm": 0.3425946533679962,
+      "learning_rate": 0.0008399523561981908,
+      "loss": 1.9844,
+      "step": 1859
+    },
+    {
+      "epoch": 0.8531131750945993,
+      "grad_norm": 0.284390926361084,
+      "learning_rate": 0.0008397707181529873,
+      "loss": 1.2471,
+      "step": 1860
+    },
+    {
+      "epoch": 0.8535718380919619,
+      "grad_norm": 0.2141527682542801,
+      "learning_rate": 0.0008395889967595766,
+      "loss": 0.8869,
+      "step": 1861
+    },
+    {
+      "epoch": 0.8540305010893247,
+      "grad_norm": 0.29586008191108704,
+      "learning_rate": 0.0008394071920625366,
+      "loss": 1.6176,
+      "step": 1862
+    },
+    {
+      "epoch": 0.8544891640866873,
+      "grad_norm": 0.25857093930244446,
+      "learning_rate": 0.0008392253041064652,
+      "loss": 1.2783,
+      "step": 1863
+    },
+    {
+      "epoch": 0.85494782708405,
+      "grad_norm": 0.33708396553993225,
+      "learning_rate": 0.0008390433329359806,
+      "loss": 1.4294,
+      "step": 1864
+    },
+    {
+      "epoch": 0.8554064900814127,
+      "grad_norm": 0.18086546659469604,
+      "learning_rate": 0.0008388612785957219,
+      "loss": 0.9876,
+      "step": 1865
+    },
+    {
+      "epoch": 0.8558651530787754,
+      "grad_norm": 0.2990560531616211,
+      "learning_rate": 0.0008386791411303482,
+      "loss": 1.2799,
+      "step": 1866
+    },
+    {
+      "epoch": 0.8563238160761381,
+      "grad_norm": 0.3266335129737854,
+      "learning_rate": 0.0008384969205845391,
+      "loss": 1.6877,
+      "step": 1867
+    },
+    {
+      "epoch": 0.8567824790735007,
+      "grad_norm": 0.21711353957653046,
+      "learning_rate": 0.0008383146170029949,
+      "loss": 0.8325,
+      "step": 1868
+    },
+    {
+      "epoch": 0.8572411420708634,
+      "grad_norm": 0.33946093916893005,
+      "learning_rate": 0.0008381322304304356,
+      "loss": 1.637,
+      "step": 1869
+    },
+    {
+      "epoch": 0.8576998050682261,
+      "grad_norm": 0.35655567049980164,
+      "learning_rate": 0.0008379497609116021,
+      "loss": 1.988,
+      "step": 1870
+    },
+    {
+      "epoch": 0.8581584680655888,
+      "grad_norm": 0.29446879029273987,
+      "learning_rate": 0.0008377672084912556,
+      "loss": 1.4559,
+      "step": 1871
+    },
+    {
+      "epoch": 0.8586171310629515,
+      "grad_norm": 0.30917033553123474,
+      "learning_rate": 0.0008375845732141773,
+      "loss": 1.2666,
+      "step": 1872
+    },
+    {
+      "epoch": 0.8590757940603142,
+      "grad_norm": 0.2967301905155182,
+      "learning_rate": 0.0008374018551251693,
+      "loss": 1.3149,
+      "step": 1873
+    },
+    {
+      "epoch": 0.8595344570576768,
+      "grad_norm": 0.37791574001312256,
+      "learning_rate": 0.0008372190542690534,
+      "loss": 1.9565,
+      "step": 1874
+    },
+    {
+      "epoch": 0.8599931200550396,
+      "grad_norm": 0.4147866368293762,
+      "learning_rate": 0.0008370361706906719,
+      "loss": 2.095,
+      "step": 1875
+    },
+    {
+      "epoch": 0.8604517830524022,
+      "grad_norm": 0.31991320848464966,
+      "learning_rate": 0.0008368532044348876,
+      "loss": 1.0417,
+      "step": 1876
+    },
+    {
+      "epoch": 0.860910446049765,
+      "grad_norm": 0.3214443624019623,
+      "learning_rate": 0.0008366701555465834,
+      "loss": 1.3086,
+      "step": 1877
+    },
+    {
+      "epoch": 0.8613691090471276,
+      "grad_norm": 0.32853764295578003,
+      "learning_rate": 0.0008364870240706626,
+      "loss": 1.4961,
+      "step": 1878
+    },
+    {
+      "epoch": 0.8618277720444903,
+      "grad_norm": 0.3698137700557709,
+      "learning_rate": 0.0008363038100520485,
+      "loss": 1.6169,
+      "step": 1879
+    },
+    {
+      "epoch": 0.862286435041853,
+      "grad_norm": 0.3349319100379944,
+      "learning_rate": 0.000836120513535685,
+      "loss": 1.6381,
+      "step": 1880
+    },
+    {
+      "epoch": 0.8627450980392157,
+      "grad_norm": 0.354120135307312,
+      "learning_rate": 0.0008359371345665359,
+      "loss": 1.7372,
+      "step": 1881
+    },
+    {
+      "epoch": 0.8632037610365784,
+      "grad_norm": 0.31368064880371094,
+      "learning_rate": 0.0008357536731895855,
+      "loss": 1.7236,
+      "step": 1882
+    },
+    {
+      "epoch": 0.8636624240339411,
+      "grad_norm": 0.33393827080726624,
+      "learning_rate": 0.0008355701294498381,
+      "loss": 1.7104,
+      "step": 1883
+    },
+    {
+      "epoch": 0.8641210870313037,
+      "grad_norm": 0.26583972573280334,
+      "learning_rate": 0.0008353865033923185,
+      "loss": 1.2689,
+      "step": 1884
+    },
+    {
+      "epoch": 0.8645797500286664,
+      "grad_norm": 0.33901429176330566,
+      "learning_rate": 0.0008352027950620714,
+      "loss": 1.652,
+      "step": 1885
+    },
+    {
+      "epoch": 0.8650384130260291,
+      "grad_norm": 0.371894508600235,
+      "learning_rate": 0.0008350190045041615,
+      "loss": 1.7666,
+      "step": 1886
+    },
+    {
+      "epoch": 0.8654970760233918,
+      "grad_norm": 0.19485050439834595,
+      "learning_rate": 0.0008348351317636742,
+      "loss": 0.8739,
+      "step": 1887
+    },
+    {
+      "epoch": 0.8659557390207545,
+      "grad_norm": 0.308272123336792,
+      "learning_rate": 0.000834651176885715,
+      "loss": 1.4182,
+      "step": 1888
+    },
+    {
+      "epoch": 0.8664144020181171,
+      "grad_norm": 0.3224412500858307,
+      "learning_rate": 0.000834467139915409,
+      "loss": 1.9642,
+      "step": 1889
+    },
+    {
+      "epoch": 0.8668730650154799,
+      "grad_norm": 0.24551476538181305,
+      "learning_rate": 0.000834283020897902,
+      "loss": 1.1545,
+      "step": 1890
+    },
+    {
+      "epoch": 0.8673317280128425,
+      "grad_norm": 0.3001308739185333,
+      "learning_rate": 0.0008340988198783597,
+      "loss": 1.6565,
+      "step": 1891
+    },
+    {
+      "epoch": 0.8677903910102053,
+      "grad_norm": 0.36925092339515686,
+      "learning_rate": 0.0008339145369019678,
+      "loss": 2.0125,
+      "step": 1892
+    },
+    {
+      "epoch": 0.8682490540075679,
+      "grad_norm": 0.3273385465145111,
+      "learning_rate": 0.0008337301720139323,
+      "loss": 1.3712,
+      "step": 1893
+    },
+    {
+      "epoch": 0.8687077170049307,
+      "grad_norm": 0.3309119939804077,
+      "learning_rate": 0.0008335457252594795,
+      "loss": 1.6698,
+      "step": 1894
+    },
+    {
+      "epoch": 0.8691663800022933,
+      "grad_norm": 0.23802553117275238,
+      "learning_rate": 0.0008333611966838552,
+      "loss": 0.9141,
+      "step": 1895
+    },
+    {
+      "epoch": 0.869625042999656,
+      "grad_norm": 0.26225998997688293,
+      "learning_rate": 0.0008331765863323255,
+      "loss": 1.04,
+      "step": 1896
+    },
+    {
+      "epoch": 0.8700837059970187,
+      "grad_norm": 0.33123326301574707,
+      "learning_rate": 0.0008329918942501772,
+      "loss": 1.7666,
+      "step": 1897
+    },
+    {
+      "epoch": 0.8705423689943814,
+      "grad_norm": 0.238576740026474,
+      "learning_rate": 0.000832807120482716,
+      "loss": 1.0255,
+      "step": 1898
+    },
+    {
+      "epoch": 0.8710010319917441,
+      "grad_norm": 0.3323553204536438,
+      "learning_rate": 0.0008326222650752686,
+      "loss": 1.4065,
+      "step": 1899
+    },
+    {
+      "epoch": 0.8714596949891068,
+      "grad_norm": 0.18547557294368744,
+      "learning_rate": 0.0008324373280731811,
+      "loss": 0.8987,
+      "step": 1900
+    },
+    {
+      "epoch": 0.8719183579864694,
+      "grad_norm": 0.3150063455104828,
+      "learning_rate": 0.0008322523095218202,
+      "loss": 1.2314,
+      "step": 1901
+    },
+    {
+      "epoch": 0.8723770209838321,
+      "grad_norm": 0.3147270977497101,
+      "learning_rate": 0.0008320672094665722,
+      "loss": 1.2318,
+      "step": 1902
+    },
+    {
+      "epoch": 0.8728356839811948,
+      "grad_norm": 0.4290883541107178,
+      "learning_rate": 0.0008318820279528432,
+      "loss": 2.0358,
+      "step": 1903
+    },
+    {
+      "epoch": 0.8732943469785575,
+      "grad_norm": 0.4178559482097626,
+      "learning_rate": 0.0008316967650260596,
+      "loss": 2.2466,
+      "step": 1904
+    },
+    {
+      "epoch": 0.8737530099759202,
+      "grad_norm": 0.18568667769432068,
+      "learning_rate": 0.0008315114207316682,
+      "loss": 0.9221,
+      "step": 1905
+    },
+    {
+      "epoch": 0.8742116729732828,
+      "grad_norm": 0.14113818109035492,
+      "learning_rate": 0.0008313259951151349,
+      "loss": 0.7635,
+      "step": 1906
+    },
+    {
+      "epoch": 0.8746703359706456,
+      "grad_norm": 0.3099023103713989,
+      "learning_rate": 0.0008311404882219458,
+      "loss": 1.3216,
+      "step": 1907
+    },
+    {
+      "epoch": 0.8751289989680082,
+      "grad_norm": 0.2004273533821106,
+      "learning_rate": 0.0008309549000976075,
+      "loss": 1.1125,
+      "step": 1908
+    },
+    {
+      "epoch": 0.875587661965371,
+      "grad_norm": 0.2782037556171417,
+      "learning_rate": 0.0008307692307876458,
+      "loss": 1.4631,
+      "step": 1909
+    },
+    {
+      "epoch": 0.8760463249627336,
+      "grad_norm": 0.29525527358055115,
+      "learning_rate": 0.0008305834803376069,
+      "loss": 1.3377,
+      "step": 1910
+    },
+    {
+      "epoch": 0.8765049879600963,
+      "grad_norm": 0.28939589858055115,
+      "learning_rate": 0.0008303976487930567,
+      "loss": 1.2637,
+      "step": 1911
+    },
+    {
+      "epoch": 0.876963650957459,
+      "grad_norm": 0.24675489962100983,
+      "learning_rate": 0.0008302117361995808,
+      "loss": 1.3734,
+      "step": 1912
+    },
+    {
+      "epoch": 0.8774223139548217,
+      "grad_norm": 0.2444259226322174,
+      "learning_rate": 0.0008300257426027851,
+      "loss": 1.0965,
+      "step": 1913
+    },
+    {
+      "epoch": 0.8778809769521844,
+      "grad_norm": 0.17775723338127136,
+      "learning_rate": 0.0008298396680482951,
+      "loss": 0.9796,
+      "step": 1914
+    },
+    {
+      "epoch": 0.8783396399495471,
+      "grad_norm": 0.24205970764160156,
+      "learning_rate": 0.0008296535125817564,
+      "loss": 1.2806,
+      "step": 1915
+    },
+    {
+      "epoch": 0.8787983029469097,
+      "grad_norm": 0.47772538661956787,
+      "learning_rate": 0.000829467276248834,
+      "loss": 1.7538,
+      "step": 1916
+    },
+    {
+      "epoch": 0.8792569659442725,
+      "grad_norm": 0.30592289566993713,
+      "learning_rate": 0.0008292809590952132,
+      "loss": 1.6431,
+      "step": 1917
+    },
+    {
+      "epoch": 0.8797156289416351,
+      "grad_norm": 0.2858329713344574,
+      "learning_rate": 0.000829094561166599,
+      "loss": 1.5801,
+      "step": 1918
+    },
+    {
+      "epoch": 0.8801742919389978,
+      "grad_norm": 0.28482332825660706,
+      "learning_rate": 0.0008289080825087158,
+      "loss": 1.093,
+      "step": 1919
+    },
+    {
+      "epoch": 0.8806329549363605,
+      "grad_norm": 0.35098031163215637,
+      "learning_rate": 0.0008287215231673087,
+      "loss": 1.6113,
+      "step": 1920
+    },
+    {
+      "epoch": 0.8810916179337231,
+      "grad_norm": 0.15006844699382782,
+      "learning_rate": 0.0008285348831881417,
+      "loss": 0.6972,
+      "step": 1921
+    },
+    {
+      "epoch": 0.8815502809310859,
+      "grad_norm": 0.393998384475708,
+      "learning_rate": 0.0008283481626169989,
+      "loss": 1.7005,
+      "step": 1922
+    },
+    {
+      "epoch": 0.8820089439284485,
+      "grad_norm": 0.2868679165840149,
+      "learning_rate": 0.0008281613614996842,
+      "loss": 0.8066,
+      "step": 1923
+    },
+    {
+      "epoch": 0.8824676069258113,
+      "grad_norm": 0.20780214667320251,
+      "learning_rate": 0.0008279744798820216,
+      "loss": 0.973,
+      "step": 1924
+    },
+    {
+      "epoch": 0.8829262699231739,
+      "grad_norm": 0.34180140495300293,
+      "learning_rate": 0.0008277875178098541,
+      "loss": 1.4466,
+      "step": 1925
+    },
+    {
+      "epoch": 0.8833849329205367,
+      "grad_norm": 0.229378804564476,
+      "learning_rate": 0.0008276004753290451,
+      "loss": 1.1042,
+      "step": 1926
+    },
+    {
+      "epoch": 0.8838435959178993,
+      "grad_norm": 0.27841222286224365,
+      "learning_rate": 0.0008274133524854773,
+      "loss": 1.3994,
+      "step": 1927
+    },
+    {
+      "epoch": 0.884302258915262,
+      "grad_norm": 0.21060162782669067,
+      "learning_rate": 0.0008272261493250533,
+      "loss": 1.0967,
+      "step": 1928
+    },
+    {
+      "epoch": 0.8847609219126247,
+      "grad_norm": 0.37685856223106384,
+      "learning_rate": 0.0008270388658936956,
+      "loss": 1.8954,
+      "step": 1929
+    },
+    {
+      "epoch": 0.8852195849099874,
+      "grad_norm": 0.44438236951828003,
+      "learning_rate": 0.000826851502237346,
+      "loss": 2.2446,
+      "step": 1930
+    },
+    {
+      "epoch": 0.8856782479073501,
+      "grad_norm": 0.4858162999153137,
+      "learning_rate": 0.0008266640584019662,
+      "loss": 2.2056,
+      "step": 1931
+    },
+    {
+      "epoch": 0.8861369109047128,
+      "grad_norm": 0.2732187509536743,
+      "learning_rate": 0.0008264765344335373,
+      "loss": 1.4948,
+      "step": 1932
+    },
+    {
+      "epoch": 0.8865955739020754,
+      "grad_norm": 0.2720052897930145,
+      "learning_rate": 0.0008262889303780607,
+      "loss": 1.3649,
+      "step": 1933
+    },
+    {
+      "epoch": 0.8870542368994382,
+      "grad_norm": 0.250740110874176,
+      "learning_rate": 0.0008261012462815567,
+      "loss": 1.2161,
+      "step": 1934
+    },
+    {
+      "epoch": 0.8875128998968008,
+      "grad_norm": 0.20551849901676178,
+      "learning_rate": 0.0008259134821900658,
+      "loss": 1.0547,
+      "step": 1935
+    },
+    {
+      "epoch": 0.8879715628941636,
+      "grad_norm": 0.23970794677734375,
+      "learning_rate": 0.0008257256381496476,
+      "loss": 1.1497,
+      "step": 1936
+    },
+    {
+      "epoch": 0.8884302258915262,
+      "grad_norm": 0.357015997171402,
+      "learning_rate": 0.0008255377142063819,
+      "loss": 1.4954,
+      "step": 1937
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 0.3159123659133911,
+      "learning_rate": 0.0008253497104063676,
+      "loss": 1.3619,
+      "step": 1938
+    },
+    {
+      "epoch": 0.8893475518862516,
+      "grad_norm": 0.2619781196117401,
+      "learning_rate": 0.0008251616267957234,
+      "loss": 1.0236,
+      "step": 1939
+    },
+    {
+      "epoch": 0.8898062148836142,
+      "grad_norm": 0.2195945680141449,
+      "learning_rate": 0.0008249734634205876,
+      "loss": 0.9583,
+      "step": 1940
+    },
+    {
+      "epoch": 0.890264877880977,
+      "grad_norm": 0.11660360544919968,
+      "learning_rate": 0.0008247852203271182,
+      "loss": 0.6041,
+      "step": 1941
+    },
+    {
+      "epoch": 0.8907235408783396,
+      "grad_norm": 0.33388739824295044,
+      "learning_rate": 0.0008245968975614924,
+      "loss": 1.6398,
+      "step": 1942
+    },
+    {
+      "epoch": 0.8911822038757024,
+      "grad_norm": 0.24743525683879852,
+      "learning_rate": 0.0008244084951699071,
+      "loss": 1.4167,
+      "step": 1943
+    },
+    {
+      "epoch": 0.891640866873065,
+      "grad_norm": 0.2996421456336975,
+      "learning_rate": 0.0008242200131985789,
+      "loss": 1.6158,
+      "step": 1944
+    },
+    {
+      "epoch": 0.8920995298704277,
+      "grad_norm": 0.24171142280101776,
+      "learning_rate": 0.0008240314516937435,
+      "loss": 1.1656,
+      "step": 1945
+    },
+    {
+      "epoch": 0.8925581928677904,
+      "grad_norm": 0.5590919256210327,
+      "learning_rate": 0.0008238428107016568,
+      "loss": 1.809,
+      "step": 1946
+    },
+    {
+      "epoch": 0.8930168558651531,
+      "grad_norm": 0.16760781407356262,
+      "learning_rate": 0.0008236540902685937,
+      "loss": 0.8538,
+      "step": 1947
+    },
+    {
+      "epoch": 0.8934755188625157,
+      "grad_norm": 0.2322719246149063,
+      "learning_rate": 0.0008234652904408482,
+      "loss": 1.0128,
+      "step": 1948
+    },
+    {
+      "epoch": 0.8939341818598785,
+      "grad_norm": 0.28307685256004333,
+      "learning_rate": 0.000823276411264735,
+      "loss": 1.377,
+      "step": 1949
+    },
+    {
+      "epoch": 0.8943928448572411,
+      "grad_norm": 0.27569177746772766,
+      "learning_rate": 0.000823087452786587,
+      "loss": 1.4318,
+      "step": 1950
+    },
+    {
+      "epoch": 0.8948515078546039,
+      "grad_norm": 0.2631165087223053,
+      "learning_rate": 0.0008228984150527574,
+      "loss": 1.4855,
+      "step": 1951
+    },
+    {
+      "epoch": 0.8953101708519665,
+      "grad_norm": 0.18486323952674866,
+      "learning_rate": 0.0008227092981096183,
+      "loss": 0.7875,
+      "step": 1952
+    },
+    {
+      "epoch": 0.8957688338493293,
+      "grad_norm": 0.20646964013576508,
+      "learning_rate": 0.0008225201020035615,
+      "loss": 0.9301,
+      "step": 1953
+    },
+    {
+      "epoch": 0.8962274968466919,
+      "grad_norm": 0.2068818062543869,
+      "learning_rate": 0.0008223308267809982,
+      "loss": 1.1328,
+      "step": 1954
+    },
+    {
+      "epoch": 0.8966861598440545,
+      "grad_norm": 0.1731792837381363,
+      "learning_rate": 0.000822141472488359,
+      "loss": 0.8079,
+      "step": 1955
+    },
+    {
+      "epoch": 0.8971448228414173,
+      "grad_norm": 0.36508551239967346,
+      "learning_rate": 0.0008219520391720938,
+      "loss": 2.1365,
+      "step": 1956
+    },
+    {
+      "epoch": 0.8976034858387799,
+      "grad_norm": 0.29660528898239136,
+      "learning_rate": 0.000821762526878672,
+      "loss": 1.6558,
+      "step": 1957
+    },
+    {
+      "epoch": 0.8980621488361427,
+      "grad_norm": 0.20313714444637299,
+      "learning_rate": 0.0008215729356545823,
+      "loss": 0.9536,
+      "step": 1958
+    },
+    {
+      "epoch": 0.8985208118335053,
+      "grad_norm": 0.24836888909339905,
+      "learning_rate": 0.0008213832655463329,
+      "loss": 1.2862,
+      "step": 1959
+    },
+    {
+      "epoch": 0.898979474830868,
+      "grad_norm": 0.26236483454704285,
+      "learning_rate": 0.000821193516600451,
+      "loss": 1.5496,
+      "step": 1960
+    },
+    {
+      "epoch": 0.8994381378282307,
+      "grad_norm": 0.2642938196659088,
+      "learning_rate": 0.0008210036888634838,
+      "loss": 1.2446,
+      "step": 1961
+    },
+    {
+      "epoch": 0.8998968008255934,
+      "grad_norm": 0.2847262918949127,
+      "learning_rate": 0.000820813782381997,
+      "loss": 1.3784,
+      "step": 1962
+    },
+    {
+      "epoch": 0.9003554638229561,
+      "grad_norm": 0.3096359372138977,
+      "learning_rate": 0.0008206237972025761,
+      "loss": 1.4408,
+      "step": 1963
+    },
+    {
+      "epoch": 0.9008141268203188,
+      "grad_norm": 0.18418483436107635,
+      "learning_rate": 0.0008204337333718261,
+      "loss": 0.8153,
+      "step": 1964
+    },
+    {
+      "epoch": 0.9012727898176814,
+      "grad_norm": 0.24402552843093872,
+      "learning_rate": 0.0008202435909363708,
+      "loss": 1.1557,
+      "step": 1965
+    },
+    {
+      "epoch": 0.9017314528150442,
+      "grad_norm": 0.2893608808517456,
+      "learning_rate": 0.0008200533699428536,
+      "loss": 1.2294,
+      "step": 1966
+    },
+    {
+      "epoch": 0.9021901158124068,
+      "grad_norm": 0.1799740344285965,
+      "learning_rate": 0.000819863070437937,
+      "loss": 0.9913,
+      "step": 1967
+    },
+    {
+      "epoch": 0.9026487788097696,
+      "grad_norm": 0.26435887813568115,
+      "learning_rate": 0.0008196726924683029,
+      "loss": 1.302,
+      "step": 1968
+    },
+    {
+      "epoch": 0.9031074418071322,
+      "grad_norm": 0.2572958767414093,
+      "learning_rate": 0.0008194822360806525,
+      "loss": 1.5272,
+      "step": 1969
+    },
+    {
+      "epoch": 0.903566104804495,
+      "grad_norm": 0.3217867612838745,
+      "learning_rate": 0.0008192917013217059,
+      "loss": 1.6104,
+      "step": 1970
+    },
+    {
+      "epoch": 0.9040247678018576,
+      "grad_norm": 0.34428954124450684,
+      "learning_rate": 0.0008191010882382027,
+      "loss": 1.8489,
+      "step": 1971
+    },
+    {
+      "epoch": 0.9044834307992202,
+      "grad_norm": 0.23394142091274261,
+      "learning_rate": 0.0008189103968769018,
+      "loss": 1.1929,
+      "step": 1972
+    },
+    {
+      "epoch": 0.904942093796583,
+      "grad_norm": 0.21039988100528717,
+      "learning_rate": 0.0008187196272845811,
+      "loss": 1.2102,
+      "step": 1973
+    },
+    {
+      "epoch": 0.9054007567939456,
+      "grad_norm": 0.36821311712265015,
+      "learning_rate": 0.0008185287795080377,
+      "loss": 2.1794,
+      "step": 1974
+    },
+    {
+      "epoch": 0.9058594197913084,
+      "grad_norm": 0.47129011154174805,
+      "learning_rate": 0.0008183378535940879,
+      "loss": 1.7034,
+      "step": 1975
+    },
+    {
+      "epoch": 0.906318082788671,
+      "grad_norm": 0.28976601362228394,
+      "learning_rate": 0.0008181468495895674,
+      "loss": 1.404,
+      "step": 1976
+    },
+    {
+      "epoch": 0.9067767457860337,
+      "grad_norm": 0.416605681180954,
+      "learning_rate": 0.0008179557675413307,
+      "loss": 1.8235,
+      "step": 1977
+    },
+    {
+      "epoch": 0.9072354087833964,
+      "grad_norm": 0.3279779255390167,
+      "learning_rate": 0.0008177646074962517,
+      "loss": 1.7274,
+      "step": 1978
+    },
+    {
+      "epoch": 0.9076940717807591,
+      "grad_norm": 0.2587118148803711,
+      "learning_rate": 0.0008175733695012231,
+      "loss": 1.1779,
+      "step": 1979
+    },
+    {
+      "epoch": 0.9081527347781218,
+      "grad_norm": 0.33901700377464294,
+      "learning_rate": 0.0008173820536031574,
+      "loss": 2.0381,
+      "step": 1980
+    },
+    {
+      "epoch": 0.9086113977754845,
+      "grad_norm": 0.22714871168136597,
+      "learning_rate": 0.0008171906598489853,
+      "loss": 1.3698,
+      "step": 1981
+    },
+    {
+      "epoch": 0.9090700607728471,
+      "grad_norm": 0.32320547103881836,
+      "learning_rate": 0.0008169991882856574,
+      "loss": 1.7041,
+      "step": 1982
+    },
+    {
+      "epoch": 0.9095287237702099,
+      "grad_norm": 0.3090137541294098,
+      "learning_rate": 0.0008168076389601427,
+      "loss": 1.3677,
+      "step": 1983
+    },
+    {
+      "epoch": 0.9099873867675725,
+      "grad_norm": 0.25512242317199707,
+      "learning_rate": 0.0008166160119194301,
+      "loss": 1.3021,
+      "step": 1984
+    },
+    {
+      "epoch": 0.9104460497649353,
+      "grad_norm": 0.328730970621109,
+      "learning_rate": 0.0008164243072105267,
+      "loss": 1.4302,
+      "step": 1985
+    },
+    {
+      "epoch": 0.9109047127622979,
+      "grad_norm": 0.30282872915267944,
+      "learning_rate": 0.0008162325248804594,
+      "loss": 1.5295,
+      "step": 1986
+    },
+    {
+      "epoch": 0.9113633757596606,
+      "grad_norm": 0.3073786795139313,
+      "learning_rate": 0.0008160406649762735,
+      "loss": 1.605,
+      "step": 1987
+    },
+    {
+      "epoch": 0.9118220387570233,
+      "grad_norm": 0.5435524582862854,
+      "learning_rate": 0.0008158487275450335,
+      "loss": 1.8843,
+      "step": 1988
+    },
+    {
+      "epoch": 0.9122807017543859,
+      "grad_norm": 0.3203667104244232,
+      "learning_rate": 0.0008156567126338236,
+      "loss": 0.826,
+      "step": 1989
+    },
+    {
+      "epoch": 0.9127393647517487,
+      "grad_norm": 0.2148309201002121,
+      "learning_rate": 0.0008154646202897459,
+      "loss": 0.9943,
+      "step": 1990
+    },
+    {
+      "epoch": 0.9131980277491113,
+      "grad_norm": 0.2307606041431427,
+      "learning_rate": 0.0008152724505599223,
+      "loss": 1.1053,
+      "step": 1991
+    },
+    {
+      "epoch": 0.913656690746474,
+      "grad_norm": 0.30222174525260925,
+      "learning_rate": 0.0008150802034914932,
+      "loss": 1.1675,
+      "step": 1992
+    },
+    {
+      "epoch": 0.9141153537438367,
+      "grad_norm": 0.3027079105377197,
+      "learning_rate": 0.0008148878791316184,
+      "loss": 1.2104,
+      "step": 1993
+    },
+    {
+      "epoch": 0.9145740167411994,
+      "grad_norm": 0.2728803753852844,
+      "learning_rate": 0.0008146954775274764,
+      "loss": 1.3286,
+      "step": 1994
+    },
+    {
+      "epoch": 0.9150326797385621,
+      "grad_norm": 0.413828045129776,
+      "learning_rate": 0.0008145029987262649,
+      "loss": 2.343,
+      "step": 1995
+    },
+    {
+      "epoch": 0.9154913427359248,
+      "grad_norm": 0.125292107462883,
+      "learning_rate": 0.0008143104427751998,
+      "loss": 0.7368,
+      "step": 1996
+    },
+    {
+      "epoch": 0.9159500057332874,
+      "grad_norm": 0.22482484579086304,
+      "learning_rate": 0.000814117809721517,
+      "loss": 1.1551,
+      "step": 1997
+    },
+    {
+      "epoch": 0.9164086687306502,
+      "grad_norm": 0.2546938955783844,
+      "learning_rate": 0.0008139250996124706,
+      "loss": 1.3978,
+      "step": 1998
+    },
+    {
+      "epoch": 0.9168673317280128,
+      "grad_norm": 0.31431275606155396,
+      "learning_rate": 0.0008137323124953335,
+      "loss": 1.5389,
+      "step": 1999
+    },
+    {
+      "epoch": 0.9173259947253756,
+      "grad_norm": 0.3111874759197235,
+      "learning_rate": 0.0008135394484173981,
+      "loss": 1.564,
+      "step": 2000
+    },
+    {
+      "epoch": 0.9177846577227382,
+      "grad_norm": 0.36270517110824585,
+      "learning_rate": 0.0008133465074259754,
+      "loss": 2.1987,
+      "step": 2001
+    },
+    {
+      "epoch": 0.918243320720101,
+      "grad_norm": 0.34382519125938416,
+      "learning_rate": 0.000813153489568395,
+      "loss": 2.0159,
+      "step": 2002
+    },
+    {
+      "epoch": 0.9187019837174636,
+      "grad_norm": 0.39224573969841003,
+      "learning_rate": 0.0008129603948920056,
+      "loss": 1.6236,
+      "step": 2003
+    },
+    {
+      "epoch": 0.9191606467148263,
+      "grad_norm": 0.2750489115715027,
+      "learning_rate": 0.000812767223444175,
+      "loss": 1.2472,
+      "step": 2004
+    },
+    {
+      "epoch": 0.919619309712189,
+      "grad_norm": 0.26714855432510376,
+      "learning_rate": 0.0008125739752722891,
+      "loss": 1.5997,
+      "step": 2005
+    },
+    {
+      "epoch": 0.9200779727095516,
+      "grad_norm": 0.24979934096336365,
+      "learning_rate": 0.0008123806504237533,
+      "loss": 1.3241,
+      "step": 2006
+    },
+    {
+      "epoch": 0.9205366357069144,
+      "grad_norm": 0.08660929650068283,
+      "learning_rate": 0.0008121872489459916,
+      "loss": 0.5286,
+      "step": 2007
+    },
+    {
+      "epoch": 0.920995298704277,
+      "grad_norm": 0.13258817791938782,
+      "learning_rate": 0.0008119937708864469,
+      "loss": 0.6686,
+      "step": 2008
+    },
+    {
+      "epoch": 0.9214539617016397,
+      "grad_norm": 0.18847809731960297,
+      "learning_rate": 0.0008118002162925804,
+      "loss": 0.839,
+      "step": 2009
+    },
+    {
+      "epoch": 0.9219126246990024,
+      "grad_norm": 0.059659842401742935,
+      "learning_rate": 0.0008116065852118728,
+      "loss": 0.3741,
+      "step": 2010
+    },
+    {
+      "epoch": 0.9223712876963651,
+      "grad_norm": 0.33892372250556946,
+      "learning_rate": 0.0008114128776918229,
+      "loss": 1.7498,
+      "step": 2011
+    },
+    {
+      "epoch": 0.9228299506937278,
+      "grad_norm": 0.2842087149620056,
+      "learning_rate": 0.0008112190937799488,
+      "loss": 1.4104,
+      "step": 2012
+    },
+    {
+      "epoch": 0.9232886136910905,
+      "grad_norm": 0.18660642206668854,
+      "learning_rate": 0.0008110252335237868,
+      "loss": 0.9644,
+      "step": 2013
+    },
+    {
+      "epoch": 0.9237472766884531,
+      "grad_norm": 0.1834762841463089,
+      "learning_rate": 0.0008108312969708928,
+      "loss": 0.8232,
+      "step": 2014
+    },
+    {
+      "epoch": 0.9242059396858159,
+      "grad_norm": 0.3446117639541626,
+      "learning_rate": 0.0008106372841688401,
+      "loss": 2.0298,
+      "step": 2015
+    },
+    {
+      "epoch": 0.9246646026831785,
+      "grad_norm": 0.3695407807826996,
+      "learning_rate": 0.0008104431951652219,
+      "loss": 1.6854,
+      "step": 2016
+    },
+    {
+      "epoch": 0.9251232656805413,
+      "grad_norm": 0.26748037338256836,
+      "learning_rate": 0.0008102490300076496,
+      "loss": 1.3953,
+      "step": 2017
+    },
+    {
+      "epoch": 0.9255819286779039,
+      "grad_norm": 0.08492042124271393,
+      "learning_rate": 0.0008100547887437531,
+      "loss": 0.5242,
+      "step": 2018
+    },
+    {
+      "epoch": 0.9260405916752666,
+      "grad_norm": 0.2597675919532776,
+      "learning_rate": 0.0008098604714211813,
+      "loss": 1.1575,
+      "step": 2019
+    },
+    {
+      "epoch": 0.9264992546726293,
+      "grad_norm": 0.29596418142318726,
+      "learning_rate": 0.0008096660780876017,
+      "loss": 1.4683,
+      "step": 2020
+    },
+    {
+      "epoch": 0.926957917669992,
+      "grad_norm": 0.2640438973903656,
+      "learning_rate": 0.0008094716087907003,
+      "loss": 1.4299,
+      "step": 2021
+    },
+    {
+      "epoch": 0.9274165806673547,
+      "grad_norm": 0.1778401881456375,
+      "learning_rate": 0.0008092770635781821,
+      "loss": 0.8381,
+      "step": 2022
+    },
+    {
+      "epoch": 0.9278752436647173,
+      "grad_norm": 0.2718510329723358,
+      "learning_rate": 0.0008090824424977699,
+      "loss": 1.5079,
+      "step": 2023
+    },
+    {
+      "epoch": 0.92833390666208,
+      "grad_norm": 0.2521864175796509,
+      "learning_rate": 0.0008088877455972062,
+      "loss": 1.0537,
+      "step": 2024
+    },
+    {
+      "epoch": 0.9287925696594427,
+      "grad_norm": 0.2869158685207367,
+      "learning_rate": 0.000808692972924251,
+      "loss": 1.6425,
+      "step": 2025
+    },
+    {
+      "epoch": 0.9292512326568054,
+      "grad_norm": 0.33233222365379333,
+      "learning_rate": 0.000808498124526684,
+      "loss": 1.9667,
+      "step": 2026
+    },
+    {
+      "epoch": 0.9297098956541681,
+      "grad_norm": 0.34734752774238586,
+      "learning_rate": 0.0008083032004523026,
+      "loss": 1.9368,
+      "step": 2027
+    },
+    {
+      "epoch": 0.9301685586515308,
+      "grad_norm": 0.37285783886909485,
+      "learning_rate": 0.0008081082007489231,
+      "loss": 1.7972,
+      "step": 2028
+    },
+    {
+      "epoch": 0.9306272216488934,
+      "grad_norm": 0.19466300308704376,
+      "learning_rate": 0.0008079131254643804,
+      "loss": 0.8481,
+      "step": 2029
+    },
+    {
+      "epoch": 0.9310858846462562,
+      "grad_norm": 0.28476589918136597,
+      "learning_rate": 0.0008077179746465278,
+      "loss": 1.4938,
+      "step": 2030
+    },
+    {
+      "epoch": 0.9315445476436188,
+      "grad_norm": 0.22528983652591705,
+      "learning_rate": 0.0008075227483432374,
+      "loss": 1.0093,
+      "step": 2031
+    },
+    {
+      "epoch": 0.9320032106409816,
+      "grad_norm": 0.33551692962646484,
+      "learning_rate": 0.0008073274466023994,
+      "loss": 1.4329,
+      "step": 2032
+    },
+    {
+      "epoch": 0.9324618736383442,
+      "grad_norm": 0.40014681220054626,
+      "learning_rate": 0.0008071320694719226,
+      "loss": 1.513,
+      "step": 2033
+    },
+    {
+      "epoch": 0.932920536635707,
+      "grad_norm": 0.35248538851737976,
+      "learning_rate": 0.000806936616999735,
+      "loss": 1.8031,
+      "step": 2034
+    },
+    {
+      "epoch": 0.9333791996330696,
+      "grad_norm": 0.20108796656131744,
+      "learning_rate": 0.0008067410892337819,
+      "loss": 0.9306,
+      "step": 2035
+    },
+    {
+      "epoch": 0.9338378626304323,
+      "grad_norm": 0.2716940939426422,
+      "learning_rate": 0.000806545486222028,
+      "loss": 1.2828,
+      "step": 2036
+    },
+    {
+      "epoch": 0.934296525627795,
+      "grad_norm": 0.3034631311893463,
+      "learning_rate": 0.0008063498080124559,
+      "loss": 1.4306,
+      "step": 2037
+    },
+    {
+      "epoch": 0.9347551886251577,
+      "grad_norm": 0.3705253005027771,
+      "learning_rate": 0.000806154054653067,
+      "loss": 1.8486,
+      "step": 2038
+    },
+    {
+      "epoch": 0.9352138516225204,
+      "grad_norm": 0.25508952140808105,
+      "learning_rate": 0.0008059582261918812,
+      "loss": 1.2751,
+      "step": 2039
+    },
+    {
+      "epoch": 0.935672514619883,
+      "grad_norm": 0.39250513911247253,
+      "learning_rate": 0.0008057623226769362,
+      "loss": 2.0559,
+      "step": 2040
+    },
+    {
+      "epoch": 0.9361311776172457,
+      "grad_norm": 0.2672523558139801,
+      "learning_rate": 0.0008055663441562889,
+      "loss": 1.037,
+      "step": 2041
+    },
+    {
+      "epoch": 0.9365898406146084,
+      "grad_norm": 0.19914977252483368,
+      "learning_rate": 0.0008053702906780142,
+      "loss": 0.812,
+      "step": 2042
+    },
+    {
+      "epoch": 0.9370485036119711,
+      "grad_norm": 0.20971417427062988,
+      "learning_rate": 0.0008051741622902052,
+      "loss": 0.8383,
+      "step": 2043
+    },
+    {
+      "epoch": 0.9375071666093338,
+      "grad_norm": 0.3085592985153198,
+      "learning_rate": 0.0008049779590409739,
+      "loss": 1.4219,
+      "step": 2044
+    },
+    {
+      "epoch": 0.9379658296066965,
+      "grad_norm": 0.361795574426651,
+      "learning_rate": 0.0008047816809784502,
+      "loss": 1.9539,
+      "step": 2045
+    },
+    {
+      "epoch": 0.9384244926040591,
+      "grad_norm": 0.37029165029525757,
+      "learning_rate": 0.0008045853281507827,
+      "loss": 1.7736,
+      "step": 2046
+    },
+    {
+      "epoch": 0.9388831556014219,
+      "grad_norm": 0.3089427053928375,
+      "learning_rate": 0.0008043889006061378,
+      "loss": 1.5316,
+      "step": 2047
+    },
+    {
+      "epoch": 0.9393418185987845,
+      "grad_norm": 0.22889067232608795,
+      "learning_rate": 0.0008041923983927009,
+      "loss": 0.8428,
+      "step": 2048
+    },
+    {
+      "epoch": 0.9398004815961473,
+      "grad_norm": 0.3076949417591095,
+      "learning_rate": 0.0008039958215586753,
+      "loss": 1.6298,
+      "step": 2049
+    },
+    {
+      "epoch": 0.9402591445935099,
+      "grad_norm": 0.3263643682003021,
+      "learning_rate": 0.0008037991701522829,
+      "loss": 1.8137,
+      "step": 2050
+    },
+    {
+      "epoch": 0.9407178075908726,
+      "grad_norm": 0.3142676055431366,
+      "learning_rate": 0.0008036024442217636,
+      "loss": 1.4615,
+      "step": 2051
+    },
+    {
+      "epoch": 0.9411764705882353,
+      "grad_norm": 0.27744585275650024,
+      "learning_rate": 0.0008034056438153753,
+      "loss": 1.2395,
+      "step": 2052
+    },
+    {
+      "epoch": 0.941635133585598,
+      "grad_norm": 0.21574388444423676,
+      "learning_rate": 0.0008032087689813952,
+      "loss": 0.914,
+      "step": 2053
+    },
+    {
+      "epoch": 0.9420937965829607,
+      "grad_norm": 0.1983022391796112,
+      "learning_rate": 0.0008030118197681178,
+      "loss": 1.0479,
+      "step": 2054
+    },
+    {
+      "epoch": 0.9425524595803234,
+      "grad_norm": 0.2801907956600189,
+      "learning_rate": 0.0008028147962238561,
+      "loss": 1.3931,
+      "step": 2055
+    },
+    {
+      "epoch": 0.943011122577686,
+      "grad_norm": 0.10424070060253143,
+      "learning_rate": 0.0008026176983969415,
+      "loss": 0.5858,
+      "step": 2056
+    },
+    {
+      "epoch": 0.9434697855750487,
+      "grad_norm": 0.260968953371048,
+      "learning_rate": 0.0008024205263357234,
+      "loss": 1.2817,
+      "step": 2057
+    },
+    {
+      "epoch": 0.9439284485724114,
+      "grad_norm": 0.24895592033863068,
+      "learning_rate": 0.0008022232800885697,
+      "loss": 1.2282,
+      "step": 2058
+    },
+    {
+      "epoch": 0.9443871115697741,
+      "grad_norm": 0.35649770498275757,
+      "learning_rate": 0.0008020259597038663,
+      "loss": 1.7548,
+      "step": 2059
+    },
+    {
+      "epoch": 0.9448457745671368,
+      "grad_norm": 0.3545326292514801,
+      "learning_rate": 0.0008018285652300173,
+      "loss": 1.819,
+      "step": 2060
+    },
+    {
+      "epoch": 0.9453044375644994,
+      "grad_norm": 0.35756614804267883,
+      "learning_rate": 0.0008016310967154448,
+      "loss": 1.8008,
+      "step": 2061
+    },
+    {
+      "epoch": 0.9457631005618622,
+      "grad_norm": 0.12270601838827133,
+      "learning_rate": 0.0008014335542085896,
+      "loss": 0.6562,
+      "step": 2062
+    },
+    {
+      "epoch": 0.9462217635592248,
+      "grad_norm": 0.3161708414554596,
+      "learning_rate": 0.0008012359377579099,
+      "loss": 1.5311,
+      "step": 2063
+    },
+    {
+      "epoch": 0.9466804265565876,
+      "grad_norm": 0.20928043127059937,
+      "learning_rate": 0.0008010382474118827,
+      "loss": 0.9492,
+      "step": 2064
+    },
+    {
+      "epoch": 0.9471390895539502,
+      "grad_norm": 0.3087616562843323,
+      "learning_rate": 0.0008008404832190028,
+      "loss": 1.3438,
+      "step": 2065
+    },
+    {
+      "epoch": 0.947597752551313,
+      "grad_norm": 0.24473027884960175,
+      "learning_rate": 0.0008006426452277833,
+      "loss": 1.151,
+      "step": 2066
+    },
+    {
+      "epoch": 0.9480564155486756,
+      "grad_norm": 0.08595964312553406,
+      "learning_rate": 0.0008004447334867551,
+      "loss": 0.43,
+      "step": 2067
+    },
+    {
+      "epoch": 0.9485150785460383,
+      "grad_norm": 0.5434077382087708,
+      "learning_rate": 0.0008002467480444675,
+      "loss": 2.1346,
+      "step": 2068
+    },
+    {
+      "epoch": 0.948973741543401,
+      "grad_norm": 0.32616284489631653,
+      "learning_rate": 0.0008000486889494877,
+      "loss": 1.3558,
+      "step": 2069
+    },
+    {
+      "epoch": 0.9494324045407637,
+      "grad_norm": 0.24117501080036163,
+      "learning_rate": 0.000799850556250401,
+      "loss": 1.0023,
+      "step": 2070
+    },
+    {
+      "epoch": 0.9498910675381264,
+      "grad_norm": 0.26819151639938354,
+      "learning_rate": 0.0007996523499958109,
+      "loss": 1.0614,
+      "step": 2071
+    },
+    {
+      "epoch": 0.9503497305354891,
+      "grad_norm": 0.28361374139785767,
+      "learning_rate": 0.0007994540702343386,
+      "loss": 1.374,
+      "step": 2072
+    },
+    {
+      "epoch": 0.9508083935328517,
+      "grad_norm": 0.4279637038707733,
+      "learning_rate": 0.000799255717014624,
+      "loss": 1.9542,
+      "step": 2073
+    },
+    {
+      "epoch": 0.9512670565302144,
+      "grad_norm": 0.4314320981502533,
+      "learning_rate": 0.0007990572903853239,
+      "loss": 1.9646,
+      "step": 2074
+    },
+    {
+      "epoch": 0.9517257195275771,
+      "grad_norm": 0.3635926842689514,
+      "learning_rate": 0.0007988587903951143,
+      "loss": 1.699,
+      "step": 2075
+    },
+    {
+      "epoch": 0.9521843825249398,
+      "grad_norm": 0.3035918176174164,
+      "learning_rate": 0.0007986602170926885,
+      "loss": 1.4184,
+      "step": 2076
+    },
+    {
+      "epoch": 0.9526430455223025,
+      "grad_norm": 0.38614422082901,
+      "learning_rate": 0.0007984615705267581,
+      "loss": 2.1599,
+      "step": 2077
+    },
+    {
+      "epoch": 0.9531017085196651,
+      "grad_norm": 0.3324802815914154,
+      "learning_rate": 0.0007982628507460523,
+      "loss": 1.2087,
+      "step": 2078
+    },
+    {
+      "epoch": 0.9535603715170279,
+      "grad_norm": 0.243191197514534,
+      "learning_rate": 0.0007980640577993187,
+      "loss": 0.9754,
+      "step": 2079
+    },
+    {
+      "epoch": 0.9540190345143905,
+      "grad_norm": 0.2958884537220001,
+      "learning_rate": 0.0007978651917353225,
+      "loss": 1.3487,
+      "step": 2080
+    },
+    {
+      "epoch": 0.9544776975117533,
+      "grad_norm": 0.17807622253894806,
+      "learning_rate": 0.0007976662526028473,
+      "loss": 0.9265,
+      "step": 2081
+    },
+    {
+      "epoch": 0.9549363605091159,
+      "grad_norm": 0.20209459960460663,
+      "learning_rate": 0.0007974672404506937,
+      "loss": 0.9037,
+      "step": 2082
+    },
+    {
+      "epoch": 0.9553950235064786,
+      "grad_norm": 0.31564533710479736,
+      "learning_rate": 0.0007972681553276813,
+      "loss": 1.7772,
+      "step": 2083
+    },
+    {
+      "epoch": 0.9558536865038413,
+      "grad_norm": 0.4125016927719116,
+      "learning_rate": 0.0007970689972826471,
+      "loss": 2.0496,
+      "step": 2084
+    },
+    {
+      "epoch": 0.956312349501204,
+      "grad_norm": 0.49603113532066345,
+      "learning_rate": 0.0007968697663644457,
+      "loss": 1.1933,
+      "step": 2085
+    },
+    {
+      "epoch": 0.9567710124985667,
+      "grad_norm": 0.2608060836791992,
+      "learning_rate": 0.0007966704626219503,
+      "loss": 1.3788,
+      "step": 2086
+    },
+    {
+      "epoch": 0.9572296754959294,
+      "grad_norm": 0.3525409400463104,
+      "learning_rate": 0.0007964710861040512,
+      "loss": 1.7322,
+      "step": 2087
+    },
+    {
+      "epoch": 0.957688338493292,
+      "grad_norm": 0.3033929169178009,
+      "learning_rate": 0.000796271636859657,
+      "loss": 1.6537,
+      "step": 2088
+    },
+    {
+      "epoch": 0.9581470014906548,
+      "grad_norm": 0.24828091263771057,
+      "learning_rate": 0.000796072114937694,
+      "loss": 1.1296,
+      "step": 2089
+    },
+    {
+      "epoch": 0.9586056644880174,
+      "grad_norm": 0.2196667641401291,
+      "learning_rate": 0.0007958725203871064,
+      "loss": 1.0247,
+      "step": 2090
+    },
+    {
+      "epoch": 0.9590643274853801,
+      "grad_norm": 0.2858245074748993,
+      "learning_rate": 0.0007956728532568563,
+      "loss": 1.3901,
+      "step": 2091
+    },
+    {
+      "epoch": 0.9595229904827428,
+      "grad_norm": 0.2958674430847168,
+      "learning_rate": 0.0007954731135959235,
+      "loss": 1.3847,
+      "step": 2092
+    },
+    {
+      "epoch": 0.9599816534801054,
+      "grad_norm": 0.25934916734695435,
+      "learning_rate": 0.0007952733014533051,
+      "loss": 1.3095,
+      "step": 2093
+    },
+    {
+      "epoch": 0.9604403164774682,
+      "grad_norm": 0.26100558042526245,
+      "learning_rate": 0.0007950734168780171,
+      "loss": 1.519,
+      "step": 2094
+    },
+    {
+      "epoch": 0.9608989794748308,
+      "grad_norm": 0.18241111934185028,
+      "learning_rate": 0.0007948734599190924,
+      "loss": 0.9885,
+      "step": 2095
+    },
+    {
+      "epoch": 0.9613576424721936,
+      "grad_norm": 1.0444318056106567,
+      "learning_rate": 0.0007946734306255816,
+      "loss": 1.6812,
+      "step": 2096
+    },
+    {
+      "epoch": 0.9618163054695562,
+      "grad_norm": 0.31861168146133423,
+      "learning_rate": 0.0007944733290465535,
+      "loss": 2.0449,
+      "step": 2097
+    },
+    {
+      "epoch": 0.962274968466919,
+      "grad_norm": 0.28558486700057983,
+      "learning_rate": 0.0007942731552310949,
+      "loss": 1.6525,
+      "step": 2098
+    },
+    {
+      "epoch": 0.9627336314642816,
+      "grad_norm": 0.23924612998962402,
+      "learning_rate": 0.0007940729092283092,
+      "loss": 1.2884,
+      "step": 2099
+    },
+    {
+      "epoch": 0.9631922944616443,
+      "grad_norm": 0.2011316865682602,
+      "learning_rate": 0.0007938725910873186,
+      "loss": 1.3885,
+      "step": 2100
+    },
+    {
+      "epoch": 0.963650957459007,
+      "grad_norm": 0.31433144211769104,
+      "learning_rate": 0.0007936722008572625,
+      "loss": 1.9054,
+      "step": 2101
+    },
+    {
+      "epoch": 0.9641096204563697,
+      "grad_norm": 0.26311978697776794,
+      "learning_rate": 0.000793471738587298,
+      "loss": 1.584,
+      "step": 2102
+    },
+    {
+      "epoch": 0.9645682834537324,
+      "grad_norm": 0.27694377303123474,
+      "learning_rate": 0.0007932712043266,
+      "loss": 1.2675,
+      "step": 2103
+    },
+    {
+      "epoch": 0.9650269464510951,
+      "grad_norm": 0.15894590318202972,
+      "learning_rate": 0.000793070598124361,
+      "loss": 0.8647,
+      "step": 2104
+    },
+    {
+      "epoch": 0.9654856094484577,
+      "grad_norm": 0.3218410909175873,
+      "learning_rate": 0.000792869920029791,
+      "loss": 1.8047,
+      "step": 2105
+    },
+    {
+      "epoch": 0.9659442724458205,
+      "grad_norm": 0.18074116110801697,
+      "learning_rate": 0.0007926691700921181,
+      "loss": 1.0374,
+      "step": 2106
+    },
+    {
+      "epoch": 0.9664029354431831,
+      "grad_norm": 0.3340591490268707,
+      "learning_rate": 0.0007924683483605875,
+      "loss": 1.8395,
+      "step": 2107
+    },
+    {
+      "epoch": 0.9668615984405458,
+      "grad_norm": 0.30720779299736023,
+      "learning_rate": 0.0007922674548844622,
+      "loss": 1.6614,
+      "step": 2108
+    },
+    {
+      "epoch": 0.9673202614379085,
+      "grad_norm": 0.275790810585022,
+      "learning_rate": 0.0007920664897130228,
+      "loss": 1.6523,
+      "step": 2109
+    },
+    {
+      "epoch": 0.9677789244352711,
+      "grad_norm": 0.0825454369187355,
+      "learning_rate": 0.0007918654528955675,
+      "loss": 0.4651,
+      "step": 2110
+    },
+    {
+      "epoch": 0.9682375874326339,
+      "grad_norm": 0.31684210896492004,
+      "learning_rate": 0.0007916643444814123,
+      "loss": 1.9248,
+      "step": 2111
+    },
+    {
+      "epoch": 0.9686962504299965,
+      "grad_norm": 0.18997028470039368,
+      "learning_rate": 0.0007914631645198904,
+      "loss": 1.0344,
+      "step": 2112
+    },
+    {
+      "epoch": 0.9691549134273593,
+      "grad_norm": 0.24045085906982422,
+      "learning_rate": 0.0007912619130603527,
+      "loss": 1.2285,
+      "step": 2113
+    },
+    {
+      "epoch": 0.9696135764247219,
+      "grad_norm": 0.3353755474090576,
+      "learning_rate": 0.0007910605901521675,
+      "loss": 1.7612,
+      "step": 2114
+    },
+    {
+      "epoch": 0.9700722394220846,
+      "grad_norm": 0.32669898867607117,
+      "learning_rate": 0.0007908591958447209,
+      "loss": 1.4136,
+      "step": 2115
+    },
+    {
+      "epoch": 0.9705309024194473,
+      "grad_norm": 0.37192410230636597,
+      "learning_rate": 0.0007906577301874163,
+      "loss": 1.8505,
+      "step": 2116
+    },
+    {
+      "epoch": 0.97098956541681,
+      "grad_norm": 0.2398930788040161,
+      "learning_rate": 0.000790456193229675,
+      "loss": 1.0386,
+      "step": 2117
+    },
+    {
+      "epoch": 0.9714482284141727,
+      "grad_norm": 0.3819679319858551,
+      "learning_rate": 0.0007902545850209349,
+      "loss": 1.7427,
+      "step": 2118
+    },
+    {
+      "epoch": 0.9719068914115354,
+      "grad_norm": 0.2345220297574997,
+      "learning_rate": 0.0007900529056106525,
+      "loss": 1.2595,
+      "step": 2119
+    },
+    {
+      "epoch": 0.972365554408898,
+      "grad_norm": 0.3607707917690277,
+      "learning_rate": 0.0007898511550483007,
+      "loss": 2.0488,
+      "step": 2120
+    },
+    {
+      "epoch": 0.9728242174062608,
+      "grad_norm": 0.33080729842185974,
+      "learning_rate": 0.0007896493333833707,
+      "loss": 1.4388,
+      "step": 2121
+    },
+    {
+      "epoch": 0.9732828804036234,
+      "grad_norm": 0.2641333341598511,
+      "learning_rate": 0.0007894474406653709,
+      "loss": 1.5211,
+      "step": 2122
+    },
+    {
+      "epoch": 0.9737415434009862,
+      "grad_norm": 0.18903343379497528,
+      "learning_rate": 0.0007892454769438268,
+      "loss": 0.7437,
+      "step": 2123
+    },
+    {
+      "epoch": 0.9742002063983488,
+      "grad_norm": 0.2841821014881134,
+      "learning_rate": 0.0007890434422682817,
+      "loss": 1.4099,
+      "step": 2124
+    },
+    {
+      "epoch": 0.9746588693957114,
+      "grad_norm": 0.18609897792339325,
+      "learning_rate": 0.0007888413366882959,
+      "loss": 0.9775,
+      "step": 2125
+    },
+    {
+      "epoch": 0.9751175323930742,
+      "grad_norm": 0.3263588845729828,
+      "learning_rate": 0.0007886391602534477,
+      "loss": 1.9374,
+      "step": 2126
+    },
+    {
+      "epoch": 0.9755761953904368,
+      "grad_norm": 0.29043829441070557,
+      "learning_rate": 0.0007884369130133325,
+      "loss": 1.6285,
+      "step": 2127
+    },
+    {
+      "epoch": 0.9760348583877996,
+      "grad_norm": 0.23733757436275482,
+      "learning_rate": 0.0007882345950175624,
+      "loss": 1.009,
+      "step": 2128
+    },
+    {
+      "epoch": 0.9764935213851622,
+      "grad_norm": 0.28636664152145386,
+      "learning_rate": 0.0007880322063157681,
+      "loss": 1.4238,
+      "step": 2129
+    },
+    {
+      "epoch": 0.976952184382525,
+      "grad_norm": 0.24077756702899933,
+      "learning_rate": 0.0007878297469575967,
+      "loss": 1.3356,
+      "step": 2130
+    },
+    {
+      "epoch": 0.9774108473798876,
+      "grad_norm": 0.21748611330986023,
+      "learning_rate": 0.000787627216992713,
+      "loss": 1.0264,
+      "step": 2131
+    },
+    {
+      "epoch": 0.9778695103772503,
+      "grad_norm": 0.17146669328212738,
+      "learning_rate": 0.0007874246164707991,
+      "loss": 0.9994,
+      "step": 2132
+    },
+    {
+      "epoch": 0.978328173374613,
+      "grad_norm": 0.28764429688453674,
+      "learning_rate": 0.0007872219454415543,
+      "loss": 1.8247,
+      "step": 2133
+    },
+    {
+      "epoch": 0.9787868363719757,
+      "grad_norm": 0.3070955276489258,
+      "learning_rate": 0.0007870192039546954,
+      "loss": 1.7511,
+      "step": 2134
+    },
+    {
+      "epoch": 0.9792454993693384,
+      "grad_norm": 0.24031320214271545,
+      "learning_rate": 0.0007868163920599563,
+      "loss": 1.4128,
+      "step": 2135
+    },
+    {
+      "epoch": 0.9797041623667011,
+      "grad_norm": 0.260254442691803,
+      "learning_rate": 0.000786613509807088,
+      "loss": 1.5773,
+      "step": 2136
+    },
+    {
+      "epoch": 0.9801628253640637,
+      "grad_norm": 0.25486305356025696,
+      "learning_rate": 0.0007864105572458592,
+      "loss": 1.146,
+      "step": 2137
+    },
+    {
+      "epoch": 0.9806214883614265,
+      "grad_norm": 0.2558663785457611,
+      "learning_rate": 0.0007862075344260555,
+      "loss": 1.3716,
+      "step": 2138
+    },
+    {
+      "epoch": 0.9810801513587891,
+      "grad_norm": 0.2597788870334625,
+      "learning_rate": 0.0007860044413974801,
+      "loss": 1.4079,
+      "step": 2139
+    },
+    {
+      "epoch": 0.9815388143561519,
+      "grad_norm": 0.2434176504611969,
+      "learning_rate": 0.000785801278209953,
+      "loss": 1.3001,
+      "step": 2140
+    },
+    {
+      "epoch": 0.9819974773535145,
+      "grad_norm": 0.24413499236106873,
+      "learning_rate": 0.0007855980449133116,
+      "loss": 1.4407,
+      "step": 2141
+    },
+    {
+      "epoch": 0.9824561403508771,
+      "grad_norm": 0.24809010326862335,
+      "learning_rate": 0.0007853947415574106,
+      "loss": 1.393,
+      "step": 2142
+    },
+    {
+      "epoch": 0.9829148033482399,
+      "grad_norm": 0.2992349863052368,
+      "learning_rate": 0.0007851913681921214,
+      "loss": 1.4247,
+      "step": 2143
+    },
+    {
+      "epoch": 0.9833734663456025,
+      "grad_norm": 0.28984925150871277,
+      "learning_rate": 0.0007849879248673336,
+      "loss": 1.7083,
+      "step": 2144
+    },
+    {
+      "epoch": 0.9838321293429653,
+      "grad_norm": 0.2734730839729309,
+      "learning_rate": 0.0007847844116329527,
+      "loss": 1.2769,
+      "step": 2145
+    },
+    {
+      "epoch": 0.9842907923403279,
+      "grad_norm": 0.23517167568206787,
+      "learning_rate": 0.0007845808285389024,
+      "loss": 1.291,
+      "step": 2146
+    },
+    {
+      "epoch": 0.9847494553376906,
+      "grad_norm": 0.09210627526044846,
+      "learning_rate": 0.0007843771756351228,
+      "loss": 0.5888,
+      "step": 2147
+    },
+    {
+      "epoch": 0.9852081183350533,
+      "grad_norm": 0.25133419036865234,
+      "learning_rate": 0.0007841734529715717,
+      "loss": 1.1768,
+      "step": 2148
+    },
+    {
+      "epoch": 0.985666781332416,
+      "grad_norm": 0.2581869065761566,
+      "learning_rate": 0.0007839696605982235,
+      "loss": 1.5143,
+      "step": 2149
+    },
+    {
+      "epoch": 0.9861254443297787,
+      "grad_norm": 0.2132362276315689,
+      "learning_rate": 0.00078376579856507,
+      "loss": 0.9654,
+      "step": 2150
+    },
+    {
+      "epoch": 0.9865841073271414,
+      "grad_norm": 0.3260076344013214,
+      "learning_rate": 0.00078356186692212,
+      "loss": 1.6854,
+      "step": 2151
+    },
+    {
+      "epoch": 0.987042770324504,
+      "grad_norm": 0.1947498768568039,
+      "learning_rate": 0.0007833578657193996,
+      "loss": 0.923,
+      "step": 2152
+    },
+    {
+      "epoch": 0.9875014333218668,
+      "grad_norm": 0.25844818353652954,
+      "learning_rate": 0.0007831537950069516,
+      "loss": 1.3283,
+      "step": 2153
+    },
+    {
+      "epoch": 0.9879600963192294,
+      "grad_norm": 0.3847392499446869,
+      "learning_rate": 0.000782949654834836,
+      "loss": 1.8961,
+      "step": 2154
+    },
+    {
+      "epoch": 0.9884187593165922,
+      "grad_norm": 0.376164048910141,
+      "learning_rate": 0.00078274544525313,
+      "loss": 2.1665,
+      "step": 2155
+    },
+    {
+      "epoch": 0.9888774223139548,
+      "grad_norm": 0.14740854501724243,
+      "learning_rate": 0.0007825411663119274,
+      "loss": 0.7855,
+      "step": 2156
+    },
+    {
+      "epoch": 0.9893360853113176,
+      "grad_norm": 0.3205507695674896,
+      "learning_rate": 0.0007823368180613395,
+      "loss": 1.5332,
+      "step": 2157
+    },
+    {
+      "epoch": 0.9897947483086802,
+      "grad_norm": 0.2271261066198349,
+      "learning_rate": 0.0007821324005514945,
+      "loss": 0.9458,
+      "step": 2158
+    },
+    {
+      "epoch": 0.9902534113060428,
+      "grad_norm": 0.2652105987071991,
+      "learning_rate": 0.0007819279138325373,
+      "loss": 1.389,
+      "step": 2159
+    },
+    {
+      "epoch": 0.9907120743034056,
+      "grad_norm": 0.2753114104270935,
+      "learning_rate": 0.00078172335795463,
+      "loss": 1.2935,
+      "step": 2160
+    },
+    {
+      "epoch": 0.9911707373007682,
+      "grad_norm": 0.10524280369281769,
+      "learning_rate": 0.0007815187329679517,
+      "loss": 0.6075,
+      "step": 2161
+    },
+    {
+      "epoch": 0.991629400298131,
+      "grad_norm": 0.22543789446353912,
+      "learning_rate": 0.0007813140389226984,
+      "loss": 1.2902,
+      "step": 2162
+    },
+    {
+      "epoch": 0.9920880632954936,
+      "grad_norm": 0.31716790795326233,
+      "learning_rate": 0.0007811092758690829,
+      "loss": 1.8091,
+      "step": 2163
+    },
+    {
+      "epoch": 0.9925467262928563,
+      "grad_norm": 0.28479596972465515,
+      "learning_rate": 0.000780904443857335,
+      "loss": 1.4618,
+      "step": 2164
+    },
+    {
+      "epoch": 0.993005389290219,
+      "grad_norm": 0.18860182166099548,
+      "learning_rate": 0.0007806995429377019,
+      "loss": 0.9116,
+      "step": 2165
+    },
+    {
+      "epoch": 0.9934640522875817,
+      "grad_norm": 0.24793879687786102,
+      "learning_rate": 0.0007804945731604467,
+      "loss": 1.2674,
+      "step": 2166
+    },
+    {
+      "epoch": 0.9939227152849444,
+      "grad_norm": 0.2623526453971863,
+      "learning_rate": 0.0007802895345758503,
+      "loss": 1.4075,
+      "step": 2167
+    },
+    {
+      "epoch": 0.9943813782823071,
+      "grad_norm": 0.3471762239933014,
+      "learning_rate": 0.0007800844272342103,
+      "loss": 2.0403,
+      "step": 2168
+    },
+    {
+      "epoch": 0.9948400412796697,
+      "grad_norm": 0.29183194041252136,
+      "learning_rate": 0.0007798792511858404,
+      "loss": 1.6896,
+      "step": 2169
+    },
+    {
+      "epoch": 0.9952987042770325,
+      "grad_norm": 0.30205872654914856,
+      "learning_rate": 0.0007796740064810724,
+      "loss": 1.6495,
+      "step": 2170
+    },
+    {
+      "epoch": 0.9957573672743951,
+      "grad_norm": 0.2651442289352417,
+      "learning_rate": 0.000779468693170254,
+      "loss": 1.3665,
+      "step": 2171
+    },
+    {
+      "epoch": 0.9962160302717579,
+      "grad_norm": 0.29746365547180176,
+      "learning_rate": 0.0007792633113037501,
+      "loss": 1.4551,
+      "step": 2172
+    },
+    {
+      "epoch": 0.9966746932691205,
+      "grad_norm": 0.30460765957832336,
+      "learning_rate": 0.0007790578609319424,
+      "loss": 1.6595,
+      "step": 2173
+    },
+    {
+      "epoch": 0.9971333562664833,
+      "grad_norm": 0.2733561396598816,
+      "learning_rate": 0.0007788523421052291,
+      "loss": 0.9736,
+      "step": 2174
+    },
+    {
+      "epoch": 0.9975920192638459,
+      "grad_norm": 0.3780529499053955,
+      "learning_rate": 0.0007786467548740259,
+      "loss": 1.7864,
+      "step": 2175
+    },
+    {
+      "epoch": 0.9980506822612085,
+      "grad_norm": 0.16774733364582062,
+      "learning_rate": 0.0007784410992887645,
+      "loss": 0.9591,
+      "step": 2176
+    },
+    {
+      "epoch": 0.9985093452585713,
+      "grad_norm": 0.31271597743034363,
+      "learning_rate": 0.0007782353753998936,
+      "loss": 1.7136,
+      "step": 2177
+    },
+    {
+      "epoch": 0.9989680082559339,
+      "grad_norm": 0.298076331615448,
+      "learning_rate": 0.0007780295832578792,
+      "loss": 1.2218,
+      "step": 2178
+    },
+    {
+      "epoch": 0.9994266712532967,
+      "grad_norm": 0.37182536721229553,
+      "learning_rate": 0.0007778237229132032,
+      "loss": 1.7005,
+      "step": 2179
+    },
+    {
+      "epoch": 0.9998853342506593,
+      "grad_norm": 0.2658548355102539,
+      "learning_rate": 0.0007776177944163648,
+      "loss": 1.2988,
+      "step": 2180
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2658548355102539,
+      "learning_rate": 0.0007776177944163648,
+      "loss": 0.5049,
+      "step": 2181
+    },
+    {
+      "epoch": 1.0004586629973626,
+      "grad_norm": 0.3670017421245575,
+      "learning_rate": 0.0007774117978178797,
+      "loss": 1.684,
+      "step": 2182
+    },
+    {
+      "epoch": 1.0009173259947253,
+      "grad_norm": 0.281902939081192,
+      "learning_rate": 0.0007772057331682802,
+      "loss": 1.1651,
+      "step": 2183
+    },
+    {
+      "epoch": 1.0013759889920881,
+      "grad_norm": 0.21292293071746826,
+      "learning_rate": 0.0007769996005181159,
+      "loss": 1.2873,
+      "step": 2184
+    },
+    {
+      "epoch": 1.0018346519894508,
+      "grad_norm": 0.3060021996498108,
+      "learning_rate": 0.0007767933999179521,
+      "loss": 1.6379,
+      "step": 2185
+    },
+    {
+      "epoch": 1.0022933149868134,
+      "grad_norm": 0.6159801483154297,
+      "learning_rate": 0.0007765871314183715,
+      "loss": 1.8337,
+      "step": 2186
+    },
+    {
+      "epoch": 1.002751977984176,
+      "grad_norm": 0.27383625507354736,
+      "learning_rate": 0.0007763807950699734,
+      "loss": 1.1533,
+      "step": 2187
+    },
+    {
+      "epoch": 1.003210640981539,
+      "grad_norm": 0.26541003584861755,
+      "learning_rate": 0.0007761743909233733,
+      "loss": 1.7667,
+      "step": 2188
+    },
+    {
+      "epoch": 1.0036693039789015,
+      "grad_norm": 0.2529599070549011,
+      "learning_rate": 0.0007759679190292039,
+      "loss": 1.2758,
+      "step": 2189
+    },
+    {
+      "epoch": 1.0041279669762642,
+      "grad_norm": 0.3762117326259613,
+      "learning_rate": 0.0007757613794381142,
+      "loss": 1.4408,
+      "step": 2190
+    },
+    {
+      "epoch": 1.0045866299736268,
+      "grad_norm": 0.17387861013412476,
+      "learning_rate": 0.0007755547722007696,
+      "loss": 1.2704,
+      "step": 2191
+    },
+    {
+      "epoch": 1.0050452929709897,
+      "grad_norm": 0.3292040228843689,
+      "learning_rate": 0.0007753480973678527,
+      "loss": 1.5397,
+      "step": 2192
+    },
+    {
+      "epoch": 1.0055039559683523,
+      "grad_norm": 0.2807086408138275,
+      "learning_rate": 0.0007751413549900621,
+      "loss": 1.2153,
+      "step": 2193
+    },
+    {
+      "epoch": 1.005962618965715,
+      "grad_norm": 0.3431303799152374,
+      "learning_rate": 0.0007749345451181132,
+      "loss": 1.6964,
+      "step": 2194
+    },
+    {
+      "epoch": 1.0064212819630776,
+      "grad_norm": 0.32860153913497925,
+      "learning_rate": 0.0007747276678027379,
+      "loss": 2.2512,
+      "step": 2195
+    },
+    {
+      "epoch": 1.0068799449604404,
+      "grad_norm": 0.4267536401748657,
+      "learning_rate": 0.000774520723094685,
+      "loss": 2.2029,
+      "step": 2196
+    },
+    {
+      "epoch": 1.007338607957803,
+      "grad_norm": 0.3573954403400421,
+      "learning_rate": 0.0007743137110447194,
+      "loss": 1.3235,
+      "step": 2197
+    },
+    {
+      "epoch": 1.0077972709551657,
+      "grad_norm": 0.31035539507865906,
+      "learning_rate": 0.0007741066317036222,
+      "loss": 1.5737,
+      "step": 2198
+    },
+    {
+      "epoch": 1.0082559339525283,
+      "grad_norm": 0.43037089705467224,
+      "learning_rate": 0.0007738994851221921,
+      "loss": 1.7621,
+      "step": 2199
+    },
+    {
+      "epoch": 1.008714596949891,
+      "grad_norm": 0.3466168940067291,
+      "learning_rate": 0.0007736922713512434,
+      "loss": 2.3148,
+      "step": 2200
+    },
+    {
+      "epoch": 1.0091732599472538,
+      "grad_norm": 0.3968213200569153,
+      "learning_rate": 0.000773484990441607,
+      "loss": 2.1924,
+      "step": 2201
+    },
+    {
+      "epoch": 1.0096319229446165,
+      "grad_norm": 0.3038836419582367,
+      "learning_rate": 0.0007732776424441307,
+      "loss": 0.9113,
+      "step": 2202
+    },
+    {
+      "epoch": 1.010090585941979,
+      "grad_norm": 0.29169583320617676,
+      "learning_rate": 0.0007730702274096782,
+      "loss": 1.7387,
+      "step": 2203
+    },
+    {
+      "epoch": 1.0105492489393417,
+      "grad_norm": 0.37365761399269104,
+      "learning_rate": 0.0007728627453891297,
+      "loss": 2.0283,
+      "step": 2204
+    },
+    {
+      "epoch": 1.0110079119367046,
+      "grad_norm": 0.35608237981796265,
+      "learning_rate": 0.0007726551964333827,
+      "loss": 0.9617,
+      "step": 2205
+    },
+    {
+      "epoch": 1.0114665749340672,
+      "grad_norm": 0.3315924406051636,
+      "learning_rate": 0.0007724475805933498,
+      "loss": 1.8494,
+      "step": 2206
+    },
+    {
+      "epoch": 1.0119252379314299,
+      "grad_norm": 0.31007975339889526,
+      "learning_rate": 0.000772239897919961,
+      "loss": 1.2838,
+      "step": 2207
+    },
+    {
+      "epoch": 1.0123839009287925,
+      "grad_norm": 0.4036838710308075,
+      "learning_rate": 0.0007720321484641625,
+      "loss": 1.7657,
+      "step": 2208
+    },
+    {
+      "epoch": 1.0128425639261553,
+      "grad_norm": 0.32576027512550354,
+      "learning_rate": 0.0007718243322769163,
+      "loss": 1.1251,
+      "step": 2209
+    },
+    {
+      "epoch": 1.013301226923518,
+      "grad_norm": 0.3255734145641327,
+      "learning_rate": 0.0007716164494092014,
+      "loss": 1.7505,
+      "step": 2210
+    },
+    {
+      "epoch": 1.0137598899208806,
+      "grad_norm": 0.2708061635494232,
+      "learning_rate": 0.0007714084999120132,
+      "loss": 1.2809,
+      "step": 2211
+    },
+    {
+      "epoch": 1.0142185529182433,
+      "grad_norm": 0.3722837269306183,
+      "learning_rate": 0.0007712004838363629,
+      "loss": 0.8855,
+      "step": 2212
+    },
+    {
+      "epoch": 1.0146772159156061,
+      "grad_norm": 0.19533422589302063,
+      "learning_rate": 0.0007709924012332784,
+      "loss": 0.9815,
+      "step": 2213
+    },
+    {
+      "epoch": 1.0151358789129687,
+      "grad_norm": 0.14171691238880157,
+      "learning_rate": 0.000770784252153804,
+      "loss": 1.119,
+      "step": 2214
+    },
+    {
+      "epoch": 1.0155945419103314,
+      "grad_norm": 0.4291399121284485,
+      "learning_rate": 0.000770576036649,
+      "loss": 1.8247,
+      "step": 2215
+    },
+    {
+      "epoch": 1.016053204907694,
+      "grad_norm": 0.35333549976348877,
+      "learning_rate": 0.0007703677547699435,
+      "loss": 1.5409,
+      "step": 2216
+    },
+    {
+      "epoch": 1.0165118679050567,
+      "grad_norm": 0.6529225707054138,
+      "learning_rate": 0.000770159406567727,
+      "loss": 1.7983,
+      "step": 2217
+    },
+    {
+      "epoch": 1.0169705309024195,
+      "grad_norm": 0.1469832956790924,
+      "learning_rate": 0.0007699509920934603,
+      "loss": 0.5275,
+      "step": 2218
+    },
+    {
+      "epoch": 1.0174291938997821,
+      "grad_norm": 0.35177433490753174,
+      "learning_rate": 0.0007697425113982688,
+      "loss": 2.0906,
+      "step": 2219
+    },
+    {
+      "epoch": 1.0178878568971448,
+      "grad_norm": 0.3718166649341583,
+      "learning_rate": 0.0007695339645332943,
+      "loss": 1.5574,
+      "step": 2220
+    },
+    {
+      "epoch": 1.0183465198945074,
+      "grad_norm": 0.4496723413467407,
+      "learning_rate": 0.0007693253515496947,
+      "loss": 1.6213,
+      "step": 2221
+    },
+    {
+      "epoch": 1.0188051828918703,
+      "grad_norm": 0.3293614089488983,
+      "learning_rate": 0.0007691166724986447,
+      "loss": 1.3462,
+      "step": 2222
+    },
+    {
+      "epoch": 1.019263845889233,
+      "grad_norm": 0.2451786994934082,
+      "learning_rate": 0.0007689079274313342,
+      "loss": 1.3479,
+      "step": 2223
+    },
+    {
+      "epoch": 1.0197225088865955,
+      "grad_norm": 0.33322182297706604,
+      "learning_rate": 0.0007686991163989704,
+      "loss": 2.0833,
+      "step": 2224
+    },
+    {
+      "epoch": 1.0201811718839582,
+      "grad_norm": 0.18467935919761658,
+      "learning_rate": 0.000768490239452776,
+      "loss": 0.8448,
+      "step": 2225
+    },
+    {
+      "epoch": 1.020639834881321,
+      "grad_norm": 0.21402765810489655,
+      "learning_rate": 0.0007682812966439896,
+      "loss": 0.7428,
+      "step": 2226
+    },
+    {
+      "epoch": 1.0210984978786837,
+      "grad_norm": 0.2612096667289734,
+      "learning_rate": 0.0007680722880238669,
+      "loss": 1.5559,
+      "step": 2227
+    },
+    {
+      "epoch": 1.0215571608760463,
+      "grad_norm": 0.2521737217903137,
+      "learning_rate": 0.0007678632136436792,
+      "loss": 1.6891,
+      "step": 2228
+    },
+    {
+      "epoch": 1.022015823873409,
+      "grad_norm": 0.2869427502155304,
+      "learning_rate": 0.0007676540735547136,
+      "loss": 0.9844,
+      "step": 2229
+    },
+    {
+      "epoch": 1.0224744868707718,
+      "grad_norm": 0.17099249362945557,
+      "learning_rate": 0.0007674448678082741,
+      "loss": 1.3577,
+      "step": 2230
+    },
+    {
+      "epoch": 1.0229331498681344,
+      "grad_norm": 0.3532930612564087,
+      "learning_rate": 0.0007672355964556799,
+      "loss": 1.4924,
+      "step": 2231
+    },
+    {
+      "epoch": 1.023391812865497,
+      "grad_norm": 0.22827592492103577,
+      "learning_rate": 0.000767026259548267,
+      "loss": 0.8428,
+      "step": 2232
+    },
+    {
+      "epoch": 1.0238504758628597,
+      "grad_norm": 0.19980807602405548,
+      "learning_rate": 0.0007668168571373875,
+      "loss": 1.0052,
+      "step": 2233
+    },
+    {
+      "epoch": 1.0243091388602223,
+      "grad_norm": 0.1704528033733368,
+      "learning_rate": 0.000766607389274409,
+      "loss": 0.8841,
+      "step": 2234
+    },
+    {
+      "epoch": 1.0247678018575852,
+      "grad_norm": 0.28992947936058044,
+      "learning_rate": 0.0007663978560107155,
+      "loss": 1.5616,
+      "step": 2235
+    },
+    {
+      "epoch": 1.0252264648549478,
+      "grad_norm": 0.29282909631729126,
+      "learning_rate": 0.000766188257397707,
+      "loss": 1.7693,
+      "step": 2236
+    },
+    {
+      "epoch": 1.0256851278523105,
+      "grad_norm": 0.4391837418079376,
+      "learning_rate": 0.0007659785934867999,
+      "loss": 2.3022,
+      "step": 2237
+    },
+    {
+      "epoch": 1.026143790849673,
+      "grad_norm": 0.3151383399963379,
+      "learning_rate": 0.0007657688643294259,
+      "loss": 1.6134,
+      "step": 2238
+    },
+    {
+      "epoch": 1.026602453847036,
+      "grad_norm": 0.264639288187027,
+      "learning_rate": 0.0007655590699770331,
+      "loss": 1.0672,
+      "step": 2239
+    },
+    {
+      "epoch": 1.0270611168443986,
+      "grad_norm": 0.24356485903263092,
+      "learning_rate": 0.0007653492104810858,
+      "loss": 1.1269,
+      "step": 2240
+    },
+    {
+      "epoch": 1.0275197798417612,
+      "grad_norm": 0.191898375749588,
+      "learning_rate": 0.000765139285893064,
+      "loss": 0.8673,
+      "step": 2241
+    },
+    {
+      "epoch": 1.0279784428391239,
+      "grad_norm": 0.2663523256778717,
+      "learning_rate": 0.0007649292962644634,
+      "loss": 1.2664,
+      "step": 2242
+    },
+    {
+      "epoch": 1.0284371058364867,
+      "grad_norm": 0.24687746167182922,
+      "learning_rate": 0.0007647192416467962,
+      "loss": 1.3866,
+      "step": 2243
+    },
+    {
+      "epoch": 1.0288957688338494,
+      "grad_norm": 0.26128000020980835,
+      "learning_rate": 0.0007645091220915904,
+      "loss": 1.1718,
+      "step": 2244
+    },
+    {
+      "epoch": 1.029354431831212,
+      "grad_norm": 0.3191836476325989,
+      "learning_rate": 0.0007642989376503897,
+      "loss": 1.4741,
+      "step": 2245
+    },
+    {
+      "epoch": 1.0298130948285746,
+      "grad_norm": 0.11389821767807007,
+      "learning_rate": 0.0007640886883747539,
+      "loss": 1.0261,
+      "step": 2246
+    },
+    {
+      "epoch": 1.0302717578259375,
+      "grad_norm": 0.4371493458747864,
+      "learning_rate": 0.0007638783743162586,
+      "loss": 1.3102,
+      "step": 2247
+    },
+    {
+      "epoch": 1.0307304208233001,
+      "grad_norm": 0.2924779951572418,
+      "learning_rate": 0.0007636679955264954,
+      "loss": 1.5544,
+      "step": 2248
+    },
+    {
+      "epoch": 1.0311890838206628,
+      "grad_norm": 0.29356637597084045,
+      "learning_rate": 0.0007634575520570719,
+      "loss": 1.4492,
+      "step": 2249
+    },
+    {
+      "epoch": 1.0316477468180254,
+      "grad_norm": 0.41005346179008484,
+      "learning_rate": 0.0007632470439596113,
+      "loss": 1.8448,
+      "step": 2250
+    },
+    {
+      "epoch": 1.032106409815388,
+      "grad_norm": 0.3051200807094574,
+      "learning_rate": 0.0007630364712857525,
+      "loss": 1.9664,
+      "step": 2251
+    },
+    {
+      "epoch": 1.032565072812751,
+      "grad_norm": 0.29568442702293396,
+      "learning_rate": 0.0007628258340871507,
+      "loss": 1.519,
+      "step": 2252
+    },
+    {
+      "epoch": 1.0330237358101135,
+      "grad_norm": 0.34316301345825195,
+      "learning_rate": 0.0007626151324154768,
+      "loss": 1.6171,
+      "step": 2253
+    },
+    {
+      "epoch": 1.0334823988074762,
+      "grad_norm": 0.34773850440979004,
+      "learning_rate": 0.0007624043663224173,
+      "loss": 1.3993,
+      "step": 2254
+    },
+    {
+      "epoch": 1.0339410618048388,
+      "grad_norm": 0.4576634466648102,
+      "learning_rate": 0.000762193535859675,
+      "loss": 1.3873,
+      "step": 2255
+    },
+    {
+      "epoch": 1.0343997248022017,
+      "grad_norm": 0.3191813826560974,
+      "learning_rate": 0.0007619826410789676,
+      "loss": 1.6381,
+      "step": 2256
+    },
+    {
+      "epoch": 1.0348583877995643,
+      "grad_norm": 0.29539400339126587,
+      "learning_rate": 0.0007617716820320293,
+      "loss": 1.3324,
+      "step": 2257
+    },
+    {
+      "epoch": 1.035317050796927,
+      "grad_norm": 0.18559886515140533,
+      "learning_rate": 0.0007615606587706101,
+      "loss": 1.2413,
+      "step": 2258
+    },
+    {
+      "epoch": 1.0357757137942896,
+      "grad_norm": 0.33654916286468506,
+      "learning_rate": 0.0007613495713464752,
+      "loss": 1.215,
+      "step": 2259
+    },
+    {
+      "epoch": 1.0362343767916524,
+      "grad_norm": 0.2632749378681183,
+      "learning_rate": 0.0007611384198114061,
+      "loss": 1.8086,
+      "step": 2260
+    },
+    {
+      "epoch": 1.036693039789015,
+      "grad_norm": 0.36246350407600403,
+      "learning_rate": 0.0007609272042171998,
+      "loss": 1.32,
+      "step": 2261
+    },
+    {
+      "epoch": 1.0371517027863777,
+      "grad_norm": 0.21696679294109344,
+      "learning_rate": 0.0007607159246156688,
+      "loss": 1.3752,
+      "step": 2262
+    },
+    {
+      "epoch": 1.0376103657837403,
+      "grad_norm": 0.20672181248664856,
+      "learning_rate": 0.0007605045810586415,
+      "loss": 0.3842,
+      "step": 2263
+    },
+    {
+      "epoch": 1.0380690287811032,
+      "grad_norm": 0.22013358771800995,
+      "learning_rate": 0.0007602931735979624,
+      "loss": 0.9232,
+      "step": 2264
+    },
+    {
+      "epoch": 1.0385276917784658,
+      "grad_norm": 0.2625974118709564,
+      "learning_rate": 0.0007600817022854908,
+      "loss": 1.6846,
+      "step": 2265
+    },
+    {
+      "epoch": 1.0389863547758285,
+      "grad_norm": 0.2270309031009674,
+      "learning_rate": 0.0007598701671731025,
+      "loss": 0.6877,
+      "step": 2266
+    },
+    {
+      "epoch": 1.039445017773191,
+      "grad_norm": 0.35904768109321594,
+      "learning_rate": 0.0007596585683126883,
+      "loss": 1.9496,
+      "step": 2267
+    },
+    {
+      "epoch": 1.0399036807705537,
+      "grad_norm": 0.31437742710113525,
+      "learning_rate": 0.0007594469057561551,
+      "loss": 1.4388,
+      "step": 2268
+    },
+    {
+      "epoch": 1.0403623437679166,
+      "grad_norm": 0.31415170431137085,
+      "learning_rate": 0.0007592351795554254,
+      "loss": 2.0281,
+      "step": 2269
+    },
+    {
+      "epoch": 1.0408210067652792,
+      "grad_norm": 0.19663023948669434,
+      "learning_rate": 0.0007590233897624367,
+      "loss": 0.5938,
+      "step": 2270
+    },
+    {
+      "epoch": 1.0412796697626419,
+      "grad_norm": 0.16666308045387268,
+      "learning_rate": 0.0007588115364291429,
+      "loss": 1.0012,
+      "step": 2271
+    },
+    {
+      "epoch": 1.0417383327600045,
+      "grad_norm": 0.10785413533449173,
+      "learning_rate": 0.0007585996196075131,
+      "loss": 0.913,
+      "step": 2272
+    },
+    {
+      "epoch": 1.0421969957573674,
+      "grad_norm": 0.34523460268974304,
+      "learning_rate": 0.000758387639349532,
+      "loss": 1.8508,
+      "step": 2273
+    },
+    {
+      "epoch": 1.04265565875473,
+      "grad_norm": 0.25581368803977966,
+      "learning_rate": 0.0007581755957072,
+      "loss": 1.0048,
+      "step": 2274
+    },
+    {
+      "epoch": 1.0431143217520926,
+      "grad_norm": 0.2878674864768982,
+      "learning_rate": 0.0007579634887325328,
+      "loss": 1.2516,
+      "step": 2275
+    },
+    {
+      "epoch": 1.0435729847494553,
+      "grad_norm": 0.43240469694137573,
+      "learning_rate": 0.0007577513184775617,
+      "loss": 1.5632,
+      "step": 2276
+    },
+    {
+      "epoch": 1.0440316477468181,
+      "grad_norm": 0.25301140546798706,
+      "learning_rate": 0.0007575390849943337,
+      "loss": 1.3549,
+      "step": 2277
+    },
+    {
+      "epoch": 1.0444903107441807,
+      "grad_norm": 0.31979095935821533,
+      "learning_rate": 0.0007573267883349114,
+      "loss": 1.1457,
+      "step": 2278
+    },
+    {
+      "epoch": 1.0449489737415434,
+      "grad_norm": 1.364774465560913,
+      "learning_rate": 0.0007571144285513723,
+      "loss": 2.123,
+      "step": 2279
+    },
+    {
+      "epoch": 1.045407636738906,
+      "grad_norm": 0.30994194746017456,
+      "learning_rate": 0.00075690200569581,
+      "loss": 1.4648,
+      "step": 2280
+    },
+    {
+      "epoch": 1.0458662997362689,
+      "grad_norm": 0.3474901616573334,
+      "learning_rate": 0.0007566895198203334,
+      "loss": 1.8733,
+      "step": 2281
+    },
+    {
+      "epoch": 1.0463249627336315,
+      "grad_norm": 0.23491498827934265,
+      "learning_rate": 0.0007564769709770667,
+      "loss": 1.5338,
+      "step": 2282
+    },
+    {
+      "epoch": 1.0467836257309941,
+      "grad_norm": 0.25132763385772705,
+      "learning_rate": 0.0007562643592181498,
+      "loss": 1.3498,
+      "step": 2283
+    },
+    {
+      "epoch": 1.0472422887283568,
+      "grad_norm": 0.33392879366874695,
+      "learning_rate": 0.0007560516845957377,
+      "loss": 1.4968,
+      "step": 2284
+    },
+    {
+      "epoch": 1.0477009517257194,
+      "grad_norm": 0.3323450982570648,
+      "learning_rate": 0.0007558389471620013,
+      "loss": 1.1117,
+      "step": 2285
+    },
+    {
+      "epoch": 1.0481596147230823,
+      "grad_norm": 0.2603180408477783,
+      "learning_rate": 0.0007556261469691264,
+      "loss": 1.6954,
+      "step": 2286
+    },
+    {
+      "epoch": 1.048618277720445,
+      "grad_norm": 0.35003596544265747,
+      "learning_rate": 0.0007554132840693145,
+      "loss": 1.53,
+      "step": 2287
+    },
+    {
+      "epoch": 1.0490769407178075,
+      "grad_norm": 0.3515700101852417,
+      "learning_rate": 0.0007552003585147823,
+      "loss": 1.696,
+      "step": 2288
+    },
+    {
+      "epoch": 1.0495356037151702,
+      "grad_norm": 0.2830311059951782,
+      "learning_rate": 0.0007549873703577622,
+      "loss": 1.8374,
+      "step": 2289
+    },
+    {
+      "epoch": 1.049994266712533,
+      "grad_norm": 0.2728405296802521,
+      "learning_rate": 0.0007547743196505014,
+      "loss": 0.9555,
+      "step": 2290
+    },
+    {
+      "epoch": 1.0504529297098957,
+      "grad_norm": 0.3142251968383789,
+      "learning_rate": 0.0007545612064452632,
+      "loss": 1.657,
+      "step": 2291
+    },
+    {
+      "epoch": 1.0509115927072583,
+      "grad_norm": 0.2731315791606903,
+      "learning_rate": 0.0007543480307943256,
+      "loss": 0.8105,
+      "step": 2292
+    },
+    {
+      "epoch": 1.051370255704621,
+      "grad_norm": 0.2212861180305481,
+      "learning_rate": 0.0007541347927499818,
+      "loss": 1.3339,
+      "step": 2293
+    },
+    {
+      "epoch": 1.0518289187019838,
+      "grad_norm": 0.29469791054725647,
+      "learning_rate": 0.0007539214923645412,
+      "loss": 1.0019,
+      "step": 2294
+    },
+    {
+      "epoch": 1.0522875816993464,
+      "grad_norm": 0.20634829998016357,
+      "learning_rate": 0.0007537081296903277,
+      "loss": 1.3472,
+      "step": 2295
+    },
+    {
+      "epoch": 1.052746244696709,
+      "grad_norm": 0.3181491792201996,
+      "learning_rate": 0.0007534947047796805,
+      "loss": 2.062,
+      "step": 2296
+    },
+    {
+      "epoch": 1.0532049076940717,
+      "grad_norm": 0.31653597950935364,
+      "learning_rate": 0.0007532812176849545,
+      "loss": 1.8488,
+      "step": 2297
+    },
+    {
+      "epoch": 1.0536635706914346,
+      "grad_norm": 0.3137333393096924,
+      "learning_rate": 0.0007530676684585194,
+      "loss": 1.6195,
+      "step": 2298
+    },
+    {
+      "epoch": 1.0541222336887972,
+      "grad_norm": 0.3669377565383911,
+      "learning_rate": 0.0007528540571527607,
+      "loss": 1.928,
+      "step": 2299
+    },
+    {
+      "epoch": 1.0545808966861598,
+      "grad_norm": 0.35595008730888367,
+      "learning_rate": 0.0007526403838200786,
+      "loss": 2.1411,
+      "step": 2300
+    },
+    {
+      "epoch": 1.0550395596835225,
+      "grad_norm": 0.4003475606441498,
+      "learning_rate": 0.0007524266485128885,
+      "loss": 1.6777,
+      "step": 2301
+    },
+    {
+      "epoch": 1.0554982226808853,
+      "grad_norm": 0.2703254818916321,
+      "learning_rate": 0.0007522128512836217,
+      "loss": 1.7441,
+      "step": 2302
+    },
+    {
+      "epoch": 1.055956885678248,
+      "grad_norm": 0.3241822123527527,
+      "learning_rate": 0.0007519989921847236,
+      "loss": 1.4656,
+      "step": 2303
+    },
+    {
+      "epoch": 1.0564155486756106,
+      "grad_norm": 0.38216519355773926,
+      "learning_rate": 0.0007517850712686561,
+      "loss": 1.7497,
+      "step": 2304
+    },
+    {
+      "epoch": 1.0568742116729732,
+      "grad_norm": 0.24903887510299683,
+      "learning_rate": 0.0007515710885878948,
+      "loss": 1.7455,
+      "step": 2305
+    },
+    {
+      "epoch": 1.0573328746703359,
+      "grad_norm": 0.36446714401245117,
+      "learning_rate": 0.0007513570441949319,
+      "loss": 2.2563,
+      "step": 2306
+    },
+    {
+      "epoch": 1.0577915376676987,
+      "grad_norm": 0.23342153429985046,
+      "learning_rate": 0.0007511429381422734,
+      "loss": 0.7739,
+      "step": 2307
+    },
+    {
+      "epoch": 1.0582502006650614,
+      "grad_norm": 0.3443793058395386,
+      "learning_rate": 0.0007509287704824415,
+      "loss": 0.9252,
+      "step": 2308
+    },
+    {
+      "epoch": 1.058708863662424,
+      "grad_norm": 0.4298572540283203,
+      "learning_rate": 0.0007507145412679728,
+      "loss": 2.2942,
+      "step": 2309
+    },
+    {
+      "epoch": 1.0591675266597866,
+      "grad_norm": 0.2859693765640259,
+      "learning_rate": 0.0007505002505514194,
+      "loss": 1.3583,
+      "step": 2310
+    },
+    {
+      "epoch": 1.0596261896571495,
+      "grad_norm": 0.35848772525787354,
+      "learning_rate": 0.0007502858983853485,
+      "loss": 1.5612,
+      "step": 2311
+    },
+    {
+      "epoch": 1.0600848526545121,
+      "grad_norm": 0.1740964949131012,
+      "learning_rate": 0.000750071484822342,
+      "loss": 1.1107,
+      "step": 2312
+    },
+    {
+      "epoch": 1.0605435156518748,
+      "grad_norm": 0.24920308589935303,
+      "learning_rate": 0.000749857009914997,
+      "loss": 0.8362,
+      "step": 2313
+    },
+    {
+      "epoch": 1.0610021786492374,
+      "grad_norm": 0.07761970907449722,
+      "learning_rate": 0.000749642473715926,
+      "loss": 0.4695,
+      "step": 2314
+    },
+    {
+      "epoch": 1.0614608416466003,
+      "grad_norm": 0.17175906896591187,
+      "learning_rate": 0.0007494278762777562,
+      "loss": 1.2153,
+      "step": 2315
+    },
+    {
+      "epoch": 1.061919504643963,
+      "grad_norm": 0.20324799418449402,
+      "learning_rate": 0.0007492132176531299,
+      "loss": 0.6542,
+      "step": 2316
+    },
+    {
+      "epoch": 1.0623781676413255,
+      "grad_norm": 0.3163895308971405,
+      "learning_rate": 0.0007489984978947044,
+      "loss": 2.1238,
+      "step": 2317
+    },
+    {
+      "epoch": 1.0628368306386882,
+      "grad_norm": 0.36171120405197144,
+      "learning_rate": 0.000748783717055152,
+      "loss": 1.8789,
+      "step": 2318
+    },
+    {
+      "epoch": 1.0632954936360508,
+      "grad_norm": 0.6624128222465515,
+      "learning_rate": 0.0007485688751871597,
+      "loss": 2.0256,
+      "step": 2319
+    },
+    {
+      "epoch": 1.0637541566334137,
+      "grad_norm": 0.3431372046470642,
+      "learning_rate": 0.0007483539723434305,
+      "loss": 1.7518,
+      "step": 2320
+    },
+    {
+      "epoch": 1.0642128196307763,
+      "grad_norm": 0.3786202371120453,
+      "learning_rate": 0.0007481390085766808,
+      "loss": 1.8253,
+      "step": 2321
+    },
+    {
+      "epoch": 1.064671482628139,
+      "grad_norm": 0.30203521251678467,
+      "learning_rate": 0.0007479239839396434,
+      "loss": 1.3458,
+      "step": 2322
+    },
+    {
+      "epoch": 1.0651301456255016,
+      "grad_norm": 0.2725338637828827,
+      "learning_rate": 0.000747708898485065,
+      "loss": 1.3139,
+      "step": 2323
+    },
+    {
+      "epoch": 1.0655888086228644,
+      "grad_norm": 0.2869611084461212,
+      "learning_rate": 0.0007474937522657076,
+      "loss": 0.9489,
+      "step": 2324
+    },
+    {
+      "epoch": 1.066047471620227,
+      "grad_norm": 0.4075610637664795,
+      "learning_rate": 0.0007472785453343485,
+      "loss": 1.8008,
+      "step": 2325
+    },
+    {
+      "epoch": 1.0665061346175897,
+      "grad_norm": 0.16893206536769867,
+      "learning_rate": 0.000747063277743779,
+      "loss": 1.1216,
+      "step": 2326
+    },
+    {
+      "epoch": 1.0669647976149523,
+      "grad_norm": 0.3267589211463928,
+      "learning_rate": 0.0007468479495468061,
+      "loss": 1.4798,
+      "step": 2327
+    },
+    {
+      "epoch": 1.0674234606123152,
+      "grad_norm": 0.3620133101940155,
+      "learning_rate": 0.0007466325607962516,
+      "loss": 1.8303,
+      "step": 2328
+    },
+    {
+      "epoch": 1.0678821236096778,
+      "grad_norm": 0.3057517409324646,
+      "learning_rate": 0.0007464171115449512,
+      "loss": 1.6888,
+      "step": 2329
+    },
+    {
+      "epoch": 1.0683407866070405,
+      "grad_norm": 0.2740839719772339,
+      "learning_rate": 0.0007462016018457568,
+      "loss": 1.4534,
+      "step": 2330
+    },
+    {
+      "epoch": 1.068799449604403,
+      "grad_norm": 0.20154106616973877,
+      "learning_rate": 0.0007459860317515344,
+      "loss": 1.3666,
+      "step": 2331
+    },
+    {
+      "epoch": 1.069258112601766,
+      "grad_norm": 0.24330544471740723,
+      "learning_rate": 0.0007457704013151645,
+      "loss": 0.856,
+      "step": 2332
+    },
+    {
+      "epoch": 1.0697167755991286,
+      "grad_norm": 0.23237231373786926,
+      "learning_rate": 0.0007455547105895432,
+      "loss": 1.5328,
+      "step": 2333
+    },
+    {
+      "epoch": 1.0701754385964912,
+      "grad_norm": 0.24399161338806152,
+      "learning_rate": 0.0007453389596275808,
+      "loss": 1.0707,
+      "step": 2334
+    },
+    {
+      "epoch": 1.0706341015938539,
+      "grad_norm": 0.2570146322250366,
+      "learning_rate": 0.0007451231484822025,
+      "loss": 0.9601,
+      "step": 2335
+    },
+    {
+      "epoch": 1.0710927645912167,
+      "grad_norm": 0.33579733967781067,
+      "learning_rate": 0.0007449072772063486,
+      "loss": 1.7599,
+      "step": 2336
+    },
+    {
+      "epoch": 1.0715514275885794,
+      "grad_norm": 0.3293912410736084,
+      "learning_rate": 0.0007446913458529738,
+      "loss": 1.2646,
+      "step": 2337
+    },
+    {
+      "epoch": 1.072010090585942,
+      "grad_norm": 0.19118434190750122,
+      "learning_rate": 0.0007444753544750475,
+      "loss": 0.8713,
+      "step": 2338
+    },
+    {
+      "epoch": 1.0724687535833046,
+      "grad_norm": 0.1878962367773056,
+      "learning_rate": 0.0007442593031255539,
+      "loss": 1.1241,
+      "step": 2339
+    },
+    {
+      "epoch": 1.0729274165806673,
+      "grad_norm": 0.3359505534172058,
+      "learning_rate": 0.000744043191857492,
+      "loss": 1.3387,
+      "step": 2340
+    },
+    {
+      "epoch": 1.0733860795780301,
+      "grad_norm": 0.2292228788137436,
+      "learning_rate": 0.0007438270207238756,
+      "loss": 1.3343,
+      "step": 2341
+    },
+    {
+      "epoch": 1.0738447425753928,
+      "grad_norm": 0.41298791766166687,
+      "learning_rate": 0.000743610789777733,
+      "loss": 2.1614,
+      "step": 2342
+    },
+    {
+      "epoch": 1.0743034055727554,
+      "grad_norm": 0.28897812962532043,
+      "learning_rate": 0.0007433944990721071,
+      "loss": 1.3058,
+      "step": 2343
+    },
+    {
+      "epoch": 1.074762068570118,
+      "grad_norm": 0.13999620079994202,
+      "learning_rate": 0.0007431781486600556,
+      "loss": 0.7708,
+      "step": 2344
+    },
+    {
+      "epoch": 1.0752207315674809,
+      "grad_norm": 0.259048193693161,
+      "learning_rate": 0.0007429617385946507,
+      "loss": 1.1908,
+      "step": 2345
+    },
+    {
+      "epoch": 1.0756793945648435,
+      "grad_norm": 0.18018385767936707,
+      "learning_rate": 0.0007427452689289795,
+      "loss": 0.8871,
+      "step": 2346
+    },
+    {
+      "epoch": 1.0761380575622062,
+      "grad_norm": 0.2923835217952728,
+      "learning_rate": 0.0007425287397161437,
+      "loss": 1.6891,
+      "step": 2347
+    },
+    {
+      "epoch": 1.0765967205595688,
+      "grad_norm": 0.3825288414955139,
+      "learning_rate": 0.0007423121510092593,
+      "loss": 1.6979,
+      "step": 2348
+    },
+    {
+      "epoch": 1.0770553835569316,
+      "grad_norm": 0.3247275948524475,
+      "learning_rate": 0.000742095502861457,
+      "loss": 1.6702,
+      "step": 2349
+    },
+    {
+      "epoch": 1.0775140465542943,
+      "grad_norm": 0.30282968282699585,
+      "learning_rate": 0.0007418787953258822,
+      "loss": 1.756,
+      "step": 2350
+    },
+    {
+      "epoch": 1.077972709551657,
+      "grad_norm": 0.3590962588787079,
+      "learning_rate": 0.000741662028455695,
+      "loss": 2.1335,
+      "step": 2351
+    },
+    {
+      "epoch": 1.0784313725490196,
+      "grad_norm": 0.3158145546913147,
+      "learning_rate": 0.0007414452023040697,
+      "loss": 1.6022,
+      "step": 2352
+    },
+    {
+      "epoch": 1.0788900355463822,
+      "grad_norm": 0.26424726843833923,
+      "learning_rate": 0.0007412283169241955,
+      "loss": 1.3064,
+      "step": 2353
+    },
+    {
+      "epoch": 1.079348698543745,
+      "grad_norm": 0.3655887544155121,
+      "learning_rate": 0.0007410113723692756,
+      "loss": 1.7661,
+      "step": 2354
+    },
+    {
+      "epoch": 1.0798073615411077,
+      "grad_norm": 0.22848142683506012,
+      "learning_rate": 0.0007407943686925282,
+      "loss": 0.8983,
+      "step": 2355
+    },
+    {
+      "epoch": 1.0802660245384703,
+      "grad_norm": 0.2924310863018036,
+      "learning_rate": 0.0007405773059471863,
+      "loss": 1.9707,
+      "step": 2356
+    },
+    {
+      "epoch": 1.080724687535833,
+      "grad_norm": 0.27466198801994324,
+      "learning_rate": 0.0007403601841864964,
+      "loss": 1.3336,
+      "step": 2357
+    },
+    {
+      "epoch": 1.0811833505331958,
+      "grad_norm": 0.2992285490036011,
+      "learning_rate": 0.0007401430034637202,
+      "loss": 0.8565,
+      "step": 2358
+    },
+    {
+      "epoch": 1.0816420135305584,
+      "grad_norm": 0.27274230122566223,
+      "learning_rate": 0.0007399257638321338,
+      "loss": 1.6908,
+      "step": 2359
+    },
+    {
+      "epoch": 1.082100676527921,
+      "grad_norm": 0.2696456015110016,
+      "learning_rate": 0.0007397084653450274,
+      "loss": 0.8956,
+      "step": 2360
+    },
+    {
+      "epoch": 1.0825593395252837,
+      "grad_norm": 0.18956485390663147,
+      "learning_rate": 0.0007394911080557063,
+      "loss": 1.3373,
+      "step": 2361
+    },
+    {
+      "epoch": 1.0830180025226466,
+      "grad_norm": 0.36281585693359375,
+      "learning_rate": 0.0007392736920174895,
+      "loss": 1.6471,
+      "step": 2362
+    },
+    {
+      "epoch": 1.0834766655200092,
+      "grad_norm": 0.30485787987709045,
+      "learning_rate": 0.0007390562172837108,
+      "loss": 1.2273,
+      "step": 2363
+    },
+    {
+      "epoch": 1.0839353285173718,
+      "grad_norm": 0.4604003131389618,
+      "learning_rate": 0.0007388386839077182,
+      "loss": 1.9656,
+      "step": 2364
+    },
+    {
+      "epoch": 1.0843939915147345,
+      "grad_norm": 0.21141566336154938,
+      "learning_rate": 0.0007386210919428744,
+      "loss": 1.2972,
+      "step": 2365
+    },
+    {
+      "epoch": 1.0848526545120973,
+      "grad_norm": 0.26035094261169434,
+      "learning_rate": 0.0007384034414425562,
+      "loss": 1.1507,
+      "step": 2366
+    },
+    {
+      "epoch": 1.08531131750946,
+      "grad_norm": 0.2898489832878113,
+      "learning_rate": 0.000738185732460155,
+      "loss": 1.4745,
+      "step": 2367
+    },
+    {
+      "epoch": 1.0857699805068226,
+      "grad_norm": 0.3080480992794037,
+      "learning_rate": 0.000737967965049076,
+      "loss": 1.7976,
+      "step": 2368
+    },
+    {
+      "epoch": 1.0862286435041852,
+      "grad_norm": 0.2572938799858093,
+      "learning_rate": 0.0007377501392627394,
+      "loss": 1.108,
+      "step": 2369
+    },
+    {
+      "epoch": 1.086687306501548,
+      "grad_norm": 0.2855343520641327,
+      "learning_rate": 0.0007375322551545794,
+      "loss": 1.2778,
+      "step": 2370
+    },
+    {
+      "epoch": 1.0871459694989107,
+      "grad_norm": 0.22971484065055847,
+      "learning_rate": 0.0007373143127780444,
+      "loss": 1.0277,
+      "step": 2371
+    },
+    {
+      "epoch": 1.0876046324962734,
+      "grad_norm": 0.41709139943122864,
+      "learning_rate": 0.0007370963121865974,
+      "loss": 2.3818,
+      "step": 2372
+    },
+    {
+      "epoch": 1.088063295493636,
+      "grad_norm": 0.21090003848075867,
+      "learning_rate": 0.0007368782534337156,
+      "loss": 1.0302,
+      "step": 2373
+    },
+    {
+      "epoch": 1.0885219584909986,
+      "grad_norm": 0.2724030315876007,
+      "learning_rate": 0.0007366601365728902,
+      "loss": 1.4246,
+      "step": 2374
+    },
+    {
+      "epoch": 1.0889806214883615,
+      "grad_norm": 0.45919355750083923,
+      "learning_rate": 0.0007364419616576268,
+      "loss": 2.0112,
+      "step": 2375
+    },
+    {
+      "epoch": 1.0894392844857241,
+      "grad_norm": 0.4057621657848358,
+      "learning_rate": 0.0007362237287414455,
+      "loss": 2.0664,
+      "step": 2376
+    },
+    {
+      "epoch": 1.0898979474830868,
+      "grad_norm": 0.2805062532424927,
+      "learning_rate": 0.0007360054378778801,
+      "loss": 1.3488,
+      "step": 2377
+    },
+    {
+      "epoch": 1.0903566104804494,
+      "grad_norm": 0.38183027505874634,
+      "learning_rate": 0.0007357870891204792,
+      "loss": 1.8075,
+      "step": 2378
+    },
+    {
+      "epoch": 1.0908152734778123,
+      "grad_norm": 0.22754809260368347,
+      "learning_rate": 0.0007355686825228053,
+      "loss": 0.8548,
+      "step": 2379
+    },
+    {
+      "epoch": 1.091273936475175,
+      "grad_norm": 0.3018571138381958,
+      "learning_rate": 0.0007353502181384349,
+      "loss": 1.9176,
+      "step": 2380
+    },
+    {
+      "epoch": 1.0917325994725375,
+      "grad_norm": 0.3934653103351593,
+      "learning_rate": 0.0007351316960209591,
+      "loss": 2.1248,
+      "step": 2381
+    },
+    {
+      "epoch": 1.0921912624699002,
+      "grad_norm": 0.25040295720100403,
+      "learning_rate": 0.0007349131162239828,
+      "loss": 1.211,
+      "step": 2382
+    },
+    {
+      "epoch": 1.092649925467263,
+      "grad_norm": 0.3955865204334259,
+      "learning_rate": 0.0007346944788011254,
+      "loss": 1.103,
+      "step": 2383
+    },
+    {
+      "epoch": 1.0931085884646257,
+      "grad_norm": 0.406610906124115,
+      "learning_rate": 0.0007344757838060203,
+      "loss": 2.2019,
+      "step": 2384
+    },
+    {
+      "epoch": 1.0935672514619883,
+      "grad_norm": 0.45408493280410767,
+      "learning_rate": 0.0007342570312923143,
+      "loss": 1.9214,
+      "step": 2385
+    },
+    {
+      "epoch": 1.094025914459351,
+      "grad_norm": 0.3225853741168976,
+      "learning_rate": 0.0007340382213136695,
+      "loss": 1.437,
+      "step": 2386
+    },
+    {
+      "epoch": 1.0944845774567136,
+      "grad_norm": 0.2830888330936432,
+      "learning_rate": 0.0007338193539237619,
+      "loss": 1.6927,
+      "step": 2387
+    },
+    {
+      "epoch": 1.0949432404540764,
+      "grad_norm": 0.3365802466869354,
+      "learning_rate": 0.0007336004291762807,
+      "loss": 1.3085,
+      "step": 2388
+    },
+    {
+      "epoch": 1.095401903451439,
+      "grad_norm": 0.2103971689939499,
+      "learning_rate": 0.0007333814471249298,
+      "loss": 1.3113,
+      "step": 2389
+    },
+    {
+      "epoch": 1.0958605664488017,
+      "grad_norm": 0.3487602472305298,
+      "learning_rate": 0.0007331624078234272,
+      "loss": 1.2643,
+      "step": 2390
+    },
+    {
+      "epoch": 1.0963192294461643,
+      "grad_norm": 0.29860758781433105,
+      "learning_rate": 0.0007329433113255047,
+      "loss": 1.6295,
+      "step": 2391
+    },
+    {
+      "epoch": 1.0967778924435272,
+      "grad_norm": 0.21186383068561554,
+      "learning_rate": 0.0007327241576849083,
+      "loss": 1.4821,
+      "step": 2392
+    },
+    {
+      "epoch": 1.0972365554408898,
+      "grad_norm": 0.2788501977920532,
+      "learning_rate": 0.0007325049469553981,
+      "loss": 1.2378,
+      "step": 2393
+    },
+    {
+      "epoch": 1.0976952184382525,
+      "grad_norm": 0.19782112538814545,
+      "learning_rate": 0.000732285679190748,
+      "loss": 0.8517,
+      "step": 2394
+    },
+    {
+      "epoch": 1.098153881435615,
+      "grad_norm": 0.3813456594944,
+      "learning_rate": 0.0007320663544447459,
+      "loss": 1.818,
+      "step": 2395
+    },
+    {
+      "epoch": 1.098612544432978,
+      "grad_norm": 0.29733943939208984,
+      "learning_rate": 0.0007318469727711936,
+      "loss": 0.8673,
+      "step": 2396
+    },
+    {
+      "epoch": 1.0990712074303406,
+      "grad_norm": 0.18940015137195587,
+      "learning_rate": 0.0007316275342239074,
+      "loss": 0.8093,
+      "step": 2397
+    },
+    {
+      "epoch": 1.0995298704277032,
+      "grad_norm": 0.3412417769432068,
+      "learning_rate": 0.0007314080388567168,
+      "loss": 1.7512,
+      "step": 2398
+    },
+    {
+      "epoch": 1.0999885334250659,
+      "grad_norm": 0.09079885482788086,
+      "learning_rate": 0.0007311884867234658,
+      "loss": 0.9669,
+      "step": 2399
+    },
+    {
+      "epoch": 1.1004471964224287,
+      "grad_norm": 0.34120801091194153,
+      "learning_rate": 0.0007309688778780121,
+      "loss": 1.6829,
+      "step": 2400
+    },
+    {
+      "epoch": 1.1009058594197914,
+      "grad_norm": 0.31036293506622314,
+      "learning_rate": 0.0007307492123742271,
+      "loss": 1.2819,
+      "step": 2401
+    },
+    {
+      "epoch": 1.101364522417154,
+      "grad_norm": 0.32201525568962097,
+      "learning_rate": 0.0007305294902659967,
+      "loss": 1.3809,
+      "step": 2402
+    },
+    {
+      "epoch": 1.1018231854145166,
+      "grad_norm": 0.09801112860441208,
+      "learning_rate": 0.0007303097116072199,
+      "loss": 1.0084,
+      "step": 2403
+    },
+    {
+      "epoch": 1.1022818484118795,
+      "grad_norm": 0.28749385476112366,
+      "learning_rate": 0.0007300898764518105,
+      "loss": 1.0966,
+      "step": 2404
+    },
+    {
+      "epoch": 1.1027405114092421,
+      "grad_norm": 0.22139480710029602,
+      "learning_rate": 0.0007298699848536953,
+      "loss": 1.0551,
+      "step": 2405
+    },
+    {
+      "epoch": 1.1031991744066048,
+      "grad_norm": 0.24073892831802368,
+      "learning_rate": 0.0007296500368668154,
+      "loss": 1.7311,
+      "step": 2406
+    },
+    {
+      "epoch": 1.1036578374039674,
+      "grad_norm": 0.3581311106681824,
+      "learning_rate": 0.0007294300325451253,
+      "loss": 1.576,
+      "step": 2407
+    },
+    {
+      "epoch": 1.10411650040133,
+      "grad_norm": 0.2959865927696228,
+      "learning_rate": 0.0007292099719425942,
+      "loss": 1.2758,
+      "step": 2408
+    },
+    {
+      "epoch": 1.1045751633986929,
+      "grad_norm": 0.24298930168151855,
+      "learning_rate": 0.0007289898551132044,
+      "loss": 1.2407,
+      "step": 2409
+    },
+    {
+      "epoch": 1.1050338263960555,
+      "grad_norm": 0.24619142711162567,
+      "learning_rate": 0.0007287696821109517,
+      "loss": 1.2528,
+      "step": 2410
+    },
+    {
+      "epoch": 1.1054924893934182,
+      "grad_norm": 0.28459224104881287,
+      "learning_rate": 0.0007285494529898468,
+      "loss": 2.0142,
+      "step": 2411
+    },
+    {
+      "epoch": 1.1059511523907808,
+      "grad_norm": 0.295964777469635,
+      "learning_rate": 0.0007283291678039129,
+      "loss": 1.2126,
+      "step": 2412
+    },
+    {
+      "epoch": 1.1064098153881436,
+      "grad_norm": 0.23746977746486664,
+      "learning_rate": 0.000728108826607188,
+      "loss": 1.3847,
+      "step": 2413
+    },
+    {
+      "epoch": 1.1068684783855063,
+      "grad_norm": 0.38938137888908386,
+      "learning_rate": 0.0007278884294537229,
+      "loss": 2.1191,
+      "step": 2414
+    },
+    {
+      "epoch": 1.107327141382869,
+      "grad_norm": 0.1922319382429123,
+      "learning_rate": 0.0007276679763975832,
+      "loss": 0.5424,
+      "step": 2415
+    },
+    {
+      "epoch": 1.1077858043802316,
+      "grad_norm": 0.2816941738128662,
+      "learning_rate": 0.0007274474674928472,
+      "loss": 1.7775,
+      "step": 2416
+    },
+    {
+      "epoch": 1.1082444673775944,
+      "grad_norm": 0.25834113359451294,
+      "learning_rate": 0.0007272269027936073,
+      "loss": 0.9534,
+      "step": 2417
+    },
+    {
+      "epoch": 1.108703130374957,
+      "grad_norm": 0.49492350220680237,
+      "learning_rate": 0.00072700628235397,
+      "loss": 1.386,
+      "step": 2418
+    },
+    {
+      "epoch": 1.1091617933723197,
+      "grad_norm": 0.18091417849063873,
+      "learning_rate": 0.0007267856062280547,
+      "loss": 0.86,
+      "step": 2419
+    },
+    {
+      "epoch": 1.1096204563696823,
+      "grad_norm": 0.19237549602985382,
+      "learning_rate": 0.0007265648744699951,
+      "loss": 1.0081,
+      "step": 2420
+    },
+    {
+      "epoch": 1.110079119367045,
+      "grad_norm": 0.17153310775756836,
+      "learning_rate": 0.0007263440871339382,
+      "loss": 0.7964,
+      "step": 2421
+    },
+    {
+      "epoch": 1.1105377823644078,
+      "grad_norm": 0.22889000177383423,
+      "learning_rate": 0.0007261232442740444,
+      "loss": 1.6779,
+      "step": 2422
+    },
+    {
+      "epoch": 1.1109964453617704,
+      "grad_norm": 0.3227975070476532,
+      "learning_rate": 0.0007259023459444887,
+      "loss": 1.6245,
+      "step": 2423
+    },
+    {
+      "epoch": 1.111455108359133,
+      "grad_norm": 0.38517293334007263,
+      "learning_rate": 0.0007256813921994585,
+      "loss": 1.1709,
+      "step": 2424
+    },
+    {
+      "epoch": 1.1119137713564957,
+      "grad_norm": 0.19292643666267395,
+      "learning_rate": 0.0007254603830931555,
+      "loss": 0.8391,
+      "step": 2425
+    },
+    {
+      "epoch": 1.1123724343538586,
+      "grad_norm": 0.18455813825130463,
+      "learning_rate": 0.000725239318679795,
+      "loss": 0.7338,
+      "step": 2426
+    },
+    {
+      "epoch": 1.1128310973512212,
+      "grad_norm": 0.27097609639167786,
+      "learning_rate": 0.0007250181990136054,
+      "loss": 1.8622,
+      "step": 2427
+    },
+    {
+      "epoch": 1.1132897603485838,
+      "grad_norm": 0.30490532517433167,
+      "learning_rate": 0.0007247970241488293,
+      "loss": 1.6636,
+      "step": 2428
+    },
+    {
+      "epoch": 1.1137484233459465,
+      "grad_norm": 0.27548128366470337,
+      "learning_rate": 0.0007245757941397223,
+      "loss": 1.4517,
+      "step": 2429
+    },
+    {
+      "epoch": 1.1142070863433093,
+      "grad_norm": 0.28075456619262695,
+      "learning_rate": 0.0007243545090405537,
+      "loss": 1.1777,
+      "step": 2430
+    },
+    {
+      "epoch": 1.114665749340672,
+      "grad_norm": 0.2504705488681793,
+      "learning_rate": 0.0007241331689056064,
+      "loss": 1.1601,
+      "step": 2431
+    },
+    {
+      "epoch": 1.1151244123380346,
+      "grad_norm": 0.25834277272224426,
+      "learning_rate": 0.0007239117737891765,
+      "loss": 1.1205,
+      "step": 2432
+    },
+    {
+      "epoch": 1.1155830753353972,
+      "grad_norm": 0.35737091302871704,
+      "learning_rate": 0.0007236903237455741,
+      "loss": 1.629,
+      "step": 2433
+    },
+    {
+      "epoch": 1.11604173833276,
+      "grad_norm": 0.3298056423664093,
+      "learning_rate": 0.0007234688188291226,
+      "loss": 1.9195,
+      "step": 2434
+    },
+    {
+      "epoch": 1.1165004013301227,
+      "grad_norm": 0.4346219003200531,
+      "learning_rate": 0.0007232472590941582,
+      "loss": 2.2007,
+      "step": 2435
+    },
+    {
+      "epoch": 1.1169590643274854,
+      "grad_norm": 0.3693140745162964,
+      "learning_rate": 0.0007230256445950316,
+      "loss": 1.0917,
+      "step": 2436
+    },
+    {
+      "epoch": 1.117417727324848,
+      "grad_norm": 0.3174065947532654,
+      "learning_rate": 0.0007228039753861062,
+      "loss": 2.0214,
+      "step": 2437
+    },
+    {
+      "epoch": 1.1178763903222109,
+      "grad_norm": 0.37119260430336,
+      "learning_rate": 0.0007225822515217593,
+      "loss": 1.564,
+      "step": 2438
+    },
+    {
+      "epoch": 1.1183350533195735,
+      "grad_norm": 0.22383145987987518,
+      "learning_rate": 0.0007223604730563811,
+      "loss": 1.2421,
+      "step": 2439
+    },
+    {
+      "epoch": 1.1187937163169361,
+      "grad_norm": 0.34915295243263245,
+      "learning_rate": 0.0007221386400443757,
+      "loss": 1.7604,
+      "step": 2440
+    },
+    {
+      "epoch": 1.1192523793142988,
+      "grad_norm": 0.31573063135147095,
+      "learning_rate": 0.00072191675254016,
+      "loss": 1.6093,
+      "step": 2441
+    },
+    {
+      "epoch": 1.1197110423116614,
+      "grad_norm": 0.3130987882614136,
+      "learning_rate": 0.0007216948105981649,
+      "loss": 1.3731,
+      "step": 2442
+    },
+    {
+      "epoch": 1.1201697053090243,
+      "grad_norm": 0.25755324959754944,
+      "learning_rate": 0.0007214728142728342,
+      "loss": 1.4297,
+      "step": 2443
+    },
+    {
+      "epoch": 1.120628368306387,
+      "grad_norm": 0.12649127840995789,
+      "learning_rate": 0.0007212507636186251,
+      "loss": 0.6196,
+      "step": 2444
+    },
+    {
+      "epoch": 1.1210870313037495,
+      "grad_norm": 0.20406757295131683,
+      "learning_rate": 0.0007210286586900086,
+      "loss": 1.4123,
+      "step": 2445
+    },
+    {
+      "epoch": 1.1215456943011122,
+      "grad_norm": 0.3635023236274719,
+      "learning_rate": 0.0007208064995414686,
+      "loss": 1.8994,
+      "step": 2446
+    },
+    {
+      "epoch": 1.122004357298475,
+      "grad_norm": 0.27390024065971375,
+      "learning_rate": 0.0007205842862275019,
+      "loss": 0.8153,
+      "step": 2447
+    },
+    {
+      "epoch": 1.1224630202958377,
+      "grad_norm": 0.18709996342658997,
+      "learning_rate": 0.0007203620188026193,
+      "loss": 1.1956,
+      "step": 2448
+    },
+    {
+      "epoch": 1.1229216832932003,
+      "grad_norm": 0.2258118987083435,
+      "learning_rate": 0.0007201396973213446,
+      "loss": 1.261,
+      "step": 2449
+    },
+    {
+      "epoch": 1.123380346290563,
+      "grad_norm": 0.3747474253177643,
+      "learning_rate": 0.000719917321838215,
+      "loss": 1.8525,
+      "step": 2450
+    },
+    {
+      "epoch": 1.1238390092879258,
+      "grad_norm": 0.4060818552970886,
+      "learning_rate": 0.0007196948924077806,
+      "loss": 1.8212,
+      "step": 2451
+    },
+    {
+      "epoch": 1.1242976722852884,
+      "grad_norm": 0.35225608944892883,
+      "learning_rate": 0.000719472409084605,
+      "loss": 1.3362,
+      "step": 2452
+    },
+    {
+      "epoch": 1.124756335282651,
+      "grad_norm": 0.29107534885406494,
+      "learning_rate": 0.0007192498719232649,
+      "loss": 1.9098,
+      "step": 2453
+    },
+    {
+      "epoch": 1.1252149982800137,
+      "grad_norm": 0.35325366258621216,
+      "learning_rate": 0.0007190272809783504,
+      "loss": 1.4185,
+      "step": 2454
+    },
+    {
+      "epoch": 1.1256736612773763,
+      "grad_norm": 0.3643795847892761,
+      "learning_rate": 0.0007188046363044646,
+      "loss": 1.0865,
+      "step": 2455
+    },
+    {
+      "epoch": 1.1261323242747392,
+      "grad_norm": 0.10707731544971466,
+      "learning_rate": 0.0007185819379562238,
+      "loss": 0.8521,
+      "step": 2456
+    },
+    {
+      "epoch": 1.1265909872721018,
+      "grad_norm": 0.9587938189506531,
+      "learning_rate": 0.0007183591859882578,
+      "loss": 2.2844,
+      "step": 2457
+    },
+    {
+      "epoch": 1.1270496502694645,
+      "grad_norm": 0.2738874554634094,
+      "learning_rate": 0.0007181363804552086,
+      "loss": 0.9316,
+      "step": 2458
+    },
+    {
+      "epoch": 1.1275083132668273,
+      "grad_norm": 0.1969241052865982,
+      "learning_rate": 0.0007179135214117327,
+      "loss": 1.2078,
+      "step": 2459
+    },
+    {
+      "epoch": 1.12796697626419,
+      "grad_norm": 0.2931908965110779,
+      "learning_rate": 0.0007176906089124989,
+      "loss": 0.901,
+      "step": 2460
+    },
+    {
+      "epoch": 1.1284256392615526,
+      "grad_norm": 0.26238539814949036,
+      "learning_rate": 0.0007174676430121889,
+      "loss": 1.839,
+      "step": 2461
+    },
+    {
+      "epoch": 1.1288843022589152,
+      "grad_norm": 0.2989310026168823,
+      "learning_rate": 0.0007172446237654981,
+      "loss": 1.1135,
+      "step": 2462
+    },
+    {
+      "epoch": 1.1293429652562779,
+      "grad_norm": 0.42721855640411377,
+      "learning_rate": 0.0007170215512271347,
+      "loss": 1.2623,
+      "step": 2463
+    },
+    {
+      "epoch": 1.1298016282536407,
+      "grad_norm": 0.29083552956581116,
+      "learning_rate": 0.0007167984254518199,
+      "loss": 1.5911,
+      "step": 2464
+    },
+    {
+      "epoch": 1.1302602912510034,
+      "grad_norm": 0.32871025800704956,
+      "learning_rate": 0.0007165752464942882,
+      "loss": 1.3376,
+      "step": 2465
+    },
+    {
+      "epoch": 1.130718954248366,
+      "grad_norm": 0.28994280099868774,
+      "learning_rate": 0.000716352014409287,
+      "loss": 1.7526,
+      "step": 2466
+    },
+    {
+      "epoch": 1.1311776172457286,
+      "grad_norm": 0.30776262283325195,
+      "learning_rate": 0.0007161287292515766,
+      "loss": 1.5325,
+      "step": 2467
+    },
+    {
+      "epoch": 1.1316362802430913,
+      "grad_norm": 0.22005335986614227,
+      "learning_rate": 0.0007159053910759304,
+      "loss": 1.5527,
+      "step": 2468
+    },
+    {
+      "epoch": 1.1320949432404541,
+      "grad_norm": 0.38737305998802185,
+      "learning_rate": 0.0007156819999371353,
+      "loss": 2.2419,
+      "step": 2469
+    },
+    {
+      "epoch": 1.1325536062378168,
+      "grad_norm": 0.243751659989357,
+      "learning_rate": 0.0007154585558899902,
+      "loss": 0.8154,
+      "step": 2470
+    },
+    {
+      "epoch": 1.1330122692351794,
+      "grad_norm": 0.18171143531799316,
+      "learning_rate": 0.0007152350589893081,
+      "loss": 1.5097,
+      "step": 2471
+    },
+    {
+      "epoch": 1.1334709322325422,
+      "grad_norm": 0.2935083508491516,
+      "learning_rate": 0.0007150115092899138,
+      "loss": 1.6631,
+      "step": 2472
+    },
+    {
+      "epoch": 1.1339295952299049,
+      "grad_norm": 0.2859863340854645,
+      "learning_rate": 0.0007147879068466462,
+      "loss": 1.7206,
+      "step": 2473
+    },
+    {
+      "epoch": 1.1343882582272675,
+      "grad_norm": 0.3790501058101654,
+      "learning_rate": 0.0007145642517143563,
+      "loss": 1.9156,
+      "step": 2474
+    },
+    {
+      "epoch": 1.1348469212246302,
+      "grad_norm": 0.23689134418964386,
+      "learning_rate": 0.0007143405439479082,
+      "loss": 0.707,
+      "step": 2475
+    },
+    {
+      "epoch": 1.1353055842219928,
+      "grad_norm": 0.3226507008075714,
+      "learning_rate": 0.0007141167836021793,
+      "loss": 1.9152,
+      "step": 2476
+    },
+    {
+      "epoch": 1.1357642472193556,
+      "grad_norm": 0.3560091257095337,
+      "learning_rate": 0.0007138929707320596,
+      "loss": 1.97,
+      "step": 2477
+    },
+    {
+      "epoch": 1.1362229102167183,
+      "grad_norm": 0.3632325828075409,
+      "learning_rate": 0.0007136691053924519,
+      "loss": 0.7859,
+      "step": 2478
+    },
+    {
+      "epoch": 1.136681573214081,
+      "grad_norm": 0.11780796200037003,
+      "learning_rate": 0.0007134451876382719,
+      "loss": 1.0535,
+      "step": 2479
+    },
+    {
+      "epoch": 1.1371402362114436,
+      "grad_norm": 0.3758339583873749,
+      "learning_rate": 0.0007132212175244484,
+      "loss": 1.8105,
+      "step": 2480
+    },
+    {
+      "epoch": 1.1375988992088064,
+      "grad_norm": 0.3397655785083771,
+      "learning_rate": 0.0007129971951059229,
+      "loss": 1.3989,
+      "step": 2481
+    },
+    {
+      "epoch": 1.138057562206169,
+      "grad_norm": 0.3613055944442749,
+      "learning_rate": 0.0007127731204376497,
+      "loss": 1.5942,
+      "step": 2482
+    },
+    {
+      "epoch": 1.1385162252035317,
+      "grad_norm": 0.3114750385284424,
+      "learning_rate": 0.0007125489935745958,
+      "loss": 1.447,
+      "step": 2483
+    },
+    {
+      "epoch": 1.1389748882008943,
+      "grad_norm": 0.4733560085296631,
+      "learning_rate": 0.0007123248145717412,
+      "loss": 2.2578,
+      "step": 2484
+    },
+    {
+      "epoch": 1.1394335511982572,
+      "grad_norm": 0.28705328702926636,
+      "learning_rate": 0.0007121005834840786,
+      "loss": 1.2581,
+      "step": 2485
+    },
+    {
+      "epoch": 1.1398922141956198,
+      "grad_norm": 0.22501793503761292,
+      "learning_rate": 0.0007118763003666137,
+      "loss": 1.1338,
+      "step": 2486
+    },
+    {
+      "epoch": 1.1403508771929824,
+      "grad_norm": 0.24394629895687103,
+      "learning_rate": 0.0007116519652743644,
+      "loss": 1.2005,
+      "step": 2487
+    },
+    {
+      "epoch": 1.140809540190345,
+      "grad_norm": 0.32340019941329956,
+      "learning_rate": 0.0007114275782623622,
+      "loss": 1.2241,
+      "step": 2488
+    },
+    {
+      "epoch": 1.1412682031877077,
+      "grad_norm": 0.29200488328933716,
+      "learning_rate": 0.0007112031393856504,
+      "loss": 0.9522,
+      "step": 2489
+    },
+    {
+      "epoch": 1.1417268661850706,
+      "grad_norm": 0.11450646817684174,
+      "learning_rate": 0.0007109786486992856,
+      "loss": 0.9667,
+      "step": 2490
+    },
+    {
+      "epoch": 1.1421855291824332,
+      "grad_norm": 0.32306355237960815,
+      "learning_rate": 0.0007107541062583372,
+      "loss": 0.917,
+      "step": 2491
+    },
+    {
+      "epoch": 1.1426441921797958,
+      "grad_norm": 0.2934865355491638,
+      "learning_rate": 0.0007105295121178867,
+      "loss": 1.5807,
+      "step": 2492
+    },
+    {
+      "epoch": 1.1431028551771587,
+      "grad_norm": 0.3221622705459595,
+      "learning_rate": 0.0007103048663330291,
+      "loss": 1.895,
+      "step": 2493
+    },
+    {
+      "epoch": 1.1435615181745213,
+      "grad_norm": 0.3509625792503357,
+      "learning_rate": 0.0007100801689588714,
+      "loss": 1.4817,
+      "step": 2494
+    },
+    {
+      "epoch": 1.144020181171884,
+      "grad_norm": 0.255881130695343,
+      "learning_rate": 0.0007098554200505334,
+      "loss": 1.564,
+      "step": 2495
+    },
+    {
+      "epoch": 1.1444788441692466,
+      "grad_norm": 0.33859267830848694,
+      "learning_rate": 0.0007096306196631478,
+      "loss": 1.1945,
+      "step": 2496
+    },
+    {
+      "epoch": 1.1449375071666092,
+      "grad_norm": 0.2730942964553833,
+      "learning_rate": 0.0007094057678518597,
+      "loss": 1.1724,
+      "step": 2497
+    },
+    {
+      "epoch": 1.145396170163972,
+      "grad_norm": 0.12389812618494034,
+      "learning_rate": 0.0007091808646718268,
+      "loss": 1.1204,
+      "step": 2498
+    },
+    {
+      "epoch": 1.1458548331613347,
+      "grad_norm": 0.3796531558036804,
+      "learning_rate": 0.0007089559101782195,
+      "loss": 1.622,
+      "step": 2499
+    },
+    {
+      "epoch": 1.1463134961586974,
+      "grad_norm": 0.2889677584171295,
+      "learning_rate": 0.0007087309044262206,
+      "loss": 0.8866,
+      "step": 2500
+    },
+    {
+      "epoch": 1.14677215915606,
+      "grad_norm": 0.3515912890434265,
+      "learning_rate": 0.0007085058474710261,
+      "loss": 1.2305,
+      "step": 2501
+    },
+    {
+      "epoch": 1.1472308221534229,
+      "grad_norm": 0.2346973717212677,
+      "learning_rate": 0.0007082807393678439,
+      "loss": 1.3991,
+      "step": 2502
+    },
+    {
+      "epoch": 1.1476894851507855,
+      "grad_norm": 0.26319047808647156,
+      "learning_rate": 0.0007080555801718943,
+      "loss": 1.3355,
+      "step": 2503
+    },
+    {
+      "epoch": 1.1481481481481481,
+      "grad_norm": 0.38340768218040466,
+      "learning_rate": 0.0007078303699384107,
+      "loss": 1.7297,
+      "step": 2504
+    },
+    {
+      "epoch": 1.1486068111455108,
+      "grad_norm": 0.27420321106910706,
+      "learning_rate": 0.0007076051087226389,
+      "loss": 1.7687,
+      "step": 2505
+    },
+    {
+      "epoch": 1.1490654741428736,
+      "grad_norm": 0.36036863923072815,
+      "learning_rate": 0.0007073797965798371,
+      "loss": 1.2501,
+      "step": 2506
+    },
+    {
+      "epoch": 1.1495241371402363,
+      "grad_norm": 0.2933008074760437,
+      "learning_rate": 0.000707154433565276,
+      "loss": 1.762,
+      "step": 2507
+    },
+    {
+      "epoch": 1.149982800137599,
+      "grad_norm": 0.6776466369628906,
+      "learning_rate": 0.0007069290197342385,
+      "loss": 1.1041,
+      "step": 2508
+    },
+    {
+      "epoch": 1.1504414631349615,
+      "grad_norm": 0.16171732544898987,
+      "learning_rate": 0.0007067035551420205,
+      "loss": 0.8364,
+      "step": 2509
+    },
+    {
+      "epoch": 1.1509001261323242,
+      "grad_norm": 0.28582391142845154,
+      "learning_rate": 0.0007064780398439299,
+      "loss": 1.4814,
+      "step": 2510
+    },
+    {
+      "epoch": 1.151358789129687,
+      "grad_norm": 0.31193310022354126,
+      "learning_rate": 0.0007062524738952875,
+      "loss": 1.2991,
+      "step": 2511
+    },
+    {
+      "epoch": 1.1518174521270497,
+      "grad_norm": 0.28195393085479736,
+      "learning_rate": 0.0007060268573514259,
+      "loss": 1.8887,
+      "step": 2512
+    },
+    {
+      "epoch": 1.1522761151244123,
+      "grad_norm": 0.30946487188339233,
+      "learning_rate": 0.0007058011902676909,
+      "loss": 1.2166,
+      "step": 2513
+    },
+    {
+      "epoch": 1.152734778121775,
+      "grad_norm": 0.3007708787918091,
+      "learning_rate": 0.0007055754726994399,
+      "loss": 1.4958,
+      "step": 2514
+    },
+    {
+      "epoch": 1.1531934411191378,
+      "grad_norm": 0.21293707191944122,
+      "learning_rate": 0.0007053497047020432,
+      "loss": 1.0453,
+      "step": 2515
+    },
+    {
+      "epoch": 1.1536521041165004,
+      "grad_norm": 0.32221704721450806,
+      "learning_rate": 0.0007051238863308832,
+      "loss": 1.4675,
+      "step": 2516
+    },
+    {
+      "epoch": 1.154110767113863,
+      "grad_norm": 0.2842558026313782,
+      "learning_rate": 0.0007048980176413549,
+      "loss": 1.9645,
+      "step": 2517
+    },
+    {
+      "epoch": 1.1545694301112257,
+      "grad_norm": 0.3128332793712616,
+      "learning_rate": 0.0007046720986888656,
+      "loss": 1.3461,
+      "step": 2518
+    },
+    {
+      "epoch": 1.1550280931085886,
+      "grad_norm": 0.3141840994358063,
+      "learning_rate": 0.0007044461295288347,
+      "loss": 2.0339,
+      "step": 2519
+    },
+    {
+      "epoch": 1.1554867561059512,
+      "grad_norm": 0.39887315034866333,
+      "learning_rate": 0.0007042201102166939,
+      "loss": 2.0115,
+      "step": 2520
+    },
+    {
+      "epoch": 1.1559454191033138,
+      "grad_norm": 0.3459334075450897,
+      "learning_rate": 0.0007039940408078878,
+      "loss": 1.922,
+      "step": 2521
+    },
+    {
+      "epoch": 1.1564040821006765,
+      "grad_norm": 0.31680259108543396,
+      "learning_rate": 0.0007037679213578725,
+      "loss": 1.6561,
+      "step": 2522
+    },
+    {
+      "epoch": 1.156862745098039,
+      "grad_norm": 0.27092963457107544,
+      "learning_rate": 0.0007035417519221168,
+      "loss": 1.2754,
+      "step": 2523
+    },
+    {
+      "epoch": 1.157321408095402,
+      "grad_norm": 0.23691010475158691,
+      "learning_rate": 0.0007033155325561018,
+      "loss": 1.4512,
+      "step": 2524
+    },
+    {
+      "epoch": 1.1577800710927646,
+      "grad_norm": 0.3520990014076233,
+      "learning_rate": 0.0007030892633153208,
+      "loss": 1.3797,
+      "step": 2525
+    },
+    {
+      "epoch": 1.1582387340901272,
+      "grad_norm": 0.19260728359222412,
+      "learning_rate": 0.0007028629442552788,
+      "loss": 1.3934,
+      "step": 2526
+    },
+    {
+      "epoch": 1.15869739708749,
+      "grad_norm": 0.31958216428756714,
+      "learning_rate": 0.0007026365754314943,
+      "loss": 1.1669,
+      "step": 2527
+    },
+    {
+      "epoch": 1.1591560600848527,
+      "grad_norm": 0.1669234186410904,
+      "learning_rate": 0.0007024101568994965,
+      "loss": 1.2692,
+      "step": 2528
+    },
+    {
+      "epoch": 1.1596147230822154,
+      "grad_norm": 0.3692995011806488,
+      "learning_rate": 0.0007021836887148278,
+      "loss": 1.7561,
+      "step": 2529
+    },
+    {
+      "epoch": 1.160073386079578,
+      "grad_norm": 0.21594902873039246,
+      "learning_rate": 0.0007019571709330425,
+      "loss": 1.4732,
+      "step": 2530
+    },
+    {
+      "epoch": 1.1605320490769406,
+      "grad_norm": 0.3213531970977783,
+      "learning_rate": 0.0007017306036097068,
+      "loss": 1.6332,
+      "step": 2531
+    },
+    {
+      "epoch": 1.1609907120743035,
+      "grad_norm": 0.34258440136909485,
+      "learning_rate": 0.0007015039868003998,
+      "loss": 1.9807,
+      "step": 2532
+    },
+    {
+      "epoch": 1.1614493750716661,
+      "grad_norm": 0.4435928165912628,
+      "learning_rate": 0.0007012773205607117,
+      "loss": 1.7593,
+      "step": 2533
+    },
+    {
+      "epoch": 1.1619080380690288,
+      "grad_norm": 0.2438933551311493,
+      "learning_rate": 0.0007010506049462456,
+      "loss": 1.1445,
+      "step": 2534
+    },
+    {
+      "epoch": 1.1623667010663914,
+      "grad_norm": 0.32633253931999207,
+      "learning_rate": 0.0007008238400126165,
+      "loss": 2.1003,
+      "step": 2535
+    },
+    {
+      "epoch": 1.1628253640637543,
+      "grad_norm": 0.3241181969642639,
+      "learning_rate": 0.0007005970258154514,
+      "loss": 1.5915,
+      "step": 2536
+    },
+    {
+      "epoch": 1.1632840270611169,
+      "grad_norm": 0.2975384593009949,
+      "learning_rate": 0.0007003701624103895,
+      "loss": 1.5494,
+      "step": 2537
+    },
+    {
+      "epoch": 1.1637426900584795,
+      "grad_norm": 0.26780200004577637,
+      "learning_rate": 0.000700143249853082,
+      "loss": 1.3053,
+      "step": 2538
+    },
+    {
+      "epoch": 1.1642013530558422,
+      "grad_norm": 0.26362505555152893,
+      "learning_rate": 0.0006999162881991922,
+      "loss": 1.3801,
+      "step": 2539
+    },
+    {
+      "epoch": 1.164660016053205,
+      "grad_norm": 0.21280516684055328,
+      "learning_rate": 0.0006996892775043955,
+      "loss": 1.0339,
+      "step": 2540
+    },
+    {
+      "epoch": 1.1651186790505677,
+      "grad_norm": 0.2693462073802948,
+      "learning_rate": 0.0006994622178243792,
+      "loss": 1.1786,
+      "step": 2541
+    },
+    {
+      "epoch": 1.1655773420479303,
+      "grad_norm": 0.18628975749015808,
+      "learning_rate": 0.0006992351092148426,
+      "loss": 1.4404,
+      "step": 2542
+    },
+    {
+      "epoch": 1.166036005045293,
+      "grad_norm": 0.41851601004600525,
+      "learning_rate": 0.0006990079517314971,
+      "loss": 1.853,
+      "step": 2543
+    },
+    {
+      "epoch": 1.1664946680426556,
+      "grad_norm": 0.3848247826099396,
+      "learning_rate": 0.0006987807454300662,
+      "loss": 1.8879,
+      "step": 2544
+    },
+    {
+      "epoch": 1.1669533310400184,
+      "grad_norm": 0.258331835269928,
+      "learning_rate": 0.0006985534903662851,
+      "loss": 1.3988,
+      "step": 2545
+    },
+    {
+      "epoch": 1.167411994037381,
+      "grad_norm": 0.30951839685440063,
+      "learning_rate": 0.0006983261865959011,
+      "loss": 1.3336,
+      "step": 2546
+    },
+    {
+      "epoch": 1.1678706570347437,
+      "grad_norm": 0.2942619025707245,
+      "learning_rate": 0.0006980988341746737,
+      "loss": 1.7388,
+      "step": 2547
+    },
+    {
+      "epoch": 1.1683293200321063,
+      "grad_norm": 0.2846639156341553,
+      "learning_rate": 0.0006978714331583739,
+      "loss": 1.2854,
+      "step": 2548
+    },
+    {
+      "epoch": 1.1687879830294692,
+      "grad_norm": 0.3299280107021332,
+      "learning_rate": 0.0006976439836027848,
+      "loss": 2.272,
+      "step": 2549
+    },
+    {
+      "epoch": 1.1692466460268318,
+      "grad_norm": 0.2958674132823944,
+      "learning_rate": 0.0006974164855637015,
+      "loss": 1.6334,
+      "step": 2550
+    },
+    {
+      "epoch": 1.1697053090241945,
+      "grad_norm": 0.43015730381011963,
+      "learning_rate": 0.0006971889390969307,
+      "loss": 1.6784,
+      "step": 2551
+    },
+    {
+      "epoch": 1.170163972021557,
+      "grad_norm": 0.1926140934228897,
+      "learning_rate": 0.0006969613442582914,
+      "loss": 1.0963,
+      "step": 2552
+    },
+    {
+      "epoch": 1.17062263501892,
+      "grad_norm": 0.26699936389923096,
+      "learning_rate": 0.0006967337011036141,
+      "loss": 1.2809,
+      "step": 2553
+    },
+    {
+      "epoch": 1.1710812980162826,
+      "grad_norm": 0.34958434104919434,
+      "learning_rate": 0.0006965060096887414,
+      "loss": 2.0635,
+      "step": 2554
+    },
+    {
+      "epoch": 1.1715399610136452,
+      "grad_norm": 0.3384473919868469,
+      "learning_rate": 0.0006962782700695278,
+      "loss": 1.3294,
+      "step": 2555
+    },
+    {
+      "epoch": 1.1719986240110079,
+      "grad_norm": 0.3625926375389099,
+      "learning_rate": 0.0006960504823018392,
+      "loss": 1.9546,
+      "step": 2556
+    },
+    {
+      "epoch": 1.1724572870083705,
+      "grad_norm": 0.23835191130638123,
+      "learning_rate": 0.0006958226464415537,
+      "loss": 1.3571,
+      "step": 2557
+    },
+    {
+      "epoch": 1.1729159500057333,
+      "grad_norm": 0.38838183879852295,
+      "learning_rate": 0.0006955947625445611,
+      "loss": 1.9731,
+      "step": 2558
+    },
+    {
+      "epoch": 1.173374613003096,
+      "grad_norm": 0.40480363368988037,
+      "learning_rate": 0.000695366830666763,
+      "loss": 2.1235,
+      "step": 2559
+    },
+    {
+      "epoch": 1.1738332760004586,
+      "grad_norm": 0.32704851031303406,
+      "learning_rate": 0.0006951388508640725,
+      "loss": 1.7461,
+      "step": 2560
+    },
+    {
+      "epoch": 1.1742919389978215,
+      "grad_norm": 0.2514033615589142,
+      "learning_rate": 0.000694910823192415,
+      "loss": 0.8705,
+      "step": 2561
+    },
+    {
+      "epoch": 1.174750601995184,
+      "grad_norm": 0.20851144194602966,
+      "learning_rate": 0.0006946827477077271,
+      "loss": 1.0728,
+      "step": 2562
+    },
+    {
+      "epoch": 1.1752092649925467,
+      "grad_norm": 0.1491597592830658,
+      "learning_rate": 0.0006944546244659575,
+      "loss": 0.7351,
+      "step": 2563
+    },
+    {
+      "epoch": 1.1756679279899094,
+      "grad_norm": 0.2879053056240082,
+      "learning_rate": 0.0006942264535230665,
+      "loss": 1.2752,
+      "step": 2564
+    },
+    {
+      "epoch": 1.176126590987272,
+      "grad_norm": 0.4318799078464508,
+      "learning_rate": 0.0006939982349350259,
+      "loss": 1.8137,
+      "step": 2565
+    },
+    {
+      "epoch": 1.1765852539846349,
+      "grad_norm": 0.23562239110469818,
+      "learning_rate": 0.0006937699687578195,
+      "loss": 1.4907,
+      "step": 2566
+    },
+    {
+      "epoch": 1.1770439169819975,
+      "grad_norm": 0.22170042991638184,
+      "learning_rate": 0.0006935416550474426,
+      "loss": 1.2719,
+      "step": 2567
+    },
+    {
+      "epoch": 1.1775025799793601,
+      "grad_norm": 0.3226730227470398,
+      "learning_rate": 0.0006933132938599022,
+      "loss": 1.6284,
+      "step": 2568
+    },
+    {
+      "epoch": 1.1779612429767228,
+      "grad_norm": 0.2588571310043335,
+      "learning_rate": 0.000693084885251217,
+      "loss": 0.9474,
+      "step": 2569
+    },
+    {
+      "epoch": 1.1784199059740856,
+      "grad_norm": 0.331263929605484,
+      "learning_rate": 0.0006928564292774173,
+      "loss": 2.2437,
+      "step": 2570
+    },
+    {
+      "epoch": 1.1788785689714483,
+      "grad_norm": 0.37038788199424744,
+      "learning_rate": 0.0006926279259945447,
+      "loss": 1.5394,
+      "step": 2571
+    },
+    {
+      "epoch": 1.179337231968811,
+      "grad_norm": 0.21661770343780518,
+      "learning_rate": 0.0006923993754586532,
+      "loss": 1.4075,
+      "step": 2572
+    },
+    {
+      "epoch": 1.1797958949661735,
+      "grad_norm": 0.24760504066944122,
+      "learning_rate": 0.0006921707777258073,
+      "loss": 0.8289,
+      "step": 2573
+    },
+    {
+      "epoch": 1.1802545579635364,
+      "grad_norm": 0.29299482703208923,
+      "learning_rate": 0.0006919421328520844,
+      "loss": 1.6813,
+      "step": 2574
+    },
+    {
+      "epoch": 1.180713220960899,
+      "grad_norm": 0.21871504187583923,
+      "learning_rate": 0.0006917134408935721,
+      "loss": 1.2732,
+      "step": 2575
+    },
+    {
+      "epoch": 1.1811718839582617,
+      "grad_norm": 0.31473806500434875,
+      "learning_rate": 0.0006914847019063705,
+      "loss": 0.5663,
+      "step": 2576
+    },
+    {
+      "epoch": 1.1816305469556243,
+      "grad_norm": 0.2423599660396576,
+      "learning_rate": 0.0006912559159465908,
+      "loss": 0.8704,
+      "step": 2577
+    },
+    {
+      "epoch": 1.182089209952987,
+      "grad_norm": 0.3309333026409149,
+      "learning_rate": 0.0006910270830703559,
+      "loss": 1.4814,
+      "step": 2578
+    },
+    {
+      "epoch": 1.1825478729503498,
+      "grad_norm": 0.281486451625824,
+      "learning_rate": 0.0006907982033338001,
+      "loss": 1.5132,
+      "step": 2579
+    },
+    {
+      "epoch": 1.1830065359477124,
+      "grad_norm": 0.3493801951408386,
+      "learning_rate": 0.0006905692767930695,
+      "loss": 1.7191,
+      "step": 2580
+    },
+    {
+      "epoch": 1.183465198945075,
+      "grad_norm": 0.3665405213832855,
+      "learning_rate": 0.000690340303504321,
+      "loss": 2.2778,
+      "step": 2581
+    },
+    {
+      "epoch": 1.1839238619424377,
+      "grad_norm": 0.36640089750289917,
+      "learning_rate": 0.0006901112835237237,
+      "loss": 1.8275,
+      "step": 2582
+    },
+    {
+      "epoch": 1.1843825249398006,
+      "grad_norm": 0.33539673686027527,
+      "learning_rate": 0.0006898822169074577,
+      "loss": 1.6515,
+      "step": 2583
+    },
+    {
+      "epoch": 1.1848411879371632,
+      "grad_norm": 0.3045608401298523,
+      "learning_rate": 0.0006896531037117148,
+      "loss": 1.5253,
+      "step": 2584
+    },
+    {
+      "epoch": 1.1852998509345258,
+      "grad_norm": 0.18760670721530914,
+      "learning_rate": 0.0006894239439926981,
+      "loss": 0.6896,
+      "step": 2585
+    },
+    {
+      "epoch": 1.1857585139318885,
+      "grad_norm": 0.11621358245611191,
+      "learning_rate": 0.0006891947378066223,
+      "loss": 0.916,
+      "step": 2586
+    },
+    {
+      "epoch": 1.1862171769292513,
+      "grad_norm": 0.24042093753814697,
+      "learning_rate": 0.000688965485209713,
+      "loss": 0.7641,
+      "step": 2587
+    },
+    {
+      "epoch": 1.186675839926614,
+      "grad_norm": 0.1797315925359726,
+      "learning_rate": 0.0006887361862582077,
+      "loss": 1.1713,
+      "step": 2588
+    },
+    {
+      "epoch": 1.1871345029239766,
+      "grad_norm": 0.305295467376709,
+      "learning_rate": 0.000688506841008355,
+      "loss": 1.2055,
+      "step": 2589
+    },
+    {
+      "epoch": 1.1875931659213392,
+      "grad_norm": 0.18166819214820862,
+      "learning_rate": 0.000688277449516415,
+      "loss": 1.0945,
+      "step": 2590
+    },
+    {
+      "epoch": 1.1880518289187019,
+      "grad_norm": 0.23136307299137115,
+      "learning_rate": 0.0006880480118386592,
+      "loss": 0.8324,
+      "step": 2591
+    },
+    {
+      "epoch": 1.1885104919160647,
+      "grad_norm": 0.22618937492370605,
+      "learning_rate": 0.00068781852803137,
+      "loss": 1.1397,
+      "step": 2592
+    },
+    {
+      "epoch": 1.1889691549134274,
+      "grad_norm": 0.30645951628685,
+      "learning_rate": 0.0006875889981508416,
+      "loss": 1.4277,
+      "step": 2593
+    },
+    {
+      "epoch": 1.18942781791079,
+      "grad_norm": 0.3529110848903656,
+      "learning_rate": 0.0006873594222533796,
+      "loss": 1.6997,
+      "step": 2594
+    },
+    {
+      "epoch": 1.1898864809081529,
+      "grad_norm": 0.32532528042793274,
+      "learning_rate": 0.0006871298003953004,
+      "loss": 1.8083,
+      "step": 2595
+    },
+    {
+      "epoch": 1.1903451439055155,
+      "grad_norm": 0.3486868143081665,
+      "learning_rate": 0.0006869001326329317,
+      "loss": 0.871,
+      "step": 2596
+    },
+    {
+      "epoch": 1.1908038069028781,
+      "grad_norm": 0.2799234092235565,
+      "learning_rate": 0.0006866704190226131,
+      "loss": 1.4758,
+      "step": 2597
+    },
+    {
+      "epoch": 1.1912624699002408,
+      "grad_norm": 0.17606878280639648,
+      "learning_rate": 0.0006864406596206945,
+      "loss": 1.1527,
+      "step": 2598
+    },
+    {
+      "epoch": 1.1917211328976034,
+      "grad_norm": 0.32030197978019714,
+      "learning_rate": 0.0006862108544835379,
+      "loss": 1.5411,
+      "step": 2599
+    },
+    {
+      "epoch": 1.1921797958949663,
+      "grad_norm": 0.2235921025276184,
+      "learning_rate": 0.000685981003667516,
+      "loss": 0.9424,
+      "step": 2600
+    },
+    {
+      "epoch": 1.192638458892329,
+      "grad_norm": 0.39481407403945923,
+      "learning_rate": 0.0006857511072290128,
+      "loss": 2.1982,
+      "step": 2601
+    },
+    {
+      "epoch": 1.1930971218896915,
+      "grad_norm": 0.2962379455566406,
+      "learning_rate": 0.0006855211652244238,
+      "loss": 1.0606,
+      "step": 2602
+    },
+    {
+      "epoch": 1.1935557848870542,
+      "grad_norm": 0.2588828504085541,
+      "learning_rate": 0.0006852911777101553,
+      "loss": 1.2921,
+      "step": 2603
+    },
+    {
+      "epoch": 1.194014447884417,
+      "grad_norm": 0.09400691092014313,
+      "learning_rate": 0.0006850611447426248,
+      "loss": 0.9105,
+      "step": 2604
+    },
+    {
+      "epoch": 1.1944731108817797,
+      "grad_norm": 0.299527645111084,
+      "learning_rate": 0.0006848310663782613,
+      "loss": 1.4135,
+      "step": 2605
+    },
+    {
+      "epoch": 1.1949317738791423,
+      "grad_norm": 0.33210527896881104,
+      "learning_rate": 0.0006846009426735045,
+      "loss": 1.6572,
+      "step": 2606
+    },
+    {
+      "epoch": 1.195390436876505,
+      "grad_norm": 0.3115682303905487,
+      "learning_rate": 0.0006843707736848052,
+      "loss": 1.3729,
+      "step": 2607
+    },
+    {
+      "epoch": 1.1958490998738678,
+      "grad_norm": 0.356736958026886,
+      "learning_rate": 0.0006841405594686259,
+      "loss": 1.7662,
+      "step": 2608
+    },
+    {
+      "epoch": 1.1963077628712304,
+      "grad_norm": 0.2792001962661743,
+      "learning_rate": 0.0006839103000814397,
+      "loss": 1.5021,
+      "step": 2609
+    },
+    {
+      "epoch": 1.196766425868593,
+      "grad_norm": 0.35029587149620056,
+      "learning_rate": 0.0006836799955797306,
+      "loss": 2.1519,
+      "step": 2610
+    },
+    {
+      "epoch": 1.1972250888659557,
+      "grad_norm": 0.24187533557415009,
+      "learning_rate": 0.0006834496460199944,
+      "loss": 0.7456,
+      "step": 2611
+    },
+    {
+      "epoch": 1.1976837518633183,
+      "grad_norm": 0.12775690853595734,
+      "learning_rate": 0.0006832192514587372,
+      "loss": 0.6849,
+      "step": 2612
+    },
+    {
+      "epoch": 1.1981424148606812,
+      "grad_norm": 0.23146137595176697,
+      "learning_rate": 0.0006829888119524765,
+      "loss": 1.2602,
+      "step": 2613
+    },
+    {
+      "epoch": 1.1986010778580438,
+      "grad_norm": 0.3723481595516205,
+      "learning_rate": 0.0006827583275577409,
+      "loss": 1.2093,
+      "step": 2614
+    },
+    {
+      "epoch": 1.1990597408554065,
+      "grad_norm": 0.2874572277069092,
+      "learning_rate": 0.0006825277983310697,
+      "loss": 1.7144,
+      "step": 2615
+    },
+    {
+      "epoch": 1.199518403852769,
+      "grad_norm": 0.32732078433036804,
+      "learning_rate": 0.0006822972243290136,
+      "loss": 1.6793,
+      "step": 2616
+    },
+    {
+      "epoch": 1.199977066850132,
+      "grad_norm": 0.28288617730140686,
+      "learning_rate": 0.0006820666056081339,
+      "loss": 1.5604,
+      "step": 2617
+    },
+    {
+      "epoch": 1.2004357298474946,
+      "grad_norm": 0.18358604609966278,
+      "learning_rate": 0.000681835942225003,
+      "loss": 0.4566,
+      "step": 2618
+    },
+    {
+      "epoch": 1.2008943928448572,
+      "grad_norm": 0.260851114988327,
+      "learning_rate": 0.0006816052342362045,
+      "loss": 0.9449,
+      "step": 2619
+    },
+    {
+      "epoch": 1.2013530558422199,
+      "grad_norm": 0.22726760804653168,
+      "learning_rate": 0.0006813744816983324,
+      "loss": 0.879,
+      "step": 2620
+    },
+    {
+      "epoch": 1.2018117188395827,
+      "grad_norm": 0.229889914393425,
+      "learning_rate": 0.0006811436846679923,
+      "loss": 0.899,
+      "step": 2621
+    },
+    {
+      "epoch": 1.2022703818369453,
+      "grad_norm": 0.30879998207092285,
+      "learning_rate": 0.0006809128432018003,
+      "loss": 1.7426,
+      "step": 2622
+    },
+    {
+      "epoch": 1.202729044834308,
+      "grad_norm": 0.1792958378791809,
+      "learning_rate": 0.0006806819573563832,
+      "loss": 0.9025,
+      "step": 2623
+    },
+    {
+      "epoch": 1.2031877078316706,
+      "grad_norm": 0.4172597825527191,
+      "learning_rate": 0.0006804510271883793,
+      "loss": 2.02,
+      "step": 2624
+    },
+    {
+      "epoch": 1.2036463708290333,
+      "grad_norm": 0.26834097504615784,
+      "learning_rate": 0.0006802200527544374,
+      "loss": 1.1736,
+      "step": 2625
+    },
+    {
+      "epoch": 1.204105033826396,
+      "grad_norm": 0.2656348645687103,
+      "learning_rate": 0.000679989034111217,
+      "loss": 1.2722,
+      "step": 2626
+    },
+    {
+      "epoch": 1.2045636968237587,
+      "grad_norm": 0.22897177934646606,
+      "learning_rate": 0.0006797579713153888,
+      "loss": 1.1406,
+      "step": 2627
+    },
+    {
+      "epoch": 1.2050223598211214,
+      "grad_norm": 0.2911318838596344,
+      "learning_rate": 0.0006795268644236341,
+      "loss": 0.839,
+      "step": 2628
+    },
+    {
+      "epoch": 1.2054810228184842,
+      "grad_norm": 1.6645565032958984,
+      "learning_rate": 0.000679295713492645,
+      "loss": 1.858,
+      "step": 2629
+    },
+    {
+      "epoch": 1.2059396858158469,
+      "grad_norm": 0.32228031754493713,
+      "learning_rate": 0.0006790645185791247,
+      "loss": 2.0117,
+      "step": 2630
+    },
+    {
+      "epoch": 1.2063983488132095,
+      "grad_norm": 0.1701250672340393,
+      "learning_rate": 0.0006788332797397868,
+      "loss": 0.4305,
+      "step": 2631
+    },
+    {
+      "epoch": 1.2068570118105721,
+      "grad_norm": 0.3705667555332184,
+      "learning_rate": 0.0006786019970313559,
+      "loss": 1.7415,
+      "step": 2632
+    },
+    {
+      "epoch": 1.2073156748079348,
+      "grad_norm": 0.12469058483839035,
+      "learning_rate": 0.0006783706705105675,
+      "loss": 1.3766,
+      "step": 2633
+    },
+    {
+      "epoch": 1.2077743378052976,
+      "grad_norm": 0.5770121216773987,
+      "learning_rate": 0.0006781393002341674,
+      "loss": 1.9946,
+      "step": 2634
+    },
+    {
+      "epoch": 1.2082330008026603,
+      "grad_norm": 0.2683452069759369,
+      "learning_rate": 0.0006779078862589126,
+      "loss": 1.5091,
+      "step": 2635
+    },
+    {
+      "epoch": 1.208691663800023,
+      "grad_norm": 0.3430030643939972,
+      "learning_rate": 0.0006776764286415704,
+      "loss": 1.6704,
+      "step": 2636
+    },
+    {
+      "epoch": 1.2091503267973855,
+      "grad_norm": 0.3244686424732208,
+      "learning_rate": 0.0006774449274389192,
+      "loss": 1.5453,
+      "step": 2637
+    },
+    {
+      "epoch": 1.2096089897947484,
+      "grad_norm": 0.31932759284973145,
+      "learning_rate": 0.0006772133827077478,
+      "loss": 1.8912,
+      "step": 2638
+    },
+    {
+      "epoch": 1.210067652792111,
+      "grad_norm": 0.3642272353172302,
+      "learning_rate": 0.0006769817945048558,
+      "loss": 1.7648,
+      "step": 2639
+    },
+    {
+      "epoch": 1.2105263157894737,
+      "grad_norm": 0.40499117970466614,
+      "learning_rate": 0.0006767501628870536,
+      "loss": 1.9568,
+      "step": 2640
+    },
+    {
+      "epoch": 1.2109849787868363,
+      "grad_norm": 0.2826281189918518,
+      "learning_rate": 0.0006765184879111621,
+      "loss": 1.1599,
+      "step": 2641
+    },
+    {
+      "epoch": 1.2114436417841992,
+      "grad_norm": 0.2369105964899063,
+      "learning_rate": 0.0006762867696340126,
+      "loss": 1.1869,
+      "step": 2642
+    },
+    {
+      "epoch": 1.2119023047815618,
+      "grad_norm": 0.34914588928222656,
+      "learning_rate": 0.0006760550081124475,
+      "loss": 1.6072,
+      "step": 2643
+    },
+    {
+      "epoch": 1.2123609677789244,
+      "grad_norm": 0.8818534016609192,
+      "learning_rate": 0.0006758232034033193,
+      "loss": 2.145,
+      "step": 2644
+    },
+    {
+      "epoch": 1.212819630776287,
+      "grad_norm": 0.3968670070171356,
+      "learning_rate": 0.0006755913555634918,
+      "loss": 1.8677,
+      "step": 2645
+    },
+    {
+      "epoch": 1.2132782937736497,
+      "grad_norm": 0.20291972160339355,
+      "learning_rate": 0.0006753594646498384,
+      "loss": 0.8382,
+      "step": 2646
+    },
+    {
+      "epoch": 1.2137369567710126,
+      "grad_norm": 0.3209551274776459,
+      "learning_rate": 0.0006751275307192442,
+      "loss": 1.3622,
+      "step": 2647
+    },
+    {
+      "epoch": 1.2141956197683752,
+      "grad_norm": 0.12200617045164108,
+      "learning_rate": 0.000674895553828604,
+      "loss": 0.7502,
+      "step": 2648
+    },
+    {
+      "epoch": 1.2146542827657378,
+      "grad_norm": 0.3413892090320587,
+      "learning_rate": 0.0006746635340348232,
+      "loss": 2.4104,
+      "step": 2649
+    },
+    {
+      "epoch": 1.2151129457631005,
+      "grad_norm": 0.3582800626754761,
+      "learning_rate": 0.0006744314713948182,
+      "loss": 1.6539,
+      "step": 2650
+    },
+    {
+      "epoch": 1.2155716087604633,
+      "grad_norm": 0.386416494846344,
+      "learning_rate": 0.0006741993659655155,
+      "loss": 1.6932,
+      "step": 2651
+    },
+    {
+      "epoch": 1.216030271757826,
+      "grad_norm": 0.3171131908893585,
+      "learning_rate": 0.0006739672178038522,
+      "loss": 1.5858,
+      "step": 2652
+    },
+    {
+      "epoch": 1.2164889347551886,
+      "grad_norm": 0.35482868552207947,
+      "learning_rate": 0.0006737350269667763,
+      "loss": 1.8237,
+      "step": 2653
+    },
+    {
+      "epoch": 1.2169475977525512,
+      "grad_norm": 0.3112960457801819,
+      "learning_rate": 0.0006735027935112454,
+      "loss": 1.3225,
+      "step": 2654
+    },
+    {
+      "epoch": 1.217406260749914,
+      "grad_norm": 0.3752141296863556,
+      "learning_rate": 0.0006732705174942283,
+      "loss": 1.3334,
+      "step": 2655
+    },
+    {
+      "epoch": 1.2178649237472767,
+      "grad_norm": 0.29117441177368164,
+      "learning_rate": 0.0006730381989727038,
+      "loss": 1.148,
+      "step": 2656
+    },
+    {
+      "epoch": 1.2183235867446394,
+      "grad_norm": 0.19909793138504028,
+      "learning_rate": 0.0006728058380036614,
+      "loss": 1.4814,
+      "step": 2657
+    },
+    {
+      "epoch": 1.218782249742002,
+      "grad_norm": 0.2802788317203522,
+      "learning_rate": 0.0006725734346441013,
+      "loss": 1.1915,
+      "step": 2658
+    },
+    {
+      "epoch": 1.2192409127393646,
+      "grad_norm": 0.34183627367019653,
+      "learning_rate": 0.000672340988951033,
+      "loss": 1.6281,
+      "step": 2659
+    },
+    {
+      "epoch": 1.2196995757367275,
+      "grad_norm": 0.09204282611608505,
+      "learning_rate": 0.0006721085009814775,
+      "loss": 0.5768,
+      "step": 2660
+    },
+    {
+      "epoch": 1.2201582387340901,
+      "grad_norm": 0.20829735696315765,
+      "learning_rate": 0.0006718759707924658,
+      "loss": 1.2005,
+      "step": 2661
+    },
+    {
+      "epoch": 1.2206169017314528,
+      "grad_norm": 0.23320595920085907,
+      "learning_rate": 0.0006716433984410391,
+      "loss": 1.4799,
+      "step": 2662
+    },
+    {
+      "epoch": 1.2210755647288156,
+      "grad_norm": 0.446114718914032,
+      "learning_rate": 0.0006714107839842493,
+      "loss": 1.2414,
+      "step": 2663
+    },
+    {
+      "epoch": 1.2215342277261783,
+      "grad_norm": 0.26882967352867126,
+      "learning_rate": 0.0006711781274791582,
+      "loss": 1.344,
+      "step": 2664
+    },
+    {
+      "epoch": 1.221992890723541,
+      "grad_norm": 0.20352056622505188,
+      "learning_rate": 0.000670945428982838,
+      "loss": 1.0079,
+      "step": 2665
+    },
+    {
+      "epoch": 1.2224515537209035,
+      "grad_norm": 0.23894716799259186,
+      "learning_rate": 0.0006707126885523717,
+      "loss": 1.8065,
+      "step": 2666
+    },
+    {
+      "epoch": 1.2229102167182662,
+      "grad_norm": 0.3363592028617859,
+      "learning_rate": 0.000670479906244852,
+      "loss": 1.2569,
+      "step": 2667
+    },
+    {
+      "epoch": 1.223368879715629,
+      "grad_norm": 0.30896368622779846,
+      "learning_rate": 0.000670247082117382,
+      "loss": 1.9127,
+      "step": 2668
+    },
+    {
+      "epoch": 1.2238275427129917,
+      "grad_norm": 0.21522556245326996,
+      "learning_rate": 0.0006700142162270753,
+      "loss": 1.5338,
+      "step": 2669
+    },
+    {
+      "epoch": 1.2242862057103543,
+      "grad_norm": 0.30787453055381775,
+      "learning_rate": 0.0006697813086310553,
+      "loss": 1.3445,
+      "step": 2670
+    },
+    {
+      "epoch": 1.224744868707717,
+      "grad_norm": 0.2880922853946686,
+      "learning_rate": 0.0006695483593864562,
+      "loss": 1.2805,
+      "step": 2671
+    },
+    {
+      "epoch": 1.2252035317050798,
+      "grad_norm": 0.2538045346736908,
+      "learning_rate": 0.0006693153685504221,
+      "loss": 1.4706,
+      "step": 2672
+    },
+    {
+      "epoch": 1.2256621947024424,
+      "grad_norm": 0.2669422924518585,
+      "learning_rate": 0.0006690823361801073,
+      "loss": 0.8488,
+      "step": 2673
+    },
+    {
+      "epoch": 1.226120857699805,
+      "grad_norm": 0.09968861192464828,
+      "learning_rate": 0.0006688492623326762,
+      "loss": 0.6257,
+      "step": 2674
+    },
+    {
+      "epoch": 1.2265795206971677,
+      "grad_norm": 0.06321458518505096,
+      "learning_rate": 0.0006686161470653036,
+      "loss": 0.8595,
+      "step": 2675
+    },
+    {
+      "epoch": 1.2270381836945305,
+      "grad_norm": 0.3460922837257385,
+      "learning_rate": 0.0006683829904351742,
+      "loss": 2.0936,
+      "step": 2676
+    },
+    {
+      "epoch": 1.2274968466918932,
+      "grad_norm": 0.28384023904800415,
+      "learning_rate": 0.0006681497924994834,
+      "loss": 0.7923,
+      "step": 2677
+    },
+    {
+      "epoch": 1.2279555096892558,
+      "grad_norm": 0.22630473971366882,
+      "learning_rate": 0.000667916553315436,
+      "loss": 1.5095,
+      "step": 2678
+    },
+    {
+      "epoch": 1.2284141726866185,
+      "grad_norm": 0.3462603688240051,
+      "learning_rate": 0.0006676832729402472,
+      "loss": 1.7954,
+      "step": 2679
+    },
+    {
+      "epoch": 1.228872835683981,
+      "grad_norm": 0.18489515781402588,
+      "learning_rate": 0.0006674499514311426,
+      "loss": 0.5131,
+      "step": 2680
+    },
+    {
+      "epoch": 1.229331498681344,
+      "grad_norm": 0.2563524544239044,
+      "learning_rate": 0.0006672165888453576,
+      "loss": 1.455,
+      "step": 2681
+    },
+    {
+      "epoch": 1.2297901616787066,
+      "grad_norm": 0.28500133752822876,
+      "learning_rate": 0.0006669831852401375,
+      "loss": 1.4278,
+      "step": 2682
+    },
+    {
+      "epoch": 1.2302488246760692,
+      "grad_norm": 0.21729940176010132,
+      "learning_rate": 0.0006667497406727382,
+      "loss": 1.6448,
+      "step": 2683
+    },
+    {
+      "epoch": 1.2307074876734319,
+      "grad_norm": 0.32196274399757385,
+      "learning_rate": 0.0006665162552004251,
+      "loss": 1.5613,
+      "step": 2684
+    },
+    {
+      "epoch": 1.2311661506707947,
+      "grad_norm": 0.2639254927635193,
+      "learning_rate": 0.0006662827288804739,
+      "loss": 1.1826,
+      "step": 2685
+    },
+    {
+      "epoch": 1.2316248136681573,
+      "grad_norm": 0.3229156732559204,
+      "learning_rate": 0.0006660491617701704,
+      "loss": 2.0408,
+      "step": 2686
+    },
+    {
+      "epoch": 1.23208347666552,
+      "grad_norm": 0.3566708266735077,
+      "learning_rate": 0.0006658155539268103,
+      "loss": 1.7527,
+      "step": 2687
+    },
+    {
+      "epoch": 1.2325421396628826,
+      "grad_norm": 0.09745471924543381,
+      "learning_rate": 0.0006655819054076991,
+      "loss": 0.9554,
+      "step": 2688
+    },
+    {
+      "epoch": 1.2330008026602455,
+      "grad_norm": 0.37350574135780334,
+      "learning_rate": 0.0006653482162701528,
+      "loss": 1.5167,
+      "step": 2689
+    },
+    {
+      "epoch": 1.2334594656576081,
+      "grad_norm": 0.3267020285129547,
+      "learning_rate": 0.0006651144865714966,
+      "loss": 2.0956,
+      "step": 2690
+    },
+    {
+      "epoch": 1.2339181286549707,
+      "grad_norm": 0.32065388560295105,
+      "learning_rate": 0.0006648807163690664,
+      "loss": 1.6762,
+      "step": 2691
+    },
+    {
+      "epoch": 1.2343767916523334,
+      "grad_norm": 0.5290999412536621,
+      "learning_rate": 0.0006646469057202076,
+      "loss": 1.6888,
+      "step": 2692
+    },
+    {
+      "epoch": 1.234835454649696,
+      "grad_norm": 0.3192092478275299,
+      "learning_rate": 0.0006644130546822757,
+      "loss": 1.5339,
+      "step": 2693
+    },
+    {
+      "epoch": 1.2352941176470589,
+      "grad_norm": 0.3551962077617645,
+      "learning_rate": 0.000664179163312636,
+      "loss": 1.5905,
+      "step": 2694
+    },
+    {
+      "epoch": 1.2357527806444215,
+      "grad_norm": 0.2710552513599396,
+      "learning_rate": 0.0006639452316686638,
+      "loss": 1.394,
+      "step": 2695
+    },
+    {
+      "epoch": 1.2362114436417841,
+      "grad_norm": 0.3438358008861542,
+      "learning_rate": 0.0006637112598077441,
+      "loss": 1.373,
+      "step": 2696
+    },
+    {
+      "epoch": 1.236670106639147,
+      "grad_norm": 0.33121544122695923,
+      "learning_rate": 0.0006634772477872719,
+      "loss": 1.8278,
+      "step": 2697
+    },
+    {
+      "epoch": 1.2371287696365096,
+      "grad_norm": 0.2892778217792511,
+      "learning_rate": 0.0006632431956646522,
+      "loss": 1.3161,
+      "step": 2698
+    },
+    {
+      "epoch": 1.2375874326338723,
+      "grad_norm": 0.3857962191104889,
+      "learning_rate": 0.0006630091034972995,
+      "loss": 1.9475,
+      "step": 2699
+    },
+    {
+      "epoch": 1.238046095631235,
+      "grad_norm": 0.2920449376106262,
+      "learning_rate": 0.0006627749713426385,
+      "loss": 1.6476,
+      "step": 2700
+    },
+    {
+      "epoch": 1.2385047586285975,
+      "grad_norm": 0.29545149207115173,
+      "learning_rate": 0.000662540799258103,
+      "loss": 1.2706,
+      "step": 2701
+    },
+    {
+      "epoch": 1.2389634216259604,
+      "grad_norm": 0.32311832904815674,
+      "learning_rate": 0.0006623065873011378,
+      "loss": 0.8628,
+      "step": 2702
+    },
+    {
+      "epoch": 1.239422084623323,
+      "grad_norm": 0.20221565663814545,
+      "learning_rate": 0.0006620723355291963,
+      "loss": 0.8695,
+      "step": 2703
+    },
+    {
+      "epoch": 1.2398807476206857,
+      "grad_norm": 0.3797789216041565,
+      "learning_rate": 0.0006618380439997423,
+      "loss": 1.8011,
+      "step": 2704
+    },
+    {
+      "epoch": 1.2403394106180483,
+      "grad_norm": 0.27936404943466187,
+      "learning_rate": 0.0006616037127702493,
+      "loss": 1.7474,
+      "step": 2705
+    },
+    {
+      "epoch": 1.2407980736154112,
+      "grad_norm": 0.305891752243042,
+      "learning_rate": 0.0006613693418982002,
+      "loss": 1.4064,
+      "step": 2706
+    },
+    {
+      "epoch": 1.2412567366127738,
+      "grad_norm": 0.22525465488433838,
+      "learning_rate": 0.0006611349314410881,
+      "loss": 1.634,
+      "step": 2707
+    },
+    {
+      "epoch": 1.2417153996101364,
+      "grad_norm": 0.2716008126735687,
+      "learning_rate": 0.0006609004814564154,
+      "loss": 1.3457,
+      "step": 2708
+    },
+    {
+      "epoch": 1.242174062607499,
+      "grad_norm": 0.28558602929115295,
+      "learning_rate": 0.0006606659920016945,
+      "loss": 1.4251,
+      "step": 2709
+    },
+    {
+      "epoch": 1.242632725604862,
+      "grad_norm": 0.36208319664001465,
+      "learning_rate": 0.0006604314631344472,
+      "loss": 1.4049,
+      "step": 2710
+    },
+    {
+      "epoch": 1.2430913886022246,
+      "grad_norm": 0.3267166316509247,
+      "learning_rate": 0.0006601968949122053,
+      "loss": 1.9824,
+      "step": 2711
+    },
+    {
+      "epoch": 1.2435500515995872,
+      "grad_norm": 0.22049808502197266,
+      "learning_rate": 0.00065996228739251,
+      "loss": 0.8636,
+      "step": 2712
+    },
+    {
+      "epoch": 1.2440087145969498,
+      "grad_norm": 0.17521077394485474,
+      "learning_rate": 0.0006597276406329122,
+      "loss": 1.3998,
+      "step": 2713
+    },
+    {
+      "epoch": 1.2444673775943125,
+      "grad_norm": 0.4500933289527893,
+      "learning_rate": 0.0006594929546909725,
+      "loss": 1.4056,
+      "step": 2714
+    },
+    {
+      "epoch": 1.2449260405916753,
+      "grad_norm": 0.37240904569625854,
+      "learning_rate": 0.0006592582296242609,
+      "loss": 1.4709,
+      "step": 2715
+    },
+    {
+      "epoch": 1.245384703589038,
+      "grad_norm": 0.3110402226448059,
+      "learning_rate": 0.0006590234654903574,
+      "loss": 1.8608,
+      "step": 2716
+    },
+    {
+      "epoch": 1.2458433665864006,
+      "grad_norm": 0.2775774896144867,
+      "learning_rate": 0.0006587886623468511,
+      "loss": 1.3253,
+      "step": 2717
+    },
+    {
+      "epoch": 1.2463020295837632,
+      "grad_norm": 0.2369251698255539,
+      "learning_rate": 0.000658553820251341,
+      "loss": 0.7903,
+      "step": 2718
+    },
+    {
+      "epoch": 1.246760692581126,
+      "grad_norm": 0.25416409969329834,
+      "learning_rate": 0.0006583189392614356,
+      "loss": 1.2426,
+      "step": 2719
+    },
+    {
+      "epoch": 1.2472193555784887,
+      "grad_norm": 0.36366981267929077,
+      "learning_rate": 0.0006580840194347529,
+      "loss": 2.1185,
+      "step": 2720
+    },
+    {
+      "epoch": 1.2476780185758514,
+      "grad_norm": 0.21823568642139435,
+      "learning_rate": 0.0006578490608289204,
+      "loss": 0.8336,
+      "step": 2721
+    },
+    {
+      "epoch": 1.248136681573214,
+      "grad_norm": 0.2580203115940094,
+      "learning_rate": 0.0006576140635015749,
+      "loss": 1.243,
+      "step": 2722
+    },
+    {
+      "epoch": 1.2485953445705769,
+      "grad_norm": 0.19676536321640015,
+      "learning_rate": 0.0006573790275103635,
+      "loss": 1.3928,
+      "step": 2723
+    },
+    {
+      "epoch": 1.2490540075679395,
+      "grad_norm": 0.3913279175758362,
+      "learning_rate": 0.0006571439529129417,
+      "loss": 1.6071,
+      "step": 2724
+    },
+    {
+      "epoch": 1.2495126705653021,
+      "grad_norm": 0.33274513483047485,
+      "learning_rate": 0.0006569088397669752,
+      "loss": 1.9368,
+      "step": 2725
+    },
+    {
+      "epoch": 1.2499713335626648,
+      "grad_norm": 0.3166465759277344,
+      "learning_rate": 0.0006566736881301389,
+      "loss": 1.193,
+      "step": 2726
+    },
+    {
+      "epoch": 1.2504299965600274,
+      "grad_norm": 0.12209037691354752,
+      "learning_rate": 0.0006564384980601172,
+      "loss": 0.7718,
+      "step": 2727
+    },
+    {
+      "epoch": 1.2508886595573903,
+      "grad_norm": 0.19902729988098145,
+      "learning_rate": 0.000656203269614604,
+      "loss": 0.83,
+      "step": 2728
+    },
+    {
+      "epoch": 1.251347322554753,
+      "grad_norm": 0.23850427567958832,
+      "learning_rate": 0.0006559680028513027,
+      "loss": 1.1411,
+      "step": 2729
+    },
+    {
+      "epoch": 1.2518059855521155,
+      "grad_norm": 0.22274386882781982,
+      "learning_rate": 0.0006557326978279255,
+      "loss": 1.2014,
+      "step": 2730
+    },
+    {
+      "epoch": 1.2522646485494784,
+      "grad_norm": 0.2648008465766907,
+      "learning_rate": 0.0006554973546021946,
+      "loss": 1.1293,
+      "step": 2731
+    },
+    {
+      "epoch": 1.252723311546841,
+      "grad_norm": 0.20041297376155853,
+      "learning_rate": 0.0006552619732318414,
+      "loss": 0.9521,
+      "step": 2732
+    },
+    {
+      "epoch": 1.2531819745442037,
+      "grad_norm": 0.3079836666584015,
+      "learning_rate": 0.0006550265537746068,
+      "loss": 1.9308,
+      "step": 2733
+    },
+    {
+      "epoch": 1.2536406375415663,
+      "grad_norm": 0.3721216917037964,
+      "learning_rate": 0.0006547910962882407,
+      "loss": 1.7333,
+      "step": 2734
+    },
+    {
+      "epoch": 1.254099300538929,
+      "grad_norm": 0.37531670928001404,
+      "learning_rate": 0.0006545556008305025,
+      "loss": 1.4659,
+      "step": 2735
+    },
+    {
+      "epoch": 1.2545579635362918,
+      "grad_norm": 0.24703913927078247,
+      "learning_rate": 0.0006543200674591611,
+      "loss": 1.6251,
+      "step": 2736
+    },
+    {
+      "epoch": 1.2550166265336544,
+      "grad_norm": 0.3122749924659729,
+      "learning_rate": 0.0006540844962319944,
+      "loss": 0.7767,
+      "step": 2737
+    },
+    {
+      "epoch": 1.255475289531017,
+      "grad_norm": 0.41236305236816406,
+      "learning_rate": 0.0006538488872067899,
+      "loss": 2.0791,
+      "step": 2738
+    },
+    {
+      "epoch": 1.2559339525283797,
+      "grad_norm": 0.3451601266860962,
+      "learning_rate": 0.000653613240441344,
+      "loss": 1.5688,
+      "step": 2739
+    },
+    {
+      "epoch": 1.2563926155257423,
+      "grad_norm": 0.27274802327156067,
+      "learning_rate": 0.0006533775559934624,
+      "loss": 1.2577,
+      "step": 2740
+    },
+    {
+      "epoch": 1.2568512785231052,
+      "grad_norm": 0.42158329486846924,
+      "learning_rate": 0.0006531418339209607,
+      "loss": 1.9504,
+      "step": 2741
+    },
+    {
+      "epoch": 1.2573099415204678,
+      "grad_norm": 0.336901992559433,
+      "learning_rate": 0.0006529060742816627,
+      "loss": 1.6164,
+      "step": 2742
+    },
+    {
+      "epoch": 1.2577686045178305,
+      "grad_norm": 0.2214047908782959,
+      "learning_rate": 0.0006526702771334023,
+      "loss": 1.2476,
+      "step": 2743
+    },
+    {
+      "epoch": 1.2582272675151933,
+      "grad_norm": 0.2765026092529297,
+      "learning_rate": 0.000652434442534022,
+      "loss": 1.2798,
+      "step": 2744
+    },
+    {
+      "epoch": 1.258685930512556,
+      "grad_norm": 0.4691556692123413,
+      "learning_rate": 0.0006521985705413741,
+      "loss": 2.1313,
+      "step": 2745
+    },
+    {
+      "epoch": 1.2591445935099186,
+      "grad_norm": 0.2610776126384735,
+      "learning_rate": 0.0006519626612133192,
+      "loss": 0.9233,
+      "step": 2746
+    },
+    {
+      "epoch": 1.2596032565072812,
+      "grad_norm": 0.26943257451057434,
+      "learning_rate": 0.0006517267146077279,
+      "loss": 1.618,
+      "step": 2747
+    },
+    {
+      "epoch": 1.2600619195046439,
+      "grad_norm": 0.3554578125476837,
+      "learning_rate": 0.0006514907307824794,
+      "loss": 1.9045,
+      "step": 2748
+    },
+    {
+      "epoch": 1.2605205825020067,
+      "grad_norm": 0.28835994005203247,
+      "learning_rate": 0.0006512547097954624,
+      "loss": 0.8852,
+      "step": 2749
+    },
+    {
+      "epoch": 1.2609792454993694,
+      "grad_norm": 0.12801244854927063,
+      "learning_rate": 0.0006510186517045744,
+      "loss": 0.7682,
+      "step": 2750
+    },
+    {
+      "epoch": 1.261437908496732,
+      "grad_norm": 0.3506546914577484,
+      "learning_rate": 0.0006507825565677225,
+      "loss": 2.2786,
+      "step": 2751
+    },
+    {
+      "epoch": 1.2618965714940948,
+      "grad_norm": 0.36096346378326416,
+      "learning_rate": 0.0006505464244428219,
+      "loss": 1.8541,
+      "step": 2752
+    },
+    {
+      "epoch": 1.2623552344914575,
+      "grad_norm": 0.18552398681640625,
+      "learning_rate": 0.0006503102553877982,
+      "loss": 0.4271,
+      "step": 2753
+    },
+    {
+      "epoch": 1.2628138974888201,
+      "grad_norm": 0.11175768822431564,
+      "learning_rate": 0.0006500740494605848,
+      "loss": 0.7008,
+      "step": 2754
+    },
+    {
+      "epoch": 1.2632725604861827,
+      "grad_norm": 0.34600672125816345,
+      "learning_rate": 0.0006498378067191252,
+      "loss": 2.1968,
+      "step": 2755
+    },
+    {
+      "epoch": 1.2637312234835454,
+      "grad_norm": 0.32166171073913574,
+      "learning_rate": 0.0006496015272213711,
+      "loss": 1.4522,
+      "step": 2756
+    },
+    {
+      "epoch": 1.2641898864809082,
+      "grad_norm": 0.3662795126438141,
+      "learning_rate": 0.0006493652110252838,
+      "loss": 2.3257,
+      "step": 2757
+    },
+    {
+      "epoch": 1.2646485494782709,
+      "grad_norm": 0.2909261882305145,
+      "learning_rate": 0.000649128858188833,
+      "loss": 1.1611,
+      "step": 2758
+    },
+    {
+      "epoch": 1.2651072124756335,
+      "grad_norm": 0.4382424056529999,
+      "learning_rate": 0.0006488924687699983,
+      "loss": 1.4101,
+      "step": 2759
+    },
+    {
+      "epoch": 1.2655658754729961,
+      "grad_norm": 0.17091111838817596,
+      "learning_rate": 0.0006486560428267674,
+      "loss": 0.8147,
+      "step": 2760
+    },
+    {
+      "epoch": 1.2660245384703588,
+      "grad_norm": 0.3439600169658661,
+      "learning_rate": 0.0006484195804171371,
+      "loss": 1.7435,
+      "step": 2761
+    },
+    {
+      "epoch": 1.2664832014677216,
+      "grad_norm": 0.3564503490924835,
+      "learning_rate": 0.0006481830815991138,
+      "loss": 1.1747,
+      "step": 2762
+    },
+    {
+      "epoch": 1.2669418644650843,
+      "grad_norm": 0.2912631034851074,
+      "learning_rate": 0.000647946546430712,
+      "loss": 1.6711,
+      "step": 2763
+    },
+    {
+      "epoch": 1.267400527462447,
+      "grad_norm": 0.4116171598434448,
+      "learning_rate": 0.0006477099749699557,
+      "loss": 2.1626,
+      "step": 2764
+    },
+    {
+      "epoch": 1.2678591904598098,
+      "grad_norm": 0.2863385081291199,
+      "learning_rate": 0.0006474733672748775,
+      "loss": 0.9852,
+      "step": 2765
+    },
+    {
+      "epoch": 1.2683178534571724,
+      "grad_norm": 0.30499163269996643,
+      "learning_rate": 0.0006472367234035189,
+      "loss": 1.5042,
+      "step": 2766
+    },
+    {
+      "epoch": 1.268776516454535,
+      "grad_norm": 0.20567883551120758,
+      "learning_rate": 0.0006470000434139306,
+      "loss": 0.7004,
+      "step": 2767
+    },
+    {
+      "epoch": 1.2692351794518977,
+      "grad_norm": 0.1941119283437729,
+      "learning_rate": 0.0006467633273641714,
+      "loss": 1.2625,
+      "step": 2768
+    },
+    {
+      "epoch": 1.2696938424492603,
+      "grad_norm": 0.33352068066596985,
+      "learning_rate": 0.00064652657531231,
+      "loss": 1.3027,
+      "step": 2769
+    },
+    {
+      "epoch": 1.2701525054466232,
+      "grad_norm": 0.23452205955982208,
+      "learning_rate": 0.0006462897873164232,
+      "loss": 1.4857,
+      "step": 2770
+    },
+    {
+      "epoch": 1.2706111684439858,
+      "grad_norm": 0.3402770757675171,
+      "learning_rate": 0.0006460529634345967,
+      "loss": 1.4496,
+      "step": 2771
+    },
+    {
+      "epoch": 1.2710698314413484,
+      "grad_norm": 0.25416603684425354,
+      "learning_rate": 0.000645816103724925,
+      "loss": 1.1967,
+      "step": 2772
+    },
+    {
+      "epoch": 1.271528494438711,
+      "grad_norm": 0.29919010400772095,
+      "learning_rate": 0.0006455792082455118,
+      "loss": 1.3418,
+      "step": 2773
+    },
+    {
+      "epoch": 1.2719871574360737,
+      "grad_norm": 0.2903623580932617,
+      "learning_rate": 0.0006453422770544692,
+      "loss": 1.3553,
+      "step": 2774
+    },
+    {
+      "epoch": 1.2724458204334366,
+      "grad_norm": 0.478496253490448,
+      "learning_rate": 0.0006451053102099181,
+      "loss": 2.0164,
+      "step": 2775
+    },
+    {
+      "epoch": 1.2729044834307992,
+      "grad_norm": 0.2910837233066559,
+      "learning_rate": 0.000644868307769988,
+      "loss": 1.1557,
+      "step": 2776
+    },
+    {
+      "epoch": 1.2733631464281618,
+      "grad_norm": 0.19006802141666412,
+      "learning_rate": 0.0006446312697928176,
+      "loss": 1.3286,
+      "step": 2777
+    },
+    {
+      "epoch": 1.2738218094255247,
+      "grad_norm": 0.23939000070095062,
+      "learning_rate": 0.0006443941963365539,
+      "loss": 0.9206,
+      "step": 2778
+    },
+    {
+      "epoch": 1.2742804724228873,
+      "grad_norm": 0.34608039259910583,
+      "learning_rate": 0.0006441570874593524,
+      "loss": 2.2271,
+      "step": 2779
+    },
+    {
+      "epoch": 1.27473913542025,
+      "grad_norm": 0.3570116460323334,
+      "learning_rate": 0.0006439199432193782,
+      "loss": 1.7172,
+      "step": 2780
+    },
+    {
+      "epoch": 1.2751977984176126,
+      "grad_norm": 0.21899329125881195,
+      "learning_rate": 0.0006436827636748042,
+      "loss": 1.1881,
+      "step": 2781
+    },
+    {
+      "epoch": 1.2756564614149752,
+      "grad_norm": 0.2552841305732727,
+      "learning_rate": 0.0006434455488838121,
+      "loss": 1.5012,
+      "step": 2782
+    },
+    {
+      "epoch": 1.276115124412338,
+      "grad_norm": 0.3439258635044098,
+      "learning_rate": 0.0006432082989045926,
+      "loss": 1.7224,
+      "step": 2783
+    },
+    {
+      "epoch": 1.2765737874097007,
+      "grad_norm": 0.12296123802661896,
+      "learning_rate": 0.000642971013795345,
+      "loss": 1.0084,
+      "step": 2784
+    },
+    {
+      "epoch": 1.2770324504070634,
+      "grad_norm": 0.37332117557525635,
+      "learning_rate": 0.0006427336936142766,
+      "loss": 1.8006,
+      "step": 2785
+    },
+    {
+      "epoch": 1.2774911134044262,
+      "grad_norm": 0.1939849704504013,
+      "learning_rate": 0.0006424963384196041,
+      "loss": 0.8541,
+      "step": 2786
+    },
+    {
+      "epoch": 1.2779497764017889,
+      "grad_norm": 0.3263346552848816,
+      "learning_rate": 0.0006422589482695523,
+      "loss": 1.8838,
+      "step": 2787
+    },
+    {
+      "epoch": 1.2784084393991515,
+      "grad_norm": 0.07417495548725128,
+      "learning_rate": 0.0006420215232223548,
+      "loss": 0.3335,
+      "step": 2788
+    },
+    {
+      "epoch": 1.2788671023965141,
+      "grad_norm": 0.27612733840942383,
+      "learning_rate": 0.0006417840633362535,
+      "loss": 1.3939,
+      "step": 2789
+    },
+    {
+      "epoch": 1.2793257653938768,
+      "grad_norm": 0.31349727511405945,
+      "learning_rate": 0.0006415465686694993,
+      "loss": 1.7898,
+      "step": 2790
+    },
+    {
+      "epoch": 1.2797844283912396,
+      "grad_norm": 0.3115302622318268,
+      "learning_rate": 0.0006413090392803511,
+      "loss": 1.4323,
+      "step": 2791
+    },
+    {
+      "epoch": 1.2802430913886023,
+      "grad_norm": 0.29694634675979614,
+      "learning_rate": 0.0006410714752270769,
+      "loss": 1.7373,
+      "step": 2792
+    },
+    {
+      "epoch": 1.280701754385965,
+      "grad_norm": 0.2605624496936798,
+      "learning_rate": 0.0006408338765679525,
+      "loss": 1.2168,
+      "step": 2793
+    },
+    {
+      "epoch": 1.2811604173833275,
+      "grad_norm": 0.4056708514690399,
+      "learning_rate": 0.0006405962433612625,
+      "loss": 1.701,
+      "step": 2794
+    },
+    {
+      "epoch": 1.2816190803806902,
+      "grad_norm": 0.30486351251602173,
+      "learning_rate": 0.0006403585756653004,
+      "loss": 0.8638,
+      "step": 2795
+    },
+    {
+      "epoch": 1.282077743378053,
+      "grad_norm": 0.27140724658966064,
+      "learning_rate": 0.0006401208735383677,
+      "loss": 1.5908,
+      "step": 2796
+    },
+    {
+      "epoch": 1.2825364063754157,
+      "grad_norm": 0.30335965752601624,
+      "learning_rate": 0.0006398831370387744,
+      "loss": 1.1826,
+      "step": 2797
+    },
+    {
+      "epoch": 1.2829950693727783,
+      "grad_norm": 0.22256600856781006,
+      "learning_rate": 0.0006396453662248391,
+      "loss": 1.0292,
+      "step": 2798
+    },
+    {
+      "epoch": 1.2834537323701412,
+      "grad_norm": 0.34704825282096863,
+      "learning_rate": 0.0006394075611548882,
+      "loss": 1.5649,
+      "step": 2799
+    },
+    {
+      "epoch": 1.2839123953675038,
+      "grad_norm": 0.4331229031085968,
+      "learning_rate": 0.0006391697218872576,
+      "loss": 1.5117,
+      "step": 2800
+    },
+    {
+      "epoch": 1.2843710583648664,
+      "grad_norm": 0.25863364338874817,
+      "learning_rate": 0.0006389318484802908,
+      "loss": 1.5621,
+      "step": 2801
+    },
+    {
+      "epoch": 1.284829721362229,
+      "grad_norm": 0.31081312894821167,
+      "learning_rate": 0.0006386939409923398,
+      "loss": 1.7057,
+      "step": 2802
+    },
+    {
+      "epoch": 1.2852883843595917,
+      "grad_norm": 0.3220570683479309,
+      "learning_rate": 0.0006384559994817649,
+      "loss": 0.9301,
+      "step": 2803
+    },
+    {
+      "epoch": 1.2857470473569546,
+      "grad_norm": 0.32526329159736633,
+      "learning_rate": 0.0006382180240069352,
+      "loss": 2.1208,
+      "step": 2804
+    },
+    {
+      "epoch": 1.2862057103543172,
+      "grad_norm": 0.33022022247314453,
+      "learning_rate": 0.0006379800146262274,
+      "loss": 1.1948,
+      "step": 2805
+    },
+    {
+      "epoch": 1.2866643733516798,
+      "grad_norm": 0.28297072649002075,
+      "learning_rate": 0.0006377419713980274,
+      "loss": 1.351,
+      "step": 2806
+    },
+    {
+      "epoch": 1.2871230363490425,
+      "grad_norm": 0.23854339122772217,
+      "learning_rate": 0.0006375038943807284,
+      "loss": 1.46,
+      "step": 2807
+    },
+    {
+      "epoch": 1.287581699346405,
+      "grad_norm": 0.2918485999107361,
+      "learning_rate": 0.0006372657836327328,
+      "loss": 0.8088,
+      "step": 2808
+    },
+    {
+      "epoch": 1.288040362343768,
+      "grad_norm": 0.19841410219669342,
+      "learning_rate": 0.0006370276392124506,
+      "loss": 1.2504,
+      "step": 2809
+    },
+    {
+      "epoch": 1.2884990253411306,
+      "grad_norm": 0.28576546907424927,
+      "learning_rate": 0.0006367894611783006,
+      "loss": 1.4295,
+      "step": 2810
+    },
+    {
+      "epoch": 1.2889576883384932,
+      "grad_norm": 0.32226133346557617,
+      "learning_rate": 0.0006365512495887094,
+      "loss": 1.6712,
+      "step": 2811
+    },
+    {
+      "epoch": 1.289416351335856,
+      "grad_norm": 0.29555824398994446,
+      "learning_rate": 0.0006363130045021121,
+      "loss": 1.2164,
+      "step": 2812
+    },
+    {
+      "epoch": 1.2898750143332187,
+      "grad_norm": 0.17073696851730347,
+      "learning_rate": 0.0006360747259769521,
+      "loss": 0.9707,
+      "step": 2813
+    },
+    {
+      "epoch": 1.2903336773305814,
+      "grad_norm": 0.19965320825576782,
+      "learning_rate": 0.0006358364140716805,
+      "loss": 1.3948,
+      "step": 2814
+    },
+    {
+      "epoch": 1.290792340327944,
+      "grad_norm": 0.2776814103126526,
+      "learning_rate": 0.0006355980688447571,
+      "loss": 1.6216,
+      "step": 2815
+    },
+    {
+      "epoch": 1.2912510033253066,
+      "grad_norm": 0.29604098200798035,
+      "learning_rate": 0.0006353596903546497,
+      "loss": 1.0638,
+      "step": 2816
+    },
+    {
+      "epoch": 1.2917096663226695,
+      "grad_norm": 0.2199414223432541,
+      "learning_rate": 0.0006351212786598341,
+      "loss": 1.8243,
+      "step": 2817
+    },
+    {
+      "epoch": 1.2921683293200321,
+      "grad_norm": 0.4345381259918213,
+      "learning_rate": 0.0006348828338187949,
+      "loss": 2.1404,
+      "step": 2818
+    },
+    {
+      "epoch": 1.2926269923173948,
+      "grad_norm": 0.303279310464859,
+      "learning_rate": 0.0006346443558900239,
+      "loss": 1.6801,
+      "step": 2819
+    },
+    {
+      "epoch": 1.2930856553147576,
+      "grad_norm": 0.32503482699394226,
+      "learning_rate": 0.0006344058449320215,
+      "loss": 1.7594,
+      "step": 2820
+    },
+    {
+      "epoch": 1.2935443183121202,
+      "grad_norm": 0.2687481641769409,
+      "learning_rate": 0.0006341673010032962,
+      "loss": 0.9268,
+      "step": 2821
+    },
+    {
+      "epoch": 1.2940029813094829,
+      "grad_norm": 0.21548904478549957,
+      "learning_rate": 0.0006339287241623646,
+      "loss": 1.2141,
+      "step": 2822
+    },
+    {
+      "epoch": 1.2944616443068455,
+      "grad_norm": 0.2961852550506592,
+      "learning_rate": 0.0006336901144677514,
+      "loss": 1.4321,
+      "step": 2823
+    },
+    {
+      "epoch": 1.2949203073042082,
+      "grad_norm": 0.3688824474811554,
+      "learning_rate": 0.0006334514719779891,
+      "loss": 1.7928,
+      "step": 2824
+    },
+    {
+      "epoch": 1.295378970301571,
+      "grad_norm": 0.3493098318576813,
+      "learning_rate": 0.0006332127967516182,
+      "loss": 1.9301,
+      "step": 2825
+    },
+    {
+      "epoch": 1.2958376332989336,
+      "grad_norm": 0.344375878572464,
+      "learning_rate": 0.0006329740888471881,
+      "loss": 1.4666,
+      "step": 2826
+    },
+    {
+      "epoch": 1.2962962962962963,
+      "grad_norm": 0.36221787333488464,
+      "learning_rate": 0.000632735348323255,
+      "loss": 1.4453,
+      "step": 2827
+    },
+    {
+      "epoch": 1.296754959293659,
+      "grad_norm": 0.3239736557006836,
+      "learning_rate": 0.000632496575238384,
+      "loss": 1.926,
+      "step": 2828
+    },
+    {
+      "epoch": 1.2972136222910216,
+      "grad_norm": 0.20216989517211914,
+      "learning_rate": 0.0006322577696511479,
+      "loss": 0.4995,
+      "step": 2829
+    },
+    {
+      "epoch": 1.2976722852883844,
+      "grad_norm": 0.713234007358551,
+      "learning_rate": 0.000632018931620127,
+      "loss": 1.2388,
+      "step": 2830
+    },
+    {
+      "epoch": 1.298130948285747,
+      "grad_norm": 0.39462095499038696,
+      "learning_rate": 0.0006317800612039103,
+      "loss": 1.3626,
+      "step": 2831
+    },
+    {
+      "epoch": 1.2985896112831097,
+      "grad_norm": 0.3811722993850708,
+      "learning_rate": 0.0006315411584610946,
+      "loss": 1.5548,
+      "step": 2832
+    },
+    {
+      "epoch": 1.2990482742804725,
+      "grad_norm": 0.1306716352701187,
+      "learning_rate": 0.0006313022234502841,
+      "loss": 1.1012,
+      "step": 2833
+    },
+    {
+      "epoch": 1.2995069372778352,
+      "grad_norm": 0.37927016615867615,
+      "learning_rate": 0.0006310632562300917,
+      "loss": 1.6273,
+      "step": 2834
+    },
+    {
+      "epoch": 1.2999656002751978,
+      "grad_norm": 0.19377583265304565,
+      "learning_rate": 0.0006308242568591371,
+      "loss": 1.2111,
+      "step": 2835
+    },
+    {
+      "epoch": 1.3004242632725604,
+      "grad_norm": 0.3456363379955292,
+      "learning_rate": 0.0006305852253960493,
+      "loss": 2.2551,
+      "step": 2836
+    },
+    {
+      "epoch": 1.300882926269923,
+      "grad_norm": 0.3247735798358917,
+      "learning_rate": 0.0006303461618994642,
+      "loss": 1.5282,
+      "step": 2837
+    },
+    {
+      "epoch": 1.301341589267286,
+      "grad_norm": 0.22547945380210876,
+      "learning_rate": 0.0006301070664280255,
+      "loss": 1.5934,
+      "step": 2838
+    },
+    {
+      "epoch": 1.3018002522646486,
+      "grad_norm": 0.26010963320732117,
+      "learning_rate": 0.0006298679390403854,
+      "loss": 1.6635,
+      "step": 2839
+    },
+    {
+      "epoch": 1.3022589152620112,
+      "grad_norm": 0.2988232672214508,
+      "learning_rate": 0.0006296287797952034,
+      "loss": 1.1585,
+      "step": 2840
+    },
+    {
+      "epoch": 1.3027175782593738,
+      "grad_norm": 0.3130618631839752,
+      "learning_rate": 0.0006293895887511472,
+      "loss": 2.3552,
+      "step": 2841
+    },
+    {
+      "epoch": 1.3031762412567365,
+      "grad_norm": 0.40085339546203613,
+      "learning_rate": 0.0006291503659668916,
+      "loss": 1.7498,
+      "step": 2842
+    },
+    {
+      "epoch": 1.3036349042540993,
+      "grad_norm": 0.3021315336227417,
+      "learning_rate": 0.0006289111115011204,
+      "loss": 1.6539,
+      "step": 2843
+    },
+    {
+      "epoch": 1.304093567251462,
+      "grad_norm": 0.32519641518592834,
+      "learning_rate": 0.0006286718254125238,
+      "loss": 1.626,
+      "step": 2844
+    },
+    {
+      "epoch": 1.3045522302488246,
+      "grad_norm": 0.34999242424964905,
+      "learning_rate": 0.0006284325077598008,
+      "loss": 1.4192,
+      "step": 2845
+    },
+    {
+      "epoch": 1.3050108932461875,
+      "grad_norm": 0.48127058148384094,
+      "learning_rate": 0.0006281931586016576,
+      "loss": 1.2933,
+      "step": 2846
+    },
+    {
+      "epoch": 1.30546955624355,
+      "grad_norm": 0.3110017776489258,
+      "learning_rate": 0.0006279537779968082,
+      "loss": 1.4093,
+      "step": 2847
+    },
+    {
+      "epoch": 1.3059282192409127,
+      "grad_norm": 0.28830286860466003,
+      "learning_rate": 0.0006277143660039746,
+      "loss": 1.553,
+      "step": 2848
+    },
+    {
+      "epoch": 1.3063868822382754,
+      "grad_norm": 0.29233160614967346,
+      "learning_rate": 0.000627474922681886,
+      "loss": 1.988,
+      "step": 2849
+    },
+    {
+      "epoch": 1.306845545235638,
+      "grad_norm": 0.3408198058605194,
+      "learning_rate": 0.0006272354480892797,
+      "loss": 1.4908,
+      "step": 2850
+    },
+    {
+      "epoch": 1.3073042082330009,
+      "grad_norm": 0.3165978491306305,
+      "learning_rate": 0.0006269959422849007,
+      "loss": 1.6967,
+      "step": 2851
+    },
+    {
+      "epoch": 1.3077628712303635,
+      "grad_norm": 0.2932069003582001,
+      "learning_rate": 0.0006267564053275012,
+      "loss": 1.9072,
+      "step": 2852
+    },
+    {
+      "epoch": 1.3082215342277261,
+      "grad_norm": 0.35463714599609375,
+      "learning_rate": 0.0006265168372758417,
+      "loss": 1.5883,
+      "step": 2853
+    },
+    {
+      "epoch": 1.308680197225089,
+      "grad_norm": 0.32562148571014404,
+      "learning_rate": 0.0006262772381886896,
+      "loss": 1.4199,
+      "step": 2854
+    },
+    {
+      "epoch": 1.3091388602224516,
+      "grad_norm": 0.3598746955394745,
+      "learning_rate": 0.0006260376081248205,
+      "loss": 1.2422,
+      "step": 2855
+    },
+    {
+      "epoch": 1.3095975232198143,
+      "grad_norm": 0.23722459375858307,
+      "learning_rate": 0.0006257979471430174,
+      "loss": 1.5364,
+      "step": 2856
+    },
+    {
+      "epoch": 1.310056186217177,
+      "grad_norm": 0.25034475326538086,
+      "learning_rate": 0.0006255582553020708,
+      "loss": 1.2079,
+      "step": 2857
+    },
+    {
+      "epoch": 1.3105148492145395,
+      "grad_norm": 0.3492996096611023,
+      "learning_rate": 0.000625318532660779,
+      "loss": 1.614,
+      "step": 2858
+    },
+    {
+      "epoch": 1.3109735122119024,
+      "grad_norm": 0.3420841097831726,
+      "learning_rate": 0.0006250787792779473,
+      "loss": 1.7344,
+      "step": 2859
+    },
+    {
+      "epoch": 1.311432175209265,
+      "grad_norm": 0.3418578505516052,
+      "learning_rate": 0.0006248389952123895,
+      "loss": 2.0039,
+      "step": 2860
+    },
+    {
+      "epoch": 1.3118908382066277,
+      "grad_norm": 0.4057673513889313,
+      "learning_rate": 0.0006245991805229256,
+      "loss": 1.4351,
+      "step": 2861
+    },
+    {
+      "epoch": 1.3123495012039903,
+      "grad_norm": 0.1766156405210495,
+      "learning_rate": 0.0006243593352683848,
+      "loss": 0.7945,
+      "step": 2862
+    },
+    {
+      "epoch": 1.312808164201353,
+      "grad_norm": 0.32402530312538147,
+      "learning_rate": 0.0006241194595076019,
+      "loss": 1.3118,
+      "step": 2863
+    },
+    {
+      "epoch": 1.3132668271987158,
+      "grad_norm": 0.24747517704963684,
+      "learning_rate": 0.000623879553299421,
+      "loss": 1.4999,
+      "step": 2864
+    },
+    {
+      "epoch": 1.3137254901960784,
+      "grad_norm": 0.4034758508205414,
+      "learning_rate": 0.0006236396167026923,
+      "loss": 1.2577,
+      "step": 2865
+    },
+    {
+      "epoch": 1.314184153193441,
+      "grad_norm": 0.2698371410369873,
+      "learning_rate": 0.0006233996497762741,
+      "loss": 1.4708,
+      "step": 2866
+    },
+    {
+      "epoch": 1.314642816190804,
+      "grad_norm": 0.26158544421195984,
+      "learning_rate": 0.000623159652579032,
+      "loss": 1.4747,
+      "step": 2867
+    },
+    {
+      "epoch": 1.3151014791881666,
+      "grad_norm": 0.3202878534793854,
+      "learning_rate": 0.0006229196251698394,
+      "loss": 1.4017,
+      "step": 2868
+    },
+    {
+      "epoch": 1.3155601421855292,
+      "grad_norm": 0.33199355006217957,
+      "learning_rate": 0.0006226795676075762,
+      "loss": 1.305,
+      "step": 2869
+    },
+    {
+      "epoch": 1.3160188051828918,
+      "grad_norm": 0.29111596941947937,
+      "learning_rate": 0.0006224394799511306,
+      "loss": 1.4327,
+      "step": 2870
+    },
+    {
+      "epoch": 1.3164774681802545,
+      "grad_norm": 0.2958594858646393,
+      "learning_rate": 0.0006221993622593975,
+      "loss": 1.2742,
+      "step": 2871
+    },
+    {
+      "epoch": 1.3169361311776173,
+      "grad_norm": 0.08179383724927902,
+      "learning_rate": 0.00062195921459128,
+      "loss": 0.8383,
+      "step": 2872
+    },
+    {
+      "epoch": 1.31739479417498,
+      "grad_norm": 0.3127470314502716,
+      "learning_rate": 0.0006217190370056876,
+      "loss": 1.7184,
+      "step": 2873
+    },
+    {
+      "epoch": 1.3178534571723426,
+      "grad_norm": 0.2972666621208191,
+      "learning_rate": 0.0006214788295615381,
+      "loss": 1.6077,
+      "step": 2874
+    },
+    {
+      "epoch": 1.3183121201697052,
+      "grad_norm": 0.3870674669742584,
+      "learning_rate": 0.0006212385923177556,
+      "loss": 1.2744,
+      "step": 2875
+    },
+    {
+      "epoch": 1.3187707831670679,
+      "grad_norm": 0.2808718979358673,
+      "learning_rate": 0.0006209983253332721,
+      "loss": 1.2932,
+      "step": 2876
+    },
+    {
+      "epoch": 1.3192294461644307,
+      "grad_norm": 0.2264634668827057,
+      "learning_rate": 0.0006207580286670271,
+      "loss": 1.3213,
+      "step": 2877
+    },
+    {
+      "epoch": 1.3196881091617934,
+      "grad_norm": 0.3030881881713867,
+      "learning_rate": 0.0006205177023779669,
+      "loss": 1.1255,
+      "step": 2878
+    },
+    {
+      "epoch": 1.320146772159156,
+      "grad_norm": 0.3098306953907013,
+      "learning_rate": 0.0006202773465250456,
+      "loss": 1.3304,
+      "step": 2879
+    },
+    {
+      "epoch": 1.3206054351565188,
+      "grad_norm": 0.26796406507492065,
+      "learning_rate": 0.0006200369611672236,
+      "loss": 1.2411,
+      "step": 2880
+    },
+    {
+      "epoch": 1.3210640981538815,
+      "grad_norm": 0.20453116297721863,
+      "learning_rate": 0.0006197965463634696,
+      "loss": 0.9436,
+      "step": 2881
+    },
+    {
+      "epoch": 1.3215227611512441,
+      "grad_norm": 0.22560381889343262,
+      "learning_rate": 0.000619556102172759,
+      "loss": 1.673,
+      "step": 2882
+    },
+    {
+      "epoch": 1.3219814241486068,
+      "grad_norm": 0.24689380824565887,
+      "learning_rate": 0.0006193156286540745,
+      "loss": 0.8455,
+      "step": 2883
+    },
+    {
+      "epoch": 1.3224400871459694,
+      "grad_norm": 0.24661903083324432,
+      "learning_rate": 0.000619075125866406,
+      "loss": 0.9642,
+      "step": 2884
+    },
+    {
+      "epoch": 1.3228987501433322,
+      "grad_norm": 0.2800322473049164,
+      "learning_rate": 0.0006188345938687506,
+      "loss": 1.3403,
+      "step": 2885
+    },
+    {
+      "epoch": 1.3233574131406949,
+      "grad_norm": 0.21044687926769257,
+      "learning_rate": 0.0006185940327201125,
+      "loss": 0.7411,
+      "step": 2886
+    },
+    {
+      "epoch": 1.3238160761380575,
+      "grad_norm": 0.319131463766098,
+      "learning_rate": 0.000618353442479503,
+      "loss": 1.756,
+      "step": 2887
+    },
+    {
+      "epoch": 1.3242747391354204,
+      "grad_norm": 0.37177905440330505,
+      "learning_rate": 0.0006181128232059407,
+      "loss": 1.2766,
+      "step": 2888
+    },
+    {
+      "epoch": 1.324733402132783,
+      "grad_norm": 0.30548402667045593,
+      "learning_rate": 0.0006178721749584512,
+      "loss": 1.2389,
+      "step": 2889
+    },
+    {
+      "epoch": 1.3251920651301456,
+      "grad_norm": 0.30189475417137146,
+      "learning_rate": 0.0006176314977960673,
+      "loss": 1.2825,
+      "step": 2890
+    },
+    {
+      "epoch": 1.3256507281275083,
+      "grad_norm": 0.11458157002925873,
+      "learning_rate": 0.0006173907917778288,
+      "loss": 0.5948,
+      "step": 2891
+    },
+    {
+      "epoch": 1.326109391124871,
+      "grad_norm": 0.2670215666294098,
+      "learning_rate": 0.0006171500569627827,
+      "loss": 0.8963,
+      "step": 2892
+    },
+    {
+      "epoch": 1.3265680541222338,
+      "grad_norm": 0.23610422015190125,
+      "learning_rate": 0.0006169092934099828,
+      "loss": 1.4451,
+      "step": 2893
+    },
+    {
+      "epoch": 1.3270267171195964,
+      "grad_norm": 0.3228674530982971,
+      "learning_rate": 0.0006166685011784904,
+      "loss": 1.6901,
+      "step": 2894
+    },
+    {
+      "epoch": 1.327485380116959,
+      "grad_norm": 0.323810875415802,
+      "learning_rate": 0.0006164276803273734,
+      "loss": 1.6255,
+      "step": 2895
+    },
+    {
+      "epoch": 1.3279440431143217,
+      "grad_norm": 0.2585465908050537,
+      "learning_rate": 0.000616186830915707,
+      "loss": 0.8431,
+      "step": 2896
+    },
+    {
+      "epoch": 1.3284027061116843,
+      "grad_norm": 0.2836966812610626,
+      "learning_rate": 0.000615945953002573,
+      "loss": 1.4153,
+      "step": 2897
+    },
+    {
+      "epoch": 1.3288613691090472,
+      "grad_norm": 0.3975668251514435,
+      "learning_rate": 0.0006157050466470608,
+      "loss": 1.78,
+      "step": 2898
+    },
+    {
+      "epoch": 1.3293200321064098,
+      "grad_norm": 0.3772067129611969,
+      "learning_rate": 0.0006154641119082666,
+      "loss": 2.2615,
+      "step": 2899
+    },
+    {
+      "epoch": 1.3297786951037724,
+      "grad_norm": 0.39629077911376953,
+      "learning_rate": 0.0006152231488452931,
+      "loss": 1.9524,
+      "step": 2900
+    },
+    {
+      "epoch": 1.3302373581011353,
+      "grad_norm": 0.26282697916030884,
+      "learning_rate": 0.0006149821575172502,
+      "loss": 0.9766,
+      "step": 2901
+    },
+    {
+      "epoch": 1.330696021098498,
+      "grad_norm": 0.22965925931930542,
+      "learning_rate": 0.0006147411379832553,
+      "loss": 1.179,
+      "step": 2902
+    },
+    {
+      "epoch": 1.3311546840958606,
+      "grad_norm": 0.2938666045665741,
+      "learning_rate": 0.0006145000903024317,
+      "loss": 1.5908,
+      "step": 2903
+    },
+    {
+      "epoch": 1.3316133470932232,
+      "grad_norm": 0.35874783992767334,
+      "learning_rate": 0.0006142590145339106,
+      "loss": 1.3643,
+      "step": 2904
+    },
+    {
+      "epoch": 1.3320720100905858,
+      "grad_norm": 0.3151531219482422,
+      "learning_rate": 0.0006140179107368291,
+      "loss": 1.5058,
+      "step": 2905
+    },
+    {
+      "epoch": 1.3325306730879487,
+      "grad_norm": 0.3742069602012634,
+      "learning_rate": 0.0006137767789703322,
+      "loss": 1.6921,
+      "step": 2906
+    },
+    {
+      "epoch": 1.3329893360853113,
+      "grad_norm": 0.19927988946437836,
+      "learning_rate": 0.0006135356192935709,
+      "loss": 1.2602,
+      "step": 2907
+    },
+    {
+      "epoch": 1.333447999082674,
+      "grad_norm": 0.31676074862480164,
+      "learning_rate": 0.0006132944317657035,
+      "loss": 1.7969,
+      "step": 2908
+    },
+    {
+      "epoch": 1.3339066620800368,
+      "grad_norm": 0.2802976965904236,
+      "learning_rate": 0.0006130532164458952,
+      "loss": 1.3497,
+      "step": 2909
+    },
+    {
+      "epoch": 1.3343653250773992,
+      "grad_norm": 0.3053966760635376,
+      "learning_rate": 0.0006128119733933179,
+      "loss": 1.7183,
+      "step": 2910
+    },
+    {
+      "epoch": 1.334823988074762,
+      "grad_norm": 0.3425951302051544,
+      "learning_rate": 0.0006125707026671498,
+      "loss": 1.333,
+      "step": 2911
+    },
+    {
+      "epoch": 1.3352826510721247,
+      "grad_norm": 0.33375629782676697,
+      "learning_rate": 0.0006123294043265768,
+      "loss": 2.2371,
+      "step": 2912
+    },
+    {
+      "epoch": 1.3357413140694874,
+      "grad_norm": 0.358460396528244,
+      "learning_rate": 0.000612088078430791,
+      "loss": 1.4041,
+      "step": 2913
+    },
+    {
+      "epoch": 1.3361999770668502,
+      "grad_norm": 0.30156955122947693,
+      "learning_rate": 0.0006118467250389914,
+      "loss": 2.1921,
+      "step": 2914
+    },
+    {
+      "epoch": 1.3366586400642129,
+      "grad_norm": 0.32735612988471985,
+      "learning_rate": 0.0006116053442103836,
+      "loss": 1.3602,
+      "step": 2915
+    },
+    {
+      "epoch": 1.3371173030615755,
+      "grad_norm": 0.2227545529603958,
+      "learning_rate": 0.0006113639360041803,
+      "loss": 0.9218,
+      "step": 2916
+    },
+    {
+      "epoch": 1.3375759660589381,
+      "grad_norm": 0.28981199860572815,
+      "learning_rate": 0.0006111225004796004,
+      "loss": 1.1524,
+      "step": 2917
+    },
+    {
+      "epoch": 1.3380346290563008,
+      "grad_norm": 0.2632843554019928,
+      "learning_rate": 0.0006108810376958699,
+      "loss": 1.2719,
+      "step": 2918
+    },
+    {
+      "epoch": 1.3384932920536636,
+      "grad_norm": 0.3473341166973114,
+      "learning_rate": 0.0006106395477122213,
+      "loss": 1.3489,
+      "step": 2919
+    },
+    {
+      "epoch": 1.3389519550510263,
+      "grad_norm": 0.10266976803541183,
+      "learning_rate": 0.000610398030587894,
+      "loss": 0.5534,
+      "step": 2920
+    },
+    {
+      "epoch": 1.339410618048389,
+      "grad_norm": 0.29631996154785156,
+      "learning_rate": 0.0006101564863821338,
+      "loss": 1.2847,
+      "step": 2921
+    },
+    {
+      "epoch": 1.3398692810457518,
+      "grad_norm": 0.2845175266265869,
+      "learning_rate": 0.0006099149151541931,
+      "loss": 1.7554,
+      "step": 2922
+    },
+    {
+      "epoch": 1.3403279440431144,
+      "grad_norm": 0.2774314284324646,
+      "learning_rate": 0.0006096733169633314,
+      "loss": 1.5494,
+      "step": 2923
+    },
+    {
+      "epoch": 1.340786607040477,
+      "grad_norm": 0.31548067927360535,
+      "learning_rate": 0.0006094316918688141,
+      "loss": 1.1424,
+      "step": 2924
+    },
+    {
+      "epoch": 1.3412452700378397,
+      "grad_norm": 0.357732892036438,
+      "learning_rate": 0.0006091900399299138,
+      "loss": 1.3782,
+      "step": 2925
+    },
+    {
+      "epoch": 1.3417039330352023,
+      "grad_norm": 0.2636417746543884,
+      "learning_rate": 0.0006089483612059093,
+      "loss": 1.6639,
+      "step": 2926
+    },
+    {
+      "epoch": 1.3421625960325652,
+      "grad_norm": 0.32517683506011963,
+      "learning_rate": 0.0006087066557560865,
+      "loss": 1.743,
+      "step": 2927
+    },
+    {
+      "epoch": 1.3426212590299278,
+      "grad_norm": 0.3475479185581207,
+      "learning_rate": 0.0006084649236397369,
+      "loss": 1.6309,
+      "step": 2928
+    },
+    {
+      "epoch": 1.3430799220272904,
+      "grad_norm": 0.3126490116119385,
+      "learning_rate": 0.0006082231649161595,
+      "loss": 1.4764,
+      "step": 2929
+    },
+    {
+      "epoch": 1.343538585024653,
+      "grad_norm": 0.24782423675060272,
+      "learning_rate": 0.0006079813796446596,
+      "loss": 0.7667,
+      "step": 2930
+    },
+    {
+      "epoch": 1.3439972480220157,
+      "grad_norm": 0.28448745608329773,
+      "learning_rate": 0.0006077395678845486,
+      "loss": 1.4008,
+      "step": 2931
+    },
+    {
+      "epoch": 1.3444559110193786,
+      "grad_norm": 0.1474987417459488,
+      "learning_rate": 0.0006074977296951448,
+      "loss": 0.5947,
+      "step": 2932
+    },
+    {
+      "epoch": 1.3449145740167412,
+      "grad_norm": 0.20980101823806763,
+      "learning_rate": 0.0006072558651357727,
+      "loss": 1.2209,
+      "step": 2933
+    },
+    {
+      "epoch": 1.3453732370141038,
+      "grad_norm": 0.377865731716156,
+      "learning_rate": 0.0006070139742657635,
+      "loss": 2.0645,
+      "step": 2934
+    },
+    {
+      "epoch": 1.3458319000114667,
+      "grad_norm": 0.47778239846229553,
+      "learning_rate": 0.0006067720571444551,
+      "loss": 1.5841,
+      "step": 2935
+    },
+    {
+      "epoch": 1.3462905630088293,
+      "grad_norm": 0.3798399865627289,
+      "learning_rate": 0.000606530113831191,
+      "loss": 1.7783,
+      "step": 2936
+    },
+    {
+      "epoch": 1.346749226006192,
+      "grad_norm": 0.21647045016288757,
+      "learning_rate": 0.0006062881443853218,
+      "loss": 0.4051,
+      "step": 2937
+    },
+    {
+      "epoch": 1.3472078890035546,
+      "grad_norm": 0.3337489664554596,
+      "learning_rate": 0.0006060461488662044,
+      "loss": 2.0281,
+      "step": 2938
+    },
+    {
+      "epoch": 1.3476665520009172,
+      "grad_norm": 0.3525756895542145,
+      "learning_rate": 0.0006058041273332021,
+      "loss": 1.5044,
+      "step": 2939
+    },
+    {
+      "epoch": 1.34812521499828,
+      "grad_norm": 0.24019671976566315,
+      "learning_rate": 0.0006055620798456845,
+      "loss": 1.3129,
+      "step": 2940
+    },
+    {
+      "epoch": 1.3485838779956427,
+      "grad_norm": 0.37379053235054016,
+      "learning_rate": 0.0006053200064630276,
+      "loss": 2.0564,
+      "step": 2941
+    },
+    {
+      "epoch": 1.3490425409930054,
+      "grad_norm": 0.2772229015827179,
+      "learning_rate": 0.0006050779072446137,
+      "loss": 1.0508,
+      "step": 2942
+    },
+    {
+      "epoch": 1.3495012039903682,
+      "grad_norm": 0.15392114222049713,
+      "learning_rate": 0.0006048357822498315,
+      "loss": 1.0146,
+      "step": 2943
+    },
+    {
+      "epoch": 1.3499598669877306,
+      "grad_norm": 0.3035467267036438,
+      "learning_rate": 0.000604593631538076,
+      "loss": 1.5189,
+      "step": 2944
+    },
+    {
+      "epoch": 1.3504185299850935,
+      "grad_norm": 0.3849334120750427,
+      "learning_rate": 0.0006043514551687484,
+      "loss": 1.7883,
+      "step": 2945
+    },
+    {
+      "epoch": 1.3508771929824561,
+      "grad_norm": 0.36944082379341125,
+      "learning_rate": 0.0006041092532012566,
+      "loss": 1.8156,
+      "step": 2946
+    },
+    {
+      "epoch": 1.3513358559798188,
+      "grad_norm": 0.5630089044570923,
+      "learning_rate": 0.0006038670256950144,
+      "loss": 1.1769,
+      "step": 2947
+    },
+    {
+      "epoch": 1.3517945189771816,
+      "grad_norm": 0.22693516314029694,
+      "learning_rate": 0.0006036247727094418,
+      "loss": 1.3827,
+      "step": 2948
+    },
+    {
+      "epoch": 1.3522531819745442,
+      "grad_norm": 0.2991299033164978,
+      "learning_rate": 0.0006033824943039651,
+      "loss": 0.9545,
+      "step": 2949
+    },
+    {
+      "epoch": 1.3527118449719069,
+      "grad_norm": 0.21848253905773163,
+      "learning_rate": 0.0006031401905380173,
+      "loss": 0.9471,
+      "step": 2950
+    },
+    {
+      "epoch": 1.3531705079692695,
+      "grad_norm": 0.2379998415708542,
+      "learning_rate": 0.000602897861471037,
+      "loss": 1.1672,
+      "step": 2951
+    },
+    {
+      "epoch": 1.3536291709666322,
+      "grad_norm": 0.28074660897254944,
+      "learning_rate": 0.0006026555071624694,
+      "loss": 1.7654,
+      "step": 2952
+    },
+    {
+      "epoch": 1.354087833963995,
+      "grad_norm": 0.4257287383079529,
+      "learning_rate": 0.0006024131276717657,
+      "loss": 2.002,
+      "step": 2953
+    },
+    {
+      "epoch": 1.3545464969613576,
+      "grad_norm": 0.9441617131233215,
+      "learning_rate": 0.0006021707230583834,
+      "loss": 1.7286,
+      "step": 2954
+    },
+    {
+      "epoch": 1.3550051599587203,
+      "grad_norm": 0.24226488173007965,
+      "learning_rate": 0.0006019282933817859,
+      "loss": 1.2914,
+      "step": 2955
+    },
+    {
+      "epoch": 1.3554638229560831,
+      "grad_norm": 0.3882712125778198,
+      "learning_rate": 0.0006016858387014432,
+      "loss": 2.168,
+      "step": 2956
+    },
+    {
+      "epoch": 1.3559224859534458,
+      "grad_norm": 0.2512166202068329,
+      "learning_rate": 0.0006014433590768311,
+      "loss": 1.1456,
+      "step": 2957
+    },
+    {
+      "epoch": 1.3563811489508084,
+      "grad_norm": 0.17546239495277405,
+      "learning_rate": 0.0006012008545674316,
+      "loss": 1.2354,
+      "step": 2958
+    },
+    {
+      "epoch": 1.356839811948171,
+      "grad_norm": 0.35493841767311096,
+      "learning_rate": 0.0006009583252327326,
+      "loss": 1.7339,
+      "step": 2959
+    },
+    {
+      "epoch": 1.3572984749455337,
+      "grad_norm": 0.381168931722641,
+      "learning_rate": 0.0006007157711322286,
+      "loss": 2.2961,
+      "step": 2960
+    },
+    {
+      "epoch": 1.3577571379428965,
+      "grad_norm": 0.229428231716156,
+      "learning_rate": 0.0006004731923254197,
+      "loss": 0.6636,
+      "step": 2961
+    },
+    {
+      "epoch": 1.3582158009402592,
+      "grad_norm": 0.3353422284126282,
+      "learning_rate": 0.0006002305888718123,
+      "loss": 1.9835,
+      "step": 2962
+    },
+    {
+      "epoch": 1.3586744639376218,
+      "grad_norm": 0.32611238956451416,
+      "learning_rate": 0.0005999879608309189,
+      "loss": 1.601,
+      "step": 2963
+    },
+    {
+      "epoch": 1.3591331269349844,
+      "grad_norm": 0.29757001996040344,
+      "learning_rate": 0.0005997453082622575,
+      "loss": 1.6636,
+      "step": 2964
+    },
+    {
+      "epoch": 1.359591789932347,
+      "grad_norm": 0.2990017235279083,
+      "learning_rate": 0.0005995026312253527,
+      "loss": 1.4233,
+      "step": 2965
+    },
+    {
+      "epoch": 1.36005045292971,
+      "grad_norm": 0.3065161108970642,
+      "learning_rate": 0.0005992599297797354,
+      "loss": 1.7319,
+      "step": 2966
+    },
+    {
+      "epoch": 1.3605091159270726,
+      "grad_norm": 0.2563233971595764,
+      "learning_rate": 0.0005990172039849413,
+      "loss": 0.6647,
+      "step": 2967
+    },
+    {
+      "epoch": 1.3609677789244352,
+      "grad_norm": 0.12097487598657608,
+      "learning_rate": 0.0005987744539005134,
+      "loss": 0.9222,
+      "step": 2968
+    },
+    {
+      "epoch": 1.361426441921798,
+      "grad_norm": 0.2720169723033905,
+      "learning_rate": 0.0005985316795859994,
+      "loss": 0.8245,
+      "step": 2969
+    },
+    {
+      "epoch": 1.3618851049191607,
+      "grad_norm": 0.2813059985637665,
+      "learning_rate": 0.0005982888811009541,
+      "loss": 1.371,
+      "step": 2970
+    },
+    {
+      "epoch": 1.3623437679165233,
+      "grad_norm": 0.32221513986587524,
+      "learning_rate": 0.0005980460585049377,
+      "loss": 1.9143,
+      "step": 2971
+    },
+    {
+      "epoch": 1.362802430913886,
+      "grad_norm": 0.3392675817012787,
+      "learning_rate": 0.0005978032118575162,
+      "loss": 1.5422,
+      "step": 2972
+    },
+    {
+      "epoch": 1.3632610939112486,
+      "grad_norm": 0.26565980911254883,
+      "learning_rate": 0.0005975603412182615,
+      "loss": 1.222,
+      "step": 2973
+    },
+    {
+      "epoch": 1.3637197569086115,
+      "grad_norm": 0.32214289903640747,
+      "learning_rate": 0.0005973174466467516,
+      "loss": 2.3055,
+      "step": 2974
+    },
+    {
+      "epoch": 1.364178419905974,
+      "grad_norm": 0.3520325720310211,
+      "learning_rate": 0.0005970745282025705,
+      "loss": 1.8467,
+      "step": 2975
+    },
+    {
+      "epoch": 1.3646370829033367,
+      "grad_norm": 0.42569711804389954,
+      "learning_rate": 0.0005968315859453075,
+      "loss": 2.2639,
+      "step": 2976
+    },
+    {
+      "epoch": 1.3650957459006996,
+      "grad_norm": 0.32188594341278076,
+      "learning_rate": 0.0005965886199345584,
+      "loss": 1.6467,
+      "step": 2977
+    },
+    {
+      "epoch": 1.365554408898062,
+      "grad_norm": 0.3434820771217346,
+      "learning_rate": 0.0005963456302299245,
+      "loss": 1.8241,
+      "step": 2978
+    },
+    {
+      "epoch": 1.3660130718954249,
+      "grad_norm": 0.4579332172870636,
+      "learning_rate": 0.0005961026168910126,
+      "loss": 1.7689,
+      "step": 2979
+    },
+    {
+      "epoch": 1.3664717348927875,
+      "grad_norm": 0.32917508482933044,
+      "learning_rate": 0.000595859579977436,
+      "loss": 1.7894,
+      "step": 2980
+    },
+    {
+      "epoch": 1.3669303978901501,
+      "grad_norm": 0.2983636260032654,
+      "learning_rate": 0.0005956165195488131,
+      "loss": 1.6581,
+      "step": 2981
+    },
+    {
+      "epoch": 1.367389060887513,
+      "grad_norm": 0.23157159984111786,
+      "learning_rate": 0.0005953734356647686,
+      "loss": 0.804,
+      "step": 2982
+    },
+    {
+      "epoch": 1.3678477238848756,
+      "grad_norm": 0.2740468680858612,
+      "learning_rate": 0.0005951303283849329,
+      "loss": 1.2906,
+      "step": 2983
+    },
+    {
+      "epoch": 1.3683063868822383,
+      "grad_norm": 0.15333612263202667,
+      "learning_rate": 0.0005948871977689414,
+      "loss": 0.7368,
+      "step": 2984
+    },
+    {
+      "epoch": 1.368765049879601,
+      "grad_norm": 0.3173515796661377,
+      "learning_rate": 0.0005946440438764362,
+      "loss": 2.1804,
+      "step": 2985
+    },
+    {
+      "epoch": 1.3692237128769635,
+      "grad_norm": 0.3675207495689392,
+      "learning_rate": 0.0005944008667670646,
+      "loss": 1.7012,
+      "step": 2986
+    },
+    {
+      "epoch": 1.3696823758743264,
+      "grad_norm": 0.42129290103912354,
+      "learning_rate": 0.0005941576665004798,
+      "loss": 1.2498,
+      "step": 2987
+    },
+    {
+      "epoch": 1.370141038871689,
+      "grad_norm": 0.24997709691524506,
+      "learning_rate": 0.0005939144431363404,
+      "loss": 1.4775,
+      "step": 2988
+    },
+    {
+      "epoch": 1.3705997018690517,
+      "grad_norm": 0.2527235448360443,
+      "learning_rate": 0.0005936711967343111,
+      "loss": 1.2377,
+      "step": 2989
+    },
+    {
+      "epoch": 1.3710583648664145,
+      "grad_norm": 0.2272566556930542,
+      "learning_rate": 0.0005934279273540616,
+      "loss": 1.5359,
+      "step": 2990
+    },
+    {
+      "epoch": 1.3715170278637772,
+      "grad_norm": 0.30363374948501587,
+      "learning_rate": 0.000593184635055268,
+      "loss": 1.6325,
+      "step": 2991
+    },
+    {
+      "epoch": 1.3719756908611398,
+      "grad_norm": 0.352542519569397,
+      "learning_rate": 0.0005929413198976115,
+      "loss": 1.656,
+      "step": 2992
+    },
+    {
+      "epoch": 1.3724343538585024,
+      "grad_norm": 0.3105866611003876,
+      "learning_rate": 0.0005926979819407791,
+      "loss": 1.9192,
+      "step": 2993
+    },
+    {
+      "epoch": 1.372893016855865,
+      "grad_norm": 0.2809374928474426,
+      "learning_rate": 0.0005924546212444634,
+      "loss": 1.3723,
+      "step": 2994
+    },
+    {
+      "epoch": 1.373351679853228,
+      "grad_norm": 0.32282719016075134,
+      "learning_rate": 0.0005922112378683624,
+      "loss": 1.6761,
+      "step": 2995
+    },
+    {
+      "epoch": 1.3738103428505906,
+      "grad_norm": 0.2368834912776947,
+      "learning_rate": 0.0005919678318721797,
+      "loss": 1.1937,
+      "step": 2996
+    },
+    {
+      "epoch": 1.3742690058479532,
+      "grad_norm": 0.2217353880405426,
+      "learning_rate": 0.000591724403315625,
+      "loss": 1.2874,
+      "step": 2997
+    },
+    {
+      "epoch": 1.3747276688453158,
+      "grad_norm": 0.2690431773662567,
+      "learning_rate": 0.0005914809522584127,
+      "loss": 1.4406,
+      "step": 2998
+    },
+    {
+      "epoch": 1.3751863318426785,
+      "grad_norm": 0.20234516263008118,
+      "learning_rate": 0.0005912374787602632,
+      "loss": 1.3164,
+      "step": 2999
+    },
+    {
+      "epoch": 1.3756449948400413,
+      "grad_norm": 0.265828937292099,
+      "learning_rate": 0.0005909939828809024,
+      "loss": 1.2651,
+      "step": 3000
+    },
+    {
+      "epoch": 1.376103657837404,
+      "grad_norm": 0.28443530201911926,
+      "learning_rate": 0.0005907504646800613,
+      "loss": 1.1567,
+      "step": 3001
+    },
+    {
+      "epoch": 1.3765623208347666,
+      "grad_norm": 0.2254990190267563,
+      "learning_rate": 0.0005905069242174769,
+      "loss": 1.4349,
+      "step": 3002
+    },
+    {
+      "epoch": 1.3770209838321295,
+      "grad_norm": 0.2833521366119385,
+      "learning_rate": 0.0005902633615528916,
+      "loss": 1.8909,
+      "step": 3003
+    },
+    {
+      "epoch": 1.377479646829492,
+      "grad_norm": 0.3044750392436981,
+      "learning_rate": 0.0005900197767460527,
+      "loss": 1.5297,
+      "step": 3004
+    },
+    {
+      "epoch": 1.3779383098268547,
+      "grad_norm": 0.29396411776542664,
+      "learning_rate": 0.0005897761698567135,
+      "loss": 1.2523,
+      "step": 3005
+    },
+    {
+      "epoch": 1.3783969728242174,
+      "grad_norm": 0.24863453209400177,
+      "learning_rate": 0.0005895325409446327,
+      "loss": 1.5853,
+      "step": 3006
+    },
+    {
+      "epoch": 1.37885563582158,
+      "grad_norm": 0.31140193343162537,
+      "learning_rate": 0.0005892888900695738,
+      "loss": 1.259,
+      "step": 3007
+    },
+    {
+      "epoch": 1.3793142988189429,
+      "grad_norm": 0.21979323029518127,
+      "learning_rate": 0.0005890452172913065,
+      "loss": 1.415,
+      "step": 3008
+    },
+    {
+      "epoch": 1.3797729618163055,
+      "grad_norm": 0.3025084435939789,
+      "learning_rate": 0.0005888015226696053,
+      "loss": 0.7619,
+      "step": 3009
+    },
+    {
+      "epoch": 1.3802316248136681,
+      "grad_norm": 0.11269453167915344,
+      "learning_rate": 0.0005885578062642502,
+      "loss": 1.0517,
+      "step": 3010
+    },
+    {
+      "epoch": 1.380690287811031,
+      "grad_norm": 0.2220744490623474,
+      "learning_rate": 0.0005883140681350269,
+      "loss": 0.8101,
+      "step": 3011
+    },
+    {
+      "epoch": 1.3811489508083934,
+      "grad_norm": 0.25374385714530945,
+      "learning_rate": 0.0005880703083417258,
+      "loss": 1.2319,
+      "step": 3012
+    },
+    {
+      "epoch": 1.3816076138057563,
+      "grad_norm": 0.26675471663475037,
+      "learning_rate": 0.0005878265269441432,
+      "loss": 1.4619,
+      "step": 3013
+    },
+    {
+      "epoch": 1.3820662768031189,
+      "grad_norm": 0.09885770082473755,
+      "learning_rate": 0.00058758272400208,
+      "loss": 0.6359,
+      "step": 3014
+    },
+    {
+      "epoch": 1.3825249398004815,
+      "grad_norm": 0.16511155664920807,
+      "learning_rate": 0.0005873388995753432,
+      "loss": 1.0938,
+      "step": 3015
+    },
+    {
+      "epoch": 1.3829836027978444,
+      "grad_norm": 0.3333290219306946,
+      "learning_rate": 0.0005870950537237446,
+      "loss": 1.7571,
+      "step": 3016
+    },
+    {
+      "epoch": 1.383442265795207,
+      "grad_norm": 0.26791778206825256,
+      "learning_rate": 0.0005868511865071013,
+      "loss": 0.9278,
+      "step": 3017
+    },
+    {
+      "epoch": 1.3839009287925697,
+      "grad_norm": 0.2120082974433899,
+      "learning_rate": 0.0005866072979852358,
+      "loss": 1.2115,
+      "step": 3018
+    },
+    {
+      "epoch": 1.3843595917899323,
+      "grad_norm": 0.41435691714286804,
+      "learning_rate": 0.0005863633882179758,
+      "loss": 2.1711,
+      "step": 3019
+    },
+    {
+      "epoch": 1.384818254787295,
+      "grad_norm": 0.2728036046028137,
+      "learning_rate": 0.0005861194572651537,
+      "loss": 1.5656,
+      "step": 3020
+    },
+    {
+      "epoch": 1.3852769177846578,
+      "grad_norm": 0.3878629803657532,
+      "learning_rate": 0.0005858755051866078,
+      "loss": 1.8513,
+      "step": 3021
+    },
+    {
+      "epoch": 1.3857355807820204,
+      "grad_norm": 0.3821435570716858,
+      "learning_rate": 0.0005856315320421812,
+      "loss": 1.7039,
+      "step": 3022
+    },
+    {
+      "epoch": 1.386194243779383,
+      "grad_norm": 0.28816846013069153,
+      "learning_rate": 0.0005853875378917225,
+      "loss": 1.8958,
+      "step": 3023
+    },
+    {
+      "epoch": 1.386652906776746,
+      "grad_norm": 0.24760308861732483,
+      "learning_rate": 0.0005851435227950851,
+      "loss": 0.9471,
+      "step": 3024
+    },
+    {
+      "epoch": 1.3871115697741085,
+      "grad_norm": 0.20163798332214355,
+      "learning_rate": 0.0005848994868121277,
+      "loss": 1.524,
+      "step": 3025
+    },
+    {
+      "epoch": 1.3875702327714712,
+      "grad_norm": 0.6160727143287659,
+      "learning_rate": 0.0005846554300027138,
+      "loss": 1.6717,
+      "step": 3026
+    },
+    {
+      "epoch": 1.3880288957688338,
+      "grad_norm": 0.2782137095928192,
+      "learning_rate": 0.0005844113524267128,
+      "loss": 0.8874,
+      "step": 3027
+    },
+    {
+      "epoch": 1.3884875587661965,
+      "grad_norm": 0.2697048783302307,
+      "learning_rate": 0.0005841672541439982,
+      "loss": 1.3064,
+      "step": 3028
+    },
+    {
+      "epoch": 1.3889462217635593,
+      "grad_norm": 0.2576363980770111,
+      "learning_rate": 0.0005839231352144493,
+      "loss": 1.8171,
+      "step": 3029
+    },
+    {
+      "epoch": 1.389404884760922,
+      "grad_norm": 0.34383904933929443,
+      "learning_rate": 0.0005836789956979503,
+      "loss": 1.5734,
+      "step": 3030
+    },
+    {
+      "epoch": 1.3898635477582846,
+      "grad_norm": 0.2631419003009796,
+      "learning_rate": 0.0005834348356543903,
+      "loss": 1.2598,
+      "step": 3031
+    },
+    {
+      "epoch": 1.3903222107556472,
+      "grad_norm": 0.3006002604961395,
+      "learning_rate": 0.0005831906551436633,
+      "loss": 1.4291,
+      "step": 3032
+    },
+    {
+      "epoch": 1.3907808737530098,
+      "grad_norm": 0.13184750080108643,
+      "learning_rate": 0.0005829464542256691,
+      "loss": 0.9445,
+      "step": 3033
+    },
+    {
+      "epoch": 1.3912395367503727,
+      "grad_norm": 0.38121703267097473,
+      "learning_rate": 0.0005827022329603114,
+      "loss": 1.3728,
+      "step": 3034
+    },
+    {
+      "epoch": 1.3916981997477353,
+      "grad_norm": 0.3366886079311371,
+      "learning_rate": 0.0005824579914074996,
+      "loss": 2.0142,
+      "step": 3035
+    },
+    {
+      "epoch": 1.392156862745098,
+      "grad_norm": 0.2520803213119507,
+      "learning_rate": 0.0005822137296271481,
+      "loss": 1.1912,
+      "step": 3036
+    },
+    {
+      "epoch": 1.3926155257424608,
+      "grad_norm": 0.2898450493812561,
+      "learning_rate": 0.0005819694476791757,
+      "loss": 1.5584,
+      "step": 3037
+    },
+    {
+      "epoch": 1.3930741887398235,
+      "grad_norm": 0.33184388279914856,
+      "learning_rate": 0.000581725145623507,
+      "loss": 1.6064,
+      "step": 3038
+    },
+    {
+      "epoch": 1.393532851737186,
+      "grad_norm": 0.40687495470046997,
+      "learning_rate": 0.0005814808235200708,
+      "loss": 2.0935,
+      "step": 3039
+    },
+    {
+      "epoch": 1.3939915147345487,
+      "grad_norm": 0.30613189935684204,
+      "learning_rate": 0.000581236481428801,
+      "loss": 1.0827,
+      "step": 3040
+    },
+    {
+      "epoch": 1.3944501777319114,
+      "grad_norm": 0.350396066904068,
+      "learning_rate": 0.0005809921194096365,
+      "loss": 1.9321,
+      "step": 3041
+    },
+    {
+      "epoch": 1.3949088407292742,
+      "grad_norm": 0.2401767075061798,
+      "learning_rate": 0.0005807477375225212,
+      "loss": 1.1565,
+      "step": 3042
+    },
+    {
+      "epoch": 1.3953675037266369,
+      "grad_norm": 0.29170459508895874,
+      "learning_rate": 0.0005805033358274037,
+      "loss": 1.3792,
+      "step": 3043
+    },
+    {
+      "epoch": 1.3958261667239995,
+      "grad_norm": 0.3244495689868927,
+      "learning_rate": 0.0005802589143842374,
+      "loss": 1.5734,
+      "step": 3044
+    },
+    {
+      "epoch": 1.3962848297213624,
+      "grad_norm": 0.35495296120643616,
+      "learning_rate": 0.0005800144732529808,
+      "loss": 1.4654,
+      "step": 3045
+    },
+    {
+      "epoch": 1.396743492718725,
+      "grad_norm": 0.31566861271858215,
+      "learning_rate": 0.0005797700124935968,
+      "loss": 2.1155,
+      "step": 3046
+    },
+    {
+      "epoch": 1.3972021557160876,
+      "grad_norm": 0.35331422090530396,
+      "learning_rate": 0.0005795255321660536,
+      "loss": 1.4407,
+      "step": 3047
+    },
+    {
+      "epoch": 1.3976608187134503,
+      "grad_norm": 0.46943673491477966,
+      "learning_rate": 0.000579281032330324,
+      "loss": 0.9726,
+      "step": 3048
+    },
+    {
+      "epoch": 1.398119481710813,
+      "grad_norm": 0.3449120819568634,
+      "learning_rate": 0.0005790365130463857,
+      "loss": 1.9149,
+      "step": 3049
+    },
+    {
+      "epoch": 1.3985781447081758,
+      "grad_norm": 0.45382195711135864,
+      "learning_rate": 0.0005787919743742207,
+      "loss": 1.5399,
+      "step": 3050
+    },
+    {
+      "epoch": 1.3990368077055384,
+      "grad_norm": 0.2743489742279053,
+      "learning_rate": 0.0005785474163738163,
+      "loss": 1.3011,
+      "step": 3051
+    },
+    {
+      "epoch": 1.399495470702901,
+      "grad_norm": 0.24921023845672607,
+      "learning_rate": 0.0005783028391051642,
+      "loss": 1.4218,
+      "step": 3052
+    },
+    {
+      "epoch": 1.3999541337002637,
+      "grad_norm": 0.25878721475601196,
+      "learning_rate": 0.0005780582426282611,
+      "loss": 1.3025,
+      "step": 3053
+    },
+    {
+      "epoch": 1.4004127966976263,
+      "grad_norm": 0.24198046326637268,
+      "learning_rate": 0.0005778136270031083,
+      "loss": 0.8168,
+      "step": 3054
+    },
+    {
+      "epoch": 1.4008714596949892,
+      "grad_norm": 0.259734183549881,
+      "learning_rate": 0.0005775689922897119,
+      "loss": 0.9978,
+      "step": 3055
+    },
+    {
+      "epoch": 1.4013301226923518,
+      "grad_norm": 0.2518146336078644,
+      "learning_rate": 0.0005773243385480822,
+      "loss": 1.4738,
+      "step": 3056
+    },
+    {
+      "epoch": 1.4017887856897144,
+      "grad_norm": 0.16398538649082184,
+      "learning_rate": 0.0005770796658382347,
+      "loss": 1.2174,
+      "step": 3057
+    },
+    {
+      "epoch": 1.4022474486870773,
+      "grad_norm": 0.2645060122013092,
+      "learning_rate": 0.0005768349742201895,
+      "loss": 1.6903,
+      "step": 3058
+    },
+    {
+      "epoch": 1.40270611168444,
+      "grad_norm": 0.30692818760871887,
+      "learning_rate": 0.0005765902637539709,
+      "loss": 1.3245,
+      "step": 3059
+    },
+    {
+      "epoch": 1.4031647746818026,
+      "grad_norm": 0.19634264707565308,
+      "learning_rate": 0.0005763455344996086,
+      "loss": 1.1637,
+      "step": 3060
+    },
+    {
+      "epoch": 1.4036234376791652,
+      "grad_norm": 0.2435206174850464,
+      "learning_rate": 0.0005761007865171361,
+      "loss": 1.6165,
+      "step": 3061
+    },
+    {
+      "epoch": 1.4040821006765278,
+      "grad_norm": 0.17027845978736877,
+      "learning_rate": 0.0005758560198665921,
+      "loss": 0.8738,
+      "step": 3062
+    },
+    {
+      "epoch": 1.4045407636738907,
+      "grad_norm": 0.23753295838832855,
+      "learning_rate": 0.0005756112346080192,
+      "loss": 1.6266,
+      "step": 3063
+    },
+    {
+      "epoch": 1.4049994266712533,
+      "grad_norm": 0.26798611879348755,
+      "learning_rate": 0.0005753664308014655,
+      "loss": 1.7124,
+      "step": 3064
+    },
+    {
+      "epoch": 1.405458089668616,
+      "grad_norm": 0.4665045738220215,
+      "learning_rate": 0.0005751216085069828,
+      "loss": 2.0098,
+      "step": 3065
+    },
+    {
+      "epoch": 1.4059167526659786,
+      "grad_norm": 0.279823362827301,
+      "learning_rate": 0.0005748767677846279,
+      "loss": 1.3131,
+      "step": 3066
+    },
+    {
+      "epoch": 1.4063754156633412,
+      "grad_norm": 0.27207767963409424,
+      "learning_rate": 0.000574631908694462,
+      "loss": 1.1439,
+      "step": 3067
+    },
+    {
+      "epoch": 1.406834078660704,
+      "grad_norm": 0.30169880390167236,
+      "learning_rate": 0.0005743870312965506,
+      "loss": 1.2415,
+      "step": 3068
+    },
+    {
+      "epoch": 1.4072927416580667,
+      "grad_norm": 0.20790642499923706,
+      "learning_rate": 0.0005741421356509645,
+      "loss": 1.0963,
+      "step": 3069
+    },
+    {
+      "epoch": 1.4077514046554294,
+      "grad_norm": 0.3159151077270508,
+      "learning_rate": 0.0005738972218177774,
+      "loss": 1.3106,
+      "step": 3070
+    },
+    {
+      "epoch": 1.4082100676527922,
+      "grad_norm": 0.2372143566608429,
+      "learning_rate": 0.0005736522898570694,
+      "loss": 1.2921,
+      "step": 3071
+    },
+    {
+      "epoch": 1.4086687306501549,
+      "grad_norm": 0.37024182081222534,
+      "learning_rate": 0.0005734073398289234,
+      "loss": 2.0061,
+      "step": 3072
+    },
+    {
+      "epoch": 1.4091273936475175,
+      "grad_norm": 0.38484346866607666,
+      "learning_rate": 0.0005731623717934275,
+      "loss": 1.5103,
+      "step": 3073
+    },
+    {
+      "epoch": 1.4095860566448801,
+      "grad_norm": 0.33676987886428833,
+      "learning_rate": 0.0005729173858106746,
+      "loss": 1.8154,
+      "step": 3074
+    },
+    {
+      "epoch": 1.4100447196422428,
+      "grad_norm": 0.35367336869239807,
+      "learning_rate": 0.0005726723819407611,
+      "loss": 1.9507,
+      "step": 3075
+    },
+    {
+      "epoch": 1.4105033826396056,
+      "grad_norm": 0.3594896197319031,
+      "learning_rate": 0.0005724273602437882,
+      "loss": 1.793,
+      "step": 3076
+    },
+    {
+      "epoch": 1.4109620456369683,
+      "grad_norm": 0.3166644275188446,
+      "learning_rate": 0.0005721823207798616,
+      "loss": 1.2424,
+      "step": 3077
+    },
+    {
+      "epoch": 1.411420708634331,
+      "grad_norm": 0.30905553698539734,
+      "learning_rate": 0.0005719372636090914,
+      "loss": 1.4936,
+      "step": 3078
+    },
+    {
+      "epoch": 1.4118793716316937,
+      "grad_norm": 0.3719440698623657,
+      "learning_rate": 0.0005716921887915916,
+      "loss": 1.8947,
+      "step": 3079
+    },
+    {
+      "epoch": 1.4123380346290564,
+      "grad_norm": 0.35400110483169556,
+      "learning_rate": 0.000571447096387481,
+      "loss": 1.325,
+      "step": 3080
+    },
+    {
+      "epoch": 1.412796697626419,
+      "grad_norm": 0.2209610939025879,
+      "learning_rate": 0.0005712019864568827,
+      "loss": 1.1089,
+      "step": 3081
+    },
+    {
+      "epoch": 1.4132553606237817,
+      "grad_norm": 0.3398480713367462,
+      "learning_rate": 0.0005709568590599234,
+      "loss": 1.7584,
+      "step": 3082
+    },
+    {
+      "epoch": 1.4137140236211443,
+      "grad_norm": 0.3765453100204468,
+      "learning_rate": 0.0005707117142567351,
+      "loss": 2.1033,
+      "step": 3083
+    },
+    {
+      "epoch": 1.4141726866185071,
+      "grad_norm": 0.2562631070613861,
+      "learning_rate": 0.0005704665521074534,
+      "loss": 1.5005,
+      "step": 3084
+    },
+    {
+      "epoch": 1.4146313496158698,
+      "grad_norm": 0.32559189200401306,
+      "learning_rate": 0.0005702213726722185,
+      "loss": 1.1359,
+      "step": 3085
+    },
+    {
+      "epoch": 1.4150900126132324,
+      "grad_norm": 0.2984718084335327,
+      "learning_rate": 0.0005699761760111745,
+      "loss": 1.4721,
+      "step": 3086
+    },
+    {
+      "epoch": 1.415548675610595,
+      "grad_norm": 0.3191785216331482,
+      "learning_rate": 0.00056973096218447,
+      "loss": 1.773,
+      "step": 3087
+    },
+    {
+      "epoch": 1.4160073386079577,
+      "grad_norm": 0.3590591251850128,
+      "learning_rate": 0.0005694857312522576,
+      "loss": 1.7094,
+      "step": 3088
+    },
+    {
+      "epoch": 1.4164660016053205,
+      "grad_norm": 0.23591914772987366,
+      "learning_rate": 0.0005692404832746944,
+      "loss": 1.0517,
+      "step": 3089
+    },
+    {
+      "epoch": 1.4169246646026832,
+      "grad_norm": 0.28452879190444946,
+      "learning_rate": 0.0005689952183119413,
+      "loss": 1.6881,
+      "step": 3090
+    },
+    {
+      "epoch": 1.4173833276000458,
+      "grad_norm": 0.32875385880470276,
+      "learning_rate": 0.0005687499364241637,
+      "loss": 1.7948,
+      "step": 3091
+    },
+    {
+      "epoch": 1.4178419905974087,
+      "grad_norm": 0.3080821931362152,
+      "learning_rate": 0.0005685046376715311,
+      "loss": 1.4707,
+      "step": 3092
+    },
+    {
+      "epoch": 1.4183006535947713,
+      "grad_norm": 0.3038639724254608,
+      "learning_rate": 0.0005682593221142168,
+      "loss": 1.2142,
+      "step": 3093
+    },
+    {
+      "epoch": 1.418759316592134,
+      "grad_norm": 0.3059542775154114,
+      "learning_rate": 0.0005680139898123985,
+      "loss": 1.2471,
+      "step": 3094
+    },
+    {
+      "epoch": 1.4192179795894966,
+      "grad_norm": 0.2539036273956299,
+      "learning_rate": 0.0005677686408262583,
+      "loss": 1.058,
+      "step": 3095
+    },
+    {
+      "epoch": 1.4196766425868592,
+      "grad_norm": 0.17899896204471588,
+      "learning_rate": 0.0005675232752159818,
+      "loss": 1.3064,
+      "step": 3096
+    },
+    {
+      "epoch": 1.420135305584222,
+      "grad_norm": 0.3230034112930298,
+      "learning_rate": 0.0005672778930417592,
+      "loss": 1.6761,
+      "step": 3097
+    },
+    {
+      "epoch": 1.4205939685815847,
+      "grad_norm": 0.2745686173439026,
+      "learning_rate": 0.0005670324943637842,
+      "loss": 0.7442,
+      "step": 3098
+    },
+    {
+      "epoch": 1.4210526315789473,
+      "grad_norm": 0.2400350570678711,
+      "learning_rate": 0.0005667870792422547,
+      "loss": 0.937,
+      "step": 3099
+    },
+    {
+      "epoch": 1.42151129457631,
+      "grad_norm": 0.3091734051704407,
+      "learning_rate": 0.0005665416477373734,
+      "loss": 1.6765,
+      "step": 3100
+    },
+    {
+      "epoch": 1.4219699575736726,
+      "grad_norm": 0.217657670378685,
+      "learning_rate": 0.0005662961999093461,
+      "loss": 0.8566,
+      "step": 3101
+    },
+    {
+      "epoch": 1.4224286205710355,
+      "grad_norm": 0.18862880766391754,
+      "learning_rate": 0.0005660507358183829,
+      "loss": 1.1898,
+      "step": 3102
+    },
+    {
+      "epoch": 1.422887283568398,
+      "grad_norm": 0.3786904513835907,
+      "learning_rate": 0.000565805255524698,
+      "loss": 1.9863,
+      "step": 3103
+    },
+    {
+      "epoch": 1.4233459465657607,
+      "grad_norm": 0.3076726496219635,
+      "learning_rate": 0.0005655597590885095,
+      "loss": 1.2256,
+      "step": 3104
+    },
+    {
+      "epoch": 1.4238046095631236,
+      "grad_norm": 0.24448339641094208,
+      "learning_rate": 0.0005653142465700392,
+      "loss": 1.7419,
+      "step": 3105
+    },
+    {
+      "epoch": 1.4242632725604862,
+      "grad_norm": 0.28757891058921814,
+      "learning_rate": 0.0005650687180295134,
+      "loss": 1.0779,
+      "step": 3106
+    },
+    {
+      "epoch": 1.4247219355578489,
+      "grad_norm": 0.0921797975897789,
+      "learning_rate": 0.0005648231735271619,
+      "loss": 0.5927,
+      "step": 3107
+    },
+    {
+      "epoch": 1.4251805985552115,
+      "grad_norm": 0.2944778800010681,
+      "learning_rate": 0.0005645776131232186,
+      "loss": 1.6512,
+      "step": 3108
+    },
+    {
+      "epoch": 1.4256392615525741,
+      "grad_norm": 0.24034199118614197,
+      "learning_rate": 0.000564332036877921,
+      "loss": 1.505,
+      "step": 3109
+    },
+    {
+      "epoch": 1.426097924549937,
+      "grad_norm": 0.245818093419075,
+      "learning_rate": 0.000564086444851511,
+      "loss": 1.2884,
+      "step": 3110
+    },
+    {
+      "epoch": 1.4265565875472996,
+      "grad_norm": 0.29974740743637085,
+      "learning_rate": 0.000563840837104234,
+      "loss": 0.8771,
+      "step": 3111
+    },
+    {
+      "epoch": 1.4270152505446623,
+      "grad_norm": 0.2733640968799591,
+      "learning_rate": 0.0005635952136963393,
+      "loss": 1.5035,
+      "step": 3112
+    },
+    {
+      "epoch": 1.4274739135420251,
+      "grad_norm": 0.2323792427778244,
+      "learning_rate": 0.0005633495746880801,
+      "loss": 1.1803,
+      "step": 3113
+    },
+    {
+      "epoch": 1.4279325765393878,
+      "grad_norm": 0.438142329454422,
+      "learning_rate": 0.0005631039201397136,
+      "loss": 2.2593,
+      "step": 3114
+    },
+    {
+      "epoch": 1.4283912395367504,
+      "grad_norm": 0.2616313099861145,
+      "learning_rate": 0.0005628582501115004,
+      "loss": 1.3368,
+      "step": 3115
+    },
+    {
+      "epoch": 1.428849902534113,
+      "grad_norm": 0.2123333215713501,
+      "learning_rate": 0.0005626125646637051,
+      "loss": 1.0031,
+      "step": 3116
+    },
+    {
+      "epoch": 1.4293085655314757,
+      "grad_norm": 0.32749924063682556,
+      "learning_rate": 0.0005623668638565964,
+      "loss": 1.4506,
+      "step": 3117
+    },
+    {
+      "epoch": 1.4297672285288385,
+      "grad_norm": 0.2560507655143738,
+      "learning_rate": 0.0005621211477504463,
+      "loss": 1.7554,
+      "step": 3118
+    },
+    {
+      "epoch": 1.4302258915262012,
+      "grad_norm": 0.3207387924194336,
+      "learning_rate": 0.0005618754164055306,
+      "loss": 1.4721,
+      "step": 3119
+    },
+    {
+      "epoch": 1.4306845545235638,
+      "grad_norm": 0.3025607764720917,
+      "learning_rate": 0.0005616296698821293,
+      "loss": 0.993,
+      "step": 3120
+    },
+    {
+      "epoch": 1.4311432175209264,
+      "grad_norm": 0.22865654528141022,
+      "learning_rate": 0.0005613839082405255,
+      "loss": 1.5714,
+      "step": 3121
+    },
+    {
+      "epoch": 1.431601880518289,
+      "grad_norm": 0.31752631068229675,
+      "learning_rate": 0.0005611381315410065,
+      "loss": 1.4345,
+      "step": 3122
+    },
+    {
+      "epoch": 1.432060543515652,
+      "grad_norm": 0.29477816820144653,
+      "learning_rate": 0.0005608923398438628,
+      "loss": 1.9834,
+      "step": 3123
+    },
+    {
+      "epoch": 1.4325192065130146,
+      "grad_norm": 0.30475521087646484,
+      "learning_rate": 0.0005606465332093892,
+      "loss": 1.5315,
+      "step": 3124
+    },
+    {
+      "epoch": 1.4329778695103772,
+      "grad_norm": 0.3452279567718506,
+      "learning_rate": 0.0005604007116978836,
+      "loss": 1.2822,
+      "step": 3125
+    },
+    {
+      "epoch": 1.43343653250774,
+      "grad_norm": 0.3388407230377197,
+      "learning_rate": 0.0005601548753696481,
+      "loss": 1.6831,
+      "step": 3126
+    },
+    {
+      "epoch": 1.4338951955051027,
+      "grad_norm": 0.3091072142124176,
+      "learning_rate": 0.0005599090242849878,
+      "loss": 0.8168,
+      "step": 3127
+    },
+    {
+      "epoch": 1.4343538585024653,
+      "grad_norm": 0.27518755197525024,
+      "learning_rate": 0.000559663158504212,
+      "loss": 1.9319,
+      "step": 3128
+    },
+    {
+      "epoch": 1.434812521499828,
+      "grad_norm": 0.2800983488559723,
+      "learning_rate": 0.0005594172780876332,
+      "loss": 1.3327,
+      "step": 3129
+    },
+    {
+      "epoch": 1.4352711844971906,
+      "grad_norm": 0.34306713938713074,
+      "learning_rate": 0.0005591713830955674,
+      "loss": 1.2191,
+      "step": 3130
+    },
+    {
+      "epoch": 1.4357298474945535,
+      "grad_norm": 0.19693058729171753,
+      "learning_rate": 0.000558925473588335,
+      "loss": 1.3058,
+      "step": 3131
+    },
+    {
+      "epoch": 1.436188510491916,
+      "grad_norm": 0.42898425459861755,
+      "learning_rate": 0.000558679549626259,
+      "loss": 1.7591,
+      "step": 3132
+    },
+    {
+      "epoch": 1.4366471734892787,
+      "grad_norm": 0.3185971975326538,
+      "learning_rate": 0.0005584336112696663,
+      "loss": 2.0237,
+      "step": 3133
+    },
+    {
+      "epoch": 1.4371058364866414,
+      "grad_norm": 0.18667162954807281,
+      "learning_rate": 0.0005581876585788875,
+      "loss": 0.9062,
+      "step": 3134
+    },
+    {
+      "epoch": 1.437564499484004,
+      "grad_norm": 0.2841646671295166,
+      "learning_rate": 0.0005579416916142564,
+      "loss": 1.2529,
+      "step": 3135
+    },
+    {
+      "epoch": 1.4380231624813669,
+      "grad_norm": 0.27004119753837585,
+      "learning_rate": 0.0005576957104361106,
+      "loss": 1.3765,
+      "step": 3136
+    },
+    {
+      "epoch": 1.4384818254787295,
+      "grad_norm": 0.2476406842470169,
+      "learning_rate": 0.0005574497151047911,
+      "loss": 1.1895,
+      "step": 3137
+    },
+    {
+      "epoch": 1.4389404884760921,
+      "grad_norm": 0.25219327211380005,
+      "learning_rate": 0.0005572037056806421,
+      "loss": 1.1415,
+      "step": 3138
+    },
+    {
+      "epoch": 1.439399151473455,
+      "grad_norm": 0.31320807337760925,
+      "learning_rate": 0.0005569576822240118,
+      "loss": 2.229,
+      "step": 3139
+    },
+    {
+      "epoch": 1.4398578144708176,
+      "grad_norm": 0.3259999752044678,
+      "learning_rate": 0.0005567116447952513,
+      "loss": 1.7386,
+      "step": 3140
+    },
+    {
+      "epoch": 1.4403164774681803,
+      "grad_norm": 0.1600116491317749,
+      "learning_rate": 0.0005564655934547154,
+      "loss": 0.7964,
+      "step": 3141
+    },
+    {
+      "epoch": 1.440775140465543,
+      "grad_norm": 0.2874891757965088,
+      "learning_rate": 0.0005562195282627624,
+      "loss": 1.3448,
+      "step": 3142
+    },
+    {
+      "epoch": 1.4412338034629055,
+      "grad_norm": 0.29961147904396057,
+      "learning_rate": 0.0005559734492797536,
+      "loss": 1.353,
+      "step": 3143
+    },
+    {
+      "epoch": 1.4416924664602684,
+      "grad_norm": 0.4548627734184265,
+      "learning_rate": 0.0005557273565660541,
+      "loss": 1.9115,
+      "step": 3144
+    },
+    {
+      "epoch": 1.442151129457631,
+      "grad_norm": 0.25489258766174316,
+      "learning_rate": 0.0005554812501820322,
+      "loss": 1.1127,
+      "step": 3145
+    },
+    {
+      "epoch": 1.4426097924549937,
+      "grad_norm": 0.3416697680950165,
+      "learning_rate": 0.0005552351301880597,
+      "loss": 2.1523,
+      "step": 3146
+    },
+    {
+      "epoch": 1.4430684554523565,
+      "grad_norm": 0.29285651445388794,
+      "learning_rate": 0.0005549889966445115,
+      "loss": 1.4188,
+      "step": 3147
+    },
+    {
+      "epoch": 1.4435271184497191,
+      "grad_norm": 0.3054216206073761,
+      "learning_rate": 0.0005547428496117659,
+      "loss": 1.0746,
+      "step": 3148
+    },
+    {
+      "epoch": 1.4439857814470818,
+      "grad_norm": 0.2599756717681885,
+      "learning_rate": 0.0005544966891502046,
+      "loss": 1.3993,
+      "step": 3149
+    },
+    {
+      "epoch": 1.4444444444444444,
+      "grad_norm": 0.3153897821903229,
+      "learning_rate": 0.0005542505153202124,
+      "loss": 1.8159,
+      "step": 3150
+    },
+    {
+      "epoch": 1.444903107441807,
+      "grad_norm": 0.5788086652755737,
+      "learning_rate": 0.0005540043281821777,
+      "loss": 2.1663,
+      "step": 3151
+    },
+    {
+      "epoch": 1.44536177043917,
+      "grad_norm": 0.2490013986825943,
+      "learning_rate": 0.0005537581277964919,
+      "loss": 1.3651,
+      "step": 3152
+    },
+    {
+      "epoch": 1.4458204334365325,
+      "grad_norm": 0.28390905261039734,
+      "learning_rate": 0.0005535119142235499,
+      "loss": 0.8854,
+      "step": 3153
+    },
+    {
+      "epoch": 1.4462790964338952,
+      "grad_norm": 0.31385406851768494,
+      "learning_rate": 0.0005532656875237492,
+      "loss": 1.4232,
+      "step": 3154
+    },
+    {
+      "epoch": 1.4467377594312578,
+      "grad_norm": 0.3778940439224243,
+      "learning_rate": 0.0005530194477574914,
+      "loss": 1.9998,
+      "step": 3155
+    },
+    {
+      "epoch": 1.4471964224286205,
+      "grad_norm": 0.18103332817554474,
+      "learning_rate": 0.0005527731949851809,
+      "loss": 0.7658,
+      "step": 3156
+    },
+    {
+      "epoch": 1.4476550854259833,
+      "grad_norm": 0.26208803057670593,
+      "learning_rate": 0.000552526929267225,
+      "loss": 1.3012,
+      "step": 3157
+    },
+    {
+      "epoch": 1.448113748423346,
+      "grad_norm": 0.23250733315944672,
+      "learning_rate": 0.0005522806506640346,
+      "loss": 1.4842,
+      "step": 3158
+    },
+    {
+      "epoch": 1.4485724114207086,
+      "grad_norm": 0.38623639941215515,
+      "learning_rate": 0.0005520343592360238,
+      "loss": 2.0955,
+      "step": 3159
+    },
+    {
+      "epoch": 1.4490310744180714,
+      "grad_norm": 0.30866172909736633,
+      "learning_rate": 0.0005517880550436094,
+      "loss": 1.6685,
+      "step": 3160
+    },
+    {
+      "epoch": 1.449489737415434,
+      "grad_norm": 0.1992468386888504,
+      "learning_rate": 0.0005515417381472117,
+      "loss": 0.969,
+      "step": 3161
+    },
+    {
+      "epoch": 1.4499484004127967,
+      "grad_norm": 0.29209160804748535,
+      "learning_rate": 0.000551295408607254,
+      "loss": 2.0913,
+      "step": 3162
+    },
+    {
+      "epoch": 1.4504070634101593,
+      "grad_norm": 0.19212748110294342,
+      "learning_rate": 0.0005510490664841626,
+      "loss": 0.8058,
+      "step": 3163
+    },
+    {
+      "epoch": 1.450865726407522,
+      "grad_norm": 0.31732991337776184,
+      "learning_rate": 0.0005508027118383673,
+      "loss": 1.6986,
+      "step": 3164
+    },
+    {
+      "epoch": 1.4513243894048848,
+      "grad_norm": 0.23571406304836273,
+      "learning_rate": 0.0005505563447303004,
+      "loss": 1.1407,
+      "step": 3165
+    },
+    {
+      "epoch": 1.4517830524022475,
+      "grad_norm": 0.24879354238510132,
+      "learning_rate": 0.0005503099652203974,
+      "loss": 1.7388,
+      "step": 3166
+    },
+    {
+      "epoch": 1.4522417153996101,
+      "grad_norm": 0.28732791543006897,
+      "learning_rate": 0.0005500635733690975,
+      "loss": 1.2711,
+      "step": 3167
+    },
+    {
+      "epoch": 1.4527003783969727,
+      "grad_norm": 0.2248416393995285,
+      "learning_rate": 0.0005498171692368419,
+      "loss": 1.1397,
+      "step": 3168
+    },
+    {
+      "epoch": 1.4531590413943354,
+      "grad_norm": 0.34044063091278076,
+      "learning_rate": 0.0005495707528840755,
+      "loss": 1.819,
+      "step": 3169
+    },
+    {
+      "epoch": 1.4536177043916982,
+      "grad_norm": 0.19185033440589905,
+      "learning_rate": 0.0005493243243712461,
+      "loss": 1.2428,
+      "step": 3170
+    },
+    {
+      "epoch": 1.4540763673890609,
+      "grad_norm": 0.2840229570865631,
+      "learning_rate": 0.0005490778837588041,
+      "loss": 1.1735,
+      "step": 3171
+    },
+    {
+      "epoch": 1.4545350303864235,
+      "grad_norm": 0.3059384822845459,
+      "learning_rate": 0.0005488314311072035,
+      "loss": 1.9674,
+      "step": 3172
+    },
+    {
+      "epoch": 1.4549936933837864,
+      "grad_norm": 0.27836617827415466,
+      "learning_rate": 0.0005485849664769008,
+      "loss": 0.7415,
+      "step": 3173
+    },
+    {
+      "epoch": 1.455452356381149,
+      "grad_norm": 0.21108005940914154,
+      "learning_rate": 0.0005483384899283554,
+      "loss": 1.1039,
+      "step": 3174
+    },
+    {
+      "epoch": 1.4559110193785116,
+      "grad_norm": 0.24077321588993073,
+      "learning_rate": 0.0005480920015220298,
+      "loss": 1.6875,
+      "step": 3175
+    },
+    {
+      "epoch": 1.4563696823758743,
+      "grad_norm": 0.3374916911125183,
+      "learning_rate": 0.0005478455013183895,
+      "loss": 1.938,
+      "step": 3176
+    },
+    {
+      "epoch": 1.456828345373237,
+      "grad_norm": 0.3088974356651306,
+      "learning_rate": 0.0005475989893779028,
+      "loss": 1.232,
+      "step": 3177
+    },
+    {
+      "epoch": 1.4572870083705998,
+      "grad_norm": 0.3156466782093048,
+      "learning_rate": 0.0005473524657610409,
+      "loss": 2.2339,
+      "step": 3178
+    },
+    {
+      "epoch": 1.4577456713679624,
+      "grad_norm": 0.373761922121048,
+      "learning_rate": 0.0005471059305282776,
+      "loss": 1.6581,
+      "step": 3179
+    },
+    {
+      "epoch": 1.458204334365325,
+      "grad_norm": 0.18877850472927094,
+      "learning_rate": 0.00054685938374009,
+      "loss": 0.885,
+      "step": 3180
+    },
+    {
+      "epoch": 1.458662997362688,
+      "grad_norm": 0.2712680995464325,
+      "learning_rate": 0.0005466128254569575,
+      "loss": 1.3269,
+      "step": 3181
+    },
+    {
+      "epoch": 1.4591216603600505,
+      "grad_norm": 0.17114239931106567,
+      "learning_rate": 0.0005463662557393628,
+      "loss": 0.8921,
+      "step": 3182
+    },
+    {
+      "epoch": 1.4595803233574132,
+      "grad_norm": 0.22322210669517517,
+      "learning_rate": 0.0005461196746477915,
+      "loss": 1.4297,
+      "step": 3183
+    },
+    {
+      "epoch": 1.4600389863547758,
+      "grad_norm": 0.24018515646457672,
+      "learning_rate": 0.0005458730822427314,
+      "loss": 0.8356,
+      "step": 3184
+    },
+    {
+      "epoch": 1.4604976493521384,
+      "grad_norm": 0.18391193449497223,
+      "learning_rate": 0.0005456264785846733,
+      "loss": 1.4681,
+      "step": 3185
+    },
+    {
+      "epoch": 1.4609563123495013,
+      "grad_norm": 0.33580252528190613,
+      "learning_rate": 0.0005453798637341112,
+      "loss": 1.371,
+      "step": 3186
+    },
+    {
+      "epoch": 1.461414975346864,
+      "grad_norm": 0.2702253460884094,
+      "learning_rate": 0.0005451332377515412,
+      "loss": 1.6213,
+      "step": 3187
+    },
+    {
+      "epoch": 1.4618736383442266,
+      "grad_norm": 0.32494962215423584,
+      "learning_rate": 0.0005448866006974625,
+      "loss": 1.9729,
+      "step": 3188
+    },
+    {
+      "epoch": 1.4623323013415892,
+      "grad_norm": 0.27085816860198975,
+      "learning_rate": 0.000544639952632377,
+      "loss": 1.0135,
+      "step": 3189
+    },
+    {
+      "epoch": 1.4627909643389518,
+      "grad_norm": 0.20625822246074677,
+      "learning_rate": 0.0005443932936167895,
+      "loss": 1.3878,
+      "step": 3190
+    },
+    {
+      "epoch": 1.4632496273363147,
+      "grad_norm": 0.38351795077323914,
+      "learning_rate": 0.0005441466237112068,
+      "loss": 1.9377,
+      "step": 3191
+    },
+    {
+      "epoch": 1.4637082903336773,
+      "grad_norm": 0.20434589684009552,
+      "learning_rate": 0.000543899942976139,
+      "loss": 1.1232,
+      "step": 3192
+    },
+    {
+      "epoch": 1.46416695333104,
+      "grad_norm": 0.28791090846061707,
+      "learning_rate": 0.0005436532514720986,
+      "loss": 0.877,
+      "step": 3193
+    },
+    {
+      "epoch": 1.4646256163284028,
+      "grad_norm": 0.27952685952186584,
+      "learning_rate": 0.000543406549259601,
+      "loss": 1.4366,
+      "step": 3194
+    },
+    {
+      "epoch": 1.4650842793257655,
+      "grad_norm": 0.34718945622444153,
+      "learning_rate": 0.0005431598363991639,
+      "loss": 1.9646,
+      "step": 3195
+    },
+    {
+      "epoch": 1.465542942323128,
+      "grad_norm": 0.3882228434085846,
+      "learning_rate": 0.0005429131129513076,
+      "loss": 1.6616,
+      "step": 3196
+    },
+    {
+      "epoch": 1.4660016053204907,
+      "grad_norm": 0.2647295296192169,
+      "learning_rate": 0.0005426663789765552,
+      "loss": 1.4854,
+      "step": 3197
+    },
+    {
+      "epoch": 1.4664602683178534,
+      "grad_norm": 0.28624817728996277,
+      "learning_rate": 0.0005424196345354326,
+      "loss": 1.9561,
+      "step": 3198
+    },
+    {
+      "epoch": 1.4669189313152162,
+      "grad_norm": 0.3558148741722107,
+      "learning_rate": 0.0005421728796884677,
+      "loss": 1.7657,
+      "step": 3199
+    },
+    {
+      "epoch": 1.4673775943125789,
+      "grad_norm": 0.29009684920310974,
+      "learning_rate": 0.0005419261144961914,
+      "loss": 1.6457,
+      "step": 3200
+    },
+    {
+      "epoch": 1.4678362573099415,
+      "grad_norm": 0.3886021375656128,
+      "learning_rate": 0.0005416793390191369,
+      "loss": 1.6,
+      "step": 3201
+    },
+    {
+      "epoch": 1.4682949203073041,
+      "grad_norm": 0.14990092813968658,
+      "learning_rate": 0.0005414325533178398,
+      "loss": 0.8758,
+      "step": 3202
+    },
+    {
+      "epoch": 1.4687535833046668,
+      "grad_norm": 0.21671035885810852,
+      "learning_rate": 0.0005411857574528389,
+      "loss": 0.9272,
+      "step": 3203
+    },
+    {
+      "epoch": 1.4692122463020296,
+      "grad_norm": 0.2458488643169403,
+      "learning_rate": 0.0005409389514846746,
+      "loss": 1.6566,
+      "step": 3204
+    },
+    {
+      "epoch": 1.4696709092993923,
+      "grad_norm": 0.38461047410964966,
+      "learning_rate": 0.0005406921354738904,
+      "loss": 1.5311,
+      "step": 3205
+    },
+    {
+      "epoch": 1.470129572296755,
+      "grad_norm": 0.31382817029953003,
+      "learning_rate": 0.000540445309481032,
+      "loss": 1.6455,
+      "step": 3206
+    },
+    {
+      "epoch": 1.4705882352941178,
+      "grad_norm": 0.23861967027187347,
+      "learning_rate": 0.0005401984735666474,
+      "loss": 1.2702,
+      "step": 3207
+    },
+    {
+      "epoch": 1.4710468982914804,
+      "grad_norm": 0.4511847496032715,
+      "learning_rate": 0.0005399516277912873,
+      "loss": 1.2289,
+      "step": 3208
+    },
+    {
+      "epoch": 1.471505561288843,
+      "grad_norm": 0.3328205645084381,
+      "learning_rate": 0.0005397047722155051,
+      "loss": 1.4151,
+      "step": 3209
+    },
+    {
+      "epoch": 1.4719642242862057,
+      "grad_norm": 0.10846666991710663,
+      "learning_rate": 0.0005394579068998559,
+      "loss": 0.8251,
+      "step": 3210
+    },
+    {
+      "epoch": 1.4724228872835683,
+      "grad_norm": 0.3210654854774475,
+      "learning_rate": 0.0005392110319048975,
+      "loss": 1.7081,
+      "step": 3211
+    },
+    {
+      "epoch": 1.4728815502809312,
+      "grad_norm": 0.1869962215423584,
+      "learning_rate": 0.0005389641472911904,
+      "loss": 1.0442,
+      "step": 3212
+    },
+    {
+      "epoch": 1.4733402132782938,
+      "grad_norm": 0.3858005404472351,
+      "learning_rate": 0.000538717253119297,
+      "loss": 1.7048,
+      "step": 3213
+    },
+    {
+      "epoch": 1.4737988762756564,
+      "grad_norm": 0.22492477297782898,
+      "learning_rate": 0.0005384703494497821,
+      "loss": 1.5041,
+      "step": 3214
+    },
+    {
+      "epoch": 1.4742575392730193,
+      "grad_norm": 0.3225257694721222,
+      "learning_rate": 0.0005382234363432134,
+      "loss": 1.8363,
+      "step": 3215
+    },
+    {
+      "epoch": 1.474716202270382,
+      "grad_norm": 0.37609943747520447,
+      "learning_rate": 0.0005379765138601598,
+      "loss": 2.0479,
+      "step": 3216
+    },
+    {
+      "epoch": 1.4751748652677446,
+      "grad_norm": 0.38782861828804016,
+      "learning_rate": 0.0005377295820611939,
+      "loss": 1.3388,
+      "step": 3217
+    },
+    {
+      "epoch": 1.4756335282651072,
+      "grad_norm": 0.40626588463783264,
+      "learning_rate": 0.0005374826410068891,
+      "loss": 1.9222,
+      "step": 3218
+    },
+    {
+      "epoch": 1.4760921912624698,
+      "grad_norm": 0.4013312757015228,
+      "learning_rate": 0.0005372356907578224,
+      "loss": 2.2166,
+      "step": 3219
+    },
+    {
+      "epoch": 1.4765508542598327,
+      "grad_norm": 0.2755921483039856,
+      "learning_rate": 0.0005369887313745723,
+      "loss": 1.2296,
+      "step": 3220
+    },
+    {
+      "epoch": 1.4770095172571953,
+      "grad_norm": 0.17257989943027496,
+      "learning_rate": 0.0005367417629177196,
+      "loss": 0.9679,
+      "step": 3221
+    },
+    {
+      "epoch": 1.477468180254558,
+      "grad_norm": 0.28084853291511536,
+      "learning_rate": 0.0005364947854478476,
+      "loss": 1.3068,
+      "step": 3222
+    },
+    {
+      "epoch": 1.4779268432519206,
+      "grad_norm": 0.24307353794574738,
+      "learning_rate": 0.0005362477990255416,
+      "loss": 0.8342,
+      "step": 3223
+    },
+    {
+      "epoch": 1.4783855062492832,
+      "grad_norm": 0.22293055057525635,
+      "learning_rate": 0.000536000803711389,
+      "loss": 1.3204,
+      "step": 3224
+    },
+    {
+      "epoch": 1.478844169246646,
+      "grad_norm": 0.2704020142555237,
+      "learning_rate": 0.0005357537995659798,
+      "loss": 1.0284,
+      "step": 3225
+    },
+    {
+      "epoch": 1.4793028322440087,
+      "grad_norm": 0.22363778948783875,
+      "learning_rate": 0.0005355067866499059,
+      "loss": 1.4955,
+      "step": 3226
+    },
+    {
+      "epoch": 1.4797614952413713,
+      "grad_norm": 0.3845861256122589,
+      "learning_rate": 0.000535259765023761,
+      "loss": 1.5793,
+      "step": 3227
+    },
+    {
+      "epoch": 1.4802201582387342,
+      "grad_norm": 0.2982635498046875,
+      "learning_rate": 0.0005350127347481414,
+      "loss": 1.2065,
+      "step": 3228
+    },
+    {
+      "epoch": 1.4806788212360968,
+      "grad_norm": 0.18739233911037445,
+      "learning_rate": 0.0005347656958836457,
+      "loss": 1.3103,
+      "step": 3229
+    },
+    {
+      "epoch": 1.4811374842334595,
+      "grad_norm": 0.34681299328804016,
+      "learning_rate": 0.0005345186484908741,
+      "loss": 1.4734,
+      "step": 3230
+    },
+    {
+      "epoch": 1.4815961472308221,
+      "grad_norm": 0.3014410734176636,
+      "learning_rate": 0.0005342715926304291,
+      "loss": 1.6099,
+      "step": 3231
+    },
+    {
+      "epoch": 1.4820548102281847,
+      "grad_norm": 0.4709073603153229,
+      "learning_rate": 0.0005340245283629155,
+      "loss": 1.7924,
+      "step": 3232
+    },
+    {
+      "epoch": 1.4825134732255476,
+      "grad_norm": 0.37753811478614807,
+      "learning_rate": 0.0005337774557489394,
+      "loss": 1.3316,
+      "step": 3233
+    },
+    {
+      "epoch": 1.4829721362229102,
+      "grad_norm": 0.31126853823661804,
+      "learning_rate": 0.0005335303748491101,
+      "loss": 0.7645,
+      "step": 3234
+    },
+    {
+      "epoch": 1.4834307992202729,
+      "grad_norm": 0.10572401434183121,
+      "learning_rate": 0.0005332832857240381,
+      "loss": 0.9926,
+      "step": 3235
+    },
+    {
+      "epoch": 1.4838894622176355,
+      "grad_norm": 0.38704633712768555,
+      "learning_rate": 0.0005330361884343361,
+      "loss": 2.0361,
+      "step": 3236
+    },
+    {
+      "epoch": 1.4843481252149981,
+      "grad_norm": 0.2931767702102661,
+      "learning_rate": 0.0005327890830406189,
+      "loss": 0.9363,
+      "step": 3237
+    },
+    {
+      "epoch": 1.484806788212361,
+      "grad_norm": 0.11565959453582764,
+      "learning_rate": 0.000532541969603503,
+      "loss": 0.9478,
+      "step": 3238
+    },
+    {
+      "epoch": 1.4852654512097236,
+      "grad_norm": 0.3610280156135559,
+      "learning_rate": 0.0005322948481836075,
+      "loss": 1.6384,
+      "step": 3239
+    },
+    {
+      "epoch": 1.4857241142070863,
+      "grad_norm": 0.32327285408973694,
+      "learning_rate": 0.0005320477188415529,
+      "loss": 0.8816,
+      "step": 3240
+    },
+    {
+      "epoch": 1.4861827772044491,
+      "grad_norm": 0.20238777995109558,
+      "learning_rate": 0.0005318005816379618,
+      "loss": 1.2931,
+      "step": 3241
+    },
+    {
+      "epoch": 1.4866414402018118,
+      "grad_norm": 0.29169613122940063,
+      "learning_rate": 0.0005315534366334587,
+      "loss": 1.1745,
+      "step": 3242
+    },
+    {
+      "epoch": 1.4871001031991744,
+      "grad_norm": 0.34999263286590576,
+      "learning_rate": 0.00053130628388867,
+      "loss": 1.7833,
+      "step": 3243
+    },
+    {
+      "epoch": 1.487558766196537,
+      "grad_norm": 0.25474709272384644,
+      "learning_rate": 0.0005310591234642242,
+      "loss": 1.2222,
+      "step": 3244
+    },
+    {
+      "epoch": 1.4880174291938997,
+      "grad_norm": 0.333258718252182,
+      "learning_rate": 0.0005308119554207515,
+      "loss": 1.3405,
+      "step": 3245
+    },
+    {
+      "epoch": 1.4884760921912625,
+      "grad_norm": 0.19669318199157715,
+      "learning_rate": 0.0005305647798188839,
+      "loss": 1.2637,
+      "step": 3246
+    },
+    {
+      "epoch": 1.4889347551886252,
+      "grad_norm": 0.25908851623535156,
+      "learning_rate": 0.0005303175967192555,
+      "loss": 1.5399,
+      "step": 3247
+    },
+    {
+      "epoch": 1.4893934181859878,
+      "grad_norm": 0.32331693172454834,
+      "learning_rate": 0.0005300704061825019,
+      "loss": 1.6418,
+      "step": 3248
+    },
+    {
+      "epoch": 1.4898520811833507,
+      "grad_norm": 0.2600635588169098,
+      "learning_rate": 0.0005298232082692609,
+      "loss": 1.0823,
+      "step": 3249
+    },
+    {
+      "epoch": 1.4903107441807133,
+      "grad_norm": 0.2755780518054962,
+      "learning_rate": 0.000529576003040172,
+      "loss": 1.6127,
+      "step": 3250
+    },
+    {
+      "epoch": 1.490769407178076,
+      "grad_norm": 0.33199846744537354,
+      "learning_rate": 0.0005293287905558762,
+      "loss": 1.6889,
+      "step": 3251
+    },
+    {
+      "epoch": 1.4912280701754386,
+      "grad_norm": 0.24848337471485138,
+      "learning_rate": 0.0005290815708770166,
+      "loss": 0.8654,
+      "step": 3252
+    },
+    {
+      "epoch": 1.4916867331728012,
+      "grad_norm": 0.2537640333175659,
+      "learning_rate": 0.0005288343440642379,
+      "loss": 1.316,
+      "step": 3253
+    },
+    {
+      "epoch": 1.492145396170164,
+      "grad_norm": 0.22786585986614227,
+      "learning_rate": 0.0005285871101781868,
+      "loss": 1.1967,
+      "step": 3254
+    },
+    {
+      "epoch": 1.4926040591675267,
+      "grad_norm": 0.30544570088386536,
+      "learning_rate": 0.0005283398692795114,
+      "loss": 1.6742,
+      "step": 3255
+    },
+    {
+      "epoch": 1.4930627221648893,
+      "grad_norm": 1.4452085494995117,
+      "learning_rate": 0.0005280926214288617,
+      "loss": 0.8225,
+      "step": 3256
+    },
+    {
+      "epoch": 1.493521385162252,
+      "grad_norm": 0.2699876129627228,
+      "learning_rate": 0.0005278453666868896,
+      "loss": 1.3572,
+      "step": 3257
+    },
+    {
+      "epoch": 1.4939800481596146,
+      "grad_norm": 0.35777711868286133,
+      "learning_rate": 0.0005275981051142481,
+      "loss": 2.1099,
+      "step": 3258
+    },
+    {
+      "epoch": 1.4944387111569775,
+      "grad_norm": 0.19279718399047852,
+      "learning_rate": 0.0005273508367715923,
+      "loss": 0.8453,
+      "step": 3259
+    },
+    {
+      "epoch": 1.49489737415434,
+      "grad_norm": 0.29106536507606506,
+      "learning_rate": 0.0005271035617195793,
+      "loss": 1.2269,
+      "step": 3260
+    },
+    {
+      "epoch": 1.4953560371517027,
+      "grad_norm": 0.286878377199173,
+      "learning_rate": 0.0005268562800188671,
+      "loss": 1.6506,
+      "step": 3261
+    },
+    {
+      "epoch": 1.4958147001490656,
+      "grad_norm": 0.3785189688205719,
+      "learning_rate": 0.0005266089917301158,
+      "loss": 1.5022,
+      "step": 3262
+    },
+    {
+      "epoch": 1.4962733631464282,
+      "grad_norm": 0.25897216796875,
+      "learning_rate": 0.0005263616969139868,
+      "loss": 1.6508,
+      "step": 3263
+    },
+    {
+      "epoch": 1.4967320261437909,
+      "grad_norm": 0.3450983166694641,
+      "learning_rate": 0.0005261143956311435,
+      "loss": 1.7249,
+      "step": 3264
+    },
+    {
+      "epoch": 1.4971906891411535,
+      "grad_norm": 0.28310757875442505,
+      "learning_rate": 0.0005258670879422508,
+      "loss": 1.2171,
+      "step": 3265
+    },
+    {
+      "epoch": 1.4976493521385161,
+      "grad_norm": 0.3658524751663208,
+      "learning_rate": 0.0005256197739079749,
+      "loss": 1.3394,
+      "step": 3266
+    },
+    {
+      "epoch": 1.498108015135879,
+      "grad_norm": 0.3043535053730011,
+      "learning_rate": 0.0005253724535889836,
+      "loss": 1.3353,
+      "step": 3267
+    },
+    {
+      "epoch": 1.4985666781332416,
+      "grad_norm": 0.2485763430595398,
+      "learning_rate": 0.0005251251270459468,
+      "loss": 1.0306,
+      "step": 3268
+    },
+    {
+      "epoch": 1.4990253411306043,
+      "grad_norm": 0.26514461636543274,
+      "learning_rate": 0.0005248777943395347,
+      "loss": 1.0359,
+      "step": 3269
+    },
+    {
+      "epoch": 1.499484004127967,
+      "grad_norm": 0.26036179065704346,
+      "learning_rate": 0.0005246304555304205,
+      "loss": 1.3007,
+      "step": 3270
+    },
+    {
+      "epoch": 1.4999426671253295,
+      "grad_norm": 0.20205332338809967,
+      "learning_rate": 0.000524383110679278,
+      "loss": 0.61,
+      "step": 3271
+    },
+    {
+      "epoch": 1.5004013301226924,
+      "grad_norm": 0.36065131425857544,
+      "learning_rate": 0.0005241357598467826,
+      "loss": 2.1804,
+      "step": 3272
+    },
+    {
+      "epoch": 1.500859993120055,
+      "grad_norm": 0.3411425054073334,
+      "learning_rate": 0.0005238884030936112,
+      "loss": 1.4267,
+      "step": 3273
+    },
+    {
+      "epoch": 1.5013186561174177,
+      "grad_norm": 0.45014554262161255,
+      "learning_rate": 0.0005236410404804425,
+      "loss": 2.0813,
+      "step": 3274
+    },
+    {
+      "epoch": 1.5017773191147805,
+      "grad_norm": 0.4235081076622009,
+      "learning_rate": 0.000523393672067956,
+      "loss": 1.2112,
+      "step": 3275
+    },
+    {
+      "epoch": 1.5022359821121432,
+      "grad_norm": 0.26777932047843933,
+      "learning_rate": 0.0005231462979168331,
+      "loss": 0.8293,
+      "step": 3276
+    },
+    {
+      "epoch": 1.5026946451095058,
+      "grad_norm": 0.20920783281326294,
+      "learning_rate": 0.0005228989180877564,
+      "loss": 1.123,
+      "step": 3277
+    },
+    {
+      "epoch": 1.5031533081068684,
+      "grad_norm": 0.27009251713752747,
+      "learning_rate": 0.0005226515326414099,
+      "loss": 1.7735,
+      "step": 3278
+    },
+    {
+      "epoch": 1.503611971104231,
+      "grad_norm": 0.3320116698741913,
+      "learning_rate": 0.0005224041416384791,
+      "loss": 1.6422,
+      "step": 3279
+    },
+    {
+      "epoch": 1.504070634101594,
+      "grad_norm": 0.2971821129322052,
+      "learning_rate": 0.0005221567451396509,
+      "loss": 0.9693,
+      "step": 3280
+    },
+    {
+      "epoch": 1.5045292970989566,
+      "grad_norm": 0.13652589917182922,
+      "learning_rate": 0.0005219093432056133,
+      "loss": 1.0251,
+      "step": 3281
+    },
+    {
+      "epoch": 1.5049879600963192,
+      "grad_norm": 0.31048429012298584,
+      "learning_rate": 0.000521661935897056,
+      "loss": 1.3032,
+      "step": 3282
+    },
+    {
+      "epoch": 1.505446623093682,
+      "grad_norm": 0.2567446827888489,
+      "learning_rate": 0.0005214145232746696,
+      "loss": 0.8873,
+      "step": 3283
+    },
+    {
+      "epoch": 1.5059052860910445,
+      "grad_norm": 0.2890402674674988,
+      "learning_rate": 0.000521167105399146,
+      "loss": 1.3135,
+      "step": 3284
+    },
+    {
+      "epoch": 1.5063639490884073,
+      "grad_norm": 0.3106888234615326,
+      "learning_rate": 0.0005209196823311791,
+      "loss": 1.3594,
+      "step": 3285
+    },
+    {
+      "epoch": 1.50682261208577,
+      "grad_norm": 0.28741034865379333,
+      "learning_rate": 0.0005206722541314631,
+      "loss": 1.7812,
+      "step": 3286
+    },
+    {
+      "epoch": 1.5072812750831326,
+      "grad_norm": 0.32389602065086365,
+      "learning_rate": 0.0005204248208606942,
+      "loss": 1.4911,
+      "step": 3287
+    },
+    {
+      "epoch": 1.5077399380804954,
+      "grad_norm": 0.26342904567718506,
+      "learning_rate": 0.0005201773825795694,
+      "loss": 1.6143,
+      "step": 3288
+    },
+    {
+      "epoch": 1.508198601077858,
+      "grad_norm": 0.21466577053070068,
+      "learning_rate": 0.0005199299393487872,
+      "loss": 0.3381,
+      "step": 3289
+    },
+    {
+      "epoch": 1.5086572640752207,
+      "grad_norm": 0.3118995428085327,
+      "learning_rate": 0.0005196824912290472,
+      "loss": 1.2741,
+      "step": 3290
+    },
+    {
+      "epoch": 1.5091159270725836,
+      "grad_norm": 0.3177085220813751,
+      "learning_rate": 0.00051943503828105,
+      "loss": 1.3895,
+      "step": 3291
+    },
+    {
+      "epoch": 1.509574590069946,
+      "grad_norm": 0.14354930818080902,
+      "learning_rate": 0.0005191875805654981,
+      "loss": 1.14,
+      "step": 3292
+    },
+    {
+      "epoch": 1.5100332530673088,
+      "grad_norm": 0.256320983171463,
+      "learning_rate": 0.0005189401181430941,
+      "loss": 0.592,
+      "step": 3293
+    },
+    {
+      "epoch": 1.5104919160646715,
+      "grad_norm": 0.33321627974510193,
+      "learning_rate": 0.0005186926510745427,
+      "loss": 2.0759,
+      "step": 3294
+    },
+    {
+      "epoch": 1.5109505790620341,
+      "grad_norm": 0.26053670048713684,
+      "learning_rate": 0.0005184451794205491,
+      "loss": 1.1047,
+      "step": 3295
+    },
+    {
+      "epoch": 1.511409242059397,
+      "grad_norm": 0.408119261264801,
+      "learning_rate": 0.00051819770324182,
+      "loss": 1.6132,
+      "step": 3296
+    },
+    {
+      "epoch": 1.5118679050567594,
+      "grad_norm": 0.2578348219394684,
+      "learning_rate": 0.0005179502225990632,
+      "loss": 0.9136,
+      "step": 3297
+    },
+    {
+      "epoch": 1.5123265680541222,
+      "grad_norm": 0.2745778560638428,
+      "learning_rate": 0.0005177027375529872,
+      "loss": 1.7166,
+      "step": 3298
+    },
+    {
+      "epoch": 1.5127852310514849,
+      "grad_norm": 0.3158656656742096,
+      "learning_rate": 0.0005174552481643023,
+      "loss": 1.1242,
+      "step": 3299
+    },
+    {
+      "epoch": 1.5132438940488475,
+      "grad_norm": 0.270859032869339,
+      "learning_rate": 0.0005172077544937188,
+      "loss": 1.9039,
+      "step": 3300
+    },
+    {
+      "epoch": 1.5137025570462104,
+      "grad_norm": 0.4259583652019501,
+      "learning_rate": 0.0005169602566019493,
+      "loss": 2.1599,
+      "step": 3301
+    },
+    {
+      "epoch": 1.514161220043573,
+      "grad_norm": 0.33669549226760864,
+      "learning_rate": 0.0005167127545497066,
+      "loss": 2.24,
+      "step": 3302
+    },
+    {
+      "epoch": 1.5146198830409356,
+      "grad_norm": 0.3532778024673462,
+      "learning_rate": 0.0005164652483977044,
+      "loss": 1.3746,
+      "step": 3303
+    },
+    {
+      "epoch": 1.5150785460382985,
+      "grad_norm": 0.22524480521678925,
+      "learning_rate": 0.0005162177382066584,
+      "loss": 0.8802,
+      "step": 3304
+    },
+    {
+      "epoch": 1.515537209035661,
+      "grad_norm": 0.21874026954174042,
+      "learning_rate": 0.000515970224037284,
+      "loss": 0.9228,
+      "step": 3305
+    },
+    {
+      "epoch": 1.5159958720330238,
+      "grad_norm": 0.2486424446105957,
+      "learning_rate": 0.0005157227059502987,
+      "loss": 1.7778,
+      "step": 3306
+    },
+    {
+      "epoch": 1.5164545350303864,
+      "grad_norm": 0.28031882643699646,
+      "learning_rate": 0.0005154751840064203,
+      "loss": 1.2578,
+      "step": 3307
+    },
+    {
+      "epoch": 1.516913198027749,
+      "grad_norm": 0.4096969664096832,
+      "learning_rate": 0.0005152276582663676,
+      "loss": 2.0068,
+      "step": 3308
+    },
+    {
+      "epoch": 1.517371861025112,
+      "grad_norm": 0.3127899467945099,
+      "learning_rate": 0.0005149801287908604,
+      "loss": 1.2456,
+      "step": 3309
+    },
+    {
+      "epoch": 1.5178305240224745,
+      "grad_norm": 0.325255811214447,
+      "learning_rate": 0.0005147325956406197,
+      "loss": 1.6535,
+      "step": 3310
+    },
+    {
+      "epoch": 1.5182891870198372,
+      "grad_norm": 0.06586357206106186,
+      "learning_rate": 0.000514485058876367,
+      "loss": 0.9711,
+      "step": 3311
+    },
+    {
+      "epoch": 1.5187478500171998,
+      "grad_norm": 0.31240516901016235,
+      "learning_rate": 0.000514237518558825,
+      "loss": 1.6185,
+      "step": 3312
+    },
+    {
+      "epoch": 1.5192065130145624,
+      "grad_norm": 0.29829588532447815,
+      "learning_rate": 0.0005139899747487171,
+      "loss": 1.2696,
+      "step": 3313
+    },
+    {
+      "epoch": 1.5196651760119253,
+      "grad_norm": 0.25990763306617737,
+      "learning_rate": 0.0005137424275067674,
+      "loss": 1.3779,
+      "step": 3314
+    },
+    {
+      "epoch": 1.520123839009288,
+      "grad_norm": 0.32825708389282227,
+      "learning_rate": 0.0005134948768937011,
+      "loss": 1.3303,
+      "step": 3315
+    },
+    {
+      "epoch": 1.5205825020066506,
+      "grad_norm": 0.2681593894958496,
+      "learning_rate": 0.0005132473229702444,
+      "loss": 1.9851,
+      "step": 3316
+    },
+    {
+      "epoch": 1.5210411650040134,
+      "grad_norm": 0.2627713680267334,
+      "learning_rate": 0.0005129997657971236,
+      "loss": 1.3954,
+      "step": 3317
+    },
+    {
+      "epoch": 1.5214998280013758,
+      "grad_norm": 0.22389861941337585,
+      "learning_rate": 0.0005127522054350667,
+      "loss": 0.7996,
+      "step": 3318
+    },
+    {
+      "epoch": 1.5219584909987387,
+      "grad_norm": 0.06943022459745407,
+      "learning_rate": 0.0005125046419448019,
+      "loss": 0.4316,
+      "step": 3319
+    },
+    {
+      "epoch": 1.5224171539961013,
+      "grad_norm": 0.16455991566181183,
+      "learning_rate": 0.0005122570753870582,
+      "loss": 0.9303,
+      "step": 3320
+    },
+    {
+      "epoch": 1.522875816993464,
+      "grad_norm": 0.22131837904453278,
+      "learning_rate": 0.0005120095058225654,
+      "loss": 1.1393,
+      "step": 3321
+    },
+    {
+      "epoch": 1.5233344799908268,
+      "grad_norm": 0.2995634973049164,
+      "learning_rate": 0.0005117619333120544,
+      "loss": 1.0248,
+      "step": 3322
+    },
+    {
+      "epoch": 1.5237931429881895,
+      "grad_norm": 0.2665209472179413,
+      "learning_rate": 0.0005115143579162561,
+      "loss": 1.1959,
+      "step": 3323
+    },
+    {
+      "epoch": 1.524251805985552,
+      "grad_norm": 0.41921502351760864,
+      "learning_rate": 0.000511266779695903,
+      "loss": 1.6565,
+      "step": 3324
+    },
+    {
+      "epoch": 1.524710468982915,
+      "grad_norm": 0.3324905037879944,
+      "learning_rate": 0.0005110191987117277,
+      "loss": 2.2944,
+      "step": 3325
+    },
+    {
+      "epoch": 1.5251691319802774,
+      "grad_norm": 0.22564005851745605,
+      "learning_rate": 0.0005107716150244634,
+      "loss": 0.761,
+      "step": 3326
+    },
+    {
+      "epoch": 1.5256277949776402,
+      "grad_norm": 0.25976645946502686,
+      "learning_rate": 0.0005105240286948442,
+      "loss": 1.4489,
+      "step": 3327
+    },
+    {
+      "epoch": 1.5260864579750029,
+      "grad_norm": 0.3188987076282501,
+      "learning_rate": 0.0005102764397836049,
+      "loss": 1.4666,
+      "step": 3328
+    },
+    {
+      "epoch": 1.5265451209723655,
+      "grad_norm": 0.1813998967409134,
+      "learning_rate": 0.0005100288483514809,
+      "loss": 0.9902,
+      "step": 3329
+    },
+    {
+      "epoch": 1.5270037839697284,
+      "grad_norm": 0.24019388854503632,
+      "learning_rate": 0.000509781254459208,
+      "loss": 1.1389,
+      "step": 3330
+    },
+    {
+      "epoch": 1.5274624469670908,
+      "grad_norm": 0.27883660793304443,
+      "learning_rate": 0.0005095336581675228,
+      "loss": 1.225,
+      "step": 3331
+    },
+    {
+      "epoch": 1.5279211099644536,
+      "grad_norm": 0.3314226269721985,
+      "learning_rate": 0.0005092860595371627,
+      "loss": 2.3223,
+      "step": 3332
+    },
+    {
+      "epoch": 1.5283797729618163,
+      "grad_norm": 0.3248099982738495,
+      "learning_rate": 0.000509038458628865,
+      "loss": 1.6426,
+      "step": 3333
+    },
+    {
+      "epoch": 1.528838435959179,
+      "grad_norm": 0.3302631378173828,
+      "learning_rate": 0.0005087908555033683,
+      "loss": 1.1074,
+      "step": 3334
+    },
+    {
+      "epoch": 1.5292970989565418,
+      "grad_norm": 0.2534301280975342,
+      "learning_rate": 0.0005085432502214116,
+      "loss": 1.6361,
+      "step": 3335
+    },
+    {
+      "epoch": 1.5297557619539044,
+      "grad_norm": 0.4103301167488098,
+      "learning_rate": 0.0005082956428437337,
+      "loss": 1.5873,
+      "step": 3336
+    },
+    {
+      "epoch": 1.530214424951267,
+      "grad_norm": 0.17628952860832214,
+      "learning_rate": 0.000508048033431075,
+      "loss": 0.789,
+      "step": 3337
+    },
+    {
+      "epoch": 1.5306730879486299,
+      "grad_norm": 0.28603991866111755,
+      "learning_rate": 0.0005078004220441756,
+      "loss": 1.2981,
+      "step": 3338
+    },
+    {
+      "epoch": 1.5311317509459923,
+      "grad_norm": 0.30459219217300415,
+      "learning_rate": 0.0005075528087437764,
+      "loss": 1.3026,
+      "step": 3339
+    },
+    {
+      "epoch": 1.5315904139433552,
+      "grad_norm": 0.09923279285430908,
+      "learning_rate": 0.0005073051935906188,
+      "loss": 0.6987,
+      "step": 3340
+    },
+    {
+      "epoch": 1.5320490769407178,
+      "grad_norm": 0.28860870003700256,
+      "learning_rate": 0.0005070575766454445,
+      "loss": 0.8572,
+      "step": 3341
+    },
+    {
+      "epoch": 1.5325077399380804,
+      "grad_norm": 0.2449285089969635,
+      "learning_rate": 0.0005068099579689958,
+      "loss": 1.2982,
+      "step": 3342
+    },
+    {
+      "epoch": 1.5329664029354433,
+      "grad_norm": 0.26312926411628723,
+      "learning_rate": 0.0005065623376220154,
+      "loss": 1.7064,
+      "step": 3343
+    },
+    {
+      "epoch": 1.533425065932806,
+      "grad_norm": 0.36778807640075684,
+      "learning_rate": 0.0005063147156652461,
+      "loss": 1.5251,
+      "step": 3344
+    },
+    {
+      "epoch": 1.5338837289301686,
+      "grad_norm": 0.2748435437679291,
+      "learning_rate": 0.0005060670921594316,
+      "loss": 1.427,
+      "step": 3345
+    },
+    {
+      "epoch": 1.5343423919275314,
+      "grad_norm": 0.18528130650520325,
+      "learning_rate": 0.0005058194671653156,
+      "loss": 0.4789,
+      "step": 3346
+    },
+    {
+      "epoch": 1.5348010549248938,
+      "grad_norm": 0.3140660524368286,
+      "learning_rate": 0.0005055718407436424,
+      "loss": 1.6775,
+      "step": 3347
+    },
+    {
+      "epoch": 1.5352597179222567,
+      "grad_norm": 0.37981608510017395,
+      "learning_rate": 0.0005053242129551564,
+      "loss": 1.5612,
+      "step": 3348
+    },
+    {
+      "epoch": 1.5357183809196193,
+      "grad_norm": 0.2672194242477417,
+      "learning_rate": 0.0005050765838606027,
+      "loss": 1.0255,
+      "step": 3349
+    },
+    {
+      "epoch": 1.536177043916982,
+      "grad_norm": 0.08587975054979324,
+      "learning_rate": 0.0005048289535207264,
+      "loss": 0.9292,
+      "step": 3350
+    },
+    {
+      "epoch": 1.5366357069143448,
+      "grad_norm": 0.3729512691497803,
+      "learning_rate": 0.0005045813219962728,
+      "loss": 1.7538,
+      "step": 3351
+    },
+    {
+      "epoch": 1.5370943699117072,
+      "grad_norm": 0.29773613810539246,
+      "learning_rate": 0.0005043336893479879,
+      "loss": 1.9409,
+      "step": 3352
+    },
+    {
+      "epoch": 1.53755303290907,
+      "grad_norm": 0.27870243787765503,
+      "learning_rate": 0.0005040860556366179,
+      "loss": 0.8977,
+      "step": 3353
+    },
+    {
+      "epoch": 1.5380116959064327,
+      "grad_norm": 0.13409045338630676,
+      "learning_rate": 0.0005038384209229089,
+      "loss": 0.6826,
+      "step": 3354
+    },
+    {
+      "epoch": 1.5384703589037954,
+      "grad_norm": 0.22798749804496765,
+      "learning_rate": 0.0005035907852676076,
+      "loss": 1.1876,
+      "step": 3355
+    },
+    {
+      "epoch": 1.5389290219011582,
+      "grad_norm": 0.3014310300350189,
+      "learning_rate": 0.0005033431487314608,
+      "loss": 1.7822,
+      "step": 3356
+    },
+    {
+      "epoch": 1.5393876848985208,
+      "grad_norm": 0.19987566769123077,
+      "learning_rate": 0.0005030955113752155,
+      "loss": 0.8876,
+      "step": 3357
+    },
+    {
+      "epoch": 1.5398463478958835,
+      "grad_norm": 0.2765864431858063,
+      "learning_rate": 0.0005028478732596189,
+      "loss": 1.7389,
+      "step": 3358
+    },
+    {
+      "epoch": 1.5403050108932463,
+      "grad_norm": 0.26134082674980164,
+      "learning_rate": 0.0005026002344454184,
+      "loss": 0.8819,
+      "step": 3359
+    },
+    {
+      "epoch": 1.5407636738906088,
+      "grad_norm": 0.1979461908340454,
+      "learning_rate": 0.0005023525949933618,
+      "loss": 1.2951,
+      "step": 3360
+    },
+    {
+      "epoch": 1.5412223368879716,
+      "grad_norm": 0.5640215873718262,
+      "learning_rate": 0.0005021049549641967,
+      "loss": 1.6044,
+      "step": 3361
+    },
+    {
+      "epoch": 1.5416809998853342,
+      "grad_norm": 0.4211662709712982,
+      "learning_rate": 0.0005018573144186708,
+      "loss": 1.8833,
+      "step": 3362
+    },
+    {
+      "epoch": 1.5421396628826969,
+      "grad_norm": 0.28429386019706726,
+      "learning_rate": 0.0005016096734175324,
+      "loss": 1.2702,
+      "step": 3363
+    },
+    {
+      "epoch": 1.5425983258800597,
+      "grad_norm": 0.20547764003276825,
+      "learning_rate": 0.0005013620320215294,
+      "loss": 1.0284,
+      "step": 3364
+    },
+    {
+      "epoch": 1.5430569888774222,
+      "grad_norm": 0.3609144985675812,
+      "learning_rate": 0.0005011143902914102,
+      "loss": 1.1247,
+      "step": 3365
+    },
+    {
+      "epoch": 1.543515651874785,
+      "grad_norm": 0.3178929388523102,
+      "learning_rate": 0.000500866748287923,
+      "loss": 1.6366,
+      "step": 3366
+    },
+    {
+      "epoch": 1.5439743148721476,
+      "grad_norm": 0.22963233292102814,
+      "learning_rate": 0.0005006191060718163,
+      "loss": 1.7095,
+      "step": 3367
+    },
+    {
+      "epoch": 1.5444329778695103,
+      "grad_norm": 0.29375985264778137,
+      "learning_rate": 0.0005003714637038381,
+      "loss": 1.2357,
+      "step": 3368
+    },
+    {
+      "epoch": 1.5448916408668731,
+      "grad_norm": 0.26059386134147644,
+      "learning_rate": 0.0005001238212447376,
+      "loss": 1.4654,
+      "step": 3369
+    },
+    {
+      "epoch": 1.5453503038642358,
+      "grad_norm": 0.2658666670322418,
+      "learning_rate": 0.0004998761787552626,
+      "loss": 1.481,
+      "step": 3370
+    },
+    {
+      "epoch": 1.5458089668615984,
+      "grad_norm": 0.26609566807746887,
+      "learning_rate": 0.0004996285362961619,
+      "loss": 1.1959,
+      "step": 3371
+    },
+    {
+      "epoch": 1.5462676298589613,
+      "grad_norm": 0.2814446985721588,
+      "learning_rate": 0.0004993808939281839,
+      "loss": 1.764,
+      "step": 3372
+    },
+    {
+      "epoch": 1.5467262928563237,
+      "grad_norm": 0.35369497537612915,
+      "learning_rate": 0.0004991332517120771,
+      "loss": 1.0037,
+      "step": 3373
+    },
+    {
+      "epoch": 1.5471849558536865,
+      "grad_norm": 0.2843606173992157,
+      "learning_rate": 0.00049888560970859,
+      "loss": 1.7404,
+      "step": 3374
+    },
+    {
+      "epoch": 1.5476436188510492,
+      "grad_norm": 0.20866908133029938,
+      "learning_rate": 0.0004986379679784707,
+      "loss": 0.5552,
+      "step": 3375
+    },
+    {
+      "epoch": 1.5481022818484118,
+      "grad_norm": 0.25553098320961,
+      "learning_rate": 0.0004983903265824677,
+      "loss": 1.1967,
+      "step": 3376
+    },
+    {
+      "epoch": 1.5485609448457747,
+      "grad_norm": 0.26248809695243835,
+      "learning_rate": 0.0004981426855813293,
+      "loss": 1.6772,
+      "step": 3377
+    },
+    {
+      "epoch": 1.5490196078431373,
+      "grad_norm": 0.35352829098701477,
+      "learning_rate": 0.0004978950450358036,
+      "loss": 1.5785,
+      "step": 3378
+    },
+    {
+      "epoch": 1.5494782708405,
+      "grad_norm": 0.26842719316482544,
+      "learning_rate": 0.0004976474050066384,
+      "loss": 1.6119,
+      "step": 3379
+    },
+    {
+      "epoch": 1.5499369338378628,
+      "grad_norm": 0.17810232937335968,
+      "learning_rate": 0.0004973997655545817,
+      "loss": 1.106,
+      "step": 3380
+    },
+    {
+      "epoch": 1.5503955968352252,
+      "grad_norm": 0.2810850143432617,
+      "learning_rate": 0.0004971521267403812,
+      "loss": 1.1019,
+      "step": 3381
+    },
+    {
+      "epoch": 1.550854259832588,
+      "grad_norm": 0.34839344024658203,
+      "learning_rate": 0.0004969044886247846,
+      "loss": 1.548,
+      "step": 3382
+    },
+    {
+      "epoch": 1.5513129228299507,
+      "grad_norm": 0.2651059925556183,
+      "learning_rate": 0.0004966568512685392,
+      "loss": 1.3282,
+      "step": 3383
+    },
+    {
+      "epoch": 1.5517715858273133,
+      "grad_norm": 0.2800137996673584,
+      "learning_rate": 0.0004964092147323925,
+      "loss": 2.1348,
+      "step": 3384
+    },
+    {
+      "epoch": 1.5522302488246762,
+      "grad_norm": 0.223674938082695,
+      "learning_rate": 0.0004961615790770912,
+      "loss": 0.8656,
+      "step": 3385
+    },
+    {
+      "epoch": 1.5526889118220386,
+      "grad_norm": 0.3796307146549225,
+      "learning_rate": 0.0004959139443633823,
+      "loss": 2.2166,
+      "step": 3386
+    },
+    {
+      "epoch": 1.5531475748194015,
+      "grad_norm": 0.3234909176826477,
+      "learning_rate": 0.0004956663106520121,
+      "loss": 1.7649,
+      "step": 3387
+    },
+    {
+      "epoch": 1.553606237816764,
+      "grad_norm": 0.3189575672149658,
+      "learning_rate": 0.0004954186780037273,
+      "loss": 1.5074,
+      "step": 3388
+    },
+    {
+      "epoch": 1.5540649008141267,
+      "grad_norm": 0.2766447365283966,
+      "learning_rate": 0.0004951710464792736,
+      "loss": 1.6479,
+      "step": 3389
+    },
+    {
+      "epoch": 1.5545235638114896,
+      "grad_norm": 0.35292649269104004,
+      "learning_rate": 0.0004949234161393974,
+      "loss": 1.7307,
+      "step": 3390
+    },
+    {
+      "epoch": 1.5549822268088522,
+      "grad_norm": 0.24963252246379852,
+      "learning_rate": 0.0004946757870448437,
+      "loss": 0.7278,
+      "step": 3391
+    },
+    {
+      "epoch": 1.5554408898062149,
+      "grad_norm": 0.17716290056705475,
+      "learning_rate": 0.0004944281592563577,
+      "loss": 0.7671,
+      "step": 3392
+    },
+    {
+      "epoch": 1.5558995528035777,
+      "grad_norm": 0.234075665473938,
+      "learning_rate": 0.0004941805328346845,
+      "loss": 1.2981,
+      "step": 3393
+    },
+    {
+      "epoch": 1.5563582158009401,
+      "grad_norm": 0.4235064685344696,
+      "learning_rate": 0.0004939329078405683,
+      "loss": 2.1018,
+      "step": 3394
+    },
+    {
+      "epoch": 1.556816878798303,
+      "grad_norm": 0.31396913528442383,
+      "learning_rate": 0.0004936852843347541,
+      "loss": 1.6549,
+      "step": 3395
+    },
+    {
+      "epoch": 1.5572755417956656,
+      "grad_norm": 0.275888592004776,
+      "learning_rate": 0.0004934376623779848,
+      "loss": 1.3459,
+      "step": 3396
+    },
+    {
+      "epoch": 1.5577342047930283,
+      "grad_norm": 0.36937257647514343,
+      "learning_rate": 0.0004931900420310042,
+      "loss": 2.2034,
+      "step": 3397
+    },
+    {
+      "epoch": 1.5581928677903911,
+      "grad_norm": 0.45881137251853943,
+      "learning_rate": 0.0004929424233545556,
+      "loss": 1.7052,
+      "step": 3398
+    },
+    {
+      "epoch": 1.5586515307877535,
+      "grad_norm": 0.26412880420684814,
+      "learning_rate": 0.0004926948064093811,
+      "loss": 1.5784,
+      "step": 3399
+    },
+    {
+      "epoch": 1.5591101937851164,
+      "grad_norm": 0.35005372762680054,
+      "learning_rate": 0.0004924471912562236,
+      "loss": 1.5,
+      "step": 3400
+    },
+    {
+      "epoch": 1.559568856782479,
+      "grad_norm": 0.27285388112068176,
+      "learning_rate": 0.0004921995779558246,
+      "loss": 0.7449,
+      "step": 3401
+    },
+    {
+      "epoch": 1.5600275197798417,
+      "grad_norm": 0.35796064138412476,
+      "learning_rate": 0.0004919519665689252,
+      "loss": 1.644,
+      "step": 3402
+    },
+    {
+      "epoch": 1.5604861827772045,
+      "grad_norm": 0.25865140557289124,
+      "learning_rate": 0.0004917043571562664,
+      "loss": 1.5549,
+      "step": 3403
+    },
+    {
+      "epoch": 1.5609448457745672,
+      "grad_norm": 0.3593021631240845,
+      "learning_rate": 0.0004914567497785885,
+      "loss": 1.1608,
+      "step": 3404
+    },
+    {
+      "epoch": 1.5614035087719298,
+      "grad_norm": 0.2613557279109955,
+      "learning_rate": 0.0004912091444966316,
+      "loss": 1.0942,
+      "step": 3405
+    },
+    {
+      "epoch": 1.5618621717692927,
+      "grad_norm": 0.20529672503471375,
+      "learning_rate": 0.0004909615413711351,
+      "loss": 1.1398,
+      "step": 3406
+    },
+    {
+      "epoch": 1.562320834766655,
+      "grad_norm": 0.3400242030620575,
+      "learning_rate": 0.0004907139404628375,
+      "loss": 1.6556,
+      "step": 3407
+    },
+    {
+      "epoch": 1.562779497764018,
+      "grad_norm": 0.19008702039718628,
+      "learning_rate": 0.0004904663418324772,
+      "loss": 0.3512,
+      "step": 3408
+    },
+    {
+      "epoch": 1.5632381607613806,
+      "grad_norm": 0.28319889307022095,
+      "learning_rate": 0.0004902187455407921,
+      "loss": 1.2299,
+      "step": 3409
+    },
+    {
+      "epoch": 1.5636968237587432,
+      "grad_norm": 0.19672320783138275,
+      "learning_rate": 0.0004899711516485192,
+      "loss": 1.1458,
+      "step": 3410
+    },
+    {
+      "epoch": 1.564155486756106,
+      "grad_norm": 0.38868579268455505,
+      "learning_rate": 0.0004897235602163952,
+      "loss": 1.691,
+      "step": 3411
+    },
+    {
+      "epoch": 1.5646141497534687,
+      "grad_norm": 0.33131128549575806,
+      "learning_rate": 0.0004894759713051561,
+      "loss": 1.6458,
+      "step": 3412
+    },
+    {
+      "epoch": 1.5650728127508313,
+      "grad_norm": 0.31250569224357605,
+      "learning_rate": 0.0004892283849755368,
+      "loss": 1.0369,
+      "step": 3413
+    },
+    {
+      "epoch": 1.5655314757481942,
+      "grad_norm": 0.3426419198513031,
+      "learning_rate": 0.0004889808012882725,
+      "loss": 1.7825,
+      "step": 3414
+    },
+    {
+      "epoch": 1.5659901387455566,
+      "grad_norm": 0.10699386149644852,
+      "learning_rate": 0.000488733220304097,
+      "loss": 0.5646,
+      "step": 3415
+    },
+    {
+      "epoch": 1.5664488017429194,
+      "grad_norm": 0.22663241624832153,
+      "learning_rate": 0.0004884856420837438,
+      "loss": 1.1749,
+      "step": 3416
+    },
+    {
+      "epoch": 1.566907464740282,
+      "grad_norm": 0.21530061960220337,
+      "learning_rate": 0.0004882380666879457,
+      "loss": 1.2376,
+      "step": 3417
+    },
+    {
+      "epoch": 1.5673661277376447,
+      "grad_norm": 0.4145222306251526,
+      "learning_rate": 0.0004879904941774347,
+      "loss": 1.5345,
+      "step": 3418
+    },
+    {
+      "epoch": 1.5678247907350076,
+      "grad_norm": 0.28199657797813416,
+      "learning_rate": 0.00048774292461294203,
+      "loss": 1.5359,
+      "step": 3419
+    },
+    {
+      "epoch": 1.56828345373237,
+      "grad_norm": 0.35698872804641724,
+      "learning_rate": 0.00048749535805519824,
+      "loss": 1.5125,
+      "step": 3420
+    },
+    {
+      "epoch": 1.5687421167297328,
+      "grad_norm": 0.08986271172761917,
+      "learning_rate": 0.00048724779456493333,
+      "loss": 0.5488,
+      "step": 3421
+    },
+    {
+      "epoch": 1.5692007797270955,
+      "grad_norm": 0.28814515471458435,
+      "learning_rate": 0.00048700023420287635,
+      "loss": 1.4119,
+      "step": 3422
+    },
+    {
+      "epoch": 1.5696594427244581,
+      "grad_norm": 0.33969786763191223,
+      "learning_rate": 0.0004867526770297558,
+      "loss": 2.0642,
+      "step": 3423
+    },
+    {
+      "epoch": 1.570118105721821,
+      "grad_norm": 0.18406681716442108,
+      "learning_rate": 0.00048650512310629895,
+      "loss": 0.7826,
+      "step": 3424
+    },
+    {
+      "epoch": 1.5705767687191836,
+      "grad_norm": 0.2164044976234436,
+      "learning_rate": 0.0004862575724932327,
+      "loss": 1.1195,
+      "step": 3425
+    },
+    {
+      "epoch": 1.5710354317165462,
+      "grad_norm": 0.3818514943122864,
+      "learning_rate": 0.00048601002525128304,
+      "loss": 1.6297,
+      "step": 3426
+    },
+    {
+      "epoch": 1.571494094713909,
+      "grad_norm": 0.27068087458610535,
+      "learning_rate": 0.000485762481441175,
+      "loss": 1.4949,
+      "step": 3427
+    },
+    {
+      "epoch": 1.5719527577112715,
+      "grad_norm": 0.2720354497432709,
+      "learning_rate": 0.000485514941123633,
+      "loss": 1.3561,
+      "step": 3428
+    },
+    {
+      "epoch": 1.5724114207086344,
+      "grad_norm": 0.273271769285202,
+      "learning_rate": 0.00048526740435938045,
+      "loss": 1.4606,
+      "step": 3429
+    },
+    {
+      "epoch": 1.572870083705997,
+      "grad_norm": 0.06212342157959938,
+      "learning_rate": 0.0004850198712091397,
+      "loss": 0.6363,
+      "step": 3430
+    },
+    {
+      "epoch": 1.5733287467033596,
+      "grad_norm": 0.17056238651275635,
+      "learning_rate": 0.0004847723417336326,
+      "loss": 0.4197,
+      "step": 3431
+    },
+    {
+      "epoch": 1.5737874097007225,
+      "grad_norm": 0.3277488648891449,
+      "learning_rate": 0.00048452481599357985,
+      "loss": 1.6985,
+      "step": 3432
+    },
+    {
+      "epoch": 1.574246072698085,
+      "grad_norm": 0.29063600301742554,
+      "learning_rate": 0.00048427729404970133,
+      "loss": 1.446,
+      "step": 3433
+    },
+    {
+      "epoch": 1.5747047356954478,
+      "grad_norm": 0.2664231061935425,
+      "learning_rate": 0.00048402977596271604,
+      "loss": 1.5747,
+      "step": 3434
+    },
+    {
+      "epoch": 1.5751633986928104,
+      "grad_norm": 0.28931164741516113,
+      "learning_rate": 0.0004837822617933417,
+      "loss": 1.275,
+      "step": 3435
+    },
+    {
+      "epoch": 1.575622061690173,
+      "grad_norm": 0.3391111195087433,
+      "learning_rate": 0.0004835347516022956,
+      "loss": 2.0002,
+      "step": 3436
+    },
+    {
+      "epoch": 1.576080724687536,
+      "grad_norm": 0.3661150336265564,
+      "learning_rate": 0.00048328724545029355,
+      "loss": 1.0988,
+      "step": 3437
+    },
+    {
+      "epoch": 1.5765393876848985,
+      "grad_norm": 0.1353694349527359,
+      "learning_rate": 0.00048303974339805074,
+      "loss": 0.6863,
+      "step": 3438
+    },
+    {
+      "epoch": 1.5769980506822612,
+      "grad_norm": 0.14725974202156067,
+      "learning_rate": 0.00048279224550628117,
+      "loss": 1.3112,
+      "step": 3439
+    },
+    {
+      "epoch": 1.577456713679624,
+      "grad_norm": 0.2503822147846222,
+      "learning_rate": 0.000482544751835698,
+      "loss": 1.1244,
+      "step": 3440
+    },
+    {
+      "epoch": 1.5779153766769864,
+      "grad_norm": 0.55783611536026,
+      "learning_rate": 0.0004822972624470128,
+      "loss": 1.9937,
+      "step": 3441
+    },
+    {
+      "epoch": 1.5783740396743493,
+      "grad_norm": 0.2923833429813385,
+      "learning_rate": 0.00048204977740093694,
+      "loss": 1.0448,
+      "step": 3442
+    },
+    {
+      "epoch": 1.578832702671712,
+      "grad_norm": 0.09032203257083893,
+      "learning_rate": 0.00048180229675817997,
+      "loss": 0.529,
+      "step": 3443
+    },
+    {
+      "epoch": 1.5792913656690746,
+      "grad_norm": 0.12156753242015839,
+      "learning_rate": 0.00048155482057945094,
+      "loss": 0.9733,
+      "step": 3444
+    },
+    {
+      "epoch": 1.5797500286664374,
+      "grad_norm": 0.3344416916370392,
+      "learning_rate": 0.00048130734892545737,
+      "loss": 1.7934,
+      "step": 3445
+    },
+    {
+      "epoch": 1.5802086916638,
+      "grad_norm": 0.35299161076545715,
+      "learning_rate": 0.00048105988185690596,
+      "loss": 1.5703,
+      "step": 3446
+    },
+    {
+      "epoch": 1.5806673546611627,
+      "grad_norm": 0.22280535101890564,
+      "learning_rate": 0.00048081241943450205,
+      "loss": 1.6415,
+      "step": 3447
+    },
+    {
+      "epoch": 1.5811260176585256,
+      "grad_norm": 0.34450089931488037,
+      "learning_rate": 0.00048056496171895,
+      "loss": 1.4304,
+      "step": 3448
+    },
+    {
+      "epoch": 1.581584680655888,
+      "grad_norm": 0.2558358311653137,
+      "learning_rate": 0.0004803175087709529,
+      "loss": 1.5538,
+      "step": 3449
+    },
+    {
+      "epoch": 1.5820433436532508,
+      "grad_norm": 0.2851111888885498,
+      "learning_rate": 0.00048007006065121287,
+      "loss": 1.8889,
+      "step": 3450
+    },
+    {
+      "epoch": 1.5825020066506135,
+      "grad_norm": 0.21149995923042297,
+      "learning_rate": 0.0004798226174204308,
+      "loss": 0.9199,
+      "step": 3451
+    },
+    {
+      "epoch": 1.582960669647976,
+      "grad_norm": 0.31764793395996094,
+      "learning_rate": 0.00047957517913930597,
+      "loss": 1.7222,
+      "step": 3452
+    },
+    {
+      "epoch": 1.583419332645339,
+      "grad_norm": 0.27936065196990967,
+      "learning_rate": 0.00047932774586853706,
+      "loss": 1.4536,
+      "step": 3453
+    },
+    {
+      "epoch": 1.5838779956427014,
+      "grad_norm": 0.35680124163627625,
+      "learning_rate": 0.0004790803176688211,
+      "loss": 1.7658,
+      "step": 3454
+    },
+    {
+      "epoch": 1.5843366586400642,
+      "grad_norm": 0.2759742736816406,
+      "learning_rate": 0.00047883289460085406,
+      "loss": 1.122,
+      "step": 3455
+    },
+    {
+      "epoch": 1.5847953216374269,
+      "grad_norm": 0.36273276805877686,
+      "learning_rate": 0.0004785854767253305,
+      "loss": 1.7786,
+      "step": 3456
+    },
+    {
+      "epoch": 1.5852539846347895,
+      "grad_norm": 0.2069203108549118,
+      "learning_rate": 0.00047833806410294417,
+      "loss": 0.7401,
+      "step": 3457
+    },
+    {
+      "epoch": 1.5857126476321524,
+      "grad_norm": 0.2959374487400055,
+      "learning_rate": 0.00047809065679438675,
+      "loss": 1.3228,
+      "step": 3458
+    },
+    {
+      "epoch": 1.586171310629515,
+      "grad_norm": 0.25699517130851746,
+      "learning_rate": 0.00047784325486034917,
+      "loss": 1.2399,
+      "step": 3459
+    },
+    {
+      "epoch": 1.5866299736268776,
+      "grad_norm": 0.22081950306892395,
+      "learning_rate": 0.00047759585836152095,
+      "loss": 1.1349,
+      "step": 3460
+    },
+    {
+      "epoch": 1.5870886366242405,
+      "grad_norm": 0.304353266954422,
+      "learning_rate": 0.0004773484673585901,
+      "loss": 1.9224,
+      "step": 3461
+    },
+    {
+      "epoch": 1.587547299621603,
+      "grad_norm": 0.3038237392902374,
+      "learning_rate": 0.00047710108191224387,
+      "loss": 1.6517,
+      "step": 3462
+    },
+    {
+      "epoch": 1.5880059626189658,
+      "grad_norm": 0.24518848955631256,
+      "learning_rate": 0.00047685370208316717,
+      "loss": 0.8,
+      "step": 3463
+    },
+    {
+      "epoch": 1.5884646256163284,
+      "grad_norm": 0.2411755472421646,
+      "learning_rate": 0.0004766063279320442,
+      "loss": 1.7107,
+      "step": 3464
+    },
+    {
+      "epoch": 1.588923288613691,
+      "grad_norm": 0.31453174352645874,
+      "learning_rate": 0.0004763589595195576,
+      "loss": 1.9206,
+      "step": 3465
+    },
+    {
+      "epoch": 1.589381951611054,
+      "grad_norm": 0.36513078212738037,
+      "learning_rate": 0.00047611159690638867,
+      "loss": 1.6099,
+      "step": 3466
+    },
+    {
+      "epoch": 1.5898406146084163,
+      "grad_norm": 0.2547532916069031,
+      "learning_rate": 0.00047586424015321735,
+      "loss": 1.4818,
+      "step": 3467
+    },
+    {
+      "epoch": 1.5902992776057792,
+      "grad_norm": 0.35492321848869324,
+      "learning_rate": 0.00047561688932072215,
+      "loss": 1.6364,
+      "step": 3468
+    },
+    {
+      "epoch": 1.5907579406031418,
+      "grad_norm": 0.2952052652835846,
+      "learning_rate": 0.00047536954446957957,
+      "loss": 1.6503,
+      "step": 3469
+    },
+    {
+      "epoch": 1.5912166036005044,
+      "grad_norm": 0.3726816773414612,
+      "learning_rate": 0.0004751222056604654,
+      "loss": 2.0603,
+      "step": 3470
+    },
+    {
+      "epoch": 1.5916752665978673,
+      "grad_norm": 0.33943477272987366,
+      "learning_rate": 0.0004748748729540534,
+      "loss": 1.558,
+      "step": 3471
+    },
+    {
+      "epoch": 1.59213392959523,
+      "grad_norm": 0.35351845622062683,
+      "learning_rate": 0.00047462754641101635,
+      "loss": 1.6615,
+      "step": 3472
+    },
+    {
+      "epoch": 1.5925925925925926,
+      "grad_norm": 0.40344950556755066,
+      "learning_rate": 0.00047438022609202536,
+      "loss": 1.6789,
+      "step": 3473
+    },
+    {
+      "epoch": 1.5930512555899554,
+      "grad_norm": 0.16976001858711243,
+      "learning_rate": 0.0004741329120577494,
+      "loss": 0.945,
+      "step": 3474
+    },
+    {
+      "epoch": 1.5935099185873178,
+      "grad_norm": 0.25279539823532104,
+      "learning_rate": 0.00047388560436885656,
+      "loss": 1.0529,
+      "step": 3475
+    },
+    {
+      "epoch": 1.5939685815846807,
+      "grad_norm": 0.22543933987617493,
+      "learning_rate": 0.0004736383030860132,
+      "loss": 1.2178,
+      "step": 3476
+    },
+    {
+      "epoch": 1.5944272445820433,
+      "grad_norm": 0.33394721150398254,
+      "learning_rate": 0.00047339100826988427,
+      "loss": 1.3074,
+      "step": 3477
+    },
+    {
+      "epoch": 1.594885907579406,
+      "grad_norm": 0.30631691217422485,
+      "learning_rate": 0.0004731437199811329,
+      "loss": 1.2521,
+      "step": 3478
+    },
+    {
+      "epoch": 1.5953445705767688,
+      "grad_norm": 0.20698705315589905,
+      "learning_rate": 0.0004728964382804209,
+      "loss": 1.071,
+      "step": 3479
+    },
+    {
+      "epoch": 1.5958032335741315,
+      "grad_norm": 0.5738452076911926,
+      "learning_rate": 0.00047264916322840774,
+      "loss": 1.965,
+      "step": 3480
+    },
+    {
+      "epoch": 1.596261896571494,
+      "grad_norm": 0.42928802967071533,
+      "learning_rate": 0.000472401894885752,
+      "loss": 1.6219,
+      "step": 3481
+    },
+    {
+      "epoch": 1.596720559568857,
+      "grad_norm": 0.30077821016311646,
+      "learning_rate": 0.00047215463331311047,
+      "loss": 1.8263,
+      "step": 3482
+    },
+    {
+      "epoch": 1.5971792225662194,
+      "grad_norm": 0.27442824840545654,
+      "learning_rate": 0.00047190737857113823,
+      "loss": 1.48,
+      "step": 3483
+    },
+    {
+      "epoch": 1.5976378855635822,
+      "grad_norm": 0.3932221531867981,
+      "learning_rate": 0.00047166013072048857,
+      "loss": 1.9976,
+      "step": 3484
+    },
+    {
+      "epoch": 1.5980965485609449,
+      "grad_norm": 0.30163654685020447,
+      "learning_rate": 0.0004714128898218133,
+      "loss": 1.4284,
+      "step": 3485
+    },
+    {
+      "epoch": 1.5985552115583075,
+      "grad_norm": 0.29906541109085083,
+      "learning_rate": 0.0004711656559357621,
+      "loss": 1.1145,
+      "step": 3486
+    },
+    {
+      "epoch": 1.5990138745556703,
+      "grad_norm": 0.2641544044017792,
+      "learning_rate": 0.0004709184291229835,
+      "loss": 1.3258,
+      "step": 3487
+    },
+    {
+      "epoch": 1.5994725375530328,
+      "grad_norm": 0.2394290268421173,
+      "learning_rate": 0.0004706712094441239,
+      "loss": 1.6715,
+      "step": 3488
+    },
+    {
+      "epoch": 1.5999312005503956,
+      "grad_norm": 0.33737921714782715,
+      "learning_rate": 0.0004704239969598281,
+      "loss": 1.3931,
+      "step": 3489
+    },
+    {
+      "epoch": 1.6003898635477583,
+      "grad_norm": 0.2610185444355011,
+      "learning_rate": 0.0004701767917307391,
+      "loss": 1.7404,
+      "step": 3490
+    },
+    {
+      "epoch": 1.6008485265451209,
+      "grad_norm": 0.22963584959506989,
+      "learning_rate": 0.00046992959381749816,
+      "loss": 1.2682,
+      "step": 3491
+    },
+    {
+      "epoch": 1.6013071895424837,
+      "grad_norm": 0.28074997663497925,
+      "learning_rate": 0.00046968240328074465,
+      "loss": 1.2201,
+      "step": 3492
+    },
+    {
+      "epoch": 1.6017658525398464,
+      "grad_norm": 0.29938602447509766,
+      "learning_rate": 0.00046943522018111616,
+      "loss": 1.4655,
+      "step": 3493
+    },
+    {
+      "epoch": 1.602224515537209,
+      "grad_norm": 0.2991687059402466,
+      "learning_rate": 0.0004691880445792486,
+      "loss": 2.1379,
+      "step": 3494
+    },
+    {
+      "epoch": 1.6026831785345719,
+      "grad_norm": 0.2304016351699829,
+      "learning_rate": 0.0004689408765357758,
+      "loss": 0.7851,
+      "step": 3495
+    },
+    {
+      "epoch": 1.6031418415319343,
+      "grad_norm": 0.3397028148174286,
+      "learning_rate": 0.0004686937161113301,
+      "loss": 1.3103,
+      "step": 3496
+    },
+    {
+      "epoch": 1.6036005045292971,
+      "grad_norm": 0.23544049263000488,
+      "learning_rate": 0.0004684465633665415,
+      "loss": 1.593,
+      "step": 3497
+    },
+    {
+      "epoch": 1.6040591675266598,
+      "grad_norm": 0.2129555195569992,
+      "learning_rate": 0.0004681994183620383,
+      "loss": 0.743,
+      "step": 3498
+    },
+    {
+      "epoch": 1.6045178305240224,
+      "grad_norm": 0.4064857065677643,
+      "learning_rate": 0.0004679522811584471,
+      "loss": 1.6642,
+      "step": 3499
+    },
+    {
+      "epoch": 1.6049764935213853,
+      "grad_norm": 0.1782543659210205,
+      "learning_rate": 0.0004677051518163925,
+      "loss": 0.8695,
+      "step": 3500
+    },
+    {
+      "epoch": 1.6054351565187477,
+      "grad_norm": 0.21518990397453308,
+      "learning_rate": 0.00046745803039649703,
+      "loss": 0.8586,
+      "step": 3501
+    },
+    {
+      "epoch": 1.6058938195161105,
+      "grad_norm": 0.3452260494232178,
+      "learning_rate": 0.0004672109169593813,
+      "loss": 1.7291,
+      "step": 3502
+    },
+    {
+      "epoch": 1.6063524825134732,
+      "grad_norm": 0.20058414340019226,
+      "learning_rate": 0.00046696381156566405,
+      "loss": 1.2682,
+      "step": 3503
+    },
+    {
+      "epoch": 1.6068111455108358,
+      "grad_norm": 0.2975291609764099,
+      "learning_rate": 0.00046671671427596194,
+      "loss": 1.2484,
+      "step": 3504
+    },
+    {
+      "epoch": 1.6072698085081987,
+      "grad_norm": 0.6816866397857666,
+      "learning_rate": 0.0004664696251508899,
+      "loss": 1.4316,
+      "step": 3505
+    },
+    {
+      "epoch": 1.6077284715055613,
+      "grad_norm": 0.27953004837036133,
+      "learning_rate": 0.00046622254425106053,
+      "loss": 1.7111,
+      "step": 3506
+    },
+    {
+      "epoch": 1.608187134502924,
+      "grad_norm": 0.34763795137405396,
+      "learning_rate": 0.0004659754716370848,
+      "loss": 1.1754,
+      "step": 3507
+    },
+    {
+      "epoch": 1.6086457975002868,
+      "grad_norm": 0.2617149353027344,
+      "learning_rate": 0.00046572840736957096,
+      "loss": 1.636,
+      "step": 3508
+    },
+    {
+      "epoch": 1.6091044604976492,
+      "grad_norm": 0.3082357943058014,
+      "learning_rate": 0.00046548135150912596,
+      "loss": 1.7413,
+      "step": 3509
+    },
+    {
+      "epoch": 1.609563123495012,
+      "grad_norm": 0.2684236764907837,
+      "learning_rate": 0.00046523430411635436,
+      "loss": 0.8466,
+      "step": 3510
+    },
+    {
+      "epoch": 1.6100217864923747,
+      "grad_norm": 0.2572011649608612,
+      "learning_rate": 0.00046498726525185866,
+      "loss": 1.6624,
+      "step": 3511
+    },
+    {
+      "epoch": 1.6104804494897373,
+      "grad_norm": 0.27827000617980957,
+      "learning_rate": 0.0004647402349762392,
+      "loss": 1.3497,
+      "step": 3512
+    },
+    {
+      "epoch": 1.6109391124871002,
+      "grad_norm": 0.2422563135623932,
+      "learning_rate": 0.00046449321335009444,
+      "loss": 0.9238,
+      "step": 3513
+    },
+    {
+      "epoch": 1.6113977754844628,
+      "grad_norm": 0.2140582948923111,
+      "learning_rate": 0.0004642462004340203,
+      "loss": 1.4498,
+      "step": 3514
+    },
+    {
+      "epoch": 1.6118564384818255,
+      "grad_norm": 0.3195685148239136,
+      "learning_rate": 0.0004639991962886111,
+      "loss": 1.6652,
+      "step": 3515
+    },
+    {
+      "epoch": 1.6123151014791883,
+      "grad_norm": 0.40083667635917664,
+      "learning_rate": 0.0004637522009744586,
+      "loss": 1.7421,
+      "step": 3516
+    },
+    {
+      "epoch": 1.6127737644765507,
+      "grad_norm": 0.2781684398651123,
+      "learning_rate": 0.0004635052145521525,
+      "loss": 1.3621,
+      "step": 3517
+    },
+    {
+      "epoch": 1.6132324274739136,
+      "grad_norm": 0.3488773703575134,
+      "learning_rate": 0.00046325823708228054,
+      "loss": 1.6398,
+      "step": 3518
+    },
+    {
+      "epoch": 1.6136910904712762,
+      "grad_norm": 0.3113149106502533,
+      "learning_rate": 0.0004630112686254279,
+      "loss": 1.9727,
+      "step": 3519
+    },
+    {
+      "epoch": 1.6141497534686389,
+      "grad_norm": 0.3670024871826172,
+      "learning_rate": 0.0004627643092421777,
+      "loss": 1.7104,
+      "step": 3520
+    },
+    {
+      "epoch": 1.6146084164660017,
+      "grad_norm": 0.2058316022157669,
+      "learning_rate": 0.00046251735899311097,
+      "loss": 1.31,
+      "step": 3521
+    },
+    {
+      "epoch": 1.6150670794633641,
+      "grad_norm": 0.2428571581840515,
+      "learning_rate": 0.0004622704179388063,
+      "loss": 1.3736,
+      "step": 3522
+    },
+    {
+      "epoch": 1.615525742460727,
+      "grad_norm": 0.4996251165866852,
+      "learning_rate": 0.00046202348613984007,
+      "loss": 1.9026,
+      "step": 3523
+    },
+    {
+      "epoch": 1.6159844054580896,
+      "grad_norm": 0.33074381947517395,
+      "learning_rate": 0.0004617765636567869,
+      "loss": 1.9189,
+      "step": 3524
+    },
+    {
+      "epoch": 1.6164430684554523,
+      "grad_norm": 0.2520156800746918,
+      "learning_rate": 0.00046152965055021794,
+      "loss": 0.8857,
+      "step": 3525
+    },
+    {
+      "epoch": 1.6169017314528151,
+      "grad_norm": 0.28665515780448914,
+      "learning_rate": 0.00046128274688070315,
+      "loss": 1.8541,
+      "step": 3526
+    },
+    {
+      "epoch": 1.6173603944501778,
+      "grad_norm": 0.18662500381469727,
+      "learning_rate": 0.0004610358527088097,
+      "loss": 0.7944,
+      "step": 3527
+    },
+    {
+      "epoch": 1.6178190574475404,
+      "grad_norm": 0.3536602258682251,
+      "learning_rate": 0.00046078896809510246,
+      "loss": 1.3347,
+      "step": 3528
+    },
+    {
+      "epoch": 1.6182777204449033,
+      "grad_norm": 0.3583596348762512,
+      "learning_rate": 0.00046054209310014433,
+      "loss": 1.6628,
+      "step": 3529
+    },
+    {
+      "epoch": 1.6187363834422657,
+      "grad_norm": 0.6140498518943787,
+      "learning_rate": 0.0004602952277844951,
+      "loss": 2.1606,
+      "step": 3530
+    },
+    {
+      "epoch": 1.6191950464396285,
+      "grad_norm": 1.5802074670791626,
+      "learning_rate": 0.00046004837220871276,
+      "loss": 1.8225,
+      "step": 3531
+    },
+    {
+      "epoch": 1.6196537094369912,
+      "grad_norm": 0.33315399289131165,
+      "learning_rate": 0.0004598015264333528,
+      "loss": 1.4759,
+      "step": 3532
+    },
+    {
+      "epoch": 1.6201123724343538,
+      "grad_norm": 0.24571913480758667,
+      "learning_rate": 0.00045955469051896816,
+      "loss": 1.5706,
+      "step": 3533
+    },
+    {
+      "epoch": 1.6205710354317167,
+      "grad_norm": 0.33706846833229065,
+      "learning_rate": 0.0004593078645261096,
+      "loss": 1.7424,
+      "step": 3534
+    },
+    {
+      "epoch": 1.621029698429079,
+      "grad_norm": 0.29877063632011414,
+      "learning_rate": 0.0004590610485153255,
+      "loss": 1.397,
+      "step": 3535
+    },
+    {
+      "epoch": 1.621488361426442,
+      "grad_norm": 0.33204445242881775,
+      "learning_rate": 0.00045881424254716127,
+      "loss": 1.1624,
+      "step": 3536
+    },
+    {
+      "epoch": 1.6219470244238046,
+      "grad_norm": 0.2581148147583008,
+      "learning_rate": 0.00045856744668216023,
+      "loss": 1.7374,
+      "step": 3537
+    },
+    {
+      "epoch": 1.6224056874211672,
+      "grad_norm": 0.33581775426864624,
+      "learning_rate": 0.00045832066098086316,
+      "loss": 1.9,
+      "step": 3538
+    },
+    {
+      "epoch": 1.62286435041853,
+      "grad_norm": 0.1745557337999344,
+      "learning_rate": 0.00045807388550380855,
+      "loss": 0.8279,
+      "step": 3539
+    },
+    {
+      "epoch": 1.6233230134158927,
+      "grad_norm": 0.31032976508140564,
+      "learning_rate": 0.00045782712031153223,
+      "loss": 1.6848,
+      "step": 3540
+    },
+    {
+      "epoch": 1.6237816764132553,
+      "grad_norm": 0.1794162392616272,
+      "learning_rate": 0.0004575803654645675,
+      "loss": 0.5406,
+      "step": 3541
+    },
+    {
+      "epoch": 1.6242403394106182,
+      "grad_norm": 0.19714608788490295,
+      "learning_rate": 0.00045733362102344483,
+      "loss": 1.3489,
+      "step": 3542
+    },
+    {
+      "epoch": 1.6246990024079806,
+      "grad_norm": 0.3311532139778137,
+      "learning_rate": 0.0004570868870486924,
+      "loss": 1.7387,
+      "step": 3543
+    },
+    {
+      "epoch": 1.6251576654053435,
+      "grad_norm": 0.3105451762676239,
+      "learning_rate": 0.00045684016360083614,
+      "loss": 1.4056,
+      "step": 3544
+    },
+    {
+      "epoch": 1.625616328402706,
+      "grad_norm": 0.14965184032917023,
+      "learning_rate": 0.00045659345074039893,
+      "loss": 0.7161,
+      "step": 3545
+    },
+    {
+      "epoch": 1.6260749914000687,
+      "grad_norm": 0.2359105944633484,
+      "learning_rate": 0.0004563467485279015,
+      "loss": 1.7198,
+      "step": 3546
+    },
+    {
+      "epoch": 1.6265336543974316,
+      "grad_norm": 0.3810403645038605,
+      "learning_rate": 0.00045610005702386114,
+      "loss": 1.8206,
+      "step": 3547
+    },
+    {
+      "epoch": 1.6269923173947942,
+      "grad_norm": 0.2406771034002304,
+      "learning_rate": 0.0004558533762887932,
+      "loss": 1.6469,
+      "step": 3548
+    },
+    {
+      "epoch": 1.6274509803921569,
+      "grad_norm": 0.29384976625442505,
+      "learning_rate": 0.00045560670638321055,
+      "loss": 1.3338,
+      "step": 3549
+    },
+    {
+      "epoch": 1.6279096433895197,
+      "grad_norm": 0.3127608597278595,
+      "learning_rate": 0.0004553600473676229,
+      "loss": 1.8749,
+      "step": 3550
+    },
+    {
+      "epoch": 1.6283683063868821,
+      "grad_norm": 0.24576979875564575,
+      "learning_rate": 0.0004551133993025374,
+      "loss": 0.7694,
+      "step": 3551
+    },
+    {
+      "epoch": 1.628826969384245,
+      "grad_norm": 0.315415620803833,
+      "learning_rate": 0.000454866762248459,
+      "loss": 1.6774,
+      "step": 3552
+    },
+    {
+      "epoch": 1.6292856323816076,
+      "grad_norm": 0.2339719533920288,
+      "learning_rate": 0.00045462013626588896,
+      "loss": 0.7991,
+      "step": 3553
+    },
+    {
+      "epoch": 1.6297442953789703,
+      "grad_norm": 0.2772282361984253,
+      "learning_rate": 0.0004543735214153267,
+      "loss": 1.3975,
+      "step": 3554
+    },
+    {
+      "epoch": 1.6302029583763331,
+      "grad_norm": 0.2241169661283493,
+      "learning_rate": 0.00045412691775726874,
+      "loss": 0.8122,
+      "step": 3555
+    },
+    {
+      "epoch": 1.6306616213736955,
+      "grad_norm": 0.10823415219783783,
+      "learning_rate": 0.0004538803253522086,
+      "loss": 0.9379,
+      "step": 3556
+    },
+    {
+      "epoch": 1.6311202843710584,
+      "grad_norm": 0.3783659040927887,
+      "learning_rate": 0.00045363374426063717,
+      "loss": 2.1521,
+      "step": 3557
+    },
+    {
+      "epoch": 1.631578947368421,
+      "grad_norm": 0.3071146309375763,
+      "learning_rate": 0.00045338717454304265,
+      "loss": 0.8087,
+      "step": 3558
+    },
+    {
+      "epoch": 1.6320376103657837,
+      "grad_norm": 0.2720809876918793,
+      "learning_rate": 0.0004531406162599102,
+      "loss": 1.369,
+      "step": 3559
+    },
+    {
+      "epoch": 1.6324962733631465,
+      "grad_norm": 0.21303652226924896,
+      "learning_rate": 0.0004528940694717225,
+      "loss": 1.4075,
+      "step": 3560
+    },
+    {
+      "epoch": 1.6329549363605091,
+      "grad_norm": 0.2808886766433716,
+      "learning_rate": 0.0004526475342389592,
+      "loss": 1.2719,
+      "step": 3561
+    },
+    {
+      "epoch": 1.6334135993578718,
+      "grad_norm": 0.21272645890712738,
+      "learning_rate": 0.0004524010106220972,
+      "loss": 1.0988,
+      "step": 3562
+    },
+    {
+      "epoch": 1.6338722623552346,
+      "grad_norm": 0.49326127767562866,
+      "learning_rate": 0.00045215449868161057,
+      "loss": 1.6325,
+      "step": 3563
+    },
+    {
+      "epoch": 1.634330925352597,
+      "grad_norm": 0.19443345069885254,
+      "learning_rate": 0.0004519079984779703,
+      "loss": 0.8355,
+      "step": 3564
+    },
+    {
+      "epoch": 1.63478958834996,
+      "grad_norm": 0.2636580169200897,
+      "learning_rate": 0.0004516615100716448,
+      "loss": 1.3528,
+      "step": 3565
+    },
+    {
+      "epoch": 1.6352482513473225,
+      "grad_norm": 0.2543386220932007,
+      "learning_rate": 0.0004514150335230994,
+      "loss": 1.15,
+      "step": 3566
+    },
+    {
+      "epoch": 1.6357069143446852,
+      "grad_norm": 0.18656910955905914,
+      "learning_rate": 0.0004511685688927966,
+      "loss": 1.1378,
+      "step": 3567
+    },
+    {
+      "epoch": 1.636165577342048,
+      "grad_norm": 0.2844998836517334,
+      "learning_rate": 0.0004509221162411959,
+      "loss": 1.1173,
+      "step": 3568
+    },
+    {
+      "epoch": 1.6366242403394105,
+      "grad_norm": 0.1812886893749237,
+      "learning_rate": 0.00045067567562875405,
+      "loss": 1.0365,
+      "step": 3569
+    },
+    {
+      "epoch": 1.6370829033367733,
+      "grad_norm": 0.20662900805473328,
+      "learning_rate": 0.0004504292471159246,
+      "loss": 1.2693,
+      "step": 3570
+    },
+    {
+      "epoch": 1.637541566334136,
+      "grad_norm": 0.2970770299434662,
+      "learning_rate": 0.00045018283076315817,
+      "loss": 1.7665,
+      "step": 3571
+    },
+    {
+      "epoch": 1.6380002293314986,
+      "grad_norm": 0.22958998382091522,
+      "learning_rate": 0.0004499364266309026,
+      "loss": 0.7569,
+      "step": 3572
+    },
+    {
+      "epoch": 1.6384588923288614,
+      "grad_norm": 0.22331257164478302,
+      "learning_rate": 0.0004496900347796025,
+      "loss": 1.5889,
+      "step": 3573
+    },
+    {
+      "epoch": 1.638917555326224,
+      "grad_norm": 0.2571808695793152,
+      "learning_rate": 0.00044944365526969983,
+      "loss": 0.8337,
+      "step": 3574
+    },
+    {
+      "epoch": 1.6393762183235867,
+      "grad_norm": 0.17887678742408752,
+      "learning_rate": 0.0004491972881616329,
+      "loss": 0.6978,
+      "step": 3575
+    },
+    {
+      "epoch": 1.6398348813209496,
+      "grad_norm": 0.24759553372859955,
+      "learning_rate": 0.00044895093351583743,
+      "loss": 1.2387,
+      "step": 3576
+    },
+    {
+      "epoch": 1.640293544318312,
+      "grad_norm": 0.3115570545196533,
+      "learning_rate": 0.0004487045913927461,
+      "loss": 1.6693,
+      "step": 3577
+    },
+    {
+      "epoch": 1.6407522073156748,
+      "grad_norm": 0.32540813088417053,
+      "learning_rate": 0.00044845826185278836,
+      "loss": 1.0903,
+      "step": 3578
+    },
+    {
+      "epoch": 1.6412108703130375,
+      "grad_norm": 0.08525694906711578,
+      "learning_rate": 0.0004482119449563906,
+      "loss": 0.5742,
+      "step": 3579
+    },
+    {
+      "epoch": 1.6416695333104,
+      "grad_norm": 0.12443814426660538,
+      "learning_rate": 0.0004479656407639763,
+      "loss": 1.0961,
+      "step": 3580
+    },
+    {
+      "epoch": 1.642128196307763,
+      "grad_norm": 0.2620854079723358,
+      "learning_rate": 0.00044771934933596544,
+      "loss": 1.0662,
+      "step": 3581
+    },
+    {
+      "epoch": 1.6425868593051256,
+      "grad_norm": 0.20561723411083221,
+      "learning_rate": 0.00044747307073277504,
+      "loss": 1.4257,
+      "step": 3582
+    },
+    {
+      "epoch": 1.6430455223024882,
+      "grad_norm": 0.42892375588417053,
+      "learning_rate": 0.00044722680501481924,
+      "loss": 1.651,
+      "step": 3583
+    },
+    {
+      "epoch": 1.643504185299851,
+      "grad_norm": 0.18103362619876862,
+      "learning_rate": 0.00044698055224250854,
+      "loss": 1.4236,
+      "step": 3584
+    },
+    {
+      "epoch": 1.6439628482972135,
+      "grad_norm": 0.3172457218170166,
+      "learning_rate": 0.0004467343124762509,
+      "loss": 1.7086,
+      "step": 3585
+    },
+    {
+      "epoch": 1.6444215112945764,
+      "grad_norm": 0.3391329348087311,
+      "learning_rate": 0.0004464880857764503,
+      "loss": 1.2327,
+      "step": 3586
+    },
+    {
+      "epoch": 1.644880174291939,
+      "grad_norm": 0.2603316307067871,
+      "learning_rate": 0.00044624187220350815,
+      "loss": 1.8528,
+      "step": 3587
+    },
+    {
+      "epoch": 1.6453388372893016,
+      "grad_norm": 0.2531551122665405,
+      "learning_rate": 0.0004459956718178224,
+      "loss": 1.1763,
+      "step": 3588
+    },
+    {
+      "epoch": 1.6457975002866645,
+      "grad_norm": 0.3372352719306946,
+      "learning_rate": 0.00044574948467978763,
+      "loss": 1.3344,
+      "step": 3589
+    },
+    {
+      "epoch": 1.646256163284027,
+      "grad_norm": 0.2342667281627655,
+      "learning_rate": 0.0004455033108497954,
+      "loss": 1.1017,
+      "step": 3590
+    },
+    {
+      "epoch": 1.6467148262813898,
+      "grad_norm": 0.21960115432739258,
+      "learning_rate": 0.0004452571503882342,
+      "loss": 0.8768,
+      "step": 3591
+    },
+    {
+      "epoch": 1.6471734892787524,
+      "grad_norm": 0.24046221375465393,
+      "learning_rate": 0.0004450110033554886,
+      "loss": 1.0574,
+      "step": 3592
+    },
+    {
+      "epoch": 1.647632152276115,
+      "grad_norm": 0.22882205247879028,
+      "learning_rate": 0.00044476486981194035,
+      "loss": 1.2656,
+      "step": 3593
+    },
+    {
+      "epoch": 1.648090815273478,
+      "grad_norm": 0.21356913447380066,
+      "learning_rate": 0.0004445187498179678,
+      "loss": 0.9254,
+      "step": 3594
+    },
+    {
+      "epoch": 1.6485494782708405,
+      "grad_norm": 0.2324090301990509,
+      "learning_rate": 0.00044427264343394583,
+      "loss": 1.6883,
+      "step": 3595
+    },
+    {
+      "epoch": 1.6490081412682032,
+      "grad_norm": 0.19994021952152252,
+      "learning_rate": 0.0004440265507202464,
+      "loss": 0.803,
+      "step": 3596
+    },
+    {
+      "epoch": 1.649466804265566,
+      "grad_norm": 0.22320851683616638,
+      "learning_rate": 0.0004437804717372378,
+      "loss": 1.3083,
+      "step": 3597
+    },
+    {
+      "epoch": 1.6499254672629284,
+      "grad_norm": 0.3010443150997162,
+      "learning_rate": 0.0004435344065452847,
+      "loss": 1.4539,
+      "step": 3598
+    },
+    {
+      "epoch": 1.6503841302602913,
+      "grad_norm": 0.27338162064552307,
+      "learning_rate": 0.0004432883552047488,
+      "loss": 1.2612,
+      "step": 3599
+    },
+    {
+      "epoch": 1.650842793257654,
+      "grad_norm": 0.2307090312242508,
+      "learning_rate": 0.0004430423177759882,
+      "loss": 0.721,
+      "step": 3600
+    },
+    {
+      "epoch": 1.6513014562550166,
+      "grad_norm": 0.26010721921920776,
+      "learning_rate": 0.0004427962943193578,
+      "loss": 1.3625,
+      "step": 3601
+    },
+    {
+      "epoch": 1.6517601192523794,
+      "grad_norm": 0.302643746137619,
+      "learning_rate": 0.0004425502848952091,
+      "loss": 1.3484,
+      "step": 3602
+    },
+    {
+      "epoch": 1.652218782249742,
+      "grad_norm": 0.32459530234336853,
+      "learning_rate": 0.0004423042895638895,
+      "loss": 1.6387,
+      "step": 3603
+    },
+    {
+      "epoch": 1.6526774452471047,
+      "grad_norm": 0.1907692402601242,
+      "learning_rate": 0.0004420583083857437,
+      "loss": 0.7533,
+      "step": 3604
+    },
+    {
+      "epoch": 1.6531361082444673,
+      "grad_norm": 0.26839619874954224,
+      "learning_rate": 0.00044181234142111255,
+      "loss": 1.3954,
+      "step": 3605
+    },
+    {
+      "epoch": 1.65359477124183,
+      "grad_norm": 0.09783615916967392,
+      "learning_rate": 0.0004415663887303337,
+      "loss": 0.8584,
+      "step": 3606
+    },
+    {
+      "epoch": 1.6540534342391928,
+      "grad_norm": 0.42067283391952515,
+      "learning_rate": 0.00044132045037374094,
+      "loss": 1.8727,
+      "step": 3607
+    },
+    {
+      "epoch": 1.6545120972365555,
+      "grad_norm": 0.32053786516189575,
+      "learning_rate": 0.00044107452641166514,
+      "loss": 1.4919,
+      "step": 3608
+    },
+    {
+      "epoch": 1.654970760233918,
+      "grad_norm": 0.2961629331111908,
+      "learning_rate": 0.0004408286169044326,
+      "loss": 1.5599,
+      "step": 3609
+    },
+    {
+      "epoch": 1.655429423231281,
+      "grad_norm": 0.16102895140647888,
+      "learning_rate": 0.0004405827219123669,
+      "loss": 0.568,
+      "step": 3610
+    },
+    {
+      "epoch": 1.6558880862286434,
+      "grad_norm": 0.2084360271692276,
+      "learning_rate": 0.0004403368414957881,
+      "loss": 1.0012,
+      "step": 3611
+    },
+    {
+      "epoch": 1.6563467492260062,
+      "grad_norm": 0.222111776471138,
+      "learning_rate": 0.00044009097571501217,
+      "loss": 1.2141,
+      "step": 3612
+    },
+    {
+      "epoch": 1.6568054122233689,
+      "grad_norm": 0.3039361536502838,
+      "learning_rate": 0.0004398451246303521,
+      "loss": 1.1281,
+      "step": 3613
+    },
+    {
+      "epoch": 1.6572640752207315,
+      "grad_norm": 0.19220460951328278,
+      "learning_rate": 0.00043959928830211655,
+      "loss": 1.2178,
+      "step": 3614
+    },
+    {
+      "epoch": 1.6577227382180943,
+      "grad_norm": 0.23995845019817352,
+      "learning_rate": 0.0004393534667906109,
+      "loss": 0.944,
+      "step": 3615
+    },
+    {
+      "epoch": 1.658181401215457,
+      "grad_norm": 0.2802288830280304,
+      "learning_rate": 0.00043910766015613727,
+      "loss": 0.8565,
+      "step": 3616
+    },
+    {
+      "epoch": 1.6586400642128196,
+      "grad_norm": 0.17965088784694672,
+      "learning_rate": 0.00043886186845899366,
+      "loss": 0.9316,
+      "step": 3617
+    },
+    {
+      "epoch": 1.6590987272101825,
+      "grad_norm": 0.33390891551971436,
+      "learning_rate": 0.0004386160917594746,
+      "loss": 1.7971,
+      "step": 3618
+    },
+    {
+      "epoch": 1.659557390207545,
+      "grad_norm": 0.19103044271469116,
+      "learning_rate": 0.00043837033011787097,
+      "loss": 0.9313,
+      "step": 3619
+    },
+    {
+      "epoch": 1.6600160532049077,
+      "grad_norm": 0.23847994208335876,
+      "learning_rate": 0.00043812458359446943,
+      "loss": 1.623,
+      "step": 3620
+    },
+    {
+      "epoch": 1.6604747162022704,
+      "grad_norm": 0.26979711651802063,
+      "learning_rate": 0.0004378788522495538,
+      "loss": 1.0571,
+      "step": 3621
+    },
+    {
+      "epoch": 1.660933379199633,
+      "grad_norm": 0.2273758053779602,
+      "learning_rate": 0.0004376331361434036,
+      "loss": 1.2567,
+      "step": 3622
+    },
+    {
+      "epoch": 1.6613920421969959,
+      "grad_norm": 0.3926750123500824,
+      "learning_rate": 0.00043738743533629486,
+      "loss": 2.0264,
+      "step": 3623
+    },
+    {
+      "epoch": 1.6618507051943583,
+      "grad_norm": 0.3111841380596161,
+      "learning_rate": 0.00043714174988849965,
+      "loss": 1.1513,
+      "step": 3624
+    },
+    {
+      "epoch": 1.6623093681917211,
+      "grad_norm": 0.12631827592849731,
+      "learning_rate": 0.0004368960798602865,
+      "loss": 0.8732,
+      "step": 3625
+    },
+    {
+      "epoch": 1.6627680311890838,
+      "grad_norm": 0.18259353935718536,
+      "learning_rate": 0.0004366504253119199,
+      "loss": 0.5492,
+      "step": 3626
+    },
+    {
+      "epoch": 1.6632266941864464,
+      "grad_norm": 0.27222898602485657,
+      "learning_rate": 0.00043640478630366074,
+      "loss": 1.4242,
+      "step": 3627
+    },
+    {
+      "epoch": 1.6636853571838093,
+      "grad_norm": 0.31462812423706055,
+      "learning_rate": 0.0004361591628957661,
+      "loss": 1.7146,
+      "step": 3628
+    },
+    {
+      "epoch": 1.664144020181172,
+      "grad_norm": 0.28236204385757446,
+      "learning_rate": 0.00043591355514848904,
+      "loss": 1.4069,
+      "step": 3629
+    },
+    {
+      "epoch": 1.6646026831785345,
+      "grad_norm": 0.23949186503887177,
+      "learning_rate": 0.0004356679631220791,
+      "loss": 0.7886,
+      "step": 3630
+    },
+    {
+      "epoch": 1.6650613461758974,
+      "grad_norm": 0.06866566091775894,
+      "learning_rate": 0.00043542238687678157,
+      "loss": 0.4383,
+      "step": 3631
+    },
+    {
+      "epoch": 1.6655200091732598,
+      "grad_norm": 0.2557963728904724,
+      "learning_rate": 0.00043517682647283815,
+      "loss": 1.4226,
+      "step": 3632
+    },
+    {
+      "epoch": 1.6659786721706227,
+      "grad_norm": 0.19061607122421265,
+      "learning_rate": 0.00043493128197048657,
+      "loss": 1.2208,
+      "step": 3633
+    },
+    {
+      "epoch": 1.6664373351679853,
+      "grad_norm": 0.3818952143192291,
+      "learning_rate": 0.00043468575342996076,
+      "loss": 2.0298,
+      "step": 3634
+    },
+    {
+      "epoch": 1.666895998165348,
+      "grad_norm": 0.3634778559207916,
+      "learning_rate": 0.0004344402409114906,
+      "loss": 1.5644,
+      "step": 3635
+    },
+    {
+      "epoch": 1.6673546611627108,
+      "grad_norm": 0.17684818804264069,
+      "learning_rate": 0.00043419474447530204,
+      "loss": 1.0081,
+      "step": 3636
+    },
+    {
+      "epoch": 1.6678133241600734,
+      "grad_norm": 0.2824413478374481,
+      "learning_rate": 0.0004339492641816171,
+      "loss": 1.5297,
+      "step": 3637
+    },
+    {
+      "epoch": 1.668271987157436,
+      "grad_norm": 0.34963420033454895,
+      "learning_rate": 0.00043370380009065396,
+      "loss": 1.3151,
+      "step": 3638
+    },
+    {
+      "epoch": 1.6687306501547987,
+      "grad_norm": 0.1694229692220688,
+      "learning_rate": 0.00043345835226262663,
+      "loss": 1.2376,
+      "step": 3639
+    },
+    {
+      "epoch": 1.6691893131521613,
+      "grad_norm": 0.26671671867370605,
+      "learning_rate": 0.00043321292075774526,
+      "loss": 1.6826,
+      "step": 3640
+    },
+    {
+      "epoch": 1.6696479761495242,
+      "grad_norm": 0.3384977877140045,
+      "learning_rate": 0.00043296750563621614,
+      "loss": 1.525,
+      "step": 3641
+    },
+    {
+      "epoch": 1.6701066391468868,
+      "grad_norm": 0.2929702699184418,
+      "learning_rate": 0.0004327221069582411,
+      "loss": 2.0645,
+      "step": 3642
+    },
+    {
+      "epoch": 1.6705653021442495,
+      "grad_norm": 0.3001018762588501,
+      "learning_rate": 0.0004324767247840183,
+      "loss": 1.6188,
+      "step": 3643
+    },
+    {
+      "epoch": 1.6710239651416123,
+      "grad_norm": 0.2780381441116333,
+      "learning_rate": 0.0004322313591737418,
+      "loss": 1.1631,
+      "step": 3644
+    },
+    {
+      "epoch": 1.6714826281389747,
+      "grad_norm": 0.19090348482131958,
+      "learning_rate": 0.00043198601018760145,
+      "loss": 0.9705,
+      "step": 3645
+    },
+    {
+      "epoch": 1.6719412911363376,
+      "grad_norm": 0.17542727291584015,
+      "learning_rate": 0.0004317406778857833,
+      "loss": 1.0067,
+      "step": 3646
+    },
+    {
+      "epoch": 1.6723999541337002,
+      "grad_norm": 0.3193509876728058,
+      "learning_rate": 0.00043149536232846915,
+      "loss": 2.0522,
+      "step": 3647
+    },
+    {
+      "epoch": 1.6728586171310629,
+      "grad_norm": 0.262650728225708,
+      "learning_rate": 0.00043125006357583643,
+      "loss": 1.2916,
+      "step": 3648
+    },
+    {
+      "epoch": 1.6733172801284257,
+      "grad_norm": 0.2200448364019394,
+      "learning_rate": 0.0004310047816880588,
+      "loss": 1.0105,
+      "step": 3649
+    },
+    {
+      "epoch": 1.6737759431257884,
+      "grad_norm": 0.31807059049606323,
+      "learning_rate": 0.00043075951672530573,
+      "loss": 1.375,
+      "step": 3650
+    },
+    {
+      "epoch": 1.674234606123151,
+      "grad_norm": 0.23543456196784973,
+      "learning_rate": 0.0004305142687477425,
+      "loss": 1.4635,
+      "step": 3651
+    },
+    {
+      "epoch": 1.6746932691205139,
+      "grad_norm": 0.7807444930076599,
+      "learning_rate": 0.00043026903781553016,
+      "loss": 0.9573,
+      "step": 3652
+    },
+    {
+      "epoch": 1.6751519321178763,
+      "grad_norm": 0.18116839230060577,
+      "learning_rate": 0.0004300238239888256,
+      "loss": 1.1865,
+      "step": 3653
+    },
+    {
+      "epoch": 1.6756105951152391,
+      "grad_norm": 0.32931381464004517,
+      "learning_rate": 0.00042977862732778154,
+      "loss": 1.6169,
+      "step": 3654
+    },
+    {
+      "epoch": 1.6760692581126018,
+      "grad_norm": 0.32929980754852295,
+      "learning_rate": 0.0004295334478925466,
+      "loss": 1.5129,
+      "step": 3655
+    },
+    {
+      "epoch": 1.6765279211099644,
+      "grad_norm": 0.26823946833610535,
+      "learning_rate": 0.0004292882857432649,
+      "loss": 1.6238,
+      "step": 3656
+    },
+    {
+      "epoch": 1.6769865841073273,
+      "grad_norm": 0.3574983775615692,
+      "learning_rate": 0.00042904314094007655,
+      "loss": 2.0161,
+      "step": 3657
+    },
+    {
+      "epoch": 1.6774452471046897,
+      "grad_norm": 0.45830219984054565,
+      "learning_rate": 0.0004287980135431175,
+      "loss": 1.4208,
+      "step": 3658
+    },
+    {
+      "epoch": 1.6779039101020525,
+      "grad_norm": 0.3208571672439575,
+      "learning_rate": 0.000428552903612519,
+      "loss": 1.8867,
+      "step": 3659
+    },
+    {
+      "epoch": 1.6783625730994152,
+      "grad_norm": 0.11624974012374878,
+      "learning_rate": 0.00042830781120840845,
+      "loss": 1.0681,
+      "step": 3660
+    },
+    {
+      "epoch": 1.6788212360967778,
+      "grad_norm": 0.2735399901866913,
+      "learning_rate": 0.0004280627363909087,
+      "loss": 0.9048,
+      "step": 3661
+    },
+    {
+      "epoch": 1.6792798990941407,
+      "grad_norm": 0.31739717721939087,
+      "learning_rate": 0.0004278176792201383,
+      "loss": 1.7313,
+      "step": 3662
+    },
+    {
+      "epoch": 1.6797385620915033,
+      "grad_norm": 0.3343389332294464,
+      "learning_rate": 0.00042757263975621174,
+      "loss": 1.8636,
+      "step": 3663
+    },
+    {
+      "epoch": 1.680197225088866,
+      "grad_norm": 0.35586023330688477,
+      "learning_rate": 0.0004273276180592391,
+      "loss": 1.7568,
+      "step": 3664
+    },
+    {
+      "epoch": 1.6806558880862288,
+      "grad_norm": 0.31270912289619446,
+      "learning_rate": 0.0004270826141893256,
+      "loss": 1.6099,
+      "step": 3665
+    },
+    {
+      "epoch": 1.6811145510835912,
+      "grad_norm": 0.4064270555973053,
+      "learning_rate": 0.0004268376282065725,
+      "loss": 1.5682,
+      "step": 3666
+    },
+    {
+      "epoch": 1.681573214080954,
+      "grad_norm": 0.24555212259292603,
+      "learning_rate": 0.00042659266017107666,
+      "loss": 0.9243,
+      "step": 3667
+    },
+    {
+      "epoch": 1.6820318770783167,
+      "grad_norm": 0.37791720032691956,
+      "learning_rate": 0.0004263477101429307,
+      "loss": 1.3367,
+      "step": 3668
+    },
+    {
+      "epoch": 1.6824905400756793,
+      "grad_norm": 0.25306499004364014,
+      "learning_rate": 0.0004261027781822227,
+      "loss": 1.5412,
+      "step": 3669
+    },
+    {
+      "epoch": 1.6829492030730422,
+      "grad_norm": 0.23409046232700348,
+      "learning_rate": 0.00042585786434903584,
+      "loss": 0.8784,
+      "step": 3670
+    },
+    {
+      "epoch": 1.6834078660704048,
+      "grad_norm": 0.23634447157382965,
+      "learning_rate": 0.00042561296870344945,
+      "loss": 1.4553,
+      "step": 3671
+    },
+    {
+      "epoch": 1.6838665290677675,
+      "grad_norm": 0.29898911714553833,
+      "learning_rate": 0.0004253680913055381,
+      "loss": 1.2293,
+      "step": 3672
+    },
+    {
+      "epoch": 1.68432519206513,
+      "grad_norm": 0.20857423543930054,
+      "learning_rate": 0.00042512323221537206,
+      "loss": 1.5924,
+      "step": 3673
+    },
+    {
+      "epoch": 1.6847838550624927,
+      "grad_norm": 0.4438040852546692,
+      "learning_rate": 0.0004248783914930172,
+      "loss": 1.8789,
+      "step": 3674
+    },
+    {
+      "epoch": 1.6852425180598556,
+      "grad_norm": 0.33179306983947754,
+      "learning_rate": 0.0004246335691985347,
+      "loss": 1.343,
+      "step": 3675
+    },
+    {
+      "epoch": 1.6857011810572182,
+      "grad_norm": 0.3002280294895172,
+      "learning_rate": 0.0004243887653919809,
+      "loss": 1.6851,
+      "step": 3676
+    },
+    {
+      "epoch": 1.6861598440545809,
+      "grad_norm": 0.2567557692527771,
+      "learning_rate": 0.00042414398013340806,
+      "loss": 1.5004,
+      "step": 3677
+    },
+    {
+      "epoch": 1.6866185070519437,
+      "grad_norm": 0.28894615173339844,
+      "learning_rate": 0.00042389921348286386,
+      "loss": 1.1944,
+      "step": 3678
+    },
+    {
+      "epoch": 1.6870771700493061,
+      "grad_norm": 0.30856838822364807,
+      "learning_rate": 0.00042365446550039136,
+      "loss": 1.746,
+      "step": 3679
+    },
+    {
+      "epoch": 1.687535833046669,
+      "grad_norm": 0.24894581735134125,
+      "learning_rate": 0.0004234097362460292,
+      "loss": 0.8261,
+      "step": 3680
+    },
+    {
+      "epoch": 1.6879944960440316,
+      "grad_norm": 0.16843147575855255,
+      "learning_rate": 0.00042316502577981077,
+      "loss": 1.1577,
+      "step": 3681
+    },
+    {
+      "epoch": 1.6884531590413943,
+      "grad_norm": 0.37144291400909424,
+      "learning_rate": 0.00042292033416176534,
+      "loss": 1.601,
+      "step": 3682
+    },
+    {
+      "epoch": 1.6889118220387571,
+      "grad_norm": 0.28090983629226685,
+      "learning_rate": 0.0004226756614519179,
+      "loss": 1.4488,
+      "step": 3683
+    },
+    {
+      "epoch": 1.6893704850361198,
+      "grad_norm": 0.2437400072813034,
+      "learning_rate": 0.0004224310077102882,
+      "loss": 1.2493,
+      "step": 3684
+    },
+    {
+      "epoch": 1.6898291480334824,
+      "grad_norm": 0.25606828927993774,
+      "learning_rate": 0.0004221863729968917,
+      "loss": 1.2882,
+      "step": 3685
+    },
+    {
+      "epoch": 1.6902878110308452,
+      "grad_norm": 0.1753174066543579,
+      "learning_rate": 0.00042194175737173906,
+      "loss": 0.7293,
+      "step": 3686
+    },
+    {
+      "epoch": 1.6907464740282077,
+      "grad_norm": 0.2955300211906433,
+      "learning_rate": 0.0004216971608948359,
+      "loss": 1.5758,
+      "step": 3687
+    },
+    {
+      "epoch": 1.6912051370255705,
+      "grad_norm": 0.3101601004600525,
+      "learning_rate": 0.00042145258362618384,
+      "loss": 1.5141,
+      "step": 3688
+    },
+    {
+      "epoch": 1.6916638000229332,
+      "grad_norm": 0.3169883191585541,
+      "learning_rate": 0.00042120802562577945,
+      "loss": 1.5436,
+      "step": 3689
+    },
+    {
+      "epoch": 1.6921224630202958,
+      "grad_norm": 0.3298269510269165,
+      "learning_rate": 0.00042096348695361437,
+      "loss": 1.8237,
+      "step": 3690
+    },
+    {
+      "epoch": 1.6925811260176586,
+      "grad_norm": 0.3161281645298004,
+      "learning_rate": 0.0004207189676696759,
+      "loss": 1.5731,
+      "step": 3691
+    },
+    {
+      "epoch": 1.693039789015021,
+      "grad_norm": 0.3234587609767914,
+      "learning_rate": 0.0004204744678339464,
+      "loss": 1.3514,
+      "step": 3692
+    },
+    {
+      "epoch": 1.693498452012384,
+      "grad_norm": 0.19994625449180603,
+      "learning_rate": 0.0004202299875064033,
+      "loss": 0.9323,
+      "step": 3693
+    },
+    {
+      "epoch": 1.6939571150097466,
+      "grad_norm": 0.2861116826534271,
+      "learning_rate": 0.0004199855267470193,
+      "loss": 1.4347,
+      "step": 3694
+    },
+    {
+      "epoch": 1.6944157780071092,
+      "grad_norm": 0.32218021154403687,
+      "learning_rate": 0.00041974108561576264,
+      "loss": 1.743,
+      "step": 3695
+    },
+    {
+      "epoch": 1.694874441004472,
+      "grad_norm": 0.33372506499290466,
+      "learning_rate": 0.00041949666417259637,
+      "loss": 1.1621,
+      "step": 3696
+    },
+    {
+      "epoch": 1.6953331040018347,
+      "grad_norm": 0.37846240401268005,
+      "learning_rate": 0.0004192522624774789,
+      "loss": 1.1866,
+      "step": 3697
+    },
+    {
+      "epoch": 1.6957917669991973,
+      "grad_norm": 0.2000287026166916,
+      "learning_rate": 0.00041900788059036354,
+      "loss": 1.3976,
+      "step": 3698
+    },
+    {
+      "epoch": 1.6962504299965602,
+      "grad_norm": 0.37890589237213135,
+      "learning_rate": 0.0004187635185711991,
+      "loss": 1.7291,
+      "step": 3699
+    },
+    {
+      "epoch": 1.6967090929939226,
+      "grad_norm": 0.30530160665512085,
+      "learning_rate": 0.0004185191764799293,
+      "loss": 1.0033,
+      "step": 3700
+    },
+    {
+      "epoch": 1.6971677559912854,
+      "grad_norm": 0.24736268818378448,
+      "learning_rate": 0.00041827485437649306,
+      "loss": 1.6701,
+      "step": 3701
+    },
+    {
+      "epoch": 1.697626418988648,
+      "grad_norm": 0.2429756373167038,
+      "learning_rate": 0.00041803055232082423,
+      "loss": 0.8325,
+      "step": 3702
+    },
+    {
+      "epoch": 1.6980850819860107,
+      "grad_norm": 0.297656774520874,
+      "learning_rate": 0.00041778627037285205,
+      "loss": 1.4572,
+      "step": 3703
+    },
+    {
+      "epoch": 1.6985437449833736,
+      "grad_norm": 0.26800161600112915,
+      "learning_rate": 0.0004175420085925005,
+      "loss": 1.7028,
+      "step": 3704
+    },
+    {
+      "epoch": 1.6990024079807362,
+      "grad_norm": 0.27825766801834106,
+      "learning_rate": 0.0004172977670396887,
+      "loss": 1.6484,
+      "step": 3705
+    },
+    {
+      "epoch": 1.6994610709780988,
+      "grad_norm": 0.2302626520395279,
+      "learning_rate": 0.0004170535457743311,
+      "loss": 0.9787,
+      "step": 3706
+    },
+    {
+      "epoch": 1.6999197339754615,
+      "grad_norm": 0.22059178352355957,
+      "learning_rate": 0.0004168093448563367,
+      "loss": 0.9163,
+      "step": 3707
+    },
+    {
+      "epoch": 1.7003783969728241,
+      "grad_norm": 0.28476452827453613,
+      "learning_rate": 0.0004165651643456099,
+      "loss": 1.9269,
+      "step": 3708
+    },
+    {
+      "epoch": 1.700837059970187,
+      "grad_norm": 0.3191598057746887,
+      "learning_rate": 0.0004163210043020499,
+      "loss": 1.7429,
+      "step": 3709
+    },
+    {
+      "epoch": 1.7012957229675496,
+      "grad_norm": 0.5909731984138489,
+      "learning_rate": 0.00041607686478555076,
+      "loss": 1.2906,
+      "step": 3710
+    },
+    {
+      "epoch": 1.7017543859649122,
+      "grad_norm": 0.2239387482404709,
+      "learning_rate": 0.00041583274585600195,
+      "loss": 0.8284,
+      "step": 3711
+    },
+    {
+      "epoch": 1.702213048962275,
+      "grad_norm": 0.3220686912536621,
+      "learning_rate": 0.0004155886475732874,
+      "loss": 1.9756,
+      "step": 3712
+    },
+    {
+      "epoch": 1.7026717119596375,
+      "grad_norm": 0.23089122772216797,
+      "learning_rate": 0.0004153445699972862,
+      "loss": 1.3217,
+      "step": 3713
+    },
+    {
+      "epoch": 1.7031303749570004,
+      "grad_norm": 0.2528243064880371,
+      "learning_rate": 0.0004151005131878725,
+      "loss": 0.8455,
+      "step": 3714
+    },
+    {
+      "epoch": 1.703589037954363,
+      "grad_norm": 0.10433992743492126,
+      "learning_rate": 0.00041485647720491503,
+      "loss": 1.0346,
+      "step": 3715
+    },
+    {
+      "epoch": 1.7040477009517256,
+      "grad_norm": 0.26862263679504395,
+      "learning_rate": 0.0004146124621082775,
+      "loss": 1.2124,
+      "step": 3716
+    },
+    {
+      "epoch": 1.7045063639490885,
+      "grad_norm": 0.3269628882408142,
+      "learning_rate": 0.0004143684679578188,
+      "loss": 1.3434,
+      "step": 3717
+    },
+    {
+      "epoch": 1.7049650269464511,
+      "grad_norm": 0.2566255033016205,
+      "learning_rate": 0.00041412449481339233,
+      "loss": 1.7516,
+      "step": 3718
+    },
+    {
+      "epoch": 1.7054236899438138,
+      "grad_norm": 0.2328820526599884,
+      "learning_rate": 0.0004138805427348464,
+      "loss": 1.1428,
+      "step": 3719
+    },
+    {
+      "epoch": 1.7058823529411766,
+      "grad_norm": 0.41982701420783997,
+      "learning_rate": 0.0004136366117820245,
+      "loss": 1.6322,
+      "step": 3720
+    },
+    {
+      "epoch": 1.706341015938539,
+      "grad_norm": 0.2226077914237976,
+      "learning_rate": 0.00041339270201476425,
+      "loss": 1.2604,
+      "step": 3721
+    },
+    {
+      "epoch": 1.706799678935902,
+      "grad_norm": 0.2776189148426056,
+      "learning_rate": 0.0004131488134928987,
+      "loss": 1.2419,
+      "step": 3722
+    },
+    {
+      "epoch": 1.7072583419332645,
+      "grad_norm": 0.39088577032089233,
+      "learning_rate": 0.0004129049462762554,
+      "loss": 1.9712,
+      "step": 3723
+    },
+    {
+      "epoch": 1.7077170049306272,
+      "grad_norm": 0.28871458768844604,
+      "learning_rate": 0.0004126611004246568,
+      "loss": 1.5851,
+      "step": 3724
+    },
+    {
+      "epoch": 1.70817566792799,
+      "grad_norm": 0.2573470175266266,
+      "learning_rate": 0.00041241727599792015,
+      "loss": 1.0604,
+      "step": 3725
+    },
+    {
+      "epoch": 1.7086343309253524,
+      "grad_norm": 0.257318913936615,
+      "learning_rate": 0.00041217347305585707,
+      "loss": 1.5261,
+      "step": 3726
+    },
+    {
+      "epoch": 1.7090929939227153,
+      "grad_norm": 0.32521334290504456,
+      "learning_rate": 0.00041192969165827433,
+      "loss": 1.9292,
+      "step": 3727
+    },
+    {
+      "epoch": 1.709551656920078,
+      "grad_norm": 0.2845672369003296,
+      "learning_rate": 0.00041168593186497317,
+      "loss": 1.2908,
+      "step": 3728
+    },
+    {
+      "epoch": 1.7100103199174406,
+      "grad_norm": 0.3328210115432739,
+      "learning_rate": 0.00041144219373574976,
+      "loss": 2.0022,
+      "step": 3729
+    },
+    {
+      "epoch": 1.7104689829148034,
+      "grad_norm": 0.3225518763065338,
+      "learning_rate": 0.0004111984773303946,
+      "loss": 1.4616,
+      "step": 3730
+    },
+    {
+      "epoch": 1.710927645912166,
+      "grad_norm": 0.14123696088790894,
+      "learning_rate": 0.0004109547827086937,
+      "loss": 0.4822,
+      "step": 3731
+    },
+    {
+      "epoch": 1.7113863089095287,
+      "grad_norm": 0.08275225758552551,
+      "learning_rate": 0.00041071110993042627,
+      "loss": 0.8438,
+      "step": 3732
+    },
+    {
+      "epoch": 1.7118449719068916,
+      "grad_norm": 0.2467496544122696,
+      "learning_rate": 0.0004104674590553675,
+      "loss": 1.2802,
+      "step": 3733
+    },
+    {
+      "epoch": 1.712303634904254,
+      "grad_norm": 0.25316253304481506,
+      "learning_rate": 0.0004102238301432865,
+      "loss": 0.9283,
+      "step": 3734
+    },
+    {
+      "epoch": 1.7127622979016168,
+      "grad_norm": 0.23193414509296417,
+      "learning_rate": 0.00040998022325394723,
+      "loss": 1.4131,
+      "step": 3735
+    },
+    {
+      "epoch": 1.7132209608989795,
+      "grad_norm": 0.22135387361049652,
+      "learning_rate": 0.0004097366384471086,
+      "loss": 0.8179,
+      "step": 3736
+    },
+    {
+      "epoch": 1.713679623896342,
+      "grad_norm": 0.26501017808914185,
+      "learning_rate": 0.00040949307578252314,
+      "loss": 1.5214,
+      "step": 3737
+    },
+    {
+      "epoch": 1.714138286893705,
+      "grad_norm": 0.2112581878900528,
+      "learning_rate": 0.0004092495353199388,
+      "loss": 1.0725,
+      "step": 3738
+    },
+    {
+      "epoch": 1.7145969498910676,
+      "grad_norm": 0.21010689437389374,
+      "learning_rate": 0.0004090060171190977,
+      "loss": 0.6734,
+      "step": 3739
+    },
+    {
+      "epoch": 1.7150556128884302,
+      "grad_norm": 0.3008427321910858,
+      "learning_rate": 0.00040876252123973677,
+      "loss": 2.1113,
+      "step": 3740
+    },
+    {
+      "epoch": 1.7155142758857929,
+      "grad_norm": 0.33156007528305054,
+      "learning_rate": 0.00040851904774158725,
+      "loss": 1.8834,
+      "step": 3741
+    },
+    {
+      "epoch": 1.7159729388831555,
+      "grad_norm": 0.2164911925792694,
+      "learning_rate": 0.0004082755966843752,
+      "loss": 0.832,
+      "step": 3742
+    },
+    {
+      "epoch": 1.7164316018805184,
+      "grad_norm": 0.2621867060661316,
+      "learning_rate": 0.0004080321681278204,
+      "loss": 0.9056,
+      "step": 3743
+    },
+    {
+      "epoch": 1.716890264877881,
+      "grad_norm": 0.2789194881916046,
+      "learning_rate": 0.0004077887621316377,
+      "loss": 1.7653,
+      "step": 3744
+    },
+    {
+      "epoch": 1.7173489278752436,
+      "grad_norm": 0.27011173963546753,
+      "learning_rate": 0.0004075453787555367,
+      "loss": 1.3661,
+      "step": 3745
+    },
+    {
+      "epoch": 1.7178075908726065,
+      "grad_norm": 0.29901739954948425,
+      "learning_rate": 0.00040730201805922096,
+      "loss": 1.5548,
+      "step": 3746
+    },
+    {
+      "epoch": 1.718266253869969,
+      "grad_norm": 0.3142026662826538,
+      "learning_rate": 0.0004070586801023885,
+      "loss": 1.79,
+      "step": 3747
+    },
+    {
+      "epoch": 1.7187249168673318,
+      "grad_norm": 0.26456326246261597,
+      "learning_rate": 0.00040681536494473224,
+      "loss": 1.305,
+      "step": 3748
+    },
+    {
+      "epoch": 1.7191835798646944,
+      "grad_norm": 0.35430580377578735,
+      "learning_rate": 0.0004065720726459385,
+      "loss": 1.6096,
+      "step": 3749
+    },
+    {
+      "epoch": 1.719642242862057,
+      "grad_norm": 0.31563955545425415,
+      "learning_rate": 0.0004063288032656891,
+      "loss": 1.6702,
+      "step": 3750
+    },
+    {
+      "epoch": 1.7201009058594199,
+      "grad_norm": 0.31197500228881836,
+      "learning_rate": 0.00040608555686365966,
+      "loss": 0.9615,
+      "step": 3751
+    },
+    {
+      "epoch": 1.7205595688567825,
+      "grad_norm": 0.26841220259666443,
+      "learning_rate": 0.00040584233349952027,
+      "loss": 1.7595,
+      "step": 3752
+    },
+    {
+      "epoch": 1.7210182318541452,
+      "grad_norm": 0.3827781677246094,
+      "learning_rate": 0.0004055991332329356,
+      "loss": 1.309,
+      "step": 3753
+    },
+    {
+      "epoch": 1.721476894851508,
+      "grad_norm": 0.30011874437332153,
+      "learning_rate": 0.00040535595612356393,
+      "loss": 1.0917,
+      "step": 3754
+    },
+    {
+      "epoch": 1.7219355578488704,
+      "grad_norm": 0.23119381070137024,
+      "learning_rate": 0.00040511280223105866,
+      "loss": 0.9667,
+      "step": 3755
+    },
+    {
+      "epoch": 1.7223942208462333,
+      "grad_norm": 0.3517080545425415,
+      "learning_rate": 0.00040486967161506725,
+      "loss": 1.3582,
+      "step": 3756
+    },
+    {
+      "epoch": 1.722852883843596,
+      "grad_norm": 0.12287653982639313,
+      "learning_rate": 0.0004046265643352313,
+      "loss": 0.7969,
+      "step": 3757
+    },
+    {
+      "epoch": 1.7233115468409586,
+      "grad_norm": 0.22328488528728485,
+      "learning_rate": 0.0004043834804511868,
+      "loss": 1.2675,
+      "step": 3758
+    },
+    {
+      "epoch": 1.7237702098383214,
+      "grad_norm": 0.3529665768146515,
+      "learning_rate": 0.0004041404200225641,
+      "loss": 1.1346,
+      "step": 3759
+    },
+    {
+      "epoch": 1.7242288728356838,
+      "grad_norm": 0.283286452293396,
+      "learning_rate": 0.00040389738310898736,
+      "loss": 1.2228,
+      "step": 3760
+    },
+    {
+      "epoch": 1.7246875358330467,
+      "grad_norm": 0.181251659989357,
+      "learning_rate": 0.0004036543697700756,
+      "loss": 1.1888,
+      "step": 3761
+    },
+    {
+      "epoch": 1.7251461988304093,
+      "grad_norm": 0.2636159062385559,
+      "learning_rate": 0.0004034113800654415,
+      "loss": 0.6633,
+      "step": 3762
+    },
+    {
+      "epoch": 1.725604861827772,
+      "grad_norm": 0.2084595412015915,
+      "learning_rate": 0.0004031684140546924,
+      "loss": 1.3781,
+      "step": 3763
+    },
+    {
+      "epoch": 1.7260635248251348,
+      "grad_norm": 0.3345155715942383,
+      "learning_rate": 0.0004029254717974297,
+      "loss": 1.3012,
+      "step": 3764
+    },
+    {
+      "epoch": 1.7265221878224974,
+      "grad_norm": 0.3305032551288605,
+      "learning_rate": 0.00040268255335324844,
+      "loss": 1.6357,
+      "step": 3765
+    },
+    {
+      "epoch": 1.72698085081986,
+      "grad_norm": 0.33602216839790344,
+      "learning_rate": 0.0004024396587817386,
+      "loss": 1.7511,
+      "step": 3766
+    },
+    {
+      "epoch": 1.727439513817223,
+      "grad_norm": 0.3443238437175751,
+      "learning_rate": 0.00040219678814248396,
+      "loss": 1.7075,
+      "step": 3767
+    },
+    {
+      "epoch": 1.7278981768145854,
+      "grad_norm": 0.3635205924510956,
+      "learning_rate": 0.00040195394149506234,
+      "loss": 1.413,
+      "step": 3768
+    },
+    {
+      "epoch": 1.7283568398119482,
+      "grad_norm": 0.2028258591890335,
+      "learning_rate": 0.00040171111889904584,
+      "loss": 0.8682,
+      "step": 3769
+    },
+    {
+      "epoch": 1.7288155028093108,
+      "grad_norm": 0.26344314217567444,
+      "learning_rate": 0.0004014683204140006,
+      "loss": 1.6746,
+      "step": 3770
+    },
+    {
+      "epoch": 1.7292741658066735,
+      "grad_norm": 0.36593350768089294,
+      "learning_rate": 0.0004012255460994868,
+      "loss": 1.3928,
+      "step": 3771
+    },
+    {
+      "epoch": 1.7297328288040363,
+      "grad_norm": 0.2758760154247284,
+      "learning_rate": 0.0004009827960150587,
+      "loss": 0.9172,
+      "step": 3772
+    },
+    {
+      "epoch": 1.730191491801399,
+      "grad_norm": 0.348369836807251,
+      "learning_rate": 0.00040074007022026473,
+      "loss": 2.3096,
+      "step": 3773
+    },
+    {
+      "epoch": 1.7306501547987616,
+      "grad_norm": 0.3024252653121948,
+      "learning_rate": 0.0004004973687746472,
+      "loss": 1.2795,
+      "step": 3774
+    },
+    {
+      "epoch": 1.7311088177961242,
+      "grad_norm": 0.3365992307662964,
+      "learning_rate": 0.00040025469173774256,
+      "loss": 1.1777,
+      "step": 3775
+    },
+    {
+      "epoch": 1.7315674807934869,
+      "grad_norm": 0.2750336229801178,
+      "learning_rate": 0.0004000120391690814,
+      "loss": 1.3168,
+      "step": 3776
+    },
+    {
+      "epoch": 1.7320261437908497,
+      "grad_norm": 0.32935917377471924,
+      "learning_rate": 0.00039976941112818777,
+      "loss": 1.9238,
+      "step": 3777
+    },
+    {
+      "epoch": 1.7324848067882124,
+      "grad_norm": 0.3010925054550171,
+      "learning_rate": 0.00039952680767458036,
+      "loss": 1.0422,
+      "step": 3778
+    },
+    {
+      "epoch": 1.732943469785575,
+      "grad_norm": 0.29777830839157104,
+      "learning_rate": 0.0003992842288677715,
+      "loss": 1.7227,
+      "step": 3779
+    },
+    {
+      "epoch": 1.7334021327829379,
+      "grad_norm": 0.3348708748817444,
+      "learning_rate": 0.00039904167476726744,
+      "loss": 1.6201,
+      "step": 3780
+    },
+    {
+      "epoch": 1.7338607957803003,
+      "grad_norm": 0.2975271940231323,
+      "learning_rate": 0.00039879914543256863,
+      "loss": 1.2951,
+      "step": 3781
+    },
+    {
+      "epoch": 1.7343194587776631,
+      "grad_norm": 0.2370220124721527,
+      "learning_rate": 0.000398556640923169,
+      "loss": 0.8168,
+      "step": 3782
+    },
+    {
+      "epoch": 1.7347781217750258,
+      "grad_norm": 0.3307121992111206,
+      "learning_rate": 0.0003983141612985569,
+      "loss": 1.6048,
+      "step": 3783
+    },
+    {
+      "epoch": 1.7352367847723884,
+      "grad_norm": 0.357295960187912,
+      "learning_rate": 0.00039807170661821414,
+      "loss": 1.0067,
+      "step": 3784
+    },
+    {
+      "epoch": 1.7356954477697513,
+      "grad_norm": 0.28571048378944397,
+      "learning_rate": 0.0003978292769416167,
+      "loss": 0.9839,
+      "step": 3785
+    },
+    {
+      "epoch": 1.736154110767114,
+      "grad_norm": 0.35537266731262207,
+      "learning_rate": 0.00039758687232823434,
+      "loss": 1.6417,
+      "step": 3786
+    },
+    {
+      "epoch": 1.7366127737644765,
+      "grad_norm": 0.28138241171836853,
+      "learning_rate": 0.0003973444928375307,
+      "loss": 1.1609,
+      "step": 3787
+    },
+    {
+      "epoch": 1.7370714367618394,
+      "grad_norm": 0.24497286975383759,
+      "learning_rate": 0.0003971021385289631,
+      "loss": 1.7524,
+      "step": 3788
+    },
+    {
+      "epoch": 1.7375300997592018,
+      "grad_norm": 0.289495587348938,
+      "learning_rate": 0.0003968598094619828,
+      "loss": 1.1777,
+      "step": 3789
+    },
+    {
+      "epoch": 1.7379887627565647,
+      "grad_norm": 0.2697417736053467,
+      "learning_rate": 0.00039661750569603495,
+      "loss": 1.42,
+      "step": 3790
+    },
+    {
+      "epoch": 1.7384474257539273,
+      "grad_norm": 0.3958907127380371,
+      "learning_rate": 0.00039637522729055836,
+      "loss": 1.7687,
+      "step": 3791
+    },
+    {
+      "epoch": 1.73890608875129,
+      "grad_norm": 0.23908482491970062,
+      "learning_rate": 0.00039613297430498586,
+      "loss": 0.9498,
+      "step": 3792
+    },
+    {
+      "epoch": 1.7393647517486528,
+      "grad_norm": 0.262422651052475,
+      "learning_rate": 0.0003958907467987435,
+      "loss": 1.0038,
+      "step": 3793
+    },
+    {
+      "epoch": 1.7398234147460152,
+      "grad_norm": 0.3177359998226166,
+      "learning_rate": 0.00039564854483125164,
+      "loss": 1.9203,
+      "step": 3794
+    },
+    {
+      "epoch": 1.740282077743378,
+      "grad_norm": 0.3895881772041321,
+      "learning_rate": 0.0003954063684619241,
+      "loss": 1.788,
+      "step": 3795
+    },
+    {
+      "epoch": 1.7407407407407407,
+      "grad_norm": 0.2772713005542755,
+      "learning_rate": 0.00039516421775016863,
+      "loss": 1.3488,
+      "step": 3796
+    },
+    {
+      "epoch": 1.7411994037381033,
+      "grad_norm": 0.2620038390159607,
+      "learning_rate": 0.00039492209275538624,
+      "loss": 1.4497,
+      "step": 3797
+    },
+    {
+      "epoch": 1.7416580667354662,
+      "grad_norm": 0.21062573790550232,
+      "learning_rate": 0.0003946799935369726,
+      "loss": 0.8097,
+      "step": 3798
+    },
+    {
+      "epoch": 1.7421167297328288,
+      "grad_norm": 0.313883900642395,
+      "learning_rate": 0.0003944379201543156,
+      "loss": 1.6286,
+      "step": 3799
+    },
+    {
+      "epoch": 1.7425753927301915,
+      "grad_norm": 0.28428205847740173,
+      "learning_rate": 0.000394195872666798,
+      "loss": 1.1285,
+      "step": 3800
+    },
+    {
+      "epoch": 1.7430340557275543,
+      "grad_norm": 0.33933934569358826,
+      "learning_rate": 0.00039395385113379566,
+      "loss": 1.5898,
+      "step": 3801
+    },
+    {
+      "epoch": 1.7434927187249167,
+      "grad_norm": 0.3156382143497467,
+      "learning_rate": 0.00039371185561467827,
+      "loss": 0.9087,
+      "step": 3802
+    },
+    {
+      "epoch": 1.7439513817222796,
+      "grad_norm": 0.38781848549842834,
+      "learning_rate": 0.0003934698861688093,
+      "loss": 1.8621,
+      "step": 3803
+    },
+    {
+      "epoch": 1.7444100447196422,
+      "grad_norm": 0.2756793200969696,
+      "learning_rate": 0.0003932279428555452,
+      "loss": 1.5569,
+      "step": 3804
+    },
+    {
+      "epoch": 1.7448687077170049,
+      "grad_norm": 0.2981017231941223,
+      "learning_rate": 0.0003929860257342366,
+      "loss": 1.1586,
+      "step": 3805
+    },
+    {
+      "epoch": 1.7453273707143677,
+      "grad_norm": 0.32057425379753113,
+      "learning_rate": 0.0003927441348642274,
+      "loss": 1.6129,
+      "step": 3806
+    },
+    {
+      "epoch": 1.7457860337117304,
+      "grad_norm": 0.359298437833786,
+      "learning_rate": 0.0003925022703048553,
+      "loss": 1.2742,
+      "step": 3807
+    },
+    {
+      "epoch": 1.746244696709093,
+      "grad_norm": 0.21599337458610535,
+      "learning_rate": 0.0003922604321154514,
+      "loss": 1.2625,
+      "step": 3808
+    },
+    {
+      "epoch": 1.7467033597064556,
+      "grad_norm": 0.25992512702941895,
+      "learning_rate": 0.00039201862035534066,
+      "loss": 1.3849,
+      "step": 3809
+    },
+    {
+      "epoch": 1.7471620227038183,
+      "grad_norm": 0.342759907245636,
+      "learning_rate": 0.0003917768350838406,
+      "loss": 1.7155,
+      "step": 3810
+    },
+    {
+      "epoch": 1.7476206857011811,
+      "grad_norm": 0.22882609069347382,
+      "learning_rate": 0.0003915350763602632,
+      "loss": 1.1857,
+      "step": 3811
+    },
+    {
+      "epoch": 1.7480793486985438,
+      "grad_norm": 0.2731204032897949,
+      "learning_rate": 0.0003912933442439137,
+      "loss": 1.0107,
+      "step": 3812
+    },
+    {
+      "epoch": 1.7485380116959064,
+      "grad_norm": 0.21852166950702667,
+      "learning_rate": 0.00039105163879409066,
+      "loss": 1.2848,
+      "step": 3813
+    },
+    {
+      "epoch": 1.7489966746932692,
+      "grad_norm": 0.6290088295936584,
+      "learning_rate": 0.00039080996007008625,
+      "loss": 1.4191,
+      "step": 3814
+    },
+    {
+      "epoch": 1.7494553376906317,
+      "grad_norm": 0.2989594042301178,
+      "learning_rate": 0.0003905683081311861,
+      "loss": 1.5646,
+      "step": 3815
+    },
+    {
+      "epoch": 1.7499140006879945,
+      "grad_norm": 0.256778746843338,
+      "learning_rate": 0.00039032668303666876,
+      "loss": 1.4916,
+      "step": 3816
+    },
+    {
+      "epoch": 1.7503726636853572,
+      "grad_norm": 0.19242458045482635,
+      "learning_rate": 0.00039008508484580684,
+      "loss": 0.338,
+      "step": 3817
+    },
+    {
+      "epoch": 1.7508313266827198,
+      "grad_norm": 0.24674712121486664,
+      "learning_rate": 0.0003898435136178662,
+      "loss": 1.1446,
+      "step": 3818
+    },
+    {
+      "epoch": 1.7512899896800826,
+      "grad_norm": 0.2414608746767044,
+      "learning_rate": 0.000389601969412106,
+      "loss": 1.5443,
+      "step": 3819
+    },
+    {
+      "epoch": 1.7517486526774453,
+      "grad_norm": 0.8330361247062683,
+      "learning_rate": 0.00038936045228777884,
+      "loss": 1.5558,
+      "step": 3820
+    },
+    {
+      "epoch": 1.752207315674808,
+      "grad_norm": 0.2938498258590698,
+      "learning_rate": 0.0003891189623041302,
+      "loss": 1.0142,
+      "step": 3821
+    },
+    {
+      "epoch": 1.7526659786721708,
+      "grad_norm": 0.3147178888320923,
+      "learning_rate": 0.0003888774995203997,
+      "loss": 1.2997,
+      "step": 3822
+    },
+    {
+      "epoch": 1.7531246416695332,
+      "grad_norm": 0.26122626662254333,
+      "learning_rate": 0.0003886360639958198,
+      "loss": 1.3746,
+      "step": 3823
+    },
+    {
+      "epoch": 1.753583304666896,
+      "grad_norm": 0.08440568298101425,
+      "learning_rate": 0.00038839465578961637,
+      "loss": 0.9028,
+      "step": 3824
+    },
+    {
+      "epoch": 1.7540419676642587,
+      "grad_norm": 0.27365854382514954,
+      "learning_rate": 0.00038815327496100863,
+      "loss": 1.2714,
+      "step": 3825
+    },
+    {
+      "epoch": 1.7545006306616213,
+      "grad_norm": 0.2890813648700714,
+      "learning_rate": 0.0003879119215692091,
+      "loss": 1.3066,
+      "step": 3826
+    },
+    {
+      "epoch": 1.7549592936589842,
+      "grad_norm": 0.315480500459671,
+      "learning_rate": 0.00038767059567342325,
+      "loss": 1.3728,
+      "step": 3827
+    },
+    {
+      "epoch": 1.7554179566563466,
+      "grad_norm": 0.22030934691429138,
+      "learning_rate": 0.0003874292973328502,
+      "loss": 1.0422,
+      "step": 3828
+    },
+    {
+      "epoch": 1.7558766196537094,
+      "grad_norm": 0.21673201024532318,
+      "learning_rate": 0.0003871880266066823,
+      "loss": 1.2688,
+      "step": 3829
+    },
+    {
+      "epoch": 1.756335282651072,
+      "grad_norm": 0.31848639249801636,
+      "learning_rate": 0.0003869467835541048,
+      "loss": 1.376,
+      "step": 3830
+    },
+    {
+      "epoch": 1.7567939456484347,
+      "grad_norm": 0.08145318180322647,
+      "learning_rate": 0.0003867055682342966,
+      "loss": 0.4348,
+      "step": 3831
+    },
+    {
+      "epoch": 1.7572526086457976,
+      "grad_norm": 0.1764577478170395,
+      "learning_rate": 0.00038646438070642926,
+      "loss": 0.7916,
+      "step": 3832
+    },
+    {
+      "epoch": 1.7577112716431602,
+      "grad_norm": 0.32474926114082336,
+      "learning_rate": 0.00038622322102966803,
+      "loss": 1.7656,
+      "step": 3833
+    },
+    {
+      "epoch": 1.7581699346405228,
+      "grad_norm": 0.2409229725599289,
+      "learning_rate": 0.00038598208926317096,
+      "loss": 1.5365,
+      "step": 3834
+    },
+    {
+      "epoch": 1.7586285976378857,
+      "grad_norm": 0.2720002830028534,
+      "learning_rate": 0.00038574098546608957,
+      "loss": 0.9022,
+      "step": 3835
+    },
+    {
+      "epoch": 1.7590872606352481,
+      "grad_norm": 1.557668924331665,
+      "learning_rate": 0.0003854999096975683,
+      "loss": 1.754,
+      "step": 3836
+    },
+    {
+      "epoch": 1.759545923632611,
+      "grad_norm": 0.3512636721134186,
+      "learning_rate": 0.00038525886201674485,
+      "loss": 1.4061,
+      "step": 3837
+    },
+    {
+      "epoch": 1.7600045866299736,
+      "grad_norm": 0.29787155985832214,
+      "learning_rate": 0.0003850178424827497,
+      "loss": 1.8271,
+      "step": 3838
+    },
+    {
+      "epoch": 1.7604632496273362,
+      "grad_norm": 0.3310008645057678,
+      "learning_rate": 0.000384776851154707,
+      "loss": 1.6361,
+      "step": 3839
+    },
+    {
+      "epoch": 1.760921912624699,
+      "grad_norm": 0.42053020000457764,
+      "learning_rate": 0.00038453588809173343,
+      "loss": 1.8173,
+      "step": 3840
+    },
+    {
+      "epoch": 1.7613805756220617,
+      "grad_norm": 0.41074061393737793,
+      "learning_rate": 0.00038429495335293905,
+      "loss": 1.7546,
+      "step": 3841
+    },
+    {
+      "epoch": 1.7618392386194244,
+      "grad_norm": 0.27962028980255127,
+      "learning_rate": 0.00038405404699742694,
+      "loss": 1.4692,
+      "step": 3842
+    },
+    {
+      "epoch": 1.762297901616787,
+      "grad_norm": 0.3928588628768921,
+      "learning_rate": 0.0003838131690842932,
+      "loss": 1.9565,
+      "step": 3843
+    },
+    {
+      "epoch": 1.7627565646141496,
+      "grad_norm": 0.4200422465801239,
+      "learning_rate": 0.0003835723196726267,
+      "loss": 1.8337,
+      "step": 3844
+    },
+    {
+      "epoch": 1.7632152276115125,
+      "grad_norm": 0.3624529540538788,
+      "learning_rate": 0.0003833314988215097,
+      "loss": 1.7949,
+      "step": 3845
+    },
+    {
+      "epoch": 1.7636738906088751,
+      "grad_norm": 0.4131540358066559,
+      "learning_rate": 0.00038309070659001723,
+      "loss": 1.7876,
+      "step": 3846
+    },
+    {
+      "epoch": 1.7641325536062378,
+      "grad_norm": 0.26267632842063904,
+      "learning_rate": 0.00038284994303721743,
+      "loss": 1.3583,
+      "step": 3847
+    },
+    {
+      "epoch": 1.7645912166036006,
+      "grad_norm": 0.31181371212005615,
+      "learning_rate": 0.0003826092082221714,
+      "loss": 1.2529,
+      "step": 3848
+    },
+    {
+      "epoch": 1.765049879600963,
+      "grad_norm": 0.2337476760149002,
+      "learning_rate": 0.00038236850220393285,
+      "loss": 1.1989,
+      "step": 3849
+    },
+    {
+      "epoch": 1.765508542598326,
+      "grad_norm": 0.2774173319339752,
+      "learning_rate": 0.000382127825041549,
+      "loss": 1.6288,
+      "step": 3850
+    },
+    {
+      "epoch": 1.7659672055956885,
+      "grad_norm": 0.3347560465335846,
+      "learning_rate": 0.0003818871767940595,
+      "loss": 1.3356,
+      "step": 3851
+    },
+    {
+      "epoch": 1.7664258685930512,
+      "grad_norm": 0.339616596698761,
+      "learning_rate": 0.00038164655752049713,
+      "loss": 1.9014,
+      "step": 3852
+    },
+    {
+      "epoch": 1.766884531590414,
+      "grad_norm": 0.3301674425601959,
+      "learning_rate": 0.0003814059672798876,
+      "loss": 1.2728,
+      "step": 3853
+    },
+    {
+      "epoch": 1.7673431945877767,
+      "grad_norm": 0.2250106781721115,
+      "learning_rate": 0.0003811654061312495,
+      "loss": 1.7408,
+      "step": 3854
+    },
+    {
+      "epoch": 1.7678018575851393,
+      "grad_norm": 0.252518892288208,
+      "learning_rate": 0.00038092487413359405,
+      "loss": 1.2438,
+      "step": 3855
+    },
+    {
+      "epoch": 1.7682605205825022,
+      "grad_norm": 0.2107822746038437,
+      "learning_rate": 0.00038068437134592553,
+      "loss": 1.3066,
+      "step": 3856
+    },
+    {
+      "epoch": 1.7687191835798646,
+      "grad_norm": 0.33197417855262756,
+      "learning_rate": 0.0003804438978272411,
+      "loss": 1.2697,
+      "step": 3857
+    },
+    {
+      "epoch": 1.7691778465772274,
+      "grad_norm": 0.22738035023212433,
+      "learning_rate": 0.0003802034536365305,
+      "loss": 0.8106,
+      "step": 3858
+    },
+    {
+      "epoch": 1.76963650957459,
+      "grad_norm": 0.22326631844043732,
+      "learning_rate": 0.0003799630388327766,
+      "loss": 1.5657,
+      "step": 3859
+    },
+    {
+      "epoch": 1.7700951725719527,
+      "grad_norm": 0.30757084488868713,
+      "learning_rate": 0.00037972265347495474,
+      "loss": 1.4067,
+      "step": 3860
+    },
+    {
+      "epoch": 1.7705538355693156,
+      "grad_norm": 0.22894053161144257,
+      "learning_rate": 0.00037948229762203313,
+      "loss": 0.7838,
+      "step": 3861
+    },
+    {
+      "epoch": 1.771012498566678,
+      "grad_norm": 0.3080291152000427,
+      "learning_rate": 0.0003792419713329729,
+      "loss": 1.3486,
+      "step": 3862
+    },
+    {
+      "epoch": 1.7714711615640408,
+      "grad_norm": 0.21413278579711914,
+      "learning_rate": 0.00037900167466672793,
+      "loss": 1.0098,
+      "step": 3863
+    },
+    {
+      "epoch": 1.7719298245614035,
+      "grad_norm": 0.2879568040370941,
+      "learning_rate": 0.00037876140768224444,
+      "loss": 1.5512,
+      "step": 3864
+    },
+    {
+      "epoch": 1.772388487558766,
+      "grad_norm": 0.28043875098228455,
+      "learning_rate": 0.0003785211704384621,
+      "loss": 1.2938,
+      "step": 3865
+    },
+    {
+      "epoch": 1.772847150556129,
+      "grad_norm": 0.3483384847640991,
+      "learning_rate": 0.0003782809629943124,
+      "loss": 1.7574,
+      "step": 3866
+    },
+    {
+      "epoch": 1.7733058135534916,
+      "grad_norm": 0.33012205362319946,
+      "learning_rate": 0.00037804078540872005,
+      "loss": 1.2911,
+      "step": 3867
+    },
+    {
+      "epoch": 1.7737644765508542,
+      "grad_norm": 0.2982887327671051,
+      "learning_rate": 0.0003778006377406025,
+      "loss": 1.8078,
+      "step": 3868
+    },
+    {
+      "epoch": 1.774223139548217,
+      "grad_norm": 0.19081206619739532,
+      "learning_rate": 0.0003775605200488694,
+      "loss": 0.4456,
+      "step": 3869
+    },
+    {
+      "epoch": 1.7746818025455795,
+      "grad_norm": 0.3630902171134949,
+      "learning_rate": 0.00037732043239242373,
+      "loss": 1.98,
+      "step": 3870
+    },
+    {
+      "epoch": 1.7751404655429424,
+      "grad_norm": 0.25359615683555603,
+      "learning_rate": 0.00037708037483016085,
+      "loss": 0.9155,
+      "step": 3871
+    },
+    {
+      "epoch": 1.775599128540305,
+      "grad_norm": 0.3517308533191681,
+      "learning_rate": 0.000376840347420968,
+      "loss": 1.8185,
+      "step": 3872
+    },
+    {
+      "epoch": 1.7760577915376676,
+      "grad_norm": 0.23731131851673126,
+      "learning_rate": 0.00037660035022372604,
+      "loss": 0.7537,
+      "step": 3873
+    },
+    {
+      "epoch": 1.7765164545350305,
+      "grad_norm": 0.2584936320781708,
+      "learning_rate": 0.0003763603832973077,
+      "loss": 1.7855,
+      "step": 3874
+    },
+    {
+      "epoch": 1.7769751175323931,
+      "grad_norm": 0.42461255192756653,
+      "learning_rate": 0.00037612044670057906,
+      "loss": 1.1224,
+      "step": 3875
+    },
+    {
+      "epoch": 1.7774337805297558,
+      "grad_norm": 0.2801918089389801,
+      "learning_rate": 0.00037588054049239817,
+      "loss": 1.6647,
+      "step": 3876
+    },
+    {
+      "epoch": 1.7778924435271186,
+      "grad_norm": 0.3888150751590729,
+      "learning_rate": 0.0003756406647316155,
+      "loss": 1.7823,
+      "step": 3877
+    },
+    {
+      "epoch": 1.778351106524481,
+      "grad_norm": 0.4158638119697571,
+      "learning_rate": 0.00037540081947707443,
+      "loss": 1.3085,
+      "step": 3878
+    },
+    {
+      "epoch": 1.7788097695218439,
+      "grad_norm": 0.055091023445129395,
+      "learning_rate": 0.0003751610047876106,
+      "loss": 0.7616,
+      "step": 3879
+    },
+    {
+      "epoch": 1.7792684325192065,
+      "grad_norm": 0.2197464108467102,
+      "learning_rate": 0.00037492122072205257,
+      "loss": 0.7161,
+      "step": 3880
+    },
+    {
+      "epoch": 1.7797270955165692,
+      "grad_norm": 0.39263108372688293,
+      "learning_rate": 0.00037468146733922106,
+      "loss": 1.9539,
+      "step": 3881
+    },
+    {
+      "epoch": 1.780185758513932,
+      "grad_norm": 0.43456920981407166,
+      "learning_rate": 0.0003744417446979293,
+      "loss": 1.8441,
+      "step": 3882
+    },
+    {
+      "epoch": 1.7806444215112944,
+      "grad_norm": 0.3044719099998474,
+      "learning_rate": 0.0003742020528569827,
+      "loss": 1.7751,
+      "step": 3883
+    },
+    {
+      "epoch": 1.7811030845086573,
+      "grad_norm": 0.44964897632598877,
+      "learning_rate": 0.0003739623918751795,
+      "loss": 1.338,
+      "step": 3884
+    },
+    {
+      "epoch": 1.78156174750602,
+      "grad_norm": 0.3019511103630066,
+      "learning_rate": 0.00037372276181131043,
+      "loss": 1.298,
+      "step": 3885
+    },
+    {
+      "epoch": 1.7820204105033826,
+      "grad_norm": 0.33377906680107117,
+      "learning_rate": 0.0003734831627241584,
+      "loss": 2.1521,
+      "step": 3886
+    },
+    {
+      "epoch": 1.7824790735007454,
+      "grad_norm": 0.4243369996547699,
+      "learning_rate": 0.000373243594672499,
+      "loss": 2.0967,
+      "step": 3887
+    },
+    {
+      "epoch": 1.782937736498108,
+      "grad_norm": 0.41644036769866943,
+      "learning_rate": 0.0003730040577150995,
+      "loss": 1.89,
+      "step": 3888
+    },
+    {
+      "epoch": 1.7833963994954707,
+      "grad_norm": 0.2524632215499878,
+      "learning_rate": 0.0003727645519107204,
+      "loss": 1.5137,
+      "step": 3889
+    },
+    {
+      "epoch": 1.7838550624928335,
+      "grad_norm": 0.4522644281387329,
+      "learning_rate": 0.0003725250773181141,
+      "loss": 1.5999,
+      "step": 3890
+    },
+    {
+      "epoch": 1.784313725490196,
+      "grad_norm": 0.25750890374183655,
+      "learning_rate": 0.0003722856339960256,
+      "loss": 1.3307,
+      "step": 3891
+    },
+    {
+      "epoch": 1.7847723884875588,
+      "grad_norm": 0.37072017788887024,
+      "learning_rate": 0.0003720462220031918,
+      "loss": 1.6609,
+      "step": 3892
+    },
+    {
+      "epoch": 1.7852310514849214,
+      "grad_norm": 0.35205331444740295,
+      "learning_rate": 0.0003718068413983425,
+      "loss": 1.0562,
+      "step": 3893
+    },
+    {
+      "epoch": 1.785689714482284,
+      "grad_norm": 0.05115019530057907,
+      "learning_rate": 0.00037156749224019923,
+      "loss": 0.6781,
+      "step": 3894
+    },
+    {
+      "epoch": 1.786148377479647,
+      "grad_norm": 0.25191041827201843,
+      "learning_rate": 0.00037132817458747613,
+      "loss": 1.4272,
+      "step": 3895
+    },
+    {
+      "epoch": 1.7866070404770094,
+      "grad_norm": 0.27612894773483276,
+      "learning_rate": 0.00037108888849887966,
+      "loss": 0.9952,
+      "step": 3896
+    },
+    {
+      "epoch": 1.7870657034743722,
+      "grad_norm": 0.2937767803668976,
+      "learning_rate": 0.0003708496340331082,
+      "loss": 1.6864,
+      "step": 3897
+    },
+    {
+      "epoch": 1.7875243664717348,
+      "grad_norm": 0.3436873257160187,
+      "learning_rate": 0.00037061041124885285,
+      "loss": 1.9355,
+      "step": 3898
+    },
+    {
+      "epoch": 1.7879830294690975,
+      "grad_norm": 0.35292357206344604,
+      "learning_rate": 0.00037037122020479665,
+      "loss": 1.7389,
+      "step": 3899
+    },
+    {
+      "epoch": 1.7884416924664603,
+      "grad_norm": 0.3429259955883026,
+      "learning_rate": 0.0003701320609596147,
+      "loss": 1.8025,
+      "step": 3900
+    },
+    {
+      "epoch": 1.788900355463823,
+      "grad_norm": 0.38343197107315063,
+      "learning_rate": 0.00036989293357197464,
+      "loss": 2.0798,
+      "step": 3901
+    },
+    {
+      "epoch": 1.7893590184611856,
+      "grad_norm": 0.30104926228523254,
+      "learning_rate": 0.00036965383810053597,
+      "loss": 1.3101,
+      "step": 3902
+    },
+    {
+      "epoch": 1.7898176814585485,
+      "grad_norm": 0.2541302740573883,
+      "learning_rate": 0.00036941477460395074,
+      "loss": 1.0455,
+      "step": 3903
+    },
+    {
+      "epoch": 1.7902763444559109,
+      "grad_norm": 0.2968154847621918,
+      "learning_rate": 0.000369175743140863,
+      "loss": 1.9026,
+      "step": 3904
+    },
+    {
+      "epoch": 1.7907350074532737,
+      "grad_norm": 0.33623620867729187,
+      "learning_rate": 0.0003689367437699086,
+      "loss": 1.1528,
+      "step": 3905
+    },
+    {
+      "epoch": 1.7911936704506364,
+      "grad_norm": 0.3630521297454834,
+      "learning_rate": 0.00036869777654971594,
+      "loss": 1.7005,
+      "step": 3906
+    },
+    {
+      "epoch": 1.791652333447999,
+      "grad_norm": 0.38895827531814575,
+      "learning_rate": 0.0003684588415389055,
+      "loss": 1.7465,
+      "step": 3907
+    },
+    {
+      "epoch": 1.7921109964453619,
+      "grad_norm": 0.38369861245155334,
+      "learning_rate": 0.0003682199387960896,
+      "loss": 1.2702,
+      "step": 3908
+    },
+    {
+      "epoch": 1.7925696594427245,
+      "grad_norm": 0.222330242395401,
+      "learning_rate": 0.00036798106837987297,
+      "loss": 1.2868,
+      "step": 3909
+    },
+    {
+      "epoch": 1.7930283224400871,
+      "grad_norm": 0.2768653631210327,
+      "learning_rate": 0.00036774223034885236,
+      "loss": 1.0325,
+      "step": 3910
+    },
+    {
+      "epoch": 1.79348698543745,
+      "grad_norm": 0.4517102837562561,
+      "learning_rate": 0.000367503424761616,
+      "loss": 1.3302,
+      "step": 3911
+    },
+    {
+      "epoch": 1.7939456484348124,
+      "grad_norm": 0.31973719596862793,
+      "learning_rate": 0.000367264651676745,
+      "loss": 1.7516,
+      "step": 3912
+    },
+    {
+      "epoch": 1.7944043114321753,
+      "grad_norm": 0.36346325278282166,
+      "learning_rate": 0.000367025911152812,
+      "loss": 1.7175,
+      "step": 3913
+    },
+    {
+      "epoch": 1.794862974429538,
+      "grad_norm": 0.3458804190158844,
+      "learning_rate": 0.00036678720324838176,
+      "loss": 1.3469,
+      "step": 3914
+    },
+    {
+      "epoch": 1.7953216374269005,
+      "grad_norm": 0.3012971580028534,
+      "learning_rate": 0.0003665485280220112,
+      "loss": 1.5734,
+      "step": 3915
+    },
+    {
+      "epoch": 1.7957803004242634,
+      "grad_norm": 0.3793402910232544,
+      "learning_rate": 0.00036630988553224887,
+      "loss": 1.6147,
+      "step": 3916
+    },
+    {
+      "epoch": 1.7962389634216258,
+      "grad_norm": 0.4072246253490448,
+      "learning_rate": 0.00036607127583763554,
+      "loss": 1.4536,
+      "step": 3917
+    },
+    {
+      "epoch": 1.7966976264189887,
+      "grad_norm": 0.32000985741615295,
+      "learning_rate": 0.0003658326989967039,
+      "loss": 1.438,
+      "step": 3918
+    },
+    {
+      "epoch": 1.7971562894163513,
+      "grad_norm": 0.2759474217891693,
+      "learning_rate": 0.00036559415506797865,
+      "loss": 1.33,
+      "step": 3919
+    },
+    {
+      "epoch": 1.797614952413714,
+      "grad_norm": 0.32986119389533997,
+      "learning_rate": 0.0003653556441099762,
+      "loss": 1.3684,
+      "step": 3920
+    },
+    {
+      "epoch": 1.7980736154110768,
+      "grad_norm": 0.23913626372814178,
+      "learning_rate": 0.0003651171661812053,
+      "loss": 1.3252,
+      "step": 3921
+    },
+    {
+      "epoch": 1.7985322784084394,
+      "grad_norm": 0.33937355875968933,
+      "learning_rate": 0.0003648787213401659,
+      "loss": 1.4227,
+      "step": 3922
+    },
+    {
+      "epoch": 1.798990941405802,
+      "grad_norm": 0.2144535481929779,
+      "learning_rate": 0.00036464030964535044,
+      "loss": 0.9513,
+      "step": 3923
+    },
+    {
+      "epoch": 1.799449604403165,
+      "grad_norm": 0.2545612156391144,
+      "learning_rate": 0.00036440193115524306,
+      "loss": 1.1305,
+      "step": 3924
+    },
+    {
+      "epoch": 1.7999082674005273,
+      "grad_norm": 0.33486199378967285,
+      "learning_rate": 0.0003641635859283197,
+      "loss": 1.4782,
+      "step": 3925
+    },
+    {
+      "epoch": 1.8003669303978902,
+      "grad_norm": 0.16233354806900024,
+      "learning_rate": 0.000363925274023048,
+      "loss": 0.611,
+      "step": 3926
+    },
+    {
+      "epoch": 1.8008255933952528,
+      "grad_norm": 0.33829358220100403,
+      "learning_rate": 0.00036368699549788795,
+      "loss": 1.8271,
+      "step": 3927
+    },
+    {
+      "epoch": 1.8012842563926155,
+      "grad_norm": 0.2723900079727173,
+      "learning_rate": 0.00036344875041129066,
+      "loss": 1.1435,
+      "step": 3928
+    },
+    {
+      "epoch": 1.8017429193899783,
+      "grad_norm": 0.2544368803501129,
+      "learning_rate": 0.00036321053882169954,
+      "loss": 0.8193,
+      "step": 3929
+    },
+    {
+      "epoch": 1.8022015823873407,
+      "grad_norm": 0.2767103910446167,
+      "learning_rate": 0.00036297236078754945,
+      "loss": 1.3754,
+      "step": 3930
+    },
+    {
+      "epoch": 1.8026602453847036,
+      "grad_norm": 0.1679062396287918,
+      "learning_rate": 0.00036273421636726723,
+      "loss": 0.8304,
+      "step": 3931
+    },
+    {
+      "epoch": 1.8031189083820662,
+      "grad_norm": 0.32275891304016113,
+      "learning_rate": 0.0003624961056192717,
+      "loss": 1.2045,
+      "step": 3932
+    },
+    {
+      "epoch": 1.8035775713794289,
+      "grad_norm": 0.1948852390050888,
+      "learning_rate": 0.00036225802860197275,
+      "loss": 0.963,
+      "step": 3933
+    },
+    {
+      "epoch": 1.8040362343767917,
+      "grad_norm": 0.39679884910583496,
+      "learning_rate": 0.00036201998537377264,
+      "loss": 1.5401,
+      "step": 3934
+    },
+    {
+      "epoch": 1.8044948973741544,
+      "grad_norm": 0.2812747657299042,
+      "learning_rate": 0.00036178197599306494,
+      "loss": 2.0823,
+      "step": 3935
+    },
+    {
+      "epoch": 1.804953560371517,
+      "grad_norm": 0.42034879326820374,
+      "learning_rate": 0.00036154400051823497,
+      "loss": 2.0599,
+      "step": 3936
+    },
+    {
+      "epoch": 1.8054122233688799,
+      "grad_norm": 0.4247654676437378,
+      "learning_rate": 0.00036130605900766024,
+      "loss": 1.8334,
+      "step": 3937
+    },
+    {
+      "epoch": 1.8058708863662423,
+      "grad_norm": 0.35334038734436035,
+      "learning_rate": 0.0003610681515197094,
+      "loss": 2.0579,
+      "step": 3938
+    },
+    {
+      "epoch": 1.8063295493636051,
+      "grad_norm": 0.3598625063896179,
+      "learning_rate": 0.0003608302781127425,
+      "loss": 2.2363,
+      "step": 3939
+    },
+    {
+      "epoch": 1.8067882123609678,
+      "grad_norm": 0.29301023483276367,
+      "learning_rate": 0.00036059243884511185,
+      "loss": 1.0839,
+      "step": 3940
+    },
+    {
+      "epoch": 1.8072468753583304,
+      "grad_norm": 0.281200110912323,
+      "learning_rate": 0.0003603546337751611,
+      "loss": 1.0665,
+      "step": 3941
+    },
+    {
+      "epoch": 1.8077055383556933,
+      "grad_norm": 0.16020144522190094,
+      "learning_rate": 0.0003601168629612256,
+      "loss": 1.1812,
+      "step": 3942
+    },
+    {
+      "epoch": 1.8081642013530559,
+      "grad_norm": 0.3196254372596741,
+      "learning_rate": 0.00035987912646163247,
+      "loss": 1.4895,
+      "step": 3943
+    },
+    {
+      "epoch": 1.8086228643504185,
+      "grad_norm": 0.34358012676239014,
+      "learning_rate": 0.0003596414243346997,
+      "loss": 1.4952,
+      "step": 3944
+    },
+    {
+      "epoch": 1.8090815273477814,
+      "grad_norm": 0.2578059136867523,
+      "learning_rate": 0.00035940375663873767,
+      "loss": 1.4763,
+      "step": 3945
+    },
+    {
+      "epoch": 1.8095401903451438,
+      "grad_norm": 0.33023521304130554,
+      "learning_rate": 0.0003591661234320477,
+      "loss": 1.0975,
+      "step": 3946
+    },
+    {
+      "epoch": 1.8099988533425067,
+      "grad_norm": 0.16670942306518555,
+      "learning_rate": 0.00035892852477292325,
+      "loss": 1.1774,
+      "step": 3947
+    },
+    {
+      "epoch": 1.8104575163398693,
+      "grad_norm": 0.3831387758255005,
+      "learning_rate": 0.00035869096071964885,
+      "loss": 2.0361,
+      "step": 3948
+    },
+    {
+      "epoch": 1.810916179337232,
+      "grad_norm": 0.3592255413532257,
+      "learning_rate": 0.0003584534313305009,
+      "loss": 1.2978,
+      "step": 3949
+    },
+    {
+      "epoch": 1.8113748423345948,
+      "grad_norm": 0.05875217914581299,
+      "learning_rate": 0.0003582159366637466,
+      "loss": 0.4578,
+      "step": 3950
+    },
+    {
+      "epoch": 1.8118335053319572,
+      "grad_norm": 0.2751762270927429,
+      "learning_rate": 0.00035797847677764526,
+      "loss": 1.256,
+      "step": 3951
+    },
+    {
+      "epoch": 1.81229216832932,
+      "grad_norm": 0.20929132401943207,
+      "learning_rate": 0.0003577410517304477,
+      "loss": 1.1679,
+      "step": 3952
+    },
+    {
+      "epoch": 1.8127508313266827,
+      "grad_norm": 0.25931280851364136,
+      "learning_rate": 0.00035750366158039594,
+      "loss": 0.9565,
+      "step": 3953
+    },
+    {
+      "epoch": 1.8132094943240453,
+      "grad_norm": 0.20596548914909363,
+      "learning_rate": 0.0003572663063857234,
+      "loss": 1.2343,
+      "step": 3954
+    },
+    {
+      "epoch": 1.8136681573214082,
+      "grad_norm": 0.2564990818500519,
+      "learning_rate": 0.0003570289862046553,
+      "loss": 1.6389,
+      "step": 3955
+    },
+    {
+      "epoch": 1.8141268203187708,
+      "grad_norm": 0.3115670382976532,
+      "learning_rate": 0.0003567917010954074,
+      "loss": 1.24,
+      "step": 3956
+    },
+    {
+      "epoch": 1.8145854833161335,
+      "grad_norm": 0.29912760853767395,
+      "learning_rate": 0.0003565544511161879,
+      "loss": 1.6694,
+      "step": 3957
+    },
+    {
+      "epoch": 1.8150441463134963,
+      "grad_norm": 0.37836483120918274,
+      "learning_rate": 0.00035631723632519594,
+      "loss": 2.1592,
+      "step": 3958
+    },
+    {
+      "epoch": 1.8155028093108587,
+      "grad_norm": 0.3637576401233673,
+      "learning_rate": 0.0003560800567806218,
+      "loss": 1.2883,
+      "step": 3959
+    },
+    {
+      "epoch": 1.8159614723082216,
+      "grad_norm": 0.2932051420211792,
+      "learning_rate": 0.0003558429125406476,
+      "loss": 1.4525,
+      "step": 3960
+    },
+    {
+      "epoch": 1.8164201353055842,
+      "grad_norm": 0.3286663591861725,
+      "learning_rate": 0.0003556058036634463,
+      "loss": 1.7548,
+      "step": 3961
+    },
+    {
+      "epoch": 1.8168787983029469,
+      "grad_norm": 0.3011263310909271,
+      "learning_rate": 0.00035536873020718254,
+      "loss": 1.6652,
+      "step": 3962
+    },
+    {
+      "epoch": 1.8173374613003097,
+      "grad_norm": 0.381115585565567,
+      "learning_rate": 0.000355131692230012,
+      "loss": 1.6675,
+      "step": 3963
+    },
+    {
+      "epoch": 1.8177961242976721,
+      "grad_norm": 0.2726176083087921,
+      "learning_rate": 0.00035489468979008195,
+      "loss": 1.4564,
+      "step": 3964
+    },
+    {
+      "epoch": 1.818254787295035,
+      "grad_norm": 0.32600149512290955,
+      "learning_rate": 0.0003546577229455308,
+      "loss": 1.9954,
+      "step": 3965
+    },
+    {
+      "epoch": 1.8187134502923976,
+      "grad_norm": 0.2766259014606476,
+      "learning_rate": 0.0003544207917544882,
+      "loss": 1.4604,
+      "step": 3966
+    },
+    {
+      "epoch": 1.8191721132897603,
+      "grad_norm": 0.3894624710083008,
+      "learning_rate": 0.000354183896275075,
+      "loss": 0.8529,
+      "step": 3967
+    },
+    {
+      "epoch": 1.819630776287123,
+      "grad_norm": 0.2296787053346634,
+      "learning_rate": 0.00035394703656540345,
+      "loss": 1.3822,
+      "step": 3968
+    },
+    {
+      "epoch": 1.8200894392844857,
+      "grad_norm": 0.3591334819793701,
+      "learning_rate": 0.00035371021268357694,
+      "loss": 1.9995,
+      "step": 3969
+    },
+    {
+      "epoch": 1.8205481022818484,
+      "grad_norm": 0.33714228868484497,
+      "learning_rate": 0.00035347342468769,
+      "loss": 1.4948,
+      "step": 3970
+    },
+    {
+      "epoch": 1.8210067652792112,
+      "grad_norm": 0.2692127227783203,
+      "learning_rate": 0.00035323667263582865,
+      "loss": 1.1647,
+      "step": 3971
+    },
+    {
+      "epoch": 1.8214654282765737,
+      "grad_norm": 0.1590389609336853,
+      "learning_rate": 0.00035299995658606963,
+      "loss": 0.9988,
+      "step": 3972
+    },
+    {
+      "epoch": 1.8219240912739365,
+      "grad_norm": 0.25324851274490356,
+      "learning_rate": 0.0003527632765964811,
+      "loss": 1.0857,
+      "step": 3973
+    },
+    {
+      "epoch": 1.8223827542712991,
+      "grad_norm": 0.29076528549194336,
+      "learning_rate": 0.00035252663272512255,
+      "loss": 1.6969,
+      "step": 3974
+    },
+    {
+      "epoch": 1.8228414172686618,
+      "grad_norm": 0.2527695894241333,
+      "learning_rate": 0.0003522900250300443,
+      "loss": 1.2706,
+      "step": 3975
+    },
+    {
+      "epoch": 1.8233000802660246,
+      "grad_norm": 0.34270545840263367,
+      "learning_rate": 0.000352053453569288,
+      "loss": 1.6218,
+      "step": 3976
+    },
+    {
+      "epoch": 1.8237587432633873,
+      "grad_norm": 0.2784097194671631,
+      "learning_rate": 0.00035181691840088623,
+      "loss": 1.166,
+      "step": 3977
+    },
+    {
+      "epoch": 1.82421740626075,
+      "grad_norm": 0.30808866024017334,
+      "learning_rate": 0.0003515804195828629,
+      "loss": 1.2634,
+      "step": 3978
+    },
+    {
+      "epoch": 1.8246760692581128,
+      "grad_norm": 0.16882003843784332,
+      "learning_rate": 0.00035134395717323276,
+      "loss": 0.8922,
+      "step": 3979
+    },
+    {
+      "epoch": 1.8251347322554752,
+      "grad_norm": 0.1895928829908371,
+      "learning_rate": 0.0003511075312300018,
+      "loss": 0.895,
+      "step": 3980
+    },
+    {
+      "epoch": 1.825593395252838,
+      "grad_norm": 0.21428383886814117,
+      "learning_rate": 0.00035087114181116697,
+      "loss": 1.3183,
+      "step": 3981
+    },
+    {
+      "epoch": 1.8260520582502007,
+      "grad_norm": 0.28207674622535706,
+      "learning_rate": 0.0003506347889747164,
+      "loss": 0.9283,
+      "step": 3982
+    },
+    {
+      "epoch": 1.8265107212475633,
+      "grad_norm": 0.24229072034358978,
+      "learning_rate": 0.000350398472778629,
+      "loss": 1.783,
+      "step": 3983
+    },
+    {
+      "epoch": 1.8269693842449262,
+      "grad_norm": 0.309413880109787,
+      "learning_rate": 0.000350162193280875,
+      "loss": 1.2617,
+      "step": 3984
+    },
+    {
+      "epoch": 1.8274280472422886,
+      "grad_norm": 0.2385943979024887,
+      "learning_rate": 0.00034992595053941525,
+      "loss": 1.2587,
+      "step": 3985
+    },
+    {
+      "epoch": 1.8278867102396514,
+      "grad_norm": 0.31749603152275085,
+      "learning_rate": 0.00034968974461220195,
+      "loss": 1.8345,
+      "step": 3986
+    },
+    {
+      "epoch": 1.828345373237014,
+      "grad_norm": 0.28810974955558777,
+      "learning_rate": 0.0003494535755571781,
+      "loss": 1.673,
+      "step": 3987
+    },
+    {
+      "epoch": 1.8288040362343767,
+      "grad_norm": 0.310904324054718,
+      "learning_rate": 0.0003492174434322778,
+      "loss": 1.5857,
+      "step": 3988
+    },
+    {
+      "epoch": 1.8292626992317396,
+      "grad_norm": 0.20826320350170135,
+      "learning_rate": 0.00034898134829542565,
+      "loss": 0.8103,
+      "step": 3989
+    },
+    {
+      "epoch": 1.8297213622291022,
+      "grad_norm": 0.3267036974430084,
+      "learning_rate": 0.0003487452902045377,
+      "loss": 1.3232,
+      "step": 3990
+    },
+    {
+      "epoch": 1.8301800252264648,
+      "grad_norm": 0.24371105432510376,
+      "learning_rate": 0.00034850926921752067,
+      "loss": 1.2933,
+      "step": 3991
+    },
+    {
+      "epoch": 1.8306386882238277,
+      "grad_norm": 0.3523206412792206,
+      "learning_rate": 0.0003482732853922722,
+      "loss": 1.2762,
+      "step": 3992
+    },
+    {
+      "epoch": 1.83109735122119,
+      "grad_norm": 0.5423739552497864,
+      "learning_rate": 0.00034803733878668077,
+      "loss": 0.9312,
+      "step": 3993
+    },
+    {
+      "epoch": 1.831556014218553,
+      "grad_norm": 0.31650495529174805,
+      "learning_rate": 0.0003478014294586261,
+      "loss": 2.0525,
+      "step": 3994
+    },
+    {
+      "epoch": 1.8320146772159156,
+      "grad_norm": 0.3087564706802368,
+      "learning_rate": 0.000347565557465978,
+      "loss": 1.3784,
+      "step": 3995
+    },
+    {
+      "epoch": 1.8324733402132782,
+      "grad_norm": 0.2533722519874573,
+      "learning_rate": 0.0003473297228665978,
+      "loss": 1.3376,
+      "step": 3996
+    },
+    {
+      "epoch": 1.832932003210641,
+      "grad_norm": 0.3357933759689331,
+      "learning_rate": 0.0003470939257183373,
+      "loss": 1.7561,
+      "step": 3997
+    },
+    {
+      "epoch": 1.8333906662080035,
+      "grad_norm": 0.36861729621887207,
+      "learning_rate": 0.0003468581660790393,
+      "loss": 1.7476,
+      "step": 3998
+    },
+    {
+      "epoch": 1.8338493292053664,
+      "grad_norm": 0.26520270109176636,
+      "learning_rate": 0.0003466224440065377,
+      "loss": 1.5038,
+      "step": 3999
+    },
+    {
+      "epoch": 1.834307992202729,
+      "grad_norm": 0.3509536385536194,
+      "learning_rate": 0.0003463867595586562,
+      "loss": 1.8193,
+      "step": 4000
+    },
+    {
+      "epoch": 1.8347666552000916,
+      "grad_norm": 0.2574576139450073,
+      "learning_rate": 0.0003461511127932103,
+      "loss": 1.2697,
+      "step": 4001
+    },
+    {
+      "epoch": 1.8352253181974545,
+      "grad_norm": 0.3160921037197113,
+      "learning_rate": 0.00034591550376800563,
+      "loss": 1.655,
+      "step": 4002
+    },
+    {
+      "epoch": 1.8356839811948171,
+      "grad_norm": 0.34427279233932495,
+      "learning_rate": 0.00034567993254083887,
+      "loss": 1.3817,
+      "step": 4003
+    },
+    {
+      "epoch": 1.8361426441921798,
+      "grad_norm": 0.31001847982406616,
+      "learning_rate": 0.00034544439916949746,
+      "loss": 1.1495,
+      "step": 4004
+    },
+    {
+      "epoch": 1.8366013071895426,
+      "grad_norm": 0.23114445805549622,
+      "learning_rate": 0.0003452089037117595,
+      "loss": 1.3843,
+      "step": 4005
+    },
+    {
+      "epoch": 1.837059970186905,
+      "grad_norm": 0.2845550775527954,
+      "learning_rate": 0.0003449734462253934,
+      "loss": 1.0899,
+      "step": 4006
+    },
+    {
+      "epoch": 1.837518633184268,
+      "grad_norm": 0.368184894323349,
+      "learning_rate": 0.0003447380267681587,
+      "loss": 1.9531,
+      "step": 4007
+    },
+    {
+      "epoch": 1.8379772961816305,
+      "grad_norm": 0.27855613827705383,
+      "learning_rate": 0.0003445026453978054,
+      "loss": 0.8792,
+      "step": 4008
+    },
+    {
+      "epoch": 1.8384359591789932,
+      "grad_norm": 0.29736289381980896,
+      "learning_rate": 0.00034426730217207457,
+      "loss": 1.3437,
+      "step": 4009
+    },
+    {
+      "epoch": 1.838894622176356,
+      "grad_norm": 0.26978176832199097,
+      "learning_rate": 0.0003440319971486976,
+      "loss": 0.9284,
+      "step": 4010
+    },
+    {
+      "epoch": 1.8393532851737187,
+      "grad_norm": 0.24370868504047394,
+      "learning_rate": 0.00034379673038539604,
+      "loss": 1.8301,
+      "step": 4011
+    },
+    {
+      "epoch": 1.8398119481710813,
+      "grad_norm": 0.31381699442863464,
+      "learning_rate": 0.0003435615019398828,
+      "loss": 1.2786,
+      "step": 4012
+    },
+    {
+      "epoch": 1.8402706111684441,
+      "grad_norm": 0.36527925729751587,
+      "learning_rate": 0.00034332631186986107,
+      "loss": 1.5788,
+      "step": 4013
+    },
+    {
+      "epoch": 1.8407292741658066,
+      "grad_norm": 0.29045170545578003,
+      "learning_rate": 0.0003430911602330248,
+      "loss": 1.7421,
+      "step": 4014
+    },
+    {
+      "epoch": 1.8411879371631694,
+      "grad_norm": 0.515656590461731,
+      "learning_rate": 0.0003428560470870583,
+      "loss": 2.0176,
+      "step": 4015
+    },
+    {
+      "epoch": 1.841646600160532,
+      "grad_norm": 0.2787885367870331,
+      "learning_rate": 0.00034262097248963675,
+      "loss": 1.4663,
+      "step": 4016
+    },
+    {
+      "epoch": 1.8421052631578947,
+      "grad_norm": 0.2862381637096405,
+      "learning_rate": 0.00034238593649842515,
+      "loss": 0.9792,
+      "step": 4017
+    },
+    {
+      "epoch": 1.8425639261552575,
+      "grad_norm": 0.30698710680007935,
+      "learning_rate": 0.0003421509391710797,
+      "loss": 1.7229,
+      "step": 4018
+    },
+    {
+      "epoch": 1.84302258915262,
+      "grad_norm": 0.3188919723033905,
+      "learning_rate": 0.0003419159805652471,
+      "loss": 1.5143,
+      "step": 4019
+    },
+    {
+      "epoch": 1.8434812521499828,
+      "grad_norm": 0.3322117328643799,
+      "learning_rate": 0.0003416810607385644,
+      "loss": 1.9916,
+      "step": 4020
+    },
+    {
+      "epoch": 1.8439399151473455,
+      "grad_norm": 0.3120747208595276,
+      "learning_rate": 0.00034144617974865896,
+      "loss": 1.4741,
+      "step": 4021
+    },
+    {
+      "epoch": 1.844398578144708,
+      "grad_norm": 0.2903960049152374,
+      "learning_rate": 0.00034121133765314905,
+      "loss": 1.6743,
+      "step": 4022
+    },
+    {
+      "epoch": 1.844857241142071,
+      "grad_norm": 0.40729033946990967,
+      "learning_rate": 0.00034097653450964265,
+      "loss": 1.9248,
+      "step": 4023
+    },
+    {
+      "epoch": 1.8453159041394336,
+      "grad_norm": 0.3693930208683014,
+      "learning_rate": 0.00034074177037573904,
+      "loss": 2.1533,
+      "step": 4024
+    },
+    {
+      "epoch": 1.8457745671367962,
+      "grad_norm": 0.2894419729709625,
+      "learning_rate": 0.00034050704530902756,
+      "loss": 1.4656,
+      "step": 4025
+    },
+    {
+      "epoch": 1.846233230134159,
+      "grad_norm": 0.32523804903030396,
+      "learning_rate": 0.00034027235936708776,
+      "loss": 1.505,
+      "step": 4026
+    },
+    {
+      "epoch": 1.8466918931315215,
+      "grad_norm": 0.2615148723125458,
+      "learning_rate": 0.00034003771260749017,
+      "loss": 1.1829,
+      "step": 4027
+    },
+    {
+      "epoch": 1.8471505561288843,
+      "grad_norm": 0.34713688492774963,
+      "learning_rate": 0.00033980310508779476,
+      "loss": 1.1238,
+      "step": 4028
+    },
+    {
+      "epoch": 1.847609219126247,
+      "grad_norm": 0.2878703773021698,
+      "learning_rate": 0.0003395685368655528,
+      "loss": 1.0275,
+      "step": 4029
+    },
+    {
+      "epoch": 1.8480678821236096,
+      "grad_norm": 0.2368079125881195,
+      "learning_rate": 0.00033933400799830563,
+      "loss": 1.4276,
+      "step": 4030
+    },
+    {
+      "epoch": 1.8485265451209725,
+      "grad_norm": 0.20480850338935852,
+      "learning_rate": 0.0003390995185435847,
+      "loss": 1.324,
+      "step": 4031
+    },
+    {
+      "epoch": 1.8489852081183349,
+      "grad_norm": 0.27292463183403015,
+      "learning_rate": 0.00033886506855891195,
+      "loss": 1.3118,
+      "step": 4032
+    },
+    {
+      "epoch": 1.8494438711156977,
+      "grad_norm": 0.2694958448410034,
+      "learning_rate": 0.00033863065810179986,
+      "loss": 1.5169,
+      "step": 4033
+    },
+    {
+      "epoch": 1.8499025341130604,
+      "grad_norm": 0.3515981137752533,
+      "learning_rate": 0.0003383962872297508,
+      "loss": 2.0571,
+      "step": 4034
+    },
+    {
+      "epoch": 1.850361197110423,
+      "grad_norm": 0.3751930594444275,
+      "learning_rate": 0.0003381619560002577,
+      "loss": 1.3073,
+      "step": 4035
+    },
+    {
+      "epoch": 1.8508198601077859,
+      "grad_norm": 0.3277819752693176,
+      "learning_rate": 0.0003379276644708037,
+      "loss": 1.6505,
+      "step": 4036
+    },
+    {
+      "epoch": 1.8512785231051485,
+      "grad_norm": 0.17637912929058075,
+      "learning_rate": 0.00033769341269886225,
+      "loss": 0.4892,
+      "step": 4037
+    },
+    {
+      "epoch": 1.8517371861025111,
+      "grad_norm": 0.2396342158317566,
+      "learning_rate": 0.00033745920074189703,
+      "loss": 1.3491,
+      "step": 4038
+    },
+    {
+      "epoch": 1.852195849099874,
+      "grad_norm": 0.3997967541217804,
+      "learning_rate": 0.0003372250286573617,
+      "loss": 2.1116,
+      "step": 4039
+    },
+    {
+      "epoch": 1.8526545120972364,
+      "grad_norm": 0.27099528908729553,
+      "learning_rate": 0.00033699089650270054,
+      "loss": 1.0897,
+      "step": 4040
+    },
+    {
+      "epoch": 1.8531131750945993,
+      "grad_norm": 0.29823604226112366,
+      "learning_rate": 0.00033675680433534785,
+      "loss": 1.5873,
+      "step": 4041
+    },
+    {
+      "epoch": 1.853571838091962,
+      "grad_norm": 0.30324751138687134,
+      "learning_rate": 0.0003365227522127281,
+      "loss": 1.3735,
+      "step": 4042
+    },
+    {
+      "epoch": 1.8540305010893245,
+      "grad_norm": 0.2946029305458069,
+      "learning_rate": 0.000336288740192256,
+      "loss": 1.4114,
+      "step": 4043
+    },
+    {
+      "epoch": 1.8544891640866874,
+      "grad_norm": 0.2893146872520447,
+      "learning_rate": 0.0003360547683313363,
+      "loss": 2.0024,
+      "step": 4044
+    },
+    {
+      "epoch": 1.85494782708405,
+      "grad_norm": 0.37933430075645447,
+      "learning_rate": 0.00033582083668736405,
+      "loss": 1.7002,
+      "step": 4045
+    },
+    {
+      "epoch": 1.8554064900814127,
+      "grad_norm": 0.27654582262039185,
+      "learning_rate": 0.0003355869453177244,
+      "loss": 0.7906,
+      "step": 4046
+    },
+    {
+      "epoch": 1.8558651530787755,
+      "grad_norm": 0.09229593724012375,
+      "learning_rate": 0.00033535309427979245,
+      "loss": 1.09,
+      "step": 4047
+    },
+    {
+      "epoch": 1.856323816076138,
+      "grad_norm": 0.3510340750217438,
+      "learning_rate": 0.0003351192836309336,
+      "loss": 1.3514,
+      "step": 4048
+    },
+    {
+      "epoch": 1.8567824790735008,
+      "grad_norm": 0.2260064333677292,
+      "learning_rate": 0.0003348855134285034,
+      "loss": 0.8711,
+      "step": 4049
+    },
+    {
+      "epoch": 1.8572411420708634,
+      "grad_norm": 0.27447935938835144,
+      "learning_rate": 0.0003346517837298474,
+      "loss": 1.7773,
+      "step": 4050
+    },
+    {
+      "epoch": 1.857699805068226,
+      "grad_norm": 0.2299553006887436,
+      "learning_rate": 0.000334418094592301,
+      "loss": 0.9055,
+      "step": 4051
+    },
+    {
+      "epoch": 1.858158468065589,
+      "grad_norm": 0.2812572121620178,
+      "learning_rate": 0.0003341844460731899,
+      "loss": 1.3986,
+      "step": 4052
+    },
+    {
+      "epoch": 1.8586171310629513,
+      "grad_norm": 0.18786439299583435,
+      "learning_rate": 0.0003339508382298297,
+      "loss": 0.7558,
+      "step": 4053
+    },
+    {
+      "epoch": 1.8590757940603142,
+      "grad_norm": 0.06244681030511856,
+      "learning_rate": 0.0003337172711195262,
+      "loss": 0.9041,
+      "step": 4054
+    },
+    {
+      "epoch": 1.8595344570576768,
+      "grad_norm": 0.3086715042591095,
+      "learning_rate": 0.00033348374479957513,
+      "loss": 1.2524,
+      "step": 4055
+    },
+    {
+      "epoch": 1.8599931200550395,
+      "grad_norm": 0.1853373497724533,
+      "learning_rate": 0.000333250259327262,
+      "loss": 0.4612,
+      "step": 4056
+    },
+    {
+      "epoch": 1.8604517830524023,
+      "grad_norm": 0.24042657017707825,
+      "learning_rate": 0.0003330168147598626,
+      "loss": 1.0255,
+      "step": 4057
+    },
+    {
+      "epoch": 1.860910446049765,
+      "grad_norm": 0.36561837792396545,
+      "learning_rate": 0.00033278341115464263,
+      "loss": 1.3232,
+      "step": 4058
+    },
+    {
+      "epoch": 1.8613691090471276,
+      "grad_norm": 0.2835361361503601,
+      "learning_rate": 0.0003325500485688575,
+      "loss": 1.1592,
+      "step": 4059
+    },
+    {
+      "epoch": 1.8618277720444905,
+      "grad_norm": 0.3773711919784546,
+      "learning_rate": 0.0003323167270597528,
+      "loss": 1.4202,
+      "step": 4060
+    },
+    {
+      "epoch": 1.8622864350418529,
+      "grad_norm": 0.28732502460479736,
+      "learning_rate": 0.00033208344668456417,
+      "loss": 1.6362,
+      "step": 4061
+    },
+    {
+      "epoch": 1.8627450980392157,
+      "grad_norm": 0.35971352458000183,
+      "learning_rate": 0.00033185020750051673,
+      "loss": 1.4979,
+      "step": 4062
+    },
+    {
+      "epoch": 1.8632037610365784,
+      "grad_norm": 0.22411774098873138,
+      "learning_rate": 0.00033161700956482574,
+      "loss": 0.8292,
+      "step": 4063
+    },
+    {
+      "epoch": 1.863662424033941,
+      "grad_norm": 0.2379181832075119,
+      "learning_rate": 0.00033138385293469654,
+      "loss": 1.1412,
+      "step": 4064
+    },
+    {
+      "epoch": 1.8641210870313039,
+      "grad_norm": 0.2763615548610687,
+      "learning_rate": 0.00033115073766732376,
+      "loss": 1.2769,
+      "step": 4065
+    },
+    {
+      "epoch": 1.8645797500286663,
+      "grad_norm": 0.23881085216999054,
+      "learning_rate": 0.0003309176638198929,
+      "loss": 0.9071,
+      "step": 4066
+    },
+    {
+      "epoch": 1.8650384130260291,
+      "grad_norm": 0.3810456097126007,
+      "learning_rate": 0.000330684631449578,
+      "loss": 1.6372,
+      "step": 4067
+    },
+    {
+      "epoch": 1.8654970760233918,
+      "grad_norm": 0.310472309589386,
+      "learning_rate": 0.0003304516406135438,
+      "loss": 1.0613,
+      "step": 4068
+    },
+    {
+      "epoch": 1.8659557390207544,
+      "grad_norm": 0.15861466526985168,
+      "learning_rate": 0.0003302186913689448,
+      "loss": 0.6324,
+      "step": 4069
+    },
+    {
+      "epoch": 1.8664144020181173,
+      "grad_norm": 0.22283266484737396,
+      "learning_rate": 0.00032998578377292474,
+      "loss": 0.9156,
+      "step": 4070
+    },
+    {
+      "epoch": 1.86687306501548,
+      "grad_norm": 1.139766812324524,
+      "learning_rate": 0.00032975291788261794,
+      "loss": 1.3506,
+      "step": 4071
+    },
+    {
+      "epoch": 1.8673317280128425,
+      "grad_norm": 0.41252392530441284,
+      "learning_rate": 0.00032952009375514815,
+      "loss": 1.6586,
+      "step": 4072
+    },
+    {
+      "epoch": 1.8677903910102054,
+      "grad_norm": 0.1836547702550888,
+      "learning_rate": 0.00032928731144762837,
+      "loss": 0.9196,
+      "step": 4073
+    },
+    {
+      "epoch": 1.8682490540075678,
+      "grad_norm": 0.32911989092826843,
+      "learning_rate": 0.00032905457101716195,
+      "loss": 1.861,
+      "step": 4074
+    },
+    {
+      "epoch": 1.8687077170049307,
+      "grad_norm": 0.3695148825645447,
+      "learning_rate": 0.00032882187252084185,
+      "loss": 2.0391,
+      "step": 4075
+    },
+    {
+      "epoch": 1.8691663800022933,
+      "grad_norm": 0.4213230311870575,
+      "learning_rate": 0.0003285892160157507,
+      "loss": 1.6511,
+      "step": 4076
+    },
+    {
+      "epoch": 1.869625042999656,
+      "grad_norm": 0.24618668854236603,
+      "learning_rate": 0.0003283566015589608,
+      "loss": 1.6064,
+      "step": 4077
+    },
+    {
+      "epoch": 1.8700837059970188,
+      "grad_norm": 0.43131116032600403,
+      "learning_rate": 0.00032812402920753434,
+      "loss": 1.8835,
+      "step": 4078
+    },
+    {
+      "epoch": 1.8705423689943814,
+      "grad_norm": 0.3532734811306,
+      "learning_rate": 0.00032789149901852265,
+      "loss": 0.9022,
+      "step": 4079
+    },
+    {
+      "epoch": 1.871001031991744,
+      "grad_norm": 0.21644118428230286,
+      "learning_rate": 0.00032765901104896714,
+      "loss": 1.0367,
+      "step": 4080
+    },
+    {
+      "epoch": 1.871459694989107,
+      "grad_norm": 0.13179314136505127,
+      "learning_rate": 0.0003274265653558989,
+      "loss": 1.123,
+      "step": 4081
+    },
+    {
+      "epoch": 1.8719183579864693,
+      "grad_norm": 0.3340977430343628,
+      "learning_rate": 0.00032719416199633843,
+      "loss": 1.6467,
+      "step": 4082
+    },
+    {
+      "epoch": 1.8723770209838322,
+      "grad_norm": 0.3839796185493469,
+      "learning_rate": 0.0003269618010272963,
+      "loss": 2.147,
+      "step": 4083
+    },
+    {
+      "epoch": 1.8728356839811948,
+      "grad_norm": 0.42947497963905334,
+      "learning_rate": 0.0003267294825057719,
+      "loss": 1.448,
+      "step": 4084
+    },
+    {
+      "epoch": 1.8732943469785575,
+      "grad_norm": 0.3551345765590668,
+      "learning_rate": 0.0003264972064887546,
+      "loss": 1.4087,
+      "step": 4085
+    },
+    {
+      "epoch": 1.8737530099759203,
+      "grad_norm": 0.20159225165843964,
+      "learning_rate": 0.0003262649730332237,
+      "loss": 0.8688,
+      "step": 4086
+    },
+    {
+      "epoch": 1.8742116729732827,
+      "grad_norm": 0.17780308425426483,
+      "learning_rate": 0.0003260327821961476,
+      "loss": 0.6803,
+      "step": 4087
+    },
+    {
+      "epoch": 1.8746703359706456,
+      "grad_norm": 0.27476388216018677,
+      "learning_rate": 0.0003258006340344845,
+      "loss": 1.5988,
+      "step": 4088
+    },
+    {
+      "epoch": 1.8751289989680082,
+      "grad_norm": 0.2936633229255676,
+      "learning_rate": 0.000325568528605182,
+      "loss": 1.5938,
+      "step": 4089
+    },
+    {
+      "epoch": 1.8755876619653709,
+      "grad_norm": 0.24134798347949982,
+      "learning_rate": 0.00032533646596517683,
+      "loss": 0.5465,
+      "step": 4090
+    },
+    {
+      "epoch": 1.8760463249627337,
+      "grad_norm": 0.23964928090572357,
+      "learning_rate": 0.0003251044461713961,
+      "loss": 1.5688,
+      "step": 4091
+    },
+    {
+      "epoch": 1.8765049879600963,
+      "grad_norm": 0.4487531781196594,
+      "learning_rate": 0.0003248724692807558,
+      "loss": 1.3783,
+      "step": 4092
+    },
+    {
+      "epoch": 1.876963650957459,
+      "grad_norm": 0.2621656358242035,
+      "learning_rate": 0.00032464053535016145,
+      "loss": 1.6387,
+      "step": 4093
+    },
+    {
+      "epoch": 1.8774223139548218,
+      "grad_norm": 0.5254698395729065,
+      "learning_rate": 0.0003244086444365085,
+      "loss": 1.5309,
+      "step": 4094
+    },
+    {
+      "epoch": 1.8778809769521843,
+      "grad_norm": 0.4006554186344147,
+      "learning_rate": 0.00032417679659668073,
+      "loss": 1.6442,
+      "step": 4095
+    },
+    {
+      "epoch": 1.8783396399495471,
+      "grad_norm": 0.44737085700035095,
+      "learning_rate": 0.00032394499188755267,
+      "loss": 1.443,
+      "step": 4096
+    },
+    {
+      "epoch": 1.8787983029469097,
+      "grad_norm": 0.5100827217102051,
+      "learning_rate": 0.0003237132303659875,
+      "loss": 2.3247,
+      "step": 4097
+    },
+    {
+      "epoch": 1.8792569659442724,
+      "grad_norm": 0.43226632475852966,
+      "learning_rate": 0.00032348151208883805,
+      "loss": 2.1605,
+      "step": 4098
+    },
+    {
+      "epoch": 1.8797156289416352,
+      "grad_norm": 0.35672900080680847,
+      "learning_rate": 0.0003232498371129464,
+      "loss": 1.7168,
+      "step": 4099
+    },
+    {
+      "epoch": 1.8801742919389977,
+      "grad_norm": 0.3012166917324066,
+      "learning_rate": 0.0003230182054951443,
+      "loss": 1.6538,
+      "step": 4100
+    },
+    {
+      "epoch": 1.8806329549363605,
+      "grad_norm": 0.31628987193107605,
+      "learning_rate": 0.00032278661729225234,
+      "loss": 0.8449,
+      "step": 4101
+    },
+    {
+      "epoch": 1.8810916179337231,
+      "grad_norm": 0.25368666648864746,
+      "learning_rate": 0.000322555072561081,
+      "loss": 1.324,
+      "step": 4102
+    },
+    {
+      "epoch": 1.8815502809310858,
+      "grad_norm": 0.21288840472698212,
+      "learning_rate": 0.0003223235713584297,
+      "loss": 0.9169,
+      "step": 4103
+    },
+    {
+      "epoch": 1.8820089439284486,
+      "grad_norm": 0.24230335652828217,
+      "learning_rate": 0.00032209211374108746,
+      "loss": 1.0717,
+      "step": 4104
+    },
+    {
+      "epoch": 1.8824676069258113,
+      "grad_norm": 0.24073179066181183,
+      "learning_rate": 0.0003218606997658326,
+      "loss": 1.5495,
+      "step": 4105
+    },
+    {
+      "epoch": 1.882926269923174,
+      "grad_norm": 0.2884657680988312,
+      "learning_rate": 0.00032162932948943257,
+      "loss": 0.7274,
+      "step": 4106
+    },
+    {
+      "epoch": 1.8833849329205368,
+      "grad_norm": 0.2240462303161621,
+      "learning_rate": 0.0003213980029686441,
+      "loss": 1.6514,
+      "step": 4107
+    },
+    {
+      "epoch": 1.8838435959178992,
+      "grad_norm": 0.3176068663597107,
+      "learning_rate": 0.0003211667202602132,
+      "loss": 1.1339,
+      "step": 4108
+    },
+    {
+      "epoch": 1.884302258915262,
+      "grad_norm": 0.2350878119468689,
+      "learning_rate": 0.0003209354814208754,
+      "loss": 0.9783,
+      "step": 4109
+    },
+    {
+      "epoch": 1.8847609219126247,
+      "grad_norm": 0.16942141950130463,
+      "learning_rate": 0.00032070428650735506,
+      "loss": 1.2051,
+      "step": 4110
+    },
+    {
+      "epoch": 1.8852195849099873,
+      "grad_norm": 0.2659427523612976,
+      "learning_rate": 0.0003204731355763661,
+      "loss": 1.3427,
+      "step": 4111
+    },
+    {
+      "epoch": 1.8856782479073502,
+      "grad_norm": 0.2091580033302307,
+      "learning_rate": 0.00032024202868461137,
+      "loss": 1.1494,
+      "step": 4112
+    },
+    {
+      "epoch": 1.8861369109047128,
+      "grad_norm": 0.3054332733154297,
+      "learning_rate": 0.0003200109658887831,
+      "loss": 1.0687,
+      "step": 4113
+    },
+    {
+      "epoch": 1.8865955739020754,
+      "grad_norm": 0.23581813275814056,
+      "learning_rate": 0.0003197799472455627,
+      "loss": 1.0554,
+      "step": 4114
+    },
+    {
+      "epoch": 1.8870542368994383,
+      "grad_norm": 0.3387794494628906,
+      "learning_rate": 0.0003195489728116207,
+      "loss": 2.1521,
+      "step": 4115
+    },
+    {
+      "epoch": 1.8875128998968007,
+      "grad_norm": 0.29953324794769287,
+      "learning_rate": 0.00031931804264361674,
+      "loss": 0.8916,
+      "step": 4116
+    },
+    {
+      "epoch": 1.8879715628941636,
+      "grad_norm": 0.2123539000749588,
+      "learning_rate": 0.0003190871567981999,
+      "loss": 1.2987,
+      "step": 4117
+    },
+    {
+      "epoch": 1.8884302258915262,
+      "grad_norm": 0.40169602632522583,
+      "learning_rate": 0.00031885631533200775,
+      "loss": 1.4036,
+      "step": 4118
+    },
+    {
+      "epoch": 1.8888888888888888,
+      "grad_norm": 0.2332458198070526,
+      "learning_rate": 0.00031862551830166765,
+      "loss": 1.162,
+      "step": 4119
+    },
+    {
+      "epoch": 1.8893475518862517,
+      "grad_norm": 0.27471116185188293,
+      "learning_rate": 0.0003183947657637957,
+      "loss": 1.3871,
+      "step": 4120
+    },
+    {
+      "epoch": 1.889806214883614,
+      "grad_norm": 0.3699858784675598,
+      "learning_rate": 0.00031816405777499704,
+      "loss": 1.5227,
+      "step": 4121
+    },
+    {
+      "epoch": 1.890264877880977,
+      "grad_norm": 0.27009811997413635,
+      "learning_rate": 0.0003179333943918663,
+      "loss": 1.3455,
+      "step": 4122
+    },
+    {
+      "epoch": 1.8907235408783396,
+      "grad_norm": 0.2861959934234619,
+      "learning_rate": 0.00031770277567098654,
+      "loss": 1.7307,
+      "step": 4123
+    },
+    {
+      "epoch": 1.8911822038757022,
+      "grad_norm": 0.31758224964141846,
+      "learning_rate": 0.0003174722016689303,
+      "loss": 1.1325,
+      "step": 4124
+    },
+    {
+      "epoch": 1.891640866873065,
+      "grad_norm": 0.32737424969673157,
+      "learning_rate": 0.0003172416724422592,
+      "loss": 1.3532,
+      "step": 4125
+    },
+    {
+      "epoch": 1.8920995298704277,
+      "grad_norm": 0.14036279916763306,
+      "learning_rate": 0.00031701118804752353,
+      "loss": 0.6664,
+      "step": 4126
+    },
+    {
+      "epoch": 1.8925581928677904,
+      "grad_norm": 0.19789431989192963,
+      "learning_rate": 0.0003167807485412629,
+      "loss": 1.0829,
+      "step": 4127
+    },
+    {
+      "epoch": 1.8930168558651532,
+      "grad_norm": 0.2573801577091217,
+      "learning_rate": 0.00031655035398000576,
+      "loss": 1.0196,
+      "step": 4128
+    },
+    {
+      "epoch": 1.8934755188625156,
+      "grad_norm": 0.20172810554504395,
+      "learning_rate": 0.00031632000442026947,
+      "loss": 1.053,
+      "step": 4129
+    },
+    {
+      "epoch": 1.8939341818598785,
+      "grad_norm": 0.33292561769485474,
+      "learning_rate": 0.00031608969991856053,
+      "loss": 1.9376,
+      "step": 4130
+    },
+    {
+      "epoch": 1.8943928448572411,
+      "grad_norm": 0.2850032150745392,
+      "learning_rate": 0.00031585944053137417,
+      "loss": 1.5627,
+      "step": 4131
+    },
+    {
+      "epoch": 1.8948515078546038,
+      "grad_norm": 0.46692344546318054,
+      "learning_rate": 0.0003156292263151949,
+      "loss": 1.7141,
+      "step": 4132
+    },
+    {
+      "epoch": 1.8953101708519666,
+      "grad_norm": 0.39476048946380615,
+      "learning_rate": 0.00031539905732649555,
+      "loss": 2.0542,
+      "step": 4133
+    },
+    {
+      "epoch": 1.8957688338493293,
+      "grad_norm": 0.4325892925262451,
+      "learning_rate": 0.00031516893362173884,
+      "loss": 1.1792,
+      "step": 4134
+    },
+    {
+      "epoch": 1.896227496846692,
+      "grad_norm": 0.22258636355400085,
+      "learning_rate": 0.0003149388552573752,
+      "loss": 1.4821,
+      "step": 4135
+    },
+    {
+      "epoch": 1.8966861598440545,
+      "grad_norm": 0.47724971175193787,
+      "learning_rate": 0.00031470882228984475,
+      "loss": 2.2678,
+      "step": 4136
+    },
+    {
+      "epoch": 1.8971448228414172,
+      "grad_norm": 0.26698240637779236,
+      "learning_rate": 0.0003144788347755763,
+      "loss": 1.0363,
+      "step": 4137
+    },
+    {
+      "epoch": 1.89760348583878,
+      "grad_norm": 0.3450622856616974,
+      "learning_rate": 0.0003142488927709871,
+      "loss": 2.0267,
+      "step": 4138
+    },
+    {
+      "epoch": 1.8980621488361427,
+      "grad_norm": 0.2508101463317871,
+      "learning_rate": 0.0003140189963324842,
+      "loss": 0.9557,
+      "step": 4139
+    },
+    {
+      "epoch": 1.8985208118335053,
+      "grad_norm": 0.2404715120792389,
+      "learning_rate": 0.0003137891455164623,
+      "loss": 1.6057,
+      "step": 4140
+    },
+    {
+      "epoch": 1.8989794748308682,
+      "grad_norm": 0.3351982533931732,
+      "learning_rate": 0.00031355934037930567,
+      "loss": 1.451,
+      "step": 4141
+    },
+    {
+      "epoch": 1.8994381378282306,
+      "grad_norm": 0.3819112479686737,
+      "learning_rate": 0.00031332958097738707,
+      "loss": 1.3151,
+      "step": 4142
+    },
+    {
+      "epoch": 1.8998968008255934,
+      "grad_norm": 0.25655218958854675,
+      "learning_rate": 0.00031309986736706826,
+      "loss": 1.1487,
+      "step": 4143
+    },
+    {
+      "epoch": 1.900355463822956,
+      "grad_norm": 0.18810351192951202,
+      "learning_rate": 0.00031287019960469966,
+      "loss": 1.4159,
+      "step": 4144
+    },
+    {
+      "epoch": 1.9008141268203187,
+      "grad_norm": 0.3258521854877472,
+      "learning_rate": 0.00031264057774662044,
+      "loss": 1.6307,
+      "step": 4145
+    },
+    {
+      "epoch": 1.9012727898176816,
+      "grad_norm": 0.309843510389328,
+      "learning_rate": 0.0003124110018491584,
+      "loss": 1.011,
+      "step": 4146
+    },
+    {
+      "epoch": 1.9017314528150442,
+      "grad_norm": 0.2843288481235504,
+      "learning_rate": 0.00031218147196863,
+      "loss": 1.2536,
+      "step": 4147
+    },
+    {
+      "epoch": 1.9021901158124068,
+      "grad_norm": 0.08047755807638168,
+      "learning_rate": 0.00031195198816134093,
+      "loss": 0.6959,
+      "step": 4148
+    },
+    {
+      "epoch": 1.9026487788097697,
+      "grad_norm": 0.4606052339076996,
+      "learning_rate": 0.000311722550483585,
+      "loss": 1.2046,
+      "step": 4149
+    },
+    {
+      "epoch": 1.903107441807132,
+      "grad_norm": 0.2677907645702362,
+      "learning_rate": 0.0003114931589916452,
+      "loss": 1.6278,
+      "step": 4150
+    },
+    {
+      "epoch": 1.903566104804495,
+      "grad_norm": 0.32580938935279846,
+      "learning_rate": 0.0003112638137417925,
+      "loss": 1.4391,
+      "step": 4151
+    },
+    {
+      "epoch": 1.9040247678018576,
+      "grad_norm": 0.23725193738937378,
+      "learning_rate": 0.00031103451479028713,
+      "loss": 1.2067,
+      "step": 4152
+    },
+    {
+      "epoch": 1.9044834307992202,
+      "grad_norm": 0.3534078299999237,
+      "learning_rate": 0.0003108052621933778,
+      "loss": 1.437,
+      "step": 4153
+    },
+    {
+      "epoch": 1.904942093796583,
+      "grad_norm": 0.395913690328598,
+      "learning_rate": 0.0003105760560073018,
+      "loss": 2.239,
+      "step": 4154
+    },
+    {
+      "epoch": 1.9054007567939455,
+      "grad_norm": 0.38620731234550476,
+      "learning_rate": 0.0003103468962882851,
+      "loss": 1.0384,
+      "step": 4155
+    },
+    {
+      "epoch": 1.9058594197913084,
+      "grad_norm": 0.27703335881233215,
+      "learning_rate": 0.00031011778309254247,
+      "loss": 1.6416,
+      "step": 4156
+    },
+    {
+      "epoch": 1.906318082788671,
+      "grad_norm": 0.23753665387630463,
+      "learning_rate": 0.00030988871647627645,
+      "loss": 0.7991,
+      "step": 4157
+    },
+    {
+      "epoch": 1.9067767457860336,
+      "grad_norm": 0.39035359025001526,
+      "learning_rate": 0.0003096596964956791,
+      "loss": 1.59,
+      "step": 4158
+    },
+    {
+      "epoch": 1.9072354087833965,
+      "grad_norm": 0.22406819462776184,
+      "learning_rate": 0.00030943072320693067,
+      "loss": 0.8857,
+      "step": 4159
+    },
+    {
+      "epoch": 1.9076940717807591,
+      "grad_norm": 0.36969655752182007,
+      "learning_rate": 0.00030920179666619986,
+      "loss": 1.712,
+      "step": 4160
+    },
+    {
+      "epoch": 1.9081527347781218,
+      "grad_norm": 0.3277752697467804,
+      "learning_rate": 0.0003089729169296444,
+      "loss": 1.5408,
+      "step": 4161
+    },
+    {
+      "epoch": 1.9086113977754846,
+      "grad_norm": 0.4305424392223358,
+      "learning_rate": 0.0003087440840534093,
+      "loss": 1.7606,
+      "step": 4162
+    },
+    {
+      "epoch": 1.909070060772847,
+      "grad_norm": 0.2649680972099304,
+      "learning_rate": 0.0003085152980936296,
+      "loss": 1.4221,
+      "step": 4163
+    },
+    {
+      "epoch": 1.9095287237702099,
+      "grad_norm": 0.34129345417022705,
+      "learning_rate": 0.00030828655910642794,
+      "loss": 1.2801,
+      "step": 4164
+    },
+    {
+      "epoch": 1.9099873867675725,
+      "grad_norm": 0.16658979654312134,
+      "learning_rate": 0.0003080578671479157,
+      "loss": 0.7952,
+      "step": 4165
+    },
+    {
+      "epoch": 1.9104460497649352,
+      "grad_norm": 0.18683701753616333,
+      "learning_rate": 0.0003078292222741925,
+      "loss": 1.1949,
+      "step": 4166
+    },
+    {
+      "epoch": 1.910904712762298,
+      "grad_norm": 0.25396600365638733,
+      "learning_rate": 0.00030760062454134697,
+      "loss": 0.8212,
+      "step": 4167
+    },
+    {
+      "epoch": 1.9113633757596606,
+      "grad_norm": 0.2024562507867813,
+      "learning_rate": 0.0003073720740054553,
+      "loss": 1.2571,
+      "step": 4168
+    },
+    {
+      "epoch": 1.9118220387570233,
+      "grad_norm": 0.3227396309375763,
+      "learning_rate": 0.0003071435707225828,
+      "loss": 1.4965,
+      "step": 4169
+    },
+    {
+      "epoch": 1.912280701754386,
+      "grad_norm": 0.29104283452033997,
+      "learning_rate": 0.000306915114748783,
+      "loss": 1.8234,
+      "step": 4170
+    },
+    {
+      "epoch": 1.9127393647517485,
+      "grad_norm": 0.3544504642486572,
+      "learning_rate": 0.00030668670614009775,
+      "loss": 1.3737,
+      "step": 4171
+    },
+    {
+      "epoch": 1.9131980277491114,
+      "grad_norm": 0.14929145574569702,
+      "learning_rate": 0.0003064583449525574,
+      "loss": 1.0649,
+      "step": 4172
+    },
+    {
+      "epoch": 1.913656690746474,
+      "grad_norm": 0.2932626008987427,
+      "learning_rate": 0.0003062300312421806,
+      "loss": 1.3504,
+      "step": 4173
+    },
+    {
+      "epoch": 1.9141153537438367,
+      "grad_norm": 0.3478389084339142,
+      "learning_rate": 0.0003060017650649742,
+      "loss": 1.5471,
+      "step": 4174
+    },
+    {
+      "epoch": 1.9145740167411995,
+      "grad_norm": 0.30949702858924866,
+      "learning_rate": 0.00030577354647693354,
+      "loss": 1.1508,
+      "step": 4175
+    },
+    {
+      "epoch": 1.915032679738562,
+      "grad_norm": 0.2801455855369568,
+      "learning_rate": 0.0003055453755340425,
+      "loss": 1.7807,
+      "step": 4176
+    },
+    {
+      "epoch": 1.9154913427359248,
+      "grad_norm": 0.360272616147995,
+      "learning_rate": 0.0003053172522922729,
+      "loss": 1.7485,
+      "step": 4177
+    },
+    {
+      "epoch": 1.9159500057332874,
+      "grad_norm": 0.34427425265312195,
+      "learning_rate": 0.0003050891768075851,
+      "loss": 1.1602,
+      "step": 4178
+    },
+    {
+      "epoch": 1.91640866873065,
+      "grad_norm": 0.3189522922039032,
+      "learning_rate": 0.00030486114913592753,
+      "loss": 1.6614,
+      "step": 4179
+    },
+    {
+      "epoch": 1.916867331728013,
+      "grad_norm": 0.3376857340335846,
+      "learning_rate": 0.00030463316933323717,
+      "loss": 1.2032,
+      "step": 4180
+    },
+    {
+      "epoch": 1.9173259947253756,
+      "grad_norm": 0.36120155453681946,
+      "learning_rate": 0.00030440523745543893,
+      "loss": 1.8868,
+      "step": 4181
+    },
+    {
+      "epoch": 1.9177846577227382,
+      "grad_norm": 0.3722556531429291,
+      "learning_rate": 0.00030417735355844634,
+      "loss": 1.3034,
+      "step": 4182
+    },
+    {
+      "epoch": 1.918243320720101,
+      "grad_norm": 0.27059370279312134,
+      "learning_rate": 0.00030394951769816084,
+      "loss": 0.9223,
+      "step": 4183
+    },
+    {
+      "epoch": 1.9187019837174635,
+      "grad_norm": 0.3722488582134247,
+      "learning_rate": 0.0003037217299304723,
+      "loss": 2.0471,
+      "step": 4184
+    },
+    {
+      "epoch": 1.9191606467148263,
+      "grad_norm": 0.42110106348991394,
+      "learning_rate": 0.00030349399031125856,
+      "loss": 1.8804,
+      "step": 4185
+    },
+    {
+      "epoch": 1.919619309712189,
+      "grad_norm": 0.33694103360176086,
+      "learning_rate": 0.00030326629889638595,
+      "loss": 1.3992,
+      "step": 4186
+    },
+    {
+      "epoch": 1.9200779727095516,
+      "grad_norm": 0.26390311121940613,
+      "learning_rate": 0.00030303865574170876,
+      "loss": 1.6186,
+      "step": 4187
+    },
+    {
+      "epoch": 1.9205366357069145,
+      "grad_norm": 0.2693273723125458,
+      "learning_rate": 0.0003028110609030694,
+      "loss": 1.4586,
+      "step": 4188
+    },
+    {
+      "epoch": 1.9209952987042769,
+      "grad_norm": 0.4307970702648163,
+      "learning_rate": 0.0003025835144362987,
+      "loss": 1.1955,
+      "step": 4189
+    },
+    {
+      "epoch": 1.9214539617016397,
+      "grad_norm": 0.11909201741218567,
+      "learning_rate": 0.00030235601639721534,
+      "loss": 1.0518,
+      "step": 4190
+    },
+    {
+      "epoch": 1.9219126246990024,
+      "grad_norm": 0.36401188373565674,
+      "learning_rate": 0.00030212856684162613,
+      "loss": 0.9025,
+      "step": 4191
+    },
+    {
+      "epoch": 1.922371287696365,
+      "grad_norm": 0.22211536765098572,
+      "learning_rate": 0.00030190116582532627,
+      "loss": 1.1021,
+      "step": 4192
+    },
+    {
+      "epoch": 1.9228299506937279,
+      "grad_norm": 0.310380220413208,
+      "learning_rate": 0.0003016738134040988,
+      "loss": 1.709,
+      "step": 4193
+    },
+    {
+      "epoch": 1.9232886136910905,
+      "grad_norm": 0.36067163944244385,
+      "learning_rate": 0.00030144650963371487,
+      "loss": 1.153,
+      "step": 4194
+    },
+    {
+      "epoch": 1.9237472766884531,
+      "grad_norm": 0.2253832072019577,
+      "learning_rate": 0.00030121925456993396,
+      "loss": 1.1205,
+      "step": 4195
+    },
+    {
+      "epoch": 1.924205939685816,
+      "grad_norm": 0.3536555767059326,
+      "learning_rate": 0.000300992048268503,
+      "loss": 1.7184,
+      "step": 4196
+    },
+    {
+      "epoch": 1.9246646026831784,
+      "grad_norm": 0.0917428508400917,
+      "learning_rate": 0.0003007648907851576,
+      "loss": 1.0483,
+      "step": 4197
+    },
+    {
+      "epoch": 1.9251232656805413,
+      "grad_norm": 0.36443841457366943,
+      "learning_rate": 0.000300537782175621,
+      "loss": 1.7299,
+      "step": 4198
+    },
+    {
+      "epoch": 1.925581928677904,
+      "grad_norm": 0.33077698945999146,
+      "learning_rate": 0.0003003107224956046,
+      "loss": 1.6121,
+      "step": 4199
+    },
+    {
+      "epoch": 1.9260405916752665,
+      "grad_norm": 0.40068572759628296,
+      "learning_rate": 0.00030008371180080773,
+      "loss": 1.2148,
+      "step": 4200
+    },
+    {
+      "epoch": 1.9264992546726294,
+      "grad_norm": 0.363941490650177,
+      "learning_rate": 0.00029985675014691814,
+      "loss": 1.9509,
+      "step": 4201
+    },
+    {
+      "epoch": 1.926957917669992,
+      "grad_norm": 0.1912655234336853,
+      "learning_rate": 0.00029962983758961067,
+      "loss": 0.9309,
+      "step": 4202
+    },
+    {
+      "epoch": 1.9274165806673547,
+      "grad_norm": 0.2495778501033783,
+      "learning_rate": 0.0002994029741845488,
+      "loss": 1.5195,
+      "step": 4203
+    },
+    {
+      "epoch": 1.9278752436647173,
+      "grad_norm": 0.29733824729919434,
+      "learning_rate": 0.00029917615998738365,
+      "loss": 1.0684,
+      "step": 4204
+    },
+    {
+      "epoch": 1.92833390666208,
+      "grad_norm": 0.36461958289146423,
+      "learning_rate": 0.0002989493950537544,
+      "loss": 1.8616,
+      "step": 4205
+    },
+    {
+      "epoch": 1.9287925696594428,
+      "grad_norm": 0.3024968206882477,
+      "learning_rate": 0.0002987226794392885,
+      "loss": 1.166,
+      "step": 4206
+    },
+    {
+      "epoch": 1.9292512326568054,
+      "grad_norm": 0.32046911120414734,
+      "learning_rate": 0.0002984960131996004,
+      "loss": 1.0956,
+      "step": 4207
+    },
+    {
+      "epoch": 1.929709895654168,
+      "grad_norm": 0.1437264233827591,
+      "learning_rate": 0.00029826939639029324,
+      "loss": 1.0188,
+      "step": 4208
+    },
+    {
+      "epoch": 1.930168558651531,
+      "grad_norm": 0.342047780752182,
+      "learning_rate": 0.00029804282906695765,
+      "loss": 1.4002,
+      "step": 4209
+    },
+    {
+      "epoch": 1.9306272216488933,
+      "grad_norm": 0.27232903242111206,
+      "learning_rate": 0.0002978163112851722,
+      "loss": 1.7488,
+      "step": 4210
+    },
+    {
+      "epoch": 1.9310858846462562,
+      "grad_norm": 0.30967265367507935,
+      "learning_rate": 0.00029758984310050354,
+      "loss": 1.4791,
+      "step": 4211
+    },
+    {
+      "epoch": 1.9315445476436188,
+      "grad_norm": 0.3274848759174347,
+      "learning_rate": 0.000297363424568506,
+      "loss": 1.6986,
+      "step": 4212
+    },
+    {
+      "epoch": 1.9320032106409815,
+      "grad_norm": 0.3882467746734619,
+      "learning_rate": 0.0002971370557447213,
+      "loss": 1.7115,
+      "step": 4213
+    },
+    {
+      "epoch": 1.9324618736383443,
+      "grad_norm": 0.35071897506713867,
+      "learning_rate": 0.0002969107366846794,
+      "loss": 1.2036,
+      "step": 4214
+    },
+    {
+      "epoch": 1.932920536635707,
+      "grad_norm": 0.3669132590293884,
+      "learning_rate": 0.0002966844674438982,
+      "loss": 1.8774,
+      "step": 4215
+    },
+    {
+      "epoch": 1.9333791996330696,
+      "grad_norm": 0.3220601975917816,
+      "learning_rate": 0.00029645824807788325,
+      "loss": 1.8994,
+      "step": 4216
+    },
+    {
+      "epoch": 1.9338378626304324,
+      "grad_norm": 0.3539913594722748,
+      "learning_rate": 0.00029623207864212775,
+      "loss": 1.5546,
+      "step": 4217
+    },
+    {
+      "epoch": 1.9342965256277949,
+      "grad_norm": 0.2368205189704895,
+      "learning_rate": 0.00029600595919211247,
+      "loss": 1.272,
+      "step": 4218
+    },
+    {
+      "epoch": 1.9347551886251577,
+      "grad_norm": 0.38960981369018555,
+      "learning_rate": 0.00029577988978330615,
+      "loss": 2.0374,
+      "step": 4219
+    },
+    {
+      "epoch": 1.9352138516225204,
+      "grad_norm": 0.3929775059223175,
+      "learning_rate": 0.00029555387047116547,
+      "loss": 2.1378,
+      "step": 4220
+    },
+    {
+      "epoch": 1.935672514619883,
+      "grad_norm": 0.34435439109802246,
+      "learning_rate": 0.00029532790131113446,
+      "loss": 1.444,
+      "step": 4221
+    },
+    {
+      "epoch": 1.9361311776172458,
+      "grad_norm": 0.27234023809432983,
+      "learning_rate": 0.00029510198235864504,
+      "loss": 1.2824,
+      "step": 4222
+    },
+    {
+      "epoch": 1.9365898406146083,
+      "grad_norm": 0.30018359422683716,
+      "learning_rate": 0.000294876113669117,
+      "loss": 1.8326,
+      "step": 4223
+    },
+    {
+      "epoch": 1.9370485036119711,
+      "grad_norm": 0.46480369567871094,
+      "learning_rate": 0.00029465029529795696,
+      "loss": 1.6801,
+      "step": 4224
+    },
+    {
+      "epoch": 1.9375071666093338,
+      "grad_norm": 0.4117535948753357,
+      "learning_rate": 0.0002944245273005602,
+      "loss": 1.7745,
+      "step": 4225
+    },
+    {
+      "epoch": 1.9379658296066964,
+      "grad_norm": 0.19903844594955444,
+      "learning_rate": 0.00029419880973230916,
+      "loss": 0.8529,
+      "step": 4226
+    },
+    {
+      "epoch": 1.9384244926040592,
+      "grad_norm": 0.3728976249694824,
+      "learning_rate": 0.00029397314264857405,
+      "loss": 2.2271,
+      "step": 4227
+    },
+    {
+      "epoch": 1.9388831556014219,
+      "grad_norm": 0.4130072593688965,
+      "learning_rate": 0.00029374752610471255,
+      "loss": 1.9492,
+      "step": 4228
+    },
+    {
+      "epoch": 1.9393418185987845,
+      "grad_norm": 0.09748774021863937,
+      "learning_rate": 0.00029352196015607014,
+      "loss": 0.4857,
+      "step": 4229
+    },
+    {
+      "epoch": 1.9398004815961474,
+      "grad_norm": 0.289890319108963,
+      "learning_rate": 0.00029329644485797963,
+      "loss": 1.7548,
+      "step": 4230
+    },
+    {
+      "epoch": 1.9402591445935098,
+      "grad_norm": 0.33195921778678894,
+      "learning_rate": 0.00029307098026576156,
+      "loss": 1.767,
+      "step": 4231
+    },
+    {
+      "epoch": 1.9407178075908726,
+      "grad_norm": 0.30616116523742676,
+      "learning_rate": 0.0002928455664347241,
+      "loss": 1.5225,
+      "step": 4232
+    },
+    {
+      "epoch": 1.9411764705882353,
+      "grad_norm": 0.3174733519554138,
+      "learning_rate": 0.0002926202034201628,
+      "loss": 1.3201,
+      "step": 4233
+    },
+    {
+      "epoch": 1.941635133585598,
+      "grad_norm": 0.3208681643009186,
+      "learning_rate": 0.00029239489127736107,
+      "loss": 1.0652,
+      "step": 4234
+    },
+    {
+      "epoch": 1.9420937965829608,
+      "grad_norm": 0.21303045749664307,
+      "learning_rate": 0.0002921696300615893,
+      "loss": 1.0555,
+      "step": 4235
+    },
+    {
+      "epoch": 1.9425524595803234,
+      "grad_norm": 0.2299261838197708,
+      "learning_rate": 0.0002919444198281058,
+      "loss": 1.4859,
+      "step": 4236
+    },
+    {
+      "epoch": 1.943011122577686,
+      "grad_norm": 0.38345012068748474,
+      "learning_rate": 0.0002917192606321563,
+      "loss": 1.7592,
+      "step": 4237
+    },
+    {
+      "epoch": 1.9434697855750487,
+      "grad_norm": 0.36465907096862793,
+      "learning_rate": 0.0002914941525289739,
+      "loss": 1.1496,
+      "step": 4238
+    },
+    {
+      "epoch": 1.9439284485724113,
+      "grad_norm": 0.4449915885925293,
+      "learning_rate": 0.00029126909557377923,
+      "loss": 1.3264,
+      "step": 4239
+    },
+    {
+      "epoch": 1.9443871115697742,
+      "grad_norm": 0.31518861651420593,
+      "learning_rate": 0.0002910440898217808,
+      "loss": 1.6181,
+      "step": 4240
+    },
+    {
+      "epoch": 1.9448457745671368,
+      "grad_norm": 0.3502573072910309,
+      "learning_rate": 0.0002908191353281735,
+      "loss": 1.5842,
+      "step": 4241
+    },
+    {
+      "epoch": 1.9453044375644994,
+      "grad_norm": 0.3748527467250824,
+      "learning_rate": 0.00029059423214814053,
+      "loss": 1.6724,
+      "step": 4242
+    },
+    {
+      "epoch": 1.9457631005618623,
+      "grad_norm": 0.39455685019493103,
+      "learning_rate": 0.0002903693803368522,
+      "loss": 1.4643,
+      "step": 4243
+    },
+    {
+      "epoch": 1.9462217635592247,
+      "grad_norm": 0.14820732176303864,
+      "learning_rate": 0.00029014457994946654,
+      "loss": 0.9309,
+      "step": 4244
+    },
+    {
+      "epoch": 1.9466804265565876,
+      "grad_norm": 0.18288370966911316,
+      "learning_rate": 0.00028991983104112874,
+      "loss": 0.9465,
+      "step": 4245
+    },
+    {
+      "epoch": 1.9471390895539502,
+      "grad_norm": 0.38087254762649536,
+      "learning_rate": 0.00028969513366697096,
+      "loss": 1.2691,
+      "step": 4246
+    },
+    {
+      "epoch": 1.9475977525513128,
+      "grad_norm": 0.14294399321079254,
+      "learning_rate": 0.0002894704878821133,
+      "loss": 1.094,
+      "step": 4247
+    },
+    {
+      "epoch": 1.9480564155486757,
+      "grad_norm": 0.4238864779472351,
+      "learning_rate": 0.00028924589374166286,
+      "loss": 1.7693,
+      "step": 4248
+    },
+    {
+      "epoch": 1.9485150785460383,
+      "grad_norm": 0.2929639220237732,
+      "learning_rate": 0.0002890213513007144,
+      "loss": 1.2274,
+      "step": 4249
+    },
+    {
+      "epoch": 1.948973741543401,
+      "grad_norm": 0.25756093859672546,
+      "learning_rate": 0.00028879686061434966,
+      "loss": 1.2784,
+      "step": 4250
+    },
+    {
+      "epoch": 1.9494324045407638,
+      "grad_norm": 0.3350880742073059,
+      "learning_rate": 0.0002885724217376381,
+      "loss": 2.0134,
+      "step": 4251
+    },
+    {
+      "epoch": 1.9498910675381262,
+      "grad_norm": 0.2928470969200134,
+      "learning_rate": 0.00028834803472563574,
+      "loss": 0.9233,
+      "step": 4252
+    },
+    {
+      "epoch": 1.950349730535489,
+      "grad_norm": 0.35262230038642883,
+      "learning_rate": 0.0002881236996333864,
+      "loss": 2.0552,
+      "step": 4253
+    },
+    {
+      "epoch": 1.9508083935328517,
+      "grad_norm": 0.3223377466201782,
+      "learning_rate": 0.00028789941651592135,
+      "loss": 1.6767,
+      "step": 4254
+    },
+    {
+      "epoch": 1.9512670565302144,
+      "grad_norm": 0.3945949375629425,
+      "learning_rate": 0.0002876751854282588,
+      "loss": 1.5204,
+      "step": 4255
+    },
+    {
+      "epoch": 1.9517257195275772,
+      "grad_norm": 0.25893688201904297,
+      "learning_rate": 0.00028745100642540425,
+      "loss": 1.2777,
+      "step": 4256
+    },
+    {
+      "epoch": 1.9521843825249396,
+      "grad_norm": 0.34024423360824585,
+      "learning_rate": 0.0002872268795623505,
+      "loss": 1.7372,
+      "step": 4257
+    },
+    {
+      "epoch": 1.9526430455223025,
+      "grad_norm": 0.2754332423210144,
+      "learning_rate": 0.00028700280489407713,
+      "loss": 1.1841,
+      "step": 4258
+    },
+    {
+      "epoch": 1.9531017085196651,
+      "grad_norm": 0.3347947299480438,
+      "learning_rate": 0.0002867787824755516,
+      "loss": 1.3132,
+      "step": 4259
+    },
+    {
+      "epoch": 1.9535603715170278,
+      "grad_norm": 0.21994498372077942,
+      "learning_rate": 0.00028655481236172813,
+      "loss": 1.2384,
+      "step": 4260
+    },
+    {
+      "epoch": 1.9540190345143906,
+      "grad_norm": 0.2327728271484375,
+      "learning_rate": 0.00028633089460754826,
+      "loss": 1.3356,
+      "step": 4261
+    },
+    {
+      "epoch": 1.9544776975117533,
+      "grad_norm": 0.2511197030544281,
+      "learning_rate": 0.00028610702926794063,
+      "loss": 0.728,
+      "step": 4262
+    },
+    {
+      "epoch": 1.954936360509116,
+      "grad_norm": 0.2759820222854614,
+      "learning_rate": 0.0002858832163978207,
+      "loss": 1.6011,
+      "step": 4263
+    },
+    {
+      "epoch": 1.9553950235064788,
+      "grad_norm": 0.3019004166126251,
+      "learning_rate": 0.00028565945605209177,
+      "loss": 1.7163,
+      "step": 4264
+    },
+    {
+      "epoch": 1.9558536865038412,
+      "grad_norm": 0.3434310555458069,
+      "learning_rate": 0.0002854357482856438,
+      "loss": 1.338,
+      "step": 4265
+    },
+    {
+      "epoch": 1.956312349501204,
+      "grad_norm": 0.2556118369102478,
+      "learning_rate": 0.0002852120931533538,
+      "loss": 1.6398,
+      "step": 4266
+    },
+    {
+      "epoch": 1.9567710124985667,
+      "grad_norm": 0.39801543951034546,
+      "learning_rate": 0.0002849884907100861,
+      "loss": 1.4964,
+      "step": 4267
+    },
+    {
+      "epoch": 1.9572296754959293,
+      "grad_norm": 0.30779194831848145,
+      "learning_rate": 0.0002847649410106921,
+      "loss": 1.7112,
+      "step": 4268
+    },
+    {
+      "epoch": 1.9576883384932922,
+      "grad_norm": 0.29870155453681946,
+      "learning_rate": 0.0002845414441100098,
+      "loss": 1.6331,
+      "step": 4269
+    },
+    {
+      "epoch": 1.9581470014906548,
+      "grad_norm": 0.2507461905479431,
+      "learning_rate": 0.0002843180000628649,
+      "loss": 1.0191,
+      "step": 4270
+    },
+    {
+      "epoch": 1.9586056644880174,
+      "grad_norm": 0.25291094183921814,
+      "learning_rate": 0.00028409460892406957,
+      "loss": 0.802,
+      "step": 4271
+    },
+    {
+      "epoch": 1.95906432748538,
+      "grad_norm": 0.06607484817504883,
+      "learning_rate": 0.00028387127074842357,
+      "loss": 0.6714,
+      "step": 4272
+    },
+    {
+      "epoch": 1.9595229904827427,
+      "grad_norm": 0.30349382758140564,
+      "learning_rate": 0.00028364798559071315,
+      "loss": 1.3984,
+      "step": 4273
+    },
+    {
+      "epoch": 1.9599816534801056,
+      "grad_norm": 0.23646502196788788,
+      "learning_rate": 0.0002834247535057118,
+      "loss": 1.5465,
+      "step": 4274
+    },
+    {
+      "epoch": 1.9604403164774682,
+      "grad_norm": 0.32397279143333435,
+      "learning_rate": 0.00028320157454818015,
+      "loss": 1.2772,
+      "step": 4275
+    },
+    {
+      "epoch": 1.9608989794748308,
+      "grad_norm": 0.2407788336277008,
+      "learning_rate": 0.00028297844877286536,
+      "loss": 1.336,
+      "step": 4276
+    },
+    {
+      "epoch": 1.9613576424721937,
+      "grad_norm": 0.3070877194404602,
+      "learning_rate": 0.00028275537623450187,
+      "loss": 1.8708,
+      "step": 4277
+    },
+    {
+      "epoch": 1.961816305469556,
+      "grad_norm": 0.3111575245857239,
+      "learning_rate": 0.0002825323569878111,
+      "loss": 1.2437,
+      "step": 4278
+    },
+    {
+      "epoch": 1.962274968466919,
+      "grad_norm": 0.1594800353050232,
+      "learning_rate": 0.0002823093910875013,
+      "loss": 0.7383,
+      "step": 4279
+    },
+    {
+      "epoch": 1.9627336314642816,
+      "grad_norm": 0.23517662286758423,
+      "learning_rate": 0.0002820864785882673,
+      "loss": 1.308,
+      "step": 4280
+    },
+    {
+      "epoch": 1.9631922944616442,
+      "grad_norm": 0.2781253755092621,
+      "learning_rate": 0.0002818636195447913,
+      "loss": 1.3627,
+      "step": 4281
+    },
+    {
+      "epoch": 1.963650957459007,
+      "grad_norm": 0.2628926634788513,
+      "learning_rate": 0.0002816408140117424,
+      "loss": 1.2191,
+      "step": 4282
+    },
+    {
+      "epoch": 1.9641096204563697,
+      "grad_norm": 0.2107929289340973,
+      "learning_rate": 0.00028141806204377617,
+      "loss": 1.2504,
+      "step": 4283
+    },
+    {
+      "epoch": 1.9645682834537324,
+      "grad_norm": 0.27736082673072815,
+      "learning_rate": 0.0002811953636955354,
+      "loss": 1.0677,
+      "step": 4284
+    },
+    {
+      "epoch": 1.9650269464510952,
+      "grad_norm": 0.23589631915092468,
+      "learning_rate": 0.00028097271902164967,
+      "loss": 1.6243,
+      "step": 4285
+    },
+    {
+      "epoch": 1.9654856094484576,
+      "grad_norm": 0.458365261554718,
+      "learning_rate": 0.00028075012807673516,
+      "loss": 1.6266,
+      "step": 4286
+    },
+    {
+      "epoch": 1.9659442724458205,
+      "grad_norm": 0.25476112961769104,
+      "learning_rate": 0.0002805275909153951,
+      "loss": 1.7578,
+      "step": 4287
+    },
+    {
+      "epoch": 1.9664029354431831,
+      "grad_norm": 0.3246137797832489,
+      "learning_rate": 0.00028030510759221943,
+      "loss": 1.1602,
+      "step": 4288
+    },
+    {
+      "epoch": 1.9668615984405458,
+      "grad_norm": 0.3675731420516968,
+      "learning_rate": 0.000280082678161785,
+      "loss": 1.2238,
+      "step": 4289
+    },
+    {
+      "epoch": 1.9673202614379086,
+      "grad_norm": 0.3241674304008484,
+      "learning_rate": 0.00027986030267865546,
+      "loss": 1.2205,
+      "step": 4290
+    },
+    {
+      "epoch": 1.967778924435271,
+      "grad_norm": 0.25529178977012634,
+      "learning_rate": 0.00027963798119738084,
+      "loss": 1.7133,
+      "step": 4291
+    },
+    {
+      "epoch": 1.9682375874326339,
+      "grad_norm": 0.2785083055496216,
+      "learning_rate": 0.0002794157137724983,
+      "loss": 1.4641,
+      "step": 4292
+    },
+    {
+      "epoch": 1.9686962504299965,
+      "grad_norm": 0.39693623781204224,
+      "learning_rate": 0.00027919350045853167,
+      "loss": 2.1711,
+      "step": 4293
+    },
+    {
+      "epoch": 1.9691549134273592,
+      "grad_norm": 0.3063134253025055,
+      "learning_rate": 0.0002789713413099914,
+      "loss": 1.2238,
+      "step": 4294
+    },
+    {
+      "epoch": 1.969613576424722,
+      "grad_norm": 0.21434158086776733,
+      "learning_rate": 0.0002787492363813748,
+      "loss": 1.2683,
+      "step": 4295
+    },
+    {
+      "epoch": 1.9700722394220846,
+      "grad_norm": 0.26843902468681335,
+      "learning_rate": 0.00027852718572716594,
+      "loss": 1.1436,
+      "step": 4296
+    },
+    {
+      "epoch": 1.9705309024194473,
+      "grad_norm": 0.26620641350746155,
+      "learning_rate": 0.00027830518940183527,
+      "loss": 1.2504,
+      "step": 4297
+    },
+    {
+      "epoch": 1.9709895654168101,
+      "grad_norm": 0.4168534278869629,
+      "learning_rate": 0.0002780832474598401,
+      "loss": 1.4581,
+      "step": 4298
+    },
+    {
+      "epoch": 1.9714482284141726,
+      "grad_norm": 0.5389671325683594,
+      "learning_rate": 0.00027786135995562446,
+      "loss": 1.738,
+      "step": 4299
+    },
+    {
+      "epoch": 1.9719068914115354,
+      "grad_norm": 0.18300795555114746,
+      "learning_rate": 0.00027763952694361894,
+      "loss": 0.9293,
+      "step": 4300
+    },
+    {
+      "epoch": 1.972365554408898,
+      "grad_norm": 0.18415942788124084,
+      "learning_rate": 0.00027741774847824094,
+      "loss": 1.2104,
+      "step": 4301
+    },
+    {
+      "epoch": 1.9728242174062607,
+      "grad_norm": 0.3821175992488861,
+      "learning_rate": 0.00027719602461389394,
+      "loss": 1.6418,
+      "step": 4302
+    },
+    {
+      "epoch": 1.9732828804036235,
+      "grad_norm": 0.34809380769729614,
+      "learning_rate": 0.0002769743554049686,
+      "loss": 1.3178,
+      "step": 4303
+    },
+    {
+      "epoch": 1.9737415434009862,
+      "grad_norm": 0.34898874163627625,
+      "learning_rate": 0.00027675274090584195,
+      "loss": 0.9742,
+      "step": 4304
+    },
+    {
+      "epoch": 1.9742002063983488,
+      "grad_norm": 0.30163076519966125,
+      "learning_rate": 0.0002765311811708775,
+      "loss": 1.6653,
+      "step": 4305
+    },
+    {
+      "epoch": 1.9746588693957114,
+      "grad_norm": 0.37521055340766907,
+      "learning_rate": 0.0002763096762544258,
+      "loss": 1.4052,
+      "step": 4306
+    },
+    {
+      "epoch": 1.975117532393074,
+      "grad_norm": 0.36267635226249695,
+      "learning_rate": 0.0002760882262108236,
+      "loss": 1.627,
+      "step": 4307
+    },
+    {
+      "epoch": 1.975576195390437,
+      "grad_norm": 0.37913310527801514,
+      "learning_rate": 0.0002758668310943938,
+      "loss": 1.8204,
+      "step": 4308
+    },
+    {
+      "epoch": 1.9760348583877996,
+      "grad_norm": 0.29660242795944214,
+      "learning_rate": 0.00027564549095944636,
+      "loss": 1.3698,
+      "step": 4309
+    },
+    {
+      "epoch": 1.9764935213851622,
+      "grad_norm": 0.3772091865539551,
+      "learning_rate": 0.00027542420586027774,
+      "loss": 1.7451,
+      "step": 4310
+    },
+    {
+      "epoch": 1.976952184382525,
+      "grad_norm": 0.32186251878738403,
+      "learning_rate": 0.0002752029758511707,
+      "loss": 1.7236,
+      "step": 4311
+    },
+    {
+      "epoch": 1.9774108473798875,
+      "grad_norm": 0.2753252685070038,
+      "learning_rate": 0.0002749818009863945,
+      "loss": 1.0989,
+      "step": 4312
+    },
+    {
+      "epoch": 1.9778695103772503,
+      "grad_norm": 0.30564045906066895,
+      "learning_rate": 0.0002747606813202052,
+      "loss": 1.3082,
+      "step": 4313
+    },
+    {
+      "epoch": 1.978328173374613,
+      "grad_norm": 0.24932487308979034,
+      "learning_rate": 0.0002745396169068447,
+      "loss": 0.9138,
+      "step": 4314
+    },
+    {
+      "epoch": 1.9787868363719756,
+      "grad_norm": 0.26185494661331177,
+      "learning_rate": 0.0002743186078005415,
+      "loss": 1.2475,
+      "step": 4315
+    },
+    {
+      "epoch": 1.9792454993693385,
+      "grad_norm": 0.2203855961561203,
+      "learning_rate": 0.00027409765405551136,
+      "loss": 1.3713,
+      "step": 4316
+    },
+    {
+      "epoch": 1.979704162366701,
+      "grad_norm": 0.3790103495121002,
+      "learning_rate": 0.0002738767557259555,
+      "loss": 1.4261,
+      "step": 4317
+    },
+    {
+      "epoch": 1.9801628253640637,
+      "grad_norm": 0.3733806312084198,
+      "learning_rate": 0.0002736559128660621,
+      "loss": 2.1038,
+      "step": 4318
+    },
+    {
+      "epoch": 1.9806214883614266,
+      "grad_norm": 0.20003962516784668,
+      "learning_rate": 0.00027343512553000505,
+      "loss": 0.7565,
+      "step": 4319
+    },
+    {
+      "epoch": 1.981080151358789,
+      "grad_norm": 0.2958119511604309,
+      "learning_rate": 0.00027321439377194537,
+      "loss": 0.9704,
+      "step": 4320
+    },
+    {
+      "epoch": 1.9815388143561519,
+      "grad_norm": 0.2841302752494812,
+      "learning_rate": 0.00027299371764603,
+      "loss": 1.327,
+      "step": 4321
+    },
+    {
+      "epoch": 1.9819974773535145,
+      "grad_norm": 0.3038768172264099,
+      "learning_rate": 0.00027277309720639265,
+      "loss": 1.2805,
+      "step": 4322
+    },
+    {
+      "epoch": 1.9824561403508771,
+      "grad_norm": 0.3406788110733032,
+      "learning_rate": 0.00027255253250715286,
+      "loss": 1.4304,
+      "step": 4323
+    },
+    {
+      "epoch": 1.98291480334824,
+      "grad_norm": 0.2855389416217804,
+      "learning_rate": 0.000272332023602417,
+      "loss": 1.6539,
+      "step": 4324
+    },
+    {
+      "epoch": 1.9833734663456024,
+      "grad_norm": 0.23860915005207062,
+      "learning_rate": 0.000272111570546277,
+      "loss": 1.3198,
+      "step": 4325
+    },
+    {
+      "epoch": 1.9838321293429653,
+      "grad_norm": 0.3688766360282898,
+      "learning_rate": 0.0002718911733928121,
+      "loss": 1.1564,
+      "step": 4326
+    },
+    {
+      "epoch": 1.984290792340328,
+      "grad_norm": 0.15739460289478302,
+      "learning_rate": 0.00027167083219608706,
+      "loss": 0.6855,
+      "step": 4327
+    },
+    {
+      "epoch": 1.9847494553376905,
+      "grad_norm": 0.2910972535610199,
+      "learning_rate": 0.0002714505470101533,
+      "loss": 1.7732,
+      "step": 4328
+    },
+    {
+      "epoch": 1.9852081183350534,
+      "grad_norm": 0.3392382562160492,
+      "learning_rate": 0.0002712303178890484,
+      "loss": 1.2592,
+      "step": 4329
+    },
+    {
+      "epoch": 1.985666781332416,
+      "grad_norm": 0.27205178141593933,
+      "learning_rate": 0.0002710101448867959,
+      "loss": 1.304,
+      "step": 4330
+    },
+    {
+      "epoch": 1.9861254443297787,
+      "grad_norm": 0.31658264994621277,
+      "learning_rate": 0.0002707900280574059,
+      "loss": 1.6132,
+      "step": 4331
+    },
+    {
+      "epoch": 1.9865841073271415,
+      "grad_norm": 0.34457242488861084,
+      "learning_rate": 0.00027056996745487475,
+      "loss": 1.6532,
+      "step": 4332
+    },
+    {
+      "epoch": 1.987042770324504,
+      "grad_norm": 0.27797314524650574,
+      "learning_rate": 0.00027034996313318483,
+      "loss": 1.3838,
+      "step": 4333
+    },
+    {
+      "epoch": 1.9875014333218668,
+      "grad_norm": 0.2671056091785431,
+      "learning_rate": 0.00027013001514630483,
+      "loss": 1.34,
+      "step": 4334
+    },
+    {
+      "epoch": 1.9879600963192294,
+      "grad_norm": 0.2947138845920563,
+      "learning_rate": 0.0002699101235481896,
+      "loss": 1.0969,
+      "step": 4335
+    },
+    {
+      "epoch": 1.988418759316592,
+      "grad_norm": 0.3621201515197754,
+      "learning_rate": 0.00026969028839278,
+      "loss": 1.7797,
+      "step": 4336
+    },
+    {
+      "epoch": 1.988877422313955,
+      "grad_norm": 0.2773393988609314,
+      "learning_rate": 0.00026947050973400333,
+      "loss": 1.5847,
+      "step": 4337
+    },
+    {
+      "epoch": 1.9893360853113176,
+      "grad_norm": 0.24329693615436554,
+      "learning_rate": 0.00026925078762577283,
+      "loss": 1.2757,
+      "step": 4338
+    },
+    {
+      "epoch": 1.9897947483086802,
+      "grad_norm": 0.3237564265727997,
+      "learning_rate": 0.00026903112212198796,
+      "loss": 1.1135,
+      "step": 4339
+    },
+    {
+      "epoch": 1.9902534113060428,
+      "grad_norm": 0.21693646907806396,
+      "learning_rate": 0.0002688115132765344,
+      "loss": 1.2356,
+      "step": 4340
+    },
+    {
+      "epoch": 1.9907120743034055,
+      "grad_norm": 0.27011099457740784,
+      "learning_rate": 0.00026859196114328333,
+      "loss": 1.6289,
+      "step": 4341
+    },
+    {
+      "epoch": 1.9911707373007683,
+      "grad_norm": 0.34364983439445496,
+      "learning_rate": 0.0002683724657760928,
+      "loss": 1.2663,
+      "step": 4342
+    },
+    {
+      "epoch": 1.991629400298131,
+      "grad_norm": 0.189200758934021,
+      "learning_rate": 0.00026815302722880643,
+      "loss": 0.7483,
+      "step": 4343
+    },
+    {
+      "epoch": 1.9920880632954936,
+      "grad_norm": 0.23307786881923676,
+      "learning_rate": 0.00026793364555525426,
+      "loss": 0.8417,
+      "step": 4344
+    },
+    {
+      "epoch": 1.9925467262928565,
+      "grad_norm": 0.23828265070915222,
+      "learning_rate": 0.00026771432080925205,
+      "loss": 0.7974,
+      "step": 4345
+    },
+    {
+      "epoch": 1.9930053892902189,
+      "grad_norm": 0.22147458791732788,
+      "learning_rate": 0.0002674950530446019,
+      "loss": 0.8344,
+      "step": 4346
+    },
+    {
+      "epoch": 1.9934640522875817,
+      "grad_norm": 0.2830180525779724,
+      "learning_rate": 0.0002672758423150916,
+      "loss": 1.2985,
+      "step": 4347
+    },
+    {
+      "epoch": 1.9939227152849444,
+      "grad_norm": 0.3324924409389496,
+      "learning_rate": 0.0002670566886744953,
+      "loss": 1.7325,
+      "step": 4348
+    },
+    {
+      "epoch": 1.994381378282307,
+      "grad_norm": 0.18930895626544952,
+      "learning_rate": 0.0002668375921765728,
+      "loss": 0.7449,
+      "step": 4349
+    },
+    {
+      "epoch": 1.9948400412796699,
+      "grad_norm": 0.28275609016418457,
+      "learning_rate": 0.0002666185528750702,
+      "loss": 1.7032,
+      "step": 4350
+    },
+    {
+      "epoch": 1.9952987042770325,
+      "grad_norm": 0.2890438735485077,
+      "learning_rate": 0.00026639957082371936,
+      "loss": 1.2205,
+      "step": 4351
+    },
+    {
+      "epoch": 1.9957573672743951,
+      "grad_norm": 0.36931562423706055,
+      "learning_rate": 0.0002661806460762383,
+      "loss": 1.7939,
+      "step": 4352
+    },
+    {
+      "epoch": 1.996216030271758,
+      "grad_norm": 0.3492710590362549,
+      "learning_rate": 0.0002659617786863304,
+      "loss": 1.6342,
+      "step": 4353
+    },
+    {
+      "epoch": 1.9966746932691204,
+      "grad_norm": 0.29765257239341736,
+      "learning_rate": 0.00026574296870768575,
+      "loss": 1.7397,
+      "step": 4354
+    },
+    {
+      "epoch": 1.9971333562664833,
+      "grad_norm": 0.37656140327453613,
+      "learning_rate": 0.00026552421619398004,
+      "loss": 2.0061,
+      "step": 4355
+    },
+    {
+      "epoch": 1.9975920192638459,
+      "grad_norm": 0.29920002818107605,
+      "learning_rate": 0.0002653055211988746,
+      "loss": 1.6586,
+      "step": 4356
+    },
+    {
+      "epoch": 1.9980506822612085,
+      "grad_norm": 0.34368252754211426,
+      "learning_rate": 0.0002650868837760172,
+      "loss": 1.3087,
+      "step": 4357
+    },
+    {
+      "epoch": 1.9985093452585714,
+      "grad_norm": 0.26091626286506653,
+      "learning_rate": 0.0002648683039790409,
+      "loss": 1.244,
+      "step": 4358
+    },
+    {
+      "epoch": 1.9989680082559338,
+      "grad_norm": 0.18554215133190155,
+      "learning_rate": 0.0002646497818615651,
+      "loss": 1.1157,
+      "step": 4359
+    },
+    {
+      "epoch": 1.9994266712532967,
+      "grad_norm": 0.23204344511032104,
+      "learning_rate": 0.00026443131747719474,
+      "loss": 1.2117,
+      "step": 4360
+    },
+    {
+      "epoch": 1.9998853342506593,
+      "grad_norm": 0.25021374225616455,
+      "learning_rate": 0.00026421291087952084,
+      "loss": 1.3657,
+      "step": 4361
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.25021374225616455,
+      "learning_rate": 0.00026421291087952084,
+      "loss": 0.5557,
+      "step": 4362
+    },
+    {
+      "epoch": 2.000458662997363,
+      "grad_norm": 0.31650856137275696,
+      "learning_rate": 0.0002639945621221199,
+      "loss": 1.4053,
+      "step": 4363
+    },
+    {
+      "epoch": 2.0009173259947253,
+      "grad_norm": 0.3828623592853546,
+      "learning_rate": 0.00026377627125855475,
+      "loss": 1.5593,
+      "step": 4364
+    },
+    {
+      "epoch": 2.001375988992088,
+      "grad_norm": 0.3297773003578186,
+      "learning_rate": 0.00026355803834237333,
+      "loss": 1.6141,
+      "step": 4365
+    },
+    {
+      "epoch": 2.0018346519894505,
+      "grad_norm": 0.20084944367408752,
+      "learning_rate": 0.00026333986342711,
+      "loss": 0.831,
+      "step": 4366
+    },
+    {
+      "epoch": 2.0022933149868134,
+      "grad_norm": 0.21461676061153412,
+      "learning_rate": 0.0002631217465662845,
+      "loss": 1.5271,
+      "step": 4367
+    },
+    {
+      "epoch": 2.0027519779841763,
+      "grad_norm": 0.3193070590496063,
+      "learning_rate": 0.00026290368781340263,
+      "loss": 1.1464,
+      "step": 4368
+    },
+    {
+      "epoch": 2.0032106409815387,
+      "grad_norm": 0.26611328125,
+      "learning_rate": 0.00026268568722195564,
+      "loss": 0.8123,
+      "step": 4369
+    },
+    {
+      "epoch": 2.0036693039789015,
+      "grad_norm": 0.07296108454465866,
+      "learning_rate": 0.0002624677448454207,
+      "loss": 0.8224,
+      "step": 4370
+    },
+    {
+      "epoch": 2.0041279669762644,
+      "grad_norm": 0.36370766162872314,
+      "learning_rate": 0.00026224986073726064,
+      "loss": 1.5906,
+      "step": 4371
+    },
+    {
+      "epoch": 2.004586629973627,
+      "grad_norm": 0.3442339301109314,
+      "learning_rate": 0.0002620320349509241,
+      "loss": 1.3951,
+      "step": 4372
+    },
+    {
+      "epoch": 2.0050452929709897,
+      "grad_norm": 0.2694573700428009,
+      "learning_rate": 0.0002618142675398451,
+      "loss": 0.851,
+      "step": 4373
+    },
+    {
+      "epoch": 2.005503955968352,
+      "grad_norm": 0.22100800275802612,
+      "learning_rate": 0.00026159655855744374,
+      "loss": 1.3334,
+      "step": 4374
+    },
+    {
+      "epoch": 2.005962618965715,
+      "grad_norm": 0.3315063714981079,
+      "learning_rate": 0.00026137890805712574,
+      "loss": 1.7025,
+      "step": 4375
+    },
+    {
+      "epoch": 2.006421281963078,
+      "grad_norm": 0.5389212369918823,
+      "learning_rate": 0.0002611613160922819,
+      "loss": 1.4756,
+      "step": 4376
+    },
+    {
+      "epoch": 2.00687994496044,
+      "grad_norm": 0.3037826120853424,
+      "learning_rate": 0.0002609437827162894,
+      "loss": 1.66,
+      "step": 4377
+    },
+    {
+      "epoch": 2.007338607957803,
+      "grad_norm": 0.3313737213611603,
+      "learning_rate": 0.0002607263079825106,
+      "loss": 1.8962,
+      "step": 4378
+    },
+    {
+      "epoch": 2.007797270955166,
+      "grad_norm": 0.3897867500782013,
+      "learning_rate": 0.0002605088919442936,
+      "loss": 1.5619,
+      "step": 4379
+    },
+    {
+      "epoch": 2.0082559339525283,
+      "grad_norm": 0.24746671319007874,
+      "learning_rate": 0.00026029153465497243,
+      "loss": 0.7236,
+      "step": 4380
+    },
+    {
+      "epoch": 2.008714596949891,
+      "grad_norm": 0.30122828483581543,
+      "learning_rate": 0.0002600742361678663,
+      "loss": 1.7644,
+      "step": 4381
+    },
+    {
+      "epoch": 2.0091732599472536,
+      "grad_norm": 0.32251477241516113,
+      "learning_rate": 0.0002598569965362799,
+      "loss": 1.2259,
+      "step": 4382
+    },
+    {
+      "epoch": 2.0096319229446165,
+      "grad_norm": 0.219418004155159,
+      "learning_rate": 0.0002596398158135037,
+      "loss": 0.5097,
+      "step": 4383
+    },
+    {
+      "epoch": 2.0100905859419793,
+      "grad_norm": 0.3621459901332855,
+      "learning_rate": 0.00025942269405281385,
+      "loss": 1.3002,
+      "step": 4384
+    },
+    {
+      "epoch": 2.0105492489393417,
+      "grad_norm": 0.06939633190631866,
+      "learning_rate": 0.00025920563130747167,
+      "loss": 0.3688,
+      "step": 4385
+    },
+    {
+      "epoch": 2.0110079119367046,
+      "grad_norm": 0.2349678874015808,
+      "learning_rate": 0.0002589886276307246,
+      "loss": 1.1327,
+      "step": 4386
+    },
+    {
+      "epoch": 2.011466574934067,
+      "grad_norm": 0.3204725384712219,
+      "learning_rate": 0.0002587716830758048,
+      "loss": 1.3671,
+      "step": 4387
+    },
+    {
+      "epoch": 2.01192523793143,
+      "grad_norm": 0.22198036313056946,
+      "learning_rate": 0.0002585547976959303,
+      "loss": 1.0752,
+      "step": 4388
+    },
+    {
+      "epoch": 2.0123839009287927,
+      "grad_norm": 0.30143803358078003,
+      "learning_rate": 0.00025833797154430494,
+      "loss": 1.792,
+      "step": 4389
+    },
+    {
+      "epoch": 2.012842563926155,
+      "grad_norm": 0.3181972801685333,
+      "learning_rate": 0.0002581212046741177,
+      "loss": 1.5635,
+      "step": 4390
+    },
+    {
+      "epoch": 2.013301226923518,
+      "grad_norm": 0.3738367557525635,
+      "learning_rate": 0.000257904497138543,
+      "loss": 1.2866,
+      "step": 4391
+    },
+    {
+      "epoch": 2.013759889920881,
+      "grad_norm": 0.256151020526886,
+      "learning_rate": 0.00025768784899074087,
+      "loss": 1.8867,
+      "step": 4392
+    },
+    {
+      "epoch": 2.0142185529182433,
+      "grad_norm": 0.28460898995399475,
+      "learning_rate": 0.00025747126028385643,
+      "loss": 0.8942,
+      "step": 4393
+    },
+    {
+      "epoch": 2.014677215915606,
+      "grad_norm": 0.27417483925819397,
+      "learning_rate": 0.0002572547310710205,
+      "loss": 1.547,
+      "step": 4394
+    },
+    {
+      "epoch": 2.0151358789129685,
+      "grad_norm": 0.3943521976470947,
+      "learning_rate": 0.00025703826140534937,
+      "loss": 1.222,
+      "step": 4395
+    },
+    {
+      "epoch": 2.0155945419103314,
+      "grad_norm": 0.3956080377101898,
+      "learning_rate": 0.00025682185133994457,
+      "loss": 1.679,
+      "step": 4396
+    },
+    {
+      "epoch": 2.0160532049076942,
+      "grad_norm": 0.41523680090904236,
+      "learning_rate": 0.0002566055009278932,
+      "loss": 0.6923,
+      "step": 4397
+    },
+    {
+      "epoch": 2.0165118679050567,
+      "grad_norm": 0.228208526968956,
+      "learning_rate": 0.00025638921022226704,
+      "loss": 1.4105,
+      "step": 4398
+    },
+    {
+      "epoch": 2.0169705309024195,
+      "grad_norm": 0.27708700299263,
+      "learning_rate": 0.00025617297927612437,
+      "loss": 1.4982,
+      "step": 4399
+    },
+    {
+      "epoch": 2.017429193899782,
+      "grad_norm": 0.34748536348342896,
+      "learning_rate": 0.0002559568081425079,
+      "loss": 1.3942,
+      "step": 4400
+    },
+    {
+      "epoch": 2.017887856897145,
+      "grad_norm": 0.25076958537101746,
+      "learning_rate": 0.00025574069687444613,
+      "loss": 1.3903,
+      "step": 4401
+    },
+    {
+      "epoch": 2.0183465198945076,
+      "grad_norm": 0.34007957577705383,
+      "learning_rate": 0.00025552464552495253,
+      "loss": 1.5933,
+      "step": 4402
+    },
+    {
+      "epoch": 2.01880518289187,
+      "grad_norm": 0.31931763887405396,
+      "learning_rate": 0.0002553086541470263,
+      "loss": 0.8987,
+      "step": 4403
+    },
+    {
+      "epoch": 2.019263845889233,
+      "grad_norm": 0.2598975598812103,
+      "learning_rate": 0.0002550927227936515,
+      "loss": 1.6906,
+      "step": 4404
+    },
+    {
+      "epoch": 2.0197225088865958,
+      "grad_norm": 0.3859318196773529,
+      "learning_rate": 0.0002548768515177975,
+      "loss": 2.1326,
+      "step": 4405
+    },
+    {
+      "epoch": 2.020181171883958,
+      "grad_norm": 0.35015788674354553,
+      "learning_rate": 0.0002546610403724193,
+      "loss": 0.8087,
+      "step": 4406
+    },
+    {
+      "epoch": 2.020639834881321,
+      "grad_norm": 0.25756731629371643,
+      "learning_rate": 0.0002544452894104569,
+      "loss": 1.2472,
+      "step": 4407
+    },
+    {
+      "epoch": 2.0210984978786835,
+      "grad_norm": 0.3442479372024536,
+      "learning_rate": 0.0002542295986848355,
+      "loss": 1.7928,
+      "step": 4408
+    },
+    {
+      "epoch": 2.0215571608760463,
+      "grad_norm": 0.41881293058395386,
+      "learning_rate": 0.00025401396824846576,
+      "loss": 1.7277,
+      "step": 4409
+    },
+    {
+      "epoch": 2.022015823873409,
+      "grad_norm": 0.28389236330986023,
+      "learning_rate": 0.0002537983981542432,
+      "loss": 1.1472,
+      "step": 4410
+    },
+    {
+      "epoch": 2.0224744868707716,
+      "grad_norm": 0.19957682490348816,
+      "learning_rate": 0.0002535828884550487,
+      "loss": 0.9713,
+      "step": 4411
+    },
+    {
+      "epoch": 2.0229331498681344,
+      "grad_norm": 0.20129139721393585,
+      "learning_rate": 0.0002533674392037485,
+      "loss": 1.2563,
+      "step": 4412
+    },
+    {
+      "epoch": 2.023391812865497,
+      "grad_norm": 0.39847332239151,
+      "learning_rate": 0.0002531520504531938,
+      "loss": 1.3053,
+      "step": 4413
+    },
+    {
+      "epoch": 2.0238504758628597,
+      "grad_norm": 0.3507583439350128,
+      "learning_rate": 0.00025293672225622113,
+      "loss": 1.7719,
+      "step": 4414
+    },
+    {
+      "epoch": 2.0243091388602226,
+      "grad_norm": 0.34539008140563965,
+      "learning_rate": 0.0002527214546656517,
+      "loss": 1.5484,
+      "step": 4415
+    },
+    {
+      "epoch": 2.024767801857585,
+      "grad_norm": 0.3002701997756958,
+      "learning_rate": 0.0002525062477342925,
+      "loss": 1.8398,
+      "step": 4416
+    },
+    {
+      "epoch": 2.025226464854948,
+      "grad_norm": 0.39402061700820923,
+      "learning_rate": 0.00025229110151493516,
+      "loss": 1.6705,
+      "step": 4417
+    },
+    {
+      "epoch": 2.0256851278523107,
+      "grad_norm": 0.28095006942749023,
+      "learning_rate": 0.0002520760160603567,
+      "loss": 1.7056,
+      "step": 4418
+    },
+    {
+      "epoch": 2.026143790849673,
+      "grad_norm": 0.33673858642578125,
+      "learning_rate": 0.0002518609914233192,
+      "loss": 1.4712,
+      "step": 4419
+    },
+    {
+      "epoch": 2.026602453847036,
+      "grad_norm": 0.3398224711418152,
+      "learning_rate": 0.00025164602765656964,
+      "loss": 1.0774,
+      "step": 4420
+    },
+    {
+      "epoch": 2.0270611168443984,
+      "grad_norm": 0.3440173268318176,
+      "learning_rate": 0.00025143112481284017,
+      "loss": 1.2208,
+      "step": 4421
+    },
+    {
+      "epoch": 2.0275197798417612,
+      "grad_norm": 0.30769407749176025,
+      "learning_rate": 0.0002512162829448481,
+      "loss": 1.2709,
+      "step": 4422
+    },
+    {
+      "epoch": 2.027978442839124,
+      "grad_norm": 0.2576143443584442,
+      "learning_rate": 0.00025100150210529565,
+      "loss": 1.1309,
+      "step": 4423
+    },
+    {
+      "epoch": 2.0284371058364865,
+      "grad_norm": 0.17252714931964874,
+      "learning_rate": 0.00025078678234687014,
+      "loss": 0.61,
+      "step": 4424
+    },
+    {
+      "epoch": 2.0288957688338494,
+      "grad_norm": 0.21611100435256958,
+      "learning_rate": 0.000250572123722244,
+      "loss": 1.1369,
+      "step": 4425
+    },
+    {
+      "epoch": 2.0293544318312122,
+      "grad_norm": 0.3421436846256256,
+      "learning_rate": 0.00025035752628407414,
+      "loss": 1.5566,
+      "step": 4426
+    },
+    {
+      "epoch": 2.0298130948285746,
+      "grad_norm": 0.29357588291168213,
+      "learning_rate": 0.00025014299008500315,
+      "loss": 1.1662,
+      "step": 4427
+    },
+    {
+      "epoch": 2.0302717578259375,
+      "grad_norm": 0.20734919607639313,
+      "learning_rate": 0.00024992851517765825,
+      "loss": 1.5304,
+      "step": 4428
+    },
+    {
+      "epoch": 2.0307304208233,
+      "grad_norm": 0.41112732887268066,
+      "learning_rate": 0.0002497141016146517,
+      "loss": 1.8782,
+      "step": 4429
+    },
+    {
+      "epoch": 2.0311890838206628,
+      "grad_norm": 0.33853259682655334,
+      "learning_rate": 0.0002494997494485806,
+      "loss": 1.2181,
+      "step": 4430
+    },
+    {
+      "epoch": 2.0316477468180256,
+      "grad_norm": 0.4708898067474365,
+      "learning_rate": 0.0002492854587320272,
+      "loss": 1.8342,
+      "step": 4431
+    },
+    {
+      "epoch": 2.032106409815388,
+      "grad_norm": 0.3891645073890686,
+      "learning_rate": 0.00024907122951755856,
+      "loss": 0.9312,
+      "step": 4432
+    },
+    {
+      "epoch": 2.032565072812751,
+      "grad_norm": 0.304002583026886,
+      "learning_rate": 0.00024885706185772663,
+      "loss": 1.071,
+      "step": 4433
+    },
+    {
+      "epoch": 2.0330237358101133,
+      "grad_norm": 0.26506680250167847,
+      "learning_rate": 0.00024864295580506816,
+      "loss": 1.4494,
+      "step": 4434
+    },
+    {
+      "epoch": 2.033482398807476,
+      "grad_norm": 0.33976414799690247,
+      "learning_rate": 0.0002484289114121051,
+      "loss": 1.2593,
+      "step": 4435
+    },
+    {
+      "epoch": 2.033941061804839,
+      "grad_norm": 0.2222852259874344,
+      "learning_rate": 0.0002482149287313439,
+      "loss": 0.7598,
+      "step": 4436
+    },
+    {
+      "epoch": 2.0343997248022014,
+      "grad_norm": 0.34920746088027954,
+      "learning_rate": 0.00024800100781527645,
+      "loss": 1.8142,
+      "step": 4437
+    },
+    {
+      "epoch": 2.0348583877995643,
+      "grad_norm": 0.3795028328895569,
+      "learning_rate": 0.00024778714871637853,
+      "loss": 1.2554,
+      "step": 4438
+    },
+    {
+      "epoch": 2.035317050796927,
+      "grad_norm": 0.0961453840136528,
+      "learning_rate": 0.0002475733514871116,
+      "loss": 0.9809,
+      "step": 4439
+    },
+    {
+      "epoch": 2.0357757137942896,
+      "grad_norm": 0.2711848020553589,
+      "learning_rate": 0.00024735961617992165,
+      "loss": 0.9706,
+      "step": 4440
+    },
+    {
+      "epoch": 2.0362343767916524,
+      "grad_norm": 0.3385700583457947,
+      "learning_rate": 0.0002471459428472393,
+      "loss": 1.4021,
+      "step": 4441
+    },
+    {
+      "epoch": 2.036693039789015,
+      "grad_norm": 0.2203783392906189,
+      "learning_rate": 0.00024693233154148063,
+      "loss": 0.8029,
+      "step": 4442
+    },
+    {
+      "epoch": 2.0371517027863777,
+      "grad_norm": 0.3024193346500397,
+      "learning_rate": 0.0002467187823150457,
+      "loss": 1.1432,
+      "step": 4443
+    },
+    {
+      "epoch": 2.0376103657837406,
+      "grad_norm": 0.3142333924770355,
+      "learning_rate": 0.0002465052952203196,
+      "loss": 1.7221,
+      "step": 4444
+    },
+    {
+      "epoch": 2.038069028781103,
+      "grad_norm": 0.3084831237792969,
+      "learning_rate": 0.0002462918703096724,
+      "loss": 1.2093,
+      "step": 4445
+    },
+    {
+      "epoch": 2.038527691778466,
+      "grad_norm": 0.38534000515937805,
+      "learning_rate": 0.0002460785076354588,
+      "loss": 1.4895,
+      "step": 4446
+    },
+    {
+      "epoch": 2.0389863547758287,
+      "grad_norm": 0.303943008184433,
+      "learning_rate": 0.0002458652072500181,
+      "loss": 0.9302,
+      "step": 4447
+    },
+    {
+      "epoch": 2.039445017773191,
+      "grad_norm": 0.26607340574264526,
+      "learning_rate": 0.0002456519692056747,
+      "loss": 1.4809,
+      "step": 4448
+    },
+    {
+      "epoch": 2.039903680770554,
+      "grad_norm": 0.3666565418243408,
+      "learning_rate": 0.0002454387935547369,
+      "loss": 1.2939,
+      "step": 4449
+    },
+    {
+      "epoch": 2.0403623437679164,
+      "grad_norm": 0.24224300682544708,
+      "learning_rate": 0.00024522568034949865,
+      "loss": 1.0746,
+      "step": 4450
+    },
+    {
+      "epoch": 2.040821006765279,
+      "grad_norm": 0.2701547145843506,
+      "learning_rate": 0.0002450126296422377,
+      "loss": 1.1681,
+      "step": 4451
+    },
+    {
+      "epoch": 2.041279669762642,
+      "grad_norm": 0.19419100880622864,
+      "learning_rate": 0.0002447996414852176,
+      "loss": 1.3045,
+      "step": 4452
+    },
+    {
+      "epoch": 2.0417383327600045,
+      "grad_norm": 0.30444204807281494,
+      "learning_rate": 0.00024458671593068564,
+      "loss": 1.4119,
+      "step": 4453
+    },
+    {
+      "epoch": 2.0421969957573674,
+      "grad_norm": 0.3836541473865509,
+      "learning_rate": 0.00024437385303087373,
+      "loss": 1.7348,
+      "step": 4454
+    },
+    {
+      "epoch": 2.0426556587547298,
+      "grad_norm": 0.312198281288147,
+      "learning_rate": 0.0002441610528379988,
+      "loss": 1.2589,
+      "step": 4455
+    },
+    {
+      "epoch": 2.0431143217520926,
+      "grad_norm": 0.29707929491996765,
+      "learning_rate": 0.00024394831540426232,
+      "loss": 0.8035,
+      "step": 4456
+    },
+    {
+      "epoch": 2.0435729847494555,
+      "grad_norm": 0.2303507924079895,
+      "learning_rate": 0.0002437356407818503,
+      "loss": 1.088,
+      "step": 4457
+    },
+    {
+      "epoch": 2.044031647746818,
+      "grad_norm": 0.3469330072402954,
+      "learning_rate": 0.00024352302902293333,
+      "loss": 1.6418,
+      "step": 4458
+    },
+    {
+      "epoch": 2.0444903107441807,
+      "grad_norm": 0.21516135334968567,
+      "learning_rate": 0.00024331048017966683,
+      "loss": 0.9941,
+      "step": 4459
+    },
+    {
+      "epoch": 2.0449489737415436,
+      "grad_norm": 0.38784271478652954,
+      "learning_rate": 0.00024309799430419,
+      "loss": 2.2803,
+      "step": 4460
+    },
+    {
+      "epoch": 2.045407636738906,
+      "grad_norm": 0.4177115261554718,
+      "learning_rate": 0.0002428855714486277,
+      "loss": 1.9731,
+      "step": 4461
+    },
+    {
+      "epoch": 2.045866299736269,
+      "grad_norm": 0.3393002152442932,
+      "learning_rate": 0.00024267321166508867,
+      "loss": 1.2382,
+      "step": 4462
+    },
+    {
+      "epoch": 2.0463249627336313,
+      "grad_norm": 0.28032952547073364,
+      "learning_rate": 0.00024246091500566619,
+      "loss": 1.1954,
+      "step": 4463
+    },
+    {
+      "epoch": 2.046783625730994,
+      "grad_norm": 0.28215348720550537,
+      "learning_rate": 0.00024224868152243823,
+      "loss": 0.8485,
+      "step": 4464
+    },
+    {
+      "epoch": 2.047242288728357,
+      "grad_norm": 0.26953309774398804,
+      "learning_rate": 0.0002420365112674674,
+      "loss": 1.4311,
+      "step": 4465
+    },
+    {
+      "epoch": 2.0477009517257194,
+      "grad_norm": 0.20407378673553467,
+      "learning_rate": 0.0002418244042928001,
+      "loss": 0.9565,
+      "step": 4466
+    },
+    {
+      "epoch": 2.0481596147230823,
+      "grad_norm": 0.27266445755958557,
+      "learning_rate": 0.00024161236065046806,
+      "loss": 1.2134,
+      "step": 4467
+    },
+    {
+      "epoch": 2.0486182777204447,
+      "grad_norm": 0.2968637943267822,
+      "learning_rate": 0.00024140038039248697,
+      "loss": 1.1802,
+      "step": 4468
+    },
+    {
+      "epoch": 2.0490769407178075,
+      "grad_norm": 0.34877461194992065,
+      "learning_rate": 0.00024118846357085717,
+      "loss": 1.5968,
+      "step": 4469
+    },
+    {
+      "epoch": 2.0495356037151704,
+      "grad_norm": 0.3103572726249695,
+      "learning_rate": 0.0002409766102375634,
+      "loss": 1.218,
+      "step": 4470
+    },
+    {
+      "epoch": 2.049994266712533,
+      "grad_norm": 0.33419668674468994,
+      "learning_rate": 0.00024076482044457477,
+      "loss": 1.781,
+      "step": 4471
+    },
+    {
+      "epoch": 2.0504529297098957,
+      "grad_norm": 0.3635629415512085,
+      "learning_rate": 0.00024055309424384486,
+      "loss": 1.4742,
+      "step": 4472
+    },
+    {
+      "epoch": 2.0509115927072585,
+      "grad_norm": 0.336867094039917,
+      "learning_rate": 0.00024034143168731172,
+      "loss": 1.2226,
+      "step": 4473
+    },
+    {
+      "epoch": 2.051370255704621,
+      "grad_norm": 0.2931895852088928,
+      "learning_rate": 0.00024012983282689754,
+      "loss": 1.1535,
+      "step": 4474
+    },
+    {
+      "epoch": 2.051828918701984,
+      "grad_norm": 0.36698460578918457,
+      "learning_rate": 0.00023991829771450912,
+      "loss": 1.2108,
+      "step": 4475
+    },
+    {
+      "epoch": 2.052287581699346,
+      "grad_norm": 0.306598037481308,
+      "learning_rate": 0.00023970682640203782,
+      "loss": 1.5089,
+      "step": 4476
+    },
+    {
+      "epoch": 2.052746244696709,
+      "grad_norm": 0.37377047538757324,
+      "learning_rate": 0.00023949541894135857,
+      "loss": 1.7544,
+      "step": 4477
+    },
+    {
+      "epoch": 2.053204907694072,
+      "grad_norm": 0.3563328981399536,
+      "learning_rate": 0.00023928407538433138,
+      "loss": 1.3446,
+      "step": 4478
+    },
+    {
+      "epoch": 2.0536635706914343,
+      "grad_norm": 0.29219722747802734,
+      "learning_rate": 0.0002390727957828004,
+      "loss": 1.7413,
+      "step": 4479
+    },
+    {
+      "epoch": 2.054122233688797,
+      "grad_norm": 0.49362048506736755,
+      "learning_rate": 0.000238861580188594,
+      "loss": 1.6321,
+      "step": 4480
+    },
+    {
+      "epoch": 2.0545808966861596,
+      "grad_norm": 0.4204590916633606,
+      "learning_rate": 0.00023865042865352487,
+      "loss": 2.0813,
+      "step": 4481
+    },
+    {
+      "epoch": 2.0550395596835225,
+      "grad_norm": 0.33447083830833435,
+      "learning_rate": 0.00023843934122938997,
+      "loss": 1.4967,
+      "step": 4482
+    },
+    {
+      "epoch": 2.0554982226808853,
+      "grad_norm": 0.3546378016471863,
+      "learning_rate": 0.0002382283179679707,
+      "loss": 0.8393,
+      "step": 4483
+    },
+    {
+      "epoch": 2.0559568856782477,
+      "grad_norm": 0.2834445536136627,
+      "learning_rate": 0.00023801735892103244,
+      "loss": 1.6294,
+      "step": 4484
+    },
+    {
+      "epoch": 2.0564155486756106,
+      "grad_norm": 0.46338558197021484,
+      "learning_rate": 0.0002378064641403251,
+      "loss": 1.1402,
+      "step": 4485
+    },
+    {
+      "epoch": 2.0568742116729735,
+      "grad_norm": 0.14141368865966797,
+      "learning_rate": 0.00023759563367758252,
+      "loss": 1.6619,
+      "step": 4486
+    },
+    {
+      "epoch": 2.057332874670336,
+      "grad_norm": 0.35884973406791687,
+      "learning_rate": 0.00023738486758452326,
+      "loss": 1.1481,
+      "step": 4487
+    },
+    {
+      "epoch": 2.0577915376676987,
+      "grad_norm": 0.35921135544776917,
+      "learning_rate": 0.0002371741659128494,
+      "loss": 1.7199,
+      "step": 4488
+    },
+    {
+      "epoch": 2.058250200665061,
+      "grad_norm": 0.35381534695625305,
+      "learning_rate": 0.00023696352871424765,
+      "loss": 0.8198,
+      "step": 4489
+    },
+    {
+      "epoch": 2.058708863662424,
+      "grad_norm": 0.29065418243408203,
+      "learning_rate": 0.00023675295604038893,
+      "loss": 1.0642,
+      "step": 4490
+    },
+    {
+      "epoch": 2.059167526659787,
+      "grad_norm": 0.30102601647377014,
+      "learning_rate": 0.00023654244794292823,
+      "loss": 1.451,
+      "step": 4491
+    },
+    {
+      "epoch": 2.0596261896571493,
+      "grad_norm": 0.37403160333633423,
+      "learning_rate": 0.00023633200447350462,
+      "loss": 1.8662,
+      "step": 4492
+    },
+    {
+      "epoch": 2.060084852654512,
+      "grad_norm": 0.45195087790489197,
+      "learning_rate": 0.00023612162568374147,
+      "loss": 0.8842,
+      "step": 4493
+    },
+    {
+      "epoch": 2.060543515651875,
+      "grad_norm": 0.06819970905780792,
+      "learning_rate": 0.0002359113116252462,
+      "loss": 0.9856,
+      "step": 4494
+    },
+    {
+      "epoch": 2.0610021786492374,
+      "grad_norm": 0.31288787722587585,
+      "learning_rate": 0.00023570106234961036,
+      "loss": 1.5755,
+      "step": 4495
+    },
+    {
+      "epoch": 2.0614608416466003,
+      "grad_norm": 0.38856643438339233,
+      "learning_rate": 0.00023549087790840966,
+      "loss": 1.3749,
+      "step": 4496
+    },
+    {
+      "epoch": 2.0619195046439627,
+      "grad_norm": 0.2518812119960785,
+      "learning_rate": 0.00023528075835320378,
+      "loss": 1.7864,
+      "step": 4497
+    },
+    {
+      "epoch": 2.0623781676413255,
+      "grad_norm": 0.4028247892856598,
+      "learning_rate": 0.0002350707037355368,
+      "loss": 1.551,
+      "step": 4498
+    },
+    {
+      "epoch": 2.0628368306386884,
+      "grad_norm": 0.34255698323249817,
+      "learning_rate": 0.00023486071410693627,
+      "loss": 1.0178,
+      "step": 4499
+    },
+    {
+      "epoch": 2.063295493636051,
+      "grad_norm": 0.08957389742136002,
+      "learning_rate": 0.0002346507895189143,
+      "loss": 1.2612,
+      "step": 4500
+    },
+    {
+      "epoch": 2.0637541566334137,
+      "grad_norm": 0.3247790038585663,
+      "learning_rate": 0.0002344409300229669,
+      "loss": 1.258,
+      "step": 4501
+    },
+    {
+      "epoch": 2.064212819630776,
+      "grad_norm": 0.39857447147369385,
+      "learning_rate": 0.0002342311356705742,
+      "loss": 1.3753,
+      "step": 4502
+    },
+    {
+      "epoch": 2.064671482628139,
+      "grad_norm": 0.13461144268512726,
+      "learning_rate": 0.00023402140651320003,
+      "loss": 0.8576,
+      "step": 4503
+    },
+    {
+      "epoch": 2.065130145625502,
+      "grad_norm": 0.33944034576416016,
+      "learning_rate": 0.000233811742602293,
+      "loss": 1.5578,
+      "step": 4504
+    },
+    {
+      "epoch": 2.065588808622864,
+      "grad_norm": 0.3041727542877197,
+      "learning_rate": 0.0002336021439892846,
+      "loss": 1.278,
+      "step": 4505
+    },
+    {
+      "epoch": 2.066047471620227,
+      "grad_norm": 0.28007856011390686,
+      "learning_rate": 0.00023339261072559116,
+      "loss": 1.2006,
+      "step": 4506
+    },
+    {
+      "epoch": 2.06650613461759,
+      "grad_norm": 0.2758924663066864,
+      "learning_rate": 0.00023318314286261262,
+      "loss": 1.5737,
+      "step": 4507
+    },
+    {
+      "epoch": 2.0669647976149523,
+      "grad_norm": 0.318847119808197,
+      "learning_rate": 0.00023297374045173298,
+      "loss": 1.3716,
+      "step": 4508
+    },
+    {
+      "epoch": 2.067423460612315,
+      "grad_norm": 0.3623434007167816,
+      "learning_rate": 0.00023276440354432038,
+      "loss": 1.8606,
+      "step": 4509
+    },
+    {
+      "epoch": 2.0678821236096776,
+      "grad_norm": 0.43447864055633545,
+      "learning_rate": 0.00023255513219172625,
+      "loss": 1.9836,
+      "step": 4510
+    },
+    {
+      "epoch": 2.0683407866070405,
+      "grad_norm": 0.38707858324050903,
+      "learning_rate": 0.00023234592644528657,
+      "loss": 1.4044,
+      "step": 4511
+    },
+    {
+      "epoch": 2.0687994496044033,
+      "grad_norm": 0.43568772077560425,
+      "learning_rate": 0.00023213678635632102,
+      "loss": 2.0554,
+      "step": 4512
+    },
+    {
+      "epoch": 2.0692581126017657,
+      "grad_norm": 0.3717525601387024,
+      "learning_rate": 0.000231927711976133,
+      "loss": 1.493,
+      "step": 4513
+    },
+    {
+      "epoch": 2.0697167755991286,
+      "grad_norm": 0.21989569067955017,
+      "learning_rate": 0.0002317187033560103,
+      "loss": 0.6022,
+      "step": 4514
+    },
+    {
+      "epoch": 2.0701754385964914,
+      "grad_norm": 0.10017600655555725,
+      "learning_rate": 0.0002315097605472243,
+      "loss": 1.3052,
+      "step": 4515
+    },
+    {
+      "epoch": 2.070634101593854,
+      "grad_norm": 0.3896404802799225,
+      "learning_rate": 0.00023130088360102968,
+      "loss": 1.0817,
+      "step": 4516
+    },
+    {
+      "epoch": 2.0710927645912167,
+      "grad_norm": 0.07273133844137192,
+      "learning_rate": 0.00023109207256866583,
+      "loss": 0.5404,
+      "step": 4517
+    },
+    {
+      "epoch": 2.071551427588579,
+      "grad_norm": 0.35691580176353455,
+      "learning_rate": 0.00023088332750135544,
+      "loss": 1.6287,
+      "step": 4518
+    },
+    {
+      "epoch": 2.072010090585942,
+      "grad_norm": 0.42362430691719055,
+      "learning_rate": 0.00023067464845030527,
+      "loss": 1.7539,
+      "step": 4519
+    },
+    {
+      "epoch": 2.072468753583305,
+      "grad_norm": 0.5343214273452759,
+      "learning_rate": 0.00023046603546670596,
+      "loss": 2.1514,
+      "step": 4520
+    },
+    {
+      "epoch": 2.0729274165806673,
+      "grad_norm": 0.4141870439052582,
+      "learning_rate": 0.0002302574886017314,
+      "loss": 1.6187,
+      "step": 4521
+    },
+    {
+      "epoch": 2.07338607957803,
+      "grad_norm": 0.2982878088951111,
+      "learning_rate": 0.00023004900790653986,
+      "loss": 1.6875,
+      "step": 4522
+    },
+    {
+      "epoch": 2.0738447425753925,
+      "grad_norm": 0.28776493668556213,
+      "learning_rate": 0.00022984059343227292,
+      "loss": 0.642,
+      "step": 4523
+    },
+    {
+      "epoch": 2.0743034055727554,
+      "grad_norm": 0.2196163535118103,
+      "learning_rate": 0.00022963224523005654,
+      "loss": 1.3851,
+      "step": 4524
+    },
+    {
+      "epoch": 2.0747620685701182,
+      "grad_norm": 0.3807169795036316,
+      "learning_rate": 0.00022942396335099986,
+      "loss": 1.2697,
+      "step": 4525
+    },
+    {
+      "epoch": 2.0752207315674807,
+      "grad_norm": 0.3772948086261749,
+      "learning_rate": 0.00022921574784619608,
+      "loss": 1.7243,
+      "step": 4526
+    },
+    {
+      "epoch": 2.0756793945648435,
+      "grad_norm": 0.532681405544281,
+      "learning_rate": 0.00022900759876672168,
+      "loss": 1.4512,
+      "step": 4527
+    },
+    {
+      "epoch": 2.0761380575622064,
+      "grad_norm": 0.36544162034988403,
+      "learning_rate": 0.00022879951616363727,
+      "loss": 1.5479,
+      "step": 4528
+    },
+    {
+      "epoch": 2.076596720559569,
+      "grad_norm": 0.2846382260322571,
+      "learning_rate": 0.0002285915000879869,
+      "loss": 1.2101,
+      "step": 4529
+    },
+    {
+      "epoch": 2.0770553835569316,
+      "grad_norm": 0.3465891480445862,
+      "learning_rate": 0.00022838355059079862,
+      "loss": 1.9592,
+      "step": 4530
+    },
+    {
+      "epoch": 2.077514046554294,
+      "grad_norm": 0.3925153315067291,
+      "learning_rate": 0.00022817566772308378,
+      "loss": 2.1309,
+      "step": 4531
+    },
+    {
+      "epoch": 2.077972709551657,
+      "grad_norm": 0.3107575476169586,
+      "learning_rate": 0.0002279678515358376,
+      "loss": 0.5652,
+      "step": 4532
+    },
+    {
+      "epoch": 2.0784313725490198,
+      "grad_norm": 0.19689220190048218,
+      "learning_rate": 0.00022776010208003895,
+      "loss": 1.6448,
+      "step": 4533
+    },
+    {
+      "epoch": 2.078890035546382,
+      "grad_norm": 0.3134257197380066,
+      "learning_rate": 0.00022755241940665018,
+      "loss": 1.3657,
+      "step": 4534
+    },
+    {
+      "epoch": 2.079348698543745,
+      "grad_norm": 0.4296601116657257,
+      "learning_rate": 0.00022734480356661736,
+      "loss": 1.34,
+      "step": 4535
+    },
+    {
+      "epoch": 2.0798073615411075,
+      "grad_norm": 0.3112638294696808,
+      "learning_rate": 0.00022713725461087015,
+      "loss": 1.282,
+      "step": 4536
+    },
+    {
+      "epoch": 2.0802660245384703,
+      "grad_norm": 0.13700056076049805,
+      "learning_rate": 0.00022692977259032205,
+      "loss": 1.3737,
+      "step": 4537
+    },
+    {
+      "epoch": 2.080724687535833,
+      "grad_norm": 0.4667096436023712,
+      "learning_rate": 0.00022672235755586952,
+      "loss": 1.8356,
+      "step": 4538
+    },
+    {
+      "epoch": 2.0811833505331956,
+      "grad_norm": 0.31745702028274536,
+      "learning_rate": 0.00022651500955839305,
+      "loss": 0.7115,
+      "step": 4539
+    },
+    {
+      "epoch": 2.0816420135305584,
+      "grad_norm": 0.3241938352584839,
+      "learning_rate": 0.0002263077286487567,
+      "loss": 1.7333,
+      "step": 4540
+    },
+    {
+      "epoch": 2.0821006765279213,
+      "grad_norm": 0.2541601359844208,
+      "learning_rate": 0.00022610051487780792,
+      "loss": 1.3767,
+      "step": 4541
+    },
+    {
+      "epoch": 2.0825593395252837,
+      "grad_norm": 0.22772157192230225,
+      "learning_rate": 0.00022589336829637776,
+      "loss": 0.4693,
+      "step": 4542
+    },
+    {
+      "epoch": 2.0830180025226466,
+      "grad_norm": 0.283128947019577,
+      "learning_rate": 0.00022568628895528077,
+      "loss": 1.7429,
+      "step": 4543
+    },
+    {
+      "epoch": 2.083476665520009,
+      "grad_norm": 0.4075316786766052,
+      "learning_rate": 0.000225479276905315,
+      "loss": 1.6674,
+      "step": 4544
+    },
+    {
+      "epoch": 2.083935328517372,
+      "grad_norm": 0.36636197566986084,
+      "learning_rate": 0.00022527233219726202,
+      "loss": 1.7224,
+      "step": 4545
+    },
+    {
+      "epoch": 2.0843939915147347,
+      "grad_norm": 0.29077982902526855,
+      "learning_rate": 0.00022506545488188678,
+      "loss": 1.3293,
+      "step": 4546
+    },
+    {
+      "epoch": 2.084852654512097,
+      "grad_norm": 0.2742144465446472,
+      "learning_rate": 0.0002248586450099379,
+      "loss": 0.9725,
+      "step": 4547
+    },
+    {
+      "epoch": 2.08531131750946,
+      "grad_norm": 0.2022017389535904,
+      "learning_rate": 0.00022465190263214747,
+      "loss": 1.2256,
+      "step": 4548
+    },
+    {
+      "epoch": 2.0857699805068224,
+      "grad_norm": 0.2726742625236511,
+      "learning_rate": 0.00022444522779923044,
+      "loss": 1.079,
+      "step": 4549
+    },
+    {
+      "epoch": 2.0862286435041852,
+      "grad_norm": 0.23730774223804474,
+      "learning_rate": 0.00022423862056188593,
+      "loss": 1.5887,
+      "step": 4550
+    },
+    {
+      "epoch": 2.086687306501548,
+      "grad_norm": 1.3285748958587646,
+      "learning_rate": 0.00022403208097079613,
+      "loss": 1.9584,
+      "step": 4551
+    },
+    {
+      "epoch": 2.0871459694989105,
+      "grad_norm": 0.32901179790496826,
+      "learning_rate": 0.00022382560907662668,
+      "loss": 1.2246,
+      "step": 4552
+    },
+    {
+      "epoch": 2.0876046324962734,
+      "grad_norm": 0.260644793510437,
+      "learning_rate": 0.00022361920493002669,
+      "loss": 1.4697,
+      "step": 4553
+    },
+    {
+      "epoch": 2.0880632954936362,
+      "grad_norm": 0.35175642371177673,
+      "learning_rate": 0.0002234128685816285,
+      "loss": 1.5698,
+      "step": 4554
+    },
+    {
+      "epoch": 2.0885219584909986,
+      "grad_norm": 0.44502708315849304,
+      "learning_rate": 0.00022320660008204795,
+      "loss": 2.0159,
+      "step": 4555
+    },
+    {
+      "epoch": 2.0889806214883615,
+      "grad_norm": 0.37410539388656616,
+      "learning_rate": 0.00022300039948188418,
+      "loss": 1.7451,
+      "step": 4556
+    },
+    {
+      "epoch": 2.089439284485724,
+      "grad_norm": 0.4350254237651825,
+      "learning_rate": 0.0002227942668317197,
+      "loss": 1.2582,
+      "step": 4557
+    },
+    {
+      "epoch": 2.0898979474830868,
+      "grad_norm": 0.23693208396434784,
+      "learning_rate": 0.00022258820218212035,
+      "loss": 1.162,
+      "step": 4558
+    },
+    {
+      "epoch": 2.0903566104804496,
+      "grad_norm": 0.33850157260894775,
+      "learning_rate": 0.0002223822055836352,
+      "loss": 1.2121,
+      "step": 4559
+    },
+    {
+      "epoch": 2.090815273477812,
+      "grad_norm": 0.2646215260028839,
+      "learning_rate": 0.00022217627708679693,
+      "loss": 0.4118,
+      "step": 4560
+    },
+    {
+      "epoch": 2.091273936475175,
+      "grad_norm": 0.26649442315101624,
+      "learning_rate": 0.00022197041674212092,
+      "loss": 2.1147,
+      "step": 4561
+    },
+    {
+      "epoch": 2.0917325994725378,
+      "grad_norm": 0.27695992588996887,
+      "learning_rate": 0.0002217646246001064,
+      "loss": 1.0989,
+      "step": 4562
+    },
+    {
+      "epoch": 2.0921912624699,
+      "grad_norm": 0.34253594279289246,
+      "learning_rate": 0.00022155890071123564,
+      "loss": 1.6344,
+      "step": 4563
+    },
+    {
+      "epoch": 2.092649925467263,
+      "grad_norm": 0.32384851574897766,
+      "learning_rate": 0.0002213532451259742,
+      "loss": 1.7167,
+      "step": 4564
+    },
+    {
+      "epoch": 2.0931085884646254,
+      "grad_norm": 0.35032370686531067,
+      "learning_rate": 0.00022114765789477088,
+      "loss": 1.7271,
+      "step": 4565
+    },
+    {
+      "epoch": 2.0935672514619883,
+      "grad_norm": 0.2640658915042877,
+      "learning_rate": 0.0002209421390680577,
+      "loss": 1.6772,
+      "step": 4566
+    },
+    {
+      "epoch": 2.094025914459351,
+      "grad_norm": 0.33520472049713135,
+      "learning_rate": 0.00022073668869624995,
+      "loss": 1.4508,
+      "step": 4567
+    },
+    {
+      "epoch": 2.0944845774567136,
+      "grad_norm": 0.31736597418785095,
+      "learning_rate": 0.00022053130682974604,
+      "loss": 1.4152,
+      "step": 4568
+    },
+    {
+      "epoch": 2.0949432404540764,
+      "grad_norm": 0.31707099080085754,
+      "learning_rate": 0.00022032599351892764,
+      "loss": 1.6255,
+      "step": 4569
+    },
+    {
+      "epoch": 2.095401903451439,
+      "grad_norm": 0.4904292821884155,
+      "learning_rate": 0.00022012074881415955,
+      "loss": 1.9502,
+      "step": 4570
+    },
+    {
+      "epoch": 2.0958605664488017,
+      "grad_norm": 0.3095618188381195,
+      "learning_rate": 0.00021991557276579,
+      "loss": 1.5995,
+      "step": 4571
+    },
+    {
+      "epoch": 2.0963192294461646,
+      "grad_norm": 0.41936981678009033,
+      "learning_rate": 0.0002197104654241498,
+      "loss": 1.7164,
+      "step": 4572
+    },
+    {
+      "epoch": 2.096777892443527,
+      "grad_norm": 0.34843704104423523,
+      "learning_rate": 0.00021950542683955344,
+      "loss": 1.2392,
+      "step": 4573
+    },
+    {
+      "epoch": 2.09723655544089,
+      "grad_norm": 0.19430193305015564,
+      "learning_rate": 0.00021930045706229835,
+      "loss": 0.9546,
+      "step": 4574
+    },
+    {
+      "epoch": 2.0976952184382527,
+      "grad_norm": 0.36258745193481445,
+      "learning_rate": 0.00021909555614266484,
+      "loss": 1.447,
+      "step": 4575
+    },
+    {
+      "epoch": 2.098153881435615,
+      "grad_norm": 0.21773214638233185,
+      "learning_rate": 0.00021889072413091727,
+      "loss": 1.3679,
+      "step": 4576
+    },
+    {
+      "epoch": 2.098612544432978,
+      "grad_norm": 0.2897671163082123,
+      "learning_rate": 0.00021868596107730176,
+      "loss": 0.8525,
+      "step": 4577
+    },
+    {
+      "epoch": 2.0990712074303404,
+      "grad_norm": 0.22055676579475403,
+      "learning_rate": 0.0002184812670320484,
+      "loss": 1.4193,
+      "step": 4578
+    },
+    {
+      "epoch": 2.0995298704277032,
+      "grad_norm": 0.24514396488666534,
+      "learning_rate": 0.00021827664204537007,
+      "loss": 0.5204,
+      "step": 4579
+    },
+    {
+      "epoch": 2.099988533425066,
+      "grad_norm": 0.11148292571306229,
+      "learning_rate": 0.00021807208616746277,
+      "loss": 0.7519,
+      "step": 4580
+    },
+    {
+      "epoch": 2.1004471964224285,
+      "grad_norm": 0.35977640748023987,
+      "learning_rate": 0.00021786759944850554,
+      "loss": 1.3285,
+      "step": 4581
+    },
+    {
+      "epoch": 2.1009058594197914,
+      "grad_norm": 0.33894822001457214,
+      "learning_rate": 0.00021766318193866064,
+      "loss": 1.6344,
+      "step": 4582
+    },
+    {
+      "epoch": 2.101364522417154,
+      "grad_norm": 0.1890052706003189,
+      "learning_rate": 0.00021745883368807278,
+      "loss": 0.8127,
+      "step": 4583
+    },
+    {
+      "epoch": 2.1018231854145166,
+      "grad_norm": 0.3900175094604492,
+      "learning_rate": 0.00021725455474687027,
+      "loss": 1.8716,
+      "step": 4584
+    },
+    {
+      "epoch": 2.1022818484118795,
+      "grad_norm": 0.2661358714103699,
+      "learning_rate": 0.00021705034516516396,
+      "loss": 0.8125,
+      "step": 4585
+    },
+    {
+      "epoch": 2.102740511409242,
+      "grad_norm": 0.34905606508255005,
+      "learning_rate": 0.00021684620499304836,
+      "loss": 1.2191,
+      "step": 4586
+    },
+    {
+      "epoch": 2.1031991744066048,
+      "grad_norm": 0.2521708905696869,
+      "learning_rate": 0.0002166421342806003,
+      "loss": 1.4932,
+      "step": 4587
+    },
+    {
+      "epoch": 2.1036578374039676,
+      "grad_norm": 0.31396129727363586,
+      "learning_rate": 0.00021643813307788002,
+      "loss": 0.9659,
+      "step": 4588
+    },
+    {
+      "epoch": 2.10411650040133,
+      "grad_norm": 0.22702209651470184,
+      "learning_rate": 0.00021623420143493006,
+      "loss": 0.6938,
+      "step": 4589
+    },
+    {
+      "epoch": 2.104575163398693,
+      "grad_norm": 0.10886117815971375,
+      "learning_rate": 0.00021603033940177657,
+      "loss": 0.9339,
+      "step": 4590
+    },
+    {
+      "epoch": 2.1050338263960553,
+      "grad_norm": 0.262203186750412,
+      "learning_rate": 0.00021582654702842835,
+      "loss": 1.6595,
+      "step": 4591
+    },
+    {
+      "epoch": 2.105492489393418,
+      "grad_norm": 0.5163483023643494,
+      "learning_rate": 0.00021562282436487717,
+      "loss": 1.8124,
+      "step": 4592
+    },
+    {
+      "epoch": 2.105951152390781,
+      "grad_norm": 0.31572282314300537,
+      "learning_rate": 0.0002154191714610978,
+      "loss": 1.6561,
+      "step": 4593
+    },
+    {
+      "epoch": 2.1064098153881434,
+      "grad_norm": 0.5046913623809814,
+      "learning_rate": 0.0002152155883670474,
+      "loss": 1.4269,
+      "step": 4594
+    },
+    {
+      "epoch": 2.1068684783855063,
+      "grad_norm": 0.27989131212234497,
+      "learning_rate": 0.0002150120751326664,
+      "loss": 0.3858,
+      "step": 4595
+    },
+    {
+      "epoch": 2.107327141382869,
+      "grad_norm": 0.8256828784942627,
+      "learning_rate": 0.00021480863180787846,
+      "loss": 1.5546,
+      "step": 4596
+    },
+    {
+      "epoch": 2.1077858043802316,
+      "grad_norm": 0.2901069223880768,
+      "learning_rate": 0.00021460525844258944,
+      "loss": 1.1568,
+      "step": 4597
+    },
+    {
+      "epoch": 2.1082444673775944,
+      "grad_norm": 0.49482643604278564,
+      "learning_rate": 0.00021440195508668836,
+      "loss": 1.4728,
+      "step": 4598
+    },
+    {
+      "epoch": 2.108703130374957,
+      "grad_norm": 0.34062066674232483,
+      "learning_rate": 0.00021419872179004714,
+      "loss": 1.5667,
+      "step": 4599
+    },
+    {
+      "epoch": 2.1091617933723197,
+      "grad_norm": 0.3349739909172058,
+      "learning_rate": 0.00021399555860251995,
+      "loss": 1.2008,
+      "step": 4600
+    },
+    {
+      "epoch": 2.1096204563696825,
+      "grad_norm": 0.26807984709739685,
+      "learning_rate": 0.0002137924655739445,
+      "loss": 0.8011,
+      "step": 4601
+    },
+    {
+      "epoch": 2.110079119367045,
+      "grad_norm": 0.17875202000141144,
+      "learning_rate": 0.0002135894427541409,
+      "loss": 1.6011,
+      "step": 4602
+    },
+    {
+      "epoch": 2.110537782364408,
+      "grad_norm": 0.3981230854988098,
+      "learning_rate": 0.00021338649019291212,
+      "loss": 1.1253,
+      "step": 4603
+    },
+    {
+      "epoch": 2.1109964453617707,
+      "grad_norm": 0.24843645095825195,
+      "learning_rate": 0.00021318360794004388,
+      "loss": 1.1508,
+      "step": 4604
+    },
+    {
+      "epoch": 2.111455108359133,
+      "grad_norm": 0.25917336344718933,
+      "learning_rate": 0.00021298079604530464,
+      "loss": 1.8315,
+      "step": 4605
+    },
+    {
+      "epoch": 2.111913771356496,
+      "grad_norm": 0.36518481373786926,
+      "learning_rate": 0.00021277805455844568,
+      "loss": 1.3478,
+      "step": 4606
+    },
+    {
+      "epoch": 2.1123724343538584,
+      "grad_norm": 0.4103847146034241,
+      "learning_rate": 0.00021257538352920091,
+      "loss": 1.0914,
+      "step": 4607
+    },
+    {
+      "epoch": 2.112831097351221,
+      "grad_norm": 0.31975114345550537,
+      "learning_rate": 0.00021237278300728697,
+      "loss": 1.1174,
+      "step": 4608
+    },
+    {
+      "epoch": 2.113289760348584,
+      "grad_norm": 0.23576036095619202,
+      "learning_rate": 0.00021217025304240327,
+      "loss": 1.0789,
+      "step": 4609
+    },
+    {
+      "epoch": 2.1137484233459465,
+      "grad_norm": 0.3097432255744934,
+      "learning_rate": 0.00021196779368423208,
+      "loss": 0.7991,
+      "step": 4610
+    },
+    {
+      "epoch": 2.1142070863433093,
+      "grad_norm": 0.1049700602889061,
+      "learning_rate": 0.00021176540498243768,
+      "loss": 1.429,
+      "step": 4611
+    },
+    {
+      "epoch": 2.1146657493406718,
+      "grad_norm": 0.3536173403263092,
+      "learning_rate": 0.00021156308698666777,
+      "loss": 1.2648,
+      "step": 4612
+    },
+    {
+      "epoch": 2.1151244123380346,
+      "grad_norm": 0.26619645953178406,
+      "learning_rate": 0.00021136083974655236,
+      "loss": 1.4557,
+      "step": 4613
+    },
+    {
+      "epoch": 2.1155830753353975,
+      "grad_norm": 0.308630108833313,
+      "learning_rate": 0.0002111586633117041,
+      "loss": 0.9568,
+      "step": 4614
+    },
+    {
+      "epoch": 2.11604173833276,
+      "grad_norm": 0.26167064905166626,
+      "learning_rate": 0.0002109565577317184,
+      "loss": 1.3277,
+      "step": 4615
+    },
+    {
+      "epoch": 2.1165004013301227,
+      "grad_norm": 0.26667726039886475,
+      "learning_rate": 0.0002107545230561732,
+      "loss": 1.1866,
+      "step": 4616
+    },
+    {
+      "epoch": 2.116959064327485,
+      "grad_norm": 0.35242959856987,
+      "learning_rate": 0.00021055255933462912,
+      "loss": 1.6272,
+      "step": 4617
+    },
+    {
+      "epoch": 2.117417727324848,
+      "grad_norm": 0.3653687536716461,
+      "learning_rate": 0.0002103506666166292,
+      "loss": 1.4598,
+      "step": 4618
+    },
+    {
+      "epoch": 2.117876390322211,
+      "grad_norm": 0.3237496614456177,
+      "learning_rate": 0.00021014884495169927,
+      "loss": 1.6118,
+      "step": 4619
+    },
+    {
+      "epoch": 2.1183350533195733,
+      "grad_norm": 0.4205828011035919,
+      "learning_rate": 0.00020994709438934756,
+      "loss": 1.92,
+      "step": 4620
+    },
+    {
+      "epoch": 2.118793716316936,
+      "grad_norm": 0.3911856710910797,
+      "learning_rate": 0.00020974541497906525,
+      "loss": 1.6688,
+      "step": 4621
+    },
+    {
+      "epoch": 2.119252379314299,
+      "grad_norm": 0.345805823802948,
+      "learning_rate": 0.00020954380677032526,
+      "loss": 0.8207,
+      "step": 4622
+    },
+    {
+      "epoch": 2.1197110423116614,
+      "grad_norm": 0.2564966082572937,
+      "learning_rate": 0.00020934226981258376,
+      "loss": 1.2849,
+      "step": 4623
+    },
+    {
+      "epoch": 2.1201697053090243,
+      "grad_norm": 0.18908265233039856,
+      "learning_rate": 0.0002091408041552792,
+      "loss": 1.1646,
+      "step": 4624
+    },
+    {
+      "epoch": 2.1206283683063867,
+      "grad_norm": 0.32488518953323364,
+      "learning_rate": 0.00020893940984783262,
+      "loss": 1.7903,
+      "step": 4625
+    },
+    {
+      "epoch": 2.1210870313037495,
+      "grad_norm": 0.25629740953445435,
+      "learning_rate": 0.00020873808693964746,
+      "loss": 0.9413,
+      "step": 4626
+    },
+    {
+      "epoch": 2.1215456943011124,
+      "grad_norm": 0.2971414029598236,
+      "learning_rate": 0.00020853683548010965,
+      "loss": 1.7205,
+      "step": 4627
+    },
+    {
+      "epoch": 2.122004357298475,
+      "grad_norm": 0.2897069454193115,
+      "learning_rate": 0.00020833565551858768,
+      "loss": 1.1088,
+      "step": 4628
+    },
+    {
+      "epoch": 2.1224630202958377,
+      "grad_norm": 0.21978043019771576,
+      "learning_rate": 0.0002081345471044324,
+      "loss": 1.2353,
+      "step": 4629
+    },
+    {
+      "epoch": 2.1229216832932005,
+      "grad_norm": 0.3789485692977905,
+      "learning_rate": 0.0002079335102869772,
+      "loss": 1.7919,
+      "step": 4630
+    },
+    {
+      "epoch": 2.123380346290563,
+      "grad_norm": 0.27863040566444397,
+      "learning_rate": 0.00020773254511553786,
+      "loss": 0.7957,
+      "step": 4631
+    },
+    {
+      "epoch": 2.123839009287926,
+      "grad_norm": 0.2726113796234131,
+      "learning_rate": 0.00020753165163941273,
+      "loss": 2.3215,
+      "step": 4632
+    },
+    {
+      "epoch": 2.124297672285288,
+      "grad_norm": 0.3731834888458252,
+      "learning_rate": 0.00020733082990788204,
+      "loss": 2.0028,
+      "step": 4633
+    },
+    {
+      "epoch": 2.124756335282651,
+      "grad_norm": 0.3649563789367676,
+      "learning_rate": 0.00020713007997020906,
+      "loss": 1.4407,
+      "step": 4634
+    },
+    {
+      "epoch": 2.125214998280014,
+      "grad_norm": 0.2045055776834488,
+      "learning_rate": 0.00020692940187563914,
+      "loss": 0.942,
+      "step": 4635
+    },
+    {
+      "epoch": 2.1256736612773763,
+      "grad_norm": 0.39572790265083313,
+      "learning_rate": 0.0002067287956734001,
+      "loss": 2.0579,
+      "step": 4636
+    },
+    {
+      "epoch": 2.126132324274739,
+      "grad_norm": 0.47954314947128296,
+      "learning_rate": 0.00020652826141270194,
+      "loss": 1.1333,
+      "step": 4637
+    },
+    {
+      "epoch": 2.1265909872721016,
+      "grad_norm": 0.30008918046951294,
+      "learning_rate": 0.00020632779914273757,
+      "loss": 0.9181,
+      "step": 4638
+    },
+    {
+      "epoch": 2.1270496502694645,
+      "grad_norm": 0.19218164682388306,
+      "learning_rate": 0.00020612740891268145,
+      "loss": 1.7064,
+      "step": 4639
+    },
+    {
+      "epoch": 2.1275083132668273,
+      "grad_norm": 0.36220747232437134,
+      "learning_rate": 0.00020592709077169082,
+      "loss": 1.2349,
+      "step": 4640
+    },
+    {
+      "epoch": 2.1279669762641897,
+      "grad_norm": 0.2821071445941925,
+      "learning_rate": 0.00020572684476890518,
+      "loss": 1.5732,
+      "step": 4641
+    },
+    {
+      "epoch": 2.1284256392615526,
+      "grad_norm": 0.3287753760814667,
+      "learning_rate": 0.00020552667095344635,
+      "loss": 1.9521,
+      "step": 4642
+    },
+    {
+      "epoch": 2.1288843022589155,
+      "grad_norm": 0.29272401332855225,
+      "learning_rate": 0.00020532656937441841,
+      "loss": 1.1461,
+      "step": 4643
+    },
+    {
+      "epoch": 2.129342965256278,
+      "grad_norm": 0.4012896716594696,
+      "learning_rate": 0.00020512654008090792,
+      "loss": 1.6783,
+      "step": 4644
+    },
+    {
+      "epoch": 2.1298016282536407,
+      "grad_norm": 0.2766939401626587,
+      "learning_rate": 0.00020492658312198304,
+      "loss": 0.8703,
+      "step": 4645
+    },
+    {
+      "epoch": 2.130260291251003,
+      "grad_norm": 0.20289787650108337,
+      "learning_rate": 0.00020472669854669495,
+      "loss": 1.5026,
+      "step": 4646
+    },
+    {
+      "epoch": 2.130718954248366,
+      "grad_norm": 0.37175294756889343,
+      "learning_rate": 0.00020452688640407656,
+      "loss": 1.6858,
+      "step": 4647
+    },
+    {
+      "epoch": 2.131177617245729,
+      "grad_norm": 0.28814128041267395,
+      "learning_rate": 0.00020432714674314362,
+      "loss": 1.4595,
+      "step": 4648
+    },
+    {
+      "epoch": 2.1316362802430913,
+      "grad_norm": 0.4047505557537079,
+      "learning_rate": 0.00020412747961289364,
+      "loss": 0.8171,
+      "step": 4649
+    },
+    {
+      "epoch": 2.132094943240454,
+      "grad_norm": 0.23079468309879303,
+      "learning_rate": 0.0002039278850623061,
+      "loss": 0.9852,
+      "step": 4650
+    },
+    {
+      "epoch": 2.132553606237817,
+      "grad_norm": 0.3248424232006073,
+      "learning_rate": 0.0002037283631403431,
+      "loss": 1.979,
+      "step": 4651
+    },
+    {
+      "epoch": 2.1330122692351794,
+      "grad_norm": 0.4883231222629547,
+      "learning_rate": 0.0002035289138959489,
+      "loss": 0.9462,
+      "step": 4652
+    },
+    {
+      "epoch": 2.1334709322325422,
+      "grad_norm": 0.09132955968379974,
+      "learning_rate": 0.00020332953737804978,
+      "loss": 0.8115,
+      "step": 4653
+    },
+    {
+      "epoch": 2.1339295952299047,
+      "grad_norm": 0.31292226910591125,
+      "learning_rate": 0.00020313023363555422,
+      "loss": 1.2776,
+      "step": 4654
+    },
+    {
+      "epoch": 2.1343882582272675,
+      "grad_norm": 0.32427236437797546,
+      "learning_rate": 0.00020293100271735303,
+      "loss": 1.2678,
+      "step": 4655
+    },
+    {
+      "epoch": 2.1348469212246304,
+      "grad_norm": 0.2879985272884369,
+      "learning_rate": 0.00020273184467231876,
+      "loss": 0.938,
+      "step": 4656
+    },
+    {
+      "epoch": 2.135305584221993,
+      "grad_norm": 0.2097647488117218,
+      "learning_rate": 0.00020253275954930621,
+      "loss": 1.1378,
+      "step": 4657
+    },
+    {
+      "epoch": 2.1357642472193556,
+      "grad_norm": 0.3992927670478821,
+      "learning_rate": 0.00020233374739715276,
+      "loss": 1.2597,
+      "step": 4658
+    },
+    {
+      "epoch": 2.136222910216718,
+      "grad_norm": 0.2745298147201538,
+      "learning_rate": 0.00020213480826467733,
+      "loss": 1.3282,
+      "step": 4659
+    },
+    {
+      "epoch": 2.136681573214081,
+      "grad_norm": 0.3620222806930542,
+      "learning_rate": 0.00020193594220068134,
+      "loss": 1.2953,
+      "step": 4660
+    },
+    {
+      "epoch": 2.1371402362114438,
+      "grad_norm": 0.4318819046020508,
+      "learning_rate": 0.00020173714925394775,
+      "loss": 2.0383,
+      "step": 4661
+    },
+    {
+      "epoch": 2.137598899208806,
+      "grad_norm": 0.4898906946182251,
+      "learning_rate": 0.00020153842947324196,
+      "loss": 1.1289,
+      "step": 4662
+    },
+    {
+      "epoch": 2.138057562206169,
+      "grad_norm": 0.32130667567253113,
+      "learning_rate": 0.00020133978290731152,
+      "loss": 1.5143,
+      "step": 4663
+    },
+    {
+      "epoch": 2.138516225203532,
+      "grad_norm": 0.196794331073761,
+      "learning_rate": 0.00020114120960488575,
+      "loss": 1.2338,
+      "step": 4664
+    },
+    {
+      "epoch": 2.1389748882008943,
+      "grad_norm": 0.3671046197414398,
+      "learning_rate": 0.00020094270961467614,
+      "loss": 0.8875,
+      "step": 4665
+    },
+    {
+      "epoch": 2.139433551198257,
+      "grad_norm": 0.3376992344856262,
+      "learning_rate": 0.00020074428298537633,
+      "loss": 1.4139,
+      "step": 4666
+    },
+    {
+      "epoch": 2.1398922141956196,
+      "grad_norm": 0.24704553186893463,
+      "learning_rate": 0.00020054592976566132,
+      "loss": 1.4949,
+      "step": 4667
+    },
+    {
+      "epoch": 2.1403508771929824,
+      "grad_norm": 0.21429981291294098,
+      "learning_rate": 0.00020034765000418914,
+      "loss": 0.863,
+      "step": 4668
+    },
+    {
+      "epoch": 2.1408095401903453,
+      "grad_norm": 0.2140214592218399,
+      "learning_rate": 0.0002001494437495989,
+      "loss": 0.8906,
+      "step": 4669
+    },
+    {
+      "epoch": 2.1412682031877077,
+      "grad_norm": 0.27707546949386597,
+      "learning_rate": 0.00019995131105051228,
+      "loss": 1.3032,
+      "step": 4670
+    },
+    {
+      "epoch": 2.1417268661850706,
+      "grad_norm": 0.7792873382568359,
+      "learning_rate": 0.00019975325195553263,
+      "loss": 1.3045,
+      "step": 4671
+    },
+    {
+      "epoch": 2.1421855291824334,
+      "grad_norm": 0.3791343569755554,
+      "learning_rate": 0.00019955526651324495,
+      "loss": 1.6859,
+      "step": 4672
+    },
+    {
+      "epoch": 2.142644192179796,
+      "grad_norm": 1.3164491653442383,
+      "learning_rate": 0.00019935735477221678,
+      "loss": 1.0486,
+      "step": 4673
+    },
+    {
+      "epoch": 2.1431028551771587,
+      "grad_norm": 0.318463534116745,
+      "learning_rate": 0.0001991595167809972,
+      "loss": 1.8174,
+      "step": 4674
+    },
+    {
+      "epoch": 2.143561518174521,
+      "grad_norm": 0.4365736246109009,
+      "learning_rate": 0.00019896175258811734,
+      "loss": 0.6856,
+      "step": 4675
+    },
+    {
+      "epoch": 2.144020181171884,
+      "grad_norm": 0.2647281289100647,
+      "learning_rate": 0.00019876406224209015,
+      "loss": 1.6011,
+      "step": 4676
+    },
+    {
+      "epoch": 2.144478844169247,
+      "grad_norm": 0.3120388090610504,
+      "learning_rate": 0.00019856644579141052,
+      "loss": 1.074,
+      "step": 4677
+    },
+    {
+      "epoch": 2.1449375071666092,
+      "grad_norm": 0.3103806674480438,
+      "learning_rate": 0.0001983689032845552,
+      "loss": 1.7906,
+      "step": 4678
+    },
+    {
+      "epoch": 2.145396170163972,
+      "grad_norm": 0.36245760321617126,
+      "learning_rate": 0.0001981714347699828,
+      "loss": 1.4742,
+      "step": 4679
+    },
+    {
+      "epoch": 2.1458548331613345,
+      "grad_norm": 0.36996036767959595,
+      "learning_rate": 0.00019797404029613368,
+      "loss": 1.2744,
+      "step": 4680
+    },
+    {
+      "epoch": 2.1463134961586974,
+      "grad_norm": 0.28153517842292786,
+      "learning_rate": 0.00019777671991143026,
+      "loss": 1.4012,
+      "step": 4681
+    },
+    {
+      "epoch": 2.1467721591560602,
+      "grad_norm": 0.3398634195327759,
+      "learning_rate": 0.00019757947366427653,
+      "loss": 1.4422,
+      "step": 4682
+    },
+    {
+      "epoch": 2.1472308221534226,
+      "grad_norm": 0.17714032530784607,
+      "learning_rate": 0.0001973823016030587,
+      "loss": 0.4065,
+      "step": 4683
+    },
+    {
+      "epoch": 2.1476894851507855,
+      "grad_norm": 0.21079567074775696,
+      "learning_rate": 0.00019718520377614407,
+      "loss": 1.353,
+      "step": 4684
+    },
+    {
+      "epoch": 2.148148148148148,
+      "grad_norm": 0.46706634759902954,
+      "learning_rate": 0.00019698818023188236,
+      "loss": 1.7202,
+      "step": 4685
+    },
+    {
+      "epoch": 2.1486068111455108,
+      "grad_norm": 0.2902261018753052,
+      "learning_rate": 0.00019679123101860491,
+      "loss": 1.7689,
+      "step": 4686
+    },
+    {
+      "epoch": 2.1490654741428736,
+      "grad_norm": 0.4393427073955536,
+      "learning_rate": 0.00019659435618462473,
+      "loss": 2.2651,
+      "step": 4687
+    },
+    {
+      "epoch": 2.149524137140236,
+      "grad_norm": 0.3855237364768982,
+      "learning_rate": 0.0001963975557782366,
+      "loss": 1.5914,
+      "step": 4688
+    },
+    {
+      "epoch": 2.149982800137599,
+      "grad_norm": 0.6791412234306335,
+      "learning_rate": 0.00019620082984771715,
+      "loss": 1.9954,
+      "step": 4689
+    },
+    {
+      "epoch": 2.1504414631349618,
+      "grad_norm": 0.38182568550109863,
+      "learning_rate": 0.00019600417844132463,
+      "loss": 1.1696,
+      "step": 4690
+    },
+    {
+      "epoch": 2.150900126132324,
+      "grad_norm": 0.2527628540992737,
+      "learning_rate": 0.0001958076016072991,
+      "loss": 0.83,
+      "step": 4691
+    },
+    {
+      "epoch": 2.151358789129687,
+      "grad_norm": 0.29251906275749207,
+      "learning_rate": 0.00019561109939386217,
+      "loss": 1.1837,
+      "step": 4692
+    },
+    {
+      "epoch": 2.1518174521270494,
+      "grad_norm": 0.3355877995491028,
+      "learning_rate": 0.0001954146718492174,
+      "loss": 1.3884,
+      "step": 4693
+    },
+    {
+      "epoch": 2.1522761151244123,
+      "grad_norm": 0.2906731367111206,
+      "learning_rate": 0.0001952183190215499,
+      "loss": 1.2325,
+      "step": 4694
+    },
+    {
+      "epoch": 2.152734778121775,
+      "grad_norm": 0.2701873779296875,
+      "learning_rate": 0.0001950220409590262,
+      "loss": 1.6547,
+      "step": 4695
+    },
+    {
+      "epoch": 2.1531934411191376,
+      "grad_norm": 0.3364790976047516,
+      "learning_rate": 0.00019482583770979485,
+      "loss": 1.2307,
+      "step": 4696
+    },
+    {
+      "epoch": 2.1536521041165004,
+      "grad_norm": 0.22372028231620789,
+      "learning_rate": 0.00019462970932198592,
+      "loss": 1.2701,
+      "step": 4697
+    },
+    {
+      "epoch": 2.1541107671138633,
+      "grad_norm": 0.34942564368247986,
+      "learning_rate": 0.00019443365584371114,
+      "loss": 1.3538,
+      "step": 4698
+    },
+    {
+      "epoch": 2.1545694301112257,
+      "grad_norm": 0.21902301907539368,
+      "learning_rate": 0.0001942376773230638,
+      "loss": 0.9939,
+      "step": 4699
+    },
+    {
+      "epoch": 2.1550280931085886,
+      "grad_norm": 0.3378940224647522,
+      "learning_rate": 0.00019404177380811895,
+      "loss": 1.5621,
+      "step": 4700
+    },
+    {
+      "epoch": 2.155486756105951,
+      "grad_norm": 0.44091933965682983,
+      "learning_rate": 0.00019384594534693295,
+      "loss": 1.6776,
+      "step": 4701
+    },
+    {
+      "epoch": 2.155945419103314,
+      "grad_norm": 0.3093583285808563,
+      "learning_rate": 0.00019365019198754413,
+      "loss": 1.556,
+      "step": 4702
+    },
+    {
+      "epoch": 2.1564040821006767,
+      "grad_norm": 0.30539020895957947,
+      "learning_rate": 0.00019345451377797207,
+      "loss": 0.9653,
+      "step": 4703
+    },
+    {
+      "epoch": 2.156862745098039,
+      "grad_norm": 0.2688155174255371,
+      "learning_rate": 0.0001932589107662181,
+      "loss": 1.2151,
+      "step": 4704
+    },
+    {
+      "epoch": 2.157321408095402,
+      "grad_norm": 0.29725632071495056,
+      "learning_rate": 0.0001930633830002652,
+      "loss": 1.8734,
+      "step": 4705
+    },
+    {
+      "epoch": 2.1577800710927644,
+      "grad_norm": 0.3029474914073944,
+      "learning_rate": 0.00019286793052807744,
+      "loss": 0.8146,
+      "step": 4706
+    },
+    {
+      "epoch": 2.1582387340901272,
+      "grad_norm": 0.4033948481082916,
+      "learning_rate": 0.00019267255339760082,
+      "loss": 2.1886,
+      "step": 4707
+    },
+    {
+      "epoch": 2.15869739708749,
+      "grad_norm": 0.4187861979007721,
+      "learning_rate": 0.00019247725165676276,
+      "loss": 1.5054,
+      "step": 4708
+    },
+    {
+      "epoch": 2.1591560600848525,
+      "grad_norm": 0.22975748777389526,
+      "learning_rate": 0.00019228202535347212,
+      "loss": 1.0834,
+      "step": 4709
+    },
+    {
+      "epoch": 2.1596147230822154,
+      "grad_norm": 0.4031769633293152,
+      "learning_rate": 0.00019208687453561957,
+      "loss": 1.534,
+      "step": 4710
+    },
+    {
+      "epoch": 2.160073386079578,
+      "grad_norm": 0.36004751920700073,
+      "learning_rate": 0.00019189179925107702,
+      "loss": 1.5185,
+      "step": 4711
+    },
+    {
+      "epoch": 2.1605320490769406,
+      "grad_norm": 0.230648472905159,
+      "learning_rate": 0.00019169679954769754,
+      "loss": 0.6958,
+      "step": 4712
+    },
+    {
+      "epoch": 2.1609907120743035,
+      "grad_norm": 0.38399839401245117,
+      "learning_rate": 0.0001915018754733161,
+      "loss": 1.9727,
+      "step": 4713
+    },
+    {
+      "epoch": 2.161449375071666,
+      "grad_norm": 1.3798997402191162,
+      "learning_rate": 0.00019130702707574905,
+      "loss": 1.0898,
+      "step": 4714
+    },
+    {
+      "epoch": 2.1619080380690288,
+      "grad_norm": 0.3109949231147766,
+      "learning_rate": 0.00019111225440279395,
+      "loss": 1.5253,
+      "step": 4715
+    },
+    {
+      "epoch": 2.1623667010663916,
+      "grad_norm": 0.24511021375656128,
+      "learning_rate": 0.00019091755750223028,
+      "loss": 1.6026,
+      "step": 4716
+    },
+    {
+      "epoch": 2.162825364063754,
+      "grad_norm": 0.4369632601737976,
+      "learning_rate": 0.00019072293642181815,
+      "loss": 1.6895,
+      "step": 4717
+    },
+    {
+      "epoch": 2.163284027061117,
+      "grad_norm": 0.4405476748943329,
+      "learning_rate": 0.00019052839120929977,
+      "loss": 1.3796,
+      "step": 4718
+    },
+    {
+      "epoch": 2.1637426900584797,
+      "grad_norm": 0.31771278381347656,
+      "learning_rate": 0.00019033392191239817,
+      "loss": 1.1156,
+      "step": 4719
+    },
+    {
+      "epoch": 2.164201353055842,
+      "grad_norm": 0.28369247913360596,
+      "learning_rate": 0.0001901395285788186,
+      "loss": 2.0259,
+      "step": 4720
+    },
+    {
+      "epoch": 2.164660016053205,
+      "grad_norm": 0.39868542551994324,
+      "learning_rate": 0.0001899452112562468,
+      "loss": 1.8127,
+      "step": 4721
+    },
+    {
+      "epoch": 2.1651186790505674,
+      "grad_norm": 0.24997852742671967,
+      "learning_rate": 0.00018975096999235052,
+      "loss": 1.468,
+      "step": 4722
+    },
+    {
+      "epoch": 2.1655773420479303,
+      "grad_norm": 0.3659355342388153,
+      "learning_rate": 0.0001895568048347781,
+      "loss": 1.2206,
+      "step": 4723
+    },
+    {
+      "epoch": 2.166036005045293,
+      "grad_norm": 0.20338742434978485,
+      "learning_rate": 0.00018936271583115994,
+      "loss": 1.3026,
+      "step": 4724
+    },
+    {
+      "epoch": 2.1664946680426556,
+      "grad_norm": 0.39747223258018494,
+      "learning_rate": 0.00018916870302910732,
+      "loss": 1.6281,
+      "step": 4725
+    },
+    {
+      "epoch": 2.1669533310400184,
+      "grad_norm": 0.33269479870796204,
+      "learning_rate": 0.00018897476647621308,
+      "loss": 1.9271,
+      "step": 4726
+    },
+    {
+      "epoch": 2.167411994037381,
+      "grad_norm": 0.2740723788738251,
+      "learning_rate": 0.00018878090622005138,
+      "loss": 1.1646,
+      "step": 4727
+    },
+    {
+      "epoch": 2.1678706570347437,
+      "grad_norm": 0.3122137188911438,
+      "learning_rate": 0.00018858712230817727,
+      "loss": 1.6887,
+      "step": 4728
+    },
+    {
+      "epoch": 2.1683293200321065,
+      "grad_norm": 0.3537803590297699,
+      "learning_rate": 0.00018839341478812726,
+      "loss": 1.1824,
+      "step": 4729
+    },
+    {
+      "epoch": 2.168787983029469,
+      "grad_norm": 0.273580402135849,
+      "learning_rate": 0.00018819978370741958,
+      "loss": 0.7199,
+      "step": 4730
+    },
+    {
+      "epoch": 2.169246646026832,
+      "grad_norm": 0.25088873505592346,
+      "learning_rate": 0.00018800622911355314,
+      "loss": 1.5972,
+      "step": 4731
+    },
+    {
+      "epoch": 2.1697053090241947,
+      "grad_norm": 0.38248297572135925,
+      "learning_rate": 0.0001878127510540083,
+      "loss": 1.4888,
+      "step": 4732
+    },
+    {
+      "epoch": 2.170163972021557,
+      "grad_norm": 0.23572513461112976,
+      "learning_rate": 0.00018761934957624675,
+      "loss": 1.2231,
+      "step": 4733
+    },
+    {
+      "epoch": 2.17062263501892,
+      "grad_norm": 0.35793864727020264,
+      "learning_rate": 0.00018742602472771104,
+      "loss": 1.3505,
+      "step": 4734
+    },
+    {
+      "epoch": 2.1710812980162824,
+      "grad_norm": 0.5893045663833618,
+      "learning_rate": 0.00018723277655582516,
+      "loss": 1.1632,
+      "step": 4735
+    },
+    {
+      "epoch": 2.171539961013645,
+      "grad_norm": 0.18537980318069458,
+      "learning_rate": 0.0001870396051079944,
+      "loss": 0.7823,
+      "step": 4736
+    },
+    {
+      "epoch": 2.171998624011008,
+      "grad_norm": 0.31003275513648987,
+      "learning_rate": 0.00018684651043160506,
+      "loss": 1.2098,
+      "step": 4737
+    },
+    {
+      "epoch": 2.1724572870083705,
+      "grad_norm": 0.40928885340690613,
+      "learning_rate": 0.00018665349257402465,
+      "loss": 1.1833,
+      "step": 4738
+    },
+    {
+      "epoch": 2.1729159500057333,
+      "grad_norm": 0.27787014842033386,
+      "learning_rate": 0.00018646055158260189,
+      "loss": 1.4373,
+      "step": 4739
+    },
+    {
+      "epoch": 2.173374613003096,
+      "grad_norm": 0.2921246886253357,
+      "learning_rate": 0.00018626768750466656,
+      "loss": 1.1901,
+      "step": 4740
+    },
+    {
+      "epoch": 2.1738332760004586,
+      "grad_norm": 0.36071351170539856,
+      "learning_rate": 0.00018607490038752956,
+      "loss": 1.4543,
+      "step": 4741
+    },
+    {
+      "epoch": 2.1742919389978215,
+      "grad_norm": 0.2633201777935028,
+      "learning_rate": 0.00018588219027848303,
+      "loss": 1.1895,
+      "step": 4742
+    },
+    {
+      "epoch": 2.174750601995184,
+      "grad_norm": 0.30791527032852173,
+      "learning_rate": 0.0001856895572248002,
+      "loss": 1.2403,
+      "step": 4743
+    },
+    {
+      "epoch": 2.1752092649925467,
+      "grad_norm": 0.2025088369846344,
+      "learning_rate": 0.00018549700127373537,
+      "loss": 1.7025,
+      "step": 4744
+    },
+    {
+      "epoch": 2.1756679279899096,
+      "grad_norm": 0.34419453144073486,
+      "learning_rate": 0.00018530452247252367,
+      "loss": 1.7213,
+      "step": 4745
+    },
+    {
+      "epoch": 2.176126590987272,
+      "grad_norm": 0.3504865765571594,
+      "learning_rate": 0.00018511212086838163,
+      "loss": 1.312,
+      "step": 4746
+    },
+    {
+      "epoch": 2.176585253984635,
+      "grad_norm": 0.2538597881793976,
+      "learning_rate": 0.00018491979650850688,
+      "loss": 0.6609,
+      "step": 4747
+    },
+    {
+      "epoch": 2.1770439169819973,
+      "grad_norm": 0.226425439119339,
+      "learning_rate": 0.00018472754944007786,
+      "loss": 1.7527,
+      "step": 4748
+    },
+    {
+      "epoch": 2.17750257997936,
+      "grad_norm": 0.3634589910507202,
+      "learning_rate": 0.0001845353797102542,
+      "loss": 0.7849,
+      "step": 4749
+    },
+    {
+      "epoch": 2.177961242976723,
+      "grad_norm": 0.19483381509780884,
+      "learning_rate": 0.00018434328736617652,
+      "loss": 1.1454,
+      "step": 4750
+    },
+    {
+      "epoch": 2.1784199059740854,
+      "grad_norm": 0.2180272936820984,
+      "learning_rate": 0.00018415127245496643,
+      "loss": 1.18,
+      "step": 4751
+    },
+    {
+      "epoch": 2.1788785689714483,
+      "grad_norm": 0.36905816197395325,
+      "learning_rate": 0.0001839593350237266,
+      "loss": 1.0663,
+      "step": 4752
+    },
+    {
+      "epoch": 2.1793372319688107,
+      "grad_norm": 0.23431743681430817,
+      "learning_rate": 0.00018376747511954068,
+      "loss": 1.481,
+      "step": 4753
+    },
+    {
+      "epoch": 2.1797958949661735,
+      "grad_norm": 0.2120712399482727,
+      "learning_rate": 0.00018357569278947323,
+      "loss": 0.5578,
+      "step": 4754
+    },
+    {
+      "epoch": 2.1802545579635364,
+      "grad_norm": 0.29886701703071594,
+      "learning_rate": 0.00018338398808057004,
+      "loss": 1.2928,
+      "step": 4755
+    },
+    {
+      "epoch": 2.180713220960899,
+      "grad_norm": 0.09233911335468292,
+      "learning_rate": 0.00018319236103985737,
+      "loss": 1.2831,
+      "step": 4756
+    },
+    {
+      "epoch": 2.1811718839582617,
+      "grad_norm": 0.33295729756355286,
+      "learning_rate": 0.00018300081171434285,
+      "loss": 1.4729,
+      "step": 4757
+    },
+    {
+      "epoch": 2.1816305469556245,
+      "grad_norm": 0.42028337717056274,
+      "learning_rate": 0.00018280934015101486,
+      "loss": 1.1091,
+      "step": 4758
+    },
+    {
+      "epoch": 2.182089209952987,
+      "grad_norm": 0.21170879900455475,
+      "learning_rate": 0.00018261794639684283,
+      "loss": 0.8437,
+      "step": 4759
+    },
+    {
+      "epoch": 2.18254787295035,
+      "grad_norm": 0.22131375968456268,
+      "learning_rate": 0.00018242663049877696,
+      "loss": 0.491,
+      "step": 4760
+    },
+    {
+      "epoch": 2.183006535947712,
+      "grad_norm": 0.22741609811782837,
+      "learning_rate": 0.00018223539250374844,
+      "loss": 1.4339,
+      "step": 4761
+    },
+    {
+      "epoch": 2.183465198945075,
+      "grad_norm": 0.353257417678833,
+      "learning_rate": 0.00018204423245866936,
+      "loss": 1.5006,
+      "step": 4762
+    },
+    {
+      "epoch": 2.183923861942438,
+      "grad_norm": 0.21219374239444733,
+      "learning_rate": 0.00018185315041043267,
+      "loss": 0.8634,
+      "step": 4763
+    },
+    {
+      "epoch": 2.1843825249398003,
+      "grad_norm": 0.3175276815891266,
+      "learning_rate": 0.00018166214640591205,
+      "loss": 1.6262,
+      "step": 4764
+    },
+    {
+      "epoch": 2.184841187937163,
+      "grad_norm": 0.4105018377304077,
+      "learning_rate": 0.0001814712204919623,
+      "loss": 1.631,
+      "step": 4765
+    },
+    {
+      "epoch": 2.185299850934526,
+      "grad_norm": 0.36518141627311707,
+      "learning_rate": 0.0001812803727154189,
+      "loss": 1.6072,
+      "step": 4766
+    },
+    {
+      "epoch": 2.1857585139318885,
+      "grad_norm": 0.2514982521533966,
+      "learning_rate": 0.0001810896031230983,
+      "loss": 1.1576,
+      "step": 4767
+    },
+    {
+      "epoch": 2.1862171769292513,
+      "grad_norm": 0.33633479475975037,
+      "learning_rate": 0.0001808989117617974,
+      "loss": 1.3829,
+      "step": 4768
+    },
+    {
+      "epoch": 2.1866758399266137,
+      "grad_norm": 0.2530759274959564,
+      "learning_rate": 0.00018070829867829425,
+      "loss": 1.0331,
+      "step": 4769
+    },
+    {
+      "epoch": 2.1871345029239766,
+      "grad_norm": 0.25773900747299194,
+      "learning_rate": 0.0001805177639193476,
+      "loss": 0.5553,
+      "step": 4770
+    },
+    {
+      "epoch": 2.1875931659213395,
+      "grad_norm": 0.28829655051231384,
+      "learning_rate": 0.00018032730753169714,
+      "loss": 2.219,
+      "step": 4771
+    },
+    {
+      "epoch": 2.188051828918702,
+      "grad_norm": 0.36159905791282654,
+      "learning_rate": 0.00018013692956206302,
+      "loss": 1.0755,
+      "step": 4772
+    },
+    {
+      "epoch": 2.1885104919160647,
+      "grad_norm": 0.40641242265701294,
+      "learning_rate": 0.00017994663005714646,
+      "loss": 1.019,
+      "step": 4773
+    },
+    {
+      "epoch": 2.188969154913427,
+      "grad_norm": 0.2221248745918274,
+      "learning_rate": 0.00017975640906362923,
+      "loss": 1.5312,
+      "step": 4774
+    },
+    {
+      "epoch": 2.18942781791079,
+      "grad_norm": 0.23253923654556274,
+      "learning_rate": 0.00017956626662817387,
+      "loss": 0.9588,
+      "step": 4775
+    },
+    {
+      "epoch": 2.189886480908153,
+      "grad_norm": 0.3233194947242737,
+      "learning_rate": 0.00017937620279742384,
+      "loss": 1.1987,
+      "step": 4776
+    },
+    {
+      "epoch": 2.1903451439055153,
+      "grad_norm": 0.4366668164730072,
+      "learning_rate": 0.0001791862176180031,
+      "loss": 1.6294,
+      "step": 4777
+    },
+    {
+      "epoch": 2.190803806902878,
+      "grad_norm": 0.4189610481262207,
+      "learning_rate": 0.00017899631113651643,
+      "loss": 1.3567,
+      "step": 4778
+    },
+    {
+      "epoch": 2.191262469900241,
+      "grad_norm": 0.28253674507141113,
+      "learning_rate": 0.00017880648339954914,
+      "loss": 1.637,
+      "step": 4779
+    },
+    {
+      "epoch": 2.1917211328976034,
+      "grad_norm": 0.43727073073387146,
+      "learning_rate": 0.00017861673445366733,
+      "loss": 1.8063,
+      "step": 4780
+    },
+    {
+      "epoch": 2.1921797958949663,
+      "grad_norm": 0.2974712550640106,
+      "learning_rate": 0.00017842706434541785,
+      "loss": 1.1469,
+      "step": 4781
+    },
+    {
+      "epoch": 2.1926384588923287,
+      "grad_norm": 0.3751980662345886,
+      "learning_rate": 0.00017823747312132798,
+      "loss": 0.908,
+      "step": 4782
+    },
+    {
+      "epoch": 2.1930971218896915,
+      "grad_norm": 0.15295343101024628,
+      "learning_rate": 0.00017804796082790636,
+      "loss": 0.8782,
+      "step": 4783
+    },
+    {
+      "epoch": 2.1935557848870544,
+      "grad_norm": 0.33498823642730713,
+      "learning_rate": 0.00017785852751164117,
+      "loss": 1.3965,
+      "step": 4784
+    },
+    {
+      "epoch": 2.194014447884417,
+      "grad_norm": 0.3449128568172455,
+      "learning_rate": 0.0001776691732190019,
+      "loss": 1.7836,
+      "step": 4785
+    },
+    {
+      "epoch": 2.1944731108817797,
+      "grad_norm": 0.3249196708202362,
+      "learning_rate": 0.0001774798979964386,
+      "loss": 1.1383,
+      "step": 4786
+    },
+    {
+      "epoch": 2.1949317738791425,
+      "grad_norm": 0.3362160325050354,
+      "learning_rate": 0.00017729070189038176,
+      "loss": 1.3211,
+      "step": 4787
+    },
+    {
+      "epoch": 2.195390436876505,
+      "grad_norm": 0.32171592116355896,
+      "learning_rate": 0.00017710158494724265,
+      "loss": 1.9521,
+      "step": 4788
+    },
+    {
+      "epoch": 2.195849099873868,
+      "grad_norm": 0.39043116569519043,
+      "learning_rate": 0.0001769125472134131,
+      "loss": 1.2834,
+      "step": 4789
+    },
+    {
+      "epoch": 2.19630776287123,
+      "grad_norm": 0.29517433047294617,
+      "learning_rate": 0.00017672358873526518,
+      "loss": 1.4635,
+      "step": 4790
+    },
+    {
+      "epoch": 2.196766425868593,
+      "grad_norm": 0.4014551639556885,
+      "learning_rate": 0.0001765347095591517,
+      "loss": 1.3101,
+      "step": 4791
+    },
+    {
+      "epoch": 2.197225088865956,
+      "grad_norm": 0.247705340385437,
+      "learning_rate": 0.0001763459097314064,
+      "loss": 1.5868,
+      "step": 4792
+    },
+    {
+      "epoch": 2.1976837518633183,
+      "grad_norm": 0.2868292033672333,
+      "learning_rate": 0.00017615718929834317,
+      "loss": 1.5704,
+      "step": 4793
+    },
+    {
+      "epoch": 2.198142414860681,
+      "grad_norm": 0.24143101274967194,
+      "learning_rate": 0.00017596854830625642,
+      "loss": 1.4587,
+      "step": 4794
+    },
+    {
+      "epoch": 2.1986010778580436,
+      "grad_norm": 0.2784029245376587,
+      "learning_rate": 0.00017577998680142132,
+      "loss": 1.2386,
+      "step": 4795
+    },
+    {
+      "epoch": 2.1990597408554065,
+      "grad_norm": 0.336953729391098,
+      "learning_rate": 0.00017559150483009302,
+      "loss": 1.296,
+      "step": 4796
+    },
+    {
+      "epoch": 2.1995184038527693,
+      "grad_norm": 0.39816170930862427,
+      "learning_rate": 0.0001754031024385077,
+      "loss": 1.5718,
+      "step": 4797
+    },
+    {
+      "epoch": 2.1999770668501317,
+      "grad_norm": 0.34481489658355713,
+      "learning_rate": 0.0001752147796728818,
+      "loss": 2.1771,
+      "step": 4798
+    },
+    {
+      "epoch": 2.2004357298474946,
+      "grad_norm": 0.3532600998878479,
+      "learning_rate": 0.0001750265365794123,
+      "loss": 1.6757,
+      "step": 4799
+    },
+    {
+      "epoch": 2.2008943928448574,
+      "grad_norm": 0.4431455135345459,
+      "learning_rate": 0.0001748383732042767,
+      "loss": 1.632,
+      "step": 4800
+    },
+    {
+      "epoch": 2.20135305584222,
+      "grad_norm": 0.3615950047969818,
+      "learning_rate": 0.00017465028959363238,
+      "loss": 1.4655,
+      "step": 4801
+    },
+    {
+      "epoch": 2.2018117188395827,
+      "grad_norm": 0.21566683053970337,
+      "learning_rate": 0.00017446228579361806,
+      "loss": 0.9168,
+      "step": 4802
+    },
+    {
+      "epoch": 2.202270381836945,
+      "grad_norm": 0.3532969057559967,
+      "learning_rate": 0.00017427436185035234,
+      "loss": 2.1003,
+      "step": 4803
+    },
+    {
+      "epoch": 2.202729044834308,
+      "grad_norm": 0.4000427722930908,
+      "learning_rate": 0.00017408651780993417,
+      "loss": 2.0688,
+      "step": 4804
+    },
+    {
+      "epoch": 2.203187707831671,
+      "grad_norm": 0.2674494981765747,
+      "learning_rate": 0.0001738987537184432,
+      "loss": 0.7786,
+      "step": 4805
+    },
+    {
+      "epoch": 2.2036463708290333,
+      "grad_norm": 0.2524220049381256,
+      "learning_rate": 0.00017371106962193938,
+      "loss": 1.5247,
+      "step": 4806
+    },
+    {
+      "epoch": 2.204105033826396,
+      "grad_norm": 0.3025605082511902,
+      "learning_rate": 0.00017352346556646277,
+      "loss": 0.6016,
+      "step": 4807
+    },
+    {
+      "epoch": 2.204563696823759,
+      "grad_norm": 0.07593965530395508,
+      "learning_rate": 0.00017333594159803397,
+      "loss": 0.8819,
+      "step": 4808
+    },
+    {
+      "epoch": 2.2050223598211214,
+      "grad_norm": 0.322405606508255,
+      "learning_rate": 0.00017314849776265412,
+      "loss": 0.8851,
+      "step": 4809
+    },
+    {
+      "epoch": 2.2054810228184842,
+      "grad_norm": 0.25816667079925537,
+      "learning_rate": 0.0001729611341063045,
+      "loss": 1.9113,
+      "step": 4810
+    },
+    {
+      "epoch": 2.2059396858158467,
+      "grad_norm": 0.31706973910331726,
+      "learning_rate": 0.00017277385067494672,
+      "loss": 0.929,
+      "step": 4811
+    },
+    {
+      "epoch": 2.2063983488132095,
+      "grad_norm": 0.16823376715183258,
+      "learning_rate": 0.0001725866475145228,
+      "loss": 1.0099,
+      "step": 4812
+    },
+    {
+      "epoch": 2.2068570118105724,
+      "grad_norm": 0.28849852085113525,
+      "learning_rate": 0.00017239952467095498,
+      "loss": 1.4562,
+      "step": 4813
+    },
+    {
+      "epoch": 2.207315674807935,
+      "grad_norm": 0.3034904897212982,
+      "learning_rate": 0.00017221248219014595,
+      "loss": 1.6464,
+      "step": 4814
+    },
+    {
+      "epoch": 2.2077743378052976,
+      "grad_norm": 0.31821388006210327,
+      "learning_rate": 0.00017202552011797852,
+      "loss": 1.2505,
+      "step": 4815
+    },
+    {
+      "epoch": 2.20823300080266,
+      "grad_norm": 0.3559507429599762,
+      "learning_rate": 0.00017183863850031572,
+      "loss": 1.3757,
+      "step": 4816
+    },
+    {
+      "epoch": 2.208691663800023,
+      "grad_norm": 0.20952913165092468,
+      "learning_rate": 0.00017165183738300133,
+      "loss": 0.3598,
+      "step": 4817
+    },
+    {
+      "epoch": 2.2091503267973858,
+      "grad_norm": 0.0632583424448967,
+      "learning_rate": 0.0001714651168118585,
+      "loss": 0.3411,
+      "step": 4818
+    },
+    {
+      "epoch": 2.209608989794748,
+      "grad_norm": 0.13204383850097656,
+      "learning_rate": 0.00017127847683269144,
+      "loss": 1.1918,
+      "step": 4819
+    },
+    {
+      "epoch": 2.210067652792111,
+      "grad_norm": 0.30054593086242676,
+      "learning_rate": 0.00017109191749128418,
+      "loss": 1.1966,
+      "step": 4820
+    },
+    {
+      "epoch": 2.2105263157894735,
+      "grad_norm": 0.35834217071533203,
+      "learning_rate": 0.00017090543883340115,
+      "loss": 2.2354,
+      "step": 4821
+    },
+    {
+      "epoch": 2.2109849787868363,
+      "grad_norm": 0.3150874674320221,
+      "learning_rate": 0.00017071904090478686,
+      "loss": 0.3315,
+      "step": 4822
+    },
+    {
+      "epoch": 2.211443641784199,
+      "grad_norm": 0.2139630913734436,
+      "learning_rate": 0.00017053272375116603,
+      "loss": 1.3162,
+      "step": 4823
+    },
+    {
+      "epoch": 2.2119023047815616,
+      "grad_norm": 0.35393640398979187,
+      "learning_rate": 0.00017034648741824366,
+      "loss": 2.103,
+      "step": 4824
+    },
+    {
+      "epoch": 2.2123609677789244,
+      "grad_norm": 0.43089064955711365,
+      "learning_rate": 0.00017016033195170488,
+      "loss": 1.7612,
+      "step": 4825
+    },
+    {
+      "epoch": 2.2128196307762873,
+      "grad_norm": 0.3777117431163788,
+      "learning_rate": 0.00016997425739721488,
+      "loss": 1.6729,
+      "step": 4826
+    },
+    {
+      "epoch": 2.2132782937736497,
+      "grad_norm": 0.3245319426059723,
+      "learning_rate": 0.00016978826380041923,
+      "loss": 1.1172,
+      "step": 4827
+    },
+    {
+      "epoch": 2.2137369567710126,
+      "grad_norm": 0.2487240731716156,
+      "learning_rate": 0.0001696023512069435,
+      "loss": 1.145,
+      "step": 4828
+    },
+    {
+      "epoch": 2.214195619768375,
+      "grad_norm": 0.1767636239528656,
+      "learning_rate": 0.00016941651966239325,
+      "loss": 0.94,
+      "step": 4829
+    },
+    {
+      "epoch": 2.214654282765738,
+      "grad_norm": 0.2936687171459198,
+      "learning_rate": 0.00016923076921235424,
+      "loss": 1.3027,
+      "step": 4830
+    },
+    {
+      "epoch": 2.2151129457631007,
+      "grad_norm": 0.5650012493133545,
+      "learning_rate": 0.00016904509990239258,
+      "loss": 1.048,
+      "step": 4831
+    },
+    {
+      "epoch": 2.215571608760463,
+      "grad_norm": 0.266467809677124,
+      "learning_rate": 0.00016885951177805425,
+      "loss": 1.7801,
+      "step": 4832
+    },
+    {
+      "epoch": 2.216030271757826,
+      "grad_norm": 0.39118704199790955,
+      "learning_rate": 0.00016867400488486528,
+      "loss": 1.2689,
+      "step": 4833
+    },
+    {
+      "epoch": 2.216488934755189,
+      "grad_norm": 0.3572444021701813,
+      "learning_rate": 0.0001684885792683319,
+      "loss": 2.2139,
+      "step": 4834
+    },
+    {
+      "epoch": 2.2169475977525512,
+      "grad_norm": 0.351784348487854,
+      "learning_rate": 0.00016830323497394033,
+      "loss": 1.1197,
+      "step": 4835
+    },
+    {
+      "epoch": 2.217406260749914,
+      "grad_norm": 0.19013693928718567,
+      "learning_rate": 0.0001681179720471569,
+      "loss": 1.0192,
+      "step": 4836
+    },
+    {
+      "epoch": 2.2178649237472765,
+      "grad_norm": 0.35627466440200806,
+      "learning_rate": 0.00016793279053342792,
+      "loss": 1.6245,
+      "step": 4837
+    },
+    {
+      "epoch": 2.2183235867446394,
+      "grad_norm": 0.34376806020736694,
+      "learning_rate": 0.00016774769047817978,
+      "loss": 1.575,
+      "step": 4838
+    },
+    {
+      "epoch": 2.218782249742002,
+      "grad_norm": 0.24683015048503876,
+      "learning_rate": 0.00016756267192681896,
+      "loss": 0.9045,
+      "step": 4839
+    },
+    {
+      "epoch": 2.2192409127393646,
+      "grad_norm": 0.19168370962142944,
+      "learning_rate": 0.00016737773492473152,
+      "loss": 0.8613,
+      "step": 4840
+    },
+    {
+      "epoch": 2.2196995757367275,
+      "grad_norm": 0.29889073967933655,
+      "learning_rate": 0.00016719287951728407,
+      "loss": 0.8257,
+      "step": 4841
+    },
+    {
+      "epoch": 2.22015823873409,
+      "grad_norm": 0.0963655561208725,
+      "learning_rate": 0.00016700810574982294,
+      "loss": 0.5904,
+      "step": 4842
+    },
+    {
+      "epoch": 2.2206169017314528,
+      "grad_norm": 0.2194061130285263,
+      "learning_rate": 0.00016682341366767444,
+      "loss": 1.1664,
+      "step": 4843
+    },
+    {
+      "epoch": 2.2210755647288156,
+      "grad_norm": 0.20055797696113586,
+      "learning_rate": 0.0001666388033161448,
+      "loss": 1.1049,
+      "step": 4844
+    },
+    {
+      "epoch": 2.221534227726178,
+      "grad_norm": 0.2792820334434509,
+      "learning_rate": 0.0001664542747405206,
+      "loss": 1.011,
+      "step": 4845
+    },
+    {
+      "epoch": 2.221992890723541,
+      "grad_norm": 0.243804931640625,
+      "learning_rate": 0.0001662698279860677,
+      "loss": 1.5193,
+      "step": 4846
+    },
+    {
+      "epoch": 2.2224515537209037,
+      "grad_norm": 0.42623358964920044,
+      "learning_rate": 0.00016608546309803229,
+      "loss": 1.7656,
+      "step": 4847
+    },
+    {
+      "epoch": 2.222910216718266,
+      "grad_norm": 0.5009300112724304,
+      "learning_rate": 0.00016590118012164046,
+      "loss": 1.7704,
+      "step": 4848
+    },
+    {
+      "epoch": 2.223368879715629,
+      "grad_norm": 0.20968489348888397,
+      "learning_rate": 0.0001657169791020981,
+      "loss": 0.8382,
+      "step": 4849
+    },
+    {
+      "epoch": 2.2238275427129914,
+      "grad_norm": 0.3352392017841339,
+      "learning_rate": 0.00016553286008459117,
+      "loss": 1.5792,
+      "step": 4850
+    },
+    {
+      "epoch": 2.2242862057103543,
+      "grad_norm": 0.27464622259140015,
+      "learning_rate": 0.00016534882311428523,
+      "loss": 0.928,
+      "step": 4851
+    },
+    {
+      "epoch": 2.224744868707717,
+      "grad_norm": 0.1357714831829071,
+      "learning_rate": 0.00016516486823632586,
+      "loss": 0.797,
+      "step": 4852
+    },
+    {
+      "epoch": 2.2252035317050796,
+      "grad_norm": 0.36050960421562195,
+      "learning_rate": 0.00016498099549583866,
+      "loss": 1.7069,
+      "step": 4853
+    },
+    {
+      "epoch": 2.2256621947024424,
+      "grad_norm": 0.367965430021286,
+      "learning_rate": 0.00016479720493792872,
+      "loss": 1.2835,
+      "step": 4854
+    },
+    {
+      "epoch": 2.2261208576998053,
+      "grad_norm": 0.2768895626068115,
+      "learning_rate": 0.00016461349660768144,
+      "loss": 1.2358,
+      "step": 4855
+    },
+    {
+      "epoch": 2.2265795206971677,
+      "grad_norm": 0.3254320025444031,
+      "learning_rate": 0.00016442987055016194,
+      "loss": 0.8148,
+      "step": 4856
+    },
+    {
+      "epoch": 2.2270381836945305,
+      "grad_norm": 0.21533845365047455,
+      "learning_rate": 0.00016424632681041456,
+      "loss": 1.7542,
+      "step": 4857
+    },
+    {
+      "epoch": 2.227496846691893,
+      "grad_norm": 0.2936849594116211,
+      "learning_rate": 0.00016406286543346415,
+      "loss": 0.4233,
+      "step": 4858
+    },
+    {
+      "epoch": 2.227955509689256,
+      "grad_norm": 0.5081586241722107,
+      "learning_rate": 0.0001638794864643151,
+      "loss": 1.5443,
+      "step": 4859
+    },
+    {
+      "epoch": 2.2284141726866187,
+      "grad_norm": 0.24850165843963623,
+      "learning_rate": 0.00016369618994795156,
+      "loss": 0.9991,
+      "step": 4860
+    },
+    {
+      "epoch": 2.228872835683981,
+      "grad_norm": 0.3500426113605499,
+      "learning_rate": 0.0001635129759293375,
+      "loss": 1.6708,
+      "step": 4861
+    },
+    {
+      "epoch": 2.229331498681344,
+      "grad_norm": 0.25747743248939514,
+      "learning_rate": 0.00016332984445341681,
+      "loss": 1.2169,
+      "step": 4862
+    },
+    {
+      "epoch": 2.2297901616787064,
+      "grad_norm": 0.31107261776924133,
+      "learning_rate": 0.0001631467955651124,
+      "loss": 1.2371,
+      "step": 4863
+    },
+    {
+      "epoch": 2.230248824676069,
+      "grad_norm": 0.25597137212753296,
+      "learning_rate": 0.00016296382930932812,
+      "loss": 0.7018,
+      "step": 4864
+    },
+    {
+      "epoch": 2.230707487673432,
+      "grad_norm": 0.21790584921836853,
+      "learning_rate": 0.00016278094573094666,
+      "loss": 1.2848,
+      "step": 4865
+    },
+    {
+      "epoch": 2.2311661506707945,
+      "grad_norm": 0.30176883935928345,
+      "learning_rate": 0.00016259814487483066,
+      "loss": 1.7701,
+      "step": 4866
+    },
+    {
+      "epoch": 2.2316248136681573,
+      "grad_norm": 0.3356066644191742,
+      "learning_rate": 0.00016241542678582268,
+      "loss": 1.2003,
+      "step": 4867
+    },
+    {
+      "epoch": 2.23208347666552,
+      "grad_norm": 0.3005014955997467,
+      "learning_rate": 0.00016223279150874448,
+      "loss": 1.0873,
+      "step": 4868
+    },
+    {
+      "epoch": 2.2325421396628826,
+      "grad_norm": 0.2405785322189331,
+      "learning_rate": 0.00016205023908839793,
+      "loss": 1.3601,
+      "step": 4869
+    },
+    {
+      "epoch": 2.2330008026602455,
+      "grad_norm": 0.3136742115020752,
+      "learning_rate": 0.00016186776956956451,
+      "loss": 1.1295,
+      "step": 4870
+    },
+    {
+      "epoch": 2.233459465657608,
+      "grad_norm": 0.18356893956661224,
+      "learning_rate": 0.00016168538299700519,
+      "loss": 1.1707,
+      "step": 4871
+    },
+    {
+      "epoch": 2.2339181286549707,
+      "grad_norm": 0.2986769676208496,
+      "learning_rate": 0.00016150307941546088,
+      "loss": 0.9671,
+      "step": 4872
+    },
+    {
+      "epoch": 2.2343767916523336,
+      "grad_norm": 0.2441634237766266,
+      "learning_rate": 0.00016132085886965187,
+      "loss": 0.5982,
+      "step": 4873
+    },
+    {
+      "epoch": 2.234835454649696,
+      "grad_norm": 0.18692080676555634,
+      "learning_rate": 0.00016113872140427815,
+      "loss": 1.5326,
+      "step": 4874
+    },
+    {
+      "epoch": 2.235294117647059,
+      "grad_norm": 0.3282175362110138,
+      "learning_rate": 0.00016095666706401941,
+      "loss": 1.2403,
+      "step": 4875
+    },
+    {
+      "epoch": 2.2357527806444217,
+      "grad_norm": 0.33327755331993103,
+      "learning_rate": 0.00016077469589353483,
+      "loss": 1.1533,
+      "step": 4876
+    },
+    {
+      "epoch": 2.236211443641784,
+      "grad_norm": 0.274987131357193,
+      "learning_rate": 0.00016059280793746333,
+      "loss": 1.6538,
+      "step": 4877
+    },
+    {
+      "epoch": 2.236670106639147,
+      "grad_norm": 0.43245795369148254,
+      "learning_rate": 0.00016041100324042345,
+      "loss": 1.9604,
+      "step": 4878
+    },
+    {
+      "epoch": 2.2371287696365094,
+      "grad_norm": 0.3761482238769531,
+      "learning_rate": 0.00016022928184701286,
+      "loss": 1.5767,
+      "step": 4879
+    },
+    {
+      "epoch": 2.2375874326338723,
+      "grad_norm": 0.3432115316390991,
+      "learning_rate": 0.0001600476438018093,
+      "loss": 1.6666,
+      "step": 4880
+    },
+    {
+      "epoch": 2.238046095631235,
+      "grad_norm": 0.3119170367717743,
+      "learning_rate": 0.00015986608914936995,
+      "loss": 1.251,
+      "step": 4881
+    },
+    {
+      "epoch": 2.2385047586285975,
+      "grad_norm": 0.4060390591621399,
+      "learning_rate": 0.0001596846179342314,
+      "loss": 0.4879,
+      "step": 4882
+    },
+    {
+      "epoch": 2.2389634216259604,
+      "grad_norm": 0.20119361579418182,
+      "learning_rate": 0.00015950323020090984,
+      "loss": 1.4327,
+      "step": 4883
+    },
+    {
+      "epoch": 2.239422084623323,
+      "grad_norm": 0.21751080453395844,
+      "learning_rate": 0.00015932192599390105,
+      "loss": 1.6125,
+      "step": 4884
+    },
+    {
+      "epoch": 2.2398807476206857,
+      "grad_norm": 0.40151306986808777,
+      "learning_rate": 0.00015914070535768022,
+      "loss": 1.9309,
+      "step": 4885
+    },
+    {
+      "epoch": 2.2403394106180485,
+      "grad_norm": 0.35669711232185364,
+      "learning_rate": 0.00015895956833670205,
+      "loss": 1.8298,
+      "step": 4886
+    },
+    {
+      "epoch": 2.240798073615411,
+      "grad_norm": 0.3669508993625641,
+      "learning_rate": 0.00015877851497540085,
+      "loss": 1.6097,
+      "step": 4887
+    },
+    {
+      "epoch": 2.241256736612774,
+      "grad_norm": 0.2940203845500946,
+      "learning_rate": 0.00015859754531819028,
+      "loss": 0.3973,
+      "step": 4888
+    },
+    {
+      "epoch": 2.241715399610136,
+      "grad_norm": 0.19039645791053772,
+      "learning_rate": 0.00015841665940946343,
+      "loss": 1.3714,
+      "step": 4889
+    },
+    {
+      "epoch": 2.242174062607499,
+      "grad_norm": 0.3116855025291443,
+      "learning_rate": 0.00015823585729359314,
+      "loss": 1.6032,
+      "step": 4890
+    },
+    {
+      "epoch": 2.242632725604862,
+      "grad_norm": 0.32782799005508423,
+      "learning_rate": 0.00015805513901493118,
+      "loss": 1.1902,
+      "step": 4891
+    },
+    {
+      "epoch": 2.2430913886022243,
+      "grad_norm": 0.1876303255558014,
+      "learning_rate": 0.0001578745046178091,
+      "loss": 0.6099,
+      "step": 4892
+    },
+    {
+      "epoch": 2.243550051599587,
+      "grad_norm": 0.32714807987213135,
+      "learning_rate": 0.00015769395414653797,
+      "loss": 1.6815,
+      "step": 4893
+    },
+    {
+      "epoch": 2.24400871459695,
+      "grad_norm": 0.2183062732219696,
+      "learning_rate": 0.00015751348764540802,
+      "loss": 0.6329,
+      "step": 4894
+    },
+    {
+      "epoch": 2.2444673775943125,
+      "grad_norm": 0.12229790538549423,
+      "learning_rate": 0.00015733310515868897,
+      "loss": 1.269,
+      "step": 4895
+    },
+    {
+      "epoch": 2.2449260405916753,
+      "grad_norm": 0.38500019907951355,
+      "learning_rate": 0.00015715280673062997,
+      "loss": 1.7126,
+      "step": 4896
+    },
+    {
+      "epoch": 2.245384703589038,
+      "grad_norm": 0.33836498856544495,
+      "learning_rate": 0.00015697259240545958,
+      "loss": 1.7375,
+      "step": 4897
+    },
+    {
+      "epoch": 2.2458433665864006,
+      "grad_norm": 0.4215908646583557,
+      "learning_rate": 0.00015679246222738562,
+      "loss": 2.0007,
+      "step": 4898
+    },
+    {
+      "epoch": 2.2463020295837635,
+      "grad_norm": 0.38347312808036804,
+      "learning_rate": 0.0001566124162405953,
+      "loss": 1.9561,
+      "step": 4899
+    },
+    {
+      "epoch": 2.246760692581126,
+      "grad_norm": 0.34876549243927,
+      "learning_rate": 0.0001564324544892553,
+      "loss": 1.7222,
+      "step": 4900
+    },
+    {
+      "epoch": 2.2472193555784887,
+      "grad_norm": 0.29349571466445923,
+      "learning_rate": 0.00015625257701751155,
+      "loss": 1.0293,
+      "step": 4901
+    },
+    {
+      "epoch": 2.2476780185758516,
+      "grad_norm": 0.3368821442127228,
+      "learning_rate": 0.00015607278386948909,
+      "loss": 1.2804,
+      "step": 4902
+    },
+    {
+      "epoch": 2.248136681573214,
+      "grad_norm": 0.3242160975933075,
+      "learning_rate": 0.00015589307508929258,
+      "loss": 1.6011,
+      "step": 4903
+    },
+    {
+      "epoch": 2.248595344570577,
+      "grad_norm": 0.2980966567993164,
+      "learning_rate": 0.0001557134507210059,
+      "loss": 1.2891,
+      "step": 4904
+    },
+    {
+      "epoch": 2.2490540075679393,
+      "grad_norm": 0.32442331314086914,
+      "learning_rate": 0.00015553391080869218,
+      "loss": 1.694,
+      "step": 4905
+    },
+    {
+      "epoch": 2.249512670565302,
+      "grad_norm": 0.328112930059433,
+      "learning_rate": 0.00015535445539639382,
+      "loss": 1.6802,
+      "step": 4906
+    },
+    {
+      "epoch": 2.249971333562665,
+      "grad_norm": 0.275643914937973,
+      "learning_rate": 0.0001551750845281326,
+      "loss": 1.1635,
+      "step": 4907
+    },
+    {
+      "epoch": 2.2504299965600274,
+      "grad_norm": 0.31056198477745056,
+      "learning_rate": 0.00015499579824790948,
+      "loss": 1.6135,
+      "step": 4908
+    },
+    {
+      "epoch": 2.2508886595573903,
+      "grad_norm": 0.31692421436309814,
+      "learning_rate": 0.0001548165965997047,
+      "loss": 1.5069,
+      "step": 4909
+    },
+    {
+      "epoch": 2.2513473225547527,
+      "grad_norm": 0.3236372768878937,
+      "learning_rate": 0.00015463747962747766,
+      "loss": 1.6261,
+      "step": 4910
+    },
+    {
+      "epoch": 2.2518059855521155,
+      "grad_norm": 0.3909819722175598,
+      "learning_rate": 0.000154458447375167,
+      "loss": 2.105,
+      "step": 4911
+    },
+    {
+      "epoch": 2.2522646485494784,
+      "grad_norm": 0.3079122304916382,
+      "learning_rate": 0.00015427949988669088,
+      "loss": 1.2053,
+      "step": 4912
+    },
+    {
+      "epoch": 2.252723311546841,
+      "grad_norm": 0.3336406946182251,
+      "learning_rate": 0.00015410063720594603,
+      "loss": 1.9375,
+      "step": 4913
+    },
+    {
+      "epoch": 2.2531819745442037,
+      "grad_norm": 0.38423246145248413,
+      "learning_rate": 0.00015392185937680898,
+      "loss": 1.4344,
+      "step": 4914
+    },
+    {
+      "epoch": 2.2536406375415665,
+      "grad_norm": 0.3887491822242737,
+      "learning_rate": 0.00015374316644313512,
+      "loss": 1.4943,
+      "step": 4915
+    },
+    {
+      "epoch": 2.254099300538929,
+      "grad_norm": 0.21756817400455475,
+      "learning_rate": 0.00015356455844875905,
+      "loss": 0.9165,
+      "step": 4916
+    },
+    {
+      "epoch": 2.254557963536292,
+      "grad_norm": 0.4763137102127075,
+      "learning_rate": 0.0001533860354374949,
+      "loss": 1.7384,
+      "step": 4917
+    },
+    {
+      "epoch": 2.2550166265336546,
+      "grad_norm": 0.3908590078353882,
+      "learning_rate": 0.00015320759745313562,
+      "loss": 1.2142,
+      "step": 4918
+    },
+    {
+      "epoch": 2.255475289531017,
+      "grad_norm": 0.38368988037109375,
+      "learning_rate": 0.000153029244539453,
+      "loss": 1.4749,
+      "step": 4919
+    },
+    {
+      "epoch": 2.25593395252838,
+      "grad_norm": 0.1392853558063507,
+      "learning_rate": 0.0001528509767401985,
+      "loss": 0.4752,
+      "step": 4920
+    },
+    {
+      "epoch": 2.2563926155257423,
+      "grad_norm": 0.2166876643896103,
+      "learning_rate": 0.00015267279409910252,
+      "loss": 1.1606,
+      "step": 4921
+    },
+    {
+      "epoch": 2.256851278523105,
+      "grad_norm": 0.2793666124343872,
+      "learning_rate": 0.0001524946966598744,
+      "loss": 1.2599,
+      "step": 4922
+    },
+    {
+      "epoch": 2.257309941520468,
+      "grad_norm": 0.2081069052219391,
+      "learning_rate": 0.0001523166844662031,
+      "loss": 1.1284,
+      "step": 4923
+    },
+    {
+      "epoch": 2.2577686045178305,
+      "grad_norm": 0.2551545202732086,
+      "learning_rate": 0.00015213875756175583,
+      "loss": 1.6474,
+      "step": 4924
+    },
+    {
+      "epoch": 2.2582272675151933,
+      "grad_norm": 0.412919819355011,
+      "learning_rate": 0.00015196091599017951,
+      "loss": 1.2211,
+      "step": 4925
+    },
+    {
+      "epoch": 2.2586859305125557,
+      "grad_norm": 0.5170382261276245,
+      "learning_rate": 0.00015178315979509988,
+      "loss": 1.3411,
+      "step": 4926
+    },
+    {
+      "epoch": 2.2591445935099186,
+      "grad_norm": 0.23865391314029694,
+      "learning_rate": 0.00015160548902012205,
+      "loss": 1.4869,
+      "step": 4927
+    },
+    {
+      "epoch": 2.2596032565072814,
+      "grad_norm": 0.31659218668937683,
+      "learning_rate": 0.00015142790370882987,
+      "loss": 0.9003,
+      "step": 4928
+    },
+    {
+      "epoch": 2.260061919504644,
+      "grad_norm": 0.09430176764726639,
+      "learning_rate": 0.00015125040390478634,
+      "loss": 0.4091,
+      "step": 4929
+    },
+    {
+      "epoch": 2.2605205825020067,
+      "grad_norm": 0.2541019320487976,
+      "learning_rate": 0.0001510729896515332,
+      "loss": 1.455,
+      "step": 4930
+    },
+    {
+      "epoch": 2.260979245499369,
+      "grad_norm": 0.25227102637290955,
+      "learning_rate": 0.00015089566099259162,
+      "loss": 1.4478,
+      "step": 4931
+    },
+    {
+      "epoch": 2.261437908496732,
+      "grad_norm": 0.37687572836875916,
+      "learning_rate": 0.00015071841797146152,
+      "loss": 1.6046,
+      "step": 4932
+    },
+    {
+      "epoch": 2.261896571494095,
+      "grad_norm": 0.25936609506607056,
+      "learning_rate": 0.000150541260631622,
+      "loss": 1.2579,
+      "step": 4933
+    },
+    {
+      "epoch": 2.2623552344914573,
+      "grad_norm": 0.3484478294849396,
+      "learning_rate": 0.0001503641890165311,
+      "loss": 1.8318,
+      "step": 4934
+    },
+    {
+      "epoch": 2.26281389748882,
+      "grad_norm": 0.33855557441711426,
+      "learning_rate": 0.00015018720316962536,
+      "loss": 1.6309,
+      "step": 4935
+    },
+    {
+      "epoch": 2.2632725604861825,
+      "grad_norm": 0.3302558958530426,
+      "learning_rate": 0.00015001030313432107,
+      "loss": 1.1351,
+      "step": 4936
+    },
+    {
+      "epoch": 2.2637312234835454,
+      "grad_norm": 0.18633894622325897,
+      "learning_rate": 0.000149833488954013,
+      "loss": 1.4874,
+      "step": 4937
+    },
+    {
+      "epoch": 2.2641898864809082,
+      "grad_norm": 0.3469991981983185,
+      "learning_rate": 0.00014965676067207496,
+      "loss": 1.4942,
+      "step": 4938
+    },
+    {
+      "epoch": 2.2646485494782707,
+      "grad_norm": 0.39552754163742065,
+      "learning_rate": 0.0001494801183318596,
+      "loss": 1.8694,
+      "step": 4939
+    },
+    {
+      "epoch": 2.2651072124756335,
+      "grad_norm": 0.26547738909721375,
+      "learning_rate": 0.0001493035619766987,
+      "loss": 0.8757,
+      "step": 4940
+    },
+    {
+      "epoch": 2.2655658754729964,
+      "grad_norm": 0.39476221799850464,
+      "learning_rate": 0.00014912709164990263,
+      "loss": 1.9812,
+      "step": 4941
+    },
+    {
+      "epoch": 2.266024538470359,
+      "grad_norm": 0.37706848978996277,
+      "learning_rate": 0.00014895070739476087,
+      "loss": 0.931,
+      "step": 4942
+    },
+    {
+      "epoch": 2.2664832014677216,
+      "grad_norm": 0.36269137263298035,
+      "learning_rate": 0.00014877440925454172,
+      "loss": 1.2654,
+      "step": 4943
+    },
+    {
+      "epoch": 2.2669418644650845,
+      "grad_norm": 0.23204423487186432,
+      "learning_rate": 0.0001485981972724925,
+      "loss": 1.6955,
+      "step": 4944
+    },
+    {
+      "epoch": 2.267400527462447,
+      "grad_norm": 0.4347366392612457,
+      "learning_rate": 0.00014842207149183922,
+      "loss": 1.5891,
+      "step": 4945
+    },
+    {
+      "epoch": 2.2678591904598098,
+      "grad_norm": 0.2783931493759155,
+      "learning_rate": 0.00014824603195578683,
+      "loss": 1.1357,
+      "step": 4946
+    },
+    {
+      "epoch": 2.268317853457172,
+      "grad_norm": 0.31447914242744446,
+      "learning_rate": 0.00014807007870751908,
+      "loss": 1.8933,
+      "step": 4947
+    },
+    {
+      "epoch": 2.268776516454535,
+      "grad_norm": 0.41312742233276367,
+      "learning_rate": 0.00014789421179019858,
+      "loss": 1.8218,
+      "step": 4948
+    },
+    {
+      "epoch": 2.269235179451898,
+      "grad_norm": 0.36053013801574707,
+      "learning_rate": 0.0001477184312469667,
+      "loss": 1.2607,
+      "step": 4949
+    },
+    {
+      "epoch": 2.2696938424492603,
+      "grad_norm": 0.28373968601226807,
+      "learning_rate": 0.00014754273712094373,
+      "loss": 1.4863,
+      "step": 4950
+    },
+    {
+      "epoch": 2.270152505446623,
+      "grad_norm": 0.3220798671245575,
+      "learning_rate": 0.00014736712945522884,
+      "loss": 1.7179,
+      "step": 4951
+    },
+    {
+      "epoch": 2.2706111684439856,
+      "grad_norm": 0.29745230078697205,
+      "learning_rate": 0.00014719160829289958,
+      "loss": 1.5435,
+      "step": 4952
+    },
+    {
+      "epoch": 2.2710698314413484,
+      "grad_norm": 0.32250505685806274,
+      "learning_rate": 0.0001470161736770127,
+      "loss": 1.0081,
+      "step": 4953
+    },
+    {
+      "epoch": 2.2715284944387113,
+      "grad_norm": 0.29016774892807007,
+      "learning_rate": 0.00014684082565060352,
+      "loss": 1.2257,
+      "step": 4954
+    },
+    {
+      "epoch": 2.2719871574360737,
+      "grad_norm": 0.3230065107345581,
+      "learning_rate": 0.00014666556425668625,
+      "loss": 1.8289,
+      "step": 4955
+    },
+    {
+      "epoch": 2.2724458204334366,
+      "grad_norm": 0.2680938243865967,
+      "learning_rate": 0.00014649038953825372,
+      "loss": 0.6338,
+      "step": 4956
+    },
+    {
+      "epoch": 2.272904483430799,
+      "grad_norm": 0.1309625804424286,
+      "learning_rate": 0.00014631530153827755,
+      "loss": 1.0811,
+      "step": 4957
+    },
+    {
+      "epoch": 2.273363146428162,
+      "grad_norm": 0.36883193254470825,
+      "learning_rate": 0.00014614030029970815,
+      "loss": 1.3616,
+      "step": 4958
+    },
+    {
+      "epoch": 2.2738218094255247,
+      "grad_norm": 0.2308487594127655,
+      "learning_rate": 0.00014596538586547454,
+      "loss": 1.6417,
+      "step": 4959
+    },
+    {
+      "epoch": 2.274280472422887,
+      "grad_norm": 0.38618066906929016,
+      "learning_rate": 0.00014579055827848448,
+      "loss": 1.0165,
+      "step": 4960
+    },
+    {
+      "epoch": 2.27473913542025,
+      "grad_norm": 0.22514605522155762,
+      "learning_rate": 0.0001456158175816245,
+      "loss": 1.1064,
+      "step": 4961
+    },
+    {
+      "epoch": 2.275197798417613,
+      "grad_norm": 0.34494972229003906,
+      "learning_rate": 0.00014544116381775985,
+      "loss": 1.1996,
+      "step": 4962
+    },
+    {
+      "epoch": 2.2756564614149752,
+      "grad_norm": 0.2711491882801056,
+      "learning_rate": 0.000145266597029734,
+      "loss": 1.4014,
+      "step": 4963
+    },
+    {
+      "epoch": 2.276115124412338,
+      "grad_norm": 0.2802002727985382,
+      "learning_rate": 0.00014509211726036975,
+      "loss": 1.1682,
+      "step": 4964
+    },
+    {
+      "epoch": 2.276573787409701,
+      "grad_norm": 0.2963082194328308,
+      "learning_rate": 0.0001449177245524681,
+      "loss": 1.387,
+      "step": 4965
+    },
+    {
+      "epoch": 2.2770324504070634,
+      "grad_norm": 0.33113130927085876,
+      "learning_rate": 0.00014474341894880888,
+      "loss": 1.2594,
+      "step": 4966
+    },
+    {
+      "epoch": 2.2774911134044262,
+      "grad_norm": 0.37912243604660034,
+      "learning_rate": 0.00014456920049215054,
+      "loss": 1.7935,
+      "step": 4967
+    },
+    {
+      "epoch": 2.2779497764017886,
+      "grad_norm": 0.3738153278827667,
+      "learning_rate": 0.00014439506922523016,
+      "loss": 1.233,
+      "step": 4968
+    },
+    {
+      "epoch": 2.2784084393991515,
+      "grad_norm": 0.280269980430603,
+      "learning_rate": 0.0001442210251907633,
+      "loss": 1.2351,
+      "step": 4969
+    },
+    {
+      "epoch": 2.2788671023965144,
+      "grad_norm": 0.3059045374393463,
+      "learning_rate": 0.00014404706843144423,
+      "loss": 1.5538,
+      "step": 4970
+    },
+    {
+      "epoch": 2.2793257653938768,
+      "grad_norm": 0.22319234907627106,
+      "learning_rate": 0.0001438731989899459,
+      "loss": 0.9915,
+      "step": 4971
+    },
+    {
+      "epoch": 2.2797844283912396,
+      "grad_norm": 0.2013273686170578,
+      "learning_rate": 0.00014369941690891959,
+      "loss": 0.5437,
+      "step": 4972
+    },
+    {
+      "epoch": 2.280243091388602,
+      "grad_norm": 0.33141711354255676,
+      "learning_rate": 0.00014352572223099542,
+      "loss": 1.1842,
+      "step": 4973
+    },
+    {
+      "epoch": 2.280701754385965,
+      "grad_norm": 0.24733178317546844,
+      "learning_rate": 0.00014335211499878203,
+      "loss": 1.5018,
+      "step": 4974
+    },
+    {
+      "epoch": 2.2811604173833278,
+      "grad_norm": 0.32721972465515137,
+      "learning_rate": 0.00014317859525486625,
+      "loss": 1.2133,
+      "step": 4975
+    },
+    {
+      "epoch": 2.28161908038069,
+      "grad_norm": 0.3684854209423065,
+      "learning_rate": 0.00014300516304181389,
+      "loss": 1.7316,
+      "step": 4976
+    },
+    {
+      "epoch": 2.282077743378053,
+      "grad_norm": 0.22372929751873016,
+      "learning_rate": 0.0001428318184021691,
+      "loss": 0.6982,
+      "step": 4977
+    },
+    {
+      "epoch": 2.2825364063754154,
+      "grad_norm": 0.31195321679115295,
+      "learning_rate": 0.00014265856137845434,
+      "loss": 1.6587,
+      "step": 4978
+    },
+    {
+      "epoch": 2.2829950693727783,
+      "grad_norm": 0.2723373472690582,
+      "learning_rate": 0.0001424853920131714,
+      "loss": 1.2561,
+      "step": 4979
+    },
+    {
+      "epoch": 2.283453732370141,
+      "grad_norm": 0.4274025559425354,
+      "learning_rate": 0.0001423123103487995,
+      "loss": 2.2205,
+      "step": 4980
+    },
+    {
+      "epoch": 2.2839123953675036,
+      "grad_norm": 0.3065417408943176,
+      "learning_rate": 0.00014213931642779686,
+      "loss": 0.9494,
+      "step": 4981
+    },
+    {
+      "epoch": 2.2843710583648664,
+      "grad_norm": 0.21199515461921692,
+      "learning_rate": 0.00014196641029260026,
+      "loss": 0.624,
+      "step": 4982
+    },
+    {
+      "epoch": 2.2848297213622293,
+      "grad_norm": 0.23199045658111572,
+      "learning_rate": 0.00014179359198562475,
+      "loss": 1.1652,
+      "step": 4983
+    },
+    {
+      "epoch": 2.2852883843595917,
+      "grad_norm": 0.29177722334861755,
+      "learning_rate": 0.00014162086154926397,
+      "loss": 1.4468,
+      "step": 4984
+    },
+    {
+      "epoch": 2.2857470473569546,
+      "grad_norm": 0.37286636233329773,
+      "learning_rate": 0.00014144821902589,
+      "loss": 1.9711,
+      "step": 4985
+    },
+    {
+      "epoch": 2.2862057103543174,
+      "grad_norm": 0.4224682152271271,
+      "learning_rate": 0.00014127566445785306,
+      "loss": 1.3417,
+      "step": 4986
+    },
+    {
+      "epoch": 2.28666437335168,
+      "grad_norm": 0.3329372704029083,
+      "learning_rate": 0.00014110319788748215,
+      "loss": 1.4227,
+      "step": 4987
+    },
+    {
+      "epoch": 2.2871230363490427,
+      "grad_norm": 0.30778220295906067,
+      "learning_rate": 0.00014093081935708445,
+      "loss": 1.3458,
+      "step": 4988
+    },
+    {
+      "epoch": 2.287581699346405,
+      "grad_norm": 0.34832048416137695,
+      "learning_rate": 0.0001407585289089459,
+      "loss": 1.2986,
+      "step": 4989
+    },
+    {
+      "epoch": 2.288040362343768,
+      "grad_norm": 0.32428908348083496,
+      "learning_rate": 0.00014058632658533072,
+      "loss": 1.9873,
+      "step": 4990
+    },
+    {
+      "epoch": 2.288499025341131,
+      "grad_norm": 0.39319339394569397,
+      "learning_rate": 0.0001404142124284809,
+      "loss": 1.6913,
+      "step": 4991
+    },
+    {
+      "epoch": 2.2889576883384932,
+      "grad_norm": 0.3040050268173218,
+      "learning_rate": 0.00014024218648061755,
+      "loss": 1.1154,
+      "step": 4992
+    },
+    {
+      "epoch": 2.289416351335856,
+      "grad_norm": 0.4540446102619171,
+      "learning_rate": 0.00014007024878393983,
+      "loss": 1.297,
+      "step": 4993
+    },
+    {
+      "epoch": 2.2898750143332185,
+      "grad_norm": 0.29186972975730896,
+      "learning_rate": 0.0001398983993806253,
+      "loss": 1.6358,
+      "step": 4994
+    },
+    {
+      "epoch": 2.2903336773305814,
+      "grad_norm": 0.25946202874183655,
+      "learning_rate": 0.0001397266383128299,
+      "loss": 1.522,
+      "step": 4995
+    },
+    {
+      "epoch": 2.290792340327944,
+      "grad_norm": 0.27808359265327454,
+      "learning_rate": 0.00013955496562268794,
+      "loss": 1.1537,
+      "step": 4996
+    },
+    {
+      "epoch": 2.2912510033253066,
+      "grad_norm": 0.34630027413368225,
+      "learning_rate": 0.0001393833813523117,
+      "loss": 1.5734,
+      "step": 4997
+    },
+    {
+      "epoch": 2.2917096663226695,
+      "grad_norm": 0.2861468195915222,
+      "learning_rate": 0.0001392118855437921,
+      "loss": 0.5451,
+      "step": 4998
+    },
+    {
+      "epoch": 2.292168329320032,
+      "grad_norm": 0.3561016917228699,
+      "learning_rate": 0.0001390404782391985,
+      "loss": 1.2978,
+      "step": 4999
+    },
+    {
+      "epoch": 2.2926269923173948,
+      "grad_norm": 0.2726607024669647,
+      "learning_rate": 0.00013886915948057825,
+      "loss": 1.256,
+      "step": 5000
+    },
+    {
+      "epoch": 2.2930856553147576,
+      "grad_norm": 0.41800275444984436,
+      "learning_rate": 0.00013869792930995724,
+      "loss": 1.5729,
+      "step": 5001
+    },
+    {
+      "epoch": 2.29354431831212,
+      "grad_norm": 0.3518107533454895,
+      "learning_rate": 0.00013852678776933914,
+      "loss": 1.1977,
+      "step": 5002
+    },
+    {
+      "epoch": 2.294002981309483,
+      "grad_norm": 0.14994703233242035,
+      "learning_rate": 0.0001383557349007063,
+      "loss": 0.6909,
+      "step": 5003
+    },
+    {
+      "epoch": 2.2944616443068457,
+      "grad_norm": 0.2874891459941864,
+      "learning_rate": 0.00013818477074601933,
+      "loss": 2.2263,
+      "step": 5004
+    },
+    {
+      "epoch": 2.294920307304208,
+      "grad_norm": 0.41052594780921936,
+      "learning_rate": 0.00013801389534721692,
+      "loss": 1.6545,
+      "step": 5005
+    },
+    {
+      "epoch": 2.295378970301571,
+      "grad_norm": 0.6844239830970764,
+      "learning_rate": 0.00013784310874621604,
+      "loss": 1.9092,
+      "step": 5006
+    },
+    {
+      "epoch": 2.2958376332989334,
+      "grad_norm": 0.21883147954940796,
+      "learning_rate": 0.00013767241098491185,
+      "loss": 0.4407,
+      "step": 5007
+    },
+    {
+      "epoch": 2.2962962962962963,
+      "grad_norm": 0.06891786307096481,
+      "learning_rate": 0.00013750180210517777,
+      "loss": 0.7355,
+      "step": 5008
+    },
+    {
+      "epoch": 2.296754959293659,
+      "grad_norm": 0.2810133099555969,
+      "learning_rate": 0.00013733128214886536,
+      "loss": 0.9875,
+      "step": 5009
+    },
+    {
+      "epoch": 2.2972136222910216,
+      "grad_norm": 0.28389817476272583,
+      "learning_rate": 0.00013716085115780447,
+      "loss": 1.6359,
+      "step": 5010
+    },
+    {
+      "epoch": 2.2976722852883844,
+      "grad_norm": 0.43318691849708557,
+      "learning_rate": 0.00013699050917380295,
+      "loss": 1.6229,
+      "step": 5011
+    },
+    {
+      "epoch": 2.2981309482857473,
+      "grad_norm": 0.36424124240875244,
+      "learning_rate": 0.00013682025623864698,
+      "loss": 1.8982,
+      "step": 5012
+    },
+    {
+      "epoch": 2.2985896112831097,
+      "grad_norm": 0.3689767122268677,
+      "learning_rate": 0.00013665009239410098,
+      "loss": 1.2045,
+      "step": 5013
+    },
+    {
+      "epoch": 2.2990482742804725,
+      "grad_norm": 0.28414589166641235,
+      "learning_rate": 0.00013648001768190699,
+      "loss": 1.5854,
+      "step": 5014
+    },
+    {
+      "epoch": 2.299506937277835,
+      "grad_norm": 0.3238101601600647,
+      "learning_rate": 0.00013631003214378584,
+      "loss": 0.8048,
+      "step": 5015
+    },
+    {
+      "epoch": 2.299965600275198,
+      "grad_norm": 0.25493547320365906,
+      "learning_rate": 0.00013614013582143614,
+      "loss": 1.5388,
+      "step": 5016
+    },
+    {
+      "epoch": 2.3004242632725607,
+      "grad_norm": 0.3472574055194855,
+      "learning_rate": 0.0001359703287565347,
+      "loss": 1.3253,
+      "step": 5017
+    },
+    {
+      "epoch": 2.300882926269923,
+      "grad_norm": 0.21940115094184875,
+      "learning_rate": 0.00013580061099073638,
+      "loss": 1.473,
+      "step": 5018
+    },
+    {
+      "epoch": 2.301341589267286,
+      "grad_norm": 0.33092114329338074,
+      "learning_rate": 0.0001356309825656742,
+      "loss": 1.5323,
+      "step": 5019
+    },
+    {
+      "epoch": 2.3018002522646483,
+      "grad_norm": 0.3832414150238037,
+      "learning_rate": 0.0001354614435229592,
+      "loss": 1.3107,
+      "step": 5020
+    },
+    {
+      "epoch": 2.302258915262011,
+      "grad_norm": 0.27697277069091797,
+      "learning_rate": 0.0001352919939041806,
+      "loss": 1.5306,
+      "step": 5021
+    },
+    {
+      "epoch": 2.302717578259374,
+      "grad_norm": 0.30245643854141235,
+      "learning_rate": 0.00013512263375090562,
+      "loss": 1.4718,
+      "step": 5022
+    },
+    {
+      "epoch": 2.3031762412567365,
+      "grad_norm": 0.3190750777721405,
+      "learning_rate": 0.00013495336310467943,
+      "loss": 1.4829,
+      "step": 5023
+    },
+    {
+      "epoch": 2.3036349042540993,
+      "grad_norm": 0.2573975622653961,
+      "learning_rate": 0.00013478418200702552,
+      "loss": 0.5235,
+      "step": 5024
+    },
+    {
+      "epoch": 2.3040935672514617,
+      "grad_norm": 0.13736076653003693,
+      "learning_rate": 0.00013461509049944497,
+      "loss": 0.5812,
+      "step": 5025
+    },
+    {
+      "epoch": 2.3045522302488246,
+      "grad_norm": 0.19605790078639984,
+      "learning_rate": 0.00013444608862341734,
+      "loss": 1.2568,
+      "step": 5026
+    },
+    {
+      "epoch": 2.3050108932461875,
+      "grad_norm": 0.25617191195487976,
+      "learning_rate": 0.00013427717642039988,
+      "loss": 1.5799,
+      "step": 5027
+    },
+    {
+      "epoch": 2.30546955624355,
+      "grad_norm": 0.3717952370643616,
+      "learning_rate": 0.00013410835393182807,
+      "loss": 1.3069,
+      "step": 5028
+    },
+    {
+      "epoch": 2.3059282192409127,
+      "grad_norm": 0.37081441283226013,
+      "learning_rate": 0.00013393962119911528,
+      "loss": 1.7627,
+      "step": 5029
+    },
+    {
+      "epoch": 2.3063868822382756,
+      "grad_norm": 0.3451120853424072,
+      "learning_rate": 0.0001337709782636528,
+      "loss": 1.7584,
+      "step": 5030
+    },
+    {
+      "epoch": 2.306845545235638,
+      "grad_norm": 0.326267272233963,
+      "learning_rate": 0.00013360242516681004,
+      "loss": 1.0427,
+      "step": 5031
+    },
+    {
+      "epoch": 2.307304208233001,
+      "grad_norm": 0.2787007987499237,
+      "learning_rate": 0.00013343396194993423,
+      "loss": 1.3569,
+      "step": 5032
+    },
+    {
+      "epoch": 2.3077628712303637,
+      "grad_norm": 0.2687190771102905,
+      "learning_rate": 0.0001332655886543506,
+      "loss": 1.8645,
+      "step": 5033
+    },
+    {
+      "epoch": 2.308221534227726,
+      "grad_norm": 0.3988434970378876,
+      "learning_rate": 0.00013309730532136245,
+      "loss": 2.0012,
+      "step": 5034
+    },
+    {
+      "epoch": 2.308680197225089,
+      "grad_norm": 0.1921907663345337,
+      "learning_rate": 0.0001329291119922509,
+      "loss": 1.2577,
+      "step": 5035
+    },
+    {
+      "epoch": 2.3091388602224514,
+      "grad_norm": 0.4280322790145874,
+      "learning_rate": 0.00013276100870827473,
+      "loss": 1.693,
+      "step": 5036
+    },
+    {
+      "epoch": 2.3095975232198143,
+      "grad_norm": 0.34726113080978394,
+      "learning_rate": 0.00013259299551067106,
+      "loss": 0.8334,
+      "step": 5037
+    },
+    {
+      "epoch": 2.310056186217177,
+      "grad_norm": 0.1919124275445938,
+      "learning_rate": 0.00013242507244065477,
+      "loss": 1.77,
+      "step": 5038
+    },
+    {
+      "epoch": 2.3105148492145395,
+      "grad_norm": 0.3817395865917206,
+      "learning_rate": 0.00013225723953941854,
+      "loss": 1.9272,
+      "step": 5039
+    },
+    {
+      "epoch": 2.3109735122119024,
+      "grad_norm": 0.41375109553337097,
+      "learning_rate": 0.00013208949684813287,
+      "loss": 1.5177,
+      "step": 5040
+    },
+    {
+      "epoch": 2.311432175209265,
+      "grad_norm": 0.29578983783721924,
+      "learning_rate": 0.00013192184440794668,
+      "loss": 1.3288,
+      "step": 5041
+    },
+    {
+      "epoch": 2.3118908382066277,
+      "grad_norm": 0.19934503734111786,
+      "learning_rate": 0.00013175428225998593,
+      "loss": 0.9299,
+      "step": 5042
+    },
+    {
+      "epoch": 2.3123495012039905,
+      "grad_norm": 0.2743481397628784,
+      "learning_rate": 0.00013158681044535487,
+      "loss": 1.5061,
+      "step": 5043
+    },
+    {
+      "epoch": 2.312808164201353,
+      "grad_norm": 0.44560080766677856,
+      "learning_rate": 0.00013141942900513564,
+      "loss": 0.9513,
+      "step": 5044
+    },
+    {
+      "epoch": 2.313266827198716,
+      "grad_norm": 0.21826134622097015,
+      "learning_rate": 0.0001312521379803881,
+      "loss": 0.941,
+      "step": 5045
+    },
+    {
+      "epoch": 2.313725490196078,
+      "grad_norm": 0.16389012336730957,
+      "learning_rate": 0.00013108493741215,
+      "loss": 1.3627,
+      "step": 5046
+    },
+    {
+      "epoch": 2.314184153193441,
+      "grad_norm": 0.3645002543926239,
+      "learning_rate": 0.00013091782734143671,
+      "loss": 1.5707,
+      "step": 5047
+    },
+    {
+      "epoch": 2.314642816190804,
+      "grad_norm": 0.3624444901943207,
+      "learning_rate": 0.00013075080780924154,
+      "loss": 2.1348,
+      "step": 5048
+    },
+    {
+      "epoch": 2.3151014791881663,
+      "grad_norm": 0.3850797712802887,
+      "learning_rate": 0.00013058387885653562,
+      "loss": 1.4394,
+      "step": 5049
+    },
+    {
+      "epoch": 2.315560142185529,
+      "grad_norm": 0.281230092048645,
+      "learning_rate": 0.00013041704052426772,
+      "loss": 1.6204,
+      "step": 5050
+    },
+    {
+      "epoch": 2.316018805182892,
+      "grad_norm": 0.2876693308353424,
+      "learning_rate": 0.00013025029285336476,
+      "loss": 1.1374,
+      "step": 5051
+    },
+    {
+      "epoch": 2.3164774681802545,
+      "grad_norm": 0.3920785188674927,
+      "learning_rate": 0.00013008363588473115,
+      "loss": 1.639,
+      "step": 5052
+    },
+    {
+      "epoch": 2.3169361311776173,
+      "grad_norm": 0.2921794652938843,
+      "learning_rate": 0.00012991706965924876,
+      "loss": 1.5347,
+      "step": 5053
+    },
+    {
+      "epoch": 2.31739479417498,
+      "grad_norm": 0.26020175218582153,
+      "learning_rate": 0.00012975059421777759,
+      "loss": 1.1937,
+      "step": 5054
+    },
+    {
+      "epoch": 2.3178534571723426,
+      "grad_norm": 0.36081263422966003,
+      "learning_rate": 0.0001295842096011553,
+      "loss": 1.3285,
+      "step": 5055
+    },
+    {
+      "epoch": 2.3183121201697054,
+      "grad_norm": 0.48956963419914246,
+      "learning_rate": 0.00012941791585019725,
+      "loss": 1.4641,
+      "step": 5056
+    },
+    {
+      "epoch": 2.318770783167068,
+      "grad_norm": 0.1845049411058426,
+      "learning_rate": 0.0001292517130056966,
+      "loss": 1.1712,
+      "step": 5057
+    },
+    {
+      "epoch": 2.3192294461644307,
+      "grad_norm": 0.35714197158813477,
+      "learning_rate": 0.00012908560110842383,
+      "loss": 1.4028,
+      "step": 5058
+    },
+    {
+      "epoch": 2.3196881091617936,
+      "grad_norm": 0.2765589654445648,
+      "learning_rate": 0.00012891958019912758,
+      "loss": 0.852,
+      "step": 5059
+    },
+    {
+      "epoch": 2.320146772159156,
+      "grad_norm": 0.32246285676956177,
+      "learning_rate": 0.00012875365031853376,
+      "loss": 1.725,
+      "step": 5060
+    },
+    {
+      "epoch": 2.320605435156519,
+      "grad_norm": 0.3229038715362549,
+      "learning_rate": 0.0001285878115073465,
+      "loss": 1.4902,
+      "step": 5061
+    },
+    {
+      "epoch": 2.3210640981538813,
+      "grad_norm": 0.24635767936706543,
+      "learning_rate": 0.0001284220638062471,
+      "loss": 0.9351,
+      "step": 5062
+    },
+    {
+      "epoch": 2.321522761151244,
+      "grad_norm": 0.2740919589996338,
+      "learning_rate": 0.00012825640725589477,
+      "loss": 1.7063,
+      "step": 5063
+    },
+    {
+      "epoch": 2.321981424148607,
+      "grad_norm": 0.3636001646518707,
+      "learning_rate": 0.00012809084189692604,
+      "loss": 1.3257,
+      "step": 5064
+    },
+    {
+      "epoch": 2.3224400871459694,
+      "grad_norm": 0.27342599630355835,
+      "learning_rate": 0.0001279253677699554,
+      "loss": 1.2865,
+      "step": 5065
+    },
+    {
+      "epoch": 2.3228987501433322,
+      "grad_norm": 0.3146760165691376,
+      "learning_rate": 0.00012775998491557485,
+      "loss": 1.28,
+      "step": 5066
+    },
+    {
+      "epoch": 2.3233574131406947,
+      "grad_norm": 0.22414685785770416,
+      "learning_rate": 0.00012759469337435397,
+      "loss": 1.2673,
+      "step": 5067
+    },
+    {
+      "epoch": 2.3238160761380575,
+      "grad_norm": 0.42230096459388733,
+      "learning_rate": 0.00012742949318684,
+      "loss": 1.0508,
+      "step": 5068
+    },
+    {
+      "epoch": 2.3242747391354204,
+      "grad_norm": 0.38766834139823914,
+      "learning_rate": 0.00012726438439355787,
+      "loss": 1.6228,
+      "step": 5069
+    },
+    {
+      "epoch": 2.324733402132783,
+      "grad_norm": 0.37745577096939087,
+      "learning_rate": 0.00012709936703500947,
+      "loss": 1.8116,
+      "step": 5070
+    },
+    {
+      "epoch": 2.3251920651301456,
+      "grad_norm": 0.4257441759109497,
+      "learning_rate": 0.0001269344411516753,
+      "loss": 1.1618,
+      "step": 5071
+    },
+    {
+      "epoch": 2.3256507281275085,
+      "grad_norm": 0.3758509159088135,
+      "learning_rate": 0.00012676960678401262,
+      "loss": 1.6102,
+      "step": 5072
+    },
+    {
+      "epoch": 2.326109391124871,
+      "grad_norm": 0.27729111909866333,
+      "learning_rate": 0.0001266048639724565,
+      "loss": 0.6324,
+      "step": 5073
+    },
+    {
+      "epoch": 2.3265680541222338,
+      "grad_norm": 0.27736055850982666,
+      "learning_rate": 0.0001264402127574198,
+      "loss": 1.5171,
+      "step": 5074
+    },
+    {
+      "epoch": 2.327026717119596,
+      "grad_norm": 0.2820827066898346,
+      "learning_rate": 0.0001262756531792922,
+      "loss": 1.6448,
+      "step": 5075
+    },
+    {
+      "epoch": 2.327485380116959,
+      "grad_norm": 0.31878018379211426,
+      "learning_rate": 0.0001261111852784416,
+      "loss": 0.9581,
+      "step": 5076
+    },
+    {
+      "epoch": 2.327944043114322,
+      "grad_norm": 0.32956355810165405,
+      "learning_rate": 0.0001259468090952131,
+      "loss": 1.3363,
+      "step": 5077
+    },
+    {
+      "epoch": 2.3284027061116843,
+      "grad_norm": 0.33068135380744934,
+      "learning_rate": 0.0001257825246699294,
+      "loss": 0.8995,
+      "step": 5078
+    },
+    {
+      "epoch": 2.328861369109047,
+      "grad_norm": 0.2901584506034851,
+      "learning_rate": 0.0001256183320428907,
+      "loss": 1.6243,
+      "step": 5079
+    },
+    {
+      "epoch": 2.32932003210641,
+      "grad_norm": 0.2884989380836487,
+      "learning_rate": 0.0001254542312543745,
+      "loss": 1.683,
+      "step": 5080
+    },
+    {
+      "epoch": 2.3297786951037724,
+      "grad_norm": 0.38220369815826416,
+      "learning_rate": 0.00012529022234463604,
+      "loss": 1.7488,
+      "step": 5081
+    },
+    {
+      "epoch": 2.3302373581011353,
+      "grad_norm": 0.3322943150997162,
+      "learning_rate": 0.00012512630535390783,
+      "loss": 1.9446,
+      "step": 5082
+    },
+    {
+      "epoch": 2.3306960210984977,
+      "grad_norm": 0.3463505804538727,
+      "learning_rate": 0.00012496248032239988,
+      "loss": 1.0688,
+      "step": 5083
+    },
+    {
+      "epoch": 2.3311546840958606,
+      "grad_norm": 0.240933358669281,
+      "learning_rate": 0.00012479874729029968,
+      "loss": 1.6491,
+      "step": 5084
+    },
+    {
+      "epoch": 2.3316133470932234,
+      "grad_norm": 0.4334370195865631,
+      "learning_rate": 0.00012463510629777226,
+      "loss": 1.3899,
+      "step": 5085
+    },
+    {
+      "epoch": 2.332072010090586,
+      "grad_norm": 0.2391756922006607,
+      "learning_rate": 0.00012447155738495963,
+      "loss": 0.608,
+      "step": 5086
+    },
+    {
+      "epoch": 2.3325306730879487,
+      "grad_norm": 0.2066376507282257,
+      "learning_rate": 0.00012430810059198166,
+      "loss": 1.1892,
+      "step": 5087
+    },
+    {
+      "epoch": 2.332989336085311,
+      "grad_norm": 0.33126476407051086,
+      "learning_rate": 0.0001241447359589355,
+      "loss": 1.22,
+      "step": 5088
+    },
+    {
+      "epoch": 2.333447999082674,
+      "grad_norm": 0.29312899708747864,
+      "learning_rate": 0.00012398146352589568,
+      "loss": 1.2554,
+      "step": 5089
+    },
+    {
+      "epoch": 2.333906662080037,
+      "grad_norm": 0.24564377963542938,
+      "learning_rate": 0.0001238182833329141,
+      "loss": 1.3128,
+      "step": 5090
+    },
+    {
+      "epoch": 2.3343653250773992,
+      "grad_norm": 0.3976064920425415,
+      "learning_rate": 0.00012365519542002,
+      "loss": 1.4039,
+      "step": 5091
+    },
+    {
+      "epoch": 2.334823988074762,
+      "grad_norm": 0.31696027517318726,
+      "learning_rate": 0.0001234921998272201,
+      "loss": 2.1157,
+      "step": 5092
+    },
+    {
+      "epoch": 2.3352826510721245,
+      "grad_norm": 0.3862619400024414,
+      "learning_rate": 0.00012332929659449827,
+      "loss": 1.1534,
+      "step": 5093
+    },
+    {
+      "epoch": 2.3357413140694874,
+      "grad_norm": 0.2641201317310333,
+      "learning_rate": 0.000123166485761816,
+      "loss": 0.9848,
+      "step": 5094
+    },
+    {
+      "epoch": 2.3361999770668502,
+      "grad_norm": 0.3019377291202545,
+      "learning_rate": 0.00012300376736911183,
+      "loss": 1.233,
+      "step": 5095
+    },
+    {
+      "epoch": 2.3366586400642126,
+      "grad_norm": 0.280788779258728,
+      "learning_rate": 0.00012284114145630183,
+      "loss": 1.6123,
+      "step": 5096
+    },
+    {
+      "epoch": 2.3371173030615755,
+      "grad_norm": 0.3243066370487213,
+      "learning_rate": 0.0001226786080632794,
+      "loss": 0.5487,
+      "step": 5097
+    },
+    {
+      "epoch": 2.3375759660589384,
+      "grad_norm": 0.18734464049339294,
+      "learning_rate": 0.00012251616722991492,
+      "loss": 0.9184,
+      "step": 5098
+    },
+    {
+      "epoch": 2.3380346290563008,
+      "grad_norm": 0.17624834179878235,
+      "learning_rate": 0.0001223538189960564,
+      "loss": 1.5011,
+      "step": 5099
+    },
+    {
+      "epoch": 2.3384932920536636,
+      "grad_norm": 0.2665694057941437,
+      "learning_rate": 0.00012219156340152908,
+      "loss": 0.8928,
+      "step": 5100
+    },
+    {
+      "epoch": 2.3389519550510265,
+      "grad_norm": 0.2912820279598236,
+      "learning_rate": 0.00012202940048613542,
+      "loss": 1.5328,
+      "step": 5101
+    },
+    {
+      "epoch": 2.339410618048389,
+      "grad_norm": 0.3863449692726135,
+      "learning_rate": 0.00012186733028965507,
+      "loss": 1.4951,
+      "step": 5102
+    },
+    {
+      "epoch": 2.3398692810457518,
+      "grad_norm": 0.40611687302589417,
+      "learning_rate": 0.00012170535285184509,
+      "loss": 2.1716,
+      "step": 5103
+    },
+    {
+      "epoch": 2.340327944043114,
+      "grad_norm": 0.4273900091648102,
+      "learning_rate": 0.00012154346821243972,
+      "loss": 1.21,
+      "step": 5104
+    },
+    {
+      "epoch": 2.340786607040477,
+      "grad_norm": 0.3556171953678131,
+      "learning_rate": 0.0001213816764111504,
+      "loss": 1.3158,
+      "step": 5105
+    },
+    {
+      "epoch": 2.34124527003784,
+      "grad_norm": 0.28761526942253113,
+      "learning_rate": 0.00012121997748766583,
+      "loss": 1.6339,
+      "step": 5106
+    },
+    {
+      "epoch": 2.3417039330352023,
+      "grad_norm": 0.2865532338619232,
+      "learning_rate": 0.00012105837148165194,
+      "loss": 1.3912,
+      "step": 5107
+    },
+    {
+      "epoch": 2.342162596032565,
+      "grad_norm": 0.30358991026878357,
+      "learning_rate": 0.000120896858432752,
+      "loss": 1.1263,
+      "step": 5108
+    },
+    {
+      "epoch": 2.3426212590299276,
+      "grad_norm": 0.07187612354755402,
+      "learning_rate": 0.00012073543838058598,
+      "loss": 1.1824,
+      "step": 5109
+    },
+    {
+      "epoch": 2.3430799220272904,
+      "grad_norm": 0.2775716483592987,
+      "learning_rate": 0.00012057411136475161,
+      "loss": 1.1338,
+      "step": 5110
+    },
+    {
+      "epoch": 2.3435385850246533,
+      "grad_norm": 0.370094358921051,
+      "learning_rate": 0.00012041287742482348,
+      "loss": 0.8284,
+      "step": 5111
+    },
+    {
+      "epoch": 2.3439972480220157,
+      "grad_norm": 0.5262327790260315,
+      "learning_rate": 0.00012025173660035338,
+      "loss": 1.6775,
+      "step": 5112
+    },
+    {
+      "epoch": 2.3444559110193786,
+      "grad_norm": 0.36328423023223877,
+      "learning_rate": 0.00012009068893087067,
+      "loss": 1.453,
+      "step": 5113
+    },
+    {
+      "epoch": 2.344914574016741,
+      "grad_norm": 1.2442697286605835,
+      "learning_rate": 0.00011992973445588112,
+      "loss": 2.0186,
+      "step": 5114
+    },
+    {
+      "epoch": 2.345373237014104,
+      "grad_norm": 0.26644742488861084,
+      "learning_rate": 0.00011976887321486813,
+      "loss": 0.9504,
+      "step": 5115
+    },
+    {
+      "epoch": 2.3458319000114667,
+      "grad_norm": 0.31766730546951294,
+      "learning_rate": 0.00011960810524729222,
+      "loss": 1.277,
+      "step": 5116
+    },
+    {
+      "epoch": 2.346290563008829,
+      "grad_norm": 0.3006284236907959,
+      "learning_rate": 0.00011944743059259078,
+      "loss": 0.7456,
+      "step": 5117
+    },
+    {
+      "epoch": 2.346749226006192,
+      "grad_norm": 0.2082628756761551,
+      "learning_rate": 0.00011928684929017859,
+      "loss": 0.9155,
+      "step": 5118
+    },
+    {
+      "epoch": 2.347207889003555,
+      "grad_norm": 0.2231052815914154,
+      "learning_rate": 0.0001191263613794475,
+      "loss": 0.882,
+      "step": 5119
+    },
+    {
+      "epoch": 2.3476665520009172,
+      "grad_norm": 0.2642306983470917,
+      "learning_rate": 0.00011896596689976602,
+      "loss": 1.3383,
+      "step": 5120
+    },
+    {
+      "epoch": 2.34812521499828,
+      "grad_norm": 0.26055455207824707,
+      "learning_rate": 0.00011880566589048031,
+      "loss": 1.2426,
+      "step": 5121
+    },
+    {
+      "epoch": 2.348583877995643,
+      "grad_norm": 0.20619313418865204,
+      "learning_rate": 0.00011864545839091312,
+      "loss": 0.7382,
+      "step": 5122
+    },
+    {
+      "epoch": 2.3490425409930054,
+      "grad_norm": 0.19531986117362976,
+      "learning_rate": 0.00011848534444036485,
+      "loss": 1.091,
+      "step": 5123
+    },
+    {
+      "epoch": 2.349501203990368,
+      "grad_norm": 0.3989899158477783,
+      "learning_rate": 0.00011832532407811247,
+      "loss": 0.8412,
+      "step": 5124
+    },
+    {
+      "epoch": 2.3499598669877306,
+      "grad_norm": 0.2260282337665558,
+      "learning_rate": 0.00011816539734341025,
+      "loss": 1.4639,
+      "step": 5125
+    },
+    {
+      "epoch": 2.3504185299850935,
+      "grad_norm": 0.26231110095977783,
+      "learning_rate": 0.00011800556427548908,
+      "loss": 1.1874,
+      "step": 5126
+    },
+    {
+      "epoch": 2.3508771929824563,
+      "grad_norm": 0.19405876100063324,
+      "learning_rate": 0.00011784582491355727,
+      "loss": 0.5713,
+      "step": 5127
+    },
+    {
+      "epoch": 2.3513358559798188,
+      "grad_norm": 0.2558598220348358,
+      "learning_rate": 0.00011768617929680014,
+      "loss": 1.4579,
+      "step": 5128
+    },
+    {
+      "epoch": 2.3517945189771816,
+      "grad_norm": 0.2595175504684448,
+      "learning_rate": 0.00011752662746437986,
+      "loss": 1.4692,
+      "step": 5129
+    },
+    {
+      "epoch": 2.352253181974544,
+      "grad_norm": 0.4527866244316101,
+      "learning_rate": 0.0001173671694554358,
+      "loss": 1.7926,
+      "step": 5130
+    },
+    {
+      "epoch": 2.352711844971907,
+      "grad_norm": 0.24965371191501617,
+      "learning_rate": 0.00011720780530908381,
+      "loss": 0.972,
+      "step": 5131
+    },
+    {
+      "epoch": 2.3531705079692697,
+      "grad_norm": 0.21504342555999756,
+      "learning_rate": 0.0001170485350644171,
+      "loss": 0.7318,
+      "step": 5132
+    },
+    {
+      "epoch": 2.353629170966632,
+      "grad_norm": 0.46009358763694763,
+      "learning_rate": 0.00011688935876050616,
+      "loss": 1.8917,
+      "step": 5133
+    },
+    {
+      "epoch": 2.354087833963995,
+      "grad_norm": 0.35450685024261475,
+      "learning_rate": 0.00011673027643639784,
+      "loss": 0.7528,
+      "step": 5134
+    },
+    {
+      "epoch": 2.3545464969613574,
+      "grad_norm": 0.10655711591243744,
+      "learning_rate": 0.00011657128813111622,
+      "loss": 0.8409,
+      "step": 5135
+    },
+    {
+      "epoch": 2.3550051599587203,
+      "grad_norm": 0.40161076188087463,
+      "learning_rate": 0.00011641239388366249,
+      "loss": 1.6137,
+      "step": 5136
+    },
+    {
+      "epoch": 2.355463822956083,
+      "grad_norm": 0.34276244044303894,
+      "learning_rate": 0.00011625359373301414,
+      "loss": 0.9094,
+      "step": 5137
+    },
+    {
+      "epoch": 2.3559224859534456,
+      "grad_norm": 0.3367175757884979,
+      "learning_rate": 0.00011609488771812621,
+      "loss": 1.2706,
+      "step": 5138
+    },
+    {
+      "epoch": 2.3563811489508084,
+      "grad_norm": 0.27182361483573914,
+      "learning_rate": 0.00011593627587793043,
+      "loss": 1.489,
+      "step": 5139
+    },
+    {
+      "epoch": 2.3568398119481713,
+      "grad_norm": 0.3773139715194702,
+      "learning_rate": 0.00011577775825133546,
+      "loss": 1.0298,
+      "step": 5140
+    },
+    {
+      "epoch": 2.3572984749455337,
+      "grad_norm": 0.11614841967821121,
+      "learning_rate": 0.00011561933487722687,
+      "loss": 0.7662,
+      "step": 5141
+    },
+    {
+      "epoch": 2.3577571379428965,
+      "grad_norm": 0.24605746567249298,
+      "learning_rate": 0.00011546100579446672,
+      "loss": 0.8525,
+      "step": 5142
+    },
+    {
+      "epoch": 2.358215800940259,
+      "grad_norm": 0.4239916503429413,
+      "learning_rate": 0.00011530277104189463,
+      "loss": 1.5721,
+      "step": 5143
+    },
+    {
+      "epoch": 2.358674463937622,
+      "grad_norm": 0.2336055487394333,
+      "learning_rate": 0.00011514463065832665,
+      "loss": 1.2539,
+      "step": 5144
+    },
+    {
+      "epoch": 2.3591331269349847,
+      "grad_norm": 0.333873987197876,
+      "learning_rate": 0.00011498658468255568,
+      "loss": 1.3782,
+      "step": 5145
+    },
+    {
+      "epoch": 2.359591789932347,
+      "grad_norm": 0.35513144731521606,
+      "learning_rate": 0.00011482863315335157,
+      "loss": 1.3682,
+      "step": 5146
+    },
+    {
+      "epoch": 2.36005045292971,
+      "grad_norm": 0.2521505057811737,
+      "learning_rate": 0.00011467077610946113,
+      "loss": 1.2587,
+      "step": 5147
+    },
+    {
+      "epoch": 2.360509115927073,
+      "grad_norm": 0.30736634135246277,
+      "learning_rate": 0.00011451301358960758,
+      "loss": 1.3173,
+      "step": 5148
+    },
+    {
+      "epoch": 2.360967778924435,
+      "grad_norm": 0.3339071273803711,
+      "learning_rate": 0.00011435534563249123,
+      "loss": 1.3926,
+      "step": 5149
+    },
+    {
+      "epoch": 2.361426441921798,
+      "grad_norm": 0.3305213749408722,
+      "learning_rate": 0.00011419777227678929,
+      "loss": 1.4092,
+      "step": 5150
+    },
+    {
+      "epoch": 2.3618851049191605,
+      "grad_norm": 0.30224788188934326,
+      "learning_rate": 0.00011404029356115558,
+      "loss": 1.4058,
+      "step": 5151
+    },
+    {
+      "epoch": 2.3623437679165233,
+      "grad_norm": 0.3267022967338562,
+      "learning_rate": 0.00011388290952422075,
+      "loss": 1.3408,
+      "step": 5152
+    },
+    {
+      "epoch": 2.362802430913886,
+      "grad_norm": 0.17730341851711273,
+      "learning_rate": 0.00011372562020459231,
+      "loss": 0.8203,
+      "step": 5153
+    },
+    {
+      "epoch": 2.3632610939112486,
+      "grad_norm": 0.2721312344074249,
+      "learning_rate": 0.00011356842564085434,
+      "loss": 0.8142,
+      "step": 5154
+    },
+    {
+      "epoch": 2.3637197569086115,
+      "grad_norm": 0.34186965227127075,
+      "learning_rate": 0.00011341132587156793,
+      "loss": 1.9951,
+      "step": 5155
+    },
+    {
+      "epoch": 2.364178419905974,
+      "grad_norm": 0.36209332942962646,
+      "learning_rate": 0.00011325432093527077,
+      "loss": 2.0428,
+      "step": 5156
+    },
+    {
+      "epoch": 2.3646370829033367,
+      "grad_norm": 0.3819536864757538,
+      "learning_rate": 0.0001130974108704772,
+      "loss": 1.578,
+      "step": 5157
+    },
+    {
+      "epoch": 2.3650957459006996,
+      "grad_norm": 0.36597928404808044,
+      "learning_rate": 0.00011294059571567861,
+      "loss": 1.5025,
+      "step": 5158
+    },
+    {
+      "epoch": 2.365554408898062,
+      "grad_norm": 0.328401654958725,
+      "learning_rate": 0.0001127838755093426,
+      "loss": 0.9453,
+      "step": 5159
+    },
+    {
+      "epoch": 2.366013071895425,
+      "grad_norm": 0.29384055733680725,
+      "learning_rate": 0.00011262725028991388,
+      "loss": 1.6815,
+      "step": 5160
+    },
+    {
+      "epoch": 2.3664717348927873,
+      "grad_norm": 0.3777378499507904,
+      "learning_rate": 0.00011247072009581383,
+      "loss": 1.2265,
+      "step": 5161
+    },
+    {
+      "epoch": 2.36693039789015,
+      "grad_norm": 0.7261742949485779,
+      "learning_rate": 0.00011231428496544033,
+      "loss": 1.8345,
+      "step": 5162
+    },
+    {
+      "epoch": 2.367389060887513,
+      "grad_norm": 0.2863733172416687,
+      "learning_rate": 0.00011215794493716808,
+      "loss": 0.9066,
+      "step": 5163
+    },
+    {
+      "epoch": 2.3678477238848754,
+      "grad_norm": 0.3603268265724182,
+      "learning_rate": 0.00011200170004934839,
+      "loss": 1.6718,
+      "step": 5164
+    },
+    {
+      "epoch": 2.3683063868822383,
+      "grad_norm": 0.2777169346809387,
+      "learning_rate": 0.00011184555034030936,
+      "loss": 1.1851,
+      "step": 5165
+    },
+    {
+      "epoch": 2.368765049879601,
+      "grad_norm": 0.24623733758926392,
+      "learning_rate": 0.00011168949584835553,
+      "loss": 1.5477,
+      "step": 5166
+    },
+    {
+      "epoch": 2.3692237128769635,
+      "grad_norm": 0.2616453766822815,
+      "learning_rate": 0.00011153353661176824,
+      "loss": 0.8338,
+      "step": 5167
+    },
+    {
+      "epoch": 2.3696823758743264,
+      "grad_norm": 0.24914288520812988,
+      "learning_rate": 0.00011137767266880538,
+      "loss": 0.859,
+      "step": 5168
+    },
+    {
+      "epoch": 2.3701410388716893,
+      "grad_norm": 0.21932218968868256,
+      "learning_rate": 0.00011122190405770172,
+      "loss": 1.8062,
+      "step": 5169
+    },
+    {
+      "epoch": 2.3705997018690517,
+      "grad_norm": 0.38597074151039124,
+      "learning_rate": 0.00011106623081666806,
+      "loss": 1.8027,
+      "step": 5170
+    },
+    {
+      "epoch": 2.3710583648664145,
+      "grad_norm": 0.33088138699531555,
+      "learning_rate": 0.0001109106529838923,
+      "loss": 1.2009,
+      "step": 5171
+    },
+    {
+      "epoch": 2.371517027863777,
+      "grad_norm": 0.2947171628475189,
+      "learning_rate": 0.00011075517059753892,
+      "loss": 1.1661,
+      "step": 5172
+    },
+    {
+      "epoch": 2.37197569086114,
+      "grad_norm": 0.26863881945610046,
+      "learning_rate": 0.00011059978369574875,
+      "loss": 1.145,
+      "step": 5173
+    },
+    {
+      "epoch": 2.3724343538585027,
+      "grad_norm": 0.2579619586467743,
+      "learning_rate": 0.00011044449231663939,
+      "loss": 1.1743,
+      "step": 5174
+    },
+    {
+      "epoch": 2.372893016855865,
+      "grad_norm": 0.4377444088459015,
+      "learning_rate": 0.00011028929649830489,
+      "loss": 1.72,
+      "step": 5175
+    },
+    {
+      "epoch": 2.373351679853228,
+      "grad_norm": 0.6905438303947449,
+      "learning_rate": 0.00011013419627881587,
+      "loss": 0.7064,
+      "step": 5176
+    },
+    {
+      "epoch": 2.3738103428505903,
+      "grad_norm": 0.1522587239742279,
+      "learning_rate": 0.00010997919169621962,
+      "loss": 1.4874,
+      "step": 5177
+    },
+    {
+      "epoch": 2.374269005847953,
+      "grad_norm": 0.3481024205684662,
+      "learning_rate": 0.00010982428278853984,
+      "loss": 0.9126,
+      "step": 5178
+    },
+    {
+      "epoch": 2.374727668845316,
+      "grad_norm": 0.24200594425201416,
+      "learning_rate": 0.00010966946959377682,
+      "loss": 1.7862,
+      "step": 5179
+    },
+    {
+      "epoch": 2.3751863318426785,
+      "grad_norm": 0.40860190987586975,
+      "learning_rate": 0.00010951475214990747,
+      "loss": 1.0853,
+      "step": 5180
+    },
+    {
+      "epoch": 2.3756449948400413,
+      "grad_norm": 0.09399288147687912,
+      "learning_rate": 0.0001093601304948848,
+      "loss": 1.2584,
+      "step": 5181
+    },
+    {
+      "epoch": 2.3761036578374037,
+      "grad_norm": 0.31636422872543335,
+      "learning_rate": 0.00010920560466663882,
+      "loss": 0.8112,
+      "step": 5182
+    },
+    {
+      "epoch": 2.3765623208347666,
+      "grad_norm": 0.3569824695587158,
+      "learning_rate": 0.00010905117470307573,
+      "loss": 1.3749,
+      "step": 5183
+    },
+    {
+      "epoch": 2.3770209838321295,
+      "grad_norm": 0.07064448297023773,
+      "learning_rate": 0.00010889684064207845,
+      "loss": 0.731,
+      "step": 5184
+    },
+    {
+      "epoch": 2.377479646829492,
+      "grad_norm": 0.2785203456878662,
+      "learning_rate": 0.00010874260252150598,
+      "loss": 1.1004,
+      "step": 5185
+    },
+    {
+      "epoch": 2.3779383098268547,
+      "grad_norm": 0.32780954241752625,
+      "learning_rate": 0.00010858846037919451,
+      "loss": 1.3779,
+      "step": 5186
+    },
+    {
+      "epoch": 2.3783969728242176,
+      "grad_norm": 0.23089632391929626,
+      "learning_rate": 0.00010843441425295575,
+      "loss": 0.9364,
+      "step": 5187
+    },
+    {
+      "epoch": 2.37885563582158,
+      "grad_norm": 0.2271614968776703,
+      "learning_rate": 0.00010828046418057858,
+      "loss": 1.3,
+      "step": 5188
+    },
+    {
+      "epoch": 2.379314298818943,
+      "grad_norm": 0.32289209961891174,
+      "learning_rate": 0.00010812661019982795,
+      "loss": 1.4732,
+      "step": 5189
+    },
+    {
+      "epoch": 2.3797729618163057,
+      "grad_norm": 0.2017325609922409,
+      "learning_rate": 0.00010797285234844539,
+      "loss": 0.9344,
+      "step": 5190
+    },
+    {
+      "epoch": 2.380231624813668,
+      "grad_norm": 0.2721211910247803,
+      "learning_rate": 0.00010781919066414881,
+      "loss": 1.6788,
+      "step": 5191
+    },
+    {
+      "epoch": 2.380690287811031,
+      "grad_norm": 0.3644427955150604,
+      "learning_rate": 0.00010766562518463268,
+      "loss": 1.307,
+      "step": 5192
+    },
+    {
+      "epoch": 2.3811489508083934,
+      "grad_norm": 0.35269877314567566,
+      "learning_rate": 0.00010751215594756741,
+      "loss": 1.3381,
+      "step": 5193
+    },
+    {
+      "epoch": 2.3816076138057563,
+      "grad_norm": 0.3497646450996399,
+      "learning_rate": 0.00010735878299060014,
+      "loss": 1.5481,
+      "step": 5194
+    },
+    {
+      "epoch": 2.382066276803119,
+      "grad_norm": 0.28884097933769226,
+      "learning_rate": 0.00010720550635135467,
+      "loss": 1.1808,
+      "step": 5195
+    },
+    {
+      "epoch": 2.3825249398004815,
+      "grad_norm": 0.43060022592544556,
+      "learning_rate": 0.00010705232606743066,
+      "loss": 1.9484,
+      "step": 5196
+    },
+    {
+      "epoch": 2.3829836027978444,
+      "grad_norm": 0.25279608368873596,
+      "learning_rate": 0.00010689924217640445,
+      "loss": 1.1418,
+      "step": 5197
+    },
+    {
+      "epoch": 2.383442265795207,
+      "grad_norm": 0.36996960639953613,
+      "learning_rate": 0.00010674625471582849,
+      "loss": 1.4287,
+      "step": 5198
+    },
+    {
+      "epoch": 2.3839009287925697,
+      "grad_norm": 0.3359069228172302,
+      "learning_rate": 0.0001065933637232317,
+      "loss": 1.0623,
+      "step": 5199
+    },
+    {
+      "epoch": 2.3843595917899325,
+      "grad_norm": 0.2985363006591797,
+      "learning_rate": 0.00010644056923611939,
+      "loss": 1.366,
+      "step": 5200
+    },
+    {
+      "epoch": 2.384818254787295,
+      "grad_norm": 0.5279024243354797,
+      "learning_rate": 0.0001062878712919732,
+      "loss": 2.0491,
+      "step": 5201
+    },
+    {
+      "epoch": 2.385276917784658,
+      "grad_norm": 0.3357357680797577,
+      "learning_rate": 0.00010613526992825096,
+      "loss": 1.1142,
+      "step": 5202
+    },
+    {
+      "epoch": 2.38573558078202,
+      "grad_norm": 0.2649446427822113,
+      "learning_rate": 0.00010598276518238709,
+      "loss": 0.9735,
+      "step": 5203
+    },
+    {
+      "epoch": 2.386194243779383,
+      "grad_norm": 0.23111626505851746,
+      "learning_rate": 0.00010583035709179162,
+      "loss": 0.9959,
+      "step": 5204
+    },
+    {
+      "epoch": 2.386652906776746,
+      "grad_norm": 0.18565475940704346,
+      "learning_rate": 0.0001056780456938518,
+      "loss": 1.4027,
+      "step": 5205
+    },
+    {
+      "epoch": 2.3871115697741083,
+      "grad_norm": 0.3480309247970581,
+      "learning_rate": 0.00010552583102593061,
+      "loss": 2.1028,
+      "step": 5206
+    },
+    {
+      "epoch": 2.387570232771471,
+      "grad_norm": 0.2654421329498291,
+      "learning_rate": 0.00010537371312536736,
+      "loss": 1.1671,
+      "step": 5207
+    },
+    {
+      "epoch": 2.388028895768834,
+      "grad_norm": 0.4123840928077698,
+      "learning_rate": 0.00010522169202947784,
+      "loss": 1.4777,
+      "step": 5208
+    },
+    {
+      "epoch": 2.3884875587661965,
+      "grad_norm": 0.28358355164527893,
+      "learning_rate": 0.0001050697677755536,
+      "loss": 0.8817,
+      "step": 5209
+    },
+    {
+      "epoch": 2.3889462217635593,
+      "grad_norm": 0.29822635650634766,
+      "learning_rate": 0.00010491794040086289,
+      "loss": 1.6391,
+      "step": 5210
+    },
+    {
+      "epoch": 2.3894048847609217,
+      "grad_norm": 0.34754717350006104,
+      "learning_rate": 0.00010476620994265013,
+      "loss": 0.8802,
+      "step": 5211
+    },
+    {
+      "epoch": 2.3898635477582846,
+      "grad_norm": 0.08379904180765152,
+      "learning_rate": 0.00010461457643813588,
+      "loss": 0.3658,
+      "step": 5212
+    },
+    {
+      "epoch": 2.3903222107556474,
+      "grad_norm": 0.06315892189741135,
+      "learning_rate": 0.00010446303992451689,
+      "loss": 0.8891,
+      "step": 5213
+    },
+    {
+      "epoch": 2.39078087375301,
+      "grad_norm": 0.31999555230140686,
+      "learning_rate": 0.00010431160043896615,
+      "loss": 1.1747,
+      "step": 5214
+    },
+    {
+      "epoch": 2.3912395367503727,
+      "grad_norm": 0.2988559901714325,
+      "learning_rate": 0.00010416025801863289,
+      "loss": 1.2582,
+      "step": 5215
+    },
+    {
+      "epoch": 2.3916981997477356,
+      "grad_norm": 0.342385858297348,
+      "learning_rate": 0.0001040090127006425,
+      "loss": 1.0219,
+      "step": 5216
+    },
+    {
+      "epoch": 2.392156862745098,
+      "grad_norm": 0.21291863918304443,
+      "learning_rate": 0.00010385786452209656,
+      "loss": 1.1656,
+      "step": 5217
+    },
+    {
+      "epoch": 2.392615525742461,
+      "grad_norm": 0.29049792885780334,
+      "learning_rate": 0.00010370681352007272,
+      "loss": 1.4101,
+      "step": 5218
+    },
+    {
+      "epoch": 2.3930741887398232,
+      "grad_norm": 0.26569485664367676,
+      "learning_rate": 0.00010355585973162501,
+      "loss": 1.0166,
+      "step": 5219
+    },
+    {
+      "epoch": 2.393532851737186,
+      "grad_norm": 0.325291246175766,
+      "learning_rate": 0.00010340500319378348,
+      "loss": 1.544,
+      "step": 5220
+    },
+    {
+      "epoch": 2.393991514734549,
+      "grad_norm": 0.30292245745658875,
+      "learning_rate": 0.00010325424394355421,
+      "loss": 1.1687,
+      "step": 5221
+    },
+    {
+      "epoch": 2.3944501777319114,
+      "grad_norm": 0.28847742080688477,
+      "learning_rate": 0.00010310358201791953,
+      "loss": 1.7224,
+      "step": 5222
+    },
+    {
+      "epoch": 2.3949088407292742,
+      "grad_norm": 0.36918914318084717,
+      "learning_rate": 0.000102953017453838,
+      "loss": 1.5133,
+      "step": 5223
+    },
+    {
+      "epoch": 2.3953675037266366,
+      "grad_norm": 0.32062050700187683,
+      "learning_rate": 0.00010280255028824414,
+      "loss": 1.5388,
+      "step": 5224
+    },
+    {
+      "epoch": 2.3958261667239995,
+      "grad_norm": 0.4374137222766876,
+      "learning_rate": 0.0001026521805580486,
+      "loss": 1.0036,
+      "step": 5225
+    },
+    {
+      "epoch": 2.3962848297213624,
+      "grad_norm": 0.24663524329662323,
+      "learning_rate": 0.00010250190830013823,
+      "loss": 1.4458,
+      "step": 5226
+    },
+    {
+      "epoch": 2.3967434927187248,
+      "grad_norm": 0.39408281445503235,
+      "learning_rate": 0.00010235173355137583,
+      "loss": 1.0691,
+      "step": 5227
+    },
+    {
+      "epoch": 2.3972021557160876,
+      "grad_norm": 0.33270618319511414,
+      "learning_rate": 0.00010220165634860041,
+      "loss": 1.4243,
+      "step": 5228
+    },
+    {
+      "epoch": 2.39766081871345,
+      "grad_norm": 0.4075172543525696,
+      "learning_rate": 0.00010205167672862691,
+      "loss": 1.5062,
+      "step": 5229
+    },
+    {
+      "epoch": 2.398119481710813,
+      "grad_norm": 0.28742605447769165,
+      "learning_rate": 0.00010190179472824651,
+      "loss": 1.4183,
+      "step": 5230
+    },
+    {
+      "epoch": 2.3985781447081758,
+      "grad_norm": 0.35692864656448364,
+      "learning_rate": 0.00010175201038422644,
+      "loss": 1.2006,
+      "step": 5231
+    },
+    {
+      "epoch": 2.399036807705538,
+      "grad_norm": 0.20757196843624115,
+      "learning_rate": 0.00010160232373330963,
+      "loss": 0.9988,
+      "step": 5232
+    },
+    {
+      "epoch": 2.399495470702901,
+      "grad_norm": 0.3633536696434021,
+      "learning_rate": 0.00010145273481221534,
+      "loss": 1.744,
+      "step": 5233
+    },
+    {
+      "epoch": 2.399954133700264,
+      "grad_norm": 0.3276228904724121,
+      "learning_rate": 0.00010130324365763894,
+      "loss": 0.7654,
+      "step": 5234
+    },
+    {
+      "epoch": 2.4004127966976263,
+      "grad_norm": 0.2344532608985901,
+      "learning_rate": 0.00010115385030625157,
+      "loss": 1.82,
+      "step": 5235
+    },
+    {
+      "epoch": 2.400871459694989,
+      "grad_norm": 0.4215475022792816,
+      "learning_rate": 0.00010100455479470055,
+      "loss": 1.3953,
+      "step": 5236
+    },
+    {
+      "epoch": 2.401330122692352,
+      "grad_norm": 0.3673091232776642,
+      "learning_rate": 0.0001008553571596092,
+      "loss": 2.074,
+      "step": 5237
+    },
+    {
+      "epoch": 2.4017887856897144,
+      "grad_norm": 0.34490370750427246,
+      "learning_rate": 0.00010070625743757666,
+      "loss": 1.7086,
+      "step": 5238
+    },
+    {
+      "epoch": 2.4022474486870773,
+      "grad_norm": 0.35969072580337524,
+      "learning_rate": 0.00010055725566517826,
+      "loss": 1.1673,
+      "step": 5239
+    },
+    {
+      "epoch": 2.4027061116844397,
+      "grad_norm": 0.36554619669914246,
+      "learning_rate": 0.00010040835187896513,
+      "loss": 1.6442,
+      "step": 5240
+    },
+    {
+      "epoch": 2.4031647746818026,
+      "grad_norm": 0.39475345611572266,
+      "learning_rate": 0.00010025954611546457,
+      "loss": 0.4012,
+      "step": 5241
+    },
+    {
+      "epoch": 2.4036234376791654,
+      "grad_norm": 0.1467778980731964,
+      "learning_rate": 0.00010011083841117968,
+      "loss": 0.8082,
+      "step": 5242
+    },
+    {
+      "epoch": 2.404082100676528,
+      "grad_norm": 0.13162125647068024,
+      "learning_rate": 9.996222880258937e-05,
+      "loss": 0.8181,
+      "step": 5243
+    },
+    {
+      "epoch": 2.4045407636738907,
+      "grad_norm": 0.20284073054790497,
+      "learning_rate": 9.98137173261488e-05,
+      "loss": 1.213,
+      "step": 5244
+    },
+    {
+      "epoch": 2.404999426671253,
+      "grad_norm": 0.303290456533432,
+      "learning_rate": 9.966530401828883e-05,
+      "loss": 1.3629,
+      "step": 5245
+    },
+    {
+      "epoch": 2.405458089668616,
+      "grad_norm": 0.4616430997848511,
+      "learning_rate": 9.951698891541633e-05,
+      "loss": 1.5586,
+      "step": 5246
+    },
+    {
+      "epoch": 2.405916752665979,
+      "grad_norm": 0.3878755569458008,
+      "learning_rate": 9.936877205391398e-05,
+      "loss": 1.2528,
+      "step": 5247
+    },
+    {
+      "epoch": 2.4063754156633412,
+      "grad_norm": 0.29884567856788635,
+      "learning_rate": 9.922065347014075e-05,
+      "loss": 0.9254,
+      "step": 5248
+    },
+    {
+      "epoch": 2.406834078660704,
+      "grad_norm": 0.2206033170223236,
+      "learning_rate": 9.907263320043092e-05,
+      "loss": 1.3166,
+      "step": 5249
+    },
+    {
+      "epoch": 2.4072927416580665,
+      "grad_norm": 0.3127782642841339,
+      "learning_rate": 9.892471128109498e-05,
+      "loss": 1.0651,
+      "step": 5250
+    },
+    {
+      "epoch": 2.4077514046554294,
+      "grad_norm": 0.3565352261066437,
+      "learning_rate": 9.877688774841931e-05,
+      "loss": 1.317,
+      "step": 5251
+    },
+    {
+      "epoch": 2.408210067652792,
+      "grad_norm": 0.35081884264945984,
+      "learning_rate": 9.8629162638666e-05,
+      "loss": 0.9689,
+      "step": 5252
+    },
+    {
+      "epoch": 2.4086687306501546,
+      "grad_norm": 0.2946298122406006,
+      "learning_rate": 9.848153598807324e-05,
+      "loss": 2.0137,
+      "step": 5253
+    },
+    {
+      "epoch": 2.4091273936475175,
+      "grad_norm": 0.3904908001422882,
+      "learning_rate": 9.833400783285474e-05,
+      "loss": 1.0215,
+      "step": 5254
+    },
+    {
+      "epoch": 2.4095860566448803,
+      "grad_norm": 0.4543945789337158,
+      "learning_rate": 9.818657820920019e-05,
+      "loss": 1.4427,
+      "step": 5255
+    },
+    {
+      "epoch": 2.4100447196422428,
+      "grad_norm": 0.2723967730998993,
+      "learning_rate": 9.803924715327528e-05,
+      "loss": 1.3513,
+      "step": 5256
+    },
+    {
+      "epoch": 2.4105033826396056,
+      "grad_norm": 0.2895095646381378,
+      "learning_rate": 9.789201470122121e-05,
+      "loss": 0.9585,
+      "step": 5257
+    },
+    {
+      "epoch": 2.4109620456369685,
+      "grad_norm": 0.27330470085144043,
+      "learning_rate": 9.774488088915539e-05,
+      "loss": 1.174,
+      "step": 5258
+    },
+    {
+      "epoch": 2.411420708634331,
+      "grad_norm": 0.31435030698776245,
+      "learning_rate": 9.759784575317082e-05,
+      "loss": 1.9413,
+      "step": 5259
+    },
+    {
+      "epoch": 2.4118793716316937,
+      "grad_norm": 0.34424829483032227,
+      "learning_rate": 9.74509093293361e-05,
+      "loss": 1.6773,
+      "step": 5260
+    },
+    {
+      "epoch": 2.412338034629056,
+      "grad_norm": 0.37348756194114685,
+      "learning_rate": 9.730407165369576e-05,
+      "loss": 1.3452,
+      "step": 5261
+    },
+    {
+      "epoch": 2.412796697626419,
+      "grad_norm": 0.3043566346168518,
+      "learning_rate": 9.715733276227029e-05,
+      "loss": 1.5723,
+      "step": 5262
+    },
+    {
+      "epoch": 2.413255360623782,
+      "grad_norm": 0.3890668451786041,
+      "learning_rate": 9.701069269105567e-05,
+      "loss": 1.2312,
+      "step": 5263
+    },
+    {
+      "epoch": 2.4137140236211443,
+      "grad_norm": 0.30525264143943787,
+      "learning_rate": 9.686415147602401e-05,
+      "loss": 1.858,
+      "step": 5264
+    },
+    {
+      "epoch": 2.414172686618507,
+      "grad_norm": 0.33891257643699646,
+      "learning_rate": 9.671770915312267e-05,
+      "loss": 0.9398,
+      "step": 5265
+    },
+    {
+      "epoch": 2.4146313496158696,
+      "grad_norm": 0.2957022488117218,
+      "learning_rate": 9.657136575827491e-05,
+      "loss": 1.884,
+      "step": 5266
+    },
+    {
+      "epoch": 2.4150900126132324,
+      "grad_norm": 0.5537558197975159,
+      "learning_rate": 9.642512132738012e-05,
+      "loss": 1.1761,
+      "step": 5267
+    },
+    {
+      "epoch": 2.4155486756105953,
+      "grad_norm": 0.39335304498672485,
+      "learning_rate": 9.627897589631301e-05,
+      "loss": 1.5337,
+      "step": 5268
+    },
+    {
+      "epoch": 2.4160073386079577,
+      "grad_norm": 0.23258255422115326,
+      "learning_rate": 9.613292950092406e-05,
+      "loss": 1.4903,
+      "step": 5269
+    },
+    {
+      "epoch": 2.4164660016053205,
+      "grad_norm": 0.39902931451797485,
+      "learning_rate": 9.598698217703972e-05,
+      "loss": 1.5256,
+      "step": 5270
+    },
+    {
+      "epoch": 2.416924664602683,
+      "grad_norm": 0.3279167711734772,
+      "learning_rate": 9.584113396046157e-05,
+      "loss": 1.2243,
+      "step": 5271
+    },
+    {
+      "epoch": 2.417383327600046,
+      "grad_norm": 0.44774487614631653,
+      "learning_rate": 9.569538488696744e-05,
+      "loss": 1.1367,
+      "step": 5272
+    },
+    {
+      "epoch": 2.4178419905974087,
+      "grad_norm": 0.29976382851600647,
+      "learning_rate": 9.55497349923105e-05,
+      "loss": 1.3715,
+      "step": 5273
+    },
+    {
+      "epoch": 2.418300653594771,
+      "grad_norm": 0.3518228232860565,
+      "learning_rate": 9.540418431221986e-05,
+      "loss": 1.4482,
+      "step": 5274
+    },
+    {
+      "epoch": 2.418759316592134,
+      "grad_norm": 0.3567742109298706,
+      "learning_rate": 9.525873288240011e-05,
+      "loss": 1.937,
+      "step": 5275
+    },
+    {
+      "epoch": 2.419217979589497,
+      "grad_norm": 0.35031193494796753,
+      "learning_rate": 9.511338073853149e-05,
+      "loss": 1.551,
+      "step": 5276
+    },
+    {
+      "epoch": 2.419676642586859,
+      "grad_norm": 0.1989692896604538,
+      "learning_rate": 9.496812791626996e-05,
+      "loss": 1.1373,
+      "step": 5277
+    },
+    {
+      "epoch": 2.420135305584222,
+      "grad_norm": 0.3707393705844879,
+      "learning_rate": 9.48229744512471e-05,
+      "loss": 1.5035,
+      "step": 5278
+    },
+    {
+      "epoch": 2.4205939685815845,
+      "grad_norm": 0.3178713023662567,
+      "learning_rate": 9.467792037907008e-05,
+      "loss": 1.8657,
+      "step": 5279
+    },
+    {
+      "epoch": 2.4210526315789473,
+      "grad_norm": 0.30721351504325867,
+      "learning_rate": 9.453296573532172e-05,
+      "loss": 1.3541,
+      "step": 5280
+    },
+    {
+      "epoch": 2.42151129457631,
+      "grad_norm": 0.4439692199230194,
+      "learning_rate": 9.438811055556057e-05,
+      "loss": 1.243,
+      "step": 5281
+    },
+    {
+      "epoch": 2.4219699575736726,
+      "grad_norm": 0.206447571516037,
+      "learning_rate": 9.424335487532037e-05,
+      "loss": 0.8625,
+      "step": 5282
+    },
+    {
+      "epoch": 2.4224286205710355,
+      "grad_norm": 0.3932112157344818,
+      "learning_rate": 9.40986987301109e-05,
+      "loss": 1.208,
+      "step": 5283
+    },
+    {
+      "epoch": 2.4228872835683983,
+      "grad_norm": 0.28267911076545715,
+      "learning_rate": 9.395414215541731e-05,
+      "loss": 1.3602,
+      "step": 5284
+    },
+    {
+      "epoch": 2.4233459465657607,
+      "grad_norm": 0.3156595528125763,
+      "learning_rate": 9.380968518670036e-05,
+      "loss": 1.6733,
+      "step": 5285
+    },
+    {
+      "epoch": 2.4238046095631236,
+      "grad_norm": 0.3946542739868164,
+      "learning_rate": 9.366532785939647e-05,
+      "loss": 1.978,
+      "step": 5286
+    },
+    {
+      "epoch": 2.424263272560486,
+      "grad_norm": 0.4107709527015686,
+      "learning_rate": 9.352107020891743e-05,
+      "loss": 2.0332,
+      "step": 5287
+    },
+    {
+      "epoch": 2.424721935557849,
+      "grad_norm": 0.3392653167247772,
+      "learning_rate": 9.337691227065075e-05,
+      "loss": 1.1757,
+      "step": 5288
+    },
+    {
+      "epoch": 2.4251805985552117,
+      "grad_norm": 0.3073945939540863,
+      "learning_rate": 9.32328540799594e-05,
+      "loss": 1.5089,
+      "step": 5289
+    },
+    {
+      "epoch": 2.425639261552574,
+      "grad_norm": 0.37497302889823914,
+      "learning_rate": 9.308889567218193e-05,
+      "loss": 1.3739,
+      "step": 5290
+    },
+    {
+      "epoch": 2.426097924549937,
+      "grad_norm": 0.2676681578159332,
+      "learning_rate": 9.29450370826323e-05,
+      "loss": 0.9049,
+      "step": 5291
+    },
+    {
+      "epoch": 2.4265565875472994,
+      "grad_norm": 0.3354664742946625,
+      "learning_rate": 9.280127834660019e-05,
+      "loss": 1.2768,
+      "step": 5292
+    },
+    {
+      "epoch": 2.4270152505446623,
+      "grad_norm": 0.2157263159751892,
+      "learning_rate": 9.265761949935048e-05,
+      "loss": 1.1697,
+      "step": 5293
+    },
+    {
+      "epoch": 2.427473913542025,
+      "grad_norm": 0.323390394449234,
+      "learning_rate": 9.251406057612378e-05,
+      "loss": 1.2113,
+      "step": 5294
+    },
+    {
+      "epoch": 2.4279325765393875,
+      "grad_norm": 0.2757461965084076,
+      "learning_rate": 9.237060161213612e-05,
+      "loss": 1.727,
+      "step": 5295
+    },
+    {
+      "epoch": 2.4283912395367504,
+      "grad_norm": 0.33096444606781006,
+      "learning_rate": 9.222724264257904e-05,
+      "loss": 1.4423,
+      "step": 5296
+    },
+    {
+      "epoch": 2.428849902534113,
+      "grad_norm": 0.31503209471702576,
+      "learning_rate": 9.208398370261956e-05,
+      "loss": 1.6621,
+      "step": 5297
+    },
+    {
+      "epoch": 2.4293085655314757,
+      "grad_norm": 0.41407549381256104,
+      "learning_rate": 9.194082482740012e-05,
+      "loss": 1.9802,
+      "step": 5298
+    },
+    {
+      "epoch": 2.4297672285288385,
+      "grad_norm": 0.31064707040786743,
+      "learning_rate": 9.17977660520386e-05,
+      "loss": 0.7767,
+      "step": 5299
+    },
+    {
+      "epoch": 2.430225891526201,
+      "grad_norm": 0.26733309030532837,
+      "learning_rate": 9.165480741162829e-05,
+      "loss": 1.0861,
+      "step": 5300
+    },
+    {
+      "epoch": 2.430684554523564,
+      "grad_norm": 0.19575932621955872,
+      "learning_rate": 9.151194894123815e-05,
+      "loss": 1.223,
+      "step": 5301
+    },
+    {
+      "epoch": 2.4311432175209267,
+      "grad_norm": 0.22654877603054047,
+      "learning_rate": 9.13691906759122e-05,
+      "loss": 0.4944,
+      "step": 5302
+    },
+    {
+      "epoch": 2.431601880518289,
+      "grad_norm": 0.35672181844711304,
+      "learning_rate": 9.122653265067022e-05,
+      "loss": 1.9861,
+      "step": 5303
+    },
+    {
+      "epoch": 2.432060543515652,
+      "grad_norm": 0.32909905910491943,
+      "learning_rate": 9.10839749005073e-05,
+      "loss": 0.9556,
+      "step": 5304
+    },
+    {
+      "epoch": 2.432519206513015,
+      "grad_norm": 0.32880130410194397,
+      "learning_rate": 9.094151746039364e-05,
+      "loss": 1.7028,
+      "step": 5305
+    },
+    {
+      "epoch": 2.432977869510377,
+      "grad_norm": 0.3756924867630005,
+      "learning_rate": 9.079916036527519e-05,
+      "loss": 1.2567,
+      "step": 5306
+    },
+    {
+      "epoch": 2.43343653250774,
+      "grad_norm": 0.39229616522789,
+      "learning_rate": 9.065690365007323e-05,
+      "loss": 1.4464,
+      "step": 5307
+    },
+    {
+      "epoch": 2.4338951955051025,
+      "grad_norm": 0.19489213824272156,
+      "learning_rate": 9.051474734968429e-05,
+      "loss": 1.6805,
+      "step": 5308
+    },
+    {
+      "epoch": 2.4343538585024653,
+      "grad_norm": 0.35982856154441833,
+      "learning_rate": 9.037269149898036e-05,
+      "loss": 0.8735,
+      "step": 5309
+    },
+    {
+      "epoch": 2.434812521499828,
+      "grad_norm": 0.30394065380096436,
+      "learning_rate": 9.02307361328088e-05,
+      "loss": 1.3712,
+      "step": 5310
+    },
+    {
+      "epoch": 2.4352711844971906,
+      "grad_norm": 0.28870540857315063,
+      "learning_rate": 9.008888128599224e-05,
+      "loss": 2.0938,
+      "step": 5311
+    },
+    {
+      "epoch": 2.4357298474945535,
+      "grad_norm": 0.37780362367630005,
+      "learning_rate": 8.994712699332875e-05,
+      "loss": 1.53,
+      "step": 5312
+    },
+    {
+      "epoch": 2.436188510491916,
+      "grad_norm": 0.41944536566734314,
+      "learning_rate": 8.98054732895916e-05,
+      "loss": 1.4752,
+      "step": 5313
+    },
+    {
+      "epoch": 2.4366471734892787,
+      "grad_norm": 0.2813193202018738,
+      "learning_rate": 8.966392020952952e-05,
+      "loss": 1.2044,
+      "step": 5314
+    },
+    {
+      "epoch": 2.4371058364866416,
+      "grad_norm": 0.38517066836357117,
+      "learning_rate": 8.95224677878666e-05,
+      "loss": 1.5339,
+      "step": 5315
+    },
+    {
+      "epoch": 2.437564499484004,
+      "grad_norm": 0.38846316933631897,
+      "learning_rate": 8.938111605930194e-05,
+      "loss": 0.9716,
+      "step": 5316
+    },
+    {
+      "epoch": 2.438023162481367,
+      "grad_norm": 0.30371803045272827,
+      "learning_rate": 8.923986505851023e-05,
+      "loss": 1.2411,
+      "step": 5317
+    },
+    {
+      "epoch": 2.4384818254787293,
+      "grad_norm": 0.28628620505332947,
+      "learning_rate": 8.909871482014132e-05,
+      "loss": 0.8832,
+      "step": 5318
+    },
+    {
+      "epoch": 2.438940488476092,
+      "grad_norm": 0.17390215396881104,
+      "learning_rate": 8.895766537882027e-05,
+      "loss": 1.0545,
+      "step": 5319
+    },
+    {
+      "epoch": 2.439399151473455,
+      "grad_norm": 0.3960818946361542,
+      "learning_rate": 8.8816716769148e-05,
+      "loss": 1.9539,
+      "step": 5320
+    },
+    {
+      "epoch": 2.4398578144708174,
+      "grad_norm": 0.4454388916492462,
+      "learning_rate": 8.867586902569968e-05,
+      "loss": 1.5828,
+      "step": 5321
+    },
+    {
+      "epoch": 2.4403164774681803,
+      "grad_norm": 0.33164292573928833,
+      "learning_rate": 8.85351221830265e-05,
+      "loss": 1.5117,
+      "step": 5322
+    },
+    {
+      "epoch": 2.440775140465543,
+      "grad_norm": 0.31885433197021484,
+      "learning_rate": 8.839447627565472e-05,
+      "loss": 1.6736,
+      "step": 5323
+    },
+    {
+      "epoch": 2.4412338034629055,
+      "grad_norm": 0.28047341108322144,
+      "learning_rate": 8.825393133808574e-05,
+      "loss": 0.3669,
+      "step": 5324
+    },
+    {
+      "epoch": 2.4416924664602684,
+      "grad_norm": 0.09215112030506134,
+      "learning_rate": 8.811348740479619e-05,
+      "loss": 1.2191,
+      "step": 5325
+    },
+    {
+      "epoch": 2.4421511294576312,
+      "grad_norm": 0.33795610070228577,
+      "learning_rate": 8.797314451023819e-05,
+      "loss": 1.0369,
+      "step": 5326
+    },
+    {
+      "epoch": 2.4426097924549937,
+      "grad_norm": 0.34105727076530457,
+      "learning_rate": 8.783290268883859e-05,
+      "loss": 1.9551,
+      "step": 5327
+    },
+    {
+      "epoch": 2.4430684554523565,
+      "grad_norm": 0.2725188732147217,
+      "learning_rate": 8.76927619749998e-05,
+      "loss": 0.8592,
+      "step": 5328
+    },
+    {
+      "epoch": 2.443527118449719,
+      "grad_norm": 0.36431559920310974,
+      "learning_rate": 8.75527224030993e-05,
+      "loss": 1.7697,
+      "step": 5329
+    },
+    {
+      "epoch": 2.443985781447082,
+      "grad_norm": 0.38820746541023254,
+      "learning_rate": 8.741278400749003e-05,
+      "loss": 0.8046,
+      "step": 5330
+    },
+    {
+      "epoch": 2.4444444444444446,
+      "grad_norm": 0.25825342535972595,
+      "learning_rate": 8.72729468224997e-05,
+      "loss": 1.3749,
+      "step": 5331
+    },
+    {
+      "epoch": 2.444903107441807,
+      "grad_norm": 0.300819993019104,
+      "learning_rate": 8.713321088243159e-05,
+      "loss": 0.8929,
+      "step": 5332
+    },
+    {
+      "epoch": 2.44536177043917,
+      "grad_norm": 0.0780341699719429,
+      "learning_rate": 8.699357622156368e-05,
+      "loss": 0.6362,
+      "step": 5333
+    },
+    {
+      "epoch": 2.4458204334365323,
+      "grad_norm": 0.28823330998420715,
+      "learning_rate": 8.685404287414939e-05,
+      "loss": 1.5396,
+      "step": 5334
+    },
+    {
+      "epoch": 2.446279096433895,
+      "grad_norm": 0.3040931522846222,
+      "learning_rate": 8.671461087441735e-05,
+      "loss": 0.73,
+      "step": 5335
+    },
+    {
+      "epoch": 2.446737759431258,
+      "grad_norm": 0.4103958010673523,
+      "learning_rate": 8.657528025657118e-05,
+      "loss": 2.0576,
+      "step": 5336
+    },
+    {
+      "epoch": 2.4471964224286205,
+      "grad_norm": 0.33495497703552246,
+      "learning_rate": 8.643605105478986e-05,
+      "loss": 0.8664,
+      "step": 5337
+    },
+    {
+      "epoch": 2.4476550854259833,
+      "grad_norm": 0.2784813344478607,
+      "learning_rate": 8.629692330322691e-05,
+      "loss": 1.8088,
+      "step": 5338
+    },
+    {
+      "epoch": 2.4481137484233457,
+      "grad_norm": 0.3490314185619354,
+      "learning_rate": 8.615789703601179e-05,
+      "loss": 1.2238,
+      "step": 5339
+    },
+    {
+      "epoch": 2.4485724114207086,
+      "grad_norm": 0.2916855812072754,
+      "learning_rate": 8.601897228724842e-05,
+      "loss": 1.0304,
+      "step": 5340
+    },
+    {
+      "epoch": 2.4490310744180714,
+      "grad_norm": 0.3161444365978241,
+      "learning_rate": 8.588014909101616e-05,
+      "loss": 1.833,
+      "step": 5341
+    },
+    {
+      "epoch": 2.449489737415434,
+      "grad_norm": 0.3349079191684723,
+      "learning_rate": 8.574142748136926e-05,
+      "loss": 0.7148,
+      "step": 5342
+    },
+    {
+      "epoch": 2.4499484004127967,
+      "grad_norm": 0.19072647392749786,
+      "learning_rate": 8.560280749233729e-05,
+      "loss": 0.9673,
+      "step": 5343
+    },
+    {
+      "epoch": 2.4504070634101596,
+      "grad_norm": 0.09794793277978897,
+      "learning_rate": 8.546428915792449e-05,
+      "loss": 0.7698,
+      "step": 5344
+    },
+    {
+      "epoch": 2.450865726407522,
+      "grad_norm": 0.209795743227005,
+      "learning_rate": 8.53258725121105e-05,
+      "loss": 1.0825,
+      "step": 5345
+    },
+    {
+      "epoch": 2.451324389404885,
+      "grad_norm": 0.39282190799713135,
+      "learning_rate": 8.518755758884988e-05,
+      "loss": 2.1384,
+      "step": 5346
+    },
+    {
+      "epoch": 2.4517830524022477,
+      "grad_norm": 0.3572220504283905,
+      "learning_rate": 8.504934442207241e-05,
+      "loss": 0.8121,
+      "step": 5347
+    },
+    {
+      "epoch": 2.45224171539961,
+      "grad_norm": 0.20404981076717377,
+      "learning_rate": 8.491123304568271e-05,
+      "loss": 0.3737,
+      "step": 5348
+    },
+    {
+      "epoch": 2.452700378396973,
+      "grad_norm": 0.1945522129535675,
+      "learning_rate": 8.477322349356042e-05,
+      "loss": 1.6915,
+      "step": 5349
+    },
+    {
+      "epoch": 2.4531590413943354,
+      "grad_norm": 0.4479086101055145,
+      "learning_rate": 8.46353157995604e-05,
+      "loss": 2.0293,
+      "step": 5350
+    },
+    {
+      "epoch": 2.4536177043916982,
+      "grad_norm": 0.3536776602268219,
+      "learning_rate": 8.449750999751238e-05,
+      "loss": 1.1595,
+      "step": 5351
+    },
+    {
+      "epoch": 2.454076367389061,
+      "grad_norm": 0.23635387420654297,
+      "learning_rate": 8.435980612122101e-05,
+      "loss": 0.7237,
+      "step": 5352
+    },
+    {
+      "epoch": 2.4545350303864235,
+      "grad_norm": 0.32154580950737,
+      "learning_rate": 8.422220420446613e-05,
+      "loss": 1.7249,
+      "step": 5353
+    },
+    {
+      "epoch": 2.4549936933837864,
+      "grad_norm": 0.34665876626968384,
+      "learning_rate": 8.408470428100262e-05,
+      "loss": 1.2853,
+      "step": 5354
+    },
+    {
+      "epoch": 2.455452356381149,
+      "grad_norm": 0.5096401572227478,
+      "learning_rate": 8.394730638455994e-05,
+      "loss": 0.9551,
+      "step": 5355
+    },
+    {
+      "epoch": 2.4559110193785116,
+      "grad_norm": 0.28207308053970337,
+      "learning_rate": 8.381001054884291e-05,
+      "loss": 1.1176,
+      "step": 5356
+    },
+    {
+      "epoch": 2.4563696823758745,
+      "grad_norm": 0.23310762643814087,
+      "learning_rate": 8.367281680753114e-05,
+      "loss": 0.9396,
+      "step": 5357
+    },
+    {
+      "epoch": 2.456828345373237,
+      "grad_norm": 0.29276517033576965,
+      "learning_rate": 8.353572519427932e-05,
+      "loss": 1.1319,
+      "step": 5358
+    },
+    {
+      "epoch": 2.4572870083705998,
+      "grad_norm": 0.18109527230262756,
+      "learning_rate": 8.339873574271694e-05,
+      "loss": 0.9314,
+      "step": 5359
+    },
+    {
+      "epoch": 2.457745671367962,
+      "grad_norm": 0.26322728395462036,
+      "learning_rate": 8.326184848644852e-05,
+      "loss": 1.3008,
+      "step": 5360
+    },
+    {
+      "epoch": 2.458204334365325,
+      "grad_norm": 0.39521169662475586,
+      "learning_rate": 8.312506345905358e-05,
+      "loss": 1.3013,
+      "step": 5361
+    },
+    {
+      "epoch": 2.458662997362688,
+      "grad_norm": 0.2466443032026291,
+      "learning_rate": 8.298838069408632e-05,
+      "loss": 1.2078,
+      "step": 5362
+    },
+    {
+      "epoch": 2.4591216603600503,
+      "grad_norm": 0.31641414761543274,
+      "learning_rate": 8.28518002250761e-05,
+      "loss": 1.4739,
+      "step": 5363
+    },
+    {
+      "epoch": 2.459580323357413,
+      "grad_norm": 0.34621062874794006,
+      "learning_rate": 8.271532208552712e-05,
+      "loss": 1.1296,
+      "step": 5364
+    },
+    {
+      "epoch": 2.4600389863547756,
+      "grad_norm": 0.2962487041950226,
+      "learning_rate": 8.25789463089185e-05,
+      "loss": 1.2483,
+      "step": 5365
+    },
+    {
+      "epoch": 2.4604976493521384,
+      "grad_norm": 0.1835666447877884,
+      "learning_rate": 8.2442672928704e-05,
+      "loss": 0.8519,
+      "step": 5366
+    },
+    {
+      "epoch": 2.4609563123495013,
+      "grad_norm": 0.290197491645813,
+      "learning_rate": 8.230650197831252e-05,
+      "loss": 0.886,
+      "step": 5367
+    },
+    {
+      "epoch": 2.4614149753468637,
+      "grad_norm": 0.27010378241539,
+      "learning_rate": 8.217043349114789e-05,
+      "loss": 1.2361,
+      "step": 5368
+    },
+    {
+      "epoch": 2.4618736383442266,
+      "grad_norm": 0.5618923306465149,
+      "learning_rate": 8.203446750058862e-05,
+      "loss": 2.2034,
+      "step": 5369
+    },
+    {
+      "epoch": 2.4623323013415894,
+      "grad_norm": 0.3638753294944763,
+      "learning_rate": 8.189860403998816e-05,
+      "loss": 1.4188,
+      "step": 5370
+    },
+    {
+      "epoch": 2.462790964338952,
+      "grad_norm": 0.22352086007595062,
+      "learning_rate": 8.176284314267479e-05,
+      "loss": 0.8657,
+      "step": 5371
+    },
+    {
+      "epoch": 2.4632496273363147,
+      "grad_norm": 0.32366839051246643,
+      "learning_rate": 8.162718484195169e-05,
+      "loss": 1.392,
+      "step": 5372
+    },
+    {
+      "epoch": 2.4637082903336776,
+      "grad_norm": 0.25489479303359985,
+      "learning_rate": 8.14916291710967e-05,
+      "loss": 1.4828,
+      "step": 5373
+    },
+    {
+      "epoch": 2.46416695333104,
+      "grad_norm": 0.34520161151885986,
+      "learning_rate": 8.135617616336272e-05,
+      "loss": 1.5386,
+      "step": 5374
+    },
+    {
+      "epoch": 2.464625616328403,
+      "grad_norm": 0.4345172643661499,
+      "learning_rate": 8.122082585197732e-05,
+      "loss": 0.7834,
+      "step": 5375
+    },
+    {
+      "epoch": 2.4650842793257652,
+      "grad_norm": 0.24464713037014008,
+      "learning_rate": 8.108557827014295e-05,
+      "loss": 0.9882,
+      "step": 5376
+    },
+    {
+      "epoch": 2.465542942323128,
+      "grad_norm": 0.2222648561000824,
+      "learning_rate": 8.09504334510367e-05,
+      "loss": 1.5798,
+      "step": 5377
+    },
+    {
+      "epoch": 2.466001605320491,
+      "grad_norm": 0.2998507618904114,
+      "learning_rate": 8.081539142781058e-05,
+      "loss": 1.2544,
+      "step": 5378
+    },
+    {
+      "epoch": 2.4664602683178534,
+      "grad_norm": 0.3242281973361969,
+      "learning_rate": 8.068045223359144e-05,
+      "loss": 0.8808,
+      "step": 5379
+    },
+    {
+      "epoch": 2.4669189313152162,
+      "grad_norm": 0.30032265186309814,
+      "learning_rate": 8.054561590148085e-05,
+      "loss": 1.9047,
+      "step": 5380
+    },
+    {
+      "epoch": 2.4673775943125786,
+      "grad_norm": 0.3668994903564453,
+      "learning_rate": 8.041088246455492e-05,
+      "loss": 0.7418,
+      "step": 5381
+    },
+    {
+      "epoch": 2.4678362573099415,
+      "grad_norm": 0.18076832592487335,
+      "learning_rate": 8.027625195586519e-05,
+      "loss": 1.2765,
+      "step": 5382
+    },
+    {
+      "epoch": 2.4682949203073044,
+      "grad_norm": 0.32863399386405945,
+      "learning_rate": 8.014172440843714e-05,
+      "loss": 1.4399,
+      "step": 5383
+    },
+    {
+      "epoch": 2.4687535833046668,
+      "grad_norm": 0.27445659041404724,
+      "learning_rate": 8.000729985527139e-05,
+      "loss": 0.8303,
+      "step": 5384
+    },
+    {
+      "epoch": 2.4692122463020296,
+      "grad_norm": 0.23894087970256805,
+      "learning_rate": 7.987297832934326e-05,
+      "loss": 1.2864,
+      "step": 5385
+    },
+    {
+      "epoch": 2.469670909299392,
+      "grad_norm": 0.2814047336578369,
+      "learning_rate": 7.973875986360285e-05,
+      "loss": 1.5797,
+      "step": 5386
+    },
+    {
+      "epoch": 2.470129572296755,
+      "grad_norm": 0.4583660364151001,
+      "learning_rate": 7.960464449097498e-05,
+      "loss": 1.522,
+      "step": 5387
+    },
+    {
+      "epoch": 2.4705882352941178,
+      "grad_norm": 0.3099602460861206,
+      "learning_rate": 7.947063224435897e-05,
+      "loss": 1.2623,
+      "step": 5388
+    },
+    {
+      "epoch": 2.47104689829148,
+      "grad_norm": 0.2514623701572418,
+      "learning_rate": 7.933672315662898e-05,
+      "loss": 1.2364,
+      "step": 5389
+    },
+    {
+      "epoch": 2.471505561288843,
+      "grad_norm": 0.27281129360198975,
+      "learning_rate": 7.920291726063395e-05,
+      "loss": 1.3347,
+      "step": 5390
+    },
+    {
+      "epoch": 2.471964224286206,
+      "grad_norm": 0.2575758099555969,
+      "learning_rate": 7.906921458919731e-05,
+      "loss": 1.0359,
+      "step": 5391
+    },
+    {
+      "epoch": 2.4724228872835683,
+      "grad_norm": 0.3928264081478119,
+      "learning_rate": 7.893561517511754e-05,
+      "loss": 1.5787,
+      "step": 5392
+    },
+    {
+      "epoch": 2.472881550280931,
+      "grad_norm": 0.2101087123155594,
+      "learning_rate": 7.880211905116747e-05,
+      "loss": 0.6325,
+      "step": 5393
+    },
+    {
+      "epoch": 2.473340213278294,
+      "grad_norm": 0.22780464589595795,
+      "learning_rate": 7.866872625009453e-05,
+      "loss": 1.2313,
+      "step": 5394
+    },
+    {
+      "epoch": 2.4737988762756564,
+      "grad_norm": 0.19158299267292023,
+      "learning_rate": 7.853543680462094e-05,
+      "loss": 0.5635,
+      "step": 5395
+    },
+    {
+      "epoch": 2.4742575392730193,
+      "grad_norm": 0.14477090537548065,
+      "learning_rate": 7.840225074744367e-05,
+      "loss": 1.1157,
+      "step": 5396
+    },
+    {
+      "epoch": 2.4747162022703817,
+      "grad_norm": 0.34960320591926575,
+      "learning_rate": 7.826916811123414e-05,
+      "loss": 1.5063,
+      "step": 5397
+    },
+    {
+      "epoch": 2.4751748652677446,
+      "grad_norm": 0.510556161403656,
+      "learning_rate": 7.813618892863849e-05,
+      "loss": 1.4665,
+      "step": 5398
+    },
+    {
+      "epoch": 2.4756335282651074,
+      "grad_norm": 0.31196311116218567,
+      "learning_rate": 7.80033132322776e-05,
+      "loss": 1.184,
+      "step": 5399
+    },
+    {
+      "epoch": 2.47609219126247,
+      "grad_norm": 0.3243480920791626,
+      "learning_rate": 7.787054105474667e-05,
+      "loss": 1.3827,
+      "step": 5400
+    },
+    {
+      "epoch": 2.4765508542598327,
+      "grad_norm": 0.21474260091781616,
+      "learning_rate": 7.773787242861557e-05,
+      "loss": 0.7858,
+      "step": 5401
+    },
+    {
+      "epoch": 2.477009517257195,
+      "grad_norm": 0.07359279692173004,
+      "learning_rate": 7.760530738642918e-05,
+      "loss": 0.9702,
+      "step": 5402
+    },
+    {
+      "epoch": 2.477468180254558,
+      "grad_norm": 0.34640324115753174,
+      "learning_rate": 7.747284596070647e-05,
+      "loss": 1.6594,
+      "step": 5403
+    },
+    {
+      "epoch": 2.477926843251921,
+      "grad_norm": 0.217728853225708,
+      "learning_rate": 7.734048818394141e-05,
+      "loss": 1.1982,
+      "step": 5404
+    },
+    {
+      "epoch": 2.478385506249283,
+      "grad_norm": 0.3077888488769531,
+      "learning_rate": 7.720823408860195e-05,
+      "loss": 1.1781,
+      "step": 5405
+    },
+    {
+      "epoch": 2.478844169246646,
+      "grad_norm": 0.39549943804740906,
+      "learning_rate": 7.707608370713116e-05,
+      "loss": 1.9365,
+      "step": 5406
+    },
+    {
+      "epoch": 2.4793028322440085,
+      "grad_norm": 0.3485013246536255,
+      "learning_rate": 7.694403707194647e-05,
+      "loss": 1.3334,
+      "step": 5407
+    },
+    {
+      "epoch": 2.4797614952413713,
+      "grad_norm": 0.290386438369751,
+      "learning_rate": 7.681209421543994e-05,
+      "loss": 1.2493,
+      "step": 5408
+    },
+    {
+      "epoch": 2.480220158238734,
+      "grad_norm": 0.3037647306919098,
+      "learning_rate": 7.668025516997795e-05,
+      "loss": 1.694,
+      "step": 5409
+    },
+    {
+      "epoch": 2.4806788212360966,
+      "grad_norm": 0.4137989282608032,
+      "learning_rate": 7.654851996790174e-05,
+      "loss": 2.085,
+      "step": 5410
+    },
+    {
+      "epoch": 2.4811374842334595,
+      "grad_norm": 0.3738190531730652,
+      "learning_rate": 7.64168886415268e-05,
+      "loss": 1.1558,
+      "step": 5411
+    },
+    {
+      "epoch": 2.4815961472308223,
+      "grad_norm": 0.3116551339626312,
+      "learning_rate": 7.628536122314328e-05,
+      "loss": 1.9504,
+      "step": 5412
+    },
+    {
+      "epoch": 2.4820548102281847,
+      "grad_norm": 0.30240756273269653,
+      "learning_rate": 7.615393774501578e-05,
+      "loss": 0.7979,
+      "step": 5413
+    },
+    {
+      "epoch": 2.4825134732255476,
+      "grad_norm": 0.29710403084754944,
+      "learning_rate": 7.602261823938339e-05,
+      "loss": 1.5649,
+      "step": 5414
+    },
+    {
+      "epoch": 2.4829721362229105,
+      "grad_norm": 0.3587922155857086,
+      "learning_rate": 7.589140273845995e-05,
+      "loss": 1.4416,
+      "step": 5415
+    },
+    {
+      "epoch": 2.483430799220273,
+      "grad_norm": 0.3009377121925354,
+      "learning_rate": 7.576029127443329e-05,
+      "loss": 1.2025,
+      "step": 5416
+    },
+    {
+      "epoch": 2.4838894622176357,
+      "grad_norm": 0.3110191226005554,
+      "learning_rate": 7.562928387946611e-05,
+      "loss": 1.5738,
+      "step": 5417
+    },
+    {
+      "epoch": 2.484348125214998,
+      "grad_norm": 0.5595056414604187,
+      "learning_rate": 7.549838058569542e-05,
+      "loss": 1.5994,
+      "step": 5418
+    },
+    {
+      "epoch": 2.484806788212361,
+      "grad_norm": 0.37332621216773987,
+      "learning_rate": 7.536758142523281e-05,
+      "loss": 1.8284,
+      "step": 5419
+    },
+    {
+      "epoch": 2.485265451209724,
+      "grad_norm": 0.34219086170196533,
+      "learning_rate": 7.523688643016424e-05,
+      "loss": 0.7883,
+      "step": 5420
+    },
+    {
+      "epoch": 2.4857241142070863,
+      "grad_norm": 0.2129797488451004,
+      "learning_rate": 7.510629563255017e-05,
+      "loss": 1.3984,
+      "step": 5421
+    },
+    {
+      "epoch": 2.486182777204449,
+      "grad_norm": 0.2923283874988556,
+      "learning_rate": 7.497580906442537e-05,
+      "loss": 1.4247,
+      "step": 5422
+    },
+    {
+      "epoch": 2.4866414402018115,
+      "grad_norm": 0.3322497606277466,
+      "learning_rate": 7.48454267577992e-05,
+      "loss": 1.6427,
+      "step": 5423
+    },
+    {
+      "epoch": 2.4871001031991744,
+      "grad_norm": 0.3204174041748047,
+      "learning_rate": 7.471514874465535e-05,
+      "loss": 0.9109,
+      "step": 5424
+    },
+    {
+      "epoch": 2.4875587661965373,
+      "grad_norm": 0.23341026902198792,
+      "learning_rate": 7.4584975056952e-05,
+      "loss": 0.7723,
+      "step": 5425
+    },
+    {
+      "epoch": 2.4880174291938997,
+      "grad_norm": 0.29962819814682007,
+      "learning_rate": 7.445490572662168e-05,
+      "loss": 1.5294,
+      "step": 5426
+    },
+    {
+      "epoch": 2.4884760921912625,
+      "grad_norm": 0.30557799339294434,
+      "learning_rate": 7.432494078557145e-05,
+      "loss": 0.6204,
+      "step": 5427
+    },
+    {
+      "epoch": 2.488934755188625,
+      "grad_norm": 0.14737457036972046,
+      "learning_rate": 7.41950802656824e-05,
+      "loss": 1.0,
+      "step": 5428
+    },
+    {
+      "epoch": 2.489393418185988,
+      "grad_norm": 0.29757335782051086,
+      "learning_rate": 7.406532419881035e-05,
+      "loss": 1.3304,
+      "step": 5429
+    },
+    {
+      "epoch": 2.4898520811833507,
+      "grad_norm": 0.3414275348186493,
+      "learning_rate": 7.39356726167854e-05,
+      "loss": 1.3593,
+      "step": 5430
+    },
+    {
+      "epoch": 2.490310744180713,
+      "grad_norm": 0.1296243965625763,
+      "learning_rate": 7.380612555141209e-05,
+      "loss": 0.3876,
+      "step": 5431
+    },
+    {
+      "epoch": 2.490769407178076,
+      "grad_norm": 0.2734202444553375,
+      "learning_rate": 7.367668303446917e-05,
+      "loss": 1.3786,
+      "step": 5432
+    },
+    {
+      "epoch": 2.4912280701754383,
+      "grad_norm": 0.35316675901412964,
+      "learning_rate": 7.354734509770983e-05,
+      "loss": 1.3578,
+      "step": 5433
+    },
+    {
+      "epoch": 2.491686733172801,
+      "grad_norm": 0.33932507038116455,
+      "learning_rate": 7.341811177286167e-05,
+      "loss": 1.4916,
+      "step": 5434
+    },
+    {
+      "epoch": 2.492145396170164,
+      "grad_norm": 0.535605788230896,
+      "learning_rate": 7.328898309162652e-05,
+      "loss": 0.992,
+      "step": 5435
+    },
+    {
+      "epoch": 2.4926040591675265,
+      "grad_norm": 0.31163036823272705,
+      "learning_rate": 7.315995908568051e-05,
+      "loss": 1.804,
+      "step": 5436
+    },
+    {
+      "epoch": 2.4930627221648893,
+      "grad_norm": 0.29368847608566284,
+      "learning_rate": 7.303103978667425e-05,
+      "loss": 1.1411,
+      "step": 5437
+    },
+    {
+      "epoch": 2.493521385162252,
+      "grad_norm": 0.40923842787742615,
+      "learning_rate": 7.290222522623263e-05,
+      "loss": 1.6904,
+      "step": 5438
+    },
+    {
+      "epoch": 2.4939800481596146,
+      "grad_norm": 0.18885545432567596,
+      "learning_rate": 7.277351543595457e-05,
+      "loss": 0.9873,
+      "step": 5439
+    },
+    {
+      "epoch": 2.4944387111569775,
+      "grad_norm": 0.31630420684814453,
+      "learning_rate": 7.264491044741367e-05,
+      "loss": 0.9632,
+      "step": 5440
+    },
+    {
+      "epoch": 2.4948973741543403,
+      "grad_norm": 0.0943269282579422,
+      "learning_rate": 7.251641029215761e-05,
+      "loss": 0.6806,
+      "step": 5441
+    },
+    {
+      "epoch": 2.4953560371517027,
+      "grad_norm": 0.38258495926856995,
+      "learning_rate": 7.238801500170838e-05,
+      "loss": 1.7886,
+      "step": 5442
+    },
+    {
+      "epoch": 2.4958147001490656,
+      "grad_norm": 0.2949416935443878,
+      "learning_rate": 7.225972460756236e-05,
+      "loss": 1.0203,
+      "step": 5443
+    },
+    {
+      "epoch": 2.496273363146428,
+      "grad_norm": 0.3290545344352722,
+      "learning_rate": 7.213153914119008e-05,
+      "loss": 1.5944,
+      "step": 5444
+    },
+    {
+      "epoch": 2.496732026143791,
+      "grad_norm": 0.19472959637641907,
+      "learning_rate": 7.200345863403629e-05,
+      "loss": 1.22,
+      "step": 5445
+    },
+    {
+      "epoch": 2.4971906891411537,
+      "grad_norm": 0.27619174122810364,
+      "learning_rate": 7.187548311752012e-05,
+      "loss": 1.4291,
+      "step": 5446
+    },
+    {
+      "epoch": 2.497649352138516,
+      "grad_norm": 0.3692777156829834,
+      "learning_rate": 7.174761262303492e-05,
+      "loss": 1.534,
+      "step": 5447
+    },
+    {
+      "epoch": 2.498108015135879,
+      "grad_norm": 0.4094838798046112,
+      "learning_rate": 7.161984718194819e-05,
+      "loss": 1.1218,
+      "step": 5448
+    },
+    {
+      "epoch": 2.4985666781332414,
+      "grad_norm": 0.4053116738796234,
+      "learning_rate": 7.149218682560182e-05,
+      "loss": 1.2167,
+      "step": 5449
+    },
+    {
+      "epoch": 2.4990253411306043,
+      "grad_norm": 0.32053858041763306,
+      "learning_rate": 7.136463158531159e-05,
+      "loss": 1.58,
+      "step": 5450
+    },
+    {
+      "epoch": 2.499484004127967,
+      "grad_norm": 0.287165105342865,
+      "learning_rate": 7.123718149236791e-05,
+      "loss": 1.4514,
+      "step": 5451
+    },
+    {
+      "epoch": 2.4999426671253295,
+      "grad_norm": 0.2848725914955139,
+      "learning_rate": 7.11098365780351e-05,
+      "loss": 1.2621,
+      "step": 5452
+    },
+    {
+      "epoch": 2.5004013301226924,
+      "grad_norm": 0.398199200630188,
+      "learning_rate": 7.098259687355174e-05,
+      "loss": 1.1001,
+      "step": 5453
+    },
+    {
+      "epoch": 2.500859993120055,
+      "grad_norm": 0.29740533232688904,
+      "learning_rate": 7.085546241013085e-05,
+      "loss": 1.9265,
+      "step": 5454
+    },
+    {
+      "epoch": 2.5013186561174177,
+      "grad_norm": 0.25716161727905273,
+      "learning_rate": 7.07284332189595e-05,
+      "loss": 0.3225,
+      "step": 5455
+    },
+    {
+      "epoch": 2.5017773191147805,
+      "grad_norm": 0.19730542600154877,
+      "learning_rate": 7.060150933119852e-05,
+      "loss": 1.6379,
+      "step": 5456
+    },
+    {
+      "epoch": 2.5022359821121434,
+      "grad_norm": 0.3309273421764374,
+      "learning_rate": 7.04746907779834e-05,
+      "loss": 1.517,
+      "step": 5457
+    },
+    {
+      "epoch": 2.502694645109506,
+      "grad_norm": 0.32735908031463623,
+      "learning_rate": 7.034797759042371e-05,
+      "loss": 1.1667,
+      "step": 5458
+    },
+    {
+      "epoch": 2.503153308106868,
+      "grad_norm": 0.3031359910964966,
+      "learning_rate": 7.022136979960303e-05,
+      "loss": 1.3101,
+      "step": 5459
+    },
+    {
+      "epoch": 2.503611971104231,
+      "grad_norm": 0.18799692392349243,
+      "learning_rate": 7.009486743657934e-05,
+      "loss": 0.7399,
+      "step": 5460
+    },
+    {
+      "epoch": 2.504070634101594,
+      "grad_norm": 0.3134501576423645,
+      "learning_rate": 6.996847053238437e-05,
+      "loss": 1.1102,
+      "step": 5461
+    },
+    {
+      "epoch": 2.5045292970989568,
+      "grad_norm": 0.2856585681438446,
+      "learning_rate": 6.98421791180242e-05,
+      "loss": 1.4576,
+      "step": 5462
+    },
+    {
+      "epoch": 2.504987960096319,
+      "grad_norm": 0.34674420952796936,
+      "learning_rate": 6.971599322447903e-05,
+      "loss": 1.7148,
+      "step": 5463
+    },
+    {
+      "epoch": 2.505446623093682,
+      "grad_norm": 0.3734150528907776,
+      "learning_rate": 6.958991288270334e-05,
+      "loss": 0.9374,
+      "step": 5464
+    },
+    {
+      "epoch": 2.5059052860910445,
+      "grad_norm": 0.27259308099746704,
+      "learning_rate": 6.94639381236254e-05,
+      "loss": 1.8166,
+      "step": 5465
+    },
+    {
+      "epoch": 2.5063639490884073,
+      "grad_norm": 0.3674778640270233,
+      "learning_rate": 6.933806897814787e-05,
+      "loss": 1.4604,
+      "step": 5466
+    },
+    {
+      "epoch": 2.50682261208577,
+      "grad_norm": 0.45165976881980896,
+      "learning_rate": 6.921230547714719e-05,
+      "loss": 1.4612,
+      "step": 5467
+    },
+    {
+      "epoch": 2.5072812750831326,
+      "grad_norm": 0.2723102271556854,
+      "learning_rate": 6.90866476514741e-05,
+      "loss": 0.8123,
+      "step": 5468
+    },
+    {
+      "epoch": 2.5077399380804954,
+      "grad_norm": 0.27866679430007935,
+      "learning_rate": 6.896109553195334e-05,
+      "loss": 1.6746,
+      "step": 5469
+    },
+    {
+      "epoch": 2.508198601077858,
+      "grad_norm": 0.4871840476989746,
+      "learning_rate": 6.883564914938384e-05,
+      "loss": 2.0107,
+      "step": 5470
+    },
+    {
+      "epoch": 2.5086572640752207,
+      "grad_norm": 0.27734771370887756,
+      "learning_rate": 6.871030853453857e-05,
+      "loss": 0.4191,
+      "step": 5471
+    },
+    {
+      "epoch": 2.5091159270725836,
+      "grad_norm": 0.3640304207801819,
+      "learning_rate": 6.858507371816424e-05,
+      "loss": 1.7487,
+      "step": 5472
+    },
+    {
+      "epoch": 2.509574590069946,
+      "grad_norm": 0.42865926027297974,
+      "learning_rate": 6.845994473098194e-05,
+      "loss": 1.7012,
+      "step": 5473
+    },
+    {
+      "epoch": 2.510033253067309,
+      "grad_norm": 0.27769333124160767,
+      "learning_rate": 6.833492160368681e-05,
+      "loss": 0.8013,
+      "step": 5474
+    },
+    {
+      "epoch": 2.5104919160646713,
+      "grad_norm": 0.08679754287004471,
+      "learning_rate": 6.821000436694791e-05,
+      "loss": 1.0261,
+      "step": 5475
+    },
+    {
+      "epoch": 2.510950579062034,
+      "grad_norm": 0.41627833247184753,
+      "learning_rate": 6.808519305140831e-05,
+      "loss": 1.1582,
+      "step": 5476
+    },
+    {
+      "epoch": 2.511409242059397,
+      "grad_norm": 0.19535160064697266,
+      "learning_rate": 6.79604876876852e-05,
+      "loss": 1.0494,
+      "step": 5477
+    },
+    {
+      "epoch": 2.5118679050567594,
+      "grad_norm": 0.31459343433380127,
+      "learning_rate": 6.783588830636956e-05,
+      "loss": 1.9631,
+      "step": 5478
+    },
+    {
+      "epoch": 2.5123265680541222,
+      "grad_norm": 0.3410641551017761,
+      "learning_rate": 6.771139493802653e-05,
+      "loss": 1.2182,
+      "step": 5479
+    },
+    {
+      "epoch": 2.5127852310514847,
+      "grad_norm": 0.3488866090774536,
+      "learning_rate": 6.75870076131953e-05,
+      "loss": 1.1641,
+      "step": 5480
+    },
+    {
+      "epoch": 2.5132438940488475,
+      "grad_norm": 0.34782347083091736,
+      "learning_rate": 6.746272636238898e-05,
+      "loss": 1.5015,
+      "step": 5481
+    },
+    {
+      "epoch": 2.5137025570462104,
+      "grad_norm": 0.41583070158958435,
+      "learning_rate": 6.733855121609467e-05,
+      "loss": 1.5594,
+      "step": 5482
+    },
+    {
+      "epoch": 2.5141612200435732,
+      "grad_norm": 0.3468913435935974,
+      "learning_rate": 6.721448220477333e-05,
+      "loss": 1.8156,
+      "step": 5483
+    },
+    {
+      "epoch": 2.5146198830409356,
+      "grad_norm": 0.324745237827301,
+      "learning_rate": 6.709051935886007e-05,
+      "loss": 1.3654,
+      "step": 5484
+    },
+    {
+      "epoch": 2.5150785460382985,
+      "grad_norm": 0.31781327724456787,
+      "learning_rate": 6.696666270876389e-05,
+      "loss": 1.3649,
+      "step": 5485
+    },
+    {
+      "epoch": 2.515537209035661,
+      "grad_norm": 0.3324166238307953,
+      "learning_rate": 6.684291228486761e-05,
+      "loss": 0.8118,
+      "step": 5486
+    },
+    {
+      "epoch": 2.5159958720330238,
+      "grad_norm": 0.3521929085254669,
+      "learning_rate": 6.67192681175282e-05,
+      "loss": 2.1743,
+      "step": 5487
+    },
+    {
+      "epoch": 2.5164545350303866,
+      "grad_norm": 0.43346625566482544,
+      "learning_rate": 6.65957302370766e-05,
+      "loss": 1.2647,
+      "step": 5488
+    },
+    {
+      "epoch": 2.516913198027749,
+      "grad_norm": 0.10324136167764664,
+      "learning_rate": 6.647229867381722e-05,
+      "loss": 0.815,
+      "step": 5489
+    },
+    {
+      "epoch": 2.517371861025112,
+      "grad_norm": 0.23839758336544037,
+      "learning_rate": 6.634897345802888e-05,
+      "loss": 1.7404,
+      "step": 5490
+    },
+    {
+      "epoch": 2.5178305240224743,
+      "grad_norm": 0.39846575260162354,
+      "learning_rate": 6.62257546199641e-05,
+      "loss": 0.859,
+      "step": 5491
+    },
+    {
+      "epoch": 2.518289187019837,
+      "grad_norm": 0.2599179148674011,
+      "learning_rate": 6.610264218984946e-05,
+      "loss": 1.4614,
+      "step": 5492
+    },
+    {
+      "epoch": 2.5187478500172,
+      "grad_norm": 0.3531731963157654,
+      "learning_rate": 6.597963619788521e-05,
+      "loss": 1.4722,
+      "step": 5493
+    },
+    {
+      "epoch": 2.5192065130145624,
+      "grad_norm": 0.28518784046173096,
+      "learning_rate": 6.585673667424563e-05,
+      "loss": 0.9644,
+      "step": 5494
+    },
+    {
+      "epoch": 2.5196651760119253,
+      "grad_norm": 0.1243680790066719,
+      "learning_rate": 6.573394364907892e-05,
+      "loss": 0.4391,
+      "step": 5495
+    },
+    {
+      "epoch": 2.5201238390092877,
+      "grad_norm": 0.11469584703445435,
+      "learning_rate": 6.561125715250704e-05,
+      "loss": 1.0112,
+      "step": 5496
+    },
+    {
+      "epoch": 2.5205825020066506,
+      "grad_norm": 0.3533564507961273,
+      "learning_rate": 6.548867721462587e-05,
+      "loss": 1.1036,
+      "step": 5497
+    },
+    {
+      "epoch": 2.5210411650040134,
+      "grad_norm": 0.23048144578933716,
+      "learning_rate": 6.536620386550518e-05,
+      "loss": 1.0601,
+      "step": 5498
+    },
+    {
+      "epoch": 2.521499828001376,
+      "grad_norm": 0.20298047363758087,
+      "learning_rate": 6.524383713518867e-05,
+      "loss": 0.7747,
+      "step": 5499
+    },
+    {
+      "epoch": 2.5219584909987387,
+      "grad_norm": 0.3512571156024933,
+      "learning_rate": 6.512157705369354e-05,
+      "loss": 1.4583,
+      "step": 5500
+    },
+    {
+      "epoch": 2.522417153996101,
+      "grad_norm": 0.3375270962715149,
+      "learning_rate": 6.49994236510112e-05,
+      "loss": 1.4264,
+      "step": 5501
+    },
+    {
+      "epoch": 2.522875816993464,
+      "grad_norm": 0.27176031470298767,
+      "learning_rate": 6.487737695710677e-05,
+      "loss": 1.6927,
+      "step": 5502
+    },
+    {
+      "epoch": 2.523334479990827,
+      "grad_norm": 0.3321675956249237,
+      "learning_rate": 6.475543700191916e-05,
+      "loss": 0.9206,
+      "step": 5503
+    },
+    {
+      "epoch": 2.5237931429881897,
+      "grad_norm": 0.19732095301151276,
+      "learning_rate": 6.46336038153611e-05,
+      "loss": 1.333,
+      "step": 5504
+    },
+    {
+      "epoch": 2.524251805985552,
+      "grad_norm": 0.3470333218574524,
+      "learning_rate": 6.451187742731923e-05,
+      "loss": 1.6497,
+      "step": 5505
+    },
+    {
+      "epoch": 2.524710468982915,
+      "grad_norm": 0.3208187222480774,
+      "learning_rate": 6.43902578676538e-05,
+      "loss": 1.3061,
+      "step": 5506
+    },
+    {
+      "epoch": 2.5251691319802774,
+      "grad_norm": 0.353569358587265,
+      "learning_rate": 6.426874516619907e-05,
+      "loss": 2.2356,
+      "step": 5507
+    },
+    {
+      "epoch": 2.5256277949776402,
+      "grad_norm": 0.3554728031158447,
+      "learning_rate": 6.41473393527629e-05,
+      "loss": 1.2403,
+      "step": 5508
+    },
+    {
+      "epoch": 2.526086457975003,
+      "grad_norm": 0.3198685050010681,
+      "learning_rate": 6.402604045712707e-05,
+      "loss": 0.8093,
+      "step": 5509
+    },
+    {
+      "epoch": 2.5265451209723655,
+      "grad_norm": 0.20116372406482697,
+      "learning_rate": 6.39048485090471e-05,
+      "loss": 1.1317,
+      "step": 5510
+    },
+    {
+      "epoch": 2.5270037839697284,
+      "grad_norm": 0.33361703157424927,
+      "learning_rate": 6.378376353825222e-05,
+      "loss": 2.0852,
+      "step": 5511
+    },
+    {
+      "epoch": 2.5274624469670908,
+      "grad_norm": 0.3906664252281189,
+      "learning_rate": 6.366278557444538e-05,
+      "loss": 1.202,
+      "step": 5512
+    },
+    {
+      "epoch": 2.5279211099644536,
+      "grad_norm": 0.2128971815109253,
+      "learning_rate": 6.354191464730335e-05,
+      "loss": 0.8657,
+      "step": 5513
+    },
+    {
+      "epoch": 2.5283797729618165,
+      "grad_norm": 0.2909519672393799,
+      "learning_rate": 6.34211507864767e-05,
+      "loss": 1.7382,
+      "step": 5514
+    },
+    {
+      "epoch": 2.528838435959179,
+      "grad_norm": 0.4050029218196869,
+      "learning_rate": 6.330049402158955e-05,
+      "loss": 1.5528,
+      "step": 5515
+    },
+    {
+      "epoch": 2.5292970989565418,
+      "grad_norm": 0.27755218744277954,
+      "learning_rate": 6.317994438224023e-05,
+      "loss": 1.2415,
+      "step": 5516
+    },
+    {
+      "epoch": 2.529755761953904,
+      "grad_norm": 0.3861069083213806,
+      "learning_rate": 6.305950189800003e-05,
+      "loss": 2.1422,
+      "step": 5517
+    },
+    {
+      "epoch": 2.530214424951267,
+      "grad_norm": 0.5897046327590942,
+      "learning_rate": 6.293916659841454e-05,
+      "loss": 2.0242,
+      "step": 5518
+    },
+    {
+      "epoch": 2.53067308794863,
+      "grad_norm": 0.27490806579589844,
+      "learning_rate": 6.281893851300285e-05,
+      "loss": 0.8,
+      "step": 5519
+    },
+    {
+      "epoch": 2.5311317509459923,
+      "grad_norm": 0.35904571413993835,
+      "learning_rate": 6.269881767125778e-05,
+      "loss": 1.6329,
+      "step": 5520
+    },
+    {
+      "epoch": 2.531590413943355,
+      "grad_norm": 0.3965526819229126,
+      "learning_rate": 6.257880410264589e-05,
+      "loss": 1.6531,
+      "step": 5521
+    },
+    {
+      "epoch": 2.5320490769407176,
+      "grad_norm": 0.3628177046775818,
+      "learning_rate": 6.245889783660735e-05,
+      "loss": 1.5156,
+      "step": 5522
+    },
+    {
+      "epoch": 2.5325077399380804,
+      "grad_norm": 0.19028045237064362,
+      "learning_rate": 6.233909890255596e-05,
+      "loss": 0.7489,
+      "step": 5523
+    },
+    {
+      "epoch": 2.5329664029354433,
+      "grad_norm": 0.30474984645843506,
+      "learning_rate": 6.22194073298793e-05,
+      "loss": 1.6149,
+      "step": 5524
+    },
+    {
+      "epoch": 2.533425065932806,
+      "grad_norm": 0.3913409411907196,
+      "learning_rate": 6.209982314793845e-05,
+      "loss": 1.6556,
+      "step": 5525
+    },
+    {
+      "epoch": 2.5338837289301686,
+      "grad_norm": 0.3876248896121979,
+      "learning_rate": 6.198034638606854e-05,
+      "loss": 1.5082,
+      "step": 5526
+    },
+    {
+      "epoch": 2.5343423919275314,
+      "grad_norm": 0.324945867061615,
+      "learning_rate": 6.186097707357802e-05,
+      "loss": 1.6067,
+      "step": 5527
+    },
+    {
+      "epoch": 2.534801054924894,
+      "grad_norm": 0.26667168736457825,
+      "learning_rate": 6.174171523974886e-05,
+      "loss": 1.201,
+      "step": 5528
+    },
+    {
+      "epoch": 2.5352597179222567,
+      "grad_norm": 0.2764081656932831,
+      "learning_rate": 6.1622560913837e-05,
+      "loss": 1.2445,
+      "step": 5529
+    },
+    {
+      "epoch": 2.5357183809196195,
+      "grad_norm": 0.39166367053985596,
+      "learning_rate": 6.150351412507176e-05,
+      "loss": 1.3779,
+      "step": 5530
+    },
+    {
+      "epoch": 2.536177043916982,
+      "grad_norm": 0.2029440701007843,
+      "learning_rate": 6.138457490265625e-05,
+      "loss": 1.1893,
+      "step": 5531
+    },
+    {
+      "epoch": 2.536635706914345,
+      "grad_norm": 0.3429381847381592,
+      "learning_rate": 6.126574327576711e-05,
+      "loss": 1.2501,
+      "step": 5532
+    },
+    {
+      "epoch": 2.5370943699117072,
+      "grad_norm": 0.41793501377105713,
+      "learning_rate": 6.11470192735547e-05,
+      "loss": 1.7427,
+      "step": 5533
+    },
+    {
+      "epoch": 2.53755303290907,
+      "grad_norm": 0.3555110692977905,
+      "learning_rate": 6.1028402925142665e-05,
+      "loss": 1.5948,
+      "step": 5534
+    },
+    {
+      "epoch": 2.538011695906433,
+      "grad_norm": 0.26082339882850647,
+      "learning_rate": 6.090989425962839e-05,
+      "loss": 1.725,
+      "step": 5535
+    },
+    {
+      "epoch": 2.5384703589037954,
+      "grad_norm": 0.3338443338871002,
+      "learning_rate": 6.079149330608319e-05,
+      "loss": 1.4976,
+      "step": 5536
+    },
+    {
+      "epoch": 2.538929021901158,
+      "grad_norm": 0.35809195041656494,
+      "learning_rate": 6.067320009355148e-05,
+      "loss": 1.532,
+      "step": 5537
+    },
+    {
+      "epoch": 2.5393876848985206,
+      "grad_norm": 0.29855048656463623,
+      "learning_rate": 6.055501465105168e-05,
+      "loss": 1.3347,
+      "step": 5538
+    },
+    {
+      "epoch": 2.5398463478958835,
+      "grad_norm": 0.3272113800048828,
+      "learning_rate": 6.043693700757519e-05,
+      "loss": 0.8456,
+      "step": 5539
+    },
+    {
+      "epoch": 2.5403050108932463,
+      "grad_norm": 0.06610861420631409,
+      "learning_rate": 6.0318967192087516e-05,
+      "loss": 1.141,
+      "step": 5540
+    },
+    {
+      "epoch": 2.5407636738906088,
+      "grad_norm": 0.3634183406829834,
+      "learning_rate": 6.0201105233527453e-05,
+      "loss": 0.9582,
+      "step": 5541
+    },
+    {
+      "epoch": 2.5412223368879716,
+      "grad_norm": 0.13566255569458008,
+      "learning_rate": 6.008335116080732e-05,
+      "loss": 1.0286,
+      "step": 5542
+    },
+    {
+      "epoch": 2.541680999885334,
+      "grad_norm": 0.33389878273010254,
+      "learning_rate": 5.9965705002813174e-05,
+      "loss": 1.2472,
+      "step": 5543
+    },
+    {
+      "epoch": 2.542139662882697,
+      "grad_norm": 0.3005581796169281,
+      "learning_rate": 5.984816678840449e-05,
+      "loss": 1.3406,
+      "step": 5544
+    },
+    {
+      "epoch": 2.5425983258800597,
+      "grad_norm": 0.44604259729385376,
+      "learning_rate": 5.973073654641392e-05,
+      "loss": 1.6216,
+      "step": 5545
+    },
+    {
+      "epoch": 2.543056988877422,
+      "grad_norm": 0.365333616733551,
+      "learning_rate": 5.9613414305648314e-05,
+      "loss": 1.1402,
+      "step": 5546
+    },
+    {
+      "epoch": 2.543515651874785,
+      "grad_norm": 0.32972466945648193,
+      "learning_rate": 5.9496200094887546e-05,
+      "loss": 1.101,
+      "step": 5547
+    },
+    {
+      "epoch": 2.5439743148721474,
+      "grad_norm": 0.24824364483356476,
+      "learning_rate": 5.937909394288504e-05,
+      "loss": 1.7358,
+      "step": 5548
+    },
+    {
+      "epoch": 2.5444329778695103,
+      "grad_norm": 0.433464378118515,
+      "learning_rate": 5.92620958783679e-05,
+      "loss": 1.2081,
+      "step": 5549
+    },
+    {
+      "epoch": 2.544891640866873,
+      "grad_norm": 0.07062580436468124,
+      "learning_rate": 5.9145205930036583e-05,
+      "loss": 0.7539,
+      "step": 5550
+    },
+    {
+      "epoch": 2.545350303864236,
+      "grad_norm": 0.3250384032726288,
+      "learning_rate": 5.9028424126564894e-05,
+      "loss": 0.9884,
+      "step": 5551
+    },
+    {
+      "epoch": 2.5458089668615984,
+      "grad_norm": 0.3027406930923462,
+      "learning_rate": 5.89117504966003e-05,
+      "loss": 1.7088,
+      "step": 5552
+    },
+    {
+      "epoch": 2.5462676298589613,
+      "grad_norm": 0.23532333970069885,
+      "learning_rate": 5.879518506876375e-05,
+      "loss": 0.5515,
+      "step": 5553
+    },
+    {
+      "epoch": 2.5467262928563237,
+      "grad_norm": 0.25035008788108826,
+      "learning_rate": 5.867872787164952e-05,
+      "loss": 1.5544,
+      "step": 5554
+    },
+    {
+      "epoch": 2.5471849558536865,
+      "grad_norm": 0.38063761591911316,
+      "learning_rate": 5.8562378933825486e-05,
+      "loss": 1.6796,
+      "step": 5555
+    },
+    {
+      "epoch": 2.5476436188510494,
+      "grad_norm": 0.35923126339912415,
+      "learning_rate": 5.8446138283832826e-05,
+      "loss": 1.0818,
+      "step": 5556
+    },
+    {
+      "epoch": 2.548102281848412,
+      "grad_norm": 0.19541509449481964,
+      "learning_rate": 5.833000595018617e-05,
+      "loss": 1.6249,
+      "step": 5557
+    },
+    {
+      "epoch": 2.5485609448457747,
+      "grad_norm": 0.38436034321784973,
+      "learning_rate": 5.821398196137368e-05,
+      "loss": 1.3776,
+      "step": 5558
+    },
+    {
+      "epoch": 2.549019607843137,
+      "grad_norm": 0.2822701930999756,
+      "learning_rate": 5.8098066345856846e-05,
+      "loss": 0.4237,
+      "step": 5559
+    },
+    {
+      "epoch": 2.5494782708405,
+      "grad_norm": 0.15349756181240082,
+      "learning_rate": 5.798225913207067e-05,
+      "loss": 0.5825,
+      "step": 5560
+    },
+    {
+      "epoch": 2.549936933837863,
+      "grad_norm": 0.13313986361026764,
+      "learning_rate": 5.786656034842347e-05,
+      "loss": 1.3864,
+      "step": 5561
+    },
+    {
+      "epoch": 2.550395596835225,
+      "grad_norm": 0.45757579803466797,
+      "learning_rate": 5.775097002329693e-05,
+      "loss": 0.76,
+      "step": 5562
+    },
+    {
+      "epoch": 2.550854259832588,
+      "grad_norm": 0.1999332308769226,
+      "learning_rate": 5.763548818504616e-05,
+      "loss": 1.2987,
+      "step": 5563
+    },
+    {
+      "epoch": 2.5513129228299505,
+      "grad_norm": 0.30660200119018555,
+      "learning_rate": 5.752011486199982e-05,
+      "loss": 1.3302,
+      "step": 5564
+    },
+    {
+      "epoch": 2.5517715858273133,
+      "grad_norm": 0.27582883834838867,
+      "learning_rate": 5.740485008245966e-05,
+      "loss": 1.0434,
+      "step": 5565
+    },
+    {
+      "epoch": 2.552230248824676,
+      "grad_norm": 0.24701207876205444,
+      "learning_rate": 5.728969387470112e-05,
+      "loss": 0.4937,
+      "step": 5566
+    },
+    {
+      "epoch": 2.5526889118220386,
+      "grad_norm": 0.25263074040412903,
+      "learning_rate": 5.7174646266972776e-05,
+      "loss": 1.2341,
+      "step": 5567
+    },
+    {
+      "epoch": 2.5531475748194015,
+      "grad_norm": 0.3954515755176544,
+      "learning_rate": 5.70597072874966e-05,
+      "loss": 1.4972,
+      "step": 5568
+    },
+    {
+      "epoch": 2.553606237816764,
+      "grad_norm": 0.4183090031147003,
+      "learning_rate": 5.694487696446804e-05,
+      "loss": 2.063,
+      "step": 5569
+    },
+    {
+      "epoch": 2.5540649008141267,
+      "grad_norm": 0.3172152638435364,
+      "learning_rate": 5.683015532605573e-05,
+      "loss": 0.9099,
+      "step": 5570
+    },
+    {
+      "epoch": 2.5545235638114896,
+      "grad_norm": 0.3616786301136017,
+      "learning_rate": 5.671554240040172e-05,
+      "loss": 1.481,
+      "step": 5571
+    },
+    {
+      "epoch": 2.5549822268088525,
+      "grad_norm": 0.30609598755836487,
+      "learning_rate": 5.660103821562151e-05,
+      "loss": 1.6557,
+      "step": 5572
+    },
+    {
+      "epoch": 2.555440889806215,
+      "grad_norm": 0.4670886993408203,
+      "learning_rate": 5.648664279980365e-05,
+      "loss": 1.2957,
+      "step": 5573
+    },
+    {
+      "epoch": 2.5558995528035777,
+      "grad_norm": 0.2856581211090088,
+      "learning_rate": 5.6372356181010164e-05,
+      "loss": 1.1547,
+      "step": 5574
+    },
+    {
+      "epoch": 2.55635821580094,
+      "grad_norm": 0.21106205880641937,
+      "learning_rate": 5.625817838727643e-05,
+      "loss": 1.188,
+      "step": 5575
+    },
+    {
+      "epoch": 2.556816878798303,
+      "grad_norm": 0.1674429029226303,
+      "learning_rate": 5.614410944661108e-05,
+      "loss": 0.5747,
+      "step": 5576
+    },
+    {
+      "epoch": 2.557275541795666,
+      "grad_norm": 0.15634135901927948,
+      "learning_rate": 5.603014938699602e-05,
+      "loss": 1.5959,
+      "step": 5577
+    },
+    {
+      "epoch": 2.5577342047930283,
+      "grad_norm": 0.2876671552658081,
+      "learning_rate": 5.591629823638655e-05,
+      "loss": 0.8638,
+      "step": 5578
+    },
+    {
+      "epoch": 2.558192867790391,
+      "grad_norm": 0.3428886830806732,
+      "learning_rate": 5.5802556022711115e-05,
+      "loss": 1.4761,
+      "step": 5579
+    },
+    {
+      "epoch": 2.5586515307877535,
+      "grad_norm": 0.33712685108184814,
+      "learning_rate": 5.5688922773871555e-05,
+      "loss": 1.5991,
+      "step": 5580
+    },
+    {
+      "epoch": 2.5591101937851164,
+      "grad_norm": 0.3278370797634125,
+      "learning_rate": 5.557539851774285e-05,
+      "loss": 1.3031,
+      "step": 5581
+    },
+    {
+      "epoch": 2.5595688567824793,
+      "grad_norm": 0.30777499079704285,
+      "learning_rate": 5.546198328217333e-05,
+      "loss": 1.088,
+      "step": 5582
+    },
+    {
+      "epoch": 2.5600275197798417,
+      "grad_norm": 0.16568246483802795,
+      "learning_rate": 5.5348677094984755e-05,
+      "loss": 0.9488,
+      "step": 5583
+    },
+    {
+      "epoch": 2.5604861827772045,
+      "grad_norm": 0.3086935877799988,
+      "learning_rate": 5.523547998397166e-05,
+      "loss": 1.6018,
+      "step": 5584
+    },
+    {
+      "epoch": 2.560944845774567,
+      "grad_norm": 0.3688352406024933,
+      "learning_rate": 5.512239197690222e-05,
+      "loss": 1.5702,
+      "step": 5585
+    },
+    {
+      "epoch": 2.56140350877193,
+      "grad_norm": 0.34201377630233765,
+      "learning_rate": 5.5009413101517804e-05,
+      "loss": 1.51,
+      "step": 5586
+    },
+    {
+      "epoch": 2.5618621717692927,
+      "grad_norm": 0.5058449506759644,
+      "learning_rate": 5.4896543385532904e-05,
+      "loss": 2.0178,
+      "step": 5587
+    },
+    {
+      "epoch": 2.562320834766655,
+      "grad_norm": 0.34568533301353455,
+      "learning_rate": 5.4783782856635156e-05,
+      "loss": 1.5913,
+      "step": 5588
+    },
+    {
+      "epoch": 2.562779497764018,
+      "grad_norm": 0.27001887559890747,
+      "learning_rate": 5.467113154248582e-05,
+      "loss": 0.7074,
+      "step": 5589
+    },
+    {
+      "epoch": 2.5632381607613803,
+      "grad_norm": 0.3905376195907593,
+      "learning_rate": 5.455858947071885e-05,
+      "loss": 2.2896,
+      "step": 5590
+    },
+    {
+      "epoch": 2.563696823758743,
+      "grad_norm": 0.34636062383651733,
+      "learning_rate": 5.444615666894165e-05,
+      "loss": 1.4116,
+      "step": 5591
+    },
+    {
+      "epoch": 2.564155486756106,
+      "grad_norm": 0.3349077105522156,
+      "learning_rate": 5.433383316473484e-05,
+      "loss": 0.8179,
+      "step": 5592
+    },
+    {
+      "epoch": 2.564614149753469,
+      "grad_norm": 0.20421668887138367,
+      "learning_rate": 5.42216189856522e-05,
+      "loss": 1.3837,
+      "step": 5593
+    },
+    {
+      "epoch": 2.5650728127508313,
+      "grad_norm": 0.36774739623069763,
+      "learning_rate": 5.410951415922072e-05,
+      "loss": 1.0772,
+      "step": 5594
+    },
+    {
+      "epoch": 2.565531475748194,
+      "grad_norm": 0.27548274397850037,
+      "learning_rate": 5.399751871294034e-05,
+      "loss": 0.7288,
+      "step": 5595
+    },
+    {
+      "epoch": 2.5659901387455566,
+      "grad_norm": 0.27638116478919983,
+      "learning_rate": 5.388563267428448e-05,
+      "loss": 1.6965,
+      "step": 5596
+    },
+    {
+      "epoch": 2.5664488017429194,
+      "grad_norm": 0.3485362231731415,
+      "learning_rate": 5.377385607069951e-05,
+      "loss": 1.7012,
+      "step": 5597
+    },
+    {
+      "epoch": 2.5669074647402823,
+      "grad_norm": 0.4059440493583679,
+      "learning_rate": 5.366218892960517e-05,
+      "loss": 1.7341,
+      "step": 5598
+    },
+    {
+      "epoch": 2.5673661277376447,
+      "grad_norm": 0.3992423415184021,
+      "learning_rate": 5.3550631278394156e-05,
+      "loss": 1.6401,
+      "step": 5599
+    },
+    {
+      "epoch": 2.5678247907350076,
+      "grad_norm": 0.34082409739494324,
+      "learning_rate": 5.3439183144432414e-05,
+      "loss": 1.1353,
+      "step": 5600
+    },
+    {
+      "epoch": 2.56828345373237,
+      "grad_norm": 0.2575600743293762,
+      "learning_rate": 5.33278445550589e-05,
+      "loss": 0.4625,
+      "step": 5601
+    },
+    {
+      "epoch": 2.568742116729733,
+      "grad_norm": 0.30971759557724,
+      "learning_rate": 5.321661553758572e-05,
+      "loss": 1.5112,
+      "step": 5602
+    },
+    {
+      "epoch": 2.5692007797270957,
+      "grad_norm": 0.16957223415374756,
+      "learning_rate": 5.3105496119298266e-05,
+      "loss": 0.7928,
+      "step": 5603
+    },
+    {
+      "epoch": 2.569659442724458,
+      "grad_norm": 0.22056140005588531,
+      "learning_rate": 5.299448632745485e-05,
+      "loss": 0.8559,
+      "step": 5604
+    },
+    {
+      "epoch": 2.570118105721821,
+      "grad_norm": 0.2757347524166107,
+      "learning_rate": 5.288358618928707e-05,
+      "loss": 1.0507,
+      "step": 5605
+    },
+    {
+      "epoch": 2.5705767687191834,
+      "grad_norm": 0.2410385012626648,
+      "learning_rate": 5.27727957319995e-05,
+      "loss": 1.358,
+      "step": 5606
+    },
+    {
+      "epoch": 2.5710354317165462,
+      "grad_norm": 0.39340847730636597,
+      "learning_rate": 5.266211498276968e-05,
+      "loss": 1.5098,
+      "step": 5607
+    },
+    {
+      "epoch": 2.571494094713909,
+      "grad_norm": 0.23687437176704407,
+      "learning_rate": 5.2551543968748564e-05,
+      "loss": 1.2888,
+      "step": 5608
+    },
+    {
+      "epoch": 2.5719527577112715,
+      "grad_norm": 0.32233813405036926,
+      "learning_rate": 5.244108271706005e-05,
+      "loss": 1.6709,
+      "step": 5609
+    },
+    {
+      "epoch": 2.5724114207086344,
+      "grad_norm": 0.40911930799484253,
+      "learning_rate": 5.233073125480098e-05,
+      "loss": 1.2322,
+      "step": 5610
+    },
+    {
+      "epoch": 2.572870083705997,
+      "grad_norm": 0.24620671570301056,
+      "learning_rate": 5.2220489609041466e-05,
+      "loss": 0.8972,
+      "step": 5611
+    },
+    {
+      "epoch": 2.5733287467033596,
+      "grad_norm": 0.2210768610239029,
+      "learning_rate": 5.2110357806824494e-05,
+      "loss": 1.2861,
+      "step": 5612
+    },
+    {
+      "epoch": 2.5737874097007225,
+      "grad_norm": 0.9936086535453796,
+      "learning_rate": 5.2000335875166117e-05,
+      "loss": 1.4037,
+      "step": 5613
+    },
+    {
+      "epoch": 2.574246072698085,
+      "grad_norm": 0.22487474977970123,
+      "learning_rate": 5.1890423841055636e-05,
+      "loss": 1.1986,
+      "step": 5614
+    },
+    {
+      "epoch": 2.5747047356954478,
+      "grad_norm": 0.36320793628692627,
+      "learning_rate": 5.17806217314552e-05,
+      "loss": 1.7546,
+      "step": 5615
+    },
+    {
+      "epoch": 2.57516339869281,
+      "grad_norm": 0.3797517418861389,
+      "learning_rate": 5.167092957330011e-05,
+      "loss": 0.8319,
+      "step": 5616
+    },
+    {
+      "epoch": 2.575622061690173,
+      "grad_norm": 0.1391095519065857,
+      "learning_rate": 5.15613473934986e-05,
+      "loss": 0.9581,
+      "step": 5617
+    },
+    {
+      "epoch": 2.576080724687536,
+      "grad_norm": 0.3511468470096588,
+      "learning_rate": 5.145187521893202e-05,
+      "loss": 1.5688,
+      "step": 5618
+    },
+    {
+      "epoch": 2.5765393876848988,
+      "grad_norm": 0.33020275831222534,
+      "learning_rate": 5.13425130764546e-05,
+      "loss": 1.8931,
+      "step": 5619
+    },
+    {
+      "epoch": 2.576998050682261,
+      "grad_norm": 0.3568331003189087,
+      "learning_rate": 5.1233260992893715e-05,
+      "loss": 1.1766,
+      "step": 5620
+    },
+    {
+      "epoch": 2.577456713679624,
+      "grad_norm": 0.3786485195159912,
+      "learning_rate": 5.112411899504971e-05,
+      "loss": 1.5772,
+      "step": 5621
+    },
+    {
+      "epoch": 2.5779153766769864,
+      "grad_norm": 0.3674635887145996,
+      "learning_rate": 5.1015087109695934e-05,
+      "loss": 2.0232,
+      "step": 5622
+    },
+    {
+      "epoch": 2.5783740396743493,
+      "grad_norm": 0.428074449300766,
+      "learning_rate": 5.090616536357856e-05,
+      "loss": 0.9435,
+      "step": 5623
+    },
+    {
+      "epoch": 2.578832702671712,
+      "grad_norm": 0.10987308621406555,
+      "learning_rate": 5.0797353783416986e-05,
+      "loss": 0.4183,
+      "step": 5624
+    },
+    {
+      "epoch": 2.5792913656690746,
+      "grad_norm": 0.29732468724250793,
+      "learning_rate": 5.068865239590342e-05,
+      "loss": 0.8677,
+      "step": 5625
+    },
+    {
+      "epoch": 2.5797500286664374,
+      "grad_norm": 0.272393137216568,
+      "learning_rate": 5.0580061227703145e-05,
+      "loss": 0.9888,
+      "step": 5626
+    },
+    {
+      "epoch": 2.5802086916638,
+      "grad_norm": 0.23563776910305023,
+      "learning_rate": 5.047158030545434e-05,
+      "loss": 1.5809,
+      "step": 5627
+    },
+    {
+      "epoch": 2.5806673546611627,
+      "grad_norm": 0.4073021411895752,
+      "learning_rate": 5.036320965576813e-05,
+      "loss": 1.1274,
+      "step": 5628
+    },
+    {
+      "epoch": 2.5811260176585256,
+      "grad_norm": 0.4311738610267639,
+      "learning_rate": 5.0254949305228614e-05,
+      "loss": 0.9954,
+      "step": 5629
+    },
+    {
+      "epoch": 2.581584680655888,
+      "grad_norm": 0.2439291924238205,
+      "learning_rate": 5.01467992803929e-05,
+      "loss": 1.2027,
+      "step": 5630
+    },
+    {
+      "epoch": 2.582043343653251,
+      "grad_norm": 0.241475909948349,
+      "learning_rate": 5.003875960779097e-05,
+      "loss": 1.0108,
+      "step": 5631
+    },
+    {
+      "epoch": 2.5825020066506132,
+      "grad_norm": 0.25994980335235596,
+      "learning_rate": 4.9930830313925645e-05,
+      "loss": 1.0152,
+      "step": 5632
+    },
+    {
+      "epoch": 2.582960669647976,
+      "grad_norm": 0.234562948346138,
+      "learning_rate": 4.982301142527279e-05,
+      "loss": 0.7739,
+      "step": 5633
+    },
+    {
+      "epoch": 2.583419332645339,
+      "grad_norm": 0.2599080801010132,
+      "learning_rate": 4.971530296828131e-05,
+      "loss": 1.2073,
+      "step": 5634
+    },
+    {
+      "epoch": 2.5838779956427014,
+      "grad_norm": 0.3854179382324219,
+      "learning_rate": 4.960770496937267e-05,
+      "loss": 1.1806,
+      "step": 5635
+    },
+    {
+      "epoch": 2.5843366586400642,
+      "grad_norm": 0.21886670589447021,
+      "learning_rate": 4.9500217454941434e-05,
+      "loss": 1.6209,
+      "step": 5636
+    },
+    {
+      "epoch": 2.5847953216374266,
+      "grad_norm": 0.4453312158584595,
+      "learning_rate": 4.939284045135517e-05,
+      "loss": 1.5825,
+      "step": 5637
+    },
+    {
+      "epoch": 2.5852539846347895,
+      "grad_norm": 0.24376074969768524,
+      "learning_rate": 4.9285573984954144e-05,
+      "loss": 1.1272,
+      "step": 5638
+    },
+    {
+      "epoch": 2.5857126476321524,
+      "grad_norm": 0.3159853219985962,
+      "learning_rate": 4.91784180820517e-05,
+      "loss": 0.9751,
+      "step": 5639
+    },
+    {
+      "epoch": 2.586171310629515,
+      "grad_norm": 0.39112159609794617,
+      "learning_rate": 4.907137276893381e-05,
+      "loss": 1.7409,
+      "step": 5640
+    },
+    {
+      "epoch": 2.5866299736268776,
+      "grad_norm": 0.39343714714050293,
+      "learning_rate": 4.896443807185963e-05,
+      "loss": 1.5584,
+      "step": 5641
+    },
+    {
+      "epoch": 2.5870886366242405,
+      "grad_norm": 0.361751526594162,
+      "learning_rate": 4.88576140170609e-05,
+      "loss": 1.7249,
+      "step": 5642
+    },
+    {
+      "epoch": 2.587547299621603,
+      "grad_norm": 0.273359090089798,
+      "learning_rate": 4.875090063074233e-05,
+      "loss": 0.8665,
+      "step": 5643
+    },
+    {
+      "epoch": 2.5880059626189658,
+      "grad_norm": 0.11545918881893158,
+      "learning_rate": 4.864429793908154e-05,
+      "loss": 1.1998,
+      "step": 5644
+    },
+    {
+      "epoch": 2.5884646256163286,
+      "grad_norm": 1.5149635076522827,
+      "learning_rate": 4.8537805968228985e-05,
+      "loss": 1.7397,
+      "step": 5645
+    },
+    {
+      "epoch": 2.588923288613691,
+      "grad_norm": 0.37379226088523865,
+      "learning_rate": 4.843142474430773e-05,
+      "loss": 1.7961,
+      "step": 5646
+    },
+    {
+      "epoch": 2.589381951611054,
+      "grad_norm": 0.41186100244522095,
+      "learning_rate": 4.8325154293414e-05,
+      "loss": 1.182,
+      "step": 5647
+    },
+    {
+      "epoch": 2.5898406146084163,
+      "grad_norm": 0.23608848452568054,
+      "learning_rate": 4.8218994641616606e-05,
+      "loss": 1.4241,
+      "step": 5648
+    },
+    {
+      "epoch": 2.590299277605779,
+      "grad_norm": 0.2904841899871826,
+      "learning_rate": 4.8112945814957396e-05,
+      "loss": 0.8572,
+      "step": 5649
+    },
+    {
+      "epoch": 2.590757940603142,
+      "grad_norm": 0.20750272274017334,
+      "learning_rate": 4.800700783945083e-05,
+      "loss": 1.0704,
+      "step": 5650
+    },
+    {
+      "epoch": 2.5912166036005044,
+      "grad_norm": 0.2611089050769806,
+      "learning_rate": 4.790118074108429e-05,
+      "loss": 1.7778,
+      "step": 5651
+    },
+    {
+      "epoch": 2.5916752665978673,
+      "grad_norm": 0.4068661332130432,
+      "learning_rate": 4.77954645458179e-05,
+      "loss": 1.3848,
+      "step": 5652
+    },
+    {
+      "epoch": 2.5921339295952297,
+      "grad_norm": 0.2891179919242859,
+      "learning_rate": 4.768985927958464e-05,
+      "loss": 1.014,
+      "step": 5653
+    },
+    {
+      "epoch": 2.5925925925925926,
+      "grad_norm": 0.3051091432571411,
+      "learning_rate": 4.758436496829016e-05,
+      "loss": 1.1055,
+      "step": 5654
+    },
+    {
+      "epoch": 2.5930512555899554,
+      "grad_norm": 0.3727298080921173,
+      "learning_rate": 4.747898163781311e-05,
+      "loss": 1.7048,
+      "step": 5655
+    },
+    {
+      "epoch": 2.593509918587318,
+      "grad_norm": 0.26991981267929077,
+      "learning_rate": 4.737370931400475e-05,
+      "loss": 0.955,
+      "step": 5656
+    },
+    {
+      "epoch": 2.5939685815846807,
+      "grad_norm": 0.29294559359550476,
+      "learning_rate": 4.726854802268909e-05,
+      "loss": 0.7957,
+      "step": 5657
+    },
+    {
+      "epoch": 2.594427244582043,
+      "grad_norm": 0.28209206461906433,
+      "learning_rate": 4.716349778966289e-05,
+      "loss": 1.6527,
+      "step": 5658
+    },
+    {
+      "epoch": 2.594885907579406,
+      "grad_norm": 0.2640170454978943,
+      "learning_rate": 4.7058558640695805e-05,
+      "loss": 0.8565,
+      "step": 5659
+    },
+    {
+      "epoch": 2.595344570576769,
+      "grad_norm": 0.13128653168678284,
+      "learning_rate": 4.695373060153013e-05,
+      "loss": 0.6372,
+      "step": 5660
+    },
+    {
+      "epoch": 2.5958032335741317,
+      "grad_norm": 0.22628602385520935,
+      "learning_rate": 4.684901369788097e-05,
+      "loss": 1.334,
+      "step": 5661
+    },
+    {
+      "epoch": 2.596261896571494,
+      "grad_norm": 0.33850499987602234,
+      "learning_rate": 4.674440795543633e-05,
+      "loss": 1.685,
+      "step": 5662
+    },
+    {
+      "epoch": 2.596720559568857,
+      "grad_norm": 0.3774524927139282,
+      "learning_rate": 4.663991339985641e-05,
+      "loss": 1.4136,
+      "step": 5663
+    },
+    {
+      "epoch": 2.5971792225662194,
+      "grad_norm": 0.38456571102142334,
+      "learning_rate": 4.653553005677463e-05,
+      "loss": 1.7705,
+      "step": 5664
+    },
+    {
+      "epoch": 2.597637885563582,
+      "grad_norm": 0.2402615249156952,
+      "learning_rate": 4.643125795179698e-05,
+      "loss": 0.9066,
+      "step": 5665
+    },
+    {
+      "epoch": 2.598096548560945,
+      "grad_norm": 0.31829625368118286,
+      "learning_rate": 4.632709711050215e-05,
+      "loss": 1.3141,
+      "step": 5666
+    },
+    {
+      "epoch": 2.5985552115583075,
+      "grad_norm": 0.2783154547214508,
+      "learning_rate": 4.622304755844164e-05,
+      "loss": 1.2182,
+      "step": 5667
+    },
+    {
+      "epoch": 2.5990138745556703,
+      "grad_norm": 0.2189006507396698,
+      "learning_rate": 4.6119109321139384e-05,
+      "loss": 0.6295,
+      "step": 5668
+    },
+    {
+      "epoch": 2.5994725375530328,
+      "grad_norm": 0.20123249292373657,
+      "learning_rate": 4.6015282424092196e-05,
+      "loss": 1.3485,
+      "step": 5669
+    },
+    {
+      "epoch": 2.5999312005503956,
+      "grad_norm": 0.4019489884376526,
+      "learning_rate": 4.591156689276971e-05,
+      "loss": 1.5854,
+      "step": 5670
+    },
+    {
+      "epoch": 2.6003898635477585,
+      "grad_norm": 0.335580438375473,
+      "learning_rate": 4.580796275261395e-05,
+      "loss": 1.5316,
+      "step": 5671
+    },
+    {
+      "epoch": 2.600848526545121,
+      "grad_norm": 0.31390464305877686,
+      "learning_rate": 4.570447002903988e-05,
+      "loss": 1.3737,
+      "step": 5672
+    },
+    {
+      "epoch": 2.6013071895424837,
+      "grad_norm": 0.34774190187454224,
+      "learning_rate": 4.560108874743507e-05,
+      "loss": 1.2221,
+      "step": 5673
+    },
+    {
+      "epoch": 2.601765852539846,
+      "grad_norm": 0.13695718348026276,
+      "learning_rate": 4.5497818933159406e-05,
+      "loss": 1.0302,
+      "step": 5674
+    },
+    {
+      "epoch": 2.602224515537209,
+      "grad_norm": 0.28607991337776184,
+      "learning_rate": 4.5394660611545955e-05,
+      "loss": 0.8901,
+      "step": 5675
+    },
+    {
+      "epoch": 2.602683178534572,
+      "grad_norm": 0.2588864266872406,
+      "learning_rate": 4.529161380790009e-05,
+      "loss": 1.4887,
+      "step": 5676
+    },
+    {
+      "epoch": 2.6031418415319343,
+      "grad_norm": 0.2673766314983368,
+      "learning_rate": 4.5188678547499976e-05,
+      "loss": 1.3279,
+      "step": 5677
+    },
+    {
+      "epoch": 2.603600504529297,
+      "grad_norm": 0.22360926866531372,
+      "learning_rate": 4.508585485559652e-05,
+      "loss": 1.3532,
+      "step": 5678
+    },
+    {
+      "epoch": 2.6040591675266596,
+      "grad_norm": 0.35530173778533936,
+      "learning_rate": 4.498314275741272e-05,
+      "loss": 1.1174,
+      "step": 5679
+    },
+    {
+      "epoch": 2.6045178305240224,
+      "grad_norm": 0.20542040467262268,
+      "learning_rate": 4.488054227814497e-05,
+      "loss": 1.0656,
+      "step": 5680
+    },
+    {
+      "epoch": 2.6049764935213853,
+      "grad_norm": 0.3311353325843811,
+      "learning_rate": 4.4778053442961764e-05,
+      "loss": 0.7792,
+      "step": 5681
+    },
+    {
+      "epoch": 2.6054351565187477,
+      "grad_norm": 0.09570083022117615,
+      "learning_rate": 4.467567627700436e-05,
+      "loss": 1.1345,
+      "step": 5682
+    },
+    {
+      "epoch": 2.6058938195161105,
+      "grad_norm": 0.26272979378700256,
+      "learning_rate": 4.4573410805386624e-05,
+      "loss": 1.1146,
+      "step": 5683
+    },
+    {
+      "epoch": 2.606352482513473,
+      "grad_norm": 0.368933767080307,
+      "learning_rate": 4.447125705319516e-05,
+      "loss": 1.2422,
+      "step": 5684
+    },
+    {
+      "epoch": 2.606811145510836,
+      "grad_norm": 0.3198513686656952,
+      "learning_rate": 4.436921504548874e-05,
+      "loss": 1.2469,
+      "step": 5685
+    },
+    {
+      "epoch": 2.6072698085081987,
+      "grad_norm": 0.3850035071372986,
+      "learning_rate": 4.426728480729914e-05,
+      "loss": 1.6652,
+      "step": 5686
+    },
+    {
+      "epoch": 2.6077284715055615,
+      "grad_norm": 0.35521817207336426,
+      "learning_rate": 4.4165466363630556e-05,
+      "loss": 1.4731,
+      "step": 5687
+    },
+    {
+      "epoch": 2.608187134502924,
+      "grad_norm": 0.2965492308139801,
+      "learning_rate": 4.406375973945981e-05,
+      "loss": 1.5766,
+      "step": 5688
+    },
+    {
+      "epoch": 2.608645797500287,
+      "grad_norm": 0.3511826694011688,
+      "learning_rate": 4.396216495973632e-05,
+      "loss": 1.9038,
+      "step": 5689
+    },
+    {
+      "epoch": 2.609104460497649,
+      "grad_norm": 0.3754914104938507,
+      "learning_rate": 4.386068204938193e-05,
+      "loss": 1.5048,
+      "step": 5690
+    },
+    {
+      "epoch": 2.609563123495012,
+      "grad_norm": 0.374553382396698,
+      "learning_rate": 4.375931103329117e-05,
+      "loss": 1.7387,
+      "step": 5691
+    },
+    {
+      "epoch": 2.610021786492375,
+      "grad_norm": 0.3759397268295288,
+      "learning_rate": 4.36580519363311e-05,
+      "loss": 1.6533,
+      "step": 5692
+    },
+    {
+      "epoch": 2.6104804494897373,
+      "grad_norm": 0.3354843258857727,
+      "learning_rate": 4.355690478334129e-05,
+      "loss": 0.8031,
+      "step": 5693
+    },
+    {
+      "epoch": 2.6109391124871,
+      "grad_norm": 0.209818497300148,
+      "learning_rate": 4.3455869599133834e-05,
+      "loss": 1.2783,
+      "step": 5694
+    },
+    {
+      "epoch": 2.6113977754844626,
+      "grad_norm": 0.36011189222335815,
+      "learning_rate": 4.3354946408493524e-05,
+      "loss": 2.0714,
+      "step": 5695
+    },
+    {
+      "epoch": 2.6118564384818255,
+      "grad_norm": 0.2897484302520752,
+      "learning_rate": 4.325413523617733e-05,
+      "loss": 1.3004,
+      "step": 5696
+    },
+    {
+      "epoch": 2.6123151014791883,
+      "grad_norm": 0.4306742548942566,
+      "learning_rate": 4.315343610691508e-05,
+      "loss": 1.3275,
+      "step": 5697
+    },
+    {
+      "epoch": 2.6127737644765507,
+      "grad_norm": 0.14981167018413544,
+      "learning_rate": 4.305284904540901e-05,
+      "loss": 1.0299,
+      "step": 5698
+    },
+    {
+      "epoch": 2.6132324274739136,
+      "grad_norm": 0.3096173405647278,
+      "learning_rate": 4.295237407633379e-05,
+      "loss": 1.6353,
+      "step": 5699
+    },
+    {
+      "epoch": 2.613691090471276,
+      "grad_norm": 0.33780768513679504,
+      "learning_rate": 4.285201122433674e-05,
+      "loss": 0.4902,
+      "step": 5700
+    },
+    {
+      "epoch": 2.614149753468639,
+      "grad_norm": 0.06309907138347626,
+      "learning_rate": 4.2751760514037506e-05,
+      "loss": 1.1826,
+      "step": 5701
+    },
+    {
+      "epoch": 2.6146084164660017,
+      "grad_norm": 0.39696255326271057,
+      "learning_rate": 4.2651621970028366e-05,
+      "loss": 1.5219,
+      "step": 5702
+    },
+    {
+      "epoch": 2.615067079463364,
+      "grad_norm": 0.4694381058216095,
+      "learning_rate": 4.2551595616874e-05,
+      "loss": 1.8362,
+      "step": 5703
+    },
+    {
+      "epoch": 2.615525742460727,
+      "grad_norm": 0.29091694951057434,
+      "learning_rate": 4.24516814791116e-05,
+      "loss": 1.6509,
+      "step": 5704
+    },
+    {
+      "epoch": 2.6159844054580894,
+      "grad_norm": 0.40303540229797363,
+      "learning_rate": 4.235187958125086e-05,
+      "loss": 1.3851,
+      "step": 5705
+    },
+    {
+      "epoch": 2.6164430684554523,
+      "grad_norm": 0.3498489558696747,
+      "learning_rate": 4.225218994777397e-05,
+      "loss": 1.513,
+      "step": 5706
+    },
+    {
+      "epoch": 2.616901731452815,
+      "grad_norm": 0.3207775354385376,
+      "learning_rate": 4.2152612603135244e-05,
+      "loss": 1.484,
+      "step": 5707
+    },
+    {
+      "epoch": 2.617360394450178,
+      "grad_norm": 0.2937675416469574,
+      "learning_rate": 4.2053147571761985e-05,
+      "loss": 1.2656,
+      "step": 5708
+    },
+    {
+      "epoch": 2.6178190574475404,
+      "grad_norm": 0.3286793529987335,
+      "learning_rate": 4.1953794878053565e-05,
+      "loss": 1.202,
+      "step": 5709
+    },
+    {
+      "epoch": 2.6182777204449033,
+      "grad_norm": 0.22947460412979126,
+      "learning_rate": 4.185455454638193e-05,
+      "loss": 1.2922,
+      "step": 5710
+    },
+    {
+      "epoch": 2.6187363834422657,
+      "grad_norm": 0.22816713154315948,
+      "learning_rate": 4.17554266010915e-05,
+      "loss": 0.9779,
+      "step": 5711
+    },
+    {
+      "epoch": 2.6191950464396285,
+      "grad_norm": 0.34349822998046875,
+      "learning_rate": 4.165641106649898e-05,
+      "loss": 1.5968,
+      "step": 5712
+    },
+    {
+      "epoch": 2.6196537094369914,
+      "grad_norm": 0.3354540169239044,
+      "learning_rate": 4.155750796689373e-05,
+      "loss": 0.8374,
+      "step": 5713
+    },
+    {
+      "epoch": 2.620112372434354,
+      "grad_norm": 0.3276011645793915,
+      "learning_rate": 4.1458717326537276e-05,
+      "loss": 1.5134,
+      "step": 5714
+    },
+    {
+      "epoch": 2.6205710354317167,
+      "grad_norm": 0.3458406925201416,
+      "learning_rate": 4.1360039169663685e-05,
+      "loss": 1.4265,
+      "step": 5715
+    },
+    {
+      "epoch": 2.621029698429079,
+      "grad_norm": 0.3710658550262451,
+      "learning_rate": 4.1261473520479483e-05,
+      "loss": 1.4791,
+      "step": 5716
+    },
+    {
+      "epoch": 2.621488361426442,
+      "grad_norm": 0.3279964029788971,
+      "learning_rate": 4.116302040316361e-05,
+      "loss": 1.191,
+      "step": 5717
+    },
+    {
+      "epoch": 2.621947024423805,
+      "grad_norm": 0.2667905390262604,
+      "learning_rate": 4.106467984186707e-05,
+      "loss": 1.1352,
+      "step": 5718
+    },
+    {
+      "epoch": 2.622405687421167,
+      "grad_norm": 0.3724762201309204,
+      "learning_rate": 4.0966451860713736e-05,
+      "loss": 1.4844,
+      "step": 5719
+    },
+    {
+      "epoch": 2.62286435041853,
+      "grad_norm": 0.3980143070220947,
+      "learning_rate": 4.0868336483799505e-05,
+      "loss": 1.5843,
+      "step": 5720
+    },
+    {
+      "epoch": 2.6233230134158925,
+      "grad_norm": 0.2862567603588104,
+      "learning_rate": 4.07703337351929e-05,
+      "loss": 1.5024,
+      "step": 5721
+    },
+    {
+      "epoch": 2.6237816764132553,
+      "grad_norm": 0.3262464106082916,
+      "learning_rate": 4.067244363893452e-05,
+      "loss": 1.1802,
+      "step": 5722
+    },
+    {
+      "epoch": 2.624240339410618,
+      "grad_norm": 0.3230985999107361,
+      "learning_rate": 4.0574666219037824e-05,
+      "loss": 1.7435,
+      "step": 5723
+    },
+    {
+      "epoch": 2.6246990024079806,
+      "grad_norm": 0.3618335425853729,
+      "learning_rate": 4.047700149948802e-05,
+      "loss": 1.3806,
+      "step": 5724
+    },
+    {
+      "epoch": 2.6251576654053435,
+      "grad_norm": 0.33226069808006287,
+      "learning_rate": 4.037944950424311e-05,
+      "loss": 1.197,
+      "step": 5725
+    },
+    {
+      "epoch": 2.625616328402706,
+      "grad_norm": 0.22331880033016205,
+      "learning_rate": 4.028201025723327e-05,
+      "loss": 1.3807,
+      "step": 5726
+    },
+    {
+      "epoch": 2.6260749914000687,
+      "grad_norm": 0.27581825852394104,
+      "learning_rate": 4.018468378236106e-05,
+      "loss": 1.0424,
+      "step": 5727
+    },
+    {
+      "epoch": 2.6265336543974316,
+      "grad_norm": 0.3891744315624237,
+      "learning_rate": 4.008747010350133e-05,
+      "loss": 1.9856,
+      "step": 5728
+    },
+    {
+      "epoch": 2.6269923173947944,
+      "grad_norm": 0.3294212222099304,
+      "learning_rate": 3.9990369244501445e-05,
+      "loss": 1.2145,
+      "step": 5729
+    },
+    {
+      "epoch": 2.627450980392157,
+      "grad_norm": 0.35058024525642395,
+      "learning_rate": 3.989338122918068e-05,
+      "loss": 1.1814,
+      "step": 5730
+    },
+    {
+      "epoch": 2.6279096433895197,
+      "grad_norm": 0.3072217106819153,
+      "learning_rate": 3.979650608133112e-05,
+      "loss": 1.6852,
+      "step": 5731
+    },
+    {
+      "epoch": 2.628368306386882,
+      "grad_norm": 0.3541480004787445,
+      "learning_rate": 3.969974382471669e-05,
+      "loss": 1.1888,
+      "step": 5732
+    },
+    {
+      "epoch": 2.628826969384245,
+      "grad_norm": 0.26909348368644714,
+      "learning_rate": 3.96030944830742e-05,
+      "loss": 1.2612,
+      "step": 5733
+    },
+    {
+      "epoch": 2.629285632381608,
+      "grad_norm": 0.33335986733436584,
+      "learning_rate": 3.950655808011233e-05,
+      "loss": 1.7208,
+      "step": 5734
+    },
+    {
+      "epoch": 2.6297442953789703,
+      "grad_norm": 0.3109550178050995,
+      "learning_rate": 3.9410134639511986e-05,
+      "loss": 1.0018,
+      "step": 5735
+    },
+    {
+      "epoch": 2.630202958376333,
+      "grad_norm": 0.3982895612716675,
+      "learning_rate": 3.931382418492663e-05,
+      "loss": 1.1909,
+      "step": 5736
+    },
+    {
+      "epoch": 2.6306616213736955,
+      "grad_norm": 0.31409791111946106,
+      "learning_rate": 3.921762673998197e-05,
+      "loss": 1.0483,
+      "step": 5737
+    },
+    {
+      "epoch": 2.6311202843710584,
+      "grad_norm": 0.24062570929527283,
+      "learning_rate": 3.912154232827581e-05,
+      "loss": 0.5005,
+      "step": 5738
+    },
+    {
+      "epoch": 2.6315789473684212,
+      "grad_norm": 0.33497846126556396,
+      "learning_rate": 3.90255709733785e-05,
+      "loss": 1.5151,
+      "step": 5739
+    },
+    {
+      "epoch": 2.6320376103657837,
+      "grad_norm": 0.2941315770149231,
+      "learning_rate": 3.8929712698832445e-05,
+      "loss": 0.8872,
+      "step": 5740
+    },
+    {
+      "epoch": 2.6324962733631465,
+      "grad_norm": 0.23103603720664978,
+      "learning_rate": 3.883396752815221e-05,
+      "loss": 0.6996,
+      "step": 5741
+    },
+    {
+      "epoch": 2.632954936360509,
+      "grad_norm": 0.2940506935119629,
+      "learning_rate": 3.873833548482503e-05,
+      "loss": 1.7249,
+      "step": 5742
+    },
+    {
+      "epoch": 2.633413599357872,
+      "grad_norm": 0.24942685663700104,
+      "learning_rate": 3.864281659231001e-05,
+      "loss": 0.5719,
+      "step": 5743
+    },
+    {
+      "epoch": 2.6338722623552346,
+      "grad_norm": 0.1974136382341385,
+      "learning_rate": 3.8547410874038625e-05,
+      "loss": 1.5941,
+      "step": 5744
+    },
+    {
+      "epoch": 2.634330925352597,
+      "grad_norm": 0.31970781087875366,
+      "learning_rate": 3.845211835341472e-05,
+      "loss": 1.0022,
+      "step": 5745
+    },
+    {
+      "epoch": 2.63478958834996,
+      "grad_norm": 0.32989394664764404,
+      "learning_rate": 3.835693905381399e-05,
+      "loss": 1.4321,
+      "step": 5746
+    },
+    {
+      "epoch": 2.6352482513473223,
+      "grad_norm": 0.3487594723701477,
+      "learning_rate": 3.826187299858474e-05,
+      "loss": 1.421,
+      "step": 5747
+    },
+    {
+      "epoch": 2.635706914344685,
+      "grad_norm": 0.2510157823562622,
+      "learning_rate": 3.816692021104734e-05,
+      "loss": 0.8154,
+      "step": 5748
+    },
+    {
+      "epoch": 2.636165577342048,
+      "grad_norm": 0.10467905551195145,
+      "learning_rate": 3.807208071449442e-05,
+      "loss": 1.0183,
+      "step": 5749
+    },
+    {
+      "epoch": 2.6366242403394105,
+      "grad_norm": 0.34240782260894775,
+      "learning_rate": 3.7977354532190754e-05,
+      "loss": 1.3793,
+      "step": 5750
+    },
+    {
+      "epoch": 2.6370829033367733,
+      "grad_norm": 0.33239027857780457,
+      "learning_rate": 3.788274168737338e-05,
+      "loss": 1.5253,
+      "step": 5751
+    },
+    {
+      "epoch": 2.6375415663341357,
+      "grad_norm": 0.3215150535106659,
+      "learning_rate": 3.77882422032515e-05,
+      "loss": 1.173,
+      "step": 5752
+    },
+    {
+      "epoch": 2.6380002293314986,
+      "grad_norm": 0.3687779903411865,
+      "learning_rate": 3.7693856103006574e-05,
+      "loss": 1.6017,
+      "step": 5753
+    },
+    {
+      "epoch": 2.6384588923288614,
+      "grad_norm": 0.3637235164642334,
+      "learning_rate": 3.759958340979208e-05,
+      "loss": 1.7386,
+      "step": 5754
+    },
+    {
+      "epoch": 2.6389175553262243,
+      "grad_norm": 0.3436271846294403,
+      "learning_rate": 3.750542414673391e-05,
+      "loss": 0.7148,
+      "step": 5755
+    },
+    {
+      "epoch": 2.6393762183235867,
+      "grad_norm": 0.10530371963977814,
+      "learning_rate": 3.7411378336929925e-05,
+      "loss": 1.4356,
+      "step": 5756
+    },
+    {
+      "epoch": 2.6398348813209496,
+      "grad_norm": 0.4608929753303528,
+      "learning_rate": 3.731744600345038e-05,
+      "loss": 2.1367,
+      "step": 5757
+    },
+    {
+      "epoch": 2.640293544318312,
+      "grad_norm": 0.31003305315971375,
+      "learning_rate": 3.722362716933736e-05,
+      "loss": 1.2718,
+      "step": 5758
+    },
+    {
+      "epoch": 2.640752207315675,
+      "grad_norm": 0.34914591908454895,
+      "learning_rate": 3.712992185760533e-05,
+      "loss": 1.3382,
+      "step": 5759
+    },
+    {
+      "epoch": 2.6412108703130377,
+      "grad_norm": 0.4381144344806671,
+      "learning_rate": 3.703633009124102e-05,
+      "loss": 1.5305,
+      "step": 5760
+    },
+    {
+      "epoch": 2.6416695333104,
+      "grad_norm": 0.20248155295848846,
+      "learning_rate": 3.694285189320301e-05,
+      "loss": 1.1224,
+      "step": 5761
+    },
+    {
+      "epoch": 2.642128196307763,
+      "grad_norm": 0.3816075325012207,
+      "learning_rate": 3.684948728642229e-05,
+      "loss": 1.5889,
+      "step": 5762
+    },
+    {
+      "epoch": 2.6425868593051254,
+      "grad_norm": 0.3584928810596466,
+      "learning_rate": 3.675623629380181e-05,
+      "loss": 1.4312,
+      "step": 5763
+    },
+    {
+      "epoch": 2.6430455223024882,
+      "grad_norm": 0.4676406681537628,
+      "learning_rate": 3.6663098938216734e-05,
+      "loss": 1.2666,
+      "step": 5764
+    },
+    {
+      "epoch": 2.643504185299851,
+      "grad_norm": 0.2964949309825897,
+      "learning_rate": 3.657007524251427e-05,
+      "loss": 1.8092,
+      "step": 5765
+    },
+    {
+      "epoch": 2.6439628482972135,
+      "grad_norm": 0.28791379928588867,
+      "learning_rate": 3.6477165229513885e-05,
+      "loss": 0.9804,
+      "step": 5766
+    },
+    {
+      "epoch": 2.6444215112945764,
+      "grad_norm": 0.3603127598762512,
+      "learning_rate": 3.638436892200708e-05,
+      "loss": 1.5343,
+      "step": 5767
+    },
+    {
+      "epoch": 2.644880174291939,
+      "grad_norm": 0.28618577122688293,
+      "learning_rate": 3.629168634275748e-05,
+      "loss": 0.933,
+      "step": 5768
+    },
+    {
+      "epoch": 2.6453388372893016,
+      "grad_norm": 0.31322944164276123,
+      "learning_rate": 3.619911751450061e-05,
+      "loss": 1.6573,
+      "step": 5769
+    },
+    {
+      "epoch": 2.6457975002866645,
+      "grad_norm": 0.3839041590690613,
+      "learning_rate": 3.610666245994443e-05,
+      "loss": 0.8116,
+      "step": 5770
+    },
+    {
+      "epoch": 2.646256163284027,
+      "grad_norm": 0.11088380962610245,
+      "learning_rate": 3.601432120176879e-05,
+      "loss": 0.9275,
+      "step": 5771
+    },
+    {
+      "epoch": 2.6467148262813898,
+      "grad_norm": 0.29426756501197815,
+      "learning_rate": 3.5922093762625674e-05,
+      "loss": 0.9716,
+      "step": 5772
+    },
+    {
+      "epoch": 2.647173489278752,
+      "grad_norm": 0.21911796927452087,
+      "learning_rate": 3.582998016513911e-05,
+      "loss": 1.243,
+      "step": 5773
+    },
+    {
+      "epoch": 2.647632152276115,
+      "grad_norm": 0.40464484691619873,
+      "learning_rate": 3.573798043190524e-05,
+      "loss": 1.4725,
+      "step": 5774
+    },
+    {
+      "epoch": 2.648090815273478,
+      "grad_norm": 0.32122787833213806,
+      "learning_rate": 3.564609458549234e-05,
+      "loss": 1.5024,
+      "step": 5775
+    },
+    {
+      "epoch": 2.6485494782708408,
+      "grad_norm": 0.28623196482658386,
+      "learning_rate": 3.555432264844055e-05,
+      "loss": 1.4161,
+      "step": 5776
+    },
+    {
+      "epoch": 2.649008141268203,
+      "grad_norm": 0.28445377945899963,
+      "learning_rate": 3.5462664643262255e-05,
+      "loss": 1.1441,
+      "step": 5777
+    },
+    {
+      "epoch": 2.649466804265566,
+      "grad_norm": 0.27459508180618286,
+      "learning_rate": 3.53711205924418e-05,
+      "loss": 1.0607,
+      "step": 5778
+    },
+    {
+      "epoch": 2.6499254672629284,
+      "grad_norm": 0.3308379352092743,
+      "learning_rate": 3.5279690518435656e-05,
+      "loss": 0.8458,
+      "step": 5779
+    },
+    {
+      "epoch": 2.6503841302602913,
+      "grad_norm": 0.23437727987766266,
+      "learning_rate": 3.518837444367223e-05,
+      "loss": 1.3428,
+      "step": 5780
+    },
+    {
+      "epoch": 2.650842793257654,
+      "grad_norm": 0.30645015835762024,
+      "learning_rate": 3.509717239055199e-05,
+      "loss": 0.9823,
+      "step": 5781
+    },
+    {
+      "epoch": 2.6513014562550166,
+      "grad_norm": 0.20318692922592163,
+      "learning_rate": 3.500608438144748e-05,
+      "loss": 1.3997,
+      "step": 5782
+    },
+    {
+      "epoch": 2.6517601192523794,
+      "grad_norm": 0.3624488115310669,
+      "learning_rate": 3.491511043870321e-05,
+      "loss": 0.991,
+      "step": 5783
+    },
+    {
+      "epoch": 2.652218782249742,
+      "grad_norm": 0.3760741949081421,
+      "learning_rate": 3.482425058463579e-05,
+      "loss": 1.937,
+      "step": 5784
+    },
+    {
+      "epoch": 2.6526774452471047,
+      "grad_norm": 0.4173126816749573,
+      "learning_rate": 3.473350484153387e-05,
+      "loss": 1.5058,
+      "step": 5785
+    },
+    {
+      "epoch": 2.6531361082444676,
+      "grad_norm": 0.2894061803817749,
+      "learning_rate": 3.4642873231657934e-05,
+      "loss": 1.5995,
+      "step": 5786
+    },
+    {
+      "epoch": 2.65359477124183,
+      "grad_norm": 0.34869733452796936,
+      "learning_rate": 3.4552355777240585e-05,
+      "loss": 1.4747,
+      "step": 5787
+    },
+    {
+      "epoch": 2.654053434239193,
+      "grad_norm": 0.3714545965194702,
+      "learning_rate": 3.446195250048639e-05,
+      "loss": 1.4193,
+      "step": 5788
+    },
+    {
+      "epoch": 2.6545120972365552,
+      "grad_norm": 0.28376656770706177,
+      "learning_rate": 3.437166342357195e-05,
+      "loss": 1.3256,
+      "step": 5789
+    },
+    {
+      "epoch": 2.654970760233918,
+      "grad_norm": 0.37970438599586487,
+      "learning_rate": 3.4281488568645934e-05,
+      "loss": 1.584,
+      "step": 5790
+    },
+    {
+      "epoch": 2.655429423231281,
+      "grad_norm": 0.3399360179901123,
+      "learning_rate": 3.4191427957828705e-05,
+      "loss": 1.4299,
+      "step": 5791
+    },
+    {
+      "epoch": 2.6558880862286434,
+      "grad_norm": 0.4417398273944855,
+      "learning_rate": 3.41014816132128e-05,
+      "loss": 1.6428,
+      "step": 5792
+    },
+    {
+      "epoch": 2.656346749226006,
+      "grad_norm": 0.3878077268600464,
+      "learning_rate": 3.401164955686281e-05,
+      "loss": 1.587,
+      "step": 5793
+    },
+    {
+      "epoch": 2.6568054122233686,
+      "grad_norm": 0.5019458532333374,
+      "learning_rate": 3.392193181081504e-05,
+      "loss": 1.2179,
+      "step": 5794
+    },
+    {
+      "epoch": 2.6572640752207315,
+      "grad_norm": 0.3489948809146881,
+      "learning_rate": 3.383232839707806e-05,
+      "loss": 1.608,
+      "step": 5795
+    },
+    {
+      "epoch": 2.6577227382180943,
+      "grad_norm": 0.43875566124916077,
+      "learning_rate": 3.3742839337632223e-05,
+      "loss": 1.3773,
+      "step": 5796
+    },
+    {
+      "epoch": 2.658181401215457,
+      "grad_norm": 0.2078205943107605,
+      "learning_rate": 3.365346465442976e-05,
+      "loss": 1.2278,
+      "step": 5797
+    },
+    {
+      "epoch": 2.6586400642128196,
+      "grad_norm": 0.36126843094825745,
+      "learning_rate": 3.356420436939489e-05,
+      "loss": 1.799,
+      "step": 5798
+    },
+    {
+      "epoch": 2.6590987272101825,
+      "grad_norm": 0.40285730361938477,
+      "learning_rate": 3.347505850442395e-05,
+      "loss": 1.5258,
+      "step": 5799
+    },
+    {
+      "epoch": 2.659557390207545,
+      "grad_norm": 0.29165950417518616,
+      "learning_rate": 3.3386027081384886e-05,
+      "loss": 1.3616,
+      "step": 5800
+    },
+    {
+      "epoch": 2.6600160532049077,
+      "grad_norm": 0.2867632210254669,
+      "learning_rate": 3.3297110122118005e-05,
+      "loss": 1.2755,
+      "step": 5801
+    },
+    {
+      "epoch": 2.6604747162022706,
+      "grad_norm": 0.33592167496681213,
+      "learning_rate": 3.3208307648434964e-05,
+      "loss": 1.5153,
+      "step": 5802
+    },
+    {
+      "epoch": 2.660933379199633,
+      "grad_norm": 0.2837589979171753,
+      "learning_rate": 3.311961968211979e-05,
+      "loss": 1.204,
+      "step": 5803
+    },
+    {
+      "epoch": 2.661392042196996,
+      "grad_norm": 0.2693648040294647,
+      "learning_rate": 3.303104624492825e-05,
+      "loss": 1.2056,
+      "step": 5804
+    },
+    {
+      "epoch": 2.6618507051943583,
+      "grad_norm": 0.206869438290596,
+      "learning_rate": 3.294258735858818e-05,
+      "loss": 0.9919,
+      "step": 5805
+    },
+    {
+      "epoch": 2.662309368191721,
+      "grad_norm": 0.3209756314754486,
+      "learning_rate": 3.2854243044799056e-05,
+      "loss": 1.6161,
+      "step": 5806
+    },
+    {
+      "epoch": 2.662768031189084,
+      "grad_norm": 0.301139235496521,
+      "learning_rate": 3.276601332523249e-05,
+      "loss": 1.2684,
+      "step": 5807
+    },
+    {
+      "epoch": 2.6632266941864464,
+      "grad_norm": 0.2972935140132904,
+      "learning_rate": 3.267789822153172e-05,
+      "loss": 1.5424,
+      "step": 5808
+    },
+    {
+      "epoch": 2.6636853571838093,
+      "grad_norm": 0.3825678527355194,
+      "learning_rate": 3.258989775531212e-05,
+      "loss": 1.5092,
+      "step": 5809
+    },
+    {
+      "epoch": 2.6641440201811717,
+      "grad_norm": 0.22569845616817474,
+      "learning_rate": 3.250201194816077e-05,
+      "loss": 1.0415,
+      "step": 5810
+    },
+    {
+      "epoch": 2.6646026831785345,
+      "grad_norm": 0.2539108991622925,
+      "learning_rate": 3.2414240821636685e-05,
+      "loss": 1.3038,
+      "step": 5811
+    },
+    {
+      "epoch": 2.6650613461758974,
+      "grad_norm": 0.3397371470928192,
+      "learning_rate": 3.232658439727082e-05,
+      "loss": 1.1019,
+      "step": 5812
+    },
+    {
+      "epoch": 2.66552000917326,
+      "grad_norm": 0.35667502880096436,
+      "learning_rate": 3.2239042696565915e-05,
+      "loss": 1.9956,
+      "step": 5813
+    },
+    {
+      "epoch": 2.6659786721706227,
+      "grad_norm": 0.37884020805358887,
+      "learning_rate": 3.2151615740996565e-05,
+      "loss": 1.2626,
+      "step": 5814
+    },
+    {
+      "epoch": 2.666437335167985,
+      "grad_norm": 0.38707906007766724,
+      "learning_rate": 3.2064303552009236e-05,
+      "loss": 1.2571,
+      "step": 5815
+    },
+    {
+      "epoch": 2.666895998165348,
+      "grad_norm": 0.3102846145629883,
+      "learning_rate": 3.1977106151022226e-05,
+      "loss": 1.2002,
+      "step": 5816
+    },
+    {
+      "epoch": 2.667354661162711,
+      "grad_norm": 0.28075534105300903,
+      "learning_rate": 3.189002355942572e-05,
+      "loss": 1.3118,
+      "step": 5817
+    },
+    {
+      "epoch": 2.6678133241600737,
+      "grad_norm": 0.2957073748111725,
+      "learning_rate": 3.1803055798581725e-05,
+      "loss": 1.1588,
+      "step": 5818
+    },
+    {
+      "epoch": 2.668271987157436,
+      "grad_norm": 0.24751010537147522,
+      "learning_rate": 3.1716202889823966e-05,
+      "loss": 0.9748,
+      "step": 5819
+    },
+    {
+      "epoch": 2.6687306501547985,
+      "grad_norm": 0.301859587430954,
+      "learning_rate": 3.162946485445817e-05,
+      "loss": 1.3465,
+      "step": 5820
+    },
+    {
+      "epoch": 2.6691893131521613,
+      "grad_norm": 0.39302027225494385,
+      "learning_rate": 3.15428417137617e-05,
+      "loss": 1.4904,
+      "step": 5821
+    },
+    {
+      "epoch": 2.669647976149524,
+      "grad_norm": 0.27386486530303955,
+      "learning_rate": 3.145633348898397e-05,
+      "loss": 1.6713,
+      "step": 5822
+    },
+    {
+      "epoch": 2.670106639146887,
+      "grad_norm": 0.33279913663864136,
+      "learning_rate": 3.136994020134598e-05,
+      "loss": 1.2917,
+      "step": 5823
+    },
+    {
+      "epoch": 2.6705653021442495,
+      "grad_norm": 0.3797074258327484,
+      "learning_rate": 3.1283661872040626e-05,
+      "loss": 1.4031,
+      "step": 5824
+    },
+    {
+      "epoch": 2.6710239651416123,
+      "grad_norm": 0.3788340091705322,
+      "learning_rate": 3.119749852223269e-05,
+      "loss": 1.2273,
+      "step": 5825
+    },
+    {
+      "epoch": 2.6714826281389747,
+      "grad_norm": 0.2667097747325897,
+      "learning_rate": 3.1111450173058553e-05,
+      "loss": 0.9149,
+      "step": 5826
+    },
+    {
+      "epoch": 2.6719412911363376,
+      "grad_norm": 0.3300883173942566,
+      "learning_rate": 3.10255168456266e-05,
+      "loss": 1.7291,
+      "step": 5827
+    },
+    {
+      "epoch": 2.6723999541337005,
+      "grad_norm": 0.33332276344299316,
+      "learning_rate": 3.093969856101686e-05,
+      "loss": 1.4482,
+      "step": 5828
+    },
+    {
+      "epoch": 2.672858617131063,
+      "grad_norm": 0.27086135745048523,
+      "learning_rate": 3.085399534028116e-05,
+      "loss": 0.9669,
+      "step": 5829
+    },
+    {
+      "epoch": 2.6733172801284257,
+      "grad_norm": 0.10022089630365372,
+      "learning_rate": 3.0768407204443126e-05,
+      "loss": 0.5217,
+      "step": 5830
+    },
+    {
+      "epoch": 2.673775943125788,
+      "grad_norm": 0.1253892332315445,
+      "learning_rate": 3.068293417449808e-05,
+      "loss": 0.8534,
+      "step": 5831
+    },
+    {
+      "epoch": 2.674234606123151,
+      "grad_norm": 0.3180748224258423,
+      "learning_rate": 3.05975762714133e-05,
+      "loss": 1.609,
+      "step": 5832
+    },
+    {
+      "epoch": 2.674693269120514,
+      "grad_norm": 0.3440166413784027,
+      "learning_rate": 3.051233351612759e-05,
+      "loss": 0.7612,
+      "step": 5833
+    },
+    {
+      "epoch": 2.6751519321178763,
+      "grad_norm": 0.3188319802284241,
+      "learning_rate": 3.0427205929551615e-05,
+      "loss": 2.0381,
+      "step": 5834
+    },
+    {
+      "epoch": 2.675610595115239,
+      "grad_norm": 0.30568626523017883,
+      "learning_rate": 3.0342193532567842e-05,
+      "loss": 0.9399,
+      "step": 5835
+    },
+    {
+      "epoch": 2.6760692581126015,
+      "grad_norm": 0.373977392911911,
+      "learning_rate": 3.0257296346030416e-05,
+      "loss": 1.6466,
+      "step": 5836
+    },
+    {
+      "epoch": 2.6765279211099644,
+      "grad_norm": 0.372895210981369,
+      "learning_rate": 3.0172514390765238e-05,
+      "loss": 1.7245,
+      "step": 5837
+    },
+    {
+      "epoch": 2.6769865841073273,
+      "grad_norm": 0.37782806158065796,
+      "learning_rate": 3.0087847687569893e-05,
+      "loss": 1.6417,
+      "step": 5838
+    },
+    {
+      "epoch": 2.6774452471046897,
+      "grad_norm": 0.3993247449398041,
+      "learning_rate": 3.0003296257213708e-05,
+      "loss": 1.9526,
+      "step": 5839
+    },
+    {
+      "epoch": 2.6779039101020525,
+      "grad_norm": 0.3877170979976654,
+      "learning_rate": 2.9918860120437873e-05,
+      "loss": 1.5638,
+      "step": 5840
+    },
+    {
+      "epoch": 2.678362573099415,
+      "grad_norm": 0.38056743144989014,
+      "learning_rate": 2.9834539297955156e-05,
+      "loss": 1.2691,
+      "step": 5841
+    },
+    {
+      "epoch": 2.678821236096778,
+      "grad_norm": 0.3735964298248291,
+      "learning_rate": 2.975033381044995e-05,
+      "loss": 2.0476,
+      "step": 5842
+    },
+    {
+      "epoch": 2.6792798990941407,
+      "grad_norm": 0.37376493215560913,
+      "learning_rate": 2.966624367857851e-05,
+      "loss": 1.5699,
+      "step": 5843
+    },
+    {
+      "epoch": 2.6797385620915035,
+      "grad_norm": 0.3251396119594574,
+      "learning_rate": 2.958226892296878e-05,
+      "loss": 0.6622,
+      "step": 5844
+    },
+    {
+      "epoch": 2.680197225088866,
+      "grad_norm": 0.20608735084533691,
+      "learning_rate": 2.9498409564220396e-05,
+      "loss": 0.7508,
+      "step": 5845
+    },
+    {
+      "epoch": 2.680655888086229,
+      "grad_norm": 0.2004690021276474,
+      "learning_rate": 2.9414665622904623e-05,
+      "loss": 0.8976,
+      "step": 5846
+    },
+    {
+      "epoch": 2.681114551083591,
+      "grad_norm": 0.30501410365104675,
+      "learning_rate": 2.9331037119564473e-05,
+      "loss": 1.0064,
+      "step": 5847
+    },
+    {
+      "epoch": 2.681573214080954,
+      "grad_norm": 0.2096373289823532,
+      "learning_rate": 2.9247524074714594e-05,
+      "loss": 1.6792,
+      "step": 5848
+    },
+    {
+      "epoch": 2.682031877078317,
+      "grad_norm": 0.4180261790752411,
+      "learning_rate": 2.9164126508841428e-05,
+      "loss": 1.6902,
+      "step": 5849
+    },
+    {
+      "epoch": 2.6824905400756793,
+      "grad_norm": 0.2524012327194214,
+      "learning_rate": 2.9080844442402887e-05,
+      "loss": 0.4348,
+      "step": 5850
+    },
+    {
+      "epoch": 2.682949203073042,
+      "grad_norm": 0.2333446592092514,
+      "learning_rate": 2.8997677895828688e-05,
+      "loss": 1.4499,
+      "step": 5851
+    },
+    {
+      "epoch": 2.6834078660704046,
+      "grad_norm": 0.4264712631702423,
+      "learning_rate": 2.891462688952029e-05,
+      "loss": 1.3736,
+      "step": 5852
+    },
+    {
+      "epoch": 2.6838665290677675,
+      "grad_norm": 0.3079976737499237,
+      "learning_rate": 2.883169144385056e-05,
+      "loss": 1.2166,
+      "step": 5853
+    },
+    {
+      "epoch": 2.6843251920651303,
+      "grad_norm": 0.18179012835025787,
+      "learning_rate": 2.874887157916417e-05,
+      "loss": 0.6641,
+      "step": 5854
+    },
+    {
+      "epoch": 2.6847838550624927,
+      "grad_norm": 0.32719552516937256,
+      "learning_rate": 2.8666167315777535e-05,
+      "loss": 1.2209,
+      "step": 5855
+    },
+    {
+      "epoch": 2.6852425180598556,
+      "grad_norm": 0.30747801065444946,
+      "learning_rate": 2.8583578673978483e-05,
+      "loss": 1.7491,
+      "step": 5856
+    },
+    {
+      "epoch": 2.685701181057218,
+      "grad_norm": 0.39954906702041626,
+      "learning_rate": 2.8501105674026808e-05,
+      "loss": 1.2025,
+      "step": 5857
+    },
+    {
+      "epoch": 2.686159844054581,
+      "grad_norm": 0.29626455903053284,
+      "learning_rate": 2.841874833615349e-05,
+      "loss": 1.8599,
+      "step": 5858
+    },
+    {
+      "epoch": 2.6866185070519437,
+      "grad_norm": 0.44469526410102844,
+      "learning_rate": 2.833650668056148e-05,
+      "loss": 1.537,
+      "step": 5859
+    },
+    {
+      "epoch": 2.687077170049306,
+      "grad_norm": 0.325447142124176,
+      "learning_rate": 2.8254380727425255e-05,
+      "loss": 1.2097,
+      "step": 5860
+    },
+    {
+      "epoch": 2.687535833046669,
+      "grad_norm": 0.32134348154067993,
+      "learning_rate": 2.817237049689092e-05,
+      "loss": 1.6572,
+      "step": 5861
+    },
+    {
+      "epoch": 2.6879944960440314,
+      "grad_norm": 0.2648102045059204,
+      "learning_rate": 2.809047600907616e-05,
+      "loss": 0.9762,
+      "step": 5862
+    },
+    {
+      "epoch": 2.6884531590413943,
+      "grad_norm": 0.3223074972629547,
+      "learning_rate": 2.800869728407035e-05,
+      "loss": 1.5692,
+      "step": 5863
+    },
+    {
+      "epoch": 2.688911822038757,
+      "grad_norm": 0.2671777307987213,
+      "learning_rate": 2.792703434193422e-05,
+      "loss": 1.4353,
+      "step": 5864
+    },
+    {
+      "epoch": 2.68937048503612,
+      "grad_norm": 0.3055770695209503,
+      "learning_rate": 2.7845487202700416e-05,
+      "loss": 0.8293,
+      "step": 5865
+    },
+    {
+      "epoch": 2.6898291480334824,
+      "grad_norm": 0.2750590145587921,
+      "learning_rate": 2.7764055886372987e-05,
+      "loss": 1.4748,
+      "step": 5866
+    },
+    {
+      "epoch": 2.6902878110308452,
+      "grad_norm": 0.3781544268131256,
+      "learning_rate": 2.768274041292762e-05,
+      "loss": 0.7972,
+      "step": 5867
+    },
+    {
+      "epoch": 2.6907464740282077,
+      "grad_norm": 0.18787944316864014,
+      "learning_rate": 2.760154080231175e-05,
+      "loss": 1.6368,
+      "step": 5868
+    },
+    {
+      "epoch": 2.6912051370255705,
+      "grad_norm": 0.41110900044441223,
+      "learning_rate": 2.7520457074444005e-05,
+      "loss": 1.1579,
+      "step": 5869
+    },
+    {
+      "epoch": 2.6916638000229334,
+      "grad_norm": 0.28122058510780334,
+      "learning_rate": 2.743948924921491e-05,
+      "loss": 0.9766,
+      "step": 5870
+    },
+    {
+      "epoch": 2.692122463020296,
+      "grad_norm": 0.1648864597082138,
+      "learning_rate": 2.7358637346486414e-05,
+      "loss": 1.1961,
+      "step": 5871
+    },
+    {
+      "epoch": 2.6925811260176586,
+      "grad_norm": 0.45734626054763794,
+      "learning_rate": 2.7277901386092096e-05,
+      "loss": 1.5893,
+      "step": 5872
+    },
+    {
+      "epoch": 2.693039789015021,
+      "grad_norm": 0.31361323595046997,
+      "learning_rate": 2.7197281387837114e-05,
+      "loss": 0.8256,
+      "step": 5873
+    },
+    {
+      "epoch": 2.693498452012384,
+      "grad_norm": 0.22070448100566864,
+      "learning_rate": 2.7116777371498145e-05,
+      "loss": 1.4343,
+      "step": 5874
+    },
+    {
+      "epoch": 2.6939571150097468,
+      "grad_norm": 0.428138792514801,
+      "learning_rate": 2.703638935682323e-05,
+      "loss": 1.6013,
+      "step": 5875
+    },
+    {
+      "epoch": 2.694415778007109,
+      "grad_norm": 0.3450068235397339,
+      "learning_rate": 2.6956117363532207e-05,
+      "loss": 1.2018,
+      "step": 5876
+    },
+    {
+      "epoch": 2.694874441004472,
+      "grad_norm": 0.34051015973091125,
+      "learning_rate": 2.687596141131654e-05,
+      "loss": 1.6506,
+      "step": 5877
+    },
+    {
+      "epoch": 2.6953331040018345,
+      "grad_norm": 0.31475841999053955,
+      "learning_rate": 2.6795921519838895e-05,
+      "loss": 1.5417,
+      "step": 5878
+    },
+    {
+      "epoch": 2.6957917669991973,
+      "grad_norm": 0.216132253408432,
+      "learning_rate": 2.6715997708733675e-05,
+      "loss": 0.3737,
+      "step": 5879
+    },
+    {
+      "epoch": 2.69625042999656,
+      "grad_norm": 0.2257094830274582,
+      "learning_rate": 2.6636189997606864e-05,
+      "loss": 1.4002,
+      "step": 5880
+    },
+    {
+      "epoch": 2.6967090929939226,
+      "grad_norm": 0.3286268413066864,
+      "learning_rate": 2.655649840603569e-05,
+      "loss": 1.5927,
+      "step": 5881
+    },
+    {
+      "epoch": 2.6971677559912854,
+      "grad_norm": 0.32692962884902954,
+      "learning_rate": 2.6476922953569127e-05,
+      "loss": 0.7494,
+      "step": 5882
+    },
+    {
+      "epoch": 2.697626418988648,
+      "grad_norm": 0.29675719141960144,
+      "learning_rate": 2.6397463659727672e-05,
+      "loss": 1.6597,
+      "step": 5883
+    },
+    {
+      "epoch": 2.6980850819860107,
+      "grad_norm": 0.3485512137413025,
+      "learning_rate": 2.6318120544003234e-05,
+      "loss": 1.2645,
+      "step": 5884
+    },
+    {
+      "epoch": 2.6985437449833736,
+      "grad_norm": 0.41130709648132324,
+      "learning_rate": 2.623889362585924e-05,
+      "loss": 1.9763,
+      "step": 5885
+    },
+    {
+      "epoch": 2.6990024079807364,
+      "grad_norm": 0.2589356005191803,
+      "learning_rate": 2.6159782924730647e-05,
+      "loss": 0.3383,
+      "step": 5886
+    },
+    {
+      "epoch": 2.699461070978099,
+      "grad_norm": 0.15324239432811737,
+      "learning_rate": 2.6080788460023875e-05,
+      "loss": 1.1049,
+      "step": 5887
+    },
+    {
+      "epoch": 2.6999197339754613,
+      "grad_norm": 0.2607985734939575,
+      "learning_rate": 2.6001910251116812e-05,
+      "loss": 1.1836,
+      "step": 5888
+    },
+    {
+      "epoch": 2.700378396972824,
+      "grad_norm": 0.29069429636001587,
+      "learning_rate": 2.5923148317358926e-05,
+      "loss": 0.8517,
+      "step": 5889
+    },
+    {
+      "epoch": 2.700837059970187,
+      "grad_norm": 0.2796778976917267,
+      "learning_rate": 2.584450267807098e-05,
+      "loss": 0.8593,
+      "step": 5890
+    },
+    {
+      "epoch": 2.70129572296755,
+      "grad_norm": 0.2640216648578644,
+      "learning_rate": 2.5765973352545436e-05,
+      "loss": 1.5907,
+      "step": 5891
+    },
+    {
+      "epoch": 2.7017543859649122,
+      "grad_norm": 0.39561182260513306,
+      "learning_rate": 2.5687560360045935e-05,
+      "loss": 1.3107,
+      "step": 5892
+    },
+    {
+      "epoch": 2.702213048962275,
+      "grad_norm": 0.3599591553211212,
+      "learning_rate": 2.5609263719807875e-05,
+      "loss": 1.8921,
+      "step": 5893
+    },
+    {
+      "epoch": 2.7026717119596375,
+      "grad_norm": 0.37715888023376465,
+      "learning_rate": 2.553108345103794e-05,
+      "loss": 1.3452,
+      "step": 5894
+    },
+    {
+      "epoch": 2.7031303749570004,
+      "grad_norm": 0.3551565110683441,
+      "learning_rate": 2.545301957291435e-05,
+      "loss": 0.9477,
+      "step": 5895
+    },
+    {
+      "epoch": 2.7035890379543632,
+      "grad_norm": 0.12380225211381912,
+      "learning_rate": 2.5375072104586726e-05,
+      "loss": 0.5937,
+      "step": 5896
+    },
+    {
+      "epoch": 2.7040477009517256,
+      "grad_norm": 0.21938225626945496,
+      "learning_rate": 2.5297241065176168e-05,
+      "loss": 1.0974,
+      "step": 5897
+    },
+    {
+      "epoch": 2.7045063639490885,
+      "grad_norm": 0.3853704333305359,
+      "learning_rate": 2.5219526473775122e-05,
+      "loss": 1.6407,
+      "step": 5898
+    },
+    {
+      "epoch": 2.704965026946451,
+      "grad_norm": 0.2925291657447815,
+      "learning_rate": 2.5141928349447563e-05,
+      "loss": 1.2876,
+      "step": 5899
+    },
+    {
+      "epoch": 2.7054236899438138,
+      "grad_norm": 0.21706360578536987,
+      "learning_rate": 2.5064446711228872e-05,
+      "loss": 0.7466,
+      "step": 5900
+    },
+    {
+      "epoch": 2.7058823529411766,
+      "grad_norm": 0.32736656069755554,
+      "learning_rate": 2.4987081578125904e-05,
+      "loss": 2.0911,
+      "step": 5901
+    },
+    {
+      "epoch": 2.706341015938539,
+      "grad_norm": 0.4860953390598297,
+      "learning_rate": 2.4909832969116918e-05,
+      "loss": 1.9189,
+      "step": 5902
+    },
+    {
+      "epoch": 2.706799678935902,
+      "grad_norm": 0.3114350736141205,
+      "learning_rate": 2.4832700903151374e-05,
+      "loss": 0.5055,
+      "step": 5903
+    },
+    {
+      "epoch": 2.7072583419332643,
+      "grad_norm": 0.12829411029815674,
+      "learning_rate": 2.4755685399150463e-05,
+      "loss": 0.6061,
+      "step": 5904
+    },
+    {
+      "epoch": 2.707717004930627,
+      "grad_norm": 0.3164483308792114,
+      "learning_rate": 2.467878647600663e-05,
+      "loss": 1.7654,
+      "step": 5905
+    },
+    {
+      "epoch": 2.70817566792799,
+      "grad_norm": 0.37488433718681335,
+      "learning_rate": 2.460200415258368e-05,
+      "loss": 1.6296,
+      "step": 5906
+    },
+    {
+      "epoch": 2.7086343309253524,
+      "grad_norm": 0.28757867217063904,
+      "learning_rate": 2.4525338447716928e-05,
+      "loss": 1.1265,
+      "step": 5907
+    },
+    {
+      "epoch": 2.7090929939227153,
+      "grad_norm": 0.32620272040367126,
+      "learning_rate": 2.4448789380213e-05,
+      "loss": 1.1835,
+      "step": 5908
+    },
+    {
+      "epoch": 2.7095516569200777,
+      "grad_norm": 0.293837308883667,
+      "learning_rate": 2.437235696884993e-05,
+      "loss": 1.1319,
+      "step": 5909
+    },
+    {
+      "epoch": 2.7100103199174406,
+      "grad_norm": 0.1874661147594452,
+      "learning_rate": 2.4296041232377165e-05,
+      "loss": 0.5175,
+      "step": 5910
+    },
+    {
+      "epoch": 2.7104689829148034,
+      "grad_norm": 0.2423352599143982,
+      "learning_rate": 2.421984218951545e-05,
+      "loss": 1.2948,
+      "step": 5911
+    },
+    {
+      "epoch": 2.7109276459121663,
+      "grad_norm": 0.32054775953292847,
+      "learning_rate": 2.414375985895706e-05,
+      "loss": 1.2799,
+      "step": 5912
+    },
+    {
+      "epoch": 2.7113863089095287,
+      "grad_norm": 0.3801712989807129,
+      "learning_rate": 2.4067794259365504e-05,
+      "loss": 1.63,
+      "step": 5913
+    },
+    {
+      "epoch": 2.7118449719068916,
+      "grad_norm": 0.29861852526664734,
+      "learning_rate": 2.3991945409375604e-05,
+      "loss": 1.3125,
+      "step": 5914
+    },
+    {
+      "epoch": 2.712303634904254,
+      "grad_norm": 0.3143099248409271,
+      "learning_rate": 2.3916213327593694e-05,
+      "loss": 0.8409,
+      "step": 5915
+    },
+    {
+      "epoch": 2.712762297901617,
+      "grad_norm": 0.09029946476221085,
+      "learning_rate": 2.3840598032597417e-05,
+      "loss": 0.7455,
+      "step": 5916
+    },
+    {
+      "epoch": 2.7132209608989797,
+      "grad_norm": 0.4967743158340454,
+      "learning_rate": 2.376509954293571e-05,
+      "loss": 1.5552,
+      "step": 5917
+    },
+    {
+      "epoch": 2.713679623896342,
+      "grad_norm": 0.37226957082748413,
+      "learning_rate": 2.3689717877128815e-05,
+      "loss": 1.7788,
+      "step": 5918
+    },
+    {
+      "epoch": 2.714138286893705,
+      "grad_norm": 0.41915664076805115,
+      "learning_rate": 2.3614453053668716e-05,
+      "loss": 1.508,
+      "step": 5919
+    },
+    {
+      "epoch": 2.7145969498910674,
+      "grad_norm": 0.30586564540863037,
+      "learning_rate": 2.3539305091018038e-05,
+      "loss": 1.7762,
+      "step": 5920
+    },
+    {
+      "epoch": 2.7150556128884302,
+      "grad_norm": 0.4077855944633484,
+      "learning_rate": 2.3464274007611364e-05,
+      "loss": 1.6046,
+      "step": 5921
+    },
+    {
+      "epoch": 2.715514275885793,
+      "grad_norm": 4.286491870880127,
+      "learning_rate": 2.338935982185425e-05,
+      "loss": 1.3848,
+      "step": 5922
+    },
+    {
+      "epoch": 2.7159729388831555,
+      "grad_norm": 0.2541790306568146,
+      "learning_rate": 2.3314562552123663e-05,
+      "loss": 1.7035,
+      "step": 5923
+    },
+    {
+      "epoch": 2.7164316018805184,
+      "grad_norm": 0.4110478162765503,
+      "learning_rate": 2.3239882216768093e-05,
+      "loss": 2.1335,
+      "step": 5924
+    },
+    {
+      "epoch": 2.7168902648778808,
+      "grad_norm": 0.4424358606338501,
+      "learning_rate": 2.3165318834106942e-05,
+      "loss": 1.9265,
+      "step": 5925
+    },
+    {
+      "epoch": 2.7173489278752436,
+      "grad_norm": 0.3610834777355194,
+      "learning_rate": 2.309087242243124e-05,
+      "loss": 1.1285,
+      "step": 5926
+    },
+    {
+      "epoch": 2.7178075908726065,
+      "grad_norm": 0.2572648525238037,
+      "learning_rate": 2.3016543000003222e-05,
+      "loss": 0.8428,
+      "step": 5927
+    },
+    {
+      "epoch": 2.718266253869969,
+      "grad_norm": 0.36045536398887634,
+      "learning_rate": 2.2942330585056347e-05,
+      "loss": 1.8323,
+      "step": 5928
+    },
+    {
+      "epoch": 2.7187249168673318,
+      "grad_norm": 0.21581125259399414,
+      "learning_rate": 2.2868235195795672e-05,
+      "loss": 1.3904,
+      "step": 5929
+    },
+    {
+      "epoch": 2.719183579864694,
+      "grad_norm": 0.38794851303100586,
+      "learning_rate": 2.279425685039721e-05,
+      "loss": 1.0128,
+      "step": 5930
+    },
+    {
+      "epoch": 2.719642242862057,
+      "grad_norm": 0.3242373466491699,
+      "learning_rate": 2.2720395567008334e-05,
+      "loss": 1.4001,
+      "step": 5931
+    },
+    {
+      "epoch": 2.72010090585942,
+      "grad_norm": 0.130849227309227,
+      "learning_rate": 2.2646651363747773e-05,
+      "loss": 1.1965,
+      "step": 5932
+    },
+    {
+      "epoch": 2.7205595688567827,
+      "grad_norm": 0.3577413856983185,
+      "learning_rate": 2.2573024258705554e-05,
+      "loss": 0.9914,
+      "step": 5933
+    },
+    {
+      "epoch": 2.721018231854145,
+      "grad_norm": 0.2112027406692505,
+      "learning_rate": 2.24995142699429e-05,
+      "loss": 1.7675,
+      "step": 5934
+    },
+    {
+      "epoch": 2.721476894851508,
+      "grad_norm": 0.28808677196502686,
+      "learning_rate": 2.242612141549233e-05,
+      "loss": 0.3857,
+      "step": 5935
+    },
+    {
+      "epoch": 2.7219355578488704,
+      "grad_norm": 0.2923356294631958,
+      "learning_rate": 2.2352845713357772e-05,
+      "loss": 1.5071,
+      "step": 5936
+    },
+    {
+      "epoch": 2.7223942208462333,
+      "grad_norm": 0.24712756276130676,
+      "learning_rate": 2.2279687181514076e-05,
+      "loss": 1.0919,
+      "step": 5937
+    },
+    {
+      "epoch": 2.722852883843596,
+      "grad_norm": 0.3377127945423126,
+      "learning_rate": 2.2206645837907602e-05,
+      "loss": 1.8531,
+      "step": 5938
+    },
+    {
+      "epoch": 2.7233115468409586,
+      "grad_norm": 0.5375214219093323,
+      "learning_rate": 2.2133721700456022e-05,
+      "loss": 1.3341,
+      "step": 5939
+    },
+    {
+      "epoch": 2.7237702098383214,
+      "grad_norm": 0.37174880504608154,
+      "learning_rate": 2.2060914787048136e-05,
+      "loss": 1.8732,
+      "step": 5940
+    },
+    {
+      "epoch": 2.724228872835684,
+      "grad_norm": 0.319526344537735,
+      "learning_rate": 2.198822511554399e-05,
+      "loss": 0.9168,
+      "step": 5941
+    },
+    {
+      "epoch": 2.7246875358330467,
+      "grad_norm": 0.32301315665245056,
+      "learning_rate": 2.1915652703774824e-05,
+      "loss": 1.5364,
+      "step": 5942
+    },
+    {
+      "epoch": 2.7251461988304095,
+      "grad_norm": 0.2971673309803009,
+      "learning_rate": 2.184319756954323e-05,
+      "loss": 0.6956,
+      "step": 5943
+    },
+    {
+      "epoch": 2.725604861827772,
+      "grad_norm": 0.1800968199968338,
+      "learning_rate": 2.177085973062293e-05,
+      "loss": 0.9738,
+      "step": 5944
+    },
+    {
+      "epoch": 2.726063524825135,
+      "grad_norm": 0.3155420124530792,
+      "learning_rate": 2.1698639204759006e-05,
+      "loss": 1.2188,
+      "step": 5945
+    },
+    {
+      "epoch": 2.7265221878224972,
+      "grad_norm": 0.2110133171081543,
+      "learning_rate": 2.1626536009667575e-05,
+      "loss": 1.0003,
+      "step": 5946
+    },
+    {
+      "epoch": 2.72698085081986,
+      "grad_norm": 0.45827966928482056,
+      "learning_rate": 2.1554550163036145e-05,
+      "loss": 1.7184,
+      "step": 5947
+    },
+    {
+      "epoch": 2.727439513817223,
+      "grad_norm": 0.25445568561553955,
+      "learning_rate": 2.1482681682523263e-05,
+      "loss": 0.6922,
+      "step": 5948
+    },
+    {
+      "epoch": 2.7278981768145854,
+      "grad_norm": 0.3437955379486084,
+      "learning_rate": 2.141093058575888e-05,
+      "loss": 1.2655,
+      "step": 5949
+    },
+    {
+      "epoch": 2.728356839811948,
+      "grad_norm": 0.12270909547805786,
+      "learning_rate": 2.133929689034403e-05,
+      "loss": 0.6745,
+      "step": 5950
+    },
+    {
+      "epoch": 2.7288155028093106,
+      "grad_norm": 0.31163090467453003,
+      "learning_rate": 2.1267780613850986e-05,
+      "loss": 2.0465,
+      "step": 5951
+    },
+    {
+      "epoch": 2.7292741658066735,
+      "grad_norm": 0.2993125021457672,
+      "learning_rate": 2.119638177382327e-05,
+      "loss": 1.076,
+      "step": 5952
+    },
+    {
+      "epoch": 2.7297328288040363,
+      "grad_norm": 0.3309202790260315,
+      "learning_rate": 2.1125100387775375e-05,
+      "loss": 0.924,
+      "step": 5953
+    },
+    {
+      "epoch": 2.730191491801399,
+      "grad_norm": 0.35655951499938965,
+      "learning_rate": 2.105393647319326e-05,
+      "loss": 1.8235,
+      "step": 5954
+    },
+    {
+      "epoch": 2.7306501547987616,
+      "grad_norm": 0.3149603307247162,
+      "learning_rate": 2.0982890047533898e-05,
+      "loss": 1.5205,
+      "step": 5955
+    },
+    {
+      "epoch": 2.731108817796124,
+      "grad_norm": 0.43384552001953125,
+      "learning_rate": 2.0911961128225466e-05,
+      "loss": 1.6622,
+      "step": 5956
+    },
+    {
+      "epoch": 2.731567480793487,
+      "grad_norm": 0.30674198269844055,
+      "learning_rate": 2.0841149732667375e-05,
+      "loss": 1.2803,
+      "step": 5957
+    },
+    {
+      "epoch": 2.7320261437908497,
+      "grad_norm": 0.2231689691543579,
+      "learning_rate": 2.0770455878230178e-05,
+      "loss": 0.6584,
+      "step": 5958
+    },
+    {
+      "epoch": 2.7324848067882126,
+      "grad_norm": 0.23369106650352478,
+      "learning_rate": 2.069987958225561e-05,
+      "loss": 1.7311,
+      "step": 5959
+    },
+    {
+      "epoch": 2.732943469785575,
+      "grad_norm": 0.4220520853996277,
+      "learning_rate": 2.062942086205649e-05,
+      "loss": 1.9563,
+      "step": 5960
+    },
+    {
+      "epoch": 2.733402132782938,
+      "grad_norm": 0.4430352747440338,
+      "learning_rate": 2.055907973491683e-05,
+      "loss": 1.5979,
+      "step": 5961
+    },
+    {
+      "epoch": 2.7338607957803003,
+      "grad_norm": 0.22902828454971313,
+      "learning_rate": 2.0488856218091935e-05,
+      "loss": 1.2991,
+      "step": 5962
+    },
+    {
+      "epoch": 2.734319458777663,
+      "grad_norm": 0.2558428943157196,
+      "learning_rate": 2.0418750328808024e-05,
+      "loss": 1.2835,
+      "step": 5963
+    },
+    {
+      "epoch": 2.734778121775026,
+      "grad_norm": 0.5350182056427002,
+      "learning_rate": 2.034876208426267e-05,
+      "loss": 0.9716,
+      "step": 5964
+    },
+    {
+      "epoch": 2.7352367847723884,
+      "grad_norm": 0.13574962317943573,
+      "learning_rate": 2.0278891501624375e-05,
+      "loss": 1.4147,
+      "step": 5965
+    },
+    {
+      "epoch": 2.7356954477697513,
+      "grad_norm": 0.3386705815792084,
+      "learning_rate": 2.0209138598033026e-05,
+      "loss": 1.1691,
+      "step": 5966
+    },
+    {
+      "epoch": 2.7361541107671137,
+      "grad_norm": 0.34220030903816223,
+      "learning_rate": 2.013950339059939e-05,
+      "loss": 2.25,
+      "step": 5967
+    },
+    {
+      "epoch": 2.7366127737644765,
+      "grad_norm": 0.47397106885910034,
+      "learning_rate": 2.0069985896405574e-05,
+      "loss": 1.6315,
+      "step": 5968
+    },
+    {
+      "epoch": 2.7370714367618394,
+      "grad_norm": 0.3405340015888214,
+      "learning_rate": 2.0000586132504662e-05,
+      "loss": 1.6804,
+      "step": 5969
+    },
+    {
+      "epoch": 2.737530099759202,
+      "grad_norm": 0.3932589292526245,
+      "learning_rate": 1.993130411592098e-05,
+      "loss": 1.1376,
+      "step": 5970
+    },
+    {
+      "epoch": 2.7379887627565647,
+      "grad_norm": 0.3252299427986145,
+      "learning_rate": 1.986213986364982e-05,
+      "loss": 1.9539,
+      "step": 5971
+    },
+    {
+      "epoch": 2.738447425753927,
+      "grad_norm": 0.38866326212882996,
+      "learning_rate": 1.979309339265778e-05,
+      "loss": 1.2502,
+      "step": 5972
+    },
+    {
+      "epoch": 2.73890608875129,
+      "grad_norm": 0.2286403626203537,
+      "learning_rate": 1.9724164719882367e-05,
+      "loss": 1.0532,
+      "step": 5973
+    },
+    {
+      "epoch": 2.739364751748653,
+      "grad_norm": 0.35474932193756104,
+      "learning_rate": 1.9655353862232326e-05,
+      "loss": 2.0171,
+      "step": 5974
+    },
+    {
+      "epoch": 2.739823414746015,
+      "grad_norm": 0.3784157633781433,
+      "learning_rate": 1.9586660836587554e-05,
+      "loss": 1.5926,
+      "step": 5975
+    },
+    {
+      "epoch": 2.740282077743378,
+      "grad_norm": 0.4135417640209198,
+      "learning_rate": 1.9518085659798734e-05,
+      "loss": 1.5249,
+      "step": 5976
+    },
+    {
+      "epoch": 2.7407407407407405,
+      "grad_norm": 0.37199753522872925,
+      "learning_rate": 1.9449628348687964e-05,
+      "loss": 1.5783,
+      "step": 5977
+    },
+    {
+      "epoch": 2.7411994037381033,
+      "grad_norm": 0.29575780034065247,
+      "learning_rate": 1.9381288920048255e-05,
+      "loss": 1.1942,
+      "step": 5978
+    },
+    {
+      "epoch": 2.741658066735466,
+      "grad_norm": 0.2783096730709076,
+      "learning_rate": 1.9313067390643866e-05,
+      "loss": 1.7344,
+      "step": 5979
+    },
+    {
+      "epoch": 2.742116729732829,
+      "grad_norm": 0.37723588943481445,
+      "learning_rate": 1.9244963777209967e-05,
+      "loss": 1.3625,
+      "step": 5980
+    },
+    {
+      "epoch": 2.7425753927301915,
+      "grad_norm": 0.26153478026390076,
+      "learning_rate": 1.917697809645291e-05,
+      "loss": 0.8029,
+      "step": 5981
+    },
+    {
+      "epoch": 2.7430340557275543,
+      "grad_norm": 0.28612053394317627,
+      "learning_rate": 1.910911036505003e-05,
+      "loss": 1.6521,
+      "step": 5982
+    },
+    {
+      "epoch": 2.7434927187249167,
+      "grad_norm": 0.23514768481254578,
+      "learning_rate": 1.9041360599649725e-05,
+      "loss": 1.1054,
+      "step": 5983
+    },
+    {
+      "epoch": 2.7439513817222796,
+      "grad_norm": 0.44542109966278076,
+      "learning_rate": 1.8973728816871592e-05,
+      "loss": 1.3016,
+      "step": 5984
+    },
+    {
+      "epoch": 2.7444100447196424,
+      "grad_norm": 0.19931384921073914,
+      "learning_rate": 1.8906215033306196e-05,
+      "loss": 1.5551,
+      "step": 5985
+    },
+    {
+      "epoch": 2.744868707717005,
+      "grad_norm": 0.2757704555988312,
+      "learning_rate": 1.8838819265515117e-05,
+      "loss": 1.2303,
+      "step": 5986
+    },
+    {
+      "epoch": 2.7453273707143677,
+      "grad_norm": 0.39500927925109863,
+      "learning_rate": 1.8771541530031023e-05,
+      "loss": 1.5353,
+      "step": 5987
+    },
+    {
+      "epoch": 2.74578603371173,
+      "grad_norm": 0.350881963968277,
+      "learning_rate": 1.8704381843357598e-05,
+      "loss": 1.248,
+      "step": 5988
+    },
+    {
+      "epoch": 2.746244696709093,
+      "grad_norm": 0.30257681012153625,
+      "learning_rate": 1.8637340221969613e-05,
+      "loss": 0.8883,
+      "step": 5989
+    },
+    {
+      "epoch": 2.746703359706456,
+      "grad_norm": 0.258948415517807,
+      "learning_rate": 1.8570416682312908e-05,
+      "loss": 1.0292,
+      "step": 5990
+    },
+    {
+      "epoch": 2.7471620227038183,
+      "grad_norm": 0.16734470427036285,
+      "learning_rate": 1.8503611240804186e-05,
+      "loss": 1.1654,
+      "step": 5991
+    },
+    {
+      "epoch": 2.747620685701181,
+      "grad_norm": 0.2848394811153412,
+      "learning_rate": 1.8436923913831506e-05,
+      "loss": 1.6372,
+      "step": 5992
+    },
+    {
+      "epoch": 2.7480793486985435,
+      "grad_norm": 0.3137653172016144,
+      "learning_rate": 1.8370354717753612e-05,
+      "loss": 0.653,
+      "step": 5993
+    },
+    {
+      "epoch": 2.7485380116959064,
+      "grad_norm": 0.21349652111530304,
+      "learning_rate": 1.8303903668900446e-05,
+      "loss": 1.3882,
+      "step": 5994
+    },
+    {
+      "epoch": 2.7489966746932692,
+      "grad_norm": 0.3199220895767212,
+      "learning_rate": 1.823757078357291e-05,
+      "loss": 0.9641,
+      "step": 5995
+    },
+    {
+      "epoch": 2.7494553376906317,
+      "grad_norm": 0.31056568026542664,
+      "learning_rate": 1.8171356078042932e-05,
+      "loss": 1.7107,
+      "step": 5996
+    },
+    {
+      "epoch": 2.7499140006879945,
+      "grad_norm": 0.3857615888118744,
+      "learning_rate": 1.8105259568553524e-05,
+      "loss": 1.0108,
+      "step": 5997
+    },
+    {
+      "epoch": 2.750372663685357,
+      "grad_norm": 0.2920821011066437,
+      "learning_rate": 1.803928127131854e-05,
+      "loss": 1.8806,
+      "step": 5998
+    },
+    {
+      "epoch": 2.75083132668272,
+      "grad_norm": 0.28004190325737,
+      "learning_rate": 1.797342120252299e-05,
+      "loss": 0.4187,
+      "step": 5999
+    },
+    {
+      "epoch": 2.7512899896800826,
+      "grad_norm": 0.26144489645957947,
+      "learning_rate": 1.7907679378322716e-05,
+      "loss": 1.446,
+      "step": 6000
+    },
+    {
+      "epoch": 2.7517486526774455,
+      "grad_norm": 0.2627753019332886,
+      "learning_rate": 1.7842055814844828e-05,
+      "loss": 1.212,
+      "step": 6001
+    },
+    {
+      "epoch": 2.752207315674808,
+      "grad_norm": 0.3469284772872925,
+      "learning_rate": 1.777655052818722e-05,
+      "loss": 0.7107,
+      "step": 6002
+    },
+    {
+      "epoch": 2.7526659786721708,
+      "grad_norm": 0.06406933069229126,
+      "learning_rate": 1.771116353441876e-05,
+      "loss": 0.7278,
+      "step": 6003
+    },
+    {
+      "epoch": 2.753124641669533,
+      "grad_norm": 0.2524787485599518,
+      "learning_rate": 1.7645894849579403e-05,
+      "loss": 0.8452,
+      "step": 6004
+    },
+    {
+      "epoch": 2.753583304666896,
+      "grad_norm": 0.2804563343524933,
+      "learning_rate": 1.7580744489679945e-05,
+      "loss": 0.7473,
+      "step": 6005
+    },
+    {
+      "epoch": 2.754041967664259,
+      "grad_norm": 0.2500944435596466,
+      "learning_rate": 1.7515712470702272e-05,
+      "loss": 1.0894,
+      "step": 6006
+    },
+    {
+      "epoch": 2.7545006306616213,
+      "grad_norm": 0.2973553538322449,
+      "learning_rate": 1.7450798808599234e-05,
+      "loss": 1.3237,
+      "step": 6007
+    },
+    {
+      "epoch": 2.754959293658984,
+      "grad_norm": 0.201811283826828,
+      "learning_rate": 1.73860035192947e-05,
+      "loss": 1.2607,
+      "step": 6008
+    },
+    {
+      "epoch": 2.7554179566563466,
+      "grad_norm": 0.4674420952796936,
+      "learning_rate": 1.7321326618683243e-05,
+      "loss": 1.6664,
+      "step": 6009
+    },
+    {
+      "epoch": 2.7558766196537094,
+      "grad_norm": 0.37687787413597107,
+      "learning_rate": 1.7256768122630607e-05,
+      "loss": 1.213,
+      "step": 6010
+    },
+    {
+      "epoch": 2.7563352826510723,
+      "grad_norm": 0.20660851895809174,
+      "learning_rate": 1.7192328046973572e-05,
+      "loss": 0.5505,
+      "step": 6011
+    },
+    {
+      "epoch": 2.7567939456484347,
+      "grad_norm": 0.573073148727417,
+      "learning_rate": 1.712800640751966e-05,
+      "loss": 1.4566,
+      "step": 6012
+    },
+    {
+      "epoch": 2.7572526086457976,
+      "grad_norm": 0.20739376544952393,
+      "learning_rate": 1.7063803220047524e-05,
+      "loss": 1.18,
+      "step": 6013
+    },
+    {
+      "epoch": 2.75771127164316,
+      "grad_norm": 0.3631076514720917,
+      "learning_rate": 1.6999718500306626e-05,
+      "loss": 1.2751,
+      "step": 6014
+    },
+    {
+      "epoch": 2.758169934640523,
+      "grad_norm": 0.3393692076206207,
+      "learning_rate": 1.6935752264017334e-05,
+      "loss": 1.3281,
+      "step": 6015
+    },
+    {
+      "epoch": 2.7586285976378857,
+      "grad_norm": 0.435544490814209,
+      "learning_rate": 1.6871904526871096e-05,
+      "loss": 1.7262,
+      "step": 6016
+    },
+    {
+      "epoch": 2.759087260635248,
+      "grad_norm": 0.20737430453300476,
+      "learning_rate": 1.680817530453016e-05,
+      "loss": 0.6926,
+      "step": 6017
+    },
+    {
+      "epoch": 2.759545923632611,
+      "grad_norm": 0.2829097807407379,
+      "learning_rate": 1.674456461262791e-05,
+      "loss": 1.6014,
+      "step": 6018
+    },
+    {
+      "epoch": 2.7600045866299734,
+      "grad_norm": 0.3296710252761841,
+      "learning_rate": 1.6681072466768367e-05,
+      "loss": 2.1915,
+      "step": 6019
+    },
+    {
+      "epoch": 2.7604632496273362,
+      "grad_norm": 0.3263249099254608,
+      "learning_rate": 1.6617698882526623e-05,
+      "loss": 1.3439,
+      "step": 6020
+    },
+    {
+      "epoch": 2.760921912624699,
+      "grad_norm": 0.34194350242614746,
+      "learning_rate": 1.6554443875448744e-05,
+      "loss": 1.4912,
+      "step": 6021
+    },
+    {
+      "epoch": 2.761380575622062,
+      "grad_norm": 0.36899879574775696,
+      "learning_rate": 1.6491307461051595e-05,
+      "loss": 1.7578,
+      "step": 6022
+    },
+    {
+      "epoch": 2.7618392386194244,
+      "grad_norm": 0.3341805934906006,
+      "learning_rate": 1.6428289654823014e-05,
+      "loss": 1.1788,
+      "step": 6023
+    },
+    {
+      "epoch": 2.762297901616787,
+      "grad_norm": 0.32973915338516235,
+      "learning_rate": 1.6365390472221742e-05,
+      "loss": 1.2469,
+      "step": 6024
+    },
+    {
+      "epoch": 2.7627565646141496,
+      "grad_norm": 0.463471919298172,
+      "learning_rate": 1.6302609928677382e-05,
+      "loss": 1.6849,
+      "step": 6025
+    },
+    {
+      "epoch": 2.7632152276115125,
+      "grad_norm": 0.40361595153808594,
+      "learning_rate": 1.6239948039590393e-05,
+      "loss": 1.0661,
+      "step": 6026
+    },
+    {
+      "epoch": 2.7636738906088754,
+      "grad_norm": 0.2925053834915161,
+      "learning_rate": 1.6177404820332253e-05,
+      "loss": 1.342,
+      "step": 6027
+    },
+    {
+      "epoch": 2.7641325536062378,
+      "grad_norm": 0.30632370710372925,
+      "learning_rate": 1.611498028624525e-05,
+      "loss": 1.151,
+      "step": 6028
+    },
+    {
+      "epoch": 2.7645912166036006,
+      "grad_norm": 0.22111092507839203,
+      "learning_rate": 1.605267445264258e-05,
+      "loss": 0.4576,
+      "step": 6029
+    },
+    {
+      "epoch": 2.765049879600963,
+      "grad_norm": 0.14749830961227417,
+      "learning_rate": 1.5990487334808292e-05,
+      "loss": 1.0352,
+      "step": 6030
+    },
+    {
+      "epoch": 2.765508542598326,
+      "grad_norm": 0.2582720220088959,
+      "learning_rate": 1.59284189479974e-05,
+      "loss": 1.6917,
+      "step": 6031
+    },
+    {
+      "epoch": 2.7659672055956888,
+      "grad_norm": 0.41496822237968445,
+      "learning_rate": 1.5866469307435626e-05,
+      "loss": 1.5566,
+      "step": 6032
+    },
+    {
+      "epoch": 2.766425868593051,
+      "grad_norm": 0.3769316077232361,
+      "learning_rate": 1.5804638428319694e-05,
+      "loss": 1.65,
+      "step": 6033
+    },
+    {
+      "epoch": 2.766884531590414,
+      "grad_norm": 0.23421497642993927,
+      "learning_rate": 1.5742926325817253e-05,
+      "loss": 1.1492,
+      "step": 6034
+    },
+    {
+      "epoch": 2.7673431945877764,
+      "grad_norm": 0.33073118329048157,
+      "learning_rate": 1.5681333015066635e-05,
+      "loss": 1.5674,
+      "step": 6035
+    },
+    {
+      "epoch": 2.7678018575851393,
+      "grad_norm": 0.48587143421173096,
+      "learning_rate": 1.56198585111772e-05,
+      "loss": 0.745,
+      "step": 6036
+    },
+    {
+      "epoch": 2.768260520582502,
+      "grad_norm": 0.20380260050296783,
+      "learning_rate": 1.5558502829228937e-05,
+      "loss": 0.9425,
+      "step": 6037
+    },
+    {
+      "epoch": 2.7687191835798646,
+      "grad_norm": 0.2204909473657608,
+      "learning_rate": 1.549726598427298e-05,
+      "loss": 1.633,
+      "step": 6038
+    },
+    {
+      "epoch": 2.7691778465772274,
+      "grad_norm": 0.34915247559547424,
+      "learning_rate": 1.5436147991331083e-05,
+      "loss": 0.5378,
+      "step": 6039
+    },
+    {
+      "epoch": 2.76963650957459,
+      "grad_norm": 0.24728168547153473,
+      "learning_rate": 1.5375148865396038e-05,
+      "loss": 1.2788,
+      "step": 6040
+    },
+    {
+      "epoch": 2.7700951725719527,
+      "grad_norm": 0.24167189002037048,
+      "learning_rate": 1.531426862143126e-05,
+      "loss": 1.3767,
+      "step": 6041
+    },
+    {
+      "epoch": 2.7705538355693156,
+      "grad_norm": 0.3961116075515747,
+      "learning_rate": 1.5253507274371137e-05,
+      "loss": 1.0845,
+      "step": 6042
+    },
+    {
+      "epoch": 2.771012498566678,
+      "grad_norm": 0.28995588421821594,
+      "learning_rate": 1.5192864839120912e-05,
+      "loss": 1.2602,
+      "step": 6043
+    },
+    {
+      "epoch": 2.771471161564041,
+      "grad_norm": 0.2699783742427826,
+      "learning_rate": 1.5132341330556576e-05,
+      "loss": 0.9651,
+      "step": 6044
+    },
+    {
+      "epoch": 2.7719298245614032,
+      "grad_norm": 0.2684682011604309,
+      "learning_rate": 1.5071936763524974e-05,
+      "loss": 1.4155,
+      "step": 6045
+    },
+    {
+      "epoch": 2.772388487558766,
+      "grad_norm": 0.27157121896743774,
+      "learning_rate": 1.5011651152843809e-05,
+      "loss": 1.6052,
+      "step": 6046
+    },
+    {
+      "epoch": 2.772847150556129,
+      "grad_norm": 0.3431642949581146,
+      "learning_rate": 1.4951484513301583e-05,
+      "loss": 1.454,
+      "step": 6047
+    },
+    {
+      "epoch": 2.773305813553492,
+      "grad_norm": 0.32993775606155396,
+      "learning_rate": 1.4891436859657604e-05,
+      "loss": 0.9063,
+      "step": 6048
+    },
+    {
+      "epoch": 2.7737644765508542,
+      "grad_norm": 0.28404250741004944,
+      "learning_rate": 1.483150820664192e-05,
+      "loss": 1.5258,
+      "step": 6049
+    },
+    {
+      "epoch": 2.774223139548217,
+      "grad_norm": 0.4539187252521515,
+      "learning_rate": 1.477169856895555e-05,
+      "loss": 1.155,
+      "step": 6050
+    },
+    {
+      "epoch": 2.7746818025455795,
+      "grad_norm": 0.20708614587783813,
+      "learning_rate": 1.4712007961270146e-05,
+      "loss": 0.9088,
+      "step": 6051
+    },
+    {
+      "epoch": 2.7751404655429424,
+      "grad_norm": 0.09351902455091476,
+      "learning_rate": 1.4652436398228385e-05,
+      "loss": 0.8234,
+      "step": 6052
+    },
+    {
+      "epoch": 2.775599128540305,
+      "grad_norm": 0.2884185016155243,
+      "learning_rate": 1.4592983894443468e-05,
+      "loss": 1.8303,
+      "step": 6053
+    },
+    {
+      "epoch": 2.7760577915376676,
+      "grad_norm": 0.3106091618537903,
+      "learning_rate": 1.4533650464499559e-05,
+      "loss": 1.0048,
+      "step": 6054
+    },
+    {
+      "epoch": 2.7765164545350305,
+      "grad_norm": 0.42278361320495605,
+      "learning_rate": 1.4474436122951572e-05,
+      "loss": 1.63,
+      "step": 6055
+    },
+    {
+      "epoch": 2.776975117532393,
+      "grad_norm": 0.37238556146621704,
+      "learning_rate": 1.4415340884325223e-05,
+      "loss": 1.6514,
+      "step": 6056
+    },
+    {
+      "epoch": 2.7774337805297558,
+      "grad_norm": 0.28060218691825867,
+      "learning_rate": 1.4356364763117024e-05,
+      "loss": 0.8222,
+      "step": 6057
+    },
+    {
+      "epoch": 2.7778924435271186,
+      "grad_norm": 0.27165743708610535,
+      "learning_rate": 1.4297507773794239e-05,
+      "loss": 0.7359,
+      "step": 6058
+    },
+    {
+      "epoch": 2.778351106524481,
+      "grad_norm": 0.22083696722984314,
+      "learning_rate": 1.4238769930794926e-05,
+      "loss": 1.2175,
+      "step": 6059
+    },
+    {
+      "epoch": 2.778809769521844,
+      "grad_norm": 0.3628489375114441,
+      "learning_rate": 1.4180151248527784e-05,
+      "loss": 2.0398,
+      "step": 6060
+    },
+    {
+      "epoch": 2.7792684325192063,
+      "grad_norm": 0.36967116594314575,
+      "learning_rate": 1.4121651741372533e-05,
+      "loss": 1.1404,
+      "step": 6061
+    },
+    {
+      "epoch": 2.779727095516569,
+      "grad_norm": 0.30434438586235046,
+      "learning_rate": 1.4063271423679469e-05,
+      "loss": 1.418,
+      "step": 6062
+    },
+    {
+      "epoch": 2.780185758513932,
+      "grad_norm": 0.34148743748664856,
+      "learning_rate": 1.4005010309769638e-05,
+      "loss": 1.3456,
+      "step": 6063
+    },
+    {
+      "epoch": 2.7806444215112944,
+      "grad_norm": 0.2640988528728485,
+      "learning_rate": 1.3946868413935055e-05,
+      "loss": 1.0164,
+      "step": 6064
+    },
+    {
+      "epoch": 2.7811030845086573,
+      "grad_norm": 0.2465277761220932,
+      "learning_rate": 1.3888845750438306e-05,
+      "loss": 0.7947,
+      "step": 6065
+    },
+    {
+      "epoch": 2.7815617475060197,
+      "grad_norm": 0.35097736120224,
+      "learning_rate": 1.3830942333512675e-05,
+      "loss": 1.9631,
+      "step": 6066
+    },
+    {
+      "epoch": 2.7820204105033826,
+      "grad_norm": 0.409355103969574,
+      "learning_rate": 1.3773158177362356e-05,
+      "loss": 1.48,
+      "step": 6067
+    },
+    {
+      "epoch": 2.7824790735007454,
+      "grad_norm": 0.324785441160202,
+      "learning_rate": 1.371549329616223e-05,
+      "loss": 1.4276,
+      "step": 6068
+    },
+    {
+      "epoch": 2.7829377364981083,
+      "grad_norm": 0.372164249420166,
+      "learning_rate": 1.3657947704057872e-05,
+      "loss": 1.144,
+      "step": 6069
+    },
+    {
+      "epoch": 2.7833963994954707,
+      "grad_norm": 0.160455122590065,
+      "learning_rate": 1.3600521415165712e-05,
+      "loss": 1.3252,
+      "step": 6070
+    },
+    {
+      "epoch": 2.7838550624928335,
+      "grad_norm": 0.3352322280406952,
+      "learning_rate": 1.354321444357276e-05,
+      "loss": 1.2853,
+      "step": 6071
+    },
+    {
+      "epoch": 2.784313725490196,
+      "grad_norm": 0.2882045805454254,
+      "learning_rate": 1.3486026803336715e-05,
+      "loss": 1.1088,
+      "step": 6072
+    },
+    {
+      "epoch": 2.784772388487559,
+      "grad_norm": 0.28998827934265137,
+      "learning_rate": 1.3428958508486355e-05,
+      "loss": 1.6669,
+      "step": 6073
+    },
+    {
+      "epoch": 2.7852310514849217,
+      "grad_norm": 0.6466904282569885,
+      "learning_rate": 1.3372009573020816e-05,
+      "loss": 1.6118,
+      "step": 6074
+    },
+    {
+      "epoch": 2.785689714482284,
+      "grad_norm": 0.3447767496109009,
+      "learning_rate": 1.3315180010910145e-05,
+      "loss": 1.3989,
+      "step": 6075
+    },
+    {
+      "epoch": 2.786148377479647,
+      "grad_norm": 0.24244464933872223,
+      "learning_rate": 1.3258469836094911e-05,
+      "loss": 0.4131,
+      "step": 6076
+    },
+    {
+      "epoch": 2.7866070404770094,
+      "grad_norm": 0.264919251203537,
+      "learning_rate": 1.3201879062486655e-05,
+      "loss": 1.6393,
+      "step": 6077
+    },
+    {
+      "epoch": 2.787065703474372,
+      "grad_norm": 0.266245037317276,
+      "learning_rate": 1.3145407703967438e-05,
+      "loss": 1.474,
+      "step": 6078
+    },
+    {
+      "epoch": 2.787524366471735,
+      "grad_norm": 0.3023591935634613,
+      "learning_rate": 1.3089055774390124e-05,
+      "loss": 1.1628,
+      "step": 6079
+    },
+    {
+      "epoch": 2.7879830294690975,
+      "grad_norm": 0.28299427032470703,
+      "learning_rate": 1.3032823287578266e-05,
+      "loss": 1.1443,
+      "step": 6080
+    },
+    {
+      "epoch": 2.7884416924664603,
+      "grad_norm": 0.30756518244743347,
+      "learning_rate": 1.2976710257326053e-05,
+      "loss": 1.1689,
+      "step": 6081
+    },
+    {
+      "epoch": 2.7889003554638228,
+      "grad_norm": 0.2923178970813751,
+      "learning_rate": 1.2920716697398416e-05,
+      "loss": 1.3088,
+      "step": 6082
+    },
+    {
+      "epoch": 2.7893590184611856,
+      "grad_norm": 0.3415941894054413,
+      "learning_rate": 1.2864842621530982e-05,
+      "loss": 1.1528,
+      "step": 6083
+    },
+    {
+      "epoch": 2.7898176814585485,
+      "grad_norm": 0.37786003947257996,
+      "learning_rate": 1.2809088043430116e-05,
+      "loss": 1.2742,
+      "step": 6084
+    },
+    {
+      "epoch": 2.790276344455911,
+      "grad_norm": 0.27256593108177185,
+      "learning_rate": 1.2753452976772773e-05,
+      "loss": 1.4647,
+      "step": 6085
+    },
+    {
+      "epoch": 2.7907350074532737,
+      "grad_norm": 0.3987545073032379,
+      "learning_rate": 1.2697937435206642e-05,
+      "loss": 1.9492,
+      "step": 6086
+    },
+    {
+      "epoch": 2.791193670450636,
+      "grad_norm": 0.36392930150032043,
+      "learning_rate": 1.2642541432350108e-05,
+      "loss": 1.2538,
+      "step": 6087
+    },
+    {
+      "epoch": 2.791652333447999,
+      "grad_norm": 0.31756749749183655,
+      "learning_rate": 1.258726498179219e-05,
+      "loss": 0.9692,
+      "step": 6088
+    },
+    {
+      "epoch": 2.792110996445362,
+      "grad_norm": 0.20181064307689667,
+      "learning_rate": 1.2532108097092598e-05,
+      "loss": 0.8022,
+      "step": 6089
+    },
+    {
+      "epoch": 2.7925696594427247,
+      "grad_norm": 0.1365479826927185,
+      "learning_rate": 1.2477070791781675e-05,
+      "loss": 0.9823,
+      "step": 6090
+    },
+    {
+      "epoch": 2.793028322440087,
+      "grad_norm": 0.20917746424674988,
+      "learning_rate": 1.242215307936051e-05,
+      "loss": 0.7463,
+      "step": 6091
+    },
+    {
+      "epoch": 2.79348698543745,
+      "grad_norm": 0.3799673318862915,
+      "learning_rate": 1.2367354973300881e-05,
+      "loss": 1.5071,
+      "step": 6092
+    },
+    {
+      "epoch": 2.7939456484348124,
+      "grad_norm": 0.3432566523551941,
+      "learning_rate": 1.2312676487045038e-05,
+      "loss": 1.736,
+      "step": 6093
+    },
+    {
+      "epoch": 2.7944043114321753,
+      "grad_norm": 0.33574342727661133,
+      "learning_rate": 1.2258117634006028e-05,
+      "loss": 0.94,
+      "step": 6094
+    },
+    {
+      "epoch": 2.794862974429538,
+      "grad_norm": 0.35387691855430603,
+      "learning_rate": 1.2203678427567588e-05,
+      "loss": 1.8061,
+      "step": 6095
+    },
+    {
+      "epoch": 2.7953216374269005,
+      "grad_norm": 0.2692587971687317,
+      "learning_rate": 1.2149358881084039e-05,
+      "loss": 1.1821,
+      "step": 6096
+    },
+    {
+      "epoch": 2.7957803004242634,
+      "grad_norm": 0.36804336309432983,
+      "learning_rate": 1.2095159007880385e-05,
+      "loss": 1.5841,
+      "step": 6097
+    },
+    {
+      "epoch": 2.796238963421626,
+      "grad_norm": 0.36808767914772034,
+      "learning_rate": 1.204107882125216e-05,
+      "loss": 1.8313,
+      "step": 6098
+    },
+    {
+      "epoch": 2.7966976264189887,
+      "grad_norm": 0.33622780442237854,
+      "learning_rate": 1.1987118334465696e-05,
+      "loss": 1.2008,
+      "step": 6099
+    },
+    {
+      "epoch": 2.7971562894163515,
+      "grad_norm": 0.27367860078811646,
+      "learning_rate": 1.1933277560757793e-05,
+      "loss": 1.3219,
+      "step": 6100
+    },
+    {
+      "epoch": 2.797614952413714,
+      "grad_norm": 0.28815925121307373,
+      "learning_rate": 1.187955651333611e-05,
+      "loss": 0.7678,
+      "step": 6101
+    },
+    {
+      "epoch": 2.798073615411077,
+      "grad_norm": 0.27640852332115173,
+      "learning_rate": 1.1825955205378713e-05,
+      "loss": 1.75,
+      "step": 6102
+    },
+    {
+      "epoch": 2.798532278408439,
+      "grad_norm": 0.32954832911491394,
+      "learning_rate": 1.1772473650034421e-05,
+      "loss": 1.0438,
+      "step": 6103
+    },
+    {
+      "epoch": 2.798990941405802,
+      "grad_norm": 0.3560253381729126,
+      "learning_rate": 1.1719111860422627e-05,
+      "loss": 1.1265,
+      "step": 6104
+    },
+    {
+      "epoch": 2.799449604403165,
+      "grad_norm": 0.11905957758426666,
+      "learning_rate": 1.1665869849633414e-05,
+      "loss": 1.2286,
+      "step": 6105
+    },
+    {
+      "epoch": 2.7999082674005273,
+      "grad_norm": 0.2836627960205078,
+      "learning_rate": 1.1612747630727394e-05,
+      "loss": 1.1157,
+      "step": 6106
+    },
+    {
+      "epoch": 2.80036693039789,
+      "grad_norm": 0.33813512325286865,
+      "learning_rate": 1.1559745216735806e-05,
+      "loss": 0.8221,
+      "step": 6107
+    },
+    {
+      "epoch": 2.8008255933952526,
+      "grad_norm": 0.24317127466201782,
+      "learning_rate": 1.1506862620660586e-05,
+      "loss": 1.5408,
+      "step": 6108
+    },
+    {
+      "epoch": 2.8012842563926155,
+      "grad_norm": 0.33810973167419434,
+      "learning_rate": 1.1454099855474242e-05,
+      "loss": 1.2968,
+      "step": 6109
+    },
+    {
+      "epoch": 2.8017429193899783,
+      "grad_norm": 0.31549063324928284,
+      "learning_rate": 1.1401456934119703e-05,
+      "loss": 1.26,
+      "step": 6110
+    },
+    {
+      "epoch": 2.8022015823873407,
+      "grad_norm": 0.2274772822856903,
+      "learning_rate": 1.1348933869510802e-05,
+      "loss": 0.9997,
+      "step": 6111
+    },
+    {
+      "epoch": 2.8026602453847036,
+      "grad_norm": 0.3890027403831482,
+      "learning_rate": 1.1296530674531735e-05,
+      "loss": 1.1818,
+      "step": 6112
+    },
+    {
+      "epoch": 2.803118908382066,
+      "grad_norm": 0.2179887443780899,
+      "learning_rate": 1.1244247362037496e-05,
+      "loss": 1.7811,
+      "step": 6113
+    },
+    {
+      "epoch": 2.803577571379429,
+      "grad_norm": 0.3312167227268219,
+      "learning_rate": 1.1192083944853438e-05,
+      "loss": 1.1515,
+      "step": 6114
+    },
+    {
+      "epoch": 2.8040362343767917,
+      "grad_norm": 0.24871665239334106,
+      "learning_rate": 1.1140040435775655e-05,
+      "loss": 1.3264,
+      "step": 6115
+    },
+    {
+      "epoch": 2.8044948973741546,
+      "grad_norm": 0.30252861976623535,
+      "learning_rate": 1.1088116847570885e-05,
+      "loss": 1.4769,
+      "step": 6116
+    },
+    {
+      "epoch": 2.804953560371517,
+      "grad_norm": 0.31379151344299316,
+      "learning_rate": 1.1036313192976266e-05,
+      "loss": 1.2689,
+      "step": 6117
+    },
+    {
+      "epoch": 2.80541222336888,
+      "grad_norm": 0.30472010374069214,
+      "learning_rate": 1.0984629484699582e-05,
+      "loss": 0.8448,
+      "step": 6118
+    },
+    {
+      "epoch": 2.8058708863662423,
+      "grad_norm": 0.3143567144870758,
+      "learning_rate": 1.09330657354193e-05,
+      "loss": 1.4705,
+      "step": 6119
+    },
+    {
+      "epoch": 2.806329549363605,
+      "grad_norm": 0.34770259261131287,
+      "learning_rate": 1.0881621957784416e-05,
+      "loss": 1.1561,
+      "step": 6120
+    },
+    {
+      "epoch": 2.806788212360968,
+      "grad_norm": 0.09801855683326721,
+      "learning_rate": 1.0830298164414331e-05,
+      "loss": 0.8337,
+      "step": 6121
+    },
+    {
+      "epoch": 2.8072468753583304,
+      "grad_norm": 0.2806740701198578,
+      "learning_rate": 1.0779094367899201e-05,
+      "loss": 1.6565,
+      "step": 6122
+    },
+    {
+      "epoch": 2.8077055383556933,
+      "grad_norm": 0.7073934674263,
+      "learning_rate": 1.0728010580799696e-05,
+      "loss": 1.5767,
+      "step": 6123
+    },
+    {
+      "epoch": 2.8081642013530557,
+      "grad_norm": 0.32200196385383606,
+      "learning_rate": 1.0677046815647018e-05,
+      "loss": 0.7768,
+      "step": 6124
+    },
+    {
+      "epoch": 2.8086228643504185,
+      "grad_norm": 0.08174686133861542,
+      "learning_rate": 1.0626203084942886e-05,
+      "loss": 1.385,
+      "step": 6125
+    },
+    {
+      "epoch": 2.8090815273477814,
+      "grad_norm": 0.3870648741722107,
+      "learning_rate": 1.0575479401159827e-05,
+      "loss": 1.4463,
+      "step": 6126
+    },
+    {
+      "epoch": 2.809540190345144,
+      "grad_norm": 0.289341539144516,
+      "learning_rate": 1.052487577674055e-05,
+      "loss": 1.167,
+      "step": 6127
+    },
+    {
+      "epoch": 2.8099988533425067,
+      "grad_norm": 0.31475889682769775,
+      "learning_rate": 1.0474392224098572e-05,
+      "loss": 0.841,
+      "step": 6128
+    },
+    {
+      "epoch": 2.810457516339869,
+      "grad_norm": 0.29922276735305786,
+      "learning_rate": 1.042402875561782e-05,
+      "loss": 2.0278,
+      "step": 6129
+    },
+    {
+      "epoch": 2.810916179337232,
+      "grad_norm": 0.4082167148590088,
+      "learning_rate": 1.0373785383652856e-05,
+      "loss": 1.6776,
+      "step": 6130
+    },
+    {
+      "epoch": 2.811374842334595,
+      "grad_norm": 0.3589903712272644,
+      "learning_rate": 1.0323662120528765e-05,
+      "loss": 1.1015,
+      "step": 6131
+    },
+    {
+      "epoch": 2.811833505331957,
+      "grad_norm": 0.24516572058200836,
+      "learning_rate": 1.0273658978541044e-05,
+      "loss": 1.232,
+      "step": 6132
+    },
+    {
+      "epoch": 2.81229216832932,
+      "grad_norm": 0.21401365101337433,
+      "learning_rate": 1.0223775969955883e-05,
+      "loss": 1.2669,
+      "step": 6133
+    },
+    {
+      "epoch": 2.8127508313266825,
+      "grad_norm": 0.4826924800872803,
+      "learning_rate": 1.0174013107009938e-05,
+      "loss": 1.4506,
+      "step": 6134
+    },
+    {
+      "epoch": 2.8132094943240453,
+      "grad_norm": 0.5444715023040771,
+      "learning_rate": 1.0124370401910388e-05,
+      "loss": 1.0889,
+      "step": 6135
+    },
+    {
+      "epoch": 2.813668157321408,
+      "grad_norm": 0.32295656204223633,
+      "learning_rate": 1.0074847866834991e-05,
+      "loss": 1.4926,
+      "step": 6136
+    },
+    {
+      "epoch": 2.814126820318771,
+      "grad_norm": 0.3280322253704071,
+      "learning_rate": 1.002544551393203e-05,
+      "loss": 1.8701,
+      "step": 6137
+    },
+    {
+      "epoch": 2.8145854833161335,
+      "grad_norm": 0.3465189039707184,
+      "learning_rate": 9.976163355320089e-06,
+      "loss": 1.6473,
+      "step": 6138
+    },
+    {
+      "epoch": 2.8150441463134963,
+      "grad_norm": 0.406650573015213,
+      "learning_rate": 9.92700140308861e-06,
+      "loss": 1.2048,
+      "step": 6139
+    },
+    {
+      "epoch": 2.8155028093108587,
+      "grad_norm": 0.32793253660202026,
+      "learning_rate": 9.87795966929722e-06,
+      "loss": 1.0226,
+      "step": 6140
+    },
+    {
+      "epoch": 2.8159614723082216,
+      "grad_norm": 0.24950462579727173,
+      "learning_rate": 9.8290381659763e-06,
+      "loss": 1.614,
+      "step": 6141
+    },
+    {
+      "epoch": 2.8164201353055844,
+      "grad_norm": 0.27943155169487,
+      "learning_rate": 9.780236905126694e-06,
+      "loss": 1.1877,
+      "step": 6142
+    },
+    {
+      "epoch": 2.816878798302947,
+      "grad_norm": 0.33849433064460754,
+      "learning_rate": 9.731555898719601e-06,
+      "loss": 1.5759,
+      "step": 6143
+    },
+    {
+      "epoch": 2.8173374613003097,
+      "grad_norm": 0.28267404437065125,
+      "learning_rate": 9.682995158696806e-06,
+      "loss": 0.7854,
+      "step": 6144
+    },
+    {
+      "epoch": 2.817796124297672,
+      "grad_norm": 0.28812381625175476,
+      "learning_rate": 9.63455469697072e-06,
+      "loss": 1.5585,
+      "step": 6145
+    },
+    {
+      "epoch": 2.818254787295035,
+      "grad_norm": 0.3284319341182709,
+      "learning_rate": 9.58623452542412e-06,
+      "loss": 1.6185,
+      "step": 6146
+    },
+    {
+      "epoch": 2.818713450292398,
+      "grad_norm": 0.391230046749115,
+      "learning_rate": 9.538034655910189e-06,
+      "loss": 1.822,
+      "step": 6147
+    },
+    {
+      "epoch": 2.8191721132897603,
+      "grad_norm": 0.38372287154197693,
+      "learning_rate": 9.489955100252855e-06,
+      "loss": 1.2379,
+      "step": 6148
+    },
+    {
+      "epoch": 2.819630776287123,
+      "grad_norm": 0.37333911657333374,
+      "learning_rate": 9.441995870246244e-06,
+      "loss": 1.638,
+      "step": 6149
+    },
+    {
+      "epoch": 2.8200894392844855,
+      "grad_norm": 0.33912524580955505,
+      "learning_rate": 9.394156977655165e-06,
+      "loss": 1.0467,
+      "step": 6150
+    },
+    {
+      "epoch": 2.8205481022818484,
+      "grad_norm": 0.24765615165233612,
+      "learning_rate": 9.346438434214843e-06,
+      "loss": 1.575,
+      "step": 6151
+    },
+    {
+      "epoch": 2.8210067652792112,
+      "grad_norm": 0.6433941125869751,
+      "learning_rate": 9.298840251630913e-06,
+      "loss": 1.5814,
+      "step": 6152
+    },
+    {
+      "epoch": 2.8214654282765737,
+      "grad_norm": 0.2799464464187622,
+      "learning_rate": 9.251362441579646e-06,
+      "loss": 1.0583,
+      "step": 6153
+    },
+    {
+      "epoch": 2.8219240912739365,
+      "grad_norm": 0.21018120646476746,
+      "learning_rate": 9.204005015707673e-06,
+      "loss": 1.2,
+      "step": 6154
+    },
+    {
+      "epoch": 2.822382754271299,
+      "grad_norm": 0.2773264944553375,
+      "learning_rate": 9.15676798563203e-06,
+      "loss": 1.0197,
+      "step": 6155
+    },
+    {
+      "epoch": 2.822841417268662,
+      "grad_norm": 0.34957289695739746,
+      "learning_rate": 9.109651362940397e-06,
+      "loss": 1.3116,
+      "step": 6156
+    },
+    {
+      "epoch": 2.8233000802660246,
+      "grad_norm": 0.3279551863670349,
+      "learning_rate": 9.062655159190802e-06,
+      "loss": 2.0532,
+      "step": 6157
+    },
+    {
+      "epoch": 2.8237587432633875,
+      "grad_norm": 0.4215746521949768,
+      "learning_rate": 9.015779385911748e-06,
+      "loss": 1.4172,
+      "step": 6158
+    },
+    {
+      "epoch": 2.82421740626075,
+      "grad_norm": 0.41953399777412415,
+      "learning_rate": 8.969024054602204e-06,
+      "loss": 1.0516,
+      "step": 6159
+    },
+    {
+      "epoch": 2.8246760692581128,
+      "grad_norm": 0.08408641070127487,
+      "learning_rate": 8.922389176731549e-06,
+      "loss": 1.189,
+      "step": 6160
+    },
+    {
+      "epoch": 2.825134732255475,
+      "grad_norm": 0.3760605752468109,
+      "learning_rate": 8.875874763739633e-06,
+      "loss": 1.4748,
+      "step": 6161
+    },
+    {
+      "epoch": 2.825593395252838,
+      "grad_norm": 0.21087396144866943,
+      "learning_rate": 8.829480827036884e-06,
+      "loss": 0.7344,
+      "step": 6162
+    },
+    {
+      "epoch": 2.826052058250201,
+      "grad_norm": 0.19473138451576233,
+      "learning_rate": 8.783207378003977e-06,
+      "loss": 0.9306,
+      "step": 6163
+    },
+    {
+      "epoch": 2.8265107212475633,
+      "grad_norm": 0.3104762136936188,
+      "learning_rate": 8.737054427992164e-06,
+      "loss": 1.9001,
+      "step": 6164
+    },
+    {
+      "epoch": 2.826969384244926,
+      "grad_norm": 0.4818089008331299,
+      "learning_rate": 8.691021988323111e-06,
+      "loss": 1.6207,
+      "step": 6165
+    },
+    {
+      "epoch": 2.8274280472422886,
+      "grad_norm": 0.31389549374580383,
+      "learning_rate": 8.645110070288897e-06,
+      "loss": 1.3694,
+      "step": 6166
+    },
+    {
+      "epoch": 2.8278867102396514,
+      "grad_norm": 0.352982759475708,
+      "learning_rate": 8.599318685152014e-06,
+      "loss": 1.1738,
+      "step": 6167
+    },
+    {
+      "epoch": 2.8283453732370143,
+      "grad_norm": 0.20497311651706696,
+      "learning_rate": 8.553647844145418e-06,
+      "loss": 0.7585,
+      "step": 6168
+    },
+    {
+      "epoch": 2.8288040362343767,
+      "grad_norm": 0.30926766991615295,
+      "learning_rate": 8.508097558472538e-06,
+      "loss": 1.0177,
+      "step": 6169
+    },
+    {
+      "epoch": 2.8292626992317396,
+      "grad_norm": 0.3021996319293976,
+      "learning_rate": 8.462667839307159e-06,
+      "loss": 1.1897,
+      "step": 6170
+    },
+    {
+      "epoch": 2.829721362229102,
+      "grad_norm": 0.2463819533586502,
+      "learning_rate": 8.417358697793587e-06,
+      "loss": 1.6089,
+      "step": 6171
+    },
+    {
+      "epoch": 2.830180025226465,
+      "grad_norm": 0.3797639012336731,
+      "learning_rate": 8.372170145046376e-06,
+      "loss": 1.718,
+      "step": 6172
+    },
+    {
+      "epoch": 2.8306386882238277,
+      "grad_norm": 0.26537811756134033,
+      "learning_rate": 8.327102192150604e-06,
+      "loss": 1.5723,
+      "step": 6173
+    },
+    {
+      "epoch": 2.83109735122119,
+      "grad_norm": 0.26074454188346863,
+      "learning_rate": 8.282154850161871e-06,
+      "loss": 0.9977,
+      "step": 6174
+    },
+    {
+      "epoch": 2.831556014218553,
+      "grad_norm": 0.29502686858177185,
+      "learning_rate": 8.23732813010597e-06,
+      "loss": 1.4637,
+      "step": 6175
+    },
+    {
+      "epoch": 2.8320146772159154,
+      "grad_norm": 0.42862004041671753,
+      "learning_rate": 8.192622042979325e-06,
+      "loss": 1.2069,
+      "step": 6176
+    },
+    {
+      "epoch": 2.8324733402132782,
+      "grad_norm": 0.197200745344162,
+      "learning_rate": 8.148036599748554e-06,
+      "loss": 1.193,
+      "step": 6177
+    },
+    {
+      "epoch": 2.832932003210641,
+      "grad_norm": 0.34668564796447754,
+      "learning_rate": 8.103571811350851e-06,
+      "loss": 1.3786,
+      "step": 6178
+    },
+    {
+      "epoch": 2.8333906662080035,
+      "grad_norm": 0.22151388227939606,
+      "learning_rate": 8.059227688693771e-06,
+      "loss": 1.4785,
+      "step": 6179
+    },
+    {
+      "epoch": 2.8338493292053664,
+      "grad_norm": 0.38517090678215027,
+      "learning_rate": 8.015004242655222e-06,
+      "loss": 1.0836,
+      "step": 6180
+    },
+    {
+      "epoch": 2.8343079922027288,
+      "grad_norm": 0.40751755237579346,
+      "learning_rate": 7.970901484083471e-06,
+      "loss": 2.0908,
+      "step": 6181
+    },
+    {
+      "epoch": 2.8347666552000916,
+      "grad_norm": 0.32812926173210144,
+      "learning_rate": 7.926919423797362e-06,
+      "loss": 0.83,
+      "step": 6182
+    },
+    {
+      "epoch": 2.8352253181974545,
+      "grad_norm": 0.36086252331733704,
+      "learning_rate": 7.883058072585935e-06,
+      "loss": 1.5337,
+      "step": 6183
+    },
+    {
+      "epoch": 2.8356839811948173,
+      "grad_norm": 0.3034031391143799,
+      "learning_rate": 7.839317441208693e-06,
+      "loss": 1.334,
+      "step": 6184
+    },
+    {
+      "epoch": 2.8361426441921798,
+      "grad_norm": 0.16254828870296478,
+      "learning_rate": 7.795697540395552e-06,
+      "loss": 0.5699,
+      "step": 6185
+    },
+    {
+      "epoch": 2.8366013071895426,
+      "grad_norm": 0.30455073714256287,
+      "learning_rate": 7.752198380846787e-06,
+      "loss": 1.2536,
+      "step": 6186
+    },
+    {
+      "epoch": 2.837059970186905,
+      "grad_norm": 0.21652016043663025,
+      "learning_rate": 7.708819973233028e-06,
+      "loss": 1.4794,
+      "step": 6187
+    },
+    {
+      "epoch": 2.837518633184268,
+      "grad_norm": 0.2286727875471115,
+      "learning_rate": 7.665562328195375e-06,
+      "loss": 0.6623,
+      "step": 6188
+    },
+    {
+      "epoch": 2.8379772961816307,
+      "grad_norm": 0.31880277395248413,
+      "learning_rate": 7.622425456345172e-06,
+      "loss": 1.2148,
+      "step": 6189
+    },
+    {
+      "epoch": 2.838435959178993,
+      "grad_norm": 0.2918204367160797,
+      "learning_rate": 7.5794093682641785e-06,
+      "loss": 1.5525,
+      "step": 6190
+    },
+    {
+      "epoch": 2.838894622176356,
+      "grad_norm": 0.3420119881629944,
+      "learning_rate": 7.536514074504675e-06,
+      "loss": 1.3251,
+      "step": 6191
+    },
+    {
+      "epoch": 2.8393532851737184,
+      "grad_norm": 0.3030669391155243,
+      "learning_rate": 7.4937395855890765e-06,
+      "loss": 1.1624,
+      "step": 6192
+    },
+    {
+      "epoch": 2.8398119481710813,
+      "grad_norm": 0.22030462324619293,
+      "learning_rate": 7.45108591201038e-06,
+      "loss": 1.4805,
+      "step": 6193
+    },
+    {
+      "epoch": 2.840270611168444,
+      "grad_norm": 0.2560308873653412,
+      "learning_rate": 7.408553064231716e-06,
+      "loss": 0.6927,
+      "step": 6194
+    },
+    {
+      "epoch": 2.8407292741658066,
+      "grad_norm": 0.28317221999168396,
+      "learning_rate": 7.366141052686737e-06,
+      "loss": 1.3214,
+      "step": 6195
+    },
+    {
+      "epoch": 2.8411879371631694,
+      "grad_norm": 0.29174163937568665,
+      "learning_rate": 7.3238498877794544e-06,
+      "loss": 1.5961,
+      "step": 6196
+    },
+    {
+      "epoch": 2.841646600160532,
+      "grad_norm": 0.4265110492706299,
+      "learning_rate": 7.281679579884126e-06,
+      "loss": 1.8915,
+      "step": 6197
+    },
+    {
+      "epoch": 2.8421052631578947,
+      "grad_norm": 0.36179837584495544,
+      "learning_rate": 7.239630139345532e-06,
+      "loss": 0.9835,
+      "step": 6198
+    },
+    {
+      "epoch": 2.8425639261552575,
+      "grad_norm": 0.1963047832250595,
+      "learning_rate": 7.197701576478699e-06,
+      "loss": 1.0536,
+      "step": 6199
+    },
+    {
+      "epoch": 2.84302258915262,
+      "grad_norm": 0.26563286781311035,
+      "learning_rate": 7.1558939015689e-06,
+      "loss": 1.058,
+      "step": 6200
+    },
+    {
+      "epoch": 2.843481252149983,
+      "grad_norm": 0.295461505651474,
+      "learning_rate": 7.114207124871874e-06,
+      "loss": 1.3314,
+      "step": 6201
+    },
+    {
+      "epoch": 2.8439399151473452,
+      "grad_norm": 0.287113219499588,
+      "learning_rate": 7.072641256613777e-06,
+      "loss": 0.7959,
+      "step": 6202
+    },
+    {
+      "epoch": 2.844398578144708,
+      "grad_norm": 0.2837468385696411,
+      "learning_rate": 7.031196306991005e-06,
+      "loss": 1.5568,
+      "step": 6203
+    },
+    {
+      "epoch": 2.844857241142071,
+      "grad_norm": 0.2658827602863312,
+      "learning_rate": 6.989872286170262e-06,
+      "loss": 0.6964,
+      "step": 6204
+    },
+    {
+      "epoch": 2.845315904139434,
+      "grad_norm": 0.3154827654361725,
+      "learning_rate": 6.948669204288604e-06,
+      "loss": 1.7673,
+      "step": 6205
+    },
+    {
+      "epoch": 2.845774567136796,
+      "grad_norm": 0.3644176423549652,
+      "learning_rate": 6.907587071453447e-06,
+      "loss": 0.5235,
+      "step": 6206
+    },
+    {
+      "epoch": 2.846233230134159,
+      "grad_norm": 0.2641429603099823,
+      "learning_rate": 6.866625897742562e-06,
+      "loss": 1.2719,
+      "step": 6207
+    },
+    {
+      "epoch": 2.8466918931315215,
+      "grad_norm": 0.34545841813087463,
+      "learning_rate": 6.825785693204023e-06,
+      "loss": 1.8203,
+      "step": 6208
+    },
+    {
+      "epoch": 2.8471505561288843,
+      "grad_norm": 0.34669947624206543,
+      "learning_rate": 6.78506646785626e-06,
+      "loss": 1.2168,
+      "step": 6209
+    },
+    {
+      "epoch": 2.847609219126247,
+      "grad_norm": 0.2993127405643463,
+      "learning_rate": 6.744468231688006e-06,
+      "loss": 1.5106,
+      "step": 6210
+    },
+    {
+      "epoch": 2.8480678821236096,
+      "grad_norm": 0.36481523513793945,
+      "learning_rate": 6.7039909946581825e-06,
+      "loss": 0.9908,
+      "step": 6211
+    },
+    {
+      "epoch": 2.8485265451209725,
+      "grad_norm": 0.24906761944293976,
+      "learning_rate": 6.663634766696236e-06,
+      "loss": 1.5602,
+      "step": 6212
+    },
+    {
+      "epoch": 2.848985208118335,
+      "grad_norm": 0.3981691002845764,
+      "learning_rate": 6.623399557701803e-06,
+      "loss": 1.237,
+      "step": 6213
+    },
+    {
+      "epoch": 2.8494438711156977,
+      "grad_norm": 0.25347137451171875,
+      "learning_rate": 6.5832853775448784e-06,
+      "loss": 1.5596,
+      "step": 6214
+    },
+    {
+      "epoch": 2.8499025341130606,
+      "grad_norm": 0.33858397603034973,
+      "learning_rate": 6.543292236065812e-06,
+      "loss": 1.1393,
+      "step": 6215
+    },
+    {
+      "epoch": 2.850361197110423,
+      "grad_norm": 0.31756460666656494,
+      "learning_rate": 6.50342014307509e-06,
+      "loss": 2.2029,
+      "step": 6216
+    },
+    {
+      "epoch": 2.850819860107786,
+      "grad_norm": 0.4458785653114319,
+      "learning_rate": 6.463669108353776e-06,
+      "loss": 1.5674,
+      "step": 6217
+    },
+    {
+      "epoch": 2.8512785231051483,
+      "grad_norm": 0.3402140438556671,
+      "learning_rate": 6.42403914165296e-06,
+      "loss": 1.3496,
+      "step": 6218
+    },
+    {
+      "epoch": 2.851737186102511,
+      "grad_norm": 0.3040880858898163,
+      "learning_rate": 6.384530252694254e-06,
+      "loss": 1.3495,
+      "step": 6219
+    },
+    {
+      "epoch": 2.852195849099874,
+      "grad_norm": 0.4113820493221283,
+      "learning_rate": 6.345142451169405e-06,
+      "loss": 1.2961,
+      "step": 6220
+    },
+    {
+      "epoch": 2.8526545120972364,
+      "grad_norm": 0.3598925471305847,
+      "learning_rate": 6.305875746740574e-06,
+      "loss": 1.4615,
+      "step": 6221
+    },
+    {
+      "epoch": 2.8531131750945993,
+      "grad_norm": 0.3074336051940918,
+      "learning_rate": 6.266730149040112e-06,
+      "loss": 1.2253,
+      "step": 6222
+    },
+    {
+      "epoch": 2.8535718380919617,
+      "grad_norm": 0.3465123474597931,
+      "learning_rate": 6.22770566767078e-06,
+      "loss": 1.5122,
+      "step": 6223
+    },
+    {
+      "epoch": 2.8540305010893245,
+      "grad_norm": 0.3549046218395233,
+      "learning_rate": 6.188802312205477e-06,
+      "loss": 1.9436,
+      "step": 6224
+    },
+    {
+      "epoch": 2.8544891640866874,
+      "grad_norm": 0.30973848700523376,
+      "learning_rate": 6.1500200921875695e-06,
+      "loss": 0.8639,
+      "step": 6225
+    },
+    {
+      "epoch": 2.8549478270840503,
+      "grad_norm": 0.21493889391422272,
+      "learning_rate": 6.111359017130558e-06,
+      "loss": 1.3644,
+      "step": 6226
+    },
+    {
+      "epoch": 2.8554064900814127,
+      "grad_norm": 0.2825007140636444,
+      "learning_rate": 6.072819096518301e-06,
+      "loss": 1.3103,
+      "step": 6227
+    },
+    {
+      "epoch": 2.8558651530787755,
+      "grad_norm": 0.3511542081832886,
+      "learning_rate": 6.034400339804902e-06,
+      "loss": 1.46,
+      "step": 6228
+    },
+    {
+      "epoch": 2.856323816076138,
+      "grad_norm": 0.27357274293899536,
+      "learning_rate": 5.996102756414823e-06,
+      "loss": 0.7502,
+      "step": 6229
+    },
+    {
+      "epoch": 2.856782479073501,
+      "grad_norm": 0.22305287420749664,
+      "learning_rate": 5.95792635574266e-06,
+      "loss": 1.3862,
+      "step": 6230
+    },
+    {
+      "epoch": 2.8572411420708637,
+      "grad_norm": 0.3165452778339386,
+      "learning_rate": 5.919871147153422e-06,
+      "loss": 1.2199,
+      "step": 6231
+    },
+    {
+      "epoch": 2.857699805068226,
+      "grad_norm": 0.2813020646572113,
+      "learning_rate": 5.881937139982307e-06,
+      "loss": 0.8148,
+      "step": 6232
+    },
+    {
+      "epoch": 2.858158468065589,
+      "grad_norm": 0.32087957859039307,
+      "learning_rate": 5.844124343534707e-06,
+      "loss": 1.5054,
+      "step": 6233
+    },
+    {
+      "epoch": 2.8586171310629513,
+      "grad_norm": 0.22067861258983612,
+      "learning_rate": 5.806432767086534e-06,
+      "loss": 0.9947,
+      "step": 6234
+    },
+    {
+      "epoch": 2.859075794060314,
+      "grad_norm": 0.28397810459136963,
+      "learning_rate": 5.768862419883669e-06,
+      "loss": 1.1807,
+      "step": 6235
+    },
+    {
+      "epoch": 2.859534457057677,
+      "grad_norm": 0.28341740369796753,
+      "learning_rate": 5.731413311142464e-06,
+      "loss": 0.7978,
+      "step": 6236
+    },
+    {
+      "epoch": 2.8599931200550395,
+      "grad_norm": 0.30934879183769226,
+      "learning_rate": 5.694085450049402e-06,
+      "loss": 1.3267,
+      "step": 6237
+    },
+    {
+      "epoch": 2.8604517830524023,
+      "grad_norm": 0.32292166352272034,
+      "learning_rate": 5.656878845761326e-06,
+      "loss": 1.7622,
+      "step": 6238
+    },
+    {
+      "epoch": 2.8609104460497647,
+      "grad_norm": 0.41772255301475525,
+      "learning_rate": 5.619793507405324e-06,
+      "loss": 1.4691,
+      "step": 6239
+    },
+    {
+      "epoch": 2.8613691090471276,
+      "grad_norm": 0.3389241695404053,
+      "learning_rate": 5.582829444078563e-06,
+      "loss": 1.1824,
+      "step": 6240
+    },
+    {
+      "epoch": 2.8618277720444905,
+      "grad_norm": 0.15310680866241455,
+      "learning_rate": 5.5459866648487345e-06,
+      "loss": 0.7682,
+      "step": 6241
+    },
+    {
+      "epoch": 2.862286435041853,
+      "grad_norm": 0.33451229333877563,
+      "learning_rate": 5.509265178753497e-06,
+      "loss": 1.092,
+      "step": 6242
+    },
+    {
+      "epoch": 2.8627450980392157,
+      "grad_norm": 0.23663324117660522,
+      "learning_rate": 5.472664994801091e-06,
+      "loss": 1.3781,
+      "step": 6243
+    },
+    {
+      "epoch": 2.863203761036578,
+      "grad_norm": 0.3743632137775421,
+      "learning_rate": 5.436186121969611e-06,
+      "loss": 1.5989,
+      "step": 6244
+    },
+    {
+      "epoch": 2.863662424033941,
+      "grad_norm": 0.3161230683326721,
+      "learning_rate": 5.3998285692076765e-06,
+      "loss": 1.1666,
+      "step": 6245
+    },
+    {
+      "epoch": 2.864121087031304,
+      "grad_norm": 0.16574889421463013,
+      "learning_rate": 5.363592345434043e-06,
+      "loss": 1.4562,
+      "step": 6246
+    },
+    {
+      "epoch": 2.8645797500286663,
+      "grad_norm": 0.35472193360328674,
+      "learning_rate": 5.327477459537711e-06,
+      "loss": 0.7368,
+      "step": 6247
+    },
+    {
+      "epoch": 2.865038413026029,
+      "grad_norm": 0.2824570834636688,
+      "learning_rate": 5.291483920377926e-06,
+      "loss": 1.7105,
+      "step": 6248
+    },
+    {
+      "epoch": 2.8654970760233915,
+      "grad_norm": 0.2135975956916809,
+      "learning_rate": 5.255611736784183e-06,
+      "loss": 1.493,
+      "step": 6249
+    },
+    {
+      "epoch": 2.8659557390207544,
+      "grad_norm": 0.29121702909469604,
+      "learning_rate": 5.219860917556163e-06,
+      "loss": 1.2036,
+      "step": 6250
+    },
+    {
+      "epoch": 2.8664144020181173,
+      "grad_norm": 0.43849441409111023,
+      "learning_rate": 5.184231471463852e-06,
+      "loss": 1.9182,
+      "step": 6251
+    },
+    {
+      "epoch": 2.86687306501548,
+      "grad_norm": 0.3722185492515564,
+      "learning_rate": 5.1487234072473135e-06,
+      "loss": 1.6022,
+      "step": 6252
+    },
+    {
+      "epoch": 2.8673317280128425,
+      "grad_norm": 0.3918931782245636,
+      "learning_rate": 5.1133367336170245e-06,
+      "loss": 1.9431,
+      "step": 6253
+    },
+    {
+      "epoch": 2.8677903910102054,
+      "grad_norm": 0.4557853937149048,
+      "learning_rate": 5.078071459253541e-06,
+      "loss": 1.2453,
+      "step": 6254
+    },
+    {
+      "epoch": 2.868249054007568,
+      "grad_norm": 0.11661559343338013,
+      "learning_rate": 5.042927592807722e-06,
+      "loss": 0.6514,
+      "step": 6255
+    },
+    {
+      "epoch": 2.8687077170049307,
+      "grad_norm": 0.3350408673286438,
+      "learning_rate": 5.00790514290056e-06,
+      "loss": 1.8324,
+      "step": 6256
+    },
+    {
+      "epoch": 2.8691663800022935,
+      "grad_norm": 0.37260258197784424,
+      "learning_rate": 4.973004118123348e-06,
+      "loss": 2.1526,
+      "step": 6257
+    },
+    {
+      "epoch": 2.869625042999656,
+      "grad_norm": 0.6885807514190674,
+      "learning_rate": 4.938224527037516e-06,
+      "loss": 1.2802,
+      "step": 6258
+    },
+    {
+      "epoch": 2.870083705997019,
+      "grad_norm": 0.27935993671417236,
+      "learning_rate": 4.903566378174795e-06,
+      "loss": 1.3413,
+      "step": 6259
+    },
+    {
+      "epoch": 2.870542368994381,
+      "grad_norm": 0.3012205958366394,
+      "learning_rate": 4.869029680037162e-06,
+      "loss": 1.4973,
+      "step": 6260
+    },
+    {
+      "epoch": 2.871001031991744,
+      "grad_norm": 0.4477826654911041,
+      "learning_rate": 4.834614441096563e-06,
+      "loss": 1.712,
+      "step": 6261
+    },
+    {
+      "epoch": 2.871459694989107,
+      "grad_norm": 0.3502928912639618,
+      "learning_rate": 4.800320669795355e-06,
+      "loss": 1.2868,
+      "step": 6262
+    },
+    {
+      "epoch": 2.8719183579864693,
+      "grad_norm": 0.19043685495853424,
+      "learning_rate": 4.766148374546087e-06,
+      "loss": 1.1846,
+      "step": 6263
+    },
+    {
+      "epoch": 2.872377020983832,
+      "grad_norm": 0.27802857756614685,
+      "learning_rate": 4.7320975637314415e-06,
+      "loss": 1.6084,
+      "step": 6264
+    },
+    {
+      "epoch": 2.8728356839811946,
+      "grad_norm": 0.3813520669937134,
+      "learning_rate": 4.698168245704349e-06,
+      "loss": 1.7174,
+      "step": 6265
+    },
+    {
+      "epoch": 2.8732943469785575,
+      "grad_norm": 0.3918079137802124,
+      "learning_rate": 4.6643604287878726e-06,
+      "loss": 1.5721,
+      "step": 6266
+    },
+    {
+      "epoch": 2.8737530099759203,
+      "grad_norm": 0.2761102020740509,
+      "learning_rate": 4.630674121275325e-06,
+      "loss": 1.1836,
+      "step": 6267
+    },
+    {
+      "epoch": 2.8742116729732827,
+      "grad_norm": 0.33837902545928955,
+      "learning_rate": 4.59710933143026e-06,
+      "loss": 1.2286,
+      "step": 6268
+    },
+    {
+      "epoch": 2.8746703359706456,
+      "grad_norm": 0.24910743534564972,
+      "learning_rate": 4.56366606748626e-06,
+      "loss": 1.6427,
+      "step": 6269
+    },
+    {
+      "epoch": 2.875128998968008,
+      "grad_norm": 0.37364476919174194,
+      "learning_rate": 4.5303443376472635e-06,
+      "loss": 1.169,
+      "step": 6270
+    },
+    {
+      "epoch": 2.875587661965371,
+      "grad_norm": 0.38252532482147217,
+      "learning_rate": 4.49714415008734e-06,
+      "loss": 1.5502,
+      "step": 6271
+    },
+    {
+      "epoch": 2.8760463249627337,
+      "grad_norm": 0.35861456394195557,
+      "learning_rate": 4.464065512950754e-06,
+      "loss": 1.1111,
+      "step": 6272
+    },
+    {
+      "epoch": 2.8765049879600966,
+      "grad_norm": 0.3056200444698334,
+      "learning_rate": 4.4311084343518496e-06,
+      "loss": 1.8339,
+      "step": 6273
+    },
+    {
+      "epoch": 2.876963650957459,
+      "grad_norm": 0.2743782103061676,
+      "learning_rate": 4.398272922375268e-06,
+      "loss": 1.0009,
+      "step": 6274
+    },
+    {
+      "epoch": 2.877422313954822,
+      "grad_norm": 0.3063773214817047,
+      "learning_rate": 4.365558985075846e-06,
+      "loss": 1.2168,
+      "step": 6275
+    },
+    {
+      "epoch": 2.8778809769521843,
+      "grad_norm": 0.2390686422586441,
+      "learning_rate": 4.332966630478497e-06,
+      "loss": 0.8118,
+      "step": 6276
+    },
+    {
+      "epoch": 2.878339639949547,
+      "grad_norm": 0.3186044991016388,
+      "learning_rate": 4.300495866578435e-06,
+      "loss": 1.162,
+      "step": 6277
+    },
+    {
+      "epoch": 2.87879830294691,
+      "grad_norm": 0.32464686036109924,
+      "learning_rate": 4.268146701340847e-06,
+      "loss": 1.9541,
+      "step": 6278
+    },
+    {
+      "epoch": 2.8792569659442724,
+      "grad_norm": 0.3670261800289154,
+      "learning_rate": 4.235919142701272e-06,
+      "loss": 1.6052,
+      "step": 6279
+    },
+    {
+      "epoch": 2.8797156289416352,
+      "grad_norm": 0.2734290361404419,
+      "learning_rate": 4.203813198565387e-06,
+      "loss": 0.9119,
+      "step": 6280
+    },
+    {
+      "epoch": 2.8801742919389977,
+      "grad_norm": 0.20262974500656128,
+      "learning_rate": 4.171828876809003e-06,
+      "loss": 1.2017,
+      "step": 6281
+    },
+    {
+      "epoch": 2.8806329549363605,
+      "grad_norm": 0.3488999307155609,
+      "learning_rate": 4.1399661852781764e-06,
+      "loss": 1.2915,
+      "step": 6282
+    },
+    {
+      "epoch": 2.8810916179337234,
+      "grad_norm": 0.39077028632164,
+      "learning_rate": 4.108225131788934e-06,
+      "loss": 2.073,
+      "step": 6283
+    },
+    {
+      "epoch": 2.881550280931086,
+      "grad_norm": 0.3423274755477905,
+      "learning_rate": 4.076605724127602e-06,
+      "loss": 1.2798,
+      "step": 6284
+    },
+    {
+      "epoch": 2.8820089439284486,
+      "grad_norm": 0.3173692226409912,
+      "learning_rate": 4.045107970050699e-06,
+      "loss": 1.0493,
+      "step": 6285
+    },
+    {
+      "epoch": 2.882467606925811,
+      "grad_norm": 0.3519616425037384,
+      "learning_rate": 4.0137318772848205e-06,
+      "loss": 1.5614,
+      "step": 6286
+    },
+    {
+      "epoch": 2.882926269923174,
+      "grad_norm": 0.3195357024669647,
+      "learning_rate": 3.982477453526756e-06,
+      "loss": 1.1449,
+      "step": 6287
+    },
+    {
+      "epoch": 2.8833849329205368,
+      "grad_norm": 0.21202579140663147,
+      "learning_rate": 3.951344706443427e-06,
+      "loss": 1.2399,
+      "step": 6288
+    },
+    {
+      "epoch": 2.883843595917899,
+      "grad_norm": 0.28453898429870605,
+      "learning_rate": 3.920333643672003e-06,
+      "loss": 1.2978,
+      "step": 6289
+    },
+    {
+      "epoch": 2.884302258915262,
+      "grad_norm": 0.3411079943180084,
+      "learning_rate": 3.88944427281962e-06,
+      "loss": 1.5203,
+      "step": 6290
+    },
+    {
+      "epoch": 2.8847609219126245,
+      "grad_norm": 0.32296404242515564,
+      "learning_rate": 3.858676601463662e-06,
+      "loss": 1.1908,
+      "step": 6291
+    },
+    {
+      "epoch": 2.8852195849099873,
+      "grad_norm": 0.3648724853992462,
+      "learning_rate": 3.828030637151758e-06,
+      "loss": 1.6689,
+      "step": 6292
+    },
+    {
+      "epoch": 2.88567824790735,
+      "grad_norm": 0.26379936933517456,
+      "learning_rate": 3.797506387401506e-06,
+      "loss": 1.2486,
+      "step": 6293
+    },
+    {
+      "epoch": 2.886136910904713,
+      "grad_norm": 0.3017299175262451,
+      "learning_rate": 3.767103859700749e-06,
+      "loss": 0.9213,
+      "step": 6294
+    },
+    {
+      "epoch": 2.8865955739020754,
+      "grad_norm": 0.3231751322746277,
+      "learning_rate": 3.7368230615074105e-06,
+      "loss": 1.9851,
+      "step": 6295
+    },
+    {
+      "epoch": 2.8870542368994383,
+      "grad_norm": 0.401769757270813,
+      "learning_rate": 3.70666400024966e-06,
+      "loss": 1.3682,
+      "step": 6296
+    },
+    {
+      "epoch": 2.8875128998968007,
+      "grad_norm": 0.3910331428050995,
+      "learning_rate": 3.6766266833256346e-06,
+      "loss": 1.2714,
+      "step": 6297
+    },
+    {
+      "epoch": 2.8879715628941636,
+      "grad_norm": 0.3329823911190033,
+      "learning_rate": 3.646711118103774e-06,
+      "loss": 1.8126,
+      "step": 6298
+    },
+    {
+      "epoch": 2.8884302258915264,
+      "grad_norm": 0.4459758996963501,
+      "learning_rate": 3.616917311922596e-06,
+      "loss": 1.793,
+      "step": 6299
+    },
+    {
+      "epoch": 2.888888888888889,
+      "grad_norm": 0.4213806986808777,
+      "learning_rate": 3.5872452720907e-06,
+      "loss": 1.9231,
+      "step": 6300
+    },
+    {
+      "epoch": 2.8893475518862517,
+      "grad_norm": 0.4440397620201111,
+      "learning_rate": 3.557695005886874e-06,
+      "loss": 1.5308,
+      "step": 6301
+    },
+    {
+      "epoch": 2.889806214883614,
+      "grad_norm": 0.263126939535141,
+      "learning_rate": 3.5282665205599306e-06,
+      "loss": 0.9987,
+      "step": 6302
+    },
+    {
+      "epoch": 2.890264877880977,
+      "grad_norm": 0.2880323529243469,
+      "learning_rate": 3.49895982332904e-06,
+      "loss": 1.6559,
+      "step": 6303
+    },
+    {
+      "epoch": 2.89072354087834,
+      "grad_norm": 0.9521461725234985,
+      "learning_rate": 3.4697749213832284e-06,
+      "loss": 1.5062,
+      "step": 6304
+    },
+    {
+      "epoch": 2.8911822038757022,
+      "grad_norm": 0.28019800782203674,
+      "learning_rate": 3.4407118218818256e-06,
+      "loss": 1.2073,
+      "step": 6305
+    },
+    {
+      "epoch": 2.891640866873065,
+      "grad_norm": 0.35234352946281433,
+      "learning_rate": 3.411770531954128e-06,
+      "loss": 1.5915,
+      "step": 6306
+    },
+    {
+      "epoch": 2.8920995298704275,
+      "grad_norm": 0.37154096364974976,
+      "learning_rate": 3.382951058699735e-06,
+      "loss": 0.9094,
+      "step": 6307
+    },
+    {
+      "epoch": 2.8925581928677904,
+      "grad_norm": 0.08573618531227112,
+      "learning_rate": 3.354253409188268e-06,
+      "loss": 1.1077,
+      "step": 6308
+    },
+    {
+      "epoch": 2.8930168558651532,
+      "grad_norm": 0.3723006844520569,
+      "learning_rate": 3.3256775904594307e-06,
+      "loss": 0.845,
+      "step": 6309
+    },
+    {
+      "epoch": 2.8934755188625156,
+      "grad_norm": 0.10502085089683533,
+      "learning_rate": 3.297223609523059e-06,
+      "loss": 1.2752,
+      "step": 6310
+    },
+    {
+      "epoch": 2.8939341818598785,
+      "grad_norm": 0.4659182131290436,
+      "learning_rate": 3.2688914733591814e-06,
+      "loss": 0.976,
+      "step": 6311
+    },
+    {
+      "epoch": 2.894392844857241,
+      "grad_norm": 0.3109420835971832,
+      "learning_rate": 3.2406811889177933e-06,
+      "loss": 0.9553,
+      "step": 6312
+    },
+    {
+      "epoch": 2.8948515078546038,
+      "grad_norm": 0.2608180344104767,
+      "learning_rate": 3.2125927631191933e-06,
+      "loss": 0.8474,
+      "step": 6313
+    },
+    {
+      "epoch": 2.8953101708519666,
+      "grad_norm": 0.21614636480808258,
+      "learning_rate": 3.184626202853591e-06,
+      "loss": 1.3879,
+      "step": 6314
+    },
+    {
+      "epoch": 2.8957688338493295,
+      "grad_norm": 0.3109268546104431,
+      "learning_rate": 3.1567815149813885e-06,
+      "loss": 1.1218,
+      "step": 6315
+    },
+    {
+      "epoch": 2.896227496846692,
+      "grad_norm": 0.2876843214035034,
+      "learning_rate": 3.129058706333121e-06,
+      "loss": 1.6882,
+      "step": 6316
+    },
+    {
+      "epoch": 2.8966861598440543,
+      "grad_norm": 0.38865727186203003,
+      "learning_rate": 3.1014577837093496e-06,
+      "loss": 0.799,
+      "step": 6317
+    },
+    {
+      "epoch": 2.897144822841417,
+      "grad_norm": 0.21412920951843262,
+      "learning_rate": 3.073978753880824e-06,
+      "loss": 1.5184,
+      "step": 6318
+    },
+    {
+      "epoch": 2.89760348583878,
+      "grad_norm": 0.4019043743610382,
+      "learning_rate": 3.046621623588375e-06,
+      "loss": 1.6321,
+      "step": 6319
+    },
+    {
+      "epoch": 2.898062148836143,
+      "grad_norm": 0.18866117298603058,
+      "learning_rate": 3.0193863995428005e-06,
+      "loss": 1.1956,
+      "step": 6320
+    },
+    {
+      "epoch": 2.8985208118335053,
+      "grad_norm": 0.3983370065689087,
+      "learning_rate": 2.992273088425146e-06,
+      "loss": 0.9323,
+      "step": 6321
+    },
+    {
+      "epoch": 2.898979474830868,
+      "grad_norm": 0.17893198132514954,
+      "learning_rate": 2.96528169688659e-06,
+      "loss": 1.5333,
+      "step": 6322
+    },
+    {
+      "epoch": 2.8994381378282306,
+      "grad_norm": 0.3908742368221283,
+      "learning_rate": 2.938412231548171e-06,
+      "loss": 1.3581,
+      "step": 6323
+    },
+    {
+      "epoch": 2.8998968008255934,
+      "grad_norm": 0.28429678082466125,
+      "learning_rate": 2.911664699001282e-06,
+      "loss": 1.7076,
+      "step": 6324
+    },
+    {
+      "epoch": 2.9003554638229563,
+      "grad_norm": 0.36891356110572815,
+      "learning_rate": 2.8850391058071747e-06,
+      "loss": 1.2607,
+      "step": 6325
+    },
+    {
+      "epoch": 2.9008141268203187,
+      "grad_norm": 0.2793447971343994,
+      "learning_rate": 2.8585354584974022e-06,
+      "loss": 1.29,
+      "step": 6326
+    },
+    {
+      "epoch": 2.9012727898176816,
+      "grad_norm": 0.29457083344459534,
+      "learning_rate": 2.832153763573486e-06,
+      "loss": 0.7886,
+      "step": 6327
+    },
+    {
+      "epoch": 2.901731452815044,
+      "grad_norm": 0.3590116798877716,
+      "learning_rate": 2.8058940275069722e-06,
+      "loss": 1.5443,
+      "step": 6328
+    },
+    {
+      "epoch": 2.902190115812407,
+      "grad_norm": 0.43476438522338867,
+      "learning_rate": 2.7797562567395963e-06,
+      "loss": 1.8802,
+      "step": 6329
+    },
+    {
+      "epoch": 2.9026487788097697,
+      "grad_norm": 0.3066580891609192,
+      "learning_rate": 2.7537404576831737e-06,
+      "loss": 0.789,
+      "step": 6330
+    },
+    {
+      "epoch": 2.903107441807132,
+      "grad_norm": 0.07547781616449356,
+      "learning_rate": 2.727846636719544e-06,
+      "loss": 0.803,
+      "step": 6331
+    },
+    {
+      "epoch": 2.903566104804495,
+      "grad_norm": 0.38495829701423645,
+      "learning_rate": 2.7020748002006824e-06,
+      "loss": 2.0801,
+      "step": 6332
+    },
+    {
+      "epoch": 2.9040247678018574,
+      "grad_norm": 0.5015888810157776,
+      "learning_rate": 2.676424954448531e-06,
+      "loss": 1.3515,
+      "step": 6333
+    },
+    {
+      "epoch": 2.9044834307992202,
+      "grad_norm": 0.4295569658279419,
+      "learning_rate": 2.650897105755279e-06,
+      "loss": 1.6154,
+      "step": 6334
+    },
+    {
+      "epoch": 2.904942093796583,
+      "grad_norm": 0.24834929406642914,
+      "learning_rate": 2.625491260382973e-06,
+      "loss": 0.7925,
+      "step": 6335
+    },
+    {
+      "epoch": 2.9054007567939455,
+      "grad_norm": 0.3073466420173645,
+      "learning_rate": 2.600207424563961e-06,
+      "loss": 1.0556,
+      "step": 6336
+    },
+    {
+      "epoch": 2.9058594197913084,
+      "grad_norm": 0.21239478886127472,
+      "learning_rate": 2.5750456045005035e-06,
+      "loss": 1.2195,
+      "step": 6337
+    },
+    {
+      "epoch": 2.9063180827886708,
+      "grad_norm": 0.32616496086120605,
+      "learning_rate": 2.5500058063649965e-06,
+      "loss": 0.7554,
+      "step": 6338
+    },
+    {
+      "epoch": 2.9067767457860336,
+      "grad_norm": 0.08231624960899353,
+      "learning_rate": 2.5250880362998607e-06,
+      "loss": 1.3146,
+      "step": 6339
+    },
+    {
+      "epoch": 2.9072354087833965,
+      "grad_norm": 0.4305974245071411,
+      "learning_rate": 2.500292300417595e-06,
+      "loss": 1.6447,
+      "step": 6340
+    },
+    {
+      "epoch": 2.9076940717807593,
+      "grad_norm": 0.27547237277030945,
+      "learning_rate": 2.4756186048007225e-06,
+      "loss": 1.3123,
+      "step": 6341
+    },
+    {
+      "epoch": 2.9081527347781218,
+      "grad_norm": 0.25122833251953125,
+      "learning_rate": 2.4510669555020125e-06,
+      "loss": 1.1544,
+      "step": 6342
+    },
+    {
+      "epoch": 2.9086113977754846,
+      "grad_norm": 0.27697306871414185,
+      "learning_rate": 2.4266373585440924e-06,
+      "loss": 0.4727,
+      "step": 6343
+    },
+    {
+      "epoch": 2.909070060772847,
+      "grad_norm": 0.3418446481227875,
+      "learning_rate": 2.402329819919724e-06,
+      "loss": 1.6129,
+      "step": 6344
+    },
+    {
+      "epoch": 2.90952872377021,
+      "grad_norm": 0.4216833710670471,
+      "learning_rate": 2.3781443455916927e-06,
+      "loss": 1.2566,
+      "step": 6345
+    },
+    {
+      "epoch": 2.9099873867675727,
+      "grad_norm": 0.3174903690814972,
+      "learning_rate": 2.3540809414929196e-06,
+      "loss": 1.3995,
+      "step": 6346
+    },
+    {
+      "epoch": 2.910446049764935,
+      "grad_norm": 0.2971450984477997,
+      "learning_rate": 2.3301396135262387e-06,
+      "loss": 1.087,
+      "step": 6347
+    },
+    {
+      "epoch": 2.910904712762298,
+      "grad_norm": 0.41538989543914795,
+      "learning_rate": 2.3063203675647848e-06,
+      "loss": 1.3513,
+      "step": 6348
+    },
+    {
+      "epoch": 2.9113633757596604,
+      "grad_norm": 0.36953091621398926,
+      "learning_rate": 2.28262320945144e-06,
+      "loss": 1.7742,
+      "step": 6349
+    },
+    {
+      "epoch": 2.9118220387570233,
+      "grad_norm": 0.31279754638671875,
+      "learning_rate": 2.259048144999387e-06,
+      "loss": 1.3542,
+      "step": 6350
+    },
+    {
+      "epoch": 2.912280701754386,
+      "grad_norm": 0.2468571960926056,
+      "learning_rate": 2.2355951799916674e-06,
+      "loss": 1.3647,
+      "step": 6351
+    },
+    {
+      "epoch": 2.9127393647517485,
+      "grad_norm": 0.577455461025238,
+      "learning_rate": 2.212264320181567e-06,
+      "loss": 1.2788,
+      "step": 6352
+    },
+    {
+      "epoch": 2.9131980277491114,
+      "grad_norm": 0.3301372230052948,
+      "learning_rate": 2.1890555712922313e-06,
+      "loss": 1.6881,
+      "step": 6353
+    },
+    {
+      "epoch": 2.913656690746474,
+      "grad_norm": 0.3213047385215759,
+      "learning_rate": 2.1659689390169934e-06,
+      "loss": 1.3643,
+      "step": 6354
+    },
+    {
+      "epoch": 2.9141153537438367,
+      "grad_norm": 0.3952171504497528,
+      "learning_rate": 2.1430044290191573e-06,
+      "loss": 1.7974,
+      "step": 6355
+    },
+    {
+      "epoch": 2.9145740167411995,
+      "grad_norm": 0.3204817771911621,
+      "learning_rate": 2.1201620469320503e-06,
+      "loss": 1.4843,
+      "step": 6356
+    },
+    {
+      "epoch": 2.915032679738562,
+      "grad_norm": 0.4052104949951172,
+      "learning_rate": 2.0974417983590787e-06,
+      "loss": 1.7517,
+      "step": 6357
+    },
+    {
+      "epoch": 2.915491342735925,
+      "grad_norm": 0.28286999464035034,
+      "learning_rate": 2.0748436888737286e-06,
+      "loss": 1.0675,
+      "step": 6358
+    },
+    {
+      "epoch": 2.915950005733287,
+      "grad_norm": 0.4025648534297943,
+      "learning_rate": 2.0523677240193994e-06,
+      "loss": 1.6526,
+      "step": 6359
+    },
+    {
+      "epoch": 2.91640866873065,
+      "grad_norm": 0.41483741998672485,
+      "learning_rate": 2.030013909309736e-06,
+      "loss": 2.1431,
+      "step": 6360
+    },
+    {
+      "epoch": 2.916867331728013,
+      "grad_norm": 0.36023521423339844,
+      "learning_rate": 2.0077822502281295e-06,
+      "loss": 1.3888,
+      "step": 6361
+    },
+    {
+      "epoch": 2.917325994725376,
+      "grad_norm": 0.3408292233943939,
+      "learning_rate": 1.9856727522282734e-06,
+      "loss": 1.2604,
+      "step": 6362
+    },
+    {
+      "epoch": 2.917784657722738,
+      "grad_norm": 0.2545235753059387,
+      "learning_rate": 1.963685420733774e-06,
+      "loss": 1.4949,
+      "step": 6363
+    },
+    {
+      "epoch": 2.918243320720101,
+      "grad_norm": 0.27326932549476624,
+      "learning_rate": 1.9418202611382607e-06,
+      "loss": 0.7846,
+      "step": 6364
+    },
+    {
+      "epoch": 2.9187019837174635,
+      "grad_norm": 0.33998316526412964,
+      "learning_rate": 1.920077278805443e-06,
+      "loss": 1.5005,
+      "step": 6365
+    },
+    {
+      "epoch": 2.9191606467148263,
+      "grad_norm": 0.3685235381126404,
+      "learning_rate": 1.8984564790689996e-06,
+      "loss": 1.4185,
+      "step": 6366
+    },
+    {
+      "epoch": 2.919619309712189,
+      "grad_norm": 0.18980297446250916,
+      "learning_rate": 1.8769578672326316e-06,
+      "loss": 1.2739,
+      "step": 6367
+    },
+    {
+      "epoch": 2.9200779727095516,
+      "grad_norm": 0.30294135212898254,
+      "learning_rate": 1.8555814485702316e-06,
+      "loss": 1.1948,
+      "step": 6368
+    },
+    {
+      "epoch": 2.9205366357069145,
+      "grad_norm": 0.3762916922569275,
+      "learning_rate": 1.8343272283254386e-06,
+      "loss": 1.7822,
+      "step": 6369
+    },
+    {
+      "epoch": 2.920995298704277,
+      "grad_norm": 0.3765801787376404,
+      "learning_rate": 1.813195211712193e-06,
+      "loss": 1.1016,
+      "step": 6370
+    },
+    {
+      "epoch": 2.9214539617016397,
+      "grad_norm": 0.22130340337753296,
+      "learning_rate": 1.792185403914237e-06,
+      "loss": 1.6736,
+      "step": 6371
+    },
+    {
+      "epoch": 2.9219126246990026,
+      "grad_norm": 0.33724433183670044,
+      "learning_rate": 1.7712978100854482e-06,
+      "loss": 1.2495,
+      "step": 6372
+    },
+    {
+      "epoch": 2.922371287696365,
+      "grad_norm": 0.30301743745803833,
+      "learning_rate": 1.7505324353497831e-06,
+      "loss": 1.7449,
+      "step": 6373
+    },
+    {
+      "epoch": 2.922829950693728,
+      "grad_norm": 0.2647053003311157,
+      "learning_rate": 1.7298892848010006e-06,
+      "loss": 1.3491,
+      "step": 6374
+    },
+    {
+      "epoch": 2.9232886136910903,
+      "grad_norm": 0.47797808051109314,
+      "learning_rate": 1.7093683635031609e-06,
+      "loss": 1.2687,
+      "step": 6375
+    },
+    {
+      "epoch": 2.923747276688453,
+      "grad_norm": 0.32065433263778687,
+      "learning_rate": 1.6889696764900708e-06,
+      "loss": 1.3747,
+      "step": 6376
+    },
+    {
+      "epoch": 2.924205939685816,
+      "grad_norm": 0.24311882257461548,
+      "learning_rate": 1.668693228765783e-06,
+      "loss": 0.6366,
+      "step": 6377
+    },
+    {
+      "epoch": 2.9246646026831784,
+      "grad_norm": 0.15887467563152313,
+      "learning_rate": 1.6485390253041521e-06,
+      "loss": 1.2029,
+      "step": 6378
+    },
+    {
+      "epoch": 2.9251232656805413,
+      "grad_norm": 0.33973756432533264,
+      "learning_rate": 1.6285070710492233e-06,
+      "loss": 1.6896,
+      "step": 6379
+    },
+    {
+      "epoch": 2.9255819286779037,
+      "grad_norm": 0.38102298974990845,
+      "learning_rate": 1.6085973709149548e-06,
+      "loss": 1.0677,
+      "step": 6380
+    },
+    {
+      "epoch": 2.9260405916752665,
+      "grad_norm": 0.3507750332355499,
+      "learning_rate": 1.5888099297853288e-06,
+      "loss": 1.2086,
+      "step": 6381
+    },
+    {
+      "epoch": 2.9264992546726294,
+      "grad_norm": 0.1782960146665573,
+      "learning_rate": 1.5691447525143509e-06,
+      "loss": 1.05,
+      "step": 6382
+    },
+    {
+      "epoch": 2.9269579176699922,
+      "grad_norm": 0.25808510184288025,
+      "learning_rate": 1.5496018439260518e-06,
+      "loss": 0.8826,
+      "step": 6383
+    },
+    {
+      "epoch": 2.9274165806673547,
+      "grad_norm": 0.2699086368083954,
+      "learning_rate": 1.5301812088144296e-06,
+      "loss": 0.7635,
+      "step": 6384
+    },
+    {
+      "epoch": 2.927875243664717,
+      "grad_norm": 0.12165196985006332,
+      "learning_rate": 1.510882851943507e-06,
+      "loss": 1.2339,
+      "step": 6385
+    },
+    {
+      "epoch": 2.92833390666208,
+      "grad_norm": 0.27769702672958374,
+      "learning_rate": 1.4917067780473304e-06,
+      "loss": 0.799,
+      "step": 6386
+    },
+    {
+      "epoch": 2.928792569659443,
+      "grad_norm": 0.29953867197036743,
+      "learning_rate": 1.4726529918299148e-06,
+      "loss": 1.9141,
+      "step": 6387
+    },
+    {
+      "epoch": 2.9292512326568056,
+      "grad_norm": 0.39282310009002686,
+      "learning_rate": 1.4537214979652435e-06,
+      "loss": 1.6138,
+      "step": 6388
+    },
+    {
+      "epoch": 2.929709895654168,
+      "grad_norm": 0.28682559728622437,
+      "learning_rate": 1.4349123010974908e-06,
+      "loss": 0.9948,
+      "step": 6389
+    },
+    {
+      "epoch": 2.930168558651531,
+      "grad_norm": 0.3742808699607849,
+      "learning_rate": 1.416225405840521e-06,
+      "loss": 1.5709,
+      "step": 6390
+    },
+    {
+      "epoch": 2.9306272216488933,
+      "grad_norm": 0.36167392134666443,
+      "learning_rate": 1.39766081677839e-06,
+      "loss": 1.3771,
+      "step": 6391
+    },
+    {
+      "epoch": 2.931085884646256,
+      "grad_norm": 0.33344385027885437,
+      "learning_rate": 1.3792185384652322e-06,
+      "loss": 1.4467,
+      "step": 6392
+    },
+    {
+      "epoch": 2.931544547643619,
+      "grad_norm": 0.3055780231952667,
+      "learning_rate": 1.3608985754249848e-06,
+      "loss": 1.4431,
+      "step": 6393
+    },
+    {
+      "epoch": 2.9320032106409815,
+      "grad_norm": 0.39428263902664185,
+      "learning_rate": 1.3427009321517191e-06,
+      "loss": 2.0354,
+      "step": 6394
+    },
+    {
+      "epoch": 2.9324618736383443,
+      "grad_norm": 0.4126293361186981,
+      "learning_rate": 1.3246256131093647e-06,
+      "loss": 1.7002,
+      "step": 6395
+    },
+    {
+      "epoch": 2.9329205366357067,
+      "grad_norm": 0.2903102934360504,
+      "learning_rate": 1.3066726227320414e-06,
+      "loss": 1.148,
+      "step": 6396
+    },
+    {
+      "epoch": 2.9333791996330696,
+      "grad_norm": 0.35992351174354553,
+      "learning_rate": 1.2888419654236706e-06,
+      "loss": 1.39,
+      "step": 6397
+    },
+    {
+      "epoch": 2.9338378626304324,
+      "grad_norm": 0.37300676107406616,
+      "learning_rate": 1.2711336455582533e-06,
+      "loss": 1.2245,
+      "step": 6398
+    },
+    {
+      "epoch": 2.934296525627795,
+      "grad_norm": 0.21308903396129608,
+      "learning_rate": 1.2535476674797596e-06,
+      "loss": 1.2061,
+      "step": 6399
+    },
+    {
+      "epoch": 2.9347551886251577,
+      "grad_norm": 0.2780902683734894,
+      "learning_rate": 1.2360840355022386e-06,
+      "loss": 1.5712,
+      "step": 6400
+    },
+    {
+      "epoch": 2.93521385162252,
+      "grad_norm": 0.36661916971206665,
+      "learning_rate": 1.2187427539094854e-06,
+      "loss": 1.1625,
+      "step": 6401
+    },
+    {
+      "epoch": 2.935672514619883,
+      "grad_norm": 0.28895169496536255,
+      "learning_rate": 1.2015238269555972e-06,
+      "loss": 0.7739,
+      "step": 6402
+    },
+    {
+      "epoch": 2.936131177617246,
+      "grad_norm": 0.2886188328266144,
+      "learning_rate": 1.184427258864418e-06,
+      "loss": 1.7198,
+      "step": 6403
+    },
+    {
+      "epoch": 2.9365898406146083,
+      "grad_norm": 0.36322250962257385,
+      "learning_rate": 1.16745305382987e-06,
+      "loss": 0.7532,
+      "step": 6404
+    },
+    {
+      "epoch": 2.937048503611971,
+      "grad_norm": 0.3152855634689331,
+      "learning_rate": 1.1506012160158452e-06,
+      "loss": 1.7031,
+      "step": 6405
+    },
+    {
+      "epoch": 2.9375071666093335,
+      "grad_norm": 0.39350810647010803,
+      "learning_rate": 1.133871749556259e-06,
+      "loss": 1.6578,
+      "step": 6406
+    },
+    {
+      "epoch": 2.9379658296066964,
+      "grad_norm": 0.23590238392353058,
+      "learning_rate": 1.1172646585549396e-06,
+      "loss": 1.6073,
+      "step": 6407
+    },
+    {
+      "epoch": 2.9384244926040592,
+      "grad_norm": 0.3189743161201477,
+      "learning_rate": 1.1007799470857393e-06,
+      "loss": 1.1992,
+      "step": 6408
+    },
+    {
+      "epoch": 2.938883155601422,
+      "grad_norm": 0.45767655968666077,
+      "learning_rate": 1.0844176191924237e-06,
+      "loss": 1.8853,
+      "step": 6409
+    },
+    {
+      "epoch": 2.9393418185987845,
+      "grad_norm": 0.3135172128677368,
+      "learning_rate": 1.068177678888893e-06,
+      "loss": 1.7242,
+      "step": 6410
+    },
+    {
+      "epoch": 2.9398004815961474,
+      "grad_norm": 0.3678114712238312,
+      "learning_rate": 1.0520601301588494e-06,
+      "loss": 1.4546,
+      "step": 6411
+    },
+    {
+      "epoch": 2.94025914459351,
+      "grad_norm": 0.33061683177948,
+      "learning_rate": 1.0360649769560193e-06,
+      "loss": 1.214,
+      "step": 6412
+    },
+    {
+      "epoch": 2.9407178075908726,
+      "grad_norm": 0.31986305117607117,
+      "learning_rate": 1.0201922232041528e-06,
+      "loss": 1.6663,
+      "step": 6413
+    },
+    {
+      "epoch": 2.9411764705882355,
+      "grad_norm": 0.4389913082122803,
+      "learning_rate": 1.0044418727970238e-06,
+      "loss": 1.4183,
+      "step": 6414
+    },
+    {
+      "epoch": 2.941635133585598,
+      "grad_norm": 0.22724516689777374,
+      "learning_rate": 9.88813929598209e-07,
+      "loss": 1.3397,
+      "step": 6415
+    },
+    {
+      "epoch": 2.9420937965829608,
+      "grad_norm": 0.41958609223365784,
+      "learning_rate": 9.733083974414193e-07,
+      "loss": 1.3082,
+      "step": 6416
+    },
+    {
+      "epoch": 2.942552459580323,
+      "grad_norm": 0.3317757546901703,
+      "learning_rate": 9.579252801302785e-07,
+      "loss": 1.5914,
+      "step": 6417
+    },
+    {
+      "epoch": 2.943011122577686,
+      "grad_norm": 0.2724321782588959,
+      "learning_rate": 9.426645814382683e-07,
+      "loss": 1.1695,
+      "step": 6418
+    },
+    {
+      "epoch": 2.943469785575049,
+      "grad_norm": 0.3204740285873413,
+      "learning_rate": 9.27526305109061e-07,
+      "loss": 1.5457,
+      "step": 6419
+    },
+    {
+      "epoch": 2.9439284485724113,
+      "grad_norm": 0.34586411714553833,
+      "learning_rate": 9.125104548561857e-07,
+      "loss": 1.2657,
+      "step": 6420
+    },
+    {
+      "epoch": 2.944387111569774,
+      "grad_norm": 0.3183816075325012,
+      "learning_rate": 8.976170343630297e-07,
+      "loss": 0.9479,
+      "step": 6421
+    },
+    {
+      "epoch": 2.9448457745671366,
+      "grad_norm": 0.23148377239704132,
+      "learning_rate": 8.828460472832256e-07,
+      "loss": 1.7264,
+      "step": 6422
+    },
+    {
+      "epoch": 2.9453044375644994,
+      "grad_norm": 0.37857529520988464,
+      "learning_rate": 8.68197497240042e-07,
+      "loss": 0.9766,
+      "step": 6423
+    },
+    {
+      "epoch": 2.9457631005618623,
+      "grad_norm": 0.10237754881381989,
+      "learning_rate": 8.536713878269376e-07,
+      "loss": 0.4709,
+      "step": 6424
+    },
+    {
+      "epoch": 2.9462217635592247,
+      "grad_norm": 0.21351243555545807,
+      "learning_rate": 8.392677226072843e-07,
+      "loss": 1.2244,
+      "step": 6425
+    },
+    {
+      "epoch": 2.9466804265565876,
+      "grad_norm": 0.2921064794063568,
+      "learning_rate": 8.249865051143668e-07,
+      "loss": 1.2155,
+      "step": 6426
+    },
+    {
+      "epoch": 2.94713908955395,
+      "grad_norm": 0.3035675585269928,
+      "learning_rate": 8.108277388515495e-07,
+      "loss": 1.0669,
+      "step": 6427
+    },
+    {
+      "epoch": 2.947597752551313,
+      "grad_norm": 0.2804781198501587,
+      "learning_rate": 7.967914272919985e-07,
+      "loss": 0.9862,
+      "step": 6428
+    },
+    {
+      "epoch": 2.9480564155486757,
+      "grad_norm": 0.34943005442619324,
+      "learning_rate": 7.828775738789595e-07,
+      "loss": 1.194,
+      "step": 6429
+    },
+    {
+      "epoch": 2.9485150785460386,
+      "grad_norm": 0.13486827909946442,
+      "learning_rate": 7.690861820255912e-07,
+      "loss": 0.9005,
+      "step": 6430
+    },
+    {
+      "epoch": 2.948973741543401,
+      "grad_norm": 0.3738475739955902,
+      "learning_rate": 7.554172551150206e-07,
+      "loss": 1.5659,
+      "step": 6431
+    },
+    {
+      "epoch": 2.949432404540764,
+      "grad_norm": 0.2654702961444855,
+      "learning_rate": 7.418707965003435e-07,
+      "loss": 1.1194,
+      "step": 6432
+    },
+    {
+      "epoch": 2.9498910675381262,
+      "grad_norm": 0.36275702714920044,
+      "learning_rate": 7.284468095045682e-07,
+      "loss": 1.5873,
+      "step": 6433
+    },
+    {
+      "epoch": 2.950349730535489,
+      "grad_norm": 0.386787474155426,
+      "learning_rate": 7.151452974207828e-07,
+      "loss": 1.2502,
+      "step": 6434
+    },
+    {
+      "epoch": 2.950808393532852,
+      "grad_norm": 0.1252823770046234,
+      "learning_rate": 7.019662635118218e-07,
+      "loss": 0.8411,
+      "step": 6435
+    },
+    {
+      "epoch": 2.9512670565302144,
+      "grad_norm": 0.24427084624767303,
+      "learning_rate": 6.8890971101071e-07,
+      "loss": 0.7842,
+      "step": 6436
+    },
+    {
+      "epoch": 2.9517257195275772,
+      "grad_norm": 0.26967909932136536,
+      "learning_rate": 6.759756431202746e-07,
+      "loss": 1.2764,
+      "step": 6437
+    },
+    {
+      "epoch": 2.9521843825249396,
+      "grad_norm": 0.3875824511051178,
+      "learning_rate": 6.631640630133106e-07,
+      "loss": 1.6353,
+      "step": 6438
+    },
+    {
+      "epoch": 2.9526430455223025,
+      "grad_norm": 0.3367568850517273,
+      "learning_rate": 6.504749738325822e-07,
+      "loss": 1.6357,
+      "step": 6439
+    },
+    {
+      "epoch": 2.9531017085196654,
+      "grad_norm": 0.36924004554748535,
+      "learning_rate": 6.379083786908768e-07,
+      "loss": 1.1762,
+      "step": 6440
+    },
+    {
+      "epoch": 2.9535603715170278,
+      "grad_norm": 0.09012544900178909,
+      "learning_rate": 6.254642806707845e-07,
+      "loss": 0.7179,
+      "step": 6441
+    },
+    {
+      "epoch": 2.9540190345143906,
+      "grad_norm": 0.40828028321266174,
+      "learning_rate": 6.131426828250297e-07,
+      "loss": 2.0659,
+      "step": 6442
+    },
+    {
+      "epoch": 2.954477697511753,
+      "grad_norm": 0.7955999374389648,
+      "learning_rate": 6.009435881760838e-07,
+      "loss": 1.0877,
+      "step": 6443
+    },
+    {
+      "epoch": 2.954936360509116,
+      "grad_norm": 0.1659400314092636,
+      "learning_rate": 5.888669997165529e-07,
+      "loss": 1.1334,
+      "step": 6444
+    },
+    {
+      "epoch": 2.9553950235064788,
+      "grad_norm": 0.3118785619735718,
+      "learning_rate": 5.769129204089007e-07,
+      "loss": 1.781,
+      "step": 6445
+    },
+    {
+      "epoch": 2.955853686503841,
+      "grad_norm": 0.4559882879257202,
+      "learning_rate": 5.650813531855592e-07,
+      "loss": 2.0151,
+      "step": 6446
+    },
+    {
+      "epoch": 2.956312349501204,
+      "grad_norm": 0.4162604808807373,
+      "learning_rate": 5.533723009488734e-07,
+      "loss": 1.5276,
+      "step": 6447
+    },
+    {
+      "epoch": 2.9567710124985664,
+      "grad_norm": 0.3055538237094879,
+      "learning_rate": 5.417857665711012e-07,
+      "loss": 0.8876,
+      "step": 6448
+    },
+    {
+      "epoch": 2.9572296754959293,
+      "grad_norm": 0.28854599595069885,
+      "learning_rate": 5.303217528945802e-07,
+      "loss": 1.6246,
+      "step": 6449
+    },
+    {
+      "epoch": 2.957688338493292,
+      "grad_norm": 0.3452959358692169,
+      "learning_rate": 5.18980262731561e-07,
+      "loss": 0.9966,
+      "step": 6450
+    },
+    {
+      "epoch": 2.958147001490655,
+      "grad_norm": 0.16069626808166504,
+      "learning_rate": 5.077612988640401e-07,
+      "loss": 0.626,
+      "step": 6451
+    },
+    {
+      "epoch": 2.9586056644880174,
+      "grad_norm": 0.2646504342556,
+      "learning_rate": 4.966648640442606e-07,
+      "loss": 1.5773,
+      "step": 6452
+    },
+    {
+      "epoch": 2.95906432748538,
+      "grad_norm": 0.40957656502723694,
+      "learning_rate": 4.856909609941562e-07,
+      "loss": 1.1276,
+      "step": 6453
+    },
+    {
+      "epoch": 2.9595229904827427,
+      "grad_norm": 0.27915382385253906,
+      "learning_rate": 4.7483959240574025e-07,
+      "loss": 1.4488,
+      "step": 6454
+    },
+    {
+      "epoch": 2.9599816534801056,
+      "grad_norm": 0.36528947949409485,
+      "learning_rate": 4.6411076094099447e-07,
+      "loss": 1.6633,
+      "step": 6455
+    },
+    {
+      "epoch": 2.9604403164774684,
+      "grad_norm": 0.3213283121585846,
+      "learning_rate": 4.53504469231647e-07,
+      "loss": 1.697,
+      "step": 6456
+    },
+    {
+      "epoch": 2.960898979474831,
+      "grad_norm": 0.3525945842266083,
+      "learning_rate": 4.430207198796166e-07,
+      "loss": 1.1332,
+      "step": 6457
+    },
+    {
+      "epoch": 2.9613576424721937,
+      "grad_norm": 0.3523452877998352,
+      "learning_rate": 4.3265951545656823e-07,
+      "loss": 1.9775,
+      "step": 6458
+    },
+    {
+      "epoch": 2.961816305469556,
+      "grad_norm": 0.3788760006427765,
+      "learning_rate": 4.224208585042466e-07,
+      "loss": 1.3544,
+      "step": 6459
+    },
+    {
+      "epoch": 2.962274968466919,
+      "grad_norm": 0.3480757772922516,
+      "learning_rate": 4.123047515341982e-07,
+      "loss": 1.3392,
+      "step": 6460
+    },
+    {
+      "epoch": 2.962733631464282,
+      "grad_norm": 0.33295026421546936,
+      "learning_rate": 4.0231119702799355e-07,
+      "loss": 1.4835,
+      "step": 6461
+    },
+    {
+      "epoch": 2.9631922944616442,
+      "grad_norm": 0.32605263590812683,
+      "learning_rate": 3.924401974371716e-07,
+      "loss": 1.1706,
+      "step": 6462
+    },
+    {
+      "epoch": 2.963650957459007,
+      "grad_norm": 0.3713498115539551,
+      "learning_rate": 3.826917551831288e-07,
+      "loss": 1.5077,
+      "step": 6463
+    },
+    {
+      "epoch": 2.9641096204563695,
+      "grad_norm": 0.41880208253860474,
+      "learning_rate": 3.7306587265717453e-07,
+      "loss": 1.3178,
+      "step": 6464
+    },
+    {
+      "epoch": 2.9645682834537324,
+      "grad_norm": 0.19970370829105377,
+      "learning_rate": 3.6356255222069756e-07,
+      "loss": 0.8712,
+      "step": 6465
+    },
+    {
+      "epoch": 2.965026946451095,
+      "grad_norm": 0.27967116236686707,
+      "learning_rate": 3.5418179620488876e-07,
+      "loss": 1.4177,
+      "step": 6466
+    },
+    {
+      "epoch": 2.9654856094484576,
+      "grad_norm": 0.43477344512939453,
+      "learning_rate": 3.449236069109074e-07,
+      "loss": 1.1676,
+      "step": 6467
+    },
+    {
+      "epoch": 2.9659442724458205,
+      "grad_norm": 0.274976909160614,
+      "learning_rate": 3.357879866098812e-07,
+      "loss": 1.6171,
+      "step": 6468
+    },
+    {
+      "epoch": 2.966402935443183,
+      "grad_norm": 0.28893306851387024,
+      "learning_rate": 3.267749375427953e-07,
+      "loss": 0.8573,
+      "step": 6469
+    },
+    {
+      "epoch": 2.9668615984405458,
+      "grad_norm": 0.12806662917137146,
+      "learning_rate": 3.178844619207144e-07,
+      "loss": 0.8589,
+      "step": 6470
+    },
+    {
+      "epoch": 2.9673202614379086,
+      "grad_norm": 0.24989382922649384,
+      "learning_rate": 3.091165619243941e-07,
+      "loss": 1.3807,
+      "step": 6471
+    },
+    {
+      "epoch": 2.967778924435271,
+      "grad_norm": 0.2839709520339966,
+      "learning_rate": 3.004712397047804e-07,
+      "loss": 1.2,
+      "step": 6472
+    },
+    {
+      "epoch": 2.968237587432634,
+      "grad_norm": 0.4096415638923645,
+      "learning_rate": 2.919484973826214e-07,
+      "loss": 1.9077,
+      "step": 6473
+    },
+    {
+      "epoch": 2.9686962504299963,
+      "grad_norm": 0.32572391629219055,
+      "learning_rate": 2.8354833704852257e-07,
+      "loss": 1.0639,
+      "step": 6474
+    },
+    {
+      "epoch": 2.969154913427359,
+      "grad_norm": 0.3826916515827179,
+      "learning_rate": 2.752707607631688e-07,
+      "loss": 1.9023,
+      "step": 6475
+    },
+    {
+      "epoch": 2.969613576424722,
+      "grad_norm": 0.48622024059295654,
+      "learning_rate": 2.6711577055710255e-07,
+      "loss": 1.61,
+      "step": 6476
+    },
+    {
+      "epoch": 2.970072239422085,
+      "grad_norm": 0.262005478143692,
+      "learning_rate": 2.590833684307792e-07,
+      "loss": 1.0063,
+      "step": 6477
+    },
+    {
+      "epoch": 2.9705309024194473,
+      "grad_norm": 0.21147648990154266,
+      "learning_rate": 2.51173556354678e-07,
+      "loss": 0.8564,
+      "step": 6478
+    },
+    {
+      "epoch": 2.97098956541681,
+      "grad_norm": 0.34185197949409485,
+      "learning_rate": 2.433863362690247e-07,
+      "loss": 1.8838,
+      "step": 6479
+    },
+    {
+      "epoch": 2.9714482284141726,
+      "grad_norm": 0.334324449300766,
+      "learning_rate": 2.357217100841802e-07,
+      "loss": 0.827,
+      "step": 6480
+    },
+    {
+      "epoch": 2.9719068914115354,
+      "grad_norm": 0.28424781560897827,
+      "learning_rate": 2.2817967968025155e-07,
+      "loss": 1.6332,
+      "step": 6481
+    },
+    {
+      "epoch": 2.9723655544088983,
+      "grad_norm": 0.31317272782325745,
+      "learning_rate": 2.2076024690742546e-07,
+      "loss": 1.1931,
+      "step": 6482
+    },
+    {
+      "epoch": 2.9728242174062607,
+      "grad_norm": 0.24583421647548676,
+      "learning_rate": 2.1346341358569054e-07,
+      "loss": 0.6048,
+      "step": 6483
+    },
+    {
+      "epoch": 2.9732828804036235,
+      "grad_norm": 0.38179898262023926,
+      "learning_rate": 2.0628918150500387e-07,
+      "loss": 1.2948,
+      "step": 6484
+    },
+    {
+      "epoch": 2.973741543400986,
+      "grad_norm": 0.18613789975643158,
+      "learning_rate": 1.9923755242529094e-07,
+      "loss": 1.0493,
+      "step": 6485
+    },
+    {
+      "epoch": 2.974200206398349,
+      "grad_norm": 0.25649017095565796,
+      "learning_rate": 1.9230852807639031e-07,
+      "loss": 1.1181,
+      "step": 6486
+    },
+    {
+      "epoch": 2.9746588693957117,
+      "grad_norm": 0.30037227272987366,
+      "learning_rate": 1.8550211015794239e-07,
+      "loss": 1.7121,
+      "step": 6487
+    },
+    {
+      "epoch": 2.975117532393074,
+      "grad_norm": 0.31024396419525146,
+      "learning_rate": 1.7881830033972258e-07,
+      "loss": 1.5497,
+      "step": 6488
+    },
+    {
+      "epoch": 2.975576195390437,
+      "grad_norm": 0.4017513692378998,
+      "learning_rate": 1.7225710026125275e-07,
+      "loss": 1.1522,
+      "step": 6489
+    },
+    {
+      "epoch": 2.9760348583877994,
+      "grad_norm": 0.34826523065567017,
+      "learning_rate": 1.658185115320787e-07,
+      "loss": 1.6506,
+      "step": 6490
+    },
+    {
+      "epoch": 2.976493521385162,
+      "grad_norm": 0.36130425333976746,
+      "learning_rate": 1.5950253573160377e-07,
+      "loss": 1.4875,
+      "step": 6491
+    },
+    {
+      "epoch": 2.976952184382525,
+      "grad_norm": 0.31285083293914795,
+      "learning_rate": 1.5330917440919967e-07,
+      "loss": 1.5449,
+      "step": 6492
+    },
+    {
+      "epoch": 2.9774108473798875,
+      "grad_norm": 0.3980162441730499,
+      "learning_rate": 1.472384290841511e-07,
+      "loss": 1.8258,
+      "step": 6493
+    },
+    {
+      "epoch": 2.9778695103772503,
+      "grad_norm": 0.3458695709705353,
+      "learning_rate": 1.4129030124560016e-07,
+      "loss": 0.7553,
+      "step": 6494
+    },
+    {
+      "epoch": 2.9783281733746128,
+      "grad_norm": 0.2076312005519867,
+      "learning_rate": 1.3546479235276854e-07,
+      "loss": 1.4727,
+      "step": 6495
+    },
+    {
+      "epoch": 2.9787868363719756,
+      "grad_norm": 0.3255481421947479,
+      "learning_rate": 1.2976190383456876e-07,
+      "loss": 0.87,
+      "step": 6496
+    },
+    {
+      "epoch": 2.9792454993693385,
+      "grad_norm": 0.2653532326221466,
+      "learning_rate": 1.2418163709004836e-07,
+      "loss": 1.4395,
+      "step": 6497
+    },
+    {
+      "epoch": 2.9797041623667013,
+      "grad_norm": 0.22563296556472778,
+      "learning_rate": 1.1872399348805685e-07,
+      "loss": 1.1741,
+      "step": 6498
+    },
+    {
+      "epoch": 2.9801628253640637,
+      "grad_norm": 0.3683829605579376,
+      "learning_rate": 1.1338897436741213e-07,
+      "loss": 0.9402,
+      "step": 6499
+    },
+    {
+      "epoch": 2.9806214883614266,
+      "grad_norm": 0.10145165026187897,
+      "learning_rate": 1.0817658103684513e-07,
+      "loss": 0.6497,
+      "step": 6500
+    },
+    {
+      "epoch": 2.981080151358789,
+      "grad_norm": 0.24345941841602325,
+      "learning_rate": 1.030868147749442e-07,
+      "loss": 1.7396,
+      "step": 6501
+    },
+    {
+      "epoch": 2.981538814356152,
+      "grad_norm": 0.37198102474212646,
+      "learning_rate": 9.811967683026613e-08,
+      "loss": 1.5526,
+      "step": 6502
+    },
+    {
+      "epoch": 2.9819974773535147,
+      "grad_norm": 0.3141441345214844,
+      "learning_rate": 9.32751684213362e-08,
+      "loss": 1.2018,
+      "step": 6503
+    },
+    {
+      "epoch": 2.982456140350877,
+      "grad_norm": 0.40361887216567993,
+      "learning_rate": 8.855329073653717e-08,
+      "loss": 1.7252,
+      "step": 6504
+    },
+    {
+      "epoch": 2.98291480334824,
+      "grad_norm": 0.4085117280483246,
+      "learning_rate": 8.395404493410919e-08,
+      "loss": 1.1441,
+      "step": 6505
+    },
+    {
+      "epoch": 2.9833734663456024,
+      "grad_norm": 0.20205476880073547,
+      "learning_rate": 7.947743214237191e-08,
+      "loss": 0.9854,
+      "step": 6506
+    },
+    {
+      "epoch": 2.9838321293429653,
+      "grad_norm": 0.31978803873062134,
+      "learning_rate": 7.512345345944693e-08,
+      "loss": 1.3225,
+      "step": 6507
+    },
+    {
+      "epoch": 2.984290792340328,
+      "grad_norm": 0.30714917182922363,
+      "learning_rate": 7.089210995336881e-08,
+      "loss": 1.1396,
+      "step": 6508
+    },
+    {
+      "epoch": 2.9847494553376905,
+      "grad_norm": 0.07466486096382141,
+      "learning_rate": 6.678340266214056e-08,
+      "loss": 1.0693,
+      "step": 6509
+    },
+    {
+      "epoch": 2.9852081183350534,
+      "grad_norm": 0.3598880469799042,
+      "learning_rate": 6.279733259362264e-08,
+      "loss": 0.9477,
+      "step": 6510
+    },
+    {
+      "epoch": 2.985666781332416,
+      "grad_norm": 0.12736907601356506,
+      "learning_rate": 5.8933900725699486e-08,
+      "loss": 0.9232,
+      "step": 6511
+    },
+    {
+      "epoch": 2.9861254443297787,
+      "grad_norm": 0.20195496082305908,
+      "learning_rate": 5.519310800600197e-08,
+      "loss": 0.8135,
+      "step": 6512
+    },
+    {
+      "epoch": 2.9865841073271415,
+      "grad_norm": 0.2941180169582367,
+      "learning_rate": 5.157495535229595e-08,
+      "loss": 1.6885,
+      "step": 6513
+    },
+    {
+      "epoch": 2.987042770324504,
+      "grad_norm": 0.5080846548080444,
+      "learning_rate": 4.807944365198269e-08,
+      "loss": 1.3666,
+      "step": 6514
+    },
+    {
+      "epoch": 2.987501433321867,
+      "grad_norm": 0.2823677659034729,
+      "learning_rate": 4.470657376265397e-08,
+      "loss": 0.9075,
+      "step": 6515
+    },
+    {
+      "epoch": 2.987960096319229,
+      "grad_norm": 0.18029870092868805,
+      "learning_rate": 4.145634651170349e-08,
+      "loss": 0.8622,
+      "step": 6516
+    },
+    {
+      "epoch": 2.988418759316592,
+      "grad_norm": 0.267039954662323,
+      "learning_rate": 3.83287626963269e-08,
+      "loss": 1.145,
+      "step": 6517
+    },
+    {
+      "epoch": 2.988877422313955,
+      "grad_norm": 0.28512319922447205,
+      "learning_rate": 3.5323823083854845e-08,
+      "loss": 1.2058,
+      "step": 6518
+    },
+    {
+      "epoch": 2.989336085311318,
+      "grad_norm": 0.27092644572257996,
+      "learning_rate": 3.24415284114199e-08,
+      "loss": 1.3474,
+      "step": 6519
+    },
+    {
+      "epoch": 2.98979474830868,
+      "grad_norm": 0.34065067768096924,
+      "learning_rate": 2.9681879385956566e-08,
+      "loss": 2.1089,
+      "step": 6520
+    },
+    {
+      "epoch": 2.9902534113060426,
+      "grad_norm": 0.4673296809196472,
+      "learning_rate": 2.704487668453437e-08,
+      "loss": 2.0068,
+      "step": 6521
+    },
+    {
+      "epoch": 2.9907120743034055,
+      "grad_norm": 0.2979157865047455,
+      "learning_rate": 2.4530520954024748e-08,
+      "loss": 0.7067,
+      "step": 6522
+    },
+    {
+      "epoch": 2.9911707373007683,
+      "grad_norm": 0.2104611098766327,
+      "learning_rate": 2.2138812811156596e-08,
+      "loss": 1.7247,
+      "step": 6523
+    },
+    {
+      "epoch": 2.991629400298131,
+      "grad_norm": 0.4868721067905426,
+      "learning_rate": 1.9869752842682777e-08,
+      "loss": 1.6491,
+      "step": 6524
+    },
+    {
+      "epoch": 2.9920880632954936,
+      "grad_norm": 0.33822518587112427,
+      "learning_rate": 1.7723341605158095e-08,
+      "loss": 2.0029,
+      "step": 6525
+    },
+    {
+      "epoch": 2.9925467262928565,
+      "grad_norm": 0.3121846616268158,
+      "learning_rate": 1.5699579625216842e-08,
+      "loss": 1.4298,
+      "step": 6526
+    },
+    {
+      "epoch": 2.993005389290219,
+      "grad_norm": 0.3756405711174011,
+      "learning_rate": 1.3798467399184222e-08,
+      "loss": 1.7493,
+      "step": 6527
+    },
+    {
+      "epoch": 2.9934640522875817,
+      "grad_norm": 0.37064328789711,
+      "learning_rate": 1.2020005393520439e-08,
+      "loss": 1.5358,
+      "step": 6528
+    },
+    {
+      "epoch": 2.9939227152849446,
+      "grad_norm": 0.2886911928653717,
+      "learning_rate": 1.0364194044432117e-08,
+      "loss": 1.3199,
+      "step": 6529
+    },
+    {
+      "epoch": 2.994381378282307,
+      "grad_norm": 0.3759532868862152,
+      "learning_rate": 8.831033758149865e-09,
+      "loss": 1.3483,
+      "step": 6530
+    },
+    {
+      "epoch": 2.99484004127967,
+      "grad_norm": 0.21223415434360504,
+      "learning_rate": 7.420524910706217e-09,
+      "loss": 0.9441,
+      "step": 6531
+    },
+    {
+      "epoch": 2.9952987042770323,
+      "grad_norm": 0.2493351846933365,
+      "learning_rate": 6.13266784810218e-09,
+      "loss": 0.9465,
+      "step": 6532
+    },
+    {
+      "epoch": 2.995757367274395,
+      "grad_norm": 0.27161523699760437,
+      "learning_rate": 4.967462886362739e-09,
+      "loss": 1.5159,
+      "step": 6533
+    },
+    {
+      "epoch": 2.996216030271758,
+      "grad_norm": 0.40845414996147156,
+      "learning_rate": 3.924910311203789e-09,
+      "loss": 1.468,
+      "step": 6534
+    },
+    {
+      "epoch": 2.9966746932691204,
+      "grad_norm": 0.281252920627594,
+      "learning_rate": 3.005010378476225e-09,
+      "loss": 0.8934,
+      "step": 6535
+    },
+    {
+      "epoch": 2.9971333562664833,
+      "grad_norm": 0.27226901054382324,
+      "learning_rate": 2.207763313777367e-09,
+      "loss": 0.8065,
+      "step": 6536
+    },
+    {
+      "epoch": 2.9975920192638457,
+      "grad_norm": 0.0754655972123146,
+      "learning_rate": 1.5331693126174884e-09,
+      "loss": 0.4985,
+      "step": 6537
+    },
+    {
+      "epoch": 2.9980506822612085,
+      "grad_norm": 0.20777086913585663,
+      "learning_rate": 9.812285406418653e-10,
+      "loss": 1.425,
+      "step": 6538
+    },
+    {
+      "epoch": 2.9985093452585714,
+      "grad_norm": 0.3428349196910858,
+      "learning_rate": 5.519411330756619e-10,
+      "loss": 1.5652,
+      "step": 6539
+    },
+    {
+      "epoch": 2.998968008255934,
+      "grad_norm": 0.3928774297237396,
+      "learning_rate": 2.4530719533455425e-10,
+      "loss": 1.4103,
+      "step": 6540
+    },
+    {
+      "epoch": 2.998968008255934,
+      "step": 6540,
+      "total_flos": 4.3151621005372293e+18,
+      "train_loss": 1.4276331583658854,
+      "train_runtime": 125872.0926,
+      "train_samples_per_second": 13.302,
+      "train_steps_per_second": 0.052
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 6540,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.3151621005372293e+18,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}