diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,7034 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.48828125,
+  "eval_steps": 500,
+  "global_step": 1000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00048828125,
+      "grad_norm": 11.68676471710205,
+      "learning_rate": 1.6129032258064518e-07,
+      "loss": 0.6894,
+      "step": 1
+    },
+    {
+      "epoch": 0.0009765625,
+      "grad_norm": 16.37053871154785,
+      "learning_rate": 3.2258064516129035e-07,
+      "loss": 0.7171,
+      "step": 2
+    },
+    {
+      "epoch": 0.00146484375,
+      "grad_norm": 23.564491271972656,
+      "learning_rate": 4.838709677419355e-07,
+      "loss": 0.7123,
+      "step": 3
+    },
+    {
+      "epoch": 0.001953125,
+      "grad_norm": 16.051462173461914,
+      "learning_rate": 6.451612903225807e-07,
+      "loss": 0.7445,
+      "step": 4
+    },
+    {
+      "epoch": 0.00244140625,
+      "grad_norm": 13.484965324401855,
+      "learning_rate": 8.064516129032258e-07,
+      "loss": 0.7697,
+      "step": 5
+    },
+    {
+      "epoch": 0.0029296875,
+      "grad_norm": 12.733880043029785,
+      "learning_rate": 9.67741935483871e-07,
+      "loss": 0.6796,
+      "step": 6
+    },
+    {
+      "epoch": 0.00341796875,
+      "grad_norm": 11.081924438476562,
+      "learning_rate": 1.1290322580645162e-06,
+      "loss": 0.6711,
+      "step": 7
+    },
+    {
+      "epoch": 0.00390625,
+      "grad_norm": 11.96164321899414,
+      "learning_rate": 1.2903225806451614e-06,
+      "loss": 0.6916,
+      "step": 8
+    },
+    {
+      "epoch": 0.00439453125,
+      "grad_norm": 8.69968318939209,
+      "learning_rate": 1.4516129032258066e-06,
+      "loss": 0.6125,
+      "step": 9
+    },
+    {
+      "epoch": 0.0048828125,
+      "grad_norm": 8.749759674072266,
+      "learning_rate": 1.6129032258064516e-06,
+      "loss": 0.5684,
+      "step": 10
+    },
+    {
+      "epoch": 0.00537109375,
+      "grad_norm": 9.206546783447266,
+      "learning_rate": 1.774193548387097e-06,
+      "loss": 0.5901,
+      "step": 11
+    },
+    {
+      "epoch": 0.005859375,
+      "grad_norm": 6.172158718109131,
+      "learning_rate": 1.935483870967742e-06,
+      "loss": 0.5147,
+      "step": 12
+    },
+    {
+      "epoch": 0.00634765625,
+      "grad_norm": 5.583189010620117,
+      "learning_rate": 2.096774193548387e-06,
+      "loss": 0.5078,
+      "step": 13
+    },
+    {
+      "epoch": 0.0068359375,
+      "grad_norm": 8.174113273620605,
+      "learning_rate": 2.2580645161290324e-06,
+      "loss": 0.5151,
+      "step": 14
+    },
+    {
+      "epoch": 0.00732421875,
+      "grad_norm": 11.44507122039795,
+      "learning_rate": 2.4193548387096776e-06,
+      "loss": 0.5215,
+      "step": 15
+    },
+    {
+      "epoch": 0.0078125,
+      "grad_norm": 4.763265132904053,
+      "learning_rate": 2.580645161290323e-06,
+      "loss": 0.5062,
+      "step": 16
+    },
+    {
+      "epoch": 0.00830078125,
+      "grad_norm": 7.144759178161621,
+      "learning_rate": 2.7419354838709676e-06,
+      "loss": 0.5313,
+      "step": 17
+    },
+    {
+      "epoch": 0.0087890625,
+      "grad_norm": 4.595753192901611,
+      "learning_rate": 2.903225806451613e-06,
+      "loss": 0.4514,
+      "step": 18
+    },
+    {
+      "epoch": 0.00927734375,
+      "grad_norm": 5.988632678985596,
+      "learning_rate": 3.0645161290322584e-06,
+      "loss": 0.468,
+      "step": 19
+    },
+    {
+      "epoch": 0.009765625,
+      "grad_norm": 5.993471145629883,
+      "learning_rate": 3.225806451612903e-06,
+      "loss": 0.4231,
+      "step": 20
+    },
+    {
+      "epoch": 0.01025390625,
+      "grad_norm": 5.629610538482666,
+      "learning_rate": 3.3870967741935484e-06,
+      "loss": 0.4748,
+      "step": 21
+    },
+    {
+      "epoch": 0.0107421875,
+      "grad_norm": 5.070748329162598,
+      "learning_rate": 3.548387096774194e-06,
+      "loss": 0.4851,
+      "step": 22
+    },
+    {
+      "epoch": 0.01123046875,
+      "grad_norm": 5.008419990539551,
+      "learning_rate": 3.7096774193548392e-06,
+      "loss": 0.4251,
+      "step": 23
+    },
+    {
+      "epoch": 0.01171875,
+      "grad_norm": 5.048961162567139,
+      "learning_rate": 3.870967741935484e-06,
+      "loss": 0.4423,
+      "step": 24
+    },
+    {
+      "epoch": 0.01220703125,
+      "grad_norm": 3.505443811416626,
+      "learning_rate": 4.032258064516129e-06,
+      "loss": 0.4165,
+      "step": 25
+    },
+    {
+      "epoch": 0.0126953125,
+      "grad_norm": 4.471498966217041,
+      "learning_rate": 4.193548387096774e-06,
+      "loss": 0.4132,
+      "step": 26
+    },
+    {
+      "epoch": 0.01318359375,
+      "grad_norm": 3.593733310699463,
+      "learning_rate": 4.35483870967742e-06,
+      "loss": 0.38,
+      "step": 27
+    },
+    {
+      "epoch": 0.013671875,
+      "grad_norm": 7.17294979095459,
+      "learning_rate": 4.516129032258065e-06,
+      "loss": 0.3956,
+      "step": 28
+    },
+    {
+      "epoch": 0.01416015625,
+      "grad_norm": 15.088685989379883,
+      "learning_rate": 4.67741935483871e-06,
+      "loss": 0.4425,
+      "step": 29
+    },
+    {
+      "epoch": 0.0146484375,
+      "grad_norm": 4.4346113204956055,
+      "learning_rate": 4.838709677419355e-06,
+      "loss": 0.3911,
+      "step": 30
+    },
+    {
+      "epoch": 0.01513671875,
+      "grad_norm": 4.740771293640137,
+      "learning_rate": 5e-06,
+      "loss": 0.423,
+      "step": 31
+    },
+    {
+      "epoch": 0.015625,
+      "grad_norm": 3.4211642742156982,
+      "learning_rate": 5.161290322580646e-06,
+      "loss": 0.4183,
+      "step": 32
+    },
+    {
+      "epoch": 0.01611328125,
+      "grad_norm": 5.500433444976807,
+      "learning_rate": 5.322580645161291e-06,
+      "loss": 0.3956,
+      "step": 33
+    },
+    {
+      "epoch": 0.0166015625,
+      "grad_norm": 4.092607021331787,
+      "learning_rate": 5.483870967741935e-06,
+      "loss": 0.4028,
+      "step": 34
+    },
+    {
+      "epoch": 0.01708984375,
+      "grad_norm": 12.963457107543945,
+      "learning_rate": 5.645161290322582e-06,
+      "loss": 0.3862,
+      "step": 35
+    },
+    {
+      "epoch": 0.017578125,
+      "grad_norm": 4.550689697265625,
+      "learning_rate": 5.806451612903226e-06,
+      "loss": 0.4078,
+      "step": 36
+    },
+    {
+      "epoch": 0.01806640625,
+      "grad_norm": 3.3017280101776123,
+      "learning_rate": 5.967741935483872e-06,
+      "loss": 0.4334,
+      "step": 37
+    },
+    {
+      "epoch": 0.0185546875,
+      "grad_norm": 4.2097954750061035,
+      "learning_rate": 6.129032258064517e-06,
+      "loss": 0.342,
+      "step": 38
+    },
+    {
+      "epoch": 0.01904296875,
+      "grad_norm": 2.9576752185821533,
+      "learning_rate": 6.290322580645162e-06,
+      "loss": 0.3824,
+      "step": 39
+    },
+    {
+      "epoch": 0.01953125,
+      "grad_norm": 6.747947692871094,
+      "learning_rate": 6.451612903225806e-06,
+      "loss": 0.3952,
+      "step": 40
+    },
+    {
+      "epoch": 0.02001953125,
+      "grad_norm": 2.851712942123413,
+      "learning_rate": 6.612903225806452e-06,
+      "loss": 0.4143,
+      "step": 41
+    },
+    {
+      "epoch": 0.0205078125,
+      "grad_norm": 3.3788578510284424,
+      "learning_rate": 6.774193548387097e-06,
+      "loss": 0.3733,
+      "step": 42
+    },
+    {
+      "epoch": 0.02099609375,
+      "grad_norm": 4.708284378051758,
+      "learning_rate": 6.935483870967743e-06,
+      "loss": 0.3955,
+      "step": 43
+    },
+    {
+      "epoch": 0.021484375,
+      "grad_norm": 3.0566701889038086,
+      "learning_rate": 7.096774193548388e-06,
+      "loss": 0.402,
+      "step": 44
+    },
+    {
+      "epoch": 0.02197265625,
+      "grad_norm": 4.440851211547852,
+      "learning_rate": 7.258064516129033e-06,
+      "loss": 0.361,
+      "step": 45
+    },
+    {
+      "epoch": 0.0224609375,
+      "grad_norm": 2.7747905254364014,
+      "learning_rate": 7.4193548387096784e-06,
+      "loss": 0.3896,
+      "step": 46
+    },
+    {
+      "epoch": 0.02294921875,
+      "grad_norm": 3.510695695877075,
+      "learning_rate": 7.580645161290323e-06,
+      "loss": 0.364,
+      "step": 47
+    },
+    {
+      "epoch": 0.0234375,
+      "grad_norm": 20.806020736694336,
+      "learning_rate": 7.741935483870968e-06,
+      "loss": 0.3849,
+      "step": 48
+    },
+    {
+      "epoch": 0.02392578125,
+      "grad_norm": 3.569124698638916,
+      "learning_rate": 7.903225806451613e-06,
+      "loss": 0.3569,
+      "step": 49
+    },
+    {
+      "epoch": 0.0244140625,
+      "grad_norm": 2.8412413597106934,
+      "learning_rate": 8.064516129032258e-06,
+      "loss": 0.362,
+      "step": 50
+    },
+    {
+      "epoch": 0.02490234375,
+      "grad_norm": 3.287231683731079,
+      "learning_rate": 8.225806451612904e-06,
+      "loss": 0.3941,
+      "step": 51
+    },
+    {
+      "epoch": 0.025390625,
+      "grad_norm": 2.849888563156128,
+      "learning_rate": 8.387096774193549e-06,
+      "loss": 0.3906,
+      "step": 52
+    },
+    {
+      "epoch": 0.02587890625,
+      "grad_norm": 6.925948619842529,
+      "learning_rate": 8.548387096774194e-06,
+      "loss": 0.3783,
+      "step": 53
+    },
+    {
+      "epoch": 0.0263671875,
+      "grad_norm": 2.9347381591796875,
+      "learning_rate": 8.70967741935484e-06,
+      "loss": 0.4156,
+      "step": 54
+    },
+    {
+      "epoch": 0.02685546875,
+      "grad_norm": 3.695150375366211,
+      "learning_rate": 8.870967741935484e-06,
+      "loss": 0.3586,
+      "step": 55
+    },
+    {
+      "epoch": 0.02734375,
+      "grad_norm": 7.241847038269043,
+      "learning_rate": 9.03225806451613e-06,
+      "loss": 0.3693,
+      "step": 56
+    },
+    {
+      "epoch": 0.02783203125,
+      "grad_norm": 2.603956699371338,
+      "learning_rate": 9.193548387096775e-06,
+      "loss": 0.4109,
+      "step": 57
+    },
+    {
+      "epoch": 0.0283203125,
+      "grad_norm": 3.0118958950042725,
+      "learning_rate": 9.35483870967742e-06,
+      "loss": 0.4096,
+      "step": 58
+    },
+    {
+      "epoch": 0.02880859375,
+      "grad_norm": 5.108702182769775,
+      "learning_rate": 9.516129032258065e-06,
+      "loss": 0.3786,
+      "step": 59
+    },
+    {
+      "epoch": 0.029296875,
+      "grad_norm": 3.0591766834259033,
+      "learning_rate": 9.67741935483871e-06,
+      "loss": 0.3979,
+      "step": 60
+    },
+    {
+      "epoch": 0.02978515625,
+      "grad_norm": 3.5517218112945557,
+      "learning_rate": 9.838709677419356e-06,
+      "loss": 0.3847,
+      "step": 61
+    },
+    {
+      "epoch": 0.0302734375,
+      "grad_norm": 3.091423988342285,
+      "learning_rate": 1e-05,
+      "loss": 0.35,
+      "step": 62
+    },
+    {
+      "epoch": 0.03076171875,
+      "grad_norm": 2.7133779525756836,
+      "learning_rate": 9.999993744224208e-06,
+      "loss": 0.3592,
+      "step": 63
+    },
+    {
+      "epoch": 0.03125,
+      "grad_norm": 2.4324684143066406,
+      "learning_rate": 9.999974976912485e-06,
+      "loss": 0.3616,
+      "step": 64
+    },
+    {
+      "epoch": 0.03173828125,
+      "grad_norm": 2.872821807861328,
+      "learning_rate": 9.999943698111792e-06,
+      "loss": 0.3741,
+      "step": 65
+    },
+    {
+      "epoch": 0.0322265625,
+      "grad_norm": 2.9383156299591064,
+      "learning_rate": 9.999899907900399e-06,
+      "loss": 0.3732,
+      "step": 66
+    },
+    {
+      "epoch": 0.03271484375,
+      "grad_norm": 3.5359489917755127,
+      "learning_rate": 9.999843606387883e-06,
+      "loss": 0.4053,
+      "step": 67
+    },
+    {
+      "epoch": 0.033203125,
+      "grad_norm": 3.5608558654785156,
+      "learning_rate": 9.999774793715126e-06,
+      "loss": 0.4197,
+      "step": 68
+    },
+    {
+      "epoch": 0.03369140625,
+      "grad_norm": 2.5407004356384277,
+      "learning_rate": 9.999693470054321e-06,
+      "loss": 0.354,
+      "step": 69
+    },
+    {
+      "epoch": 0.0341796875,
+      "grad_norm": 3.4264254570007324,
+      "learning_rate": 9.999599635608964e-06,
+      "loss": 0.3936,
+      "step": 70
+    },
+    {
+      "epoch": 0.03466796875,
+      "grad_norm": 3.0363235473632812,
+      "learning_rate": 9.999493290613859e-06,
+      "loss": 0.3753,
+      "step": 71
+    },
+    {
+      "epoch": 0.03515625,
+      "grad_norm": 2.2824833393096924,
+      "learning_rate": 9.999374435335113e-06,
+      "loss": 0.3813,
+      "step": 72
+    },
+    {
+      "epoch": 0.03564453125,
+      "grad_norm": 2.445328712463379,
+      "learning_rate": 9.999243070070137e-06,
+      "loss": 0.4237,
+      "step": 73
+    },
+    {
+      "epoch": 0.0361328125,
+      "grad_norm": 5.150700092315674,
+      "learning_rate": 9.99909919514765e-06,
+      "loss": 0.3892,
+      "step": 74
+    },
+    {
+      "epoch": 0.03662109375,
+      "grad_norm": 4.1412272453308105,
+      "learning_rate": 9.998942810927673e-06,
+      "loss": 0.3675,
+      "step": 75
+    },
+    {
+      "epoch": 0.037109375,
+      "grad_norm": 5.456881999969482,
+      "learning_rate": 9.998773917801526e-06,
+      "loss": 0.3955,
+      "step": 76
+    },
+    {
+      "epoch": 0.03759765625,
+      "grad_norm": 2.2837321758270264,
+      "learning_rate": 9.998592516191832e-06,
+      "loss": 0.3477,
+      "step": 77
+    },
+    {
+      "epoch": 0.0380859375,
+      "grad_norm": 2.237900972366333,
+      "learning_rate": 9.998398606552513e-06,
+      "loss": 0.3771,
+      "step": 78
+    },
+    {
+      "epoch": 0.03857421875,
+      "grad_norm": 2.6276211738586426,
+      "learning_rate": 9.998192189368795e-06,
+      "loss": 0.3989,
+      "step": 79
+    },
+    {
+      "epoch": 0.0390625,
+      "grad_norm": 3.5280210971832275,
+      "learning_rate": 9.997973265157192e-06,
+      "loss": 0.3726,
+      "step": 80
+    },
+    {
+      "epoch": 0.03955078125,
+      "grad_norm": 8.555140495300293,
+      "learning_rate": 9.997741834465526e-06,
+      "loss": 0.397,
+      "step": 81
+    },
+    {
+      "epoch": 0.0400390625,
+      "grad_norm": 2.1885085105895996,
+      "learning_rate": 9.997497897872904e-06,
+      "loss": 0.4058,
+      "step": 82
+    },
+    {
+      "epoch": 0.04052734375,
+      "grad_norm": 3.0636098384857178,
+      "learning_rate": 9.997241455989735e-06,
+      "loss": 0.3866,
+      "step": 83
+    },
+    {
+      "epoch": 0.041015625,
+      "grad_norm": 3.7982375621795654,
+      "learning_rate": 9.996972509457711e-06,
+      "loss": 0.3877,
+      "step": 84
+    },
+    {
+      "epoch": 0.04150390625,
+      "grad_norm": 2.4791505336761475,
+      "learning_rate": 9.996691058949826e-06,
+      "loss": 0.3789,
+      "step": 85
+    },
+    {
+      "epoch": 0.0419921875,
+      "grad_norm": 3.917693614959717,
+      "learning_rate": 9.996397105170353e-06,
+      "loss": 0.3737,
+      "step": 86
+    },
+    {
+      "epoch": 0.04248046875,
+      "grad_norm": 2.3083252906799316,
+      "learning_rate": 9.996090648854856e-06,
+      "loss": 0.3658,
+      "step": 87
+    },
+    {
+      "epoch": 0.04296875,
+      "grad_norm": 1.9872547388076782,
+      "learning_rate": 9.995771690770184e-06,
+      "loss": 0.3819,
+      "step": 88
+    },
+    {
+      "epoch": 0.04345703125,
+      "grad_norm": 1.8703923225402832,
+      "learning_rate": 9.995440231714469e-06,
+      "loss": 0.37,
+      "step": 89
+    },
+    {
+      "epoch": 0.0439453125,
+      "grad_norm": 2.7573578357696533,
+      "learning_rate": 9.995096272517122e-06,
+      "loss": 0.3876,
+      "step": 90
+    },
+    {
+      "epoch": 0.04443359375,
+      "grad_norm": 2.177542209625244,
+      "learning_rate": 9.99473981403884e-06,
+      "loss": 0.434,
+      "step": 91
+    },
+    {
+      "epoch": 0.044921875,
+      "grad_norm": 1.9339114427566528,
+      "learning_rate": 9.99437085717159e-06,
+      "loss": 0.333,
+      "step": 92
+    },
+    {
+      "epoch": 0.04541015625,
+      "grad_norm": 2.9820590019226074,
+      "learning_rate": 9.993989402838618e-06,
+      "loss": 0.3321,
+      "step": 93
+    },
+    {
+      "epoch": 0.0458984375,
+      "grad_norm": 2.0244717597961426,
+      "learning_rate": 9.99359545199444e-06,
+      "loss": 0.3153,
+      "step": 94
+    },
+    {
+      "epoch": 0.04638671875,
+      "grad_norm": 2.0268101692199707,
+      "learning_rate": 9.993189005624842e-06,
+      "loss": 0.3663,
+      "step": 95
+    },
+    {
+      "epoch": 0.046875,
+      "grad_norm": 1.920785903930664,
+      "learning_rate": 9.992770064746882e-06,
+      "loss": 0.3419,
+      "step": 96
+    },
+    {
+      "epoch": 0.04736328125,
+      "grad_norm": 3.2875781059265137,
+      "learning_rate": 9.992338630408877e-06,
+      "loss": 0.3406,
+      "step": 97
+    },
+    {
+      "epoch": 0.0478515625,
+      "grad_norm": 3.7749016284942627,
+      "learning_rate": 9.991894703690414e-06,
+      "loss": 0.3555,
+      "step": 98
+    },
+    {
+      "epoch": 0.04833984375,
+      "grad_norm": 4.618077754974365,
+      "learning_rate": 9.991438285702332e-06,
+      "loss": 0.4001,
+      "step": 99
+    },
+    {
+      "epoch": 0.048828125,
+      "grad_norm": 2.468576192855835,
+      "learning_rate": 9.99096937758673e-06,
+      "loss": 0.4258,
+      "step": 100
+    },
+    {
+      "epoch": 0.04931640625,
+      "grad_norm": 5.204842567443848,
+      "learning_rate": 9.990487980516962e-06,
+      "loss": 0.4107,
+      "step": 101
+    },
+    {
+      "epoch": 0.0498046875,
+      "grad_norm": 3.3488011360168457,
+      "learning_rate": 9.989994095697636e-06,
+      "loss": 0.3658,
+      "step": 102
+    },
+    {
+      "epoch": 0.05029296875,
+      "grad_norm": 4.41386079788208,
+      "learning_rate": 9.989487724364602e-06,
+      "loss": 0.3705,
+      "step": 103
+    },
+    {
+      "epoch": 0.05078125,
+      "grad_norm": 2.9542033672332764,
+      "learning_rate": 9.988968867784958e-06,
+      "loss": 0.3955,
+      "step": 104
+    },
+    {
+      "epoch": 0.05126953125,
+      "grad_norm": 2.3820998668670654,
+      "learning_rate": 9.988437527257044e-06,
+      "loss": 0.3652,
+      "step": 105
+    },
+    {
+      "epoch": 0.0517578125,
+      "grad_norm": 2.352477550506592,
+      "learning_rate": 9.987893704110441e-06,
+      "loss": 0.3545,
+      "step": 106
+    },
+    {
+      "epoch": 0.05224609375,
+      "grad_norm": 7.221553802490234,
+      "learning_rate": 9.987337399705964e-06,
+      "loss": 0.3616,
+      "step": 107
+    },
+    {
+      "epoch": 0.052734375,
+      "grad_norm": 2.3267176151275635,
+      "learning_rate": 9.986768615435655e-06,
+      "loss": 0.3868,
+      "step": 108
+    },
+    {
+      "epoch": 0.05322265625,
+      "grad_norm": 1.9337338209152222,
+      "learning_rate": 9.986187352722792e-06,
+      "loss": 0.3664,
+      "step": 109
+    },
+    {
+      "epoch": 0.0537109375,
+      "grad_norm": 2.2121877670288086,
+      "learning_rate": 9.985593613021873e-06,
+      "loss": 0.3731,
+      "step": 110
+    },
+    {
+      "epoch": 0.05419921875,
+      "grad_norm": 1.9584633111953735,
+      "learning_rate": 9.98498739781862e-06,
+      "loss": 0.3805,
+      "step": 111
+    },
+    {
+      "epoch": 0.0546875,
+      "grad_norm": 2.3601884841918945,
+      "learning_rate": 9.984368708629972e-06,
+      "loss": 0.3328,
+      "step": 112
+    },
+    {
+      "epoch": 0.05517578125,
+      "grad_norm": 2.705298662185669,
+      "learning_rate": 9.98373754700408e-06,
+      "loss": 0.3573,
+      "step": 113
+    },
+    {
+      "epoch": 0.0556640625,
+      "grad_norm": 4.535929203033447,
+      "learning_rate": 9.98309391452031e-06,
+      "loss": 0.3853,
+      "step": 114
+    },
+    {
+      "epoch": 0.05615234375,
+      "grad_norm": 2.4388949871063232,
+      "learning_rate": 9.982437812789224e-06,
+      "loss": 0.3389,
+      "step": 115
+    },
+    {
+      "epoch": 0.056640625,
+      "grad_norm": 3.7873549461364746,
+      "learning_rate": 9.981769243452595e-06,
+      "loss": 0.3745,
+      "step": 116
+    },
+    {
+      "epoch": 0.05712890625,
+      "grad_norm": 2.1249921321868896,
+      "learning_rate": 9.981088208183392e-06,
+      "loss": 0.3854,
+      "step": 117
+    },
+    {
+      "epoch": 0.0576171875,
+      "grad_norm": 3.2426087856292725,
+      "learning_rate": 9.980394708685777e-06,
+      "loss": 0.3743,
+      "step": 118
+    },
+    {
+      "epoch": 0.05810546875,
+      "grad_norm": 2.349886178970337,
+      "learning_rate": 9.979688746695099e-06,
+      "loss": 0.3477,
+      "step": 119
+    },
+    {
+      "epoch": 0.05859375,
+      "grad_norm": 2.6616315841674805,
+      "learning_rate": 9.978970323977895e-06,
+      "loss": 0.3497,
+      "step": 120
+    },
+    {
+      "epoch": 0.05908203125,
+      "grad_norm": 2.284364938735962,
+      "learning_rate": 9.978239442331881e-06,
+      "loss": 0.3987,
+      "step": 121
+    },
+    {
+      "epoch": 0.0595703125,
+      "grad_norm": 2.347794532775879,
+      "learning_rate": 9.977496103585949e-06,
+      "loss": 0.3375,
+      "step": 122
+    },
+    {
+      "epoch": 0.06005859375,
+      "grad_norm": 2.8935320377349854,
+      "learning_rate": 9.976740309600166e-06,
+      "loss": 0.3943,
+      "step": 123
+    },
+    {
+      "epoch": 0.060546875,
+      "grad_norm": 2.3763160705566406,
+      "learning_rate": 9.97597206226576e-06,
+      "loss": 0.3703,
+      "step": 124
+    },
+    {
+      "epoch": 0.06103515625,
+      "grad_norm": 2.1485118865966797,
+      "learning_rate": 9.975191363505127e-06,
+      "loss": 0.3604,
+      "step": 125
+    },
+    {
+      "epoch": 0.0615234375,
+      "grad_norm": 4.019608020782471,
+      "learning_rate": 9.974398215271814e-06,
+      "loss": 0.3345,
+      "step": 126
+    },
+    {
+      "epoch": 0.06201171875,
+      "grad_norm": 4.793520450592041,
+      "learning_rate": 9.973592619550528e-06,
+      "loss": 0.3583,
+      "step": 127
+    },
+    {
+      "epoch": 0.0625,
+      "grad_norm": 2.3743088245391846,
+      "learning_rate": 9.972774578357118e-06,
+      "loss": 0.3612,
+      "step": 128
+    },
+    {
+      "epoch": 0.06298828125,
+      "grad_norm": 2.3221397399902344,
+      "learning_rate": 9.971944093738575e-06,
+      "loss": 0.3759,
+      "step": 129
+    },
+    {
+      "epoch": 0.0634765625,
+      "grad_norm": 2.639760971069336,
+      "learning_rate": 9.971101167773032e-06,
+      "loss": 0.3749,
+      "step": 130
+    },
+    {
+      "epoch": 0.06396484375,
+      "grad_norm": 2.3176326751708984,
+      "learning_rate": 9.97024580256975e-06,
+      "loss": 0.3324,
+      "step": 131
+    },
+    {
+      "epoch": 0.064453125,
+      "grad_norm": 2.5662341117858887,
+      "learning_rate": 9.969378000269117e-06,
+      "loss": 0.3956,
+      "step": 132
+    },
+    {
+      "epoch": 0.06494140625,
+      "grad_norm": 3.271336793899536,
+      "learning_rate": 9.968497763042644e-06,
+      "loss": 0.3702,
+      "step": 133
+    },
+    {
+      "epoch": 0.0654296875,
+      "grad_norm": 2.0121848583221436,
+      "learning_rate": 9.96760509309296e-06,
+      "loss": 0.3644,
+      "step": 134
+    },
+    {
+      "epoch": 0.06591796875,
+      "grad_norm": 2.1467254161834717,
+      "learning_rate": 9.9666999926538e-06,
+      "loss": 0.3444,
+      "step": 135
+    },
+    {
+      "epoch": 0.06640625,
+      "grad_norm": 2.985793113708496,
+      "learning_rate": 9.96578246399001e-06,
+      "loss": 0.4015,
+      "step": 136
+    },
+    {
+      "epoch": 0.06689453125,
+      "grad_norm": 2.158658504486084,
+      "learning_rate": 9.964852509397527e-06,
+      "loss": 0.3809,
+      "step": 137
+    },
+    {
+      "epoch": 0.0673828125,
+      "grad_norm": 4.1197919845581055,
+      "learning_rate": 9.963910131203386e-06,
+      "loss": 0.3874,
+      "step": 138
+    },
+    {
+      "epoch": 0.06787109375,
+      "grad_norm": 2.2979846000671387,
+      "learning_rate": 9.962955331765712e-06,
+      "loss": 0.342,
+      "step": 139
+    },
+    {
+      "epoch": 0.068359375,
+      "grad_norm": 2.2568418979644775,
+      "learning_rate": 9.961988113473708e-06,
+      "loss": 0.3223,
+      "step": 140
+    },
+    {
+      "epoch": 0.06884765625,
+      "grad_norm": 2.358520030975342,
+      "learning_rate": 9.961008478747655e-06,
+      "loss": 0.374,
+      "step": 141
+    },
+    {
+      "epoch": 0.0693359375,
+      "grad_norm": 2.6409096717834473,
+      "learning_rate": 9.960016430038903e-06,
+      "loss": 0.3705,
+      "step": 142
+    },
+    {
+      "epoch": 0.06982421875,
+      "grad_norm": 2.167280673980713,
+      "learning_rate": 9.959011969829867e-06,
+      "loss": 0.3302,
+      "step": 143
+    },
+    {
+      "epoch": 0.0703125,
+      "grad_norm": 2.3867969512939453,
+      "learning_rate": 9.957995100634016e-06,
+      "loss": 0.3251,
+      "step": 144
+    },
+    {
+      "epoch": 0.07080078125,
+      "grad_norm": 2.305117130279541,
+      "learning_rate": 9.956965824995873e-06,
+      "loss": 0.3593,
+      "step": 145
+    },
+    {
+      "epoch": 0.0712890625,
+      "grad_norm": 2.1817824840545654,
+      "learning_rate": 9.955924145491005e-06,
+      "loss": 0.3371,
+      "step": 146
+    },
+    {
+      "epoch": 0.07177734375,
+      "grad_norm": 4.12109375,
+      "learning_rate": 9.954870064726017e-06,
+      "loss": 0.3771,
+      "step": 147
+    },
+    {
+      "epoch": 0.072265625,
+      "grad_norm": 3.0079329013824463,
+      "learning_rate": 9.953803585338548e-06,
+      "loss": 0.3636,
+      "step": 148
+    },
+    {
+      "epoch": 0.07275390625,
+      "grad_norm": 2.473532199859619,
+      "learning_rate": 9.95272470999726e-06,
+      "loss": 0.3692,
+      "step": 149
+    },
+    {
+      "epoch": 0.0732421875,
+      "grad_norm": 3.1922385692596436,
+      "learning_rate": 9.95163344140183e-06,
+      "loss": 0.3773,
+      "step": 150
+    },
+    {
+      "epoch": 0.07373046875,
+      "grad_norm": 6.991460800170898,
+      "learning_rate": 9.950529782282955e-06,
+      "loss": 0.2813,
+      "step": 151
+    },
+    {
+      "epoch": 0.07421875,
+      "grad_norm": 2.9967305660247803,
+      "learning_rate": 9.949413735402332e-06,
+      "loss": 0.3565,
+      "step": 152
+    },
+    {
+      "epoch": 0.07470703125,
+      "grad_norm": 1.8642289638519287,
+      "learning_rate": 9.948285303552654e-06,
+      "loss": 0.3715,
+      "step": 153
+    },
+    {
+      "epoch": 0.0751953125,
+      "grad_norm": 2.169416904449463,
+      "learning_rate": 9.947144489557612e-06,
+      "loss": 0.3507,
+      "step": 154
+    },
+    {
+      "epoch": 0.07568359375,
+      "grad_norm": 2.5897326469421387,
+      "learning_rate": 9.945991296271874e-06,
+      "loss": 0.3508,
+      "step": 155
+    },
+    {
+      "epoch": 0.076171875,
+      "grad_norm": 1.8967130184173584,
+      "learning_rate": 9.944825726581085e-06,
+      "loss": 0.318,
+      "step": 156
+    },
+    {
+      "epoch": 0.07666015625,
+      "grad_norm": 1.998544454574585,
+      "learning_rate": 9.943647783401867e-06,
+      "loss": 0.3757,
+      "step": 157
+    },
+    {
+      "epoch": 0.0771484375,
+      "grad_norm": 2.5188403129577637,
+      "learning_rate": 9.942457469681794e-06,
+      "loss": 0.3551,
+      "step": 158
+    },
+    {
+      "epoch": 0.07763671875,
+      "grad_norm": 2.2102835178375244,
+      "learning_rate": 9.941254788399406e-06,
+      "loss": 0.3499,
+      "step": 159
+    },
+    {
+      "epoch": 0.078125,
+      "grad_norm": 3.3190438747406006,
+      "learning_rate": 9.940039742564182e-06,
+      "loss": 0.3586,
+      "step": 160
+    },
+    {
+      "epoch": 0.07861328125,
+      "grad_norm": 6.675033092498779,
+      "learning_rate": 9.938812335216543e-06,
+      "loss": 0.3892,
+      "step": 161
+    },
+    {
+      "epoch": 0.0791015625,
+      "grad_norm": 3.091517925262451,
+      "learning_rate": 9.937572569427844e-06,
+      "loss": 0.3434,
+      "step": 162
+    },
+    {
+      "epoch": 0.07958984375,
+      "grad_norm": 2.7739408016204834,
+      "learning_rate": 9.936320448300364e-06,
+      "loss": 0.3366,
+      "step": 163
+    },
+    {
+      "epoch": 0.080078125,
+      "grad_norm": 4.218409538269043,
+      "learning_rate": 9.935055974967299e-06,
+      "loss": 0.3129,
+      "step": 164
+    },
+    {
+      "epoch": 0.08056640625,
+      "grad_norm": 2.2632052898406982,
+      "learning_rate": 9.933779152592752e-06,
+      "loss": 0.3507,
+      "step": 165
+    },
+    {
+      "epoch": 0.0810546875,
+      "grad_norm": 2.3607664108276367,
+      "learning_rate": 9.93248998437173e-06,
+      "loss": 0.3598,
+      "step": 166
+    },
+    {
+      "epoch": 0.08154296875,
+      "grad_norm": 2.2539124488830566,
+      "learning_rate": 9.931188473530132e-06,
+      "loss": 0.404,
+      "step": 167
+    },
+    {
+      "epoch": 0.08203125,
+      "grad_norm": 2.049994945526123,
+      "learning_rate": 9.929874623324741e-06,
+      "loss": 0.3534,
+      "step": 168
+    },
+    {
+      "epoch": 0.08251953125,
+      "grad_norm": 4.720448017120361,
+      "learning_rate": 9.92854843704322e-06,
+      "loss": 0.3492,
+      "step": 169
+    },
+    {
+      "epoch": 0.0830078125,
+      "grad_norm": 2.1875171661376953,
+      "learning_rate": 9.927209918004095e-06,
+      "loss": 0.3765,
+      "step": 170
+    },
+    {
+      "epoch": 0.08349609375,
+      "grad_norm": 6.087578773498535,
+      "learning_rate": 9.92585906955676e-06,
+      "loss": 0.3519,
+      "step": 171
+    },
+    {
+      "epoch": 0.083984375,
+      "grad_norm": 6.033719539642334,
+      "learning_rate": 9.924495895081455e-06,
+      "loss": 0.3493,
+      "step": 172
+    },
+    {
+      "epoch": 0.08447265625,
+      "grad_norm": 4.239842414855957,
+      "learning_rate": 9.923120397989265e-06,
+      "loss": 0.3566,
+      "step": 173
+    },
+    {
+      "epoch": 0.0849609375,
+      "grad_norm": 3.4344899654388428,
+      "learning_rate": 9.92173258172211e-06,
+      "loss": 0.3291,
+      "step": 174
+    },
+    {
+      "epoch": 0.08544921875,
+      "grad_norm": 2.5044116973876953,
+      "learning_rate": 9.920332449752741e-06,
+      "loss": 0.368,
+      "step": 175
+    },
+    {
+      "epoch": 0.0859375,
+      "grad_norm": 2.5513086318969727,
+      "learning_rate": 9.91892000558472e-06,
+      "loss": 0.3715,
+      "step": 176
+    },
+    {
+      "epoch": 0.08642578125,
+      "grad_norm": 3.1087024211883545,
+      "learning_rate": 9.917495252752418e-06,
+      "loss": 0.3421,
+      "step": 177
+    },
+    {
+      "epoch": 0.0869140625,
+      "grad_norm": 4.5129194259643555,
+      "learning_rate": 9.916058194821013e-06,
+      "loss": 0.3348,
+      "step": 178
+    },
+    {
+      "epoch": 0.08740234375,
+      "grad_norm": 2.54546856880188,
+      "learning_rate": 9.914608835386468e-06,
+      "loss": 0.3741,
+      "step": 179
+    },
+    {
+      "epoch": 0.087890625,
+      "grad_norm": 3.379059314727783,
+      "learning_rate": 9.913147178075531e-06,
+      "loss": 0.3633,
+      "step": 180
+    },
+    {
+      "epoch": 0.08837890625,
+      "grad_norm": 2.6582908630371094,
+      "learning_rate": 9.911673226545721e-06,
+      "loss": 0.3626,
+      "step": 181
+    },
+    {
+      "epoch": 0.0888671875,
+      "grad_norm": 2.116603374481201,
+      "learning_rate": 9.910186984485321e-06,
+      "loss": 0.3627,
+      "step": 182
+    },
+    {
+      "epoch": 0.08935546875,
+      "grad_norm": 3.2947633266448975,
+      "learning_rate": 9.908688455613374e-06,
+      "loss": 0.3264,
+      "step": 183
+    },
+    {
+      "epoch": 0.08984375,
+      "grad_norm": 2.313702344894409,
+      "learning_rate": 9.90717764367966e-06,
+      "loss": 0.3285,
+      "step": 184
+    },
+    {
+      "epoch": 0.09033203125,
+      "grad_norm": 2.2801687717437744,
+      "learning_rate": 9.9056545524647e-06,
+      "loss": 0.3573,
+      "step": 185
+    },
+    {
+      "epoch": 0.0908203125,
+      "grad_norm": 3.657966375350952,
+      "learning_rate": 9.904119185779744e-06,
+      "loss": 0.3711,
+      "step": 186
+    },
+    {
+      "epoch": 0.09130859375,
+      "grad_norm": 22.30857276916504,
+      "learning_rate": 9.902571547466753e-06,
+      "loss": 0.3995,
+      "step": 187
+    },
+    {
+      "epoch": 0.091796875,
+      "grad_norm": 2.184039831161499,
+      "learning_rate": 9.901011641398398e-06,
+      "loss": 0.3654,
+      "step": 188
+    },
+    {
+      "epoch": 0.09228515625,
+      "grad_norm": 4.786393165588379,
+      "learning_rate": 9.89943947147805e-06,
+      "loss": 0.3859,
+      "step": 189
+    },
+    {
+      "epoch": 0.0927734375,
+      "grad_norm": 2.666750431060791,
+      "learning_rate": 9.897855041639764e-06,
+      "loss": 0.3888,
+      "step": 190
+    },
+    {
+      "epoch": 0.09326171875,
+      "grad_norm": 2.0390570163726807,
+      "learning_rate": 9.896258355848277e-06,
+      "loss": 0.3488,
+      "step": 191
+    },
+    {
+      "epoch": 0.09375,
+      "grad_norm": 2.618748188018799,
+      "learning_rate": 9.894649418098992e-06,
+      "loss": 0.3513,
+      "step": 192
+    },
+    {
+      "epoch": 0.09423828125,
+      "grad_norm": 2.525346040725708,
+      "learning_rate": 9.89302823241797e-06,
+      "loss": 0.3689,
+      "step": 193
+    },
+    {
+      "epoch": 0.0947265625,
+      "grad_norm": 2.0813663005828857,
+      "learning_rate": 9.89139480286192e-06,
+      "loss": 0.3718,
+      "step": 194
+    },
+    {
+      "epoch": 0.09521484375,
+      "grad_norm": 3.025359630584717,
+      "learning_rate": 9.88974913351819e-06,
+      "loss": 0.3786,
+      "step": 195
+    },
+    {
+      "epoch": 0.095703125,
+      "grad_norm": 2.8500590324401855,
+      "learning_rate": 9.888091228504757e-06,
+      "loss": 0.3481,
+      "step": 196
+    },
+    {
+      "epoch": 0.09619140625,
+      "grad_norm": 2.450500249862671,
+      "learning_rate": 9.88642109197021e-06,
+      "loss": 0.383,
+      "step": 197
+    },
+    {
+      "epoch": 0.0966796875,
+      "grad_norm": 1.9162877798080444,
+      "learning_rate": 9.884738728093754e-06,
+      "loss": 0.3698,
+      "step": 198
+    },
+    {
+      "epoch": 0.09716796875,
+      "grad_norm": 14.184158325195312,
+      "learning_rate": 9.883044141085183e-06,
+      "loss": 0.3327,
+      "step": 199
+    },
+    {
+      "epoch": 0.09765625,
+      "grad_norm": 3.0886130332946777,
+      "learning_rate": 9.881337335184879e-06,
+      "loss": 0.3767,
+      "step": 200
+    },
+    {
+      "epoch": 0.09814453125,
+      "grad_norm": 2.5864577293395996,
+      "learning_rate": 9.879618314663799e-06,
+      "loss": 0.3498,
+      "step": 201
+    },
+    {
+      "epoch": 0.0986328125,
+      "grad_norm": 3.3661086559295654,
+      "learning_rate": 9.87788708382347e-06,
+      "loss": 0.3487,
+      "step": 202
+    },
+    {
+      "epoch": 0.09912109375,
+      "grad_norm": 2.543836832046509,
+      "learning_rate": 9.876143646995964e-06,
+      "loss": 0.3611,
+      "step": 203
+    },
+    {
+      "epoch": 0.099609375,
+      "grad_norm": 2.209348201751709,
+      "learning_rate": 9.874388008543903e-06,
+      "loss": 0.3303,
+      "step": 204
+    },
+    {
+      "epoch": 0.10009765625,
+      "grad_norm": 8.464391708374023,
+      "learning_rate": 9.87262017286044e-06,
+      "loss": 0.3915,
+      "step": 205
+    },
+    {
+      "epoch": 0.1005859375,
+      "grad_norm": 2.339383125305176,
+      "learning_rate": 9.870840144369247e-06,
+      "loss": 0.3386,
+      "step": 206
+    },
+    {
+      "epoch": 0.10107421875,
+      "grad_norm": 4.952784538269043,
+      "learning_rate": 9.869047927524508e-06,
+      "loss": 0.3189,
+      "step": 207
+    },
+    {
+      "epoch": 0.1015625,
+      "grad_norm": 2.147639036178589,
+      "learning_rate": 9.867243526810909e-06,
+      "loss": 0.325,
+      "step": 208
+    },
+    {
+      "epoch": 0.10205078125,
+      "grad_norm": 2.364194393157959,
+      "learning_rate": 9.865426946743614e-06,
+      "loss": 0.3728,
+      "step": 209
+    },
+    {
+      "epoch": 0.1025390625,
+      "grad_norm": 2.0875487327575684,
+      "learning_rate": 9.863598191868275e-06,
+      "loss": 0.3493,
+      "step": 210
+    },
+    {
+      "epoch": 0.10302734375,
+      "grad_norm": 3.100674629211426,
+      "learning_rate": 9.861757266761002e-06,
+      "loss": 0.3503,
+      "step": 211
+    },
+    {
+      "epoch": 0.103515625,
+      "grad_norm": 3.1530754566192627,
+      "learning_rate": 9.859904176028364e-06,
+      "loss": 0.3635,
+      "step": 212
+    },
+    {
+      "epoch": 0.10400390625,
+      "grad_norm": 2.373269557952881,
+      "learning_rate": 9.858038924307363e-06,
+      "loss": 0.316,
+      "step": 213
+    },
+    {
+      "epoch": 0.1044921875,
+      "grad_norm": 2.517578125,
+      "learning_rate": 9.856161516265445e-06,
+      "loss": 0.3729,
+      "step": 214
+    },
+    {
+      "epoch": 0.10498046875,
+      "grad_norm": 3.9366421699523926,
+      "learning_rate": 9.854271956600463e-06,
+      "loss": 0.3119,
+      "step": 215
+    },
+    {
+      "epoch": 0.10546875,
+      "grad_norm": 3.0418357849121094,
+      "learning_rate": 9.852370250040682e-06,
+      "loss": 0.3799,
+      "step": 216
+    },
+    {
+      "epoch": 0.10595703125,
+      "grad_norm": 2.486046314239502,
+      "learning_rate": 9.85045640134476e-06,
+      "loss": 0.3761,
+      "step": 217
+    },
+    {
+      "epoch": 0.1064453125,
+      "grad_norm": 3.757772207260132,
+      "learning_rate": 9.848530415301748e-06,
+      "loss": 0.3281,
+      "step": 218
+    },
+    {
+      "epoch": 0.10693359375,
+      "grad_norm": 5.470198631286621,
+      "learning_rate": 9.846592296731052e-06,
+      "loss": 0.3626,
+      "step": 219
+    },
+    {
+      "epoch": 0.107421875,
+      "grad_norm": 2.6514899730682373,
+      "learning_rate": 9.84464205048245e-06,
+      "loss": 0.3312,
+      "step": 220
+    },
+    {
+      "epoch": 0.10791015625,
+      "grad_norm": 2.359720230102539,
+      "learning_rate": 9.842679681436062e-06,
+      "loss": 0.3332,
+      "step": 221
+    },
+    {
+      "epoch": 0.1083984375,
+      "grad_norm": 2.7306034564971924,
+      "learning_rate": 9.840705194502349e-06,
+      "loss": 0.3623,
+      "step": 222
+    },
+    {
+      "epoch": 0.10888671875,
+      "grad_norm": 2.2408559322357178,
+      "learning_rate": 9.838718594622083e-06,
+      "loss": 0.3579,
+      "step": 223
+    },
+    {
+      "epoch": 0.109375,
+      "grad_norm": 1.9728875160217285,
+      "learning_rate": 9.836719886766357e-06,
+      "loss": 0.3411,
+      "step": 224
+    },
+    {
+      "epoch": 0.10986328125,
+      "grad_norm": 2.826547861099243,
+      "learning_rate": 9.83470907593656e-06,
+      "loss": 0.2803,
+      "step": 225
+    },
+    {
+      "epoch": 0.1103515625,
+      "grad_norm": 2.5550942420959473,
+      "learning_rate": 9.832686167164361e-06,
+      "loss": 0.3537,
+      "step": 226
+    },
+    {
+      "epoch": 0.11083984375,
+      "grad_norm": 2.6079165935516357,
+      "learning_rate": 9.830651165511707e-06,
+      "loss": 0.3527,
+      "step": 227
+    },
+    {
+      "epoch": 0.111328125,
+      "grad_norm": 2.2585561275482178,
+      "learning_rate": 9.828604076070805e-06,
+      "loss": 0.3741,
+      "step": 228
+    },
+    {
+      "epoch": 0.11181640625,
+      "grad_norm": 2.335930585861206,
+      "learning_rate": 9.826544903964105e-06,
+      "loss": 0.34,
+      "step": 229
+    },
+    {
+      "epoch": 0.1123046875,
+      "grad_norm": 2.3235063552856445,
+      "learning_rate": 9.824473654344297e-06,
+      "loss": 0.3691,
+      "step": 230
+    },
+    {
+      "epoch": 0.11279296875,
+      "grad_norm": 3.584376811981201,
+      "learning_rate": 9.82239033239429e-06,
+      "loss": 0.3548,
+      "step": 231
+    },
+    {
+      "epoch": 0.11328125,
+      "grad_norm": 3.483834743499756,
+      "learning_rate": 9.820294943327202e-06,
+      "loss": 0.3905,
+      "step": 232
+    },
+    {
+      "epoch": 0.11376953125,
+      "grad_norm": 2.4160964488983154,
+      "learning_rate": 9.818187492386346e-06,
+      "loss": 0.3723,
+      "step": 233
+    },
+    {
+      "epoch": 0.1142578125,
+      "grad_norm": 2.206505298614502,
+      "learning_rate": 9.816067984845218e-06,
+      "loss": 0.3572,
+      "step": 234
+    },
+    {
+      "epoch": 0.11474609375,
+      "grad_norm": 2.8877620697021484,
+      "learning_rate": 9.813936426007487e-06,
+      "loss": 0.3486,
+      "step": 235
+    },
+    {
+      "epoch": 0.115234375,
+      "grad_norm": 2.2150516510009766,
+      "learning_rate": 9.81179282120697e-06,
+      "loss": 0.3431,
+      "step": 236
+    },
+    {
+      "epoch": 0.11572265625,
+      "grad_norm": 4.500147819519043,
+      "learning_rate": 9.809637175807634e-06,
+      "loss": 0.3465,
+      "step": 237
+    },
+    {
+      "epoch": 0.1162109375,
+      "grad_norm": 2.428119659423828,
+      "learning_rate": 9.80746949520357e-06,
+      "loss": 0.3193,
+      "step": 238
+    },
+    {
+      "epoch": 0.11669921875,
+      "grad_norm": 4.387357711791992,
+      "learning_rate": 9.805289784818991e-06,
+      "loss": 0.3789,
+      "step": 239
+    },
+    {
+      "epoch": 0.1171875,
+      "grad_norm": 2.6022865772247314,
+      "learning_rate": 9.803098050108206e-06,
+      "loss": 0.3744,
+      "step": 240
+    },
+    {
+      "epoch": 0.11767578125,
+      "grad_norm": 2.3189945220947266,
+      "learning_rate": 9.800894296555618e-06,
+      "loss": 0.3542,
+      "step": 241
+    },
+    {
+      "epoch": 0.1181640625,
+      "grad_norm": 2.428673505783081,
+      "learning_rate": 9.798678529675702e-06,
+      "loss": 0.354,
+      "step": 242
+    },
+    {
+      "epoch": 0.11865234375,
+      "grad_norm": 2.112927198410034,
+      "learning_rate": 9.796450755012992e-06,
+      "loss": 0.3541,
+      "step": 243
+    },
+    {
+      "epoch": 0.119140625,
+      "grad_norm": 3.9023051261901855,
+      "learning_rate": 9.794210978142073e-06,
+      "loss": 0.3902,
+      "step": 244
+    },
+    {
+      "epoch": 0.11962890625,
+      "grad_norm": 2.621843099594116,
+      "learning_rate": 9.79195920466756e-06,
+      "loss": 0.35,
+      "step": 245
+    },
+    {
+      "epoch": 0.1201171875,
+      "grad_norm": 2.8156723976135254,
+      "learning_rate": 9.789695440224094e-06,
+      "loss": 0.3562,
+      "step": 246
+    },
+    {
+      "epoch": 0.12060546875,
+      "grad_norm": 4.237185001373291,
+      "learning_rate": 9.78741969047631e-06,
+      "loss": 0.3596,
+      "step": 247
+    },
+    {
+      "epoch": 0.12109375,
+      "grad_norm": 2.050010919570923,
+      "learning_rate": 9.785131961118843e-06,
+      "loss": 0.3562,
+      "step": 248
+    },
+    {
+      "epoch": 0.12158203125,
+      "grad_norm": 2.1943752765655518,
+      "learning_rate": 9.782832257876302e-06,
+      "loss": 0.3147,
+      "step": 249
+    },
+    {
+      "epoch": 0.1220703125,
+      "grad_norm": 3.3409993648529053,
+      "learning_rate": 9.780520586503258e-06,
+      "loss": 0.4023,
+      "step": 250
+    },
+    {
+      "epoch": 0.12255859375,
+      "grad_norm": 2.073791027069092,
+      "learning_rate": 9.77819695278423e-06,
+      "loss": 0.3323,
+      "step": 251
+    },
+    {
+      "epoch": 0.123046875,
+      "grad_norm": 2.773463010787964,
+      "learning_rate": 9.77586136253367e-06,
+      "loss": 0.3461,
+      "step": 252
+    },
+    {
+      "epoch": 0.12353515625,
+      "grad_norm": 2.2921154499053955,
+      "learning_rate": 9.773513821595951e-06,
+      "loss": 0.3344,
+      "step": 253
+    },
+    {
+      "epoch": 0.1240234375,
+      "grad_norm": 2.6613571643829346,
+      "learning_rate": 9.771154335845345e-06,
+      "loss": 0.348,
+      "step": 254
+    },
+    {
+      "epoch": 0.12451171875,
+      "grad_norm": 8.336869239807129,
+      "learning_rate": 9.768782911186023e-06,
+      "loss": 0.3726,
+      "step": 255
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 2.428882360458374,
+      "learning_rate": 9.766399553552022e-06,
+      "loss": 0.3765,
+      "step": 256
+    },
+    {
+      "epoch": 0.12548828125,
+      "grad_norm": 1.8940154314041138,
+      "learning_rate": 9.764004268907244e-06,
+      "loss": 0.3407,
+      "step": 257
+    },
+    {
+      "epoch": 0.1259765625,
+      "grad_norm": 2.5715792179107666,
+      "learning_rate": 9.761597063245434e-06,
+      "loss": 0.3679,
+      "step": 258
+    },
+    {
+      "epoch": 0.12646484375,
+      "grad_norm": 2.1206367015838623,
+      "learning_rate": 9.759177942590166e-06,
+      "loss": 0.3409,
+      "step": 259
+    },
+    {
+      "epoch": 0.126953125,
+      "grad_norm": 2.5495412349700928,
+      "learning_rate": 9.756746912994832e-06,
+      "loss": 0.3499,
+      "step": 260
+    },
+    {
+      "epoch": 0.12744140625,
+      "grad_norm": 2.9602348804473877,
+      "learning_rate": 9.754303980542623e-06,
+      "loss": 0.3706,
+      "step": 261
+    },
+    {
+      "epoch": 0.1279296875,
+      "grad_norm": 2.7507028579711914,
+      "learning_rate": 9.751849151346513e-06,
+      "loss": 0.3767,
+      "step": 262
+    },
+    {
+      "epoch": 0.12841796875,
+      "grad_norm": 2.539034843444824,
+      "learning_rate": 9.749382431549247e-06,
+      "loss": 0.3406,
+      "step": 263
+    },
+    {
+      "epoch": 0.12890625,
+      "grad_norm": 2.833279848098755,
+      "learning_rate": 9.746903827323324e-06,
+      "loss": 0.3522,
+      "step": 264
+    },
+    {
+      "epoch": 0.12939453125,
+      "grad_norm": 2.5430469512939453,
+      "learning_rate": 9.74441334487098e-06,
+      "loss": 0.3406,
+      "step": 265
+    },
+    {
+      "epoch": 0.1298828125,
+      "grad_norm": 2.858895778656006,
+      "learning_rate": 9.741910990424173e-06,
+      "loss": 0.3396,
+      "step": 266
+    },
+    {
+      "epoch": 0.13037109375,
+      "grad_norm": 3.113898515701294,
+      "learning_rate": 9.739396770244575e-06,
+      "loss": 0.3779,
+      "step": 267
+    },
+    {
+      "epoch": 0.130859375,
+      "grad_norm": 2.812479257583618,
+      "learning_rate": 9.736870690623541e-06,
+      "loss": 0.3581,
+      "step": 268
+    },
+    {
+      "epoch": 0.13134765625,
+      "grad_norm": 4.137664318084717,
+      "learning_rate": 9.734332757882108e-06,
+      "loss": 0.3731,
+      "step": 269
+    },
+    {
+      "epoch": 0.1318359375,
+      "grad_norm": 2.346695899963379,
+      "learning_rate": 9.73178297837097e-06,
+      "loss": 0.3499,
+      "step": 270
+    },
+    {
+      "epoch": 0.13232421875,
+      "grad_norm": 3.5724024772644043,
+      "learning_rate": 9.729221358470468e-06,
+      "loss": 0.346,
+      "step": 271
+    },
+    {
+      "epoch": 0.1328125,
+      "grad_norm": 2.5001883506774902,
+      "learning_rate": 9.726647904590572e-06,
+      "loss": 0.3371,
+      "step": 272
+    },
+    {
+      "epoch": 0.13330078125,
+      "grad_norm": 1.8020128011703491,
+      "learning_rate": 9.724062623170855e-06,
+      "loss": 0.3632,
+      "step": 273
+    },
+    {
+      "epoch": 0.1337890625,
+      "grad_norm": 2.486666679382324,
+      "learning_rate": 9.721465520680501e-06,
+      "loss": 0.3505,
+      "step": 274
+    },
+    {
+      "epoch": 0.13427734375,
+      "grad_norm": 2.269751787185669,
+      "learning_rate": 9.718856603618263e-06,
+      "loss": 0.3718,
+      "step": 275
+    },
+    {
+      "epoch": 0.134765625,
+      "grad_norm": 2.7286322116851807,
+      "learning_rate": 9.716235878512462e-06,
+      "loss": 0.3462,
+      "step": 276
+    },
+    {
+      "epoch": 0.13525390625,
+      "grad_norm": 2.535698175430298,
+      "learning_rate": 9.713603351920964e-06,
+      "loss": 0.3451,
+      "step": 277
+    },
+    {
+      "epoch": 0.1357421875,
+      "grad_norm": 1.9008198976516724,
+      "learning_rate": 9.710959030431167e-06,
+      "loss": 0.3924,
+      "step": 278
+    },
+    {
+      "epoch": 0.13623046875,
+      "grad_norm": 2.339395046234131,
+      "learning_rate": 9.708302920659987e-06,
+      "loss": 0.3331,
+      "step": 279
+    },
+    {
+      "epoch": 0.13671875,
+      "grad_norm": 2.376002550125122,
+      "learning_rate": 9.705635029253833e-06,
+      "loss": 0.3815,
+      "step": 280
+    },
+    {
+      "epoch": 0.13720703125,
+      "grad_norm": 2.245027780532837,
+      "learning_rate": 9.702955362888595e-06,
+      "loss": 0.3548,
+      "step": 281
+    },
+    {
+      "epoch": 0.1376953125,
+      "grad_norm": 2.206878900527954,
+      "learning_rate": 9.700263928269636e-06,
+      "loss": 0.3204,
+      "step": 282
+    },
+    {
+      "epoch": 0.13818359375,
+      "grad_norm": 2.0215516090393066,
+      "learning_rate": 9.697560732131753e-06,
+      "loss": 0.3387,
+      "step": 283
+    },
+    {
+      "epoch": 0.138671875,
+      "grad_norm": 2.9142580032348633,
+      "learning_rate": 9.694845781239188e-06,
+      "loss": 0.3336,
+      "step": 284
+    },
+    {
+      "epoch": 0.13916015625,
+      "grad_norm": 2.0387048721313477,
+      "learning_rate": 9.692119082385588e-06,
+      "loss": 0.3342,
+      "step": 285
+    },
+    {
+      "epoch": 0.1396484375,
+      "grad_norm": 2.3236615657806396,
+      "learning_rate": 9.689380642393998e-06,
+      "loss": 0.3773,
+      "step": 286
+    },
+    {
+      "epoch": 0.14013671875,
+      "grad_norm": 3.4590189456939697,
+      "learning_rate": 9.686630468116846e-06,
+      "loss": 0.3358,
+      "step": 287
+    },
+    {
+      "epoch": 0.140625,
+      "grad_norm": 1.6319761276245117,
+      "learning_rate": 9.683868566435922e-06,
+      "loss": 0.2913,
+      "step": 288
+    },
+    {
+      "epoch": 0.14111328125,
+      "grad_norm": 6.874841690063477,
+      "learning_rate": 9.681094944262361e-06,
+      "loss": 0.3259,
+      "step": 289
+    },
+    {
+      "epoch": 0.1416015625,
+      "grad_norm": 4.962515830993652,
+      "learning_rate": 9.678309608536626e-06,
+      "loss": 0.3455,
+      "step": 290
+    },
+    {
+      "epoch": 0.14208984375,
+      "grad_norm": 3.334455966949463,
+      "learning_rate": 9.675512566228493e-06,
+      "loss": 0.3561,
+      "step": 291
+    },
+    {
+      "epoch": 0.142578125,
+      "grad_norm": 3.891530990600586,
+      "learning_rate": 9.672703824337026e-06,
+      "loss": 0.3627,
+      "step": 292
+    },
+    {
+      "epoch": 0.14306640625,
+      "grad_norm": 2.2160141468048096,
+      "learning_rate": 9.669883389890572e-06,
+      "loss": 0.312,
+      "step": 293
+    },
+    {
+      "epoch": 0.1435546875,
+      "grad_norm": 3.7108445167541504,
+      "learning_rate": 9.667051269946734e-06,
+      "loss": 0.338,
+      "step": 294
+    },
+    {
+      "epoch": 0.14404296875,
+      "grad_norm": 2.138221025466919,
+      "learning_rate": 9.664207471592353e-06,
+      "loss": 0.3767,
+      "step": 295
+    },
+    {
+      "epoch": 0.14453125,
+      "grad_norm": 11.57601547241211,
+      "learning_rate": 9.661352001943494e-06,
+      "loss": 0.3481,
+      "step": 296
+    },
+    {
+      "epoch": 0.14501953125,
+      "grad_norm": 2.1737406253814697,
+      "learning_rate": 9.658484868145428e-06,
+      "loss": 0.3319,
+      "step": 297
+    },
+    {
+      "epoch": 0.1455078125,
+      "grad_norm": 4.048387050628662,
+      "learning_rate": 9.655606077372619e-06,
+      "loss": 0.3061,
+      "step": 298
+    },
+    {
+      "epoch": 0.14599609375,
+      "grad_norm": 2.4968268871307373,
+      "learning_rate": 9.652715636828687e-06,
+      "loss": 0.333,
+      "step": 299
+    },
+    {
+      "epoch": 0.146484375,
+      "grad_norm": 2.2704763412475586,
+      "learning_rate": 9.649813553746416e-06,
+      "loss": 0.3307,
+      "step": 300
+    },
+    {
+      "epoch": 0.14697265625,
+      "grad_norm": 1.9303852319717407,
+      "learning_rate": 9.646899835387718e-06,
+      "loss": 0.3342,
+      "step": 301
+    },
+    {
+      "epoch": 0.1474609375,
+      "grad_norm": 2.8917553424835205,
+      "learning_rate": 9.64397448904362e-06,
+      "loss": 0.3595,
+      "step": 302
+    },
+    {
+      "epoch": 0.14794921875,
+      "grad_norm": 2.193105697631836,
+      "learning_rate": 9.641037522034246e-06,
+      "loss": 0.3675,
+      "step": 303
+    },
+    {
+      "epoch": 0.1484375,
+      "grad_norm": 1.9201539754867554,
+      "learning_rate": 9.638088941708799e-06,
+      "loss": 0.353,
+      "step": 304
+    },
+    {
+      "epoch": 0.14892578125,
+      "grad_norm": 2.513864517211914,
+      "learning_rate": 9.635128755445542e-06,
+      "loss": 0.3669,
+      "step": 305
+    },
+    {
+      "epoch": 0.1494140625,
+      "grad_norm": 2.397608518600464,
+      "learning_rate": 9.63215697065178e-06,
+      "loss": 0.3439,
+      "step": 306
+    },
+    {
+      "epoch": 0.14990234375,
+      "grad_norm": 2.335594654083252,
+      "learning_rate": 9.62917359476384e-06,
+      "loss": 0.3558,
+      "step": 307
+    },
+    {
+      "epoch": 0.150390625,
+      "grad_norm": 2.5134353637695312,
+      "learning_rate": 9.626178635247054e-06,
+      "loss": 0.3923,
+      "step": 308
+    },
+    {
+      "epoch": 0.15087890625,
+      "grad_norm": 2.9013524055480957,
+      "learning_rate": 9.623172099595743e-06,
+      "loss": 0.3748,
+      "step": 309
+    },
+    {
+      "epoch": 0.1513671875,
+      "grad_norm": 3.2646868228912354,
+      "learning_rate": 9.620153995333188e-06,
+      "loss": 0.3268,
+      "step": 310
+    },
+    {
+      "epoch": 0.15185546875,
+      "grad_norm": 2.843632459640503,
+      "learning_rate": 9.617124330011624e-06,
+      "loss": 0.3392,
+      "step": 311
+    },
+    {
+      "epoch": 0.15234375,
+      "grad_norm": 2.5182275772094727,
+      "learning_rate": 9.614083111212216e-06,
+      "loss": 0.3849,
+      "step": 312
+    },
+    {
+      "epoch": 0.15283203125,
+      "grad_norm": 2.9543368816375732,
+      "learning_rate": 9.611030346545035e-06,
+      "loss": 0.3784,
+      "step": 313
+    },
+    {
+      "epoch": 0.1533203125,
+      "grad_norm": 3.7902252674102783,
+      "learning_rate": 9.607966043649047e-06,
+      "loss": 0.3466,
+      "step": 314
+    },
+    {
+      "epoch": 0.15380859375,
+      "grad_norm": 2.4927687644958496,
+      "learning_rate": 9.604890210192084e-06,
+      "loss": 0.3638,
+      "step": 315
+    },
+    {
+      "epoch": 0.154296875,
+      "grad_norm": 4.722542762756348,
+      "learning_rate": 9.601802853870843e-06,
+      "loss": 0.3439,
+      "step": 316
+    },
+    {
+      "epoch": 0.15478515625,
+      "grad_norm": 2.0797646045684814,
+      "learning_rate": 9.598703982410842e-06,
+      "loss": 0.373,
+      "step": 317
+    },
+    {
+      "epoch": 0.1552734375,
+      "grad_norm": 2.1771399974823,
+      "learning_rate": 9.595593603566423e-06,
+      "loss": 0.3112,
+      "step": 318
+    },
+    {
+      "epoch": 0.15576171875,
+      "grad_norm": 2.621591091156006,
+      "learning_rate": 9.592471725120714e-06,
+      "loss": 0.3384,
+      "step": 319
+    },
+    {
+      "epoch": 0.15625,
+      "grad_norm": 4.34113883972168,
+      "learning_rate": 9.58933835488563e-06,
+      "loss": 0.3488,
+      "step": 320
+    },
+    {
+      "epoch": 0.15673828125,
+      "grad_norm": 3.58477783203125,
+      "learning_rate": 9.58619350070183e-06,
+      "loss": 0.3329,
+      "step": 321
+    },
+    {
+      "epoch": 0.1572265625,
+      "grad_norm": 2.657738208770752,
+      "learning_rate": 9.583037170438719e-06,
+      "loss": 0.3371,
+      "step": 322
+    },
+    {
+      "epoch": 0.15771484375,
+      "grad_norm": 2.3004322052001953,
+      "learning_rate": 9.579869371994412e-06,
+      "loss": 0.3658,
+      "step": 323
+    },
+    {
+      "epoch": 0.158203125,
+      "grad_norm": 3.4922330379486084,
+      "learning_rate": 9.576690113295726e-06,
+      "loss": 0.3713,
+      "step": 324
+    },
+    {
+      "epoch": 0.15869140625,
+      "grad_norm": 4.173436641693115,
+      "learning_rate": 9.573499402298152e-06,
+      "loss": 0.3349,
+      "step": 325
+    },
+    {
+      "epoch": 0.1591796875,
+      "grad_norm": 12.521305084228516,
+      "learning_rate": 9.570297246985838e-06,
+      "loss": 0.3411,
+      "step": 326
+    },
+    {
+      "epoch": 0.15966796875,
+      "grad_norm": 3.122694253921509,
+      "learning_rate": 9.567083655371572e-06,
+      "loss": 0.3644,
+      "step": 327
+    },
+    {
+      "epoch": 0.16015625,
+      "grad_norm": 1.6851651668548584,
+      "learning_rate": 9.563858635496755e-06,
+      "loss": 0.3567,
+      "step": 328
+    },
+    {
+      "epoch": 0.16064453125,
+      "grad_norm": 2.407923698425293,
+      "learning_rate": 9.56062219543139e-06,
+      "loss": 0.3298,
+      "step": 329
+    },
+    {
+      "epoch": 0.1611328125,
+      "grad_norm": 1.9536917209625244,
+      "learning_rate": 9.557374343274056e-06,
+      "loss": 0.352,
+      "step": 330
+    },
+    {
+      "epoch": 0.16162109375,
+      "grad_norm": 2.042382001876831,
+      "learning_rate": 9.55411508715188e-06,
+      "loss": 0.3249,
+      "step": 331
+    },
+    {
+      "epoch": 0.162109375,
+      "grad_norm": 1.9811147451400757,
+      "learning_rate": 9.55084443522054e-06,
+      "loss": 0.3341,
+      "step": 332
+    },
+    {
+      "epoch": 0.16259765625,
+      "grad_norm": 2.6401963233947754,
+      "learning_rate": 9.547562395664219e-06,
+      "loss": 0.3296,
+      "step": 333
+    },
+    {
+      "epoch": 0.1630859375,
+      "grad_norm": 2.3292157649993896,
+      "learning_rate": 9.544268976695596e-06,
+      "loss": 0.3446,
+      "step": 334
+    },
+    {
+      "epoch": 0.16357421875,
+      "grad_norm": 3.5120034217834473,
+      "learning_rate": 9.54096418655583e-06,
+      "loss": 0.3796,
+      "step": 335
+    },
+    {
+      "epoch": 0.1640625,
+      "grad_norm": 2.3993301391601562,
+      "learning_rate": 9.53764803351453e-06,
+      "loss": 0.3544,
+      "step": 336
+    },
+    {
+      "epoch": 0.16455078125,
+      "grad_norm": 2.403285264968872,
+      "learning_rate": 9.534320525869742e-06,
+      "loss": 0.3734,
+      "step": 337
+    },
+    {
+      "epoch": 0.1650390625,
+      "grad_norm": 1.878564476966858,
+      "learning_rate": 9.530981671947924e-06,
+      "loss": 0.3334,
+      "step": 338
+    },
+    {
+      "epoch": 0.16552734375,
+      "grad_norm": 3.3280200958251953,
+      "learning_rate": 9.527631480103919e-06,
+      "loss": 0.3282,
+      "step": 339
+    },
+    {
+      "epoch": 0.166015625,
+      "grad_norm": 2.304945230484009,
+      "learning_rate": 9.524269958720951e-06,
+      "loss": 0.3422,
+      "step": 340
+    },
+    {
+      "epoch": 0.16650390625,
+      "grad_norm": 2.0590991973876953,
+      "learning_rate": 9.520897116210588e-06,
+      "loss": 0.355,
+      "step": 341
+    },
+    {
+      "epoch": 0.1669921875,
+      "grad_norm": 1.660049557685852,
+      "learning_rate": 9.517512961012729e-06,
+      "loss": 0.3499,
+      "step": 342
+    },
+    {
+      "epoch": 0.16748046875,
+      "grad_norm": 1.8652247190475464,
+      "learning_rate": 9.514117501595582e-06,
+      "loss": 0.3594,
+      "step": 343
+    },
+    {
+      "epoch": 0.16796875,
+      "grad_norm": 1.7373839616775513,
+      "learning_rate": 9.510710746455636e-06,
+      "loss": 0.3447,
+      "step": 344
+    },
+    {
+      "epoch": 0.16845703125,
+      "grad_norm": 2.8204782009124756,
+      "learning_rate": 9.507292704117655e-06,
+      "loss": 0.362,
+      "step": 345
+    },
+    {
+      "epoch": 0.1689453125,
+      "grad_norm": 1.6446189880371094,
+      "learning_rate": 9.503863383134636e-06,
+      "loss": 0.3752,
+      "step": 346
+    },
+    {
+      "epoch": 0.16943359375,
+      "grad_norm": 3.4714109897613525,
+      "learning_rate": 9.500422792087809e-06,
+      "loss": 0.3358,
+      "step": 347
+    },
+    {
+      "epoch": 0.169921875,
+      "grad_norm": 2.125108003616333,
+      "learning_rate": 9.496970939586598e-06,
+      "loss": 0.3822,
+      "step": 348
+    },
+    {
+      "epoch": 0.17041015625,
+      "grad_norm": 2.7372467517852783,
+      "learning_rate": 9.493507834268609e-06,
+      "loss": 0.3513,
+      "step": 349
+    },
+    {
+      "epoch": 0.1708984375,
+      "grad_norm": 2.562140941619873,
+      "learning_rate": 9.490033484799608e-06,
+      "loss": 0.3727,
+      "step": 350
+    },
+    {
+      "epoch": 0.17138671875,
+      "grad_norm": 2.868966817855835,
+      "learning_rate": 9.486547899873495e-06,
+      "loss": 0.3309,
+      "step": 351
+    },
+    {
+      "epoch": 0.171875,
+      "grad_norm": 2.5418648719787598,
+      "learning_rate": 9.483051088212283e-06,
+      "loss": 0.3826,
+      "step": 352
+    },
+    {
+      "epoch": 0.17236328125,
+      "grad_norm": 1.7842854261398315,
+      "learning_rate": 9.479543058566081e-06,
+      "loss": 0.3404,
+      "step": 353
+    },
+    {
+      "epoch": 0.1728515625,
+      "grad_norm": 1.8991374969482422,
+      "learning_rate": 9.47602381971307e-06,
+      "loss": 0.3946,
+      "step": 354
+    },
+    {
+      "epoch": 0.17333984375,
+      "grad_norm": 1.9261831045150757,
+      "learning_rate": 9.472493380459474e-06,
+      "loss": 0.3579,
+      "step": 355
+    },
+    {
+      "epoch": 0.173828125,
+      "grad_norm": 1.6657100915908813,
+      "learning_rate": 9.468951749639552e-06,
+      "loss": 0.3405,
+      "step": 356
+    },
+    {
+      "epoch": 0.17431640625,
+      "grad_norm": 2.1538491249084473,
+      "learning_rate": 9.465398936115557e-06,
+      "loss": 0.3657,
+      "step": 357
+    },
+    {
+      "epoch": 0.1748046875,
+      "grad_norm": 1.8424322605133057,
+      "learning_rate": 9.461834948777738e-06,
+      "loss": 0.3685,
+      "step": 358
+    },
+    {
+      "epoch": 0.17529296875,
+      "grad_norm": 3.16018009185791,
+      "learning_rate": 9.458259796544293e-06,
+      "loss": 0.3225,
+      "step": 359
+    },
+    {
+      "epoch": 0.17578125,
+      "grad_norm": 1.7529760599136353,
+      "learning_rate": 9.454673488361363e-06,
+      "loss": 0.3428,
+      "step": 360
+    },
+    {
+      "epoch": 0.17626953125,
+      "grad_norm": 1.6713848114013672,
+      "learning_rate": 9.451076033203003e-06,
+      "loss": 0.3383,
+      "step": 361
+    },
+    {
+      "epoch": 0.1767578125,
+      "grad_norm": 2.688614845275879,
+      "learning_rate": 9.447467440071165e-06,
+      "loss": 0.3553,
+      "step": 362
+    },
+    {
+      "epoch": 0.17724609375,
+      "grad_norm": 2.0093319416046143,
+      "learning_rate": 9.443847717995666e-06,
+      "loss": 0.3689,
+      "step": 363
+    },
+    {
+      "epoch": 0.177734375,
+      "grad_norm": 5.026141166687012,
+      "learning_rate": 9.440216876034177e-06,
+      "loss": 0.3072,
+      "step": 364
+    },
+    {
+      "epoch": 0.17822265625,
+      "grad_norm": 2.687075138092041,
+      "learning_rate": 9.436574923272188e-06,
+      "loss": 0.3624,
+      "step": 365
+    },
+    {
+      "epoch": 0.1787109375,
+      "grad_norm": 1.9798976182937622,
+      "learning_rate": 9.432921868822997e-06,
+      "loss": 0.3355,
+      "step": 366
+    },
+    {
+      "epoch": 0.17919921875,
+      "grad_norm": 2.060910701751709,
+      "learning_rate": 9.42925772182768e-06,
+      "loss": 0.3435,
+      "step": 367
+    },
+    {
+      "epoch": 0.1796875,
+      "grad_norm": 1.7003917694091797,
+      "learning_rate": 9.425582491455068e-06,
+      "loss": 0.3659,
+      "step": 368
+    },
+    {
+      "epoch": 0.18017578125,
+      "grad_norm": 2.026036262512207,
+      "learning_rate": 9.421896186901729e-06,
+      "loss": 0.3523,
+      "step": 369
+    },
+    {
+      "epoch": 0.1806640625,
+      "grad_norm": 1.9931825399398804,
+      "learning_rate": 9.418198817391941e-06,
+      "loss": 0.3654,
+      "step": 370
+    },
+    {
+      "epoch": 0.18115234375,
+      "grad_norm": 2.7290432453155518,
+      "learning_rate": 9.41449039217767e-06,
+      "loss": 0.3599,
+      "step": 371
+    },
+    {
+      "epoch": 0.181640625,
+      "grad_norm": 1.5444127321243286,
+      "learning_rate": 9.410770920538545e-06,
+      "loss": 0.2991,
+      "step": 372
+    },
+    {
+      "epoch": 0.18212890625,
+      "grad_norm": 2.319566011428833,
+      "learning_rate": 9.407040411781843e-06,
+      "loss": 0.3724,
+      "step": 373
+    },
+    {
+      "epoch": 0.1826171875,
+      "grad_norm": 1.9856535196304321,
+      "learning_rate": 9.403298875242448e-06,
+      "loss": 0.348,
+      "step": 374
+    },
+    {
+      "epoch": 0.18310546875,
+      "grad_norm": 1.9270925521850586,
+      "learning_rate": 9.39954632028285e-06,
+      "loss": 0.3766,
+      "step": 375
+    },
+    {
+      "epoch": 0.18359375,
+      "grad_norm": 2.2769391536712646,
+      "learning_rate": 9.395782756293104e-06,
+      "loss": 0.3563,
+      "step": 376
+    },
+    {
+      "epoch": 0.18408203125,
+      "grad_norm": 2.2026526927948,
+      "learning_rate": 9.392008192690816e-06,
+      "loss": 0.3213,
+      "step": 377
+    },
+    {
+      "epoch": 0.1845703125,
+      "grad_norm": 2.3757741451263428,
+      "learning_rate": 9.388222638921116e-06,
+      "loss": 0.3595,
+      "step": 378
+    },
+    {
+      "epoch": 0.18505859375,
+      "grad_norm": 1.9485424757003784,
+      "learning_rate": 9.384426104456632e-06,
+      "loss": 0.3561,
+      "step": 379
+    },
+    {
+      "epoch": 0.185546875,
+      "grad_norm": 2.7337324619293213,
+      "learning_rate": 9.380618598797473e-06,
+      "loss": 0.38,
+      "step": 380
+    },
+    {
+      "epoch": 0.18603515625,
+      "grad_norm": 2.1130242347717285,
+      "learning_rate": 9.3768001314712e-06,
+      "loss": 0.3533,
+      "step": 381
+    },
+    {
+      "epoch": 0.1865234375,
+      "grad_norm": 1.831874966621399,
+      "learning_rate": 9.372970712032803e-06,
+      "loss": 0.332,
+      "step": 382
+    },
+    {
+      "epoch": 0.18701171875,
+      "grad_norm": 2.3811991214752197,
+      "learning_rate": 9.369130350064677e-06,
+      "loss": 0.3798,
+      "step": 383
+    },
+    {
+      "epoch": 0.1875,
+      "grad_norm": 1.8242988586425781,
+      "learning_rate": 9.3652790551766e-06,
+      "loss": 0.3634,
+      "step": 384
+    },
+    {
+      "epoch": 0.18798828125,
+      "grad_norm": 3.14345645904541,
+      "learning_rate": 9.361416837005705e-06,
+      "loss": 0.3513,
+      "step": 385
+    },
+    {
+      "epoch": 0.1884765625,
+      "grad_norm": 1.9473716020584106,
+      "learning_rate": 9.357543705216465e-06,
+      "loss": 0.3687,
+      "step": 386
+    },
+    {
+      "epoch": 0.18896484375,
+      "grad_norm": 1.982612133026123,
+      "learning_rate": 9.353659669500652e-06,
+      "loss": 0.3803,
+      "step": 387
+    },
+    {
+      "epoch": 0.189453125,
+      "grad_norm": 1.774999976158142,
+      "learning_rate": 9.349764739577334e-06,
+      "loss": 0.3331,
+      "step": 388
+    },
+    {
+      "epoch": 0.18994140625,
+      "grad_norm": 1.5273141860961914,
+      "learning_rate": 9.34585892519283e-06,
+      "loss": 0.3599,
+      "step": 389
+    },
+    {
+      "epoch": 0.1904296875,
+      "grad_norm": 1.8035123348236084,
+      "learning_rate": 9.3419422361207e-06,
+      "loss": 0.3771,
+      "step": 390
+    },
+    {
+      "epoch": 0.19091796875,
+      "grad_norm": 1.789610505104065,
+      "learning_rate": 9.338014682161719e-06,
+      "loss": 0.3236,
+      "step": 391
+    },
+    {
+      "epoch": 0.19140625,
+      "grad_norm": 1.9845644235610962,
+      "learning_rate": 9.334076273143843e-06,
+      "loss": 0.3274,
+      "step": 392
+    },
+    {
+      "epoch": 0.19189453125,
+      "grad_norm": 2.072159767150879,
+      "learning_rate": 9.330127018922195e-06,
+      "loss": 0.3416,
+      "step": 393
+    },
+    {
+      "epoch": 0.1923828125,
+      "grad_norm": 1.8441466093063354,
+      "learning_rate": 9.326166929379032e-06,
+      "loss": 0.3352,
+      "step": 394
+    },
+    {
+      "epoch": 0.19287109375,
+      "grad_norm": 2.479971170425415,
+      "learning_rate": 9.322196014423729e-06,
+      "loss": 0.3472,
+      "step": 395
+    },
+    {
+      "epoch": 0.193359375,
+      "grad_norm": 2.514597177505493,
+      "learning_rate": 9.318214283992747e-06,
+      "loss": 0.3544,
+      "step": 396
+    },
+    {
+      "epoch": 0.19384765625,
+      "grad_norm": 2.048144578933716,
+      "learning_rate": 9.314221748049613e-06,
+      "loss": 0.3869,
+      "step": 397
+    },
+    {
+      "epoch": 0.1943359375,
+      "grad_norm": 2.8453140258789062,
+      "learning_rate": 9.310218416584887e-06,
+      "loss": 0.3734,
+      "step": 398
+    },
+    {
+      "epoch": 0.19482421875,
+      "grad_norm": 1.6406381130218506,
+      "learning_rate": 9.306204299616148e-06,
+      "loss": 0.3507,
+      "step": 399
+    },
+    {
+      "epoch": 0.1953125,
+      "grad_norm": 2.275040626525879,
+      "learning_rate": 9.302179407187965e-06,
+      "loss": 0.3787,
+      "step": 400
+    },
+    {
+      "epoch": 0.19580078125,
+      "grad_norm": 1.522905945777893,
+      "learning_rate": 9.298143749371865e-06,
+      "loss": 0.341,
+      "step": 401
+    },
+    {
+      "epoch": 0.1962890625,
+      "grad_norm": 2.3068466186523438,
+      "learning_rate": 9.294097336266317e-06,
+      "loss": 0.3686,
+      "step": 402
+    },
+    {
+      "epoch": 0.19677734375,
+      "grad_norm": 2.8621833324432373,
+      "learning_rate": 9.290040177996703e-06,
+      "loss": 0.3331,
+      "step": 403
+    },
+    {
+      "epoch": 0.197265625,
+      "grad_norm": 2.339892864227295,
+      "learning_rate": 9.285972284715291e-06,
+      "loss": 0.3889,
+      "step": 404
+    },
+    {
+      "epoch": 0.19775390625,
+      "grad_norm": 1.7295536994934082,
+      "learning_rate": 9.281893666601214e-06,
+      "loss": 0.3692,
+      "step": 405
+    },
+    {
+      "epoch": 0.1982421875,
+      "grad_norm": 4.145984649658203,
+      "learning_rate": 9.277804333860435e-06,
+      "loss": 0.3387,
+      "step": 406
+    },
+    {
+      "epoch": 0.19873046875,
+      "grad_norm": 1.866166114807129,
+      "learning_rate": 9.273704296725741e-06,
+      "loss": 0.3503,
+      "step": 407
+    },
+    {
+      "epoch": 0.19921875,
+      "grad_norm": 1.8600391149520874,
+      "learning_rate": 9.269593565456691e-06,
+      "loss": 0.347,
+      "step": 408
+    },
+    {
+      "epoch": 0.19970703125,
+      "grad_norm": 1.990860104560852,
+      "learning_rate": 9.265472150339615e-06,
+      "loss": 0.3642,
+      "step": 409
+    },
+    {
+      "epoch": 0.2001953125,
+      "grad_norm": 1.4612618684768677,
+      "learning_rate": 9.26134006168757e-06,
+      "loss": 0.3624,
+      "step": 410
+    },
+    {
+      "epoch": 0.20068359375,
+      "grad_norm": 1.4518144130706787,
+      "learning_rate": 9.257197309840322e-06,
+      "loss": 0.3374,
+      "step": 411
+    },
+    {
+      "epoch": 0.201171875,
+      "grad_norm": 1.5550000667572021,
+      "learning_rate": 9.253043905164327e-06,
+      "loss": 0.3651,
+      "step": 412
+    },
+    {
+      "epoch": 0.20166015625,
+      "grad_norm": 1.9353028535842896,
+      "learning_rate": 9.248879858052688e-06,
+      "loss": 0.3111,
+      "step": 413
+    },
+    {
+      "epoch": 0.2021484375,
+      "grad_norm": 1.5865511894226074,
+      "learning_rate": 9.244705178925146e-06,
+      "loss": 0.3734,
+      "step": 414
+    },
+    {
+      "epoch": 0.20263671875,
+      "grad_norm": 1.9505976438522339,
+      "learning_rate": 9.24051987822804e-06,
+      "loss": 0.3294,
+      "step": 415
+    },
+    {
+      "epoch": 0.203125,
+      "grad_norm": 1.7402981519699097,
+      "learning_rate": 9.236323966434296e-06,
+      "loss": 0.3664,
+      "step": 416
+    },
+    {
+      "epoch": 0.20361328125,
+      "grad_norm": 2.2276546955108643,
+      "learning_rate": 9.232117454043383e-06,
+      "loss": 0.3943,
+      "step": 417
+    },
+    {
+      "epoch": 0.2041015625,
+      "grad_norm": 2.5883917808532715,
+      "learning_rate": 9.227900351581303e-06,
+      "loss": 0.3759,
+      "step": 418
+    },
+    {
+      "epoch": 0.20458984375,
+      "grad_norm": 2.116527795791626,
+      "learning_rate": 9.223672669600552e-06,
+      "loss": 0.371,
+      "step": 419
+    },
+    {
+      "epoch": 0.205078125,
+      "grad_norm": 1.890336036682129,
+      "learning_rate": 9.219434418680107e-06,
+      "loss": 0.3208,
+      "step": 420
+    },
+    {
+      "epoch": 0.20556640625,
+      "grad_norm": 2.831151247024536,
+      "learning_rate": 9.215185609425383e-06,
+      "loss": 0.3283,
+      "step": 421
+    },
+    {
+      "epoch": 0.2060546875,
+      "grad_norm": 1.890857458114624,
+      "learning_rate": 9.21092625246822e-06,
+      "loss": 0.3634,
+      "step": 422
+    },
+    {
+      "epoch": 0.20654296875,
+      "grad_norm": 1.4543401002883911,
+      "learning_rate": 9.206656358466851e-06,
+      "loss": 0.3615,
+      "step": 423
+    },
+    {
+      "epoch": 0.20703125,
+      "grad_norm": 1.9577465057373047,
+      "learning_rate": 9.202375938105876e-06,
+      "loss": 0.364,
+      "step": 424
+    },
+    {
+      "epoch": 0.20751953125,
+      "grad_norm": 1.5794016122817993,
+      "learning_rate": 9.198085002096237e-06,
+      "loss": 0.34,
+      "step": 425
+    },
+    {
+      "epoch": 0.2080078125,
+      "grad_norm": 1.8114027976989746,
+      "learning_rate": 9.193783561175184e-06,
+      "loss": 0.3413,
+      "step": 426
+    },
+    {
+      "epoch": 0.20849609375,
+      "grad_norm": 1.5112391710281372,
+      "learning_rate": 9.189471626106261e-06,
+      "loss": 0.3558,
+      "step": 427
+    },
+    {
+      "epoch": 0.208984375,
+      "grad_norm": 1.5750012397766113,
+      "learning_rate": 9.185149207679263e-06,
+      "loss": 0.3211,
+      "step": 428
+    },
+    {
+      "epoch": 0.20947265625,
+      "grad_norm": 1.5355925559997559,
+      "learning_rate": 9.180816316710226e-06,
+      "loss": 0.316,
+      "step": 429
+    },
+    {
+      "epoch": 0.2099609375,
+      "grad_norm": 1.7540535926818848,
+      "learning_rate": 9.176472964041385e-06,
+      "loss": 0.3446,
+      "step": 430
+    },
+    {
+      "epoch": 0.21044921875,
+      "grad_norm": 1.94683837890625,
+      "learning_rate": 9.172119160541158e-06,
+      "loss": 0.3894,
+      "step": 431
+    },
+    {
+      "epoch": 0.2109375,
+      "grad_norm": 2.1505014896392822,
+      "learning_rate": 9.167754917104112e-06,
+      "loss": 0.3516,
+      "step": 432
+    },
+    {
+      "epoch": 0.21142578125,
+      "grad_norm": 3.6382253170013428,
+      "learning_rate": 9.163380244650938e-06,
+      "loss": 0.3766,
+      "step": 433
+    },
+    {
+      "epoch": 0.2119140625,
+      "grad_norm": 1.4218906164169312,
+      "learning_rate": 9.158995154128425e-06,
+      "loss": 0.3377,
+      "step": 434
+    },
+    {
+      "epoch": 0.21240234375,
+      "grad_norm": 1.6487233638763428,
+      "learning_rate": 9.15459965650943e-06,
+      "loss": 0.3198,
+      "step": 435
+    },
+    {
+      "epoch": 0.212890625,
+      "grad_norm": 6.333557605743408,
+      "learning_rate": 9.15019376279285e-06,
+      "loss": 0.3336,
+      "step": 436
+    },
+    {
+      "epoch": 0.21337890625,
+      "grad_norm": 1.746251106262207,
+      "learning_rate": 9.1457774840036e-06,
+      "loss": 0.3434,
+      "step": 437
+    },
+    {
+      "epoch": 0.2138671875,
+      "grad_norm": 2.1596200466156006,
+      "learning_rate": 9.14135083119258e-06,
+      "loss": 0.3496,
+      "step": 438
+    },
+    {
+      "epoch": 0.21435546875,
+      "grad_norm": 1.7951174974441528,
+      "learning_rate": 9.13691381543665e-06,
+      "loss": 0.3589,
+      "step": 439
+    },
+    {
+      "epoch": 0.21484375,
+      "grad_norm": 1.7067686319351196,
+      "learning_rate": 9.132466447838598e-06,
+      "loss": 0.3367,
+      "step": 440
+    },
+    {
+      "epoch": 0.21533203125,
+      "grad_norm": 2.095935344696045,
+      "learning_rate": 9.128008739527119e-06,
+      "loss": 0.3305,
+      "step": 441
+    },
+    {
+      "epoch": 0.2158203125,
+      "grad_norm": 2.011528968811035,
+      "learning_rate": 9.123540701656782e-06,
+      "loss": 0.368,
+      "step": 442
+    },
+    {
+      "epoch": 0.21630859375,
+      "grad_norm": 1.4319236278533936,
+      "learning_rate": 9.119062345408005e-06,
+      "loss": 0.3288,
+      "step": 443
+    },
+    {
+      "epoch": 0.216796875,
+      "grad_norm": 1.8978536128997803,
+      "learning_rate": 9.114573681987024e-06,
+      "loss": 0.3222,
+      "step": 444
+    },
+    {
+      "epoch": 0.21728515625,
+      "grad_norm": 1.8402870893478394,
+      "learning_rate": 9.11007472262587e-06,
+      "loss": 0.3286,
+      "step": 445
+    },
+    {
+      "epoch": 0.2177734375,
+      "grad_norm": 1.8938474655151367,
+      "learning_rate": 9.105565478582335e-06,
+      "loss": 0.3725,
+      "step": 446
+    },
+    {
+      "epoch": 0.21826171875,
+      "grad_norm": 1.723388433456421,
+      "learning_rate": 9.101045961139945e-06,
+      "loss": 0.3634,
+      "step": 447
+    },
+    {
+      "epoch": 0.21875,
+      "grad_norm": 1.8326998949050903,
+      "learning_rate": 9.096516181607935e-06,
+      "loss": 0.3276,
+      "step": 448
+    },
+    {
+      "epoch": 0.21923828125,
+      "grad_norm": 1.6433813571929932,
+      "learning_rate": 9.09197615132122e-06,
+      "loss": 0.3637,
+      "step": 449
+    },
+    {
+      "epoch": 0.2197265625,
+      "grad_norm": 1.482116460800171,
+      "learning_rate": 9.087425881640366e-06,
+      "loss": 0.3413,
+      "step": 450
+    },
+    {
+      "epoch": 0.22021484375,
+      "grad_norm": 5.252507209777832,
+      "learning_rate": 9.082865383951558e-06,
+      "loss": 0.35,
+      "step": 451
+    },
+    {
+      "epoch": 0.220703125,
+      "grad_norm": 1.4982550144195557,
+      "learning_rate": 9.078294669666577e-06,
+      "loss": 0.3354,
+      "step": 452
+    },
+    {
+      "epoch": 0.22119140625,
+      "grad_norm": 2.408413887023926,
+      "learning_rate": 9.073713750222766e-06,
+      "loss": 0.3376,
+      "step": 453
+    },
+    {
+      "epoch": 0.2216796875,
+      "grad_norm": 1.682771921157837,
+      "learning_rate": 9.069122637083012e-06,
+      "loss": 0.3131,
+      "step": 454
+    },
+    {
+      "epoch": 0.22216796875,
+      "grad_norm": 1.6665334701538086,
+      "learning_rate": 9.064521341735702e-06,
+      "loss": 0.3348,
+      "step": 455
+    },
+    {
+      "epoch": 0.22265625,
+      "grad_norm": 1.3198261260986328,
+      "learning_rate": 9.059909875694703e-06,
+      "loss": 0.3087,
+      "step": 456
+    },
+    {
+      "epoch": 0.22314453125,
+      "grad_norm": 2.0489742755889893,
+      "learning_rate": 9.055288250499339e-06,
+      "loss": 0.3549,
+      "step": 457
+    },
+    {
+      "epoch": 0.2236328125,
+      "grad_norm": 1.4335616827011108,
+      "learning_rate": 9.050656477714345e-06,
+      "loss": 0.3859,
+      "step": 458
+    },
+    {
+      "epoch": 0.22412109375,
+      "grad_norm": 1.9734736680984497,
+      "learning_rate": 9.046014568929856e-06,
+      "loss": 0.358,
+      "step": 459
+    },
+    {
+      "epoch": 0.224609375,
+      "grad_norm": 1.8493421077728271,
+      "learning_rate": 9.04136253576137e-06,
+      "loss": 0.3306,
+      "step": 460
+    },
+    {
+      "epoch": 0.22509765625,
+      "grad_norm": 2.6172261238098145,
+      "learning_rate": 9.036700389849717e-06,
+      "loss": 0.3481,
+      "step": 461
+    },
+    {
+      "epoch": 0.2255859375,
+      "grad_norm": 1.538042664527893,
+      "learning_rate": 9.03202814286103e-06,
+      "loss": 0.3154,
+      "step": 462
+    },
+    {
+      "epoch": 0.22607421875,
+      "grad_norm": 2.418534278869629,
+      "learning_rate": 9.027345806486722e-06,
+      "loss": 0.3247,
+      "step": 463
+    },
+    {
+      "epoch": 0.2265625,
+      "grad_norm": 1.7823346853256226,
+      "learning_rate": 9.022653392443455e-06,
+      "loss": 0.338,
+      "step": 464
+    },
+    {
+      "epoch": 0.22705078125,
+      "grad_norm": 1.9469126462936401,
+      "learning_rate": 9.0179509124731e-06,
+      "loss": 0.3377,
+      "step": 465
+    },
+    {
+      "epoch": 0.2275390625,
+      "grad_norm": 1.985723614692688,
+      "learning_rate": 9.013238378342725e-06,
+      "loss": 0.3438,
+      "step": 466
+    },
+    {
+      "epoch": 0.22802734375,
+      "grad_norm": 1.5227419137954712,
+      "learning_rate": 9.008515801844552e-06,
+      "loss": 0.3392,
+      "step": 467
+    },
+    {
+      "epoch": 0.228515625,
+      "grad_norm": 2.764451026916504,
+      "learning_rate": 9.003783194795931e-06,
+      "loss": 0.3439,
+      "step": 468
+    },
+    {
+      "epoch": 0.22900390625,
+      "grad_norm": 1.489700198173523,
+      "learning_rate": 8.999040569039315e-06,
+      "loss": 0.3654,
+      "step": 469
+    },
+    {
+      "epoch": 0.2294921875,
+      "grad_norm": 2.0311126708984375,
+      "learning_rate": 8.994287936442226e-06,
+      "loss": 0.3312,
+      "step": 470
+    },
+    {
+      "epoch": 0.22998046875,
+      "grad_norm": 1.7580716609954834,
+      "learning_rate": 8.989525308897223e-06,
+      "loss": 0.3573,
+      "step": 471
+    },
+    {
+      "epoch": 0.23046875,
+      "grad_norm": 1.7429345846176147,
+      "learning_rate": 8.98475269832188e-06,
+      "loss": 0.3757,
+      "step": 472
+    },
+    {
+      "epoch": 0.23095703125,
+      "grad_norm": 1.544498085975647,
+      "learning_rate": 8.97997011665875e-06,
+      "loss": 0.2787,
+      "step": 473
+    },
+    {
+      "epoch": 0.2314453125,
+      "grad_norm": 1.6220890283584595,
+      "learning_rate": 8.975177575875335e-06,
+      "loss": 0.3597,
+      "step": 474
+    },
+    {
+      "epoch": 0.23193359375,
+      "grad_norm": 1.598620057106018,
+      "learning_rate": 8.97037508796406e-06,
+      "loss": 0.3615,
+      "step": 475
+    },
+    {
+      "epoch": 0.232421875,
+      "grad_norm": 1.567460298538208,
+      "learning_rate": 8.96556266494224e-06,
+      "loss": 0.3613,
+      "step": 476
+    },
+    {
+      "epoch": 0.23291015625,
+      "grad_norm": 1.5737589597702026,
+      "learning_rate": 8.960740318852051e-06,
+      "loss": 0.3699,
+      "step": 477
+    },
+    {
+      "epoch": 0.2333984375,
+      "grad_norm": 1.9563899040222168,
+      "learning_rate": 8.9559080617605e-06,
+      "loss": 0.3578,
+      "step": 478
+    },
+    {
+      "epoch": 0.23388671875,
+      "grad_norm": 2.225196599960327,
+      "learning_rate": 8.951065905759392e-06,
+      "loss": 0.3346,
+      "step": 479
+    },
+    {
+      "epoch": 0.234375,
+      "grad_norm": 1.5860683917999268,
+      "learning_rate": 8.946213862965306e-06,
+      "loss": 0.3741,
+      "step": 480
+    },
+    {
+      "epoch": 0.23486328125,
+      "grad_norm": 1.289207935333252,
+      "learning_rate": 8.941351945519557e-06,
+      "loss": 0.3434,
+      "step": 481
+    },
+    {
+      "epoch": 0.2353515625,
+      "grad_norm": 1.576648235321045,
+      "learning_rate": 8.936480165588174e-06,
+      "loss": 0.3513,
+      "step": 482
+    },
+    {
+      "epoch": 0.23583984375,
+      "grad_norm": 1.5328677892684937,
+      "learning_rate": 8.931598535361855e-06,
+      "loss": 0.3299,
+      "step": 483
+    },
+    {
+      "epoch": 0.236328125,
+      "grad_norm": 1.439266562461853,
+      "learning_rate": 8.926707067055963e-06,
+      "loss": 0.3077,
+      "step": 484
+    },
+    {
+      "epoch": 0.23681640625,
+      "grad_norm": 1.6571671962738037,
+      "learning_rate": 8.921805772910463e-06,
+      "loss": 0.3666,
+      "step": 485
+    },
+    {
+      "epoch": 0.2373046875,
+      "grad_norm": 2.0075385570526123,
+      "learning_rate": 8.916894665189918e-06,
+      "loss": 0.3695,
+      "step": 486
+    },
+    {
+      "epoch": 0.23779296875,
+      "grad_norm": 1.3680145740509033,
+      "learning_rate": 8.91197375618344e-06,
+      "loss": 0.3393,
+      "step": 487
+    },
+    {
+      "epoch": 0.23828125,
+      "grad_norm": 1.9149501323699951,
+      "learning_rate": 8.907043058204674e-06,
+      "loss": 0.3374,
+      "step": 488
+    },
+    {
+      "epoch": 0.23876953125,
+      "grad_norm": 1.5481083393096924,
+      "learning_rate": 8.902102583591755e-06,
+      "loss": 0.3263,
+      "step": 489
+    },
+    {
+      "epoch": 0.2392578125,
+      "grad_norm": 1.8688881397247314,
+      "learning_rate": 8.89715234470728e-06,
+      "loss": 0.3207,
+      "step": 490
+    },
+    {
+      "epoch": 0.23974609375,
+      "grad_norm": 1.846941351890564,
+      "learning_rate": 8.892192353938288e-06,
+      "loss": 0.3677,
+      "step": 491
+    },
+    {
+      "epoch": 0.240234375,
+      "grad_norm": 1.4003583192825317,
+      "learning_rate": 8.887222623696213e-06,
+      "loss": 0.3281,
+      "step": 492
+    },
+    {
+      "epoch": 0.24072265625,
+      "grad_norm": 1.9100502729415894,
+      "learning_rate": 8.882243166416862e-06,
+      "loss": 0.3685,
+      "step": 493
+    },
+    {
+      "epoch": 0.2412109375,
+      "grad_norm": 1.6730045080184937,
+      "learning_rate": 8.877253994560381e-06,
+      "loss": 0.3482,
+      "step": 494
+    },
+    {
+      "epoch": 0.24169921875,
+      "grad_norm": 1.4065086841583252,
+      "learning_rate": 8.87225512061123e-06,
+      "loss": 0.3404,
+      "step": 495
+    },
+    {
+      "epoch": 0.2421875,
+      "grad_norm": 1.5349781513214111,
+      "learning_rate": 8.867246557078141e-06,
+      "loss": 0.3279,
+      "step": 496
+    },
+    {
+      "epoch": 0.24267578125,
+      "grad_norm": 1.376725196838379,
+      "learning_rate": 8.862228316494094e-06,
+      "loss": 0.3384,
+      "step": 497
+    },
+    {
+      "epoch": 0.2431640625,
+      "grad_norm": 1.5585695505142212,
+      "learning_rate": 8.857200411416283e-06,
+      "loss": 0.3638,
+      "step": 498
+    },
+    {
+      "epoch": 0.24365234375,
+      "grad_norm": 3.5493311882019043,
+      "learning_rate": 8.852162854426087e-06,
+      "loss": 0.3561,
+      "step": 499
+    },
+    {
+      "epoch": 0.244140625,
+      "grad_norm": 2.1406612396240234,
+      "learning_rate": 8.84711565812904e-06,
+      "loss": 0.3097,
+      "step": 500
+    },
+    {
+      "epoch": 0.24462890625,
+      "grad_norm": 1.5322456359863281,
+      "learning_rate": 8.842058835154789e-06,
+      "loss": 0.36,
+      "step": 501
+    },
+    {
+      "epoch": 0.2451171875,
+      "grad_norm": 2.3245677947998047,
+      "learning_rate": 8.836992398157076e-06,
+      "loss": 0.3479,
+      "step": 502
+    },
+    {
+      "epoch": 0.24560546875,
+      "grad_norm": 1.8092581033706665,
+      "learning_rate": 8.831916359813702e-06,
+      "loss": 0.3292,
+      "step": 503
+    },
+    {
+      "epoch": 0.24609375,
+      "grad_norm": 1.6669384241104126,
+      "learning_rate": 8.826830732826484e-06,
+      "loss": 0.357,
+      "step": 504
+    },
+    {
+      "epoch": 0.24658203125,
+      "grad_norm": 1.3617286682128906,
+      "learning_rate": 8.821735529921243e-06,
+      "loss": 0.3434,
+      "step": 505
+    },
+    {
+      "epoch": 0.2470703125,
+      "grad_norm": 5.754039287567139,
+      "learning_rate": 8.816630763847756e-06,
+      "loss": 0.3677,
+      "step": 506
+    },
+    {
+      "epoch": 0.24755859375,
+      "grad_norm": 1.2652654647827148,
+      "learning_rate": 8.811516447379734e-06,
+      "loss": 0.3573,
+      "step": 507
+    },
+    {
+      "epoch": 0.248046875,
+      "grad_norm": 1.6732009649276733,
+      "learning_rate": 8.806392593314781e-06,
+      "loss": 0.3398,
+      "step": 508
+    },
+    {
+      "epoch": 0.24853515625,
+      "grad_norm": 1.280765175819397,
+      "learning_rate": 8.801259214474371e-06,
+      "loss": 0.3371,
+      "step": 509
+    },
+    {
+      "epoch": 0.2490234375,
+      "grad_norm": 1.2774041891098022,
+      "learning_rate": 8.796116323703811e-06,
+      "loss": 0.3466,
+      "step": 510
+    },
+    {
+      "epoch": 0.24951171875,
+      "grad_norm": 1.4741958379745483,
+      "learning_rate": 8.790963933872212e-06,
+      "loss": 0.3506,
+      "step": 511
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4504543542861938,
+      "learning_rate": 8.785802057872447e-06,
+      "loss": 0.4083,
+      "step": 512
+    },
+    {
+      "epoch": 0.25048828125,
+      "grad_norm": 1.4813644886016846,
+      "learning_rate": 8.780630708621135e-06,
+      "loss": 0.382,
+      "step": 513
+    },
+    {
+      "epoch": 0.2509765625,
+      "grad_norm": 1.6617738008499146,
+      "learning_rate": 8.775449899058597e-06,
+      "loss": 0.3387,
+      "step": 514
+    },
+    {
+      "epoch": 0.25146484375,
+      "grad_norm": 1.8677629232406616,
+      "learning_rate": 8.770259642148826e-06,
+      "loss": 0.3422,
+      "step": 515
+    },
+    {
+      "epoch": 0.251953125,
+      "grad_norm": 1.4123599529266357,
+      "learning_rate": 8.765059950879454e-06,
+      "loss": 0.3621,
+      "step": 516
+    },
+    {
+      "epoch": 0.25244140625,
+      "grad_norm": 1.966430902481079,
+      "learning_rate": 8.759850838261723e-06,
+      "loss": 0.3475,
+      "step": 517
+    },
+    {
+      "epoch": 0.2529296875,
+      "grad_norm": 1.3296693563461304,
+      "learning_rate": 8.754632317330448e-06,
+      "loss": 0.3938,
+      "step": 518
+    },
+    {
+      "epoch": 0.25341796875,
+      "grad_norm": 1.4010918140411377,
+      "learning_rate": 8.749404401143991e-06,
+      "loss": 0.3474,
+      "step": 519
+    },
+    {
+      "epoch": 0.25390625,
+      "grad_norm": 1.5129917860031128,
+      "learning_rate": 8.744167102784216e-06,
+      "loss": 0.3783,
+      "step": 520
+    },
+    {
+      "epoch": 0.25439453125,
+      "grad_norm": 1.7624212503433228,
+      "learning_rate": 8.738920435356473e-06,
+      "loss": 0.3272,
+      "step": 521
+    },
+    {
+      "epoch": 0.2548828125,
+      "grad_norm": 1.4559099674224854,
+      "learning_rate": 8.733664411989548e-06,
+      "loss": 0.3526,
+      "step": 522
+    },
+    {
+      "epoch": 0.25537109375,
+      "grad_norm": 1.8239963054656982,
+      "learning_rate": 8.728399045835648e-06,
+      "loss": 0.3385,
+      "step": 523
+    },
+    {
+      "epoch": 0.255859375,
+      "grad_norm": 1.4369486570358276,
+      "learning_rate": 8.723124350070347e-06,
+      "loss": 0.3193,
+      "step": 524
+    },
+    {
+      "epoch": 0.25634765625,
+      "grad_norm": 4.341763496398926,
+      "learning_rate": 8.717840337892575e-06,
+      "loss": 0.3256,
+      "step": 525
+    },
+    {
+      "epoch": 0.2568359375,
+      "grad_norm": 2.0711512565612793,
+      "learning_rate": 8.712547022524566e-06,
+      "loss": 0.3639,
+      "step": 526
+    },
+    {
+      "epoch": 0.25732421875,
+      "grad_norm": 1.4793862104415894,
+      "learning_rate": 8.707244417211844e-06,
+      "loss": 0.3166,
+      "step": 527
+    },
+    {
+      "epoch": 0.2578125,
+      "grad_norm": 1.742661476135254,
+      "learning_rate": 8.701932535223168e-06,
+      "loss": 0.3533,
+      "step": 528
+    },
+    {
+      "epoch": 0.25830078125,
+      "grad_norm": 1.4166213274002075,
+      "learning_rate": 8.696611389850516e-06,
+      "loss": 0.3436,
+      "step": 529
+    },
+    {
+      "epoch": 0.2587890625,
+      "grad_norm": 1.362882137298584,
+      "learning_rate": 8.691280994409044e-06,
+      "loss": 0.3165,
+      "step": 530
+    },
+    {
+      "epoch": 0.25927734375,
+      "grad_norm": 2.5286190509796143,
+      "learning_rate": 8.685941362237058e-06,
+      "loss": 0.3438,
+      "step": 531
+    },
+    {
+      "epoch": 0.259765625,
+      "grad_norm": 2.232900381088257,
+      "learning_rate": 8.680592506695972e-06,
+      "loss": 0.3389,
+      "step": 532
+    },
+    {
+      "epoch": 0.26025390625,
+      "grad_norm": 1.2126928567886353,
+      "learning_rate": 8.675234441170286e-06,
+      "loss": 0.306,
+      "step": 533
+    },
+    {
+      "epoch": 0.2607421875,
+      "grad_norm": 1.480934977531433,
+      "learning_rate": 8.669867179067538e-06,
+      "loss": 0.3696,
+      "step": 534
+    },
+    {
+      "epoch": 0.26123046875,
+      "grad_norm": 2.439810037612915,
+      "learning_rate": 8.664490733818289e-06,
+      "loss": 0.3628,
+      "step": 535
+    },
+    {
+      "epoch": 0.26171875,
+      "grad_norm": 1.3664276599884033,
+      "learning_rate": 8.659105118876068e-06,
+      "loss": 0.3534,
+      "step": 536
+    },
+    {
+      "epoch": 0.26220703125,
+      "grad_norm": 1.8439381122589111,
+      "learning_rate": 8.65371034771736e-06,
+      "loss": 0.3539,
+      "step": 537
+    },
+    {
+      "epoch": 0.2626953125,
+      "grad_norm": 2.1068308353424072,
+      "learning_rate": 8.64830643384155e-06,
+      "loss": 0.4281,
+      "step": 538
+    },
+    {
+      "epoch": 0.26318359375,
+      "grad_norm": 1.847388505935669,
+      "learning_rate": 8.642893390770912e-06,
+      "loss": 0.3624,
+      "step": 539
+    },
+    {
+      "epoch": 0.263671875,
+      "grad_norm": 2.783621311187744,
+      "learning_rate": 8.63747123205056e-06,
+      "loss": 0.3501,
+      "step": 540
+    },
+    {
+      "epoch": 0.26416015625,
+      "grad_norm": 5.078010559082031,
+      "learning_rate": 8.632039971248416e-06,
+      "loss": 0.3423,
+      "step": 541
+    },
+    {
+      "epoch": 0.2646484375,
+      "grad_norm": 1.461103916168213,
+      "learning_rate": 8.626599621955179e-06,
+      "loss": 0.3505,
+      "step": 542
+    },
+    {
+      "epoch": 0.26513671875,
+      "grad_norm": 1.512221336364746,
+      "learning_rate": 8.621150197784293e-06,
+      "loss": 0.344,
+      "step": 543
+    },
+    {
+      "epoch": 0.265625,
+      "grad_norm": 2.6210267543792725,
+      "learning_rate": 8.615691712371907e-06,
+      "loss": 0.3192,
+      "step": 544
+    },
+    {
+      "epoch": 0.26611328125,
+      "grad_norm": 1.5492252111434937,
+      "learning_rate": 8.610224179376847e-06,
+      "loss": 0.3217,
+      "step": 545
+    },
+    {
+      "epoch": 0.2666015625,
+      "grad_norm": 1.4719685316085815,
+      "learning_rate": 8.604747612480577e-06,
+      "loss": 0.3251,
+      "step": 546
+    },
+    {
+      "epoch": 0.26708984375,
+      "grad_norm": 1.9413729906082153,
+      "learning_rate": 8.599262025387165e-06,
+      "loss": 0.3658,
+      "step": 547
+    },
+    {
+      "epoch": 0.267578125,
+      "grad_norm": 1.8121291399002075,
+      "learning_rate": 8.593767431823255e-06,
+      "loss": 0.3274,
+      "step": 548
+    },
+    {
+      "epoch": 0.26806640625,
+      "grad_norm": 1.7863436937332153,
+      "learning_rate": 8.588263845538021e-06,
+      "loss": 0.3586,
+      "step": 549
+    },
+    {
+      "epoch": 0.2685546875,
+      "grad_norm": 2.253500461578369,
+      "learning_rate": 8.582751280303148e-06,
+      "loss": 0.383,
+      "step": 550
+    },
+    {
+      "epoch": 0.26904296875,
+      "grad_norm": 1.9108343124389648,
+      "learning_rate": 8.577229749912782e-06,
+      "loss": 0.3188,
+      "step": 551
+    },
+    {
+      "epoch": 0.26953125,
+      "grad_norm": 1.4474389553070068,
+      "learning_rate": 8.571699268183506e-06,
+      "loss": 0.3239,
+      "step": 552
+    },
+    {
+      "epoch": 0.27001953125,
+      "grad_norm": 1.6433511972427368,
+      "learning_rate": 8.566159848954305e-06,
+      "loss": 0.3565,
+      "step": 553
+    },
+    {
+      "epoch": 0.2705078125,
+      "grad_norm": 2.9185471534729004,
+      "learning_rate": 8.560611506086518e-06,
+      "loss": 0.3916,
+      "step": 554
+    },
+    {
+      "epoch": 0.27099609375,
+      "grad_norm": 1.6128103733062744,
+      "learning_rate": 8.555054253463828e-06,
+      "loss": 0.3518,
+      "step": 555
+    },
+    {
+      "epoch": 0.271484375,
+      "grad_norm": 1.3888630867004395,
+      "learning_rate": 8.549488104992201e-06,
+      "loss": 0.3772,
+      "step": 556
+    },
+    {
+      "epoch": 0.27197265625,
+      "grad_norm": 1.7909587621688843,
+      "learning_rate": 8.543913074599867e-06,
+      "loss": 0.3313,
+      "step": 557
+    },
+    {
+      "epoch": 0.2724609375,
+      "grad_norm": 1.6241544485092163,
+      "learning_rate": 8.538329176237287e-06,
+      "loss": 0.3535,
+      "step": 558
+    },
+    {
+      "epoch": 0.27294921875,
+      "grad_norm": 1.4434620141983032,
+      "learning_rate": 8.532736423877102e-06,
+      "loss": 0.3329,
+      "step": 559
+    },
+    {
+      "epoch": 0.2734375,
+      "grad_norm": 1.8953794240951538,
+      "learning_rate": 8.527134831514116e-06,
+      "loss": 0.3318,
+      "step": 560
+    },
+    {
+      "epoch": 0.27392578125,
+      "grad_norm": 1.287680983543396,
+      "learning_rate": 8.521524413165254e-06,
+      "loss": 0.3187,
+      "step": 561
+    },
+    {
+      "epoch": 0.2744140625,
+      "grad_norm": 1.6521981954574585,
+      "learning_rate": 8.51590518286952e-06,
+      "loss": 0.3509,
+      "step": 562
+    },
+    {
+      "epoch": 0.27490234375,
+      "grad_norm": 1.4679384231567383,
+      "learning_rate": 8.510277154687973e-06,
+      "loss": 0.3598,
+      "step": 563
+    },
+    {
+      "epoch": 0.275390625,
+      "grad_norm": 2.19455885887146,
+      "learning_rate": 8.504640342703687e-06,
+      "loss": 0.3371,
+      "step": 564
+    },
+    {
+      "epoch": 0.27587890625,
+      "grad_norm": 1.4917466640472412,
+      "learning_rate": 8.498994761021715e-06,
+      "loss": 0.3086,
+      "step": 565
+    },
+    {
+      "epoch": 0.2763671875,
+      "grad_norm": 2.3828556537628174,
+      "learning_rate": 8.493340423769054e-06,
+      "loss": 0.328,
+      "step": 566
+    },
+    {
+      "epoch": 0.27685546875,
+      "grad_norm": 2.0100631713867188,
+      "learning_rate": 8.487677345094606e-06,
+      "loss": 0.3497,
+      "step": 567
+    },
+    {
+      "epoch": 0.27734375,
+      "grad_norm": 2.037872552871704,
+      "learning_rate": 8.482005539169158e-06,
+      "loss": 0.3649,
+      "step": 568
+    },
+    {
+      "epoch": 0.27783203125,
+      "grad_norm": 1.3535383939743042,
+      "learning_rate": 8.476325020185326e-06,
+      "loss": 0.3321,
+      "step": 569
+    },
+    {
+      "epoch": 0.2783203125,
+      "grad_norm": 1.4872392416000366,
+      "learning_rate": 8.47063580235753e-06,
+      "loss": 0.3775,
+      "step": 570
+    },
+    {
+      "epoch": 0.27880859375,
+      "grad_norm": 2.482274293899536,
+      "learning_rate": 8.46493789992196e-06,
+      "loss": 0.3518,
+      "step": 571
+    },
+    {
+      "epoch": 0.279296875,
+      "grad_norm": 1.4444823265075684,
+      "learning_rate": 8.459231327136532e-06,
+      "loss": 0.3503,
+      "step": 572
+    },
+    {
+      "epoch": 0.27978515625,
+      "grad_norm": 1.3315978050231934,
+      "learning_rate": 8.453516098280869e-06,
+      "loss": 0.3408,
+      "step": 573
+    },
+    {
+      "epoch": 0.2802734375,
+      "grad_norm": 2.0306880474090576,
+      "learning_rate": 8.447792227656241e-06,
+      "loss": 0.3751,
+      "step": 574
+    },
+    {
+      "epoch": 0.28076171875,
+      "grad_norm": 1.3674098253250122,
+      "learning_rate": 8.442059729585552e-06,
+      "loss": 0.3307,
+      "step": 575
+    },
+    {
+      "epoch": 0.28125,
+      "grad_norm": 2.2325830459594727,
+      "learning_rate": 8.43631861841329e-06,
+      "loss": 0.3168,
+      "step": 576
+    },
+    {
+      "epoch": 0.28173828125,
+      "grad_norm": 1.956121802330017,
+      "learning_rate": 8.430568908505497e-06,
+      "loss": 0.3317,
+      "step": 577
+    },
+    {
+      "epoch": 0.2822265625,
+      "grad_norm": 2.0539493560791016,
+      "learning_rate": 8.42481061424973e-06,
+      "loss": 0.3172,
+      "step": 578
+    },
+    {
+      "epoch": 0.28271484375,
+      "grad_norm": 1.3269410133361816,
+      "learning_rate": 8.41904375005503e-06,
+      "loss": 0.3726,
+      "step": 579
+    },
+    {
+      "epoch": 0.283203125,
+      "grad_norm": 2.887756586074829,
+      "learning_rate": 8.413268330351881e-06,
+      "loss": 0.342,
+      "step": 580
+    },
+    {
+      "epoch": 0.28369140625,
+      "grad_norm": 1.640519618988037,
+      "learning_rate": 8.40748436959217e-06,
+      "loss": 0.3418,
+      "step": 581
+    },
+    {
+      "epoch": 0.2841796875,
+      "grad_norm": 2.179222583770752,
+      "learning_rate": 8.40169188224917e-06,
+      "loss": 0.368,
+      "step": 582
+    },
+    {
+      "epoch": 0.28466796875,
+      "grad_norm": 2.25158429145813,
+      "learning_rate": 8.395890882817478e-06,
+      "loss": 0.3555,
+      "step": 583
+    },
+    {
+      "epoch": 0.28515625,
+      "grad_norm": 1.5757050514221191,
+      "learning_rate": 8.390081385812993e-06,
+      "loss": 0.3453,
+      "step": 584
+    },
+    {
+      "epoch": 0.28564453125,
+      "grad_norm": 1.5802643299102783,
+      "learning_rate": 8.38426340577288e-06,
+      "loss": 0.3635,
+      "step": 585
+    },
+    {
+      "epoch": 0.2861328125,
+      "grad_norm": 1.5654072761535645,
+      "learning_rate": 8.378436957255535e-06,
+      "loss": 0.3304,
+      "step": 586
+    },
+    {
+      "epoch": 0.28662109375,
+      "grad_norm": 1.2622393369674683,
+      "learning_rate": 8.372602054840532e-06,
+      "loss": 0.3468,
+      "step": 587
+    },
+    {
+      "epoch": 0.287109375,
+      "grad_norm": 2.9419167041778564,
+      "learning_rate": 8.366758713128617e-06,
+      "loss": 0.3286,
+      "step": 588
+    },
+    {
+      "epoch": 0.28759765625,
+      "grad_norm": 1.6033565998077393,
+      "learning_rate": 8.360906946741635e-06,
+      "loss": 0.3375,
+      "step": 589
+    },
+    {
+      "epoch": 0.2880859375,
+      "grad_norm": 1.5381578207015991,
+      "learning_rate": 8.355046770322528e-06,
+      "loss": 0.3531,
+      "step": 590
+    },
+    {
+      "epoch": 0.28857421875,
+      "grad_norm": 1.7467304468154907,
+      "learning_rate": 8.349178198535273e-06,
+      "loss": 0.305,
+      "step": 591
+    },
+    {
+      "epoch": 0.2890625,
+      "grad_norm": 1.3759098052978516,
+      "learning_rate": 8.343301246064858e-06,
+      "loss": 0.3643,
+      "step": 592
+    },
+    {
+      "epoch": 0.28955078125,
+      "grad_norm": 1.3180525302886963,
+      "learning_rate": 8.337415927617243e-06,
+      "loss": 0.3468,
+      "step": 593
+    },
+    {
+      "epoch": 0.2900390625,
+      "grad_norm": 1.3249021768569946,
+      "learning_rate": 8.33152225791932e-06,
+      "loss": 0.3502,
+      "step": 594
+    },
+    {
+      "epoch": 0.29052734375,
+      "grad_norm": 1.9022133350372314,
+      "learning_rate": 8.32562025171888e-06,
+      "loss": 0.3842,
+      "step": 595
+    },
+    {
+      "epoch": 0.291015625,
+      "grad_norm": 1.4465323686599731,
+      "learning_rate": 8.319709923784573e-06,
+      "loss": 0.3247,
+      "step": 596
+    },
+    {
+      "epoch": 0.29150390625,
+      "grad_norm": 2.4993956089019775,
+      "learning_rate": 8.313791288905874e-06,
+      "loss": 0.3826,
+      "step": 597
+    },
+    {
+      "epoch": 0.2919921875,
+      "grad_norm": 1.842347264289856,
+      "learning_rate": 8.307864361893045e-06,
+      "loss": 0.329,
+      "step": 598
+    },
+    {
+      "epoch": 0.29248046875,
+      "grad_norm": 1.5460954904556274,
+      "learning_rate": 8.301929157577097e-06,
+      "loss": 0.3453,
+      "step": 599
+    },
+    {
+      "epoch": 0.29296875,
+      "grad_norm": 3.255307912826538,
+      "learning_rate": 8.295985690809752e-06,
+      "loss": 0.3358,
+      "step": 600
+    },
+    {
+      "epoch": 0.29345703125,
+      "grad_norm": 1.4224542379379272,
+      "learning_rate": 8.290033976463407e-06,
+      "loss": 0.3683,
+      "step": 601
+    },
+    {
+      "epoch": 0.2939453125,
+      "grad_norm": 1.4209293127059937,
+      "learning_rate": 8.2840740294311e-06,
+      "loss": 0.315,
+      "step": 602
+    },
+    {
+      "epoch": 0.29443359375,
+      "grad_norm": 2.0559093952178955,
+      "learning_rate": 8.278105864626467e-06,
+      "loss": 0.3801,
+      "step": 603
+    },
+    {
+      "epoch": 0.294921875,
+      "grad_norm": 1.880486249923706,
+      "learning_rate": 8.27212949698371e-06,
+      "loss": 0.3713,
+      "step": 604
+    },
+    {
+      "epoch": 0.29541015625,
+      "grad_norm": 3.0988686084747314,
+      "learning_rate": 8.266144941457552e-06,
+      "loss": 0.3917,
+      "step": 605
+    },
+    {
+      "epoch": 0.2958984375,
+      "grad_norm": 1.6043518781661987,
+      "learning_rate": 8.26015221302321e-06,
+      "loss": 0.3678,
+      "step": 606
+    },
+    {
+      "epoch": 0.29638671875,
+      "grad_norm": 1.520564079284668,
+      "learning_rate": 8.254151326676354e-06,
+      "loss": 0.3259,
+      "step": 607
+    },
+    {
+      "epoch": 0.296875,
+      "grad_norm": 1.9146232604980469,
+      "learning_rate": 8.248142297433058e-06,
+      "loss": 0.3291,
+      "step": 608
+    },
+    {
+      "epoch": 0.29736328125,
+      "grad_norm": 2.2928895950317383,
+      "learning_rate": 8.24212514032978e-06,
+      "loss": 0.3828,
+      "step": 609
+    },
+    {
+      "epoch": 0.2978515625,
+      "grad_norm": 1.9419975280761719,
+      "learning_rate": 8.236099870423314e-06,
+      "loss": 0.3287,
+      "step": 610
+    },
+    {
+      "epoch": 0.29833984375,
+      "grad_norm": 1.7183066606521606,
+      "learning_rate": 8.230066502790756e-06,
+      "loss": 0.3121,
+      "step": 611
+    },
+    {
+      "epoch": 0.298828125,
+      "grad_norm": 1.5658105611801147,
+      "learning_rate": 8.224025052529463e-06,
+      "loss": 0.3501,
+      "step": 612
+    },
+    {
+      "epoch": 0.29931640625,
+      "grad_norm": 1.9759196043014526,
+      "learning_rate": 8.21797553475702e-06,
+      "loss": 0.3345,
+      "step": 613
+    },
+    {
+      "epoch": 0.2998046875,
+      "grad_norm": 2.0763461589813232,
+      "learning_rate": 8.211917964611197e-06,
+      "loss": 0.3187,
+      "step": 614
+    },
+    {
+      "epoch": 0.30029296875,
+      "grad_norm": 1.4480257034301758,
+      "learning_rate": 8.205852357249912e-06,
+      "loss": 0.2866,
+      "step": 615
+    },
+    {
+      "epoch": 0.30078125,
+      "grad_norm": 1.9418996572494507,
+      "learning_rate": 8.1997787278512e-06,
+      "loss": 0.3125,
+      "step": 616
+    },
+    {
+      "epoch": 0.30126953125,
+      "grad_norm": 1.726302146911621,
+      "learning_rate": 8.193697091613163e-06,
+      "loss": 0.3663,
+      "step": 617
+    },
+    {
+      "epoch": 0.3017578125,
+      "grad_norm": 1.622819423675537,
+      "learning_rate": 8.187607463753946e-06,
+      "loss": 0.3385,
+      "step": 618
+    },
+    {
+      "epoch": 0.30224609375,
+      "grad_norm": 2.375453472137451,
+      "learning_rate": 8.181509859511686e-06,
+      "loss": 0.3314,
+      "step": 619
+    },
+    {
+      "epoch": 0.302734375,
+      "grad_norm": 1.6941611766815186,
+      "learning_rate": 8.175404294144482e-06,
+      "loss": 0.3152,
+      "step": 620
+    },
+    {
+      "epoch": 0.30322265625,
+      "grad_norm": 1.6905850172042847,
+      "learning_rate": 8.16929078293035e-06,
+      "loss": 0.3352,
+      "step": 621
+    },
+    {
+      "epoch": 0.3037109375,
+      "grad_norm": 1.9776393175125122,
+      "learning_rate": 8.163169341167196e-06,
+      "loss": 0.39,
+      "step": 622
+    },
+    {
+      "epoch": 0.30419921875,
+      "grad_norm": 1.4409841299057007,
+      "learning_rate": 8.157039984172764e-06,
+      "loss": 0.3445,
+      "step": 623
+    },
+    {
+      "epoch": 0.3046875,
+      "grad_norm": 1.7097798585891724,
+      "learning_rate": 8.150902727284609e-06,
+      "loss": 0.3583,
+      "step": 624
+    },
+    {
+      "epoch": 0.30517578125,
+      "grad_norm": 1.5705921649932861,
+      "learning_rate": 8.144757585860053e-06,
+      "loss": 0.355,
+      "step": 625
+    },
+    {
+      "epoch": 0.3056640625,
+      "grad_norm": 1.5804706811904907,
+      "learning_rate": 8.138604575276143e-06,
+      "loss": 0.3615,
+      "step": 626
+    },
+    {
+      "epoch": 0.30615234375,
+      "grad_norm": 1.7296881675720215,
+      "learning_rate": 8.132443710929624e-06,
+      "loss": 0.381,
+      "step": 627
+    },
+    {
+      "epoch": 0.306640625,
+      "grad_norm": 1.3139718770980835,
+      "learning_rate": 8.126275008236891e-06,
+      "loss": 0.3296,
+      "step": 628
+    },
+    {
+      "epoch": 0.30712890625,
+      "grad_norm": 1.339277744293213,
+      "learning_rate": 8.12009848263395e-06,
+      "loss": 0.3262,
+      "step": 629
+    },
+    {
+      "epoch": 0.3076171875,
+      "grad_norm": 5.439074516296387,
+      "learning_rate": 8.113914149576388e-06,
+      "loss": 0.361,
+      "step": 630
+    },
+    {
+      "epoch": 0.30810546875,
+      "grad_norm": 1.8875752687454224,
+      "learning_rate": 8.107722024539321e-06,
+      "loss": 0.3419,
+      "step": 631
+    },
+    {
+      "epoch": 0.30859375,
+      "grad_norm": 1.3780957460403442,
+      "learning_rate": 8.10152212301737e-06,
+      "loss": 0.3398,
+      "step": 632
+    },
+    {
+      "epoch": 0.30908203125,
+      "grad_norm": 2.1425485610961914,
+      "learning_rate": 8.095314460524612e-06,
+      "loss": 0.3473,
+      "step": 633
+    },
+    {
+      "epoch": 0.3095703125,
+      "grad_norm": 2.3225300312042236,
+      "learning_rate": 8.089099052594545e-06,
+      "loss": 0.3757,
+      "step": 634
+    },
+    {
+      "epoch": 0.31005859375,
+      "grad_norm": 1.4518051147460938,
+      "learning_rate": 8.08287591478005e-06,
+      "loss": 0.3112,
+      "step": 635
+    },
+    {
+      "epoch": 0.310546875,
+      "grad_norm": 2.2762012481689453,
+      "learning_rate": 8.076645062653346e-06,
+      "loss": 0.3642,
+      "step": 636
+    },
+    {
+      "epoch": 0.31103515625,
+      "grad_norm": 1.6947425603866577,
+      "learning_rate": 8.070406511805961e-06,
+      "loss": 0.35,
+      "step": 637
+    },
+    {
+      "epoch": 0.3115234375,
+      "grad_norm": 1.5694466829299927,
+      "learning_rate": 8.064160277848683e-06,
+      "loss": 0.3458,
+      "step": 638
+    },
+    {
+      "epoch": 0.31201171875,
+      "grad_norm": 1.9441496133804321,
+      "learning_rate": 8.05790637641153e-06,
+      "loss": 0.3698,
+      "step": 639
+    },
+    {
+      "epoch": 0.3125,
+      "grad_norm": 1.6394853591918945,
+      "learning_rate": 8.051644823143702e-06,
+      "loss": 0.3515,
+      "step": 640
+    },
+    {
+      "epoch": 0.31298828125,
+      "grad_norm": 1.8157254457473755,
+      "learning_rate": 8.04537563371355e-06,
+      "loss": 0.3278,
+      "step": 641
+    },
+    {
+      "epoch": 0.3134765625,
+      "grad_norm": 1.6162160634994507,
+      "learning_rate": 8.03909882380853e-06,
+      "loss": 0.3586,
+      "step": 642
+    },
+    {
+      "epoch": 0.31396484375,
+      "grad_norm": 1.7346367835998535,
+      "learning_rate": 8.03281440913517e-06,
+      "loss": 0.3194,
+      "step": 643
+    },
+    {
+      "epoch": 0.314453125,
+      "grad_norm": 1.593997836112976,
+      "learning_rate": 8.026522405419024e-06,
+      "loss": 0.3205,
+      "step": 644
+    },
+    {
+      "epoch": 0.31494140625,
+      "grad_norm": 1.3535056114196777,
+      "learning_rate": 8.020222828404638e-06,
+      "loss": 0.3382,
+      "step": 645
+    },
+    {
+      "epoch": 0.3154296875,
+      "grad_norm": 2.354459524154663,
+      "learning_rate": 8.01391569385551e-06,
+      "loss": 0.3041,
+      "step": 646
+    },
+    {
+      "epoch": 0.31591796875,
+      "grad_norm": 1.6168910264968872,
+      "learning_rate": 8.007601017554045e-06,
+      "loss": 0.392,
+      "step": 647
+    },
+    {
+      "epoch": 0.31640625,
+      "grad_norm": 1.7411466836929321,
+      "learning_rate": 8.001278815301525e-06,
+      "loss": 0.319,
+      "step": 648
+    },
+    {
+      "epoch": 0.31689453125,
+      "grad_norm": 2.3402931690216064,
+      "learning_rate": 7.994949102918062e-06,
+      "loss": 0.3657,
+      "step": 649
+    },
+    {
+      "epoch": 0.3173828125,
+      "grad_norm": 1.2933272123336792,
+      "learning_rate": 7.98861189624256e-06,
+      "loss": 0.3049,
+      "step": 650
+    },
+    {
+      "epoch": 0.31787109375,
+      "grad_norm": 1.6581286191940308,
+      "learning_rate": 7.982267211132675e-06,
+      "loss": 0.354,
+      "step": 651
+    },
+    {
+      "epoch": 0.318359375,
+      "grad_norm": 2.0283968448638916,
+      "learning_rate": 7.97591506346478e-06,
+      "loss": 0.3521,
+      "step": 652
+    },
+    {
+      "epoch": 0.31884765625,
+      "grad_norm": 1.6676313877105713,
+      "learning_rate": 7.96955546913392e-06,
+      "loss": 0.3237,
+      "step": 653
+    },
+    {
+      "epoch": 0.3193359375,
+      "grad_norm": 1.548922061920166,
+      "learning_rate": 7.963188444053772e-06,
+      "loss": 0.3145,
+      "step": 654
+    },
+    {
+      "epoch": 0.31982421875,
+      "grad_norm": 2.61688232421875,
+      "learning_rate": 7.95681400415661e-06,
+      "loss": 0.3159,
+      "step": 655
+    },
+    {
+      "epoch": 0.3203125,
+      "grad_norm": 2.0864787101745605,
+      "learning_rate": 7.95043216539326e-06,
+      "loss": 0.3394,
+      "step": 656
+    },
+    {
+      "epoch": 0.32080078125,
+      "grad_norm": 1.82245934009552,
+      "learning_rate": 7.944042943733061e-06,
+      "loss": 0.355,
+      "step": 657
+    },
+    {
+      "epoch": 0.3212890625,
+      "grad_norm": 1.6342824697494507,
+      "learning_rate": 7.937646355163833e-06,
+      "loss": 0.3407,
+      "step": 658
+    },
+    {
+      "epoch": 0.32177734375,
+      "grad_norm": 1.7688589096069336,
+      "learning_rate": 7.931242415691822e-06,
+      "loss": 0.3936,
+      "step": 659
+    },
+    {
+      "epoch": 0.322265625,
+      "grad_norm": 1.5749949216842651,
+      "learning_rate": 7.924831141341671e-06,
+      "loss": 0.3226,
+      "step": 660
+    },
+    {
+      "epoch": 0.32275390625,
+      "grad_norm": 4.079642295837402,
+      "learning_rate": 7.918412548156382e-06,
+      "loss": 0.3478,
+      "step": 661
+    },
+    {
+      "epoch": 0.3232421875,
+      "grad_norm": 1.564584732055664,
+      "learning_rate": 7.911986652197263e-06,
+      "loss": 0.345,
+      "step": 662
+    },
+    {
+      "epoch": 0.32373046875,
+      "grad_norm": 1.9359629154205322,
+      "learning_rate": 7.905553469543903e-06,
+      "loss": 0.3478,
+      "step": 663
+    },
+    {
+      "epoch": 0.32421875,
+      "grad_norm": 1.3265938758850098,
+      "learning_rate": 7.899113016294118e-06,
+      "loss": 0.3789,
+      "step": 664
+    },
+    {
+      "epoch": 0.32470703125,
+      "grad_norm": 1.617301106452942,
+      "learning_rate": 7.892665308563922e-06,
+      "loss": 0.3182,
+      "step": 665
+    },
+    {
+      "epoch": 0.3251953125,
+      "grad_norm": 2.50874924659729,
+      "learning_rate": 7.88621036248748e-06,
+      "loss": 0.3269,
+      "step": 666
+    },
+    {
+      "epoch": 0.32568359375,
+      "grad_norm": 2.0309231281280518,
+      "learning_rate": 7.879748194217074e-06,
+      "loss": 0.3294,
+      "step": 667
+    },
+    {
+      "epoch": 0.326171875,
+      "grad_norm": 1.6182068586349487,
+      "learning_rate": 7.873278819923047e-06,
+      "loss": 0.3269,
+      "step": 668
+    },
+    {
+      "epoch": 0.32666015625,
+      "grad_norm": 2.3924951553344727,
+      "learning_rate": 7.866802255793788e-06,
+      "loss": 0.3498,
+      "step": 669
+    },
+    {
+      "epoch": 0.3271484375,
+      "grad_norm": 2.816044330596924,
+      "learning_rate": 7.860318518035668e-06,
+      "loss": 0.3231,
+      "step": 670
+    },
+    {
+      "epoch": 0.32763671875,
+      "grad_norm": 1.9277939796447754,
+      "learning_rate": 7.853827622873011e-06,
+      "loss": 0.3236,
+      "step": 671
+    },
+    {
+      "epoch": 0.328125,
+      "grad_norm": 1.364225149154663,
+      "learning_rate": 7.847329586548049e-06,
+      "loss": 0.3807,
+      "step": 672
+    },
+    {
+      "epoch": 0.32861328125,
+      "grad_norm": 1.443907380104065,
+      "learning_rate": 7.840824425320888e-06,
+      "loss": 0.4092,
+      "step": 673
+    },
+    {
+      "epoch": 0.3291015625,
+      "grad_norm": 1.670778512954712,
+      "learning_rate": 7.834312155469457e-06,
+      "loss": 0.3653,
+      "step": 674
+    },
+    {
+      "epoch": 0.32958984375,
+      "grad_norm": 1.510043740272522,
+      "learning_rate": 7.827792793289477e-06,
+      "loss": 0.3463,
+      "step": 675
+    },
+    {
+      "epoch": 0.330078125,
+      "grad_norm": 2.1872780323028564,
+      "learning_rate": 7.821266355094419e-06,
+      "loss": 0.3479,
+      "step": 676
+    },
+    {
+      "epoch": 0.33056640625,
+      "grad_norm": 1.6790423393249512,
+      "learning_rate": 7.814732857215453e-06,
+      "loss": 0.3476,
+      "step": 677
+    },
+    {
+      "epoch": 0.3310546875,
+      "grad_norm": 1.3476860523223877,
+      "learning_rate": 7.808192316001417e-06,
+      "loss": 0.3333,
+      "step": 678
+    },
+    {
+      "epoch": 0.33154296875,
+      "grad_norm": 1.752164602279663,
+      "learning_rate": 7.801644747818777e-06,
+      "loss": 0.3341,
+      "step": 679
+    },
+    {
+      "epoch": 0.33203125,
+      "grad_norm": 2.4022326469421387,
+      "learning_rate": 7.79509016905158e-06,
+      "loss": 0.357,
+      "step": 680
+    },
+    {
+      "epoch": 0.33251953125,
+      "grad_norm": 1.3659697771072388,
+      "learning_rate": 7.788528596101419e-06,
+      "loss": 0.3073,
+      "step": 681
+    },
+    {
+      "epoch": 0.3330078125,
+      "grad_norm": 1.4519615173339844,
+      "learning_rate": 7.78196004538738e-06,
+      "loss": 0.3052,
+      "step": 682
+    },
+    {
+      "epoch": 0.33349609375,
+      "grad_norm": 2.08927583694458,
+      "learning_rate": 7.775384533346018e-06,
+      "loss": 0.3242,
+      "step": 683
+    },
+    {
+      "epoch": 0.333984375,
+      "grad_norm": 1.4538501501083374,
+      "learning_rate": 7.768802076431304e-06,
+      "loss": 0.3495,
+      "step": 684
+    },
+    {
+      "epoch": 0.33447265625,
+      "grad_norm": 2.239643096923828,
+      "learning_rate": 7.76221269111459e-06,
+      "loss": 0.3554,
+      "step": 685
+    },
+    {
+      "epoch": 0.3349609375,
+      "grad_norm": 1.8009265661239624,
+      "learning_rate": 7.755616393884562e-06,
+      "loss": 0.3652,
+      "step": 686
+    },
+    {
+      "epoch": 0.33544921875,
+      "grad_norm": 1.5794439315795898,
+      "learning_rate": 7.7490132012472e-06,
+      "loss": 0.3321,
+      "step": 687
+    },
+    {
+      "epoch": 0.3359375,
+      "grad_norm": 1.737437129020691,
+      "learning_rate": 7.742403129725742e-06,
+      "loss": 0.3138,
+      "step": 688
+    },
+    {
+      "epoch": 0.33642578125,
+      "grad_norm": 1.7152299880981445,
+      "learning_rate": 7.735786195860641e-06,
+      "loss": 0.3582,
+      "step": 689
+    },
+    {
+      "epoch": 0.3369140625,
+      "grad_norm": 1.3847858905792236,
+      "learning_rate": 7.729162416209518e-06,
+      "loss": 0.3396,
+      "step": 690
+    },
+    {
+      "epoch": 0.33740234375,
+      "grad_norm": 1.6747031211853027,
+      "learning_rate": 7.722531807347122e-06,
+      "loss": 0.3474,
+      "step": 691
+    },
+    {
+      "epoch": 0.337890625,
+      "grad_norm": 1.3016866445541382,
+      "learning_rate": 7.715894385865299e-06,
+      "loss": 0.3391,
+      "step": 692
+    },
+    {
+      "epoch": 0.33837890625,
+      "grad_norm": 1.3648223876953125,
+      "learning_rate": 7.709250168372932e-06,
+      "loss": 0.3298,
+      "step": 693
+    },
+    {
+      "epoch": 0.3388671875,
+      "grad_norm": 1.5124351978302002,
+      "learning_rate": 7.702599171495919e-06,
+      "loss": 0.3334,
+      "step": 694
+    },
+    {
+      "epoch": 0.33935546875,
+      "grad_norm": 37.46984100341797,
+      "learning_rate": 7.695941411877115e-06,
+      "loss": 0.3342,
+      "step": 695
+    },
+    {
+      "epoch": 0.33984375,
+      "grad_norm": 1.4970625638961792,
+      "learning_rate": 7.689276906176302e-06,
+      "loss": 0.3436,
+      "step": 696
+    },
+    {
+      "epoch": 0.34033203125,
+      "grad_norm": 3.098925828933716,
+      "learning_rate": 7.682605671070142e-06,
+      "loss": 0.3437,
+      "step": 697
+    },
+    {
+      "epoch": 0.3408203125,
+      "grad_norm": 1.7555867433547974,
+      "learning_rate": 7.675927723252134e-06,
+      "loss": 0.322,
+      "step": 698
+    },
+    {
+      "epoch": 0.34130859375,
+      "grad_norm": 1.5935651063919067,
+      "learning_rate": 7.669243079432578e-06,
+      "loss": 0.2998,
+      "step": 699
+    },
+    {
+      "epoch": 0.341796875,
+      "grad_norm": 1.506208896636963,
+      "learning_rate": 7.662551756338525e-06,
+      "loss": 0.3612,
+      "step": 700
+    },
+    {
+      "epoch": 0.34228515625,
+      "grad_norm": 1.923596978187561,
+      "learning_rate": 7.655853770713744e-06,
+      "loss": 0.3593,
+      "step": 701
+    },
+    {
+      "epoch": 0.3427734375,
+      "grad_norm": 1.9344090223312378,
+      "learning_rate": 7.64914913931867e-06,
+      "loss": 0.3156,
+      "step": 702
+    },
+    {
+      "epoch": 0.34326171875,
+      "grad_norm": 1.7808047533035278,
+      "learning_rate": 7.642437878930376e-06,
+      "loss": 0.3419,
+      "step": 703
+    },
+    {
+      "epoch": 0.34375,
+      "grad_norm": 1.5053675174713135,
+      "learning_rate": 7.635720006342513e-06,
+      "loss": 0.3539,
+      "step": 704
+    },
+    {
+      "epoch": 0.34423828125,
+      "grad_norm": 1.5963175296783447,
+      "learning_rate": 7.628995538365287e-06,
+      "loss": 0.3562,
+      "step": 705
+    },
+    {
+      "epoch": 0.3447265625,
+      "grad_norm": 1.4388726949691772,
+      "learning_rate": 7.6222644918254005e-06,
+      "loss": 0.3413,
+      "step": 706
+    },
+    {
+      "epoch": 0.34521484375,
+      "grad_norm": 3.6217451095581055,
+      "learning_rate": 7.615526883566023e-06,
+      "loss": 0.3584,
+      "step": 707
+    },
+    {
+      "epoch": 0.345703125,
+      "grad_norm": 1.6617943048477173,
+      "learning_rate": 7.608782730446741e-06,
+      "loss": 0.3675,
+      "step": 708
+    },
+    {
+      "epoch": 0.34619140625,
+      "grad_norm": 3.6505870819091797,
+      "learning_rate": 7.6020320493435175e-06,
+      "loss": 0.3028,
+      "step": 709
+    },
+    {
+      "epoch": 0.3466796875,
+      "grad_norm": 1.5057923793792725,
+      "learning_rate": 7.595274857148651e-06,
+      "loss": 0.3601,
+      "step": 710
+    },
+    {
+      "epoch": 0.34716796875,
+      "grad_norm": 1.775791049003601,
+      "learning_rate": 7.588511170770736e-06,
+      "loss": 0.3561,
+      "step": 711
+    },
+    {
+      "epoch": 0.34765625,
+      "grad_norm": 2.0912845134735107,
+      "learning_rate": 7.581741007134611e-06,
+      "loss": 0.3211,
+      "step": 712
+    },
+    {
+      "epoch": 0.34814453125,
+      "grad_norm": 1.4719021320343018,
+      "learning_rate": 7.574964383181329e-06,
+      "loss": 0.3571,
+      "step": 713
+    },
+    {
+      "epoch": 0.3486328125,
+      "grad_norm": 1.5099034309387207,
+      "learning_rate": 7.568181315868104e-06,
+      "loss": 0.3773,
+      "step": 714
+    },
+    {
+      "epoch": 0.34912109375,
+      "grad_norm": 1.797803282737732,
+      "learning_rate": 7.561391822168277e-06,
+      "loss": 0.3305,
+      "step": 715
+    },
+    {
+      "epoch": 0.349609375,
+      "grad_norm": 1.5316636562347412,
+      "learning_rate": 7.554595919071268e-06,
+      "loss": 0.3692,
+      "step": 716
+    },
+    {
+      "epoch": 0.35009765625,
+      "grad_norm": 1.332055926322937,
+      "learning_rate": 7.5477936235825344e-06,
+      "loss": 0.2998,
+      "step": 717
+    },
+    {
+      "epoch": 0.3505859375,
+      "grad_norm": 1.538785457611084,
+      "learning_rate": 7.540984952723531e-06,
+      "loss": 0.3325,
+      "step": 718
+    },
+    {
+      "epoch": 0.35107421875,
+      "grad_norm": 2.884404420852661,
+      "learning_rate": 7.534169923531665e-06,
+      "loss": 0.3036,
+      "step": 719
+    },
+    {
+      "epoch": 0.3515625,
+      "grad_norm": 1.7468745708465576,
+      "learning_rate": 7.527348553060254e-06,
+      "loss": 0.3199,
+      "step": 720
+    },
+    {
+      "epoch": 0.35205078125,
+      "grad_norm": 2.015227794647217,
+      "learning_rate": 7.520520858378486e-06,
+      "loss": 0.3884,
+      "step": 721
+    },
+    {
+      "epoch": 0.3525390625,
+      "grad_norm": 1.3880223035812378,
+      "learning_rate": 7.513686856571367e-06,
+      "loss": 0.336,
+      "step": 722
+    },
+    {
+      "epoch": 0.35302734375,
+      "grad_norm": 1.297411561012268,
+      "learning_rate": 7.506846564739694e-06,
+      "loss": 0.3306,
+      "step": 723
+    },
+    {
+      "epoch": 0.353515625,
+      "grad_norm": 1.55870521068573,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.3056,
+      "step": 724
+    },
+    {
+      "epoch": 0.35400390625,
+      "grad_norm": 2.036909818649292,
+      "learning_rate": 7.493147179484514e-06,
+      "loss": 0.3273,
+      "step": 725
+    },
+    {
+      "epoch": 0.3544921875,
+      "grad_norm": 1.3678783178329468,
+      "learning_rate": 7.486288120341118e-06,
+      "loss": 0.345,
+      "step": 726
+    },
+    {
+      "epoch": 0.35498046875,
+      "grad_norm": 2.0894579887390137,
+      "learning_rate": 7.479422839733307e-06,
+      "loss": 0.359,
+      "step": 727
+    },
+    {
+      "epoch": 0.35546875,
+      "grad_norm": 1.6823246479034424,
+      "learning_rate": 7.4725513548401455e-06,
+      "loss": 0.3563,
+      "step": 728
+    },
+    {
+      "epoch": 0.35595703125,
+      "grad_norm": 1.351969838142395,
+      "learning_rate": 7.4656736828562186e-06,
+      "loss": 0.3017,
+      "step": 729
+    },
+    {
+      "epoch": 0.3564453125,
+      "grad_norm": 1.6686972379684448,
+      "learning_rate": 7.458789840991596e-06,
+      "loss": 0.3478,
+      "step": 730
+    },
+    {
+      "epoch": 0.35693359375,
+      "grad_norm": 1.3534908294677734,
+      "learning_rate": 7.4518998464717874e-06,
+      "loss": 0.3244,
+      "step": 731
+    },
+    {
+      "epoch": 0.357421875,
+      "grad_norm": 1.4082777500152588,
+      "learning_rate": 7.445003716537698e-06,
+      "loss": 0.3251,
+      "step": 732
+    },
+    {
+      "epoch": 0.35791015625,
+      "grad_norm": 2.0288498401641846,
+      "learning_rate": 7.438101468445582e-06,
+      "loss": 0.3379,
+      "step": 733
+    },
+    {
+      "epoch": 0.3583984375,
+      "grad_norm": 1.6891510486602783,
+      "learning_rate": 7.4311931194670085e-06,
+      "loss": 0.3576,
+      "step": 734
+    },
+    {
+      "epoch": 0.35888671875,
+      "grad_norm": 1.3616983890533447,
+      "learning_rate": 7.42427868688881e-06,
+      "loss": 0.3439,
+      "step": 735
+    },
+    {
+      "epoch": 0.359375,
+      "grad_norm": 1.5869650840759277,
+      "learning_rate": 7.417358188013042e-06,
+      "loss": 0.3389,
+      "step": 736
+    },
+    {
+      "epoch": 0.35986328125,
+      "grad_norm": 1.3705356121063232,
+      "learning_rate": 7.410431640156937e-06,
+      "loss": 0.346,
+      "step": 737
+    },
+    {
+      "epoch": 0.3603515625,
+      "grad_norm": 2.2622792720794678,
+      "learning_rate": 7.403499060652874e-06,
+      "loss": 0.3535,
+      "step": 738
+    },
+    {
+      "epoch": 0.36083984375,
+      "grad_norm": 1.719897747039795,
+      "learning_rate": 7.3965604668483145e-06,
+      "loss": 0.382,
+      "step": 739
+    },
+    {
+      "epoch": 0.361328125,
+      "grad_norm": 1.3844950199127197,
+      "learning_rate": 7.389615876105773e-06,
+      "loss": 0.3481,
+      "step": 740
+    },
+    {
+      "epoch": 0.36181640625,
+      "grad_norm": 1.6294703483581543,
+      "learning_rate": 7.38266530580277e-06,
+      "loss": 0.3656,
+      "step": 741
+    },
+    {
+      "epoch": 0.3623046875,
+      "grad_norm": 2.908967971801758,
+      "learning_rate": 7.375708773331791e-06,
+      "loss": 0.3457,
+      "step": 742
+    },
+    {
+      "epoch": 0.36279296875,
+      "grad_norm": 1.473132848739624,
+      "learning_rate": 7.36874629610024e-06,
+      "loss": 0.3385,
+      "step": 743
+    },
+    {
+      "epoch": 0.36328125,
+      "grad_norm": 2.919328451156616,
+      "learning_rate": 7.361777891530392e-06,
+      "loss": 0.3336,
+      "step": 744
+    },
+    {
+      "epoch": 0.36376953125,
+      "grad_norm": 2.563336133956909,
+      "learning_rate": 7.354803577059359e-06,
+      "loss": 0.3357,
+      "step": 745
+    },
+    {
+      "epoch": 0.3642578125,
+      "grad_norm": 1.4097625017166138,
+      "learning_rate": 7.347823370139042e-06,
+      "loss": 0.3559,
+      "step": 746
+    },
+    {
+      "epoch": 0.36474609375,
+      "grad_norm": 1.3321950435638428,
+      "learning_rate": 7.340837288236085e-06,
+      "loss": 0.3626,
+      "step": 747
+    },
+    {
+      "epoch": 0.365234375,
+      "grad_norm": 1.6507295370101929,
+      "learning_rate": 7.3338453488318284e-06,
+      "loss": 0.3095,
+      "step": 748
+    },
+    {
+      "epoch": 0.36572265625,
+      "grad_norm": 1.8008859157562256,
+      "learning_rate": 7.326847569422278e-06,
+      "loss": 0.3193,
+      "step": 749
+    },
+    {
+      "epoch": 0.3662109375,
+      "grad_norm": 1.4755789041519165,
+      "learning_rate": 7.3198439675180484e-06,
+      "loss": 0.2986,
+      "step": 750
+    },
+    {
+      "epoch": 0.36669921875,
+      "grad_norm": 1.7474323511123657,
+      "learning_rate": 7.312834560644327e-06,
+      "loss": 0.3936,
+      "step": 751
+    },
+    {
+      "epoch": 0.3671875,
+      "grad_norm": 1.6639896631240845,
+      "learning_rate": 7.30581936634082e-06,
+      "loss": 0.3673,
+      "step": 752
+    },
+    {
+      "epoch": 0.36767578125,
+      "grad_norm": 1.3790712356567383,
+      "learning_rate": 7.298798402161725e-06,
+      "loss": 0.3639,
+      "step": 753
+    },
+    {
+      "epoch": 0.3681640625,
+      "grad_norm": 1.9777040481567383,
+      "learning_rate": 7.291771685675673e-06,
+      "loss": 0.3299,
+      "step": 754
+    },
+    {
+      "epoch": 0.36865234375,
+      "grad_norm": 1.7995957136154175,
+      "learning_rate": 7.284739234465686e-06,
+      "loss": 0.3605,
+      "step": 755
+    },
+    {
+      "epoch": 0.369140625,
+      "grad_norm": 1.9671039581298828,
+      "learning_rate": 7.277701066129141e-06,
+      "loss": 0.3792,
+      "step": 756
+    },
+    {
+      "epoch": 0.36962890625,
+      "grad_norm": 2.719590187072754,
+      "learning_rate": 7.27065719827772e-06,
+      "loss": 0.3318,
+      "step": 757
+    },
+    {
+      "epoch": 0.3701171875,
+      "grad_norm": 1.9835278987884521,
+      "learning_rate": 7.2636076485373645e-06,
+      "loss": 0.3286,
+      "step": 758
+    },
+    {
+      "epoch": 0.37060546875,
+      "grad_norm": 1.2610225677490234,
+      "learning_rate": 7.256552434548236e-06,
+      "loss": 0.3274,
+      "step": 759
+    },
+    {
+      "epoch": 0.37109375,
+      "grad_norm": 1.2788983583450317,
+      "learning_rate": 7.249491573964671e-06,
+      "loss": 0.3622,
+      "step": 760
+    },
+    {
+      "epoch": 0.37158203125,
+      "grad_norm": 1.2974728345870972,
+      "learning_rate": 7.242425084455132e-06,
+      "loss": 0.3253,
+      "step": 761
+    },
+    {
+      "epoch": 0.3720703125,
+      "grad_norm": 1.8051031827926636,
+      "learning_rate": 7.23535298370217e-06,
+      "loss": 0.3486,
+      "step": 762
+    },
+    {
+      "epoch": 0.37255859375,
+      "grad_norm": 1.7785935401916504,
+      "learning_rate": 7.228275289402373e-06,
+      "loss": 0.3195,
+      "step": 763
+    },
+    {
+      "epoch": 0.373046875,
+      "grad_norm": 1.2360249757766724,
+      "learning_rate": 7.221192019266332e-06,
+      "loss": 0.3005,
+      "step": 764
+    },
+    {
+      "epoch": 0.37353515625,
+      "grad_norm": 1.5772784948349,
+      "learning_rate": 7.214103191018584e-06,
+      "loss": 0.3319,
+      "step": 765
+    },
+    {
+      "epoch": 0.3740234375,
+      "grad_norm": 1.5777393579483032,
+      "learning_rate": 7.2070088223975784e-06,
+      "loss": 0.3412,
+      "step": 766
+    },
+    {
+      "epoch": 0.37451171875,
+      "grad_norm": 1.2442673444747925,
+      "learning_rate": 7.199908931155628e-06,
+      "loss": 0.3236,
+      "step": 767
+    },
+    {
+      "epoch": 0.375,
+      "grad_norm": 1.1323033571243286,
+      "learning_rate": 7.192803535058861e-06,
+      "loss": 0.3236,
+      "step": 768
+    },
+    {
+      "epoch": 0.37548828125,
+      "grad_norm": 1.316483974456787,
+      "learning_rate": 7.185692651887186e-06,
+      "loss": 0.3295,
+      "step": 769
+    },
+    {
+      "epoch": 0.3759765625,
+      "grad_norm": 1.5371990203857422,
+      "learning_rate": 7.178576299434239e-06,
+      "loss": 0.3711,
+      "step": 770
+    },
+    {
+      "epoch": 0.37646484375,
+      "grad_norm": 1.7177865505218506,
+      "learning_rate": 7.171454495507341e-06,
+      "loss": 0.3294,
+      "step": 771
+    },
+    {
+      "epoch": 0.376953125,
+      "grad_norm": 1.4074996709823608,
+      "learning_rate": 7.164327257927456e-06,
+      "loss": 0.3472,
+      "step": 772
+    },
+    {
+      "epoch": 0.37744140625,
+      "grad_norm": 1.3459590673446655,
+      "learning_rate": 7.157194604529143e-06,
+      "loss": 0.3268,
+      "step": 773
+    },
+    {
+      "epoch": 0.3779296875,
+      "grad_norm": 1.3509142398834229,
+      "learning_rate": 7.150056553160517e-06,
+      "loss": 0.3258,
+      "step": 774
+    },
+    {
+      "epoch": 0.37841796875,
+      "grad_norm": 1.3562768697738647,
+      "learning_rate": 7.142913121683195e-06,
+      "loss": 0.3301,
+      "step": 775
+    },
+    {
+      "epoch": 0.37890625,
+      "grad_norm": 1.815333604812622,
+      "learning_rate": 7.135764327972261e-06,
+      "loss": 0.3653,
+      "step": 776
+    },
+    {
+      "epoch": 0.37939453125,
+      "grad_norm": 1.3162930011749268,
+      "learning_rate": 7.128610189916213e-06,
+      "loss": 0.376,
+      "step": 777
+    },
+    {
+      "epoch": 0.3798828125,
+      "grad_norm": 1.7800266742706299,
+      "learning_rate": 7.121450725416928e-06,
+      "loss": 0.3662,
+      "step": 778
+    },
+    {
+      "epoch": 0.38037109375,
+      "grad_norm": 1.5096458196640015,
+      "learning_rate": 7.114285952389604e-06,
+      "loss": 0.3588,
+      "step": 779
+    },
+    {
+      "epoch": 0.380859375,
+      "grad_norm": 2.538273334503174,
+      "learning_rate": 7.1071158887627304e-06,
+      "loss": 0.3312,
+      "step": 780
+    },
+    {
+      "epoch": 0.38134765625,
+      "grad_norm": 1.3077067136764526,
+      "learning_rate": 7.0999405524780266e-06,
+      "loss": 0.3344,
+      "step": 781
+    },
+    {
+      "epoch": 0.3818359375,
+      "grad_norm": 1.3059022426605225,
+      "learning_rate": 7.092759961490415e-06,
+      "loss": 0.3259,
+      "step": 782
+    },
+    {
+      "epoch": 0.38232421875,
+      "grad_norm": 2.276553153991699,
+      "learning_rate": 7.08557413376796e-06,
+      "loss": 0.3331,
+      "step": 783
+    },
+    {
+      "epoch": 0.3828125,
+      "grad_norm": 1.3777782917022705,
+      "learning_rate": 7.078383087291833e-06,
+      "loss": 0.3211,
+      "step": 784
+    },
+    {
+      "epoch": 0.38330078125,
+      "grad_norm": 1.3232738971710205,
+      "learning_rate": 7.071186840056264e-06,
+      "loss": 0.2928,
+      "step": 785
+    },
+    {
+      "epoch": 0.3837890625,
+      "grad_norm": 1.1360565423965454,
+      "learning_rate": 7.063985410068499e-06,
+      "loss": 0.3291,
+      "step": 786
+    },
+    {
+      "epoch": 0.38427734375,
+      "grad_norm": 1.5104074478149414,
+      "learning_rate": 7.056778815348746e-06,
+      "loss": 0.3388,
+      "step": 787
+    },
+    {
+      "epoch": 0.384765625,
+      "grad_norm": 1.3837941884994507,
+      "learning_rate": 7.0495670739301435e-06,
+      "loss": 0.3802,
+      "step": 788
+    },
+    {
+      "epoch": 0.38525390625,
+      "grad_norm": 2.0784964561462402,
+      "learning_rate": 7.042350203858706e-06,
+      "loss": 0.3153,
+      "step": 789
+    },
+    {
+      "epoch": 0.3857421875,
+      "grad_norm": 1.4472565650939941,
+      "learning_rate": 7.035128223193286e-06,
+      "loss": 0.3145,
+      "step": 790
+    },
+    {
+      "epoch": 0.38623046875,
+      "grad_norm": 1.729691505432129,
+      "learning_rate": 7.0279011500055136e-06,
+      "loss": 0.393,
+      "step": 791
+    },
+    {
+      "epoch": 0.38671875,
+      "grad_norm": 1.4967801570892334,
+      "learning_rate": 7.020669002379772e-06,
+      "loss": 0.3344,
+      "step": 792
+    },
+    {
+      "epoch": 0.38720703125,
+      "grad_norm": 1.322029948234558,
+      "learning_rate": 7.0134317984131395e-06,
+      "loss": 0.3319,
+      "step": 793
+    },
+    {
+      "epoch": 0.3876953125,
+      "grad_norm": 2.8917009830474854,
+      "learning_rate": 7.006189556215346e-06,
+      "loss": 0.3152,
+      "step": 794
+    },
+    {
+      "epoch": 0.38818359375,
+      "grad_norm": 1.581947922706604,
+      "learning_rate": 6.998942293908725e-06,
+      "loss": 0.3606,
+      "step": 795
+    },
+    {
+      "epoch": 0.388671875,
+      "grad_norm": 2.658916711807251,
+      "learning_rate": 6.991690029628181e-06,
+      "loss": 0.3451,
+      "step": 796
+    },
+    {
+      "epoch": 0.38916015625,
+      "grad_norm": 2.3201754093170166,
+      "learning_rate": 6.9844327815211275e-06,
+      "loss": 0.333,
+      "step": 797
+    },
+    {
+      "epoch": 0.3896484375,
+      "grad_norm": 1.4934650659561157,
+      "learning_rate": 6.977170567747452e-06,
+      "loss": 0.3336,
+      "step": 798
+    },
+    {
+      "epoch": 0.39013671875,
+      "grad_norm": 1.4863629341125488,
+      "learning_rate": 6.969903406479465e-06,
+      "loss": 0.3347,
+      "step": 799
+    },
+    {
+      "epoch": 0.390625,
+      "grad_norm": 1.3552590608596802,
+      "learning_rate": 6.962631315901861e-06,
+      "loss": 0.3623,
+      "step": 800
+    },
+    {
+      "epoch": 0.39111328125,
+      "grad_norm": 2.2949376106262207,
+      "learning_rate": 6.955354314211669e-06,
+      "loss": 0.2987,
+      "step": 801
+    },
+    {
+      "epoch": 0.3916015625,
+      "grad_norm": 1.3013123273849487,
+      "learning_rate": 6.948072419618201e-06,
+      "loss": 0.3307,
+      "step": 802
+    },
+    {
+      "epoch": 0.39208984375,
+      "grad_norm": 1.4084373712539673,
+      "learning_rate": 6.940785650343019e-06,
+      "loss": 0.3119,
+      "step": 803
+    },
+    {
+      "epoch": 0.392578125,
+      "grad_norm": 2.596653461456299,
+      "learning_rate": 6.93349402461988e-06,
+      "loss": 0.3228,
+      "step": 804
+    },
+    {
+      "epoch": 0.39306640625,
+      "grad_norm": 1.5036858320236206,
+      "learning_rate": 6.926197560694699e-06,
+      "loss": 0.3463,
+      "step": 805
+    },
+    {
+      "epoch": 0.3935546875,
+      "grad_norm": 1.8642725944519043,
+      "learning_rate": 6.918896276825485e-06,
+      "loss": 0.368,
+      "step": 806
+    },
+    {
+      "epoch": 0.39404296875,
+      "grad_norm": 1.289711356163025,
+      "learning_rate": 6.9115901912823226e-06,
+      "loss": 0.3582,
+      "step": 807
+    },
+    {
+      "epoch": 0.39453125,
+      "grad_norm": 1.507915735244751,
+      "learning_rate": 6.9042793223473024e-06,
+      "loss": 0.3829,
+      "step": 808
+    },
+    {
+      "epoch": 0.39501953125,
+      "grad_norm": 1.7021656036376953,
+      "learning_rate": 6.896963688314489e-06,
+      "loss": 0.3668,
+      "step": 809
+    },
+    {
+      "epoch": 0.3955078125,
+      "grad_norm": 1.2955149412155151,
+      "learning_rate": 6.889643307489865e-06,
+      "loss": 0.3344,
+      "step": 810
+    },
+    {
+      "epoch": 0.39599609375,
+      "grad_norm": 1.183563232421875,
+      "learning_rate": 6.882318198191298e-06,
+      "loss": 0.3191,
+      "step": 811
+    },
+    {
+      "epoch": 0.396484375,
+      "grad_norm": 1.458882451057434,
+      "learning_rate": 6.874988378748484e-06,
+      "loss": 0.3531,
+      "step": 812
+    },
+    {
+      "epoch": 0.39697265625,
+      "grad_norm": 1.6540387868881226,
+      "learning_rate": 6.8676538675029054e-06,
+      "loss": 0.3399,
+      "step": 813
+    },
+    {
+      "epoch": 0.3974609375,
+      "grad_norm": 1.2130305767059326,
+      "learning_rate": 6.860314682807786e-06,
+      "loss": 0.3387,
+      "step": 814
+    },
+    {
+      "epoch": 0.39794921875,
+      "grad_norm": 1.3185558319091797,
+      "learning_rate": 6.852970843028043e-06,
+      "loss": 0.3389,
+      "step": 815
+    },
+    {
+      "epoch": 0.3984375,
+      "grad_norm": 1.6620187759399414,
+      "learning_rate": 6.845622366540242e-06,
+      "loss": 0.3041,
+      "step": 816
+    },
+    {
+      "epoch": 0.39892578125,
+      "grad_norm": 1.1920667886734009,
+      "learning_rate": 6.8382692717325525e-06,
+      "loss": 0.3047,
+      "step": 817
+    },
+    {
+      "epoch": 0.3994140625,
+      "grad_norm": 1.4352617263793945,
+      "learning_rate": 6.8309115770046986e-06,
+      "loss": 0.3276,
+      "step": 818
+    },
+    {
+      "epoch": 0.39990234375,
+      "grad_norm": 1.6452810764312744,
+      "learning_rate": 6.8235493007679155e-06,
+      "loss": 0.3243,
+      "step": 819
+    },
+    {
+      "epoch": 0.400390625,
+      "grad_norm": 1.6612956523895264,
+      "learning_rate": 6.816182461444905e-06,
+      "loss": 0.342,
+      "step": 820
+    },
+    {
+      "epoch": 0.40087890625,
+      "grad_norm": 1.2954360246658325,
+      "learning_rate": 6.8088110774697825e-06,
+      "loss": 0.3117,
+      "step": 821
+    },
+    {
+      "epoch": 0.4013671875,
+      "grad_norm": 2.189624786376953,
+      "learning_rate": 6.8014351672880395e-06,
+      "loss": 0.3069,
+      "step": 822
+    },
+    {
+      "epoch": 0.40185546875,
+      "grad_norm": 1.4809291362762451,
+      "learning_rate": 6.794054749356492e-06,
+      "loss": 0.3355,
+      "step": 823
+    },
+    {
+      "epoch": 0.40234375,
+      "grad_norm": 1.6851189136505127,
+      "learning_rate": 6.786669842143236e-06,
+      "loss": 0.3435,
+      "step": 824
+    },
+    {
+      "epoch": 0.40283203125,
+      "grad_norm": 1.401813268661499,
+      "learning_rate": 6.779280464127601e-06,
+      "loss": 0.326,
+      "step": 825
+    },
+    {
+      "epoch": 0.4033203125,
+      "grad_norm": 1.7311843633651733,
+      "learning_rate": 6.771886633800104e-06,
+      "loss": 0.3281,
+      "step": 826
+    },
+    {
+      "epoch": 0.40380859375,
+      "grad_norm": 2.936901092529297,
+      "learning_rate": 6.764488369662403e-06,
+      "loss": 0.3727,
+      "step": 827
+    },
+    {
+      "epoch": 0.404296875,
+      "grad_norm": 1.319385051727295,
+      "learning_rate": 6.75708569022725e-06,
+      "loss": 0.344,
+      "step": 828
+    },
+    {
+      "epoch": 0.40478515625,
+      "grad_norm": 1.9358359575271606,
+      "learning_rate": 6.749678614018446e-06,
+      "loss": 0.3622,
+      "step": 829
+    },
+    {
+      "epoch": 0.4052734375,
+      "grad_norm": 1.1188249588012695,
+      "learning_rate": 6.742267159570796e-06,
+      "loss": 0.3299,
+      "step": 830
+    },
+    {
+      "epoch": 0.40576171875,
+      "grad_norm": 1.3562527894973755,
+      "learning_rate": 6.734851345430057e-06,
+      "loss": 0.319,
+      "step": 831
+    },
+    {
+      "epoch": 0.40625,
+      "grad_norm": 1.2941495180130005,
+      "learning_rate": 6.727431190152898e-06,
+      "loss": 0.3323,
+      "step": 832
+    },
+    {
+      "epoch": 0.40673828125,
+      "grad_norm": 2.1621103286743164,
+      "learning_rate": 6.720006712306849e-06,
+      "loss": 0.3409,
+      "step": 833
+    },
+    {
+      "epoch": 0.4072265625,
+      "grad_norm": 1.3561265468597412,
+      "learning_rate": 6.712577930470258e-06,
+      "loss": 0.3549,
+      "step": 834
+    },
+    {
+      "epoch": 0.40771484375,
+      "grad_norm": 1.2518807649612427,
+      "learning_rate": 6.705144863232246e-06,
+      "loss": 0.3279,
+      "step": 835
+    },
+    {
+      "epoch": 0.408203125,
+      "grad_norm": 1.1951934099197388,
+      "learning_rate": 6.697707529192648e-06,
+      "loss": 0.3146,
+      "step": 836
+    },
+    {
+      "epoch": 0.40869140625,
+      "grad_norm": 1.2976142168045044,
+      "learning_rate": 6.6902659469619855e-06,
+      "loss": 0.3151,
+      "step": 837
+    },
+    {
+      "epoch": 0.4091796875,
+      "grad_norm": 1.554851770401001,
+      "learning_rate": 6.682820135161405e-06,
+      "loss": 0.2972,
+      "step": 838
+    },
+    {
+      "epoch": 0.40966796875,
+      "grad_norm": 1.467674732208252,
+      "learning_rate": 6.675370112422639e-06,
+      "loss": 0.3538,
+      "step": 839
+    },
+    {
+      "epoch": 0.41015625,
+      "grad_norm": 2.0394184589385986,
+      "learning_rate": 6.667915897387957e-06,
+      "loss": 0.3124,
+      "step": 840
+    },
+    {
+      "epoch": 0.41064453125,
+      "grad_norm": 1.458815097808838,
+      "learning_rate": 6.6604575087101165e-06,
+      "loss": 0.3073,
+      "step": 841
+    },
+    {
+      "epoch": 0.4111328125,
+      "grad_norm": 1.2343790531158447,
+      "learning_rate": 6.6529949650523195e-06,
+      "loss": 0.3224,
+      "step": 842
+    },
+    {
+      "epoch": 0.41162109375,
+      "grad_norm": 1.307780385017395,
+      "learning_rate": 6.645528285088169e-06,
+      "loss": 0.3139,
+      "step": 843
+    },
+    {
+      "epoch": 0.412109375,
+      "grad_norm": 1.187071681022644,
+      "learning_rate": 6.638057487501613e-06,
+      "loss": 0.3316,
+      "step": 844
+    },
+    {
+      "epoch": 0.41259765625,
+      "grad_norm": 1.9509886503219604,
+      "learning_rate": 6.630582590986907e-06,
+      "loss": 0.3381,
+      "step": 845
+    },
+    {
+      "epoch": 0.4130859375,
+      "grad_norm": 1.5562846660614014,
+      "learning_rate": 6.623103614248561e-06,
+      "loss": 0.3648,
+      "step": 846
+    },
+    {
+      "epoch": 0.41357421875,
+      "grad_norm": 1.423948049545288,
+      "learning_rate": 6.615620576001293e-06,
+      "loss": 0.3163,
+      "step": 847
+    },
+    {
+      "epoch": 0.4140625,
+      "grad_norm": 1.5273832082748413,
+      "learning_rate": 6.608133494969993e-06,
+      "loss": 0.3002,
+      "step": 848
+    },
+    {
+      "epoch": 0.41455078125,
+      "grad_norm": 1.2620773315429688,
+      "learning_rate": 6.600642389889657e-06,
+      "loss": 0.3599,
+      "step": 849
+    },
+    {
+      "epoch": 0.4150390625,
+      "grad_norm": 1.283124566078186,
+      "learning_rate": 6.593147279505352e-06,
+      "loss": 0.3348,
+      "step": 850
+    },
+    {
+      "epoch": 0.41552734375,
+      "grad_norm": 1.2876836061477661,
+      "learning_rate": 6.585648182572176e-06,
+      "loss": 0.347,
+      "step": 851
+    },
+    {
+      "epoch": 0.416015625,
+      "grad_norm": 2.6049535274505615,
+      "learning_rate": 6.578145117855192e-06,
+      "loss": 0.3305,
+      "step": 852
+    },
+    {
+      "epoch": 0.41650390625,
+      "grad_norm": 1.7834153175354004,
+      "learning_rate": 6.570638104129399e-06,
+      "loss": 0.323,
+      "step": 853
+    },
+    {
+      "epoch": 0.4169921875,
+      "grad_norm": 1.3892278671264648,
+      "learning_rate": 6.563127160179672e-06,
+      "loss": 0.3475,
+      "step": 854
+    },
+    {
+      "epoch": 0.41748046875,
+      "grad_norm": 1.4540331363677979,
+      "learning_rate": 6.555612304800727e-06,
+      "loss": 0.3442,
+      "step": 855
+    },
+    {
+      "epoch": 0.41796875,
+      "grad_norm": 1.058359146118164,
+      "learning_rate": 6.548093556797063e-06,
+      "loss": 0.3398,
+      "step": 856
+    },
+    {
+      "epoch": 0.41845703125,
+      "grad_norm": 1.587546706199646,
+      "learning_rate": 6.540570934982917e-06,
+      "loss": 0.3261,
+      "step": 857
+    },
+    {
+      "epoch": 0.4189453125,
+      "grad_norm": 2.1293222904205322,
+      "learning_rate": 6.533044458182229e-06,
+      "loss": 0.3755,
+      "step": 858
+    },
+    {
+      "epoch": 0.41943359375,
+      "grad_norm": 1.2648324966430664,
+      "learning_rate": 6.5255141452285765e-06,
+      "loss": 0.3001,
+      "step": 859
+    },
+    {
+      "epoch": 0.419921875,
+      "grad_norm": 1.4118512868881226,
+      "learning_rate": 6.51798001496514e-06,
+      "loss": 0.3376,
+      "step": 860
+    },
+    {
+      "epoch": 0.42041015625,
+      "grad_norm": 1.4707554578781128,
+      "learning_rate": 6.510442086244649e-06,
+      "loss": 0.3247,
+      "step": 861
+    },
+    {
+      "epoch": 0.4208984375,
+      "grad_norm": 1.3729053735733032,
+      "learning_rate": 6.502900377929344e-06,
+      "loss": 0.3039,
+      "step": 862
+    },
+    {
+      "epoch": 0.42138671875,
+      "grad_norm": 3.840740442276001,
+      "learning_rate": 6.4953549088909194e-06,
+      "loss": 0.3567,
+      "step": 863
+    },
+    {
+      "epoch": 0.421875,
+      "grad_norm": 1.3986668586730957,
+      "learning_rate": 6.487805698010476e-06,
+      "loss": 0.3313,
+      "step": 864
+    },
+    {
+      "epoch": 0.42236328125,
+      "grad_norm": 3.7465996742248535,
+      "learning_rate": 6.4802527641784866e-06,
+      "loss": 0.3357,
+      "step": 865
+    },
+    {
+      "epoch": 0.4228515625,
+      "grad_norm": 1.7644517421722412,
+      "learning_rate": 6.472696126294733e-06,
+      "loss": 0.3662,
+      "step": 866
+    },
+    {
+      "epoch": 0.42333984375,
+      "grad_norm": 1.2544833421707153,
+      "learning_rate": 6.4651358032682694e-06,
+      "loss": 0.3371,
+      "step": 867
+    },
+    {
+      "epoch": 0.423828125,
+      "grad_norm": 1.500871181488037,
+      "learning_rate": 6.457571814017368e-06,
+      "loss": 0.3224,
+      "step": 868
+    },
+    {
+      "epoch": 0.42431640625,
+      "grad_norm": 1.3260788917541504,
+      "learning_rate": 6.45000417746948e-06,
+      "loss": 0.3161,
+      "step": 869
+    },
+    {
+      "epoch": 0.4248046875,
+      "grad_norm": 1.334038257598877,
+      "learning_rate": 6.442432912561178e-06,
+      "loss": 0.3423,
+      "step": 870
+    },
+    {
+      "epoch": 0.42529296875,
+      "grad_norm": 1.378933310508728,
+      "learning_rate": 6.434858038238118e-06,
+      "loss": 0.3492,
+      "step": 871
+    },
+    {
+      "epoch": 0.42578125,
+      "grad_norm": 1.5512367486953735,
+      "learning_rate": 6.427279573454985e-06,
+      "loss": 0.3731,
+      "step": 872
+    },
+    {
+      "epoch": 0.42626953125,
+      "grad_norm": 1.4665623903274536,
+      "learning_rate": 6.4196975371754514e-06,
+      "loss": 0.3481,
+      "step": 873
+    },
+    {
+      "epoch": 0.4267578125,
+      "grad_norm": 1.5259501934051514,
+      "learning_rate": 6.412111948372122e-06,
+      "loss": 0.3439,
+      "step": 874
+    },
+    {
+      "epoch": 0.42724609375,
+      "grad_norm": 1.465909719467163,
+      "learning_rate": 6.404522826026496e-06,
+      "loss": 0.33,
+      "step": 875
+    },
+    {
+      "epoch": 0.427734375,
+      "grad_norm": 1.357045292854309,
+      "learning_rate": 6.396930189128912e-06,
+      "loss": 0.344,
+      "step": 876
+    },
+    {
+      "epoch": 0.42822265625,
+      "grad_norm": 1.352899193763733,
+      "learning_rate": 6.3893340566785046e-06,
+      "loss": 0.3021,
+      "step": 877
+    },
+    {
+      "epoch": 0.4287109375,
+      "grad_norm": 1.3821226358413696,
+      "learning_rate": 6.381734447683152e-06,
+      "loss": 0.3326,
+      "step": 878
+    },
+    {
+      "epoch": 0.42919921875,
+      "grad_norm": 1.675229787826538,
+      "learning_rate": 6.374131381159436e-06,
+      "loss": 0.4357,
+      "step": 879
+    },
+    {
+      "epoch": 0.4296875,
+      "grad_norm": 1.7067149877548218,
+      "learning_rate": 6.366524876132589e-06,
+      "loss": 0.3018,
+      "step": 880
+    },
+    {
+      "epoch": 0.43017578125,
+      "grad_norm": 1.4271488189697266,
+      "learning_rate": 6.358914951636444e-06,
+      "loss": 0.3468,
+      "step": 881
+    },
+    {
+      "epoch": 0.4306640625,
+      "grad_norm": 1.3299568891525269,
+      "learning_rate": 6.351301626713398e-06,
+      "loss": 0.3466,
+      "step": 882
+    },
+    {
+      "epoch": 0.43115234375,
+      "grad_norm": 1.6695646047592163,
+      "learning_rate": 6.343684920414348e-06,
+      "loss": 0.3214,
+      "step": 883
+    },
+    {
+      "epoch": 0.431640625,
+      "grad_norm": 1.3570027351379395,
+      "learning_rate": 6.3360648517986605e-06,
+      "loss": 0.3382,
+      "step": 884
+    },
+    {
+      "epoch": 0.43212890625,
+      "grad_norm": 1.385907769203186,
+      "learning_rate": 6.32844143993411e-06,
+      "loss": 0.3092,
+      "step": 885
+    },
+    {
+      "epoch": 0.4326171875,
+      "grad_norm": 1.5601329803466797,
+      "learning_rate": 6.320814703896838e-06,
+      "loss": 0.3587,
+      "step": 886
+    },
+    {
+      "epoch": 0.43310546875,
+      "grad_norm": 1.39394211769104,
+      "learning_rate": 6.313184662771305e-06,
+      "loss": 0.3404,
+      "step": 887
+    },
+    {
+      "epoch": 0.43359375,
+      "grad_norm": 1.2028573751449585,
+      "learning_rate": 6.305551335650244e-06,
+      "loss": 0.3548,
+      "step": 888
+    },
+    {
+      "epoch": 0.43408203125,
+      "grad_norm": 4.250852108001709,
+      "learning_rate": 6.297914741634605e-06,
+      "loss": 0.3454,
+      "step": 889
+    },
+    {
+      "epoch": 0.4345703125,
+      "grad_norm": 1.5344691276550293,
+      "learning_rate": 6.290274899833517e-06,
+      "loss": 0.3176,
+      "step": 890
+    },
+    {
+      "epoch": 0.43505859375,
+      "grad_norm": 1.7602498531341553,
+      "learning_rate": 6.2826318293642385e-06,
+      "loss": 0.339,
+      "step": 891
+    },
+    {
+      "epoch": 0.435546875,
+      "grad_norm": 1.1949964761734009,
+      "learning_rate": 6.274985549352098e-06,
+      "loss": 0.304,
+      "step": 892
+    },
+    {
+      "epoch": 0.43603515625,
+      "grad_norm": 1.1564438343048096,
+      "learning_rate": 6.267336078930464e-06,
+      "loss": 0.3145,
+      "step": 893
+    },
+    {
+      "epoch": 0.4365234375,
+      "grad_norm": 1.3757606744766235,
+      "learning_rate": 6.259683437240683e-06,
+      "loss": 0.3385,
+      "step": 894
+    },
+    {
+      "epoch": 0.43701171875,
+      "grad_norm": 1.8371174335479736,
+      "learning_rate": 6.252027643432044e-06,
+      "loss": 0.3355,
+      "step": 895
+    },
+    {
+      "epoch": 0.4375,
+      "grad_norm": 1.334598422050476,
+      "learning_rate": 6.244368716661714e-06,
+      "loss": 0.3276,
+      "step": 896
+    },
+    {
+      "epoch": 0.43798828125,
+      "grad_norm": 1.5038282871246338,
+      "learning_rate": 6.236706676094705e-06,
+      "loss": 0.3522,
+      "step": 897
+    },
+    {
+      "epoch": 0.4384765625,
+      "grad_norm": 3.6733760833740234,
+      "learning_rate": 6.229041540903823e-06,
+      "loss": 0.3431,
+      "step": 898
+    },
+    {
+      "epoch": 0.43896484375,
+      "grad_norm": 1.5863288640975952,
+      "learning_rate": 6.221373330269613e-06,
+      "loss": 0.3324,
+      "step": 899
+    },
+    {
+      "epoch": 0.439453125,
+      "grad_norm": 1.4606237411499023,
+      "learning_rate": 6.213702063380317e-06,
+      "loss": 0.3226,
+      "step": 900
+    },
+    {
+      "epoch": 0.43994140625,
+      "grad_norm": 1.8370083570480347,
+      "learning_rate": 6.206027759431825e-06,
+      "loss": 0.3294,
+      "step": 901
+    },
+    {
+      "epoch": 0.4404296875,
+      "grad_norm": 1.6841802597045898,
+      "learning_rate": 6.198350437627631e-06,
+      "loss": 0.3238,
+      "step": 902
+    },
+    {
+      "epoch": 0.44091796875,
+      "grad_norm": 1.9791240692138672,
+      "learning_rate": 6.190670117178772e-06,
+      "loss": 0.3326,
+      "step": 903
+    },
+    {
+      "epoch": 0.44140625,
+      "grad_norm": 1.4503194093704224,
+      "learning_rate": 6.182986817303794e-06,
+      "loss": 0.3544,
+      "step": 904
+    },
+    {
+      "epoch": 0.44189453125,
+      "grad_norm": 1.9381232261657715,
+      "learning_rate": 6.175300557228698e-06,
+      "loss": 0.3278,
+      "step": 905
+    },
+    {
+      "epoch": 0.4423828125,
+      "grad_norm": 4.399080753326416,
+      "learning_rate": 6.167611356186895e-06,
+      "loss": 0.3367,
+      "step": 906
+    },
+    {
+      "epoch": 0.44287109375,
+      "grad_norm": 1.4784455299377441,
+      "learning_rate": 6.159919233419147e-06,
+      "loss": 0.3559,
+      "step": 907
+    },
+    {
+      "epoch": 0.443359375,
+      "grad_norm": 1.9754478931427002,
+      "learning_rate": 6.152224208173533e-06,
+      "loss": 0.3311,
+      "step": 908
+    },
+    {
+      "epoch": 0.44384765625,
+      "grad_norm": 1.5615670680999756,
+      "learning_rate": 6.144526299705396e-06,
+      "loss": 0.4023,
+      "step": 909
+    },
+    {
+      "epoch": 0.4443359375,
+      "grad_norm": 1.461332082748413,
+      "learning_rate": 6.136825527277295e-06,
+      "loss": 0.3026,
+      "step": 910
+    },
+    {
+      "epoch": 0.44482421875,
+      "grad_norm": 1.4366703033447266,
+      "learning_rate": 6.129121910158945e-06,
+      "loss": 0.336,
+      "step": 911
+    },
+    {
+      "epoch": 0.4453125,
+      "grad_norm": 2.06691575050354,
+      "learning_rate": 6.12141546762719e-06,
+      "loss": 0.342,
+      "step": 912
+    },
+    {
+      "epoch": 0.44580078125,
+      "grad_norm": 1.7794272899627686,
+      "learning_rate": 6.11370621896594e-06,
+      "loss": 0.3532,
+      "step": 913
+    },
+    {
+      "epoch": 0.4462890625,
+      "grad_norm": 1.4335381984710693,
+      "learning_rate": 6.105994183466131e-06,
+      "loss": 0.3471,
+      "step": 914
+    },
+    {
+      "epoch": 0.44677734375,
+      "grad_norm": 5.071071147918701,
+      "learning_rate": 6.0982793804256636e-06,
+      "loss": 0.336,
+      "step": 915
+    },
+    {
+      "epoch": 0.447265625,
+      "grad_norm": 1.2241181135177612,
+      "learning_rate": 6.090561829149373e-06,
+      "loss": 0.3232,
+      "step": 916
+    },
+    {
+      "epoch": 0.44775390625,
+      "grad_norm": 1.267858624458313,
+      "learning_rate": 6.082841548948966e-06,
+      "loss": 0.3556,
+      "step": 917
+    },
+    {
+      "epoch": 0.4482421875,
+      "grad_norm": 1.1905056238174438,
+      "learning_rate": 6.07511855914298e-06,
+      "loss": 0.2941,
+      "step": 918
+    },
+    {
+      "epoch": 0.44873046875,
+      "grad_norm": 1.2715431451797485,
+      "learning_rate": 6.067392879056729e-06,
+      "loss": 0.3159,
+      "step": 919
+    },
+    {
+      "epoch": 0.44921875,
+      "grad_norm": 1.2241966724395752,
+      "learning_rate": 6.059664528022267e-06,
+      "loss": 0.3141,
+      "step": 920
+    },
+    {
+      "epoch": 0.44970703125,
+      "grad_norm": 1.6341863870620728,
+      "learning_rate": 6.051933525378323e-06,
+      "loss": 0.3319,
+      "step": 921
+    },
+    {
+      "epoch": 0.4501953125,
+      "grad_norm": 3.6661813259124756,
+      "learning_rate": 6.044199890470267e-06,
+      "loss": 0.3482,
+      "step": 922
+    },
+    {
+      "epoch": 0.45068359375,
+      "grad_norm": 1.4551990032196045,
+      "learning_rate": 6.036463642650049e-06,
+      "loss": 0.3899,
+      "step": 923
+    },
+    {
+      "epoch": 0.451171875,
+      "grad_norm": 1.8738077878952026,
+      "learning_rate": 6.028724801276167e-06,
+      "loss": 0.3412,
+      "step": 924
+    },
+    {
+      "epoch": 0.45166015625,
+      "grad_norm": 1.3348729610443115,
+      "learning_rate": 6.020983385713601e-06,
+      "loss": 0.3194,
+      "step": 925
+    },
+    {
+      "epoch": 0.4521484375,
+      "grad_norm": 1.675868034362793,
+      "learning_rate": 6.013239415333776e-06,
+      "loss": 0.338,
+      "step": 926
+    },
+    {
+      "epoch": 0.45263671875,
+      "grad_norm": 1.5089606046676636,
+      "learning_rate": 6.005492909514507e-06,
+      "loss": 0.3502,
+      "step": 927
+    },
+    {
+      "epoch": 0.453125,
+      "grad_norm": 1.6367465257644653,
+      "learning_rate": 5.997743887639959e-06,
+      "loss": 0.3356,
+      "step": 928
+    },
+    {
+      "epoch": 0.45361328125,
+      "grad_norm": 1.5445111989974976,
+      "learning_rate": 5.989992369100586e-06,
+      "loss": 0.3192,
+      "step": 929
+    },
+    {
+      "epoch": 0.4541015625,
+      "grad_norm": 1.2671817541122437,
+      "learning_rate": 5.982238373293093e-06,
+      "loss": 0.3282,
+      "step": 930
+    },
+    {
+      "epoch": 0.45458984375,
+      "grad_norm": 1.2266660928726196,
+      "learning_rate": 5.974481919620386e-06,
+      "loss": 0.3202,
+      "step": 931
+    },
+    {
+      "epoch": 0.455078125,
+      "grad_norm": 1.5652544498443604,
+      "learning_rate": 5.966723027491518e-06,
+      "loss": 0.3502,
+      "step": 932
+    },
+    {
+      "epoch": 0.45556640625,
+      "grad_norm": 1.2947496175765991,
+      "learning_rate": 5.958961716321644e-06,
+      "loss": 0.317,
+      "step": 933
+    },
+    {
+      "epoch": 0.4560546875,
+      "grad_norm": 2.053834915161133,
+      "learning_rate": 5.951198005531974e-06,
+      "loss": 0.308,
+      "step": 934
+    },
+    {
+      "epoch": 0.45654296875,
+      "grad_norm": 2.342907428741455,
+      "learning_rate": 5.943431914549721e-06,
+      "loss": 0.3314,
+      "step": 935
+    },
+    {
+      "epoch": 0.45703125,
+      "grad_norm": 1.5535999536514282,
+      "learning_rate": 5.9356634628080555e-06,
+      "loss": 0.3362,
+      "step": 936
+    },
+    {
+      "epoch": 0.45751953125,
+      "grad_norm": 1.607968807220459,
+      "learning_rate": 5.927892669746054e-06,
+      "loss": 0.317,
+      "step": 937
+    },
+    {
+      "epoch": 0.4580078125,
+      "grad_norm": 1.268129825592041,
+      "learning_rate": 5.920119554808651e-06,
+      "loss": 0.3278,
+      "step": 938
+    },
+    {
+      "epoch": 0.45849609375,
+      "grad_norm": 4.848256587982178,
+      "learning_rate": 5.912344137446593e-06,
+      "loss": 0.3448,
+      "step": 939
+    },
+    {
+      "epoch": 0.458984375,
+      "grad_norm": 1.1670955419540405,
+      "learning_rate": 5.904566437116388e-06,
+      "loss": 0.2967,
+      "step": 940
+    },
+    {
+      "epoch": 0.45947265625,
+      "grad_norm": 2.250368595123291,
+      "learning_rate": 5.896786473280255e-06,
+      "loss": 0.32,
+      "step": 941
+    },
+    {
+      "epoch": 0.4599609375,
+      "grad_norm": 1.5156008005142212,
+      "learning_rate": 5.889004265406077e-06,
+      "loss": 0.2914,
+      "step": 942
+    },
+    {
+      "epoch": 0.46044921875,
+      "grad_norm": 1.0980958938598633,
+      "learning_rate": 5.8812198329673545e-06,
+      "loss": 0.304,
+      "step": 943
+    },
+    {
+      "epoch": 0.4609375,
+      "grad_norm": 1.7652188539505005,
+      "learning_rate": 5.873433195443152e-06,
+      "loss": 0.3497,
+      "step": 944
+    },
+    {
+      "epoch": 0.46142578125,
+      "grad_norm": 1.977793574333191,
+      "learning_rate": 5.865644372318053e-06,
+      "loss": 0.3598,
+      "step": 945
+    },
+    {
+      "epoch": 0.4619140625,
+      "grad_norm": 1.490369200706482,
+      "learning_rate": 5.857853383082112e-06,
+      "loss": 0.3433,
+      "step": 946
+    },
+    {
+      "epoch": 0.46240234375,
+      "grad_norm": 5.214506149291992,
+      "learning_rate": 5.8500602472307974e-06,
+      "loss": 0.3506,
+      "step": 947
+    },
+    {
+      "epoch": 0.462890625,
+      "grad_norm": 1.304093837738037,
+      "learning_rate": 5.842264984264958e-06,
+      "loss": 0.3035,
+      "step": 948
+    },
+    {
+      "epoch": 0.46337890625,
+      "grad_norm": 1.2441211938858032,
+      "learning_rate": 5.834467613690759e-06,
+      "loss": 0.3308,
+      "step": 949
+    },
+    {
+      "epoch": 0.4638671875,
+      "grad_norm": 1.0881738662719727,
+      "learning_rate": 5.82666815501964e-06,
+      "loss": 0.3163,
+      "step": 950
+    },
+    {
+      "epoch": 0.46435546875,
+      "grad_norm": 1.4398066997528076,
+      "learning_rate": 5.8188666277682695e-06,
+      "loss": 0.327,
+      "step": 951
+    },
+    {
+      "epoch": 0.46484375,
+      "grad_norm": 1.81572425365448,
+      "learning_rate": 5.8110630514584854e-06,
+      "loss": 0.3328,
+      "step": 952
+    },
+    {
+      "epoch": 0.46533203125,
+      "grad_norm": 1.5575212240219116,
+      "learning_rate": 5.803257445617263e-06,
+      "loss": 0.3495,
+      "step": 953
+    },
+    {
+      "epoch": 0.4658203125,
+      "grad_norm": 1.3975605964660645,
+      "learning_rate": 5.795449829776645e-06,
+      "loss": 0.3448,
+      "step": 954
+    },
+    {
+      "epoch": 0.46630859375,
+      "grad_norm": 1.2950125932693481,
+      "learning_rate": 5.787640223473713e-06,
+      "loss": 0.3617,
+      "step": 955
+    },
+    {
+      "epoch": 0.466796875,
+      "grad_norm": 1.3984689712524414,
+      "learning_rate": 5.779828646250522e-06,
+      "loss": 0.3608,
+      "step": 956
+    },
+    {
+      "epoch": 0.46728515625,
+      "grad_norm": 1.0765591859817505,
+      "learning_rate": 5.772015117654065e-06,
+      "loss": 0.3093,
+      "step": 957
+    },
+    {
+      "epoch": 0.4677734375,
+      "grad_norm": 1.5954604148864746,
+      "learning_rate": 5.764199657236214e-06,
+      "loss": 0.3504,
+      "step": 958
+    },
+    {
+      "epoch": 0.46826171875,
+      "grad_norm": 1.6604746580123901,
+      "learning_rate": 5.756382284553675e-06,
+      "loss": 0.3096,
+      "step": 959
+    },
+    {
+      "epoch": 0.46875,
+      "grad_norm": 1.3618206977844238,
+      "learning_rate": 5.7485630191679456e-06,
+      "loss": 0.3057,
+      "step": 960
+    },
+    {
+      "epoch": 0.46923828125,
+      "grad_norm": 1.217523217201233,
+      "learning_rate": 5.740741880645248e-06,
+      "loss": 0.3708,
+      "step": 961
+    },
+    {
+      "epoch": 0.4697265625,
+      "grad_norm": 1.2130963802337646,
+      "learning_rate": 5.7329188885565e-06,
+      "loss": 0.321,
+      "step": 962
+    },
+    {
+      "epoch": 0.47021484375,
+      "grad_norm": 1.3064903020858765,
+      "learning_rate": 5.725094062477256e-06,
+      "loss": 0.3211,
+      "step": 963
+    },
+    {
+      "epoch": 0.470703125,
+      "grad_norm": 1.5063132047653198,
+      "learning_rate": 5.717267421987659e-06,
+      "loss": 0.3307,
+      "step": 964
+    },
+    {
+      "epoch": 0.47119140625,
+      "grad_norm": 1.3585816621780396,
+      "learning_rate": 5.7094389866723905e-06,
+      "loss": 0.3631,
+      "step": 965
+    },
+    {
+      "epoch": 0.4716796875,
+      "grad_norm": 1.5815399885177612,
+      "learning_rate": 5.701608776120627e-06,
+      "loss": 0.352,
+      "step": 966
+    },
+    {
+      "epoch": 0.47216796875,
+      "grad_norm": 1.4560235738754272,
+      "learning_rate": 5.6937768099259845e-06,
+      "loss": 0.3109,
+      "step": 967
+    },
+    {
+      "epoch": 0.47265625,
+      "grad_norm": 1.8057149648666382,
+      "learning_rate": 5.685943107686476e-06,
+      "loss": 0.3218,
+      "step": 968
+    },
+    {
+      "epoch": 0.47314453125,
+      "grad_norm": 1.4362132549285889,
+      "learning_rate": 5.678107689004449e-06,
+      "loss": 0.3293,
+      "step": 969
+    },
+    {
+      "epoch": 0.4736328125,
+      "grad_norm": 2.0112991333007812,
+      "learning_rate": 5.670270573486555e-06,
+      "loss": 0.356,
+      "step": 970
+    },
+    {
+      "epoch": 0.47412109375,
+      "grad_norm": 1.2395293712615967,
+      "learning_rate": 5.662431780743691e-06,
+      "loss": 0.3439,
+      "step": 971
+    },
+    {
+      "epoch": 0.474609375,
+      "grad_norm": 1.4867768287658691,
+      "learning_rate": 5.6545913303909495e-06,
+      "loss": 0.3767,
+      "step": 972
+    },
+    {
+      "epoch": 0.47509765625,
+      "grad_norm": 1.210928201675415,
+      "learning_rate": 5.646749242047567e-06,
+      "loss": 0.3259,
+      "step": 973
+    },
+    {
+      "epoch": 0.4755859375,
+      "grad_norm": 1.157676100730896,
+      "learning_rate": 5.6389055353368826e-06,
+      "loss": 0.336,
+      "step": 974
+    },
+    {
+      "epoch": 0.47607421875,
+      "grad_norm": 1.485719919204712,
+      "learning_rate": 5.631060229886287e-06,
+      "loss": 0.3121,
+      "step": 975
+    },
+    {
+      "epoch": 0.4765625,
+      "grad_norm": 1.1137949228286743,
+      "learning_rate": 5.6232133453271676e-06,
+      "loss": 0.3362,
+      "step": 976
+    },
+    {
+      "epoch": 0.47705078125,
+      "grad_norm": 1.213346004486084,
+      "learning_rate": 5.615364901294863e-06,
+      "loss": 0.3194,
+      "step": 977
+    },
+    {
+      "epoch": 0.4775390625,
+      "grad_norm": 1.3590606451034546,
+      "learning_rate": 5.607514917428618e-06,
+      "loss": 0.3484,
+      "step": 978
+    },
+    {
+      "epoch": 0.47802734375,
+      "grad_norm": 2.0311455726623535,
+      "learning_rate": 5.599663413371527e-06,
+      "loss": 0.3419,
+      "step": 979
+    },
+    {
+      "epoch": 0.478515625,
+      "grad_norm": 1.195672869682312,
+      "learning_rate": 5.5918104087704925e-06,
+      "loss": 0.339,
+      "step": 980
+    },
+    {
+      "epoch": 0.47900390625,
+      "grad_norm": 1.8912562131881714,
+      "learning_rate": 5.583955923276163e-06,
+      "loss": 0.3427,
+      "step": 981
+    },
+    {
+      "epoch": 0.4794921875,
+      "grad_norm": 2.002305030822754,
+      "learning_rate": 5.576099976542904e-06,
+      "loss": 0.3595,
+      "step": 982
+    },
+    {
+      "epoch": 0.47998046875,
+      "grad_norm": 1.4438331127166748,
+      "learning_rate": 5.56824258822873e-06,
+      "loss": 0.3632,
+      "step": 983
+    },
+    {
+      "epoch": 0.48046875,
+      "grad_norm": 1.366222620010376,
+      "learning_rate": 5.560383777995264e-06,
+      "loss": 0.3188,
+      "step": 984
+    },
+    {
+      "epoch": 0.48095703125,
+      "grad_norm": 1.3330532312393188,
+      "learning_rate": 5.552523565507689e-06,
+      "loss": 0.3262,
+      "step": 985
+    },
+    {
+      "epoch": 0.4814453125,
+      "grad_norm": 1.5084117650985718,
+      "learning_rate": 5.544661970434696e-06,
+      "loss": 0.325,
+      "step": 986
+    },
+    {
+      "epoch": 0.48193359375,
+      "grad_norm": 1.0425949096679688,
+      "learning_rate": 5.536799012448435e-06,
+      "loss": 0.315,
+      "step": 987
+    },
+    {
+      "epoch": 0.482421875,
+      "grad_norm": 2.695110559463501,
+      "learning_rate": 5.528934711224467e-06,
+      "loss": 0.3166,
+      "step": 988
+    },
+    {
+      "epoch": 0.48291015625,
+      "grad_norm": 1.3446696996688843,
+      "learning_rate": 5.521069086441715e-06,
+      "loss": 0.3437,
+      "step": 989
+    },
+    {
+      "epoch": 0.4833984375,
+      "grad_norm": 1.360203742980957,
+      "learning_rate": 5.513202157782411e-06,
+      "loss": 0.3472,
+      "step": 990
+    },
+    {
+      "epoch": 0.48388671875,
+      "grad_norm": 1.3492072820663452,
+      "learning_rate": 5.505333944932053e-06,
+      "loss": 0.3363,
+      "step": 991
+    },
+    {
+      "epoch": 0.484375,
+      "grad_norm": 1.1588752269744873,
+      "learning_rate": 5.497464467579351e-06,
+      "loss": 0.338,
+      "step": 992
+    },
+    {
+      "epoch": 0.48486328125,
+      "grad_norm": 1.4233770370483398,
+      "learning_rate": 5.48959374541618e-06,
+      "loss": 0.336,
+      "step": 993
+    },
+    {
+      "epoch": 0.4853515625,
+      "grad_norm": 1.3421063423156738,
+      "learning_rate": 5.4817217981375286e-06,
+      "loss": 0.324,
+      "step": 994
+    },
+    {
+      "epoch": 0.48583984375,
+      "grad_norm": 1.6678565740585327,
+      "learning_rate": 5.473848645441452e-06,
+      "loss": 0.3189,
+      "step": 995
+    },
+    {
+      "epoch": 0.486328125,
+      "grad_norm": 1.912955641746521,
+      "learning_rate": 5.465974307029021e-06,
+      "loss": 0.3643,
+      "step": 996
+    },
+    {
+      "epoch": 0.48681640625,
+      "grad_norm": 2.0670387744903564,
+      "learning_rate": 5.458098802604273e-06,
+      "loss": 0.332,
+      "step": 997
+    },
+    {
+      "epoch": 0.4873046875,
+      "grad_norm": 2.6159446239471436,
+      "learning_rate": 5.450222151874166e-06,
+      "loss": 0.3674,
+      "step": 998
+    },
+    {
+      "epoch": 0.48779296875,
+      "grad_norm": 1.3627862930297852,
+      "learning_rate": 5.442344374548524e-06,
+      "loss": 0.3496,
+      "step": 999
+    },
+    {
+      "epoch": 0.48828125,
+      "grad_norm": 1.4907851219177246,
+      "learning_rate": 5.43446549033999e-06,
+      "loss": 0.3475,
+      "step": 1000
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2048,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.000488202654974e+18,
+  "train_batch_size": 24,
+  "trial_name": null,
+  "trial_params": null
+}