checkpoint-sft / trainer_state.json
LemTenku's picture
Upload folder using huggingface_hub
a193479 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.340744709349049,
"eval_steps": 500,
"global_step": 14932,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.000669702652022502,
"grad_norm": 0.0,
"learning_rate": 6.693440428380188e-07,
"loss": 2.3902,
"step": 10
},
{
"epoch": 0.001339405304045004,
"grad_norm": 0.0,
"learning_rate": 1.3386880856760376e-06,
"loss": 2.4415,
"step": 20
},
{
"epoch": 0.002009107956067506,
"grad_norm": 0.0,
"learning_rate": 2.0080321285140564e-06,
"loss": 2.3648,
"step": 30
},
{
"epoch": 0.002678810608090008,
"grad_norm": 0.0,
"learning_rate": 2.6773761713520752e-06,
"loss": 2.2811,
"step": 40
},
{
"epoch": 0.00334851326011251,
"grad_norm": 0.0,
"learning_rate": 3.346720214190094e-06,
"loss": 2.4023,
"step": 50
},
{
"epoch": 0.004018215912135012,
"grad_norm": 0.0,
"learning_rate": 4.016064257028113e-06,
"loss": 2.3196,
"step": 60
},
{
"epoch": 0.004687918564157514,
"grad_norm": 0.0,
"learning_rate": 4.685408299866132e-06,
"loss": 2.2711,
"step": 70
},
{
"epoch": 0.005357621216180016,
"grad_norm": 0.0,
"learning_rate": 5.3547523427041504e-06,
"loss": 2.3224,
"step": 80
},
{
"epoch": 0.006027323868202518,
"grad_norm": 0.0,
"learning_rate": 6.024096385542169e-06,
"loss": 2.287,
"step": 90
},
{
"epoch": 0.00669702652022502,
"grad_norm": 0.0,
"learning_rate": 6.693440428380188e-06,
"loss": 2.2395,
"step": 100
},
{
"epoch": 0.0073667291722475225,
"grad_norm": 0.0,
"learning_rate": 7.362784471218207e-06,
"loss": 2.3105,
"step": 110
},
{
"epoch": 0.008036431824270024,
"grad_norm": 0.0,
"learning_rate": 8.032128514056226e-06,
"loss": 2.4276,
"step": 120
},
{
"epoch": 0.008706134476292525,
"grad_norm": 0.0,
"learning_rate": 8.701472556894244e-06,
"loss": 2.3688,
"step": 130
},
{
"epoch": 0.009375837128315028,
"grad_norm": 0.0,
"learning_rate": 9.370816599732263e-06,
"loss": 2.3826,
"step": 140
},
{
"epoch": 0.01004553978033753,
"grad_norm": 0.0,
"learning_rate": 1.0040160642570281e-05,
"loss": 2.383,
"step": 150
},
{
"epoch": 0.010715242432360031,
"grad_norm": 0.0,
"learning_rate": 1.0709504685408301e-05,
"loss": 2.3427,
"step": 160
},
{
"epoch": 0.011384945084382534,
"grad_norm": 0.0,
"learning_rate": 1.1378848728246319e-05,
"loss": 2.2256,
"step": 170
},
{
"epoch": 0.012054647736405036,
"grad_norm": 0.0,
"learning_rate": 1.2048192771084338e-05,
"loss": 2.3393,
"step": 180
},
{
"epoch": 0.012724350388427539,
"grad_norm": 0.0,
"learning_rate": 1.2717536813922356e-05,
"loss": 2.3954,
"step": 190
},
{
"epoch": 0.01339405304045004,
"grad_norm": 0.0,
"learning_rate": 1.3386880856760376e-05,
"loss": 2.2947,
"step": 200
},
{
"epoch": 0.014063755692472542,
"grad_norm": 0.0,
"learning_rate": 1.4056224899598394e-05,
"loss": 2.2528,
"step": 210
},
{
"epoch": 0.014733458344495045,
"grad_norm": 0.0,
"learning_rate": 1.4725568942436414e-05,
"loss": 2.2722,
"step": 220
},
{
"epoch": 0.015403160996517546,
"grad_norm": 0.0,
"learning_rate": 1.5394912985274433e-05,
"loss": 2.3249,
"step": 230
},
{
"epoch": 0.016072863648540048,
"grad_norm": 0.0,
"learning_rate": 1.606425702811245e-05,
"loss": 2.3077,
"step": 240
},
{
"epoch": 0.01674256630056255,
"grad_norm": 0.0,
"learning_rate": 1.673360107095047e-05,
"loss": 2.3056,
"step": 250
},
{
"epoch": 0.01741226895258505,
"grad_norm": 0.0,
"learning_rate": 1.7402945113788487e-05,
"loss": 2.3655,
"step": 260
},
{
"epoch": 0.018081971604607554,
"grad_norm": 0.0,
"learning_rate": 1.8072289156626505e-05,
"loss": 2.2622,
"step": 270
},
{
"epoch": 0.018751674256630057,
"grad_norm": 0.0,
"learning_rate": 1.8741633199464527e-05,
"loss": 2.2843,
"step": 280
},
{
"epoch": 0.019421376908652557,
"grad_norm": 0.0,
"learning_rate": 1.9410977242302544e-05,
"loss": 2.1858,
"step": 290
},
{
"epoch": 0.02009107956067506,
"grad_norm": 0.0,
"learning_rate": 2.0080321285140562e-05,
"loss": 2.3016,
"step": 300
},
{
"epoch": 0.020760782212697563,
"grad_norm": 0.0,
"learning_rate": 2.074966532797858e-05,
"loss": 2.2798,
"step": 310
},
{
"epoch": 0.021430484864720063,
"grad_norm": 0.0,
"learning_rate": 2.1419009370816602e-05,
"loss": 2.2639,
"step": 320
},
{
"epoch": 0.022100187516742566,
"grad_norm": 0.0,
"learning_rate": 2.208835341365462e-05,
"loss": 2.2364,
"step": 330
},
{
"epoch": 0.02276989016876507,
"grad_norm": 0.0,
"learning_rate": 2.2757697456492638e-05,
"loss": 2.2894,
"step": 340
},
{
"epoch": 0.023439592820787572,
"grad_norm": 0.0,
"learning_rate": 2.3427041499330656e-05,
"loss": 2.3324,
"step": 350
},
{
"epoch": 0.02410929547281007,
"grad_norm": 0.0,
"learning_rate": 2.4096385542168677e-05,
"loss": 2.3137,
"step": 360
},
{
"epoch": 0.024778998124832575,
"grad_norm": 0.0,
"learning_rate": 2.4765729585006695e-05,
"loss": 2.2814,
"step": 370
},
{
"epoch": 0.025448700776855078,
"grad_norm": 0.0,
"learning_rate": 2.5435073627844713e-05,
"loss": 2.3252,
"step": 380
},
{
"epoch": 0.026118403428877578,
"grad_norm": 0.0,
"learning_rate": 2.6104417670682734e-05,
"loss": 2.3306,
"step": 390
},
{
"epoch": 0.02678810608090008,
"grad_norm": 0.0,
"learning_rate": 2.6773761713520752e-05,
"loss": 2.3344,
"step": 400
},
{
"epoch": 0.027457808732922584,
"grad_norm": 0.0,
"learning_rate": 2.7443105756358774e-05,
"loss": 2.2686,
"step": 410
},
{
"epoch": 0.028127511384945084,
"grad_norm": 0.0,
"learning_rate": 2.8112449799196788e-05,
"loss": 2.2904,
"step": 420
},
{
"epoch": 0.028797214036967587,
"grad_norm": 0.0,
"learning_rate": 2.878179384203481e-05,
"loss": 2.3387,
"step": 430
},
{
"epoch": 0.02946691668899009,
"grad_norm": 0.0,
"learning_rate": 2.9451137884872827e-05,
"loss": 2.2865,
"step": 440
},
{
"epoch": 0.03013661934101259,
"grad_norm": 0.0,
"learning_rate": 3.012048192771085e-05,
"loss": 2.2472,
"step": 450
},
{
"epoch": 0.030806321993035093,
"grad_norm": 0.0,
"learning_rate": 3.078982597054887e-05,
"loss": 2.2773,
"step": 460
},
{
"epoch": 0.031476024645057596,
"grad_norm": 0.0,
"learning_rate": 3.1459170013386885e-05,
"loss": 2.3224,
"step": 470
},
{
"epoch": 0.032145727297080096,
"grad_norm": 0.0,
"learning_rate": 3.21285140562249e-05,
"loss": 2.1539,
"step": 480
},
{
"epoch": 0.032815429949102595,
"grad_norm": 0.0,
"learning_rate": 3.279785809906292e-05,
"loss": 2.3025,
"step": 490
},
{
"epoch": 0.0334851326011251,
"grad_norm": 0.0,
"learning_rate": 3.346720214190094e-05,
"loss": 2.2328,
"step": 500
},
{
"epoch": 0.0341548352531476,
"grad_norm": 0.0,
"learning_rate": 3.413654618473896e-05,
"loss": 2.2389,
"step": 510
},
{
"epoch": 0.0348245379051701,
"grad_norm": 0.0,
"learning_rate": 3.4805890227576974e-05,
"loss": 2.2465,
"step": 520
},
{
"epoch": 0.03549424055719261,
"grad_norm": 0.0,
"learning_rate": 3.5475234270415e-05,
"loss": 2.2608,
"step": 530
},
{
"epoch": 0.03616394320921511,
"grad_norm": 0.0,
"learning_rate": 3.614457831325301e-05,
"loss": 2.2254,
"step": 540
},
{
"epoch": 0.03683364586123761,
"grad_norm": 0.0,
"learning_rate": 3.6813922356091035e-05,
"loss": 2.2041,
"step": 550
},
{
"epoch": 0.037503348513260114,
"grad_norm": 0.0,
"learning_rate": 3.748326639892905e-05,
"loss": 2.3156,
"step": 560
},
{
"epoch": 0.038173051165282613,
"grad_norm": 0.0,
"learning_rate": 3.815261044176707e-05,
"loss": 2.3493,
"step": 570
},
{
"epoch": 0.03884275381730511,
"grad_norm": 0.0,
"learning_rate": 3.882195448460509e-05,
"loss": 2.2966,
"step": 580
},
{
"epoch": 0.03951245646932762,
"grad_norm": 0.0,
"learning_rate": 3.949129852744311e-05,
"loss": 2.2741,
"step": 590
},
{
"epoch": 0.04018215912135012,
"grad_norm": 0.0,
"learning_rate": 4.0160642570281125e-05,
"loss": 2.198,
"step": 600
},
{
"epoch": 0.04085186177337262,
"grad_norm": 0.0,
"learning_rate": 4.082998661311915e-05,
"loss": 2.2987,
"step": 610
},
{
"epoch": 0.041521564425395126,
"grad_norm": 0.0,
"learning_rate": 4.149933065595716e-05,
"loss": 2.3219,
"step": 620
},
{
"epoch": 0.042191267077417625,
"grad_norm": 0.0,
"learning_rate": 4.2168674698795186e-05,
"loss": 2.2609,
"step": 630
},
{
"epoch": 0.042860969729440125,
"grad_norm": 0.0,
"learning_rate": 4.2838018741633203e-05,
"loss": 2.1874,
"step": 640
},
{
"epoch": 0.04353067238146263,
"grad_norm": 0.0,
"learning_rate": 4.350736278447122e-05,
"loss": 2.2077,
"step": 650
},
{
"epoch": 0.04420037503348513,
"grad_norm": 0.0,
"learning_rate": 4.417670682730924e-05,
"loss": 2.187,
"step": 660
},
{
"epoch": 0.04487007768550764,
"grad_norm": 0.0,
"learning_rate": 4.484605087014726e-05,
"loss": 2.2983,
"step": 670
},
{
"epoch": 0.04553978033753014,
"grad_norm": 0.0,
"learning_rate": 4.5515394912985275e-05,
"loss": 2.3472,
"step": 680
},
{
"epoch": 0.04620948298955264,
"grad_norm": 0.0,
"learning_rate": 4.61847389558233e-05,
"loss": 2.2707,
"step": 690
},
{
"epoch": 0.046879185641575144,
"grad_norm": 0.0,
"learning_rate": 4.685408299866131e-05,
"loss": 2.3201,
"step": 700
},
{
"epoch": 0.047548888293597644,
"grad_norm": 0.0,
"learning_rate": 4.7523427041499336e-05,
"loss": 2.2628,
"step": 710
},
{
"epoch": 0.04821859094562014,
"grad_norm": 0.0,
"learning_rate": 4.8192771084337354e-05,
"loss": 2.3217,
"step": 720
},
{
"epoch": 0.04888829359764265,
"grad_norm": 0.0,
"learning_rate": 4.886211512717537e-05,
"loss": 2.3157,
"step": 730
},
{
"epoch": 0.04955799624966515,
"grad_norm": 0.0,
"learning_rate": 4.953145917001339e-05,
"loss": 2.2407,
"step": 740
},
{
"epoch": 0.05022769890168765,
"grad_norm": 0.0,
"learning_rate": 5.020080321285141e-05,
"loss": 2.2607,
"step": 750
},
{
"epoch": 0.050897401553710156,
"grad_norm": 0.0,
"learning_rate": 5.0870147255689426e-05,
"loss": 2.2192,
"step": 760
},
{
"epoch": 0.051567104205732656,
"grad_norm": 0.0,
"learning_rate": 5.1539491298527444e-05,
"loss": 2.2501,
"step": 770
},
{
"epoch": 0.052236806857755155,
"grad_norm": 0.0,
"learning_rate": 5.220883534136547e-05,
"loss": 2.2901,
"step": 780
},
{
"epoch": 0.05290650950977766,
"grad_norm": 0.0,
"learning_rate": 5.2878179384203486e-05,
"loss": 2.2391,
"step": 790
},
{
"epoch": 0.05357621216180016,
"grad_norm": 0.0,
"learning_rate": 5.3547523427041504e-05,
"loss": 2.2306,
"step": 800
},
{
"epoch": 0.05424591481382266,
"grad_norm": 0.0,
"learning_rate": 5.4216867469879516e-05,
"loss": 2.322,
"step": 810
},
{
"epoch": 0.05491561746584517,
"grad_norm": 0.0,
"learning_rate": 5.488621151271755e-05,
"loss": 2.2978,
"step": 820
},
{
"epoch": 0.05558532011786767,
"grad_norm": 0.0,
"learning_rate": 5.555555555555556e-05,
"loss": 2.3016,
"step": 830
},
{
"epoch": 0.05625502276989017,
"grad_norm": 0.0,
"learning_rate": 5.6224899598393576e-05,
"loss": 2.1525,
"step": 840
},
{
"epoch": 0.056924725421912674,
"grad_norm": 0.0,
"learning_rate": 5.6894243641231594e-05,
"loss": 2.2663,
"step": 850
},
{
"epoch": 0.057594428073935174,
"grad_norm": 0.0,
"learning_rate": 5.756358768406962e-05,
"loss": 2.2755,
"step": 860
},
{
"epoch": 0.05826413072595767,
"grad_norm": 0.0,
"learning_rate": 5.823293172690764e-05,
"loss": 2.2842,
"step": 870
},
{
"epoch": 0.05893383337798018,
"grad_norm": 0.0,
"learning_rate": 5.8902275769745655e-05,
"loss": 2.2635,
"step": 880
},
{
"epoch": 0.05960353603000268,
"grad_norm": 0.0,
"learning_rate": 5.9571619812583666e-05,
"loss": 2.2726,
"step": 890
},
{
"epoch": 0.06027323868202518,
"grad_norm": 0.0,
"learning_rate": 6.02409638554217e-05,
"loss": 2.2318,
"step": 900
},
{
"epoch": 0.060942941334047686,
"grad_norm": 0.0,
"learning_rate": 6.091030789825971e-05,
"loss": 2.3523,
"step": 910
},
{
"epoch": 0.061612643986070185,
"grad_norm": 0.0,
"learning_rate": 6.157965194109773e-05,
"loss": 2.317,
"step": 920
},
{
"epoch": 0.062282346638092685,
"grad_norm": 0.0,
"learning_rate": 6.224899598393574e-05,
"loss": 2.1225,
"step": 930
},
{
"epoch": 0.06295204929011519,
"grad_norm": 0.0,
"learning_rate": 6.291834002677377e-05,
"loss": 2.2357,
"step": 940
},
{
"epoch": 0.06362175194213769,
"grad_norm": 0.0,
"learning_rate": 6.358768406961179e-05,
"loss": 2.2447,
"step": 950
},
{
"epoch": 0.06429145459416019,
"grad_norm": 0.0,
"learning_rate": 6.42570281124498e-05,
"loss": 2.1638,
"step": 960
},
{
"epoch": 0.06496115724618269,
"grad_norm": 0.0,
"learning_rate": 6.492637215528782e-05,
"loss": 2.1353,
"step": 970
},
{
"epoch": 0.06563085989820519,
"grad_norm": 0.0,
"learning_rate": 6.559571619812584e-05,
"loss": 2.1564,
"step": 980
},
{
"epoch": 0.0663005625502277,
"grad_norm": 0.0,
"learning_rate": 6.626506024096386e-05,
"loss": 2.1921,
"step": 990
},
{
"epoch": 0.0669702652022502,
"grad_norm": 0.0,
"learning_rate": 6.693440428380188e-05,
"loss": 2.2064,
"step": 1000
},
{
"epoch": 0.0676399678542727,
"grad_norm": 0.0,
"learning_rate": 6.76037483266399e-05,
"loss": 2.223,
"step": 1010
},
{
"epoch": 0.0683096705062952,
"grad_norm": 0.0,
"learning_rate": 6.827309236947793e-05,
"loss": 2.3054,
"step": 1020
},
{
"epoch": 0.0689793731583177,
"grad_norm": 0.0,
"learning_rate": 6.894243641231593e-05,
"loss": 2.2002,
"step": 1030
},
{
"epoch": 0.0696490758103402,
"grad_norm": 0.0,
"learning_rate": 6.961178045515395e-05,
"loss": 2.249,
"step": 1040
},
{
"epoch": 0.07031877846236272,
"grad_norm": 0.0,
"learning_rate": 7.028112449799197e-05,
"loss": 2.1752,
"step": 1050
},
{
"epoch": 0.07098848111438522,
"grad_norm": 0.0,
"learning_rate": 7.095046854083e-05,
"loss": 2.2828,
"step": 1060
},
{
"epoch": 0.07165818376640772,
"grad_norm": 0.0,
"learning_rate": 7.161981258366802e-05,
"loss": 2.2456,
"step": 1070
},
{
"epoch": 0.07232788641843022,
"grad_norm": 0.0,
"learning_rate": 7.228915662650602e-05,
"loss": 2.3329,
"step": 1080
},
{
"epoch": 0.07299758907045271,
"grad_norm": 0.0,
"learning_rate": 7.295850066934404e-05,
"loss": 2.2911,
"step": 1090
},
{
"epoch": 0.07366729172247521,
"grad_norm": 0.0,
"learning_rate": 7.362784471218207e-05,
"loss": 2.3092,
"step": 1100
},
{
"epoch": 0.07433699437449773,
"grad_norm": 0.0,
"learning_rate": 7.429718875502009e-05,
"loss": 2.3058,
"step": 1110
},
{
"epoch": 0.07500669702652023,
"grad_norm": 0.0,
"learning_rate": 7.49665327978581e-05,
"loss": 2.3024,
"step": 1120
},
{
"epoch": 0.07567639967854273,
"grad_norm": 0.0,
"learning_rate": 7.563587684069612e-05,
"loss": 2.3491,
"step": 1130
},
{
"epoch": 0.07634610233056523,
"grad_norm": 0.0,
"learning_rate": 7.630522088353414e-05,
"loss": 2.3675,
"step": 1140
},
{
"epoch": 0.07701580498258773,
"grad_norm": 0.0,
"learning_rate": 7.697456492637216e-05,
"loss": 2.3257,
"step": 1150
},
{
"epoch": 0.07768550763461023,
"grad_norm": 0.0,
"learning_rate": 7.764390896921018e-05,
"loss": 2.3037,
"step": 1160
},
{
"epoch": 0.07835521028663274,
"grad_norm": 0.0,
"learning_rate": 7.83132530120482e-05,
"loss": 2.2627,
"step": 1170
},
{
"epoch": 0.07902491293865524,
"grad_norm": 0.0,
"learning_rate": 7.898259705488621e-05,
"loss": 2.3351,
"step": 1180
},
{
"epoch": 0.07969461559067774,
"grad_norm": 0.0,
"learning_rate": 7.965194109772423e-05,
"loss": 2.2829,
"step": 1190
},
{
"epoch": 0.08036431824270024,
"grad_norm": 0.0,
"learning_rate": 8.032128514056225e-05,
"loss": 2.2612,
"step": 1200
},
{
"epoch": 0.08103402089472274,
"grad_norm": 0.0,
"learning_rate": 8.099062918340027e-05,
"loss": 2.2377,
"step": 1210
},
{
"epoch": 0.08170372354674524,
"grad_norm": 0.0,
"learning_rate": 8.16599732262383e-05,
"loss": 2.2353,
"step": 1220
},
{
"epoch": 0.08237342619876775,
"grad_norm": 0.0,
"learning_rate": 8.232931726907632e-05,
"loss": 2.2509,
"step": 1230
},
{
"epoch": 0.08304312885079025,
"grad_norm": 0.0,
"learning_rate": 8.299866131191432e-05,
"loss": 2.2449,
"step": 1240
},
{
"epoch": 0.08371283150281275,
"grad_norm": 0.0,
"learning_rate": 8.366800535475234e-05,
"loss": 2.2076,
"step": 1250
},
{
"epoch": 0.08438253415483525,
"grad_norm": 0.0,
"learning_rate": 8.433734939759037e-05,
"loss": 2.2714,
"step": 1260
},
{
"epoch": 0.08505223680685775,
"grad_norm": 0.0,
"learning_rate": 8.500669344042839e-05,
"loss": 2.2424,
"step": 1270
},
{
"epoch": 0.08572193945888025,
"grad_norm": 0.0,
"learning_rate": 8.567603748326641e-05,
"loss": 2.2711,
"step": 1280
},
{
"epoch": 0.08639164211090276,
"grad_norm": 0.0,
"learning_rate": 8.634538152610442e-05,
"loss": 2.2266,
"step": 1290
},
{
"epoch": 0.08706134476292526,
"grad_norm": 0.0,
"learning_rate": 8.701472556894244e-05,
"loss": 2.2329,
"step": 1300
},
{
"epoch": 0.08773104741494776,
"grad_norm": 0.0,
"learning_rate": 8.768406961178046e-05,
"loss": 2.2388,
"step": 1310
},
{
"epoch": 0.08840075006697026,
"grad_norm": 0.0,
"learning_rate": 8.835341365461848e-05,
"loss": 2.3047,
"step": 1320
},
{
"epoch": 0.08907045271899276,
"grad_norm": 0.0,
"learning_rate": 8.90227576974565e-05,
"loss": 2.2679,
"step": 1330
},
{
"epoch": 0.08974015537101528,
"grad_norm": 0.0,
"learning_rate": 8.969210174029451e-05,
"loss": 2.3128,
"step": 1340
},
{
"epoch": 0.09040985802303778,
"grad_norm": 0.0,
"learning_rate": 9.036144578313253e-05,
"loss": 2.2134,
"step": 1350
},
{
"epoch": 0.09107956067506028,
"grad_norm": 0.0,
"learning_rate": 9.103078982597055e-05,
"loss": 2.2883,
"step": 1360
},
{
"epoch": 0.09174926332708278,
"grad_norm": 0.0,
"learning_rate": 9.170013386880857e-05,
"loss": 2.2464,
"step": 1370
},
{
"epoch": 0.09241896597910527,
"grad_norm": 0.0,
"learning_rate": 9.23694779116466e-05,
"loss": 2.2318,
"step": 1380
},
{
"epoch": 0.09308866863112777,
"grad_norm": 0.0,
"learning_rate": 9.303882195448462e-05,
"loss": 2.229,
"step": 1390
},
{
"epoch": 0.09375837128315029,
"grad_norm": 0.0,
"learning_rate": 9.370816599732262e-05,
"loss": 2.2054,
"step": 1400
},
{
"epoch": 0.09442807393517279,
"grad_norm": 0.0,
"learning_rate": 9.437751004016064e-05,
"loss": 2.2667,
"step": 1410
},
{
"epoch": 0.09509777658719529,
"grad_norm": 0.0,
"learning_rate": 9.504685408299867e-05,
"loss": 2.2674,
"step": 1420
},
{
"epoch": 0.09576747923921779,
"grad_norm": 0.0,
"learning_rate": 9.571619812583669e-05,
"loss": 2.2865,
"step": 1430
},
{
"epoch": 0.09643718189124029,
"grad_norm": 0.0,
"learning_rate": 9.638554216867471e-05,
"loss": 2.2337,
"step": 1440
},
{
"epoch": 0.09710688454326279,
"grad_norm": 0.0,
"learning_rate": 9.705488621151271e-05,
"loss": 2.291,
"step": 1450
},
{
"epoch": 0.0977765871952853,
"grad_norm": 0.0,
"learning_rate": 9.772423025435074e-05,
"loss": 2.3411,
"step": 1460
},
{
"epoch": 0.0984462898473078,
"grad_norm": 0.0,
"learning_rate": 9.839357429718876e-05,
"loss": 2.1827,
"step": 1470
},
{
"epoch": 0.0991159924993303,
"grad_norm": 0.0,
"learning_rate": 9.906291834002678e-05,
"loss": 2.1987,
"step": 1480
},
{
"epoch": 0.0997856951513528,
"grad_norm": 0.0,
"learning_rate": 9.97322623828648e-05,
"loss": 2.2079,
"step": 1490
},
{
"epoch": 0.1004553978033753,
"grad_norm": 0.0,
"learning_rate": 9.999995081044314e-05,
"loss": 2.2484,
"step": 1500
},
{
"epoch": 0.1011251004553978,
"grad_norm": 0.0,
"learning_rate": 9.999965020794615e-05,
"loss": 2.3123,
"step": 1510
},
{
"epoch": 0.10179480310742031,
"grad_norm": 0.0,
"learning_rate": 9.99990763321247e-05,
"loss": 2.2429,
"step": 1520
},
{
"epoch": 0.10246450575944281,
"grad_norm": 0.0,
"learning_rate": 9.999822918611533e-05,
"loss": 2.2205,
"step": 1530
},
{
"epoch": 0.10313420841146531,
"grad_norm": 0.0,
"learning_rate": 9.999710877454811e-05,
"loss": 2.2077,
"step": 1540
},
{
"epoch": 0.10380391106348781,
"grad_norm": 0.0,
"learning_rate": 9.999571510354664e-05,
"loss": 2.3028,
"step": 1550
},
{
"epoch": 0.10447361371551031,
"grad_norm": 0.0,
"learning_rate": 9.999404818072808e-05,
"loss": 2.252,
"step": 1560
},
{
"epoch": 0.10514331636753281,
"grad_norm": 0.0,
"learning_rate": 9.999210801520296e-05,
"loss": 2.3754,
"step": 1570
},
{
"epoch": 0.10581301901955532,
"grad_norm": 0.0,
"learning_rate": 9.998989461757526e-05,
"loss": 2.2761,
"step": 1580
},
{
"epoch": 0.10648272167157782,
"grad_norm": 0.0,
"learning_rate": 9.998740799994235e-05,
"loss": 2.2168,
"step": 1590
},
{
"epoch": 0.10715242432360032,
"grad_norm": 0.0,
"learning_rate": 9.998464817589484e-05,
"loss": 2.2639,
"step": 1600
},
{
"epoch": 0.10782212697562282,
"grad_norm": 0.0,
"learning_rate": 9.998161516051656e-05,
"loss": 2.2687,
"step": 1610
},
{
"epoch": 0.10849182962764532,
"grad_norm": 0.0,
"learning_rate": 9.997830897038446e-05,
"loss": 2.2239,
"step": 1620
},
{
"epoch": 0.10916153227966782,
"grad_norm": 0.0,
"learning_rate": 9.997472962356854e-05,
"loss": 2.1802,
"step": 1630
},
{
"epoch": 0.10983123493169034,
"grad_norm": 0.0,
"learning_rate": 9.997087713963174e-05,
"loss": 2.2154,
"step": 1640
},
{
"epoch": 0.11050093758371284,
"grad_norm": 0.0,
"learning_rate": 9.996675153962984e-05,
"loss": 2.212,
"step": 1650
},
{
"epoch": 0.11117064023573534,
"grad_norm": 0.0,
"learning_rate": 9.996235284611131e-05,
"loss": 2.251,
"step": 1660
},
{
"epoch": 0.11184034288775783,
"grad_norm": 0.0,
"learning_rate": 9.995768108311722e-05,
"loss": 2.299,
"step": 1670
},
{
"epoch": 0.11251004553978033,
"grad_norm": 0.0,
"learning_rate": 9.995273627618116e-05,
"loss": 2.2021,
"step": 1680
},
{
"epoch": 0.11317974819180283,
"grad_norm": 0.0,
"learning_rate": 9.994751845232894e-05,
"loss": 2.244,
"step": 1690
},
{
"epoch": 0.11384945084382535,
"grad_norm": 0.0,
"learning_rate": 9.994202764007865e-05,
"loss": 2.3063,
"step": 1700
},
{
"epoch": 0.11451915349584785,
"grad_norm": 0.0,
"learning_rate": 9.993626386944031e-05,
"loss": 2.2809,
"step": 1710
},
{
"epoch": 0.11518885614787035,
"grad_norm": 0.0,
"learning_rate": 9.993022717191584e-05,
"loss": 2.2767,
"step": 1720
},
{
"epoch": 0.11585855879989285,
"grad_norm": 0.0,
"learning_rate": 9.992391758049889e-05,
"loss": 2.2354,
"step": 1730
},
{
"epoch": 0.11652826145191535,
"grad_norm": 0.0,
"learning_rate": 9.99173351296745e-05,
"loss": 2.2889,
"step": 1740
},
{
"epoch": 0.11719796410393785,
"grad_norm": 0.0,
"learning_rate": 9.99104798554191e-05,
"loss": 2.22,
"step": 1750
},
{
"epoch": 0.11786766675596036,
"grad_norm": 0.0,
"learning_rate": 9.990335179520023e-05,
"loss": 2.2941,
"step": 1760
},
{
"epoch": 0.11853736940798286,
"grad_norm": 0.0,
"learning_rate": 9.989595098797635e-05,
"loss": 2.2584,
"step": 1770
},
{
"epoch": 0.11920707206000536,
"grad_norm": 0.0,
"learning_rate": 9.988827747419659e-05,
"loss": 2.2408,
"step": 1780
},
{
"epoch": 0.11987677471202786,
"grad_norm": 0.0,
"learning_rate": 9.988033129580059e-05,
"loss": 2.1831,
"step": 1790
},
{
"epoch": 0.12054647736405036,
"grad_norm": 0.0,
"learning_rate": 9.98721124962182e-05,
"loss": 2.2853,
"step": 1800
},
{
"epoch": 0.12121618001607286,
"grad_norm": 0.0,
"learning_rate": 9.986362112036935e-05,
"loss": 2.2537,
"step": 1810
},
{
"epoch": 0.12188588266809537,
"grad_norm": 0.0,
"learning_rate": 9.985485721466366e-05,
"loss": 2.2257,
"step": 1820
},
{
"epoch": 0.12255558532011787,
"grad_norm": 0.0,
"learning_rate": 9.984582082700029e-05,
"loss": 2.3031,
"step": 1830
},
{
"epoch": 0.12322528797214037,
"grad_norm": 0.0,
"learning_rate": 9.98365120067677e-05,
"loss": 2.2435,
"step": 1840
},
{
"epoch": 0.12389499062416287,
"grad_norm": 0.0,
"learning_rate": 9.982693080484327e-05,
"loss": 2.1834,
"step": 1850
},
{
"epoch": 0.12456469327618537,
"grad_norm": 0.0,
"learning_rate": 9.981707727359308e-05,
"loss": 2.2215,
"step": 1860
},
{
"epoch": 0.12523439592820787,
"grad_norm": 0.0,
"learning_rate": 9.98069514668717e-05,
"loss": 2.3128,
"step": 1870
},
{
"epoch": 0.12590409858023038,
"grad_norm": 0.0,
"learning_rate": 9.979655344002172e-05,
"loss": 2.2512,
"step": 1880
},
{
"epoch": 0.12657380123225287,
"grad_norm": 0.0,
"learning_rate": 9.978588324987365e-05,
"loss": 2.2278,
"step": 1890
},
{
"epoch": 0.12724350388427538,
"grad_norm": 0.0,
"learning_rate": 9.977494095474546e-05,
"loss": 2.2445,
"step": 1900
},
{
"epoch": 0.1279132065362979,
"grad_norm": 0.0,
"learning_rate": 9.97637266144423e-05,
"loss": 2.2834,
"step": 1910
},
{
"epoch": 0.12858290918832038,
"grad_norm": 0.0,
"learning_rate": 9.975224029025619e-05,
"loss": 2.2807,
"step": 1920
},
{
"epoch": 0.1292526118403429,
"grad_norm": 0.0,
"learning_rate": 9.974048204496572e-05,
"loss": 2.3837,
"step": 1930
},
{
"epoch": 0.12992231449236538,
"grad_norm": 0.0,
"learning_rate": 9.97284519428356e-05,
"loss": 2.2318,
"step": 1940
},
{
"epoch": 0.1305920171443879,
"grad_norm": 0.0,
"learning_rate": 9.971615004961645e-05,
"loss": 2.1961,
"step": 1950
},
{
"epoch": 0.13126171979641038,
"grad_norm": 0.0,
"learning_rate": 9.970357643254429e-05,
"loss": 2.301,
"step": 1960
},
{
"epoch": 0.1319314224484329,
"grad_norm": 0.0,
"learning_rate": 9.96907311603403e-05,
"loss": 2.2454,
"step": 1970
},
{
"epoch": 0.1326011251004554,
"grad_norm": 0.0,
"learning_rate": 9.967761430321037e-05,
"loss": 2.2505,
"step": 1980
},
{
"epoch": 0.1332708277524779,
"grad_norm": 0.0,
"learning_rate": 9.966422593284474e-05,
"loss": 2.2103,
"step": 1990
},
{
"epoch": 0.1339405304045004,
"grad_norm": 0.0,
"learning_rate": 9.965056612241764e-05,
"loss": 2.3467,
"step": 2000
},
{
"epoch": 0.1346102330565229,
"grad_norm": 0.0,
"learning_rate": 9.963663494658681e-05,
"loss": 2.174,
"step": 2010
},
{
"epoch": 0.1352799357085454,
"grad_norm": 0.0,
"learning_rate": 9.962243248149314e-05,
"loss": 2.3771,
"step": 2020
},
{
"epoch": 0.13594963836056792,
"grad_norm": 0.0,
"learning_rate": 9.960795880476029e-05,
"loss": 2.3623,
"step": 2030
},
{
"epoch": 0.1366193410125904,
"grad_norm": 0.0,
"learning_rate": 9.95932139954942e-05,
"loss": 2.27,
"step": 2040
},
{
"epoch": 0.13728904366461292,
"grad_norm": 0.0,
"learning_rate": 9.95781981342827e-05,
"loss": 2.2534,
"step": 2050
},
{
"epoch": 0.1379587463166354,
"grad_norm": 0.0,
"learning_rate": 9.956291130319506e-05,
"loss": 2.2726,
"step": 2060
},
{
"epoch": 0.13862844896865792,
"grad_norm": 0.0,
"learning_rate": 9.95473535857815e-05,
"loss": 2.3065,
"step": 2070
},
{
"epoch": 0.1392981516206804,
"grad_norm": 0.0,
"learning_rate": 9.953152506707283e-05,
"loss": 2.1472,
"step": 2080
},
{
"epoch": 0.13996785427270292,
"grad_norm": 0.0,
"learning_rate": 9.951542583357986e-05,
"loss": 2.2893,
"step": 2090
},
{
"epoch": 0.14063755692472543,
"grad_norm": 0.0,
"learning_rate": 9.949905597329306e-05,
"loss": 2.1606,
"step": 2100
},
{
"epoch": 0.14130725957674792,
"grad_norm": 0.0,
"learning_rate": 9.948241557568196e-05,
"loss": 2.2155,
"step": 2110
},
{
"epoch": 0.14197696222877043,
"grad_norm": 0.0,
"learning_rate": 9.946550473169474e-05,
"loss": 2.2544,
"step": 2120
},
{
"epoch": 0.14264666488079292,
"grad_norm": 0.0,
"learning_rate": 9.944832353375769e-05,
"loss": 2.2661,
"step": 2130
},
{
"epoch": 0.14331636753281543,
"grad_norm": 0.0,
"learning_rate": 9.943087207577473e-05,
"loss": 2.2873,
"step": 2140
},
{
"epoch": 0.14398607018483794,
"grad_norm": 0.0,
"learning_rate": 9.941315045312689e-05,
"loss": 2.2044,
"step": 2150
},
{
"epoch": 0.14465577283686043,
"grad_norm": 0.0,
"learning_rate": 9.939515876267178e-05,
"loss": 2.3315,
"step": 2160
},
{
"epoch": 0.14532547548888294,
"grad_norm": 0.0,
"learning_rate": 9.937689710274308e-05,
"loss": 2.2413,
"step": 2170
},
{
"epoch": 0.14599517814090543,
"grad_norm": 0.0,
"learning_rate": 9.935836557314995e-05,
"loss": 2.2494,
"step": 2180
},
{
"epoch": 0.14666488079292794,
"grad_norm": 0.0,
"learning_rate": 9.933956427517657e-05,
"loss": 2.2623,
"step": 2190
},
{
"epoch": 0.14733458344495043,
"grad_norm": 0.0,
"learning_rate": 9.932049331158153e-05,
"loss": 2.2775,
"step": 2200
},
{
"epoch": 0.14800428609697294,
"grad_norm": 0.0,
"learning_rate": 9.930115278659727e-05,
"loss": 2.1811,
"step": 2210
},
{
"epoch": 0.14867398874899546,
"grad_norm": 0.0,
"learning_rate": 9.928154280592952e-05,
"loss": 2.2731,
"step": 2220
},
{
"epoch": 0.14934369140101794,
"grad_norm": 0.0,
"learning_rate": 9.92616634767567e-05,
"loss": 2.2588,
"step": 2230
},
{
"epoch": 0.15001339405304046,
"grad_norm": 0.0,
"learning_rate": 9.924151490772942e-05,
"loss": 2.2073,
"step": 2240
},
{
"epoch": 0.15068309670506294,
"grad_norm": 0.0,
"learning_rate": 9.922109720896973e-05,
"loss": 2.2965,
"step": 2250
},
{
"epoch": 0.15135279935708545,
"grad_norm": 0.0,
"learning_rate": 9.92004104920707e-05,
"loss": 2.1766,
"step": 2260
},
{
"epoch": 0.15202250200910797,
"grad_norm": 0.0,
"learning_rate": 9.917945487009566e-05,
"loss": 2.2915,
"step": 2270
},
{
"epoch": 0.15269220466113045,
"grad_norm": 0.0,
"learning_rate": 9.915823045757765e-05,
"loss": 2.2717,
"step": 2280
},
{
"epoch": 0.15336190731315297,
"grad_norm": 0.0,
"learning_rate": 9.913673737051882e-05,
"loss": 2.2301,
"step": 2290
},
{
"epoch": 0.15403160996517545,
"grad_norm": 0.0,
"learning_rate": 9.911497572638973e-05,
"loss": 2.2289,
"step": 2300
},
{
"epoch": 0.15470131261719797,
"grad_norm": 0.0,
"learning_rate": 9.909294564412874e-05,
"loss": 2.3655,
"step": 2310
},
{
"epoch": 0.15537101526922045,
"grad_norm": 0.0,
"learning_rate": 9.907064724414139e-05,
"loss": 2.1503,
"step": 2320
},
{
"epoch": 0.15604071792124297,
"grad_norm": 0.0,
"learning_rate": 9.904808064829967e-05,
"loss": 2.2254,
"step": 2330
},
{
"epoch": 0.15671042057326548,
"grad_norm": 0.0,
"learning_rate": 9.902524597994143e-05,
"loss": 2.28,
"step": 2340
},
{
"epoch": 0.15738012322528797,
"grad_norm": 0.0,
"learning_rate": 9.900214336386964e-05,
"loss": 2.18,
"step": 2350
},
{
"epoch": 0.15804982587731048,
"grad_norm": 0.0,
"learning_rate": 9.897877292635179e-05,
"loss": 2.3032,
"step": 2360
},
{
"epoch": 0.15871952852933296,
"grad_norm": 0.0,
"learning_rate": 9.895513479511907e-05,
"loss": 2.2472,
"step": 2370
},
{
"epoch": 0.15938923118135548,
"grad_norm": 0.0,
"learning_rate": 9.893122909936583e-05,
"loss": 2.1331,
"step": 2380
},
{
"epoch": 0.160058933833378,
"grad_norm": 0.0,
"learning_rate": 9.890705596974875e-05,
"loss": 2.1931,
"step": 2390
},
{
"epoch": 0.16072863648540048,
"grad_norm": 0.0,
"learning_rate": 9.88826155383862e-05,
"loss": 2.2551,
"step": 2400
},
{
"epoch": 0.161398339137423,
"grad_norm": 0.0,
"learning_rate": 9.885790793885746e-05,
"loss": 2.1717,
"step": 2410
},
{
"epoch": 0.16206804178944548,
"grad_norm": 0.0,
"learning_rate": 9.883293330620205e-05,
"loss": 2.2582,
"step": 2420
},
{
"epoch": 0.162737744441468,
"grad_norm": 0.0,
"learning_rate": 9.880769177691892e-05,
"loss": 2.2466,
"step": 2430
},
{
"epoch": 0.16340744709349048,
"grad_norm": 0.0,
"learning_rate": 9.878218348896577e-05,
"loss": 2.2734,
"step": 2440
},
{
"epoch": 0.164077149745513,
"grad_norm": 0.0,
"learning_rate": 9.875640858175827e-05,
"loss": 2.3381,
"step": 2450
},
{
"epoch": 0.1647468523975355,
"grad_norm": 0.0,
"learning_rate": 9.87303671961693e-05,
"loss": 2.2529,
"step": 2460
},
{
"epoch": 0.165416555049558,
"grad_norm": 0.0,
"learning_rate": 9.870405947452819e-05,
"loss": 2.2889,
"step": 2470
},
{
"epoch": 0.1660862577015805,
"grad_norm": 0.0,
"learning_rate": 9.86774855606199e-05,
"loss": 2.1831,
"step": 2480
},
{
"epoch": 0.166755960353603,
"grad_norm": 0.0,
"learning_rate": 9.86506455996843e-05,
"loss": 2.2241,
"step": 2490
},
{
"epoch": 0.1674256630056255,
"grad_norm": 0.0,
"learning_rate": 9.862353973841526e-05,
"loss": 2.2539,
"step": 2500
},
{
"epoch": 0.16809536565764802,
"grad_norm": 0.0,
"learning_rate": 9.859616812496008e-05,
"loss": 2.263,
"step": 2510
},
{
"epoch": 0.1687650683096705,
"grad_norm": 0.0,
"learning_rate": 9.856853090891843e-05,
"loss": 2.2179,
"step": 2520
},
{
"epoch": 0.16943477096169302,
"grad_norm": 0.0,
"learning_rate": 9.854062824134159e-05,
"loss": 2.2326,
"step": 2530
},
{
"epoch": 0.1701044736137155,
"grad_norm": 0.0,
"learning_rate": 9.851246027473173e-05,
"loss": 2.3374,
"step": 2540
},
{
"epoch": 0.17077417626573801,
"grad_norm": 0.0,
"learning_rate": 9.848402716304106e-05,
"loss": 2.2265,
"step": 2550
},
{
"epoch": 0.1714438789177605,
"grad_norm": 0.0,
"learning_rate": 9.845532906167083e-05,
"loss": 2.1868,
"step": 2560
},
{
"epoch": 0.17211358156978301,
"grad_norm": 0.0,
"learning_rate": 9.842636612747069e-05,
"loss": 2.2097,
"step": 2570
},
{
"epoch": 0.17278328422180553,
"grad_norm": 0.0,
"learning_rate": 9.839713851873766e-05,
"loss": 2.2601,
"step": 2580
},
{
"epoch": 0.173452986873828,
"grad_norm": 0.0,
"learning_rate": 9.836764639521539e-05,
"loss": 2.2559,
"step": 2590
},
{
"epoch": 0.17412268952585053,
"grad_norm": 0.0,
"learning_rate": 9.833788991809323e-05,
"loss": 2.2896,
"step": 2600
},
{
"epoch": 0.174792392177873,
"grad_norm": 0.0,
"learning_rate": 9.830786925000533e-05,
"loss": 2.2095,
"step": 2610
},
{
"epoch": 0.17546209482989553,
"grad_norm": 0.0,
"learning_rate": 9.827758455502978e-05,
"loss": 2.206,
"step": 2620
},
{
"epoch": 0.17613179748191804,
"grad_norm": 0.0,
"learning_rate": 9.824703599868776e-05,
"loss": 2.3201,
"step": 2630
},
{
"epoch": 0.17680150013394053,
"grad_norm": 0.0,
"learning_rate": 9.821622374794253e-05,
"loss": 2.2804,
"step": 2640
},
{
"epoch": 0.17747120278596304,
"grad_norm": 0.0,
"learning_rate": 9.81851479711986e-05,
"loss": 2.1435,
"step": 2650
},
{
"epoch": 0.17814090543798553,
"grad_norm": 0.0,
"learning_rate": 9.815380883830076e-05,
"loss": 2.1878,
"step": 2660
},
{
"epoch": 0.17881060809000804,
"grad_norm": 0.0,
"learning_rate": 9.812220652053318e-05,
"loss": 2.3302,
"step": 2670
},
{
"epoch": 0.17948031074203055,
"grad_norm": 0.0,
"learning_rate": 9.80903411906185e-05,
"loss": 2.2151,
"step": 2680
},
{
"epoch": 0.18015001339405304,
"grad_norm": 0.0,
"learning_rate": 9.805821302271685e-05,
"loss": 2.2155,
"step": 2690
},
{
"epoch": 0.18081971604607555,
"grad_norm": 0.0,
"learning_rate": 9.802582219242484e-05,
"loss": 2.2942,
"step": 2700
},
{
"epoch": 0.18148941869809804,
"grad_norm": 0.0,
"learning_rate": 9.799316887677471e-05,
"loss": 2.2229,
"step": 2710
},
{
"epoch": 0.18215912135012055,
"grad_norm": 0.0,
"learning_rate": 9.796025325423334e-05,
"loss": 2.2516,
"step": 2720
},
{
"epoch": 0.18282882400214304,
"grad_norm": 0.0,
"learning_rate": 9.792707550470122e-05,
"loss": 2.1446,
"step": 2730
},
{
"epoch": 0.18349852665416555,
"grad_norm": 0.0,
"learning_rate": 9.789363580951145e-05,
"loss": 2.2074,
"step": 2740
},
{
"epoch": 0.18416822930618806,
"grad_norm": 0.0,
"learning_rate": 9.785993435142891e-05,
"loss": 2.236,
"step": 2750
},
{
"epoch": 0.18483793195821055,
"grad_norm": 0.0,
"learning_rate": 9.782597131464901e-05,
"loss": 2.2744,
"step": 2760
},
{
"epoch": 0.18550763461023306,
"grad_norm": 0.0,
"learning_rate": 9.779174688479693e-05,
"loss": 2.2472,
"step": 2770
},
{
"epoch": 0.18617733726225555,
"grad_norm": 0.0,
"learning_rate": 9.775726124892646e-05,
"loss": 2.4463,
"step": 2780
},
{
"epoch": 0.18684703991427806,
"grad_norm": 0.0,
"learning_rate": 9.772251459551897e-05,
"loss": 2.1792,
"step": 2790
},
{
"epoch": 0.18751674256630058,
"grad_norm": 0.0,
"learning_rate": 9.768750711448249e-05,
"loss": 2.2459,
"step": 2800
},
{
"epoch": 0.18818644521832306,
"grad_norm": 0.0,
"learning_rate": 9.765223899715054e-05,
"loss": 2.1656,
"step": 2810
},
{
"epoch": 0.18885614787034558,
"grad_norm": 0.0,
"learning_rate": 9.761671043628124e-05,
"loss": 2.3268,
"step": 2820
},
{
"epoch": 0.18952585052236806,
"grad_norm": 0.0,
"learning_rate": 9.758092162605604e-05,
"loss": 2.2815,
"step": 2830
},
{
"epoch": 0.19019555317439057,
"grad_norm": 0.0,
"learning_rate": 9.754487276207889e-05,
"loss": 2.2414,
"step": 2840
},
{
"epoch": 0.19086525582641306,
"grad_norm": 0.0,
"learning_rate": 9.750856404137502e-05,
"loss": 2.1398,
"step": 2850
},
{
"epoch": 0.19153495847843557,
"grad_norm": 0.0,
"learning_rate": 9.747199566238991e-05,
"loss": 2.2683,
"step": 2860
},
{
"epoch": 0.1922046611304581,
"grad_norm": 0.0,
"learning_rate": 9.743516782498822e-05,
"loss": 2.1707,
"step": 2870
},
{
"epoch": 0.19287436378248057,
"grad_norm": 0.0,
"learning_rate": 9.739808073045264e-05,
"loss": 2.3202,
"step": 2880
},
{
"epoch": 0.1935440664345031,
"grad_norm": 0.0,
"learning_rate": 9.73607345814829e-05,
"loss": 2.2712,
"step": 2890
},
{
"epoch": 0.19421376908652557,
"grad_norm": 0.0,
"learning_rate": 9.732312958219453e-05,
"loss": 2.2749,
"step": 2900
},
{
"epoch": 0.1948834717385481,
"grad_norm": 0.0,
"learning_rate": 9.728526593811783e-05,
"loss": 2.1855,
"step": 2910
},
{
"epoch": 0.1955531743905706,
"grad_norm": 0.0,
"learning_rate": 9.724714385619673e-05,
"loss": 2.1736,
"step": 2920
},
{
"epoch": 0.19622287704259309,
"grad_norm": 0.0,
"learning_rate": 9.720876354478765e-05,
"loss": 2.2247,
"step": 2930
},
{
"epoch": 0.1968925796946156,
"grad_norm": 0.0,
"learning_rate": 9.717012521365836e-05,
"loss": 2.2102,
"step": 2940
},
{
"epoch": 0.19756228234663809,
"grad_norm": 0.0,
"learning_rate": 9.713122907398686e-05,
"loss": 2.275,
"step": 2950
},
{
"epoch": 0.1982319849986606,
"grad_norm": 0.0,
"learning_rate": 9.709207533836016e-05,
"loss": 2.0444,
"step": 2960
},
{
"epoch": 0.19890168765068308,
"grad_norm": 0.0,
"learning_rate": 9.70526642207732e-05,
"loss": 2.2226,
"step": 2970
},
{
"epoch": 0.1995713903027056,
"grad_norm": 0.0,
"learning_rate": 9.701299593662763e-05,
"loss": 2.2936,
"step": 2980
},
{
"epoch": 0.2002410929547281,
"grad_norm": 0.0,
"learning_rate": 9.697307070273062e-05,
"loss": 2.2361,
"step": 2990
},
{
"epoch": 0.2009107956067506,
"grad_norm": 0.0,
"learning_rate": 9.693288873729376e-05,
"loss": 2.1928,
"step": 3000
},
{
"epoch": 0.2015804982587731,
"grad_norm": 0.0,
"learning_rate": 9.689245025993175e-05,
"loss": 2.185,
"step": 3010
},
{
"epoch": 0.2022502009107956,
"grad_norm": 0.0,
"learning_rate": 9.68517554916613e-05,
"loss": 2.1397,
"step": 3020
},
{
"epoch": 0.2029199035628181,
"grad_norm": 0.0,
"learning_rate": 9.681080465489983e-05,
"loss": 2.3028,
"step": 3030
},
{
"epoch": 0.20358960621484062,
"grad_norm": 0.0,
"learning_rate": 9.676959797346435e-05,
"loss": 2.2324,
"step": 3040
},
{
"epoch": 0.2042593088668631,
"grad_norm": 0.0,
"learning_rate": 9.672813567257017e-05,
"loss": 2.2146,
"step": 3050
},
{
"epoch": 0.20492901151888562,
"grad_norm": 0.0,
"learning_rate": 9.668641797882969e-05,
"loss": 2.1497,
"step": 3060
},
{
"epoch": 0.2055987141709081,
"grad_norm": 0.0,
"learning_rate": 9.664444512025116e-05,
"loss": 2.2361,
"step": 3070
},
{
"epoch": 0.20626841682293062,
"grad_norm": 0.0,
"learning_rate": 9.660221732623744e-05,
"loss": 2.2336,
"step": 3080
},
{
"epoch": 0.2069381194749531,
"grad_norm": 0.0,
"learning_rate": 9.655973482758473e-05,
"loss": 2.2446,
"step": 3090
},
{
"epoch": 0.20760782212697562,
"grad_norm": 0.0,
"learning_rate": 9.651699785648135e-05,
"loss": 2.2724,
"step": 3100
},
{
"epoch": 0.20827752477899814,
"grad_norm": 0.0,
"learning_rate": 9.647400664650638e-05,
"loss": 2.2362,
"step": 3110
},
{
"epoch": 0.20894722743102062,
"grad_norm": 0.0,
"learning_rate": 9.643076143262851e-05,
"loss": 2.2481,
"step": 3120
},
{
"epoch": 0.20961693008304313,
"grad_norm": 0.0,
"learning_rate": 9.638726245120466e-05,
"loss": 2.1208,
"step": 3130
},
{
"epoch": 0.21028663273506562,
"grad_norm": 0.0,
"learning_rate": 9.634350993997871e-05,
"loss": 2.2085,
"step": 3140
},
{
"epoch": 0.21095633538708813,
"grad_norm": 0.0,
"learning_rate": 9.629950413808022e-05,
"loss": 2.2974,
"step": 3150
},
{
"epoch": 0.21162603803911065,
"grad_norm": 0.0,
"learning_rate": 9.62552452860231e-05,
"loss": 2.2781,
"step": 3160
},
{
"epoch": 0.21229574069113313,
"grad_norm": 0.0,
"learning_rate": 9.621073362570432e-05,
"loss": 2.3054,
"step": 3170
},
{
"epoch": 0.21296544334315565,
"grad_norm": 0.0,
"learning_rate": 9.616596940040257e-05,
"loss": 2.1255,
"step": 3180
},
{
"epoch": 0.21363514599517813,
"grad_norm": 0.0,
"learning_rate": 9.612095285477694e-05,
"loss": 2.1918,
"step": 3190
},
{
"epoch": 0.21430484864720065,
"grad_norm": 0.0,
"learning_rate": 9.607568423486558e-05,
"loss": 2.2317,
"step": 3200
},
{
"epoch": 0.21497455129922313,
"grad_norm": 0.0,
"learning_rate": 9.603016378808432e-05,
"loss": 2.0927,
"step": 3210
},
{
"epoch": 0.21564425395124565,
"grad_norm": 0.0,
"learning_rate": 9.59843917632254e-05,
"loss": 2.2694,
"step": 3220
},
{
"epoch": 0.21631395660326816,
"grad_norm": 0.0,
"learning_rate": 9.593836841045602e-05,
"loss": 2.2625,
"step": 3230
},
{
"epoch": 0.21698365925529065,
"grad_norm": 0.0,
"learning_rate": 9.589209398131706e-05,
"loss": 2.3238,
"step": 3240
},
{
"epoch": 0.21765336190731316,
"grad_norm": 0.0,
"learning_rate": 9.584556872872159e-05,
"loss": 2.2813,
"step": 3250
},
{
"epoch": 0.21832306455933564,
"grad_norm": 0.0,
"learning_rate": 9.579879290695364e-05,
"loss": 2.2298,
"step": 3260
},
{
"epoch": 0.21899276721135816,
"grad_norm": 0.0,
"learning_rate": 9.575176677166667e-05,
"loss": 2.2254,
"step": 3270
},
{
"epoch": 0.21966246986338067,
"grad_norm": 0.0,
"learning_rate": 9.570449057988222e-05,
"loss": 2.3366,
"step": 3280
},
{
"epoch": 0.22033217251540316,
"grad_norm": 0.0,
"learning_rate": 9.565696458998858e-05,
"loss": 2.2601,
"step": 3290
},
{
"epoch": 0.22100187516742567,
"grad_norm": 0.0,
"learning_rate": 9.560918906173923e-05,
"loss": 2.231,
"step": 3300
},
{
"epoch": 0.22167157781944816,
"grad_norm": 0.0,
"learning_rate": 9.556116425625159e-05,
"loss": 2.2668,
"step": 3310
},
{
"epoch": 0.22234128047147067,
"grad_norm": 0.0,
"learning_rate": 9.551289043600542e-05,
"loss": 2.2073,
"step": 3320
},
{
"epoch": 0.22301098312349316,
"grad_norm": 0.0,
"learning_rate": 9.546436786484155e-05,
"loss": 2.3307,
"step": 3330
},
{
"epoch": 0.22368068577551567,
"grad_norm": 0.0,
"learning_rate": 9.541559680796029e-05,
"loss": 2.2272,
"step": 3340
},
{
"epoch": 0.22435038842753818,
"grad_norm": 0.0,
"learning_rate": 9.536657753192011e-05,
"loss": 2.2672,
"step": 3350
},
{
"epoch": 0.22502009107956067,
"grad_norm": 0.0,
"learning_rate": 9.531731030463607e-05,
"loss": 2.2473,
"step": 3360
},
{
"epoch": 0.22568979373158318,
"grad_norm": 0.0,
"learning_rate": 9.526779539537845e-05,
"loss": 2.2187,
"step": 3370
},
{
"epoch": 0.22635949638360567,
"grad_norm": 0.0,
"learning_rate": 9.52180330747712e-05,
"loss": 2.3373,
"step": 3380
},
{
"epoch": 0.22702919903562818,
"grad_norm": 0.0,
"learning_rate": 9.516802361479056e-05,
"loss": 2.1831,
"step": 3390
},
{
"epoch": 0.2276989016876507,
"grad_norm": 0.0,
"learning_rate": 9.511776728876341e-05,
"loss": 2.1993,
"step": 3400
},
{
"epoch": 0.22836860433967318,
"grad_norm": 0.0,
"learning_rate": 9.506726437136599e-05,
"loss": 2.1827,
"step": 3410
},
{
"epoch": 0.2290383069916957,
"grad_norm": 0.0,
"learning_rate": 9.501651513862222e-05,
"loss": 2.2418,
"step": 3420
},
{
"epoch": 0.22970800964371818,
"grad_norm": 0.0,
"learning_rate": 9.496551986790225e-05,
"loss": 2.1535,
"step": 3430
},
{
"epoch": 0.2303777122957407,
"grad_norm": 0.0,
"learning_rate": 9.4914278837921e-05,
"loss": 2.37,
"step": 3440
},
{
"epoch": 0.23104741494776318,
"grad_norm": 0.0,
"learning_rate": 9.486279232873654e-05,
"loss": 2.1839,
"step": 3450
},
{
"epoch": 0.2317171175997857,
"grad_norm": 0.0,
"learning_rate": 9.481106062174863e-05,
"loss": 2.2899,
"step": 3460
},
{
"epoch": 0.2323868202518082,
"grad_norm": 0.0,
"learning_rate": 9.475908399969718e-05,
"loss": 2.2152,
"step": 3470
},
{
"epoch": 0.2330565229038307,
"grad_norm": 0.0,
"learning_rate": 9.470686274666065e-05,
"loss": 2.2035,
"step": 3480
},
{
"epoch": 0.2337262255558532,
"grad_norm": 0.0,
"learning_rate": 9.465439714805455e-05,
"loss": 2.2696,
"step": 3490
},
{
"epoch": 0.2343959282078757,
"grad_norm": 0.0,
"learning_rate": 9.460168749062985e-05,
"loss": 2.2141,
"step": 3500
},
{
"epoch": 0.2350656308598982,
"grad_norm": 0.0,
"learning_rate": 9.454873406247143e-05,
"loss": 2.2786,
"step": 3510
},
{
"epoch": 0.23573533351192072,
"grad_norm": 0.0,
"learning_rate": 9.449553715299652e-05,
"loss": 2.2578,
"step": 3520
},
{
"epoch": 0.2364050361639432,
"grad_norm": 0.0,
"learning_rate": 9.444209705295305e-05,
"loss": 2.2044,
"step": 3530
},
{
"epoch": 0.23707473881596572,
"grad_norm": 0.0,
"learning_rate": 9.438841405441816e-05,
"loss": 2.2658,
"step": 3540
},
{
"epoch": 0.2377444414679882,
"grad_norm": 0.0,
"learning_rate": 9.433448845079653e-05,
"loss": 2.2988,
"step": 3550
},
{
"epoch": 0.23841414412001072,
"grad_norm": 0.0,
"learning_rate": 9.42803205368188e-05,
"loss": 2.2158,
"step": 3560
},
{
"epoch": 0.2390838467720332,
"grad_norm": 0.0,
"learning_rate": 9.422591060853997e-05,
"loss": 2.3033,
"step": 3570
},
{
"epoch": 0.23975354942405572,
"grad_norm": 0.0,
"learning_rate": 9.417125896333774e-05,
"loss": 2.3241,
"step": 3580
},
{
"epoch": 0.24042325207607823,
"grad_norm": 0.0,
"learning_rate": 9.411636589991095e-05,
"loss": 2.2695,
"step": 3590
},
{
"epoch": 0.24109295472810072,
"grad_norm": 0.0,
"learning_rate": 9.406123171827789e-05,
"loss": 2.2711,
"step": 3600
},
{
"epoch": 0.24176265738012323,
"grad_norm": 0.0,
"learning_rate": 9.400585671977469e-05,
"loss": 2.1886,
"step": 3610
},
{
"epoch": 0.24243236003214572,
"grad_norm": 0.0,
"learning_rate": 9.395024120705367e-05,
"loss": 2.2764,
"step": 3620
},
{
"epoch": 0.24310206268416823,
"grad_norm": 0.0,
"learning_rate": 9.389438548408167e-05,
"loss": 2.1141,
"step": 3630
},
{
"epoch": 0.24377176533619074,
"grad_norm": 0.0,
"learning_rate": 9.383828985613843e-05,
"loss": 2.2824,
"step": 3640
},
{
"epoch": 0.24444146798821323,
"grad_norm": 0.0,
"learning_rate": 9.378195462981484e-05,
"loss": 2.2951,
"step": 3650
},
{
"epoch": 0.24511117064023574,
"grad_norm": 0.0,
"learning_rate": 9.372538011301135e-05,
"loss": 2.2363,
"step": 3660
},
{
"epoch": 0.24578087329225823,
"grad_norm": 0.0,
"learning_rate": 9.366856661493628e-05,
"loss": 2.2701,
"step": 3670
},
{
"epoch": 0.24645057594428074,
"grad_norm": 0.0,
"learning_rate": 9.361151444610404e-05,
"loss": 2.2683,
"step": 3680
},
{
"epoch": 0.24712027859630323,
"grad_norm": 0.0,
"learning_rate": 9.355422391833353e-05,
"loss": 2.3463,
"step": 3690
},
{
"epoch": 0.24778998124832574,
"grad_norm": 0.0,
"learning_rate": 9.349669534474641e-05,
"loss": 2.2302,
"step": 3700
},
{
"epoch": 0.24845968390034825,
"grad_norm": 0.0,
"learning_rate": 9.343892903976539e-05,
"loss": 2.2496,
"step": 3710
},
{
"epoch": 0.24912938655237074,
"grad_norm": 0.0,
"learning_rate": 9.338092531911245e-05,
"loss": 2.2797,
"step": 3720
},
{
"epoch": 0.24979908920439325,
"grad_norm": 0.0,
"learning_rate": 9.33226844998072e-05,
"loss": 2.2403,
"step": 3730
},
{
"epoch": 0.25046879185641574,
"grad_norm": 0.0,
"learning_rate": 9.326420690016513e-05,
"loss": 2.1988,
"step": 3740
},
{
"epoch": 0.2511384945084382,
"grad_norm": 0.0,
"learning_rate": 9.320549283979584e-05,
"loss": 2.249,
"step": 3750
},
{
"epoch": 0.25180819716046077,
"grad_norm": 0.0,
"learning_rate": 9.314654263960128e-05,
"loss": 2.2201,
"step": 3760
},
{
"epoch": 0.25247789981248325,
"grad_norm": 0.0,
"learning_rate": 9.308735662177407e-05,
"loss": 2.2789,
"step": 3770
},
{
"epoch": 0.25314760246450574,
"grad_norm": 0.0,
"learning_rate": 9.302793510979568e-05,
"loss": 2.2515,
"step": 3780
},
{
"epoch": 0.2538173051165283,
"grad_norm": 0.0,
"learning_rate": 9.296827842843463e-05,
"loss": 2.2134,
"step": 3790
},
{
"epoch": 0.25448700776855077,
"grad_norm": 0.0,
"learning_rate": 9.290838690374483e-05,
"loss": 2.2498,
"step": 3800
},
{
"epoch": 0.25515671042057325,
"grad_norm": 0.0,
"learning_rate": 9.284826086306366e-05,
"loss": 2.1802,
"step": 3810
},
{
"epoch": 0.2558264130725958,
"grad_norm": 0.0,
"learning_rate": 9.278790063501029e-05,
"loss": 2.2067,
"step": 3820
},
{
"epoch": 0.2564961157246183,
"grad_norm": 0.0,
"learning_rate": 9.272730654948384e-05,
"loss": 2.2824,
"step": 3830
},
{
"epoch": 0.25716581837664076,
"grad_norm": 0.0,
"learning_rate": 9.266647893766157e-05,
"loss": 2.2489,
"step": 3840
},
{
"epoch": 0.25783552102866325,
"grad_norm": 0.0,
"learning_rate": 9.260541813199706e-05,
"loss": 2.2607,
"step": 3850
},
{
"epoch": 0.2585052236806858,
"grad_norm": 0.0,
"learning_rate": 9.254412446621845e-05,
"loss": 2.3607,
"step": 3860
},
{
"epoch": 0.2591749263327083,
"grad_norm": 0.0,
"learning_rate": 9.248259827532656e-05,
"loss": 2.2326,
"step": 3870
},
{
"epoch": 0.25984462898473076,
"grad_norm": 0.0,
"learning_rate": 9.242083989559308e-05,
"loss": 2.2679,
"step": 3880
},
{
"epoch": 0.2605143316367533,
"grad_norm": 0.0,
"learning_rate": 9.235884966455872e-05,
"loss": 2.1288,
"step": 3890
},
{
"epoch": 0.2611840342887758,
"grad_norm": 0.0,
"learning_rate": 9.229662792103137e-05,
"loss": 2.2834,
"step": 3900
},
{
"epoch": 0.2618537369407983,
"grad_norm": 0.0,
"learning_rate": 9.223417500508427e-05,
"loss": 2.0785,
"step": 3910
},
{
"epoch": 0.26252343959282076,
"grad_norm": 0.0,
"learning_rate": 9.217149125805416e-05,
"loss": 2.2012,
"step": 3920
},
{
"epoch": 0.2631931422448433,
"grad_norm": 0.0,
"learning_rate": 9.21085770225393e-05,
"loss": 2.2774,
"step": 3930
},
{
"epoch": 0.2638628448968658,
"grad_norm": 0.0,
"learning_rate": 9.204543264239778e-05,
"loss": 2.2201,
"step": 3940
},
{
"epoch": 0.2645325475488883,
"grad_norm": 0.0,
"learning_rate": 9.198205846274548e-05,
"loss": 2.3057,
"step": 3950
},
{
"epoch": 0.2652022502009108,
"grad_norm": 0.0,
"learning_rate": 9.191845482995431e-05,
"loss": 2.2332,
"step": 3960
},
{
"epoch": 0.2658719528529333,
"grad_norm": 0.0,
"learning_rate": 9.185462209165021e-05,
"loss": 2.1952,
"step": 3970
},
{
"epoch": 0.2665416555049558,
"grad_norm": 0.0,
"learning_rate": 9.179056059671129e-05,
"loss": 2.3655,
"step": 3980
},
{
"epoch": 0.26721135815697833,
"grad_norm": 0.0,
"learning_rate": 9.1726270695266e-05,
"loss": 2.2544,
"step": 3990
},
{
"epoch": 0.2678810608090008,
"grad_norm": 0.0,
"learning_rate": 9.166175273869107e-05,
"loss": 2.2741,
"step": 4000
},
{
"epoch": 0.2685507634610233,
"grad_norm": 0.0,
"learning_rate": 9.15970070796097e-05,
"loss": 2.2881,
"step": 4010
},
{
"epoch": 0.2692204661130458,
"grad_norm": 0.0,
"learning_rate": 9.15320340718896e-05,
"loss": 2.2729,
"step": 4020
},
{
"epoch": 0.26989016876506833,
"grad_norm": 0.0,
"learning_rate": 9.146683407064105e-05,
"loss": 2.238,
"step": 4030
},
{
"epoch": 0.2705598714170908,
"grad_norm": 0.0,
"learning_rate": 9.140140743221496e-05,
"loss": 2.3336,
"step": 4040
},
{
"epoch": 0.2712295740691133,
"grad_norm": 0.0,
"learning_rate": 9.133575451420093e-05,
"loss": 2.2249,
"step": 4050
},
{
"epoch": 0.27189927672113584,
"grad_norm": 0.0,
"learning_rate": 9.126987567542532e-05,
"loss": 2.2385,
"step": 4060
},
{
"epoch": 0.2725689793731583,
"grad_norm": 0.0,
"learning_rate": 9.12037712759492e-05,
"loss": 2.2343,
"step": 4070
},
{
"epoch": 0.2732386820251808,
"grad_norm": 0.0,
"learning_rate": 9.113744167706648e-05,
"loss": 2.2202,
"step": 4080
},
{
"epoch": 0.2739083846772033,
"grad_norm": 0.0,
"learning_rate": 9.107088724130192e-05,
"loss": 2.2392,
"step": 4090
},
{
"epoch": 0.27457808732922584,
"grad_norm": 0.0,
"learning_rate": 9.100410833240908e-05,
"loss": 2.2371,
"step": 4100
},
{
"epoch": 0.2752477899812483,
"grad_norm": 0.0,
"learning_rate": 9.093710531536842e-05,
"loss": 2.3052,
"step": 4110
},
{
"epoch": 0.2759174926332708,
"grad_norm": 0.0,
"learning_rate": 9.086987855638525e-05,
"loss": 2.1853,
"step": 4120
},
{
"epoch": 0.27658719528529335,
"grad_norm": 0.0,
"learning_rate": 9.08024284228877e-05,
"loss": 2.2073,
"step": 4130
},
{
"epoch": 0.27725689793731584,
"grad_norm": 0.0,
"learning_rate": 9.073475528352484e-05,
"loss": 2.1914,
"step": 4140
},
{
"epoch": 0.2779266005893383,
"grad_norm": 0.0,
"learning_rate": 9.066685950816451e-05,
"loss": 2.2104,
"step": 4150
},
{
"epoch": 0.2785963032413608,
"grad_norm": 0.0,
"learning_rate": 9.059874146789139e-05,
"loss": 2.1314,
"step": 4160
},
{
"epoch": 0.27926600589338335,
"grad_norm": 0.0,
"learning_rate": 9.053040153500496e-05,
"loss": 2.1836,
"step": 4170
},
{
"epoch": 0.27993570854540584,
"grad_norm": 0.0,
"learning_rate": 9.046184008301743e-05,
"loss": 2.2723,
"step": 4180
},
{
"epoch": 0.2806054111974283,
"grad_norm": 0.0,
"learning_rate": 9.039305748665176e-05,
"loss": 2.2827,
"step": 4190
},
{
"epoch": 0.28127511384945086,
"grad_norm": 0.0,
"learning_rate": 9.032405412183956e-05,
"loss": 2.1977,
"step": 4200
},
{
"epoch": 0.28194481650147335,
"grad_norm": 0.0,
"learning_rate": 9.025483036571902e-05,
"loss": 2.2147,
"step": 4210
},
{
"epoch": 0.28261451915349584,
"grad_norm": 0.0,
"learning_rate": 9.018538659663293e-05,
"loss": 2.2783,
"step": 4220
},
{
"epoch": 0.2832842218055184,
"grad_norm": 0.0,
"learning_rate": 9.01157231941265e-05,
"loss": 2.1797,
"step": 4230
},
{
"epoch": 0.28395392445754086,
"grad_norm": 0.0,
"learning_rate": 9.004584053894545e-05,
"loss": 2.2945,
"step": 4240
},
{
"epoch": 0.28462362710956335,
"grad_norm": 0.0,
"learning_rate": 8.997573901303372e-05,
"loss": 2.2139,
"step": 4250
},
{
"epoch": 0.28529332976158583,
"grad_norm": 0.0,
"learning_rate": 8.990541899953151e-05,
"loss": 2.2501,
"step": 4260
},
{
"epoch": 0.2859630324136084,
"grad_norm": 0.0,
"learning_rate": 8.983488088277323e-05,
"loss": 2.233,
"step": 4270
},
{
"epoch": 0.28663273506563086,
"grad_norm": 0.0,
"learning_rate": 8.976412504828526e-05,
"loss": 2.2957,
"step": 4280
},
{
"epoch": 0.28730243771765335,
"grad_norm": 0.0,
"learning_rate": 8.969315188278396e-05,
"loss": 2.2512,
"step": 4290
},
{
"epoch": 0.2879721403696759,
"grad_norm": 0.0,
"learning_rate": 8.962196177417353e-05,
"loss": 2.1994,
"step": 4300
},
{
"epoch": 0.2886418430216984,
"grad_norm": 0.0,
"learning_rate": 8.955055511154378e-05,
"loss": 2.189,
"step": 4310
},
{
"epoch": 0.28931154567372086,
"grad_norm": 0.0,
"learning_rate": 8.947893228516821e-05,
"loss": 2.2342,
"step": 4320
},
{
"epoch": 0.28998124832574335,
"grad_norm": 0.0,
"learning_rate": 8.940709368650173e-05,
"loss": 2.1649,
"step": 4330
},
{
"epoch": 0.2906509509777659,
"grad_norm": 0.0,
"learning_rate": 8.933503970817849e-05,
"loss": 2.232,
"step": 4340
},
{
"epoch": 0.2913206536297884,
"grad_norm": 0.0,
"learning_rate": 8.926277074400987e-05,
"loss": 2.2633,
"step": 4350
},
{
"epoch": 0.29199035628181086,
"grad_norm": 0.0,
"learning_rate": 8.919028718898226e-05,
"loss": 2.2506,
"step": 4360
},
{
"epoch": 0.2926600589338334,
"grad_norm": 0.0,
"learning_rate": 8.911758943925483e-05,
"loss": 2.2919,
"step": 4370
},
{
"epoch": 0.2933297615858559,
"grad_norm": 0.0,
"learning_rate": 8.904467789215751e-05,
"loss": 2.2114,
"step": 4380
},
{
"epoch": 0.29399946423787837,
"grad_norm": 0.0,
"learning_rate": 8.897155294618869e-05,
"loss": 2.2352,
"step": 4390
},
{
"epoch": 0.29466916688990086,
"grad_norm": 0.0,
"learning_rate": 8.88982150010131e-05,
"loss": 2.2521,
"step": 4400
},
{
"epoch": 0.2953388695419234,
"grad_norm": 0.0,
"learning_rate": 8.882466445745964e-05,
"loss": 2.2084,
"step": 4410
},
{
"epoch": 0.2960085721939459,
"grad_norm": 0.0,
"learning_rate": 8.875090171751915e-05,
"loss": 2.2102,
"step": 4420
},
{
"epoch": 0.29667827484596837,
"grad_norm": 0.0,
"learning_rate": 8.867692718434223e-05,
"loss": 2.1578,
"step": 4430
},
{
"epoch": 0.2973479774979909,
"grad_norm": 0.0,
"learning_rate": 8.860274126223705e-05,
"loss": 2.2352,
"step": 4440
},
{
"epoch": 0.2980176801500134,
"grad_norm": 0.0,
"learning_rate": 8.85283443566671e-05,
"loss": 2.2075,
"step": 4450
},
{
"epoch": 0.2986873828020359,
"grad_norm": 0.0,
"learning_rate": 8.845373687424903e-05,
"loss": 2.3332,
"step": 4460
},
{
"epoch": 0.2993570854540584,
"grad_norm": 0.0,
"learning_rate": 8.83789192227504e-05,
"loss": 2.1775,
"step": 4470
},
{
"epoch": 0.3000267881060809,
"grad_norm": 0.0,
"learning_rate": 8.83038918110874e-05,
"loss": 2.2348,
"step": 4480
},
{
"epoch": 0.3006964907581034,
"grad_norm": 0.0,
"learning_rate": 8.822865504932275e-05,
"loss": 2.1897,
"step": 4490
},
{
"epoch": 0.3013661934101259,
"grad_norm": 0.0,
"learning_rate": 8.815320934866329e-05,
"loss": 2.2539,
"step": 4500
},
{
"epoch": 0.3020358960621484,
"grad_norm": 0.0,
"learning_rate": 8.807755512145788e-05,
"loss": 2.2351,
"step": 4510
},
{
"epoch": 0.3027055987141709,
"grad_norm": 0.0,
"learning_rate": 8.800169278119503e-05,
"loss": 2.3542,
"step": 4520
},
{
"epoch": 0.3033753013661934,
"grad_norm": 0.0,
"learning_rate": 8.792562274250075e-05,
"loss": 2.2724,
"step": 4530
},
{
"epoch": 0.30404500401821594,
"grad_norm": 0.0,
"learning_rate": 8.784934542113617e-05,
"loss": 2.204,
"step": 4540
},
{
"epoch": 0.3047147066702384,
"grad_norm": 0.0,
"learning_rate": 8.777286123399536e-05,
"loss": 2.2944,
"step": 4550
},
{
"epoch": 0.3053844093222609,
"grad_norm": 0.0,
"learning_rate": 8.7696170599103e-05,
"loss": 2.1318,
"step": 4560
},
{
"epoch": 0.3060541119742834,
"grad_norm": 0.0,
"learning_rate": 8.761927393561214e-05,
"loss": 2.1441,
"step": 4570
},
{
"epoch": 0.30672381462630594,
"grad_norm": 0.0,
"learning_rate": 8.754217166380184e-05,
"loss": 2.303,
"step": 4580
},
{
"epoch": 0.3073935172783284,
"grad_norm": 0.0,
"learning_rate": 8.746486420507491e-05,
"loss": 2.2769,
"step": 4590
},
{
"epoch": 0.3080632199303509,
"grad_norm": 0.0,
"learning_rate": 8.738735198195566e-05,
"loss": 2.3048,
"step": 4600
},
{
"epoch": 0.30873292258237345,
"grad_norm": 0.0,
"learning_rate": 8.73096354180875e-05,
"loss": 2.2582,
"step": 4610
},
{
"epoch": 0.30940262523439593,
"grad_norm": 0.0,
"learning_rate": 8.72317149382307e-05,
"loss": 2.3533,
"step": 4620
},
{
"epoch": 0.3100723278864184,
"grad_norm": 0.0,
"learning_rate": 8.715359096825999e-05,
"loss": 2.2417,
"step": 4630
},
{
"epoch": 0.3107420305384409,
"grad_norm": 0.0,
"learning_rate": 8.707526393516228e-05,
"loss": 2.215,
"step": 4640
},
{
"epoch": 0.31141173319046345,
"grad_norm": 0.0,
"learning_rate": 8.699673426703436e-05,
"loss": 2.1881,
"step": 4650
},
{
"epoch": 0.31208143584248593,
"grad_norm": 0.0,
"learning_rate": 8.691800239308052e-05,
"loss": 2.2907,
"step": 4660
},
{
"epoch": 0.3127511384945084,
"grad_norm": 0.0,
"learning_rate": 8.683906874361017e-05,
"loss": 2.221,
"step": 4670
},
{
"epoch": 0.31342084114653096,
"grad_norm": 0.0,
"learning_rate": 8.675993375003553e-05,
"loss": 2.1868,
"step": 4680
},
{
"epoch": 0.31409054379855345,
"grad_norm": 0.0,
"learning_rate": 8.668059784486929e-05,
"loss": 2.244,
"step": 4690
},
{
"epoch": 0.31476024645057593,
"grad_norm": 0.0,
"learning_rate": 8.660106146172223e-05,
"loss": 2.2294,
"step": 4700
},
{
"epoch": 0.3154299491025985,
"grad_norm": 0.0,
"learning_rate": 8.652132503530082e-05,
"loss": 2.2347,
"step": 4710
},
{
"epoch": 0.31609965175462096,
"grad_norm": 0.0,
"learning_rate": 8.644138900140485e-05,
"loss": 2.2633,
"step": 4720
},
{
"epoch": 0.31676935440664344,
"grad_norm": 0.0,
"learning_rate": 8.636125379692515e-05,
"loss": 2.2383,
"step": 4730
},
{
"epoch": 0.31743905705866593,
"grad_norm": 0.0,
"learning_rate": 8.628091985984099e-05,
"loss": 2.2282,
"step": 4740
},
{
"epoch": 0.31810875971068847,
"grad_norm": 0.0,
"learning_rate": 8.620038762921794e-05,
"loss": 2.172,
"step": 4750
},
{
"epoch": 0.31877846236271096,
"grad_norm": 0.0,
"learning_rate": 8.611965754520526e-05,
"loss": 2.2849,
"step": 4760
},
{
"epoch": 0.31944816501473344,
"grad_norm": 0.0,
"learning_rate": 8.60387300490336e-05,
"loss": 2.3451,
"step": 4770
},
{
"epoch": 0.320117867666756,
"grad_norm": 0.0,
"learning_rate": 8.595760558301257e-05,
"loss": 2.2314,
"step": 4780
},
{
"epoch": 0.32078757031877847,
"grad_norm": 0.0,
"learning_rate": 8.587628459052834e-05,
"loss": 2.155,
"step": 4790
},
{
"epoch": 0.32145727297080096,
"grad_norm": 0.0,
"learning_rate": 8.579476751604119e-05,
"loss": 2.2874,
"step": 4800
},
{
"epoch": 0.32212697562282344,
"grad_norm": 0.0,
"learning_rate": 8.571305480508302e-05,
"loss": 2.2094,
"step": 4810
},
{
"epoch": 0.322796678274846,
"grad_norm": 0.0,
"learning_rate": 8.563114690425511e-05,
"loss": 2.2631,
"step": 4820
},
{
"epoch": 0.32346638092686847,
"grad_norm": 0.0,
"learning_rate": 8.554904426122543e-05,
"loss": 2.287,
"step": 4830
},
{
"epoch": 0.32413608357889095,
"grad_norm": 0.0,
"learning_rate": 8.546674732472638e-05,
"loss": 2.2728,
"step": 4840
},
{
"epoch": 0.3248057862309135,
"grad_norm": 0.0,
"learning_rate": 8.538425654455225e-05,
"loss": 2.2106,
"step": 4850
},
{
"epoch": 0.325475488882936,
"grad_norm": 0.0,
"learning_rate": 8.530157237155681e-05,
"loss": 2.2396,
"step": 4860
},
{
"epoch": 0.32614519153495847,
"grad_norm": 0.0,
"learning_rate": 8.521869525765076e-05,
"loss": 2.2975,
"step": 4870
},
{
"epoch": 0.32681489418698095,
"grad_norm": 0.0,
"learning_rate": 8.513562565579936e-05,
"loss": 2.2552,
"step": 4880
},
{
"epoch": 0.3274845968390035,
"grad_norm": 0.0,
"learning_rate": 8.505236402001996e-05,
"loss": 2.2631,
"step": 4890
},
{
"epoch": 0.328154299491026,
"grad_norm": 0.0,
"learning_rate": 8.496891080537939e-05,
"loss": 2.1826,
"step": 4900
},
{
"epoch": 0.32882400214304847,
"grad_norm": 0.0,
"learning_rate": 8.488526646799158e-05,
"loss": 2.2127,
"step": 4910
},
{
"epoch": 0.329493704795071,
"grad_norm": 0.0,
"learning_rate": 8.480143146501506e-05,
"loss": 2.1991,
"step": 4920
},
{
"epoch": 0.3301634074470935,
"grad_norm": 0.0,
"learning_rate": 8.471740625465044e-05,
"loss": 2.2428,
"step": 4930
},
{
"epoch": 0.330833110099116,
"grad_norm": 0.0,
"learning_rate": 8.463319129613791e-05,
"loss": 2.2601,
"step": 4940
},
{
"epoch": 0.3315028127511385,
"grad_norm": 0.0,
"learning_rate": 8.45487870497547e-05,
"loss": 2.2814,
"step": 4950
},
{
"epoch": 0.332172515403161,
"grad_norm": 0.0,
"learning_rate": 8.446419397681265e-05,
"loss": 2.1667,
"step": 4960
},
{
"epoch": 0.3328422180551835,
"grad_norm": 0.0,
"learning_rate": 8.437941253965558e-05,
"loss": 2.3395,
"step": 4970
},
{
"epoch": 0.333511920707206,
"grad_norm": 0.0,
"learning_rate": 8.429444320165683e-05,
"loss": 2.2234,
"step": 4980
},
{
"epoch": 0.3341816233592285,
"grad_norm": 0.0,
"learning_rate": 8.420928642721672e-05,
"loss": 2.3486,
"step": 4990
},
{
"epoch": 0.334851326011251,
"grad_norm": 0.0,
"learning_rate": 8.412394268176003e-05,
"loss": 2.2212,
"step": 5000
},
{
"epoch": 0.3355210286632735,
"grad_norm": 0.0,
"learning_rate": 8.403841243173338e-05,
"loss": 2.2387,
"step": 5010
},
{
"epoch": 0.33619073131529603,
"grad_norm": 0.0,
"learning_rate": 8.395269614460275e-05,
"loss": 2.1898,
"step": 5020
},
{
"epoch": 0.3368604339673185,
"grad_norm": 0.0,
"learning_rate": 8.386679428885092e-05,
"loss": 2.245,
"step": 5030
},
{
"epoch": 0.337530136619341,
"grad_norm": 0.0,
"learning_rate": 8.37807073339749e-05,
"loss": 2.2294,
"step": 5040
},
{
"epoch": 0.3381998392713635,
"grad_norm": 0.0,
"learning_rate": 8.369443575048332e-05,
"loss": 2.2182,
"step": 5050
},
{
"epoch": 0.33886954192338603,
"grad_norm": 0.0,
"learning_rate": 8.360798000989394e-05,
"loss": 2.2475,
"step": 5060
},
{
"epoch": 0.3395392445754085,
"grad_norm": 0.0,
"learning_rate": 8.352134058473106e-05,
"loss": 2.2286,
"step": 5070
},
{
"epoch": 0.340208947227431,
"grad_norm": 0.0,
"learning_rate": 8.343451794852282e-05,
"loss": 2.2071,
"step": 5080
},
{
"epoch": 0.34087864987945354,
"grad_norm": 0.0,
"learning_rate": 8.334751257579874e-05,
"loss": 2.2365,
"step": 5090
},
{
"epoch": 0.34154835253147603,
"grad_norm": 0.0,
"learning_rate": 8.326032494208713e-05,
"loss": 2.3365,
"step": 5100
},
{
"epoch": 0.3422180551834985,
"grad_norm": 0.0,
"learning_rate": 8.31729555239124e-05,
"loss": 2.2691,
"step": 5110
},
{
"epoch": 0.342887757835521,
"grad_norm": 0.0,
"learning_rate": 8.308540479879252e-05,
"loss": 2.1833,
"step": 5120
},
{
"epoch": 0.34355746048754354,
"grad_norm": 0.0,
"learning_rate": 8.299767324523638e-05,
"loss": 2.1165,
"step": 5130
},
{
"epoch": 0.34422716313956603,
"grad_norm": 0.0,
"learning_rate": 8.290976134274123e-05,
"loss": 2.3461,
"step": 5140
},
{
"epoch": 0.3448968657915885,
"grad_norm": 0.0,
"learning_rate": 8.282166957178995e-05,
"loss": 2.2149,
"step": 5150
},
{
"epoch": 0.34556656844361106,
"grad_norm": 0.0,
"learning_rate": 8.273339841384855e-05,
"loss": 2.2644,
"step": 5160
},
{
"epoch": 0.34623627109563354,
"grad_norm": 0.0,
"learning_rate": 8.264494835136347e-05,
"loss": 2.2684,
"step": 5170
},
{
"epoch": 0.346905973747656,
"grad_norm": 0.0,
"learning_rate": 8.255631986775894e-05,
"loss": 2.2436,
"step": 5180
},
{
"epoch": 0.34757567639967857,
"grad_norm": 0.0,
"learning_rate": 8.246751344743433e-05,
"loss": 2.1691,
"step": 5190
},
{
"epoch": 0.34824537905170105,
"grad_norm": 0.0,
"learning_rate": 8.237852957576158e-05,
"loss": 2.2467,
"step": 5200
},
{
"epoch": 0.34891508170372354,
"grad_norm": 0.0,
"learning_rate": 8.228936873908244e-05,
"loss": 2.2633,
"step": 5210
},
{
"epoch": 0.349584784355746,
"grad_norm": 0.0,
"learning_rate": 8.220003142470592e-05,
"loss": 2.247,
"step": 5220
},
{
"epoch": 0.35025448700776857,
"grad_norm": 0.0,
"learning_rate": 8.211051812090548e-05,
"loss": 2.3088,
"step": 5230
},
{
"epoch": 0.35092418965979105,
"grad_norm": 0.0,
"learning_rate": 8.202082931691655e-05,
"loss": 2.2103,
"step": 5240
},
{
"epoch": 0.35159389231181354,
"grad_norm": 0.0,
"learning_rate": 8.193096550293369e-05,
"loss": 2.0959,
"step": 5250
},
{
"epoch": 0.3522635949638361,
"grad_norm": 0.0,
"learning_rate": 8.184092717010801e-05,
"loss": 2.2747,
"step": 5260
},
{
"epoch": 0.35293329761585857,
"grad_norm": 0.0,
"learning_rate": 8.175071481054444e-05,
"loss": 2.2294,
"step": 5270
},
{
"epoch": 0.35360300026788105,
"grad_norm": 0.0,
"learning_rate": 8.166032891729905e-05,
"loss": 2.1772,
"step": 5280
},
{
"epoch": 0.35427270291990354,
"grad_norm": 0.0,
"learning_rate": 8.15697699843764e-05,
"loss": 2.2499,
"step": 5290
},
{
"epoch": 0.3549424055719261,
"grad_norm": 0.0,
"learning_rate": 8.147903850672671e-05,
"loss": 2.2,
"step": 5300
},
{
"epoch": 0.35561210822394856,
"grad_norm": 0.0,
"learning_rate": 8.138813498024332e-05,
"loss": 2.2583,
"step": 5310
},
{
"epoch": 0.35628181087597105,
"grad_norm": 0.0,
"learning_rate": 8.129705990175991e-05,
"loss": 2.2661,
"step": 5320
},
{
"epoch": 0.3569515135279936,
"grad_norm": 0.0,
"learning_rate": 8.120581376904773e-05,
"loss": 2.2114,
"step": 5330
},
{
"epoch": 0.3576212161800161,
"grad_norm": 0.0,
"learning_rate": 8.111439708081297e-05,
"loss": 2.2011,
"step": 5340
},
{
"epoch": 0.35829091883203856,
"grad_norm": 0.0,
"learning_rate": 8.102281033669393e-05,
"loss": 2.252,
"step": 5350
},
{
"epoch": 0.3589606214840611,
"grad_norm": 0.0,
"learning_rate": 8.093105403725842e-05,
"loss": 2.1967,
"step": 5360
},
{
"epoch": 0.3596303241360836,
"grad_norm": 0.0,
"learning_rate": 8.083912868400094e-05,
"loss": 2.2036,
"step": 5370
},
{
"epoch": 0.3603000267881061,
"grad_norm": 0.0,
"learning_rate": 8.074703477933991e-05,
"loss": 2.2294,
"step": 5380
},
{
"epoch": 0.36096972944012856,
"grad_norm": 0.0,
"learning_rate": 8.065477282661504e-05,
"loss": 2.2663,
"step": 5390
},
{
"epoch": 0.3616394320921511,
"grad_norm": 0.0,
"learning_rate": 8.056234333008445e-05,
"loss": 2.2182,
"step": 5400
},
{
"epoch": 0.3623091347441736,
"grad_norm": 0.0,
"learning_rate": 8.046974679492197e-05,
"loss": 2.2519,
"step": 5410
},
{
"epoch": 0.3629788373961961,
"grad_norm": 0.0,
"learning_rate": 8.037698372721442e-05,
"loss": 2.1965,
"step": 5420
},
{
"epoch": 0.3636485400482186,
"grad_norm": 0.0,
"learning_rate": 8.028405463395878e-05,
"loss": 2.1936,
"step": 5430
},
{
"epoch": 0.3643182427002411,
"grad_norm": 0.0,
"learning_rate": 8.019096002305946e-05,
"loss": 2.2061,
"step": 5440
},
{
"epoch": 0.3649879453522636,
"grad_norm": 0.0,
"learning_rate": 8.00977004033255e-05,
"loss": 2.2627,
"step": 5450
},
{
"epoch": 0.3656576480042861,
"grad_norm": 0.0,
"learning_rate": 8.000427628446776e-05,
"loss": 2.3286,
"step": 5460
},
{
"epoch": 0.3663273506563086,
"grad_norm": 0.0,
"learning_rate": 7.991068817709624e-05,
"loss": 2.1929,
"step": 5470
},
{
"epoch": 0.3669970533083311,
"grad_norm": 0.0,
"learning_rate": 7.981693659271716e-05,
"loss": 2.2085,
"step": 5480
},
{
"epoch": 0.3676667559603536,
"grad_norm": 0.0,
"learning_rate": 7.972302204373024e-05,
"loss": 2.2468,
"step": 5490
},
{
"epoch": 0.3683364586123761,
"grad_norm": 0.0,
"learning_rate": 7.962894504342591e-05,
"loss": 2.2503,
"step": 5500
},
{
"epoch": 0.3690061612643986,
"grad_norm": 0.0,
"learning_rate": 7.953470610598244e-05,
"loss": 2.2457,
"step": 5510
},
{
"epoch": 0.3696758639164211,
"grad_norm": 0.0,
"learning_rate": 7.944030574646323e-05,
"loss": 2.2607,
"step": 5520
},
{
"epoch": 0.3703455665684436,
"grad_norm": 0.0,
"learning_rate": 7.934574448081385e-05,
"loss": 2.4296,
"step": 5530
},
{
"epoch": 0.3710152692204661,
"grad_norm": 0.0,
"learning_rate": 7.925102282585936e-05,
"loss": 2.2071,
"step": 5540
},
{
"epoch": 0.3716849718724886,
"grad_norm": 0.0,
"learning_rate": 7.91561412993014e-05,
"loss": 2.2177,
"step": 5550
},
{
"epoch": 0.3723546745245111,
"grad_norm": 0.0,
"learning_rate": 7.906110041971541e-05,
"loss": 2.2693,
"step": 5560
},
{
"epoch": 0.37302437717653364,
"grad_norm": 0.0,
"learning_rate": 7.896590070654777e-05,
"loss": 2.1922,
"step": 5570
},
{
"epoch": 0.3736940798285561,
"grad_norm": 0.0,
"learning_rate": 7.887054268011297e-05,
"loss": 2.216,
"step": 5580
},
{
"epoch": 0.3743637824805786,
"grad_norm": 0.0,
"learning_rate": 7.877502686159074e-05,
"loss": 2.2372,
"step": 5590
},
{
"epoch": 0.37503348513260115,
"grad_norm": 0.0,
"learning_rate": 7.867935377302324e-05,
"loss": 2.2107,
"step": 5600
},
{
"epoch": 0.37570318778462364,
"grad_norm": 0.0,
"learning_rate": 7.858352393731219e-05,
"loss": 2.2494,
"step": 5610
},
{
"epoch": 0.3763728904366461,
"grad_norm": 0.0,
"learning_rate": 7.8487537878216e-05,
"loss": 2.1675,
"step": 5620
},
{
"epoch": 0.3770425930886686,
"grad_norm": 0.0,
"learning_rate": 7.839139612034695e-05,
"loss": 2.2335,
"step": 5630
},
{
"epoch": 0.37771229574069115,
"grad_norm": 0.0,
"learning_rate": 7.829509918916825e-05,
"loss": 2.2172,
"step": 5640
},
{
"epoch": 0.37838199839271364,
"grad_norm": 0.0,
"learning_rate": 7.819864761099125e-05,
"loss": 2.2317,
"step": 5650
},
{
"epoch": 0.3790517010447361,
"grad_norm": 0.0,
"learning_rate": 7.81020419129725e-05,
"loss": 2.2232,
"step": 5660
},
{
"epoch": 0.37972140369675866,
"grad_norm": 0.0,
"learning_rate": 7.800528262311089e-05,
"loss": 2.2933,
"step": 5670
},
{
"epoch": 0.38039110634878115,
"grad_norm": 0.0,
"learning_rate": 7.790837027024478e-05,
"loss": 2.1793,
"step": 5680
},
{
"epoch": 0.38106080900080364,
"grad_norm": 0.0,
"learning_rate": 7.781130538404911e-05,
"loss": 2.1571,
"step": 5690
},
{
"epoch": 0.3817305116528261,
"grad_norm": 0.0,
"learning_rate": 7.771408849503245e-05,
"loss": 2.1802,
"step": 5700
},
{
"epoch": 0.38240021430484866,
"grad_norm": 0.0,
"learning_rate": 7.761672013453418e-05,
"loss": 2.247,
"step": 5710
},
{
"epoch": 0.38306991695687115,
"grad_norm": 0.0,
"learning_rate": 7.751920083472153e-05,
"loss": 2.2638,
"step": 5720
},
{
"epoch": 0.38373961960889363,
"grad_norm": 0.0,
"learning_rate": 7.74215311285867e-05,
"loss": 2.2909,
"step": 5730
},
{
"epoch": 0.3844093222609162,
"grad_norm": 0.0,
"learning_rate": 7.732371154994392e-05,
"loss": 2.1795,
"step": 5740
},
{
"epoch": 0.38507902491293866,
"grad_norm": 0.0,
"learning_rate": 7.722574263342656e-05,
"loss": 2.2015,
"step": 5750
},
{
"epoch": 0.38574872756496115,
"grad_norm": 0.0,
"learning_rate": 7.71276249144842e-05,
"loss": 2.2313,
"step": 5760
},
{
"epoch": 0.38641843021698363,
"grad_norm": 0.0,
"learning_rate": 7.70293589293797e-05,
"loss": 2.2416,
"step": 5770
},
{
"epoch": 0.3870881328690062,
"grad_norm": 0.0,
"learning_rate": 7.693094521518627e-05,
"loss": 2.2684,
"step": 5780
},
{
"epoch": 0.38775783552102866,
"grad_norm": 0.0,
"learning_rate": 7.683238430978452e-05,
"loss": 2.237,
"step": 5790
},
{
"epoch": 0.38842753817305115,
"grad_norm": 0.0,
"learning_rate": 7.673367675185953e-05,
"loss": 2.053,
"step": 5800
},
{
"epoch": 0.3890972408250737,
"grad_norm": 0.0,
"learning_rate": 7.663482308089795e-05,
"loss": 2.228,
"step": 5810
},
{
"epoch": 0.3897669434770962,
"grad_norm": 0.0,
"learning_rate": 7.653582383718498e-05,
"loss": 2.1679,
"step": 5820
},
{
"epoch": 0.39043664612911866,
"grad_norm": 0.0,
"learning_rate": 7.643667956180146e-05,
"loss": 2.2186,
"step": 5830
},
{
"epoch": 0.3911063487811412,
"grad_norm": 0.0,
"learning_rate": 7.633739079662088e-05,
"loss": 2.2065,
"step": 5840
},
{
"epoch": 0.3917760514331637,
"grad_norm": 0.0,
"learning_rate": 7.623795808430645e-05,
"loss": 2.2311,
"step": 5850
},
{
"epoch": 0.39244575408518617,
"grad_norm": 0.0,
"learning_rate": 7.613838196830816e-05,
"loss": 2.222,
"step": 5860
},
{
"epoch": 0.39311545673720866,
"grad_norm": 0.0,
"learning_rate": 7.603866299285972e-05,
"loss": 2.2572,
"step": 5870
},
{
"epoch": 0.3937851593892312,
"grad_norm": 0.0,
"learning_rate": 7.593880170297564e-05,
"loss": 2.209,
"step": 5880
},
{
"epoch": 0.3944548620412537,
"grad_norm": 0.0,
"learning_rate": 7.583879864444832e-05,
"loss": 2.259,
"step": 5890
},
{
"epoch": 0.39512456469327617,
"grad_norm": 0.0,
"learning_rate": 7.573865436384491e-05,
"loss": 2.1778,
"step": 5900
},
{
"epoch": 0.3957942673452987,
"grad_norm": 0.0,
"learning_rate": 7.563836940850449e-05,
"loss": 2.2173,
"step": 5910
},
{
"epoch": 0.3964639699973212,
"grad_norm": 0.0,
"learning_rate": 7.55379443265349e-05,
"loss": 2.2584,
"step": 5920
},
{
"epoch": 0.3971336726493437,
"grad_norm": 0.0,
"learning_rate": 7.543737966680994e-05,
"loss": 2.2348,
"step": 5930
},
{
"epoch": 0.39780337530136617,
"grad_norm": 0.0,
"learning_rate": 7.533667597896623e-05,
"loss": 2.1596,
"step": 5940
},
{
"epoch": 0.3984730779533887,
"grad_norm": 0.0,
"learning_rate": 7.523583381340025e-05,
"loss": 2.2926,
"step": 5950
},
{
"epoch": 0.3991427806054112,
"grad_norm": 0.0,
"learning_rate": 7.513485372126531e-05,
"loss": 2.2573,
"step": 5960
},
{
"epoch": 0.3998124832574337,
"grad_norm": 0.0,
"learning_rate": 7.503373625446862e-05,
"loss": 2.2771,
"step": 5970
},
{
"epoch": 0.4004821859094562,
"grad_norm": 0.0,
"learning_rate": 7.493248196566816e-05,
"loss": 2.2364,
"step": 5980
},
{
"epoch": 0.4011518885614787,
"grad_norm": 0.0,
"learning_rate": 7.483109140826968e-05,
"loss": 2.163,
"step": 5990
},
{
"epoch": 0.4018215912135012,
"grad_norm": 0.0,
"learning_rate": 7.472956513642379e-05,
"loss": 2.1557,
"step": 6000
},
{
"epoch": 0.4024912938655237,
"grad_norm": 0.0,
"learning_rate": 7.462790370502284e-05,
"loss": 2.2514,
"step": 6010
},
{
"epoch": 0.4031609965175462,
"grad_norm": 0.0,
"learning_rate": 7.452610766969781e-05,
"loss": 2.1986,
"step": 6020
},
{
"epoch": 0.4038306991695687,
"grad_norm": 0.0,
"learning_rate": 7.442417758681542e-05,
"loss": 2.2523,
"step": 6030
},
{
"epoch": 0.4045004018215912,
"grad_norm": 0.0,
"learning_rate": 7.432211401347504e-05,
"loss": 2.3855,
"step": 6040
},
{
"epoch": 0.40517010447361373,
"grad_norm": 0.0,
"learning_rate": 7.421991750750559e-05,
"loss": 2.1914,
"step": 6050
},
{
"epoch": 0.4058398071256362,
"grad_norm": 0.0,
"learning_rate": 7.411758862746258e-05,
"loss": 2.122,
"step": 6060
},
{
"epoch": 0.4065095097776587,
"grad_norm": 0.0,
"learning_rate": 7.401512793262496e-05,
"loss": 2.1607,
"step": 6070
},
{
"epoch": 0.40717921242968125,
"grad_norm": 0.0,
"learning_rate": 7.391253598299217e-05,
"loss": 2.1664,
"step": 6080
},
{
"epoch": 0.40784891508170373,
"grad_norm": 0.0,
"learning_rate": 7.380981333928097e-05,
"loss": 2.2631,
"step": 6090
},
{
"epoch": 0.4085186177337262,
"grad_norm": 0.0,
"learning_rate": 7.370696056292249e-05,
"loss": 2.2071,
"step": 6100
},
{
"epoch": 0.4091883203857487,
"grad_norm": 0.0,
"learning_rate": 7.360397821605902e-05,
"loss": 2.2336,
"step": 6110
},
{
"epoch": 0.40985802303777125,
"grad_norm": 0.0,
"learning_rate": 7.350086686154111e-05,
"loss": 2.2513,
"step": 6120
},
{
"epoch": 0.41052772568979373,
"grad_norm": 0.0,
"learning_rate": 7.33976270629243e-05,
"loss": 2.2542,
"step": 6130
},
{
"epoch": 0.4111974283418162,
"grad_norm": 0.0,
"learning_rate": 7.329425938446625e-05,
"loss": 2.2058,
"step": 6140
},
{
"epoch": 0.41186713099383876,
"grad_norm": 0.0,
"learning_rate": 7.319076439112347e-05,
"loss": 2.2086,
"step": 6150
},
{
"epoch": 0.41253683364586125,
"grad_norm": 0.0,
"learning_rate": 7.308714264854833e-05,
"loss": 2.1696,
"step": 6160
},
{
"epoch": 0.41320653629788373,
"grad_norm": 0.0,
"learning_rate": 7.298339472308598e-05,
"loss": 2.2503,
"step": 6170
},
{
"epoch": 0.4138762389499062,
"grad_norm": 0.0,
"learning_rate": 7.287952118177117e-05,
"loss": 2.1965,
"step": 6180
},
{
"epoch": 0.41454594160192876,
"grad_norm": 0.0,
"learning_rate": 7.277552259232522e-05,
"loss": 2.1826,
"step": 6190
},
{
"epoch": 0.41521564425395124,
"grad_norm": 0.0,
"learning_rate": 7.267139952315295e-05,
"loss": 2.1862,
"step": 6200
},
{
"epoch": 0.41588534690597373,
"grad_norm": 0.0,
"learning_rate": 7.256715254333946e-05,
"loss": 2.1264,
"step": 6210
},
{
"epoch": 0.41655504955799627,
"grad_norm": 0.0,
"learning_rate": 7.246278222264713e-05,
"loss": 2.2027,
"step": 6220
},
{
"epoch": 0.41722475221001876,
"grad_norm": 0.0,
"learning_rate": 7.235828913151242e-05,
"loss": 2.2236,
"step": 6230
},
{
"epoch": 0.41789445486204124,
"grad_norm": 0.0,
"learning_rate": 7.225367384104282e-05,
"loss": 2.2782,
"step": 6240
},
{
"epoch": 0.41856415751406373,
"grad_norm": 0.0,
"learning_rate": 7.21489369230137e-05,
"loss": 2.2659,
"step": 6250
},
{
"epoch": 0.41923386016608627,
"grad_norm": 0.0,
"learning_rate": 7.204407894986518e-05,
"loss": 2.1987,
"step": 6260
},
{
"epoch": 0.41990356281810876,
"grad_norm": 0.0,
"learning_rate": 7.1939100494699e-05,
"loss": 2.2765,
"step": 6270
},
{
"epoch": 0.42057326547013124,
"grad_norm": 0.0,
"learning_rate": 7.183400213127543e-05,
"loss": 2.1841,
"step": 6280
},
{
"epoch": 0.4212429681221538,
"grad_norm": 0.0,
"learning_rate": 7.172878443401003e-05,
"loss": 2.2169,
"step": 6290
},
{
"epoch": 0.42191267077417627,
"grad_norm": 0.0,
"learning_rate": 7.162344797797065e-05,
"loss": 2.1275,
"step": 6300
},
{
"epoch": 0.42258237342619875,
"grad_norm": 0.0,
"learning_rate": 7.151799333887415e-05,
"loss": 2.1387,
"step": 6310
},
{
"epoch": 0.4232520760782213,
"grad_norm": 0.0,
"learning_rate": 7.141242109308343e-05,
"loss": 2.3116,
"step": 6320
},
{
"epoch": 0.4239217787302438,
"grad_norm": 0.0,
"learning_rate": 7.130673181760403e-05,
"loss": 2.2473,
"step": 6330
},
{
"epoch": 0.42459148138226627,
"grad_norm": 0.0,
"learning_rate": 7.120092609008122e-05,
"loss": 2.2514,
"step": 6340
},
{
"epoch": 0.42526118403428875,
"grad_norm": 0.0,
"learning_rate": 7.10950044887967e-05,
"loss": 2.2596,
"step": 6350
},
{
"epoch": 0.4259308866863113,
"grad_norm": 0.0,
"learning_rate": 7.098896759266547e-05,
"loss": 2.1879,
"step": 6360
},
{
"epoch": 0.4266005893383338,
"grad_norm": 0.0,
"learning_rate": 7.088281598123267e-05,
"loss": 2.1676,
"step": 6370
},
{
"epoch": 0.42727029199035627,
"grad_norm": 0.0,
"learning_rate": 7.077655023467049e-05,
"loss": 2.2725,
"step": 6380
},
{
"epoch": 0.4279399946423788,
"grad_norm": 0.0,
"learning_rate": 7.067017093377484e-05,
"loss": 2.1804,
"step": 6390
},
{
"epoch": 0.4286096972944013,
"grad_norm": 0.0,
"learning_rate": 7.056367865996226e-05,
"loss": 2.2629,
"step": 6400
},
{
"epoch": 0.4292793999464238,
"grad_norm": 0.0,
"learning_rate": 7.045707399526682e-05,
"loss": 2.1842,
"step": 6410
},
{
"epoch": 0.42994910259844626,
"grad_norm": 0.0,
"learning_rate": 7.035035752233681e-05,
"loss": 2.217,
"step": 6420
},
{
"epoch": 0.4306188052504688,
"grad_norm": 0.0,
"learning_rate": 7.024352982443159e-05,
"loss": 2.2961,
"step": 6430
},
{
"epoch": 0.4312885079024913,
"grad_norm": 0.0,
"learning_rate": 7.013659148541849e-05,
"loss": 2.2467,
"step": 6440
},
{
"epoch": 0.4319582105545138,
"grad_norm": 0.0,
"learning_rate": 7.002954308976948e-05,
"loss": 2.3547,
"step": 6450
},
{
"epoch": 0.4326279132065363,
"grad_norm": 0.0,
"learning_rate": 6.992238522255805e-05,
"loss": 2.2718,
"step": 6460
},
{
"epoch": 0.4332976158585588,
"grad_norm": 0.0,
"learning_rate": 6.981511846945608e-05,
"loss": 2.2483,
"step": 6470
},
{
"epoch": 0.4339673185105813,
"grad_norm": 0.0,
"learning_rate": 6.97077434167305e-05,
"loss": 2.2501,
"step": 6480
},
{
"epoch": 0.43463702116260383,
"grad_norm": 0.0,
"learning_rate": 6.960026065124013e-05,
"loss": 2.3018,
"step": 6490
},
{
"epoch": 0.4353067238146263,
"grad_norm": 0.0,
"learning_rate": 6.94926707604326e-05,
"loss": 2.2755,
"step": 6500
},
{
"epoch": 0.4359764264666488,
"grad_norm": 0.0,
"learning_rate": 6.938497433234091e-05,
"loss": 2.2152,
"step": 6510
},
{
"epoch": 0.4366461291186713,
"grad_norm": 0.0,
"learning_rate": 6.927717195558041e-05,
"loss": 2.2302,
"step": 6520
},
{
"epoch": 0.43731583177069383,
"grad_norm": 0.0,
"learning_rate": 6.916926421934553e-05,
"loss": 2.2201,
"step": 6530
},
{
"epoch": 0.4379855344227163,
"grad_norm": 0.0,
"learning_rate": 6.906125171340646e-05,
"loss": 2.2863,
"step": 6540
},
{
"epoch": 0.4386552370747388,
"grad_norm": 0.0,
"learning_rate": 6.89531350281061e-05,
"loss": 2.2234,
"step": 6550
},
{
"epoch": 0.43932493972676134,
"grad_norm": 0.0,
"learning_rate": 6.88449147543567e-05,
"loss": 2.1963,
"step": 6560
},
{
"epoch": 0.43999464237878383,
"grad_norm": 0.0,
"learning_rate": 6.873659148363667e-05,
"loss": 2.2119,
"step": 6570
},
{
"epoch": 0.4406643450308063,
"grad_norm": 0.0,
"learning_rate": 6.862816580798734e-05,
"loss": 2.2322,
"step": 6580
},
{
"epoch": 0.4413340476828288,
"grad_norm": 0.0,
"learning_rate": 6.85196383200098e-05,
"loss": 2.2327,
"step": 6590
},
{
"epoch": 0.44200375033485134,
"grad_norm": 0.0,
"learning_rate": 6.841100961286151e-05,
"loss": 2.2449,
"step": 6600
},
{
"epoch": 0.4426734529868738,
"grad_norm": 0.0,
"learning_rate": 6.830228028025319e-05,
"loss": 2.1634,
"step": 6610
},
{
"epoch": 0.4433431556388963,
"grad_norm": 0.0,
"learning_rate": 6.819345091644552e-05,
"loss": 2.2752,
"step": 6620
},
{
"epoch": 0.44401285829091885,
"grad_norm": 0.0,
"learning_rate": 6.80845221162459e-05,
"loss": 2.2577,
"step": 6630
},
{
"epoch": 0.44468256094294134,
"grad_norm": 0.0,
"learning_rate": 6.797549447500522e-05,
"loss": 2.2371,
"step": 6640
},
{
"epoch": 0.4453522635949638,
"grad_norm": 0.0,
"learning_rate": 6.786636858861456e-05,
"loss": 2.1673,
"step": 6650
},
{
"epoch": 0.4460219662469863,
"grad_norm": 0.0,
"learning_rate": 6.775714505350195e-05,
"loss": 2.162,
"step": 6660
},
{
"epoch": 0.44669166889900885,
"grad_norm": 0.0,
"learning_rate": 6.764782446662915e-05,
"loss": 2.3001,
"step": 6670
},
{
"epoch": 0.44736137155103134,
"grad_norm": 0.0,
"learning_rate": 6.753840742548836e-05,
"loss": 2.2371,
"step": 6680
},
{
"epoch": 0.4480310742030538,
"grad_norm": 0.0,
"learning_rate": 6.742889452809891e-05,
"loss": 2.214,
"step": 6690
},
{
"epoch": 0.44870077685507637,
"grad_norm": 0.0,
"learning_rate": 6.731928637300407e-05,
"loss": 2.1805,
"step": 6700
},
{
"epoch": 0.44937047950709885,
"grad_norm": 0.0,
"learning_rate": 6.720958355926773e-05,
"loss": 2.2062,
"step": 6710
},
{
"epoch": 0.45004018215912134,
"grad_norm": 0.0,
"learning_rate": 6.709978668647112e-05,
"loss": 2.2897,
"step": 6720
},
{
"epoch": 0.4507098848111439,
"grad_norm": 0.0,
"learning_rate": 6.698989635470959e-05,
"loss": 2.2066,
"step": 6730
},
{
"epoch": 0.45137958746316637,
"grad_norm": 0.0,
"learning_rate": 6.687991316458927e-05,
"loss": 2.1965,
"step": 6740
},
{
"epoch": 0.45204929011518885,
"grad_norm": 0.0,
"learning_rate": 6.676983771722379e-05,
"loss": 2.2724,
"step": 6750
},
{
"epoch": 0.45271899276721134,
"grad_norm": 0.0,
"learning_rate": 6.665967061423103e-05,
"loss": 2.2437,
"step": 6760
},
{
"epoch": 0.4533886954192339,
"grad_norm": 0.0,
"learning_rate": 6.654941245772981e-05,
"loss": 2.2341,
"step": 6770
},
{
"epoch": 0.45405839807125636,
"grad_norm": 0.0,
"learning_rate": 6.643906385033663e-05,
"loss": 2.2412,
"step": 6780
},
{
"epoch": 0.45472810072327885,
"grad_norm": 0.0,
"learning_rate": 6.632862539516227e-05,
"loss": 2.183,
"step": 6790
},
{
"epoch": 0.4553978033753014,
"grad_norm": 0.0,
"learning_rate": 6.621809769580867e-05,
"loss": 2.3024,
"step": 6800
},
{
"epoch": 0.4560675060273239,
"grad_norm": 0.0,
"learning_rate": 6.61074813563655e-05,
"loss": 2.1565,
"step": 6810
},
{
"epoch": 0.45673720867934636,
"grad_norm": 0.0,
"learning_rate": 6.599677698140688e-05,
"loss": 2.2655,
"step": 6820
},
{
"epoch": 0.45740691133136885,
"grad_norm": 0.0,
"learning_rate": 6.588598517598808e-05,
"loss": 2.2271,
"step": 6830
},
{
"epoch": 0.4580766139833914,
"grad_norm": 0.0,
"learning_rate": 6.577510654564222e-05,
"loss": 2.1505,
"step": 6840
},
{
"epoch": 0.4587463166354139,
"grad_norm": 0.0,
"learning_rate": 6.566414169637702e-05,
"loss": 2.2722,
"step": 6850
},
{
"epoch": 0.45941601928743636,
"grad_norm": 0.0,
"learning_rate": 6.555309123467136e-05,
"loss": 2.3141,
"step": 6860
},
{
"epoch": 0.4600857219394589,
"grad_norm": 0.0,
"learning_rate": 6.544195576747208e-05,
"loss": 2.2745,
"step": 6870
},
{
"epoch": 0.4607554245914814,
"grad_norm": 0.0,
"learning_rate": 6.533073590219058e-05,
"loss": 2.2846,
"step": 6880
},
{
"epoch": 0.4614251272435039,
"grad_norm": 0.0,
"learning_rate": 6.52194322466996e-05,
"loss": 2.2694,
"step": 6890
},
{
"epoch": 0.46209482989552636,
"grad_norm": 0.0,
"learning_rate": 6.510804540932974e-05,
"loss": 2.2835,
"step": 6900
},
{
"epoch": 0.4627645325475489,
"grad_norm": 0.0,
"learning_rate": 6.49965759988663e-05,
"loss": 2.2306,
"step": 6910
},
{
"epoch": 0.4634342351995714,
"grad_norm": 0.0,
"learning_rate": 6.48850246245459e-05,
"loss": 2.2775,
"step": 6920
},
{
"epoch": 0.4641039378515939,
"grad_norm": 0.0,
"learning_rate": 6.477339189605311e-05,
"loss": 2.3575,
"step": 6930
},
{
"epoch": 0.4647736405036164,
"grad_norm": 0.0,
"learning_rate": 6.466167842351708e-05,
"loss": 2.2569,
"step": 6940
},
{
"epoch": 0.4654433431556389,
"grad_norm": 0.0,
"learning_rate": 6.454988481750836e-05,
"loss": 2.207,
"step": 6950
},
{
"epoch": 0.4661130458076614,
"grad_norm": 0.0,
"learning_rate": 6.443801168903542e-05,
"loss": 2.2561,
"step": 6960
},
{
"epoch": 0.4667827484596839,
"grad_norm": 0.0,
"learning_rate": 6.432605964954139e-05,
"loss": 2.1582,
"step": 6970
},
{
"epoch": 0.4674524511117064,
"grad_norm": 0.0,
"learning_rate": 6.421402931090064e-05,
"loss": 2.2948,
"step": 6980
},
{
"epoch": 0.4681221537637289,
"grad_norm": 0.0,
"learning_rate": 6.410192128541554e-05,
"loss": 2.1836,
"step": 6990
},
{
"epoch": 0.4687918564157514,
"grad_norm": 0.0,
"learning_rate": 6.398973618581303e-05,
"loss": 2.2952,
"step": 7000
},
{
"epoch": 0.4694615590677739,
"grad_norm": 0.0,
"learning_rate": 6.38774746252413e-05,
"loss": 2.2567,
"step": 7010
},
{
"epoch": 0.4701312617197964,
"grad_norm": 0.0,
"learning_rate": 6.376513721726642e-05,
"loss": 2.1953,
"step": 7020
},
{
"epoch": 0.4708009643718189,
"grad_norm": 0.0,
"learning_rate": 6.365272457586906e-05,
"loss": 2.2673,
"step": 7030
},
{
"epoch": 0.47147066702384144,
"grad_norm": 0.0,
"learning_rate": 6.354023731544101e-05,
"loss": 2.1898,
"step": 7040
},
{
"epoch": 0.4721403696758639,
"grad_norm": 0.0,
"learning_rate": 6.342767605078195e-05,
"loss": 2.2221,
"step": 7050
},
{
"epoch": 0.4728100723278864,
"grad_norm": 0.0,
"learning_rate": 6.331504139709597e-05,
"loss": 2.3085,
"step": 7060
},
{
"epoch": 0.4734797749799089,
"grad_norm": 0.0,
"learning_rate": 6.320233396998835e-05,
"loss": 2.2103,
"step": 7070
},
{
"epoch": 0.47414947763193144,
"grad_norm": 0.0,
"learning_rate": 6.308955438546202e-05,
"loss": 2.2137,
"step": 7080
},
{
"epoch": 0.4748191802839539,
"grad_norm": 0.0,
"learning_rate": 6.297670325991436e-05,
"loss": 2.2429,
"step": 7090
},
{
"epoch": 0.4754888829359764,
"grad_norm": 0.0,
"learning_rate": 6.286378121013372e-05,
"loss": 2.1838,
"step": 7100
},
{
"epoch": 0.47615858558799895,
"grad_norm": 0.0,
"learning_rate": 6.275078885329609e-05,
"loss": 2.2697,
"step": 7110
},
{
"epoch": 0.47682828824002144,
"grad_norm": 0.0,
"learning_rate": 6.263772680696173e-05,
"loss": 2.2319,
"step": 7120
},
{
"epoch": 0.4774979908920439,
"grad_norm": 0.0,
"learning_rate": 6.252459568907181e-05,
"loss": 2.2726,
"step": 7130
},
{
"epoch": 0.4781676935440664,
"grad_norm": 0.0,
"learning_rate": 6.241139611794499e-05,
"loss": 2.2054,
"step": 7140
},
{
"epoch": 0.47883739619608895,
"grad_norm": 0.0,
"learning_rate": 6.229812871227401e-05,
"loss": 2.1353,
"step": 7150
},
{
"epoch": 0.47950709884811143,
"grad_norm": 0.0,
"learning_rate": 6.218479409112246e-05,
"loss": 2.2005,
"step": 7160
},
{
"epoch": 0.4801768015001339,
"grad_norm": 0.0,
"learning_rate": 6.207139287392126e-05,
"loss": 2.1865,
"step": 7170
},
{
"epoch": 0.48084650415215646,
"grad_norm": 0.0,
"learning_rate": 6.195792568046524e-05,
"loss": 2.1478,
"step": 7180
},
{
"epoch": 0.48151620680417895,
"grad_norm": 0.0,
"learning_rate": 6.184439313090992e-05,
"loss": 2.1964,
"step": 7190
},
{
"epoch": 0.48218590945620143,
"grad_norm": 0.0,
"learning_rate": 6.1730795845768e-05,
"loss": 2.2652,
"step": 7200
},
{
"epoch": 0.482855612108224,
"grad_norm": 0.0,
"learning_rate": 6.161713444590594e-05,
"loss": 2.1486,
"step": 7210
},
{
"epoch": 0.48352531476024646,
"grad_norm": 0.0,
"learning_rate": 6.150340955254067e-05,
"loss": 2.3471,
"step": 7220
},
{
"epoch": 0.48419501741226895,
"grad_norm": 0.0,
"learning_rate": 6.138962178723615e-05,
"loss": 2.3191,
"step": 7230
},
{
"epoch": 0.48486472006429143,
"grad_norm": 0.0,
"learning_rate": 6.127577177189994e-05,
"loss": 2.1875,
"step": 7240
},
{
"epoch": 0.485534422716314,
"grad_norm": 0.0,
"learning_rate": 6.116186012877984e-05,
"loss": 2.2136,
"step": 7250
},
{
"epoch": 0.48620412536833646,
"grad_norm": 0.0,
"learning_rate": 6.104788748046046e-05,
"loss": 2.1942,
"step": 7260
},
{
"epoch": 0.48687382802035895,
"grad_norm": 0.0,
"learning_rate": 6.093385444985985e-05,
"loss": 2.2411,
"step": 7270
},
{
"epoch": 0.4875435306723815,
"grad_norm": 0.0,
"learning_rate": 6.0819761660226095e-05,
"loss": 2.221,
"step": 7280
},
{
"epoch": 0.488213233324404,
"grad_norm": 0.0,
"learning_rate": 6.070560973513386e-05,
"loss": 2.2098,
"step": 7290
},
{
"epoch": 0.48888293597642646,
"grad_norm": 0.0,
"learning_rate": 6.0591399298481036e-05,
"loss": 2.2021,
"step": 7300
},
{
"epoch": 0.48955263862844894,
"grad_norm": 0.0,
"learning_rate": 6.047713097448531e-05,
"loss": 2.3027,
"step": 7310
},
{
"epoch": 0.4902223412804715,
"grad_norm": 0.0,
"learning_rate": 6.036280538768073e-05,
"loss": 2.1895,
"step": 7320
},
{
"epoch": 0.49089204393249397,
"grad_norm": 0.0,
"learning_rate": 6.0248423162914346e-05,
"loss": 2.1976,
"step": 7330
},
{
"epoch": 0.49156174658451646,
"grad_norm": 0.0,
"learning_rate": 6.013398492534274e-05,
"loss": 2.2668,
"step": 7340
},
{
"epoch": 0.492231449236539,
"grad_norm": 0.0,
"learning_rate": 6.001949130042864e-05,
"loss": 2.1931,
"step": 7350
},
{
"epoch": 0.4929011518885615,
"grad_norm": 0.0,
"learning_rate": 5.99049429139375e-05,
"loss": 2.2773,
"step": 7360
},
{
"epoch": 0.49357085454058397,
"grad_norm": 0.0,
"learning_rate": 5.9790340391934076e-05,
"loss": 2.1634,
"step": 7370
},
{
"epoch": 0.49424055719260646,
"grad_norm": 0.0,
"learning_rate": 5.9675684360778984e-05,
"loss": 2.3121,
"step": 7380
},
{
"epoch": 0.494910259844629,
"grad_norm": 0.0,
"learning_rate": 5.956097544712529e-05,
"loss": 2.289,
"step": 7390
},
{
"epoch": 0.4955799624966515,
"grad_norm": 0.0,
"learning_rate": 5.944621427791513e-05,
"loss": 2.3509,
"step": 7400
},
{
"epoch": 0.49624966514867397,
"grad_norm": 0.0,
"learning_rate": 5.933140148037617e-05,
"loss": 2.2824,
"step": 7410
},
{
"epoch": 0.4969193678006965,
"grad_norm": 0.0,
"learning_rate": 5.9216537682018345e-05,
"loss": 2.2424,
"step": 7420
},
{
"epoch": 0.497589070452719,
"grad_norm": 0.0,
"learning_rate": 5.910162351063029e-05,
"loss": 2.2644,
"step": 7430
},
{
"epoch": 0.4982587731047415,
"grad_norm": 0.0,
"learning_rate": 5.898665959427593e-05,
"loss": 2.2801,
"step": 7440
},
{
"epoch": 0.498928475756764,
"grad_norm": 0.0,
"learning_rate": 5.8871646561291085e-05,
"loss": 2.2242,
"step": 7450
},
{
"epoch": 0.4995981784087865,
"grad_norm": 0.0,
"learning_rate": 5.875658504028004e-05,
"loss": 2.2587,
"step": 7460
},
{
"epoch": 0.500267881060809,
"grad_norm": 0.0,
"learning_rate": 5.864147566011209e-05,
"loss": 2.3066,
"step": 7470
},
{
"epoch": 0.5009375837128315,
"grad_norm": 0.0,
"learning_rate": 5.8526319049918085e-05,
"loss": 2.2228,
"step": 7480
},
{
"epoch": 0.501607286364854,
"grad_norm": 0.0,
"learning_rate": 5.841111583908705e-05,
"loss": 2.1845,
"step": 7490
},
{
"epoch": 0.5022769890168765,
"grad_norm": 0.0,
"learning_rate": 5.8295866657262634e-05,
"loss": 2.2567,
"step": 7500
},
{
"epoch": 0.502946691668899,
"grad_norm": 0.0,
"learning_rate": 5.8180572134339805e-05,
"loss": 2.2037,
"step": 7510
},
{
"epoch": 0.5036163943209215,
"grad_norm": 0.0,
"learning_rate": 5.806523290046134e-05,
"loss": 2.1284,
"step": 7520
},
{
"epoch": 0.504286096972944,
"grad_norm": 0.0,
"learning_rate": 5.7949849586014337e-05,
"loss": 2.1658,
"step": 7530
},
{
"epoch": 0.5049557996249665,
"grad_norm": 0.0,
"learning_rate": 5.7834422821626855e-05,
"loss": 2.2954,
"step": 7540
},
{
"epoch": 0.505625502276989,
"grad_norm": 0.0,
"learning_rate": 5.771895323816444e-05,
"loss": 2.1094,
"step": 7550
},
{
"epoch": 0.5062952049290115,
"grad_norm": 0.0,
"learning_rate": 5.7603441466726637e-05,
"loss": 2.2384,
"step": 7560
},
{
"epoch": 0.5069649075810341,
"grad_norm": 0.0,
"learning_rate": 5.7487888138643555e-05,
"loss": 2.2502,
"step": 7570
},
{
"epoch": 0.5076346102330566,
"grad_norm": 0.0,
"learning_rate": 5.737229388547248e-05,
"loss": 2.0901,
"step": 7580
},
{
"epoch": 0.508304312885079,
"grad_norm": 0.0,
"learning_rate": 5.7256659338994334e-05,
"loss": 2.2689,
"step": 7590
},
{
"epoch": 0.5089740155371015,
"grad_norm": 0.0,
"learning_rate": 5.714098513121029e-05,
"loss": 2.2226,
"step": 7600
},
{
"epoch": 0.509643718189124,
"grad_norm": 0.0,
"learning_rate": 5.702527189433827e-05,
"loss": 2.2495,
"step": 7610
},
{
"epoch": 0.5103134208411465,
"grad_norm": 0.0,
"learning_rate": 5.690952026080954e-05,
"loss": 2.2461,
"step": 7620
},
{
"epoch": 0.510983123493169,
"grad_norm": 0.0,
"learning_rate": 5.6793730863265134e-05,
"loss": 2.205,
"step": 7630
},
{
"epoch": 0.5116528261451916,
"grad_norm": 0.0,
"learning_rate": 5.6677904334552636e-05,
"loss": 2.1968,
"step": 7640
},
{
"epoch": 0.5123225287972141,
"grad_norm": 0.0,
"learning_rate": 5.656204130772244e-05,
"loss": 2.1998,
"step": 7650
},
{
"epoch": 0.5129922314492366,
"grad_norm": 0.0,
"learning_rate": 5.644614241602445e-05,
"loss": 2.1372,
"step": 7660
},
{
"epoch": 0.513661934101259,
"grad_norm": 0.0,
"learning_rate": 5.633020829290463e-05,
"loss": 2.2227,
"step": 7670
},
{
"epoch": 0.5143316367532815,
"grad_norm": 0.0,
"learning_rate": 5.621423957200149e-05,
"loss": 2.2054,
"step": 7680
},
{
"epoch": 0.515001339405304,
"grad_norm": 0.0,
"learning_rate": 5.609823688714261e-05,
"loss": 2.1844,
"step": 7690
},
{
"epoch": 0.5156710420573265,
"grad_norm": 0.0,
"learning_rate": 5.598220087234121e-05,
"loss": 2.1766,
"step": 7700
},
{
"epoch": 0.5163407447093491,
"grad_norm": 0.0,
"learning_rate": 5.5866132161792676e-05,
"loss": 2.2252,
"step": 7710
},
{
"epoch": 0.5170104473613716,
"grad_norm": 0.0,
"learning_rate": 5.575003138987108e-05,
"loss": 2.2189,
"step": 7720
},
{
"epoch": 0.5176801500133941,
"grad_norm": 0.0,
"learning_rate": 5.563389919112576e-05,
"loss": 2.2178,
"step": 7730
},
{
"epoch": 0.5183498526654166,
"grad_norm": 0.0,
"learning_rate": 5.55177362002778e-05,
"loss": 2.2234,
"step": 7740
},
{
"epoch": 0.519019555317439,
"grad_norm": 0.0,
"learning_rate": 5.5401543052216565e-05,
"loss": 2.3178,
"step": 7750
},
{
"epoch": 0.5196892579694615,
"grad_norm": 0.0,
"learning_rate": 5.5285320381996266e-05,
"loss": 2.3021,
"step": 7760
},
{
"epoch": 0.520358960621484,
"grad_norm": 0.0,
"learning_rate": 5.5169068824832436e-05,
"loss": 2.2454,
"step": 7770
},
{
"epoch": 0.5210286632735066,
"grad_norm": 0.0,
"learning_rate": 5.505278901609852e-05,
"loss": 2.2416,
"step": 7780
},
{
"epoch": 0.5216983659255291,
"grad_norm": 0.0,
"learning_rate": 5.493648159132237e-05,
"loss": 2.196,
"step": 7790
},
{
"epoch": 0.5223680685775516,
"grad_norm": 0.0,
"learning_rate": 5.4820147186182743e-05,
"loss": 2.3359,
"step": 7800
},
{
"epoch": 0.5230377712295741,
"grad_norm": 0.0,
"learning_rate": 5.470378643650591e-05,
"loss": 2.2629,
"step": 7810
},
{
"epoch": 0.5237074738815966,
"grad_norm": 0.0,
"learning_rate": 5.458739997826209e-05,
"loss": 2.2305,
"step": 7820
},
{
"epoch": 0.524377176533619,
"grad_norm": 0.0,
"learning_rate": 5.4470988447562e-05,
"loss": 2.1962,
"step": 7830
},
{
"epoch": 0.5250468791856415,
"grad_norm": 0.0,
"learning_rate": 5.435455248065343e-05,
"loss": 2.1846,
"step": 7840
},
{
"epoch": 0.5257165818376641,
"grad_norm": 0.0,
"learning_rate": 5.42380927139177e-05,
"loss": 2.2579,
"step": 7850
},
{
"epoch": 0.5263862844896866,
"grad_norm": 0.0,
"learning_rate": 5.412160978386622e-05,
"loss": 2.1761,
"step": 7860
},
{
"epoch": 0.5270559871417091,
"grad_norm": 0.0,
"learning_rate": 5.400510432713699e-05,
"loss": 2.138,
"step": 7870
},
{
"epoch": 0.5277256897937316,
"grad_norm": 0.0,
"learning_rate": 5.3888576980491145e-05,
"loss": 2.2061,
"step": 7880
},
{
"epoch": 0.5283953924457541,
"grad_norm": 0.0,
"learning_rate": 5.3772028380809434e-05,
"loss": 2.2719,
"step": 7890
},
{
"epoch": 0.5290650950977765,
"grad_norm": 0.0,
"learning_rate": 5.3655459165088784e-05,
"loss": 2.3146,
"step": 7900
},
{
"epoch": 0.529734797749799,
"grad_norm": 0.0,
"learning_rate": 5.3538869970438795e-05,
"loss": 2.268,
"step": 7910
},
{
"epoch": 0.5304045004018216,
"grad_norm": 0.0,
"learning_rate": 5.342226143407827e-05,
"loss": 2.0887,
"step": 7920
},
{
"epoch": 0.5310742030538441,
"grad_norm": 0.0,
"learning_rate": 5.33056341933317e-05,
"loss": 2.1917,
"step": 7930
},
{
"epoch": 0.5317439057058666,
"grad_norm": 0.0,
"learning_rate": 5.318898888562584e-05,
"loss": 2.2067,
"step": 7940
},
{
"epoch": 0.5324136083578891,
"grad_norm": 0.0,
"learning_rate": 5.307232614848615e-05,
"loss": 2.2597,
"step": 7950
},
{
"epoch": 0.5330833110099116,
"grad_norm": 0.0,
"learning_rate": 5.295564661953337e-05,
"loss": 2.2602,
"step": 7960
},
{
"epoch": 0.5337530136619341,
"grad_norm": 0.0,
"learning_rate": 5.283895093648002e-05,
"loss": 2.2489,
"step": 7970
},
{
"epoch": 0.5344227163139567,
"grad_norm": 0.0,
"learning_rate": 5.272223973712694e-05,
"loss": 2.2289,
"step": 7980
},
{
"epoch": 0.5350924189659791,
"grad_norm": 0.0,
"learning_rate": 5.26055136593597e-05,
"loss": 2.3036,
"step": 7990
},
{
"epoch": 0.5357621216180016,
"grad_norm": 0.0,
"learning_rate": 5.248877334114525e-05,
"loss": 2.2411,
"step": 8000
},
{
"epoch": 0.5364318242700241,
"grad_norm": 0.0,
"learning_rate": 5.237201942052834e-05,
"loss": 2.196,
"step": 8010
},
{
"epoch": 0.5371015269220466,
"grad_norm": 0.0,
"learning_rate": 5.2255252535628084e-05,
"loss": 2.2915,
"step": 8020
},
{
"epoch": 0.5377712295740691,
"grad_norm": 0.0,
"learning_rate": 5.213847332463443e-05,
"loss": 2.2728,
"step": 8030
},
{
"epoch": 0.5384409322260916,
"grad_norm": 0.0,
"learning_rate": 5.202168242580472e-05,
"loss": 2.2747,
"step": 8040
},
{
"epoch": 0.5391106348781142,
"grad_norm": 0.0,
"learning_rate": 5.190488047746017e-05,
"loss": 2.2406,
"step": 8050
},
{
"epoch": 0.5397803375301367,
"grad_norm": 0.0,
"learning_rate": 5.178806811798236e-05,
"loss": 2.2754,
"step": 8060
},
{
"epoch": 0.5404500401821591,
"grad_norm": 0.0,
"learning_rate": 5.167124598580981e-05,
"loss": 2.2857,
"step": 8070
},
{
"epoch": 0.5411197428341816,
"grad_norm": 0.0,
"learning_rate": 5.155441471943442e-05,
"loss": 2.2487,
"step": 8080
},
{
"epoch": 0.5417894454862041,
"grad_norm": 0.0,
"learning_rate": 5.1437574957398026e-05,
"loss": 2.2719,
"step": 8090
},
{
"epoch": 0.5424591481382266,
"grad_norm": 0.0,
"learning_rate": 5.132072733828893e-05,
"loss": 2.2708,
"step": 8100
},
{
"epoch": 0.5431288507902491,
"grad_norm": 0.0,
"learning_rate": 5.120387250073833e-05,
"loss": 2.3129,
"step": 8110
},
{
"epoch": 0.5437985534422717,
"grad_norm": 0.0,
"learning_rate": 5.1087011083416866e-05,
"loss": 2.2324,
"step": 8120
},
{
"epoch": 0.5444682560942942,
"grad_norm": 0.0,
"learning_rate": 5.097014372503121e-05,
"loss": 2.258,
"step": 8130
},
{
"epoch": 0.5451379587463167,
"grad_norm": 0.0,
"learning_rate": 5.0853271064320427e-05,
"loss": 2.2671,
"step": 8140
},
{
"epoch": 0.5458076613983391,
"grad_norm": 0.0,
"learning_rate": 5.073639374005261e-05,
"loss": 2.1627,
"step": 8150
},
{
"epoch": 0.5464773640503616,
"grad_norm": 0.0,
"learning_rate": 5.0619512391021354e-05,
"loss": 2.1862,
"step": 8160
},
{
"epoch": 0.5471470667023841,
"grad_norm": 0.0,
"learning_rate": 5.0502627656042177e-05,
"loss": 2.2419,
"step": 8170
},
{
"epoch": 0.5478167693544066,
"grad_norm": 0.0,
"learning_rate": 5.038574017394917e-05,
"loss": 2.2802,
"step": 8180
},
{
"epoch": 0.5484864720064292,
"grad_norm": 0.0,
"learning_rate": 5.026885058359142e-05,
"loss": 2.2084,
"step": 8190
},
{
"epoch": 0.5491561746584517,
"grad_norm": 0.0,
"learning_rate": 5.015195952382954e-05,
"loss": 2.1723,
"step": 8200
},
{
"epoch": 0.5498258773104742,
"grad_norm": 0.0,
"learning_rate": 5.003506763353215e-05,
"loss": 2.2422,
"step": 8210
},
{
"epoch": 0.5504955799624967,
"grad_norm": 0.0,
"learning_rate": 4.991817555157246e-05,
"loss": 2.2529,
"step": 8220
},
{
"epoch": 0.5511652826145191,
"grad_norm": 0.0,
"learning_rate": 4.980128391682468e-05,
"loss": 2.1813,
"step": 8230
},
{
"epoch": 0.5518349852665416,
"grad_norm": 0.0,
"learning_rate": 4.968439336816058e-05,
"loss": 2.2105,
"step": 8240
},
{
"epoch": 0.5525046879185641,
"grad_norm": 0.0,
"learning_rate": 4.956750454444603e-05,
"loss": 2.2075,
"step": 8250
},
{
"epoch": 0.5531743905705867,
"grad_norm": 0.0,
"learning_rate": 4.9450618084537425e-05,
"loss": 2.1755,
"step": 8260
},
{
"epoch": 0.5538440932226092,
"grad_norm": 0.0,
"learning_rate": 4.9333734627278284e-05,
"loss": 2.1517,
"step": 8270
},
{
"epoch": 0.5545137958746317,
"grad_norm": 0.0,
"learning_rate": 4.921685481149569e-05,
"loss": 2.1668,
"step": 8280
},
{
"epoch": 0.5551834985266542,
"grad_norm": 0.0,
"learning_rate": 4.9099979275996806e-05,
"loss": 2.205,
"step": 8290
},
{
"epoch": 0.5558532011786766,
"grad_norm": 0.0,
"learning_rate": 4.898310865956544e-05,
"loss": 2.249,
"step": 8300
},
{
"epoch": 0.5565229038306991,
"grad_norm": 0.0,
"learning_rate": 4.886624360095851e-05,
"loss": 2.2103,
"step": 8310
},
{
"epoch": 0.5571926064827216,
"grad_norm": 0.0,
"learning_rate": 4.8749384738902545e-05,
"loss": 2.2059,
"step": 8320
},
{
"epoch": 0.5578623091347442,
"grad_norm": 0.0,
"learning_rate": 4.863253271209018e-05,
"loss": 2.2839,
"step": 8330
},
{
"epoch": 0.5585320117867667,
"grad_norm": 0.0,
"learning_rate": 4.851568815917675e-05,
"loss": 2.1547,
"step": 8340
},
{
"epoch": 0.5592017144387892,
"grad_norm": 0.0,
"learning_rate": 4.8398851718776703e-05,
"loss": 2.2948,
"step": 8350
},
{
"epoch": 0.5598714170908117,
"grad_norm": 0.0,
"learning_rate": 4.828202402946014e-05,
"loss": 2.1743,
"step": 8360
},
{
"epoch": 0.5605411197428342,
"grad_norm": 0.0,
"learning_rate": 4.8165205729749375e-05,
"loss": 2.2389,
"step": 8370
},
{
"epoch": 0.5612108223948566,
"grad_norm": 0.0,
"learning_rate": 4.804839745811535e-05,
"loss": 2.2072,
"step": 8380
},
{
"epoch": 0.5618805250468791,
"grad_norm": 0.0,
"learning_rate": 4.7931599852974244e-05,
"loss": 2.1628,
"step": 8390
},
{
"epoch": 0.5625502276989017,
"grad_norm": 0.0,
"learning_rate": 4.781481355268393e-05,
"loss": 2.1809,
"step": 8400
},
{
"epoch": 0.5632199303509242,
"grad_norm": 0.0,
"learning_rate": 4.769803919554044e-05,
"loss": 2.2693,
"step": 8410
},
{
"epoch": 0.5638896330029467,
"grad_norm": 0.0,
"learning_rate": 4.7581277419774616e-05,
"loss": 2.1879,
"step": 8420
},
{
"epoch": 0.5645593356549692,
"grad_norm": 0.0,
"learning_rate": 4.7464528863548466e-05,
"loss": 2.1736,
"step": 8430
},
{
"epoch": 0.5652290383069917,
"grad_norm": 0.0,
"learning_rate": 4.734779416495182e-05,
"loss": 2.2325,
"step": 8440
},
{
"epoch": 0.5658987409590142,
"grad_norm": 0.0,
"learning_rate": 4.723107396199867e-05,
"loss": 2.1428,
"step": 8450
},
{
"epoch": 0.5665684436110368,
"grad_norm": 0.0,
"learning_rate": 4.711436889262388e-05,
"loss": 2.1846,
"step": 8460
},
{
"epoch": 0.5672381462630592,
"grad_norm": 0.0,
"learning_rate": 4.699767959467955e-05,
"loss": 2.1416,
"step": 8470
},
{
"epoch": 0.5679078489150817,
"grad_norm": 0.0,
"learning_rate": 4.688100670593156e-05,
"loss": 2.2282,
"step": 8480
},
{
"epoch": 0.5685775515671042,
"grad_norm": 0.0,
"learning_rate": 4.6764350864056174e-05,
"loss": 2.2048,
"step": 8490
},
{
"epoch": 0.5692472542191267,
"grad_norm": 0.0,
"learning_rate": 4.6647712706636427e-05,
"loss": 2.204,
"step": 8500
},
{
"epoch": 0.5699169568711492,
"grad_norm": 0.0,
"learning_rate": 4.6531092871158706e-05,
"loss": 2.1583,
"step": 8510
},
{
"epoch": 0.5705866595231717,
"grad_norm": 0.0,
"learning_rate": 4.64144919950093e-05,
"loss": 2.2748,
"step": 8520
},
{
"epoch": 0.5712563621751943,
"grad_norm": 0.0,
"learning_rate": 4.6297910715470824e-05,
"loss": 2.2072,
"step": 8530
},
{
"epoch": 0.5719260648272168,
"grad_norm": 0.0,
"learning_rate": 4.6181349669718836e-05,
"loss": 2.2438,
"step": 8540
},
{
"epoch": 0.5725957674792392,
"grad_norm": 0.0,
"learning_rate": 4.606480949481825e-05,
"loss": 2.2714,
"step": 8550
},
{
"epoch": 0.5732654701312617,
"grad_norm": 0.0,
"learning_rate": 4.594829082771999e-05,
"loss": 2.1909,
"step": 8560
},
{
"epoch": 0.5739351727832842,
"grad_norm": 0.0,
"learning_rate": 4.583179430525734e-05,
"loss": 2.2508,
"step": 8570
},
{
"epoch": 0.5746048754353067,
"grad_norm": 0.0,
"learning_rate": 4.571532056414261e-05,
"loss": 2.2314,
"step": 8580
},
{
"epoch": 0.5752745780873292,
"grad_norm": 0.0,
"learning_rate": 4.559887024096359e-05,
"loss": 2.2437,
"step": 8590
},
{
"epoch": 0.5759442807393518,
"grad_norm": 0.0,
"learning_rate": 4.548244397218005e-05,
"loss": 2.216,
"step": 8600
},
{
"epoch": 0.5766139833913743,
"grad_norm": 0.0,
"learning_rate": 4.536604239412034e-05,
"loss": 2.2446,
"step": 8610
},
{
"epoch": 0.5772836860433967,
"grad_norm": 0.0,
"learning_rate": 4.5249666142977815e-05,
"loss": 2.1948,
"step": 8620
},
{
"epoch": 0.5779533886954192,
"grad_norm": 0.0,
"learning_rate": 4.513331585480742e-05,
"loss": 2.1927,
"step": 8630
},
{
"epoch": 0.5786230913474417,
"grad_norm": 0.0,
"learning_rate": 4.501699216552222e-05,
"loss": 2.2892,
"step": 8640
},
{
"epoch": 0.5792927939994642,
"grad_norm": 0.0,
"learning_rate": 4.4900695710889885e-05,
"loss": 2.222,
"step": 8650
},
{
"epoch": 0.5799624966514867,
"grad_norm": 0.0,
"learning_rate": 4.478442712652924e-05,
"loss": 2.2294,
"step": 8660
},
{
"epoch": 0.5806321993035093,
"grad_norm": 0.0,
"learning_rate": 4.466818704790675e-05,
"loss": 2.2264,
"step": 8670
},
{
"epoch": 0.5813019019555318,
"grad_norm": 0.0,
"learning_rate": 4.455197611033317e-05,
"loss": 2.1831,
"step": 8680
},
{
"epoch": 0.5819716046075543,
"grad_norm": 0.0,
"learning_rate": 4.443579494895989e-05,
"loss": 2.3108,
"step": 8690
},
{
"epoch": 0.5826413072595767,
"grad_norm": 0.0,
"learning_rate": 4.431964419877561e-05,
"loss": 2.094,
"step": 8700
},
{
"epoch": 0.5833110099115992,
"grad_norm": 0.0,
"learning_rate": 4.42035244946028e-05,
"loss": 2.1602,
"step": 8710
},
{
"epoch": 0.5839807125636217,
"grad_norm": 0.0,
"learning_rate": 4.408743647109425e-05,
"loss": 2.2146,
"step": 8720
},
{
"epoch": 0.5846504152156442,
"grad_norm": 0.0,
"learning_rate": 4.397138076272961e-05,
"loss": 2.1913,
"step": 8730
},
{
"epoch": 0.5853201178676668,
"grad_norm": 0.0,
"learning_rate": 4.38553580038119e-05,
"loss": 2.1041,
"step": 8740
},
{
"epoch": 0.5859898205196893,
"grad_norm": 0.0,
"learning_rate": 4.3739368828464036e-05,
"loss": 2.1198,
"step": 8750
},
{
"epoch": 0.5866595231717118,
"grad_norm": 0.0,
"learning_rate": 4.362341387062543e-05,
"loss": 2.1642,
"step": 8760
},
{
"epoch": 0.5873292258237343,
"grad_norm": 0.0,
"learning_rate": 4.350749376404844e-05,
"loss": 2.2242,
"step": 8770
},
{
"epoch": 0.5879989284757567,
"grad_norm": 0.0,
"learning_rate": 4.339160914229495e-05,
"loss": 2.1732,
"step": 8780
},
{
"epoch": 0.5886686311277792,
"grad_norm": 0.0,
"learning_rate": 4.327576063873293e-05,
"loss": 2.2439,
"step": 8790
},
{
"epoch": 0.5893383337798017,
"grad_norm": 0.0,
"learning_rate": 4.315994888653288e-05,
"loss": 2.2989,
"step": 8800
},
{
"epoch": 0.5900080364318243,
"grad_norm": 0.0,
"learning_rate": 4.304417451866454e-05,
"loss": 2.2347,
"step": 8810
},
{
"epoch": 0.5906777390838468,
"grad_norm": 0.0,
"learning_rate": 4.2928438167893224e-05,
"loss": 2.2218,
"step": 8820
},
{
"epoch": 0.5913474417358693,
"grad_norm": 0.0,
"learning_rate": 4.281274046677653e-05,
"loss": 2.1046,
"step": 8830
},
{
"epoch": 0.5920171443878918,
"grad_norm": 0.0,
"learning_rate": 4.269708204766074e-05,
"loss": 2.19,
"step": 8840
},
{
"epoch": 0.5926868470399143,
"grad_norm": 0.0,
"learning_rate": 4.2581463542677554e-05,
"loss": 2.1805,
"step": 8850
},
{
"epoch": 0.5933565496919367,
"grad_norm": 0.0,
"learning_rate": 4.2465885583740415e-05,
"loss": 2.2131,
"step": 8860
},
{
"epoch": 0.5940262523439592,
"grad_norm": 0.0,
"learning_rate": 4.2350348802541205e-05,
"loss": 2.1286,
"step": 8870
},
{
"epoch": 0.5946959549959818,
"grad_norm": 0.0,
"learning_rate": 4.2234853830546775e-05,
"loss": 2.1751,
"step": 8880
},
{
"epoch": 0.5953656576480043,
"grad_norm": 0.0,
"learning_rate": 4.211940129899542e-05,
"loss": 2.2916,
"step": 8890
},
{
"epoch": 0.5960353603000268,
"grad_norm": 0.0,
"learning_rate": 4.200399183889349e-05,
"loss": 2.213,
"step": 8900
},
{
"epoch": 0.5967050629520493,
"grad_norm": 0.0,
"learning_rate": 4.1888626081011954e-05,
"loss": 2.2073,
"step": 8910
},
{
"epoch": 0.5973747656040718,
"grad_norm": 0.0,
"learning_rate": 4.177330465588288e-05,
"loss": 2.2888,
"step": 8920
},
{
"epoch": 0.5980444682560943,
"grad_norm": 0.0,
"learning_rate": 4.165802819379609e-05,
"loss": 2.2332,
"step": 8930
},
{
"epoch": 0.5987141709081168,
"grad_norm": 0.0,
"learning_rate": 4.154279732479562e-05,
"loss": 2.2595,
"step": 8940
},
{
"epoch": 0.5993838735601393,
"grad_norm": 0.0,
"learning_rate": 4.142761267867634e-05,
"loss": 2.2216,
"step": 8950
},
{
"epoch": 0.6000535762121618,
"grad_norm": 0.0,
"learning_rate": 4.1312474884980454e-05,
"loss": 2.23,
"step": 8960
},
{
"epoch": 0.6007232788641843,
"grad_norm": 0.0,
"learning_rate": 4.119738457299415e-05,
"loss": 2.2446,
"step": 8970
},
{
"epoch": 0.6013929815162068,
"grad_norm": 0.0,
"learning_rate": 4.1082342371744044e-05,
"loss": 2.2392,
"step": 8980
},
{
"epoch": 0.6020626841682293,
"grad_norm": 0.0,
"learning_rate": 4.096734890999382e-05,
"loss": 2.1591,
"step": 8990
},
{
"epoch": 0.6027323868202518,
"grad_norm": 0.0,
"learning_rate": 4.085240481624083e-05,
"loss": 2.277,
"step": 9000
},
{
"epoch": 0.6034020894722744,
"grad_norm": 0.0,
"learning_rate": 4.073751071871254e-05,
"loss": 2.1811,
"step": 9010
},
{
"epoch": 0.6040717921242968,
"grad_norm": 0.0,
"learning_rate": 4.062266724536317e-05,
"loss": 2.2649,
"step": 9020
},
{
"epoch": 0.6047414947763193,
"grad_norm": 0.0,
"learning_rate": 4.050787502387028e-05,
"loss": 2.2494,
"step": 9030
},
{
"epoch": 0.6054111974283418,
"grad_norm": 0.0,
"learning_rate": 4.039313468163128e-05,
"loss": 2.1891,
"step": 9040
},
{
"epoch": 0.6060809000803643,
"grad_norm": 0.0,
"learning_rate": 4.0278446845760077e-05,
"loss": 2.248,
"step": 9050
},
{
"epoch": 0.6067506027323868,
"grad_norm": 0.0,
"learning_rate": 4.0163812143083566e-05,
"loss": 2.1845,
"step": 9060
},
{
"epoch": 0.6074203053844093,
"grad_norm": 0.0,
"learning_rate": 4.004923120013828e-05,
"loss": 2.2027,
"step": 9070
},
{
"epoch": 0.6080900080364319,
"grad_norm": 0.0,
"learning_rate": 3.993470464316686e-05,
"loss": 2.2485,
"step": 9080
},
{
"epoch": 0.6087597106884544,
"grad_norm": 0.0,
"learning_rate": 3.98202330981148e-05,
"loss": 2.1538,
"step": 9090
},
{
"epoch": 0.6094294133404768,
"grad_norm": 0.0,
"learning_rate": 3.9705817190626827e-05,
"loss": 2.2527,
"step": 9100
},
{
"epoch": 0.6100991159924993,
"grad_norm": 0.0,
"learning_rate": 3.959145754604363e-05,
"loss": 2.2026,
"step": 9110
},
{
"epoch": 0.6107688186445218,
"grad_norm": 0.0,
"learning_rate": 3.94771547893984e-05,
"loss": 2.1733,
"step": 9120
},
{
"epoch": 0.6114385212965443,
"grad_norm": 0.0,
"learning_rate": 3.93629095454134e-05,
"loss": 2.3497,
"step": 9130
},
{
"epoch": 0.6121082239485668,
"grad_norm": 0.0,
"learning_rate": 3.924872243849651e-05,
"loss": 2.2275,
"step": 9140
},
{
"epoch": 0.6127779266005894,
"grad_norm": 0.0,
"learning_rate": 3.9134594092737944e-05,
"loss": 2.3132,
"step": 9150
},
{
"epoch": 0.6134476292526119,
"grad_norm": 0.0,
"learning_rate": 3.902052513190671e-05,
"loss": 2.2949,
"step": 9160
},
{
"epoch": 0.6141173319046344,
"grad_norm": 0.0,
"learning_rate": 3.890651617944722e-05,
"loss": 2.199,
"step": 9170
},
{
"epoch": 0.6147870345566568,
"grad_norm": 0.0,
"learning_rate": 3.879256785847597e-05,
"loss": 2.1144,
"step": 9180
},
{
"epoch": 0.6154567372086793,
"grad_norm": 0.0,
"learning_rate": 3.867868079177805e-05,
"loss": 2.1892,
"step": 9190
},
{
"epoch": 0.6161264398607018,
"grad_norm": 0.0,
"learning_rate": 3.8564855601803754e-05,
"loss": 2.221,
"step": 9200
},
{
"epoch": 0.6167961425127243,
"grad_norm": 0.0,
"learning_rate": 3.8451092910665206e-05,
"loss": 2.2979,
"step": 9210
},
{
"epoch": 0.6174658451647469,
"grad_norm": 0.0,
"learning_rate": 3.833739334013292e-05,
"loss": 2.2331,
"step": 9220
},
{
"epoch": 0.6181355478167694,
"grad_norm": 0.0,
"learning_rate": 3.8223757511632424e-05,
"loss": 2.2855,
"step": 9230
},
{
"epoch": 0.6188052504687919,
"grad_norm": 0.0,
"learning_rate": 3.81101860462409e-05,
"loss": 2.1975,
"step": 9240
},
{
"epoch": 0.6194749531208144,
"grad_norm": 0.0,
"learning_rate": 3.7996679564683714e-05,
"loss": 2.1869,
"step": 9250
},
{
"epoch": 0.6201446557728368,
"grad_norm": 0.0,
"learning_rate": 3.788323868733107e-05,
"loss": 2.2632,
"step": 9260
},
{
"epoch": 0.6208143584248593,
"grad_norm": 0.0,
"learning_rate": 3.7769864034194646e-05,
"loss": 2.1872,
"step": 9270
},
{
"epoch": 0.6214840610768818,
"grad_norm": 0.0,
"learning_rate": 3.7656556224924125e-05,
"loss": 2.2888,
"step": 9280
},
{
"epoch": 0.6221537637289044,
"grad_norm": 0.0,
"learning_rate": 3.7543315878803863e-05,
"loss": 2.1669,
"step": 9290
},
{
"epoch": 0.6228234663809269,
"grad_norm": 0.0,
"learning_rate": 3.743014361474953e-05,
"loss": 2.2148,
"step": 9300
},
{
"epoch": 0.6234931690329494,
"grad_norm": 0.0,
"learning_rate": 3.731704005130466e-05,
"loss": 2.0846,
"step": 9310
},
{
"epoch": 0.6241628716849719,
"grad_norm": 0.0,
"learning_rate": 3.720400580663732e-05,
"loss": 2.319,
"step": 9320
},
{
"epoch": 0.6248325743369944,
"grad_norm": 0.0,
"learning_rate": 3.709104149853669e-05,
"loss": 2.2351,
"step": 9330
},
{
"epoch": 0.6255022769890168,
"grad_norm": 0.0,
"learning_rate": 3.697814774440976e-05,
"loss": 2.1733,
"step": 9340
},
{
"epoch": 0.6261719796410394,
"grad_norm": 0.0,
"learning_rate": 3.686532516127785e-05,
"loss": 2.3272,
"step": 9350
},
{
"epoch": 0.6268416822930619,
"grad_norm": 0.0,
"learning_rate": 3.675257436577333e-05,
"loss": 2.1916,
"step": 9360
},
{
"epoch": 0.6275113849450844,
"grad_norm": 0.0,
"learning_rate": 3.663989597413621e-05,
"loss": 2.2904,
"step": 9370
},
{
"epoch": 0.6281810875971069,
"grad_norm": 0.0,
"learning_rate": 3.6527290602210763e-05,
"loss": 2.2809,
"step": 9380
},
{
"epoch": 0.6288507902491294,
"grad_norm": 0.0,
"learning_rate": 3.6414758865442186e-05,
"loss": 2.1766,
"step": 9390
},
{
"epoch": 0.6295204929011519,
"grad_norm": 0.0,
"learning_rate": 3.6302301378873225e-05,
"loss": 2.223,
"step": 9400
},
{
"epoch": 0.6301901955531743,
"grad_norm": 0.0,
"learning_rate": 3.618991875714078e-05,
"loss": 2.2235,
"step": 9410
},
{
"epoch": 0.630859898205197,
"grad_norm": 0.0,
"learning_rate": 3.6077611614472616e-05,
"loss": 2.2017,
"step": 9420
},
{
"epoch": 0.6315296008572194,
"grad_norm": 0.0,
"learning_rate": 3.596538056468397e-05,
"loss": 2.3,
"step": 9430
},
{
"epoch": 0.6321993035092419,
"grad_norm": 0.0,
"learning_rate": 3.5853226221174155e-05,
"loss": 2.1629,
"step": 9440
},
{
"epoch": 0.6328690061612644,
"grad_norm": 0.0,
"learning_rate": 3.5741149196923264e-05,
"loss": 2.1419,
"step": 9450
},
{
"epoch": 0.6335387088132869,
"grad_norm": 0.0,
"learning_rate": 3.5629150104488816e-05,
"loss": 2.1754,
"step": 9460
},
{
"epoch": 0.6342084114653094,
"grad_norm": 0.0,
"learning_rate": 3.5517229556002366e-05,
"loss": 2.2461,
"step": 9470
},
{
"epoch": 0.6348781141173319,
"grad_norm": 0.0,
"learning_rate": 3.540538816316621e-05,
"loss": 2.1987,
"step": 9480
},
{
"epoch": 0.6355478167693545,
"grad_norm": 0.0,
"learning_rate": 3.529362653725001e-05,
"loss": 2.1122,
"step": 9490
},
{
"epoch": 0.6362175194213769,
"grad_norm": 0.0,
"learning_rate": 3.5181945289087445e-05,
"loss": 2.2948,
"step": 9500
},
{
"epoch": 0.6368872220733994,
"grad_norm": 0.0,
"learning_rate": 3.507034502907292e-05,
"loss": 2.2711,
"step": 9510
},
{
"epoch": 0.6375569247254219,
"grad_norm": 0.0,
"learning_rate": 3.495882636715818e-05,
"loss": 2.1809,
"step": 9520
},
{
"epoch": 0.6382266273774444,
"grad_norm": 0.0,
"learning_rate": 3.484738991284899e-05,
"loss": 2.2134,
"step": 9530
},
{
"epoch": 0.6388963300294669,
"grad_norm": 0.0,
"learning_rate": 3.473603627520181e-05,
"loss": 2.2256,
"step": 9540
},
{
"epoch": 0.6395660326814894,
"grad_norm": 0.0,
"learning_rate": 3.462476606282051e-05,
"loss": 2.2067,
"step": 9550
},
{
"epoch": 0.640235735333512,
"grad_norm": 0.0,
"learning_rate": 3.4513579883852945e-05,
"loss": 2.2316,
"step": 9560
},
{
"epoch": 0.6409054379855345,
"grad_norm": 0.0,
"learning_rate": 3.440247834598768e-05,
"loss": 2.2769,
"step": 9570
},
{
"epoch": 0.6415751406375569,
"grad_norm": 0.0,
"learning_rate": 3.429146205645072e-05,
"loss": 2.2183,
"step": 9580
},
{
"epoch": 0.6422448432895794,
"grad_norm": 0.0,
"learning_rate": 3.41805316220021e-05,
"loss": 2.154,
"step": 9590
},
{
"epoch": 0.6429145459416019,
"grad_norm": 0.0,
"learning_rate": 3.406968764893266e-05,
"loss": 2.2405,
"step": 9600
},
{
"epoch": 0.6435842485936244,
"grad_norm": 0.0,
"learning_rate": 3.3958930743060635e-05,
"loss": 2.2594,
"step": 9610
},
{
"epoch": 0.6442539512456469,
"grad_norm": 0.0,
"learning_rate": 3.384826150972842e-05,
"loss": 2.1802,
"step": 9620
},
{
"epoch": 0.6449236538976695,
"grad_norm": 0.0,
"learning_rate": 3.373768055379924e-05,
"loss": 2.171,
"step": 9630
},
{
"epoch": 0.645593356549692,
"grad_norm": 0.0,
"learning_rate": 3.3627188479653836e-05,
"loss": 2.3012,
"step": 9640
},
{
"epoch": 0.6462630592017145,
"grad_norm": 0.0,
"learning_rate": 3.351678589118714e-05,
"loss": 2.2603,
"step": 9650
},
{
"epoch": 0.6469327618537369,
"grad_norm": 0.0,
"learning_rate": 3.340647339180503e-05,
"loss": 2.2594,
"step": 9660
},
{
"epoch": 0.6476024645057594,
"grad_norm": 0.0,
"learning_rate": 3.329625158442102e-05,
"loss": 2.2487,
"step": 9670
},
{
"epoch": 0.6482721671577819,
"grad_norm": 0.0,
"learning_rate": 3.3186121071452895e-05,
"loss": 2.2431,
"step": 9680
},
{
"epoch": 0.6489418698098044,
"grad_norm": 0.0,
"learning_rate": 3.307608245481949e-05,
"loss": 2.2311,
"step": 9690
},
{
"epoch": 0.649611572461827,
"grad_norm": 0.0,
"learning_rate": 3.2966136335937395e-05,
"loss": 2.2746,
"step": 9700
},
{
"epoch": 0.6502812751138495,
"grad_norm": 0.0,
"learning_rate": 3.285628331571764e-05,
"loss": 2.1033,
"step": 9710
},
{
"epoch": 0.650950977765872,
"grad_norm": 0.0,
"learning_rate": 3.2746523994562416e-05,
"loss": 2.2116,
"step": 9720
},
{
"epoch": 0.6516206804178944,
"grad_norm": 0.0,
"learning_rate": 3.263685897236183e-05,
"loss": 2.2938,
"step": 9730
},
{
"epoch": 0.6522903830699169,
"grad_norm": 0.0,
"learning_rate": 3.252728884849056e-05,
"loss": 2.1125,
"step": 9740
},
{
"epoch": 0.6529600857219394,
"grad_norm": 0.0,
"learning_rate": 3.241781422180464e-05,
"loss": 2.2638,
"step": 9750
},
{
"epoch": 0.6536297883739619,
"grad_norm": 0.0,
"learning_rate": 3.2308435690638186e-05,
"loss": 2.1999,
"step": 9760
},
{
"epoch": 0.6542994910259845,
"grad_norm": 0.0,
"learning_rate": 3.219915385280002e-05,
"loss": 2.2453,
"step": 9770
},
{
"epoch": 0.654969193678007,
"grad_norm": 0.0,
"learning_rate": 3.2089969305570606e-05,
"loss": 2.2918,
"step": 9780
},
{
"epoch": 0.6556388963300295,
"grad_norm": 0.0,
"learning_rate": 3.198088264569854e-05,
"loss": 2.2353,
"step": 9790
},
{
"epoch": 0.656308598982052,
"grad_norm": 0.0,
"learning_rate": 3.1871894469397544e-05,
"loss": 2.1873,
"step": 9800
},
{
"epoch": 0.6569783016340744,
"grad_norm": 0.0,
"learning_rate": 3.1763005372342966e-05,
"loss": 2.2251,
"step": 9810
},
{
"epoch": 0.6576480042860969,
"grad_norm": 0.0,
"learning_rate": 3.1654215949668695e-05,
"loss": 2.296,
"step": 9820
},
{
"epoch": 0.6583177069381195,
"grad_norm": 0.0,
"learning_rate": 3.154552679596383e-05,
"loss": 2.3183,
"step": 9830
},
{
"epoch": 0.658987409590142,
"grad_norm": 0.0,
"learning_rate": 3.143693850526945e-05,
"loss": 2.1321,
"step": 9840
},
{
"epoch": 1.0004018215912136,
"grad_norm": 0.0,
"learning_rate": 3.132845167107539e-05,
"loss": 2.1889,
"step": 9850
},
{
"epoch": 1.001071524243236,
"grad_norm": 0.0,
"learning_rate": 3.122006688631693e-05,
"loss": 2.1428,
"step": 9860
},
{
"epoch": 1.0017412268952586,
"grad_norm": 0.0,
"learning_rate": 3.1111784743371644e-05,
"loss": 2.0989,
"step": 9870
},
{
"epoch": 1.002410929547281,
"grad_norm": 0.0,
"learning_rate": 3.100360583405608e-05,
"loss": 2.1649,
"step": 9880
},
{
"epoch": 1.0030806321993035,
"grad_norm": 0.0,
"learning_rate": 3.0895530749622584e-05,
"loss": 2.2272,
"step": 9890
},
{
"epoch": 1.003750334851326,
"grad_norm": 0.0,
"learning_rate": 3.078756008075605e-05,
"loss": 2.0914,
"step": 9900
},
{
"epoch": 1.0044200375033485,
"grad_norm": 0.0,
"learning_rate": 3.0679694417570635e-05,
"loss": 2.0899,
"step": 9910
},
{
"epoch": 1.0050897401553711,
"grad_norm": 0.0,
"learning_rate": 3.05719343496067e-05,
"loss": 1.9538,
"step": 9920
},
{
"epoch": 1.0057594428073935,
"grad_norm": 0.0,
"learning_rate": 3.046428046582736e-05,
"loss": 2.155,
"step": 9930
},
{
"epoch": 1.006429145459416,
"grad_norm": 0.0,
"learning_rate": 3.0356733354615446e-05,
"loss": 2.1708,
"step": 9940
},
{
"epoch": 1.0070988481114385,
"grad_norm": 0.0,
"learning_rate": 3.024929360377019e-05,
"loss": 2.2296,
"step": 9950
},
{
"epoch": 1.007768550763461,
"grad_norm": 0.0,
"learning_rate": 3.014196180050406e-05,
"loss": 2.0675,
"step": 9960
},
{
"epoch": 1.0084382534154834,
"grad_norm": 0.0,
"learning_rate": 3.0034738531439533e-05,
"loss": 2.118,
"step": 9970
},
{
"epoch": 1.009107956067506,
"grad_norm": 0.0,
"learning_rate": 2.992762438260589e-05,
"loss": 2.1154,
"step": 9980
},
{
"epoch": 1.0097776587195286,
"grad_norm": 0.0,
"learning_rate": 2.9820619939436035e-05,
"loss": 2.1501,
"step": 9990
},
{
"epoch": 1.010447361371551,
"grad_norm": 0.0,
"learning_rate": 2.9713725786763248e-05,
"loss": 2.0916,
"step": 10000
},
{
"epoch": 1.0111170640235736,
"grad_norm": 0.0,
"learning_rate": 2.9606942508818e-05,
"loss": 2.0645,
"step": 10010
},
{
"epoch": 1.011786766675596,
"grad_norm": 0.0,
"learning_rate": 2.950027068922484e-05,
"loss": 2.179,
"step": 10020
},
{
"epoch": 1.0124564693276186,
"grad_norm": 0.0,
"learning_rate": 2.9393710910999073e-05,
"loss": 2.1498,
"step": 10030
},
{
"epoch": 1.013126171979641,
"grad_norm": 0.0,
"learning_rate": 2.928726375654369e-05,
"loss": 2.1339,
"step": 10040
},
{
"epoch": 1.0137958746316635,
"grad_norm": 0.0,
"learning_rate": 2.9180929807646118e-05,
"loss": 2.1826,
"step": 10050
},
{
"epoch": 1.0144655772836861,
"grad_norm": 0.0,
"learning_rate": 2.9074709645475036e-05,
"loss": 2.1725,
"step": 10060
},
{
"epoch": 1.0151352799357085,
"grad_norm": 0.0,
"learning_rate": 2.896860385057726e-05,
"loss": 2.1184,
"step": 10070
},
{
"epoch": 1.015804982587731,
"grad_norm": 0.0,
"learning_rate": 2.8862613002874517e-05,
"loss": 2.1758,
"step": 10080
},
{
"epoch": 1.0164746852397535,
"grad_norm": 0.0,
"learning_rate": 2.875673768166026e-05,
"loss": 2.1563,
"step": 10090
},
{
"epoch": 1.017144387891776,
"grad_norm": 0.0,
"learning_rate": 2.865097846559658e-05,
"loss": 2.1865,
"step": 10100
},
{
"epoch": 1.0178140905437985,
"grad_norm": 0.0,
"learning_rate": 2.8545335932710985e-05,
"loss": 2.1202,
"step": 10110
},
{
"epoch": 1.018483793195821,
"grad_norm": 0.0,
"learning_rate": 2.8439810660393197e-05,
"loss": 2.2371,
"step": 10120
},
{
"epoch": 1.0191534958478436,
"grad_norm": 0.0,
"learning_rate": 2.8334403225392116e-05,
"loss": 2.2004,
"step": 10130
},
{
"epoch": 1.019823198499866,
"grad_norm": 0.0,
"learning_rate": 2.822911420381259e-05,
"loss": 2.236,
"step": 10140
},
{
"epoch": 1.0204929011518886,
"grad_norm": 0.0,
"learning_rate": 2.8123944171112226e-05,
"loss": 2.1106,
"step": 10150
},
{
"epoch": 1.021162603803911,
"grad_norm": 0.0,
"learning_rate": 2.8018893702098337e-05,
"loss": 2.1454,
"step": 10160
},
{
"epoch": 1.0218323064559336,
"grad_norm": 0.0,
"learning_rate": 2.7913963370924817e-05,
"loss": 2.2404,
"step": 10170
},
{
"epoch": 1.022502009107956,
"grad_norm": 0.0,
"learning_rate": 2.7809153751088823e-05,
"loss": 2.2138,
"step": 10180
},
{
"epoch": 1.0231717117599786,
"grad_norm": 0.0,
"learning_rate": 2.7704465415427854e-05,
"loss": 2.1251,
"step": 10190
},
{
"epoch": 1.0238414144120012,
"grad_norm": 0.0,
"learning_rate": 2.7599898936116518e-05,
"loss": 2.2301,
"step": 10200
},
{
"epoch": 1.0245111170640235,
"grad_norm": 0.0,
"learning_rate": 2.7495454884663364e-05,
"loss": 2.1531,
"step": 10210
},
{
"epoch": 1.0251808197160461,
"grad_norm": 0.0,
"learning_rate": 2.7391133831907862e-05,
"loss": 2.1389,
"step": 10220
},
{
"epoch": 1.0258505223680685,
"grad_norm": 0.0,
"learning_rate": 2.7286936348017245e-05,
"loss": 2.1855,
"step": 10230
},
{
"epoch": 1.026520225020091,
"grad_norm": 0.0,
"learning_rate": 2.7182863002483317e-05,
"loss": 2.1678,
"step": 10240
},
{
"epoch": 1.0271899276721135,
"grad_norm": 0.0,
"learning_rate": 2.7078914364119458e-05,
"loss": 2.106,
"step": 10250
},
{
"epoch": 1.027859630324136,
"grad_norm": 0.0,
"learning_rate": 2.697509100105745e-05,
"loss": 2.2005,
"step": 10260
},
{
"epoch": 1.0285293329761587,
"grad_norm": 0.0,
"learning_rate": 2.6871393480744356e-05,
"loss": 2.1844,
"step": 10270
},
{
"epoch": 1.029199035628181,
"grad_norm": 0.0,
"learning_rate": 2.6767822369939442e-05,
"loss": 2.1796,
"step": 10280
},
{
"epoch": 1.0298687382802036,
"grad_norm": 0.0,
"learning_rate": 2.6664378234711185e-05,
"loss": 2.1288,
"step": 10290
},
{
"epoch": 1.030538440932226,
"grad_norm": 0.0,
"learning_rate": 2.6561061640433916e-05,
"loss": 2.1717,
"step": 10300
},
{
"epoch": 1.0312081435842486,
"grad_norm": 0.0,
"learning_rate": 2.645787315178499e-05,
"loss": 2.0873,
"step": 10310
},
{
"epoch": 1.031877846236271,
"grad_norm": 0.0,
"learning_rate": 2.6354813332741612e-05,
"loss": 2.1526,
"step": 10320
},
{
"epoch": 1.0325475488882936,
"grad_norm": 0.0,
"learning_rate": 2.625188274657766e-05,
"loss": 2.0779,
"step": 10330
},
{
"epoch": 1.0332172515403162,
"grad_norm": 0.0,
"learning_rate": 2.6149081955860765e-05,
"loss": 2.0766,
"step": 10340
},
{
"epoch": 1.0338869541923386,
"grad_norm": 0.0,
"learning_rate": 2.6046411522449148e-05,
"loss": 2.1073,
"step": 10350
},
{
"epoch": 1.0345566568443612,
"grad_norm": 0.0,
"learning_rate": 2.594387200748851e-05,
"loss": 2.1058,
"step": 10360
},
{
"epoch": 1.0352263594963835,
"grad_norm": 0.0,
"learning_rate": 2.5841463971409085e-05,
"loss": 2.1599,
"step": 10370
},
{
"epoch": 1.0358960621484061,
"grad_norm": 0.0,
"learning_rate": 2.5739187973922484e-05,
"loss": 1.9751,
"step": 10380
},
{
"epoch": 1.0365657648004287,
"grad_norm": 0.0,
"learning_rate": 2.563704457401862e-05,
"loss": 2.0324,
"step": 10390
},
{
"epoch": 1.037235467452451,
"grad_norm": 0.0,
"learning_rate": 2.5535034329962726e-05,
"loss": 2.1225,
"step": 10400
},
{
"epoch": 1.0379051701044737,
"grad_norm": 0.0,
"learning_rate": 2.543315779929233e-05,
"loss": 2.0986,
"step": 10410
},
{
"epoch": 1.038574872756496,
"grad_norm": 0.0,
"learning_rate": 2.533141553881404e-05,
"loss": 2.0317,
"step": 10420
},
{
"epoch": 1.0392445754085187,
"grad_norm": 0.0,
"learning_rate": 2.5229808104600684e-05,
"loss": 2.0995,
"step": 10430
},
{
"epoch": 1.039914278060541,
"grad_norm": 0.0,
"learning_rate": 2.512833605198818e-05,
"loss": 2.1134,
"step": 10440
},
{
"epoch": 1.0405839807125636,
"grad_norm": 0.0,
"learning_rate": 2.502699993557247e-05,
"loss": 2.1163,
"step": 10450
},
{
"epoch": 1.041253683364586,
"grad_norm": 0.0,
"learning_rate": 2.4925800309206614e-05,
"loss": 2.0981,
"step": 10460
},
{
"epoch": 1.0419233860166086,
"grad_norm": 0.0,
"learning_rate": 2.482473772599766e-05,
"loss": 2.1944,
"step": 10470
},
{
"epoch": 1.0425930886686312,
"grad_norm": 0.0,
"learning_rate": 2.472381273830359e-05,
"loss": 2.0236,
"step": 10480
},
{
"epoch": 1.0432627913206536,
"grad_norm": 0.0,
"learning_rate": 2.4623025897730427e-05,
"loss": 2.1266,
"step": 10490
},
{
"epoch": 1.0439324939726762,
"grad_norm": 0.0,
"learning_rate": 2.452237775512914e-05,
"loss": 2.1275,
"step": 10500
},
{
"epoch": 1.0446021966246986,
"grad_norm": 0.0,
"learning_rate": 2.442186886059259e-05,
"loss": 2.1504,
"step": 10510
},
{
"epoch": 1.0452718992767212,
"grad_norm": 0.0,
"learning_rate": 2.4321499763452637e-05,
"loss": 2.1843,
"step": 10520
},
{
"epoch": 1.0459416019287437,
"grad_norm": 0.0,
"learning_rate": 2.4221271012277047e-05,
"loss": 2.1632,
"step": 10530
},
{
"epoch": 1.0466113045807661,
"grad_norm": 0.0,
"learning_rate": 2.4121183154866546e-05,
"loss": 2.1631,
"step": 10540
},
{
"epoch": 1.0472810072327887,
"grad_norm": 0.0,
"learning_rate": 2.402123673825178e-05,
"loss": 2.0182,
"step": 10550
},
{
"epoch": 1.047950709884811,
"grad_norm": 0.0,
"learning_rate": 2.392143230869039e-05,
"loss": 2.1462,
"step": 10560
},
{
"epoch": 1.0486204125368337,
"grad_norm": 0.0,
"learning_rate": 2.382177041166391e-05,
"loss": 2.1991,
"step": 10570
},
{
"epoch": 1.049290115188856,
"grad_norm": 0.0,
"learning_rate": 2.3722251591874932e-05,
"loss": 2.1854,
"step": 10580
},
{
"epoch": 1.0499598178408787,
"grad_norm": 0.0,
"learning_rate": 2.3622876393244053e-05,
"loss": 2.1261,
"step": 10590
},
{
"epoch": 1.0506295204929013,
"grad_norm": 0.0,
"learning_rate": 2.352364535890686e-05,
"loss": 2.2759,
"step": 10600
},
{
"epoch": 1.0512992231449236,
"grad_norm": 0.0,
"learning_rate": 2.342455903121105e-05,
"loss": 2.1393,
"step": 10610
},
{
"epoch": 1.0519689257969462,
"grad_norm": 0.0,
"learning_rate": 2.3325617951713445e-05,
"loss": 2.2647,
"step": 10620
},
{
"epoch": 1.0526386284489686,
"grad_norm": 0.0,
"learning_rate": 2.3226822661176933e-05,
"loss": 2.1644,
"step": 10630
},
{
"epoch": 1.0533083311009912,
"grad_norm": 0.0,
"learning_rate": 2.312817369956766e-05,
"loss": 2.1273,
"step": 10640
},
{
"epoch": 1.0539780337530136,
"grad_norm": 0.0,
"learning_rate": 2.3029671606052004e-05,
"loss": 2.1682,
"step": 10650
},
{
"epoch": 1.0546477364050362,
"grad_norm": 0.0,
"learning_rate": 2.2931316918993613e-05,
"loss": 2.1651,
"step": 10660
},
{
"epoch": 1.0553174390570588,
"grad_norm": 0.0,
"learning_rate": 2.2833110175950496e-05,
"loss": 2.2218,
"step": 10670
},
{
"epoch": 1.0559871417090811,
"grad_norm": 0.0,
"learning_rate": 2.273505191367209e-05,
"loss": 2.1568,
"step": 10680
},
{
"epoch": 1.0566568443611037,
"grad_norm": 0.0,
"learning_rate": 2.263714266809625e-05,
"loss": 2.1827,
"step": 10690
},
{
"epoch": 1.0573265470131261,
"grad_norm": 0.0,
"learning_rate": 2.2539382974346445e-05,
"loss": 2.2167,
"step": 10700
},
{
"epoch": 1.0579962496651487,
"grad_norm": 0.0,
"learning_rate": 2.2441773366728758e-05,
"loss": 2.0347,
"step": 10710
},
{
"epoch": 1.058665952317171,
"grad_norm": 0.0,
"learning_rate": 2.2344314378728926e-05,
"loss": 2.1217,
"step": 10720
},
{
"epoch": 1.0593356549691937,
"grad_norm": 0.0,
"learning_rate": 2.2247006543009518e-05,
"loss": 2.0515,
"step": 10730
},
{
"epoch": 1.0600053576212163,
"grad_norm": 0.0,
"learning_rate": 2.2149850391407002e-05,
"loss": 2.0677,
"step": 10740
},
{
"epoch": 1.0606750602732387,
"grad_norm": 0.0,
"learning_rate": 2.2052846454928737e-05,
"loss": 2.1297,
"step": 10750
},
{
"epoch": 1.0613447629252613,
"grad_norm": 0.0,
"learning_rate": 2.1955995263750225e-05,
"loss": 2.1283,
"step": 10760
},
{
"epoch": 1.0620144655772836,
"grad_norm": 0.0,
"learning_rate": 2.1859297347212094e-05,
"loss": 2.0728,
"step": 10770
},
{
"epoch": 1.0626841682293062,
"grad_norm": 0.0,
"learning_rate": 2.1762753233817274e-05,
"loss": 2.1211,
"step": 10780
},
{
"epoch": 1.0633538708813286,
"grad_norm": 0.0,
"learning_rate": 2.1666363451228073e-05,
"loss": 2.2635,
"step": 10790
},
{
"epoch": 1.0640235735333512,
"grad_norm": 0.0,
"learning_rate": 2.1570128526263317e-05,
"loss": 2.1496,
"step": 10800
},
{
"epoch": 1.0646932761853738,
"grad_norm": 0.0,
"learning_rate": 2.1474048984895406e-05,
"loss": 2.1991,
"step": 10810
},
{
"epoch": 1.0653629788373962,
"grad_norm": 0.0,
"learning_rate": 2.137812535224755e-05,
"loss": 2.1058,
"step": 10820
},
{
"epoch": 1.0660326814894188,
"grad_norm": 0.0,
"learning_rate": 2.128235815259083e-05,
"loss": 2.2513,
"step": 10830
},
{
"epoch": 1.0667023841414411,
"grad_norm": 0.0,
"learning_rate": 2.1186747909341298e-05,
"loss": 2.0678,
"step": 10840
},
{
"epoch": 1.0673720867934637,
"grad_norm": 0.0,
"learning_rate": 2.1091295145057207e-05,
"loss": 2.0552,
"step": 10850
},
{
"epoch": 1.0680417894454861,
"grad_norm": 0.0,
"learning_rate": 2.0996000381436104e-05,
"loss": 2.1286,
"step": 10860
},
{
"epoch": 1.0687114920975087,
"grad_norm": 0.0,
"learning_rate": 2.090086413931194e-05,
"loss": 2.2085,
"step": 10870
},
{
"epoch": 1.0693811947495313,
"grad_norm": 0.0,
"learning_rate": 2.0805886938652302e-05,
"loss": 2.1108,
"step": 10880
},
{
"epoch": 1.0700508974015537,
"grad_norm": 0.0,
"learning_rate": 2.0711069298555564e-05,
"loss": 2.1562,
"step": 10890
},
{
"epoch": 1.0707206000535763,
"grad_norm": 0.0,
"learning_rate": 2.061641173724792e-05,
"loss": 2.1488,
"step": 10900
},
{
"epoch": 1.0713903027055987,
"grad_norm": 0.0,
"learning_rate": 2.0521914772080776e-05,
"loss": 2.1766,
"step": 10910
},
{
"epoch": 1.0720600053576212,
"grad_norm": 0.0,
"learning_rate": 2.0427578919527747e-05,
"loss": 2.2053,
"step": 10920
},
{
"epoch": 1.0727297080096436,
"grad_norm": 0.0,
"learning_rate": 2.0333404695181836e-05,
"loss": 2.1427,
"step": 10930
},
{
"epoch": 1.0733994106616662,
"grad_norm": 0.0,
"learning_rate": 2.0239392613752728e-05,
"loss": 2.1092,
"step": 10940
},
{
"epoch": 1.0740691133136888,
"grad_norm": 0.0,
"learning_rate": 2.0145543189063927e-05,
"loss": 2.153,
"step": 10950
},
{
"epoch": 1.0747388159657112,
"grad_norm": 0.0,
"learning_rate": 2.005185693404986e-05,
"loss": 2.143,
"step": 10960
},
{
"epoch": 1.0754085186177338,
"grad_norm": 0.0,
"learning_rate": 1.9958334360753228e-05,
"loss": 2.1103,
"step": 10970
},
{
"epoch": 1.0760782212697562,
"grad_norm": 0.0,
"learning_rate": 1.9864975980322098e-05,
"loss": 2.1149,
"step": 10980
},
{
"epoch": 1.0767479239217788,
"grad_norm": 0.0,
"learning_rate": 1.9771782303007118e-05,
"loss": 2.2324,
"step": 10990
},
{
"epoch": 1.0774176265738011,
"grad_norm": 0.0,
"learning_rate": 1.967875383815878e-05,
"loss": 2.1108,
"step": 11000
},
{
"epoch": 1.0780873292258237,
"grad_norm": 0.0,
"learning_rate": 1.9585891094224623e-05,
"loss": 2.2484,
"step": 11010
},
{
"epoch": 1.0787570318778463,
"grad_norm": 0.0,
"learning_rate": 1.9493194578746356e-05,
"loss": 2.1761,
"step": 11020
},
{
"epoch": 1.0794267345298687,
"grad_norm": 0.0,
"learning_rate": 1.9400664798357278e-05,
"loss": 2.1615,
"step": 11030
},
{
"epoch": 1.0800964371818913,
"grad_norm": 0.0,
"learning_rate": 1.930830225877933e-05,
"loss": 2.1355,
"step": 11040
},
{
"epoch": 1.0807661398339137,
"grad_norm": 0.0,
"learning_rate": 1.921610746482039e-05,
"loss": 2.0492,
"step": 11050
},
{
"epoch": 1.0814358424859363,
"grad_norm": 0.0,
"learning_rate": 1.9124080920371544e-05,
"loss": 2.1128,
"step": 11060
},
{
"epoch": 1.0821055451379586,
"grad_norm": 0.0,
"learning_rate": 1.903222312840433e-05,
"loss": 2.195,
"step": 11070
},
{
"epoch": 1.0827752477899812,
"grad_norm": 0.0,
"learning_rate": 1.8940534590967907e-05,
"loss": 2.1437,
"step": 11080
},
{
"epoch": 1.0834449504420038,
"grad_norm": 0.0,
"learning_rate": 1.8849015809186427e-05,
"loss": 2.0666,
"step": 11090
},
{
"epoch": 1.0841146530940262,
"grad_norm": 0.0,
"learning_rate": 1.875766728325625e-05,
"loss": 2.1769,
"step": 11100
},
{
"epoch": 1.0847843557460488,
"grad_norm": 0.0,
"learning_rate": 1.866648951244312e-05,
"loss": 2.109,
"step": 11110
},
{
"epoch": 1.0854540583980712,
"grad_norm": 0.0,
"learning_rate": 1.857548299507961e-05,
"loss": 2.162,
"step": 11120
},
{
"epoch": 1.0861237610500938,
"grad_norm": 0.0,
"learning_rate": 1.848464822856229e-05,
"loss": 2.1181,
"step": 11130
},
{
"epoch": 1.0867934637021162,
"grad_norm": 0.0,
"learning_rate": 1.8393985709348954e-05,
"loss": 2.1641,
"step": 11140
},
{
"epoch": 1.0874631663541388,
"grad_norm": 0.0,
"learning_rate": 1.8303495932956045e-05,
"loss": 2.0839,
"step": 11150
},
{
"epoch": 1.0881328690061614,
"grad_norm": 0.0,
"learning_rate": 1.8213179393955908e-05,
"loss": 2.1487,
"step": 11160
},
{
"epoch": 1.0888025716581837,
"grad_norm": 0.0,
"learning_rate": 1.812303658597396e-05,
"loss": 2.2031,
"step": 11170
},
{
"epoch": 1.0894722743102063,
"grad_norm": 0.0,
"learning_rate": 1.803306800168617e-05,
"loss": 2.0955,
"step": 11180
},
{
"epoch": 1.0901419769622287,
"grad_norm": 0.0,
"learning_rate": 1.7943274132816267e-05,
"loss": 2.1206,
"step": 11190
},
{
"epoch": 1.0908116796142513,
"grad_norm": 0.0,
"learning_rate": 1.7853655470133042e-05,
"loss": 2.0943,
"step": 11200
},
{
"epoch": 1.0914813822662737,
"grad_norm": 0.0,
"learning_rate": 1.7764212503447725e-05,
"loss": 2.2108,
"step": 11210
},
{
"epoch": 1.0921510849182963,
"grad_norm": 0.0,
"learning_rate": 1.7674945721611297e-05,
"loss": 2.1896,
"step": 11220
},
{
"epoch": 1.0928207875703189,
"grad_norm": 0.0,
"learning_rate": 1.7585855612511737e-05,
"loss": 2.1392,
"step": 11230
},
{
"epoch": 1.0934904902223412,
"grad_norm": 0.0,
"learning_rate": 1.749694266307148e-05,
"loss": 2.098,
"step": 11240
},
{
"epoch": 1.0941601928743638,
"grad_norm": 0.0,
"learning_rate": 1.7408207359244687e-05,
"loss": 2.1777,
"step": 11250
},
{
"epoch": 1.0948298955263862,
"grad_norm": 0.0,
"learning_rate": 1.7319650186014547e-05,
"loss": 2.0452,
"step": 11260
},
{
"epoch": 1.0954995981784088,
"grad_norm": 0.0,
"learning_rate": 1.723127162739071e-05,
"loss": 2.1313,
"step": 11270
},
{
"epoch": 1.0961693008304314,
"grad_norm": 0.0,
"learning_rate": 1.7143072166406648e-05,
"loss": 2.0861,
"step": 11280
},
{
"epoch": 1.0968390034824538,
"grad_norm": 0.0,
"learning_rate": 1.705505228511689e-05,
"loss": 2.1566,
"step": 11290
},
{
"epoch": 1.0975087061344764,
"grad_norm": 0.0,
"learning_rate": 1.6967212464594524e-05,
"loss": 2.1025,
"step": 11300
},
{
"epoch": 1.0981784087864987,
"grad_norm": 0.0,
"learning_rate": 1.687955318492852e-05,
"loss": 2.1918,
"step": 11310
},
{
"epoch": 1.0988481114385213,
"grad_norm": 0.0,
"learning_rate": 1.679207492522106e-05,
"loss": 2.0869,
"step": 11320
},
{
"epoch": 1.0995178140905437,
"grad_norm": 0.0,
"learning_rate": 1.6704778163585e-05,
"loss": 2.1169,
"step": 11330
},
{
"epoch": 1.1001875167425663,
"grad_norm": 0.0,
"learning_rate": 1.6617663377141214e-05,
"loss": 2.1397,
"step": 11340
},
{
"epoch": 1.1008572193945887,
"grad_norm": 0.0,
"learning_rate": 1.653073104201595e-05,
"loss": 2.0632,
"step": 11350
},
{
"epoch": 1.1015269220466113,
"grad_norm": 0.0,
"learning_rate": 1.6443981633338314e-05,
"loss": 2.1112,
"step": 11360
},
{
"epoch": 1.1021966246986339,
"grad_norm": 0.0,
"learning_rate": 1.6357415625237622e-05,
"loss": 2.2122,
"step": 11370
},
{
"epoch": 1.1028663273506563,
"grad_norm": 0.0,
"learning_rate": 1.6271033490840764e-05,
"loss": 2.1889,
"step": 11380
},
{
"epoch": 1.1035360300026789,
"grad_norm": 0.0,
"learning_rate": 1.6184835702269728e-05,
"loss": 2.1106,
"step": 11390
},
{
"epoch": 1.1042057326547012,
"grad_norm": 0.0,
"learning_rate": 1.6098822730638934e-05,
"loss": 2.1479,
"step": 11400
},
{
"epoch": 1.1048754353067238,
"grad_norm": 0.0,
"learning_rate": 1.6012995046052687e-05,
"loss": 2.1759,
"step": 11410
},
{
"epoch": 1.1055451379587464,
"grad_norm": 0.0,
"learning_rate": 1.592735311760259e-05,
"loss": 2.1718,
"step": 11420
},
{
"epoch": 1.1062148406107688,
"grad_norm": 0.0,
"learning_rate": 1.5841897413365038e-05,
"loss": 2.1863,
"step": 11430
},
{
"epoch": 1.1068845432627914,
"grad_norm": 0.0,
"learning_rate": 1.5756628400398543e-05,
"loss": 2.1882,
"step": 11440
},
{
"epoch": 1.1075542459148138,
"grad_norm": 0.0,
"learning_rate": 1.567154654474132e-05,
"loss": 2.1132,
"step": 11450
},
{
"epoch": 1.1082239485668364,
"grad_norm": 0.0,
"learning_rate": 1.558665231140867e-05,
"loss": 2.1689,
"step": 11460
},
{
"epoch": 1.1088936512188587,
"grad_norm": 0.0,
"learning_rate": 1.5501946164390397e-05,
"loss": 2.1887,
"step": 11470
},
{
"epoch": 1.1095633538708813,
"grad_norm": 0.0,
"learning_rate": 1.541742856664835e-05,
"loss": 2.1643,
"step": 11480
},
{
"epoch": 1.1102330565229037,
"grad_norm": 0.0,
"learning_rate": 1.533309998011389e-05,
"loss": 2.1496,
"step": 11490
},
{
"epoch": 1.1109027591749263,
"grad_norm": 0.0,
"learning_rate": 1.5248960865685268e-05,
"loss": 2.0605,
"step": 11500
},
{
"epoch": 1.111572461826949,
"grad_norm": 0.0,
"learning_rate": 1.5165011683225228e-05,
"loss": 2.1085,
"step": 11510
},
{
"epoch": 1.1122421644789713,
"grad_norm": 0.0,
"learning_rate": 1.5081252891558418e-05,
"loss": 2.039,
"step": 11520
},
{
"epoch": 1.1129118671309939,
"grad_norm": 0.0,
"learning_rate": 1.499768494846892e-05,
"loss": 2.189,
"step": 11530
},
{
"epoch": 1.1135815697830163,
"grad_norm": 0.0,
"learning_rate": 1.4914308310697712e-05,
"loss": 2.0645,
"step": 11540
},
{
"epoch": 1.1142512724350389,
"grad_norm": 0.0,
"learning_rate": 1.483112343394022e-05,
"loss": 2.1368,
"step": 11550
},
{
"epoch": 1.1149209750870614,
"grad_norm": 0.0,
"learning_rate": 1.4748130772843755e-05,
"loss": 2.097,
"step": 11560
},
{
"epoch": 1.1155906777390838,
"grad_norm": 0.0,
"learning_rate": 1.4665330781005105e-05,
"loss": 2.1421,
"step": 11570
},
{
"epoch": 1.1162603803911064,
"grad_norm": 0.0,
"learning_rate": 1.4582723910968037e-05,
"loss": 2.1087,
"step": 11580
},
{
"epoch": 1.1169300830431288,
"grad_norm": 0.0,
"learning_rate": 1.4500310614220763e-05,
"loss": 2.1139,
"step": 11590
},
{
"epoch": 1.1175997856951514,
"grad_norm": 0.0,
"learning_rate": 1.4418091341193552e-05,
"loss": 2.1825,
"step": 11600
},
{
"epoch": 1.1182694883471738,
"grad_norm": 0.0,
"learning_rate": 1.4336066541256249e-05,
"loss": 2.1644,
"step": 11610
},
{
"epoch": 1.1189391909991964,
"grad_norm": 0.0,
"learning_rate": 1.4254236662715759e-05,
"loss": 2.1694,
"step": 11620
},
{
"epoch": 1.119608893651219,
"grad_norm": 0.0,
"learning_rate": 1.4172602152813679e-05,
"loss": 2.1323,
"step": 11630
},
{
"epoch": 1.1202785963032413,
"grad_norm": 0.0,
"learning_rate": 1.409116345772381e-05,
"loss": 2.1212,
"step": 11640
},
{
"epoch": 1.120948298955264,
"grad_norm": 0.0,
"learning_rate": 1.4009921022549727e-05,
"loss": 2.2011,
"step": 11650
},
{
"epoch": 1.1216180016072863,
"grad_norm": 0.0,
"learning_rate": 1.392887529132234e-05,
"loss": 2.137,
"step": 11660
},
{
"epoch": 1.122287704259309,
"grad_norm": 0.0,
"learning_rate": 1.3848026706997496e-05,
"loss": 2.0782,
"step": 11670
},
{
"epoch": 1.1229574069113313,
"grad_norm": 0.0,
"learning_rate": 1.3767375711453479e-05,
"loss": 2.1234,
"step": 11680
},
{
"epoch": 1.1236271095633539,
"grad_norm": 0.0,
"learning_rate": 1.36869227454887e-05,
"loss": 2.0887,
"step": 11690
},
{
"epoch": 1.1242968122153765,
"grad_norm": 0.0,
"learning_rate": 1.3606668248819249e-05,
"loss": 2.1095,
"step": 11700
},
{
"epoch": 1.1249665148673988,
"grad_norm": 0.0,
"learning_rate": 1.3526612660076421e-05,
"loss": 2.0477,
"step": 11710
},
{
"epoch": 1.1256362175194214,
"grad_norm": 0.0,
"learning_rate": 1.344675641680443e-05,
"loss": 2.1671,
"step": 11720
},
{
"epoch": 1.1263059201714438,
"grad_norm": 0.0,
"learning_rate": 1.3367099955457962e-05,
"loss": 2.18,
"step": 11730
},
{
"epoch": 1.1269756228234664,
"grad_norm": 0.0,
"learning_rate": 1.3287643711399756e-05,
"loss": 2.1475,
"step": 11740
},
{
"epoch": 1.1276453254754888,
"grad_norm": 0.0,
"learning_rate": 1.3208388118898307e-05,
"loss": 2.1707,
"step": 11750
},
{
"epoch": 1.1283150281275114,
"grad_norm": 0.0,
"learning_rate": 1.3129333611125427e-05,
"loss": 2.2242,
"step": 11760
},
{
"epoch": 1.1289847307795338,
"grad_norm": 0.0,
"learning_rate": 1.3050480620153904e-05,
"loss": 2.2287,
"step": 11770
},
{
"epoch": 1.1296544334315564,
"grad_norm": 0.0,
"learning_rate": 1.2971829576955141e-05,
"loss": 2.15,
"step": 11780
},
{
"epoch": 1.130324136083579,
"grad_norm": 0.0,
"learning_rate": 1.2893380911396768e-05,
"loss": 2.0854,
"step": 11790
},
{
"epoch": 1.1309938387356013,
"grad_norm": 0.0,
"learning_rate": 1.2815135052240346e-05,
"loss": 2.1068,
"step": 11800
},
{
"epoch": 1.131663541387624,
"grad_norm": 0.0,
"learning_rate": 1.273709242713898e-05,
"loss": 2.1606,
"step": 11810
},
{
"epoch": 1.1323332440396463,
"grad_norm": 0.0,
"learning_rate": 1.2659253462635034e-05,
"loss": 2.2733,
"step": 11820
},
{
"epoch": 1.133002946691669,
"grad_norm": 0.0,
"learning_rate": 1.2581618584157706e-05,
"loss": 2.2681,
"step": 11830
},
{
"epoch": 1.1336726493436915,
"grad_norm": 0.0,
"learning_rate": 1.2504188216020796e-05,
"loss": 2.133,
"step": 11840
},
{
"epoch": 1.1343423519957139,
"grad_norm": 0.0,
"learning_rate": 1.2426962781420365e-05,
"loss": 2.1469,
"step": 11850
},
{
"epoch": 1.1350120546477365,
"grad_norm": 0.0,
"learning_rate": 1.2349942702432365e-05,
"loss": 2.098,
"step": 11860
},
{
"epoch": 1.1356817572997588,
"grad_norm": 0.0,
"learning_rate": 1.227312840001042e-05,
"loss": 2.1423,
"step": 11870
},
{
"epoch": 1.1363514599517814,
"grad_norm": 0.0,
"learning_rate": 1.2196520293983466e-05,
"loss": 2.1606,
"step": 11880
},
{
"epoch": 1.1370211626038038,
"grad_norm": 0.0,
"learning_rate": 1.2120118803053432e-05,
"loss": 2.1523,
"step": 11890
},
{
"epoch": 1.1376908652558264,
"grad_norm": 0.0,
"learning_rate": 1.2043924344793083e-05,
"loss": 2.0652,
"step": 11900
},
{
"epoch": 1.138360567907849,
"grad_norm": 0.0,
"learning_rate": 1.1967937335643553e-05,
"loss": 2.1047,
"step": 11910
},
{
"epoch": 1.1390302705598714,
"grad_norm": 0.0,
"learning_rate": 1.1892158190912207e-05,
"loss": 2.1319,
"step": 11920
},
{
"epoch": 1.139699973211894,
"grad_norm": 0.0,
"learning_rate": 1.1816587324770317e-05,
"loss": 2.0914,
"step": 11930
},
{
"epoch": 1.1403696758639164,
"grad_norm": 0.0,
"learning_rate": 1.1741225150250835e-05,
"loss": 2.1607,
"step": 11940
},
{
"epoch": 1.141039378515939,
"grad_norm": 0.0,
"learning_rate": 1.1666072079246043e-05,
"loss": 2.1669,
"step": 11950
},
{
"epoch": 1.1417090811679613,
"grad_norm": 0.0,
"learning_rate": 1.1591128522505423e-05,
"loss": 2.1226,
"step": 11960
},
{
"epoch": 1.142378783819984,
"grad_norm": 0.0,
"learning_rate": 1.1516394889633347e-05,
"loss": 2.2758,
"step": 11970
},
{
"epoch": 1.1430484864720065,
"grad_norm": 0.0,
"learning_rate": 1.1441871589086822e-05,
"loss": 2.1448,
"step": 11980
},
{
"epoch": 1.143718189124029,
"grad_norm": 0.0,
"learning_rate": 1.1367559028173307e-05,
"loss": 2.1372,
"step": 11990
},
{
"epoch": 1.1443878917760515,
"grad_norm": 0.0,
"learning_rate": 1.1293457613048465e-05,
"loss": 2.1874,
"step": 12000
},
{
"epoch": 1.1450575944280739,
"grad_norm": 0.0,
"learning_rate": 1.1219567748713894e-05,
"loss": 2.1447,
"step": 12010
},
{
"epoch": 1.1457272970800965,
"grad_norm": 0.0,
"learning_rate": 1.1145889839015056e-05,
"loss": 2.1918,
"step": 12020
},
{
"epoch": 1.146396999732119,
"grad_norm": 0.0,
"learning_rate": 1.1072424286638861e-05,
"loss": 2.1311,
"step": 12030
},
{
"epoch": 1.1470667023841414,
"grad_norm": 0.0,
"learning_rate": 1.0999171493111654e-05,
"loss": 2.2333,
"step": 12040
},
{
"epoch": 1.147736405036164,
"grad_norm": 0.0,
"learning_rate": 1.0926131858796924e-05,
"loss": 2.1357,
"step": 12050
},
{
"epoch": 1.1484061076881864,
"grad_norm": 0.0,
"learning_rate": 1.0853305782893153e-05,
"loss": 2.1856,
"step": 12060
},
{
"epoch": 1.149075810340209,
"grad_norm": 0.0,
"learning_rate": 1.0780693663431569e-05,
"loss": 2.1265,
"step": 12070
},
{
"epoch": 1.1497455129922314,
"grad_norm": 0.0,
"learning_rate": 1.0708295897274073e-05,
"loss": 2.0987,
"step": 12080
},
{
"epoch": 1.150415215644254,
"grad_norm": 0.0,
"learning_rate": 1.063611288011101e-05,
"loss": 2.2172,
"step": 12090
},
{
"epoch": 1.1510849182962763,
"grad_norm": 0.0,
"learning_rate": 1.0564145006458976e-05,
"loss": 2.1524,
"step": 12100
},
{
"epoch": 1.151754620948299,
"grad_norm": 0.0,
"learning_rate": 1.0492392669658735e-05,
"loss": 2.1419,
"step": 12110
},
{
"epoch": 1.1524243236003215,
"grad_norm": 0.0,
"learning_rate": 1.0420856261873041e-05,
"loss": 2.1354,
"step": 12120
},
{
"epoch": 1.153094026252344,
"grad_norm": 0.0,
"learning_rate": 1.0349536174084417e-05,
"loss": 2.2442,
"step": 12130
},
{
"epoch": 1.1537637289043665,
"grad_norm": 0.0,
"learning_rate": 1.0278432796093206e-05,
"loss": 2.0998,
"step": 12140
},
{
"epoch": 1.154433431556389,
"grad_norm": 0.0,
"learning_rate": 1.0207546516515215e-05,
"loss": 2.1719,
"step": 12150
},
{
"epoch": 1.1551031342084115,
"grad_norm": 0.0,
"learning_rate": 1.0136877722779748e-05,
"loss": 2.18,
"step": 12160
},
{
"epoch": 1.155772836860434,
"grad_norm": 0.0,
"learning_rate": 1.0066426801127448e-05,
"loss": 2.1423,
"step": 12170
},
{
"epoch": 1.1564425395124565,
"grad_norm": 0.0,
"learning_rate": 9.996194136608128e-06,
"loss": 2.1923,
"step": 12180
},
{
"epoch": 1.157112242164479,
"grad_norm": 0.0,
"learning_rate": 9.92618011307877e-06,
"loss": 2.0779,
"step": 12190
},
{
"epoch": 1.1577819448165014,
"grad_norm": 0.0,
"learning_rate": 9.856385113201344e-06,
"loss": 2.1922,
"step": 12200
},
{
"epoch": 1.158451647468524,
"grad_norm": 0.0,
"learning_rate": 9.786809518440765e-06,
"loss": 2.141,
"step": 12210
},
{
"epoch": 1.1591213501205464,
"grad_norm": 0.0,
"learning_rate": 9.717453709062752e-06,
"loss": 2.242,
"step": 12220
},
{
"epoch": 1.159791052772569,
"grad_norm": 0.0,
"learning_rate": 9.648318064131817e-06,
"loss": 2.1027,
"step": 12230
},
{
"epoch": 1.1604607554245914,
"grad_norm": 0.0,
"learning_rate": 9.579402961509171e-06,
"loss": 2.1743,
"step": 12240
},
{
"epoch": 1.161130458076614,
"grad_norm": 0.0,
"learning_rate": 9.510708777850602e-06,
"loss": 2.1378,
"step": 12250
},
{
"epoch": 1.1618001607286366,
"grad_norm": 0.0,
"learning_rate": 9.442235888604518e-06,
"loss": 2.1996,
"step": 12260
},
{
"epoch": 1.162469863380659,
"grad_norm": 0.0,
"learning_rate": 9.373984668009811e-06,
"loss": 2.1038,
"step": 12270
},
{
"epoch": 1.1631395660326815,
"grad_norm": 0.0,
"learning_rate": 9.305955489093853e-06,
"loss": 2.1266,
"step": 12280
},
{
"epoch": 1.163809268684704,
"grad_norm": 0.0,
"learning_rate": 9.238148723670454e-06,
"loss": 2.2141,
"step": 12290
},
{
"epoch": 1.1644789713367265,
"grad_norm": 0.0,
"learning_rate": 9.170564742337784e-06,
"loss": 2.1873,
"step": 12300
},
{
"epoch": 1.165148673988749,
"grad_norm": 0.0,
"learning_rate": 9.103203914476427e-06,
"loss": 2.1729,
"step": 12310
},
{
"epoch": 1.1658183766407715,
"grad_norm": 0.0,
"learning_rate": 9.0360666082473e-06,
"loss": 2.1411,
"step": 12320
},
{
"epoch": 1.166488079292794,
"grad_norm": 0.0,
"learning_rate": 8.969153190589691e-06,
"loss": 2.2096,
"step": 12330
},
{
"epoch": 1.1671577819448165,
"grad_norm": 0.0,
"learning_rate": 8.902464027219171e-06,
"loss": 2.1246,
"step": 12340
},
{
"epoch": 1.167827484596839,
"grad_norm": 0.0,
"learning_rate": 8.835999482625684e-06,
"loss": 2.1925,
"step": 12350
},
{
"epoch": 1.1684971872488614,
"grad_norm": 0.0,
"learning_rate": 8.769759920071536e-06,
"loss": 2.0945,
"step": 12360
},
{
"epoch": 1.169166889900884,
"grad_norm": 0.0,
"learning_rate": 8.703745701589332e-06,
"loss": 2.098,
"step": 12370
},
{
"epoch": 1.1698365925529064,
"grad_norm": 0.0,
"learning_rate": 8.637957187980116e-06,
"loss": 2.1349,
"step": 12380
},
{
"epoch": 1.170506295204929,
"grad_norm": 0.0,
"learning_rate": 8.572394738811318e-06,
"loss": 2.1894,
"step": 12390
},
{
"epoch": 1.1711759978569516,
"grad_norm": 0.0,
"learning_rate": 8.507058712414811e-06,
"loss": 2.1568,
"step": 12400
},
{
"epoch": 1.171845700508974,
"grad_norm": 0.0,
"learning_rate": 8.441949465884969e-06,
"loss": 2.1275,
"step": 12410
},
{
"epoch": 1.1725154031609966,
"grad_norm": 0.0,
"learning_rate": 8.377067355076668e-06,
"loss": 2.1202,
"step": 12420
},
{
"epoch": 1.173185105813019,
"grad_norm": 0.0,
"learning_rate": 8.312412734603392e-06,
"loss": 2.1644,
"step": 12430
},
{
"epoch": 1.1738548084650415,
"grad_norm": 0.0,
"learning_rate": 8.247985957835286e-06,
"loss": 2.2205,
"step": 12440
},
{
"epoch": 1.1745245111170641,
"grad_norm": 0.0,
"learning_rate": 8.183787376897212e-06,
"loss": 2.1565,
"step": 12450
},
{
"epoch": 1.1751942137690865,
"grad_norm": 0.0,
"learning_rate": 8.119817342666785e-06,
"loss": 2.0464,
"step": 12460
},
{
"epoch": 1.175863916421109,
"grad_norm": 0.0,
"learning_rate": 8.056076204772556e-06,
"loss": 2.0674,
"step": 12470
},
{
"epoch": 1.1765336190731315,
"grad_norm": 0.0,
"learning_rate": 7.99256431159201e-06,
"loss": 2.0096,
"step": 12480
},
{
"epoch": 1.177203321725154,
"grad_norm": 0.0,
"learning_rate": 7.92928201024969e-06,
"loss": 2.1404,
"step": 12490
},
{
"epoch": 1.1778730243771764,
"grad_norm": 0.0,
"learning_rate": 7.86622964661532e-06,
"loss": 2.1675,
"step": 12500
},
{
"epoch": 1.178542727029199,
"grad_norm": 0.0,
"learning_rate": 7.803407565301891e-06,
"loss": 2.1568,
"step": 12510
},
{
"epoch": 1.1792124296812214,
"grad_norm": 0.0,
"learning_rate": 7.740816109663785e-06,
"loss": 2.1097,
"step": 12520
},
{
"epoch": 1.179882132333244,
"grad_norm": 0.0,
"learning_rate": 7.678455621794906e-06,
"loss": 2.1512,
"step": 12530
},
{
"epoch": 1.1805518349852666,
"grad_norm": 0.0,
"learning_rate": 7.616326442526784e-06,
"loss": 2.0985,
"step": 12540
},
{
"epoch": 1.181221537637289,
"grad_norm": 0.0,
"learning_rate": 7.5544289114267365e-06,
"loss": 2.175,
"step": 12550
},
{
"epoch": 1.1818912402893116,
"grad_norm": 0.0,
"learning_rate": 7.492763366796041e-06,
"loss": 2.0826,
"step": 12560
},
{
"epoch": 1.182560942941334,
"grad_norm": 0.0,
"learning_rate": 7.431330145667986e-06,
"loss": 2.2003,
"step": 12570
},
{
"epoch": 1.1832306455933566,
"grad_norm": 0.0,
"learning_rate": 7.370129583806151e-06,
"loss": 1.9417,
"step": 12580
},
{
"epoch": 1.1839003482453792,
"grad_norm": 0.0,
"learning_rate": 7.309162015702492e-06,
"loss": 2.2006,
"step": 12590
},
{
"epoch": 1.1845700508974015,
"grad_norm": 0.0,
"learning_rate": 7.2484277745755615e-06,
"loss": 2.1589,
"step": 12600
},
{
"epoch": 1.1852397535494241,
"grad_norm": 0.0,
"learning_rate": 7.187927192368604e-06,
"loss": 2.1157,
"step": 12610
},
{
"epoch": 1.1859094562014465,
"grad_norm": 0.0,
"learning_rate": 7.127660599747859e-06,
"loss": 2.1022,
"step": 12620
},
{
"epoch": 1.186579158853469,
"grad_norm": 0.0,
"learning_rate": 7.067628326100667e-06,
"loss": 2.0934,
"step": 12630
},
{
"epoch": 1.1872488615054915,
"grad_norm": 0.0,
"learning_rate": 7.0078306995337076e-06,
"loss": 2.202,
"step": 12640
},
{
"epoch": 1.187918564157514,
"grad_norm": 0.0,
"learning_rate": 6.948268046871203e-06,
"loss": 2.1258,
"step": 12650
},
{
"epoch": 1.1885882668095364,
"grad_norm": 0.0,
"learning_rate": 6.88894069365309e-06,
"loss": 2.1314,
"step": 12660
},
{
"epoch": 1.189257969461559,
"grad_norm": 0.0,
"learning_rate": 6.8298489641333005e-06,
"loss": 2.2186,
"step": 12670
},
{
"epoch": 1.1899276721135816,
"grad_norm": 0.0,
"learning_rate": 6.77099318127799e-06,
"loss": 2.1139,
"step": 12680
},
{
"epoch": 1.190597374765604,
"grad_norm": 0.0,
"learning_rate": 6.7123736667636776e-06,
"loss": 2.1249,
"step": 12690
},
{
"epoch": 1.1912670774176266,
"grad_norm": 0.0,
"learning_rate": 6.6539907409756185e-06,
"loss": 2.0937,
"step": 12700
},
{
"epoch": 1.191936780069649,
"grad_norm": 0.0,
"learning_rate": 6.5958447230059675e-06,
"loss": 2.1653,
"step": 12710
},
{
"epoch": 1.1926064827216716,
"grad_norm": 0.0,
"learning_rate": 6.537935930652067e-06,
"loss": 2.1334,
"step": 12720
},
{
"epoch": 1.1932761853736942,
"grad_norm": 0.0,
"learning_rate": 6.480264680414683e-06,
"loss": 2.1211,
"step": 12730
},
{
"epoch": 1.1939458880257166,
"grad_norm": 0.0,
"learning_rate": 6.422831287496306e-06,
"loss": 2.1794,
"step": 12740
},
{
"epoch": 1.1946155906777391,
"grad_norm": 0.0,
"learning_rate": 6.365636065799424e-06,
"loss": 2.1811,
"step": 12750
},
{
"epoch": 1.1952852933297615,
"grad_norm": 0.0,
"learning_rate": 6.308679327924788e-06,
"loss": 2.1791,
"step": 12760
},
{
"epoch": 1.1959549959817841,
"grad_norm": 0.0,
"learning_rate": 6.25196138516973e-06,
"loss": 1.9809,
"step": 12770
},
{
"epoch": 1.1966246986338065,
"grad_norm": 0.0,
"learning_rate": 6.19548254752641e-06,
"loss": 2.0739,
"step": 12780
},
{
"epoch": 1.197294401285829,
"grad_norm": 0.0,
"learning_rate": 6.139243123680194e-06,
"loss": 2.1205,
"step": 12790
},
{
"epoch": 1.1979641039378517,
"grad_norm": 0.0,
"learning_rate": 6.083243421007934e-06,
"loss": 2.1492,
"step": 12800
},
{
"epoch": 1.198633806589874,
"grad_norm": 0.0,
"learning_rate": 6.027483745576246e-06,
"loss": 2.0709,
"step": 12810
},
{
"epoch": 1.1993035092418967,
"grad_norm": 0.0,
"learning_rate": 5.971964402139907e-06,
"loss": 2.1815,
"step": 12820
},
{
"epoch": 1.199973211893919,
"grad_norm": 0.0,
"learning_rate": 5.9166856941401505e-06,
"loss": 2.0973,
"step": 12830
},
{
"epoch": 1.2006429145459416,
"grad_norm": 0.0,
"learning_rate": 5.861647923703028e-06,
"loss": 2.1153,
"step": 12840
},
{
"epoch": 1.201312617197964,
"grad_norm": 0.0,
"learning_rate": 5.806851391637719e-06,
"loss": 2.1091,
"step": 12850
},
{
"epoch": 1.2019823198499866,
"grad_norm": 0.0,
"learning_rate": 5.752296397434931e-06,
"loss": 2.1239,
"step": 12860
},
{
"epoch": 1.2026520225020092,
"grad_norm": 0.0,
"learning_rate": 5.697983239265242e-06,
"loss": 2.0781,
"step": 12870
},
{
"epoch": 1.2033217251540316,
"grad_norm": 0.0,
"learning_rate": 5.6439122139774815e-06,
"loss": 2.0746,
"step": 12880
},
{
"epoch": 1.2039914278060542,
"grad_norm": 0.0,
"learning_rate": 5.590083617097097e-06,
"loss": 2.1967,
"step": 12890
},
{
"epoch": 1.2046611304580765,
"grad_norm": 0.0,
"learning_rate": 5.5364977428245255e-06,
"loss": 2.0922,
"step": 12900
},
{
"epoch": 1.2053308331100991,
"grad_norm": 0.0,
"learning_rate": 5.48315488403362e-06,
"loss": 2.138,
"step": 12910
},
{
"epoch": 1.2060005357621217,
"grad_norm": 0.0,
"learning_rate": 5.430055332270045e-06,
"loss": 2.0713,
"step": 12920
},
{
"epoch": 1.2066702384141441,
"grad_norm": 0.0,
"learning_rate": 5.37719937774962e-06,
"loss": 2.0803,
"step": 12930
},
{
"epoch": 1.2073399410661667,
"grad_norm": 0.0,
"learning_rate": 5.324587309356832e-06,
"loss": 2.1192,
"step": 12940
},
{
"epoch": 1.208009643718189,
"grad_norm": 0.0,
"learning_rate": 5.272219414643203e-06,
"loss": 2.1146,
"step": 12950
},
{
"epoch": 1.2086793463702117,
"grad_norm": 0.0,
"learning_rate": 5.2200959798256785e-06,
"loss": 2.175,
"step": 12960
},
{
"epoch": 1.209349049022234,
"grad_norm": 0.0,
"learning_rate": 5.168217289785154e-06,
"loss": 2.1732,
"step": 12970
},
{
"epoch": 1.2100187516742567,
"grad_norm": 0.0,
"learning_rate": 5.116583628064858e-06,
"loss": 2.115,
"step": 12980
},
{
"epoch": 1.210688454326279,
"grad_norm": 0.0,
"learning_rate": 5.0651952768688115e-06,
"loss": 2.0194,
"step": 12990
},
{
"epoch": 1.2113581569783016,
"grad_norm": 0.0,
"learning_rate": 5.0140525170602744e-06,
"loss": 2.2069,
"step": 13000
},
{
"epoch": 1.2120278596303242,
"grad_norm": 0.0,
"learning_rate": 4.963155628160276e-06,
"loss": 2.0356,
"step": 13010
},
{
"epoch": 1.2126975622823466,
"grad_norm": 0.0,
"learning_rate": 4.9125048883459705e-06,
"loss": 2.1624,
"step": 13020
},
{
"epoch": 1.2133672649343692,
"grad_norm": 0.0,
"learning_rate": 4.862100574449224e-06,
"loss": 2.1486,
"step": 13030
},
{
"epoch": 1.2140369675863916,
"grad_norm": 0.0,
"learning_rate": 4.811942961955074e-06,
"loss": 2.1325,
"step": 13040
},
{
"epoch": 1.2147066702384142,
"grad_norm": 0.0,
"learning_rate": 4.762032325000154e-06,
"loss": 2.215,
"step": 13050
},
{
"epoch": 1.2153763728904368,
"grad_norm": 0.0,
"learning_rate": 4.712368936371309e-06,
"loss": 2.1144,
"step": 13060
},
{
"epoch": 1.2160460755424591,
"grad_norm": 0.0,
"learning_rate": 4.662953067504022e-06,
"loss": 2.1361,
"step": 13070
},
{
"epoch": 1.2167157781944817,
"grad_norm": 0.0,
"learning_rate": 4.613784988480946e-06,
"loss": 2.1634,
"step": 13080
},
{
"epoch": 1.217385480846504,
"grad_norm": 0.0,
"learning_rate": 4.564864968030464e-06,
"loss": 2.1118,
"step": 13090
},
{
"epoch": 1.2180551834985267,
"grad_norm": 0.0,
"learning_rate": 4.516193273525165e-06,
"loss": 2.1313,
"step": 13100
},
{
"epoch": 1.218724886150549,
"grad_norm": 0.0,
"learning_rate": 4.46777017098044e-06,
"loss": 2.1132,
"step": 13110
},
{
"epoch": 1.2193945888025717,
"grad_norm": 0.0,
"learning_rate": 4.41959592505296e-06,
"loss": 2.1188,
"step": 13120
},
{
"epoch": 1.220064291454594,
"grad_norm": 0.0,
"learning_rate": 4.371670799039335e-06,
"loss": 2.17,
"step": 13130
},
{
"epoch": 1.2207339941066166,
"grad_norm": 0.0,
"learning_rate": 4.32399505487453e-06,
"loss": 2.1296,
"step": 13140
},
{
"epoch": 1.2214036967586392,
"grad_norm": 0.0,
"learning_rate": 4.276568953130561e-06,
"loss": 2.1647,
"step": 13150
},
{
"epoch": 1.2220733994106616,
"grad_norm": 0.0,
"learning_rate": 4.229392753015027e-06,
"loss": 2.1463,
"step": 13160
},
{
"epoch": 1.2227431020626842,
"grad_norm": 0.0,
"learning_rate": 4.182466712369648e-06,
"loss": 2.1315,
"step": 13170
},
{
"epoch": 1.2234128047147066,
"grad_norm": 0.0,
"learning_rate": 4.135791087668933e-06,
"loss": 2.1308,
"step": 13180
},
{
"epoch": 1.2240825073667292,
"grad_norm": 0.0,
"learning_rate": 4.089366134018735e-06,
"loss": 2.1655,
"step": 13190
},
{
"epoch": 1.2247522100187518,
"grad_norm": 0.0,
"learning_rate": 4.043192105154842e-06,
"loss": 2.0646,
"step": 13200
},
{
"epoch": 1.2254219126707742,
"grad_norm": 0.0,
"learning_rate": 3.997269253441638e-06,
"loss": 2.1805,
"step": 13210
},
{
"epoch": 1.2260916153227968,
"grad_norm": 0.0,
"learning_rate": 3.951597829870685e-06,
"loss": 2.0898,
"step": 13220
},
{
"epoch": 1.2267613179748191,
"grad_norm": 0.0,
"learning_rate": 3.90617808405937e-06,
"loss": 2.2064,
"step": 13230
},
{
"epoch": 1.2274310206268417,
"grad_norm": 0.0,
"learning_rate": 3.861010264249504e-06,
"loss": 2.1149,
"step": 13240
},
{
"epoch": 1.228100723278864,
"grad_norm": 0.0,
"learning_rate": 3.816094617306054e-06,
"loss": 2.2301,
"step": 13250
},
{
"epoch": 1.2287704259308867,
"grad_norm": 0.0,
"learning_rate": 3.7714313887156793e-06,
"loss": 2.2351,
"step": 13260
},
{
"epoch": 1.229440128582909,
"grad_norm": 0.0,
"learning_rate": 3.7270208225854696e-06,
"loss": 2.1545,
"step": 13270
},
{
"epoch": 1.2301098312349317,
"grad_norm": 0.0,
"learning_rate": 3.682863161641592e-06,
"loss": 2.2179,
"step": 13280
},
{
"epoch": 1.2307795338869543,
"grad_norm": 0.0,
"learning_rate": 3.638958647227936e-06,
"loss": 2.063,
"step": 13290
},
{
"epoch": 1.2314492365389766,
"grad_norm": 0.0,
"learning_rate": 3.595307519304841e-06,
"loss": 2.1517,
"step": 13300
},
{
"epoch": 1.2321189391909992,
"grad_norm": 0.0,
"learning_rate": 3.5519100164477625e-06,
"loss": 2.0889,
"step": 13310
},
{
"epoch": 1.2327886418430216,
"grad_norm": 0.0,
"learning_rate": 3.5087663758459397e-06,
"loss": 2.2184,
"step": 13320
},
{
"epoch": 1.2334583444950442,
"grad_norm": 0.0,
"learning_rate": 3.4658768333011514e-06,
"loss": 2.075,
"step": 13330
},
{
"epoch": 1.2341280471470668,
"grad_norm": 0.0,
"learning_rate": 3.423241623226403e-06,
"loss": 2.1403,
"step": 13340
},
{
"epoch": 1.2347977497990892,
"grad_norm": 0.0,
"learning_rate": 3.380860978644623e-06,
"loss": 2.2063,
"step": 13350
},
{
"epoch": 1.2354674524511118,
"grad_norm": 0.0,
"learning_rate": 3.3387351311874314e-06,
"loss": 2.1884,
"step": 13360
},
{
"epoch": 1.2361371551031342,
"grad_norm": 0.0,
"learning_rate": 3.2968643110938444e-06,
"loss": 2.2006,
"step": 13370
},
{
"epoch": 1.2368068577551568,
"grad_norm": 0.0,
"learning_rate": 3.255248747209033e-06,
"loss": 2.0195,
"step": 13380
},
{
"epoch": 1.2374765604071791,
"grad_norm": 0.0,
"learning_rate": 3.2138886669830504e-06,
"loss": 2.0719,
"step": 13390
},
{
"epoch": 1.2381462630592017,
"grad_norm": 0.0,
"learning_rate": 3.1727842964696274e-06,
"loss": 2.1849,
"step": 13400
},
{
"epoch": 1.238815965711224,
"grad_norm": 0.0,
"learning_rate": 3.1319358603248694e-06,
"loss": 2.071,
"step": 13410
},
{
"epoch": 1.2394856683632467,
"grad_norm": 0.0,
"learning_rate": 3.091343581806111e-06,
"loss": 2.0483,
"step": 13420
},
{
"epoch": 1.2401553710152693,
"grad_norm": 0.0,
"learning_rate": 3.0510076827706456e-06,
"loss": 2.0878,
"step": 13430
},
{
"epoch": 1.2408250736672917,
"grad_norm": 0.0,
"learning_rate": 3.010928383674516e-06,
"loss": 2.1341,
"step": 13440
},
{
"epoch": 1.2414947763193143,
"grad_norm": 0.0,
"learning_rate": 2.971105903571314e-06,
"loss": 2.1568,
"step": 13450
},
{
"epoch": 1.2421644789713366,
"grad_norm": 0.0,
"learning_rate": 2.9315404601110096e-06,
"loss": 2.1335,
"step": 13460
},
{
"epoch": 1.2428341816233592,
"grad_norm": 0.0,
"learning_rate": 2.8922322695386974e-06,
"loss": 2.1339,
"step": 13470
},
{
"epoch": 1.2435038842753818,
"grad_norm": 0.0,
"learning_rate": 2.853181546693501e-06,
"loss": 2.203,
"step": 13480
},
{
"epoch": 1.2441735869274042,
"grad_norm": 0.0,
"learning_rate": 2.8143885050073148e-06,
"loss": 2.0968,
"step": 13490
},
{
"epoch": 1.2448432895794268,
"grad_norm": 0.0,
"learning_rate": 2.7758533565036993e-06,
"loss": 2.1765,
"step": 13500
},
{
"epoch": 1.2455129922314492,
"grad_norm": 0.0,
"learning_rate": 2.7375763117966857e-06,
"loss": 2.088,
"step": 13510
},
{
"epoch": 1.2461826948834718,
"grad_norm": 0.0,
"learning_rate": 2.699557580089662e-06,
"loss": 2.1812,
"step": 13520
},
{
"epoch": 1.2468523975354942,
"grad_norm": 0.0,
"learning_rate": 2.661797369174157e-06,
"loss": 2.1773,
"step": 13530
},
{
"epoch": 1.2475221001875167,
"grad_norm": 0.0,
"learning_rate": 2.624295885428796e-06,
"loss": 2.131,
"step": 13540
},
{
"epoch": 1.2481918028395391,
"grad_norm": 0.0,
"learning_rate": 2.5870533338181235e-06,
"loss": 2.1509,
"step": 13550
},
{
"epoch": 1.2488615054915617,
"grad_norm": 0.0,
"learning_rate": 2.5500699178914623e-06,
"loss": 2.1224,
"step": 13560
},
{
"epoch": 1.2495312081435843,
"grad_norm": 0.0,
"learning_rate": 2.513345839781861e-06,
"loss": 2.1377,
"step": 13570
},
{
"epoch": 1.2502009107956067,
"grad_norm": 0.0,
"learning_rate": 2.4768813002049416e-06,
"loss": 2.0525,
"step": 13580
},
{
"epoch": 1.2508706134476293,
"grad_norm": 0.0,
"learning_rate": 2.440676498457811e-06,
"loss": 2.1756,
"step": 13590
},
{
"epoch": 1.2515403160996517,
"grad_norm": 0.0,
"learning_rate": 2.4047316324179837e-06,
"loss": 2.1819,
"step": 13600
},
{
"epoch": 1.2522100187516743,
"grad_norm": 0.0,
"learning_rate": 2.3690468985423096e-06,
"loss": 2.0861,
"step": 13610
},
{
"epoch": 1.2528797214036969,
"grad_norm": 0.0,
"learning_rate": 2.3336224918658446e-06,
"loss": 2.1533,
"step": 13620
},
{
"epoch": 1.2535494240557192,
"grad_norm": 0.0,
"learning_rate": 2.2984586060008696e-06,
"loss": 2.1075,
"step": 13630
},
{
"epoch": 1.2542191267077418,
"grad_norm": 0.0,
"learning_rate": 2.2635554331357724e-06,
"loss": 2.158,
"step": 13640
},
{
"epoch": 1.2548888293597642,
"grad_norm": 0.0,
"learning_rate": 2.2289131640339913e-06,
"loss": 2.0522,
"step": 13650
},
{
"epoch": 1.2555585320117868,
"grad_norm": 0.0,
"learning_rate": 2.1945319880330274e-06,
"loss": 2.1394,
"step": 13660
},
{
"epoch": 1.2562282346638094,
"grad_norm": 0.0,
"learning_rate": 2.160412093043357e-06,
"loss": 2.1565,
"step": 13670
},
{
"epoch": 1.2568979373158318,
"grad_norm": 0.0,
"learning_rate": 2.126553665547426e-06,
"loss": 2.1117,
"step": 13680
},
{
"epoch": 1.2575676399678541,
"grad_norm": 0.0,
"learning_rate": 2.0929568905986242e-06,
"loss": 2.1721,
"step": 13690
},
{
"epoch": 1.2582373426198767,
"grad_norm": 0.0,
"learning_rate": 2.0596219518202953e-06,
"loss": 2.1621,
"step": 13700
},
{
"epoch": 1.2589070452718993,
"grad_norm": 0.0,
"learning_rate": 2.02654903140469e-06,
"loss": 2.1366,
"step": 13710
},
{
"epoch": 1.2595767479239217,
"grad_norm": 0.0,
"learning_rate": 1.9937383101120255e-06,
"loss": 2.0949,
"step": 13720
},
{
"epoch": 1.2602464505759443,
"grad_norm": 0.0,
"learning_rate": 1.96118996726945e-06,
"loss": 2.1662,
"step": 13730
},
{
"epoch": 1.2609161532279667,
"grad_norm": 0.0,
"learning_rate": 1.928904180770069e-06,
"loss": 2.1523,
"step": 13740
},
{
"epoch": 1.2615858558799893,
"grad_norm": 0.0,
"learning_rate": 1.8968811270720256e-06,
"loss": 2.1152,
"step": 13750
},
{
"epoch": 1.2622555585320119,
"grad_norm": 0.0,
"learning_rate": 1.8651209811974657e-06,
"loss": 2.1852,
"step": 13760
},
{
"epoch": 1.2629252611840343,
"grad_norm": 0.0,
"learning_rate": 1.8336239167316027e-06,
"loss": 2.1504,
"step": 13770
},
{
"epoch": 1.2635949638360569,
"grad_norm": 0.0,
"learning_rate": 1.802390105821805e-06,
"loss": 2.1188,
"step": 13780
},
{
"epoch": 1.2642646664880792,
"grad_norm": 0.0,
"learning_rate": 1.771419719176609e-06,
"loss": 2.1576,
"step": 13790
},
{
"epoch": 1.2649343691401018,
"grad_norm": 0.0,
"learning_rate": 1.740712926064808e-06,
"loss": 2.1234,
"step": 13800
},
{
"epoch": 1.2656040717921244,
"grad_norm": 0.0,
"learning_rate": 1.7102698943145256e-06,
"loss": 2.1045,
"step": 13810
},
{
"epoch": 1.2662737744441468,
"grad_norm": 0.0,
"learning_rate": 1.6800907903122998e-06,
"loss": 2.15,
"step": 13820
},
{
"epoch": 1.2669434770961692,
"grad_norm": 0.0,
"learning_rate": 1.6501757790021555e-06,
"loss": 2.1931,
"step": 13830
},
{
"epoch": 1.2676131797481918,
"grad_norm": 0.0,
"learning_rate": 1.620525023884739e-06,
"loss": 2.1509,
"step": 13840
},
{
"epoch": 1.2682828824002144,
"grad_norm": 0.0,
"learning_rate": 1.5911386870163913e-06,
"loss": 2.1728,
"step": 13850
},
{
"epoch": 1.2689525850522367,
"grad_norm": 0.0,
"learning_rate": 1.5620169290082642e-06,
"loss": 2.1134,
"step": 13860
},
{
"epoch": 1.2696222877042593,
"grad_norm": 0.0,
"learning_rate": 1.5331599090254778e-06,
"loss": 2.1672,
"step": 13870
},
{
"epoch": 1.2702919903562817,
"grad_norm": 0.0,
"learning_rate": 1.5045677847862094e-06,
"loss": 2.1091,
"step": 13880
},
{
"epoch": 1.2709616930083043,
"grad_norm": 0.0,
"learning_rate": 1.476240712560839e-06,
"loss": 2.169,
"step": 13890
},
{
"epoch": 1.271631395660327,
"grad_norm": 0.0,
"learning_rate": 1.4481788471711222e-06,
"loss": 2.2051,
"step": 13900
},
{
"epoch": 1.2723010983123493,
"grad_norm": 0.0,
"learning_rate": 1.4203823419893237e-06,
"loss": 2.1638,
"step": 13910
},
{
"epoch": 1.2729708009643719,
"grad_norm": 0.0,
"learning_rate": 1.3928513489373518e-06,
"loss": 2.1165,
"step": 13920
},
{
"epoch": 1.2736405036163942,
"grad_norm": 0.0,
"learning_rate": 1.365586018485998e-06,
"loss": 2.182,
"step": 13930
},
{
"epoch": 1.2743102062684168,
"grad_norm": 0.0,
"learning_rate": 1.3385864996540488e-06,
"loss": 2.1222,
"step": 13940
},
{
"epoch": 1.2749799089204394,
"grad_norm": 0.0,
"learning_rate": 1.311852940007502e-06,
"loss": 2.2216,
"step": 13950
},
{
"epoch": 1.2756496115724618,
"grad_norm": 0.0,
"learning_rate": 1.2853854856587521e-06,
"loss": 2.1033,
"step": 13960
},
{
"epoch": 1.2763193142244842,
"grad_norm": 0.0,
"learning_rate": 1.2591842812658017e-06,
"loss": 2.0526,
"step": 13970
},
{
"epoch": 1.2769890168765068,
"grad_norm": 0.0,
"learning_rate": 1.2332494700314556e-06,
"loss": 2.1177,
"step": 13980
},
{
"epoch": 1.2776587195285294,
"grad_norm": 0.0,
"learning_rate": 1.207581193702534e-06,
"loss": 2.1358,
"step": 13990
},
{
"epoch": 1.2783284221805518,
"grad_norm": 0.0,
"learning_rate": 1.1821795925691436e-06,
"loss": 2.104,
"step": 14000
},
{
"epoch": 1.2789981248325744,
"grad_norm": 0.0,
"learning_rate": 1.1570448054638417e-06,
"loss": 2.1815,
"step": 14010
},
{
"epoch": 1.2796678274845967,
"grad_norm": 0.0,
"learning_rate": 1.1321769697609286e-06,
"loss": 2.1766,
"step": 14020
},
{
"epoch": 1.2803375301366193,
"grad_norm": 0.0,
"learning_rate": 1.1075762213756835e-06,
"loss": 2.1828,
"step": 14030
},
{
"epoch": 1.281007232788642,
"grad_norm": 0.0,
"learning_rate": 1.083242694763603e-06,
"loss": 2.0712,
"step": 14040
},
{
"epoch": 1.2816769354406643,
"grad_norm": 0.0,
"learning_rate": 1.0591765229197026e-06,
"loss": 2.1697,
"step": 14050
},
{
"epoch": 1.282346638092687,
"grad_norm": 0.0,
"learning_rate": 1.035377837377749e-06,
"loss": 2.21,
"step": 14060
},
{
"epoch": 1.2830163407447093,
"grad_norm": 0.0,
"learning_rate": 1.0118467682095733e-06,
"loss": 2.1687,
"step": 14070
},
{
"epoch": 1.2836860433967319,
"grad_norm": 0.0,
"learning_rate": 9.88583444024349e-07,
"loss": 2.1748,
"step": 14080
},
{
"epoch": 1.2843557460487545,
"grad_norm": 0.0,
"learning_rate": 9.655879919678811e-07,
"loss": 2.1528,
"step": 14090
},
{
"epoch": 1.2850254487007768,
"grad_norm": 0.0,
"learning_rate": 9.428605377219235e-07,
"loss": 2.1177,
"step": 14100
},
{
"epoch": 1.2856951513527994,
"grad_norm": 0.0,
"learning_rate": 9.204012055034739e-07,
"loss": 2.1339,
"step": 14110
},
{
"epoch": 1.2863648540048218,
"grad_norm": 0.0,
"learning_rate": 8.982101180641355e-07,
"loss": 2.2136,
"step": 14120
},
{
"epoch": 1.2870345566568444,
"grad_norm": 0.0,
"learning_rate": 8.762873966893959e-07,
"loss": 2.1921,
"step": 14130
},
{
"epoch": 1.2877042593088668,
"grad_norm": 0.0,
"learning_rate": 8.546331611979818e-07,
"loss": 2.204,
"step": 14140
},
{
"epoch": 1.2883739619608894,
"grad_norm": 0.0,
"learning_rate": 8.33247529941239e-07,
"loss": 2.1083,
"step": 14150
},
{
"epoch": 1.2890436646129118,
"grad_norm": 0.0,
"learning_rate": 8.121306198024204e-07,
"loss": 2.0582,
"step": 14160
},
{
"epoch": 1.2897133672649344,
"grad_norm": 0.0,
"learning_rate": 7.912825461960982e-07,
"loss": 2.1152,
"step": 14170
},
{
"epoch": 1.290383069916957,
"grad_norm": 0.0,
"learning_rate": 7.707034230675314e-07,
"loss": 2.1319,
"step": 14180
},
{
"epoch": 1.2910527725689793,
"grad_norm": 0.0,
"learning_rate": 7.50393362891999e-07,
"loss": 2.1414,
"step": 14190
},
{
"epoch": 1.291722475221002,
"grad_norm": 0.0,
"learning_rate": 7.303524766742398e-07,
"loss": 2.0645,
"step": 14200
},
{
"epoch": 1.2923921778730243,
"grad_norm": 0.0,
"learning_rate": 7.105808739478082e-07,
"loss": 2.0771,
"step": 14210
},
{
"epoch": 1.293061880525047,
"grad_norm": 0.0,
"learning_rate": 6.910786627744914e-07,
"loss": 2.2023,
"step": 14220
},
{
"epoch": 1.2937315831770695,
"grad_norm": 0.0,
"learning_rate": 6.718459497437213e-07,
"loss": 2.0567,
"step": 14230
},
{
"epoch": 1.2944012858290919,
"grad_norm": 0.0,
"learning_rate": 6.5288283997198e-07,
"loss": 2.1083,
"step": 14240
},
{
"epoch": 1.2950709884811145,
"grad_norm": 0.0,
"learning_rate": 6.341894371022339e-07,
"loss": 2.1249,
"step": 14250
},
{
"epoch": 1.2957406911331368,
"grad_norm": 0.0,
"learning_rate": 6.15765843303362e-07,
"loss": 2.1732,
"step": 14260
},
{
"epoch": 1.2964103937851594,
"grad_norm": 0.0,
"learning_rate": 5.97612159269606e-07,
"loss": 2.1813,
"step": 14270
},
{
"epoch": 1.297080096437182,
"grad_norm": 0.0,
"learning_rate": 5.797284842200102e-07,
"loss": 2.0738,
"step": 14280
},
{
"epoch": 1.2977497990892044,
"grad_norm": 0.0,
"learning_rate": 5.621149158978822e-07,
"loss": 2.1295,
"step": 14290
},
{
"epoch": 1.2984195017412268,
"grad_norm": 0.0,
"learning_rate": 5.447715505702722e-07,
"loss": 2.1226,
"step": 14300
},
{
"epoch": 1.2990892043932494,
"grad_norm": 0.0,
"learning_rate": 5.276984830274167e-07,
"loss": 2.1751,
"step": 14310
},
{
"epoch": 1.299758907045272,
"grad_norm": 0.0,
"learning_rate": 5.108958065822511e-07,
"loss": 2.1895,
"step": 14320
},
{
"epoch": 1.3004286096972943,
"grad_norm": 0.0,
"learning_rate": 4.943636130698814e-07,
"loss": 2.1437,
"step": 14330
},
{
"epoch": 1.301098312349317,
"grad_norm": 0.0,
"learning_rate": 4.781019928470909e-07,
"loss": 2.1998,
"step": 14340
},
{
"epoch": 1.3017680150013393,
"grad_norm": 0.0,
"learning_rate": 4.6211103479184006e-07,
"loss": 2.1283,
"step": 14350
},
{
"epoch": 1.302437717653362,
"grad_norm": 0.0,
"learning_rate": 4.4639082630278383e-07,
"loss": 2.1758,
"step": 14360
},
{
"epoch": 1.3031074203053845,
"grad_norm": 0.0,
"learning_rate": 4.3094145329879987e-07,
"loss": 2.1329,
"step": 14370
},
{
"epoch": 1.3037771229574069,
"grad_norm": 0.0,
"learning_rate": 4.157630002185053e-07,
"loss": 2.175,
"step": 14380
},
{
"epoch": 1.3044468256094295,
"grad_norm": 0.0,
"learning_rate": 4.00855550019813e-07,
"loss": 2.1512,
"step": 14390
},
{
"epoch": 1.3051165282614519,
"grad_norm": 0.0,
"learning_rate": 3.862191841794538e-07,
"loss": 2.1549,
"step": 14400
},
{
"epoch": 1.3057862309134745,
"grad_norm": 0.0,
"learning_rate": 3.71853982692566e-07,
"loss": 2.1426,
"step": 14410
},
{
"epoch": 1.306455933565497,
"grad_norm": 0.0,
"learning_rate": 3.5776002407221785e-07,
"loss": 2.1325,
"step": 14420
},
{
"epoch": 1.3071256362175194,
"grad_norm": 0.0,
"learning_rate": 3.43937385349008e-07,
"loss": 2.16,
"step": 14430
},
{
"epoch": 1.3077953388695418,
"grad_norm": 0.0,
"learning_rate": 3.3038614207063224e-07,
"loss": 2.0809,
"step": 14440
},
{
"epoch": 1.3084650415215644,
"grad_norm": 0.0,
"learning_rate": 3.171063683014841e-07,
"loss": 2.0655,
"step": 14450
},
{
"epoch": 1.309134744173587,
"grad_norm": 0.0,
"learning_rate": 3.040981366222162e-07,
"loss": 2.1228,
"step": 14460
},
{
"epoch": 1.3098044468256094,
"grad_norm": 0.0,
"learning_rate": 2.913615181293905e-07,
"loss": 2.1572,
"step": 14470
},
{
"epoch": 1.310474149477632,
"grad_norm": 0.0,
"learning_rate": 2.7889658243504536e-07,
"loss": 2.1039,
"step": 14480
},
{
"epoch": 1.3111438521296543,
"grad_norm": 0.0,
"learning_rate": 2.6670339766635686e-07,
"loss": 2.2135,
"step": 14490
},
{
"epoch": 1.311813554781677,
"grad_norm": 0.0,
"learning_rate": 2.547820304652282e-07,
"loss": 2.1868,
"step": 14500
},
{
"epoch": 1.3124832574336995,
"grad_norm": 0.0,
"learning_rate": 2.4313254598795075e-07,
"loss": 2.1494,
"step": 14510
},
{
"epoch": 1.313152960085722,
"grad_norm": 0.0,
"learning_rate": 2.3175500790483252e-07,
"loss": 2.1583,
"step": 14520
},
{
"epoch": 1.3138226627377445,
"grad_norm": 0.0,
"learning_rate": 2.2064947839987027e-07,
"loss": 2.2234,
"step": 14530
},
{
"epoch": 1.3144923653897669,
"grad_norm": 0.0,
"learning_rate": 2.098160181703779e-07,
"loss": 2.0454,
"step": 14540
},
{
"epoch": 1.3151620680417895,
"grad_norm": 0.0,
"learning_rate": 1.9925468642668643e-07,
"loss": 2.1914,
"step": 14550
},
{
"epoch": 1.315831770693812,
"grad_norm": 0.0,
"learning_rate": 1.8896554089181118e-07,
"loss": 2.0893,
"step": 14560
},
{
"epoch": 1.3165014733458344,
"grad_norm": 0.0,
"learning_rate": 1.7894863780111847e-07,
"loss": 2.1605,
"step": 14570
},
{
"epoch": 1.3171711759978568,
"grad_norm": 0.0,
"learning_rate": 1.6920403190203714e-07,
"loss": 2.0833,
"step": 14580
},
{
"epoch": 1.3178408786498794,
"grad_norm": 0.0,
"learning_rate": 1.5973177645376426e-07,
"loss": 2.17,
"step": 14590
},
{
"epoch": 1.318510581301902,
"grad_norm": 0.0,
"learning_rate": 1.5053192322694863e-07,
"loss": 2.1729,
"step": 14600
},
{
"epoch": 1.3191802839539244,
"grad_norm": 0.0,
"learning_rate": 1.4160452250344125e-07,
"loss": 2.1936,
"step": 14610
},
{
"epoch": 1.319849986605947,
"grad_norm": 0.0,
"learning_rate": 1.3294962307598967e-07,
"loss": 2.1658,
"step": 14620
},
{
"epoch": 1.3205196892579694,
"grad_norm": 0.0,
"learning_rate": 1.245672722479829e-07,
"loss": 2.1066,
"step": 14630
},
{
"epoch": 1.321189391909992,
"grad_norm": 0.0,
"learning_rate": 1.1645751583320152e-07,
"loss": 2.1153,
"step": 14640
},
{
"epoch": 1.3218590945620146,
"grad_norm": 0.0,
"learning_rate": 1.0862039815555669e-07,
"loss": 2.1155,
"step": 14650
},
{
"epoch": 1.322528797214037,
"grad_norm": 0.0,
"learning_rate": 1.0105596204885715e-07,
"loss": 2.0831,
"step": 14660
},
{
"epoch": 1.3231984998660595,
"grad_norm": 0.0,
"learning_rate": 9.376424885655932e-08,
"loss": 2.2167,
"step": 14670
},
{
"epoch": 1.323868202518082,
"grad_norm": 0.0,
"learning_rate": 8.674529843155643e-08,
"loss": 2.2026,
"step": 14680
},
{
"epoch": 1.3245379051701045,
"grad_norm": 0.0,
"learning_rate": 7.999914913595641e-08,
"loss": 2.2008,
"step": 14690
},
{
"epoch": 1.325207607822127,
"grad_norm": 0.0,
"learning_rate": 7.352583784087096e-08,
"loss": 2.2153,
"step": 14700
},
{
"epoch": 1.3258773104741495,
"grad_norm": 0.0,
"learning_rate": 6.732539992621023e-08,
"loss": 2.12,
"step": 14710
},
{
"epoch": 1.3265470131261718,
"grad_norm": 0.0,
"learning_rate": 6.139786928050506e-08,
"loss": 2.1566,
"step": 14720
},
{
"epoch": 1.3272167157781944,
"grad_norm": 0.0,
"learning_rate": 5.5743278300690596e-08,
"loss": 2.1574,
"step": 14730
},
{
"epoch": 1.327886418430217,
"grad_norm": 0.0,
"learning_rate": 5.0361657891967454e-08,
"loss": 2.0934,
"step": 14740
},
{
"epoch": 1.3285561210822394,
"grad_norm": 0.0,
"learning_rate": 4.52530374676019e-08,
"loss": 2.2051,
"step": 14750
},
{
"epoch": 1.329225823734262,
"grad_norm": 0.0,
"learning_rate": 4.0417444948792627e-08,
"loss": 2.2382,
"step": 14760
},
{
"epoch": 1.3298955263862844,
"grad_norm": 0.0,
"learning_rate": 3.585490676448755e-08,
"loss": 2.2488,
"step": 14770
},
{
"epoch": 1.330565229038307,
"grad_norm": 0.0,
"learning_rate": 3.15654478512617e-08,
"loss": 2.1103,
"step": 14780
},
{
"epoch": 1.3312349316903296,
"grad_norm": 0.0,
"learning_rate": 2.7549091653167325e-08,
"loss": 2.1718,
"step": 14790
},
{
"epoch": 1.331904634342352,
"grad_norm": 0.0,
"learning_rate": 2.3805860121628442e-08,
"loss": 2.0955,
"step": 14800
},
{
"epoch": 1.3325743369943746,
"grad_norm": 0.0,
"learning_rate": 2.0335773715285388e-08,
"loss": 2.1328,
"step": 14810
},
{
"epoch": 1.333244039646397,
"grad_norm": 0.0,
"learning_rate": 1.713885139991711e-08,
"loss": 2.2143,
"step": 14820
},
{
"epoch": 1.3339137422984195,
"grad_norm": 0.0,
"learning_rate": 1.4215110648319042e-08,
"loss": 2.0206,
"step": 14830
},
{
"epoch": 1.3345834449504421,
"grad_norm": 0.0,
"learning_rate": 1.156456744020873e-08,
"loss": 2.1667,
"step": 14840
},
{
"epoch": 1.3352531476024645,
"grad_norm": 0.0,
"learning_rate": 9.187236262137022e-09,
"loss": 2.1365,
"step": 14850
},
{
"epoch": 1.3359228502544869,
"grad_norm": 0.0,
"learning_rate": 7.083130107421454e-09,
"loss": 2.1685,
"step": 14860
},
{
"epoch": 1.3365925529065095,
"grad_norm": 0.0,
"learning_rate": 5.252260476074078e-09,
"loss": 2.2017,
"step": 14870
},
{
"epoch": 1.337262255558532,
"grad_norm": 0.0,
"learning_rate": 3.6946373747070994e-09,
"loss": 2.1203,
"step": 14880
},
{
"epoch": 1.3379319582105544,
"grad_norm": 0.0,
"learning_rate": 2.410269316521774e-09,
"loss": 2.1974,
"step": 14890
},
{
"epoch": 1.338601660862577,
"grad_norm": 0.0,
"learning_rate": 1.3991633212417921e-09,
"loss": 2.152,
"step": 14900
},
{
"epoch": 1.3392713635145994,
"grad_norm": 0.0,
"learning_rate": 6.613249150688727e-10,
"loss": 2.137,
"step": 14910
},
{
"epoch": 1.339941066166622,
"grad_norm": 0.0,
"learning_rate": 1.9675813066055704e-10,
"loss": 2.1729,
"step": 14920
},
{
"epoch": 1.3406107688186446,
"grad_norm": 0.0,
"learning_rate": 5.465507113555645e-12,
"loss": 2.1984,
"step": 14930
},
{
"epoch": 1.340744709349049,
"step": 14932,
"total_flos": 5.508146882263646e+18,
"train_loss": 2.206626670404634,
"train_runtime": 64853.4008,
"train_samples_per_second": 0.23,
"train_steps_per_second": 0.23
}
],
"logging_steps": 10,
"max_steps": 14932,
"num_input_tokens_seen": 0,
"num_train_epochs": 9223372036854775807,
"save_steps": 1660,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 5.508146882263646e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}