Files
TR_CPT/last-checkpoint/trainer_state.json
ModelHub XC 7ad1c843a7 初始化项目,由ModelHub XC社区提供模型
Model: Ba2han/TR_CPT
Source: Original Platform
2026-10-01 01:31:19 +08:00

7043 lines
155 KiB
JSON

{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.2,
"eval_steps": 2000,
"global_step": 2000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0002,
"grad_norm": 0.24895025789737701,
"learning_rate": 2e-05,
"loss": 2.5380594730377197,
"step": 2
},
{
"epoch": 0.0004,
"grad_norm": 0.29820701479911804,
"learning_rate": 6e-05,
"loss": 2.706937074661255,
"step": 4
},
{
"epoch": 0.0006,
"grad_norm": 0.2516429126262665,
"learning_rate": 0.0001,
"loss": 2.7765252590179443,
"step": 6
},
{
"epoch": 0.0008,
"grad_norm": 0.33202970027923584,
"learning_rate": 0.00014000000000000001,
"loss": 2.544926404953003,
"step": 8
},
{
"epoch": 0.001,
"grad_norm": 0.2849053740501404,
"learning_rate": 0.00017999999999999998,
"loss": 2.5805275440216064,
"step": 10
},
{
"epoch": 0.0012,
"grad_norm": 0.24798542261123657,
"learning_rate": 0.00022,
"loss": 2.5191028118133545,
"step": 12
},
{
"epoch": 0.0014,
"grad_norm": 0.275317519903183,
"learning_rate": 0.00026000000000000003,
"loss": 2.5663347244262695,
"step": 14
},
{
"epoch": 0.0016,
"grad_norm": 0.2572816014289856,
"learning_rate": 0.0003,
"loss": 2.6948204040527344,
"step": 16
},
{
"epoch": 0.0018,
"grad_norm": 0.2944325804710388,
"learning_rate": 0.00034,
"loss": 2.5620336532592773,
"step": 18
},
{
"epoch": 0.002,
"grad_norm": 0.22850075364112854,
"learning_rate": 0.00038,
"loss": 2.6055848598480225,
"step": 20
},
{
"epoch": 0.0022,
"grad_norm": 0.2550802230834961,
"learning_rate": 0.00042,
"loss": 2.592937469482422,
"step": 22
},
{
"epoch": 0.0024,
"grad_norm": 0.18905332684516907,
"learning_rate": 0.00046,
"loss": 2.573544979095459,
"step": 24
},
{
"epoch": 0.0026,
"grad_norm": 0.19383828341960907,
"learning_rate": 0.0005,
"loss": 2.505801200866699,
"step": 26
},
{
"epoch": 0.0028,
"grad_norm": 0.1747068166732788,
"learning_rate": 0.00054,
"loss": 2.6725902557373047,
"step": 28
},
{
"epoch": 0.003,
"grad_norm": 0.21107226610183716,
"learning_rate": 0.00058,
"loss": 2.598604202270508,
"step": 30
},
{
"epoch": 0.0032,
"grad_norm": 0.24685928225517273,
"learning_rate": 0.00062,
"loss": 2.5880119800567627,
"step": 32
},
{
"epoch": 0.0034,
"grad_norm": 0.16816066205501556,
"learning_rate": 0.00066,
"loss": 2.585233211517334,
"step": 34
},
{
"epoch": 0.0036,
"grad_norm": 0.19224759936332703,
"learning_rate": 0.0007,
"loss": 2.6399312019348145,
"step": 36
},
{
"epoch": 0.0038,
"grad_norm": 0.18157117068767548,
"learning_rate": 0.00074,
"loss": 2.650930881500244,
"step": 38
},
{
"epoch": 0.004,
"grad_norm": 0.17778383195400238,
"learning_rate": 0.0007800000000000001,
"loss": 2.652998208999634,
"step": 40
},
{
"epoch": 0.0042,
"grad_norm": 0.18349920213222504,
"learning_rate": 0.00082,
"loss": 2.542379379272461,
"step": 42
},
{
"epoch": 0.0044,
"grad_norm": 0.1516871303319931,
"learning_rate": 0.00086,
"loss": 2.544069528579712,
"step": 44
},
{
"epoch": 0.0046,
"grad_norm": 0.1675216257572174,
"learning_rate": 0.0009000000000000001,
"loss": 2.4357407093048096,
"step": 46
},
{
"epoch": 0.0048,
"grad_norm": 0.15476682782173157,
"learning_rate": 0.00094,
"loss": 2.550136089324951,
"step": 48
},
{
"epoch": 0.005,
"grad_norm": 0.14530816674232483,
"learning_rate": 0.00098,
"loss": 2.5942604541778564,
"step": 50
},
{
"epoch": 0.0052,
"grad_norm": 0.14011730253696442,
"learning_rate": 0.00102,
"loss": 2.646078586578369,
"step": 52
},
{
"epoch": 0.0054,
"grad_norm": 0.13879209756851196,
"learning_rate": 0.0010600000000000002,
"loss": 2.5419538021087646,
"step": 54
},
{
"epoch": 0.0056,
"grad_norm": 0.12999264895915985,
"learning_rate": 0.0011,
"loss": 2.4900166988372803,
"step": 56
},
{
"epoch": 0.0058,
"grad_norm": 0.11466622352600098,
"learning_rate": 0.00114,
"loss": 2.5115253925323486,
"step": 58
},
{
"epoch": 0.006,
"grad_norm": 0.10732487589120865,
"learning_rate": 0.00118,
"loss": 2.525278329849243,
"step": 60
},
{
"epoch": 0.0062,
"grad_norm": 0.11611727625131607,
"learning_rate": 0.00122,
"loss": 2.558769464492798,
"step": 62
},
{
"epoch": 0.0064,
"grad_norm": 0.10347674041986465,
"learning_rate": 0.00126,
"loss": 2.5731728076934814,
"step": 64
},
{
"epoch": 0.0066,
"grad_norm": 0.1129438579082489,
"learning_rate": 0.0013000000000000002,
"loss": 2.6827847957611084,
"step": 66
},
{
"epoch": 0.0068,
"grad_norm": 0.10339458286762238,
"learning_rate": 0.00134,
"loss": 2.638899326324463,
"step": 68
},
{
"epoch": 0.007,
"grad_norm": 0.13137196004390717,
"learning_rate": 0.00138,
"loss": 2.5533316135406494,
"step": 70
},
{
"epoch": 0.0072,
"grad_norm": 0.1640232503414154,
"learning_rate": 0.00142,
"loss": 2.4588212966918945,
"step": 72
},
{
"epoch": 0.0074,
"grad_norm": 0.13227109611034393,
"learning_rate": 0.00146,
"loss": 2.529284715652466,
"step": 74
},
{
"epoch": 0.0076,
"grad_norm": 0.09697236865758896,
"learning_rate": 0.0015,
"loss": 2.496910572052002,
"step": 76
},
{
"epoch": 0.0078,
"grad_norm": 0.09535796195268631,
"learning_rate": 0.0015400000000000001,
"loss": 2.3877482414245605,
"step": 78
},
{
"epoch": 0.008,
"grad_norm": 0.09408529102802277,
"learning_rate": 0.00158,
"loss": 2.390780448913574,
"step": 80
},
{
"epoch": 0.0082,
"grad_norm": 0.10365808755159378,
"learning_rate": 0.0016200000000000001,
"loss": 2.7170658111572266,
"step": 82
},
{
"epoch": 0.0084,
"grad_norm": 0.09149951487779617,
"learning_rate": 0.00166,
"loss": 2.367365598678589,
"step": 84
},
{
"epoch": 0.0086,
"grad_norm": 0.09350724518299103,
"learning_rate": 0.0017,
"loss": 2.3851242065429688,
"step": 86
},
{
"epoch": 0.0088,
"grad_norm": 0.1038142666220665,
"learning_rate": 0.00174,
"loss": 2.5150654315948486,
"step": 88
},
{
"epoch": 0.009,
"grad_norm": 0.0919538363814354,
"learning_rate": 0.0017800000000000001,
"loss": 2.6233551502227783,
"step": 90
},
{
"epoch": 0.0092,
"grad_norm": 0.10591281205415726,
"learning_rate": 0.00182,
"loss": 2.478425979614258,
"step": 92
},
{
"epoch": 0.0094,
"grad_norm": 0.08705190569162369,
"learning_rate": 0.00186,
"loss": 2.470477819442749,
"step": 94
},
{
"epoch": 0.0096,
"grad_norm": 0.09547323733568192,
"learning_rate": 0.0019,
"loss": 2.4050943851470947,
"step": 96
},
{
"epoch": 0.0098,
"grad_norm": 0.08529368042945862,
"learning_rate": 0.0019399999999999999,
"loss": 2.475706100463867,
"step": 98
},
{
"epoch": 0.01,
"grad_norm": 0.08629848808050156,
"learning_rate": 0.00198,
"loss": 2.482457160949707,
"step": 100
},
{
"epoch": 0.0102,
"grad_norm": 0.09109193831682205,
"learning_rate": 0.002,
"loss": 2.3759820461273193,
"step": 102
},
{
"epoch": 0.0104,
"grad_norm": 0.08126833289861679,
"learning_rate": 0.002,
"loss": 2.3447344303131104,
"step": 104
},
{
"epoch": 0.0106,
"grad_norm": 0.08956345915794373,
"learning_rate": 0.002,
"loss": 2.545990467071533,
"step": 106
},
{
"epoch": 0.0108,
"grad_norm": 0.09673260897397995,
"learning_rate": 0.002,
"loss": 2.337914228439331,
"step": 108
},
{
"epoch": 0.011,
"grad_norm": 0.08681271225214005,
"learning_rate": 0.002,
"loss": 2.372396469116211,
"step": 110
},
{
"epoch": 0.0112,
"grad_norm": 0.0875004306435585,
"learning_rate": 0.002,
"loss": 2.381185531616211,
"step": 112
},
{
"epoch": 0.0114,
"grad_norm": 0.08569031953811646,
"learning_rate": 0.002,
"loss": 2.364795446395874,
"step": 114
},
{
"epoch": 0.0116,
"grad_norm": 0.0811447948217392,
"learning_rate": 0.002,
"loss": 2.320478916168213,
"step": 116
},
{
"epoch": 0.0118,
"grad_norm": 0.08084925264120102,
"learning_rate": 0.002,
"loss": 2.3602778911590576,
"step": 118
},
{
"epoch": 0.012,
"grad_norm": 0.08097168058156967,
"learning_rate": 0.002,
"loss": 2.344461679458618,
"step": 120
},
{
"epoch": 0.0122,
"grad_norm": 0.08191122114658356,
"learning_rate": 0.002,
"loss": 2.4383997917175293,
"step": 122
},
{
"epoch": 0.0124,
"grad_norm": 0.09045127779245377,
"learning_rate": 0.002,
"loss": 2.341723680496216,
"step": 124
},
{
"epoch": 0.0126,
"grad_norm": 0.08509522676467896,
"learning_rate": 0.002,
"loss": 2.4552154541015625,
"step": 126
},
{
"epoch": 0.0128,
"grad_norm": 0.07946667075157166,
"learning_rate": 0.002,
"loss": 2.2917001247406006,
"step": 128
},
{
"epoch": 0.013,
"grad_norm": 0.09372023493051529,
"learning_rate": 0.002,
"loss": 2.409574031829834,
"step": 130
},
{
"epoch": 0.0132,
"grad_norm": 0.08136238157749176,
"learning_rate": 0.002,
"loss": 2.4375526905059814,
"step": 132
},
{
"epoch": 0.0134,
"grad_norm": 0.08563613891601562,
"learning_rate": 0.002,
"loss": 2.3880152702331543,
"step": 134
},
{
"epoch": 0.0136,
"grad_norm": 0.08194541931152344,
"learning_rate": 0.002,
"loss": 2.3403987884521484,
"step": 136
},
{
"epoch": 0.0138,
"grad_norm": 0.08111666142940521,
"learning_rate": 0.002,
"loss": 2.2349488735198975,
"step": 138
},
{
"epoch": 0.014,
"grad_norm": 0.08328550308942795,
"learning_rate": 0.002,
"loss": 2.512582540512085,
"step": 140
},
{
"epoch": 0.0142,
"grad_norm": 0.08633063733577728,
"learning_rate": 0.002,
"loss": 2.3220415115356445,
"step": 142
},
{
"epoch": 0.0144,
"grad_norm": 0.0758780762553215,
"learning_rate": 0.002,
"loss": 2.4241178035736084,
"step": 144
},
{
"epoch": 0.0146,
"grad_norm": 0.08827359974384308,
"learning_rate": 0.002,
"loss": 2.3007774353027344,
"step": 146
},
{
"epoch": 0.0148,
"grad_norm": 0.08051057904958725,
"learning_rate": 0.002,
"loss": 2.3448150157928467,
"step": 148
},
{
"epoch": 0.015,
"grad_norm": 0.09375292807817459,
"learning_rate": 0.002,
"loss": 2.327587366104126,
"step": 150
},
{
"epoch": 0.0152,
"grad_norm": 0.0823783427476883,
"learning_rate": 0.002,
"loss": 2.3877604007720947,
"step": 152
},
{
"epoch": 0.0154,
"grad_norm": 0.08136262744665146,
"learning_rate": 0.002,
"loss": 2.3504531383514404,
"step": 154
},
{
"epoch": 0.0156,
"grad_norm": 0.0795651227235794,
"learning_rate": 0.002,
"loss": 2.440415620803833,
"step": 156
},
{
"epoch": 0.0158,
"grad_norm": 0.08610543608665466,
"learning_rate": 0.002,
"loss": 2.3680200576782227,
"step": 158
},
{
"epoch": 0.016,
"grad_norm": 0.08510268479585648,
"learning_rate": 0.002,
"loss": 2.3274452686309814,
"step": 160
},
{
"epoch": 0.0162,
"grad_norm": 0.08430306613445282,
"learning_rate": 0.002,
"loss": 2.2722561359405518,
"step": 162
},
{
"epoch": 0.0164,
"grad_norm": 0.08521681278944016,
"learning_rate": 0.002,
"loss": 2.2904763221740723,
"step": 164
},
{
"epoch": 0.0166,
"grad_norm": 0.08444844186306,
"learning_rate": 0.002,
"loss": 2.237799644470215,
"step": 166
},
{
"epoch": 0.0168,
"grad_norm": 0.08309835940599442,
"learning_rate": 0.002,
"loss": 2.476332187652588,
"step": 168
},
{
"epoch": 0.017,
"grad_norm": 0.09012819081544876,
"learning_rate": 0.002,
"loss": 2.506131410598755,
"step": 170
},
{
"epoch": 0.0172,
"grad_norm": 0.08156735450029373,
"learning_rate": 0.002,
"loss": 2.3241539001464844,
"step": 172
},
{
"epoch": 0.0174,
"grad_norm": 0.08085346221923828,
"learning_rate": 0.002,
"loss": 2.324756622314453,
"step": 174
},
{
"epoch": 0.0176,
"grad_norm": 0.08306079357862473,
"learning_rate": 0.002,
"loss": 2.2054829597473145,
"step": 176
},
{
"epoch": 0.0178,
"grad_norm": 0.08217490464448929,
"learning_rate": 0.002,
"loss": 2.1210339069366455,
"step": 178
},
{
"epoch": 0.018,
"grad_norm": 0.08472347259521484,
"learning_rate": 0.002,
"loss": 2.334371566772461,
"step": 180
},
{
"epoch": 0.0182,
"grad_norm": 0.0811653882265091,
"learning_rate": 0.002,
"loss": 2.325976848602295,
"step": 182
},
{
"epoch": 0.0184,
"grad_norm": 0.08189306408166885,
"learning_rate": 0.002,
"loss": 2.4059085845947266,
"step": 184
},
{
"epoch": 0.0186,
"grad_norm": 0.09424543380737305,
"learning_rate": 0.002,
"loss": 2.336735725402832,
"step": 186
},
{
"epoch": 0.0188,
"grad_norm": 0.08703279495239258,
"learning_rate": 0.002,
"loss": 2.1647582054138184,
"step": 188
},
{
"epoch": 0.019,
"grad_norm": 0.08056046068668365,
"learning_rate": 0.002,
"loss": 2.3341431617736816,
"step": 190
},
{
"epoch": 0.0192,
"grad_norm": 0.09091464430093765,
"learning_rate": 0.002,
"loss": 2.337998628616333,
"step": 192
},
{
"epoch": 0.0194,
"grad_norm": 0.08341612666845322,
"learning_rate": 0.002,
"loss": 2.191992998123169,
"step": 194
},
{
"epoch": 0.0196,
"grad_norm": 0.0801461935043335,
"learning_rate": 0.002,
"loss": 2.4013006687164307,
"step": 196
},
{
"epoch": 0.0198,
"grad_norm": 0.07991934567689896,
"learning_rate": 0.002,
"loss": 2.2777493000030518,
"step": 198
},
{
"epoch": 0.02,
"grad_norm": 0.08411742001771927,
"learning_rate": 0.002,
"loss": 2.195572853088379,
"step": 200
},
{
"epoch": 0.0202,
"grad_norm": 0.08740992844104767,
"learning_rate": 0.002,
"loss": 2.249514102935791,
"step": 202
},
{
"epoch": 0.0204,
"grad_norm": 0.08233100920915604,
"learning_rate": 0.002,
"loss": 2.3361704349517822,
"step": 204
},
{
"epoch": 0.0206,
"grad_norm": 0.07927953451871872,
"learning_rate": 0.002,
"loss": 2.368013620376587,
"step": 206
},
{
"epoch": 0.0208,
"grad_norm": 0.08175376057624817,
"learning_rate": 0.002,
"loss": 2.288491725921631,
"step": 208
},
{
"epoch": 0.021,
"grad_norm": 0.0927237719297409,
"learning_rate": 0.002,
"loss": 2.353435516357422,
"step": 210
},
{
"epoch": 0.0212,
"grad_norm": 0.08352026343345642,
"learning_rate": 0.002,
"loss": 2.2495923042297363,
"step": 212
},
{
"epoch": 0.0214,
"grad_norm": 0.08100274205207825,
"learning_rate": 0.002,
"loss": 2.203218936920166,
"step": 214
},
{
"epoch": 0.0216,
"grad_norm": 0.08187449723482132,
"learning_rate": 0.002,
"loss": 2.297393321990967,
"step": 216
},
{
"epoch": 0.0218,
"grad_norm": 0.08764786273241043,
"learning_rate": 0.002,
"loss": 2.3698809146881104,
"step": 218
},
{
"epoch": 0.022,
"grad_norm": 0.08565348386764526,
"learning_rate": 0.002,
"loss": 2.3626506328582764,
"step": 220
},
{
"epoch": 0.0222,
"grad_norm": 0.08267498016357422,
"learning_rate": 0.002,
"loss": 2.281557559967041,
"step": 222
},
{
"epoch": 0.0224,
"grad_norm": 0.08042807877063751,
"learning_rate": 0.002,
"loss": 2.212425708770752,
"step": 224
},
{
"epoch": 0.0226,
"grad_norm": 0.0761464312672615,
"learning_rate": 0.002,
"loss": 2.3434534072875977,
"step": 226
},
{
"epoch": 0.0228,
"grad_norm": 0.08358408510684967,
"learning_rate": 0.002,
"loss": 2.379551887512207,
"step": 228
},
{
"epoch": 0.023,
"grad_norm": 0.08283611387014389,
"learning_rate": 0.002,
"loss": 2.3552372455596924,
"step": 230
},
{
"epoch": 0.0232,
"grad_norm": 0.08272962272167206,
"learning_rate": 0.002,
"loss": 2.2771637439727783,
"step": 232
},
{
"epoch": 0.0234,
"grad_norm": 0.08492973446846008,
"learning_rate": 0.002,
"loss": 2.336639642715454,
"step": 234
},
{
"epoch": 0.0236,
"grad_norm": 0.0840471163392067,
"learning_rate": 0.002,
"loss": 2.2582857608795166,
"step": 236
},
{
"epoch": 0.0238,
"grad_norm": 0.08327464759349823,
"learning_rate": 0.002,
"loss": 2.266247510910034,
"step": 238
},
{
"epoch": 0.024,
"grad_norm": 0.08937662094831467,
"learning_rate": 0.002,
"loss": 2.157350778579712,
"step": 240
},
{
"epoch": 0.0242,
"grad_norm": 0.07938210666179657,
"learning_rate": 0.002,
"loss": 2.3311309814453125,
"step": 242
},
{
"epoch": 0.0244,
"grad_norm": 0.08206083625555038,
"learning_rate": 0.002,
"loss": 2.2125890254974365,
"step": 244
},
{
"epoch": 0.0246,
"grad_norm": 0.08916025608778,
"learning_rate": 0.002,
"loss": 2.3561928272247314,
"step": 246
},
{
"epoch": 0.0248,
"grad_norm": 0.08466347306966782,
"learning_rate": 0.002,
"loss": 2.293194532394409,
"step": 248
},
{
"epoch": 0.025,
"grad_norm": 0.0937456488609314,
"learning_rate": 0.002,
"loss": 2.2553927898406982,
"step": 250
},
{
"epoch": 0.0252,
"grad_norm": 0.11129719018936157,
"learning_rate": 0.002,
"loss": 2.3435614109039307,
"step": 252
},
{
"epoch": 0.0254,
"grad_norm": 0.07488962262868881,
"learning_rate": 0.002,
"loss": 2.063786506652832,
"step": 254
},
{
"epoch": 0.0256,
"grad_norm": 0.08457907289266586,
"learning_rate": 0.002,
"loss": 2.2499372959136963,
"step": 256
},
{
"epoch": 0.0258,
"grad_norm": 0.09276195615530014,
"learning_rate": 0.002,
"loss": 2.166271686553955,
"step": 258
},
{
"epoch": 0.026,
"grad_norm": 0.08557724952697754,
"learning_rate": 0.002,
"loss": 2.2800700664520264,
"step": 260
},
{
"epoch": 0.0262,
"grad_norm": 0.09652607142925262,
"learning_rate": 0.002,
"loss": 2.328278064727783,
"step": 262
},
{
"epoch": 0.0264,
"grad_norm": 0.09041029959917068,
"learning_rate": 0.002,
"loss": 2.34018611907959,
"step": 264
},
{
"epoch": 0.0266,
"grad_norm": 0.07893089950084686,
"learning_rate": 0.002,
"loss": 2.2603254318237305,
"step": 266
},
{
"epoch": 0.0268,
"grad_norm": 0.08397849649190903,
"learning_rate": 0.002,
"loss": 2.2433888912200928,
"step": 268
},
{
"epoch": 0.027,
"grad_norm": 0.0883750393986702,
"learning_rate": 0.002,
"loss": 2.3365883827209473,
"step": 270
},
{
"epoch": 0.0272,
"grad_norm": 0.08821607381105423,
"learning_rate": 0.002,
"loss": 2.178830862045288,
"step": 272
},
{
"epoch": 0.0274,
"grad_norm": 0.08053535223007202,
"learning_rate": 0.002,
"loss": 2.1993696689605713,
"step": 274
},
{
"epoch": 0.0276,
"grad_norm": 0.08129502832889557,
"learning_rate": 0.002,
"loss": 2.179894208908081,
"step": 276
},
{
"epoch": 0.0278,
"grad_norm": 0.08053018152713776,
"learning_rate": 0.002,
"loss": 2.2494606971740723,
"step": 278
},
{
"epoch": 0.028,
"grad_norm": 0.08439638465642929,
"learning_rate": 0.002,
"loss": 2.1246116161346436,
"step": 280
},
{
"epoch": 0.0282,
"grad_norm": 0.07741652429103851,
"learning_rate": 0.002,
"loss": 2.2139837741851807,
"step": 282
},
{
"epoch": 0.0284,
"grad_norm": 0.08243114501237869,
"learning_rate": 0.002,
"loss": 2.2016916275024414,
"step": 284
},
{
"epoch": 0.0286,
"grad_norm": 0.0877411887049675,
"learning_rate": 0.002,
"loss": 2.2535932064056396,
"step": 286
},
{
"epoch": 0.0288,
"grad_norm": 0.07781921327114105,
"learning_rate": 0.002,
"loss": 2.226775646209717,
"step": 288
},
{
"epoch": 0.029,
"grad_norm": 0.08718002587556839,
"learning_rate": 0.002,
"loss": 2.1132407188415527,
"step": 290
},
{
"epoch": 0.0292,
"grad_norm": 0.08359642326831818,
"learning_rate": 0.002,
"loss": 2.1846566200256348,
"step": 292
},
{
"epoch": 0.0294,
"grad_norm": 0.08507300168275833,
"learning_rate": 0.002,
"loss": 2.2220089435577393,
"step": 294
},
{
"epoch": 0.0296,
"grad_norm": 0.08341013640165329,
"learning_rate": 0.002,
"loss": 2.1871256828308105,
"step": 296
},
{
"epoch": 0.0298,
"grad_norm": 0.08784069865942001,
"learning_rate": 0.002,
"loss": 2.2707948684692383,
"step": 298
},
{
"epoch": 0.03,
"grad_norm": 0.09680568426847458,
"learning_rate": 0.002,
"loss": 2.1852164268493652,
"step": 300
},
{
"epoch": 0.0302,
"grad_norm": 0.08098499476909637,
"learning_rate": 0.002,
"loss": 2.0984084606170654,
"step": 302
},
{
"epoch": 0.0304,
"grad_norm": 0.08306451886892319,
"learning_rate": 0.002,
"loss": 2.0187599658966064,
"step": 304
},
{
"epoch": 0.0306,
"grad_norm": 0.08087552338838577,
"learning_rate": 0.002,
"loss": 2.104374647140503,
"step": 306
},
{
"epoch": 0.0308,
"grad_norm": 0.08561889827251434,
"learning_rate": 0.002,
"loss": 2.2318408489227295,
"step": 308
},
{
"epoch": 0.031,
"grad_norm": 0.09089436382055283,
"learning_rate": 0.002,
"loss": 2.252009868621826,
"step": 310
},
{
"epoch": 0.0312,
"grad_norm": 0.08175069093704224,
"learning_rate": 0.002,
"loss": 2.127460241317749,
"step": 312
},
{
"epoch": 0.0314,
"grad_norm": 0.07938393950462341,
"learning_rate": 0.002,
"loss": 2.308382749557495,
"step": 314
},
{
"epoch": 0.0316,
"grad_norm": 0.07617771625518799,
"learning_rate": 0.002,
"loss": 2.193786144256592,
"step": 316
},
{
"epoch": 0.0318,
"grad_norm": 0.08023491501808167,
"learning_rate": 0.002,
"loss": 2.289525032043457,
"step": 318
},
{
"epoch": 0.032,
"grad_norm": 0.08281378448009491,
"learning_rate": 0.002,
"loss": 2.280076026916504,
"step": 320
},
{
"epoch": 0.0322,
"grad_norm": 0.08946441113948822,
"learning_rate": 0.002,
"loss": 2.185492515563965,
"step": 322
},
{
"epoch": 0.0324,
"grad_norm": 0.08579517900943756,
"learning_rate": 0.002,
"loss": 2.254850387573242,
"step": 324
},
{
"epoch": 0.0326,
"grad_norm": 0.09209157526493073,
"learning_rate": 0.002,
"loss": 2.1961588859558105,
"step": 326
},
{
"epoch": 0.0328,
"grad_norm": 0.08560938388109207,
"learning_rate": 0.002,
"loss": 2.1954500675201416,
"step": 328
},
{
"epoch": 0.033,
"grad_norm": 0.08744245767593384,
"learning_rate": 0.002,
"loss": 2.3917150497436523,
"step": 330
},
{
"epoch": 0.0332,
"grad_norm": 0.08644507080316544,
"learning_rate": 0.002,
"loss": 2.18365478515625,
"step": 332
},
{
"epoch": 0.0334,
"grad_norm": 0.08469928056001663,
"learning_rate": 0.002,
"loss": 2.282428741455078,
"step": 334
},
{
"epoch": 0.0336,
"grad_norm": 0.0834391713142395,
"learning_rate": 0.002,
"loss": 2.189535617828369,
"step": 336
},
{
"epoch": 0.0338,
"grad_norm": 0.08007276803255081,
"learning_rate": 0.002,
"loss": 2.2560901641845703,
"step": 338
},
{
"epoch": 0.034,
"grad_norm": 0.08669793605804443,
"learning_rate": 0.002,
"loss": 2.3364031314849854,
"step": 340
},
{
"epoch": 0.0342,
"grad_norm": 0.0828932449221611,
"learning_rate": 0.002,
"loss": 2.3209688663482666,
"step": 342
},
{
"epoch": 0.0344,
"grad_norm": 0.08443323522806168,
"learning_rate": 0.002,
"loss": 2.141862630844116,
"step": 344
},
{
"epoch": 0.0346,
"grad_norm": 0.08589595556259155,
"learning_rate": 0.002,
"loss": 2.265019655227661,
"step": 346
},
{
"epoch": 0.0348,
"grad_norm": 0.08147434145212173,
"learning_rate": 0.002,
"loss": 2.0417823791503906,
"step": 348
},
{
"epoch": 0.035,
"grad_norm": 0.07960567623376846,
"learning_rate": 0.002,
"loss": 2.2330386638641357,
"step": 350
},
{
"epoch": 0.0352,
"grad_norm": 0.08081433922052383,
"learning_rate": 0.002,
"loss": 2.261106014251709,
"step": 352
},
{
"epoch": 0.0354,
"grad_norm": 0.08448409289121628,
"learning_rate": 0.002,
"loss": 2.274304151535034,
"step": 354
},
{
"epoch": 0.0356,
"grad_norm": 0.07975075393915176,
"learning_rate": 0.002,
"loss": 2.154627561569214,
"step": 356
},
{
"epoch": 0.0358,
"grad_norm": 0.08375726640224457,
"learning_rate": 0.002,
"loss": 2.1503400802612305,
"step": 358
},
{
"epoch": 0.036,
"grad_norm": 0.08201711624860764,
"learning_rate": 0.002,
"loss": 2.231780767440796,
"step": 360
},
{
"epoch": 0.0362,
"grad_norm": 0.07838217169046402,
"learning_rate": 0.002,
"loss": 2.237990617752075,
"step": 362
},
{
"epoch": 0.0364,
"grad_norm": 0.08470814675092697,
"learning_rate": 0.002,
"loss": 2.2229793071746826,
"step": 364
},
{
"epoch": 0.0366,
"grad_norm": 0.0849805697798729,
"learning_rate": 0.002,
"loss": 2.193814277648926,
"step": 366
},
{
"epoch": 0.0368,
"grad_norm": 0.07920201122760773,
"learning_rate": 0.002,
"loss": 2.078490734100342,
"step": 368
},
{
"epoch": 0.037,
"grad_norm": 0.08708656579256058,
"learning_rate": 0.002,
"loss": 2.2713921070098877,
"step": 370
},
{
"epoch": 0.0372,
"grad_norm": 0.08646319061517715,
"learning_rate": 0.002,
"loss": 2.1741607189178467,
"step": 372
},
{
"epoch": 0.0374,
"grad_norm": 0.07915737479925156,
"learning_rate": 0.002,
"loss": 2.13606858253479,
"step": 374
},
{
"epoch": 0.0376,
"grad_norm": 0.08452604711055756,
"learning_rate": 0.002,
"loss": 2.2122349739074707,
"step": 376
},
{
"epoch": 0.0378,
"grad_norm": 0.08610899746417999,
"learning_rate": 0.002,
"loss": 2.2641565799713135,
"step": 378
},
{
"epoch": 0.038,
"grad_norm": 0.08267128467559814,
"learning_rate": 0.002,
"loss": 2.098310947418213,
"step": 380
},
{
"epoch": 0.0382,
"grad_norm": 0.08457912504673004,
"learning_rate": 0.002,
"loss": 2.2084639072418213,
"step": 382
},
{
"epoch": 0.0384,
"grad_norm": 0.08544643223285675,
"learning_rate": 0.002,
"loss": 2.210813283920288,
"step": 384
},
{
"epoch": 0.0386,
"grad_norm": 0.08054296672344208,
"learning_rate": 0.002,
"loss": 2.0772266387939453,
"step": 386
},
{
"epoch": 0.0388,
"grad_norm": 0.08014437556266785,
"learning_rate": 0.002,
"loss": 2.076784610748291,
"step": 388
},
{
"epoch": 0.039,
"grad_norm": 0.08763397485017776,
"learning_rate": 0.002,
"loss": 2.201284170150757,
"step": 390
},
{
"epoch": 0.0392,
"grad_norm": 0.08363637328147888,
"learning_rate": 0.002,
"loss": 2.280714511871338,
"step": 392
},
{
"epoch": 0.0394,
"grad_norm": 0.08890454471111298,
"learning_rate": 0.002,
"loss": 2.2450146675109863,
"step": 394
},
{
"epoch": 0.0396,
"grad_norm": 0.07954143732786179,
"learning_rate": 0.002,
"loss": 2.146226406097412,
"step": 396
},
{
"epoch": 0.0398,
"grad_norm": 0.09244970977306366,
"learning_rate": 0.002,
"loss": 2.164762496948242,
"step": 398
},
{
"epoch": 0.04,
"grad_norm": 0.0843396931886673,
"learning_rate": 0.002,
"loss": 2.1567463874816895,
"step": 400
},
{
"epoch": 0.0402,
"grad_norm": 0.0928284302353859,
"learning_rate": 0.002,
"loss": 2.289367198944092,
"step": 402
},
{
"epoch": 0.0404,
"grad_norm": 0.08285237848758698,
"learning_rate": 0.002,
"loss": 2.124049425125122,
"step": 404
},
{
"epoch": 0.0406,
"grad_norm": 0.09418705850839615,
"learning_rate": 0.002,
"loss": 2.2794275283813477,
"step": 406
},
{
"epoch": 0.0408,
"grad_norm": 0.07908788323402405,
"learning_rate": 0.002,
"loss": 2.2018861770629883,
"step": 408
},
{
"epoch": 0.041,
"grad_norm": 0.09090521931648254,
"learning_rate": 0.002,
"loss": 2.1399567127227783,
"step": 410
},
{
"epoch": 0.0412,
"grad_norm": 0.07676365226507187,
"learning_rate": 0.002,
"loss": 2.143925189971924,
"step": 412
},
{
"epoch": 0.0414,
"grad_norm": 0.08761579543352127,
"learning_rate": 0.002,
"loss": 2.1854593753814697,
"step": 414
},
{
"epoch": 0.0416,
"grad_norm": 0.08117334544658661,
"learning_rate": 0.002,
"loss": 2.2653510570526123,
"step": 416
},
{
"epoch": 0.0418,
"grad_norm": 0.09029056876897812,
"learning_rate": 0.002,
"loss": 2.1869001388549805,
"step": 418
},
{
"epoch": 0.042,
"grad_norm": 0.08253905177116394,
"learning_rate": 0.002,
"loss": 2.2279319763183594,
"step": 420
},
{
"epoch": 0.0422,
"grad_norm": 0.07928216457366943,
"learning_rate": 0.002,
"loss": 2.1313045024871826,
"step": 422
},
{
"epoch": 0.0424,
"grad_norm": 0.08245409280061722,
"learning_rate": 0.002,
"loss": 2.193098545074463,
"step": 424
},
{
"epoch": 0.0426,
"grad_norm": 0.08652161061763763,
"learning_rate": 0.002,
"loss": 2.2768752574920654,
"step": 426
},
{
"epoch": 0.0428,
"grad_norm": 0.08524289727210999,
"learning_rate": 0.002,
"loss": 2.2626569271087646,
"step": 428
},
{
"epoch": 0.043,
"grad_norm": 0.08207456767559052,
"learning_rate": 0.002,
"loss": 2.1923787593841553,
"step": 430
},
{
"epoch": 0.0432,
"grad_norm": 0.07999079674482346,
"learning_rate": 0.002,
"loss": 1.9743129014968872,
"step": 432
},
{
"epoch": 0.0434,
"grad_norm": 0.09092952311038971,
"learning_rate": 0.002,
"loss": 2.262089490890503,
"step": 434
},
{
"epoch": 0.0436,
"grad_norm": 0.09337197989225388,
"learning_rate": 0.002,
"loss": 2.1411046981811523,
"step": 436
},
{
"epoch": 0.0438,
"grad_norm": 0.08668949455022812,
"learning_rate": 0.002,
"loss": 2.2864744663238525,
"step": 438
},
{
"epoch": 0.044,
"grad_norm": 0.08115921169519424,
"learning_rate": 0.002,
"loss": 2.1516242027282715,
"step": 440
},
{
"epoch": 0.0442,
"grad_norm": 0.0895581766963005,
"learning_rate": 0.002,
"loss": 2.2562949657440186,
"step": 442
},
{
"epoch": 0.0444,
"grad_norm": 0.07889360934495926,
"learning_rate": 0.002,
"loss": 2.240028142929077,
"step": 444
},
{
"epoch": 0.0446,
"grad_norm": 0.07827291637659073,
"learning_rate": 0.002,
"loss": 2.160885810852051,
"step": 446
},
{
"epoch": 0.0448,
"grad_norm": 0.08347690850496292,
"learning_rate": 0.002,
"loss": 2.2001569271087646,
"step": 448
},
{
"epoch": 0.045,
"grad_norm": 0.08106813579797745,
"learning_rate": 0.002,
"loss": 2.2011420726776123,
"step": 450
},
{
"epoch": 0.0452,
"grad_norm": 0.08012887835502625,
"learning_rate": 0.002,
"loss": 2.279381036758423,
"step": 452
},
{
"epoch": 0.0454,
"grad_norm": 0.08830226957798004,
"learning_rate": 0.002,
"loss": 2.1910595893859863,
"step": 454
},
{
"epoch": 0.0456,
"grad_norm": 0.0870884582400322,
"learning_rate": 0.002,
"loss": 2.18572735786438,
"step": 456
},
{
"epoch": 0.0458,
"grad_norm": 0.07899375259876251,
"learning_rate": 0.002,
"loss": 2.0611331462860107,
"step": 458
},
{
"epoch": 0.046,
"grad_norm": 0.0819762796163559,
"learning_rate": 0.002,
"loss": 2.172961950302124,
"step": 460
},
{
"epoch": 0.0462,
"grad_norm": 0.08505290746688843,
"learning_rate": 0.002,
"loss": 2.1397311687469482,
"step": 462
},
{
"epoch": 0.0464,
"grad_norm": 0.09050805121660233,
"learning_rate": 0.002,
"loss": 2.267570734024048,
"step": 464
},
{
"epoch": 0.0466,
"grad_norm": 0.08763068914413452,
"learning_rate": 0.002,
"loss": 2.2150211334228516,
"step": 466
},
{
"epoch": 0.0468,
"grad_norm": 0.08520094305276871,
"learning_rate": 0.002,
"loss": 2.2994608879089355,
"step": 468
},
{
"epoch": 0.047,
"grad_norm": 0.0786767303943634,
"learning_rate": 0.002,
"loss": 2.0375938415527344,
"step": 470
},
{
"epoch": 0.0472,
"grad_norm": 0.0806039571762085,
"learning_rate": 0.002,
"loss": 2.1081976890563965,
"step": 472
},
{
"epoch": 0.0474,
"grad_norm": 0.08359119296073914,
"learning_rate": 0.002,
"loss": 2.2876009941101074,
"step": 474
},
{
"epoch": 0.0476,
"grad_norm": 0.07881803810596466,
"learning_rate": 0.002,
"loss": 2.126938819885254,
"step": 476
},
{
"epoch": 0.0478,
"grad_norm": 0.0793905034661293,
"learning_rate": 0.002,
"loss": 2.1410984992980957,
"step": 478
},
{
"epoch": 0.048,
"grad_norm": 0.08302482217550278,
"learning_rate": 0.002,
"loss": 2.137707471847534,
"step": 480
},
{
"epoch": 0.0482,
"grad_norm": 0.08349647372961044,
"learning_rate": 0.002,
"loss": 2.2077903747558594,
"step": 482
},
{
"epoch": 0.0484,
"grad_norm": 0.08352367579936981,
"learning_rate": 0.002,
"loss": 2.2567689418792725,
"step": 484
},
{
"epoch": 0.0486,
"grad_norm": 0.08130789548158646,
"learning_rate": 0.002,
"loss": 2.140805721282959,
"step": 486
},
{
"epoch": 0.0488,
"grad_norm": 0.0914766862988472,
"learning_rate": 0.002,
"loss": 2.268686294555664,
"step": 488
},
{
"epoch": 0.049,
"grad_norm": 0.07911514490842819,
"learning_rate": 0.002,
"loss": 2.0743894577026367,
"step": 490
},
{
"epoch": 0.0492,
"grad_norm": 0.08769986778497696,
"learning_rate": 0.002,
"loss": 2.2632336616516113,
"step": 492
},
{
"epoch": 0.0494,
"grad_norm": 0.08505533635616302,
"learning_rate": 0.002,
"loss": 2.1295549869537354,
"step": 494
},
{
"epoch": 0.0496,
"grad_norm": 0.08464609086513519,
"learning_rate": 0.002,
"loss": 2.1045517921447754,
"step": 496
},
{
"epoch": 0.0498,
"grad_norm": 0.08616667240858078,
"learning_rate": 0.002,
"loss": 2.1572630405426025,
"step": 498
},
{
"epoch": 0.05,
"grad_norm": 0.0773443877696991,
"learning_rate": 0.002,
"loss": 2.19515323638916,
"step": 500
},
{
"epoch": 0.0502,
"grad_norm": 0.09228549897670746,
"learning_rate": 0.002,
"loss": 2.3968870639801025,
"step": 502
},
{
"epoch": 0.0504,
"grad_norm": 0.091270312666893,
"learning_rate": 0.002,
"loss": 2.030975580215454,
"step": 504
},
{
"epoch": 0.0506,
"grad_norm": 0.08151127398014069,
"learning_rate": 0.002,
"loss": 2.1822268962860107,
"step": 506
},
{
"epoch": 0.0508,
"grad_norm": 0.08774775266647339,
"learning_rate": 0.002,
"loss": 2.1789495944976807,
"step": 508
},
{
"epoch": 0.051,
"grad_norm": 0.0810612142086029,
"learning_rate": 0.002,
"loss": 2.0437979698181152,
"step": 510
},
{
"epoch": 0.0512,
"grad_norm": 0.07760541141033173,
"learning_rate": 0.002,
"loss": 2.170164108276367,
"step": 512
},
{
"epoch": 0.0514,
"grad_norm": 0.08637677133083344,
"learning_rate": 0.002,
"loss": 2.2719333171844482,
"step": 514
},
{
"epoch": 0.0516,
"grad_norm": 0.07816094160079956,
"learning_rate": 0.002,
"loss": 2.0255773067474365,
"step": 516
},
{
"epoch": 0.0518,
"grad_norm": 0.08268273621797562,
"learning_rate": 0.002,
"loss": 2.2123360633850098,
"step": 518
},
{
"epoch": 0.052,
"grad_norm": 0.08146864175796509,
"learning_rate": 0.002,
"loss": 2.1841495037078857,
"step": 520
},
{
"epoch": 0.0522,
"grad_norm": 0.07715478539466858,
"learning_rate": 0.002,
"loss": 2.180346727371216,
"step": 522
},
{
"epoch": 0.0524,
"grad_norm": 0.08240430802106857,
"learning_rate": 0.002,
"loss": 2.3442516326904297,
"step": 524
},
{
"epoch": 0.0526,
"grad_norm": 0.0884537398815155,
"learning_rate": 0.002,
"loss": 2.1030502319335938,
"step": 526
},
{
"epoch": 0.0528,
"grad_norm": 0.08892269432544708,
"learning_rate": 0.002,
"loss": 2.0394692420959473,
"step": 528
},
{
"epoch": 0.053,
"grad_norm": 0.08169779181480408,
"learning_rate": 0.002,
"loss": 2.1839816570281982,
"step": 530
},
{
"epoch": 0.0532,
"grad_norm": 0.08319083601236343,
"learning_rate": 0.002,
"loss": 2.20290470123291,
"step": 532
},
{
"epoch": 0.0534,
"grad_norm": 0.08176832646131516,
"learning_rate": 0.002,
"loss": 2.1976611614227295,
"step": 534
},
{
"epoch": 0.0536,
"grad_norm": 0.08160136640071869,
"learning_rate": 0.002,
"loss": 2.1640212535858154,
"step": 536
},
{
"epoch": 0.0538,
"grad_norm": 0.08101780712604523,
"learning_rate": 0.002,
"loss": 2.1780595779418945,
"step": 538
},
{
"epoch": 0.054,
"grad_norm": 0.09154853224754333,
"learning_rate": 0.002,
"loss": 2.047353506088257,
"step": 540
},
{
"epoch": 0.0542,
"grad_norm": 0.07948208600282669,
"learning_rate": 0.002,
"loss": 2.0542216300964355,
"step": 542
},
{
"epoch": 0.0544,
"grad_norm": 0.08215510100126266,
"learning_rate": 0.002,
"loss": 2.145672082901001,
"step": 544
},
{
"epoch": 0.0546,
"grad_norm": 0.0831441879272461,
"learning_rate": 0.002,
"loss": 2.3219962120056152,
"step": 546
},
{
"epoch": 0.0548,
"grad_norm": 0.08922858536243439,
"learning_rate": 0.002,
"loss": 2.1470274925231934,
"step": 548
},
{
"epoch": 0.055,
"grad_norm": 0.08093307167291641,
"learning_rate": 0.002,
"loss": 2.231550931930542,
"step": 550
},
{
"epoch": 0.0552,
"grad_norm": 0.08403438329696655,
"learning_rate": 0.002,
"loss": 2.1990110874176025,
"step": 552
},
{
"epoch": 0.0554,
"grad_norm": 0.08278726041316986,
"learning_rate": 0.002,
"loss": 2.0193846225738525,
"step": 554
},
{
"epoch": 0.0556,
"grad_norm": 0.07895541191101074,
"learning_rate": 0.002,
"loss": 2.1149227619171143,
"step": 556
},
{
"epoch": 0.0558,
"grad_norm": 0.0833519846200943,
"learning_rate": 0.002,
"loss": 2.2097527980804443,
"step": 558
},
{
"epoch": 0.056,
"grad_norm": 0.08651773631572723,
"learning_rate": 0.002,
"loss": 2.105790138244629,
"step": 560
},
{
"epoch": 0.0562,
"grad_norm": 0.08364711701869965,
"learning_rate": 0.002,
"loss": 2.079378366470337,
"step": 562
},
{
"epoch": 0.0564,
"grad_norm": 0.08137064427137375,
"learning_rate": 0.002,
"loss": 2.1865992546081543,
"step": 564
},
{
"epoch": 0.0566,
"grad_norm": 0.07938656955957413,
"learning_rate": 0.002,
"loss": 2.030928373336792,
"step": 566
},
{
"epoch": 0.0568,
"grad_norm": 0.08281201869249344,
"learning_rate": 0.002,
"loss": 2.0662636756896973,
"step": 568
},
{
"epoch": 0.057,
"grad_norm": 0.07970619946718216,
"learning_rate": 0.002,
"loss": 2.1972622871398926,
"step": 570
},
{
"epoch": 0.0572,
"grad_norm": 0.08183321356773376,
"learning_rate": 0.002,
"loss": 2.09775972366333,
"step": 572
},
{
"epoch": 0.0574,
"grad_norm": 0.08047384023666382,
"learning_rate": 0.002,
"loss": 2.1719322204589844,
"step": 574
},
{
"epoch": 0.0576,
"grad_norm": 0.09200158715248108,
"learning_rate": 0.002,
"loss": 2.1225814819335938,
"step": 576
},
{
"epoch": 0.0578,
"grad_norm": 0.08861026167869568,
"learning_rate": 0.002,
"loss": 2.118323564529419,
"step": 578
},
{
"epoch": 0.058,
"grad_norm": 0.08256974816322327,
"learning_rate": 0.002,
"loss": 2.078427791595459,
"step": 580
},
{
"epoch": 0.0582,
"grad_norm": 0.08179314434528351,
"learning_rate": 0.002,
"loss": 2.0977330207824707,
"step": 582
},
{
"epoch": 0.0584,
"grad_norm": 0.08758915215730667,
"learning_rate": 0.002,
"loss": 2.152599334716797,
"step": 584
},
{
"epoch": 0.0586,
"grad_norm": 0.08636537194252014,
"learning_rate": 0.002,
"loss": 2.138857364654541,
"step": 586
},
{
"epoch": 0.0588,
"grad_norm": 0.08421066403388977,
"learning_rate": 0.002,
"loss": 2.1526072025299072,
"step": 588
},
{
"epoch": 0.059,
"grad_norm": 0.08765074610710144,
"learning_rate": 0.002,
"loss": 2.213284969329834,
"step": 590
},
{
"epoch": 0.0592,
"grad_norm": 0.08737796545028687,
"learning_rate": 0.002,
"loss": 2.060230016708374,
"step": 592
},
{
"epoch": 0.0594,
"grad_norm": 0.08773931115865707,
"learning_rate": 0.002,
"loss": 2.241100549697876,
"step": 594
},
{
"epoch": 0.0596,
"grad_norm": 0.08091755956411362,
"learning_rate": 0.002,
"loss": 2.0961341857910156,
"step": 596
},
{
"epoch": 0.0598,
"grad_norm": 0.08538051694631577,
"learning_rate": 0.002,
"loss": 2.054225206375122,
"step": 598
},
{
"epoch": 0.06,
"grad_norm": 0.08402661979198456,
"learning_rate": 0.002,
"loss": 2.155684232711792,
"step": 600
},
{
"epoch": 0.0602,
"grad_norm": 0.08178598433732986,
"learning_rate": 0.002,
"loss": 2.0045166015625,
"step": 602
},
{
"epoch": 0.0604,
"grad_norm": 0.08665584772825241,
"learning_rate": 0.002,
"loss": 2.236240863800049,
"step": 604
},
{
"epoch": 0.0606,
"grad_norm": 0.0843273401260376,
"learning_rate": 0.002,
"loss": 2.2379565238952637,
"step": 606
},
{
"epoch": 0.0608,
"grad_norm": 0.08679091185331345,
"learning_rate": 0.002,
"loss": 2.1263556480407715,
"step": 608
},
{
"epoch": 0.061,
"grad_norm": 0.08113043010234833,
"learning_rate": 0.002,
"loss": 2.0564451217651367,
"step": 610
},
{
"epoch": 0.0612,
"grad_norm": 0.08990326523780823,
"learning_rate": 0.002,
"loss": 2.158761978149414,
"step": 612
},
{
"epoch": 0.0614,
"grad_norm": 0.08801403641700745,
"learning_rate": 0.002,
"loss": 2.0868020057678223,
"step": 614
},
{
"epoch": 0.0616,
"grad_norm": 0.08505520224571228,
"learning_rate": 0.002,
"loss": 2.126589775085449,
"step": 616
},
{
"epoch": 0.0618,
"grad_norm": 0.0785745307803154,
"learning_rate": 0.002,
"loss": 2.1435048580169678,
"step": 618
},
{
"epoch": 0.062,
"grad_norm": 0.08204290270805359,
"learning_rate": 0.002,
"loss": 2.014315366744995,
"step": 620
},
{
"epoch": 0.0622,
"grad_norm": 0.08544141054153442,
"learning_rate": 0.002,
"loss": 2.120809555053711,
"step": 622
},
{
"epoch": 0.0624,
"grad_norm": 0.08234025537967682,
"learning_rate": 0.002,
"loss": 2.0794384479522705,
"step": 624
},
{
"epoch": 0.0626,
"grad_norm": 0.0837748721241951,
"learning_rate": 0.002,
"loss": 2.1955528259277344,
"step": 626
},
{
"epoch": 0.0628,
"grad_norm": 0.08219987154006958,
"learning_rate": 0.002,
"loss": 2.1821634769439697,
"step": 628
},
{
"epoch": 0.063,
"grad_norm": 0.0797683373093605,
"learning_rate": 0.002,
"loss": 2.0721824169158936,
"step": 630
},
{
"epoch": 0.0632,
"grad_norm": 0.08416537195444107,
"learning_rate": 0.002,
"loss": 2.0227813720703125,
"step": 632
},
{
"epoch": 0.0634,
"grad_norm": 0.08238273859024048,
"learning_rate": 0.002,
"loss": 2.1617636680603027,
"step": 634
},
{
"epoch": 0.0636,
"grad_norm": 0.0783059373497963,
"learning_rate": 0.002,
"loss": 2.0238099098205566,
"step": 636
},
{
"epoch": 0.0638,
"grad_norm": 0.08266188204288483,
"learning_rate": 0.002,
"loss": 2.1875689029693604,
"step": 638
},
{
"epoch": 0.064,
"grad_norm": 0.09303969144821167,
"learning_rate": 0.002,
"loss": 2.1887948513031006,
"step": 640
},
{
"epoch": 0.0642,
"grad_norm": 0.0803394764661789,
"learning_rate": 0.002,
"loss": 2.019932985305786,
"step": 642
},
{
"epoch": 0.0644,
"grad_norm": 0.07706098258495331,
"learning_rate": 0.002,
"loss": 1.9300638437271118,
"step": 644
},
{
"epoch": 0.0646,
"grad_norm": 0.09266029298305511,
"learning_rate": 0.002,
"loss": 2.0977258682250977,
"step": 646
},
{
"epoch": 0.0648,
"grad_norm": 0.08937431871891022,
"learning_rate": 0.002,
"loss": 2.1930906772613525,
"step": 648
},
{
"epoch": 0.065,
"grad_norm": 0.08725675195455551,
"learning_rate": 0.002,
"loss": 2.2145559787750244,
"step": 650
},
{
"epoch": 0.0652,
"grad_norm": 0.08516485244035721,
"learning_rate": 0.002,
"loss": 2.1822683811187744,
"step": 652
},
{
"epoch": 0.0654,
"grad_norm": 0.08013498038053513,
"learning_rate": 0.002,
"loss": 2.1743760108947754,
"step": 654
},
{
"epoch": 0.0656,
"grad_norm": 0.09914427250623703,
"learning_rate": 0.002,
"loss": 2.083150625228882,
"step": 656
},
{
"epoch": 0.0658,
"grad_norm": 0.08586171269416809,
"learning_rate": 0.002,
"loss": 2.095839023590088,
"step": 658
},
{
"epoch": 0.066,
"grad_norm": 0.08430416136980057,
"learning_rate": 0.002,
"loss": 2.066486120223999,
"step": 660
},
{
"epoch": 0.0662,
"grad_norm": 0.08449099212884903,
"learning_rate": 0.002,
"loss": 2.2718098163604736,
"step": 662
},
{
"epoch": 0.0664,
"grad_norm": 0.08435925096273422,
"learning_rate": 0.002,
"loss": 2.073138475418091,
"step": 664
},
{
"epoch": 0.0666,
"grad_norm": 0.08241667598485947,
"learning_rate": 0.002,
"loss": 2.0648231506347656,
"step": 666
},
{
"epoch": 0.0668,
"grad_norm": 0.08146832883358002,
"learning_rate": 0.002,
"loss": 2.207123279571533,
"step": 668
},
{
"epoch": 0.067,
"grad_norm": 0.0855381116271019,
"learning_rate": 0.002,
"loss": 2.183933734893799,
"step": 670
},
{
"epoch": 0.0672,
"grad_norm": 0.08441325277090073,
"learning_rate": 0.002,
"loss": 2.2200825214385986,
"step": 672
},
{
"epoch": 0.0674,
"grad_norm": 0.08546078205108643,
"learning_rate": 0.002,
"loss": 2.2786710262298584,
"step": 674
},
{
"epoch": 0.0676,
"grad_norm": 0.08342700451612473,
"learning_rate": 0.002,
"loss": 2.1871044635772705,
"step": 676
},
{
"epoch": 0.0678,
"grad_norm": 0.08333663642406464,
"learning_rate": 0.002,
"loss": 2.140864372253418,
"step": 678
},
{
"epoch": 0.068,
"grad_norm": 0.08477053046226501,
"learning_rate": 0.002,
"loss": 2.2283010482788086,
"step": 680
},
{
"epoch": 0.0682,
"grad_norm": 0.0860549658536911,
"learning_rate": 0.002,
"loss": 2.1384453773498535,
"step": 682
},
{
"epoch": 0.0684,
"grad_norm": 0.08343090116977692,
"learning_rate": 0.002,
"loss": 2.193233013153076,
"step": 684
},
{
"epoch": 0.0686,
"grad_norm": 0.08568835258483887,
"learning_rate": 0.002,
"loss": 2.2944540977478027,
"step": 686
},
{
"epoch": 0.0688,
"grad_norm": 0.07981055974960327,
"learning_rate": 0.002,
"loss": 2.0588438510894775,
"step": 688
},
{
"epoch": 0.069,
"grad_norm": 0.08547872304916382,
"learning_rate": 0.002,
"loss": 2.1976232528686523,
"step": 690
},
{
"epoch": 0.0692,
"grad_norm": 0.08215370029211044,
"learning_rate": 0.002,
"loss": 2.078406572341919,
"step": 692
},
{
"epoch": 0.0694,
"grad_norm": 0.08741050958633423,
"learning_rate": 0.002,
"loss": 2.054339647293091,
"step": 694
},
{
"epoch": 0.0696,
"grad_norm": 0.08312851190567017,
"learning_rate": 0.002,
"loss": 2.1488704681396484,
"step": 696
},
{
"epoch": 0.0698,
"grad_norm": 0.08758502453565598,
"learning_rate": 0.002,
"loss": 2.1626691818237305,
"step": 698
},
{
"epoch": 0.07,
"grad_norm": 0.08032335340976715,
"learning_rate": 0.002,
"loss": 2.0651960372924805,
"step": 700
},
{
"epoch": 0.0702,
"grad_norm": 0.08148928731679916,
"learning_rate": 0.002,
"loss": 2.2758400440216064,
"step": 702
},
{
"epoch": 0.0704,
"grad_norm": 0.08187469840049744,
"learning_rate": 0.002,
"loss": 2.035198926925659,
"step": 704
},
{
"epoch": 0.0706,
"grad_norm": 0.08114253729581833,
"learning_rate": 0.002,
"loss": 2.0897557735443115,
"step": 706
},
{
"epoch": 0.0708,
"grad_norm": 0.08664634823799133,
"learning_rate": 0.002,
"loss": 2.111149787902832,
"step": 708
},
{
"epoch": 0.071,
"grad_norm": 0.08153948187828064,
"learning_rate": 0.002,
"loss": 2.1473336219787598,
"step": 710
},
{
"epoch": 0.0712,
"grad_norm": 0.07831638306379318,
"learning_rate": 0.002,
"loss": 2.207933187484741,
"step": 712
},
{
"epoch": 0.0714,
"grad_norm": 0.0807049423456192,
"learning_rate": 0.002,
"loss": 2.0551791191101074,
"step": 714
},
{
"epoch": 0.0716,
"grad_norm": 0.0798383504152298,
"learning_rate": 0.002,
"loss": 2.056682825088501,
"step": 716
},
{
"epoch": 0.0718,
"grad_norm": 0.07966647297143936,
"learning_rate": 0.002,
"loss": 2.0972092151641846,
"step": 718
},
{
"epoch": 0.072,
"grad_norm": 0.07914122194051743,
"learning_rate": 0.002,
"loss": 2.0922281742095947,
"step": 720
},
{
"epoch": 0.0722,
"grad_norm": 0.08131422847509384,
"learning_rate": 0.002,
"loss": 2.0901072025299072,
"step": 722
},
{
"epoch": 0.0724,
"grad_norm": 0.0859156921505928,
"learning_rate": 0.002,
"loss": 2.2403104305267334,
"step": 724
},
{
"epoch": 0.0726,
"grad_norm": 0.08557098358869553,
"learning_rate": 0.002,
"loss": 2.0302278995513916,
"step": 726
},
{
"epoch": 0.0728,
"grad_norm": 0.08160679787397385,
"learning_rate": 0.002,
"loss": 2.2086362838745117,
"step": 728
},
{
"epoch": 0.073,
"grad_norm": 0.09040449559688568,
"learning_rate": 0.002,
"loss": 2.1949613094329834,
"step": 730
},
{
"epoch": 0.0732,
"grad_norm": 0.09031929075717926,
"learning_rate": 0.002,
"loss": 2.130150318145752,
"step": 732
},
{
"epoch": 0.0734,
"grad_norm": 0.07965083420276642,
"learning_rate": 0.002,
"loss": 2.178861141204834,
"step": 734
},
{
"epoch": 0.0736,
"grad_norm": 0.0887785479426384,
"learning_rate": 0.002,
"loss": 2.233626127243042,
"step": 736
},
{
"epoch": 0.0738,
"grad_norm": 0.0869516059756279,
"learning_rate": 0.002,
"loss": 2.218959093093872,
"step": 738
},
{
"epoch": 0.074,
"grad_norm": 0.08595601469278336,
"learning_rate": 0.002,
"loss": 2.0451598167419434,
"step": 740
},
{
"epoch": 0.0742,
"grad_norm": 0.08284755051136017,
"learning_rate": 0.002,
"loss": 2.213379383087158,
"step": 742
},
{
"epoch": 0.0744,
"grad_norm": 0.08810687810182571,
"learning_rate": 0.002,
"loss": 2.143005132675171,
"step": 744
},
{
"epoch": 0.0746,
"grad_norm": 0.08963578194379807,
"learning_rate": 0.002,
"loss": 2.181699514389038,
"step": 746
},
{
"epoch": 0.0748,
"grad_norm": 0.08201977610588074,
"learning_rate": 0.002,
"loss": 2.1216704845428467,
"step": 748
},
{
"epoch": 0.075,
"grad_norm": 0.08557361364364624,
"learning_rate": 0.002,
"loss": 2.0755794048309326,
"step": 750
},
{
"epoch": 0.0752,
"grad_norm": 0.08696344494819641,
"learning_rate": 0.002,
"loss": 2.167257070541382,
"step": 752
},
{
"epoch": 0.0754,
"grad_norm": 0.08058693259954453,
"learning_rate": 0.002,
"loss": 2.0220694541931152,
"step": 754
},
{
"epoch": 0.0756,
"grad_norm": 0.0869988203048706,
"learning_rate": 0.002,
"loss": 2.136317014694214,
"step": 756
},
{
"epoch": 0.0758,
"grad_norm": 0.08319912850856781,
"learning_rate": 0.002,
"loss": 2.085160255432129,
"step": 758
},
{
"epoch": 0.076,
"grad_norm": 0.08059532195329666,
"learning_rate": 0.002,
"loss": 2.246485710144043,
"step": 760
},
{
"epoch": 0.0762,
"grad_norm": 0.08649477362632751,
"learning_rate": 0.002,
"loss": 2.065037965774536,
"step": 762
},
{
"epoch": 0.0764,
"grad_norm": 0.08678106218576431,
"learning_rate": 0.002,
"loss": 2.117459535598755,
"step": 764
},
{
"epoch": 0.0766,
"grad_norm": 0.08034630864858627,
"learning_rate": 0.002,
"loss": 2.2410354614257812,
"step": 766
},
{
"epoch": 0.0768,
"grad_norm": 0.08014260232448578,
"learning_rate": 0.002,
"loss": 1.9938665628433228,
"step": 768
},
{
"epoch": 0.077,
"grad_norm": 0.07953570038080215,
"learning_rate": 0.002,
"loss": 2.079798460006714,
"step": 770
},
{
"epoch": 0.0772,
"grad_norm": 0.08154676109552383,
"learning_rate": 0.002,
"loss": 2.1257548332214355,
"step": 772
},
{
"epoch": 0.0774,
"grad_norm": 0.08637973666191101,
"learning_rate": 0.002,
"loss": 2.071361780166626,
"step": 774
},
{
"epoch": 0.0776,
"grad_norm": 0.08340935409069061,
"learning_rate": 0.002,
"loss": 2.0494585037231445,
"step": 776
},
{
"epoch": 0.0778,
"grad_norm": 0.08452790975570679,
"learning_rate": 0.002,
"loss": 2.11816668510437,
"step": 778
},
{
"epoch": 0.078,
"grad_norm": 0.08489201217889786,
"learning_rate": 0.002,
"loss": 2.263950824737549,
"step": 780
},
{
"epoch": 0.0782,
"grad_norm": 0.08214303851127625,
"learning_rate": 0.002,
"loss": 2.110708475112915,
"step": 782
},
{
"epoch": 0.0784,
"grad_norm": 0.0857962816953659,
"learning_rate": 0.002,
"loss": 2.1584527492523193,
"step": 784
},
{
"epoch": 0.0786,
"grad_norm": 0.09196900576353073,
"learning_rate": 0.002,
"loss": 2.220926284790039,
"step": 786
},
{
"epoch": 0.0788,
"grad_norm": 0.08537618815898895,
"learning_rate": 0.002,
"loss": 1.9997313022613525,
"step": 788
},
{
"epoch": 0.079,
"grad_norm": 0.08131696283817291,
"learning_rate": 0.002,
"loss": 2.1834237575531006,
"step": 790
},
{
"epoch": 0.0792,
"grad_norm": 0.08764319121837616,
"learning_rate": 0.002,
"loss": 2.1230621337890625,
"step": 792
},
{
"epoch": 0.0794,
"grad_norm": 0.08338648825883865,
"learning_rate": 0.002,
"loss": 2.2192790508270264,
"step": 794
},
{
"epoch": 0.0796,
"grad_norm": 0.08298825472593307,
"learning_rate": 0.002,
"loss": 2.2677149772644043,
"step": 796
},
{
"epoch": 0.0798,
"grad_norm": 0.08725271373987198,
"learning_rate": 0.002,
"loss": 2.1610474586486816,
"step": 798
},
{
"epoch": 0.08,
"grad_norm": 0.08009359240531921,
"learning_rate": 0.002,
"loss": 1.9144237041473389,
"step": 800
},
{
"epoch": 0.0802,
"grad_norm": 0.09149710088968277,
"learning_rate": 0.002,
"loss": 2.0341458320617676,
"step": 802
},
{
"epoch": 0.0804,
"grad_norm": 0.08096542954444885,
"learning_rate": 0.002,
"loss": 1.964756727218628,
"step": 804
},
{
"epoch": 0.0806,
"grad_norm": 0.08604897558689117,
"learning_rate": 0.002,
"loss": 2.282911777496338,
"step": 806
},
{
"epoch": 0.0808,
"grad_norm": 0.08179110288619995,
"learning_rate": 0.002,
"loss": 2.0278642177581787,
"step": 808
},
{
"epoch": 0.081,
"grad_norm": 0.08169227093458176,
"learning_rate": 0.002,
"loss": 2.140594720840454,
"step": 810
},
{
"epoch": 0.0812,
"grad_norm": 0.08215657621622086,
"learning_rate": 0.002,
"loss": 2.201343059539795,
"step": 812
},
{
"epoch": 0.0814,
"grad_norm": 0.08443048596382141,
"learning_rate": 0.002,
"loss": 2.028852939605713,
"step": 814
},
{
"epoch": 0.0816,
"grad_norm": 0.07630164921283722,
"learning_rate": 0.002,
"loss": 1.9512213468551636,
"step": 816
},
{
"epoch": 0.0818,
"grad_norm": 0.08184944838285446,
"learning_rate": 0.002,
"loss": 2.151747941970825,
"step": 818
},
{
"epoch": 0.082,
"grad_norm": 0.0822487398982048,
"learning_rate": 0.002,
"loss": 2.098198890686035,
"step": 820
},
{
"epoch": 0.0822,
"grad_norm": 0.08129869401454926,
"learning_rate": 0.002,
"loss": 2.0526070594787598,
"step": 822
},
{
"epoch": 0.0824,
"grad_norm": 0.08036435395479202,
"learning_rate": 0.002,
"loss": 1.9685380458831787,
"step": 824
},
{
"epoch": 0.0826,
"grad_norm": 0.0947558656334877,
"learning_rate": 0.002,
"loss": 2.161348581314087,
"step": 826
},
{
"epoch": 0.0828,
"grad_norm": 0.08436311781406403,
"learning_rate": 0.002,
"loss": 2.073805093765259,
"step": 828
},
{
"epoch": 0.083,
"grad_norm": 0.08767018467187881,
"learning_rate": 0.002,
"loss": 2.122269630432129,
"step": 830
},
{
"epoch": 0.0832,
"grad_norm": 0.08740667253732681,
"learning_rate": 0.002,
"loss": 2.2189927101135254,
"step": 832
},
{
"epoch": 0.0834,
"grad_norm": 0.08330400288105011,
"learning_rate": 0.002,
"loss": 1.9072109460830688,
"step": 834
},
{
"epoch": 0.0836,
"grad_norm": 0.08315642178058624,
"learning_rate": 0.002,
"loss": 2.1749982833862305,
"step": 836
},
{
"epoch": 0.0838,
"grad_norm": 0.08645769208669662,
"learning_rate": 0.002,
"loss": 2.106210231781006,
"step": 838
},
{
"epoch": 0.084,
"grad_norm": 0.08705995231866837,
"learning_rate": 0.002,
"loss": 1.9553714990615845,
"step": 840
},
{
"epoch": 0.0842,
"grad_norm": 0.08447366207838058,
"learning_rate": 0.002,
"loss": 1.9656076431274414,
"step": 842
},
{
"epoch": 0.0844,
"grad_norm": 0.08096335083246231,
"learning_rate": 0.002,
"loss": 2.0138046741485596,
"step": 844
},
{
"epoch": 0.0846,
"grad_norm": 0.08387453109025955,
"learning_rate": 0.002,
"loss": 2.111412763595581,
"step": 846
},
{
"epoch": 0.0848,
"grad_norm": 0.0926416888833046,
"learning_rate": 0.002,
"loss": 1.963688850402832,
"step": 848
},
{
"epoch": 0.085,
"grad_norm": 0.08636775612831116,
"learning_rate": 0.002,
"loss": 2.143697500228882,
"step": 850
},
{
"epoch": 0.0852,
"grad_norm": 0.07966245710849762,
"learning_rate": 0.002,
"loss": 2.0497334003448486,
"step": 852
},
{
"epoch": 0.0854,
"grad_norm": 0.08662667125463486,
"learning_rate": 0.002,
"loss": 2.122199296951294,
"step": 854
},
{
"epoch": 0.0856,
"grad_norm": 0.0864025205373764,
"learning_rate": 0.002,
"loss": 1.9273364543914795,
"step": 856
},
{
"epoch": 0.0858,
"grad_norm": 0.08271387219429016,
"learning_rate": 0.002,
"loss": 1.9567370414733887,
"step": 858
},
{
"epoch": 0.086,
"grad_norm": 0.07819661498069763,
"learning_rate": 0.002,
"loss": 2.027707099914551,
"step": 860
},
{
"epoch": 0.0862,
"grad_norm": 0.07887491583824158,
"learning_rate": 0.002,
"loss": 2.119666576385498,
"step": 862
},
{
"epoch": 0.0864,
"grad_norm": 0.08716923743486404,
"learning_rate": 0.002,
"loss": 2.034428358078003,
"step": 864
},
{
"epoch": 0.0866,
"grad_norm": 0.08342289179563522,
"learning_rate": 0.002,
"loss": 2.031719923019409,
"step": 866
},
{
"epoch": 0.0868,
"grad_norm": 0.08829818665981293,
"learning_rate": 0.002,
"loss": 2.0985631942749023,
"step": 868
},
{
"epoch": 0.087,
"grad_norm": 0.08318227529525757,
"learning_rate": 0.002,
"loss": 2.04681396484375,
"step": 870
},
{
"epoch": 0.0872,
"grad_norm": 0.07766366750001907,
"learning_rate": 0.002,
"loss": 2.121795415878296,
"step": 872
},
{
"epoch": 0.0874,
"grad_norm": 0.08091418445110321,
"learning_rate": 0.002,
"loss": 2.047576904296875,
"step": 874
},
{
"epoch": 0.0876,
"grad_norm": 0.08225125074386597,
"learning_rate": 0.002,
"loss": 2.079455614089966,
"step": 876
},
{
"epoch": 0.0878,
"grad_norm": 0.09061382710933685,
"learning_rate": 0.002,
"loss": 2.0008864402770996,
"step": 878
},
{
"epoch": 0.088,
"grad_norm": 0.08554644137620926,
"learning_rate": 0.002,
"loss": 2.025676965713501,
"step": 880
},
{
"epoch": 0.0882,
"grad_norm": 0.08290041238069534,
"learning_rate": 0.002,
"loss": 2.0943896770477295,
"step": 882
},
{
"epoch": 0.0884,
"grad_norm": 0.07951974868774414,
"learning_rate": 0.002,
"loss": 2.121314764022827,
"step": 884
},
{
"epoch": 0.0886,
"grad_norm": 0.08930447697639465,
"learning_rate": 0.002,
"loss": 1.9818767309188843,
"step": 886
},
{
"epoch": 0.0888,
"grad_norm": 0.08254047483205795,
"learning_rate": 0.002,
"loss": 2.1701200008392334,
"step": 888
},
{
"epoch": 0.089,
"grad_norm": 0.09686768800020218,
"learning_rate": 0.002,
"loss": 2.038996934890747,
"step": 890
},
{
"epoch": 0.0892,
"grad_norm": 0.08204664289951324,
"learning_rate": 0.002,
"loss": 1.9888279438018799,
"step": 892
},
{
"epoch": 0.0894,
"grad_norm": 0.08473488688468933,
"learning_rate": 0.002,
"loss": 2.216445207595825,
"step": 894
},
{
"epoch": 0.0896,
"grad_norm": 0.08001922070980072,
"learning_rate": 0.002,
"loss": 2.15483021736145,
"step": 896
},
{
"epoch": 0.0898,
"grad_norm": 0.0844753161072731,
"learning_rate": 0.002,
"loss": 2.171548843383789,
"step": 898
},
{
"epoch": 0.09,
"grad_norm": 0.08674822747707367,
"learning_rate": 0.002,
"loss": 2.2151987552642822,
"step": 900
},
{
"epoch": 0.0902,
"grad_norm": 0.0803431048989296,
"learning_rate": 0.002,
"loss": 1.9038811922073364,
"step": 902
},
{
"epoch": 0.0904,
"grad_norm": 0.0898161306977272,
"learning_rate": 0.002,
"loss": 2.0242724418640137,
"step": 904
},
{
"epoch": 0.0906,
"grad_norm": 0.08193943649530411,
"learning_rate": 0.002,
"loss": 1.988626480102539,
"step": 906
},
{
"epoch": 0.0908,
"grad_norm": 0.07968460023403168,
"learning_rate": 0.002,
"loss": 2.1576478481292725,
"step": 908
},
{
"epoch": 0.091,
"grad_norm": 0.09334853291511536,
"learning_rate": 0.002,
"loss": 2.1346182823181152,
"step": 910
},
{
"epoch": 0.0912,
"grad_norm": 0.08479738980531693,
"learning_rate": 0.002,
"loss": 2.0742697715759277,
"step": 912
},
{
"epoch": 0.0914,
"grad_norm": 0.09049994498491287,
"learning_rate": 0.002,
"loss": 2.054737091064453,
"step": 914
},
{
"epoch": 0.0916,
"grad_norm": 0.0903417244553566,
"learning_rate": 0.002,
"loss": 2.0382070541381836,
"step": 916
},
{
"epoch": 0.0918,
"grad_norm": 0.0855557844042778,
"learning_rate": 0.002,
"loss": 2.1551735401153564,
"step": 918
},
{
"epoch": 0.092,
"grad_norm": 0.09852523356676102,
"learning_rate": 0.002,
"loss": 2.151409149169922,
"step": 920
},
{
"epoch": 0.0922,
"grad_norm": 0.09058385342359543,
"learning_rate": 0.002,
"loss": 2.147742986679077,
"step": 922
},
{
"epoch": 0.0924,
"grad_norm": 0.08544475585222244,
"learning_rate": 0.002,
"loss": 2.033942222595215,
"step": 924
},
{
"epoch": 0.0926,
"grad_norm": 0.08427257090806961,
"learning_rate": 0.002,
"loss": 1.859273076057434,
"step": 926
},
{
"epoch": 0.0928,
"grad_norm": 0.08283957093954086,
"learning_rate": 0.002,
"loss": 2.099193811416626,
"step": 928
},
{
"epoch": 0.093,
"grad_norm": 0.08537071198225021,
"learning_rate": 0.002,
"loss": 2.1653969287872314,
"step": 930
},
{
"epoch": 0.0932,
"grad_norm": 0.08579947799444199,
"learning_rate": 0.002,
"loss": 2.0872714519500732,
"step": 932
},
{
"epoch": 0.0934,
"grad_norm": 0.082401342689991,
"learning_rate": 0.002,
"loss": 2.078909158706665,
"step": 934
},
{
"epoch": 0.0936,
"grad_norm": 0.0790775865316391,
"learning_rate": 0.002,
"loss": 2.0969159603118896,
"step": 936
},
{
"epoch": 0.0938,
"grad_norm": 0.08064674586057663,
"learning_rate": 0.002,
"loss": 2.0534610748291016,
"step": 938
},
{
"epoch": 0.094,
"grad_norm": 0.08048015832901001,
"learning_rate": 0.002,
"loss": 2.054018974304199,
"step": 940
},
{
"epoch": 0.0942,
"grad_norm": 0.0787394717335701,
"learning_rate": 0.002,
"loss": 2.0691468715667725,
"step": 942
},
{
"epoch": 0.0944,
"grad_norm": 0.08421686291694641,
"learning_rate": 0.002,
"loss": 2.0654053688049316,
"step": 944
},
{
"epoch": 0.0946,
"grad_norm": 0.08260521292686462,
"learning_rate": 0.002,
"loss": 2.090667724609375,
"step": 946
},
{
"epoch": 0.0948,
"grad_norm": 0.08113668859004974,
"learning_rate": 0.002,
"loss": 2.1073262691497803,
"step": 948
},
{
"epoch": 0.095,
"grad_norm": 0.08405879139900208,
"learning_rate": 0.002,
"loss": 2.0921452045440674,
"step": 950
},
{
"epoch": 0.0952,
"grad_norm": 0.08539978414773941,
"learning_rate": 0.002,
"loss": 2.1568100452423096,
"step": 952
},
{
"epoch": 0.0954,
"grad_norm": 0.09084311872720718,
"learning_rate": 0.002,
"loss": 1.9966379404067993,
"step": 954
},
{
"epoch": 0.0956,
"grad_norm": 0.08043288439512253,
"learning_rate": 0.002,
"loss": 2.0659050941467285,
"step": 956
},
{
"epoch": 0.0958,
"grad_norm": 0.08921446651220322,
"learning_rate": 0.002,
"loss": 2.1390604972839355,
"step": 958
},
{
"epoch": 0.096,
"grad_norm": 0.088004469871521,
"learning_rate": 0.002,
"loss": 2.1213271617889404,
"step": 960
},
{
"epoch": 0.0962,
"grad_norm": 0.08578295260667801,
"learning_rate": 0.002,
"loss": 2.1044929027557373,
"step": 962
},
{
"epoch": 0.0964,
"grad_norm": 0.08526584506034851,
"learning_rate": 0.002,
"loss": 2.1497132778167725,
"step": 964
},
{
"epoch": 0.0966,
"grad_norm": 0.08785363286733627,
"learning_rate": 0.002,
"loss": 2.1348042488098145,
"step": 966
},
{
"epoch": 0.0968,
"grad_norm": 0.08324627578258514,
"learning_rate": 0.002,
"loss": 2.1350347995758057,
"step": 968
},
{
"epoch": 0.097,
"grad_norm": 0.0847279503941536,
"learning_rate": 0.002,
"loss": 2.237224578857422,
"step": 970
},
{
"epoch": 0.0972,
"grad_norm": 0.08324167132377625,
"learning_rate": 0.002,
"loss": 2.147554397583008,
"step": 972
},
{
"epoch": 0.0974,
"grad_norm": 0.08121193945407867,
"learning_rate": 0.002,
"loss": 2.091959238052368,
"step": 974
},
{
"epoch": 0.0976,
"grad_norm": 0.09089396148920059,
"learning_rate": 0.002,
"loss": 2.1985602378845215,
"step": 976
},
{
"epoch": 0.0978,
"grad_norm": 0.08862462639808655,
"learning_rate": 0.002,
"loss": 1.9854496717453003,
"step": 978
},
{
"epoch": 0.098,
"grad_norm": 0.07896845042705536,
"learning_rate": 0.002,
"loss": 2.0448176860809326,
"step": 980
},
{
"epoch": 0.0982,
"grad_norm": 0.07985764741897583,
"learning_rate": 0.002,
"loss": 2.1397225856781006,
"step": 982
},
{
"epoch": 0.0984,
"grad_norm": 0.08435620367527008,
"learning_rate": 0.002,
"loss": 2.2018702030181885,
"step": 984
},
{
"epoch": 0.0986,
"grad_norm": 0.08832981437444687,
"learning_rate": 0.002,
"loss": 2.234189748764038,
"step": 986
},
{
"epoch": 0.0988,
"grad_norm": 0.08233267068862915,
"learning_rate": 0.002,
"loss": 2.041668176651001,
"step": 988
},
{
"epoch": 0.099,
"grad_norm": 0.08580370247364044,
"learning_rate": 0.002,
"loss": 2.0748517513275146,
"step": 990
},
{
"epoch": 0.0992,
"grad_norm": 0.08231016248464584,
"learning_rate": 0.002,
"loss": 2.0949463844299316,
"step": 992
},
{
"epoch": 0.0994,
"grad_norm": 0.08033733814954758,
"learning_rate": 0.002,
"loss": 2.0531740188598633,
"step": 994
},
{
"epoch": 0.0996,
"grad_norm": 0.08302079886198044,
"learning_rate": 0.002,
"loss": 2.0564210414886475,
"step": 996
},
{
"epoch": 0.0998,
"grad_norm": 0.08644754439592361,
"learning_rate": 0.002,
"loss": 2.107149124145508,
"step": 998
},
{
"epoch": 0.1,
"grad_norm": 0.08344193547964096,
"learning_rate": 0.002,
"loss": 2.0810580253601074,
"step": 1000
},
{
"epoch": 0.1002,
"grad_norm": 0.08144240826368332,
"learning_rate": 0.002,
"loss": 2.040194511413574,
"step": 1002
},
{
"epoch": 0.1004,
"grad_norm": 0.08272895216941833,
"learning_rate": 0.002,
"loss": 2.0771665573120117,
"step": 1004
},
{
"epoch": 0.1006,
"grad_norm": 0.08309302479028702,
"learning_rate": 0.002,
"loss": 2.1700282096862793,
"step": 1006
},
{
"epoch": 0.1008,
"grad_norm": 0.0821627601981163,
"learning_rate": 0.002,
"loss": 1.9310054779052734,
"step": 1008
},
{
"epoch": 0.101,
"grad_norm": 0.08486980199813843,
"learning_rate": 0.002,
"loss": 1.9665870666503906,
"step": 1010
},
{
"epoch": 0.1012,
"grad_norm": 0.07736771553754807,
"learning_rate": 0.002,
"loss": 2.1191978454589844,
"step": 1012
},
{
"epoch": 0.1014,
"grad_norm": 0.08738190680742264,
"learning_rate": 0.002,
"loss": 2.1413991451263428,
"step": 1014
},
{
"epoch": 0.1016,
"grad_norm": 0.08060958236455917,
"learning_rate": 0.002,
"loss": 2.053593158721924,
"step": 1016
},
{
"epoch": 0.1018,
"grad_norm": 0.08549308031797409,
"learning_rate": 0.002,
"loss": 2.2351629734039307,
"step": 1018
},
{
"epoch": 0.102,
"grad_norm": 0.0951957181096077,
"learning_rate": 0.002,
"loss": 2.0900161266326904,
"step": 1020
},
{
"epoch": 0.1022,
"grad_norm": 0.09305122494697571,
"learning_rate": 0.002,
"loss": 2.0782663822174072,
"step": 1022
},
{
"epoch": 0.1024,
"grad_norm": 0.08745758980512619,
"learning_rate": 0.002,
"loss": 2.112008810043335,
"step": 1024
},
{
"epoch": 0.1026,
"grad_norm": 0.09555797278881073,
"learning_rate": 0.002,
"loss": 2.0057218074798584,
"step": 1026
},
{
"epoch": 0.1028,
"grad_norm": 0.08803076297044754,
"learning_rate": 0.002,
"loss": 1.9368822574615479,
"step": 1028
},
{
"epoch": 0.103,
"grad_norm": 0.08392704278230667,
"learning_rate": 0.002,
"loss": 2.203640937805176,
"step": 1030
},
{
"epoch": 0.1032,
"grad_norm": 0.09330909699201584,
"learning_rate": 0.002,
"loss": 1.9989039897918701,
"step": 1032
},
{
"epoch": 0.1034,
"grad_norm": 0.08549216389656067,
"learning_rate": 0.002,
"loss": 2.0902061462402344,
"step": 1034
},
{
"epoch": 0.1036,
"grad_norm": 0.08370019495487213,
"learning_rate": 0.002,
"loss": 2.1310465335845947,
"step": 1036
},
{
"epoch": 0.1038,
"grad_norm": 0.09530872106552124,
"learning_rate": 0.002,
"loss": 2.1101088523864746,
"step": 1038
},
{
"epoch": 0.104,
"grad_norm": 0.08453866839408875,
"learning_rate": 0.002,
"loss": 2.044961929321289,
"step": 1040
},
{
"epoch": 0.1042,
"grad_norm": 0.08468504995107651,
"learning_rate": 0.002,
"loss": 2.267702102661133,
"step": 1042
},
{
"epoch": 0.1044,
"grad_norm": 0.08485621213912964,
"learning_rate": 0.002,
"loss": 2.112591028213501,
"step": 1044
},
{
"epoch": 0.1046,
"grad_norm": 0.08785756677389145,
"learning_rate": 0.002,
"loss": 1.9623355865478516,
"step": 1046
},
{
"epoch": 0.1048,
"grad_norm": 0.09318473190069199,
"learning_rate": 0.002,
"loss": 2.1089303493499756,
"step": 1048
},
{
"epoch": 0.105,
"grad_norm": 0.08206820487976074,
"learning_rate": 0.002,
"loss": 2.2320778369903564,
"step": 1050
},
{
"epoch": 0.1052,
"grad_norm": 0.08517453819513321,
"learning_rate": 0.002,
"loss": 2.005950927734375,
"step": 1052
},
{
"epoch": 0.1054,
"grad_norm": 0.0872499942779541,
"learning_rate": 0.002,
"loss": 2.032996654510498,
"step": 1054
},
{
"epoch": 0.1056,
"grad_norm": 0.08218297362327576,
"learning_rate": 0.002,
"loss": 1.9941152334213257,
"step": 1056
},
{
"epoch": 0.1058,
"grad_norm": 0.08805368840694427,
"learning_rate": 0.002,
"loss": 2.1603922843933105,
"step": 1058
},
{
"epoch": 0.106,
"grad_norm": 0.0835268646478653,
"learning_rate": 0.002,
"loss": 1.8843207359313965,
"step": 1060
},
{
"epoch": 0.1062,
"grad_norm": 0.08661171793937683,
"learning_rate": 0.002,
"loss": 2.143033504486084,
"step": 1062
},
{
"epoch": 0.1064,
"grad_norm": 0.08741480112075806,
"learning_rate": 0.002,
"loss": 1.9643219709396362,
"step": 1064
},
{
"epoch": 0.1066,
"grad_norm": 0.08474089205265045,
"learning_rate": 0.002,
"loss": 1.9824506044387817,
"step": 1066
},
{
"epoch": 0.1068,
"grad_norm": 0.08600416034460068,
"learning_rate": 0.002,
"loss": 2.0218515396118164,
"step": 1068
},
{
"epoch": 0.107,
"grad_norm": 0.09156841784715652,
"learning_rate": 0.002,
"loss": 1.9574592113494873,
"step": 1070
},
{
"epoch": 0.1072,
"grad_norm": 0.07863934338092804,
"learning_rate": 0.002,
"loss": 1.9684548377990723,
"step": 1072
},
{
"epoch": 0.1074,
"grad_norm": 0.08361982554197311,
"learning_rate": 0.002,
"loss": 2.0725059509277344,
"step": 1074
},
{
"epoch": 0.1076,
"grad_norm": 0.08357243984937668,
"learning_rate": 0.002,
"loss": 2.0255863666534424,
"step": 1076
},
{
"epoch": 0.1078,
"grad_norm": 0.08284170925617218,
"learning_rate": 0.002,
"loss": 2.123167037963867,
"step": 1078
},
{
"epoch": 0.108,
"grad_norm": 0.08012061566114426,
"learning_rate": 0.002,
"loss": 1.9576153755187988,
"step": 1080
},
{
"epoch": 0.1082,
"grad_norm": 0.08395293354988098,
"learning_rate": 0.002,
"loss": 2.124877691268921,
"step": 1082
},
{
"epoch": 0.1084,
"grad_norm": 0.08838649839162827,
"learning_rate": 0.002,
"loss": 2.0684261322021484,
"step": 1084
},
{
"epoch": 0.1086,
"grad_norm": 0.08425189554691315,
"learning_rate": 0.002,
"loss": 1.8929017782211304,
"step": 1086
},
{
"epoch": 0.1088,
"grad_norm": 0.08429091423749924,
"learning_rate": 0.002,
"loss": 2.0337469577789307,
"step": 1088
},
{
"epoch": 0.109,
"grad_norm": 0.08154532313346863,
"learning_rate": 0.002,
"loss": 2.051666498184204,
"step": 1090
},
{
"epoch": 0.1092,
"grad_norm": 0.09230396896600723,
"learning_rate": 0.002,
"loss": 2.0522727966308594,
"step": 1092
},
{
"epoch": 0.1094,
"grad_norm": 0.08772566169500351,
"learning_rate": 0.002,
"loss": 2.1167776584625244,
"step": 1094
},
{
"epoch": 0.1096,
"grad_norm": 0.08110581338405609,
"learning_rate": 0.002,
"loss": 2.104281425476074,
"step": 1096
},
{
"epoch": 0.1098,
"grad_norm": 0.08904805779457092,
"learning_rate": 0.002,
"loss": 2.062429189682007,
"step": 1098
},
{
"epoch": 0.11,
"grad_norm": 0.08452310413122177,
"learning_rate": 0.002,
"loss": 2.104130744934082,
"step": 1100
},
{
"epoch": 0.1102,
"grad_norm": 0.0857018306851387,
"learning_rate": 0.002,
"loss": 2.093200206756592,
"step": 1102
},
{
"epoch": 0.1104,
"grad_norm": 0.08385004103183746,
"learning_rate": 0.002,
"loss": 2.0353457927703857,
"step": 1104
},
{
"epoch": 0.1106,
"grad_norm": 0.08778984844684601,
"learning_rate": 0.002,
"loss": 2.123830556869507,
"step": 1106
},
{
"epoch": 0.1108,
"grad_norm": 0.08281433582305908,
"learning_rate": 0.002,
"loss": 2.0778861045837402,
"step": 1108
},
{
"epoch": 0.111,
"grad_norm": 0.07921630889177322,
"learning_rate": 0.002,
"loss": 2.048490285873413,
"step": 1110
},
{
"epoch": 0.1112,
"grad_norm": 0.08667758107185364,
"learning_rate": 0.002,
"loss": 2.0113866329193115,
"step": 1112
},
{
"epoch": 0.1114,
"grad_norm": 0.08297854661941528,
"learning_rate": 0.002,
"loss": 1.9370574951171875,
"step": 1114
},
{
"epoch": 0.1116,
"grad_norm": 0.08787955343723297,
"learning_rate": 0.002,
"loss": 2.1908116340637207,
"step": 1116
},
{
"epoch": 0.1118,
"grad_norm": 0.09227648377418518,
"learning_rate": 0.002,
"loss": 2.018409490585327,
"step": 1118
},
{
"epoch": 0.112,
"grad_norm": 0.08470617979764938,
"learning_rate": 0.002,
"loss": 2.2633841037750244,
"step": 1120
},
{
"epoch": 0.1122,
"grad_norm": 0.08956170082092285,
"learning_rate": 0.002,
"loss": 1.9378591775894165,
"step": 1122
},
{
"epoch": 0.1124,
"grad_norm": 0.08724003285169601,
"learning_rate": 0.002,
"loss": 2.1270406246185303,
"step": 1124
},
{
"epoch": 0.1126,
"grad_norm": 0.08309338986873627,
"learning_rate": 0.002,
"loss": 2.024172067642212,
"step": 1126
},
{
"epoch": 0.1128,
"grad_norm": 0.08448407799005508,
"learning_rate": 0.002,
"loss": 2.034794807434082,
"step": 1128
},
{
"epoch": 0.113,
"grad_norm": 0.07977596670389175,
"learning_rate": 0.002,
"loss": 2.0115578174591064,
"step": 1130
},
{
"epoch": 0.1132,
"grad_norm": 0.08561068028211594,
"learning_rate": 0.002,
"loss": 2.022027015686035,
"step": 1132
},
{
"epoch": 0.1134,
"grad_norm": 0.08134664595127106,
"learning_rate": 0.002,
"loss": 2.1603844165802,
"step": 1134
},
{
"epoch": 0.1136,
"grad_norm": 0.0817975252866745,
"learning_rate": 0.002,
"loss": 2.0459768772125244,
"step": 1136
},
{
"epoch": 0.1138,
"grad_norm": 0.08369145542383194,
"learning_rate": 0.002,
"loss": 2.014500856399536,
"step": 1138
},
{
"epoch": 0.114,
"grad_norm": 0.08317965269088745,
"learning_rate": 0.002,
"loss": 2.0618274211883545,
"step": 1140
},
{
"epoch": 0.1142,
"grad_norm": 0.08403006196022034,
"learning_rate": 0.002,
"loss": 2.056919813156128,
"step": 1142
},
{
"epoch": 0.1144,
"grad_norm": 0.08624644577503204,
"learning_rate": 0.002,
"loss": 2.1261184215545654,
"step": 1144
},
{
"epoch": 0.1146,
"grad_norm": 0.09393399953842163,
"learning_rate": 0.002,
"loss": 1.9532134532928467,
"step": 1146
},
{
"epoch": 0.1148,
"grad_norm": 0.08186756819486618,
"learning_rate": 0.002,
"loss": 1.9701474905014038,
"step": 1148
},
{
"epoch": 0.115,
"grad_norm": 0.08925191313028336,
"learning_rate": 0.002,
"loss": 2.02519154548645,
"step": 1150
},
{
"epoch": 0.1152,
"grad_norm": 0.08585000783205032,
"learning_rate": 0.002,
"loss": 1.9593058824539185,
"step": 1152
},
{
"epoch": 0.1154,
"grad_norm": 0.08077649772167206,
"learning_rate": 0.002,
"loss": 1.9867432117462158,
"step": 1154
},
{
"epoch": 0.1156,
"grad_norm": 0.0841299295425415,
"learning_rate": 0.002,
"loss": 2.0073344707489014,
"step": 1156
},
{
"epoch": 0.1158,
"grad_norm": 0.08617623150348663,
"learning_rate": 0.002,
"loss": 2.055337905883789,
"step": 1158
},
{
"epoch": 0.116,
"grad_norm": 0.08616787195205688,
"learning_rate": 0.002,
"loss": 2.165081739425659,
"step": 1160
},
{
"epoch": 0.1162,
"grad_norm": 0.08450275659561157,
"learning_rate": 0.002,
"loss": 1.9545855522155762,
"step": 1162
},
{
"epoch": 0.1164,
"grad_norm": 0.09259680658578873,
"learning_rate": 0.002,
"loss": 2.0461883544921875,
"step": 1164
},
{
"epoch": 0.1166,
"grad_norm": 0.08266383409500122,
"learning_rate": 0.002,
"loss": 1.9070090055465698,
"step": 1166
},
{
"epoch": 0.1168,
"grad_norm": 0.08327124267816544,
"learning_rate": 0.002,
"loss": 1.9365736246109009,
"step": 1168
},
{
"epoch": 0.117,
"grad_norm": 0.0869339257478714,
"learning_rate": 0.002,
"loss": 2.068601131439209,
"step": 1170
},
{
"epoch": 0.1172,
"grad_norm": 0.08477666229009628,
"learning_rate": 0.002,
"loss": 2.068366527557373,
"step": 1172
},
{
"epoch": 0.1174,
"grad_norm": 0.08698020130395889,
"learning_rate": 0.002,
"loss": 2.003343105316162,
"step": 1174
},
{
"epoch": 0.1176,
"grad_norm": 0.08435402065515518,
"learning_rate": 0.002,
"loss": 1.9344985485076904,
"step": 1176
},
{
"epoch": 0.1178,
"grad_norm": 0.08306105434894562,
"learning_rate": 0.002,
"loss": 2.047485828399658,
"step": 1178
},
{
"epoch": 0.118,
"grad_norm": 0.08435844630002975,
"learning_rate": 0.002,
"loss": 1.8901934623718262,
"step": 1180
},
{
"epoch": 0.1182,
"grad_norm": 0.09012391418218613,
"learning_rate": 0.002,
"loss": 2.106781005859375,
"step": 1182
},
{
"epoch": 0.1184,
"grad_norm": 0.08325701206922531,
"learning_rate": 0.002,
"loss": 2.0709898471832275,
"step": 1184
},
{
"epoch": 0.1186,
"grad_norm": 0.08456530421972275,
"learning_rate": 0.002,
"loss": 2.1262314319610596,
"step": 1186
},
{
"epoch": 0.1188,
"grad_norm": 0.08413276821374893,
"learning_rate": 0.002,
"loss": 1.9871206283569336,
"step": 1188
},
{
"epoch": 0.119,
"grad_norm": 0.08718832582235336,
"learning_rate": 0.002,
"loss": 2.084408760070801,
"step": 1190
},
{
"epoch": 0.1192,
"grad_norm": 0.08340590447187424,
"learning_rate": 0.002,
"loss": 1.8883227109909058,
"step": 1192
},
{
"epoch": 0.1194,
"grad_norm": 0.07996539771556854,
"learning_rate": 0.002,
"loss": 2.039456367492676,
"step": 1194
},
{
"epoch": 0.1196,
"grad_norm": 0.08217329531908035,
"learning_rate": 0.002,
"loss": 2.024759531021118,
"step": 1196
},
{
"epoch": 0.1198,
"grad_norm": 0.08313747495412827,
"learning_rate": 0.002,
"loss": 2.103900194168091,
"step": 1198
},
{
"epoch": 0.12,
"grad_norm": 0.0881541520357132,
"learning_rate": 0.002,
"loss": 2.222304344177246,
"step": 1200
},
{
"epoch": 0.1202,
"grad_norm": 0.08509214967489243,
"learning_rate": 0.002,
"loss": 2.061307907104492,
"step": 1202
},
{
"epoch": 0.1204,
"grad_norm": 0.08412045985460281,
"learning_rate": 0.002,
"loss": 2.0424652099609375,
"step": 1204
},
{
"epoch": 0.1206,
"grad_norm": 0.08259782195091248,
"learning_rate": 0.002,
"loss": 1.8773162364959717,
"step": 1206
},
{
"epoch": 0.1208,
"grad_norm": 0.08027885109186172,
"learning_rate": 0.002,
"loss": 2.0441079139709473,
"step": 1208
},
{
"epoch": 0.121,
"grad_norm": 0.08164530247449875,
"learning_rate": 0.002,
"loss": 1.7741857767105103,
"step": 1210
},
{
"epoch": 0.1212,
"grad_norm": 0.07959095388650894,
"learning_rate": 0.002,
"loss": 1.986232876777649,
"step": 1212
},
{
"epoch": 0.1214,
"grad_norm": 0.0858709067106247,
"learning_rate": 0.002,
"loss": 2.1754837036132812,
"step": 1214
},
{
"epoch": 0.1216,
"grad_norm": 0.08982837945222855,
"learning_rate": 0.002,
"loss": 2.0495615005493164,
"step": 1216
},
{
"epoch": 0.1218,
"grad_norm": 0.08699710667133331,
"learning_rate": 0.002,
"loss": 2.140305995941162,
"step": 1218
},
{
"epoch": 0.122,
"grad_norm": 0.08187000453472137,
"learning_rate": 0.002,
"loss": 2.2325439453125,
"step": 1220
},
{
"epoch": 0.1222,
"grad_norm": 0.08354407548904419,
"learning_rate": 0.002,
"loss": 1.9623830318450928,
"step": 1222
},
{
"epoch": 0.1224,
"grad_norm": 0.0819743424654007,
"learning_rate": 0.002,
"loss": 1.944139003753662,
"step": 1224
},
{
"epoch": 0.1226,
"grad_norm": 0.0813278779387474,
"learning_rate": 0.002,
"loss": 2.115978479385376,
"step": 1226
},
{
"epoch": 0.1228,
"grad_norm": 0.07694555073976517,
"learning_rate": 0.002,
"loss": 2.039741277694702,
"step": 1228
},
{
"epoch": 0.123,
"grad_norm": 0.085311658680439,
"learning_rate": 0.002,
"loss": 2.0899102687835693,
"step": 1230
},
{
"epoch": 0.1232,
"grad_norm": 0.09176570177078247,
"learning_rate": 0.002,
"loss": 2.0086252689361572,
"step": 1232
},
{
"epoch": 0.1234,
"grad_norm": 0.09619509428739548,
"learning_rate": 0.002,
"loss": 2.032975196838379,
"step": 1234
},
{
"epoch": 0.1236,
"grad_norm": 0.08549220860004425,
"learning_rate": 0.002,
"loss": 2.0882456302642822,
"step": 1236
},
{
"epoch": 0.1238,
"grad_norm": 0.08163252472877502,
"learning_rate": 0.002,
"loss": 2.1956636905670166,
"step": 1238
},
{
"epoch": 0.124,
"grad_norm": 0.090960793197155,
"learning_rate": 0.002,
"loss": 1.989876627922058,
"step": 1240
},
{
"epoch": 0.1242,
"grad_norm": 0.08690346032381058,
"learning_rate": 0.002,
"loss": 1.9766404628753662,
"step": 1242
},
{
"epoch": 0.1244,
"grad_norm": 0.08295867592096329,
"learning_rate": 0.002,
"loss": 1.9767091274261475,
"step": 1244
},
{
"epoch": 0.1246,
"grad_norm": 0.09051886945962906,
"learning_rate": 0.002,
"loss": 1.9998339414596558,
"step": 1246
},
{
"epoch": 0.1248,
"grad_norm": 0.08345075696706772,
"learning_rate": 0.002,
"loss": 2.077249050140381,
"step": 1248
},
{
"epoch": 0.125,
"grad_norm": 0.08784466981887817,
"learning_rate": 0.002,
"loss": 1.9706625938415527,
"step": 1250
},
{
"epoch": 0.1252,
"grad_norm": 0.08191345632076263,
"learning_rate": 0.002,
"loss": 1.9752776622772217,
"step": 1252
},
{
"epoch": 0.1254,
"grad_norm": 0.08956123888492584,
"learning_rate": 0.002,
"loss": 2.182732105255127,
"step": 1254
},
{
"epoch": 0.1256,
"grad_norm": 0.08047401160001755,
"learning_rate": 0.002,
"loss": 1.9972858428955078,
"step": 1256
},
{
"epoch": 0.1258,
"grad_norm": 0.08208698779344559,
"learning_rate": 0.002,
"loss": 2.0774693489074707,
"step": 1258
},
{
"epoch": 0.126,
"grad_norm": 0.08476417511701584,
"learning_rate": 0.002,
"loss": 1.952588677406311,
"step": 1260
},
{
"epoch": 0.1262,
"grad_norm": 0.0856880471110344,
"learning_rate": 0.002,
"loss": 1.926142692565918,
"step": 1262
},
{
"epoch": 0.1264,
"grad_norm": 0.07638447731733322,
"learning_rate": 0.002,
"loss": 2.017592430114746,
"step": 1264
},
{
"epoch": 0.1266,
"grad_norm": 0.09431228041648865,
"learning_rate": 0.002,
"loss": 2.0062355995178223,
"step": 1266
},
{
"epoch": 0.1268,
"grad_norm": 0.08278736472129822,
"learning_rate": 0.002,
"loss": 2.0274674892425537,
"step": 1268
},
{
"epoch": 0.127,
"grad_norm": 0.07975926250219345,
"learning_rate": 0.002,
"loss": 2.0235726833343506,
"step": 1270
},
{
"epoch": 0.1272,
"grad_norm": 0.08381231129169464,
"learning_rate": 0.002,
"loss": 1.9192882776260376,
"step": 1272
},
{
"epoch": 0.1274,
"grad_norm": 0.08344686776399612,
"learning_rate": 0.002,
"loss": 2.1065878868103027,
"step": 1274
},
{
"epoch": 0.1276,
"grad_norm": 0.0879516452550888,
"learning_rate": 0.002,
"loss": 2.0292720794677734,
"step": 1276
},
{
"epoch": 0.1278,
"grad_norm": 0.07877160608768463,
"learning_rate": 0.002,
"loss": 1.9498127698898315,
"step": 1278
},
{
"epoch": 0.128,
"grad_norm": 0.08284289389848709,
"learning_rate": 0.002,
"loss": 1.961515188217163,
"step": 1280
},
{
"epoch": 0.1282,
"grad_norm": 0.09120652824640274,
"learning_rate": 0.002,
"loss": 1.9631229639053345,
"step": 1282
},
{
"epoch": 0.1284,
"grad_norm": 0.07842068374156952,
"learning_rate": 0.002,
"loss": 1.9782634973526,
"step": 1284
},
{
"epoch": 0.1286,
"grad_norm": 0.0847662091255188,
"learning_rate": 0.002,
"loss": 2.046393394470215,
"step": 1286
},
{
"epoch": 0.1288,
"grad_norm": 0.08472472429275513,
"learning_rate": 0.002,
"loss": 2.0080883502960205,
"step": 1288
},
{
"epoch": 0.129,
"grad_norm": 0.08054573088884354,
"learning_rate": 0.002,
"loss": 1.985576868057251,
"step": 1290
},
{
"epoch": 0.1292,
"grad_norm": 0.07891428470611572,
"learning_rate": 0.002,
"loss": 1.9087384939193726,
"step": 1292
},
{
"epoch": 0.1294,
"grad_norm": 0.08551960438489914,
"learning_rate": 0.002,
"loss": 2.0102670192718506,
"step": 1294
},
{
"epoch": 0.1296,
"grad_norm": 0.08134318888187408,
"learning_rate": 0.002,
"loss": 2.0879855155944824,
"step": 1296
},
{
"epoch": 0.1298,
"grad_norm": 0.08065108954906464,
"learning_rate": 0.002,
"loss": 2.0096890926361084,
"step": 1298
},
{
"epoch": 0.13,
"grad_norm": 0.08614367246627808,
"learning_rate": 0.002,
"loss": 1.9494773149490356,
"step": 1300
},
{
"epoch": 0.1302,
"grad_norm": 0.08957385271787643,
"learning_rate": 0.002,
"loss": 1.9771714210510254,
"step": 1302
},
{
"epoch": 0.1304,
"grad_norm": 0.07945297658443451,
"learning_rate": 0.002,
"loss": 1.8320209980010986,
"step": 1304
},
{
"epoch": 0.1306,
"grad_norm": 0.08779644221067429,
"learning_rate": 0.002,
"loss": 1.9819495677947998,
"step": 1306
},
{
"epoch": 0.1308,
"grad_norm": 0.08483578264713287,
"learning_rate": 0.002,
"loss": 2.0804686546325684,
"step": 1308
},
{
"epoch": 0.131,
"grad_norm": 0.08246662467718124,
"learning_rate": 0.002,
"loss": 1.9303864240646362,
"step": 1310
},
{
"epoch": 0.1312,
"grad_norm": 0.08496952056884766,
"learning_rate": 0.002,
"loss": 2.158707618713379,
"step": 1312
},
{
"epoch": 0.1314,
"grad_norm": 0.08043855428695679,
"learning_rate": 0.002,
"loss": 2.0037665367126465,
"step": 1314
},
{
"epoch": 0.1316,
"grad_norm": 0.08313608914613724,
"learning_rate": 0.002,
"loss": 2.0118744373321533,
"step": 1316
},
{
"epoch": 0.1318,
"grad_norm": 0.08287588506937027,
"learning_rate": 0.002,
"loss": 2.1368191242218018,
"step": 1318
},
{
"epoch": 0.132,
"grad_norm": 0.08581621944904327,
"learning_rate": 0.002,
"loss": 2.2238590717315674,
"step": 1320
},
{
"epoch": 0.1322,
"grad_norm": 0.08289754390716553,
"learning_rate": 0.002,
"loss": 1.9536819458007812,
"step": 1322
},
{
"epoch": 0.1324,
"grad_norm": 0.08500684797763824,
"learning_rate": 0.002,
"loss": 1.9097071886062622,
"step": 1324
},
{
"epoch": 0.1326,
"grad_norm": 0.08194470405578613,
"learning_rate": 0.002,
"loss": 2.009051561355591,
"step": 1326
},
{
"epoch": 0.1328,
"grad_norm": 0.08581238985061646,
"learning_rate": 0.002,
"loss": 2.0748465061187744,
"step": 1328
},
{
"epoch": 0.133,
"grad_norm": 0.08034423738718033,
"learning_rate": 0.002,
"loss": 2.1352503299713135,
"step": 1330
},
{
"epoch": 0.1332,
"grad_norm": 0.08761461079120636,
"learning_rate": 0.002,
"loss": 1.940513253211975,
"step": 1332
},
{
"epoch": 0.1334,
"grad_norm": 0.09479259699583054,
"learning_rate": 0.002,
"loss": 2.0859739780426025,
"step": 1334
},
{
"epoch": 0.1336,
"grad_norm": 0.08276405930519104,
"learning_rate": 0.002,
"loss": 1.9383325576782227,
"step": 1336
},
{
"epoch": 0.1338,
"grad_norm": 0.08897998183965683,
"learning_rate": 0.002,
"loss": 1.9012813568115234,
"step": 1338
},
{
"epoch": 0.134,
"grad_norm": 0.08546336740255356,
"learning_rate": 0.002,
"loss": 2.0785326957702637,
"step": 1340
},
{
"epoch": 0.1342,
"grad_norm": 0.07834289968013763,
"learning_rate": 0.002,
"loss": 1.9747636318206787,
"step": 1342
},
{
"epoch": 0.1344,
"grad_norm": 0.08041142672300339,
"learning_rate": 0.002,
"loss": 1.8900039196014404,
"step": 1344
},
{
"epoch": 0.1346,
"grad_norm": 0.08464096486568451,
"learning_rate": 0.002,
"loss": 1.9109103679656982,
"step": 1346
},
{
"epoch": 0.1348,
"grad_norm": 0.08302085101604462,
"learning_rate": 0.002,
"loss": 2.083608865737915,
"step": 1348
},
{
"epoch": 0.135,
"grad_norm": 0.08918166160583496,
"learning_rate": 0.002,
"loss": 2.1519148349761963,
"step": 1350
},
{
"epoch": 0.1352,
"grad_norm": 0.08706078678369522,
"learning_rate": 0.002,
"loss": 1.8809285163879395,
"step": 1352
},
{
"epoch": 0.1354,
"grad_norm": 0.08329781889915466,
"learning_rate": 0.002,
"loss": 1.9737176895141602,
"step": 1354
},
{
"epoch": 0.1356,
"grad_norm": 0.081386499106884,
"learning_rate": 0.002,
"loss": 2.104400873184204,
"step": 1356
},
{
"epoch": 0.1358,
"grad_norm": 0.08352888375520706,
"learning_rate": 0.002,
"loss": 2.0278191566467285,
"step": 1358
},
{
"epoch": 0.136,
"grad_norm": 0.09237174689769745,
"learning_rate": 0.002,
"loss": 2.112232208251953,
"step": 1360
},
{
"epoch": 0.1362,
"grad_norm": 0.08363782614469528,
"learning_rate": 0.002,
"loss": 1.9792081117630005,
"step": 1362
},
{
"epoch": 0.1364,
"grad_norm": 0.08287493884563446,
"learning_rate": 0.002,
"loss": 2.1652090549468994,
"step": 1364
},
{
"epoch": 0.1366,
"grad_norm": 0.08788807690143585,
"learning_rate": 0.002,
"loss": 2.1167266368865967,
"step": 1366
},
{
"epoch": 0.1368,
"grad_norm": 0.08260025829076767,
"learning_rate": 0.002,
"loss": 1.980812430381775,
"step": 1368
},
{
"epoch": 0.137,
"grad_norm": 0.08492080122232437,
"learning_rate": 0.002,
"loss": 2.1833114624023438,
"step": 1370
},
{
"epoch": 0.1372,
"grad_norm": 0.09278244525194168,
"learning_rate": 0.002,
"loss": 2.0212106704711914,
"step": 1372
},
{
"epoch": 0.1374,
"grad_norm": 0.08107468485832214,
"learning_rate": 0.002,
"loss": 2.0436851978302,
"step": 1374
},
{
"epoch": 0.1376,
"grad_norm": 0.0862237960100174,
"learning_rate": 0.002,
"loss": 2.1253888607025146,
"step": 1376
},
{
"epoch": 0.1378,
"grad_norm": 0.08353643864393234,
"learning_rate": 0.002,
"loss": 2.0461161136627197,
"step": 1378
},
{
"epoch": 0.138,
"grad_norm": 0.08285964280366898,
"learning_rate": 0.002,
"loss": 1.9251466989517212,
"step": 1380
},
{
"epoch": 0.1382,
"grad_norm": 0.0818663164973259,
"learning_rate": 0.002,
"loss": 2.0516092777252197,
"step": 1382
},
{
"epoch": 0.1384,
"grad_norm": 0.08050210773944855,
"learning_rate": 0.002,
"loss": 1.8944146633148193,
"step": 1384
},
{
"epoch": 0.1386,
"grad_norm": 0.08588985353708267,
"learning_rate": 0.002,
"loss": 2.0791821479797363,
"step": 1386
},
{
"epoch": 0.1388,
"grad_norm": 0.08128977566957474,
"learning_rate": 0.002,
"loss": 2.0822651386260986,
"step": 1388
},
{
"epoch": 0.139,
"grad_norm": 0.08936209976673126,
"learning_rate": 0.002,
"loss": 2.119364023208618,
"step": 1390
},
{
"epoch": 0.1392,
"grad_norm": 0.082890585064888,
"learning_rate": 0.002,
"loss": 1.938192367553711,
"step": 1392
},
{
"epoch": 0.1394,
"grad_norm": 0.07707413285970688,
"learning_rate": 0.002,
"loss": 2.1069889068603516,
"step": 1394
},
{
"epoch": 0.1396,
"grad_norm": 0.09440362453460693,
"learning_rate": 0.002,
"loss": 2.0704145431518555,
"step": 1396
},
{
"epoch": 0.1398,
"grad_norm": 0.08753009885549545,
"learning_rate": 0.002,
"loss": 1.9484301805496216,
"step": 1398
},
{
"epoch": 0.14,
"grad_norm": 0.08215557783842087,
"learning_rate": 0.002,
"loss": 2.124666452407837,
"step": 1400
},
{
"epoch": 0.1402,
"grad_norm": 0.0817045196890831,
"learning_rate": 0.002,
"loss": 1.984005093574524,
"step": 1402
},
{
"epoch": 0.1404,
"grad_norm": 0.08075842261314392,
"learning_rate": 0.002,
"loss": 2.00469708442688,
"step": 1404
},
{
"epoch": 0.1406,
"grad_norm": 0.08940353244543076,
"learning_rate": 0.002,
"loss": 2.082493543624878,
"step": 1406
},
{
"epoch": 0.1408,
"grad_norm": 0.08476462215185165,
"learning_rate": 0.002,
"loss": 1.9702262878417969,
"step": 1408
},
{
"epoch": 0.141,
"grad_norm": 0.08214747905731201,
"learning_rate": 0.002,
"loss": 1.8237375020980835,
"step": 1410
},
{
"epoch": 0.1412,
"grad_norm": 0.08178984373807907,
"learning_rate": 0.002,
"loss": 2.025057077407837,
"step": 1412
},
{
"epoch": 0.1414,
"grad_norm": 0.08408454805612564,
"learning_rate": 0.002,
"loss": 1.9185508489608765,
"step": 1414
},
{
"epoch": 0.1416,
"grad_norm": 0.0835493877530098,
"learning_rate": 0.002,
"loss": 1.8612854480743408,
"step": 1416
},
{
"epoch": 0.1418,
"grad_norm": 0.08780626207590103,
"learning_rate": 0.002,
"loss": 1.9259387254714966,
"step": 1418
},
{
"epoch": 0.142,
"grad_norm": 0.08390364050865173,
"learning_rate": 0.002,
"loss": 1.9215179681777954,
"step": 1420
},
{
"epoch": 0.1422,
"grad_norm": 0.07897157967090607,
"learning_rate": 0.002,
"loss": 1.8806076049804688,
"step": 1422
},
{
"epoch": 0.1424,
"grad_norm": 0.08310697227716446,
"learning_rate": 0.002,
"loss": 2.0307650566101074,
"step": 1424
},
{
"epoch": 0.1426,
"grad_norm": 0.08411696553230286,
"learning_rate": 0.002,
"loss": 1.9524784088134766,
"step": 1426
},
{
"epoch": 0.1428,
"grad_norm": 0.09483884274959564,
"learning_rate": 0.002,
"loss": 2.1354057788848877,
"step": 1428
},
{
"epoch": 0.143,
"grad_norm": 0.08339402079582214,
"learning_rate": 0.002,
"loss": 2.0050485134124756,
"step": 1430
},
{
"epoch": 0.1432,
"grad_norm": 0.08105912059545517,
"learning_rate": 0.002,
"loss": 2.0065789222717285,
"step": 1432
},
{
"epoch": 0.1434,
"grad_norm": 0.08590013533830643,
"learning_rate": 0.002,
"loss": 2.036484479904175,
"step": 1434
},
{
"epoch": 0.1436,
"grad_norm": 0.08551734685897827,
"learning_rate": 0.002,
"loss": 2.053032875061035,
"step": 1436
},
{
"epoch": 0.1438,
"grad_norm": 0.08437639474868774,
"learning_rate": 0.002,
"loss": 1.9124197959899902,
"step": 1438
},
{
"epoch": 0.144,
"grad_norm": 0.08546862751245499,
"learning_rate": 0.002,
"loss": 2.0219709873199463,
"step": 1440
},
{
"epoch": 0.1442,
"grad_norm": 0.09397388249635696,
"learning_rate": 0.002,
"loss": 2.005190849304199,
"step": 1442
},
{
"epoch": 0.1444,
"grad_norm": 0.08086714148521423,
"learning_rate": 0.002,
"loss": 1.977594256401062,
"step": 1444
},
{
"epoch": 0.1446,
"grad_norm": 0.08263556659221649,
"learning_rate": 0.002,
"loss": 2.159130573272705,
"step": 1446
},
{
"epoch": 0.1448,
"grad_norm": 0.08463070541620255,
"learning_rate": 0.002,
"loss": 2.046889305114746,
"step": 1448
},
{
"epoch": 0.145,
"grad_norm": 0.08332529664039612,
"learning_rate": 0.002,
"loss": 1.8495128154754639,
"step": 1450
},
{
"epoch": 0.1452,
"grad_norm": 0.08365868777036667,
"learning_rate": 0.002,
"loss": 1.9837555885314941,
"step": 1452
},
{
"epoch": 0.1454,
"grad_norm": 0.0867474228143692,
"learning_rate": 0.002,
"loss": 2.0188663005828857,
"step": 1454
},
{
"epoch": 0.1456,
"grad_norm": 0.09054825454950333,
"learning_rate": 0.002,
"loss": 1.9647122621536255,
"step": 1456
},
{
"epoch": 0.1458,
"grad_norm": 0.08483801037073135,
"learning_rate": 0.002,
"loss": 1.855425477027893,
"step": 1458
},
{
"epoch": 0.146,
"grad_norm": 0.09346556663513184,
"learning_rate": 0.002,
"loss": 2.191694736480713,
"step": 1460
},
{
"epoch": 0.1462,
"grad_norm": 0.08465056121349335,
"learning_rate": 0.002,
"loss": 2.0250799655914307,
"step": 1462
},
{
"epoch": 0.1464,
"grad_norm": 0.08501332998275757,
"learning_rate": 0.002,
"loss": 2.0752594470977783,
"step": 1464
},
{
"epoch": 0.1466,
"grad_norm": 0.08634602278470993,
"learning_rate": 0.002,
"loss": 2.071413993835449,
"step": 1466
},
{
"epoch": 0.1468,
"grad_norm": 0.08813812583684921,
"learning_rate": 0.002,
"loss": 2.019972324371338,
"step": 1468
},
{
"epoch": 0.147,
"grad_norm": 0.08231979608535767,
"learning_rate": 0.002,
"loss": 2.0380661487579346,
"step": 1470
},
{
"epoch": 0.1472,
"grad_norm": 0.08754221349954605,
"learning_rate": 0.002,
"loss": 1.8944451808929443,
"step": 1472
},
{
"epoch": 0.1474,
"grad_norm": 0.081061452627182,
"learning_rate": 0.002,
"loss": 2.084580898284912,
"step": 1474
},
{
"epoch": 0.1476,
"grad_norm": 0.08811423182487488,
"learning_rate": 0.002,
"loss": 2.010695695877075,
"step": 1476
},
{
"epoch": 0.1478,
"grad_norm": 0.0883941501379013,
"learning_rate": 0.002,
"loss": 2.0035157203674316,
"step": 1478
},
{
"epoch": 0.148,
"grad_norm": 0.08566522598266602,
"learning_rate": 0.002,
"loss": 2.150222063064575,
"step": 1480
},
{
"epoch": 0.1482,
"grad_norm": 0.0847812071442604,
"learning_rate": 0.002,
"loss": 2.1005401611328125,
"step": 1482
},
{
"epoch": 0.1484,
"grad_norm": 0.0855879932641983,
"learning_rate": 0.002,
"loss": 1.962319254875183,
"step": 1484
},
{
"epoch": 0.1486,
"grad_norm": 0.08677064627408981,
"learning_rate": 0.002,
"loss": 1.96921968460083,
"step": 1486
},
{
"epoch": 0.1488,
"grad_norm": 0.08283451199531555,
"learning_rate": 0.002,
"loss": 1.9546277523040771,
"step": 1488
},
{
"epoch": 0.149,
"grad_norm": 0.08457811921834946,
"learning_rate": 0.002,
"loss": 2.0715556144714355,
"step": 1490
},
{
"epoch": 0.1492,
"grad_norm": 0.08524119108915329,
"learning_rate": 0.002,
"loss": 1.8723537921905518,
"step": 1492
},
{
"epoch": 0.1494,
"grad_norm": 0.08556007593870163,
"learning_rate": 0.002,
"loss": 1.972347617149353,
"step": 1494
},
{
"epoch": 0.1496,
"grad_norm": 0.08538345992565155,
"learning_rate": 0.002,
"loss": 2.0747456550598145,
"step": 1496
},
{
"epoch": 0.1498,
"grad_norm": 0.08342566341161728,
"learning_rate": 0.002,
"loss": 1.9905140399932861,
"step": 1498
},
{
"epoch": 0.15,
"grad_norm": 0.08465108275413513,
"learning_rate": 0.002,
"loss": 2.118013620376587,
"step": 1500
},
{
"epoch": 0.1502,
"grad_norm": 0.09234253317117691,
"learning_rate": 0.002,
"loss": 1.91473388671875,
"step": 1502
},
{
"epoch": 0.1504,
"grad_norm": 0.08265455812215805,
"learning_rate": 0.002,
"loss": 2.0202364921569824,
"step": 1504
},
{
"epoch": 0.1506,
"grad_norm": 0.08589425683021545,
"learning_rate": 0.002,
"loss": 2.0076608657836914,
"step": 1506
},
{
"epoch": 0.1508,
"grad_norm": 0.08468682318925858,
"learning_rate": 0.002,
"loss": 2.0685858726501465,
"step": 1508
},
{
"epoch": 0.151,
"grad_norm": 0.08672576397657394,
"learning_rate": 0.002,
"loss": 2.0356192588806152,
"step": 1510
},
{
"epoch": 0.1512,
"grad_norm": 0.08560921996831894,
"learning_rate": 0.002,
"loss": 1.9504189491271973,
"step": 1512
},
{
"epoch": 0.1514,
"grad_norm": 0.08880069106817245,
"learning_rate": 0.002,
"loss": 1.9090538024902344,
"step": 1514
},
{
"epoch": 0.1516,
"grad_norm": 0.08352338522672653,
"learning_rate": 0.002,
"loss": 1.9770448207855225,
"step": 1516
},
{
"epoch": 0.1518,
"grad_norm": 0.0834403857588768,
"learning_rate": 0.002,
"loss": 2.032268762588501,
"step": 1518
},
{
"epoch": 0.152,
"grad_norm": 0.08182430267333984,
"learning_rate": 0.002,
"loss": 1.7237893342971802,
"step": 1520
},
{
"epoch": 0.1522,
"grad_norm": 0.0823986604809761,
"learning_rate": 0.002,
"loss": 2.0870168209075928,
"step": 1522
},
{
"epoch": 0.1524,
"grad_norm": 0.08288682997226715,
"learning_rate": 0.002,
"loss": 2.04171085357666,
"step": 1524
},
{
"epoch": 0.1526,
"grad_norm": 0.09022911638021469,
"learning_rate": 0.002,
"loss": 2.058595895767212,
"step": 1526
},
{
"epoch": 0.1528,
"grad_norm": 0.08884043991565704,
"learning_rate": 0.002,
"loss": 2.112304449081421,
"step": 1528
},
{
"epoch": 0.153,
"grad_norm": 0.08846046030521393,
"learning_rate": 0.002,
"loss": 2.1000938415527344,
"step": 1530
},
{
"epoch": 0.1532,
"grad_norm": 0.08225948363542557,
"learning_rate": 0.002,
"loss": 1.9290692806243896,
"step": 1532
},
{
"epoch": 0.1534,
"grad_norm": 0.08358252048492432,
"learning_rate": 0.002,
"loss": 1.880184531211853,
"step": 1534
},
{
"epoch": 0.1536,
"grad_norm": 0.08781662583351135,
"learning_rate": 0.002,
"loss": 2.03822660446167,
"step": 1536
},
{
"epoch": 0.1538,
"grad_norm": 0.0874786451458931,
"learning_rate": 0.002,
"loss": 2.003692626953125,
"step": 1538
},
{
"epoch": 0.154,
"grad_norm": 0.08314626663923264,
"learning_rate": 0.002,
"loss": 2.103372812271118,
"step": 1540
},
{
"epoch": 0.1542,
"grad_norm": 0.08071058243513107,
"learning_rate": 0.002,
"loss": 1.988526463508606,
"step": 1542
},
{
"epoch": 0.1544,
"grad_norm": 0.08185788989067078,
"learning_rate": 0.002,
"loss": 1.8805160522460938,
"step": 1544
},
{
"epoch": 0.1546,
"grad_norm": 0.08340204507112503,
"learning_rate": 0.002,
"loss": 2.0312440395355225,
"step": 1546
},
{
"epoch": 0.1548,
"grad_norm": 0.08513487875461578,
"learning_rate": 0.002,
"loss": 1.9365514516830444,
"step": 1548
},
{
"epoch": 0.155,
"grad_norm": 0.0839838907122612,
"learning_rate": 0.002,
"loss": 1.9162694215774536,
"step": 1550
},
{
"epoch": 0.1552,
"grad_norm": 0.08493292331695557,
"learning_rate": 0.002,
"loss": 1.9749780893325806,
"step": 1552
},
{
"epoch": 0.1554,
"grad_norm": 0.08738046139478683,
"learning_rate": 0.002,
"loss": 1.8714990615844727,
"step": 1554
},
{
"epoch": 0.1556,
"grad_norm": 0.08110321313142776,
"learning_rate": 0.002,
"loss": 1.947487235069275,
"step": 1556
},
{
"epoch": 0.1558,
"grad_norm": 0.08131778240203857,
"learning_rate": 0.002,
"loss": 1.9373154640197754,
"step": 1558
},
{
"epoch": 0.156,
"grad_norm": 0.0785713717341423,
"learning_rate": 0.002,
"loss": 2.0550003051757812,
"step": 1560
},
{
"epoch": 0.1562,
"grad_norm": 0.08419571071863174,
"learning_rate": 0.002,
"loss": 1.9248937368392944,
"step": 1562
},
{
"epoch": 0.1564,
"grad_norm": 0.08059526234865189,
"learning_rate": 0.002,
"loss": 1.9841457605361938,
"step": 1564
},
{
"epoch": 0.1566,
"grad_norm": 0.08866710215806961,
"learning_rate": 0.002,
"loss": 1.9777973890304565,
"step": 1566
},
{
"epoch": 0.1568,
"grad_norm": 0.08271918445825577,
"learning_rate": 0.002,
"loss": 2.074117422103882,
"step": 1568
},
{
"epoch": 0.157,
"grad_norm": 0.08722631633281708,
"learning_rate": 0.002,
"loss": 2.1816694736480713,
"step": 1570
},
{
"epoch": 0.1572,
"grad_norm": 0.08447971940040588,
"learning_rate": 0.002,
"loss": 2.109995126724243,
"step": 1572
},
{
"epoch": 0.1574,
"grad_norm": 0.08193506300449371,
"learning_rate": 0.002,
"loss": 1.8912419080734253,
"step": 1574
},
{
"epoch": 0.1576,
"grad_norm": 0.0820985659956932,
"learning_rate": 0.002,
"loss": 1.9356868267059326,
"step": 1576
},
{
"epoch": 0.1578,
"grad_norm": 0.08184309303760529,
"learning_rate": 0.002,
"loss": 1.9752860069274902,
"step": 1578
},
{
"epoch": 0.158,
"grad_norm": 0.08817871659994125,
"learning_rate": 0.002,
"loss": 2.0186655521392822,
"step": 1580
},
{
"epoch": 0.1582,
"grad_norm": 0.08116264641284943,
"learning_rate": 0.002,
"loss": 2.0638322830200195,
"step": 1582
},
{
"epoch": 0.1584,
"grad_norm": 0.10066129267215729,
"learning_rate": 0.002,
"loss": 1.9643419981002808,
"step": 1584
},
{
"epoch": 0.1586,
"grad_norm": 0.07976362854242325,
"learning_rate": 0.002,
"loss": 1.9309545755386353,
"step": 1586
},
{
"epoch": 0.1588,
"grad_norm": 0.08252105861902237,
"learning_rate": 0.002,
"loss": 1.942662239074707,
"step": 1588
},
{
"epoch": 0.159,
"grad_norm": 0.08732359856367111,
"learning_rate": 0.002,
"loss": 1.936599850654602,
"step": 1590
},
{
"epoch": 0.1592,
"grad_norm": 0.08256518840789795,
"learning_rate": 0.002,
"loss": 2.0580368041992188,
"step": 1592
},
{
"epoch": 0.1594,
"grad_norm": 0.0870324894785881,
"learning_rate": 0.002,
"loss": 2.0031468868255615,
"step": 1594
},
{
"epoch": 0.1596,
"grad_norm": 0.08515676110982895,
"learning_rate": 0.002,
"loss": 2.0998167991638184,
"step": 1596
},
{
"epoch": 0.1598,
"grad_norm": 0.10461036115884781,
"learning_rate": 0.002,
"loss": 1.9901013374328613,
"step": 1598
},
{
"epoch": 0.16,
"grad_norm": 0.08584758639335632,
"learning_rate": 0.002,
"loss": 1.888537049293518,
"step": 1600
},
{
"epoch": 0.1602,
"grad_norm": 0.08241167664527893,
"learning_rate": 0.002,
"loss": 2.0399234294891357,
"step": 1602
},
{
"epoch": 0.1604,
"grad_norm": 0.08698111772537231,
"learning_rate": 0.002,
"loss": 1.869358777999878,
"step": 1604
},
{
"epoch": 0.1606,
"grad_norm": 0.08399704098701477,
"learning_rate": 0.002,
"loss": 1.9496454000473022,
"step": 1606
},
{
"epoch": 0.1608,
"grad_norm": 0.08519057184457779,
"learning_rate": 0.002,
"loss": 2.175718307495117,
"step": 1608
},
{
"epoch": 0.161,
"grad_norm": 0.07982669025659561,
"learning_rate": 0.002,
"loss": 2.018760919570923,
"step": 1610
},
{
"epoch": 0.1612,
"grad_norm": 0.08624474704265594,
"learning_rate": 0.002,
"loss": 1.9684149026870728,
"step": 1612
},
{
"epoch": 0.1614,
"grad_norm": 0.08299598097801208,
"learning_rate": 0.002,
"loss": 1.9885144233703613,
"step": 1614
},
{
"epoch": 0.1616,
"grad_norm": 0.08168502897024155,
"learning_rate": 0.002,
"loss": 1.9772814512252808,
"step": 1616
},
{
"epoch": 0.1618,
"grad_norm": 0.09290925413370132,
"learning_rate": 0.002,
"loss": 2.070359468460083,
"step": 1618
},
{
"epoch": 0.162,
"grad_norm": 0.0804649293422699,
"learning_rate": 0.002,
"loss": 1.988921046257019,
"step": 1620
},
{
"epoch": 0.1622,
"grad_norm": 0.08479253202676773,
"learning_rate": 0.002,
"loss": 1.9732486009597778,
"step": 1622
},
{
"epoch": 0.1624,
"grad_norm": 0.08966785669326782,
"learning_rate": 0.002,
"loss": 1.9530940055847168,
"step": 1624
},
{
"epoch": 0.1626,
"grad_norm": 0.08252689987421036,
"learning_rate": 0.002,
"loss": 2.0290355682373047,
"step": 1626
},
{
"epoch": 0.1628,
"grad_norm": 0.08560115844011307,
"learning_rate": 0.002,
"loss": 1.955880045890808,
"step": 1628
},
{
"epoch": 0.163,
"grad_norm": 0.08271026611328125,
"learning_rate": 0.002,
"loss": 2.0511937141418457,
"step": 1630
},
{
"epoch": 0.1632,
"grad_norm": 0.08341914415359497,
"learning_rate": 0.002,
"loss": 2.062037706375122,
"step": 1632
},
{
"epoch": 0.1634,
"grad_norm": 0.08394568413496017,
"learning_rate": 0.002,
"loss": 1.9848227500915527,
"step": 1634
},
{
"epoch": 0.1636,
"grad_norm": 0.08560135960578918,
"learning_rate": 0.002,
"loss": 2.127549409866333,
"step": 1636
},
{
"epoch": 0.1638,
"grad_norm": 0.08927592635154724,
"learning_rate": 0.002,
"loss": 1.9887274503707886,
"step": 1638
},
{
"epoch": 0.164,
"grad_norm": 0.08311126381158829,
"learning_rate": 0.002,
"loss": 2.021876335144043,
"step": 1640
},
{
"epoch": 0.1642,
"grad_norm": 0.08075287938117981,
"learning_rate": 0.002,
"loss": 2.0108084678649902,
"step": 1642
},
{
"epoch": 0.1644,
"grad_norm": 0.08535883575677872,
"learning_rate": 0.002,
"loss": 2.025994300842285,
"step": 1644
},
{
"epoch": 0.1646,
"grad_norm": 0.08394721895456314,
"learning_rate": 0.002,
"loss": 1.9767261743545532,
"step": 1646
},
{
"epoch": 0.1648,
"grad_norm": 0.08777830749750137,
"learning_rate": 0.002,
"loss": 1.9552806615829468,
"step": 1648
},
{
"epoch": 0.165,
"grad_norm": 0.08747299760580063,
"learning_rate": 0.002,
"loss": 2.0201032161712646,
"step": 1650
},
{
"epoch": 0.1652,
"grad_norm": 0.08579709380865097,
"learning_rate": 0.002,
"loss": 1.9406492710113525,
"step": 1652
},
{
"epoch": 0.1654,
"grad_norm": 0.08180826902389526,
"learning_rate": 0.002,
"loss": 2.0655133724212646,
"step": 1654
},
{
"epoch": 0.1656,
"grad_norm": 0.08200948685407639,
"learning_rate": 0.002,
"loss": 1.9462840557098389,
"step": 1656
},
{
"epoch": 0.1658,
"grad_norm": 0.07699384540319443,
"learning_rate": 0.002,
"loss": 1.9783786535263062,
"step": 1658
},
{
"epoch": 0.166,
"grad_norm": 0.08740390837192535,
"learning_rate": 0.002,
"loss": 2.0780999660491943,
"step": 1660
},
{
"epoch": 0.1662,
"grad_norm": 0.08262025564908981,
"learning_rate": 0.002,
"loss": 1.9985356330871582,
"step": 1662
},
{
"epoch": 0.1664,
"grad_norm": 0.08615583926439285,
"learning_rate": 0.002,
"loss": 2.0267257690429688,
"step": 1664
},
{
"epoch": 0.1666,
"grad_norm": 0.08822714537382126,
"learning_rate": 0.002,
"loss": 2.11911678314209,
"step": 1666
},
{
"epoch": 0.1668,
"grad_norm": 0.09044399857521057,
"learning_rate": 0.002,
"loss": 1.91670823097229,
"step": 1668
},
{
"epoch": 0.167,
"grad_norm": 0.08889217674732208,
"learning_rate": 0.002,
"loss": 1.8854286670684814,
"step": 1670
},
{
"epoch": 0.1672,
"grad_norm": 0.08416656404733658,
"learning_rate": 0.002,
"loss": 1.9102697372436523,
"step": 1672
},
{
"epoch": 0.1674,
"grad_norm": 0.0856962502002716,
"learning_rate": 0.002,
"loss": 1.95755934715271,
"step": 1674
},
{
"epoch": 0.1676,
"grad_norm": 0.08149226754903793,
"learning_rate": 0.002,
"loss": 1.8925155401229858,
"step": 1676
},
{
"epoch": 0.1678,
"grad_norm": 0.08411500602960587,
"learning_rate": 0.002,
"loss": 2.0812156200408936,
"step": 1678
},
{
"epoch": 0.168,
"grad_norm": 0.09518945217132568,
"learning_rate": 0.002,
"loss": 1.850506067276001,
"step": 1680
},
{
"epoch": 0.1682,
"grad_norm": 0.08547048270702362,
"learning_rate": 0.002,
"loss": 2.069958209991455,
"step": 1682
},
{
"epoch": 0.1684,
"grad_norm": 0.08539139479398727,
"learning_rate": 0.002,
"loss": 2.0619094371795654,
"step": 1684
},
{
"epoch": 0.1686,
"grad_norm": 0.0835799127817154,
"learning_rate": 0.002,
"loss": 1.986047387123108,
"step": 1686
},
{
"epoch": 0.1688,
"grad_norm": 0.08559209108352661,
"learning_rate": 0.002,
"loss": 2.079071283340454,
"step": 1688
},
{
"epoch": 0.169,
"grad_norm": 0.08535871654748917,
"learning_rate": 0.002,
"loss": 1.9463937282562256,
"step": 1690
},
{
"epoch": 0.1692,
"grad_norm": 0.08502212911844254,
"learning_rate": 0.002,
"loss": 1.9606218338012695,
"step": 1692
},
{
"epoch": 0.1694,
"grad_norm": 0.0909137949347496,
"learning_rate": 0.002,
"loss": 1.8967794179916382,
"step": 1694
},
{
"epoch": 0.1696,
"grad_norm": 0.08144690096378326,
"learning_rate": 0.002,
"loss": 1.8653541803359985,
"step": 1696
},
{
"epoch": 0.1698,
"grad_norm": 0.08554151654243469,
"learning_rate": 0.002,
"loss": 1.82559072971344,
"step": 1698
},
{
"epoch": 0.17,
"grad_norm": 0.08355498313903809,
"learning_rate": 0.002,
"loss": 1.9472588300704956,
"step": 1700
},
{
"epoch": 0.1702,
"grad_norm": 0.08412764966487885,
"learning_rate": 0.002,
"loss": 2.054037094116211,
"step": 1702
},
{
"epoch": 0.1704,
"grad_norm": 0.08119752258062363,
"learning_rate": 0.002,
"loss": 1.98397696018219,
"step": 1704
},
{
"epoch": 0.1706,
"grad_norm": 0.08186545222997665,
"learning_rate": 0.002,
"loss": 1.811954140663147,
"step": 1706
},
{
"epoch": 0.1708,
"grad_norm": 0.0800834372639656,
"learning_rate": 0.002,
"loss": 2.0531461238861084,
"step": 1708
},
{
"epoch": 0.171,
"grad_norm": 0.0843147560954094,
"learning_rate": 0.002,
"loss": 2.086811065673828,
"step": 1710
},
{
"epoch": 0.1712,
"grad_norm": 0.08326831459999084,
"learning_rate": 0.002,
"loss": 1.9017562866210938,
"step": 1712
},
{
"epoch": 0.1714,
"grad_norm": 0.0864068940281868,
"learning_rate": 0.002,
"loss": 2.0999181270599365,
"step": 1714
},
{
"epoch": 0.1716,
"grad_norm": 0.08121897280216217,
"learning_rate": 0.002,
"loss": 1.8443279266357422,
"step": 1716
},
{
"epoch": 0.1718,
"grad_norm": 0.09560268372297287,
"learning_rate": 0.002,
"loss": 2.0472049713134766,
"step": 1718
},
{
"epoch": 0.172,
"grad_norm": 0.08162263035774231,
"learning_rate": 0.002,
"loss": 1.9566751718521118,
"step": 1720
},
{
"epoch": 0.1722,
"grad_norm": 0.08066337555646896,
"learning_rate": 0.002,
"loss": 2.077460289001465,
"step": 1722
},
{
"epoch": 0.1724,
"grad_norm": 0.08657451719045639,
"learning_rate": 0.002,
"loss": 2.007951498031616,
"step": 1724
},
{
"epoch": 0.1726,
"grad_norm": 0.08407124876976013,
"learning_rate": 0.002,
"loss": 1.9007164239883423,
"step": 1726
},
{
"epoch": 0.1728,
"grad_norm": 0.0846768319606781,
"learning_rate": 0.002,
"loss": 1.9592881202697754,
"step": 1728
},
{
"epoch": 0.173,
"grad_norm": 0.0836617574095726,
"learning_rate": 0.002,
"loss": 1.9015003442764282,
"step": 1730
},
{
"epoch": 0.1732,
"grad_norm": 0.08796169608831406,
"learning_rate": 0.002,
"loss": 2.019059658050537,
"step": 1732
},
{
"epoch": 0.1734,
"grad_norm": 0.08226469159126282,
"learning_rate": 0.002,
"loss": 1.9710384607315063,
"step": 1734
},
{
"epoch": 0.1736,
"grad_norm": 0.08608261495828629,
"learning_rate": 0.002,
"loss": 2.0204579830169678,
"step": 1736
},
{
"epoch": 0.1738,
"grad_norm": 0.09337099641561508,
"learning_rate": 0.002,
"loss": 1.914751648902893,
"step": 1738
},
{
"epoch": 0.174,
"grad_norm": 0.08730588853359222,
"learning_rate": 0.002,
"loss": 2.0349481105804443,
"step": 1740
},
{
"epoch": 0.1742,
"grad_norm": 0.08293639868497849,
"learning_rate": 0.002,
"loss": 1.9420349597930908,
"step": 1742
},
{
"epoch": 0.1744,
"grad_norm": 0.08442080765962601,
"learning_rate": 0.002,
"loss": 1.8389413356781006,
"step": 1744
},
{
"epoch": 0.1746,
"grad_norm": 0.0882704108953476,
"learning_rate": 0.002,
"loss": 2.055248498916626,
"step": 1746
},
{
"epoch": 0.1748,
"grad_norm": 0.0814896747469902,
"learning_rate": 0.002,
"loss": 1.9105103015899658,
"step": 1748
},
{
"epoch": 0.175,
"grad_norm": 0.09156211465597153,
"learning_rate": 0.002,
"loss": 1.9891000986099243,
"step": 1750
},
{
"epoch": 0.1752,
"grad_norm": 0.09001124650239944,
"learning_rate": 0.002,
"loss": 1.9349809885025024,
"step": 1752
},
{
"epoch": 0.1754,
"grad_norm": 0.087516650557518,
"learning_rate": 0.002,
"loss": 1.8842822313308716,
"step": 1754
},
{
"epoch": 0.1756,
"grad_norm": 0.0836789458990097,
"learning_rate": 0.002,
"loss": 1.9222761392593384,
"step": 1756
},
{
"epoch": 0.1758,
"grad_norm": 0.08474660664796829,
"learning_rate": 0.002,
"loss": 2.075972557067871,
"step": 1758
},
{
"epoch": 0.176,
"grad_norm": 0.6000994443893433,
"learning_rate": 0.002,
"loss": 2.0548179149627686,
"step": 1760
},
{
"epoch": 0.1762,
"grad_norm": 0.08651186525821686,
"learning_rate": 0.002,
"loss": 2.043919563293457,
"step": 1762
},
{
"epoch": 0.1764,
"grad_norm": 0.08672037720680237,
"learning_rate": 0.002,
"loss": 1.9761316776275635,
"step": 1764
},
{
"epoch": 0.1766,
"grad_norm": 0.08266716450452805,
"learning_rate": 0.002,
"loss": 1.9101964235305786,
"step": 1766
},
{
"epoch": 0.1768,
"grad_norm": 0.08747704327106476,
"learning_rate": 0.002,
"loss": 1.858526349067688,
"step": 1768
},
{
"epoch": 0.177,
"grad_norm": 0.08086052536964417,
"learning_rate": 0.002,
"loss": 2.04201602935791,
"step": 1770
},
{
"epoch": 0.1772,
"grad_norm": 0.08061601221561432,
"learning_rate": 0.002,
"loss": 1.9362319707870483,
"step": 1772
},
{
"epoch": 0.1774,
"grad_norm": 0.08680877089500427,
"learning_rate": 0.002,
"loss": 1.8955398797988892,
"step": 1774
},
{
"epoch": 0.1776,
"grad_norm": 0.08626703172922134,
"learning_rate": 0.002,
"loss": 1.9597290754318237,
"step": 1776
},
{
"epoch": 0.1778,
"grad_norm": 0.08292295783758163,
"learning_rate": 0.002,
"loss": 2.015136957168579,
"step": 1778
},
{
"epoch": 0.178,
"grad_norm": 0.09201980382204056,
"learning_rate": 0.002,
"loss": 1.9670177698135376,
"step": 1780
},
{
"epoch": 0.1782,
"grad_norm": 0.08161337673664093,
"learning_rate": 0.002,
"loss": 1.9454771280288696,
"step": 1782
},
{
"epoch": 0.1784,
"grad_norm": 0.08982264995574951,
"learning_rate": 0.002,
"loss": 1.9867963790893555,
"step": 1784
},
{
"epoch": 0.1786,
"grad_norm": 0.08044017106294632,
"learning_rate": 0.002,
"loss": 1.960343599319458,
"step": 1786
},
{
"epoch": 0.1788,
"grad_norm": 0.08217635005712509,
"learning_rate": 0.002,
"loss": 1.8296501636505127,
"step": 1788
},
{
"epoch": 0.179,
"grad_norm": 0.08359440416097641,
"learning_rate": 0.002,
"loss": 1.9807384014129639,
"step": 1790
},
{
"epoch": 0.1792,
"grad_norm": 0.08547817915678024,
"learning_rate": 0.002,
"loss": 1.9460605382919312,
"step": 1792
},
{
"epoch": 0.1794,
"grad_norm": 0.08134095370769501,
"learning_rate": 0.002,
"loss": 2.0722501277923584,
"step": 1794
},
{
"epoch": 0.1796,
"grad_norm": 0.084647998213768,
"learning_rate": 0.002,
"loss": 1.8832693099975586,
"step": 1796
},
{
"epoch": 0.1798,
"grad_norm": 0.08631309121847153,
"learning_rate": 0.002,
"loss": 1.9553685188293457,
"step": 1798
},
{
"epoch": 0.18,
"grad_norm": 0.08567328006029129,
"learning_rate": 0.002,
"loss": 2.0348782539367676,
"step": 1800
},
{
"epoch": 0.1802,
"grad_norm": 0.08318687230348587,
"learning_rate": 0.002,
"loss": 1.779697060585022,
"step": 1802
},
{
"epoch": 0.1804,
"grad_norm": 0.08443592488765717,
"learning_rate": 0.002,
"loss": 2.09885573387146,
"step": 1804
},
{
"epoch": 0.1806,
"grad_norm": 0.09795618057250977,
"learning_rate": 0.002,
"loss": 2.0944602489471436,
"step": 1806
},
{
"epoch": 0.1808,
"grad_norm": 0.09287634491920471,
"learning_rate": 0.002,
"loss": 1.956834077835083,
"step": 1808
},
{
"epoch": 0.181,
"grad_norm": 0.08185828477144241,
"learning_rate": 0.002,
"loss": 1.877648949623108,
"step": 1810
},
{
"epoch": 0.1812,
"grad_norm": 0.08287682384252548,
"learning_rate": 0.002,
"loss": 1.947015404701233,
"step": 1812
},
{
"epoch": 0.1814,
"grad_norm": 0.08119559288024902,
"learning_rate": 0.002,
"loss": 1.9950679540634155,
"step": 1814
},
{
"epoch": 0.1816,
"grad_norm": 0.08255063742399216,
"learning_rate": 0.002,
"loss": 1.9064756631851196,
"step": 1816
},
{
"epoch": 0.1818,
"grad_norm": 0.07887885719537735,
"learning_rate": 0.002,
"loss": 1.8976937532424927,
"step": 1818
},
{
"epoch": 0.182,
"grad_norm": 0.08447431772947311,
"learning_rate": 0.002,
"loss": 1.8944346904754639,
"step": 1820
},
{
"epoch": 0.1822,
"grad_norm": 0.08535470813512802,
"learning_rate": 0.002,
"loss": 2.0893661975860596,
"step": 1822
},
{
"epoch": 0.1824,
"grad_norm": 0.08468759804964066,
"learning_rate": 0.002,
"loss": 2.001993417739868,
"step": 1824
},
{
"epoch": 0.1826,
"grad_norm": 0.08959305286407471,
"learning_rate": 0.002,
"loss": 2.005653142929077,
"step": 1826
},
{
"epoch": 0.1828,
"grad_norm": 0.08373583853244781,
"learning_rate": 0.002,
"loss": 2.0966317653656006,
"step": 1828
},
{
"epoch": 0.183,
"grad_norm": 0.09113795310258865,
"learning_rate": 0.002,
"loss": 1.9518117904663086,
"step": 1830
},
{
"epoch": 0.1832,
"grad_norm": 0.08594139665365219,
"learning_rate": 0.002,
"loss": 2.1065845489501953,
"step": 1832
},
{
"epoch": 0.1834,
"grad_norm": 0.08325822651386261,
"learning_rate": 0.002,
"loss": 1.9954397678375244,
"step": 1834
},
{
"epoch": 0.1836,
"grad_norm": 0.08779548108577728,
"learning_rate": 0.002,
"loss": 1.9716862440109253,
"step": 1836
},
{
"epoch": 0.1838,
"grad_norm": 0.08372428268194199,
"learning_rate": 0.002,
"loss": 1.9748488664627075,
"step": 1838
},
{
"epoch": 0.184,
"grad_norm": 0.08295018970966339,
"learning_rate": 0.002,
"loss": 2.0002617835998535,
"step": 1840
},
{
"epoch": 0.1842,
"grad_norm": 0.08405480533838272,
"learning_rate": 0.002,
"loss": 1.9223436117172241,
"step": 1842
},
{
"epoch": 0.1844,
"grad_norm": 0.08446697890758514,
"learning_rate": 0.002,
"loss": 1.8754009008407593,
"step": 1844
},
{
"epoch": 0.1846,
"grad_norm": 0.08543042838573456,
"learning_rate": 0.002,
"loss": 1.8577792644500732,
"step": 1846
},
{
"epoch": 0.1848,
"grad_norm": 0.08950234204530716,
"learning_rate": 0.002,
"loss": 2.096989870071411,
"step": 1848
},
{
"epoch": 0.185,
"grad_norm": 0.0939083844423294,
"learning_rate": 0.002,
"loss": 1.8748059272766113,
"step": 1850
},
{
"epoch": 0.1852,
"grad_norm": 0.09873612225055695,
"learning_rate": 0.002,
"loss": 1.9742906093597412,
"step": 1852
},
{
"epoch": 0.1854,
"grad_norm": 0.08476125448942184,
"learning_rate": 0.002,
"loss": 2.1047325134277344,
"step": 1854
},
{
"epoch": 0.1856,
"grad_norm": 0.0880674496293068,
"learning_rate": 0.002,
"loss": 1.8781228065490723,
"step": 1856
},
{
"epoch": 0.1858,
"grad_norm": 0.0865429937839508,
"learning_rate": 0.002,
"loss": 1.9451115131378174,
"step": 1858
},
{
"epoch": 0.186,
"grad_norm": 0.08469100296497345,
"learning_rate": 0.002,
"loss": 2.0146076679229736,
"step": 1860
},
{
"epoch": 0.1862,
"grad_norm": 0.08361531794071198,
"learning_rate": 0.002,
"loss": 1.9248909950256348,
"step": 1862
},
{
"epoch": 0.1864,
"grad_norm": 0.08243401348590851,
"learning_rate": 0.002,
"loss": 2.044677257537842,
"step": 1864
},
{
"epoch": 0.1866,
"grad_norm": 0.08730989694595337,
"learning_rate": 0.002,
"loss": 1.9625319242477417,
"step": 1866
},
{
"epoch": 0.1868,
"grad_norm": 0.0863344743847847,
"learning_rate": 0.002,
"loss": 1.8846096992492676,
"step": 1868
},
{
"epoch": 0.187,
"grad_norm": 0.08576243370771408,
"learning_rate": 0.002,
"loss": 1.9552323818206787,
"step": 1870
},
{
"epoch": 0.1872,
"grad_norm": 0.08861047029495239,
"learning_rate": 0.002,
"loss": 2.0286762714385986,
"step": 1872
},
{
"epoch": 0.1874,
"grad_norm": 0.08303575962781906,
"learning_rate": 0.002,
"loss": 1.9901646375656128,
"step": 1874
},
{
"epoch": 0.1876,
"grad_norm": 0.08088395744562149,
"learning_rate": 0.002,
"loss": 1.7853816747665405,
"step": 1876
},
{
"epoch": 0.1878,
"grad_norm": 0.08346857875585556,
"learning_rate": 0.002,
"loss": 1.9030942916870117,
"step": 1878
},
{
"epoch": 0.188,
"grad_norm": 0.08730462193489075,
"learning_rate": 0.002,
"loss": 1.9499236345291138,
"step": 1880
},
{
"epoch": 0.1882,
"grad_norm": 0.08578626811504364,
"learning_rate": 0.002,
"loss": 1.9349133968353271,
"step": 1882
},
{
"epoch": 0.1884,
"grad_norm": 0.08277959376573563,
"learning_rate": 0.002,
"loss": 1.8892171382904053,
"step": 1884
},
{
"epoch": 0.1886,
"grad_norm": 0.08430566638708115,
"learning_rate": 0.002,
"loss": 1.9678326845169067,
"step": 1886
},
{
"epoch": 0.1888,
"grad_norm": 0.08673564344644547,
"learning_rate": 0.002,
"loss": 1.9515163898468018,
"step": 1888
},
{
"epoch": 0.189,
"grad_norm": 0.0811837762594223,
"learning_rate": 0.002,
"loss": 1.9885120391845703,
"step": 1890
},
{
"epoch": 0.1892,
"grad_norm": 0.09070060402154922,
"learning_rate": 0.002,
"loss": 1.9701941013336182,
"step": 1892
},
{
"epoch": 0.1894,
"grad_norm": 0.0852934792637825,
"learning_rate": 0.002,
"loss": 1.8665810823440552,
"step": 1894
},
{
"epoch": 0.1896,
"grad_norm": 0.08353486657142639,
"learning_rate": 0.002,
"loss": 1.9298522472381592,
"step": 1896
},
{
"epoch": 0.1898,
"grad_norm": 0.09107336401939392,
"learning_rate": 0.002,
"loss": 1.899435043334961,
"step": 1898
},
{
"epoch": 0.19,
"grad_norm": 0.08662459999322891,
"learning_rate": 0.002,
"loss": 2.028043031692505,
"step": 1900
},
{
"epoch": 0.1902,
"grad_norm": 0.08985409140586853,
"learning_rate": 0.002,
"loss": 2.043958902359009,
"step": 1902
},
{
"epoch": 0.1904,
"grad_norm": 0.08701593428850174,
"learning_rate": 0.002,
"loss": 1.9687647819519043,
"step": 1904
},
{
"epoch": 0.1906,
"grad_norm": 0.08502478152513504,
"learning_rate": 0.002,
"loss": 2.1415886878967285,
"step": 1906
},
{
"epoch": 0.1908,
"grad_norm": 0.0901290699839592,
"learning_rate": 0.002,
"loss": 1.940739393234253,
"step": 1908
},
{
"epoch": 0.191,
"grad_norm": 0.08363977819681168,
"learning_rate": 0.002,
"loss": 1.9836229085922241,
"step": 1910
},
{
"epoch": 0.1912,
"grad_norm": 0.08620919287204742,
"learning_rate": 0.002,
"loss": 2.012991189956665,
"step": 1912
},
{
"epoch": 0.1914,
"grad_norm": 0.08760473877191544,
"learning_rate": 0.002,
"loss": 2.0166289806365967,
"step": 1914
},
{
"epoch": 0.1916,
"grad_norm": 0.08473231643438339,
"learning_rate": 0.002,
"loss": 2.0080785751342773,
"step": 1916
},
{
"epoch": 0.1918,
"grad_norm": 0.08816628158092499,
"learning_rate": 0.002,
"loss": 2.0298595428466797,
"step": 1918
},
{
"epoch": 0.192,
"grad_norm": 0.08431993424892426,
"learning_rate": 0.002,
"loss": 1.9413989782333374,
"step": 1920
},
{
"epoch": 0.1922,
"grad_norm": 0.08576439321041107,
"learning_rate": 0.002,
"loss": 1.836694598197937,
"step": 1922
},
{
"epoch": 0.1924,
"grad_norm": 0.08376266062259674,
"learning_rate": 0.002,
"loss": 1.9192601442337036,
"step": 1924
},
{
"epoch": 0.1926,
"grad_norm": 0.0852290689945221,
"learning_rate": 0.002,
"loss": 1.9260629415512085,
"step": 1926
},
{
"epoch": 0.1928,
"grad_norm": 0.0883747860789299,
"learning_rate": 0.002,
"loss": 1.9678843021392822,
"step": 1928
},
{
"epoch": 0.193,
"grad_norm": 0.08285381644964218,
"learning_rate": 0.002,
"loss": 2.0911788940429688,
"step": 1930
},
{
"epoch": 0.1932,
"grad_norm": 0.08180463314056396,
"learning_rate": 0.002,
"loss": 1.877135157585144,
"step": 1932
},
{
"epoch": 0.1934,
"grad_norm": 0.08843019604682922,
"learning_rate": 0.002,
"loss": 1.8937883377075195,
"step": 1934
},
{
"epoch": 0.1936,
"grad_norm": 0.08965593576431274,
"learning_rate": 0.002,
"loss": 2.018563985824585,
"step": 1936
},
{
"epoch": 0.1938,
"grad_norm": 0.09083735197782516,
"learning_rate": 0.002,
"loss": 1.9771754741668701,
"step": 1938
},
{
"epoch": 0.194,
"grad_norm": 0.08669350296258926,
"learning_rate": 0.002,
"loss": 1.9471850395202637,
"step": 1940
},
{
"epoch": 0.1942,
"grad_norm": 0.08626765757799149,
"learning_rate": 0.002,
"loss": 1.8741486072540283,
"step": 1942
},
{
"epoch": 0.1944,
"grad_norm": 0.08252197504043579,
"learning_rate": 0.002,
"loss": 1.8809633255004883,
"step": 1944
},
{
"epoch": 0.1946,
"grad_norm": 0.0845436081290245,
"learning_rate": 0.002,
"loss": 1.8249372243881226,
"step": 1946
},
{
"epoch": 0.1948,
"grad_norm": 0.0836493968963623,
"learning_rate": 0.002,
"loss": 1.8964265584945679,
"step": 1948
},
{
"epoch": 0.195,
"grad_norm": 0.08754970878362656,
"learning_rate": 0.002,
"loss": 1.9884169101715088,
"step": 1950
},
{
"epoch": 0.1952,
"grad_norm": 0.08162931352853775,
"learning_rate": 0.002,
"loss": 1.9248230457305908,
"step": 1952
},
{
"epoch": 0.1954,
"grad_norm": 0.08138412982225418,
"learning_rate": 0.002,
"loss": 1.9708248376846313,
"step": 1954
},
{
"epoch": 0.1956,
"grad_norm": 0.08538922667503357,
"learning_rate": 0.002,
"loss": 1.8737009763717651,
"step": 1956
},
{
"epoch": 0.1958,
"grad_norm": 0.08595019578933716,
"learning_rate": 0.002,
"loss": 1.902658224105835,
"step": 1958
},
{
"epoch": 0.196,
"grad_norm": 0.08747034519910812,
"learning_rate": 0.002,
"loss": 2.1205339431762695,
"step": 1960
},
{
"epoch": 0.1962,
"grad_norm": 0.08311095833778381,
"learning_rate": 0.002,
"loss": 2.0103437900543213,
"step": 1962
},
{
"epoch": 0.1964,
"grad_norm": 0.08260630071163177,
"learning_rate": 0.002,
"loss": 2.08880352973938,
"step": 1964
},
{
"epoch": 0.1966,
"grad_norm": 0.08510100841522217,
"learning_rate": 0.002,
"loss": 1.9506934881210327,
"step": 1966
},
{
"epoch": 0.1968,
"grad_norm": 0.08761399984359741,
"learning_rate": 0.002,
"loss": 1.925706386566162,
"step": 1968
},
{
"epoch": 0.197,
"grad_norm": 0.08779874444007874,
"learning_rate": 0.002,
"loss": 1.9873325824737549,
"step": 1970
},
{
"epoch": 0.1972,
"grad_norm": 0.08690716326236725,
"learning_rate": 0.002,
"loss": 1.9406440258026123,
"step": 1972
},
{
"epoch": 0.1974,
"grad_norm": 0.09347079694271088,
"learning_rate": 0.002,
"loss": 1.909558653831482,
"step": 1974
},
{
"epoch": 0.1976,
"grad_norm": 0.08548922091722488,
"learning_rate": 0.002,
"loss": 1.9357054233551025,
"step": 1976
},
{
"epoch": 0.1978,
"grad_norm": 0.08342369645833969,
"learning_rate": 0.002,
"loss": 1.9169145822525024,
"step": 1978
},
{
"epoch": 0.198,
"grad_norm": 0.09005551040172577,
"learning_rate": 0.002,
"loss": 2.041494846343994,
"step": 1980
},
{
"epoch": 0.1982,
"grad_norm": 0.0824834331870079,
"learning_rate": 0.002,
"loss": 1.9692611694335938,
"step": 1982
},
{
"epoch": 0.1984,
"grad_norm": 0.08805745840072632,
"learning_rate": 0.002,
"loss": 1.8638404607772827,
"step": 1984
},
{
"epoch": 0.1986,
"grad_norm": 0.08590634167194366,
"learning_rate": 0.002,
"loss": 2.159425735473633,
"step": 1986
},
{
"epoch": 0.1988,
"grad_norm": 0.09385378658771515,
"learning_rate": 0.002,
"loss": 1.94144868850708,
"step": 1988
},
{
"epoch": 0.199,
"grad_norm": 0.09556061029434204,
"learning_rate": 0.002,
"loss": 1.9435380697250366,
"step": 1990
},
{
"epoch": 0.1992,
"grad_norm": 0.08348517119884491,
"learning_rate": 0.002,
"loss": 2.0118563175201416,
"step": 1992
},
{
"epoch": 0.1994,
"grad_norm": 0.08558572828769684,
"learning_rate": 0.002,
"loss": 2.076636791229248,
"step": 1994
},
{
"epoch": 0.1996,
"grad_norm": 0.0863652303814888,
"learning_rate": 0.002,
"loss": 1.9437497854232788,
"step": 1996
},
{
"epoch": 0.1998,
"grad_norm": 0.08715331554412842,
"learning_rate": 0.002,
"loss": 2.138975143432617,
"step": 1998
},
{
"epoch": 0.2,
"grad_norm": 0.08320055156946182,
"learning_rate": 0.002,
"loss": 1.9396672248840332,
"step": 2000
},
{
"epoch": 0.2,
"eval_loss": 2.0632307529449463,
"eval_runtime": 204.5116,
"eval_samples_per_second": 3.056,
"eval_steps_per_second": 0.386,
"step": 2000
}
],
"logging_steps": 2,
"max_steps": 10000,
"num_input_tokens_seen": 0,
"num_train_epochs": 9223372036854775807,
"save_steps": 1000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 1.7568125032652122e+18,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}