|
{
|
|
"best_metric": 0.17312079668045044,
|
|
"best_model_checkpoint": "vit_epochs5_batch32_lr5e-05_size224_tiles9_seed2_q2_complexity\\checkpoint-469",
|
|
"epoch": 5.0,
|
|
"eval_steps": 500,
|
|
"global_step": 2345,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"epoch": 0.010660980810234541,
|
|
"grad_norm": 4.213649749755859,
|
|
"learning_rate": 4.989339019189766e-05,
|
|
"loss": 0.1615,
|
|
"step": 5
|
|
},
|
|
{
|
|
"epoch": 0.021321961620469083,
|
|
"grad_norm": 1.5957719087600708,
|
|
"learning_rate": 4.978678038379531e-05,
|
|
"loss": 0.2217,
|
|
"step": 10
|
|
},
|
|
{
|
|
"epoch": 0.031982942430703626,
|
|
"grad_norm": 9.061514854431152,
|
|
"learning_rate": 4.9680170575692967e-05,
|
|
"loss": 0.106,
|
|
"step": 15
|
|
},
|
|
{
|
|
"epoch": 0.042643923240938165,
|
|
"grad_norm": 7.132235527038574,
|
|
"learning_rate": 4.957356076759062e-05,
|
|
"loss": 0.2051,
|
|
"step": 20
|
|
},
|
|
{
|
|
"epoch": 0.053304904051172705,
|
|
"grad_norm": 3.497645616531372,
|
|
"learning_rate": 4.9466950959488276e-05,
|
|
"loss": 0.1147,
|
|
"step": 25
|
|
},
|
|
{
|
|
"epoch": 0.06396588486140725,
|
|
"grad_norm": 11.90768051147461,
|
|
"learning_rate": 4.936034115138593e-05,
|
|
"loss": 0.1967,
|
|
"step": 30
|
|
},
|
|
{
|
|
"epoch": 0.07462686567164178,
|
|
"grad_norm": 6.974191188812256,
|
|
"learning_rate": 4.9253731343283586e-05,
|
|
"loss": 0.1531,
|
|
"step": 35
|
|
},
|
|
{
|
|
"epoch": 0.08528784648187633,
|
|
"grad_norm": 2.5099663734436035,
|
|
"learning_rate": 4.914712153518124e-05,
|
|
"loss": 0.1764,
|
|
"step": 40
|
|
},
|
|
{
|
|
"epoch": 0.09594882729211088,
|
|
"grad_norm": 8.150910377502441,
|
|
"learning_rate": 4.904051172707889e-05,
|
|
"loss": 0.159,
|
|
"step": 45
|
|
},
|
|
{
|
|
"epoch": 0.10660980810234541,
|
|
"grad_norm": 0.8478419780731201,
|
|
"learning_rate": 4.893390191897655e-05,
|
|
"loss": 0.1091,
|
|
"step": 50
|
|
},
|
|
{
|
|
"epoch": 0.11727078891257996,
|
|
"grad_norm": 7.68576192855835,
|
|
"learning_rate": 4.88272921108742e-05,
|
|
"loss": 0.1553,
|
|
"step": 55
|
|
},
|
|
{
|
|
"epoch": 0.1279317697228145,
|
|
"grad_norm": 3.7954885959625244,
|
|
"learning_rate": 4.872068230277186e-05,
|
|
"loss": 0.1131,
|
|
"step": 60
|
|
},
|
|
{
|
|
"epoch": 0.13859275053304904,
|
|
"grad_norm": 0.5815512537956238,
|
|
"learning_rate": 4.861407249466951e-05,
|
|
"loss": 0.091,
|
|
"step": 65
|
|
},
|
|
{
|
|
"epoch": 0.14925373134328357,
|
|
"grad_norm": 10.776674270629883,
|
|
"learning_rate": 4.850746268656717e-05,
|
|
"loss": 0.1171,
|
|
"step": 70
|
|
},
|
|
{
|
|
"epoch": 0.15991471215351813,
|
|
"grad_norm": 7.9250969886779785,
|
|
"learning_rate": 4.840085287846482e-05,
|
|
"loss": 0.0976,
|
|
"step": 75
|
|
},
|
|
{
|
|
"epoch": 0.17057569296375266,
|
|
"grad_norm": 2.3389108180999756,
|
|
"learning_rate": 4.829424307036248e-05,
|
|
"loss": 0.1336,
|
|
"step": 80
|
|
},
|
|
{
|
|
"epoch": 0.1812366737739872,
|
|
"grad_norm": 3.41266131401062,
|
|
"learning_rate": 4.8187633262260126e-05,
|
|
"loss": 0.069,
|
|
"step": 85
|
|
},
|
|
{
|
|
"epoch": 0.19189765458422176,
|
|
"grad_norm": 2.4709668159484863,
|
|
"learning_rate": 4.808102345415779e-05,
|
|
"loss": 0.1309,
|
|
"step": 90
|
|
},
|
|
{
|
|
"epoch": 0.2025586353944563,
|
|
"grad_norm": 8.687853813171387,
|
|
"learning_rate": 4.7974413646055436e-05,
|
|
"loss": 0.1397,
|
|
"step": 95
|
|
},
|
|
{
|
|
"epoch": 0.21321961620469082,
|
|
"grad_norm": 3.623166561126709,
|
|
"learning_rate": 4.78678038379531e-05,
|
|
"loss": 0.147,
|
|
"step": 100
|
|
},
|
|
{
|
|
"epoch": 0.22388059701492538,
|
|
"grad_norm": 4.238902568817139,
|
|
"learning_rate": 4.7761194029850745e-05,
|
|
"loss": 0.1205,
|
|
"step": 105
|
|
},
|
|
{
|
|
"epoch": 0.2345415778251599,
|
|
"grad_norm": 5.488494873046875,
|
|
"learning_rate": 4.765458422174841e-05,
|
|
"loss": 0.1389,
|
|
"step": 110
|
|
},
|
|
{
|
|
"epoch": 0.24520255863539445,
|
|
"grad_norm": 4.744454860687256,
|
|
"learning_rate": 4.7547974413646055e-05,
|
|
"loss": 0.1626,
|
|
"step": 115
|
|
},
|
|
{
|
|
"epoch": 0.255863539445629,
|
|
"grad_norm": 4.200260162353516,
|
|
"learning_rate": 4.7441364605543716e-05,
|
|
"loss": 0.0701,
|
|
"step": 120
|
|
},
|
|
{
|
|
"epoch": 0.26652452025586354,
|
|
"grad_norm": 3.702713966369629,
|
|
"learning_rate": 4.7334754797441364e-05,
|
|
"loss": 0.0985,
|
|
"step": 125
|
|
},
|
|
{
|
|
"epoch": 0.2771855010660981,
|
|
"grad_norm": 4.4238715171813965,
|
|
"learning_rate": 4.7228144989339026e-05,
|
|
"loss": 0.1493,
|
|
"step": 130
|
|
},
|
|
{
|
|
"epoch": 0.2878464818763326,
|
|
"grad_norm": 2.5960519313812256,
|
|
"learning_rate": 4.7121535181236674e-05,
|
|
"loss": 0.0376,
|
|
"step": 135
|
|
},
|
|
{
|
|
"epoch": 0.29850746268656714,
|
|
"grad_norm": 5.107735633850098,
|
|
"learning_rate": 4.7014925373134335e-05,
|
|
"loss": 0.1315,
|
|
"step": 140
|
|
},
|
|
{
|
|
"epoch": 0.3091684434968017,
|
|
"grad_norm": 3.4888722896575928,
|
|
"learning_rate": 4.690831556503198e-05,
|
|
"loss": 0.0732,
|
|
"step": 145
|
|
},
|
|
{
|
|
"epoch": 0.31982942430703626,
|
|
"grad_norm": 3.210352659225464,
|
|
"learning_rate": 4.6801705756929645e-05,
|
|
"loss": 0.0833,
|
|
"step": 150
|
|
},
|
|
{
|
|
"epoch": 0.3304904051172708,
|
|
"grad_norm": 3.754889726638794,
|
|
"learning_rate": 4.669509594882729e-05,
|
|
"loss": 0.0888,
|
|
"step": 155
|
|
},
|
|
{
|
|
"epoch": 0.3411513859275053,
|
|
"grad_norm": 14.111072540283203,
|
|
"learning_rate": 4.658848614072495e-05,
|
|
"loss": 0.1112,
|
|
"step": 160
|
|
},
|
|
{
|
|
"epoch": 0.35181236673773986,
|
|
"grad_norm": 4.763731956481934,
|
|
"learning_rate": 4.64818763326226e-05,
|
|
"loss": 0.0836,
|
|
"step": 165
|
|
},
|
|
{
|
|
"epoch": 0.3624733475479744,
|
|
"grad_norm": 5.260176658630371,
|
|
"learning_rate": 4.637526652452026e-05,
|
|
"loss": 0.1355,
|
|
"step": 170
|
|
},
|
|
{
|
|
"epoch": 0.373134328358209,
|
|
"grad_norm": 5.902770042419434,
|
|
"learning_rate": 4.626865671641791e-05,
|
|
"loss": 0.0921,
|
|
"step": 175
|
|
},
|
|
{
|
|
"epoch": 0.3837953091684435,
|
|
"grad_norm": 6.20088529586792,
|
|
"learning_rate": 4.6162046908315566e-05,
|
|
"loss": 0.1773,
|
|
"step": 180
|
|
},
|
|
{
|
|
"epoch": 0.39445628997867804,
|
|
"grad_norm": 5.141183376312256,
|
|
"learning_rate": 4.605543710021322e-05,
|
|
"loss": 0.1065,
|
|
"step": 185
|
|
},
|
|
{
|
|
"epoch": 0.4051172707889126,
|
|
"grad_norm": 1.2287838459014893,
|
|
"learning_rate": 4.5948827292110876e-05,
|
|
"loss": 0.0689,
|
|
"step": 190
|
|
},
|
|
{
|
|
"epoch": 0.4157782515991471,
|
|
"grad_norm": 1.0071206092834473,
|
|
"learning_rate": 4.584221748400853e-05,
|
|
"loss": 0.0615,
|
|
"step": 195
|
|
},
|
|
{
|
|
"epoch": 0.42643923240938164,
|
|
"grad_norm": 0.6955390572547913,
|
|
"learning_rate": 4.5735607675906185e-05,
|
|
"loss": 0.1116,
|
|
"step": 200
|
|
},
|
|
{
|
|
"epoch": 0.43710021321961623,
|
|
"grad_norm": 8.547406196594238,
|
|
"learning_rate": 4.562899786780384e-05,
|
|
"loss": 0.1041,
|
|
"step": 205
|
|
},
|
|
{
|
|
"epoch": 0.44776119402985076,
|
|
"grad_norm": 0.12464961409568787,
|
|
"learning_rate": 4.5522388059701495e-05,
|
|
"loss": 0.0551,
|
|
"step": 210
|
|
},
|
|
{
|
|
"epoch": 0.4584221748400853,
|
|
"grad_norm": 9.267840385437012,
|
|
"learning_rate": 4.541577825159915e-05,
|
|
"loss": 0.1245,
|
|
"step": 215
|
|
},
|
|
{
|
|
"epoch": 0.4690831556503198,
|
|
"grad_norm": 5.183570384979248,
|
|
"learning_rate": 4.5309168443496804e-05,
|
|
"loss": 0.1303,
|
|
"step": 220
|
|
},
|
|
{
|
|
"epoch": 0.47974413646055436,
|
|
"grad_norm": 5.079310417175293,
|
|
"learning_rate": 4.520255863539446e-05,
|
|
"loss": 0.1154,
|
|
"step": 225
|
|
},
|
|
{
|
|
"epoch": 0.4904051172707889,
|
|
"grad_norm": 1.7525018453598022,
|
|
"learning_rate": 4.5095948827292114e-05,
|
|
"loss": 0.166,
|
|
"step": 230
|
|
},
|
|
{
|
|
"epoch": 0.5010660980810234,
|
|
"grad_norm": 3.7788939476013184,
|
|
"learning_rate": 4.498933901918977e-05,
|
|
"loss": 0.1074,
|
|
"step": 235
|
|
},
|
|
{
|
|
"epoch": 0.511727078891258,
|
|
"grad_norm": 8.968123435974121,
|
|
"learning_rate": 4.488272921108742e-05,
|
|
"loss": 0.2202,
|
|
"step": 240
|
|
},
|
|
{
|
|
"epoch": 0.5223880597014925,
|
|
"grad_norm": 0.9381376504898071,
|
|
"learning_rate": 4.477611940298508e-05,
|
|
"loss": 0.0574,
|
|
"step": 245
|
|
},
|
|
{
|
|
"epoch": 0.5330490405117271,
|
|
"grad_norm": 3.7572848796844482,
|
|
"learning_rate": 4.466950959488273e-05,
|
|
"loss": 0.1386,
|
|
"step": 250
|
|
},
|
|
{
|
|
"epoch": 0.5437100213219617,
|
|
"grad_norm": 3.773059129714966,
|
|
"learning_rate": 4.456289978678039e-05,
|
|
"loss": 0.0807,
|
|
"step": 255
|
|
},
|
|
{
|
|
"epoch": 0.5543710021321961,
|
|
"grad_norm": 3.1569793224334717,
|
|
"learning_rate": 4.445628997867804e-05,
|
|
"loss": 0.1055,
|
|
"step": 260
|
|
},
|
|
{
|
|
"epoch": 0.5650319829424307,
|
|
"grad_norm": 8.580225944519043,
|
|
"learning_rate": 4.43496801705757e-05,
|
|
"loss": 0.1352,
|
|
"step": 265
|
|
},
|
|
{
|
|
"epoch": 0.5756929637526652,
|
|
"grad_norm": 6.328425884246826,
|
|
"learning_rate": 4.424307036247335e-05,
|
|
"loss": 0.0949,
|
|
"step": 270
|
|
},
|
|
{
|
|
"epoch": 0.5863539445628998,
|
|
"grad_norm": 6.301974296569824,
|
|
"learning_rate": 4.4136460554371006e-05,
|
|
"loss": 0.1122,
|
|
"step": 275
|
|
},
|
|
{
|
|
"epoch": 0.5970149253731343,
|
|
"grad_norm": 3.1882998943328857,
|
|
"learning_rate": 4.402985074626866e-05,
|
|
"loss": 0.0663,
|
|
"step": 280
|
|
},
|
|
{
|
|
"epoch": 0.6076759061833689,
|
|
"grad_norm": 5.099627494812012,
|
|
"learning_rate": 4.3923240938166316e-05,
|
|
"loss": 0.0888,
|
|
"step": 285
|
|
},
|
|
{
|
|
"epoch": 0.6183368869936035,
|
|
"grad_norm": 3.622894525527954,
|
|
"learning_rate": 4.381663113006397e-05,
|
|
"loss": 0.08,
|
|
"step": 290
|
|
},
|
|
{
|
|
"epoch": 0.6289978678038379,
|
|
"grad_norm": 3.7854912281036377,
|
|
"learning_rate": 4.3710021321961625e-05,
|
|
"loss": 0.1065,
|
|
"step": 295
|
|
},
|
|
{
|
|
"epoch": 0.6396588486140725,
|
|
"grad_norm": 4.7853593826293945,
|
|
"learning_rate": 4.360341151385928e-05,
|
|
"loss": 0.0858,
|
|
"step": 300
|
|
},
|
|
{
|
|
"epoch": 0.650319829424307,
|
|
"grad_norm": 4.2659759521484375,
|
|
"learning_rate": 4.3496801705756935e-05,
|
|
"loss": 0.0745,
|
|
"step": 305
|
|
},
|
|
{
|
|
"epoch": 0.6609808102345416,
|
|
"grad_norm": 5.468679904937744,
|
|
"learning_rate": 4.339019189765459e-05,
|
|
"loss": 0.1323,
|
|
"step": 310
|
|
},
|
|
{
|
|
"epoch": 0.6716417910447762,
|
|
"grad_norm": 4.176182746887207,
|
|
"learning_rate": 4.328358208955224e-05,
|
|
"loss": 0.0689,
|
|
"step": 315
|
|
},
|
|
{
|
|
"epoch": 0.6823027718550106,
|
|
"grad_norm": 0.18954375386238098,
|
|
"learning_rate": 4.31769722814499e-05,
|
|
"loss": 0.0402,
|
|
"step": 320
|
|
},
|
|
{
|
|
"epoch": 0.6929637526652452,
|
|
"grad_norm": 10.576562881469727,
|
|
"learning_rate": 4.307036247334755e-05,
|
|
"loss": 0.0965,
|
|
"step": 325
|
|
},
|
|
{
|
|
"epoch": 0.7036247334754797,
|
|
"grad_norm": 5.207491874694824,
|
|
"learning_rate": 4.29637526652452e-05,
|
|
"loss": 0.0932,
|
|
"step": 330
|
|
},
|
|
{
|
|
"epoch": 0.7142857142857143,
|
|
"grad_norm": 2.177462577819824,
|
|
"learning_rate": 4.2857142857142856e-05,
|
|
"loss": 0.0731,
|
|
"step": 335
|
|
},
|
|
{
|
|
"epoch": 0.7249466950959488,
|
|
"grad_norm": 4.819252967834473,
|
|
"learning_rate": 4.275053304904051e-05,
|
|
"loss": 0.0728,
|
|
"step": 340
|
|
},
|
|
{
|
|
"epoch": 0.7356076759061834,
|
|
"grad_norm": 5.519009113311768,
|
|
"learning_rate": 4.2643923240938166e-05,
|
|
"loss": 0.135,
|
|
"step": 345
|
|
},
|
|
{
|
|
"epoch": 0.746268656716418,
|
|
"grad_norm": 3.2497637271881104,
|
|
"learning_rate": 4.253731343283582e-05,
|
|
"loss": 0.2316,
|
|
"step": 350
|
|
},
|
|
{
|
|
"epoch": 0.7569296375266524,
|
|
"grad_norm": 2.4187228679656982,
|
|
"learning_rate": 4.2430703624733475e-05,
|
|
"loss": 0.0343,
|
|
"step": 355
|
|
},
|
|
{
|
|
"epoch": 0.767590618336887,
|
|
"grad_norm": 0.9253553152084351,
|
|
"learning_rate": 4.232409381663113e-05,
|
|
"loss": 0.1067,
|
|
"step": 360
|
|
},
|
|
{
|
|
"epoch": 0.7782515991471215,
|
|
"grad_norm": 1.4930566549301147,
|
|
"learning_rate": 4.2217484008528785e-05,
|
|
"loss": 0.0688,
|
|
"step": 365
|
|
},
|
|
{
|
|
"epoch": 0.7889125799573561,
|
|
"grad_norm": 3.6474721431732178,
|
|
"learning_rate": 4.211087420042644e-05,
|
|
"loss": 0.078,
|
|
"step": 370
|
|
},
|
|
{
|
|
"epoch": 0.7995735607675906,
|
|
"grad_norm": 6.267404556274414,
|
|
"learning_rate": 4.2004264392324094e-05,
|
|
"loss": 0.1332,
|
|
"step": 375
|
|
},
|
|
{
|
|
"epoch": 0.8102345415778252,
|
|
"grad_norm": 0.7626522779464722,
|
|
"learning_rate": 4.189765458422175e-05,
|
|
"loss": 0.0597,
|
|
"step": 380
|
|
},
|
|
{
|
|
"epoch": 0.8208955223880597,
|
|
"grad_norm": 3.383026123046875,
|
|
"learning_rate": 4.1791044776119404e-05,
|
|
"loss": 0.1053,
|
|
"step": 385
|
|
},
|
|
{
|
|
"epoch": 0.8315565031982942,
|
|
"grad_norm": 4.386033058166504,
|
|
"learning_rate": 4.168443496801706e-05,
|
|
"loss": 0.2217,
|
|
"step": 390
|
|
},
|
|
{
|
|
"epoch": 0.8422174840085288,
|
|
"grad_norm": 3.1787729263305664,
|
|
"learning_rate": 4.157782515991471e-05,
|
|
"loss": 0.1962,
|
|
"step": 395
|
|
},
|
|
{
|
|
"epoch": 0.8528784648187633,
|
|
"grad_norm": 4.633083343505859,
|
|
"learning_rate": 4.147121535181237e-05,
|
|
"loss": 0.1686,
|
|
"step": 400
|
|
},
|
|
{
|
|
"epoch": 0.8635394456289979,
|
|
"grad_norm": 3.869473457336426,
|
|
"learning_rate": 4.136460554371002e-05,
|
|
"loss": 0.047,
|
|
"step": 405
|
|
},
|
|
{
|
|
"epoch": 0.8742004264392325,
|
|
"grad_norm": 0.5081688165664673,
|
|
"learning_rate": 4.125799573560768e-05,
|
|
"loss": 0.0767,
|
|
"step": 410
|
|
},
|
|
{
|
|
"epoch": 0.8848614072494669,
|
|
"grad_norm": 0.7017682790756226,
|
|
"learning_rate": 4.115138592750533e-05,
|
|
"loss": 0.1073,
|
|
"step": 415
|
|
},
|
|
{
|
|
"epoch": 0.8955223880597015,
|
|
"grad_norm": 0.2812543213367462,
|
|
"learning_rate": 4.104477611940299e-05,
|
|
"loss": 0.0665,
|
|
"step": 420
|
|
},
|
|
{
|
|
"epoch": 0.906183368869936,
|
|
"grad_norm": 3.224841594696045,
|
|
"learning_rate": 4.093816631130064e-05,
|
|
"loss": 0.1028,
|
|
"step": 425
|
|
},
|
|
{
|
|
"epoch": 0.9168443496801706,
|
|
"grad_norm": 1.9396395683288574,
|
|
"learning_rate": 4.0831556503198296e-05,
|
|
"loss": 0.0858,
|
|
"step": 430
|
|
},
|
|
{
|
|
"epoch": 0.9275053304904051,
|
|
"grad_norm": 1.766357183456421,
|
|
"learning_rate": 4.072494669509595e-05,
|
|
"loss": 0.105,
|
|
"step": 435
|
|
},
|
|
{
|
|
"epoch": 0.9381663113006397,
|
|
"grad_norm": 7.229098796844482,
|
|
"learning_rate": 4.0618336886993606e-05,
|
|
"loss": 0.102,
|
|
"step": 440
|
|
},
|
|
{
|
|
"epoch": 0.9488272921108742,
|
|
"grad_norm": 6.790981769561768,
|
|
"learning_rate": 4.051172707889126e-05,
|
|
"loss": 0.1531,
|
|
"step": 445
|
|
},
|
|
{
|
|
"epoch": 0.9594882729211087,
|
|
"grad_norm": 3.6507298946380615,
|
|
"learning_rate": 4.0405117270788915e-05,
|
|
"loss": 0.0739,
|
|
"step": 450
|
|
},
|
|
{
|
|
"epoch": 0.9701492537313433,
|
|
"grad_norm": 0.277474969625473,
|
|
"learning_rate": 4.029850746268657e-05,
|
|
"loss": 0.0705,
|
|
"step": 455
|
|
},
|
|
{
|
|
"epoch": 0.9808102345415778,
|
|
"grad_norm": 4.2118072509765625,
|
|
"learning_rate": 4.0191897654584225e-05,
|
|
"loss": 0.0377,
|
|
"step": 460
|
|
},
|
|
{
|
|
"epoch": 0.9914712153518124,
|
|
"grad_norm": 2.435720443725586,
|
|
"learning_rate": 4.008528784648188e-05,
|
|
"loss": 0.124,
|
|
"step": 465
|
|
},
|
|
{
|
|
"epoch": 1.0,
|
|
"eval_accuracy": 0.9429333333333333,
|
|
"eval_loss": 0.17312079668045044,
|
|
"eval_runtime": 18.4652,
|
|
"eval_samples_per_second": 203.084,
|
|
"eval_steps_per_second": 6.39,
|
|
"step": 469
|
|
},
|
|
{
|
|
"epoch": 1.0021321961620469,
|
|
"grad_norm": 4.489440441131592,
|
|
"learning_rate": 3.997867803837953e-05,
|
|
"loss": 0.0623,
|
|
"step": 470
|
|
},
|
|
{
|
|
"epoch": 1.0127931769722816,
|
|
"grad_norm": 0.3635157644748688,
|
|
"learning_rate": 3.987206823027719e-05,
|
|
"loss": 0.0136,
|
|
"step": 475
|
|
},
|
|
{
|
|
"epoch": 1.023454157782516,
|
|
"grad_norm": 0.04086770862340927,
|
|
"learning_rate": 3.976545842217484e-05,
|
|
"loss": 0.0336,
|
|
"step": 480
|
|
},
|
|
{
|
|
"epoch": 1.0341151385927505,
|
|
"grad_norm": 0.5033191442489624,
|
|
"learning_rate": 3.96588486140725e-05,
|
|
"loss": 0.0088,
|
|
"step": 485
|
|
},
|
|
{
|
|
"epoch": 1.044776119402985,
|
|
"grad_norm": 0.07028315961360931,
|
|
"learning_rate": 3.9552238805970146e-05,
|
|
"loss": 0.0158,
|
|
"step": 490
|
|
},
|
|
{
|
|
"epoch": 1.0554371002132197,
|
|
"grad_norm": 0.04735887050628662,
|
|
"learning_rate": 3.944562899786781e-05,
|
|
"loss": 0.0177,
|
|
"step": 495
|
|
},
|
|
{
|
|
"epoch": 1.0660980810234542,
|
|
"grad_norm": 0.43348121643066406,
|
|
"learning_rate": 3.9339019189765456e-05,
|
|
"loss": 0.0131,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 1.0767590618336886,
|
|
"grad_norm": 8.368108749389648,
|
|
"learning_rate": 3.923240938166312e-05,
|
|
"loss": 0.0313,
|
|
"step": 505
|
|
},
|
|
{
|
|
"epoch": 1.0874200426439233,
|
|
"grad_norm": 7.312335014343262,
|
|
"learning_rate": 3.9125799573560765e-05,
|
|
"loss": 0.0846,
|
|
"step": 510
|
|
},
|
|
{
|
|
"epoch": 1.0980810234541578,
|
|
"grad_norm": 2.763037919998169,
|
|
"learning_rate": 3.901918976545843e-05,
|
|
"loss": 0.0513,
|
|
"step": 515
|
|
},
|
|
{
|
|
"epoch": 1.1087420042643923,
|
|
"grad_norm": 0.4378064274787903,
|
|
"learning_rate": 3.8912579957356075e-05,
|
|
"loss": 0.0024,
|
|
"step": 520
|
|
},
|
|
{
|
|
"epoch": 1.1194029850746268,
|
|
"grad_norm": 0.37240126729011536,
|
|
"learning_rate": 3.8805970149253736e-05,
|
|
"loss": 0.0015,
|
|
"step": 525
|
|
},
|
|
{
|
|
"epoch": 1.1300639658848615,
|
|
"grad_norm": 0.02618749998509884,
|
|
"learning_rate": 3.8699360341151384e-05,
|
|
"loss": 0.0145,
|
|
"step": 530
|
|
},
|
|
{
|
|
"epoch": 1.140724946695096,
|
|
"grad_norm": 0.013713640160858631,
|
|
"learning_rate": 3.8592750533049046e-05,
|
|
"loss": 0.042,
|
|
"step": 535
|
|
},
|
|
{
|
|
"epoch": 1.1513859275053304,
|
|
"grad_norm": 6.348232269287109,
|
|
"learning_rate": 3.8486140724946694e-05,
|
|
"loss": 0.0765,
|
|
"step": 540
|
|
},
|
|
{
|
|
"epoch": 1.1620469083155651,
|
|
"grad_norm": 0.2921755015850067,
|
|
"learning_rate": 3.8379530916844355e-05,
|
|
"loss": 0.0032,
|
|
"step": 545
|
|
},
|
|
{
|
|
"epoch": 1.1727078891257996,
|
|
"grad_norm": 0.028481731191277504,
|
|
"learning_rate": 3.8272921108742e-05,
|
|
"loss": 0.0114,
|
|
"step": 550
|
|
},
|
|
{
|
|
"epoch": 1.183368869936034,
|
|
"grad_norm": 0.2265009731054306,
|
|
"learning_rate": 3.8166311300639665e-05,
|
|
"loss": 0.005,
|
|
"step": 555
|
|
},
|
|
{
|
|
"epoch": 1.1940298507462686,
|
|
"grad_norm": 3.0314130783081055,
|
|
"learning_rate": 3.805970149253731e-05,
|
|
"loss": 0.0331,
|
|
"step": 560
|
|
},
|
|
{
|
|
"epoch": 1.2046908315565032,
|
|
"grad_norm": 0.7699183821678162,
|
|
"learning_rate": 3.7953091684434974e-05,
|
|
"loss": 0.0086,
|
|
"step": 565
|
|
},
|
|
{
|
|
"epoch": 1.2153518123667377,
|
|
"grad_norm": 0.0020996136590838432,
|
|
"learning_rate": 3.784648187633262e-05,
|
|
"loss": 0.0101,
|
|
"step": 570
|
|
},
|
|
{
|
|
"epoch": 1.2260127931769722,
|
|
"grad_norm": 0.049583759158849716,
|
|
"learning_rate": 3.7739872068230284e-05,
|
|
"loss": 0.0293,
|
|
"step": 575
|
|
},
|
|
{
|
|
"epoch": 1.236673773987207,
|
|
"grad_norm": 11.139657974243164,
|
|
"learning_rate": 3.763326226012793e-05,
|
|
"loss": 0.0087,
|
|
"step": 580
|
|
},
|
|
{
|
|
"epoch": 1.2473347547974414,
|
|
"grad_norm": 0.0236153993755579,
|
|
"learning_rate": 3.752665245202559e-05,
|
|
"loss": 0.0252,
|
|
"step": 585
|
|
},
|
|
{
|
|
"epoch": 1.2579957356076759,
|
|
"grad_norm": 5.523359775543213,
|
|
"learning_rate": 3.742004264392324e-05,
|
|
"loss": 0.0212,
|
|
"step": 590
|
|
},
|
|
{
|
|
"epoch": 1.2686567164179103,
|
|
"grad_norm": 0.021226538345217705,
|
|
"learning_rate": 3.73134328358209e-05,
|
|
"loss": 0.0014,
|
|
"step": 595
|
|
},
|
|
{
|
|
"epoch": 1.279317697228145,
|
|
"grad_norm": 0.0022214530035853386,
|
|
"learning_rate": 3.720682302771855e-05,
|
|
"loss": 0.0226,
|
|
"step": 600
|
|
},
|
|
{
|
|
"epoch": 1.2899786780383795,
|
|
"grad_norm": 0.6694089770317078,
|
|
"learning_rate": 3.710021321961621e-05,
|
|
"loss": 0.0023,
|
|
"step": 605
|
|
},
|
|
{
|
|
"epoch": 1.3006396588486142,
|
|
"grad_norm": 1.18583083152771,
|
|
"learning_rate": 3.699360341151386e-05,
|
|
"loss": 0.0024,
|
|
"step": 610
|
|
},
|
|
{
|
|
"epoch": 1.3113006396588487,
|
|
"grad_norm": 0.03705647215247154,
|
|
"learning_rate": 3.6886993603411515e-05,
|
|
"loss": 0.045,
|
|
"step": 615
|
|
},
|
|
{
|
|
"epoch": 1.3219616204690832,
|
|
"grad_norm": 0.0837084949016571,
|
|
"learning_rate": 3.678038379530917e-05,
|
|
"loss": 0.0047,
|
|
"step": 620
|
|
},
|
|
{
|
|
"epoch": 1.3326226012793176,
|
|
"grad_norm": 20.727052688598633,
|
|
"learning_rate": 3.6673773987206824e-05,
|
|
"loss": 0.0889,
|
|
"step": 625
|
|
},
|
|
{
|
|
"epoch": 1.3432835820895521,
|
|
"grad_norm": 0.22105655074119568,
|
|
"learning_rate": 3.656716417910448e-05,
|
|
"loss": 0.0145,
|
|
"step": 630
|
|
},
|
|
{
|
|
"epoch": 1.3539445628997868,
|
|
"grad_norm": 0.25451749563217163,
|
|
"learning_rate": 3.6460554371002134e-05,
|
|
"loss": 0.0237,
|
|
"step": 635
|
|
},
|
|
{
|
|
"epoch": 1.3646055437100213,
|
|
"grad_norm": 0.5099952220916748,
|
|
"learning_rate": 3.635394456289979e-05,
|
|
"loss": 0.0283,
|
|
"step": 640
|
|
},
|
|
{
|
|
"epoch": 1.375266524520256,
|
|
"grad_norm": 11.225927352905273,
|
|
"learning_rate": 3.624733475479744e-05,
|
|
"loss": 0.0439,
|
|
"step": 645
|
|
},
|
|
{
|
|
"epoch": 1.3859275053304905,
|
|
"grad_norm": 0.03543882444500923,
|
|
"learning_rate": 3.61407249466951e-05,
|
|
"loss": 0.029,
|
|
"step": 650
|
|
},
|
|
{
|
|
"epoch": 1.396588486140725,
|
|
"grad_norm": 0.300640732049942,
|
|
"learning_rate": 3.603411513859275e-05,
|
|
"loss": 0.0404,
|
|
"step": 655
|
|
},
|
|
{
|
|
"epoch": 1.4072494669509594,
|
|
"grad_norm": 0.09688057750463486,
|
|
"learning_rate": 3.592750533049041e-05,
|
|
"loss": 0.0141,
|
|
"step": 660
|
|
},
|
|
{
|
|
"epoch": 1.417910447761194,
|
|
"grad_norm": 0.15651468932628632,
|
|
"learning_rate": 3.582089552238806e-05,
|
|
"loss": 0.0014,
|
|
"step": 665
|
|
},
|
|
{
|
|
"epoch": 1.4285714285714286,
|
|
"grad_norm": 16.413286209106445,
|
|
"learning_rate": 3.571428571428572e-05,
|
|
"loss": 0.061,
|
|
"step": 670
|
|
},
|
|
{
|
|
"epoch": 1.439232409381663,
|
|
"grad_norm": 0.009997420944273472,
|
|
"learning_rate": 3.560767590618337e-05,
|
|
"loss": 0.0032,
|
|
"step": 675
|
|
},
|
|
{
|
|
"epoch": 1.4498933901918978,
|
|
"grad_norm": 0.11960610747337341,
|
|
"learning_rate": 3.5501066098081026e-05,
|
|
"loss": 0.0698,
|
|
"step": 680
|
|
},
|
|
{
|
|
"epoch": 1.4605543710021323,
|
|
"grad_norm": 0.18982942402362823,
|
|
"learning_rate": 3.539445628997868e-05,
|
|
"loss": 0.0512,
|
|
"step": 685
|
|
},
|
|
{
|
|
"epoch": 1.4712153518123667,
|
|
"grad_norm": 6.774919033050537,
|
|
"learning_rate": 3.5287846481876336e-05,
|
|
"loss": 0.068,
|
|
"step": 690
|
|
},
|
|
{
|
|
"epoch": 1.4818763326226012,
|
|
"grad_norm": 0.07752835005521774,
|
|
"learning_rate": 3.518123667377399e-05,
|
|
"loss": 0.0137,
|
|
"step": 695
|
|
},
|
|
{
|
|
"epoch": 1.4925373134328357,
|
|
"grad_norm": 4.090426445007324,
|
|
"learning_rate": 3.5074626865671645e-05,
|
|
"loss": 0.0177,
|
|
"step": 700
|
|
},
|
|
{
|
|
"epoch": 1.5031982942430704,
|
|
"grad_norm": 0.6071398854255676,
|
|
"learning_rate": 3.496801705756929e-05,
|
|
"loss": 0.0766,
|
|
"step": 705
|
|
},
|
|
{
|
|
"epoch": 1.5138592750533049,
|
|
"grad_norm": 0.14962276816368103,
|
|
"learning_rate": 3.4861407249466955e-05,
|
|
"loss": 0.0136,
|
|
"step": 710
|
|
},
|
|
{
|
|
"epoch": 1.5245202558635396,
|
|
"grad_norm": 0.09204737842082977,
|
|
"learning_rate": 3.47547974413646e-05,
|
|
"loss": 0.0494,
|
|
"step": 715
|
|
},
|
|
{
|
|
"epoch": 1.535181236673774,
|
|
"grad_norm": 7.531642913818359,
|
|
"learning_rate": 3.4648187633262264e-05,
|
|
"loss": 0.0161,
|
|
"step": 720
|
|
},
|
|
{
|
|
"epoch": 1.5458422174840085,
|
|
"grad_norm": 6.497248649597168,
|
|
"learning_rate": 3.454157782515991e-05,
|
|
"loss": 0.0414,
|
|
"step": 725
|
|
},
|
|
{
|
|
"epoch": 1.556503198294243,
|
|
"grad_norm": 0.034405265003442764,
|
|
"learning_rate": 3.4434968017057574e-05,
|
|
"loss": 0.0188,
|
|
"step": 730
|
|
},
|
|
{
|
|
"epoch": 1.5671641791044775,
|
|
"grad_norm": 1.0338650941848755,
|
|
"learning_rate": 3.432835820895522e-05,
|
|
"loss": 0.0088,
|
|
"step": 735
|
|
},
|
|
{
|
|
"epoch": 1.5778251599147122,
|
|
"grad_norm": 0.007859860546886921,
|
|
"learning_rate": 3.422174840085288e-05,
|
|
"loss": 0.0008,
|
|
"step": 740
|
|
},
|
|
{
|
|
"epoch": 1.5884861407249466,
|
|
"grad_norm": 9.049701690673828,
|
|
"learning_rate": 3.411513859275053e-05,
|
|
"loss": 0.0313,
|
|
"step": 745
|
|
},
|
|
{
|
|
"epoch": 1.5991471215351813,
|
|
"grad_norm": 0.3167734444141388,
|
|
"learning_rate": 3.400852878464819e-05,
|
|
"loss": 0.0304,
|
|
"step": 750
|
|
},
|
|
{
|
|
"epoch": 1.6098081023454158,
|
|
"grad_norm": 0.03937331959605217,
|
|
"learning_rate": 3.390191897654584e-05,
|
|
"loss": 0.0081,
|
|
"step": 755
|
|
},
|
|
{
|
|
"epoch": 1.6204690831556503,
|
|
"grad_norm": 0.39047786593437195,
|
|
"learning_rate": 3.37953091684435e-05,
|
|
"loss": 0.0249,
|
|
"step": 760
|
|
},
|
|
{
|
|
"epoch": 1.6311300639658848,
|
|
"grad_norm": 0.007541940547525883,
|
|
"learning_rate": 3.368869936034115e-05,
|
|
"loss": 0.0004,
|
|
"step": 765
|
|
},
|
|
{
|
|
"epoch": 1.6417910447761193,
|
|
"grad_norm": 0.5652676820755005,
|
|
"learning_rate": 3.358208955223881e-05,
|
|
"loss": 0.0015,
|
|
"step": 770
|
|
},
|
|
{
|
|
"epoch": 1.652452025586354,
|
|
"grad_norm": 0.6932991743087769,
|
|
"learning_rate": 3.347547974413646e-05,
|
|
"loss": 0.0009,
|
|
"step": 775
|
|
},
|
|
{
|
|
"epoch": 1.6631130063965884,
|
|
"grad_norm": 10.74230670928955,
|
|
"learning_rate": 3.336886993603412e-05,
|
|
"loss": 0.0427,
|
|
"step": 780
|
|
},
|
|
{
|
|
"epoch": 1.6737739872068231,
|
|
"grad_norm": 0.015600375831127167,
|
|
"learning_rate": 3.326226012793177e-05,
|
|
"loss": 0.0908,
|
|
"step": 785
|
|
},
|
|
{
|
|
"epoch": 1.6844349680170576,
|
|
"grad_norm": 0.032702818512916565,
|
|
"learning_rate": 3.3155650319829424e-05,
|
|
"loss": 0.0105,
|
|
"step": 790
|
|
},
|
|
{
|
|
"epoch": 1.695095948827292,
|
|
"grad_norm": 0.011013907380402088,
|
|
"learning_rate": 3.304904051172708e-05,
|
|
"loss": 0.0115,
|
|
"step": 795
|
|
},
|
|
{
|
|
"epoch": 1.7057569296375266,
|
|
"grad_norm": 0.4279489517211914,
|
|
"learning_rate": 3.294243070362473e-05,
|
|
"loss": 0.0425,
|
|
"step": 800
|
|
},
|
|
{
|
|
"epoch": 1.716417910447761,
|
|
"grad_norm": 0.6606729030609131,
|
|
"learning_rate": 3.283582089552239e-05,
|
|
"loss": 0.0161,
|
|
"step": 805
|
|
},
|
|
{
|
|
"epoch": 1.7270788912579957,
|
|
"grad_norm": 0.06445536762475967,
|
|
"learning_rate": 3.272921108742004e-05,
|
|
"loss": 0.0148,
|
|
"step": 810
|
|
},
|
|
{
|
|
"epoch": 1.7377398720682304,
|
|
"grad_norm": 8.382514953613281,
|
|
"learning_rate": 3.26226012793177e-05,
|
|
"loss": 0.0202,
|
|
"step": 815
|
|
},
|
|
{
|
|
"epoch": 1.748400852878465,
|
|
"grad_norm": 0.0198732428252697,
|
|
"learning_rate": 3.251599147121535e-05,
|
|
"loss": 0.0426,
|
|
"step": 820
|
|
},
|
|
{
|
|
"epoch": 1.7590618336886994,
|
|
"grad_norm": 4.975524425506592,
|
|
"learning_rate": 3.240938166311301e-05,
|
|
"loss": 0.077,
|
|
"step": 825
|
|
},
|
|
{
|
|
"epoch": 1.7697228144989339,
|
|
"grad_norm": 9.508905410766602,
|
|
"learning_rate": 3.230277185501066e-05,
|
|
"loss": 0.0164,
|
|
"step": 830
|
|
},
|
|
{
|
|
"epoch": 1.7803837953091683,
|
|
"grad_norm": 14.851171493530273,
|
|
"learning_rate": 3.2196162046908317e-05,
|
|
"loss": 0.0194,
|
|
"step": 835
|
|
},
|
|
{
|
|
"epoch": 1.7910447761194028,
|
|
"grad_norm": 0.01751009188592434,
|
|
"learning_rate": 3.208955223880597e-05,
|
|
"loss": 0.0014,
|
|
"step": 840
|
|
},
|
|
{
|
|
"epoch": 1.8017057569296375,
|
|
"grad_norm": 0.02330157533288002,
|
|
"learning_rate": 3.1982942430703626e-05,
|
|
"loss": 0.002,
|
|
"step": 845
|
|
},
|
|
{
|
|
"epoch": 1.8123667377398722,
|
|
"grad_norm": 0.02951938658952713,
|
|
"learning_rate": 3.187633262260128e-05,
|
|
"loss": 0.0015,
|
|
"step": 850
|
|
},
|
|
{
|
|
"epoch": 1.8230277185501067,
|
|
"grad_norm": 0.0671420693397522,
|
|
"learning_rate": 3.1769722814498935e-05,
|
|
"loss": 0.066,
|
|
"step": 855
|
|
},
|
|
{
|
|
"epoch": 1.8336886993603412,
|
|
"grad_norm": 6.726079940795898,
|
|
"learning_rate": 3.166311300639659e-05,
|
|
"loss": 0.0298,
|
|
"step": 860
|
|
},
|
|
{
|
|
"epoch": 1.8443496801705757,
|
|
"grad_norm": 5.813134670257568,
|
|
"learning_rate": 3.1556503198294245e-05,
|
|
"loss": 0.0225,
|
|
"step": 865
|
|
},
|
|
{
|
|
"epoch": 1.8550106609808101,
|
|
"grad_norm": 0.009066685102880001,
|
|
"learning_rate": 3.14498933901919e-05,
|
|
"loss": 0.0033,
|
|
"step": 870
|
|
},
|
|
{
|
|
"epoch": 1.8656716417910446,
|
|
"grad_norm": 0.13737043738365173,
|
|
"learning_rate": 3.1343283582089554e-05,
|
|
"loss": 0.0009,
|
|
"step": 875
|
|
},
|
|
{
|
|
"epoch": 1.8763326226012793,
|
|
"grad_norm": 8.68946647644043,
|
|
"learning_rate": 3.123667377398721e-05,
|
|
"loss": 0.0107,
|
|
"step": 880
|
|
},
|
|
{
|
|
"epoch": 1.886993603411514,
|
|
"grad_norm": 1.1588736772537231,
|
|
"learning_rate": 3.1130063965884864e-05,
|
|
"loss": 0.0018,
|
|
"step": 885
|
|
},
|
|
{
|
|
"epoch": 1.8976545842217485,
|
|
"grad_norm": 7.715778827667236,
|
|
"learning_rate": 3.102345415778252e-05,
|
|
"loss": 0.0773,
|
|
"step": 890
|
|
},
|
|
{
|
|
"epoch": 1.908315565031983,
|
|
"grad_norm": 7.597550392150879,
|
|
"learning_rate": 3.0916844349680173e-05,
|
|
"loss": 0.0232,
|
|
"step": 895
|
|
},
|
|
{
|
|
"epoch": 1.9189765458422174,
|
|
"grad_norm": 0.0018281742231920362,
|
|
"learning_rate": 3.081023454157783e-05,
|
|
"loss": 0.0043,
|
|
"step": 900
|
|
},
|
|
{
|
|
"epoch": 1.929637526652452,
|
|
"grad_norm": 0.057870108634233475,
|
|
"learning_rate": 3.070362473347548e-05,
|
|
"loss": 0.0003,
|
|
"step": 905
|
|
},
|
|
{
|
|
"epoch": 1.9402985074626866,
|
|
"grad_norm": 0.06244092807173729,
|
|
"learning_rate": 3.059701492537314e-05,
|
|
"loss": 0.012,
|
|
"step": 910
|
|
},
|
|
{
|
|
"epoch": 1.950959488272921,
|
|
"grad_norm": 0.005397165194153786,
|
|
"learning_rate": 3.0490405117270792e-05,
|
|
"loss": 0.0007,
|
|
"step": 915
|
|
},
|
|
{
|
|
"epoch": 1.9616204690831558,
|
|
"grad_norm": 0.13206999003887177,
|
|
"learning_rate": 3.0383795309168444e-05,
|
|
"loss": 0.0005,
|
|
"step": 920
|
|
},
|
|
{
|
|
"epoch": 1.9722814498933903,
|
|
"grad_norm": 17.099491119384766,
|
|
"learning_rate": 3.0277185501066102e-05,
|
|
"loss": 0.0543,
|
|
"step": 925
|
|
},
|
|
{
|
|
"epoch": 1.9829424307036247,
|
|
"grad_norm": 8.941192626953125,
|
|
"learning_rate": 3.0170575692963753e-05,
|
|
"loss": 0.0189,
|
|
"step": 930
|
|
},
|
|
{
|
|
"epoch": 1.9936034115138592,
|
|
"grad_norm": 0.5419250130653381,
|
|
"learning_rate": 3.006396588486141e-05,
|
|
"loss": 0.0215,
|
|
"step": 935
|
|
},
|
|
{
|
|
"epoch": 2.0,
|
|
"eval_accuracy": 0.952,
|
|
"eval_loss": 0.23366907238960266,
|
|
"eval_runtime": 17.9114,
|
|
"eval_samples_per_second": 209.364,
|
|
"eval_steps_per_second": 6.588,
|
|
"step": 938
|
|
},
|
|
{
|
|
"epoch": 2.0042643923240937,
|
|
"grad_norm": 0.23212935030460358,
|
|
"learning_rate": 2.9957356076759063e-05,
|
|
"loss": 0.1364,
|
|
"step": 940
|
|
},
|
|
{
|
|
"epoch": 2.014925373134328,
|
|
"grad_norm": 0.006510720122605562,
|
|
"learning_rate": 2.9850746268656714e-05,
|
|
"loss": 0.0002,
|
|
"step": 945
|
|
},
|
|
{
|
|
"epoch": 2.025586353944563,
|
|
"grad_norm": 0.2070126086473465,
|
|
"learning_rate": 2.9744136460554372e-05,
|
|
"loss": 0.0054,
|
|
"step": 950
|
|
},
|
|
{
|
|
"epoch": 2.0362473347547976,
|
|
"grad_norm": 0.0019345590844750404,
|
|
"learning_rate": 2.9637526652452023e-05,
|
|
"loss": 0.0018,
|
|
"step": 955
|
|
},
|
|
{
|
|
"epoch": 2.046908315565032,
|
|
"grad_norm": 0.04831007495522499,
|
|
"learning_rate": 2.953091684434968e-05,
|
|
"loss": 0.0059,
|
|
"step": 960
|
|
},
|
|
{
|
|
"epoch": 2.0575692963752665,
|
|
"grad_norm": 0.012780869379639626,
|
|
"learning_rate": 2.9424307036247333e-05,
|
|
"loss": 0.0006,
|
|
"step": 965
|
|
},
|
|
{
|
|
"epoch": 2.068230277185501,
|
|
"grad_norm": 0.006146378815174103,
|
|
"learning_rate": 2.931769722814499e-05,
|
|
"loss": 0.0008,
|
|
"step": 970
|
|
},
|
|
{
|
|
"epoch": 2.0788912579957355,
|
|
"grad_norm": 0.0021279011853039265,
|
|
"learning_rate": 2.9211087420042642e-05,
|
|
"loss": 0.0001,
|
|
"step": 975
|
|
},
|
|
{
|
|
"epoch": 2.08955223880597,
|
|
"grad_norm": 7.1951584815979,
|
|
"learning_rate": 2.91044776119403e-05,
|
|
"loss": 0.0273,
|
|
"step": 980
|
|
},
|
|
{
|
|
"epoch": 2.100213219616205,
|
|
"grad_norm": 0.006574519444257021,
|
|
"learning_rate": 2.8997867803837952e-05,
|
|
"loss": 0.0019,
|
|
"step": 985
|
|
},
|
|
{
|
|
"epoch": 2.1108742004264394,
|
|
"grad_norm": 0.02979465015232563,
|
|
"learning_rate": 2.889125799573561e-05,
|
|
"loss": 0.0005,
|
|
"step": 990
|
|
},
|
|
{
|
|
"epoch": 2.121535181236674,
|
|
"grad_norm": 0.03725764900445938,
|
|
"learning_rate": 2.878464818763326e-05,
|
|
"loss": 0.0071,
|
|
"step": 995
|
|
},
|
|
{
|
|
"epoch": 2.1321961620469083,
|
|
"grad_norm": 9.669961929321289,
|
|
"learning_rate": 2.867803837953092e-05,
|
|
"loss": 0.0297,
|
|
"step": 1000
|
|
},
|
|
{
|
|
"epoch": 2.142857142857143,
|
|
"grad_norm": 0.05215339735150337,
|
|
"learning_rate": 2.857142857142857e-05,
|
|
"loss": 0.0003,
|
|
"step": 1005
|
|
},
|
|
{
|
|
"epoch": 2.1535181236673773,
|
|
"grad_norm": 0.005723474081605673,
|
|
"learning_rate": 2.846481876332623e-05,
|
|
"loss": 0.0005,
|
|
"step": 1010
|
|
},
|
|
{
|
|
"epoch": 2.1641791044776117,
|
|
"grad_norm": 0.12211333960294724,
|
|
"learning_rate": 2.835820895522388e-05,
|
|
"loss": 0.0003,
|
|
"step": 1015
|
|
},
|
|
{
|
|
"epoch": 2.1748400852878467,
|
|
"grad_norm": 0.26316601037979126,
|
|
"learning_rate": 2.825159914712154e-05,
|
|
"loss": 0.0008,
|
|
"step": 1020
|
|
},
|
|
{
|
|
"epoch": 2.185501066098081,
|
|
"grad_norm": 0.001690951525233686,
|
|
"learning_rate": 2.814498933901919e-05,
|
|
"loss": 0.0021,
|
|
"step": 1025
|
|
},
|
|
{
|
|
"epoch": 2.1961620469083156,
|
|
"grad_norm": 0.0034065067302435637,
|
|
"learning_rate": 2.8038379530916848e-05,
|
|
"loss": 0.0004,
|
|
"step": 1030
|
|
},
|
|
{
|
|
"epoch": 2.20682302771855,
|
|
"grad_norm": 0.0010407333029434085,
|
|
"learning_rate": 2.79317697228145e-05,
|
|
"loss": 0.0254,
|
|
"step": 1035
|
|
},
|
|
{
|
|
"epoch": 2.2174840085287846,
|
|
"grad_norm": 0.07637695968151093,
|
|
"learning_rate": 2.7825159914712157e-05,
|
|
"loss": 0.0237,
|
|
"step": 1040
|
|
},
|
|
{
|
|
"epoch": 2.228144989339019,
|
|
"grad_norm": 0.002068418078124523,
|
|
"learning_rate": 2.771855010660981e-05,
|
|
"loss": 0.0002,
|
|
"step": 1045
|
|
},
|
|
{
|
|
"epoch": 2.2388059701492535,
|
|
"grad_norm": 0.003188133705407381,
|
|
"learning_rate": 2.7611940298507467e-05,
|
|
"loss": 0.0002,
|
|
"step": 1050
|
|
},
|
|
{
|
|
"epoch": 2.2494669509594885,
|
|
"grad_norm": 2.3322396278381348,
|
|
"learning_rate": 2.7505330490405118e-05,
|
|
"loss": 0.008,
|
|
"step": 1055
|
|
},
|
|
{
|
|
"epoch": 2.260127931769723,
|
|
"grad_norm": 0.007329651620239019,
|
|
"learning_rate": 2.7398720682302776e-05,
|
|
"loss": 0.0002,
|
|
"step": 1060
|
|
},
|
|
{
|
|
"epoch": 2.2707889125799574,
|
|
"grad_norm": 0.02505979686975479,
|
|
"learning_rate": 2.7292110874200428e-05,
|
|
"loss": 0.0002,
|
|
"step": 1065
|
|
},
|
|
{
|
|
"epoch": 2.281449893390192,
|
|
"grad_norm": 0.0007982216193340719,
|
|
"learning_rate": 2.7185501066098086e-05,
|
|
"loss": 0.0001,
|
|
"step": 1070
|
|
},
|
|
{
|
|
"epoch": 2.2921108742004264,
|
|
"grad_norm": 0.009859717451035976,
|
|
"learning_rate": 2.7078891257995737e-05,
|
|
"loss": 0.004,
|
|
"step": 1075
|
|
},
|
|
{
|
|
"epoch": 2.302771855010661,
|
|
"grad_norm": 0.031057290732860565,
|
|
"learning_rate": 2.6972281449893395e-05,
|
|
"loss": 0.0008,
|
|
"step": 1080
|
|
},
|
|
{
|
|
"epoch": 2.3134328358208958,
|
|
"grad_norm": 0.002461077645421028,
|
|
"learning_rate": 2.6865671641791047e-05,
|
|
"loss": 0.0001,
|
|
"step": 1085
|
|
},
|
|
{
|
|
"epoch": 2.3240938166311302,
|
|
"grad_norm": 0.0019033419666811824,
|
|
"learning_rate": 2.6759061833688705e-05,
|
|
"loss": 0.0005,
|
|
"step": 1090
|
|
},
|
|
{
|
|
"epoch": 2.3347547974413647,
|
|
"grad_norm": 0.020137565210461617,
|
|
"learning_rate": 2.6652452025586356e-05,
|
|
"loss": 0.0001,
|
|
"step": 1095
|
|
},
|
|
{
|
|
"epoch": 2.345415778251599,
|
|
"grad_norm": 0.01116106379777193,
|
|
"learning_rate": 2.6545842217484007e-05,
|
|
"loss": 0.0002,
|
|
"step": 1100
|
|
},
|
|
{
|
|
"epoch": 2.3560767590618337,
|
|
"grad_norm": 0.06040339916944504,
|
|
"learning_rate": 2.6439232409381666e-05,
|
|
"loss": 0.0002,
|
|
"step": 1105
|
|
},
|
|
{
|
|
"epoch": 2.366737739872068,
|
|
"grad_norm": 1.5732898712158203,
|
|
"learning_rate": 2.6332622601279317e-05,
|
|
"loss": 0.0032,
|
|
"step": 1110
|
|
},
|
|
{
|
|
"epoch": 2.3773987206823026,
|
|
"grad_norm": 0.007247257046401501,
|
|
"learning_rate": 2.6226012793176975e-05,
|
|
"loss": 0.0002,
|
|
"step": 1115
|
|
},
|
|
{
|
|
"epoch": 2.388059701492537,
|
|
"grad_norm": 0.021347081288695335,
|
|
"learning_rate": 2.6119402985074626e-05,
|
|
"loss": 0.0212,
|
|
"step": 1120
|
|
},
|
|
{
|
|
"epoch": 2.398720682302772,
|
|
"grad_norm": 1.4401729106903076,
|
|
"learning_rate": 2.6012793176972285e-05,
|
|
"loss": 0.0011,
|
|
"step": 1125
|
|
},
|
|
{
|
|
"epoch": 2.4093816631130065,
|
|
"grad_norm": 0.009736673906445503,
|
|
"learning_rate": 2.5906183368869936e-05,
|
|
"loss": 0.0001,
|
|
"step": 1130
|
|
},
|
|
{
|
|
"epoch": 2.420042643923241,
|
|
"grad_norm": 0.0005613169632852077,
|
|
"learning_rate": 2.5799573560767594e-05,
|
|
"loss": 0.0038,
|
|
"step": 1135
|
|
},
|
|
{
|
|
"epoch": 2.4307036247334755,
|
|
"grad_norm": 0.006597707513719797,
|
|
"learning_rate": 2.5692963752665245e-05,
|
|
"loss": 0.0001,
|
|
"step": 1140
|
|
},
|
|
{
|
|
"epoch": 2.44136460554371,
|
|
"grad_norm": 0.5111905932426453,
|
|
"learning_rate": 2.5586353944562904e-05,
|
|
"loss": 0.0409,
|
|
"step": 1145
|
|
},
|
|
{
|
|
"epoch": 2.4520255863539444,
|
|
"grad_norm": 0.014620088040828705,
|
|
"learning_rate": 2.5479744136460555e-05,
|
|
"loss": 0.0003,
|
|
"step": 1150
|
|
},
|
|
{
|
|
"epoch": 2.4626865671641793,
|
|
"grad_norm": 0.5755995512008667,
|
|
"learning_rate": 2.537313432835821e-05,
|
|
"loss": 0.0006,
|
|
"step": 1155
|
|
},
|
|
{
|
|
"epoch": 2.473347547974414,
|
|
"grad_norm": 0.014837886206805706,
|
|
"learning_rate": 2.5266524520255864e-05,
|
|
"loss": 0.0003,
|
|
"step": 1160
|
|
},
|
|
{
|
|
"epoch": 2.4840085287846483,
|
|
"grad_norm": 0.0005985779571346939,
|
|
"learning_rate": 2.515991471215352e-05,
|
|
"loss": 0.0001,
|
|
"step": 1165
|
|
},
|
|
{
|
|
"epoch": 2.4946695095948828,
|
|
"grad_norm": 0.0027303595561534166,
|
|
"learning_rate": 2.5053304904051174e-05,
|
|
"loss": 0.0001,
|
|
"step": 1170
|
|
},
|
|
{
|
|
"epoch": 2.5053304904051172,
|
|
"grad_norm": 0.16521170735359192,
|
|
"learning_rate": 2.494669509594883e-05,
|
|
"loss": 0.0105,
|
|
"step": 1175
|
|
},
|
|
{
|
|
"epoch": 2.5159914712153517,
|
|
"grad_norm": 0.004499639384448528,
|
|
"learning_rate": 2.4840085287846483e-05,
|
|
"loss": 0.0001,
|
|
"step": 1180
|
|
},
|
|
{
|
|
"epoch": 2.526652452025586,
|
|
"grad_norm": 0.026896901428699493,
|
|
"learning_rate": 2.4733475479744138e-05,
|
|
"loss": 0.0026,
|
|
"step": 1185
|
|
},
|
|
{
|
|
"epoch": 2.5373134328358207,
|
|
"grad_norm": 0.5625950694084167,
|
|
"learning_rate": 2.4626865671641793e-05,
|
|
"loss": 0.0006,
|
|
"step": 1190
|
|
},
|
|
{
|
|
"epoch": 2.5479744136460556,
|
|
"grad_norm": 1.4240341186523438,
|
|
"learning_rate": 2.4520255863539444e-05,
|
|
"loss": 0.0161,
|
|
"step": 1195
|
|
},
|
|
{
|
|
"epoch": 2.55863539445629,
|
|
"grad_norm": 0.0059285759925842285,
|
|
"learning_rate": 2.44136460554371e-05,
|
|
"loss": 0.0001,
|
|
"step": 1200
|
|
},
|
|
{
|
|
"epoch": 2.5692963752665245,
|
|
"grad_norm": 3.4125325679779053,
|
|
"learning_rate": 2.4307036247334754e-05,
|
|
"loss": 0.0016,
|
|
"step": 1205
|
|
},
|
|
{
|
|
"epoch": 2.579957356076759,
|
|
"grad_norm": 0.09839680790901184,
|
|
"learning_rate": 2.420042643923241e-05,
|
|
"loss": 0.0004,
|
|
"step": 1210
|
|
},
|
|
{
|
|
"epoch": 2.5906183368869935,
|
|
"grad_norm": 0.0010177026269957423,
|
|
"learning_rate": 2.4093816631130063e-05,
|
|
"loss": 0.001,
|
|
"step": 1215
|
|
},
|
|
{
|
|
"epoch": 2.6012793176972284,
|
|
"grad_norm": 0.002969444030895829,
|
|
"learning_rate": 2.3987206823027718e-05,
|
|
"loss": 0.0003,
|
|
"step": 1220
|
|
},
|
|
{
|
|
"epoch": 2.611940298507463,
|
|
"grad_norm": 0.0015349864261224866,
|
|
"learning_rate": 2.3880597014925373e-05,
|
|
"loss": 0.0004,
|
|
"step": 1225
|
|
},
|
|
{
|
|
"epoch": 2.6226012793176974,
|
|
"grad_norm": 0.0008386714034713805,
|
|
"learning_rate": 2.3773987206823027e-05,
|
|
"loss": 0.0001,
|
|
"step": 1230
|
|
},
|
|
{
|
|
"epoch": 2.633262260127932,
|
|
"grad_norm": 0.0006430671201087534,
|
|
"learning_rate": 2.3667377398720682e-05,
|
|
"loss": 0.0001,
|
|
"step": 1235
|
|
},
|
|
{
|
|
"epoch": 2.6439232409381663,
|
|
"grad_norm": 0.021702909842133522,
|
|
"learning_rate": 2.3560767590618337e-05,
|
|
"loss": 0.002,
|
|
"step": 1240
|
|
},
|
|
{
|
|
"epoch": 2.654584221748401,
|
|
"grad_norm": 0.0046903654001653194,
|
|
"learning_rate": 2.345415778251599e-05,
|
|
"loss": 0.0004,
|
|
"step": 1245
|
|
},
|
|
{
|
|
"epoch": 2.6652452025586353,
|
|
"grad_norm": 0.06749894469976425,
|
|
"learning_rate": 2.3347547974413646e-05,
|
|
"loss": 0.0002,
|
|
"step": 1250
|
|
},
|
|
{
|
|
"epoch": 2.6759061833688698,
|
|
"grad_norm": 0.10783130675554276,
|
|
"learning_rate": 2.32409381663113e-05,
|
|
"loss": 0.0077,
|
|
"step": 1255
|
|
},
|
|
{
|
|
"epoch": 2.6865671641791042,
|
|
"grad_norm": 0.32774871587753296,
|
|
"learning_rate": 2.3134328358208956e-05,
|
|
"loss": 0.0004,
|
|
"step": 1260
|
|
},
|
|
{
|
|
"epoch": 2.697228144989339,
|
|
"grad_norm": 0.003545227460563183,
|
|
"learning_rate": 2.302771855010661e-05,
|
|
"loss": 0.0001,
|
|
"step": 1265
|
|
},
|
|
{
|
|
"epoch": 2.7078891257995736,
|
|
"grad_norm": 0.01444152370095253,
|
|
"learning_rate": 2.2921108742004265e-05,
|
|
"loss": 0.112,
|
|
"step": 1270
|
|
},
|
|
{
|
|
"epoch": 2.718550106609808,
|
|
"grad_norm": 0.0018640354974195361,
|
|
"learning_rate": 2.281449893390192e-05,
|
|
"loss": 0.0094,
|
|
"step": 1275
|
|
},
|
|
{
|
|
"epoch": 2.7292110874200426,
|
|
"grad_norm": 0.01793220080435276,
|
|
"learning_rate": 2.2707889125799575e-05,
|
|
"loss": 0.0274,
|
|
"step": 1280
|
|
},
|
|
{
|
|
"epoch": 2.739872068230277,
|
|
"grad_norm": 11.778922080993652,
|
|
"learning_rate": 2.260127931769723e-05,
|
|
"loss": 0.0365,
|
|
"step": 1285
|
|
},
|
|
{
|
|
"epoch": 2.750533049040512,
|
|
"grad_norm": 0.0011289332760497928,
|
|
"learning_rate": 2.2494669509594884e-05,
|
|
"loss": 0.0004,
|
|
"step": 1290
|
|
},
|
|
{
|
|
"epoch": 2.7611940298507465,
|
|
"grad_norm": 0.010555271059274673,
|
|
"learning_rate": 2.238805970149254e-05,
|
|
"loss": 0.0001,
|
|
"step": 1295
|
|
},
|
|
{
|
|
"epoch": 2.771855010660981,
|
|
"grad_norm": 0.0015511485980823636,
|
|
"learning_rate": 2.2281449893390194e-05,
|
|
"loss": 0.0002,
|
|
"step": 1300
|
|
},
|
|
{
|
|
"epoch": 2.7825159914712154,
|
|
"grad_norm": 0.0005575595423579216,
|
|
"learning_rate": 2.217484008528785e-05,
|
|
"loss": 0.0001,
|
|
"step": 1305
|
|
},
|
|
{
|
|
"epoch": 2.79317697228145,
|
|
"grad_norm": 0.1889764964580536,
|
|
"learning_rate": 2.2068230277185503e-05,
|
|
"loss": 0.0004,
|
|
"step": 1310
|
|
},
|
|
{
|
|
"epoch": 2.8038379530916844,
|
|
"grad_norm": 0.023335302248597145,
|
|
"learning_rate": 2.1961620469083158e-05,
|
|
"loss": 0.0004,
|
|
"step": 1315
|
|
},
|
|
{
|
|
"epoch": 2.814498933901919,
|
|
"grad_norm": 0.003003540216013789,
|
|
"learning_rate": 2.1855010660980813e-05,
|
|
"loss": 0.0001,
|
|
"step": 1320
|
|
},
|
|
{
|
|
"epoch": 2.8251599147121533,
|
|
"grad_norm": 0.028375841677188873,
|
|
"learning_rate": 2.1748400852878467e-05,
|
|
"loss": 0.0006,
|
|
"step": 1325
|
|
},
|
|
{
|
|
"epoch": 2.835820895522388,
|
|
"grad_norm": 0.012598455883562565,
|
|
"learning_rate": 2.164179104477612e-05,
|
|
"loss": 0.0001,
|
|
"step": 1330
|
|
},
|
|
{
|
|
"epoch": 2.8464818763326227,
|
|
"grad_norm": 0.003832779824733734,
|
|
"learning_rate": 2.1535181236673773e-05,
|
|
"loss": 0.0032,
|
|
"step": 1335
|
|
},
|
|
{
|
|
"epoch": 2.857142857142857,
|
|
"grad_norm": 0.0008711920818313956,
|
|
"learning_rate": 2.1428571428571428e-05,
|
|
"loss": 0.0021,
|
|
"step": 1340
|
|
},
|
|
{
|
|
"epoch": 2.8678038379530917,
|
|
"grad_norm": 0.0016271332278847694,
|
|
"learning_rate": 2.1321961620469083e-05,
|
|
"loss": 0.0001,
|
|
"step": 1345
|
|
},
|
|
{
|
|
"epoch": 2.878464818763326,
|
|
"grad_norm": 0.0009944208431988955,
|
|
"learning_rate": 2.1215351812366738e-05,
|
|
"loss": 0.0001,
|
|
"step": 1350
|
|
},
|
|
{
|
|
"epoch": 2.8891257995735606,
|
|
"grad_norm": 0.023777911439538002,
|
|
"learning_rate": 2.1108742004264392e-05,
|
|
"loss": 0.0426,
|
|
"step": 1355
|
|
},
|
|
{
|
|
"epoch": 2.8997867803837956,
|
|
"grad_norm": 0.8271000981330872,
|
|
"learning_rate": 2.1002132196162047e-05,
|
|
"loss": 0.033,
|
|
"step": 1360
|
|
},
|
|
{
|
|
"epoch": 2.91044776119403,
|
|
"grad_norm": 0.005003150552511215,
|
|
"learning_rate": 2.0895522388059702e-05,
|
|
"loss": 0.0044,
|
|
"step": 1365
|
|
},
|
|
{
|
|
"epoch": 2.9211087420042645,
|
|
"grad_norm": 0.03347320854663849,
|
|
"learning_rate": 2.0788912579957357e-05,
|
|
"loss": 0.0015,
|
|
"step": 1370
|
|
},
|
|
{
|
|
"epoch": 2.931769722814499,
|
|
"grad_norm": 0.000696173869073391,
|
|
"learning_rate": 2.068230277185501e-05,
|
|
"loss": 0.0001,
|
|
"step": 1375
|
|
},
|
|
{
|
|
"epoch": 2.9424307036247335,
|
|
"grad_norm": 19.805208206176758,
|
|
"learning_rate": 2.0575692963752666e-05,
|
|
"loss": 0.0099,
|
|
"step": 1380
|
|
},
|
|
{
|
|
"epoch": 2.953091684434968,
|
|
"grad_norm": 0.0019483945798128843,
|
|
"learning_rate": 2.046908315565032e-05,
|
|
"loss": 0.001,
|
|
"step": 1385
|
|
},
|
|
{
|
|
"epoch": 2.9637526652452024,
|
|
"grad_norm": 10.088868141174316,
|
|
"learning_rate": 2.0362473347547976e-05,
|
|
"loss": 0.0066,
|
|
"step": 1390
|
|
},
|
|
{
|
|
"epoch": 2.974413646055437,
|
|
"grad_norm": 0.003981050103902817,
|
|
"learning_rate": 2.025586353944563e-05,
|
|
"loss": 0.0001,
|
|
"step": 1395
|
|
},
|
|
{
|
|
"epoch": 2.9850746268656714,
|
|
"grad_norm": 0.0004683963197749108,
|
|
"learning_rate": 2.0149253731343285e-05,
|
|
"loss": 0.0008,
|
|
"step": 1400
|
|
},
|
|
{
|
|
"epoch": 2.9957356076759063,
|
|
"grad_norm": 5.499639987945557,
|
|
"learning_rate": 2.004264392324094e-05,
|
|
"loss": 0.0021,
|
|
"step": 1405
|
|
},
|
|
{
|
|
"epoch": 3.0,
|
|
"eval_accuracy": 0.9546666666666667,
|
|
"eval_loss": 0.248244509100914,
|
|
"eval_runtime": 18.2234,
|
|
"eval_samples_per_second": 205.78,
|
|
"eval_steps_per_second": 6.475,
|
|
"step": 1407
|
|
},
|
|
{
|
|
"epoch": 3.0063965884861408,
|
|
"grad_norm": 0.000801321177277714,
|
|
"learning_rate": 1.9936034115138594e-05,
|
|
"loss": 0.0635,
|
|
"step": 1410
|
|
},
|
|
{
|
|
"epoch": 3.0170575692963753,
|
|
"grad_norm": 0.0018451112555339932,
|
|
"learning_rate": 1.982942430703625e-05,
|
|
"loss": 0.0004,
|
|
"step": 1415
|
|
},
|
|
{
|
|
"epoch": 3.0277185501066097,
|
|
"grad_norm": 0.002699748147279024,
|
|
"learning_rate": 1.9722814498933904e-05,
|
|
"loss": 0.0001,
|
|
"step": 1420
|
|
},
|
|
{
|
|
"epoch": 3.038379530916844,
|
|
"grad_norm": 0.016597602516412735,
|
|
"learning_rate": 1.961620469083156e-05,
|
|
"loss": 0.0001,
|
|
"step": 1425
|
|
},
|
|
{
|
|
"epoch": 3.0490405117270787,
|
|
"grad_norm": 0.002888405928388238,
|
|
"learning_rate": 1.9509594882729213e-05,
|
|
"loss": 0.0001,
|
|
"step": 1430
|
|
},
|
|
{
|
|
"epoch": 3.0597014925373136,
|
|
"grad_norm": 0.001081654685549438,
|
|
"learning_rate": 1.9402985074626868e-05,
|
|
"loss": 0.0001,
|
|
"step": 1435
|
|
},
|
|
{
|
|
"epoch": 3.070362473347548,
|
|
"grad_norm": 0.01322456356137991,
|
|
"learning_rate": 1.9296375266524523e-05,
|
|
"loss": 0.0003,
|
|
"step": 1440
|
|
},
|
|
{
|
|
"epoch": 3.0810234541577826,
|
|
"grad_norm": 0.006288584321737289,
|
|
"learning_rate": 1.9189765458422178e-05,
|
|
"loss": 0.0001,
|
|
"step": 1445
|
|
},
|
|
{
|
|
"epoch": 3.091684434968017,
|
|
"grad_norm": 0.0008831746526993811,
|
|
"learning_rate": 1.9083155650319832e-05,
|
|
"loss": 0.0188,
|
|
"step": 1450
|
|
},
|
|
{
|
|
"epoch": 3.1023454157782515,
|
|
"grad_norm": 1.561693787574768,
|
|
"learning_rate": 1.8976545842217487e-05,
|
|
"loss": 0.0008,
|
|
"step": 1455
|
|
},
|
|
{
|
|
"epoch": 3.113006396588486,
|
|
"grad_norm": 0.001299654715694487,
|
|
"learning_rate": 1.8869936034115142e-05,
|
|
"loss": 0.0001,
|
|
"step": 1460
|
|
},
|
|
{
|
|
"epoch": 3.1236673773987205,
|
|
"grad_norm": 0.008705204352736473,
|
|
"learning_rate": 1.8763326226012797e-05,
|
|
"loss": 0.0001,
|
|
"step": 1465
|
|
},
|
|
{
|
|
"epoch": 3.1343283582089554,
|
|
"grad_norm": 0.017434099689126015,
|
|
"learning_rate": 1.865671641791045e-05,
|
|
"loss": 0.0001,
|
|
"step": 1470
|
|
},
|
|
{
|
|
"epoch": 3.14498933901919,
|
|
"grad_norm": 0.0013582018436864018,
|
|
"learning_rate": 1.8550106609808106e-05,
|
|
"loss": 0.0001,
|
|
"step": 1475
|
|
},
|
|
{
|
|
"epoch": 3.1556503198294243,
|
|
"grad_norm": 0.004880212247371674,
|
|
"learning_rate": 1.8443496801705757e-05,
|
|
"loss": 0.0023,
|
|
"step": 1480
|
|
},
|
|
{
|
|
"epoch": 3.166311300639659,
|
|
"grad_norm": 0.007500508800148964,
|
|
"learning_rate": 1.8336886993603412e-05,
|
|
"loss": 0.0001,
|
|
"step": 1485
|
|
},
|
|
{
|
|
"epoch": 3.1769722814498933,
|
|
"grad_norm": 0.002009419957175851,
|
|
"learning_rate": 1.8230277185501067e-05,
|
|
"loss": 0.0001,
|
|
"step": 1490
|
|
},
|
|
{
|
|
"epoch": 3.1876332622601278,
|
|
"grad_norm": 0.009095050394535065,
|
|
"learning_rate": 1.812366737739872e-05,
|
|
"loss": 0.0001,
|
|
"step": 1495
|
|
},
|
|
{
|
|
"epoch": 3.1982942430703627,
|
|
"grad_norm": 0.001300916774198413,
|
|
"learning_rate": 1.8017057569296376e-05,
|
|
"loss": 0.0005,
|
|
"step": 1500
|
|
},
|
|
{
|
|
"epoch": 3.208955223880597,
|
|
"grad_norm": 0.0024559053126722574,
|
|
"learning_rate": 1.791044776119403e-05,
|
|
"loss": 0.0001,
|
|
"step": 1505
|
|
},
|
|
{
|
|
"epoch": 3.2196162046908317,
|
|
"grad_norm": 0.004607733339071274,
|
|
"learning_rate": 1.7803837953091686e-05,
|
|
"loss": 0.0001,
|
|
"step": 1510
|
|
},
|
|
{
|
|
"epoch": 3.230277185501066,
|
|
"grad_norm": 0.0005166428163647652,
|
|
"learning_rate": 1.769722814498934e-05,
|
|
"loss": 0.0009,
|
|
"step": 1515
|
|
},
|
|
{
|
|
"epoch": 3.2409381663113006,
|
|
"grad_norm": 0.0011023251572623849,
|
|
"learning_rate": 1.7590618336886995e-05,
|
|
"loss": 0.0001,
|
|
"step": 1520
|
|
},
|
|
{
|
|
"epoch": 3.251599147121535,
|
|
"grad_norm": 0.0347328707575798,
|
|
"learning_rate": 1.7484008528784647e-05,
|
|
"loss": 0.0001,
|
|
"step": 1525
|
|
},
|
|
{
|
|
"epoch": 3.2622601279317696,
|
|
"grad_norm": 0.03133232146501541,
|
|
"learning_rate": 1.73773987206823e-05,
|
|
"loss": 0.0002,
|
|
"step": 1530
|
|
},
|
|
{
|
|
"epoch": 3.272921108742004,
|
|
"grad_norm": 0.002386990701779723,
|
|
"learning_rate": 1.7270788912579956e-05,
|
|
"loss": 0.0001,
|
|
"step": 1535
|
|
},
|
|
{
|
|
"epoch": 3.283582089552239,
|
|
"grad_norm": 0.0023339290637522936,
|
|
"learning_rate": 1.716417910447761e-05,
|
|
"loss": 0.0001,
|
|
"step": 1540
|
|
},
|
|
{
|
|
"epoch": 3.2942430703624734,
|
|
"grad_norm": 0.0022606707643717527,
|
|
"learning_rate": 1.7057569296375266e-05,
|
|
"loss": 0.0002,
|
|
"step": 1545
|
|
},
|
|
{
|
|
"epoch": 3.304904051172708,
|
|
"grad_norm": 0.021408548578619957,
|
|
"learning_rate": 1.695095948827292e-05,
|
|
"loss": 0.0001,
|
|
"step": 1550
|
|
},
|
|
{
|
|
"epoch": 3.3155650319829424,
|
|
"grad_norm": 18.459163665771484,
|
|
"learning_rate": 1.6844349680170575e-05,
|
|
"loss": 0.0075,
|
|
"step": 1555
|
|
},
|
|
{
|
|
"epoch": 3.326226012793177,
|
|
"grad_norm": 0.00619714567437768,
|
|
"learning_rate": 1.673773987206823e-05,
|
|
"loss": 0.0001,
|
|
"step": 1560
|
|
},
|
|
{
|
|
"epoch": 3.3368869936034113,
|
|
"grad_norm": 0.135460764169693,
|
|
"learning_rate": 1.6631130063965885e-05,
|
|
"loss": 0.0049,
|
|
"step": 1565
|
|
},
|
|
{
|
|
"epoch": 3.3475479744136463,
|
|
"grad_norm": 0.004021491855382919,
|
|
"learning_rate": 1.652452025586354e-05,
|
|
"loss": 0.0001,
|
|
"step": 1570
|
|
},
|
|
{
|
|
"epoch": 3.3582089552238807,
|
|
"grad_norm": 0.001629222882911563,
|
|
"learning_rate": 1.6417910447761194e-05,
|
|
"loss": 0.0011,
|
|
"step": 1575
|
|
},
|
|
{
|
|
"epoch": 3.368869936034115,
|
|
"grad_norm": 0.00044148892629891634,
|
|
"learning_rate": 1.631130063965885e-05,
|
|
"loss": 0.0001,
|
|
"step": 1580
|
|
},
|
|
{
|
|
"epoch": 3.3795309168443497,
|
|
"grad_norm": 0.014529965817928314,
|
|
"learning_rate": 1.6204690831556504e-05,
|
|
"loss": 0.0001,
|
|
"step": 1585
|
|
},
|
|
{
|
|
"epoch": 3.390191897654584,
|
|
"grad_norm": 0.0005223302287049592,
|
|
"learning_rate": 1.6098081023454158e-05,
|
|
"loss": 0.0001,
|
|
"step": 1590
|
|
},
|
|
{
|
|
"epoch": 3.4008528784648187,
|
|
"grad_norm": 0.0013045272789895535,
|
|
"learning_rate": 1.5991471215351813e-05,
|
|
"loss": 0.0002,
|
|
"step": 1595
|
|
},
|
|
{
|
|
"epoch": 3.411513859275053,
|
|
"grad_norm": 0.0005001061945222318,
|
|
"learning_rate": 1.5884861407249468e-05,
|
|
"loss": 0.0001,
|
|
"step": 1600
|
|
},
|
|
{
|
|
"epoch": 3.4221748400852876,
|
|
"grad_norm": 0.009848849847912788,
|
|
"learning_rate": 1.5778251599147122e-05,
|
|
"loss": 0.0027,
|
|
"step": 1605
|
|
},
|
|
{
|
|
"epoch": 3.4328358208955225,
|
|
"grad_norm": 0.0005870772292837501,
|
|
"learning_rate": 1.5671641791044777e-05,
|
|
"loss": 0.0001,
|
|
"step": 1610
|
|
},
|
|
{
|
|
"epoch": 3.443496801705757,
|
|
"grad_norm": 0.0008098568068817258,
|
|
"learning_rate": 1.5565031982942432e-05,
|
|
"loss": 0.0001,
|
|
"step": 1615
|
|
},
|
|
{
|
|
"epoch": 3.4541577825159915,
|
|
"grad_norm": 0.012052051723003387,
|
|
"learning_rate": 1.5458422174840087e-05,
|
|
"loss": 0.0001,
|
|
"step": 1620
|
|
},
|
|
{
|
|
"epoch": 3.464818763326226,
|
|
"grad_norm": 0.011413906700909138,
|
|
"learning_rate": 1.535181236673774e-05,
|
|
"loss": 0.0002,
|
|
"step": 1625
|
|
},
|
|
{
|
|
"epoch": 3.4754797441364604,
|
|
"grad_norm": 0.010221424512565136,
|
|
"learning_rate": 1.5245202558635396e-05,
|
|
"loss": 0.0001,
|
|
"step": 1630
|
|
},
|
|
{
|
|
"epoch": 3.486140724946695,
|
|
"grad_norm": 0.000859771273098886,
|
|
"learning_rate": 1.5138592750533051e-05,
|
|
"loss": 0.0003,
|
|
"step": 1635
|
|
},
|
|
{
|
|
"epoch": 3.49680170575693,
|
|
"grad_norm": 0.0017335901502519846,
|
|
"learning_rate": 1.5031982942430706e-05,
|
|
"loss": 0.0001,
|
|
"step": 1640
|
|
},
|
|
{
|
|
"epoch": 3.5074626865671643,
|
|
"grad_norm": 0.0005336115136742592,
|
|
"learning_rate": 1.4925373134328357e-05,
|
|
"loss": 0.0001,
|
|
"step": 1645
|
|
},
|
|
{
|
|
"epoch": 3.518123667377399,
|
|
"grad_norm": 0.011401413939893246,
|
|
"learning_rate": 1.4818763326226012e-05,
|
|
"loss": 0.0001,
|
|
"step": 1650
|
|
},
|
|
{
|
|
"epoch": 3.5287846481876333,
|
|
"grad_norm": 0.00043319741962477565,
|
|
"learning_rate": 1.4712153518123666e-05,
|
|
"loss": 0.0001,
|
|
"step": 1655
|
|
},
|
|
{
|
|
"epoch": 3.5394456289978677,
|
|
"grad_norm": 0.004011583514511585,
|
|
"learning_rate": 1.4605543710021321e-05,
|
|
"loss": 0.0002,
|
|
"step": 1660
|
|
},
|
|
{
|
|
"epoch": 3.550106609808102,
|
|
"grad_norm": 0.0025535374879837036,
|
|
"learning_rate": 1.4498933901918976e-05,
|
|
"loss": 0.0001,
|
|
"step": 1665
|
|
},
|
|
{
|
|
"epoch": 3.5607675906183367,
|
|
"grad_norm": 0.0008301076595671475,
|
|
"learning_rate": 1.439232409381663e-05,
|
|
"loss": 0.0001,
|
|
"step": 1670
|
|
},
|
|
{
|
|
"epoch": 3.571428571428571,
|
|
"grad_norm": 0.004071046132594347,
|
|
"learning_rate": 1.4285714285714285e-05,
|
|
"loss": 0.0001,
|
|
"step": 1675
|
|
},
|
|
{
|
|
"epoch": 3.582089552238806,
|
|
"grad_norm": 0.005094863008707762,
|
|
"learning_rate": 1.417910447761194e-05,
|
|
"loss": 0.0001,
|
|
"step": 1680
|
|
},
|
|
{
|
|
"epoch": 3.5927505330490406,
|
|
"grad_norm": 11.434825897216797,
|
|
"learning_rate": 1.4072494669509595e-05,
|
|
"loss": 0.0054,
|
|
"step": 1685
|
|
},
|
|
{
|
|
"epoch": 3.603411513859275,
|
|
"grad_norm": 0.0005070870975032449,
|
|
"learning_rate": 1.396588486140725e-05,
|
|
"loss": 0.0001,
|
|
"step": 1690
|
|
},
|
|
{
|
|
"epoch": 3.6140724946695095,
|
|
"grad_norm": 0.0168480034917593,
|
|
"learning_rate": 1.3859275053304904e-05,
|
|
"loss": 0.0001,
|
|
"step": 1695
|
|
},
|
|
{
|
|
"epoch": 3.624733475479744,
|
|
"grad_norm": 0.0008461446850560606,
|
|
"learning_rate": 1.3752665245202559e-05,
|
|
"loss": 0.0001,
|
|
"step": 1700
|
|
},
|
|
{
|
|
"epoch": 3.635394456289979,
|
|
"grad_norm": 0.008455686271190643,
|
|
"learning_rate": 1.3646055437100214e-05,
|
|
"loss": 0.0001,
|
|
"step": 1705
|
|
},
|
|
{
|
|
"epoch": 3.6460554371002134,
|
|
"grad_norm": 0.00041681021684780717,
|
|
"learning_rate": 1.3539445628997869e-05,
|
|
"loss": 0.0001,
|
|
"step": 1710
|
|
},
|
|
{
|
|
"epoch": 3.656716417910448,
|
|
"grad_norm": 0.0087368693202734,
|
|
"learning_rate": 1.3432835820895523e-05,
|
|
"loss": 0.0001,
|
|
"step": 1715
|
|
},
|
|
{
|
|
"epoch": 3.6673773987206824,
|
|
"grad_norm": 0.0003989397664554417,
|
|
"learning_rate": 1.3326226012793178e-05,
|
|
"loss": 0.0001,
|
|
"step": 1720
|
|
},
|
|
{
|
|
"epoch": 3.678038379530917,
|
|
"grad_norm": 0.0028573633171617985,
|
|
"learning_rate": 1.3219616204690833e-05,
|
|
"loss": 0.0001,
|
|
"step": 1725
|
|
},
|
|
{
|
|
"epoch": 3.6886993603411513,
|
|
"grad_norm": 0.010743173770606518,
|
|
"learning_rate": 1.3113006396588488e-05,
|
|
"loss": 0.0001,
|
|
"step": 1730
|
|
},
|
|
{
|
|
"epoch": 3.699360341151386,
|
|
"grad_norm": 0.0005059782997705042,
|
|
"learning_rate": 1.3006396588486142e-05,
|
|
"loss": 0.0074,
|
|
"step": 1735
|
|
},
|
|
{
|
|
"epoch": 3.7100213219616203,
|
|
"grad_norm": 0.004534687381237745,
|
|
"learning_rate": 1.2899786780383797e-05,
|
|
"loss": 0.0,
|
|
"step": 1740
|
|
},
|
|
{
|
|
"epoch": 3.7206823027718547,
|
|
"grad_norm": 0.013566154055297375,
|
|
"learning_rate": 1.2793176972281452e-05,
|
|
"loss": 0.0001,
|
|
"step": 1745
|
|
},
|
|
{
|
|
"epoch": 3.7313432835820897,
|
|
"grad_norm": 0.0014023985713720322,
|
|
"learning_rate": 1.2686567164179105e-05,
|
|
"loss": 0.0001,
|
|
"step": 1750
|
|
},
|
|
{
|
|
"epoch": 3.742004264392324,
|
|
"grad_norm": 0.0015459961723536253,
|
|
"learning_rate": 1.257995735607676e-05,
|
|
"loss": 0.0001,
|
|
"step": 1755
|
|
},
|
|
{
|
|
"epoch": 3.7526652452025586,
|
|
"grad_norm": 0.0004811576509382576,
|
|
"learning_rate": 1.2473347547974414e-05,
|
|
"loss": 0.0001,
|
|
"step": 1760
|
|
},
|
|
{
|
|
"epoch": 3.763326226012793,
|
|
"grad_norm": 0.0011233583791181445,
|
|
"learning_rate": 1.2366737739872069e-05,
|
|
"loss": 0.0001,
|
|
"step": 1765
|
|
},
|
|
{
|
|
"epoch": 3.7739872068230276,
|
|
"grad_norm": 0.0005417008651420474,
|
|
"learning_rate": 1.2260127931769722e-05,
|
|
"loss": 0.0001,
|
|
"step": 1770
|
|
},
|
|
{
|
|
"epoch": 3.7846481876332625,
|
|
"grad_norm": 0.0008258482557721436,
|
|
"learning_rate": 1.2153518123667377e-05,
|
|
"loss": 0.0002,
|
|
"step": 1775
|
|
},
|
|
{
|
|
"epoch": 3.795309168443497,
|
|
"grad_norm": 0.0006673976313322783,
|
|
"learning_rate": 1.2046908315565032e-05,
|
|
"loss": 0.0002,
|
|
"step": 1780
|
|
},
|
|
{
|
|
"epoch": 3.8059701492537314,
|
|
"grad_norm": 0.05592097342014313,
|
|
"learning_rate": 1.1940298507462686e-05,
|
|
"loss": 0.0001,
|
|
"step": 1785
|
|
},
|
|
{
|
|
"epoch": 3.816631130063966,
|
|
"grad_norm": 0.012964832596480846,
|
|
"learning_rate": 1.1833688699360341e-05,
|
|
"loss": 0.0001,
|
|
"step": 1790
|
|
},
|
|
{
|
|
"epoch": 3.8272921108742004,
|
|
"grad_norm": 0.0033881776034832,
|
|
"learning_rate": 1.1727078891257996e-05,
|
|
"loss": 0.0001,
|
|
"step": 1795
|
|
},
|
|
{
|
|
"epoch": 3.837953091684435,
|
|
"grad_norm": 0.003563040867447853,
|
|
"learning_rate": 1.162046908315565e-05,
|
|
"loss": 0.0001,
|
|
"step": 1800
|
|
},
|
|
{
|
|
"epoch": 3.8486140724946694,
|
|
"grad_norm": 0.0017364048399031162,
|
|
"learning_rate": 1.1513859275053305e-05,
|
|
"loss": 0.0018,
|
|
"step": 1805
|
|
},
|
|
{
|
|
"epoch": 3.859275053304904,
|
|
"grad_norm": 0.0875784158706665,
|
|
"learning_rate": 1.140724946695096e-05,
|
|
"loss": 0.0001,
|
|
"step": 1810
|
|
},
|
|
{
|
|
"epoch": 3.8699360341151388,
|
|
"grad_norm": 0.002513446379452944,
|
|
"learning_rate": 1.1300639658848615e-05,
|
|
"loss": 0.0001,
|
|
"step": 1815
|
|
},
|
|
{
|
|
"epoch": 3.8805970149253732,
|
|
"grad_norm": 0.004740020260214806,
|
|
"learning_rate": 1.119402985074627e-05,
|
|
"loss": 0.0001,
|
|
"step": 1820
|
|
},
|
|
{
|
|
"epoch": 3.8912579957356077,
|
|
"grad_norm": 0.00046836529509164393,
|
|
"learning_rate": 1.1087420042643924e-05,
|
|
"loss": 0.0001,
|
|
"step": 1825
|
|
},
|
|
{
|
|
"epoch": 3.901918976545842,
|
|
"grad_norm": 0.0053132628090679646,
|
|
"learning_rate": 1.0980810234541579e-05,
|
|
"loss": 0.0001,
|
|
"step": 1830
|
|
},
|
|
{
|
|
"epoch": 3.9125799573560767,
|
|
"grad_norm": 0.001051435829140246,
|
|
"learning_rate": 1.0874200426439234e-05,
|
|
"loss": 0.0001,
|
|
"step": 1835
|
|
},
|
|
{
|
|
"epoch": 3.923240938166311,
|
|
"grad_norm": 16.337718963623047,
|
|
"learning_rate": 1.0767590618336887e-05,
|
|
"loss": 0.0112,
|
|
"step": 1840
|
|
},
|
|
{
|
|
"epoch": 3.933901918976546,
|
|
"grad_norm": 0.08996672928333282,
|
|
"learning_rate": 1.0660980810234541e-05,
|
|
"loss": 0.0001,
|
|
"step": 1845
|
|
},
|
|
{
|
|
"epoch": 3.9445628997867805,
|
|
"grad_norm": 0.0006036468548700213,
|
|
"learning_rate": 1.0554371002132196e-05,
|
|
"loss": 0.0001,
|
|
"step": 1850
|
|
},
|
|
{
|
|
"epoch": 3.955223880597015,
|
|
"grad_norm": 0.0012560180621221662,
|
|
"learning_rate": 1.0447761194029851e-05,
|
|
"loss": 0.0001,
|
|
"step": 1855
|
|
},
|
|
{
|
|
"epoch": 3.9658848614072495,
|
|
"grad_norm": 0.002333889715373516,
|
|
"learning_rate": 1.0341151385927506e-05,
|
|
"loss": 0.0001,
|
|
"step": 1860
|
|
},
|
|
{
|
|
"epoch": 3.976545842217484,
|
|
"grad_norm": 0.002112707355991006,
|
|
"learning_rate": 1.023454157782516e-05,
|
|
"loss": 0.0001,
|
|
"step": 1865
|
|
},
|
|
{
|
|
"epoch": 3.9872068230277184,
|
|
"grad_norm": 0.0009484985494054854,
|
|
"learning_rate": 1.0127931769722815e-05,
|
|
"loss": 0.0001,
|
|
"step": 1870
|
|
},
|
|
{
|
|
"epoch": 3.997867803837953,
|
|
"grad_norm": 0.000689346925355494,
|
|
"learning_rate": 1.002132196162047e-05,
|
|
"loss": 0.0001,
|
|
"step": 1875
|
|
},
|
|
{
|
|
"epoch": 4.0,
|
|
"eval_accuracy": 0.9562666666666667,
|
|
"eval_loss": 0.2534376382827759,
|
|
"eval_runtime": 17.7051,
|
|
"eval_samples_per_second": 211.803,
|
|
"eval_steps_per_second": 6.665,
|
|
"step": 1876
|
|
},
|
|
{
|
|
"epoch": 4.008528784648187,
|
|
"grad_norm": 0.0005996975814923644,
|
|
"learning_rate": 9.914712153518125e-06,
|
|
"loss": 0.002,
|
|
"step": 1880
|
|
},
|
|
{
|
|
"epoch": 4.019189765458422,
|
|
"grad_norm": 0.0005832866299897432,
|
|
"learning_rate": 9.80810234541578e-06,
|
|
"loss": 0.0001,
|
|
"step": 1885
|
|
},
|
|
{
|
|
"epoch": 4.029850746268656,
|
|
"grad_norm": 0.0008000755333341658,
|
|
"learning_rate": 9.701492537313434e-06,
|
|
"loss": 0.0,
|
|
"step": 1890
|
|
},
|
|
{
|
|
"epoch": 4.040511727078891,
|
|
"grad_norm": 0.0008381507941521704,
|
|
"learning_rate": 9.594882729211089e-06,
|
|
"loss": 0.0001,
|
|
"step": 1895
|
|
},
|
|
{
|
|
"epoch": 4.051172707889126,
|
|
"grad_norm": 0.0007523804670199752,
|
|
"learning_rate": 9.488272921108744e-06,
|
|
"loss": 0.0001,
|
|
"step": 1900
|
|
},
|
|
{
|
|
"epoch": 4.061833688699361,
|
|
"grad_norm": 0.000499650661367923,
|
|
"learning_rate": 9.381663113006398e-06,
|
|
"loss": 0.0001,
|
|
"step": 1905
|
|
},
|
|
{
|
|
"epoch": 4.072494669509595,
|
|
"grad_norm": 0.0004529411962721497,
|
|
"learning_rate": 9.275053304904053e-06,
|
|
"loss": 0.0,
|
|
"step": 1910
|
|
},
|
|
{
|
|
"epoch": 4.08315565031983,
|
|
"grad_norm": 0.0015674469759687781,
|
|
"learning_rate": 9.168443496801706e-06,
|
|
"loss": 0.0001,
|
|
"step": 1915
|
|
},
|
|
{
|
|
"epoch": 4.093816631130064,
|
|
"grad_norm": 0.0015955445123836398,
|
|
"learning_rate": 9.06183368869936e-06,
|
|
"loss": 0.0001,
|
|
"step": 1920
|
|
},
|
|
{
|
|
"epoch": 4.104477611940299,
|
|
"grad_norm": 0.0005198908620513976,
|
|
"learning_rate": 8.955223880597016e-06,
|
|
"loss": 0.0001,
|
|
"step": 1925
|
|
},
|
|
{
|
|
"epoch": 4.115138592750533,
|
|
"grad_norm": 0.0006500064628198743,
|
|
"learning_rate": 8.84861407249467e-06,
|
|
"loss": 0.0001,
|
|
"step": 1930
|
|
},
|
|
{
|
|
"epoch": 4.1257995735607675,
|
|
"grad_norm": 0.01806180737912655,
|
|
"learning_rate": 8.742004264392323e-06,
|
|
"loss": 0.0001,
|
|
"step": 1935
|
|
},
|
|
{
|
|
"epoch": 4.136460554371002,
|
|
"grad_norm": 0.004434024915099144,
|
|
"learning_rate": 8.635394456289978e-06,
|
|
"loss": 0.0001,
|
|
"step": 1940
|
|
},
|
|
{
|
|
"epoch": 4.1471215351812365,
|
|
"grad_norm": 0.0009599188924767077,
|
|
"learning_rate": 8.528784648187633e-06,
|
|
"loss": 0.0001,
|
|
"step": 1945
|
|
},
|
|
{
|
|
"epoch": 4.157782515991471,
|
|
"grad_norm": 0.0004996213829144835,
|
|
"learning_rate": 8.422174840085288e-06,
|
|
"loss": 0.0001,
|
|
"step": 1950
|
|
},
|
|
{
|
|
"epoch": 4.1684434968017055,
|
|
"grad_norm": 0.0008770860731601715,
|
|
"learning_rate": 8.315565031982942e-06,
|
|
"loss": 0.0001,
|
|
"step": 1955
|
|
},
|
|
{
|
|
"epoch": 4.17910447761194,
|
|
"grad_norm": 0.0005104989977553487,
|
|
"learning_rate": 8.208955223880597e-06,
|
|
"loss": 0.0001,
|
|
"step": 1960
|
|
},
|
|
{
|
|
"epoch": 4.189765458422174,
|
|
"grad_norm": 0.00043773301877081394,
|
|
"learning_rate": 8.102345415778252e-06,
|
|
"loss": 0.0,
|
|
"step": 1965
|
|
},
|
|
{
|
|
"epoch": 4.20042643923241,
|
|
"grad_norm": 0.00038476119516417384,
|
|
"learning_rate": 7.995735607675907e-06,
|
|
"loss": 0.0001,
|
|
"step": 1970
|
|
},
|
|
{
|
|
"epoch": 4.211087420042644,
|
|
"grad_norm": 0.023430561646819115,
|
|
"learning_rate": 7.889125799573561e-06,
|
|
"loss": 0.0001,
|
|
"step": 1975
|
|
},
|
|
{
|
|
"epoch": 4.221748400852879,
|
|
"grad_norm": 0.00048294186126440763,
|
|
"learning_rate": 7.782515991471216e-06,
|
|
"loss": 0.0001,
|
|
"step": 1980
|
|
},
|
|
{
|
|
"epoch": 4.232409381663113,
|
|
"grad_norm": 0.0015317806974053383,
|
|
"learning_rate": 7.67590618336887e-06,
|
|
"loss": 0.0001,
|
|
"step": 1985
|
|
},
|
|
{
|
|
"epoch": 4.243070362473348,
|
|
"grad_norm": 0.0007306035258807242,
|
|
"learning_rate": 7.5692963752665255e-06,
|
|
"loss": 0.0001,
|
|
"step": 1990
|
|
},
|
|
{
|
|
"epoch": 4.253731343283582,
|
|
"grad_norm": 0.0006193820736370981,
|
|
"learning_rate": 7.4626865671641785e-06,
|
|
"loss": 0.0001,
|
|
"step": 1995
|
|
},
|
|
{
|
|
"epoch": 4.264392324093817,
|
|
"grad_norm": 0.001915331231430173,
|
|
"learning_rate": 7.356076759061833e-06,
|
|
"loss": 0.0001,
|
|
"step": 2000
|
|
},
|
|
{
|
|
"epoch": 4.275053304904051,
|
|
"grad_norm": 0.0028169583529233932,
|
|
"learning_rate": 7.249466950959488e-06,
|
|
"loss": 0.0001,
|
|
"step": 2005
|
|
},
|
|
{
|
|
"epoch": 4.285714285714286,
|
|
"grad_norm": 0.00524992635473609,
|
|
"learning_rate": 7.142857142857143e-06,
|
|
"loss": 0.0001,
|
|
"step": 2010
|
|
},
|
|
{
|
|
"epoch": 4.29637526652452,
|
|
"grad_norm": 0.00454280711710453,
|
|
"learning_rate": 7.0362473347547975e-06,
|
|
"loss": 0.0001,
|
|
"step": 2015
|
|
},
|
|
{
|
|
"epoch": 4.3070362473347545,
|
|
"grad_norm": 0.000466930097900331,
|
|
"learning_rate": 6.929637526652452e-06,
|
|
"loss": 0.0001,
|
|
"step": 2020
|
|
},
|
|
{
|
|
"epoch": 4.317697228144989,
|
|
"grad_norm": 0.18847821652889252,
|
|
"learning_rate": 6.823027718550107e-06,
|
|
"loss": 0.001,
|
|
"step": 2025
|
|
},
|
|
{
|
|
"epoch": 4.3283582089552235,
|
|
"grad_norm": 0.0003990318800788373,
|
|
"learning_rate": 6.716417910447762e-06,
|
|
"loss": 0.0,
|
|
"step": 2030
|
|
},
|
|
{
|
|
"epoch": 4.339019189765459,
|
|
"grad_norm": 0.009269189089536667,
|
|
"learning_rate": 6.609808102345416e-06,
|
|
"loss": 0.0001,
|
|
"step": 2035
|
|
},
|
|
{
|
|
"epoch": 4.349680170575693,
|
|
"grad_norm": 0.0008212586981244385,
|
|
"learning_rate": 6.503198294243071e-06,
|
|
"loss": 0.0001,
|
|
"step": 2040
|
|
},
|
|
{
|
|
"epoch": 4.360341151385928,
|
|
"grad_norm": 0.0005204948829486966,
|
|
"learning_rate": 6.396588486140726e-06,
|
|
"loss": 0.0001,
|
|
"step": 2045
|
|
},
|
|
{
|
|
"epoch": 4.371002132196162,
|
|
"grad_norm": 0.00046012879465706646,
|
|
"learning_rate": 6.28997867803838e-06,
|
|
"loss": 0.0,
|
|
"step": 2050
|
|
},
|
|
{
|
|
"epoch": 4.381663113006397,
|
|
"grad_norm": 0.0005494586657732725,
|
|
"learning_rate": 6.1833688699360345e-06,
|
|
"loss": 0.0,
|
|
"step": 2055
|
|
},
|
|
{
|
|
"epoch": 4.392324093816631,
|
|
"grad_norm": 0.0007372510735876858,
|
|
"learning_rate": 6.076759061833688e-06,
|
|
"loss": 0.0,
|
|
"step": 2060
|
|
},
|
|
{
|
|
"epoch": 4.402985074626866,
|
|
"grad_norm": 0.0009128697565756738,
|
|
"learning_rate": 5.970149253731343e-06,
|
|
"loss": 0.0001,
|
|
"step": 2065
|
|
},
|
|
{
|
|
"epoch": 4.4136460554371,
|
|
"grad_norm": 0.0007230183109641075,
|
|
"learning_rate": 5.863539445628998e-06,
|
|
"loss": 0.0001,
|
|
"step": 2070
|
|
},
|
|
{
|
|
"epoch": 4.424307036247335,
|
|
"grad_norm": 0.001352477353066206,
|
|
"learning_rate": 5.756929637526653e-06,
|
|
"loss": 0.0001,
|
|
"step": 2075
|
|
},
|
|
{
|
|
"epoch": 4.434968017057569,
|
|
"grad_norm": 0.000400946126319468,
|
|
"learning_rate": 5.650319829424307e-06,
|
|
"loss": 0.0,
|
|
"step": 2080
|
|
},
|
|
{
|
|
"epoch": 4.445628997867804,
|
|
"grad_norm": 0.003582812612876296,
|
|
"learning_rate": 5.543710021321962e-06,
|
|
"loss": 0.0001,
|
|
"step": 2085
|
|
},
|
|
{
|
|
"epoch": 4.456289978678038,
|
|
"grad_norm": 0.0007582177640870214,
|
|
"learning_rate": 5.437100213219617e-06,
|
|
"loss": 0.0001,
|
|
"step": 2090
|
|
},
|
|
{
|
|
"epoch": 4.466950959488273,
|
|
"grad_norm": 0.0006859108107164502,
|
|
"learning_rate": 5.330490405117271e-06,
|
|
"loss": 0.0001,
|
|
"step": 2095
|
|
},
|
|
{
|
|
"epoch": 4.477611940298507,
|
|
"grad_norm": 0.00044654577504843473,
|
|
"learning_rate": 5.2238805970149255e-06,
|
|
"loss": 0.0001,
|
|
"step": 2100
|
|
},
|
|
{
|
|
"epoch": 4.4882729211087415,
|
|
"grad_norm": 0.0007140070665627718,
|
|
"learning_rate": 5.11727078891258e-06,
|
|
"loss": 0.0001,
|
|
"step": 2105
|
|
},
|
|
{
|
|
"epoch": 4.498933901918977,
|
|
"grad_norm": 0.0004241761635057628,
|
|
"learning_rate": 5.010660980810235e-06,
|
|
"loss": 0.0,
|
|
"step": 2110
|
|
},
|
|
{
|
|
"epoch": 4.509594882729211,
|
|
"grad_norm": 0.0005627896753139794,
|
|
"learning_rate": 4.90405117270789e-06,
|
|
"loss": 0.0,
|
|
"step": 2115
|
|
},
|
|
{
|
|
"epoch": 4.520255863539446,
|
|
"grad_norm": 0.0005950676859356463,
|
|
"learning_rate": 4.797441364605544e-06,
|
|
"loss": 0.0001,
|
|
"step": 2120
|
|
},
|
|
{
|
|
"epoch": 4.53091684434968,
|
|
"grad_norm": 0.0014802158111706376,
|
|
"learning_rate": 4.690831556503199e-06,
|
|
"loss": 0.0001,
|
|
"step": 2125
|
|
},
|
|
{
|
|
"epoch": 4.541577825159915,
|
|
"grad_norm": 0.001410114113241434,
|
|
"learning_rate": 4.584221748400853e-06,
|
|
"loss": 0.0001,
|
|
"step": 2130
|
|
},
|
|
{
|
|
"epoch": 4.552238805970149,
|
|
"grad_norm": 0.0006104848580434918,
|
|
"learning_rate": 4.477611940298508e-06,
|
|
"loss": 0.0001,
|
|
"step": 2135
|
|
},
|
|
{
|
|
"epoch": 4.562899786780384,
|
|
"grad_norm": 0.00041335076093673706,
|
|
"learning_rate": 4.371002132196162e-06,
|
|
"loss": 0.0001,
|
|
"step": 2140
|
|
},
|
|
{
|
|
"epoch": 4.573560767590618,
|
|
"grad_norm": 0.0010959588689729571,
|
|
"learning_rate": 4.264392324093816e-06,
|
|
"loss": 0.0001,
|
|
"step": 2145
|
|
},
|
|
{
|
|
"epoch": 4.584221748400853,
|
|
"grad_norm": 0.0004437138559296727,
|
|
"learning_rate": 4.157782515991471e-06,
|
|
"loss": 0.0,
|
|
"step": 2150
|
|
},
|
|
{
|
|
"epoch": 4.594882729211087,
|
|
"grad_norm": 0.0006708800792694092,
|
|
"learning_rate": 4.051172707889126e-06,
|
|
"loss": 0.0,
|
|
"step": 2155
|
|
},
|
|
{
|
|
"epoch": 4.605543710021322,
|
|
"grad_norm": 0.0019771638326346874,
|
|
"learning_rate": 3.944562899786781e-06,
|
|
"loss": 0.0001,
|
|
"step": 2160
|
|
},
|
|
{
|
|
"epoch": 4.616204690831556,
|
|
"grad_norm": 0.00046450068475678563,
|
|
"learning_rate": 3.837953091684435e-06,
|
|
"loss": 0.0,
|
|
"step": 2165
|
|
},
|
|
{
|
|
"epoch": 4.6268656716417915,
|
|
"grad_norm": 0.003939396236091852,
|
|
"learning_rate": 3.7313432835820893e-06,
|
|
"loss": 0.0,
|
|
"step": 2170
|
|
},
|
|
{
|
|
"epoch": 4.637526652452026,
|
|
"grad_norm": 0.003692534053698182,
|
|
"learning_rate": 3.624733475479744e-06,
|
|
"loss": 0.0001,
|
|
"step": 2175
|
|
},
|
|
{
|
|
"epoch": 4.6481876332622605,
|
|
"grad_norm": 0.001608129357919097,
|
|
"learning_rate": 3.5181236673773987e-06,
|
|
"loss": 0.0,
|
|
"step": 2180
|
|
},
|
|
{
|
|
"epoch": 4.658848614072495,
|
|
"grad_norm": 0.0026127141900360584,
|
|
"learning_rate": 3.4115138592750535e-06,
|
|
"loss": 0.0001,
|
|
"step": 2185
|
|
},
|
|
{
|
|
"epoch": 4.669509594882729,
|
|
"grad_norm": 0.0008662310428917408,
|
|
"learning_rate": 3.304904051172708e-06,
|
|
"loss": 0.0,
|
|
"step": 2190
|
|
},
|
|
{
|
|
"epoch": 4.680170575692964,
|
|
"grad_norm": 0.0004079797654412687,
|
|
"learning_rate": 3.198294243070363e-06,
|
|
"loss": 0.0001,
|
|
"step": 2195
|
|
},
|
|
{
|
|
"epoch": 4.690831556503198,
|
|
"grad_norm": 0.0014223584439605474,
|
|
"learning_rate": 3.0916844349680173e-06,
|
|
"loss": 0.0001,
|
|
"step": 2200
|
|
},
|
|
{
|
|
"epoch": 4.701492537313433,
|
|
"grad_norm": 0.001200292375870049,
|
|
"learning_rate": 2.9850746268656716e-06,
|
|
"loss": 0.0001,
|
|
"step": 2205
|
|
},
|
|
{
|
|
"epoch": 4.712153518123667,
|
|
"grad_norm": 0.0006543201743625104,
|
|
"learning_rate": 2.8784648187633263e-06,
|
|
"loss": 0.0,
|
|
"step": 2210
|
|
},
|
|
{
|
|
"epoch": 4.722814498933902,
|
|
"grad_norm": 0.3227696120738983,
|
|
"learning_rate": 2.771855010660981e-06,
|
|
"loss": 0.0057,
|
|
"step": 2215
|
|
},
|
|
{
|
|
"epoch": 4.733475479744136,
|
|
"grad_norm": 0.0005573455709964037,
|
|
"learning_rate": 2.6652452025586354e-06,
|
|
"loss": 0.0001,
|
|
"step": 2220
|
|
},
|
|
{
|
|
"epoch": 4.744136460554371,
|
|
"grad_norm": 0.0008872459293343127,
|
|
"learning_rate": 2.55863539445629e-06,
|
|
"loss": 0.0,
|
|
"step": 2225
|
|
},
|
|
{
|
|
"epoch": 4.754797441364605,
|
|
"grad_norm": 0.006936908699572086,
|
|
"learning_rate": 2.452025586353945e-06,
|
|
"loss": 0.0001,
|
|
"step": 2230
|
|
},
|
|
{
|
|
"epoch": 4.76545842217484,
|
|
"grad_norm": 0.0006477853166870773,
|
|
"learning_rate": 2.3454157782515996e-06,
|
|
"loss": 0.0001,
|
|
"step": 2235
|
|
},
|
|
{
|
|
"epoch": 4.776119402985074,
|
|
"grad_norm": 0.004170337691903114,
|
|
"learning_rate": 2.238805970149254e-06,
|
|
"loss": 0.0001,
|
|
"step": 2240
|
|
},
|
|
{
|
|
"epoch": 4.786780383795309,
|
|
"grad_norm": 0.0004751793167088181,
|
|
"learning_rate": 2.132196162046908e-06,
|
|
"loss": 0.0001,
|
|
"step": 2245
|
|
},
|
|
{
|
|
"epoch": 4.797441364605544,
|
|
"grad_norm": 0.00041296033305116,
|
|
"learning_rate": 2.025586353944563e-06,
|
|
"loss": 0.0001,
|
|
"step": 2250
|
|
},
|
|
{
|
|
"epoch": 4.8081023454157785,
|
|
"grad_norm": 0.0022082251962274313,
|
|
"learning_rate": 1.9189765458422177e-06,
|
|
"loss": 0.0001,
|
|
"step": 2255
|
|
},
|
|
{
|
|
"epoch": 4.818763326226013,
|
|
"grad_norm": 0.0007396464352495968,
|
|
"learning_rate": 1.812366737739872e-06,
|
|
"loss": 0.0001,
|
|
"step": 2260
|
|
},
|
|
{
|
|
"epoch": 4.8294243070362475,
|
|
"grad_norm": 0.0009258220670744777,
|
|
"learning_rate": 1.7057569296375267e-06,
|
|
"loss": 0.0001,
|
|
"step": 2265
|
|
},
|
|
{
|
|
"epoch": 4.840085287846482,
|
|
"grad_norm": 0.001538619282655418,
|
|
"learning_rate": 1.5991471215351815e-06,
|
|
"loss": 0.0,
|
|
"step": 2270
|
|
},
|
|
{
|
|
"epoch": 4.850746268656716,
|
|
"grad_norm": 0.0015768733574077487,
|
|
"learning_rate": 1.4925373134328358e-06,
|
|
"loss": 0.0,
|
|
"step": 2275
|
|
},
|
|
{
|
|
"epoch": 4.861407249466951,
|
|
"grad_norm": 0.0008061281987465918,
|
|
"learning_rate": 1.3859275053304905e-06,
|
|
"loss": 0.0001,
|
|
"step": 2280
|
|
},
|
|
{
|
|
"epoch": 4.872068230277185,
|
|
"grad_norm": 0.010045230388641357,
|
|
"learning_rate": 1.279317697228145e-06,
|
|
"loss": 0.0001,
|
|
"step": 2285
|
|
},
|
|
{
|
|
"epoch": 4.88272921108742,
|
|
"grad_norm": 0.002255336381494999,
|
|
"learning_rate": 1.1727078891257998e-06,
|
|
"loss": 0.0001,
|
|
"step": 2290
|
|
},
|
|
{
|
|
"epoch": 4.893390191897654,
|
|
"grad_norm": 0.0019368295324966311,
|
|
"learning_rate": 1.066098081023454e-06,
|
|
"loss": 0.0001,
|
|
"step": 2295
|
|
},
|
|
{
|
|
"epoch": 4.904051172707889,
|
|
"grad_norm": 0.00044618244282901287,
|
|
"learning_rate": 9.594882729211088e-07,
|
|
"loss": 0.0001,
|
|
"step": 2300
|
|
},
|
|
{
|
|
"epoch": 4.914712153518123,
|
|
"grad_norm": 0.0007794931880198419,
|
|
"learning_rate": 8.528784648187634e-07,
|
|
"loss": 0.0,
|
|
"step": 2305
|
|
},
|
|
{
|
|
"epoch": 4.925373134328359,
|
|
"grad_norm": 0.0007738995482213795,
|
|
"learning_rate": 7.462686567164179e-07,
|
|
"loss": 0.0001,
|
|
"step": 2310
|
|
},
|
|
{
|
|
"epoch": 4.936034115138593,
|
|
"grad_norm": 0.0018532798858359456,
|
|
"learning_rate": 6.396588486140725e-07,
|
|
"loss": 0.0001,
|
|
"step": 2315
|
|
},
|
|
{
|
|
"epoch": 4.946695095948828,
|
|
"grad_norm": 0.0026583250146359205,
|
|
"learning_rate": 5.33049040511727e-07,
|
|
"loss": 0.0001,
|
|
"step": 2320
|
|
},
|
|
{
|
|
"epoch": 4.957356076759062,
|
|
"grad_norm": 0.0007867804961279035,
|
|
"learning_rate": 4.264392324093817e-07,
|
|
"loss": 0.0001,
|
|
"step": 2325
|
|
},
|
|
{
|
|
"epoch": 4.968017057569297,
|
|
"grad_norm": 0.0018914755200967193,
|
|
"learning_rate": 3.1982942430703626e-07,
|
|
"loss": 0.0001,
|
|
"step": 2330
|
|
},
|
|
{
|
|
"epoch": 4.978678038379531,
|
|
"grad_norm": 0.011659308336675167,
|
|
"learning_rate": 2.1321961620469084e-07,
|
|
"loss": 0.0001,
|
|
"step": 2335
|
|
},
|
|
{
|
|
"epoch": 4.9893390191897655,
|
|
"grad_norm": 0.0004196056688670069,
|
|
"learning_rate": 1.0660980810234542e-07,
|
|
"loss": 0.0001,
|
|
"step": 2340
|
|
},
|
|
{
|
|
"epoch": 5.0,
|
|
"grad_norm": 0.0008831822779029608,
|
|
"learning_rate": 0.0,
|
|
"loss": 0.0,
|
|
"step": 2345
|
|
},
|
|
{
|
|
"epoch": 5.0,
|
|
"eval_accuracy": 0.9544,
|
|
"eval_loss": 0.25770530104637146,
|
|
"eval_runtime": 18.0906,
|
|
"eval_samples_per_second": 207.29,
|
|
"eval_steps_per_second": 6.523,
|
|
"step": 2345
|
|
},
|
|
{
|
|
"epoch": 5.0,
|
|
"step": 2345,
|
|
"total_flos": 5.8118992210944e+18,
|
|
"train_loss": 0.029037296106406638,
|
|
"train_runtime": 835.0958,
|
|
"train_samples_per_second": 89.81,
|
|
"train_steps_per_second": 2.808
|
|
}
|
|
],
|
|
"logging_steps": 5,
|
|
"max_steps": 2345,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 5,
|
|
"save_steps": 500,
|
|
"stateful_callbacks": {
|
|
"TrainerControl": {
|
|
"args": {
|
|
"should_epoch_stop": false,
|
|
"should_evaluate": false,
|
|
"should_log": false,
|
|
"should_save": true,
|
|
"should_training_stop": true
|
|
},
|
|
"attributes": {}
|
|
}
|
|
},
|
|
"total_flos": 5.8118992210944e+18,
|
|
"train_batch_size": 32,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|
|
|