2464 lines
64 KiB
JSON
2464 lines
64 KiB
JSON
{
|
|
"best_global_step": 284,
|
|
"best_metric": 0.11889845132827759,
|
|
"best_model_checkpoint": "saves_bts_preliminary/base/llama-3.2-1b-instruct/train_rte_42_1776331559/checkpoint-284",
|
|
"epoch": 5.0,
|
|
"eval_steps": 71,
|
|
"global_step": 1405,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"epoch": 0.017793594306049824,
|
|
"grad_norm": 214.55776977539062,
|
|
"learning_rate": 1.4184397163120568e-07,
|
|
"loss": 0.7401,
|
|
"num_input_tokens_seen": 7872,
|
|
"step": 5
|
|
},
|
|
{
|
|
"epoch": 0.03558718861209965,
|
|
"grad_norm": 161.54965209960938,
|
|
"learning_rate": 3.1914893617021275e-07,
|
|
"loss": 0.6171,
|
|
"num_input_tokens_seen": 14784,
|
|
"step": 10
|
|
},
|
|
{
|
|
"epoch": 0.05338078291814947,
|
|
"grad_norm": 70.1294937133789,
|
|
"learning_rate": 4.964539007092199e-07,
|
|
"loss": 0.4092,
|
|
"num_input_tokens_seen": 23424,
|
|
"step": 15
|
|
},
|
|
{
|
|
"epoch": 0.0711743772241993,
|
|
"grad_norm": 83.8577880859375,
|
|
"learning_rate": 6.73758865248227e-07,
|
|
"loss": 0.2659,
|
|
"num_input_tokens_seen": 29824,
|
|
"step": 20
|
|
},
|
|
{
|
|
"epoch": 0.08896797153024912,
|
|
"grad_norm": 70.67394256591797,
|
|
"learning_rate": 8.510638297872341e-07,
|
|
"loss": 0.2666,
|
|
"num_input_tokens_seen": 37824,
|
|
"step": 25
|
|
},
|
|
{
|
|
"epoch": 0.10676156583629894,
|
|
"grad_norm": 35.297691345214844,
|
|
"learning_rate": 1.0283687943262412e-06,
|
|
"loss": 0.2573,
|
|
"num_input_tokens_seen": 44608,
|
|
"step": 30
|
|
},
|
|
{
|
|
"epoch": 0.12455516014234876,
|
|
"grad_norm": 8.436112403869629,
|
|
"learning_rate": 1.2056737588652482e-06,
|
|
"loss": 0.3322,
|
|
"num_input_tokens_seen": 51968,
|
|
"step": 35
|
|
},
|
|
{
|
|
"epoch": 0.1423487544483986,
|
|
"grad_norm": 13.455877304077148,
|
|
"learning_rate": 1.3829787234042555e-06,
|
|
"loss": 0.1441,
|
|
"num_input_tokens_seen": 59456,
|
|
"step": 40
|
|
},
|
|
{
|
|
"epoch": 0.1601423487544484,
|
|
"grad_norm": 32.384830474853516,
|
|
"learning_rate": 1.5602836879432626e-06,
|
|
"loss": 0.163,
|
|
"num_input_tokens_seen": 66496,
|
|
"step": 45
|
|
},
|
|
{
|
|
"epoch": 0.17793594306049823,
|
|
"grad_norm": 26.830564498901367,
|
|
"learning_rate": 1.7375886524822697e-06,
|
|
"loss": 0.1815,
|
|
"num_input_tokens_seen": 73408,
|
|
"step": 50
|
|
},
|
|
{
|
|
"epoch": 0.19572953736654805,
|
|
"grad_norm": 7.963785171508789,
|
|
"learning_rate": 1.9148936170212767e-06,
|
|
"loss": 0.1484,
|
|
"num_input_tokens_seen": 80576,
|
|
"step": 55
|
|
},
|
|
{
|
|
"epoch": 0.21352313167259787,
|
|
"grad_norm": 8.084745407104492,
|
|
"learning_rate": 2.092198581560284e-06,
|
|
"loss": 0.1828,
|
|
"num_input_tokens_seen": 88256,
|
|
"step": 60
|
|
},
|
|
{
|
|
"epoch": 0.2313167259786477,
|
|
"grad_norm": 11.574740409851074,
|
|
"learning_rate": 2.269503546099291e-06,
|
|
"loss": 0.1709,
|
|
"num_input_tokens_seen": 96256,
|
|
"step": 65
|
|
},
|
|
{
|
|
"epoch": 0.2491103202846975,
|
|
"grad_norm": 20.736160278320312,
|
|
"learning_rate": 2.446808510638298e-06,
|
|
"loss": 0.2309,
|
|
"num_input_tokens_seen": 103424,
|
|
"step": 70
|
|
},
|
|
{
|
|
"epoch": 0.2526690391459075,
|
|
"eval_loss": 0.18017518520355225,
|
|
"eval_runtime": 0.6321,
|
|
"eval_samples_per_second": 393.895,
|
|
"eval_steps_per_second": 50.621,
|
|
"num_input_tokens_seen": 105024,
|
|
"step": 71
|
|
},
|
|
{
|
|
"epoch": 0.2669039145907473,
|
|
"grad_norm": 20.432262420654297,
|
|
"learning_rate": 2.624113475177305e-06,
|
|
"loss": 0.1733,
|
|
"num_input_tokens_seen": 110528,
|
|
"step": 75
|
|
},
|
|
{
|
|
"epoch": 0.2846975088967972,
|
|
"grad_norm": 4.497078895568848,
|
|
"learning_rate": 2.8014184397163125e-06,
|
|
"loss": 0.1469,
|
|
"num_input_tokens_seen": 117440,
|
|
"step": 80
|
|
},
|
|
{
|
|
"epoch": 0.302491103202847,
|
|
"grad_norm": 13.809417724609375,
|
|
"learning_rate": 2.978723404255319e-06,
|
|
"loss": 0.1417,
|
|
"num_input_tokens_seen": 125504,
|
|
"step": 85
|
|
},
|
|
{
|
|
"epoch": 0.3202846975088968,
|
|
"grad_norm": 47.335994720458984,
|
|
"learning_rate": 3.1560283687943267e-06,
|
|
"loss": 0.1887,
|
|
"num_input_tokens_seen": 132352,
|
|
"step": 90
|
|
},
|
|
{
|
|
"epoch": 0.33807829181494664,
|
|
"grad_norm": 33.13681411743164,
|
|
"learning_rate": 3.3333333333333333e-06,
|
|
"loss": 0.1507,
|
|
"num_input_tokens_seen": 139200,
|
|
"step": 95
|
|
},
|
|
{
|
|
"epoch": 0.35587188612099646,
|
|
"grad_norm": 21.207801818847656,
|
|
"learning_rate": 3.510638297872341e-06,
|
|
"loss": 0.0966,
|
|
"num_input_tokens_seen": 147904,
|
|
"step": 100
|
|
},
|
|
{
|
|
"epoch": 0.3736654804270463,
|
|
"grad_norm": 27.387420654296875,
|
|
"learning_rate": 3.6879432624113475e-06,
|
|
"loss": 0.2218,
|
|
"num_input_tokens_seen": 154240,
|
|
"step": 105
|
|
},
|
|
{
|
|
"epoch": 0.3914590747330961,
|
|
"grad_norm": 14.321903228759766,
|
|
"learning_rate": 3.865248226950355e-06,
|
|
"loss": 0.1246,
|
|
"num_input_tokens_seen": 161472,
|
|
"step": 110
|
|
},
|
|
{
|
|
"epoch": 0.4092526690391459,
|
|
"grad_norm": 24.01630401611328,
|
|
"learning_rate": 4.042553191489362e-06,
|
|
"loss": 0.1689,
|
|
"num_input_tokens_seen": 168192,
|
|
"step": 115
|
|
},
|
|
{
|
|
"epoch": 0.42704626334519574,
|
|
"grad_norm": 5.953591346740723,
|
|
"learning_rate": 4.219858156028369e-06,
|
|
"loss": 0.1818,
|
|
"num_input_tokens_seen": 174656,
|
|
"step": 120
|
|
},
|
|
{
|
|
"epoch": 0.44483985765124556,
|
|
"grad_norm": 7.9348039627075195,
|
|
"learning_rate": 4.397163120567377e-06,
|
|
"loss": 0.1189,
|
|
"num_input_tokens_seen": 181632,
|
|
"step": 125
|
|
},
|
|
{
|
|
"epoch": 0.4626334519572954,
|
|
"grad_norm": 31.200048446655273,
|
|
"learning_rate": 4.574468085106383e-06,
|
|
"loss": 0.0973,
|
|
"num_input_tokens_seen": 191488,
|
|
"step": 130
|
|
},
|
|
{
|
|
"epoch": 0.4804270462633452,
|
|
"grad_norm": 20.26383399963379,
|
|
"learning_rate": 4.751773049645391e-06,
|
|
"loss": 0.1978,
|
|
"num_input_tokens_seen": 198848,
|
|
"step": 135
|
|
},
|
|
{
|
|
"epoch": 0.498220640569395,
|
|
"grad_norm": 34.11137771606445,
|
|
"learning_rate": 4.929078014184397e-06,
|
|
"loss": 0.1861,
|
|
"num_input_tokens_seen": 207232,
|
|
"step": 140
|
|
},
|
|
{
|
|
"epoch": 0.505338078291815,
|
|
"eval_loss": 0.2461702525615692,
|
|
"eval_runtime": 0.5958,
|
|
"eval_samples_per_second": 417.954,
|
|
"eval_steps_per_second": 53.713,
|
|
"num_input_tokens_seen": 209536,
|
|
"step": 142
|
|
},
|
|
{
|
|
"epoch": 0.5160142348754448,
|
|
"grad_norm": 5.17560338973999,
|
|
"learning_rate": 4.999930504592181e-06,
|
|
"loss": 0.2676,
|
|
"num_input_tokens_seen": 213952,
|
|
"step": 145
|
|
},
|
|
{
|
|
"epoch": 0.5338078291814946,
|
|
"grad_norm": 4.064759731292725,
|
|
"learning_rate": 4.999505824425164e-06,
|
|
"loss": 0.1686,
|
|
"num_input_tokens_seen": 221376,
|
|
"step": 150
|
|
},
|
|
{
|
|
"epoch": 0.5516014234875445,
|
|
"grad_norm": 6.850118160247803,
|
|
"learning_rate": 4.998695138156149e-06,
|
|
"loss": 0.1074,
|
|
"num_input_tokens_seen": 228928,
|
|
"step": 155
|
|
},
|
|
{
|
|
"epoch": 0.5693950177935944,
|
|
"grad_norm": 12.681975364685059,
|
|
"learning_rate": 4.997498570981822e-06,
|
|
"loss": 0.1216,
|
|
"num_input_tokens_seen": 236352,
|
|
"step": 160
|
|
},
|
|
{
|
|
"epoch": 0.5871886120996441,
|
|
"grad_norm": 20.49947166442871,
|
|
"learning_rate": 4.995916307691601e-06,
|
|
"loss": 0.1426,
|
|
"num_input_tokens_seen": 244416,
|
|
"step": 165
|
|
},
|
|
{
|
|
"epoch": 0.604982206405694,
|
|
"grad_norm": 17.282085418701172,
|
|
"learning_rate": 4.993948592639105e-06,
|
|
"loss": 0.175,
|
|
"num_input_tokens_seen": 251456,
|
|
"step": 170
|
|
},
|
|
{
|
|
"epoch": 0.6227758007117438,
|
|
"grad_norm": 12.41250991821289,
|
|
"learning_rate": 4.991595729704405e-06,
|
|
"loss": 0.1179,
|
|
"num_input_tokens_seen": 258880,
|
|
"step": 175
|
|
},
|
|
{
|
|
"epoch": 0.6405693950177936,
|
|
"grad_norm": 28.129770278930664,
|
|
"learning_rate": 4.988858082247109e-06,
|
|
"loss": 0.1235,
|
|
"num_input_tokens_seen": 265152,
|
|
"step": 180
|
|
},
|
|
{
|
|
"epoch": 0.6583629893238434,
|
|
"grad_norm": 27.979137420654297,
|
|
"learning_rate": 4.985736073050237e-06,
|
|
"loss": 0.1596,
|
|
"num_input_tokens_seen": 272576,
|
|
"step": 185
|
|
},
|
|
{
|
|
"epoch": 0.6761565836298933,
|
|
"grad_norm": 5.348273754119873,
|
|
"learning_rate": 4.982230184254934e-06,
|
|
"loss": 0.1188,
|
|
"num_input_tokens_seen": 279744,
|
|
"step": 190
|
|
},
|
|
{
|
|
"epoch": 0.693950177935943,
|
|
"grad_norm": 21.296180725097656,
|
|
"learning_rate": 4.9783409572860105e-06,
|
|
"loss": 0.1255,
|
|
"num_input_tokens_seen": 287680,
|
|
"step": 195
|
|
},
|
|
{
|
|
"epoch": 0.7117437722419929,
|
|
"grad_norm": 8.518939018249512,
|
|
"learning_rate": 4.9740689927683314e-06,
|
|
"loss": 0.0801,
|
|
"num_input_tokens_seen": 294592,
|
|
"step": 200
|
|
},
|
|
{
|
|
"epoch": 0.7295373665480427,
|
|
"grad_norm": 22.954431533813477,
|
|
"learning_rate": 4.9694149504340515e-06,
|
|
"loss": 0.0902,
|
|
"num_input_tokens_seen": 301440,
|
|
"step": 205
|
|
},
|
|
{
|
|
"epoch": 0.7473309608540926,
|
|
"grad_norm": 6.020153999328613,
|
|
"learning_rate": 4.964379549020741e-06,
|
|
"loss": 0.0658,
|
|
"num_input_tokens_seen": 308416,
|
|
"step": 210
|
|
},
|
|
{
|
|
"epoch": 0.7580071174377224,
|
|
"eval_loss": 0.1589389145374298,
|
|
"eval_runtime": 1.701,
|
|
"eval_samples_per_second": 146.387,
|
|
"eval_steps_per_second": 18.813,
|
|
"num_input_tokens_seen": 312576,
|
|
"step": 213
|
|
},
|
|
{
|
|
"epoch": 0.7651245551601423,
|
|
"grad_norm": 22.559890747070312,
|
|
"learning_rate": 4.9589635661603845e-06,
|
|
"loss": 0.1047,
|
|
"num_input_tokens_seen": 315328,
|
|
"step": 215
|
|
},
|
|
{
|
|
"epoch": 0.7829181494661922,
|
|
"grad_norm": 21.644763946533203,
|
|
"learning_rate": 4.953167838259285e-06,
|
|
"loss": 0.0899,
|
|
"num_input_tokens_seen": 322688,
|
|
"step": 220
|
|
},
|
|
{
|
|
"epoch": 0.800711743772242,
|
|
"grad_norm": 21.93659782409668,
|
|
"learning_rate": 4.946993260368904e-06,
|
|
"loss": 0.1884,
|
|
"num_input_tokens_seen": 329280,
|
|
"step": 225
|
|
},
|
|
{
|
|
"epoch": 0.8185053380782918,
|
|
"grad_norm": 4.014484405517578,
|
|
"learning_rate": 4.9404407860476275e-06,
|
|
"loss": 0.0862,
|
|
"num_input_tokens_seen": 336896,
|
|
"step": 230
|
|
},
|
|
{
|
|
"epoch": 0.8362989323843416,
|
|
"grad_norm": 9.570290565490723,
|
|
"learning_rate": 4.933511427213511e-06,
|
|
"loss": 0.1129,
|
|
"num_input_tokens_seen": 344128,
|
|
"step": 235
|
|
},
|
|
{
|
|
"epoch": 0.8540925266903915,
|
|
"grad_norm": 8.622743606567383,
|
|
"learning_rate": 4.926206253988001e-06,
|
|
"loss": 0.0612,
|
|
"num_input_tokens_seen": 350912,
|
|
"step": 240
|
|
},
|
|
{
|
|
"epoch": 0.8718861209964412,
|
|
"grad_norm": 20.623458862304688,
|
|
"learning_rate": 4.91852639453068e-06,
|
|
"loss": 0.1364,
|
|
"num_input_tokens_seen": 358016,
|
|
"step": 245
|
|
},
|
|
{
|
|
"epoch": 0.8896797153024911,
|
|
"grad_norm": 20.129985809326172,
|
|
"learning_rate": 4.910473034865033e-06,
|
|
"loss": 0.0751,
|
|
"num_input_tokens_seen": 364736,
|
|
"step": 250
|
|
},
|
|
{
|
|
"epoch": 0.9074733096085409,
|
|
"grad_norm": 38.003440856933594,
|
|
"learning_rate": 4.902047418695293e-06,
|
|
"loss": 0.1051,
|
|
"num_input_tokens_seen": 371648,
|
|
"step": 255
|
|
},
|
|
{
|
|
"epoch": 0.9252669039145908,
|
|
"grad_norm": 12.691219329833984,
|
|
"learning_rate": 4.893250847214369e-06,
|
|
"loss": 0.0675,
|
|
"num_input_tokens_seen": 379200,
|
|
"step": 260
|
|
},
|
|
{
|
|
"epoch": 0.9430604982206405,
|
|
"grad_norm": 13.248903274536133,
|
|
"learning_rate": 4.884084678902898e-06,
|
|
"loss": 0.1611,
|
|
"num_input_tokens_seen": 387200,
|
|
"step": 265
|
|
},
|
|
{
|
|
"epoch": 0.9608540925266904,
|
|
"grad_norm": 19.279905319213867,
|
|
"learning_rate": 4.874550329319457e-06,
|
|
"loss": 0.1232,
|
|
"num_input_tokens_seen": 395264,
|
|
"step": 270
|
|
},
|
|
{
|
|
"epoch": 0.9786476868327402,
|
|
"grad_norm": 10.716254234313965,
|
|
"learning_rate": 4.864649270881944e-06,
|
|
"loss": 0.135,
|
|
"num_input_tokens_seen": 402176,
|
|
"step": 275
|
|
},
|
|
{
|
|
"epoch": 0.99644128113879,
|
|
"grad_norm": 10.358949661254883,
|
|
"learning_rate": 4.854383032640196e-06,
|
|
"loss": 0.0765,
|
|
"num_input_tokens_seen": 409984,
|
|
"step": 280
|
|
},
|
|
{
|
|
"epoch": 1.01067615658363,
|
|
"eval_loss": 0.11889845132827759,
|
|
"eval_runtime": 0.6175,
|
|
"eval_samples_per_second": 403.229,
|
|
"eval_steps_per_second": 51.821,
|
|
"num_input_tokens_seen": 414040,
|
|
"step": 284
|
|
},
|
|
{
|
|
"epoch": 1.0142348754448398,
|
|
"grad_norm": 19.193450927734375,
|
|
"learning_rate": 4.843753200039851e-06,
|
|
"loss": 0.0754,
|
|
"num_input_tokens_seen": 415256,
|
|
"step": 285
|
|
},
|
|
{
|
|
"epoch": 1.0320284697508897,
|
|
"grad_norm": 15.135278701782227,
|
|
"learning_rate": 4.832761414677502e-06,
|
|
"loss": 0.067,
|
|
"num_input_tokens_seen": 422808,
|
|
"step": 290
|
|
},
|
|
{
|
|
"epoch": 1.0498220640569396,
|
|
"grad_norm": 39.43471145629883,
|
|
"learning_rate": 4.821409374047184e-06,
|
|
"loss": 0.0342,
|
|
"num_input_tokens_seen": 430104,
|
|
"step": 295
|
|
},
|
|
{
|
|
"epoch": 1.0676156583629894,
|
|
"grad_norm": 88.81342315673828,
|
|
"learning_rate": 4.809698831278217e-06,
|
|
"loss": 0.0683,
|
|
"num_input_tokens_seen": 436760,
|
|
"step": 300
|
|
},
|
|
{
|
|
"epoch": 1.085409252669039,
|
|
"grad_norm": 0.4617443382740021,
|
|
"learning_rate": 4.797631594864475e-06,
|
|
"loss": 0.1073,
|
|
"num_input_tokens_seen": 444952,
|
|
"step": 305
|
|
},
|
|
{
|
|
"epoch": 1.103202846975089,
|
|
"grad_norm": 40.835601806640625,
|
|
"learning_rate": 4.785209528385087e-06,
|
|
"loss": 0.1453,
|
|
"num_input_tokens_seen": 452760,
|
|
"step": 310
|
|
},
|
|
{
|
|
"epoch": 1.1209964412811388,
|
|
"grad_norm": 47.22534942626953,
|
|
"learning_rate": 4.7724345502166435e-06,
|
|
"loss": 0.2851,
|
|
"num_input_tokens_seen": 458392,
|
|
"step": 315
|
|
},
|
|
{
|
|
"epoch": 1.1387900355871885,
|
|
"grad_norm": 22.309940338134766,
|
|
"learning_rate": 4.759308633236934e-06,
|
|
"loss": 0.0427,
|
|
"num_input_tokens_seen": 465112,
|
|
"step": 320
|
|
},
|
|
{
|
|
"epoch": 1.1565836298932384,
|
|
"grad_norm": 0.2836686968803406,
|
|
"learning_rate": 4.74583380452027e-06,
|
|
"loss": 0.0369,
|
|
"num_input_tokens_seen": 472216,
|
|
"step": 325
|
|
},
|
|
{
|
|
"epoch": 1.1743772241992882,
|
|
"grad_norm": 20.73933219909668,
|
|
"learning_rate": 4.7320121450244395e-06,
|
|
"loss": 0.1237,
|
|
"num_input_tokens_seen": 479576,
|
|
"step": 330
|
|
},
|
|
{
|
|
"epoch": 1.1921708185053381,
|
|
"grad_norm": 0.6483802199363708,
|
|
"learning_rate": 4.717845789269333e-06,
|
|
"loss": 0.0822,
|
|
"num_input_tokens_seen": 486552,
|
|
"step": 335
|
|
},
|
|
{
|
|
"epoch": 1.209964412811388,
|
|
"grad_norm": 10.158080101013184,
|
|
"learning_rate": 4.703336925007311e-06,
|
|
"loss": 0.0835,
|
|
"num_input_tokens_seen": 494616,
|
|
"step": 340
|
|
},
|
|
{
|
|
"epoch": 1.2277580071174377,
|
|
"grad_norm": 22.884260177612305,
|
|
"learning_rate": 4.68848779288534e-06,
|
|
"loss": 0.0413,
|
|
"num_input_tokens_seen": 501400,
|
|
"step": 345
|
|
},
|
|
{
|
|
"epoch": 1.2455516014234875,
|
|
"grad_norm": 0.23994407057762146,
|
|
"learning_rate": 4.673300686098957e-06,
|
|
"loss": 0.0284,
|
|
"num_input_tokens_seen": 508888,
|
|
"step": 350
|
|
},
|
|
{
|
|
"epoch": 1.2633451957295374,
|
|
"grad_norm": 28.652849197387695,
|
|
"learning_rate": 4.657777950038133e-06,
|
|
"loss": 0.1848,
|
|
"num_input_tokens_seen": 517656,
|
|
"step": 355
|
|
},
|
|
{
|
|
"epoch": 1.2633451957295374,
|
|
"eval_loss": 0.21280953288078308,
|
|
"eval_runtime": 0.6154,
|
|
"eval_samples_per_second": 404.643,
|
|
"eval_steps_per_second": 52.002,
|
|
"num_input_tokens_seen": 517656,
|
|
"step": 355
|
|
},
|
|
{
|
|
"epoch": 1.281138790035587,
|
|
"grad_norm": 29.27073860168457,
|
|
"learning_rate": 4.641921981925064e-06,
|
|
"loss": 0.0684,
|
|
"num_input_tokens_seen": 526232,
|
|
"step": 360
|
|
},
|
|
{
|
|
"epoch": 1.298932384341637,
|
|
"grad_norm": 6.552826404571533,
|
|
"learning_rate": 4.625735230443959e-06,
|
|
"loss": 0.0556,
|
|
"num_input_tokens_seen": 533400,
|
|
"step": 365
|
|
},
|
|
{
|
|
"epoch": 1.3167259786476868,
|
|
"grad_norm": 19.556528091430664,
|
|
"learning_rate": 4.609220195362886e-06,
|
|
"loss": 0.1059,
|
|
"num_input_tokens_seen": 542168,
|
|
"step": 370
|
|
},
|
|
{
|
|
"epoch": 1.3345195729537367,
|
|
"grad_norm": 0.45224103331565857,
|
|
"learning_rate": 4.592379427147722e-06,
|
|
"loss": 0.0525,
|
|
"num_input_tokens_seen": 549976,
|
|
"step": 375
|
|
},
|
|
{
|
|
"epoch": 1.3523131672597866,
|
|
"grad_norm": 7.306344032287598,
|
|
"learning_rate": 4.575215526568278e-06,
|
|
"loss": 0.118,
|
|
"num_input_tokens_seen": 557016,
|
|
"step": 380
|
|
},
|
|
{
|
|
"epoch": 1.3701067615658362,
|
|
"grad_norm": 17.069351196289062,
|
|
"learning_rate": 4.557731144296659e-06,
|
|
"loss": 0.0831,
|
|
"num_input_tokens_seen": 564504,
|
|
"step": 385
|
|
},
|
|
{
|
|
"epoch": 1.387900355871886,
|
|
"grad_norm": 2.3357198238372803,
|
|
"learning_rate": 4.539928980497903e-06,
|
|
"loss": 0.0283,
|
|
"num_input_tokens_seen": 571864,
|
|
"step": 390
|
|
},
|
|
{
|
|
"epoch": 1.405693950177936,
|
|
"grad_norm": 18.61362648010254,
|
|
"learning_rate": 4.521811784412996e-06,
|
|
"loss": 0.0483,
|
|
"num_input_tokens_seen": 578456,
|
|
"step": 395
|
|
},
|
|
{
|
|
"epoch": 1.4234875444839858,
|
|
"grad_norm": 51.09469223022461,
|
|
"learning_rate": 4.503382353934295e-06,
|
|
"loss": 0.1155,
|
|
"num_input_tokens_seen": 584600,
|
|
"step": 400
|
|
},
|
|
{
|
|
"epoch": 1.4412811387900355,
|
|
"grad_norm": 0.02246745675802231,
|
|
"learning_rate": 4.484643535173438e-06,
|
|
"loss": 0.0086,
|
|
"num_input_tokens_seen": 591128,
|
|
"step": 405
|
|
},
|
|
{
|
|
"epoch": 1.4590747330960854,
|
|
"grad_norm": 0.0882764533162117,
|
|
"learning_rate": 4.465598222021818e-06,
|
|
"loss": 0.0969,
|
|
"num_input_tokens_seen": 598552,
|
|
"step": 410
|
|
},
|
|
{
|
|
"epoch": 1.4768683274021353,
|
|
"grad_norm": 33.67585372924805,
|
|
"learning_rate": 4.446249355703661e-06,
|
|
"loss": 0.0715,
|
|
"num_input_tokens_seen": 607320,
|
|
"step": 415
|
|
},
|
|
{
|
|
"epoch": 1.4946619217081851,
|
|
"grad_norm": 24.378616333007812,
|
|
"learning_rate": 4.426599924321815e-06,
|
|
"loss": 0.094,
|
|
"num_input_tokens_seen": 614744,
|
|
"step": 420
|
|
},
|
|
{
|
|
"epoch": 1.512455516014235,
|
|
"grad_norm": 9.61978530883789,
|
|
"learning_rate": 4.406652962396278e-06,
|
|
"loss": 0.0306,
|
|
"num_input_tokens_seen": 622808,
|
|
"step": 425
|
|
},
|
|
{
|
|
"epoch": 1.5160142348754448,
|
|
"eval_loss": 0.17913997173309326,
|
|
"eval_runtime": 0.6176,
|
|
"eval_samples_per_second": 403.205,
|
|
"eval_steps_per_second": 51.817,
|
|
"num_input_tokens_seen": 624344,
|
|
"step": 426
|
|
},
|
|
{
|
|
"epoch": 1.5302491103202847,
|
|
"grad_norm": 18.067750930786133,
|
|
"learning_rate": 4.386411550395576e-06,
|
|
"loss": 0.103,
|
|
"num_input_tokens_seen": 630488,
|
|
"step": 430
|
|
},
|
|
{
|
|
"epoch": 1.5480427046263345,
|
|
"grad_norm": 2.518416404724121,
|
|
"learning_rate": 4.365878814261032e-06,
|
|
"loss": 0.0775,
|
|
"num_input_tokens_seen": 638424,
|
|
"step": 435
|
|
},
|
|
{
|
|
"epoch": 1.5658362989323842,
|
|
"grad_norm": 17.197315216064453,
|
|
"learning_rate": 4.34505792492402e-06,
|
|
"loss": 0.0767,
|
|
"num_input_tokens_seen": 645208,
|
|
"step": 440
|
|
},
|
|
{
|
|
"epoch": 1.583629893238434,
|
|
"grad_norm": 1.7272682189941406,
|
|
"learning_rate": 4.3239520978162685e-06,
|
|
"loss": 0.013,
|
|
"num_input_tokens_seen": 653016,
|
|
"step": 445
|
|
},
|
|
{
|
|
"epoch": 1.601423487544484,
|
|
"grad_norm": 2.9054439067840576,
|
|
"learning_rate": 4.302564592373293e-06,
|
|
"loss": 0.0129,
|
|
"num_input_tokens_seen": 659992,
|
|
"step": 450
|
|
},
|
|
{
|
|
"epoch": 1.6192170818505338,
|
|
"grad_norm": 0.1965407431125641,
|
|
"learning_rate": 4.280898711531026e-06,
|
|
"loss": 0.1234,
|
|
"num_input_tokens_seen": 667224,
|
|
"step": 455
|
|
},
|
|
{
|
|
"epoch": 1.6370106761565837,
|
|
"grad_norm": 1.4352664947509766,
|
|
"learning_rate": 4.258957801215743e-06,
|
|
"loss": 0.1334,
|
|
"num_input_tokens_seen": 675160,
|
|
"step": 460
|
|
},
|
|
{
|
|
"epoch": 1.6548042704626336,
|
|
"grad_norm": 7.041714191436768,
|
|
"learning_rate": 4.236745249827336e-06,
|
|
"loss": 0.1297,
|
|
"num_input_tokens_seen": 683544,
|
|
"step": 465
|
|
},
|
|
{
|
|
"epoch": 1.6725978647686834,
|
|
"grad_norm": 22.762420654296875,
|
|
"learning_rate": 4.2142644877160334e-06,
|
|
"loss": 0.0334,
|
|
"num_input_tokens_seen": 689368,
|
|
"step": 470
|
|
},
|
|
{
|
|
"epoch": 1.690391459074733,
|
|
"grad_norm": 15.38829231262207,
|
|
"learning_rate": 4.191518986652642e-06,
|
|
"loss": 0.0779,
|
|
"num_input_tokens_seen": 695832,
|
|
"step": 475
|
|
},
|
|
{
|
|
"epoch": 1.708185053380783,
|
|
"grad_norm": 5.282833576202393,
|
|
"learning_rate": 4.168512259292391e-06,
|
|
"loss": 0.0085,
|
|
"num_input_tokens_seen": 703128,
|
|
"step": 480
|
|
},
|
|
{
|
|
"epoch": 1.7259786476868326,
|
|
"grad_norm": 0.2967665493488312,
|
|
"learning_rate": 4.14524785863246e-06,
|
|
"loss": 0.0617,
|
|
"num_input_tokens_seen": 709528,
|
|
"step": 485
|
|
},
|
|
{
|
|
"epoch": 1.7437722419928825,
|
|
"grad_norm": 18.050338745117188,
|
|
"learning_rate": 4.121729377463285e-06,
|
|
"loss": 0.0537,
|
|
"num_input_tokens_seen": 716312,
|
|
"step": 490
|
|
},
|
|
{
|
|
"epoch": 1.7615658362989324,
|
|
"grad_norm": 34.60630416870117,
|
|
"learning_rate": 4.0979604478137045e-06,
|
|
"loss": 0.1029,
|
|
"num_input_tokens_seen": 722776,
|
|
"step": 495
|
|
},
|
|
{
|
|
"epoch": 1.7686832740213523,
|
|
"eval_loss": 0.13597266376018524,
|
|
"eval_runtime": 0.6055,
|
|
"eval_samples_per_second": 411.239,
|
|
"eval_steps_per_second": 52.85,
|
|
"num_input_tokens_seen": 725656,
|
|
"step": 497
|
|
},
|
|
{
|
|
"epoch": 1.7793594306049823,
|
|
"grad_norm": 19.958894729614258,
|
|
"learning_rate": 4.0739447403900605e-06,
|
|
"loss": 0.1142,
|
|
"num_input_tokens_seen": 729944,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 1.7971530249110321,
|
|
"grad_norm": 13.6007719039917,
|
|
"learning_rate": 4.0496859640093215e-06,
|
|
"loss": 0.0837,
|
|
"num_input_tokens_seen": 737112,
|
|
"step": 505
|
|
},
|
|
{
|
|
"epoch": 1.814946619217082,
|
|
"grad_norm": 2.8377349376678467,
|
|
"learning_rate": 4.025187865026311e-06,
|
|
"loss": 0.0124,
|
|
"num_input_tokens_seen": 744408,
|
|
"step": 510
|
|
},
|
|
{
|
|
"epoch": 1.8327402135231317,
|
|
"grad_norm": 4.554283618927002,
|
|
"learning_rate": 4.0004542267551585e-06,
|
|
"loss": 0.059,
|
|
"num_input_tokens_seen": 750488,
|
|
"step": 515
|
|
},
|
|
{
|
|
"epoch": 1.8505338078291815,
|
|
"grad_norm": 20.501800537109375,
|
|
"learning_rate": 3.975488868885022e-06,
|
|
"loss": 0.0662,
|
|
"num_input_tokens_seen": 757528,
|
|
"step": 520
|
|
},
|
|
{
|
|
"epoch": 1.8683274021352312,
|
|
"grad_norm": 37.35130310058594,
|
|
"learning_rate": 3.950295646890202e-06,
|
|
"loss": 0.0299,
|
|
"num_input_tokens_seen": 763736,
|
|
"step": 525
|
|
},
|
|
{
|
|
"epoch": 1.886120996441281,
|
|
"grad_norm": 39.31193161010742,
|
|
"learning_rate": 3.924878451434736e-06,
|
|
"loss": 0.0666,
|
|
"num_input_tokens_seen": 771864,
|
|
"step": 530
|
|
},
|
|
{
|
|
"epoch": 1.903914590747331,
|
|
"grad_norm": 1.3388175964355469,
|
|
"learning_rate": 3.899241207771546e-06,
|
|
"loss": 0.072,
|
|
"num_input_tokens_seen": 778712,
|
|
"step": 535
|
|
},
|
|
{
|
|
"epoch": 1.9217081850533808,
|
|
"grad_norm": 12.562256813049316,
|
|
"learning_rate": 3.873387875136252e-06,
|
|
"loss": 0.0475,
|
|
"num_input_tokens_seen": 784280,
|
|
"step": 540
|
|
},
|
|
{
|
|
"epoch": 1.9395017793594307,
|
|
"grad_norm": 20.693452835083008,
|
|
"learning_rate": 3.847322446135736e-06,
|
|
"loss": 0.1443,
|
|
"num_input_tokens_seen": 792280,
|
|
"step": 545
|
|
},
|
|
{
|
|
"epoch": 1.9572953736654806,
|
|
"grad_norm": 10.482217788696289,
|
|
"learning_rate": 3.821048946131549e-06,
|
|
"loss": 0.1501,
|
|
"num_input_tokens_seen": 798488,
|
|
"step": 550
|
|
},
|
|
{
|
|
"epoch": 1.9750889679715302,
|
|
"grad_norm": 0.7575608491897583,
|
|
"learning_rate": 3.794571432618267e-06,
|
|
"loss": 0.0502,
|
|
"num_input_tokens_seen": 806104,
|
|
"step": 555
|
|
},
|
|
{
|
|
"epoch": 1.99288256227758,
|
|
"grad_norm": 13.330556869506836,
|
|
"learning_rate": 3.767893994596876e-06,
|
|
"loss": 0.0142,
|
|
"num_input_tokens_seen": 813336,
|
|
"step": 560
|
|
},
|
|
{
|
|
"epoch": 2.0106761565836297,
|
|
"grad_norm": 0.28736042976379395,
|
|
"learning_rate": 3.7410207519432972e-06,
|
|
"loss": 0.1868,
|
|
"num_input_tokens_seen": 817576,
|
|
"step": 565
|
|
},
|
|
{
|
|
"epoch": 2.02135231316726,
|
|
"eval_loss": 0.16060441732406616,
|
|
"eval_runtime": 0.6313,
|
|
"eval_samples_per_second": 394.424,
|
|
"eval_steps_per_second": 50.689,
|
|
"num_input_tokens_seen": 821416,
|
|
"step": 568
|
|
},
|
|
{
|
|
"epoch": 2.0284697508896796,
|
|
"grad_norm": 35.438270568847656,
|
|
"learning_rate": 3.713955854772144e-06,
|
|
"loss": 0.0138,
|
|
"num_input_tokens_seen": 823848,
|
|
"step": 570
|
|
},
|
|
{
|
|
"epoch": 2.0462633451957295,
|
|
"grad_norm": 23.84980010986328,
|
|
"learning_rate": 3.686703482795802e-06,
|
|
"loss": 0.1069,
|
|
"num_input_tokens_seen": 832232,
|
|
"step": 575
|
|
},
|
|
{
|
|
"epoch": 2.0640569395017794,
|
|
"grad_norm": 0.2273331582546234,
|
|
"learning_rate": 3.6592678446789516e-06,
|
|
"loss": 0.042,
|
|
"num_input_tokens_seen": 840424,
|
|
"step": 580
|
|
},
|
|
{
|
|
"epoch": 2.0818505338078293,
|
|
"grad_norm": 0.2911248505115509,
|
|
"learning_rate": 3.631653177388605e-06,
|
|
"loss": 0.0325,
|
|
"num_input_tokens_seen": 846824,
|
|
"step": 585
|
|
},
|
|
{
|
|
"epoch": 2.099644128113879,
|
|
"grad_norm": 0.03189073130488396,
|
|
"learning_rate": 3.6038637455397802e-06,
|
|
"loss": 0.003,
|
|
"num_input_tokens_seen": 853608,
|
|
"step": 590
|
|
},
|
|
{
|
|
"epoch": 2.117437722419929,
|
|
"grad_norm": 0.06126879155635834,
|
|
"learning_rate": 3.575903840736906e-06,
|
|
"loss": 0.1002,
|
|
"num_input_tokens_seen": 860968,
|
|
"step": 595
|
|
},
|
|
{
|
|
"epoch": 2.135231316725979,
|
|
"grad_norm": 0.12086429446935654,
|
|
"learning_rate": 3.547777780911055e-06,
|
|
"loss": 0.0251,
|
|
"num_input_tokens_seen": 868904,
|
|
"step": 600
|
|
},
|
|
{
|
|
"epoch": 2.1530249110320283,
|
|
"grad_norm": 0.013016794808208942,
|
|
"learning_rate": 3.519489909653113e-06,
|
|
"loss": 0.0005,
|
|
"num_input_tokens_seen": 876072,
|
|
"step": 605
|
|
},
|
|
{
|
|
"epoch": 2.170818505338078,
|
|
"grad_norm": 41.4961051940918,
|
|
"learning_rate": 3.4910445955429856e-06,
|
|
"loss": 0.0155,
|
|
"num_input_tokens_seen": 883752,
|
|
"step": 610
|
|
},
|
|
{
|
|
"epoch": 2.188612099644128,
|
|
"grad_norm": 0.10037070512771606,
|
|
"learning_rate": 3.4624462314749447e-06,
|
|
"loss": 0.0004,
|
|
"num_input_tokens_seen": 891304,
|
|
"step": 615
|
|
},
|
|
{
|
|
"epoch": 2.206405693950178,
|
|
"grad_norm": 0.04163924232125282,
|
|
"learning_rate": 3.433699233979222e-06,
|
|
"loss": 0.0163,
|
|
"num_input_tokens_seen": 899176,
|
|
"step": 620
|
|
},
|
|
{
|
|
"epoch": 2.224199288256228,
|
|
"grad_norm": 0.09111510962247849,
|
|
"learning_rate": 3.4048080425399506e-06,
|
|
"loss": 0.0001,
|
|
"num_input_tokens_seen": 907560,
|
|
"step": 625
|
|
},
|
|
{
|
|
"epoch": 2.2419928825622777,
|
|
"grad_norm": 0.023458287119865417,
|
|
"learning_rate": 3.375777118909561e-06,
|
|
"loss": 0.0027,
|
|
"num_input_tokens_seen": 915240,
|
|
"step": 630
|
|
},
|
|
{
|
|
"epoch": 2.2597864768683276,
|
|
"grad_norm": 0.007639422547072172,
|
|
"learning_rate": 3.346610946419743e-06,
|
|
"loss": 0.0259,
|
|
"num_input_tokens_seen": 921384,
|
|
"step": 635
|
|
},
|
|
{
|
|
"epoch": 2.2740213523131674,
|
|
"eval_loss": 0.25424328446388245,
|
|
"eval_runtime": 0.622,
|
|
"eval_samples_per_second": 400.338,
|
|
"eval_steps_per_second": 51.449,
|
|
"num_input_tokens_seen": 926760,
|
|
"step": 639
|
|
},
|
|
{
|
|
"epoch": 2.277580071174377,
|
|
"grad_norm": 0.02619522623717785,
|
|
"learning_rate": 3.3173140292890673e-06,
|
|
"loss": 0.0233,
|
|
"num_input_tokens_seen": 927528,
|
|
"step": 640
|
|
},
|
|
{
|
|
"epoch": 2.295373665480427,
|
|
"grad_norm": 0.031525298953056335,
|
|
"learning_rate": 3.2878908919273867e-06,
|
|
"loss": 0.0266,
|
|
"num_input_tokens_seen": 934568,
|
|
"step": 645
|
|
},
|
|
{
|
|
"epoch": 2.3131672597864767,
|
|
"grad_norm": 0.08830317109823227,
|
|
"learning_rate": 3.2583460782371217e-06,
|
|
"loss": 0.0006,
|
|
"num_input_tokens_seen": 942248,
|
|
"step": 650
|
|
},
|
|
{
|
|
"epoch": 2.3309608540925266,
|
|
"grad_norm": 0.005108945071697235,
|
|
"learning_rate": 3.228684150911527e-06,
|
|
"loss": 0.0003,
|
|
"num_input_tokens_seen": 949096,
|
|
"step": 655
|
|
},
|
|
{
|
|
"epoch": 2.3487544483985765,
|
|
"grad_norm": 0.04930766299366951,
|
|
"learning_rate": 3.1989096907300634e-06,
|
|
"loss": 0.0082,
|
|
"num_input_tokens_seen": 955752,
|
|
"step": 660
|
|
},
|
|
{
|
|
"epoch": 2.3665480427046264,
|
|
"grad_norm": 0.06299162656068802,
|
|
"learning_rate": 3.1690272958509772e-06,
|
|
"loss": 0.0486,
|
|
"num_input_tokens_seen": 963176,
|
|
"step": 665
|
|
},
|
|
{
|
|
"epoch": 2.3843416370106763,
|
|
"grad_norm": 0.6633197665214539,
|
|
"learning_rate": 3.139041581101187e-06,
|
|
"loss": 0.0001,
|
|
"num_input_tokens_seen": 968232,
|
|
"step": 670
|
|
},
|
|
{
|
|
"epoch": 2.402135231316726,
|
|
"grad_norm": 59.2174186706543,
|
|
"learning_rate": 3.108957177263608e-06,
|
|
"loss": 0.0222,
|
|
"num_input_tokens_seen": 976552,
|
|
"step": 675
|
|
},
|
|
{
|
|
"epoch": 2.419928825622776,
|
|
"grad_norm": 0.0013539546635001898,
|
|
"learning_rate": 3.078778730362003e-06,
|
|
"loss": 0.0075,
|
|
"num_input_tokens_seen": 983720,
|
|
"step": 680
|
|
},
|
|
{
|
|
"epoch": 2.4377224199288254,
|
|
"grad_norm": 0.01872558705508709,
|
|
"learning_rate": 3.0485109009434844e-06,
|
|
"loss": 0.0003,
|
|
"num_input_tokens_seen": 991976,
|
|
"step": 685
|
|
},
|
|
{
|
|
"epoch": 2.4555160142348753,
|
|
"grad_norm": 0.06228525564074516,
|
|
"learning_rate": 3.018158363358773e-06,
|
|
"loss": 0.061,
|
|
"num_input_tokens_seen": 998184,
|
|
"step": 690
|
|
},
|
|
{
|
|
"epoch": 2.473309608540925,
|
|
"grad_norm": 0.015886032953858376,
|
|
"learning_rate": 2.9877258050403214e-06,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1005672,
|
|
"step": 695
|
|
},
|
|
{
|
|
"epoch": 2.491103202846975,
|
|
"grad_norm": 0.49660900235176086,
|
|
"learning_rate": 2.9572179257784215e-06,
|
|
"loss": 0.0152,
|
|
"num_input_tokens_seen": 1013096,
|
|
"step": 700
|
|
},
|
|
{
|
|
"epoch": 2.508896797153025,
|
|
"grad_norm": 0.01401636004447937,
|
|
"learning_rate": 2.9266394369954056e-06,
|
|
"loss": 0.0006,
|
|
"num_input_tokens_seen": 1019304,
|
|
"step": 705
|
|
},
|
|
{
|
|
"epoch": 2.526690391459075,
|
|
"grad_norm": 0.011517788283526897,
|
|
"learning_rate": 2.8959950610180376e-06,
|
|
"loss": 0.029,
|
|
"num_input_tokens_seen": 1025320,
|
|
"step": 710
|
|
},
|
|
{
|
|
"epoch": 2.526690391459075,
|
|
"eval_loss": 0.23608553409576416,
|
|
"eval_runtime": 0.6224,
|
|
"eval_samples_per_second": 400.086,
|
|
"eval_steps_per_second": 51.417,
|
|
"num_input_tokens_seen": 1025320,
|
|
"step": 710
|
|
},
|
|
{
|
|
"epoch": 2.5444839857651247,
|
|
"grad_norm": 0.09000097960233688,
|
|
"learning_rate": 2.865289530348243e-06,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1032552,
|
|
"step": 715
|
|
},
|
|
{
|
|
"epoch": 2.562277580071174,
|
|
"grad_norm": 0.0029357271268963814,
|
|
"learning_rate": 2.8345275869322432e-06,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1039912,
|
|
"step": 720
|
|
},
|
|
{
|
|
"epoch": 2.580071174377224,
|
|
"grad_norm": 0.00977697316557169,
|
|
"learning_rate": 2.8037139814282494e-06,
|
|
"loss": 0.0092,
|
|
"num_input_tokens_seen": 1047208,
|
|
"step": 725
|
|
},
|
|
{
|
|
"epoch": 2.597864768683274,
|
|
"grad_norm": 0.03362439572811127,
|
|
"learning_rate": 2.7728534724728027e-06,
|
|
"loss": 0.0001,
|
|
"num_input_tokens_seen": 1053928,
|
|
"step": 730
|
|
},
|
|
{
|
|
"epoch": 2.6156583629893237,
|
|
"grad_norm": 109.98116302490234,
|
|
"learning_rate": 2.741950825945881e-06,
|
|
"loss": 0.0806,
|
|
"num_input_tokens_seen": 1061608,
|
|
"step": 735
|
|
},
|
|
{
|
|
"epoch": 2.6334519572953736,
|
|
"grad_norm": 0.026239968836307526,
|
|
"learning_rate": 2.7110108142348962e-06,
|
|
"loss": 0.0747,
|
|
"num_input_tokens_seen": 1067560,
|
|
"step": 740
|
|
},
|
|
{
|
|
"epoch": 2.6512455516014235,
|
|
"grad_norm": 0.13907551765441895,
|
|
"learning_rate": 2.6800382154976734e-06,
|
|
"loss": 0.0001,
|
|
"num_input_tokens_seen": 1074152,
|
|
"step": 745
|
|
},
|
|
{
|
|
"epoch": 2.6690391459074734,
|
|
"grad_norm": 0.02416655234992504,
|
|
"learning_rate": 2.64903781292455e-06,
|
|
"loss": 0.0005,
|
|
"num_input_tokens_seen": 1082856,
|
|
"step": 750
|
|
},
|
|
{
|
|
"epoch": 2.6868327402135233,
|
|
"grad_norm": 1.5046820640563965,
|
|
"learning_rate": 2.6180143939996926e-06,
|
|
"loss": 0.0003,
|
|
"num_input_tokens_seen": 1089512,
|
|
"step": 755
|
|
},
|
|
{
|
|
"epoch": 2.704626334519573,
|
|
"grad_norm": 0.027851078659296036,
|
|
"learning_rate": 2.5869727497617495e-06,
|
|
"loss": 0.0433,
|
|
"num_input_tokens_seen": 1096232,
|
|
"step": 760
|
|
},
|
|
{
|
|
"epoch": 2.722419928825623,
|
|
"grad_norm": 0.0055125863291323185,
|
|
"learning_rate": 2.55591767406396e-06,
|
|
"loss": 0.0004,
|
|
"num_input_tokens_seen": 1104168,
|
|
"step": 765
|
|
},
|
|
{
|
|
"epoch": 2.7402135231316724,
|
|
"grad_norm": 26.093463897705078,
|
|
"learning_rate": 2.524853962833825e-06,
|
|
"loss": 0.1194,
|
|
"num_input_tokens_seen": 1112232,
|
|
"step": 770
|
|
},
|
|
{
|
|
"epoch": 2.7580071174377223,
|
|
"grad_norm": 0.31872352957725525,
|
|
"learning_rate": 2.4937864133324514e-06,
|
|
"loss": 0.0031,
|
|
"num_input_tokens_seen": 1119016,
|
|
"step": 775
|
|
},
|
|
{
|
|
"epoch": 2.775800711743772,
|
|
"grad_norm": 0.7696136832237244,
|
|
"learning_rate": 2.462719823413707e-06,
|
|
"loss": 0.0005,
|
|
"num_input_tokens_seen": 1126696,
|
|
"step": 780
|
|
},
|
|
{
|
|
"epoch": 2.7793594306049823,
|
|
"eval_loss": 0.23524385690689087,
|
|
"eval_runtime": 1.141,
|
|
"eval_samples_per_second": 218.224,
|
|
"eval_steps_per_second": 28.045,
|
|
"num_input_tokens_seen": 1128104,
|
|
"step": 781
|
|
},
|
|
{
|
|
"epoch": 2.793594306049822,
|
|
"grad_norm": 0.014183886349201202,
|
|
"learning_rate": 2.4316589907832654e-06,
|
|
"loss": 0.0423,
|
|
"num_input_tokens_seen": 1134184,
|
|
"step": 785
|
|
},
|
|
{
|
|
"epoch": 2.811387900355872,
|
|
"grad_norm": 0.04211263358592987,
|
|
"learning_rate": 2.4006087122576867e-06,
|
|
"loss": 0.001,
|
|
"num_input_tokens_seen": 1140392,
|
|
"step": 790
|
|
},
|
|
{
|
|
"epoch": 2.829181494661922,
|
|
"grad_norm": 0.09497106820344925,
|
|
"learning_rate": 2.3695737830236263e-06,
|
|
"loss": 0.031,
|
|
"num_input_tokens_seen": 1148328,
|
|
"step": 795
|
|
},
|
|
{
|
|
"epoch": 2.8469750889679717,
|
|
"grad_norm": 0.040660299360752106,
|
|
"learning_rate": 2.3385589958973073e-06,
|
|
"loss": 0.0007,
|
|
"num_input_tokens_seen": 1154024,
|
|
"step": 800
|
|
},
|
|
{
|
|
"epoch": 2.864768683274021,
|
|
"grad_norm": 0.06477546691894531,
|
|
"learning_rate": 2.3075691405843435e-06,
|
|
"loss": 0.0003,
|
|
"num_input_tokens_seen": 1160808,
|
|
"step": 805
|
|
},
|
|
{
|
|
"epoch": 2.882562277580071,
|
|
"grad_norm": 0.022568199783563614,
|
|
"learning_rate": 2.2766090029400573e-06,
|
|
"loss": 0.0299,
|
|
"num_input_tokens_seen": 1167912,
|
|
"step": 810
|
|
},
|
|
{
|
|
"epoch": 2.900355871886121,
|
|
"grad_norm": 0.012872009538114071,
|
|
"learning_rate": 2.2456833642303825e-06,
|
|
"loss": 0.0001,
|
|
"num_input_tokens_seen": 1174568,
|
|
"step": 815
|
|
},
|
|
{
|
|
"epoch": 2.9181494661921707,
|
|
"grad_norm": 0.0069219921715557575,
|
|
"learning_rate": 2.214797000393479e-06,
|
|
"loss": 0.0001,
|
|
"num_input_tokens_seen": 1181480,
|
|
"step": 820
|
|
},
|
|
{
|
|
"epoch": 2.9359430604982206,
|
|
"grad_norm": 74.77655792236328,
|
|
"learning_rate": 2.183954681302173e-06,
|
|
"loss": 0.0251,
|
|
"num_input_tokens_seen": 1189928,
|
|
"step": 825
|
|
},
|
|
{
|
|
"epoch": 2.9537366548042705,
|
|
"grad_norm": 0.00504131056368351,
|
|
"learning_rate": 2.15316117002733e-06,
|
|
"loss": 0.0001,
|
|
"num_input_tokens_seen": 1197480,
|
|
"step": 830
|
|
},
|
|
{
|
|
"epoch": 2.9715302491103204,
|
|
"grad_norm": 0.014680106192827225,
|
|
"learning_rate": 2.122421222102278e-06,
|
|
"loss": 0.039,
|
|
"num_input_tokens_seen": 1204584,
|
|
"step": 835
|
|
},
|
|
{
|
|
"epoch": 2.9893238434163703,
|
|
"grad_norm": 32.642799377441406,
|
|
"learning_rate": 2.0917395847884e-06,
|
|
"loss": 0.0311,
|
|
"num_input_tokens_seen": 1212584,
|
|
"step": 840
|
|
},
|
|
{
|
|
"epoch": 3.00711743772242,
|
|
"grad_norm": 55.226280212402344,
|
|
"learning_rate": 2.061120996341996e-06,
|
|
"loss": 0.0104,
|
|
"num_input_tokens_seen": 1217856,
|
|
"step": 845
|
|
},
|
|
{
|
|
"epoch": 3.0249110320284696,
|
|
"grad_norm": 0.005825079046189785,
|
|
"learning_rate": 2.030570185282544e-06,
|
|
"loss": 0.0001,
|
|
"num_input_tokens_seen": 1226624,
|
|
"step": 850
|
|
},
|
|
{
|
|
"epoch": 3.0320284697508897,
|
|
"eval_loss": 0.25802624225616455,
|
|
"eval_runtime": 0.6232,
|
|
"eval_samples_per_second": 399.564,
|
|
"eval_steps_per_second": 51.35,
|
|
"num_input_tokens_seen": 1229440,
|
|
"step": 852
|
|
},
|
|
{
|
|
"epoch": 3.0427046263345194,
|
|
"grad_norm": 0.0032935605850070715,
|
|
"learning_rate": 2.0000918696624587e-06,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1233152,
|
|
"step": 855
|
|
},
|
|
{
|
|
"epoch": 3.0604982206405693,
|
|
"grad_norm": 0.024462368339300156,
|
|
"learning_rate": 1.9696907563384687e-06,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1240128,
|
|
"step": 860
|
|
},
|
|
{
|
|
"epoch": 3.078291814946619,
|
|
"grad_norm": 0.004567572381347418,
|
|
"learning_rate": 1.9393715402447228e-06,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1248064,
|
|
"step": 865
|
|
},
|
|
{
|
|
"epoch": 3.096085409252669,
|
|
"grad_norm": 0.008781103417277336,
|
|
"learning_rate": 1.9091389036677384e-06,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1255232,
|
|
"step": 870
|
|
},
|
|
{
|
|
"epoch": 3.113879003558719,
|
|
"grad_norm": 0.003714508144184947,
|
|
"learning_rate": 1.878997515523299e-06,
|
|
"loss": 0.0486,
|
|
"num_input_tokens_seen": 1262272,
|
|
"step": 875
|
|
},
|
|
{
|
|
"epoch": 3.131672597864769,
|
|
"grad_norm": 0.0036656211595982313,
|
|
"learning_rate": 1.8489520306354243e-06,
|
|
"loss": 0.0001,
|
|
"num_input_tokens_seen": 1269632,
|
|
"step": 880
|
|
},
|
|
{
|
|
"epoch": 3.1494661921708187,
|
|
"grad_norm": 0.06997384876012802,
|
|
"learning_rate": 1.8190070890175082e-06,
|
|
"loss": 0.0001,
|
|
"num_input_tokens_seen": 1277312,
|
|
"step": 885
|
|
},
|
|
{
|
|
"epoch": 3.167259786476868,
|
|
"grad_norm": 0.0681416317820549,
|
|
"learning_rate": 1.7891673151557493e-06,
|
|
"loss": 0.0502,
|
|
"num_input_tokens_seen": 1284096,
|
|
"step": 890
|
|
},
|
|
{
|
|
"epoch": 3.185053380782918,
|
|
"grad_norm": 0.008159984834492207,
|
|
"learning_rate": 1.7594373172949786e-06,
|
|
"loss": 0.0001,
|
|
"num_input_tokens_seen": 1291648,
|
|
"step": 895
|
|
},
|
|
{
|
|
"epoch": 3.202846975088968,
|
|
"grad_norm": 0.008854905143380165,
|
|
"learning_rate": 1.7298216867269906e-06,
|
|
"loss": 0.0001,
|
|
"num_input_tokens_seen": 1299712,
|
|
"step": 900
|
|
},
|
|
{
|
|
"epoch": 3.2206405693950177,
|
|
"grad_norm": 0.016451064497232437,
|
|
"learning_rate": 1.7003249970815028e-06,
|
|
"loss": 0.0001,
|
|
"num_input_tokens_seen": 1306176,
|
|
"step": 905
|
|
},
|
|
{
|
|
"epoch": 3.2384341637010676,
|
|
"grad_norm": 0.010798790492117405,
|
|
"learning_rate": 1.6709518036198307e-06,
|
|
"loss": 0.0001,
|
|
"num_input_tokens_seen": 1314112,
|
|
"step": 910
|
|
},
|
|
{
|
|
"epoch": 3.2562277580071175,
|
|
"grad_norm": 13.17015266418457,
|
|
"learning_rate": 1.6417066425314088e-06,
|
|
"loss": 0.0251,
|
|
"num_input_tokens_seen": 1321088,
|
|
"step": 915
|
|
},
|
|
{
|
|
"epoch": 3.2740213523131674,
|
|
"grad_norm": 0.006120134145021439,
|
|
"learning_rate": 1.612594030233252e-06,
|
|
"loss": 0.0001,
|
|
"num_input_tokens_seen": 1328512,
|
|
"step": 920
|
|
},
|
|
{
|
|
"epoch": 3.284697508896797,
|
|
"eval_loss": 0.22950421273708344,
|
|
"eval_runtime": 0.6162,
|
|
"eval_samples_per_second": 404.1,
|
|
"eval_steps_per_second": 51.932,
|
|
"num_input_tokens_seen": 1332544,
|
|
"step": 923
|
|
},
|
|
{
|
|
"epoch": 3.2918149466192173,
|
|
"grad_norm": 0.008929496631026268,
|
|
"learning_rate": 1.5836184626724722e-06,
|
|
"loss": 0.0005,
|
|
"num_input_tokens_seen": 1336128,
|
|
"step": 925
|
|
},
|
|
{
|
|
"epoch": 3.309608540925267,
|
|
"grad_norm": 0.004387512803077698,
|
|
"learning_rate": 1.5547844146319547e-06,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1343552,
|
|
"step": 930
|
|
},
|
|
{
|
|
"epoch": 3.3274021352313166,
|
|
"grad_norm": 0.011066491715610027,
|
|
"learning_rate": 1.5260963390393075e-06,
|
|
"loss": 0.0383,
|
|
"num_input_tokens_seen": 1351552,
|
|
"step": 935
|
|
},
|
|
{
|
|
"epoch": 3.3451957295373664,
|
|
"grad_norm": 0.006294109858572483,
|
|
"learning_rate": 1.4975586662791783e-06,
|
|
"loss": 0.0002,
|
|
"num_input_tokens_seen": 1358272,
|
|
"step": 940
|
|
},
|
|
{
|
|
"epoch": 3.3629893238434163,
|
|
"grad_norm": 0.01710674725472927,
|
|
"learning_rate": 1.4691758035090603e-06,
|
|
"loss": 0.0001,
|
|
"num_input_tokens_seen": 1366784,
|
|
"step": 945
|
|
},
|
|
{
|
|
"epoch": 3.380782918149466,
|
|
"grad_norm": 0.006381936836987734,
|
|
"learning_rate": 1.4409521339786809e-06,
|
|
"loss": 0.0001,
|
|
"num_input_tokens_seen": 1373312,
|
|
"step": 950
|
|
},
|
|
{
|
|
"epoch": 3.398576512455516,
|
|
"grad_norm": 0.012110439129173756,
|
|
"learning_rate": 1.41289201635308e-06,
|
|
"loss": 0.0001,
|
|
"num_input_tokens_seen": 1380736,
|
|
"step": 955
|
|
},
|
|
{
|
|
"epoch": 3.416370106761566,
|
|
"grad_norm": 0.03375524654984474,
|
|
"learning_rate": 1.3849997840394943e-06,
|
|
"loss": 0.0001,
|
|
"num_input_tokens_seen": 1388544,
|
|
"step": 960
|
|
},
|
|
{
|
|
"epoch": 3.434163701067616,
|
|
"grad_norm": 0.010791816748678684,
|
|
"learning_rate": 1.3572797445181346e-06,
|
|
"loss": 0.0001,
|
|
"num_input_tokens_seen": 1396160,
|
|
"step": 965
|
|
},
|
|
{
|
|
"epoch": 3.4519572953736652,
|
|
"grad_norm": 0.006806841120123863,
|
|
"learning_rate": 1.3297361786769654e-06,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1404096,
|
|
"step": 970
|
|
},
|
|
{
|
|
"epoch": 3.469750889679715,
|
|
"grad_norm": 0.006986913271248341,
|
|
"learning_rate": 1.302373340150598e-06,
|
|
"loss": 0.0004,
|
|
"num_input_tokens_seen": 1411008,
|
|
"step": 975
|
|
},
|
|
{
|
|
"epoch": 3.487544483985765,
|
|
"grad_norm": 0.006954096723347902,
|
|
"learning_rate": 1.2751954546633872e-06,
|
|
"loss": 0.0001,
|
|
"num_input_tokens_seen": 1418880,
|
|
"step": 980
|
|
},
|
|
{
|
|
"epoch": 3.505338078291815,
|
|
"grad_norm": 0.0037973152939230204,
|
|
"learning_rate": 1.2482067193768419e-06,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1426048,
|
|
"step": 985
|
|
},
|
|
{
|
|
"epoch": 3.5231316725978647,
|
|
"grad_norm": 0.1800023317337036,
|
|
"learning_rate": 1.2214113022414448e-06,
|
|
"loss": 0.0001,
|
|
"num_input_tokens_seen": 1432064,
|
|
"step": 990
|
|
},
|
|
{
|
|
"epoch": 3.5373665480427046,
|
|
"eval_loss": 0.24046999216079712,
|
|
"eval_runtime": 0.6864,
|
|
"eval_samples_per_second": 362.784,
|
|
"eval_steps_per_second": 46.623,
|
|
"num_input_tokens_seen": 1438336,
|
|
"step": 994
|
|
},
|
|
{
|
|
"epoch": 3.5409252669039146,
|
|
"grad_norm": 0.008008265867829323,
|
|
"learning_rate": 1.1948133413529817e-06,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1439808,
|
|
"step": 995
|
|
},
|
|
{
|
|
"epoch": 3.5587188612099645,
|
|
"grad_norm": 0.004162625875324011,
|
|
"learning_rate": 1.168416944313486e-06,
|
|
"loss": 0.0001,
|
|
"num_input_tokens_seen": 1447616,
|
|
"step": 1000
|
|
},
|
|
{
|
|
"epoch": 3.5765124555160144,
|
|
"grad_norm": 0.0032155588269233704,
|
|
"learning_rate": 1.1422261875968845e-06,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1454208,
|
|
"step": 1005
|
|
},
|
|
{
|
|
"epoch": 3.5943060498220643,
|
|
"grad_norm": 0.007771800272166729,
|
|
"learning_rate": 1.1162451159194615e-06,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1463296,
|
|
"step": 1010
|
|
},
|
|
{
|
|
"epoch": 3.612099644128114,
|
|
"grad_norm": 0.12433820962905884,
|
|
"learning_rate": 1.0904777416152166e-06,
|
|
"loss": 0.0009,
|
|
"num_input_tokens_seen": 1469952,
|
|
"step": 1015
|
|
},
|
|
{
|
|
"epoch": 3.6298932384341636,
|
|
"grad_norm": 0.003224864834919572,
|
|
"learning_rate": 1.0649280440162326e-06,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1477184,
|
|
"step": 1020
|
|
},
|
|
{
|
|
"epoch": 3.6476868327402134,
|
|
"grad_norm": 0.00848240777850151,
|
|
"learning_rate": 1.0395999688381313e-06,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1484160,
|
|
"step": 1025
|
|
},
|
|
{
|
|
"epoch": 3.6654804270462633,
|
|
"grad_norm": 0.005719688255339861,
|
|
"learning_rate": 1.0144974275707243e-06,
|
|
"loss": 0.0001,
|
|
"num_input_tokens_seen": 1491200,
|
|
"step": 1030
|
|
},
|
|
{
|
|
"epoch": 3.683274021352313,
|
|
"grad_norm": 0.0028636339120566845,
|
|
"learning_rate": 9.896242968739538e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1498368,
|
|
"step": 1035
|
|
},
|
|
{
|
|
"epoch": 3.701067615658363,
|
|
"grad_norm": 0.0037352785002440214,
|
|
"learning_rate": 9.649844179792082e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1505984,
|
|
"step": 1040
|
|
},
|
|
{
|
|
"epoch": 3.718861209964413,
|
|
"grad_norm": 0.0014299631584435701,
|
|
"learning_rate": 9.405815960961054e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1511680,
|
|
"step": 1045
|
|
},
|
|
{
|
|
"epoch": 3.7366548042704624,
|
|
"grad_norm": 0.002295982325449586,
|
|
"learning_rate": 9.164195998248471e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1517888,
|
|
"step": 1050
|
|
},
|
|
{
|
|
"epoch": 3.7544483985765122,
|
|
"grad_norm": 0.0017496125074103475,
|
|
"learning_rate": 8.925021605742212e-07,
|
|
"loss": 0.0109,
|
|
"num_input_tokens_seen": 1525568,
|
|
"step": 1055
|
|
},
|
|
{
|
|
"epoch": 3.772241992882562,
|
|
"grad_norm": 0.0030260924249887466,
|
|
"learning_rate": 8.68832971985347e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1532480,
|
|
"step": 1060
|
|
},
|
|
{
|
|
"epoch": 3.790035587188612,
|
|
"grad_norm": 0.0027729899156838655,
|
|
"learning_rate": 8.454156893612592e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1539072,
|
|
"step": 1065
|
|
},
|
|
{
|
|
"epoch": 3.790035587188612,
|
|
"eval_loss": 0.2512344419956207,
|
|
"eval_runtime": 0.6676,
|
|
"eval_samples_per_second": 372.986,
|
|
"eval_steps_per_second": 47.934,
|
|
"num_input_tokens_seen": 1539072,
|
|
"step": 1065
|
|
},
|
|
{
|
|
"epoch": 3.807829181494662,
|
|
"grad_norm": 0.010402582585811615,
|
|
"learning_rate": 8.222539291024079e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1547584,
|
|
"step": 1070
|
|
},
|
|
{
|
|
"epoch": 3.8256227758007118,
|
|
"grad_norm": 0.0076858168467879295,
|
|
"learning_rate": 7.993512681481638e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1554304,
|
|
"step": 1075
|
|
},
|
|
{
|
|
"epoch": 3.8434163701067616,
|
|
"grad_norm": 0.005163070745766163,
|
|
"learning_rate": 7.767112434244254e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1560896,
|
|
"step": 1080
|
|
},
|
|
{
|
|
"epoch": 3.8612099644128115,
|
|
"grad_norm": 0.007071156986057758,
|
|
"learning_rate": 7.543373512973947e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1567744,
|
|
"step": 1085
|
|
},
|
|
{
|
|
"epoch": 3.8790035587188614,
|
|
"grad_norm": 0.004036191385239363,
|
|
"learning_rate": 7.322330470336314e-07,
|
|
"loss": 0.032,
|
|
"num_input_tokens_seen": 1574400,
|
|
"step": 1090
|
|
},
|
|
{
|
|
"epoch": 3.8967971530249113,
|
|
"grad_norm": 0.001716184546239674,
|
|
"learning_rate": 7.104017442664393e-07,
|
|
"loss": 0.0187,
|
|
"num_input_tokens_seen": 1581504,
|
|
"step": 1095
|
|
},
|
|
{
|
|
"epoch": 3.914590747330961,
|
|
"grad_norm": 0.03370984271168709,
|
|
"learning_rate": 6.88846814468691e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1589504,
|
|
"step": 1100
|
|
},
|
|
{
|
|
"epoch": 3.9323843416370106,
|
|
"grad_norm": 19.512357711791992,
|
|
"learning_rate": 6.67571586432163e-07,
|
|
"loss": 0.0369,
|
|
"num_input_tokens_seen": 1597696,
|
|
"step": 1105
|
|
},
|
|
{
|
|
"epoch": 3.9501779359430604,
|
|
"grad_norm": 0.0021878909319639206,
|
|
"learning_rate": 6.465793457534553e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1605248,
|
|
"step": 1110
|
|
},
|
|
{
|
|
"epoch": 3.9679715302491103,
|
|
"grad_norm": 0.0022776706609874964,
|
|
"learning_rate": 6.258733343265933e-07,
|
|
"loss": 0.0002,
|
|
"num_input_tokens_seen": 1613952,
|
|
"step": 1115
|
|
},
|
|
{
|
|
"epoch": 3.98576512455516,
|
|
"grad_norm": 0.010679907165467739,
|
|
"learning_rate": 6.054567498423683e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1620224,
|
|
"step": 1120
|
|
},
|
|
{
|
|
"epoch": 4.00355871886121,
|
|
"grad_norm": 0.003633465152233839,
|
|
"learning_rate": 5.853327452945115e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1625800,
|
|
"step": 1125
|
|
},
|
|
{
|
|
"epoch": 4.0213523131672595,
|
|
"grad_norm": 0.0048196627758443356,
|
|
"learning_rate": 5.655044284927658e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1633352,
|
|
"step": 1130
|
|
},
|
|
{
|
|
"epoch": 4.039145907473309,
|
|
"grad_norm": 0.0030679679475724697,
|
|
"learning_rate": 5.459748615829355e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1640840,
|
|
"step": 1135
|
|
},
|
|
{
|
|
"epoch": 4.04270462633452,
|
|
"eval_loss": 0.2551669180393219,
|
|
"eval_runtime": 1.6934,
|
|
"eval_samples_per_second": 147.043,
|
|
"eval_steps_per_second": 18.897,
|
|
"num_input_tokens_seen": 1642696,
|
|
"step": 1136
|
|
},
|
|
{
|
|
"epoch": 4.056939501779359,
|
|
"grad_norm": 0.00447492441162467,
|
|
"learning_rate": 5.267470605739953e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1648520,
|
|
"step": 1140
|
|
},
|
|
{
|
|
"epoch": 4.074733096085409,
|
|
"grad_norm": 0.0019011611584573984,
|
|
"learning_rate": 5.078239948723154e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1655752,
|
|
"step": 1145
|
|
},
|
|
{
|
|
"epoch": 4.092526690391459,
|
|
"grad_norm": 0.005783023778349161,
|
|
"learning_rate": 4.892085868230881e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1662920,
|
|
"step": 1150
|
|
},
|
|
{
|
|
"epoch": 4.110320284697509,
|
|
"grad_norm": 0.0052789985202252865,
|
|
"learning_rate": 4.7090371125902175e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1669896,
|
|
"step": 1155
|
|
},
|
|
{
|
|
"epoch": 4.128113879003559,
|
|
"grad_norm": 0.005814805161207914,
|
|
"learning_rate": 4.529121950563717e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1675400,
|
|
"step": 1160
|
|
},
|
|
{
|
|
"epoch": 4.145907473309609,
|
|
"grad_norm": 0.010795537382364273,
|
|
"learning_rate": 4.352368166983753e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1682952,
|
|
"step": 1165
|
|
},
|
|
{
|
|
"epoch": 4.1637010676156585,
|
|
"grad_norm": 0.013921362347900867,
|
|
"learning_rate": 4.178803058461664e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1690248,
|
|
"step": 1170
|
|
},
|
|
{
|
|
"epoch": 4.181494661921708,
|
|
"grad_norm": 0.00226654764264822,
|
|
"learning_rate": 4.0084534291722375e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1696840,
|
|
"step": 1175
|
|
},
|
|
{
|
|
"epoch": 4.199288256227758,
|
|
"grad_norm": 0.005677198059856892,
|
|
"learning_rate": 3.8413455867142513e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1703624,
|
|
"step": 1180
|
|
},
|
|
{
|
|
"epoch": 4.217081850533808,
|
|
"grad_norm": 0.004136247094720602,
|
|
"learning_rate": 3.6775053380477296e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1710024,
|
|
"step": 1185
|
|
},
|
|
{
|
|
"epoch": 4.234875444839858,
|
|
"grad_norm": 0.0022235463839024305,
|
|
"learning_rate": 3.516957985508476e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1717768,
|
|
"step": 1190
|
|
},
|
|
{
|
|
"epoch": 4.252669039145908,
|
|
"grad_norm": 0.0016525887185707688,
|
|
"learning_rate": 3.3597283229005877e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1727240,
|
|
"step": 1195
|
|
},
|
|
{
|
|
"epoch": 4.270462633451958,
|
|
"grad_norm": 0.0030577515717595816,
|
|
"learning_rate": 3.2058406316674563e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1734408,
|
|
"step": 1200
|
|
},
|
|
{
|
|
"epoch": 4.288256227758007,
|
|
"grad_norm": 0.004889196250587702,
|
|
"learning_rate": 3.055318677141916e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1740936,
|
|
"step": 1205
|
|
},
|
|
{
|
|
"epoch": 4.295373665480427,
|
|
"eval_loss": 0.257210373878479,
|
|
"eval_runtime": 0.6184,
|
|
"eval_samples_per_second": 402.654,
|
|
"eval_steps_per_second": 51.747,
|
|
"num_input_tokens_seen": 1743624,
|
|
"step": 1207
|
|
},
|
|
{
|
|
"epoch": 4.306049822064057,
|
|
"grad_norm": 0.010472727008163929,
|
|
"learning_rate": 2.9081857048761014e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1747784,
|
|
"step": 1210
|
|
},
|
|
{
|
|
"epoch": 4.3238434163701065,
|
|
"grad_norm": 0.002632361836731434,
|
|
"learning_rate": 2.764464437051537e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1754888,
|
|
"step": 1215
|
|
},
|
|
{
|
|
"epoch": 4.341637010676156,
|
|
"grad_norm": 0.003168675350025296,
|
|
"learning_rate": 2.624177068970124e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1762632,
|
|
"step": 1220
|
|
},
|
|
{
|
|
"epoch": 4.359430604982206,
|
|
"grad_norm": 0.001775842742063105,
|
|
"learning_rate": 2.4873452656264316e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1769928,
|
|
"step": 1225
|
|
},
|
|
{
|
|
"epoch": 4.377224199288256,
|
|
"grad_norm": 0.0023636040277779102,
|
|
"learning_rate": 2.3539901583619186e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1777480,
|
|
"step": 1230
|
|
},
|
|
{
|
|
"epoch": 4.395017793594306,
|
|
"grad_norm": 0.003756535705178976,
|
|
"learning_rate": 2.2241323416015452e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1784840,
|
|
"step": 1235
|
|
},
|
|
{
|
|
"epoch": 4.412811387900356,
|
|
"grad_norm": 0.0014317089226096869,
|
|
"learning_rate": 2.0977918696733103e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1792584,
|
|
"step": 1240
|
|
},
|
|
{
|
|
"epoch": 4.430604982206406,
|
|
"grad_norm": 0.0015214212471619248,
|
|
"learning_rate": 1.9749882537112297e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1800968,
|
|
"step": 1245
|
|
},
|
|
{
|
|
"epoch": 4.448398576512456,
|
|
"grad_norm": 0.003403919516131282,
|
|
"learning_rate": 1.8557404586421413e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1808456,
|
|
"step": 1250
|
|
},
|
|
{
|
|
"epoch": 4.4661921708185055,
|
|
"grad_norm": 0.013262175023555756,
|
|
"learning_rate": 1.7400669002569233e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1816136,
|
|
"step": 1255
|
|
},
|
|
{
|
|
"epoch": 4.483985765124555,
|
|
"grad_norm": 0.001888920902274549,
|
|
"learning_rate": 1.62798544236647e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1824136,
|
|
"step": 1260
|
|
},
|
|
{
|
|
"epoch": 4.501779359430605,
|
|
"grad_norm": 0.00637847138568759,
|
|
"learning_rate": 1.5195133940429345e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1831304,
|
|
"step": 1265
|
|
},
|
|
{
|
|
"epoch": 4.519572953736655,
|
|
"grad_norm": 0.0010983651736751199,
|
|
"learning_rate": 1.4146675069466403e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1837512,
|
|
"step": 1270
|
|
},
|
|
{
|
|
"epoch": 4.537366548042705,
|
|
"grad_norm": 0.0015767315635457635,
|
|
"learning_rate": 1.313463972739068e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1844296,
|
|
"step": 1275
|
|
},
|
|
{
|
|
"epoch": 4.548042704626335,
|
|
"eval_loss": 0.259037584066391,
|
|
"eval_runtime": 0.6197,
|
|
"eval_samples_per_second": 401.818,
|
|
"eval_steps_per_second": 51.639,
|
|
"num_input_tokens_seen": 1849416,
|
|
"step": 1278
|
|
},
|
|
{
|
|
"epoch": 4.555160142348754,
|
|
"grad_norm": 0.003252014284953475,
|
|
"learning_rate": 1.215918420582343e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1851720,
|
|
"step": 1280
|
|
},
|
|
{
|
|
"epoch": 4.572953736654805,
|
|
"grad_norm": 0.0038456034380942583,
|
|
"learning_rate": 1.1220459147255642e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1858120,
|
|
"step": 1285
|
|
},
|
|
{
|
|
"epoch": 4.590747330960854,
|
|
"grad_norm": 0.002327981637790799,
|
|
"learning_rate": 1.0318609521783818e-07,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1865928,
|
|
"step": 1290
|
|
},
|
|
{
|
|
"epoch": 4.608540925266904,
|
|
"grad_norm": 0.002249909332022071,
|
|
"learning_rate": 9.453774604721937e-08,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1873800,
|
|
"step": 1295
|
|
},
|
|
{
|
|
"epoch": 4.6263345195729535,
|
|
"grad_norm": 0.00320955878123641,
|
|
"learning_rate": 8.62608795509276e-08,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1881800,
|
|
"step": 1300
|
|
},
|
|
{
|
|
"epoch": 4.644128113879003,
|
|
"grad_norm": 0.010727422311902046,
|
|
"learning_rate": 7.835677395001795e-08,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1888648,
|
|
"step": 1305
|
|
},
|
|
{
|
|
"epoch": 4.661921708185053,
|
|
"grad_norm": 0.0015563094057142735,
|
|
"learning_rate": 7.082664989897486e-08,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1895432,
|
|
"step": 1310
|
|
},
|
|
{
|
|
"epoch": 4.679715302491103,
|
|
"grad_norm": 0.0029733225237578154,
|
|
"learning_rate": 6.367167029720234e-08,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1902408,
|
|
"step": 1315
|
|
},
|
|
{
|
|
"epoch": 4.697508896797153,
|
|
"grad_norm": 0.0011107242899015546,
|
|
"learning_rate": 5.68929401094323e-08,
|
|
"loss": 0.0277,
|
|
"num_input_tokens_seen": 1910344,
|
|
"step": 1320
|
|
},
|
|
{
|
|
"epoch": 4.715302491103203,
|
|
"grad_norm": 0.0024087605997920036,
|
|
"learning_rate": 5.049150619508503e-08,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1918472,
|
|
"step": 1325
|
|
},
|
|
{
|
|
"epoch": 4.733096085409253,
|
|
"grad_norm": 0.002195857698097825,
|
|
"learning_rate": 4.446835714659647e-08,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1924744,
|
|
"step": 1330
|
|
},
|
|
{
|
|
"epoch": 4.750889679715303,
|
|
"grad_norm": 0.0017717446899041533,
|
|
"learning_rate": 3.882442313674878e-08,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1932872,
|
|
"step": 1335
|
|
},
|
|
{
|
|
"epoch": 4.7686832740213525,
|
|
"grad_norm": 0.0018043630989268422,
|
|
"learning_rate": 3.3560575775019866e-08,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1940040,
|
|
"step": 1340
|
|
},
|
|
{
|
|
"epoch": 4.786476868327402,
|
|
"grad_norm": 0.0031781333964318037,
|
|
"learning_rate": 2.8677627972978905e-08,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1948936,
|
|
"step": 1345
|
|
},
|
|
{
|
|
"epoch": 4.800711743772242,
|
|
"eval_loss": 0.2602100372314453,
|
|
"eval_runtime": 0.6213,
|
|
"eval_samples_per_second": 400.766,
|
|
"eval_steps_per_second": 51.504,
|
|
"num_input_tokens_seen": 1954568,
|
|
"step": 1349
|
|
},
|
|
{
|
|
"epoch": 4.804270462633452,
|
|
"grad_norm": 0.001851449953392148,
|
|
"learning_rate": 2.4176333818745347e-08,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1955912,
|
|
"step": 1350
|
|
},
|
|
{
|
|
"epoch": 4.822064056939502,
|
|
"grad_norm": 0.003628035541623831,
|
|
"learning_rate": 2.0057388460533733e-08,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1962760,
|
|
"step": 1355
|
|
},
|
|
{
|
|
"epoch": 4.839857651245552,
|
|
"grad_norm": 0.0011652238899841905,
|
|
"learning_rate": 1.6321427999298754e-08,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1969160,
|
|
"step": 1360
|
|
},
|
|
{
|
|
"epoch": 4.857651245551601,
|
|
"grad_norm": 0.003991110250353813,
|
|
"learning_rate": 1.2969029390501597e-08,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1975752,
|
|
"step": 1365
|
|
},
|
|
{
|
|
"epoch": 4.875444839857651,
|
|
"grad_norm": 0.004909892100840807,
|
|
"learning_rate": 1.000071035500816e-08,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1983240,
|
|
"step": 1370
|
|
},
|
|
{
|
|
"epoch": 4.893238434163701,
|
|
"grad_norm": 0.00255265599116683,
|
|
"learning_rate": 7.416929299135511e-09,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1990792,
|
|
"step": 1375
|
|
},
|
|
{
|
|
"epoch": 4.911032028469751,
|
|
"grad_norm": 0.0013633174821734428,
|
|
"learning_rate": 5.218085243859639e-09,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 1998728,
|
|
"step": 1380
|
|
},
|
|
{
|
|
"epoch": 4.9288256227758005,
|
|
"grad_norm": 0.005876463372260332,
|
|
"learning_rate": 3.4045177631936154e-09,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 2006920,
|
|
"step": 1385
|
|
},
|
|
{
|
|
"epoch": 4.94661921708185,
|
|
"grad_norm": 0.0021194759756326675,
|
|
"learning_rate": 1.976506931745392e-09,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 2013128,
|
|
"step": 1390
|
|
},
|
|
{
|
|
"epoch": 4.9644128113879,
|
|
"grad_norm": 0.0024136609863489866,
|
|
"learning_rate": 9.3427328146517e-10,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 2021704,
|
|
"step": 1395
|
|
},
|
|
{
|
|
"epoch": 4.98220640569395,
|
|
"grad_norm": 0.16585078835487366,
|
|
"learning_rate": 2.7797776758903274e-10,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 2028872,
|
|
"step": 1400
|
|
},
|
|
{
|
|
"epoch": 5.0,
|
|
"grad_norm": 0.002416080329567194,
|
|
"learning_rate": 7.72174378022017e-12,
|
|
"loss": 0.0,
|
|
"num_input_tokens_seen": 2035272,
|
|
"step": 1405
|
|
},
|
|
{
|
|
"epoch": 5.0,
|
|
"num_input_tokens_seen": 2035272,
|
|
"step": 1405,
|
|
"total_flos": 1.1883702201974784e+16,
|
|
"train_loss": 0.05568007128206763,
|
|
"train_runtime": 1085.6649,
|
|
"train_samples_per_second": 10.321,
|
|
"train_steps_per_second": 1.294
|
|
}
|
|
],
|
|
"logging_steps": 5,
|
|
"max_steps": 1405,
|
|
"num_input_tokens_seen": 2035272,
|
|
"num_train_epochs": 5,
|
|
"save_steps": 71,
|
|
"stateful_callbacks": {
|
|
"TrainerControl": {
|
|
"args": {
|
|
"should_epoch_stop": false,
|
|
"should_evaluate": false,
|
|
"should_log": false,
|
|
"should_save": true,
|
|
"should_training_stop": true
|
|
},
|
|
"attributes": {}
|
|
}
|
|
},
|
|
"total_flos": 1.1883702201974784e+16,
|
|
"train_batch_size": 8,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|