mistral-7b-ift / trainer_state.json
NicholasCorrado's picture
Model save
cc241fa verified
raw
history blame
17.1 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.9989023051591658,
"eval_steps": 500,
"global_step": 455,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0021953896816684962,
"grad_norm": 37.23542178718614,
"learning_rate": 1.0869565217391303e-08,
"loss": 1.4277,
"step": 1
},
{
"epoch": 0.010976948408342482,
"grad_norm": 31.768758474830744,
"learning_rate": 5.434782608695652e-08,
"loss": 1.4875,
"step": 5
},
{
"epoch": 0.021953896816684963,
"grad_norm": 37.21672974282255,
"learning_rate": 1.0869565217391303e-07,
"loss": 1.4405,
"step": 10
},
{
"epoch": 0.03293084522502744,
"grad_norm": 28.728369641123845,
"learning_rate": 1.6304347826086955e-07,
"loss": 1.4495,
"step": 15
},
{
"epoch": 0.043907793633369926,
"grad_norm": 24.14103448999392,
"learning_rate": 2.1739130434782607e-07,
"loss": 1.348,
"step": 20
},
{
"epoch": 0.054884742041712405,
"grad_norm": 17.84361561407524,
"learning_rate": 2.717391304347826e-07,
"loss": 1.2803,
"step": 25
},
{
"epoch": 0.06586169045005488,
"grad_norm": 8.32310884222812,
"learning_rate": 3.260869565217391e-07,
"loss": 1.2081,
"step": 30
},
{
"epoch": 0.07683863885839737,
"grad_norm": 4.7849107325475435,
"learning_rate": 3.8043478260869567e-07,
"loss": 1.1701,
"step": 35
},
{
"epoch": 0.08781558726673985,
"grad_norm": 4.142884139685247,
"learning_rate": 4.3478260869565214e-07,
"loss": 1.1529,
"step": 40
},
{
"epoch": 0.09879253567508232,
"grad_norm": 4.1545233643304105,
"learning_rate": 4.891304347826087e-07,
"loss": 1.1303,
"step": 45
},
{
"epoch": 0.10976948408342481,
"grad_norm": 3.10192089815295,
"learning_rate": 4.998820089787287e-07,
"loss": 1.1088,
"step": 50
},
{
"epoch": 0.1207464324917673,
"grad_norm": 3.1563385745763677,
"learning_rate": 4.99402861330474e-07,
"loss": 1.1042,
"step": 55
},
{
"epoch": 0.13172338090010977,
"grad_norm": 2.5435814181022827,
"learning_rate": 4.985558887302488e-07,
"loss": 1.0956,
"step": 60
},
{
"epoch": 0.14270032930845225,
"grad_norm": 2.640027183329368,
"learning_rate": 4.973423403123028e-07,
"loss": 1.079,
"step": 65
},
{
"epoch": 0.15367727771679474,
"grad_norm": 3.037284920443539,
"learning_rate": 4.957640058451501e-07,
"loss": 1.0675,
"step": 70
},
{
"epoch": 0.16465422612513722,
"grad_norm": 3.1427488556058223,
"learning_rate": 4.938232130919785e-07,
"loss": 1.0734,
"step": 75
},
{
"epoch": 0.1756311745334797,
"grad_norm": 2.2891126109812543,
"learning_rate": 4.915228243776124e-07,
"loss": 1.0663,
"step": 80
},
{
"epoch": 0.18660812294182216,
"grad_norm": 3.1770515308884497,
"learning_rate": 4.888662323670913e-07,
"loss": 1.0556,
"step": 85
},
{
"epoch": 0.19758507135016465,
"grad_norm": 2.497553823382758,
"learning_rate": 4.858573550620908e-07,
"loss": 1.067,
"step": 90
},
{
"epoch": 0.20856201975850713,
"grad_norm": 2.2786389579554225,
"learning_rate": 4.825006300225665e-07,
"loss": 1.0713,
"step": 95
},
{
"epoch": 0.21953896816684962,
"grad_norm": 2.4565400673449314,
"learning_rate": 4.7880100782214e-07,
"loss": 1.0594,
"step": 100
},
{
"epoch": 0.2305159165751921,
"grad_norm": 2.2057241212988274,
"learning_rate": 4.747639447468816e-07,
"loss": 1.0622,
"step": 105
},
{
"epoch": 0.2414928649835346,
"grad_norm": 2.699908595698097,
"learning_rate": 4.7039539474825683e-07,
"loss": 1.0263,
"step": 110
},
{
"epoch": 0.2524698133918771,
"grad_norm": 2.595572602014253,
"learning_rate": 4.657018006621053e-07,
"loss": 1.0289,
"step": 115
},
{
"epoch": 0.26344676180021953,
"grad_norm": 2.367781794798575,
"learning_rate": 4.6069008470660057e-07,
"loss": 1.0671,
"step": 120
},
{
"epoch": 0.27442371020856204,
"grad_norm": 2.191786685372748,
"learning_rate": 4.5536763827320803e-07,
"loss": 1.0331,
"step": 125
},
{
"epoch": 0.2854006586169045,
"grad_norm": 2.2558364905064097,
"learning_rate": 4.4974231102569355e-07,
"loss": 1.0247,
"step": 130
},
{
"epoch": 0.29637760702524696,
"grad_norm": 2.2396909107202134,
"learning_rate": 4.438223993232634e-07,
"loss": 1.0097,
"step": 135
},
{
"epoch": 0.30735455543358947,
"grad_norm": 2.2953815342987007,
"learning_rate": 4.3761663398490634e-07,
"loss": 1.0162,
"step": 140
},
{
"epoch": 0.31833150384193193,
"grad_norm": 2.469492662807899,
"learning_rate": 4.3113416741298616e-07,
"loss": 1.0233,
"step": 145
},
{
"epoch": 0.32930845225027444,
"grad_norm": 2.3585547924597052,
"learning_rate": 4.2438456009507195e-07,
"loss": 1.0095,
"step": 150
},
{
"epoch": 0.3402854006586169,
"grad_norm": 2.3421663442140837,
"learning_rate": 4.1737776650391625e-07,
"loss": 1.0187,
"step": 155
},
{
"epoch": 0.3512623490669594,
"grad_norm": 2.5398466939682733,
"learning_rate": 4.101241204163748e-07,
"loss": 0.9977,
"step": 160
},
{
"epoch": 0.36223929747530187,
"grad_norm": 2.2242476659757275,
"learning_rate": 4.0263431967291934e-07,
"loss": 1.027,
"step": 165
},
{
"epoch": 0.3732162458836443,
"grad_norm": 2.1753753105769227,
"learning_rate": 3.949194104002224e-07,
"loss": 1.008,
"step": 170
},
{
"epoch": 0.38419319429198684,
"grad_norm": 2.300275831812692,
"learning_rate": 3.8699077072008085e-07,
"loss": 1.0106,
"step": 175
},
{
"epoch": 0.3951701427003293,
"grad_norm": 2.369231981384555,
"learning_rate": 3.7886009396870564e-07,
"loss": 1.0034,
"step": 180
},
{
"epoch": 0.4061470911086718,
"grad_norm": 3.342159958597614,
"learning_rate": 3.705393714511268e-07,
"loss": 1.0008,
"step": 185
},
{
"epoch": 0.41712403951701427,
"grad_norm": 2.2742818660752344,
"learning_rate": 3.6204087475614676e-07,
"loss": 1.0163,
"step": 190
},
{
"epoch": 0.4281009879253567,
"grad_norm": 2.224871070156181,
"learning_rate": 3.533771376579249e-07,
"loss": 1.0116,
"step": 195
},
{
"epoch": 0.43907793633369924,
"grad_norm": 2.201217019563197,
"learning_rate": 3.445609376308857e-07,
"loss": 0.9992,
"step": 200
},
{
"epoch": 0.4500548847420417,
"grad_norm": 2.244052004363795,
"learning_rate": 3.356052770052119e-07,
"loss": 1.0002,
"step": 205
},
{
"epoch": 0.4610318331503842,
"grad_norm": 2.1869532116505437,
"learning_rate": 3.2652336379071506e-07,
"loss": 1.003,
"step": 210
},
{
"epoch": 0.47200878155872666,
"grad_norm": 2.25200716108438,
"learning_rate": 3.173285921973657e-07,
"loss": 0.9967,
"step": 215
},
{
"epoch": 0.4829857299670692,
"grad_norm": 2.1977221613580915,
"learning_rate": 3.0803452288121113e-07,
"loss": 0.9933,
"step": 220
},
{
"epoch": 0.49396267837541163,
"grad_norm": 2.1857074442694064,
"learning_rate": 2.986548629448146e-07,
"loss": 1.0118,
"step": 225
},
{
"epoch": 0.5049396267837541,
"grad_norm": 2.176224528870358,
"learning_rate": 2.892034457217119e-07,
"loss": 0.9803,
"step": 230
},
{
"epoch": 0.5159165751920965,
"grad_norm": 2.1323657558564895,
"learning_rate": 2.7969421037470033e-07,
"loss": 1.011,
"step": 235
},
{
"epoch": 0.5268935236004391,
"grad_norm": 2.210400681860978,
"learning_rate": 2.70141181338047e-07,
"loss": 0.9835,
"step": 240
},
{
"epoch": 0.5378704720087816,
"grad_norm": 2.294325436997366,
"learning_rate": 2.60558447633938e-07,
"loss": 0.9994,
"step": 245
},
{
"epoch": 0.5488474204171241,
"grad_norm": 2.269474512577858,
"learning_rate": 2.509601420936717e-07,
"loss": 1.0056,
"step": 250
},
{
"epoch": 0.5598243688254665,
"grad_norm": 2.350884633568068,
"learning_rate": 2.413604205142404e-07,
"loss": 0.9975,
"step": 255
},
{
"epoch": 0.570801317233809,
"grad_norm": 2.2618918862021387,
"learning_rate": 2.3177344078104305e-07,
"loss": 1.0226,
"step": 260
},
{
"epoch": 0.5817782656421515,
"grad_norm": 2.335669763682869,
"learning_rate": 2.2221334198751717e-07,
"loss": 0.9996,
"step": 265
},
{
"epoch": 0.5927552140504939,
"grad_norm": 2.218697106476552,
"learning_rate": 2.1269422358248534e-07,
"loss": 0.9764,
"step": 270
},
{
"epoch": 0.6037321624588364,
"grad_norm": 2.294055527983993,
"learning_rate": 2.0323012457597113e-07,
"loss": 0.9911,
"step": 275
},
{
"epoch": 0.6147091108671789,
"grad_norm": 2.131671305300942,
"learning_rate": 1.9383500283415127e-07,
"loss": 1.0084,
"step": 280
},
{
"epoch": 0.6256860592755215,
"grad_norm": 2.170441232463747,
"learning_rate": 1.8452271449398015e-07,
"loss": 0.9935,
"step": 285
},
{
"epoch": 0.6366630076838639,
"grad_norm": 2.2800412432483994,
"learning_rate": 1.753069935278477e-07,
"loss": 0.9981,
"step": 290
},
{
"epoch": 0.6476399560922064,
"grad_norm": 2.2112002252829233,
"learning_rate": 1.662014314884074e-07,
"loss": 0.9821,
"step": 295
},
{
"epoch": 0.6586169045005489,
"grad_norm": 2.1523725296369696,
"learning_rate": 1.5721945746344914e-07,
"loss": 1.0028,
"step": 300
},
{
"epoch": 0.6695938529088913,
"grad_norm": 2.2685448996190476,
"learning_rate": 1.4837431827037786e-07,
"loss": 0.9843,
"step": 305
},
{
"epoch": 0.6805708013172338,
"grad_norm": 2.222219631378057,
"learning_rate": 1.3967905891950936e-07,
"loss": 0.9892,
"step": 310
},
{
"epoch": 0.6915477497255763,
"grad_norm": 2.17757738873566,
"learning_rate": 1.3114650337499578e-07,
"loss": 0.993,
"step": 315
},
{
"epoch": 0.7025246981339188,
"grad_norm": 2.202207960410075,
"learning_rate": 1.227892356417542e-07,
"loss": 0.98,
"step": 320
},
{
"epoch": 0.7135016465422612,
"grad_norm": 2.2364238502729887,
"learning_rate": 1.1461958120629345e-07,
"loss": 0.9747,
"step": 325
},
{
"epoch": 0.7244785949506037,
"grad_norm": 2.2007426350936314,
"learning_rate": 1.0664958885880901e-07,
"loss": 0.9993,
"step": 330
},
{
"epoch": 0.7354555433589463,
"grad_norm": 2.4965113706058433,
"learning_rate": 9.889101292335625e-08,
"loss": 0.9881,
"step": 335
},
{
"epoch": 0.7464324917672887,
"grad_norm": 2.1808970559228955,
"learning_rate": 9.13552959223089e-08,
"loss": 0.9961,
"step": 340
},
{
"epoch": 0.7574094401756312,
"grad_norm": 2.1878061482574336,
"learning_rate": 8.405355170066925e-08,
"loss": 0.9943,
"step": 345
},
{
"epoch": 0.7683863885839737,
"grad_norm": 2.1845807299852034,
"learning_rate": 7.699654903511971e-08,
"loss": 0.9866,
"step": 350
},
{
"epoch": 0.7793633369923162,
"grad_norm": 2.1414775634398375,
"learning_rate": 7.01946957519886e-08,
"loss": 0.9759,
"step": 355
},
{
"epoch": 0.7903402854006586,
"grad_norm": 2.122717025599496,
"learning_rate": 6.365802337755364e-08,
"loss": 1.0019,
"step": 360
},
{
"epoch": 0.8013172338090011,
"grad_norm": 2.14871365458936,
"learning_rate": 5.739617234332131e-08,
"loss": 0.9961,
"step": 365
},
{
"epoch": 0.8122941822173436,
"grad_norm": 2.1563921789431233,
"learning_rate": 5.14183777681014e-08,
"loss": 0.979,
"step": 370
},
{
"epoch": 0.823271130625686,
"grad_norm": 2.2409621573734566,
"learning_rate": 4.5733455837846325e-08,
"loss": 1.0026,
"step": 375
},
{
"epoch": 0.8342480790340285,
"grad_norm": 2.2773154227725296,
"learning_rate": 4.0349790803341274e-08,
"loss": 1.0058,
"step": 380
},
{
"epoch": 0.845225027442371,
"grad_norm": 2.117720539958819,
"learning_rate": 3.527532261492272e-08,
"loss": 0.9875,
"step": 385
},
{
"epoch": 0.8562019758507134,
"grad_norm": 2.152957888451534,
"learning_rate": 3.0517535212460946e-08,
"loss": 0.9784,
"step": 390
},
{
"epoch": 0.867178924259056,
"grad_norm": 2.1225713025355692,
"learning_rate": 2.608344548787722e-08,
"loss": 0.9713,
"step": 395
},
{
"epoch": 0.8781558726673985,
"grad_norm": 2.1615502933761643,
"learning_rate": 2.19795929364735e-08,
"loss": 0.9895,
"step": 400
},
{
"epoch": 0.889132821075741,
"grad_norm": 2.167176721017015,
"learning_rate": 1.8212030012337704e-08,
"loss": 0.9982,
"step": 405
},
{
"epoch": 0.9001097694840834,
"grad_norm": 2.1328147082553994,
"learning_rate": 1.4786313202048456e-08,
"loss": 1.0076,
"step": 410
},
{
"epoch": 0.9110867178924259,
"grad_norm": 2.249033557726424,
"learning_rate": 1.1707494829843207e-08,
"loss": 0.9855,
"step": 415
},
{
"epoch": 0.9220636663007684,
"grad_norm": 2.8747995493229372,
"learning_rate": 8.980115606337046e-09,
"loss": 0.9957,
"step": 420
},
{
"epoch": 0.9330406147091108,
"grad_norm": 2.124614068925926,
"learning_rate": 6.608197931780496e-09,
"loss": 0.9769,
"step": 425
},
{
"epoch": 0.9440175631174533,
"grad_norm": 2.176873597616659,
"learning_rate": 4.595239963733011e-09,
"loss": 0.9853,
"step": 430
},
{
"epoch": 0.9549945115257958,
"grad_norm": 2.1342884606915202,
"learning_rate": 2.9442104579016356e-09,
"loss": 0.9917,
"step": 435
},
{
"epoch": 0.9659714599341384,
"grad_norm": 2.1374930773738487,
"learning_rate": 1.6575443897531294e-09,
"loss": 0.9909,
"step": 440
},
{
"epoch": 0.9769484083424808,
"grad_norm": 2.1416862097657865,
"learning_rate": 7.371393633574252e-10,
"loss": 0.9845,
"step": 445
},
{
"epoch": 0.9879253567508233,
"grad_norm": 2.156534272117916,
"learning_rate": 1.843528127584981e-10,
"loss": 0.9959,
"step": 450
},
{
"epoch": 0.9989023051591658,
"grad_norm": 2.1792993284280864,
"learning_rate": 0.0,
"loss": 0.9717,
"step": 455
},
{
"epoch": 0.9989023051591658,
"eval_loss": 1.0698977708816528,
"eval_runtime": 3.5449,
"eval_samples_per_second": 75.601,
"eval_steps_per_second": 1.41,
"step": 455
},
{
"epoch": 0.9989023051591658,
"step": 455,
"total_flos": 190430796840960.0,
"train_loss": 1.0385934979051024,
"train_runtime": 5380.501,
"train_samples_per_second": 21.667,
"train_steps_per_second": 0.085
}
],
"logging_steps": 5,
"max_steps": 455,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 190430796840960.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}