terry69's picture
Model save
6f2fce9 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.9983579638752053,
"eval_steps": 500,
"global_step": 456,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0021893814997263274,
"grad_norm": 1.099035580545203,
"learning_rate": 4.347826086956522e-06,
"loss": 1.3884,
"step": 1
},
{
"epoch": 0.010946907498631636,
"grad_norm": 1.0633282817267398,
"learning_rate": 2.173913043478261e-05,
"loss": 1.3896,
"step": 5
},
{
"epoch": 0.021893814997263273,
"grad_norm": 0.2753570266695847,
"learning_rate": 4.347826086956522e-05,
"loss": 1.3001,
"step": 10
},
{
"epoch": 0.03284072249589491,
"grad_norm": 0.23486140029795313,
"learning_rate": 6.521739130434783e-05,
"loss": 1.2641,
"step": 15
},
{
"epoch": 0.043787629994526546,
"grad_norm": 0.17661059915753233,
"learning_rate": 8.695652173913044e-05,
"loss": 1.244,
"step": 20
},
{
"epoch": 0.05473453749315818,
"grad_norm": 0.13588516740211107,
"learning_rate": 0.00010869565217391305,
"loss": 1.2062,
"step": 25
},
{
"epoch": 0.06568144499178982,
"grad_norm": 0.09952038895662578,
"learning_rate": 0.00013043478260869567,
"loss": 1.2036,
"step": 30
},
{
"epoch": 0.07662835249042145,
"grad_norm": 0.08662921860732123,
"learning_rate": 0.00015217391304347827,
"loss": 1.1697,
"step": 35
},
{
"epoch": 0.08757525998905309,
"grad_norm": 0.08556102174203851,
"learning_rate": 0.00017391304347826088,
"loss": 1.168,
"step": 40
},
{
"epoch": 0.09852216748768473,
"grad_norm": 0.08239874598331373,
"learning_rate": 0.0001956521739130435,
"loss": 1.1752,
"step": 45
},
{
"epoch": 0.10946907498631636,
"grad_norm": 0.08011130479266197,
"learning_rate": 0.0001999530335191093,
"loss": 1.1317,
"step": 50
},
{
"epoch": 0.120415982484948,
"grad_norm": 0.09091690871262173,
"learning_rate": 0.00019976230779866525,
"loss": 1.1235,
"step": 55
},
{
"epoch": 0.13136288998357964,
"grad_norm": 0.10018228425429193,
"learning_rate": 0.000199425167134466,
"loss": 1.14,
"step": 60
},
{
"epoch": 0.1423097974822113,
"grad_norm": 0.08568976262387419,
"learning_rate": 0.00019894210632692745,
"loss": 1.1184,
"step": 65
},
{
"epoch": 0.1532567049808429,
"grad_norm": 0.09058910078299229,
"learning_rate": 0.00019831383433439797,
"loss": 1.1182,
"step": 70
},
{
"epoch": 0.16420361247947454,
"grad_norm": 0.07075651635331383,
"learning_rate": 0.00019754127323266428,
"loss": 1.1107,
"step": 75
},
{
"epoch": 0.17515051997810618,
"grad_norm": 0.08205972533630575,
"learning_rate": 0.00019662555686167808,
"loss": 1.1146,
"step": 80
},
{
"epoch": 0.18609742747673783,
"grad_norm": 0.09822244044433728,
"learning_rate": 0.00019556802916148962,
"loss": 1.1206,
"step": 85
},
{
"epoch": 0.19704433497536947,
"grad_norm": 0.08491467009411377,
"learning_rate": 0.00019437024219983028,
"loss": 1.1091,
"step": 90
},
{
"epoch": 0.20799124247400108,
"grad_norm": 0.08166420957630437,
"learning_rate": 0.00019303395389423918,
"loss": 1.1115,
"step": 95
},
{
"epoch": 0.21893814997263272,
"grad_norm": 0.08260201243872915,
"learning_rate": 0.00019156112543207673,
"loss": 1.1175,
"step": 100
},
{
"epoch": 0.22988505747126436,
"grad_norm": 0.08255167959693703,
"learning_rate": 0.0001899539183922119,
"loss": 1.0968,
"step": 105
},
{
"epoch": 0.240831964969896,
"grad_norm": 0.07836100445915845,
"learning_rate": 0.00018821469157260685,
"loss": 1.0888,
"step": 110
},
{
"epoch": 0.25177887246852765,
"grad_norm": 0.07207251942982666,
"learning_rate": 0.00018634599752845592,
"loss": 1.1074,
"step": 115
},
{
"epoch": 0.2627257799671593,
"grad_norm": 0.08720514170327426,
"learning_rate": 0.00018435057882595882,
"loss": 1.1052,
"step": 120
},
{
"epoch": 0.27367268746579093,
"grad_norm": 0.0853293654672411,
"learning_rate": 0.0001822313640172265,
"loss": 1.0952,
"step": 125
},
{
"epoch": 0.2846195949644226,
"grad_norm": 0.09875940431911555,
"learning_rate": 0.00017999146334222695,
"loss": 1.1227,
"step": 130
},
{
"epoch": 0.2955665024630542,
"grad_norm": 0.07021884431606892,
"learning_rate": 0.00017763416416407952,
"loss": 1.0942,
"step": 135
},
{
"epoch": 0.3065134099616858,
"grad_norm": 0.06842699832201858,
"learning_rate": 0.00017516292614439585,
"loss": 1.0953,
"step": 140
},
{
"epoch": 0.31746031746031744,
"grad_norm": 0.07208694079619522,
"learning_rate": 0.0001725813761657495,
"loss": 1.0967,
"step": 145
},
{
"epoch": 0.3284072249589491,
"grad_norm": 0.06921789319582454,
"learning_rate": 0.00016989330300872576,
"loss": 1.0801,
"step": 150
},
{
"epoch": 0.3393541324575807,
"grad_norm": 0.07271469189791151,
"learning_rate": 0.0001671026517913634,
"loss": 1.0833,
"step": 155
},
{
"epoch": 0.35030103995621237,
"grad_norm": 0.06997352100806448,
"learning_rate": 0.00016421351817915024,
"loss": 1.0965,
"step": 160
},
{
"epoch": 0.361247947454844,
"grad_norm": 0.07856963780328721,
"learning_rate": 0.00016123014237406913,
"loss": 1.1091,
"step": 165
},
{
"epoch": 0.37219485495347565,
"grad_norm": 0.07209466893505154,
"learning_rate": 0.00015815690289151658,
"loss": 1.0899,
"step": 170
},
{
"epoch": 0.3831417624521073,
"grad_norm": 0.07481315197731042,
"learning_rate": 0.00015499831013422803,
"loss": 1.1041,
"step": 175
},
{
"epoch": 0.39408866995073893,
"grad_norm": 0.08084183121949147,
"learning_rate": 0.00015175899977263964,
"loss": 1.0955,
"step": 180
},
{
"epoch": 0.4050355774493706,
"grad_norm": 0.0818129087053093,
"learning_rate": 0.0001484437259414027,
"loss": 1.1145,
"step": 185
},
{
"epoch": 0.41598248494800216,
"grad_norm": 0.06950621743678526,
"learning_rate": 0.00014505735426203543,
"loss": 1.1149,
"step": 190
},
{
"epoch": 0.4269293924466338,
"grad_norm": 0.07548210193666667,
"learning_rate": 0.00014160485470195245,
"loss": 1.11,
"step": 195
},
{
"epoch": 0.43787629994526545,
"grad_norm": 0.07058243846026176,
"learning_rate": 0.00013809129428035227,
"loss": 1.0846,
"step": 200
},
{
"epoch": 0.4488232074438971,
"grad_norm": 0.07020891379155109,
"learning_rate": 0.00013452182963166794,
"loss": 1.0848,
"step": 205
},
{
"epoch": 0.45977011494252873,
"grad_norm": 0.06835760784196979,
"learning_rate": 0.00013090169943749476,
"loss": 1.0851,
"step": 210
},
{
"epoch": 0.47071702244116037,
"grad_norm": 0.07605646093102703,
"learning_rate": 0.00012723621673810275,
"loss": 1.1047,
"step": 215
},
{
"epoch": 0.481663929939792,
"grad_norm": 0.0705995475810648,
"learning_rate": 0.0001235307611348174,
"loss": 1.0808,
"step": 220
},
{
"epoch": 0.49261083743842365,
"grad_norm": 0.07066082603796292,
"learning_rate": 0.00011979077089471287,
"loss": 1.0823,
"step": 225
},
{
"epoch": 0.5035577449370553,
"grad_norm": 0.06843040733212367,
"learning_rate": 0.0001160217349692051,
"loss": 1.093,
"step": 230
},
{
"epoch": 0.5145046524356869,
"grad_norm": 0.07581398076733512,
"learning_rate": 0.00011222918493825876,
"loss": 1.0799,
"step": 235
},
{
"epoch": 0.5254515599343186,
"grad_norm": 0.06701479053423436,
"learning_rate": 0.00010841868689203071,
"loss": 1.0984,
"step": 240
},
{
"epoch": 0.5363984674329502,
"grad_norm": 0.06919851726403416,
"learning_rate": 0.00010459583326186533,
"loss": 1.0919,
"step": 245
},
{
"epoch": 0.5473453749315819,
"grad_norm": 0.07632131476135244,
"learning_rate": 0.00010076623461263018,
"loss": 1.0801,
"step": 250
},
{
"epoch": 0.5582922824302134,
"grad_norm": 0.0661733510360967,
"learning_rate": 9.693551140843847e-05,
"loss": 1.0824,
"step": 255
},
{
"epoch": 0.5692391899288451,
"grad_norm": 0.06851059826679573,
"learning_rate": 9.310928576384293e-05,
"loss": 1.0956,
"step": 260
},
{
"epoch": 0.5801860974274767,
"grad_norm": 0.07225957964642454,
"learning_rate": 8.929317319260726e-05,
"loss": 1.0818,
"step": 265
},
{
"epoch": 0.5911330049261084,
"grad_norm": 0.07526177174581719,
"learning_rate": 8.549277436616551e-05,
"loss": 1.0876,
"step": 270
},
{
"epoch": 0.60207991242474,
"grad_norm": 0.0688828763506549,
"learning_rate": 8.171366689386432e-05,
"loss": 1.0834,
"step": 275
},
{
"epoch": 0.6130268199233716,
"grad_norm": 0.07233185605708406,
"learning_rate": 7.796139713705213e-05,
"loss": 1.0904,
"step": 280
},
{
"epoch": 0.6239737274220033,
"grad_norm": 0.07238689140745656,
"learning_rate": 7.424147206902891e-05,
"loss": 1.078,
"step": 285
},
{
"epoch": 0.6349206349206349,
"grad_norm": 0.07045252374699128,
"learning_rate": 7.055935119280369e-05,
"loss": 1.0746,
"step": 290
},
{
"epoch": 0.6458675424192666,
"grad_norm": 0.06885925177718841,
"learning_rate": 6.692043852852079e-05,
"loss": 1.0914,
"step": 295
},
{
"epoch": 0.6568144499178982,
"grad_norm": 0.0714964117404916,
"learning_rate": 6.333007468231521e-05,
"loss": 1.0829,
"step": 300
},
{
"epoch": 0.6677613574165299,
"grad_norm": 0.0670654539976087,
"learning_rate": 5.9793529008236625e-05,
"loss": 1.0756,
"step": 305
},
{
"epoch": 0.6787082649151615,
"grad_norm": 0.06833244959359351,
"learning_rate": 5.63159918747457e-05,
"loss": 1.0779,
"step": 310
},
{
"epoch": 0.6896551724137931,
"grad_norm": 0.07144684011257411,
"learning_rate": 5.29025670471325e-05,
"loss": 1.0983,
"step": 315
},
{
"epoch": 0.7006020799124247,
"grad_norm": 0.07142805023654634,
"learning_rate": 4.955826419703735e-05,
"loss": 1.0867,
"step": 320
},
{
"epoch": 0.7115489874110563,
"grad_norm": 0.07674467638840121,
"learning_rate": 4.628799155006669e-05,
"loss": 1.0651,
"step": 325
},
{
"epoch": 0.722495894909688,
"grad_norm": 0.06634612204586297,
"learning_rate": 4.30965486822953e-05,
"loss": 1.0938,
"step": 330
},
{
"epoch": 0.7334428024083196,
"grad_norm": 0.06579811491660853,
"learning_rate": 3.9988619476226355e-05,
"loss": 1.0781,
"step": 335
},
{
"epoch": 0.7443897099069513,
"grad_norm": 0.06839286253175804,
"learning_rate": 3.69687652465482e-05,
"loss": 1.0807,
"step": 340
},
{
"epoch": 0.7553366174055829,
"grad_norm": 0.06644270765658328,
"learning_rate": 3.4041418045775895e-05,
"loss": 1.074,
"step": 345
},
{
"epoch": 0.7662835249042146,
"grad_norm": 0.06772025228639957,
"learning_rate": 3.121087415960304e-05,
"loss": 1.0762,
"step": 350
},
{
"epoch": 0.7772304324028462,
"grad_norm": 0.06615084228702964,
"learning_rate": 2.848128780150995e-05,
"loss": 1.0779,
"step": 355
},
{
"epoch": 0.7881773399014779,
"grad_norm": 0.0706561964889078,
"learning_rate": 2.5856665015882685e-05,
"loss": 1.0904,
"step": 360
},
{
"epoch": 0.7991242474001095,
"grad_norm": 0.06748598590158515,
"learning_rate": 2.334085779859041e-05,
"loss": 1.0654,
"step": 365
},
{
"epoch": 0.8100711548987412,
"grad_norm": 0.06669809967764875,
"learning_rate": 2.0937558443650206e-05,
"loss": 1.0835,
"step": 370
},
{
"epoch": 0.8210180623973727,
"grad_norm": 0.06776479655939238,
"learning_rate": 1.8650294124276556e-05,
"loss": 1.0836,
"step": 375
},
{
"epoch": 0.8319649698960043,
"grad_norm": 0.06672866236853266,
"learning_rate": 1.6482421716268215e-05,
"loss": 1.0909,
"step": 380
},
{
"epoch": 0.842911877394636,
"grad_norm": 0.0670546035798045,
"learning_rate": 1.4437122871329955e-05,
"loss": 1.0984,
"step": 385
},
{
"epoch": 0.8538587848932676,
"grad_norm": 0.06639094648741257,
"learning_rate": 1.2517399347560167e-05,
"loss": 1.1008,
"step": 390
},
{
"epoch": 0.8648056923918993,
"grad_norm": 0.06852807789172832,
"learning_rate": 1.0726068603956741e-05,
"loss": 1.0783,
"step": 395
},
{
"epoch": 0.8757525998905309,
"grad_norm": 0.06914284979355585,
"learning_rate": 9.065759665407515e-06,
"loss": 1.0828,
"step": 400
},
{
"epoch": 0.8866995073891626,
"grad_norm": 0.06687442460085294,
"learning_rate": 7.538909264233751e-06,
"loss": 1.0845,
"step": 405
},
{
"epoch": 0.8976464148877942,
"grad_norm": 0.07046699389113141,
"learning_rate": 6.147758263949321e-06,
"loss": 1.0784,
"step": 410
},
{
"epoch": 0.9085933223864259,
"grad_norm": 0.06774076270250588,
"learning_rate": 4.8943483704846475e-06,
"loss": 1.0914,
"step": 415
},
{
"epoch": 0.9195402298850575,
"grad_norm": 0.06693453050067996,
"learning_rate": 3.7805191357015657e-06,
"loss": 1.0986,
"step": 420
},
{
"epoch": 0.9304871373836892,
"grad_norm": 0.06389085654103847,
"learning_rate": 2.8079052575973765e-06,
"loss": 1.087,
"step": 425
},
{
"epoch": 0.9414340448823207,
"grad_norm": 0.06811361952720432,
"learning_rate": 1.9779341811600794e-06,
"loss": 1.07,
"step": 430
},
{
"epoch": 0.9523809523809523,
"grad_norm": 0.06592264001597521,
"learning_rate": 1.2918240033960027e-06,
"loss": 1.0825,
"step": 435
},
{
"epoch": 0.963327859879584,
"grad_norm": 0.06657689313692897,
"learning_rate": 7.505816856045012e-07,
"loss": 1.0843,
"step": 440
},
{
"epoch": 0.9742747673782156,
"grad_norm": 0.06488852108049653,
"learning_rate": 3.5500157552337264e-07,
"loss": 1.0923,
"step": 445
},
{
"epoch": 0.9852216748768473,
"grad_norm": 0.06657099771817772,
"learning_rate": 1.0566424151401411e-07,
"loss": 1.0882,
"step": 450
},
{
"epoch": 0.9961685823754789,
"grad_norm": 0.06558284673944498,
"learning_rate": 2.935620497301894e-09,
"loss": 1.0893,
"step": 455
},
{
"epoch": 0.9983579638752053,
"eval_loss": 0.9926520586013794,
"eval_runtime": 2.1266,
"eval_samples_per_second": 3.292,
"eval_steps_per_second": 0.94,
"step": 456
},
{
"epoch": 0.9983579638752053,
"step": 456,
"total_flos": 1.447234595979264e+16,
"train_loss": 1.1072515829613334,
"train_runtime": 17333.1907,
"train_samples_per_second": 3.373,
"train_steps_per_second": 0.026
}
],
"logging_steps": 5,
"max_steps": 456,
"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": 1.447234595979264e+16,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}