sanity_syntax_33p_60k / trainer_state.json
terry69's picture
Model save
e1e4f78 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 533,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.001876172607879925,
"grad_norm": 0.4121309108650883,
"learning_rate": 3.7037037037037037e-06,
"loss": 1.6355,
"step": 1
},
{
"epoch": 0.009380863039399626,
"grad_norm": 0.4145535276322867,
"learning_rate": 1.8518518518518518e-05,
"loss": 1.6509,
"step": 5
},
{
"epoch": 0.01876172607879925,
"grad_norm": 0.499090421739371,
"learning_rate": 3.7037037037037037e-05,
"loss": 1.6194,
"step": 10
},
{
"epoch": 0.028142589118198873,
"grad_norm": 0.2168002082363555,
"learning_rate": 5.555555555555556e-05,
"loss": 1.6102,
"step": 15
},
{
"epoch": 0.0375234521575985,
"grad_norm": 0.19182026954015316,
"learning_rate": 7.407407407407407e-05,
"loss": 1.5826,
"step": 20
},
{
"epoch": 0.04690431519699812,
"grad_norm": 0.17104214986412603,
"learning_rate": 9.25925925925926e-05,
"loss": 1.4983,
"step": 25
},
{
"epoch": 0.05628517823639775,
"grad_norm": 0.1314352337398472,
"learning_rate": 0.00011111111111111112,
"loss": 1.4705,
"step": 30
},
{
"epoch": 0.06566604127579738,
"grad_norm": 0.13683731835169527,
"learning_rate": 0.00012962962962962963,
"loss": 1.4351,
"step": 35
},
{
"epoch": 0.075046904315197,
"grad_norm": 0.12592565344109571,
"learning_rate": 0.00014814814814814815,
"loss": 1.4154,
"step": 40
},
{
"epoch": 0.08442776735459662,
"grad_norm": 0.10747815841874438,
"learning_rate": 0.0001666666666666667,
"loss": 1.3925,
"step": 45
},
{
"epoch": 0.09380863039399624,
"grad_norm": 0.09736495320865902,
"learning_rate": 0.0001851851851851852,
"loss": 1.3452,
"step": 50
},
{
"epoch": 0.10318949343339587,
"grad_norm": 0.09026377977279064,
"learning_rate": 0.00019999784921417228,
"loss": 1.3658,
"step": 55
},
{
"epoch": 0.1125703564727955,
"grad_norm": 0.08191150814769302,
"learning_rate": 0.00019992258142410334,
"loss": 1.3526,
"step": 60
},
{
"epoch": 0.12195121951219512,
"grad_norm": 0.0987761994127843,
"learning_rate": 0.00019973986684235418,
"loss": 1.3457,
"step": 65
},
{
"epoch": 0.13133208255159476,
"grad_norm": 0.08470882333862399,
"learning_rate": 0.00019944990194198758,
"loss": 1.3237,
"step": 70
},
{
"epoch": 0.14071294559099437,
"grad_norm": 0.08215471773448521,
"learning_rate": 0.00019905299852237654,
"loss": 1.3171,
"step": 75
},
{
"epoch": 0.150093808630394,
"grad_norm": 0.07920641876308229,
"learning_rate": 0.00019854958337392654,
"loss": 1.2913,
"step": 80
},
{
"epoch": 0.15947467166979362,
"grad_norm": 0.09467087341552041,
"learning_rate": 0.00019794019781914766,
"loss": 1.3156,
"step": 85
},
{
"epoch": 0.16885553470919323,
"grad_norm": 0.08313358090040693,
"learning_rate": 0.0001972254971305701,
"loss": 1.3032,
"step": 90
},
{
"epoch": 0.17823639774859287,
"grad_norm": 0.08181396678268484,
"learning_rate": 0.00019640624982612942,
"loss": 1.3047,
"step": 95
},
{
"epoch": 0.18761726078799248,
"grad_norm": 0.08648614872822097,
"learning_rate": 0.00019548333684277919,
"loss": 1.2938,
"step": 100
},
{
"epoch": 0.19699812382739212,
"grad_norm": 0.08114011033820992,
"learning_rate": 0.00019445775058921853,
"loss": 1.2919,
"step": 105
},
{
"epoch": 0.20637898686679174,
"grad_norm": 0.08069454171446046,
"learning_rate": 0.00019333059387875525,
"loss": 1.3007,
"step": 110
},
{
"epoch": 0.21575984990619138,
"grad_norm": 0.1055930521893841,
"learning_rate": 0.0001921030787434499,
"loss": 1.3022,
"step": 115
},
{
"epoch": 0.225140712945591,
"grad_norm": 0.1039770749041834,
"learning_rate": 0.0001907765251308173,
"loss": 1.2945,
"step": 120
},
{
"epoch": 0.23452157598499063,
"grad_norm": 0.08612620245766023,
"learning_rate": 0.0001893523594844865,
"loss": 1.2734,
"step": 125
},
{
"epoch": 0.24390243902439024,
"grad_norm": 0.09453075060390269,
"learning_rate": 0.00018783211321034534,
"loss": 1.2903,
"step": 130
},
{
"epoch": 0.25328330206378985,
"grad_norm": 0.09115708333629854,
"learning_rate": 0.00018621742102981905,
"loss": 1.3021,
"step": 135
},
{
"epoch": 0.2626641651031895,
"grad_norm": 0.09246301111274602,
"learning_rate": 0.0001845100192220537,
"loss": 1.2994,
"step": 140
},
{
"epoch": 0.27204502814258913,
"grad_norm": 0.08863879008659571,
"learning_rate": 0.00018271174375689454,
"loss": 1.3049,
"step": 145
},
{
"epoch": 0.28142589118198874,
"grad_norm": 0.09121089281314407,
"learning_rate": 0.00018082452832066687,
"loss": 1.2836,
"step": 150
},
{
"epoch": 0.29080675422138835,
"grad_norm": 0.08883133842744517,
"learning_rate": 0.00017885040223688235,
"loss": 1.2643,
"step": 155
},
{
"epoch": 0.300187617260788,
"grad_norm": 0.0889208798307562,
"learning_rate": 0.0001767914882841067,
"loss": 1.2902,
"step": 160
},
{
"epoch": 0.30956848030018763,
"grad_norm": 0.10746121860126627,
"learning_rate": 0.00017465000041333494,
"loss": 1.2968,
"step": 165
},
{
"epoch": 0.31894934333958724,
"grad_norm": 0.0832629700388369,
"learning_rate": 0.0001724282413673291,
"loss": 1.3031,
"step": 170
},
{
"epoch": 0.32833020637898686,
"grad_norm": 0.09626519128659516,
"learning_rate": 0.00017012860020447796,
"loss": 1.3223,
"step": 175
},
{
"epoch": 0.33771106941838647,
"grad_norm": 0.08434970867620112,
"learning_rate": 0.0001677535497298416,
"loss": 1.2684,
"step": 180
},
{
"epoch": 0.34709193245778613,
"grad_norm": 0.08136406881743141,
"learning_rate": 0.0001653056438361432,
"loss": 1.265,
"step": 185
},
{
"epoch": 0.35647279549718575,
"grad_norm": 0.0897096956349179,
"learning_rate": 0.0001627875147575671,
"loss": 1.2936,
"step": 190
},
{
"epoch": 0.36585365853658536,
"grad_norm": 0.08876653720245482,
"learning_rate": 0.00016020187023931638,
"loss": 1.2961,
"step": 195
},
{
"epoch": 0.37523452157598497,
"grad_norm": 0.08532928831285669,
"learning_rate": 0.00015755149062597333,
"loss": 1.2634,
"step": 200
},
{
"epoch": 0.38461538461538464,
"grad_norm": 0.09494232614320906,
"learning_rate": 0.00015483922587179386,
"loss": 1.2794,
"step": 205
},
{
"epoch": 0.39399624765478425,
"grad_norm": 0.0938403596147772,
"learning_rate": 0.00015206799247615037,
"loss": 1.2668,
"step": 210
},
{
"epoch": 0.40337711069418386,
"grad_norm": 0.09298814066383618,
"learning_rate": 0.00014924077034741924,
"loss": 1.2704,
"step": 215
},
{
"epoch": 0.41275797373358347,
"grad_norm": 0.09381021867831099,
"learning_rate": 0.0001463605995986836,
"loss": 1.2743,
"step": 220
},
{
"epoch": 0.42213883677298314,
"grad_norm": 0.10478061496717873,
"learning_rate": 0.000143430577278699,
"loss": 1.3064,
"step": 225
},
{
"epoch": 0.43151969981238275,
"grad_norm": 0.0967412755428019,
"learning_rate": 0.0001404538540416353,
"loss": 1.2749,
"step": 230
},
{
"epoch": 0.44090056285178236,
"grad_norm": 0.10245748178767897,
"learning_rate": 0.00013743363075917724,
"loss": 1.2879,
"step": 235
},
{
"epoch": 0.450281425891182,
"grad_norm": 0.09494585437974697,
"learning_rate": 0.00013437315507862566,
"loss": 1.268,
"step": 240
},
{
"epoch": 0.4596622889305816,
"grad_norm": 0.10070892164674464,
"learning_rate": 0.0001312757179307012,
"loss": 1.2741,
"step": 245
},
{
"epoch": 0.46904315196998125,
"grad_norm": 0.08590715074864536,
"learning_rate": 0.000128144649990805,
"loss": 1.28,
"step": 250
},
{
"epoch": 0.47842401500938087,
"grad_norm": 0.0921370657647381,
"learning_rate": 0.00012498331809754243,
"loss": 1.3099,
"step": 255
},
{
"epoch": 0.4878048780487805,
"grad_norm": 0.09411198985391388,
"learning_rate": 0.00012179512163235974,
"loss": 1.2402,
"step": 260
},
{
"epoch": 0.4971857410881801,
"grad_norm": 0.0955041592156138,
"learning_rate": 0.0001185834888641883,
"loss": 1.2637,
"step": 265
},
{
"epoch": 0.5065666041275797,
"grad_norm": 0.08843739914811306,
"learning_rate": 0.0001153518732630253,
"loss": 1.2524,
"step": 270
},
{
"epoch": 0.5159474671669794,
"grad_norm": 0.08894269736301577,
"learning_rate": 0.00011210374978641631,
"loss": 1.3058,
"step": 275
},
{
"epoch": 0.525328330206379,
"grad_norm": 0.09583708909351138,
"learning_rate": 0.0001088426111428319,
"loss": 1.2462,
"step": 280
},
{
"epoch": 0.5347091932457786,
"grad_norm": 0.0934542252776399,
"learning_rate": 0.00010557196403595688,
"loss": 1.2641,
"step": 285
},
{
"epoch": 0.5440900562851783,
"grad_norm": 0.08688918325528941,
"learning_rate": 0.00010229532539393049,
"loss": 1.2593,
"step": 290
},
{
"epoch": 0.5534709193245778,
"grad_norm": 0.08796474911729359,
"learning_rate": 9.901621858759203e-05,
"loss": 1.2639,
"step": 295
},
{
"epoch": 0.5628517823639775,
"grad_norm": 0.09427974834689387,
"learning_rate": 9.57381696417989e-05,
"loss": 1.257,
"step": 300
},
{
"epoch": 0.5722326454033771,
"grad_norm": 0.08656751800419889,
"learning_rate": 9.246470344389065e-05,
"loss": 1.2545,
"step": 305
},
{
"epoch": 0.5816135084427767,
"grad_norm": 0.09133977950050422,
"learning_rate": 8.919933995337624e-05,
"loss": 1.2629,
"step": 310
},
{
"epoch": 0.5909943714821764,
"grad_norm": 0.0881803403964572,
"learning_rate": 8.594559041692003e-05,
"loss": 1.2668,
"step": 315
},
{
"epoch": 0.600375234521576,
"grad_norm": 0.08194022763623,
"learning_rate": 8.270695359269698e-05,
"loss": 1.2426,
"step": 320
},
{
"epoch": 0.6097560975609756,
"grad_norm": 0.08404336869271818,
"learning_rate": 7.948691198817666e-05,
"loss": 1.2608,
"step": 325
},
{
"epoch": 0.6191369606003753,
"grad_norm": 0.0911226278784459,
"learning_rate": 7.628892811538137e-05,
"loss": 1.2446,
"step": 330
},
{
"epoch": 0.6285178236397748,
"grad_norm": 0.0866722940291979,
"learning_rate": 7.311644076764564e-05,
"loss": 1.2661,
"step": 335
},
{
"epoch": 0.6378986866791745,
"grad_norm": 0.08492545007005736,
"learning_rate": 6.997286132188057e-05,
"loss": 1.2476,
"step": 340
},
{
"epoch": 0.6472795497185742,
"grad_norm": 0.08729514750397552,
"learning_rate": 6.68615700703186e-05,
"loss": 1.2471,
"step": 345
},
{
"epoch": 0.6566604127579737,
"grad_norm": 0.08416435736232936,
"learning_rate": 6.37859125856842e-05,
"loss": 1.2773,
"step": 350
},
{
"epoch": 0.6660412757973734,
"grad_norm": 0.09054925125258262,
"learning_rate": 6.074919612369787e-05,
"loss": 1.2854,
"step": 355
},
{
"epoch": 0.6754221388367729,
"grad_norm": 0.0840484253491192,
"learning_rate": 5.7754686066783045e-05,
"loss": 1.2321,
"step": 360
},
{
"epoch": 0.6848030018761726,
"grad_norm": 0.09091692153105359,
"learning_rate": 5.4805602412798906e-05,
"loss": 1.2519,
"step": 365
},
{
"epoch": 0.6941838649155723,
"grad_norm": 0.0996295825165971,
"learning_rate": 5.1905116312575475e-05,
"loss": 1.2422,
"step": 370
},
{
"epoch": 0.7035647279549718,
"grad_norm": 0.09167775721865827,
"learning_rate": 4.905634665997371e-05,
"loss": 1.274,
"step": 375
},
{
"epoch": 0.7129455909943715,
"grad_norm": 0.09342003802996314,
"learning_rate": 4.6262356738137937e-05,
"loss": 1.2727,
"step": 380
},
{
"epoch": 0.7223264540337712,
"grad_norm": 0.08770871963126985,
"learning_rate": 4.3526150925546e-05,
"loss": 1.2558,
"step": 385
},
{
"epoch": 0.7317073170731707,
"grad_norm": 0.09069696254933435,
"learning_rate": 4.0850671465400144e-05,
"loss": 1.2593,
"step": 390
},
{
"epoch": 0.7410881801125704,
"grad_norm": 0.08665209061432325,
"learning_rate": 3.823879530183154e-05,
"loss": 1.273,
"step": 395
},
{
"epoch": 0.7504690431519699,
"grad_norm": 0.08862136722443326,
"learning_rate": 3.569333098632109e-05,
"loss": 1.2613,
"step": 400
},
{
"epoch": 0.7598499061913696,
"grad_norm": 0.08931630016078942,
"learning_rate": 3.3217015657663145e-05,
"loss": 1.255,
"step": 405
},
{
"epoch": 0.7692307692307693,
"grad_norm": 0.08603421106367509,
"learning_rate": 3.081251209871872e-05,
"loss": 1.2956,
"step": 410
},
{
"epoch": 0.7786116322701688,
"grad_norm": 0.08449087840242356,
"learning_rate": 2.848240587312433e-05,
"loss": 1.2554,
"step": 415
},
{
"epoch": 0.7879924953095685,
"grad_norm": 0.09050119999554507,
"learning_rate": 2.622920254503416e-05,
"loss": 1.2555,
"step": 420
},
{
"epoch": 0.797373358348968,
"grad_norm": 0.08603303013311027,
"learning_rate": 2.405532498488612e-05,
"loss": 1.2445,
"step": 425
},
{
"epoch": 0.8067542213883677,
"grad_norm": 0.0917770059937159,
"learning_rate": 2.196311076408808e-05,
"loss": 1.2499,
"step": 430
},
{
"epoch": 0.8161350844277674,
"grad_norm": 0.09260597797294362,
"learning_rate": 1.995480964142663e-05,
"loss": 1.2643,
"step": 435
},
{
"epoch": 0.8255159474671669,
"grad_norm": 0.08638741822571576,
"learning_rate": 1.803258114390034e-05,
"loss": 1.2381,
"step": 440
},
{
"epoch": 0.8348968105065666,
"grad_norm": 0.08438081141056018,
"learning_rate": 1.6198492244579722e-05,
"loss": 1.2422,
"step": 445
},
{
"epoch": 0.8442776735459663,
"grad_norm": 0.08400750233223092,
"learning_rate": 1.445451513999012e-05,
"loss": 1.2178,
"step": 450
},
{
"epoch": 0.8536585365853658,
"grad_norm": 0.08554689021587018,
"learning_rate": 1.2802525129408038e-05,
"loss": 1.2406,
"step": 455
},
{
"epoch": 0.8630393996247655,
"grad_norm": 0.08737158991700927,
"learning_rate": 1.1244298598351077e-05,
"loss": 1.2434,
"step": 460
},
{
"epoch": 0.8724202626641651,
"grad_norm": 0.08730413757640722,
"learning_rate": 9.781511108429909e-06,
"loss": 1.2653,
"step": 465
},
{
"epoch": 0.8818011257035647,
"grad_norm": 0.0881674821245138,
"learning_rate": 8.415735595616203e-06,
"loss": 1.2524,
"step": 470
},
{
"epoch": 0.8911819887429644,
"grad_norm": 0.09353121525052323,
"learning_rate": 7.1484406788639215e-06,
"loss": 1.2523,
"step": 475
},
{
"epoch": 0.900562851782364,
"grad_norm": 0.08427560236702278,
"learning_rate": 5.980989080902721e-06,
"loss": 1.2812,
"step": 480
},
{
"epoch": 0.9099437148217636,
"grad_norm": 0.08684819671006813,
"learning_rate": 4.914636162901798e-06,
"loss": 1.2549,
"step": 485
},
{
"epoch": 0.9193245778611632,
"grad_norm": 0.08436176928753658,
"learning_rate": 3.950528574579415e-06,
"loss": 1.2724,
"step": 490
},
{
"epoch": 0.9287054409005628,
"grad_norm": 0.0881012431213454,
"learning_rate": 3.089703021210033e-06,
"loss": 1.2375,
"step": 495
},
{
"epoch": 0.9380863039399625,
"grad_norm": 0.10177275096404065,
"learning_rate": 2.333085148854708e-06,
"loss": 1.2587,
"step": 500
},
{
"epoch": 0.9474671669793621,
"grad_norm": 0.08489096282767854,
"learning_rate": 1.6814885490135102e-06,
"loss": 1.2745,
"step": 505
},
{
"epoch": 0.9568480300187617,
"grad_norm": 0.08633339614238927,
"learning_rate": 1.1356138837702702e-06,
"loss": 1.2495,
"step": 510
},
{
"epoch": 0.9662288930581614,
"grad_norm": 0.0865458432922912,
"learning_rate": 6.960481323703638e-07,
"loss": 1.2604,
"step": 515
},
{
"epoch": 0.975609756097561,
"grad_norm": 0.0850043491504774,
"learning_rate": 3.632639600416932e-07,
"loss": 1.2568,
"step": 520
},
{
"epoch": 0.9849906191369606,
"grad_norm": 0.08694509913799078,
"learning_rate": 1.376192097375495e-07,
"loss": 1.2397,
"step": 525
},
{
"epoch": 0.9943714821763602,
"grad_norm": 0.08603902256571394,
"learning_rate": 1.9356517347990378e-08,
"loss": 1.2735,
"step": 530
},
{
"epoch": 1.0,
"eval_loss": 1.1848469972610474,
"eval_runtime": 1561.8445,
"eval_samples_per_second": 8.566,
"eval_steps_per_second": 0.536,
"step": 533
},
{
"epoch": 1.0,
"step": 533,
"total_flos": 6701415728676864.0,
"train_loss": 1.2950908849208038,
"train_runtime": 12657.6418,
"train_samples_per_second": 2.694,
"train_steps_per_second": 0.042
}
],
"logging_steps": 5,
"max_steps": 533,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 25,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 6701415728676864.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}