Files
tofu_Llama-2-7b-chat-hf_ret…/trainer_state.json

869 lines
21 KiB
JSON
Raw Normal View History

{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 4.968421052631579,
"eval_steps": 500,
"global_step": 590,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.042105263157894736,
"grad_norm": 37.56959101479875,
"learning_rate": 4.2372881355932204e-07,
"loss": 2.7684,
"step": 5
},
{
"epoch": 0.08421052631578947,
"grad_norm": 31.289451739550167,
"learning_rate": 8.474576271186441e-07,
"loss": 2.7059,
"step": 10
},
{
"epoch": 0.12631578947368421,
"grad_norm": 31.193323882177836,
"learning_rate": 1.2711864406779662e-06,
"loss": 2.5582,
"step": 15
},
{
"epoch": 0.16842105263157894,
"grad_norm": 19.559857035516696,
"learning_rate": 1.6949152542372882e-06,
"loss": 2.1059,
"step": 20
},
{
"epoch": 0.21052631578947367,
"grad_norm": 9.217597906831552,
"learning_rate": 2.11864406779661e-06,
"loss": 1.9037,
"step": 25
},
{
"epoch": 0.25263157894736843,
"grad_norm": 10.294915413914124,
"learning_rate": 2.5423728813559323e-06,
"loss": 1.7266,
"step": 30
},
{
"epoch": 0.29473684210526313,
"grad_norm": 9.327403442633655,
"learning_rate": 2.9661016949152545e-06,
"loss": 1.6559,
"step": 35
},
{
"epoch": 0.3368421052631579,
"grad_norm": 7.710428440860352,
"learning_rate": 3.3898305084745763e-06,
"loss": 1.5544,
"step": 40
},
{
"epoch": 0.37894736842105264,
"grad_norm": 5.69482613560667,
"learning_rate": 3.8135593220338985e-06,
"loss": 1.4948,
"step": 45
},
{
"epoch": 0.42105263157894735,
"grad_norm": 5.210507367207414,
"learning_rate": 4.23728813559322e-06,
"loss": 1.4226,
"step": 50
},
{
"epoch": 0.4631578947368421,
"grad_norm": 4.682240928918827,
"learning_rate": 4.6610169491525425e-06,
"loss": 1.4354,
"step": 55
},
{
"epoch": 0.5052631578947369,
"grad_norm": 4.966102260395402,
"learning_rate": 5.084745762711865e-06,
"loss": 1.3991,
"step": 60
},
{
"epoch": 0.5473684210526316,
"grad_norm": 5.75566327559658,
"learning_rate": 5.508474576271187e-06,
"loss": 1.3451,
"step": 65
},
{
"epoch": 0.5894736842105263,
"grad_norm": 4.675561876572965,
"learning_rate": 5.932203389830509e-06,
"loss": 1.3689,
"step": 70
},
{
"epoch": 0.631578947368421,
"grad_norm": 4.7464790635795335,
"learning_rate": 6.3559322033898304e-06,
"loss": 1.3307,
"step": 75
},
{
"epoch": 0.6736842105263158,
"grad_norm": 5.101350767057404,
"learning_rate": 6.779661016949153e-06,
"loss": 1.3383,
"step": 80
},
{
"epoch": 0.7157894736842105,
"grad_norm": 4.817887469772435,
"learning_rate": 7.203389830508475e-06,
"loss": 1.2957,
"step": 85
},
{
"epoch": 0.7578947368421053,
"grad_norm": 4.644863620183367,
"learning_rate": 7.627118644067797e-06,
"loss": 1.2931,
"step": 90
},
{
"epoch": 0.8,
"grad_norm": 4.682998891732904,
"learning_rate": 8.050847457627118e-06,
"loss": 1.2658,
"step": 95
},
{
"epoch": 0.8421052631578947,
"grad_norm": 4.432257920161264,
"learning_rate": 8.47457627118644e-06,
"loss": 1.2024,
"step": 100
},
{
"epoch": 0.8842105263157894,
"grad_norm": 4.511031868472057,
"learning_rate": 8.898305084745763e-06,
"loss": 1.2439,
"step": 105
},
{
"epoch": 0.9263157894736842,
"grad_norm": 4.627996159355769,
"learning_rate": 9.322033898305085e-06,
"loss": 1.2899,
"step": 110
},
{
"epoch": 0.968421052631579,
"grad_norm": 5.098335008818089,
"learning_rate": 9.745762711864407e-06,
"loss": 1.241,
"step": 115
},
{
"epoch": 1.0105263157894737,
"grad_norm": 3.9650974779884938,
"learning_rate": 9.957627118644069e-06,
"loss": 1.1802,
"step": 120
},
{
"epoch": 1.0526315789473684,
"grad_norm": 3.954303638319163,
"learning_rate": 9.851694915254239e-06,
"loss": 1.09,
"step": 125
},
{
"epoch": 1.0947368421052632,
"grad_norm": 4.077358237651598,
"learning_rate": 9.745762711864407e-06,
"loss": 0.9745,
"step": 130
},
{
"epoch": 1.1368421052631579,
"grad_norm": 4.91764657692823,
"learning_rate": 9.639830508474577e-06,
"loss": 1.0262,
"step": 135
},
{
"epoch": 1.1789473684210527,
"grad_norm": 5.3221038272450905,
"learning_rate": 9.533898305084747e-06,
"loss": 1.0054,
"step": 140
},
{
"epoch": 1.2210526315789474,
"grad_norm": 4.764950986365931,
"learning_rate": 9.427966101694917e-06,
"loss": 1.0034,
"step": 145
},
{
"epoch": 1.263157894736842,
"grad_norm": 4.397956284754463,
"learning_rate": 9.322033898305085e-06,
"loss": 0.9692,
"step": 150
},
{
"epoch": 1.305263157894737,
"grad_norm": 5.576640017388553,
"learning_rate": 9.216101694915255e-06,
"loss": 0.9714,
"step": 155
},
{
"epoch": 1.3473684210526315,
"grad_norm": 4.252510823881436,
"learning_rate": 9.110169491525425e-06,
"loss": 0.9341,
"step": 160
},
{
"epoch": 1.3894736842105262,
"grad_norm": 4.722393724784491,
"learning_rate": 9.004237288135595e-06,
"loss": 0.9678,
"step": 165
},
{
"epoch": 1.431578947368421,
"grad_norm": 4.563250808335276,
"learning_rate": 8.898305084745763e-06,
"loss": 1.0039,
"step": 170
},
{
"epoch": 1.4736842105263157,
"grad_norm": 4.552862851357299,
"learning_rate": 8.792372881355933e-06,
"loss": 0.9707,
"step": 175
},
{
"epoch": 1.5157894736842106,
"grad_norm": 5.468925608374858,
"learning_rate": 8.686440677966103e-06,
"loss": 0.9662,
"step": 180
},
{
"epoch": 1.5578947368421052,
"grad_norm": 5.100218329702008,
"learning_rate": 8.580508474576272e-06,
"loss": 0.9582,
"step": 185
},
{
"epoch": 1.6,
"grad_norm": 4.621347397165775,
"learning_rate": 8.47457627118644e-06,
"loss": 0.955,
"step": 190
},
{
"epoch": 1.6421052631578947,
"grad_norm": 4.884622562017621,
"learning_rate": 8.36864406779661e-06,
"loss": 1.009,
"step": 195
},
{
"epoch": 1.6842105263157894,
"grad_norm": 5.2175253410954285,
"learning_rate": 8.26271186440678e-06,
"loss": 0.9902,
"step": 200
},
{
"epoch": 1.7263157894736842,
"grad_norm": 6.359889758676588,
"learning_rate": 8.15677966101695e-06,
"loss": 0.9538,
"step": 205
},
{
"epoch": 1.768421052631579,
"grad_norm": 4.9855484178179665,
"learning_rate": 8.050847457627118e-06,
"loss": 0.94,
"step": 210
},
{
"epoch": 1.8105263157894735,
"grad_norm": 5.172183531187905,
"learning_rate": 7.944915254237288e-06,
"loss": 1.0018,
"step": 215
},
{
"epoch": 1.8526315789473684,
"grad_norm": 4.828811604350556,
"learning_rate": 7.838983050847458e-06,
"loss": 0.9867,
"step": 220
},
{
"epoch": 1.8947368421052633,
"grad_norm": 5.144596691264576,
"learning_rate": 7.733050847457628e-06,
"loss": 0.9916,
"step": 225
},
{
"epoch": 1.936842105263158,
"grad_norm": 4.970464184071689,
"learning_rate": 7.627118644067797e-06,
"loss": 0.9417,
"step": 230
},
{
"epoch": 1.9789473684210526,
"grad_norm": 5.180493471003968,
"learning_rate": 7.521186440677967e-06,
"loss": 0.9583,
"step": 235
},
{
"epoch": 2.0210526315789474,
"grad_norm": 3.9416734314093285,
"learning_rate": 7.415254237288137e-06,
"loss": 0.7218,
"step": 240
},
{
"epoch": 2.0631578947368423,
"grad_norm": 4.619095489451604,
"learning_rate": 7.309322033898306e-06,
"loss": 0.4864,
"step": 245
},
{
"epoch": 2.1052631578947367,
"grad_norm": 5.701097605509544,
"learning_rate": 7.203389830508475e-06,
"loss": 0.4446,
"step": 250
},
{
"epoch": 2.1473684210526316,
"grad_norm": 5.792761603770171,
"learning_rate": 7.097457627118645e-06,
"loss": 0.4579,
"step": 255
},
{
"epoch": 2.1894736842105265,
"grad_norm": 6.270955149892091,
"learning_rate": 6.9915254237288146e-06,
"loss": 0.4519,
"step": 260
},
{
"epoch": 2.231578947368421,
"grad_norm": 5.323086408298113,
"learning_rate": 6.8855932203389844e-06,
"loss": 0.4534,
"step": 265
},
{
"epoch": 2.2736842105263158,
"grad_norm": 6.420773603363344,
"learning_rate": 6.779661016949153e-06,
"loss": 0.4518,
"step": 270
},
{
"epoch": 2.3157894736842106,
"grad_norm": 4.889737887828383,
"learning_rate": 6.6737288135593225e-06,
"loss": 0.4519,
"step": 275
},
{
"epoch": 2.3578947368421055,
"grad_norm": 5.198662251137422,
"learning_rate": 6.567796610169492e-06,
"loss": 0.4843,
"step": 280
},
{
"epoch": 2.4,
"grad_norm": 4.902609500348769,
"learning_rate": 6.461864406779662e-06,
"loss": 0.4617,
"step": 285
},
{
"epoch": 2.442105263157895,
"grad_norm": 5.890700840686636,
"learning_rate": 6.3559322033898304e-06,
"loss": 0.4611,
"step": 290
},
{
"epoch": 2.4842105263157896,
"grad_norm": 5.266072088286826,
"learning_rate": 6.25e-06,
"loss": 0.4678,
"step": 295
},
{
"epoch": 2.526315789473684,
"grad_norm": 5.001862146562411,
"learning_rate": 6.14406779661017e-06,
"loss": 0.4793,
"step": 300
},
{
"epoch": 2.568421052631579,
"grad_norm": 5.885958149708538,
"learning_rate": 6.038135593220339e-06,
"loss": 0.4558,
"step": 305
},
{
"epoch": 2.610526315789474,
"grad_norm": 4.677322542622177,
"learning_rate": 5.932203389830509e-06,
"loss": 0.4429,
"step": 310
},
{
"epoch": 2.6526315789473687,
"grad_norm": 6.29067162665123,
"learning_rate": 5.826271186440678e-06,
"loss": 0.4426,
"step": 315
},
{
"epoch": 2.694736842105263,
"grad_norm": 5.118881563294215,
"learning_rate": 5.720338983050848e-06,
"loss": 0.4886,
"step": 320
},
{
"epoch": 2.736842105263158,
"grad_norm": 5.576669473381538,
"learning_rate": 5.614406779661017e-06,
"loss": 0.4706,
"step": 325
},
{
"epoch": 2.7789473684210524,
"grad_norm": 5.915973784144362,
"learning_rate": 5.508474576271187e-06,
"loss": 0.445,
"step": 330
},
{
"epoch": 2.8210526315789473,
"grad_norm": 5.242589320967943,
"learning_rate": 5.402542372881357e-06,
"loss": 0.4302,
"step": 335
},
{
"epoch": 2.863157894736842,
"grad_norm": 7.920945983664561,
"learning_rate": 5.296610169491526e-06,
"loss": 0.4726,
"step": 340
},
{
"epoch": 2.905263157894737,
"grad_norm": 5.0546857622133405,
"learning_rate": 5.190677966101695e-06,
"loss": 0.4685,
"step": 345
},
{
"epoch": 2.9473684210526314,
"grad_norm": 5.612595665000495,
"learning_rate": 5.084745762711865e-06,
"loss": 0.5103,
"step": 350
},
{
"epoch": 2.9894736842105263,
"grad_norm": 4.728389290445434,
"learning_rate": 4.9788135593220346e-06,
"loss": 0.4377,
"step": 355
},
{
"epoch": 3.031578947368421,
"grad_norm": 3.754417559778345,
"learning_rate": 4.872881355932204e-06,
"loss": 0.2314,
"step": 360
},
{
"epoch": 3.0736842105263156,
"grad_norm": 4.311941194829004,
"learning_rate": 4.7669491525423735e-06,
"loss": 0.1246,
"step": 365
},
{
"epoch": 3.1157894736842104,
"grad_norm": 5.409314123339806,
"learning_rate": 4.6610169491525425e-06,
"loss": 0.1201,
"step": 370
},
{
"epoch": 3.1578947368421053,
"grad_norm": 4.754979765258319,
"learning_rate": 4.555084745762712e-06,
"loss": 0.1123,
"step": 375
},
{
"epoch": 3.2,
"grad_norm": 4.274523577322417,
"learning_rate": 4.449152542372881e-06,
"loss": 0.1091,
"step": 380
},
{
"epoch": 3.2421052631578946,
"grad_norm": 4.34830357359911,
"learning_rate": 4.343220338983051e-06,
"loss": 0.1264,
"step": 385
},
{
"epoch": 3.2842105263157895,
"grad_norm": 5.21825342106621,
"learning_rate": 4.23728813559322e-06,
"loss": 0.1267,
"step": 390
},
{
"epoch": 3.3263157894736843,
"grad_norm": 4.382327715489745,
"learning_rate": 4.13135593220339e-06,
"loss": 0.1198,
"step": 395
},
{
"epoch": 3.3684210526315788,
"grad_norm": 4.736638684494241,
"learning_rate": 4.025423728813559e-06,
"loss": 0.1115,
"step": 400
},
{
"epoch": 3.4105263157894736,
"grad_norm": 3.0871403823899217,
"learning_rate": 3.919491525423729e-06,
"loss": 0.1047,
"step": 405
},
{
"epoch": 3.4526315789473685,
"grad_norm": 3.7883466738808753,
"learning_rate": 3.8135593220338985e-06,
"loss": 0.1151,
"step": 410
},
{
"epoch": 3.4947368421052634,
"grad_norm": 4.58550598062961,
"learning_rate": 3.7076271186440684e-06,
"loss": 0.1121,
"step": 415
},
{
"epoch": 3.536842105263158,
"grad_norm": 4.294675453038568,
"learning_rate": 3.6016949152542374e-06,
"loss": 0.1307,
"step": 420
},
{
"epoch": 3.5789473684210527,
"grad_norm": 5.924960221535346,
"learning_rate": 3.4957627118644073e-06,
"loss": 0.1126,
"step": 425
},
{
"epoch": 3.6210526315789475,
"grad_norm": 4.746834849766813,
"learning_rate": 3.3898305084745763e-06,
"loss": 0.1173,
"step": 430
},
{
"epoch": 3.663157894736842,
"grad_norm": 4.636641100831445,
"learning_rate": 3.283898305084746e-06,
"loss": 0.1136,
"step": 435
},
{
"epoch": 3.705263157894737,
"grad_norm": 3.8235166072561233,
"learning_rate": 3.1779661016949152e-06,
"loss": 0.1103,
"step": 440
},
{
"epoch": 3.7473684210526317,
"grad_norm": 5.689985079126845,
"learning_rate": 3.072033898305085e-06,
"loss": 0.1205,
"step": 445
},
{
"epoch": 3.7894736842105265,
"grad_norm": 4.323007571024886,
"learning_rate": 2.9661016949152545e-06,
"loss": 0.1106,
"step": 450
},
{
"epoch": 3.831578947368421,
"grad_norm": 4.526137345222685,
"learning_rate": 2.860169491525424e-06,
"loss": 0.1076,
"step": 455
},
{
"epoch": 3.873684210526316,
"grad_norm": 3.8633371298181136,
"learning_rate": 2.7542372881355934e-06,
"loss": 0.1124,
"step": 460
},
{
"epoch": 3.9157894736842103,
"grad_norm": 4.477884926304517,
"learning_rate": 2.648305084745763e-06,
"loss": 0.1035,
"step": 465
},
{
"epoch": 3.957894736842105,
"grad_norm": 4.352237219667765,
"learning_rate": 2.5423728813559323e-06,
"loss": 0.109,
"step": 470
},
{
"epoch": 4.0,
"grad_norm": 4.851565652375144,
"learning_rate": 2.436440677966102e-06,
"loss": 0.1075,
"step": 475
},
{
"epoch": 4.042105263157895,
"grad_norm": 1.226350758570929,
"learning_rate": 2.3305084745762712e-06,
"loss": 0.0309,
"step": 480
},
{
"epoch": 4.08421052631579,
"grad_norm": 2.52804331820356,
"learning_rate": 2.2245762711864407e-06,
"loss": 0.0225,
"step": 485
},
{
"epoch": 4.126315789473685,
"grad_norm": 2.1045864497648235,
"learning_rate": 2.11864406779661e-06,
"loss": 0.0258,
"step": 490
},
{
"epoch": 4.168421052631579,
"grad_norm": 1.8501880996749807,
"learning_rate": 2.0127118644067796e-06,
"loss": 0.0208,
"step": 495
},
{
"epoch": 4.2105263157894735,
"grad_norm": 1.7462459629213352,
"learning_rate": 1.9067796610169493e-06,
"loss": 0.0253,
"step": 500
},
{
"epoch": 4.252631578947368,
"grad_norm": 2.769692681231935,
"learning_rate": 1.8008474576271187e-06,
"loss": 0.024,
"step": 505
},
{
"epoch": 4.294736842105263,
"grad_norm": 2.1844146550375747,
"learning_rate": 1.6949152542372882e-06,
"loss": 0.0237,
"step": 510
},
{
"epoch": 4.336842105263158,
"grad_norm": 2.9011926226307443,
"learning_rate": 1.5889830508474576e-06,
"loss": 0.0261,
"step": 515
},
{
"epoch": 4.378947368421053,
"grad_norm": 1.2834311898595507,
"learning_rate": 1.4830508474576273e-06,
"loss": 0.019,
"step": 520
},
{
"epoch": 4.421052631578947,
"grad_norm": 1.6648114999094412,
"learning_rate": 1.3771186440677967e-06,
"loss": 0.02,
"step": 525
},
{
"epoch": 4.463157894736842,
"grad_norm": 1.9849693679452145,
"learning_rate": 1.2711864406779662e-06,
"loss": 0.0219,
"step": 530
},
{
"epoch": 4.505263157894737,
"grad_norm": 3.3305963220896397,
"learning_rate": 1.1652542372881356e-06,
"loss": 0.0206,
"step": 535
},
{
"epoch": 4.5473684210526315,
"grad_norm": 2.311798625936221,
"learning_rate": 1.059322033898305e-06,
"loss": 0.024,
"step": 540
},
{
"epoch": 4.589473684210526,
"grad_norm": 2.247474855883875,
"learning_rate": 9.533898305084746e-07,
"loss": 0.0206,
"step": 545
},
{
"epoch": 4.631578947368421,
"grad_norm": 1.3852190774929651,
"learning_rate": 8.474576271186441e-07,
"loss": 0.0171,
"step": 550
},
{
"epoch": 4.673684210526316,
"grad_norm": 2.057844065940414,
"learning_rate": 7.415254237288136e-07,
"loss": 0.0212,
"step": 555
},
{
"epoch": 4.715789473684211,
"grad_norm": 1.6775650562869804,
"learning_rate": 6.355932203389831e-07,
"loss": 0.0203,
"step": 560
},
{
"epoch": 4.757894736842105,
"grad_norm": 1.0412183276644529,
"learning_rate": 5.296610169491525e-07,
"loss": 0.0207,
"step": 565
},
{
"epoch": 4.8,
"grad_norm": 1.729200767880448,
"learning_rate": 4.2372881355932204e-07,
"loss": 0.0169,
"step": 570
},
{
"epoch": 4.842105263157895,
"grad_norm": 1.5343563742912298,
"learning_rate": 3.1779661016949154e-07,
"loss": 0.015,
"step": 575
},
{
"epoch": 4.88421052631579,
"grad_norm": 2.574691284471041,
"learning_rate": 2.1186440677966102e-07,
"loss": 0.0204,
"step": 580
},
{
"epoch": 4.926315789473684,
"grad_norm": 1.9138929482866247,
"learning_rate": 1.0593220338983051e-07,
"loss": 0.0224,
"step": 585
},
{
"epoch": 4.968421052631579,
"grad_norm": 1.5037996680827372,
"learning_rate": 0.0,
"loss": 0.0165,
"step": 590
},
{
"epoch": 4.968421052631579,
"step": 590,
"total_flos": 2615373496320.0,
"train_loss": 0.6389699679941444,
"train_runtime": 7590.116,
"train_samples_per_second": 2.503,
"train_steps_per_second": 0.078
}
],
"logging_steps": 5,
"max_steps": 590,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": false,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 2615373496320.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}