{
  "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": 59.186122635836746,
      "learning_rate": 4.2372881355932204e-07,
      "loss": 2.4603,
      "step": 5
    },
    {
      "epoch": 0.08421052631578947,
      "grad_norm": 36.87393417015288,
      "learning_rate": 8.474576271186441e-07,
      "loss": 2.3388,
      "step": 10
    },
    {
      "epoch": 0.12631578947368421,
      "grad_norm": 20.043296735403747,
      "learning_rate": 1.2711864406779662e-06,
      "loss": 2.0338,
      "step": 15
    },
    {
      "epoch": 0.16842105263157894,
      "grad_norm": 11.121390904551564,
      "learning_rate": 1.6949152542372882e-06,
      "loss": 1.777,
      "step": 20
    },
    {
      "epoch": 0.21052631578947367,
      "grad_norm": 10.659251053648834,
      "learning_rate": 2.11864406779661e-06,
      "loss": 1.8444,
      "step": 25
    },
    {
      "epoch": 0.25263157894736843,
      "grad_norm": 9.894165657596236,
      "learning_rate": 2.5423728813559323e-06,
      "loss": 1.6908,
      "step": 30
    },
    {
      "epoch": 0.29473684210526313,
      "grad_norm": 8.499052596494911,
      "learning_rate": 2.9661016949152545e-06,
      "loss": 1.6764,
      "step": 35
    },
    {
      "epoch": 0.3368421052631579,
      "grad_norm": 8.631987871555012,
      "learning_rate": 3.3898305084745763e-06,
      "loss": 1.6804,
      "step": 40
    },
    {
      "epoch": 0.37894736842105264,
      "grad_norm": 8.944725403455173,
      "learning_rate": 3.8135593220338985e-06,
      "loss": 1.6147,
      "step": 45
    },
    {
      "epoch": 0.42105263157894735,
      "grad_norm": 7.451416155855185,
      "learning_rate": 4.23728813559322e-06,
      "loss": 1.5405,
      "step": 50
    },
    {
      "epoch": 0.4631578947368421,
      "grad_norm": 7.589628411776622,
      "learning_rate": 4.6610169491525425e-06,
      "loss": 1.6085,
      "step": 55
    },
    {
      "epoch": 0.5052631578947369,
      "grad_norm": 7.44768625468831,
      "learning_rate": 5.084745762711865e-06,
      "loss": 1.5564,
      "step": 60
    },
    {
      "epoch": 0.5473684210526316,
      "grad_norm": 7.352672033290564,
      "learning_rate": 5.508474576271187e-06,
      "loss": 1.4954,
      "step": 65
    },
    {
      "epoch": 0.5894736842105263,
      "grad_norm": 7.616918006430889,
      "learning_rate": 5.932203389830509e-06,
      "loss": 1.5416,
      "step": 70
    },
    {
      "epoch": 0.631578947368421,
      "grad_norm": 6.877320604922845,
      "learning_rate": 6.3559322033898304e-06,
      "loss": 1.5066,
      "step": 75
    },
    {
      "epoch": 0.6736842105263158,
      "grad_norm": 8.546887321959106,
      "learning_rate": 6.779661016949153e-06,
      "loss": 1.5177,
      "step": 80
    },
    {
      "epoch": 0.7157894736842105,
      "grad_norm": 7.627777339872312,
      "learning_rate": 7.203389830508475e-06,
      "loss": 1.4555,
      "step": 85
    },
    {
      "epoch": 0.7578947368421053,
      "grad_norm": 7.255044791164478,
      "learning_rate": 7.627118644067797e-06,
      "loss": 1.484,
      "step": 90
    },
    {
      "epoch": 0.8,
      "grad_norm": 7.216022671329812,
      "learning_rate": 8.050847457627118e-06,
      "loss": 1.4583,
      "step": 95
    },
    {
      "epoch": 0.8421052631578947,
      "grad_norm": 8.229563678387398,
      "learning_rate": 8.47457627118644e-06,
      "loss": 1.3882,
      "step": 100
    },
    {
      "epoch": 0.8842105263157894,
      "grad_norm": 7.843597233096358,
      "learning_rate": 8.898305084745763e-06,
      "loss": 1.438,
      "step": 105
    },
    {
      "epoch": 0.9263157894736842,
      "grad_norm": 9.215162860382458,
      "learning_rate": 9.322033898305085e-06,
      "loss": 1.5054,
      "step": 110
    },
    {
      "epoch": 0.968421052631579,
      "grad_norm": 7.140118548550846,
      "learning_rate": 9.745762711864407e-06,
      "loss": 1.4503,
      "step": 115
    },
    {
      "epoch": 1.0105263157894737,
      "grad_norm": 6.748671626763195,
      "learning_rate": 9.957627118644069e-06,
      "loss": 1.3367,
      "step": 120
    },
    {
      "epoch": 1.0526315789473684,
      "grad_norm": 6.73463664484534,
      "learning_rate": 9.851694915254239e-06,
      "loss": 1.1651,
      "step": 125
    },
    {
      "epoch": 1.0947368421052632,
      "grad_norm": 7.515253638739764,
      "learning_rate": 9.745762711864407e-06,
      "loss": 1.0404,
      "step": 130
    },
    {
      "epoch": 1.1368421052631579,
      "grad_norm": 7.378542745254653,
      "learning_rate": 9.639830508474577e-06,
      "loss": 1.1182,
      "step": 135
    },
    {
      "epoch": 1.1789473684210527,
      "grad_norm": 7.079657658037281,
      "learning_rate": 9.533898305084747e-06,
      "loss": 1.0773,
      "step": 140
    },
    {
      "epoch": 1.2210526315789474,
      "grad_norm": 7.588488747311591,
      "learning_rate": 9.427966101694917e-06,
      "loss": 1.0879,
      "step": 145
    },
    {
      "epoch": 1.263157894736842,
      "grad_norm": 7.1962496775988845,
      "learning_rate": 9.322033898305085e-06,
      "loss": 1.0458,
      "step": 150
    },
    {
      "epoch": 1.305263157894737,
      "grad_norm": 7.122446608522058,
      "learning_rate": 9.216101694915255e-06,
      "loss": 1.0797,
      "step": 155
    },
    {
      "epoch": 1.3473684210526315,
      "grad_norm": 7.266127947775064,
      "learning_rate": 9.110169491525425e-06,
      "loss": 1.0191,
      "step": 160
    },
    {
      "epoch": 1.3894736842105262,
      "grad_norm": 7.198858912559754,
      "learning_rate": 9.004237288135595e-06,
      "loss": 1.0868,
      "step": 165
    },
    {
      "epoch": 1.431578947368421,
      "grad_norm": 7.458407619830278,
      "learning_rate": 8.898305084745763e-06,
      "loss": 1.1146,
      "step": 170
    },
    {
      "epoch": 1.4736842105263157,
      "grad_norm": 8.340366309991538,
      "learning_rate": 8.792372881355933e-06,
      "loss": 1.1261,
      "step": 175
    },
    {
      "epoch": 1.5157894736842106,
      "grad_norm": 7.372927263845646,
      "learning_rate": 8.686440677966103e-06,
      "loss": 1.084,
      "step": 180
    },
    {
      "epoch": 1.5578947368421052,
      "grad_norm": 8.197317712148422,
      "learning_rate": 8.580508474576272e-06,
      "loss": 1.0843,
      "step": 185
    },
    {
      "epoch": 1.6,
      "grad_norm": 6.76287693989083,
      "learning_rate": 8.47457627118644e-06,
      "loss": 1.0717,
      "step": 190
    },
    {
      "epoch": 1.6421052631578947,
      "grad_norm": 7.450549901628139,
      "learning_rate": 8.36864406779661e-06,
      "loss": 1.1449,
      "step": 195
    },
    {
      "epoch": 1.6842105263157894,
      "grad_norm": 7.4530358617487735,
      "learning_rate": 8.26271186440678e-06,
      "loss": 1.1142,
      "step": 200
    },
    {
      "epoch": 1.7263157894736842,
      "grad_norm": 6.831433157315665,
      "learning_rate": 8.15677966101695e-06,
      "loss": 1.0701,
      "step": 205
    },
    {
      "epoch": 1.768421052631579,
      "grad_norm": 7.590684893896947,
      "learning_rate": 8.050847457627118e-06,
      "loss": 1.0692,
      "step": 210
    },
    {
      "epoch": 1.8105263157894735,
      "grad_norm": 7.640124682684029,
      "learning_rate": 7.944915254237288e-06,
      "loss": 1.1,
      "step": 215
    },
    {
      "epoch": 1.8526315789473684,
      "grad_norm": 6.6088440760543135,
      "learning_rate": 7.838983050847458e-06,
      "loss": 1.134,
      "step": 220
    },
    {
      "epoch": 1.8947368421052633,
      "grad_norm": 6.633641946174649,
      "learning_rate": 7.733050847457628e-06,
      "loss": 1.1132,
      "step": 225
    },
    {
      "epoch": 1.936842105263158,
      "grad_norm": 7.227378697083851,
      "learning_rate": 7.627118644067797e-06,
      "loss": 1.0571,
      "step": 230
    },
    {
      "epoch": 1.9789473684210526,
      "grad_norm": 7.640709938150539,
      "learning_rate": 7.521186440677967e-06,
      "loss": 1.0688,
      "step": 235
    },
    {
      "epoch": 2.0210526315789474,
      "grad_norm": 6.35537239406857,
      "learning_rate": 7.415254237288137e-06,
      "loss": 0.7965,
      "step": 240
    },
    {
      "epoch": 2.0631578947368423,
      "grad_norm": 10.91779748014116,
      "learning_rate": 7.309322033898306e-06,
      "loss": 0.4459,
      "step": 245
    },
    {
      "epoch": 2.1052631578947367,
      "grad_norm": 6.517221908196971,
      "learning_rate": 7.203389830508475e-06,
      "loss": 0.4122,
      "step": 250
    },
    {
      "epoch": 2.1473684210526316,
      "grad_norm": 7.5764961638398916,
      "learning_rate": 7.097457627118645e-06,
      "loss": 0.4398,
      "step": 255
    },
    {
      "epoch": 2.1894736842105265,
      "grad_norm": 7.088390575223766,
      "learning_rate": 6.9915254237288146e-06,
      "loss": 0.4203,
      "step": 260
    },
    {
      "epoch": 2.231578947368421,
      "grad_norm": 7.310749803148054,
      "learning_rate": 6.8855932203389844e-06,
      "loss": 0.4501,
      "step": 265
    },
    {
      "epoch": 2.2736842105263158,
      "grad_norm": 7.818080048925102,
      "learning_rate": 6.779661016949153e-06,
      "loss": 0.4338,
      "step": 270
    },
    {
      "epoch": 2.3157894736842106,
      "grad_norm": 6.872060479838489,
      "learning_rate": 6.6737288135593225e-06,
      "loss": 0.4372,
      "step": 275
    },
    {
      "epoch": 2.3578947368421055,
      "grad_norm": 7.0092458158559845,
      "learning_rate": 6.567796610169492e-06,
      "loss": 0.4641,
      "step": 280
    },
    {
      "epoch": 2.4,
      "grad_norm": 7.911914367732163,
      "learning_rate": 6.461864406779662e-06,
      "loss": 0.4657,
      "step": 285
    },
    {
      "epoch": 2.442105263157895,
      "grad_norm": 7.33176850956155,
      "learning_rate": 6.3559322033898304e-06,
      "loss": 0.4287,
      "step": 290
    },
    {
      "epoch": 2.4842105263157896,
      "grad_norm": 7.951572525840537,
      "learning_rate": 6.25e-06,
      "loss": 0.4459,
      "step": 295
    },
    {
      "epoch": 2.526315789473684,
      "grad_norm": 6.964773507356797,
      "learning_rate": 6.14406779661017e-06,
      "loss": 0.4839,
      "step": 300
    },
    {
      "epoch": 2.568421052631579,
      "grad_norm": 6.830391571684712,
      "learning_rate": 6.038135593220339e-06,
      "loss": 0.4466,
      "step": 305
    },
    {
      "epoch": 2.610526315789474,
      "grad_norm": 7.024093265865361,
      "learning_rate": 5.932203389830509e-06,
      "loss": 0.4279,
      "step": 310
    },
    {
      "epoch": 2.6526315789473687,
      "grad_norm": 8.158523666799287,
      "learning_rate": 5.826271186440678e-06,
      "loss": 0.4464,
      "step": 315
    },
    {
      "epoch": 2.694736842105263,
      "grad_norm": 7.764203015817906,
      "learning_rate": 5.720338983050848e-06,
      "loss": 0.4783,
      "step": 320
    },
    {
      "epoch": 2.736842105263158,
      "grad_norm": 7.571384359513728,
      "learning_rate": 5.614406779661017e-06,
      "loss": 0.4518,
      "step": 325
    },
    {
      "epoch": 2.7789473684210524,
      "grad_norm": 7.140383846371695,
      "learning_rate": 5.508474576271187e-06,
      "loss": 0.4421,
      "step": 330
    },
    {
      "epoch": 2.8210526315789473,
      "grad_norm": 7.492523856895696,
      "learning_rate": 5.402542372881357e-06,
      "loss": 0.4312,
      "step": 335
    },
    {
      "epoch": 2.863157894736842,
      "grad_norm": 8.670841798951404,
      "learning_rate": 5.296610169491526e-06,
      "loss": 0.4816,
      "step": 340
    },
    {
      "epoch": 2.905263157894737,
      "grad_norm": 7.834608417874871,
      "learning_rate": 5.190677966101695e-06,
      "loss": 0.4717,
      "step": 345
    },
    {
      "epoch": 2.9473684210526314,
      "grad_norm": 7.033203687936457,
      "learning_rate": 5.084745762711865e-06,
      "loss": 0.501,
      "step": 350
    },
    {
      "epoch": 2.9894736842105263,
      "grad_norm": 6.319066496006612,
      "learning_rate": 4.9788135593220346e-06,
      "loss": 0.4465,
      "step": 355
    },
    {
      "epoch": 3.031578947368421,
      "grad_norm": 4.626859620584807,
      "learning_rate": 4.872881355932204e-06,
      "loss": 0.2172,
      "step": 360
    },
    {
      "epoch": 3.0736842105263156,
      "grad_norm": 9.187991562407264,
      "learning_rate": 4.7669491525423735e-06,
      "loss": 0.1255,
      "step": 365
    },
    {
      "epoch": 3.1157894736842104,
      "grad_norm": 6.865170285628022,
      "learning_rate": 4.6610169491525425e-06,
      "loss": 0.1119,
      "step": 370
    },
    {
      "epoch": 3.1578947368421053,
      "grad_norm": 5.838119353640726,
      "learning_rate": 4.555084745762712e-06,
      "loss": 0.1147,
      "step": 375
    },
    {
      "epoch": 3.2,
      "grad_norm": 4.973038738057853,
      "learning_rate": 4.449152542372881e-06,
      "loss": 0.1026,
      "step": 380
    },
    {
      "epoch": 3.2421052631578946,
      "grad_norm": 5.902818726202673,
      "learning_rate": 4.343220338983051e-06,
      "loss": 0.125,
      "step": 385
    },
    {
      "epoch": 3.2842105263157895,
      "grad_norm": 6.369141263627721,
      "learning_rate": 4.23728813559322e-06,
      "loss": 0.1183,
      "step": 390
    },
    {
      "epoch": 3.3263157894736843,
      "grad_norm": 5.731894647562693,
      "learning_rate": 4.13135593220339e-06,
      "loss": 0.1147,
      "step": 395
    },
    {
      "epoch": 3.3684210526315788,
      "grad_norm": 5.534376819814475,
      "learning_rate": 4.025423728813559e-06,
      "loss": 0.1192,
      "step": 400
    },
    {
      "epoch": 3.4105263157894736,
      "grad_norm": 5.109827709031106,
      "learning_rate": 3.919491525423729e-06,
      "loss": 0.1131,
      "step": 405
    },
    {
      "epoch": 3.4526315789473685,
      "grad_norm": 5.655774073483852,
      "learning_rate": 3.8135593220338985e-06,
      "loss": 0.1122,
      "step": 410
    },
    {
      "epoch": 3.4947368421052634,
      "grad_norm": 6.1563038513596355,
      "learning_rate": 3.7076271186440684e-06,
      "loss": 0.1026,
      "step": 415
    },
    {
      "epoch": 3.536842105263158,
      "grad_norm": 6.189125401975805,
      "learning_rate": 3.6016949152542374e-06,
      "loss": 0.1266,
      "step": 420
    },
    {
      "epoch": 3.5789473684210527,
      "grad_norm": 6.695657527189328,
      "learning_rate": 3.4957627118644073e-06,
      "loss": 0.116,
      "step": 425
    },
    {
      "epoch": 3.6210526315789475,
      "grad_norm": 5.146839355811748,
      "learning_rate": 3.3898305084745763e-06,
      "loss": 0.1094,
      "step": 430
    },
    {
      "epoch": 3.663157894736842,
      "grad_norm": 4.63918516753777,
      "learning_rate": 3.283898305084746e-06,
      "loss": 0.104,
      "step": 435
    },
    {
      "epoch": 3.705263157894737,
      "grad_norm": 5.387327579196708,
      "learning_rate": 3.1779661016949152e-06,
      "loss": 0.1197,
      "step": 440
    },
    {
      "epoch": 3.7473684210526317,
      "grad_norm": 5.514817746762858,
      "learning_rate": 3.072033898305085e-06,
      "loss": 0.1076,
      "step": 445
    },
    {
      "epoch": 3.7894736842105265,
      "grad_norm": 5.6423366322507436,
      "learning_rate": 2.9661016949152545e-06,
      "loss": 0.1115,
      "step": 450
    },
    {
      "epoch": 3.831578947368421,
      "grad_norm": 5.950468486954442,
      "learning_rate": 2.860169491525424e-06,
      "loss": 0.1152,
      "step": 455
    },
    {
      "epoch": 3.873684210526316,
      "grad_norm": 4.498597911195406,
      "learning_rate": 2.7542372881355934e-06,
      "loss": 0.1019,
      "step": 460
    },
    {
      "epoch": 3.9157894736842103,
      "grad_norm": 6.20055694554923,
      "learning_rate": 2.648305084745763e-06,
      "loss": 0.1042,
      "step": 465
    },
    {
      "epoch": 3.957894736842105,
      "grad_norm": 4.907554759496963,
      "learning_rate": 2.5423728813559323e-06,
      "loss": 0.1024,
      "step": 470
    },
    {
      "epoch": 4.0,
      "grad_norm": 6.0664642679300735,
      "learning_rate": 2.436440677966102e-06,
      "loss": 0.1124,
      "step": 475
    },
    {
      "epoch": 4.042105263157895,
      "grad_norm": 2.0320210838584862,
      "learning_rate": 2.3305084745762712e-06,
      "loss": 0.0324,
      "step": 480
    },
    {
      "epoch": 4.08421052631579,
      "grad_norm": 2.0587678590102607,
      "learning_rate": 2.2245762711864407e-06,
      "loss": 0.0243,
      "step": 485
    },
    {
      "epoch": 4.126315789473685,
      "grad_norm": 2.6001000088906228,
      "learning_rate": 2.11864406779661e-06,
      "loss": 0.0234,
      "step": 490
    },
    {
      "epoch": 4.168421052631579,
      "grad_norm": 2.9546541617587123,
      "learning_rate": 2.0127118644067796e-06,
      "loss": 0.0198,
      "step": 495
    },
    {
      "epoch": 4.2105263157894735,
      "grad_norm": 2.7479394162518616,
      "learning_rate": 1.9067796610169493e-06,
      "loss": 0.0191,
      "step": 500
    },
    {
      "epoch": 4.252631578947368,
      "grad_norm": 2.585400769515734,
      "learning_rate": 1.8008474576271187e-06,
      "loss": 0.026,
      "step": 505
    },
    {
      "epoch": 4.294736842105263,
      "grad_norm": 3.45010915438713,
      "learning_rate": 1.6949152542372882e-06,
      "loss": 0.0231,
      "step": 510
    },
    {
      "epoch": 4.336842105263158,
      "grad_norm": 3.3324197388715024,
      "learning_rate": 1.5889830508474576e-06,
      "loss": 0.022,
      "step": 515
    },
    {
      "epoch": 4.378947368421053,
      "grad_norm": 2.5027087319574775,
      "learning_rate": 1.4830508474576273e-06,
      "loss": 0.0213,
      "step": 520
    },
    {
      "epoch": 4.421052631578947,
      "grad_norm": 2.0395665317290166,
      "learning_rate": 1.3771186440677967e-06,
      "loss": 0.0228,
      "step": 525
    },
    {
      "epoch": 4.463157894736842,
      "grad_norm": 3.0293656616479736,
      "learning_rate": 1.2711864406779662e-06,
      "loss": 0.0137,
      "step": 530
    },
    {
      "epoch": 4.505263157894737,
      "grad_norm": 1.8109024541694492,
      "learning_rate": 1.1652542372881356e-06,
      "loss": 0.0135,
      "step": 535
    },
    {
      "epoch": 4.5473684210526315,
      "grad_norm": 5.2315694190949005,
      "learning_rate": 1.059322033898305e-06,
      "loss": 0.0196,
      "step": 540
    },
    {
      "epoch": 4.589473684210526,
      "grad_norm": 2.573193328478675,
      "learning_rate": 9.533898305084746e-07,
      "loss": 0.0167,
      "step": 545
    },
    {
      "epoch": 4.631578947368421,
      "grad_norm": 2.468444546799618,
      "learning_rate": 8.474576271186441e-07,
      "loss": 0.0158,
      "step": 550
    },
    {
      "epoch": 4.673684210526316,
      "grad_norm": 3.04193368871112,
      "learning_rate": 7.415254237288136e-07,
      "loss": 0.0187,
      "step": 555
    },
    {
      "epoch": 4.715789473684211,
      "grad_norm": 3.4162557440577874,
      "learning_rate": 6.355932203389831e-07,
      "loss": 0.0179,
      "step": 560
    },
    {
      "epoch": 4.757894736842105,
      "grad_norm": 2.5544455082391826,
      "learning_rate": 5.296610169491525e-07,
      "loss": 0.0213,
      "step": 565
    },
    {
      "epoch": 4.8,
      "grad_norm": 1.519910879343402,
      "learning_rate": 4.2372881355932204e-07,
      "loss": 0.0133,
      "step": 570
    },
    {
      "epoch": 4.842105263157895,
      "grad_norm": 2.822666300583857,
      "learning_rate": 3.1779661016949154e-07,
      "loss": 0.0179,
      "step": 575
    },
    {
      "epoch": 4.88421052631579,
      "grad_norm": 4.704566025745165,
      "learning_rate": 2.1186440677966102e-07,
      "loss": 0.0216,
      "step": 580
    },
    {
      "epoch": 4.926315789473684,
      "grad_norm": 1.5555788465593001,
      "learning_rate": 1.0593220338983051e-07,
      "loss": 0.0142,
      "step": 585
    },
    {
      "epoch": 4.968421052631579,
      "grad_norm": 2.454426736258744,
      "learning_rate": 0.0,
      "loss": 0.018,
      "step": 590
    },
    {
      "epoch": 4.968421052631579,
      "step": 590,
      "total_flos": 3206125977600.0,
      "train_loss": 0.6685241217709194,
      "train_runtime": 7829.319,
      "train_samples_per_second": 2.427,
      "train_steps_per_second": 0.075
    }
  ],
  "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": 3206125977600.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}