opt-peter-1.3B / trainer_state.json
pszemraj's picture
add model files straight-outta-trainer
5fea7c5
raw
history blame
No virus
35.7 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.1828941378787112,
"global_step": 1461,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"learning_rate": 6e-06,
"loss": 1.9846,
"step": 5
},
{
"epoch": 0.01,
"learning_rate": 1.2e-05,
"loss": 1.8345,
"step": 10
},
{
"epoch": 0.01,
"learning_rate": 1.8e-05,
"loss": 1.7106,
"step": 15
},
{
"epoch": 0.02,
"learning_rate": 2.4e-05,
"loss": 1.6318,
"step": 20
},
{
"epoch": 0.02,
"learning_rate": 3e-05,
"loss": 1.5838,
"step": 25
},
{
"epoch": 0.02,
"learning_rate": 2.9999690442167746e-05,
"loss": 1.6664,
"step": 30
},
{
"epoch": 0.03,
"learning_rate": 2.999876178144779e-05,
"loss": 1.5785,
"step": 35
},
{
"epoch": 0.03,
"learning_rate": 2.9997214056170024e-05,
"loss": 1.6536,
"step": 40
},
{
"epoch": 0.04,
"learning_rate": 2.9995047330215847e-05,
"loss": 1.5866,
"step": 45
},
{
"epoch": 0.04,
"learning_rate": 2.9992261693015524e-05,
"loss": 1.6324,
"step": 50
},
{
"epoch": 0.04,
"learning_rate": 2.9988857259544498e-05,
"loss": 1.6263,
"step": 55
},
{
"epoch": 0.05,
"learning_rate": 2.9984834170318635e-05,
"loss": 1.6481,
"step": 60
},
{
"epoch": 0.05,
"learning_rate": 2.9980192591388452e-05,
"loss": 1.6703,
"step": 65
},
{
"epoch": 0.06,
"learning_rate": 2.997493271433222e-05,
"loss": 1.6566,
"step": 70
},
{
"epoch": 0.06,
"learning_rate": 2.9969054756248093e-05,
"loss": 1.6168,
"step": 75
},
{
"epoch": 0.06,
"learning_rate": 2.9962558959745133e-05,
"loss": 1.6836,
"step": 80
},
{
"epoch": 0.07,
"learning_rate": 2.9955445592933296e-05,
"loss": 1.7089,
"step": 85
},
{
"epoch": 0.07,
"learning_rate": 2.994771494941238e-05,
"loss": 1.6573,
"step": 90
},
{
"epoch": 0.08,
"learning_rate": 2.9939367348259873e-05,
"loss": 1.5558,
"step": 95
},
{
"epoch": 0.08,
"learning_rate": 2.9930403134017823e-05,
"loss": 1.6069,
"step": 100
},
{
"epoch": 0.08,
"learning_rate": 2.992082267667859e-05,
"loss": 1.6357,
"step": 105
},
{
"epoch": 0.09,
"learning_rate": 2.9910626371669593e-05,
"loss": 1.5968,
"step": 110
},
{
"epoch": 0.09,
"learning_rate": 2.9899814639836972e-05,
"loss": 1.6244,
"step": 115
},
{
"epoch": 0.1,
"learning_rate": 2.9888387927428234e-05,
"loss": 1.5168,
"step": 120
},
{
"epoch": 0.1,
"learning_rate": 2.987634670607381e-05,
"loss": 1.6162,
"step": 125
},
{
"epoch": 0.11,
"learning_rate": 2.9863691472767633e-05,
"loss": 1.6014,
"step": 130
},
{
"epoch": 0.11,
"learning_rate": 2.9850422749846577e-05,
"loss": 1.6596,
"step": 135
},
{
"epoch": 0.11,
"learning_rate": 2.9836541084968914e-05,
"loss": 1.6175,
"step": 140
},
{
"epoch": 0.12,
"learning_rate": 2.9822047051091735e-05,
"loss": 1.6219,
"step": 145
},
{
"epoch": 0.12,
"learning_rate": 2.9806941246447258e-05,
"loss": 1.5605,
"step": 150
},
{
"epoch": 0.13,
"learning_rate": 2.9791224294518173e-05,
"loss": 1.6062,
"step": 155
},
{
"epoch": 0.13,
"learning_rate": 2.9774896844011887e-05,
"loss": 1.5561,
"step": 160
},
{
"epoch": 0.13,
"learning_rate": 2.9757959568833758e-05,
"loss": 1.5929,
"step": 165
},
{
"epoch": 0.14,
"learning_rate": 2.9740413168059278e-05,
"loss": 1.6087,
"step": 170
},
{
"epoch": 0.14,
"learning_rate": 2.9722258365905223e-05,
"loss": 1.603,
"step": 175
},
{
"epoch": 0.15,
"learning_rate": 2.9703495911699746e-05,
"loss": 1.5866,
"step": 180
},
{
"epoch": 0.15,
"learning_rate": 2.9684126579851468e-05,
"loss": 1.696,
"step": 185
},
{
"epoch": 0.15,
"learning_rate": 2.9664151169817515e-05,
"loss": 1.59,
"step": 190
},
{
"epoch": 0.16,
"learning_rate": 2.9643570506070493e-05,
"loss": 1.5724,
"step": 195
},
{
"epoch": 0.16,
"learning_rate": 2.9622385438064493e-05,
"loss": 1.5482,
"step": 200
},
{
"epoch": 0.17,
"learning_rate": 2.9600596840200022e-05,
"loss": 1.6198,
"step": 205
},
{
"epoch": 0.17,
"learning_rate": 2.9578205611787877e-05,
"loss": 1.6246,
"step": 210
},
{
"epoch": 0.17,
"learning_rate": 2.955521267701209e-05,
"loss": 1.6953,
"step": 215
},
{
"epoch": 0.18,
"learning_rate": 2.9531618984891737e-05,
"loss": 1.6397,
"step": 220
},
{
"epoch": 0.18,
"learning_rate": 2.9507425509241757e-05,
"loss": 1.5797,
"step": 225
},
{
"epoch": 0.19,
"learning_rate": 2.948263324863282e-05,
"loss": 1.63,
"step": 230
},
{
"epoch": 0.19,
"learning_rate": 2.945724322635004e-05,
"loss": 1.5258,
"step": 235
},
{
"epoch": 0.19,
"learning_rate": 2.9431256490350795e-05,
"loss": 1.6107,
"step": 240
},
{
"epoch": 0.2,
"learning_rate": 2.9404674113221433e-05,
"loss": 1.5577,
"step": 245
},
{
"epoch": 0.2,
"learning_rate": 2.937749719213303e-05,
"loss": 1.6001,
"step": 250
},
{
"epoch": 0.21,
"learning_rate": 2.9349726848796083e-05,
"loss": 1.6326,
"step": 255
},
{
"epoch": 0.21,
"learning_rate": 2.932136422941424e-05,
"loss": 1.5575,
"step": 260
},
{
"epoch": 0.21,
"learning_rate": 2.929241050463696e-05,
"loss": 1.692,
"step": 265
},
{
"epoch": 0.22,
"learning_rate": 2.926286686951121e-05,
"loss": 1.5985,
"step": 270
},
{
"epoch": 0.22,
"learning_rate": 2.9232734543432146e-05,
"loss": 1.5635,
"step": 275
},
{
"epoch": 0.23,
"learning_rate": 2.920201477009277e-05,
"loss": 1.6745,
"step": 280
},
{
"epoch": 0.23,
"learning_rate": 2.9170708817432612e-05,
"loss": 1.5613,
"step": 285
},
{
"epoch": 0.23,
"learning_rate": 2.9138817977585383e-05,
"loss": 1.6063,
"step": 290
},
{
"epoch": 0.24,
"learning_rate": 2.910634356682565e-05,
"loss": 1.6118,
"step": 295
},
{
"epoch": 0.24,
"learning_rate": 2.9073286925514504e-05,
"loss": 1.5912,
"step": 300
},
{
"epoch": 0.25,
"learning_rate": 2.9039649418044247e-05,
"loss": 1.596,
"step": 305
},
{
"epoch": 0.25,
"learning_rate": 2.900543243278206e-05,
"loss": 1.5501,
"step": 310
},
{
"epoch": 0.25,
"learning_rate": 2.8970637382012714e-05,
"loss": 1.6077,
"step": 315
},
{
"epoch": 0.26,
"learning_rate": 2.8935265701880277e-05,
"loss": 1.5393,
"step": 320
},
{
"epoch": 0.26,
"learning_rate": 2.8899318852328833e-05,
"loss": 1.5622,
"step": 325
},
{
"epoch": 0.27,
"learning_rate": 2.8862798317042222e-05,
"loss": 1.5683,
"step": 330
},
{
"epoch": 0.27,
"learning_rate": 2.882570560338281e-05,
"loss": 1.5644,
"step": 335
},
{
"epoch": 0.28,
"learning_rate": 2.878804224232927e-05,
"loss": 1.6027,
"step": 340
},
{
"epoch": 0.28,
"learning_rate": 2.8749809788413383e-05,
"loss": 1.5418,
"step": 345
},
{
"epoch": 0.28,
"learning_rate": 2.87110098196559e-05,
"loss": 1.5934,
"step": 350
},
{
"epoch": 0.29,
"learning_rate": 2.8671643937501375e-05,
"loss": 1.5265,
"step": 355
},
{
"epoch": 0.29,
"learning_rate": 2.8631713766752097e-05,
"loss": 1.6044,
"step": 360
},
{
"epoch": 0.3,
"learning_rate": 2.859122095550101e-05,
"loss": 1.6244,
"step": 365
},
{
"epoch": 0.3,
"learning_rate": 2.8550167175063705e-05,
"loss": 1.5721,
"step": 370
},
{
"epoch": 0.3,
"learning_rate": 2.850855411990941e-05,
"loss": 1.5572,
"step": 375
},
{
"epoch": 0.31,
"learning_rate": 2.8466383507591083e-05,
"loss": 1.5034,
"step": 380
},
{
"epoch": 0.31,
"learning_rate": 2.84236570786745e-05,
"loss": 1.6026,
"step": 385
},
{
"epoch": 0.32,
"learning_rate": 2.8380376596666425e-05,
"loss": 1.6428,
"step": 390
},
{
"epoch": 0.32,
"learning_rate": 2.833654384794181e-05,
"loss": 1.5364,
"step": 395
},
{
"epoch": 0.32,
"learning_rate": 2.8292160641670088e-05,
"loss": 1.5064,
"step": 400
},
{
"epoch": 0.33,
"learning_rate": 2.8247228809740468e-05,
"loss": 1.6338,
"step": 405
},
{
"epoch": 0.33,
"learning_rate": 2.820175020668635e-05,
"loss": 1.5189,
"step": 410
},
{
"epoch": 0.34,
"learning_rate": 2.8155726709608777e-05,
"loss": 1.5359,
"step": 415
},
{
"epoch": 0.34,
"learning_rate": 2.810916021809894e-05,
"loss": 1.6113,
"step": 420
},
{
"epoch": 0.34,
"learning_rate": 2.8062052654159797e-05,
"loss": 1.5724,
"step": 425
},
{
"epoch": 0.35,
"learning_rate": 2.8014405962126735e-05,
"loss": 1.5174,
"step": 430
},
{
"epoch": 0.35,
"learning_rate": 2.7966222108587307e-05,
"loss": 1.5741,
"step": 435
},
{
"epoch": 0.36,
"learning_rate": 2.7917503082300086e-05,
"loss": 1.595,
"step": 440
},
{
"epoch": 0.36,
"learning_rate": 2.7868250894112555e-05,
"loss": 1.6298,
"step": 445
},
{
"epoch": 0.36,
"learning_rate": 2.7818467576878136e-05,
"loss": 1.5871,
"step": 450
},
{
"epoch": 0.37,
"learning_rate": 2.776815518537226e-05,
"loss": 1.5841,
"step": 455
},
{
"epoch": 0.37,
"learning_rate": 2.7717315796207576e-05,
"loss": 1.5128,
"step": 460
},
{
"epoch": 0.38,
"learning_rate": 2.7665951507748223e-05,
"loss": 1.5055,
"step": 465
},
{
"epoch": 0.38,
"learning_rate": 2.7614064440023254e-05,
"loss": 1.6022,
"step": 470
},
{
"epoch": 0.38,
"learning_rate": 2.7561656734639085e-05,
"loss": 1.5673,
"step": 475
},
{
"epoch": 0.39,
"learning_rate": 2.7508730554691145e-05,
"loss": 1.5504,
"step": 480
},
{
"epoch": 0.39,
"learning_rate": 2.7455288084674565e-05,
"loss": 1.6085,
"step": 485
},
{
"epoch": 0.4,
"learning_rate": 2.7401331530394037e-05,
"loss": 1.5598,
"step": 490
},
{
"epoch": 0.4,
"learning_rate": 2.7346863118872766e-05,
"loss": 1.5559,
"step": 495
},
{
"epoch": 0.4,
"learning_rate": 2.729188509826053e-05,
"loss": 1.5366,
"step": 500
},
{
"epoch": 0.41,
"learning_rate": 2.7236399737740912e-05,
"loss": 1.5431,
"step": 505
},
{
"epoch": 0.41,
"learning_rate": 2.7180409327437648e-05,
"loss": 1.58,
"step": 510
},
{
"epoch": 0.42,
"learning_rate": 2.712391617832006e-05,
"loss": 1.5656,
"step": 515
},
{
"epoch": 0.42,
"learning_rate": 2.7066922622107726e-05,
"loss": 1.5488,
"step": 520
},
{
"epoch": 0.42,
"learning_rate": 2.700943101117421e-05,
"loss": 1.563,
"step": 525
},
{
"epoch": 0.43,
"learning_rate": 2.6951443718449966e-05,
"loss": 1.5845,
"step": 530
},
{
"epoch": 0.43,
"learning_rate": 2.689296313732442e-05,
"loss": 1.5782,
"step": 535
},
{
"epoch": 0.44,
"learning_rate": 2.6833991681547158e-05,
"loss": 1.6135,
"step": 540
},
{
"epoch": 0.44,
"learning_rate": 2.677453178512831e-05,
"loss": 1.5745,
"step": 545
},
{
"epoch": 0.45,
"learning_rate": 2.6714585902238105e-05,
"loss": 1.6138,
"step": 550
},
{
"epoch": 0.45,
"learning_rate": 2.6654156507105543e-05,
"loss": 1.5663,
"step": 555
},
{
"epoch": 0.45,
"learning_rate": 2.6593246093916307e-05,
"loss": 1.5842,
"step": 560
},
{
"epoch": 0.46,
"learning_rate": 2.653185717670978e-05,
"loss": 1.5424,
"step": 565
},
{
"epoch": 0.46,
"learning_rate": 2.6469992289275325e-05,
"loss": 1.5542,
"step": 570
},
{
"epoch": 0.47,
"learning_rate": 2.6407653985047665e-05,
"loss": 1.5571,
"step": 575
},
{
"epoch": 0.47,
"learning_rate": 2.6344844837001508e-05,
"loss": 1.5838,
"step": 580
},
{
"epoch": 0.47,
"learning_rate": 2.6281567437545347e-05,
"loss": 1.5626,
"step": 585
},
{
"epoch": 0.48,
"learning_rate": 2.6217824398414473e-05,
"loss": 1.5316,
"step": 590
},
{
"epoch": 0.48,
"learning_rate": 2.615361835056314e-05,
"loss": 1.6038,
"step": 595
},
{
"epoch": 0.49,
"learning_rate": 2.6088951944056024e-05,
"loss": 1.6042,
"step": 600
},
{
"epoch": 0.49,
"learning_rate": 2.6023827847958802e-05,
"loss": 1.5683,
"step": 605
},
{
"epoch": 0.49,
"learning_rate": 2.5958248750228018e-05,
"loss": 1.5166,
"step": 610
},
{
"epoch": 0.5,
"learning_rate": 2.5892217357600104e-05,
"loss": 1.5522,
"step": 615
},
{
"epoch": 0.5,
"learning_rate": 2.5825736395479708e-05,
"loss": 1.6122,
"step": 620
},
{
"epoch": 0.51,
"learning_rate": 2.5758808607827153e-05,
"loss": 1.5722,
"step": 625
},
{
"epoch": 0.51,
"learning_rate": 2.5691436757045232e-05,
"loss": 1.5333,
"step": 630
},
{
"epoch": 0.51,
"learning_rate": 2.5623623623865152e-05,
"loss": 1.563,
"step": 635
},
{
"epoch": 0.52,
"learning_rate": 2.5555372007231777e-05,
"loss": 1.5318,
"step": 640
},
{
"epoch": 0.52,
"learning_rate": 2.548668472418811e-05,
"loss": 1.548,
"step": 645
},
{
"epoch": 0.53,
"learning_rate": 2.5417564609759005e-05,
"loss": 1.5749,
"step": 650
},
{
"epoch": 0.53,
"learning_rate": 2.5348014516834175e-05,
"loss": 1.6166,
"step": 655
},
{
"epoch": 0.53,
"learning_rate": 2.5278037316050417e-05,
"loss": 1.61,
"step": 660
},
{
"epoch": 0.54,
"learning_rate": 2.5207635895673138e-05,
"loss": 1.4852,
"step": 665
},
{
"epoch": 0.54,
"learning_rate": 2.513681316147715e-05,
"loss": 1.5476,
"step": 670
},
{
"epoch": 0.55,
"learning_rate": 2.506557203662673e-05,
"loss": 1.6291,
"step": 675
},
{
"epoch": 0.55,
"learning_rate": 2.4993915461554974e-05,
"loss": 1.5232,
"step": 680
},
{
"epoch": 0.55,
"learning_rate": 2.4921846393842414e-05,
"loss": 1.5827,
"step": 685
},
{
"epoch": 0.56,
"learning_rate": 2.484936780809497e-05,
"loss": 1.5019,
"step": 690
},
{
"epoch": 0.56,
"learning_rate": 2.4776482695821154e-05,
"loss": 1.5315,
"step": 695
},
{
"epoch": 0.57,
"learning_rate": 2.470319406530862e-05,
"loss": 1.5725,
"step": 700
},
{
"epoch": 0.57,
"learning_rate": 2.4629504941499984e-05,
"loss": 1.6135,
"step": 705
},
{
"epoch": 0.57,
"learning_rate": 2.4555418365867965e-05,
"loss": 1.556,
"step": 710
},
{
"epoch": 0.58,
"learning_rate": 2.4480937396289856e-05,
"loss": 1.5323,
"step": 715
},
{
"epoch": 0.58,
"learning_rate": 2.4406065106921332e-05,
"loss": 1.5726,
"step": 720
},
{
"epoch": 0.59,
"learning_rate": 2.4330804588069536e-05,
"loss": 1.6077,
"step": 725
},
{
"epoch": 0.59,
"learning_rate": 2.4255158946065542e-05,
"loss": 1.5959,
"step": 730
},
{
"epoch": 0.59,
"learning_rate": 2.4179131303136146e-05,
"loss": 1.5373,
"step": 735
},
{
"epoch": 0.6,
"learning_rate": 2.4102724797274994e-05,
"loss": 1.5624,
"step": 740
},
{
"epoch": 0.6,
"learning_rate": 2.4025942582113067e-05,
"loss": 1.5431,
"step": 745
},
{
"epoch": 0.61,
"learning_rate": 2.3948787826788495e-05,
"loss": 1.5807,
"step": 750
},
{
"epoch": 0.61,
"learning_rate": 2.3871263715815802e-05,
"loss": 1.5659,
"step": 755
},
{
"epoch": 0.62,
"learning_rate": 2.3793373448954406e-05,
"loss": 1.5356,
"step": 760
},
{
"epoch": 0.62,
"learning_rate": 2.3715120241076602e-05,
"loss": 1.5181,
"step": 765
},
{
"epoch": 0.62,
"learning_rate": 2.3636507322034844e-05,
"loss": 1.6218,
"step": 770
},
{
"epoch": 0.63,
"learning_rate": 2.355753793652844e-05,
"loss": 1.6171,
"step": 775
},
{
"epoch": 0.63,
"learning_rate": 2.3478215343969623e-05,
"loss": 1.5306,
"step": 780
},
{
"epoch": 0.64,
"learning_rate": 2.3398542818349042e-05,
"loss": 1.5147,
"step": 785
},
{
"epoch": 0.64,
"learning_rate": 2.33185236481006e-05,
"loss": 1.5031,
"step": 790
},
{
"epoch": 0.64,
"learning_rate": 2.323816113596575e-05,
"loss": 1.5471,
"step": 795
},
{
"epoch": 0.65,
"learning_rate": 2.3157458598857164e-05,
"loss": 1.5904,
"step": 800
},
{
"epoch": 0.65,
"learning_rate": 2.3076419367721834e-05,
"loss": 1.6055,
"step": 805
},
{
"epoch": 0.66,
"learning_rate": 2.299504678740359e-05,
"loss": 1.6248,
"step": 810
},
{
"epoch": 0.66,
"learning_rate": 2.2913344216505043e-05,
"loss": 1.5905,
"step": 815
},
{
"epoch": 0.66,
"learning_rate": 2.283131502724896e-05,
"loss": 1.4956,
"step": 820
},
{
"epoch": 0.67,
"learning_rate": 2.2748962605339066e-05,
"loss": 1.537,
"step": 825
},
{
"epoch": 0.67,
"learning_rate": 2.266629034982033e-05,
"loss": 1.5609,
"step": 830
},
{
"epoch": 0.68,
"learning_rate": 2.2583301672938648e-05,
"loss": 1.6342,
"step": 835
},
{
"epoch": 0.68,
"learning_rate": 2.25e-05,
"loss": 1.6218,
"step": 840
},
{
"epoch": 0.68,
"learning_rate": 2.24163887692291e-05,
"loss": 1.6035,
"step": 845
},
{
"epoch": 0.69,
"learning_rate": 2.233247143162746e-05,
"loss": 1.4957,
"step": 850
},
{
"epoch": 0.69,
"learning_rate": 2.224825145083096e-05,
"loss": 1.4736,
"step": 855
},
{
"epoch": 0.7,
"learning_rate": 2.216373230296689e-05,
"loss": 1.581,
"step": 860
},
{
"epoch": 0.7,
"learning_rate": 2.2078917476510483e-05,
"loss": 1.6488,
"step": 865
},
{
"epoch": 0.7,
"learning_rate": 2.1993810472140908e-05,
"loss": 1.5308,
"step": 870
},
{
"epoch": 0.71,
"learning_rate": 2.190841480259681e-05,
"loss": 1.4948,
"step": 875
},
{
"epoch": 0.71,
"learning_rate": 2.1822733992531294e-05,
"loss": 1.5796,
"step": 880
},
{
"epoch": 0.72,
"learning_rate": 2.1736771578366472e-05,
"loss": 1.6163,
"step": 885
},
{
"epoch": 0.72,
"learning_rate": 2.1650531108147493e-05,
"loss": 1.5086,
"step": 890
},
{
"epoch": 0.72,
"learning_rate": 2.1564016141396093e-05,
"loss": 1.564,
"step": 895
},
{
"epoch": 0.73,
"learning_rate": 2.1477230248963675e-05,
"loss": 1.6207,
"step": 900
},
{
"epoch": 0.73,
"learning_rate": 2.139017701288394e-05,
"loss": 1.5649,
"step": 905
},
{
"epoch": 0.74,
"learning_rate": 2.1302860026225027e-05,
"loss": 1.5835,
"step": 910
},
{
"epoch": 0.74,
"learning_rate": 2.121528289294122e-05,
"loss": 1.5632,
"step": 915
},
{
"epoch": 0.74,
"learning_rate": 2.1127449227724186e-05,
"loss": 1.5381,
"step": 920
},
{
"epoch": 0.75,
"learning_rate": 2.1039362655853796e-05,
"loss": 1.5546,
"step": 925
},
{
"epoch": 0.75,
"learning_rate": 2.0951026813048475e-05,
"loss": 1.4658,
"step": 930
},
{
"epoch": 0.76,
"learning_rate": 2.0862445345315165e-05,
"loss": 1.5073,
"step": 935
},
{
"epoch": 0.76,
"learning_rate": 2.0773621908798818e-05,
"loss": 1.5564,
"step": 940
},
{
"epoch": 0.76,
"learning_rate": 2.068456016963149e-05,
"loss": 1.6254,
"step": 945
},
{
"epoch": 0.77,
"learning_rate": 2.0595263803781037e-05,
"loss": 1.5497,
"step": 950
},
{
"epoch": 0.77,
"learning_rate": 2.050573649689938e-05,
"loss": 1.5791,
"step": 955
},
{
"epoch": 0.78,
"learning_rate": 2.0415981944170405e-05,
"loss": 1.4777,
"step": 960
},
{
"epoch": 0.78,
"learning_rate": 2.0326003850157408e-05,
"loss": 1.5316,
"step": 965
},
{
"epoch": 0.78,
"learning_rate": 2.0235805928650214e-05,
"loss": 1.5288,
"step": 970
},
{
"epoch": 0.79,
"learning_rate": 2.0145391902511905e-05,
"loss": 1.5519,
"step": 975
},
{
"epoch": 0.79,
"learning_rate": 2.0054765503525136e-05,
"loss": 1.5359,
"step": 980
},
{
"epoch": 0.8,
"learning_rate": 1.9963930472238126e-05,
"loss": 1.5367,
"step": 985
},
{
"epoch": 0.8,
"learning_rate": 1.9872890557810258e-05,
"loss": 1.5804,
"step": 990
},
{
"epoch": 0.81,
"learning_rate": 1.978164951785733e-05,
"loss": 1.6531,
"step": 995
},
{
"epoch": 0.81,
"learning_rate": 1.96902111182965e-05,
"loss": 1.6579,
"step": 1000
},
{
"epoch": 0.81,
"learning_rate": 1.959857913319078e-05,
"loss": 1.6308,
"step": 1005
},
{
"epoch": 0.82,
"learning_rate": 1.9506757344593345e-05,
"loss": 1.6248,
"step": 1010
},
{
"epoch": 0.82,
"learning_rate": 1.941474954239136e-05,
"loss": 1.5152,
"step": 1015
},
{
"epoch": 0.83,
"learning_rate": 1.9322559524149603e-05,
"loss": 1.5966,
"step": 1020
},
{
"epoch": 0.83,
"learning_rate": 1.92301910949537e-05,
"loss": 1.5624,
"step": 1025
},
{
"epoch": 0.83,
"learning_rate": 1.9137648067253087e-05,
"loss": 1.6706,
"step": 1030
},
{
"epoch": 0.84,
"learning_rate": 1.9044934260703623e-05,
"loss": 1.566,
"step": 1035
},
{
"epoch": 0.84,
"learning_rate": 1.895205350200998e-05,
"loss": 1.508,
"step": 1040
},
{
"epoch": 0.85,
"learning_rate": 1.885900962476767e-05,
"loss": 1.6144,
"step": 1045
},
{
"epoch": 0.85,
"learning_rate": 1.8765806469304814e-05,
"loss": 1.6291,
"step": 1050
},
{
"epoch": 0.85,
"learning_rate": 1.8672447882523644e-05,
"loss": 1.6088,
"step": 1055
},
{
"epoch": 0.86,
"learning_rate": 1.8578937717741727e-05,
"loss": 1.5956,
"step": 1060
},
{
"epoch": 0.86,
"learning_rate": 1.8485279834532923e-05,
"loss": 1.6554,
"step": 1065
},
{
"epoch": 0.87,
"learning_rate": 1.839147809856807e-05,
"loss": 1.6333,
"step": 1070
},
{
"epoch": 0.87,
"learning_rate": 1.8297536381455434e-05,
"loss": 1.6016,
"step": 1075
},
{
"epoch": 0.87,
"learning_rate": 1.8203458560580934e-05,
"loss": 1.5765,
"step": 1080
},
{
"epoch": 0.88,
"learning_rate": 1.810924851894807e-05,
"loss": 1.4411,
"step": 1085
},
{
"epoch": 0.88,
"learning_rate": 1.801491014501768e-05,
"loss": 1.6064,
"step": 1090
},
{
"epoch": 0.89,
"learning_rate": 1.7920447332547423e-05,
"loss": 1.5165,
"step": 1095
},
{
"epoch": 0.89,
"learning_rate": 1.7825863980431106e-05,
"loss": 1.628,
"step": 1100
},
{
"epoch": 0.89,
"learning_rate": 1.773116399253772e-05,
"loss": 1.5792,
"step": 1105
},
{
"epoch": 0.9,
"learning_rate": 1.7636351277550324e-05,
"loss": 1.5908,
"step": 1110
},
{
"epoch": 0.9,
"learning_rate": 1.7541429748804722e-05,
"loss": 1.5477,
"step": 1115
},
{
"epoch": 0.91,
"learning_rate": 1.744640332412794e-05,
"loss": 1.6118,
"step": 1120
},
{
"epoch": 0.91,
"learning_rate": 1.7351275925676517e-05,
"loss": 1.6034,
"step": 1125
},
{
"epoch": 0.91,
"learning_rate": 1.725605147977461e-05,
"loss": 1.6053,
"step": 1130
},
{
"epoch": 0.92,
"learning_rate": 1.716073391675197e-05,
"loss": 1.5638,
"step": 1135
},
{
"epoch": 0.92,
"learning_rate": 1.7065327170781678e-05,
"loss": 1.5809,
"step": 1140
},
{
"epoch": 0.93,
"learning_rate": 1.69698351797178e-05,
"loss": 1.5822,
"step": 1145
},
{
"epoch": 0.93,
"learning_rate": 1.6874261884932842e-05,
"loss": 1.5981,
"step": 1150
},
{
"epoch": 0.93,
"learning_rate": 1.677861123115506e-05,
"loss": 1.5995,
"step": 1155
},
{
"epoch": 0.94,
"learning_rate": 1.668288716630566e-05,
"loss": 1.5323,
"step": 1160
},
{
"epoch": 0.94,
"learning_rate": 1.6587093641335844e-05,
"loss": 1.5981,
"step": 1165
},
{
"epoch": 0.95,
"learning_rate": 1.6491234610063732e-05,
"loss": 1.6448,
"step": 1170
},
{
"epoch": 0.95,
"learning_rate": 1.6395314029011184e-05,
"loss": 1.63,
"step": 1175
},
{
"epoch": 0.95,
"learning_rate": 1.6299335857240484e-05,
"loss": 1.6529,
"step": 1180
},
{
"epoch": 0.96,
"learning_rate": 1.6203304056190938e-05,
"loss": 1.6172,
"step": 1185
},
{
"epoch": 0.96,
"learning_rate": 1.6107222589515376e-05,
"loss": 1.5247,
"step": 1190
},
{
"epoch": 0.97,
"learning_rate": 1.601109542291652e-05,
"loss": 1.5932,
"step": 1195
},
{
"epoch": 0.97,
"learning_rate": 1.591492652398336e-05,
"loss": 1.6186,
"step": 1200
},
{
"epoch": 0.98,
"learning_rate": 1.5818719862027338e-05,
"loss": 1.6524,
"step": 1205
},
{
"epoch": 0.98,
"learning_rate": 1.572247940791856e-05,
"loss": 1.611,
"step": 1210
},
{
"epoch": 0.98,
"learning_rate": 1.5626209133921863e-05,
"loss": 1.5913,
"step": 1215
},
{
"epoch": 0.99,
"learning_rate": 1.5529913013532894e-05,
"loss": 1.7348,
"step": 1220
},
{
"epoch": 0.99,
"learning_rate": 1.543359502131409e-05,
"loss": 1.719,
"step": 1225
},
{
"epoch": 1.0,
"learning_rate": 1.533725913273064e-05,
"loss": 1.6274,
"step": 1230
},
{
"epoch": 1.0,
"learning_rate": 1.5240909323986384e-05,
"loss": 1.5266,
"step": 1235
},
{
"epoch": 1.0,
"eval_loss": 3.4220237731933594,
"eval_runtime": 954.9678,
"eval_samples_per_second": 9.114,
"eval_steps_per_second": 2.279,
"step": 1235
},
{
"epoch": 1.0,
"learning_rate": 1.5144549571859711e-05,
"loss": 1.3304,
"step": 1240
},
{
"epoch": 1.01,
"learning_rate": 1.5048183853539425e-05,
"loss": 1.0536,
"step": 1245
},
{
"epoch": 1.01,
"learning_rate": 1.4951816146460574e-05,
"loss": 1.0564,
"step": 1250
},
{
"epoch": 1.02,
"learning_rate": 1.4855450428140291e-05,
"loss": 1.0105,
"step": 1255
},
{
"epoch": 1.02,
"learning_rate": 1.4759090676013616e-05,
"loss": 1.0252,
"step": 1260
},
{
"epoch": 1.02,
"learning_rate": 1.4662740867269361e-05,
"loss": 1.0029,
"step": 1265
},
{
"epoch": 1.03,
"learning_rate": 1.4566404978685912e-05,
"loss": 0.9685,
"step": 1270
},
{
"epoch": 1.03,
"learning_rate": 1.447008698646711e-05,
"loss": 1.0095,
"step": 1275
},
{
"epoch": 1.04,
"learning_rate": 1.4373790866078143e-05,
"loss": 0.9759,
"step": 1280
},
{
"epoch": 1.04,
"learning_rate": 1.4277520592081442e-05,
"loss": 1.0217,
"step": 1285
},
{
"epoch": 1.04,
"learning_rate": 1.4181280137972666e-05,
"loss": 0.9472,
"step": 1290
},
{
"epoch": 1.05,
"learning_rate": 1.4085073476016642e-05,
"loss": 1.0432,
"step": 1295
},
{
"epoch": 1.05,
"learning_rate": 1.3988904577083481e-05,
"loss": 1.0355,
"step": 1300
},
{
"epoch": 1.06,
"learning_rate": 1.389277741048463e-05,
"loss": 0.9971,
"step": 1305
},
{
"epoch": 1.06,
"learning_rate": 1.3796695943809063e-05,
"loss": 0.9947,
"step": 1310
},
{
"epoch": 1.06,
"learning_rate": 1.3700664142759521e-05,
"loss": 1.0532,
"step": 1315
},
{
"epoch": 1.07,
"learning_rate": 1.3604685970988817e-05,
"loss": 0.9791,
"step": 1320
},
{
"epoch": 1.07,
"learning_rate": 1.3508765389936272e-05,
"loss": 1.0299,
"step": 1325
},
{
"epoch": 1.08,
"learning_rate": 1.341290635866416e-05,
"loss": 1.036,
"step": 1330
},
{
"epoch": 1.08,
"learning_rate": 1.3317112833694344e-05,
"loss": 1.0534,
"step": 1335
},
{
"epoch": 1.08,
"learning_rate": 1.3221388768844937e-05,
"loss": 1.0144,
"step": 1340
},
{
"epoch": 1.09,
"learning_rate": 1.3125738115067159e-05,
"loss": 1.0696,
"step": 1345
},
{
"epoch": 1.09,
"learning_rate": 1.30301648202822e-05,
"loss": 1.0651,
"step": 1350
},
{
"epoch": 1.1,
"learning_rate": 1.2934672829218327e-05,
"loss": 1.0598,
"step": 1355
},
{
"epoch": 1.1,
"learning_rate": 1.283926608324804e-05,
"loss": 1.0514,
"step": 1360
},
{
"epoch": 1.11,
"learning_rate": 1.2743948520225391e-05,
"loss": 0.9963,
"step": 1365
},
{
"epoch": 1.11,
"learning_rate": 1.2648724074323492e-05,
"loss": 1.0307,
"step": 1370
},
{
"epoch": 1.11,
"learning_rate": 1.255359667587206e-05,
"loss": 1.0354,
"step": 1375
},
{
"epoch": 1.12,
"learning_rate": 1.2458570251195279e-05,
"loss": 1.0965,
"step": 1380
},
{
"epoch": 1.12,
"learning_rate": 1.2363648722449679e-05,
"loss": 1.0484,
"step": 1385
},
{
"epoch": 1.13,
"learning_rate": 1.2268836007462284e-05,
"loss": 1.0421,
"step": 1390
},
{
"epoch": 1.13,
"learning_rate": 1.21741360195689e-05,
"loss": 1.0695,
"step": 1395
},
{
"epoch": 1.13,
"learning_rate": 1.207955266745258e-05,
"loss": 1.0323,
"step": 1400
},
{
"epoch": 1.14,
"learning_rate": 1.198508985498233e-05,
"loss": 1.0874,
"step": 1405
},
{
"epoch": 1.14,
"learning_rate": 1.1890751481051933e-05,
"loss": 1.0702,
"step": 1410
},
{
"epoch": 1.15,
"learning_rate": 1.179654143941907e-05,
"loss": 1.0211,
"step": 1415
},
{
"epoch": 1.15,
"learning_rate": 1.1702463618544562e-05,
"loss": 1.0252,
"step": 1420
},
{
"epoch": 1.15,
"learning_rate": 1.1608521901431932e-05,
"loss": 1.0588,
"step": 1425
},
{
"epoch": 1.16,
"learning_rate": 1.1514720165467076e-05,
"loss": 1.0644,
"step": 1430
},
{
"epoch": 1.16,
"learning_rate": 1.1421062282258276e-05,
"loss": 1.0831,
"step": 1435
},
{
"epoch": 1.17,
"learning_rate": 1.1327552117476363e-05,
"loss": 1.0262,
"step": 1440
},
{
"epoch": 1.17,
"learning_rate": 1.1234193530695189e-05,
"loss": 1.0775,
"step": 1445
},
{
"epoch": 1.17,
"learning_rate": 1.1140990375232336e-05,
"loss": 1.0476,
"step": 1450
},
{
"epoch": 1.18,
"learning_rate": 1.1047946497990016e-05,
"loss": 1.0373,
"step": 1455
},
{
"epoch": 1.18,
"learning_rate": 1.0955065739296378e-05,
"loss": 1.0734,
"step": 1460
}
],
"max_steps": 2470,
"num_train_epochs": 2,
"total_flos": 3.470813417302917e+17,
"trial_name": null,
"trial_params": null
}