{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.9987096774193547,
"eval_steps": 500,
"global_step": 3486,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.008602150537634409,
"grad_norm": 1.435989601453093,
"learning_rate": 5.730659025787966e-06,
"loss": 1.2534,
"step": 10
},
{
"epoch": 0.017204301075268817,
"grad_norm": 0.956311142058486,
"learning_rate": 1.1461318051575932e-05,
"loss": 1.1941,
"step": 20
},
{
"epoch": 0.025806451612903226,
"grad_norm": 0.5955272986463238,
"learning_rate": 1.7191977077363898e-05,
"loss": 1.0416,
"step": 30
},
{
"epoch": 0.034408602150537634,
"grad_norm": 0.5371220864077184,
"learning_rate": 2.2922636103151864e-05,
"loss": 0.9252,
"step": 40
},
{
"epoch": 0.043010752688172046,
"grad_norm": 0.3298242735866968,
"learning_rate": 2.8653295128939826e-05,
"loss": 0.8582,
"step": 50
},
{
"epoch": 0.05161290322580645,
"grad_norm": 0.3017426075394883,
"learning_rate": 3.4383954154727795e-05,
"loss": 0.7998,
"step": 60
},
{
"epoch": 0.060215053763440864,
"grad_norm": 0.3375526814194791,
"learning_rate": 4.011461318051576e-05,
"loss": 0.7638,
"step": 70
},
{
"epoch": 0.06881720430107527,
"grad_norm": 0.2705319023822148,
"learning_rate": 4.584527220630373e-05,
"loss": 0.731,
"step": 80
},
{
"epoch": 0.07741935483870968,
"grad_norm": 0.2857694097360681,
"learning_rate": 5.157593123209169e-05,
"loss": 0.7083,
"step": 90
},
{
"epoch": 0.08602150537634409,
"grad_norm": 0.3098635398347818,
"learning_rate": 5.730659025787965e-05,
"loss": 0.6841,
"step": 100
},
{
"epoch": 0.09462365591397849,
"grad_norm": 0.3198686339975185,
"learning_rate": 6.303724928366761e-05,
"loss": 0.6606,
"step": 110
},
{
"epoch": 0.1032258064516129,
"grad_norm": 0.3190855908911574,
"learning_rate": 6.876790830945559e-05,
"loss": 0.6557,
"step": 120
},
{
"epoch": 0.11182795698924732,
"grad_norm": 0.3381628025718819,
"learning_rate": 7.449856733524355e-05,
"loss": 0.649,
"step": 130
},
{
"epoch": 0.12043010752688173,
"grad_norm": 0.31598591910637613,
"learning_rate": 8.022922636103152e-05,
"loss": 0.6497,
"step": 140
},
{
"epoch": 0.12903225806451613,
"grad_norm": 0.3128921107543011,
"learning_rate": 8.595988538681948e-05,
"loss": 0.6392,
"step": 150
},
{
"epoch": 0.13763440860215054,
"grad_norm": 0.32601075629670023,
"learning_rate": 9.169054441260745e-05,
"loss": 0.6329,
"step": 160
},
{
"epoch": 0.14623655913978495,
"grad_norm": 0.3810140632755748,
"learning_rate": 9.742120343839543e-05,
"loss": 0.6246,
"step": 170
},
{
"epoch": 0.15483870967741936,
"grad_norm": 0.32108686980228696,
"learning_rate": 0.00010315186246418338,
"loss": 0.6341,
"step": 180
},
{
"epoch": 0.16344086021505377,
"grad_norm": 0.35140482524130334,
"learning_rate": 0.00010888252148997136,
"loss": 0.6341,
"step": 190
},
{
"epoch": 0.17204301075268819,
"grad_norm": 0.29928696988494263,
"learning_rate": 0.0001146131805157593,
"loss": 0.6256,
"step": 200
},
{
"epoch": 0.18064516129032257,
"grad_norm": 0.29203351329015653,
"learning_rate": 0.0001203438395415473,
"loss": 0.6206,
"step": 210
},
{
"epoch": 0.18924731182795698,
"grad_norm": 0.269908780529402,
"learning_rate": 0.00012607449856733523,
"loss": 0.6201,
"step": 220
},
{
"epoch": 0.1978494623655914,
"grad_norm": 0.2918951896599236,
"learning_rate": 0.0001318051575931232,
"loss": 0.6181,
"step": 230
},
{
"epoch": 0.2064516129032258,
"grad_norm": 0.2708453980927917,
"learning_rate": 0.00013753581661891118,
"loss": 0.6053,
"step": 240
},
{
"epoch": 0.21505376344086022,
"grad_norm": 0.2662594416135698,
"learning_rate": 0.00014326647564469916,
"loss": 0.6033,
"step": 250
},
{
"epoch": 0.22365591397849463,
"grad_norm": 0.2776129880103733,
"learning_rate": 0.0001489971346704871,
"loss": 0.618,
"step": 260
},
{
"epoch": 0.23225806451612904,
"grad_norm": 0.279068656400945,
"learning_rate": 0.00015472779369627508,
"loss": 0.6111,
"step": 270
},
{
"epoch": 0.24086021505376345,
"grad_norm": 0.2524990953581738,
"learning_rate": 0.00016045845272206303,
"loss": 0.5888,
"step": 280
},
{
"epoch": 0.24946236559139784,
"grad_norm": 0.27628464942088166,
"learning_rate": 0.000166189111747851,
"loss": 0.6051,
"step": 290
},
{
"epoch": 0.25806451612903225,
"grad_norm": 0.2738432329342113,
"learning_rate": 0.00017191977077363896,
"loss": 0.6025,
"step": 300
},
{
"epoch": 0.26666666666666666,
"grad_norm": 0.25607005690614404,
"learning_rate": 0.00017765042979942693,
"loss": 0.6035,
"step": 310
},
{
"epoch": 0.2752688172043011,
"grad_norm": 0.247801611216806,
"learning_rate": 0.0001833810888252149,
"loss": 0.6118,
"step": 320
},
{
"epoch": 0.2838709677419355,
"grad_norm": 0.24971595808191485,
"learning_rate": 0.00018911174785100288,
"loss": 0.6003,
"step": 330
},
{
"epoch": 0.2924731182795699,
"grad_norm": 0.2411785853482113,
"learning_rate": 0.00019484240687679086,
"loss": 0.6019,
"step": 340
},
{
"epoch": 0.3010752688172043,
"grad_norm": 0.24693597842223552,
"learning_rate": 0.0001999999498534943,
"loss": 0.5912,
"step": 350
},
{
"epoch": 0.3096774193548387,
"grad_norm": 0.23649676383007667,
"learning_rate": 0.0001999939323336644,
"loss": 0.6032,
"step": 360
},
{
"epoch": 0.31827956989247314,
"grad_norm": 0.23340962816367694,
"learning_rate": 0.0001999778862042167,
"loss": 0.5867,
"step": 370
},
{
"epoch": 0.32688172043010755,
"grad_norm": 0.228292747503904,
"learning_rate": 0.0001999518130744525,
"loss": 0.5869,
"step": 380
},
{
"epoch": 0.33548387096774196,
"grad_norm": 0.23546242335669995,
"learning_rate": 0.0001999157155593029,
"loss": 0.585,
"step": 390
},
{
"epoch": 0.34408602150537637,
"grad_norm": 0.23135215802858206,
"learning_rate": 0.0001998695972790664,
"loss": 0.5929,
"step": 400
},
{
"epoch": 0.35268817204301073,
"grad_norm": 0.22524299635343834,
"learning_rate": 0.00019981346285904595,
"loss": 0.5976,
"step": 410
},
{
"epoch": 0.36129032258064514,
"grad_norm": 0.21756192058074578,
"learning_rate": 0.00019974731792908506,
"loss": 0.587,
"step": 420
},
{
"epoch": 0.36989247311827955,
"grad_norm": 0.23763473038875538,
"learning_rate": 0.000199671169123003,
"loss": 0.5934,
"step": 430
},
{
"epoch": 0.37849462365591396,
"grad_norm": 0.23701431523985933,
"learning_rate": 0.00019958502407792963,
"loss": 0.6009,
"step": 440
},
{
"epoch": 0.3870967741935484,
"grad_norm": 0.22134697304917678,
"learning_rate": 0.00019948889143353948,
"loss": 0.5994,
"step": 450
},
{
"epoch": 0.3956989247311828,
"grad_norm": 0.2268467235014266,
"learning_rate": 0.00019938278083118517,
"loss": 0.5982,
"step": 460
},
{
"epoch": 0.4043010752688172,
"grad_norm": 0.2247581060586749,
"learning_rate": 0.00019926670291293055,
"loss": 0.5782,
"step": 470
},
{
"epoch": 0.4129032258064516,
"grad_norm": 0.21243655986167898,
"learning_rate": 0.00019914066932048317,
"loss": 0.5874,
"step": 480
},
{
"epoch": 0.421505376344086,
"grad_norm": 0.21385338375703783,
"learning_rate": 0.00019900469269402708,
"loss": 0.5813,
"step": 490
},
{
"epoch": 0.43010752688172044,
"grad_norm": 0.21415772228830113,
"learning_rate": 0.00019885878667095472,
"loss": 0.5934,
"step": 500
},
{
"epoch": 0.43870967741935485,
"grad_norm": 0.23153816801674154,
"learning_rate": 0.00019870296588449948,
"loss": 0.5766,
"step": 510
},
{
"epoch": 0.44731182795698926,
"grad_norm": 0.21768913976009108,
"learning_rate": 0.00019853724596226795,
"loss": 0.5964,
"step": 520
},
{
"epoch": 0.4559139784946237,
"grad_norm": 0.21481929780656356,
"learning_rate": 0.00019836164352467263,
"loss": 0.5748,
"step": 530
},
{
"epoch": 0.4645161290322581,
"grad_norm": 0.29328457757364856,
"learning_rate": 0.00019817617618326516,
"loss": 0.5863,
"step": 540
},
{
"epoch": 0.4731182795698925,
"grad_norm": 0.2140715456768234,
"learning_rate": 0.00019798086253896974,
"loss": 0.5787,
"step": 550
},
{
"epoch": 0.4817204301075269,
"grad_norm": 0.22386132897333755,
"learning_rate": 0.00019777572218021789,
"loss": 0.5863,
"step": 560
},
{
"epoch": 0.49032258064516127,
"grad_norm": 0.23289220288123677,
"learning_rate": 0.00019756077568098372,
"loss": 0.5638,
"step": 570
},
{
"epoch": 0.4989247311827957,
"grad_norm": 0.21115266625450407,
"learning_rate": 0.00019733604459872055,
"loss": 0.5744,
"step": 580
},
{
"epoch": 0.5075268817204301,
"grad_norm": 0.2207246383657776,
"learning_rate": 0.00019710155147219886,
"loss": 0.5789,
"step": 590
},
{
"epoch": 0.5161290322580645,
"grad_norm": 0.2127412426650313,
"learning_rate": 0.00019685731981924592,
"loss": 0.5764,
"step": 600
},
{
"epoch": 0.524731182795699,
"grad_norm": 0.2256297779595538,
"learning_rate": 0.00019660337413438697,
"loss": 0.5798,
"step": 610
},
{
"epoch": 0.5333333333333333,
"grad_norm": 0.22243638154383857,
"learning_rate": 0.00019633973988638877,
"loss": 0.5767,
"step": 620
},
{
"epoch": 0.5419354838709678,
"grad_norm": 0.21425852582870433,
"learning_rate": 0.00019606644351570518,
"loss": 0.5736,
"step": 630
},
{
"epoch": 0.5505376344086022,
"grad_norm": 0.20506928092936336,
"learning_rate": 0.00019578351243182545,
"loss": 0.5766,
"step": 640
},
{
"epoch": 0.5591397849462365,
"grad_norm": 0.2175808639953357,
"learning_rate": 0.0001954909750105252,
"loss": 0.5883,
"step": 650
},
{
"epoch": 0.567741935483871,
"grad_norm": 0.21972174494018643,
"learning_rate": 0.00019518886059102062,
"loss": 0.5774,
"step": 660
},
{
"epoch": 0.5763440860215053,
"grad_norm": 0.2222552681428519,
"learning_rate": 0.0001948771994730259,
"loss": 0.5694,
"step": 670
},
{
"epoch": 0.5849462365591398,
"grad_norm": 0.21552094411185982,
"learning_rate": 0.0001945560229137145,
"loss": 0.5657,
"step": 680
},
{
"epoch": 0.5935483870967742,
"grad_norm": 0.21243053577901072,
"learning_rate": 0.0001942253631245842,
"loss": 0.5776,
"step": 690
},
{
"epoch": 0.6021505376344086,
"grad_norm": 0.21257424919590467,
"learning_rate": 0.00019388525326822665,
"loss": 0.5764,
"step": 700
},
{
"epoch": 0.610752688172043,
"grad_norm": 0.21485742957719725,
"learning_rate": 0.0001935357274550013,
"loss": 0.5821,
"step": 710
},
{
"epoch": 0.6193548387096774,
"grad_norm": 0.2184569847068338,
"learning_rate": 0.00019317682073961444,
"loss": 0.5674,
"step": 720
},
{
"epoch": 0.6279569892473118,
"grad_norm": 0.20912150886711225,
"learning_rate": 0.00019280856911760354,
"loss": 0.559,
"step": 730
},
{
"epoch": 0.6365591397849463,
"grad_norm": 0.21433879312995557,
"learning_rate": 0.00019243100952172723,
"loss": 0.5615,
"step": 740
},
{
"epoch": 0.6451612903225806,
"grad_norm": 0.22214855400737676,
"learning_rate": 0.00019204417981826091,
"loss": 0.5761,
"step": 750
},
{
"epoch": 0.6537634408602151,
"grad_norm": 0.2196194509755129,
"learning_rate": 0.00019164811880319958,
"loss": 0.5718,
"step": 760
},
{
"epoch": 0.6623655913978495,
"grad_norm": 0.2089872234950888,
"learning_rate": 0.00019124286619836637,
"loss": 0.5819,
"step": 770
},
{
"epoch": 0.6709677419354839,
"grad_norm": 0.20958335214398927,
"learning_rate": 0.00019082846264742917,
"loss": 0.5663,
"step": 780
},
{
"epoch": 0.6795698924731183,
"grad_norm": 0.22250020784958888,
"learning_rate": 0.00019040494971182413,
"loss": 0.572,
"step": 790
},
{
"epoch": 0.6881720430107527,
"grad_norm": 0.2162793985278103,
"learning_rate": 0.00018997236986658753,
"loss": 0.5578,
"step": 800
},
{
"epoch": 0.6967741935483871,
"grad_norm": 0.21650281403806723,
"learning_rate": 0.00018953076649609564,
"loss": 0.5698,
"step": 810
},
{
"epoch": 0.7053763440860215,
"grad_norm": 0.21882852535131322,
"learning_rate": 0.000189080183889714,
"loss": 0.5712,
"step": 820
},
{
"epoch": 0.7139784946236559,
"grad_norm": 0.21577728486568806,
"learning_rate": 0.00018862066723735512,
"loss": 0.5676,
"step": 830
},
{
"epoch": 0.7225806451612903,
"grad_norm": 0.20731869739137426,
"learning_rate": 0.00018815226262494655,
"loss": 0.567,
"step": 840
},
{
"epoch": 0.7311827956989247,
"grad_norm": 0.21966746890763253,
"learning_rate": 0.00018767501702980874,
"loss": 0.5738,
"step": 850
},
{
"epoch": 0.7397849462365591,
"grad_norm": 0.2199440815107056,
"learning_rate": 0.00018718897831594355,
"loss": 0.5691,
"step": 860
},
{
"epoch": 0.7483870967741936,
"grad_norm": 0.21722437739899045,
"learning_rate": 0.00018669419522923393,
"loss": 0.5753,
"step": 870
},
{
"epoch": 0.7569892473118279,
"grad_norm": 0.2218916153172574,
"learning_rate": 0.00018619071739255506,
"loss": 0.5572,
"step": 880
},
{
"epoch": 0.7655913978494624,
"grad_norm": 0.20612913226427296,
"learning_rate": 0.00018567859530079753,
"loss": 0.567,
"step": 890
},
{
"epoch": 0.7741935483870968,
"grad_norm": 0.2296891108499081,
"learning_rate": 0.00018515788031580317,
"loss": 0.5735,
"step": 900
},
{
"epoch": 0.7827956989247312,
"grad_norm": 0.20791733581224098,
"learning_rate": 0.0001846286246612138,
"loss": 0.56,
"step": 910
},
{
"epoch": 0.7913978494623656,
"grad_norm": 0.2130914822501822,
"learning_rate": 0.00018409088141723364,
"loss": 0.5588,
"step": 920
},
{
"epoch": 0.8,
"grad_norm": 0.21931892060132505,
"learning_rate": 0.00018354470451530574,
"loss": 0.5718,
"step": 930
},
{
"epoch": 0.8086021505376344,
"grad_norm": 0.20510223609898656,
"learning_rate": 0.00018299014873270314,
"loss": 0.5693,
"step": 940
},
{
"epoch": 0.8172043010752689,
"grad_norm": 0.20754250893604087,
"learning_rate": 0.00018242726968703505,
"loss": 0.5657,
"step": 950
},
{
"epoch": 0.8258064516129032,
"grad_norm": 0.2119047697743998,
"learning_rate": 0.00018185612383066893,
"loss": 0.5523,
"step": 960
},
{
"epoch": 0.8344086021505376,
"grad_norm": 0.21665086623022414,
"learning_rate": 0.00018127676844506874,
"loss": 0.5564,
"step": 970
},
{
"epoch": 0.843010752688172,
"grad_norm": 0.22240006490076028,
"learning_rate": 0.00018068926163505,
"loss": 0.5661,
"step": 980
},
{
"epoch": 0.8516129032258064,
"grad_norm": 0.20224509449303205,
"learning_rate": 0.00018009366232295235,
"loss": 0.5709,
"step": 990
},
{
"epoch": 0.8602150537634409,
"grad_norm": 0.20943438346218987,
"learning_rate": 0.00017949003024273015,
"loss": 0.5567,
"step": 1000
},
{
"epoch": 0.8688172043010752,
"grad_norm": 0.2119577392060311,
"learning_rate": 0.0001788784259339616,
"loss": 0.5726,
"step": 1010
},
{
"epoch": 0.8774193548387097,
"grad_norm": 0.21671277914581005,
"learning_rate": 0.0001782589107357771,
"loss": 0.5721,
"step": 1020
},
{
"epoch": 0.886021505376344,
"grad_norm": 0.21042958546072746,
"learning_rate": 0.00017763154678070733,
"loss": 0.5682,
"step": 1030
},
{
"epoch": 0.8946236559139785,
"grad_norm": 0.21134103487495276,
"learning_rate": 0.0001769963969884521,
"loss": 0.5578,
"step": 1040
},
{
"epoch": 0.9032258064516129,
"grad_norm": 0.2138330925142033,
"learning_rate": 0.0001763535250595696,
"loss": 0.5622,
"step": 1050
},
{
"epoch": 0.9118279569892473,
"grad_norm": 0.20191022832768318,
"learning_rate": 0.00017570299546908812,
"loss": 0.5543,
"step": 1060
},
{
"epoch": 0.9204301075268817,
"grad_norm": 0.20993539174391937,
"learning_rate": 0.0001750448734600394,
"loss": 0.5687,
"step": 1070
},
{
"epoch": 0.9290322580645162,
"grad_norm": 0.21448241595075324,
"learning_rate": 0.0001743792250369155,
"loss": 0.5628,
"step": 1080
},
{
"epoch": 0.9376344086021505,
"grad_norm": 0.21562115673441815,
"learning_rate": 0.00017370611695904895,
"loss": 0.5707,
"step": 1090
},
{
"epoch": 0.946236559139785,
"grad_norm": 0.20800257947078069,
"learning_rate": 0.00017302561673391732,
"loss": 0.5682,
"step": 1100
},
{
"epoch": 0.9548387096774194,
"grad_norm": 0.2142950321860787,
"learning_rate": 0.00017233779261037268,
"loss": 0.5644,
"step": 1110
},
{
"epoch": 0.9634408602150538,
"grad_norm": 0.22988301024268806,
"learning_rate": 0.00017164271357179698,
"loss": 0.5747,
"step": 1120
},
{
"epoch": 0.9720430107526882,
"grad_norm": 0.2044546345524181,
"learning_rate": 0.00017094044932918336,
"loss": 0.5627,
"step": 1130
},
{
"epoch": 0.9806451612903225,
"grad_norm": 0.20801733651601362,
"learning_rate": 0.00017023107031414477,
"loss": 0.5655,
"step": 1140
},
{
"epoch": 0.989247311827957,
"grad_norm": 0.21348629669321226,
"learning_rate": 0.00016951464767185013,
"loss": 0.5569,
"step": 1150
},
{
"epoch": 0.9978494623655914,
"grad_norm": 0.21098078403488563,
"learning_rate": 0.0001687912532538892,
"loss": 0.5686,
"step": 1160
},
{
"epoch": 1.0064516129032257,
"grad_norm": 0.23232328451409845,
"learning_rate": 0.00016806095961106632,
"loss": 0.5352,
"step": 1170
},
{
"epoch": 1.0150537634408603,
"grad_norm": 0.21849412389861325,
"learning_rate": 0.00016732383998612407,
"loss": 0.5166,
"step": 1180
},
{
"epoch": 1.0236559139784946,
"grad_norm": 0.22255257432573036,
"learning_rate": 0.00016657996830639774,
"loss": 0.5286,
"step": 1190
},
{
"epoch": 1.032258064516129,
"grad_norm": 0.23162311614487038,
"learning_rate": 0.0001658294191764009,
"loss": 0.5209,
"step": 1200
},
{
"epoch": 1.0408602150537634,
"grad_norm": 0.23679830183057837,
"learning_rate": 0.0001650722678703432,
"loss": 0.5155,
"step": 1210
},
{
"epoch": 1.049462365591398,
"grad_norm": 0.24431579126710756,
"learning_rate": 0.00016430859032458086,
"loss": 0.5239,
"step": 1220
},
{
"epoch": 1.0580645161290323,
"grad_norm": 0.21925666538005387,
"learning_rate": 0.00016353846313000098,
"loss": 0.5142,
"step": 1230
},
{
"epoch": 1.0666666666666667,
"grad_norm": 0.23676074055821078,
"learning_rate": 0.00016276196352434,
"loss": 0.5267,
"step": 1240
},
{
"epoch": 1.075268817204301,
"grad_norm": 0.23532815378000083,
"learning_rate": 0.00016197916938443733,
"loss": 0.5211,
"step": 1250
},
{
"epoch": 1.0838709677419356,
"grad_norm": 0.23216395268938977,
"learning_rate": 0.00016119015921842503,
"loss": 0.5161,
"step": 1260
},
{
"epoch": 1.09247311827957,
"grad_norm": 0.23822646088799643,
"learning_rate": 0.0001603950121578539,
"loss": 0.5077,
"step": 1270
},
{
"epoch": 1.1010752688172043,
"grad_norm": 0.2290007557059369,
"learning_rate": 0.00015959380794975734,
"loss": 0.5156,
"step": 1280
},
{
"epoch": 1.1096774193548387,
"grad_norm": 0.22484506558131512,
"learning_rate": 0.00015878662694865327,
"loss": 0.5205,
"step": 1290
},
{
"epoch": 1.118279569892473,
"grad_norm": 0.22971689352325197,
"learning_rate": 0.00015797355010848519,
"loss": 0.5218,
"step": 1300
},
{
"epoch": 1.1268817204301076,
"grad_norm": 0.23491219834804952,
"learning_rate": 0.00015715465897450317,
"loss": 0.5248,
"step": 1310
},
{
"epoch": 1.135483870967742,
"grad_norm": 0.23555948179310204,
"learning_rate": 0.00015633003567508557,
"loss": 0.524,
"step": 1320
},
{
"epoch": 1.1440860215053763,
"grad_norm": 0.2275140567256555,
"learning_rate": 0.00015549976291350204,
"loss": 0.517,
"step": 1330
},
{
"epoch": 1.1526881720430107,
"grad_norm": 0.22712302146976637,
"learning_rate": 0.00015466392395961908,
"loss": 0.5243,
"step": 1340
},
{
"epoch": 1.1612903225806452,
"grad_norm": 0.22097671231695554,
"learning_rate": 0.0001538226026415489,
"loss": 0.5118,
"step": 1350
},
{
"epoch": 1.1698924731182796,
"grad_norm": 0.22706370204760404,
"learning_rate": 0.00015297588333724183,
"loss": 0.5235,
"step": 1360
},
{
"epoch": 1.178494623655914,
"grad_norm": 0.2309770159886724,
"learning_rate": 0.00015212385096602415,
"loss": 0.5202,
"step": 1370
},
{
"epoch": 1.1870967741935483,
"grad_norm": 0.23885726185135214,
"learning_rate": 0.0001512665909800811,
"loss": 0.5201,
"step": 1380
},
{
"epoch": 1.1956989247311829,
"grad_norm": 0.23447118491147034,
"learning_rate": 0.00015040418935588682,
"loss": 0.5103,
"step": 1390
},
{
"epoch": 1.2043010752688172,
"grad_norm": 0.24148993682260503,
"learning_rate": 0.0001495367325855816,
"loss": 0.5229,
"step": 1400
},
{
"epoch": 1.2129032258064516,
"grad_norm": 0.22483070355440393,
"learning_rate": 0.00014866430766829743,
"loss": 0.5132,
"step": 1410
},
{
"epoch": 1.221505376344086,
"grad_norm": 0.2254845362019036,
"learning_rate": 0.00014778700210143242,
"loss": 0.5231,
"step": 1420
},
{
"epoch": 1.2301075268817203,
"grad_norm": 0.23952864821588327,
"learning_rate": 0.00014690490387187584,
"loss": 0.5157,
"step": 1430
},
{
"epoch": 1.238709677419355,
"grad_norm": 0.2257464446574845,
"learning_rate": 0.00014601810144718345,
"loss": 0.537,
"step": 1440
},
{
"epoch": 1.2473118279569892,
"grad_norm": 0.23617654028249638,
"learning_rate": 0.0001451266837667051,
"loss": 0.5187,
"step": 1450
},
{
"epoch": 1.2559139784946236,
"grad_norm": 0.23682708876074962,
"learning_rate": 0.00014423074023266463,
"loss": 0.52,
"step": 1460
},
{
"epoch": 1.2645161290322582,
"grad_norm": 0.22896552654447108,
"learning_rate": 0.00014333036070119363,
"loss": 0.5163,
"step": 1470
},
{
"epoch": 1.2731182795698925,
"grad_norm": 0.24027250250612717,
"learning_rate": 0.00014242563547331956,
"loss": 0.517,
"step": 1480
},
{
"epoch": 1.281720430107527,
"grad_norm": 0.23395780934976781,
"learning_rate": 0.00014151665528590925,
"loss": 0.5215,
"step": 1490
},
{
"epoch": 1.2903225806451613,
"grad_norm": 0.24728067879450083,
"learning_rate": 0.0001406035113025687,
"loss": 0.5208,
"step": 1500
},
{
"epoch": 1.2989247311827956,
"grad_norm": 0.22394688366493132,
"learning_rate": 0.0001396862951045001,
"loss": 0.5274,
"step": 1510
},
{
"epoch": 1.3075268817204302,
"grad_norm": 0.23212309124084482,
"learning_rate": 0.00013876509868131692,
"loss": 0.5172,
"step": 1520
},
{
"epoch": 1.3161290322580645,
"grad_norm": 0.23245525514987117,
"learning_rate": 0.0001378400144218181,
"loss": 0.521,
"step": 1530
},
{
"epoch": 1.324731182795699,
"grad_norm": 0.24340497246468923,
"learning_rate": 0.00013691113510472212,
"loss": 0.533,
"step": 1540
},
{
"epoch": 1.3333333333333333,
"grad_norm": 0.2321928544760521,
"learning_rate": 0.00013597855388936217,
"loss": 0.5283,
"step": 1550
},
{
"epoch": 1.3419354838709676,
"grad_norm": 0.23326523784187517,
"learning_rate": 0.00013504236430634286,
"loss": 0.5416,
"step": 1560
},
{
"epoch": 1.3505376344086022,
"grad_norm": 0.2262845978211386,
"learning_rate": 0.0001341026602481597,
"loss": 0.5231,
"step": 1570
},
{
"epoch": 1.3591397849462366,
"grad_norm": 0.2371091538942524,
"learning_rate": 0.00013315953595978287,
"loss": 0.5328,
"step": 1580
},
{
"epoch": 1.367741935483871,
"grad_norm": 0.2404491874213034,
"learning_rate": 0.00013221308602920468,
"loss": 0.5273,
"step": 1590
},
{
"epoch": 1.3763440860215055,
"grad_norm": 0.2382492165329928,
"learning_rate": 0.00013126340537795343,
"loss": 0.529,
"step": 1600
},
{
"epoch": 1.3849462365591398,
"grad_norm": 0.22921228109841923,
"learning_rate": 0.0001303105892515734,
"loss": 0.5198,
"step": 1610
},
{
"epoch": 1.3935483870967742,
"grad_norm": 0.22823020260902263,
"learning_rate": 0.00012935473321007255,
"loss": 0.5215,
"step": 1620
},
{
"epoch": 1.4021505376344086,
"grad_norm": 0.23341543795313835,
"learning_rate": 0.0001283959331183386,
"loss": 0.5179,
"step": 1630
},
{
"epoch": 1.410752688172043,
"grad_norm": 0.23483174318348113,
"learning_rate": 0.00012743428513652442,
"loss": 0.5242,
"step": 1640
},
{
"epoch": 1.4193548387096775,
"grad_norm": 0.23015296236307278,
"learning_rate": 0.00012646988571040398,
"loss": 0.522,
"step": 1650
},
{
"epoch": 1.4279569892473118,
"grad_norm": 0.24213814226406005,
"learning_rate": 0.00012550283156169965,
"loss": 0.5238,
"step": 1660
},
{
"epoch": 1.4365591397849462,
"grad_norm": 0.23239320809561437,
"learning_rate": 0.0001245332196783817,
"loss": 0.5156,
"step": 1670
},
{
"epoch": 1.4451612903225808,
"grad_norm": 0.23180542434985962,
"learning_rate": 0.00012356114730494123,
"loss": 0.5184,
"step": 1680
},
{
"epoch": 1.453763440860215,
"grad_norm": 0.23588148083202282,
"learning_rate": 0.00012258671193263716,
"loss": 0.5155,
"step": 1690
},
{
"epoch": 1.4623655913978495,
"grad_norm": 0.23435001337205247,
"learning_rate": 0.00012161001128971879,
"loss": 0.5293,
"step": 1700
},
{
"epoch": 1.4709677419354839,
"grad_norm": 0.22453229677185074,
"learning_rate": 0.00012063114333162438,
"loss": 0.5151,
"step": 1710
},
{
"epoch": 1.4795698924731182,
"grad_norm": 0.2513082837299489,
"learning_rate": 0.00011965020623115688,
"loss": 0.5416,
"step": 1720
},
{
"epoch": 1.4881720430107528,
"grad_norm": 0.23485488435380322,
"learning_rate": 0.0001186672983686381,
"loss": 0.5114,
"step": 1730
},
{
"epoch": 1.4967741935483871,
"grad_norm": 0.2339236059203922,
"learning_rate": 0.00011768251832204187,
"loss": 0.5189,
"step": 1740
},
{
"epoch": 1.5053763440860215,
"grad_norm": 0.2363738308894046,
"learning_rate": 0.00011669596485710741,
"loss": 0.5236,
"step": 1750
},
{
"epoch": 1.513978494623656,
"grad_norm": 0.23458902400188686,
"learning_rate": 0.00011570773691743397,
"loss": 0.5284,
"step": 1760
},
{
"epoch": 1.5225806451612902,
"grad_norm": 0.23836545061013703,
"learning_rate": 0.0001147179336145575,
"loss": 0.5121,
"step": 1770
},
{
"epoch": 1.5311827956989248,
"grad_norm": 0.23790373487214236,
"learning_rate": 0.00011372665421801056,
"loss": 0.5165,
"step": 1780
},
{
"epoch": 1.5397849462365591,
"grad_norm": 0.23929874324218836,
"learning_rate": 0.0001127339981453664,
"loss": 0.5257,
"step": 1790
},
{
"epoch": 1.5483870967741935,
"grad_norm": 0.23994815774490452,
"learning_rate": 0.00011174006495226812,
"loss": 0.5125,
"step": 1800
},
{
"epoch": 1.556989247311828,
"grad_norm": 0.22986043036890028,
"learning_rate": 0.00011074495432244397,
"loss": 0.5142,
"step": 1810
},
{
"epoch": 1.5655913978494622,
"grad_norm": 0.2276160657387438,
"learning_rate": 0.0001097487660577099,
"loss": 0.5198,
"step": 1820
},
{
"epoch": 1.5741935483870968,
"grad_norm": 0.23590750768735258,
"learning_rate": 0.00010875160006796024,
"loss": 0.5203,
"step": 1830
},
{
"epoch": 1.5827956989247312,
"grad_norm": 0.24238494113010198,
"learning_rate": 0.00010775355636114755,
"loss": 0.5288,
"step": 1840
},
{
"epoch": 1.5913978494623655,
"grad_norm": 0.23079601762605023,
"learning_rate": 0.00010675473503325245,
"loss": 0.516,
"step": 1850
},
{
"epoch": 1.6,
"grad_norm": 0.23473655518093983,
"learning_rate": 0.00010575523625824488,
"loss": 0.5245,
"step": 1860
},
{
"epoch": 1.6086021505376344,
"grad_norm": 0.2297499982621768,
"learning_rate": 0.00010475516027803751,
"loss": 0.5162,
"step": 1870
},
{
"epoch": 1.6172043010752688,
"grad_norm": 0.23292165198339548,
"learning_rate": 0.00010375460739243215,
"loss": 0.5246,
"step": 1880
},
{
"epoch": 1.6258064516129034,
"grad_norm": 0.24340787369255004,
"learning_rate": 0.00010275367794906044,
"loss": 0.5099,
"step": 1890
},
{
"epoch": 1.6344086021505375,
"grad_norm": 0.24374482359627925,
"learning_rate": 0.00010175247233331989,
"loss": 0.5097,
"step": 1900
},
{
"epoch": 1.643010752688172,
"grad_norm": 0.23616305839556445,
"learning_rate": 0.00010075109095830584,
"loss": 0.5231,
"step": 1910
},
{
"epoch": 1.6516129032258065,
"grad_norm": 0.23231972508384582,
"learning_rate": 9.974963425474106e-05,
"loss": 0.5213,
"step": 1920
},
{
"epoch": 1.6602150537634408,
"grad_norm": 0.24023523508198666,
"learning_rate": 9.874820266090303e-05,
"loss": 0.5145,
"step": 1930
},
{
"epoch": 1.6688172043010754,
"grad_norm": 0.24186555744758984,
"learning_rate": 9.774689661255106e-05,
"loss": 0.5256,
"step": 1940
},
{
"epoch": 1.6774193548387095,
"grad_norm": 0.2372332084270041,
"learning_rate": 9.67458165328531e-05,
"loss": 0.5094,
"step": 1950
},
{
"epoch": 1.686021505376344,
"grad_norm": 0.2350639603238352,
"learning_rate": 9.574506282231433e-05,
"loss": 0.5119,
"step": 1960
},
{
"epoch": 1.6946236559139785,
"grad_norm": 0.23559422057113213,
"learning_rate": 9.474473584870757e-05,
"loss": 0.5155,
"step": 1970
},
{
"epoch": 1.7032258064516128,
"grad_norm": 0.2336022634685686,
"learning_rate": 9.374493593700723e-05,
"loss": 0.5113,
"step": 1980
},
{
"epoch": 1.7118279569892474,
"grad_norm": 0.23539681381001779,
"learning_rate": 9.274576335932767e-05,
"loss": 0.5189,
"step": 1990
},
{
"epoch": 1.7204301075268817,
"grad_norm": 0.23650024951926094,
"learning_rate": 9.174731832486648e-05,
"loss": 0.5082,
"step": 2000
},
{
"epoch": 1.729032258064516,
"grad_norm": 0.23916533165546086,
"learning_rate": 9.074970096985427e-05,
"loss": 0.5284,
"step": 2010
},
{
"epoch": 1.7376344086021507,
"grad_norm": 0.23608187656307292,
"learning_rate": 8.975301134751202e-05,
"loss": 0.5172,
"step": 2020
},
{
"epoch": 1.7462365591397848,
"grad_norm": 0.23344322072994478,
"learning_rate": 8.87573494180163e-05,
"loss": 0.511,
"step": 2030
},
{
"epoch": 1.7548387096774194,
"grad_norm": 0.24030220361622895,
"learning_rate": 8.77628150384741e-05,
"loss": 0.5254,
"step": 2040
},
{
"epoch": 1.7634408602150538,
"grad_norm": 0.23404925065530963,
"learning_rate": 8.676950795290802e-05,
"loss": 0.5189,
"step": 2050
},
{
"epoch": 1.772043010752688,
"grad_norm": 0.23053178874827976,
"learning_rate": 8.57775277822526e-05,
"loss": 0.5252,
"step": 2060
},
{
"epoch": 1.7806451612903227,
"grad_norm": 0.23781516582822665,
"learning_rate": 8.478697401436323e-05,
"loss": 0.5146,
"step": 2070
},
{
"epoch": 1.789247311827957,
"grad_norm": 0.2276955393985772,
"learning_rate": 8.379794599403836e-05,
"loss": 0.5169,
"step": 2080
},
{
"epoch": 1.7978494623655914,
"grad_norm": 0.23946966468353473,
"learning_rate": 8.281054291305566e-05,
"loss": 0.5181,
"step": 2090
},
{
"epoch": 1.8064516129032258,
"grad_norm": 0.22844184267941248,
"learning_rate": 8.182486380022426e-05,
"loss": 0.5009,
"step": 2100
},
{
"epoch": 1.8150537634408601,
"grad_norm": 0.2386146449957716,
"learning_rate": 8.084100751145277e-05,
"loss": 0.5167,
"step": 2110
},
{
"epoch": 1.8236559139784947,
"grad_norm": 0.22996861067382443,
"learning_rate": 7.985907271983467e-05,
"loss": 0.5172,
"step": 2120
},
{
"epoch": 1.832258064516129,
"grad_norm": 0.23122069068423765,
"learning_rate": 7.887915790575241e-05,
"loss": 0.512,
"step": 2130
},
{
"epoch": 1.8408602150537634,
"grad_norm": 0.226577342689351,
"learning_rate": 7.790136134700042e-05,
"loss": 0.5181,
"step": 2140
},
{
"epoch": 1.849462365591398,
"grad_norm": 0.24205636496242264,
"learning_rate": 7.692578110892876e-05,
"loss": 0.5206,
"step": 2150
},
{
"epoch": 1.8580645161290321,
"grad_norm": 0.24349873078348327,
"learning_rate": 7.595251503460778e-05,
"loss": 0.5226,
"step": 2160
},
{
"epoch": 1.8666666666666667,
"grad_norm": 0.23758597762780134,
"learning_rate": 7.498166073501529e-05,
"loss": 0.5244,
"step": 2170
},
{
"epoch": 1.875268817204301,
"grad_norm": 0.2284958254909026,
"learning_rate": 7.401331557924707e-05,
"loss": 0.505,
"step": 2180
},
{
"epoch": 1.8838709677419354,
"grad_norm": 0.24598875919420438,
"learning_rate": 7.304757668475122e-05,
"loss": 0.5317,
"step": 2190
},
{
"epoch": 1.89247311827957,
"grad_norm": 0.24113280579774263,
"learning_rate": 7.208454090758832e-05,
"loss": 0.5178,
"step": 2200
},
{
"epoch": 1.9010752688172043,
"grad_norm": 0.2362286661628526,
"learning_rate": 7.112430483271746e-05,
"loss": 0.5202,
"step": 2210
},
{
"epoch": 1.9096774193548387,
"grad_norm": 0.23413327678165968,
"learning_rate": 7.016696476430931e-05,
"loss": 0.4944,
"step": 2220
},
{
"epoch": 1.9182795698924733,
"grad_norm": 0.23501629015109748,
"learning_rate": 6.921261671608791e-05,
"loss": 0.5153,
"step": 2230
},
{
"epoch": 1.9268817204301074,
"grad_norm": 0.27066815871872996,
"learning_rate": 6.826135640170101e-05,
"loss": 0.5252,
"step": 2240
},
{
"epoch": 1.935483870967742,
"grad_norm": 0.2439503871040344,
"learning_rate": 6.731327922512074e-05,
"loss": 0.5112,
"step": 2250
},
{
"epoch": 1.9440860215053763,
"grad_norm": 0.23161029568643865,
"learning_rate": 6.636848027107544e-05,
"loss": 0.5307,
"step": 2260
},
{
"epoch": 1.9526881720430107,
"grad_norm": 0.24210935717094484,
"learning_rate": 6.54270542955134e-05,
"loss": 0.5011,
"step": 2270
},
{
"epoch": 1.9612903225806453,
"grad_norm": 0.24063211639329316,
"learning_rate": 6.44890957160994e-05,
"loss": 0.5234,
"step": 2280
},
{
"epoch": 1.9698924731182794,
"grad_norm": 0.23615678793407338,
"learning_rate": 6.355469860274574e-05,
"loss": 0.5046,
"step": 2290
},
{
"epoch": 1.978494623655914,
"grad_norm": 0.22842985809994715,
"learning_rate": 6.262395666817724e-05,
"loss": 0.5012,
"step": 2300
},
{
"epoch": 1.9870967741935484,
"grad_norm": 0.23784853454359128,
"learning_rate": 6.169696325853312e-05,
"loss": 0.519,
"step": 2310
},
{
"epoch": 1.9956989247311827,
"grad_norm": 0.2342469521451372,
"learning_rate": 6.077381134400462e-05,
"loss": 0.5002,
"step": 2320
},
{
"epoch": 2.0043010752688173,
"grad_norm": 0.22833969780513486,
"learning_rate": 5.985459350951121e-05,
"loss": 0.4845,
"step": 2330
},
{
"epoch": 2.0129032258064514,
"grad_norm": 0.2544744796985262,
"learning_rate": 5.893940194541492e-05,
"loss": 0.464,
"step": 2340
},
{
"epoch": 2.021505376344086,
"grad_norm": 0.24382320744489763,
"learning_rate": 5.802832843827419e-05,
"loss": 0.455,
"step": 2350
},
{
"epoch": 2.0301075268817206,
"grad_norm": 0.24697875997895144,
"learning_rate": 5.712146436163863e-05,
"loss": 0.4598,
"step": 2360
},
{
"epoch": 2.0387096774193547,
"grad_norm": 0.2513303656347981,
"learning_rate": 5.6218900666884975e-05,
"loss": 0.4515,
"step": 2370
},
{
"epoch": 2.0473118279569893,
"grad_norm": 0.26236895102262536,
"learning_rate": 5.5320727874095014e-05,
"loss": 0.4598,
"step": 2380
},
{
"epoch": 2.055913978494624,
"grad_norm": 0.25978280232463796,
"learning_rate": 5.4427036062977744e-05,
"loss": 0.451,
"step": 2390
},
{
"epoch": 2.064516129032258,
"grad_norm": 0.26162376545024385,
"learning_rate": 5.3537914863834374e-05,
"loss": 0.4563,
"step": 2400
},
{
"epoch": 2.0731182795698926,
"grad_norm": 0.27200603174186766,
"learning_rate": 5.265345344856979e-05,
"loss": 0.4595,
"step": 2410
},
{
"epoch": 2.0817204301075267,
"grad_norm": 0.2516133186216482,
"learning_rate": 5.1773740521748793e-05,
"loss": 0.4543,
"step": 2420
},
{
"epoch": 2.0903225806451613,
"grad_norm": 0.2559266512535566,
"learning_rate": 5.089886431169999e-05,
"loss": 0.4576,
"step": 2430
},
{
"epoch": 2.098924731182796,
"grad_norm": 0.26643890557668787,
"learning_rate": 5.0028912561667104e-05,
"loss": 0.4651,
"step": 2440
},
{
"epoch": 2.10752688172043,
"grad_norm": 0.2594035563202495,
"learning_rate": 4.916397252100892e-05,
"loss": 0.4581,
"step": 2450
},
{
"epoch": 2.1161290322580646,
"grad_norm": 0.2589087763537241,
"learning_rate": 4.830413093644913e-05,
"loss": 0.4508,
"step": 2460
},
{
"epoch": 2.1247311827956987,
"grad_norm": 0.25564217019549307,
"learning_rate": 4.744947404337605e-05,
"loss": 0.4498,
"step": 2470
},
{
"epoch": 2.1333333333333333,
"grad_norm": 0.25846858564022296,
"learning_rate": 4.660008755719397e-05,
"loss": 0.4439,
"step": 2480
},
{
"epoch": 2.141935483870968,
"grad_norm": 0.2615309859057683,
"learning_rate": 4.5756056664726554e-05,
"loss": 0.4621,
"step": 2490
},
{
"epoch": 2.150537634408602,
"grad_norm": 0.2751137583558904,
"learning_rate": 4.491746601567343e-05,
"loss": 0.4542,
"step": 2500
},
{
"epoch": 2.1591397849462366,
"grad_norm": 0.2640093266145403,
"learning_rate": 4.408439971412013e-05,
"loss": 0.4709,
"step": 2510
},
{
"epoch": 2.167741935483871,
"grad_norm": 0.26804533689921306,
"learning_rate": 4.325694131010346e-05,
"loss": 0.4545,
"step": 2520
},
{
"epoch": 2.1763440860215053,
"grad_norm": 0.26953837584010415,
"learning_rate": 4.243517379123193e-05,
"loss": 0.447,
"step": 2530
},
{
"epoch": 2.18494623655914,
"grad_norm": 0.26331494847521747,
"learning_rate": 4.161917957436271e-05,
"loss": 0.4454,
"step": 2540
},
{
"epoch": 2.193548387096774,
"grad_norm": 0.26650178799391877,
"learning_rate": 4.080904049733607e-05,
"loss": 0.4452,
"step": 2550
},
{
"epoch": 2.2021505376344086,
"grad_norm": 0.2596397199813554,
"learning_rate": 4.0004837810767294e-05,
"loss": 0.455,
"step": 2560
},
{
"epoch": 2.210752688172043,
"grad_norm": 0.2855359315189352,
"learning_rate": 3.9206652169898364e-05,
"loss": 0.4619,
"step": 2570
},
{
"epoch": 2.2193548387096773,
"grad_norm": 0.2715169650377671,
"learning_rate": 3.841456362650837e-05,
"loss": 0.4615,
"step": 2580
},
{
"epoch": 2.227956989247312,
"grad_norm": 0.2628755048951637,
"learning_rate": 3.7628651620885444e-05,
"loss": 0.4645,
"step": 2590
},
{
"epoch": 2.236559139784946,
"grad_norm": 0.28592361434576347,
"learning_rate": 3.6848994973859105e-05,
"loss": 0.4692,
"step": 2600
},
{
"epoch": 2.2451612903225806,
"grad_norm": 0.26764492546621405,
"learning_rate": 3.607567187889538e-05,
"loss": 0.4563,
"step": 2610
},
{
"epoch": 2.253763440860215,
"grad_norm": 0.271421407542708,
"learning_rate": 3.5308759894254496e-05,
"loss": 0.4671,
"step": 2620
},
{
"epoch": 2.2623655913978493,
"grad_norm": 0.2653346076557134,
"learning_rate": 3.45483359352125e-05,
"loss": 0.4535,
"step": 2630
},
{
"epoch": 2.270967741935484,
"grad_norm": 0.2672124362703016,
"learning_rate": 3.379447626634712e-05,
"loss": 0.4523,
"step": 2640
},
{
"epoch": 2.279569892473118,
"grad_norm": 0.2815570906050588,
"learning_rate": 3.304725649388919e-05,
"loss": 0.468,
"step": 2650
},
{
"epoch": 2.2881720430107526,
"grad_norm": 0.27533793807613155,
"learning_rate": 3.230675155813979e-05,
"loss": 0.4587,
"step": 2660
},
{
"epoch": 2.296774193548387,
"grad_norm": 0.2718010567194757,
"learning_rate": 3.1573035725954344e-05,
"loss": 0.4518,
"step": 2670
},
{
"epoch": 2.3053763440860213,
"grad_norm": 0.26881948505806064,
"learning_rate": 3.084618258329443e-05,
"loss": 0.4511,
"step": 2680
},
{
"epoch": 2.313978494623656,
"grad_norm": 0.2735905144552066,
"learning_rate": 3.012626502784729e-05,
"loss": 0.4437,
"step": 2690
},
{
"epoch": 2.3225806451612905,
"grad_norm": 0.2753569674239171,
"learning_rate": 2.9413355261715192e-05,
"loss": 0.457,
"step": 2700
},
{
"epoch": 2.3311827956989246,
"grad_norm": 0.27264140236415646,
"learning_rate": 2.87075247841738e-05,
"loss": 0.4562,
"step": 2710
},
{
"epoch": 2.339784946236559,
"grad_norm": 0.2750362048193341,
"learning_rate": 2.8008844384501566e-05,
"loss": 0.4578,
"step": 2720
},
{
"epoch": 2.3483870967741938,
"grad_norm": 0.2743768587791329,
"learning_rate": 2.7317384134879965e-05,
"loss": 0.4569,
"step": 2730
},
{
"epoch": 2.356989247311828,
"grad_norm": 0.27994254494622967,
"learning_rate": 2.6633213383365906e-05,
"loss": 0.4554,
"step": 2740
},
{
"epoch": 2.3655913978494625,
"grad_norm": 0.273412382006339,
"learning_rate": 2.595640074693664e-05,
"loss": 0.4521,
"step": 2750
},
{
"epoch": 2.3741935483870966,
"grad_norm": 0.26693176903025,
"learning_rate": 2.5287014104607975e-05,
"loss": 0.4472,
"step": 2760
},
{
"epoch": 2.382795698924731,
"grad_norm": 0.26563176753886897,
"learning_rate": 2.4625120590626595e-05,
"loss": 0.4487,
"step": 2770
},
{
"epoch": 2.3913978494623658,
"grad_norm": 0.27008804180981577,
"learning_rate": 2.397078658773699e-05,
"loss": 0.4611,
"step": 2780
},
{
"epoch": 2.4,
"grad_norm": 0.2624172113992881,
"learning_rate": 2.3324077720523785e-05,
"loss": 0.4461,
"step": 2790
},
{
"epoch": 2.4086021505376345,
"grad_norm": 0.27677007427727796,
"learning_rate": 2.2685058848830076e-05,
"loss": 0.4536,
"step": 2800
},
{
"epoch": 2.4172043010752686,
"grad_norm": 0.27569949386731046,
"learning_rate": 2.2053794061252675e-05,
"loss": 0.4474,
"step": 2810
},
{
"epoch": 2.425806451612903,
"grad_norm": 0.267987384900051,
"learning_rate": 2.1430346668714175e-05,
"loss": 0.4487,
"step": 2820
},
{
"epoch": 2.434408602150538,
"grad_norm": 0.2669869480120869,
"learning_rate": 2.0814779198113687e-05,
"loss": 0.455,
"step": 2830
},
{
"epoch": 2.443010752688172,
"grad_norm": 0.26354591329030985,
"learning_rate": 2.020715338605581e-05,
"loss": 0.4616,
"step": 2840
},
{
"epoch": 2.4516129032258065,
"grad_norm": 0.2785720489510887,
"learning_rate": 1.9607530172658715e-05,
"loss": 0.4419,
"step": 2850
},
{
"epoch": 2.4602150537634406,
"grad_norm": 0.27851326696395934,
"learning_rate": 1.9015969695442704e-05,
"loss": 0.4572,
"step": 2860
},
{
"epoch": 2.468817204301075,
"grad_norm": 0.27236418612561275,
"learning_rate": 1.8432531283298458e-05,
"loss": 0.453,
"step": 2870
},
{
"epoch": 2.47741935483871,
"grad_norm": 0.2816067593156165,
"learning_rate": 1.7857273450537227e-05,
"loss": 0.4607,
"step": 2880
},
{
"epoch": 2.486021505376344,
"grad_norm": 0.27424768533392857,
"learning_rate": 1.7290253891022e-05,
"loss": 0.4385,
"step": 2890
},
{
"epoch": 2.4946236559139785,
"grad_norm": 0.2766562755277204,
"learning_rate": 1.673152947238139e-05,
"loss": 0.4567,
"step": 2900
},
{
"epoch": 2.5032258064516126,
"grad_norm": 0.2614813998070769,
"learning_rate": 1.618115623030625e-05,
"loss": 0.4428,
"step": 2910
},
{
"epoch": 2.511827956989247,
"grad_norm": 0.272212168117098,
"learning_rate": 1.5639189362929695e-05,
"loss": 0.453,
"step": 2920
},
{
"epoch": 2.520430107526882,
"grad_norm": 0.2747575768340182,
"learning_rate": 1.5105683225291211e-05,
"loss": 0.4586,
"step": 2930
},
{
"epoch": 2.5290322580645164,
"grad_norm": 0.2711344401379175,
"learning_rate": 1.4580691323885209e-05,
"loss": 0.4506,
"step": 2940
},
{
"epoch": 2.5376344086021505,
"grad_norm": 0.27821711929840565,
"learning_rate": 1.4064266311294793e-05,
"loss": 0.4669,
"step": 2950
},
{
"epoch": 2.546236559139785,
"grad_norm": 0.2807569627250837,
"learning_rate": 1.3556459980911085e-05,
"loss": 0.4567,
"step": 2960
},
{
"epoch": 2.554838709677419,
"grad_norm": 0.27478294823892596,
"learning_rate": 1.305732326173882e-05,
"loss": 0.4629,
"step": 2970
},
{
"epoch": 2.563440860215054,
"grad_norm": 0.2694458161205595,
"learning_rate": 1.2566906213288388e-05,
"loss": 0.4518,
"step": 2980
},
{
"epoch": 2.5720430107526884,
"grad_norm": 0.26379184313364523,
"learning_rate": 1.2085258020555556e-05,
"loss": 0.4489,
"step": 2990
},
{
"epoch": 2.5806451612903225,
"grad_norm": 0.27978762666210216,
"learning_rate": 1.1612426989088232e-05,
"loss": 0.4493,
"step": 3000
},
{
"epoch": 2.589247311827957,
"grad_norm": 0.28005079300575625,
"learning_rate": 1.1148460540142125e-05,
"loss": 0.4501,
"step": 3010
},
{
"epoch": 2.5978494623655912,
"grad_norm": 0.2765800997942907,
"learning_rate": 1.0693405205924579e-05,
"loss": 0.4599,
"step": 3020
},
{
"epoch": 2.606451612903226,
"grad_norm": 0.26856411845690764,
"learning_rate": 1.0247306624927789e-05,
"loss": 0.4489,
"step": 3030
},
{
"epoch": 2.6150537634408604,
"grad_norm": 0.2758438510594412,
"learning_rate": 9.810209537351645e-06,
"loss": 0.4427,
"step": 3040
},
{
"epoch": 2.6236559139784945,
"grad_norm": 0.2808673768423207,
"learning_rate": 9.382157780616606e-06,
"loss": 0.4369,
"step": 3050
},
{
"epoch": 2.632258064516129,
"grad_norm": 0.26355013415952777,
"learning_rate": 8.963194284967202e-06,
"loss": 0.4543,
"step": 3060
},
{
"epoch": 2.6408602150537632,
"grad_norm": 0.26682463219677827,
"learning_rate": 8.553361069166388e-06,
"loss": 0.4448,
"step": 3070
},
{
"epoch": 2.649462365591398,
"grad_norm": 0.2824919005760602,
"learning_rate": 8.15269923628147e-06,
"loss": 0.463,
"step": 3080
},
{
"epoch": 2.6580645161290324,
"grad_norm": 0.2606472077942361,
"learning_rate": 7.761248969561729e-06,
"loss": 0.4435,
"step": 3090
},
{
"epoch": 2.6666666666666665,
"grad_norm": 0.2716672911173004,
"learning_rate": 7.379049528408433e-06,
"loss": 0.4436,
"step": 3100
},
{
"epoch": 2.675268817204301,
"grad_norm": 0.2692189798747163,
"learning_rate": 7.00613924443726e-06,
"loss": 0.456,
"step": 3110
},
{
"epoch": 2.6838709677419352,
"grad_norm": 0.2751219149255623,
"learning_rate": 6.642555517634197e-06,
"loss": 0.4486,
"step": 3120
},
{
"epoch": 2.69247311827957,
"grad_norm": 0.26558006981794996,
"learning_rate": 6.288334812604324e-06,
"loss": 0.448,
"step": 3130
},
{
"epoch": 2.7010752688172044,
"grad_norm": 0.2859552474881587,
"learning_rate": 5.943512654914951e-06,
"loss": 0.4357,
"step": 3140
},
{
"epoch": 2.709677419354839,
"grad_norm": 0.27548323875619696,
"learning_rate": 5.6081236275325355e-06,
"loss": 0.4624,
"step": 3150
},
{
"epoch": 2.718279569892473,
"grad_norm": 0.2806542534666262,
"learning_rate": 5.282201367354245e-06,
"loss": 0.4505,
"step": 3160
},
{
"epoch": 2.7268817204301077,
"grad_norm": 0.26475293341781525,
"learning_rate": 4.965778561834644e-06,
"loss": 0.4386,
"step": 3170
},
{
"epoch": 2.735483870967742,
"grad_norm": 0.28440947626379526,
"learning_rate": 4.658886945707164e-06,
"loss": 0.4542,
"step": 3180
},
{
"epoch": 2.7440860215053764,
"grad_norm": 0.2696410246213196,
"learning_rate": 4.361557297801499e-06,
"loss": 0.4416,
"step": 3190
},
{
"epoch": 2.752688172043011,
"grad_norm": 0.27708040093077324,
"learning_rate": 4.073819437956694e-06,
"loss": 0.4471,
"step": 3200
},
{
"epoch": 2.761290322580645,
"grad_norm": 0.2615573424318419,
"learning_rate": 3.7957022240304173e-06,
"loss": 0.4505,
"step": 3210
},
{
"epoch": 2.7698924731182797,
"grad_norm": 0.2776744589099636,
"learning_rate": 3.5272335490047937e-06,
"loss": 0.4416,
"step": 3220
},
{
"epoch": 2.778494623655914,
"grad_norm": 0.28362883428775265,
"learning_rate": 3.2684403381889272e-06,
"loss": 0.4562,
"step": 3230
},
{
"epoch": 2.7870967741935484,
"grad_norm": 0.2659089658104915,
"learning_rate": 3.019348546518508e-06,
"loss": 0.4542,
"step": 3240
},
{
"epoch": 2.795698924731183,
"grad_norm": 0.27817617897785124,
"learning_rate": 2.7799831559527258e-06,
"loss": 0.4535,
"step": 3250
},
{
"epoch": 2.804301075268817,
"grad_norm": 0.2767257855534767,
"learning_rate": 2.550368172968809e-06,
"loss": 0.4438,
"step": 3260
},
{
"epoch": 2.8129032258064517,
"grad_norm": 0.2659520383981123,
"learning_rate": 2.3305266261542945e-06,
"loss": 0.4648,
"step": 3270
},
{
"epoch": 2.821505376344086,
"grad_norm": 0.26860959961537617,
"learning_rate": 2.1204805638975646e-06,
"loss": 0.4478,
"step": 3280
},
{
"epoch": 2.8301075268817204,
"grad_norm": 0.2772290842479072,
"learning_rate": 1.9202510521763696e-06,
"loss": 0.4556,
"step": 3290
},
{
"epoch": 2.838709677419355,
"grad_norm": 0.2773729267266774,
"learning_rate": 1.7298581724452978e-06,
"loss": 0.4546,
"step": 3300
},
{
"epoch": 2.847311827956989,
"grad_norm": 0.288243164539903,
"learning_rate": 1.5493210196216079e-06,
"loss": 0.457,
"step": 3310
},
{
"epoch": 2.8559139784946237,
"grad_norm": 0.27792696429990915,
"learning_rate": 1.378657700170205e-06,
"loss": 0.4585,
"step": 3320
},
{
"epoch": 2.864516129032258,
"grad_norm": 0.27882422779333554,
"learning_rate": 1.2178853302877159e-06,
"loss": 0.4486,
"step": 3330
},
{
"epoch": 2.8731182795698924,
"grad_norm": 0.28133695329057534,
"learning_rate": 1.0670200341858394e-06,
"loss": 0.4502,
"step": 3340
},
{
"epoch": 2.881720430107527,
"grad_norm": 0.2859131700306455,
"learning_rate": 9.260769424742633e-07,
"loss": 0.4452,
"step": 3350
},
{
"epoch": 2.8903225806451616,
"grad_norm": 0.26707708683830167,
"learning_rate": 7.950701906431324e-07,
"loss": 0.4498,
"step": 3360
},
{
"epoch": 2.8989247311827957,
"grad_norm": 0.2716338526967156,
"learning_rate": 6.740129176453725e-07,
"loss": 0.4568,
"step": 3370
},
{
"epoch": 2.90752688172043,
"grad_norm": 0.29044410591630965,
"learning_rate": 5.629172645789882e-07,
"loss": 0.4517,
"step": 3380
},
{
"epoch": 2.9161290322580644,
"grad_norm": 0.27054199049752875,
"learning_rate": 4.617943734694152e-07,
"loss": 0.4559,
"step": 3390
},
{
"epoch": 2.924731182795699,
"grad_norm": 0.27515263341966606,
"learning_rate": 3.7065438615198066e-07,
"loss": 0.4576,
"step": 3400
},
{
"epoch": 2.9333333333333336,
"grad_norm": 0.2653013119464242,
"learning_rate": 2.8950644325485e-07,
"loss": 0.449,
"step": 3410
},
{
"epoch": 2.9419354838709677,
"grad_norm": 0.2791971697586357,
"learning_rate": 2.183586832822493e-07,
"loss": 0.4562,
"step": 3420
},
{
"epoch": 2.9505376344086023,
"grad_norm": 0.2668878384341789,
"learning_rate": 1.572182417982515e-07,
"loss": 0.447,
"step": 3430
},
{
"epoch": 2.9591397849462364,
"grad_norm": 0.2759015749441583,
"learning_rate": 1.0609125071109338e-07,
"loss": 0.4567,
"step": 3440
},
{
"epoch": 2.967741935483871,
"grad_norm": 0.283562302360861,
"learning_rate": 6.49828376582673e-08,
"loss": 0.4503,
"step": 3450
},
{
"epoch": 2.9763440860215056,
"grad_norm": 0.28059419780497596,
"learning_rate": 3.3897125492188266e-08,
"loss": 0.4525,
"step": 3460
},
{
"epoch": 2.9849462365591397,
"grad_norm": 0.2757143204657877,
"learning_rate": 1.2837231866746902e-08,
"loss": 0.4455,
"step": 3470
},
{
"epoch": 2.9935483870967743,
"grad_norm": 0.26729797929321086,
"learning_rate": 1.8052689246150779e-09,
"loss": 0.4443,
"step": 3480
},
{
"epoch": 2.9987096774193547,
"step": 3486,
"total_flos": 6514990399881216.0,
"train_loss": 0.5279938033170147,
"train_runtime": 9440.7969,
"train_samples_per_second": 5.91,
"train_steps_per_second": 0.369
}
],
"logging_steps": 10,
"max_steps": 3486,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 6514990399881216.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}