ljcamargo's picture
Training in progress, step 3104, checkpoint
0a1a01f verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.9987925621830476,
"eval_steps": 500,
"global_step": 3104,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.000643966835707961,
"grad_norm": 225.0,
"learning_rate": 0.0,
"loss": 13.2867,
"step": 1
},
{
"epoch": 0.006439668357079611,
"grad_norm": 186.0,
"learning_rate": 2.903225806451613e-06,
"loss": 11.3314,
"step": 10
},
{
"epoch": 0.012879336714159221,
"grad_norm": 127.5,
"learning_rate": 6.129032258064516e-06,
"loss": 10.924,
"step": 20
},
{
"epoch": 0.01931900507123883,
"grad_norm": 143.0,
"learning_rate": 9.35483870967742e-06,
"loss": 7.8773,
"step": 30
},
{
"epoch": 0.025758673428318443,
"grad_norm": 89.5,
"learning_rate": 1.2580645161290322e-05,
"loss": 6.4675,
"step": 40
},
{
"epoch": 0.03219834178539805,
"grad_norm": 60.75,
"learning_rate": 1.5806451612903226e-05,
"loss": 4.5193,
"step": 50
},
{
"epoch": 0.03863801014247766,
"grad_norm": 42.25,
"learning_rate": 1.9032258064516127e-05,
"loss": 3.6873,
"step": 60
},
{
"epoch": 0.045077678499557274,
"grad_norm": 42.0,
"learning_rate": 2.2258064516129034e-05,
"loss": 2.923,
"step": 70
},
{
"epoch": 0.051517346856636885,
"grad_norm": 35.5,
"learning_rate": 2.5483870967741935e-05,
"loss": 2.6102,
"step": 80
},
{
"epoch": 0.057957015213716497,
"grad_norm": 37.25,
"learning_rate": 2.8709677419354843e-05,
"loss": 2.1082,
"step": 90
},
{
"epoch": 0.0643966835707961,
"grad_norm": 36.25,
"learning_rate": 3.193548387096774e-05,
"loss": 2.022,
"step": 100
},
{
"epoch": 0.07083635192787571,
"grad_norm": 35.5,
"learning_rate": 3.516129032258065e-05,
"loss": 1.8552,
"step": 110
},
{
"epoch": 0.07727602028495532,
"grad_norm": 29.625,
"learning_rate": 3.838709677419355e-05,
"loss": 1.585,
"step": 120
},
{
"epoch": 0.08371568864203494,
"grad_norm": 37.75,
"learning_rate": 4.161290322580645e-05,
"loss": 1.8271,
"step": 130
},
{
"epoch": 0.09015535699911455,
"grad_norm": 37.5,
"learning_rate": 4.4838709677419356e-05,
"loss": 1.4702,
"step": 140
},
{
"epoch": 0.09659502535619416,
"grad_norm": 35.25,
"learning_rate": 4.806451612903226e-05,
"loss": 1.8647,
"step": 150
},
{
"epoch": 0.10303469371327377,
"grad_norm": 38.75,
"learning_rate": 4.999977302420873e-05,
"loss": 1.4572,
"step": 160
},
{
"epoch": 0.10947436207035338,
"grad_norm": 31.875,
"learning_rate": 4.999721959388888e-05,
"loss": 1.4709,
"step": 170
},
{
"epoch": 0.11591403042743299,
"grad_norm": 45.5,
"learning_rate": 4.9991829304257186e-05,
"loss": 1.3333,
"step": 180
},
{
"epoch": 0.1223536987845126,
"grad_norm": 43.25,
"learning_rate": 4.998360276704142e-05,
"loss": 1.383,
"step": 190
},
{
"epoch": 0.1287933671415922,
"grad_norm": 22.375,
"learning_rate": 4.9972540915846554e-05,
"loss": 1.3452,
"step": 200
},
{
"epoch": 0.13523303549867183,
"grad_norm": 18.0,
"learning_rate": 4.9958645006048835e-05,
"loss": 1.2438,
"step": 210
},
{
"epoch": 0.14167270385575143,
"grad_norm": 22.375,
"learning_rate": 4.9941916614653294e-05,
"loss": 1.2204,
"step": 220
},
{
"epoch": 0.14811237221283105,
"grad_norm": 27.625,
"learning_rate": 4.992235764011478e-05,
"loss": 1.1272,
"step": 230
},
{
"epoch": 0.15455204056991065,
"grad_norm": 37.5,
"learning_rate": 4.9899970302122535e-05,
"loss": 1.1991,
"step": 240
},
{
"epoch": 0.16099170892699027,
"grad_norm": 33.0,
"learning_rate": 4.987475714134825e-05,
"loss": 1.3587,
"step": 250
},
{
"epoch": 0.16743137728406987,
"grad_norm": 25.0,
"learning_rate": 4.984672101915774e-05,
"loss": 1.2084,
"step": 260
},
{
"epoch": 0.17387104564114947,
"grad_norm": 22.0,
"learning_rate": 4.981586511728624e-05,
"loss": 1.0549,
"step": 270
},
{
"epoch": 0.1803107139982291,
"grad_norm": 41.25,
"learning_rate": 4.978219293747731e-05,
"loss": 1.1439,
"step": 280
},
{
"epoch": 0.1867503823553087,
"grad_norm": 18.25,
"learning_rate": 4.9745708301085426e-05,
"loss": 1.1516,
"step": 290
},
{
"epoch": 0.19319005071238832,
"grad_norm": 21.0,
"learning_rate": 4.970641534864229e-05,
"loss": 1.0755,
"step": 300
},
{
"epoch": 0.19962971906946791,
"grad_norm": 17.75,
"learning_rate": 4.966431853938697e-05,
"loss": 0.9986,
"step": 310
},
{
"epoch": 0.20606938742654754,
"grad_norm": 34.0,
"learning_rate": 4.96194226507598e-05,
"loss": 1.1859,
"step": 320
},
{
"epoch": 0.21250905578362714,
"grad_norm": 30.5,
"learning_rate": 4.957173277786024e-05,
"loss": 1.1676,
"step": 330
},
{
"epoch": 0.21894872414070676,
"grad_norm": 25.75,
"learning_rate": 4.95212543328686e-05,
"loss": 1.0089,
"step": 340
},
{
"epoch": 0.22538839249778636,
"grad_norm": 25.0,
"learning_rate": 4.946799304443186e-05,
"loss": 0.9871,
"step": 350
},
{
"epoch": 0.23182806085486599,
"grad_norm": 27.0,
"learning_rate": 4.941195495701357e-05,
"loss": 0.9072,
"step": 360
},
{
"epoch": 0.23826772921194558,
"grad_norm": 23.875,
"learning_rate": 4.935314643020781e-05,
"loss": 1.0019,
"step": 370
},
{
"epoch": 0.2447073975690252,
"grad_norm": 21.625,
"learning_rate": 4.929157413801752e-05,
"loss": 0.9848,
"step": 380
},
{
"epoch": 0.2511470659261048,
"grad_norm": 42.75,
"learning_rate": 4.922724506809705e-05,
"loss": 1.1026,
"step": 390
},
{
"epoch": 0.2575867342831844,
"grad_norm": 18.5,
"learning_rate": 4.9160166520959175e-05,
"loss": 0.933,
"step": 400
},
{
"epoch": 0.264026402640264,
"grad_norm": 38.25,
"learning_rate": 4.909034610914659e-05,
"loss": 1.1057,
"step": 410
},
{
"epoch": 0.27046607099734365,
"grad_norm": 17.25,
"learning_rate": 4.9017791756367935e-05,
"loss": 0.8572,
"step": 420
},
{
"epoch": 0.27690573935442325,
"grad_norm": 19.125,
"learning_rate": 4.894251169659863e-05,
"loss": 0.9424,
"step": 430
},
{
"epoch": 0.28334540771150285,
"grad_norm": 26.25,
"learning_rate": 4.886451447314638e-05,
"loss": 0.8728,
"step": 440
},
{
"epoch": 0.28978507606858245,
"grad_norm": 22.375,
"learning_rate": 4.878380893768161e-05,
"loss": 0.9554,
"step": 450
},
{
"epoch": 0.2962247444256621,
"grad_norm": 24.625,
"learning_rate": 4.8700404249232936e-05,
"loss": 0.8738,
"step": 460
},
{
"epoch": 0.3026644127827417,
"grad_norm": 32.25,
"learning_rate": 4.861430987314774e-05,
"loss": 1.0007,
"step": 470
},
{
"epoch": 0.3091040811398213,
"grad_norm": 36.0,
"learning_rate": 4.852553558001799e-05,
"loss": 0.8721,
"step": 480
},
{
"epoch": 0.3155437494969009,
"grad_norm": 25.5,
"learning_rate": 4.8434091444571354e-05,
"loss": 0.8288,
"step": 490
},
{
"epoch": 0.32198341785398055,
"grad_norm": 14.9375,
"learning_rate": 4.833998784452789e-05,
"loss": 0.8136,
"step": 500
},
{
"epoch": 0.32842308621106014,
"grad_norm": 21.0,
"learning_rate": 4.824323545942229e-05,
"loss": 0.8847,
"step": 510
},
{
"epoch": 0.33486275456813974,
"grad_norm": 29.0,
"learning_rate": 4.814384526939192e-05,
"loss": 0.7825,
"step": 520
},
{
"epoch": 0.34130242292521934,
"grad_norm": 22.125,
"learning_rate": 4.804182855393068e-05,
"loss": 0.8063,
"step": 530
},
{
"epoch": 0.34774209128229894,
"grad_norm": 17.5,
"learning_rate": 4.793719689060896e-05,
"loss": 0.7661,
"step": 540
},
{
"epoch": 0.3541817596393786,
"grad_norm": 21.375,
"learning_rate": 4.7829962153759706e-05,
"loss": 0.7366,
"step": 550
},
{
"epoch": 0.3606214279964582,
"grad_norm": 21.125,
"learning_rate": 4.772013651313086e-05,
"loss": 0.7042,
"step": 560
},
{
"epoch": 0.3670610963535378,
"grad_norm": 13.1875,
"learning_rate": 4.760773243250427e-05,
"loss": 0.8331,
"step": 570
},
{
"epoch": 0.3735007647106174,
"grad_norm": 27.5,
"learning_rate": 4.749276266828115e-05,
"loss": 0.774,
"step": 580
},
{
"epoch": 0.37994043306769704,
"grad_norm": 25.375,
"learning_rate": 4.737524026803445e-05,
"loss": 0.8389,
"step": 590
},
{
"epoch": 0.38638010142477663,
"grad_norm": 16.75,
"learning_rate": 4.7255178569028116e-05,
"loss": 0.7925,
"step": 600
},
{
"epoch": 0.39281976978185623,
"grad_norm": 25.125,
"learning_rate": 4.713259119670347e-05,
"loss": 0.7335,
"step": 610
},
{
"epoch": 0.39925943813893583,
"grad_norm": 15.75,
"learning_rate": 4.700749206313291e-05,
"loss": 0.788,
"step": 620
},
{
"epoch": 0.4056991064960155,
"grad_norm": 15.375,
"learning_rate": 4.687989536544106e-05,
"loss": 0.7405,
"step": 630
},
{
"epoch": 0.4121387748530951,
"grad_norm": 15.0625,
"learning_rate": 4.67498155841936e-05,
"loss": 0.7865,
"step": 640
},
{
"epoch": 0.4185784432101747,
"grad_norm": 17.25,
"learning_rate": 4.661726748175389e-05,
"loss": 0.7056,
"step": 650
},
{
"epoch": 0.4250181115672543,
"grad_norm": 34.0,
"learning_rate": 4.6482266100607654e-05,
"loss": 0.8696,
"step": 660
},
{
"epoch": 0.4314577799243339,
"grad_norm": 13.0625,
"learning_rate": 4.6344826761655816e-05,
"loss": 0.737,
"step": 670
},
{
"epoch": 0.4378974482814135,
"grad_norm": 16.375,
"learning_rate": 4.6204965062475835e-05,
"loss": 0.8578,
"step": 680
},
{
"epoch": 0.4443371166384931,
"grad_norm": 13.375,
"learning_rate": 4.6062696875551535e-05,
"loss": 0.601,
"step": 690
},
{
"epoch": 0.4507767849955727,
"grad_norm": 19.375,
"learning_rate": 4.59180383464718e-05,
"loss": 0.8887,
"step": 700
},
{
"epoch": 0.4572164533526523,
"grad_norm": 14.5,
"learning_rate": 4.5771005892098264e-05,
"loss": 0.6319,
"step": 710
},
{
"epoch": 0.46365612170973197,
"grad_norm": 17.375,
"learning_rate": 4.56216161987022e-05,
"loss": 0.692,
"step": 720
},
{
"epoch": 0.47009579006681157,
"grad_norm": 97.0,
"learning_rate": 4.546988622007086e-05,
"loss": 0.6242,
"step": 730
},
{
"epoch": 0.47653545842389117,
"grad_norm": 36.5,
"learning_rate": 4.531583317558343e-05,
"loss": 0.7309,
"step": 740
},
{
"epoch": 0.48297512678097076,
"grad_norm": 29.875,
"learning_rate": 4.515947454825685e-05,
"loss": 0.7129,
"step": 750
},
{
"epoch": 0.4894147951380504,
"grad_norm": 17.75,
"learning_rate": 4.500082808276175e-05,
"loss": 0.6915,
"step": 760
},
{
"epoch": 0.49585446349513,
"grad_norm": 27.375,
"learning_rate": 4.48399117834086e-05,
"loss": 0.6255,
"step": 770
},
{
"epoch": 0.5022941318522096,
"grad_norm": 27.125,
"learning_rate": 4.467674391210451e-05,
"loss": 0.6425,
"step": 780
},
{
"epoch": 0.5087338002092893,
"grad_norm": 19.375,
"learning_rate": 4.4511342986280736e-05,
"loss": 0.9493,
"step": 790
},
{
"epoch": 0.5151734685663688,
"grad_norm": 16.75,
"learning_rate": 4.4343727776791155e-05,
"loss": 0.6505,
"step": 800
},
{
"epoch": 0.5216131369234485,
"grad_norm": 13.0,
"learning_rate": 4.4173917305782066e-05,
"loss": 0.6181,
"step": 810
},
{
"epoch": 0.528052805280528,
"grad_norm": 30.0,
"learning_rate": 4.400193084453338e-05,
"loss": 0.6455,
"step": 820
},
{
"epoch": 0.5344924736376077,
"grad_norm": 22.5,
"learning_rate": 4.382778791127159e-05,
"loss": 0.8168,
"step": 830
},
{
"epoch": 0.5409321419946873,
"grad_norm": 11.4375,
"learning_rate": 4.3651508268954756e-05,
"loss": 0.5452,
"step": 840
},
{
"epoch": 0.5473718103517669,
"grad_norm": 21.0,
"learning_rate": 4.347311192302958e-05,
"loss": 0.5801,
"step": 850
},
{
"epoch": 0.5538114787088465,
"grad_norm": 13.375,
"learning_rate": 4.329261911916113e-05,
"loss": 0.5966,
"step": 860
},
{
"epoch": 0.5602511470659262,
"grad_norm": 9.125,
"learning_rate": 4.311005034093516e-05,
"loss": 0.6723,
"step": 870
},
{
"epoch": 0.5666908154230057,
"grad_norm": 20.125,
"learning_rate": 4.292542630753352e-05,
"loss": 0.6408,
"step": 880
},
{
"epoch": 0.5731304837800854,
"grad_norm": 45.25,
"learning_rate": 4.27387679713828e-05,
"loss": 0.6709,
"step": 890
},
{
"epoch": 0.5795701521371649,
"grad_norm": 15.625,
"learning_rate": 4.2550096515776475e-05,
"loss": 0.5812,
"step": 900
},
{
"epoch": 0.5860098204942445,
"grad_norm": 13.5,
"learning_rate": 4.2359433352470924e-05,
"loss": 0.5777,
"step": 910
},
{
"epoch": 0.5924494888513242,
"grad_norm": 12.25,
"learning_rate": 4.2166800119255426e-05,
"loss": 0.582,
"step": 920
},
{
"epoch": 0.5988891572084037,
"grad_norm": 18.25,
"learning_rate": 4.197221867749658e-05,
"loss": 0.6012,
"step": 930
},
{
"epoch": 0.6053288255654834,
"grad_norm": 13.9375,
"learning_rate": 4.17757111096573e-05,
"loss": 0.5634,
"step": 940
},
{
"epoch": 0.6117684939225629,
"grad_norm": 13.8125,
"learning_rate": 4.157729971679079e-05,
"loss": 0.4827,
"step": 950
},
{
"epoch": 0.6182081622796426,
"grad_norm": 28.75,
"learning_rate": 4.137700701600961e-05,
"loss": 0.4897,
"step": 960
},
{
"epoch": 0.6246478306367222,
"grad_norm": 15.875,
"learning_rate": 4.1174855737930294e-05,
"loss": 0.6102,
"step": 970
},
{
"epoch": 0.6310874989938018,
"grad_norm": 13.8125,
"learning_rate": 4.097086882409375e-05,
"loss": 0.6418,
"step": 980
},
{
"epoch": 0.6375271673508814,
"grad_norm": 21.625,
"learning_rate": 4.0765069424361666e-05,
"loss": 0.5357,
"step": 990
},
{
"epoch": 0.6439668357079611,
"grad_norm": 15.875,
"learning_rate": 4.055748089428928e-05,
"loss": 0.6451,
"step": 1000
},
{
"epoch": 0.6504065040650406,
"grad_norm": 18.375,
"learning_rate": 4.034812679247489e-05,
"loss": 0.5809,
"step": 1010
},
{
"epoch": 0.6568461724221203,
"grad_norm": 14.6875,
"learning_rate": 4.01370308778862e-05,
"loss": 0.5031,
"step": 1020
},
{
"epoch": 0.6632858407791998,
"grad_norm": 10.6875,
"learning_rate": 3.992421710716402e-05,
"loss": 0.5555,
"step": 1030
},
{
"epoch": 0.6697255091362795,
"grad_norm": 18.75,
"learning_rate": 3.970970963190348e-05,
"loss": 0.7092,
"step": 1040
},
{
"epoch": 0.6761651774933591,
"grad_norm": 12.5,
"learning_rate": 3.9493532795913165e-05,
"loss": 0.4889,
"step": 1050
},
{
"epoch": 0.6826048458504387,
"grad_norm": 17.375,
"learning_rate": 3.927571113245236e-05,
"loss": 0.4842,
"step": 1060
},
{
"epoch": 0.6890445142075183,
"grad_norm": 12.8125,
"learning_rate": 3.9056269361446925e-05,
"loss": 0.6244,
"step": 1070
},
{
"epoch": 0.6954841825645979,
"grad_norm": 15.5,
"learning_rate": 3.883523238668379e-05,
"loss": 0.5288,
"step": 1080
},
{
"epoch": 0.7019238509216775,
"grad_norm": 18.5,
"learning_rate": 3.8612625292984826e-05,
"loss": 0.5128,
"step": 1090
},
{
"epoch": 0.7083635192787572,
"grad_norm": 16.125,
"learning_rate": 3.8388473343359957e-05,
"loss": 0.521,
"step": 1100
},
{
"epoch": 0.7148031876358367,
"grad_norm": 15.0625,
"learning_rate": 3.816280197614013e-05,
"loss": 0.5865,
"step": 1110
},
{
"epoch": 0.7212428559929164,
"grad_norm": 13.3125,
"learning_rate": 3.793563680209049e-05,
"loss": 0.5033,
"step": 1120
},
{
"epoch": 0.727682524349996,
"grad_norm": 37.0,
"learning_rate": 3.770700360150379e-05,
"loss": 0.4805,
"step": 1130
},
{
"epoch": 0.7341221927070756,
"grad_norm": 13.8125,
"learning_rate": 3.7476928321274694e-05,
"loss": 0.529,
"step": 1140
},
{
"epoch": 0.7405618610641552,
"grad_norm": 18.625,
"learning_rate": 3.724543707195518e-05,
"loss": 0.4636,
"step": 1150
},
{
"epoch": 0.7470015294212348,
"grad_norm": 11.625,
"learning_rate": 3.701255612479129e-05,
"loss": 0.4675,
"step": 1160
},
{
"epoch": 0.7534411977783144,
"grad_norm": 10.4375,
"learning_rate": 3.6778311908741686e-05,
"loss": 0.6616,
"step": 1170
},
{
"epoch": 0.7598808661353941,
"grad_norm": 11.9375,
"learning_rate": 3.654273100747833e-05,
"loss": 0.5331,
"step": 1180
},
{
"epoch": 0.7663205344924736,
"grad_norm": 14.1875,
"learning_rate": 3.630584015636956e-05,
"loss": 0.4434,
"step": 1190
},
{
"epoch": 0.7727602028495533,
"grad_norm": 10.4375,
"learning_rate": 3.6067666239446e-05,
"loss": 0.5477,
"step": 1200
},
{
"epoch": 0.7791998712066328,
"grad_norm": 10.5,
"learning_rate": 3.5828236286349546e-05,
"loss": 0.5726,
"step": 1210
},
{
"epoch": 0.7856395395637125,
"grad_norm": 9.0,
"learning_rate": 3.558757746926586e-05,
"loss": 0.4298,
"step": 1220
},
{
"epoch": 0.7920792079207921,
"grad_norm": 13.1875,
"learning_rate": 3.53457170998407e-05,
"loss": 0.6079,
"step": 1230
},
{
"epoch": 0.7985188762778717,
"grad_norm": 12.625,
"learning_rate": 3.510268262608036e-05,
"loss": 0.4401,
"step": 1240
},
{
"epoch": 0.8049585446349513,
"grad_norm": 23.5,
"learning_rate": 3.485850162923674e-05,
"loss": 0.5457,
"step": 1250
},
{
"epoch": 0.811398212992031,
"grad_norm": 14.375,
"learning_rate": 3.461320182067716e-05,
"loss": 0.6705,
"step": 1260
},
{
"epoch": 0.8178378813491105,
"grad_norm": 13.75,
"learning_rate": 3.4366811038739575e-05,
"loss": 0.3858,
"step": 1270
},
{
"epoch": 0.8242775497061902,
"grad_norm": 11.375,
"learning_rate": 3.411935724557317e-05,
"loss": 0.4403,
"step": 1280
},
{
"epoch": 0.8307172180632697,
"grad_norm": 13.625,
"learning_rate": 3.3870868523965095e-05,
"loss": 0.5022,
"step": 1290
},
{
"epoch": 0.8371568864203494,
"grad_norm": 12.1875,
"learning_rate": 3.362137307415344e-05,
"loss": 0.4955,
"step": 1300
},
{
"epoch": 0.843596554777429,
"grad_norm": 12.0625,
"learning_rate": 3.337089921062683e-05,
"loss": 0.401,
"step": 1310
},
{
"epoch": 0.8500362231345086,
"grad_norm": 9.9375,
"learning_rate": 3.311947535891114e-05,
"loss": 0.4773,
"step": 1320
},
{
"epoch": 0.8564758914915882,
"grad_norm": 17.125,
"learning_rate": 3.286713005234353e-05,
"loss": 0.416,
"step": 1330
},
{
"epoch": 0.8629155598486677,
"grad_norm": 7.84375,
"learning_rate": 3.261389192883432e-05,
"loss": 0.327,
"step": 1340
},
{
"epoch": 0.8693552282057474,
"grad_norm": 10.9375,
"learning_rate": 3.235978972761692e-05,
"loss": 0.381,
"step": 1350
},
{
"epoch": 0.875794896562827,
"grad_norm": 10.375,
"learning_rate": 3.2104852285986336e-05,
"loss": 0.4152,
"step": 1360
},
{
"epoch": 0.8822345649199066,
"grad_norm": 12.9375,
"learning_rate": 3.18491085360265e-05,
"loss": 0.4955,
"step": 1370
},
{
"epoch": 0.8886742332769862,
"grad_norm": 12.375,
"learning_rate": 3.159258750132685e-05,
"loss": 0.4323,
"step": 1380
},
{
"epoch": 0.8951139016340659,
"grad_norm": 9.5,
"learning_rate": 3.1335318293688516e-05,
"loss": 0.3519,
"step": 1390
},
{
"epoch": 0.9015535699911454,
"grad_norm": 13.875,
"learning_rate": 3.107733010982055e-05,
"loss": 0.4079,
"step": 1400
},
{
"epoch": 0.9079932383482251,
"grad_norm": 10.375,
"learning_rate": 3.081865222802645e-05,
"loss": 0.4186,
"step": 1410
},
{
"epoch": 0.9144329067053046,
"grad_norm": 9.6875,
"learning_rate": 3.055931400488144e-05,
"loss": 0.4845,
"step": 1420
},
{
"epoch": 0.9208725750623843,
"grad_norm": 10.3125,
"learning_rate": 3.0299344871900913e-05,
"loss": 0.3626,
"step": 1430
},
{
"epoch": 0.9273122434194639,
"grad_norm": 13.75,
"learning_rate": 3.0038774332200342e-05,
"loss": 0.3498,
"step": 1440
},
{
"epoch": 0.9337519117765435,
"grad_norm": 7.6875,
"learning_rate": 2.9777631957147002e-05,
"loss": 0.3391,
"step": 1450
},
{
"epoch": 0.9401915801336231,
"grad_norm": 8.4375,
"learning_rate": 2.9515947383004062e-05,
"loss": 0.3531,
"step": 1460
},
{
"epoch": 0.9466312484907027,
"grad_norm": 13.125,
"learning_rate": 2.9253750307567252e-05,
"loss": 0.4493,
"step": 1470
},
{
"epoch": 0.9530709168477823,
"grad_norm": 11.1875,
"learning_rate": 2.8991070486794503e-05,
"loss": 0.3705,
"step": 1480
},
{
"epoch": 0.959510585204862,
"grad_norm": 12.3125,
"learning_rate": 2.872793773142909e-05,
"loss": 0.3807,
"step": 1490
},
{
"epoch": 0.9659502535619415,
"grad_norm": 20.75,
"learning_rate": 2.8464381903616438e-05,
"loss": 0.3255,
"step": 1500
},
{
"epoch": 0.9723899219190212,
"grad_norm": 10.125,
"learning_rate": 2.8200432913515235e-05,
"loss": 0.3507,
"step": 1510
},
{
"epoch": 0.9788295902761008,
"grad_norm": 17.0,
"learning_rate": 2.7936120715902957e-05,
"loss": 0.4079,
"step": 1520
},
{
"epoch": 0.9852692586331804,
"grad_norm": 11.6875,
"learning_rate": 2.7671475306776396e-05,
"loss": 0.4972,
"step": 1530
},
{
"epoch": 0.99170892699026,
"grad_norm": 10.4375,
"learning_rate": 2.740652671994754e-05,
"loss": 0.3757,
"step": 1540
},
{
"epoch": 0.9981485953473396,
"grad_norm": 8.0625,
"learning_rate": 2.7141305023635083e-05,
"loss": 0.3437,
"step": 1550
},
{
"epoch": 1.0045077678499557,
"grad_norm": 11.25,
"learning_rate": 2.687584031705211e-05,
"loss": 0.2757,
"step": 1560
},
{
"epoch": 1.0109474362070354,
"grad_norm": 7.03125,
"learning_rate": 2.6610162726990185e-05,
"loss": 0.2032,
"step": 1570
},
{
"epoch": 1.017387104564115,
"grad_norm": 10.4375,
"learning_rate": 2.6344302404400417e-05,
"loss": 0.2548,
"step": 1580
},
{
"epoch": 1.0238267729211945,
"grad_norm": 8.75,
"learning_rate": 2.607828952097165e-05,
"loss": 0.2578,
"step": 1590
},
{
"epoch": 1.0302664412782743,
"grad_norm": 10.75,
"learning_rate": 2.5812154265706413e-05,
"loss": 0.2078,
"step": 1600
},
{
"epoch": 1.0367061096353538,
"grad_norm": 11.375,
"learning_rate": 2.5545926841494826e-05,
"loss": 0.1952,
"step": 1610
},
{
"epoch": 1.0431457779924334,
"grad_norm": 7.6875,
"learning_rate": 2.5279637461687026e-05,
"loss": 0.1958,
"step": 1620
},
{
"epoch": 1.049585446349513,
"grad_norm": 9.9375,
"learning_rate": 2.5013316346664273e-05,
"loss": 0.2133,
"step": 1630
},
{
"epoch": 1.0560251147065927,
"grad_norm": 8.75,
"learning_rate": 2.474699372040936e-05,
"loss": 0.2355,
"step": 1640
},
{
"epoch": 1.0624647830636722,
"grad_norm": 9.9375,
"learning_rate": 2.4480699807076592e-05,
"loss": 0.225,
"step": 1650
},
{
"epoch": 1.0689044514207517,
"grad_norm": 9.125,
"learning_rate": 2.4214464827561752e-05,
"loss": 0.272,
"step": 1660
},
{
"epoch": 1.0753441197778315,
"grad_norm": 12.4375,
"learning_rate": 2.3948318996072363e-05,
"loss": 0.2001,
"step": 1670
},
{
"epoch": 1.081783788134911,
"grad_norm": 7.375,
"learning_rate": 2.3682292516698832e-05,
"loss": 0.183,
"step": 1680
},
{
"epoch": 1.0882234564919906,
"grad_norm": 7.71875,
"learning_rate": 2.3416415579986643e-05,
"loss": 0.1954,
"step": 1690
},
{
"epoch": 1.0946631248490704,
"grad_norm": 6.625,
"learning_rate": 2.3150718359510142e-05,
"loss": 0.1848,
"step": 1700
},
{
"epoch": 1.10110279320615,
"grad_norm": 8.5,
"learning_rate": 2.2885231008448216e-05,
"loss": 0.1769,
"step": 1710
},
{
"epoch": 1.1075424615632294,
"grad_norm": 17.875,
"learning_rate": 2.261998365616228e-05,
"loss": 0.2485,
"step": 1720
},
{
"epoch": 1.1139821299203092,
"grad_norm": 9.9375,
"learning_rate": 2.2355006404777057e-05,
"loss": 0.1795,
"step": 1730
},
{
"epoch": 1.1204217982773887,
"grad_norm": 8.375,
"learning_rate": 2.2090329325764294e-05,
"loss": 0.2424,
"step": 1740
},
{
"epoch": 1.1268614666344683,
"grad_norm": 8.5,
"learning_rate": 2.1825982456530086e-05,
"loss": 0.203,
"step": 1750
},
{
"epoch": 1.133301134991548,
"grad_norm": 6.0,
"learning_rate": 2.1561995797006017e-05,
"loss": 0.2003,
"step": 1760
},
{
"epoch": 1.1397408033486276,
"grad_norm": 18.5,
"learning_rate": 2.129839930624454e-05,
"loss": 0.1599,
"step": 1770
},
{
"epoch": 1.1461804717057071,
"grad_norm": 11.375,
"learning_rate": 2.1035222899019057e-05,
"loss": 0.2198,
"step": 1780
},
{
"epoch": 1.1526201400627867,
"grad_norm": 6.65625,
"learning_rate": 2.0772496442428914e-05,
"loss": 0.1721,
"step": 1790
},
{
"epoch": 1.1590598084198664,
"grad_norm": 6.625,
"learning_rate": 2.051024975250996e-05,
"loss": 0.1557,
"step": 1800
},
{
"epoch": 1.165499476776946,
"grad_norm": 7.96875,
"learning_rate": 2.0248512590850716e-05,
"loss": 0.205,
"step": 1810
},
{
"epoch": 1.1719391451340255,
"grad_norm": 5.53125,
"learning_rate": 1.9987314661214904e-05,
"loss": 0.203,
"step": 1820
},
{
"epoch": 1.1783788134911053,
"grad_norm": 20.5,
"learning_rate": 1.9726685606170415e-05,
"loss": 0.2571,
"step": 1830
},
{
"epoch": 1.1848184818481848,
"grad_norm": 8.25,
"learning_rate": 1.9466655003725273e-05,
"loss": 0.2108,
"step": 1840
},
{
"epoch": 1.1912581502052644,
"grad_norm": 10.375,
"learning_rate": 1.920725236397091e-05,
"loss": 0.2667,
"step": 1850
},
{
"epoch": 1.197697818562344,
"grad_norm": 5.71875,
"learning_rate": 1.8948507125733177e-05,
"loss": 0.1615,
"step": 1860
},
{
"epoch": 1.2041374869194237,
"grad_norm": 5.5625,
"learning_rate": 1.869044865323138e-05,
"loss": 0.2647,
"step": 1870
},
{
"epoch": 1.2105771552765032,
"grad_norm": 5.90625,
"learning_rate": 1.843310623274587e-05,
"loss": 0.1771,
"step": 1880
},
{
"epoch": 1.2170168236335828,
"grad_norm": 5.28125,
"learning_rate": 1.8176509069294396e-05,
"loss": 0.1781,
"step": 1890
},
{
"epoch": 1.2234564919906625,
"grad_norm": 6.78125,
"learning_rate": 1.7920686283317712e-05,
"loss": 0.167,
"step": 1900
},
{
"epoch": 1.229896160347742,
"grad_norm": 8.0625,
"learning_rate": 1.7665666907374822e-05,
"loss": 0.1785,
"step": 1910
},
{
"epoch": 1.2363358287048216,
"grad_norm": 6.71875,
"learning_rate": 1.741147988284813e-05,
"loss": 0.1816,
"step": 1920
},
{
"epoch": 1.2427754970619014,
"grad_norm": 11.25,
"learning_rate": 1.715815405665897e-05,
"loss": 0.1934,
"step": 1930
},
{
"epoch": 1.249215165418981,
"grad_norm": 8.125,
"learning_rate": 1.6905718177993897e-05,
"loss": 0.1683,
"step": 1940
},
{
"epoch": 1.2556548337760605,
"grad_norm": 8.4375,
"learning_rate": 1.6654200895041962e-05,
"loss": 0.1812,
"step": 1950
},
{
"epoch": 1.2620945021331402,
"grad_norm": 7.25,
"learning_rate": 1.6403630751743576e-05,
"loss": 0.1787,
"step": 1960
},
{
"epoch": 1.2685341704902198,
"grad_norm": 6.0625,
"learning_rate": 1.6154036184551098e-05,
"loss": 0.1483,
"step": 1970
},
{
"epoch": 1.2749738388472993,
"grad_norm": 9.0,
"learning_rate": 1.5905445519201694e-05,
"loss": 0.3355,
"step": 1980
},
{
"epoch": 1.281413507204379,
"grad_norm": 9.0,
"learning_rate": 1.5657886967502733e-05,
"loss": 0.3235,
"step": 1990
},
{
"epoch": 1.2878531755614586,
"grad_norm": 4.15625,
"learning_rate": 1.541138862413009e-05,
"loss": 0.1555,
"step": 2000
},
{
"epoch": 1.2942928439185382,
"grad_norm": 2.5,
"learning_rate": 1.5165978463439823e-05,
"loss": 0.129,
"step": 2010
},
{
"epoch": 1.300732512275618,
"grad_norm": 7.9375,
"learning_rate": 1.4921684336293367e-05,
"loss": 0.1734,
"step": 2020
},
{
"epoch": 1.3071721806326975,
"grad_norm": 6.53125,
"learning_rate": 1.467853396689688e-05,
"loss": 0.1907,
"step": 2030
},
{
"epoch": 1.313611848989777,
"grad_norm": 11.125,
"learning_rate": 1.4436554949654906e-05,
"loss": 0.139,
"step": 2040
},
{
"epoch": 1.3200515173468568,
"grad_norm": 4.5625,
"learning_rate": 1.4195774746038723e-05,
"loss": 0.1053,
"step": 2050
},
{
"epoch": 1.3264911857039363,
"grad_norm": 6.40625,
"learning_rate": 1.3956220681469866e-05,
"loss": 0.1758,
"step": 2060
},
{
"epoch": 1.3329308540610159,
"grad_norm": 5.40625,
"learning_rate": 1.3717919942219067e-05,
"loss": 0.1186,
"step": 2070
},
{
"epoch": 1.3393705224180954,
"grad_norm": 9.5625,
"learning_rate": 1.3480899572320893e-05,
"loss": 0.1283,
"step": 2080
},
{
"epoch": 1.345810190775175,
"grad_norm": 7.375,
"learning_rate": 1.3245186470504647e-05,
"loss": 0.1294,
"step": 2090
},
{
"epoch": 1.3522498591322547,
"grad_norm": 8.875,
"learning_rate": 1.3010807387141738e-05,
"loss": 0.1039,
"step": 2100
},
{
"epoch": 1.3586895274893342,
"grad_norm": 9.75,
"learning_rate": 1.2777788921209805e-05,
"loss": 0.1128,
"step": 2110
},
{
"epoch": 1.3651291958464138,
"grad_norm": 4.4375,
"learning_rate": 1.25461575172741e-05,
"loss": 0.1235,
"step": 2120
},
{
"epoch": 1.3715688642034936,
"grad_norm": 9.1875,
"learning_rate": 1.2315939462486395e-05,
"loss": 0.1536,
"step": 2130
},
{
"epoch": 1.378008532560573,
"grad_norm": 4.75,
"learning_rate": 1.2087160883601695e-05,
"loss": 0.1011,
"step": 2140
},
{
"epoch": 1.3844482009176526,
"grad_norm": 7.0,
"learning_rate": 1.185984774401325e-05,
"loss": 0.1795,
"step": 2150
},
{
"epoch": 1.3908878692747324,
"grad_norm": 4.25,
"learning_rate": 1.163402584080597e-05,
"loss": 0.089,
"step": 2160
},
{
"epoch": 1.397327537631812,
"grad_norm": 6.28125,
"learning_rate": 1.1409720801828849e-05,
"loss": 0.1405,
"step": 2170
},
{
"epoch": 1.4037672059888915,
"grad_norm": 2.84375,
"learning_rate": 1.1186958082786517e-05,
"loss": 0.1533,
"step": 2180
},
{
"epoch": 1.4102068743459713,
"grad_norm": 4.75,
"learning_rate": 1.096576296435034e-05,
"loss": 0.1359,
"step": 2190
},
{
"epoch": 1.4166465427030508,
"grad_norm": 13.875,
"learning_rate": 1.0746160549289424e-05,
"loss": 0.1046,
"step": 2200
},
{
"epoch": 1.4230862110601303,
"grad_norm": 5.375,
"learning_rate": 1.0528175759621727e-05,
"loss": 0.1285,
"step": 2210
},
{
"epoch": 1.42952587941721,
"grad_norm": 4.375,
"learning_rate": 1.0311833333785807e-05,
"loss": 0.163,
"step": 2220
},
{
"epoch": 1.4359655477742896,
"grad_norm": 7.3125,
"learning_rate": 1.0097157823833248e-05,
"loss": 0.2187,
"step": 2230
},
{
"epoch": 1.4424052161313692,
"grad_norm": 5.9375,
"learning_rate": 9.884173592642374e-06,
"loss": 0.1137,
"step": 2240
},
{
"epoch": 1.448844884488449,
"grad_norm": 12.625,
"learning_rate": 9.672904811153363e-06,
"loss": 0.1985,
"step": 2250
},
{
"epoch": 1.4552845528455285,
"grad_norm": 3.96875,
"learning_rate": 9.463375455625183e-06,
"loss": 0.0877,
"step": 2260
},
{
"epoch": 1.461724221202608,
"grad_norm": 5.5625,
"learning_rate": 9.255609304914556e-06,
"loss": 0.1086,
"step": 2270
},
{
"epoch": 1.4681638895596878,
"grad_norm": 3.0,
"learning_rate": 9.049629937777382e-06,
"loss": 0.1154,
"step": 2280
},
{
"epoch": 1.4746035579167673,
"grad_norm": 3.671875,
"learning_rate": 8.8454607301929e-06,
"loss": 0.1317,
"step": 2290
},
{
"epoch": 1.4810432262738469,
"grad_norm": 4.5625,
"learning_rate": 8.643124852710749e-06,
"loss": 0.0975,
"step": 2300
},
{
"epoch": 1.4874828946309266,
"grad_norm": 6.28125,
"learning_rate": 8.442645267821458e-06,
"loss": 0.1085,
"step": 2310
},
{
"epoch": 1.4939225629880062,
"grad_norm": 4.84375,
"learning_rate": 8.244044727350492e-06,
"loss": 0.1105,
"step": 2320
},
{
"epoch": 1.5003622313450857,
"grad_norm": 4.5625,
"learning_rate": 8.047345769876204e-06,
"loss": 0.1523,
"step": 2330
},
{
"epoch": 1.5068018997021655,
"grad_norm": 12.5,
"learning_rate": 7.85257071817202e-06,
"loss": 0.1637,
"step": 2340
},
{
"epoch": 1.5132415680592448,
"grad_norm": 4.125,
"learning_rate": 7.659741676673066e-06,
"loss": 0.1124,
"step": 2350
},
{
"epoch": 1.5196812364163246,
"grad_norm": 9.0,
"learning_rate": 7.468880528967651e-06,
"loss": 0.0965,
"step": 2360
},
{
"epoch": 1.5261209047734041,
"grad_norm": 5.53125,
"learning_rate": 7.280008935313701e-06,
"loss": 0.0999,
"step": 2370
},
{
"epoch": 1.5325605731304837,
"grad_norm": 3.578125,
"learning_rate": 7.093148330180649e-06,
"loss": 0.1199,
"step": 2380
},
{
"epoch": 1.5390002414875634,
"grad_norm": 4.71875,
"learning_rate": 6.908319919816877e-06,
"loss": 0.1595,
"step": 2390
},
{
"epoch": 1.545439909844643,
"grad_norm": 5.4375,
"learning_rate": 6.725544679843101e-06,
"loss": 0.0966,
"step": 2400
},
{
"epoch": 1.5518795782017225,
"grad_norm": 2.765625,
"learning_rate": 6.5448433528718876e-06,
"loss": 0.0993,
"step": 2410
},
{
"epoch": 1.5583192465588023,
"grad_norm": 7.875,
"learning_rate": 6.3662364461536725e-06,
"loss": 0.1303,
"step": 2420
},
{
"epoch": 1.5647589149158818,
"grad_norm": 4.625,
"learning_rate": 6.189744229249406e-06,
"loss": 0.1044,
"step": 2430
},
{
"epoch": 1.5711985832729614,
"grad_norm": 5.15625,
"learning_rate": 6.015386731730277e-06,
"loss": 0.1186,
"step": 2440
},
{
"epoch": 1.5776382516300411,
"grad_norm": 5.96875,
"learning_rate": 5.843183740904564e-06,
"loss": 0.1427,
"step": 2450
},
{
"epoch": 1.5840779199871207,
"grad_norm": 8.5625,
"learning_rate": 5.673154799572064e-06,
"loss": 0.0971,
"step": 2460
},
{
"epoch": 1.5905175883442002,
"grad_norm": 4.9375,
"learning_rate": 5.505319203806239e-06,
"loss": 0.0823,
"step": 2470
},
{
"epoch": 1.59695725670128,
"grad_norm": 8.625,
"learning_rate": 5.339696000764346e-06,
"loss": 0.0963,
"step": 2480
},
{
"epoch": 1.6033969250583595,
"grad_norm": 4.25,
"learning_rate": 5.17630398652584e-06,
"loss": 0.1075,
"step": 2490
},
{
"epoch": 1.609836593415439,
"grad_norm": 5.6875,
"learning_rate": 5.01516170395929e-06,
"loss": 0.0782,
"step": 2500
},
{
"epoch": 1.6162762617725188,
"grad_norm": 2.03125,
"learning_rate": 4.856287440617952e-06,
"loss": 0.0793,
"step": 2510
},
{
"epoch": 1.6227159301295984,
"grad_norm": 12.5625,
"learning_rate": 4.699699226664409e-06,
"loss": 0.1177,
"step": 2520
},
{
"epoch": 1.629155598486678,
"grad_norm": 4.5,
"learning_rate": 4.545414832824368e-06,
"loss": 0.0996,
"step": 2530
},
{
"epoch": 1.6355952668437577,
"grad_norm": 4.1875,
"learning_rate": 4.393451768369919e-06,
"loss": 0.119,
"step": 2540
},
{
"epoch": 1.6420349352008372,
"grad_norm": 4.625,
"learning_rate": 4.243827279132451e-06,
"loss": 0.0866,
"step": 2550
},
{
"epoch": 1.6484746035579168,
"grad_norm": 13.1875,
"learning_rate": 4.096558345545498e-06,
"loss": 0.1736,
"step": 2560
},
{
"epoch": 1.6549142719149965,
"grad_norm": 6.96875,
"learning_rate": 3.951661680717647e-06,
"loss": 0.0881,
"step": 2570
},
{
"epoch": 1.6613539402720758,
"grad_norm": 5.5625,
"learning_rate": 3.8091537285358413e-06,
"loss": 0.1013,
"step": 2580
},
{
"epoch": 1.6677936086291556,
"grad_norm": 7.25,
"learning_rate": 3.6690506617992004e-06,
"loss": 0.1896,
"step": 2590
},
{
"epoch": 1.6742332769862354,
"grad_norm": 3.78125,
"learning_rate": 3.531368380383637e-06,
"loss": 0.0874,
"step": 2600
},
{
"epoch": 1.6806729453433147,
"grad_norm": 4.28125,
"learning_rate": 3.396122509437402e-06,
"loss": 0.076,
"step": 2610
},
{
"epoch": 1.6871126137003944,
"grad_norm": 6.71875,
"learning_rate": 3.263328397607843e-06,
"loss": 0.1053,
"step": 2620
},
{
"epoch": 1.693552282057474,
"grad_norm": 3.203125,
"learning_rate": 3.1330011152995526e-06,
"loss": 0.1015,
"step": 2630
},
{
"epoch": 1.6999919504145535,
"grad_norm": 5.875,
"learning_rate": 3.0051554529640423e-06,
"loss": 0.1095,
"step": 2640
},
{
"epoch": 1.7064316187716333,
"grad_norm": 4.9375,
"learning_rate": 2.879805919421233e-06,
"loss": 0.0834,
"step": 2650
},
{
"epoch": 1.7128712871287128,
"grad_norm": 13.6875,
"learning_rate": 2.7569667402128963e-06,
"loss": 0.0952,
"step": 2660
},
{
"epoch": 1.7193109554857924,
"grad_norm": 5.6875,
"learning_rate": 2.6366518559882342e-06,
"loss": 0.1263,
"step": 2670
},
{
"epoch": 1.7257506238428721,
"grad_norm": 5.03125,
"learning_rate": 2.5188749209217986e-06,
"loss": 0.0844,
"step": 2680
},
{
"epoch": 1.7321902921999517,
"grad_norm": 3.71875,
"learning_rate": 2.4036493011639154e-06,
"loss": 0.0908,
"step": 2690
},
{
"epoch": 1.7386299605570312,
"grad_norm": 3.59375,
"learning_rate": 2.290988073323824e-06,
"loss": 0.0963,
"step": 2700
},
{
"epoch": 1.745069628914111,
"grad_norm": 3.953125,
"learning_rate": 2.1809040229856136e-06,
"loss": 0.0826,
"step": 2710
},
{
"epoch": 1.7515092972711905,
"grad_norm": 5.09375,
"learning_rate": 2.0734096432572514e-06,
"loss": 0.0921,
"step": 2720
},
{
"epoch": 1.75794896562827,
"grad_norm": 2.25,
"learning_rate": 1.9685171333527736e-06,
"loss": 0.2672,
"step": 2730
},
{
"epoch": 1.7643886339853498,
"grad_norm": 5.9375,
"learning_rate": 1.8662383972078285e-06,
"loss": 0.0808,
"step": 2740
},
{
"epoch": 1.7708283023424294,
"grad_norm": 8.125,
"learning_rate": 1.766585042128735e-06,
"loss": 0.0998,
"step": 2750
},
{
"epoch": 1.777267970699509,
"grad_norm": 4.6875,
"learning_rate": 1.6695683774752196e-06,
"loss": 0.1619,
"step": 2760
},
{
"epoch": 1.7837076390565887,
"grad_norm": 2.4375,
"learning_rate": 1.5751994133769238e-06,
"loss": 0.0764,
"step": 2770
},
{
"epoch": 1.7901473074136682,
"grad_norm": 4.9375,
"learning_rate": 1.4834888594839114e-06,
"loss": 0.0799,
"step": 2780
},
{
"epoch": 1.7965869757707478,
"grad_norm": 4.1875,
"learning_rate": 1.3944471237512774e-06,
"loss": 0.1941,
"step": 2790
},
{
"epoch": 1.8030266441278275,
"grad_norm": 4.1875,
"learning_rate": 1.308084311257954e-06,
"loss": 0.1054,
"step": 2800
},
{
"epoch": 1.809466312484907,
"grad_norm": 4.65625,
"learning_rate": 1.2244102230599297e-06,
"loss": 0.1428,
"step": 2810
},
{
"epoch": 1.8159059808419866,
"grad_norm": 4.6875,
"learning_rate": 1.143434355077952e-06,
"loss": 0.1656,
"step": 2820
},
{
"epoch": 1.8223456491990664,
"grad_norm": 5.28125,
"learning_rate": 1.0651658970198736e-06,
"loss": 0.0876,
"step": 2830
},
{
"epoch": 1.8287853175561457,
"grad_norm": 6.6875,
"learning_rate": 9.896137313377323e-07,
"loss": 0.1129,
"step": 2840
},
{
"epoch": 1.8352249859132255,
"grad_norm": 4.875,
"learning_rate": 9.167864322197073e-07,
"loss": 0.1034,
"step": 2850
},
{
"epoch": 1.8416646542703052,
"grad_norm": 3.921875,
"learning_rate": 8.466922646170578e-07,
"loss": 0.0829,
"step": 2860
},
{
"epoch": 1.8481043226273846,
"grad_norm": 4.0625,
"learning_rate": 7.793391833061686e-07,
"loss": 0.0985,
"step": 2870
},
{
"epoch": 1.8545439909844643,
"grad_norm": 5.8125,
"learning_rate": 7.147348319857805e-07,
"loss": 0.0725,
"step": 2880
},
{
"epoch": 1.8609836593415439,
"grad_norm": 4.8125,
"learning_rate": 6.528865424095254e-07,
"loss": 0.0686,
"step": 2890
},
{
"epoch": 1.8674233276986234,
"grad_norm": 7.625,
"learning_rate": 5.938013335538756e-07,
"loss": 0.0804,
"step": 2900
},
{
"epoch": 1.8738629960557032,
"grad_norm": 5.8125,
"learning_rate": 5.374859108215835e-07,
"loss": 0.0869,
"step": 2910
},
{
"epoch": 1.8803026644127827,
"grad_norm": 6.46875,
"learning_rate": 4.839466652806934e-07,
"loss": 0.1161,
"step": 2920
},
{
"epoch": 1.8867423327698623,
"grad_norm": 4.4375,
"learning_rate": 4.3318967293924096e-07,
"loss": 0.1037,
"step": 2930
},
{
"epoch": 1.893182001126942,
"grad_norm": 3.15625,
"learning_rate": 3.852206940557185e-07,
"loss": 0.0818,
"step": 2940
},
{
"epoch": 1.8996216694840216,
"grad_norm": 7.40625,
"learning_rate": 3.4004517248535373e-07,
"loss": 0.0894,
"step": 2950
},
{
"epoch": 1.906061337841101,
"grad_norm": 6.1875,
"learning_rate": 2.9766823506228725e-07,
"loss": 0.0989,
"step": 2960
},
{
"epoch": 1.9125010061981809,
"grad_norm": 4.96875,
"learning_rate": 2.5809469101776564e-07,
"loss": 0.0877,
"step": 2970
},
{
"epoch": 1.9189406745552604,
"grad_norm": 5.15625,
"learning_rate": 2.21329031434353e-07,
"loss": 0.1298,
"step": 2980
},
{
"epoch": 1.92538034291234,
"grad_norm": 1.765625,
"learning_rate": 1.8737542873623615e-07,
"loss": 0.0685,
"step": 2990
},
{
"epoch": 1.9318200112694197,
"grad_norm": 3.671875,
"learning_rate": 1.5623773621572523e-07,
"loss": 0.0739,
"step": 3000
},
{
"epoch": 1.9382596796264993,
"grad_norm": 6.40625,
"learning_rate": 1.2791948759594553e-07,
"loss": 0.0698,
"step": 3010
},
{
"epoch": 1.9446993479835788,
"grad_norm": 9.8125,
"learning_rate": 1.0242389662981078e-07,
"loss": 0.0931,
"step": 3020
},
{
"epoch": 1.9511390163406586,
"grad_norm": 5.6875,
"learning_rate": 7.975385673529844e-08,
"loss": 0.1264,
"step": 3030
},
{
"epoch": 1.957578684697738,
"grad_norm": 3.484375,
"learning_rate": 5.991194066709838e-08,
"loss": 0.0872,
"step": 3040
},
{
"epoch": 1.9640183530548176,
"grad_norm": 3.140625,
"learning_rate": 4.2900400224615835e-08,
"loss": 0.0666,
"step": 3050
},
{
"epoch": 1.9704580214118974,
"grad_norm": 5.03125,
"learning_rate": 2.872116599644259e-08,
"loss": 0.0956,
"step": 3060
},
{
"epoch": 1.976897689768977,
"grad_norm": 3.734375,
"learning_rate": 1.7375847141248846e-08,
"loss": 0.0917,
"step": 3070
},
{
"epoch": 1.9833373581260565,
"grad_norm": 9.375,
"learning_rate": 8.865731205168205e-09,
"loss": 0.1889,
"step": 3080
},
{
"epoch": 1.9897770264831363,
"grad_norm": 2.46875,
"learning_rate": 3.1917839756784307e-09,
"loss": 0.0837,
"step": 3090
},
{
"epoch": 1.9962166948402156,
"grad_norm": 4.28125,
"learning_rate": 3.546493719946975e-10,
"loss": 0.0915,
"step": 3100
}
],
"logging_steps": 10,
"max_steps": 3104,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"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": 2.4232838290224538e+17,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}