terry69's picture
Model save
6045a28 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 2415,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00041407867494824016,
"grad_norm": 25.48398351356511,
"learning_rate": 4.132231404958678e-08,
"loss": 1.4285,
"step": 1
},
{
"epoch": 0.002070393374741201,
"grad_norm": 23.634565130522244,
"learning_rate": 2.066115702479339e-07,
"loss": 1.4082,
"step": 5
},
{
"epoch": 0.004140786749482402,
"grad_norm": 15.268223520597987,
"learning_rate": 4.132231404958678e-07,
"loss": 1.3598,
"step": 10
},
{
"epoch": 0.006211180124223602,
"grad_norm": 8.33387680086649,
"learning_rate": 6.198347107438018e-07,
"loss": 1.2482,
"step": 15
},
{
"epoch": 0.008281573498964804,
"grad_norm": 10.65881740361,
"learning_rate": 8.264462809917356e-07,
"loss": 1.1385,
"step": 20
},
{
"epoch": 0.010351966873706004,
"grad_norm": 4.798283184854408,
"learning_rate": 1.0330578512396695e-06,
"loss": 1.025,
"step": 25
},
{
"epoch": 0.012422360248447204,
"grad_norm": 3.4050882247520438,
"learning_rate": 1.2396694214876035e-06,
"loss": 0.9491,
"step": 30
},
{
"epoch": 0.014492753623188406,
"grad_norm": 3.3131324457414197,
"learning_rate": 1.4462809917355372e-06,
"loss": 0.9402,
"step": 35
},
{
"epoch": 0.016563146997929608,
"grad_norm": 3.04248997117286,
"learning_rate": 1.6528925619834712e-06,
"loss": 0.9174,
"step": 40
},
{
"epoch": 0.018633540372670808,
"grad_norm": 3.2595288563535565,
"learning_rate": 1.859504132231405e-06,
"loss": 0.9048,
"step": 45
},
{
"epoch": 0.020703933747412008,
"grad_norm": 3.1946044332124526,
"learning_rate": 2.066115702479339e-06,
"loss": 0.9094,
"step": 50
},
{
"epoch": 0.022774327122153208,
"grad_norm": 2.9341337515961667,
"learning_rate": 2.2727272727272728e-06,
"loss": 0.9012,
"step": 55
},
{
"epoch": 0.024844720496894408,
"grad_norm": 3.2463091625906437,
"learning_rate": 2.479338842975207e-06,
"loss": 0.8822,
"step": 60
},
{
"epoch": 0.026915113871635612,
"grad_norm": 3.1629249082104702,
"learning_rate": 2.6859504132231405e-06,
"loss": 0.8837,
"step": 65
},
{
"epoch": 0.028985507246376812,
"grad_norm": 3.086547904362823,
"learning_rate": 2.8925619834710743e-06,
"loss": 0.8905,
"step": 70
},
{
"epoch": 0.031055900621118012,
"grad_norm": 3.076953490144745,
"learning_rate": 3.0991735537190086e-06,
"loss": 0.8751,
"step": 75
},
{
"epoch": 0.033126293995859216,
"grad_norm": 3.164660216307417,
"learning_rate": 3.3057851239669424e-06,
"loss": 0.8598,
"step": 80
},
{
"epoch": 0.035196687370600416,
"grad_norm": 3.0630037088555335,
"learning_rate": 3.5123966942148763e-06,
"loss": 0.8655,
"step": 85
},
{
"epoch": 0.037267080745341616,
"grad_norm": 2.921268821171763,
"learning_rate": 3.71900826446281e-06,
"loss": 0.849,
"step": 90
},
{
"epoch": 0.039337474120082816,
"grad_norm": 3.269353569725898,
"learning_rate": 3.925619834710744e-06,
"loss": 0.8543,
"step": 95
},
{
"epoch": 0.041407867494824016,
"grad_norm": 3.041956717819935,
"learning_rate": 4.132231404958678e-06,
"loss": 0.8452,
"step": 100
},
{
"epoch": 0.043478260869565216,
"grad_norm": 3.025520462247771,
"learning_rate": 4.338842975206612e-06,
"loss": 0.8364,
"step": 105
},
{
"epoch": 0.045548654244306416,
"grad_norm": 3.2019266158326425,
"learning_rate": 4.5454545454545455e-06,
"loss": 0.8486,
"step": 110
},
{
"epoch": 0.047619047619047616,
"grad_norm": 3.3314737381858235,
"learning_rate": 4.75206611570248e-06,
"loss": 0.8553,
"step": 115
},
{
"epoch": 0.049689440993788817,
"grad_norm": 3.1886929174900924,
"learning_rate": 4.958677685950414e-06,
"loss": 0.8373,
"step": 120
},
{
"epoch": 0.051759834368530024,
"grad_norm": 3.262410661922915,
"learning_rate": 5.165289256198347e-06,
"loss": 0.8354,
"step": 125
},
{
"epoch": 0.053830227743271224,
"grad_norm": 3.0012878653602133,
"learning_rate": 5.371900826446281e-06,
"loss": 0.8298,
"step": 130
},
{
"epoch": 0.055900621118012424,
"grad_norm": 3.207772081919496,
"learning_rate": 5.578512396694216e-06,
"loss": 0.8265,
"step": 135
},
{
"epoch": 0.057971014492753624,
"grad_norm": 2.9057159664531205,
"learning_rate": 5.785123966942149e-06,
"loss": 0.8542,
"step": 140
},
{
"epoch": 0.060041407867494824,
"grad_norm": 3.0332228429135797,
"learning_rate": 5.991735537190083e-06,
"loss": 0.8308,
"step": 145
},
{
"epoch": 0.062111801242236024,
"grad_norm": 3.157618831687654,
"learning_rate": 6.198347107438017e-06,
"loss": 0.8433,
"step": 150
},
{
"epoch": 0.06418219461697723,
"grad_norm": 3.4410763976889105,
"learning_rate": 6.404958677685951e-06,
"loss": 0.8281,
"step": 155
},
{
"epoch": 0.06625258799171843,
"grad_norm": 2.987274288267847,
"learning_rate": 6.611570247933885e-06,
"loss": 0.8113,
"step": 160
},
{
"epoch": 0.06832298136645963,
"grad_norm": 3.0989786327956512,
"learning_rate": 6.818181818181818e-06,
"loss": 0.8232,
"step": 165
},
{
"epoch": 0.07039337474120083,
"grad_norm": 3.0223369344557187,
"learning_rate": 7.0247933884297525e-06,
"loss": 0.8111,
"step": 170
},
{
"epoch": 0.07246376811594203,
"grad_norm": 3.0787523746730634,
"learning_rate": 7.231404958677687e-06,
"loss": 0.826,
"step": 175
},
{
"epoch": 0.07453416149068323,
"grad_norm": 3.0796301224865306,
"learning_rate": 7.43801652892562e-06,
"loss": 0.8006,
"step": 180
},
{
"epoch": 0.07660455486542443,
"grad_norm": 2.9934570197468204,
"learning_rate": 7.644628099173555e-06,
"loss": 0.8239,
"step": 185
},
{
"epoch": 0.07867494824016563,
"grad_norm": 3.053342444537995,
"learning_rate": 7.851239669421489e-06,
"loss": 0.8007,
"step": 190
},
{
"epoch": 0.08074534161490683,
"grad_norm": 3.0585363849661316,
"learning_rate": 8.057851239669421e-06,
"loss": 0.8212,
"step": 195
},
{
"epoch": 0.08281573498964803,
"grad_norm": 3.06780155535708,
"learning_rate": 8.264462809917356e-06,
"loss": 0.8243,
"step": 200
},
{
"epoch": 0.08488612836438923,
"grad_norm": 3.0294810469044866,
"learning_rate": 8.47107438016529e-06,
"loss": 0.8214,
"step": 205
},
{
"epoch": 0.08695652173913043,
"grad_norm": 3.189981572883342,
"learning_rate": 8.677685950413224e-06,
"loss": 0.8136,
"step": 210
},
{
"epoch": 0.08902691511387163,
"grad_norm": 2.969698067997633,
"learning_rate": 8.884297520661158e-06,
"loss": 0.8114,
"step": 215
},
{
"epoch": 0.09109730848861283,
"grad_norm": 3.005971493770243,
"learning_rate": 9.090909090909091e-06,
"loss": 0.8195,
"step": 220
},
{
"epoch": 0.09316770186335403,
"grad_norm": 5.537871207131585,
"learning_rate": 9.297520661157025e-06,
"loss": 0.8084,
"step": 225
},
{
"epoch": 0.09523809523809523,
"grad_norm": 3.3480322996327283,
"learning_rate": 9.50413223140496e-06,
"loss": 0.803,
"step": 230
},
{
"epoch": 0.09730848861283643,
"grad_norm": 3.129349449910694,
"learning_rate": 9.710743801652894e-06,
"loss": 0.8147,
"step": 235
},
{
"epoch": 0.09937888198757763,
"grad_norm": 3.134829156442627,
"learning_rate": 9.917355371900828e-06,
"loss": 0.7872,
"step": 240
},
{
"epoch": 0.10144927536231885,
"grad_norm": 3.0505826906934685,
"learning_rate": 9.999952971391373e-06,
"loss": 0.7796,
"step": 245
},
{
"epoch": 0.10351966873706005,
"grad_norm": 2.7288596400170135,
"learning_rate": 9.999665577542406e-06,
"loss": 0.8013,
"step": 250
},
{
"epoch": 0.10559006211180125,
"grad_norm": 4.615387300012688,
"learning_rate": 9.999116931848504e-06,
"loss": 0.7979,
"step": 255
},
{
"epoch": 0.10766045548654245,
"grad_norm": 3.151946299993893,
"learning_rate": 9.99830706297852e-06,
"loss": 0.8054,
"step": 260
},
{
"epoch": 0.10973084886128365,
"grad_norm": 3.165203386627693,
"learning_rate": 9.997236013251234e-06,
"loss": 0.7877,
"step": 265
},
{
"epoch": 0.11180124223602485,
"grad_norm": 3.134193615549586,
"learning_rate": 9.995903838633133e-06,
"loss": 0.7921,
"step": 270
},
{
"epoch": 0.11387163561076605,
"grad_norm": 3.123080650651501,
"learning_rate": 9.994310608735492e-06,
"loss": 0.7904,
"step": 275
},
{
"epoch": 0.11594202898550725,
"grad_norm": 3.0072109199490153,
"learning_rate": 9.99245640681072e-06,
"loss": 0.7829,
"step": 280
},
{
"epoch": 0.11801242236024845,
"grad_norm": 2.8989093329142963,
"learning_rate": 9.99034132974804e-06,
"loss": 0.7851,
"step": 285
},
{
"epoch": 0.12008281573498965,
"grad_norm": 2.8200476693657737,
"learning_rate": 9.987965488068398e-06,
"loss": 0.7958,
"step": 290
},
{
"epoch": 0.12215320910973085,
"grad_norm": 2.908655629752229,
"learning_rate": 9.985329005918702e-06,
"loss": 0.7799,
"step": 295
},
{
"epoch": 0.12422360248447205,
"grad_norm": 2.7843856702869507,
"learning_rate": 9.982432021065334e-06,
"loss": 0.7629,
"step": 300
},
{
"epoch": 0.12629399585921325,
"grad_norm": 2.8481441541172705,
"learning_rate": 9.979274684886943e-06,
"loss": 0.7917,
"step": 305
},
{
"epoch": 0.12836438923395446,
"grad_norm": 3.04542025762207,
"learning_rate": 9.975857162366547e-06,
"loss": 0.7804,
"step": 310
},
{
"epoch": 0.13043478260869565,
"grad_norm": 2.742499960287768,
"learning_rate": 9.972179632082899e-06,
"loss": 0.7848,
"step": 315
},
{
"epoch": 0.13250517598343686,
"grad_norm": 2.797474350667642,
"learning_rate": 9.968242286201171e-06,
"loss": 0.7763,
"step": 320
},
{
"epoch": 0.13457556935817805,
"grad_norm": 2.807858326433423,
"learning_rate": 9.964045330462896e-06,
"loss": 0.7691,
"step": 325
},
{
"epoch": 0.13664596273291926,
"grad_norm": 2.9394568611848437,
"learning_rate": 9.959588984175228e-06,
"loss": 0.747,
"step": 330
},
{
"epoch": 0.13871635610766045,
"grad_norm": 2.762412597751286,
"learning_rate": 9.954873480199483e-06,
"loss": 0.7503,
"step": 335
},
{
"epoch": 0.14078674948240166,
"grad_norm": 2.852641471166556,
"learning_rate": 9.94989906493896e-06,
"loss": 0.7533,
"step": 340
},
{
"epoch": 0.14285714285714285,
"grad_norm": 2.7404614960141433,
"learning_rate": 9.94466599832608e-06,
"loss": 0.7567,
"step": 345
},
{
"epoch": 0.14492753623188406,
"grad_norm": 2.8172847569820916,
"learning_rate": 9.939174553808793e-06,
"loss": 0.7537,
"step": 350
},
{
"epoch": 0.14699792960662525,
"grad_norm": 2.906331041232371,
"learning_rate": 9.933425018336292e-06,
"loss": 0.7383,
"step": 355
},
{
"epoch": 0.14906832298136646,
"grad_norm": 2.914371730456789,
"learning_rate": 9.927417692344025e-06,
"loss": 0.7521,
"step": 360
},
{
"epoch": 0.15113871635610765,
"grad_norm": 2.9757616427699163,
"learning_rate": 9.921152889737985e-06,
"loss": 0.763,
"step": 365
},
{
"epoch": 0.15320910973084886,
"grad_norm": 2.755878928013863,
"learning_rate": 9.914630937878315e-06,
"loss": 0.7513,
"step": 370
},
{
"epoch": 0.15527950310559005,
"grad_norm": 2.83127691695805,
"learning_rate": 9.907852177562201e-06,
"loss": 0.7417,
"step": 375
},
{
"epoch": 0.15734989648033126,
"grad_norm": 2.7370551817714937,
"learning_rate": 9.900816963006063e-06,
"loss": 0.7435,
"step": 380
},
{
"epoch": 0.15942028985507245,
"grad_norm": 2.9184370004367506,
"learning_rate": 9.893525661827043e-06,
"loss": 0.7254,
"step": 385
},
{
"epoch": 0.16149068322981366,
"grad_norm": 3.1457402689401337,
"learning_rate": 9.885978655023805e-06,
"loss": 0.738,
"step": 390
},
{
"epoch": 0.16356107660455488,
"grad_norm": 2.886300460794287,
"learning_rate": 9.87817633695661e-06,
"loss": 0.7261,
"step": 395
},
{
"epoch": 0.16563146997929606,
"grad_norm": 2.7768908729707866,
"learning_rate": 9.87011911532673e-06,
"loss": 0.7271,
"step": 400
},
{
"epoch": 0.16770186335403728,
"grad_norm": 2.8134670378613555,
"learning_rate": 9.861807411155126e-06,
"loss": 0.7212,
"step": 405
},
{
"epoch": 0.16977225672877846,
"grad_norm": 3.240613649618056,
"learning_rate": 9.853241658760457e-06,
"loss": 0.7142,
"step": 410
},
{
"epoch": 0.17184265010351968,
"grad_norm": 2.9010186853108793,
"learning_rate": 9.844422305736383e-06,
"loss": 0.7083,
"step": 415
},
{
"epoch": 0.17391304347826086,
"grad_norm": 2.7263222245103402,
"learning_rate": 9.835349812928178e-06,
"loss": 0.7166,
"step": 420
},
{
"epoch": 0.17598343685300208,
"grad_norm": 2.699501803535939,
"learning_rate": 9.826024654408645e-06,
"loss": 0.7249,
"step": 425
},
{
"epoch": 0.17805383022774326,
"grad_norm": 2.677451080059726,
"learning_rate": 9.816447317453353e-06,
"loss": 0.7055,
"step": 430
},
{
"epoch": 0.18012422360248448,
"grad_norm": 2.7704923962373633,
"learning_rate": 9.80661830251516e-06,
"loss": 0.7302,
"step": 435
},
{
"epoch": 0.18219461697722567,
"grad_norm": 3.0782524005334144,
"learning_rate": 9.796538123198077e-06,
"loss": 0.7135,
"step": 440
},
{
"epoch": 0.18426501035196688,
"grad_norm": 2.8689303081488906,
"learning_rate": 9.786207306230422e-06,
"loss": 0.7045,
"step": 445
},
{
"epoch": 0.18633540372670807,
"grad_norm": 2.7057146782800214,
"learning_rate": 9.775626391437303e-06,
"loss": 0.7073,
"step": 450
},
{
"epoch": 0.18840579710144928,
"grad_norm": 2.7761506153361704,
"learning_rate": 9.764795931712396e-06,
"loss": 0.7106,
"step": 455
},
{
"epoch": 0.19047619047619047,
"grad_norm": 2.75024991673497,
"learning_rate": 9.753716492989076e-06,
"loss": 0.6945,
"step": 460
},
{
"epoch": 0.19254658385093168,
"grad_norm": 2.711751260405265,
"learning_rate": 9.742388654210822e-06,
"loss": 0.6891,
"step": 465
},
{
"epoch": 0.19461697722567287,
"grad_norm": 2.74776351432592,
"learning_rate": 9.730813007300984e-06,
"loss": 0.6927,
"step": 470
},
{
"epoch": 0.19668737060041408,
"grad_norm": 2.930524598389639,
"learning_rate": 9.718990157131841e-06,
"loss": 0.6837,
"step": 475
},
{
"epoch": 0.19875776397515527,
"grad_norm": 2.663427028498334,
"learning_rate": 9.706920721492995e-06,
"loss": 0.6921,
"step": 480
},
{
"epoch": 0.20082815734989648,
"grad_norm": 2.946349388670218,
"learning_rate": 9.694605331059094e-06,
"loss": 0.7024,
"step": 485
},
{
"epoch": 0.2028985507246377,
"grad_norm": 3.074051605057056,
"learning_rate": 9.682044629356874e-06,
"loss": 0.6989,
"step": 490
},
{
"epoch": 0.20496894409937888,
"grad_norm": 3.2483495395680895,
"learning_rate": 9.669239272731538e-06,
"loss": 0.6998,
"step": 495
},
{
"epoch": 0.2070393374741201,
"grad_norm": 2.7724751765124704,
"learning_rate": 9.656189930312443e-06,
"loss": 0.707,
"step": 500
},
{
"epoch": 0.20910973084886128,
"grad_norm": 2.733330862817159,
"learning_rate": 9.642897283978157e-06,
"loss": 0.6886,
"step": 505
},
{
"epoch": 0.2111801242236025,
"grad_norm": 2.7359051389317934,
"learning_rate": 9.629362028320808e-06,
"loss": 0.6872,
"step": 510
},
{
"epoch": 0.21325051759834368,
"grad_norm": 2.7753346003774837,
"learning_rate": 9.615584870609809e-06,
"loss": 0.6608,
"step": 515
},
{
"epoch": 0.2153209109730849,
"grad_norm": 2.9308393451947548,
"learning_rate": 9.601566530754882e-06,
"loss": 0.672,
"step": 520
},
{
"epoch": 0.21739130434782608,
"grad_norm": 2.6456013984132203,
"learning_rate": 9.587307741268452e-06,
"loss": 0.67,
"step": 525
},
{
"epoch": 0.2194616977225673,
"grad_norm": 2.8187605300094885,
"learning_rate": 9.572809247227366e-06,
"loss": 0.6847,
"step": 530
},
{
"epoch": 0.22153209109730848,
"grad_norm": 2.582586417504312,
"learning_rate": 9.558071806233955e-06,
"loss": 0.6975,
"step": 535
},
{
"epoch": 0.2236024844720497,
"grad_norm": 2.837996829252072,
"learning_rate": 9.54309618837646e-06,
"loss": 0.6723,
"step": 540
},
{
"epoch": 0.22567287784679088,
"grad_norm": 2.5260261905707995,
"learning_rate": 9.52788317618878e-06,
"loss": 0.662,
"step": 545
},
{
"epoch": 0.2277432712215321,
"grad_norm": 2.758202910230446,
"learning_rate": 9.512433564609578e-06,
"loss": 0.6509,
"step": 550
},
{
"epoch": 0.22981366459627328,
"grad_norm": 2.7395905119594963,
"learning_rate": 9.496748160940762e-06,
"loss": 0.6665,
"step": 555
},
{
"epoch": 0.2318840579710145,
"grad_norm": 3.209701454856102,
"learning_rate": 9.480827784805278e-06,
"loss": 0.6509,
"step": 560
},
{
"epoch": 0.23395445134575568,
"grad_norm": 2.8037458956866548,
"learning_rate": 9.464673268104299e-06,
"loss": 0.6495,
"step": 565
},
{
"epoch": 0.2360248447204969,
"grad_norm": 3.291940633880938,
"learning_rate": 9.448285454973739e-06,
"loss": 0.6559,
"step": 570
},
{
"epoch": 0.23809523809523808,
"grad_norm": 2.9106054638621788,
"learning_rate": 9.431665201740154e-06,
"loss": 0.6715,
"step": 575
},
{
"epoch": 0.2401656314699793,
"grad_norm": 2.7543839474692473,
"learning_rate": 9.414813376876003e-06,
"loss": 0.6541,
"step": 580
},
{
"epoch": 0.2422360248447205,
"grad_norm": 2.546528761748931,
"learning_rate": 9.397730860954242e-06,
"loss": 0.6455,
"step": 585
},
{
"epoch": 0.2443064182194617,
"grad_norm": 2.8640205697823955,
"learning_rate": 9.38041854660234e-06,
"loss": 0.6378,
"step": 590
},
{
"epoch": 0.2463768115942029,
"grad_norm": 2.7615551478506943,
"learning_rate": 9.362877338455611e-06,
"loss": 0.6247,
"step": 595
},
{
"epoch": 0.2484472049689441,
"grad_norm": 2.7349208406994814,
"learning_rate": 9.345108153109963e-06,
"loss": 0.6457,
"step": 600
},
{
"epoch": 0.2505175983436853,
"grad_norm": 2.65857562232247,
"learning_rate": 9.327111919073988e-06,
"loss": 0.6472,
"step": 605
},
{
"epoch": 0.2525879917184265,
"grad_norm": 2.6711842868048157,
"learning_rate": 9.308889576720453e-06,
"loss": 0.6359,
"step": 610
},
{
"epoch": 0.2546583850931677,
"grad_norm": 2.806256632319445,
"learning_rate": 9.290442078237154e-06,
"loss": 0.6349,
"step": 615
},
{
"epoch": 0.2567287784679089,
"grad_norm": 2.7211278153243903,
"learning_rate": 9.271770387577168e-06,
"loss": 0.6296,
"step": 620
},
{
"epoch": 0.2587991718426501,
"grad_norm": 2.698035662303335,
"learning_rate": 9.252875480408479e-06,
"loss": 0.6494,
"step": 625
},
{
"epoch": 0.2608695652173913,
"grad_norm": 2.6742482278589295,
"learning_rate": 9.233758344062996e-06,
"loss": 0.6181,
"step": 630
},
{
"epoch": 0.2629399585921325,
"grad_norm": 2.584886493729489,
"learning_rate": 9.21441997748496e-06,
"loss": 0.6227,
"step": 635
},
{
"epoch": 0.2650103519668737,
"grad_norm": 2.5986395063169097,
"learning_rate": 9.194861391178749e-06,
"loss": 0.6183,
"step": 640
},
{
"epoch": 0.2670807453416149,
"grad_norm": 2.730659443072697,
"learning_rate": 9.175083607156067e-06,
"loss": 0.6384,
"step": 645
},
{
"epoch": 0.2691511387163561,
"grad_norm": 2.6656425234132115,
"learning_rate": 9.155087658882555e-06,
"loss": 0.625,
"step": 650
},
{
"epoch": 0.2712215320910973,
"grad_norm": 2.719633309943985,
"learning_rate": 9.134874591223773e-06,
"loss": 0.6489,
"step": 655
},
{
"epoch": 0.2732919254658385,
"grad_norm": 2.562172351229719,
"learning_rate": 9.114445460390605e-06,
"loss": 0.6168,
"step": 660
},
{
"epoch": 0.2753623188405797,
"grad_norm": 2.79458189618646,
"learning_rate": 9.093801333884076e-06,
"loss": 0.622,
"step": 665
},
{
"epoch": 0.2774327122153209,
"grad_norm": 2.9446754093469414,
"learning_rate": 9.072943290439566e-06,
"loss": 0.6329,
"step": 670
},
{
"epoch": 0.2795031055900621,
"grad_norm": 2.669323125514949,
"learning_rate": 9.051872419970439e-06,
"loss": 0.6275,
"step": 675
},
{
"epoch": 0.2815734989648033,
"grad_norm": 2.7534932048487253,
"learning_rate": 9.03058982351109e-06,
"loss": 0.6222,
"step": 680
},
{
"epoch": 0.2836438923395445,
"grad_norm": 2.739073018134251,
"learning_rate": 9.009096613159426e-06,
"loss": 0.6271,
"step": 685
},
{
"epoch": 0.2857142857142857,
"grad_norm": 2.6395459712038045,
"learning_rate": 8.987393912018732e-06,
"loss": 0.5984,
"step": 690
},
{
"epoch": 0.28778467908902694,
"grad_norm": 2.8269084788441496,
"learning_rate": 8.965482854139003e-06,
"loss": 0.6261,
"step": 695
},
{
"epoch": 0.2898550724637681,
"grad_norm": 2.7977196897678462,
"learning_rate": 8.943364584457675e-06,
"loss": 0.6088,
"step": 700
},
{
"epoch": 0.2919254658385093,
"grad_norm": 2.9923642118440945,
"learning_rate": 8.921040258739804e-06,
"loss": 0.6022,
"step": 705
},
{
"epoch": 0.2939958592132505,
"grad_norm": 2.627437601273443,
"learning_rate": 8.898511043517668e-06,
"loss": 0.6219,
"step": 710
},
{
"epoch": 0.29606625258799174,
"grad_norm": 2.7868086520844555,
"learning_rate": 8.875778116029816e-06,
"loss": 0.5918,
"step": 715
},
{
"epoch": 0.2981366459627329,
"grad_norm": 2.606031020009266,
"learning_rate": 8.85284266415955e-06,
"loss": 0.6065,
"step": 720
},
{
"epoch": 0.3002070393374741,
"grad_norm": 2.750031567522505,
"learning_rate": 8.829705886372845e-06,
"loss": 0.5911,
"step": 725
},
{
"epoch": 0.3022774327122153,
"grad_norm": 2.7966252505773217,
"learning_rate": 8.806368991655747e-06,
"loss": 0.5894,
"step": 730
},
{
"epoch": 0.30434782608695654,
"grad_norm": 2.6229941778509445,
"learning_rate": 8.782833199451177e-06,
"loss": 0.5856,
"step": 735
},
{
"epoch": 0.3064182194616977,
"grad_norm": 2.8015988683663418,
"learning_rate": 8.759099739595215e-06,
"loss": 0.5912,
"step": 740
},
{
"epoch": 0.3084886128364389,
"grad_norm": 2.625885402915092,
"learning_rate": 8.735169852252848e-06,
"loss": 0.5804,
"step": 745
},
{
"epoch": 0.3105590062111801,
"grad_norm": 2.540350772295356,
"learning_rate": 8.71104478785315e-06,
"loss": 0.5926,
"step": 750
},
{
"epoch": 0.31262939958592134,
"grad_norm": 2.564374062239284,
"learning_rate": 8.686725807023955e-06,
"loss": 0.5971,
"step": 755
},
{
"epoch": 0.3146997929606625,
"grad_norm": 2.715740615089575,
"learning_rate": 8.662214180525982e-06,
"loss": 0.5913,
"step": 760
},
{
"epoch": 0.3167701863354037,
"grad_norm": 2.8185377445047357,
"learning_rate": 8.637511189186425e-06,
"loss": 0.5847,
"step": 765
},
{
"epoch": 0.3188405797101449,
"grad_norm": 2.863541643896997,
"learning_rate": 8.612618123832033e-06,
"loss": 0.5701,
"step": 770
},
{
"epoch": 0.32091097308488614,
"grad_norm": 2.738443786397783,
"learning_rate": 8.587536285221656e-06,
"loss": 0.5856,
"step": 775
},
{
"epoch": 0.32298136645962733,
"grad_norm": 2.5499957989100035,
"learning_rate": 8.562266983978278e-06,
"loss": 0.575,
"step": 780
},
{
"epoch": 0.3250517598343685,
"grad_norm": 2.8607463866712304,
"learning_rate": 8.536811540520529e-06,
"loss": 0.603,
"step": 785
},
{
"epoch": 0.32712215320910976,
"grad_norm": 2.6831325082403246,
"learning_rate": 8.511171284993686e-06,
"loss": 0.5623,
"step": 790
},
{
"epoch": 0.32919254658385094,
"grad_norm": 2.78324604062969,
"learning_rate": 8.485347557200177e-06,
"loss": 0.5831,
"step": 795
},
{
"epoch": 0.33126293995859213,
"grad_norm": 2.6518381005152767,
"learning_rate": 8.459341706529557e-06,
"loss": 0.5835,
"step": 800
},
{
"epoch": 0.3333333333333333,
"grad_norm": 2.526247556926589,
"learning_rate": 8.43315509188801e-06,
"loss": 0.5693,
"step": 805
},
{
"epoch": 0.33540372670807456,
"grad_norm": 2.548647817128937,
"learning_rate": 8.406789081627337e-06,
"loss": 0.5526,
"step": 810
},
{
"epoch": 0.33747412008281574,
"grad_norm": 2.7267333364156685,
"learning_rate": 8.380245053473452e-06,
"loss": 0.5789,
"step": 815
},
{
"epoch": 0.33954451345755693,
"grad_norm": 2.556948662121955,
"learning_rate": 8.353524394454388e-06,
"loss": 0.5759,
"step": 820
},
{
"epoch": 0.3416149068322981,
"grad_norm": 2.5311685618660063,
"learning_rate": 8.326628500827826e-06,
"loss": 0.5805,
"step": 825
},
{
"epoch": 0.34368530020703936,
"grad_norm": 2.5946584747576362,
"learning_rate": 8.299558778008137e-06,
"loss": 0.563,
"step": 830
},
{
"epoch": 0.34575569358178054,
"grad_norm": 2.646570817849741,
"learning_rate": 8.272316640492932e-06,
"loss": 0.5678,
"step": 835
},
{
"epoch": 0.34782608695652173,
"grad_norm": 2.821727892284469,
"learning_rate": 8.244903511789158e-06,
"loss": 0.5761,
"step": 840
},
{
"epoch": 0.3498964803312629,
"grad_norm": 2.7282522742752082,
"learning_rate": 8.217320824338713e-06,
"loss": 0.5651,
"step": 845
},
{
"epoch": 0.35196687370600416,
"grad_norm": 2.7696477607458125,
"learning_rate": 8.189570019443597e-06,
"loss": 0.5424,
"step": 850
},
{
"epoch": 0.35403726708074534,
"grad_norm": 2.6050036703951136,
"learning_rate": 8.161652547190593e-06,
"loss": 0.5539,
"step": 855
},
{
"epoch": 0.35610766045548653,
"grad_norm": 2.4695019509374543,
"learning_rate": 8.133569866375497e-06,
"loss": 0.5528,
"step": 860
},
{
"epoch": 0.3581780538302277,
"grad_norm": 2.470537409860689,
"learning_rate": 8.105323444426891e-06,
"loss": 0.5439,
"step": 865
},
{
"epoch": 0.36024844720496896,
"grad_norm": 2.7767736485208787,
"learning_rate": 8.076914757329467e-06,
"loss": 0.5517,
"step": 870
},
{
"epoch": 0.36231884057971014,
"grad_norm": 2.560603862220981,
"learning_rate": 8.048345289546895e-06,
"loss": 0.5443,
"step": 875
},
{
"epoch": 0.36438923395445133,
"grad_norm": 2.7353218383955413,
"learning_rate": 8.01961653394426e-06,
"loss": 0.5505,
"step": 880
},
{
"epoch": 0.36645962732919257,
"grad_norm": 2.7011596762796724,
"learning_rate": 7.990729991710046e-06,
"loss": 0.5356,
"step": 885
},
{
"epoch": 0.36853002070393376,
"grad_norm": 2.6971330973859424,
"learning_rate": 7.9616871722777e-06,
"loss": 0.5453,
"step": 890
},
{
"epoch": 0.37060041407867494,
"grad_norm": 2.649246125385984,
"learning_rate": 7.932489593246764e-06,
"loss": 0.5493,
"step": 895
},
{
"epoch": 0.37267080745341613,
"grad_norm": 2.549015737650051,
"learning_rate": 7.903138780303556e-06,
"loss": 0.5385,
"step": 900
},
{
"epoch": 0.3747412008281574,
"grad_norm": 2.480703168873863,
"learning_rate": 7.873636267141463e-06,
"loss": 0.5511,
"step": 905
},
{
"epoch": 0.37681159420289856,
"grad_norm": 2.6486663220810045,
"learning_rate": 7.843983595380793e-06,
"loss": 0.5352,
"step": 910
},
{
"epoch": 0.37888198757763975,
"grad_norm": 2.613493626929314,
"learning_rate": 7.814182314488225e-06,
"loss": 0.5156,
"step": 915
},
{
"epoch": 0.38095238095238093,
"grad_norm": 2.6464127187191666,
"learning_rate": 7.784233981695835e-06,
"loss": 0.5439,
"step": 920
},
{
"epoch": 0.3830227743271222,
"grad_norm": 2.55830268587909,
"learning_rate": 7.754140161919732e-06,
"loss": 0.5392,
"step": 925
},
{
"epoch": 0.38509316770186336,
"grad_norm": 2.656861734311093,
"learning_rate": 7.72390242767828e-06,
"loss": 0.534,
"step": 930
},
{
"epoch": 0.38716356107660455,
"grad_norm": 2.5258339890051946,
"learning_rate": 7.693522359009931e-06,
"loss": 0.5162,
"step": 935
},
{
"epoch": 0.38923395445134573,
"grad_norm": 2.7625823646948153,
"learning_rate": 7.663001543390657e-06,
"loss": 0.5384,
"step": 940
},
{
"epoch": 0.391304347826087,
"grad_norm": 2.632702792337692,
"learning_rate": 7.63234157565101e-06,
"loss": 0.5297,
"step": 945
},
{
"epoch": 0.39337474120082816,
"grad_norm": 2.596119453009663,
"learning_rate": 7.601544057892769e-06,
"loss": 0.5262,
"step": 950
},
{
"epoch": 0.39544513457556935,
"grad_norm": 2.5571401053606104,
"learning_rate": 7.570610599405242e-06,
"loss": 0.5197,
"step": 955
},
{
"epoch": 0.39751552795031053,
"grad_norm": 2.701165663688961,
"learning_rate": 7.539542816581157e-06,
"loss": 0.5253,
"step": 960
},
{
"epoch": 0.3995859213250518,
"grad_norm": 2.7726491038816303,
"learning_rate": 7.508342332832213e-06,
"loss": 0.5227,
"step": 965
},
{
"epoch": 0.40165631469979296,
"grad_norm": 2.5386809718123278,
"learning_rate": 7.477010778504241e-06,
"loss": 0.5224,
"step": 970
},
{
"epoch": 0.40372670807453415,
"grad_norm": 2.4742367722323966,
"learning_rate": 7.445549790792021e-06,
"loss": 0.5203,
"step": 975
},
{
"epoch": 0.4057971014492754,
"grad_norm": 2.8506892960931274,
"learning_rate": 7.413961013653725e-06,
"loss": 0.5202,
"step": 980
},
{
"epoch": 0.4078674948240166,
"grad_norm": 2.6703829692626604,
"learning_rate": 7.3822460977250145e-06,
"loss": 0.5086,
"step": 985
},
{
"epoch": 0.40993788819875776,
"grad_norm": 2.9218133813167464,
"learning_rate": 7.350406700232794e-06,
"loss": 0.5331,
"step": 990
},
{
"epoch": 0.41200828157349895,
"grad_norm": 2.571377229519111,
"learning_rate": 7.318444484908606e-06,
"loss": 0.5081,
"step": 995
},
{
"epoch": 0.4140786749482402,
"grad_norm": 2.5673985663075785,
"learning_rate": 7.286361121901706e-06,
"loss": 0.4955,
"step": 1000
},
{
"epoch": 0.4161490683229814,
"grad_norm": 2.569124677451494,
"learning_rate": 7.254158287691775e-06,
"loss": 0.4957,
"step": 1005
},
{
"epoch": 0.41821946169772256,
"grad_norm": 2.6528729097787216,
"learning_rate": 7.221837665001335e-06,
"loss": 0.4919,
"step": 1010
},
{
"epoch": 0.42028985507246375,
"grad_norm": 2.63506957174704,
"learning_rate": 7.189400942707804e-06,
"loss": 0.4927,
"step": 1015
},
{
"epoch": 0.422360248447205,
"grad_norm": 2.5314116752795517,
"learning_rate": 7.1568498157552576e-06,
"loss": 0.5035,
"step": 1020
},
{
"epoch": 0.4244306418219462,
"grad_norm": 2.6039647624314086,
"learning_rate": 7.124185985065856e-06,
"loss": 0.5094,
"step": 1025
},
{
"epoch": 0.42650103519668736,
"grad_norm": 2.578643056036445,
"learning_rate": 7.091411157450965e-06,
"loss": 0.5045,
"step": 1030
},
{
"epoch": 0.42857142857142855,
"grad_norm": 2.600639040701803,
"learning_rate": 7.0585270455219654e-06,
"loss": 0.5097,
"step": 1035
},
{
"epoch": 0.4306418219461698,
"grad_norm": 2.5075344250253595,
"learning_rate": 7.025535367600771e-06,
"loss": 0.496,
"step": 1040
},
{
"epoch": 0.432712215320911,
"grad_norm": 2.57235419280248,
"learning_rate": 6.992437847630031e-06,
"loss": 0.4909,
"step": 1045
},
{
"epoch": 0.43478260869565216,
"grad_norm": 2.6861552032678517,
"learning_rate": 6.95923621508305e-06,
"loss": 0.4983,
"step": 1050
},
{
"epoch": 0.43685300207039335,
"grad_norm": 2.4950775602086104,
"learning_rate": 6.92593220487342e-06,
"loss": 0.4829,
"step": 1055
},
{
"epoch": 0.4389233954451346,
"grad_norm": 2.6103728536901527,
"learning_rate": 6.892527557264358e-06,
"loss": 0.4949,
"step": 1060
},
{
"epoch": 0.4409937888198758,
"grad_norm": 2.6118896191138483,
"learning_rate": 6.859024017777779e-06,
"loss": 0.5013,
"step": 1065
},
{
"epoch": 0.44306418219461696,
"grad_norm": 2.589122083174414,
"learning_rate": 6.825423337103074e-06,
"loss": 0.4869,
"step": 1070
},
{
"epoch": 0.4451345755693582,
"grad_norm": 2.5698623127213756,
"learning_rate": 6.791727271005642e-06,
"loss": 0.4823,
"step": 1075
},
{
"epoch": 0.4472049689440994,
"grad_norm": 2.54401757372677,
"learning_rate": 6.757937580235138e-06,
"loss": 0.4762,
"step": 1080
},
{
"epoch": 0.4492753623188406,
"grad_norm": 2.4698273066911915,
"learning_rate": 6.724056030433464e-06,
"loss": 0.4979,
"step": 1085
},
{
"epoch": 0.45134575569358176,
"grad_norm": 2.5498254174223853,
"learning_rate": 6.690084392042514e-06,
"loss": 0.4962,
"step": 1090
},
{
"epoch": 0.453416149068323,
"grad_norm": 2.598437906345615,
"learning_rate": 6.656024440211662e-06,
"loss": 0.4882,
"step": 1095
},
{
"epoch": 0.4554865424430642,
"grad_norm": 2.593321661322683,
"learning_rate": 6.621877954704996e-06,
"loss": 0.4855,
"step": 1100
},
{
"epoch": 0.4575569358178054,
"grad_norm": 2.7023638187383723,
"learning_rate": 6.5876467198083235e-06,
"loss": 0.4755,
"step": 1105
},
{
"epoch": 0.45962732919254656,
"grad_norm": 2.466568026339382,
"learning_rate": 6.553332524235937e-06,
"loss": 0.4868,
"step": 1110
},
{
"epoch": 0.4616977225672878,
"grad_norm": 2.4075337930670346,
"learning_rate": 6.518937161037144e-06,
"loss": 0.4766,
"step": 1115
},
{
"epoch": 0.463768115942029,
"grad_norm": 2.464658783047603,
"learning_rate": 6.484462427502572e-06,
"loss": 0.4765,
"step": 1120
},
{
"epoch": 0.4658385093167702,
"grad_norm": 2.5380751865960627,
"learning_rate": 6.44991012507026e-06,
"loss": 0.4664,
"step": 1125
},
{
"epoch": 0.46790890269151136,
"grad_norm": 2.49633781309852,
"learning_rate": 6.415282059231518e-06,
"loss": 0.4917,
"step": 1130
},
{
"epoch": 0.4699792960662526,
"grad_norm": 2.525786318450524,
"learning_rate": 6.380580039436586e-06,
"loss": 0.4683,
"step": 1135
},
{
"epoch": 0.4720496894409938,
"grad_norm": 2.5166460347854875,
"learning_rate": 6.345805879000087e-06,
"loss": 0.4862,
"step": 1140
},
{
"epoch": 0.474120082815735,
"grad_norm": 2.5450358590109565,
"learning_rate": 6.31096139500627e-06,
"loss": 0.4822,
"step": 1145
},
{
"epoch": 0.47619047619047616,
"grad_norm": 2.549418842471916,
"learning_rate": 6.2760484082140604e-06,
"loss": 0.478,
"step": 1150
},
{
"epoch": 0.4782608695652174,
"grad_norm": 2.4958659856500547,
"learning_rate": 6.24106874296192e-06,
"loss": 0.4699,
"step": 1155
},
{
"epoch": 0.4803312629399586,
"grad_norm": 2.626527624116192,
"learning_rate": 6.20602422707252e-06,
"loss": 0.4773,
"step": 1160
},
{
"epoch": 0.4824016563146998,
"grad_norm": 2.4609499575626805,
"learning_rate": 6.1709166917572264e-06,
"loss": 0.4603,
"step": 1165
},
{
"epoch": 0.484472049689441,
"grad_norm": 2.515241573825442,
"learning_rate": 6.135747971520412e-06,
"loss": 0.4576,
"step": 1170
},
{
"epoch": 0.4865424430641822,
"grad_norm": 2.524964073406838,
"learning_rate": 6.100519904063597e-06,
"loss": 0.462,
"step": 1175
},
{
"epoch": 0.4886128364389234,
"grad_norm": 2.5148293286686294,
"learning_rate": 6.0652343301894345e-06,
"loss": 0.4712,
"step": 1180
},
{
"epoch": 0.4906832298136646,
"grad_norm": 2.499516169022048,
"learning_rate": 6.029893093705492e-06,
"loss": 0.466,
"step": 1185
},
{
"epoch": 0.4927536231884058,
"grad_norm": 2.601142578827688,
"learning_rate": 5.99449804132794e-06,
"loss": 0.4583,
"step": 1190
},
{
"epoch": 0.494824016563147,
"grad_norm": 2.65823373640086,
"learning_rate": 5.959051022585025e-06,
"loss": 0.4716,
"step": 1195
},
{
"epoch": 0.4968944099378882,
"grad_norm": 2.5481027342051097,
"learning_rate": 5.923553889720447e-06,
"loss": 0.4721,
"step": 1200
},
{
"epoch": 0.4989648033126294,
"grad_norm": 2.6077505115765067,
"learning_rate": 5.888008497596553e-06,
"loss": 0.4477,
"step": 1205
},
{
"epoch": 0.5010351966873706,
"grad_norm": 2.4861547912595503,
"learning_rate": 5.852416703597431e-06,
"loss": 0.4522,
"step": 1210
},
{
"epoch": 0.5031055900621118,
"grad_norm": 2.6356091425155146,
"learning_rate": 5.816780367531841e-06,
"loss": 0.4518,
"step": 1215
},
{
"epoch": 0.505175983436853,
"grad_norm": 2.5485400203148694,
"learning_rate": 5.781101351536041e-06,
"loss": 0.453,
"step": 1220
},
{
"epoch": 0.5072463768115942,
"grad_norm": 2.4857617753584136,
"learning_rate": 5.745381519976477e-06,
"loss": 0.4687,
"step": 1225
},
{
"epoch": 0.5093167701863354,
"grad_norm": 2.5435865401804847,
"learning_rate": 5.7096227393523716e-06,
"loss": 0.4528,
"step": 1230
},
{
"epoch": 0.5113871635610766,
"grad_norm": 2.4857102655021666,
"learning_rate": 5.673826878198181e-06,
"loss": 0.4461,
"step": 1235
},
{
"epoch": 0.5134575569358178,
"grad_norm": 2.548923181016961,
"learning_rate": 5.637995806985961e-06,
"loss": 0.4461,
"step": 1240
},
{
"epoch": 0.515527950310559,
"grad_norm": 2.754257387829331,
"learning_rate": 5.602131398027637e-06,
"loss": 0.4623,
"step": 1245
},
{
"epoch": 0.5175983436853002,
"grad_norm": 2.5739827258600316,
"learning_rate": 5.566235525377155e-06,
"loss": 0.4587,
"step": 1250
},
{
"epoch": 0.5196687370600414,
"grad_norm": 2.6199240088350617,
"learning_rate": 5.530310064732559e-06,
"loss": 0.4423,
"step": 1255
},
{
"epoch": 0.5217391304347826,
"grad_norm": 2.5017468082526206,
"learning_rate": 5.494356893337985e-06,
"loss": 0.447,
"step": 1260
},
{
"epoch": 0.5238095238095238,
"grad_norm": 2.4475756497025962,
"learning_rate": 5.4583778898855576e-06,
"loss": 0.4425,
"step": 1265
},
{
"epoch": 0.525879917184265,
"grad_norm": 2.4797051803703334,
"learning_rate": 5.422374934417228e-06,
"loss": 0.4459,
"step": 1270
},
{
"epoch": 0.5279503105590062,
"grad_norm": 2.511185008743559,
"learning_rate": 5.386349908226538e-06,
"loss": 0.4506,
"step": 1275
},
{
"epoch": 0.5300207039337475,
"grad_norm": 2.5202140865028935,
"learning_rate": 5.350304693760301e-06,
"loss": 0.4514,
"step": 1280
},
{
"epoch": 0.5320910973084886,
"grad_norm": 2.420044445615974,
"learning_rate": 5.314241174520251e-06,
"loss": 0.438,
"step": 1285
},
{
"epoch": 0.5341614906832298,
"grad_norm": 2.592254989392042,
"learning_rate": 5.2781612349646175e-06,
"loss": 0.4282,
"step": 1290
},
{
"epoch": 0.5362318840579711,
"grad_norm": 2.6667125497088637,
"learning_rate": 5.242066760409653e-06,
"loss": 0.4396,
"step": 1295
},
{
"epoch": 0.5383022774327122,
"grad_norm": 2.450530941059547,
"learning_rate": 5.205959636931121e-06,
"loss": 0.4398,
"step": 1300
},
{
"epoch": 0.5403726708074534,
"grad_norm": 2.6473206241164084,
"learning_rate": 5.169841751265738e-06,
"loss": 0.4313,
"step": 1305
},
{
"epoch": 0.5424430641821946,
"grad_norm": 2.39954046846744,
"learning_rate": 5.13371499071259e-06,
"loss": 0.4264,
"step": 1310
},
{
"epoch": 0.5445134575569358,
"grad_norm": 2.398590287398949,
"learning_rate": 5.09758124303451e-06,
"loss": 0.4421,
"step": 1315
},
{
"epoch": 0.546583850931677,
"grad_norm": 2.3836291636458413,
"learning_rate": 5.0614423963594295e-06,
"loss": 0.4439,
"step": 1320
},
{
"epoch": 0.5486542443064182,
"grad_norm": 2.4595686235032637,
"learning_rate": 5.0253003390817264e-06,
"loss": 0.4252,
"step": 1325
},
{
"epoch": 0.5507246376811594,
"grad_norm": 2.3809656145760845,
"learning_rate": 4.989156959763551e-06,
"loss": 0.4236,
"step": 1330
},
{
"epoch": 0.5527950310559007,
"grad_norm": 2.610878920840408,
"learning_rate": 4.953014147036121e-06,
"loss": 0.432,
"step": 1335
},
{
"epoch": 0.5548654244306418,
"grad_norm": 2.380176771141565,
"learning_rate": 4.9168737895010615e-06,
"loss": 0.4328,
"step": 1340
},
{
"epoch": 0.556935817805383,
"grad_norm": 2.5994301919115586,
"learning_rate": 4.880737775631698e-06,
"loss": 0.4322,
"step": 1345
},
{
"epoch": 0.5590062111801242,
"grad_norm": 2.3183856073728704,
"learning_rate": 4.844607993674382e-06,
"loss": 0.4284,
"step": 1350
},
{
"epoch": 0.5610766045548654,
"grad_norm": 2.4001641944781262,
"learning_rate": 4.808486331549824e-06,
"loss": 0.4306,
"step": 1355
},
{
"epoch": 0.5631469979296067,
"grad_norm": 2.3398849256972674,
"learning_rate": 4.772374676754444e-06,
"loss": 0.4191,
"step": 1360
},
{
"epoch": 0.5652173913043478,
"grad_norm": 2.6034835600121835,
"learning_rate": 4.736274916261741e-06,
"loss": 0.426,
"step": 1365
},
{
"epoch": 0.567287784679089,
"grad_norm": 2.4449880511295725,
"learning_rate": 4.700188936423683e-06,
"loss": 0.4219,
"step": 1370
},
{
"epoch": 0.5693581780538303,
"grad_norm": 2.5020980844015352,
"learning_rate": 4.664118622872157e-06,
"loss": 0.4166,
"step": 1375
},
{
"epoch": 0.5714285714285714,
"grad_norm": 2.4138847963837633,
"learning_rate": 4.628065860420417e-06,
"loss": 0.43,
"step": 1380
},
{
"epoch": 0.5734989648033126,
"grad_norm": 2.6510892381332143,
"learning_rate": 4.592032532964611e-06,
"loss": 0.4211,
"step": 1385
},
{
"epoch": 0.5755693581780539,
"grad_norm": 2.4290310057749407,
"learning_rate": 4.556020523385326e-06,
"loss": 0.436,
"step": 1390
},
{
"epoch": 0.577639751552795,
"grad_norm": 2.6379248287348,
"learning_rate": 4.520031713449215e-06,
"loss": 0.4323,
"step": 1395
},
{
"epoch": 0.5797101449275363,
"grad_norm": 2.4152181161558866,
"learning_rate": 4.484067983710653e-06,
"loss": 0.4176,
"step": 1400
},
{
"epoch": 0.5817805383022774,
"grad_norm": 2.456186660123717,
"learning_rate": 4.448131213413485e-06,
"loss": 0.4286,
"step": 1405
},
{
"epoch": 0.5838509316770186,
"grad_norm": 2.4619064984612633,
"learning_rate": 4.4122232803928145e-06,
"loss": 0.4181,
"step": 1410
},
{
"epoch": 0.5859213250517599,
"grad_norm": 2.515165708545968,
"learning_rate": 4.376346060976888e-06,
"loss": 0.4107,
"step": 1415
},
{
"epoch": 0.587991718426501,
"grad_norm": 2.3855953284018456,
"learning_rate": 4.340501429889053e-06,
"loss": 0.4158,
"step": 1420
},
{
"epoch": 0.5900621118012422,
"grad_norm": 2.459583861923721,
"learning_rate": 4.30469126014978e-06,
"loss": 0.4043,
"step": 1425
},
{
"epoch": 0.5921325051759835,
"grad_norm": 2.4792502340500118,
"learning_rate": 4.268917422978811e-06,
"loss": 0.4003,
"step": 1430
},
{
"epoch": 0.5942028985507246,
"grad_norm": 2.6243677883306584,
"learning_rate": 4.233181787697364e-06,
"loss": 0.425,
"step": 1435
},
{
"epoch": 0.5962732919254659,
"grad_norm": 2.4060445038223253,
"learning_rate": 4.197486221630467e-06,
"loss": 0.4125,
"step": 1440
},
{
"epoch": 0.598343685300207,
"grad_norm": 2.333192916225142,
"learning_rate": 4.161832590009371e-06,
"loss": 0.4107,
"step": 1445
},
{
"epoch": 0.6004140786749482,
"grad_norm": 2.5065369175591363,
"learning_rate": 4.1262227558740915e-06,
"loss": 0.4081,
"step": 1450
},
{
"epoch": 0.6024844720496895,
"grad_norm": 2.487509717097525,
"learning_rate": 4.090658579976058e-06,
"loss": 0.4084,
"step": 1455
},
{
"epoch": 0.6045548654244306,
"grad_norm": 2.421454432071604,
"learning_rate": 4.055141920680883e-06,
"loss": 0.4093,
"step": 1460
},
{
"epoch": 0.6066252587991718,
"grad_norm": 2.4393555404832985,
"learning_rate": 4.019674633871246e-06,
"loss": 0.3985,
"step": 1465
},
{
"epoch": 0.6086956521739131,
"grad_norm": 2.3774733169227593,
"learning_rate": 3.984258572849926e-06,
"loss": 0.4037,
"step": 1470
},
{
"epoch": 0.6107660455486542,
"grad_norm": 2.6261521528226206,
"learning_rate": 3.9488955882429605e-06,
"loss": 0.3974,
"step": 1475
},
{
"epoch": 0.6128364389233955,
"grad_norm": 2.4240876700348006,
"learning_rate": 3.9135875279029366e-06,
"loss": 0.4001,
"step": 1480
},
{
"epoch": 0.6149068322981367,
"grad_norm": 2.5984545817951363,
"learning_rate": 3.878336236812432e-06,
"loss": 0.4112,
"step": 1485
},
{
"epoch": 0.6169772256728778,
"grad_norm": 2.3207269695983554,
"learning_rate": 3.84314355698762e-06,
"loss": 0.404,
"step": 1490
},
{
"epoch": 0.6190476190476191,
"grad_norm": 2.5152997136614035,
"learning_rate": 3.8080113273820024e-06,
"loss": 0.3984,
"step": 1495
},
{
"epoch": 0.6211180124223602,
"grad_norm": 2.4014319734474894,
"learning_rate": 3.7729413837903285e-06,
"loss": 0.4057,
"step": 1500
},
{
"epoch": 0.6231884057971014,
"grad_norm": 2.4898621079186727,
"learning_rate": 3.7379355587526543e-06,
"loss": 0.4001,
"step": 1505
},
{
"epoch": 0.6252587991718427,
"grad_norm": 2.4097787375747806,
"learning_rate": 3.702995681458605e-06,
"loss": 0.3939,
"step": 1510
},
{
"epoch": 0.6273291925465838,
"grad_norm": 2.33402918436361,
"learning_rate": 3.668123577651773e-06,
"loss": 0.3914,
"step": 1515
},
{
"epoch": 0.629399585921325,
"grad_norm": 2.2997377648959563,
"learning_rate": 3.633321069534325e-06,
"loss": 0.4028,
"step": 1520
},
{
"epoch": 0.6314699792960663,
"grad_norm": 2.4755804811958853,
"learning_rate": 3.5985899756717872e-06,
"loss": 0.3947,
"step": 1525
},
{
"epoch": 0.6335403726708074,
"grad_norm": 2.322186263857896,
"learning_rate": 3.563932110898015e-06,
"loss": 0.3989,
"step": 1530
},
{
"epoch": 0.6356107660455487,
"grad_norm": 2.366214150205079,
"learning_rate": 3.5293492862203595e-06,
"loss": 0.3926,
"step": 1535
},
{
"epoch": 0.6376811594202898,
"grad_norm": 2.343281760821824,
"learning_rate": 3.494843308725032e-06,
"loss": 0.3921,
"step": 1540
},
{
"epoch": 0.639751552795031,
"grad_norm": 2.3470709571665656,
"learning_rate": 3.460415981482691e-06,
"loss": 0.3911,
"step": 1545
},
{
"epoch": 0.6418219461697723,
"grad_norm": 2.3172908944396724,
"learning_rate": 3.426069103454209e-06,
"loss": 0.3934,
"step": 1550
},
{
"epoch": 0.6438923395445134,
"grad_norm": 2.364757829175207,
"learning_rate": 3.3918044693966802e-06,
"loss": 0.4034,
"step": 1555
},
{
"epoch": 0.6459627329192547,
"grad_norm": 2.389601354219861,
"learning_rate": 3.357623869769628e-06,
"loss": 0.3893,
"step": 1560
},
{
"epoch": 0.6480331262939959,
"grad_norm": 2.3563368486879193,
"learning_rate": 3.3235290906414575e-06,
"loss": 0.3978,
"step": 1565
},
{
"epoch": 0.650103519668737,
"grad_norm": 2.3745641368870647,
"learning_rate": 3.289521913596117e-06,
"loss": 0.3678,
"step": 1570
},
{
"epoch": 0.6521739130434783,
"grad_norm": 2.3280706561193747,
"learning_rate": 3.255604115640014e-06,
"loss": 0.3832,
"step": 1575
},
{
"epoch": 0.6542443064182195,
"grad_norm": 2.548296350832268,
"learning_rate": 3.2217774691091437e-06,
"loss": 0.3948,
"step": 1580
},
{
"epoch": 0.6563146997929606,
"grad_norm": 2.4059277091927203,
"learning_rate": 3.1880437415764936e-06,
"loss": 0.3809,
"step": 1585
},
{
"epoch": 0.6583850931677019,
"grad_norm": 2.3552963444757795,
"learning_rate": 3.1544046957596746e-06,
"loss": 0.3767,
"step": 1590
},
{
"epoch": 0.660455486542443,
"grad_norm": 2.3941281162623325,
"learning_rate": 3.1208620894288105e-06,
"loss": 0.3844,
"step": 1595
},
{
"epoch": 0.6625258799171843,
"grad_norm": 2.4133908197122422,
"learning_rate": 3.087417675314687e-06,
"loss": 0.3911,
"step": 1600
},
{
"epoch": 0.6645962732919255,
"grad_norm": 2.240517074625605,
"learning_rate": 3.0540732010171704e-06,
"loss": 0.3791,
"step": 1605
},
{
"epoch": 0.6666666666666666,
"grad_norm": 2.3846289931823006,
"learning_rate": 3.0208304089138873e-06,
"loss": 0.396,
"step": 1610
},
{
"epoch": 0.6687370600414079,
"grad_norm": 2.373051106938624,
"learning_rate": 2.9876910360691707e-06,
"loss": 0.3827,
"step": 1615
},
{
"epoch": 0.6708074534161491,
"grad_norm": 2.4319638143644675,
"learning_rate": 2.9546568141433007e-06,
"loss": 0.3864,
"step": 1620
},
{
"epoch": 0.6728778467908902,
"grad_norm": 2.299736837108655,
"learning_rate": 2.921729469302018e-06,
"loss": 0.3787,
"step": 1625
},
{
"epoch": 0.6749482401656315,
"grad_norm": 2.515282668787657,
"learning_rate": 2.8889107221263203e-06,
"loss": 0.3824,
"step": 1630
},
{
"epoch": 0.6770186335403726,
"grad_norm": 2.3694301433037097,
"learning_rate": 2.856202287522556e-06,
"loss": 0.3698,
"step": 1635
},
{
"epoch": 0.6790890269151139,
"grad_norm": 2.2983265677100375,
"learning_rate": 2.8236058746328203e-06,
"loss": 0.3767,
"step": 1640
},
{
"epoch": 0.6811594202898551,
"grad_norm": 2.43315902164643,
"learning_rate": 2.7911231867456345e-06,
"loss": 0.3713,
"step": 1645
},
{
"epoch": 0.6832298136645962,
"grad_norm": 2.34581734858733,
"learning_rate": 2.7587559212069554e-06,
"loss": 0.3847,
"step": 1650
},
{
"epoch": 0.6853002070393375,
"grad_norm": 2.398643950665849,
"learning_rate": 2.72650576933147e-06,
"loss": 0.3773,
"step": 1655
},
{
"epoch": 0.6873706004140787,
"grad_norm": 2.5934976031234944,
"learning_rate": 2.6943744163142292e-06,
"loss": 0.378,
"step": 1660
},
{
"epoch": 0.6894409937888198,
"grad_norm": 2.3745645368451025,
"learning_rate": 2.662363541142585e-06,
"loss": 0.3767,
"step": 1665
},
{
"epoch": 0.6915113871635611,
"grad_norm": 2.3509071760503635,
"learning_rate": 2.6304748165084503e-06,
"loss": 0.3724,
"step": 1670
},
{
"epoch": 0.6935817805383023,
"grad_norm": 2.408913615796527,
"learning_rate": 2.5987099087209035e-06,
"loss": 0.3738,
"step": 1675
},
{
"epoch": 0.6956521739130435,
"grad_norm": 2.3856450307685546,
"learning_rate": 2.5670704776191158e-06,
"loss": 0.3785,
"step": 1680
},
{
"epoch": 0.6977225672877847,
"grad_norm": 2.3751824485200217,
"learning_rate": 2.5355581764856186e-06,
"loss": 0.3713,
"step": 1685
},
{
"epoch": 0.6997929606625258,
"grad_norm": 2.3816709293947405,
"learning_rate": 2.5041746519599036e-06,
"loss": 0.3616,
"step": 1690
},
{
"epoch": 0.7018633540372671,
"grad_norm": 2.3626374585937695,
"learning_rate": 2.4729215439523945e-06,
"loss": 0.3709,
"step": 1695
},
{
"epoch": 0.7039337474120083,
"grad_norm": 2.3888912591393225,
"learning_rate": 2.441800485558748e-06,
"loss": 0.3671,
"step": 1700
},
{
"epoch": 0.7060041407867494,
"grad_norm": 2.3661474288759887,
"learning_rate": 2.410813102974512e-06,
"loss": 0.372,
"step": 1705
},
{
"epoch": 0.7080745341614907,
"grad_norm": 2.3498650327579025,
"learning_rate": 2.3799610154101603e-06,
"loss": 0.3706,
"step": 1710
},
{
"epoch": 0.7101449275362319,
"grad_norm": 2.4919895353280204,
"learning_rate": 2.3492458350064805e-06,
"loss": 0.3669,
"step": 1715
},
{
"epoch": 0.7122153209109731,
"grad_norm": 2.3794025385578954,
"learning_rate": 2.3186691667503315e-06,
"loss": 0.3657,
"step": 1720
},
{
"epoch": 0.7142857142857143,
"grad_norm": 2.3251567273214766,
"learning_rate": 2.288232608390778e-06,
"loss": 0.3646,
"step": 1725
},
{
"epoch": 0.7163561076604554,
"grad_norm": 2.407368271969709,
"learning_rate": 2.2579377503555967e-06,
"loss": 0.3677,
"step": 1730
},
{
"epoch": 0.7184265010351967,
"grad_norm": 2.3071018559559637,
"learning_rate": 2.227786175668181e-06,
"loss": 0.3557,
"step": 1735
},
{
"epoch": 0.7204968944099379,
"grad_norm": 2.4893163462904457,
"learning_rate": 2.1977794598648106e-06,
"loss": 0.3563,
"step": 1740
},
{
"epoch": 0.722567287784679,
"grad_norm": 2.291371824281609,
"learning_rate": 2.167919170912333e-06,
"loss": 0.3689,
"step": 1745
},
{
"epoch": 0.7246376811594203,
"grad_norm": 2.3380292076800755,
"learning_rate": 2.138206869126222e-06,
"loss": 0.3573,
"step": 1750
},
{
"epoch": 0.7267080745341615,
"grad_norm": 2.4453849756112334,
"learning_rate": 2.108644107089055e-06,
"loss": 0.3502,
"step": 1755
},
{
"epoch": 0.7287784679089027,
"grad_norm": 2.443139761591236,
"learning_rate": 2.0792324295693782e-06,
"loss": 0.3561,
"step": 1760
},
{
"epoch": 0.7308488612836439,
"grad_norm": 2.302265166915135,
"learning_rate": 2.0499733734409884e-06,
"loss": 0.3621,
"step": 1765
},
{
"epoch": 0.7329192546583851,
"grad_norm": 2.3838622887226535,
"learning_rate": 2.020868467602623e-06,
"loss": 0.3588,
"step": 1770
},
{
"epoch": 0.7349896480331263,
"grad_norm": 2.1793179653263293,
"learning_rate": 1.9919192328980703e-06,
"loss": 0.3513,
"step": 1775
},
{
"epoch": 0.7370600414078675,
"grad_norm": 2.4112782824350143,
"learning_rate": 1.9631271820367058e-06,
"loss": 0.3498,
"step": 1780
},
{
"epoch": 0.7391304347826086,
"grad_norm": 2.3296442055108133,
"learning_rate": 1.9344938195144353e-06,
"loss": 0.3636,
"step": 1785
},
{
"epoch": 0.7412008281573499,
"grad_norm": 2.3808963961562064,
"learning_rate": 1.9060206415350896e-06,
"loss": 0.3418,
"step": 1790
},
{
"epoch": 0.7432712215320911,
"grad_norm": 2.374085044882124,
"learning_rate": 1.8777091359322386e-06,
"loss": 0.3535,
"step": 1795
},
{
"epoch": 0.7453416149068323,
"grad_norm": 2.333661092958635,
"learning_rate": 1.8495607820914451e-06,
"loss": 0.3595,
"step": 1800
},
{
"epoch": 0.7474120082815735,
"grad_norm": 2.3514710917271464,
"learning_rate": 1.8215770508729602e-06,
"loss": 0.3531,
"step": 1805
},
{
"epoch": 0.7494824016563147,
"grad_norm": 2.3663746446172773,
"learning_rate": 1.7937594045348666e-06,
"loss": 0.3589,
"step": 1810
},
{
"epoch": 0.7515527950310559,
"grad_norm": 2.3919416697461866,
"learning_rate": 1.7661092966566728e-06,
"loss": 0.356,
"step": 1815
},
{
"epoch": 0.7536231884057971,
"grad_norm": 2.3148061424323836,
"learning_rate": 1.7386281720633569e-06,
"loss": 0.3513,
"step": 1820
},
{
"epoch": 0.7556935817805382,
"grad_norm": 2.3375605825384627,
"learning_rate": 1.7113174667498612e-06,
"loss": 0.3513,
"step": 1825
},
{
"epoch": 0.7577639751552795,
"grad_norm": 2.2470720742363004,
"learning_rate": 1.6841786078060674e-06,
"loss": 0.344,
"step": 1830
},
{
"epoch": 0.7598343685300207,
"grad_norm": 2.519539431999456,
"learning_rate": 1.6572130133422204e-06,
"loss": 0.344,
"step": 1835
},
{
"epoch": 0.7619047619047619,
"grad_norm": 2.2357554640401127,
"learning_rate": 1.630422092414823e-06,
"loss": 0.3401,
"step": 1840
},
{
"epoch": 0.7639751552795031,
"grad_norm": 2.393897715845976,
"learning_rate": 1.6038072449530118e-06,
"loss": 0.3515,
"step": 1845
},
{
"epoch": 0.7660455486542443,
"grad_norm": 2.284176351570047,
"learning_rate": 1.5773698616854065e-06,
"loss": 0.3432,
"step": 1850
},
{
"epoch": 0.7681159420289855,
"grad_norm": 2.4493329198592755,
"learning_rate": 1.5511113240674364e-06,
"loss": 0.345,
"step": 1855
},
{
"epoch": 0.7701863354037267,
"grad_norm": 2.2770793247132883,
"learning_rate": 1.5250330042091543e-06,
"loss": 0.3388,
"step": 1860
},
{
"epoch": 0.772256728778468,
"grad_norm": 2.353942648756986,
"learning_rate": 1.4991362648035368e-06,
"loss": 0.3402,
"step": 1865
},
{
"epoch": 0.7743271221532091,
"grad_norm": 2.4082953312930915,
"learning_rate": 1.4734224590552844e-06,
"loss": 0.3359,
"step": 1870
},
{
"epoch": 0.7763975155279503,
"grad_norm": 2.299065278453441,
"learning_rate": 1.4478929306101025e-06,
"loss": 0.3439,
"step": 1875
},
{
"epoch": 0.7784679089026915,
"grad_norm": 2.3482929230842737,
"learning_rate": 1.4225490134844967e-06,
"loss": 0.3416,
"step": 1880
},
{
"epoch": 0.7805383022774327,
"grad_norm": 2.526873271439544,
"learning_rate": 1.3973920319960654e-06,
"loss": 0.3486,
"step": 1885
},
{
"epoch": 0.782608695652174,
"grad_norm": 2.542295229828987,
"learning_rate": 1.3724233006942972e-06,
"loss": 0.3422,
"step": 1890
},
{
"epoch": 0.7846790890269151,
"grad_norm": 2.4872596787592407,
"learning_rate": 1.347644124291883e-06,
"loss": 0.3456,
"step": 1895
},
{
"epoch": 0.7867494824016563,
"grad_norm": 2.416999777859718,
"learning_rate": 1.323055797596532e-06,
"loss": 0.3394,
"step": 1900
},
{
"epoch": 0.7888198757763976,
"grad_norm": 2.5786050379863603,
"learning_rate": 1.2986596054433255e-06,
"loss": 0.3399,
"step": 1905
},
{
"epoch": 0.7908902691511387,
"grad_norm": 2.415140501526238,
"learning_rate": 1.2744568226275662e-06,
"loss": 0.348,
"step": 1910
},
{
"epoch": 0.7929606625258799,
"grad_norm": 2.4822020986040796,
"learning_rate": 1.2504487138381782e-06,
"loss": 0.3526,
"step": 1915
},
{
"epoch": 0.7950310559006211,
"grad_norm": 2.499492182218722,
"learning_rate": 1.2266365335916086e-06,
"loss": 0.3423,
"step": 1920
},
{
"epoch": 0.7971014492753623,
"grad_norm": 2.349828543569679,
"learning_rate": 1.203021526166287e-06,
"loss": 0.3423,
"step": 1925
},
{
"epoch": 0.7991718426501035,
"grad_norm": 2.32632190317854,
"learning_rate": 1.179604925537598e-06,
"loss": 0.3445,
"step": 1930
},
{
"epoch": 0.8012422360248447,
"grad_norm": 2.3378844582471254,
"learning_rate": 1.1563879553134072e-06,
"loss": 0.342,
"step": 1935
},
{
"epoch": 0.8033126293995859,
"grad_norm": 2.275806455849109,
"learning_rate": 1.1333718286701162e-06,
"loss": 0.3368,
"step": 1940
},
{
"epoch": 0.8053830227743272,
"grad_norm": 2.428108558665187,
"learning_rate": 1.110557748289275e-06,
"loss": 0.337,
"step": 1945
},
{
"epoch": 0.8074534161490683,
"grad_norm": 2.4293733020976203,
"learning_rate": 1.0879469062947369e-06,
"loss": 0.3476,
"step": 1950
},
{
"epoch": 0.8095238095238095,
"grad_norm": 2.372514823833049,
"learning_rate": 1.0655404841903648e-06,
"loss": 0.3424,
"step": 1955
},
{
"epoch": 0.8115942028985508,
"grad_norm": 2.271912403283504,
"learning_rate": 1.0433396527982893e-06,
"loss": 0.3352,
"step": 1960
},
{
"epoch": 0.8136645962732919,
"grad_norm": 2.3071195447516017,
"learning_rate": 1.0213455721977356e-06,
"loss": 0.3298,
"step": 1965
},
{
"epoch": 0.8157349896480331,
"grad_norm": 2.2832195173760432,
"learning_rate": 9.995593916643981e-07,
"loss": 0.3388,
"step": 1970
},
{
"epoch": 0.8178053830227743,
"grad_norm": 2.5002011517189775,
"learning_rate": 9.779822496103902e-07,
"loss": 0.3495,
"step": 1975
},
{
"epoch": 0.8198757763975155,
"grad_norm": 2.417662236034124,
"learning_rate": 9.566152735247576e-07,
"loss": 0.3351,
"step": 1980
},
{
"epoch": 0.8219461697722568,
"grad_norm": 2.2908891885326828,
"learning_rate": 9.354595799145627e-07,
"loss": 0.3346,
"step": 1985
},
{
"epoch": 0.8240165631469979,
"grad_norm": 2.1858040123308897,
"learning_rate": 9.145162742465424e-07,
"loss": 0.3364,
"step": 1990
},
{
"epoch": 0.8260869565217391,
"grad_norm": 2.440309599040629,
"learning_rate": 8.937864508893396e-07,
"loss": 0.3367,
"step": 1995
},
{
"epoch": 0.8281573498964804,
"grad_norm": 2.534190418443867,
"learning_rate": 8.732711930563259e-07,
"loss": 0.3371,
"step": 2000
},
{
"epoch": 0.8302277432712215,
"grad_norm": 2.3183100371463086,
"learning_rate": 8.529715727489912e-07,
"loss": 0.3305,
"step": 2005
},
{
"epoch": 0.8322981366459627,
"grad_norm": 2.4530932407380397,
"learning_rate": 8.32888650700936e-07,
"loss": 0.3226,
"step": 2010
},
{
"epoch": 0.8343685300207039,
"grad_norm": 2.291593647377325,
"learning_rate": 8.130234763224342e-07,
"loss": 0.3309,
"step": 2015
},
{
"epoch": 0.8364389233954451,
"grad_norm": 2.3812779417617667,
"learning_rate": 7.93377087645607e-07,
"loss": 0.3281,
"step": 2020
},
{
"epoch": 0.8385093167701864,
"grad_norm": 2.3726129395455056,
"learning_rate": 7.739505112701751e-07,
"loss": 0.3279,
"step": 2025
},
{
"epoch": 0.8405797101449275,
"grad_norm": 2.390816271029597,
"learning_rate": 7.547447623098191e-07,
"loss": 0.3246,
"step": 2030
},
{
"epoch": 0.8426501035196687,
"grad_norm": 2.4115520577321257,
"learning_rate": 7.357608443391312e-07,
"loss": 0.3163,
"step": 2035
},
{
"epoch": 0.84472049689441,
"grad_norm": 2.3696838099867152,
"learning_rate": 7.169997493411774e-07,
"loss": 0.3303,
"step": 2040
},
{
"epoch": 0.8467908902691511,
"grad_norm": 2.3355725998372616,
"learning_rate": 6.984624576556647e-07,
"loss": 0.3291,
"step": 2045
},
{
"epoch": 0.8488612836438924,
"grad_norm": 2.247548659269975,
"learning_rate": 6.801499379277115e-07,
"loss": 0.3294,
"step": 2050
},
{
"epoch": 0.8509316770186336,
"grad_norm": 2.226491669420491,
"learning_rate": 6.620631470572314e-07,
"loss": 0.3315,
"step": 2055
},
{
"epoch": 0.8530020703933747,
"grad_norm": 2.269738584663688,
"learning_rate": 6.442030301489338e-07,
"loss": 0.3285,
"step": 2060
},
{
"epoch": 0.855072463768116,
"grad_norm": 2.457363477733083,
"learning_rate": 6.265705204629402e-07,
"loss": 0.3295,
"step": 2065
},
{
"epoch": 0.8571428571428571,
"grad_norm": 2.3865924747276632,
"learning_rate": 6.091665393660112e-07,
"loss": 0.3313,
"step": 2070
},
{
"epoch": 0.8592132505175983,
"grad_norm": 2.4350093655112,
"learning_rate": 5.919919962834069e-07,
"loss": 0.3279,
"step": 2075
},
{
"epoch": 0.8612836438923396,
"grad_norm": 2.337875938278272,
"learning_rate": 5.750477886513656e-07,
"loss": 0.3296,
"step": 2080
},
{
"epoch": 0.8633540372670807,
"grad_norm": 2.4100527367913798,
"learning_rate": 5.58334801870209e-07,
"loss": 0.3306,
"step": 2085
},
{
"epoch": 0.865424430641822,
"grad_norm": 2.458257552095922,
"learning_rate": 5.418539092580727e-07,
"loss": 0.3309,
"step": 2090
},
{
"epoch": 0.8674948240165632,
"grad_norm": 2.357226102821327,
"learning_rate": 5.256059720052787e-07,
"loss": 0.3165,
"step": 2095
},
{
"epoch": 0.8695652173913043,
"grad_norm": 2.33070943252138,
"learning_rate": 5.095918391293303e-07,
"loss": 0.3407,
"step": 2100
},
{
"epoch": 0.8716356107660456,
"grad_norm": 2.396693307063821,
"learning_rate": 4.938123474305473e-07,
"loss": 0.3235,
"step": 2105
},
{
"epoch": 0.8737060041407867,
"grad_norm": 2.3760024597980682,
"learning_rate": 4.782683214483413e-07,
"loss": 0.3192,
"step": 2110
},
{
"epoch": 0.8757763975155279,
"grad_norm": 2.3982385649751317,
"learning_rate": 4.629605734181331e-07,
"loss": 0.3251,
"step": 2115
},
{
"epoch": 0.8778467908902692,
"grad_norm": 2.4181252670052027,
"learning_rate": 4.478899032289047e-07,
"loss": 0.3253,
"step": 2120
},
{
"epoch": 0.8799171842650103,
"grad_norm": 2.4582637585511073,
"learning_rate": 4.3305709838140755e-07,
"loss": 0.3236,
"step": 2125
},
{
"epoch": 0.8819875776397516,
"grad_norm": 2.460567347509424,
"learning_rate": 4.1846293394700645e-07,
"loss": 0.3219,
"step": 2130
},
{
"epoch": 0.8840579710144928,
"grad_norm": 2.3074837080255906,
"learning_rate": 4.041081725271856e-07,
"loss": 0.3073,
"step": 2135
},
{
"epoch": 0.8861283643892339,
"grad_norm": 2.3320789676611744,
"learning_rate": 3.8999356421369426e-07,
"loss": 0.322,
"step": 2140
},
{
"epoch": 0.8881987577639752,
"grad_norm": 2.3476061224867664,
"learning_rate": 3.7611984654935707e-07,
"loss": 0.3277,
"step": 2145
},
{
"epoch": 0.8902691511387164,
"grad_norm": 2.29668781545797,
"learning_rate": 3.6248774448952695e-07,
"loss": 0.3199,
"step": 2150
},
{
"epoch": 0.8923395445134575,
"grad_norm": 2.300471663322285,
"learning_rate": 3.490979703642117e-07,
"loss": 0.3109,
"step": 2155
},
{
"epoch": 0.8944099378881988,
"grad_norm": 2.4965315302153623,
"learning_rate": 3.35951223840848e-07,
"loss": 0.3224,
"step": 2160
},
{
"epoch": 0.8964803312629399,
"grad_norm": 2.5138290884080545,
"learning_rate": 3.2304819188773815e-07,
"loss": 0.3328,
"step": 2165
},
{
"epoch": 0.8985507246376812,
"grad_norm": 2.2178193008905307,
"learning_rate": 3.1038954873816027e-07,
"loss": 0.3086,
"step": 2170
},
{
"epoch": 0.9006211180124224,
"grad_norm": 2.4481560944381378,
"learning_rate": 2.9797595585512986e-07,
"loss": 0.324,
"step": 2175
},
{
"epoch": 0.9026915113871635,
"grad_norm": 2.4170021197708413,
"learning_rate": 2.858080618968423e-07,
"loss": 0.3212,
"step": 2180
},
{
"epoch": 0.9047619047619048,
"grad_norm": 2.3081360765132195,
"learning_rate": 2.738865026827714e-07,
"loss": 0.3191,
"step": 2185
},
{
"epoch": 0.906832298136646,
"grad_norm": 2.296864380825452,
"learning_rate": 2.6221190116045126e-07,
"loss": 0.319,
"step": 2190
},
{
"epoch": 0.9089026915113871,
"grad_norm": 2.444406353598231,
"learning_rate": 2.507848673729224e-07,
"loss": 0.3258,
"step": 2195
},
{
"epoch": 0.9109730848861284,
"grad_norm": 2.44272390295649,
"learning_rate": 2.3960599842685394e-07,
"loss": 0.3248,
"step": 2200
},
{
"epoch": 0.9130434782608695,
"grad_norm": 2.579957250218807,
"learning_rate": 2.2867587846134188e-07,
"loss": 0.3175,
"step": 2205
},
{
"epoch": 0.9151138716356108,
"grad_norm": 2.6035512857713297,
"learning_rate": 2.179950786173879e-07,
"loss": 0.3175,
"step": 2210
},
{
"epoch": 0.917184265010352,
"grad_norm": 2.347129005103096,
"learning_rate": 2.07564157008055e-07,
"loss": 0.3111,
"step": 2215
},
{
"epoch": 0.9192546583850931,
"grad_norm": 2.2665660329686275,
"learning_rate": 1.9738365868930188e-07,
"loss": 0.3196,
"step": 2220
},
{
"epoch": 0.9213250517598344,
"grad_norm": 2.4024556708707174,
"learning_rate": 1.8745411563150128e-07,
"loss": 0.3162,
"step": 2225
},
{
"epoch": 0.9233954451345756,
"grad_norm": 2.504734801787616,
"learning_rate": 1.7777604669164727e-07,
"loss": 0.3236,
"step": 2230
},
{
"epoch": 0.9254658385093167,
"grad_norm": 2.3547703025432627,
"learning_rate": 1.6834995758623685e-07,
"loss": 0.3206,
"step": 2235
},
{
"epoch": 0.927536231884058,
"grad_norm": 2.4065147698763454,
"learning_rate": 1.5917634086484778e-07,
"loss": 0.321,
"step": 2240
},
{
"epoch": 0.9296066252587992,
"grad_norm": 2.4537451005611146,
"learning_rate": 1.5025567588439925e-07,
"loss": 0.3158,
"step": 2245
},
{
"epoch": 0.9316770186335404,
"grad_norm": 2.5088572289429827,
"learning_rate": 1.415884287841074e-07,
"loss": 0.3193,
"step": 2250
},
{
"epoch": 0.9337474120082816,
"grad_norm": 2.3420956664158035,
"learning_rate": 1.331750524611225e-07,
"loss": 0.3137,
"step": 2255
},
{
"epoch": 0.9358178053830227,
"grad_norm": 2.3067560021581053,
"learning_rate": 1.2501598654686532e-07,
"loss": 0.3188,
"step": 2260
},
{
"epoch": 0.937888198757764,
"grad_norm": 2.298348040797609,
"learning_rate": 1.1711165738405749e-07,
"loss": 0.3208,
"step": 2265
},
{
"epoch": 0.9399585921325052,
"grad_norm": 2.3629624611235087,
"learning_rate": 1.0946247800443899e-07,
"loss": 0.3164,
"step": 2270
},
{
"epoch": 0.9420289855072463,
"grad_norm": 2.406102252286578,
"learning_rate": 1.0206884810718964e-07,
"loss": 0.3168,
"step": 2275
},
{
"epoch": 0.9440993788819876,
"grad_norm": 2.4590908048099256,
"learning_rate": 9.493115403803876e-08,
"loss": 0.314,
"step": 2280
},
{
"epoch": 0.9461697722567288,
"grad_norm": 2.678524242947405,
"learning_rate": 8.804976876908178e-08,
"loss": 0.3379,
"step": 2285
},
{
"epoch": 0.94824016563147,
"grad_norm": 2.4441412038736288,
"learning_rate": 8.142505187928861e-08,
"loss": 0.312,
"step": 2290
},
{
"epoch": 0.9503105590062112,
"grad_norm": 2.4918799370418236,
"learning_rate": 7.505734953571486e-08,
"loss": 0.3129,
"step": 2295
},
{
"epoch": 0.9523809523809523,
"grad_norm": 2.344102107154711,
"learning_rate": 6.894699447541231e-08,
"loss": 0.3167,
"step": 2300
},
{
"epoch": 0.9544513457556936,
"grad_norm": 2.3552934454835985,
"learning_rate": 6.309430598804234e-08,
"loss": 0.3142,
"step": 2305
},
{
"epoch": 0.9565217391304348,
"grad_norm": 2.5507428589353065,
"learning_rate": 5.7499589899193134e-08,
"loss": 0.3194,
"step": 2310
},
{
"epoch": 0.9585921325051759,
"grad_norm": 2.4570869813826848,
"learning_rate": 5.216313855439858e-08,
"loss": 0.3126,
"step": 2315
},
{
"epoch": 0.9606625258799172,
"grad_norm": 2.4667994829779536,
"learning_rate": 4.7085230803860513e-08,
"loss": 0.3195,
"step": 2320
},
{
"epoch": 0.9627329192546584,
"grad_norm": 2.2886057343616675,
"learning_rate": 4.2266131987880876e-08,
"loss": 0.3283,
"step": 2325
},
{
"epoch": 0.9648033126293996,
"grad_norm": 2.574246461332851,
"learning_rate": 3.770609392299285e-08,
"loss": 0.3195,
"step": 2330
},
{
"epoch": 0.9668737060041408,
"grad_norm": 2.3426062000880563,
"learning_rate": 3.340535488880525e-08,
"loss": 0.3054,
"step": 2335
},
{
"epoch": 0.968944099378882,
"grad_norm": 2.325141607110564,
"learning_rate": 2.9364139615550846e-08,
"loss": 0.3135,
"step": 2340
},
{
"epoch": 0.9710144927536232,
"grad_norm": 2.33098584703444,
"learning_rate": 2.5582659272341827e-08,
"loss": 0.3215,
"step": 2345
},
{
"epoch": 0.9730848861283644,
"grad_norm": 2.390777367949656,
"learning_rate": 2.206111145613865e-08,
"loss": 0.3137,
"step": 2350
},
{
"epoch": 0.9751552795031055,
"grad_norm": 2.5248325357907047,
"learning_rate": 1.8799680181421087e-08,
"loss": 0.3121,
"step": 2355
},
{
"epoch": 0.9772256728778468,
"grad_norm": 2.424251800700087,
"learning_rate": 1.5798535870575893e-08,
"loss": 0.3215,
"step": 2360
},
{
"epoch": 0.979296066252588,
"grad_norm": 2.4119823545769354,
"learning_rate": 1.30578353449895e-08,
"loss": 0.3247,
"step": 2365
},
{
"epoch": 0.9813664596273292,
"grad_norm": 2.3598051537797002,
"learning_rate": 1.0577721816854015e-08,
"loss": 0.3173,
"step": 2370
},
{
"epoch": 0.9834368530020704,
"grad_norm": 2.486323280726411,
"learning_rate": 8.35832488168431e-09,
"loss": 0.3192,
"step": 2375
},
{
"epoch": 0.9855072463768116,
"grad_norm": 2.5058191872081585,
"learning_rate": 6.399760511546227e-09,
"loss": 0.3146,
"step": 2380
},
{
"epoch": 0.9875776397515528,
"grad_norm": 2.4215679763279243,
"learning_rate": 4.7021310489953065e-09,
"loss": 0.323,
"step": 2385
},
{
"epoch": 0.989648033126294,
"grad_norm": 2.4158190639068278,
"learning_rate": 3.265525201731623e-09,
"loss": 0.3235,
"step": 2390
},
{
"epoch": 0.9917184265010351,
"grad_norm": 2.5061110103967885,
"learning_rate": 2.090018037960717e-09,
"loss": 0.318,
"step": 2395
},
{
"epoch": 0.9937888198757764,
"grad_norm": 2.4079071252683106,
"learning_rate": 1.1756709824750633e-09,
"loss": 0.3175,
"step": 2400
},
{
"epoch": 0.9958592132505176,
"grad_norm": 2.3201834622838238,
"learning_rate": 5.225318134399749e-10,
"loss": 0.3163,
"step": 2405
},
{
"epoch": 0.9979296066252588,
"grad_norm": 2.484615601276709,
"learning_rate": 1.306346599011521e-10,
"loss": 0.3174,
"step": 2410
},
{
"epoch": 1.0,
"grad_norm": 2.460009245744386,
"learning_rate": 0.0,
"loss": 0.314,
"step": 2415
},
{
"epoch": 1.0,
"eval_runtime": 3.3938,
"eval_samples_per_second": 2.947,
"eval_steps_per_second": 0.884,
"step": 2415
},
{
"epoch": 1.0,
"step": 2415,
"total_flos": 252825934233600.0,
"train_loss": 0.5143075758379192,
"train_runtime": 21526.6935,
"train_samples_per_second": 1.795,
"train_steps_per_second": 0.112
}
],
"logging_steps": 5,
"max_steps": 2415,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 252825934233600.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}