zephyr-7b-gemma-sft-5p / trainer_state.json
Qin Liu
Model save
60ecd17 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 978,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"grad_norm": 71.33833396142957,
"learning_rate": 2.040816326530612e-06,
"loss": 16.3689,
"step": 1
},
{
"epoch": 0.02,
"grad_norm": 71.89278720209892,
"learning_rate": 1.0204081632653061e-05,
"loss": 16.1499,
"step": 5
},
{
"epoch": 0.03,
"grad_norm": 65.72662945467931,
"learning_rate": 2.0408163265306123e-05,
"loss": 15.7133,
"step": 10
},
{
"epoch": 0.05,
"grad_norm": 53.593107939489194,
"learning_rate": 3.061224489795919e-05,
"loss": 13.3181,
"step": 15
},
{
"epoch": 0.06,
"grad_norm": 44.297380165268656,
"learning_rate": 4.0816326530612245e-05,
"loss": 9.8593,
"step": 20
},
{
"epoch": 0.08,
"grad_norm": 19.985681179385026,
"learning_rate": 5.102040816326531e-05,
"loss": 6.3007,
"step": 25
},
{
"epoch": 0.09,
"grad_norm": 12.900325144421931,
"learning_rate": 6.122448979591838e-05,
"loss": 4.5204,
"step": 30
},
{
"epoch": 0.11,
"grad_norm": 9.486242109363502,
"learning_rate": 7.142857142857143e-05,
"loss": 3.3761,
"step": 35
},
{
"epoch": 0.12,
"grad_norm": 7.1186109989022945,
"learning_rate": 8.163265306122449e-05,
"loss": 2.3811,
"step": 40
},
{
"epoch": 0.14,
"grad_norm": 5.023418237589821,
"learning_rate": 9.183673469387756e-05,
"loss": 1.851,
"step": 45
},
{
"epoch": 0.15,
"grad_norm": 4.554190050004734,
"learning_rate": 0.00010204081632653062,
"loss": 1.6038,
"step": 50
},
{
"epoch": 0.17,
"grad_norm": 2.8804678557363896,
"learning_rate": 0.00011224489795918367,
"loss": 1.4206,
"step": 55
},
{
"epoch": 0.18,
"grad_norm": 3.9529502182289398,
"learning_rate": 0.00012244897959183676,
"loss": 1.3664,
"step": 60
},
{
"epoch": 0.2,
"grad_norm": 2.248441902031195,
"learning_rate": 0.0001326530612244898,
"loss": 1.274,
"step": 65
},
{
"epoch": 0.21,
"grad_norm": 1.4413020771523257,
"learning_rate": 0.00014285714285714287,
"loss": 1.2439,
"step": 70
},
{
"epoch": 0.23,
"grad_norm": 1.493793739973276,
"learning_rate": 0.0001530612244897959,
"loss": 1.1924,
"step": 75
},
{
"epoch": 0.25,
"grad_norm": 1.458616609136318,
"learning_rate": 0.00016326530612244898,
"loss": 1.0841,
"step": 80
},
{
"epoch": 0.26,
"grad_norm": 1.6382650720922636,
"learning_rate": 0.00017346938775510205,
"loss": 1.1002,
"step": 85
},
{
"epoch": 0.28,
"grad_norm": 1.326596598600597,
"learning_rate": 0.00018367346938775512,
"loss": 1.0797,
"step": 90
},
{
"epoch": 0.29,
"grad_norm": 1.2754617583755954,
"learning_rate": 0.00019387755102040816,
"loss": 1.1313,
"step": 95
},
{
"epoch": 0.31,
"grad_norm": 1.399995470448795,
"learning_rate": 0.00019999745104274993,
"loss": 1.0745,
"step": 100
},
{
"epoch": 0.32,
"grad_norm": 1.5754181414386559,
"learning_rate": 0.00019996877676598733,
"loss": 1.0513,
"step": 105
},
{
"epoch": 0.34,
"grad_norm": 1.9062455011264745,
"learning_rate": 0.00019990825118233957,
"loss": 1.0973,
"step": 110
},
{
"epoch": 0.35,
"grad_norm": 1.2370608192620602,
"learning_rate": 0.00019981589357601727,
"loss": 1.0174,
"step": 115
},
{
"epoch": 0.37,
"grad_norm": 1.4957785022540953,
"learning_rate": 0.0001996917333733128,
"loss": 1.0495,
"step": 120
},
{
"epoch": 0.38,
"grad_norm": 2.0317868737729614,
"learning_rate": 0.00019953581013322502,
"loss": 1.0631,
"step": 125
},
{
"epoch": 0.4,
"grad_norm": 2.3640165600919016,
"learning_rate": 0.00019934817353485501,
"loss": 1.0597,
"step": 130
},
{
"epoch": 0.41,
"grad_norm": 1.4377618154920253,
"learning_rate": 0.00019912888336157792,
"loss": 1.032,
"step": 135
},
{
"epoch": 0.43,
"grad_norm": 1.6672769945555381,
"learning_rate": 0.00019887800948199496,
"loss": 1.0587,
"step": 140
},
{
"epoch": 0.44,
"grad_norm": 1.7175054641214251,
"learning_rate": 0.00019859563182767268,
"loss": 0.9808,
"step": 145
},
{
"epoch": 0.46,
"grad_norm": 2.055095058889153,
"learning_rate": 0.00019828184036767556,
"loss": 1.0301,
"step": 150
},
{
"epoch": 0.48,
"grad_norm": 1.1983943009571292,
"learning_rate": 0.00019793673507990088,
"loss": 0.9844,
"step": 155
},
{
"epoch": 0.49,
"grad_norm": 1.3919505852905427,
"learning_rate": 0.00019756042591922434,
"loss": 1.0325,
"step": 160
},
{
"epoch": 0.51,
"grad_norm": 1.0559448337045312,
"learning_rate": 0.00019715303278246724,
"loss": 0.9838,
"step": 165
},
{
"epoch": 0.52,
"grad_norm": 1.2948165594267842,
"learning_rate": 0.00019671468547019573,
"loss": 1.0184,
"step": 170
},
{
"epoch": 0.54,
"grad_norm": 1.757117551491023,
"learning_rate": 0.00019624552364536473,
"loss": 0.9951,
"step": 175
},
{
"epoch": 0.55,
"grad_norm": 1.0463120070964065,
"learning_rate": 0.00019574569678881964,
"loss": 0.9886,
"step": 180
},
{
"epoch": 0.57,
"grad_norm": 1.3316373554522543,
"learning_rate": 0.00019521536415166978,
"loss": 1.029,
"step": 185
},
{
"epoch": 0.58,
"grad_norm": 1.206556314151547,
"learning_rate": 0.000194654694704549,
"loss": 1.0185,
"step": 190
},
{
"epoch": 0.6,
"grad_norm": 1.4497941420805274,
"learning_rate": 0.00019406386708377955,
"loss": 0.9952,
"step": 195
},
{
"epoch": 0.61,
"grad_norm": 1.5111306426761502,
"learning_rate": 0.00019344306953445633,
"loss": 0.9533,
"step": 200
},
{
"epoch": 0.63,
"grad_norm": 1.5977185668213372,
"learning_rate": 0.00019279249985046948,
"loss": 0.9654,
"step": 205
},
{
"epoch": 0.64,
"grad_norm": 1.4919134437495485,
"learning_rate": 0.000192112365311485,
"loss": 0.9859,
"step": 210
},
{
"epoch": 0.66,
"grad_norm": 1.5403119826226057,
"learning_rate": 0.00019140288261690276,
"loss": 1.0116,
"step": 215
},
{
"epoch": 0.67,
"grad_norm": 1.3349640842488235,
"learning_rate": 0.00019066427781681315,
"loss": 0.9787,
"step": 220
},
{
"epoch": 0.69,
"grad_norm": 1.3368327515154408,
"learning_rate": 0.00018989678623997503,
"loss": 0.9729,
"step": 225
},
{
"epoch": 0.71,
"grad_norm": 1.3225697881784793,
"learning_rate": 0.0001891006524188368,
"loss": 0.9581,
"step": 230
},
{
"epoch": 0.72,
"grad_norm": 1.2111399093084592,
"learning_rate": 0.00018827613001162532,
"loss": 0.9993,
"step": 235
},
{
"epoch": 0.74,
"grad_norm": 1.4099026636155603,
"learning_rate": 0.00018742348172152726,
"loss": 0.9647,
"step": 240
},
{
"epoch": 0.75,
"grad_norm": 1.2230878261166291,
"learning_rate": 0.00018654297921298863,
"loss": 0.9514,
"step": 245
},
{
"epoch": 0.77,
"grad_norm": 1.376607259051782,
"learning_rate": 0.0001856349030251589,
"loss": 1.0106,
"step": 250
},
{
"epoch": 0.78,
"grad_norm": 1.5573343816575618,
"learning_rate": 0.0001846995424825079,
"loss": 0.9622,
"step": 255
},
{
"epoch": 0.8,
"grad_norm": 1.0889975412225616,
"learning_rate": 0.00018373719560264327,
"loss": 1.0253,
"step": 260
},
{
"epoch": 0.81,
"grad_norm": 1.2977152056359567,
"learning_rate": 0.0001827481690013584,
"loss": 1.0099,
"step": 265
},
{
"epoch": 0.83,
"grad_norm": 1.3195763372373182,
"learning_rate": 0.0001817327777949407,
"loss": 0.9993,
"step": 270
},
{
"epoch": 0.84,
"grad_norm": 1.2289773734368,
"learning_rate": 0.00018069134549977172,
"loss": 0.9763,
"step": 275
},
{
"epoch": 0.86,
"grad_norm": 1.2177961464095397,
"learning_rate": 0.00017962420392925066,
"loss": 0.9791,
"step": 280
},
{
"epoch": 0.87,
"grad_norm": 1.1607055291626518,
"learning_rate": 0.00017853169308807448,
"loss": 0.9801,
"step": 285
},
{
"epoch": 0.89,
"grad_norm": 1.2705107901850918,
"learning_rate": 0.00017741416106390826,
"loss": 0.9328,
"step": 290
},
{
"epoch": 0.9,
"grad_norm": 1.3432344503611726,
"learning_rate": 0.00017627196391647982,
"loss": 0.955,
"step": 295
},
{
"epoch": 0.92,
"grad_norm": 0.9736443478187509,
"learning_rate": 0.00017510546556413498,
"loss": 1.0068,
"step": 300
},
{
"epoch": 0.94,
"grad_norm": 1.2298616794502824,
"learning_rate": 0.00017391503766788828,
"loss": 0.9365,
"step": 305
},
{
"epoch": 0.95,
"grad_norm": 1.2782580681474314,
"learning_rate": 0.00017270105951300738,
"loss": 0.9398,
"step": 310
},
{
"epoch": 0.97,
"grad_norm": 1.043708677429849,
"learning_rate": 0.0001714639178881678,
"loss": 1.0062,
"step": 315
},
{
"epoch": 0.98,
"grad_norm": 1.042074622226514,
"learning_rate": 0.00017020400696221737,
"loss": 0.9572,
"step": 320
},
{
"epoch": 1.0,
"grad_norm": 1.238677723678181,
"learning_rate": 0.00016892172815858894,
"loss": 0.9772,
"step": 325
},
{
"epoch": 1.0,
"eval_loss": 1.159922480583191,
"eval_runtime": 156.0746,
"eval_samples_per_second": 14.801,
"eval_steps_per_second": 0.468,
"step": 326
},
{
"epoch": 1.01,
"grad_norm": 1.0217319894407395,
"learning_rate": 0.00016761749002740193,
"loss": 0.946,
"step": 330
},
{
"epoch": 1.03,
"grad_norm": 1.1246569533018882,
"learning_rate": 0.00016629170811529318,
"loss": 0.918,
"step": 335
},
{
"epoch": 1.04,
"grad_norm": 1.002004590448949,
"learning_rate": 0.00016494480483301836,
"loss": 0.9497,
"step": 340
},
{
"epoch": 1.06,
"grad_norm": 1.212522253836943,
"learning_rate": 0.00016357720932086688,
"loss": 0.8865,
"step": 345
},
{
"epoch": 1.07,
"grad_norm": 1.1180679319761555,
"learning_rate": 0.00016218935731193224,
"loss": 0.9137,
"step": 350
},
{
"epoch": 1.09,
"grad_norm": 1.243690740596817,
"learning_rate": 0.00016078169099328197,
"loss": 0.928,
"step": 355
},
{
"epoch": 1.1,
"grad_norm": 1.0059412266802719,
"learning_rate": 0.00015935465886507142,
"loss": 0.9132,
"step": 360
},
{
"epoch": 1.12,
"grad_norm": 1.449853291817741,
"learning_rate": 0.0001579087155976459,
"loss": 0.9236,
"step": 365
},
{
"epoch": 1.13,
"grad_norm": 1.200969298535005,
"learning_rate": 0.00015644432188667695,
"loss": 0.9191,
"step": 370
},
{
"epoch": 1.15,
"grad_norm": 1.3541813597090206,
"learning_rate": 0.00015496194430637903,
"loss": 0.9882,
"step": 375
},
{
"epoch": 1.17,
"grad_norm": 1.2265813596883026,
"learning_rate": 0.00015346205516085306,
"loss": 0.9468,
"step": 380
},
{
"epoch": 1.18,
"grad_norm": 1.4522657029483004,
"learning_rate": 0.0001519451323336044,
"loss": 0.8801,
"step": 385
},
{
"epoch": 1.2,
"grad_norm": 1.0083988226706067,
"learning_rate": 0.0001504116591352832,
"loss": 0.9001,
"step": 390
},
{
"epoch": 1.21,
"grad_norm": 1.2286558458574148,
"learning_rate": 0.00014886212414969553,
"loss": 0.9405,
"step": 395
},
{
"epoch": 1.23,
"grad_norm": 1.269365799457322,
"learning_rate": 0.00014729702107813436,
"loss": 1.0354,
"step": 400
},
{
"epoch": 1.24,
"grad_norm": 1.5775962375858563,
"learning_rate": 0.00014571684858208044,
"loss": 0.9391,
"step": 405
},
{
"epoch": 1.26,
"grad_norm": 1.5593272986635398,
"learning_rate": 0.00014412211012432212,
"loss": 0.8933,
"step": 410
},
{
"epoch": 1.27,
"grad_norm": 2.0000926705831166,
"learning_rate": 0.00014251331380854603,
"loss": 0.9003,
"step": 415
},
{
"epoch": 1.29,
"grad_norm": 1.3265098276554033,
"learning_rate": 0.00014089097221744868,
"loss": 0.9548,
"step": 420
},
{
"epoch": 1.3,
"grad_norm": 0.9964611399554132,
"learning_rate": 0.00013925560224942144,
"loss": 0.9493,
"step": 425
},
{
"epoch": 1.32,
"grad_norm": 1.0573997683696912,
"learning_rate": 0.00013760772495385998,
"loss": 0.9212,
"step": 430
},
{
"epoch": 1.33,
"grad_norm": 1.066450746302158,
"learning_rate": 0.00013594786536515153,
"loss": 0.9229,
"step": 435
},
{
"epoch": 1.35,
"grad_norm": 1.030770134357563,
"learning_rate": 0.00013427655233539228,
"loss": 0.9116,
"step": 440
},
{
"epoch": 1.37,
"grad_norm": 1.0183082642006178,
"learning_rate": 0.00013259431836588843,
"loss": 0.9292,
"step": 445
},
{
"epoch": 1.38,
"grad_norm": 1.132995542647628,
"learning_rate": 0.00013090169943749476,
"loss": 0.9155,
"step": 450
},
{
"epoch": 1.4,
"grad_norm": 1.0945930214994195,
"learning_rate": 0.00012919923483984414,
"loss": 0.9098,
"step": 455
},
{
"epoch": 1.41,
"grad_norm": 0.7625306665500079,
"learning_rate": 0.00012748746699952338,
"loss": 0.8722,
"step": 460
},
{
"epoch": 1.43,
"grad_norm": 1.2062480398980977,
"learning_rate": 0.00012576694130724905,
"loss": 0.8827,
"step": 465
},
{
"epoch": 1.44,
"grad_norm": 1.0298307072907715,
"learning_rate": 0.00012403820594409924,
"loss": 0.9001,
"step": 470
},
{
"epoch": 1.46,
"grad_norm": 1.1672438274286052,
"learning_rate": 0.00012230181170685636,
"loss": 0.9191,
"step": 475
},
{
"epoch": 1.47,
"grad_norm": 0.9217837301990276,
"learning_rate": 0.00012055831183251607,
"loss": 0.9126,
"step": 480
},
{
"epoch": 1.49,
"grad_norm": 0.9859571731260532,
"learning_rate": 0.00011880826182201926,
"loss": 0.8884,
"step": 485
},
{
"epoch": 1.5,
"grad_norm": 0.9295473746329976,
"learning_rate": 0.0001170522192632624,
"loss": 0.9002,
"step": 490
},
{
"epoch": 1.52,
"grad_norm": 0.9211214138607775,
"learning_rate": 0.00011529074365344301,
"loss": 0.8813,
"step": 495
},
{
"epoch": 1.53,
"grad_norm": 0.9277063802264193,
"learning_rate": 0.00011352439622079689,
"loss": 0.9304,
"step": 500
},
{
"epoch": 1.55,
"grad_norm": 0.9836858148590184,
"learning_rate": 0.00011175373974578378,
"loss": 0.9336,
"step": 505
},
{
"epoch": 1.56,
"grad_norm": 0.9201365394724025,
"learning_rate": 0.00010997933838177827,
"loss": 0.9007,
"step": 510
},
{
"epoch": 1.58,
"grad_norm": 1.0294642137453307,
"learning_rate": 0.00010820175747532373,
"loss": 0.9182,
"step": 515
},
{
"epoch": 1.6,
"grad_norm": 0.9815519204802536,
"learning_rate": 0.00010642156338600551,
"loss": 0.95,
"step": 520
},
{
"epoch": 1.61,
"grad_norm": 1.00497732793873,
"learning_rate": 0.00010463932330600196,
"loss": 0.905,
"step": 525
},
{
"epoch": 1.63,
"grad_norm": 1.0176550178508887,
"learning_rate": 0.00010285560507936961,
"loss": 0.9369,
"step": 530
},
{
"epoch": 1.64,
"grad_norm": 0.9691655097650015,
"learning_rate": 0.0001010709770211212,
"loss": 0.9151,
"step": 535
},
{
"epoch": 1.66,
"grad_norm": 1.043651874834748,
"learning_rate": 9.928600773615307e-05,
"loss": 0.9027,
"step": 540
},
{
"epoch": 1.67,
"grad_norm": 1.008769921338687,
"learning_rate": 9.750126593808082e-05,
"loss": 0.8884,
"step": 545
},
{
"epoch": 1.69,
"grad_norm": 1.0140163426042983,
"learning_rate": 9.571732026803977e-05,
"loss": 0.9461,
"step": 550
},
{
"epoch": 1.7,
"grad_norm": 0.8868172626319482,
"learning_rate": 9.393473911350893e-05,
"loss": 0.9176,
"step": 555
},
{
"epoch": 1.72,
"grad_norm": 0.9481462007119397,
"learning_rate": 9.215409042721552e-05,
"loss": 0.8978,
"step": 560
},
{
"epoch": 1.73,
"grad_norm": 0.976465698662806,
"learning_rate": 9.037594154617812e-05,
"loss": 0.9459,
"step": 565
},
{
"epoch": 1.75,
"grad_norm": 1.0361077003461703,
"learning_rate": 8.860085901094595e-05,
"loss": 0.8965,
"step": 570
},
{
"epoch": 1.76,
"grad_norm": 0.9744756453463735,
"learning_rate": 8.682940838509207e-05,
"loss": 0.9142,
"step": 575
},
{
"epoch": 1.78,
"grad_norm": 1.1336224000575972,
"learning_rate": 8.50621540750175e-05,
"loss": 0.8745,
"step": 580
},
{
"epoch": 1.79,
"grad_norm": 0.9837911952285221,
"learning_rate": 8.329965915012451e-05,
"loss": 0.9572,
"step": 585
},
{
"epoch": 1.81,
"grad_norm": 1.0097764972768413,
"learning_rate": 8.154248516341548e-05,
"loss": 0.9534,
"step": 590
},
{
"epoch": 1.83,
"grad_norm": 0.9690942378398057,
"learning_rate": 7.979119197257505e-05,
"loss": 0.9227,
"step": 595
},
{
"epoch": 1.84,
"grad_norm": 1.0127982484169298,
"learning_rate": 7.804633756159259e-05,
"loss": 0.9084,
"step": 600
},
{
"epoch": 1.86,
"grad_norm": 0.7943207513052626,
"learning_rate": 7.63084778629813e-05,
"loss": 0.9248,
"step": 605
},
{
"epoch": 1.87,
"grad_norm": 0.9647056317342992,
"learning_rate": 7.457816658065134e-05,
"loss": 0.9431,
"step": 610
},
{
"epoch": 1.89,
"grad_norm": 1.1403897570938433,
"learning_rate": 7.285595501349258e-05,
"loss": 0.9263,
"step": 615
},
{
"epoch": 1.9,
"grad_norm": 1.0404617304378672,
"learning_rate": 7.114239187972416e-05,
"loss": 0.9103,
"step": 620
},
{
"epoch": 1.92,
"grad_norm": 1.128471625008042,
"learning_rate": 6.94380231420656e-05,
"loss": 0.9093,
"step": 625
},
{
"epoch": 1.93,
"grad_norm": 0.9916742401674975,
"learning_rate": 6.774339183378663e-05,
"loss": 0.9331,
"step": 630
},
{
"epoch": 1.95,
"grad_norm": 0.8859012774344858,
"learning_rate": 6.60590378856896e-05,
"loss": 0.8951,
"step": 635
},
{
"epoch": 1.96,
"grad_norm": 0.9040358847526676,
"learning_rate": 6.438549795408106e-05,
"loss": 0.9024,
"step": 640
},
{
"epoch": 1.98,
"grad_norm": 0.9311904528713643,
"learning_rate": 6.272330524978613e-05,
"loss": 0.9135,
"step": 645
},
{
"epoch": 1.99,
"grad_norm": 0.9264414330433429,
"learning_rate": 6.107298936826086e-05,
"loss": 0.9241,
"step": 650
},
{
"epoch": 2.0,
"eval_loss": 1.123016595840454,
"eval_runtime": 155.2791,
"eval_samples_per_second": 14.876,
"eval_steps_per_second": 0.47,
"step": 652
},
{
"epoch": 2.01,
"grad_norm": 2.0610368602503826,
"learning_rate": 5.9435076120856616e-05,
"loss": 0.8977,
"step": 655
},
{
"epoch": 2.02,
"grad_norm": 1.1812695122093289,
"learning_rate": 5.7810087367289744e-05,
"loss": 0.8808,
"step": 660
},
{
"epoch": 2.04,
"grad_norm": 1.1330049331313394,
"learning_rate": 5.619854084937085e-05,
"loss": 0.8697,
"step": 665
},
{
"epoch": 2.06,
"grad_norm": 1.2350886571017046,
"learning_rate": 5.4600950026045326e-05,
"loss": 0.8605,
"step": 670
},
{
"epoch": 2.07,
"grad_norm": 1.0239529925811819,
"learning_rate": 5.301782390979929e-05,
"loss": 0.9032,
"step": 675
},
{
"epoch": 2.09,
"grad_norm": 1.1559449179497776,
"learning_rate": 5.1449666904481585e-05,
"loss": 0.8576,
"step": 680
},
{
"epoch": 2.1,
"grad_norm": 1.0036439705356706,
"learning_rate": 4.989697864459452e-05,
"loss": 0.8811,
"step": 685
},
{
"epoch": 2.12,
"grad_norm": 0.9058518668863318,
"learning_rate": 4.836025383610382e-05,
"loss": 0.8795,
"step": 690
},
{
"epoch": 2.13,
"grad_norm": 1.098710487990691,
"learning_rate": 4.683998209881943e-05,
"loss": 0.8806,
"step": 695
},
{
"epoch": 2.15,
"grad_norm": 1.0781974748771004,
"learning_rate": 4.5336647810396215e-05,
"loss": 0.8598,
"step": 700
},
{
"epoch": 2.16,
"grad_norm": 1.0706338013924643,
"learning_rate": 4.385072995200532e-05,
"loss": 0.8635,
"step": 705
},
{
"epoch": 2.18,
"grad_norm": 0.9548080567529033,
"learning_rate": 4.238270195572472e-05,
"loss": 0.8757,
"step": 710
},
{
"epoch": 2.19,
"grad_norm": 1.001721295347845,
"learning_rate": 4.093303155369771e-05,
"loss": 0.8919,
"step": 715
},
{
"epoch": 2.21,
"grad_norm": 1.0089807158142916,
"learning_rate": 3.9502180629107756e-05,
"loss": 0.8483,
"step": 720
},
{
"epoch": 2.22,
"grad_norm": 0.903535471960529,
"learning_rate": 3.8090605069016595e-05,
"loss": 0.9141,
"step": 725
},
{
"epoch": 2.24,
"grad_norm": 0.9151685871279929,
"learning_rate": 3.669875461911297e-05,
"loss": 0.8872,
"step": 730
},
{
"epoch": 2.25,
"grad_norm": 1.1222577897364008,
"learning_rate": 3.53270727404179e-05,
"loss": 0.8424,
"step": 735
},
{
"epoch": 2.27,
"grad_norm": 0.9362543386209857,
"learning_rate": 3.397599646799256e-05,
"loss": 0.9175,
"step": 740
},
{
"epoch": 2.29,
"grad_norm": 1.0462736025526818,
"learning_rate": 3.2645956271693257e-05,
"loss": 0.8897,
"step": 745
},
{
"epoch": 2.3,
"grad_norm": 0.9259326236334475,
"learning_rate": 3.133737591901864e-05,
"loss": 0.8671,
"step": 750
},
{
"epoch": 2.32,
"grad_norm": 0.9219118148042651,
"learning_rate": 3.0050672340091725e-05,
"loss": 0.9011,
"step": 755
},
{
"epoch": 2.33,
"grad_norm": 0.8476861060157034,
"learning_rate": 2.8786255494820835e-05,
"loss": 0.9027,
"step": 760
},
{
"epoch": 2.35,
"grad_norm": 0.847331969421794,
"learning_rate": 2.754452824228132e-05,
"loss": 0.8339,
"step": 765
},
{
"epoch": 2.36,
"grad_norm": 0.9426248637438077,
"learning_rate": 2.6325886212359498e-05,
"loss": 0.9129,
"step": 770
},
{
"epoch": 2.38,
"grad_norm": 1.0772706903168978,
"learning_rate": 2.51307176797001e-05,
"loss": 0.9141,
"step": 775
},
{
"epoch": 2.39,
"grad_norm": 0.921621759803798,
"learning_rate": 2.3959403439996907e-05,
"loss": 0.8866,
"step": 780
},
{
"epoch": 2.41,
"grad_norm": 1.0228032129160294,
"learning_rate": 2.2812316688666737e-05,
"loss": 0.8411,
"step": 785
},
{
"epoch": 2.42,
"grad_norm": 0.9794026584875383,
"learning_rate": 2.1689822901944457e-05,
"loss": 0.8148,
"step": 790
},
{
"epoch": 2.44,
"grad_norm": 1.1116773627967562,
"learning_rate": 2.0592279720437858e-05,
"loss": 0.9003,
"step": 795
},
{
"epoch": 2.45,
"grad_norm": 0.9563546052669486,
"learning_rate": 1.9520036835178668e-05,
"loss": 0.8883,
"step": 800
},
{
"epoch": 2.47,
"grad_norm": 0.9044636934616683,
"learning_rate": 1.847343587620679e-05,
"loss": 0.8769,
"step": 805
},
{
"epoch": 2.48,
"grad_norm": 0.9221646859533941,
"learning_rate": 1.74528103037226e-05,
"loss": 0.8033,
"step": 810
},
{
"epoch": 2.5,
"grad_norm": 1.0009290473276315,
"learning_rate": 1.645848530184233e-05,
"loss": 0.8381,
"step": 815
},
{
"epoch": 2.52,
"grad_norm": 0.9671119344154833,
"learning_rate": 1.5490777674990376e-05,
"loss": 0.8645,
"step": 820
},
{
"epoch": 2.53,
"grad_norm": 0.9041738041322639,
"learning_rate": 1.4549995746961332e-05,
"loss": 0.8624,
"step": 825
},
{
"epoch": 2.55,
"grad_norm": 0.9947155263280221,
"learning_rate": 1.3636439262684298e-05,
"loss": 0.8862,
"step": 830
},
{
"epoch": 2.56,
"grad_norm": 1.1355384323430027,
"learning_rate": 1.2750399292720283e-05,
"loss": 0.8914,
"step": 835
},
{
"epoch": 2.58,
"grad_norm": 0.9481996675285544,
"learning_rate": 1.1892158140523546e-05,
"loss": 0.8962,
"step": 840
},
{
"epoch": 2.59,
"grad_norm": 0.963461677617234,
"learning_rate": 1.1061989252496053e-05,
"loss": 0.8738,
"step": 845
},
{
"epoch": 2.61,
"grad_norm": 0.9731419353271179,
"learning_rate": 1.026015713086418e-05,
"loss": 0.8955,
"step": 850
},
{
"epoch": 2.62,
"grad_norm": 0.9744266823325466,
"learning_rate": 9.486917249404815e-06,
"loss": 0.9313,
"step": 855
},
{
"epoch": 2.64,
"grad_norm": 0.8667111998887408,
"learning_rate": 8.742515972048404e-06,
"loss": 0.83,
"step": 860
},
{
"epoch": 2.65,
"grad_norm": 1.007522159578116,
"learning_rate": 8.027190474384127e-06,
"loss": 0.9093,
"step": 865
},
{
"epoch": 2.67,
"grad_norm": 0.9497316726005834,
"learning_rate": 7.341168668092857e-06,
"loss": 0.8619,
"step": 870
},
{
"epoch": 2.68,
"grad_norm": 0.9890888660929842,
"learning_rate": 6.684669128331655e-06,
"loss": 0.8994,
"step": 875
},
{
"epoch": 2.7,
"grad_norm": 0.9637211127886939,
"learning_rate": 6.057901024092949e-06,
"loss": 0.8815,
"step": 880
},
{
"epoch": 2.71,
"grad_norm": 0.9454153835142218,
"learning_rate": 5.461064051560705e-06,
"loss": 0.8748,
"step": 885
},
{
"epoch": 2.73,
"grad_norm": 0.9656325292151272,
"learning_rate": 4.8943483704846475e-06,
"loss": 0.9164,
"step": 890
},
{
"epoch": 2.75,
"grad_norm": 1.167721756716136,
"learning_rate": 4.357934543593045e-06,
"loss": 0.8695,
"step": 895
},
{
"epoch": 2.76,
"grad_norm": 1.0842862872661083,
"learning_rate": 3.851993479063154e-06,
"loss": 0.864,
"step": 900
},
{
"epoch": 2.78,
"grad_norm": 0.9136231470411893,
"learning_rate": 3.376686376067695e-06,
"loss": 0.8723,
"step": 905
},
{
"epoch": 2.79,
"grad_norm": 0.8333735277456533,
"learning_rate": 2.9321646734147502e-06,
"loss": 0.8476,
"step": 910
},
{
"epoch": 2.81,
"grad_norm": 0.9013760974476497,
"learning_rate": 2.51857000129756e-06,
"loss": 0.8046,
"step": 915
},
{
"epoch": 2.82,
"grad_norm": 0.9502325737857031,
"learning_rate": 2.1360341361692517e-06,
"loss": 0.8794,
"step": 920
},
{
"epoch": 2.84,
"grad_norm": 1.0665384348019042,
"learning_rate": 1.784678958757291e-06,
"loss": 0.8471,
"step": 925
},
{
"epoch": 2.85,
"grad_norm": 0.867152754297921,
"learning_rate": 1.4646164152307018e-06,
"loss": 0.8471,
"step": 930
},
{
"epoch": 2.87,
"grad_norm": 0.8765872693616653,
"learning_rate": 1.1759484815326294e-06,
"loss": 0.8778,
"step": 935
},
{
"epoch": 2.88,
"grad_norm": 0.9977478387159145,
"learning_rate": 9.187671308895418e-07,
"loss": 0.8814,
"step": 940
},
{
"epoch": 2.9,
"grad_norm": 1.0928634095287688,
"learning_rate": 6.931543045073708e-07,
"loss": 0.8518,
"step": 945
},
{
"epoch": 2.91,
"grad_norm": 0.9615275585544613,
"learning_rate": 4.991818854640395e-07,
"loss": 0.8465,
"step": 950
},
{
"epoch": 2.93,
"grad_norm": 0.946638261056933,
"learning_rate": 3.369116758066171e-07,
"loss": 0.8591,
"step": 955
},
{
"epoch": 2.94,
"grad_norm": 0.9273052910302554,
"learning_rate": 2.0639537686037991e-07,
"loss": 0.8652,
"step": 960
},
{
"epoch": 2.96,
"grad_norm": 0.9396149477675896,
"learning_rate": 1.0767457275615567e-07,
"loss": 0.8919,
"step": 965
},
{
"epoch": 2.98,
"grad_norm": 0.9064011249970909,
"learning_rate": 4.078071718107701e-08,
"loss": 0.8436,
"step": 970
},
{
"epoch": 2.99,
"grad_norm": 0.9696335526663976,
"learning_rate": 5.735123357042404e-09,
"loss": 0.8687,
"step": 975
},
{
"epoch": 3.0,
"eval_loss": 1.123045563697815,
"eval_runtime": 155.2182,
"eval_samples_per_second": 14.882,
"eval_steps_per_second": 0.47,
"step": 978
},
{
"epoch": 3.0,
"step": 978,
"total_flos": 2513689986990080.0,
"train_loss": 1.2729851507214194,
"train_runtime": 11657.2618,
"train_samples_per_second": 5.364,
"train_steps_per_second": 0.084
}
],
"logging_steps": 5,
"max_steps": 978,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"total_flos": 2513689986990080.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}