zephyr-7b-gemma-sft / trainer_state.json
ale-bay's picture
Model save
e3b96c3 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.994991652754591,
"eval_steps": 500,
"global_step": 897,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00333889816360601,
"grad_norm": 1633.0780614838177,
"learning_rate": 2.2222222222222224e-07,
"loss": 14.7478,
"step": 1
},
{
"epoch": 0.01669449081803005,
"grad_norm": 1232.852374646303,
"learning_rate": 1.111111111111111e-06,
"loss": 13.7637,
"step": 5
},
{
"epoch": 0.0333889816360601,
"grad_norm": 160.62292486481945,
"learning_rate": 2.222222222222222e-06,
"loss": 6.5941,
"step": 10
},
{
"epoch": 0.05008347245409015,
"grad_norm": 54.67306425849189,
"learning_rate": 3.3333333333333333e-06,
"loss": 2.4942,
"step": 15
},
{
"epoch": 0.0667779632721202,
"grad_norm": 36.70425670339494,
"learning_rate": 4.444444444444444e-06,
"loss": 1.7415,
"step": 20
},
{
"epoch": 0.08347245409015025,
"grad_norm": 31.233563869291352,
"learning_rate": 5.555555555555557e-06,
"loss": 1.5583,
"step": 25
},
{
"epoch": 0.1001669449081803,
"grad_norm": 11.808038087592866,
"learning_rate": 6.666666666666667e-06,
"loss": 1.465,
"step": 30
},
{
"epoch": 0.11686143572621036,
"grad_norm": 11.136891922183436,
"learning_rate": 7.77777777777778e-06,
"loss": 1.4276,
"step": 35
},
{
"epoch": 0.1335559265442404,
"grad_norm": 35.559026101254695,
"learning_rate": 8.888888888888888e-06,
"loss": 1.2878,
"step": 40
},
{
"epoch": 0.15025041736227046,
"grad_norm": 24.433680953972228,
"learning_rate": 1e-05,
"loss": 1.2278,
"step": 45
},
{
"epoch": 0.1669449081803005,
"grad_norm": 124.38615525639153,
"learning_rate": 1.1111111111111113e-05,
"loss": 1.1305,
"step": 50
},
{
"epoch": 0.18363939899833054,
"grad_norm": 123.27724020667802,
"learning_rate": 1.2222222222222224e-05,
"loss": 1.2199,
"step": 55
},
{
"epoch": 0.2003338898163606,
"grad_norm": 57.91104969220954,
"learning_rate": 1.3333333333333333e-05,
"loss": 1.2104,
"step": 60
},
{
"epoch": 0.21702838063439064,
"grad_norm": 18.21646289641897,
"learning_rate": 1.4444444444444446e-05,
"loss": 1.1268,
"step": 65
},
{
"epoch": 0.2337228714524207,
"grad_norm": 9.526734443694286,
"learning_rate": 1.555555555555556e-05,
"loss": 1.1197,
"step": 70
},
{
"epoch": 0.25041736227045075,
"grad_norm": 82.75544838534852,
"learning_rate": 1.6666666666666667e-05,
"loss": 1.1559,
"step": 75
},
{
"epoch": 0.2671118530884808,
"grad_norm": 49.643780627497094,
"learning_rate": 1.7777777777777777e-05,
"loss": 1.172,
"step": 80
},
{
"epoch": 0.2838063439065108,
"grad_norm": 27.77568045936215,
"learning_rate": 1.888888888888889e-05,
"loss": 1.1498,
"step": 85
},
{
"epoch": 0.3005008347245409,
"grad_norm": 85.83380293575804,
"learning_rate": 2e-05,
"loss": 1.1418,
"step": 90
},
{
"epoch": 0.31719532554257096,
"grad_norm": 9.632034343502017,
"learning_rate": 1.9998105699049984e-05,
"loss": 1.1007,
"step": 95
},
{
"epoch": 0.333889816360601,
"grad_norm": 37.22916540635626,
"learning_rate": 1.9992423513875158e-05,
"loss": 1.0916,
"step": 100
},
{
"epoch": 0.35058430717863104,
"grad_norm": 49.23626406084266,
"learning_rate": 1.9982955597229275e-05,
"loss": 1.1476,
"step": 105
},
{
"epoch": 0.3672787979966611,
"grad_norm": 41.53491105452397,
"learning_rate": 1.9969705536129033e-05,
"loss": 1.1108,
"step": 110
},
{
"epoch": 0.38397328881469117,
"grad_norm": 52.40878364354266,
"learning_rate": 1.9952678350495104e-05,
"loss": 1.0695,
"step": 115
},
{
"epoch": 0.4006677796327212,
"grad_norm": 35.305587442500695,
"learning_rate": 1.9931880491250263e-05,
"loss": 1.0682,
"step": 120
},
{
"epoch": 0.41736227045075125,
"grad_norm": 37.14178936776283,
"learning_rate": 1.990731983787542e-05,
"loss": 1.0501,
"step": 125
},
{
"epoch": 0.4340567612687813,
"grad_norm": 67.45550897017867,
"learning_rate": 1.987900569542438e-05,
"loss": 1.054,
"step": 130
},
{
"epoch": 0.4507512520868113,
"grad_norm": 14.900523343688862,
"learning_rate": 1.9846948790998532e-05,
"loss": 1.047,
"step": 135
},
{
"epoch": 0.4674457429048414,
"grad_norm": 26.97945797218196,
"learning_rate": 1.9811161269682776e-05,
"loss": 1.0417,
"step": 140
},
{
"epoch": 0.48414023372287146,
"grad_norm": 31.768395936058706,
"learning_rate": 1.9771656689944238e-05,
"loss": 1.0622,
"step": 145
},
{
"epoch": 0.5008347245409015,
"grad_norm": 17.22340247377649,
"learning_rate": 1.9728450018495506e-05,
"loss": 1.0399,
"step": 150
},
{
"epoch": 0.5175292153589316,
"grad_norm": 20.785198792338,
"learning_rate": 1.968155762462433e-05,
"loss": 1.0335,
"step": 155
},
{
"epoch": 0.5342237061769616,
"grad_norm": 12.868219548980708,
"learning_rate": 1.9630997273991964e-05,
"loss": 1.0585,
"step": 160
},
{
"epoch": 0.5509181969949917,
"grad_norm": 16.52094280846651,
"learning_rate": 1.9576788121902457e-05,
"loss": 1.0563,
"step": 165
},
{
"epoch": 0.5676126878130217,
"grad_norm": 15.154512778467309,
"learning_rate": 1.951895070604547e-05,
"loss": 1.052,
"step": 170
},
{
"epoch": 0.5843071786310517,
"grad_norm": 15.109216310230805,
"learning_rate": 1.9457506938715357e-05,
"loss": 1.0353,
"step": 175
},
{
"epoch": 0.6010016694490818,
"grad_norm": 21.543252819256704,
"learning_rate": 1.9392480098509488e-05,
"loss": 1.0182,
"step": 180
},
{
"epoch": 0.6176961602671118,
"grad_norm": 15.449989929271629,
"learning_rate": 1.93238948215089e-05,
"loss": 1.0341,
"step": 185
},
{
"epoch": 0.6343906510851419,
"grad_norm": 23.23349123860765,
"learning_rate": 1.9251777091944665e-05,
"loss": 1.0261,
"step": 190
},
{
"epoch": 0.6510851419031719,
"grad_norm": 16.41219778505035,
"learning_rate": 1.9176154232353513e-05,
"loss": 1.0388,
"step": 195
},
{
"epoch": 0.667779632721202,
"grad_norm": 18.296566123425983,
"learning_rate": 1.9097054893226395e-05,
"loss": 1.0254,
"step": 200
},
{
"epoch": 0.6844741235392321,
"grad_norm": 36.60365240585806,
"learning_rate": 1.9014509042153964e-05,
"loss": 1.0338,
"step": 205
},
{
"epoch": 0.7011686143572621,
"grad_norm": 14.989102480397309,
"learning_rate": 1.8928547952473037e-05,
"loss": 1.052,
"step": 210
},
{
"epoch": 0.7178631051752922,
"grad_norm": 16.72561045227947,
"learning_rate": 1.8839204191418386e-05,
"loss": 1.0237,
"step": 215
},
{
"epoch": 0.7345575959933222,
"grad_norm": 31.90133374330817,
"learning_rate": 1.8746511607784298e-05,
"loss": 1.0203,
"step": 220
},
{
"epoch": 0.7512520868113522,
"grad_norm": 19.165418054306723,
"learning_rate": 1.865050531910062e-05,
"loss": 1.0213,
"step": 225
},
{
"epoch": 0.7679465776293823,
"grad_norm": 34.80379707869377,
"learning_rate": 1.855122169832813e-05,
"loss": 1.0056,
"step": 230
},
{
"epoch": 0.7846410684474123,
"grad_norm": 24.103854680859914,
"learning_rate": 1.844869836007825e-05,
"loss": 1.0089,
"step": 235
},
{
"epoch": 0.8013355592654424,
"grad_norm": 54.705273129166606,
"learning_rate": 1.8342974146362397e-05,
"loss": 0.9912,
"step": 240
},
{
"epoch": 0.8180300500834724,
"grad_norm": 15.340510497628932,
"learning_rate": 1.8234089111876256e-05,
"loss": 0.9708,
"step": 245
},
{
"epoch": 0.8347245409015025,
"grad_norm": 22.865867473032072,
"learning_rate": 1.8122084508824692e-05,
"loss": 0.9491,
"step": 250
},
{
"epoch": 0.8514190317195326,
"grad_norm": 28.125454120215036,
"learning_rate": 1.80070027712929e-05,
"loss": 0.9687,
"step": 255
},
{
"epoch": 0.8681135225375626,
"grad_norm": 48.320081372428405,
"learning_rate": 1.7888887499169816e-05,
"loss": 0.993,
"step": 260
},
{
"epoch": 0.8848080133555927,
"grad_norm": 17.216083695624388,
"learning_rate": 1.7767783441629883e-05,
"loss": 1.0175,
"step": 265
},
{
"epoch": 0.9015025041736227,
"grad_norm": 20.542693182690996,
"learning_rate": 1.7643736480179353e-05,
"loss": 0.9942,
"step": 270
},
{
"epoch": 0.9181969949916527,
"grad_norm": 12.90866880913044,
"learning_rate": 1.7516793611273614e-05,
"loss": 1.0009,
"step": 275
},
{
"epoch": 0.9348914858096828,
"grad_norm": 32.0014460896073,
"learning_rate": 1.7387002928512093e-05,
"loss": 0.9653,
"step": 280
},
{
"epoch": 0.9515859766277128,
"grad_norm": 17.082534238862834,
"learning_rate": 1.725441360441752e-05,
"loss": 0.9329,
"step": 285
},
{
"epoch": 0.9682804674457429,
"grad_norm": 18.406642188469615,
"learning_rate": 1.711907587180642e-05,
"loss": 0.9704,
"step": 290
},
{
"epoch": 0.9849749582637729,
"grad_norm": 25.006696471743908,
"learning_rate": 1.698104100475788e-05,
"loss": 0.9246,
"step": 295
},
{
"epoch": 0.998330550918197,
"eval_loss": 1.026776909828186,
"eval_runtime": 21.5058,
"eval_samples_per_second": 88.953,
"eval_steps_per_second": 2.79,
"step": 299
},
{
"epoch": 1.001669449081803,
"grad_norm": 15.244399722195986,
"learning_rate": 1.684036129918786e-05,
"loss": 0.9362,
"step": 300
},
{
"epoch": 1.018363939899833,
"grad_norm": 5.806089115410544,
"learning_rate": 1.6697090053036344e-05,
"loss": 0.7907,
"step": 305
},
{
"epoch": 1.0350584307178632,
"grad_norm": 5.440155995126969,
"learning_rate": 1.6551281546074863e-05,
"loss": 0.8424,
"step": 310
},
{
"epoch": 1.0517529215358932,
"grad_norm": 7.168040489651867,
"learning_rate": 1.6402991019342073e-05,
"loss": 0.7992,
"step": 315
},
{
"epoch": 1.0684474123539232,
"grad_norm": 11.878722031675423,
"learning_rate": 1.625227465421511e-05,
"loss": 0.847,
"step": 320
},
{
"epoch": 1.0851419031719534,
"grad_norm": 7.545679524982675,
"learning_rate": 1.60991895511247e-05,
"loss": 0.8243,
"step": 325
},
{
"epoch": 1.1018363939899833,
"grad_norm": 13.016594535944764,
"learning_rate": 1.5943793707922086e-05,
"loss": 0.806,
"step": 330
},
{
"epoch": 1.1185308848080133,
"grad_norm": 7.3498644599599885,
"learning_rate": 1.5786145997905952e-05,
"loss": 0.8011,
"step": 335
},
{
"epoch": 1.1352253756260433,
"grad_norm": 9.105183728580222,
"learning_rate": 1.5626306147517665e-05,
"loss": 0.8177,
"step": 340
},
{
"epoch": 1.1519198664440735,
"grad_norm": 25.008986087097295,
"learning_rate": 1.5464334713713312e-05,
"loss": 0.8145,
"step": 345
},
{
"epoch": 1.1686143572621035,
"grad_norm": 12.377325631412589,
"learning_rate": 1.5300293061021084e-05,
"loss": 0.8642,
"step": 350
},
{
"epoch": 1.1853088480801335,
"grad_norm": 11.191522601640232,
"learning_rate": 1.5134243338292686e-05,
"loss": 0.8359,
"step": 355
},
{
"epoch": 1.2020033388981637,
"grad_norm": 17.12714029774076,
"learning_rate": 1.4966248455157622e-05,
"loss": 0.833,
"step": 360
},
{
"epoch": 1.2186978297161937,
"grad_norm": 6.348885873172548,
"learning_rate": 1.4796372058189235e-05,
"loss": 0.7872,
"step": 365
},
{
"epoch": 1.2353923205342237,
"grad_norm": 9.944628643661803,
"learning_rate": 1.4624678506791556e-05,
"loss": 0.8057,
"step": 370
},
{
"epoch": 1.2520868113522536,
"grad_norm": 18.07552847892337,
"learning_rate": 1.445123284881609e-05,
"loss": 0.7829,
"step": 375
},
{
"epoch": 1.2687813021702838,
"grad_norm": 30.335726590753946,
"learning_rate": 1.4276100795917777e-05,
"loss": 0.8013,
"step": 380
},
{
"epoch": 1.2854757929883138,
"grad_norm": 16.608038839902616,
"learning_rate": 1.409934869865945e-05,
"loss": 0.831,
"step": 385
},
{
"epoch": 1.302170283806344,
"grad_norm": 14.333469737425801,
"learning_rate": 1.392104352137426e-05,
"loss": 0.8015,
"step": 390
},
{
"epoch": 1.318864774624374,
"grad_norm": 13.451868406991572,
"learning_rate": 1.3741252816795552e-05,
"loss": 0.8014,
"step": 395
},
{
"epoch": 1.335559265442404,
"grad_norm": 5.9058948036424,
"learning_rate": 1.3560044700463824e-05,
"loss": 0.8,
"step": 400
},
{
"epoch": 1.352253756260434,
"grad_norm": 53.944440853433036,
"learning_rate": 1.3377487824920459e-05,
"loss": 0.8174,
"step": 405
},
{
"epoch": 1.3689482470784642,
"grad_norm": 18.071190540649788,
"learning_rate": 1.3193651353698012e-05,
"loss": 0.8252,
"step": 410
},
{
"epoch": 1.3856427378964942,
"grad_norm": 35.62084029605276,
"learning_rate": 1.30086049351169e-05,
"loss": 0.8221,
"step": 415
},
{
"epoch": 1.4023372287145242,
"grad_norm": 19.99178320201263,
"learning_rate": 1.2822418675898428e-05,
"loss": 0.806,
"step": 420
},
{
"epoch": 1.4190317195325544,
"grad_norm": 27.2927325619504,
"learning_rate": 1.2635163114604131e-05,
"loss": 0.7997,
"step": 425
},
{
"epoch": 1.4357262103505843,
"grad_norm": 28.98127275070747,
"learning_rate": 1.2446909194911552e-05,
"loss": 0.8117,
"step": 430
},
{
"epoch": 1.4524207011686143,
"grad_norm": 12.946012794793734,
"learning_rate": 1.2257728238736468e-05,
"loss": 0.7836,
"step": 435
},
{
"epoch": 1.4691151919866443,
"grad_norm": 11.264077156777665,
"learning_rate": 1.2067691919211879e-05,
"loss": 0.7604,
"step": 440
},
{
"epoch": 1.4858096828046745,
"grad_norm": 21.934680069997075,
"learning_rate": 1.1876872233533909e-05,
"loss": 0.7897,
"step": 445
},
{
"epoch": 1.5025041736227045,
"grad_norm": 15.129234971208845,
"learning_rate": 1.1685341475684935e-05,
"loss": 0.7706,
"step": 450
},
{
"epoch": 1.5191986644407347,
"grad_norm": 18.77833994006315,
"learning_rate": 1.1493172209044259e-05,
"loss": 0.7391,
"step": 455
},
{
"epoch": 1.5358931552587647,
"grad_norm": 34.24893404758549,
"learning_rate": 1.1300437238896758e-05,
"loss": 0.7497,
"step": 460
},
{
"epoch": 1.5525876460767947,
"grad_norm": 20.076138032953008,
"learning_rate": 1.1107209584849845e-05,
"loss": 0.7466,
"step": 465
},
{
"epoch": 1.5692821368948247,
"grad_norm": 16.836099480285057,
"learning_rate": 1.0913562453169241e-05,
"loss": 0.7659,
"step": 470
},
{
"epoch": 1.5859766277128546,
"grad_norm": 22.86729702065199,
"learning_rate": 1.0719569209044047e-05,
"loss": 0.734,
"step": 475
},
{
"epoch": 1.6026711185308848,
"grad_norm": 9.979667643277459,
"learning_rate": 1.0525303348791599e-05,
"loss": 0.7673,
"step": 480
},
{
"epoch": 1.6193656093489148,
"grad_norm": 11.926547743437098,
"learning_rate": 1.0330838472012617e-05,
"loss": 0.7677,
"step": 485
},
{
"epoch": 1.636060100166945,
"grad_norm": 20.884548842874665,
"learning_rate": 1.0136248253707267e-05,
"loss": 0.7589,
"step": 490
},
{
"epoch": 1.652754590984975,
"grad_norm": 20.320143837413003,
"learning_rate": 9.94160641636263e-06,
"loss": 0.7317,
"step": 495
},
{
"epoch": 1.669449081803005,
"grad_norm": 9.784131496022562,
"learning_rate": 9.74698670202218e-06,
"loss": 0.7471,
"step": 500
},
{
"epoch": 1.686143572621035,
"grad_norm": 20.927503924872337,
"learning_rate": 9.552462844347883e-06,
"loss": 0.7476,
"step": 505
},
{
"epoch": 1.702838063439065,
"grad_norm": 16.01855549874665,
"learning_rate": 9.358108540685406e-06,
"loss": 0.7219,
"step": 510
},
{
"epoch": 1.7195325542570952,
"grad_norm": 7.286899337512383,
"learning_rate": 9.163997424143167e-06,
"loss": 0.7291,
"step": 515
},
{
"epoch": 1.7362270450751254,
"grad_norm": 26.516549504841837,
"learning_rate": 8.970203035695662e-06,
"loss": 0.7392,
"step": 520
},
{
"epoch": 1.7529215358931554,
"grad_norm": 17.76687131779593,
"learning_rate": 8.776798796321715e-06,
"loss": 0.7315,
"step": 525
},
{
"epoch": 1.7696160267111853,
"grad_norm": 11.57930379098666,
"learning_rate": 8.583857979188203e-06,
"loss": 0.7289,
"step": 530
},
{
"epoch": 1.7863105175292153,
"grad_norm": 17.537227925038266,
"learning_rate": 8.391453681889772e-06,
"loss": 0.7469,
"step": 535
},
{
"epoch": 1.8030050083472453,
"grad_norm": 10.621593546803865,
"learning_rate": 8.199658798755048e-06,
"loss": 0.777,
"step": 540
},
{
"epoch": 1.8196994991652755,
"grad_norm": 9.251016723609432,
"learning_rate": 8.008545993229897e-06,
"loss": 0.7286,
"step": 545
},
{
"epoch": 1.8363939899833055,
"grad_norm": 15.490160829475185,
"learning_rate": 7.818187670348133e-06,
"loss": 0.7222,
"step": 550
},
{
"epoch": 1.8530884808013357,
"grad_norm": 6.9953239485991325,
"learning_rate": 7.628655949300133e-06,
"loss": 0.7537,
"step": 555
},
{
"epoch": 1.8697829716193657,
"grad_norm": 3.102208689939814,
"learning_rate": 7.440022636109742e-06,
"loss": 0.7422,
"step": 560
},
{
"epoch": 1.8864774624373957,
"grad_norm": 5.758208330036847,
"learning_rate": 7.2523591964298345e-06,
"loss": 0.7179,
"step": 565
},
{
"epoch": 1.9031719532554257,
"grad_norm": 3.812195776537815,
"learning_rate": 7.065736728466832e-06,
"loss": 0.7451,
"step": 570
},
{
"epoch": 1.9198664440734556,
"grad_norm": 6.263849267436109,
"learning_rate": 6.880225936044402e-06,
"loss": 0.7557,
"step": 575
},
{
"epoch": 1.9365609348914858,
"grad_norm": 6.282114744972934,
"learning_rate": 6.695897101816606e-06,
"loss": 0.7483,
"step": 580
},
{
"epoch": 1.9532554257095158,
"grad_norm": 11.855358236520228,
"learning_rate": 6.512820060640608e-06,
"loss": 0.776,
"step": 585
},
{
"epoch": 1.969949916527546,
"grad_norm": 6.706340476505237,
"learning_rate": 6.331064173119008e-06,
"loss": 0.7526,
"step": 590
},
{
"epoch": 1.986644407345576,
"grad_norm": 2.315599430989982,
"learning_rate": 6.150698299321889e-06,
"loss": 0.7512,
"step": 595
},
{
"epoch": 2.0,
"eval_loss": 1.042004108428955,
"eval_runtime": 21.4365,
"eval_samples_per_second": 89.24,
"eval_steps_per_second": 2.799,
"step": 599
},
{
"epoch": 2.003338898163606,
"grad_norm": 7.306932166126154,
"learning_rate": 5.971790772698467e-06,
"loss": 0.6991,
"step": 600
},
{
"epoch": 2.020033388981636,
"grad_norm": 4.3017516619955645,
"learning_rate": 5.794409374188272e-06,
"loss": 0.496,
"step": 605
},
{
"epoch": 2.036727879799666,
"grad_norm": 3.6156424881801286,
"learning_rate": 5.61862130654165e-06,
"loss": 0.4997,
"step": 610
},
{
"epoch": 2.053422370617696,
"grad_norm": 2.694317641326709,
"learning_rate": 5.444493168859304e-06,
"loss": 0.4718,
"step": 615
},
{
"epoch": 2.0701168614357264,
"grad_norm": 2.613652033920123,
"learning_rate": 5.272090931360564e-06,
"loss": 0.4831,
"step": 620
},
{
"epoch": 2.0868113522537564,
"grad_norm": 2.813633749303979,
"learning_rate": 5.101479910389888e-06,
"loss": 0.5005,
"step": 625
},
{
"epoch": 2.1035058430717863,
"grad_norm": 2.5181028219855235,
"learning_rate": 4.932724743671089e-06,
"loss": 0.5251,
"step": 630
},
{
"epoch": 2.1202003338898163,
"grad_norm": 2.4641627321242785,
"learning_rate": 4.765889365818708e-06,
"loss": 0.4994,
"step": 635
},
{
"epoch": 2.1368948247078463,
"grad_norm": 2.2239295194273097,
"learning_rate": 4.601036984115684e-06,
"loss": 0.51,
"step": 640
},
{
"epoch": 2.1535893155258763,
"grad_norm": 4.030891483235892,
"learning_rate": 4.438230054566678e-06,
"loss": 0.4823,
"step": 645
},
{
"epoch": 2.1702838063439067,
"grad_norm": 3.4317624941352762,
"learning_rate": 4.277530258235955e-06,
"loss": 0.5121,
"step": 650
},
{
"epoch": 2.1869782971619367,
"grad_norm": 3.4339442550558377,
"learning_rate": 4.118998477878879e-06,
"loss": 0.5038,
"step": 655
},
{
"epoch": 2.2036727879799667,
"grad_norm": 3.7518155349947526,
"learning_rate": 3.96269477487588e-06,
"loss": 0.5255,
"step": 660
},
{
"epoch": 2.2203672787979967,
"grad_norm": 3.208806245499332,
"learning_rate": 3.8086783664775827e-06,
"loss": 0.4813,
"step": 665
},
{
"epoch": 2.2370617696160267,
"grad_norm": 2.2156022295938635,
"learning_rate": 3.657007603369728e-06,
"loss": 0.4832,
"step": 670
},
{
"epoch": 2.2537562604340566,
"grad_norm": 3.0012443950273804,
"learning_rate": 3.5077399475664474e-06,
"loss": 0.4841,
"step": 675
},
{
"epoch": 2.2704507512520866,
"grad_norm": 3.2803718521147323,
"learning_rate": 3.360931950640185e-06,
"loss": 0.5116,
"step": 680
},
{
"epoch": 2.287145242070117,
"grad_norm": 2.0114178970483985,
"learning_rate": 3.2166392322965423e-06,
"loss": 0.493,
"step": 685
},
{
"epoch": 2.303839732888147,
"grad_norm": 2.409104814481249,
"learning_rate": 3.074916459302211e-06,
"loss": 0.4808,
"step": 690
},
{
"epoch": 2.320534223706177,
"grad_norm": 1.8952913155646967,
"learning_rate": 2.935817324773893e-06,
"loss": 0.485,
"step": 695
},
{
"epoch": 2.337228714524207,
"grad_norm": 2.091074617432718,
"learning_rate": 2.799394527836129e-06,
"loss": 0.482,
"step": 700
},
{
"epoch": 2.353923205342237,
"grad_norm": 2.286915096478497,
"learning_rate": 2.665699753655684e-06,
"loss": 0.4816,
"step": 705
},
{
"epoch": 2.370617696160267,
"grad_norm": 2.9960367822706475,
"learning_rate": 2.5347836538601113e-06,
"loss": 0.4669,
"step": 710
},
{
"epoch": 2.3873121869782974,
"grad_norm": 4.617799542764574,
"learning_rate": 2.406695827347848e-06,
"loss": 0.4802,
"step": 715
},
{
"epoch": 2.4040066777963274,
"grad_norm": 2.6788734186896077,
"learning_rate": 2.281484801497186e-06,
"loss": 0.4777,
"step": 720
},
{
"epoch": 2.4207011686143574,
"grad_norm": 2.217425451473198,
"learning_rate": 2.1591980137811684e-06,
"loss": 0.4762,
"step": 725
},
{
"epoch": 2.4373956594323873,
"grad_norm": 2.5079137151103863,
"learning_rate": 2.0398817937954275e-06,
"loss": 0.4791,
"step": 730
},
{
"epoch": 2.4540901502504173,
"grad_norm": 1.832302823771755,
"learning_rate": 1.923581345705736e-06,
"loss": 0.4604,
"step": 735
},
{
"epoch": 2.4707846410684473,
"grad_norm": 3.3464829987759375,
"learning_rate": 1.8103407311219523e-06,
"loss": 0.4773,
"step": 740
},
{
"epoch": 2.4874791318864773,
"grad_norm": 3.1999456095425995,
"learning_rate": 1.7002028524048354e-06,
"loss": 0.4628,
"step": 745
},
{
"epoch": 2.5041736227045073,
"grad_norm": 2.4738950163595024,
"learning_rate": 1.5932094364120453e-06,
"loss": 0.4759,
"step": 750
},
{
"epoch": 2.5208681135225377,
"grad_norm": 5.463944004312446,
"learning_rate": 1.489401018689488e-06,
"loss": 0.4692,
"step": 755
},
{
"epoch": 2.5375626043405677,
"grad_norm": 3.785050339076787,
"learning_rate": 1.3888169281140284e-06,
"loss": 0.4605,
"step": 760
},
{
"epoch": 2.5542570951585977,
"grad_norm": 2.672219415465402,
"learning_rate": 1.291495271993337e-06,
"loss": 0.4636,
"step": 765
},
{
"epoch": 2.5709515859766277,
"grad_norm": 1.815658692342384,
"learning_rate": 1.1974729216285386e-06,
"loss": 0.4639,
"step": 770
},
{
"epoch": 2.5876460767946576,
"grad_norm": 2.721770682379103,
"learning_rate": 1.1067854983451575e-06,
"loss": 0.4564,
"step": 775
},
{
"epoch": 2.604340567612688,
"grad_norm": 2.722974978900685,
"learning_rate": 1.0194673599976134e-06,
"loss": 0.4504,
"step": 780
},
{
"epoch": 2.621035058430718,
"grad_norm": 2.5630900969446837,
"learning_rate": 9.355515879523858e-07,
"loss": 0.4536,
"step": 785
},
{
"epoch": 2.637729549248748,
"grad_norm": 2.1055399141098032,
"learning_rate": 8.550699745548196e-07,
"loss": 0.4639,
"step": 790
},
{
"epoch": 2.654424040066778,
"grad_norm": 1.7279709418461875,
"learning_rate": 7.780530110842566e-07,
"loss": 0.4796,
"step": 795
},
{
"epoch": 2.671118530884808,
"grad_norm": 1.7404953519601456,
"learning_rate": 7.045298762021125e-07,
"loss": 0.4459,
"step": 800
},
{
"epoch": 2.687813021702838,
"grad_norm": 1.9897437836650076,
"learning_rate": 6.345284248972383e-07,
"loss": 0.4573,
"step": 805
},
{
"epoch": 2.704507512520868,
"grad_norm": 1.7374239194544907,
"learning_rate": 5.680751779327742e-07,
"loss": 0.4481,
"step": 810
},
{
"epoch": 2.721202003338898,
"grad_norm": 1.9063993408126028,
"learning_rate": 5.05195311798491e-07,
"loss": 0.4476,
"step": 815
},
{
"epoch": 2.7378964941569284,
"grad_norm": 2.134180540126151,
"learning_rate": 4.4591264917242195e-07,
"loss": 0.4463,
"step": 820
},
{
"epoch": 2.7545909849749584,
"grad_norm": 1.70679451789591,
"learning_rate": 3.9024964989539227e-07,
"loss": 0.4731,
"step": 825
},
{
"epoch": 2.7712854757929883,
"grad_norm": 1.8624687631277015,
"learning_rate": 3.3822740246188477e-07,
"loss": 0.441,
"step": 830
},
{
"epoch": 2.7879799666110183,
"grad_norm": 1.7487434265005612,
"learning_rate": 2.8986561603044694e-07,
"loss": 0.4634,
"step": 835
},
{
"epoch": 2.8046744574290483,
"grad_norm": 1.7206438184487136,
"learning_rate": 2.4518261295667255e-07,
"loss": 0.4544,
"step": 840
},
{
"epoch": 2.8213689482470787,
"grad_norm": 1.6616178181435346,
"learning_rate": 2.0419532185159796e-07,
"loss": 0.4605,
"step": 845
},
{
"epoch": 2.8380634390651087,
"grad_norm": 2.151120696939262,
"learning_rate": 1.6691927116812002e-07,
"loss": 0.446,
"step": 850
},
{
"epoch": 2.8547579298831387,
"grad_norm": 1.7230509982028626,
"learning_rate": 1.3336858331787993e-07,
"loss": 0.4519,
"step": 855
},
{
"epoch": 2.8714524207011687,
"grad_norm": 1.7412338171439155,
"learning_rate": 1.0355596932085432e-07,
"loss": 0.459,
"step": 860
},
{
"epoch": 2.8881469115191987,
"grad_norm": 2.021167741282332,
"learning_rate": 7.749272398964613e-08,
"loss": 0.4382,
"step": 865
},
{
"epoch": 2.9048414023372287,
"grad_norm": 1.7263112044181872,
"learning_rate": 5.518872165033329e-08,
"loss": 0.4381,
"step": 870
},
{
"epoch": 2.9215358931552586,
"grad_norm": 1.8622740089312937,
"learning_rate": 3.6652412401478875e-08,
"loss": 0.4348,
"step": 875
},
{
"epoch": 2.9382303839732886,
"grad_norm": 1.6946497921936468,
"learning_rate": 2.1890818912728706e-08,
"loss": 0.4498,
"step": 880
},
{
"epoch": 2.9549248747913186,
"grad_norm": 1.6754781633135705,
"learning_rate": 1.0909533764194013e-08,
"loss": 0.4424,
"step": 885
},
{
"epoch": 2.971619365609349,
"grad_norm": 1.7509693982942596,
"learning_rate": 3.7127173276563234e-09,
"loss": 0.4396,
"step": 890
},
{
"epoch": 2.988313856427379,
"grad_norm": 1.688514503466763,
"learning_rate": 3.0309619035495675e-10,
"loss": 0.4573,
"step": 895
},
{
"epoch": 2.994991652754591,
"eval_loss": 1.0773704051971436,
"eval_runtime": 21.4317,
"eval_samples_per_second": 89.26,
"eval_steps_per_second": 2.8,
"step": 897
},
{
"epoch": 2.994991652754591,
"step": 897,
"total_flos": 246978202042368.0,
"train_loss": 0.8908544659880359,
"train_runtime": 5202.4388,
"train_samples_per_second": 22.101,
"train_steps_per_second": 0.172
}
],
"logging_steps": 5,
"max_steps": 897,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": false,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 246978202042368.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}