RALL_RGBCROP_Aug16F-WD01 / trainer_state.json
TanAlexanderlz's picture
Model save
b7b3ded verified
{
"best_global_step": 2601,
"best_metric": 0.8139059304703476,
"best_model_checkpoint": "RALL_RGBCROP_Aug16F-WD01/checkpoint-2601",
"epoch": 11.081744656268054,
"eval_steps": 500,
"global_step": 3462,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0028885037550548816,
"grad_norm": 5.42613410949707,
"learning_rate": 1.2968299711815564e-07,
"loss": 0.6979,
"step": 10
},
{
"epoch": 0.005777007510109763,
"grad_norm": 7.1514387130737305,
"learning_rate": 2.7377521613832854e-07,
"loss": 0.7302,
"step": 20
},
{
"epoch": 0.008665511265164644,
"grad_norm": 5.011861801147461,
"learning_rate": 4.1786743515850145e-07,
"loss": 0.7362,
"step": 30
},
{
"epoch": 0.011554015020219527,
"grad_norm": 5.10536527633667,
"learning_rate": 5.619596541786745e-07,
"loss": 0.6468,
"step": 40
},
{
"epoch": 0.014442518775274409,
"grad_norm": 4.414210796356201,
"learning_rate": 7.060518731988474e-07,
"loss": 0.6643,
"step": 50
},
{
"epoch": 0.01733102253032929,
"grad_norm": 3.883697748184204,
"learning_rate": 8.501440922190203e-07,
"loss": 0.7357,
"step": 60
},
{
"epoch": 0.020219526285384173,
"grad_norm": 6.038342475891113,
"learning_rate": 9.94236311239193e-07,
"loss": 0.713,
"step": 70
},
{
"epoch": 0.023108030040439053,
"grad_norm": 4.572866439819336,
"learning_rate": 1.138328530259366e-06,
"loss": 0.6776,
"step": 80
},
{
"epoch": 0.025996533795493933,
"grad_norm": 4.689591407775879,
"learning_rate": 1.282420749279539e-06,
"loss": 0.7337,
"step": 90
},
{
"epoch": 0.028885037550548817,
"grad_norm": 4.543818950653076,
"learning_rate": 1.426512968299712e-06,
"loss": 0.6663,
"step": 100
},
{
"epoch": 0.0317735413056037,
"grad_norm": 4.83401346206665,
"learning_rate": 1.5706051873198849e-06,
"loss": 0.6594,
"step": 110
},
{
"epoch": 0.03466204506065858,
"grad_norm": 4.072539806365967,
"learning_rate": 1.7146974063400579e-06,
"loss": 0.6605,
"step": 120
},
{
"epoch": 0.03755054881571346,
"grad_norm": 5.373439311981201,
"learning_rate": 1.8587896253602309e-06,
"loss": 0.6367,
"step": 130
},
{
"epoch": 0.040439052570768345,
"grad_norm": 5.04488468170166,
"learning_rate": 2.0028818443804035e-06,
"loss": 0.6649,
"step": 140
},
{
"epoch": 0.043327556325823226,
"grad_norm": 3.912029504776001,
"learning_rate": 2.1469740634005763e-06,
"loss": 0.6293,
"step": 150
},
{
"epoch": 0.046216060080878106,
"grad_norm": 4.424826622009277,
"learning_rate": 2.2910662824207495e-06,
"loss": 0.6533,
"step": 160
},
{
"epoch": 0.049104563835932986,
"grad_norm": 4.445134162902832,
"learning_rate": 2.4351585014409223e-06,
"loss": 0.616,
"step": 170
},
{
"epoch": 0.05199306759098787,
"grad_norm": 5.449381351470947,
"learning_rate": 2.579250720461095e-06,
"loss": 0.6339,
"step": 180
},
{
"epoch": 0.05488157134604275,
"grad_norm": 3.924225330352783,
"learning_rate": 2.7233429394812683e-06,
"loss": 0.5626,
"step": 190
},
{
"epoch": 0.057770075101097634,
"grad_norm": 5.135080814361572,
"learning_rate": 2.867435158501441e-06,
"loss": 0.5723,
"step": 200
},
{
"epoch": 0.060658578856152515,
"grad_norm": 4.920149326324463,
"learning_rate": 3.0115273775216143e-06,
"loss": 0.5679,
"step": 210
},
{
"epoch": 0.0635470826112074,
"grad_norm": 7.151332378387451,
"learning_rate": 3.1556195965417867e-06,
"loss": 0.58,
"step": 220
},
{
"epoch": 0.06643558636626228,
"grad_norm": 5.037742614746094,
"learning_rate": 3.29971181556196e-06,
"loss": 0.5527,
"step": 230
},
{
"epoch": 0.06932409012131716,
"grad_norm": 4.7811102867126465,
"learning_rate": 3.4438040345821327e-06,
"loss": 0.512,
"step": 240
},
{
"epoch": 0.07221259387637204,
"grad_norm": 4.887838840484619,
"learning_rate": 3.587896253602306e-06,
"loss": 0.4884,
"step": 250
},
{
"epoch": 0.07510109763142692,
"grad_norm": 4.093018054962158,
"learning_rate": 3.7319884726224787e-06,
"loss": 0.4964,
"step": 260
},
{
"epoch": 0.0779896013864818,
"grad_norm": 4.6939826011657715,
"learning_rate": 3.876080691642652e-06,
"loss": 0.4429,
"step": 270
},
{
"epoch": 0.08087810514153669,
"grad_norm": 4.340346813201904,
"learning_rate": 4.020172910662825e-06,
"loss": 0.4949,
"step": 280
},
{
"epoch": 0.08347775852108608,
"eval_accuracy": 0.6973415132924335,
"eval_loss": 0.5839780569076538,
"eval_runtime": 83.801,
"eval_samples_per_second": 5.835,
"eval_steps_per_second": 0.74,
"step": 289
},
{
"epoch": 1.0002888503755054,
"grad_norm": 7.082851886749268,
"learning_rate": 4.1642651296829975e-06,
"loss": 0.462,
"step": 290
},
{
"epoch": 1.0031773541305604,
"grad_norm": 8.205607414245605,
"learning_rate": 4.30835734870317e-06,
"loss": 0.443,
"step": 300
},
{
"epoch": 1.0060658578856152,
"grad_norm": 6.79085111618042,
"learning_rate": 4.452449567723343e-06,
"loss": 0.4279,
"step": 310
},
{
"epoch": 1.0089543616406702,
"grad_norm": 5.912543773651123,
"learning_rate": 4.596541786743517e-06,
"loss": 0.4448,
"step": 320
},
{
"epoch": 1.011842865395725,
"grad_norm": 7.633585453033447,
"learning_rate": 4.740634005763689e-06,
"loss": 0.3472,
"step": 330
},
{
"epoch": 1.01473136915078,
"grad_norm": 2.95974063873291,
"learning_rate": 4.884726224783862e-06,
"loss": 0.3795,
"step": 340
},
{
"epoch": 1.0176198729058348,
"grad_norm": 9.113896369934082,
"learning_rate": 4.996789727126807e-06,
"loss": 0.3192,
"step": 350
},
{
"epoch": 1.0205083766608896,
"grad_norm": 6.554640769958496,
"learning_rate": 4.980738362760835e-06,
"loss": 0.3585,
"step": 360
},
{
"epoch": 1.0233968804159446,
"grad_norm": 3.178997278213501,
"learning_rate": 4.9646869983948645e-06,
"loss": 0.3187,
"step": 370
},
{
"epoch": 1.0262853841709993,
"grad_norm": 9.070796012878418,
"learning_rate": 4.948635634028893e-06,
"loss": 0.3489,
"step": 380
},
{
"epoch": 1.0291738879260544,
"grad_norm": 8.085728645324707,
"learning_rate": 4.9325842696629215e-06,
"loss": 0.2994,
"step": 390
},
{
"epoch": 1.0320623916811091,
"grad_norm": 3.467472791671753,
"learning_rate": 4.916532905296951e-06,
"loss": 0.2839,
"step": 400
},
{
"epoch": 1.0349508954361641,
"grad_norm": 11.494449615478516,
"learning_rate": 4.90048154093098e-06,
"loss": 0.2662,
"step": 410
},
{
"epoch": 1.037839399191219,
"grad_norm": 10.576976776123047,
"learning_rate": 4.884430176565008e-06,
"loss": 0.3153,
"step": 420
},
{
"epoch": 1.040727902946274,
"grad_norm": 10.300637245178223,
"learning_rate": 4.868378812199037e-06,
"loss": 0.295,
"step": 430
},
{
"epoch": 1.0436164067013287,
"grad_norm": 2.993881940841675,
"learning_rate": 4.8523274478330665e-06,
"loss": 0.2122,
"step": 440
},
{
"epoch": 1.0465049104563835,
"grad_norm": 7.386906623840332,
"learning_rate": 4.836276083467095e-06,
"loss": 0.3317,
"step": 450
},
{
"epoch": 1.0493934142114385,
"grad_norm": 2.5367188453674316,
"learning_rate": 4.820224719101124e-06,
"loss": 0.2116,
"step": 460
},
{
"epoch": 1.0522819179664933,
"grad_norm": 9.584460258483887,
"learning_rate": 4.804173354735153e-06,
"loss": 0.2566,
"step": 470
},
{
"epoch": 1.0551704217215483,
"grad_norm": 8.785528182983398,
"learning_rate": 4.788121990369181e-06,
"loss": 0.2509,
"step": 480
},
{
"epoch": 1.058058925476603,
"grad_norm": 8.40345573425293,
"learning_rate": 4.772070626003211e-06,
"loss": 0.3175,
"step": 490
},
{
"epoch": 1.060947429231658,
"grad_norm": 5.987034320831299,
"learning_rate": 4.75601926163724e-06,
"loss": 0.2683,
"step": 500
},
{
"epoch": 1.0638359329867129,
"grad_norm": 4.399767875671387,
"learning_rate": 4.7399678972712685e-06,
"loss": 0.1927,
"step": 510
},
{
"epoch": 1.0667244367417679,
"grad_norm": 12.556426048278809,
"learning_rate": 4.723916532905297e-06,
"loss": 0.3256,
"step": 520
},
{
"epoch": 1.0696129404968227,
"grad_norm": 6.317582607269287,
"learning_rate": 4.707865168539326e-06,
"loss": 0.1805,
"step": 530
},
{
"epoch": 1.0725014442518774,
"grad_norm": 7.772289752960205,
"learning_rate": 4.691813804173355e-06,
"loss": 0.192,
"step": 540
},
{
"epoch": 1.0753899480069324,
"grad_norm": 7.010458946228027,
"learning_rate": 4.675762439807384e-06,
"loss": 0.2454,
"step": 550
},
{
"epoch": 1.0782784517619872,
"grad_norm": 6.772615909576416,
"learning_rate": 4.659711075441413e-06,
"loss": 0.1515,
"step": 560
},
{
"epoch": 1.0811669555170422,
"grad_norm": 9.612960815429688,
"learning_rate": 4.643659711075442e-06,
"loss": 0.2616,
"step": 570
},
{
"epoch": 1.0834777585210862,
"eval_accuracy": 0.7955010224948875,
"eval_loss": 0.530440628528595,
"eval_runtime": 77.7344,
"eval_samples_per_second": 6.291,
"eval_steps_per_second": 0.798,
"step": 578
},
{
"epoch": 2.000577700751011,
"grad_norm": 4.917853832244873,
"learning_rate": 4.6276083467094705e-06,
"loss": 0.1349,
"step": 580
},
{
"epoch": 2.003466204506066,
"grad_norm": 4.1992974281311035,
"learning_rate": 4.6115569823435e-06,
"loss": 0.1807,
"step": 590
},
{
"epoch": 2.006354708261121,
"grad_norm": 18.553295135498047,
"learning_rate": 4.595505617977528e-06,
"loss": 0.2002,
"step": 600
},
{
"epoch": 2.009243212016176,
"grad_norm": 7.829350471496582,
"learning_rate": 4.579454253611557e-06,
"loss": 0.1349,
"step": 610
},
{
"epoch": 2.0121317157712304,
"grad_norm": 3.751248836517334,
"learning_rate": 4.563402889245586e-06,
"loss": 0.1752,
"step": 620
},
{
"epoch": 2.0150202195262854,
"grad_norm": 12.89821720123291,
"learning_rate": 4.5473515248796155e-06,
"loss": 0.1294,
"step": 630
},
{
"epoch": 2.0179087232813404,
"grad_norm": 11.471810340881348,
"learning_rate": 4.531300160513644e-06,
"loss": 0.0896,
"step": 640
},
{
"epoch": 2.020797227036395,
"grad_norm": 0.9516859650611877,
"learning_rate": 4.515248796147673e-06,
"loss": 0.141,
"step": 650
},
{
"epoch": 2.02368573079145,
"grad_norm": 11.20914363861084,
"learning_rate": 4.499197431781702e-06,
"loss": 0.1688,
"step": 660
},
{
"epoch": 2.026574234546505,
"grad_norm": 11.292973518371582,
"learning_rate": 4.48314606741573e-06,
"loss": 0.1823,
"step": 670
},
{
"epoch": 2.02946273830156,
"grad_norm": 9.095600128173828,
"learning_rate": 4.46709470304976e-06,
"loss": 0.1181,
"step": 680
},
{
"epoch": 2.0323512420566145,
"grad_norm": 0.7727231383323669,
"learning_rate": 4.451043338683789e-06,
"loss": 0.0527,
"step": 690
},
{
"epoch": 2.0352397458116696,
"grad_norm": 3.2032511234283447,
"learning_rate": 4.4349919743178176e-06,
"loss": 0.077,
"step": 700
},
{
"epoch": 2.0381282495667246,
"grad_norm": 10.461459159851074,
"learning_rate": 4.418940609951846e-06,
"loss": 0.0952,
"step": 710
},
{
"epoch": 2.041016753321779,
"grad_norm": 7.884979248046875,
"learning_rate": 4.402889245585875e-06,
"loss": 0.0498,
"step": 720
},
{
"epoch": 2.043905257076834,
"grad_norm": 1.1473121643066406,
"learning_rate": 4.386837881219904e-06,
"loss": 0.1644,
"step": 730
},
{
"epoch": 2.046793760831889,
"grad_norm": 3.6056792736053467,
"learning_rate": 4.370786516853933e-06,
"loss": 0.1515,
"step": 740
},
{
"epoch": 2.049682264586944,
"grad_norm": 1.1584240198135376,
"learning_rate": 4.354735152487962e-06,
"loss": 0.0563,
"step": 750
},
{
"epoch": 2.0525707683419987,
"grad_norm": 6.910070896148682,
"learning_rate": 4.33868378812199e-06,
"loss": 0.0525,
"step": 760
},
{
"epoch": 2.0554592720970537,
"grad_norm": 1.379238486289978,
"learning_rate": 4.32263242375602e-06,
"loss": 0.0915,
"step": 770
},
{
"epoch": 2.0583477758521087,
"grad_norm": 23.958478927612305,
"learning_rate": 4.306581059390049e-06,
"loss": 0.0828,
"step": 780
},
{
"epoch": 2.0612362796071633,
"grad_norm": 0.20569536089897156,
"learning_rate": 4.290529695024077e-06,
"loss": 0.0821,
"step": 790
},
{
"epoch": 2.0641247833622183,
"grad_norm": 0.5140540599822998,
"learning_rate": 4.274478330658106e-06,
"loss": 0.0866,
"step": 800
},
{
"epoch": 2.0670132871172733,
"grad_norm": 10.027976989746094,
"learning_rate": 4.258426966292135e-06,
"loss": 0.217,
"step": 810
},
{
"epoch": 2.0699017908723283,
"grad_norm": 6.995995998382568,
"learning_rate": 4.242375601926164e-06,
"loss": 0.0857,
"step": 820
},
{
"epoch": 2.072790294627383,
"grad_norm": 0.4062481224536896,
"learning_rate": 4.226324237560193e-06,
"loss": 0.0597,
"step": 830
},
{
"epoch": 2.075678798382438,
"grad_norm": 4.28646993637085,
"learning_rate": 4.210272873194222e-06,
"loss": 0.0872,
"step": 840
},
{
"epoch": 2.078567302137493,
"grad_norm": 9.371219635009766,
"learning_rate": 4.194221508828251e-06,
"loss": 0.0373,
"step": 850
},
{
"epoch": 2.081455805892548,
"grad_norm": 13.69465446472168,
"learning_rate": 4.1781701444622794e-06,
"loss": 0.078,
"step": 860
},
{
"epoch": 2.083477758521086,
"eval_accuracy": 0.7975460122699386,
"eval_loss": 0.7102476358413696,
"eval_runtime": 69.3294,
"eval_samples_per_second": 7.053,
"eval_steps_per_second": 0.894,
"step": 867
},
{
"epoch": 3.0008665511265167,
"grad_norm": 0.9676089882850647,
"learning_rate": 4.162118780096309e-06,
"loss": 0.1413,
"step": 870
},
{
"epoch": 3.0037550548815712,
"grad_norm": 0.3187597095966339,
"learning_rate": 4.146067415730337e-06,
"loss": 0.0649,
"step": 880
},
{
"epoch": 3.0066435586366262,
"grad_norm": 10.905332565307617,
"learning_rate": 4.130016051364366e-06,
"loss": 0.0185,
"step": 890
},
{
"epoch": 3.0095320623916813,
"grad_norm": 2.329608917236328,
"learning_rate": 4.113964686998395e-06,
"loss": 0.0381,
"step": 900
},
{
"epoch": 3.012420566146736,
"grad_norm": 11.575812339782715,
"learning_rate": 4.0979133226324245e-06,
"loss": 0.0537,
"step": 910
},
{
"epoch": 3.015309069901791,
"grad_norm": 1.2669545412063599,
"learning_rate": 4.081861958266453e-06,
"loss": 0.0144,
"step": 920
},
{
"epoch": 3.018197573656846,
"grad_norm": 0.1274038404226303,
"learning_rate": 4.0658105939004815e-06,
"loss": 0.0468,
"step": 930
},
{
"epoch": 3.021086077411901,
"grad_norm": 0.11354996263980865,
"learning_rate": 4.049759229534511e-06,
"loss": 0.0452,
"step": 940
},
{
"epoch": 3.0239745811669554,
"grad_norm": 0.11091677844524384,
"learning_rate": 4.033707865168539e-06,
"loss": 0.0251,
"step": 950
},
{
"epoch": 3.0268630849220104,
"grad_norm": 0.12637047469615936,
"learning_rate": 4.017656500802569e-06,
"loss": 0.054,
"step": 960
},
{
"epoch": 3.0297515886770654,
"grad_norm": 3.8349571228027344,
"learning_rate": 4.001605136436598e-06,
"loss": 0.0724,
"step": 970
},
{
"epoch": 3.03264009243212,
"grad_norm": 2.915048360824585,
"learning_rate": 3.9855537720706265e-06,
"loss": 0.0171,
"step": 980
},
{
"epoch": 3.035528596187175,
"grad_norm": 0.14977912604808807,
"learning_rate": 3.969502407704655e-06,
"loss": 0.0054,
"step": 990
},
{
"epoch": 3.03841709994223,
"grad_norm": 0.3599390983581543,
"learning_rate": 3.953451043338684e-06,
"loss": 0.0537,
"step": 1000
},
{
"epoch": 3.041305603697285,
"grad_norm": 0.08736338466405869,
"learning_rate": 3.937399678972713e-06,
"loss": 0.0324,
"step": 1010
},
{
"epoch": 3.0441941074523395,
"grad_norm": 4.6888322830200195,
"learning_rate": 3.921348314606742e-06,
"loss": 0.0271,
"step": 1020
},
{
"epoch": 3.0470826112073945,
"grad_norm": 0.21701326966285706,
"learning_rate": 3.905296950240771e-06,
"loss": 0.0253,
"step": 1030
},
{
"epoch": 3.0499711149624495,
"grad_norm": 1.4320465326309204,
"learning_rate": 3.8892455858748e-06,
"loss": 0.0311,
"step": 1040
},
{
"epoch": 3.0528596187175046,
"grad_norm": 0.6287254095077515,
"learning_rate": 3.8731942215088285e-06,
"loss": 0.0027,
"step": 1050
},
{
"epoch": 3.055748122472559,
"grad_norm": 0.17181923985481262,
"learning_rate": 3.857142857142858e-06,
"loss": 0.063,
"step": 1060
},
{
"epoch": 3.058636626227614,
"grad_norm": 0.20554514229297638,
"learning_rate": 3.841091492776886e-06,
"loss": 0.0136,
"step": 1070
},
{
"epoch": 3.061525129982669,
"grad_norm": 0.05833299085497856,
"learning_rate": 3.825040128410915e-06,
"loss": 0.0192,
"step": 1080
},
{
"epoch": 3.0644136337377237,
"grad_norm": 0.08491459488868713,
"learning_rate": 3.808988764044944e-06,
"loss": 0.0216,
"step": 1090
},
{
"epoch": 3.0673021374927787,
"grad_norm": 30.636640548706055,
"learning_rate": 3.792937399678973e-06,
"loss": 0.0334,
"step": 1100
},
{
"epoch": 3.0701906412478337,
"grad_norm": 0.24978935718536377,
"learning_rate": 3.776886035313002e-06,
"loss": 0.0338,
"step": 1110
},
{
"epoch": 3.0730791450028887,
"grad_norm": 6.24122428894043,
"learning_rate": 3.7608346709470305e-06,
"loss": 0.0096,
"step": 1120
},
{
"epoch": 3.0759676487579433,
"grad_norm": 0.10055310279130936,
"learning_rate": 3.7447833065810594e-06,
"loss": 0.0078,
"step": 1130
},
{
"epoch": 3.0788561525129983,
"grad_norm": 0.07958950102329254,
"learning_rate": 3.7287319422150888e-06,
"loss": 0.0153,
"step": 1140
},
{
"epoch": 3.0817446562680533,
"grad_norm": 0.2819121479988098,
"learning_rate": 3.7126805778491177e-06,
"loss": 0.0089,
"step": 1150
},
{
"epoch": 3.083477758521086,
"eval_accuracy": 0.7934560327198364,
"eval_loss": 0.9156056046485901,
"eval_runtime": 77.6168,
"eval_samples_per_second": 6.3,
"eval_steps_per_second": 0.799,
"step": 1156
},
{
"epoch": 4.001155401502022,
"grad_norm": 0.28028979897499084,
"learning_rate": 3.6966292134831466e-06,
"loss": 0.0059,
"step": 1160
},
{
"epoch": 4.004043905257077,
"grad_norm": 0.09656143933534622,
"learning_rate": 3.680577849117175e-06,
"loss": 0.026,
"step": 1170
},
{
"epoch": 4.006932409012132,
"grad_norm": 3.490778923034668,
"learning_rate": 3.664526484751204e-06,
"loss": 0.0105,
"step": 1180
},
{
"epoch": 4.009820912767187,
"grad_norm": 0.14486035704612732,
"learning_rate": 3.648475120385233e-06,
"loss": 0.0015,
"step": 1190
},
{
"epoch": 4.012709416522242,
"grad_norm": 0.04033642262220383,
"learning_rate": 3.6324237560192623e-06,
"loss": 0.015,
"step": 1200
},
{
"epoch": 4.015597920277297,
"grad_norm": 0.35192564129829407,
"learning_rate": 3.6163723916532904e-06,
"loss": 0.0015,
"step": 1210
},
{
"epoch": 4.018486424032352,
"grad_norm": 1.2167606353759766,
"learning_rate": 3.6003210272873197e-06,
"loss": 0.0062,
"step": 1220
},
{
"epoch": 4.021374927787406,
"grad_norm": 0.30561357736587524,
"learning_rate": 3.5842696629213486e-06,
"loss": 0.0014,
"step": 1230
},
{
"epoch": 4.024263431542461,
"grad_norm": 0.07887325435876846,
"learning_rate": 3.5682182985553776e-06,
"loss": 0.0017,
"step": 1240
},
{
"epoch": 4.027151935297516,
"grad_norm": 0.08028972148895264,
"learning_rate": 3.5521669341894065e-06,
"loss": 0.0013,
"step": 1250
},
{
"epoch": 4.030040439052571,
"grad_norm": 0.026913689449429512,
"learning_rate": 3.536115569823435e-06,
"loss": 0.0634,
"step": 1260
},
{
"epoch": 4.032928942807626,
"grad_norm": 0.028358453884720802,
"learning_rate": 3.520064205457464e-06,
"loss": 0.0012,
"step": 1270
},
{
"epoch": 4.035817446562681,
"grad_norm": 0.3259369134902954,
"learning_rate": 3.5040128410914932e-06,
"loss": 0.0018,
"step": 1280
},
{
"epoch": 4.038705950317736,
"grad_norm": 0.033729150891304016,
"learning_rate": 3.487961476725522e-06,
"loss": 0.0252,
"step": 1290
},
{
"epoch": 4.04159445407279,
"grad_norm": 0.07054167240858078,
"learning_rate": 3.471910112359551e-06,
"loss": 0.001,
"step": 1300
},
{
"epoch": 4.044482957827845,
"grad_norm": 0.13403773307800293,
"learning_rate": 3.4558587479935796e-06,
"loss": 0.011,
"step": 1310
},
{
"epoch": 4.0473714615829,
"grad_norm": 0.036013245582580566,
"learning_rate": 3.4398073836276085e-06,
"loss": 0.0031,
"step": 1320
},
{
"epoch": 4.050259965337955,
"grad_norm": 0.3583677411079407,
"learning_rate": 3.4237560192616374e-06,
"loss": 0.0012,
"step": 1330
},
{
"epoch": 4.05314846909301,
"grad_norm": 0.03582717850804329,
"learning_rate": 3.4077046548956668e-06,
"loss": 0.0018,
"step": 1340
},
{
"epoch": 4.056036972848065,
"grad_norm": 0.021955780684947968,
"learning_rate": 3.391653290529695e-06,
"loss": 0.0013,
"step": 1350
},
{
"epoch": 4.05892547660312,
"grad_norm": 0.03652675449848175,
"learning_rate": 3.375601926163724e-06,
"loss": 0.0011,
"step": 1360
},
{
"epoch": 4.061813980358174,
"grad_norm": 0.02935628965497017,
"learning_rate": 3.359550561797753e-06,
"loss": 0.0007,
"step": 1370
},
{
"epoch": 4.064702484113229,
"grad_norm": 0.020217472687363625,
"learning_rate": 3.343499197431782e-06,
"loss": 0.0012,
"step": 1380
},
{
"epoch": 4.067590987868284,
"grad_norm": 13.410699844360352,
"learning_rate": 3.327447833065811e-06,
"loss": 0.0043,
"step": 1390
},
{
"epoch": 4.070479491623339,
"grad_norm": 0.04453463479876518,
"learning_rate": 3.3113964686998394e-06,
"loss": 0.0009,
"step": 1400
},
{
"epoch": 4.073367995378394,
"grad_norm": 0.05309749022126198,
"learning_rate": 3.2953451043338684e-06,
"loss": 0.0015,
"step": 1410
},
{
"epoch": 4.076256499133449,
"grad_norm": 0.15033993124961853,
"learning_rate": 3.2792937399678977e-06,
"loss": 0.0011,
"step": 1420
},
{
"epoch": 4.079145002888504,
"grad_norm": 0.044659897685050964,
"learning_rate": 3.2632423756019266e-06,
"loss": 0.0289,
"step": 1430
},
{
"epoch": 4.082033506643558,
"grad_norm": 0.10378444194793701,
"learning_rate": 3.2471910112359555e-06,
"loss": 0.0012,
"step": 1440
},
{
"epoch": 4.083477758521086,
"eval_accuracy": 0.7914110429447853,
"eval_loss": 1.0430103540420532,
"eval_runtime": 66.2163,
"eval_samples_per_second": 7.385,
"eval_steps_per_second": 0.936,
"step": 1445
},
{
"epoch": 5.001444251877527,
"grad_norm": 0.016393378376960754,
"learning_rate": 3.231139646869984e-06,
"loss": 0.0007,
"step": 1450
},
{
"epoch": 5.004332755632582,
"grad_norm": 0.023797810077667236,
"learning_rate": 3.215088282504013e-06,
"loss": 0.0009,
"step": 1460
},
{
"epoch": 5.007221259387637,
"grad_norm": 0.03091173619031906,
"learning_rate": 3.199036918138042e-06,
"loss": 0.0155,
"step": 1470
},
{
"epoch": 5.010109763142692,
"grad_norm": 0.07352596521377563,
"learning_rate": 3.1829855537720712e-06,
"loss": 0.0036,
"step": 1480
},
{
"epoch": 5.012998266897747,
"grad_norm": 0.01384744606912136,
"learning_rate": 3.1669341894060997e-06,
"loss": 0.0006,
"step": 1490
},
{
"epoch": 5.015886770652802,
"grad_norm": 0.048492494970560074,
"learning_rate": 3.1508828250401286e-06,
"loss": 0.0035,
"step": 1500
},
{
"epoch": 5.018775274407857,
"grad_norm": 0.05079851299524307,
"learning_rate": 3.1348314606741576e-06,
"loss": 0.0074,
"step": 1510
},
{
"epoch": 5.021663778162911,
"grad_norm": 0.08019035309553146,
"learning_rate": 3.1187800963081865e-06,
"loss": 0.001,
"step": 1520
},
{
"epoch": 5.024552281917966,
"grad_norm": 0.014834980480372906,
"learning_rate": 3.1027287319422154e-06,
"loss": 0.0008,
"step": 1530
},
{
"epoch": 5.027440785673021,
"grad_norm": 0.05213601887226105,
"learning_rate": 3.086677367576244e-06,
"loss": 0.0007,
"step": 1540
},
{
"epoch": 5.030329289428076,
"grad_norm": 0.03506625443696976,
"learning_rate": 3.0706260032102732e-06,
"loss": 0.0006,
"step": 1550
},
{
"epoch": 5.033217793183131,
"grad_norm": 0.02491496317088604,
"learning_rate": 3.054574638844302e-06,
"loss": 0.0007,
"step": 1560
},
{
"epoch": 5.036106296938186,
"grad_norm": 8.909887313842773,
"learning_rate": 3.038523274478331e-06,
"loss": 0.0098,
"step": 1570
},
{
"epoch": 5.038994800693241,
"grad_norm": 0.026988506317138672,
"learning_rate": 3.02247191011236e-06,
"loss": 0.0006,
"step": 1580
},
{
"epoch": 5.041883304448295,
"grad_norm": 0.09933661669492722,
"learning_rate": 3.0064205457463885e-06,
"loss": 0.0006,
"step": 1590
},
{
"epoch": 5.04477180820335,
"grad_norm": 0.019388403743505478,
"learning_rate": 2.9903691813804174e-06,
"loss": 0.0006,
"step": 1600
},
{
"epoch": 5.047660311958405,
"grad_norm": 0.04846403747797012,
"learning_rate": 2.9743178170144463e-06,
"loss": 0.0011,
"step": 1610
},
{
"epoch": 5.05054881571346,
"grad_norm": 0.019278401508927345,
"learning_rate": 2.9582664526484757e-06,
"loss": 0.0006,
"step": 1620
},
{
"epoch": 5.053437319468515,
"grad_norm": 0.023772340267896652,
"learning_rate": 2.942215088282504e-06,
"loss": 0.0007,
"step": 1630
},
{
"epoch": 5.05632582322357,
"grad_norm": 0.13470448553562164,
"learning_rate": 2.926163723916533e-06,
"loss": 0.0006,
"step": 1640
},
{
"epoch": 5.059214326978625,
"grad_norm": 0.016751592978835106,
"learning_rate": 2.910112359550562e-06,
"loss": 0.0006,
"step": 1650
},
{
"epoch": 5.06210283073368,
"grad_norm": 0.05011456087231636,
"learning_rate": 2.894060995184591e-06,
"loss": 0.0004,
"step": 1660
},
{
"epoch": 5.0649913344887345,
"grad_norm": 3.883164405822754,
"learning_rate": 2.87800963081862e-06,
"loss": 0.0048,
"step": 1670
},
{
"epoch": 5.0678798382437895,
"grad_norm": 0.014229238964617252,
"learning_rate": 2.8619582664526484e-06,
"loss": 0.0004,
"step": 1680
},
{
"epoch": 5.0707683419988445,
"grad_norm": 0.029299357905983925,
"learning_rate": 2.8459069020866777e-06,
"loss": 0.0004,
"step": 1690
},
{
"epoch": 5.0736568457538995,
"grad_norm": 0.0175183042883873,
"learning_rate": 2.8298555377207066e-06,
"loss": 0.0005,
"step": 1700
},
{
"epoch": 5.0765453495089545,
"grad_norm": 0.34429457783699036,
"learning_rate": 2.8138041733547356e-06,
"loss": 0.0008,
"step": 1710
},
{
"epoch": 5.0794338532640095,
"grad_norm": 0.033084671944379807,
"learning_rate": 2.797752808988764e-06,
"loss": 0.0008,
"step": 1720
},
{
"epoch": 5.0823223570190645,
"grad_norm": 0.10312939435243607,
"learning_rate": 2.781701444622793e-06,
"loss": 0.0005,
"step": 1730
},
{
"epoch": 5.083477758521086,
"eval_accuracy": 0.803680981595092,
"eval_loss": 1.052091121673584,
"eval_runtime": 70.6742,
"eval_samples_per_second": 6.919,
"eval_steps_per_second": 0.877,
"step": 1734
},
{
"epoch": 6.001733102253033,
"grad_norm": 0.024799736216664314,
"learning_rate": 2.765650080256822e-06,
"loss": 0.0004,
"step": 1740
},
{
"epoch": 6.0046216060080875,
"grad_norm": 0.007654332090169191,
"learning_rate": 2.7495987158908512e-06,
"loss": 0.0004,
"step": 1750
},
{
"epoch": 6.0075101097631425,
"grad_norm": 0.04988357052206993,
"learning_rate": 2.73354735152488e-06,
"loss": 0.0004,
"step": 1760
},
{
"epoch": 6.0103986135181975,
"grad_norm": 0.020988399162888527,
"learning_rate": 2.7174959871589087e-06,
"loss": 0.0094,
"step": 1770
},
{
"epoch": 6.0132871172732525,
"grad_norm": 0.01697859913110733,
"learning_rate": 2.7014446227929376e-06,
"loss": 0.0014,
"step": 1780
},
{
"epoch": 6.0161756210283075,
"grad_norm": 0.0390143021941185,
"learning_rate": 2.6853932584269665e-06,
"loss": 0.0003,
"step": 1790
},
{
"epoch": 6.0190641247833625,
"grad_norm": 0.11428093165159225,
"learning_rate": 2.6693418940609954e-06,
"loss": 0.0076,
"step": 1800
},
{
"epoch": 6.0219526285384175,
"grad_norm": 0.008388683199882507,
"learning_rate": 2.6532905296950243e-06,
"loss": 0.0006,
"step": 1810
},
{
"epoch": 6.024841132293472,
"grad_norm": 0.01707868091762066,
"learning_rate": 2.637239165329053e-06,
"loss": 0.0004,
"step": 1820
},
{
"epoch": 6.027729636048527,
"grad_norm": 0.02858269028365612,
"learning_rate": 2.621187800963082e-06,
"loss": 0.0006,
"step": 1830
},
{
"epoch": 6.030618139803582,
"grad_norm": 0.018801476806402206,
"learning_rate": 2.605136436597111e-06,
"loss": 0.0004,
"step": 1840
},
{
"epoch": 6.033506643558637,
"grad_norm": 0.038629643619060516,
"learning_rate": 2.58908507223114e-06,
"loss": 0.0004,
"step": 1850
},
{
"epoch": 6.036395147313692,
"grad_norm": 0.01919536292552948,
"learning_rate": 2.5730337078651685e-06,
"loss": 0.0005,
"step": 1860
},
{
"epoch": 6.039283651068747,
"grad_norm": 0.01720685325562954,
"learning_rate": 2.5569823434991974e-06,
"loss": 0.0004,
"step": 1870
},
{
"epoch": 6.042172154823802,
"grad_norm": 0.07143541425466537,
"learning_rate": 2.5409309791332264e-06,
"loss": 0.0004,
"step": 1880
},
{
"epoch": 6.045060658578856,
"grad_norm": 0.026604874059557915,
"learning_rate": 2.5248796147672557e-06,
"loss": 0.0004,
"step": 1890
},
{
"epoch": 6.047949162333911,
"grad_norm": 0.012191304937005043,
"learning_rate": 2.5088282504012846e-06,
"loss": 0.0003,
"step": 1900
},
{
"epoch": 6.050837666088966,
"grad_norm": 0.021905358880758286,
"learning_rate": 2.492776886035313e-06,
"loss": 0.0004,
"step": 1910
},
{
"epoch": 6.053726169844021,
"grad_norm": 0.006233502645045519,
"learning_rate": 2.476725521669342e-06,
"loss": 0.0011,
"step": 1920
},
{
"epoch": 6.056614673599076,
"grad_norm": 0.04028209298849106,
"learning_rate": 2.460674157303371e-06,
"loss": 0.0003,
"step": 1930
},
{
"epoch": 6.059503177354131,
"grad_norm": 0.2698580026626587,
"learning_rate": 2.4446227929374e-06,
"loss": 0.0004,
"step": 1940
},
{
"epoch": 6.062391681109186,
"grad_norm": 0.01763250306248665,
"learning_rate": 2.428571428571429e-06,
"loss": 0.0006,
"step": 1950
},
{
"epoch": 6.06528018486424,
"grad_norm": 0.04466104507446289,
"learning_rate": 2.4125200642054577e-06,
"loss": 0.0018,
"step": 1960
},
{
"epoch": 6.068168688619295,
"grad_norm": 0.024800406768918037,
"learning_rate": 2.3964686998394866e-06,
"loss": 0.0072,
"step": 1970
},
{
"epoch": 6.07105719237435,
"grad_norm": 0.2319115251302719,
"learning_rate": 2.3804173354735156e-06,
"loss": 0.0004,
"step": 1980
},
{
"epoch": 6.073945696129405,
"grad_norm": 0.04877489432692528,
"learning_rate": 2.364365971107544e-06,
"loss": 0.0003,
"step": 1990
},
{
"epoch": 6.07683419988446,
"grad_norm": 0.4406111538410187,
"learning_rate": 2.3483146067415734e-06,
"loss": 0.0007,
"step": 2000
},
{
"epoch": 6.079722703639515,
"grad_norm": 0.013905012980103493,
"learning_rate": 2.3322632423756023e-06,
"loss": 0.0003,
"step": 2010
},
{
"epoch": 6.08261120739457,
"grad_norm": 0.02120201103389263,
"learning_rate": 2.316211878009631e-06,
"loss": 0.0006,
"step": 2020
},
{
"epoch": 6.083477758521086,
"eval_accuracy": 0.8118609406952966,
"eval_loss": 1.0801777839660645,
"eval_runtime": 71.8007,
"eval_samples_per_second": 6.811,
"eval_steps_per_second": 0.864,
"step": 2023
},
{
"epoch": 7.002021952628539,
"grad_norm": 0.08827612549066544,
"learning_rate": 2.30016051364366e-06,
"loss": 0.0003,
"step": 2030
},
{
"epoch": 7.004910456383593,
"grad_norm": 0.2227260172367096,
"learning_rate": 2.2841091492776887e-06,
"loss": 0.0004,
"step": 2040
},
{
"epoch": 7.007798960138648,
"grad_norm": 0.010085574351251125,
"learning_rate": 2.2680577849117176e-06,
"loss": 0.0003,
"step": 2050
},
{
"epoch": 7.010687463893703,
"grad_norm": 0.07205743342638016,
"learning_rate": 2.2520064205457465e-06,
"loss": 0.0014,
"step": 2060
},
{
"epoch": 7.013575967648758,
"grad_norm": 0.1408742517232895,
"learning_rate": 2.2359550561797754e-06,
"loss": 0.0004,
"step": 2070
},
{
"epoch": 7.016464471403813,
"grad_norm": 0.033056750893592834,
"learning_rate": 2.2199036918138043e-06,
"loss": 0.0003,
"step": 2080
},
{
"epoch": 7.019352975158868,
"grad_norm": 0.008353917859494686,
"learning_rate": 2.2038523274478333e-06,
"loss": 0.0002,
"step": 2090
},
{
"epoch": 7.022241478913923,
"grad_norm": 0.007533412892371416,
"learning_rate": 2.187800963081862e-06,
"loss": 0.0003,
"step": 2100
},
{
"epoch": 7.025129982668977,
"grad_norm": 0.009329171851277351,
"learning_rate": 2.171749598715891e-06,
"loss": 0.0003,
"step": 2110
},
{
"epoch": 7.028018486424032,
"grad_norm": 0.006963707972317934,
"learning_rate": 2.15569823434992e-06,
"loss": 0.0003,
"step": 2120
},
{
"epoch": 7.030906990179087,
"grad_norm": 0.01631193608045578,
"learning_rate": 2.1396468699839485e-06,
"loss": 0.0003,
"step": 2130
},
{
"epoch": 7.033795493934142,
"grad_norm": 0.015248533338308334,
"learning_rate": 2.123595505617978e-06,
"loss": 0.0003,
"step": 2140
},
{
"epoch": 7.036683997689197,
"grad_norm": 0.005624765995889902,
"learning_rate": 2.1075441412520064e-06,
"loss": 0.0013,
"step": 2150
},
{
"epoch": 7.039572501444252,
"grad_norm": 0.007767071016132832,
"learning_rate": 2.0914927768860353e-06,
"loss": 0.0002,
"step": 2160
},
{
"epoch": 7.042461005199307,
"grad_norm": 0.05685372278094292,
"learning_rate": 2.0754414125200646e-06,
"loss": 0.0019,
"step": 2170
},
{
"epoch": 7.045349508954362,
"grad_norm": 0.010567046701908112,
"learning_rate": 2.059390048154093e-06,
"loss": 0.0003,
"step": 2180
},
{
"epoch": 7.048238012709416,
"grad_norm": 0.00699876993894577,
"learning_rate": 2.043338683788122e-06,
"loss": 0.0006,
"step": 2190
},
{
"epoch": 7.051126516464471,
"grad_norm": 0.023216497153043747,
"learning_rate": 2.027287319422151e-06,
"loss": 0.0006,
"step": 2200
},
{
"epoch": 7.054015020219526,
"grad_norm": 0.3829960525035858,
"learning_rate": 2.01123595505618e-06,
"loss": 0.0003,
"step": 2210
},
{
"epoch": 7.056903523974581,
"grad_norm": 0.01023771334439516,
"learning_rate": 1.995184590690209e-06,
"loss": 0.0005,
"step": 2220
},
{
"epoch": 7.059792027729636,
"grad_norm": 0.013453644700348377,
"learning_rate": 1.9791332263242377e-06,
"loss": 0.0003,
"step": 2230
},
{
"epoch": 7.062680531484691,
"grad_norm": 0.009664042852818966,
"learning_rate": 1.9630818619582666e-06,
"loss": 0.0003,
"step": 2240
},
{
"epoch": 7.065569035239746,
"grad_norm": 0.02549542300403118,
"learning_rate": 1.9470304975922956e-06,
"loss": 0.0008,
"step": 2250
},
{
"epoch": 7.0684575389948,
"grad_norm": 0.016215378418564796,
"learning_rate": 1.9309791332263245e-06,
"loss": 0.0002,
"step": 2260
},
{
"epoch": 7.071346042749855,
"grad_norm": 0.014920054003596306,
"learning_rate": 1.9149277688603534e-06,
"loss": 0.0002,
"step": 2270
},
{
"epoch": 7.07423454650491,
"grad_norm": 0.00620607752352953,
"learning_rate": 1.8988764044943821e-06,
"loss": 0.0003,
"step": 2280
},
{
"epoch": 7.077123050259965,
"grad_norm": 0.015935564413666725,
"learning_rate": 1.882825040128411e-06,
"loss": 0.0003,
"step": 2290
},
{
"epoch": 7.08001155401502,
"grad_norm": 0.0060450248420238495,
"learning_rate": 1.86677367576244e-06,
"loss": 0.0002,
"step": 2300
},
{
"epoch": 7.082900057770075,
"grad_norm": 0.012453804723918438,
"learning_rate": 1.8507223113964689e-06,
"loss": 0.0002,
"step": 2310
},
{
"epoch": 7.083477758521086,
"eval_accuracy": 0.8077709611451943,
"eval_loss": 1.1695784330368042,
"eval_runtime": 65.8882,
"eval_samples_per_second": 7.422,
"eval_steps_per_second": 0.941,
"step": 2312
},
{
"epoch": 8.002310803004043,
"grad_norm": 0.6690245866775513,
"learning_rate": 1.8346709470304978e-06,
"loss": 0.0007,
"step": 2320
},
{
"epoch": 8.0051993067591,
"grad_norm": 0.0172465480864048,
"learning_rate": 1.8186195826645267e-06,
"loss": 0.0002,
"step": 2330
},
{
"epoch": 8.008087810514153,
"grad_norm": 0.04959592595696449,
"learning_rate": 1.8025682182985554e-06,
"loss": 0.0003,
"step": 2340
},
{
"epoch": 8.01097631426921,
"grad_norm": 0.008038152940571308,
"learning_rate": 1.7865168539325846e-06,
"loss": 0.0002,
"step": 2350
},
{
"epoch": 8.013864818024263,
"grad_norm": 0.009810101240873337,
"learning_rate": 1.7704654895666133e-06,
"loss": 0.0002,
"step": 2360
},
{
"epoch": 8.016753321779317,
"grad_norm": 0.00564626744017005,
"learning_rate": 1.7544141252006422e-06,
"loss": 0.0002,
"step": 2370
},
{
"epoch": 8.019641825534373,
"grad_norm": 0.010218638926744461,
"learning_rate": 1.738362760834671e-06,
"loss": 0.0002,
"step": 2380
},
{
"epoch": 8.022530329289427,
"grad_norm": 0.008607611060142517,
"learning_rate": 1.7223113964687e-06,
"loss": 0.0002,
"step": 2390
},
{
"epoch": 8.025418833044483,
"grad_norm": 0.006128065288066864,
"learning_rate": 1.706260032102729e-06,
"loss": 0.0003,
"step": 2400
},
{
"epoch": 8.028307336799537,
"grad_norm": 0.0036342665553092957,
"learning_rate": 1.6902086677367576e-06,
"loss": 0.0007,
"step": 2410
},
{
"epoch": 8.031195840554593,
"grad_norm": 0.008177458308637142,
"learning_rate": 1.6741573033707868e-06,
"loss": 0.0003,
"step": 2420
},
{
"epoch": 8.034084344309647,
"grad_norm": 0.009102354757487774,
"learning_rate": 1.6581059390048155e-06,
"loss": 0.0006,
"step": 2430
},
{
"epoch": 8.036972848064703,
"grad_norm": 0.012548357248306274,
"learning_rate": 1.6420545746388444e-06,
"loss": 0.0002,
"step": 2440
},
{
"epoch": 8.039861351819757,
"grad_norm": 0.012381269596517086,
"learning_rate": 1.6260032102728735e-06,
"loss": 0.0002,
"step": 2450
},
{
"epoch": 8.042749855574812,
"grad_norm": 0.008270743303000927,
"learning_rate": 1.6099518459069023e-06,
"loss": 0.0002,
"step": 2460
},
{
"epoch": 8.045638359329867,
"grad_norm": 0.008984532207250595,
"learning_rate": 1.5939004815409312e-06,
"loss": 0.0003,
"step": 2470
},
{
"epoch": 8.048526863084922,
"grad_norm": 0.007146380841732025,
"learning_rate": 1.5778491171749599e-06,
"loss": 0.0002,
"step": 2480
},
{
"epoch": 8.051415366839977,
"grad_norm": 0.013866865076124668,
"learning_rate": 1.561797752808989e-06,
"loss": 0.0002,
"step": 2490
},
{
"epoch": 8.054303870595032,
"grad_norm": 0.01275590155273676,
"learning_rate": 1.5457463884430177e-06,
"loss": 0.0002,
"step": 2500
},
{
"epoch": 8.057192374350088,
"grad_norm": 0.008473788388073444,
"learning_rate": 1.5296950240770466e-06,
"loss": 0.0002,
"step": 2510
},
{
"epoch": 8.060080878105142,
"grad_norm": 0.0103828813880682,
"learning_rate": 1.5136436597110758e-06,
"loss": 0.0002,
"step": 2520
},
{
"epoch": 8.062969381860196,
"grad_norm": 0.005070853512734175,
"learning_rate": 1.4975922953451045e-06,
"loss": 0.0002,
"step": 2530
},
{
"epoch": 8.065857885615252,
"grad_norm": 0.009254480712115765,
"learning_rate": 1.4815409309791334e-06,
"loss": 0.0002,
"step": 2540
},
{
"epoch": 8.068746389370306,
"grad_norm": 0.05558139830827713,
"learning_rate": 1.4654895666131621e-06,
"loss": 0.0002,
"step": 2550
},
{
"epoch": 8.071634893125362,
"grad_norm": 0.006696054711937904,
"learning_rate": 1.4494382022471912e-06,
"loss": 0.0002,
"step": 2560
},
{
"epoch": 8.074523396880416,
"grad_norm": 0.00565418740734458,
"learning_rate": 1.43338683788122e-06,
"loss": 0.0002,
"step": 2570
},
{
"epoch": 8.077411900635472,
"grad_norm": 0.011022423394024372,
"learning_rate": 1.4173354735152489e-06,
"loss": 0.0002,
"step": 2580
},
{
"epoch": 8.080300404390526,
"grad_norm": 0.006783703807741404,
"learning_rate": 1.401284109149278e-06,
"loss": 0.0002,
"step": 2590
},
{
"epoch": 8.08318890814558,
"grad_norm": 0.03814297169446945,
"learning_rate": 1.3852327447833067e-06,
"loss": 0.0003,
"step": 2600
},
{
"epoch": 8.083477758521086,
"eval_accuracy": 0.8139059304703476,
"eval_loss": 1.1929619312286377,
"eval_runtime": 66.8221,
"eval_samples_per_second": 7.318,
"eval_steps_per_second": 0.928,
"step": 2601
},
{
"epoch": 9.00259965337955,
"grad_norm": 0.004687231034040451,
"learning_rate": 1.3691813804173356e-06,
"loss": 0.0002,
"step": 2610
},
{
"epoch": 9.005488157134604,
"grad_norm": 0.01752406731247902,
"learning_rate": 1.3531300160513643e-06,
"loss": 0.0002,
"step": 2620
},
{
"epoch": 9.00837666088966,
"grad_norm": 0.005811620503664017,
"learning_rate": 1.3370786516853935e-06,
"loss": 0.0002,
"step": 2630
},
{
"epoch": 9.011265164644714,
"grad_norm": 0.0183991938829422,
"learning_rate": 1.3210272873194222e-06,
"loss": 0.0002,
"step": 2640
},
{
"epoch": 9.01415366839977,
"grad_norm": 0.00602333527058363,
"learning_rate": 1.304975922953451e-06,
"loss": 0.0002,
"step": 2650
},
{
"epoch": 9.017042172154824,
"grad_norm": 0.06442410498857498,
"learning_rate": 1.2889245585874802e-06,
"loss": 0.0002,
"step": 2660
},
{
"epoch": 9.019930675909878,
"grad_norm": 0.006975897122174501,
"learning_rate": 1.272873194221509e-06,
"loss": 0.0002,
"step": 2670
},
{
"epoch": 9.022819179664934,
"grad_norm": 0.006208993028849363,
"learning_rate": 1.2568218298555379e-06,
"loss": 0.0002,
"step": 2680
},
{
"epoch": 9.025707683419988,
"grad_norm": 0.006232697516679764,
"learning_rate": 1.2407704654895668e-06,
"loss": 0.0004,
"step": 2690
},
{
"epoch": 9.028596187175044,
"grad_norm": 0.008961373008787632,
"learning_rate": 1.2247191011235957e-06,
"loss": 0.0002,
"step": 2700
},
{
"epoch": 9.031484690930098,
"grad_norm": 0.017206748947501183,
"learning_rate": 1.2086677367576246e-06,
"loss": 0.0002,
"step": 2710
},
{
"epoch": 9.034373194685154,
"grad_norm": 0.024422556161880493,
"learning_rate": 1.1926163723916533e-06,
"loss": 0.0002,
"step": 2720
},
{
"epoch": 9.037261698440208,
"grad_norm": 0.0045550218783319,
"learning_rate": 1.1765650080256823e-06,
"loss": 0.0002,
"step": 2730
},
{
"epoch": 9.040150202195262,
"grad_norm": 0.030436644330620766,
"learning_rate": 1.1605136436597112e-06,
"loss": 0.0002,
"step": 2740
},
{
"epoch": 9.043038705950318,
"grad_norm": 0.018208695575594902,
"learning_rate": 1.14446227929374e-06,
"loss": 0.0003,
"step": 2750
},
{
"epoch": 9.045927209705372,
"grad_norm": 0.005186667200177908,
"learning_rate": 1.1284109149277688e-06,
"loss": 0.0002,
"step": 2760
},
{
"epoch": 9.048815713460428,
"grad_norm": 0.021143196150660515,
"learning_rate": 1.112359550561798e-06,
"loss": 0.0002,
"step": 2770
},
{
"epoch": 9.051704217215482,
"grad_norm": 0.312775194644928,
"learning_rate": 1.0963081861958269e-06,
"loss": 0.0002,
"step": 2780
},
{
"epoch": 9.054592720970538,
"grad_norm": 0.009823737666010857,
"learning_rate": 1.0802568218298556e-06,
"loss": 0.0002,
"step": 2790
},
{
"epoch": 9.057481224725592,
"grad_norm": 0.01661614701151848,
"learning_rate": 1.0642054574638845e-06,
"loss": 0.0002,
"step": 2800
},
{
"epoch": 9.060369728480646,
"grad_norm": 0.010930007323622704,
"learning_rate": 1.0481540930979134e-06,
"loss": 0.0001,
"step": 2810
},
{
"epoch": 9.063258232235702,
"grad_norm": 0.005455450154840946,
"learning_rate": 1.0321027287319423e-06,
"loss": 0.0002,
"step": 2820
},
{
"epoch": 9.066146735990756,
"grad_norm": 0.010016432963311672,
"learning_rate": 1.016051364365971e-06,
"loss": 0.0002,
"step": 2830
},
{
"epoch": 9.069035239745812,
"grad_norm": 0.0051452540792524815,
"learning_rate": 1.0000000000000002e-06,
"loss": 0.0002,
"step": 2840
},
{
"epoch": 9.071923743500866,
"grad_norm": 0.008383591659367085,
"learning_rate": 9.83948635634029e-07,
"loss": 0.0002,
"step": 2850
},
{
"epoch": 9.074812247255922,
"grad_norm": 0.005880800541490316,
"learning_rate": 9.678972712680578e-07,
"loss": 0.0003,
"step": 2860
},
{
"epoch": 9.077700751010976,
"grad_norm": 0.004441461060196161,
"learning_rate": 9.518459069020867e-07,
"loss": 0.0002,
"step": 2870
},
{
"epoch": 9.080589254766032,
"grad_norm": 0.018957389518618584,
"learning_rate": 9.357945425361156e-07,
"loss": 0.0002,
"step": 2880
},
{
"epoch": 9.083477758521086,
"grad_norm": 0.0041592782363295555,
"learning_rate": 9.197431781701445e-07,
"loss": 0.0002,
"step": 2890
},
{
"epoch": 9.083477758521086,
"eval_accuracy": 0.8077709611451943,
"eval_loss": 1.2366055250167847,
"eval_runtime": 69.8127,
"eval_samples_per_second": 7.004,
"eval_steps_per_second": 0.888,
"step": 2890
},
{
"epoch": 10.002888503755054,
"grad_norm": 0.0036611612886190414,
"learning_rate": 9.036918138041734e-07,
"loss": 0.0002,
"step": 2900
},
{
"epoch": 10.00577700751011,
"grad_norm": 0.010820758529007435,
"learning_rate": 8.876404494382024e-07,
"loss": 0.0002,
"step": 2910
},
{
"epoch": 10.008665511265164,
"grad_norm": 0.18064051866531372,
"learning_rate": 8.715890850722312e-07,
"loss": 0.0002,
"step": 2920
},
{
"epoch": 10.01155401502022,
"grad_norm": 0.011522581800818443,
"learning_rate": 8.555377207062601e-07,
"loss": 0.0002,
"step": 2930
},
{
"epoch": 10.014442518775274,
"grad_norm": 0.02415822632610798,
"learning_rate": 8.394863563402889e-07,
"loss": 0.0002,
"step": 2940
},
{
"epoch": 10.01733102253033,
"grad_norm": 0.005263539496809244,
"learning_rate": 8.234349919743179e-07,
"loss": 0.0002,
"step": 2950
},
{
"epoch": 10.020219526285384,
"grad_norm": 0.008322079665958881,
"learning_rate": 8.073836276083468e-07,
"loss": 0.0002,
"step": 2960
},
{
"epoch": 10.023108030040438,
"grad_norm": 0.0035860182251781225,
"learning_rate": 7.913322632423756e-07,
"loss": 0.0002,
"step": 2970
},
{
"epoch": 10.025996533795494,
"grad_norm": 0.026575535535812378,
"learning_rate": 7.752808988764046e-07,
"loss": 0.0002,
"step": 2980
},
{
"epoch": 10.028885037550548,
"grad_norm": 0.006455577909946442,
"learning_rate": 7.592295345104334e-07,
"loss": 0.0002,
"step": 2990
},
{
"epoch": 10.031773541305604,
"grad_norm": 0.023962225764989853,
"learning_rate": 7.431781701444624e-07,
"loss": 0.0002,
"step": 3000
},
{
"epoch": 10.034662045060658,
"grad_norm": 0.004179599694907665,
"learning_rate": 7.271268057784913e-07,
"loss": 0.0001,
"step": 3010
},
{
"epoch": 10.037550548815714,
"grad_norm": 0.0037284549325704575,
"learning_rate": 7.110754414125201e-07,
"loss": 0.0002,
"step": 3020
},
{
"epoch": 10.040439052570768,
"grad_norm": 0.006995455361902714,
"learning_rate": 6.95024077046549e-07,
"loss": 0.0002,
"step": 3030
},
{
"epoch": 10.043327556325822,
"grad_norm": 0.005864481907337904,
"learning_rate": 6.789727126805778e-07,
"loss": 0.0002,
"step": 3040
},
{
"epoch": 10.046216060080878,
"grad_norm": 0.007472215220332146,
"learning_rate": 6.629213483146069e-07,
"loss": 0.0002,
"step": 3050
},
{
"epoch": 10.049104563835932,
"grad_norm": 0.00438402034342289,
"learning_rate": 6.468699839486358e-07,
"loss": 0.0002,
"step": 3060
},
{
"epoch": 10.051993067590988,
"grad_norm": 0.029750341549515724,
"learning_rate": 6.308186195826646e-07,
"loss": 0.0001,
"step": 3070
},
{
"epoch": 10.054881571346042,
"grad_norm": 0.010507237166166306,
"learning_rate": 6.147672552166935e-07,
"loss": 0.0002,
"step": 3080
},
{
"epoch": 10.057770075101098,
"grad_norm": 0.006501095835119486,
"learning_rate": 5.987158908507223e-07,
"loss": 0.0001,
"step": 3090
},
{
"epoch": 10.060658578856152,
"grad_norm": 0.013263998553156853,
"learning_rate": 5.826645264847512e-07,
"loss": 0.0001,
"step": 3100
},
{
"epoch": 10.063547082611207,
"grad_norm": 0.06579581648111343,
"learning_rate": 5.666131621187802e-07,
"loss": 0.0002,
"step": 3110
},
{
"epoch": 10.066435586366262,
"grad_norm": 0.024195345118641853,
"learning_rate": 5.50561797752809e-07,
"loss": 0.0002,
"step": 3120
},
{
"epoch": 10.069324090121317,
"grad_norm": 0.00511367479339242,
"learning_rate": 5.345104333868379e-07,
"loss": 0.0003,
"step": 3130
},
{
"epoch": 10.072212593876372,
"grad_norm": 0.008129936642944813,
"learning_rate": 5.184590690208668e-07,
"loss": 0.0002,
"step": 3140
},
{
"epoch": 10.075101097631427,
"grad_norm": 0.0060712313279509544,
"learning_rate": 5.024077046548957e-07,
"loss": 0.0002,
"step": 3150
},
{
"epoch": 10.077989601386482,
"grad_norm": 0.016326796263456345,
"learning_rate": 4.863563402889246e-07,
"loss": 0.0001,
"step": 3160
},
{
"epoch": 10.080878105141537,
"grad_norm": 0.0037833349779248238,
"learning_rate": 4.703049759229535e-07,
"loss": 0.0002,
"step": 3170
},
{
"epoch": 10.083477758521086,
"eval_accuracy": 0.8098159509202454,
"eval_loss": 1.2292110919952393,
"eval_runtime": 69.0823,
"eval_samples_per_second": 7.079,
"eval_steps_per_second": 0.897,
"step": 3179
},
{
"epoch": 11.000288850375506,
"grad_norm": 0.007949397899210453,
"learning_rate": 4.542536115569824e-07,
"loss": 0.0001,
"step": 3180
},
{
"epoch": 11.00317735413056,
"grad_norm": 0.005895549897104502,
"learning_rate": 4.3820224719101127e-07,
"loss": 0.0004,
"step": 3190
},
{
"epoch": 11.006065857885615,
"grad_norm": 0.004202838055789471,
"learning_rate": 4.2215088282504014e-07,
"loss": 0.0001,
"step": 3200
},
{
"epoch": 11.00895436164067,
"grad_norm": 0.005272583104670048,
"learning_rate": 4.060995184590691e-07,
"loss": 0.0002,
"step": 3210
},
{
"epoch": 11.011842865395725,
"grad_norm": 0.009343048557639122,
"learning_rate": 3.90048154093098e-07,
"loss": 0.0002,
"step": 3220
},
{
"epoch": 11.01473136915078,
"grad_norm": 0.004706119187176228,
"learning_rate": 3.7399678972712684e-07,
"loss": 0.0002,
"step": 3230
},
{
"epoch": 11.017619872905835,
"grad_norm": 0.03129461407661438,
"learning_rate": 3.579454253611557e-07,
"loss": 0.0002,
"step": 3240
},
{
"epoch": 11.02050837666089,
"grad_norm": 0.004606070462614298,
"learning_rate": 3.4189406099518463e-07,
"loss": 0.0002,
"step": 3250
},
{
"epoch": 11.023396880415945,
"grad_norm": 0.0045063793659210205,
"learning_rate": 3.258426966292135e-07,
"loss": 0.0002,
"step": 3260
},
{
"epoch": 11.026285384170999,
"grad_norm": 0.0062253838405013084,
"learning_rate": 3.0979133226324237e-07,
"loss": 0.0003,
"step": 3270
},
{
"epoch": 11.029173887926055,
"grad_norm": 0.0034174402244389057,
"learning_rate": 2.937399678972713e-07,
"loss": 0.0002,
"step": 3280
},
{
"epoch": 11.032062391681109,
"grad_norm": 0.004293691832572222,
"learning_rate": 2.776886035313002e-07,
"loss": 0.0001,
"step": 3290
},
{
"epoch": 11.034950895436165,
"grad_norm": 0.0049605295062065125,
"learning_rate": 2.616372391653291e-07,
"loss": 0.0002,
"step": 3300
},
{
"epoch": 11.037839399191219,
"grad_norm": 0.005613707937300205,
"learning_rate": 2.45585874799358e-07,
"loss": 0.0002,
"step": 3310
},
{
"epoch": 11.040727902946275,
"grad_norm": 0.18399178981781006,
"learning_rate": 2.2953451043338686e-07,
"loss": 0.0002,
"step": 3320
},
{
"epoch": 11.043616406701329,
"grad_norm": 0.004424693062901497,
"learning_rate": 2.1348314606741576e-07,
"loss": 0.0001,
"step": 3330
},
{
"epoch": 11.046504910456383,
"grad_norm": 0.020336680114269257,
"learning_rate": 1.9743178170144462e-07,
"loss": 0.0001,
"step": 3340
},
{
"epoch": 11.049393414211439,
"grad_norm": 0.015212628059089184,
"learning_rate": 1.8138041733547354e-07,
"loss": 0.0001,
"step": 3350
},
{
"epoch": 11.052281917966493,
"grad_norm": 0.008279630914330482,
"learning_rate": 1.653290529695024e-07,
"loss": 0.0001,
"step": 3360
},
{
"epoch": 11.055170421721549,
"grad_norm": 0.0034426343627274036,
"learning_rate": 1.492776886035313e-07,
"loss": 0.0002,
"step": 3370
},
{
"epoch": 11.058058925476603,
"grad_norm": 0.010223750956356525,
"learning_rate": 1.332263242375602e-07,
"loss": 0.0002,
"step": 3380
},
{
"epoch": 11.060947429231659,
"grad_norm": 0.004876590799540281,
"learning_rate": 1.1717495987158909e-07,
"loss": 0.0001,
"step": 3390
},
{
"epoch": 11.063835932986713,
"grad_norm": 0.006909272633492947,
"learning_rate": 1.0112359550561797e-07,
"loss": 0.0002,
"step": 3400
},
{
"epoch": 11.066724436741767,
"grad_norm": 0.017502378672361374,
"learning_rate": 8.507223113964687e-08,
"loss": 0.0002,
"step": 3410
},
{
"epoch": 11.069612940496823,
"grad_norm": 0.020532017573714256,
"learning_rate": 6.902086677367576e-08,
"loss": 0.0001,
"step": 3420
},
{
"epoch": 11.072501444251877,
"grad_norm": 0.01509220339357853,
"learning_rate": 5.2969502407704655e-08,
"loss": 0.0002,
"step": 3430
},
{
"epoch": 11.075389948006933,
"grad_norm": 0.0059640249237418175,
"learning_rate": 3.691813804173355e-08,
"loss": 0.0002,
"step": 3440
},
{
"epoch": 11.078278451761987,
"grad_norm": 0.0043342639692127705,
"learning_rate": 2.0866773675762443e-08,
"loss": 0.0001,
"step": 3450
},
{
"epoch": 11.081166955517043,
"grad_norm": 0.010837825946509838,
"learning_rate": 4.815409309791333e-09,
"loss": 0.0002,
"step": 3460
},
{
"epoch": 11.081744656268054,
"eval_accuracy": 0.8098159509202454,
"eval_loss": 1.2295476198196411,
"eval_runtime": 73.4564,
"eval_samples_per_second": 6.657,
"eval_steps_per_second": 0.844,
"step": 3462
},
{
"epoch": 11.081744656268054,
"step": 3462,
"total_flos": 3.4483591449041043e+19,
"train_loss": 0.08882198683258911,
"train_runtime": 9253.1851,
"train_samples_per_second": 2.993,
"train_steps_per_second": 0.374
},
{
"epoch": 11.081744656268054,
"eval_accuracy": 0.8413654618473896,
"eval_loss": 0.7867043018341064,
"eval_runtime": 76.5589,
"eval_samples_per_second": 6.505,
"eval_steps_per_second": 0.823,
"step": 3462
},
{
"epoch": 11.081744656268054,
"eval_accuracy": 0.8413654618473896,
"eval_loss": 0.7867043018341064,
"eval_runtime": 74.8858,
"eval_samples_per_second": 6.65,
"eval_steps_per_second": 0.841,
"step": 3462
}
],
"logging_steps": 10,
"max_steps": 3462,
"num_input_tokens_seen": 0,
"num_train_epochs": 9223372036854775807,
"save_steps": 500,
"stateful_callbacks": {
"EarlyStoppingCallback": {
"args": {
"early_stopping_patience": 4,
"early_stopping_threshold": 0.0
},
"attributes": {
"early_stopping_patience_counter": 3
}
},
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 3.4483591449041043e+19,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}