atsuki-yamaguchi's picture
Upload folder using huggingface_hub
0f69a64 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.1882686776623717,
"eval_steps": 500,
"global_step": 12208,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0004780741323340042,
"grad_norm": 3.1866345405578613,
"learning_rate": 1.0157273918741808e-06,
"loss": 1.194,
"step": 31
},
{
"epoch": 0.0009561482646680084,
"grad_norm": 1.7708723545074463,
"learning_rate": 2.0314547837483616e-06,
"loss": 1.1124,
"step": 62
},
{
"epoch": 0.0014342223970020126,
"grad_norm": 1.6835601329803467,
"learning_rate": 3.0471821756225426e-06,
"loss": 1.0585,
"step": 93
},
{
"epoch": 0.0019122965293360168,
"grad_norm": 1.642537236213684,
"learning_rate": 4.062909567496723e-06,
"loss": 1.0213,
"step": 124
},
{
"epoch": 0.002390370661670021,
"grad_norm": 1.8476150035858154,
"learning_rate": 5.078636959370905e-06,
"loss": 1.0024,
"step": 155
},
{
"epoch": 0.0028684447940040252,
"grad_norm": 1.96336829662323,
"learning_rate": 6.094364351245085e-06,
"loss": 0.9938,
"step": 186
},
{
"epoch": 0.0033465189263380294,
"grad_norm": 1.8158528804779053,
"learning_rate": 7.110091743119267e-06,
"loss": 0.969,
"step": 217
},
{
"epoch": 0.0038245930586720336,
"grad_norm": 1.9796783924102783,
"learning_rate": 8.125819134993446e-06,
"loss": 0.9573,
"step": 248
},
{
"epoch": 0.004302667191006038,
"grad_norm": 1.8192514181137085,
"learning_rate": 9.141546526867629e-06,
"loss": 0.9498,
"step": 279
},
{
"epoch": 0.004780741323340042,
"grad_norm": 2.1396358013153076,
"learning_rate": 1.015727391874181e-05,
"loss": 0.9294,
"step": 310
},
{
"epoch": 0.005258815455674046,
"grad_norm": 2.0807156562805176,
"learning_rate": 1.117300131061599e-05,
"loss": 0.9186,
"step": 341
},
{
"epoch": 0.0057368895880080505,
"grad_norm": 2.002887010574341,
"learning_rate": 1.218872870249017e-05,
"loss": 0.9149,
"step": 372
},
{
"epoch": 0.006214963720342055,
"grad_norm": 2.240675687789917,
"learning_rate": 1.3204456094364351e-05,
"loss": 0.9139,
"step": 403
},
{
"epoch": 0.006693037852676059,
"grad_norm": 1.9016591310501099,
"learning_rate": 1.4220183486238533e-05,
"loss": 0.8946,
"step": 434
},
{
"epoch": 0.007171111985010063,
"grad_norm": 1.6729310750961304,
"learning_rate": 1.5235910878112714e-05,
"loss": 0.8959,
"step": 465
},
{
"epoch": 0.007649186117344067,
"grad_norm": 2.2607779502868652,
"learning_rate": 1.6251638269986893e-05,
"loss": 0.8847,
"step": 496
},
{
"epoch": 0.00812726024967807,
"grad_norm": 1.699507236480713,
"learning_rate": 1.7267365661861077e-05,
"loss": 0.8773,
"step": 527
},
{
"epoch": 0.008605334382012076,
"grad_norm": 1.93502938747406,
"learning_rate": 1.8283093053735257e-05,
"loss": 0.8773,
"step": 558
},
{
"epoch": 0.009083408514346079,
"grad_norm": 2.1234281063079834,
"learning_rate": 1.9298820445609438e-05,
"loss": 0.8669,
"step": 589
},
{
"epoch": 0.009561482646680084,
"grad_norm": 1.7465944290161133,
"learning_rate": 2.031454783748362e-05,
"loss": 0.8536,
"step": 620
},
{
"epoch": 0.010039556779014087,
"grad_norm": 1.7039932012557983,
"learning_rate": 2.13302752293578e-05,
"loss": 0.8624,
"step": 651
},
{
"epoch": 0.010517630911348093,
"grad_norm": 1.551837682723999,
"learning_rate": 2.234600262123198e-05,
"loss": 0.8555,
"step": 682
},
{
"epoch": 0.010995705043682096,
"grad_norm": 1.7622497081756592,
"learning_rate": 2.336173001310616e-05,
"loss": 0.8475,
"step": 713
},
{
"epoch": 0.011473779176016101,
"grad_norm": 2.000936269760132,
"learning_rate": 2.437745740498034e-05,
"loss": 0.852,
"step": 744
},
{
"epoch": 0.011951853308350104,
"grad_norm": 1.6770379543304443,
"learning_rate": 2.5393184796854525e-05,
"loss": 0.8455,
"step": 775
},
{
"epoch": 0.01242992744068411,
"grad_norm": 1.746506929397583,
"learning_rate": 2.6408912188728702e-05,
"loss": 0.8406,
"step": 806
},
{
"epoch": 0.012908001573018113,
"grad_norm": 1.507759690284729,
"learning_rate": 2.7424639580602886e-05,
"loss": 0.8556,
"step": 837
},
{
"epoch": 0.013386075705352118,
"grad_norm": 1.6984492540359497,
"learning_rate": 2.8440366972477066e-05,
"loss": 0.8281,
"step": 868
},
{
"epoch": 0.013864149837686121,
"grad_norm": 1.5145736932754517,
"learning_rate": 2.9456094364351244e-05,
"loss": 0.8359,
"step": 899
},
{
"epoch": 0.014342223970020126,
"grad_norm": 1.5700000524520874,
"learning_rate": 3.0471821756225428e-05,
"loss": 0.8305,
"step": 930
},
{
"epoch": 0.01482029810235413,
"grad_norm": 1.6657596826553345,
"learning_rate": 3.148754914809961e-05,
"loss": 0.8263,
"step": 961
},
{
"epoch": 0.015298372234688135,
"grad_norm": 1.527628779411316,
"learning_rate": 3.2503276539973785e-05,
"loss": 0.8318,
"step": 992
},
{
"epoch": 0.015776446367022138,
"grad_norm": 1.8353967666625977,
"learning_rate": 3.351900393184797e-05,
"loss": 0.8192,
"step": 1023
},
{
"epoch": 0.01625452049935614,
"grad_norm": 1.2118042707443237,
"learning_rate": 3.453473132372215e-05,
"loss": 0.819,
"step": 1054
},
{
"epoch": 0.016732594631690145,
"grad_norm": 1.3235970735549927,
"learning_rate": 3.555045871559633e-05,
"loss": 0.8176,
"step": 1085
},
{
"epoch": 0.01721066876402415,
"grad_norm": 1.6729888916015625,
"learning_rate": 3.6566186107470514e-05,
"loss": 0.8202,
"step": 1116
},
{
"epoch": 0.017688742896358155,
"grad_norm": 1.3251298666000366,
"learning_rate": 3.7581913499344695e-05,
"loss": 0.8104,
"step": 1147
},
{
"epoch": 0.018166817028692158,
"grad_norm": 1.5231342315673828,
"learning_rate": 3.8597640891218876e-05,
"loss": 0.8113,
"step": 1178
},
{
"epoch": 0.01864489116102616,
"grad_norm": 1.3263883590698242,
"learning_rate": 3.9613368283093056e-05,
"loss": 0.7981,
"step": 1209
},
{
"epoch": 0.019122965293360168,
"grad_norm": 1.317791223526001,
"learning_rate": 4.062909567496724e-05,
"loss": 0.7962,
"step": 1240
},
{
"epoch": 0.01960103942569417,
"grad_norm": 1.566698670387268,
"learning_rate": 4.164482306684142e-05,
"loss": 0.8072,
"step": 1271
},
{
"epoch": 0.020079113558028175,
"grad_norm": 1.2935110330581665,
"learning_rate": 4.26605504587156e-05,
"loss": 0.7818,
"step": 1302
},
{
"epoch": 0.020557187690362178,
"grad_norm": 1.336227536201477,
"learning_rate": 4.367627785058978e-05,
"loss": 0.7917,
"step": 1333
},
{
"epoch": 0.021035261822696185,
"grad_norm": 1.5656548738479614,
"learning_rate": 4.469200524246396e-05,
"loss": 0.7785,
"step": 1364
},
{
"epoch": 0.02151333595503019,
"grad_norm": 1.5208159685134888,
"learning_rate": 4.570773263433814e-05,
"loss": 0.7839,
"step": 1395
},
{
"epoch": 0.02199141008736419,
"grad_norm": 1.4816375970840454,
"learning_rate": 4.672346002621232e-05,
"loss": 0.7789,
"step": 1426
},
{
"epoch": 0.022469484219698195,
"grad_norm": 1.402166485786438,
"learning_rate": 4.77391874180865e-05,
"loss": 0.7803,
"step": 1457
},
{
"epoch": 0.022947558352032202,
"grad_norm": 1.2986265420913696,
"learning_rate": 4.875491480996068e-05,
"loss": 0.7739,
"step": 1488
},
{
"epoch": 0.023425632484366205,
"grad_norm": 1.3171271085739136,
"learning_rate": 4.977064220183487e-05,
"loss": 0.7694,
"step": 1519
},
{
"epoch": 0.02390370661670021,
"grad_norm": 1.4870030879974365,
"learning_rate": 4.9999915451558777e-05,
"loss": 0.7663,
"step": 1550
},
{
"epoch": 0.024381780749034212,
"grad_norm": 1.7419476509094238,
"learning_rate": 4.999955597496219e-05,
"loss": 0.7731,
"step": 1581
},
{
"epoch": 0.02485985488136822,
"grad_norm": 1.109238624572754,
"learning_rate": 4.9998914381774255e-05,
"loss": 0.7704,
"step": 1612
},
{
"epoch": 0.025337929013702222,
"grad_norm": 1.3478920459747314,
"learning_rate": 4.999799067923527e-05,
"loss": 0.7723,
"step": 1643
},
{
"epoch": 0.025816003146036225,
"grad_norm": 1.2739650011062622,
"learning_rate": 4.999678487776908e-05,
"loss": 0.7699,
"step": 1674
},
{
"epoch": 0.02629407727837023,
"grad_norm": 1.3892935514450073,
"learning_rate": 4.9995296990983006e-05,
"loss": 0.7709,
"step": 1705
},
{
"epoch": 0.026772151410704235,
"grad_norm": 1.2031512260437012,
"learning_rate": 4.999352703566763e-05,
"loss": 0.7557,
"step": 1736
},
{
"epoch": 0.02725022554303824,
"grad_norm": 1.185471773147583,
"learning_rate": 4.999147503179668e-05,
"loss": 0.7645,
"step": 1767
},
{
"epoch": 0.027728299675372242,
"grad_norm": 1.2331740856170654,
"learning_rate": 4.998914100252672e-05,
"loss": 0.7655,
"step": 1798
},
{
"epoch": 0.028206373807706245,
"grad_norm": 1.1817049980163574,
"learning_rate": 4.998652497419696e-05,
"loss": 0.7617,
"step": 1829
},
{
"epoch": 0.028684447940040252,
"grad_norm": 1.2225860357284546,
"learning_rate": 4.9983626976328927e-05,
"loss": 0.7558,
"step": 1860
},
{
"epoch": 0.029162522072374256,
"grad_norm": 1.6452528238296509,
"learning_rate": 4.998044704162613e-05,
"loss": 0.7566,
"step": 1891
},
{
"epoch": 0.02964059620470826,
"grad_norm": 1.3764828443527222,
"learning_rate": 4.9976985205973705e-05,
"loss": 0.7545,
"step": 1922
},
{
"epoch": 0.030118670337042262,
"grad_norm": 1.6581465005874634,
"learning_rate": 4.997324150843799e-05,
"loss": 0.7519,
"step": 1953
},
{
"epoch": 0.03059674446937627,
"grad_norm": 1.1065036058425903,
"learning_rate": 4.99692159912661e-05,
"loss": 0.7579,
"step": 1984
},
{
"epoch": 0.031074818601710272,
"grad_norm": 2.6465444564819336,
"learning_rate": 4.996490869988546e-05,
"loss": 0.7538,
"step": 2015
},
{
"epoch": 0.031552892734044276,
"grad_norm": 1.3028968572616577,
"learning_rate": 4.996031968290326e-05,
"loss": 0.7522,
"step": 2046
},
{
"epoch": 0.03203096686637828,
"grad_norm": 1.0450382232666016,
"learning_rate": 4.995544899210594e-05,
"loss": 0.742,
"step": 2077
},
{
"epoch": 0.03250904099871228,
"grad_norm": 0.9411994814872742,
"learning_rate": 4.9950296682458583e-05,
"loss": 0.7401,
"step": 2108
},
{
"epoch": 0.03298711513104629,
"grad_norm": 1.0501608848571777,
"learning_rate": 4.994486281210429e-05,
"loss": 0.7455,
"step": 2139
},
{
"epoch": 0.03346518926338029,
"grad_norm": 1.266400694847107,
"learning_rate": 4.9939147442363566e-05,
"loss": 0.7391,
"step": 2170
},
{
"epoch": 0.033943263395714296,
"grad_norm": 1.2862213850021362,
"learning_rate": 4.9933150637733574e-05,
"loss": 0.7397,
"step": 2201
},
{
"epoch": 0.0344213375280483,
"grad_norm": 1.2715409994125366,
"learning_rate": 4.992687246588743e-05,
"loss": 0.7467,
"step": 2232
},
{
"epoch": 0.0348994116603823,
"grad_norm": 1.3290003538131714,
"learning_rate": 4.992031299767347e-05,
"loss": 0.7432,
"step": 2263
},
{
"epoch": 0.03537748579271631,
"grad_norm": 1.0561761856079102,
"learning_rate": 4.9913472307114386e-05,
"loss": 0.7336,
"step": 2294
},
{
"epoch": 0.035855559925050316,
"grad_norm": 1.3272422552108765,
"learning_rate": 4.9906350471406446e-05,
"loss": 0.7251,
"step": 2325
},
{
"epoch": 0.036333634057384316,
"grad_norm": 1.0749491453170776,
"learning_rate": 4.989894757091861e-05,
"loss": 0.7205,
"step": 2356
},
{
"epoch": 0.03681170818971832,
"grad_norm": 1.1581461429595947,
"learning_rate": 4.989126368919158e-05,
"loss": 0.7311,
"step": 2387
},
{
"epoch": 0.03728978232205232,
"grad_norm": 1.0796961784362793,
"learning_rate": 4.988329891293693e-05,
"loss": 0.7259,
"step": 2418
},
{
"epoch": 0.03776785645438633,
"grad_norm": 1.1916818618774414,
"learning_rate": 4.987505333203608e-05,
"loss": 0.7258,
"step": 2449
},
{
"epoch": 0.038245930586720336,
"grad_norm": 0.9515872001647949,
"learning_rate": 4.9866527039539276e-05,
"loss": 0.7273,
"step": 2480
},
{
"epoch": 0.038724004719054336,
"grad_norm": 1.1217246055603027,
"learning_rate": 4.9857720131664594e-05,
"loss": 0.7269,
"step": 2511
},
{
"epoch": 0.03920207885138834,
"grad_norm": 0.9570199847221375,
"learning_rate": 4.9848632707796773e-05,
"loss": 0.7289,
"step": 2542
},
{
"epoch": 0.03968015298372235,
"grad_norm": 1.2980682849884033,
"learning_rate": 4.9839264870486155e-05,
"loss": 0.7382,
"step": 2573
},
{
"epoch": 0.04015822711605635,
"grad_norm": 2.1257143020629883,
"learning_rate": 4.9829616725447526e-05,
"loss": 0.8112,
"step": 2604
},
{
"epoch": 0.04063630124839036,
"grad_norm": 1.5091110467910767,
"learning_rate": 4.981968838155888e-05,
"loss": 0.7451,
"step": 2635
},
{
"epoch": 0.041114375380724356,
"grad_norm": 2.4548749923706055,
"learning_rate": 4.980947995086024e-05,
"loss": 0.7358,
"step": 2666
},
{
"epoch": 0.04159244951305836,
"grad_norm": 1.176115870475769,
"learning_rate": 4.979899154855234e-05,
"loss": 0.7368,
"step": 2697
},
{
"epoch": 0.04207052364539237,
"grad_norm": 1.063672661781311,
"learning_rate": 4.9788223292995386e-05,
"loss": 0.7465,
"step": 2728
},
{
"epoch": 0.04254859777772637,
"grad_norm": 1.3062779903411865,
"learning_rate": 4.977717530570768e-05,
"loss": 0.7413,
"step": 2759
},
{
"epoch": 0.04302667191006038,
"grad_norm": 1.5550280809402466,
"learning_rate": 4.976584771136425e-05,
"loss": 0.724,
"step": 2790
},
{
"epoch": 0.043504746042394384,
"grad_norm": 1.2525728940963745,
"learning_rate": 4.975424063779547e-05,
"loss": 0.7216,
"step": 2821
},
{
"epoch": 0.04398282017472838,
"grad_norm": 1.158134937286377,
"learning_rate": 4.974235421598557e-05,
"loss": 0.728,
"step": 2852
},
{
"epoch": 0.04446089430706239,
"grad_norm": 1.262291669845581,
"learning_rate": 4.973018858007122e-05,
"loss": 0.7191,
"step": 2883
},
{
"epoch": 0.04493896843939639,
"grad_norm": 2.826028347015381,
"learning_rate": 4.9717743867339963e-05,
"loss": 0.7211,
"step": 2914
},
{
"epoch": 0.0454170425717304,
"grad_norm": 1.0346958637237549,
"learning_rate": 4.9705020218228695e-05,
"loss": 0.7268,
"step": 2945
},
{
"epoch": 0.045895116704064404,
"grad_norm": 1.4338330030441284,
"learning_rate": 4.969201777632205e-05,
"loss": 0.7154,
"step": 2976
},
{
"epoch": 0.046373190836398404,
"grad_norm": 0.9223676919937134,
"learning_rate": 4.9678736688350846e-05,
"loss": 0.7195,
"step": 3007
},
{
"epoch": 0.04685126496873241,
"grad_norm": 0.9936623573303223,
"learning_rate": 4.966517710419033e-05,
"loss": 0.7194,
"step": 3038
},
{
"epoch": 0.04732933910106642,
"grad_norm": 1.017962098121643,
"learning_rate": 4.965133917685858e-05,
"loss": 0.713,
"step": 3069
},
{
"epoch": 0.04780741323340042,
"grad_norm": 0.9654473662376404,
"learning_rate": 4.9637223062514714e-05,
"loss": 0.7096,
"step": 3100
},
{
"epoch": 0.048285487365734424,
"grad_norm": 0.9372695684432983,
"learning_rate": 4.962282892045718e-05,
"loss": 0.7116,
"step": 3131
},
{
"epoch": 0.048763561498068424,
"grad_norm": 0.9450846910476685,
"learning_rate": 4.9608156913121904e-05,
"loss": 0.7129,
"step": 3162
},
{
"epoch": 0.04924163563040243,
"grad_norm": 1.0803054571151733,
"learning_rate": 4.959320720608049e-05,
"loss": 0.706,
"step": 3193
},
{
"epoch": 0.04971970976273644,
"grad_norm": 1.2218348979949951,
"learning_rate": 4.9577979968038354e-05,
"loss": 0.7076,
"step": 3224
},
{
"epoch": 0.05019778389507044,
"grad_norm": 1.0429767370224,
"learning_rate": 4.956247537083282e-05,
"loss": 0.7089,
"step": 3255
},
{
"epoch": 0.050675858027404444,
"grad_norm": 0.9912049770355225,
"learning_rate": 4.9546693589431145e-05,
"loss": 0.7016,
"step": 3286
},
{
"epoch": 0.051153932159738444,
"grad_norm": 0.9687103033065796,
"learning_rate": 4.9530634801928595e-05,
"loss": 0.7071,
"step": 3317
},
{
"epoch": 0.05163200629207245,
"grad_norm": 0.8178670406341553,
"learning_rate": 4.9514299189546395e-05,
"loss": 0.6991,
"step": 3348
},
{
"epoch": 0.05211008042440646,
"grad_norm": 0.8601382374763489,
"learning_rate": 4.949768693662973e-05,
"loss": 0.6978,
"step": 3379
},
{
"epoch": 0.05258815455674046,
"grad_norm": 1.076370120048523,
"learning_rate": 4.948079823064559e-05,
"loss": 0.7044,
"step": 3410
},
{
"epoch": 0.053066228689074464,
"grad_norm": 3.9457356929779053,
"learning_rate": 4.946363326218074e-05,
"loss": 0.6978,
"step": 3441
},
{
"epoch": 0.05354430282140847,
"grad_norm": 0.8186474442481995,
"learning_rate": 4.9446192224939525e-05,
"loss": 0.6974,
"step": 3472
},
{
"epoch": 0.05402237695374247,
"grad_norm": 0.9643816947937012,
"learning_rate": 4.942847531574167e-05,
"loss": 0.7025,
"step": 3503
},
{
"epoch": 0.05450045108607648,
"grad_norm": 1.024248480796814,
"learning_rate": 4.941048273452008e-05,
"loss": 0.7006,
"step": 3534
},
{
"epoch": 0.05497852521841048,
"grad_norm": 0.7718949317932129,
"learning_rate": 4.9392214684318605e-05,
"loss": 0.7024,
"step": 3565
},
{
"epoch": 0.055456599350744484,
"grad_norm": 1.1313899755477905,
"learning_rate": 4.93736713712897e-05,
"loss": 0.701,
"step": 3596
},
{
"epoch": 0.05593467348307849,
"grad_norm": 1.0118827819824219,
"learning_rate": 4.9354853004692124e-05,
"loss": 0.7036,
"step": 3627
},
{
"epoch": 0.05641274761541249,
"grad_norm": 0.9465724229812622,
"learning_rate": 4.93357597968886e-05,
"loss": 0.6869,
"step": 3658
},
{
"epoch": 0.0568908217477465,
"grad_norm": 1.0233882665634155,
"learning_rate": 4.931639196334338e-05,
"loss": 0.6944,
"step": 3689
},
{
"epoch": 0.057368895880080505,
"grad_norm": 1.008609652519226,
"learning_rate": 4.9296749722619826e-05,
"loss": 0.6916,
"step": 3720
},
{
"epoch": 0.057846970012414504,
"grad_norm": 1.0083181858062744,
"learning_rate": 4.9276833296377966e-05,
"loss": 0.6886,
"step": 3751
},
{
"epoch": 0.05832504414474851,
"grad_norm": 0.9374220967292786,
"learning_rate": 4.925664290937196e-05,
"loss": 0.6976,
"step": 3782
},
{
"epoch": 0.05880311827708251,
"grad_norm": 0.9066904783248901,
"learning_rate": 4.9236178789447576e-05,
"loss": 0.6911,
"step": 3813
},
{
"epoch": 0.05928119240941652,
"grad_norm": 0.9702699780464172,
"learning_rate": 4.921544116753962e-05,
"loss": 0.6959,
"step": 3844
},
{
"epoch": 0.059759266541750525,
"grad_norm": 0.8405037522315979,
"learning_rate": 4.919443027766935e-05,
"loss": 0.6896,
"step": 3875
},
{
"epoch": 0.060237340674084525,
"grad_norm": 0.9460383653640747,
"learning_rate": 4.91731463569418e-05,
"loss": 0.6871,
"step": 3906
},
{
"epoch": 0.06071541480641853,
"grad_norm": 0.9355078339576721,
"learning_rate": 4.915158964554312e-05,
"loss": 0.6843,
"step": 3937
},
{
"epoch": 0.06119348893875254,
"grad_norm": 0.7211058735847473,
"learning_rate": 4.912976038673786e-05,
"loss": 0.6861,
"step": 3968
},
{
"epoch": 0.06167156307108654,
"grad_norm": 0.8674766421318054,
"learning_rate": 4.9107658826866254e-05,
"loss": 0.6939,
"step": 3999
},
{
"epoch": 0.062149637203420545,
"grad_norm": 0.8166181445121765,
"learning_rate": 4.908528521534139e-05,
"loss": 0.692,
"step": 4030
},
{
"epoch": 0.06262771133575455,
"grad_norm": 0.841305136680603,
"learning_rate": 4.906263980464644e-05,
"loss": 0.6855,
"step": 4061
},
{
"epoch": 0.06310578546808855,
"grad_norm": 0.8942857384681702,
"learning_rate": 4.903972285033178e-05,
"loss": 0.6946,
"step": 4092
},
{
"epoch": 0.06358385960042255,
"grad_norm": 0.8595120310783386,
"learning_rate": 4.901653461101213e-05,
"loss": 0.6825,
"step": 4123
},
{
"epoch": 0.06406193373275657,
"grad_norm": 0.8155812621116638,
"learning_rate": 4.8993075348363626e-05,
"loss": 0.6821,
"step": 4154
},
{
"epoch": 0.06454000786509057,
"grad_norm": 0.8901113271713257,
"learning_rate": 4.896934532712084e-05,
"loss": 0.6898,
"step": 4185
},
{
"epoch": 0.06501808199742456,
"grad_norm": 0.9637976288795471,
"learning_rate": 4.8945344815073846e-05,
"loss": 0.6829,
"step": 4216
},
{
"epoch": 0.06549615612975858,
"grad_norm": 0.7048139572143555,
"learning_rate": 4.892107408306516e-05,
"loss": 0.6834,
"step": 4247
},
{
"epoch": 0.06597423026209258,
"grad_norm": 0.8655612468719482,
"learning_rate": 4.889653340498669e-05,
"loss": 0.6778,
"step": 4278
},
{
"epoch": 0.06645230439442658,
"grad_norm": 0.8204261064529419,
"learning_rate": 4.8871723057776664e-05,
"loss": 0.6672,
"step": 4309
},
{
"epoch": 0.06693037852676058,
"grad_norm": 0.9259466528892517,
"learning_rate": 4.8846643321416476e-05,
"loss": 0.6778,
"step": 4340
},
{
"epoch": 0.06740845265909459,
"grad_norm": 0.9826278686523438,
"learning_rate": 4.882129447892753e-05,
"loss": 0.6882,
"step": 4371
},
{
"epoch": 0.06788652679142859,
"grad_norm": 0.7376586198806763,
"learning_rate": 4.8795676816368076e-05,
"loss": 0.6802,
"step": 4402
},
{
"epoch": 0.06836460092376259,
"grad_norm": 0.6778755784034729,
"learning_rate": 4.876979062282995e-05,
"loss": 0.6814,
"step": 4433
},
{
"epoch": 0.0688426750560966,
"grad_norm": 0.8117042183876038,
"learning_rate": 4.8743636190435325e-05,
"loss": 0.6759,
"step": 4464
},
{
"epoch": 0.0693207491884306,
"grad_norm": 0.9493227005004883,
"learning_rate": 4.871721381433344e-05,
"loss": 0.6744,
"step": 4495
},
{
"epoch": 0.0697988233207646,
"grad_norm": 1.3356900215148926,
"learning_rate": 4.869052379269719e-05,
"loss": 0.6763,
"step": 4526
},
{
"epoch": 0.07027689745309862,
"grad_norm": 1.1542543172836304,
"learning_rate": 4.866356642671985e-05,
"loss": 0.6788,
"step": 4557
},
{
"epoch": 0.07075497158543262,
"grad_norm": 0.785176157951355,
"learning_rate": 4.8636342020611634e-05,
"loss": 0.6853,
"step": 4588
},
{
"epoch": 0.07123304571776662,
"grad_norm": 0.8102776408195496,
"learning_rate": 4.860885088159626e-05,
"loss": 0.6794,
"step": 4619
},
{
"epoch": 0.07171111985010063,
"grad_norm": 0.9685975313186646,
"learning_rate": 4.858109331990751e-05,
"loss": 0.6823,
"step": 4650
},
{
"epoch": 0.07218919398243463,
"grad_norm": 0.7929072976112366,
"learning_rate": 4.855306964878567e-05,
"loss": 0.6781,
"step": 4681
},
{
"epoch": 0.07266726811476863,
"grad_norm": 0.8171564340591431,
"learning_rate": 4.8524780184474084e-05,
"loss": 0.6685,
"step": 4712
},
{
"epoch": 0.07314534224710265,
"grad_norm": 0.9218119382858276,
"learning_rate": 4.8496225246215496e-05,
"loss": 0.6796,
"step": 4743
},
{
"epoch": 0.07362341637943665,
"grad_norm": 3.1681158542633057,
"learning_rate": 4.8467405156248505e-05,
"loss": 0.676,
"step": 4774
},
{
"epoch": 0.07410149051177065,
"grad_norm": 0.9336240887641907,
"learning_rate": 4.843832023980392e-05,
"loss": 0.6734,
"step": 4805
},
{
"epoch": 0.07457956464410465,
"grad_norm": 0.8264138102531433,
"learning_rate": 4.840897082510106e-05,
"loss": 0.6905,
"step": 4836
},
{
"epoch": 0.07505763877643866,
"grad_norm": 0.819663405418396,
"learning_rate": 4.8379357243344084e-05,
"loss": 0.6814,
"step": 4867
},
{
"epoch": 0.07553571290877266,
"grad_norm": 3.124502420425415,
"learning_rate": 4.8349479828718236e-05,
"loss": 0.6704,
"step": 4898
},
{
"epoch": 0.07601378704110666,
"grad_norm": 0.7769860625267029,
"learning_rate": 4.8319338918386075e-05,
"loss": 0.6728,
"step": 4929
},
{
"epoch": 0.07649186117344067,
"grad_norm": 0.8232171535491943,
"learning_rate": 4.828893485248369e-05,
"loss": 0.6798,
"step": 4960
},
{
"epoch": 0.07696993530577467,
"grad_norm": 0.8771420121192932,
"learning_rate": 4.825826797411682e-05,
"loss": 0.6722,
"step": 4991
},
{
"epoch": 0.07744800943810867,
"grad_norm": 0.8321033716201782,
"learning_rate": 4.822733862935702e-05,
"loss": 0.6724,
"step": 5022
},
{
"epoch": 0.07792608357044269,
"grad_norm": 0.9346029162406921,
"learning_rate": 4.819614716723775e-05,
"loss": 0.6764,
"step": 5053
},
{
"epoch": 0.07840415770277669,
"grad_norm": 0.9193580150604248,
"learning_rate": 4.8164693939750425e-05,
"loss": 0.6789,
"step": 5084
},
{
"epoch": 0.07888223183511069,
"grad_norm": 1.0850661993026733,
"learning_rate": 4.813297930184042e-05,
"loss": 0.6685,
"step": 5115
},
{
"epoch": 0.0793603059674447,
"grad_norm": 0.6851856708526611,
"learning_rate": 4.810100361140314e-05,
"loss": 0.6621,
"step": 5146
},
{
"epoch": 0.0798383800997787,
"grad_norm": 0.7950114607810974,
"learning_rate": 4.8068767229279885e-05,
"loss": 0.6642,
"step": 5177
},
{
"epoch": 0.0803164542321127,
"grad_norm": 0.9632556438446045,
"learning_rate": 4.8036270519253854e-05,
"loss": 0.6628,
"step": 5208
},
{
"epoch": 0.0807945283644467,
"grad_norm": 0.7066652178764343,
"learning_rate": 4.8003513848046e-05,
"loss": 0.6642,
"step": 5239
},
{
"epoch": 0.08127260249678071,
"grad_norm": 0.7749651670455933,
"learning_rate": 4.79704975853109e-05,
"loss": 0.6633,
"step": 5270
},
{
"epoch": 0.08175067662911471,
"grad_norm": 1.702022671699524,
"learning_rate": 4.793722210363262e-05,
"loss": 0.6745,
"step": 5301
},
{
"epoch": 0.08222875076144871,
"grad_norm": 0.8134759068489075,
"learning_rate": 4.7903687778520414e-05,
"loss": 0.659,
"step": 5332
},
{
"epoch": 0.08270682489378273,
"grad_norm": 0.8805097341537476,
"learning_rate": 4.7869894988404593e-05,
"loss": 0.6779,
"step": 5363
},
{
"epoch": 0.08318489902611673,
"grad_norm": 0.9312698841094971,
"learning_rate": 4.783584411463221e-05,
"loss": 0.6634,
"step": 5394
},
{
"epoch": 0.08366297315845073,
"grad_norm": 0.7013948559761047,
"learning_rate": 4.780153554146274e-05,
"loss": 0.6589,
"step": 5425
},
{
"epoch": 0.08414104729078474,
"grad_norm": 0.8027428388595581,
"learning_rate": 4.7766969656063766e-05,
"loss": 0.6603,
"step": 5456
},
{
"epoch": 0.08461912142311874,
"grad_norm": 1.0179306268692017,
"learning_rate": 4.773214684850662e-05,
"loss": 0.6595,
"step": 5487
},
{
"epoch": 0.08509719555545274,
"grad_norm": 0.8512592911720276,
"learning_rate": 4.769706751176193e-05,
"loss": 0.6659,
"step": 5518
},
{
"epoch": 0.08557526968778675,
"grad_norm": 0.8476304411888123,
"learning_rate": 4.7661732041695264e-05,
"loss": 0.6658,
"step": 5549
},
{
"epoch": 0.08605334382012075,
"grad_norm": 0.771584153175354,
"learning_rate": 4.762614083706258e-05,
"loss": 0.6622,
"step": 5580
},
{
"epoch": 0.08653141795245475,
"grad_norm": 0.8719794154167175,
"learning_rate": 4.759029429950581e-05,
"loss": 0.661,
"step": 5611
},
{
"epoch": 0.08700949208478877,
"grad_norm": 0.9247443675994873,
"learning_rate": 4.7554192833548235e-05,
"loss": 0.6475,
"step": 5642
},
{
"epoch": 0.08748756621712277,
"grad_norm": 0.9540857672691345,
"learning_rate": 4.751783684659e-05,
"loss": 0.662,
"step": 5673
},
{
"epoch": 0.08796564034945677,
"grad_norm": 0.7939008474349976,
"learning_rate": 4.748122674890348e-05,
"loss": 0.6514,
"step": 5704
},
{
"epoch": 0.08844371448179077,
"grad_norm": 0.613531768321991,
"learning_rate": 4.7444362953628654e-05,
"loss": 0.6513,
"step": 5735
},
{
"epoch": 0.08892178861412478,
"grad_norm": 0.8084924817085266,
"learning_rate": 4.7407245876768424e-05,
"loss": 0.6465,
"step": 5766
},
{
"epoch": 0.08939986274645878,
"grad_norm": 0.7709007263183594,
"learning_rate": 4.736987593718397e-05,
"loss": 0.6618,
"step": 5797
},
{
"epoch": 0.08987793687879278,
"grad_norm": 0.6461811065673828,
"learning_rate": 4.733225355658999e-05,
"loss": 0.6516,
"step": 5828
},
{
"epoch": 0.0903560110111268,
"grad_norm": 0.6879326105117798,
"learning_rate": 4.7294379159549926e-05,
"loss": 0.655,
"step": 5859
},
{
"epoch": 0.0908340851434608,
"grad_norm": 0.7594075798988342,
"learning_rate": 4.725625317347119e-05,
"loss": 0.655,
"step": 5890
},
{
"epoch": 0.0913121592757948,
"grad_norm": 0.774758517742157,
"learning_rate": 4.7217876028600374e-05,
"loss": 0.6697,
"step": 5921
},
{
"epoch": 0.09179023340812881,
"grad_norm": 0.9164844155311584,
"learning_rate": 4.717924815801832e-05,
"loss": 0.6483,
"step": 5952
},
{
"epoch": 0.09226830754046281,
"grad_norm": 0.792631208896637,
"learning_rate": 4.714036999763532e-05,
"loss": 0.6644,
"step": 5983
},
{
"epoch": 0.09274638167279681,
"grad_norm": 0.7735174298286438,
"learning_rate": 4.7101241986186116e-05,
"loss": 0.6545,
"step": 6014
},
{
"epoch": 0.09322445580513082,
"grad_norm": 0.725825309753418,
"learning_rate": 4.7061864565225e-05,
"loss": 0.6579,
"step": 6045
},
{
"epoch": 0.09370252993746482,
"grad_norm": 0.7191294431686401,
"learning_rate": 4.702223817912081e-05,
"loss": 0.6509,
"step": 6076
},
{
"epoch": 0.09418060406979882,
"grad_norm": 0.8470766544342041,
"learning_rate": 4.698236327505195e-05,
"loss": 0.6515,
"step": 6107
},
{
"epoch": 0.09465867820213283,
"grad_norm": 0.7250267863273621,
"learning_rate": 4.694224030300127e-05,
"loss": 0.6484,
"step": 6138
},
{
"epoch": 0.09513675233446683,
"grad_norm": 0.7370029091835022,
"learning_rate": 4.690186971575107e-05,
"loss": 0.6639,
"step": 6169
},
{
"epoch": 0.09561482646680083,
"grad_norm": 0.7768808007240295,
"learning_rate": 4.6861251968877916e-05,
"loss": 0.657,
"step": 6200
},
{
"epoch": 0.09609290059913483,
"grad_norm": 0.770039439201355,
"learning_rate": 4.68203875207476e-05,
"loss": 0.6588,
"step": 6231
},
{
"epoch": 0.09657097473146885,
"grad_norm": 0.637853741645813,
"learning_rate": 4.677927683250983e-05,
"loss": 0.6438,
"step": 6262
},
{
"epoch": 0.09704904886380285,
"grad_norm": 0.6896365284919739,
"learning_rate": 4.6737920368093156e-05,
"loss": 0.6493,
"step": 6293
},
{
"epoch": 0.09752712299613685,
"grad_norm": 0.721318781375885,
"learning_rate": 4.669631859419965e-05,
"loss": 0.6427,
"step": 6324
},
{
"epoch": 0.09800519712847086,
"grad_norm": 0.7492154836654663,
"learning_rate": 4.6654471980299676e-05,
"loss": 0.6474,
"step": 6355
},
{
"epoch": 0.09848327126080486,
"grad_norm": 0.8512872457504272,
"learning_rate": 4.661238099862658e-05,
"loss": 0.6423,
"step": 6386
},
{
"epoch": 0.09896134539313886,
"grad_norm": 0.6349690556526184,
"learning_rate": 4.657004612417138e-05,
"loss": 0.6499,
"step": 6417
},
{
"epoch": 0.09943941952547287,
"grad_norm": 0.8040255904197693,
"learning_rate": 4.6527467834677374e-05,
"loss": 0.6495,
"step": 6448
},
{
"epoch": 0.09991749365780687,
"grad_norm": 0.6428426504135132,
"learning_rate": 4.648464661063478e-05,
"loss": 0.6565,
"step": 6479
},
{
"epoch": 0.10039556779014087,
"grad_norm": 0.8141711950302124,
"learning_rate": 4.6441582935275264e-05,
"loss": 0.6592,
"step": 6510
},
{
"epoch": 0.10087364192247489,
"grad_norm": 0.8665099740028381,
"learning_rate": 4.6398277294566586e-05,
"loss": 0.6497,
"step": 6541
},
{
"epoch": 0.10135171605480889,
"grad_norm": 0.709786057472229,
"learning_rate": 4.6354730177207e-05,
"loss": 0.6546,
"step": 6572
},
{
"epoch": 0.10182979018714289,
"grad_norm": 0.8345597982406616,
"learning_rate": 4.6310942074619787e-05,
"loss": 0.6547,
"step": 6603
},
{
"epoch": 0.10230786431947689,
"grad_norm": 0.746769905090332,
"learning_rate": 4.626691348094777e-05,
"loss": 0.6526,
"step": 6634
},
{
"epoch": 0.1027859384518109,
"grad_norm": 0.7869583368301392,
"learning_rate": 4.622264489304762e-05,
"loss": 0.6458,
"step": 6665
},
{
"epoch": 0.1032640125841449,
"grad_norm": 0.5438331365585327,
"learning_rate": 4.617813681048434e-05,
"loss": 0.6468,
"step": 6696
},
{
"epoch": 0.1037420867164789,
"grad_norm": 0.7230411171913147,
"learning_rate": 4.61333897355256e-05,
"loss": 0.6435,
"step": 6727
},
{
"epoch": 0.10422016084881291,
"grad_norm": 0.9861361384391785,
"learning_rate": 4.608840417313604e-05,
"loss": 0.6488,
"step": 6758
},
{
"epoch": 0.10469823498114691,
"grad_norm": 0.7873183488845825,
"learning_rate": 4.6043180630971646e-05,
"loss": 0.6485,
"step": 6789
},
{
"epoch": 0.10517630911348091,
"grad_norm": 0.7746345400810242,
"learning_rate": 4.599771961937391e-05,
"loss": 0.6384,
"step": 6820
},
{
"epoch": 0.10565438324581493,
"grad_norm": 0.8218130469322205,
"learning_rate": 4.5952021651364204e-05,
"loss": 0.6434,
"step": 6851
},
{
"epoch": 0.10613245737814893,
"grad_norm": 0.7047727108001709,
"learning_rate": 4.590608724263786e-05,
"loss": 0.6562,
"step": 6882
},
{
"epoch": 0.10661053151048293,
"grad_norm": 0.9560827612876892,
"learning_rate": 4.585991691155845e-05,
"loss": 0.6422,
"step": 6913
},
{
"epoch": 0.10708860564281694,
"grad_norm": 1.1834834814071655,
"learning_rate": 4.581351117915188e-05,
"loss": 0.6525,
"step": 6944
},
{
"epoch": 0.10756667977515094,
"grad_norm": 0.8007238507270813,
"learning_rate": 4.5766870569100534e-05,
"loss": 0.6575,
"step": 6975
},
{
"epoch": 0.10804475390748494,
"grad_norm": 0.6816396713256836,
"learning_rate": 4.571999560773736e-05,
"loss": 0.6453,
"step": 7006
},
{
"epoch": 0.10852282803981896,
"grad_norm": 0.7240257859230042,
"learning_rate": 4.5672886824039915e-05,
"loss": 0.6542,
"step": 7037
},
{
"epoch": 0.10900090217215296,
"grad_norm": 0.6335741877555847,
"learning_rate": 4.5625544749624435e-05,
"loss": 0.6449,
"step": 7068
},
{
"epoch": 0.10947897630448696,
"grad_norm": 0.6071763038635254,
"learning_rate": 4.5577969918739794e-05,
"loss": 0.6572,
"step": 7099
},
{
"epoch": 0.10995705043682096,
"grad_norm": 0.8242653608322144,
"learning_rate": 4.5530162868261486e-05,
"loss": 0.6483,
"step": 7130
},
{
"epoch": 0.11043512456915497,
"grad_norm": 0.7106221318244934,
"learning_rate": 4.548212413768558e-05,
"loss": 0.6376,
"step": 7161
},
{
"epoch": 0.11091319870148897,
"grad_norm": 0.6587203145027161,
"learning_rate": 4.543385426912261e-05,
"loss": 0.6426,
"step": 7192
},
{
"epoch": 0.11139127283382297,
"grad_norm": 0.8035815954208374,
"learning_rate": 4.53853538072915e-05,
"loss": 0.6428,
"step": 7223
},
{
"epoch": 0.11186934696615698,
"grad_norm": 0.6661431193351746,
"learning_rate": 4.533662329951336e-05,
"loss": 0.6431,
"step": 7254
},
{
"epoch": 0.11234742109849098,
"grad_norm": 0.5891786813735962,
"learning_rate": 4.528766329570536e-05,
"loss": 0.6356,
"step": 7285
},
{
"epoch": 0.11282549523082498,
"grad_norm": 0.7475966811180115,
"learning_rate": 4.523847434837447e-05,
"loss": 0.6348,
"step": 7316
},
{
"epoch": 0.113303569363159,
"grad_norm": 0.7461502552032471,
"learning_rate": 4.518905701261128e-05,
"loss": 0.639,
"step": 7347
},
{
"epoch": 0.113781643495493,
"grad_norm": 0.7351141571998596,
"learning_rate": 4.5139411846083715e-05,
"loss": 0.6473,
"step": 7378
},
{
"epoch": 0.114259717627827,
"grad_norm": 0.7713771462440491,
"learning_rate": 4.508953940903073e-05,
"loss": 0.6433,
"step": 7409
},
{
"epoch": 0.11473779176016101,
"grad_norm": 0.639986515045166,
"learning_rate": 4.5039440264255994e-05,
"loss": 0.6411,
"step": 7440
},
{
"epoch": 0.11521586589249501,
"grad_norm": 0.8399244546890259,
"learning_rate": 4.498911497712155e-05,
"loss": 0.6421,
"step": 7471
},
{
"epoch": 0.11569394002482901,
"grad_norm": 0.7681542634963989,
"learning_rate": 4.493856411554142e-05,
"loss": 0.6371,
"step": 7502
},
{
"epoch": 0.11617201415716301,
"grad_norm": 0.6907945275306702,
"learning_rate": 4.4887788249975206e-05,
"loss": 0.6446,
"step": 7533
},
{
"epoch": 0.11665008828949702,
"grad_norm": 0.7653492093086243,
"learning_rate": 4.4836787953421656e-05,
"loss": 0.6407,
"step": 7564
},
{
"epoch": 0.11712816242183102,
"grad_norm": 0.6535690426826477,
"learning_rate": 4.478556380141218e-05,
"loss": 0.6388,
"step": 7595
},
{
"epoch": 0.11760623655416502,
"grad_norm": 2.6184701919555664,
"learning_rate": 4.4734116372004375e-05,
"loss": 0.6382,
"step": 7626
},
{
"epoch": 0.11808431068649904,
"grad_norm": 0.7846981287002563,
"learning_rate": 4.4682446245775477e-05,
"loss": 0.6364,
"step": 7657
},
{
"epoch": 0.11856238481883304,
"grad_norm": 0.8151688575744629,
"learning_rate": 4.463055400581586e-05,
"loss": 0.6427,
"step": 7688
},
{
"epoch": 0.11904045895116704,
"grad_norm": 0.7663447856903076,
"learning_rate": 4.4578440237722374e-05,
"loss": 0.643,
"step": 7719
},
{
"epoch": 0.11951853308350105,
"grad_norm": 0.7225657105445862,
"learning_rate": 4.452610552959183e-05,
"loss": 0.6395,
"step": 7750
},
{
"epoch": 0.11999660721583505,
"grad_norm": 0.7033439874649048,
"learning_rate": 4.447355047201428e-05,
"loss": 0.6354,
"step": 7781
},
{
"epoch": 0.12047468134816905,
"grad_norm": 0.7693649530410767,
"learning_rate": 4.4420775658066414e-05,
"loss": 0.6436,
"step": 7812
},
{
"epoch": 0.12095275548050306,
"grad_norm": 0.7597894668579102,
"learning_rate": 4.436778168330484e-05,
"loss": 0.6411,
"step": 7843
},
{
"epoch": 0.12143082961283706,
"grad_norm": 0.7899183630943298,
"learning_rate": 4.4314569145759353e-05,
"loss": 0.6381,
"step": 7874
},
{
"epoch": 0.12190890374517106,
"grad_norm": 0.7656906247138977,
"learning_rate": 4.42611386459262e-05,
"loss": 0.6386,
"step": 7905
},
{
"epoch": 0.12238697787750508,
"grad_norm": 0.8353867530822754,
"learning_rate": 4.420749078676133e-05,
"loss": 0.6386,
"step": 7936
},
{
"epoch": 0.12286505200983908,
"grad_norm": 0.6712770462036133,
"learning_rate": 4.4153626173673516e-05,
"loss": 0.6438,
"step": 7967
},
{
"epoch": 0.12334312614217308,
"grad_norm": 0.8008986115455627,
"learning_rate": 4.409954541451762e-05,
"loss": 0.6384,
"step": 7998
},
{
"epoch": 0.12382120027450708,
"grad_norm": 0.6632928252220154,
"learning_rate": 4.404524911958764e-05,
"loss": 0.629,
"step": 8029
},
{
"epoch": 0.12429927440684109,
"grad_norm": 0.6943359971046448,
"learning_rate": 4.399073790160989e-05,
"loss": 0.6456,
"step": 8060
},
{
"epoch": 0.12477734853917509,
"grad_norm": 0.6177495718002319,
"learning_rate": 4.393601237573607e-05,
"loss": 0.6426,
"step": 8091
},
{
"epoch": 0.1252554226715091,
"grad_norm": 0.6089790463447571,
"learning_rate": 4.388107315953628e-05,
"loss": 0.6396,
"step": 8122
},
{
"epoch": 0.1257334968038431,
"grad_norm": 0.6960969567298889,
"learning_rate": 4.382592087299212e-05,
"loss": 0.6293,
"step": 8153
},
{
"epoch": 0.1262115709361771,
"grad_norm": 0.6519069075584412,
"learning_rate": 4.377055613848964e-05,
"loss": 0.6284,
"step": 8184
},
{
"epoch": 0.1266896450685111,
"grad_norm": 0.6861445903778076,
"learning_rate": 4.3714979580812355e-05,
"loss": 0.6386,
"step": 8215
},
{
"epoch": 0.1271677192008451,
"grad_norm": 0.7522263526916504,
"learning_rate": 4.365919182713416e-05,
"loss": 0.6424,
"step": 8246
},
{
"epoch": 0.1276457933331791,
"grad_norm": 0.6868363618850708,
"learning_rate": 4.360319350701226e-05,
"loss": 0.6367,
"step": 8277
},
{
"epoch": 0.12812386746551313,
"grad_norm": 0.5939040780067444,
"learning_rate": 4.3546985252380115e-05,
"loss": 0.6406,
"step": 8308
},
{
"epoch": 0.12860194159784713,
"grad_norm": 0.7341668605804443,
"learning_rate": 4.349056769754021e-05,
"loss": 0.6346,
"step": 8339
},
{
"epoch": 0.12908001573018113,
"grad_norm": 0.7468376755714417,
"learning_rate": 4.3433941479156994e-05,
"loss": 0.6398,
"step": 8370
},
{
"epoch": 0.12955808986251513,
"grad_norm": 0.6253494024276733,
"learning_rate": 4.3377107236249647e-05,
"loss": 0.6375,
"step": 8401
},
{
"epoch": 0.13003616399484913,
"grad_norm": 0.6635178923606873,
"learning_rate": 4.332006561018488e-05,
"loss": 0.6299,
"step": 8432
},
{
"epoch": 0.13051423812718313,
"grad_norm": 0.584648609161377,
"learning_rate": 4.3262817244669683e-05,
"loss": 0.6277,
"step": 8463
},
{
"epoch": 0.13099231225951716,
"grad_norm": 0.5907359719276428,
"learning_rate": 4.3205362785744083e-05,
"loss": 0.6328,
"step": 8494
},
{
"epoch": 0.13147038639185116,
"grad_norm": 0.7864153385162354,
"learning_rate": 4.314770288177384e-05,
"loss": 0.6365,
"step": 8525
},
{
"epoch": 0.13194846052418516,
"grad_norm": 0.6572133302688599,
"learning_rate": 4.308983818344313e-05,
"loss": 0.6312,
"step": 8556
},
{
"epoch": 0.13242653465651916,
"grad_norm": 0.7108786702156067,
"learning_rate": 4.3031769343747206e-05,
"loss": 0.6355,
"step": 8587
},
{
"epoch": 0.13290460878885316,
"grad_norm": 0.6737267374992371,
"learning_rate": 4.297349701798505e-05,
"loss": 0.6256,
"step": 8618
},
{
"epoch": 0.13338268292118716,
"grad_norm": 0.6710911989212036,
"learning_rate": 4.2915021863751916e-05,
"loss": 0.632,
"step": 8649
},
{
"epoch": 0.13386075705352116,
"grad_norm": 0.6295626759529114,
"learning_rate": 4.285634454093198e-05,
"loss": 0.6372,
"step": 8680
},
{
"epoch": 0.13433883118585518,
"grad_norm": 0.6405526399612427,
"learning_rate": 4.279746571169086e-05,
"loss": 0.6364,
"step": 8711
},
{
"epoch": 0.13481690531818918,
"grad_norm": 0.6847979426383972,
"learning_rate": 4.2738386040468136e-05,
"loss": 0.6246,
"step": 8742
},
{
"epoch": 0.13529497945052318,
"grad_norm": 0.7338352799415588,
"learning_rate": 4.2679106193969866e-05,
"loss": 0.6373,
"step": 8773
},
{
"epoch": 0.13577305358285718,
"grad_norm": 0.6530189514160156,
"learning_rate": 4.261962684116106e-05,
"loss": 0.6217,
"step": 8804
},
{
"epoch": 0.13625112771519118,
"grad_norm": 0.6450487375259399,
"learning_rate": 4.2559948653258145e-05,
"loss": 0.6365,
"step": 8835
},
{
"epoch": 0.13672920184752518,
"grad_norm": 0.706732988357544,
"learning_rate": 4.250007230372134e-05,
"loss": 0.6294,
"step": 8866
},
{
"epoch": 0.1372072759798592,
"grad_norm": 0.687240481376648,
"learning_rate": 4.2439998468247126e-05,
"loss": 0.6332,
"step": 8897
},
{
"epoch": 0.1376853501121932,
"grad_norm": 0.6686804890632629,
"learning_rate": 4.2379727824760566e-05,
"loss": 0.6334,
"step": 8928
},
{
"epoch": 0.1381634242445272,
"grad_norm": 0.8371458649635315,
"learning_rate": 4.231926105340768e-05,
"loss": 0.6249,
"step": 8959
},
{
"epoch": 0.1386414983768612,
"grad_norm": 0.7204101085662842,
"learning_rate": 4.225859883654776e-05,
"loss": 0.6309,
"step": 8990
},
{
"epoch": 0.1391195725091952,
"grad_norm": 0.8539360165596008,
"learning_rate": 4.219774185874569e-05,
"loss": 0.6326,
"step": 9021
},
{
"epoch": 0.1395976466415292,
"grad_norm": 0.7264754176139832,
"learning_rate": 4.213669080676418e-05,
"loss": 0.6342,
"step": 9052
},
{
"epoch": 0.14007572077386324,
"grad_norm": 0.6681591868400574,
"learning_rate": 4.2075446369556056e-05,
"loss": 0.6295,
"step": 9083
},
{
"epoch": 0.14055379490619724,
"grad_norm": 0.6293045282363892,
"learning_rate": 4.201400923825648e-05,
"loss": 0.6304,
"step": 9114
},
{
"epoch": 0.14103186903853124,
"grad_norm": 0.6432914137840271,
"learning_rate": 4.195238010617511e-05,
"loss": 0.6215,
"step": 9145
},
{
"epoch": 0.14150994317086524,
"grad_norm": 0.7524629235267639,
"learning_rate": 4.1890559668788344e-05,
"loss": 0.625,
"step": 9176
},
{
"epoch": 0.14198801730319924,
"grad_norm": 0.7128170728683472,
"learning_rate": 4.1828548623731405e-05,
"loss": 0.6342,
"step": 9207
},
{
"epoch": 0.14246609143553324,
"grad_norm": 0.651841938495636,
"learning_rate": 4.1766347670790506e-05,
"loss": 0.6303,
"step": 9238
},
{
"epoch": 0.14294416556786724,
"grad_norm": 0.6290706396102905,
"learning_rate": 4.170395751189495e-05,
"loss": 0.6297,
"step": 9269
},
{
"epoch": 0.14342223970020127,
"grad_norm": 0.6372231841087341,
"learning_rate": 4.164137885110921e-05,
"loss": 0.6244,
"step": 9300
},
{
"epoch": 0.14390031383253526,
"grad_norm": 0.8424429297447205,
"learning_rate": 4.157861239462495e-05,
"loss": 0.6363,
"step": 9331
},
{
"epoch": 0.14437838796486926,
"grad_norm": 0.7121688723564148,
"learning_rate": 4.1515658850753114e-05,
"loss": 0.63,
"step": 9362
},
{
"epoch": 0.14485646209720326,
"grad_norm": 0.7196950316429138,
"learning_rate": 4.145251892991588e-05,
"loss": 0.6254,
"step": 9393
},
{
"epoch": 0.14533453622953726,
"grad_norm": 0.6595038175582886,
"learning_rate": 4.138919334463868e-05,
"loss": 0.6207,
"step": 9424
},
{
"epoch": 0.14581261036187126,
"grad_norm": 0.6118171811103821,
"learning_rate": 4.1325682809542124e-05,
"loss": 0.6233,
"step": 9455
},
{
"epoch": 0.1462906844942053,
"grad_norm": 0.7232321500778198,
"learning_rate": 4.126198804133398e-05,
"loss": 0.6328,
"step": 9486
},
{
"epoch": 0.1467687586265393,
"grad_norm": 0.691681981086731,
"learning_rate": 4.1198109758801055e-05,
"loss": 0.634,
"step": 9517
},
{
"epoch": 0.1472468327588733,
"grad_norm": 0.6887443661689758,
"learning_rate": 4.113404868280107e-05,
"loss": 0.6275,
"step": 9548
},
{
"epoch": 0.1477249068912073,
"grad_norm": 0.6932326555252075,
"learning_rate": 4.106980553625457e-05,
"loss": 0.6233,
"step": 9579
},
{
"epoch": 0.1482029810235413,
"grad_norm": 0.7189227342605591,
"learning_rate": 4.100538104413674e-05,
"loss": 0.6229,
"step": 9610
},
{
"epoch": 0.1486810551558753,
"grad_norm": 0.6898639798164368,
"learning_rate": 4.09407759334692e-05,
"loss": 0.6251,
"step": 9641
},
{
"epoch": 0.1491591292882093,
"grad_norm": 0.6311922669410706,
"learning_rate": 4.087599093331186e-05,
"loss": 0.6249,
"step": 9672
},
{
"epoch": 0.14963720342054332,
"grad_norm": 0.5781577825546265,
"learning_rate": 4.081102677475462e-05,
"loss": 0.6337,
"step": 9703
},
{
"epoch": 0.15011527755287732,
"grad_norm": 0.6525286436080933,
"learning_rate": 4.0745884190909194e-05,
"loss": 0.6347,
"step": 9734
},
{
"epoch": 0.15059335168521132,
"grad_norm": 0.6761188507080078,
"learning_rate": 4.0680563916900796e-05,
"loss": 0.6267,
"step": 9765
},
{
"epoch": 0.15107142581754532,
"grad_norm": 0.674532413482666,
"learning_rate": 4.0615066689859815e-05,
"loss": 0.6303,
"step": 9796
},
{
"epoch": 0.15154949994987932,
"grad_norm": 0.5923493504524231,
"learning_rate": 4.0549393248913584e-05,
"loss": 0.6241,
"step": 9827
},
{
"epoch": 0.15202757408221332,
"grad_norm": 0.6309007406234741,
"learning_rate": 4.048354433517794e-05,
"loss": 0.6231,
"step": 9858
},
{
"epoch": 0.15250564821454735,
"grad_norm": 0.7521491050720215,
"learning_rate": 4.0417520691748916e-05,
"loss": 0.6214,
"step": 9889
},
{
"epoch": 0.15298372234688135,
"grad_norm": 0.8196119666099548,
"learning_rate": 4.035132306369438e-05,
"loss": 0.6254,
"step": 9920
},
{
"epoch": 0.15346179647921535,
"grad_norm": 0.6895152926445007,
"learning_rate": 4.028495219804555e-05,
"loss": 0.6227,
"step": 9951
},
{
"epoch": 0.15393987061154935,
"grad_norm": 0.6823534369468689,
"learning_rate": 4.021840884378864e-05,
"loss": 0.6281,
"step": 9982
},
{
"epoch": 0.15441794474388335,
"grad_norm": 0.6531606912612915,
"learning_rate": 4.015169375185633e-05,
"loss": 0.6259,
"step": 10013
},
{
"epoch": 0.15489601887621735,
"grad_norm": 0.5898056626319885,
"learning_rate": 4.0084807675119396e-05,
"loss": 0.6184,
"step": 10044
},
{
"epoch": 0.15537409300855134,
"grad_norm": 0.6368978023529053,
"learning_rate": 4.0017751368378106e-05,
"loss": 0.6196,
"step": 10075
},
{
"epoch": 0.15585216714088537,
"grad_norm": 0.8086898922920227,
"learning_rate": 3.995052558835377e-05,
"loss": 0.6226,
"step": 10106
},
{
"epoch": 0.15633024127321937,
"grad_norm": 0.6620404124259949,
"learning_rate": 3.988313109368017e-05,
"loss": 0.6146,
"step": 10137
},
{
"epoch": 0.15680831540555337,
"grad_norm": 0.6264041662216187,
"learning_rate": 3.981556864489504e-05,
"loss": 0.6254,
"step": 10168
},
{
"epoch": 0.15728638953788737,
"grad_norm": 0.6619133353233337,
"learning_rate": 3.974783900443142e-05,
"loss": 0.6233,
"step": 10199
},
{
"epoch": 0.15776446367022137,
"grad_norm": 0.6223868131637573,
"learning_rate": 3.9679942936609095e-05,
"loss": 0.6232,
"step": 10230
},
{
"epoch": 0.15824253780255537,
"grad_norm": 0.7436427474021912,
"learning_rate": 3.961188120762596e-05,
"loss": 0.6194,
"step": 10261
},
{
"epoch": 0.1587206119348894,
"grad_norm": 0.7166286110877991,
"learning_rate": 3.954365458554938e-05,
"loss": 0.6267,
"step": 10292
},
{
"epoch": 0.1591986860672234,
"grad_norm": 0.6377813220024109,
"learning_rate": 3.947526384030751e-05,
"loss": 0.6145,
"step": 10323
},
{
"epoch": 0.1596767601995574,
"grad_norm": 1.5197011232376099,
"learning_rate": 3.9406709743680624e-05,
"loss": 0.6261,
"step": 10354
},
{
"epoch": 0.1601548343318914,
"grad_norm": 0.6208730340003967,
"learning_rate": 3.9337993069292366e-05,
"loss": 0.6201,
"step": 10385
},
{
"epoch": 0.1606329084642254,
"grad_norm": 0.6887170076370239,
"learning_rate": 3.926911459260109e-05,
"loss": 0.6235,
"step": 10416
},
{
"epoch": 0.1611109825965594,
"grad_norm": 0.6381047368049622,
"learning_rate": 3.920007509089102e-05,
"loss": 0.6189,
"step": 10447
},
{
"epoch": 0.1615890567288934,
"grad_norm": 0.6249068379402161,
"learning_rate": 3.913087534326357e-05,
"loss": 0.6222,
"step": 10478
},
{
"epoch": 0.16206713086122743,
"grad_norm": 0.659757137298584,
"learning_rate": 3.9061516130628475e-05,
"loss": 0.6345,
"step": 10509
},
{
"epoch": 0.16254520499356143,
"grad_norm": 0.6268470883369446,
"learning_rate": 3.8991998235695025e-05,
"loss": 0.6197,
"step": 10540
},
{
"epoch": 0.16302327912589543,
"grad_norm": 0.6520307660102844,
"learning_rate": 3.8922322442963224e-05,
"loss": 0.6122,
"step": 10571
},
{
"epoch": 0.16350135325822943,
"grad_norm": 0.5919711589813232,
"learning_rate": 3.885248953871491e-05,
"loss": 0.6237,
"step": 10602
},
{
"epoch": 0.16397942739056343,
"grad_norm": 0.6761168837547302,
"learning_rate": 3.8782500311004915e-05,
"loss": 0.615,
"step": 10633
},
{
"epoch": 0.16445750152289743,
"grad_norm": 0.7123913168907166,
"learning_rate": 3.871235554965218e-05,
"loss": 0.6178,
"step": 10664
},
{
"epoch": 0.16493557565523145,
"grad_norm": 0.5414535403251648,
"learning_rate": 3.864205604623078e-05,
"loss": 0.6319,
"step": 10695
},
{
"epoch": 0.16541364978756545,
"grad_norm": 0.741256594657898,
"learning_rate": 3.857160259406107e-05,
"loss": 0.6102,
"step": 10726
},
{
"epoch": 0.16589172391989945,
"grad_norm": 0.6938264966011047,
"learning_rate": 3.8500995988200674e-05,
"loss": 0.6127,
"step": 10757
},
{
"epoch": 0.16636979805223345,
"grad_norm": 0.7152143716812134,
"learning_rate": 3.843023702543556e-05,
"loss": 0.6119,
"step": 10788
},
{
"epoch": 0.16684787218456745,
"grad_norm": 0.6516321897506714,
"learning_rate": 3.8359326504270984e-05,
"loss": 0.6224,
"step": 10819
},
{
"epoch": 0.16732594631690145,
"grad_norm": 0.6300469040870667,
"learning_rate": 3.828826522492255e-05,
"loss": 0.6229,
"step": 10850
},
{
"epoch": 0.16780402044923548,
"grad_norm": 0.645636260509491,
"learning_rate": 3.821705398930713e-05,
"loss": 0.6284,
"step": 10881
},
{
"epoch": 0.16828209458156948,
"grad_norm": 0.5895135402679443,
"learning_rate": 3.814569360103385e-05,
"loss": 0.6142,
"step": 10912
},
{
"epoch": 0.16876016871390348,
"grad_norm": 0.6440445184707642,
"learning_rate": 3.807418486539499e-05,
"loss": 0.6158,
"step": 10943
},
{
"epoch": 0.16923824284623748,
"grad_norm": 0.6739510893821716,
"learning_rate": 3.80025285893569e-05,
"loss": 0.6149,
"step": 10974
},
{
"epoch": 0.16971631697857148,
"grad_norm": 0.627185046672821,
"learning_rate": 3.793072558155093e-05,
"loss": 0.614,
"step": 11005
},
{
"epoch": 0.17019439111090548,
"grad_norm": 0.710350513458252,
"learning_rate": 3.785877665226426e-05,
"loss": 0.6214,
"step": 11036
},
{
"epoch": 0.17067246524323948,
"grad_norm": 0.6784375309944153,
"learning_rate": 3.778668261343079e-05,
"loss": 0.6123,
"step": 11067
},
{
"epoch": 0.1711505393755735,
"grad_norm": 0.6646184325218201,
"learning_rate": 3.771444427862192e-05,
"loss": 0.6178,
"step": 11098
},
{
"epoch": 0.1716286135079075,
"grad_norm": 0.7227056622505188,
"learning_rate": 3.7642062463037465e-05,
"loss": 0.6158,
"step": 11129
},
{
"epoch": 0.1721066876402415,
"grad_norm": 0.5937723517417908,
"learning_rate": 3.7569537983496373e-05,
"loss": 0.6194,
"step": 11160
},
{
"epoch": 0.1725847617725755,
"grad_norm": 0.6352291703224182,
"learning_rate": 3.749687165842753e-05,
"loss": 0.6286,
"step": 11191
},
{
"epoch": 0.1730628359049095,
"grad_norm": 0.5864464044570923,
"learning_rate": 3.7424064307860536e-05,
"loss": 0.6186,
"step": 11222
},
{
"epoch": 0.1735409100372435,
"grad_norm": 0.5402609705924988,
"learning_rate": 3.735111675341645e-05,
"loss": 0.6072,
"step": 11253
},
{
"epoch": 0.17401898416957753,
"grad_norm": 0.6429933905601501,
"learning_rate": 3.7278029818298524e-05,
"loss": 0.608,
"step": 11284
},
{
"epoch": 0.17449705830191153,
"grad_norm": 0.5420451164245605,
"learning_rate": 3.720480432728287e-05,
"loss": 0.6174,
"step": 11315
},
{
"epoch": 0.17497513243424553,
"grad_norm": 0.5691730380058289,
"learning_rate": 3.71314411067092e-05,
"loss": 0.6138,
"step": 11346
},
{
"epoch": 0.17545320656657953,
"grad_norm": 0.6198572516441345,
"learning_rate": 3.70579409844715e-05,
"loss": 0.6166,
"step": 11377
},
{
"epoch": 0.17593128069891353,
"grad_norm": 0.6781229972839355,
"learning_rate": 3.698430479000865e-05,
"loss": 0.6083,
"step": 11408
},
{
"epoch": 0.17640935483124753,
"grad_norm": 0.6572268605232239,
"learning_rate": 3.691053335429509e-05,
"loss": 0.6179,
"step": 11439
},
{
"epoch": 0.17688742896358153,
"grad_norm": 0.6720935702323914,
"learning_rate": 3.683662750983147e-05,
"loss": 0.6039,
"step": 11470
},
{
"epoch": 0.17736550309591556,
"grad_norm": 0.6098059415817261,
"learning_rate": 3.676258809063518e-05,
"loss": 0.608,
"step": 11501
},
{
"epoch": 0.17784357722824956,
"grad_norm": 0.6108371615409851,
"learning_rate": 3.6688415932231004e-05,
"loss": 0.6147,
"step": 11532
},
{
"epoch": 0.17832165136058356,
"grad_norm": 0.8087718486785889,
"learning_rate": 3.661411187164166e-05,
"loss": 0.6091,
"step": 11563
},
{
"epoch": 0.17879972549291756,
"grad_norm": 0.719902753829956,
"learning_rate": 3.65396767473784e-05,
"loss": 0.6107,
"step": 11594
},
{
"epoch": 0.17927779962525156,
"grad_norm": 0.5758282542228699,
"learning_rate": 3.6465111399431465e-05,
"loss": 0.6236,
"step": 11625
},
{
"epoch": 0.17975587375758556,
"grad_norm": 0.5602071285247803,
"learning_rate": 3.6390416669260674e-05,
"loss": 0.611,
"step": 11656
},
{
"epoch": 0.1802339478899196,
"grad_norm": 0.6070683002471924,
"learning_rate": 3.63155933997859e-05,
"loss": 0.609,
"step": 11687
},
{
"epoch": 0.1807120220222536,
"grad_norm": 0.64506596326828,
"learning_rate": 3.624064243537758e-05,
"loss": 0.605,
"step": 11718
},
{
"epoch": 0.1811900961545876,
"grad_norm": 0.6906920075416565,
"learning_rate": 3.616556462184716e-05,
"loss": 0.6168,
"step": 11749
},
{
"epoch": 0.1816681702869216,
"grad_norm": 0.6724758148193359,
"learning_rate": 3.609036080643755e-05,
"loss": 0.6143,
"step": 11780
},
{
"epoch": 0.1821462444192556,
"grad_norm": 0.6033251881599426,
"learning_rate": 3.60150318378136e-05,
"loss": 0.6161,
"step": 11811
},
{
"epoch": 0.1826243185515896,
"grad_norm": 0.6713966727256775,
"learning_rate": 3.5939578566052465e-05,
"loss": 0.619,
"step": 11842
},
{
"epoch": 0.1831023926839236,
"grad_norm": 0.6794861555099487,
"learning_rate": 3.586400184263408e-05,
"loss": 0.6116,
"step": 11873
},
{
"epoch": 0.18358046681625761,
"grad_norm": 0.5780847668647766,
"learning_rate": 3.578830252043148e-05,
"loss": 0.6159,
"step": 11904
},
{
"epoch": 0.18405854094859161,
"grad_norm": 0.6031593680381775,
"learning_rate": 3.571248145370125e-05,
"loss": 0.6135,
"step": 11935
},
{
"epoch": 0.18453661508092561,
"grad_norm": 0.6339271664619446,
"learning_rate": 3.5636539498073794e-05,
"loss": 0.6113,
"step": 11966
},
{
"epoch": 0.18501468921325961,
"grad_norm": 0.6663223505020142,
"learning_rate": 3.556047751054378e-05,
"loss": 0.6184,
"step": 11997
},
{
"epoch": 0.18549276334559361,
"grad_norm": 0.9799377918243408,
"learning_rate": 3.548429634946039e-05,
"loss": 0.6103,
"step": 12028
},
{
"epoch": 0.18597083747792761,
"grad_norm": 0.7260358929634094,
"learning_rate": 3.540799687451768e-05,
"loss": 0.6245,
"step": 12059
},
{
"epoch": 0.18644891161026164,
"grad_norm": 0.7097154259681702,
"learning_rate": 3.533157994674485e-05,
"loss": 0.6194,
"step": 12090
},
{
"epoch": 0.18692698574259564,
"grad_norm": 0.5897482633590698,
"learning_rate": 3.5255046428496546e-05,
"loss": 0.6171,
"step": 12121
},
{
"epoch": 0.18740505987492964,
"grad_norm": 0.6374841332435608,
"learning_rate": 3.517839718344311e-05,
"loss": 0.6141,
"step": 12152
},
{
"epoch": 0.18788313400726364,
"grad_norm": 0.6854335069656372,
"learning_rate": 3.510163307656086e-05,
"loss": 0.6077,
"step": 12183
}
],
"logging_steps": 31,
"max_steps": 30517,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 3052,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 8.48541631240156e+18,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}