atsuki-yamaguchi's picture
Upload folder using huggingface_hub
fcc5f9f verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.3862802176939628,
"eval_steps": 500,
"global_step": 9156,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0013078513268362654,
"grad_norm": 31.18772315979004,
"learning_rate": 1.0157273918741808e-06,
"loss": 9.3832,
"step": 31
},
{
"epoch": 0.0026157026536725307,
"grad_norm": 24.045747756958008,
"learning_rate": 2.0314547837483616e-06,
"loss": 7.9011,
"step": 62
},
{
"epoch": 0.003923553980508796,
"grad_norm": 16.459999084472656,
"learning_rate": 3.0471821756225426e-06,
"loss": 6.3655,
"step": 93
},
{
"epoch": 0.0052314053073450615,
"grad_norm": 5.892439365386963,
"learning_rate": 4.062909567496723e-06,
"loss": 5.1577,
"step": 124
},
{
"epoch": 0.0065392566341813275,
"grad_norm": 7.445889949798584,
"learning_rate": 5.078636959370905e-06,
"loss": 4.4767,
"step": 155
},
{
"epoch": 0.007847107961017593,
"grad_norm": 8.36017894744873,
"learning_rate": 6.094364351245085e-06,
"loss": 4.0756,
"step": 186
},
{
"epoch": 0.009154959287853858,
"grad_norm": 13.23757266998291,
"learning_rate": 7.110091743119267e-06,
"loss": 3.7893,
"step": 217
},
{
"epoch": 0.010462810614690123,
"grad_norm": 10.12263011932373,
"learning_rate": 8.125819134993446e-06,
"loss": 3.6154,
"step": 248
},
{
"epoch": 0.01177066194152639,
"grad_norm": 6.559150218963623,
"learning_rate": 9.141546526867629e-06,
"loss": 3.4391,
"step": 279
},
{
"epoch": 0.013078513268362655,
"grad_norm": 6.898993015289307,
"learning_rate": 1.015727391874181e-05,
"loss": 3.3229,
"step": 310
},
{
"epoch": 0.01438636459519892,
"grad_norm": 8.434200286865234,
"learning_rate": 1.117300131061599e-05,
"loss": 3.225,
"step": 341
},
{
"epoch": 0.015694215922035185,
"grad_norm": 7.735746383666992,
"learning_rate": 1.218872870249017e-05,
"loss": 3.1381,
"step": 372
},
{
"epoch": 0.01700206724887145,
"grad_norm": 7.964824676513672,
"learning_rate": 1.3204456094364351e-05,
"loss": 3.06,
"step": 403
},
{
"epoch": 0.018309918575707716,
"grad_norm": 6.081613063812256,
"learning_rate": 1.4220183486238533e-05,
"loss": 3.0052,
"step": 434
},
{
"epoch": 0.01961776990254398,
"grad_norm": 6.740341663360596,
"learning_rate": 1.5235910878112714e-05,
"loss": 2.9035,
"step": 465
},
{
"epoch": 0.020925621229380246,
"grad_norm": 4.908828258514404,
"learning_rate": 1.6251638269986893e-05,
"loss": 2.8602,
"step": 496
},
{
"epoch": 0.02223347255621651,
"grad_norm": 5.6150360107421875,
"learning_rate": 1.7267365661861077e-05,
"loss": 2.8223,
"step": 527
},
{
"epoch": 0.02354132388305278,
"grad_norm": 3.5981647968292236,
"learning_rate": 1.8283093053735257e-05,
"loss": 2.7488,
"step": 558
},
{
"epoch": 0.024849175209889045,
"grad_norm": 4.921584606170654,
"learning_rate": 1.9298820445609438e-05,
"loss": 2.741,
"step": 589
},
{
"epoch": 0.02615702653672531,
"grad_norm": 3.8303065299987793,
"learning_rate": 2.031454783748362e-05,
"loss": 2.6923,
"step": 620
},
{
"epoch": 0.027464877863561575,
"grad_norm": 3.3674638271331787,
"learning_rate": 2.13302752293578e-05,
"loss": 2.6402,
"step": 651
},
{
"epoch": 0.02877272919039784,
"grad_norm": 4.096519947052002,
"learning_rate": 2.234600262123198e-05,
"loss": 2.6372,
"step": 682
},
{
"epoch": 0.030080580517234105,
"grad_norm": 3.10866641998291,
"learning_rate": 2.336173001310616e-05,
"loss": 2.5979,
"step": 713
},
{
"epoch": 0.03138843184407037,
"grad_norm": 3.3147943019866943,
"learning_rate": 2.437745740498034e-05,
"loss": 2.549,
"step": 744
},
{
"epoch": 0.032696283170906636,
"grad_norm": 3.045280933380127,
"learning_rate": 2.5393184796854525e-05,
"loss": 2.5487,
"step": 775
},
{
"epoch": 0.0340041344977429,
"grad_norm": 3.2536299228668213,
"learning_rate": 2.6408912188728702e-05,
"loss": 2.5138,
"step": 806
},
{
"epoch": 0.035311985824579166,
"grad_norm": 3.108548641204834,
"learning_rate": 2.7424639580602886e-05,
"loss": 2.5006,
"step": 837
},
{
"epoch": 0.03661983715141543,
"grad_norm": 3.048689603805542,
"learning_rate": 2.8440366972477066e-05,
"loss": 2.4434,
"step": 868
},
{
"epoch": 0.037927688478251696,
"grad_norm": 2.8378429412841797,
"learning_rate": 2.9456094364351244e-05,
"loss": 2.4403,
"step": 899
},
{
"epoch": 0.03923553980508796,
"grad_norm": 2.731419324874878,
"learning_rate": 3.0471821756225428e-05,
"loss": 2.4028,
"step": 930
},
{
"epoch": 0.04054339113192423,
"grad_norm": 2.4825587272644043,
"learning_rate": 3.148754914809961e-05,
"loss": 2.3695,
"step": 961
},
{
"epoch": 0.04185124245876049,
"grad_norm": 2.5020129680633545,
"learning_rate": 3.2503276539973785e-05,
"loss": 2.3513,
"step": 992
},
{
"epoch": 0.04315909378559676,
"grad_norm": 2.374310255050659,
"learning_rate": 3.351900393184797e-05,
"loss": 2.3221,
"step": 1023
},
{
"epoch": 0.04446694511243302,
"grad_norm": 2.4097862243652344,
"learning_rate": 3.453473132372215e-05,
"loss": 2.3366,
"step": 1054
},
{
"epoch": 0.045774796439269294,
"grad_norm": 2.573655128479004,
"learning_rate": 3.555045871559633e-05,
"loss": 2.2997,
"step": 1085
},
{
"epoch": 0.04708264776610556,
"grad_norm": 2.5408434867858887,
"learning_rate": 3.6566186107470514e-05,
"loss": 2.2729,
"step": 1116
},
{
"epoch": 0.048390499092941824,
"grad_norm": 2.222527265548706,
"learning_rate": 3.7581913499344695e-05,
"loss": 2.2716,
"step": 1147
},
{
"epoch": 0.04969835041977809,
"grad_norm": 2.342715263366699,
"learning_rate": 3.8597640891218876e-05,
"loss": 2.2661,
"step": 1178
},
{
"epoch": 0.051006201746614355,
"grad_norm": 2.134208917617798,
"learning_rate": 3.9613368283093056e-05,
"loss": 2.2206,
"step": 1209
},
{
"epoch": 0.05231405307345062,
"grad_norm": 2.2482688426971436,
"learning_rate": 4.062909567496724e-05,
"loss": 2.2207,
"step": 1240
},
{
"epoch": 0.053621904400286885,
"grad_norm": 2.2316832542419434,
"learning_rate": 4.164482306684142e-05,
"loss": 2.1847,
"step": 1271
},
{
"epoch": 0.05492975572712315,
"grad_norm": 2.2321085929870605,
"learning_rate": 4.26605504587156e-05,
"loss": 2.2079,
"step": 1302
},
{
"epoch": 0.056237607053959415,
"grad_norm": 2.115349054336548,
"learning_rate": 4.367627785058978e-05,
"loss": 2.1458,
"step": 1333
},
{
"epoch": 0.05754545838079568,
"grad_norm": 2.210914373397827,
"learning_rate": 4.469200524246396e-05,
"loss": 2.1435,
"step": 1364
},
{
"epoch": 0.058853309707631946,
"grad_norm": 2.1882164478302,
"learning_rate": 4.570773263433814e-05,
"loss": 2.1389,
"step": 1395
},
{
"epoch": 0.06016116103446821,
"grad_norm": 2.01705265045166,
"learning_rate": 4.672346002621232e-05,
"loss": 2.1599,
"step": 1426
},
{
"epoch": 0.061469012361304476,
"grad_norm": 1.9447391033172607,
"learning_rate": 4.77391874180865e-05,
"loss": 2.1254,
"step": 1457
},
{
"epoch": 0.06277686368814074,
"grad_norm": 1.9278348684310913,
"learning_rate": 4.875491480996068e-05,
"loss": 2.1109,
"step": 1488
},
{
"epoch": 0.064084715014977,
"grad_norm": 167.2142791748047,
"learning_rate": 4.977064220183487e-05,
"loss": 2.12,
"step": 1519
},
{
"epoch": 0.06539256634181327,
"grad_norm": 15.058513641357422,
"learning_rate": 4.9999915451558777e-05,
"loss": 3.2174,
"step": 1550
},
{
"epoch": 0.06670041766864954,
"grad_norm": 2.8855061531066895,
"learning_rate": 4.999955597496219e-05,
"loss": 2.191,
"step": 1581
},
{
"epoch": 0.0680082689954858,
"grad_norm": 1.9484020471572876,
"learning_rate": 4.9998914381774255e-05,
"loss": 2.1053,
"step": 1612
},
{
"epoch": 0.06931612032232207,
"grad_norm": 5.792853832244873,
"learning_rate": 4.999799067923527e-05,
"loss": 2.1058,
"step": 1643
},
{
"epoch": 0.07062397164915833,
"grad_norm": 5.8536553382873535,
"learning_rate": 4.999678487776908e-05,
"loss": 2.0906,
"step": 1674
},
{
"epoch": 0.0719318229759946,
"grad_norm": 2.8606152534484863,
"learning_rate": 4.9995296990983006e-05,
"loss": 2.1356,
"step": 1705
},
{
"epoch": 0.07323967430283086,
"grad_norm": 2.423034191131592,
"learning_rate": 4.999352703566763e-05,
"loss": 2.0468,
"step": 1736
},
{
"epoch": 0.07454752562966713,
"grad_norm": 1.7297399044036865,
"learning_rate": 4.999147503179668e-05,
"loss": 2.0687,
"step": 1767
},
{
"epoch": 0.07585537695650339,
"grad_norm": 4.29679536819458,
"learning_rate": 4.998914100252672e-05,
"loss": 2.026,
"step": 1798
},
{
"epoch": 0.07716322828333966,
"grad_norm": 1.942807674407959,
"learning_rate": 4.998652497419696e-05,
"loss": 2.034,
"step": 1829
},
{
"epoch": 0.07847107961017592,
"grad_norm": 1.8823152780532837,
"learning_rate": 4.9983626976328927e-05,
"loss": 2.0186,
"step": 1860
},
{
"epoch": 0.07977893093701219,
"grad_norm": 2.135873556137085,
"learning_rate": 4.998044704162613e-05,
"loss": 2.0047,
"step": 1891
},
{
"epoch": 0.08108678226384845,
"grad_norm": 3.113102436065674,
"learning_rate": 4.9976985205973705e-05,
"loss": 1.9954,
"step": 1922
},
{
"epoch": 0.08239463359068472,
"grad_norm": 2.253596305847168,
"learning_rate": 4.997324150843799e-05,
"loss": 1.9744,
"step": 1953
},
{
"epoch": 0.08370248491752098,
"grad_norm": 5.765998363494873,
"learning_rate": 4.99692159912661e-05,
"loss": 1.9908,
"step": 1984
},
{
"epoch": 0.08501033624435725,
"grad_norm": 2.8442440032958984,
"learning_rate": 4.996490869988546e-05,
"loss": 1.9803,
"step": 2015
},
{
"epoch": 0.08631818757119351,
"grad_norm": 1.7528963088989258,
"learning_rate": 4.996031968290326e-05,
"loss": 1.9419,
"step": 2046
},
{
"epoch": 0.08762603889802978,
"grad_norm": 1.7754414081573486,
"learning_rate": 4.995544899210594e-05,
"loss": 1.9485,
"step": 2077
},
{
"epoch": 0.08893389022486604,
"grad_norm": 1.6553430557250977,
"learning_rate": 4.9950296682458583e-05,
"loss": 1.9469,
"step": 2108
},
{
"epoch": 0.09024174155170232,
"grad_norm": 1.7296561002731323,
"learning_rate": 4.994486281210429e-05,
"loss": 1.9195,
"step": 2139
},
{
"epoch": 0.09154959287853859,
"grad_norm": 1.7001283168792725,
"learning_rate": 4.9939147442363566e-05,
"loss": 1.9324,
"step": 2170
},
{
"epoch": 0.09285744420537485,
"grad_norm": 7.332984447479248,
"learning_rate": 4.9933150637733574e-05,
"loss": 1.9171,
"step": 2201
},
{
"epoch": 0.09416529553221112,
"grad_norm": 1.6546969413757324,
"learning_rate": 4.992687246588743e-05,
"loss": 1.9152,
"step": 2232
},
{
"epoch": 0.09547314685904738,
"grad_norm": 1.843203067779541,
"learning_rate": 4.992031299767347e-05,
"loss": 1.9015,
"step": 2263
},
{
"epoch": 0.09678099818588365,
"grad_norm": 4.51094388961792,
"learning_rate": 4.9913472307114386e-05,
"loss": 1.917,
"step": 2294
},
{
"epoch": 0.09808884951271991,
"grad_norm": 1.5267099142074585,
"learning_rate": 4.9906350471406446e-05,
"loss": 1.8985,
"step": 2325
},
{
"epoch": 0.09939670083955618,
"grad_norm": 2.0826663970947266,
"learning_rate": 4.989894757091861e-05,
"loss": 1.8968,
"step": 2356
},
{
"epoch": 0.10070455216639244,
"grad_norm": 1.4929964542388916,
"learning_rate": 4.989126368919158e-05,
"loss": 1.8711,
"step": 2387
},
{
"epoch": 0.10201240349322871,
"grad_norm": 1.478498935699463,
"learning_rate": 4.988329891293693e-05,
"loss": 1.8672,
"step": 2418
},
{
"epoch": 0.10332025482006497,
"grad_norm": 1.4785969257354736,
"learning_rate": 4.987505333203608e-05,
"loss": 1.8435,
"step": 2449
},
{
"epoch": 0.10462810614690124,
"grad_norm": 1.524807333946228,
"learning_rate": 4.9866527039539276e-05,
"loss": 1.8543,
"step": 2480
},
{
"epoch": 0.1059359574737375,
"grad_norm": 1.4232748746871948,
"learning_rate": 4.9857720131664594e-05,
"loss": 1.8463,
"step": 2511
},
{
"epoch": 0.10724380880057377,
"grad_norm": 1.7178603410720825,
"learning_rate": 4.9848632707796773e-05,
"loss": 1.8512,
"step": 2542
},
{
"epoch": 0.10855166012741004,
"grad_norm": 1.4624053239822388,
"learning_rate": 4.9839264870486155e-05,
"loss": 1.8421,
"step": 2573
},
{
"epoch": 0.1098595114542463,
"grad_norm": 2.0191211700439453,
"learning_rate": 4.9829616725447526e-05,
"loss": 1.8393,
"step": 2604
},
{
"epoch": 0.11116736278108257,
"grad_norm": 1.4346847534179688,
"learning_rate": 4.981968838155888e-05,
"loss": 1.8149,
"step": 2635
},
{
"epoch": 0.11247521410791883,
"grad_norm": 1.4674941301345825,
"learning_rate": 4.980947995086024e-05,
"loss": 1.8215,
"step": 2666
},
{
"epoch": 0.1137830654347551,
"grad_norm": 1.4122453927993774,
"learning_rate": 4.979899154855234e-05,
"loss": 1.8265,
"step": 2697
},
{
"epoch": 0.11509091676159136,
"grad_norm": 1.4748708009719849,
"learning_rate": 4.9788223292995386e-05,
"loss": 1.8149,
"step": 2728
},
{
"epoch": 0.11639876808842763,
"grad_norm": 1.40476655960083,
"learning_rate": 4.977717530570768e-05,
"loss": 1.8316,
"step": 2759
},
{
"epoch": 0.11770661941526389,
"grad_norm": 1.4909706115722656,
"learning_rate": 4.976584771136425e-05,
"loss": 1.8208,
"step": 2790
},
{
"epoch": 0.11901447074210016,
"grad_norm": 1.3661363124847412,
"learning_rate": 4.975424063779547e-05,
"loss": 1.8038,
"step": 2821
},
{
"epoch": 0.12032232206893642,
"grad_norm": 1.422453761100769,
"learning_rate": 4.974235421598557e-05,
"loss": 1.7926,
"step": 2852
},
{
"epoch": 0.12163017339577269,
"grad_norm": 1.3766013383865356,
"learning_rate": 4.973018858007122e-05,
"loss": 1.8192,
"step": 2883
},
{
"epoch": 0.12293802472260895,
"grad_norm": 1.3904545307159424,
"learning_rate": 4.9717743867339963e-05,
"loss": 1.7653,
"step": 2914
},
{
"epoch": 0.12424587604944522,
"grad_norm": 1.369018316268921,
"learning_rate": 4.9705020218228695e-05,
"loss": 1.7879,
"step": 2945
},
{
"epoch": 0.12555372737628148,
"grad_norm": 1.5040982961654663,
"learning_rate": 4.969201777632205e-05,
"loss": 1.7777,
"step": 2976
},
{
"epoch": 0.12686157870311776,
"grad_norm": 1.3161810636520386,
"learning_rate": 4.9678736688350846e-05,
"loss": 1.7654,
"step": 3007
},
{
"epoch": 0.128169430029954,
"grad_norm": 1.3379234075546265,
"learning_rate": 4.966517710419033e-05,
"loss": 1.7699,
"step": 3038
},
{
"epoch": 0.1294772813567903,
"grad_norm": 1.3384974002838135,
"learning_rate": 4.965133917685858e-05,
"loss": 1.7915,
"step": 3069
},
{
"epoch": 0.13078513268362654,
"grad_norm": 1.2991316318511963,
"learning_rate": 4.9637223062514714e-05,
"loss": 1.7636,
"step": 3100
},
{
"epoch": 0.13209298401046282,
"grad_norm": 2.500105857849121,
"learning_rate": 4.962282892045718e-05,
"loss": 1.7573,
"step": 3131
},
{
"epoch": 0.13340083533729907,
"grad_norm": 1.2635610103607178,
"learning_rate": 4.9608156913121904e-05,
"loss": 1.7537,
"step": 3162
},
{
"epoch": 0.13470868666413535,
"grad_norm": 1.2823940515518188,
"learning_rate": 4.959320720608049e-05,
"loss": 1.7709,
"step": 3193
},
{
"epoch": 0.1360165379909716,
"grad_norm": 1.2606983184814453,
"learning_rate": 4.9577979968038354e-05,
"loss": 1.7611,
"step": 3224
},
{
"epoch": 0.13732438931780788,
"grad_norm": 1.305782437324524,
"learning_rate": 4.956247537083282e-05,
"loss": 1.7547,
"step": 3255
},
{
"epoch": 0.13863224064464413,
"grad_norm": 1.2986756563186646,
"learning_rate": 4.9546693589431145e-05,
"loss": 1.7241,
"step": 3286
},
{
"epoch": 0.1399400919714804,
"grad_norm": 1.3163083791732788,
"learning_rate": 4.9530634801928595e-05,
"loss": 1.7352,
"step": 3317
},
{
"epoch": 0.14124794329831666,
"grad_norm": 1.2844524383544922,
"learning_rate": 4.9514299189546395e-05,
"loss": 1.7316,
"step": 3348
},
{
"epoch": 0.14255579462515294,
"grad_norm": 1.2861442565917969,
"learning_rate": 4.949768693662973e-05,
"loss": 1.7233,
"step": 3379
},
{
"epoch": 0.1438636459519892,
"grad_norm": 1.2419766187667847,
"learning_rate": 4.948079823064559e-05,
"loss": 1.7278,
"step": 3410
},
{
"epoch": 0.14517149727882547,
"grad_norm": 1.2518367767333984,
"learning_rate": 4.946363326218074e-05,
"loss": 1.7399,
"step": 3441
},
{
"epoch": 0.14647934860566172,
"grad_norm": 1.217373251914978,
"learning_rate": 4.9446192224939525e-05,
"loss": 1.7386,
"step": 3472
},
{
"epoch": 0.147787199932498,
"grad_norm": 1.2229036092758179,
"learning_rate": 4.942847531574167e-05,
"loss": 1.7096,
"step": 3503
},
{
"epoch": 0.14909505125933425,
"grad_norm": 1.2012135982513428,
"learning_rate": 4.941048273452008e-05,
"loss": 1.7195,
"step": 3534
},
{
"epoch": 0.15040290258617053,
"grad_norm": 1.2632161378860474,
"learning_rate": 4.9392214684318605e-05,
"loss": 1.7377,
"step": 3565
},
{
"epoch": 0.15171075391300679,
"grad_norm": 1.2119094133377075,
"learning_rate": 4.93736713712897e-05,
"loss": 1.7276,
"step": 3596
},
{
"epoch": 0.15301860523984306,
"grad_norm": 1.2516822814941406,
"learning_rate": 4.9354853004692124e-05,
"loss": 1.7224,
"step": 3627
},
{
"epoch": 0.15432645656667932,
"grad_norm": 1.2437561750411987,
"learning_rate": 4.93357597968886e-05,
"loss": 1.703,
"step": 3658
},
{
"epoch": 0.1556343078935156,
"grad_norm": 1.4856292009353638,
"learning_rate": 4.931639196334338e-05,
"loss": 1.7139,
"step": 3689
},
{
"epoch": 0.15694215922035185,
"grad_norm": 1.202919840812683,
"learning_rate": 4.9296749722619826e-05,
"loss": 1.7081,
"step": 3720
},
{
"epoch": 0.15825001054718812,
"grad_norm": 1.1720086336135864,
"learning_rate": 4.9276833296377966e-05,
"loss": 1.7086,
"step": 3751
},
{
"epoch": 0.15955786187402438,
"grad_norm": 1.2187249660491943,
"learning_rate": 4.925664290937196e-05,
"loss": 1.6871,
"step": 3782
},
{
"epoch": 0.16086571320086065,
"grad_norm": 1.1918401718139648,
"learning_rate": 4.9236178789447576e-05,
"loss": 1.7058,
"step": 3813
},
{
"epoch": 0.1621735645276969,
"grad_norm": 1.1956909894943237,
"learning_rate": 4.921544116753962e-05,
"loss": 1.6986,
"step": 3844
},
{
"epoch": 0.16348141585453319,
"grad_norm": 1.1646822690963745,
"learning_rate": 4.919443027766935e-05,
"loss": 1.6797,
"step": 3875
},
{
"epoch": 0.16478926718136944,
"grad_norm": 1.2145137786865234,
"learning_rate": 4.91731463569418e-05,
"loss": 1.678,
"step": 3906
},
{
"epoch": 0.16609711850820572,
"grad_norm": 1.2347453832626343,
"learning_rate": 4.915158964554312e-05,
"loss": 1.672,
"step": 3937
},
{
"epoch": 0.16740496983504197,
"grad_norm": 1.1764047145843506,
"learning_rate": 4.912976038673786e-05,
"loss": 1.6864,
"step": 3968
},
{
"epoch": 0.16871282116187825,
"grad_norm": 1.1730256080627441,
"learning_rate": 4.9107658826866254e-05,
"loss": 1.7021,
"step": 3999
},
{
"epoch": 0.1700206724887145,
"grad_norm": 1.1880205869674683,
"learning_rate": 4.908528521534139e-05,
"loss": 1.6832,
"step": 4030
},
{
"epoch": 0.17132852381555078,
"grad_norm": 1.1810024976730347,
"learning_rate": 4.906263980464644e-05,
"loss": 1.6674,
"step": 4061
},
{
"epoch": 0.17263637514238703,
"grad_norm": 1.2133933305740356,
"learning_rate": 4.903972285033178e-05,
"loss": 1.6887,
"step": 4092
},
{
"epoch": 0.1739442264692233,
"grad_norm": 1.1253775358200073,
"learning_rate": 4.901653461101213e-05,
"loss": 1.6884,
"step": 4123
},
{
"epoch": 0.17525207779605956,
"grad_norm": 1.215759038925171,
"learning_rate": 4.8993075348363626e-05,
"loss": 1.7013,
"step": 4154
},
{
"epoch": 0.17655992912289584,
"grad_norm": 1.1305285692214966,
"learning_rate": 4.896934532712084e-05,
"loss": 1.6573,
"step": 4185
},
{
"epoch": 0.1778677804497321,
"grad_norm": 1.127212405204773,
"learning_rate": 4.8945344815073846e-05,
"loss": 1.6769,
"step": 4216
},
{
"epoch": 0.17917563177656837,
"grad_norm": 1.1379367113113403,
"learning_rate": 4.892107408306516e-05,
"loss": 1.6583,
"step": 4247
},
{
"epoch": 0.18048348310340465,
"grad_norm": 1.0926482677459717,
"learning_rate": 4.889653340498669e-05,
"loss": 1.6674,
"step": 4278
},
{
"epoch": 0.1817913344302409,
"grad_norm": 1.1098601818084717,
"learning_rate": 4.8871723057776664e-05,
"loss": 1.6598,
"step": 4309
},
{
"epoch": 0.18309918575707718,
"grad_norm": 1.129784345626831,
"learning_rate": 4.8846643321416476e-05,
"loss": 1.6786,
"step": 4340
},
{
"epoch": 0.18440703708391343,
"grad_norm": 1.1000723838806152,
"learning_rate": 4.882129447892753e-05,
"loss": 1.6623,
"step": 4371
},
{
"epoch": 0.1857148884107497,
"grad_norm": 1.2229663133621216,
"learning_rate": 4.8795676816368076e-05,
"loss": 1.6675,
"step": 4402
},
{
"epoch": 0.18702273973758596,
"grad_norm": 1.1608648300170898,
"learning_rate": 4.876979062282995e-05,
"loss": 1.6499,
"step": 4433
},
{
"epoch": 0.18833059106442224,
"grad_norm": 1.1060220003128052,
"learning_rate": 4.8743636190435325e-05,
"loss": 1.6453,
"step": 4464
},
{
"epoch": 0.1896384423912585,
"grad_norm": 1.1028501987457275,
"learning_rate": 4.871721381433344e-05,
"loss": 1.6393,
"step": 4495
},
{
"epoch": 0.19094629371809477,
"grad_norm": 1.1736187934875488,
"learning_rate": 4.869052379269719e-05,
"loss": 1.6563,
"step": 4526
},
{
"epoch": 0.19225414504493102,
"grad_norm": 1.1284164190292358,
"learning_rate": 4.866356642671985e-05,
"loss": 1.6535,
"step": 4557
},
{
"epoch": 0.1935619963717673,
"grad_norm": 1.1155307292938232,
"learning_rate": 4.8636342020611634e-05,
"loss": 1.6417,
"step": 4588
},
{
"epoch": 0.19486984769860355,
"grad_norm": 1.1379483938217163,
"learning_rate": 4.860885088159626e-05,
"loss": 1.6432,
"step": 4619
},
{
"epoch": 0.19617769902543983,
"grad_norm": 1.1166200637817383,
"learning_rate": 4.858109331990751e-05,
"loss": 1.6545,
"step": 4650
},
{
"epoch": 0.19748555035227608,
"grad_norm": 1.081626296043396,
"learning_rate": 4.855306964878567e-05,
"loss": 1.6402,
"step": 4681
},
{
"epoch": 0.19879340167911236,
"grad_norm": 1.0875532627105713,
"learning_rate": 4.8524780184474084e-05,
"loss": 1.6498,
"step": 4712
},
{
"epoch": 0.2001012530059486,
"grad_norm": 1.0928492546081543,
"learning_rate": 4.8496225246215496e-05,
"loss": 1.6341,
"step": 4743
},
{
"epoch": 0.2014091043327849,
"grad_norm": 1.1116279363632202,
"learning_rate": 4.8467405156248505e-05,
"loss": 1.6389,
"step": 4774
},
{
"epoch": 0.20271695565962114,
"grad_norm": 1.0984513759613037,
"learning_rate": 4.843832023980392e-05,
"loss": 1.6277,
"step": 4805
},
{
"epoch": 0.20402480698645742,
"grad_norm": 1.09713876247406,
"learning_rate": 4.840897082510106e-05,
"loss": 1.6244,
"step": 4836
},
{
"epoch": 0.20533265831329367,
"grad_norm": 1.089237928390503,
"learning_rate": 4.8379357243344084e-05,
"loss": 1.6136,
"step": 4867
},
{
"epoch": 0.20664050964012995,
"grad_norm": 1.1337491273880005,
"learning_rate": 4.8349479828718236e-05,
"loss": 1.641,
"step": 4898
},
{
"epoch": 0.2079483609669662,
"grad_norm": 1.1337246894836426,
"learning_rate": 4.8319338918386075e-05,
"loss": 1.6163,
"step": 4929
},
{
"epoch": 0.20925621229380248,
"grad_norm": 1.112273097038269,
"learning_rate": 4.828893485248369e-05,
"loss": 1.6384,
"step": 4960
},
{
"epoch": 0.21056406362063873,
"grad_norm": 1.049856424331665,
"learning_rate": 4.825826797411682e-05,
"loss": 1.6245,
"step": 4991
},
{
"epoch": 0.211871914947475,
"grad_norm": 2.468637704849243,
"learning_rate": 4.822733862935702e-05,
"loss": 1.6165,
"step": 5022
},
{
"epoch": 0.21317976627431126,
"grad_norm": 1.0854454040527344,
"learning_rate": 4.819614716723775e-05,
"loss": 1.6252,
"step": 5053
},
{
"epoch": 0.21448761760114754,
"grad_norm": 1.0815082788467407,
"learning_rate": 4.8164693939750425e-05,
"loss": 1.6096,
"step": 5084
},
{
"epoch": 0.2157954689279838,
"grad_norm": 1.053945541381836,
"learning_rate": 4.813297930184042e-05,
"loss": 1.6296,
"step": 5115
},
{
"epoch": 0.21710332025482007,
"grad_norm": 1.06475031375885,
"learning_rate": 4.810100361140314e-05,
"loss": 1.629,
"step": 5146
},
{
"epoch": 0.21841117158165632,
"grad_norm": 1.1072969436645508,
"learning_rate": 4.8068767229279885e-05,
"loss": 1.6244,
"step": 5177
},
{
"epoch": 0.2197190229084926,
"grad_norm": 1.0536466836929321,
"learning_rate": 4.8036270519253854e-05,
"loss": 1.6474,
"step": 5208
},
{
"epoch": 0.22102687423532885,
"grad_norm": 1.1039215326309204,
"learning_rate": 4.8003513848046e-05,
"loss": 1.6354,
"step": 5239
},
{
"epoch": 0.22233472556216513,
"grad_norm": 1.0463244915008545,
"learning_rate": 4.79704975853109e-05,
"loss": 1.6154,
"step": 5270
},
{
"epoch": 0.22364257688900138,
"grad_norm": 1.0931519269943237,
"learning_rate": 4.793722210363262e-05,
"loss": 1.6037,
"step": 5301
},
{
"epoch": 0.22495042821583766,
"grad_norm": 1.0838814973831177,
"learning_rate": 4.7903687778520414e-05,
"loss": 1.6352,
"step": 5332
},
{
"epoch": 0.2262582795426739,
"grad_norm": 1.0400842428207397,
"learning_rate": 4.7869894988404593e-05,
"loss": 1.5924,
"step": 5363
},
{
"epoch": 0.2275661308695102,
"grad_norm": 1.0707345008850098,
"learning_rate": 4.783584411463221e-05,
"loss": 1.6211,
"step": 5394
},
{
"epoch": 0.22887398219634644,
"grad_norm": 1.0422611236572266,
"learning_rate": 4.780153554146274e-05,
"loss": 1.5969,
"step": 5425
},
{
"epoch": 0.23018183352318272,
"grad_norm": 1.0401543378829956,
"learning_rate": 4.7766969656063766e-05,
"loss": 1.5974,
"step": 5456
},
{
"epoch": 0.23148968485001897,
"grad_norm": 1.0429288148880005,
"learning_rate": 4.773214684850662e-05,
"loss": 1.6043,
"step": 5487
},
{
"epoch": 0.23279753617685525,
"grad_norm": 1.098827600479126,
"learning_rate": 4.769706751176193e-05,
"loss": 1.5984,
"step": 5518
},
{
"epoch": 0.2341053875036915,
"grad_norm": 1.0090724229812622,
"learning_rate": 4.7661732041695264e-05,
"loss": 1.6012,
"step": 5549
},
{
"epoch": 0.23541323883052778,
"grad_norm": 1.0732818841934204,
"learning_rate": 4.762614083706258e-05,
"loss": 1.5996,
"step": 5580
},
{
"epoch": 0.23672109015736403,
"grad_norm": 1.082144856452942,
"learning_rate": 4.759029429950581e-05,
"loss": 1.6127,
"step": 5611
},
{
"epoch": 0.2380289414842003,
"grad_norm": 1.0642670392990112,
"learning_rate": 4.7554192833548235e-05,
"loss": 1.5962,
"step": 5642
},
{
"epoch": 0.23933679281103656,
"grad_norm": 1.1351134777069092,
"learning_rate": 4.751783684659e-05,
"loss": 1.5983,
"step": 5673
},
{
"epoch": 0.24064464413787284,
"grad_norm": 1.09076726436615,
"learning_rate": 4.748122674890348e-05,
"loss": 1.6162,
"step": 5704
},
{
"epoch": 0.24195249546470912,
"grad_norm": 1.0234689712524414,
"learning_rate": 4.7444362953628654e-05,
"loss": 1.596,
"step": 5735
},
{
"epoch": 0.24326034679154537,
"grad_norm": 1.1209464073181152,
"learning_rate": 4.7407245876768424e-05,
"loss": 1.602,
"step": 5766
},
{
"epoch": 0.24456819811838165,
"grad_norm": 1.0940710306167603,
"learning_rate": 4.736987593718397e-05,
"loss": 1.5992,
"step": 5797
},
{
"epoch": 0.2458760494452179,
"grad_norm": 1.0460273027420044,
"learning_rate": 4.733225355658999e-05,
"loss": 1.5881,
"step": 5828
},
{
"epoch": 0.24718390077205418,
"grad_norm": 1.0842856168746948,
"learning_rate": 4.7294379159549926e-05,
"loss": 1.594,
"step": 5859
},
{
"epoch": 0.24849175209889043,
"grad_norm": 1.0244876146316528,
"learning_rate": 4.725625317347119e-05,
"loss": 1.583,
"step": 5890
},
{
"epoch": 0.2497996034257267,
"grad_norm": 1.0182656049728394,
"learning_rate": 4.7217876028600374e-05,
"loss": 1.5839,
"step": 5921
},
{
"epoch": 0.25110745475256296,
"grad_norm": 1.0451878309249878,
"learning_rate": 4.717924815801832e-05,
"loss": 1.6051,
"step": 5952
},
{
"epoch": 0.2524153060793992,
"grad_norm": 1.0340505838394165,
"learning_rate": 4.714036999763532e-05,
"loss": 1.5969,
"step": 5983
},
{
"epoch": 0.2537231574062355,
"grad_norm": 1.0457735061645508,
"learning_rate": 4.7101241986186116e-05,
"loss": 1.5971,
"step": 6014
},
{
"epoch": 0.2550310087330718,
"grad_norm": 1.034166693687439,
"learning_rate": 4.7061864565225e-05,
"loss": 1.5982,
"step": 6045
},
{
"epoch": 0.256338860059908,
"grad_norm": 1.0326176881790161,
"learning_rate": 4.702223817912081e-05,
"loss": 1.5691,
"step": 6076
},
{
"epoch": 0.2576467113867443,
"grad_norm": 1.0339311361312866,
"learning_rate": 4.698236327505195e-05,
"loss": 1.5726,
"step": 6107
},
{
"epoch": 0.2589545627135806,
"grad_norm": 1.022299885749817,
"learning_rate": 4.694224030300127e-05,
"loss": 1.5935,
"step": 6138
},
{
"epoch": 0.26026241404041683,
"grad_norm": 1.009255051612854,
"learning_rate": 4.690186971575107e-05,
"loss": 1.574,
"step": 6169
},
{
"epoch": 0.2615702653672531,
"grad_norm": 1.0181437730789185,
"learning_rate": 4.6861251968877916e-05,
"loss": 1.57,
"step": 6200
},
{
"epoch": 0.26287811669408934,
"grad_norm": 1.0302594900131226,
"learning_rate": 4.68203875207476e-05,
"loss": 1.5785,
"step": 6231
},
{
"epoch": 0.26418596802092564,
"grad_norm": 1.006404161453247,
"learning_rate": 4.677927683250983e-05,
"loss": 1.5798,
"step": 6262
},
{
"epoch": 0.2654938193477619,
"grad_norm": 0.993584394454956,
"learning_rate": 4.6737920368093156e-05,
"loss": 1.5558,
"step": 6293
},
{
"epoch": 0.26680167067459815,
"grad_norm": 0.9635722637176514,
"learning_rate": 4.669631859419965e-05,
"loss": 1.5716,
"step": 6324
},
{
"epoch": 0.2681095220014344,
"grad_norm": 0.9815465211868286,
"learning_rate": 4.6654471980299676e-05,
"loss": 1.5472,
"step": 6355
},
{
"epoch": 0.2694173733282707,
"grad_norm": 0.9921912550926208,
"learning_rate": 4.661238099862658e-05,
"loss": 1.5786,
"step": 6386
},
{
"epoch": 0.27072522465510696,
"grad_norm": 1.0193076133728027,
"learning_rate": 4.657004612417138e-05,
"loss": 1.5672,
"step": 6417
},
{
"epoch": 0.2720330759819432,
"grad_norm": 0.9979498982429504,
"learning_rate": 4.6527467834677374e-05,
"loss": 1.5549,
"step": 6448
},
{
"epoch": 0.27334092730877946,
"grad_norm": 1.0342402458190918,
"learning_rate": 4.648464661063478e-05,
"loss": 1.5808,
"step": 6479
},
{
"epoch": 0.27464877863561576,
"grad_norm": 0.9821639060974121,
"learning_rate": 4.6441582935275264e-05,
"loss": 1.5642,
"step": 6510
},
{
"epoch": 0.275956629962452,
"grad_norm": 1.002773404121399,
"learning_rate": 4.6398277294566586e-05,
"loss": 1.5646,
"step": 6541
},
{
"epoch": 0.27726448128928827,
"grad_norm": 1.0241411924362183,
"learning_rate": 4.6354730177207e-05,
"loss": 1.5744,
"step": 6572
},
{
"epoch": 0.2785723326161245,
"grad_norm": 1.0101004838943481,
"learning_rate": 4.6310942074619787e-05,
"loss": 1.5905,
"step": 6603
},
{
"epoch": 0.2798801839429608,
"grad_norm": 1.0355374813079834,
"learning_rate": 4.626691348094777e-05,
"loss": 1.5713,
"step": 6634
},
{
"epoch": 0.2811880352697971,
"grad_norm": 1.0100233554840088,
"learning_rate": 4.622264489304762e-05,
"loss": 1.5491,
"step": 6665
},
{
"epoch": 0.2824958865966333,
"grad_norm": 1.0172356367111206,
"learning_rate": 4.617813681048434e-05,
"loss": 1.562,
"step": 6696
},
{
"epoch": 0.2838037379234696,
"grad_norm": 0.994799017906189,
"learning_rate": 4.61333897355256e-05,
"loss": 1.5526,
"step": 6727
},
{
"epoch": 0.2851115892503059,
"grad_norm": 1.1320250034332275,
"learning_rate": 4.608840417313604e-05,
"loss": 1.5616,
"step": 6758
},
{
"epoch": 0.28641944057714214,
"grad_norm": 1.013725996017456,
"learning_rate": 4.6043180630971646e-05,
"loss": 1.5345,
"step": 6789
},
{
"epoch": 0.2877272919039784,
"grad_norm": 1.045462727546692,
"learning_rate": 4.599771961937391e-05,
"loss": 1.56,
"step": 6820
},
{
"epoch": 0.28903514323081464,
"grad_norm": 1.0075098276138306,
"learning_rate": 4.5952021651364204e-05,
"loss": 1.5825,
"step": 6851
},
{
"epoch": 0.29034299455765095,
"grad_norm": 0.9952128529548645,
"learning_rate": 4.590608724263786e-05,
"loss": 1.5581,
"step": 6882
},
{
"epoch": 0.2916508458844872,
"grad_norm": 0.9492574334144592,
"learning_rate": 4.585991691155845e-05,
"loss": 1.554,
"step": 6913
},
{
"epoch": 0.29295869721132345,
"grad_norm": 0.9850384593009949,
"learning_rate": 4.581351117915188e-05,
"loss": 1.5614,
"step": 6944
},
{
"epoch": 0.2942665485381597,
"grad_norm": 0.9840113520622253,
"learning_rate": 4.5766870569100534e-05,
"loss": 1.5631,
"step": 6975
},
{
"epoch": 0.295574399864996,
"grad_norm": 0.9830542802810669,
"learning_rate": 4.571999560773736e-05,
"loss": 1.5597,
"step": 7006
},
{
"epoch": 0.29688225119183226,
"grad_norm": 0.9737412929534912,
"learning_rate": 4.5672886824039915e-05,
"loss": 1.5412,
"step": 7037
},
{
"epoch": 0.2981901025186685,
"grad_norm": 1.0272278785705566,
"learning_rate": 4.5625544749624435e-05,
"loss": 1.5724,
"step": 7068
},
{
"epoch": 0.29949795384550476,
"grad_norm": 0.973682165145874,
"learning_rate": 4.5577969918739794e-05,
"loss": 1.5739,
"step": 7099
},
{
"epoch": 0.30080580517234107,
"grad_norm": 0.9709253907203674,
"learning_rate": 4.5530162868261486e-05,
"loss": 1.5666,
"step": 7130
},
{
"epoch": 0.3021136564991773,
"grad_norm": 0.9426018595695496,
"learning_rate": 4.548212413768558e-05,
"loss": 1.5542,
"step": 7161
},
{
"epoch": 0.30342150782601357,
"grad_norm": 1.026181697845459,
"learning_rate": 4.543385426912261e-05,
"loss": 1.5235,
"step": 7192
},
{
"epoch": 0.3047293591528499,
"grad_norm": 0.9783868193626404,
"learning_rate": 4.53853538072915e-05,
"loss": 1.5591,
"step": 7223
},
{
"epoch": 0.30603721047968613,
"grad_norm": 0.9816028475761414,
"learning_rate": 4.533662329951336e-05,
"loss": 1.5484,
"step": 7254
},
{
"epoch": 0.3073450618065224,
"grad_norm": 0.9516949653625488,
"learning_rate": 4.528766329570536e-05,
"loss": 1.5251,
"step": 7285
},
{
"epoch": 0.30865291313335863,
"grad_norm": 0.9601557850837708,
"learning_rate": 4.523847434837447e-05,
"loss": 1.5668,
"step": 7316
},
{
"epoch": 0.30996076446019494,
"grad_norm": 0.9607630372047424,
"learning_rate": 4.518905701261128e-05,
"loss": 1.5576,
"step": 7347
},
{
"epoch": 0.3112686157870312,
"grad_norm": 0.9606896638870239,
"learning_rate": 4.5139411846083715e-05,
"loss": 1.5441,
"step": 7378
},
{
"epoch": 0.31257646711386744,
"grad_norm": 1.0111651420593262,
"learning_rate": 4.508953940903073e-05,
"loss": 1.5586,
"step": 7409
},
{
"epoch": 0.3138843184407037,
"grad_norm": 0.9388339519500732,
"learning_rate": 4.5039440264255994e-05,
"loss": 1.5472,
"step": 7440
},
{
"epoch": 0.31519216976754,
"grad_norm": 0.9507540464401245,
"learning_rate": 4.498911497712155e-05,
"loss": 1.5579,
"step": 7471
},
{
"epoch": 0.31650002109437625,
"grad_norm": 0.9572455286979675,
"learning_rate": 4.493856411554142e-05,
"loss": 1.5433,
"step": 7502
},
{
"epoch": 0.3178078724212125,
"grad_norm": 0.9485601782798767,
"learning_rate": 4.4887788249975206e-05,
"loss": 1.5513,
"step": 7533
},
{
"epoch": 0.31911572374804875,
"grad_norm": 0.9995166063308716,
"learning_rate": 4.4836787953421656e-05,
"loss": 1.5289,
"step": 7564
},
{
"epoch": 0.32042357507488506,
"grad_norm": 1.1004894971847534,
"learning_rate": 4.478556380141218e-05,
"loss": 1.5306,
"step": 7595
},
{
"epoch": 0.3217314264017213,
"grad_norm": 0.9642030000686646,
"learning_rate": 4.4734116372004375e-05,
"loss": 1.5497,
"step": 7626
},
{
"epoch": 0.32303927772855756,
"grad_norm": 0.9486343860626221,
"learning_rate": 4.4682446245775477e-05,
"loss": 1.5458,
"step": 7657
},
{
"epoch": 0.3243471290553938,
"grad_norm": 0.942319929599762,
"learning_rate": 4.463055400581586e-05,
"loss": 1.5306,
"step": 7688
},
{
"epoch": 0.3256549803822301,
"grad_norm": 0.9433887004852295,
"learning_rate": 4.4578440237722374e-05,
"loss": 1.5394,
"step": 7719
},
{
"epoch": 0.32696283170906637,
"grad_norm": 0.9954056739807129,
"learning_rate": 4.452610552959183e-05,
"loss": 1.551,
"step": 7750
},
{
"epoch": 0.3282706830359026,
"grad_norm": 0.9406622052192688,
"learning_rate": 4.447355047201428e-05,
"loss": 1.5481,
"step": 7781
},
{
"epoch": 0.3295785343627389,
"grad_norm": 0.9743403792381287,
"learning_rate": 4.4420775658066414e-05,
"loss": 1.533,
"step": 7812
},
{
"epoch": 0.3308863856895752,
"grad_norm": 0.9561812877655029,
"learning_rate": 4.436778168330484e-05,
"loss": 1.5615,
"step": 7843
},
{
"epoch": 0.33219423701641143,
"grad_norm": 0.9623595476150513,
"learning_rate": 4.4314569145759353e-05,
"loss": 1.5463,
"step": 7874
},
{
"epoch": 0.3335020883432477,
"grad_norm": 0.9475456476211548,
"learning_rate": 4.42611386459262e-05,
"loss": 1.558,
"step": 7905
},
{
"epoch": 0.33480993967008393,
"grad_norm": 0.9283867478370667,
"learning_rate": 4.420749078676133e-05,
"loss": 1.5237,
"step": 7936
},
{
"epoch": 0.33611779099692024,
"grad_norm": 0.9494192004203796,
"learning_rate": 4.4153626173673516e-05,
"loss": 1.5315,
"step": 7967
},
{
"epoch": 0.3374256423237565,
"grad_norm": 0.9378790855407715,
"learning_rate": 4.409954541451762e-05,
"loss": 1.5378,
"step": 7998
},
{
"epoch": 0.33873349365059274,
"grad_norm": 0.9318397045135498,
"learning_rate": 4.404524911958764e-05,
"loss": 1.5199,
"step": 8029
},
{
"epoch": 0.340041344977429,
"grad_norm": 0.9440187215805054,
"learning_rate": 4.399073790160989e-05,
"loss": 1.5313,
"step": 8060
},
{
"epoch": 0.3413491963042653,
"grad_norm": 0.9643428921699524,
"learning_rate": 4.393601237573607e-05,
"loss": 1.5267,
"step": 8091
},
{
"epoch": 0.34265704763110155,
"grad_norm": 0.9626632928848267,
"learning_rate": 4.388107315953628e-05,
"loss": 1.5312,
"step": 8122
},
{
"epoch": 0.3439648989579378,
"grad_norm": 0.9867373704910278,
"learning_rate": 4.382592087299212e-05,
"loss": 1.5162,
"step": 8153
},
{
"epoch": 0.34527275028477405,
"grad_norm": 0.9584757685661316,
"learning_rate": 4.377055613848964e-05,
"loss": 1.5201,
"step": 8184
},
{
"epoch": 0.34658060161161036,
"grad_norm": 0.9587807655334473,
"learning_rate": 4.3714979580812355e-05,
"loss": 1.55,
"step": 8215
},
{
"epoch": 0.3478884529384466,
"grad_norm": 0.948785662651062,
"learning_rate": 4.365919182713416e-05,
"loss": 1.5392,
"step": 8246
},
{
"epoch": 0.34919630426528286,
"grad_norm": 0.9717795252799988,
"learning_rate": 4.360319350701226e-05,
"loss": 1.5112,
"step": 8277
},
{
"epoch": 0.3505041555921191,
"grad_norm": 0.9338726997375488,
"learning_rate": 4.3546985252380115e-05,
"loss": 1.5371,
"step": 8308
},
{
"epoch": 0.3518120069189554,
"grad_norm": 0.9878547191619873,
"learning_rate": 4.349056769754021e-05,
"loss": 1.5349,
"step": 8339
},
{
"epoch": 0.3531198582457917,
"grad_norm": 1.0211338996887207,
"learning_rate": 4.3433941479156994e-05,
"loss": 1.5234,
"step": 8370
},
{
"epoch": 0.3544277095726279,
"grad_norm": 0.944474458694458,
"learning_rate": 4.3377107236249647e-05,
"loss": 1.5306,
"step": 8401
},
{
"epoch": 0.3557355608994642,
"grad_norm": 0.9218102097511292,
"learning_rate": 4.332006561018488e-05,
"loss": 1.5113,
"step": 8432
},
{
"epoch": 0.3570434122263005,
"grad_norm": 0.9429653286933899,
"learning_rate": 4.3262817244669683e-05,
"loss": 1.5095,
"step": 8463
},
{
"epoch": 0.35835126355313673,
"grad_norm": 0.8979545831680298,
"learning_rate": 4.3205362785744083e-05,
"loss": 1.522,
"step": 8494
},
{
"epoch": 0.359659114879973,
"grad_norm": 0.9040619134902954,
"learning_rate": 4.314770288177384e-05,
"loss": 1.5197,
"step": 8525
},
{
"epoch": 0.3609669662068093,
"grad_norm": 0.9038030505180359,
"learning_rate": 4.308983818344313e-05,
"loss": 1.5218,
"step": 8556
},
{
"epoch": 0.36227481753364554,
"grad_norm": 0.9493038654327393,
"learning_rate": 4.3031769343747206e-05,
"loss": 1.5306,
"step": 8587
},
{
"epoch": 0.3635826688604818,
"grad_norm": 0.9084991812705994,
"learning_rate": 4.297349701798505e-05,
"loss": 1.5146,
"step": 8618
},
{
"epoch": 0.36489052018731805,
"grad_norm": 0.9597145915031433,
"learning_rate": 4.2915021863751916e-05,
"loss": 1.5202,
"step": 8649
},
{
"epoch": 0.36619837151415435,
"grad_norm": 0.9139339923858643,
"learning_rate": 4.285634454093198e-05,
"loss": 1.5082,
"step": 8680
},
{
"epoch": 0.3675062228409906,
"grad_norm": 0.955664336681366,
"learning_rate": 4.279746571169086e-05,
"loss": 1.5073,
"step": 8711
},
{
"epoch": 0.36881407416782686,
"grad_norm": 0.975068986415863,
"learning_rate": 4.2738386040468136e-05,
"loss": 1.5086,
"step": 8742
},
{
"epoch": 0.3701219254946631,
"grad_norm": 0.9280393719673157,
"learning_rate": 4.2679106193969866e-05,
"loss": 1.5342,
"step": 8773
},
{
"epoch": 0.3714297768214994,
"grad_norm": 0.9098469614982605,
"learning_rate": 4.261962684116106e-05,
"loss": 1.5305,
"step": 8804
},
{
"epoch": 0.37273762814833566,
"grad_norm": 0.9359835386276245,
"learning_rate": 4.2559948653258145e-05,
"loss": 1.5172,
"step": 8835
},
{
"epoch": 0.3740454794751719,
"grad_norm": 0.9431541562080383,
"learning_rate": 4.250007230372134e-05,
"loss": 1.5128,
"step": 8866
},
{
"epoch": 0.37535333080200817,
"grad_norm": 0.8999470472335815,
"learning_rate": 4.2439998468247126e-05,
"loss": 1.5025,
"step": 8897
},
{
"epoch": 0.3766611821288445,
"grad_norm": 0.9414424300193787,
"learning_rate": 4.2379727824760566e-05,
"loss": 1.5135,
"step": 8928
},
{
"epoch": 0.3779690334556807,
"grad_norm": 0.9347614049911499,
"learning_rate": 4.231926105340768e-05,
"loss": 1.5252,
"step": 8959
},
{
"epoch": 0.379276884782517,
"grad_norm": 0.9166226983070374,
"learning_rate": 4.225859883654776e-05,
"loss": 1.519,
"step": 8990
},
{
"epoch": 0.3805847361093532,
"grad_norm": 0.9614241719245911,
"learning_rate": 4.219774185874569e-05,
"loss": 1.494,
"step": 9021
},
{
"epoch": 0.38189258743618953,
"grad_norm": 0.9002947211265564,
"learning_rate": 4.213669080676418e-05,
"loss": 1.5093,
"step": 9052
},
{
"epoch": 0.3832004387630258,
"grad_norm": 0.8749356269836426,
"learning_rate": 4.2075446369556056e-05,
"loss": 1.512,
"step": 9083
},
{
"epoch": 0.38450829008986204,
"grad_norm": 0.954454243183136,
"learning_rate": 4.201400923825648e-05,
"loss": 1.4987,
"step": 9114
},
{
"epoch": 0.3858161414166983,
"grad_norm": 0.9184893369674683,
"learning_rate": 4.195238010617511e-05,
"loss": 1.5195,
"step": 9145
}
],
"logging_steps": 31,
"max_steps": 30517,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"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": 6.395056258591752e+18,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}