{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 2.0, "eval_steps": 500, "global_step": 3738, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.005350454788657036, "grad_norm": 14.125, "learning_rate": 1.9951845906902087e-05, "loss": 0.8081, "mean_token_accuracy": 0.7898837089538574, "num_tokens": 6502.0, "step": 10 }, { "epoch": 0.010700909577314071, "grad_norm": 15.5625, "learning_rate": 1.989834135901552e-05, "loss": 0.6171, "mean_token_accuracy": 0.8267314732074738, "num_tokens": 12955.0, "step": 20 }, { "epoch": 0.016051364365971106, "grad_norm": 11.3125, "learning_rate": 1.9844836811128947e-05, "loss": 0.6067, "mean_token_accuracy": 0.8393934667110443, "num_tokens": 20188.0, "step": 30 }, { "epoch": 0.021401819154628143, "grad_norm": 16.625, "learning_rate": 1.979133226324238e-05, "loss": 0.629, "mean_token_accuracy": 0.8308047413825989, "num_tokens": 26883.0, "step": 40 }, { "epoch": 0.02675227394328518, "grad_norm": 9.25, "learning_rate": 1.9737827715355807e-05, "loss": 0.665, "mean_token_accuracy": 0.8247407853603363, "num_tokens": 32972.0, "step": 50 }, { "epoch": 0.03210272873194221, "grad_norm": 12.6875, "learning_rate": 1.9684323167469236e-05, "loss": 0.6403, "mean_token_accuracy": 0.8298830628395081, "num_tokens": 39159.0, "step": 60 }, { "epoch": 0.03745318352059925, "grad_norm": 10.9375, "learning_rate": 1.9630818619582667e-05, "loss": 0.5836, "mean_token_accuracy": 0.8472412645816803, "num_tokens": 45393.0, "step": 70 }, { "epoch": 0.042803638309256285, "grad_norm": 10.125, "learning_rate": 1.9577314071696096e-05, "loss": 0.5471, "mean_token_accuracy": 0.8438372075557709, "num_tokens": 51641.0, "step": 80 }, { "epoch": 0.048154093097913325, "grad_norm": 9.75, "learning_rate": 1.9523809523809524e-05, "loss": 0.5661, "mean_token_accuracy": 0.839038759469986, "num_tokens": 58507.0, "step": 90 }, { "epoch": 0.05350454788657036, "grad_norm": 11.4375, "learning_rate": 1.9470304975922956e-05, "loss": 0.5204, "mean_token_accuracy": 0.8511396527290345, "num_tokens": 64759.0, "step": 100 }, { "epoch": 0.05885500267522739, "grad_norm": 13.0, "learning_rate": 1.9416800428036384e-05, "loss": 0.5552, "mean_token_accuracy": 0.845803028345108, "num_tokens": 71240.0, "step": 110 }, { "epoch": 0.06420545746388442, "grad_norm": 12.25, "learning_rate": 1.9363295880149816e-05, "loss": 0.5781, "mean_token_accuracy": 0.8335193932056427, "num_tokens": 77470.0, "step": 120 }, { "epoch": 0.06955591225254147, "grad_norm": 13.8125, "learning_rate": 1.9309791332263244e-05, "loss": 0.6074, "mean_token_accuracy": 0.8274448573589325, "num_tokens": 84887.0, "step": 130 }, { "epoch": 0.0749063670411985, "grad_norm": 10.0, "learning_rate": 1.9256286784376676e-05, "loss": 0.5014, "mean_token_accuracy": 0.8621269166469574, "num_tokens": 90792.0, "step": 140 }, { "epoch": 0.08025682182985554, "grad_norm": 10.8125, "learning_rate": 1.9202782236490104e-05, "loss": 0.534, "mean_token_accuracy": 0.8480484664440155, "num_tokens": 96995.0, "step": 150 }, { "epoch": 0.08560727661851257, "grad_norm": 14.3125, "learning_rate": 1.9149277688603532e-05, "loss": 0.5377, "mean_token_accuracy": 0.8530115425586701, "num_tokens": 103449.0, "step": 160 }, { "epoch": 0.0909577314071696, "grad_norm": 12.625, "learning_rate": 1.909577314071696e-05, "loss": 0.5344, "mean_token_accuracy": 0.8539405584335327, "num_tokens": 109584.0, "step": 170 }, { "epoch": 0.09630818619582665, "grad_norm": 12.5625, "learning_rate": 1.9042268592830392e-05, "loss": 0.6131, "mean_token_accuracy": 0.8290488123893738, "num_tokens": 116298.0, "step": 180 }, { "epoch": 0.10165864098448368, "grad_norm": 9.3125, "learning_rate": 1.898876404494382e-05, "loss": 0.556, "mean_token_accuracy": 0.8458720922470093, "num_tokens": 123306.0, "step": 190 }, { "epoch": 0.10700909577314072, "grad_norm": 11.125, "learning_rate": 1.8935259497057252e-05, "loss": 0.554, "mean_token_accuracy": 0.836270147562027, "num_tokens": 129919.0, "step": 200 }, { "epoch": 0.11235955056179775, "grad_norm": 9.6875, "learning_rate": 1.888175494917068e-05, "loss": 0.5265, "mean_token_accuracy": 0.8515542447566986, "num_tokens": 135935.0, "step": 210 }, { "epoch": 0.11771000535045478, "grad_norm": 9.9375, "learning_rate": 1.8828250401284112e-05, "loss": 0.6605, "mean_token_accuracy": 0.8285135805606842, "num_tokens": 142921.0, "step": 220 }, { "epoch": 0.12306046013911183, "grad_norm": 14.4375, "learning_rate": 1.877474585339754e-05, "loss": 0.5226, "mean_token_accuracy": 0.8500208795070648, "num_tokens": 149545.0, "step": 230 }, { "epoch": 0.12841091492776885, "grad_norm": 9.6875, "learning_rate": 1.8721241305510972e-05, "loss": 0.596, "mean_token_accuracy": 0.8368849396705628, "num_tokens": 156260.0, "step": 240 }, { "epoch": 0.1337613697164259, "grad_norm": 9.625, "learning_rate": 1.86677367576244e-05, "loss": 0.5068, "mean_token_accuracy": 0.8535505294799804, "num_tokens": 162945.0, "step": 250 }, { "epoch": 0.13911182450508294, "grad_norm": 10.0, "learning_rate": 1.861423220973783e-05, "loss": 0.4634, "mean_token_accuracy": 0.8602178037166596, "num_tokens": 168808.0, "step": 260 }, { "epoch": 0.14446227929373998, "grad_norm": 9.1875, "learning_rate": 1.8560727661851257e-05, "loss": 0.5033, "mean_token_accuracy": 0.8550495684146882, "num_tokens": 174844.0, "step": 270 }, { "epoch": 0.149812734082397, "grad_norm": 12.125, "learning_rate": 1.8507223113964686e-05, "loss": 0.5881, "mean_token_accuracy": 0.8327823579311371, "num_tokens": 180772.0, "step": 280 }, { "epoch": 0.15516318887105404, "grad_norm": 10.3125, "learning_rate": 1.8453718566078117e-05, "loss": 0.5553, "mean_token_accuracy": 0.8426111042499542, "num_tokens": 187267.0, "step": 290 }, { "epoch": 0.16051364365971107, "grad_norm": 13.9375, "learning_rate": 1.8400214018191546e-05, "loss": 0.5659, "mean_token_accuracy": 0.8400906562805176, "num_tokens": 193710.0, "step": 300 }, { "epoch": 0.1658640984483681, "grad_norm": 10.125, "learning_rate": 1.8346709470304977e-05, "loss": 0.4957, "mean_token_accuracy": 0.8593023598194123, "num_tokens": 200857.0, "step": 310 }, { "epoch": 0.17121455323702514, "grad_norm": 9.5625, "learning_rate": 1.8293204922418406e-05, "loss": 0.5723, "mean_token_accuracy": 0.8425621449947357, "num_tokens": 207079.0, "step": 320 }, { "epoch": 0.17656500802568217, "grad_norm": 9.625, "learning_rate": 1.8239700374531838e-05, "loss": 0.5732, "mean_token_accuracy": 0.8408381104469299, "num_tokens": 213687.0, "step": 330 }, { "epoch": 0.1819154628143392, "grad_norm": 14.0, "learning_rate": 1.8186195826645266e-05, "loss": 0.5892, "mean_token_accuracy": 0.8376498997211457, "num_tokens": 219813.0, "step": 340 }, { "epoch": 0.18726591760299627, "grad_norm": 9.75, "learning_rate": 1.8132691278758698e-05, "loss": 0.5856, "mean_token_accuracy": 0.8350161492824555, "num_tokens": 226169.0, "step": 350 }, { "epoch": 0.1926163723916533, "grad_norm": 9.625, "learning_rate": 1.8079186730872126e-05, "loss": 0.5054, "mean_token_accuracy": 0.8569566071033478, "num_tokens": 232745.0, "step": 360 }, { "epoch": 0.19796682718031033, "grad_norm": 9.25, "learning_rate": 1.8025682182985554e-05, "loss": 0.5859, "mean_token_accuracy": 0.8372735023498535, "num_tokens": 239421.0, "step": 370 }, { "epoch": 0.20331728196896737, "grad_norm": 11.125, "learning_rate": 1.7972177635098983e-05, "loss": 0.5413, "mean_token_accuracy": 0.8501188158988953, "num_tokens": 246018.0, "step": 380 }, { "epoch": 0.2086677367576244, "grad_norm": 9.0, "learning_rate": 1.7918673087212414e-05, "loss": 0.5353, "mean_token_accuracy": 0.853459233045578, "num_tokens": 252606.0, "step": 390 }, { "epoch": 0.21401819154628143, "grad_norm": 10.0625, "learning_rate": 1.7865168539325843e-05, "loss": 0.4907, "mean_token_accuracy": 0.8613495647907257, "num_tokens": 258868.0, "step": 400 }, { "epoch": 0.21936864633493847, "grad_norm": 7.6875, "learning_rate": 1.7811663991439274e-05, "loss": 0.5388, "mean_token_accuracy": 0.8516083061695099, "num_tokens": 264952.0, "step": 410 }, { "epoch": 0.2247191011235955, "grad_norm": 9.4375, "learning_rate": 1.7758159443552703e-05, "loss": 0.6146, "mean_token_accuracy": 0.8291812062263488, "num_tokens": 271874.0, "step": 420 }, { "epoch": 0.23006955591225253, "grad_norm": 11.0625, "learning_rate": 1.7704654895666134e-05, "loss": 0.5411, "mean_token_accuracy": 0.8542669951915741, "num_tokens": 278190.0, "step": 430 }, { "epoch": 0.23542001070090957, "grad_norm": 9.8125, "learning_rate": 1.7651150347779563e-05, "loss": 0.5449, "mean_token_accuracy": 0.849567049741745, "num_tokens": 284605.0, "step": 440 }, { "epoch": 0.24077046548956663, "grad_norm": 12.625, "learning_rate": 1.7597645799892994e-05, "loss": 0.4891, "mean_token_accuracy": 0.8632073998451233, "num_tokens": 290661.0, "step": 450 }, { "epoch": 0.24612092027822366, "grad_norm": 13.875, "learning_rate": 1.7544141252006423e-05, "loss": 0.5757, "mean_token_accuracy": 0.8453100442886352, "num_tokens": 296627.0, "step": 460 }, { "epoch": 0.25147137506688066, "grad_norm": 9.0625, "learning_rate": 1.749063670411985e-05, "loss": 0.5463, "mean_token_accuracy": 0.842570561170578, "num_tokens": 303643.0, "step": 470 }, { "epoch": 0.2568218298555377, "grad_norm": 10.5625, "learning_rate": 1.743713215623328e-05, "loss": 0.5236, "mean_token_accuracy": 0.8542482018470764, "num_tokens": 310033.0, "step": 480 }, { "epoch": 0.26217228464419473, "grad_norm": 9.5625, "learning_rate": 1.738362760834671e-05, "loss": 0.5389, "mean_token_accuracy": 0.84858078956604, "num_tokens": 316827.0, "step": 490 }, { "epoch": 0.2675227394328518, "grad_norm": 10.3125, "learning_rate": 1.733012306046014e-05, "loss": 0.5375, "mean_token_accuracy": 0.8425202608108521, "num_tokens": 323621.0, "step": 500 }, { "epoch": 0.27287319422150885, "grad_norm": 14.625, "learning_rate": 1.727661851257357e-05, "loss": 0.5325, "mean_token_accuracy": 0.8490587711334229, "num_tokens": 330247.0, "step": 510 }, { "epoch": 0.2782236490101659, "grad_norm": 9.9375, "learning_rate": 1.7223113964687e-05, "loss": 0.5204, "mean_token_accuracy": 0.8513540923595428, "num_tokens": 337357.0, "step": 520 }, { "epoch": 0.2835741037988229, "grad_norm": 9.5625, "learning_rate": 1.716960941680043e-05, "loss": 0.5946, "mean_token_accuracy": 0.8314260065555572, "num_tokens": 343657.0, "step": 530 }, { "epoch": 0.28892455858747995, "grad_norm": 16.25, "learning_rate": 1.711610486891386e-05, "loss": 0.5195, "mean_token_accuracy": 0.8544752418994903, "num_tokens": 350536.0, "step": 540 }, { "epoch": 0.294275013376137, "grad_norm": 12.6875, "learning_rate": 1.706260032102729e-05, "loss": 0.5459, "mean_token_accuracy": 0.8454052388668061, "num_tokens": 357171.0, "step": 550 }, { "epoch": 0.299625468164794, "grad_norm": 9.9375, "learning_rate": 1.700909577314072e-05, "loss": 0.4781, "mean_token_accuracy": 0.8612410366535187, "num_tokens": 363433.0, "step": 560 }, { "epoch": 0.30497592295345105, "grad_norm": 7.5625, "learning_rate": 1.6955591225254148e-05, "loss": 0.5892, "mean_token_accuracy": 0.8377074897289276, "num_tokens": 369893.0, "step": 570 }, { "epoch": 0.3103263777421081, "grad_norm": 9.75, "learning_rate": 1.6902086677367576e-05, "loss": 0.5467, "mean_token_accuracy": 0.8436422348022461, "num_tokens": 376210.0, "step": 580 }, { "epoch": 0.3156768325307651, "grad_norm": 10.5, "learning_rate": 1.6848582129481008e-05, "loss": 0.5209, "mean_token_accuracy": 0.8535410046577454, "num_tokens": 381765.0, "step": 590 }, { "epoch": 0.32102728731942215, "grad_norm": 11.75, "learning_rate": 1.6795077581594436e-05, "loss": 0.4805, "mean_token_accuracy": 0.8601600766181946, "num_tokens": 388754.0, "step": 600 }, { "epoch": 0.3263777421080792, "grad_norm": 10.4375, "learning_rate": 1.6741573033707868e-05, "loss": 0.4905, "mean_token_accuracy": 0.8509059309959411, "num_tokens": 395249.0, "step": 610 }, { "epoch": 0.3317281968967362, "grad_norm": 13.75, "learning_rate": 1.6688068485821296e-05, "loss": 0.4774, "mean_token_accuracy": 0.8670556247234344, "num_tokens": 401373.0, "step": 620 }, { "epoch": 0.33707865168539325, "grad_norm": 7.96875, "learning_rate": 1.6634563937934728e-05, "loss": 0.5774, "mean_token_accuracy": 0.8421377062797546, "num_tokens": 408011.0, "step": 630 }, { "epoch": 0.3424291064740503, "grad_norm": 8.8125, "learning_rate": 1.6581059390048156e-05, "loss": 0.4924, "mean_token_accuracy": 0.857643461227417, "num_tokens": 414915.0, "step": 640 }, { "epoch": 0.3477795612627073, "grad_norm": 10.0, "learning_rate": 1.6527554842161584e-05, "loss": 0.5237, "mean_token_accuracy": 0.8495872795581818, "num_tokens": 421459.0, "step": 650 }, { "epoch": 0.35313001605136435, "grad_norm": 7.4375, "learning_rate": 1.6474050294275016e-05, "loss": 0.5691, "mean_token_accuracy": 0.846283620595932, "num_tokens": 428457.0, "step": 660 }, { "epoch": 0.3584804708400214, "grad_norm": 9.5, "learning_rate": 1.6420545746388445e-05, "loss": 0.4969, "mean_token_accuracy": 0.8526872992515564, "num_tokens": 435238.0, "step": 670 }, { "epoch": 0.3638309256286784, "grad_norm": 7.59375, "learning_rate": 1.6367041198501873e-05, "loss": 0.4456, "mean_token_accuracy": 0.8758604288101196, "num_tokens": 441591.0, "step": 680 }, { "epoch": 0.36918138041733545, "grad_norm": 10.0625, "learning_rate": 1.63135366506153e-05, "loss": 0.5724, "mean_token_accuracy": 0.8439893245697021, "num_tokens": 447991.0, "step": 690 }, { "epoch": 0.37453183520599254, "grad_norm": 9.6875, "learning_rate": 1.6260032102728733e-05, "loss": 0.4868, "mean_token_accuracy": 0.8588210284709931, "num_tokens": 454514.0, "step": 700 }, { "epoch": 0.37988228999464957, "grad_norm": 10.5625, "learning_rate": 1.620652755484216e-05, "loss": 0.5282, "mean_token_accuracy": 0.8578278422355652, "num_tokens": 460925.0, "step": 710 }, { "epoch": 0.3852327447833066, "grad_norm": 8.5625, "learning_rate": 1.6153023006955593e-05, "loss": 0.4948, "mean_token_accuracy": 0.8606979012489319, "num_tokens": 467596.0, "step": 720 }, { "epoch": 0.39058319957196364, "grad_norm": 8.9375, "learning_rate": 1.609951845906902e-05, "loss": 0.4643, "mean_token_accuracy": 0.8703970611095428, "num_tokens": 473866.0, "step": 730 }, { "epoch": 0.39593365436062067, "grad_norm": 8.0625, "learning_rate": 1.6046013911182453e-05, "loss": 0.4953, "mean_token_accuracy": 0.8542136013507843, "num_tokens": 480311.0, "step": 740 }, { "epoch": 0.4012841091492777, "grad_norm": 7.84375, "learning_rate": 1.599250936329588e-05, "loss": 0.4899, "mean_token_accuracy": 0.8637793123722076, "num_tokens": 486720.0, "step": 750 }, { "epoch": 0.40663456393793473, "grad_norm": 7.4375, "learning_rate": 1.5939004815409313e-05, "loss": 0.5478, "mean_token_accuracy": 0.8467224180698395, "num_tokens": 493448.0, "step": 760 }, { "epoch": 0.41198501872659177, "grad_norm": 10.5, "learning_rate": 1.588550026752274e-05, "loss": 0.4872, "mean_token_accuracy": 0.861944979429245, "num_tokens": 499902.0, "step": 770 }, { "epoch": 0.4173354735152488, "grad_norm": 9.8125, "learning_rate": 1.583199571963617e-05, "loss": 0.5355, "mean_token_accuracy": 0.845079094171524, "num_tokens": 506365.0, "step": 780 }, { "epoch": 0.42268592830390583, "grad_norm": 11.6875, "learning_rate": 1.5778491171749598e-05, "loss": 0.5154, "mean_token_accuracy": 0.8498620450496673, "num_tokens": 513065.0, "step": 790 }, { "epoch": 0.42803638309256287, "grad_norm": 9.1875, "learning_rate": 1.572498662386303e-05, "loss": 0.5474, "mean_token_accuracy": 0.851120263338089, "num_tokens": 519347.0, "step": 800 }, { "epoch": 0.4333868378812199, "grad_norm": 8.5, "learning_rate": 1.5671482075976458e-05, "loss": 0.5405, "mean_token_accuracy": 0.8517447054386139, "num_tokens": 525664.0, "step": 810 }, { "epoch": 0.43873729266987693, "grad_norm": 7.84375, "learning_rate": 1.561797752808989e-05, "loss": 0.4647, "mean_token_accuracy": 0.8698622226715088, "num_tokens": 532239.0, "step": 820 }, { "epoch": 0.44408774745853397, "grad_norm": 11.9375, "learning_rate": 1.5564472980203318e-05, "loss": 0.5381, "mean_token_accuracy": 0.8493444561958313, "num_tokens": 539312.0, "step": 830 }, { "epoch": 0.449438202247191, "grad_norm": 8.5, "learning_rate": 1.551096843231675e-05, "loss": 0.4776, "mean_token_accuracy": 0.8628087341785431, "num_tokens": 545961.0, "step": 840 }, { "epoch": 0.45478865703584803, "grad_norm": 11.0, "learning_rate": 1.5457463884430178e-05, "loss": 0.5359, "mean_token_accuracy": 0.8486712396144866, "num_tokens": 552915.0, "step": 850 }, { "epoch": 0.46013911182450506, "grad_norm": 12.25, "learning_rate": 1.5403959336543606e-05, "loss": 0.5779, "mean_token_accuracy": 0.8374400615692139, "num_tokens": 559078.0, "step": 860 }, { "epoch": 0.4654895666131621, "grad_norm": 8.4375, "learning_rate": 1.5350454788657038e-05, "loss": 0.4708, "mean_token_accuracy": 0.858840000629425, "num_tokens": 565415.0, "step": 870 }, { "epoch": 0.47084002140181913, "grad_norm": 10.1875, "learning_rate": 1.5296950240770466e-05, "loss": 0.4559, "mean_token_accuracy": 0.870689183473587, "num_tokens": 571434.0, "step": 880 }, { "epoch": 0.47619047619047616, "grad_norm": 7.375, "learning_rate": 1.5243445692883896e-05, "loss": 0.5225, "mean_token_accuracy": 0.8468902468681335, "num_tokens": 577901.0, "step": 890 }, { "epoch": 0.48154093097913325, "grad_norm": 9.8125, "learning_rate": 1.5189941144997326e-05, "loss": 0.4875, "mean_token_accuracy": 0.8613629460334777, "num_tokens": 584621.0, "step": 900 }, { "epoch": 0.4868913857677903, "grad_norm": 8.125, "learning_rate": 1.5136436597110755e-05, "loss": 0.5044, "mean_token_accuracy": 0.8571871101856232, "num_tokens": 591414.0, "step": 910 }, { "epoch": 0.4922418405564473, "grad_norm": 8.1875, "learning_rate": 1.5082932049224186e-05, "loss": 0.495, "mean_token_accuracy": 0.8607097446918488, "num_tokens": 598320.0, "step": 920 }, { "epoch": 0.49759229534510435, "grad_norm": 9.625, "learning_rate": 1.5029427501337615e-05, "loss": 0.5292, "mean_token_accuracy": 0.8526637494564057, "num_tokens": 605223.0, "step": 930 }, { "epoch": 0.5029427501337613, "grad_norm": 10.0625, "learning_rate": 1.4975922953451045e-05, "loss": 0.496, "mean_token_accuracy": 0.8612302124500275, "num_tokens": 611910.0, "step": 940 }, { "epoch": 0.5082932049224184, "grad_norm": 9.75, "learning_rate": 1.4922418405564473e-05, "loss": 0.5256, "mean_token_accuracy": 0.8496457397937774, "num_tokens": 618060.0, "step": 950 }, { "epoch": 0.5136436597110754, "grad_norm": 8.8125, "learning_rate": 1.4868913857677905e-05, "loss": 0.4723, "mean_token_accuracy": 0.8637391149997711, "num_tokens": 624574.0, "step": 960 }, { "epoch": 0.5189941144997324, "grad_norm": 13.1875, "learning_rate": 1.4815409309791333e-05, "loss": 0.542, "mean_token_accuracy": 0.8487633049488068, "num_tokens": 631491.0, "step": 970 }, { "epoch": 0.5243445692883895, "grad_norm": 7.8125, "learning_rate": 1.4761904761904763e-05, "loss": 0.5317, "mean_token_accuracy": 0.850472503900528, "num_tokens": 638853.0, "step": 980 }, { "epoch": 0.5296950240770465, "grad_norm": 7.4375, "learning_rate": 1.4708400214018193e-05, "loss": 0.5616, "mean_token_accuracy": 0.8456982135772705, "num_tokens": 645344.0, "step": 990 }, { "epoch": 0.5350454788657036, "grad_norm": 11.4375, "learning_rate": 1.4654895666131623e-05, "loss": 0.4459, "mean_token_accuracy": 0.8731405317783356, "num_tokens": 651308.0, "step": 1000 }, { "epoch": 0.5403959336543607, "grad_norm": 9.1875, "learning_rate": 1.4601391118245052e-05, "loss": 0.5351, "mean_token_accuracy": 0.8496191263198852, "num_tokens": 657754.0, "step": 1010 }, { "epoch": 0.5457463884430177, "grad_norm": 6.9375, "learning_rate": 1.4547886570358483e-05, "loss": 0.4688, "mean_token_accuracy": 0.8611973404884339, "num_tokens": 664175.0, "step": 1020 }, { "epoch": 0.5510968432316747, "grad_norm": 9.4375, "learning_rate": 1.4494382022471912e-05, "loss": 0.4508, "mean_token_accuracy": 0.8671613514423371, "num_tokens": 671074.0, "step": 1030 }, { "epoch": 0.5564472980203318, "grad_norm": 8.4375, "learning_rate": 1.4440877474585342e-05, "loss": 0.5119, "mean_token_accuracy": 0.8499213218688965, "num_tokens": 677776.0, "step": 1040 }, { "epoch": 0.5617977528089888, "grad_norm": 6.0, "learning_rate": 1.438737292669877e-05, "loss": 0.5044, "mean_token_accuracy": 0.8619873523712158, "num_tokens": 684314.0, "step": 1050 }, { "epoch": 0.5671482075976458, "grad_norm": 8.375, "learning_rate": 1.4333868378812202e-05, "loss": 0.5054, "mean_token_accuracy": 0.848728209733963, "num_tokens": 690881.0, "step": 1060 }, { "epoch": 0.5724986623863029, "grad_norm": 11.5, "learning_rate": 1.428036383092563e-05, "loss": 0.5052, "mean_token_accuracy": 0.8564326465129852, "num_tokens": 696657.0, "step": 1070 }, { "epoch": 0.5778491171749599, "grad_norm": 11.6875, "learning_rate": 1.4226859283039058e-05, "loss": 0.5434, "mean_token_accuracy": 0.847975218296051, "num_tokens": 702770.0, "step": 1080 }, { "epoch": 0.5831995719636169, "grad_norm": 10.125, "learning_rate": 1.417335473515249e-05, "loss": 0.4924, "mean_token_accuracy": 0.8593003988265991, "num_tokens": 709335.0, "step": 1090 }, { "epoch": 0.588550026752274, "grad_norm": 7.59375, "learning_rate": 1.4119850187265918e-05, "loss": 0.5006, "mean_token_accuracy": 0.8596942722797394, "num_tokens": 715235.0, "step": 1100 }, { "epoch": 0.593900481540931, "grad_norm": 9.1875, "learning_rate": 1.4066345639379348e-05, "loss": 0.5074, "mean_token_accuracy": 0.8601271152496338, "num_tokens": 721004.0, "step": 1110 }, { "epoch": 0.599250936329588, "grad_norm": 6.96875, "learning_rate": 1.4012841091492777e-05, "loss": 0.4979, "mean_token_accuracy": 0.8663940608501435, "num_tokens": 727411.0, "step": 1120 }, { "epoch": 0.6046013911182451, "grad_norm": 8.6875, "learning_rate": 1.3959336543606208e-05, "loss": 0.5381, "mean_token_accuracy": 0.8473880052566528, "num_tokens": 734648.0, "step": 1130 }, { "epoch": 0.6099518459069021, "grad_norm": 8.875, "learning_rate": 1.3905831995719637e-05, "loss": 0.4691, "mean_token_accuracy": 0.8657299041748047, "num_tokens": 741439.0, "step": 1140 }, { "epoch": 0.6153023006955591, "grad_norm": 6.71875, "learning_rate": 1.3852327447833067e-05, "loss": 0.4473, "mean_token_accuracy": 0.8780945777893067, "num_tokens": 747475.0, "step": 1150 }, { "epoch": 0.6206527554842162, "grad_norm": 9.1875, "learning_rate": 1.3798822899946495e-05, "loss": 0.4477, "mean_token_accuracy": 0.8737396836280823, "num_tokens": 753825.0, "step": 1160 }, { "epoch": 0.6260032102728732, "grad_norm": 10.3125, "learning_rate": 1.3745318352059927e-05, "loss": 0.5464, "mean_token_accuracy": 0.8504304826259613, "num_tokens": 760665.0, "step": 1170 }, { "epoch": 0.6313536650615302, "grad_norm": 8.625, "learning_rate": 1.3691813804173355e-05, "loss": 0.5125, "mean_token_accuracy": 0.8584359765052796, "num_tokens": 768177.0, "step": 1180 }, { "epoch": 0.6367041198501873, "grad_norm": 5.625, "learning_rate": 1.3638309256286785e-05, "loss": 0.4683, "mean_token_accuracy": 0.8621358156204224, "num_tokens": 774416.0, "step": 1190 }, { "epoch": 0.6420545746388443, "grad_norm": 7.625, "learning_rate": 1.3584804708400215e-05, "loss": 0.4527, "mean_token_accuracy": 0.8733129918575286, "num_tokens": 781075.0, "step": 1200 }, { "epoch": 0.6474050294275013, "grad_norm": 9.375, "learning_rate": 1.3531300160513645e-05, "loss": 0.4839, "mean_token_accuracy": 0.8580621004104614, "num_tokens": 788095.0, "step": 1210 }, { "epoch": 0.6527554842161584, "grad_norm": 7.9375, "learning_rate": 1.3477795612627073e-05, "loss": 0.4827, "mean_token_accuracy": 0.8657938718795777, "num_tokens": 793967.0, "step": 1220 }, { "epoch": 0.6581059390048154, "grad_norm": 8.25, "learning_rate": 1.3424291064740505e-05, "loss": 0.5781, "mean_token_accuracy": 0.8424267470836639, "num_tokens": 800555.0, "step": 1230 }, { "epoch": 0.6634563937934724, "grad_norm": 7.78125, "learning_rate": 1.3370786516853933e-05, "loss": 0.5029, "mean_token_accuracy": 0.8587170779705048, "num_tokens": 806577.0, "step": 1240 }, { "epoch": 0.6688068485821295, "grad_norm": 7.15625, "learning_rate": 1.3317281968967363e-05, "loss": 0.5126, "mean_token_accuracy": 0.8580023169517517, "num_tokens": 813220.0, "step": 1250 }, { "epoch": 0.6741573033707865, "grad_norm": 9.3125, "learning_rate": 1.3263777421080792e-05, "loss": 0.4942, "mean_token_accuracy": 0.8643537580966949, "num_tokens": 819351.0, "step": 1260 }, { "epoch": 0.6795077581594435, "grad_norm": 9.125, "learning_rate": 1.3210272873194224e-05, "loss": 0.5066, "mean_token_accuracy": 0.8472048044204712, "num_tokens": 825655.0, "step": 1270 }, { "epoch": 0.6848582129481006, "grad_norm": 8.0, "learning_rate": 1.3156768325307652e-05, "loss": 0.5565, "mean_token_accuracy": 0.8363802790641784, "num_tokens": 831838.0, "step": 1280 }, { "epoch": 0.6902086677367576, "grad_norm": 11.875, "learning_rate": 1.3103263777421082e-05, "loss": 0.4575, "mean_token_accuracy": 0.8693145513534546, "num_tokens": 838393.0, "step": 1290 }, { "epoch": 0.6955591225254146, "grad_norm": 8.875, "learning_rate": 1.304975922953451e-05, "loss": 0.4696, "mean_token_accuracy": 0.8697144985198975, "num_tokens": 844761.0, "step": 1300 }, { "epoch": 0.7009095773140717, "grad_norm": 6.46875, "learning_rate": 1.2996254681647942e-05, "loss": 0.4838, "mean_token_accuracy": 0.8624867856502533, "num_tokens": 851304.0, "step": 1310 }, { "epoch": 0.7062600321027287, "grad_norm": 10.8125, "learning_rate": 1.294275013376137e-05, "loss": 0.4388, "mean_token_accuracy": 0.872041517496109, "num_tokens": 857241.0, "step": 1320 }, { "epoch": 0.7116104868913857, "grad_norm": 5.46875, "learning_rate": 1.2889245585874802e-05, "loss": 0.4736, "mean_token_accuracy": 0.8685207903385163, "num_tokens": 864157.0, "step": 1330 }, { "epoch": 0.7169609416800428, "grad_norm": 5.96875, "learning_rate": 1.283574103798823e-05, "loss": 0.5193, "mean_token_accuracy": 0.851599931716919, "num_tokens": 871101.0, "step": 1340 }, { "epoch": 0.7223113964686998, "grad_norm": 10.4375, "learning_rate": 1.278223649010166e-05, "loss": 0.4777, "mean_token_accuracy": 0.8594719052314759, "num_tokens": 877637.0, "step": 1350 }, { "epoch": 0.7276618512573568, "grad_norm": 9.75, "learning_rate": 1.2728731942215089e-05, "loss": 0.5032, "mean_token_accuracy": 0.8476529121398926, "num_tokens": 884080.0, "step": 1360 }, { "epoch": 0.7330123060460139, "grad_norm": 9.8125, "learning_rate": 1.267522739432852e-05, "loss": 0.516, "mean_token_accuracy": 0.8532926142215729, "num_tokens": 890701.0, "step": 1370 }, { "epoch": 0.7383627608346709, "grad_norm": 8.9375, "learning_rate": 1.2621722846441949e-05, "loss": 0.4795, "mean_token_accuracy": 0.8652157545089721, "num_tokens": 896880.0, "step": 1380 }, { "epoch": 0.7437132156233279, "grad_norm": 8.8125, "learning_rate": 1.2568218298555379e-05, "loss": 0.4782, "mean_token_accuracy": 0.8685885846614838, "num_tokens": 903104.0, "step": 1390 }, { "epoch": 0.7490636704119851, "grad_norm": 9.5, "learning_rate": 1.2514713750668807e-05, "loss": 0.4377, "mean_token_accuracy": 0.8747394144535064, "num_tokens": 909330.0, "step": 1400 }, { "epoch": 0.7544141252006421, "grad_norm": 9.875, "learning_rate": 1.2461209202782239e-05, "loss": 0.5199, "mean_token_accuracy": 0.8472451269626617, "num_tokens": 916445.0, "step": 1410 }, { "epoch": 0.7597645799892991, "grad_norm": 10.3125, "learning_rate": 1.2407704654895667e-05, "loss": 0.4614, "mean_token_accuracy": 0.8695728182792664, "num_tokens": 922411.0, "step": 1420 }, { "epoch": 0.7651150347779562, "grad_norm": 7.09375, "learning_rate": 1.2354200107009097e-05, "loss": 0.4577, "mean_token_accuracy": 0.8714372098445893, "num_tokens": 929386.0, "step": 1430 }, { "epoch": 0.7704654895666132, "grad_norm": 10.125, "learning_rate": 1.2300695559122527e-05, "loss": 0.4607, "mean_token_accuracy": 0.8719231188297272, "num_tokens": 935645.0, "step": 1440 }, { "epoch": 0.7758159443552702, "grad_norm": 8.5, "learning_rate": 1.2247191011235957e-05, "loss": 0.465, "mean_token_accuracy": 0.863955694437027, "num_tokens": 942398.0, "step": 1450 }, { "epoch": 0.7811663991439273, "grad_norm": 7.0625, "learning_rate": 1.2193686463349385e-05, "loss": 0.474, "mean_token_accuracy": 0.861868017911911, "num_tokens": 949005.0, "step": 1460 }, { "epoch": 0.7865168539325843, "grad_norm": 7.8125, "learning_rate": 1.2140181915462817e-05, "loss": 0.4895, "mean_token_accuracy": 0.8581187665462494, "num_tokens": 956417.0, "step": 1470 }, { "epoch": 0.7918673087212413, "grad_norm": 9.4375, "learning_rate": 1.2086677367576245e-05, "loss": 0.5182, "mean_token_accuracy": 0.8575788676738739, "num_tokens": 963644.0, "step": 1480 }, { "epoch": 0.7972177635098984, "grad_norm": 7.0625, "learning_rate": 1.2033172819689674e-05, "loss": 0.5234, "mean_token_accuracy": 0.8501484274864197, "num_tokens": 970812.0, "step": 1490 }, { "epoch": 0.8025682182985554, "grad_norm": 6.71875, "learning_rate": 1.1979668271803104e-05, "loss": 0.5019, "mean_token_accuracy": 0.8565720319747925, "num_tokens": 977053.0, "step": 1500 }, { "epoch": 0.8079186730872124, "grad_norm": 8.1875, "learning_rate": 1.1926163723916532e-05, "loss": 0.44, "mean_token_accuracy": 0.8646707773208618, "num_tokens": 983762.0, "step": 1510 }, { "epoch": 0.8132691278758695, "grad_norm": 9.625, "learning_rate": 1.1872659176029964e-05, "loss": 0.5464, "mean_token_accuracy": 0.8449681103229523, "num_tokens": 990288.0, "step": 1520 }, { "epoch": 0.8186195826645265, "grad_norm": 8.875, "learning_rate": 1.1819154628143392e-05, "loss": 0.5198, "mean_token_accuracy": 0.8555871546268463, "num_tokens": 997438.0, "step": 1530 }, { "epoch": 0.8239700374531835, "grad_norm": 8.4375, "learning_rate": 1.1765650080256824e-05, "loss": 0.4492, "mean_token_accuracy": 0.8678657710552216, "num_tokens": 1004098.0, "step": 1540 }, { "epoch": 0.8293204922418406, "grad_norm": 11.9375, "learning_rate": 1.1712145532370252e-05, "loss": 0.5576, "mean_token_accuracy": 0.8431281268596649, "num_tokens": 1010380.0, "step": 1550 }, { "epoch": 0.8346709470304976, "grad_norm": 11.625, "learning_rate": 1.1658640984483682e-05, "loss": 0.4539, "mean_token_accuracy": 0.8704251766204834, "num_tokens": 1016789.0, "step": 1560 }, { "epoch": 0.8400214018191546, "grad_norm": 9.0, "learning_rate": 1.160513643659711e-05, "loss": 0.4357, "mean_token_accuracy": 0.8751893520355225, "num_tokens": 1022967.0, "step": 1570 }, { "epoch": 0.8453718566078117, "grad_norm": 10.0625, "learning_rate": 1.1551631888710542e-05, "loss": 0.5466, "mean_token_accuracy": 0.8374965190887451, "num_tokens": 1029135.0, "step": 1580 }, { "epoch": 0.8507223113964687, "grad_norm": 8.125, "learning_rate": 1.149812734082397e-05, "loss": 0.525, "mean_token_accuracy": 0.8522794008255005, "num_tokens": 1035221.0, "step": 1590 }, { "epoch": 0.8560727661851257, "grad_norm": 8.625, "learning_rate": 1.14446227929374e-05, "loss": 0.4801, "mean_token_accuracy": 0.8593502819538117, "num_tokens": 1042125.0, "step": 1600 }, { "epoch": 0.8614232209737828, "grad_norm": 9.0, "learning_rate": 1.1391118245050829e-05, "loss": 0.4982, "mean_token_accuracy": 0.8598998129367829, "num_tokens": 1048953.0, "step": 1610 }, { "epoch": 0.8667736757624398, "grad_norm": 6.625, "learning_rate": 1.133761369716426e-05, "loss": 0.4641, "mean_token_accuracy": 0.8693908751010895, "num_tokens": 1055355.0, "step": 1620 }, { "epoch": 0.8721241305510968, "grad_norm": 10.5, "learning_rate": 1.1284109149277689e-05, "loss": 0.4705, "mean_token_accuracy": 0.8653159081935883, "num_tokens": 1062272.0, "step": 1630 }, { "epoch": 0.8774745853397539, "grad_norm": 9.25, "learning_rate": 1.1230604601391119e-05, "loss": 0.4551, "mean_token_accuracy": 0.871656346321106, "num_tokens": 1069123.0, "step": 1640 }, { "epoch": 0.8828250401284109, "grad_norm": 8.1875, "learning_rate": 1.1177100053504549e-05, "loss": 0.4293, "mean_token_accuracy": 0.8720015764236451, "num_tokens": 1075375.0, "step": 1650 }, { "epoch": 0.8881754949170679, "grad_norm": 14.5625, "learning_rate": 1.1123595505617979e-05, "loss": 0.4974, "mean_token_accuracy": 0.8535413622856141, "num_tokens": 1081597.0, "step": 1660 }, { "epoch": 0.893525949705725, "grad_norm": 6.53125, "learning_rate": 1.1070090957731407e-05, "loss": 0.5454, "mean_token_accuracy": 0.8464123189449311, "num_tokens": 1088245.0, "step": 1670 }, { "epoch": 0.898876404494382, "grad_norm": 10.1875, "learning_rate": 1.1016586409844839e-05, "loss": 0.4541, "mean_token_accuracy": 0.8754329204559326, "num_tokens": 1095246.0, "step": 1680 }, { "epoch": 0.904226859283039, "grad_norm": 7.78125, "learning_rate": 1.0963081861958267e-05, "loss": 0.4978, "mean_token_accuracy": 0.8644464790821076, "num_tokens": 1101925.0, "step": 1690 }, { "epoch": 0.9095773140716961, "grad_norm": 9.25, "learning_rate": 1.0909577314071697e-05, "loss": 0.5421, "mean_token_accuracy": 0.8480099380016327, "num_tokens": 1108818.0, "step": 1700 }, { "epoch": 0.9149277688603531, "grad_norm": 9.4375, "learning_rate": 1.0856072766185126e-05, "loss": 0.4705, "mean_token_accuracy": 0.8590927422046661, "num_tokens": 1115131.0, "step": 1710 }, { "epoch": 0.9202782236490101, "grad_norm": 13.75, "learning_rate": 1.0802568218298557e-05, "loss": 0.4575, "mean_token_accuracy": 0.8650537014007569, "num_tokens": 1121195.0, "step": 1720 }, { "epoch": 0.9256286784376672, "grad_norm": 6.53125, "learning_rate": 1.0749063670411986e-05, "loss": 0.5212, "mean_token_accuracy": 0.852067494392395, "num_tokens": 1127987.0, "step": 1730 }, { "epoch": 0.9309791332263242, "grad_norm": 9.3125, "learning_rate": 1.0695559122525416e-05, "loss": 0.4579, "mean_token_accuracy": 0.866095632314682, "num_tokens": 1134805.0, "step": 1740 }, { "epoch": 0.9363295880149812, "grad_norm": 6.4375, "learning_rate": 1.0642054574638844e-05, "loss": 0.4645, "mean_token_accuracy": 0.8721861898899078, "num_tokens": 1140962.0, "step": 1750 }, { "epoch": 0.9416800428036383, "grad_norm": 7.5, "learning_rate": 1.0588550026752276e-05, "loss": 0.502, "mean_token_accuracy": 0.8602936387062072, "num_tokens": 1147523.0, "step": 1760 }, { "epoch": 0.9470304975922953, "grad_norm": 10.625, "learning_rate": 1.0535045478865704e-05, "loss": 0.4933, "mean_token_accuracy": 0.8610971391201019, "num_tokens": 1154959.0, "step": 1770 }, { "epoch": 0.9523809523809523, "grad_norm": 6.3125, "learning_rate": 1.0481540930979136e-05, "loss": 0.4575, "mean_token_accuracy": 0.8689629137516022, "num_tokens": 1162475.0, "step": 1780 }, { "epoch": 0.9577314071696095, "grad_norm": 7.125, "learning_rate": 1.0428036383092564e-05, "loss": 0.437, "mean_token_accuracy": 0.8738560259342194, "num_tokens": 1168603.0, "step": 1790 }, { "epoch": 0.9630818619582665, "grad_norm": 12.6875, "learning_rate": 1.0374531835205994e-05, "loss": 0.4615, "mean_token_accuracy": 0.8618976831436157, "num_tokens": 1174779.0, "step": 1800 }, { "epoch": 0.9684323167469235, "grad_norm": 10.25, "learning_rate": 1.0321027287319422e-05, "loss": 0.4797, "mean_token_accuracy": 0.8645634591579437, "num_tokens": 1181476.0, "step": 1810 }, { "epoch": 0.9737827715355806, "grad_norm": 7.03125, "learning_rate": 1.0267522739432854e-05, "loss": 0.4296, "mean_token_accuracy": 0.8760972023010254, "num_tokens": 1188089.0, "step": 1820 }, { "epoch": 0.9791332263242376, "grad_norm": 6.65625, "learning_rate": 1.0214018191546282e-05, "loss": 0.5225, "mean_token_accuracy": 0.8593319177627563, "num_tokens": 1195160.0, "step": 1830 }, { "epoch": 0.9844836811128946, "grad_norm": 11.625, "learning_rate": 1.0160513643659712e-05, "loss": 0.489, "mean_token_accuracy": 0.8619667589664459, "num_tokens": 1201876.0, "step": 1840 }, { "epoch": 0.9898341359015517, "grad_norm": 9.25, "learning_rate": 1.010700909577314e-05, "loss": 0.4675, "mean_token_accuracy": 0.8764336228370666, "num_tokens": 1208135.0, "step": 1850 }, { "epoch": 0.9951845906902087, "grad_norm": 8.6875, "learning_rate": 1.0053504547886572e-05, "loss": 0.4956, "mean_token_accuracy": 0.8619385719299316, "num_tokens": 1214656.0, "step": 1860 }, { "epoch": 1.0005350454788657, "grad_norm": 8.125, "learning_rate": 1e-05, "loss": 0.4559, "mean_token_accuracy": 0.8742705702781677, "num_tokens": 1220654.0, "step": 1870 }, { "epoch": 1.0058855002675227, "grad_norm": 8.8125, "learning_rate": 9.94649545211343e-06, "loss": 0.3394, "mean_token_accuracy": 0.8988647699356079, "num_tokens": 1226661.0, "step": 1880 }, { "epoch": 1.0112359550561798, "grad_norm": 5.96875, "learning_rate": 9.892990904226861e-06, "loss": 0.3926, "mean_token_accuracy": 0.8898398280143738, "num_tokens": 1233536.0, "step": 1890 }, { "epoch": 1.0165864098448367, "grad_norm": 7.09375, "learning_rate": 9.83948635634029e-06, "loss": 0.3577, "mean_token_accuracy": 0.8911078989505767, "num_tokens": 1240167.0, "step": 1900 }, { "epoch": 1.0219368646334939, "grad_norm": 6.25, "learning_rate": 9.78598180845372e-06, "loss": 0.3717, "mean_token_accuracy": 0.8895718336105347, "num_tokens": 1246552.0, "step": 1910 }, { "epoch": 1.0272873194221508, "grad_norm": 10.1875, "learning_rate": 9.73247726056715e-06, "loss": 0.3604, "mean_token_accuracy": 0.8959209024906158, "num_tokens": 1253408.0, "step": 1920 }, { "epoch": 1.032637774210808, "grad_norm": 6.0625, "learning_rate": 9.67897271268058e-06, "loss": 0.3521, "mean_token_accuracy": 0.8941415071487426, "num_tokens": 1260313.0, "step": 1930 }, { "epoch": 1.0379882289994649, "grad_norm": 6.34375, "learning_rate": 9.62546816479401e-06, "loss": 0.3065, "mean_token_accuracy": 0.9043578207492828, "num_tokens": 1267225.0, "step": 1940 }, { "epoch": 1.043338683788122, "grad_norm": 11.125, "learning_rate": 9.571963616907438e-06, "loss": 0.404, "mean_token_accuracy": 0.8826900124549866, "num_tokens": 1273953.0, "step": 1950 }, { "epoch": 1.048689138576779, "grad_norm": 9.625, "learning_rate": 9.518459069020868e-06, "loss": 0.3026, "mean_token_accuracy": 0.9042790353298187, "num_tokens": 1280335.0, "step": 1960 }, { "epoch": 1.054039593365436, "grad_norm": 10.8125, "learning_rate": 9.464954521134298e-06, "loss": 0.3693, "mean_token_accuracy": 0.8848277688026428, "num_tokens": 1286700.0, "step": 1970 }, { "epoch": 1.0593900481540932, "grad_norm": 9.25, "learning_rate": 9.411449973247728e-06, "loss": 0.3215, "mean_token_accuracy": 0.8999672591686249, "num_tokens": 1293204.0, "step": 1980 }, { "epoch": 1.0647405029427501, "grad_norm": 8.75, "learning_rate": 9.357945425361158e-06, "loss": 0.3674, "mean_token_accuracy": 0.8879241943359375, "num_tokens": 1299594.0, "step": 1990 }, { "epoch": 1.070090957731407, "grad_norm": 8.8125, "learning_rate": 9.304440877474586e-06, "loss": 0.363, "mean_token_accuracy": 0.8884270906448364, "num_tokens": 1306020.0, "step": 2000 }, { "epoch": 1.0754414125200642, "grad_norm": 9.625, "learning_rate": 9.250936329588016e-06, "loss": 0.3956, "mean_token_accuracy": 0.8869378864765167, "num_tokens": 1312861.0, "step": 2010 }, { "epoch": 1.0807918673087213, "grad_norm": 7.5625, "learning_rate": 9.197431781701446e-06, "loss": 0.3518, "mean_token_accuracy": 0.8959320962429047, "num_tokens": 1319753.0, "step": 2020 }, { "epoch": 1.0861423220973783, "grad_norm": 9.8125, "learning_rate": 9.143927233814876e-06, "loss": 0.4075, "mean_token_accuracy": 0.8801172077655792, "num_tokens": 1326224.0, "step": 2030 }, { "epoch": 1.0914927768860354, "grad_norm": 10.3125, "learning_rate": 9.090422685928304e-06, "loss": 0.3709, "mean_token_accuracy": 0.8898184835910797, "num_tokens": 1333039.0, "step": 2040 }, { "epoch": 1.0968432316746923, "grad_norm": 6.625, "learning_rate": 9.036918138041734e-06, "loss": 0.3034, "mean_token_accuracy": 0.9072488963603973, "num_tokens": 1340236.0, "step": 2050 }, { "epoch": 1.1021936864633495, "grad_norm": 12.0, "learning_rate": 8.983413590155164e-06, "loss": 0.3418, "mean_token_accuracy": 0.9017502725124359, "num_tokens": 1346973.0, "step": 2060 }, { "epoch": 1.1075441412520064, "grad_norm": 7.78125, "learning_rate": 8.929909042268593e-06, "loss": 0.3795, "mean_token_accuracy": 0.8840995252132415, "num_tokens": 1353427.0, "step": 2070 }, { "epoch": 1.1128945960406635, "grad_norm": 9.125, "learning_rate": 8.876404494382023e-06, "loss": 0.3918, "mean_token_accuracy": 0.8862473785877227, "num_tokens": 1360613.0, "step": 2080 }, { "epoch": 1.1182450508293205, "grad_norm": 8.75, "learning_rate": 8.822899946495453e-06, "loss": 0.4017, "mean_token_accuracy": 0.8799244403839112, "num_tokens": 1367228.0, "step": 2090 }, { "epoch": 1.1235955056179776, "grad_norm": 8.9375, "learning_rate": 8.769395398608883e-06, "loss": 0.3047, "mean_token_accuracy": 0.9094751477241516, "num_tokens": 1373735.0, "step": 2100 }, { "epoch": 1.1289459604066345, "grad_norm": 9.125, "learning_rate": 8.715890850722311e-06, "loss": 0.3501, "mean_token_accuracy": 0.8966971039772034, "num_tokens": 1380530.0, "step": 2110 }, { "epoch": 1.1342964151952917, "grad_norm": 8.8125, "learning_rate": 8.662386302835741e-06, "loss": 0.3691, "mean_token_accuracy": 0.8842532277107239, "num_tokens": 1387385.0, "step": 2120 }, { "epoch": 1.1396468699839486, "grad_norm": 10.6875, "learning_rate": 8.608881754949171e-06, "loss": 0.3716, "mean_token_accuracy": 0.895151424407959, "num_tokens": 1394662.0, "step": 2130 }, { "epoch": 1.1449973247726057, "grad_norm": 7.1875, "learning_rate": 8.555377207062601e-06, "loss": 0.3864, "mean_token_accuracy": 0.8837956130504608, "num_tokens": 1401323.0, "step": 2140 }, { "epoch": 1.1503477795612627, "grad_norm": 8.125, "learning_rate": 8.501872659176031e-06, "loss": 0.368, "mean_token_accuracy": 0.8879690945148468, "num_tokens": 1407852.0, "step": 2150 }, { "epoch": 1.1556982343499198, "grad_norm": 6.65625, "learning_rate": 8.44836811128946e-06, "loss": 0.3, "mean_token_accuracy": 0.9060183703899384, "num_tokens": 1414137.0, "step": 2160 }, { "epoch": 1.1610486891385767, "grad_norm": 11.3125, "learning_rate": 8.39486356340289e-06, "loss": 0.3253, "mean_token_accuracy": 0.9028343200683594, "num_tokens": 1421113.0, "step": 2170 }, { "epoch": 1.1663991439272339, "grad_norm": 7.5625, "learning_rate": 8.34135901551632e-06, "loss": 0.2847, "mean_token_accuracy": 0.9114493012428284, "num_tokens": 1427489.0, "step": 2180 }, { "epoch": 1.1717495987158908, "grad_norm": 6.46875, "learning_rate": 8.28785446762975e-06, "loss": 0.2842, "mean_token_accuracy": 0.9132417678833008, "num_tokens": 1433978.0, "step": 2190 }, { "epoch": 1.177100053504548, "grad_norm": 10.5625, "learning_rate": 8.234349919743178e-06, "loss": 0.3206, "mean_token_accuracy": 0.9075594186782837, "num_tokens": 1440172.0, "step": 2200 }, { "epoch": 1.1824505082932049, "grad_norm": 8.25, "learning_rate": 8.180845371856608e-06, "loss": 0.3455, "mean_token_accuracy": 0.8966793894767762, "num_tokens": 1446504.0, "step": 2210 }, { "epoch": 1.187800963081862, "grad_norm": 6.9375, "learning_rate": 8.127340823970038e-06, "loss": 0.3489, "mean_token_accuracy": 0.8949692785739899, "num_tokens": 1453079.0, "step": 2220 }, { "epoch": 1.193151417870519, "grad_norm": 10.6875, "learning_rate": 8.073836276083468e-06, "loss": 0.3881, "mean_token_accuracy": 0.8886881649494172, "num_tokens": 1459908.0, "step": 2230 }, { "epoch": 1.198501872659176, "grad_norm": 7.75, "learning_rate": 8.020331728196898e-06, "loss": 0.3661, "mean_token_accuracy": 0.8884704887866974, "num_tokens": 1467254.0, "step": 2240 }, { "epoch": 1.203852327447833, "grad_norm": 9.0, "learning_rate": 7.966827180310326e-06, "loss": 0.3155, "mean_token_accuracy": 0.8988059878349304, "num_tokens": 1473470.0, "step": 2250 }, { "epoch": 1.2092027822364901, "grad_norm": 13.5, "learning_rate": 7.913322632423756e-06, "loss": 0.3903, "mean_token_accuracy": 0.8906476199626923, "num_tokens": 1480222.0, "step": 2260 }, { "epoch": 1.214553237025147, "grad_norm": 10.4375, "learning_rate": 7.859818084537186e-06, "loss": 0.3134, "mean_token_accuracy": 0.9039566814899445, "num_tokens": 1486782.0, "step": 2270 }, { "epoch": 1.2199036918138042, "grad_norm": 8.0625, "learning_rate": 7.806313536650616e-06, "loss": 0.3347, "mean_token_accuracy": 0.8967584848403931, "num_tokens": 1493113.0, "step": 2280 }, { "epoch": 1.2252541466024611, "grad_norm": 16.75, "learning_rate": 7.752808988764046e-06, "loss": 0.3924, "mean_token_accuracy": 0.8869854748249054, "num_tokens": 1499765.0, "step": 2290 }, { "epoch": 1.2306046013911183, "grad_norm": 12.375, "learning_rate": 7.699304440877475e-06, "loss": 0.3403, "mean_token_accuracy": 0.9002821922302247, "num_tokens": 1506003.0, "step": 2300 }, { "epoch": 1.2359550561797752, "grad_norm": 7.25, "learning_rate": 7.645799892990905e-06, "loss": 0.381, "mean_token_accuracy": 0.878502094745636, "num_tokens": 1512806.0, "step": 2310 }, { "epoch": 1.2413055109684323, "grad_norm": 10.0, "learning_rate": 7.592295345104335e-06, "loss": 0.4031, "mean_token_accuracy": 0.8795210957527161, "num_tokens": 1518946.0, "step": 2320 }, { "epoch": 1.2466559657570893, "grad_norm": 8.4375, "learning_rate": 7.538790797217765e-06, "loss": 0.3466, "mean_token_accuracy": 0.8933668613433838, "num_tokens": 1524768.0, "step": 2330 }, { "epoch": 1.2520064205457464, "grad_norm": 8.6875, "learning_rate": 7.485286249331194e-06, "loss": 0.3471, "mean_token_accuracy": 0.8912493407726287, "num_tokens": 1530975.0, "step": 2340 }, { "epoch": 1.2573568753344033, "grad_norm": 9.0, "learning_rate": 7.431781701444624e-06, "loss": 0.4103, "mean_token_accuracy": 0.877688080072403, "num_tokens": 1537454.0, "step": 2350 }, { "epoch": 1.2627073301230605, "grad_norm": 5.6875, "learning_rate": 7.378277153558053e-06, "loss": 0.3769, "mean_token_accuracy": 0.8930474102497101, "num_tokens": 1544340.0, "step": 2360 }, { "epoch": 1.2680577849117176, "grad_norm": 9.125, "learning_rate": 7.324772605671483e-06, "loss": 0.3346, "mean_token_accuracy": 0.8972549378871918, "num_tokens": 1550488.0, "step": 2370 }, { "epoch": 1.2734082397003745, "grad_norm": 8.0625, "learning_rate": 7.271268057784913e-06, "loss": 0.3049, "mean_token_accuracy": 0.9082387566566468, "num_tokens": 1556730.0, "step": 2380 }, { "epoch": 1.2787586944890315, "grad_norm": 9.375, "learning_rate": 7.217763509898342e-06, "loss": 0.356, "mean_token_accuracy": 0.89334836602211, "num_tokens": 1562581.0, "step": 2390 }, { "epoch": 1.2841091492776886, "grad_norm": 10.1875, "learning_rate": 7.164258962011772e-06, "loss": 0.372, "mean_token_accuracy": 0.8907365202903748, "num_tokens": 1569389.0, "step": 2400 }, { "epoch": 1.2894596040663457, "grad_norm": 10.25, "learning_rate": 7.110754414125201e-06, "loss": 0.3498, "mean_token_accuracy": 0.8915913522243499, "num_tokens": 1575712.0, "step": 2410 }, { "epoch": 1.2948100588550027, "grad_norm": 5.75, "learning_rate": 7.057249866238631e-06, "loss": 0.3118, "mean_token_accuracy": 0.9068877398967743, "num_tokens": 1582020.0, "step": 2420 }, { "epoch": 1.3001605136436596, "grad_norm": 7.84375, "learning_rate": 7.003745318352061e-06, "loss": 0.3317, "mean_token_accuracy": 0.9070441126823425, "num_tokens": 1588164.0, "step": 2430 }, { "epoch": 1.3055109684323167, "grad_norm": 10.875, "learning_rate": 6.950240770465491e-06, "loss": 0.3442, "mean_token_accuracy": 0.8882237911224365, "num_tokens": 1594360.0, "step": 2440 }, { "epoch": 1.3108614232209739, "grad_norm": 9.5625, "learning_rate": 6.896736222578921e-06, "loss": 0.3709, "mean_token_accuracy": 0.889040058851242, "num_tokens": 1600479.0, "step": 2450 }, { "epoch": 1.3162118780096308, "grad_norm": 10.1875, "learning_rate": 6.84323167469235e-06, "loss": 0.37, "mean_token_accuracy": 0.8896639764308929, "num_tokens": 1607273.0, "step": 2460 }, { "epoch": 1.3215623327982877, "grad_norm": 6.625, "learning_rate": 6.789727126805778e-06, "loss": 0.3328, "mean_token_accuracy": 0.8940999627113342, "num_tokens": 1613208.0, "step": 2470 }, { "epoch": 1.3269127875869449, "grad_norm": 9.625, "learning_rate": 6.736222578919208e-06, "loss": 0.3422, "mean_token_accuracy": 0.9023346900939941, "num_tokens": 1620301.0, "step": 2480 }, { "epoch": 1.332263242375602, "grad_norm": 8.25, "learning_rate": 6.682718031032638e-06, "loss": 0.3418, "mean_token_accuracy": 0.9014999985694885, "num_tokens": 1626597.0, "step": 2490 }, { "epoch": 1.337613697164259, "grad_norm": 10.625, "learning_rate": 6.629213483146067e-06, "loss": 0.338, "mean_token_accuracy": 0.9060708105564117, "num_tokens": 1632506.0, "step": 2500 }, { "epoch": 1.342964151952916, "grad_norm": 8.125, "learning_rate": 6.575708935259497e-06, "loss": 0.3491, "mean_token_accuracy": 0.8955333709716797, "num_tokens": 1638948.0, "step": 2510 }, { "epoch": 1.348314606741573, "grad_norm": 10.125, "learning_rate": 6.5222043873729265e-06, "loss": 0.339, "mean_token_accuracy": 0.8978384971618653, "num_tokens": 1645643.0, "step": 2520 }, { "epoch": 1.3536650615302301, "grad_norm": 8.625, "learning_rate": 6.4686998394863565e-06, "loss": 0.3217, "mean_token_accuracy": 0.9035401582717896, "num_tokens": 1652275.0, "step": 2530 }, { "epoch": 1.359015516318887, "grad_norm": 9.1875, "learning_rate": 6.4151952915997865e-06, "loss": 0.3454, "mean_token_accuracy": 0.8946024060249329, "num_tokens": 1658829.0, "step": 2540 }, { "epoch": 1.3643659711075442, "grad_norm": 8.875, "learning_rate": 6.361690743713216e-06, "loss": 0.3405, "mean_token_accuracy": 0.8921943783760071, "num_tokens": 1664662.0, "step": 2550 }, { "epoch": 1.3697164258962011, "grad_norm": 6.1875, "learning_rate": 6.308186195826646e-06, "loss": 0.3327, "mean_token_accuracy": 0.9018294215202332, "num_tokens": 1671085.0, "step": 2560 }, { "epoch": 1.3750668806848583, "grad_norm": 8.0, "learning_rate": 6.254681647940075e-06, "loss": 0.3413, "mean_token_accuracy": 0.9011593520641327, "num_tokens": 1677920.0, "step": 2570 }, { "epoch": 1.3804173354735152, "grad_norm": 10.9375, "learning_rate": 6.201177100053505e-06, "loss": 0.3833, "mean_token_accuracy": 0.8926679432392121, "num_tokens": 1683811.0, "step": 2580 }, { "epoch": 1.3857677902621723, "grad_norm": 8.75, "learning_rate": 6.147672552166934e-06, "loss": 0.2923, "mean_token_accuracy": 0.9132450461387634, "num_tokens": 1689979.0, "step": 2590 }, { "epoch": 1.3911182450508293, "grad_norm": 7.59375, "learning_rate": 6.094168004280364e-06, "loss": 0.2582, "mean_token_accuracy": 0.9228733658790589, "num_tokens": 1696264.0, "step": 2600 }, { "epoch": 1.3964686998394864, "grad_norm": 9.625, "learning_rate": 6.040663456393794e-06, "loss": 0.3403, "mean_token_accuracy": 0.8928385436534881, "num_tokens": 1702973.0, "step": 2610 }, { "epoch": 1.4018191546281433, "grad_norm": 7.90625, "learning_rate": 5.987158908507223e-06, "loss": 0.2963, "mean_token_accuracy": 0.9104150712490082, "num_tokens": 1709602.0, "step": 2620 }, { "epoch": 1.4071696094168005, "grad_norm": 7.6875, "learning_rate": 5.933654360620653e-06, "loss": 0.3216, "mean_token_accuracy": 0.9025378346443176, "num_tokens": 1716470.0, "step": 2630 }, { "epoch": 1.4125200642054574, "grad_norm": 10.3125, "learning_rate": 5.8801498127340825e-06, "loss": 0.3962, "mean_token_accuracy": 0.878995281457901, "num_tokens": 1723392.0, "step": 2640 }, { "epoch": 1.4178705189941145, "grad_norm": 6.40625, "learning_rate": 5.8266452648475125e-06, "loss": 0.3892, "mean_token_accuracy": 0.8873062074184418, "num_tokens": 1730010.0, "step": 2650 }, { "epoch": 1.4232209737827715, "grad_norm": 7.625, "learning_rate": 5.7731407169609425e-06, "loss": 0.3133, "mean_token_accuracy": 0.9009443700313569, "num_tokens": 1736820.0, "step": 2660 }, { "epoch": 1.4285714285714286, "grad_norm": 7.09375, "learning_rate": 5.719636169074372e-06, "loss": 0.3286, "mean_token_accuracy": 0.896684056520462, "num_tokens": 1743860.0, "step": 2670 }, { "epoch": 1.4339218833600855, "grad_norm": 7.28125, "learning_rate": 5.666131621187802e-06, "loss": 0.348, "mean_token_accuracy": 0.8956638753414154, "num_tokens": 1750364.0, "step": 2680 }, { "epoch": 1.4392723381487427, "grad_norm": 8.6875, "learning_rate": 5.612627073301231e-06, "loss": 0.3551, "mean_token_accuracy": 0.8925686001777648, "num_tokens": 1756391.0, "step": 2690 }, { "epoch": 1.4446227929373996, "grad_norm": 9.0625, "learning_rate": 5.559122525414661e-06, "loss": 0.3404, "mean_token_accuracy": 0.8996007680892945, "num_tokens": 1763019.0, "step": 2700 }, { "epoch": 1.4499732477260567, "grad_norm": 8.4375, "learning_rate": 5.50561797752809e-06, "loss": 0.3093, "mean_token_accuracy": 0.9115182936191559, "num_tokens": 1769392.0, "step": 2710 }, { "epoch": 1.4553237025147139, "grad_norm": 8.1875, "learning_rate": 5.45211342964152e-06, "loss": 0.3389, "mean_token_accuracy": 0.8998614251613617, "num_tokens": 1775905.0, "step": 2720 }, { "epoch": 1.4606741573033708, "grad_norm": 7.90625, "learning_rate": 5.39860888175495e-06, "loss": 0.3237, "mean_token_accuracy": 0.9044057548046112, "num_tokens": 1782214.0, "step": 2730 }, { "epoch": 1.4660246120920277, "grad_norm": 10.0, "learning_rate": 5.345104333868379e-06, "loss": 0.3756, "mean_token_accuracy": 0.8828893363475799, "num_tokens": 1788446.0, "step": 2740 }, { "epoch": 1.4713750668806849, "grad_norm": 10.875, "learning_rate": 5.291599785981809e-06, "loss": 0.3584, "mean_token_accuracy": 0.8971490979194641, "num_tokens": 1794711.0, "step": 2750 }, { "epoch": 1.476725521669342, "grad_norm": 8.8125, "learning_rate": 5.2380952380952384e-06, "loss": 0.3247, "mean_token_accuracy": 0.9000103354454041, "num_tokens": 1800841.0, "step": 2760 }, { "epoch": 1.482075976457999, "grad_norm": 7.75, "learning_rate": 5.1845906902086685e-06, "loss": 0.3371, "mean_token_accuracy": 0.8988662958145142, "num_tokens": 1807228.0, "step": 2770 }, { "epoch": 1.4874264312466559, "grad_norm": 9.875, "learning_rate": 5.1310861423220985e-06, "loss": 0.3404, "mean_token_accuracy": 0.9001451551914215, "num_tokens": 1814225.0, "step": 2780 }, { "epoch": 1.492776886035313, "grad_norm": 10.3125, "learning_rate": 5.077581594435528e-06, "loss": 0.4084, "mean_token_accuracy": 0.8779936671257019, "num_tokens": 1820905.0, "step": 2790 }, { "epoch": 1.4981273408239701, "grad_norm": 12.5625, "learning_rate": 5.024077046548958e-06, "loss": 0.3701, "mean_token_accuracy": 0.8849872767925262, "num_tokens": 1827102.0, "step": 2800 }, { "epoch": 1.503477795612627, "grad_norm": 9.4375, "learning_rate": 4.970572498662387e-06, "loss": 0.3414, "mean_token_accuracy": 0.897352808713913, "num_tokens": 1833523.0, "step": 2810 }, { "epoch": 1.508828250401284, "grad_norm": 8.25, "learning_rate": 4.917067950775816e-06, "loss": 0.3864, "mean_token_accuracy": 0.8879909336566925, "num_tokens": 1840350.0, "step": 2820 }, { "epoch": 1.5141787051899411, "grad_norm": 6.5, "learning_rate": 4.863563402889246e-06, "loss": 0.3325, "mean_token_accuracy": 0.9002113997936249, "num_tokens": 1847160.0, "step": 2830 }, { "epoch": 1.5195291599785983, "grad_norm": 9.5, "learning_rate": 4.810058855002675e-06, "loss": 0.3482, "mean_token_accuracy": 0.8907867848873139, "num_tokens": 1853521.0, "step": 2840 }, { "epoch": 1.5248796147672552, "grad_norm": 7.53125, "learning_rate": 4.756554307116105e-06, "loss": 0.3298, "mean_token_accuracy": 0.9050326466560363, "num_tokens": 1859824.0, "step": 2850 }, { "epoch": 1.5302300695559121, "grad_norm": 10.6875, "learning_rate": 4.703049759229535e-06, "loss": 0.3128, "mean_token_accuracy": 0.906907033920288, "num_tokens": 1865776.0, "step": 2860 }, { "epoch": 1.5355805243445693, "grad_norm": 7.5625, "learning_rate": 4.649545211342964e-06, "loss": 0.3428, "mean_token_accuracy": 0.8972375392913818, "num_tokens": 1872781.0, "step": 2870 }, { "epoch": 1.5409309791332264, "grad_norm": 6.75, "learning_rate": 4.596040663456394e-06, "loss": 0.3074, "mean_token_accuracy": 0.9071969211101532, "num_tokens": 1879491.0, "step": 2880 }, { "epoch": 1.5462814339218833, "grad_norm": 23.25, "learning_rate": 4.542536115569824e-06, "loss": 0.3335, "mean_token_accuracy": 0.9018124461174011, "num_tokens": 1885428.0, "step": 2890 }, { "epoch": 1.5516318887105403, "grad_norm": 8.25, "learning_rate": 4.489031567683254e-06, "loss": 0.3902, "mean_token_accuracy": 0.8826347589492798, "num_tokens": 1892532.0, "step": 2900 }, { "epoch": 1.5569823434991974, "grad_norm": 7.75, "learning_rate": 4.435527019796684e-06, "loss": 0.3117, "mean_token_accuracy": 0.9055287480354309, "num_tokens": 1899423.0, "step": 2910 }, { "epoch": 1.5623327982878545, "grad_norm": 10.3125, "learning_rate": 4.382022471910113e-06, "loss": 0.3462, "mean_token_accuracy": 0.8927427113056183, "num_tokens": 1906348.0, "step": 2920 }, { "epoch": 1.5676832530765115, "grad_norm": 9.8125, "learning_rate": 4.328517924023543e-06, "loss": 0.3395, "mean_token_accuracy": 0.8988035976886749, "num_tokens": 1913174.0, "step": 2930 }, { "epoch": 1.5730337078651684, "grad_norm": 7.25, "learning_rate": 4.275013376136972e-06, "loss": 0.3363, "mean_token_accuracy": 0.9001219630241394, "num_tokens": 1919532.0, "step": 2940 }, { "epoch": 1.5783841626538255, "grad_norm": 9.1875, "learning_rate": 4.221508828250402e-06, "loss": 0.3718, "mean_token_accuracy": 0.8894855201244354, "num_tokens": 1925669.0, "step": 2950 }, { "epoch": 1.5837346174424827, "grad_norm": 6.5, "learning_rate": 4.168004280363831e-06, "loss": 0.3727, "mean_token_accuracy": 0.8913348078727722, "num_tokens": 1933020.0, "step": 2960 }, { "epoch": 1.5890850722311396, "grad_norm": 8.9375, "learning_rate": 4.11449973247726e-06, "loss": 0.3519, "mean_token_accuracy": 0.89510737657547, "num_tokens": 1939268.0, "step": 2970 }, { "epoch": 1.5944355270197965, "grad_norm": 8.8125, "learning_rate": 4.06099518459069e-06, "loss": 0.3251, "mean_token_accuracy": 0.9008099377155304, "num_tokens": 1946433.0, "step": 2980 }, { "epoch": 1.5997859818084539, "grad_norm": 10.0, "learning_rate": 4.00749063670412e-06, "loss": 0.3386, "mean_token_accuracy": 0.8937528669834137, "num_tokens": 1952254.0, "step": 2990 }, { "epoch": 1.6051364365971108, "grad_norm": 9.9375, "learning_rate": 3.9539860888175495e-06, "loss": 0.3478, "mean_token_accuracy": 0.8926220834255219, "num_tokens": 1958313.0, "step": 3000 }, { "epoch": 1.6104868913857677, "grad_norm": 10.5, "learning_rate": 3.9004815409309795e-06, "loss": 0.3792, "mean_token_accuracy": 0.8869211077690125, "num_tokens": 1964738.0, "step": 3010 }, { "epoch": 1.6158373461744249, "grad_norm": 8.5625, "learning_rate": 3.846976993044409e-06, "loss": 0.3428, "mean_token_accuracy": 0.9007344305515289, "num_tokens": 1971092.0, "step": 3020 }, { "epoch": 1.621187800963082, "grad_norm": 12.5, "learning_rate": 3.7934724451578387e-06, "loss": 0.2979, "mean_token_accuracy": 0.9057847261428833, "num_tokens": 1977152.0, "step": 3030 }, { "epoch": 1.626538255751739, "grad_norm": 11.0625, "learning_rate": 3.7399678972712683e-06, "loss": 0.3014, "mean_token_accuracy": 0.910237044095993, "num_tokens": 1982979.0, "step": 3040 }, { "epoch": 1.6318887105403959, "grad_norm": 10.75, "learning_rate": 3.686463349384698e-06, "loss": 0.3689, "mean_token_accuracy": 0.8931089162826538, "num_tokens": 1989823.0, "step": 3050 }, { "epoch": 1.637239165329053, "grad_norm": 10.3125, "learning_rate": 3.6329588014981275e-06, "loss": 0.362, "mean_token_accuracy": 0.8893911719322205, "num_tokens": 1997073.0, "step": 3060 }, { "epoch": 1.6425896201177101, "grad_norm": 10.1875, "learning_rate": 3.579454253611557e-06, "loss": 0.344, "mean_token_accuracy": 0.9022652387619019, "num_tokens": 2003752.0, "step": 3070 }, { "epoch": 1.647940074906367, "grad_norm": 7.90625, "learning_rate": 3.525949705724987e-06, "loss": 0.3392, "mean_token_accuracy": 0.8962293148040772, "num_tokens": 2011479.0, "step": 3080 }, { "epoch": 1.653290529695024, "grad_norm": 6.0, "learning_rate": 3.4724451578384167e-06, "loss": 0.2623, "mean_token_accuracy": 0.9199820160865784, "num_tokens": 2017807.0, "step": 3090 }, { "epoch": 1.6586409844836811, "grad_norm": 6.28125, "learning_rate": 3.4189406099518463e-06, "loss": 0.3305, "mean_token_accuracy": 0.901260507106781, "num_tokens": 2024158.0, "step": 3100 }, { "epoch": 1.6639914392723383, "grad_norm": 10.25, "learning_rate": 3.365436062065276e-06, "loss": 0.3646, "mean_token_accuracy": 0.8901061832904815, "num_tokens": 2030992.0, "step": 3110 }, { "epoch": 1.6693418940609952, "grad_norm": 12.5625, "learning_rate": 3.3119315141787055e-06, "loss": 0.3578, "mean_token_accuracy": 0.8952379763126374, "num_tokens": 2037390.0, "step": 3120 }, { "epoch": 1.6746923488496521, "grad_norm": 8.5625, "learning_rate": 3.258426966292135e-06, "loss": 0.322, "mean_token_accuracy": 0.9057238042354584, "num_tokens": 2044386.0, "step": 3130 }, { "epoch": 1.6800428036383093, "grad_norm": 7.625, "learning_rate": 3.204922418405565e-06, "loss": 0.3176, "mean_token_accuracy": 0.9069669425487519, "num_tokens": 2050619.0, "step": 3140 }, { "epoch": 1.6853932584269664, "grad_norm": 9.625, "learning_rate": 3.1514178705189947e-06, "loss": 0.3639, "mean_token_accuracy": 0.8924400031566619, "num_tokens": 2057077.0, "step": 3150 }, { "epoch": 1.6907437132156233, "grad_norm": 8.25, "learning_rate": 3.0979133226324243e-06, "loss": 0.3445, "mean_token_accuracy": 0.8952427864074707, "num_tokens": 2063253.0, "step": 3160 }, { "epoch": 1.6960941680042803, "grad_norm": 9.3125, "learning_rate": 3.0444087747458535e-06, "loss": 0.3174, "mean_token_accuracy": 0.9069586932659149, "num_tokens": 2069729.0, "step": 3170 }, { "epoch": 1.7014446227929374, "grad_norm": 9.375, "learning_rate": 2.990904226859283e-06, "loss": 0.3848, "mean_token_accuracy": 0.888233357667923, "num_tokens": 2076469.0, "step": 3180 }, { "epoch": 1.7067950775815945, "grad_norm": 7.75, "learning_rate": 2.9373996789727127e-06, "loss": 0.3392, "mean_token_accuracy": 0.9015300810337067, "num_tokens": 2082871.0, "step": 3190 }, { "epoch": 1.7121455323702515, "grad_norm": 15.4375, "learning_rate": 2.8838951310861422e-06, "loss": 0.3588, "mean_token_accuracy": 0.8898575723171234, "num_tokens": 2089392.0, "step": 3200 }, { "epoch": 1.7174959871589084, "grad_norm": 8.875, "learning_rate": 2.830390583199572e-06, "loss": 0.3714, "mean_token_accuracy": 0.8909191906452179, "num_tokens": 2096174.0, "step": 3210 }, { "epoch": 1.7228464419475655, "grad_norm": 8.9375, "learning_rate": 2.776886035313002e-06, "loss": 0.3347, "mean_token_accuracy": 0.8951489269733429, "num_tokens": 2102666.0, "step": 3220 }, { "epoch": 1.7281968967362227, "grad_norm": 7.6875, "learning_rate": 2.7233814874264315e-06, "loss": 0.3282, "mean_token_accuracy": 0.8995968997478485, "num_tokens": 2109330.0, "step": 3230 }, { "epoch": 1.7335473515248796, "grad_norm": 10.375, "learning_rate": 2.669876939539861e-06, "loss": 0.3086, "mean_token_accuracy": 0.9100795567035675, "num_tokens": 2115781.0, "step": 3240 }, { "epoch": 1.7388978063135365, "grad_norm": 8.3125, "learning_rate": 2.6163723916532906e-06, "loss": 0.3324, "mean_token_accuracy": 0.8953737199306488, "num_tokens": 2122021.0, "step": 3250 }, { "epoch": 1.7442482611021937, "grad_norm": 8.5, "learning_rate": 2.5628678437667202e-06, "loss": 0.2891, "mean_token_accuracy": 0.9065665066242218, "num_tokens": 2128633.0, "step": 3260 }, { "epoch": 1.7495987158908508, "grad_norm": 9.8125, "learning_rate": 2.5093632958801502e-06, "loss": 0.4166, "mean_token_accuracy": 0.8839025139808655, "num_tokens": 2135025.0, "step": 3270 }, { "epoch": 1.7549491706795077, "grad_norm": 9.125, "learning_rate": 2.45585874799358e-06, "loss": 0.3554, "mean_token_accuracy": 0.8929996490478516, "num_tokens": 2141295.0, "step": 3280 }, { "epoch": 1.7602996254681647, "grad_norm": 8.4375, "learning_rate": 2.4023542001070094e-06, "loss": 0.3447, "mean_token_accuracy": 0.9023141860961914, "num_tokens": 2147616.0, "step": 3290 }, { "epoch": 1.7656500802568218, "grad_norm": 8.25, "learning_rate": 2.348849652220439e-06, "loss": 0.2771, "mean_token_accuracy": 0.9172522783279419, "num_tokens": 2153542.0, "step": 3300 }, { "epoch": 1.771000535045479, "grad_norm": 8.6875, "learning_rate": 2.2953451043338686e-06, "loss": 0.3327, "mean_token_accuracy": 0.8964531362056732, "num_tokens": 2160023.0, "step": 3310 }, { "epoch": 1.7763509898341359, "grad_norm": 12.375, "learning_rate": 2.2418405564472982e-06, "loss": 0.3264, "mean_token_accuracy": 0.8989507377147674, "num_tokens": 2166244.0, "step": 3320 }, { "epoch": 1.7817014446227928, "grad_norm": 7.9375, "learning_rate": 2.188336008560728e-06, "loss": 0.3699, "mean_token_accuracy": 0.888497507572174, "num_tokens": 2173440.0, "step": 3330 }, { "epoch": 1.78705189941145, "grad_norm": 11.25, "learning_rate": 2.1348314606741574e-06, "loss": 0.3742, "mean_token_accuracy": 0.8912931084632874, "num_tokens": 2179711.0, "step": 3340 }, { "epoch": 1.792402354200107, "grad_norm": 7.4375, "learning_rate": 2.081326912787587e-06, "loss": 0.3864, "mean_token_accuracy": 0.8808314979076386, "num_tokens": 2186921.0, "step": 3350 }, { "epoch": 1.797752808988764, "grad_norm": 8.1875, "learning_rate": 2.0278223649010166e-06, "loss": 0.3362, "mean_token_accuracy": 0.9031544148921966, "num_tokens": 2193385.0, "step": 3360 }, { "epoch": 1.803103263777421, "grad_norm": 11.4375, "learning_rate": 1.9743178170144466e-06, "loss": 0.3126, "mean_token_accuracy": 0.9044448614120484, "num_tokens": 2199762.0, "step": 3370 }, { "epoch": 1.808453718566078, "grad_norm": 9.0, "learning_rate": 1.920813269127876e-06, "loss": 0.3434, "mean_token_accuracy": 0.8923964023590087, "num_tokens": 2206344.0, "step": 3380 }, { "epoch": 1.8138041733547352, "grad_norm": 6.65625, "learning_rate": 1.8673087212413058e-06, "loss": 0.2742, "mean_token_accuracy": 0.9249617278575897, "num_tokens": 2213149.0, "step": 3390 }, { "epoch": 1.8191546281433921, "grad_norm": 11.8125, "learning_rate": 1.8138041733547354e-06, "loss": 0.3236, "mean_token_accuracy": 0.9012238442897796, "num_tokens": 2220066.0, "step": 3400 }, { "epoch": 1.8245050829320493, "grad_norm": 7.96875, "learning_rate": 1.760299625468165e-06, "loss": 0.325, "mean_token_accuracy": 0.9002729892730713, "num_tokens": 2226983.0, "step": 3410 }, { "epoch": 1.8298555377207064, "grad_norm": 9.125, "learning_rate": 1.7067950775815944e-06, "loss": 0.3375, "mean_token_accuracy": 0.8952729225158691, "num_tokens": 2232785.0, "step": 3420 }, { "epoch": 1.8352059925093633, "grad_norm": 10.6875, "learning_rate": 1.6532905296950242e-06, "loss": 0.3058, "mean_token_accuracy": 0.9085634052753448, "num_tokens": 2239708.0, "step": 3430 }, { "epoch": 1.8405564472980203, "grad_norm": 10.125, "learning_rate": 1.5997859818084538e-06, "loss": 0.3692, "mean_token_accuracy": 0.885639077425003, "num_tokens": 2246306.0, "step": 3440 }, { "epoch": 1.8459069020866774, "grad_norm": 9.5, "learning_rate": 1.5462814339218834e-06, "loss": 0.3302, "mean_token_accuracy": 0.8959932744503021, "num_tokens": 2252583.0, "step": 3450 }, { "epoch": 1.8512573568753345, "grad_norm": 9.9375, "learning_rate": 1.4927768860353132e-06, "loss": 0.3236, "mean_token_accuracy": 0.9030149042606354, "num_tokens": 2258814.0, "step": 3460 }, { "epoch": 1.8566078116639915, "grad_norm": 8.375, "learning_rate": 1.4392723381487428e-06, "loss": 0.2901, "mean_token_accuracy": 0.9094631016254425, "num_tokens": 2264764.0, "step": 3470 }, { "epoch": 1.8619582664526484, "grad_norm": 7.03125, "learning_rate": 1.3857677902621726e-06, "loss": 0.3744, "mean_token_accuracy": 0.8912467360496521, "num_tokens": 2271073.0, "step": 3480 }, { "epoch": 1.8673087212413055, "grad_norm": 7.125, "learning_rate": 1.3322632423756022e-06, "loss": 0.3651, "mean_token_accuracy": 0.8913474678993225, "num_tokens": 2277607.0, "step": 3490 }, { "epoch": 1.8726591760299627, "grad_norm": 10.125, "learning_rate": 1.2787586944890317e-06, "loss": 0.3505, "mean_token_accuracy": 0.8946048021316528, "num_tokens": 2284061.0, "step": 3500 }, { "epoch": 1.8780096308186196, "grad_norm": 10.6875, "learning_rate": 1.2252541466024613e-06, "loss": 0.32, "mean_token_accuracy": 0.9035404205322266, "num_tokens": 2290550.0, "step": 3510 }, { "epoch": 1.8833600856072765, "grad_norm": 10.5625, "learning_rate": 1.171749598715891e-06, "loss": 0.3797, "mean_token_accuracy": 0.8915813207626343, "num_tokens": 2296968.0, "step": 3520 }, { "epoch": 1.8887105403959337, "grad_norm": 8.125, "learning_rate": 1.1182450508293207e-06, "loss": 0.343, "mean_token_accuracy": 0.8994002282619477, "num_tokens": 2303909.0, "step": 3530 }, { "epoch": 1.8940609951845908, "grad_norm": 6.46875, "learning_rate": 1.0647405029427501e-06, "loss": 0.3846, "mean_token_accuracy": 0.8919515788555146, "num_tokens": 2310432.0, "step": 3540 }, { "epoch": 1.8994114499732477, "grad_norm": 8.25, "learning_rate": 1.01123595505618e-06, "loss": 0.3954, "mean_token_accuracy": 0.8824563086032867, "num_tokens": 2317407.0, "step": 3550 }, { "epoch": 1.9047619047619047, "grad_norm": 8.8125, "learning_rate": 9.577314071696095e-07, "loss": 0.3479, "mean_token_accuracy": 0.8875617921352387, "num_tokens": 2324712.0, "step": 3560 }, { "epoch": 1.9101123595505618, "grad_norm": 8.0625, "learning_rate": 9.042268592830391e-07, "loss": 0.3082, "mean_token_accuracy": 0.9076863348484039, "num_tokens": 2330868.0, "step": 3570 }, { "epoch": 1.915462814339219, "grad_norm": 6.8125, "learning_rate": 8.507223113964688e-07, "loss": 0.3674, "mean_token_accuracy": 0.8886627376079559, "num_tokens": 2337924.0, "step": 3580 }, { "epoch": 1.9208132691278759, "grad_norm": 8.75, "learning_rate": 7.972177635098984e-07, "loss": 0.3268, "mean_token_accuracy": 0.9019134938716888, "num_tokens": 2344581.0, "step": 3590 }, { "epoch": 1.9261637239165328, "grad_norm": 9.1875, "learning_rate": 7.43713215623328e-07, "loss": 0.3517, "mean_token_accuracy": 0.8964671671390534, "num_tokens": 2351482.0, "step": 3600 }, { "epoch": 1.93151417870519, "grad_norm": 6.03125, "learning_rate": 6.902086677367577e-07, "loss": 0.2813, "mean_token_accuracy": 0.9128317832946777, "num_tokens": 2359060.0, "step": 3610 }, { "epoch": 1.936864633493847, "grad_norm": 6.53125, "learning_rate": 6.367041198501874e-07, "loss": 0.3894, "mean_token_accuracy": 0.8853586316108704, "num_tokens": 2365528.0, "step": 3620 }, { "epoch": 1.942215088282504, "grad_norm": 10.0625, "learning_rate": 5.83199571963617e-07, "loss": 0.364, "mean_token_accuracy": 0.887633866071701, "num_tokens": 2371856.0, "step": 3630 }, { "epoch": 1.947565543071161, "grad_norm": 6.71875, "learning_rate": 5.296950240770466e-07, "loss": 0.3765, "mean_token_accuracy": 0.8906560599803924, "num_tokens": 2378509.0, "step": 3640 }, { "epoch": 1.952915997859818, "grad_norm": 6.0625, "learning_rate": 4.7619047619047623e-07, "loss": 0.349, "mean_token_accuracy": 0.8941151320934295, "num_tokens": 2385085.0, "step": 3650 }, { "epoch": 1.9582664526484752, "grad_norm": 12.8125, "learning_rate": 4.2268592830390587e-07, "loss": 0.3443, "mean_token_accuracy": 0.894556188583374, "num_tokens": 2391771.0, "step": 3660 }, { "epoch": 1.9636169074371321, "grad_norm": 11.5, "learning_rate": 3.691813804173355e-07, "loss": 0.3169, "mean_token_accuracy": 0.9032965779304505, "num_tokens": 2397758.0, "step": 3670 }, { "epoch": 1.968967362225789, "grad_norm": 10.125, "learning_rate": 3.156768325307651e-07, "loss": 0.3207, "mean_token_accuracy": 0.9104713261127472, "num_tokens": 2403999.0, "step": 3680 }, { "epoch": 1.9743178170144462, "grad_norm": 7.8125, "learning_rate": 2.621722846441948e-07, "loss": 0.3217, "mean_token_accuracy": 0.9018463790416718, "num_tokens": 2410165.0, "step": 3690 }, { "epoch": 1.9796682718031033, "grad_norm": 8.5, "learning_rate": 2.0866773675762443e-07, "loss": 0.3518, "mean_token_accuracy": 0.891678249835968, "num_tokens": 2417068.0, "step": 3700 }, { "epoch": 1.9850187265917603, "grad_norm": 8.875, "learning_rate": 1.5516318887105405e-07, "loss": 0.3323, "mean_token_accuracy": 0.8979846119880677, "num_tokens": 2423856.0, "step": 3710 }, { "epoch": 1.9903691813804172, "grad_norm": 10.5, "learning_rate": 1.0165864098448369e-07, "loss": 0.3551, "mean_token_accuracy": 0.8831677317619324, "num_tokens": 2429846.0, "step": 3720 }, { "epoch": 1.9957196361690743, "grad_norm": 7.75, "learning_rate": 4.8154093097913325e-08, "loss": 0.3107, "mean_token_accuracy": 0.9064607203006745, "num_tokens": 2435968.0, "step": 3730 } ], "logging_steps": 10, "max_steps": 3738, "num_input_tokens_seen": 0, "num_train_epochs": 2, "save_steps": 5000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 1.978366146705408e+16, "train_batch_size": 4, "trial_name": null, "trial_params": null }