{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 2.0, "eval_steps": 500, "global_step": 5504, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.003633720930232558, "grad_norm": 26.975276947021484, "learning_rate": 1.6304347826086958e-07, "loss": 2.3138734817504885, "step": 10 }, { "epoch": 0.007267441860465116, "grad_norm": 23.218563079833984, "learning_rate": 3.442028985507247e-07, "loss": 2.209037017822266, "step": 20 }, { "epoch": 0.010901162790697675, "grad_norm": 12.365008354187012, "learning_rate": 5.253623188405797e-07, "loss": 1.977761459350586, "step": 30 }, { "epoch": 0.014534883720930232, "grad_norm": 7.03855562210083, "learning_rate": 7.065217391304348e-07, "loss": 1.6725656509399414, "step": 40 }, { "epoch": 0.018168604651162792, "grad_norm": 5.794276237487793, "learning_rate": 8.8768115942029e-07, "loss": 1.4673602104187011, "step": 50 }, { "epoch": 0.02180232558139535, "grad_norm": 4.949296474456787, "learning_rate": 1.068840579710145e-06, "loss": 1.354590892791748, "step": 60 }, { "epoch": 0.025436046511627907, "grad_norm": 4.834152698516846, "learning_rate": 1.25e-06, "loss": 1.2498680114746095, "step": 70 }, { "epoch": 0.029069767441860465, "grad_norm": 4.502780437469482, "learning_rate": 1.4311594202898552e-06, "loss": 1.137081241607666, "step": 80 }, { "epoch": 0.032703488372093026, "grad_norm": 3.9782040119171143, "learning_rate": 1.6123188405797101e-06, "loss": 1.1200174331665038, "step": 90 }, { "epoch": 0.036337209302325583, "grad_norm": 4.458466529846191, "learning_rate": 1.7934782608695653e-06, "loss": 1.091425323486328, "step": 100 }, { "epoch": 0.03997093023255814, "grad_norm": 4.185495853424072, "learning_rate": 1.9746376811594204e-06, "loss": 1.0495548248291016, "step": 110 }, { "epoch": 0.0436046511627907, "grad_norm": 4.143903732299805, "learning_rate": 2.1557971014492757e-06, "loss": 1.012381362915039, "step": 120 }, { "epoch": 0.047238372093023256, "grad_norm": 4.220444202423096, "learning_rate": 2.3369565217391307e-06, "loss": 1.004845428466797, "step": 130 }, { "epoch": 0.050872093023255814, "grad_norm": 4.035683631896973, "learning_rate": 2.518115942028986e-06, "loss": 0.9986296653747558, "step": 140 }, { "epoch": 0.05450581395348837, "grad_norm": 4.283889293670654, "learning_rate": 2.6992753623188405e-06, "loss": 1.011301040649414, "step": 150 }, { "epoch": 0.05813953488372093, "grad_norm": 4.427198886871338, "learning_rate": 2.880434782608696e-06, "loss": 1.0010372161865235, "step": 160 }, { "epoch": 0.06177325581395349, "grad_norm": 4.012505054473877, "learning_rate": 3.061594202898551e-06, "loss": 0.9743292808532715, "step": 170 }, { "epoch": 0.06540697674418605, "grad_norm": 3.958740472793579, "learning_rate": 3.242753623188406e-06, "loss": 0.9683514595031738, "step": 180 }, { "epoch": 0.0690406976744186, "grad_norm": 3.877817153930664, "learning_rate": 3.4239130434782614e-06, "loss": 0.9519264221191406, "step": 190 }, { "epoch": 0.07267441860465117, "grad_norm": 4.2695841789245605, "learning_rate": 3.6050724637681163e-06, "loss": 0.9914299011230469, "step": 200 }, { "epoch": 0.07630813953488372, "grad_norm": 4.172298908233643, "learning_rate": 3.7862318840579717e-06, "loss": 0.9663248062133789, "step": 210 }, { "epoch": 0.07994186046511628, "grad_norm": 3.850515127182007, "learning_rate": 3.967391304347827e-06, "loss": 0.9332950592041016, "step": 220 }, { "epoch": 0.08357558139534883, "grad_norm": 4.158744812011719, "learning_rate": 4.148550724637682e-06, "loss": 0.9559157371520997, "step": 230 }, { "epoch": 0.0872093023255814, "grad_norm": 3.918304920196533, "learning_rate": 4.329710144927536e-06, "loss": 0.9728614807128906, "step": 240 }, { "epoch": 0.09084302325581395, "grad_norm": 3.7205450534820557, "learning_rate": 4.510869565217392e-06, "loss": 0.9517169952392578, "step": 250 }, { "epoch": 0.09447674418604651, "grad_norm": 4.09927225112915, "learning_rate": 4.692028985507246e-06, "loss": 0.924904727935791, "step": 260 }, { "epoch": 0.09811046511627906, "grad_norm": 3.5890684127807617, "learning_rate": 4.873188405797102e-06, "loss": 0.9166353225708008, "step": 270 }, { "epoch": 0.10174418604651163, "grad_norm": 3.7735586166381836, "learning_rate": 4.999995937615819e-06, "loss": 0.9599678993225098, "step": 280 }, { "epoch": 0.10537790697674419, "grad_norm": 4.090644836425781, "learning_rate": 4.999923717819873e-06, "loss": 0.9784866333007812, "step": 290 }, { "epoch": 0.10901162790697674, "grad_norm": 3.6386289596557617, "learning_rate": 4.999761225821655e-06, "loss": 0.9279603004455567, "step": 300 }, { "epoch": 0.11264534883720931, "grad_norm": 3.862521171569824, "learning_rate": 4.999508467488746e-06, "loss": 0.9072933197021484, "step": 310 }, { "epoch": 0.11627906976744186, "grad_norm": 3.708317756652832, "learning_rate": 4.999165451948247e-06, "loss": 0.9534824371337891, "step": 320 }, { "epoch": 0.11991279069767442, "grad_norm": 3.7813735008239746, "learning_rate": 4.998732191586442e-06, "loss": 0.9028387069702148, "step": 330 }, { "epoch": 0.12354651162790697, "grad_norm": 3.4658145904541016, "learning_rate": 4.998208702048356e-06, "loss": 0.9283824920654297, "step": 340 }, { "epoch": 0.12718023255813954, "grad_norm": 3.554983615875244, "learning_rate": 4.997595002237186e-06, "loss": 0.9344646453857421, "step": 350 }, { "epoch": 0.1308139534883721, "grad_norm": 3.5096824169158936, "learning_rate": 4.9968911143136225e-06, "loss": 0.8727184295654297, "step": 360 }, { "epoch": 0.13444767441860464, "grad_norm": 4.025424957275391, "learning_rate": 4.996097063695047e-06, "loss": 0.9354368209838867, "step": 370 }, { "epoch": 0.1380813953488372, "grad_norm": 3.9307773113250732, "learning_rate": 4.9952128790546115e-06, "loss": 0.9102288246154785, "step": 380 }, { "epoch": 0.14171511627906977, "grad_norm": 3.7353551387786865, "learning_rate": 4.9942385923202096e-06, "loss": 0.8842922210693359, "step": 390 }, { "epoch": 0.14534883720930233, "grad_norm": 3.4209423065185547, "learning_rate": 4.993174238673318e-06, "loss": 0.9148771286010742, "step": 400 }, { "epoch": 0.14898255813953487, "grad_norm": 3.5949199199676514, "learning_rate": 4.9920198565477295e-06, "loss": 0.9135640144348145, "step": 410 }, { "epoch": 0.15261627906976744, "grad_norm": 3.3128015995025635, "learning_rate": 4.990775487628162e-06, "loss": 0.8838287353515625, "step": 420 }, { "epoch": 0.15625, "grad_norm": 3.315894603729248, "learning_rate": 4.989441176848755e-06, "loss": 0.9229852676391601, "step": 430 }, { "epoch": 0.15988372093023256, "grad_norm": 3.24360728263855, "learning_rate": 4.988016972391448e-06, "loss": 0.8780896186828613, "step": 440 }, { "epoch": 0.16351744186046513, "grad_norm": 3.412320613861084, "learning_rate": 4.98650292568424e-06, "loss": 0.9195170402526855, "step": 450 }, { "epoch": 0.16715116279069767, "grad_norm": 3.335376739501953, "learning_rate": 4.984899091399331e-06, "loss": 0.897549057006836, "step": 460 }, { "epoch": 0.17078488372093023, "grad_norm": 3.4164929389953613, "learning_rate": 4.98320552745115e-06, "loss": 0.8903985977172851, "step": 470 }, { "epoch": 0.1744186046511628, "grad_norm": 3.1281023025512695, "learning_rate": 4.981422294994261e-06, "loss": 0.862525749206543, "step": 480 }, { "epoch": 0.17805232558139536, "grad_norm": 3.1380455493927, "learning_rate": 4.9795494584211575e-06, "loss": 0.8879472732543945, "step": 490 }, { "epoch": 0.1816860465116279, "grad_norm": 3.2967255115509033, "learning_rate": 4.9775870853599365e-06, "loss": 0.9352281570434571, "step": 500 }, { "epoch": 0.18531976744186046, "grad_norm": 3.2854440212249756, "learning_rate": 4.975535246671853e-06, "loss": 0.8670001029968262, "step": 510 }, { "epoch": 0.18895348837209303, "grad_norm": 3.250467300415039, "learning_rate": 4.97339401644877e-06, "loss": 0.8876986503601074, "step": 520 }, { "epoch": 0.1925872093023256, "grad_norm": 3.321596622467041, "learning_rate": 4.9711634720104714e-06, "loss": 0.9065715789794921, "step": 530 }, { "epoch": 0.19622093023255813, "grad_norm": 3.3532400131225586, "learning_rate": 4.968843693901879e-06, "loss": 0.9148740768432617, "step": 540 }, { "epoch": 0.1998546511627907, "grad_norm": 3.0000295639038086, "learning_rate": 4.96643476589014e-06, "loss": 0.8900104522705078, "step": 550 }, { "epoch": 0.20348837209302326, "grad_norm": 3.054938554763794, "learning_rate": 4.963936774961605e-06, "loss": 0.8990900993347168, "step": 560 }, { "epoch": 0.20712209302325582, "grad_norm": 3.737675666809082, "learning_rate": 4.961349811318681e-06, "loss": 0.8763975143432617, "step": 570 }, { "epoch": 0.21075581395348839, "grad_norm": 3.054027795791626, "learning_rate": 4.958673968376582e-06, "loss": 0.8707597732543946, "step": 580 }, { "epoch": 0.21438953488372092, "grad_norm": 3.250688314437866, "learning_rate": 4.955909342759949e-06, "loss": 0.8907978057861328, "step": 590 }, { "epoch": 0.2180232558139535, "grad_norm": 3.243490695953369, "learning_rate": 4.953056034299367e-06, "loss": 0.8893247604370117, "step": 600 }, { "epoch": 0.22165697674418605, "grad_norm": 3.19518780708313, "learning_rate": 4.9501141460277536e-06, "loss": 0.8762750625610352, "step": 610 }, { "epoch": 0.22529069767441862, "grad_norm": 3.316575050354004, "learning_rate": 4.9470837841766446e-06, "loss": 0.8776466369628906, "step": 620 }, { "epoch": 0.22892441860465115, "grad_norm": 3.0390732288360596, "learning_rate": 4.943965058172354e-06, "loss": 0.8640607833862305, "step": 630 }, { "epoch": 0.23255813953488372, "grad_norm": 3.0142030715942383, "learning_rate": 4.940758080632023e-06, "loss": 0.8418485641479492, "step": 640 }, { "epoch": 0.23619186046511628, "grad_norm": 3.2164595127105713, "learning_rate": 4.937462967359556e-06, "loss": 0.8713783264160156, "step": 650 }, { "epoch": 0.23982558139534885, "grad_norm": 3.233011484146118, "learning_rate": 4.934079837341436e-06, "loss": 0.8551886558532715, "step": 660 }, { "epoch": 0.24345930232558138, "grad_norm": 2.8924005031585693, "learning_rate": 4.93060881274243e-06, "loss": 0.8672443389892578, "step": 670 }, { "epoch": 0.24709302325581395, "grad_norm": 3.2440953254699707, "learning_rate": 4.927050018901176e-06, "loss": 0.8703802108764649, "step": 680 }, { "epoch": 0.2507267441860465, "grad_norm": 3.142500400543213, "learning_rate": 4.923403584325659e-06, "loss": 0.8739217758178711, "step": 690 }, { "epoch": 0.2543604651162791, "grad_norm": 3.116825580596924, "learning_rate": 4.919669640688567e-06, "loss": 0.848177146911621, "step": 700 }, { "epoch": 0.25799418604651164, "grad_norm": 2.9440503120422363, "learning_rate": 4.91584832282254e-06, "loss": 0.8814225196838379, "step": 710 }, { "epoch": 0.2616279069767442, "grad_norm": 3.111051559448242, "learning_rate": 4.911939768715302e-06, "loss": 0.8624879837036132, "step": 720 }, { "epoch": 0.26526162790697677, "grad_norm": 3.0284619331359863, "learning_rate": 4.907944119504674e-06, "loss": 0.8661399841308594, "step": 730 }, { "epoch": 0.2688953488372093, "grad_norm": 3.13392972946167, "learning_rate": 4.903861519473479e-06, "loss": 0.8655827522277832, "step": 740 }, { "epoch": 0.27252906976744184, "grad_norm": 3.3833975791931152, "learning_rate": 4.899692116044334e-06, "loss": 0.8533415794372559, "step": 750 }, { "epoch": 0.2761627906976744, "grad_norm": 3.6023480892181396, "learning_rate": 4.895436059774325e-06, "loss": 0.8765420913696289, "step": 760 }, { "epoch": 0.279796511627907, "grad_norm": 2.986733913421631, "learning_rate": 4.89109350434957e-06, "loss": 0.8727610588073731, "step": 770 }, { "epoch": 0.28343023255813954, "grad_norm": 3.0566656589508057, "learning_rate": 4.886664606579671e-06, "loss": 0.8449860572814941, "step": 780 }, { "epoch": 0.2870639534883721, "grad_norm": 3.382549524307251, "learning_rate": 4.88214952639205e-06, "loss": 0.8713051795959472, "step": 790 }, { "epoch": 0.29069767441860467, "grad_norm": 3.041335105895996, "learning_rate": 4.877548426826173e-06, "loss": 0.8494526863098144, "step": 800 }, { "epoch": 0.29433139534883723, "grad_norm": 2.935906410217285, "learning_rate": 4.872861474027667e-06, "loss": 0.8570301055908203, "step": 810 }, { "epoch": 0.29796511627906974, "grad_norm": 2.9906163215637207, "learning_rate": 4.868088837242318e-06, "loss": 0.8500057220458984, "step": 820 }, { "epoch": 0.3015988372093023, "grad_norm": 3.039278984069824, "learning_rate": 4.863230688809955e-06, "loss": 0.8411312103271484, "step": 830 }, { "epoch": 0.30523255813953487, "grad_norm": 3.0760066509246826, "learning_rate": 4.858287204158236e-06, "loss": 0.8711600303649902, "step": 840 }, { "epoch": 0.30886627906976744, "grad_norm": 3.267594814300537, "learning_rate": 4.853258561796307e-06, "loss": 0.8397054672241211, "step": 850 }, { "epoch": 0.3125, "grad_norm": 2.9856038093566895, "learning_rate": 4.848144943308355e-06, "loss": 0.8396913528442382, "step": 860 }, { "epoch": 0.31613372093023256, "grad_norm": 2.778498888015747, "learning_rate": 4.842946533347056e-06, "loss": 0.8382930755615234, "step": 870 }, { "epoch": 0.31976744186046513, "grad_norm": 3.0495941638946533, "learning_rate": 4.837663519626904e-06, "loss": 0.8362157821655274, "step": 880 }, { "epoch": 0.3234011627906977, "grad_norm": 3.0766844749450684, "learning_rate": 4.8322960929174325e-06, "loss": 0.8268529891967773, "step": 890 }, { "epoch": 0.32703488372093026, "grad_norm": 3.0890631675720215, "learning_rate": 4.826844447036329e-06, "loss": 0.8377861022949219, "step": 900 }, { "epoch": 0.33066860465116277, "grad_norm": 2.9986603260040283, "learning_rate": 4.82130877884243e-06, "loss": 0.8358505249023438, "step": 910 }, { "epoch": 0.33430232558139533, "grad_norm": 2.8856871128082275, "learning_rate": 4.8156892882286206e-06, "loss": 0.8571210861206054, "step": 920 }, { "epoch": 0.3379360465116279, "grad_norm": 3.0375850200653076, "learning_rate": 4.809986178114608e-06, "loss": 0.8279001235961914, "step": 930 }, { "epoch": 0.34156976744186046, "grad_norm": 2.7761526107788086, "learning_rate": 4.804199654439601e-06, "loss": 0.8522439956665039, "step": 940 }, { "epoch": 0.345203488372093, "grad_norm": 3.0612213611602783, "learning_rate": 4.798329926154868e-06, "loss": 0.843962574005127, "step": 950 }, { "epoch": 0.3488372093023256, "grad_norm": 3.3076977729797363, "learning_rate": 4.792377205216198e-06, "loss": 0.8473945617675781, "step": 960 }, { "epoch": 0.35247093023255816, "grad_norm": 2.7613766193389893, "learning_rate": 4.78634170657624e-06, "loss": 0.8338375091552734, "step": 970 }, { "epoch": 0.3561046511627907, "grad_norm": 2.9154269695281982, "learning_rate": 4.780223648176746e-06, "loss": 0.8274489402770996, "step": 980 }, { "epoch": 0.35973837209302323, "grad_norm": 2.8659121990203857, "learning_rate": 4.774023250940699e-06, "loss": 0.8512365341186523, "step": 990 }, { "epoch": 0.3633720930232558, "grad_norm": 2.993288516998291, "learning_rate": 4.767740738764337e-06, "loss": 0.8384992599487304, "step": 1000 }, { "epoch": 0.36700581395348836, "grad_norm": 2.8574283123016357, "learning_rate": 4.761376338509065e-06, "loss": 0.8402500152587891, "step": 1010 }, { "epoch": 0.3706395348837209, "grad_norm": 3.0336382389068604, "learning_rate": 4.754930279993266e-06, "loss": 0.8413254737854003, "step": 1020 }, { "epoch": 0.3742732558139535, "grad_norm": 2.7382609844207764, "learning_rate": 4.748402795984004e-06, "loss": 0.8280426025390625, "step": 1030 }, { "epoch": 0.37790697674418605, "grad_norm": 2.9250476360321045, "learning_rate": 4.741794122188611e-06, "loss": 0.8249115943908691, "step": 1040 }, { "epoch": 0.3815406976744186, "grad_norm": 2.767603874206543, "learning_rate": 4.735104497246185e-06, "loss": 0.8396032333374024, "step": 1050 }, { "epoch": 0.3851744186046512, "grad_norm": 2.853214740753174, "learning_rate": 4.728334162718967e-06, "loss": 0.8256677627563477, "step": 1060 }, { "epoch": 0.38880813953488375, "grad_norm": 2.9439098834991455, "learning_rate": 4.72148336308362e-06, "loss": 0.8689069747924805, "step": 1070 }, { "epoch": 0.39244186046511625, "grad_norm": 2.757331371307373, "learning_rate": 4.7145523457224005e-06, "loss": 0.8378227233886719, "step": 1080 }, { "epoch": 0.3960755813953488, "grad_norm": 2.884744882583618, "learning_rate": 4.707541360914225e-06, "loss": 0.8407644271850586, "step": 1090 }, { "epoch": 0.3997093023255814, "grad_norm": 2.771481513977051, "learning_rate": 4.700450661825635e-06, "loss": 0.8069744110107422, "step": 1100 }, { "epoch": 0.40334302325581395, "grad_norm": 2.8814430236816406, "learning_rate": 4.693280504501651e-06, "loss": 0.8310678482055665, "step": 1110 }, { "epoch": 0.4069767441860465, "grad_norm": 2.7411293983459473, "learning_rate": 4.686031147856531e-06, "loss": 0.8619399070739746, "step": 1120 }, { "epoch": 0.4106104651162791, "grad_norm": 2.7865939140319824, "learning_rate": 4.678702853664418e-06, "loss": 0.8180102348327637, "step": 1130 }, { "epoch": 0.41424418604651164, "grad_norm": 2.827937602996826, "learning_rate": 4.671295886549887e-06, "loss": 0.8340296745300293, "step": 1140 }, { "epoch": 0.4178779069767442, "grad_norm": 2.8981363773345947, "learning_rate": 4.6638105139783955e-06, "loss": 0.8189412117004394, "step": 1150 }, { "epoch": 0.42151162790697677, "grad_norm": 2.6089940071105957, "learning_rate": 4.656247006246617e-06, "loss": 0.8204338073730468, "step": 1160 }, { "epoch": 0.4251453488372093, "grad_norm": 2.7830424308776855, "learning_rate": 4.648605636472683e-06, "loss": 0.8322567939758301, "step": 1170 }, { "epoch": 0.42877906976744184, "grad_norm": 2.8253374099731445, "learning_rate": 4.640886680586326e-06, "loss": 0.7853239059448243, "step": 1180 }, { "epoch": 0.4324127906976744, "grad_norm": 2.7395074367523193, "learning_rate": 4.6330904173189105e-06, "loss": 0.8274772644042969, "step": 1190 }, { "epoch": 0.436046511627907, "grad_norm": 2.8378896713256836, "learning_rate": 4.625217128193366e-06, "loss": 0.8572607040405273, "step": 1200 }, { "epoch": 0.43968023255813954, "grad_norm": 2.8448286056518555, "learning_rate": 4.617267097514028e-06, "loss": 0.822886848449707, "step": 1210 }, { "epoch": 0.4433139534883721, "grad_norm": 2.8581748008728027, "learning_rate": 4.609240612356367e-06, "loss": 0.7989314556121826, "step": 1220 }, { "epoch": 0.44694767441860467, "grad_norm": 2.6748061180114746, "learning_rate": 4.6011379625566245e-06, "loss": 0.8521600723266601, "step": 1230 }, { "epoch": 0.45058139534883723, "grad_norm": 2.6777000427246094, "learning_rate": 4.5929594407013434e-06, "loss": 0.8159719467163086, "step": 1240 }, { "epoch": 0.45421511627906974, "grad_norm": 2.781126022338867, "learning_rate": 4.584705342116807e-06, "loss": 0.8263021469116211, "step": 1250 }, { "epoch": 0.4578488372093023, "grad_norm": 3.040696144104004, "learning_rate": 4.576375964858373e-06, "loss": 0.7871081352233886, "step": 1260 }, { "epoch": 0.46148255813953487, "grad_norm": 2.9158823490142822, "learning_rate": 4.567971609699709e-06, "loss": 0.828523826599121, "step": 1270 }, { "epoch": 0.46511627906976744, "grad_norm": 2.79797625541687, "learning_rate": 4.559492580121936e-06, "loss": 0.8283788681030273, "step": 1280 }, { "epoch": 0.46875, "grad_norm": 2.6865875720977783, "learning_rate": 4.550939182302666e-06, "loss": 0.8271783828735352, "step": 1290 }, { "epoch": 0.47238372093023256, "grad_norm": 2.9567787647247314, "learning_rate": 4.5423117251049455e-06, "loss": 0.8017131805419921, "step": 1300 }, { "epoch": 0.47601744186046513, "grad_norm": 2.9435689449310303, "learning_rate": 4.533610520066104e-06, "loss": 0.8426065444946289, "step": 1310 }, { "epoch": 0.4796511627906977, "grad_norm": 3.003634452819824, "learning_rate": 4.524835881386508e-06, "loss": 0.8319045066833496, "step": 1320 }, { "epoch": 0.48328488372093026, "grad_norm": 2.7824082374572754, "learning_rate": 4.515988125918206e-06, "loss": 0.8165316581726074, "step": 1330 }, { "epoch": 0.48691860465116277, "grad_norm": 2.6757547855377197, "learning_rate": 4.507067573153496e-06, "loss": 0.8275221824645996, "step": 1340 }, { "epoch": 0.49055232558139533, "grad_norm": 2.7333762645721436, "learning_rate": 4.498074545213386e-06, "loss": 0.7960930824279785, "step": 1350 }, { "epoch": 0.4941860465116279, "grad_norm": 2.8366189002990723, "learning_rate": 4.489009366835956e-06, "loss": 0.8204254150390625, "step": 1360 }, { "epoch": 0.49781976744186046, "grad_norm": 2.8708670139312744, "learning_rate": 4.479872365364644e-06, "loss": 0.7996141910552979, "step": 1370 }, { "epoch": 0.501453488372093, "grad_norm": 2.927445411682129, "learning_rate": 4.470663870736414e-06, "loss": 0.7884037017822265, "step": 1380 }, { "epoch": 0.5050872093023255, "grad_norm": 2.7692763805389404, "learning_rate": 4.461384215469847e-06, "loss": 0.8215972900390625, "step": 1390 }, { "epoch": 0.5087209302325582, "grad_norm": 2.714712619781494, "learning_rate": 4.452033734653136e-06, "loss": 0.8339604377746582, "step": 1400 }, { "epoch": 0.5123546511627907, "grad_norm": 2.8098700046539307, "learning_rate": 4.442612765931981e-06, "loss": 0.819193172454834, "step": 1410 }, { "epoch": 0.5159883720930233, "grad_norm": 2.9055378437042236, "learning_rate": 4.433121649497401e-06, "loss": 0.8303899765014648, "step": 1420 }, { "epoch": 0.5196220930232558, "grad_norm": 2.8050930500030518, "learning_rate": 4.423560728073444e-06, "loss": 0.8256185531616211, "step": 1430 }, { "epoch": 0.5232558139534884, "grad_norm": 2.776355266571045, "learning_rate": 4.413930346904819e-06, "loss": 0.7909624576568604, "step": 1440 }, { "epoch": 0.5268895348837209, "grad_norm": 2.799743175506592, "learning_rate": 4.404230853744424e-06, "loss": 0.8219989776611328, "step": 1450 }, { "epoch": 0.5305232558139535, "grad_norm": 2.6980018615722656, "learning_rate": 4.3944625988407896e-06, "loss": 0.8094476699829102, "step": 1460 }, { "epoch": 0.534156976744186, "grad_norm": 2.7240099906921387, "learning_rate": 4.384625934925431e-06, "loss": 0.7854260444641114, "step": 1470 }, { "epoch": 0.5377906976744186, "grad_norm": 2.8290176391601562, "learning_rate": 4.374721217200113e-06, "loss": 0.7866772174835205, "step": 1480 }, { "epoch": 0.5414244186046512, "grad_norm": 2.822491407394409, "learning_rate": 4.364748803324021e-06, "loss": 0.8133050918579101, "step": 1490 }, { "epoch": 0.5450581395348837, "grad_norm": 2.792834520339966, "learning_rate": 4.354709053400846e-06, "loss": 0.8320375442504883, "step": 1500 }, { "epoch": 0.5486918604651163, "grad_norm": 2.802419900894165, "learning_rate": 4.344602329965788e-06, "loss": 0.7835256576538085, "step": 1510 }, { "epoch": 0.5523255813953488, "grad_norm": 2.7649500370025635, "learning_rate": 4.334428997972454e-06, "loss": 0.8077914237976074, "step": 1520 }, { "epoch": 0.5559593023255814, "grad_norm": 2.6321909427642822, "learning_rate": 4.324189424779686e-06, "loss": 0.7951157093048096, "step": 1530 }, { "epoch": 0.559593023255814, "grad_norm": 2.90681529045105, "learning_rate": 4.313883980138296e-06, "loss": 0.7936089038848877, "step": 1540 }, { "epoch": 0.5632267441860465, "grad_norm": 2.783231258392334, "learning_rate": 4.303513036177711e-06, "loss": 0.8397782325744629, "step": 1550 }, { "epoch": 0.5668604651162791, "grad_norm": 2.735781669616699, "learning_rate": 4.293076967392539e-06, "loss": 0.7950201034545898, "step": 1560 }, { "epoch": 0.5704941860465116, "grad_norm": 2.7377381324768066, "learning_rate": 4.282576150629044e-06, "loss": 0.8493624687194824, "step": 1570 }, { "epoch": 0.5741279069767442, "grad_norm": 2.8512473106384277, "learning_rate": 4.272010965071538e-06, "loss": 0.8232376098632812, "step": 1580 }, { "epoch": 0.5777616279069767, "grad_norm": 2.664661407470703, "learning_rate": 4.2613817922286875e-06, "loss": 0.8027572631835938, "step": 1590 }, { "epoch": 0.5813953488372093, "grad_norm": 2.655877113342285, "learning_rate": 4.250689015919741e-06, "loss": 0.7969479560852051, "step": 1600 }, { "epoch": 0.5850290697674418, "grad_norm": 2.7822320461273193, "learning_rate": 4.239933022260668e-06, "loss": 0.83875732421875, "step": 1610 }, { "epoch": 0.5886627906976745, "grad_norm": 2.707385778427124, "learning_rate": 4.2291141996502115e-06, "loss": 0.7893309593200684, "step": 1620 }, { "epoch": 0.592296511627907, "grad_norm": 2.7490038871765137, "learning_rate": 4.21823293875587e-06, "loss": 0.8146431922912598, "step": 1630 }, { "epoch": 0.5959302325581395, "grad_norm": 2.7448737621307373, "learning_rate": 4.207289632499787e-06, "loss": 0.8016767501831055, "step": 1640 }, { "epoch": 0.5995639534883721, "grad_norm": 2.762751817703247, "learning_rate": 4.1962846760445606e-06, "loss": 0.7814556121826172, "step": 1650 }, { "epoch": 0.6031976744186046, "grad_norm": 2.8780152797698975, "learning_rate": 4.185218466778981e-06, "loss": 0.8208811759948731, "step": 1660 }, { "epoch": 0.6068313953488372, "grad_norm": 2.860222816467285, "learning_rate": 4.174091404303671e-06, "loss": 0.8105660438537597, "step": 1670 }, { "epoch": 0.6104651162790697, "grad_norm": 2.866760730743408, "learning_rate": 4.162903890416666e-06, "loss": 0.8159380912780761, "step": 1680 }, { "epoch": 0.6140988372093024, "grad_norm": 2.6297552585601807, "learning_rate": 4.151656329098899e-06, "loss": 0.7768171310424805, "step": 1690 }, { "epoch": 0.6177325581395349, "grad_norm": 2.73301100730896, "learning_rate": 4.1403491264996144e-06, "loss": 0.7993753910064697, "step": 1700 }, { "epoch": 0.6213662790697675, "grad_norm": 2.8257224559783936, "learning_rate": 4.1289826909217034e-06, "loss": 0.7895632266998291, "step": 1710 }, { "epoch": 0.625, "grad_norm": 2.745051860809326, "learning_rate": 4.117557432806958e-06, "loss": 0.8250743865966796, "step": 1720 }, { "epoch": 0.6286337209302325, "grad_norm": 2.557063102722168, "learning_rate": 4.106073764721252e-06, "loss": 0.8024697303771973, "step": 1730 }, { "epoch": 0.6322674418604651, "grad_norm": 2.8046607971191406, "learning_rate": 4.094532101339641e-06, "loss": 0.7836231231689453, "step": 1740 }, { "epoch": 0.6359011627906976, "grad_norm": 2.7214012145996094, "learning_rate": 4.08293285943139e-06, "loss": 0.8137659072875977, "step": 1750 }, { "epoch": 0.6395348837209303, "grad_norm": 2.6622800827026367, "learning_rate": 4.071276457844925e-06, "loss": 0.817494010925293, "step": 1760 }, { "epoch": 0.6431686046511628, "grad_norm": 2.906050443649292, "learning_rate": 4.059563317492705e-06, "loss": 0.7947888851165772, "step": 1770 }, { "epoch": 0.6468023255813954, "grad_norm": 2.7529661655426025, "learning_rate": 4.0477938613360265e-06, "loss": 0.7987423896789551, "step": 1780 }, { "epoch": 0.6504360465116279, "grad_norm": 2.8821957111358643, "learning_rate": 4.035968514369748e-06, "loss": 0.8163650512695313, "step": 1790 }, { "epoch": 0.6540697674418605, "grad_norm": 2.667881727218628, "learning_rate": 4.024087703606945e-06, "loss": 0.8074198722839355, "step": 1800 }, { "epoch": 0.657703488372093, "grad_norm": 2.6779751777648926, "learning_rate": 4.012151858063487e-06, "loss": 0.7849567413330079, "step": 1810 }, { "epoch": 0.6613372093023255, "grad_norm": 2.7572410106658936, "learning_rate": 4.0001614087425515e-06, "loss": 0.8169970512390137, "step": 1820 }, { "epoch": 0.6649709302325582, "grad_norm": 2.729264736175537, "learning_rate": 3.988116788619055e-06, "loss": 0.7919125080108642, "step": 1830 }, { "epoch": 0.6686046511627907, "grad_norm": 2.716473340988159, "learning_rate": 3.976018432624025e-06, "loss": 0.8070649147033692, "step": 1840 }, { "epoch": 0.6722383720930233, "grad_norm": 2.732853651046753, "learning_rate": 3.963866777628885e-06, "loss": 0.7902655601501465, "step": 1850 }, { "epoch": 0.6758720930232558, "grad_norm": 2.834977865219116, "learning_rate": 3.95166226242969e-06, "loss": 0.7799157619476318, "step": 1860 }, { "epoch": 0.6795058139534884, "grad_norm": 2.7858405113220215, "learning_rate": 3.939405327731268e-06, "loss": 0.8153091430664062, "step": 1870 }, { "epoch": 0.6831395348837209, "grad_norm": 2.7912445068359375, "learning_rate": 3.927096416131324e-06, "loss": 0.7729724884033203, "step": 1880 }, { "epoch": 0.6867732558139535, "grad_norm": 2.8135998249053955, "learning_rate": 3.914735972104442e-06, "loss": 0.798687744140625, "step": 1890 }, { "epoch": 0.690406976744186, "grad_norm": 2.7605013847351074, "learning_rate": 3.902324441986041e-06, "loss": 0.8077116012573242, "step": 1900 }, { "epoch": 0.6940406976744186, "grad_norm": 2.720306158065796, "learning_rate": 3.889862273956261e-06, "loss": 0.7985315799713135, "step": 1910 }, { "epoch": 0.6976744186046512, "grad_norm": 2.7768609523773193, "learning_rate": 3.877349918023772e-06, "loss": 0.7771415710449219, "step": 1920 }, { "epoch": 0.7013081395348837, "grad_norm": 2.8074111938476562, "learning_rate": 3.864787826009533e-06, "loss": 0.7958961486816406, "step": 1930 }, { "epoch": 0.7049418604651163, "grad_norm": 2.90981388092041, "learning_rate": 3.852176451530468e-06, "loss": 0.7736949920654297, "step": 1940 }, { "epoch": 0.7085755813953488, "grad_norm": 2.7982535362243652, "learning_rate": 3.839516249983093e-06, "loss": 0.751534366607666, "step": 1950 }, { "epoch": 0.7122093023255814, "grad_norm": 2.6651105880737305, "learning_rate": 3.826807678527067e-06, "loss": 0.8033027648925781, "step": 1960 }, { "epoch": 0.715843023255814, "grad_norm": 2.684173107147217, "learning_rate": 3.814051196068684e-06, "loss": 0.7902488708496094, "step": 1970 }, { "epoch": 0.7194767441860465, "grad_norm": 2.5975940227508545, "learning_rate": 3.8012472632443065e-06, "loss": 0.7830523490905762, "step": 1980 }, { "epoch": 0.7231104651162791, "grad_norm": 3.0428762435913086, "learning_rate": 3.7883963424037286e-06, "loss": 0.7818472385406494, "step": 1990 }, { "epoch": 0.7267441860465116, "grad_norm": 2.6959846019744873, "learning_rate": 3.775498897593478e-06, "loss": 0.8127883911132813, "step": 2000 }, { "epoch": 0.7303779069767442, "grad_norm": 2.730274200439453, "learning_rate": 3.762555394540066e-06, "loss": 0.7654043674468994, "step": 2010 }, { "epoch": 0.7340116279069767, "grad_norm": 2.6885862350463867, "learning_rate": 3.7495663006331633e-06, "loss": 0.7917684555053711, "step": 2020 }, { "epoch": 0.7376453488372093, "grad_norm": 2.7472004890441895, "learning_rate": 3.7365320849087262e-06, "loss": 0.7998331546783447, "step": 2030 }, { "epoch": 0.7412790697674418, "grad_norm": 2.767969846725464, "learning_rate": 3.7234532180320605e-06, "loss": 0.7952752590179444, "step": 2040 }, { "epoch": 0.7449127906976745, "grad_norm": 2.6142473220825195, "learning_rate": 3.710330172280821e-06, "loss": 0.8077480316162109, "step": 2050 }, { "epoch": 0.748546511627907, "grad_norm": 2.7542755603790283, "learning_rate": 3.6971634215279664e-06, "loss": 0.7876755237579346, "step": 2060 }, { "epoch": 0.7521802325581395, "grad_norm": 2.6669540405273438, "learning_rate": 3.683953441224636e-06, "loss": 0.7639244079589844, "step": 2070 }, { "epoch": 0.7558139534883721, "grad_norm": 2.7806105613708496, "learning_rate": 3.6707007083829913e-06, "loss": 0.8007266998291016, "step": 2080 }, { "epoch": 0.7594476744186046, "grad_norm": 2.7667555809020996, "learning_rate": 3.6574057015589877e-06, "loss": 0.7894309520721435, "step": 2090 }, { "epoch": 0.7630813953488372, "grad_norm": 2.709882974624634, "learning_rate": 3.6440689008350904e-06, "loss": 0.7834427356719971, "step": 2100 }, { "epoch": 0.7667151162790697, "grad_norm": 2.6989541053771973, "learning_rate": 3.630690787802944e-06, "loss": 0.8200445175170898, "step": 2110 }, { "epoch": 0.7703488372093024, "grad_norm": 2.769083261489868, "learning_rate": 3.6172718455459783e-06, "loss": 0.7756868362426758, "step": 2120 }, { "epoch": 0.7739825581395349, "grad_norm": 2.8316144943237305, "learning_rate": 3.6038125586219663e-06, "loss": 0.7860795974731445, "step": 2130 }, { "epoch": 0.7776162790697675, "grad_norm": 2.8543038368225098, "learning_rate": 3.5903134130455285e-06, "loss": 0.7748101234436036, "step": 2140 }, { "epoch": 0.78125, "grad_norm": 3.5973920822143555, "learning_rate": 3.5767748962705756e-06, "loss": 0.7883114814758301, "step": 2150 }, { "epoch": 0.7848837209302325, "grad_norm": 2.770141124725342, "learning_rate": 3.563197497172717e-06, "loss": 0.8006394386291504, "step": 2160 }, { "epoch": 0.7885174418604651, "grad_norm": 2.8230247497558594, "learning_rate": 3.5495817060315975e-06, "loss": 0.7962004661560058, "step": 2170 }, { "epoch": 0.7921511627906976, "grad_norm": 2.885927677154541, "learning_rate": 3.535928014513201e-06, "loss": 0.7753909587860107, "step": 2180 }, { "epoch": 0.7957848837209303, "grad_norm": 2.7065513134002686, "learning_rate": 3.5222369156520927e-06, "loss": 0.7820100784301758, "step": 2190 }, { "epoch": 0.7994186046511628, "grad_norm": 2.6883857250213623, "learning_rate": 3.5085089038336157e-06, "loss": 0.7736979007720948, "step": 2200 }, { "epoch": 0.8030523255813954, "grad_norm": 2.665839195251465, "learning_rate": 3.49474447477604e-06, "loss": 0.7771726131439209, "step": 2210 }, { "epoch": 0.8066860465116279, "grad_norm": 2.775423765182495, "learning_rate": 3.4809441255126598e-06, "loss": 0.7675837993621826, "step": 2220 }, { "epoch": 0.8103197674418605, "grad_norm": 2.7208096981048584, "learning_rate": 3.4671083543738505e-06, "loss": 0.7966558456420898, "step": 2230 }, { "epoch": 0.813953488372093, "grad_norm": 2.8475911617279053, "learning_rate": 3.453237660969068e-06, "loss": 0.772910213470459, "step": 2240 }, { "epoch": 0.8175872093023255, "grad_norm": 2.780776023864746, "learning_rate": 3.439332546168812e-06, "loss": 0.7731334209442139, "step": 2250 }, { "epoch": 0.8212209302325582, "grad_norm": 2.746427536010742, "learning_rate": 3.425393512086538e-06, "loss": 0.7510773658752441, "step": 2260 }, { "epoch": 0.8248546511627907, "grad_norm": 2.705579996109009, "learning_rate": 3.4114210620605248e-06, "loss": 0.7795310974121094, "step": 2270 }, { "epoch": 0.8284883720930233, "grad_norm": 2.866590738296509, "learning_rate": 3.3974157006357013e-06, "loss": 0.7611992835998536, "step": 2280 }, { "epoch": 0.8321220930232558, "grad_norm": 2.670555591583252, "learning_rate": 3.383377933545429e-06, "loss": 0.808956527709961, "step": 2290 }, { "epoch": 0.8357558139534884, "grad_norm": 2.7718398571014404, "learning_rate": 3.3693082676932326e-06, "loss": 0.7825285434722901, "step": 2300 }, { "epoch": 0.8393895348837209, "grad_norm": 2.710367441177368, "learning_rate": 3.3552072111345047e-06, "loss": 0.8079341888427735, "step": 2310 }, { "epoch": 0.8430232558139535, "grad_norm": 2.7639403343200684, "learning_rate": 3.341075273058153e-06, "loss": 0.7904543876647949, "step": 2320 }, { "epoch": 0.846656976744186, "grad_norm": 2.618574380874634, "learning_rate": 3.3269129637682178e-06, "loss": 0.79985671043396, "step": 2330 }, { "epoch": 0.8502906976744186, "grad_norm": 2.6449429988861084, "learning_rate": 3.3127207946654434e-06, "loss": 0.7599962234497071, "step": 2340 }, { "epoch": 0.8539244186046512, "grad_norm": 2.722358226776123, "learning_rate": 3.29849927822881e-06, "loss": 0.7602685451507568, "step": 2350 }, { "epoch": 0.8575581395348837, "grad_norm": 2.695909023284912, "learning_rate": 3.2842489279970314e-06, "loss": 0.7823375701904297, "step": 2360 }, { "epoch": 0.8611918604651163, "grad_norm": 2.84134840965271, "learning_rate": 3.2699702585500072e-06, "loss": 0.8109968185424805, "step": 2370 }, { "epoch": 0.8648255813953488, "grad_norm": 2.819700241088867, "learning_rate": 3.2556637854902456e-06, "loss": 0.7682777404785156, "step": 2380 }, { "epoch": 0.8684593023255814, "grad_norm": 2.7188804149627686, "learning_rate": 3.2413300254242417e-06, "loss": 0.7601710796356201, "step": 2390 }, { "epoch": 0.872093023255814, "grad_norm": 2.7124922275543213, "learning_rate": 3.2269694959438236e-06, "loss": 0.7514790534973145, "step": 2400 }, { "epoch": 0.8757267441860465, "grad_norm": 2.796386241912842, "learning_rate": 3.212582715607463e-06, "loss": 0.8110128402709961, "step": 2410 }, { "epoch": 0.8793604651162791, "grad_norm": 2.6567811965942383, "learning_rate": 3.1981702039215483e-06, "loss": 0.7716230392456055, "step": 2420 }, { "epoch": 0.8829941860465116, "grad_norm": 2.7610013484954834, "learning_rate": 3.183732481321629e-06, "loss": 0.7623504638671875, "step": 2430 }, { "epoch": 0.8866279069767442, "grad_norm": 2.900007963180542, "learning_rate": 3.1692700691536184e-06, "loss": 0.777961540222168, "step": 2440 }, { "epoch": 0.8902616279069767, "grad_norm": 2.6529910564422607, "learning_rate": 3.154783489654969e-06, "loss": 0.7695446491241456, "step": 2450 }, { "epoch": 0.8938953488372093, "grad_norm": 2.8663878440856934, "learning_rate": 3.1402732659358187e-06, "loss": 0.7680397033691406, "step": 2460 }, { "epoch": 0.8975290697674418, "grad_norm": 2.582566022872925, "learning_rate": 3.1257399219600926e-06, "loss": 0.7840737342834473, "step": 2470 }, { "epoch": 0.9011627906976745, "grad_norm": 2.6422550678253174, "learning_rate": 3.1111839825265933e-06, "loss": 0.7447628974914551, "step": 2480 }, { "epoch": 0.904796511627907, "grad_norm": 2.8522748947143555, "learning_rate": 3.0966059732500396e-06, "loss": 0.7944939136505127, "step": 2490 }, { "epoch": 0.9084302325581395, "grad_norm": 2.84637188911438, "learning_rate": 3.0820064205420967e-06, "loss": 0.7697494506835938, "step": 2500 }, { "epoch": 0.9120639534883721, "grad_norm": 2.6845076084136963, "learning_rate": 3.0673858515923603e-06, "loss": 0.7460027694702148, "step": 2510 }, { "epoch": 0.9156976744186046, "grad_norm": 2.8455183506011963, "learning_rate": 3.052744794349323e-06, "loss": 0.7743155479431152, "step": 2520 }, { "epoch": 0.9193313953488372, "grad_norm": 2.6090633869171143, "learning_rate": 3.0380837775013072e-06, "loss": 0.7766456604003906, "step": 2530 }, { "epoch": 0.9229651162790697, "grad_norm": 2.987428903579712, "learning_rate": 3.0234033304573797e-06, "loss": 0.7843479633331298, "step": 2540 }, { "epoch": 0.9265988372093024, "grad_norm": 2.794109582901001, "learning_rate": 3.0087039833282293e-06, "loss": 0.7880901336669922, "step": 2550 }, { "epoch": 0.9302325581395349, "grad_norm": 2.7973132133483887, "learning_rate": 2.9939862669070274e-06, "loss": 0.7407669067382813, "step": 2560 }, { "epoch": 0.9338662790697675, "grad_norm": 2.6931514739990234, "learning_rate": 2.9792507126502595e-06, "loss": 0.7737750053405762, "step": 2570 }, { "epoch": 0.9375, "grad_norm": 2.7191874980926514, "learning_rate": 2.964497852658535e-06, "loss": 0.7814858436584473, "step": 2580 }, { "epoch": 0.9411337209302325, "grad_norm": 2.7388243675231934, "learning_rate": 2.9497282196573733e-06, "loss": 0.7539449214935303, "step": 2590 }, { "epoch": 0.9447674418604651, "grad_norm": 2.594496250152588, "learning_rate": 2.934942346977967e-06, "loss": 0.7599133491516114, "step": 2600 }, { "epoch": 0.9484011627906976, "grad_norm": 2.6111090183258057, "learning_rate": 2.9201407685379242e-06, "loss": 0.7895750999450684, "step": 2610 }, { "epoch": 0.9520348837209303, "grad_norm": 2.7242159843444824, "learning_rate": 2.905324018821985e-06, "loss": 0.7981988430023194, "step": 2620 }, { "epoch": 0.9556686046511628, "grad_norm": 2.847475290298462, "learning_rate": 2.890492632862728e-06, "loss": 0.7751128196716308, "step": 2630 }, { "epoch": 0.9593023255813954, "grad_norm": 2.7806103229522705, "learning_rate": 2.875647146221241e-06, "loss": 0.7592604637145997, "step": 2640 }, { "epoch": 0.9629360465116279, "grad_norm": 3.4675211906433105, "learning_rate": 2.860788094967793e-06, "loss": 0.7890193939208985, "step": 2650 }, { "epoch": 0.9665697674418605, "grad_norm": 2.7082395553588867, "learning_rate": 2.845916015662467e-06, "loss": 0.7717397212982178, "step": 2660 }, { "epoch": 0.970203488372093, "grad_norm": 2.694770574569702, "learning_rate": 2.8310314453357902e-06, "loss": 0.7812085151672363, "step": 2670 }, { "epoch": 0.9738372093023255, "grad_norm": 2.8651087284088135, "learning_rate": 2.816134921469339e-06, "loss": 0.794677734375, "step": 2680 }, { "epoch": 0.9774709302325582, "grad_norm": 2.712221384048462, "learning_rate": 2.8012269819763337e-06, "loss": 0.7712698459625245, "step": 2690 }, { "epoch": 0.9811046511627907, "grad_norm": 2.781355381011963, "learning_rate": 2.7863081651822122e-06, "loss": 0.7705037593841553, "step": 2700 }, { "epoch": 0.9847383720930233, "grad_norm": 2.675858974456787, "learning_rate": 2.7713790098051906e-06, "loss": 0.7476750373840332, "step": 2710 }, { "epoch": 0.9883720930232558, "grad_norm": 2.681861639022827, "learning_rate": 2.7564400549368127e-06, "loss": 0.7630274295806885, "step": 2720 }, { "epoch": 0.9920058139534884, "grad_norm": 2.920302629470825, "learning_rate": 2.74149184002248e-06, "loss": 0.7798184871673584, "step": 2730 }, { "epoch": 0.9956395348837209, "grad_norm": 2.8120718002319336, "learning_rate": 2.726534904841974e-06, "loss": 0.781275749206543, "step": 2740 }, { "epoch": 0.9992732558139535, "grad_norm": 2.6712262630462646, "learning_rate": 2.7115697894899675e-06, "loss": 0.7707889556884766, "step": 2750 }, { "epoch": 1.002906976744186, "grad_norm": 2.6090970039367676, "learning_rate": 2.696597034356517e-06, "loss": 0.6963653564453125, "step": 2760 }, { "epoch": 1.0065406976744187, "grad_norm": 2.7611663341522217, "learning_rate": 2.6816171801075507e-06, "loss": 0.6414544105529785, "step": 2770 }, { "epoch": 1.010174418604651, "grad_norm": 2.7799036502838135, "learning_rate": 2.6666307676653485e-06, "loss": 0.6255013465881347, "step": 2780 }, { "epoch": 1.0138081395348837, "grad_norm": 2.7116854190826416, "learning_rate": 2.651638338189003e-06, "loss": 0.6427628517150878, "step": 2790 }, { "epoch": 1.0174418604651163, "grad_norm": 2.8874623775482178, "learning_rate": 2.6366404330548847e-06, "loss": 0.6351820945739746, "step": 2800 }, { "epoch": 1.021075581395349, "grad_norm": 2.93046498298645, "learning_rate": 2.6216375938370887e-06, "loss": 0.6491835117340088, "step": 2810 }, { "epoch": 1.0247093023255813, "grad_norm": 2.82824969291687, "learning_rate": 2.6066303622878802e-06, "loss": 0.6247024536132812, "step": 2820 }, { "epoch": 1.028343023255814, "grad_norm": 2.7935800552368164, "learning_rate": 2.591619280318131e-06, "loss": 0.6559598922729493, "step": 2830 }, { "epoch": 1.0319767441860466, "grad_norm": 2.8328874111175537, "learning_rate": 2.5766048899777524e-06, "loss": 0.6576976776123047, "step": 2840 }, { "epoch": 1.035610465116279, "grad_norm": 2.6769254207611084, "learning_rate": 2.561587733436119e-06, "loss": 0.6029273986816406, "step": 2850 }, { "epoch": 1.0392441860465116, "grad_norm": 2.838825225830078, "learning_rate": 2.5465683529624953e-06, "loss": 0.6157190322875976, "step": 2860 }, { "epoch": 1.0428779069767442, "grad_norm": 2.6545779705047607, "learning_rate": 2.5315472909064502e-06, "loss": 0.633302640914917, "step": 2870 }, { "epoch": 1.0465116279069768, "grad_norm": 2.840080976486206, "learning_rate": 2.5165250896782763e-06, "loss": 0.6440884113311768, "step": 2880 }, { "epoch": 1.0501453488372092, "grad_norm": 2.7946488857269287, "learning_rate": 2.5015022917293983e-06, "loss": 0.6195040702819824, "step": 2890 }, { "epoch": 1.0537790697674418, "grad_norm": 2.7816784381866455, "learning_rate": 2.486479439532793e-06, "loss": 0.6428169250488281, "step": 2900 }, { "epoch": 1.0574127906976745, "grad_norm": 2.7808825969696045, "learning_rate": 2.4714570755633915e-06, "loss": 0.6089080810546875, "step": 2910 }, { "epoch": 1.0610465116279069, "grad_norm": 2.9815218448638916, "learning_rate": 2.456435742278499e-06, "loss": 0.6232911109924316, "step": 2920 }, { "epoch": 1.0646802325581395, "grad_norm": 2.8019156455993652, "learning_rate": 2.441415982098201e-06, "loss": 0.644099760055542, "step": 2930 }, { "epoch": 1.068313953488372, "grad_norm": 2.777388095855713, "learning_rate": 2.4263983373857773e-06, "loss": 0.6139822483062745, "step": 2940 }, { "epoch": 1.0719476744186047, "grad_norm": 2.5681188106536865, "learning_rate": 2.4113833504281187e-06, "loss": 0.6084999084472656, "step": 2950 }, { "epoch": 1.0755813953488371, "grad_norm": 2.812208652496338, "learning_rate": 2.396371563416146e-06, "loss": 0.6468010902404785, "step": 2960 }, { "epoch": 1.0792151162790697, "grad_norm": 2.7401721477508545, "learning_rate": 2.3813635184252285e-06, "loss": 0.6259693622589111, "step": 2970 }, { "epoch": 1.0828488372093024, "grad_norm": 2.7435302734375, "learning_rate": 2.3663597573956113e-06, "loss": 0.6502357482910156, "step": 2980 }, { "epoch": 1.086482558139535, "grad_norm": 2.9690780639648438, "learning_rate": 2.3513608221128477e-06, "loss": 0.6185351848602295, "step": 2990 }, { "epoch": 1.0901162790697674, "grad_norm": 2.996502161026001, "learning_rate": 2.336367254188229e-06, "loss": 0.6417867183685303, "step": 3000 }, { "epoch": 1.09375, "grad_norm": 2.8609607219696045, "learning_rate": 2.3213795950392355e-06, "loss": 0.6164373397827149, "step": 3010 }, { "epoch": 1.0973837209302326, "grad_norm": 2.8891489505767822, "learning_rate": 2.3063983858699808e-06, "loss": 0.6243292808532714, "step": 3020 }, { "epoch": 1.101017441860465, "grad_norm": 2.7142832279205322, "learning_rate": 2.291424167651668e-06, "loss": 0.6159169197082519, "step": 3030 }, { "epoch": 1.1046511627906976, "grad_norm": 2.84442400932312, "learning_rate": 2.2764574811030597e-06, "loss": 0.6410219192504882, "step": 3040 }, { "epoch": 1.1082848837209303, "grad_norm": 2.816093921661377, "learning_rate": 2.2614988666709464e-06, "loss": 0.6166658401489258, "step": 3050 }, { "epoch": 1.1119186046511629, "grad_norm": 2.8112494945526123, "learning_rate": 2.24654886451064e-06, "loss": 0.6123160362243653, "step": 3060 }, { "epoch": 1.1155523255813953, "grad_norm": 2.8347339630126953, "learning_rate": 2.2316080144664577e-06, "loss": 0.6261919498443603, "step": 3070 }, { "epoch": 1.119186046511628, "grad_norm": 3.0708513259887695, "learning_rate": 2.216676856052237e-06, "loss": 0.6543515205383301, "step": 3080 }, { "epoch": 1.1228197674418605, "grad_norm": 2.624223470687866, "learning_rate": 2.2017559284318517e-06, "loss": 0.647164249420166, "step": 3090 }, { "epoch": 1.1264534883720931, "grad_norm": 2.8848209381103516, "learning_rate": 2.18684577039974e-06, "loss": 0.6185880661010742, "step": 3100 }, { "epoch": 1.1300872093023255, "grad_norm": 2.8103585243225098, "learning_rate": 2.171946920361449e-06, "loss": 0.6510479927062989, "step": 3110 }, { "epoch": 1.1337209302325582, "grad_norm": 2.793445348739624, "learning_rate": 2.157059916314199e-06, "loss": 0.6501373291015625, "step": 3120 }, { "epoch": 1.1373546511627908, "grad_norm": 2.8822362422943115, "learning_rate": 2.142185295827448e-06, "loss": 0.6411583900451661, "step": 3130 }, { "epoch": 1.1409883720930232, "grad_norm": 2.8682944774627686, "learning_rate": 2.1273235960234844e-06, "loss": 0.616573715209961, "step": 3140 }, { "epoch": 1.1446220930232558, "grad_norm": 2.8865904808044434, "learning_rate": 2.1124753535580308e-06, "loss": 0.6400615215301514, "step": 3150 }, { "epoch": 1.1482558139534884, "grad_norm": 2.9084205627441406, "learning_rate": 2.0976411046008683e-06, "loss": 0.6371217727661133, "step": 3160 }, { "epoch": 1.151889534883721, "grad_norm": 2.7648379802703857, "learning_rate": 2.0828213848164693e-06, "loss": 0.6138511180877686, "step": 3170 }, { "epoch": 1.1555232558139534, "grad_norm": 2.8824777603149414, "learning_rate": 2.0680167293446585e-06, "loss": 0.6189014434814453, "step": 3180 }, { "epoch": 1.159156976744186, "grad_norm": 2.6267261505126953, "learning_rate": 2.05322767278129e-06, "loss": 0.6297481060028076, "step": 3190 }, { "epoch": 1.1627906976744187, "grad_norm": 2.7822842597961426, "learning_rate": 2.0384547491589416e-06, "loss": 0.6191432476043701, "step": 3200 }, { "epoch": 1.166424418604651, "grad_norm": 6.852417469024658, "learning_rate": 2.023698491927629e-06, "loss": 0.619715404510498, "step": 3210 }, { "epoch": 1.1700581395348837, "grad_norm": 2.771573781967163, "learning_rate": 2.008959433935547e-06, "loss": 0.6091080665588379, "step": 3220 }, { "epoch": 1.1736918604651163, "grad_norm": 2.977626085281372, "learning_rate": 1.9942381074098248e-06, "loss": 0.6197806358337402, "step": 3230 }, { "epoch": 1.177325581395349, "grad_norm": 2.740381956100464, "learning_rate": 1.9795350439373085e-06, "loss": 0.6083920001983643, "step": 3240 }, { "epoch": 1.1809593023255813, "grad_norm": 2.862666606903076, "learning_rate": 1.9648507744453657e-06, "loss": 0.6217282295227051, "step": 3250 }, { "epoch": 1.184593023255814, "grad_norm": 2.7649455070495605, "learning_rate": 1.9501858291827157e-06, "loss": 0.637911319732666, "step": 3260 }, { "epoch": 1.1882267441860466, "grad_norm": 2.8812625408172607, "learning_rate": 1.935540737700278e-06, "loss": 0.6026611804962159, "step": 3270 }, { "epoch": 1.191860465116279, "grad_norm": 2.8181381225585938, "learning_rate": 1.9209160288320517e-06, "loss": 0.6029317378997803, "step": 3280 }, { "epoch": 1.1954941860465116, "grad_norm": 2.935403823852539, "learning_rate": 1.906312230676024e-06, "loss": 0.645781135559082, "step": 3290 }, { "epoch": 1.1991279069767442, "grad_norm": 2.8169732093811035, "learning_rate": 1.891729870575092e-06, "loss": 0.6284616470336915, "step": 3300 }, { "epoch": 1.2027616279069768, "grad_norm": 2.8880012035369873, "learning_rate": 1.8771694750980277e-06, "loss": 0.6369534492492676, "step": 3310 }, { "epoch": 1.2063953488372092, "grad_norm": 2.939326286315918, "learning_rate": 1.8626315700204606e-06, "loss": 0.6241387367248535, "step": 3320 }, { "epoch": 1.2100290697674418, "grad_norm": 2.9609262943267822, "learning_rate": 1.848116680305892e-06, "loss": 0.6426369190216065, "step": 3330 }, { "epoch": 1.2136627906976745, "grad_norm": 2.8644490242004395, "learning_rate": 1.833625330086739e-06, "loss": 0.6207024574279785, "step": 3340 }, { "epoch": 1.2172965116279069, "grad_norm": 2.843106508255005, "learning_rate": 1.8191580426454064e-06, "loss": 0.6336426734924316, "step": 3350 }, { "epoch": 1.2209302325581395, "grad_norm": 3.1315815448760986, "learning_rate": 1.8047153403953948e-06, "loss": 0.6345052719116211, "step": 3360 }, { "epoch": 1.224563953488372, "grad_norm": 2.949122905731201, "learning_rate": 1.7902977448624334e-06, "loss": 0.6444235801696777, "step": 3370 }, { "epoch": 1.2281976744186047, "grad_norm": 2.835570812225342, "learning_rate": 1.7759057766656468e-06, "loss": 0.6098015785217286, "step": 3380 }, { "epoch": 1.2318313953488371, "grad_norm": 2.9221482276916504, "learning_rate": 1.7615399554987588e-06, "loss": 0.6236421585083007, "step": 3390 }, { "epoch": 1.2354651162790697, "grad_norm": 2.8423030376434326, "learning_rate": 1.7472008001113244e-06, "loss": 0.6142833709716797, "step": 3400 }, { "epoch": 1.2390988372093024, "grad_norm": 2.8265531063079834, "learning_rate": 1.7328888282899956e-06, "loss": 0.6144848346710206, "step": 3410 }, { "epoch": 1.2427325581395348, "grad_norm": 2.9081876277923584, "learning_rate": 1.7186045568398274e-06, "loss": 0.6280595779418945, "step": 3420 }, { "epoch": 1.2463662790697674, "grad_norm": 3.0004329681396484, "learning_rate": 1.7043485015656164e-06, "loss": 0.6482262134552002, "step": 3430 }, { "epoch": 1.25, "grad_norm": 2.7952747344970703, "learning_rate": 1.6901211772532717e-06, "loss": 0.6031205654144287, "step": 3440 }, { "epoch": 1.2536337209302326, "grad_norm": 3.007476329803467, "learning_rate": 1.675923097651228e-06, "loss": 0.648679256439209, "step": 3450 }, { "epoch": 1.2572674418604652, "grad_norm": 2.8375749588012695, "learning_rate": 1.6617547754518964e-06, "loss": 0.6462221622467041, "step": 3460 }, { "epoch": 1.2609011627906976, "grad_norm": 2.735497236251831, "learning_rate": 1.6476167222731448e-06, "loss": 0.6184795379638672, "step": 3470 }, { "epoch": 1.2645348837209303, "grad_norm": 3.0197677612304688, "learning_rate": 1.6335094486398295e-06, "loss": 0.6254825115203857, "step": 3480 }, { "epoch": 1.2681686046511627, "grad_norm": 2.9058151245117188, "learning_rate": 1.6194334639653588e-06, "loss": 0.6179644584655761, "step": 3490 }, { "epoch": 1.2718023255813953, "grad_norm": 2.8195924758911133, "learning_rate": 1.6053892765332957e-06, "loss": 0.6161976814270019, "step": 3500 }, { "epoch": 1.275436046511628, "grad_norm": 2.803330421447754, "learning_rate": 1.5913773934790062e-06, "loss": 0.6237835884094238, "step": 3510 }, { "epoch": 1.2790697674418605, "grad_norm": 2.789128303527832, "learning_rate": 1.577398320771344e-06, "loss": 0.6449883460998536, "step": 3520 }, { "epoch": 1.2827034883720931, "grad_norm": 2.9578816890716553, "learning_rate": 1.5634525631943844e-06, "loss": 0.6117482662200928, "step": 3530 }, { "epoch": 1.2863372093023255, "grad_norm": 2.8928158283233643, "learning_rate": 1.5495406243291925e-06, "loss": 0.6140883445739747, "step": 3540 }, { "epoch": 1.2899709302325582, "grad_norm": 2.741936206817627, "learning_rate": 1.5356630065356395e-06, "loss": 0.6419960498809815, "step": 3550 }, { "epoch": 1.2936046511627908, "grad_norm": 2.890014171600342, "learning_rate": 1.5218202109342655e-06, "loss": 0.5909524440765381, "step": 3560 }, { "epoch": 1.2972383720930232, "grad_norm": 2.7969582080841064, "learning_rate": 1.508012737388181e-06, "loss": 0.5948357582092285, "step": 3570 }, { "epoch": 1.3008720930232558, "grad_norm": 2.8763949871063232, "learning_rate": 1.4942410844850144e-06, "loss": 0.6192843914031982, "step": 3580 }, { "epoch": 1.3045058139534884, "grad_norm": 2.978353977203369, "learning_rate": 1.4805057495189178e-06, "loss": 0.614830493927002, "step": 3590 }, { "epoch": 1.308139534883721, "grad_norm": 2.9594194889068604, "learning_rate": 1.4668072284725974e-06, "loss": 0.6267194271087646, "step": 3600 }, { "epoch": 1.3117732558139534, "grad_norm": 2.777967691421509, "learning_rate": 1.4531460159994125e-06, "loss": 0.6032927513122559, "step": 3610 }, { "epoch": 1.315406976744186, "grad_norm": 2.823629856109619, "learning_rate": 1.4395226054055084e-06, "loss": 0.6197312355041504, "step": 3620 }, { "epoch": 1.3190406976744187, "grad_norm": 2.972522735595703, "learning_rate": 1.4259374886320064e-06, "loss": 0.6175290107727051, "step": 3630 }, { "epoch": 1.322674418604651, "grad_norm": 3.864912271499634, "learning_rate": 1.4123911562372411e-06, "loss": 0.5972983360290527, "step": 3640 }, { "epoch": 1.3263081395348837, "grad_norm": 2.642355442047119, "learning_rate": 1.398884097379038e-06, "loss": 0.6334139823913574, "step": 3650 }, { "epoch": 1.3299418604651163, "grad_norm": 2.843428134918213, "learning_rate": 1.3854167997970607e-06, "loss": 0.6349598407745362, "step": 3660 }, { "epoch": 1.333575581395349, "grad_norm": 2.9344217777252197, "learning_rate": 1.3719897497951934e-06, "loss": 0.6205593585968018, "step": 3670 }, { "epoch": 1.3372093023255813, "grad_norm": 2.964674949645996, "learning_rate": 1.3586034322239777e-06, "loss": 0.6372932910919189, "step": 3680 }, { "epoch": 1.340843023255814, "grad_norm": 2.9677414894104004, "learning_rate": 1.3452583304631118e-06, "loss": 0.5919643402099609, "step": 3690 }, { "epoch": 1.3444767441860466, "grad_norm": 2.970325469970703, "learning_rate": 1.3319549264039874e-06, "loss": 0.6072146415710449, "step": 3700 }, { "epoch": 1.348110465116279, "grad_norm": 2.8650684356689453, "learning_rate": 1.318693700432298e-06, "loss": 0.6300556182861328, "step": 3710 }, { "epoch": 1.3517441860465116, "grad_norm": 3.107119083404541, "learning_rate": 1.305475131410682e-06, "loss": 0.6029188156127929, "step": 3720 }, { "epoch": 1.3553779069767442, "grad_norm": 2.9079434871673584, "learning_rate": 1.2922996966614407e-06, "loss": 0.6184733390808106, "step": 3730 }, { "epoch": 1.3590116279069768, "grad_norm": 2.755070924758911, "learning_rate": 1.2791678719492931e-06, "loss": 0.6277907371520997, "step": 3740 }, { "epoch": 1.3626453488372092, "grad_norm": 3.044586658477783, "learning_rate": 1.2660801314642038e-06, "loss": 0.6256103515625, "step": 3750 }, { "epoch": 1.3662790697674418, "grad_norm": 2.9525163173675537, "learning_rate": 1.2530369478042567e-06, "loss": 0.6190742015838623, "step": 3760 }, { "epoch": 1.3699127906976745, "grad_norm": 2.899778366088867, "learning_rate": 1.2400387919585876e-06, "loss": 0.6287983894348145, "step": 3770 }, { "epoch": 1.3735465116279069, "grad_norm": 2.8771963119506836, "learning_rate": 1.2270861332903774e-06, "loss": 0.6238520622253418, "step": 3780 }, { "epoch": 1.3771802325581395, "grad_norm": 2.7630653381347656, "learning_rate": 1.2141794395199112e-06, "loss": 0.6069451332092285, "step": 3790 }, { "epoch": 1.380813953488372, "grad_norm": 2.847613573074341, "learning_rate": 1.2013191767076756e-06, "loss": 0.6269810676574707, "step": 3800 }, { "epoch": 1.3844476744186047, "grad_norm": 2.879697799682617, "learning_rate": 1.1885058092375385e-06, "loss": 0.6266268730163574, "step": 3810 }, { "epoch": 1.3880813953488373, "grad_norm": 2.9201645851135254, "learning_rate": 1.1757397997999778e-06, "loss": 0.6043568611145019, "step": 3820 }, { "epoch": 1.3917151162790697, "grad_norm": 3.0528650283813477, "learning_rate": 1.1630216093753756e-06, "loss": 0.6078732490539551, "step": 3830 }, { "epoch": 1.3953488372093024, "grad_norm": 2.8742434978485107, "learning_rate": 1.1503516972173675e-06, "loss": 0.6376693248748779, "step": 3840 }, { "epoch": 1.3989825581395348, "grad_norm": 2.934420585632324, "learning_rate": 1.1377305208362613e-06, "loss": 0.594962215423584, "step": 3850 }, { "epoch": 1.4026162790697674, "grad_norm": 2.923924684524536, "learning_rate": 1.1251585359825193e-06, "loss": 0.6222631454467773, "step": 3860 }, { "epoch": 1.40625, "grad_norm": 2.8981270790100098, "learning_rate": 1.1126361966302983e-06, "loss": 0.6129872322082519, "step": 3870 }, { "epoch": 1.4098837209302326, "grad_norm": 2.969998598098755, "learning_rate": 1.1001639549610546e-06, "loss": 0.6390999794006348, "step": 3880 }, { "epoch": 1.4135174418604652, "grad_norm": 2.8989439010620117, "learning_rate": 1.0877422613472196e-06, "loss": 0.5968554496765137, "step": 3890 }, { "epoch": 1.4171511627906976, "grad_norm": 2.928755044937134, "learning_rate": 1.075371564335937e-06, "loss": 0.6382357597351074, "step": 3900 }, { "epoch": 1.4207848837209303, "grad_norm": 2.8408379554748535, "learning_rate": 1.0630523106328605e-06, "loss": 0.6284653186798096, "step": 3910 }, { "epoch": 1.4244186046511627, "grad_norm": 3.0078012943267822, "learning_rate": 1.0507849450860277e-06, "loss": 0.6212825775146484, "step": 3920 }, { "epoch": 1.4280523255813953, "grad_norm": 2.981374979019165, "learning_rate": 1.0385699106697983e-06, "loss": 0.6222080230712891, "step": 3930 }, { "epoch": 1.431686046511628, "grad_norm": 2.9854984283447266, "learning_rate": 1.026407648468854e-06, "loss": 0.6164888858795166, "step": 3940 }, { "epoch": 1.4353197674418605, "grad_norm": 2.661363124847412, "learning_rate": 1.0142985976622718e-06, "loss": 0.6039540767669678, "step": 3950 }, { "epoch": 1.4389534883720931, "grad_norm": 3.1405701637268066, "learning_rate": 1.0022431955076686e-06, "loss": 0.6292922973632813, "step": 3960 }, { "epoch": 1.4425872093023255, "grad_norm": 2.916566848754883, "learning_rate": 9.902418773254064e-07, "loss": 0.6347232341766358, "step": 3970 }, { "epoch": 1.4462209302325582, "grad_norm": 2.7745325565338135, "learning_rate": 9.782950764828788e-07, "loss": 0.6333943367004394, "step": 3980 }, { "epoch": 1.4498546511627908, "grad_norm": 2.7898178100585938, "learning_rate": 9.664032243788565e-07, "loss": 0.6126363754272461, "step": 3990 }, { "epoch": 1.4534883720930232, "grad_norm": 3.018996238708496, "learning_rate": 9.545667504279147e-07, "loss": 0.6129265785217285, "step": 4000 }, { "epoch": 1.4571220930232558, "grad_norm": 2.880807638168335, "learning_rate": 9.427860820449236e-07, "loss": 0.6031558990478516, "step": 4010 }, { "epoch": 1.4607558139534884, "grad_norm": 2.9041407108306885, "learning_rate": 9.310616446296139e-07, "loss": 0.6169293880462646, "step": 4020 }, { "epoch": 1.464389534883721, "grad_norm": 2.7830722332000732, "learning_rate": 9.193938615512193e-07, "loss": 0.6310886383056641, "step": 4030 }, { "epoch": 1.4680232558139534, "grad_norm": 2.8672258853912354, "learning_rate": 9.077831541331842e-07, "loss": 0.6097919464111328, "step": 4040 }, { "epoch": 1.471656976744186, "grad_norm": 3.0425376892089844, "learning_rate": 8.962299416379533e-07, "loss": 0.6108634948730469, "step": 4050 }, { "epoch": 1.4752906976744187, "grad_norm": 3.1145591735839844, "learning_rate": 8.847346412518312e-07, "loss": 0.6188015937805176, "step": 4060 }, { "epoch": 1.478924418604651, "grad_norm": 2.893653631210327, "learning_rate": 8.732976680699153e-07, "loss": 0.6100796699523926, "step": 4070 }, { "epoch": 1.4825581395348837, "grad_norm": 2.7886803150177, "learning_rate": 8.619194350811094e-07, "loss": 0.6053628921508789, "step": 4080 }, { "epoch": 1.4861918604651163, "grad_norm": 3.087308168411255, "learning_rate": 8.506003531532103e-07, "loss": 0.6025334358215332, "step": 4090 }, { "epoch": 1.489825581395349, "grad_norm": 2.927269220352173, "learning_rate": 8.393408310180722e-07, "loss": 0.6248285293579101, "step": 4100 }, { "epoch": 1.4934593023255813, "grad_norm": 2.8830156326293945, "learning_rate": 8.281412752568435e-07, "loss": 0.601197338104248, "step": 4110 }, { "epoch": 1.497093023255814, "grad_norm": 2.9723641872406006, "learning_rate": 8.170020902852896e-07, "loss": 0.6211593627929688, "step": 4120 }, { "epoch": 1.5007267441860463, "grad_norm": 2.9226901531219482, "learning_rate": 8.059236783391885e-07, "loss": 0.6135400295257568, "step": 4130 }, { "epoch": 1.504360465116279, "grad_norm": 2.7875003814697266, "learning_rate": 7.949064394598033e-07, "loss": 0.6013899326324463, "step": 4140 }, { "epoch": 1.5079941860465116, "grad_norm": 2.9814796447753906, "learning_rate": 7.839507714794389e-07, "loss": 0.5883074283599854, "step": 4150 }, { "epoch": 1.5116279069767442, "grad_norm": 2.743788242340088, "learning_rate": 7.73057070007078e-07, "loss": 0.6389053344726563, "step": 4160 }, { "epoch": 1.5152616279069768, "grad_norm": 2.8575990200042725, "learning_rate": 7.622257284140927e-07, "loss": 0.6024910926818847, "step": 4170 }, { "epoch": 1.5188953488372094, "grad_norm": 2.818049907684326, "learning_rate": 7.514571378200408e-07, "loss": 0.6210941314697266, "step": 4180 }, { "epoch": 1.5225290697674418, "grad_norm": 2.9408257007598877, "learning_rate": 7.407516870785405e-07, "loss": 0.6397895812988281, "step": 4190 }, { "epoch": 1.5261627906976745, "grad_norm": 3.026486396789551, "learning_rate": 7.301097627632367e-07, "loss": 0.607673454284668, "step": 4200 }, { "epoch": 1.5297965116279069, "grad_norm": 2.9318480491638184, "learning_rate": 7.195317491538308e-07, "loss": 0.5777746677398682, "step": 4210 }, { "epoch": 1.5334302325581395, "grad_norm": 2.8912055492401123, "learning_rate": 7.090180282222109e-07, "loss": 0.6106159210205078, "step": 4220 }, { "epoch": 1.537063953488372, "grad_norm": 2.863584280014038, "learning_rate": 6.985689796186593e-07, "loss": 0.6015161037445068, "step": 4230 }, { "epoch": 1.5406976744186047, "grad_norm": 2.8779637813568115, "learning_rate": 6.881849806581405e-07, "loss": 0.5897069931030273, "step": 4240 }, { "epoch": 1.5443313953488373, "grad_norm": 2.8327910900115967, "learning_rate": 6.778664063066773e-07, "loss": 0.5938574790954589, "step": 4250 }, { "epoch": 1.5479651162790697, "grad_norm": 2.9977948665618896, "learning_rate": 6.676136291678118e-07, "loss": 0.583928394317627, "step": 4260 }, { "epoch": 1.5515988372093024, "grad_norm": 3.029289722442627, "learning_rate": 6.574270194691485e-07, "loss": 0.6218618392944336, "step": 4270 }, { "epoch": 1.5552325581395348, "grad_norm": 2.750098943710327, "learning_rate": 6.473069450489888e-07, "loss": 0.6067587375640869, "step": 4280 }, { "epoch": 1.5588662790697674, "grad_norm": 3.033493757247925, "learning_rate": 6.37253771343044e-07, "loss": 0.6148815631866456, "step": 4290 }, { "epoch": 1.5625, "grad_norm": 2.889744758605957, "learning_rate": 6.272678613712437e-07, "loss": 0.6183005332946777, "step": 4300 }, { "epoch": 1.5661337209302326, "grad_norm": 3.086649179458618, "learning_rate": 6.173495757246251e-07, "loss": 0.5992136001586914, "step": 4310 }, { "epoch": 1.5697674418604652, "grad_norm": 2.90238618850708, "learning_rate": 6.074992725523102e-07, "loss": 0.6303570747375489, "step": 4320 }, { "epoch": 1.5734011627906976, "grad_norm": 2.995821475982666, "learning_rate": 5.977173075485778e-07, "loss": 0.5939804077148437, "step": 4330 }, { "epoch": 1.5770348837209303, "grad_norm": 2.9475557804107666, "learning_rate": 5.880040339400142e-07, "loss": 0.6166177749633789, "step": 4340 }, { "epoch": 1.5806686046511627, "grad_norm": 3.1026580333709717, "learning_rate": 5.783598024727618e-07, "loss": 0.6242012500762939, "step": 4350 }, { "epoch": 1.5843023255813953, "grad_norm": 2.9524173736572266, "learning_rate": 5.687849613998534e-07, "loss": 0.6128497123718262, "step": 4360 }, { "epoch": 1.587936046511628, "grad_norm": 3.0528416633605957, "learning_rate": 5.592798564686339e-07, "loss": 0.6200154304504395, "step": 4370 }, { "epoch": 1.5915697674418605, "grad_norm": 3.027564525604248, "learning_rate": 5.498448309082777e-07, "loss": 0.6254747390747071, "step": 4380 }, { "epoch": 1.5952034883720931, "grad_norm": 2.8900911808013916, "learning_rate": 5.40480225417395e-07, "loss": 0.6331692695617676, "step": 4390 }, { "epoch": 1.5988372093023255, "grad_norm": 2.9777119159698486, "learning_rate": 5.311863781517296e-07, "loss": 0.5979508399963379, "step": 4400 }, { "epoch": 1.6024709302325582, "grad_norm": 2.78714919090271, "learning_rate": 5.219636247119448e-07, "loss": 0.6151116371154786, "step": 4410 }, { "epoch": 1.6061046511627906, "grad_norm": 2.840519666671753, "learning_rate": 5.128122981315068e-07, "loss": 0.6057453155517578, "step": 4420 }, { "epoch": 1.6097383720930232, "grad_norm": 2.6660914421081543, "learning_rate": 5.03732728864663e-07, "loss": 0.6372000694274902, "step": 4430 }, { "epoch": 1.6133720930232558, "grad_norm": 2.752408504486084, "learning_rate": 4.947252447745013e-07, "loss": 0.6181222438812256, "step": 4440 }, { "epoch": 1.6170058139534884, "grad_norm": 2.874816417694092, "learning_rate": 4.857901711211161e-07, "loss": 0.6298707962036133, "step": 4450 }, { "epoch": 1.620639534883721, "grad_norm": 2.794626474380493, "learning_rate": 4.769278305498632e-07, "loss": 0.6108518600463867, "step": 4460 }, { "epoch": 1.6242732558139537, "grad_norm": 2.8354408740997314, "learning_rate": 4.681385430797086e-07, "loss": 0.6353400230407715, "step": 4470 }, { "epoch": 1.627906976744186, "grad_norm": 2.9009554386138916, "learning_rate": 4.594226260916701e-07, "loss": 0.6042340278625489, "step": 4480 }, { "epoch": 1.6315406976744184, "grad_norm": 2.8754899501800537, "learning_rate": 4.507803943173592e-07, "loss": 0.6066765308380127, "step": 4490 }, { "epoch": 1.635174418604651, "grad_norm": 2.9461255073547363, "learning_rate": 4.422121598276161e-07, "loss": 0.6181964874267578, "step": 4500 }, { "epoch": 1.6388081395348837, "grad_norm": 2.949796438217163, "learning_rate": 4.3371823202124156e-07, "loss": 0.602027177810669, "step": 4510 }, { "epoch": 1.6424418604651163, "grad_norm": 2.9036405086517334, "learning_rate": 4.2529891761382e-07, "loss": 0.5935526847839355, "step": 4520 }, { "epoch": 1.646075581395349, "grad_norm": 2.956812620162964, "learning_rate": 4.1695452062665067e-07, "loss": 0.6340709686279297, "step": 4530 }, { "epoch": 1.6497093023255816, "grad_norm": 2.9781763553619385, "learning_rate": 4.0868534237576476e-07, "loss": 0.6047008514404297, "step": 4540 }, { "epoch": 1.653343023255814, "grad_norm": 2.7890918254852295, "learning_rate": 4.0049168146104487e-07, "loss": 0.6058784484863281, "step": 4550 }, { "epoch": 1.6569767441860463, "grad_norm": 2.9034600257873535, "learning_rate": 3.923738337554467e-07, "loss": 0.6007349967956543, "step": 4560 }, { "epoch": 1.660610465116279, "grad_norm": 2.934140205383301, "learning_rate": 3.843320923943089e-07, "loss": 0.5974663734436035, "step": 4570 }, { "epoch": 1.6642441860465116, "grad_norm": 2.7771620750427246, "learning_rate": 3.7636674776477507e-07, "loss": 0.6011982440948487, "step": 4580 }, { "epoch": 1.6678779069767442, "grad_norm": 2.8858895301818848, "learning_rate": 3.6847808749530005e-07, "loss": 0.6234261512756347, "step": 4590 }, { "epoch": 1.6715116279069768, "grad_norm": 2.909925937652588, "learning_rate": 3.6066639644527195e-07, "loss": 0.6237231254577636, "step": 4600 }, { "epoch": 1.6751453488372094, "grad_norm": 2.912471055984497, "learning_rate": 3.529319566947184e-07, "loss": 0.6141771793365478, "step": 4610 }, { "epoch": 1.6787790697674418, "grad_norm": 2.846323013305664, "learning_rate": 3.4527504753412517e-07, "loss": 0.6363746643066406, "step": 4620 }, { "epoch": 1.6824127906976745, "grad_norm": 2.9305002689361572, "learning_rate": 3.3769594545435103e-07, "loss": 0.5971851348876953, "step": 4630 }, { "epoch": 1.6860465116279069, "grad_norm": 2.954364776611328, "learning_rate": 3.301949241366398e-07, "loss": 0.5849172592163085, "step": 4640 }, { "epoch": 1.6896802325581395, "grad_norm": 2.9237112998962402, "learning_rate": 3.227722544427428e-07, "loss": 0.5754762649536133, "step": 4650 }, { "epoch": 1.693313953488372, "grad_norm": 2.8475701808929443, "learning_rate": 3.154282044051335e-07, "loss": 0.570661211013794, "step": 4660 }, { "epoch": 1.6969476744186047, "grad_norm": 3.0029714107513428, "learning_rate": 3.081630392173329e-07, "loss": 0.6232757091522216, "step": 4670 }, { "epoch": 1.7005813953488373, "grad_norm": 3.1450560092926025, "learning_rate": 3.009770212243293e-07, "loss": 0.6245338439941406, "step": 4680 }, { "epoch": 1.7042151162790697, "grad_norm": 2.9455108642578125, "learning_rate": 2.93870409913109e-07, "loss": 0.6269196510314942, "step": 4690 }, { "epoch": 1.7078488372093024, "grad_norm": 2.832939386367798, "learning_rate": 2.8684346190328393e-07, "loss": 0.6224913597106934, "step": 4700 }, { "epoch": 1.7114825581395348, "grad_norm": 2.767312526702881, "learning_rate": 2.798964309378255e-07, "loss": 0.6045539379119873, "step": 4710 }, { "epoch": 1.7151162790697674, "grad_norm": 2.9070537090301514, "learning_rate": 2.7302956787390006e-07, "loss": 0.6086003303527832, "step": 4720 }, { "epoch": 1.71875, "grad_norm": 2.816274642944336, "learning_rate": 2.662431206738167e-07, "loss": 0.6211147308349609, "step": 4730 }, { "epoch": 1.7223837209302326, "grad_norm": 2.8480284214019775, "learning_rate": 2.595373343960658e-07, "loss": 0.6072659492492676, "step": 4740 }, { "epoch": 1.7260174418604652, "grad_norm": 2.88503098487854, "learning_rate": 2.5291245118647304e-07, "loss": 0.6151592254638671, "step": 4750 }, { "epoch": 1.7296511627906976, "grad_norm": 3.040351629257202, "learning_rate": 2.4636871026945757e-07, "loss": 0.5649903297424317, "step": 4760 }, { "epoch": 1.7332848837209303, "grad_norm": 3.088498830795288, "learning_rate": 2.399063479393912e-07, "loss": 0.6478461265563965, "step": 4770 }, { "epoch": 1.7369186046511627, "grad_norm": 2.9244418144226074, "learning_rate": 2.335255975520656e-07, "loss": 0.6145503997802735, "step": 4780 }, { "epoch": 1.7405523255813953, "grad_norm": 2.8953909873962402, "learning_rate": 2.272266895162664e-07, "loss": 0.596315574645996, "step": 4790 }, { "epoch": 1.744186046511628, "grad_norm": 3.0355396270751953, "learning_rate": 2.2100985128545498e-07, "loss": 0.6153911590576172, "step": 4800 }, { "epoch": 1.7478197674418605, "grad_norm": 2.813180685043335, "learning_rate": 2.1487530734955242e-07, "loss": 0.6165101051330566, "step": 4810 }, { "epoch": 1.7514534883720931, "grad_norm": 2.9147789478302, "learning_rate": 2.0882327922683354e-07, "loss": 0.591987133026123, "step": 4820 }, { "epoch": 1.7550872093023255, "grad_norm": 2.9788706302642822, "learning_rate": 2.0285398545592999e-07, "loss": 0.6493700981140137, "step": 4830 }, { "epoch": 1.7587209302325582, "grad_norm": 2.7692644596099854, "learning_rate": 1.969676415879368e-07, "loss": 0.621481990814209, "step": 4840 }, { "epoch": 1.7623546511627906, "grad_norm": 2.873155355453491, "learning_rate": 1.9116446017862917e-07, "loss": 0.6019724845886231, "step": 4850 }, { "epoch": 1.7659883720930232, "grad_norm": 2.8992841243743896, "learning_rate": 1.8544465078078693e-07, "loss": 0.6230618953704834, "step": 4860 }, { "epoch": 1.7696220930232558, "grad_norm": 3.053983211517334, "learning_rate": 1.798084199366293e-07, "loss": 0.5850800514221192, "step": 4870 }, { "epoch": 1.7732558139534884, "grad_norm": 2.929288387298584, "learning_rate": 1.7425597117035537e-07, "loss": 0.597361946105957, "step": 4880 }, { "epoch": 1.776889534883721, "grad_norm": 2.917435884475708, "learning_rate": 1.6878750498079273e-07, "loss": 0.5993693351745606, "step": 4890 }, { "epoch": 1.7805232558139537, "grad_norm": 3.0222082138061523, "learning_rate": 1.6340321883416222e-07, "loss": 0.6234514236450195, "step": 4900 }, { "epoch": 1.784156976744186, "grad_norm": 3.007539749145508, "learning_rate": 1.5810330715694266e-07, "loss": 0.5682544708251953, "step": 4910 }, { "epoch": 1.7877906976744184, "grad_norm": 2.8434793949127197, "learning_rate": 1.5288796132885393e-07, "loss": 0.5935199737548829, "step": 4920 }, { "epoch": 1.791424418604651, "grad_norm": 2.786289691925049, "learning_rate": 1.4775736967594379e-07, "loss": 0.5871915817260742, "step": 4930 }, { "epoch": 1.7950581395348837, "grad_norm": 2.984264850616455, "learning_rate": 1.4271171746378764e-07, "loss": 0.6204954147338867, "step": 4940 }, { "epoch": 1.7986918604651163, "grad_norm": 2.8578007221221924, "learning_rate": 1.3775118689080048e-07, "loss": 0.601482343673706, "step": 4950 }, { "epoch": 1.802325581395349, "grad_norm": 2.9476382732391357, "learning_rate": 1.3287595708165457e-07, "loss": 0.6299497127532959, "step": 4960 }, { "epoch": 1.8059593023255816, "grad_norm": 2.946737051010132, "learning_rate": 1.2808620408081536e-07, "loss": 0.6067971229553223, "step": 4970 }, { "epoch": 1.809593023255814, "grad_norm": 2.9323055744171143, "learning_rate": 1.2338210084617996e-07, "loss": 0.6230051040649414, "step": 4980 }, { "epoch": 1.8132267441860463, "grad_norm": 3.0436692237854004, "learning_rate": 1.1876381724283515e-07, "loss": 0.6139970302581788, "step": 4990 }, { "epoch": 1.816860465116279, "grad_norm": 2.9199883937835693, "learning_rate": 1.142315200369229e-07, "loss": 0.6280738830566406, "step": 5000 }, { "epoch": 1.8204941860465116, "grad_norm": 2.92120361328125, "learning_rate": 1.0978537288961616e-07, "loss": 0.5834029197692872, "step": 5010 }, { "epoch": 1.8241279069767442, "grad_norm": 3.087002754211426, "learning_rate": 1.0542553635121134e-07, "loss": 0.6059079647064209, "step": 5020 }, { "epoch": 1.8277616279069768, "grad_norm": 2.9495768547058105, "learning_rate": 1.0115216785533149e-07, "loss": 0.6083863258361817, "step": 5030 }, { "epoch": 1.8313953488372094, "grad_norm": 2.7307910919189453, "learning_rate": 9.696542171323902e-08, "loss": 0.5967412948608398, "step": 5040 }, { "epoch": 1.8350290697674418, "grad_norm": 2.919420003890991, "learning_rate": 9.286544910826395e-08, "loss": 0.6107944488525391, "step": 5050 }, { "epoch": 1.8386627906976745, "grad_norm": 3.077995538711548, "learning_rate": 8.885239809034718e-08, "loss": 0.6117441654205322, "step": 5060 }, { "epoch": 1.8422965116279069, "grad_norm": 2.8363027572631836, "learning_rate": 8.492641357069148e-08, "loss": 0.5940214633941651, "step": 5070 }, { "epoch": 1.8459302325581395, "grad_norm": 2.935504913330078, "learning_rate": 8.10876373165298e-08, "loss": 0.5961288452148438, "step": 5080 }, { "epoch": 1.849563953488372, "grad_norm": 2.998866081237793, "learning_rate": 7.733620794600604e-08, "loss": 0.6110012054443359, "step": 5090 }, { "epoch": 1.8531976744186047, "grad_norm": 3.069027900695801, "learning_rate": 7.367226092317076e-08, "loss": 0.5856106281280518, "step": 5100 }, { "epoch": 1.8568313953488373, "grad_norm": 2.948878049850464, "learning_rate": 7.009592855308722e-08, "loss": 0.6236933708190918, "step": 5110 }, { "epoch": 1.8604651162790697, "grad_norm": 2.9042975902557373, "learning_rate": 6.66073399770556e-08, "loss": 0.5904771327972412, "step": 5120 }, { "epoch": 1.8640988372093024, "grad_norm": 3.013841390609741, "learning_rate": 6.320662116794912e-08, "loss": 0.610911750793457, "step": 5130 }, { "epoch": 1.8677325581395348, "grad_norm": 2.9105255603790283, "learning_rate": 5.989389492566527e-08, "loss": 0.5993082046508789, "step": 5140 }, { "epoch": 1.8713662790697674, "grad_norm": 2.9587931632995605, "learning_rate": 5.66692808726918e-08, "loss": 0.6032283306121826, "step": 5150 }, { "epoch": 1.875, "grad_norm": 2.8542747497558594, "learning_rate": 5.3532895449785735e-08, "loss": 0.6127349376678467, "step": 5160 }, { "epoch": 1.8786337209302326, "grad_norm": 2.926576614379883, "learning_rate": 5.0484851911770664e-08, "loss": 0.5979854583740234, "step": 5170 }, { "epoch": 1.8822674418604652, "grad_norm": 2.8693761825561523, "learning_rate": 4.7525260323446897e-08, "loss": 0.5823795318603515, "step": 5180 }, { "epoch": 1.8859011627906976, "grad_norm": 2.8152878284454346, "learning_rate": 4.4654227555614716e-08, "loss": 0.5919192790985107, "step": 5190 }, { "epoch": 1.8895348837209303, "grad_norm": 2.8572585582733154, "learning_rate": 4.187185728121823e-08, "loss": 0.588810396194458, "step": 5200 }, { "epoch": 1.8931686046511627, "grad_norm": 2.9437289237976074, "learning_rate": 3.9178249971599523e-08, "loss": 0.6189647674560547, "step": 5210 }, { "epoch": 1.8968023255813953, "grad_norm": 2.8459627628326416, "learning_rate": 3.657350289287265e-08, "loss": 0.6121656894683838, "step": 5220 }, { "epoch": 1.900436046511628, "grad_norm": 3.0236427783966064, "learning_rate": 3.4057710102409215e-08, "loss": 0.6005516529083252, "step": 5230 }, { "epoch": 1.9040697674418605, "grad_norm": 3.0145411491394043, "learning_rate": 3.163096244544361e-08, "loss": 0.6206085205078125, "step": 5240 }, { "epoch": 1.9077034883720931, "grad_norm": 2.7869176864624023, "learning_rate": 2.929334755179175e-08, "loss": 0.6017412185668946, "step": 5250 }, { "epoch": 1.9113372093023255, "grad_norm": 2.7438442707061768, "learning_rate": 2.7044949832686908e-08, "loss": 0.611904239654541, "step": 5260 }, { "epoch": 1.9149709302325582, "grad_norm": 2.8678460121154785, "learning_rate": 2.4885850477732178e-08, "loss": 0.6094479084014892, "step": 5270 }, { "epoch": 1.9186046511627906, "grad_norm": 2.9781479835510254, "learning_rate": 2.281612745196782e-08, "loss": 0.6219481945037841, "step": 5280 }, { "epoch": 1.9222383720930232, "grad_norm": 3.0608043670654297, "learning_rate": 2.083585549305628e-08, "loss": 0.5890862464904785, "step": 5290 }, { "epoch": 1.9258720930232558, "grad_norm": 3.0372581481933594, "learning_rate": 1.8945106108584634e-08, "loss": 0.6225212574005127, "step": 5300 }, { "epoch": 1.9295058139534884, "grad_norm": 3.0238840579986572, "learning_rate": 1.7143947573479704e-08, "loss": 0.6276484489440918, "step": 5310 }, { "epoch": 1.933139534883721, "grad_norm": 2.9078598022460938, "learning_rate": 1.543244492754531e-08, "loss": 0.5980105400085449, "step": 5320 }, { "epoch": 1.9367732558139537, "grad_norm": 2.8863484859466553, "learning_rate": 1.3810659973112206e-08, "loss": 0.6007534980773925, "step": 5330 }, { "epoch": 1.940406976744186, "grad_norm": 2.7025346755981445, "learning_rate": 1.227865127280764e-08, "loss": 0.5999039649963379, "step": 5340 }, { "epoch": 1.9440406976744184, "grad_norm": 2.943773031234741, "learning_rate": 1.0836474147438991e-08, "loss": 0.614153242111206, "step": 5350 }, { "epoch": 1.947674418604651, "grad_norm": 2.920337438583374, "learning_rate": 9.484180673997312e-09, "loss": 0.6102373123168945, "step": 5360 }, { "epoch": 1.9513081395348837, "grad_norm": 2.9423022270202637, "learning_rate": 8.22181968377661e-09, "loss": 0.5994505882263184, "step": 5370 }, { "epoch": 1.9549418604651163, "grad_norm": 2.965574264526367, "learning_rate": 7.0494367606105396e-09, "loss": 0.5969588279724121, "step": 5380 }, { "epoch": 1.958575581395349, "grad_norm": 2.9373302459716797, "learning_rate": 5.967074239226212e-09, "loss": 0.6102699279785156, "step": 5390 }, { "epoch": 1.9622093023255816, "grad_norm": 3.050691604614258, "learning_rate": 4.974771203715145e-09, "loss": 0.607047176361084, "step": 5400 }, { "epoch": 1.965843023255814, "grad_norm": 2.996528387069702, "learning_rate": 4.07256348612356e-09, "loss": 0.6196224212646484, "step": 5410 }, { "epoch": 1.9694767441860463, "grad_norm": 2.9037375450134277, "learning_rate": 3.260483665155911e-09, "loss": 0.5886441230773926, "step": 5420 }, { "epoch": 1.973110465116279, "grad_norm": 2.8899309635162354, "learning_rate": 2.538561065000833e-09, "loss": 0.6017139434814454, "step": 5430 }, { "epoch": 1.9767441860465116, "grad_norm": 2.9255728721618652, "learning_rate": 1.906821754270871e-09, "loss": 0.6277134895324707, "step": 5440 }, { "epoch": 1.9803779069767442, "grad_norm": 3.010143995285034, "learning_rate": 1.365288545061849e-09, "loss": 0.6320951461791993, "step": 5450 }, { "epoch": 1.9840116279069768, "grad_norm": 3.0056440830230713, "learning_rate": 9.139809921282494e-10, "loss": 0.5893740177154541, "step": 5460 }, { "epoch": 1.9876453488372094, "grad_norm": 2.968355894088745, "learning_rate": 5.529153921779439e-10, "loss": 0.6041777610778809, "step": 5470 }, { "epoch": 1.9912790697674418, "grad_norm": 2.8749659061431885, "learning_rate": 2.821047832837764e-10, "loss": 0.5875132083892822, "step": 5480 }, { "epoch": 1.9949127906976745, "grad_norm": 2.902491331100464, "learning_rate": 1.0155894441227221e-10, "loss": 0.5986234664916992, "step": 5490 }, { "epoch": 1.9985465116279069, "grad_norm": 2.9591474533081055, "learning_rate": 1.1284395070310095e-11, "loss": 0.6058923721313476, "step": 5500 }, { "epoch": 2.0, "step": 5504, "total_flos": 2.0144853494769123e+19, "train_loss": 0.736562816166254, "train_runtime": 18605.1295, "train_samples_per_second": 9.464, "train_steps_per_second": 0.296 } ], "logging_steps": 10, "max_steps": 5504, "num_input_tokens_seen": 0, "num_train_epochs": 2, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 2.0144853494769123e+19, "train_batch_size": 2, "trial_name": null, "trial_params": null }