diff --git "a/qwen-full-sft-few-shot/trainer_state.json" "b/qwen-full-sft-few-shot/trainer_state.json" new file mode 100644--- /dev/null +++ "b/qwen-full-sft-few-shot/trainer_state.json" @@ -0,0 +1,32739 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 3.0, + "eval_steps": 500, + "global_step": 4671, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0006422607578676942, + "grad_norm": 35.19259286597093, + "learning_rate": 2.136752136752137e-08, + "loss": 0.9316, + "step": 1 + }, + { + "epoch": 0.0012845215157353885, + "grad_norm": 36.73974467039917, + "learning_rate": 4.273504273504274e-08, + "loss": 0.8993, + "step": 2 + }, + { + "epoch": 0.0019267822736030828, + "grad_norm": 30.242473219086172, + "learning_rate": 6.41025641025641e-08, + "loss": 1.0401, + "step": 3 + }, + { + "epoch": 0.002569043031470777, + "grad_norm": 29.331775210557844, + "learning_rate": 8.547008547008549e-08, + "loss": 0.9108, + "step": 4 + }, + { + "epoch": 0.0032113037893384713, + "grad_norm": 35.15269258099587, + "learning_rate": 1.0683760683760685e-07, + "loss": 0.8122, + "step": 5 + }, + { + "epoch": 0.0038535645472061657, + "grad_norm": 31.978604015533335, + "learning_rate": 1.282051282051282e-07, + "loss": 0.9301, + "step": 6 + }, + { + "epoch": 0.00449582530507386, + "grad_norm": 36.722282595711, + "learning_rate": 1.495726495726496e-07, + "loss": 0.8837, + "step": 7 + }, + { + "epoch": 0.005138086062941554, + "grad_norm": 37.54297987885547, + "learning_rate": 1.7094017094017097e-07, + "loss": 0.7932, + "step": 8 + }, + { + "epoch": 0.005780346820809248, + "grad_norm": 33.532768433730475, + "learning_rate": 1.9230769230769234e-07, + "loss": 0.9353, + "step": 9 + }, + { + "epoch": 0.006422607578676943, + "grad_norm": 32.44704896986305, + "learning_rate": 2.136752136752137e-07, + "loss": 0.7511, + "step": 10 + }, + { + "epoch": 0.007064868336544637, + "grad_norm": 32.45586854189174, + "learning_rate": 2.3504273504273505e-07, + "loss": 0.8832, + "step": 11 + }, + { + "epoch": 0.007707129094412331, + "grad_norm": 37.21909470015766, + "learning_rate": 2.564102564102564e-07, + "loss": 0.8438, + "step": 12 + }, + { + "epoch": 0.008349389852280027, + "grad_norm": 30.392063955129288, + "learning_rate": 2.7777777777777776e-07, + "loss": 0.7719, + "step": 13 + }, + { + "epoch": 0.00899165061014772, + "grad_norm": 21.346693252371978, + "learning_rate": 2.991452991452992e-07, + "loss": 0.8025, + "step": 14 + }, + { + "epoch": 0.009633911368015413, + "grad_norm": 24.888078023359032, + "learning_rate": 3.205128205128205e-07, + "loss": 1.0109, + "step": 15 + }, + { + "epoch": 0.010276172125883108, + "grad_norm": 17.000787298052348, + "learning_rate": 3.4188034188034194e-07, + "loss": 0.8232, + "step": 16 + }, + { + "epoch": 0.010918432883750802, + "grad_norm": 17.347936733456407, + "learning_rate": 3.632478632478633e-07, + "loss": 0.9089, + "step": 17 + }, + { + "epoch": 0.011560693641618497, + "grad_norm": 15.871988522478407, + "learning_rate": 3.846153846153847e-07, + "loss": 0.8894, + "step": 18 + }, + { + "epoch": 0.012202954399486191, + "grad_norm": 17.18231446470477, + "learning_rate": 4.05982905982906e-07, + "loss": 0.714, + "step": 19 + }, + { + "epoch": 0.012845215157353885, + "grad_norm": 16.870803101909946, + "learning_rate": 4.273504273504274e-07, + "loss": 0.806, + "step": 20 + }, + { + "epoch": 0.01348747591522158, + "grad_norm": 10.761977633101118, + "learning_rate": 4.4871794871794876e-07, + "loss": 0.7073, + "step": 21 + }, + { + "epoch": 0.014129736673089274, + "grad_norm": 11.51232138469463, + "learning_rate": 4.700854700854701e-07, + "loss": 0.6548, + "step": 22 + }, + { + "epoch": 0.014771997430956968, + "grad_norm": 9.567911545762493, + "learning_rate": 4.914529914529914e-07, + "loss": 0.691, + "step": 23 + }, + { + "epoch": 0.015414258188824663, + "grad_norm": 10.89845526559224, + "learning_rate": 5.128205128205128e-07, + "loss": 0.771, + "step": 24 + }, + { + "epoch": 0.01605651894669236, + "grad_norm": 8.791870275554759, + "learning_rate": 5.341880341880342e-07, + "loss": 0.7682, + "step": 25 + }, + { + "epoch": 0.016698779704560053, + "grad_norm": 9.664810731968254, + "learning_rate": 5.555555555555555e-07, + "loss": 0.5903, + "step": 26 + }, + { + "epoch": 0.017341040462427744, + "grad_norm": 10.910272329104139, + "learning_rate": 5.76923076923077e-07, + "loss": 0.5897, + "step": 27 + }, + { + "epoch": 0.01798330122029544, + "grad_norm": 8.66206870591788, + "learning_rate": 5.982905982905984e-07, + "loss": 0.6694, + "step": 28 + }, + { + "epoch": 0.018625561978163133, + "grad_norm": 7.204536987549252, + "learning_rate": 6.196581196581197e-07, + "loss": 0.6246, + "step": 29 + }, + { + "epoch": 0.019267822736030827, + "grad_norm": 7.257037922653615, + "learning_rate": 6.41025641025641e-07, + "loss": 0.5171, + "step": 30 + }, + { + "epoch": 0.01991008349389852, + "grad_norm": 6.437552649386079, + "learning_rate": 6.623931623931625e-07, + "loss": 0.5833, + "step": 31 + }, + { + "epoch": 0.020552344251766216, + "grad_norm": 6.84638738933486, + "learning_rate": 6.837606837606839e-07, + "loss": 0.4933, + "step": 32 + }, + { + "epoch": 0.02119460500963391, + "grad_norm": 6.053972768128161, + "learning_rate": 7.051282051282052e-07, + "loss": 0.558, + "step": 33 + }, + { + "epoch": 0.021836865767501604, + "grad_norm": 6.231069073349385, + "learning_rate": 7.264957264957266e-07, + "loss": 0.5507, + "step": 34 + }, + { + "epoch": 0.0224791265253693, + "grad_norm": 7.915522502450035, + "learning_rate": 7.478632478632479e-07, + "loss": 0.4972, + "step": 35 + }, + { + "epoch": 0.023121387283236993, + "grad_norm": 6.647877833074881, + "learning_rate": 7.692307692307694e-07, + "loss": 0.5764, + "step": 36 + }, + { + "epoch": 0.023763648041104687, + "grad_norm": 5.246114652695355, + "learning_rate": 7.905982905982906e-07, + "loss": 0.3652, + "step": 37 + }, + { + "epoch": 0.024405908798972382, + "grad_norm": 6.12624663479125, + "learning_rate": 8.11965811965812e-07, + "loss": 0.5029, + "step": 38 + }, + { + "epoch": 0.025048169556840076, + "grad_norm": 5.12762636227482, + "learning_rate": 8.333333333333333e-07, + "loss": 0.4538, + "step": 39 + }, + { + "epoch": 0.02569043031470777, + "grad_norm": 5.083673741306729, + "learning_rate": 8.547008547008548e-07, + "loss": 0.3536, + "step": 40 + }, + { + "epoch": 0.026332691072575465, + "grad_norm": 4.856620509782038, + "learning_rate": 8.760683760683761e-07, + "loss": 0.2366, + "step": 41 + }, + { + "epoch": 0.02697495183044316, + "grad_norm": 4.589799584307676, + "learning_rate": 8.974358974358975e-07, + "loss": 0.3751, + "step": 42 + }, + { + "epoch": 0.027617212588310854, + "grad_norm": 4.694131621092587, + "learning_rate": 9.188034188034189e-07, + "loss": 0.2916, + "step": 43 + }, + { + "epoch": 0.028259473346178548, + "grad_norm": 4.366662489213961, + "learning_rate": 9.401709401709402e-07, + "loss": 0.3179, + "step": 44 + }, + { + "epoch": 0.028901734104046242, + "grad_norm": 4.156908696758279, + "learning_rate": 9.615384615384617e-07, + "loss": 0.2808, + "step": 45 + }, + { + "epoch": 0.029543994861913937, + "grad_norm": 3.8919619862564008, + "learning_rate": 9.829059829059829e-07, + "loss": 0.2242, + "step": 46 + }, + { + "epoch": 0.03018625561978163, + "grad_norm": 3.466613359537815, + "learning_rate": 1.0042735042735045e-06, + "loss": 0.2154, + "step": 47 + }, + { + "epoch": 0.030828516377649325, + "grad_norm": 3.510017002978137, + "learning_rate": 1.0256410256410257e-06, + "loss": 0.2011, + "step": 48 + }, + { + "epoch": 0.03147077713551702, + "grad_norm": 4.688924237460008, + "learning_rate": 1.047008547008547e-06, + "loss": 0.3437, + "step": 49 + }, + { + "epoch": 0.03211303789338472, + "grad_norm": 4.116392722752709, + "learning_rate": 1.0683760683760685e-06, + "loss": 0.2274, + "step": 50 + }, + { + "epoch": 0.03275529865125241, + "grad_norm": 3.2874856451887275, + "learning_rate": 1.0897435897435899e-06, + "loss": 0.1667, + "step": 51 + }, + { + "epoch": 0.033397559409120106, + "grad_norm": 3.2054734938689875, + "learning_rate": 1.111111111111111e-06, + "loss": 0.1143, + "step": 52 + }, + { + "epoch": 0.0340398201669878, + "grad_norm": 4.130128310705924, + "learning_rate": 1.1324786324786326e-06, + "loss": 0.2945, + "step": 53 + }, + { + "epoch": 0.03468208092485549, + "grad_norm": 2.8720936551230496, + "learning_rate": 1.153846153846154e-06, + "loss": 0.1716, + "step": 54 + }, + { + "epoch": 0.035324341682723186, + "grad_norm": 3.503252662389985, + "learning_rate": 1.1752136752136752e-06, + "loss": 0.1672, + "step": 55 + }, + { + "epoch": 0.03596660244059088, + "grad_norm": 3.341332977498924, + "learning_rate": 1.1965811965811968e-06, + "loss": 0.2963, + "step": 56 + }, + { + "epoch": 0.036608863198458574, + "grad_norm": 3.2451060769769065, + "learning_rate": 1.217948717948718e-06, + "loss": 0.2387, + "step": 57 + }, + { + "epoch": 0.037251123956326265, + "grad_norm": 3.0762271945640363, + "learning_rate": 1.2393162393162394e-06, + "loss": 0.2973, + "step": 58 + }, + { + "epoch": 0.03789338471419396, + "grad_norm": 2.29247026903211, + "learning_rate": 1.2606837606837608e-06, + "loss": 0.0969, + "step": 59 + }, + { + "epoch": 0.038535645472061654, + "grad_norm": 3.7483830870589174, + "learning_rate": 1.282051282051282e-06, + "loss": 0.185, + "step": 60 + }, + { + "epoch": 0.03917790622992935, + "grad_norm": 2.713638297351357, + "learning_rate": 1.3034188034188036e-06, + "loss": 0.1599, + "step": 61 + }, + { + "epoch": 0.03982016698779704, + "grad_norm": 2.6910639940045327, + "learning_rate": 1.324786324786325e-06, + "loss": 0.1703, + "step": 62 + }, + { + "epoch": 0.04046242774566474, + "grad_norm": 2.430564855900542, + "learning_rate": 1.3461538461538462e-06, + "loss": 0.1175, + "step": 63 + }, + { + "epoch": 0.04110468850353243, + "grad_norm": 1.979220672171525, + "learning_rate": 1.3675213675213678e-06, + "loss": 0.1108, + "step": 64 + }, + { + "epoch": 0.04174694926140013, + "grad_norm": 2.5174913732390793, + "learning_rate": 1.3888888888888892e-06, + "loss": 0.1147, + "step": 65 + }, + { + "epoch": 0.04238921001926782, + "grad_norm": 2.598003144259982, + "learning_rate": 1.4102564102564104e-06, + "loss": 0.2354, + "step": 66 + }, + { + "epoch": 0.04303147077713552, + "grad_norm": 2.3242669945435384, + "learning_rate": 1.4316239316239317e-06, + "loss": 0.1215, + "step": 67 + }, + { + "epoch": 0.04367373153500321, + "grad_norm": 2.3027095564077533, + "learning_rate": 1.4529914529914531e-06, + "loss": 0.1223, + "step": 68 + }, + { + "epoch": 0.04431599229287091, + "grad_norm": 3.2407908079258227, + "learning_rate": 1.4743589743589745e-06, + "loss": 0.2775, + "step": 69 + }, + { + "epoch": 0.0449582530507386, + "grad_norm": 2.696160743533787, + "learning_rate": 1.4957264957264957e-06, + "loss": 0.2421, + "step": 70 + }, + { + "epoch": 0.045600513808606295, + "grad_norm": 1.9391359194557212, + "learning_rate": 1.5170940170940171e-06, + "loss": 0.1179, + "step": 71 + }, + { + "epoch": 0.046242774566473986, + "grad_norm": 2.378611993063526, + "learning_rate": 1.5384615384615387e-06, + "loss": 0.0984, + "step": 72 + }, + { + "epoch": 0.046885035324341684, + "grad_norm": 2.5508985936569273, + "learning_rate": 1.55982905982906e-06, + "loss": 0.1715, + "step": 73 + }, + { + "epoch": 0.047527296082209375, + "grad_norm": 2.551878251863097, + "learning_rate": 1.5811965811965813e-06, + "loss": 0.1268, + "step": 74 + }, + { + "epoch": 0.04816955684007707, + "grad_norm": 2.8587654019658437, + "learning_rate": 1.602564102564103e-06, + "loss": 0.1961, + "step": 75 + }, + { + "epoch": 0.048811817597944764, + "grad_norm": 2.6628398689761834, + "learning_rate": 1.623931623931624e-06, + "loss": 0.1741, + "step": 76 + }, + { + "epoch": 0.04945407835581246, + "grad_norm": 3.6978803092492214, + "learning_rate": 1.6452991452991455e-06, + "loss": 0.2075, + "step": 77 + }, + { + "epoch": 0.05009633911368015, + "grad_norm": 2.4801266652239256, + "learning_rate": 1.6666666666666667e-06, + "loss": 0.1289, + "step": 78 + }, + { + "epoch": 0.05073859987154785, + "grad_norm": 2.283502467913068, + "learning_rate": 1.6880341880341883e-06, + "loss": 0.1686, + "step": 79 + }, + { + "epoch": 0.05138086062941554, + "grad_norm": 2.5801829934501006, + "learning_rate": 1.7094017094017097e-06, + "loss": 0.2076, + "step": 80 + }, + { + "epoch": 0.05202312138728324, + "grad_norm": 2.4323959667830537, + "learning_rate": 1.7307692307692308e-06, + "loss": 0.2343, + "step": 81 + }, + { + "epoch": 0.05266538214515093, + "grad_norm": 2.2440963200922868, + "learning_rate": 1.7521367521367522e-06, + "loss": 0.1649, + "step": 82 + }, + { + "epoch": 0.05330764290301863, + "grad_norm": 2.0381190988472317, + "learning_rate": 1.7735042735042736e-06, + "loss": 0.1448, + "step": 83 + }, + { + "epoch": 0.05394990366088632, + "grad_norm": 2.3301801679161414, + "learning_rate": 1.794871794871795e-06, + "loss": 0.172, + "step": 84 + }, + { + "epoch": 0.054592164418754016, + "grad_norm": 2.9483854190091714, + "learning_rate": 1.8162393162393164e-06, + "loss": 0.2127, + "step": 85 + }, + { + "epoch": 0.05523442517662171, + "grad_norm": 2.691987087341222, + "learning_rate": 1.8376068376068378e-06, + "loss": 0.1748, + "step": 86 + }, + { + "epoch": 0.055876685934489405, + "grad_norm": 2.3248580845918814, + "learning_rate": 1.8589743589743592e-06, + "loss": 0.1547, + "step": 87 + }, + { + "epoch": 0.056518946692357096, + "grad_norm": 2.0420649490192586, + "learning_rate": 1.8803418803418804e-06, + "loss": 0.1499, + "step": 88 + }, + { + "epoch": 0.057161207450224794, + "grad_norm": 2.4769362259843675, + "learning_rate": 1.9017094017094018e-06, + "loss": 0.1912, + "step": 89 + }, + { + "epoch": 0.057803468208092484, + "grad_norm": 2.249712526265293, + "learning_rate": 1.9230769230769234e-06, + "loss": 0.1575, + "step": 90 + }, + { + "epoch": 0.05844572896596018, + "grad_norm": 2.0108836603533757, + "learning_rate": 1.944444444444445e-06, + "loss": 0.1715, + "step": 91 + }, + { + "epoch": 0.05908798972382787, + "grad_norm": 1.6473269625078992, + "learning_rate": 1.9658119658119658e-06, + "loss": 0.0922, + "step": 92 + }, + { + "epoch": 0.05973025048169557, + "grad_norm": 3.2467145432264184, + "learning_rate": 1.987179487179487e-06, + "loss": 0.1805, + "step": 93 + }, + { + "epoch": 0.06037251123956326, + "grad_norm": 1.8798932771824444, + "learning_rate": 2.008547008547009e-06, + "loss": 0.1375, + "step": 94 + }, + { + "epoch": 0.06101477199743096, + "grad_norm": 2.3718840006195294, + "learning_rate": 2.02991452991453e-06, + "loss": 0.201, + "step": 95 + }, + { + "epoch": 0.06165703275529865, + "grad_norm": 2.0986421379794455, + "learning_rate": 2.0512820512820513e-06, + "loss": 0.1483, + "step": 96 + }, + { + "epoch": 0.06229929351316635, + "grad_norm": 2.5214274967718002, + "learning_rate": 2.072649572649573e-06, + "loss": 0.2154, + "step": 97 + }, + { + "epoch": 0.06294155427103404, + "grad_norm": 2.2019744100918306, + "learning_rate": 2.094017094017094e-06, + "loss": 0.2204, + "step": 98 + }, + { + "epoch": 0.06358381502890173, + "grad_norm": 2.3054441821399716, + "learning_rate": 2.1153846153846155e-06, + "loss": 0.1584, + "step": 99 + }, + { + "epoch": 0.06422607578676943, + "grad_norm": 2.5626121120028644, + "learning_rate": 2.136752136752137e-06, + "loss": 0.3444, + "step": 100 + }, + { + "epoch": 0.06486833654463713, + "grad_norm": 2.393468174154522, + "learning_rate": 2.1581196581196583e-06, + "loss": 0.2468, + "step": 101 + }, + { + "epoch": 0.06551059730250482, + "grad_norm": 2.2344945848441538, + "learning_rate": 2.1794871794871797e-06, + "loss": 0.1022, + "step": 102 + }, + { + "epoch": 0.06615285806037251, + "grad_norm": 2.0044381429234424, + "learning_rate": 2.200854700854701e-06, + "loss": 0.122, + "step": 103 + }, + { + "epoch": 0.06679511881824021, + "grad_norm": 1.960459600252525, + "learning_rate": 2.222222222222222e-06, + "loss": 0.0943, + "step": 104 + }, + { + "epoch": 0.0674373795761079, + "grad_norm": 2.013018792147706, + "learning_rate": 2.243589743589744e-06, + "loss": 0.1801, + "step": 105 + }, + { + "epoch": 0.0680796403339756, + "grad_norm": 2.6336286120275045, + "learning_rate": 2.2649572649572653e-06, + "loss": 0.1518, + "step": 106 + }, + { + "epoch": 0.06872190109184328, + "grad_norm": 2.3462492482310062, + "learning_rate": 2.2863247863247863e-06, + "loss": 0.2332, + "step": 107 + }, + { + "epoch": 0.06936416184971098, + "grad_norm": 2.035147822288555, + "learning_rate": 2.307692307692308e-06, + "loss": 0.1448, + "step": 108 + }, + { + "epoch": 0.07000642260757868, + "grad_norm": 2.070809389869112, + "learning_rate": 2.3290598290598295e-06, + "loss": 0.2116, + "step": 109 + }, + { + "epoch": 0.07064868336544637, + "grad_norm": 1.7305906653720025, + "learning_rate": 2.3504273504273504e-06, + "loss": 0.1148, + "step": 110 + }, + { + "epoch": 0.07129094412331406, + "grad_norm": 2.153498989419596, + "learning_rate": 2.371794871794872e-06, + "loss": 0.1406, + "step": 111 + }, + { + "epoch": 0.07193320488118175, + "grad_norm": 1.7212044122854997, + "learning_rate": 2.3931623931623937e-06, + "loss": 0.0919, + "step": 112 + }, + { + "epoch": 0.07257546563904946, + "grad_norm": 3.261405345750642, + "learning_rate": 2.4145299145299146e-06, + "loss": 0.1227, + "step": 113 + }, + { + "epoch": 0.07321772639691715, + "grad_norm": 1.7287494872695404, + "learning_rate": 2.435897435897436e-06, + "loss": 0.1452, + "step": 114 + }, + { + "epoch": 0.07385998715478484, + "grad_norm": 1.8556796686042356, + "learning_rate": 2.4572649572649574e-06, + "loss": 0.0906, + "step": 115 + }, + { + "epoch": 0.07450224791265253, + "grad_norm": 1.9510469582311796, + "learning_rate": 2.478632478632479e-06, + "loss": 0.1656, + "step": 116 + }, + { + "epoch": 0.07514450867052024, + "grad_norm": 1.940414188359992, + "learning_rate": 2.5e-06, + "loss": 0.1131, + "step": 117 + }, + { + "epoch": 0.07578676942838793, + "grad_norm": 1.8157991481791267, + "learning_rate": 2.5213675213675216e-06, + "loss": 0.111, + "step": 118 + }, + { + "epoch": 0.07642903018625562, + "grad_norm": 2.174510625555878, + "learning_rate": 2.542735042735043e-06, + "loss": 0.1458, + "step": 119 + }, + { + "epoch": 0.07707129094412331, + "grad_norm": 2.2356956922388287, + "learning_rate": 2.564102564102564e-06, + "loss": 0.1546, + "step": 120 + }, + { + "epoch": 0.07771355170199101, + "grad_norm": 3.4163808844665193, + "learning_rate": 2.5854700854700858e-06, + "loss": 0.1641, + "step": 121 + }, + { + "epoch": 0.0783558124598587, + "grad_norm": 1.979847787695695, + "learning_rate": 2.606837606837607e-06, + "loss": 0.1635, + "step": 122 + }, + { + "epoch": 0.0789980732177264, + "grad_norm": 1.667018750980996, + "learning_rate": 2.6282051282051286e-06, + "loss": 0.0851, + "step": 123 + }, + { + "epoch": 0.07964033397559409, + "grad_norm": 2.280950468510416, + "learning_rate": 2.64957264957265e-06, + "loss": 0.1275, + "step": 124 + }, + { + "epoch": 0.08028259473346179, + "grad_norm": 2.098402240828742, + "learning_rate": 2.670940170940171e-06, + "loss": 0.2196, + "step": 125 + }, + { + "epoch": 0.08092485549132948, + "grad_norm": 2.4733244699518777, + "learning_rate": 2.6923076923076923e-06, + "loss": 0.2235, + "step": 126 + }, + { + "epoch": 0.08156711624919717, + "grad_norm": 1.7223502653799212, + "learning_rate": 2.7136752136752137e-06, + "loss": 0.0931, + "step": 127 + }, + { + "epoch": 0.08220937700706486, + "grad_norm": 1.4128394393212629, + "learning_rate": 2.7350427350427355e-06, + "loss": 0.0668, + "step": 128 + }, + { + "epoch": 0.08285163776493257, + "grad_norm": 1.7842381441066664, + "learning_rate": 2.756410256410257e-06, + "loss": 0.1133, + "step": 129 + }, + { + "epoch": 0.08349389852280026, + "grad_norm": 2.236571069142531, + "learning_rate": 2.7777777777777783e-06, + "loss": 0.1889, + "step": 130 + }, + { + "epoch": 0.08413615928066795, + "grad_norm": 1.920631905288938, + "learning_rate": 2.7991452991452993e-06, + "loss": 0.1124, + "step": 131 + }, + { + "epoch": 0.08477842003853564, + "grad_norm": 1.8948295202312186, + "learning_rate": 2.8205128205128207e-06, + "loss": 0.1086, + "step": 132 + }, + { + "epoch": 0.08542068079640334, + "grad_norm": 1.729465042473488, + "learning_rate": 2.841880341880342e-06, + "loss": 0.1558, + "step": 133 + }, + { + "epoch": 0.08606294155427104, + "grad_norm": 4.4564741023823045, + "learning_rate": 2.8632478632478635e-06, + "loss": 0.0906, + "step": 134 + }, + { + "epoch": 0.08670520231213873, + "grad_norm": 1.6043924671112675, + "learning_rate": 2.8846153846153845e-06, + "loss": 0.0893, + "step": 135 + }, + { + "epoch": 0.08734746307000642, + "grad_norm": 1.792554886381474, + "learning_rate": 2.9059829059829063e-06, + "loss": 0.1029, + "step": 136 + }, + { + "epoch": 0.08798972382787412, + "grad_norm": 2.060841639439536, + "learning_rate": 2.9273504273504277e-06, + "loss": 0.1711, + "step": 137 + }, + { + "epoch": 0.08863198458574181, + "grad_norm": 1.9647360594524994, + "learning_rate": 2.948717948717949e-06, + "loss": 0.2083, + "step": 138 + }, + { + "epoch": 0.0892742453436095, + "grad_norm": 1.5075697234085015, + "learning_rate": 2.9700854700854705e-06, + "loss": 0.0976, + "step": 139 + }, + { + "epoch": 0.0899165061014772, + "grad_norm": 1.945935371056808, + "learning_rate": 2.9914529914529914e-06, + "loss": 0.2101, + "step": 140 + }, + { + "epoch": 0.0905587668593449, + "grad_norm": 1.8279252591585475, + "learning_rate": 3.012820512820513e-06, + "loss": 0.1832, + "step": 141 + }, + { + "epoch": 0.09120102761721259, + "grad_norm": 1.9651116329510832, + "learning_rate": 3.0341880341880342e-06, + "loss": 0.2119, + "step": 142 + }, + { + "epoch": 0.09184328837508028, + "grad_norm": 1.5130420985242192, + "learning_rate": 3.055555555555556e-06, + "loss": 0.0843, + "step": 143 + }, + { + "epoch": 0.09248554913294797, + "grad_norm": 2.29372462365056, + "learning_rate": 3.0769230769230774e-06, + "loss": 0.2519, + "step": 144 + }, + { + "epoch": 0.09312780989081568, + "grad_norm": 1.848544542333652, + "learning_rate": 3.098290598290599e-06, + "loss": 0.179, + "step": 145 + }, + { + "epoch": 0.09377007064868337, + "grad_norm": 1.5365220465489862, + "learning_rate": 3.11965811965812e-06, + "loss": 0.1309, + "step": 146 + }, + { + "epoch": 0.09441233140655106, + "grad_norm": 2.2012519353847466, + "learning_rate": 3.141025641025641e-06, + "loss": 0.2352, + "step": 147 + }, + { + "epoch": 0.09505459216441875, + "grad_norm": 2.080377950107224, + "learning_rate": 3.1623931623931626e-06, + "loss": 0.1189, + "step": 148 + }, + { + "epoch": 0.09569685292228645, + "grad_norm": 1.519857922262211, + "learning_rate": 3.183760683760684e-06, + "loss": 0.0945, + "step": 149 + }, + { + "epoch": 0.09633911368015415, + "grad_norm": 1.5841948321405324, + "learning_rate": 3.205128205128206e-06, + "loss": 0.1135, + "step": 150 + }, + { + "epoch": 0.09698137443802184, + "grad_norm": 1.8414122030936437, + "learning_rate": 3.2264957264957268e-06, + "loss": 0.1584, + "step": 151 + }, + { + "epoch": 0.09762363519588953, + "grad_norm": 1.7719909314576712, + "learning_rate": 3.247863247863248e-06, + "loss": 0.1243, + "step": 152 + }, + { + "epoch": 0.09826589595375723, + "grad_norm": 1.598755407729178, + "learning_rate": 3.2692307692307696e-06, + "loss": 0.1104, + "step": 153 + }, + { + "epoch": 0.09890815671162492, + "grad_norm": 1.787293353969307, + "learning_rate": 3.290598290598291e-06, + "loss": 0.1741, + "step": 154 + }, + { + "epoch": 0.09955041746949261, + "grad_norm": 1.7022944386458903, + "learning_rate": 3.311965811965812e-06, + "loss": 0.1329, + "step": 155 + }, + { + "epoch": 0.1001926782273603, + "grad_norm": 2.4198012657115178, + "learning_rate": 3.3333333333333333e-06, + "loss": 0.3257, + "step": 156 + }, + { + "epoch": 0.10083493898522801, + "grad_norm": 1.8617510959987191, + "learning_rate": 3.3547008547008547e-06, + "loss": 0.1441, + "step": 157 + }, + { + "epoch": 0.1014771997430957, + "grad_norm": 1.4471085657339249, + "learning_rate": 3.3760683760683765e-06, + "loss": 0.0772, + "step": 158 + }, + { + "epoch": 0.10211946050096339, + "grad_norm": 2.3713219921358806, + "learning_rate": 3.397435897435898e-06, + "loss": 0.099, + "step": 159 + }, + { + "epoch": 0.10276172125883108, + "grad_norm": 1.4956781610458112, + "learning_rate": 3.4188034188034193e-06, + "loss": 0.114, + "step": 160 + }, + { + "epoch": 0.10340398201669879, + "grad_norm": 1.9198892943214496, + "learning_rate": 3.4401709401709403e-06, + "loss": 0.2583, + "step": 161 + }, + { + "epoch": 0.10404624277456648, + "grad_norm": 1.648926323879495, + "learning_rate": 3.4615384615384617e-06, + "loss": 0.1207, + "step": 162 + }, + { + "epoch": 0.10468850353243417, + "grad_norm": 1.6766852972531172, + "learning_rate": 3.482905982905983e-06, + "loss": 0.1426, + "step": 163 + }, + { + "epoch": 0.10533076429030186, + "grad_norm": 1.862773934401278, + "learning_rate": 3.5042735042735045e-06, + "loss": 0.142, + "step": 164 + }, + { + "epoch": 0.10597302504816955, + "grad_norm": 1.3957400438952294, + "learning_rate": 3.5256410256410263e-06, + "loss": 0.0538, + "step": 165 + }, + { + "epoch": 0.10661528580603725, + "grad_norm": 1.585426607808311, + "learning_rate": 3.5470085470085473e-06, + "loss": 0.1415, + "step": 166 + }, + { + "epoch": 0.10725754656390495, + "grad_norm": 2.297839563337332, + "learning_rate": 3.5683760683760687e-06, + "loss": 0.1839, + "step": 167 + }, + { + "epoch": 0.10789980732177264, + "grad_norm": 1.5126184529621065, + "learning_rate": 3.58974358974359e-06, + "loss": 0.1147, + "step": 168 + }, + { + "epoch": 0.10854206807964033, + "grad_norm": 1.3768981083324678, + "learning_rate": 3.6111111111111115e-06, + "loss": 0.0982, + "step": 169 + }, + { + "epoch": 0.10918432883750803, + "grad_norm": 1.4637676787476834, + "learning_rate": 3.632478632478633e-06, + "loss": 0.0817, + "step": 170 + }, + { + "epoch": 0.10982658959537572, + "grad_norm": 1.717327394729676, + "learning_rate": 3.653846153846154e-06, + "loss": 0.1729, + "step": 171 + }, + { + "epoch": 0.11046885035324341, + "grad_norm": 1.6598260629755948, + "learning_rate": 3.6752136752136756e-06, + "loss": 0.1239, + "step": 172 + }, + { + "epoch": 0.1111111111111111, + "grad_norm": 1.4969181344966704, + "learning_rate": 3.696581196581197e-06, + "loss": 0.121, + "step": 173 + }, + { + "epoch": 0.11175337186897881, + "grad_norm": 1.434763460049462, + "learning_rate": 3.7179487179487184e-06, + "loss": 0.0696, + "step": 174 + }, + { + "epoch": 0.1123956326268465, + "grad_norm": 1.9555253622140758, + "learning_rate": 3.73931623931624e-06, + "loss": 0.1864, + "step": 175 + }, + { + "epoch": 0.11303789338471419, + "grad_norm": 1.6759355309139339, + "learning_rate": 3.760683760683761e-06, + "loss": 0.1236, + "step": 176 + }, + { + "epoch": 0.11368015414258188, + "grad_norm": 2.1313224943165503, + "learning_rate": 3.782051282051282e-06, + "loss": 0.2276, + "step": 177 + }, + { + "epoch": 0.11432241490044959, + "grad_norm": 1.6650772801971252, + "learning_rate": 3.8034188034188036e-06, + "loss": 0.1558, + "step": 178 + }, + { + "epoch": 0.11496467565831728, + "grad_norm": 1.726881841057934, + "learning_rate": 3.8247863247863246e-06, + "loss": 0.137, + "step": 179 + }, + { + "epoch": 0.11560693641618497, + "grad_norm": 1.8217133376644281, + "learning_rate": 3.846153846153847e-06, + "loss": 0.1272, + "step": 180 + }, + { + "epoch": 0.11624919717405266, + "grad_norm": 1.8804213210960286, + "learning_rate": 3.867521367521368e-06, + "loss": 0.1604, + "step": 181 + }, + { + "epoch": 0.11689145793192036, + "grad_norm": 1.580234231383413, + "learning_rate": 3.88888888888889e-06, + "loss": 0.0961, + "step": 182 + }, + { + "epoch": 0.11753371868978806, + "grad_norm": 1.9667580463739611, + "learning_rate": 3.910256410256411e-06, + "loss": 0.1666, + "step": 183 + }, + { + "epoch": 0.11817597944765575, + "grad_norm": 1.4936629668468109, + "learning_rate": 3.9316239316239315e-06, + "loss": 0.1081, + "step": 184 + }, + { + "epoch": 0.11881824020552344, + "grad_norm": 1.8766404510052073, + "learning_rate": 3.952991452991453e-06, + "loss": 0.1733, + "step": 185 + }, + { + "epoch": 0.11946050096339114, + "grad_norm": 1.1813276962057888, + "learning_rate": 3.974358974358974e-06, + "loss": 0.0761, + "step": 186 + }, + { + "epoch": 0.12010276172125883, + "grad_norm": 1.625503347030402, + "learning_rate": 3.9957264957264966e-06, + "loss": 0.173, + "step": 187 + }, + { + "epoch": 0.12074502247912652, + "grad_norm": 2.1234155070471883, + "learning_rate": 4.017094017094018e-06, + "loss": 0.196, + "step": 188 + }, + { + "epoch": 0.12138728323699421, + "grad_norm": 1.6041520972629528, + "learning_rate": 4.0384615384615385e-06, + "loss": 0.1259, + "step": 189 + }, + { + "epoch": 0.12202954399486192, + "grad_norm": 1.7415486343233557, + "learning_rate": 4.05982905982906e-06, + "loss": 0.1527, + "step": 190 + }, + { + "epoch": 0.12267180475272961, + "grad_norm": 1.2871025441427424, + "learning_rate": 4.081196581196581e-06, + "loss": 0.0998, + "step": 191 + }, + { + "epoch": 0.1233140655105973, + "grad_norm": 1.9754689362913986, + "learning_rate": 4.102564102564103e-06, + "loss": 0.2126, + "step": 192 + }, + { + "epoch": 0.12395632626846499, + "grad_norm": 1.3176965649979262, + "learning_rate": 4.123931623931624e-06, + "loss": 0.0744, + "step": 193 + }, + { + "epoch": 0.1245985870263327, + "grad_norm": 1.7737230091359477, + "learning_rate": 4.145299145299146e-06, + "loss": 0.1948, + "step": 194 + }, + { + "epoch": 0.1252408477842004, + "grad_norm": 1.356427814430456, + "learning_rate": 4.166666666666667e-06, + "loss": 0.0859, + "step": 195 + }, + { + "epoch": 0.12588310854206808, + "grad_norm": 3.1356513989906847, + "learning_rate": 4.188034188034188e-06, + "loss": 0.1072, + "step": 196 + }, + { + "epoch": 0.12652536929993577, + "grad_norm": 2.99206479138713, + "learning_rate": 4.20940170940171e-06, + "loss": 0.2567, + "step": 197 + }, + { + "epoch": 0.12716763005780346, + "grad_norm": 1.3712704588660682, + "learning_rate": 4.230769230769231e-06, + "loss": 0.0667, + "step": 198 + }, + { + "epoch": 0.12780989081567115, + "grad_norm": 1.6583969845861979, + "learning_rate": 4.2521367521367524e-06, + "loss": 0.1099, + "step": 199 + }, + { + "epoch": 0.12845215157353887, + "grad_norm": 1.7795450131784256, + "learning_rate": 4.273504273504274e-06, + "loss": 0.1157, + "step": 200 + }, + { + "epoch": 0.12909441233140656, + "grad_norm": 1.9539431330597965, + "learning_rate": 4.294871794871795e-06, + "loss": 0.1799, + "step": 201 + }, + { + "epoch": 0.12973667308927425, + "grad_norm": 1.638613905658282, + "learning_rate": 4.316239316239317e-06, + "loss": 0.1245, + "step": 202 + }, + { + "epoch": 0.13037893384714194, + "grad_norm": 1.9868168389836431, + "learning_rate": 4.337606837606838e-06, + "loss": 0.1772, + "step": 203 + }, + { + "epoch": 0.13102119460500963, + "grad_norm": 1.8617868502553154, + "learning_rate": 4.358974358974359e-06, + "loss": 0.1693, + "step": 204 + }, + { + "epoch": 0.13166345536287732, + "grad_norm": 1.6509896435823757, + "learning_rate": 4.380341880341881e-06, + "loss": 0.1306, + "step": 205 + }, + { + "epoch": 0.13230571612074501, + "grad_norm": 1.5076348088753515, + "learning_rate": 4.401709401709402e-06, + "loss": 0.148, + "step": 206 + }, + { + "epoch": 0.1329479768786127, + "grad_norm": 1.400500325750214, + "learning_rate": 4.423076923076924e-06, + "loss": 0.086, + "step": 207 + }, + { + "epoch": 0.13359023763648042, + "grad_norm": 1.529696025136291, + "learning_rate": 4.444444444444444e-06, + "loss": 0.1032, + "step": 208 + }, + { + "epoch": 0.13423249839434812, + "grad_norm": 2.0916751119510346, + "learning_rate": 4.465811965811966e-06, + "loss": 0.2442, + "step": 209 + }, + { + "epoch": 0.1348747591522158, + "grad_norm": 1.872168665102132, + "learning_rate": 4.487179487179488e-06, + "loss": 0.2411, + "step": 210 + }, + { + "epoch": 0.1355170199100835, + "grad_norm": 2.072214862437571, + "learning_rate": 4.508547008547009e-06, + "loss": 0.1627, + "step": 211 + }, + { + "epoch": 0.1361592806679512, + "grad_norm": 1.9546646303566961, + "learning_rate": 4.5299145299145306e-06, + "loss": 0.164, + "step": 212 + }, + { + "epoch": 0.13680154142581888, + "grad_norm": 1.90752523631071, + "learning_rate": 4.551282051282052e-06, + "loss": 0.1139, + "step": 213 + }, + { + "epoch": 0.13744380218368657, + "grad_norm": 2.3848903185647883, + "learning_rate": 4.5726495726495725e-06, + "loss": 0.1834, + "step": 214 + }, + { + "epoch": 0.13808606294155426, + "grad_norm": 1.7360179924642094, + "learning_rate": 4.594017094017094e-06, + "loss": 0.1449, + "step": 215 + }, + { + "epoch": 0.13872832369942195, + "grad_norm": 1.8639447067373993, + "learning_rate": 4.615384615384616e-06, + "loss": 0.1657, + "step": 216 + }, + { + "epoch": 0.13937058445728967, + "grad_norm": 2.0681033317171194, + "learning_rate": 4.6367521367521375e-06, + "loss": 0.169, + "step": 217 + }, + { + "epoch": 0.14001284521515736, + "grad_norm": 1.3511779535908643, + "learning_rate": 4.658119658119659e-06, + "loss": 0.0843, + "step": 218 + }, + { + "epoch": 0.14065510597302505, + "grad_norm": 3.5613163328674973, + "learning_rate": 4.6794871794871795e-06, + "loss": 0.1523, + "step": 219 + }, + { + "epoch": 0.14129736673089274, + "grad_norm": 1.4149225098520677, + "learning_rate": 4.700854700854701e-06, + "loss": 0.1066, + "step": 220 + }, + { + "epoch": 0.14193962748876043, + "grad_norm": 1.499363363541115, + "learning_rate": 4.722222222222222e-06, + "loss": 0.1018, + "step": 221 + }, + { + "epoch": 0.14258188824662812, + "grad_norm": 1.6486181628830234, + "learning_rate": 4.743589743589744e-06, + "loss": 0.2188, + "step": 222 + }, + { + "epoch": 0.14322414900449582, + "grad_norm": 2.062862510421735, + "learning_rate": 4.764957264957265e-06, + "loss": 0.3651, + "step": 223 + }, + { + "epoch": 0.1438664097623635, + "grad_norm": 1.567189768643777, + "learning_rate": 4.786324786324787e-06, + "loss": 0.1457, + "step": 224 + }, + { + "epoch": 0.14450867052023122, + "grad_norm": 1.9622014959984064, + "learning_rate": 4.807692307692308e-06, + "loss": 0.2455, + "step": 225 + }, + { + "epoch": 0.14515093127809892, + "grad_norm": 1.635370501150973, + "learning_rate": 4.829059829059829e-06, + "loss": 0.1638, + "step": 226 + }, + { + "epoch": 0.1457931920359666, + "grad_norm": 1.899711700056473, + "learning_rate": 4.850427350427351e-06, + "loss": 0.1345, + "step": 227 + }, + { + "epoch": 0.1464354527938343, + "grad_norm": 1.5292574901121858, + "learning_rate": 4.871794871794872e-06, + "loss": 0.1256, + "step": 228 + }, + { + "epoch": 0.147077713551702, + "grad_norm": 1.5988875780637921, + "learning_rate": 4.8931623931623934e-06, + "loss": 0.1329, + "step": 229 + }, + { + "epoch": 0.14771997430956968, + "grad_norm": 1.680962677214432, + "learning_rate": 4.914529914529915e-06, + "loss": 0.133, + "step": 230 + }, + { + "epoch": 0.14836223506743737, + "grad_norm": 1.661184530347413, + "learning_rate": 4.935897435897436e-06, + "loss": 0.1077, + "step": 231 + }, + { + "epoch": 0.14900449582530506, + "grad_norm": 1.7646340077369083, + "learning_rate": 4.957264957264958e-06, + "loss": 0.1726, + "step": 232 + }, + { + "epoch": 0.14964675658317278, + "grad_norm": 1.681590689278435, + "learning_rate": 4.978632478632479e-06, + "loss": 0.1435, + "step": 233 + }, + { + "epoch": 0.15028901734104047, + "grad_norm": 1.3720044352947176, + "learning_rate": 5e-06, + "loss": 0.0658, + "step": 234 + }, + { + "epoch": 0.15093127809890816, + "grad_norm": 1.5848386678848516, + "learning_rate": 5.021367521367522e-06, + "loss": 0.1077, + "step": 235 + }, + { + "epoch": 0.15157353885677585, + "grad_norm": 1.7575285649029946, + "learning_rate": 5.042735042735043e-06, + "loss": 0.1734, + "step": 236 + }, + { + "epoch": 0.15221579961464354, + "grad_norm": 1.5083555427741269, + "learning_rate": 5.064102564102565e-06, + "loss": 0.1487, + "step": 237 + }, + { + "epoch": 0.15285806037251123, + "grad_norm": 1.7224145948920861, + "learning_rate": 5.085470085470086e-06, + "loss": 0.1588, + "step": 238 + }, + { + "epoch": 0.15350032113037893, + "grad_norm": 1.2716102942297274, + "learning_rate": 5.1068376068376065e-06, + "loss": 0.0658, + "step": 239 + }, + { + "epoch": 0.15414258188824662, + "grad_norm": 2.072265069171194, + "learning_rate": 5.128205128205128e-06, + "loss": 0.1855, + "step": 240 + }, + { + "epoch": 0.15478484264611433, + "grad_norm": 1.18830328169265, + "learning_rate": 5.149572649572649e-06, + "loss": 0.0919, + "step": 241 + }, + { + "epoch": 0.15542710340398203, + "grad_norm": 1.5456195792794907, + "learning_rate": 5.1709401709401716e-06, + "loss": 0.1302, + "step": 242 + }, + { + "epoch": 0.15606936416184972, + "grad_norm": 1.7610893893512871, + "learning_rate": 5.192307692307693e-06, + "loss": 0.1754, + "step": 243 + }, + { + "epoch": 0.1567116249197174, + "grad_norm": 1.6113441243770732, + "learning_rate": 5.213675213675214e-06, + "loss": 0.1537, + "step": 244 + }, + { + "epoch": 0.1573538856775851, + "grad_norm": 1.2970707266466295, + "learning_rate": 5.235042735042736e-06, + "loss": 0.0568, + "step": 245 + }, + { + "epoch": 0.1579961464354528, + "grad_norm": 1.1501471627868498, + "learning_rate": 5.256410256410257e-06, + "loss": 0.0479, + "step": 246 + }, + { + "epoch": 0.15863840719332048, + "grad_norm": 2.3212031996645424, + "learning_rate": 5.2777777777777785e-06, + "loss": 0.2114, + "step": 247 + }, + { + "epoch": 0.15928066795118817, + "grad_norm": 1.8654773746777564, + "learning_rate": 5.2991452991453e-06, + "loss": 0.1907, + "step": 248 + }, + { + "epoch": 0.1599229287090559, + "grad_norm": 1.8298794059859385, + "learning_rate": 5.320512820512821e-06, + "loss": 0.1744, + "step": 249 + }, + { + "epoch": 0.16056518946692358, + "grad_norm": 1.4329150203491794, + "learning_rate": 5.341880341880342e-06, + "loss": 0.1077, + "step": 250 + }, + { + "epoch": 0.16120745022479127, + "grad_norm": 1.7550504564902967, + "learning_rate": 5.363247863247863e-06, + "loss": 0.1526, + "step": 251 + }, + { + "epoch": 0.16184971098265896, + "grad_norm": 2.2654092107945814, + "learning_rate": 5.384615384615385e-06, + "loss": 0.2076, + "step": 252 + }, + { + "epoch": 0.16249197174052665, + "grad_norm": 1.5062468368375765, + "learning_rate": 5.405982905982906e-06, + "loss": 0.1384, + "step": 253 + }, + { + "epoch": 0.16313423249839434, + "grad_norm": 1.8983251211471166, + "learning_rate": 5.4273504273504275e-06, + "loss": 0.1657, + "step": 254 + }, + { + "epoch": 0.16377649325626203, + "grad_norm": 2.1095319250262663, + "learning_rate": 5.448717948717949e-06, + "loss": 0.1869, + "step": 255 + }, + { + "epoch": 0.16441875401412973, + "grad_norm": 2.1928072937453487, + "learning_rate": 5.470085470085471e-06, + "loss": 0.3637, + "step": 256 + }, + { + "epoch": 0.16506101477199744, + "grad_norm": 12.777414485836376, + "learning_rate": 5.4914529914529925e-06, + "loss": 0.1224, + "step": 257 + }, + { + "epoch": 0.16570327552986513, + "grad_norm": 2.0604028373086916, + "learning_rate": 5.512820512820514e-06, + "loss": 0.1283, + "step": 258 + }, + { + "epoch": 0.16634553628773283, + "grad_norm": 1.7679641034520164, + "learning_rate": 5.534188034188035e-06, + "loss": 0.2142, + "step": 259 + }, + { + "epoch": 0.16698779704560052, + "grad_norm": 1.6470959982656463, + "learning_rate": 5.555555555555557e-06, + "loss": 0.156, + "step": 260 + }, + { + "epoch": 0.1676300578034682, + "grad_norm": 1.7469781122532952, + "learning_rate": 5.576923076923077e-06, + "loss": 0.1922, + "step": 261 + }, + { + "epoch": 0.1682723185613359, + "grad_norm": 1.3802872675846534, + "learning_rate": 5.598290598290599e-06, + "loss": 0.1381, + "step": 262 + }, + { + "epoch": 0.1689145793192036, + "grad_norm": 1.6191011184642785, + "learning_rate": 5.61965811965812e-06, + "loss": 0.1542, + "step": 263 + }, + { + "epoch": 0.16955684007707128, + "grad_norm": 1.4701793072335358, + "learning_rate": 5.641025641025641e-06, + "loss": 0.1389, + "step": 264 + }, + { + "epoch": 0.170199100834939, + "grad_norm": 1.4853434044171392, + "learning_rate": 5.662393162393163e-06, + "loss": 0.1393, + "step": 265 + }, + { + "epoch": 0.1708413615928067, + "grad_norm": 1.440762727078009, + "learning_rate": 5.683760683760684e-06, + "loss": 0.134, + "step": 266 + }, + { + "epoch": 0.17148362235067438, + "grad_norm": 1.7534768307087218, + "learning_rate": 5.705128205128206e-06, + "loss": 0.1674, + "step": 267 + }, + { + "epoch": 0.17212588310854207, + "grad_norm": 1.1786964747886273, + "learning_rate": 5.726495726495727e-06, + "loss": 0.0843, + "step": 268 + }, + { + "epoch": 0.17276814386640976, + "grad_norm": 1.5611714251181592, + "learning_rate": 5.7478632478632475e-06, + "loss": 0.1932, + "step": 269 + }, + { + "epoch": 0.17341040462427745, + "grad_norm": 1.5545669739695835, + "learning_rate": 5.769230769230769e-06, + "loss": 0.1543, + "step": 270 + }, + { + "epoch": 0.17405266538214514, + "grad_norm": 1.65052680717033, + "learning_rate": 5.790598290598292e-06, + "loss": 0.1865, + "step": 271 + }, + { + "epoch": 0.17469492614001284, + "grad_norm": 1.5803966562460312, + "learning_rate": 5.8119658119658126e-06, + "loss": 0.1381, + "step": 272 + }, + { + "epoch": 0.17533718689788053, + "grad_norm": 1.1416086953110978, + "learning_rate": 5.833333333333334e-06, + "loss": 0.0631, + "step": 273 + }, + { + "epoch": 0.17597944765574824, + "grad_norm": 2.3910523761902, + "learning_rate": 5.854700854700855e-06, + "loss": 0.1875, + "step": 274 + }, + { + "epoch": 0.17662170841361594, + "grad_norm": 1.769467808598136, + "learning_rate": 5.876068376068377e-06, + "loss": 0.1697, + "step": 275 + }, + { + "epoch": 0.17726396917148363, + "grad_norm": 1.7958403904746811, + "learning_rate": 5.897435897435898e-06, + "loss": 0.1848, + "step": 276 + }, + { + "epoch": 0.17790622992935132, + "grad_norm": 1.5951385414551633, + "learning_rate": 5.9188034188034195e-06, + "loss": 0.1492, + "step": 277 + }, + { + "epoch": 0.178548490687219, + "grad_norm": 1.779018840256742, + "learning_rate": 5.940170940170941e-06, + "loss": 0.1941, + "step": 278 + }, + { + "epoch": 0.1791907514450867, + "grad_norm": 1.8883978259080483, + "learning_rate": 5.961538461538462e-06, + "loss": 0.2482, + "step": 279 + }, + { + "epoch": 0.1798330122029544, + "grad_norm": 2.042190537856212, + "learning_rate": 5.982905982905983e-06, + "loss": 0.2333, + "step": 280 + }, + { + "epoch": 0.18047527296082208, + "grad_norm": 1.4104208151599262, + "learning_rate": 6.004273504273504e-06, + "loss": 0.0826, + "step": 281 + }, + { + "epoch": 0.1811175337186898, + "grad_norm": 1.8234187834288005, + "learning_rate": 6.025641025641026e-06, + "loss": 0.2395, + "step": 282 + }, + { + "epoch": 0.1817597944765575, + "grad_norm": 1.8421636499908145, + "learning_rate": 6.047008547008547e-06, + "loss": 0.2081, + "step": 283 + }, + { + "epoch": 0.18240205523442518, + "grad_norm": 1.6480163683437627, + "learning_rate": 6.0683760683760684e-06, + "loss": 0.1783, + "step": 284 + }, + { + "epoch": 0.18304431599229287, + "grad_norm": 1.7078551098988308, + "learning_rate": 6.08974358974359e-06, + "loss": 0.2132, + "step": 285 + }, + { + "epoch": 0.18368657675016056, + "grad_norm": 1.7644810949649952, + "learning_rate": 6.111111111111112e-06, + "loss": 0.1499, + "step": 286 + }, + { + "epoch": 0.18432883750802825, + "grad_norm": 1.4179911484567107, + "learning_rate": 6.1324786324786335e-06, + "loss": 0.1108, + "step": 287 + }, + { + "epoch": 0.18497109826589594, + "grad_norm": 1.5016450663560537, + "learning_rate": 6.153846153846155e-06, + "loss": 0.1836, + "step": 288 + }, + { + "epoch": 0.18561335902376364, + "grad_norm": 1.274277318503297, + "learning_rate": 6.175213675213676e-06, + "loss": 0.1148, + "step": 289 + }, + { + "epoch": 0.18625561978163135, + "grad_norm": 1.158452561355008, + "learning_rate": 6.196581196581198e-06, + "loss": 0.0826, + "step": 290 + }, + { + "epoch": 0.18689788053949905, + "grad_norm": 2.0890002209678937, + "learning_rate": 6.217948717948718e-06, + "loss": 0.1638, + "step": 291 + }, + { + "epoch": 0.18754014129736674, + "grad_norm": 1.3997815238988176, + "learning_rate": 6.23931623931624e-06, + "loss": 0.1036, + "step": 292 + }, + { + "epoch": 0.18818240205523443, + "grad_norm": 1.772204656493012, + "learning_rate": 6.260683760683761e-06, + "loss": 0.166, + "step": 293 + }, + { + "epoch": 0.18882466281310212, + "grad_norm": 1.444213686059476, + "learning_rate": 6.282051282051282e-06, + "loss": 0.1272, + "step": 294 + }, + { + "epoch": 0.1894669235709698, + "grad_norm": 1.607361072514269, + "learning_rate": 6.303418803418804e-06, + "loss": 0.1675, + "step": 295 + }, + { + "epoch": 0.1901091843288375, + "grad_norm": 1.8930501117698346, + "learning_rate": 6.324786324786325e-06, + "loss": 0.1661, + "step": 296 + }, + { + "epoch": 0.1907514450867052, + "grad_norm": 1.7261234067226432, + "learning_rate": 6.3461538461538466e-06, + "loss": 0.2234, + "step": 297 + }, + { + "epoch": 0.1913937058445729, + "grad_norm": 1.4498899035774084, + "learning_rate": 6.367521367521368e-06, + "loss": 0.1263, + "step": 298 + }, + { + "epoch": 0.1920359666024406, + "grad_norm": 1.3173783235873033, + "learning_rate": 6.3888888888888885e-06, + "loss": 0.1154, + "step": 299 + }, + { + "epoch": 0.1926782273603083, + "grad_norm": 1.269200025514882, + "learning_rate": 6.410256410256412e-06, + "loss": 0.1038, + "step": 300 + }, + { + "epoch": 0.19332048811817598, + "grad_norm": 1.9055564730921966, + "learning_rate": 6.431623931623933e-06, + "loss": 0.2875, + "step": 301 + }, + { + "epoch": 0.19396274887604367, + "grad_norm": 1.6399315955799212, + "learning_rate": 6.4529914529914535e-06, + "loss": 0.1802, + "step": 302 + }, + { + "epoch": 0.19460500963391136, + "grad_norm": 1.320280215499846, + "learning_rate": 6.474358974358975e-06, + "loss": 0.0819, + "step": 303 + }, + { + "epoch": 0.19524727039177905, + "grad_norm": 1.2523648917028196, + "learning_rate": 6.495726495726496e-06, + "loss": 0.1001, + "step": 304 + }, + { + "epoch": 0.19588953114964675, + "grad_norm": 1.182319240612418, + "learning_rate": 6.517094017094018e-06, + "loss": 0.0758, + "step": 305 + }, + { + "epoch": 0.19653179190751446, + "grad_norm": 1.6653672583216186, + "learning_rate": 6.538461538461539e-06, + "loss": 0.1937, + "step": 306 + }, + { + "epoch": 0.19717405266538215, + "grad_norm": 1.4068003967390916, + "learning_rate": 6.5598290598290605e-06, + "loss": 0.1322, + "step": 307 + }, + { + "epoch": 0.19781631342324985, + "grad_norm": 2.6240912255141997, + "learning_rate": 6.581196581196582e-06, + "loss": 0.2059, + "step": 308 + }, + { + "epoch": 0.19845857418111754, + "grad_norm": 1.4616747710078037, + "learning_rate": 6.602564102564103e-06, + "loss": 0.1283, + "step": 309 + }, + { + "epoch": 0.19910083493898523, + "grad_norm": 1.6655852879561668, + "learning_rate": 6.623931623931624e-06, + "loss": 0.1626, + "step": 310 + }, + { + "epoch": 0.19974309569685292, + "grad_norm": 1.3719658918865063, + "learning_rate": 6.645299145299145e-06, + "loss": 0.0651, + "step": 311 + }, + { + "epoch": 0.2003853564547206, + "grad_norm": 1.4198611438155977, + "learning_rate": 6.666666666666667e-06, + "loss": 0.1135, + "step": 312 + }, + { + "epoch": 0.2010276172125883, + "grad_norm": 2.520573388358363, + "learning_rate": 6.688034188034188e-06, + "loss": 0.0883, + "step": 313 + }, + { + "epoch": 0.20166987797045602, + "grad_norm": 1.0909216946385627, + "learning_rate": 6.7094017094017094e-06, + "loss": 0.0765, + "step": 314 + }, + { + "epoch": 0.2023121387283237, + "grad_norm": 1.6058592705998596, + "learning_rate": 6.730769230769232e-06, + "loss": 0.1364, + "step": 315 + }, + { + "epoch": 0.2029543994861914, + "grad_norm": 2.3462474171898293, + "learning_rate": 6.752136752136753e-06, + "loss": 0.1289, + "step": 316 + }, + { + "epoch": 0.2035966602440591, + "grad_norm": 0.8655341661203175, + "learning_rate": 6.7735042735042745e-06, + "loss": 0.0502, + "step": 317 + }, + { + "epoch": 0.20423892100192678, + "grad_norm": 1.341685277356201, + "learning_rate": 6.794871794871796e-06, + "loss": 0.1068, + "step": 318 + }, + { + "epoch": 0.20488118175979447, + "grad_norm": 1.6319171099706442, + "learning_rate": 6.816239316239317e-06, + "loss": 0.2271, + "step": 319 + }, + { + "epoch": 0.20552344251766216, + "grad_norm": 1.4949032409394551, + "learning_rate": 6.837606837606839e-06, + "loss": 0.1877, + "step": 320 + }, + { + "epoch": 0.20616570327552985, + "grad_norm": 1.4404884764676187, + "learning_rate": 6.858974358974359e-06, + "loss": 0.133, + "step": 321 + }, + { + "epoch": 0.20680796403339757, + "grad_norm": 1.411963894885179, + "learning_rate": 6.880341880341881e-06, + "loss": 0.1455, + "step": 322 + }, + { + "epoch": 0.20745022479126526, + "grad_norm": 1.6035411349439346, + "learning_rate": 6.901709401709402e-06, + "loss": 0.2057, + "step": 323 + }, + { + "epoch": 0.20809248554913296, + "grad_norm": 1.5411810622997264, + "learning_rate": 6.923076923076923e-06, + "loss": 0.2004, + "step": 324 + }, + { + "epoch": 0.20873474630700065, + "grad_norm": 1.4956952741171343, + "learning_rate": 6.944444444444445e-06, + "loss": 0.1541, + "step": 325 + }, + { + "epoch": 0.20937700706486834, + "grad_norm": 1.7107743636181876, + "learning_rate": 6.965811965811966e-06, + "loss": 0.214, + "step": 326 + }, + { + "epoch": 0.21001926782273603, + "grad_norm": 1.3118870197413472, + "learning_rate": 6.9871794871794876e-06, + "loss": 0.1125, + "step": 327 + }, + { + "epoch": 0.21066152858060372, + "grad_norm": 2.674730049788286, + "learning_rate": 7.008547008547009e-06, + "loss": 0.0677, + "step": 328 + }, + { + "epoch": 0.2113037893384714, + "grad_norm": 1.3691109504736463, + "learning_rate": 7.02991452991453e-06, + "loss": 0.1496, + "step": 329 + }, + { + "epoch": 0.2119460500963391, + "grad_norm": 1.8608001386662654, + "learning_rate": 7.051282051282053e-06, + "loss": 0.2869, + "step": 330 + }, + { + "epoch": 0.21258831085420682, + "grad_norm": 1.6316638505532686, + "learning_rate": 7.072649572649574e-06, + "loss": 0.1617, + "step": 331 + }, + { + "epoch": 0.2132305716120745, + "grad_norm": 2.336940901389275, + "learning_rate": 7.0940170940170945e-06, + "loss": 0.1381, + "step": 332 + }, + { + "epoch": 0.2138728323699422, + "grad_norm": 1.8223630798607844, + "learning_rate": 7.115384615384616e-06, + "loss": 0.2815, + "step": 333 + }, + { + "epoch": 0.2145150931278099, + "grad_norm": 1.824237192762063, + "learning_rate": 7.136752136752137e-06, + "loss": 0.1667, + "step": 334 + }, + { + "epoch": 0.21515735388567758, + "grad_norm": 1.6594371034745137, + "learning_rate": 7.158119658119659e-06, + "loss": 0.1943, + "step": 335 + }, + { + "epoch": 0.21579961464354527, + "grad_norm": 1.450243509539777, + "learning_rate": 7.17948717948718e-06, + "loss": 0.1298, + "step": 336 + }, + { + "epoch": 0.21644187540141296, + "grad_norm": 1.2869690037510484, + "learning_rate": 7.2008547008547015e-06, + "loss": 0.0939, + "step": 337 + }, + { + "epoch": 0.21708413615928066, + "grad_norm": 1.7748751146501742, + "learning_rate": 7.222222222222223e-06, + "loss": 0.1476, + "step": 338 + }, + { + "epoch": 0.21772639691714837, + "grad_norm": 1.9251573915723108, + "learning_rate": 7.243589743589744e-06, + "loss": 0.1017, + "step": 339 + }, + { + "epoch": 0.21836865767501606, + "grad_norm": 1.7096716187676682, + "learning_rate": 7.264957264957266e-06, + "loss": 0.2364, + "step": 340 + }, + { + "epoch": 0.21901091843288376, + "grad_norm": 4.93694886204962, + "learning_rate": 7.286324786324786e-06, + "loss": 0.219, + "step": 341 + }, + { + "epoch": 0.21965317919075145, + "grad_norm": 1.3941018421946567, + "learning_rate": 7.307692307692308e-06, + "loss": 0.1275, + "step": 342 + }, + { + "epoch": 0.22029543994861914, + "grad_norm": 2.1275128813873883, + "learning_rate": 7.329059829059829e-06, + "loss": 0.1541, + "step": 343 + }, + { + "epoch": 0.22093770070648683, + "grad_norm": 1.8164039677564348, + "learning_rate": 7.350427350427351e-06, + "loss": 0.1451, + "step": 344 + }, + { + "epoch": 0.22157996146435452, + "grad_norm": 1.3182499320288614, + "learning_rate": 7.371794871794873e-06, + "loss": 0.092, + "step": 345 + }, + { + "epoch": 0.2222222222222222, + "grad_norm": 1.968586468610248, + "learning_rate": 7.393162393162394e-06, + "loss": 0.2054, + "step": 346 + }, + { + "epoch": 0.22286448298008993, + "grad_norm": 1.6554949714580813, + "learning_rate": 7.4145299145299155e-06, + "loss": 0.2381, + "step": 347 + }, + { + "epoch": 0.22350674373795762, + "grad_norm": 1.7264156308856657, + "learning_rate": 7.435897435897437e-06, + "loss": 0.1323, + "step": 348 + }, + { + "epoch": 0.2241490044958253, + "grad_norm": 1.2435707590981013, + "learning_rate": 7.457264957264958e-06, + "loss": 0.0983, + "step": 349 + }, + { + "epoch": 0.224791265253693, + "grad_norm": 1.606141249950526, + "learning_rate": 7.47863247863248e-06, + "loss": 0.1154, + "step": 350 + }, + { + "epoch": 0.2254335260115607, + "grad_norm": 1.4865051275780938, + "learning_rate": 7.500000000000001e-06, + "loss": 0.158, + "step": 351 + }, + { + "epoch": 0.22607578676942838, + "grad_norm": 1.350037041024085, + "learning_rate": 7.521367521367522e-06, + "loss": 0.1447, + "step": 352 + }, + { + "epoch": 0.22671804752729607, + "grad_norm": 1.732733981376795, + "learning_rate": 7.542735042735043e-06, + "loss": 0.224, + "step": 353 + }, + { + "epoch": 0.22736030828516376, + "grad_norm": 1.7184568858883003, + "learning_rate": 7.564102564102564e-06, + "loss": 0.2473, + "step": 354 + }, + { + "epoch": 0.22800256904303148, + "grad_norm": 6.4986011817980485, + "learning_rate": 7.585470085470086e-06, + "loss": 0.0692, + "step": 355 + }, + { + "epoch": 0.22864482980089917, + "grad_norm": 1.6912526494372149, + "learning_rate": 7.606837606837607e-06, + "loss": 0.228, + "step": 356 + }, + { + "epoch": 0.22928709055876687, + "grad_norm": 1.437502547019552, + "learning_rate": 7.6282051282051286e-06, + "loss": 0.1185, + "step": 357 + }, + { + "epoch": 0.22992935131663456, + "grad_norm": 1.4963294585891511, + "learning_rate": 7.649572649572649e-06, + "loss": 0.166, + "step": 358 + }, + { + "epoch": 0.23057161207450225, + "grad_norm": 1.602314601329608, + "learning_rate": 7.670940170940172e-06, + "loss": 0.1277, + "step": 359 + }, + { + "epoch": 0.23121387283236994, + "grad_norm": 1.862462340403427, + "learning_rate": 7.692307692307694e-06, + "loss": 0.2839, + "step": 360 + }, + { + "epoch": 0.23185613359023763, + "grad_norm": 1.4144391532349165, + "learning_rate": 7.713675213675215e-06, + "loss": 0.1437, + "step": 361 + }, + { + "epoch": 0.23249839434810532, + "grad_norm": 1.3939944093534622, + "learning_rate": 7.735042735042736e-06, + "loss": 0.0847, + "step": 362 + }, + { + "epoch": 0.23314065510597304, + "grad_norm": 1.3678832447397904, + "learning_rate": 7.756410256410258e-06, + "loss": 0.1198, + "step": 363 + }, + { + "epoch": 0.23378291586384073, + "grad_norm": 1.6359126556994885, + "learning_rate": 7.77777777777778e-06, + "loss": 0.1982, + "step": 364 + }, + { + "epoch": 0.23442517662170842, + "grad_norm": 1.6569013496157579, + "learning_rate": 7.7991452991453e-06, + "loss": 0.1689, + "step": 365 + }, + { + "epoch": 0.2350674373795761, + "grad_norm": 1.4000277206868814, + "learning_rate": 7.820512820512822e-06, + "loss": 0.1519, + "step": 366 + }, + { + "epoch": 0.2357096981374438, + "grad_norm": 1.6377542747794123, + "learning_rate": 7.841880341880342e-06, + "loss": 0.2413, + "step": 367 + }, + { + "epoch": 0.2363519588953115, + "grad_norm": 1.4672147063036414, + "learning_rate": 7.863247863247863e-06, + "loss": 0.1058, + "step": 368 + }, + { + "epoch": 0.23699421965317918, + "grad_norm": 1.7274082607375925, + "learning_rate": 7.884615384615384e-06, + "loss": 0.2485, + "step": 369 + }, + { + "epoch": 0.23763648041104687, + "grad_norm": 1.479011511276184, + "learning_rate": 7.905982905982906e-06, + "loss": 0.197, + "step": 370 + }, + { + "epoch": 0.2382787411689146, + "grad_norm": 1.4946202486101767, + "learning_rate": 7.927350427350427e-06, + "loss": 0.1502, + "step": 371 + }, + { + "epoch": 0.23892100192678228, + "grad_norm": 1.2650020424197992, + "learning_rate": 7.948717948717949e-06, + "loss": 0.0951, + "step": 372 + }, + { + "epoch": 0.23956326268464997, + "grad_norm": 1.764864067008259, + "learning_rate": 7.970085470085472e-06, + "loss": 0.2278, + "step": 373 + }, + { + "epoch": 0.24020552344251767, + "grad_norm": 1.5378629428189603, + "learning_rate": 7.991452991452993e-06, + "loss": 0.2117, + "step": 374 + }, + { + "epoch": 0.24084778420038536, + "grad_norm": 1.4133744493337237, + "learning_rate": 8.012820512820515e-06, + "loss": 0.1084, + "step": 375 + }, + { + "epoch": 0.24149004495825305, + "grad_norm": 1.5519757846048787, + "learning_rate": 8.034188034188036e-06, + "loss": 0.1265, + "step": 376 + }, + { + "epoch": 0.24213230571612074, + "grad_norm": 1.2378055061815647, + "learning_rate": 8.055555555555557e-06, + "loss": 0.0949, + "step": 377 + }, + { + "epoch": 0.24277456647398843, + "grad_norm": 1.099868001885845, + "learning_rate": 8.076923076923077e-06, + "loss": 0.0855, + "step": 378 + }, + { + "epoch": 0.24341682723185612, + "grad_norm": 1.1206557639309074, + "learning_rate": 8.098290598290598e-06, + "loss": 0.088, + "step": 379 + }, + { + "epoch": 0.24405908798972384, + "grad_norm": 1.931399131274482, + "learning_rate": 8.11965811965812e-06, + "loss": 0.252, + "step": 380 + }, + { + "epoch": 0.24470134874759153, + "grad_norm": 1.4012795754715972, + "learning_rate": 8.141025641025641e-06, + "loss": 0.068, + "step": 381 + }, + { + "epoch": 0.24534360950545922, + "grad_norm": 1.1220573371085383, + "learning_rate": 8.162393162393163e-06, + "loss": 0.069, + "step": 382 + }, + { + "epoch": 0.2459858702633269, + "grad_norm": 1.7292756456940992, + "learning_rate": 8.183760683760684e-06, + "loss": 0.2011, + "step": 383 + }, + { + "epoch": 0.2466281310211946, + "grad_norm": 1.6948568748411632, + "learning_rate": 8.205128205128205e-06, + "loss": 0.2686, + "step": 384 + }, + { + "epoch": 0.2472703917790623, + "grad_norm": 1.6716572175943611, + "learning_rate": 8.226495726495727e-06, + "loss": 0.1312, + "step": 385 + }, + { + "epoch": 0.24791265253692998, + "grad_norm": 1.5604701702023958, + "learning_rate": 8.247863247863248e-06, + "loss": 0.192, + "step": 386 + }, + { + "epoch": 0.24855491329479767, + "grad_norm": 1.457568100977, + "learning_rate": 8.26923076923077e-06, + "loss": 0.1892, + "step": 387 + }, + { + "epoch": 0.2491971740526654, + "grad_norm": 1.4245998142377962, + "learning_rate": 8.290598290598293e-06, + "loss": 0.1548, + "step": 388 + }, + { + "epoch": 0.24983943481053308, + "grad_norm": 1.3140061526908098, + "learning_rate": 8.311965811965812e-06, + "loss": 0.1257, + "step": 389 + }, + { + "epoch": 0.2504816955684008, + "grad_norm": 1.3765208084977496, + "learning_rate": 8.333333333333334e-06, + "loss": 0.13, + "step": 390 + }, + { + "epoch": 0.25112395632626844, + "grad_norm": 1.3783349922976273, + "learning_rate": 8.354700854700855e-06, + "loss": 0.1678, + "step": 391 + }, + { + "epoch": 0.25176621708413616, + "grad_norm": 1.074435758944751, + "learning_rate": 8.376068376068377e-06, + "loss": 0.095, + "step": 392 + }, + { + "epoch": 0.2524084778420039, + "grad_norm": 1.137624741326991, + "learning_rate": 8.397435897435898e-06, + "loss": 0.078, + "step": 393 + }, + { + "epoch": 0.25305073859987154, + "grad_norm": 1.5972292429541488, + "learning_rate": 8.41880341880342e-06, + "loss": 0.2041, + "step": 394 + }, + { + "epoch": 0.25369299935773926, + "grad_norm": 1.708820992323593, + "learning_rate": 8.44017094017094e-06, + "loss": 0.1972, + "step": 395 + }, + { + "epoch": 0.2543352601156069, + "grad_norm": 1.6452209621534704, + "learning_rate": 8.461538461538462e-06, + "loss": 0.1847, + "step": 396 + }, + { + "epoch": 0.25497752087347464, + "grad_norm": 1.441134018305627, + "learning_rate": 8.482905982905983e-06, + "loss": 0.1455, + "step": 397 + }, + { + "epoch": 0.2556197816313423, + "grad_norm": 1.3903304248103567, + "learning_rate": 8.504273504273505e-06, + "loss": 0.127, + "step": 398 + }, + { + "epoch": 0.25626204238921, + "grad_norm": 1.0747613386140034, + "learning_rate": 8.525641025641026e-06, + "loss": 0.0659, + "step": 399 + }, + { + "epoch": 0.25690430314707774, + "grad_norm": 1.5812469612273417, + "learning_rate": 8.547008547008548e-06, + "loss": 0.1494, + "step": 400 + }, + { + "epoch": 0.2575465639049454, + "grad_norm": 1.1677638195746958, + "learning_rate": 8.568376068376069e-06, + "loss": 0.126, + "step": 401 + }, + { + "epoch": 0.2581888246628131, + "grad_norm": 1.8425411857865168, + "learning_rate": 8.58974358974359e-06, + "loss": 0.0574, + "step": 402 + }, + { + "epoch": 0.2588310854206808, + "grad_norm": 1.4718472971767633, + "learning_rate": 8.611111111111112e-06, + "loss": 0.2152, + "step": 403 + }, + { + "epoch": 0.2594733461785485, + "grad_norm": 1.3015174611492357, + "learning_rate": 8.632478632478633e-06, + "loss": 0.0778, + "step": 404 + }, + { + "epoch": 0.26011560693641617, + "grad_norm": 1.1129610260152418, + "learning_rate": 8.653846153846155e-06, + "loss": 0.0764, + "step": 405 + }, + { + "epoch": 0.2607578676942839, + "grad_norm": 1.181575633243009, + "learning_rate": 8.675213675213676e-06, + "loss": 0.1077, + "step": 406 + }, + { + "epoch": 0.26140012845215155, + "grad_norm": 1.168279725190337, + "learning_rate": 8.696581196581197e-06, + "loss": 0.1149, + "step": 407 + }, + { + "epoch": 0.26204238921001927, + "grad_norm": 1.5883146729718909, + "learning_rate": 8.717948717948719e-06, + "loss": 0.2193, + "step": 408 + }, + { + "epoch": 0.262684649967887, + "grad_norm": 1.4761780515829104, + "learning_rate": 8.73931623931624e-06, + "loss": 0.1964, + "step": 409 + }, + { + "epoch": 0.26332691072575465, + "grad_norm": 1.4312624720149225, + "learning_rate": 8.760683760683762e-06, + "loss": 0.1979, + "step": 410 + }, + { + "epoch": 0.26396917148362237, + "grad_norm": 1.1612375928374254, + "learning_rate": 8.782051282051283e-06, + "loss": 0.1223, + "step": 411 + }, + { + "epoch": 0.26461143224149003, + "grad_norm": 1.2504976877386023, + "learning_rate": 8.803418803418804e-06, + "loss": 0.0883, + "step": 412 + }, + { + "epoch": 0.26525369299935775, + "grad_norm": 1.3669964915280857, + "learning_rate": 8.824786324786326e-06, + "loss": 0.1658, + "step": 413 + }, + { + "epoch": 0.2658959537572254, + "grad_norm": 1.246626762743697, + "learning_rate": 8.846153846153847e-06, + "loss": 0.0903, + "step": 414 + }, + { + "epoch": 0.26653821451509313, + "grad_norm": 1.471586050484201, + "learning_rate": 8.867521367521369e-06, + "loss": 0.1876, + "step": 415 + }, + { + "epoch": 0.26718047527296085, + "grad_norm": 1.3059646699535103, + "learning_rate": 8.888888888888888e-06, + "loss": 0.148, + "step": 416 + }, + { + "epoch": 0.2678227360308285, + "grad_norm": 1.9199535786274742, + "learning_rate": 8.910256410256411e-06, + "loss": 0.18, + "step": 417 + }, + { + "epoch": 0.26846499678869623, + "grad_norm": 1.4503377613626816, + "learning_rate": 8.931623931623933e-06, + "loss": 0.1405, + "step": 418 + }, + { + "epoch": 0.2691072575465639, + "grad_norm": 1.341258451984921, + "learning_rate": 8.952991452991454e-06, + "loss": 0.0922, + "step": 419 + }, + { + "epoch": 0.2697495183044316, + "grad_norm": 1.496202658656103, + "learning_rate": 8.974358974358976e-06, + "loss": 0.1012, + "step": 420 + }, + { + "epoch": 0.2703917790622993, + "grad_norm": 1.7885525806947915, + "learning_rate": 8.995726495726497e-06, + "loss": 0.2544, + "step": 421 + }, + { + "epoch": 0.271034039820167, + "grad_norm": 1.611489777810217, + "learning_rate": 9.017094017094018e-06, + "loss": 0.152, + "step": 422 + }, + { + "epoch": 0.27167630057803466, + "grad_norm": 1.7377041174924726, + "learning_rate": 9.03846153846154e-06, + "loss": 0.1408, + "step": 423 + }, + { + "epoch": 0.2723185613359024, + "grad_norm": 1.4941202857365428, + "learning_rate": 9.059829059829061e-06, + "loss": 0.1757, + "step": 424 + }, + { + "epoch": 0.2729608220937701, + "grad_norm": 1.3577308212984789, + "learning_rate": 9.081196581196583e-06, + "loss": 0.1595, + "step": 425 + }, + { + "epoch": 0.27360308285163776, + "grad_norm": 1.2364349338735192, + "learning_rate": 9.102564102564104e-06, + "loss": 0.1216, + "step": 426 + }, + { + "epoch": 0.2742453436095055, + "grad_norm": 1.0864806677120173, + "learning_rate": 9.123931623931624e-06, + "loss": 0.0954, + "step": 427 + }, + { + "epoch": 0.27488760436737314, + "grad_norm": 1.424320443479475, + "learning_rate": 9.145299145299145e-06, + "loss": 0.1967, + "step": 428 + }, + { + "epoch": 0.27552986512524086, + "grad_norm": 1.7617157791516171, + "learning_rate": 9.166666666666666e-06, + "loss": 0.2637, + "step": 429 + }, + { + "epoch": 0.2761721258831085, + "grad_norm": 1.5983834891463615, + "learning_rate": 9.188034188034188e-06, + "loss": 0.1913, + "step": 430 + }, + { + "epoch": 0.27681438664097624, + "grad_norm": 1.3427639494263461, + "learning_rate": 9.20940170940171e-06, + "loss": 0.1356, + "step": 431 + }, + { + "epoch": 0.2774566473988439, + "grad_norm": 1.3006393064474155, + "learning_rate": 9.230769230769232e-06, + "loss": 0.1482, + "step": 432 + }, + { + "epoch": 0.2780989081567116, + "grad_norm": 0.9882839580171652, + "learning_rate": 9.252136752136754e-06, + "loss": 0.0477, + "step": 433 + }, + { + "epoch": 0.27874116891457934, + "grad_norm": 1.8187733059590476, + "learning_rate": 9.273504273504275e-06, + "loss": 0.2519, + "step": 434 + }, + { + "epoch": 0.279383429672447, + "grad_norm": 0.9719417895364293, + "learning_rate": 9.294871794871796e-06, + "loss": 0.0587, + "step": 435 + }, + { + "epoch": 0.2800256904303147, + "grad_norm": 0.9611551936768272, + "learning_rate": 9.316239316239318e-06, + "loss": 0.0703, + "step": 436 + }, + { + "epoch": 0.2806679511881824, + "grad_norm": 1.2005337560511329, + "learning_rate": 9.33760683760684e-06, + "loss": 0.1335, + "step": 437 + }, + { + "epoch": 0.2813102119460501, + "grad_norm": 1.445078163400631, + "learning_rate": 9.358974358974359e-06, + "loss": 0.0901, + "step": 438 + }, + { + "epoch": 0.28195247270391777, + "grad_norm": 1.2986045611189219, + "learning_rate": 9.38034188034188e-06, + "loss": 0.1363, + "step": 439 + }, + { + "epoch": 0.2825947334617855, + "grad_norm": 1.1780904099312453, + "learning_rate": 9.401709401709402e-06, + "loss": 0.1002, + "step": 440 + }, + { + "epoch": 0.2832369942196532, + "grad_norm": 1.501825130344284, + "learning_rate": 9.423076923076923e-06, + "loss": 0.239, + "step": 441 + }, + { + "epoch": 0.28387925497752087, + "grad_norm": 1.221974843214199, + "learning_rate": 9.444444444444445e-06, + "loss": 0.0878, + "step": 442 + }, + { + "epoch": 0.2845215157353886, + "grad_norm": 1.148163463377832, + "learning_rate": 9.465811965811966e-06, + "loss": 0.1141, + "step": 443 + }, + { + "epoch": 0.28516377649325625, + "grad_norm": 1.2539764140306715, + "learning_rate": 9.487179487179487e-06, + "loss": 0.1203, + "step": 444 + }, + { + "epoch": 0.28580603725112397, + "grad_norm": 1.0432465969684281, + "learning_rate": 9.508547008547009e-06, + "loss": 0.0678, + "step": 445 + }, + { + "epoch": 0.28644829800899163, + "grad_norm": 1.5151836409753747, + "learning_rate": 9.52991452991453e-06, + "loss": 0.2166, + "step": 446 + }, + { + "epoch": 0.28709055876685935, + "grad_norm": 1.2708015760019096, + "learning_rate": 9.551282051282053e-06, + "loss": 0.1797, + "step": 447 + }, + { + "epoch": 0.287732819524727, + "grad_norm": 1.3805512569355416, + "learning_rate": 9.572649572649575e-06, + "loss": 0.1555, + "step": 448 + }, + { + "epoch": 0.28837508028259473, + "grad_norm": 1.4017862937544694, + "learning_rate": 9.594017094017094e-06, + "loss": 0.1685, + "step": 449 + }, + { + "epoch": 0.28901734104046245, + "grad_norm": 1.0961149756637536, + "learning_rate": 9.615384615384616e-06, + "loss": 0.1213, + "step": 450 + }, + { + "epoch": 0.2896596017983301, + "grad_norm": 1.2581924122542334, + "learning_rate": 9.636752136752137e-06, + "loss": 0.1538, + "step": 451 + }, + { + "epoch": 0.29030186255619783, + "grad_norm": 1.4840630171670743, + "learning_rate": 9.658119658119659e-06, + "loss": 0.2085, + "step": 452 + }, + { + "epoch": 0.2909441233140655, + "grad_norm": 1.6339930805873184, + "learning_rate": 9.67948717948718e-06, + "loss": 0.2609, + "step": 453 + }, + { + "epoch": 0.2915863840719332, + "grad_norm": 1.6926241097340726, + "learning_rate": 9.700854700854701e-06, + "loss": 0.2112, + "step": 454 + }, + { + "epoch": 0.2922286448298009, + "grad_norm": 1.0785834064375057, + "learning_rate": 9.722222222222223e-06, + "loss": 0.0687, + "step": 455 + }, + { + "epoch": 0.2928709055876686, + "grad_norm": 1.47185548650305, + "learning_rate": 9.743589743589744e-06, + "loss": 0.1892, + "step": 456 + }, + { + "epoch": 0.2935131663455363, + "grad_norm": 1.125183030179888, + "learning_rate": 9.764957264957265e-06, + "loss": 0.0931, + "step": 457 + }, + { + "epoch": 0.294155427103404, + "grad_norm": 1.9252187971657972, + "learning_rate": 9.786324786324787e-06, + "loss": 0.0956, + "step": 458 + }, + { + "epoch": 0.2947976878612717, + "grad_norm": 1.0478027161706631, + "learning_rate": 9.807692307692308e-06, + "loss": 0.1176, + "step": 459 + }, + { + "epoch": 0.29543994861913936, + "grad_norm": 1.1090720049749356, + "learning_rate": 9.82905982905983e-06, + "loss": 0.0999, + "step": 460 + }, + { + "epoch": 0.2960822093770071, + "grad_norm": 1.337368159325035, + "learning_rate": 9.850427350427351e-06, + "loss": 0.2018, + "step": 461 + }, + { + "epoch": 0.29672447013487474, + "grad_norm": 1.4811737430990377, + "learning_rate": 9.871794871794872e-06, + "loss": 0.1312, + "step": 462 + }, + { + "epoch": 0.29736673089274246, + "grad_norm": 1.7469628988344323, + "learning_rate": 9.893162393162394e-06, + "loss": 0.1787, + "step": 463 + }, + { + "epoch": 0.2980089916506101, + "grad_norm": 1.3806845330962965, + "learning_rate": 9.914529914529915e-06, + "loss": 0.1677, + "step": 464 + }, + { + "epoch": 0.29865125240847784, + "grad_norm": 2.6083786801920814, + "learning_rate": 9.935897435897437e-06, + "loss": 0.2024, + "step": 465 + }, + { + "epoch": 0.29929351316634556, + "grad_norm": 2.193669273334277, + "learning_rate": 9.957264957264958e-06, + "loss": 0.1777, + "step": 466 + }, + { + "epoch": 0.2999357739242132, + "grad_norm": 2.798173199809623, + "learning_rate": 9.97863247863248e-06, + "loss": 0.1414, + "step": 467 + }, + { + "epoch": 0.30057803468208094, + "grad_norm": 1.4766123394900963, + "learning_rate": 1e-05, + "loss": 0.1822, + "step": 468 + }, + { + "epoch": 0.3012202954399486, + "grad_norm": 1.401679781282655, + "learning_rate": 9.99999860324269e-06, + "loss": 0.1951, + "step": 469 + }, + { + "epoch": 0.3018625561978163, + "grad_norm": 1.4230573969133684, + "learning_rate": 9.99999441297153e-06, + "loss": 0.1862, + "step": 470 + }, + { + "epoch": 0.302504816955684, + "grad_norm": 1.2268884855613509, + "learning_rate": 9.999987429188867e-06, + "loss": 0.1682, + "step": 471 + }, + { + "epoch": 0.3031470777135517, + "grad_norm": 1.529970228280339, + "learning_rate": 9.999977651898602e-06, + "loss": 0.212, + "step": 472 + }, + { + "epoch": 0.3037893384714194, + "grad_norm": 1.3615182611023045, + "learning_rate": 9.999965081106198e-06, + "loss": 0.1671, + "step": 473 + }, + { + "epoch": 0.3044315992292871, + "grad_norm": 1.7511658649948065, + "learning_rate": 9.999949716818677e-06, + "loss": 0.1919, + "step": 474 + }, + { + "epoch": 0.3050738599871548, + "grad_norm": 1.3365835479559633, + "learning_rate": 9.999931559044624e-06, + "loss": 0.1387, + "step": 475 + }, + { + "epoch": 0.30571612074502247, + "grad_norm": 1.1576187648941112, + "learning_rate": 9.999910607794183e-06, + "loss": 0.1013, + "step": 476 + }, + { + "epoch": 0.3063583815028902, + "grad_norm": 1.636853089018074, + "learning_rate": 9.99988686307906e-06, + "loss": 0.2124, + "step": 477 + }, + { + "epoch": 0.30700064226075785, + "grad_norm": 1.0900175438611053, + "learning_rate": 9.99986032491252e-06, + "loss": 0.0837, + "step": 478 + }, + { + "epoch": 0.30764290301862557, + "grad_norm": 1.3956496713792836, + "learning_rate": 9.999830993309394e-06, + "loss": 0.1795, + "step": 479 + }, + { + "epoch": 0.30828516377649323, + "grad_norm": 1.2783016042576147, + "learning_rate": 9.999798868286064e-06, + "loss": 0.1129, + "step": 480 + }, + { + "epoch": 0.30892742453436095, + "grad_norm": 1.2922579313890805, + "learning_rate": 9.999763949860483e-06, + "loss": 0.0709, + "step": 481 + }, + { + "epoch": 0.30956968529222867, + "grad_norm": 1.2716726750327352, + "learning_rate": 9.999726238052158e-06, + "loss": 0.1292, + "step": 482 + }, + { + "epoch": 0.31021194605009633, + "grad_norm": 1.5425711562279572, + "learning_rate": 9.999685732882158e-06, + "loss": 0.1658, + "step": 483 + }, + { + "epoch": 0.31085420680796405, + "grad_norm": 1.2845394580692517, + "learning_rate": 9.999642434373115e-06, + "loss": 0.1115, + "step": 484 + }, + { + "epoch": 0.3114964675658317, + "grad_norm": 1.2513061537141321, + "learning_rate": 9.999596342549218e-06, + "loss": 0.1377, + "step": 485 + }, + { + "epoch": 0.31213872832369943, + "grad_norm": 1.6123966384888107, + "learning_rate": 9.999547457436221e-06, + "loss": 0.1292, + "step": 486 + }, + { + "epoch": 0.3127809890815671, + "grad_norm": 1.2788726958721703, + "learning_rate": 9.999495779061436e-06, + "loss": 0.1164, + "step": 487 + }, + { + "epoch": 0.3134232498394348, + "grad_norm": 1.158111561653726, + "learning_rate": 9.999441307453734e-06, + "loss": 0.085, + "step": 488 + }, + { + "epoch": 0.3140655105973025, + "grad_norm": 1.6895618556999503, + "learning_rate": 9.99938404264355e-06, + "loss": 0.161, + "step": 489 + }, + { + "epoch": 0.3147077713551702, + "grad_norm": 1.1444790096569362, + "learning_rate": 9.999323984662876e-06, + "loss": 0.1307, + "step": 490 + }, + { + "epoch": 0.3153500321130379, + "grad_norm": 0.9364744968670801, + "learning_rate": 9.999261133545269e-06, + "loss": 0.1012, + "step": 491 + }, + { + "epoch": 0.3159922928709056, + "grad_norm": 1.1354282538479976, + "learning_rate": 9.999195489325842e-06, + "loss": 0.1168, + "step": 492 + }, + { + "epoch": 0.3166345536287733, + "grad_norm": 1.4201947885058752, + "learning_rate": 9.999127052041273e-06, + "loss": 0.1757, + "step": 493 + }, + { + "epoch": 0.31727681438664096, + "grad_norm": 0.8816678580376328, + "learning_rate": 9.999055821729797e-06, + "loss": 0.0682, + "step": 494 + }, + { + "epoch": 0.3179190751445087, + "grad_norm": 1.3575909065547067, + "learning_rate": 9.99898179843121e-06, + "loss": 0.1602, + "step": 495 + }, + { + "epoch": 0.31856133590237634, + "grad_norm": 1.3605449999357655, + "learning_rate": 9.99890498218687e-06, + "loss": 0.148, + "step": 496 + }, + { + "epoch": 0.31920359666024406, + "grad_norm": 1.3515922978634687, + "learning_rate": 9.998825373039693e-06, + "loss": 0.2073, + "step": 497 + }, + { + "epoch": 0.3198458574181118, + "grad_norm": 1.359661295572572, + "learning_rate": 9.998742971034159e-06, + "loss": 0.1993, + "step": 498 + }, + { + "epoch": 0.32048811817597944, + "grad_norm": 1.1326034391980486, + "learning_rate": 9.998657776216304e-06, + "loss": 0.1389, + "step": 499 + }, + { + "epoch": 0.32113037893384716, + "grad_norm": 1.1261644489908365, + "learning_rate": 9.998569788633727e-06, + "loss": 0.1189, + "step": 500 + }, + { + "epoch": 0.3217726396917148, + "grad_norm": 1.3835648665700406, + "learning_rate": 9.998479008335589e-06, + "loss": 0.2095, + "step": 501 + }, + { + "epoch": 0.32241490044958254, + "grad_norm": 1.433533897810241, + "learning_rate": 9.998385435372607e-06, + "loss": 0.2582, + "step": 502 + }, + { + "epoch": 0.3230571612074502, + "grad_norm": 1.9025972970972944, + "learning_rate": 9.998289069797062e-06, + "loss": 0.1461, + "step": 503 + }, + { + "epoch": 0.3236994219653179, + "grad_norm": 1.2412422539346373, + "learning_rate": 9.998189911662793e-06, + "loss": 0.1426, + "step": 504 + }, + { + "epoch": 0.3243416827231856, + "grad_norm": 1.6305437702298338, + "learning_rate": 9.998087961025199e-06, + "loss": 0.2954, + "step": 505 + }, + { + "epoch": 0.3249839434810533, + "grad_norm": 1.2870669824322503, + "learning_rate": 9.997983217941242e-06, + "loss": 0.1751, + "step": 506 + }, + { + "epoch": 0.325626204238921, + "grad_norm": 1.5613616033879356, + "learning_rate": 9.99787568246944e-06, + "loss": 0.2866, + "step": 507 + }, + { + "epoch": 0.3262684649967887, + "grad_norm": 1.2795705913690616, + "learning_rate": 9.997765354669877e-06, + "loss": 0.1541, + "step": 508 + }, + { + "epoch": 0.3269107257546564, + "grad_norm": 1.1755117160196136, + "learning_rate": 9.997652234604189e-06, + "loss": 0.1006, + "step": 509 + }, + { + "epoch": 0.32755298651252407, + "grad_norm": 1.3599214675742322, + "learning_rate": 9.99753632233558e-06, + "loss": 0.1527, + "step": 510 + }, + { + "epoch": 0.3281952472703918, + "grad_norm": 1.262709216353945, + "learning_rate": 9.997417617928807e-06, + "loss": 0.1589, + "step": 511 + }, + { + "epoch": 0.32883750802825945, + "grad_norm": 1.6814671797849643, + "learning_rate": 9.997296121450195e-06, + "loss": 0.2748, + "step": 512 + }, + { + "epoch": 0.32947976878612717, + "grad_norm": 1.4463982840157266, + "learning_rate": 9.99717183296762e-06, + "loss": 0.1723, + "step": 513 + }, + { + "epoch": 0.3301220295439949, + "grad_norm": 1.0079089599589608, + "learning_rate": 9.997044752550526e-06, + "loss": 0.0535, + "step": 514 + }, + { + "epoch": 0.33076429030186255, + "grad_norm": 1.15956156589621, + "learning_rate": 9.996914880269913e-06, + "loss": 0.1285, + "step": 515 + }, + { + "epoch": 0.33140655105973027, + "grad_norm": 0.9670380926919758, + "learning_rate": 9.996782216198337e-06, + "loss": 0.0921, + "step": 516 + }, + { + "epoch": 0.33204881181759793, + "grad_norm": 1.214737861964431, + "learning_rate": 9.996646760409923e-06, + "loss": 0.0892, + "step": 517 + }, + { + "epoch": 0.33269107257546565, + "grad_norm": 1.2130664819513364, + "learning_rate": 9.996508512980346e-06, + "loss": 0.1739, + "step": 518 + }, + { + "epoch": 0.3333333333333333, + "grad_norm": 1.240641127168716, + "learning_rate": 9.996367473986848e-06, + "loss": 0.1458, + "step": 519 + }, + { + "epoch": 0.33397559409120103, + "grad_norm": 1.2083839868473745, + "learning_rate": 9.996223643508227e-06, + "loss": 0.0706, + "step": 520 + }, + { + "epoch": 0.3346178548490687, + "grad_norm": 1.2933708273414886, + "learning_rate": 9.996077021624844e-06, + "loss": 0.1435, + "step": 521 + }, + { + "epoch": 0.3352601156069364, + "grad_norm": 0.9337229959166998, + "learning_rate": 9.995927608418612e-06, + "loss": 0.0907, + "step": 522 + }, + { + "epoch": 0.33590237636480413, + "grad_norm": 1.9252042744978477, + "learning_rate": 9.995775403973011e-06, + "loss": 0.1971, + "step": 523 + }, + { + "epoch": 0.3365446371226718, + "grad_norm": 1.6031072445812466, + "learning_rate": 9.99562040837308e-06, + "loss": 0.3139, + "step": 524 + }, + { + "epoch": 0.3371868978805395, + "grad_norm": 0.996559553947966, + "learning_rate": 9.995462621705414e-06, + "loss": 0.0686, + "step": 525 + }, + { + "epoch": 0.3378291586384072, + "grad_norm": 1.3360344689136863, + "learning_rate": 9.995302044058167e-06, + "loss": 0.1931, + "step": 526 + }, + { + "epoch": 0.3384714193962749, + "grad_norm": 1.3686569476537527, + "learning_rate": 9.995138675521058e-06, + "loss": 0.1955, + "step": 527 + }, + { + "epoch": 0.33911368015414256, + "grad_norm": 3.10414644225421, + "learning_rate": 9.99497251618536e-06, + "loss": 0.1412, + "step": 528 + }, + { + "epoch": 0.3397559409120103, + "grad_norm": 1.3277375882360216, + "learning_rate": 9.994803566143904e-06, + "loss": 0.0979, + "step": 529 + }, + { + "epoch": 0.340398201669878, + "grad_norm": 2.345629999106305, + "learning_rate": 9.994631825491086e-06, + "loss": 0.2156, + "step": 530 + }, + { + "epoch": 0.34104046242774566, + "grad_norm": 1.2713130743285694, + "learning_rate": 9.994457294322858e-06, + "loss": 0.1361, + "step": 531 + }, + { + "epoch": 0.3416827231856134, + "grad_norm": 1.5520305815831397, + "learning_rate": 9.994279972736729e-06, + "loss": 0.1664, + "step": 532 + }, + { + "epoch": 0.34232498394348104, + "grad_norm": 1.078252405688029, + "learning_rate": 9.99409986083177e-06, + "loss": 0.0938, + "step": 533 + }, + { + "epoch": 0.34296724470134876, + "grad_norm": 1.0721508234623964, + "learning_rate": 9.993916958708612e-06, + "loss": 0.0614, + "step": 534 + }, + { + "epoch": 0.3436095054592164, + "grad_norm": 1.9299032923934651, + "learning_rate": 9.99373126646944e-06, + "loss": 0.2264, + "step": 535 + }, + { + "epoch": 0.34425176621708414, + "grad_norm": 1.6020477482577367, + "learning_rate": 9.993542784218002e-06, + "loss": 0.1962, + "step": 536 + }, + { + "epoch": 0.3448940269749518, + "grad_norm": 1.3516642782561228, + "learning_rate": 9.993351512059603e-06, + "loss": 0.1901, + "step": 537 + }, + { + "epoch": 0.3455362877328195, + "grad_norm": 1.369367142639903, + "learning_rate": 9.99315745010111e-06, + "loss": 0.2197, + "step": 538 + }, + { + "epoch": 0.34617854849068724, + "grad_norm": 1.1828900212908962, + "learning_rate": 9.992960598450942e-06, + "loss": 0.1305, + "step": 539 + }, + { + "epoch": 0.3468208092485549, + "grad_norm": 1.202302586671264, + "learning_rate": 9.992760957219083e-06, + "loss": 0.1504, + "step": 540 + }, + { + "epoch": 0.3474630700064226, + "grad_norm": 1.3769535203374934, + "learning_rate": 9.992558526517072e-06, + "loss": 0.1118, + "step": 541 + }, + { + "epoch": 0.3481053307642903, + "grad_norm": 1.4308686376873956, + "learning_rate": 9.992353306458009e-06, + "loss": 0.1867, + "step": 542 + }, + { + "epoch": 0.348747591522158, + "grad_norm": 1.2989920932643353, + "learning_rate": 9.992145297156551e-06, + "loss": 0.1753, + "step": 543 + }, + { + "epoch": 0.34938985228002567, + "grad_norm": 1.1726411489427073, + "learning_rate": 9.991934498728913e-06, + "loss": 0.1372, + "step": 544 + }, + { + "epoch": 0.3500321130378934, + "grad_norm": 1.48598570301162, + "learning_rate": 9.991720911292867e-06, + "loss": 0.2053, + "step": 545 + }, + { + "epoch": 0.35067437379576105, + "grad_norm": 1.6170291068282865, + "learning_rate": 9.991504534967746e-06, + "loss": 0.1606, + "step": 546 + }, + { + "epoch": 0.35131663455362877, + "grad_norm": 1.3227915552042522, + "learning_rate": 9.991285369874442e-06, + "loss": 0.143, + "step": 547 + }, + { + "epoch": 0.3519588953114965, + "grad_norm": 1.7252581230053776, + "learning_rate": 9.991063416135402e-06, + "loss": 0.3148, + "step": 548 + }, + { + "epoch": 0.35260115606936415, + "grad_norm": 1.2925090494434022, + "learning_rate": 9.99083867387463e-06, + "loss": 0.1788, + "step": 549 + }, + { + "epoch": 0.35324341682723187, + "grad_norm": 1.2650423288597406, + "learning_rate": 9.990611143217693e-06, + "loss": 0.1652, + "step": 550 + }, + { + "epoch": 0.35388567758509953, + "grad_norm": 0.9930051178452783, + "learning_rate": 9.990380824291712e-06, + "loss": 0.0597, + "step": 551 + }, + { + "epoch": 0.35452793834296725, + "grad_norm": 1.377143119783634, + "learning_rate": 9.990147717225366e-06, + "loss": 0.1982, + "step": 552 + }, + { + "epoch": 0.3551701991008349, + "grad_norm": 1.2822023713305273, + "learning_rate": 9.989911822148895e-06, + "loss": 0.1713, + "step": 553 + }, + { + "epoch": 0.35581245985870263, + "grad_norm": 1.3976526931978772, + "learning_rate": 9.989673139194091e-06, + "loss": 0.2449, + "step": 554 + }, + { + "epoch": 0.35645472061657035, + "grad_norm": 1.4033696178825852, + "learning_rate": 9.989431668494312e-06, + "loss": 0.2642, + "step": 555 + }, + { + "epoch": 0.357096981374438, + "grad_norm": 1.589838054546976, + "learning_rate": 9.989187410184462e-06, + "loss": 0.2113, + "step": 556 + }, + { + "epoch": 0.35773924213230573, + "grad_norm": 1.437515180863922, + "learning_rate": 9.988940364401012e-06, + "loss": 0.153, + "step": 557 + }, + { + "epoch": 0.3583815028901734, + "grad_norm": 1.0451428605389297, + "learning_rate": 9.988690531281988e-06, + "loss": 0.1162, + "step": 558 + }, + { + "epoch": 0.3590237636480411, + "grad_norm": 1.0419337667490602, + "learning_rate": 9.98843791096697e-06, + "loss": 0.0932, + "step": 559 + }, + { + "epoch": 0.3596660244059088, + "grad_norm": 1.1085564021109642, + "learning_rate": 9.9881825035971e-06, + "loss": 0.0986, + "step": 560 + }, + { + "epoch": 0.3603082851637765, + "grad_norm": 1.5079202078821257, + "learning_rate": 9.987924309315076e-06, + "loss": 0.2373, + "step": 561 + }, + { + "epoch": 0.36095054592164416, + "grad_norm": 1.4911852808458281, + "learning_rate": 9.987663328265147e-06, + "loss": 0.2272, + "step": 562 + }, + { + "epoch": 0.3615928066795119, + "grad_norm": 1.14967677511692, + "learning_rate": 9.98739956059313e-06, + "loss": 0.1085, + "step": 563 + }, + { + "epoch": 0.3622350674373796, + "grad_norm": 1.4753686709067333, + "learning_rate": 9.987133006446387e-06, + "loss": 0.1975, + "step": 564 + }, + { + "epoch": 0.36287732819524726, + "grad_norm": 1.5354446505205828, + "learning_rate": 9.986863665973845e-06, + "loss": 0.2895, + "step": 565 + }, + { + "epoch": 0.363519588953115, + "grad_norm": 1.0342595097846297, + "learning_rate": 9.986591539325987e-06, + "loss": 0.1259, + "step": 566 + }, + { + "epoch": 0.36416184971098264, + "grad_norm": 1.0965265528094366, + "learning_rate": 9.986316626654851e-06, + "loss": 0.0779, + "step": 567 + }, + { + "epoch": 0.36480411046885036, + "grad_norm": 1.070035873044907, + "learning_rate": 9.986038928114027e-06, + "loss": 0.1222, + "step": 568 + }, + { + "epoch": 0.365446371226718, + "grad_norm": 1.0644998872854032, + "learning_rate": 9.985758443858671e-06, + "loss": 0.1004, + "step": 569 + }, + { + "epoch": 0.36608863198458574, + "grad_norm": 0.9623932246852366, + "learning_rate": 9.985475174045487e-06, + "loss": 0.0829, + "step": 570 + }, + { + "epoch": 0.36673089274245346, + "grad_norm": 1.160648669898989, + "learning_rate": 9.98518911883274e-06, + "loss": 0.1065, + "step": 571 + }, + { + "epoch": 0.3673731535003211, + "grad_norm": 1.1926541575831076, + "learning_rate": 9.98490027838025e-06, + "loss": 0.1465, + "step": 572 + }, + { + "epoch": 0.36801541425818884, + "grad_norm": 1.1342123650396483, + "learning_rate": 9.984608652849393e-06, + "loss": 0.0862, + "step": 573 + }, + { + "epoch": 0.3686576750160565, + "grad_norm": 0.9527064110133098, + "learning_rate": 9.984314242403101e-06, + "loss": 0.0809, + "step": 574 + }, + { + "epoch": 0.3692999357739242, + "grad_norm": 1.301322789300309, + "learning_rate": 9.984017047205862e-06, + "loss": 0.1875, + "step": 575 + }, + { + "epoch": 0.3699421965317919, + "grad_norm": 1.0528957071742613, + "learning_rate": 9.983717067423721e-06, + "loss": 0.0775, + "step": 576 + }, + { + "epoch": 0.3705844572896596, + "grad_norm": 0.9998573478826787, + "learning_rate": 9.983414303224275e-06, + "loss": 0.1108, + "step": 577 + }, + { + "epoch": 0.37122671804752727, + "grad_norm": 0.8154045866660634, + "learning_rate": 9.983108754776681e-06, + "loss": 0.0535, + "step": 578 + }, + { + "epoch": 0.371868978805395, + "grad_norm": 1.5863089509997181, + "learning_rate": 9.982800422251649e-06, + "loss": 0.1804, + "step": 579 + }, + { + "epoch": 0.3725112395632627, + "grad_norm": 1.2374581339811792, + "learning_rate": 9.982489305821447e-06, + "loss": 0.2103, + "step": 580 + }, + { + "epoch": 0.37315350032113037, + "grad_norm": 1.2719196446670826, + "learning_rate": 9.982175405659893e-06, + "loss": 0.1813, + "step": 581 + }, + { + "epoch": 0.3737957610789981, + "grad_norm": 1.382435426987582, + "learning_rate": 9.981858721942368e-06, + "loss": 0.2144, + "step": 582 + }, + { + "epoch": 0.37443802183686575, + "grad_norm": 1.0022546941070802, + "learning_rate": 9.981539254845803e-06, + "loss": 0.0804, + "step": 583 + }, + { + "epoch": 0.37508028259473347, + "grad_norm": 1.167990480453959, + "learning_rate": 9.981217004548682e-06, + "loss": 0.0986, + "step": 584 + }, + { + "epoch": 0.37572254335260113, + "grad_norm": 1.4004633302098262, + "learning_rate": 9.980891971231052e-06, + "loss": 0.1962, + "step": 585 + }, + { + "epoch": 0.37636480411046885, + "grad_norm": 1.5990158880389318, + "learning_rate": 9.980564155074508e-06, + "loss": 0.1459, + "step": 586 + }, + { + "epoch": 0.37700706486833657, + "grad_norm": 1.202888887960096, + "learning_rate": 9.9802335562622e-06, + "loss": 0.0705, + "step": 587 + }, + { + "epoch": 0.37764932562620424, + "grad_norm": 1.168835224164736, + "learning_rate": 9.979900174978839e-06, + "loss": 0.1026, + "step": 588 + }, + { + "epoch": 0.37829158638407195, + "grad_norm": 1.511607107884817, + "learning_rate": 9.97956401141068e-06, + "loss": 0.2384, + "step": 589 + }, + { + "epoch": 0.3789338471419396, + "grad_norm": 1.2182761475085255, + "learning_rate": 9.979225065745543e-06, + "loss": 0.1372, + "step": 590 + }, + { + "epoch": 0.37957610789980734, + "grad_norm": 0.9767699667754075, + "learning_rate": 9.978883338172797e-06, + "loss": 0.0824, + "step": 591 + }, + { + "epoch": 0.380218368657675, + "grad_norm": 1.7803236219987635, + "learning_rate": 9.978538828883366e-06, + "loss": 0.1222, + "step": 592 + }, + { + "epoch": 0.3808606294155427, + "grad_norm": 1.1148822549888213, + "learning_rate": 9.978191538069729e-06, + "loss": 0.1331, + "step": 593 + }, + { + "epoch": 0.3815028901734104, + "grad_norm": 0.9318069668534283, + "learning_rate": 9.977841465925918e-06, + "loss": 0.0753, + "step": 594 + }, + { + "epoch": 0.3821451509312781, + "grad_norm": 1.1629310966662918, + "learning_rate": 9.977488612647517e-06, + "loss": 0.1414, + "step": 595 + }, + { + "epoch": 0.3827874116891458, + "grad_norm": 1.3349589686128955, + "learning_rate": 9.977132978431671e-06, + "loss": 0.2012, + "step": 596 + }, + { + "epoch": 0.3834296724470135, + "grad_norm": 1.401464006845661, + "learning_rate": 9.97677456347707e-06, + "loss": 0.117, + "step": 597 + }, + { + "epoch": 0.3840719332048812, + "grad_norm": 1.120605901294766, + "learning_rate": 9.97641336798396e-06, + "loss": 0.1285, + "step": 598 + }, + { + "epoch": 0.38471419396274886, + "grad_norm": 1.392929688056288, + "learning_rate": 9.976049392154148e-06, + "loss": 0.2291, + "step": 599 + }, + { + "epoch": 0.3853564547206166, + "grad_norm": 1.066103415277783, + "learning_rate": 9.975682636190984e-06, + "loss": 0.1281, + "step": 600 + }, + { + "epoch": 0.38599871547848424, + "grad_norm": 1.2119678847853228, + "learning_rate": 9.975313100299377e-06, + "loss": 0.1587, + "step": 601 + }, + { + "epoch": 0.38664097623635196, + "grad_norm": 2.6101016057604927, + "learning_rate": 9.974940784685787e-06, + "loss": 0.2458, + "step": 602 + }, + { + "epoch": 0.3872832369942196, + "grad_norm": 1.2994255404615325, + "learning_rate": 9.974565689558228e-06, + "loss": 0.1546, + "step": 603 + }, + { + "epoch": 0.38792549775208734, + "grad_norm": 1.0641453479458467, + "learning_rate": 9.974187815126266e-06, + "loss": 0.1389, + "step": 604 + }, + { + "epoch": 0.38856775850995506, + "grad_norm": 1.4978990795618343, + "learning_rate": 9.973807161601023e-06, + "loss": 0.2078, + "step": 605 + }, + { + "epoch": 0.3892100192678227, + "grad_norm": 0.9998347623298768, + "learning_rate": 9.973423729195168e-06, + "loss": 0.0829, + "step": 606 + }, + { + "epoch": 0.38985228002569045, + "grad_norm": 2.1086621515492836, + "learning_rate": 9.97303751812293e-06, + "loss": 0.0859, + "step": 607 + }, + { + "epoch": 0.3904945407835581, + "grad_norm": 1.4399130993342173, + "learning_rate": 9.972648528600082e-06, + "loss": 0.1669, + "step": 608 + }, + { + "epoch": 0.3911368015414258, + "grad_norm": 1.0928586079869564, + "learning_rate": 9.972256760843956e-06, + "loss": 0.148, + "step": 609 + }, + { + "epoch": 0.3917790622992935, + "grad_norm": 1.0926464575636161, + "learning_rate": 9.971862215073432e-06, + "loss": 0.0786, + "step": 610 + }, + { + "epoch": 0.3924213230571612, + "grad_norm": 1.4326346206830427, + "learning_rate": 9.971464891508946e-06, + "loss": 0.188, + "step": 611 + }, + { + "epoch": 0.3930635838150289, + "grad_norm": 1.0870657741471015, + "learning_rate": 9.971064790372484e-06, + "loss": 0.1373, + "step": 612 + }, + { + "epoch": 0.3937058445728966, + "grad_norm": 1.030009776232205, + "learning_rate": 9.970661911887581e-06, + "loss": 0.0825, + "step": 613 + }, + { + "epoch": 0.3943481053307643, + "grad_norm": 1.2325589115296487, + "learning_rate": 9.970256256279328e-06, + "loss": 0.1463, + "step": 614 + }, + { + "epoch": 0.394990366088632, + "grad_norm": 1.117405946374561, + "learning_rate": 9.969847823774365e-06, + "loss": 0.1656, + "step": 615 + }, + { + "epoch": 0.3956326268464997, + "grad_norm": 1.3785072473682172, + "learning_rate": 9.969436614600888e-06, + "loss": 0.204, + "step": 616 + }, + { + "epoch": 0.39627488760436735, + "grad_norm": 0.9881583322453033, + "learning_rate": 9.969022628988636e-06, + "loss": 0.1152, + "step": 617 + }, + { + "epoch": 0.3969171483622351, + "grad_norm": 1.113758800990811, + "learning_rate": 9.968605867168906e-06, + "loss": 0.1387, + "step": 618 + }, + { + "epoch": 0.39755940912010274, + "grad_norm": 1.2563101004836599, + "learning_rate": 9.968186329374546e-06, + "loss": 0.2171, + "step": 619 + }, + { + "epoch": 0.39820166987797045, + "grad_norm": 1.1680972398202725, + "learning_rate": 9.96776401583995e-06, + "loss": 0.1482, + "step": 620 + }, + { + "epoch": 0.3988439306358382, + "grad_norm": 1.1361200100668198, + "learning_rate": 9.967338926801066e-06, + "loss": 0.1464, + "step": 621 + }, + { + "epoch": 0.39948619139370584, + "grad_norm": 1.0688797239750034, + "learning_rate": 9.966911062495395e-06, + "loss": 0.1148, + "step": 622 + }, + { + "epoch": 0.40012845215157355, + "grad_norm": 1.4659899315730192, + "learning_rate": 9.966480423161984e-06, + "loss": 0.2452, + "step": 623 + }, + { + "epoch": 0.4007707129094412, + "grad_norm": 1.1854649465413343, + "learning_rate": 9.966047009041432e-06, + "loss": 0.1798, + "step": 624 + }, + { + "epoch": 0.40141297366730894, + "grad_norm": 1.369892981847824, + "learning_rate": 9.965610820375891e-06, + "loss": 0.2009, + "step": 625 + }, + { + "epoch": 0.4020552344251766, + "grad_norm": 1.248983879216194, + "learning_rate": 9.965171857409058e-06, + "loss": 0.1199, + "step": 626 + }, + { + "epoch": 0.4026974951830443, + "grad_norm": 1.3317090482490885, + "learning_rate": 9.964730120386186e-06, + "loss": 0.1764, + "step": 627 + }, + { + "epoch": 0.40333975594091204, + "grad_norm": 1.3437180314324562, + "learning_rate": 9.964285609554072e-06, + "loss": 0.1705, + "step": 628 + }, + { + "epoch": 0.4039820166987797, + "grad_norm": 1.211084722091283, + "learning_rate": 9.963838325161068e-06, + "loss": 0.1325, + "step": 629 + }, + { + "epoch": 0.4046242774566474, + "grad_norm": 0.9110068755957377, + "learning_rate": 9.963388267457071e-06, + "loss": 0.0733, + "step": 630 + }, + { + "epoch": 0.4052665382145151, + "grad_norm": 0.9684702450491173, + "learning_rate": 9.962935436693528e-06, + "loss": 0.0796, + "step": 631 + }, + { + "epoch": 0.4059087989723828, + "grad_norm": 1.1669994749947439, + "learning_rate": 9.962479833123442e-06, + "loss": 0.1741, + "step": 632 + }, + { + "epoch": 0.40655105973025046, + "grad_norm": 1.0659707071100517, + "learning_rate": 9.962021457001358e-06, + "loss": 0.0823, + "step": 633 + }, + { + "epoch": 0.4071933204881182, + "grad_norm": 1.1882478610742946, + "learning_rate": 9.96156030858337e-06, + "loss": 0.1555, + "step": 634 + }, + { + "epoch": 0.40783558124598585, + "grad_norm": 1.2098332671300456, + "learning_rate": 9.961096388127125e-06, + "loss": 0.1637, + "step": 635 + }, + { + "epoch": 0.40847784200385356, + "grad_norm": 1.1047282983158175, + "learning_rate": 9.960629695891815e-06, + "loss": 0.1293, + "step": 636 + }, + { + "epoch": 0.4091201027617213, + "grad_norm": 1.169169051546672, + "learning_rate": 9.960160232138184e-06, + "loss": 0.1505, + "step": 637 + }, + { + "epoch": 0.40976236351958895, + "grad_norm": 1.0815389910270992, + "learning_rate": 9.95968799712852e-06, + "loss": 0.1316, + "step": 638 + }, + { + "epoch": 0.41040462427745666, + "grad_norm": 1.0895311832493424, + "learning_rate": 9.959212991126668e-06, + "loss": 0.1226, + "step": 639 + }, + { + "epoch": 0.4110468850353243, + "grad_norm": 1.4664622999516055, + "learning_rate": 9.958735214398009e-06, + "loss": 0.1765, + "step": 640 + }, + { + "epoch": 0.41168914579319205, + "grad_norm": 1.1443262993373904, + "learning_rate": 9.958254667209482e-06, + "loss": 0.1391, + "step": 641 + }, + { + "epoch": 0.4123314065510597, + "grad_norm": 1.372209795290251, + "learning_rate": 9.957771349829566e-06, + "loss": 0.2396, + "step": 642 + }, + { + "epoch": 0.41297366730892743, + "grad_norm": 1.1850853795691276, + "learning_rate": 9.957285262528298e-06, + "loss": 0.1789, + "step": 643 + }, + { + "epoch": 0.41361592806679515, + "grad_norm": 1.0700652617240567, + "learning_rate": 9.956796405577251e-06, + "loss": 0.1039, + "step": 644 + }, + { + "epoch": 0.4142581888246628, + "grad_norm": 1.209904110355157, + "learning_rate": 9.956304779249555e-06, + "loss": 0.1732, + "step": 645 + }, + { + "epoch": 0.41490044958253053, + "grad_norm": 1.2014699999769767, + "learning_rate": 9.95581038381988e-06, + "loss": 0.1921, + "step": 646 + }, + { + "epoch": 0.4155427103403982, + "grad_norm": 1.0638521029914398, + "learning_rate": 9.955313219564447e-06, + "loss": 0.1225, + "step": 647 + }, + { + "epoch": 0.4161849710982659, + "grad_norm": 1.2985055692253444, + "learning_rate": 9.954813286761021e-06, + "loss": 0.2235, + "step": 648 + }, + { + "epoch": 0.4168272318561336, + "grad_norm": 1.2787365728060542, + "learning_rate": 9.95431058568892e-06, + "loss": 0.2082, + "step": 649 + }, + { + "epoch": 0.4174694926140013, + "grad_norm": 1.1762193640482639, + "learning_rate": 9.953805116629003e-06, + "loss": 0.1634, + "step": 650 + }, + { + "epoch": 0.41811175337186895, + "grad_norm": 1.5450964950399455, + "learning_rate": 9.953296879863676e-06, + "loss": 0.2769, + "step": 651 + }, + { + "epoch": 0.4187540141297367, + "grad_norm": 1.4521029434520185, + "learning_rate": 9.952785875676894e-06, + "loss": 0.3136, + "step": 652 + }, + { + "epoch": 0.4193962748876044, + "grad_norm": 1.2768980505465175, + "learning_rate": 9.952272104354156e-06, + "loss": 0.1976, + "step": 653 + }, + { + "epoch": 0.42003853564547206, + "grad_norm": 1.0200784435994328, + "learning_rate": 9.951755566182506e-06, + "loss": 0.0775, + "step": 654 + }, + { + "epoch": 0.4206807964033398, + "grad_norm": 1.015288120179267, + "learning_rate": 9.951236261450538e-06, + "loss": 0.1175, + "step": 655 + }, + { + "epoch": 0.42132305716120744, + "grad_norm": 1.250772871161883, + "learning_rate": 9.950714190448387e-06, + "loss": 0.1551, + "step": 656 + }, + { + "epoch": 0.42196531791907516, + "grad_norm": 1.3349097749533692, + "learning_rate": 9.950189353467735e-06, + "loss": 0.2068, + "step": 657 + }, + { + "epoch": 0.4226075786769428, + "grad_norm": 1.5030410244540928, + "learning_rate": 9.949661750801813e-06, + "loss": 0.2759, + "step": 658 + }, + { + "epoch": 0.42324983943481054, + "grad_norm": 1.1001832802997873, + "learning_rate": 9.94913138274539e-06, + "loss": 0.136, + "step": 659 + }, + { + "epoch": 0.4238921001926782, + "grad_norm": 1.0730040644130878, + "learning_rate": 9.948598249594788e-06, + "loss": 0.1085, + "step": 660 + }, + { + "epoch": 0.4245343609505459, + "grad_norm": 0.9064725180801447, + "learning_rate": 9.94806235164787e-06, + "loss": 0.0929, + "step": 661 + }, + { + "epoch": 0.42517662170841364, + "grad_norm": 1.079634889618215, + "learning_rate": 9.94752368920404e-06, + "loss": 0.1415, + "step": 662 + }, + { + "epoch": 0.4258188824662813, + "grad_norm": 1.3093125889824533, + "learning_rate": 9.946982262564254e-06, + "loss": 0.177, + "step": 663 + }, + { + "epoch": 0.426461143224149, + "grad_norm": 0.9532911950135895, + "learning_rate": 9.946438072031007e-06, + "loss": 0.1096, + "step": 664 + }, + { + "epoch": 0.4271034039820167, + "grad_norm": 1.3264873204520973, + "learning_rate": 9.94589111790834e-06, + "loss": 0.1905, + "step": 665 + }, + { + "epoch": 0.4277456647398844, + "grad_norm": 0.9975834475948804, + "learning_rate": 9.945341400501838e-06, + "loss": 0.0971, + "step": 666 + }, + { + "epoch": 0.42838792549775206, + "grad_norm": 1.1420759923724544, + "learning_rate": 9.944788920118628e-06, + "loss": 0.1068, + "step": 667 + }, + { + "epoch": 0.4290301862556198, + "grad_norm": 1.4990942282951418, + "learning_rate": 9.944233677067388e-06, + "loss": 0.1623, + "step": 668 + }, + { + "epoch": 0.4296724470134875, + "grad_norm": 1.1525614000473254, + "learning_rate": 9.943675671658327e-06, + "loss": 0.1494, + "step": 669 + }, + { + "epoch": 0.43031470777135516, + "grad_norm": 1.2058602291539142, + "learning_rate": 9.943114904203209e-06, + "loss": 0.1954, + "step": 670 + }, + { + "epoch": 0.4309569685292229, + "grad_norm": 1.1875999659660574, + "learning_rate": 9.942551375015333e-06, + "loss": 0.1339, + "step": 671 + }, + { + "epoch": 0.43159922928709055, + "grad_norm": 1.210368723549343, + "learning_rate": 9.941985084409547e-06, + "loss": 0.0856, + "step": 672 + }, + { + "epoch": 0.43224149004495827, + "grad_norm": 1.3688874643258395, + "learning_rate": 9.941416032702238e-06, + "loss": 0.239, + "step": 673 + }, + { + "epoch": 0.43288375080282593, + "grad_norm": 1.007265494799531, + "learning_rate": 9.940844220211338e-06, + "loss": 0.1165, + "step": 674 + }, + { + "epoch": 0.43352601156069365, + "grad_norm": 0.9014081653060163, + "learning_rate": 9.940269647256319e-06, + "loss": 0.0863, + "step": 675 + }, + { + "epoch": 0.4341682723185613, + "grad_norm": 1.5569994444758755, + "learning_rate": 9.939692314158195e-06, + "loss": 0.1774, + "step": 676 + }, + { + "epoch": 0.43481053307642903, + "grad_norm": 0.94783521445833, + "learning_rate": 9.939112221239528e-06, + "loss": 0.0854, + "step": 677 + }, + { + "epoch": 0.43545279383429675, + "grad_norm": 0.9681285370530087, + "learning_rate": 9.938529368824416e-06, + "loss": 0.0979, + "step": 678 + }, + { + "epoch": 0.4360950545921644, + "grad_norm": 1.2192879874497826, + "learning_rate": 9.9379437572385e-06, + "loss": 0.1725, + "step": 679 + }, + { + "epoch": 0.43673731535003213, + "grad_norm": 1.0874164597054057, + "learning_rate": 9.93735538680896e-06, + "loss": 0.1236, + "step": 680 + }, + { + "epoch": 0.4373795761078998, + "grad_norm": 1.2964261223518025, + "learning_rate": 9.936764257864525e-06, + "loss": 0.2335, + "step": 681 + }, + { + "epoch": 0.4380218368657675, + "grad_norm": 1.274856455810396, + "learning_rate": 9.936170370735458e-06, + "loss": 0.1937, + "step": 682 + }, + { + "epoch": 0.4386640976236352, + "grad_norm": 1.1855924206615143, + "learning_rate": 9.935573725753565e-06, + "loss": 0.1671, + "step": 683 + }, + { + "epoch": 0.4393063583815029, + "grad_norm": 1.0405064207221113, + "learning_rate": 9.934974323252195e-06, + "loss": 0.0875, + "step": 684 + }, + { + "epoch": 0.4399486191393706, + "grad_norm": 1.35682085136994, + "learning_rate": 9.934372163566236e-06, + "loss": 0.2646, + "step": 685 + }, + { + "epoch": 0.4405908798972383, + "grad_norm": 1.1588403749907583, + "learning_rate": 9.933767247032114e-06, + "loss": 0.1724, + "step": 686 + }, + { + "epoch": 0.441233140655106, + "grad_norm": 1.1844246603190718, + "learning_rate": 9.933159573987799e-06, + "loss": 0.1236, + "step": 687 + }, + { + "epoch": 0.44187540141297366, + "grad_norm": 1.0565907305571356, + "learning_rate": 9.9325491447728e-06, + "loss": 0.1294, + "step": 688 + }, + { + "epoch": 0.4425176621708414, + "grad_norm": 1.147425489921103, + "learning_rate": 9.931935959728167e-06, + "loss": 0.151, + "step": 689 + }, + { + "epoch": 0.44315992292870904, + "grad_norm": 1.109070710062008, + "learning_rate": 9.931320019196485e-06, + "loss": 0.1059, + "step": 690 + }, + { + "epoch": 0.44380218368657676, + "grad_norm": 1.2667259234737038, + "learning_rate": 9.930701323521884e-06, + "loss": 0.1886, + "step": 691 + }, + { + "epoch": 0.4444444444444444, + "grad_norm": 1.2767289800146495, + "learning_rate": 9.93007987305003e-06, + "loss": 0.1783, + "step": 692 + }, + { + "epoch": 0.44508670520231214, + "grad_norm": 1.4469699659123147, + "learning_rate": 9.929455668128129e-06, + "loss": 0.285, + "step": 693 + }, + { + "epoch": 0.44572896596017986, + "grad_norm": 1.1224010218372236, + "learning_rate": 9.92882870910493e-06, + "loss": 0.137, + "step": 694 + }, + { + "epoch": 0.4463712267180475, + "grad_norm": 1.1818084985386024, + "learning_rate": 9.92819899633071e-06, + "loss": 0.1532, + "step": 695 + }, + { + "epoch": 0.44701348747591524, + "grad_norm": 1.0327746296446123, + "learning_rate": 9.927566530157298e-06, + "loss": 0.1112, + "step": 696 + }, + { + "epoch": 0.4476557482337829, + "grad_norm": 1.159184895805368, + "learning_rate": 9.92693131093805e-06, + "loss": 0.1401, + "step": 697 + }, + { + "epoch": 0.4482980089916506, + "grad_norm": 1.1786277519667105, + "learning_rate": 9.926293339027866e-06, + "loss": 0.1556, + "step": 698 + }, + { + "epoch": 0.4489402697495183, + "grad_norm": 1.2233840113064118, + "learning_rate": 9.925652614783185e-06, + "loss": 0.1421, + "step": 699 + }, + { + "epoch": 0.449582530507386, + "grad_norm": 1.229010093764339, + "learning_rate": 9.92500913856198e-06, + "loss": 0.1992, + "step": 700 + }, + { + "epoch": 0.45022479126525367, + "grad_norm": 0.8901407534248555, + "learning_rate": 9.924362910723762e-06, + "loss": 0.0715, + "step": 701 + }, + { + "epoch": 0.4508670520231214, + "grad_norm": 1.2678129474982192, + "learning_rate": 9.923713931629582e-06, + "loss": 0.1635, + "step": 702 + }, + { + "epoch": 0.4515093127809891, + "grad_norm": 1.218780277735721, + "learning_rate": 9.923062201642023e-06, + "loss": 0.1539, + "step": 703 + }, + { + "epoch": 0.45215157353885677, + "grad_norm": 2.1260446635252834, + "learning_rate": 9.922407721125216e-06, + "loss": 0.072, + "step": 704 + }, + { + "epoch": 0.4527938342967245, + "grad_norm": 1.0902827713423608, + "learning_rate": 9.921750490444813e-06, + "loss": 0.1197, + "step": 705 + }, + { + "epoch": 0.45343609505459215, + "grad_norm": 1.1389851195059886, + "learning_rate": 9.921090509968015e-06, + "loss": 0.1311, + "step": 706 + }, + { + "epoch": 0.45407835581245987, + "grad_norm": 1.1833854694769887, + "learning_rate": 9.920427780063556e-06, + "loss": 0.1194, + "step": 707 + }, + { + "epoch": 0.45472061657032753, + "grad_norm": 1.059190453243874, + "learning_rate": 9.9197623011017e-06, + "loss": 0.0945, + "step": 708 + }, + { + "epoch": 0.45536287732819525, + "grad_norm": 1.071300978181373, + "learning_rate": 9.919094073454258e-06, + "loss": 0.0747, + "step": 709 + }, + { + "epoch": 0.45600513808606297, + "grad_norm": 0.8127102602424098, + "learning_rate": 9.918423097494567e-06, + "loss": 0.0634, + "step": 710 + }, + { + "epoch": 0.45664739884393063, + "grad_norm": 0.950989516464118, + "learning_rate": 9.917749373597506e-06, + "loss": 0.0595, + "step": 711 + }, + { + "epoch": 0.45728965960179835, + "grad_norm": 1.8289975154681404, + "learning_rate": 9.917072902139481e-06, + "loss": 0.1141, + "step": 712 + }, + { + "epoch": 0.457931920359666, + "grad_norm": 1.2794849607398417, + "learning_rate": 9.916393683498447e-06, + "loss": 0.1766, + "step": 713 + }, + { + "epoch": 0.45857418111753373, + "grad_norm": 1.2440638630880883, + "learning_rate": 9.915711718053877e-06, + "loss": 0.2115, + "step": 714 + }, + { + "epoch": 0.4592164418754014, + "grad_norm": 0.965129264091513, + "learning_rate": 9.915027006186794e-06, + "loss": 0.0821, + "step": 715 + }, + { + "epoch": 0.4598587026332691, + "grad_norm": 1.374174214204627, + "learning_rate": 9.914339548279744e-06, + "loss": 0.1926, + "step": 716 + }, + { + "epoch": 0.4605009633911368, + "grad_norm": 1.2147828534936043, + "learning_rate": 9.913649344716814e-06, + "loss": 0.1584, + "step": 717 + }, + { + "epoch": 0.4611432241490045, + "grad_norm": 0.9595973028374526, + "learning_rate": 9.91295639588362e-06, + "loss": 0.1174, + "step": 718 + }, + { + "epoch": 0.4617854849068722, + "grad_norm": 1.3154117450193346, + "learning_rate": 9.912260702167319e-06, + "loss": 0.2119, + "step": 719 + }, + { + "epoch": 0.4624277456647399, + "grad_norm": 0.7727766930015317, + "learning_rate": 9.911562263956593e-06, + "loss": 0.0599, + "step": 720 + }, + { + "epoch": 0.4630700064226076, + "grad_norm": 1.3819096042382115, + "learning_rate": 9.910861081641662e-06, + "loss": 0.2483, + "step": 721 + }, + { + "epoch": 0.46371226718047526, + "grad_norm": 1.4905339087257363, + "learning_rate": 9.910157155614284e-06, + "loss": 0.2753, + "step": 722 + }, + { + "epoch": 0.464354527938343, + "grad_norm": 1.4190281631210864, + "learning_rate": 9.909450486267736e-06, + "loss": 0.1836, + "step": 723 + }, + { + "epoch": 0.46499678869621064, + "grad_norm": 1.229808469786657, + "learning_rate": 9.908741073996843e-06, + "loss": 0.2001, + "step": 724 + }, + { + "epoch": 0.46563904945407836, + "grad_norm": 1.2562870782482787, + "learning_rate": 9.908028919197951e-06, + "loss": 0.126, + "step": 725 + }, + { + "epoch": 0.4662813102119461, + "grad_norm": 1.270877524644976, + "learning_rate": 9.907314022268948e-06, + "loss": 0.1624, + "step": 726 + }, + { + "epoch": 0.46692357096981374, + "grad_norm": 0.9596131646948522, + "learning_rate": 9.906596383609243e-06, + "loss": 0.1242, + "step": 727 + }, + { + "epoch": 0.46756583172768146, + "grad_norm": 1.2425771044590748, + "learning_rate": 9.905876003619787e-06, + "loss": 0.2029, + "step": 728 + }, + { + "epoch": 0.4682080924855491, + "grad_norm": 1.2387828614085783, + "learning_rate": 9.90515288270306e-06, + "loss": 0.1779, + "step": 729 + }, + { + "epoch": 0.46885035324341684, + "grad_norm": 1.0916157119176912, + "learning_rate": 9.904427021263065e-06, + "loss": 0.1361, + "step": 730 + }, + { + "epoch": 0.4694926140012845, + "grad_norm": 1.066309807948377, + "learning_rate": 9.90369841970535e-06, + "loss": 0.1556, + "step": 731 + }, + { + "epoch": 0.4701348747591522, + "grad_norm": 1.2469458174544685, + "learning_rate": 9.902967078436982e-06, + "loss": 0.2402, + "step": 732 + }, + { + "epoch": 0.4707771355170199, + "grad_norm": 0.9686845782441501, + "learning_rate": 9.902232997866568e-06, + "loss": 0.0711, + "step": 733 + }, + { + "epoch": 0.4714193962748876, + "grad_norm": 1.1107320549329236, + "learning_rate": 9.901496178404237e-06, + "loss": 0.1184, + "step": 734 + }, + { + "epoch": 0.4720616570327553, + "grad_norm": 1.0460784160284242, + "learning_rate": 9.900756620461654e-06, + "loss": 0.119, + "step": 735 + }, + { + "epoch": 0.472703917790623, + "grad_norm": 1.388824954378202, + "learning_rate": 9.90001432445201e-06, + "loss": 0.2321, + "step": 736 + }, + { + "epoch": 0.4733461785484907, + "grad_norm": 1.0752517932723047, + "learning_rate": 9.89926929079003e-06, + "loss": 0.1219, + "step": 737 + }, + { + "epoch": 0.47398843930635837, + "grad_norm": 1.3933764551221566, + "learning_rate": 9.898521519891968e-06, + "loss": 0.1183, + "step": 738 + }, + { + "epoch": 0.4746307000642261, + "grad_norm": 1.0583209912812217, + "learning_rate": 9.897771012175602e-06, + "loss": 0.1475, + "step": 739 + }, + { + "epoch": 0.47527296082209375, + "grad_norm": 1.3541203344403927, + "learning_rate": 9.897017768060246e-06, + "loss": 0.2247, + "step": 740 + }, + { + "epoch": 0.47591522157996147, + "grad_norm": 1.1525796288560013, + "learning_rate": 9.896261787966738e-06, + "loss": 0.126, + "step": 741 + }, + { + "epoch": 0.4765574823378292, + "grad_norm": 1.406491564805693, + "learning_rate": 9.895503072317447e-06, + "loss": 0.2468, + "step": 742 + }, + { + "epoch": 0.47719974309569685, + "grad_norm": 0.9674583441142586, + "learning_rate": 9.89474162153627e-06, + "loss": 0.1099, + "step": 743 + }, + { + "epoch": 0.47784200385356457, + "grad_norm": 1.0314356279321963, + "learning_rate": 9.89397743604863e-06, + "loss": 0.1246, + "step": 744 + }, + { + "epoch": 0.47848426461143223, + "grad_norm": 1.0547011505364061, + "learning_rate": 9.893210516281482e-06, + "loss": 0.1504, + "step": 745 + }, + { + "epoch": 0.47912652536929995, + "grad_norm": 1.5206024423794595, + "learning_rate": 9.892440862663304e-06, + "loss": 0.3054, + "step": 746 + }, + { + "epoch": 0.4797687861271676, + "grad_norm": 0.7699650453355072, + "learning_rate": 9.891668475624106e-06, + "loss": 0.0665, + "step": 747 + }, + { + "epoch": 0.48041104688503533, + "grad_norm": 0.9152985701289573, + "learning_rate": 9.890893355595422e-06, + "loss": 0.1046, + "step": 748 + }, + { + "epoch": 0.481053307642903, + "grad_norm": 1.2805160113207577, + "learning_rate": 9.890115503010314e-06, + "loss": 0.223, + "step": 749 + }, + { + "epoch": 0.4816955684007707, + "grad_norm": 1.104263629722884, + "learning_rate": 9.889334918303368e-06, + "loss": 0.0889, + "step": 750 + }, + { + "epoch": 0.48233782915863843, + "grad_norm": 1.1290776634784445, + "learning_rate": 9.888551601910702e-06, + "loss": 0.1295, + "step": 751 + }, + { + "epoch": 0.4829800899165061, + "grad_norm": 1.089379124771661, + "learning_rate": 9.887765554269958e-06, + "loss": 0.1137, + "step": 752 + }, + { + "epoch": 0.4836223506743738, + "grad_norm": 1.1930221124246585, + "learning_rate": 9.8869767758203e-06, + "loss": 0.1212, + "step": 753 + }, + { + "epoch": 0.4842646114322415, + "grad_norm": 0.8690173870738976, + "learning_rate": 9.886185267002421e-06, + "loss": 0.0904, + "step": 754 + }, + { + "epoch": 0.4849068721901092, + "grad_norm": 1.1081888698761926, + "learning_rate": 9.885391028258543e-06, + "loss": 0.2018, + "step": 755 + }, + { + "epoch": 0.48554913294797686, + "grad_norm": 0.9889175308043536, + "learning_rate": 9.884594060032407e-06, + "loss": 0.1073, + "step": 756 + }, + { + "epoch": 0.4861913937058446, + "grad_norm": 1.0300025920341696, + "learning_rate": 9.883794362769279e-06, + "loss": 0.1044, + "step": 757 + }, + { + "epoch": 0.48683365446371224, + "grad_norm": 0.9801104641727205, + "learning_rate": 9.882991936915957e-06, + "loss": 0.1035, + "step": 758 + }, + { + "epoch": 0.48747591522157996, + "grad_norm": 1.0928734034484588, + "learning_rate": 9.882186782920756e-06, + "loss": 0.1317, + "step": 759 + }, + { + "epoch": 0.4881181759794477, + "grad_norm": 0.9114043823112452, + "learning_rate": 9.881378901233518e-06, + "loss": 0.118, + "step": 760 + }, + { + "epoch": 0.48876043673731534, + "grad_norm": 1.1671967352474493, + "learning_rate": 9.880568292305607e-06, + "loss": 0.1674, + "step": 761 + }, + { + "epoch": 0.48940269749518306, + "grad_norm": 1.2768209964300596, + "learning_rate": 9.879754956589919e-06, + "loss": 0.1737, + "step": 762 + }, + { + "epoch": 0.4900449582530507, + "grad_norm": 0.9681191662759554, + "learning_rate": 9.878938894540859e-06, + "loss": 0.108, + "step": 763 + }, + { + "epoch": 0.49068721901091844, + "grad_norm": 1.2668188219875824, + "learning_rate": 9.87812010661437e-06, + "loss": 0.1539, + "step": 764 + }, + { + "epoch": 0.4913294797687861, + "grad_norm": 0.8800750879899154, + "learning_rate": 9.877298593267906e-06, + "loss": 0.1002, + "step": 765 + }, + { + "epoch": 0.4919717405266538, + "grad_norm": 1.0146628859879874, + "learning_rate": 9.876474354960453e-06, + "loss": 0.1213, + "step": 766 + }, + { + "epoch": 0.49261400128452154, + "grad_norm": 1.1391403644147629, + "learning_rate": 9.875647392152512e-06, + "loss": 0.0861, + "step": 767 + }, + { + "epoch": 0.4932562620423892, + "grad_norm": 1.0953428083633854, + "learning_rate": 9.874817705306111e-06, + "loss": 0.1466, + "step": 768 + }, + { + "epoch": 0.4938985228002569, + "grad_norm": 1.184335866832412, + "learning_rate": 9.8739852948848e-06, + "loss": 0.195, + "step": 769 + }, + { + "epoch": 0.4945407835581246, + "grad_norm": 1.0451558651874406, + "learning_rate": 9.873150161353646e-06, + "loss": 0.2032, + "step": 770 + }, + { + "epoch": 0.4951830443159923, + "grad_norm": 1.2719053046304922, + "learning_rate": 9.872312305179244e-06, + "loss": 0.1829, + "step": 771 + }, + { + "epoch": 0.49582530507385997, + "grad_norm": 1.1628580990352344, + "learning_rate": 9.871471726829703e-06, + "loss": 0.1623, + "step": 772 + }, + { + "epoch": 0.4964675658317277, + "grad_norm": 0.9165119134362926, + "learning_rate": 9.870628426774659e-06, + "loss": 0.0729, + "step": 773 + }, + { + "epoch": 0.49710982658959535, + "grad_norm": 1.1415588568877013, + "learning_rate": 9.869782405485267e-06, + "loss": 0.184, + "step": 774 + }, + { + "epoch": 0.49775208734746307, + "grad_norm": 0.9925632166260551, + "learning_rate": 9.868933663434197e-06, + "loss": 0.1009, + "step": 775 + }, + { + "epoch": 0.4983943481053308, + "grad_norm": 0.9051634452020365, + "learning_rate": 9.868082201095651e-06, + "loss": 0.1202, + "step": 776 + }, + { + "epoch": 0.49903660886319845, + "grad_norm": 1.2389892288430444, + "learning_rate": 9.867228018945336e-06, + "loss": 0.2521, + "step": 777 + }, + { + "epoch": 0.49967886962106617, + "grad_norm": 1.1679428685586002, + "learning_rate": 9.866371117460491e-06, + "loss": 0.174, + "step": 778 + }, + { + "epoch": 0.5003211303789339, + "grad_norm": 1.212602783607426, + "learning_rate": 9.865511497119868e-06, + "loss": 0.1191, + "step": 779 + }, + { + "epoch": 0.5009633911368016, + "grad_norm": 0.9425830982698171, + "learning_rate": 9.86464915840374e-06, + "loss": 0.0787, + "step": 780 + }, + { + "epoch": 0.5016056518946692, + "grad_norm": 1.00643798295582, + "learning_rate": 9.863784101793895e-06, + "loss": 0.1591, + "step": 781 + }, + { + "epoch": 0.5022479126525369, + "grad_norm": 0.9445193294444623, + "learning_rate": 9.862916327773647e-06, + "loss": 0.112, + "step": 782 + }, + { + "epoch": 0.5028901734104047, + "grad_norm": 1.3699670270418685, + "learning_rate": 9.862045836827821e-06, + "loss": 0.2191, + "step": 783 + }, + { + "epoch": 0.5035324341682723, + "grad_norm": 0.9074541917105808, + "learning_rate": 9.861172629442763e-06, + "loss": 0.0746, + "step": 784 + }, + { + "epoch": 0.50417469492614, + "grad_norm": 1.0981871709583084, + "learning_rate": 9.86029670610634e-06, + "loss": 0.0967, + "step": 785 + }, + { + "epoch": 0.5048169556840078, + "grad_norm": 1.0659813339903383, + "learning_rate": 9.859418067307928e-06, + "loss": 0.1223, + "step": 786 + }, + { + "epoch": 0.5054592164418754, + "grad_norm": 1.17219100855256, + "learning_rate": 9.858536713538428e-06, + "loss": 0.183, + "step": 787 + }, + { + "epoch": 0.5061014771997431, + "grad_norm": 1.5218405461738584, + "learning_rate": 9.857652645290254e-06, + "loss": 0.2425, + "step": 788 + }, + { + "epoch": 0.5067437379576107, + "grad_norm": 1.0641987662173262, + "learning_rate": 9.856765863057338e-06, + "loss": 0.1582, + "step": 789 + }, + { + "epoch": 0.5073859987154785, + "grad_norm": 1.1454031692342046, + "learning_rate": 9.855876367335127e-06, + "loss": 0.153, + "step": 790 + }, + { + "epoch": 0.5080282594733462, + "grad_norm": 1.1984595716530422, + "learning_rate": 9.854984158620586e-06, + "loss": 0.228, + "step": 791 + }, + { + "epoch": 0.5086705202312138, + "grad_norm": 0.9508236781329201, + "learning_rate": 9.854089237412194e-06, + "loss": 0.0884, + "step": 792 + }, + { + "epoch": 0.5093127809890816, + "grad_norm": 1.3079803275012836, + "learning_rate": 9.853191604209946e-06, + "loss": 0.2067, + "step": 793 + }, + { + "epoch": 0.5099550417469493, + "grad_norm": 1.2662973359954535, + "learning_rate": 9.852291259515355e-06, + "loss": 0.1871, + "step": 794 + }, + { + "epoch": 0.5105973025048169, + "grad_norm": 1.3237485481455735, + "learning_rate": 9.85138820383144e-06, + "loss": 0.1934, + "step": 795 + }, + { + "epoch": 0.5112395632626846, + "grad_norm": 0.8535459961293539, + "learning_rate": 9.850482437662745e-06, + "loss": 0.0973, + "step": 796 + }, + { + "epoch": 0.5118818240205524, + "grad_norm": 1.2055621037886322, + "learning_rate": 9.849573961515326e-06, + "loss": 0.1694, + "step": 797 + }, + { + "epoch": 0.51252408477842, + "grad_norm": 1.1646071301868992, + "learning_rate": 9.848662775896747e-06, + "loss": 0.1458, + "step": 798 + }, + { + "epoch": 0.5131663455362877, + "grad_norm": 0.7941156512435993, + "learning_rate": 9.847748881316093e-06, + "loss": 0.085, + "step": 799 + }, + { + "epoch": 0.5138086062941555, + "grad_norm": 0.8588325845589845, + "learning_rate": 9.846832278283958e-06, + "loss": 0.0623, + "step": 800 + }, + { + "epoch": 0.5144508670520231, + "grad_norm": 0.8807326365878326, + "learning_rate": 9.84591296731245e-06, + "loss": 0.0748, + "step": 801 + }, + { + "epoch": 0.5150931278098908, + "grad_norm": 1.4859347098156772, + "learning_rate": 9.844990948915196e-06, + "loss": 0.3235, + "step": 802 + }, + { + "epoch": 0.5157353885677585, + "grad_norm": 1.2895084704900024, + "learning_rate": 9.844066223607323e-06, + "loss": 0.1744, + "step": 803 + }, + { + "epoch": 0.5163776493256262, + "grad_norm": 1.6114766521183463, + "learning_rate": 9.843138791905483e-06, + "loss": 0.2879, + "step": 804 + }, + { + "epoch": 0.5170199100834939, + "grad_norm": 1.2011075246921803, + "learning_rate": 9.842208654327833e-06, + "loss": 0.1718, + "step": 805 + }, + { + "epoch": 0.5176621708413616, + "grad_norm": 1.2579681311967643, + "learning_rate": 9.841275811394043e-06, + "loss": 0.1334, + "step": 806 + }, + { + "epoch": 0.5183044315992292, + "grad_norm": 1.1982028012329926, + "learning_rate": 9.840340263625297e-06, + "loss": 0.138, + "step": 807 + }, + { + "epoch": 0.518946692357097, + "grad_norm": 1.180684006832618, + "learning_rate": 9.839402011544286e-06, + "loss": 0.1416, + "step": 808 + }, + { + "epoch": 0.5195889531149647, + "grad_norm": 0.9211098663869206, + "learning_rate": 9.838461055675215e-06, + "loss": 0.1072, + "step": 809 + }, + { + "epoch": 0.5202312138728323, + "grad_norm": 1.2441807359753703, + "learning_rate": 9.837517396543799e-06, + "loss": 0.1994, + "step": 810 + }, + { + "epoch": 0.5208734746307001, + "grad_norm": 1.5337934244388542, + "learning_rate": 9.836571034677264e-06, + "loss": 0.0992, + "step": 811 + }, + { + "epoch": 0.5215157353885678, + "grad_norm": 1.2596649115271839, + "learning_rate": 9.835621970604343e-06, + "loss": 0.2069, + "step": 812 + }, + { + "epoch": 0.5221579961464354, + "grad_norm": 1.3831539352827265, + "learning_rate": 9.834670204855283e-06, + "loss": 0.1402, + "step": 813 + }, + { + "epoch": 0.5228002569043031, + "grad_norm": 1.1614473182751541, + "learning_rate": 9.833715737961837e-06, + "loss": 0.1895, + "step": 814 + }, + { + "epoch": 0.5234425176621709, + "grad_norm": 1.4207169568727465, + "learning_rate": 9.832758570457271e-06, + "loss": 0.2214, + "step": 815 + }, + { + "epoch": 0.5240847784200385, + "grad_norm": 1.62044864382908, + "learning_rate": 9.831798702876352e-06, + "loss": 0.2647, + "step": 816 + }, + { + "epoch": 0.5247270391779062, + "grad_norm": 0.9925594514368269, + "learning_rate": 9.830836135755366e-06, + "loss": 0.1165, + "step": 817 + }, + { + "epoch": 0.525369299935774, + "grad_norm": 1.0516655345418544, + "learning_rate": 9.829870869632098e-06, + "loss": 0.1131, + "step": 818 + }, + { + "epoch": 0.5260115606936416, + "grad_norm": 1.1228896103746269, + "learning_rate": 9.82890290504585e-06, + "loss": 0.1019, + "step": 819 + }, + { + "epoch": 0.5266538214515093, + "grad_norm": 1.2664555683581542, + "learning_rate": 9.827932242537422e-06, + "loss": 0.1718, + "step": 820 + }, + { + "epoch": 0.527296082209377, + "grad_norm": 0.9823982588982624, + "learning_rate": 9.826958882649129e-06, + "loss": 0.0813, + "step": 821 + }, + { + "epoch": 0.5279383429672447, + "grad_norm": 1.3555750806985507, + "learning_rate": 9.825982825924787e-06, + "loss": 0.2869, + "step": 822 + }, + { + "epoch": 0.5285806037251124, + "grad_norm": 1.4059149292519715, + "learning_rate": 9.825004072909724e-06, + "loss": 0.2783, + "step": 823 + }, + { + "epoch": 0.5292228644829801, + "grad_norm": 1.037182504861691, + "learning_rate": 9.824022624150772e-06, + "loss": 0.1475, + "step": 824 + }, + { + "epoch": 0.5298651252408478, + "grad_norm": 0.9400765044942545, + "learning_rate": 9.823038480196269e-06, + "loss": 0.1199, + "step": 825 + }, + { + "epoch": 0.5305073859987155, + "grad_norm": 1.2800364039404664, + "learning_rate": 9.822051641596058e-06, + "loss": 0.2038, + "step": 826 + }, + { + "epoch": 0.5311496467565832, + "grad_norm": 1.1068426977165122, + "learning_rate": 9.82106210890149e-06, + "loss": 0.1338, + "step": 827 + }, + { + "epoch": 0.5317919075144508, + "grad_norm": 0.9479583615953208, + "learning_rate": 9.82006988266542e-06, + "loss": 0.0977, + "step": 828 + }, + { + "epoch": 0.5324341682723186, + "grad_norm": 1.1832637720809327, + "learning_rate": 9.819074963442208e-06, + "loss": 0.1492, + "step": 829 + }, + { + "epoch": 0.5330764290301863, + "grad_norm": 0.8806417887666761, + "learning_rate": 9.818077351787717e-06, + "loss": 0.091, + "step": 830 + }, + { + "epoch": 0.5337186897880539, + "grad_norm": 1.3513972370811422, + "learning_rate": 9.817077048259314e-06, + "loss": 0.197, + "step": 831 + }, + { + "epoch": 0.5343609505459217, + "grad_norm": 0.9248557891364547, + "learning_rate": 9.816074053415874e-06, + "loss": 0.1039, + "step": 832 + }, + { + "epoch": 0.5350032113037894, + "grad_norm": 1.0778900589787808, + "learning_rate": 9.815068367817775e-06, + "loss": 0.148, + "step": 833 + }, + { + "epoch": 0.535645472061657, + "grad_norm": 1.3641094771680995, + "learning_rate": 9.814059992026893e-06, + "loss": 0.1401, + "step": 834 + }, + { + "epoch": 0.5362877328195247, + "grad_norm": 1.3229923584323375, + "learning_rate": 9.81304892660661e-06, + "loss": 0.23, + "step": 835 + }, + { + "epoch": 0.5369299935773925, + "grad_norm": 0.9874981816801378, + "learning_rate": 9.812035172121814e-06, + "loss": 0.1047, + "step": 836 + }, + { + "epoch": 0.5375722543352601, + "grad_norm": 0.9421152703869154, + "learning_rate": 9.811018729138892e-06, + "loss": 0.1004, + "step": 837 + }, + { + "epoch": 0.5382145150931278, + "grad_norm": 1.1843041301348378, + "learning_rate": 9.809999598225733e-06, + "loss": 0.1355, + "step": 838 + }, + { + "epoch": 0.5388567758509955, + "grad_norm": 1.095347182935344, + "learning_rate": 9.808977779951728e-06, + "loss": 0.1026, + "step": 839 + }, + { + "epoch": 0.5394990366088632, + "grad_norm": 0.9700681364772046, + "learning_rate": 9.807953274887771e-06, + "loss": 0.1181, + "step": 840 + }, + { + "epoch": 0.5401412973667309, + "grad_norm": 0.9601865222766974, + "learning_rate": 9.806926083606255e-06, + "loss": 0.0825, + "step": 841 + }, + { + "epoch": 0.5407835581245986, + "grad_norm": 1.2781083968038764, + "learning_rate": 9.805896206681075e-06, + "loss": 0.1606, + "step": 842 + }, + { + "epoch": 0.5414258188824663, + "grad_norm": 1.0705659272115575, + "learning_rate": 9.804863644687627e-06, + "loss": 0.1353, + "step": 843 + }, + { + "epoch": 0.542068079640334, + "grad_norm": 1.3492575003395122, + "learning_rate": 9.803828398202805e-06, + "loss": 0.2631, + "step": 844 + }, + { + "epoch": 0.5427103403982017, + "grad_norm": 1.3083724763072613, + "learning_rate": 9.802790467805006e-06, + "loss": 0.2021, + "step": 845 + }, + { + "epoch": 0.5433526011560693, + "grad_norm": 1.1998508850105545, + "learning_rate": 9.801749854074122e-06, + "loss": 0.2195, + "step": 846 + }, + { + "epoch": 0.5439948619139371, + "grad_norm": 1.2031173389916363, + "learning_rate": 9.800706557591551e-06, + "loss": 0.1885, + "step": 847 + }, + { + "epoch": 0.5446371226718048, + "grad_norm": 0.9984585485565872, + "learning_rate": 9.799660578940181e-06, + "loss": 0.1179, + "step": 848 + }, + { + "epoch": 0.5452793834296724, + "grad_norm": 1.0842096784505513, + "learning_rate": 9.798611918704408e-06, + "loss": 0.1432, + "step": 849 + }, + { + "epoch": 0.5459216441875402, + "grad_norm": 1.2366526583364414, + "learning_rate": 9.797560577470119e-06, + "loss": 0.2207, + "step": 850 + }, + { + "epoch": 0.5465639049454079, + "grad_norm": 1.0556901215823955, + "learning_rate": 9.7965065558247e-06, + "loss": 0.155, + "step": 851 + }, + { + "epoch": 0.5472061657032755, + "grad_norm": 0.7915842403594214, + "learning_rate": 9.79544985435704e-06, + "loss": 0.0723, + "step": 852 + }, + { + "epoch": 0.5478484264611432, + "grad_norm": 1.1921864999616434, + "learning_rate": 9.794390473657517e-06, + "loss": 0.1488, + "step": 853 + }, + { + "epoch": 0.548490687219011, + "grad_norm": 0.9295155782655121, + "learning_rate": 9.793328414318016e-06, + "loss": 0.1163, + "step": 854 + }, + { + "epoch": 0.5491329479768786, + "grad_norm": 1.240069114962376, + "learning_rate": 9.792263676931906e-06, + "loss": 0.2336, + "step": 855 + }, + { + "epoch": 0.5497752087347463, + "grad_norm": 0.9644391213102116, + "learning_rate": 9.79119626209406e-06, + "loss": 0.1443, + "step": 856 + }, + { + "epoch": 0.550417469492614, + "grad_norm": 0.8822600596256954, + "learning_rate": 9.79012617040085e-06, + "loss": 0.1041, + "step": 857 + }, + { + "epoch": 0.5510597302504817, + "grad_norm": 0.8690380758183222, + "learning_rate": 9.789053402450136e-06, + "loss": 0.0892, + "step": 858 + }, + { + "epoch": 0.5517019910083494, + "grad_norm": 1.347360712261439, + "learning_rate": 9.787977958841277e-06, + "loss": 0.2123, + "step": 859 + }, + { + "epoch": 0.552344251766217, + "grad_norm": 1.0732584036624615, + "learning_rate": 9.786899840175128e-06, + "loss": 0.1129, + "step": 860 + }, + { + "epoch": 0.5529865125240848, + "grad_norm": 1.257367957432509, + "learning_rate": 9.785819047054034e-06, + "loss": 0.2143, + "step": 861 + }, + { + "epoch": 0.5536287732819525, + "grad_norm": 0.9903342585918155, + "learning_rate": 9.78473558008184e-06, + "loss": 0.1403, + "step": 862 + }, + { + "epoch": 0.5542710340398201, + "grad_norm": 1.2037820956142335, + "learning_rate": 9.783649439863882e-06, + "loss": 0.1848, + "step": 863 + }, + { + "epoch": 0.5549132947976878, + "grad_norm": 1.0900215088980703, + "learning_rate": 9.78256062700699e-06, + "loss": 0.1314, + "step": 864 + }, + { + "epoch": 0.5555555555555556, + "grad_norm": 1.1001433182415752, + "learning_rate": 9.781469142119484e-06, + "loss": 0.1508, + "step": 865 + }, + { + "epoch": 0.5561978163134232, + "grad_norm": 1.6337940439715593, + "learning_rate": 9.780374985811184e-06, + "loss": 0.1416, + "step": 866 + }, + { + "epoch": 0.5568400770712909, + "grad_norm": 1.0391960974026144, + "learning_rate": 9.779278158693392e-06, + "loss": 0.1235, + "step": 867 + }, + { + "epoch": 0.5574823378291587, + "grad_norm": 1.0700345531999431, + "learning_rate": 9.778178661378916e-06, + "loss": 0.1485, + "step": 868 + }, + { + "epoch": 0.5581245985870263, + "grad_norm": 1.0760948998960185, + "learning_rate": 9.777076494482045e-06, + "loss": 0.1488, + "step": 869 + }, + { + "epoch": 0.558766859344894, + "grad_norm": 0.8753387270572457, + "learning_rate": 9.775971658618562e-06, + "loss": 0.0872, + "step": 870 + }, + { + "epoch": 0.5594091201027617, + "grad_norm": 1.3190956051614235, + "learning_rate": 9.774864154405743e-06, + "loss": 0.2109, + "step": 871 + }, + { + "epoch": 0.5600513808606294, + "grad_norm": 1.0421190836000214, + "learning_rate": 9.773753982462354e-06, + "loss": 0.1634, + "step": 872 + }, + { + "epoch": 0.5606936416184971, + "grad_norm": 1.0545509936977797, + "learning_rate": 9.772641143408652e-06, + "loss": 0.1655, + "step": 873 + }, + { + "epoch": 0.5613359023763648, + "grad_norm": 1.1229361183051547, + "learning_rate": 9.771525637866382e-06, + "loss": 0.1383, + "step": 874 + }, + { + "epoch": 0.5619781631342325, + "grad_norm": 1.235034194022337, + "learning_rate": 9.77040746645878e-06, + "loss": 0.2377, + "step": 875 + }, + { + "epoch": 0.5626204238921002, + "grad_norm": 1.0067402084230999, + "learning_rate": 9.769286629810572e-06, + "loss": 0.1772, + "step": 876 + }, + { + "epoch": 0.5632626846499679, + "grad_norm": 0.9406126413626174, + "learning_rate": 9.768163128547974e-06, + "loss": 0.113, + "step": 877 + }, + { + "epoch": 0.5639049454078355, + "grad_norm": 0.7794636841360506, + "learning_rate": 9.767036963298689e-06, + "loss": 0.0627, + "step": 878 + }, + { + "epoch": 0.5645472061657033, + "grad_norm": 1.0817618865946153, + "learning_rate": 9.765908134691907e-06, + "loss": 0.1112, + "step": 879 + }, + { + "epoch": 0.565189466923571, + "grad_norm": 1.192198117392412, + "learning_rate": 9.76477664335831e-06, + "loss": 0.2109, + "step": 880 + }, + { + "epoch": 0.5658317276814386, + "grad_norm": 1.1566294553675587, + "learning_rate": 9.763642489930064e-06, + "loss": 0.1602, + "step": 881 + }, + { + "epoch": 0.5664739884393064, + "grad_norm": 0.8628129981142905, + "learning_rate": 9.762505675040826e-06, + "loss": 0.0949, + "step": 882 + }, + { + "epoch": 0.5671162491971741, + "grad_norm": 1.302376908312678, + "learning_rate": 9.761366199325734e-06, + "loss": 0.231, + "step": 883 + }, + { + "epoch": 0.5677585099550417, + "grad_norm": 1.021462751224718, + "learning_rate": 9.76022406342142e-06, + "loss": 0.1153, + "step": 884 + }, + { + "epoch": 0.5684007707129094, + "grad_norm": 1.4798418530497832, + "learning_rate": 9.759079267965998e-06, + "loss": 0.2858, + "step": 885 + }, + { + "epoch": 0.5690430314707772, + "grad_norm": 1.3893496747743672, + "learning_rate": 9.75793181359907e-06, + "loss": 0.2572, + "step": 886 + }, + { + "epoch": 0.5696852922286448, + "grad_norm": 1.1496365987984296, + "learning_rate": 9.756781700961717e-06, + "loss": 0.1975, + "step": 887 + }, + { + "epoch": 0.5703275529865125, + "grad_norm": 1.0511187568186795, + "learning_rate": 9.755628930696513e-06, + "loss": 0.1488, + "step": 888 + }, + { + "epoch": 0.5709698137443803, + "grad_norm": 1.0038991614221606, + "learning_rate": 9.754473503447516e-06, + "loss": 0.1312, + "step": 889 + }, + { + "epoch": 0.5716120745022479, + "grad_norm": 1.0414644856021211, + "learning_rate": 9.753315419860264e-06, + "loss": 0.1771, + "step": 890 + }, + { + "epoch": 0.5722543352601156, + "grad_norm": 0.9403589918731176, + "learning_rate": 9.752154680581783e-06, + "loss": 0.0792, + "step": 891 + }, + { + "epoch": 0.5728965960179833, + "grad_norm": 1.8726259981838906, + "learning_rate": 9.750991286260582e-06, + "loss": 0.1794, + "step": 892 + }, + { + "epoch": 0.573538856775851, + "grad_norm": 1.1923342628119789, + "learning_rate": 9.749825237546652e-06, + "loss": 0.1518, + "step": 893 + }, + { + "epoch": 0.5741811175337187, + "grad_norm": 0.8775054122935616, + "learning_rate": 9.748656535091467e-06, + "loss": 0.0913, + "step": 894 + }, + { + "epoch": 0.5748233782915864, + "grad_norm": 1.162708700489589, + "learning_rate": 9.747485179547986e-06, + "loss": 0.1831, + "step": 895 + }, + { + "epoch": 0.575465639049454, + "grad_norm": 1.0856513767721323, + "learning_rate": 9.746311171570649e-06, + "loss": 0.1514, + "step": 896 + }, + { + "epoch": 0.5761078998073218, + "grad_norm": 1.549948780717054, + "learning_rate": 9.745134511815375e-06, + "loss": 0.1648, + "step": 897 + }, + { + "epoch": 0.5767501605651895, + "grad_norm": 1.0511816212932321, + "learning_rate": 9.74395520093957e-06, + "loss": 0.1237, + "step": 898 + }, + { + "epoch": 0.5773924213230571, + "grad_norm": 0.8874710448578135, + "learning_rate": 9.742773239602118e-06, + "loss": 0.1081, + "step": 899 + }, + { + "epoch": 0.5780346820809249, + "grad_norm": 1.1054078549597617, + "learning_rate": 9.741588628463384e-06, + "loss": 0.1675, + "step": 900 + }, + { + "epoch": 0.5786769428387926, + "grad_norm": 0.853557484544828, + "learning_rate": 9.740401368185212e-06, + "loss": 0.0959, + "step": 901 + }, + { + "epoch": 0.5793192035966602, + "grad_norm": 0.9516078508705335, + "learning_rate": 9.73921145943093e-06, + "loss": 0.1407, + "step": 902 + }, + { + "epoch": 0.5799614643545279, + "grad_norm": 1.0872239231584235, + "learning_rate": 9.738018902865342e-06, + "loss": 0.1315, + "step": 903 + }, + { + "epoch": 0.5806037251123957, + "grad_norm": 1.0481867484176843, + "learning_rate": 9.736823699154734e-06, + "loss": 0.1173, + "step": 904 + }, + { + "epoch": 0.5812459858702633, + "grad_norm": 1.0579886376070933, + "learning_rate": 9.73562584896687e-06, + "loss": 0.1293, + "step": 905 + }, + { + "epoch": 0.581888246628131, + "grad_norm": 1.0922923486001086, + "learning_rate": 9.73442535297099e-06, + "loss": 0.1475, + "step": 906 + }, + { + "epoch": 0.5825305073859988, + "grad_norm": 1.2187262498596858, + "learning_rate": 9.733222211837818e-06, + "loss": 0.164, + "step": 907 + }, + { + "epoch": 0.5831727681438664, + "grad_norm": 0.9626899102235692, + "learning_rate": 9.732016426239552e-06, + "loss": 0.1114, + "step": 908 + }, + { + "epoch": 0.5838150289017341, + "grad_norm": 1.2317450494510949, + "learning_rate": 9.730807996849864e-06, + "loss": 0.1666, + "step": 909 + }, + { + "epoch": 0.5844572896596018, + "grad_norm": 1.2083336475204853, + "learning_rate": 9.729596924343913e-06, + "loss": 0.2069, + "step": 910 + }, + { + "epoch": 0.5850995504174695, + "grad_norm": 1.094133666998947, + "learning_rate": 9.728383209398325e-06, + "loss": 0.1709, + "step": 911 + }, + { + "epoch": 0.5857418111753372, + "grad_norm": 0.837583155963877, + "learning_rate": 9.727166852691205e-06, + "loss": 0.0903, + "step": 912 + }, + { + "epoch": 0.5863840719332049, + "grad_norm": 1.0372615308710607, + "learning_rate": 9.725947854902137e-06, + "loss": 0.1176, + "step": 913 + }, + { + "epoch": 0.5870263326910726, + "grad_norm": 1.1781096439791983, + "learning_rate": 9.72472621671218e-06, + "loss": 0.1194, + "step": 914 + }, + { + "epoch": 0.5876685934489403, + "grad_norm": 1.1994544176565363, + "learning_rate": 9.723501938803863e-06, + "loss": 0.264, + "step": 915 + }, + { + "epoch": 0.588310854206808, + "grad_norm": 1.0907721954643648, + "learning_rate": 9.722275021861195e-06, + "loss": 0.1768, + "step": 916 + }, + { + "epoch": 0.5889531149646756, + "grad_norm": 1.24367256932414, + "learning_rate": 9.721045466569663e-06, + "loss": 0.2009, + "step": 917 + }, + { + "epoch": 0.5895953757225434, + "grad_norm": 1.0384838752263739, + "learning_rate": 9.719813273616216e-06, + "loss": 0.191, + "step": 918 + }, + { + "epoch": 0.590237636480411, + "grad_norm": 1.0935467220899417, + "learning_rate": 9.718578443689289e-06, + "loss": 0.1287, + "step": 919 + }, + { + "epoch": 0.5908798972382787, + "grad_norm": 0.8053864026691624, + "learning_rate": 9.717340977478781e-06, + "loss": 0.0953, + "step": 920 + }, + { + "epoch": 0.5915221579961464, + "grad_norm": 1.001051240065868, + "learning_rate": 9.716100875676071e-06, + "loss": 0.1375, + "step": 921 + }, + { + "epoch": 0.5921644187540142, + "grad_norm": 0.669959329908838, + "learning_rate": 9.714858138974006e-06, + "loss": 0.0576, + "step": 922 + }, + { + "epoch": 0.5928066795118818, + "grad_norm": 1.251400928119315, + "learning_rate": 9.713612768066907e-06, + "loss": 0.2548, + "step": 923 + }, + { + "epoch": 0.5934489402697495, + "grad_norm": 0.7584594689935764, + "learning_rate": 9.712364763650568e-06, + "loss": 0.0757, + "step": 924 + }, + { + "epoch": 0.5940912010276173, + "grad_norm": 1.4697276317066048, + "learning_rate": 9.71111412642225e-06, + "loss": 0.2143, + "step": 925 + }, + { + "epoch": 0.5947334617854849, + "grad_norm": 1.209515507072994, + "learning_rate": 9.709860857080688e-06, + "loss": 0.2036, + "step": 926 + }, + { + "epoch": 0.5953757225433526, + "grad_norm": 1.0757338077809466, + "learning_rate": 9.70860495632609e-06, + "loss": 0.0795, + "step": 927 + }, + { + "epoch": 0.5960179833012202, + "grad_norm": 0.9626580730209493, + "learning_rate": 9.707346424860129e-06, + "loss": 0.1399, + "step": 928 + }, + { + "epoch": 0.596660244059088, + "grad_norm": 0.7963443376040112, + "learning_rate": 9.706085263385951e-06, + "loss": 0.0866, + "step": 929 + }, + { + "epoch": 0.5973025048169557, + "grad_norm": 1.292544972326149, + "learning_rate": 9.704821472608169e-06, + "loss": 0.1952, + "step": 930 + }, + { + "epoch": 0.5979447655748233, + "grad_norm": 1.0729881395643057, + "learning_rate": 9.70355505323287e-06, + "loss": 0.1129, + "step": 931 + }, + { + "epoch": 0.5985870263326911, + "grad_norm": 1.1980813531494205, + "learning_rate": 9.702286005967602e-06, + "loss": 0.1657, + "step": 932 + }, + { + "epoch": 0.5992292870905588, + "grad_norm": 1.1021095720539393, + "learning_rate": 9.701014331521388e-06, + "loss": 0.1495, + "step": 933 + }, + { + "epoch": 0.5998715478484264, + "grad_norm": 1.2764903402778158, + "learning_rate": 9.699740030604716e-06, + "loss": 0.1279, + "step": 934 + }, + { + "epoch": 0.6005138086062941, + "grad_norm": 1.0830658466915088, + "learning_rate": 9.698463103929542e-06, + "loss": 0.1657, + "step": 935 + }, + { + "epoch": 0.6011560693641619, + "grad_norm": 0.95260112623732, + "learning_rate": 9.697183552209289e-06, + "loss": 0.1076, + "step": 936 + }, + { + "epoch": 0.6017983301220295, + "grad_norm": 0.7870484500378778, + "learning_rate": 9.695901376158845e-06, + "loss": 0.104, + "step": 937 + }, + { + "epoch": 0.6024405908798972, + "grad_norm": 1.156314182043536, + "learning_rate": 9.694616576494565e-06, + "loss": 0.1757, + "step": 938 + }, + { + "epoch": 0.603082851637765, + "grad_norm": 1.3114747816864243, + "learning_rate": 9.693329153934272e-06, + "loss": 0.2385, + "step": 939 + }, + { + "epoch": 0.6037251123956326, + "grad_norm": 0.9775773273958014, + "learning_rate": 9.692039109197251e-06, + "loss": 0.1299, + "step": 940 + }, + { + "epoch": 0.6043673731535003, + "grad_norm": 0.9260243116227516, + "learning_rate": 9.690746443004255e-06, + "loss": 0.1417, + "step": 941 + }, + { + "epoch": 0.605009633911368, + "grad_norm": 0.9757921700912374, + "learning_rate": 9.689451156077501e-06, + "loss": 0.1258, + "step": 942 + }, + { + "epoch": 0.6056518946692357, + "grad_norm": 0.8687882489822156, + "learning_rate": 9.68815324914067e-06, + "loss": 0.1135, + "step": 943 + }, + { + "epoch": 0.6062941554271034, + "grad_norm": 0.9404324173366121, + "learning_rate": 9.686852722918903e-06, + "loss": 0.1247, + "step": 944 + }, + { + "epoch": 0.6069364161849711, + "grad_norm": 0.7923358465689067, + "learning_rate": 9.68554957813881e-06, + "loss": 0.1032, + "step": 945 + }, + { + "epoch": 0.6075786769428388, + "grad_norm": 1.1115250780456123, + "learning_rate": 9.684243815528464e-06, + "loss": 0.2016, + "step": 946 + }, + { + "epoch": 0.6082209377007065, + "grad_norm": 0.8951700039822591, + "learning_rate": 9.682935435817394e-06, + "loss": 0.1024, + "step": 947 + }, + { + "epoch": 0.6088631984585742, + "grad_norm": 1.1677227716948888, + "learning_rate": 9.681624439736598e-06, + "loss": 0.1816, + "step": 948 + }, + { + "epoch": 0.6095054592164418, + "grad_norm": 1.0777006739792885, + "learning_rate": 9.680310828018535e-06, + "loss": 0.1704, + "step": 949 + }, + { + "epoch": 0.6101477199743096, + "grad_norm": 1.1035927484173884, + "learning_rate": 9.67899460139712e-06, + "loss": 0.1708, + "step": 950 + }, + { + "epoch": 0.6107899807321773, + "grad_norm": 1.1446044026862339, + "learning_rate": 9.677675760607735e-06, + "loss": 0.1722, + "step": 951 + }, + { + "epoch": 0.6114322414900449, + "grad_norm": 1.2319253978236746, + "learning_rate": 9.67635430638722e-06, + "loss": 0.1877, + "step": 952 + }, + { + "epoch": 0.6120745022479126, + "grad_norm": 1.0079545655368707, + "learning_rate": 9.675030239473873e-06, + "loss": 0.1814, + "step": 953 + }, + { + "epoch": 0.6127167630057804, + "grad_norm": 0.9621814994173092, + "learning_rate": 9.673703560607459e-06, + "loss": 0.1255, + "step": 954 + }, + { + "epoch": 0.613359023763648, + "grad_norm": 1.77545636182655, + "learning_rate": 9.672374270529192e-06, + "loss": 0.2768, + "step": 955 + }, + { + "epoch": 0.6140012845215157, + "grad_norm": 1.0322178724224103, + "learning_rate": 9.671042369981753e-06, + "loss": 0.1423, + "step": 956 + }, + { + "epoch": 0.6146435452793835, + "grad_norm": 0.8902711578045052, + "learning_rate": 9.669707859709279e-06, + "loss": 0.0942, + "step": 957 + }, + { + "epoch": 0.6152858060372511, + "grad_norm": 0.8624264676254073, + "learning_rate": 9.668370740457365e-06, + "loss": 0.0799, + "step": 958 + }, + { + "epoch": 0.6159280667951188, + "grad_norm": 1.1264212692167437, + "learning_rate": 9.667031012973062e-06, + "loss": 0.1381, + "step": 959 + }, + { + "epoch": 0.6165703275529865, + "grad_norm": 1.3767403145143575, + "learning_rate": 9.66568867800488e-06, + "loss": 0.2483, + "step": 960 + }, + { + "epoch": 0.6172125883108542, + "grad_norm": 1.1869158481742519, + "learning_rate": 9.664343736302786e-06, + "loss": 0.2153, + "step": 961 + }, + { + "epoch": 0.6178548490687219, + "grad_norm": 1.2977781070476957, + "learning_rate": 9.662996188618203e-06, + "loss": 0.2218, + "step": 962 + }, + { + "epoch": 0.6184971098265896, + "grad_norm": 1.0876501027797623, + "learning_rate": 9.661646035704009e-06, + "loss": 0.1801, + "step": 963 + }, + { + "epoch": 0.6191393705844573, + "grad_norm": 1.0264847705172773, + "learning_rate": 9.660293278314539e-06, + "loss": 0.144, + "step": 964 + }, + { + "epoch": 0.619781631342325, + "grad_norm": 1.0479699246147174, + "learning_rate": 9.658937917205583e-06, + "loss": 0.1975, + "step": 965 + }, + { + "epoch": 0.6204238921001927, + "grad_norm": 0.91272628667858, + "learning_rate": 9.657579953134384e-06, + "loss": 0.0997, + "step": 966 + }, + { + "epoch": 0.6210661528580603, + "grad_norm": 1.344013298973952, + "learning_rate": 9.656219386859642e-06, + "loss": 0.2899, + "step": 967 + }, + { + "epoch": 0.6217084136159281, + "grad_norm": 0.9358184227693378, + "learning_rate": 9.654856219141506e-06, + "loss": 0.1366, + "step": 968 + }, + { + "epoch": 0.6223506743737958, + "grad_norm": 1.1375747642002498, + "learning_rate": 9.653490450741587e-06, + "loss": 0.1493, + "step": 969 + }, + { + "epoch": 0.6229929351316634, + "grad_norm": 0.7928327789067786, + "learning_rate": 9.652122082422939e-06, + "loss": 0.0806, + "step": 970 + }, + { + "epoch": 0.6236351958895312, + "grad_norm": 0.9922142879704628, + "learning_rate": 9.650751114950077e-06, + "loss": 0.1081, + "step": 971 + }, + { + "epoch": 0.6242774566473989, + "grad_norm": 1.006388788964449, + "learning_rate": 9.649377549088962e-06, + "loss": 0.1289, + "step": 972 + }, + { + "epoch": 0.6249197174052665, + "grad_norm": 1.1409124565630917, + "learning_rate": 9.64800138560701e-06, + "loss": 0.1774, + "step": 973 + }, + { + "epoch": 0.6255619781631342, + "grad_norm": 3.5684411334249515, + "learning_rate": 9.646622625273088e-06, + "loss": 0.122, + "step": 974 + }, + { + "epoch": 0.626204238921002, + "grad_norm": 0.802634531380375, + "learning_rate": 9.645241268857515e-06, + "loss": 0.0795, + "step": 975 + }, + { + "epoch": 0.6268464996788696, + "grad_norm": 0.9540480028523273, + "learning_rate": 9.643857317132054e-06, + "loss": 0.1394, + "step": 976 + }, + { + "epoch": 0.6274887604367373, + "grad_norm": 0.8849055351687981, + "learning_rate": 9.64247077086993e-06, + "loss": 0.0911, + "step": 977 + }, + { + "epoch": 0.628131021194605, + "grad_norm": 0.9042268853871692, + "learning_rate": 9.641081630845803e-06, + "loss": 0.1162, + "step": 978 + }, + { + "epoch": 0.6287732819524727, + "grad_norm": 1.1494998872862188, + "learning_rate": 9.639689897835793e-06, + "loss": 0.1928, + "step": 979 + }, + { + "epoch": 0.6294155427103404, + "grad_norm": 1.091399157499931, + "learning_rate": 9.638295572617466e-06, + "loss": 0.1607, + "step": 980 + }, + { + "epoch": 0.630057803468208, + "grad_norm": 0.9888738959070698, + "learning_rate": 9.636898655969837e-06, + "loss": 0.0708, + "step": 981 + }, + { + "epoch": 0.6307000642260758, + "grad_norm": 1.082686866171021, + "learning_rate": 9.635499148673363e-06, + "loss": 0.1121, + "step": 982 + }, + { + "epoch": 0.6313423249839435, + "grad_norm": 1.1483937731268308, + "learning_rate": 9.634097051509956e-06, + "loss": 0.1618, + "step": 983 + }, + { + "epoch": 0.6319845857418112, + "grad_norm": 1.332412575125169, + "learning_rate": 9.63269236526297e-06, + "loss": 0.197, + "step": 984 + }, + { + "epoch": 0.6326268464996788, + "grad_norm": 1.0582503219970023, + "learning_rate": 9.631285090717208e-06, + "loss": 0.1669, + "step": 985 + }, + { + "epoch": 0.6332691072575466, + "grad_norm": 0.8786908289609685, + "learning_rate": 9.62987522865892e-06, + "loss": 0.1161, + "step": 986 + }, + { + "epoch": 0.6339113680154143, + "grad_norm": 1.3671576422210583, + "learning_rate": 9.6284627798758e-06, + "loss": 0.2036, + "step": 987 + }, + { + "epoch": 0.6345536287732819, + "grad_norm": 1.2086758018522596, + "learning_rate": 9.627047745156983e-06, + "loss": 0.2092, + "step": 988 + }, + { + "epoch": 0.6351958895311497, + "grad_norm": 1.3986477944747486, + "learning_rate": 9.625630125293058e-06, + "loss": 0.1817, + "step": 989 + }, + { + "epoch": 0.6358381502890174, + "grad_norm": 0.886777924348422, + "learning_rate": 9.62420992107605e-06, + "loss": 0.1087, + "step": 990 + }, + { + "epoch": 0.636480411046885, + "grad_norm": 1.2142923088525408, + "learning_rate": 9.622787133299433e-06, + "loss": 0.2225, + "step": 991 + }, + { + "epoch": 0.6371226718047527, + "grad_norm": 1.0256973328899646, + "learning_rate": 9.621361762758123e-06, + "loss": 0.0992, + "step": 992 + }, + { + "epoch": 0.6377649325626205, + "grad_norm": 1.1213859651508733, + "learning_rate": 9.619933810248476e-06, + "loss": 0.1942, + "step": 993 + }, + { + "epoch": 0.6384071933204881, + "grad_norm": 1.1251477197217472, + "learning_rate": 9.618503276568298e-06, + "loss": 0.1149, + "step": 994 + }, + { + "epoch": 0.6390494540783558, + "grad_norm": 0.9257754236366759, + "learning_rate": 9.617070162516828e-06, + "loss": 0.1013, + "step": 995 + }, + { + "epoch": 0.6396917148362236, + "grad_norm": 1.2240439049407885, + "learning_rate": 9.615634468894752e-06, + "loss": 0.1911, + "step": 996 + }, + { + "epoch": 0.6403339755940912, + "grad_norm": 0.813166702723631, + "learning_rate": 9.614196196504198e-06, + "loss": 0.0785, + "step": 997 + }, + { + "epoch": 0.6409762363519589, + "grad_norm": 1.097368305531494, + "learning_rate": 9.612755346148732e-06, + "loss": 0.1708, + "step": 998 + }, + { + "epoch": 0.6416184971098265, + "grad_norm": 0.9051889485819121, + "learning_rate": 9.61131191863336e-06, + "loss": 0.0941, + "step": 999 + }, + { + "epoch": 0.6422607578676943, + "grad_norm": 1.1590943878393984, + "learning_rate": 9.609865914764531e-06, + "loss": 0.2092, + "step": 1000 + }, + { + "epoch": 0.642903018625562, + "grad_norm": 0.9188821246350991, + "learning_rate": 9.608417335350131e-06, + "loss": 0.0974, + "step": 1001 + }, + { + "epoch": 0.6435452793834296, + "grad_norm": 0.7437662360831014, + "learning_rate": 9.606966181199487e-06, + "loss": 0.0511, + "step": 1002 + }, + { + "epoch": 0.6441875401412974, + "grad_norm": 1.119959487396456, + "learning_rate": 9.605512453123358e-06, + "loss": 0.1559, + "step": 1003 + }, + { + "epoch": 0.6448298008991651, + "grad_norm": 1.0575081939456799, + "learning_rate": 9.604056151933951e-06, + "loss": 0.1589, + "step": 1004 + }, + { + "epoch": 0.6454720616570327, + "grad_norm": 1.031532616356056, + "learning_rate": 9.602597278444906e-06, + "loss": 0.1351, + "step": 1005 + }, + { + "epoch": 0.6461143224149004, + "grad_norm": 0.8628524513034725, + "learning_rate": 9.601135833471299e-06, + "loss": 0.1123, + "step": 1006 + }, + { + "epoch": 0.6467565831727682, + "grad_norm": 0.9364331409070383, + "learning_rate": 9.59967181782964e-06, + "loss": 0.0943, + "step": 1007 + }, + { + "epoch": 0.6473988439306358, + "grad_norm": 1.0654681266414854, + "learning_rate": 9.598205232337882e-06, + "loss": 0.1367, + "step": 1008 + }, + { + "epoch": 0.6480411046885035, + "grad_norm": 0.9997186565371793, + "learning_rate": 9.596736077815412e-06, + "loss": 0.1418, + "step": 1009 + }, + { + "epoch": 0.6486833654463712, + "grad_norm": 0.9609644901111831, + "learning_rate": 9.595264355083046e-06, + "loss": 0.0974, + "step": 1010 + }, + { + "epoch": 0.649325626204239, + "grad_norm": 0.9954671279985368, + "learning_rate": 9.593790064963047e-06, + "loss": 0.1318, + "step": 1011 + }, + { + "epoch": 0.6499678869621066, + "grad_norm": 1.2434558821314847, + "learning_rate": 9.592313208279098e-06, + "loss": 0.2371, + "step": 1012 + }, + { + "epoch": 0.6506101477199743, + "grad_norm": 0.9423253174492263, + "learning_rate": 9.590833785856328e-06, + "loss": 0.1149, + "step": 1013 + }, + { + "epoch": 0.651252408477842, + "grad_norm": 0.9953903719605768, + "learning_rate": 9.589351798521293e-06, + "loss": 0.1082, + "step": 1014 + }, + { + "epoch": 0.6518946692357097, + "grad_norm": 0.9836298533982698, + "learning_rate": 9.587867247101984e-06, + "loss": 0.0872, + "step": 1015 + }, + { + "epoch": 0.6525369299935774, + "grad_norm": 1.1852611970242932, + "learning_rate": 9.586380132427823e-06, + "loss": 0.1287, + "step": 1016 + }, + { + "epoch": 0.653179190751445, + "grad_norm": 0.8051981887815424, + "learning_rate": 9.584890455329667e-06, + "loss": 0.0873, + "step": 1017 + }, + { + "epoch": 0.6538214515093128, + "grad_norm": 0.9821296823732244, + "learning_rate": 9.583398216639801e-06, + "loss": 0.1094, + "step": 1018 + }, + { + "epoch": 0.6544637122671805, + "grad_norm": 1.1317184308405006, + "learning_rate": 9.581903417191945e-06, + "loss": 0.1526, + "step": 1019 + }, + { + "epoch": 0.6551059730250481, + "grad_norm": 1.2426800576348953, + "learning_rate": 9.580406057821248e-06, + "loss": 0.1904, + "step": 1020 + }, + { + "epoch": 0.6557482337829159, + "grad_norm": 0.7910845474382636, + "learning_rate": 9.578906139364286e-06, + "loss": 0.0894, + "step": 1021 + }, + { + "epoch": 0.6563904945407836, + "grad_norm": 0.9190092700101974, + "learning_rate": 9.577403662659071e-06, + "loss": 0.0846, + "step": 1022 + }, + { + "epoch": 0.6570327552986512, + "grad_norm": 1.3246438586432672, + "learning_rate": 9.575898628545041e-06, + "loss": 0.2524, + "step": 1023 + }, + { + "epoch": 0.6576750160565189, + "grad_norm": 1.4227385150567993, + "learning_rate": 9.574391037863062e-06, + "loss": 0.2722, + "step": 1024 + }, + { + "epoch": 0.6583172768143867, + "grad_norm": 1.3354068212204362, + "learning_rate": 9.572880891455427e-06, + "loss": 0.2711, + "step": 1025 + }, + { + "epoch": 0.6589595375722543, + "grad_norm": 1.2135235850687478, + "learning_rate": 9.571368190165865e-06, + "loss": 0.196, + "step": 1026 + }, + { + "epoch": 0.659601798330122, + "grad_norm": 0.7130833174562866, + "learning_rate": 9.56985293483952e-06, + "loss": 0.0821, + "step": 1027 + }, + { + "epoch": 0.6602440590879898, + "grad_norm": 1.338755370962934, + "learning_rate": 9.568335126322975e-06, + "loss": 0.1347, + "step": 1028 + }, + { + "epoch": 0.6608863198458574, + "grad_norm": 1.135732556654184, + "learning_rate": 9.566814765464231e-06, + "loss": 0.1498, + "step": 1029 + }, + { + "epoch": 0.6615285806037251, + "grad_norm": 0.940625030508889, + "learning_rate": 9.56529185311272e-06, + "loss": 0.1135, + "step": 1030 + }, + { + "epoch": 0.6621708413615928, + "grad_norm": 1.2669037585325353, + "learning_rate": 9.563766390119296e-06, + "loss": 0.1855, + "step": 1031 + }, + { + "epoch": 0.6628131021194605, + "grad_norm": 1.0299904270562736, + "learning_rate": 9.56223837733624e-06, + "loss": 0.137, + "step": 1032 + }, + { + "epoch": 0.6634553628773282, + "grad_norm": 1.283219932642663, + "learning_rate": 9.560707815617256e-06, + "loss": 0.2131, + "step": 1033 + }, + { + "epoch": 0.6640976236351959, + "grad_norm": 1.3077209967372898, + "learning_rate": 9.559174705817477e-06, + "loss": 0.1985, + "step": 1034 + }, + { + "epoch": 0.6647398843930635, + "grad_norm": 0.6524570754120375, + "learning_rate": 9.557639048793453e-06, + "loss": 0.0573, + "step": 1035 + }, + { + "epoch": 0.6653821451509313, + "grad_norm": 0.9402783393599362, + "learning_rate": 9.556100845403161e-06, + "loss": 0.1384, + "step": 1036 + }, + { + "epoch": 0.666024405908799, + "grad_norm": 1.3107280371250611, + "learning_rate": 9.554560096505998e-06, + "loss": 0.2313, + "step": 1037 + }, + { + "epoch": 0.6666666666666666, + "grad_norm": 0.9277436045048655, + "learning_rate": 9.553016802962787e-06, + "loss": 0.0864, + "step": 1038 + }, + { + "epoch": 0.6673089274245344, + "grad_norm": 1.0395113943430794, + "learning_rate": 9.551470965635769e-06, + "loss": 0.154, + "step": 1039 + }, + { + "epoch": 0.6679511881824021, + "grad_norm": 1.0659262986875824, + "learning_rate": 9.54992258538861e-06, + "loss": 0.1676, + "step": 1040 + }, + { + "epoch": 0.6685934489402697, + "grad_norm": 1.2460719732683727, + "learning_rate": 9.548371663086392e-06, + "loss": 0.2057, + "step": 1041 + }, + { + "epoch": 0.6692357096981374, + "grad_norm": 1.2884029134619464, + "learning_rate": 9.546818199595624e-06, + "loss": 0.2132, + "step": 1042 + }, + { + "epoch": 0.6698779704560052, + "grad_norm": 0.9413836871529913, + "learning_rate": 9.545262195784226e-06, + "loss": 0.1178, + "step": 1043 + }, + { + "epoch": 0.6705202312138728, + "grad_norm": 1.8295789500015063, + "learning_rate": 9.543703652521543e-06, + "loss": 0.0999, + "step": 1044 + }, + { + "epoch": 0.6711624919717405, + "grad_norm": 1.0779248880945, + "learning_rate": 9.54214257067834e-06, + "loss": 0.161, + "step": 1045 + }, + { + "epoch": 0.6718047527296083, + "grad_norm": 0.8758804774255083, + "learning_rate": 9.540578951126796e-06, + "loss": 0.0897, + "step": 1046 + }, + { + "epoch": 0.6724470134874759, + "grad_norm": 1.084124401210672, + "learning_rate": 9.53901279474051e-06, + "loss": 0.1472, + "step": 1047 + }, + { + "epoch": 0.6730892742453436, + "grad_norm": 1.236937227145937, + "learning_rate": 9.537444102394498e-06, + "loss": 0.2219, + "step": 1048 + }, + { + "epoch": 0.6737315350032113, + "grad_norm": 0.8567293624886494, + "learning_rate": 9.535872874965193e-06, + "loss": 0.1074, + "step": 1049 + }, + { + "epoch": 0.674373795761079, + "grad_norm": 1.2501118166188878, + "learning_rate": 9.534299113330444e-06, + "loss": 0.2155, + "step": 1050 + }, + { + "epoch": 0.6750160565189467, + "grad_norm": 1.2004963800610435, + "learning_rate": 9.532722818369518e-06, + "loss": 0.2036, + "step": 1051 + }, + { + "epoch": 0.6756583172768144, + "grad_norm": 1.1242475678519175, + "learning_rate": 9.531143990963095e-06, + "loss": 0.1368, + "step": 1052 + }, + { + "epoch": 0.6763005780346821, + "grad_norm": 1.0122983130302812, + "learning_rate": 9.52956263199327e-06, + "loss": 0.174, + "step": 1053 + }, + { + "epoch": 0.6769428387925498, + "grad_norm": 0.9145305870209676, + "learning_rate": 9.527978742343552e-06, + "loss": 0.0825, + "step": 1054 + }, + { + "epoch": 0.6775850995504175, + "grad_norm": 1.1654421172149816, + "learning_rate": 9.526392322898865e-06, + "loss": 0.1668, + "step": 1055 + }, + { + "epoch": 0.6782273603082851, + "grad_norm": 1.0812849990942222, + "learning_rate": 9.524803374545549e-06, + "loss": 0.1623, + "step": 1056 + }, + { + "epoch": 0.6788696210661529, + "grad_norm": 0.9348283924688445, + "learning_rate": 9.52321189817135e-06, + "loss": 0.1126, + "step": 1057 + }, + { + "epoch": 0.6795118818240206, + "grad_norm": 5.638951924032475, + "learning_rate": 9.521617894665434e-06, + "loss": 0.27, + "step": 1058 + }, + { + "epoch": 0.6801541425818882, + "grad_norm": 1.1402789995858948, + "learning_rate": 9.520021364918373e-06, + "loss": 0.1425, + "step": 1059 + }, + { + "epoch": 0.680796403339756, + "grad_norm": 1.026196507835089, + "learning_rate": 9.518422309822154e-06, + "loss": 0.166, + "step": 1060 + }, + { + "epoch": 0.6814386640976237, + "grad_norm": 1.0913321862867231, + "learning_rate": 9.516820730270172e-06, + "loss": 0.1598, + "step": 1061 + }, + { + "epoch": 0.6820809248554913, + "grad_norm": 0.8404584325654257, + "learning_rate": 9.515216627157238e-06, + "loss": 0.1147, + "step": 1062 + }, + { + "epoch": 0.682723185613359, + "grad_norm": 0.899933938706567, + "learning_rate": 9.513610001379565e-06, + "loss": 0.0906, + "step": 1063 + }, + { + "epoch": 0.6833654463712268, + "grad_norm": 0.8819436359630632, + "learning_rate": 9.512000853834782e-06, + "loss": 0.1138, + "step": 1064 + }, + { + "epoch": 0.6840077071290944, + "grad_norm": 0.9822052076417924, + "learning_rate": 9.510389185421924e-06, + "loss": 0.139, + "step": 1065 + }, + { + "epoch": 0.6846499678869621, + "grad_norm": 1.340179788444082, + "learning_rate": 9.508774997041434e-06, + "loss": 0.2156, + "step": 1066 + }, + { + "epoch": 0.6852922286448297, + "grad_norm": 0.8343927554572682, + "learning_rate": 9.507158289595163e-06, + "loss": 0.0965, + "step": 1067 + }, + { + "epoch": 0.6859344894026975, + "grad_norm": 0.9626955276488728, + "learning_rate": 9.505539063986374e-06, + "loss": 0.0929, + "step": 1068 + }, + { + "epoch": 0.6865767501605652, + "grad_norm": 0.9094889143409799, + "learning_rate": 9.50391732111973e-06, + "loss": 0.1184, + "step": 1069 + }, + { + "epoch": 0.6872190109184328, + "grad_norm": 1.0300770468769698, + "learning_rate": 9.502293061901305e-06, + "loss": 0.1599, + "step": 1070 + }, + { + "epoch": 0.6878612716763006, + "grad_norm": 1.1722652383550893, + "learning_rate": 9.500666287238573e-06, + "loss": 0.2048, + "step": 1071 + }, + { + "epoch": 0.6885035324341683, + "grad_norm": 1.2197588282688034, + "learning_rate": 9.499036998040424e-06, + "loss": 0.2233, + "step": 1072 + }, + { + "epoch": 0.689145793192036, + "grad_norm": 1.0733784070324415, + "learning_rate": 9.497405195217144e-06, + "loss": 0.189, + "step": 1073 + }, + { + "epoch": 0.6897880539499036, + "grad_norm": 0.8402306500182662, + "learning_rate": 9.495770879680424e-06, + "loss": 0.087, + "step": 1074 + }, + { + "epoch": 0.6904303147077714, + "grad_norm": 0.8505401848535494, + "learning_rate": 9.494134052343364e-06, + "loss": 0.0921, + "step": 1075 + }, + { + "epoch": 0.691072575465639, + "grad_norm": 0.9071299143420757, + "learning_rate": 9.492494714120464e-06, + "loss": 0.0791, + "step": 1076 + }, + { + "epoch": 0.6917148362235067, + "grad_norm": 1.0767124198248752, + "learning_rate": 9.490852865927624e-06, + "loss": 0.1499, + "step": 1077 + }, + { + "epoch": 0.6923570969813745, + "grad_norm": 1.1925844306311673, + "learning_rate": 9.489208508682153e-06, + "loss": 0.1746, + "step": 1078 + }, + { + "epoch": 0.6929993577392421, + "grad_norm": 1.0898672130403697, + "learning_rate": 9.487561643302757e-06, + "loss": 0.1747, + "step": 1079 + }, + { + "epoch": 0.6936416184971098, + "grad_norm": 0.9537973263697137, + "learning_rate": 9.485912270709542e-06, + "loss": 0.1155, + "step": 1080 + }, + { + "epoch": 0.6942838792549775, + "grad_norm": 0.7544116680267429, + "learning_rate": 9.48426039182402e-06, + "loss": 0.0773, + "step": 1081 + }, + { + "epoch": 0.6949261400128453, + "grad_norm": 4.650751192818743, + "learning_rate": 9.482606007569101e-06, + "loss": 0.2573, + "step": 1082 + }, + { + "epoch": 0.6955684007707129, + "grad_norm": 1.0159200106147777, + "learning_rate": 9.480949118869092e-06, + "loss": 0.1538, + "step": 1083 + }, + { + "epoch": 0.6962106615285806, + "grad_norm": 1.4649501732212897, + "learning_rate": 9.479289726649703e-06, + "loss": 0.2513, + "step": 1084 + }, + { + "epoch": 0.6968529222864484, + "grad_norm": 1.2391928055029844, + "learning_rate": 9.477627831838043e-06, + "loss": 0.1646, + "step": 1085 + }, + { + "epoch": 0.697495183044316, + "grad_norm": 0.9516494209462489, + "learning_rate": 9.475963435362615e-06, + "loss": 0.1154, + "step": 1086 + }, + { + "epoch": 0.6981374438021837, + "grad_norm": 1.08800139728435, + "learning_rate": 9.47429653815332e-06, + "loss": 0.1433, + "step": 1087 + }, + { + "epoch": 0.6987797045600513, + "grad_norm": 0.8835456937321332, + "learning_rate": 9.472627141141464e-06, + "loss": 0.0865, + "step": 1088 + }, + { + "epoch": 0.6994219653179191, + "grad_norm": 1.2631140414965862, + "learning_rate": 9.470955245259742e-06, + "loss": 0.2409, + "step": 1089 + }, + { + "epoch": 0.7000642260757868, + "grad_norm": 0.9420738376373716, + "learning_rate": 9.469280851442245e-06, + "loss": 0.1396, + "step": 1090 + }, + { + "epoch": 0.7007064868336544, + "grad_norm": 1.3049239787871145, + "learning_rate": 9.467603960624464e-06, + "loss": 0.1504, + "step": 1091 + }, + { + "epoch": 0.7013487475915221, + "grad_norm": 1.0420175740074522, + "learning_rate": 9.465924573743279e-06, + "loss": 0.1489, + "step": 1092 + }, + { + "epoch": 0.7019910083493899, + "grad_norm": 1.1119524634159388, + "learning_rate": 9.464242691736973e-06, + "loss": 0.1856, + "step": 1093 + }, + { + "epoch": 0.7026332691072575, + "grad_norm": 0.9374210296295482, + "learning_rate": 9.462558315545216e-06, + "loss": 0.1064, + "step": 1094 + }, + { + "epoch": 0.7032755298651252, + "grad_norm": 1.2073126133238223, + "learning_rate": 9.460871446109075e-06, + "loss": 0.2161, + "step": 1095 + }, + { + "epoch": 0.703917790622993, + "grad_norm": 1.1069655904502935, + "learning_rate": 9.459182084371009e-06, + "loss": 0.2036, + "step": 1096 + }, + { + "epoch": 0.7045600513808606, + "grad_norm": 1.3650267260165574, + "learning_rate": 9.457490231274867e-06, + "loss": 0.1788, + "step": 1097 + }, + { + "epoch": 0.7052023121387283, + "grad_norm": 0.8593399138244499, + "learning_rate": 9.455795887765896e-06, + "loss": 0.1095, + "step": 1098 + }, + { + "epoch": 0.705844572896596, + "grad_norm": 1.1036398627310826, + "learning_rate": 9.454099054790728e-06, + "loss": 0.2071, + "step": 1099 + }, + { + "epoch": 0.7064868336544637, + "grad_norm": 0.9124357233219126, + "learning_rate": 9.452399733297388e-06, + "loss": 0.073, + "step": 1100 + }, + { + "epoch": 0.7071290944123314, + "grad_norm": 0.8518519918790073, + "learning_rate": 9.450697924235295e-06, + "loss": 0.0622, + "step": 1101 + }, + { + "epoch": 0.7077713551701991, + "grad_norm": 1.2249681134022647, + "learning_rate": 9.44899362855525e-06, + "loss": 0.2412, + "step": 1102 + }, + { + "epoch": 0.7084136159280668, + "grad_norm": 1.1153663591921912, + "learning_rate": 9.447286847209454e-06, + "loss": 0.2005, + "step": 1103 + }, + { + "epoch": 0.7090558766859345, + "grad_norm": 0.9666004872754912, + "learning_rate": 9.445577581151486e-06, + "loss": 0.1255, + "step": 1104 + }, + { + "epoch": 0.7096981374438022, + "grad_norm": 2.5615987882910716, + "learning_rate": 9.44386583133632e-06, + "loss": 0.1878, + "step": 1105 + }, + { + "epoch": 0.7103403982016698, + "grad_norm": 0.7593176696100421, + "learning_rate": 9.442151598720314e-06, + "loss": 0.0557, + "step": 1106 + }, + { + "epoch": 0.7109826589595376, + "grad_norm": 1.0253580892616723, + "learning_rate": 9.440434884261216e-06, + "loss": 0.1672, + "step": 1107 + }, + { + "epoch": 0.7116249197174053, + "grad_norm": 1.2562200270215007, + "learning_rate": 9.43871568891816e-06, + "loss": 0.1948, + "step": 1108 + }, + { + "epoch": 0.7122671804752729, + "grad_norm": 1.0093367851638595, + "learning_rate": 9.436994013651664e-06, + "loss": 0.1635, + "step": 1109 + }, + { + "epoch": 0.7129094412331407, + "grad_norm": 1.285696386126316, + "learning_rate": 9.435269859423634e-06, + "loss": 0.1923, + "step": 1110 + }, + { + "epoch": 0.7135517019910084, + "grad_norm": 0.7424366972966346, + "learning_rate": 9.43354322719736e-06, + "loss": 0.0634, + "step": 1111 + }, + { + "epoch": 0.714193962748876, + "grad_norm": 1.799276903370221, + "learning_rate": 9.431814117937517e-06, + "loss": 0.1764, + "step": 1112 + }, + { + "epoch": 0.7148362235067437, + "grad_norm": 1.5100441944972343, + "learning_rate": 9.43008253261016e-06, + "loss": 0.1012, + "step": 1113 + }, + { + "epoch": 0.7154784842646115, + "grad_norm": 1.2328566560819456, + "learning_rate": 9.428348472182737e-06, + "loss": 0.2301, + "step": 1114 + }, + { + "epoch": 0.7161207450224791, + "grad_norm": 0.9692457205984065, + "learning_rate": 9.426611937624067e-06, + "loss": 0.1524, + "step": 1115 + }, + { + "epoch": 0.7167630057803468, + "grad_norm": 1.0412194930886174, + "learning_rate": 9.424872929904359e-06, + "loss": 0.1474, + "step": 1116 + }, + { + "epoch": 0.7174052665382146, + "grad_norm": 0.9014702097306766, + "learning_rate": 9.423131449995202e-06, + "loss": 0.113, + "step": 1117 + }, + { + "epoch": 0.7180475272960822, + "grad_norm": 1.280781702711283, + "learning_rate": 9.421387498869565e-06, + "loss": 0.2518, + "step": 1118 + }, + { + "epoch": 0.7186897880539499, + "grad_norm": 1.0183333040780571, + "learning_rate": 9.4196410775018e-06, + "loss": 0.169, + "step": 1119 + }, + { + "epoch": 0.7193320488118176, + "grad_norm": 1.102061461694542, + "learning_rate": 9.417892186867636e-06, + "loss": 0.1676, + "step": 1120 + }, + { + "epoch": 0.7199743095696853, + "grad_norm": 1.2955317370518697, + "learning_rate": 9.416140827944185e-06, + "loss": 0.2666, + "step": 1121 + }, + { + "epoch": 0.720616570327553, + "grad_norm": 0.9618973784169228, + "learning_rate": 9.414387001709935e-06, + "loss": 0.1283, + "step": 1122 + }, + { + "epoch": 0.7212588310854207, + "grad_norm": 1.380661014290273, + "learning_rate": 9.412630709144754e-06, + "loss": 0.2781, + "step": 1123 + }, + { + "epoch": 0.7219010918432883, + "grad_norm": 1.0243270682278454, + "learning_rate": 9.41087195122989e-06, + "loss": 0.1484, + "step": 1124 + }, + { + "epoch": 0.7225433526011561, + "grad_norm": 0.7628211535294924, + "learning_rate": 9.409110728947964e-06, + "loss": 0.0691, + "step": 1125 + }, + { + "epoch": 0.7231856133590238, + "grad_norm": 1.5417022187516212, + "learning_rate": 9.407347043282974e-06, + "loss": 0.2141, + "step": 1126 + }, + { + "epoch": 0.7238278741168914, + "grad_norm": 0.788087051549884, + "learning_rate": 9.405580895220303e-06, + "loss": 0.0825, + "step": 1127 + }, + { + "epoch": 0.7244701348747592, + "grad_norm": 1.1912492643945813, + "learning_rate": 9.403812285746696e-06, + "loss": 0.1941, + "step": 1128 + }, + { + "epoch": 0.7251123956326269, + "grad_norm": 1.0529177814285908, + "learning_rate": 9.402041215850283e-06, + "loss": 0.091, + "step": 1129 + }, + { + "epoch": 0.7257546563904945, + "grad_norm": 0.8123232932657238, + "learning_rate": 9.400267686520568e-06, + "loss": 0.0777, + "step": 1130 + }, + { + "epoch": 0.7263969171483622, + "grad_norm": 1.311998777994997, + "learning_rate": 9.398491698748424e-06, + "loss": 0.2498, + "step": 1131 + }, + { + "epoch": 0.72703917790623, + "grad_norm": 0.9000697598192511, + "learning_rate": 9.3967132535261e-06, + "loss": 0.122, + "step": 1132 + }, + { + "epoch": 0.7276814386640976, + "grad_norm": 0.9934246558827003, + "learning_rate": 9.394932351847223e-06, + "loss": 0.1567, + "step": 1133 + }, + { + "epoch": 0.7283236994219653, + "grad_norm": 0.8602618283802776, + "learning_rate": 9.393148994706785e-06, + "loss": 0.068, + "step": 1134 + }, + { + "epoch": 0.7289659601798331, + "grad_norm": 1.064200468487511, + "learning_rate": 9.39136318310115e-06, + "loss": 0.1377, + "step": 1135 + }, + { + "epoch": 0.7296082209377007, + "grad_norm": 1.0117550864545968, + "learning_rate": 9.389574918028062e-06, + "loss": 0.1583, + "step": 1136 + }, + { + "epoch": 0.7302504816955684, + "grad_norm": 0.8502998883098006, + "learning_rate": 9.387784200486627e-06, + "loss": 0.0594, + "step": 1137 + }, + { + "epoch": 0.730892742453436, + "grad_norm": 1.1314837802892432, + "learning_rate": 9.385991031477323e-06, + "loss": 0.1706, + "step": 1138 + }, + { + "epoch": 0.7315350032113038, + "grad_norm": 1.1007658051071214, + "learning_rate": 9.384195412002001e-06, + "loss": 0.1824, + "step": 1139 + }, + { + "epoch": 0.7321772639691715, + "grad_norm": 0.8834771638305274, + "learning_rate": 9.382397343063879e-06, + "loss": 0.0951, + "step": 1140 + }, + { + "epoch": 0.7328195247270392, + "grad_norm": 1.0666253618179977, + "learning_rate": 9.380596825667541e-06, + "loss": 0.1733, + "step": 1141 + }, + { + "epoch": 0.7334617854849069, + "grad_norm": 1.0303245975415805, + "learning_rate": 9.378793860818942e-06, + "loss": 0.1562, + "step": 1142 + }, + { + "epoch": 0.7341040462427746, + "grad_norm": 1.01673866985881, + "learning_rate": 9.376988449525405e-06, + "loss": 0.1718, + "step": 1143 + }, + { + "epoch": 0.7347463070006423, + "grad_norm": 1.0193517580804194, + "learning_rate": 9.375180592795618e-06, + "loss": 0.1037, + "step": 1144 + }, + { + "epoch": 0.7353885677585099, + "grad_norm": 0.8764427832478728, + "learning_rate": 9.373370291639636e-06, + "loss": 0.0856, + "step": 1145 + }, + { + "epoch": 0.7360308285163777, + "grad_norm": 1.3729976487315505, + "learning_rate": 9.37155754706888e-06, + "loss": 0.2295, + "step": 1146 + }, + { + "epoch": 0.7366730892742454, + "grad_norm": 0.9647142079254503, + "learning_rate": 9.369742360096133e-06, + "loss": 0.0998, + "step": 1147 + }, + { + "epoch": 0.737315350032113, + "grad_norm": 0.9974384340508284, + "learning_rate": 9.367924731735549e-06, + "loss": 0.1354, + "step": 1148 + }, + { + "epoch": 0.7379576107899807, + "grad_norm": 1.0674753085419335, + "learning_rate": 9.36610466300264e-06, + "loss": 0.1346, + "step": 1149 + }, + { + "epoch": 0.7385998715478485, + "grad_norm": 0.7273705076280171, + "learning_rate": 9.364282154914285e-06, + "loss": 0.068, + "step": 1150 + }, + { + "epoch": 0.7392421323057161, + "grad_norm": 0.6512220129324607, + "learning_rate": 9.362457208488721e-06, + "loss": 0.0597, + "step": 1151 + }, + { + "epoch": 0.7398843930635838, + "grad_norm": 1.069026323797546, + "learning_rate": 9.360629824745558e-06, + "loss": 0.1343, + "step": 1152 + }, + { + "epoch": 0.7405266538214516, + "grad_norm": 1.0834711231092908, + "learning_rate": 9.358800004705753e-06, + "loss": 0.133, + "step": 1153 + }, + { + "epoch": 0.7411689145793192, + "grad_norm": 0.8145090872690106, + "learning_rate": 9.356967749391636e-06, + "loss": 0.0707, + "step": 1154 + }, + { + "epoch": 0.7418111753371869, + "grad_norm": 1.0770409850633695, + "learning_rate": 9.355133059826896e-06, + "loss": 0.178, + "step": 1155 + }, + { + "epoch": 0.7424534360950545, + "grad_norm": 0.6535914216829156, + "learning_rate": 9.353295937036571e-06, + "loss": 0.047, + "step": 1156 + }, + { + "epoch": 0.7430956968529223, + "grad_norm": 0.8528528199577822, + "learning_rate": 9.351456382047075e-06, + "loss": 0.1245, + "step": 1157 + }, + { + "epoch": 0.74373795761079, + "grad_norm": 1.3243267137784565, + "learning_rate": 9.349614395886168e-06, + "loss": 0.2077, + "step": 1158 + }, + { + "epoch": 0.7443802183686576, + "grad_norm": 1.2461379887073862, + "learning_rate": 9.347769979582977e-06, + "loss": 0.1561, + "step": 1159 + }, + { + "epoch": 0.7450224791265254, + "grad_norm": 1.089226356162001, + "learning_rate": 9.34592313416798e-06, + "loss": 0.1708, + "step": 1160 + }, + { + "epoch": 0.7456647398843931, + "grad_norm": 1.214445050269601, + "learning_rate": 9.344073860673016e-06, + "loss": 0.1877, + "step": 1161 + }, + { + "epoch": 0.7463070006422607, + "grad_norm": 1.0413591831542308, + "learning_rate": 9.342222160131276e-06, + "loss": 0.1398, + "step": 1162 + }, + { + "epoch": 0.7469492614001284, + "grad_norm": 1.1072982358858139, + "learning_rate": 9.340368033577316e-06, + "loss": 0.1996, + "step": 1163 + }, + { + "epoch": 0.7475915221579962, + "grad_norm": 1.0099717314439922, + "learning_rate": 9.338511482047039e-06, + "loss": 0.0948, + "step": 1164 + }, + { + "epoch": 0.7482337829158638, + "grad_norm": 0.922739088007312, + "learning_rate": 9.336652506577705e-06, + "loss": 0.1214, + "step": 1165 + }, + { + "epoch": 0.7488760436737315, + "grad_norm": 1.1404253161637756, + "learning_rate": 9.334791108207932e-06, + "loss": 0.2067, + "step": 1166 + }, + { + "epoch": 0.7495183044315993, + "grad_norm": 1.142986114743479, + "learning_rate": 9.332927287977687e-06, + "loss": 0.1747, + "step": 1167 + }, + { + "epoch": 0.7501605651894669, + "grad_norm": 0.8493432923131912, + "learning_rate": 9.33106104692829e-06, + "loss": 0.0992, + "step": 1168 + }, + { + "epoch": 0.7508028259473346, + "grad_norm": 0.9896625408197464, + "learning_rate": 9.329192386102419e-06, + "loss": 0.1487, + "step": 1169 + }, + { + "epoch": 0.7514450867052023, + "grad_norm": 1.2678909249733685, + "learning_rate": 9.327321306544097e-06, + "loss": 0.2078, + "step": 1170 + }, + { + "epoch": 0.75208734746307, + "grad_norm": 0.9269606011475018, + "learning_rate": 9.325447809298704e-06, + "loss": 0.1441, + "step": 1171 + }, + { + "epoch": 0.7527296082209377, + "grad_norm": 1.0539323821426094, + "learning_rate": 9.323571895412966e-06, + "loss": 0.1738, + "step": 1172 + }, + { + "epoch": 0.7533718689788054, + "grad_norm": 1.0006752901476197, + "learning_rate": 9.321693565934964e-06, + "loss": 0.1047, + "step": 1173 + }, + { + "epoch": 0.7540141297366731, + "grad_norm": 0.9276411377436402, + "learning_rate": 9.319812821914126e-06, + "loss": 0.1011, + "step": 1174 + }, + { + "epoch": 0.7546563904945408, + "grad_norm": 1.0323801393093919, + "learning_rate": 9.317929664401226e-06, + "loss": 0.135, + "step": 1175 + }, + { + "epoch": 0.7552986512524085, + "grad_norm": 1.261548565872805, + "learning_rate": 9.316044094448392e-06, + "loss": 0.2374, + "step": 1176 + }, + { + "epoch": 0.7559409120102761, + "grad_norm": 1.267439378584496, + "learning_rate": 9.314156113109101e-06, + "loss": 0.2032, + "step": 1177 + }, + { + "epoch": 0.7565831727681439, + "grad_norm": 1.0977099982651302, + "learning_rate": 9.312265721438168e-06, + "loss": 0.1645, + "step": 1178 + }, + { + "epoch": 0.7572254335260116, + "grad_norm": 1.0799443431690012, + "learning_rate": 9.310372920491761e-06, + "loss": 0.1509, + "step": 1179 + }, + { + "epoch": 0.7578676942838792, + "grad_norm": 1.0778306853399517, + "learning_rate": 9.308477711327398e-06, + "loss": 0.1546, + "step": 1180 + }, + { + "epoch": 0.7585099550417469, + "grad_norm": 1.2396715345746019, + "learning_rate": 9.306580095003932e-06, + "loss": 0.2236, + "step": 1181 + }, + { + "epoch": 0.7591522157996147, + "grad_norm": 0.8882978606116798, + "learning_rate": 9.30468007258157e-06, + "loss": 0.1465, + "step": 1182 + }, + { + "epoch": 0.7597944765574823, + "grad_norm": 1.2473240782712065, + "learning_rate": 9.30277764512186e-06, + "loss": 0.2342, + "step": 1183 + }, + { + "epoch": 0.76043673731535, + "grad_norm": 1.070579924525578, + "learning_rate": 9.300872813687694e-06, + "loss": 0.1352, + "step": 1184 + }, + { + "epoch": 0.7610789980732178, + "grad_norm": 0.8940926488192394, + "learning_rate": 9.298965579343305e-06, + "loss": 0.0874, + "step": 1185 + }, + { + "epoch": 0.7617212588310854, + "grad_norm": 1.0006599672890082, + "learning_rate": 9.297055943154273e-06, + "loss": 0.1534, + "step": 1186 + }, + { + "epoch": 0.7623635195889531, + "grad_norm": 1.015963349865826, + "learning_rate": 9.295143906187515e-06, + "loss": 0.1509, + "step": 1187 + }, + { + "epoch": 0.7630057803468208, + "grad_norm": 0.903957605200016, + "learning_rate": 9.293229469511293e-06, + "loss": 0.098, + "step": 1188 + }, + { + "epoch": 0.7636480411046885, + "grad_norm": 1.0621563833456404, + "learning_rate": 9.291312634195209e-06, + "loss": 0.1167, + "step": 1189 + }, + { + "epoch": 0.7642903018625562, + "grad_norm": 0.862090287122913, + "learning_rate": 9.289393401310203e-06, + "loss": 0.1134, + "step": 1190 + }, + { + "epoch": 0.7649325626204239, + "grad_norm": 0.9207921877614919, + "learning_rate": 9.287471771928556e-06, + "loss": 0.1202, + "step": 1191 + }, + { + "epoch": 0.7655748233782916, + "grad_norm": 0.8352489907311909, + "learning_rate": 9.285547747123888e-06, + "loss": 0.0681, + "step": 1192 + }, + { + "epoch": 0.7662170841361593, + "grad_norm": 1.178372497309497, + "learning_rate": 9.28362132797116e-06, + "loss": 0.1718, + "step": 1193 + }, + { + "epoch": 0.766859344894027, + "grad_norm": 0.9924716753751075, + "learning_rate": 9.281692515546665e-06, + "loss": 0.1294, + "step": 1194 + }, + { + "epoch": 0.7675016056518946, + "grad_norm": 1.0214496143603453, + "learning_rate": 9.279761310928037e-06, + "loss": 0.1745, + "step": 1195 + }, + { + "epoch": 0.7681438664097624, + "grad_norm": 0.8485960518326046, + "learning_rate": 9.277827715194246e-06, + "loss": 0.0937, + "step": 1196 + }, + { + "epoch": 0.7687861271676301, + "grad_norm": 1.0050761435144433, + "learning_rate": 9.275891729425595e-06, + "loss": 0.1192, + "step": 1197 + }, + { + "epoch": 0.7694283879254977, + "grad_norm": 1.1423061983598608, + "learning_rate": 9.273953354703728e-06, + "loss": 0.2437, + "step": 1198 + }, + { + "epoch": 0.7700706486833655, + "grad_norm": 1.0918293870143372, + "learning_rate": 9.27201259211162e-06, + "loss": 0.1791, + "step": 1199 + }, + { + "epoch": 0.7707129094412332, + "grad_norm": 1.5402955433265606, + "learning_rate": 9.270069442733583e-06, + "loss": 0.3318, + "step": 1200 + }, + { + "epoch": 0.7713551701991008, + "grad_norm": 1.3075433434263541, + "learning_rate": 9.268123907655256e-06, + "loss": 0.2321, + "step": 1201 + }, + { + "epoch": 0.7719974309569685, + "grad_norm": 0.999947753575331, + "learning_rate": 9.266175987963617e-06, + "loss": 0.1581, + "step": 1202 + }, + { + "epoch": 0.7726396917148363, + "grad_norm": 1.0384410670130042, + "learning_rate": 9.264225684746974e-06, + "loss": 0.1309, + "step": 1203 + }, + { + "epoch": 0.7732819524727039, + "grad_norm": 0.8617137193455207, + "learning_rate": 9.262272999094968e-06, + "loss": 0.1382, + "step": 1204 + }, + { + "epoch": 0.7739242132305716, + "grad_norm": 1.1483798024313954, + "learning_rate": 9.260317932098569e-06, + "loss": 0.254, + "step": 1205 + }, + { + "epoch": 0.7745664739884393, + "grad_norm": 0.5827058931940031, + "learning_rate": 9.25836048485008e-06, + "loss": 0.0303, + "step": 1206 + }, + { + "epoch": 0.775208734746307, + "grad_norm": 0.8155027100011236, + "learning_rate": 9.25640065844313e-06, + "loss": 0.0541, + "step": 1207 + }, + { + "epoch": 0.7758509955041747, + "grad_norm": 1.0575062237314468, + "learning_rate": 9.254438453972683e-06, + "loss": 0.1718, + "step": 1208 + }, + { + "epoch": 0.7764932562620424, + "grad_norm": 1.0610208630550635, + "learning_rate": 9.252473872535027e-06, + "loss": 0.1534, + "step": 1209 + }, + { + "epoch": 0.7771355170199101, + "grad_norm": 0.8730437035471011, + "learning_rate": 9.250506915227779e-06, + "loss": 0.1104, + "step": 1210 + }, + { + "epoch": 0.7777777777777778, + "grad_norm": 1.259264774372249, + "learning_rate": 9.248537583149884e-06, + "loss": 0.1996, + "step": 1211 + }, + { + "epoch": 0.7784200385356455, + "grad_norm": 1.1265469672359882, + "learning_rate": 9.246565877401614e-06, + "loss": 0.1941, + "step": 1212 + }, + { + "epoch": 0.7790622992935131, + "grad_norm": 1.0750030989337274, + "learning_rate": 9.244591799084567e-06, + "loss": 0.133, + "step": 1213 + }, + { + "epoch": 0.7797045600513809, + "grad_norm": 0.9987082786733124, + "learning_rate": 9.242615349301664e-06, + "loss": 0.1264, + "step": 1214 + }, + { + "epoch": 0.7803468208092486, + "grad_norm": 0.8373860218823954, + "learning_rate": 9.240636529157158e-06, + "loss": 0.081, + "step": 1215 + }, + { + "epoch": 0.7809890815671162, + "grad_norm": 1.0823779005115266, + "learning_rate": 9.238655339756616e-06, + "loss": 0.1766, + "step": 1216 + }, + { + "epoch": 0.781631342324984, + "grad_norm": 0.9970361139454955, + "learning_rate": 9.236671782206939e-06, + "loss": 0.1398, + "step": 1217 + }, + { + "epoch": 0.7822736030828517, + "grad_norm": 1.3372829417889462, + "learning_rate": 9.234685857616343e-06, + "loss": 0.222, + "step": 1218 + }, + { + "epoch": 0.7829158638407193, + "grad_norm": 0.7128549667590867, + "learning_rate": 9.232697567094371e-06, + "loss": 0.0715, + "step": 1219 + }, + { + "epoch": 0.783558124598587, + "grad_norm": 0.8741382935264439, + "learning_rate": 9.230706911751888e-06, + "loss": 0.1036, + "step": 1220 + }, + { + "epoch": 0.7842003853564548, + "grad_norm": 0.9824802159252779, + "learning_rate": 9.228713892701078e-06, + "loss": 0.148, + "step": 1221 + }, + { + "epoch": 0.7848426461143224, + "grad_norm": 0.9402918866355995, + "learning_rate": 9.226718511055447e-06, + "loss": 0.1158, + "step": 1222 + }, + { + "epoch": 0.7854849068721901, + "grad_norm": 0.7101807936528077, + "learning_rate": 9.224720767929819e-06, + "loss": 0.0501, + "step": 1223 + }, + { + "epoch": 0.7861271676300579, + "grad_norm": 1.2768524564781467, + "learning_rate": 9.22272066444034e-06, + "loss": 0.2038, + "step": 1224 + }, + { + "epoch": 0.7867694283879255, + "grad_norm": 1.3774635870460448, + "learning_rate": 9.220718201704475e-06, + "loss": 0.2589, + "step": 1225 + }, + { + "epoch": 0.7874116891457932, + "grad_norm": 1.4956447690798778, + "learning_rate": 9.218713380841003e-06, + "loss": 0.0934, + "step": 1226 + }, + { + "epoch": 0.7880539499036608, + "grad_norm": 1.17459337563262, + "learning_rate": 9.216706202970027e-06, + "loss": 0.2154, + "step": 1227 + }, + { + "epoch": 0.7886962106615286, + "grad_norm": 0.9467555565662488, + "learning_rate": 9.21469666921296e-06, + "loss": 0.1107, + "step": 1228 + }, + { + "epoch": 0.7893384714193963, + "grad_norm": 1.085492982071229, + "learning_rate": 9.212684780692534e-06, + "loss": 0.1473, + "step": 1229 + }, + { + "epoch": 0.789980732177264, + "grad_norm": 1.0114987930198804, + "learning_rate": 9.210670538532798e-06, + "loss": 0.1075, + "step": 1230 + }, + { + "epoch": 0.7906229929351317, + "grad_norm": 1.3749761533779445, + "learning_rate": 9.208653943859117e-06, + "loss": 0.334, + "step": 1231 + }, + { + "epoch": 0.7912652536929994, + "grad_norm": 0.8846055195197449, + "learning_rate": 9.206634997798165e-06, + "loss": 0.098, + "step": 1232 + }, + { + "epoch": 0.791907514450867, + "grad_norm": 1.145956230568199, + "learning_rate": 9.204613701477935e-06, + "loss": 0.1906, + "step": 1233 + }, + { + "epoch": 0.7925497752087347, + "grad_norm": 1.2337682019369627, + "learning_rate": 9.202590056027733e-06, + "loss": 0.2025, + "step": 1234 + }, + { + "epoch": 0.7931920359666025, + "grad_norm": 1.1725146090637935, + "learning_rate": 9.200564062578172e-06, + "loss": 0.1951, + "step": 1235 + }, + { + "epoch": 0.7938342967244701, + "grad_norm": 0.929768670356068, + "learning_rate": 9.198535722261182e-06, + "loss": 0.1388, + "step": 1236 + }, + { + "epoch": 0.7944765574823378, + "grad_norm": 1.230614843602799, + "learning_rate": 9.196505036210002e-06, + "loss": 0.2157, + "step": 1237 + }, + { + "epoch": 0.7951188182402055, + "grad_norm": 1.1247216132626723, + "learning_rate": 9.194472005559184e-06, + "loss": 0.1954, + "step": 1238 + }, + { + "epoch": 0.7957610789980732, + "grad_norm": 0.7070227650514903, + "learning_rate": 9.192436631444587e-06, + "loss": 0.0604, + "step": 1239 + }, + { + "epoch": 0.7964033397559409, + "grad_norm": 1.0793909628046772, + "learning_rate": 9.19039891500338e-06, + "loss": 0.1958, + "step": 1240 + }, + { + "epoch": 0.7970456005138086, + "grad_norm": 0.8962315455153058, + "learning_rate": 9.188358857374042e-06, + "loss": 0.1353, + "step": 1241 + }, + { + "epoch": 0.7976878612716763, + "grad_norm": 0.9565626820746468, + "learning_rate": 9.186316459696359e-06, + "loss": 0.1245, + "step": 1242 + }, + { + "epoch": 0.798330122029544, + "grad_norm": 1.079775253302687, + "learning_rate": 9.184271723111424e-06, + "loss": 0.1459, + "step": 1243 + }, + { + "epoch": 0.7989723827874117, + "grad_norm": 1.3555988911925048, + "learning_rate": 9.182224648761637e-06, + "loss": 0.2875, + "step": 1244 + }, + { + "epoch": 0.7996146435452793, + "grad_norm": 0.9651371142902591, + "learning_rate": 9.180175237790707e-06, + "loss": 0.1141, + "step": 1245 + }, + { + "epoch": 0.8002569043031471, + "grad_norm": 0.6282130646518761, + "learning_rate": 9.178123491343642e-06, + "loss": 0.0635, + "step": 1246 + }, + { + "epoch": 0.8008991650610148, + "grad_norm": 1.1762789134237592, + "learning_rate": 9.176069410566763e-06, + "loss": 0.2163, + "step": 1247 + }, + { + "epoch": 0.8015414258188824, + "grad_norm": 1.014148479910808, + "learning_rate": 9.174012996607687e-06, + "loss": 0.1284, + "step": 1248 + }, + { + "epoch": 0.8021836865767502, + "grad_norm": 1.019369412776965, + "learning_rate": 9.17195425061534e-06, + "loss": 0.1541, + "step": 1249 + }, + { + "epoch": 0.8028259473346179, + "grad_norm": 1.061869126570766, + "learning_rate": 9.16989317373995e-06, + "loss": 0.1801, + "step": 1250 + }, + { + "epoch": 0.8034682080924855, + "grad_norm": 1.0421510731000936, + "learning_rate": 9.167829767133047e-06, + "loss": 0.2031, + "step": 1251 + }, + { + "epoch": 0.8041104688503532, + "grad_norm": 1.0502704201962996, + "learning_rate": 9.165764031947461e-06, + "loss": 0.141, + "step": 1252 + }, + { + "epoch": 0.804752729608221, + "grad_norm": 1.0327450397377718, + "learning_rate": 9.163695969337325e-06, + "loss": 0.1261, + "step": 1253 + }, + { + "epoch": 0.8053949903660886, + "grad_norm": 1.4224032067220604, + "learning_rate": 9.161625580458073e-06, + "loss": 0.1727, + "step": 1254 + }, + { + "epoch": 0.8060372511239563, + "grad_norm": 0.8116406484415809, + "learning_rate": 9.159552866466434e-06, + "loss": 0.0844, + "step": 1255 + }, + { + "epoch": 0.8066795118818241, + "grad_norm": 1.0418899750226482, + "learning_rate": 9.15747782852044e-06, + "loss": 0.1497, + "step": 1256 + }, + { + "epoch": 0.8073217726396917, + "grad_norm": 1.277828988075714, + "learning_rate": 9.155400467779424e-06, + "loss": 0.1683, + "step": 1257 + }, + { + "epoch": 0.8079640333975594, + "grad_norm": 1.2123812652677632, + "learning_rate": 9.153320785404011e-06, + "loss": 0.226, + "step": 1258 + }, + { + "epoch": 0.8086062941554271, + "grad_norm": 1.2199929847008792, + "learning_rate": 9.151238782556127e-06, + "loss": 0.1911, + "step": 1259 + }, + { + "epoch": 0.8092485549132948, + "grad_norm": 0.9537092011433437, + "learning_rate": 9.149154460398993e-06, + "loss": 0.1234, + "step": 1260 + }, + { + "epoch": 0.8098908156711625, + "grad_norm": 1.0105981663834387, + "learning_rate": 9.147067820097122e-06, + "loss": 0.0986, + "step": 1261 + }, + { + "epoch": 0.8105330764290302, + "grad_norm": 1.0880914090450433, + "learning_rate": 9.144978862816332e-06, + "loss": 0.1724, + "step": 1262 + }, + { + "epoch": 0.8111753371868978, + "grad_norm": 1.1579329104382463, + "learning_rate": 9.142887589723724e-06, + "loss": 0.2577, + "step": 1263 + }, + { + "epoch": 0.8118175979447656, + "grad_norm": 0.8591359227604711, + "learning_rate": 9.140794001987703e-06, + "loss": 0.0762, + "step": 1264 + }, + { + "epoch": 0.8124598587026333, + "grad_norm": 4.668532816009121, + "learning_rate": 9.138698100777961e-06, + "loss": 0.104, + "step": 1265 + }, + { + "epoch": 0.8131021194605009, + "grad_norm": 1.2447955590489104, + "learning_rate": 9.136599887265483e-06, + "loss": 0.1915, + "step": 1266 + }, + { + "epoch": 0.8137443802183687, + "grad_norm": 1.146554723174779, + "learning_rate": 9.134499362622546e-06, + "loss": 0.2096, + "step": 1267 + }, + { + "epoch": 0.8143866409762364, + "grad_norm": 1.1413126292156996, + "learning_rate": 9.132396528022724e-06, + "loss": 0.2197, + "step": 1268 + }, + { + "epoch": 0.815028901734104, + "grad_norm": 1.0163107621323442, + "learning_rate": 9.130291384640873e-06, + "loss": 0.1574, + "step": 1269 + }, + { + "epoch": 0.8156711624919717, + "grad_norm": 1.0045227124839218, + "learning_rate": 9.128183933653142e-06, + "loss": 0.1709, + "step": 1270 + }, + { + "epoch": 0.8163134232498395, + "grad_norm": 0.6599679720892464, + "learning_rate": 9.126074176236975e-06, + "loss": 0.0552, + "step": 1271 + }, + { + "epoch": 0.8169556840077071, + "grad_norm": 0.963822068168109, + "learning_rate": 9.123962113571095e-06, + "loss": 0.1131, + "step": 1272 + }, + { + "epoch": 0.8175979447655748, + "grad_norm": 1.2431541996042454, + "learning_rate": 9.121847746835517e-06, + "loss": 0.2345, + "step": 1273 + }, + { + "epoch": 0.8182402055234426, + "grad_norm": 0.9857226548906117, + "learning_rate": 9.119731077211549e-06, + "loss": 0.1086, + "step": 1274 + }, + { + "epoch": 0.8188824662813102, + "grad_norm": 1.699538088260216, + "learning_rate": 9.117612105881775e-06, + "loss": 0.1724, + "step": 1275 + }, + { + "epoch": 0.8195247270391779, + "grad_norm": 1.0150655484343105, + "learning_rate": 9.115490834030074e-06, + "loss": 0.1649, + "step": 1276 + }, + { + "epoch": 0.8201669877970456, + "grad_norm": 1.1694839718542325, + "learning_rate": 9.113367262841606e-06, + "loss": 0.1706, + "step": 1277 + }, + { + "epoch": 0.8208092485549133, + "grad_norm": 1.1216797313074818, + "learning_rate": 9.111241393502814e-06, + "loss": 0.1674, + "step": 1278 + }, + { + "epoch": 0.821451509312781, + "grad_norm": 0.9714223409032657, + "learning_rate": 9.10911322720143e-06, + "loss": 0.1473, + "step": 1279 + }, + { + "epoch": 0.8220937700706487, + "grad_norm": 0.9777454338267505, + "learning_rate": 9.106982765126469e-06, + "loss": 0.138, + "step": 1280 + }, + { + "epoch": 0.8227360308285164, + "grad_norm": 1.1663264774459041, + "learning_rate": 9.104850008468222e-06, + "loss": 0.1822, + "step": 1281 + }, + { + "epoch": 0.8233782915863841, + "grad_norm": 0.7177652857060814, + "learning_rate": 9.102714958418266e-06, + "loss": 0.0641, + "step": 1282 + }, + { + "epoch": 0.8240205523442518, + "grad_norm": 1.0389597424515906, + "learning_rate": 9.100577616169463e-06, + "loss": 0.1499, + "step": 1283 + }, + { + "epoch": 0.8246628131021194, + "grad_norm": 1.5815002568339922, + "learning_rate": 9.098437982915953e-06, + "loss": 0.3782, + "step": 1284 + }, + { + "epoch": 0.8253050738599872, + "grad_norm": 1.0311370369678012, + "learning_rate": 9.096296059853151e-06, + "loss": 0.1301, + "step": 1285 + }, + { + "epoch": 0.8259473346178549, + "grad_norm": 0.8474660896710217, + "learning_rate": 9.094151848177759e-06, + "loss": 0.0904, + "step": 1286 + }, + { + "epoch": 0.8265895953757225, + "grad_norm": 0.822360249129622, + "learning_rate": 9.092005349087754e-06, + "loss": 0.0663, + "step": 1287 + }, + { + "epoch": 0.8272318561335903, + "grad_norm": 0.9554733603829817, + "learning_rate": 9.089856563782391e-06, + "loss": 0.1439, + "step": 1288 + }, + { + "epoch": 0.827874116891458, + "grad_norm": 1.1415784457525275, + "learning_rate": 9.0877054934622e-06, + "loss": 0.1834, + "step": 1289 + }, + { + "epoch": 0.8285163776493256, + "grad_norm": 0.9575884743630886, + "learning_rate": 9.085552139328995e-06, + "loss": 0.1119, + "step": 1290 + }, + { + "epoch": 0.8291586384071933, + "grad_norm": 1.1092182536621367, + "learning_rate": 9.083396502585858e-06, + "loss": 0.1658, + "step": 1291 + }, + { + "epoch": 0.8298008991650611, + "grad_norm": 1.0580832900903085, + "learning_rate": 9.081238584437151e-06, + "loss": 0.169, + "step": 1292 + }, + { + "epoch": 0.8304431599229287, + "grad_norm": 1.3799887381585936, + "learning_rate": 9.079078386088508e-06, + "loss": 0.2191, + "step": 1293 + }, + { + "epoch": 0.8310854206807964, + "grad_norm": 0.540702475139385, + "learning_rate": 9.07691590874684e-06, + "loss": 0.0413, + "step": 1294 + }, + { + "epoch": 0.831727681438664, + "grad_norm": 0.7410465148853834, + "learning_rate": 9.074751153620327e-06, + "loss": 0.0612, + "step": 1295 + }, + { + "epoch": 0.8323699421965318, + "grad_norm": 1.137386061759738, + "learning_rate": 9.072584121918426e-06, + "loss": 0.1983, + "step": 1296 + }, + { + "epoch": 0.8330122029543995, + "grad_norm": 0.8004099976833032, + "learning_rate": 9.070414814851862e-06, + "loss": 0.0903, + "step": 1297 + }, + { + "epoch": 0.8336544637122671, + "grad_norm": 1.0153106009307113, + "learning_rate": 9.068243233632634e-06, + "loss": 0.1429, + "step": 1298 + }, + { + "epoch": 0.8342967244701349, + "grad_norm": 1.1488966500213853, + "learning_rate": 9.066069379474014e-06, + "loss": 0.2558, + "step": 1299 + }, + { + "epoch": 0.8349389852280026, + "grad_norm": 1.1996218645295322, + "learning_rate": 9.063893253590535e-06, + "loss": 0.2291, + "step": 1300 + }, + { + "epoch": 0.8355812459858702, + "grad_norm": 0.9530199571160267, + "learning_rate": 9.06171485719801e-06, + "loss": 0.1515, + "step": 1301 + }, + { + "epoch": 0.8362235067437379, + "grad_norm": 1.0354631408957948, + "learning_rate": 9.059534191513512e-06, + "loss": 0.1029, + "step": 1302 + }, + { + "epoch": 0.8368657675016057, + "grad_norm": 0.9252110144715718, + "learning_rate": 9.057351257755387e-06, + "loss": 0.1357, + "step": 1303 + }, + { + "epoch": 0.8375080282594733, + "grad_norm": 0.9446821536683686, + "learning_rate": 9.055166057143246e-06, + "loss": 0.1005, + "step": 1304 + }, + { + "epoch": 0.838150289017341, + "grad_norm": 0.9190652644984173, + "learning_rate": 9.052978590897964e-06, + "loss": 0.1329, + "step": 1305 + }, + { + "epoch": 0.8387925497752088, + "grad_norm": 0.7390644321461484, + "learning_rate": 9.050788860241691e-06, + "loss": 0.1011, + "step": 1306 + }, + { + "epoch": 0.8394348105330764, + "grad_norm": 0.8189896647112787, + "learning_rate": 9.048596866397832e-06, + "loss": 0.1032, + "step": 1307 + }, + { + "epoch": 0.8400770712909441, + "grad_norm": 1.4373790445391856, + "learning_rate": 9.046402610591062e-06, + "loss": 0.1622, + "step": 1308 + }, + { + "epoch": 0.8407193320488118, + "grad_norm": 0.8911391631403239, + "learning_rate": 9.044206094047314e-06, + "loss": 0.0861, + "step": 1309 + }, + { + "epoch": 0.8413615928066795, + "grad_norm": 1.0409064616673682, + "learning_rate": 9.042007317993795e-06, + "loss": 0.1367, + "step": 1310 + }, + { + "epoch": 0.8420038535645472, + "grad_norm": 0.9856482686418405, + "learning_rate": 9.039806283658961e-06, + "loss": 0.1305, + "step": 1311 + }, + { + "epoch": 0.8426461143224149, + "grad_norm": 0.925191955159981, + "learning_rate": 9.037602992272542e-06, + "loss": 0.1381, + "step": 1312 + }, + { + "epoch": 0.8432883750802826, + "grad_norm": 1.0209814813339682, + "learning_rate": 9.035397445065519e-06, + "loss": 0.1381, + "step": 1313 + }, + { + "epoch": 0.8439306358381503, + "grad_norm": 0.9773143358166686, + "learning_rate": 9.033189643270139e-06, + "loss": 0.1086, + "step": 1314 + }, + { + "epoch": 0.844572896596018, + "grad_norm": 0.863821928043678, + "learning_rate": 9.030979588119908e-06, + "loss": 0.0727, + "step": 1315 + }, + { + "epoch": 0.8452151573538856, + "grad_norm": 7.610467732286724, + "learning_rate": 9.028767280849592e-06, + "loss": 0.1036, + "step": 1316 + }, + { + "epoch": 0.8458574181117534, + "grad_norm": 1.2132644177405671, + "learning_rate": 9.026552722695209e-06, + "loss": 0.2698, + "step": 1317 + }, + { + "epoch": 0.8464996788696211, + "grad_norm": 1.2301511375735814, + "learning_rate": 9.024335914894042e-06, + "loss": 0.1915, + "step": 1318 + }, + { + "epoch": 0.8471419396274887, + "grad_norm": 0.9444323725185091, + "learning_rate": 9.022116858684627e-06, + "loss": 0.1561, + "step": 1319 + }, + { + "epoch": 0.8477842003853564, + "grad_norm": 1.149429427347774, + "learning_rate": 9.01989555530676e-06, + "loss": 0.1889, + "step": 1320 + }, + { + "epoch": 0.8484264611432242, + "grad_norm": 1.0054826296960193, + "learning_rate": 9.017672006001486e-06, + "loss": 0.1305, + "step": 1321 + }, + { + "epoch": 0.8490687219010918, + "grad_norm": 1.1060638228396733, + "learning_rate": 9.01544621201111e-06, + "loss": 0.2521, + "step": 1322 + }, + { + "epoch": 0.8497109826589595, + "grad_norm": 0.9518183316485344, + "learning_rate": 9.013218174579189e-06, + "loss": 0.1163, + "step": 1323 + }, + { + "epoch": 0.8503532434168273, + "grad_norm": 2.48905222639571, + "learning_rate": 9.010987894950536e-06, + "loss": 0.2249, + "step": 1324 + }, + { + "epoch": 0.8509955041746949, + "grad_norm": 0.7931767480699543, + "learning_rate": 9.008755374371211e-06, + "loss": 0.0865, + "step": 1325 + }, + { + "epoch": 0.8516377649325626, + "grad_norm": 1.1735986701128842, + "learning_rate": 9.006520614088535e-06, + "loss": 0.1777, + "step": 1326 + }, + { + "epoch": 0.8522800256904303, + "grad_norm": 0.8117532413727453, + "learning_rate": 9.004283615351072e-06, + "loss": 0.0455, + "step": 1327 + }, + { + "epoch": 0.852922286448298, + "grad_norm": 0.7178032962300069, + "learning_rate": 9.002044379408639e-06, + "loss": 0.0512, + "step": 1328 + }, + { + "epoch": 0.8535645472061657, + "grad_norm": 1.007007926773158, + "learning_rate": 8.999802907512304e-06, + "loss": 0.1346, + "step": 1329 + }, + { + "epoch": 0.8542068079640334, + "grad_norm": 0.8554750679973299, + "learning_rate": 8.997559200914385e-06, + "loss": 0.0887, + "step": 1330 + }, + { + "epoch": 0.8548490687219011, + "grad_norm": 1.286781093155533, + "learning_rate": 8.995313260868447e-06, + "loss": 0.201, + "step": 1331 + }, + { + "epoch": 0.8554913294797688, + "grad_norm": 0.9869751414064104, + "learning_rate": 8.993065088629304e-06, + "loss": 0.1229, + "step": 1332 + }, + { + "epoch": 0.8561335902376365, + "grad_norm": 0.8018481071940321, + "learning_rate": 8.990814685453015e-06, + "loss": 0.0786, + "step": 1333 + }, + { + "epoch": 0.8567758509955041, + "grad_norm": 0.9786298005040996, + "learning_rate": 8.988562052596889e-06, + "loss": 0.0662, + "step": 1334 + }, + { + "epoch": 0.8574181117533719, + "grad_norm": 0.9188558119047617, + "learning_rate": 8.986307191319477e-06, + "loss": 0.1202, + "step": 1335 + }, + { + "epoch": 0.8580603725112396, + "grad_norm": 0.9139510743422228, + "learning_rate": 8.984050102880575e-06, + "loss": 0.1136, + "step": 1336 + }, + { + "epoch": 0.8587026332691072, + "grad_norm": 1.1369295499100593, + "learning_rate": 8.981790788541226e-06, + "loss": 0.2323, + "step": 1337 + }, + { + "epoch": 0.859344894026975, + "grad_norm": 0.9854504215139602, + "learning_rate": 8.979529249563718e-06, + "loss": 0.1519, + "step": 1338 + }, + { + "epoch": 0.8599871547848427, + "grad_norm": 0.9750944451873974, + "learning_rate": 8.977265487211576e-06, + "loss": 0.0928, + "step": 1339 + }, + { + "epoch": 0.8606294155427103, + "grad_norm": 1.0020299826676005, + "learning_rate": 8.974999502749576e-06, + "loss": 0.1563, + "step": 1340 + }, + { + "epoch": 0.861271676300578, + "grad_norm": 0.8036649716024065, + "learning_rate": 8.972731297443722e-06, + "loss": 0.1117, + "step": 1341 + }, + { + "epoch": 0.8619139370584458, + "grad_norm": 1.0148681578187133, + "learning_rate": 8.970460872561272e-06, + "loss": 0.1494, + "step": 1342 + }, + { + "epoch": 0.8625561978163134, + "grad_norm": 1.0539129594735868, + "learning_rate": 8.96818822937072e-06, + "loss": 0.1215, + "step": 1343 + }, + { + "epoch": 0.8631984585741811, + "grad_norm": 0.9492362791937698, + "learning_rate": 8.965913369141796e-06, + "loss": 0.1307, + "step": 1344 + }, + { + "epoch": 0.8638407193320488, + "grad_norm": 0.9045458836329427, + "learning_rate": 8.963636293145472e-06, + "loss": 0.1086, + "step": 1345 + }, + { + "epoch": 0.8644829800899165, + "grad_norm": 0.7935673264522606, + "learning_rate": 8.961357002653957e-06, + "loss": 0.0889, + "step": 1346 + }, + { + "epoch": 0.8651252408477842, + "grad_norm": 1.0218945511824098, + "learning_rate": 8.959075498940695e-06, + "loss": 0.1934, + "step": 1347 + }, + { + "epoch": 0.8657675016056519, + "grad_norm": 0.8819279670508089, + "learning_rate": 8.956791783280373e-06, + "loss": 0.1255, + "step": 1348 + }, + { + "epoch": 0.8664097623635196, + "grad_norm": 0.954653514075135, + "learning_rate": 8.954505856948907e-06, + "loss": 0.1332, + "step": 1349 + }, + { + "epoch": 0.8670520231213873, + "grad_norm": 1.2434110701862098, + "learning_rate": 8.95221772122345e-06, + "loss": 0.1213, + "step": 1350 + }, + { + "epoch": 0.867694283879255, + "grad_norm": 1.1308584639341668, + "learning_rate": 8.949927377382395e-06, + "loss": 0.1929, + "step": 1351 + }, + { + "epoch": 0.8683365446371226, + "grad_norm": 0.763276059502055, + "learning_rate": 8.947634826705356e-06, + "loss": 0.0884, + "step": 1352 + }, + { + "epoch": 0.8689788053949904, + "grad_norm": 1.1266842800862014, + "learning_rate": 8.945340070473194e-06, + "loss": 0.1916, + "step": 1353 + }, + { + "epoch": 0.8696210661528581, + "grad_norm": 1.1977255299568905, + "learning_rate": 8.943043109967994e-06, + "loss": 0.1206, + "step": 1354 + }, + { + "epoch": 0.8702633269107257, + "grad_norm": 1.2181677983330728, + "learning_rate": 8.940743946473074e-06, + "loss": 0.1885, + "step": 1355 + }, + { + "epoch": 0.8709055876685935, + "grad_norm": 1.1119425806982315, + "learning_rate": 8.938442581272984e-06, + "loss": 0.153, + "step": 1356 + }, + { + "epoch": 0.8715478484264612, + "grad_norm": 0.7373790742852725, + "learning_rate": 8.936139015653503e-06, + "loss": 0.0761, + "step": 1357 + }, + { + "epoch": 0.8721901091843288, + "grad_norm": 0.9312933767387477, + "learning_rate": 8.93383325090164e-06, + "loss": 0.1343, + "step": 1358 + }, + { + "epoch": 0.8728323699421965, + "grad_norm": 0.9778310089215294, + "learning_rate": 8.931525288305633e-06, + "loss": 0.1371, + "step": 1359 + }, + { + "epoch": 0.8734746307000643, + "grad_norm": 1.181923691081034, + "learning_rate": 8.929215129154947e-06, + "loss": 0.1727, + "step": 1360 + }, + { + "epoch": 0.8741168914579319, + "grad_norm": 1.4036928235947597, + "learning_rate": 8.926902774740276e-06, + "loss": 0.3065, + "step": 1361 + }, + { + "epoch": 0.8747591522157996, + "grad_norm": 1.1948571498684224, + "learning_rate": 8.924588226353536e-06, + "loss": 0.2484, + "step": 1362 + }, + { + "epoch": 0.8754014129736674, + "grad_norm": 1.3120399038253971, + "learning_rate": 8.922271485287876e-06, + "loss": 0.1215, + "step": 1363 + }, + { + "epoch": 0.876043673731535, + "grad_norm": 0.8104278413977286, + "learning_rate": 8.919952552837662e-06, + "loss": 0.1108, + "step": 1364 + }, + { + "epoch": 0.8766859344894027, + "grad_norm": 1.0635984680168182, + "learning_rate": 8.917631430298493e-06, + "loss": 0.1283, + "step": 1365 + }, + { + "epoch": 0.8773281952472703, + "grad_norm": 1.102540376027644, + "learning_rate": 8.915308118967181e-06, + "loss": 0.1497, + "step": 1366 + }, + { + "epoch": 0.8779704560051381, + "grad_norm": 0.7741100647167495, + "learning_rate": 8.912982620141772e-06, + "loss": 0.0648, + "step": 1367 + }, + { + "epoch": 0.8786127167630058, + "grad_norm": 0.9985749489930345, + "learning_rate": 8.910654935121528e-06, + "loss": 0.1213, + "step": 1368 + }, + { + "epoch": 0.8792549775208734, + "grad_norm": 0.8857367306185182, + "learning_rate": 8.908325065206932e-06, + "loss": 0.1191, + "step": 1369 + }, + { + "epoch": 0.8798972382787412, + "grad_norm": 1.0201863883587856, + "learning_rate": 8.905993011699687e-06, + "loss": 0.1085, + "step": 1370 + }, + { + "epoch": 0.8805394990366089, + "grad_norm": 0.9321955845533133, + "learning_rate": 8.903658775902725e-06, + "loss": 0.0789, + "step": 1371 + }, + { + "epoch": 0.8811817597944765, + "grad_norm": 0.9853176606773614, + "learning_rate": 8.901322359120183e-06, + "loss": 0.1329, + "step": 1372 + }, + { + "epoch": 0.8818240205523442, + "grad_norm": 1.0946521641987619, + "learning_rate": 8.898983762657426e-06, + "loss": 0.1743, + "step": 1373 + }, + { + "epoch": 0.882466281310212, + "grad_norm": 1.0840329431082198, + "learning_rate": 8.89664298782104e-06, + "loss": 0.1377, + "step": 1374 + }, + { + "epoch": 0.8831085420680796, + "grad_norm": 0.8967918087108261, + "learning_rate": 8.894300035918816e-06, + "loss": 0.1261, + "step": 1375 + }, + { + "epoch": 0.8837508028259473, + "grad_norm": 1.2182172860712042, + "learning_rate": 8.89195490825977e-06, + "loss": 0.2164, + "step": 1376 + }, + { + "epoch": 0.884393063583815, + "grad_norm": 0.8236403713287697, + "learning_rate": 8.889607606154132e-06, + "loss": 0.0841, + "step": 1377 + }, + { + "epoch": 0.8850353243416827, + "grad_norm": 0.8317792444331302, + "learning_rate": 8.887258130913347e-06, + "loss": 0.1066, + "step": 1378 + }, + { + "epoch": 0.8856775850995504, + "grad_norm": 0.9922995795667331, + "learning_rate": 8.884906483850075e-06, + "loss": 0.1386, + "step": 1379 + }, + { + "epoch": 0.8863198458574181, + "grad_norm": 0.8998857910300798, + "learning_rate": 8.882552666278185e-06, + "loss": 0.0734, + "step": 1380 + }, + { + "epoch": 0.8869621066152859, + "grad_norm": 1.0989937837362636, + "learning_rate": 8.880196679512765e-06, + "loss": 0.2112, + "step": 1381 + }, + { + "epoch": 0.8876043673731535, + "grad_norm": 0.8252010893001313, + "learning_rate": 8.877838524870109e-06, + "loss": 0.1302, + "step": 1382 + }, + { + "epoch": 0.8882466281310212, + "grad_norm": 1.0609365065432856, + "learning_rate": 8.875478203667725e-06, + "loss": 0.1511, + "step": 1383 + }, + { + "epoch": 0.8888888888888888, + "grad_norm": 1.2974801170565333, + "learning_rate": 8.873115717224334e-06, + "loss": 0.2907, + "step": 1384 + }, + { + "epoch": 0.8895311496467566, + "grad_norm": 0.6283777881995973, + "learning_rate": 8.870751066859862e-06, + "loss": 0.0586, + "step": 1385 + }, + { + "epoch": 0.8901734104046243, + "grad_norm": 1.4727797345690126, + "learning_rate": 8.868384253895445e-06, + "loss": 0.206, + "step": 1386 + }, + { + "epoch": 0.8908156711624919, + "grad_norm": 0.8618085507686493, + "learning_rate": 8.866015279653432e-06, + "loss": 0.1133, + "step": 1387 + }, + { + "epoch": 0.8914579319203597, + "grad_norm": 1.1001922722587156, + "learning_rate": 8.863644145457371e-06, + "loss": 0.178, + "step": 1388 + }, + { + "epoch": 0.8921001926782274, + "grad_norm": 0.9019606921271982, + "learning_rate": 8.861270852632027e-06, + "loss": 0.1483, + "step": 1389 + }, + { + "epoch": 0.892742453436095, + "grad_norm": 1.0359987123733958, + "learning_rate": 8.85889540250336e-06, + "loss": 0.16, + "step": 1390 + }, + { + "epoch": 0.8933847141939627, + "grad_norm": 1.0515386679038083, + "learning_rate": 8.856517796398546e-06, + "loss": 0.1673, + "step": 1391 + }, + { + "epoch": 0.8940269749518305, + "grad_norm": 1.0576341334797563, + "learning_rate": 8.854138035645958e-06, + "loss": 0.1358, + "step": 1392 + }, + { + "epoch": 0.8946692357096981, + "grad_norm": 0.993796925480109, + "learning_rate": 8.851756121575174e-06, + "loss": 0.1195, + "step": 1393 + }, + { + "epoch": 0.8953114964675658, + "grad_norm": 0.9548043113115483, + "learning_rate": 8.84937205551698e-06, + "loss": 0.1317, + "step": 1394 + }, + { + "epoch": 0.8959537572254336, + "grad_norm": 0.9034147997959365, + "learning_rate": 8.846985838803357e-06, + "loss": 0.1187, + "step": 1395 + }, + { + "epoch": 0.8965960179833012, + "grad_norm": 0.9453301439627194, + "learning_rate": 8.844597472767492e-06, + "loss": 0.0954, + "step": 1396 + }, + { + "epoch": 0.8972382787411689, + "grad_norm": 0.7540811600670605, + "learning_rate": 8.842206958743777e-06, + "loss": 0.0906, + "step": 1397 + }, + { + "epoch": 0.8978805394990366, + "grad_norm": 1.051416053966678, + "learning_rate": 8.839814298067792e-06, + "loss": 0.102, + "step": 1398 + }, + { + "epoch": 0.8985228002569043, + "grad_norm": 0.9265871792970176, + "learning_rate": 8.837419492076326e-06, + "loss": 0.0771, + "step": 1399 + }, + { + "epoch": 0.899165061014772, + "grad_norm": 0.8790229782761598, + "learning_rate": 8.835022542107367e-06, + "loss": 0.1115, + "step": 1400 + }, + { + "epoch": 0.8998073217726397, + "grad_norm": 1.0846711895740078, + "learning_rate": 8.832623449500096e-06, + "loss": 0.2119, + "step": 1401 + }, + { + "epoch": 0.9004495825305073, + "grad_norm": 0.6927762579279171, + "learning_rate": 8.83022221559489e-06, + "loss": 0.0667, + "step": 1402 + }, + { + "epoch": 0.9010918432883751, + "grad_norm": 1.030076641279745, + "learning_rate": 8.827818841733331e-06, + "loss": 0.1444, + "step": 1403 + }, + { + "epoch": 0.9017341040462428, + "grad_norm": 0.736826422785276, + "learning_rate": 8.825413329258187e-06, + "loss": 0.0709, + "step": 1404 + }, + { + "epoch": 0.9023763648041104, + "grad_norm": 1.1773572635017528, + "learning_rate": 8.823005679513426e-06, + "loss": 0.2761, + "step": 1405 + }, + { + "epoch": 0.9030186255619782, + "grad_norm": 0.9941069007636847, + "learning_rate": 8.820595893844209e-06, + "loss": 0.1667, + "step": 1406 + }, + { + "epoch": 0.9036608863198459, + "grad_norm": 0.8173506558841139, + "learning_rate": 8.81818397359689e-06, + "loss": 0.0955, + "step": 1407 + }, + { + "epoch": 0.9043031470777135, + "grad_norm": 0.7959404072021629, + "learning_rate": 8.815769920119017e-06, + "loss": 0.1043, + "step": 1408 + }, + { + "epoch": 0.9049454078355812, + "grad_norm": 1.1639191761659684, + "learning_rate": 8.813353734759327e-06, + "loss": 0.2038, + "step": 1409 + }, + { + "epoch": 0.905587668593449, + "grad_norm": 0.8482515440715065, + "learning_rate": 8.810935418867752e-06, + "loss": 0.1063, + "step": 1410 + }, + { + "epoch": 0.9062299293513166, + "grad_norm": 1.3685418882119331, + "learning_rate": 8.80851497379541e-06, + "loss": 0.3537, + "step": 1411 + }, + { + "epoch": 0.9068721901091843, + "grad_norm": 1.0494432969080132, + "learning_rate": 8.806092400894612e-06, + "loss": 0.2, + "step": 1412 + }, + { + "epoch": 0.9075144508670521, + "grad_norm": 0.8692865580663636, + "learning_rate": 8.803667701518857e-06, + "loss": 0.1097, + "step": 1413 + }, + { + "epoch": 0.9081567116249197, + "grad_norm": 0.8520689096985141, + "learning_rate": 8.801240877022829e-06, + "loss": 0.1121, + "step": 1414 + }, + { + "epoch": 0.9087989723827874, + "grad_norm": 1.0752990330430645, + "learning_rate": 8.798811928762405e-06, + "loss": 0.0941, + "step": 1415 + }, + { + "epoch": 0.9094412331406551, + "grad_norm": 0.9526567265259605, + "learning_rate": 8.796380858094645e-06, + "loss": 0.1512, + "step": 1416 + }, + { + "epoch": 0.9100834938985228, + "grad_norm": 1.0775214415755474, + "learning_rate": 8.793947666377793e-06, + "loss": 0.1269, + "step": 1417 + }, + { + "epoch": 0.9107257546563905, + "grad_norm": 1.1386943169598496, + "learning_rate": 8.791512354971283e-06, + "loss": 0.2154, + "step": 1418 + }, + { + "epoch": 0.9113680154142582, + "grad_norm": 1.2475357004740217, + "learning_rate": 8.789074925235729e-06, + "loss": 0.2771, + "step": 1419 + }, + { + "epoch": 0.9120102761721259, + "grad_norm": 1.5175156018436209, + "learning_rate": 8.786635378532933e-06, + "loss": 0.0903, + "step": 1420 + }, + { + "epoch": 0.9126525369299936, + "grad_norm": 1.2033254620024585, + "learning_rate": 8.784193716225872e-06, + "loss": 0.2603, + "step": 1421 + }, + { + "epoch": 0.9132947976878613, + "grad_norm": 1.087272857901764, + "learning_rate": 8.781749939678712e-06, + "loss": 0.1693, + "step": 1422 + }, + { + "epoch": 0.9139370584457289, + "grad_norm": 1.817649578314051, + "learning_rate": 8.779304050256801e-06, + "loss": 0.2137, + "step": 1423 + }, + { + "epoch": 0.9145793192035967, + "grad_norm": 1.2092645857568134, + "learning_rate": 8.77685604932666e-06, + "loss": 0.2079, + "step": 1424 + }, + { + "epoch": 0.9152215799614644, + "grad_norm": 1.0481748691060706, + "learning_rate": 8.774405938255997e-06, + "loss": 0.1508, + "step": 1425 + }, + { + "epoch": 0.915863840719332, + "grad_norm": 0.9661336761187245, + "learning_rate": 8.771953718413698e-06, + "loss": 0.1445, + "step": 1426 + }, + { + "epoch": 0.9165061014771998, + "grad_norm": 0.8949902511409732, + "learning_rate": 8.769499391169819e-06, + "loss": 0.1211, + "step": 1427 + }, + { + "epoch": 0.9171483622350675, + "grad_norm": 1.3479658702974189, + "learning_rate": 8.767042957895607e-06, + "loss": 0.2322, + "step": 1428 + }, + { + "epoch": 0.9177906229929351, + "grad_norm": 1.238122739986945, + "learning_rate": 8.764584419963474e-06, + "loss": 0.3118, + "step": 1429 + }, + { + "epoch": 0.9184328837508028, + "grad_norm": 0.9887353874093554, + "learning_rate": 8.762123778747011e-06, + "loss": 0.1303, + "step": 1430 + }, + { + "epoch": 0.9190751445086706, + "grad_norm": 0.9813288624861485, + "learning_rate": 8.759661035620992e-06, + "loss": 0.1165, + "step": 1431 + }, + { + "epoch": 0.9197174052665382, + "grad_norm": 1.0171464871477869, + "learning_rate": 8.757196191961353e-06, + "loss": 0.1838, + "step": 1432 + }, + { + "epoch": 0.9203596660244059, + "grad_norm": 0.9797229248040947, + "learning_rate": 8.754729249145211e-06, + "loss": 0.096, + "step": 1433 + }, + { + "epoch": 0.9210019267822736, + "grad_norm": 1.0529287735118944, + "learning_rate": 8.752260208550855e-06, + "loss": 0.2047, + "step": 1434 + }, + { + "epoch": 0.9216441875401413, + "grad_norm": 0.9569453575573863, + "learning_rate": 8.749789071557745e-06, + "loss": 0.0766, + "step": 1435 + }, + { + "epoch": 0.922286448298009, + "grad_norm": 1.1118716155961232, + "learning_rate": 8.747315839546512e-06, + "loss": 0.1919, + "step": 1436 + }, + { + "epoch": 0.9229287090558767, + "grad_norm": 0.9843114217794342, + "learning_rate": 8.744840513898957e-06, + "loss": 0.1476, + "step": 1437 + }, + { + "epoch": 0.9235709698137444, + "grad_norm": 0.8472650330463956, + "learning_rate": 8.742363095998054e-06, + "loss": 0.1029, + "step": 1438 + }, + { + "epoch": 0.9242132305716121, + "grad_norm": 0.9854372375863621, + "learning_rate": 8.739883587227941e-06, + "loss": 0.1194, + "step": 1439 + }, + { + "epoch": 0.9248554913294798, + "grad_norm": 1.1432176793565958, + "learning_rate": 8.73740198897393e-06, + "loss": 0.2345, + "step": 1440 + }, + { + "epoch": 0.9254977520873474, + "grad_norm": 1.2058361612462867, + "learning_rate": 8.734918302622493e-06, + "loss": 0.2256, + "step": 1441 + }, + { + "epoch": 0.9261400128452152, + "grad_norm": 1.148134028711643, + "learning_rate": 8.732432529561277e-06, + "loss": 0.2487, + "step": 1442 + }, + { + "epoch": 0.9267822736030829, + "grad_norm": 0.8902599093884933, + "learning_rate": 8.729944671179088e-06, + "loss": 0.1011, + "step": 1443 + }, + { + "epoch": 0.9274245343609505, + "grad_norm": 1.1131786144763074, + "learning_rate": 8.727454728865901e-06, + "loss": 0.2023, + "step": 1444 + }, + { + "epoch": 0.9280667951188183, + "grad_norm": 1.223761086975885, + "learning_rate": 8.724962704012853e-06, + "loss": 0.2025, + "step": 1445 + }, + { + "epoch": 0.928709055876686, + "grad_norm": 1.1711557390881737, + "learning_rate": 8.722468598012246e-06, + "loss": 0.1909, + "step": 1446 + }, + { + "epoch": 0.9293513166345536, + "grad_norm": 0.9586194940832241, + "learning_rate": 8.719972412257545e-06, + "loss": 0.151, + "step": 1447 + }, + { + "epoch": 0.9299935773924213, + "grad_norm": 0.9636933053701001, + "learning_rate": 8.717474148143377e-06, + "loss": 0.0952, + "step": 1448 + }, + { + "epoch": 0.930635838150289, + "grad_norm": 0.9534114370069181, + "learning_rate": 8.714973807065525e-06, + "loss": 0.1013, + "step": 1449 + }, + { + "epoch": 0.9312780989081567, + "grad_norm": 0.9150603795212835, + "learning_rate": 8.712471390420944e-06, + "loss": 0.137, + "step": 1450 + }, + { + "epoch": 0.9319203596660244, + "grad_norm": 0.7872411944832919, + "learning_rate": 8.709966899607735e-06, + "loss": 0.0691, + "step": 1451 + }, + { + "epoch": 0.9325626204238922, + "grad_norm": 1.214019204988242, + "learning_rate": 8.707460336025168e-06, + "loss": 0.2279, + "step": 1452 + }, + { + "epoch": 0.9332048811817598, + "grad_norm": 1.0437364325637781, + "learning_rate": 8.704951701073668e-06, + "loss": 0.1327, + "step": 1453 + }, + { + "epoch": 0.9338471419396275, + "grad_norm": 0.9328905776966507, + "learning_rate": 8.702440996154813e-06, + "loss": 0.0751, + "step": 1454 + }, + { + "epoch": 0.9344894026974951, + "grad_norm": 1.1398656587853022, + "learning_rate": 8.699928222671343e-06, + "loss": 0.2268, + "step": 1455 + }, + { + "epoch": 0.9351316634553629, + "grad_norm": 0.9399799886563431, + "learning_rate": 8.697413382027153e-06, + "loss": 0.1053, + "step": 1456 + }, + { + "epoch": 0.9357739242132306, + "grad_norm": 1.0149721789364357, + "learning_rate": 8.694896475627292e-06, + "loss": 0.115, + "step": 1457 + }, + { + "epoch": 0.9364161849710982, + "grad_norm": 0.921777692505786, + "learning_rate": 8.69237750487796e-06, + "loss": 0.1364, + "step": 1458 + }, + { + "epoch": 0.9370584457289659, + "grad_norm": 0.7922275709090735, + "learning_rate": 8.689856471186518e-06, + "loss": 0.0889, + "step": 1459 + }, + { + "epoch": 0.9377007064868337, + "grad_norm": 0.861724480976533, + "learning_rate": 8.687333375961468e-06, + "loss": 0.0793, + "step": 1460 + }, + { + "epoch": 0.9383429672447013, + "grad_norm": 1.2981340964889576, + "learning_rate": 8.684808220612478e-06, + "loss": 0.2187, + "step": 1461 + }, + { + "epoch": 0.938985228002569, + "grad_norm": 0.916155714248433, + "learning_rate": 8.682281006550356e-06, + "loss": 0.1336, + "step": 1462 + }, + { + "epoch": 0.9396274887604368, + "grad_norm": 1.2258869735980638, + "learning_rate": 8.679751735187065e-06, + "loss": 0.2007, + "step": 1463 + }, + { + "epoch": 0.9402697495183044, + "grad_norm": 0.8110078569786703, + "learning_rate": 8.677220407935714e-06, + "loss": 0.1017, + "step": 1464 + }, + { + "epoch": 0.9409120102761721, + "grad_norm": 0.9665908923103776, + "learning_rate": 8.674687026210567e-06, + "loss": 0.1225, + "step": 1465 + }, + { + "epoch": 0.9415542710340398, + "grad_norm": 1.2095006075600234, + "learning_rate": 8.672151591427027e-06, + "loss": 0.2109, + "step": 1466 + }, + { + "epoch": 0.9421965317919075, + "grad_norm": 1.025029759902557, + "learning_rate": 8.669614105001652e-06, + "loss": 0.175, + "step": 1467 + }, + { + "epoch": 0.9428387925497752, + "grad_norm": 1.048239482576429, + "learning_rate": 8.667074568352143e-06, + "loss": 0.1474, + "step": 1468 + }, + { + "epoch": 0.9434810533076429, + "grad_norm": 1.143545180812073, + "learning_rate": 8.664532982897346e-06, + "loss": 0.1634, + "step": 1469 + }, + { + "epoch": 0.9441233140655106, + "grad_norm": 1.4214947812443137, + "learning_rate": 8.661989350057251e-06, + "loss": 0.3369, + "step": 1470 + }, + { + "epoch": 0.9447655748233783, + "grad_norm": 1.0213494203628708, + "learning_rate": 8.659443671252995e-06, + "loss": 0.1551, + "step": 1471 + }, + { + "epoch": 0.945407835581246, + "grad_norm": 1.3267463980922412, + "learning_rate": 8.656895947906856e-06, + "loss": 0.3236, + "step": 1472 + }, + { + "epoch": 0.9460500963391136, + "grad_norm": 0.9249831600558804, + "learning_rate": 8.654346181442255e-06, + "loss": 0.1248, + "step": 1473 + }, + { + "epoch": 0.9466923570969814, + "grad_norm": 0.9776426254155072, + "learning_rate": 8.65179437328375e-06, + "loss": 0.1766, + "step": 1474 + }, + { + "epoch": 0.9473346178548491, + "grad_norm": 1.065001686054923, + "learning_rate": 8.64924052485705e-06, + "loss": 0.1811, + "step": 1475 + }, + { + "epoch": 0.9479768786127167, + "grad_norm": 0.9013496821294367, + "learning_rate": 8.646684637588992e-06, + "loss": 0.0795, + "step": 1476 + }, + { + "epoch": 0.9486191393705845, + "grad_norm": 0.86168346523351, + "learning_rate": 8.64412671290756e-06, + "loss": 0.1403, + "step": 1477 + }, + { + "epoch": 0.9492614001284522, + "grad_norm": 0.9836247440217983, + "learning_rate": 8.641566752241874e-06, + "loss": 0.1639, + "step": 1478 + }, + { + "epoch": 0.9499036608863198, + "grad_norm": 0.9332372250579678, + "learning_rate": 8.63900475702219e-06, + "loss": 0.1125, + "step": 1479 + }, + { + "epoch": 0.9505459216441875, + "grad_norm": 0.669044969897982, + "learning_rate": 8.636440728679907e-06, + "loss": 0.0725, + "step": 1480 + }, + { + "epoch": 0.9511881824020553, + "grad_norm": 0.8764447340932578, + "learning_rate": 8.63387466864755e-06, + "loss": 0.101, + "step": 1481 + }, + { + "epoch": 0.9518304431599229, + "grad_norm": 0.7446069531152378, + "learning_rate": 8.631306578358787e-06, + "loss": 0.0687, + "step": 1482 + }, + { + "epoch": 0.9524727039177906, + "grad_norm": 1.1348050027394023, + "learning_rate": 8.628736459248414e-06, + "loss": 0.1355, + "step": 1483 + }, + { + "epoch": 0.9531149646756584, + "grad_norm": 0.8437211530814774, + "learning_rate": 8.62616431275237e-06, + "loss": 0.0873, + "step": 1484 + }, + { + "epoch": 0.953757225433526, + "grad_norm": 0.8607207845163786, + "learning_rate": 8.623590140307715e-06, + "loss": 0.1228, + "step": 1485 + }, + { + "epoch": 0.9543994861913937, + "grad_norm": 0.7959645827081309, + "learning_rate": 8.62101394335265e-06, + "loss": 0.1134, + "step": 1486 + }, + { + "epoch": 0.9550417469492614, + "grad_norm": 1.1238164071983126, + "learning_rate": 8.618435723326502e-06, + "loss": 0.1675, + "step": 1487 + }, + { + "epoch": 0.9556840077071291, + "grad_norm": 0.882069098596952, + "learning_rate": 8.615855481669733e-06, + "loss": 0.0674, + "step": 1488 + }, + { + "epoch": 0.9563262684649968, + "grad_norm": 1.1107246641812767, + "learning_rate": 8.613273219823927e-06, + "loss": 0.2057, + "step": 1489 + }, + { + "epoch": 0.9569685292228645, + "grad_norm": 0.7611599504748768, + "learning_rate": 8.610688939231805e-06, + "loss": 0.0814, + "step": 1490 + }, + { + "epoch": 0.9576107899807321, + "grad_norm": 1.0264517049054607, + "learning_rate": 8.60810264133721e-06, + "loss": 0.1397, + "step": 1491 + }, + { + "epoch": 0.9582530507385999, + "grad_norm": 0.863848272467658, + "learning_rate": 8.605514327585116e-06, + "loss": 0.0768, + "step": 1492 + }, + { + "epoch": 0.9588953114964676, + "grad_norm": 1.038540233833422, + "learning_rate": 8.602923999421621e-06, + "loss": 0.0923, + "step": 1493 + }, + { + "epoch": 0.9595375722543352, + "grad_norm": 0.8074774249217516, + "learning_rate": 8.600331658293948e-06, + "loss": 0.0724, + "step": 1494 + }, + { + "epoch": 0.960179833012203, + "grad_norm": 0.9886053195607486, + "learning_rate": 8.597737305650445e-06, + "loss": 0.123, + "step": 1495 + }, + { + "epoch": 0.9608220937700707, + "grad_norm": 1.1219336990661222, + "learning_rate": 8.595140942940587e-06, + "loss": 0.2003, + "step": 1496 + }, + { + "epoch": 0.9614643545279383, + "grad_norm": 0.9025005974058088, + "learning_rate": 8.592542571614966e-06, + "loss": 0.1202, + "step": 1497 + }, + { + "epoch": 0.962106615285806, + "grad_norm": 1.407690304817839, + "learning_rate": 8.589942193125303e-06, + "loss": 0.2554, + "step": 1498 + }, + { + "epoch": 0.9627488760436738, + "grad_norm": 0.8136377219084238, + "learning_rate": 8.587339808924435e-06, + "loss": 0.0794, + "step": 1499 + }, + { + "epoch": 0.9633911368015414, + "grad_norm": 1.2185516480832415, + "learning_rate": 8.584735420466323e-06, + "loss": 0.1319, + "step": 1500 + }, + { + "epoch": 0.9640333975594091, + "grad_norm": 1.0387463542548847, + "learning_rate": 8.582129029206045e-06, + "loss": 0.1817, + "step": 1501 + }, + { + "epoch": 0.9646756583172769, + "grad_norm": 1.0116076643324627, + "learning_rate": 8.579520636599801e-06, + "loss": 0.1479, + "step": 1502 + }, + { + "epoch": 0.9653179190751445, + "grad_norm": 0.9107143010432545, + "learning_rate": 8.576910244104905e-06, + "loss": 0.124, + "step": 1503 + }, + { + "epoch": 0.9659601798330122, + "grad_norm": 0.9997514476572634, + "learning_rate": 8.574297853179795e-06, + "loss": 0.0879, + "step": 1504 + }, + { + "epoch": 0.9666024405908799, + "grad_norm": 0.9007639168707042, + "learning_rate": 8.571683465284017e-06, + "loss": 0.1016, + "step": 1505 + }, + { + "epoch": 0.9672447013487476, + "grad_norm": 0.9742110704253959, + "learning_rate": 8.56906708187824e-06, + "loss": 0.1258, + "step": 1506 + }, + { + "epoch": 0.9678869621066153, + "grad_norm": 0.8563633967898144, + "learning_rate": 8.566448704424244e-06, + "loss": 0.1277, + "step": 1507 + }, + { + "epoch": 0.968529222864483, + "grad_norm": 0.9442275346506881, + "learning_rate": 8.563828334384925e-06, + "loss": 0.1113, + "step": 1508 + }, + { + "epoch": 0.9691714836223507, + "grad_norm": 2.2478597973580614, + "learning_rate": 8.56120597322429e-06, + "loss": 0.1524, + "step": 1509 + }, + { + "epoch": 0.9698137443802184, + "grad_norm": 0.7813510437323623, + "learning_rate": 8.558581622407463e-06, + "loss": 0.0917, + "step": 1510 + }, + { + "epoch": 0.970456005138086, + "grad_norm": 0.8041738122390029, + "learning_rate": 8.555955283400672e-06, + "loss": 0.0891, + "step": 1511 + }, + { + "epoch": 0.9710982658959537, + "grad_norm": 0.9700133325306851, + "learning_rate": 8.553326957671264e-06, + "loss": 0.1324, + "step": 1512 + }, + { + "epoch": 0.9717405266538215, + "grad_norm": 1.0809723156501034, + "learning_rate": 8.55069664668769e-06, + "loss": 0.1085, + "step": 1513 + }, + { + "epoch": 0.9723827874116892, + "grad_norm": 0.9956564503927868, + "learning_rate": 8.548064351919513e-06, + "loss": 0.1818, + "step": 1514 + }, + { + "epoch": 0.9730250481695568, + "grad_norm": 0.9445478298812199, + "learning_rate": 8.545430074837404e-06, + "loss": 0.1367, + "step": 1515 + }, + { + "epoch": 0.9736673089274245, + "grad_norm": 0.7388019869864687, + "learning_rate": 8.542793816913142e-06, + "loss": 0.0529, + "step": 1516 + }, + { + "epoch": 0.9743095696852923, + "grad_norm": 1.230425417708672, + "learning_rate": 8.54015557961961e-06, + "loss": 0.1755, + "step": 1517 + }, + { + "epoch": 0.9749518304431599, + "grad_norm": 1.0722344209018604, + "learning_rate": 8.537515364430801e-06, + "loss": 0.1113, + "step": 1518 + }, + { + "epoch": 0.9755940912010276, + "grad_norm": 0.7368097239782148, + "learning_rate": 8.534873172821809e-06, + "loss": 0.067, + "step": 1519 + }, + { + "epoch": 0.9762363519588954, + "grad_norm": 0.8264553368486212, + "learning_rate": 8.532229006268837e-06, + "loss": 0.1114, + "step": 1520 + }, + { + "epoch": 0.976878612716763, + "grad_norm": 1.020370511998055, + "learning_rate": 8.529582866249187e-06, + "loss": 0.2062, + "step": 1521 + }, + { + "epoch": 0.9775208734746307, + "grad_norm": 1.0561559248835797, + "learning_rate": 8.526934754241263e-06, + "loss": 0.1484, + "step": 1522 + }, + { + "epoch": 0.9781631342324983, + "grad_norm": 0.9391269842597171, + "learning_rate": 8.524284671724576e-06, + "loss": 0.1625, + "step": 1523 + }, + { + "epoch": 0.9788053949903661, + "grad_norm": 1.0128255350978894, + "learning_rate": 8.521632620179735e-06, + "loss": 0.1315, + "step": 1524 + }, + { + "epoch": 0.9794476557482338, + "grad_norm": 1.0742161090261613, + "learning_rate": 8.518978601088448e-06, + "loss": 0.1977, + "step": 1525 + }, + { + "epoch": 0.9800899165061014, + "grad_norm": 0.9270108228443203, + "learning_rate": 8.516322615933521e-06, + "loss": 0.1353, + "step": 1526 + }, + { + "epoch": 0.9807321772639692, + "grad_norm": 1.04943814196786, + "learning_rate": 8.513664666198865e-06, + "loss": 0.1519, + "step": 1527 + }, + { + "epoch": 0.9813744380218369, + "grad_norm": 0.8482590005649712, + "learning_rate": 8.51100475336948e-06, + "loss": 0.1027, + "step": 1528 + }, + { + "epoch": 0.9820166987797045, + "grad_norm": 0.9383469258155767, + "learning_rate": 8.50834287893147e-06, + "loss": 0.1409, + "step": 1529 + }, + { + "epoch": 0.9826589595375722, + "grad_norm": 1.0280222052168781, + "learning_rate": 8.50567904437203e-06, + "loss": 0.137, + "step": 1530 + }, + { + "epoch": 0.98330122029544, + "grad_norm": 1.0566142115946089, + "learning_rate": 8.503013251179455e-06, + "loss": 0.1965, + "step": 1531 + }, + { + "epoch": 0.9839434810533076, + "grad_norm": 0.7071957057269431, + "learning_rate": 8.50034550084313e-06, + "loss": 0.079, + "step": 1532 + }, + { + "epoch": 0.9845857418111753, + "grad_norm": 1.2323818935407846, + "learning_rate": 8.497675794853532e-06, + "loss": 0.1192, + "step": 1533 + }, + { + "epoch": 0.9852280025690431, + "grad_norm": 0.9738686845958974, + "learning_rate": 8.49500413470224e-06, + "loss": 0.1434, + "step": 1534 + }, + { + "epoch": 0.9858702633269107, + "grad_norm": 0.8062867387295259, + "learning_rate": 8.492330521881914e-06, + "loss": 0.0787, + "step": 1535 + }, + { + "epoch": 0.9865125240847784, + "grad_norm": 0.8522564649879029, + "learning_rate": 8.489654957886306e-06, + "loss": 0.0739, + "step": 1536 + }, + { + "epoch": 0.9871547848426461, + "grad_norm": 0.9945756484159634, + "learning_rate": 8.486977444210268e-06, + "loss": 0.1315, + "step": 1537 + }, + { + "epoch": 0.9877970456005138, + "grad_norm": 1.0843915919595541, + "learning_rate": 8.484297982349731e-06, + "loss": 0.1779, + "step": 1538 + }, + { + "epoch": 0.9884393063583815, + "grad_norm": 1.0320844738734043, + "learning_rate": 8.48161657380172e-06, + "loss": 0.1575, + "step": 1539 + }, + { + "epoch": 0.9890815671162492, + "grad_norm": 1.0335663724517776, + "learning_rate": 8.478933220064347e-06, + "loss": 0.1299, + "step": 1540 + }, + { + "epoch": 0.989723827874117, + "grad_norm": 1.0439530737802356, + "learning_rate": 8.476247922636805e-06, + "loss": 0.1596, + "step": 1541 + }, + { + "epoch": 0.9903660886319846, + "grad_norm": 0.930598709571447, + "learning_rate": 8.47356068301938e-06, + "loss": 0.1058, + "step": 1542 + }, + { + "epoch": 0.9910083493898523, + "grad_norm": 0.8922627815866374, + "learning_rate": 8.470871502713442e-06, + "loss": 0.1303, + "step": 1543 + }, + { + "epoch": 0.9916506101477199, + "grad_norm": 1.1067390301576283, + "learning_rate": 8.468180383221441e-06, + "loss": 0.1838, + "step": 1544 + }, + { + "epoch": 0.9922928709055877, + "grad_norm": 0.8685488190904991, + "learning_rate": 8.465487326046917e-06, + "loss": 0.1071, + "step": 1545 + }, + { + "epoch": 0.9929351316634554, + "grad_norm": 0.9422258816673255, + "learning_rate": 8.462792332694485e-06, + "loss": 0.1042, + "step": 1546 + }, + { + "epoch": 0.993577392421323, + "grad_norm": 0.953377873623732, + "learning_rate": 8.460095404669848e-06, + "loss": 0.1501, + "step": 1547 + }, + { + "epoch": 0.9942196531791907, + "grad_norm": 1.005047325439977, + "learning_rate": 8.457396543479787e-06, + "loss": 0.1554, + "step": 1548 + }, + { + "epoch": 0.9948619139370585, + "grad_norm": 0.8856386891782324, + "learning_rate": 8.454695750632165e-06, + "loss": 0.0968, + "step": 1549 + }, + { + "epoch": 0.9955041746949261, + "grad_norm": 0.9146415804719286, + "learning_rate": 8.451993027635921e-06, + "loss": 0.1398, + "step": 1550 + }, + { + "epoch": 0.9961464354527938, + "grad_norm": 1.0205362710956871, + "learning_rate": 8.449288376001074e-06, + "loss": 0.1584, + "step": 1551 + }, + { + "epoch": 0.9967886962106616, + "grad_norm": 0.9628961947159539, + "learning_rate": 8.446581797238722e-06, + "loss": 0.1275, + "step": 1552 + }, + { + "epoch": 0.9974309569685292, + "grad_norm": 0.9726948406144199, + "learning_rate": 8.44387329286104e-06, + "loss": 0.1332, + "step": 1553 + }, + { + "epoch": 0.9980732177263969, + "grad_norm": 1.102278602297612, + "learning_rate": 8.441162864381274e-06, + "loss": 0.2142, + "step": 1554 + }, + { + "epoch": 0.9987154784842646, + "grad_norm": 0.9839770067428314, + "learning_rate": 8.438450513313749e-06, + "loss": 0.1531, + "step": 1555 + }, + { + "epoch": 0.9993577392421323, + "grad_norm": 0.9619377442711502, + "learning_rate": 8.435736241173866e-06, + "loss": 0.1338, + "step": 1556 + }, + { + "epoch": 1.0, + "grad_norm": 1.1064843932217125, + "learning_rate": 8.433020049478093e-06, + "loss": 0.1785, + "step": 1557 + }, + { + "epoch": 1.0006422607578678, + "grad_norm": 0.8690875198623533, + "learning_rate": 8.430301939743979e-06, + "loss": 0.0739, + "step": 1558 + }, + { + "epoch": 1.0012845215157353, + "grad_norm": 0.6744350494858311, + "learning_rate": 8.427581913490134e-06, + "loss": 0.0631, + "step": 1559 + }, + { + "epoch": 1.001926782273603, + "grad_norm": 0.9874436279002307, + "learning_rate": 8.42485997223625e-06, + "loss": 0.0826, + "step": 1560 + }, + { + "epoch": 1.0025690430314709, + "grad_norm": 9.449369721389886, + "learning_rate": 8.42213611750308e-06, + "loss": 0.1768, + "step": 1561 + }, + { + "epoch": 1.0032113037893384, + "grad_norm": 0.7309174596416331, + "learning_rate": 8.419410350812453e-06, + "loss": 0.0872, + "step": 1562 + }, + { + "epoch": 1.0038535645472062, + "grad_norm": 0.9837561215495054, + "learning_rate": 8.41668267368726e-06, + "loss": 0.0926, + "step": 1563 + }, + { + "epoch": 1.0044958253050738, + "grad_norm": 0.4868233699528552, + "learning_rate": 8.413953087651462e-06, + "loss": 0.0262, + "step": 1564 + }, + { + "epoch": 1.0051380860629415, + "grad_norm": 0.6542454773241456, + "learning_rate": 8.411221594230085e-06, + "loss": 0.0673, + "step": 1565 + }, + { + "epoch": 1.0057803468208093, + "grad_norm": 0.6865297414484731, + "learning_rate": 8.408488194949229e-06, + "loss": 0.0514, + "step": 1566 + }, + { + "epoch": 1.0064226075786769, + "grad_norm": 0.8811827478581947, + "learning_rate": 8.405752891336047e-06, + "loss": 0.1044, + "step": 1567 + }, + { + "epoch": 1.0070648683365446, + "grad_norm": 0.9808466083734804, + "learning_rate": 8.403015684918763e-06, + "loss": 0.0994, + "step": 1568 + }, + { + "epoch": 1.0077071290944124, + "grad_norm": 0.710808964937678, + "learning_rate": 8.40027657722666e-06, + "loss": 0.0612, + "step": 1569 + }, + { + "epoch": 1.00834938985228, + "grad_norm": 0.81212499220956, + "learning_rate": 8.397535569790088e-06, + "loss": 0.0783, + "step": 1570 + }, + { + "epoch": 1.0089916506101477, + "grad_norm": 0.8733884753710924, + "learning_rate": 8.394792664140454e-06, + "loss": 0.107, + "step": 1571 + }, + { + "epoch": 1.0096339113680155, + "grad_norm": 1.1065568681197386, + "learning_rate": 8.39204786181023e-06, + "loss": 0.1914, + "step": 1572 + }, + { + "epoch": 1.010276172125883, + "grad_norm": 0.6406437680343652, + "learning_rate": 8.38930116433294e-06, + "loss": 0.0612, + "step": 1573 + }, + { + "epoch": 1.0109184328837508, + "grad_norm": 0.889994269244562, + "learning_rate": 8.386552573243182e-06, + "loss": 0.083, + "step": 1574 + }, + { + "epoch": 1.0115606936416186, + "grad_norm": 0.7891882291020146, + "learning_rate": 8.383802090076589e-06, + "loss": 0.0617, + "step": 1575 + }, + { + "epoch": 1.0122029543994862, + "grad_norm": 0.8494248709998803, + "learning_rate": 8.38104971636987e-06, + "loss": 0.0963, + "step": 1576 + }, + { + "epoch": 1.012845215157354, + "grad_norm": 0.8841060151121751, + "learning_rate": 8.378295453660788e-06, + "loss": 0.0906, + "step": 1577 + }, + { + "epoch": 1.0134874759152215, + "grad_norm": 0.8163116497876214, + "learning_rate": 8.375539303488152e-06, + "loss": 0.0671, + "step": 1578 + }, + { + "epoch": 1.0141297366730893, + "grad_norm": 0.8180104708212611, + "learning_rate": 8.372781267391834e-06, + "loss": 0.0754, + "step": 1579 + }, + { + "epoch": 1.014771997430957, + "grad_norm": 0.9518327222655849, + "learning_rate": 8.370021346912756e-06, + "loss": 0.1047, + "step": 1580 + }, + { + "epoch": 1.0154142581888246, + "grad_norm": 0.8209981027489729, + "learning_rate": 8.367259543592892e-06, + "loss": 0.0733, + "step": 1581 + }, + { + "epoch": 1.0160565189466924, + "grad_norm": 0.8083816229172686, + "learning_rate": 8.364495858975274e-06, + "loss": 0.0434, + "step": 1582 + }, + { + "epoch": 1.0166987797045601, + "grad_norm": 0.8533363555954434, + "learning_rate": 8.361730294603975e-06, + "loss": 0.0931, + "step": 1583 + }, + { + "epoch": 1.0173410404624277, + "grad_norm": 0.7879773111026449, + "learning_rate": 8.358962852024128e-06, + "loss": 0.0641, + "step": 1584 + }, + { + "epoch": 1.0179833012202955, + "grad_norm": 0.8602874985069248, + "learning_rate": 8.356193532781912e-06, + "loss": 0.0915, + "step": 1585 + }, + { + "epoch": 1.0186255619781632, + "grad_norm": 1.3293924885326538, + "learning_rate": 8.35342233842455e-06, + "loss": 0.0769, + "step": 1586 + }, + { + "epoch": 1.0192678227360308, + "grad_norm": 1.0819701640625277, + "learning_rate": 8.350649270500319e-06, + "loss": 0.057, + "step": 1587 + }, + { + "epoch": 1.0199100834938986, + "grad_norm": 0.7747878834840397, + "learning_rate": 8.347874330558539e-06, + "loss": 0.0545, + "step": 1588 + }, + { + "epoch": 1.020552344251766, + "grad_norm": 1.1058713321511315, + "learning_rate": 8.345097520149578e-06, + "loss": 0.0579, + "step": 1589 + }, + { + "epoch": 1.0211946050096339, + "grad_norm": 0.7570708192687001, + "learning_rate": 8.342318840824849e-06, + "loss": 0.0882, + "step": 1590 + }, + { + "epoch": 1.0218368657675017, + "grad_norm": 0.9996462306724013, + "learning_rate": 8.339538294136805e-06, + "loss": 0.0915, + "step": 1591 + }, + { + "epoch": 1.0224791265253692, + "grad_norm": 1.0515005890259814, + "learning_rate": 8.336755881638946e-06, + "loss": 0.0887, + "step": 1592 + }, + { + "epoch": 1.023121387283237, + "grad_norm": 0.7722009897870397, + "learning_rate": 8.333971604885817e-06, + "loss": 0.0555, + "step": 1593 + }, + { + "epoch": 1.0237636480411048, + "grad_norm": 0.8460542976439686, + "learning_rate": 8.331185465433e-06, + "loss": 0.065, + "step": 1594 + }, + { + "epoch": 1.0244059087989723, + "grad_norm": 0.7097591119020523, + "learning_rate": 8.328397464837119e-06, + "loss": 0.0589, + "step": 1595 + }, + { + "epoch": 1.02504816955684, + "grad_norm": 0.9688364346165554, + "learning_rate": 8.32560760465584e-06, + "loss": 0.0853, + "step": 1596 + }, + { + "epoch": 1.0256904303147079, + "grad_norm": 0.8859038105652981, + "learning_rate": 8.32281588644786e-06, + "loss": 0.0857, + "step": 1597 + }, + { + "epoch": 1.0263326910725754, + "grad_norm": 0.8865531794768287, + "learning_rate": 8.32002231177293e-06, + "loss": 0.0921, + "step": 1598 + }, + { + "epoch": 1.0269749518304432, + "grad_norm": 1.1611035571306632, + "learning_rate": 8.31722688219182e-06, + "loss": 0.1578, + "step": 1599 + }, + { + "epoch": 1.027617212588311, + "grad_norm": 0.7920034167243027, + "learning_rate": 8.314429599266348e-06, + "loss": 0.0613, + "step": 1600 + }, + { + "epoch": 1.0282594733461785, + "grad_norm": 0.9096511351524285, + "learning_rate": 8.311630464559364e-06, + "loss": 0.0916, + "step": 1601 + }, + { + "epoch": 1.0289017341040463, + "grad_norm": 0.6232037341596393, + "learning_rate": 8.308829479634753e-06, + "loss": 0.0536, + "step": 1602 + }, + { + "epoch": 1.0295439948619138, + "grad_norm": 1.10817395806942, + "learning_rate": 8.30602664605743e-06, + "loss": 0.1906, + "step": 1603 + }, + { + "epoch": 1.0301862556197816, + "grad_norm": 0.8965779728903662, + "learning_rate": 8.303221965393351e-06, + "loss": 0.0839, + "step": 1604 + }, + { + "epoch": 1.0308285163776494, + "grad_norm": 0.9520831104533575, + "learning_rate": 8.3004154392095e-06, + "loss": 0.1108, + "step": 1605 + }, + { + "epoch": 1.031470777135517, + "grad_norm": 0.9665529867978834, + "learning_rate": 8.297607069073886e-06, + "loss": 0.0906, + "step": 1606 + }, + { + "epoch": 1.0321130378933847, + "grad_norm": 0.7452428866526217, + "learning_rate": 8.294796856555557e-06, + "loss": 0.0566, + "step": 1607 + }, + { + "epoch": 1.0327552986512525, + "grad_norm": 0.8244942856370264, + "learning_rate": 8.291984803224587e-06, + "loss": 0.0668, + "step": 1608 + }, + { + "epoch": 1.03339755940912, + "grad_norm": 0.6997649628330803, + "learning_rate": 8.289170910652077e-06, + "loss": 0.0552, + "step": 1609 + }, + { + "epoch": 1.0340398201669878, + "grad_norm": 0.6738184644198982, + "learning_rate": 8.28635518041016e-06, + "loss": 0.0478, + "step": 1610 + }, + { + "epoch": 1.0346820809248556, + "grad_norm": 0.9816967830989383, + "learning_rate": 8.283537614071987e-06, + "loss": 0.0899, + "step": 1611 + }, + { + "epoch": 1.0353243416827231, + "grad_norm": 0.927790920389692, + "learning_rate": 8.280718213211746e-06, + "loss": 0.0712, + "step": 1612 + }, + { + "epoch": 1.035966602440591, + "grad_norm": 1.1658485824589961, + "learning_rate": 8.277896979404646e-06, + "loss": 0.1435, + "step": 1613 + }, + { + "epoch": 1.0366088631984587, + "grad_norm": 0.5764232468555162, + "learning_rate": 8.275073914226911e-06, + "loss": 0.0434, + "step": 1614 + }, + { + "epoch": 1.0372511239563262, + "grad_norm": 0.9878944396425638, + "learning_rate": 8.272249019255802e-06, + "loss": 0.0969, + "step": 1615 + }, + { + "epoch": 1.037893384714194, + "grad_norm": 0.7591416114981941, + "learning_rate": 8.269422296069593e-06, + "loss": 0.0606, + "step": 1616 + }, + { + "epoch": 1.0385356454720616, + "grad_norm": 0.7461188220341565, + "learning_rate": 8.266593746247586e-06, + "loss": 0.0665, + "step": 1617 + }, + { + "epoch": 1.0391779062299293, + "grad_norm": 1.0277330808730267, + "learning_rate": 8.263763371370096e-06, + "loss": 0.0906, + "step": 1618 + }, + { + "epoch": 1.0398201669877971, + "grad_norm": 0.7913824136308275, + "learning_rate": 8.260931173018462e-06, + "loss": 0.0592, + "step": 1619 + }, + { + "epoch": 1.0404624277456647, + "grad_norm": 0.9945501861775824, + "learning_rate": 8.258097152775045e-06, + "loss": 0.0965, + "step": 1620 + }, + { + "epoch": 1.0411046885035324, + "grad_norm": 0.9843828667352952, + "learning_rate": 8.255261312223218e-06, + "loss": 0.0779, + "step": 1621 + }, + { + "epoch": 1.0417469492614002, + "grad_norm": 1.1699642309962293, + "learning_rate": 8.252423652947374e-06, + "loss": 0.0997, + "step": 1622 + }, + { + "epoch": 1.0423892100192678, + "grad_norm": 0.6233568947269924, + "learning_rate": 8.24958417653292e-06, + "loss": 0.049, + "step": 1623 + }, + { + "epoch": 1.0430314707771355, + "grad_norm": 0.8153907609026799, + "learning_rate": 8.246742884566282e-06, + "loss": 0.0583, + "step": 1624 + }, + { + "epoch": 1.0436737315350033, + "grad_norm": 0.7291668168959851, + "learning_rate": 8.243899778634898e-06, + "loss": 0.0583, + "step": 1625 + }, + { + "epoch": 1.0443159922928709, + "grad_norm": 0.7327533316376731, + "learning_rate": 8.241054860327216e-06, + "loss": 0.0583, + "step": 1626 + }, + { + "epoch": 1.0449582530507386, + "grad_norm": 0.9844433379835259, + "learning_rate": 8.238208131232705e-06, + "loss": 0.0988, + "step": 1627 + }, + { + "epoch": 1.0456005138086062, + "grad_norm": 0.9194927548485805, + "learning_rate": 8.235359592941839e-06, + "loss": 0.061, + "step": 1628 + }, + { + "epoch": 1.046242774566474, + "grad_norm": 0.9417099065798842, + "learning_rate": 8.232509247046106e-06, + "loss": 0.0928, + "step": 1629 + }, + { + "epoch": 1.0468850353243417, + "grad_norm": 1.1448364620257103, + "learning_rate": 8.229657095137998e-06, + "loss": 0.1053, + "step": 1630 + }, + { + "epoch": 1.0475272960822093, + "grad_norm": 0.7320118867566875, + "learning_rate": 8.226803138811025e-06, + "loss": 0.0521, + "step": 1631 + }, + { + "epoch": 1.048169556840077, + "grad_norm": 1.0286390581027092, + "learning_rate": 8.2239473796597e-06, + "loss": 0.0947, + "step": 1632 + }, + { + "epoch": 1.0488118175979448, + "grad_norm": 0.9410436719197245, + "learning_rate": 8.221089819279543e-06, + "loss": 0.0916, + "step": 1633 + }, + { + "epoch": 1.0494540783558124, + "grad_norm": 0.7893802720663737, + "learning_rate": 8.218230459267082e-06, + "loss": 0.0612, + "step": 1634 + }, + { + "epoch": 1.0500963391136802, + "grad_norm": 0.9847366868783571, + "learning_rate": 8.215369301219849e-06, + "loss": 0.1209, + "step": 1635 + }, + { + "epoch": 1.050738599871548, + "grad_norm": 1.317198449944792, + "learning_rate": 8.212506346736382e-06, + "loss": 0.1517, + "step": 1636 + }, + { + "epoch": 1.0513808606294155, + "grad_norm": 0.9062704365431127, + "learning_rate": 8.209641597416223e-06, + "loss": 0.1009, + "step": 1637 + }, + { + "epoch": 1.0520231213872833, + "grad_norm": 1.105794363123062, + "learning_rate": 8.206775054859914e-06, + "loss": 0.1133, + "step": 1638 + }, + { + "epoch": 1.0526653821451508, + "grad_norm": 0.8196196562174829, + "learning_rate": 8.203906720669002e-06, + "loss": 0.0555, + "step": 1639 + }, + { + "epoch": 1.0533076429030186, + "grad_norm": 0.8818979060051427, + "learning_rate": 8.201036596446034e-06, + "loss": 0.0522, + "step": 1640 + }, + { + "epoch": 1.0539499036608864, + "grad_norm": 1.1497200897920536, + "learning_rate": 8.198164683794555e-06, + "loss": 0.135, + "step": 1641 + }, + { + "epoch": 1.054592164418754, + "grad_norm": 0.9064235999671275, + "learning_rate": 8.195290984319112e-06, + "loss": 0.0733, + "step": 1642 + }, + { + "epoch": 1.0552344251766217, + "grad_norm": 1.0981539940217846, + "learning_rate": 8.192415499625251e-06, + "loss": 0.1265, + "step": 1643 + }, + { + "epoch": 1.0558766859344895, + "grad_norm": 0.945108772877543, + "learning_rate": 8.18953823131951e-06, + "loss": 0.0839, + "step": 1644 + }, + { + "epoch": 1.056518946692357, + "grad_norm": 0.7540443129792732, + "learning_rate": 8.186659181009432e-06, + "loss": 0.0561, + "step": 1645 + }, + { + "epoch": 1.0571612074502248, + "grad_norm": 0.8519654279750808, + "learning_rate": 8.183778350303544e-06, + "loss": 0.103, + "step": 1646 + }, + { + "epoch": 1.0578034682080926, + "grad_norm": 0.6473100835025298, + "learning_rate": 8.180895740811381e-06, + "loss": 0.0367, + "step": 1647 + }, + { + "epoch": 1.0584457289659601, + "grad_norm": 0.7716428529780791, + "learning_rate": 8.178011354143463e-06, + "loss": 0.0503, + "step": 1648 + }, + { + "epoch": 1.059087989723828, + "grad_norm": 0.8933913885322488, + "learning_rate": 8.175125191911307e-06, + "loss": 0.0667, + "step": 1649 + }, + { + "epoch": 1.0597302504816957, + "grad_norm": 0.6859647493656814, + "learning_rate": 8.172237255727416e-06, + "loss": 0.0578, + "step": 1650 + }, + { + "epoch": 1.0603725112395632, + "grad_norm": 1.00642638150843, + "learning_rate": 8.16934754720529e-06, + "loss": 0.1102, + "step": 1651 + }, + { + "epoch": 1.061014771997431, + "grad_norm": 0.7964036438339401, + "learning_rate": 8.166456067959422e-06, + "loss": 0.0576, + "step": 1652 + }, + { + "epoch": 1.0616570327552985, + "grad_norm": 1.0356213762880595, + "learning_rate": 8.163562819605285e-06, + "loss": 0.1131, + "step": 1653 + }, + { + "epoch": 1.0622992935131663, + "grad_norm": 0.8620818418937429, + "learning_rate": 8.160667803759346e-06, + "loss": 0.0743, + "step": 1654 + }, + { + "epoch": 1.062941554271034, + "grad_norm": 0.7210469298307979, + "learning_rate": 8.157771022039061e-06, + "loss": 0.0551, + "step": 1655 + }, + { + "epoch": 1.0635838150289016, + "grad_norm": 1.049253781027747, + "learning_rate": 8.154872476062868e-06, + "loss": 0.133, + "step": 1656 + }, + { + "epoch": 1.0642260757867694, + "grad_norm": 0.7372357968571619, + "learning_rate": 8.151972167450195e-06, + "loss": 0.056, + "step": 1657 + }, + { + "epoch": 1.0648683365446372, + "grad_norm": 0.9317802427930875, + "learning_rate": 8.149070097821452e-06, + "loss": 0.0991, + "step": 1658 + }, + { + "epoch": 1.0655105973025047, + "grad_norm": 0.6327309216744317, + "learning_rate": 8.146166268798032e-06, + "loss": 0.0367, + "step": 1659 + }, + { + "epoch": 1.0661528580603725, + "grad_norm": 1.0901492275839868, + "learning_rate": 8.143260682002317e-06, + "loss": 0.1273, + "step": 1660 + }, + { + "epoch": 1.0667951188182403, + "grad_norm": 1.0042930468201092, + "learning_rate": 8.140353339057665e-06, + "loss": 0.1295, + "step": 1661 + }, + { + "epoch": 1.0674373795761078, + "grad_norm": 0.8427839934212317, + "learning_rate": 8.137444241588414e-06, + "loss": 0.0784, + "step": 1662 + }, + { + "epoch": 1.0680796403339756, + "grad_norm": 1.1659644676338952, + "learning_rate": 8.13453339121989e-06, + "loss": 0.1688, + "step": 1663 + }, + { + "epoch": 1.0687219010918434, + "grad_norm": 0.6838778269199262, + "learning_rate": 8.131620789578389e-06, + "loss": 0.0445, + "step": 1664 + }, + { + "epoch": 1.069364161849711, + "grad_norm": 1.0271364161524965, + "learning_rate": 8.128706438291193e-06, + "loss": 0.1105, + "step": 1665 + }, + { + "epoch": 1.0700064226075787, + "grad_norm": 0.8339178488619832, + "learning_rate": 8.12579033898656e-06, + "loss": 0.0657, + "step": 1666 + }, + { + "epoch": 1.0706486833654463, + "grad_norm": 0.979980491226989, + "learning_rate": 8.122872493293718e-06, + "loss": 0.0982, + "step": 1667 + }, + { + "epoch": 1.071290944123314, + "grad_norm": 0.7775147727609594, + "learning_rate": 8.119952902842883e-06, + "loss": 0.0742, + "step": 1668 + }, + { + "epoch": 1.0719332048811818, + "grad_norm": 0.7919396159527546, + "learning_rate": 8.117031569265233e-06, + "loss": 0.0297, + "step": 1669 + }, + { + "epoch": 1.0725754656390494, + "grad_norm": 1.0639623234074336, + "learning_rate": 8.114108494192927e-06, + "loss": 0.0966, + "step": 1670 + }, + { + "epoch": 1.0732177263969171, + "grad_norm": 0.9860127838296883, + "learning_rate": 8.111183679259097e-06, + "loss": 0.0829, + "step": 1671 + }, + { + "epoch": 1.073859987154785, + "grad_norm": 0.9645512137707806, + "learning_rate": 8.108257126097843e-06, + "loss": 0.0904, + "step": 1672 + }, + { + "epoch": 1.0745022479126525, + "grad_norm": 0.9915876560287973, + "learning_rate": 8.105328836344244e-06, + "loss": 0.1109, + "step": 1673 + }, + { + "epoch": 1.0751445086705202, + "grad_norm": 1.110687044867034, + "learning_rate": 8.102398811634338e-06, + "loss": 0.1476, + "step": 1674 + }, + { + "epoch": 1.075786769428388, + "grad_norm": 0.8112658872921406, + "learning_rate": 8.09946705360514e-06, + "loss": 0.0822, + "step": 1675 + }, + { + "epoch": 1.0764290301862556, + "grad_norm": 1.0080722709952104, + "learning_rate": 8.096533563894636e-06, + "loss": 0.1117, + "step": 1676 + }, + { + "epoch": 1.0770712909441233, + "grad_norm": 0.7825368660421339, + "learning_rate": 8.09359834414177e-06, + "loss": 0.0615, + "step": 1677 + }, + { + "epoch": 1.077713551701991, + "grad_norm": 0.7904270248618978, + "learning_rate": 8.090661395986458e-06, + "loss": 0.0507, + "step": 1678 + }, + { + "epoch": 1.0783558124598587, + "grad_norm": 0.7714971225213003, + "learning_rate": 8.087722721069586e-06, + "loss": 0.0443, + "step": 1679 + }, + { + "epoch": 1.0789980732177264, + "grad_norm": 0.8656229704866235, + "learning_rate": 8.084782321032999e-06, + "loss": 0.0757, + "step": 1680 + }, + { + "epoch": 1.079640333975594, + "grad_norm": 0.6615579796279215, + "learning_rate": 8.081840197519502e-06, + "loss": 0.0413, + "step": 1681 + }, + { + "epoch": 1.0802825947334618, + "grad_norm": 1.035832234415695, + "learning_rate": 8.078896352172873e-06, + "loss": 0.0918, + "step": 1682 + }, + { + "epoch": 1.0809248554913296, + "grad_norm": 0.8420638092229705, + "learning_rate": 8.075950786637847e-06, + "loss": 0.0766, + "step": 1683 + }, + { + "epoch": 1.081567116249197, + "grad_norm": 1.3437124853943387, + "learning_rate": 8.07300350256012e-06, + "loss": 0.1104, + "step": 1684 + }, + { + "epoch": 1.0822093770070649, + "grad_norm": 0.8745168658319908, + "learning_rate": 8.070054501586345e-06, + "loss": 0.0599, + "step": 1685 + }, + { + "epoch": 1.0828516377649327, + "grad_norm": 1.1554943584544277, + "learning_rate": 8.06710378536414e-06, + "loss": 0.0747, + "step": 1686 + }, + { + "epoch": 1.0834938985228002, + "grad_norm": 0.69380945382017, + "learning_rate": 8.064151355542079e-06, + "loss": 0.0313, + "step": 1687 + }, + { + "epoch": 1.084136159280668, + "grad_norm": 0.9252193594612751, + "learning_rate": 8.061197213769693e-06, + "loss": 0.0704, + "step": 1688 + }, + { + "epoch": 1.0847784200385355, + "grad_norm": 1.2390711733508364, + "learning_rate": 8.058241361697467e-06, + "loss": 0.1669, + "step": 1689 + }, + { + "epoch": 1.0854206807964033, + "grad_norm": 0.8763232312744421, + "learning_rate": 8.055283800976848e-06, + "loss": 0.0786, + "step": 1690 + }, + { + "epoch": 1.086062941554271, + "grad_norm": 0.9477155588979019, + "learning_rate": 8.052324533260232e-06, + "loss": 0.091, + "step": 1691 + }, + { + "epoch": 1.0867052023121386, + "grad_norm": 1.0012976304821042, + "learning_rate": 8.049363560200972e-06, + "loss": 0.0904, + "step": 1692 + }, + { + "epoch": 1.0873474630700064, + "grad_norm": 1.2607489489449983, + "learning_rate": 8.04640088345337e-06, + "loss": 0.1781, + "step": 1693 + }, + { + "epoch": 1.0879897238278742, + "grad_norm": 1.1689420370879888, + "learning_rate": 8.043436504672684e-06, + "loss": 0.141, + "step": 1694 + }, + { + "epoch": 1.0886319845857417, + "grad_norm": 1.1332704706729728, + "learning_rate": 8.040470425515119e-06, + "loss": 0.1069, + "step": 1695 + }, + { + "epoch": 1.0892742453436095, + "grad_norm": 0.7490032418521179, + "learning_rate": 8.037502647637835e-06, + "loss": 0.0521, + "step": 1696 + }, + { + "epoch": 1.0899165061014773, + "grad_norm": 0.9066433753524058, + "learning_rate": 8.034533172698935e-06, + "loss": 0.0685, + "step": 1697 + }, + { + "epoch": 1.0905587668593448, + "grad_norm": 0.6352673978293079, + "learning_rate": 8.031562002357476e-06, + "loss": 0.0361, + "step": 1698 + }, + { + "epoch": 1.0912010276172126, + "grad_norm": 1.0645319931019421, + "learning_rate": 8.028589138273458e-06, + "loss": 0.1187, + "step": 1699 + }, + { + "epoch": 1.0918432883750804, + "grad_norm": 1.1053435632829567, + "learning_rate": 8.02561458210783e-06, + "loss": 0.1521, + "step": 1700 + }, + { + "epoch": 1.092485549132948, + "grad_norm": 0.5941362856184945, + "learning_rate": 8.022638335522484e-06, + "loss": 0.032, + "step": 1701 + }, + { + "epoch": 1.0931278098908157, + "grad_norm": 1.0656880713404009, + "learning_rate": 8.019660400180258e-06, + "loss": 0.1007, + "step": 1702 + }, + { + "epoch": 1.0937700706486835, + "grad_norm": 0.6425453441453549, + "learning_rate": 8.016680777744935e-06, + "loss": 0.0367, + "step": 1703 + }, + { + "epoch": 1.094412331406551, + "grad_norm": 1.0918650966681016, + "learning_rate": 8.013699469881236e-06, + "loss": 0.0909, + "step": 1704 + }, + { + "epoch": 1.0950545921644188, + "grad_norm": 0.8437352797841428, + "learning_rate": 8.010716478254827e-06, + "loss": 0.0679, + "step": 1705 + }, + { + "epoch": 1.0956968529222864, + "grad_norm": 0.8764579534167968, + "learning_rate": 8.007731804532316e-06, + "loss": 0.0974, + "step": 1706 + }, + { + "epoch": 1.0963391136801541, + "grad_norm": 0.749208785458932, + "learning_rate": 8.004745450381248e-06, + "loss": 0.062, + "step": 1707 + }, + { + "epoch": 1.096981374438022, + "grad_norm": 0.8315137641458129, + "learning_rate": 8.001757417470106e-06, + "loss": 0.0518, + "step": 1708 + }, + { + "epoch": 1.0976236351958895, + "grad_norm": 0.6068027654964601, + "learning_rate": 7.998767707468316e-06, + "loss": 0.0341, + "step": 1709 + }, + { + "epoch": 1.0982658959537572, + "grad_norm": 0.9092913550547734, + "learning_rate": 7.995776322046236e-06, + "loss": 0.0725, + "step": 1710 + }, + { + "epoch": 1.098908156711625, + "grad_norm": 0.9324314643246812, + "learning_rate": 7.992783262875162e-06, + "loss": 0.0869, + "step": 1711 + }, + { + "epoch": 1.0995504174694926, + "grad_norm": 1.5403990000945962, + "learning_rate": 7.989788531627322e-06, + "loss": 0.211, + "step": 1712 + }, + { + "epoch": 1.1001926782273603, + "grad_norm": 0.9477454022671199, + "learning_rate": 7.986792129975887e-06, + "loss": 0.1109, + "step": 1713 + }, + { + "epoch": 1.100834938985228, + "grad_norm": 0.8279018273544081, + "learning_rate": 7.983794059594951e-06, + "loss": 0.0647, + "step": 1714 + }, + { + "epoch": 1.1014771997430957, + "grad_norm": 0.9085031162074043, + "learning_rate": 7.980794322159544e-06, + "loss": 0.0744, + "step": 1715 + }, + { + "epoch": 1.1021194605009634, + "grad_norm": 0.8906010958197875, + "learning_rate": 7.977792919345632e-06, + "loss": 0.0658, + "step": 1716 + }, + { + "epoch": 1.102761721258831, + "grad_norm": 1.0550082684772664, + "learning_rate": 7.974789852830105e-06, + "loss": 0.1347, + "step": 1717 + }, + { + "epoch": 1.1034039820166988, + "grad_norm": 1.3285970140125054, + "learning_rate": 7.971785124290787e-06, + "loss": 0.2257, + "step": 1718 + }, + { + "epoch": 1.1040462427745665, + "grad_norm": 1.0781839718119215, + "learning_rate": 7.968778735406426e-06, + "loss": 0.1045, + "step": 1719 + }, + { + "epoch": 1.104688503532434, + "grad_norm": 0.8634254877820572, + "learning_rate": 7.965770687856702e-06, + "loss": 0.0731, + "step": 1720 + }, + { + "epoch": 1.1053307642903019, + "grad_norm": 1.1843741292554744, + "learning_rate": 7.962760983322217e-06, + "loss": 0.1262, + "step": 1721 + }, + { + "epoch": 1.1059730250481696, + "grad_norm": 0.8423492822334274, + "learning_rate": 7.959749623484507e-06, + "loss": 0.0486, + "step": 1722 + }, + { + "epoch": 1.1066152858060372, + "grad_norm": 0.793314810009429, + "learning_rate": 7.956736610026023e-06, + "loss": 0.0736, + "step": 1723 + }, + { + "epoch": 1.107257546563905, + "grad_norm": 0.811503286809147, + "learning_rate": 7.953721944630146e-06, + "loss": 0.0684, + "step": 1724 + }, + { + "epoch": 1.1078998073217727, + "grad_norm": 1.0881693679602837, + "learning_rate": 7.95070562898118e-06, + "loss": 0.1007, + "step": 1725 + }, + { + "epoch": 1.1085420680796403, + "grad_norm": 0.4958035173027681, + "learning_rate": 7.947687664764347e-06, + "loss": 0.0222, + "step": 1726 + }, + { + "epoch": 1.109184328837508, + "grad_norm": 0.549757874184289, + "learning_rate": 7.944668053665795e-06, + "loss": 0.0428, + "step": 1727 + }, + { + "epoch": 1.1098265895953756, + "grad_norm": 0.8140750535670787, + "learning_rate": 7.941646797372584e-06, + "loss": 0.077, + "step": 1728 + }, + { + "epoch": 1.1104688503532434, + "grad_norm": 0.7461520482770923, + "learning_rate": 7.938623897572705e-06, + "loss": 0.0534, + "step": 1729 + }, + { + "epoch": 1.1111111111111112, + "grad_norm": 0.8234119152589169, + "learning_rate": 7.935599355955059e-06, + "loss": 0.057, + "step": 1730 + }, + { + "epoch": 1.1117533718689787, + "grad_norm": 0.9966157795398628, + "learning_rate": 7.932573174209463e-06, + "loss": 0.0924, + "step": 1731 + }, + { + "epoch": 1.1123956326268465, + "grad_norm": 0.7984108917766537, + "learning_rate": 7.92954535402666e-06, + "loss": 0.0585, + "step": 1732 + }, + { + "epoch": 1.1130378933847143, + "grad_norm": 0.6612841610780037, + "learning_rate": 7.926515897098295e-06, + "loss": 0.0405, + "step": 1733 + }, + { + "epoch": 1.1136801541425818, + "grad_norm": 1.1395540408356275, + "learning_rate": 7.92348480511694e-06, + "loss": 0.1409, + "step": 1734 + }, + { + "epoch": 1.1143224149004496, + "grad_norm": 0.8586507940739067, + "learning_rate": 7.920452079776074e-06, + "loss": 0.075, + "step": 1735 + }, + { + "epoch": 1.1149646756583174, + "grad_norm": 1.0018932825968316, + "learning_rate": 7.917417722770086e-06, + "loss": 0.051, + "step": 1736 + }, + { + "epoch": 1.115606936416185, + "grad_norm": 0.8363477048706425, + "learning_rate": 7.914381735794282e-06, + "loss": 0.0728, + "step": 1737 + }, + { + "epoch": 1.1162491971740527, + "grad_norm": 1.15554241576607, + "learning_rate": 7.911344120544878e-06, + "loss": 0.1068, + "step": 1738 + }, + { + "epoch": 1.1168914579319205, + "grad_norm": 0.7250332310433686, + "learning_rate": 7.908304878718999e-06, + "loss": 0.0483, + "step": 1739 + }, + { + "epoch": 1.117533718689788, + "grad_norm": 0.8161574708409931, + "learning_rate": 7.905264012014675e-06, + "loss": 0.0766, + "step": 1740 + }, + { + "epoch": 1.1181759794476558, + "grad_norm": 1.0436350855143581, + "learning_rate": 7.90222152213085e-06, + "loss": 0.0854, + "step": 1741 + }, + { + "epoch": 1.1188182402055233, + "grad_norm": 0.7851945214788396, + "learning_rate": 7.899177410767372e-06, + "loss": 0.0672, + "step": 1742 + }, + { + "epoch": 1.1194605009633911, + "grad_norm": 1.0798399608162743, + "learning_rate": 7.896131679624995e-06, + "loss": 0.1054, + "step": 1743 + }, + { + "epoch": 1.1201027617212589, + "grad_norm": 2.8047317189216745, + "learning_rate": 7.893084330405374e-06, + "loss": 0.0991, + "step": 1744 + }, + { + "epoch": 1.1207450224791264, + "grad_norm": 1.050884612634703, + "learning_rate": 7.890035364811076e-06, + "loss": 0.0751, + "step": 1745 + }, + { + "epoch": 1.1213872832369942, + "grad_norm": 0.6164409610473328, + "learning_rate": 7.886984784545565e-06, + "loss": 0.0494, + "step": 1746 + }, + { + "epoch": 1.122029543994862, + "grad_norm": 3.6081075999626875, + "learning_rate": 7.883932591313213e-06, + "loss": 0.0547, + "step": 1747 + }, + { + "epoch": 1.1226718047527295, + "grad_norm": 0.8554460175746543, + "learning_rate": 7.880878786819283e-06, + "loss": 0.0835, + "step": 1748 + }, + { + "epoch": 1.1233140655105973, + "grad_norm": 0.6425708596240447, + "learning_rate": 7.877823372769951e-06, + "loss": 0.0426, + "step": 1749 + }, + { + "epoch": 1.123956326268465, + "grad_norm": 1.0555309384942348, + "learning_rate": 7.87476635087228e-06, + "loss": 0.1019, + "step": 1750 + }, + { + "epoch": 1.1245985870263326, + "grad_norm": 1.0187086108424273, + "learning_rate": 7.871707722834241e-06, + "loss": 0.125, + "step": 1751 + }, + { + "epoch": 1.1252408477842004, + "grad_norm": 0.865770285106466, + "learning_rate": 7.868647490364696e-06, + "loss": 0.0944, + "step": 1752 + }, + { + "epoch": 1.1258831085420682, + "grad_norm": 0.8786859599604007, + "learning_rate": 7.865585655173409e-06, + "loss": 0.0725, + "step": 1753 + }, + { + "epoch": 1.1265253692999357, + "grad_norm": 0.9328824526085445, + "learning_rate": 7.862522218971032e-06, + "loss": 0.0636, + "step": 1754 + }, + { + "epoch": 1.1271676300578035, + "grad_norm": 0.750708482607536, + "learning_rate": 7.859457183469119e-06, + "loss": 0.0737, + "step": 1755 + }, + { + "epoch": 1.127809890815671, + "grad_norm": 0.7007417706857112, + "learning_rate": 7.856390550380112e-06, + "loss": 0.0385, + "step": 1756 + }, + { + "epoch": 1.1284521515735388, + "grad_norm": 0.7079750048330997, + "learning_rate": 7.853322321417352e-06, + "loss": 0.0566, + "step": 1757 + }, + { + "epoch": 1.1290944123314066, + "grad_norm": 0.8469136393897354, + "learning_rate": 7.850252498295062e-06, + "loss": 0.0897, + "step": 1758 + }, + { + "epoch": 1.1297366730892742, + "grad_norm": 1.1687680621460212, + "learning_rate": 7.847181082728366e-06, + "loss": 0.1329, + "step": 1759 + }, + { + "epoch": 1.130378933847142, + "grad_norm": 0.742543617745753, + "learning_rate": 7.844108076433268e-06, + "loss": 0.0595, + "step": 1760 + }, + { + "epoch": 1.1310211946050097, + "grad_norm": 0.6263137009037589, + "learning_rate": 7.84103348112667e-06, + "loss": 0.0397, + "step": 1761 + }, + { + "epoch": 1.1316634553628773, + "grad_norm": 0.8295528354347866, + "learning_rate": 7.837957298526355e-06, + "loss": 0.0691, + "step": 1762 + }, + { + "epoch": 1.132305716120745, + "grad_norm": 1.0021801956210092, + "learning_rate": 7.834879530350997e-06, + "loss": 0.1246, + "step": 1763 + }, + { + "epoch": 1.1329479768786128, + "grad_norm": 0.9854546636575826, + "learning_rate": 7.831800178320153e-06, + "loss": 0.1181, + "step": 1764 + }, + { + "epoch": 1.1335902376364804, + "grad_norm": 0.7304061799201995, + "learning_rate": 7.828719244154264e-06, + "loss": 0.0512, + "step": 1765 + }, + { + "epoch": 1.1342324983943481, + "grad_norm": 1.2924264464484665, + "learning_rate": 7.82563672957466e-06, + "loss": 0.1382, + "step": 1766 + }, + { + "epoch": 1.1348747591522157, + "grad_norm": 0.6327038720885589, + "learning_rate": 7.822552636303551e-06, + "loss": 0.047, + "step": 1767 + }, + { + "epoch": 1.1355170199100835, + "grad_norm": 2.2013849971436303, + "learning_rate": 7.819466966064026e-06, + "loss": 0.0986, + "step": 1768 + }, + { + "epoch": 1.1361592806679512, + "grad_norm": 0.8517040699964673, + "learning_rate": 7.81637972058006e-06, + "loss": 0.0892, + "step": 1769 + }, + { + "epoch": 1.1368015414258188, + "grad_norm": 1.0025433170814737, + "learning_rate": 7.813290901576505e-06, + "loss": 0.0806, + "step": 1770 + }, + { + "epoch": 1.1374438021836866, + "grad_norm": 0.6358681928475501, + "learning_rate": 7.810200510779096e-06, + "loss": 0.0482, + "step": 1771 + }, + { + "epoch": 1.1380860629415543, + "grad_norm": 0.8476768541482775, + "learning_rate": 7.80710854991444e-06, + "loss": 0.068, + "step": 1772 + }, + { + "epoch": 1.138728323699422, + "grad_norm": 0.869007601730534, + "learning_rate": 7.804015020710028e-06, + "loss": 0.0799, + "step": 1773 + }, + { + "epoch": 1.1393705844572897, + "grad_norm": 1.0710427804392426, + "learning_rate": 7.800919924894219e-06, + "loss": 0.1069, + "step": 1774 + }, + { + "epoch": 1.1400128452151574, + "grad_norm": 0.9538960634318889, + "learning_rate": 7.797823264196256e-06, + "loss": 0.0711, + "step": 1775 + }, + { + "epoch": 1.140655105973025, + "grad_norm": 0.9191236152284031, + "learning_rate": 7.794725040346251e-06, + "loss": 0.0797, + "step": 1776 + }, + { + "epoch": 1.1412973667308928, + "grad_norm": 0.9762214153217911, + "learning_rate": 7.791625255075192e-06, + "loss": 0.0894, + "step": 1777 + }, + { + "epoch": 1.1419396274887603, + "grad_norm": 0.9221600114285532, + "learning_rate": 7.788523910114936e-06, + "loss": 0.0779, + "step": 1778 + }, + { + "epoch": 1.142581888246628, + "grad_norm": 1.0304690317633893, + "learning_rate": 7.785421007198216e-06, + "loss": 0.107, + "step": 1779 + }, + { + "epoch": 1.1432241490044959, + "grad_norm": 0.9144372956586918, + "learning_rate": 7.78231654805863e-06, + "loss": 0.0996, + "step": 1780 + }, + { + "epoch": 1.1438664097623634, + "grad_norm": 1.0018686930417724, + "learning_rate": 7.77921053443065e-06, + "loss": 0.0874, + "step": 1781 + }, + { + "epoch": 1.1445086705202312, + "grad_norm": 1.0244078268680723, + "learning_rate": 7.776102968049616e-06, + "loss": 0.1237, + "step": 1782 + }, + { + "epoch": 1.145150931278099, + "grad_norm": 1.023769535193666, + "learning_rate": 7.772993850651732e-06, + "loss": 0.083, + "step": 1783 + }, + { + "epoch": 1.1457931920359665, + "grad_norm": 0.8898152971233237, + "learning_rate": 7.769883183974072e-06, + "loss": 0.0646, + "step": 1784 + }, + { + "epoch": 1.1464354527938343, + "grad_norm": 0.7531420248034049, + "learning_rate": 7.766770969754575e-06, + "loss": 0.0559, + "step": 1785 + }, + { + "epoch": 1.147077713551702, + "grad_norm": 1.1643705136483513, + "learning_rate": 7.763657209732047e-06, + "loss": 0.1248, + "step": 1786 + }, + { + "epoch": 1.1477199743095696, + "grad_norm": 1.2217737423727366, + "learning_rate": 7.760541905646148e-06, + "loss": 0.1469, + "step": 1787 + }, + { + "epoch": 1.1483622350674374, + "grad_norm": 0.8299658122959415, + "learning_rate": 7.757425059237413e-06, + "loss": 0.074, + "step": 1788 + }, + { + "epoch": 1.149004495825305, + "grad_norm": 1.2748419104940807, + "learning_rate": 7.754306672247232e-06, + "loss": 0.1559, + "step": 1789 + }, + { + "epoch": 1.1496467565831727, + "grad_norm": 0.5985923601945059, + "learning_rate": 7.751186746417855e-06, + "loss": 0.0437, + "step": 1790 + }, + { + "epoch": 1.1502890173410405, + "grad_norm": 0.8740419198800066, + "learning_rate": 7.748065283492397e-06, + "loss": 0.1005, + "step": 1791 + }, + { + "epoch": 1.1509312780989083, + "grad_norm": 0.5395407708120417, + "learning_rate": 7.744942285214825e-06, + "loss": 0.0393, + "step": 1792 + }, + { + "epoch": 1.1515735388567758, + "grad_norm": 0.8383141071241305, + "learning_rate": 7.741817753329967e-06, + "loss": 0.0674, + "step": 1793 + }, + { + "epoch": 1.1522157996146436, + "grad_norm": 1.0384856810557186, + "learning_rate": 7.738691689583513e-06, + "loss": 0.0804, + "step": 1794 + }, + { + "epoch": 1.1528580603725112, + "grad_norm": 1.2248161505310835, + "learning_rate": 7.735564095722002e-06, + "loss": 0.0812, + "step": 1795 + }, + { + "epoch": 1.153500321130379, + "grad_norm": 0.7792596658261444, + "learning_rate": 7.732434973492825e-06, + "loss": 0.0732, + "step": 1796 + }, + { + "epoch": 1.1541425818882467, + "grad_norm": 1.3248954977878633, + "learning_rate": 7.729304324644239e-06, + "loss": 0.1681, + "step": 1797 + }, + { + "epoch": 1.1547848426461143, + "grad_norm": 0.6018294720296186, + "learning_rate": 7.726172150925341e-06, + "loss": 0.0402, + "step": 1798 + }, + { + "epoch": 1.155427103403982, + "grad_norm": 0.9138428676579793, + "learning_rate": 7.723038454086089e-06, + "loss": 0.101, + "step": 1799 + }, + { + "epoch": 1.1560693641618498, + "grad_norm": 0.8731165975032444, + "learning_rate": 7.719903235877289e-06, + "loss": 0.0705, + "step": 1800 + }, + { + "epoch": 1.1567116249197174, + "grad_norm": 1.0159645541024613, + "learning_rate": 7.716766498050591e-06, + "loss": 0.1043, + "step": 1801 + }, + { + "epoch": 1.1573538856775851, + "grad_norm": 0.9439306985543571, + "learning_rate": 7.713628242358507e-06, + "loss": 0.0894, + "step": 1802 + }, + { + "epoch": 1.157996146435453, + "grad_norm": 1.0357266282254027, + "learning_rate": 7.710488470554384e-06, + "loss": 0.0987, + "step": 1803 + }, + { + "epoch": 1.1586384071933205, + "grad_norm": 1.133958610798814, + "learning_rate": 7.707347184392427e-06, + "loss": 0.1093, + "step": 1804 + }, + { + "epoch": 1.1592806679511882, + "grad_norm": 0.8428266232825568, + "learning_rate": 7.704204385627674e-06, + "loss": 0.0605, + "step": 1805 + }, + { + "epoch": 1.1599229287090558, + "grad_norm": 0.752595811868998, + "learning_rate": 7.701060076016025e-06, + "loss": 0.0404, + "step": 1806 + }, + { + "epoch": 1.1605651894669236, + "grad_norm": 0.9568667807505531, + "learning_rate": 7.697914257314208e-06, + "loss": 0.0736, + "step": 1807 + }, + { + "epoch": 1.1612074502247913, + "grad_norm": 0.6339027990196642, + "learning_rate": 7.694766931279805e-06, + "loss": 0.0519, + "step": 1808 + }, + { + "epoch": 1.1618497109826589, + "grad_norm": 0.9860916601761012, + "learning_rate": 7.691618099671235e-06, + "loss": 0.0965, + "step": 1809 + }, + { + "epoch": 1.1624919717405267, + "grad_norm": 0.7664271152543983, + "learning_rate": 7.688467764247758e-06, + "loss": 0.0471, + "step": 1810 + }, + { + "epoch": 1.1631342324983944, + "grad_norm": 0.8448595399170072, + "learning_rate": 7.685315926769476e-06, + "loss": 0.0699, + "step": 1811 + }, + { + "epoch": 1.163776493256262, + "grad_norm": 0.8048405249465317, + "learning_rate": 7.682162588997333e-06, + "loss": 0.0673, + "step": 1812 + }, + { + "epoch": 1.1644187540141298, + "grad_norm": 0.9676909082066097, + "learning_rate": 7.679007752693106e-06, + "loss": 0.072, + "step": 1813 + }, + { + "epoch": 1.1650610147719975, + "grad_norm": 1.3528397043542524, + "learning_rate": 7.67585141961941e-06, + "loss": 0.0758, + "step": 1814 + }, + { + "epoch": 1.165703275529865, + "grad_norm": 0.949812215605421, + "learning_rate": 7.6726935915397e-06, + "loss": 0.1117, + "step": 1815 + }, + { + "epoch": 1.1663455362877329, + "grad_norm": 0.7272133005251402, + "learning_rate": 7.669534270218262e-06, + "loss": 0.0667, + "step": 1816 + }, + { + "epoch": 1.1669877970456004, + "grad_norm": 0.877021800222529, + "learning_rate": 7.666373457420217e-06, + "loss": 0.0942, + "step": 1817 + }, + { + "epoch": 1.1676300578034682, + "grad_norm": 0.6697078218418884, + "learning_rate": 7.663211154911523e-06, + "loss": 0.0498, + "step": 1818 + }, + { + "epoch": 1.168272318561336, + "grad_norm": 1.0891795486097446, + "learning_rate": 7.660047364458965e-06, + "loss": 0.0967, + "step": 1819 + }, + { + "epoch": 1.1689145793192035, + "grad_norm": 0.77409057320584, + "learning_rate": 7.656882087830165e-06, + "loss": 0.0737, + "step": 1820 + }, + { + "epoch": 1.1695568400770713, + "grad_norm": 0.8018288660892492, + "learning_rate": 7.65371532679357e-06, + "loss": 0.0649, + "step": 1821 + }, + { + "epoch": 1.170199100834939, + "grad_norm": 0.8832494174985813, + "learning_rate": 7.65054708311846e-06, + "loss": 0.0768, + "step": 1822 + }, + { + "epoch": 1.1708413615928066, + "grad_norm": 0.9841236316641001, + "learning_rate": 7.647377358574941e-06, + "loss": 0.0952, + "step": 1823 + }, + { + "epoch": 1.1714836223506744, + "grad_norm": 1.051414740974429, + "learning_rate": 7.644206154933948e-06, + "loss": 0.1104, + "step": 1824 + }, + { + "epoch": 1.1721258831085422, + "grad_norm": 1.144838438404055, + "learning_rate": 7.64103347396724e-06, + "loss": 0.0856, + "step": 1825 + }, + { + "epoch": 1.1727681438664097, + "grad_norm": 0.8366637464498605, + "learning_rate": 7.637859317447408e-06, + "loss": 0.0738, + "step": 1826 + }, + { + "epoch": 1.1734104046242775, + "grad_norm": 0.9417241922441494, + "learning_rate": 7.634683687147857e-06, + "loss": 0.0971, + "step": 1827 + }, + { + "epoch": 1.174052665382145, + "grad_norm": 0.9674552117661509, + "learning_rate": 7.631506584842822e-06, + "loss": 0.074, + "step": 1828 + }, + { + "epoch": 1.1746949261400128, + "grad_norm": 1.0252888555992203, + "learning_rate": 7.6283280123073635e-06, + "loss": 0.1201, + "step": 1829 + }, + { + "epoch": 1.1753371868978806, + "grad_norm": 0.6248984223409967, + "learning_rate": 7.625147971317355e-06, + "loss": 0.0475, + "step": 1830 + }, + { + "epoch": 1.1759794476557484, + "grad_norm": 0.9323087631474425, + "learning_rate": 7.6219664636494964e-06, + "loss": 0.0661, + "step": 1831 + }, + { + "epoch": 1.176621708413616, + "grad_norm": 0.9610776504459786, + "learning_rate": 7.618783491081304e-06, + "loss": 0.1012, + "step": 1832 + }, + { + "epoch": 1.1772639691714837, + "grad_norm": 1.1421628617009298, + "learning_rate": 7.615599055391115e-06, + "loss": 0.1391, + "step": 1833 + }, + { + "epoch": 1.1779062299293512, + "grad_norm": 0.8015459683683235, + "learning_rate": 7.612413158358083e-06, + "loss": 0.0794, + "step": 1834 + }, + { + "epoch": 1.178548490687219, + "grad_norm": 0.7334556995651812, + "learning_rate": 7.609225801762178e-06, + "loss": 0.0532, + "step": 1835 + }, + { + "epoch": 1.1791907514450868, + "grad_norm": 0.6878545265539194, + "learning_rate": 7.606036987384185e-06, + "loss": 0.0349, + "step": 1836 + }, + { + "epoch": 1.1798330122029543, + "grad_norm": 0.9665059301063568, + "learning_rate": 7.602846717005705e-06, + "loss": 0.0744, + "step": 1837 + }, + { + "epoch": 1.180475272960822, + "grad_norm": 1.359261117913379, + "learning_rate": 7.599654992409149e-06, + "loss": 0.0683, + "step": 1838 + }, + { + "epoch": 1.1811175337186899, + "grad_norm": 0.8137056325274447, + "learning_rate": 7.596461815377744e-06, + "loss": 0.0534, + "step": 1839 + }, + { + "epoch": 1.1817597944765574, + "grad_norm": 1.0366370589265115, + "learning_rate": 7.593267187695528e-06, + "loss": 0.039, + "step": 1840 + }, + { + "epoch": 1.1824020552344252, + "grad_norm": 1.188751636867195, + "learning_rate": 7.59007111114735e-06, + "loss": 0.1126, + "step": 1841 + }, + { + "epoch": 1.183044315992293, + "grad_norm": 0.8508327597390983, + "learning_rate": 7.586873587518866e-06, + "loss": 0.0716, + "step": 1842 + }, + { + "epoch": 1.1836865767501605, + "grad_norm": 0.8429074540420973, + "learning_rate": 7.583674618596541e-06, + "loss": 0.0716, + "step": 1843 + }, + { + "epoch": 1.1843288375080283, + "grad_norm": 0.7155525833551245, + "learning_rate": 7.580474206167648e-06, + "loss": 0.0489, + "step": 1844 + }, + { + "epoch": 1.1849710982658959, + "grad_norm": 0.8143485912139184, + "learning_rate": 7.577272352020269e-06, + "loss": 0.0737, + "step": 1845 + }, + { + "epoch": 1.1856133590237636, + "grad_norm": 0.8999314602590146, + "learning_rate": 7.574069057943287e-06, + "loss": 0.0757, + "step": 1846 + }, + { + "epoch": 1.1862556197816314, + "grad_norm": 1.0053725169247039, + "learning_rate": 7.5708643257263925e-06, + "loss": 0.1066, + "step": 1847 + }, + { + "epoch": 1.186897880539499, + "grad_norm": 0.9798026566351051, + "learning_rate": 7.56765815716008e-06, + "loss": 0.0858, + "step": 1848 + }, + { + "epoch": 1.1875401412973667, + "grad_norm": 0.7251043065519515, + "learning_rate": 7.564450554035645e-06, + "loss": 0.0625, + "step": 1849 + }, + { + "epoch": 1.1881824020552345, + "grad_norm": 0.783535812005694, + "learning_rate": 7.561241518145183e-06, + "loss": 0.0512, + "step": 1850 + }, + { + "epoch": 1.188824662813102, + "grad_norm": 1.0028298468613468, + "learning_rate": 7.558031051281592e-06, + "loss": 0.0995, + "step": 1851 + }, + { + "epoch": 1.1894669235709698, + "grad_norm": 1.1691378237612822, + "learning_rate": 7.55481915523857e-06, + "loss": 0.0894, + "step": 1852 + }, + { + "epoch": 1.1901091843288376, + "grad_norm": 0.7105570296281883, + "learning_rate": 7.551605831810616e-06, + "loss": 0.0521, + "step": 1853 + }, + { + "epoch": 1.1907514450867052, + "grad_norm": 1.0590492208629414, + "learning_rate": 7.5483910827930186e-06, + "loss": 0.1248, + "step": 1854 + }, + { + "epoch": 1.191393705844573, + "grad_norm": 1.08266733208137, + "learning_rate": 7.545174909981868e-06, + "loss": 0.1307, + "step": 1855 + }, + { + "epoch": 1.1920359666024405, + "grad_norm": 1.1204422914045935, + "learning_rate": 7.5419573151740486e-06, + "loss": 0.111, + "step": 1856 + }, + { + "epoch": 1.1926782273603083, + "grad_norm": 0.8003207271838995, + "learning_rate": 7.538738300167244e-06, + "loss": 0.0568, + "step": 1857 + }, + { + "epoch": 1.193320488118176, + "grad_norm": 0.8977566184868212, + "learning_rate": 7.535517866759925e-06, + "loss": 0.0578, + "step": 1858 + }, + { + "epoch": 1.1939627488760436, + "grad_norm": 1.2222197819727454, + "learning_rate": 7.5322960167513545e-06, + "loss": 0.1566, + "step": 1859 + }, + { + "epoch": 1.1946050096339114, + "grad_norm": 1.0565366157246248, + "learning_rate": 7.529072751941595e-06, + "loss": 0.1169, + "step": 1860 + }, + { + "epoch": 1.1952472703917791, + "grad_norm": 0.7953888050157238, + "learning_rate": 7.52584807413149e-06, + "loss": 0.0762, + "step": 1861 + }, + { + "epoch": 1.1958895311496467, + "grad_norm": 1.008434223090326, + "learning_rate": 7.522621985122678e-06, + "loss": 0.1184, + "step": 1862 + }, + { + "epoch": 1.1965317919075145, + "grad_norm": 1.0494102545488513, + "learning_rate": 7.519394486717583e-06, + "loss": 0.1523, + "step": 1863 + }, + { + "epoch": 1.1971740526653822, + "grad_norm": 0.9829531427275998, + "learning_rate": 7.516165580719419e-06, + "loss": 0.1121, + "step": 1864 + }, + { + "epoch": 1.1978163134232498, + "grad_norm": 1.139247355157383, + "learning_rate": 7.512935268932185e-06, + "loss": 0.1401, + "step": 1865 + }, + { + "epoch": 1.1984585741811176, + "grad_norm": 0.7804815175451997, + "learning_rate": 7.509703553160665e-06, + "loss": 0.0676, + "step": 1866 + }, + { + "epoch": 1.1991008349389851, + "grad_norm": 1.0521010314999584, + "learning_rate": 7.506470435210429e-06, + "loss": 0.109, + "step": 1867 + }, + { + "epoch": 1.199743095696853, + "grad_norm": 0.7299536782452448, + "learning_rate": 7.50323591688783e-06, + "loss": 0.0921, + "step": 1868 + }, + { + "epoch": 1.2003853564547207, + "grad_norm": 0.933470366380917, + "learning_rate": 7.500000000000001e-06, + "loss": 0.0833, + "step": 1869 + }, + { + "epoch": 1.2010276172125882, + "grad_norm": 1.0152602710631011, + "learning_rate": 7.496762686354859e-06, + "loss": 0.104, + "step": 1870 + }, + { + "epoch": 1.201669877970456, + "grad_norm": 0.9949026930802316, + "learning_rate": 7.493523977761101e-06, + "loss": 0.1196, + "step": 1871 + }, + { + "epoch": 1.2023121387283238, + "grad_norm": 1.0035555978414856, + "learning_rate": 7.4902838760282024e-06, + "loss": 0.102, + "step": 1872 + }, + { + "epoch": 1.2029543994861913, + "grad_norm": 0.9942214320089614, + "learning_rate": 7.487042382966418e-06, + "loss": 0.1148, + "step": 1873 + }, + { + "epoch": 1.203596660244059, + "grad_norm": 0.9995978115219636, + "learning_rate": 7.4837995003867795e-06, + "loss": 0.0867, + "step": 1874 + }, + { + "epoch": 1.2042389210019269, + "grad_norm": 0.959333521604008, + "learning_rate": 7.480555230101094e-06, + "loss": 0.0773, + "step": 1875 + }, + { + "epoch": 1.2048811817597944, + "grad_norm": 0.8550861952842208, + "learning_rate": 7.477309573921947e-06, + "loss": 0.0481, + "step": 1876 + }, + { + "epoch": 1.2055234425176622, + "grad_norm": 0.8984826262719301, + "learning_rate": 7.474062533662694e-06, + "loss": 0.061, + "step": 1877 + }, + { + "epoch": 1.2061657032755297, + "grad_norm": 0.9358423919413767, + "learning_rate": 7.4708141111374656e-06, + "loss": 0.0991, + "step": 1878 + }, + { + "epoch": 1.2068079640333975, + "grad_norm": 1.1192731914400071, + "learning_rate": 7.467564308161166e-06, + "loss": 0.1039, + "step": 1879 + }, + { + "epoch": 1.2074502247912653, + "grad_norm": 0.6550228616893234, + "learning_rate": 7.46431312654947e-06, + "loss": 0.0477, + "step": 1880 + }, + { + "epoch": 1.208092485549133, + "grad_norm": 1.1203488165285576, + "learning_rate": 7.461060568118822e-06, + "loss": 0.0625, + "step": 1881 + }, + { + "epoch": 1.2087347463070006, + "grad_norm": 0.8583195140303844, + "learning_rate": 7.457806634686436e-06, + "loss": 0.0599, + "step": 1882 + }, + { + "epoch": 1.2093770070648684, + "grad_norm": 0.8293883623685259, + "learning_rate": 7.454551328070292e-06, + "loss": 0.0687, + "step": 1883 + }, + { + "epoch": 1.210019267822736, + "grad_norm": 0.9473059429466036, + "learning_rate": 7.45129465008914e-06, + "loss": 0.0882, + "step": 1884 + }, + { + "epoch": 1.2106615285806037, + "grad_norm": 1.342515023413128, + "learning_rate": 7.4480366025624994e-06, + "loss": 0.149, + "step": 1885 + }, + { + "epoch": 1.2113037893384715, + "grad_norm": 0.9036244526066303, + "learning_rate": 7.4447771873106474e-06, + "loss": 0.0844, + "step": 1886 + }, + { + "epoch": 1.211946050096339, + "grad_norm": 1.0874426558975836, + "learning_rate": 7.4415164061546295e-06, + "loss": 0.079, + "step": 1887 + }, + { + "epoch": 1.2125883108542068, + "grad_norm": 0.9182158063529867, + "learning_rate": 7.4382542609162534e-06, + "loss": 0.0824, + "step": 1888 + }, + { + "epoch": 1.2132305716120746, + "grad_norm": 1.012747402508562, + "learning_rate": 7.43499075341809e-06, + "loss": 0.0971, + "step": 1889 + }, + { + "epoch": 1.2138728323699421, + "grad_norm": 0.8394452404215468, + "learning_rate": 7.43172588548347e-06, + "loss": 0.0555, + "step": 1890 + }, + { + "epoch": 1.21451509312781, + "grad_norm": 1.2016963107706882, + "learning_rate": 7.428459658936483e-06, + "loss": 0.107, + "step": 1891 + }, + { + "epoch": 1.2151573538856777, + "grad_norm": 1.2102566897664986, + "learning_rate": 7.425192075601984e-06, + "loss": 0.0673, + "step": 1892 + }, + { + "epoch": 1.2157996146435452, + "grad_norm": 0.5874738303992444, + "learning_rate": 7.421923137305575e-06, + "loss": 0.0365, + "step": 1893 + }, + { + "epoch": 1.216441875401413, + "grad_norm": 1.3231668922763649, + "learning_rate": 7.418652845873626e-06, + "loss": 0.1295, + "step": 1894 + }, + { + "epoch": 1.2170841361592806, + "grad_norm": 0.8161202245767644, + "learning_rate": 7.415381203133258e-06, + "loss": 0.0474, + "step": 1895 + }, + { + "epoch": 1.2177263969171483, + "grad_norm": 1.0607846979022066, + "learning_rate": 7.412108210912346e-06, + "loss": 0.0972, + "step": 1896 + }, + { + "epoch": 1.2183686576750161, + "grad_norm": 1.2000842635080673, + "learning_rate": 7.40883387103952e-06, + "loss": 0.126, + "step": 1897 + }, + { + "epoch": 1.2190109184328837, + "grad_norm": 0.9288793241211495, + "learning_rate": 7.405558185344165e-06, + "loss": 0.0846, + "step": 1898 + }, + { + "epoch": 1.2196531791907514, + "grad_norm": 1.2169249328487153, + "learning_rate": 7.402281155656415e-06, + "loss": 0.1443, + "step": 1899 + }, + { + "epoch": 1.2202954399486192, + "grad_norm": 1.0172677932437333, + "learning_rate": 7.399002783807157e-06, + "loss": 0.0605, + "step": 1900 + }, + { + "epoch": 1.2209377007064868, + "grad_norm": 0.8383423061767192, + "learning_rate": 7.395723071628025e-06, + "loss": 0.0745, + "step": 1901 + }, + { + "epoch": 1.2215799614643545, + "grad_norm": 1.159954606048721, + "learning_rate": 7.3924420209514045e-06, + "loss": 0.1145, + "step": 1902 + }, + { + "epoch": 1.2222222222222223, + "grad_norm": 1.145120195009556, + "learning_rate": 7.3891596336104295e-06, + "loss": 0.1586, + "step": 1903 + }, + { + "epoch": 1.2228644829800899, + "grad_norm": 0.7343532397875392, + "learning_rate": 7.38587591143898e-06, + "loss": 0.0729, + "step": 1904 + }, + { + "epoch": 1.2235067437379576, + "grad_norm": 0.7699898670119336, + "learning_rate": 7.3825908562716775e-06, + "loss": 0.0402, + "step": 1905 + }, + { + "epoch": 1.2241490044958252, + "grad_norm": 0.9730968544931636, + "learning_rate": 7.379304469943895e-06, + "loss": 0.1125, + "step": 1906 + }, + { + "epoch": 1.224791265253693, + "grad_norm": 0.7866084852659747, + "learning_rate": 7.376016754291746e-06, + "loss": 0.0585, + "step": 1907 + }, + { + "epoch": 1.2254335260115607, + "grad_norm": 0.893854723253194, + "learning_rate": 7.3727277111520864e-06, + "loss": 0.0907, + "step": 1908 + }, + { + "epoch": 1.2260757867694283, + "grad_norm": 0.8598457820385006, + "learning_rate": 7.3694373423625134e-06, + "loss": 0.0855, + "step": 1909 + }, + { + "epoch": 1.226718047527296, + "grad_norm": 0.8352140439577914, + "learning_rate": 7.366145649761366e-06, + "loss": 0.0864, + "step": 1910 + }, + { + "epoch": 1.2273603082851638, + "grad_norm": 1.1972619505323605, + "learning_rate": 7.362852635187722e-06, + "loss": 0.1773, + "step": 1911 + }, + { + "epoch": 1.2280025690430314, + "grad_norm": 0.967448559401531, + "learning_rate": 7.359558300481401e-06, + "loss": 0.088, + "step": 1912 + }, + { + "epoch": 1.2286448298008992, + "grad_norm": 0.8627677246951763, + "learning_rate": 7.356262647482954e-06, + "loss": 0.0888, + "step": 1913 + }, + { + "epoch": 1.229287090558767, + "grad_norm": 0.9421862651215358, + "learning_rate": 7.352965678033675e-06, + "loss": 0.0956, + "step": 1914 + }, + { + "epoch": 1.2299293513166345, + "grad_norm": 0.953125875085199, + "learning_rate": 7.349667393975587e-06, + "loss": 0.0839, + "step": 1915 + }, + { + "epoch": 1.2305716120745023, + "grad_norm": 1.0569350593050135, + "learning_rate": 7.346367797151453e-06, + "loss": 0.1314, + "step": 1916 + }, + { + "epoch": 1.2312138728323698, + "grad_norm": 0.8886082867578347, + "learning_rate": 7.343066889404769e-06, + "loss": 0.0727, + "step": 1917 + }, + { + "epoch": 1.2318561335902376, + "grad_norm": 0.852101056377998, + "learning_rate": 7.339764672579759e-06, + "loss": 0.0784, + "step": 1918 + }, + { + "epoch": 1.2324983943481054, + "grad_norm": 0.9499422525856999, + "learning_rate": 7.336461148521381e-06, + "loss": 0.1181, + "step": 1919 + }, + { + "epoch": 1.2331406551059731, + "grad_norm": 1.0293326950311628, + "learning_rate": 7.3331563190753275e-06, + "loss": 0.1146, + "step": 1920 + }, + { + "epoch": 1.2337829158638407, + "grad_norm": 0.8806772529864645, + "learning_rate": 7.329850186088012e-06, + "loss": 0.0891, + "step": 1921 + }, + { + "epoch": 1.2344251766217085, + "grad_norm": 1.0754550286873792, + "learning_rate": 7.326542751406582e-06, + "loss": 0.1256, + "step": 1922 + }, + { + "epoch": 1.235067437379576, + "grad_norm": 0.8177429140082272, + "learning_rate": 7.323234016878913e-06, + "loss": 0.0771, + "step": 1923 + }, + { + "epoch": 1.2357096981374438, + "grad_norm": 0.9464066097921429, + "learning_rate": 7.319923984353602e-06, + "loss": 0.0884, + "step": 1924 + }, + { + "epoch": 1.2363519588953116, + "grad_norm": 1.0578828348111824, + "learning_rate": 7.316612655679975e-06, + "loss": 0.0989, + "step": 1925 + }, + { + "epoch": 1.2369942196531791, + "grad_norm": 1.110755949041399, + "learning_rate": 7.313300032708081e-06, + "loss": 0.1138, + "step": 1926 + }, + { + "epoch": 1.237636480411047, + "grad_norm": 1.301726807084896, + "learning_rate": 7.309986117288692e-06, + "loss": 0.1855, + "step": 1927 + }, + { + "epoch": 1.2382787411689147, + "grad_norm": 1.214267499242661, + "learning_rate": 7.306670911273301e-06, + "loss": 0.1002, + "step": 1928 + }, + { + "epoch": 1.2389210019267822, + "grad_norm": 0.994304255914533, + "learning_rate": 7.303354416514125e-06, + "loss": 0.0388, + "step": 1929 + }, + { + "epoch": 1.23956326268465, + "grad_norm": 0.6206954735245218, + "learning_rate": 7.300036634864101e-06, + "loss": 0.0367, + "step": 1930 + }, + { + "epoch": 1.2402055234425178, + "grad_norm": 1.0436227950952555, + "learning_rate": 7.296717568176878e-06, + "loss": 0.105, + "step": 1931 + }, + { + "epoch": 1.2408477842003853, + "grad_norm": 0.7288214232191876, + "learning_rate": 7.2933972183068345e-06, + "loss": 0.0687, + "step": 1932 + }, + { + "epoch": 1.241490044958253, + "grad_norm": 0.9870744170801093, + "learning_rate": 7.290075587109053e-06, + "loss": 0.0844, + "step": 1933 + }, + { + "epoch": 1.2421323057161207, + "grad_norm": 0.9067996916782837, + "learning_rate": 7.286752676439345e-06, + "loss": 0.0892, + "step": 1934 + }, + { + "epoch": 1.2427745664739884, + "grad_norm": 1.0643280991357493, + "learning_rate": 7.283428488154227e-06, + "loss": 0.0666, + "step": 1935 + }, + { + "epoch": 1.2434168272318562, + "grad_norm": 0.8248973899555548, + "learning_rate": 7.280103024110934e-06, + "loss": 0.0641, + "step": 1936 + }, + { + "epoch": 1.2440590879897238, + "grad_norm": 0.5325119328537025, + "learning_rate": 7.276776286167411e-06, + "loss": 0.0326, + "step": 1937 + }, + { + "epoch": 1.2447013487475915, + "grad_norm": 1.2029795835732988, + "learning_rate": 7.273448276182317e-06, + "loss": 0.1816, + "step": 1938 + }, + { + "epoch": 1.2453436095054593, + "grad_norm": 1.0069467024689327, + "learning_rate": 7.2701189960150195e-06, + "loss": 0.1311, + "step": 1939 + }, + { + "epoch": 1.2459858702633269, + "grad_norm": 0.8937665234341711, + "learning_rate": 7.2667884475256015e-06, + "loss": 0.0918, + "step": 1940 + }, + { + "epoch": 1.2466281310211946, + "grad_norm": 1.0094296658644968, + "learning_rate": 7.263456632574844e-06, + "loss": 0.1151, + "step": 1941 + }, + { + "epoch": 1.2472703917790624, + "grad_norm": 0.9096577028068853, + "learning_rate": 7.2601235530242475e-06, + "loss": 0.0801, + "step": 1942 + }, + { + "epoch": 1.24791265253693, + "grad_norm": 0.8369769148621901, + "learning_rate": 7.256789210736009e-06, + "loss": 0.0716, + "step": 1943 + }, + { + "epoch": 1.2485549132947977, + "grad_norm": 1.0020857562770842, + "learning_rate": 7.253453607573037e-06, + "loss": 0.1058, + "step": 1944 + }, + { + "epoch": 1.2491971740526653, + "grad_norm": 0.7440634374245119, + "learning_rate": 7.250116745398945e-06, + "loss": 0.0776, + "step": 1945 + }, + { + "epoch": 1.249839434810533, + "grad_norm": 1.5022774023358014, + "learning_rate": 7.246778626078044e-06, + "loss": 0.0631, + "step": 1946 + }, + { + "epoch": 1.2504816955684008, + "grad_norm": 0.8695866836489263, + "learning_rate": 7.2434392514753525e-06, + "loss": 0.0746, + "step": 1947 + }, + { + "epoch": 1.2511239563262684, + "grad_norm": 1.030435940702472, + "learning_rate": 7.24009862345659e-06, + "loss": 0.0968, + "step": 1948 + }, + { + "epoch": 1.2517662170841362, + "grad_norm": 0.7551295719305243, + "learning_rate": 7.236756743888171e-06, + "loss": 0.0649, + "step": 1949 + }, + { + "epoch": 1.252408477842004, + "grad_norm": 0.9571927448763171, + "learning_rate": 7.233413614637218e-06, + "loss": 0.0925, + "step": 1950 + }, + { + "epoch": 1.2530507385998715, + "grad_norm": 0.8669215220449411, + "learning_rate": 7.230069237571546e-06, + "loss": 0.0619, + "step": 1951 + }, + { + "epoch": 1.2536929993577393, + "grad_norm": 0.9609759603215217, + "learning_rate": 7.226723614559668e-06, + "loss": 0.0816, + "step": 1952 + }, + { + "epoch": 1.254335260115607, + "grad_norm": 0.5964264232503164, + "learning_rate": 7.223376747470792e-06, + "loss": 0.0348, + "step": 1953 + }, + { + "epoch": 1.2549775208734746, + "grad_norm": 1.388077823737577, + "learning_rate": 7.220028638174823e-06, + "loss": 0.2207, + "step": 1954 + }, + { + "epoch": 1.2556197816313424, + "grad_norm": 1.0890663798748577, + "learning_rate": 7.216679288542361e-06, + "loss": 0.1106, + "step": 1955 + }, + { + "epoch": 1.25626204238921, + "grad_norm": 1.2058676822731667, + "learning_rate": 7.213328700444696e-06, + "loss": 0.1766, + "step": 1956 + }, + { + "epoch": 1.2569043031470777, + "grad_norm": 0.9740301770991936, + "learning_rate": 7.209976875753812e-06, + "loss": 0.0653, + "step": 1957 + }, + { + "epoch": 1.2575465639049455, + "grad_norm": 0.9835320173312294, + "learning_rate": 7.206623816342383e-06, + "loss": 0.088, + "step": 1958 + }, + { + "epoch": 1.2581888246628132, + "grad_norm": 0.8926456609656203, + "learning_rate": 7.203269524083773e-06, + "loss": 0.0762, + "step": 1959 + }, + { + "epoch": 1.2588310854206808, + "grad_norm": 1.033785207532834, + "learning_rate": 7.199914000852035e-06, + "loss": 0.118, + "step": 1960 + }, + { + "epoch": 1.2594733461785486, + "grad_norm": 1.0206736202350386, + "learning_rate": 7.196557248521909e-06, + "loss": 0.0547, + "step": 1961 + }, + { + "epoch": 1.260115606936416, + "grad_norm": 0.9414793153724339, + "learning_rate": 7.193199268968825e-06, + "loss": 0.1174, + "step": 1962 + }, + { + "epoch": 1.2607578676942839, + "grad_norm": 0.962815602045181, + "learning_rate": 7.189840064068892e-06, + "loss": 0.0756, + "step": 1963 + }, + { + "epoch": 1.2614001284521517, + "grad_norm": 0.9458163523875287, + "learning_rate": 7.186479635698911e-06, + "loss": 0.0994, + "step": 1964 + }, + { + "epoch": 1.2620423892100192, + "grad_norm": 1.071434090101969, + "learning_rate": 7.183117985736361e-06, + "loss": 0.1281, + "step": 1965 + }, + { + "epoch": 1.262684649967887, + "grad_norm": 0.8872249804866033, + "learning_rate": 7.1797551160594055e-06, + "loss": 0.069, + "step": 1966 + }, + { + "epoch": 1.2633269107257545, + "grad_norm": 0.8240116731209584, + "learning_rate": 7.1763910285468916e-06, + "loss": 0.077, + "step": 1967 + }, + { + "epoch": 1.2639691714836223, + "grad_norm": 0.9677892303247316, + "learning_rate": 7.173025725078344e-06, + "loss": 0.1042, + "step": 1968 + }, + { + "epoch": 1.26461143224149, + "grad_norm": 1.0552285127784764, + "learning_rate": 7.169659207533968e-06, + "loss": 0.0646, + "step": 1969 + }, + { + "epoch": 1.2652536929993579, + "grad_norm": 0.8186929377421571, + "learning_rate": 7.166291477794645e-06, + "loss": 0.0518, + "step": 1970 + }, + { + "epoch": 1.2658959537572254, + "grad_norm": 0.933557421137671, + "learning_rate": 7.162922537741937e-06, + "loss": 0.1012, + "step": 1971 + }, + { + "epoch": 1.2665382145150932, + "grad_norm": 0.8207545977365654, + "learning_rate": 7.15955238925808e-06, + "loss": 0.0586, + "step": 1972 + }, + { + "epoch": 1.2671804752729607, + "grad_norm": 0.9600589450512114, + "learning_rate": 7.1561810342259865e-06, + "loss": 0.0728, + "step": 1973 + }, + { + "epoch": 1.2678227360308285, + "grad_norm": 1.0129319417855027, + "learning_rate": 7.152808474529242e-06, + "loss": 0.1161, + "step": 1974 + }, + { + "epoch": 1.2684649967886963, + "grad_norm": 0.9016797437620927, + "learning_rate": 7.1494347120521055e-06, + "loss": 0.0945, + "step": 1975 + }, + { + "epoch": 1.2691072575465638, + "grad_norm": 1.047268827201089, + "learning_rate": 7.146059748679508e-06, + "loss": 0.0981, + "step": 1976 + }, + { + "epoch": 1.2697495183044316, + "grad_norm": 0.9606952081686321, + "learning_rate": 7.14268358629705e-06, + "loss": 0.0807, + "step": 1977 + }, + { + "epoch": 1.2703917790622992, + "grad_norm": 1.0467826262043765, + "learning_rate": 7.139306226791006e-06, + "loss": 0.0833, + "step": 1978 + }, + { + "epoch": 1.271034039820167, + "grad_norm": 0.6238349177853587, + "learning_rate": 7.1359276720483175e-06, + "loss": 0.034, + "step": 1979 + }, + { + "epoch": 1.2716763005780347, + "grad_norm": 0.6534767935506667, + "learning_rate": 7.1325479239565875e-06, + "loss": 0.0498, + "step": 1980 + }, + { + "epoch": 1.2723185613359025, + "grad_norm": 0.8910962997042163, + "learning_rate": 7.129166984404096e-06, + "loss": 0.0691, + "step": 1981 + }, + { + "epoch": 1.27296082209377, + "grad_norm": 0.7406657718093724, + "learning_rate": 7.12578485527978e-06, + "loss": 0.0636, + "step": 1982 + }, + { + "epoch": 1.2736030828516378, + "grad_norm": 0.7531576758737506, + "learning_rate": 7.122401538473249e-06, + "loss": 0.0442, + "step": 1983 + }, + { + "epoch": 1.2742453436095054, + "grad_norm": 1.3427362835276127, + "learning_rate": 7.11901703587477e-06, + "loss": 0.1753, + "step": 1984 + }, + { + "epoch": 1.2748876043673731, + "grad_norm": 0.6584314966763306, + "learning_rate": 7.115631349375275e-06, + "loss": 0.0486, + "step": 1985 + }, + { + "epoch": 1.275529865125241, + "grad_norm": 0.7137059939268097, + "learning_rate": 7.112244480866356e-06, + "loss": 0.0417, + "step": 1986 + }, + { + "epoch": 1.2761721258831085, + "grad_norm": 0.8537998637135729, + "learning_rate": 7.108856432240268e-06, + "loss": 0.0746, + "step": 1987 + }, + { + "epoch": 1.2768143866409762, + "grad_norm": 0.5952978548659066, + "learning_rate": 7.10546720538992e-06, + "loss": 0.0325, + "step": 1988 + }, + { + "epoch": 1.2774566473988438, + "grad_norm": 1.0584841674844279, + "learning_rate": 7.102076802208887e-06, + "loss": 0.1152, + "step": 1989 + }, + { + "epoch": 1.2780989081567116, + "grad_norm": 0.965006746998432, + "learning_rate": 7.098685224591396e-06, + "loss": 0.1319, + "step": 1990 + }, + { + "epoch": 1.2787411689145793, + "grad_norm": 1.1705036535212825, + "learning_rate": 7.095292474432332e-06, + "loss": 0.0738, + "step": 1991 + }, + { + "epoch": 1.2793834296724471, + "grad_norm": 0.8391211300345003, + "learning_rate": 7.091898553627232e-06, + "loss": 0.0549, + "step": 1992 + }, + { + "epoch": 1.2800256904303147, + "grad_norm": 0.9200814068641787, + "learning_rate": 7.0885034640722916e-06, + "loss": 0.1029, + "step": 1993 + }, + { + "epoch": 1.2806679511881824, + "grad_norm": 0.9035897475500713, + "learning_rate": 7.085107207664356e-06, + "loss": 0.0835, + "step": 1994 + }, + { + "epoch": 1.28131021194605, + "grad_norm": 1.1566424241629103, + "learning_rate": 7.081709786300925e-06, + "loss": 0.1652, + "step": 1995 + }, + { + "epoch": 1.2819524727039178, + "grad_norm": 0.5757112172384358, + "learning_rate": 7.078311201880147e-06, + "loss": 0.0294, + "step": 1996 + }, + { + "epoch": 1.2825947334617855, + "grad_norm": 0.62971184380133, + "learning_rate": 7.07491145630082e-06, + "loss": 0.0282, + "step": 1997 + }, + { + "epoch": 1.2832369942196533, + "grad_norm": 0.9967155404916319, + "learning_rate": 7.071510551462395e-06, + "loss": 0.0808, + "step": 1998 + }, + { + "epoch": 1.2838792549775209, + "grad_norm": 1.0463364829555573, + "learning_rate": 7.0681084892649645e-06, + "loss": 0.0991, + "step": 1999 + }, + { + "epoch": 1.2845215157353886, + "grad_norm": 0.7477571527907513, + "learning_rate": 7.064705271609271e-06, + "loss": 0.0553, + "step": 2000 + }, + { + "epoch": 1.2851637764932562, + "grad_norm": 1.2297208724047466, + "learning_rate": 7.061300900396703e-06, + "loss": 0.1676, + "step": 2001 + }, + { + "epoch": 1.285806037251124, + "grad_norm": 1.027680323650241, + "learning_rate": 7.057895377529293e-06, + "loss": 0.1148, + "step": 2002 + }, + { + "epoch": 1.2864482980089917, + "grad_norm": 0.7386092114776515, + "learning_rate": 7.0544887049097156e-06, + "loss": 0.0602, + "step": 2003 + }, + { + "epoch": 1.2870905587668593, + "grad_norm": 1.0647004928971615, + "learning_rate": 7.051080884441288e-06, + "loss": 0.1093, + "step": 2004 + }, + { + "epoch": 1.287732819524727, + "grad_norm": 1.334395943526202, + "learning_rate": 7.047671918027971e-06, + "loss": 0.1734, + "step": 2005 + }, + { + "epoch": 1.2883750802825946, + "grad_norm": 0.6942070180212508, + "learning_rate": 7.044261807574364e-06, + "loss": 0.055, + "step": 2006 + }, + { + "epoch": 1.2890173410404624, + "grad_norm": 0.9816874663741085, + "learning_rate": 7.040850554985706e-06, + "loss": 0.1118, + "step": 2007 + }, + { + "epoch": 1.2896596017983302, + "grad_norm": 0.682417091172141, + "learning_rate": 7.0374381621678725e-06, + "loss": 0.044, + "step": 2008 + }, + { + "epoch": 1.290301862556198, + "grad_norm": 0.9414605609508245, + "learning_rate": 7.034024631027379e-06, + "loss": 0.0878, + "step": 2009 + }, + { + "epoch": 1.2909441233140655, + "grad_norm": 0.8816656151866333, + "learning_rate": 7.030609963471372e-06, + "loss": 0.0932, + "step": 2010 + }, + { + "epoch": 1.2915863840719333, + "grad_norm": 1.091797584309094, + "learning_rate": 7.02719416140764e-06, + "loss": 0.1137, + "step": 2011 + }, + { + "epoch": 1.2922286448298008, + "grad_norm": 0.8132967144573243, + "learning_rate": 7.0237772267446e-06, + "loss": 0.0463, + "step": 2012 + }, + { + "epoch": 1.2928709055876686, + "grad_norm": 1.321021939717652, + "learning_rate": 7.020359161391304e-06, + "loss": 0.131, + "step": 2013 + }, + { + "epoch": 1.2935131663455364, + "grad_norm": 0.8881695969098092, + "learning_rate": 7.016939967257435e-06, + "loss": 0.0888, + "step": 2014 + }, + { + "epoch": 1.294155427103404, + "grad_norm": 0.8092432112918432, + "learning_rate": 7.013519646253305e-06, + "loss": 0.0744, + "step": 2015 + }, + { + "epoch": 1.2947976878612717, + "grad_norm": 0.7762131673777549, + "learning_rate": 7.01009820028986e-06, + "loss": 0.0643, + "step": 2016 + }, + { + "epoch": 1.2954399486191392, + "grad_norm": 1.0434332713963899, + "learning_rate": 7.00667563127867e-06, + "loss": 0.0896, + "step": 2017 + }, + { + "epoch": 1.296082209377007, + "grad_norm": 1.1950209835411212, + "learning_rate": 7.003251941131937e-06, + "loss": 0.128, + "step": 2018 + }, + { + "epoch": 1.2967244701348748, + "grad_norm": 0.9655776068617546, + "learning_rate": 6.999827131762482e-06, + "loss": 0.0858, + "step": 2019 + }, + { + "epoch": 1.2973667308927426, + "grad_norm": 0.7571201247498233, + "learning_rate": 6.99640120508376e-06, + "loss": 0.0643, + "step": 2020 + }, + { + "epoch": 1.2980089916506101, + "grad_norm": 0.49123760110408454, + "learning_rate": 6.992974163009844e-06, + "loss": 0.0294, + "step": 2021 + }, + { + "epoch": 1.298651252408478, + "grad_norm": 1.0122464005353102, + "learning_rate": 6.989546007455435e-06, + "loss": 0.103, + "step": 2022 + }, + { + "epoch": 1.2992935131663454, + "grad_norm": 1.0490383647011547, + "learning_rate": 6.986116740335851e-06, + "loss": 0.1002, + "step": 2023 + }, + { + "epoch": 1.2999357739242132, + "grad_norm": 1.1624177236142963, + "learning_rate": 6.982686363567034e-06, + "loss": 0.1211, + "step": 2024 + }, + { + "epoch": 1.300578034682081, + "grad_norm": 1.1138046878402588, + "learning_rate": 6.9792548790655465e-06, + "loss": 0.1291, + "step": 2025 + }, + { + "epoch": 1.3012202954399485, + "grad_norm": 0.5732662459057926, + "learning_rate": 6.9758222887485685e-06, + "loss": 0.0262, + "step": 2026 + }, + { + "epoch": 1.3018625561978163, + "grad_norm": 0.7381610892060717, + "learning_rate": 6.972388594533898e-06, + "loss": 0.0621, + "step": 2027 + }, + { + "epoch": 1.3025048169556839, + "grad_norm": 1.2300901703779412, + "learning_rate": 6.96895379833995e-06, + "loss": 0.1473, + "step": 2028 + }, + { + "epoch": 1.3031470777135516, + "grad_norm": 0.9497433955509799, + "learning_rate": 6.965517902085756e-06, + "loss": 0.0761, + "step": 2029 + }, + { + "epoch": 1.3037893384714194, + "grad_norm": 0.9632363610306138, + "learning_rate": 6.962080907690961e-06, + "loss": 0.0781, + "step": 2030 + }, + { + "epoch": 1.3044315992292872, + "grad_norm": 1.099017109937648, + "learning_rate": 6.958642817075823e-06, + "loss": 0.0657, + "step": 2031 + }, + { + "epoch": 1.3050738599871547, + "grad_norm": 0.7985567034834263, + "learning_rate": 6.955203632161213e-06, + "loss": 0.0661, + "step": 2032 + }, + { + "epoch": 1.3057161207450225, + "grad_norm": 1.1733850466967564, + "learning_rate": 6.9517633548686144e-06, + "loss": 0.1099, + "step": 2033 + }, + { + "epoch": 1.30635838150289, + "grad_norm": 1.2001217610745893, + "learning_rate": 6.948321987120122e-06, + "loss": 0.1703, + "step": 2034 + }, + { + "epoch": 1.3070006422607579, + "grad_norm": 0.9439391631263568, + "learning_rate": 6.944879530838435e-06, + "loss": 0.1025, + "step": 2035 + }, + { + "epoch": 1.3076429030186256, + "grad_norm": 0.9035729654232298, + "learning_rate": 6.941435987946866e-06, + "loss": 0.1108, + "step": 2036 + }, + { + "epoch": 1.3082851637764932, + "grad_norm": 0.7050729727299764, + "learning_rate": 6.937991360369331e-06, + "loss": 0.0508, + "step": 2037 + }, + { + "epoch": 1.308927424534361, + "grad_norm": 1.0399882383923829, + "learning_rate": 6.934545650030354e-06, + "loss": 0.1181, + "step": 2038 + }, + { + "epoch": 1.3095696852922287, + "grad_norm": 0.9275263147459245, + "learning_rate": 6.931098858855063e-06, + "loss": 0.1051, + "step": 2039 + }, + { + "epoch": 1.3102119460500963, + "grad_norm": 0.9542672158785227, + "learning_rate": 6.927650988769193e-06, + "loss": 0.1082, + "step": 2040 + }, + { + "epoch": 1.310854206807964, + "grad_norm": 0.8974591298535695, + "learning_rate": 6.924202041699074e-06, + "loss": 0.0705, + "step": 2041 + }, + { + "epoch": 1.3114964675658318, + "grad_norm": 0.7475704729519317, + "learning_rate": 6.9207520195716486e-06, + "loss": 0.0666, + "step": 2042 + }, + { + "epoch": 1.3121387283236994, + "grad_norm": 0.7158599120076077, + "learning_rate": 6.9173009243144485e-06, + "loss": 0.053, + "step": 2043 + }, + { + "epoch": 1.3127809890815672, + "grad_norm": 0.7061732220222038, + "learning_rate": 6.913848757855615e-06, + "loss": 0.0435, + "step": 2044 + }, + { + "epoch": 1.3134232498394347, + "grad_norm": 0.8275314172829192, + "learning_rate": 6.9103955221238815e-06, + "loss": 0.0784, + "step": 2045 + }, + { + "epoch": 1.3140655105973025, + "grad_norm": 0.986672303540467, + "learning_rate": 6.906941219048584e-06, + "loss": 0.1267, + "step": 2046 + }, + { + "epoch": 1.3147077713551703, + "grad_norm": 0.8216865945963588, + "learning_rate": 6.903485850559647e-06, + "loss": 0.1068, + "step": 2047 + }, + { + "epoch": 1.315350032113038, + "grad_norm": 0.9418934859045016, + "learning_rate": 6.900029418587597e-06, + "loss": 0.0898, + "step": 2048 + }, + { + "epoch": 1.3159922928709056, + "grad_norm": 0.9318447421928299, + "learning_rate": 6.896571925063554e-06, + "loss": 0.099, + "step": 2049 + }, + { + "epoch": 1.3166345536287734, + "grad_norm": 1.0033038351563488, + "learning_rate": 6.8931133719192286e-06, + "loss": 0.1198, + "step": 2050 + }, + { + "epoch": 1.317276814386641, + "grad_norm": 0.9523842324991022, + "learning_rate": 6.889653761086925e-06, + "loss": 0.1025, + "step": 2051 + }, + { + "epoch": 1.3179190751445087, + "grad_norm": 0.7542868554125702, + "learning_rate": 6.886193094499537e-06, + "loss": 0.0555, + "step": 2052 + }, + { + "epoch": 1.3185613359023765, + "grad_norm": 0.6709191935052061, + "learning_rate": 6.882731374090551e-06, + "loss": 0.0337, + "step": 2053 + }, + { + "epoch": 1.319203596660244, + "grad_norm": 1.0508692223595537, + "learning_rate": 6.879268601794037e-06, + "loss": 0.1344, + "step": 2054 + }, + { + "epoch": 1.3198458574181118, + "grad_norm": 0.8606583893125916, + "learning_rate": 6.875804779544659e-06, + "loss": 0.0672, + "step": 2055 + }, + { + "epoch": 1.3204881181759793, + "grad_norm": 1.3554524913951072, + "learning_rate": 6.872339909277663e-06, + "loss": 0.1992, + "step": 2056 + }, + { + "epoch": 1.321130378933847, + "grad_norm": 0.7209753812100523, + "learning_rate": 6.868873992928885e-06, + "loss": 0.053, + "step": 2057 + }, + { + "epoch": 1.3217726396917149, + "grad_norm": 0.751292662436427, + "learning_rate": 6.8654070324347384e-06, + "loss": 0.0615, + "step": 2058 + }, + { + "epoch": 1.3224149004495827, + "grad_norm": 0.892900768298336, + "learning_rate": 6.861939029732227e-06, + "loss": 0.0874, + "step": 2059 + }, + { + "epoch": 1.3230571612074502, + "grad_norm": 1.1290244938754375, + "learning_rate": 6.858469986758932e-06, + "loss": 0.1392, + "step": 2060 + }, + { + "epoch": 1.323699421965318, + "grad_norm": 0.9145709070706507, + "learning_rate": 6.854999905453022e-06, + "loss": 0.0757, + "step": 2061 + }, + { + "epoch": 1.3243416827231855, + "grad_norm": 0.6715554247762501, + "learning_rate": 6.851528787753237e-06, + "loss": 0.05, + "step": 2062 + }, + { + "epoch": 1.3249839434810533, + "grad_norm": 0.7007642693272322, + "learning_rate": 6.848056635598901e-06, + "loss": 0.0506, + "step": 2063 + }, + { + "epoch": 1.325626204238921, + "grad_norm": 0.8603356873386094, + "learning_rate": 6.844583450929918e-06, + "loss": 0.0642, + "step": 2064 + }, + { + "epoch": 1.3262684649967886, + "grad_norm": 1.2852281820860372, + "learning_rate": 6.841109235686765e-06, + "loss": 0.1712, + "step": 2065 + }, + { + "epoch": 1.3269107257546564, + "grad_norm": 1.0344072483924236, + "learning_rate": 6.837633991810498e-06, + "loss": 0.0906, + "step": 2066 + }, + { + "epoch": 1.327552986512524, + "grad_norm": 0.7374786170623946, + "learning_rate": 6.834157721242743e-06, + "loss": 0.0707, + "step": 2067 + }, + { + "epoch": 1.3281952472703917, + "grad_norm": 0.8493463994470482, + "learning_rate": 6.830680425925703e-06, + "loss": 0.0769, + "step": 2068 + }, + { + "epoch": 1.3288375080282595, + "grad_norm": 0.776652738990363, + "learning_rate": 6.827202107802155e-06, + "loss": 0.0594, + "step": 2069 + }, + { + "epoch": 1.3294797687861273, + "grad_norm": 1.0452664277835237, + "learning_rate": 6.823722768815446e-06, + "loss": 0.1075, + "step": 2070 + }, + { + "epoch": 1.3301220295439948, + "grad_norm": 0.9044770709224118, + "learning_rate": 6.820242410909489e-06, + "loss": 0.0878, + "step": 2071 + }, + { + "epoch": 1.3307642903018626, + "grad_norm": 0.8598291744051263, + "learning_rate": 6.816761036028774e-06, + "loss": 0.0672, + "step": 2072 + }, + { + "epoch": 1.3314065510597302, + "grad_norm": 1.1596859103667514, + "learning_rate": 6.8132786461183545e-06, + "loss": 0.1443, + "step": 2073 + }, + { + "epoch": 1.332048811817598, + "grad_norm": 1.4237845969796077, + "learning_rate": 6.80979524312385e-06, + "loss": 0.1901, + "step": 2074 + }, + { + "epoch": 1.3326910725754657, + "grad_norm": 0.7493463833827652, + "learning_rate": 6.806310828991449e-06, + "loss": 0.0622, + "step": 2075 + }, + { + "epoch": 1.3333333333333333, + "grad_norm": 0.7555313263707497, + "learning_rate": 6.802825405667906e-06, + "loss": 0.0534, + "step": 2076 + }, + { + "epoch": 1.333975594091201, + "grad_norm": 0.9684833109896859, + "learning_rate": 6.799338975100533e-06, + "loss": 0.0907, + "step": 2077 + }, + { + "epoch": 1.3346178548490686, + "grad_norm": 1.136011897427221, + "learning_rate": 6.7958515392372125e-06, + "loss": 0.1116, + "step": 2078 + }, + { + "epoch": 1.3352601156069364, + "grad_norm": 0.8119424568736229, + "learning_rate": 6.792363100026383e-06, + "loss": 0.0623, + "step": 2079 + }, + { + "epoch": 1.3359023763648041, + "grad_norm": 0.943173132379713, + "learning_rate": 6.788873659417048e-06, + "loss": 0.1102, + "step": 2080 + }, + { + "epoch": 1.336544637122672, + "grad_norm": 0.8648648304760076, + "learning_rate": 6.7853832193587655e-06, + "loss": 0.083, + "step": 2081 + }, + { + "epoch": 1.3371868978805395, + "grad_norm": 0.8027005747012124, + "learning_rate": 6.781891781801657e-06, + "loss": 0.0511, + "step": 2082 + }, + { + "epoch": 1.3378291586384072, + "grad_norm": 1.0719044937798885, + "learning_rate": 6.778399348696397e-06, + "loss": 0.1083, + "step": 2083 + }, + { + "epoch": 1.3384714193962748, + "grad_norm": 0.9673729692530899, + "learning_rate": 6.774905921994219e-06, + "loss": 0.0998, + "step": 2084 + }, + { + "epoch": 1.3391136801541426, + "grad_norm": 0.789499158227662, + "learning_rate": 6.771411503646912e-06, + "loss": 0.0676, + "step": 2085 + }, + { + "epoch": 1.3397559409120103, + "grad_norm": 0.9333454123809533, + "learning_rate": 6.767916095606815e-06, + "loss": 0.0906, + "step": 2086 + }, + { + "epoch": 1.340398201669878, + "grad_norm": 0.7540901475114912, + "learning_rate": 6.764419699826823e-06, + "loss": 0.0683, + "step": 2087 + }, + { + "epoch": 1.3410404624277457, + "grad_norm": 1.0066649008554904, + "learning_rate": 6.760922318260384e-06, + "loss": 0.1254, + "step": 2088 + }, + { + "epoch": 1.3416827231856134, + "grad_norm": 1.0344234874630327, + "learning_rate": 6.757423952861495e-06, + "loss": 0.0879, + "step": 2089 + }, + { + "epoch": 1.342324983943481, + "grad_norm": 1.1421055889307634, + "learning_rate": 6.753924605584703e-06, + "loss": 0.1844, + "step": 2090 + }, + { + "epoch": 1.3429672447013488, + "grad_norm": 0.9216184169322399, + "learning_rate": 6.7504242783851035e-06, + "loss": 0.0778, + "step": 2091 + }, + { + "epoch": 1.3436095054592165, + "grad_norm": 1.0344989146593455, + "learning_rate": 6.74692297321834e-06, + "loss": 0.1362, + "step": 2092 + }, + { + "epoch": 1.344251766217084, + "grad_norm": 0.9941955199693586, + "learning_rate": 6.7434206920406e-06, + "loss": 0.1132, + "step": 2093 + }, + { + "epoch": 1.3448940269749519, + "grad_norm": 1.2071134629927824, + "learning_rate": 6.739917436808621e-06, + "loss": 0.1041, + "step": 2094 + }, + { + "epoch": 1.3455362877328194, + "grad_norm": 0.9636109369218294, + "learning_rate": 6.7364132094796785e-06, + "loss": 0.1087, + "step": 2095 + }, + { + "epoch": 1.3461785484906872, + "grad_norm": 0.7197158465744669, + "learning_rate": 6.732908012011599e-06, + "loss": 0.0491, + "step": 2096 + }, + { + "epoch": 1.346820809248555, + "grad_norm": 0.7244027261707074, + "learning_rate": 6.729401846362743e-06, + "loss": 0.0449, + "step": 2097 + }, + { + "epoch": 1.3474630700064227, + "grad_norm": 1.1130403352740765, + "learning_rate": 6.725894714492016e-06, + "loss": 0.0913, + "step": 2098 + }, + { + "epoch": 1.3481053307642903, + "grad_norm": 0.7860005635053052, + "learning_rate": 6.722386618358865e-06, + "loss": 0.068, + "step": 2099 + }, + { + "epoch": 1.348747591522158, + "grad_norm": 0.7548754888894406, + "learning_rate": 6.71887755992327e-06, + "loss": 0.04, + "step": 2100 + }, + { + "epoch": 1.3493898522800256, + "grad_norm": 0.6230975816170977, + "learning_rate": 6.7153675411457574e-06, + "loss": 0.0295, + "step": 2101 + }, + { + "epoch": 1.3500321130378934, + "grad_norm": 1.061673588971549, + "learning_rate": 6.71185656398738e-06, + "loss": 0.101, + "step": 2102 + }, + { + "epoch": 1.3506743737957612, + "grad_norm": 1.220260880582269, + "learning_rate": 6.708344630409734e-06, + "loss": 0.1333, + "step": 2103 + }, + { + "epoch": 1.3513166345536287, + "grad_norm": 0.7693515430040306, + "learning_rate": 6.704831742374944e-06, + "loss": 0.0498, + "step": 2104 + }, + { + "epoch": 1.3519588953114965, + "grad_norm": 1.0097065049252207, + "learning_rate": 6.701317901845673e-06, + "loss": 0.1231, + "step": 2105 + }, + { + "epoch": 1.352601156069364, + "grad_norm": 0.8323647779694834, + "learning_rate": 6.697803110785115e-06, + "loss": 0.0631, + "step": 2106 + }, + { + "epoch": 1.3532434168272318, + "grad_norm": 0.943889229408959, + "learning_rate": 6.694287371156991e-06, + "loss": 0.0899, + "step": 2107 + }, + { + "epoch": 1.3538856775850996, + "grad_norm": 0.786524158947243, + "learning_rate": 6.690770684925558e-06, + "loss": 0.0644, + "step": 2108 + }, + { + "epoch": 1.3545279383429674, + "grad_norm": 0.9015680967846735, + "learning_rate": 6.687253054055597e-06, + "loss": 0.0935, + "step": 2109 + }, + { + "epoch": 1.355170199100835, + "grad_norm": 0.9462555488790048, + "learning_rate": 6.683734480512419e-06, + "loss": 0.0597, + "step": 2110 + }, + { + "epoch": 1.3558124598587027, + "grad_norm": 0.8755762557309074, + "learning_rate": 6.680214966261862e-06, + "loss": 0.082, + "step": 2111 + }, + { + "epoch": 1.3564547206165702, + "grad_norm": 0.8317553277858178, + "learning_rate": 6.67669451327029e-06, + "loss": 0.09, + "step": 2112 + }, + { + "epoch": 1.357096981374438, + "grad_norm": 1.24485195141571, + "learning_rate": 6.673173123504586e-06, + "loss": 0.1298, + "step": 2113 + }, + { + "epoch": 1.3577392421323058, + "grad_norm": 0.7115742289181398, + "learning_rate": 6.669650798932166e-06, + "loss": 0.0427, + "step": 2114 + }, + { + "epoch": 1.3583815028901733, + "grad_norm": 1.046830971051245, + "learning_rate": 6.666127541520958e-06, + "loss": 0.1245, + "step": 2115 + }, + { + "epoch": 1.3590237636480411, + "grad_norm": 1.1299379193601966, + "learning_rate": 6.662603353239421e-06, + "loss": 0.1418, + "step": 2116 + }, + { + "epoch": 1.3596660244059087, + "grad_norm": 0.7855834595036391, + "learning_rate": 6.6590782360565265e-06, + "loss": 0.0507, + "step": 2117 + }, + { + "epoch": 1.3603082851637764, + "grad_norm": 0.9016737349038043, + "learning_rate": 6.655552191941769e-06, + "loss": 0.0843, + "step": 2118 + }, + { + "epoch": 1.3609505459216442, + "grad_norm": 1.0694420821187753, + "learning_rate": 6.65202522286516e-06, + "loss": 0.1361, + "step": 2119 + }, + { + "epoch": 1.361592806679512, + "grad_norm": 0.793444685957149, + "learning_rate": 6.648497330797223e-06, + "loss": 0.0398, + "step": 2120 + }, + { + "epoch": 1.3622350674373795, + "grad_norm": 1.0238119344049363, + "learning_rate": 6.644968517709008e-06, + "loss": 0.1079, + "step": 2121 + }, + { + "epoch": 1.3628773281952473, + "grad_norm": 0.9289233135785565, + "learning_rate": 6.641438785572071e-06, + "loss": 0.1009, + "step": 2122 + }, + { + "epoch": 1.3635195889531149, + "grad_norm": 0.9920068541079807, + "learning_rate": 6.637908136358482e-06, + "loss": 0.0757, + "step": 2123 + }, + { + "epoch": 1.3641618497109826, + "grad_norm": 0.9272735864295257, + "learning_rate": 6.634376572040826e-06, + "loss": 0.0853, + "step": 2124 + }, + { + "epoch": 1.3648041104688504, + "grad_norm": 0.8161230464722454, + "learning_rate": 6.630844094592199e-06, + "loss": 0.064, + "step": 2125 + }, + { + "epoch": 1.365446371226718, + "grad_norm": 1.09242012849082, + "learning_rate": 6.627310705986204e-06, + "loss": 0.1176, + "step": 2126 + }, + { + "epoch": 1.3660886319845857, + "grad_norm": 0.6538564243840373, + "learning_rate": 6.623776408196959e-06, + "loss": 0.0531, + "step": 2127 + }, + { + "epoch": 1.3667308927424535, + "grad_norm": 1.1809622271356053, + "learning_rate": 6.620241203199086e-06, + "loss": 0.1376, + "step": 2128 + }, + { + "epoch": 1.367373153500321, + "grad_norm": 0.7719810630594396, + "learning_rate": 6.616705092967712e-06, + "loss": 0.0495, + "step": 2129 + }, + { + "epoch": 1.3680154142581888, + "grad_norm": 0.8478541449882684, + "learning_rate": 6.613168079478473e-06, + "loss": 0.0691, + "step": 2130 + }, + { + "epoch": 1.3686576750160566, + "grad_norm": 1.179800149331646, + "learning_rate": 6.6096301647075105e-06, + "loss": 0.1303, + "step": 2131 + }, + { + "epoch": 1.3692999357739242, + "grad_norm": 1.2154135335814982, + "learning_rate": 6.606091350631465e-06, + "loss": 0.1336, + "step": 2132 + }, + { + "epoch": 1.369942196531792, + "grad_norm": 0.7260943444620505, + "learning_rate": 6.602551639227486e-06, + "loss": 0.0501, + "step": 2133 + }, + { + "epoch": 1.3705844572896595, + "grad_norm": 0.5819560464575696, + "learning_rate": 6.599011032473217e-06, + "loss": 0.0334, + "step": 2134 + }, + { + "epoch": 1.3712267180475273, + "grad_norm": 0.9448472055630799, + "learning_rate": 6.5954695323468064e-06, + "loss": 0.0824, + "step": 2135 + }, + { + "epoch": 1.371868978805395, + "grad_norm": 1.0464122898345245, + "learning_rate": 6.5919271408269025e-06, + "loss": 0.1088, + "step": 2136 + }, + { + "epoch": 1.3725112395632628, + "grad_norm": 1.0465738151017905, + "learning_rate": 6.588383859892646e-06, + "loss": 0.1071, + "step": 2137 + }, + { + "epoch": 1.3731535003211304, + "grad_norm": 0.9608265557107227, + "learning_rate": 6.584839691523683e-06, + "loss": 0.094, + "step": 2138 + }, + { + "epoch": 1.3737957610789981, + "grad_norm": 0.8356476604870376, + "learning_rate": 6.581294637700149e-06, + "loss": 0.0662, + "step": 2139 + }, + { + "epoch": 1.3744380218368657, + "grad_norm": 1.0183005715982874, + "learning_rate": 6.577748700402674e-06, + "loss": 0.0983, + "step": 2140 + }, + { + "epoch": 1.3750802825947335, + "grad_norm": 1.1559698963368636, + "learning_rate": 6.574201881612385e-06, + "loss": 0.1042, + "step": 2141 + }, + { + "epoch": 1.3757225433526012, + "grad_norm": 0.7342889327096004, + "learning_rate": 6.570654183310901e-06, + "loss": 0.0548, + "step": 2142 + }, + { + "epoch": 1.3763648041104688, + "grad_norm": 1.3712751060574042, + "learning_rate": 6.5671056074803286e-06, + "loss": 0.1843, + "step": 2143 + }, + { + "epoch": 1.3770070648683366, + "grad_norm": 0.9040124393410874, + "learning_rate": 6.56355615610327e-06, + "loss": 0.0693, + "step": 2144 + }, + { + "epoch": 1.3776493256262041, + "grad_norm": 0.7770506591845158, + "learning_rate": 6.5600058311628145e-06, + "loss": 0.0401, + "step": 2145 + }, + { + "epoch": 1.378291586384072, + "grad_norm": 0.762812239603604, + "learning_rate": 6.556454634642537e-06, + "loss": 0.0578, + "step": 2146 + }, + { + "epoch": 1.3789338471419397, + "grad_norm": 1.0854930349074818, + "learning_rate": 6.5529025685265025e-06, + "loss": 0.1224, + "step": 2147 + }, + { + "epoch": 1.3795761078998074, + "grad_norm": 1.106804115163449, + "learning_rate": 6.54934963479926e-06, + "loss": 0.0927, + "step": 2148 + }, + { + "epoch": 1.380218368657675, + "grad_norm": 1.0368569969186636, + "learning_rate": 6.5457958354458454e-06, + "loss": 0.1025, + "step": 2149 + }, + { + "epoch": 1.3808606294155428, + "grad_norm": 0.8862151743804627, + "learning_rate": 6.542241172451774e-06, + "loss": 0.0841, + "step": 2150 + }, + { + "epoch": 1.3815028901734103, + "grad_norm": 0.9214287959818305, + "learning_rate": 6.538685647803049e-06, + "loss": 0.0961, + "step": 2151 + }, + { + "epoch": 1.382145150931278, + "grad_norm": 1.2358609435321974, + "learning_rate": 6.535129263486154e-06, + "loss": 0.1319, + "step": 2152 + }, + { + "epoch": 1.3827874116891459, + "grad_norm": 1.260720360090106, + "learning_rate": 6.531572021488047e-06, + "loss": 0.118, + "step": 2153 + }, + { + "epoch": 1.3834296724470134, + "grad_norm": 0.8659433728445008, + "learning_rate": 6.52801392379617e-06, + "loss": 0.0872, + "step": 2154 + }, + { + "epoch": 1.3840719332048812, + "grad_norm": 0.9974693790359878, + "learning_rate": 6.5244549723984465e-06, + "loss": 0.1099, + "step": 2155 + }, + { + "epoch": 1.3847141939627488, + "grad_norm": 0.6366262885015973, + "learning_rate": 6.520895169283271e-06, + "loss": 0.0476, + "step": 2156 + }, + { + "epoch": 1.3853564547206165, + "grad_norm": 0.929740757464616, + "learning_rate": 6.517334516439514e-06, + "loss": 0.0882, + "step": 2157 + }, + { + "epoch": 1.3859987154784843, + "grad_norm": 0.8972479703900603, + "learning_rate": 6.513773015856525e-06, + "loss": 0.0698, + "step": 2158 + }, + { + "epoch": 1.386640976236352, + "grad_norm": 1.0718625786731295, + "learning_rate": 6.510210669524124e-06, + "loss": 0.1116, + "step": 2159 + }, + { + "epoch": 1.3872832369942196, + "grad_norm": 0.8422278980309889, + "learning_rate": 6.506647479432604e-06, + "loss": 0.0645, + "step": 2160 + }, + { + "epoch": 1.3879254977520874, + "grad_norm": 0.8356038236126607, + "learning_rate": 6.503083447572731e-06, + "loss": 0.0734, + "step": 2161 + }, + { + "epoch": 1.388567758509955, + "grad_norm": 0.6423306381735284, + "learning_rate": 6.499518575935738e-06, + "loss": 0.0368, + "step": 2162 + }, + { + "epoch": 1.3892100192678227, + "grad_norm": 0.7428664159791573, + "learning_rate": 6.495952866513332e-06, + "loss": 0.0623, + "step": 2163 + }, + { + "epoch": 1.3898522800256905, + "grad_norm": 0.8324190652695574, + "learning_rate": 6.492386321297681e-06, + "loss": 0.0724, + "step": 2164 + }, + { + "epoch": 1.390494540783558, + "grad_norm": 0.8718993708888649, + "learning_rate": 6.4888189422814275e-06, + "loss": 0.0859, + "step": 2165 + }, + { + "epoch": 1.3911368015414258, + "grad_norm": 1.16678568166408, + "learning_rate": 6.485250731457677e-06, + "loss": 0.1073, + "step": 2166 + }, + { + "epoch": 1.3917790622992934, + "grad_norm": 0.6932670781855392, + "learning_rate": 6.481681690819999e-06, + "loss": 0.0543, + "step": 2167 + }, + { + "epoch": 1.3924213230571612, + "grad_norm": 0.9204420821794012, + "learning_rate": 6.478111822362425e-06, + "loss": 0.0865, + "step": 2168 + }, + { + "epoch": 1.393063583815029, + "grad_norm": 0.9911324180582838, + "learning_rate": 6.474541128079452e-06, + "loss": 0.1097, + "step": 2169 + }, + { + "epoch": 1.3937058445728967, + "grad_norm": 0.9187343349116788, + "learning_rate": 6.4709696099660366e-06, + "loss": 0.0721, + "step": 2170 + }, + { + "epoch": 1.3943481053307643, + "grad_norm": 0.8234366130036513, + "learning_rate": 6.467397270017599e-06, + "loss": 0.079, + "step": 2171 + }, + { + "epoch": 1.394990366088632, + "grad_norm": 0.7273345821121369, + "learning_rate": 6.463824110230015e-06, + "loss": 0.0397, + "step": 2172 + }, + { + "epoch": 1.3956326268464996, + "grad_norm": 0.772026727477735, + "learning_rate": 6.460250132599617e-06, + "loss": 0.0597, + "step": 2173 + }, + { + "epoch": 1.3962748876043674, + "grad_norm": 0.735245254167195, + "learning_rate": 6.456675339123198e-06, + "loss": 0.0558, + "step": 2174 + }, + { + "epoch": 1.3969171483622351, + "grad_norm": 0.7792945072844418, + "learning_rate": 6.453099731798007e-06, + "loss": 0.057, + "step": 2175 + }, + { + "epoch": 1.3975594091201027, + "grad_norm": 0.9289605460722784, + "learning_rate": 6.4495233126217464e-06, + "loss": 0.0707, + "step": 2176 + }, + { + "epoch": 1.3982016698779705, + "grad_norm": 0.9817531131057058, + "learning_rate": 6.445946083592569e-06, + "loss": 0.091, + "step": 2177 + }, + { + "epoch": 1.3988439306358382, + "grad_norm": 0.8449643251243204, + "learning_rate": 6.442368046709087e-06, + "loss": 0.1061, + "step": 2178 + }, + { + "epoch": 1.3994861913937058, + "grad_norm": 1.1331527941386912, + "learning_rate": 6.438789203970357e-06, + "loss": 0.1501, + "step": 2179 + }, + { + "epoch": 1.4001284521515736, + "grad_norm": 0.9292921319401408, + "learning_rate": 6.435209557375891e-06, + "loss": 0.0657, + "step": 2180 + }, + { + "epoch": 1.4007707129094413, + "grad_norm": 1.0248031224441219, + "learning_rate": 6.431629108925646e-06, + "loss": 0.0929, + "step": 2181 + }, + { + "epoch": 1.4014129736673089, + "grad_norm": 1.2145238893621466, + "learning_rate": 6.4280478606200295e-06, + "loss": 0.1549, + "step": 2182 + }, + { + "epoch": 1.4020552344251767, + "grad_norm": 0.99636644917779, + "learning_rate": 6.424465814459898e-06, + "loss": 0.0913, + "step": 2183 + }, + { + "epoch": 1.4026974951830442, + "grad_norm": 1.1250039531750833, + "learning_rate": 6.4208829724465484e-06, + "loss": 0.1232, + "step": 2184 + }, + { + "epoch": 1.403339755940912, + "grad_norm": 0.7222804363212983, + "learning_rate": 6.417299336581725e-06, + "loss": 0.0511, + "step": 2185 + }, + { + "epoch": 1.4039820166987798, + "grad_norm": 0.7280426337093224, + "learning_rate": 6.413714908867616e-06, + "loss": 0.0576, + "step": 2186 + }, + { + "epoch": 1.4046242774566475, + "grad_norm": 0.8549798370624361, + "learning_rate": 6.410129691306855e-06, + "loss": 0.0846, + "step": 2187 + }, + { + "epoch": 1.405266538214515, + "grad_norm": 0.9795009070624504, + "learning_rate": 6.406543685902508e-06, + "loss": 0.0844, + "step": 2188 + }, + { + "epoch": 1.4059087989723829, + "grad_norm": 1.1138983490762635, + "learning_rate": 6.40295689465809e-06, + "loss": 0.1212, + "step": 2189 + }, + { + "epoch": 1.4065510597302504, + "grad_norm": 0.9355819209726621, + "learning_rate": 6.3993693195775505e-06, + "loss": 0.0947, + "step": 2190 + }, + { + "epoch": 1.4071933204881182, + "grad_norm": 0.7975323163157346, + "learning_rate": 6.395780962665281e-06, + "loss": 0.054, + "step": 2191 + }, + { + "epoch": 1.407835581245986, + "grad_norm": 0.9634647915779471, + "learning_rate": 6.392191825926102e-06, + "loss": 0.0998, + "step": 2192 + }, + { + "epoch": 1.4084778420038535, + "grad_norm": 0.9414986146328217, + "learning_rate": 6.388601911365279e-06, + "loss": 0.0564, + "step": 2193 + }, + { + "epoch": 1.4091201027617213, + "grad_norm": 1.0351339576242409, + "learning_rate": 6.385011220988506e-06, + "loss": 0.082, + "step": 2194 + }, + { + "epoch": 1.4097623635195888, + "grad_norm": 0.7759844236899036, + "learning_rate": 6.3814197568019146e-06, + "loss": 0.0593, + "step": 2195 + }, + { + "epoch": 1.4104046242774566, + "grad_norm": 0.9970435121044673, + "learning_rate": 6.377827520812061e-06, + "loss": 0.091, + "step": 2196 + }, + { + "epoch": 1.4110468850353244, + "grad_norm": 0.9543262737546481, + "learning_rate": 6.374234515025942e-06, + "loss": 0.1067, + "step": 2197 + }, + { + "epoch": 1.4116891457931922, + "grad_norm": 0.9233896278470534, + "learning_rate": 6.37064074145098e-06, + "loss": 0.0991, + "step": 2198 + }, + { + "epoch": 1.4123314065510597, + "grad_norm": 1.096464069197187, + "learning_rate": 6.367046202095026e-06, + "loss": 0.1041, + "step": 2199 + }, + { + "epoch": 1.4129736673089275, + "grad_norm": 0.8711922886087791, + "learning_rate": 6.363450898966361e-06, + "loss": 0.0581, + "step": 2200 + }, + { + "epoch": 1.413615928066795, + "grad_norm": 0.7172294745068762, + "learning_rate": 6.3598548340736895e-06, + "loss": 0.0387, + "step": 2201 + }, + { + "epoch": 1.4142581888246628, + "grad_norm": 1.4035117619765312, + "learning_rate": 6.356258009426145e-06, + "loss": 0.1585, + "step": 2202 + }, + { + "epoch": 1.4149004495825306, + "grad_norm": 1.0340296041293235, + "learning_rate": 6.352660427033284e-06, + "loss": 0.1216, + "step": 2203 + }, + { + "epoch": 1.4155427103403981, + "grad_norm": 1.5271954150706037, + "learning_rate": 6.349062088905085e-06, + "loss": 0.1023, + "step": 2204 + }, + { + "epoch": 1.416184971098266, + "grad_norm": 0.9599235709102967, + "learning_rate": 6.34546299705195e-06, + "loss": 0.099, + "step": 2205 + }, + { + "epoch": 1.4168272318561335, + "grad_norm": 0.7913125146570862, + "learning_rate": 6.3418631534847034e-06, + "loss": 0.0698, + "step": 2206 + }, + { + "epoch": 1.4174694926140012, + "grad_norm": 0.7541021774258977, + "learning_rate": 6.338262560214589e-06, + "loss": 0.0453, + "step": 2207 + }, + { + "epoch": 1.418111753371869, + "grad_norm": 0.9013578270921734, + "learning_rate": 6.334661219253267e-06, + "loss": 0.0881, + "step": 2208 + }, + { + "epoch": 1.4187540141297368, + "grad_norm": 0.7772459520187764, + "learning_rate": 6.3310591326128154e-06, + "loss": 0.0655, + "step": 2209 + }, + { + "epoch": 1.4193962748876043, + "grad_norm": 0.8755061309745025, + "learning_rate": 6.327456302305734e-06, + "loss": 0.0838, + "step": 2210 + }, + { + "epoch": 1.420038535645472, + "grad_norm": 1.0080132472718406, + "learning_rate": 6.323852730344936e-06, + "loss": 0.1154, + "step": 2211 + }, + { + "epoch": 1.4206807964033397, + "grad_norm": 0.7849868767996742, + "learning_rate": 6.320248418743744e-06, + "loss": 0.0507, + "step": 2212 + }, + { + "epoch": 1.4213230571612074, + "grad_norm": 0.8846974131827929, + "learning_rate": 6.316643369515898e-06, + "loss": 0.0847, + "step": 2213 + }, + { + "epoch": 1.4219653179190752, + "grad_norm": 0.7934260203834512, + "learning_rate": 6.31303758467555e-06, + "loss": 0.0644, + "step": 2214 + }, + { + "epoch": 1.4226075786769428, + "grad_norm": 1.5536221127464476, + "learning_rate": 6.3094310662372615e-06, + "loss": 0.1696, + "step": 2215 + }, + { + "epoch": 1.4232498394348105, + "grad_norm": 0.7327980282679302, + "learning_rate": 6.305823816216008e-06, + "loss": 0.0583, + "step": 2216 + }, + { + "epoch": 1.423892100192678, + "grad_norm": 0.8307080654707376, + "learning_rate": 6.302215836627167e-06, + "loss": 0.0618, + "step": 2217 + }, + { + "epoch": 1.4245343609505459, + "grad_norm": 0.8954881917105146, + "learning_rate": 6.298607129486531e-06, + "loss": 0.0742, + "step": 2218 + }, + { + "epoch": 1.4251766217084136, + "grad_norm": 0.9952381004372782, + "learning_rate": 6.29499769681029e-06, + "loss": 0.0888, + "step": 2219 + }, + { + "epoch": 1.4258188824662814, + "grad_norm": 1.0217439031651914, + "learning_rate": 6.291387540615051e-06, + "loss": 0.0834, + "step": 2220 + }, + { + "epoch": 1.426461143224149, + "grad_norm": 0.8367837314481584, + "learning_rate": 6.287776662917813e-06, + "loss": 0.058, + "step": 2221 + }, + { + "epoch": 1.4271034039820167, + "grad_norm": 0.5603579395598867, + "learning_rate": 6.284165065735988e-06, + "loss": 0.0295, + "step": 2222 + }, + { + "epoch": 1.4277456647398843, + "grad_norm": 0.766424532522384, + "learning_rate": 6.280552751087384e-06, + "loss": 0.0559, + "step": 2223 + }, + { + "epoch": 1.428387925497752, + "grad_norm": 0.7267651151009717, + "learning_rate": 6.276939720990213e-06, + "loss": 0.0657, + "step": 2224 + }, + { + "epoch": 1.4290301862556198, + "grad_norm": 0.7801883631389273, + "learning_rate": 6.2733259774630825e-06, + "loss": 0.0665, + "step": 2225 + }, + { + "epoch": 1.4296724470134876, + "grad_norm": 1.2302285669936148, + "learning_rate": 6.269711522525006e-06, + "loss": 0.1445, + "step": 2226 + }, + { + "epoch": 1.4303147077713552, + "grad_norm": 0.8551261247983682, + "learning_rate": 6.266096358195386e-06, + "loss": 0.0537, + "step": 2227 + }, + { + "epoch": 1.430956968529223, + "grad_norm": 1.3713255806705673, + "learning_rate": 6.262480486494029e-06, + "loss": 0.0828, + "step": 2228 + }, + { + "epoch": 1.4315992292870905, + "grad_norm": 1.1129248761683899, + "learning_rate": 6.2588639094411294e-06, + "loss": 0.0782, + "step": 2229 + }, + { + "epoch": 1.4322414900449583, + "grad_norm": 1.8016748303983552, + "learning_rate": 6.255246629057282e-06, + "loss": 0.0826, + "step": 2230 + }, + { + "epoch": 1.432883750802826, + "grad_norm": 0.9132000074353761, + "learning_rate": 6.251628647363471e-06, + "loss": 0.0689, + "step": 2231 + }, + { + "epoch": 1.4335260115606936, + "grad_norm": 1.1512581663884631, + "learning_rate": 6.248009966381074e-06, + "loss": 0.1208, + "step": 2232 + }, + { + "epoch": 1.4341682723185614, + "grad_norm": 0.8540395801171295, + "learning_rate": 6.244390588131858e-06, + "loss": 0.0677, + "step": 2233 + }, + { + "epoch": 1.434810533076429, + "grad_norm": 0.9341508091535463, + "learning_rate": 6.240770514637982e-06, + "loss": 0.0751, + "step": 2234 + }, + { + "epoch": 1.4354527938342967, + "grad_norm": 0.9305662928536732, + "learning_rate": 6.237149747921988e-06, + "loss": 0.0782, + "step": 2235 + }, + { + "epoch": 1.4360950545921645, + "grad_norm": 0.9168605887402552, + "learning_rate": 6.233528290006811e-06, + "loss": 0.0795, + "step": 2236 + }, + { + "epoch": 1.4367373153500322, + "grad_norm": 0.8754959878029164, + "learning_rate": 6.2299061429157696e-06, + "loss": 0.0814, + "step": 2237 + }, + { + "epoch": 1.4373795761078998, + "grad_norm": 0.9155839351633956, + "learning_rate": 6.2262833086725704e-06, + "loss": 0.0642, + "step": 2238 + }, + { + "epoch": 1.4380218368657676, + "grad_norm": 0.7395489628666092, + "learning_rate": 6.222659789301298e-06, + "loss": 0.0429, + "step": 2239 + }, + { + "epoch": 1.4386640976236351, + "grad_norm": 0.9794070264764378, + "learning_rate": 6.2190355868264265e-06, + "loss": 0.0685, + "step": 2240 + }, + { + "epoch": 1.439306358381503, + "grad_norm": 0.8899799686620685, + "learning_rate": 6.215410703272805e-06, + "loss": 0.0646, + "step": 2241 + }, + { + "epoch": 1.4399486191393707, + "grad_norm": 0.8737388505439423, + "learning_rate": 6.211785140665669e-06, + "loss": 0.0921, + "step": 2242 + }, + { + "epoch": 1.4405908798972382, + "grad_norm": 0.6830145395131616, + "learning_rate": 6.208158901030631e-06, + "loss": 0.0453, + "step": 2243 + }, + { + "epoch": 1.441233140655106, + "grad_norm": 1.1057403691343817, + "learning_rate": 6.204531986393679e-06, + "loss": 0.1214, + "step": 2244 + }, + { + "epoch": 1.4418754014129735, + "grad_norm": 1.1538627152615304, + "learning_rate": 6.200904398781185e-06, + "loss": 0.1289, + "step": 2245 + }, + { + "epoch": 1.4425176621708413, + "grad_norm": 1.076317616686391, + "learning_rate": 6.19727614021989e-06, + "loss": 0.1317, + "step": 2246 + }, + { + "epoch": 1.443159922928709, + "grad_norm": 1.0005204803636791, + "learning_rate": 6.193647212736912e-06, + "loss": 0.1029, + "step": 2247 + }, + { + "epoch": 1.4438021836865769, + "grad_norm": 0.907848706969716, + "learning_rate": 6.190017618359746e-06, + "loss": 0.0754, + "step": 2248 + }, + { + "epoch": 1.4444444444444444, + "grad_norm": 0.6070562034805955, + "learning_rate": 6.186387359116255e-06, + "loss": 0.0433, + "step": 2249 + }, + { + "epoch": 1.4450867052023122, + "grad_norm": 0.85472029329856, + "learning_rate": 6.182756437034677e-06, + "loss": 0.0556, + "step": 2250 + }, + { + "epoch": 1.4457289659601797, + "grad_norm": 0.7699572603440714, + "learning_rate": 6.179124854143617e-06, + "loss": 0.0539, + "step": 2251 + }, + { + "epoch": 1.4463712267180475, + "grad_norm": 1.117558993665517, + "learning_rate": 6.175492612472053e-06, + "loss": 0.1075, + "step": 2252 + }, + { + "epoch": 1.4470134874759153, + "grad_norm": 0.951040610757111, + "learning_rate": 6.171859714049326e-06, + "loss": 0.0977, + "step": 2253 + }, + { + "epoch": 1.4476557482337828, + "grad_norm": 0.9288760833554776, + "learning_rate": 6.168226160905151e-06, + "loss": 0.0901, + "step": 2254 + }, + { + "epoch": 1.4482980089916506, + "grad_norm": 0.9786217118855848, + "learning_rate": 6.164591955069601e-06, + "loss": 0.0988, + "step": 2255 + }, + { + "epoch": 1.4489402697495182, + "grad_norm": 0.9399010195495431, + "learning_rate": 6.160957098573119e-06, + "loss": 0.1013, + "step": 2256 + }, + { + "epoch": 1.449582530507386, + "grad_norm": 0.9121764794020174, + "learning_rate": 6.157321593446511e-06, + "loss": 0.0816, + "step": 2257 + }, + { + "epoch": 1.4502247912652537, + "grad_norm": 1.081029719260921, + "learning_rate": 6.153685441720944e-06, + "loss": 0.1074, + "step": 2258 + }, + { + "epoch": 1.4508670520231215, + "grad_norm": 0.8850788658294789, + "learning_rate": 6.150048645427944e-06, + "loss": 0.0795, + "step": 2259 + }, + { + "epoch": 1.451509312780989, + "grad_norm": 0.7614073508378881, + "learning_rate": 6.146411206599403e-06, + "loss": 0.0543, + "step": 2260 + }, + { + "epoch": 1.4521515735388568, + "grad_norm": 1.0439487811091714, + "learning_rate": 6.142773127267567e-06, + "loss": 0.108, + "step": 2261 + }, + { + "epoch": 1.4527938342967244, + "grad_norm": 0.914719534879976, + "learning_rate": 6.139134409465044e-06, + "loss": 0.0819, + "step": 2262 + }, + { + "epoch": 1.4534360950545921, + "grad_norm": 0.7925325044610398, + "learning_rate": 6.135495055224793e-06, + "loss": 0.0483, + "step": 2263 + }, + { + "epoch": 1.45407835581246, + "grad_norm": 0.6098631714086404, + "learning_rate": 6.131855066580131e-06, + "loss": 0.0446, + "step": 2264 + }, + { + "epoch": 1.4547206165703275, + "grad_norm": 0.9830728086937686, + "learning_rate": 6.128214445564735e-06, + "loss": 0.0855, + "step": 2265 + }, + { + "epoch": 1.4553628773281952, + "grad_norm": 0.9479398927473712, + "learning_rate": 6.124573194212629e-06, + "loss": 0.0767, + "step": 2266 + }, + { + "epoch": 1.456005138086063, + "grad_norm": 0.9207281878515835, + "learning_rate": 6.120931314558187e-06, + "loss": 0.0766, + "step": 2267 + }, + { + "epoch": 1.4566473988439306, + "grad_norm": 0.8033855759437369, + "learning_rate": 6.117288808636141e-06, + "loss": 0.0697, + "step": 2268 + }, + { + "epoch": 1.4572896596017983, + "grad_norm": 0.6630124972009606, + "learning_rate": 6.113645678481571e-06, + "loss": 0.0374, + "step": 2269 + }, + { + "epoch": 1.4579319203596661, + "grad_norm": 0.8745576360288166, + "learning_rate": 6.110001926129903e-06, + "loss": 0.0672, + "step": 2270 + }, + { + "epoch": 1.4585741811175337, + "grad_norm": 0.7762634052761895, + "learning_rate": 6.10635755361691e-06, + "loss": 0.0764, + "step": 2271 + }, + { + "epoch": 1.4592164418754014, + "grad_norm": 0.9277168675311283, + "learning_rate": 6.102712562978717e-06, + "loss": 0.0918, + "step": 2272 + }, + { + "epoch": 1.459858702633269, + "grad_norm": 0.8744948820388941, + "learning_rate": 6.0990669562517905e-06, + "loss": 0.0867, + "step": 2273 + }, + { + "epoch": 1.4605009633911368, + "grad_norm": 0.8018968139904173, + "learning_rate": 6.095420735472939e-06, + "loss": 0.0678, + "step": 2274 + }, + { + "epoch": 1.4611432241490045, + "grad_norm": 1.0419090747413877, + "learning_rate": 6.091773902679317e-06, + "loss": 0.1073, + "step": 2275 + }, + { + "epoch": 1.4617854849068723, + "grad_norm": 0.9249103583795196, + "learning_rate": 6.088126459908423e-06, + "loss": 0.1001, + "step": 2276 + }, + { + "epoch": 1.4624277456647399, + "grad_norm": 0.8430087276510038, + "learning_rate": 6.0844784091980955e-06, + "loss": 0.0621, + "step": 2277 + }, + { + "epoch": 1.4630700064226076, + "grad_norm": 1.085421646909818, + "learning_rate": 6.080829752586507e-06, + "loss": 0.1106, + "step": 2278 + }, + { + "epoch": 1.4637122671804752, + "grad_norm": 0.9689839518787565, + "learning_rate": 6.077180492112173e-06, + "loss": 0.0921, + "step": 2279 + }, + { + "epoch": 1.464354527938343, + "grad_norm": 0.8882264040029048, + "learning_rate": 6.073530629813947e-06, + "loss": 0.0771, + "step": 2280 + }, + { + "epoch": 1.4649967886962107, + "grad_norm": 1.1113707827043262, + "learning_rate": 6.069880167731019e-06, + "loss": 0.1293, + "step": 2281 + }, + { + "epoch": 1.4656390494540783, + "grad_norm": 0.7343414531072708, + "learning_rate": 6.066229107902912e-06, + "loss": 0.0407, + "step": 2282 + }, + { + "epoch": 1.466281310211946, + "grad_norm": 0.7395078884101455, + "learning_rate": 6.062577452369483e-06, + "loss": 0.0543, + "step": 2283 + }, + { + "epoch": 1.4669235709698136, + "grad_norm": 0.958141821345352, + "learning_rate": 6.0589252031709225e-06, + "loss": 0.0561, + "step": 2284 + }, + { + "epoch": 1.4675658317276814, + "grad_norm": 1.1776916850947126, + "learning_rate": 6.055272362347754e-06, + "loss": 0.1267, + "step": 2285 + }, + { + "epoch": 1.4682080924855492, + "grad_norm": 1.0133225571887212, + "learning_rate": 6.05161893194083e-06, + "loss": 0.0998, + "step": 2286 + }, + { + "epoch": 1.468850353243417, + "grad_norm": 0.8050009462950727, + "learning_rate": 6.047964913991333e-06, + "loss": 0.0584, + "step": 2287 + }, + { + "epoch": 1.4694926140012845, + "grad_norm": 0.8623761024055804, + "learning_rate": 6.0443103105407715e-06, + "loss": 0.0775, + "step": 2288 + }, + { + "epoch": 1.4701348747591523, + "grad_norm": 0.958374206412463, + "learning_rate": 6.0406551236309874e-06, + "loss": 0.0893, + "step": 2289 + }, + { + "epoch": 1.4707771355170198, + "grad_norm": 1.1465528021108478, + "learning_rate": 6.036999355304141e-06, + "loss": 0.1105, + "step": 2290 + }, + { + "epoch": 1.4714193962748876, + "grad_norm": 1.0441017436781026, + "learning_rate": 6.03334300760272e-06, + "loss": 0.0911, + "step": 2291 + }, + { + "epoch": 1.4720616570327554, + "grad_norm": 1.0059614908950305, + "learning_rate": 6.029686082569537e-06, + "loss": 0.1092, + "step": 2292 + }, + { + "epoch": 1.472703917790623, + "grad_norm": 0.8210220757210566, + "learning_rate": 6.02602858224773e-06, + "loss": 0.0507, + "step": 2293 + }, + { + "epoch": 1.4733461785484907, + "grad_norm": 0.8790640991991753, + "learning_rate": 6.022370508680753e-06, + "loss": 0.0878, + "step": 2294 + }, + { + "epoch": 1.4739884393063583, + "grad_norm": 0.7172474482958701, + "learning_rate": 6.018711863912381e-06, + "loss": 0.0376, + "step": 2295 + }, + { + "epoch": 1.474630700064226, + "grad_norm": 0.6683272240425024, + "learning_rate": 6.01505264998671e-06, + "loss": 0.047, + "step": 2296 + }, + { + "epoch": 1.4752729608220938, + "grad_norm": 0.6484885901427536, + "learning_rate": 6.011392868948156e-06, + "loss": 0.0504, + "step": 2297 + }, + { + "epoch": 1.4759152215799616, + "grad_norm": 1.2275074090885378, + "learning_rate": 6.007732522841448e-06, + "loss": 0.1332, + "step": 2298 + }, + { + "epoch": 1.4765574823378291, + "grad_norm": 0.7391821617681847, + "learning_rate": 6.00407161371163e-06, + "loss": 0.0562, + "step": 2299 + }, + { + "epoch": 1.477199743095697, + "grad_norm": 1.1624665870980246, + "learning_rate": 6.000410143604064e-06, + "loss": 0.1366, + "step": 2300 + }, + { + "epoch": 1.4778420038535645, + "grad_norm": 0.9664061908769372, + "learning_rate": 5.996748114564427e-06, + "loss": 0.0735, + "step": 2301 + }, + { + "epoch": 1.4784842646114322, + "grad_norm": 1.0176047154403, + "learning_rate": 5.9930855286387e-06, + "loss": 0.0937, + "step": 2302 + }, + { + "epoch": 1.4791265253693, + "grad_norm": 1.3249010632511256, + "learning_rate": 5.989422387873185e-06, + "loss": 0.1916, + "step": 2303 + }, + { + "epoch": 1.4797687861271676, + "grad_norm": 0.9320395855038782, + "learning_rate": 5.985758694314486e-06, + "loss": 0.0625, + "step": 2304 + }, + { + "epoch": 1.4804110468850353, + "grad_norm": 0.8202790031703491, + "learning_rate": 5.982094450009524e-06, + "loss": 0.0783, + "step": 2305 + }, + { + "epoch": 1.4810533076429029, + "grad_norm": 0.5422659984846843, + "learning_rate": 5.978429657005518e-06, + "loss": 0.0493, + "step": 2306 + }, + { + "epoch": 1.4816955684007707, + "grad_norm": 0.9046496348499162, + "learning_rate": 5.97476431735e-06, + "loss": 0.0774, + "step": 2307 + }, + { + "epoch": 1.4823378291586384, + "grad_norm": 1.1067461007485653, + "learning_rate": 5.971098433090807e-06, + "loss": 0.1486, + "step": 2308 + }, + { + "epoch": 1.4829800899165062, + "grad_norm": 1.1432795959114073, + "learning_rate": 5.9674320062760795e-06, + "loss": 0.1428, + "step": 2309 + }, + { + "epoch": 1.4836223506743738, + "grad_norm": 0.5567516631903877, + "learning_rate": 5.96376503895426e-06, + "loss": 0.0333, + "step": 2310 + }, + { + "epoch": 1.4842646114322415, + "grad_norm": 0.7097088860080615, + "learning_rate": 5.960097533174094e-06, + "loss": 0.0504, + "step": 2311 + }, + { + "epoch": 1.484906872190109, + "grad_norm": 1.36139806385571, + "learning_rate": 5.956429490984628e-06, + "loss": 0.2132, + "step": 2312 + }, + { + "epoch": 1.4855491329479769, + "grad_norm": 0.9486442735110158, + "learning_rate": 5.9527609144352075e-06, + "loss": 0.079, + "step": 2313 + }, + { + "epoch": 1.4861913937058446, + "grad_norm": 0.8894112984185956, + "learning_rate": 5.949091805575477e-06, + "loss": 0.0607, + "step": 2314 + }, + { + "epoch": 1.4868336544637122, + "grad_norm": 0.7117348784904726, + "learning_rate": 5.945422166455378e-06, + "loss": 0.0489, + "step": 2315 + }, + { + "epoch": 1.48747591522158, + "grad_norm": 0.8400155918067866, + "learning_rate": 5.94175199912515e-06, + "loss": 0.0664, + "step": 2316 + }, + { + "epoch": 1.4881181759794477, + "grad_norm": 0.8576105525864969, + "learning_rate": 5.938081305635326e-06, + "loss": 0.0738, + "step": 2317 + }, + { + "epoch": 1.4887604367373153, + "grad_norm": 0.9485900516320922, + "learning_rate": 5.934410088036732e-06, + "loss": 0.103, + "step": 2318 + }, + { + "epoch": 1.489402697495183, + "grad_norm": 0.9282872006094114, + "learning_rate": 5.930738348380486e-06, + "loss": 0.0894, + "step": 2319 + }, + { + "epoch": 1.4900449582530508, + "grad_norm": 0.7344284134748736, + "learning_rate": 5.927066088718005e-06, + "loss": 0.0559, + "step": 2320 + }, + { + "epoch": 1.4906872190109184, + "grad_norm": 0.9107645005257416, + "learning_rate": 5.923393311100987e-06, + "loss": 0.1035, + "step": 2321 + }, + { + "epoch": 1.4913294797687862, + "grad_norm": 1.024059279614878, + "learning_rate": 5.9197200175814275e-06, + "loss": 0.0671, + "step": 2322 + }, + { + "epoch": 1.4919717405266537, + "grad_norm": 1.0425210977098627, + "learning_rate": 5.916046210211601e-06, + "loss": 0.0857, + "step": 2323 + }, + { + "epoch": 1.4926140012845215, + "grad_norm": 1.1126358090151292, + "learning_rate": 5.912371891044079e-06, + "loss": 0.1321, + "step": 2324 + }, + { + "epoch": 1.4932562620423893, + "grad_norm": 0.6914002988294197, + "learning_rate": 5.9086970621317115e-06, + "loss": 0.0549, + "step": 2325 + }, + { + "epoch": 1.493898522800257, + "grad_norm": 0.947639827451324, + "learning_rate": 5.905021725527637e-06, + "loss": 0.1044, + "step": 2326 + }, + { + "epoch": 1.4945407835581246, + "grad_norm": 0.8383525777921229, + "learning_rate": 5.901345883285278e-06, + "loss": 0.0502, + "step": 2327 + }, + { + "epoch": 1.4951830443159924, + "grad_norm": 1.2493995341640067, + "learning_rate": 5.897669537458339e-06, + "loss": 0.1703, + "step": 2328 + }, + { + "epoch": 1.49582530507386, + "grad_norm": 1.023339561303631, + "learning_rate": 5.893992690100803e-06, + "loss": 0.0953, + "step": 2329 + }, + { + "epoch": 1.4964675658317277, + "grad_norm": 0.8569576183092266, + "learning_rate": 5.8903153432669336e-06, + "loss": 0.0846, + "step": 2330 + }, + { + "epoch": 1.4971098265895955, + "grad_norm": 0.8898892160692433, + "learning_rate": 5.8866374990112785e-06, + "loss": 0.0726, + "step": 2331 + }, + { + "epoch": 1.497752087347463, + "grad_norm": 0.8123160546054627, + "learning_rate": 5.882959159388661e-06, + "loss": 0.0637, + "step": 2332 + }, + { + "epoch": 1.4983943481053308, + "grad_norm": 0.7257309980633548, + "learning_rate": 5.879280326454178e-06, + "loss": 0.0551, + "step": 2333 + }, + { + "epoch": 1.4990366088631983, + "grad_norm": 0.9100436574953781, + "learning_rate": 5.875601002263203e-06, + "loss": 0.0834, + "step": 2334 + }, + { + "epoch": 1.4996788696210661, + "grad_norm": 1.1123442961488355, + "learning_rate": 5.871921188871387e-06, + "loss": 0.0726, + "step": 2335 + }, + { + "epoch": 1.5003211303789339, + "grad_norm": 0.9601851797435734, + "learning_rate": 5.8682408883346535e-06, + "loss": 0.0886, + "step": 2336 + }, + { + "epoch": 1.5009633911368017, + "grad_norm": 0.8535604356953856, + "learning_rate": 5.864560102709194e-06, + "loss": 0.0927, + "step": 2337 + }, + { + "epoch": 1.5016056518946692, + "grad_norm": 1.1322262997528996, + "learning_rate": 5.860878834051478e-06, + "loss": 0.1177, + "step": 2338 + }, + { + "epoch": 1.5022479126525368, + "grad_norm": 0.7761945954162286, + "learning_rate": 5.857197084418238e-06, + "loss": 0.0724, + "step": 2339 + }, + { + "epoch": 1.5028901734104045, + "grad_norm": 1.245913790672488, + "learning_rate": 5.853514855866481e-06, + "loss": 0.1611, + "step": 2340 + }, + { + "epoch": 1.5035324341682723, + "grad_norm": 1.4174896646431934, + "learning_rate": 5.849832150453475e-06, + "loss": 0.1668, + "step": 2341 + }, + { + "epoch": 1.50417469492614, + "grad_norm": 0.8379155541961623, + "learning_rate": 5.846148970236762e-06, + "loss": 0.0582, + "step": 2342 + }, + { + "epoch": 1.5048169556840079, + "grad_norm": 0.8175933729465001, + "learning_rate": 5.842465317274144e-06, + "loss": 0.0748, + "step": 2343 + }, + { + "epoch": 1.5054592164418754, + "grad_norm": 0.8303939011433196, + "learning_rate": 5.8387811936236904e-06, + "loss": 0.0647, + "step": 2344 + }, + { + "epoch": 1.506101477199743, + "grad_norm": 1.0505967530365128, + "learning_rate": 5.8350966013437295e-06, + "loss": 0.1293, + "step": 2345 + }, + { + "epoch": 1.5067437379576107, + "grad_norm": 0.9016018194114268, + "learning_rate": 5.831411542492854e-06, + "loss": 0.0953, + "step": 2346 + }, + { + "epoch": 1.5073859987154785, + "grad_norm": 0.9068201798217378, + "learning_rate": 5.827726019129918e-06, + "loss": 0.0944, + "step": 2347 + }, + { + "epoch": 1.5080282594733463, + "grad_norm": 0.7948203705725466, + "learning_rate": 5.824040033314034e-06, + "loss": 0.0586, + "step": 2348 + }, + { + "epoch": 1.5086705202312138, + "grad_norm": 1.0730633788821515, + "learning_rate": 5.8203535871045735e-06, + "loss": 0.117, + "step": 2349 + }, + { + "epoch": 1.5093127809890816, + "grad_norm": 1.024244787795477, + "learning_rate": 5.8166666825611625e-06, + "loss": 0.1186, + "step": 2350 + }, + { + "epoch": 1.5099550417469492, + "grad_norm": 1.1146634480656248, + "learning_rate": 5.812979321743689e-06, + "loss": 0.101, + "step": 2351 + }, + { + "epoch": 1.510597302504817, + "grad_norm": 0.8389882454193822, + "learning_rate": 5.809291506712289e-06, + "loss": 0.0824, + "step": 2352 + }, + { + "epoch": 1.5112395632626847, + "grad_norm": 0.8142443663730433, + "learning_rate": 5.805603239527358e-06, + "loss": 0.0652, + "step": 2353 + }, + { + "epoch": 1.5118818240205525, + "grad_norm": 0.9835387308482662, + "learning_rate": 5.8019145222495395e-06, + "loss": 0.0831, + "step": 2354 + }, + { + "epoch": 1.51252408477842, + "grad_norm": 0.6971082178410736, + "learning_rate": 5.798225356939733e-06, + "loss": 0.0598, + "step": 2355 + }, + { + "epoch": 1.5131663455362876, + "grad_norm": 0.815998161636863, + "learning_rate": 5.7945357456590845e-06, + "loss": 0.0801, + "step": 2356 + }, + { + "epoch": 1.5138086062941554, + "grad_norm": 0.7136266043034908, + "learning_rate": 5.79084569046899e-06, + "loss": 0.0514, + "step": 2357 + }, + { + "epoch": 1.5144508670520231, + "grad_norm": 0.9361696393921768, + "learning_rate": 5.787155193431095e-06, + "loss": 0.0946, + "step": 2358 + }, + { + "epoch": 1.515093127809891, + "grad_norm": 0.9131864087514201, + "learning_rate": 5.783464256607289e-06, + "loss": 0.1107, + "step": 2359 + }, + { + "epoch": 1.5157353885677585, + "grad_norm": 0.937874038884966, + "learning_rate": 5.779772882059716e-06, + "loss": 0.0736, + "step": 2360 + }, + { + "epoch": 1.5163776493256262, + "grad_norm": 0.7565660414812377, + "learning_rate": 5.776081071850749e-06, + "loss": 0.0562, + "step": 2361 + }, + { + "epoch": 1.5170199100834938, + "grad_norm": 1.8386203137443393, + "learning_rate": 5.772388828043017e-06, + "loss": 0.0653, + "step": 2362 + }, + { + "epoch": 1.5176621708413616, + "grad_norm": 1.0161280917360114, + "learning_rate": 5.768696152699389e-06, + "loss": 0.1159, + "step": 2363 + }, + { + "epoch": 1.5183044315992293, + "grad_norm": 1.018020363872039, + "learning_rate": 5.765003047882971e-06, + "loss": 0.1055, + "step": 2364 + }, + { + "epoch": 1.5189466923570971, + "grad_norm": 1.138429280114679, + "learning_rate": 5.761309515657113e-06, + "loss": 0.1552, + "step": 2365 + }, + { + "epoch": 1.5195889531149647, + "grad_norm": 0.7844532971314132, + "learning_rate": 5.757615558085401e-06, + "loss": 0.0764, + "step": 2366 + }, + { + "epoch": 1.5202312138728322, + "grad_norm": 0.7120524395970675, + "learning_rate": 5.75392117723166e-06, + "loss": 0.0474, + "step": 2367 + }, + { + "epoch": 1.5208734746307, + "grad_norm": 1.2788180433335192, + "learning_rate": 5.750226375159954e-06, + "loss": 0.1519, + "step": 2368 + }, + { + "epoch": 1.5215157353885678, + "grad_norm": 0.6265939712840872, + "learning_rate": 5.746531153934576e-06, + "loss": 0.0281, + "step": 2369 + }, + { + "epoch": 1.5221579961464355, + "grad_norm": 1.254100473687995, + "learning_rate": 5.74283551562006e-06, + "loss": 0.1536, + "step": 2370 + }, + { + "epoch": 1.522800256904303, + "grad_norm": 0.549407177452487, + "learning_rate": 5.739139462281167e-06, + "loss": 0.0211, + "step": 2371 + }, + { + "epoch": 1.5234425176621709, + "grad_norm": 0.6321438155932302, + "learning_rate": 5.735442995982896e-06, + "loss": 0.0372, + "step": 2372 + }, + { + "epoch": 1.5240847784200384, + "grad_norm": 0.7284742337781135, + "learning_rate": 5.731746118790471e-06, + "loss": 0.0534, + "step": 2373 + }, + { + "epoch": 1.5247270391779062, + "grad_norm": 0.8574166546967341, + "learning_rate": 5.7280488327693475e-06, + "loss": 0.0823, + "step": 2374 + }, + { + "epoch": 1.525369299935774, + "grad_norm": 1.0688616756811073, + "learning_rate": 5.7243511399852125e-06, + "loss": 0.1106, + "step": 2375 + }, + { + "epoch": 1.5260115606936417, + "grad_norm": 0.8524137093810826, + "learning_rate": 5.7206530425039785e-06, + "loss": 0.0627, + "step": 2376 + }, + { + "epoch": 1.5266538214515093, + "grad_norm": 0.7645781523509348, + "learning_rate": 5.716954542391779e-06, + "loss": 0.0524, + "step": 2377 + }, + { + "epoch": 1.5272960822093768, + "grad_norm": 0.7931037015062281, + "learning_rate": 5.713255641714981e-06, + "loss": 0.0671, + "step": 2378 + }, + { + "epoch": 1.5279383429672446, + "grad_norm": 0.6652598569194805, + "learning_rate": 5.7095563425401704e-06, + "loss": 0.0588, + "step": 2379 + }, + { + "epoch": 1.5285806037251124, + "grad_norm": 1.2781195510508843, + "learning_rate": 5.705856646934154e-06, + "loss": 0.159, + "step": 2380 + }, + { + "epoch": 1.5292228644829802, + "grad_norm": 0.9846267412835382, + "learning_rate": 5.702156556963965e-06, + "loss": 0.1046, + "step": 2381 + }, + { + "epoch": 1.529865125240848, + "grad_norm": 1.0674116102686988, + "learning_rate": 5.6984560746968534e-06, + "loss": 0.0776, + "step": 2382 + }, + { + "epoch": 1.5305073859987155, + "grad_norm": 0.7826103438856983, + "learning_rate": 5.69475520220029e-06, + "loss": 0.0551, + "step": 2383 + }, + { + "epoch": 1.531149646756583, + "grad_norm": 0.6288780247865919, + "learning_rate": 5.691053941541963e-06, + "loss": 0.0319, + "step": 2384 + }, + { + "epoch": 1.5317919075144508, + "grad_norm": 0.6015403168817224, + "learning_rate": 5.687352294789776e-06, + "loss": 0.0404, + "step": 2385 + }, + { + "epoch": 1.5324341682723186, + "grad_norm": 1.00939042664293, + "learning_rate": 5.683650264011853e-06, + "loss": 0.1003, + "step": 2386 + }, + { + "epoch": 1.5330764290301864, + "grad_norm": 1.0881890517182442, + "learning_rate": 5.679947851276528e-06, + "loss": 0.1044, + "step": 2387 + }, + { + "epoch": 1.533718689788054, + "grad_norm": 0.9342887938041088, + "learning_rate": 5.676245058652349e-06, + "loss": 0.071, + "step": 2388 + }, + { + "epoch": 1.5343609505459217, + "grad_norm": 0.9120012743571765, + "learning_rate": 5.672541888208076e-06, + "loss": 0.0765, + "step": 2389 + }, + { + "epoch": 1.5350032113037893, + "grad_norm": 1.0818071197462793, + "learning_rate": 5.6688383420126845e-06, + "loss": 0.1394, + "step": 2390 + }, + { + "epoch": 1.535645472061657, + "grad_norm": 0.8448104774285105, + "learning_rate": 5.665134422135354e-06, + "loss": 0.0708, + "step": 2391 + }, + { + "epoch": 1.5362877328195248, + "grad_norm": 0.9846593659317479, + "learning_rate": 5.661430130645478e-06, + "loss": 0.1151, + "step": 2392 + }, + { + "epoch": 1.5369299935773926, + "grad_norm": 0.6758235062633663, + "learning_rate": 5.657725469612651e-06, + "loss": 0.0365, + "step": 2393 + }, + { + "epoch": 1.5375722543352601, + "grad_norm": 1.0683075529318997, + "learning_rate": 5.654020441106682e-06, + "loss": 0.1226, + "step": 2394 + }, + { + "epoch": 1.5382145150931277, + "grad_norm": 0.8865579713203171, + "learning_rate": 5.650315047197579e-06, + "loss": 0.076, + "step": 2395 + }, + { + "epoch": 1.5388567758509955, + "grad_norm": 0.7366294405029165, + "learning_rate": 5.646609289955556e-06, + "loss": 0.0538, + "step": 2396 + }, + { + "epoch": 1.5394990366088632, + "grad_norm": 0.7721945376725075, + "learning_rate": 5.642903171451031e-06, + "loss": 0.0619, + "step": 2397 + }, + { + "epoch": 1.540141297366731, + "grad_norm": 0.8825316842729521, + "learning_rate": 5.639196693754624e-06, + "loss": 0.0975, + "step": 2398 + }, + { + "epoch": 1.5407835581245986, + "grad_norm": 0.8814274946432239, + "learning_rate": 5.6354898589371555e-06, + "loss": 0.0756, + "step": 2399 + }, + { + "epoch": 1.5414258188824663, + "grad_norm": 0.825635410105419, + "learning_rate": 5.631782669069643e-06, + "loss": 0.0708, + "step": 2400 + }, + { + "epoch": 1.5420680796403339, + "grad_norm": 0.9606639638947237, + "learning_rate": 5.628075126223306e-06, + "loss": 0.0985, + "step": 2401 + }, + { + "epoch": 1.5427103403982017, + "grad_norm": 0.8061929638623027, + "learning_rate": 5.6243672324695566e-06, + "loss": 0.0682, + "step": 2402 + }, + { + "epoch": 1.5433526011560694, + "grad_norm": 0.8294774106083764, + "learning_rate": 5.620658989880011e-06, + "loss": 0.0618, + "step": 2403 + }, + { + "epoch": 1.5439948619139372, + "grad_norm": 0.7236538125469231, + "learning_rate": 5.616950400526471e-06, + "loss": 0.0601, + "step": 2404 + }, + { + "epoch": 1.5446371226718048, + "grad_norm": 0.7572471662863025, + "learning_rate": 5.613241466480936e-06, + "loss": 0.0658, + "step": 2405 + }, + { + "epoch": 1.5452793834296723, + "grad_norm": 0.5072631676492364, + "learning_rate": 5.609532189815602e-06, + "loss": 0.0317, + "step": 2406 + }, + { + "epoch": 1.54592164418754, + "grad_norm": 0.8739036823391617, + "learning_rate": 5.60582257260285e-06, + "loss": 0.0839, + "step": 2407 + }, + { + "epoch": 1.5465639049454079, + "grad_norm": 0.7545967082420214, + "learning_rate": 5.602112616915254e-06, + "loss": 0.051, + "step": 2408 + }, + { + "epoch": 1.5472061657032756, + "grad_norm": 1.0065012774439344, + "learning_rate": 5.5984023248255785e-06, + "loss": 0.1148, + "step": 2409 + }, + { + "epoch": 1.5478484264611432, + "grad_norm": 0.9619453266507889, + "learning_rate": 5.594691698406773e-06, + "loss": 0.0644, + "step": 2410 + }, + { + "epoch": 1.548490687219011, + "grad_norm": 1.0519870562042395, + "learning_rate": 5.590980739731977e-06, + "loss": 0.1022, + "step": 2411 + }, + { + "epoch": 1.5491329479768785, + "grad_norm": 1.0549285603412297, + "learning_rate": 5.587269450874513e-06, + "loss": 0.1005, + "step": 2412 + }, + { + "epoch": 1.5497752087347463, + "grad_norm": 1.2138953446202818, + "learning_rate": 5.583557833907886e-06, + "loss": 0.0472, + "step": 2413 + }, + { + "epoch": 1.550417469492614, + "grad_norm": 1.2952721049289673, + "learning_rate": 5.579845890905793e-06, + "loss": 0.1891, + "step": 2414 + }, + { + "epoch": 1.5510597302504818, + "grad_norm": 1.0628954379800784, + "learning_rate": 5.576133623942105e-06, + "loss": 0.1233, + "step": 2415 + }, + { + "epoch": 1.5517019910083494, + "grad_norm": 0.8670099780958455, + "learning_rate": 5.572421035090874e-06, + "loss": 0.0732, + "step": 2416 + }, + { + "epoch": 1.552344251766217, + "grad_norm": 0.8935570046768339, + "learning_rate": 5.5687081264263386e-06, + "loss": 0.098, + "step": 2417 + }, + { + "epoch": 1.5529865125240847, + "grad_norm": 0.9339259063274699, + "learning_rate": 5.564994900022909e-06, + "loss": 0.0983, + "step": 2418 + }, + { + "epoch": 1.5536287732819525, + "grad_norm": 1.0057567849142717, + "learning_rate": 5.561281357955174e-06, + "loss": 0.1124, + "step": 2419 + }, + { + "epoch": 1.5542710340398203, + "grad_norm": 0.8190796888807017, + "learning_rate": 5.557567502297904e-06, + "loss": 0.0759, + "step": 2420 + }, + { + "epoch": 1.5549132947976878, + "grad_norm": 1.1742580629289352, + "learning_rate": 5.5538533351260395e-06, + "loss": 0.1363, + "step": 2421 + }, + { + "epoch": 1.5555555555555556, + "grad_norm": 0.7807646964793004, + "learning_rate": 5.550138858514695e-06, + "loss": 0.0722, + "step": 2422 + }, + { + "epoch": 1.5561978163134231, + "grad_norm": 0.7292377411230687, + "learning_rate": 5.5464240745391626e-06, + "loss": 0.0385, + "step": 2423 + }, + { + "epoch": 1.556840077071291, + "grad_norm": 0.8914288778120056, + "learning_rate": 5.542708985274901e-06, + "loss": 0.0807, + "step": 2424 + }, + { + "epoch": 1.5574823378291587, + "grad_norm": 1.03502778779881, + "learning_rate": 5.538993592797543e-06, + "loss": 0.102, + "step": 2425 + }, + { + "epoch": 1.5581245985870265, + "grad_norm": 1.6665839391528536, + "learning_rate": 5.535277899182887e-06, + "loss": 0.1084, + "step": 2426 + }, + { + "epoch": 1.558766859344894, + "grad_norm": 0.7537946473151713, + "learning_rate": 5.531561906506904e-06, + "loss": 0.0489, + "step": 2427 + }, + { + "epoch": 1.5594091201027616, + "grad_norm": 1.1273507150811712, + "learning_rate": 5.527845616845727e-06, + "loss": 0.1281, + "step": 2428 + }, + { + "epoch": 1.5600513808606293, + "grad_norm": 1.0137963107532364, + "learning_rate": 5.524129032275661e-06, + "loss": 0.114, + "step": 2429 + }, + { + "epoch": 1.560693641618497, + "grad_norm": 0.9812966945853163, + "learning_rate": 5.52041215487317e-06, + "loss": 0.0603, + "step": 2430 + }, + { + "epoch": 1.5613359023763649, + "grad_norm": 1.0373162631184072, + "learning_rate": 5.516694986714888e-06, + "loss": 0.1362, + "step": 2431 + }, + { + "epoch": 1.5619781631342327, + "grad_norm": 0.8317131311225232, + "learning_rate": 5.512977529877605e-06, + "loss": 0.0748, + "step": 2432 + }, + { + "epoch": 1.5626204238921002, + "grad_norm": 0.5871166498930168, + "learning_rate": 5.509259786438275e-06, + "loss": 0.0359, + "step": 2433 + }, + { + "epoch": 1.5632626846499678, + "grad_norm": 0.6425988590582232, + "learning_rate": 5.5055417584740125e-06, + "loss": 0.0441, + "step": 2434 + }, + { + "epoch": 1.5639049454078355, + "grad_norm": 0.8783674507560337, + "learning_rate": 5.501823448062092e-06, + "loss": 0.0642, + "step": 2435 + }, + { + "epoch": 1.5645472061657033, + "grad_norm": 0.8590509745278608, + "learning_rate": 5.498104857279941e-06, + "loss": 0.0827, + "step": 2436 + }, + { + "epoch": 1.565189466923571, + "grad_norm": 1.2678590286307239, + "learning_rate": 5.494385988205151e-06, + "loss": 0.1482, + "step": 2437 + }, + { + "epoch": 1.5658317276814386, + "grad_norm": 1.137274555302496, + "learning_rate": 5.490666842915463e-06, + "loss": 0.1167, + "step": 2438 + }, + { + "epoch": 1.5664739884393064, + "grad_norm": 0.5454869221714858, + "learning_rate": 5.486947423488774e-06, + "loss": 0.0288, + "step": 2439 + }, + { + "epoch": 1.567116249197174, + "grad_norm": 1.2568170181288814, + "learning_rate": 5.4832277320031345e-06, + "loss": 0.121, + "step": 2440 + }, + { + "epoch": 1.5677585099550417, + "grad_norm": 0.7360199473136229, + "learning_rate": 5.4795077705367485e-06, + "loss": 0.0573, + "step": 2441 + }, + { + "epoch": 1.5684007707129095, + "grad_norm": 0.7727084657185417, + "learning_rate": 5.47578754116797e-06, + "loss": 0.063, + "step": 2442 + }, + { + "epoch": 1.5690430314707773, + "grad_norm": 1.0288467747310435, + "learning_rate": 5.472067045975298e-06, + "loss": 0.0927, + "step": 2443 + }, + { + "epoch": 1.5696852922286448, + "grad_norm": 1.0860662190334738, + "learning_rate": 5.4683462870373886e-06, + "loss": 0.112, + "step": 2444 + }, + { + "epoch": 1.5703275529865124, + "grad_norm": 0.7411886148073429, + "learning_rate": 5.464625266433036e-06, + "loss": 0.0561, + "step": 2445 + }, + { + "epoch": 1.5709698137443802, + "grad_norm": 0.7869218151865294, + "learning_rate": 5.4609039862411905e-06, + "loss": 0.0578, + "step": 2446 + }, + { + "epoch": 1.571612074502248, + "grad_norm": 1.1844446976065337, + "learning_rate": 5.457182448540939e-06, + "loss": 0.125, + "step": 2447 + }, + { + "epoch": 1.5722543352601157, + "grad_norm": 1.324138013549287, + "learning_rate": 5.453460655411515e-06, + "loss": 0.2478, + "step": 2448 + }, + { + "epoch": 1.5728965960179833, + "grad_norm": 1.469186603682891, + "learning_rate": 5.449738608932297e-06, + "loss": 0.1692, + "step": 2449 + }, + { + "epoch": 1.573538856775851, + "grad_norm": 0.9584912355927517, + "learning_rate": 5.446016311182804e-06, + "loss": 0.07, + "step": 2450 + }, + { + "epoch": 1.5741811175337186, + "grad_norm": 0.817557532958863, + "learning_rate": 5.442293764242691e-06, + "loss": 0.068, + "step": 2451 + }, + { + "epoch": 1.5748233782915864, + "grad_norm": 1.061194155679966, + "learning_rate": 5.438570970191759e-06, + "loss": 0.1048, + "step": 2452 + }, + { + "epoch": 1.5754656390494541, + "grad_norm": 0.7384844211637026, + "learning_rate": 5.4348479311099436e-06, + "loss": 0.0451, + "step": 2453 + }, + { + "epoch": 1.576107899807322, + "grad_norm": 1.0675451421477373, + "learning_rate": 5.4311246490773175e-06, + "loss": 0.0718, + "step": 2454 + }, + { + "epoch": 1.5767501605651895, + "grad_norm": 0.8496030994086603, + "learning_rate": 5.427401126174089e-06, + "loss": 0.0617, + "step": 2455 + }, + { + "epoch": 1.577392421323057, + "grad_norm": 1.2626141930244457, + "learning_rate": 5.4236773644806e-06, + "loss": 0.1357, + "step": 2456 + }, + { + "epoch": 1.5780346820809248, + "grad_norm": 0.7763263649453679, + "learning_rate": 5.4199533660773276e-06, + "loss": 0.0406, + "step": 2457 + }, + { + "epoch": 1.5786769428387926, + "grad_norm": 0.9032962896144214, + "learning_rate": 5.416229133044882e-06, + "loss": 0.0871, + "step": 2458 + }, + { + "epoch": 1.5793192035966603, + "grad_norm": 0.9762188822648629, + "learning_rate": 5.4125046674640025e-06, + "loss": 0.089, + "step": 2459 + }, + { + "epoch": 1.579961464354528, + "grad_norm": 0.8907457130198709, + "learning_rate": 5.408779971415559e-06, + "loss": 0.0569, + "step": 2460 + }, + { + "epoch": 1.5806037251123957, + "grad_norm": 0.849678423245804, + "learning_rate": 5.405055046980548e-06, + "loss": 0.0775, + "step": 2461 + }, + { + "epoch": 1.5812459858702632, + "grad_norm": 1.0688772246895155, + "learning_rate": 5.4013298962400986e-06, + "loss": 0.1232, + "step": 2462 + }, + { + "epoch": 1.581888246628131, + "grad_norm": 0.8432305885796736, + "learning_rate": 5.397604521275462e-06, + "loss": 0.0727, + "step": 2463 + }, + { + "epoch": 1.5825305073859988, + "grad_norm": 0.977857269006205, + "learning_rate": 5.393878924168017e-06, + "loss": 0.0885, + "step": 2464 + }, + { + "epoch": 1.5831727681438665, + "grad_norm": 1.001128743546471, + "learning_rate": 5.390153106999264e-06, + "loss": 0.0641, + "step": 2465 + }, + { + "epoch": 1.583815028901734, + "grad_norm": 0.9244462139431996, + "learning_rate": 5.3864270718508305e-06, + "loss": 0.0784, + "step": 2466 + }, + { + "epoch": 1.5844572896596016, + "grad_norm": 0.8436416964609141, + "learning_rate": 5.38270082080446e-06, + "loss": 0.0603, + "step": 2467 + }, + { + "epoch": 1.5850995504174694, + "grad_norm": 1.0881207010834446, + "learning_rate": 5.378974355942021e-06, + "loss": 0.1459, + "step": 2468 + }, + { + "epoch": 1.5857418111753372, + "grad_norm": 0.8710520487284668, + "learning_rate": 5.375247679345501e-06, + "loss": 0.0686, + "step": 2469 + }, + { + "epoch": 1.586384071933205, + "grad_norm": 0.48067051388811566, + "learning_rate": 5.371520793097007e-06, + "loss": 0.0192, + "step": 2470 + }, + { + "epoch": 1.5870263326910727, + "grad_norm": 0.7827926879944196, + "learning_rate": 5.3677936992787575e-06, + "loss": 0.0479, + "step": 2471 + }, + { + "epoch": 1.5876685934489403, + "grad_norm": 1.093769468117166, + "learning_rate": 5.364066399973092e-06, + "loss": 0.1187, + "step": 2472 + }, + { + "epoch": 1.5883108542068078, + "grad_norm": 0.9575158862353197, + "learning_rate": 5.360338897262465e-06, + "loss": 0.0816, + "step": 2473 + }, + { + "epoch": 1.5889531149646756, + "grad_norm": 0.8305396099355754, + "learning_rate": 5.356611193229441e-06, + "loss": 0.0418, + "step": 2474 + }, + { + "epoch": 1.5895953757225434, + "grad_norm": 0.9173444133326648, + "learning_rate": 5.352883289956701e-06, + "loss": 0.0634, + "step": 2475 + }, + { + "epoch": 1.5902376364804112, + "grad_norm": 1.272955557815624, + "learning_rate": 5.349155189527034e-06, + "loss": 0.1626, + "step": 2476 + }, + { + "epoch": 1.5908798972382787, + "grad_norm": 0.924966076723742, + "learning_rate": 5.345426894023341e-06, + "loss": 0.0718, + "step": 2477 + }, + { + "epoch": 1.5915221579961463, + "grad_norm": 1.002270594910355, + "learning_rate": 5.341698405528633e-06, + "loss": 0.0932, + "step": 2478 + }, + { + "epoch": 1.592164418754014, + "grad_norm": 0.8549058386227276, + "learning_rate": 5.337969726126026e-06, + "loss": 0.0782, + "step": 2479 + }, + { + "epoch": 1.5928066795118818, + "grad_norm": 0.8699664668870759, + "learning_rate": 5.3342408578987434e-06, + "loss": 0.0604, + "step": 2480 + }, + { + "epoch": 1.5934489402697496, + "grad_norm": 0.7115648919122013, + "learning_rate": 5.330511802930117e-06, + "loss": 0.0628, + "step": 2481 + }, + { + "epoch": 1.5940912010276174, + "grad_norm": 1.0288966741482861, + "learning_rate": 5.32678256330358e-06, + "loss": 0.1166, + "step": 2482 + }, + { + "epoch": 1.594733461785485, + "grad_norm": 1.0387534852026958, + "learning_rate": 5.323053141102667e-06, + "loss": 0.1148, + "step": 2483 + }, + { + "epoch": 1.5953757225433525, + "grad_norm": 0.8994785904353468, + "learning_rate": 5.319323538411021e-06, + "loss": 0.0567, + "step": 2484 + }, + { + "epoch": 1.5960179833012202, + "grad_norm": 0.8332616130982741, + "learning_rate": 5.315593757312379e-06, + "loss": 0.0593, + "step": 2485 + }, + { + "epoch": 1.596660244059088, + "grad_norm": 0.7284795949975664, + "learning_rate": 5.311863799890582e-06, + "loss": 0.0502, + "step": 2486 + }, + { + "epoch": 1.5973025048169558, + "grad_norm": 0.6271109669139749, + "learning_rate": 5.308133668229568e-06, + "loss": 0.0294, + "step": 2487 + }, + { + "epoch": 1.5979447655748233, + "grad_norm": 1.0554797313137052, + "learning_rate": 5.304403364413372e-06, + "loss": 0.1147, + "step": 2488 + }, + { + "epoch": 1.5985870263326911, + "grad_norm": 1.0773741028476616, + "learning_rate": 5.300672890526125e-06, + "loss": 0.1026, + "step": 2489 + }, + { + "epoch": 1.5992292870905587, + "grad_norm": 0.8715546338016165, + "learning_rate": 5.296942248652055e-06, + "loss": 0.0701, + "step": 2490 + }, + { + "epoch": 1.5998715478484264, + "grad_norm": 0.9764559010797103, + "learning_rate": 5.293211440875482e-06, + "loss": 0.0928, + "step": 2491 + }, + { + "epoch": 1.6005138086062942, + "grad_norm": 0.9482139474769916, + "learning_rate": 5.28948046928082e-06, + "loss": 0.1143, + "step": 2492 + }, + { + "epoch": 1.601156069364162, + "grad_norm": 1.0744658254063313, + "learning_rate": 5.285749335952573e-06, + "loss": 0.1002, + "step": 2493 + }, + { + "epoch": 1.6017983301220295, + "grad_norm": 1.136163941481303, + "learning_rate": 5.282018042975337e-06, + "loss": 0.1619, + "step": 2494 + }, + { + "epoch": 1.602440590879897, + "grad_norm": 0.7477446930070136, + "learning_rate": 5.278286592433793e-06, + "loss": 0.0669, + "step": 2495 + }, + { + "epoch": 1.6030828516377649, + "grad_norm": 1.0882765358232187, + "learning_rate": 5.274554986412716e-06, + "loss": 0.0777, + "step": 2496 + }, + { + "epoch": 1.6037251123956326, + "grad_norm": 0.8213458662348936, + "learning_rate": 5.270823226996967e-06, + "loss": 0.0786, + "step": 2497 + }, + { + "epoch": 1.6043673731535004, + "grad_norm": 0.7922880153238369, + "learning_rate": 5.267091316271489e-06, + "loss": 0.0575, + "step": 2498 + }, + { + "epoch": 1.605009633911368, + "grad_norm": 0.9318454975278878, + "learning_rate": 5.26335925632131e-06, + "loss": 0.0788, + "step": 2499 + }, + { + "epoch": 1.6056518946692357, + "grad_norm": 1.165804308999641, + "learning_rate": 5.259627049231544e-06, + "loss": 0.1132, + "step": 2500 + }, + { + "epoch": 1.6062941554271033, + "grad_norm": 0.8927474185969305, + "learning_rate": 5.2558946970873855e-06, + "loss": 0.0795, + "step": 2501 + }, + { + "epoch": 1.606936416184971, + "grad_norm": 0.8405226602893533, + "learning_rate": 5.252162201974112e-06, + "loss": 0.0617, + "step": 2502 + }, + { + "epoch": 1.6075786769428388, + "grad_norm": 0.8768640028507471, + "learning_rate": 5.2484295659770805e-06, + "loss": 0.0837, + "step": 2503 + }, + { + "epoch": 1.6082209377007066, + "grad_norm": 0.7602332942585135, + "learning_rate": 5.2446967911817214e-06, + "loss": 0.0563, + "step": 2504 + }, + { + "epoch": 1.6088631984585742, + "grad_norm": 0.7863017036258634, + "learning_rate": 5.240963879673551e-06, + "loss": 0.0702, + "step": 2505 + }, + { + "epoch": 1.6095054592164417, + "grad_norm": 0.742311862686101, + "learning_rate": 5.237230833538154e-06, + "loss": 0.0503, + "step": 2506 + }, + { + "epoch": 1.6101477199743095, + "grad_norm": 0.8590797616065403, + "learning_rate": 5.233497654861197e-06, + "loss": 0.0807, + "step": 2507 + }, + { + "epoch": 1.6107899807321773, + "grad_norm": 0.7974231187029515, + "learning_rate": 5.229764345728419e-06, + "loss": 0.0746, + "step": 2508 + }, + { + "epoch": 1.611432241490045, + "grad_norm": 0.8052902556221347, + "learning_rate": 5.226030908225628e-06, + "loss": 0.0664, + "step": 2509 + }, + { + "epoch": 1.6120745022479126, + "grad_norm": 1.1831890570267614, + "learning_rate": 5.222297344438707e-06, + "loss": 0.1319, + "step": 2510 + }, + { + "epoch": 1.6127167630057804, + "grad_norm": 0.8667517563129914, + "learning_rate": 5.218563656453609e-06, + "loss": 0.0684, + "step": 2511 + }, + { + "epoch": 1.613359023763648, + "grad_norm": 1.1466022221216483, + "learning_rate": 5.2148298463563574e-06, + "loss": 0.1451, + "step": 2512 + }, + { + "epoch": 1.6140012845215157, + "grad_norm": 0.9140358771118396, + "learning_rate": 5.2110959162330425e-06, + "loss": 0.0927, + "step": 2513 + }, + { + "epoch": 1.6146435452793835, + "grad_norm": 1.0332775678768678, + "learning_rate": 5.207361868169821e-06, + "loss": 0.1071, + "step": 2514 + }, + { + "epoch": 1.6152858060372512, + "grad_norm": 1.0106644101165445, + "learning_rate": 5.203627704252918e-06, + "loss": 0.1, + "step": 2515 + }, + { + "epoch": 1.6159280667951188, + "grad_norm": 1.0770077797659845, + "learning_rate": 5.1998934265686205e-06, + "loss": 0.1228, + "step": 2516 + }, + { + "epoch": 1.6165703275529864, + "grad_norm": 1.2345557270653276, + "learning_rate": 5.196159037203281e-06, + "loss": 0.1728, + "step": 2517 + }, + { + "epoch": 1.6172125883108541, + "grad_norm": 0.7088267148892694, + "learning_rate": 5.192424538243314e-06, + "loss": 0.0491, + "step": 2518 + }, + { + "epoch": 1.617854849068722, + "grad_norm": 1.1122974857988561, + "learning_rate": 5.1886899317751935e-06, + "loss": 0.1347, + "step": 2519 + }, + { + "epoch": 1.6184971098265897, + "grad_norm": 1.0350954736383073, + "learning_rate": 5.184955219885457e-06, + "loss": 0.0933, + "step": 2520 + }, + { + "epoch": 1.6191393705844574, + "grad_norm": 0.6656286273119784, + "learning_rate": 5.181220404660699e-06, + "loss": 0.0484, + "step": 2521 + }, + { + "epoch": 1.619781631342325, + "grad_norm": 0.625320529405032, + "learning_rate": 5.177485488187569e-06, + "loss": 0.0433, + "step": 2522 + }, + { + "epoch": 1.6204238921001926, + "grad_norm": 1.0483732014695686, + "learning_rate": 5.173750472552777e-06, + "loss": 0.0992, + "step": 2523 + }, + { + "epoch": 1.6210661528580603, + "grad_norm": 1.026031862487748, + "learning_rate": 5.17001535984309e-06, + "loss": 0.0995, + "step": 2524 + }, + { + "epoch": 1.621708413615928, + "grad_norm": 1.2283031811394194, + "learning_rate": 5.166280152145325e-06, + "loss": 0.1503, + "step": 2525 + }, + { + "epoch": 1.6223506743737959, + "grad_norm": 0.8345164208547958, + "learning_rate": 5.162544851546349e-06, + "loss": 0.0795, + "step": 2526 + }, + { + "epoch": 1.6229929351316634, + "grad_norm": 1.1133447792784144, + "learning_rate": 5.158809460133091e-06, + "loss": 0.1506, + "step": 2527 + }, + { + "epoch": 1.6236351958895312, + "grad_norm": 0.6519968269413088, + "learning_rate": 5.155073979992523e-06, + "loss": 0.0485, + "step": 2528 + }, + { + "epoch": 1.6242774566473988, + "grad_norm": 0.9673985917514083, + "learning_rate": 5.15133841321167e-06, + "loss": 0.0947, + "step": 2529 + }, + { + "epoch": 1.6249197174052665, + "grad_norm": 0.7507477468071365, + "learning_rate": 5.1476027618776005e-06, + "loss": 0.0487, + "step": 2530 + }, + { + "epoch": 1.6255619781631343, + "grad_norm": 0.9788402397238566, + "learning_rate": 5.143867028077437e-06, + "loss": 0.0921, + "step": 2531 + }, + { + "epoch": 1.626204238921002, + "grad_norm": 0.8176346171306484, + "learning_rate": 5.140131213898345e-06, + "loss": 0.0786, + "step": 2532 + }, + { + "epoch": 1.6268464996788696, + "grad_norm": 0.9642782269893894, + "learning_rate": 5.136395321427534e-06, + "loss": 0.0845, + "step": 2533 + }, + { + "epoch": 1.6274887604367372, + "grad_norm": 1.0281289538349614, + "learning_rate": 5.132659352752256e-06, + "loss": 0.1275, + "step": 2534 + }, + { + "epoch": 1.628131021194605, + "grad_norm": 0.6427343498849506, + "learning_rate": 5.12892330995981e-06, + "loss": 0.0434, + "step": 2535 + }, + { + "epoch": 1.6287732819524727, + "grad_norm": 1.0431189113028712, + "learning_rate": 5.125187195137534e-06, + "loss": 0.1196, + "step": 2536 + }, + { + "epoch": 1.6294155427103405, + "grad_norm": 1.0125971470353103, + "learning_rate": 5.121451010372808e-06, + "loss": 0.0867, + "step": 2537 + }, + { + "epoch": 1.630057803468208, + "grad_norm": 0.8200896520131143, + "learning_rate": 5.117714757753045e-06, + "loss": 0.0673, + "step": 2538 + }, + { + "epoch": 1.6307000642260758, + "grad_norm": 0.8780513510380562, + "learning_rate": 5.113978439365703e-06, + "loss": 0.0631, + "step": 2539 + }, + { + "epoch": 1.6313423249839434, + "grad_norm": 0.8622535870545113, + "learning_rate": 5.110242057298273e-06, + "loss": 0.0799, + "step": 2540 + }, + { + "epoch": 1.6319845857418112, + "grad_norm": 0.7117605073169677, + "learning_rate": 5.106505613638284e-06, + "loss": 0.0369, + "step": 2541 + }, + { + "epoch": 1.632626846499679, + "grad_norm": 1.014011770907915, + "learning_rate": 5.102769110473297e-06, + "loss": 0.0881, + "step": 2542 + }, + { + "epoch": 1.6332691072575467, + "grad_norm": 0.9210942888812278, + "learning_rate": 5.099032549890907e-06, + "loss": 0.0773, + "step": 2543 + }, + { + "epoch": 1.6339113680154143, + "grad_norm": 0.9847405000227711, + "learning_rate": 5.095295933978742e-06, + "loss": 0.0869, + "step": 2544 + }, + { + "epoch": 1.6345536287732818, + "grad_norm": 0.8914189821183908, + "learning_rate": 5.09155926482446e-06, + "loss": 0.0593, + "step": 2545 + }, + { + "epoch": 1.6351958895311496, + "grad_norm": 1.0236878908252893, + "learning_rate": 5.0878225445157485e-06, + "loss": 0.1233, + "step": 2546 + }, + { + "epoch": 1.6358381502890174, + "grad_norm": 1.0994908531953902, + "learning_rate": 5.084085775140324e-06, + "loss": 0.1192, + "step": 2547 + }, + { + "epoch": 1.6364804110468851, + "grad_norm": 0.9686384832823647, + "learning_rate": 5.0803489587859325e-06, + "loss": 0.0706, + "step": 2548 + }, + { + "epoch": 1.6371226718047527, + "grad_norm": 0.8645545074867597, + "learning_rate": 5.076612097540341e-06, + "loss": 0.0734, + "step": 2549 + }, + { + "epoch": 1.6377649325626205, + "grad_norm": 1.0743499895724644, + "learning_rate": 5.072875193491348e-06, + "loss": 0.123, + "step": 2550 + }, + { + "epoch": 1.638407193320488, + "grad_norm": 0.9324733984690039, + "learning_rate": 5.06913824872677e-06, + "loss": 0.0758, + "step": 2551 + }, + { + "epoch": 1.6390494540783558, + "grad_norm": 1.083872516624285, + "learning_rate": 5.065401265334451e-06, + "loss": 0.1081, + "step": 2552 + }, + { + "epoch": 1.6396917148362236, + "grad_norm": 1.1117027650038103, + "learning_rate": 5.061664245402253e-06, + "loss": 0.133, + "step": 2553 + }, + { + "epoch": 1.6403339755940913, + "grad_norm": 0.6125159306378712, + "learning_rate": 5.057927191018062e-06, + "loss": 0.0259, + "step": 2554 + }, + { + "epoch": 1.6409762363519589, + "grad_norm": 0.7386908439074424, + "learning_rate": 5.054190104269779e-06, + "loss": 0.0491, + "step": 2555 + }, + { + "epoch": 1.6416184971098264, + "grad_norm": 0.979212516877111, + "learning_rate": 5.050452987245325e-06, + "loss": 0.0854, + "step": 2556 + }, + { + "epoch": 1.6422607578676942, + "grad_norm": 0.83708984575502, + "learning_rate": 5.046715842032641e-06, + "loss": 0.0616, + "step": 2557 + }, + { + "epoch": 1.642903018625562, + "grad_norm": 0.744869648097497, + "learning_rate": 5.04297867071968e-06, + "loss": 0.0655, + "step": 2558 + }, + { + "epoch": 1.6435452793834298, + "grad_norm": 0.8100806661083423, + "learning_rate": 5.039241475394409e-06, + "loss": 0.0679, + "step": 2559 + }, + { + "epoch": 1.6441875401412975, + "grad_norm": 1.0399372868716443, + "learning_rate": 5.035504258144811e-06, + "loss": 0.0852, + "step": 2560 + }, + { + "epoch": 1.644829800899165, + "grad_norm": 0.874217645318543, + "learning_rate": 5.03176702105888e-06, + "loss": 0.0704, + "step": 2561 + }, + { + "epoch": 1.6454720616570326, + "grad_norm": 1.0343410540136175, + "learning_rate": 5.02802976622462e-06, + "loss": 0.1286, + "step": 2562 + }, + { + "epoch": 1.6461143224149004, + "grad_norm": 0.8762721470480768, + "learning_rate": 5.02429249573005e-06, + "loss": 0.0798, + "step": 2563 + }, + { + "epoch": 1.6467565831727682, + "grad_norm": 0.89137051984304, + "learning_rate": 5.020555211663191e-06, + "loss": 0.074, + "step": 2564 + }, + { + "epoch": 1.647398843930636, + "grad_norm": 0.6490542089384422, + "learning_rate": 5.016817916112075e-06, + "loss": 0.0434, + "step": 2565 + }, + { + "epoch": 1.6480411046885035, + "grad_norm": 0.8170585095879481, + "learning_rate": 5.01308061116474e-06, + "loss": 0.0761, + "step": 2566 + }, + { + "epoch": 1.648683365446371, + "grad_norm": 0.881126299631547, + "learning_rate": 5.0093432989092295e-06, + "loss": 0.077, + "step": 2567 + }, + { + "epoch": 1.6493256262042388, + "grad_norm": 0.9910233201051475, + "learning_rate": 5.00560598143359e-06, + "loss": 0.096, + "step": 2568 + }, + { + "epoch": 1.6499678869621066, + "grad_norm": 0.7677651192579191, + "learning_rate": 5.001868660825873e-06, + "loss": 0.0714, + "step": 2569 + }, + { + "epoch": 1.6506101477199744, + "grad_norm": 1.2581824000386734, + "learning_rate": 4.9981313391741296e-06, + "loss": 0.1539, + "step": 2570 + }, + { + "epoch": 1.6512524084778422, + "grad_norm": 1.2614726329368577, + "learning_rate": 4.994394018566413e-06, + "loss": 0.1986, + "step": 2571 + }, + { + "epoch": 1.6518946692357097, + "grad_norm": 0.7940267014116091, + "learning_rate": 4.990656701090772e-06, + "loss": 0.0523, + "step": 2572 + }, + { + "epoch": 1.6525369299935773, + "grad_norm": 0.8825656102839631, + "learning_rate": 4.986919388835261e-06, + "loss": 0.0888, + "step": 2573 + }, + { + "epoch": 1.653179190751445, + "grad_norm": 1.0548818614618982, + "learning_rate": 4.9831820838879255e-06, + "loss": 0.1474, + "step": 2574 + }, + { + "epoch": 1.6538214515093128, + "grad_norm": 1.033731236389926, + "learning_rate": 4.9794447883368095e-06, + "loss": 0.0922, + "step": 2575 + }, + { + "epoch": 1.6544637122671806, + "grad_norm": 0.8178993239529568, + "learning_rate": 4.97570750426995e-06, + "loss": 0.0689, + "step": 2576 + }, + { + "epoch": 1.6551059730250481, + "grad_norm": 0.8408689635151863, + "learning_rate": 4.97197023377538e-06, + "loss": 0.0883, + "step": 2577 + }, + { + "epoch": 1.655748233782916, + "grad_norm": 0.7417855888107352, + "learning_rate": 4.968232978941122e-06, + "loss": 0.0509, + "step": 2578 + }, + { + "epoch": 1.6563904945407835, + "grad_norm": 0.7561078228534049, + "learning_rate": 4.964495741855192e-06, + "loss": 0.0686, + "step": 2579 + }, + { + "epoch": 1.6570327552986512, + "grad_norm": 0.8535217171876011, + "learning_rate": 4.960758524605594e-06, + "loss": 0.0673, + "step": 2580 + }, + { + "epoch": 1.657675016056519, + "grad_norm": 0.9753456797366218, + "learning_rate": 4.957021329280322e-06, + "loss": 0.0974, + "step": 2581 + }, + { + "epoch": 1.6583172768143868, + "grad_norm": 1.0497994341154067, + "learning_rate": 4.9532841579673604e-06, + "loss": 0.1128, + "step": 2582 + }, + { + "epoch": 1.6589595375722543, + "grad_norm": 0.5951938892937291, + "learning_rate": 4.949547012754676e-06, + "loss": 0.0424, + "step": 2583 + }, + { + "epoch": 1.659601798330122, + "grad_norm": 1.1793423862189243, + "learning_rate": 4.945809895730223e-06, + "loss": 0.1105, + "step": 2584 + }, + { + "epoch": 1.6602440590879897, + "grad_norm": 0.8921315328969313, + "learning_rate": 4.942072808981939e-06, + "loss": 0.0827, + "step": 2585 + }, + { + "epoch": 1.6608863198458574, + "grad_norm": 0.9665540696879545, + "learning_rate": 4.938335754597749e-06, + "loss": 0.0886, + "step": 2586 + }, + { + "epoch": 1.6615285806037252, + "grad_norm": 0.7583640323819589, + "learning_rate": 4.934598734665552e-06, + "loss": 0.066, + "step": 2587 + }, + { + "epoch": 1.6621708413615928, + "grad_norm": 1.050898091700616, + "learning_rate": 4.930861751273233e-06, + "loss": 0.1254, + "step": 2588 + }, + { + "epoch": 1.6628131021194605, + "grad_norm": 0.6518222008317779, + "learning_rate": 4.9271248065086545e-06, + "loss": 0.0486, + "step": 2589 + }, + { + "epoch": 1.663455362877328, + "grad_norm": 0.8809045931924311, + "learning_rate": 4.92338790245966e-06, + "loss": 0.0752, + "step": 2590 + }, + { + "epoch": 1.6640976236351959, + "grad_norm": 0.8146813522392098, + "learning_rate": 4.919651041214069e-06, + "loss": 0.0626, + "step": 2591 + }, + { + "epoch": 1.6647398843930636, + "grad_norm": 0.9558497565923565, + "learning_rate": 4.915914224859677e-06, + "loss": 0.0964, + "step": 2592 + }, + { + "epoch": 1.6653821451509314, + "grad_norm": 1.1587250695110107, + "learning_rate": 4.912177455484252e-06, + "loss": 0.1318, + "step": 2593 + }, + { + "epoch": 1.666024405908799, + "grad_norm": 1.057990471062008, + "learning_rate": 4.908440735175543e-06, + "loss": 0.1008, + "step": 2594 + }, + { + "epoch": 1.6666666666666665, + "grad_norm": 0.998388602001487, + "learning_rate": 4.9047040660212605e-06, + "loss": 0.112, + "step": 2595 + }, + { + "epoch": 1.6673089274245343, + "grad_norm": 0.8442174206849938, + "learning_rate": 4.900967450109095e-06, + "loss": 0.0709, + "step": 2596 + }, + { + "epoch": 1.667951188182402, + "grad_norm": 1.0522116548095302, + "learning_rate": 4.897230889526705e-06, + "loss": 0.122, + "step": 2597 + }, + { + "epoch": 1.6685934489402698, + "grad_norm": 0.7243733610504605, + "learning_rate": 4.893494386361717e-06, + "loss": 0.059, + "step": 2598 + }, + { + "epoch": 1.6692357096981374, + "grad_norm": 0.9928928055910503, + "learning_rate": 4.889757942701728e-06, + "loss": 0.0648, + "step": 2599 + }, + { + "epoch": 1.6698779704560052, + "grad_norm": 1.0078219636080543, + "learning_rate": 4.886021560634298e-06, + "loss": 0.0864, + "step": 2600 + }, + { + "epoch": 1.6705202312138727, + "grad_norm": 0.561062051698104, + "learning_rate": 4.882285242246958e-06, + "loss": 0.0331, + "step": 2601 + }, + { + "epoch": 1.6711624919717405, + "grad_norm": 1.319179606532432, + "learning_rate": 4.8785489896271946e-06, + "loss": 0.1849, + "step": 2602 + }, + { + "epoch": 1.6718047527296083, + "grad_norm": 1.0354901079505117, + "learning_rate": 4.8748128048624665e-06, + "loss": 0.1157, + "step": 2603 + }, + { + "epoch": 1.672447013487476, + "grad_norm": 0.9685773542424772, + "learning_rate": 4.8710766900401915e-06, + "loss": 0.1003, + "step": 2604 + }, + { + "epoch": 1.6730892742453436, + "grad_norm": 1.0430127156929594, + "learning_rate": 4.867340647247746e-06, + "loss": 0.103, + "step": 2605 + }, + { + "epoch": 1.6737315350032111, + "grad_norm": 1.1908436131054685, + "learning_rate": 4.863604678572468e-06, + "loss": 0.1256, + "step": 2606 + }, + { + "epoch": 1.674373795761079, + "grad_norm": 0.9898286986803368, + "learning_rate": 4.859868786101656e-06, + "loss": 0.0734, + "step": 2607 + }, + { + "epoch": 1.6750160565189467, + "grad_norm": 0.8843549298881799, + "learning_rate": 4.856132971922563e-06, + "loss": 0.0849, + "step": 2608 + }, + { + "epoch": 1.6756583172768145, + "grad_norm": 0.8816382775420666, + "learning_rate": 4.852397238122401e-06, + "loss": 0.043, + "step": 2609 + }, + { + "epoch": 1.6763005780346822, + "grad_norm": 1.0705278129795623, + "learning_rate": 4.848661586788334e-06, + "loss": 0.0919, + "step": 2610 + }, + { + "epoch": 1.6769428387925498, + "grad_norm": 0.7238135149706897, + "learning_rate": 4.8449260200074775e-06, + "loss": 0.0581, + "step": 2611 + }, + { + "epoch": 1.6775850995504173, + "grad_norm": 0.8975101020967027, + "learning_rate": 4.8411905398669094e-06, + "loss": 0.0933, + "step": 2612 + }, + { + "epoch": 1.6782273603082851, + "grad_norm": 0.8656806188433845, + "learning_rate": 4.8374551484536516e-06, + "loss": 0.0646, + "step": 2613 + }, + { + "epoch": 1.678869621066153, + "grad_norm": 0.9610405882136059, + "learning_rate": 4.833719847854677e-06, + "loss": 0.0819, + "step": 2614 + }, + { + "epoch": 1.6795118818240207, + "grad_norm": 1.3140367012726486, + "learning_rate": 4.829984640156911e-06, + "loss": 0.1812, + "step": 2615 + }, + { + "epoch": 1.6801541425818882, + "grad_norm": 0.9724528094722434, + "learning_rate": 4.826249527447223e-06, + "loss": 0.0684, + "step": 2616 + }, + { + "epoch": 1.680796403339756, + "grad_norm": 0.9209576275934365, + "learning_rate": 4.822514511812433e-06, + "loss": 0.0898, + "step": 2617 + }, + { + "epoch": 1.6814386640976235, + "grad_norm": 0.9697805431760274, + "learning_rate": 4.818779595339304e-06, + "loss": 0.1051, + "step": 2618 + }, + { + "epoch": 1.6820809248554913, + "grad_norm": 0.9793087885884718, + "learning_rate": 4.815044780114544e-06, + "loss": 0.0842, + "step": 2619 + }, + { + "epoch": 1.682723185613359, + "grad_norm": 0.9636410454852998, + "learning_rate": 4.811310068224808e-06, + "loss": 0.0974, + "step": 2620 + }, + { + "epoch": 1.6833654463712269, + "grad_norm": 1.3127364255215552, + "learning_rate": 4.807575461756689e-06, + "loss": 0.1263, + "step": 2621 + }, + { + "epoch": 1.6840077071290944, + "grad_norm": 1.1009344754849872, + "learning_rate": 4.803840962796721e-06, + "loss": 0.1199, + "step": 2622 + }, + { + "epoch": 1.684649967886962, + "grad_norm": 1.1501298403834963, + "learning_rate": 4.80010657343138e-06, + "loss": 0.1404, + "step": 2623 + }, + { + "epoch": 1.6852922286448297, + "grad_norm": 1.0002998037565403, + "learning_rate": 4.7963722957470826e-06, + "loss": 0.0942, + "step": 2624 + }, + { + "epoch": 1.6859344894026975, + "grad_norm": 0.9209434401164717, + "learning_rate": 4.792638131830181e-06, + "loss": 0.0987, + "step": 2625 + }, + { + "epoch": 1.6865767501605653, + "grad_norm": 0.8535285063345646, + "learning_rate": 4.78890408376696e-06, + "loss": 0.075, + "step": 2626 + }, + { + "epoch": 1.6872190109184328, + "grad_norm": 0.7517646659041429, + "learning_rate": 4.785170153643643e-06, + "loss": 0.0469, + "step": 2627 + }, + { + "epoch": 1.6878612716763006, + "grad_norm": 0.8178684574248792, + "learning_rate": 4.781436343546392e-06, + "loss": 0.0683, + "step": 2628 + }, + { + "epoch": 1.6885035324341682, + "grad_norm": 0.6481394291277055, + "learning_rate": 4.777702655561294e-06, + "loss": 0.0418, + "step": 2629 + }, + { + "epoch": 1.689145793192036, + "grad_norm": 0.8973543225771171, + "learning_rate": 4.773969091774374e-06, + "loss": 0.1049, + "step": 2630 + }, + { + "epoch": 1.6897880539499037, + "grad_norm": 1.3996720304726822, + "learning_rate": 4.770235654271582e-06, + "loss": 0.0916, + "step": 2631 + }, + { + "epoch": 1.6904303147077715, + "grad_norm": 1.3216834011469172, + "learning_rate": 4.766502345138803e-06, + "loss": 0.1456, + "step": 2632 + }, + { + "epoch": 1.691072575465639, + "grad_norm": 0.6890929429722153, + "learning_rate": 4.762769166461848e-06, + "loss": 0.0571, + "step": 2633 + }, + { + "epoch": 1.6917148362235066, + "grad_norm": 0.8733722947643605, + "learning_rate": 4.759036120326452e-06, + "loss": 0.0902, + "step": 2634 + }, + { + "epoch": 1.6923570969813744, + "grad_norm": 0.8205610831908638, + "learning_rate": 4.75530320881828e-06, + "loss": 0.0667, + "step": 2635 + }, + { + "epoch": 1.6929993577392421, + "grad_norm": 0.9762273627647619, + "learning_rate": 4.751570434022922e-06, + "loss": 0.0965, + "step": 2636 + }, + { + "epoch": 1.69364161849711, + "grad_norm": 0.6176903866314576, + "learning_rate": 4.7478377980258885e-06, + "loss": 0.0388, + "step": 2637 + }, + { + "epoch": 1.6942838792549775, + "grad_norm": 0.8110798716568209, + "learning_rate": 4.744105302912615e-06, + "loss": 0.0613, + "step": 2638 + }, + { + "epoch": 1.6949261400128453, + "grad_norm": 0.7950077621085739, + "learning_rate": 4.740372950768458e-06, + "loss": 0.064, + "step": 2639 + }, + { + "epoch": 1.6955684007707128, + "grad_norm": 0.9696533278281613, + "learning_rate": 4.736640743678691e-06, + "loss": 0.0882, + "step": 2640 + }, + { + "epoch": 1.6962106615285806, + "grad_norm": 0.6211810549835589, + "learning_rate": 4.732908683728515e-06, + "loss": 0.0405, + "step": 2641 + }, + { + "epoch": 1.6968529222864484, + "grad_norm": 0.7224338361080352, + "learning_rate": 4.729176773003035e-06, + "loss": 0.037, + "step": 2642 + }, + { + "epoch": 1.6974951830443161, + "grad_norm": 0.7934100117832089, + "learning_rate": 4.725445013587285e-06, + "loss": 0.06, + "step": 2643 + }, + { + "epoch": 1.6981374438021837, + "grad_norm": 1.714531373502681, + "learning_rate": 4.721713407566208e-06, + "loss": 0.0766, + "step": 2644 + }, + { + "epoch": 1.6987797045600512, + "grad_norm": 0.6941137953664485, + "learning_rate": 4.717981957024665e-06, + "loss": 0.039, + "step": 2645 + }, + { + "epoch": 1.699421965317919, + "grad_norm": 0.7518633288448674, + "learning_rate": 4.714250664047428e-06, + "loss": 0.0601, + "step": 2646 + }, + { + "epoch": 1.7000642260757868, + "grad_norm": 1.2553040132879125, + "learning_rate": 4.7105195307191805e-06, + "loss": 0.1462, + "step": 2647 + }, + { + "epoch": 1.7007064868336546, + "grad_norm": 0.7329267744508255, + "learning_rate": 4.706788559124517e-06, + "loss": 0.0409, + "step": 2648 + }, + { + "epoch": 1.701348747591522, + "grad_norm": 1.008964808042988, + "learning_rate": 4.703057751347946e-06, + "loss": 0.0969, + "step": 2649 + }, + { + "epoch": 1.7019910083493899, + "grad_norm": 0.6835375136769264, + "learning_rate": 4.699327109473876e-06, + "loss": 0.0497, + "step": 2650 + }, + { + "epoch": 1.7026332691072574, + "grad_norm": 1.0218641421242323, + "learning_rate": 4.69559663558663e-06, + "loss": 0.1043, + "step": 2651 + }, + { + "epoch": 1.7032755298651252, + "grad_norm": 1.0255911170910603, + "learning_rate": 4.691866331770434e-06, + "loss": 0.0951, + "step": 2652 + }, + { + "epoch": 1.703917790622993, + "grad_norm": 0.8433781841041844, + "learning_rate": 4.68813620010942e-06, + "loss": 0.0646, + "step": 2653 + }, + { + "epoch": 1.7045600513808608, + "grad_norm": 1.0956903137679745, + "learning_rate": 4.684406242687622e-06, + "loss": 0.1111, + "step": 2654 + }, + { + "epoch": 1.7052023121387283, + "grad_norm": 0.7127169191378271, + "learning_rate": 4.68067646158898e-06, + "loss": 0.0459, + "step": 2655 + }, + { + "epoch": 1.7058445728965959, + "grad_norm": 1.24642057943183, + "learning_rate": 4.676946858897334e-06, + "loss": 0.1318, + "step": 2656 + }, + { + "epoch": 1.7064868336544636, + "grad_norm": 1.0240920324994003, + "learning_rate": 4.673217436696423e-06, + "loss": 0.0846, + "step": 2657 + }, + { + "epoch": 1.7071290944123314, + "grad_norm": 0.6646131260439231, + "learning_rate": 4.669488197069885e-06, + "loss": 0.0375, + "step": 2658 + }, + { + "epoch": 1.7077713551701992, + "grad_norm": 0.9765430387210629, + "learning_rate": 4.665759142101257e-06, + "loss": 0.0976, + "step": 2659 + }, + { + "epoch": 1.708413615928067, + "grad_norm": 0.7338539157170114, + "learning_rate": 4.662030273873976e-06, + "loss": 0.0659, + "step": 2660 + }, + { + "epoch": 1.7090558766859345, + "grad_norm": 0.8514143803680261, + "learning_rate": 4.6583015944713675e-06, + "loss": 0.0644, + "step": 2661 + }, + { + "epoch": 1.709698137443802, + "grad_norm": 0.8690372196963918, + "learning_rate": 4.654573105976658e-06, + "loss": 0.091, + "step": 2662 + }, + { + "epoch": 1.7103403982016698, + "grad_norm": 0.7353546464472509, + "learning_rate": 4.650844810472966e-06, + "loss": 0.0699, + "step": 2663 + }, + { + "epoch": 1.7109826589595376, + "grad_norm": 1.0812425592842871, + "learning_rate": 4.647116710043302e-06, + "loss": 0.1165, + "step": 2664 + }, + { + "epoch": 1.7116249197174054, + "grad_norm": 1.0081428572566151, + "learning_rate": 4.643388806770561e-06, + "loss": 0.0951, + "step": 2665 + }, + { + "epoch": 1.712267180475273, + "grad_norm": 1.0554925225624887, + "learning_rate": 4.6396611027375365e-06, + "loss": 0.1126, + "step": 2666 + }, + { + "epoch": 1.7129094412331407, + "grad_norm": 0.8401688933002057, + "learning_rate": 4.635933600026909e-06, + "loss": 0.0636, + "step": 2667 + }, + { + "epoch": 1.7135517019910083, + "grad_norm": 1.1088752096993604, + "learning_rate": 4.632206300721243e-06, + "loss": 0.1133, + "step": 2668 + }, + { + "epoch": 1.714193962748876, + "grad_norm": 0.8171946296067703, + "learning_rate": 4.628479206902995e-06, + "loss": 0.0553, + "step": 2669 + }, + { + "epoch": 1.7148362235067438, + "grad_norm": 0.9686076135067523, + "learning_rate": 4.624752320654499e-06, + "loss": 0.0884, + "step": 2670 + }, + { + "epoch": 1.7154784842646116, + "grad_norm": 0.5515573477573871, + "learning_rate": 4.621025644057979e-06, + "loss": 0.0289, + "step": 2671 + }, + { + "epoch": 1.7161207450224791, + "grad_norm": 0.7984422141990577, + "learning_rate": 4.617299179195542e-06, + "loss": 0.081, + "step": 2672 + }, + { + "epoch": 1.7167630057803467, + "grad_norm": 1.1813011905309863, + "learning_rate": 4.613572928149172e-06, + "loss": 0.135, + "step": 2673 + }, + { + "epoch": 1.7174052665382145, + "grad_norm": 0.902651048451246, + "learning_rate": 4.609846893000737e-06, + "loss": 0.082, + "step": 2674 + }, + { + "epoch": 1.7180475272960822, + "grad_norm": 0.9096959010339405, + "learning_rate": 4.606121075831985e-06, + "loss": 0.067, + "step": 2675 + }, + { + "epoch": 1.71868978805395, + "grad_norm": 0.9660700278767133, + "learning_rate": 4.602395478724539e-06, + "loss": 0.089, + "step": 2676 + }, + { + "epoch": 1.7193320488118176, + "grad_norm": 1.1240298311687313, + "learning_rate": 4.598670103759902e-06, + "loss": 0.068, + "step": 2677 + }, + { + "epoch": 1.7199743095696853, + "grad_norm": 1.0464884579923774, + "learning_rate": 4.594944953019452e-06, + "loss": 0.0985, + "step": 2678 + }, + { + "epoch": 1.7206165703275529, + "grad_norm": 0.5831403733192766, + "learning_rate": 4.591220028584442e-06, + "loss": 0.0338, + "step": 2679 + }, + { + "epoch": 1.7212588310854207, + "grad_norm": 0.8222924932259298, + "learning_rate": 4.587495332536e-06, + "loss": 0.074, + "step": 2680 + }, + { + "epoch": 1.7219010918432884, + "grad_norm": 1.0970585121162812, + "learning_rate": 4.58377086695512e-06, + "loss": 0.1328, + "step": 2681 + }, + { + "epoch": 1.7225433526011562, + "grad_norm": 0.80778147850311, + "learning_rate": 4.580046633922675e-06, + "loss": 0.0719, + "step": 2682 + }, + { + "epoch": 1.7231856133590238, + "grad_norm": 0.9004885430734002, + "learning_rate": 4.576322635519402e-06, + "loss": 0.0917, + "step": 2683 + }, + { + "epoch": 1.7238278741168913, + "grad_norm": 0.717062595269379, + "learning_rate": 4.572598873825913e-06, + "loss": 0.0585, + "step": 2684 + }, + { + "epoch": 1.724470134874759, + "grad_norm": 1.0871576094358528, + "learning_rate": 4.568875350922683e-06, + "loss": 0.1155, + "step": 2685 + }, + { + "epoch": 1.7251123956326269, + "grad_norm": 1.0351258166444097, + "learning_rate": 4.565152068890057e-06, + "loss": 0.0956, + "step": 2686 + }, + { + "epoch": 1.7257546563904946, + "grad_norm": 1.0981191092227003, + "learning_rate": 4.561429029808241e-06, + "loss": 0.1618, + "step": 2687 + }, + { + "epoch": 1.7263969171483622, + "grad_norm": 1.0371958125358505, + "learning_rate": 4.55770623575731e-06, + "loss": 0.0862, + "step": 2688 + }, + { + "epoch": 1.72703917790623, + "grad_norm": 0.9941762521687011, + "learning_rate": 4.553983688817199e-06, + "loss": 0.0823, + "step": 2689 + }, + { + "epoch": 1.7276814386640975, + "grad_norm": 0.8729693599721254, + "learning_rate": 4.550261391067704e-06, + "loss": 0.0839, + "step": 2690 + }, + { + "epoch": 1.7283236994219653, + "grad_norm": 1.202360186098218, + "learning_rate": 4.546539344588486e-06, + "loss": 0.1522, + "step": 2691 + }, + { + "epoch": 1.728965960179833, + "grad_norm": 0.8227802141762643, + "learning_rate": 4.542817551459064e-06, + "loss": 0.0621, + "step": 2692 + }, + { + "epoch": 1.7296082209377008, + "grad_norm": 0.9119574449292119, + "learning_rate": 4.539096013758811e-06, + "loss": 0.0926, + "step": 2693 + }, + { + "epoch": 1.7302504816955684, + "grad_norm": 0.6986900256131805, + "learning_rate": 4.535374733566964e-06, + "loss": 0.0464, + "step": 2694 + }, + { + "epoch": 1.730892742453436, + "grad_norm": 0.962697153269985, + "learning_rate": 4.531653712962612e-06, + "loss": 0.0952, + "step": 2695 + }, + { + "epoch": 1.7315350032113037, + "grad_norm": 1.0503479538503244, + "learning_rate": 4.527932954024704e-06, + "loss": 0.1257, + "step": 2696 + }, + { + "epoch": 1.7321772639691715, + "grad_norm": 0.9958681774867991, + "learning_rate": 4.524212458832033e-06, + "loss": 0.1104, + "step": 2697 + }, + { + "epoch": 1.7328195247270393, + "grad_norm": 1.0522361422813211, + "learning_rate": 4.520492229463253e-06, + "loss": 0.1204, + "step": 2698 + }, + { + "epoch": 1.733461785484907, + "grad_norm": 1.5713678277618952, + "learning_rate": 4.516772267996867e-06, + "loss": 0.0431, + "step": 2699 + }, + { + "epoch": 1.7341040462427746, + "grad_norm": 0.7436862259198687, + "learning_rate": 4.513052576511227e-06, + "loss": 0.0418, + "step": 2700 + }, + { + "epoch": 1.7347463070006421, + "grad_norm": 0.6187177753330486, + "learning_rate": 4.509333157084538e-06, + "loss": 0.0329, + "step": 2701 + }, + { + "epoch": 1.73538856775851, + "grad_norm": 1.1632316388629, + "learning_rate": 4.50561401179485e-06, + "loss": 0.1189, + "step": 2702 + }, + { + "epoch": 1.7360308285163777, + "grad_norm": 0.8367103149030662, + "learning_rate": 4.501895142720059e-06, + "loss": 0.0719, + "step": 2703 + }, + { + "epoch": 1.7366730892742455, + "grad_norm": 0.6968413882851457, + "learning_rate": 4.498176551937911e-06, + "loss": 0.0429, + "step": 2704 + }, + { + "epoch": 1.737315350032113, + "grad_norm": 1.1106652839164781, + "learning_rate": 4.494458241525988e-06, + "loss": 0.1574, + "step": 2705 + }, + { + "epoch": 1.7379576107899806, + "grad_norm": 0.7511273716860024, + "learning_rate": 4.490740213561727e-06, + "loss": 0.0554, + "step": 2706 + }, + { + "epoch": 1.7385998715478483, + "grad_norm": 0.7991734821750702, + "learning_rate": 4.487022470122396e-06, + "loss": 0.0556, + "step": 2707 + }, + { + "epoch": 1.7392421323057161, + "grad_norm": 0.8392357915866722, + "learning_rate": 4.483305013285113e-06, + "loss": 0.0934, + "step": 2708 + }, + { + "epoch": 1.739884393063584, + "grad_norm": 0.8725123293058199, + "learning_rate": 4.47958784512683e-06, + "loss": 0.0501, + "step": 2709 + }, + { + "epoch": 1.7405266538214517, + "grad_norm": 0.9371284992518085, + "learning_rate": 4.47587096772434e-06, + "loss": 0.0603, + "step": 2710 + }, + { + "epoch": 1.7411689145793192, + "grad_norm": 0.9383201331746686, + "learning_rate": 4.4721543831542756e-06, + "loss": 0.0734, + "step": 2711 + }, + { + "epoch": 1.7418111753371868, + "grad_norm": 0.9176709409859056, + "learning_rate": 4.468438093493099e-06, + "loss": 0.0641, + "step": 2712 + }, + { + "epoch": 1.7424534360950545, + "grad_norm": 1.1580960389736616, + "learning_rate": 4.464722100817115e-06, + "loss": 0.1145, + "step": 2713 + }, + { + "epoch": 1.7430956968529223, + "grad_norm": 0.9604786780766579, + "learning_rate": 4.461006407202459e-06, + "loss": 0.0795, + "step": 2714 + }, + { + "epoch": 1.74373795761079, + "grad_norm": 0.632998372740083, + "learning_rate": 4.4572910147250995e-06, + "loss": 0.0456, + "step": 2715 + }, + { + "epoch": 1.7443802183686576, + "grad_norm": 0.9701840732407767, + "learning_rate": 4.453575925460838e-06, + "loss": 0.0733, + "step": 2716 + }, + { + "epoch": 1.7450224791265254, + "grad_norm": 0.8193874058592129, + "learning_rate": 4.449861141485304e-06, + "loss": 0.0714, + "step": 2717 + }, + { + "epoch": 1.745664739884393, + "grad_norm": 0.7908670083475055, + "learning_rate": 4.446146664873961e-06, + "loss": 0.0547, + "step": 2718 + }, + { + "epoch": 1.7463070006422607, + "grad_norm": 0.8233700858660167, + "learning_rate": 4.4424324977020985e-06, + "loss": 0.0622, + "step": 2719 + }, + { + "epoch": 1.7469492614001285, + "grad_norm": 0.7594702457177446, + "learning_rate": 4.438718642044828e-06, + "loss": 0.0607, + "step": 2720 + }, + { + "epoch": 1.7475915221579963, + "grad_norm": 1.0789498982598489, + "learning_rate": 4.435005099977093e-06, + "loss": 0.1272, + "step": 2721 + }, + { + "epoch": 1.7482337829158638, + "grad_norm": 1.0186865325707868, + "learning_rate": 4.431291873573663e-06, + "loss": 0.116, + "step": 2722 + }, + { + "epoch": 1.7488760436737314, + "grad_norm": 0.8098003540384353, + "learning_rate": 4.4275789649091265e-06, + "loss": 0.0575, + "step": 2723 + }, + { + "epoch": 1.7495183044315992, + "grad_norm": 1.1918001104036686, + "learning_rate": 4.4238663760578965e-06, + "loss": 0.0887, + "step": 2724 + }, + { + "epoch": 1.750160565189467, + "grad_norm": 0.7167732989105998, + "learning_rate": 4.420154109094208e-06, + "loss": 0.0453, + "step": 2725 + }, + { + "epoch": 1.7508028259473347, + "grad_norm": 1.5289612413649596, + "learning_rate": 4.416442166092114e-06, + "loss": 0.2388, + "step": 2726 + }, + { + "epoch": 1.7514450867052023, + "grad_norm": 0.869953547001251, + "learning_rate": 4.41273054912549e-06, + "loss": 0.05, + "step": 2727 + }, + { + "epoch": 1.75208734746307, + "grad_norm": 0.930348294560596, + "learning_rate": 4.409019260268025e-06, + "loss": 0.0852, + "step": 2728 + }, + { + "epoch": 1.7527296082209376, + "grad_norm": 0.993545760453377, + "learning_rate": 4.405308301593228e-06, + "loss": 0.0927, + "step": 2729 + }, + { + "epoch": 1.7533718689788054, + "grad_norm": 1.110499625987169, + "learning_rate": 4.401597675174423e-06, + "loss": 0.118, + "step": 2730 + }, + { + "epoch": 1.7540141297366731, + "grad_norm": 0.8204908741709559, + "learning_rate": 4.397887383084748e-06, + "loss": 0.0599, + "step": 2731 + }, + { + "epoch": 1.754656390494541, + "grad_norm": 0.8622665386069284, + "learning_rate": 4.3941774273971515e-06, + "loss": 0.0577, + "step": 2732 + }, + { + "epoch": 1.7552986512524085, + "grad_norm": 0.6708335933606963, + "learning_rate": 4.390467810184399e-06, + "loss": 0.0552, + "step": 2733 + }, + { + "epoch": 1.755940912010276, + "grad_norm": 0.9953387652922825, + "learning_rate": 4.386758533519063e-06, + "loss": 0.1062, + "step": 2734 + }, + { + "epoch": 1.7565831727681438, + "grad_norm": 0.8482630692535189, + "learning_rate": 4.383049599473532e-06, + "loss": 0.0632, + "step": 2735 + }, + { + "epoch": 1.7572254335260116, + "grad_norm": 0.695569995637514, + "learning_rate": 4.379341010119992e-06, + "loss": 0.0444, + "step": 2736 + }, + { + "epoch": 1.7578676942838793, + "grad_norm": 1.1233694854374467, + "learning_rate": 4.375632767530444e-06, + "loss": 0.1088, + "step": 2737 + }, + { + "epoch": 1.758509955041747, + "grad_norm": 0.6934273362419551, + "learning_rate": 4.371924873776696e-06, + "loss": 0.0482, + "step": 2738 + }, + { + "epoch": 1.7591522157996147, + "grad_norm": 0.7868037309466654, + "learning_rate": 4.368217330930358e-06, + "loss": 0.0574, + "step": 2739 + }, + { + "epoch": 1.7597944765574822, + "grad_norm": 0.73655468442881, + "learning_rate": 4.364510141062845e-06, + "loss": 0.0595, + "step": 2740 + }, + { + "epoch": 1.76043673731535, + "grad_norm": 0.8663329929942345, + "learning_rate": 4.360803306245377e-06, + "loss": 0.0723, + "step": 2741 + }, + { + "epoch": 1.7610789980732178, + "grad_norm": 0.6231497207703355, + "learning_rate": 4.35709682854897e-06, + "loss": 0.0331, + "step": 2742 + }, + { + "epoch": 1.7617212588310855, + "grad_norm": 0.9279410048499611, + "learning_rate": 4.3533907100444464e-06, + "loss": 0.094, + "step": 2743 + }, + { + "epoch": 1.762363519588953, + "grad_norm": 0.8892978515951684, + "learning_rate": 4.349684952802424e-06, + "loss": 0.0901, + "step": 2744 + }, + { + "epoch": 1.7630057803468207, + "grad_norm": 1.1274395184371442, + "learning_rate": 4.34597955889332e-06, + "loss": 0.1327, + "step": 2745 + }, + { + "epoch": 1.7636480411046884, + "grad_norm": 1.137725828748674, + "learning_rate": 4.34227453038735e-06, + "loss": 0.1439, + "step": 2746 + }, + { + "epoch": 1.7642903018625562, + "grad_norm": 0.6351297827293876, + "learning_rate": 4.338569869354524e-06, + "loss": 0.0391, + "step": 2747 + }, + { + "epoch": 1.764932562620424, + "grad_norm": 0.8797940259427235, + "learning_rate": 4.334865577864647e-06, + "loss": 0.086, + "step": 2748 + }, + { + "epoch": 1.7655748233782917, + "grad_norm": 1.021164324328881, + "learning_rate": 4.331161657987316e-06, + "loss": 0.111, + "step": 2749 + }, + { + "epoch": 1.7662170841361593, + "grad_norm": 0.6264033219464432, + "learning_rate": 4.327458111791923e-06, + "loss": 0.0276, + "step": 2750 + }, + { + "epoch": 1.7668593448940269, + "grad_norm": 1.0486211646124437, + "learning_rate": 4.3237549413476544e-06, + "loss": 0.0718, + "step": 2751 + }, + { + "epoch": 1.7675016056518946, + "grad_norm": 0.7601732072331402, + "learning_rate": 4.3200521487234746e-06, + "loss": 0.0393, + "step": 2752 + }, + { + "epoch": 1.7681438664097624, + "grad_norm": 0.6759910529793498, + "learning_rate": 4.316349735988149e-06, + "loss": 0.0537, + "step": 2753 + }, + { + "epoch": 1.7687861271676302, + "grad_norm": 0.7343278260757544, + "learning_rate": 4.312647705210226e-06, + "loss": 0.0613, + "step": 2754 + }, + { + "epoch": 1.7694283879254977, + "grad_norm": 0.7190531585225993, + "learning_rate": 4.3089460584580385e-06, + "loss": 0.0473, + "step": 2755 + }, + { + "epoch": 1.7700706486833655, + "grad_norm": 1.0346979625574426, + "learning_rate": 4.3052447977997106e-06, + "loss": 0.1108, + "step": 2756 + }, + { + "epoch": 1.770712909441233, + "grad_norm": 0.9247527761928694, + "learning_rate": 4.301543925303147e-06, + "loss": 0.0905, + "step": 2757 + }, + { + "epoch": 1.7713551701991008, + "grad_norm": 0.9998769610516567, + "learning_rate": 4.297843443036036e-06, + "loss": 0.1003, + "step": 2758 + }, + { + "epoch": 1.7719974309569686, + "grad_norm": 1.294320901787597, + "learning_rate": 4.294143353065848e-06, + "loss": 0.2012, + "step": 2759 + }, + { + "epoch": 1.7726396917148364, + "grad_norm": 1.0611659027024583, + "learning_rate": 4.290443657459832e-06, + "loss": 0.1355, + "step": 2760 + }, + { + "epoch": 1.773281952472704, + "grad_norm": 1.1073135417492348, + "learning_rate": 4.28674435828502e-06, + "loss": 0.1452, + "step": 2761 + }, + { + "epoch": 1.7739242132305715, + "grad_norm": 0.9007000979638777, + "learning_rate": 4.283045457608222e-06, + "loss": 0.0916, + "step": 2762 + }, + { + "epoch": 1.7745664739884393, + "grad_norm": 0.9288493561012447, + "learning_rate": 4.279346957496023e-06, + "loss": 0.0943, + "step": 2763 + }, + { + "epoch": 1.775208734746307, + "grad_norm": 0.8571355146561566, + "learning_rate": 4.2756488600147875e-06, + "loss": 0.0856, + "step": 2764 + }, + { + "epoch": 1.7758509955041748, + "grad_norm": 0.7607807610920553, + "learning_rate": 4.271951167230653e-06, + "loss": 0.0604, + "step": 2765 + }, + { + "epoch": 1.7764932562620424, + "grad_norm": 1.1943970811879758, + "learning_rate": 4.268253881209532e-06, + "loss": 0.1499, + "step": 2766 + }, + { + "epoch": 1.7771355170199101, + "grad_norm": 1.079976576503248, + "learning_rate": 4.264557004017107e-06, + "loss": 0.0819, + "step": 2767 + }, + { + "epoch": 1.7777777777777777, + "grad_norm": 1.0009925829628343, + "learning_rate": 4.260860537718834e-06, + "loss": 0.0982, + "step": 2768 + }, + { + "epoch": 1.7784200385356455, + "grad_norm": 1.2637920421595767, + "learning_rate": 4.257164484379942e-06, + "loss": 0.1553, + "step": 2769 + }, + { + "epoch": 1.7790622992935132, + "grad_norm": 1.2460681695799907, + "learning_rate": 4.253468846065425e-06, + "loss": 0.1721, + "step": 2770 + }, + { + "epoch": 1.779704560051381, + "grad_norm": 0.8579811071064475, + "learning_rate": 4.249773624840048e-06, + "loss": 0.0615, + "step": 2771 + }, + { + "epoch": 1.7803468208092486, + "grad_norm": 0.7548254539583403, + "learning_rate": 4.246078822768339e-06, + "loss": 0.0528, + "step": 2772 + }, + { + "epoch": 1.780989081567116, + "grad_norm": 0.7402104325396233, + "learning_rate": 4.242384441914599e-06, + "loss": 0.0611, + "step": 2773 + }, + { + "epoch": 1.7816313423249839, + "grad_norm": 0.9001028186190229, + "learning_rate": 4.2386904843428895e-06, + "loss": 0.0784, + "step": 2774 + }, + { + "epoch": 1.7822736030828517, + "grad_norm": 1.268223683085791, + "learning_rate": 4.234996952117031e-06, + "loss": 0.1401, + "step": 2775 + }, + { + "epoch": 1.7829158638407194, + "grad_norm": 0.9549007596830709, + "learning_rate": 4.2313038473006125e-06, + "loss": 0.0801, + "step": 2776 + }, + { + "epoch": 1.783558124598587, + "grad_norm": 0.8139257416137722, + "learning_rate": 4.227611171956984e-06, + "loss": 0.063, + "step": 2777 + }, + { + "epoch": 1.7842003853564548, + "grad_norm": 1.1904086792351827, + "learning_rate": 4.223918928149253e-06, + "loss": 0.1428, + "step": 2778 + }, + { + "epoch": 1.7848426461143223, + "grad_norm": 0.8821451522630378, + "learning_rate": 4.220227117940287e-06, + "loss": 0.0846, + "step": 2779 + }, + { + "epoch": 1.78548490687219, + "grad_norm": 1.0331777460081002, + "learning_rate": 4.21653574339271e-06, + "loss": 0.0919, + "step": 2780 + }, + { + "epoch": 1.7861271676300579, + "grad_norm": 1.0452674654425416, + "learning_rate": 4.212844806568906e-06, + "loss": 0.1105, + "step": 2781 + }, + { + "epoch": 1.7867694283879256, + "grad_norm": 0.8781867245159302, + "learning_rate": 4.209154309531012e-06, + "loss": 0.076, + "step": 2782 + }, + { + "epoch": 1.7874116891457932, + "grad_norm": 1.1513500473543374, + "learning_rate": 4.205464254340918e-06, + "loss": 0.1263, + "step": 2783 + }, + { + "epoch": 1.7880539499036607, + "grad_norm": 0.7139756907566776, + "learning_rate": 4.201774643060269e-06, + "loss": 0.0449, + "step": 2784 + }, + { + "epoch": 1.7886962106615285, + "grad_norm": 1.0818375004672314, + "learning_rate": 4.198085477750461e-06, + "loss": 0.1222, + "step": 2785 + }, + { + "epoch": 1.7893384714193963, + "grad_norm": 1.3811199825311984, + "learning_rate": 4.194396760472643e-06, + "loss": 0.1909, + "step": 2786 + }, + { + "epoch": 1.789980732177264, + "grad_norm": 0.901845418512504, + "learning_rate": 4.1907084932877114e-06, + "loss": 0.0844, + "step": 2787 + }, + { + "epoch": 1.7906229929351318, + "grad_norm": 0.7572374667842952, + "learning_rate": 4.187020678256313e-06, + "loss": 0.0565, + "step": 2788 + }, + { + "epoch": 1.7912652536929994, + "grad_norm": 1.0007025864837533, + "learning_rate": 4.183333317438837e-06, + "loss": 0.0975, + "step": 2789 + }, + { + "epoch": 1.791907514450867, + "grad_norm": 0.6249489668037379, + "learning_rate": 4.17964641289543e-06, + "loss": 0.0373, + "step": 2790 + }, + { + "epoch": 1.7925497752087347, + "grad_norm": 1.2105628854424444, + "learning_rate": 4.175959966685968e-06, + "loss": 0.1232, + "step": 2791 + }, + { + "epoch": 1.7931920359666025, + "grad_norm": 1.1915414893482563, + "learning_rate": 4.172273980870084e-06, + "loss": 0.1556, + "step": 2792 + }, + { + "epoch": 1.7938342967244703, + "grad_norm": 1.0114874314820979, + "learning_rate": 4.168588457507147e-06, + "loss": 0.0873, + "step": 2793 + }, + { + "epoch": 1.7944765574823378, + "grad_norm": 0.8798019764371976, + "learning_rate": 4.164903398656271e-06, + "loss": 0.0623, + "step": 2794 + }, + { + "epoch": 1.7951188182402054, + "grad_norm": 0.9019752478742994, + "learning_rate": 4.16121880637631e-06, + "loss": 0.0661, + "step": 2795 + }, + { + "epoch": 1.7957610789980731, + "grad_norm": 0.662045292080635, + "learning_rate": 4.157534682725855e-06, + "loss": 0.047, + "step": 2796 + }, + { + "epoch": 1.796403339755941, + "grad_norm": 0.8362141512617, + "learning_rate": 4.153851029763238e-06, + "loss": 0.0918, + "step": 2797 + }, + { + "epoch": 1.7970456005138087, + "grad_norm": 0.7048924721309592, + "learning_rate": 4.150167849546526e-06, + "loss": 0.0568, + "step": 2798 + }, + { + "epoch": 1.7976878612716765, + "grad_norm": 1.0095089252126883, + "learning_rate": 4.1464851441335215e-06, + "loss": 0.057, + "step": 2799 + }, + { + "epoch": 1.798330122029544, + "grad_norm": 0.9944473661643998, + "learning_rate": 4.1428029155817635e-06, + "loss": 0.0869, + "step": 2800 + }, + { + "epoch": 1.7989723827874116, + "grad_norm": 0.6386103129627198, + "learning_rate": 4.139121165948524e-06, + "loss": 0.0308, + "step": 2801 + }, + { + "epoch": 1.7996146435452793, + "grad_norm": 0.8404958083388783, + "learning_rate": 4.135439897290807e-06, + "loss": 0.0684, + "step": 2802 + }, + { + "epoch": 1.800256904303147, + "grad_norm": 0.8336419452609478, + "learning_rate": 4.131759111665349e-06, + "loss": 0.07, + "step": 2803 + }, + { + "epoch": 1.8008991650610149, + "grad_norm": 0.837482446763888, + "learning_rate": 4.128078811128615e-06, + "loss": 0.0679, + "step": 2804 + }, + { + "epoch": 1.8015414258188824, + "grad_norm": 1.1119754808549287, + "learning_rate": 4.124398997736799e-06, + "loss": 0.1154, + "step": 2805 + }, + { + "epoch": 1.8021836865767502, + "grad_norm": 0.7784093897965786, + "learning_rate": 4.120719673545825e-06, + "loss": 0.0634, + "step": 2806 + }, + { + "epoch": 1.8028259473346178, + "grad_norm": 0.8280380485548, + "learning_rate": 4.1170408406113414e-06, + "loss": 0.0658, + "step": 2807 + }, + { + "epoch": 1.8034682080924855, + "grad_norm": 1.0692225085276317, + "learning_rate": 4.113362500988722e-06, + "loss": 0.1027, + "step": 2808 + }, + { + "epoch": 1.8041104688503533, + "grad_norm": 0.7636097908595061, + "learning_rate": 4.109684656733068e-06, + "loss": 0.0613, + "step": 2809 + }, + { + "epoch": 1.804752729608221, + "grad_norm": 0.9079226543648736, + "learning_rate": 4.1060073098991995e-06, + "loss": 0.0675, + "step": 2810 + }, + { + "epoch": 1.8053949903660886, + "grad_norm": 0.8425103696820072, + "learning_rate": 4.102330462541662e-06, + "loss": 0.0689, + "step": 2811 + }, + { + "epoch": 1.8060372511239562, + "grad_norm": 0.5225284428972002, + "learning_rate": 4.0986541167147216e-06, + "loss": 0.0202, + "step": 2812 + }, + { + "epoch": 1.806679511881824, + "grad_norm": 1.047075762456711, + "learning_rate": 4.094978274472362e-06, + "loss": 0.1114, + "step": 2813 + }, + { + "epoch": 1.8073217726396917, + "grad_norm": 0.9197015675277593, + "learning_rate": 4.09130293786829e-06, + "loss": 0.071, + "step": 2814 + }, + { + "epoch": 1.8079640333975595, + "grad_norm": 0.9608579851322943, + "learning_rate": 4.087628108955923e-06, + "loss": 0.0968, + "step": 2815 + }, + { + "epoch": 1.808606294155427, + "grad_norm": 0.826378093962077, + "learning_rate": 4.0839537897884e-06, + "loss": 0.0613, + "step": 2816 + }, + { + "epoch": 1.8092485549132948, + "grad_norm": 0.8346198200058661, + "learning_rate": 4.080279982418574e-06, + "loss": 0.0689, + "step": 2817 + }, + { + "epoch": 1.8098908156711624, + "grad_norm": 1.0319219448476813, + "learning_rate": 4.0766066888990134e-06, + "loss": 0.1124, + "step": 2818 + }, + { + "epoch": 1.8105330764290302, + "grad_norm": 0.8580054474531352, + "learning_rate": 4.072933911281996e-06, + "loss": 0.0712, + "step": 2819 + }, + { + "epoch": 1.811175337186898, + "grad_norm": 1.2124056802976324, + "learning_rate": 4.069261651619514e-06, + "loss": 0.1418, + "step": 2820 + }, + { + "epoch": 1.8118175979447657, + "grad_norm": 0.7542969737146059, + "learning_rate": 4.065589911963271e-06, + "loss": 0.0353, + "step": 2821 + }, + { + "epoch": 1.8124598587026333, + "grad_norm": 0.9539911262671769, + "learning_rate": 4.061918694364677e-06, + "loss": 0.1029, + "step": 2822 + }, + { + "epoch": 1.8131021194605008, + "grad_norm": 0.7272250482042449, + "learning_rate": 4.058248000874851e-06, + "loss": 0.0487, + "step": 2823 + }, + { + "epoch": 1.8137443802183686, + "grad_norm": 0.7162301353441413, + "learning_rate": 4.054577833544623e-06, + "loss": 0.0632, + "step": 2824 + }, + { + "epoch": 1.8143866409762364, + "grad_norm": 0.5660677550907155, + "learning_rate": 4.050908194424525e-06, + "loss": 0.0234, + "step": 2825 + }, + { + "epoch": 1.8150289017341041, + "grad_norm": 0.7435087780648758, + "learning_rate": 4.047239085564794e-06, + "loss": 0.05, + "step": 2826 + }, + { + "epoch": 1.8156711624919717, + "grad_norm": 0.6661991656321133, + "learning_rate": 4.043570509015373e-06, + "loss": 0.0461, + "step": 2827 + }, + { + "epoch": 1.8163134232498395, + "grad_norm": 1.0908690883610583, + "learning_rate": 4.039902466825907e-06, + "loss": 0.1166, + "step": 2828 + }, + { + "epoch": 1.816955684007707, + "grad_norm": 1.3047786208973167, + "learning_rate": 4.036234961045743e-06, + "loss": 0.1359, + "step": 2829 + }, + { + "epoch": 1.8175979447655748, + "grad_norm": 0.9934437815230956, + "learning_rate": 4.032567993723922e-06, + "loss": 0.0846, + "step": 2830 + }, + { + "epoch": 1.8182402055234426, + "grad_norm": 0.840155079344863, + "learning_rate": 4.028901566909194e-06, + "loss": 0.0645, + "step": 2831 + }, + { + "epoch": 1.8188824662813103, + "grad_norm": 0.9332585521584158, + "learning_rate": 4.025235682650001e-06, + "loss": 0.0862, + "step": 2832 + }, + { + "epoch": 1.819524727039178, + "grad_norm": 1.0371288997699248, + "learning_rate": 4.0215703429944835e-06, + "loss": 0.1268, + "step": 2833 + }, + { + "epoch": 1.8201669877970454, + "grad_norm": 0.9425044655509793, + "learning_rate": 4.017905549990477e-06, + "loss": 0.0889, + "step": 2834 + }, + { + "epoch": 1.8208092485549132, + "grad_norm": 0.8469720625968871, + "learning_rate": 4.014241305685514e-06, + "loss": 0.0855, + "step": 2835 + }, + { + "epoch": 1.821451509312781, + "grad_norm": 1.269176653273299, + "learning_rate": 4.010577612126816e-06, + "loss": 0.1514, + "step": 2836 + }, + { + "epoch": 1.8220937700706488, + "grad_norm": 1.0965385215514825, + "learning_rate": 4.006914471361301e-06, + "loss": 0.0987, + "step": 2837 + }, + { + "epoch": 1.8227360308285165, + "grad_norm": 0.9282850507445154, + "learning_rate": 4.003251885435576e-06, + "loss": 0.0595, + "step": 2838 + }, + { + "epoch": 1.823378291586384, + "grad_norm": 0.976942765864037, + "learning_rate": 3.9995898563959365e-06, + "loss": 0.1018, + "step": 2839 + }, + { + "epoch": 1.8240205523442516, + "grad_norm": 0.614235021548916, + "learning_rate": 3.995928386288372e-06, + "loss": 0.0333, + "step": 2840 + }, + { + "epoch": 1.8246628131021194, + "grad_norm": 0.6407384203974772, + "learning_rate": 3.992267477158555e-06, + "loss": 0.0455, + "step": 2841 + }, + { + "epoch": 1.8253050738599872, + "grad_norm": 0.8216997556584227, + "learning_rate": 3.988607131051845e-06, + "loss": 0.0781, + "step": 2842 + }, + { + "epoch": 1.825947334617855, + "grad_norm": 1.1069420120473414, + "learning_rate": 3.9849473500132906e-06, + "loss": 0.1378, + "step": 2843 + }, + { + "epoch": 1.8265895953757225, + "grad_norm": 0.8436593413977237, + "learning_rate": 3.98128813608762e-06, + "loss": 0.0747, + "step": 2844 + }, + { + "epoch": 1.8272318561335903, + "grad_norm": 1.230782267322386, + "learning_rate": 3.9776294913192494e-06, + "loss": 0.1369, + "step": 2845 + }, + { + "epoch": 1.8278741168914578, + "grad_norm": 0.9714007321588101, + "learning_rate": 3.973971417752272e-06, + "loss": 0.0804, + "step": 2846 + }, + { + "epoch": 1.8285163776493256, + "grad_norm": 0.9715643011353605, + "learning_rate": 3.970313917430464e-06, + "loss": 0.0745, + "step": 2847 + }, + { + "epoch": 1.8291586384071934, + "grad_norm": 0.971895503044114, + "learning_rate": 3.966656992397282e-06, + "loss": 0.0865, + "step": 2848 + }, + { + "epoch": 1.8298008991650612, + "grad_norm": 0.7135507871804022, + "learning_rate": 3.9630006446958615e-06, + "loss": 0.038, + "step": 2849 + }, + { + "epoch": 1.8304431599229287, + "grad_norm": 0.6860170296509038, + "learning_rate": 3.959344876369013e-06, + "loss": 0.0343, + "step": 2850 + }, + { + "epoch": 1.8310854206807963, + "grad_norm": 0.974686524756167, + "learning_rate": 3.955689689459228e-06, + "loss": 0.0993, + "step": 2851 + }, + { + "epoch": 1.831727681438664, + "grad_norm": 0.8399489133186544, + "learning_rate": 3.952035086008668e-06, + "loss": 0.0752, + "step": 2852 + }, + { + "epoch": 1.8323699421965318, + "grad_norm": 0.8128392982265039, + "learning_rate": 3.948381068059171e-06, + "loss": 0.0628, + "step": 2853 + }, + { + "epoch": 1.8330122029543996, + "grad_norm": 0.8793398058434629, + "learning_rate": 3.944727637652247e-06, + "loss": 0.0877, + "step": 2854 + }, + { + "epoch": 1.8336544637122671, + "grad_norm": 0.9653413917723964, + "learning_rate": 3.941074796829078e-06, + "loss": 0.0572, + "step": 2855 + }, + { + "epoch": 1.834296724470135, + "grad_norm": 0.6252655162717176, + "learning_rate": 3.937422547630519e-06, + "loss": 0.0371, + "step": 2856 + }, + { + "epoch": 1.8349389852280025, + "grad_norm": 0.9205368845061058, + "learning_rate": 3.93377089209709e-06, + "loss": 0.0866, + "step": 2857 + }, + { + "epoch": 1.8355812459858702, + "grad_norm": 1.1324157736764076, + "learning_rate": 3.930119832268982e-06, + "loss": 0.124, + "step": 2858 + }, + { + "epoch": 1.836223506743738, + "grad_norm": 0.9116738557505435, + "learning_rate": 3.926469370186053e-06, + "loss": 0.0764, + "step": 2859 + }, + { + "epoch": 1.8368657675016058, + "grad_norm": 1.032501116872323, + "learning_rate": 3.922819507887828e-06, + "loss": 0.1262, + "step": 2860 + }, + { + "epoch": 1.8375080282594733, + "grad_norm": 0.8937975613132532, + "learning_rate": 3.919170247413496e-06, + "loss": 0.0827, + "step": 2861 + }, + { + "epoch": 1.838150289017341, + "grad_norm": 1.334569441602669, + "learning_rate": 3.915521590801907e-06, + "loss": 0.145, + "step": 2862 + }, + { + "epoch": 1.8387925497752087, + "grad_norm": 1.1345416059410611, + "learning_rate": 3.911873540091578e-06, + "loss": 0.0898, + "step": 2863 + }, + { + "epoch": 1.8394348105330764, + "grad_norm": 1.2524131679229493, + "learning_rate": 3.908226097320684e-06, + "loss": 0.148, + "step": 2864 + }, + { + "epoch": 1.8400770712909442, + "grad_norm": 0.9058234126220995, + "learning_rate": 3.904579264527063e-06, + "loss": 0.0817, + "step": 2865 + }, + { + "epoch": 1.8407193320488118, + "grad_norm": 1.1177485528625706, + "learning_rate": 3.900933043748211e-06, + "loss": 0.137, + "step": 2866 + }, + { + "epoch": 1.8413615928066795, + "grad_norm": 0.592428492553499, + "learning_rate": 3.897287437021283e-06, + "loss": 0.0409, + "step": 2867 + }, + { + "epoch": 1.842003853564547, + "grad_norm": 0.9787299436996255, + "learning_rate": 3.89364244638309e-06, + "loss": 0.0873, + "step": 2868 + }, + { + "epoch": 1.8426461143224149, + "grad_norm": 0.8561947915332607, + "learning_rate": 3.8899980738700995e-06, + "loss": 0.0557, + "step": 2869 + }, + { + "epoch": 1.8432883750802826, + "grad_norm": 0.9566942956416693, + "learning_rate": 3.88635432151843e-06, + "loss": 0.0819, + "step": 2870 + }, + { + "epoch": 1.8439306358381504, + "grad_norm": 0.7713049110626984, + "learning_rate": 3.88271119136386e-06, + "loss": 0.06, + "step": 2871 + }, + { + "epoch": 1.844572896596018, + "grad_norm": 1.0285820011381877, + "learning_rate": 3.879068685441814e-06, + "loss": 0.1331, + "step": 2872 + }, + { + "epoch": 1.8452151573538855, + "grad_norm": 0.7725288121213438, + "learning_rate": 3.875426805787374e-06, + "loss": 0.0444, + "step": 2873 + }, + { + "epoch": 1.8458574181117533, + "grad_norm": 1.0650959624911789, + "learning_rate": 3.8717855544352655e-06, + "loss": 0.1198, + "step": 2874 + }, + { + "epoch": 1.846499678869621, + "grad_norm": 0.6761516362838701, + "learning_rate": 3.868144933419869e-06, + "loss": 0.0402, + "step": 2875 + }, + { + "epoch": 1.8471419396274888, + "grad_norm": 1.162706099531168, + "learning_rate": 3.86450494477521e-06, + "loss": 0.1209, + "step": 2876 + }, + { + "epoch": 1.8477842003853564, + "grad_norm": 0.9252926143355593, + "learning_rate": 3.8608655905349586e-06, + "loss": 0.0736, + "step": 2877 + }, + { + "epoch": 1.8484264611432242, + "grad_norm": 0.8967288768016477, + "learning_rate": 3.857226872732434e-06, + "loss": 0.0859, + "step": 2878 + }, + { + "epoch": 1.8490687219010917, + "grad_norm": 0.6009408908092582, + "learning_rate": 3.8535887934005985e-06, + "loss": 0.0398, + "step": 2879 + }, + { + "epoch": 1.8497109826589595, + "grad_norm": 0.8015514560828653, + "learning_rate": 3.849951354572057e-06, + "loss": 0.0757, + "step": 2880 + }, + { + "epoch": 1.8503532434168273, + "grad_norm": 0.9081676339655608, + "learning_rate": 3.846314558279059e-06, + "loss": 0.0692, + "step": 2881 + }, + { + "epoch": 1.850995504174695, + "grad_norm": 0.7884090138780456, + "learning_rate": 3.84267840655349e-06, + "loss": 0.0488, + "step": 2882 + }, + { + "epoch": 1.8516377649325626, + "grad_norm": 0.6710015495502027, + "learning_rate": 3.8390429014268806e-06, + "loss": 0.0496, + "step": 2883 + }, + { + "epoch": 1.8522800256904302, + "grad_norm": 0.8405908465548937, + "learning_rate": 3.835408044930402e-06, + "loss": 0.0749, + "step": 2884 + }, + { + "epoch": 1.852922286448298, + "grad_norm": 0.9853143106495316, + "learning_rate": 3.831773839094852e-06, + "loss": 0.0937, + "step": 2885 + }, + { + "epoch": 1.8535645472061657, + "grad_norm": 0.9121024259265514, + "learning_rate": 3.828140285950676e-06, + "loss": 0.0756, + "step": 2886 + }, + { + "epoch": 1.8542068079640335, + "grad_norm": 0.9028997007420334, + "learning_rate": 3.824507387527949e-06, + "loss": 0.0829, + "step": 2887 + }, + { + "epoch": 1.8548490687219013, + "grad_norm": 0.7614085431129581, + "learning_rate": 3.820875145856384e-06, + "loss": 0.0559, + "step": 2888 + }, + { + "epoch": 1.8554913294797688, + "grad_norm": 0.9187902867688061, + "learning_rate": 3.817243562965324e-06, + "loss": 0.0725, + "step": 2889 + }, + { + "epoch": 1.8561335902376364, + "grad_norm": 0.8049916398676852, + "learning_rate": 3.8136126408837454e-06, + "loss": 0.065, + "step": 2890 + }, + { + "epoch": 1.8567758509955041, + "grad_norm": 0.908717358392128, + "learning_rate": 3.8099823816402547e-06, + "loss": 0.068, + "step": 2891 + }, + { + "epoch": 1.857418111753372, + "grad_norm": 1.1068059208420769, + "learning_rate": 3.806352787263089e-06, + "loss": 0.0923, + "step": 2892 + }, + { + "epoch": 1.8580603725112397, + "grad_norm": 0.9197499268135894, + "learning_rate": 3.802723859780112e-06, + "loss": 0.0604, + "step": 2893 + }, + { + "epoch": 1.8587026332691072, + "grad_norm": 1.1043271697725603, + "learning_rate": 3.7990956012188168e-06, + "loss": 0.1039, + "step": 2894 + }, + { + "epoch": 1.859344894026975, + "grad_norm": 0.7957793125550808, + "learning_rate": 3.7954680136063217e-06, + "loss": 0.0662, + "step": 2895 + }, + { + "epoch": 1.8599871547848426, + "grad_norm": 0.9450460445767436, + "learning_rate": 3.791841098969371e-06, + "loss": 0.0745, + "step": 2896 + }, + { + "epoch": 1.8606294155427103, + "grad_norm": 1.022712986117439, + "learning_rate": 3.788214859334332e-06, + "loss": 0.0758, + "step": 2897 + }, + { + "epoch": 1.861271676300578, + "grad_norm": 1.1060549276497935, + "learning_rate": 3.7845892967271963e-06, + "loss": 0.111, + "step": 2898 + }, + { + "epoch": 1.8619139370584459, + "grad_norm": 0.9200827102031615, + "learning_rate": 3.7809644131735747e-06, + "loss": 0.0741, + "step": 2899 + }, + { + "epoch": 1.8625561978163134, + "grad_norm": 0.7362591433913377, + "learning_rate": 3.777340210698704e-06, + "loss": 0.0393, + "step": 2900 + }, + { + "epoch": 1.863198458574181, + "grad_norm": 0.7981321274437443, + "learning_rate": 3.773716691327432e-06, + "loss": 0.0618, + "step": 2901 + }, + { + "epoch": 1.8638407193320488, + "grad_norm": 0.9080021013644838, + "learning_rate": 3.7700938570842317e-06, + "loss": 0.0601, + "step": 2902 + }, + { + "epoch": 1.8644829800899165, + "grad_norm": 1.378651016322129, + "learning_rate": 3.7664717099931915e-06, + "loss": 0.1328, + "step": 2903 + }, + { + "epoch": 1.8651252408477843, + "grad_norm": 1.1200556744519565, + "learning_rate": 3.7628502520780134e-06, + "loss": 0.1211, + "step": 2904 + }, + { + "epoch": 1.8657675016056519, + "grad_norm": 0.7476856972188421, + "learning_rate": 3.75922948536202e-06, + "loss": 0.0255, + "step": 2905 + }, + { + "epoch": 1.8664097623635196, + "grad_norm": 1.2376701482018746, + "learning_rate": 3.755609411868142e-06, + "loss": 0.1228, + "step": 2906 + }, + { + "epoch": 1.8670520231213872, + "grad_norm": 1.1022534973099245, + "learning_rate": 3.7519900336189265e-06, + "loss": 0.1058, + "step": 2907 + }, + { + "epoch": 1.867694283879255, + "grad_norm": 0.7301494517243078, + "learning_rate": 3.7483713526365294e-06, + "loss": 0.0442, + "step": 2908 + }, + { + "epoch": 1.8683365446371227, + "grad_norm": 1.0301221612474736, + "learning_rate": 3.744753370942719e-06, + "loss": 0.0947, + "step": 2909 + }, + { + "epoch": 1.8689788053949905, + "grad_norm": 1.1936497390875087, + "learning_rate": 3.741136090558872e-06, + "loss": 0.1473, + "step": 2910 + }, + { + "epoch": 1.869621066152858, + "grad_norm": 0.892474777064154, + "learning_rate": 3.7375195135059728e-06, + "loss": 0.0714, + "step": 2911 + }, + { + "epoch": 1.8702633269107256, + "grad_norm": 1.1603471540302668, + "learning_rate": 3.7339036418046147e-06, + "loss": 0.1371, + "step": 2912 + }, + { + "epoch": 1.8709055876685934, + "grad_norm": 0.7518116761626168, + "learning_rate": 3.730288477474996e-06, + "loss": 0.0513, + "step": 2913 + }, + { + "epoch": 1.8715478484264612, + "grad_norm": 0.818025525854742, + "learning_rate": 3.726674022536918e-06, + "loss": 0.0535, + "step": 2914 + }, + { + "epoch": 1.872190109184329, + "grad_norm": 1.194043894707273, + "learning_rate": 3.723060279009789e-06, + "loss": 0.1566, + "step": 2915 + }, + { + "epoch": 1.8728323699421965, + "grad_norm": 1.0103650691102344, + "learning_rate": 3.7194472489126176e-06, + "loss": 0.1098, + "step": 2916 + }, + { + "epoch": 1.8734746307000643, + "grad_norm": 1.0225067731904154, + "learning_rate": 3.715834934264014e-06, + "loss": 0.1141, + "step": 2917 + }, + { + "epoch": 1.8741168914579318, + "grad_norm": 0.9829488129602236, + "learning_rate": 3.7122233370821885e-06, + "loss": 0.1205, + "step": 2918 + }, + { + "epoch": 1.8747591522157996, + "grad_norm": 0.9154352424371087, + "learning_rate": 3.7086124593849514e-06, + "loss": 0.1022, + "step": 2919 + }, + { + "epoch": 1.8754014129736674, + "grad_norm": 0.9020285967092594, + "learning_rate": 3.705002303189711e-06, + "loss": 0.0854, + "step": 2920 + }, + { + "epoch": 1.8760436737315351, + "grad_norm": 0.7837477562041723, + "learning_rate": 3.7013928705134705e-06, + "loss": 0.0771, + "step": 2921 + }, + { + "epoch": 1.8766859344894027, + "grad_norm": 0.9582313272300912, + "learning_rate": 3.6977841633728332e-06, + "loss": 0.064, + "step": 2922 + }, + { + "epoch": 1.8773281952472702, + "grad_norm": 0.804073066054215, + "learning_rate": 3.6941761837839928e-06, + "loss": 0.0537, + "step": 2923 + }, + { + "epoch": 1.877970456005138, + "grad_norm": 0.9578477179426967, + "learning_rate": 3.69056893376274e-06, + "loss": 0.0734, + "step": 2924 + }, + { + "epoch": 1.8786127167630058, + "grad_norm": 0.7056846970968517, + "learning_rate": 3.686962415324452e-06, + "loss": 0.0378, + "step": 2925 + }, + { + "epoch": 1.8792549775208736, + "grad_norm": 0.7017132543115379, + "learning_rate": 3.683356630484104e-06, + "loss": 0.0508, + "step": 2926 + }, + { + "epoch": 1.8798972382787413, + "grad_norm": 0.869556232709636, + "learning_rate": 3.679751581256258e-06, + "loss": 0.078, + "step": 2927 + }, + { + "epoch": 1.8805394990366089, + "grad_norm": 0.8254341032411013, + "learning_rate": 3.676147269655065e-06, + "loss": 0.0556, + "step": 2928 + }, + { + "epoch": 1.8811817597944764, + "grad_norm": 1.2872367134691984, + "learning_rate": 3.672543697694265e-06, + "loss": 0.1657, + "step": 2929 + }, + { + "epoch": 1.8818240205523442, + "grad_norm": 1.1696223424460004, + "learning_rate": 3.6689408673871845e-06, + "loss": 0.1347, + "step": 2930 + }, + { + "epoch": 1.882466281310212, + "grad_norm": 0.7540730103182514, + "learning_rate": 3.665338780746736e-06, + "loss": 0.0567, + "step": 2931 + }, + { + "epoch": 1.8831085420680798, + "grad_norm": 1.1645483270969201, + "learning_rate": 3.6617374397854134e-06, + "loss": 0.1209, + "step": 2932 + }, + { + "epoch": 1.8837508028259473, + "grad_norm": 0.9187914269431178, + "learning_rate": 3.658136846515298e-06, + "loss": 0.0821, + "step": 2933 + }, + { + "epoch": 1.8843930635838149, + "grad_norm": 0.7510272544322159, + "learning_rate": 3.6545370029480515e-06, + "loss": 0.0625, + "step": 2934 + }, + { + "epoch": 1.8850353243416826, + "grad_norm": 0.9648888112831114, + "learning_rate": 3.6509379110949173e-06, + "loss": 0.0959, + "step": 2935 + }, + { + "epoch": 1.8856775850995504, + "grad_norm": 0.8079007406627348, + "learning_rate": 3.647339572966718e-06, + "loss": 0.0547, + "step": 2936 + }, + { + "epoch": 1.8863198458574182, + "grad_norm": 0.8352525484839789, + "learning_rate": 3.6437419905738547e-06, + "loss": 0.0518, + "step": 2937 + }, + { + "epoch": 1.886962106615286, + "grad_norm": 0.8821929484817066, + "learning_rate": 3.6401451659263105e-06, + "loss": 0.0578, + "step": 2938 + }, + { + "epoch": 1.8876043673731535, + "grad_norm": 0.9651903530654986, + "learning_rate": 3.636549101033642e-06, + "loss": 0.0759, + "step": 2939 + }, + { + "epoch": 1.888246628131021, + "grad_norm": 1.1399744994798855, + "learning_rate": 3.6329537979049757e-06, + "loss": 0.1142, + "step": 2940 + }, + { + "epoch": 1.8888888888888888, + "grad_norm": 1.0844014477307582, + "learning_rate": 3.6293592585490223e-06, + "loss": 0.1139, + "step": 2941 + }, + { + "epoch": 1.8895311496467566, + "grad_norm": 0.7247904264968166, + "learning_rate": 3.6257654849740585e-06, + "loss": 0.0373, + "step": 2942 + }, + { + "epoch": 1.8901734104046244, + "grad_norm": 1.0081611837318836, + "learning_rate": 3.6221724791879406e-06, + "loss": 0.0834, + "step": 2943 + }, + { + "epoch": 1.890815671162492, + "grad_norm": 0.8729877566384475, + "learning_rate": 3.6185802431980875e-06, + "loss": 0.0603, + "step": 2944 + }, + { + "epoch": 1.8914579319203597, + "grad_norm": 1.1924952754027454, + "learning_rate": 3.6149887790114944e-06, + "loss": 0.117, + "step": 2945 + }, + { + "epoch": 1.8921001926782273, + "grad_norm": 0.9743398393834182, + "learning_rate": 3.6113980886347213e-06, + "loss": 0.0929, + "step": 2946 + }, + { + "epoch": 1.892742453436095, + "grad_norm": 0.7441596403959851, + "learning_rate": 3.607808174073899e-06, + "loss": 0.0624, + "step": 2947 + }, + { + "epoch": 1.8933847141939628, + "grad_norm": 1.3147336493424235, + "learning_rate": 3.6042190373347212e-06, + "loss": 0.1681, + "step": 2948 + }, + { + "epoch": 1.8940269749518306, + "grad_norm": 0.8188628006547808, + "learning_rate": 3.60063068042245e-06, + "loss": 0.0569, + "step": 2949 + }, + { + "epoch": 1.8946692357096981, + "grad_norm": 0.997516990989186, + "learning_rate": 3.5970431053419112e-06, + "loss": 0.0812, + "step": 2950 + }, + { + "epoch": 1.8953114964675657, + "grad_norm": 0.8270233612162645, + "learning_rate": 3.593456314097493e-06, + "loss": 0.0722, + "step": 2951 + }, + { + "epoch": 1.8959537572254335, + "grad_norm": 0.6947286234307778, + "learning_rate": 3.5898703086931474e-06, + "loss": 0.0437, + "step": 2952 + }, + { + "epoch": 1.8965960179833012, + "grad_norm": 0.8560546867606786, + "learning_rate": 3.586285091132384e-06, + "loss": 0.0617, + "step": 2953 + }, + { + "epoch": 1.897238278741169, + "grad_norm": 0.9971338941087963, + "learning_rate": 3.5827006634182754e-06, + "loss": 0.0796, + "step": 2954 + }, + { + "epoch": 1.8978805394990366, + "grad_norm": 1.0606471963050332, + "learning_rate": 3.579117027553455e-06, + "loss": 0.0815, + "step": 2955 + }, + { + "epoch": 1.8985228002569043, + "grad_norm": 0.8742948970776608, + "learning_rate": 3.5755341855401048e-06, + "loss": 0.0729, + "step": 2956 + }, + { + "epoch": 1.899165061014772, + "grad_norm": 0.5891678070616618, + "learning_rate": 3.571952139379972e-06, + "loss": 0.0259, + "step": 2957 + }, + { + "epoch": 1.8998073217726397, + "grad_norm": 1.0407520990558075, + "learning_rate": 3.5683708910743563e-06, + "loss": 0.1185, + "step": 2958 + }, + { + "epoch": 1.9004495825305074, + "grad_norm": 0.8385522781987236, + "learning_rate": 3.564790442624111e-06, + "loss": 0.0549, + "step": 2959 + }, + { + "epoch": 1.9010918432883752, + "grad_norm": 0.7848691913706894, + "learning_rate": 3.5612107960296437e-06, + "loss": 0.0455, + "step": 2960 + }, + { + "epoch": 1.9017341040462428, + "grad_norm": 0.5691087262268305, + "learning_rate": 3.557631953290914e-06, + "loss": 0.033, + "step": 2961 + }, + { + "epoch": 1.9023763648041103, + "grad_norm": 0.9703810836631783, + "learning_rate": 3.5540539164074306e-06, + "loss": 0.1041, + "step": 2962 + }, + { + "epoch": 1.903018625561978, + "grad_norm": 0.7426836365652704, + "learning_rate": 3.5504766873782552e-06, + "loss": 0.0566, + "step": 2963 + }, + { + "epoch": 1.9036608863198459, + "grad_norm": 0.8919138642202195, + "learning_rate": 3.5469002682019936e-06, + "loss": 0.0871, + "step": 2964 + }, + { + "epoch": 1.9043031470777136, + "grad_norm": 1.0802190008510666, + "learning_rate": 3.5433246608768025e-06, + "loss": 0.1191, + "step": 2965 + }, + { + "epoch": 1.9049454078355812, + "grad_norm": 1.244990694057629, + "learning_rate": 3.5397498674003843e-06, + "loss": 0.1667, + "step": 2966 + }, + { + "epoch": 1.905587668593449, + "grad_norm": 0.9460299791742679, + "learning_rate": 3.536175889769987e-06, + "loss": 0.0915, + "step": 2967 + }, + { + "epoch": 1.9062299293513165, + "grad_norm": 1.277990506154646, + "learning_rate": 3.532602729982402e-06, + "loss": 0.1601, + "step": 2968 + }, + { + "epoch": 1.9068721901091843, + "grad_norm": 1.1416643035835832, + "learning_rate": 3.5290303900339634e-06, + "loss": 0.1289, + "step": 2969 + }, + { + "epoch": 1.907514450867052, + "grad_norm": 0.7648913627149118, + "learning_rate": 3.5254588719205494e-06, + "loss": 0.0631, + "step": 2970 + }, + { + "epoch": 1.9081567116249198, + "grad_norm": 0.9375149199875578, + "learning_rate": 3.5218881776375773e-06, + "loss": 0.0815, + "step": 2971 + }, + { + "epoch": 1.9087989723827874, + "grad_norm": 0.8548005700531414, + "learning_rate": 3.5183183091800035e-06, + "loss": 0.0604, + "step": 2972 + }, + { + "epoch": 1.909441233140655, + "grad_norm": 0.6759498405945674, + "learning_rate": 3.5147492685423247e-06, + "loss": 0.0418, + "step": 2973 + }, + { + "epoch": 1.9100834938985227, + "grad_norm": 1.223606847678229, + "learning_rate": 3.5111810577185734e-06, + "loss": 0.1081, + "step": 2974 + }, + { + "epoch": 1.9107257546563905, + "grad_norm": 0.7340860193717454, + "learning_rate": 3.5076136787023207e-06, + "loss": 0.0582, + "step": 2975 + }, + { + "epoch": 1.9113680154142583, + "grad_norm": 0.7458352431908826, + "learning_rate": 3.50404713348667e-06, + "loss": 0.0565, + "step": 2976 + }, + { + "epoch": 1.912010276172126, + "grad_norm": 0.759579233442881, + "learning_rate": 3.5004814240642624e-06, + "loss": 0.0584, + "step": 2977 + }, + { + "epoch": 1.9126525369299936, + "grad_norm": 1.1608841131952232, + "learning_rate": 3.4969165524272695e-06, + "loss": 0.1138, + "step": 2978 + }, + { + "epoch": 1.9132947976878611, + "grad_norm": 1.0847349352955125, + "learning_rate": 3.4933525205673977e-06, + "loss": 0.1088, + "step": 2979 + }, + { + "epoch": 1.913937058445729, + "grad_norm": 0.7832321413855684, + "learning_rate": 3.489789330475877e-06, + "loss": 0.0675, + "step": 2980 + }, + { + "epoch": 1.9145793192035967, + "grad_norm": 0.8205021949836777, + "learning_rate": 3.4862269841434758e-06, + "loss": 0.0629, + "step": 2981 + }, + { + "epoch": 1.9152215799614645, + "grad_norm": 1.0885176026566035, + "learning_rate": 3.4826654835604867e-06, + "loss": 0.083, + "step": 2982 + }, + { + "epoch": 1.915863840719332, + "grad_norm": 1.2138320784850376, + "learning_rate": 3.4791048307167307e-06, + "loss": 0.1502, + "step": 2983 + }, + { + "epoch": 1.9165061014771998, + "grad_norm": 1.048206220218184, + "learning_rate": 3.475545027601554e-06, + "loss": 0.1167, + "step": 2984 + }, + { + "epoch": 1.9171483622350673, + "grad_norm": 0.9035219984433737, + "learning_rate": 3.47198607620383e-06, + "loss": 0.0695, + "step": 2985 + }, + { + "epoch": 1.9177906229929351, + "grad_norm": 1.1223702910894837, + "learning_rate": 3.4684279785119558e-06, + "loss": 0.113, + "step": 2986 + }, + { + "epoch": 1.918432883750803, + "grad_norm": 0.793499960300989, + "learning_rate": 3.464870736513849e-06, + "loss": 0.0545, + "step": 2987 + }, + { + "epoch": 1.9190751445086707, + "grad_norm": 0.9616741669677115, + "learning_rate": 3.461314352196952e-06, + "loss": 0.1027, + "step": 2988 + }, + { + "epoch": 1.9197174052665382, + "grad_norm": 0.8495336854934573, + "learning_rate": 3.4577588275482272e-06, + "loss": 0.0546, + "step": 2989 + }, + { + "epoch": 1.9203596660244058, + "grad_norm": 0.7224287837482255, + "learning_rate": 3.4542041645541567e-06, + "loss": 0.0605, + "step": 2990 + }, + { + "epoch": 1.9210019267822736, + "grad_norm": 1.0553905927891516, + "learning_rate": 3.450650365200742e-06, + "loss": 0.1106, + "step": 2991 + }, + { + "epoch": 1.9216441875401413, + "grad_norm": 0.9650588432856437, + "learning_rate": 3.4470974314734996e-06, + "loss": 0.0863, + "step": 2992 + }, + { + "epoch": 1.922286448298009, + "grad_norm": 0.7902869952930555, + "learning_rate": 3.4435453653574634e-06, + "loss": 0.0468, + "step": 2993 + }, + { + "epoch": 1.9229287090558767, + "grad_norm": 0.8737163333462142, + "learning_rate": 3.4399941688371884e-06, + "loss": 0.0613, + "step": 2994 + }, + { + "epoch": 1.9235709698137444, + "grad_norm": 0.44557376991981923, + "learning_rate": 3.436443843896732e-06, + "loss": 0.0222, + "step": 2995 + }, + { + "epoch": 1.924213230571612, + "grad_norm": 0.9692525810210455, + "learning_rate": 3.4328943925196735e-06, + "loss": 0.0795, + "step": 2996 + }, + { + "epoch": 1.9248554913294798, + "grad_norm": 0.9485501706648779, + "learning_rate": 3.429345816689101e-06, + "loss": 0.0951, + "step": 2997 + }, + { + "epoch": 1.9254977520873475, + "grad_norm": 0.721984090365319, + "learning_rate": 3.425798118387616e-06, + "loss": 0.0347, + "step": 2998 + }, + { + "epoch": 1.9261400128452153, + "grad_norm": 0.7010590596639742, + "learning_rate": 3.422251299597327e-06, + "loss": 0.0454, + "step": 2999 + }, + { + "epoch": 1.9267822736030829, + "grad_norm": 0.8757104964853932, + "learning_rate": 3.4187053622998524e-06, + "loss": 0.0587, + "step": 3000 + }, + { + "epoch": 1.9274245343609504, + "grad_norm": 0.5870944283703782, + "learning_rate": 3.4151603084763175e-06, + "loss": 0.0271, + "step": 3001 + }, + { + "epoch": 1.9280667951188182, + "grad_norm": 1.1245945601502039, + "learning_rate": 3.4116161401073544e-06, + "loss": 0.1305, + "step": 3002 + }, + { + "epoch": 1.928709055876686, + "grad_norm": 1.0304621568514758, + "learning_rate": 3.408072859173099e-06, + "loss": 0.096, + "step": 3003 + }, + { + "epoch": 1.9293513166345537, + "grad_norm": 1.0216619863669494, + "learning_rate": 3.4045304676531952e-06, + "loss": 0.0738, + "step": 3004 + }, + { + "epoch": 1.9299935773924213, + "grad_norm": 0.7684091605200478, + "learning_rate": 3.400988967526785e-06, + "loss": 0.0497, + "step": 3005 + }, + { + "epoch": 1.930635838150289, + "grad_norm": 0.873466111978884, + "learning_rate": 3.397448360772516e-06, + "loss": 0.0672, + "step": 3006 + }, + { + "epoch": 1.9312780989081566, + "grad_norm": 0.6885089979095547, + "learning_rate": 3.3939086493685357e-06, + "loss": 0.0502, + "step": 3007 + }, + { + "epoch": 1.9319203596660244, + "grad_norm": 0.889652108739319, + "learning_rate": 3.390369835292491e-06, + "loss": 0.0697, + "step": 3008 + }, + { + "epoch": 1.9325626204238922, + "grad_norm": 0.9885087915030093, + "learning_rate": 3.3868319205215284e-06, + "loss": 0.1114, + "step": 3009 + }, + { + "epoch": 1.93320488118176, + "grad_norm": 1.0915287374024283, + "learning_rate": 3.3832949070322907e-06, + "loss": 0.1092, + "step": 3010 + }, + { + "epoch": 1.9338471419396275, + "grad_norm": 0.710787963522594, + "learning_rate": 3.379758796800917e-06, + "loss": 0.04, + "step": 3011 + }, + { + "epoch": 1.934489402697495, + "grad_norm": 1.1055199431225844, + "learning_rate": 3.376223591803043e-06, + "loss": 0.0773, + "step": 3012 + }, + { + "epoch": 1.9351316634553628, + "grad_norm": 1.1746307988197398, + "learning_rate": 3.372689294013797e-06, + "loss": 0.1927, + "step": 3013 + }, + { + "epoch": 1.9357739242132306, + "grad_norm": 1.0574958745397423, + "learning_rate": 3.3691559054078026e-06, + "loss": 0.0948, + "step": 3014 + }, + { + "epoch": 1.9364161849710984, + "grad_norm": 0.9660479171717982, + "learning_rate": 3.365623427959175e-06, + "loss": 0.0795, + "step": 3015 + }, + { + "epoch": 1.937058445728966, + "grad_norm": 0.8176105563707229, + "learning_rate": 3.362091863641519e-06, + "loss": 0.0751, + "step": 3016 + }, + { + "epoch": 1.9377007064868337, + "grad_norm": 0.7880594813062464, + "learning_rate": 3.35856121442793e-06, + "loss": 0.0682, + "step": 3017 + }, + { + "epoch": 1.9383429672447012, + "grad_norm": 0.8853236175436879, + "learning_rate": 3.355031482290994e-06, + "loss": 0.0783, + "step": 3018 + }, + { + "epoch": 1.938985228002569, + "grad_norm": 1.196799107513488, + "learning_rate": 3.3515026692027774e-06, + "loss": 0.1084, + "step": 3019 + }, + { + "epoch": 1.9396274887604368, + "grad_norm": 0.7948565474131816, + "learning_rate": 3.3479747771348424e-06, + "loss": 0.0614, + "step": 3020 + }, + { + "epoch": 1.9402697495183046, + "grad_norm": 0.8387227778585516, + "learning_rate": 3.344447808058232e-06, + "loss": 0.0597, + "step": 3021 + }, + { + "epoch": 1.940912010276172, + "grad_norm": 1.1240240624197733, + "learning_rate": 3.3409217639434743e-06, + "loss": 0.109, + "step": 3022 + }, + { + "epoch": 1.9415542710340397, + "grad_norm": 0.7358147873921663, + "learning_rate": 3.33739664676058e-06, + "loss": 0.051, + "step": 3023 + }, + { + "epoch": 1.9421965317919074, + "grad_norm": 0.5246407886783181, + "learning_rate": 3.3338724584790427e-06, + "loss": 0.0229, + "step": 3024 + }, + { + "epoch": 1.9428387925497752, + "grad_norm": 0.8853457716947386, + "learning_rate": 3.330349201067836e-06, + "loss": 0.0741, + "step": 3025 + }, + { + "epoch": 1.943481053307643, + "grad_norm": 0.8647141862763005, + "learning_rate": 3.3268268764954157e-06, + "loss": 0.0667, + "step": 3026 + }, + { + "epoch": 1.9441233140655108, + "grad_norm": 1.0058510408967094, + "learning_rate": 3.323305486729713e-06, + "loss": 0.0903, + "step": 3027 + }, + { + "epoch": 1.9447655748233783, + "grad_norm": 0.9888351121105367, + "learning_rate": 3.31978503373814e-06, + "loss": 0.0895, + "step": 3028 + }, + { + "epoch": 1.9454078355812459, + "grad_norm": 0.5966726995366075, + "learning_rate": 3.3162655194875825e-06, + "loss": 0.044, + "step": 3029 + }, + { + "epoch": 1.9460500963391136, + "grad_norm": 0.7623476921269912, + "learning_rate": 3.312746945944405e-06, + "loss": 0.0619, + "step": 3030 + }, + { + "epoch": 1.9466923570969814, + "grad_norm": 0.7473332936360979, + "learning_rate": 3.3092293150744425e-06, + "loss": 0.0509, + "step": 3031 + }, + { + "epoch": 1.9473346178548492, + "grad_norm": 0.8562338078416403, + "learning_rate": 3.305712628843009e-06, + "loss": 0.0651, + "step": 3032 + }, + { + "epoch": 1.9479768786127167, + "grad_norm": 1.1626462014423602, + "learning_rate": 3.302196889214886e-06, + "loss": 0.1129, + "step": 3033 + }, + { + "epoch": 1.9486191393705845, + "grad_norm": 0.8502845894294964, + "learning_rate": 3.2986820981543293e-06, + "loss": 0.0677, + "step": 3034 + }, + { + "epoch": 1.949261400128452, + "grad_norm": 0.9534292634433363, + "learning_rate": 3.2951682576250575e-06, + "loss": 0.0865, + "step": 3035 + }, + { + "epoch": 1.9499036608863198, + "grad_norm": 0.7136736964217026, + "learning_rate": 3.291655369590269e-06, + "loss": 0.044, + "step": 3036 + }, + { + "epoch": 1.9505459216441876, + "grad_norm": 1.2206935381596162, + "learning_rate": 3.288143436012622e-06, + "loss": 0.1124, + "step": 3037 + }, + { + "epoch": 1.9511881824020554, + "grad_norm": 0.6617883435867118, + "learning_rate": 3.2846324588542442e-06, + "loss": 0.0489, + "step": 3038 + }, + { + "epoch": 1.951830443159923, + "grad_norm": 1.0330939982047485, + "learning_rate": 3.2811224400767292e-06, + "loss": 0.0976, + "step": 3039 + }, + { + "epoch": 1.9524727039177905, + "grad_norm": 0.8496776711025215, + "learning_rate": 3.277613381641136e-06, + "loss": 0.0483, + "step": 3040 + }, + { + "epoch": 1.9531149646756583, + "grad_norm": 1.171093887638817, + "learning_rate": 3.2741052855079857e-06, + "loss": 0.0851, + "step": 3041 + }, + { + "epoch": 1.953757225433526, + "grad_norm": 0.9881188849982768, + "learning_rate": 3.270598153637259e-06, + "loss": 0.0863, + "step": 3042 + }, + { + "epoch": 1.9543994861913938, + "grad_norm": 0.9010586998858546, + "learning_rate": 3.2670919879884035e-06, + "loss": 0.1001, + "step": 3043 + }, + { + "epoch": 1.9550417469492614, + "grad_norm": 0.6406371910921136, + "learning_rate": 3.2635867905203223e-06, + "loss": 0.0392, + "step": 3044 + }, + { + "epoch": 1.9556840077071291, + "grad_norm": 0.771840230038381, + "learning_rate": 3.2600825631913815e-06, + "loss": 0.0584, + "step": 3045 + }, + { + "epoch": 1.9563262684649967, + "grad_norm": 0.901398763416549, + "learning_rate": 3.256579307959401e-06, + "loss": 0.0677, + "step": 3046 + }, + { + "epoch": 1.9569685292228645, + "grad_norm": 1.0945554681814689, + "learning_rate": 3.253077026781662e-06, + "loss": 0.1068, + "step": 3047 + }, + { + "epoch": 1.9576107899807322, + "grad_norm": 1.0681621936774695, + "learning_rate": 3.2495757216148965e-06, + "loss": 0.1197, + "step": 3048 + }, + { + "epoch": 1.9582530507386, + "grad_norm": 0.7443500089618831, + "learning_rate": 3.246075394415299e-06, + "loss": 0.0494, + "step": 3049 + }, + { + "epoch": 1.9588953114964676, + "grad_norm": 1.0314275763468157, + "learning_rate": 3.242576047138507e-06, + "loss": 0.1025, + "step": 3050 + }, + { + "epoch": 1.9595375722543351, + "grad_norm": 0.809605069897798, + "learning_rate": 3.239077681739618e-06, + "loss": 0.0751, + "step": 3051 + }, + { + "epoch": 1.9601798330122029, + "grad_norm": 0.8903189681533896, + "learning_rate": 3.2355803001731785e-06, + "loss": 0.0587, + "step": 3052 + }, + { + "epoch": 1.9608220937700707, + "grad_norm": 1.0686289128718203, + "learning_rate": 3.232083904393187e-06, + "loss": 0.0906, + "step": 3053 + }, + { + "epoch": 1.9614643545279384, + "grad_norm": 1.052586712384587, + "learning_rate": 3.22858849635309e-06, + "loss": 0.1126, + "step": 3054 + }, + { + "epoch": 1.962106615285806, + "grad_norm": 0.9005632157109555, + "learning_rate": 3.2250940780057816e-06, + "loss": 0.0723, + "step": 3055 + }, + { + "epoch": 1.9627488760436738, + "grad_norm": 0.9540117353547968, + "learning_rate": 3.2216006513036034e-06, + "loss": 0.0645, + "step": 3056 + }, + { + "epoch": 1.9633911368015413, + "grad_norm": 1.043390090400645, + "learning_rate": 3.2181082181983447e-06, + "loss": 0.0677, + "step": 3057 + }, + { + "epoch": 1.964033397559409, + "grad_norm": 0.9481914705745825, + "learning_rate": 3.2146167806412353e-06, + "loss": 0.0794, + "step": 3058 + }, + { + "epoch": 1.9646756583172769, + "grad_norm": 1.1602293271550956, + "learning_rate": 3.2111263405829534e-06, + "loss": 0.1204, + "step": 3059 + }, + { + "epoch": 1.9653179190751446, + "grad_norm": 0.9852408818775543, + "learning_rate": 3.2076368999736175e-06, + "loss": 0.1065, + "step": 3060 + }, + { + "epoch": 1.9659601798330122, + "grad_norm": 1.063614698490368, + "learning_rate": 3.204148460762789e-06, + "loss": 0.1416, + "step": 3061 + }, + { + "epoch": 1.9666024405908797, + "grad_norm": 0.7662231162194328, + "learning_rate": 3.2006610248994687e-06, + "loss": 0.0433, + "step": 3062 + }, + { + "epoch": 1.9672447013487475, + "grad_norm": 1.0209429747989074, + "learning_rate": 3.1971745943320962e-06, + "loss": 0.0861, + "step": 3063 + }, + { + "epoch": 1.9678869621066153, + "grad_norm": 0.9692772241946914, + "learning_rate": 3.1936891710085514e-06, + "loss": 0.1027, + "step": 3064 + }, + { + "epoch": 1.968529222864483, + "grad_norm": 0.9852731370120219, + "learning_rate": 3.190204756876153e-06, + "loss": 0.1103, + "step": 3065 + }, + { + "epoch": 1.9691714836223508, + "grad_norm": 1.1178686112168534, + "learning_rate": 3.1867213538816484e-06, + "loss": 0.1052, + "step": 3066 + }, + { + "epoch": 1.9698137443802184, + "grad_norm": 0.987394697378942, + "learning_rate": 3.183238963971228e-06, + "loss": 0.0829, + "step": 3067 + }, + { + "epoch": 1.970456005138086, + "grad_norm": 1.129583798837321, + "learning_rate": 3.1797575890905125e-06, + "loss": 0.1135, + "step": 3068 + }, + { + "epoch": 1.9710982658959537, + "grad_norm": 0.9516603476137947, + "learning_rate": 3.176277231184556e-06, + "loss": 0.0747, + "step": 3069 + }, + { + "epoch": 1.9717405266538215, + "grad_norm": 0.7287596214226449, + "learning_rate": 3.172797892197845e-06, + "loss": 0.0587, + "step": 3070 + }, + { + "epoch": 1.9723827874116893, + "grad_norm": 0.6736062155810427, + "learning_rate": 3.169319574074298e-06, + "loss": 0.0455, + "step": 3071 + }, + { + "epoch": 1.9730250481695568, + "grad_norm": 0.9492669177561809, + "learning_rate": 3.1658422787572586e-06, + "loss": 0.0674, + "step": 3072 + }, + { + "epoch": 1.9736673089274244, + "grad_norm": 0.750867560769236, + "learning_rate": 3.1623660081895058e-06, + "loss": 0.0625, + "step": 3073 + }, + { + "epoch": 1.9743095696852921, + "grad_norm": 1.1662126184614732, + "learning_rate": 3.158890764313236e-06, + "loss": 0.1112, + "step": 3074 + }, + { + "epoch": 1.97495183044316, + "grad_norm": 0.7509793895536151, + "learning_rate": 3.155416549070083e-06, + "loss": 0.0329, + "step": 3075 + }, + { + "epoch": 1.9755940912010277, + "grad_norm": 0.8173590708995909, + "learning_rate": 3.1519433644010998e-06, + "loss": 0.0636, + "step": 3076 + }, + { + "epoch": 1.9762363519588955, + "grad_norm": 1.0440367810165103, + "learning_rate": 3.1484712122467655e-06, + "loss": 0.0968, + "step": 3077 + }, + { + "epoch": 1.976878612716763, + "grad_norm": 0.6397155828831599, + "learning_rate": 3.14500009454698e-06, + "loss": 0.0243, + "step": 3078 + }, + { + "epoch": 1.9775208734746306, + "grad_norm": 1.0916541463414409, + "learning_rate": 3.141530013241068e-06, + "loss": 0.1268, + "step": 3079 + }, + { + "epoch": 1.9781631342324983, + "grad_norm": 0.932381222126332, + "learning_rate": 3.138060970267774e-06, + "loss": 0.087, + "step": 3080 + }, + { + "epoch": 1.9788053949903661, + "grad_norm": 0.9105463064317432, + "learning_rate": 3.1345929675652632e-06, + "loss": 0.0712, + "step": 3081 + }, + { + "epoch": 1.979447655748234, + "grad_norm": 0.9800016088443863, + "learning_rate": 3.131126007071117e-06, + "loss": 0.09, + "step": 3082 + }, + { + "epoch": 1.9800899165061014, + "grad_norm": 1.074548084003827, + "learning_rate": 3.1276600907223374e-06, + "loss": 0.1081, + "step": 3083 + }, + { + "epoch": 1.9807321772639692, + "grad_norm": 0.993701943139973, + "learning_rate": 3.1241952204553426e-06, + "loss": 0.1074, + "step": 3084 + }, + { + "epoch": 1.9813744380218368, + "grad_norm": 0.6480168317987296, + "learning_rate": 3.120731398205964e-06, + "loss": 0.041, + "step": 3085 + }, + { + "epoch": 1.9820166987797045, + "grad_norm": 1.3042022064255845, + "learning_rate": 3.1172686259094497e-06, + "loss": 0.1761, + "step": 3086 + }, + { + "epoch": 1.9826589595375723, + "grad_norm": 1.1077370346095292, + "learning_rate": 3.1138069055004628e-06, + "loss": 0.1352, + "step": 3087 + }, + { + "epoch": 1.98330122029544, + "grad_norm": 0.8979775527481259, + "learning_rate": 3.1103462389130756e-06, + "loss": 0.0685, + "step": 3088 + }, + { + "epoch": 1.9839434810533076, + "grad_norm": 0.6605605312543426, + "learning_rate": 3.106886628080773e-06, + "loss": 0.0276, + "step": 3089 + }, + { + "epoch": 1.9845857418111752, + "grad_norm": 1.6025037443206882, + "learning_rate": 3.1034280749364476e-06, + "loss": 0.0492, + "step": 3090 + }, + { + "epoch": 1.985228002569043, + "grad_norm": 0.6902334276179038, + "learning_rate": 3.0999705814124037e-06, + "loss": 0.0397, + "step": 3091 + }, + { + "epoch": 1.9858702633269107, + "grad_norm": 0.7537042309208735, + "learning_rate": 3.0965141494403545e-06, + "loss": 0.0417, + "step": 3092 + }, + { + "epoch": 1.9865125240847785, + "grad_norm": 0.9948615295937988, + "learning_rate": 3.093058780951418e-06, + "loss": 0.0987, + "step": 3093 + }, + { + "epoch": 1.987154784842646, + "grad_norm": 0.9276783610493667, + "learning_rate": 3.0896044778761184e-06, + "loss": 0.0975, + "step": 3094 + }, + { + "epoch": 1.9877970456005138, + "grad_norm": 1.0605985375198292, + "learning_rate": 3.086151242144386e-06, + "loss": 0.1069, + "step": 3095 + }, + { + "epoch": 1.9884393063583814, + "grad_norm": 0.9738614650259233, + "learning_rate": 3.0826990756855528e-06, + "loss": 0.0868, + "step": 3096 + }, + { + "epoch": 1.9890815671162492, + "grad_norm": 1.1657114707310217, + "learning_rate": 3.0792479804283544e-06, + "loss": 0.1489, + "step": 3097 + }, + { + "epoch": 1.989723827874117, + "grad_norm": 1.0168192197013752, + "learning_rate": 3.075797958300927e-06, + "loss": 0.0932, + "step": 3098 + }, + { + "epoch": 1.9903660886319847, + "grad_norm": 0.7212672837088828, + "learning_rate": 3.0723490112308095e-06, + "loss": 0.0574, + "step": 3099 + }, + { + "epoch": 1.9910083493898523, + "grad_norm": 1.02116055138548, + "learning_rate": 3.0689011411449377e-06, + "loss": 0.0576, + "step": 3100 + }, + { + "epoch": 1.9916506101477198, + "grad_norm": 0.867720114049946, + "learning_rate": 3.065454349969647e-06, + "loss": 0.0772, + "step": 3101 + }, + { + "epoch": 1.9922928709055876, + "grad_norm": 0.8873709616412891, + "learning_rate": 3.0620086396306704e-06, + "loss": 0.0613, + "step": 3102 + }, + { + "epoch": 1.9929351316634554, + "grad_norm": 0.962523297812572, + "learning_rate": 3.058564012053134e-06, + "loss": 0.0814, + "step": 3103 + }, + { + "epoch": 1.9935773924213231, + "grad_norm": 1.029292335750507, + "learning_rate": 3.0551204691615667e-06, + "loss": 0.1153, + "step": 3104 + }, + { + "epoch": 1.9942196531791907, + "grad_norm": 1.2240019359810652, + "learning_rate": 3.0516780128798794e-06, + "loss": 0.1202, + "step": 3105 + }, + { + "epoch": 1.9948619139370585, + "grad_norm": 1.0354195696558302, + "learning_rate": 3.048236645131386e-06, + "loss": 0.089, + "step": 3106 + }, + { + "epoch": 1.995504174694926, + "grad_norm": 0.8750631744853701, + "learning_rate": 3.044796367838788e-06, + "loss": 0.057, + "step": 3107 + }, + { + "epoch": 1.9961464354527938, + "grad_norm": 1.0156618451460366, + "learning_rate": 3.0413571829241783e-06, + "loss": 0.0972, + "step": 3108 + }, + { + "epoch": 1.9967886962106616, + "grad_norm": 0.8071403419489109, + "learning_rate": 3.03791909230904e-06, + "loss": 0.0637, + "step": 3109 + }, + { + "epoch": 1.9974309569685293, + "grad_norm": 1.314657987912778, + "learning_rate": 3.0344820979142443e-06, + "loss": 0.1117, + "step": 3110 + }, + { + "epoch": 1.998073217726397, + "grad_norm": 1.0549607790716307, + "learning_rate": 3.0310462016600505e-06, + "loss": 0.0989, + "step": 3111 + }, + { + "epoch": 1.9987154784842645, + "grad_norm": 1.0309227519753943, + "learning_rate": 3.027611405466104e-06, + "loss": 0.0864, + "step": 3112 + }, + { + "epoch": 1.9993577392421322, + "grad_norm": 0.8140631319181374, + "learning_rate": 3.0241777112514336e-06, + "loss": 0.0511, + "step": 3113 + }, + { + "epoch": 2.0, + "grad_norm": 0.6347421106069666, + "learning_rate": 3.020745120934455e-06, + "loss": 0.0524, + "step": 3114 + }, + { + "epoch": 2.0006422607578678, + "grad_norm": 0.791001041454552, + "learning_rate": 3.017313636432968e-06, + "loss": 0.0499, + "step": 3115 + }, + { + "epoch": 2.0012845215157355, + "grad_norm": 0.7673935124414003, + "learning_rate": 3.013883259664151e-06, + "loss": 0.0586, + "step": 3116 + }, + { + "epoch": 2.001926782273603, + "grad_norm": 0.6348900429165952, + "learning_rate": 3.010453992544567e-06, + "loss": 0.0412, + "step": 3117 + }, + { + "epoch": 2.0025690430314707, + "grad_norm": 0.5030949675115707, + "learning_rate": 3.0070258369901573e-06, + "loss": 0.0322, + "step": 3118 + }, + { + "epoch": 2.0032113037893384, + "grad_norm": 0.7176021751403286, + "learning_rate": 3.0035987949162415e-06, + "loss": 0.0246, + "step": 3119 + }, + { + "epoch": 2.003853564547206, + "grad_norm": 1.1538559192338844, + "learning_rate": 3.0001728682375202e-06, + "loss": 0.0521, + "step": 3120 + }, + { + "epoch": 2.004495825305074, + "grad_norm": 0.5199152456876868, + "learning_rate": 2.9967480588680666e-06, + "loss": 0.0263, + "step": 3121 + }, + { + "epoch": 2.0051380860629417, + "grad_norm": 0.6124086342051501, + "learning_rate": 2.9933243687213317e-06, + "loss": 0.0201, + "step": 3122 + }, + { + "epoch": 2.005780346820809, + "grad_norm": 0.5193420288922929, + "learning_rate": 2.989901799710142e-06, + "loss": 0.0238, + "step": 3123 + }, + { + "epoch": 2.006422607578677, + "grad_norm": 0.589423314475855, + "learning_rate": 2.9864803537466963e-06, + "loss": 0.034, + "step": 3124 + }, + { + "epoch": 2.0070648683365446, + "grad_norm": 0.6586719957243019, + "learning_rate": 2.9830600327425664e-06, + "loss": 0.0312, + "step": 3125 + }, + { + "epoch": 2.0077071290944124, + "grad_norm": 0.6446941878160275, + "learning_rate": 2.979640838608696e-06, + "loss": 0.0394, + "step": 3126 + }, + { + "epoch": 2.00834938985228, + "grad_norm": 0.5913147319084, + "learning_rate": 2.9762227732554007e-06, + "loss": 0.0296, + "step": 3127 + }, + { + "epoch": 2.0089916506101475, + "grad_norm": 0.8119621203091677, + "learning_rate": 2.972805838592362e-06, + "loss": 0.0385, + "step": 3128 + }, + { + "epoch": 2.0096339113680153, + "grad_norm": 0.6710612536701301, + "learning_rate": 2.9693900365286288e-06, + "loss": 0.0431, + "step": 3129 + }, + { + "epoch": 2.010276172125883, + "grad_norm": 0.5465817115061097, + "learning_rate": 2.9659753689726234e-06, + "loss": 0.0277, + "step": 3130 + }, + { + "epoch": 2.010918432883751, + "grad_norm": 0.6383346047462043, + "learning_rate": 2.9625618378321287e-06, + "loss": 0.0282, + "step": 3131 + }, + { + "epoch": 2.0115606936416186, + "grad_norm": 0.7905384120975517, + "learning_rate": 2.9591494450142957e-06, + "loss": 0.0379, + "step": 3132 + }, + { + "epoch": 2.0122029543994864, + "grad_norm": 0.8727695950949885, + "learning_rate": 2.9557381924256366e-06, + "loss": 0.0478, + "step": 3133 + }, + { + "epoch": 2.0128452151573537, + "grad_norm": 0.9667526498428343, + "learning_rate": 2.9523280819720294e-06, + "loss": 0.0682, + "step": 3134 + }, + { + "epoch": 2.0134874759152215, + "grad_norm": 0.5666648137894439, + "learning_rate": 2.9489191155587125e-06, + "loss": 0.0347, + "step": 3135 + }, + { + "epoch": 2.0141297366730893, + "grad_norm": 0.9158358867173219, + "learning_rate": 2.9455112950902874e-06, + "loss": 0.0418, + "step": 3136 + }, + { + "epoch": 2.014771997430957, + "grad_norm": 0.420574096948687, + "learning_rate": 2.9421046224707095e-06, + "loss": 0.0146, + "step": 3137 + }, + { + "epoch": 2.015414258188825, + "grad_norm": 0.9769026202432285, + "learning_rate": 2.938699099603299e-06, + "loss": 0.0506, + "step": 3138 + }, + { + "epoch": 2.0160565189466926, + "grad_norm": 0.7020344523020101, + "learning_rate": 2.935294728390731e-06, + "loss": 0.0213, + "step": 3139 + }, + { + "epoch": 2.01669877970456, + "grad_norm": 0.6435896550703812, + "learning_rate": 2.9318915107350376e-06, + "loss": 0.0328, + "step": 3140 + }, + { + "epoch": 2.0173410404624277, + "grad_norm": 0.5022596731951449, + "learning_rate": 2.9284894485376057e-06, + "loss": 0.0157, + "step": 3141 + }, + { + "epoch": 2.0179833012202955, + "grad_norm": 0.6633520455313245, + "learning_rate": 2.9250885436991793e-06, + "loss": 0.024, + "step": 3142 + }, + { + "epoch": 2.0186255619781632, + "grad_norm": 0.7167269959785598, + "learning_rate": 2.9216887981198553e-06, + "loss": 0.0322, + "step": 3143 + }, + { + "epoch": 2.019267822736031, + "grad_norm": 0.7091134766038767, + "learning_rate": 2.9182902136990767e-06, + "loss": 0.0272, + "step": 3144 + }, + { + "epoch": 2.0199100834938983, + "grad_norm": 0.7277480648026479, + "learning_rate": 2.914892792335645e-06, + "loss": 0.0382, + "step": 3145 + }, + { + "epoch": 2.020552344251766, + "grad_norm": 0.6299819106666618, + "learning_rate": 2.9114965359277097e-06, + "loss": 0.0308, + "step": 3146 + }, + { + "epoch": 2.021194605009634, + "grad_norm": 0.5802015793143854, + "learning_rate": 2.908101446372769e-06, + "loss": 0.0281, + "step": 3147 + }, + { + "epoch": 2.0218368657675017, + "grad_norm": 1.2078599902856055, + "learning_rate": 2.904707525567669e-06, + "loss": 0.0479, + "step": 3148 + }, + { + "epoch": 2.0224791265253694, + "grad_norm": 0.46093345069764297, + "learning_rate": 2.9013147754086044e-06, + "loss": 0.0247, + "step": 3149 + }, + { + "epoch": 2.023121387283237, + "grad_norm": 0.6535775171790454, + "learning_rate": 2.8979231977911127e-06, + "loss": 0.0162, + "step": 3150 + }, + { + "epoch": 2.0237636480411045, + "grad_norm": 1.0481064495884176, + "learning_rate": 2.894532794610082e-06, + "loss": 0.0512, + "step": 3151 + }, + { + "epoch": 2.0244059087989723, + "grad_norm": 0.9554100366785427, + "learning_rate": 2.8911435677597355e-06, + "loss": 0.0518, + "step": 3152 + }, + { + "epoch": 2.02504816955684, + "grad_norm": 0.5534164862943641, + "learning_rate": 2.8877555191336464e-06, + "loss": 0.0302, + "step": 3153 + }, + { + "epoch": 2.025690430314708, + "grad_norm": 0.9928787504177474, + "learning_rate": 2.884368650624725e-06, + "loss": 0.0402, + "step": 3154 + }, + { + "epoch": 2.0263326910725756, + "grad_norm": 0.8423096176376066, + "learning_rate": 2.8809829641252297e-06, + "loss": 0.0345, + "step": 3155 + }, + { + "epoch": 2.026974951830443, + "grad_norm": 0.7221432793975352, + "learning_rate": 2.8775984615267503e-06, + "loss": 0.0302, + "step": 3156 + }, + { + "epoch": 2.0276172125883107, + "grad_norm": 0.8896846058216423, + "learning_rate": 2.8742151447202192e-06, + "loss": 0.0463, + "step": 3157 + }, + { + "epoch": 2.0282594733461785, + "grad_norm": 0.7170069994914806, + "learning_rate": 2.870833015595905e-06, + "loss": 0.0208, + "step": 3158 + }, + { + "epoch": 2.0289017341040463, + "grad_norm": 0.9373110047397465, + "learning_rate": 2.8674520760434145e-06, + "loss": 0.0407, + "step": 3159 + }, + { + "epoch": 2.029543994861914, + "grad_norm": 0.7978844455242573, + "learning_rate": 2.864072327951686e-06, + "loss": 0.0357, + "step": 3160 + }, + { + "epoch": 2.030186255619782, + "grad_norm": 0.6277573327508668, + "learning_rate": 2.8606937732089945e-06, + "loss": 0.0219, + "step": 3161 + }, + { + "epoch": 2.030828516377649, + "grad_norm": 0.9935377583070452, + "learning_rate": 2.857316413702951e-06, + "loss": 0.0583, + "step": 3162 + }, + { + "epoch": 2.031470777135517, + "grad_norm": 0.5323709935891954, + "learning_rate": 2.853940251320494e-06, + "loss": 0.027, + "step": 3163 + }, + { + "epoch": 2.0321130378933847, + "grad_norm": 0.7328226419189565, + "learning_rate": 2.850565287947896e-06, + "loss": 0.0208, + "step": 3164 + }, + { + "epoch": 2.0327552986512525, + "grad_norm": 0.6566679441573162, + "learning_rate": 2.8471915254707594e-06, + "loss": 0.0288, + "step": 3165 + }, + { + "epoch": 2.0333975594091203, + "grad_norm": 1.0808836744791666, + "learning_rate": 2.843818965774015e-06, + "loss": 0.0407, + "step": 3166 + }, + { + "epoch": 2.0340398201669876, + "grad_norm": 0.7822165987728327, + "learning_rate": 2.8404476107419217e-06, + "loss": 0.0355, + "step": 3167 + }, + { + "epoch": 2.0346820809248554, + "grad_norm": 0.7063560890039912, + "learning_rate": 2.8370774622580644e-06, + "loss": 0.0321, + "step": 3168 + }, + { + "epoch": 2.035324341682723, + "grad_norm": 0.9479025664751141, + "learning_rate": 2.8337085222053562e-06, + "loss": 0.0404, + "step": 3169 + }, + { + "epoch": 2.035966602440591, + "grad_norm": 0.4888889841908648, + "learning_rate": 2.830340792466034e-06, + "loss": 0.0181, + "step": 3170 + }, + { + "epoch": 2.0366088631984587, + "grad_norm": 1.264009174156846, + "learning_rate": 2.826974274921657e-06, + "loss": 0.0273, + "step": 3171 + }, + { + "epoch": 2.0372511239563265, + "grad_norm": 0.8076143947216825, + "learning_rate": 2.823608971453109e-06, + "loss": 0.0409, + "step": 3172 + }, + { + "epoch": 2.037893384714194, + "grad_norm": 0.8027348491882942, + "learning_rate": 2.820244883940595e-06, + "loss": 0.0409, + "step": 3173 + }, + { + "epoch": 2.0385356454720616, + "grad_norm": 0.6831473200065632, + "learning_rate": 2.8168820142636405e-06, + "loss": 0.0166, + "step": 3174 + }, + { + "epoch": 2.0391779062299293, + "grad_norm": 0.5878728855240748, + "learning_rate": 2.8135203643010923e-06, + "loss": 0.0272, + "step": 3175 + }, + { + "epoch": 2.039820166987797, + "grad_norm": 0.9847362140813842, + "learning_rate": 2.810159935931109e-06, + "loss": 0.052, + "step": 3176 + }, + { + "epoch": 2.040462427745665, + "grad_norm": 0.7566198067300856, + "learning_rate": 2.806800731031176e-06, + "loss": 0.0397, + "step": 3177 + }, + { + "epoch": 2.041104688503532, + "grad_norm": 0.6626836347686486, + "learning_rate": 2.803442751478091e-06, + "loss": 0.0269, + "step": 3178 + }, + { + "epoch": 2.0417469492614, + "grad_norm": 0.7899276367290219, + "learning_rate": 2.8000859991479655e-06, + "loss": 0.0437, + "step": 3179 + }, + { + "epoch": 2.0423892100192678, + "grad_norm": 0.7695300690905394, + "learning_rate": 2.7967304759162274e-06, + "loss": 0.0367, + "step": 3180 + }, + { + "epoch": 2.0430314707771355, + "grad_norm": 0.9013114225543752, + "learning_rate": 2.793376183657618e-06, + "loss": 0.0374, + "step": 3181 + }, + { + "epoch": 2.0436737315350033, + "grad_norm": 0.942732068928622, + "learning_rate": 2.7900231242461884e-06, + "loss": 0.0402, + "step": 3182 + }, + { + "epoch": 2.044315992292871, + "grad_norm": 0.9670173381453222, + "learning_rate": 2.7866712995553056e-06, + "loss": 0.022, + "step": 3183 + }, + { + "epoch": 2.0449582530507384, + "grad_norm": 1.0033937849292704, + "learning_rate": 2.783320711457641e-06, + "loss": 0.0584, + "step": 3184 + }, + { + "epoch": 2.045600513808606, + "grad_norm": 1.0746852773528048, + "learning_rate": 2.779971361825179e-06, + "loss": 0.031, + "step": 3185 + }, + { + "epoch": 2.046242774566474, + "grad_norm": 0.8074936760094681, + "learning_rate": 2.7766232525292104e-06, + "loss": 0.0479, + "step": 3186 + }, + { + "epoch": 2.0468850353243417, + "grad_norm": 0.7325590261054015, + "learning_rate": 2.773276385440335e-06, + "loss": 0.0239, + "step": 3187 + }, + { + "epoch": 2.0475272960822095, + "grad_norm": 0.9729421853668642, + "learning_rate": 2.769930762428454e-06, + "loss": 0.039, + "step": 3188 + }, + { + "epoch": 2.0481695568400773, + "grad_norm": 0.6550351039676715, + "learning_rate": 2.7665863853627816e-06, + "loss": 0.0322, + "step": 3189 + }, + { + "epoch": 2.0488118175979446, + "grad_norm": 0.887223210997878, + "learning_rate": 2.763243256111828e-06, + "loss": 0.0308, + "step": 3190 + }, + { + "epoch": 2.0494540783558124, + "grad_norm": 0.5917025019940955, + "learning_rate": 2.7599013765434135e-06, + "loss": 0.0182, + "step": 3191 + }, + { + "epoch": 2.05009633911368, + "grad_norm": 0.7122721230848131, + "learning_rate": 2.756560748524649e-06, + "loss": 0.0322, + "step": 3192 + }, + { + "epoch": 2.050738599871548, + "grad_norm": 0.8856072714760063, + "learning_rate": 2.7532213739219578e-06, + "loss": 0.0247, + "step": 3193 + }, + { + "epoch": 2.0513808606294157, + "grad_norm": 0.6947844650532712, + "learning_rate": 2.749883254601057e-06, + "loss": 0.0259, + "step": 3194 + }, + { + "epoch": 2.052023121387283, + "grad_norm": 0.8045353885288288, + "learning_rate": 2.746546392426963e-06, + "loss": 0.026, + "step": 3195 + }, + { + "epoch": 2.052665382145151, + "grad_norm": 0.6236062021600042, + "learning_rate": 2.743210789263992e-06, + "loss": 0.0398, + "step": 3196 + }, + { + "epoch": 2.0533076429030186, + "grad_norm": 0.5292059331113167, + "learning_rate": 2.7398764469757546e-06, + "loss": 0.0178, + "step": 3197 + }, + { + "epoch": 2.0539499036608864, + "grad_norm": 0.8083917079330777, + "learning_rate": 2.736543367425156e-06, + "loss": 0.032, + "step": 3198 + }, + { + "epoch": 2.054592164418754, + "grad_norm": 0.9324362824548952, + "learning_rate": 2.7332115524744006e-06, + "loss": 0.0493, + "step": 3199 + }, + { + "epoch": 2.055234425176622, + "grad_norm": 1.059119816404513, + "learning_rate": 2.729881003984981e-06, + "loss": 0.0523, + "step": 3200 + }, + { + "epoch": 2.0558766859344892, + "grad_norm": 1.1204486857102613, + "learning_rate": 2.7265517238176847e-06, + "loss": 0.0365, + "step": 3201 + }, + { + "epoch": 2.056518946692357, + "grad_norm": 0.940624183456716, + "learning_rate": 2.7232237138325906e-06, + "loss": 0.0283, + "step": 3202 + }, + { + "epoch": 2.057161207450225, + "grad_norm": 0.8954075997551559, + "learning_rate": 2.7198969758890675e-06, + "loss": 0.0318, + "step": 3203 + }, + { + "epoch": 2.0578034682080926, + "grad_norm": 0.9521074019465717, + "learning_rate": 2.7165715118457735e-06, + "loss": 0.0379, + "step": 3204 + }, + { + "epoch": 2.0584457289659603, + "grad_norm": 0.6970749526305088, + "learning_rate": 2.7132473235606553e-06, + "loss": 0.016, + "step": 3205 + }, + { + "epoch": 2.0590879897238277, + "grad_norm": 0.5911466228026242, + "learning_rate": 2.7099244128909483e-06, + "loss": 0.0198, + "step": 3206 + }, + { + "epoch": 2.0597302504816954, + "grad_norm": 0.8874399572579849, + "learning_rate": 2.706602781693169e-06, + "loss": 0.0434, + "step": 3207 + }, + { + "epoch": 2.060372511239563, + "grad_norm": 0.8733472097441007, + "learning_rate": 2.7032824318231237e-06, + "loss": 0.0366, + "step": 3208 + }, + { + "epoch": 2.061014771997431, + "grad_norm": 0.8209025274037063, + "learning_rate": 2.6999633651359026e-06, + "loss": 0.0389, + "step": 3209 + }, + { + "epoch": 2.0616570327552988, + "grad_norm": 0.7944574753996563, + "learning_rate": 2.6966455834858746e-06, + "loss": 0.0253, + "step": 3210 + }, + { + "epoch": 2.0622992935131665, + "grad_norm": 0.6761493675004547, + "learning_rate": 2.6933290887266994e-06, + "loss": 0.0266, + "step": 3211 + }, + { + "epoch": 2.062941554271034, + "grad_norm": 0.8257638393021541, + "learning_rate": 2.690013882711309e-06, + "loss": 0.0427, + "step": 3212 + }, + { + "epoch": 2.0635838150289016, + "grad_norm": 1.1713994754930077, + "learning_rate": 2.6866999672919198e-06, + "loss": 0.0321, + "step": 3213 + }, + { + "epoch": 2.0642260757867694, + "grad_norm": 0.8550147846149614, + "learning_rate": 2.6833873443200275e-06, + "loss": 0.0198, + "step": 3214 + }, + { + "epoch": 2.064868336544637, + "grad_norm": 0.7774189063042299, + "learning_rate": 2.6800760156464e-06, + "loss": 0.0325, + "step": 3215 + }, + { + "epoch": 2.065510597302505, + "grad_norm": 0.49322642251716997, + "learning_rate": 2.676765983121089e-06, + "loss": 0.0219, + "step": 3216 + }, + { + "epoch": 2.0661528580603723, + "grad_norm": 0.5960986775668093, + "learning_rate": 2.6734572485934192e-06, + "loss": 0.0238, + "step": 3217 + }, + { + "epoch": 2.06679511881824, + "grad_norm": 0.7710943442204922, + "learning_rate": 2.6701498139119896e-06, + "loss": 0.0279, + "step": 3218 + }, + { + "epoch": 2.067437379576108, + "grad_norm": 0.6660753714668857, + "learning_rate": 2.6668436809246746e-06, + "loss": 0.0274, + "step": 3219 + }, + { + "epoch": 2.0680796403339756, + "grad_norm": 0.6947138918324126, + "learning_rate": 2.6635388514786198e-06, + "loss": 0.0423, + "step": 3220 + }, + { + "epoch": 2.0687219010918434, + "grad_norm": 0.702779011575125, + "learning_rate": 2.660235327420243e-06, + "loss": 0.0439, + "step": 3221 + }, + { + "epoch": 2.069364161849711, + "grad_norm": 0.9358540720090645, + "learning_rate": 2.656933110595233e-06, + "loss": 0.0452, + "step": 3222 + }, + { + "epoch": 2.0700064226075785, + "grad_norm": 0.7642817110502402, + "learning_rate": 2.653632202848548e-06, + "loss": 0.0315, + "step": 3223 + }, + { + "epoch": 2.0706486833654463, + "grad_norm": 0.6262336974897339, + "learning_rate": 2.6503326060244146e-06, + "loss": 0.0316, + "step": 3224 + }, + { + "epoch": 2.071290944123314, + "grad_norm": 0.7304313473519799, + "learning_rate": 2.6470343219663275e-06, + "loss": 0.0353, + "step": 3225 + }, + { + "epoch": 2.071933204881182, + "grad_norm": 0.7235574426262332, + "learning_rate": 2.6437373525170473e-06, + "loss": 0.0375, + "step": 3226 + }, + { + "epoch": 2.0725754656390496, + "grad_norm": 0.5763856696070511, + "learning_rate": 2.6404416995186005e-06, + "loss": 0.0221, + "step": 3227 + }, + { + "epoch": 2.0732177263969174, + "grad_norm": 0.8440343756666407, + "learning_rate": 2.637147364812278e-06, + "loss": 0.0436, + "step": 3228 + }, + { + "epoch": 2.0738599871547847, + "grad_norm": 0.7326919506766564, + "learning_rate": 2.6338543502386355e-06, + "loss": 0.0389, + "step": 3229 + }, + { + "epoch": 2.0745022479126525, + "grad_norm": 0.7649779795354801, + "learning_rate": 2.63056265763749e-06, + "loss": 0.0336, + "step": 3230 + }, + { + "epoch": 2.0751445086705202, + "grad_norm": 0.8381629236742201, + "learning_rate": 2.6272722888479152e-06, + "loss": 0.0413, + "step": 3231 + }, + { + "epoch": 2.075786769428388, + "grad_norm": 0.6657312096917367, + "learning_rate": 2.623983245708255e-06, + "loss": 0.0236, + "step": 3232 + }, + { + "epoch": 2.076429030186256, + "grad_norm": 0.9326807695745203, + "learning_rate": 2.6206955300561048e-06, + "loss": 0.0539, + "step": 3233 + }, + { + "epoch": 2.077071290944123, + "grad_norm": 1.0417869804498314, + "learning_rate": 2.617409143728322e-06, + "loss": 0.0458, + "step": 3234 + }, + { + "epoch": 2.077713551701991, + "grad_norm": 0.5265759405679188, + "learning_rate": 2.6141240885610206e-06, + "loss": 0.015, + "step": 3235 + }, + { + "epoch": 2.0783558124598587, + "grad_norm": 0.6435768103049353, + "learning_rate": 2.6108403663895705e-06, + "loss": 0.0222, + "step": 3236 + }, + { + "epoch": 2.0789980732177264, + "grad_norm": 0.6855087138169853, + "learning_rate": 2.607557979048595e-06, + "loss": 0.0256, + "step": 3237 + }, + { + "epoch": 2.0796403339755942, + "grad_norm": 0.7708045099716588, + "learning_rate": 2.6042769283719783e-06, + "loss": 0.0403, + "step": 3238 + }, + { + "epoch": 2.080282594733462, + "grad_norm": 0.7784413273462673, + "learning_rate": 2.6009972161928463e-06, + "loss": 0.0417, + "step": 3239 + }, + { + "epoch": 2.0809248554913293, + "grad_norm": 0.5432318893939191, + "learning_rate": 2.5977188443435874e-06, + "loss": 0.0215, + "step": 3240 + }, + { + "epoch": 2.081567116249197, + "grad_norm": 0.8894938726879936, + "learning_rate": 2.594441814655837e-06, + "loss": 0.0383, + "step": 3241 + }, + { + "epoch": 2.082209377007065, + "grad_norm": 0.8013999896910246, + "learning_rate": 2.591166128960482e-06, + "loss": 0.0255, + "step": 3242 + }, + { + "epoch": 2.0828516377649327, + "grad_norm": 0.5600036923140306, + "learning_rate": 2.5878917890876544e-06, + "loss": 0.0276, + "step": 3243 + }, + { + "epoch": 2.0834938985228004, + "grad_norm": 0.8003045265988253, + "learning_rate": 2.5846187968667423e-06, + "loss": 0.0332, + "step": 3244 + }, + { + "epoch": 2.0841361592806678, + "grad_norm": 0.37095417018587007, + "learning_rate": 2.5813471541263734e-06, + "loss": 0.0111, + "step": 3245 + }, + { + "epoch": 2.0847784200385355, + "grad_norm": 0.736040611300248, + "learning_rate": 2.5780768626944265e-06, + "loss": 0.0265, + "step": 3246 + }, + { + "epoch": 2.0854206807964033, + "grad_norm": 0.8492352307441081, + "learning_rate": 2.574807924398019e-06, + "loss": 0.0512, + "step": 3247 + }, + { + "epoch": 2.086062941554271, + "grad_norm": 0.668526488612227, + "learning_rate": 2.571540341063518e-06, + "loss": 0.0279, + "step": 3248 + }, + { + "epoch": 2.086705202312139, + "grad_norm": 0.6198412953419515, + "learning_rate": 2.5682741145165325e-06, + "loss": 0.0259, + "step": 3249 + }, + { + "epoch": 2.0873474630700066, + "grad_norm": 1.0983906073696925, + "learning_rate": 2.565009246581912e-06, + "loss": 0.0505, + "step": 3250 + }, + { + "epoch": 2.087989723827874, + "grad_norm": 1.1681187722766342, + "learning_rate": 2.5617457390837474e-06, + "loss": 0.0568, + "step": 3251 + }, + { + "epoch": 2.0886319845857417, + "grad_norm": 0.7010817642269688, + "learning_rate": 2.558483593845372e-06, + "loss": 0.0262, + "step": 3252 + }, + { + "epoch": 2.0892742453436095, + "grad_norm": 1.1885641239216413, + "learning_rate": 2.5552228126893542e-06, + "loss": 0.046, + "step": 3253 + }, + { + "epoch": 2.0899165061014773, + "grad_norm": 0.9942304550528109, + "learning_rate": 2.5519633974375014e-06, + "loss": 0.0455, + "step": 3254 + }, + { + "epoch": 2.090558766859345, + "grad_norm": 0.8811285679457127, + "learning_rate": 2.54870534991086e-06, + "loss": 0.0392, + "step": 3255 + }, + { + "epoch": 2.0912010276172124, + "grad_norm": 0.5990528349657622, + "learning_rate": 2.54544867192971e-06, + "loss": 0.0238, + "step": 3256 + }, + { + "epoch": 2.09184328837508, + "grad_norm": 0.688321119808742, + "learning_rate": 2.5421933653135665e-06, + "loss": 0.0312, + "step": 3257 + }, + { + "epoch": 2.092485549132948, + "grad_norm": 0.6274746313800298, + "learning_rate": 2.53893943188118e-06, + "loss": 0.036, + "step": 3258 + }, + { + "epoch": 2.0931278098908157, + "grad_norm": 0.7559365545255198, + "learning_rate": 2.535686873450531e-06, + "loss": 0.0199, + "step": 3259 + }, + { + "epoch": 2.0937700706486835, + "grad_norm": 0.8374927459915017, + "learning_rate": 2.532435691838835e-06, + "loss": 0.0323, + "step": 3260 + }, + { + "epoch": 2.0944123314065513, + "grad_norm": 0.935524977011823, + "learning_rate": 2.529185888862537e-06, + "loss": 0.0268, + "step": 3261 + }, + { + "epoch": 2.0950545921644186, + "grad_norm": 0.7121938611669341, + "learning_rate": 2.5259374663373093e-06, + "loss": 0.027, + "step": 3262 + }, + { + "epoch": 2.0956968529222864, + "grad_norm": 0.8631565860150349, + "learning_rate": 2.5226904260780556e-06, + "loss": 0.0527, + "step": 3263 + }, + { + "epoch": 2.096339113680154, + "grad_norm": 0.5667771439714584, + "learning_rate": 2.519444769898908e-06, + "loss": 0.0202, + "step": 3264 + }, + { + "epoch": 2.096981374438022, + "grad_norm": 0.5209136984841487, + "learning_rate": 2.5162004996132205e-06, + "loss": 0.0111, + "step": 3265 + }, + { + "epoch": 2.0976236351958897, + "grad_norm": 0.8060006447775923, + "learning_rate": 2.512957617033582e-06, + "loss": 0.0331, + "step": 3266 + }, + { + "epoch": 2.098265895953757, + "grad_norm": 0.8784275388721444, + "learning_rate": 2.5097161239717975e-06, + "loss": 0.0397, + "step": 3267 + }, + { + "epoch": 2.098908156711625, + "grad_norm": 0.6553636662832458, + "learning_rate": 2.5064760222388994e-06, + "loss": 0.0172, + "step": 3268 + }, + { + "epoch": 2.0995504174694926, + "grad_norm": 0.6662753177262106, + "learning_rate": 2.503237313645143e-06, + "loss": 0.0283, + "step": 3269 + }, + { + "epoch": 2.1001926782273603, + "grad_norm": 1.0069740316211448, + "learning_rate": 2.5000000000000015e-06, + "loss": 0.0465, + "step": 3270 + }, + { + "epoch": 2.100834938985228, + "grad_norm": 0.8229920901197224, + "learning_rate": 2.496764083112172e-06, + "loss": 0.0388, + "step": 3271 + }, + { + "epoch": 2.101477199743096, + "grad_norm": 0.8050210919554506, + "learning_rate": 2.4935295647895724e-06, + "loss": 0.0347, + "step": 3272 + }, + { + "epoch": 2.102119460500963, + "grad_norm": 0.609784796002111, + "learning_rate": 2.4902964468393365e-06, + "loss": 0.0218, + "step": 3273 + }, + { + "epoch": 2.102761721258831, + "grad_norm": 0.5571728039982891, + "learning_rate": 2.4870647310678168e-06, + "loss": 0.0117, + "step": 3274 + }, + { + "epoch": 2.1034039820166988, + "grad_norm": 0.6361101737091563, + "learning_rate": 2.483834419280583e-06, + "loss": 0.0301, + "step": 3275 + }, + { + "epoch": 2.1040462427745665, + "grad_norm": 1.1136175591519657, + "learning_rate": 2.4806055132824186e-06, + "loss": 0.0397, + "step": 3276 + }, + { + "epoch": 2.1046885035324343, + "grad_norm": 0.7954211892903895, + "learning_rate": 2.477378014877324e-06, + "loss": 0.0326, + "step": 3277 + }, + { + "epoch": 2.1053307642903016, + "grad_norm": 0.816137998831274, + "learning_rate": 2.4741519258685115e-06, + "loss": 0.0335, + "step": 3278 + }, + { + "epoch": 2.1059730250481694, + "grad_norm": 1.0159603485666862, + "learning_rate": 2.4709272480584064e-06, + "loss": 0.0469, + "step": 3279 + }, + { + "epoch": 2.106615285806037, + "grad_norm": 0.6460823339248619, + "learning_rate": 2.4677039832486455e-06, + "loss": 0.0262, + "step": 3280 + }, + { + "epoch": 2.107257546563905, + "grad_norm": 0.85428826903129, + "learning_rate": 2.464482133240077e-06, + "loss": 0.037, + "step": 3281 + }, + { + "epoch": 2.1078998073217727, + "grad_norm": 0.75419428460334, + "learning_rate": 2.461261699832757e-06, + "loss": 0.0301, + "step": 3282 + }, + { + "epoch": 2.1085420680796405, + "grad_norm": 1.1679496023690235, + "learning_rate": 2.4580426848259514e-06, + "loss": 0.066, + "step": 3283 + }, + { + "epoch": 2.109184328837508, + "grad_norm": 0.6186389277475125, + "learning_rate": 2.4548250900181336e-06, + "loss": 0.0291, + "step": 3284 + }, + { + "epoch": 2.1098265895953756, + "grad_norm": 0.844945129994765, + "learning_rate": 2.4516089172069852e-06, + "loss": 0.0223, + "step": 3285 + }, + { + "epoch": 2.1104688503532434, + "grad_norm": 0.7716692142606174, + "learning_rate": 2.4483941681893853e-06, + "loss": 0.0298, + "step": 3286 + }, + { + "epoch": 2.111111111111111, + "grad_norm": 1.0053756034306576, + "learning_rate": 2.445180844761429e-06, + "loss": 0.041, + "step": 3287 + }, + { + "epoch": 2.111753371868979, + "grad_norm": 0.9059160313048512, + "learning_rate": 2.4419689487184077e-06, + "loss": 0.0533, + "step": 3288 + }, + { + "epoch": 2.1123956326268467, + "grad_norm": 0.542266941081418, + "learning_rate": 2.4387584818548175e-06, + "loss": 0.0148, + "step": 3289 + }, + { + "epoch": 2.113037893384714, + "grad_norm": 0.5735368357886154, + "learning_rate": 2.4355494459643562e-06, + "loss": 0.0183, + "step": 3290 + }, + { + "epoch": 2.113680154142582, + "grad_norm": 0.6218576484705493, + "learning_rate": 2.4323418428399204e-06, + "loss": 0.0131, + "step": 3291 + }, + { + "epoch": 2.1143224149004496, + "grad_norm": 0.6893925705474475, + "learning_rate": 2.429135674273608e-06, + "loss": 0.0276, + "step": 3292 + }, + { + "epoch": 2.1149646756583174, + "grad_norm": 0.6987123738913608, + "learning_rate": 2.4259309420567158e-06, + "loss": 0.0286, + "step": 3293 + }, + { + "epoch": 2.115606936416185, + "grad_norm": 0.5789053802217684, + "learning_rate": 2.422727647979734e-06, + "loss": 0.0237, + "step": 3294 + }, + { + "epoch": 2.1162491971740525, + "grad_norm": 0.9553667626117904, + "learning_rate": 2.4195257938323544e-06, + "loss": 0.0516, + "step": 3295 + }, + { + "epoch": 2.1168914579319202, + "grad_norm": 0.7195061482787911, + "learning_rate": 2.4163253814034622e-06, + "loss": 0.0344, + "step": 3296 + }, + { + "epoch": 2.117533718689788, + "grad_norm": 1.0348289018794208, + "learning_rate": 2.4131264124811366e-06, + "loss": 0.0394, + "step": 3297 + }, + { + "epoch": 2.118175979447656, + "grad_norm": 0.8201871072513327, + "learning_rate": 2.409928888852652e-06, + "loss": 0.0237, + "step": 3298 + }, + { + "epoch": 2.1188182402055236, + "grad_norm": 0.6626903693249929, + "learning_rate": 2.406732812304471e-06, + "loss": 0.0277, + "step": 3299 + }, + { + "epoch": 2.1194605009633913, + "grad_norm": 1.0233486125075397, + "learning_rate": 2.4035381846222554e-06, + "loss": 0.0397, + "step": 3300 + }, + { + "epoch": 2.1201027617212587, + "grad_norm": 0.8496028209700652, + "learning_rate": 2.4003450075908534e-06, + "loss": 0.0274, + "step": 3301 + }, + { + "epoch": 2.1207450224791264, + "grad_norm": 1.1166765390867956, + "learning_rate": 2.397153282994298e-06, + "loss": 0.0547, + "step": 3302 + }, + { + "epoch": 2.121387283236994, + "grad_norm": 1.0735989600776017, + "learning_rate": 2.393963012615817e-06, + "loss": 0.0422, + "step": 3303 + }, + { + "epoch": 2.122029543994862, + "grad_norm": 0.6392117266897839, + "learning_rate": 2.3907741982378234e-06, + "loss": 0.0231, + "step": 3304 + }, + { + "epoch": 2.1226718047527298, + "grad_norm": 1.0744173126462635, + "learning_rate": 2.3875868416419187e-06, + "loss": 0.0566, + "step": 3305 + }, + { + "epoch": 2.123314065510597, + "grad_norm": 0.6434439708009836, + "learning_rate": 2.3844009446088863e-06, + "loss": 0.0254, + "step": 3306 + }, + { + "epoch": 2.123956326268465, + "grad_norm": 0.7222502609901142, + "learning_rate": 2.3812165089186974e-06, + "loss": 0.0322, + "step": 3307 + }, + { + "epoch": 2.1245985870263326, + "grad_norm": 0.6669376506931214, + "learning_rate": 2.3780335363505052e-06, + "loss": 0.0296, + "step": 3308 + }, + { + "epoch": 2.1252408477842004, + "grad_norm": 1.0489790991262267, + "learning_rate": 2.374852028682646e-06, + "loss": 0.0416, + "step": 3309 + }, + { + "epoch": 2.125883108542068, + "grad_norm": 0.8820871856997933, + "learning_rate": 2.3716719876926377e-06, + "loss": 0.048, + "step": 3310 + }, + { + "epoch": 2.126525369299936, + "grad_norm": 0.9675081574225268, + "learning_rate": 2.3684934151571778e-06, + "loss": 0.0333, + "step": 3311 + }, + { + "epoch": 2.1271676300578033, + "grad_norm": 1.020469665855334, + "learning_rate": 2.365316312852144e-06, + "loss": 0.0521, + "step": 3312 + }, + { + "epoch": 2.127809890815671, + "grad_norm": 0.9351519693262244, + "learning_rate": 2.3621406825525935e-06, + "loss": 0.0476, + "step": 3313 + }, + { + "epoch": 2.128452151573539, + "grad_norm": 0.5051716625705412, + "learning_rate": 2.3589665260327595e-06, + "loss": 0.0124, + "step": 3314 + }, + { + "epoch": 2.1290944123314066, + "grad_norm": 0.6624969080967191, + "learning_rate": 2.355793845066053e-06, + "loss": 0.0236, + "step": 3315 + }, + { + "epoch": 2.1297366730892744, + "grad_norm": 0.664570979840964, + "learning_rate": 2.3526226414250615e-06, + "loss": 0.0338, + "step": 3316 + }, + { + "epoch": 2.130378933847142, + "grad_norm": 0.7668762713003453, + "learning_rate": 2.3494529168815428e-06, + "loss": 0.0327, + "step": 3317 + }, + { + "epoch": 2.1310211946050095, + "grad_norm": 0.703564852588013, + "learning_rate": 2.346284673206432e-06, + "loss": 0.0305, + "step": 3318 + }, + { + "epoch": 2.1316634553628773, + "grad_norm": 0.937132664657854, + "learning_rate": 2.3431179121698375e-06, + "loss": 0.0467, + "step": 3319 + }, + { + "epoch": 2.132305716120745, + "grad_norm": 1.04219912172079, + "learning_rate": 2.339952635541035e-06, + "loss": 0.0427, + "step": 3320 + }, + { + "epoch": 2.132947976878613, + "grad_norm": 0.748415328836577, + "learning_rate": 2.336788845088478e-06, + "loss": 0.0163, + "step": 3321 + }, + { + "epoch": 2.1335902376364806, + "grad_norm": 0.8773673532658661, + "learning_rate": 2.333626542579784e-06, + "loss": 0.0554, + "step": 3322 + }, + { + "epoch": 2.134232498394348, + "grad_norm": 1.0115624589900905, + "learning_rate": 2.3304657297817395e-06, + "loss": 0.0451, + "step": 3323 + }, + { + "epoch": 2.1348747591522157, + "grad_norm": 0.7664023618911923, + "learning_rate": 2.3273064084603024e-06, + "loss": 0.0253, + "step": 3324 + }, + { + "epoch": 2.1355170199100835, + "grad_norm": 0.6787077487019514, + "learning_rate": 2.3241485803805913e-06, + "loss": 0.0311, + "step": 3325 + }, + { + "epoch": 2.1361592806679512, + "grad_norm": 0.7874633557455805, + "learning_rate": 2.320992247306896e-06, + "loss": 0.0313, + "step": 3326 + }, + { + "epoch": 2.136801541425819, + "grad_norm": 0.8321066714695884, + "learning_rate": 2.317837411002668e-06, + "loss": 0.029, + "step": 3327 + }, + { + "epoch": 2.137443802183687, + "grad_norm": 0.8349058359013206, + "learning_rate": 2.3146840732305244e-06, + "loss": 0.0412, + "step": 3328 + }, + { + "epoch": 2.138086062941554, + "grad_norm": 0.8748540727405004, + "learning_rate": 2.3115322357522436e-06, + "loss": 0.0284, + "step": 3329 + }, + { + "epoch": 2.138728323699422, + "grad_norm": 0.9164733435337286, + "learning_rate": 2.308381900328767e-06, + "loss": 0.0213, + "step": 3330 + }, + { + "epoch": 2.1393705844572897, + "grad_norm": 0.8894024828906777, + "learning_rate": 2.305233068720196e-06, + "loss": 0.0246, + "step": 3331 + }, + { + "epoch": 2.1400128452151574, + "grad_norm": 0.7352891933169691, + "learning_rate": 2.3020857426857927e-06, + "loss": 0.0222, + "step": 3332 + }, + { + "epoch": 2.140655105973025, + "grad_norm": 0.9358371668114502, + "learning_rate": 2.298939923983977e-06, + "loss": 0.0465, + "step": 3333 + }, + { + "epoch": 2.1412973667308925, + "grad_norm": 0.7425570376278798, + "learning_rate": 2.2957956143723257e-06, + "loss": 0.0331, + "step": 3334 + }, + { + "epoch": 2.1419396274887603, + "grad_norm": 0.8204296469223049, + "learning_rate": 2.2926528156075756e-06, + "loss": 0.0482, + "step": 3335 + }, + { + "epoch": 2.142581888246628, + "grad_norm": 0.40847659504742934, + "learning_rate": 2.289511529445616e-06, + "loss": 0.012, + "step": 3336 + }, + { + "epoch": 2.143224149004496, + "grad_norm": 0.950999165413407, + "learning_rate": 2.286371757641494e-06, + "loss": 0.0587, + "step": 3337 + }, + { + "epoch": 2.1438664097623636, + "grad_norm": 0.7850281011866069, + "learning_rate": 2.283233501949409e-06, + "loss": 0.0329, + "step": 3338 + }, + { + "epoch": 2.1445086705202314, + "grad_norm": 1.5078958683168942, + "learning_rate": 2.2800967641227127e-06, + "loss": 0.0383, + "step": 3339 + }, + { + "epoch": 2.1451509312780987, + "grad_norm": 1.0990699434494673, + "learning_rate": 2.2769615459139127e-06, + "loss": 0.0624, + "step": 3340 + }, + { + "epoch": 2.1457931920359665, + "grad_norm": 0.8949089871008897, + "learning_rate": 2.2738278490746607e-06, + "loss": 0.048, + "step": 3341 + }, + { + "epoch": 2.1464354527938343, + "grad_norm": 0.6216171433741213, + "learning_rate": 2.2706956753557617e-06, + "loss": 0.0146, + "step": 3342 + }, + { + "epoch": 2.147077713551702, + "grad_norm": 0.841728456896803, + "learning_rate": 2.267565026507174e-06, + "loss": 0.0282, + "step": 3343 + }, + { + "epoch": 2.14771997430957, + "grad_norm": 0.9508789028792253, + "learning_rate": 2.2644359042779997e-06, + "loss": 0.0457, + "step": 3344 + }, + { + "epoch": 2.148362235067437, + "grad_norm": 0.7851981025050849, + "learning_rate": 2.2613083104164867e-06, + "loss": 0.0332, + "step": 3345 + }, + { + "epoch": 2.149004495825305, + "grad_norm": 1.0346223627774966, + "learning_rate": 2.258182246670032e-06, + "loss": 0.0505, + "step": 3346 + }, + { + "epoch": 2.1496467565831727, + "grad_norm": 1.2988894842576666, + "learning_rate": 2.2550577147851766e-06, + "loss": 0.0386, + "step": 3347 + }, + { + "epoch": 2.1502890173410405, + "grad_norm": 0.5277806688281964, + "learning_rate": 2.2519347165076067e-06, + "loss": 0.0221, + "step": 3348 + }, + { + "epoch": 2.1509312780989083, + "grad_norm": 0.6678932940159716, + "learning_rate": 2.248813253582147e-06, + "loss": 0.031, + "step": 3349 + }, + { + "epoch": 2.151573538856776, + "grad_norm": 1.0205846568622083, + "learning_rate": 2.245693327752771e-06, + "loss": 0.0355, + "step": 3350 + }, + { + "epoch": 2.1522157996146434, + "grad_norm": 0.7274140843597501, + "learning_rate": 2.242574940762589e-06, + "loss": 0.0235, + "step": 3351 + }, + { + "epoch": 2.152858060372511, + "grad_norm": 0.7193504524393092, + "learning_rate": 2.239458094353854e-06, + "loss": 0.0244, + "step": 3352 + }, + { + "epoch": 2.153500321130379, + "grad_norm": 0.8760847199650159, + "learning_rate": 2.2363427902679564e-06, + "loss": 0.0315, + "step": 3353 + }, + { + "epoch": 2.1541425818882467, + "grad_norm": 0.6329459387652011, + "learning_rate": 2.2332290302454237e-06, + "loss": 0.0232, + "step": 3354 + }, + { + "epoch": 2.1547848426461145, + "grad_norm": 0.8763386119394713, + "learning_rate": 2.2301168160259273e-06, + "loss": 0.0382, + "step": 3355 + }, + { + "epoch": 2.155427103403982, + "grad_norm": 0.9119085065194911, + "learning_rate": 2.2270061493482697e-06, + "loss": 0.0359, + "step": 3356 + }, + { + "epoch": 2.1560693641618496, + "grad_norm": 0.849090831924262, + "learning_rate": 2.223897031950386e-06, + "loss": 0.0336, + "step": 3357 + }, + { + "epoch": 2.1567116249197174, + "grad_norm": 0.8723687465619504, + "learning_rate": 2.220789465569351e-06, + "loss": 0.045, + "step": 3358 + }, + { + "epoch": 2.157353885677585, + "grad_norm": 0.8824083080130471, + "learning_rate": 2.2176834519413714e-06, + "loss": 0.0476, + "step": 3359 + }, + { + "epoch": 2.157996146435453, + "grad_norm": 0.854896287257943, + "learning_rate": 2.2145789928017865e-06, + "loss": 0.0457, + "step": 3360 + }, + { + "epoch": 2.1586384071933207, + "grad_norm": 0.8166108077828871, + "learning_rate": 2.211476089885065e-06, + "loss": 0.0234, + "step": 3361 + }, + { + "epoch": 2.159280667951188, + "grad_norm": 0.6288078432756016, + "learning_rate": 2.2083747449248094e-06, + "loss": 0.0226, + "step": 3362 + }, + { + "epoch": 2.159922928709056, + "grad_norm": 1.0559793374116295, + "learning_rate": 2.205274959653749e-06, + "loss": 0.0489, + "step": 3363 + }, + { + "epoch": 2.1605651894669236, + "grad_norm": 0.7371614418524727, + "learning_rate": 2.2021767358037448e-06, + "loss": 0.0299, + "step": 3364 + }, + { + "epoch": 2.1612074502247913, + "grad_norm": 0.49096389724503736, + "learning_rate": 2.199080075105782e-06, + "loss": 0.0142, + "step": 3365 + }, + { + "epoch": 2.161849710982659, + "grad_norm": 1.616868763030457, + "learning_rate": 2.195984979289974e-06, + "loss": 0.0599, + "step": 3366 + }, + { + "epoch": 2.1624919717405264, + "grad_norm": 0.9556446095164549, + "learning_rate": 2.1928914500855604e-06, + "loss": 0.0301, + "step": 3367 + }, + { + "epoch": 2.163134232498394, + "grad_norm": 0.6457057628666302, + "learning_rate": 2.1897994892209044e-06, + "loss": 0.0266, + "step": 3368 + }, + { + "epoch": 2.163776493256262, + "grad_norm": 0.7896007597861923, + "learning_rate": 2.1867090984234944e-06, + "loss": 0.045, + "step": 3369 + }, + { + "epoch": 2.1644187540141298, + "grad_norm": 0.8882911011929814, + "learning_rate": 2.183620279419941e-06, + "loss": 0.0384, + "step": 3370 + }, + { + "epoch": 2.1650610147719975, + "grad_norm": 0.9311844003561217, + "learning_rate": 2.1805330339359765e-06, + "loss": 0.0402, + "step": 3371 + }, + { + "epoch": 2.1657032755298653, + "grad_norm": 0.41811158281107974, + "learning_rate": 2.177447363696452e-06, + "loss": 0.0131, + "step": 3372 + }, + { + "epoch": 2.1663455362877326, + "grad_norm": 0.8073366474192268, + "learning_rate": 2.174363270425342e-06, + "loss": 0.0456, + "step": 3373 + }, + { + "epoch": 2.1669877970456004, + "grad_norm": 1.0628927046551817, + "learning_rate": 2.171280755845738e-06, + "loss": 0.0537, + "step": 3374 + }, + { + "epoch": 2.167630057803468, + "grad_norm": 0.682854995547217, + "learning_rate": 2.1681998216798476e-06, + "loss": 0.0284, + "step": 3375 + }, + { + "epoch": 2.168272318561336, + "grad_norm": 0.8565445233891327, + "learning_rate": 2.165120469649003e-06, + "loss": 0.0353, + "step": 3376 + }, + { + "epoch": 2.1689145793192037, + "grad_norm": 0.6165111872568804, + "learning_rate": 2.1620427014736443e-06, + "loss": 0.0283, + "step": 3377 + }, + { + "epoch": 2.169556840077071, + "grad_norm": 0.9507460244293467, + "learning_rate": 2.1589665188733294e-06, + "loss": 0.0413, + "step": 3378 + }, + { + "epoch": 2.170199100834939, + "grad_norm": 0.7390473076350449, + "learning_rate": 2.1558919235667332e-06, + "loss": 0.0159, + "step": 3379 + }, + { + "epoch": 2.1708413615928066, + "grad_norm": 0.6490392382739006, + "learning_rate": 2.152818917271636e-06, + "loss": 0.0103, + "step": 3380 + }, + { + "epoch": 2.1714836223506744, + "grad_norm": 1.1115823018051438, + "learning_rate": 2.149747501704939e-06, + "loss": 0.0485, + "step": 3381 + }, + { + "epoch": 2.172125883108542, + "grad_norm": 0.6481088024858652, + "learning_rate": 2.1466776785826497e-06, + "loss": 0.025, + "step": 3382 + }, + { + "epoch": 2.17276814386641, + "grad_norm": 0.824373677620185, + "learning_rate": 2.143609449619888e-06, + "loss": 0.0405, + "step": 3383 + }, + { + "epoch": 2.1734104046242773, + "grad_norm": 0.7538701192157629, + "learning_rate": 2.140542816530882e-06, + "loss": 0.0356, + "step": 3384 + }, + { + "epoch": 2.174052665382145, + "grad_norm": 0.7966909184879134, + "learning_rate": 2.1374777810289686e-06, + "loss": 0.0312, + "step": 3385 + }, + { + "epoch": 2.174694926140013, + "grad_norm": 1.0878953539930893, + "learning_rate": 2.1344143448265926e-06, + "loss": 0.028, + "step": 3386 + }, + { + "epoch": 2.1753371868978806, + "grad_norm": 1.0313884834790805, + "learning_rate": 2.1313525096353044e-06, + "loss": 0.0525, + "step": 3387 + }, + { + "epoch": 2.1759794476557484, + "grad_norm": 0.6928568314344778, + "learning_rate": 2.12829227716576e-06, + "loss": 0.0267, + "step": 3388 + }, + { + "epoch": 2.176621708413616, + "grad_norm": 0.9651054335086483, + "learning_rate": 2.1252336491277207e-06, + "loss": 0.0431, + "step": 3389 + }, + { + "epoch": 2.1772639691714835, + "grad_norm": 0.9242254766646552, + "learning_rate": 2.1221766272300506e-06, + "loss": 0.039, + "step": 3390 + }, + { + "epoch": 2.1779062299293512, + "grad_norm": 0.6378023555357174, + "learning_rate": 2.119121213180717e-06, + "loss": 0.0315, + "step": 3391 + }, + { + "epoch": 2.178548490687219, + "grad_norm": 0.6492331925132532, + "learning_rate": 2.116067408686788e-06, + "loss": 0.0211, + "step": 3392 + }, + { + "epoch": 2.179190751445087, + "grad_norm": 0.5382730150697728, + "learning_rate": 2.1130152154544346e-06, + "loss": 0.0188, + "step": 3393 + }, + { + "epoch": 2.1798330122029546, + "grad_norm": 0.8653634586614749, + "learning_rate": 2.1099646351889246e-06, + "loss": 0.043, + "step": 3394 + }, + { + "epoch": 2.180475272960822, + "grad_norm": 0.5244173896709422, + "learning_rate": 2.106915669594629e-06, + "loss": 0.0173, + "step": 3395 + }, + { + "epoch": 2.1811175337186897, + "grad_norm": 0.7019932937387369, + "learning_rate": 2.103868320375009e-06, + "loss": 0.0408, + "step": 3396 + }, + { + "epoch": 2.1817597944765574, + "grad_norm": 0.8888885083985522, + "learning_rate": 2.100822589232629e-06, + "loss": 0.0501, + "step": 3397 + }, + { + "epoch": 2.182402055234425, + "grad_norm": 0.8326509608129963, + "learning_rate": 2.09777847786915e-06, + "loss": 0.0516, + "step": 3398 + }, + { + "epoch": 2.183044315992293, + "grad_norm": 0.7013872837249128, + "learning_rate": 2.0947359879853253e-06, + "loss": 0.034, + "step": 3399 + }, + { + "epoch": 2.1836865767501608, + "grad_norm": 1.1960039768281583, + "learning_rate": 2.091695121281002e-06, + "loss": 0.0528, + "step": 3400 + }, + { + "epoch": 2.184328837508028, + "grad_norm": 0.7745009345251076, + "learning_rate": 2.088655879455122e-06, + "loss": 0.0458, + "step": 3401 + }, + { + "epoch": 2.184971098265896, + "grad_norm": 1.1930771833074858, + "learning_rate": 2.0856182642057182e-06, + "loss": 0.0463, + "step": 3402 + }, + { + "epoch": 2.1856133590237636, + "grad_norm": 0.6775949064316608, + "learning_rate": 2.082582277229917e-06, + "loss": 0.0248, + "step": 3403 + }, + { + "epoch": 2.1862556197816314, + "grad_norm": 0.8494854203302948, + "learning_rate": 2.079547920223929e-06, + "loss": 0.0303, + "step": 3404 + }, + { + "epoch": 2.186897880539499, + "grad_norm": 0.8602355080127675, + "learning_rate": 2.0765151948830613e-06, + "loss": 0.0333, + "step": 3405 + }, + { + "epoch": 2.187540141297367, + "grad_norm": 1.1917802567165223, + "learning_rate": 2.0734841029017062e-06, + "loss": 0.0657, + "step": 3406 + }, + { + "epoch": 2.1881824020552343, + "grad_norm": 0.9629222030151535, + "learning_rate": 2.0704546459733425e-06, + "loss": 0.0408, + "step": 3407 + }, + { + "epoch": 2.188824662813102, + "grad_norm": 0.5900200256532747, + "learning_rate": 2.0674268257905377e-06, + "loss": 0.0244, + "step": 3408 + }, + { + "epoch": 2.18946692357097, + "grad_norm": 0.7137322318314039, + "learning_rate": 2.064400644044942e-06, + "loss": 0.031, + "step": 3409 + }, + { + "epoch": 2.1901091843288376, + "grad_norm": 0.6128650268121317, + "learning_rate": 2.061376102427295e-06, + "loss": 0.0213, + "step": 3410 + }, + { + "epoch": 2.1907514450867054, + "grad_norm": 1.029100053668458, + "learning_rate": 2.058353202627417e-06, + "loss": 0.0503, + "step": 3411 + }, + { + "epoch": 2.1913937058445727, + "grad_norm": 0.7562791820181293, + "learning_rate": 2.055331946334209e-06, + "loss": 0.0241, + "step": 3412 + }, + { + "epoch": 2.1920359666024405, + "grad_norm": 0.8985785627394717, + "learning_rate": 2.0523123352356545e-06, + "loss": 0.0353, + "step": 3413 + }, + { + "epoch": 2.1926782273603083, + "grad_norm": 0.3697827602862209, + "learning_rate": 2.049294371018821e-06, + "loss": 0.0122, + "step": 3414 + }, + { + "epoch": 2.193320488118176, + "grad_norm": 1.0859446554888192, + "learning_rate": 2.0462780553698547e-06, + "loss": 0.0671, + "step": 3415 + }, + { + "epoch": 2.193962748876044, + "grad_norm": 0.5555944880748478, + "learning_rate": 2.043263389973978e-06, + "loss": 0.0206, + "step": 3416 + }, + { + "epoch": 2.1946050096339116, + "grad_norm": 0.49514909560028686, + "learning_rate": 2.0402503765154947e-06, + "loss": 0.021, + "step": 3417 + }, + { + "epoch": 2.195247270391779, + "grad_norm": 1.1088662342376352, + "learning_rate": 2.0372390166777833e-06, + "loss": 0.0558, + "step": 3418 + }, + { + "epoch": 2.1958895311496467, + "grad_norm": 0.6564438198926077, + "learning_rate": 2.0342293121432997e-06, + "loss": 0.0352, + "step": 3419 + }, + { + "epoch": 2.1965317919075145, + "grad_norm": 1.0782726226846262, + "learning_rate": 2.0312212645935755e-06, + "loss": 0.0486, + "step": 3420 + }, + { + "epoch": 2.1971740526653822, + "grad_norm": 0.6761269324017947, + "learning_rate": 2.028214875709214e-06, + "loss": 0.0364, + "step": 3421 + }, + { + "epoch": 2.19781631342325, + "grad_norm": 0.5108917167797521, + "learning_rate": 2.0252101471698947e-06, + "loss": 0.022, + "step": 3422 + }, + { + "epoch": 2.1984585741811173, + "grad_norm": 0.46313979161928354, + "learning_rate": 2.0222070806543683e-06, + "loss": 0.0162, + "step": 3423 + }, + { + "epoch": 2.199100834938985, + "grad_norm": 0.8550768349633692, + "learning_rate": 2.0192056778404556e-06, + "loss": 0.0378, + "step": 3424 + }, + { + "epoch": 2.199743095696853, + "grad_norm": 0.7167330651773485, + "learning_rate": 2.0162059404050505e-06, + "loss": 0.0243, + "step": 3425 + }, + { + "epoch": 2.2003853564547207, + "grad_norm": 0.8744408240507189, + "learning_rate": 2.013207870024116e-06, + "loss": 0.0351, + "step": 3426 + }, + { + "epoch": 2.2010276172125884, + "grad_norm": 1.100445379007514, + "learning_rate": 2.0102114683726796e-06, + "loss": 0.0545, + "step": 3427 + }, + { + "epoch": 2.201669877970456, + "grad_norm": 0.6340604061660352, + "learning_rate": 2.0072167371248413e-06, + "loss": 0.0221, + "step": 3428 + }, + { + "epoch": 2.2023121387283235, + "grad_norm": 0.9869561456030391, + "learning_rate": 2.0042236779537668e-06, + "loss": 0.0595, + "step": 3429 + }, + { + "epoch": 2.2029543994861913, + "grad_norm": 0.9522928798290743, + "learning_rate": 2.001232292531686e-06, + "loss": 0.0496, + "step": 3430 + }, + { + "epoch": 2.203596660244059, + "grad_norm": 1.0279244104462357, + "learning_rate": 1.9982425825298935e-06, + "loss": 0.0588, + "step": 3431 + }, + { + "epoch": 2.204238921001927, + "grad_norm": 0.7808942937077825, + "learning_rate": 1.9952545496187524e-06, + "loss": 0.0325, + "step": 3432 + }, + { + "epoch": 2.2048811817597946, + "grad_norm": 0.8664086376240301, + "learning_rate": 1.9922681954676836e-06, + "loss": 0.0325, + "step": 3433 + }, + { + "epoch": 2.205523442517662, + "grad_norm": 0.6640118095418897, + "learning_rate": 1.9892835217451743e-06, + "loss": 0.0316, + "step": 3434 + }, + { + "epoch": 2.2061657032755297, + "grad_norm": 0.6810982706586521, + "learning_rate": 1.986300530118766e-06, + "loss": 0.0254, + "step": 3435 + }, + { + "epoch": 2.2068079640333975, + "grad_norm": 0.8514544881996448, + "learning_rate": 1.9833192222550672e-06, + "loss": 0.0488, + "step": 3436 + }, + { + "epoch": 2.2074502247912653, + "grad_norm": 0.6283424935905798, + "learning_rate": 1.9803395998197433e-06, + "loss": 0.0286, + "step": 3437 + }, + { + "epoch": 2.208092485549133, + "grad_norm": 0.5036607812395709, + "learning_rate": 1.977361664477518e-06, + "loss": 0.0108, + "step": 3438 + }, + { + "epoch": 2.208734746307001, + "grad_norm": 0.6243254774708696, + "learning_rate": 1.9743854178921722e-06, + "loss": 0.017, + "step": 3439 + }, + { + "epoch": 2.209377007064868, + "grad_norm": 0.9636515678433271, + "learning_rate": 1.9714108617265436e-06, + "loss": 0.0488, + "step": 3440 + }, + { + "epoch": 2.210019267822736, + "grad_norm": 0.9524286565620961, + "learning_rate": 1.968437997642526e-06, + "loss": 0.0485, + "step": 3441 + }, + { + "epoch": 2.2106615285806037, + "grad_norm": 0.9704808627935139, + "learning_rate": 1.965466827301066e-06, + "loss": 0.0311, + "step": 3442 + }, + { + "epoch": 2.2113037893384715, + "grad_norm": 0.6630026810332749, + "learning_rate": 1.962497352362167e-06, + "loss": 0.0296, + "step": 3443 + }, + { + "epoch": 2.2119460500963393, + "grad_norm": 0.8677760822964737, + "learning_rate": 1.9595295744848823e-06, + "loss": 0.0461, + "step": 3444 + }, + { + "epoch": 2.2125883108542066, + "grad_norm": 0.8749082133887317, + "learning_rate": 1.9565634953273175e-06, + "loss": 0.0346, + "step": 3445 + }, + { + "epoch": 2.2132305716120744, + "grad_norm": 0.7875202405420995, + "learning_rate": 1.953599116546631e-06, + "loss": 0.0298, + "step": 3446 + }, + { + "epoch": 2.213872832369942, + "grad_norm": 0.9122886665032248, + "learning_rate": 1.950636439799029e-06, + "loss": 0.0529, + "step": 3447 + }, + { + "epoch": 2.21451509312781, + "grad_norm": 0.9542494404363567, + "learning_rate": 1.947675466739768e-06, + "loss": 0.0449, + "step": 3448 + }, + { + "epoch": 2.2151573538856777, + "grad_norm": 0.7630008877593005, + "learning_rate": 1.944716199023152e-06, + "loss": 0.0242, + "step": 3449 + }, + { + "epoch": 2.2157996146435455, + "grad_norm": 0.6491776966584802, + "learning_rate": 1.941758638302535e-06, + "loss": 0.0183, + "step": 3450 + }, + { + "epoch": 2.216441875401413, + "grad_norm": 0.7246476292528455, + "learning_rate": 1.9388027862303105e-06, + "loss": 0.0389, + "step": 3451 + }, + { + "epoch": 2.2170841361592806, + "grad_norm": 0.7739165556751358, + "learning_rate": 1.9358486444579217e-06, + "loss": 0.0351, + "step": 3452 + }, + { + "epoch": 2.2177263969171483, + "grad_norm": 0.8127821181709407, + "learning_rate": 1.93289621463586e-06, + "loss": 0.0298, + "step": 3453 + }, + { + "epoch": 2.218368657675016, + "grad_norm": 0.901147645058521, + "learning_rate": 1.9299454984136558e-06, + "loss": 0.0545, + "step": 3454 + }, + { + "epoch": 2.219010918432884, + "grad_norm": 0.6795236931646185, + "learning_rate": 1.9269964974398814e-06, + "loss": 0.0421, + "step": 3455 + }, + { + "epoch": 2.2196531791907512, + "grad_norm": 0.5242071231492563, + "learning_rate": 1.924049213362153e-06, + "loss": 0.0232, + "step": 3456 + }, + { + "epoch": 2.220295439948619, + "grad_norm": 0.9244514605573413, + "learning_rate": 1.9211036478271263e-06, + "loss": 0.0483, + "step": 3457 + }, + { + "epoch": 2.2209377007064868, + "grad_norm": 0.8858477730200254, + "learning_rate": 1.9181598024805e-06, + "loss": 0.0364, + "step": 3458 + }, + { + "epoch": 2.2215799614643545, + "grad_norm": 0.9378331066219359, + "learning_rate": 1.9152176789670044e-06, + "loss": 0.0248, + "step": 3459 + }, + { + "epoch": 2.2222222222222223, + "grad_norm": 1.035007948310722, + "learning_rate": 1.912277278930416e-06, + "loss": 0.065, + "step": 3460 + }, + { + "epoch": 2.22286448298009, + "grad_norm": 0.5382458668550198, + "learning_rate": 1.9093386040135436e-06, + "loss": 0.023, + "step": 3461 + }, + { + "epoch": 2.2235067437379574, + "grad_norm": 0.9852990923227671, + "learning_rate": 1.9064016558582333e-06, + "loss": 0.049, + "step": 3462 + }, + { + "epoch": 2.224149004495825, + "grad_norm": 0.5740061066859634, + "learning_rate": 1.9034664361053674e-06, + "loss": 0.0213, + "step": 3463 + }, + { + "epoch": 2.224791265253693, + "grad_norm": 0.47160467594703237, + "learning_rate": 1.9005329463948596e-06, + "loss": 0.0229, + "step": 3464 + }, + { + "epoch": 2.2254335260115607, + "grad_norm": 0.9387234816626314, + "learning_rate": 1.8976011883656632e-06, + "loss": 0.0412, + "step": 3465 + }, + { + "epoch": 2.2260757867694285, + "grad_norm": 0.7766851790028053, + "learning_rate": 1.894671163655759e-06, + "loss": 0.028, + "step": 3466 + }, + { + "epoch": 2.226718047527296, + "grad_norm": 0.7010866007289762, + "learning_rate": 1.8917428739021577e-06, + "loss": 0.0253, + "step": 3467 + }, + { + "epoch": 2.2273603082851636, + "grad_norm": 0.6420945469067814, + "learning_rate": 1.888816320740905e-06, + "loss": 0.0171, + "step": 3468 + }, + { + "epoch": 2.2280025690430314, + "grad_norm": 0.7080408357211803, + "learning_rate": 1.8858915058070742e-06, + "loss": 0.0267, + "step": 3469 + }, + { + "epoch": 2.228644829800899, + "grad_norm": 0.985447759394598, + "learning_rate": 1.8829684307347685e-06, + "loss": 0.0444, + "step": 3470 + }, + { + "epoch": 2.229287090558767, + "grad_norm": 0.6632317773268016, + "learning_rate": 1.8800470971571183e-06, + "loss": 0.0326, + "step": 3471 + }, + { + "epoch": 2.2299293513166347, + "grad_norm": 0.7665075878426243, + "learning_rate": 1.8771275067062817e-06, + "loss": 0.0232, + "step": 3472 + }, + { + "epoch": 2.230571612074502, + "grad_norm": 0.3604364283437707, + "learning_rate": 1.8742096610134413e-06, + "loss": 0.0108, + "step": 3473 + }, + { + "epoch": 2.23121387283237, + "grad_norm": 0.8235088792409154, + "learning_rate": 1.8712935617088067e-06, + "loss": 0.0351, + "step": 3474 + }, + { + "epoch": 2.2318561335902376, + "grad_norm": 0.7137266005517531, + "learning_rate": 1.8683792104216114e-06, + "loss": 0.0401, + "step": 3475 + }, + { + "epoch": 2.2324983943481054, + "grad_norm": 0.6193279021076784, + "learning_rate": 1.8654666087801116e-06, + "loss": 0.0218, + "step": 3476 + }, + { + "epoch": 2.233140655105973, + "grad_norm": 0.5217102233723345, + "learning_rate": 1.8625557584115866e-06, + "loss": 0.0173, + "step": 3477 + }, + { + "epoch": 2.233782915863841, + "grad_norm": 0.7806000419084921, + "learning_rate": 1.8596466609423364e-06, + "loss": 0.0327, + "step": 3478 + }, + { + "epoch": 2.2344251766217083, + "grad_norm": 0.8653249651917589, + "learning_rate": 1.8567393179976834e-06, + "loss": 0.0416, + "step": 3479 + }, + { + "epoch": 2.235067437379576, + "grad_norm": 0.5861520490216368, + "learning_rate": 1.8538337312019672e-06, + "loss": 0.024, + "step": 3480 + }, + { + "epoch": 2.235709698137444, + "grad_norm": 0.8289461329713538, + "learning_rate": 1.8509299021785505e-06, + "loss": 0.0369, + "step": 3481 + }, + { + "epoch": 2.2363519588953116, + "grad_norm": 0.5189075839856792, + "learning_rate": 1.8480278325498068e-06, + "loss": 0.0086, + "step": 3482 + }, + { + "epoch": 2.2369942196531793, + "grad_norm": 0.9539863882351455, + "learning_rate": 1.8451275239371337e-06, + "loss": 0.0235, + "step": 3483 + }, + { + "epoch": 2.2376364804110467, + "grad_norm": 0.639719164054563, + "learning_rate": 1.842228977960941e-06, + "loss": 0.0215, + "step": 3484 + }, + { + "epoch": 2.2382787411689145, + "grad_norm": 1.0185775027620005, + "learning_rate": 1.8393321962406552e-06, + "loss": 0.0636, + "step": 3485 + }, + { + "epoch": 2.2389210019267822, + "grad_norm": 0.8559533109242572, + "learning_rate": 1.8364371803947151e-06, + "loss": 0.0386, + "step": 3486 + }, + { + "epoch": 2.23956326268465, + "grad_norm": 0.5355255565991427, + "learning_rate": 1.833543932040578e-06, + "loss": 0.0213, + "step": 3487 + }, + { + "epoch": 2.2402055234425178, + "grad_norm": 0.7093676657854533, + "learning_rate": 1.8306524527947083e-06, + "loss": 0.0225, + "step": 3488 + }, + { + "epoch": 2.2408477842003856, + "grad_norm": 0.7175054861537781, + "learning_rate": 1.8277627442725858e-06, + "loss": 0.0416, + "step": 3489 + }, + { + "epoch": 2.241490044958253, + "grad_norm": 0.697095436941284, + "learning_rate": 1.8248748080886958e-06, + "loss": 0.0247, + "step": 3490 + }, + { + "epoch": 2.2421323057161207, + "grad_norm": 0.691491329053507, + "learning_rate": 1.8219886458565378e-06, + "loss": 0.0232, + "step": 3491 + }, + { + "epoch": 2.2427745664739884, + "grad_norm": 0.9981050387863222, + "learning_rate": 1.8191042591886198e-06, + "loss": 0.0517, + "step": 3492 + }, + { + "epoch": 2.243416827231856, + "grad_norm": 0.7178046596046346, + "learning_rate": 1.816221649696457e-06, + "loss": 0.0261, + "step": 3493 + }, + { + "epoch": 2.244059087989724, + "grad_norm": 0.647129185070157, + "learning_rate": 1.813340818990571e-06, + "loss": 0.0177, + "step": 3494 + }, + { + "epoch": 2.2447013487475918, + "grad_norm": 0.8357769361363427, + "learning_rate": 1.810461768680491e-06, + "loss": 0.032, + "step": 3495 + }, + { + "epoch": 2.245343609505459, + "grad_norm": 0.9135537029828213, + "learning_rate": 1.807584500374751e-06, + "loss": 0.0306, + "step": 3496 + }, + { + "epoch": 2.245985870263327, + "grad_norm": 0.8028765750272937, + "learning_rate": 1.8047090156808888e-06, + "loss": 0.0368, + "step": 3497 + }, + { + "epoch": 2.2466281310211946, + "grad_norm": 1.0549332243691092, + "learning_rate": 1.8018353162054458e-06, + "loss": 0.0565, + "step": 3498 + }, + { + "epoch": 2.2472703917790624, + "grad_norm": 0.9225294291357655, + "learning_rate": 1.7989634035539671e-06, + "loss": 0.0398, + "step": 3499 + }, + { + "epoch": 2.24791265253693, + "grad_norm": 0.7733750366849621, + "learning_rate": 1.7960932793309988e-06, + "loss": 0.0308, + "step": 3500 + }, + { + "epoch": 2.2485549132947975, + "grad_norm": 0.8058426188736364, + "learning_rate": 1.7932249451400863e-06, + "loss": 0.0372, + "step": 3501 + }, + { + "epoch": 2.2491971740526653, + "grad_norm": 0.8022434055087563, + "learning_rate": 1.7903584025837778e-06, + "loss": 0.034, + "step": 3502 + }, + { + "epoch": 2.249839434810533, + "grad_norm": 0.7043254714487627, + "learning_rate": 1.7874936532636182e-06, + "loss": 0.0275, + "step": 3503 + }, + { + "epoch": 2.250481695568401, + "grad_norm": 0.8866214533003941, + "learning_rate": 1.7846306987801514e-06, + "loss": 0.0359, + "step": 3504 + }, + { + "epoch": 2.2511239563262686, + "grad_norm": 0.542838845693405, + "learning_rate": 1.7817695407329205e-06, + "loss": 0.0141, + "step": 3505 + }, + { + "epoch": 2.2517662170841364, + "grad_norm": 0.991773893876819, + "learning_rate": 1.7789101807204594e-06, + "loss": 0.0572, + "step": 3506 + }, + { + "epoch": 2.2524084778420037, + "grad_norm": 0.7767896650706603, + "learning_rate": 1.776052620340301e-06, + "loss": 0.0231, + "step": 3507 + }, + { + "epoch": 2.2530507385998715, + "grad_norm": 0.8895084018437627, + "learning_rate": 1.7731968611889756e-06, + "loss": 0.0396, + "step": 3508 + }, + { + "epoch": 2.2536929993577393, + "grad_norm": 0.7008220238963826, + "learning_rate": 1.7703429048620024e-06, + "loss": 0.0302, + "step": 3509 + }, + { + "epoch": 2.254335260115607, + "grad_norm": 1.075892377518874, + "learning_rate": 1.767490752953896e-06, + "loss": 0.0567, + "step": 3510 + }, + { + "epoch": 2.254977520873475, + "grad_norm": 1.0163609879175337, + "learning_rate": 1.7646404070581613e-06, + "loss": 0.0538, + "step": 3511 + }, + { + "epoch": 2.255619781631342, + "grad_norm": 0.7168361174523952, + "learning_rate": 1.7617918687672947e-06, + "loss": 0.0297, + "step": 3512 + }, + { + "epoch": 2.25626204238921, + "grad_norm": 0.65633860620476, + "learning_rate": 1.7589451396727853e-06, + "loss": 0.0218, + "step": 3513 + }, + { + "epoch": 2.2569043031470777, + "grad_norm": 0.8025854148748303, + "learning_rate": 1.7561002213651052e-06, + "loss": 0.039, + "step": 3514 + }, + { + "epoch": 2.2575465639049455, + "grad_norm": 0.8468375552416947, + "learning_rate": 1.7532571154337202e-06, + "loss": 0.0352, + "step": 3515 + }, + { + "epoch": 2.2581888246628132, + "grad_norm": 0.6456444824641062, + "learning_rate": 1.750415823467082e-06, + "loss": 0.0218, + "step": 3516 + }, + { + "epoch": 2.258831085420681, + "grad_norm": 0.7816843169755874, + "learning_rate": 1.7475763470526286e-06, + "loss": 0.0314, + "step": 3517 + }, + { + "epoch": 2.2594733461785483, + "grad_norm": 0.8352784216669287, + "learning_rate": 1.744738687776784e-06, + "loss": 0.0312, + "step": 3518 + }, + { + "epoch": 2.260115606936416, + "grad_norm": 0.7515897404363285, + "learning_rate": 1.7419028472249566e-06, + "loss": 0.0333, + "step": 3519 + }, + { + "epoch": 2.260757867694284, + "grad_norm": 0.6521494569428792, + "learning_rate": 1.7390688269815375e-06, + "loss": 0.0231, + "step": 3520 + }, + { + "epoch": 2.2614001284521517, + "grad_norm": 0.9551563062410857, + "learning_rate": 1.7362366286299065e-06, + "loss": 0.0433, + "step": 3521 + }, + { + "epoch": 2.2620423892100194, + "grad_norm": 0.8845377193871722, + "learning_rate": 1.7334062537524165e-06, + "loss": 0.0429, + "step": 3522 + }, + { + "epoch": 2.2626846499678868, + "grad_norm": 0.7124970554387072, + "learning_rate": 1.7305777039304078e-06, + "loss": 0.0311, + "step": 3523 + }, + { + "epoch": 2.2633269107257545, + "grad_norm": 0.8435454840362326, + "learning_rate": 1.7277509807441994e-06, + "loss": 0.0429, + "step": 3524 + }, + { + "epoch": 2.2639691714836223, + "grad_norm": 1.1198593545392712, + "learning_rate": 1.7249260857730898e-06, + "loss": 0.0685, + "step": 3525 + }, + { + "epoch": 2.26461143224149, + "grad_norm": 1.0189173449019602, + "learning_rate": 1.7221030205953565e-06, + "loss": 0.0492, + "step": 3526 + }, + { + "epoch": 2.265253692999358, + "grad_norm": 0.606656670662427, + "learning_rate": 1.7192817867882533e-06, + "loss": 0.0317, + "step": 3527 + }, + { + "epoch": 2.2658959537572256, + "grad_norm": 0.6174747931288639, + "learning_rate": 1.7164623859280144e-06, + "loss": 0.0302, + "step": 3528 + }, + { + "epoch": 2.266538214515093, + "grad_norm": 0.9542653055229993, + "learning_rate": 1.713644819589842e-06, + "loss": 0.0359, + "step": 3529 + }, + { + "epoch": 2.2671804752729607, + "grad_norm": 0.9567485918935901, + "learning_rate": 1.7108290893479234e-06, + "loss": 0.0455, + "step": 3530 + }, + { + "epoch": 2.2678227360308285, + "grad_norm": 0.36517362236748896, + "learning_rate": 1.7080151967754145e-06, + "loss": 0.0144, + "step": 3531 + }, + { + "epoch": 2.2684649967886963, + "grad_norm": 0.42214402734192774, + "learning_rate": 1.7052031434444438e-06, + "loss": 0.0087, + "step": 3532 + }, + { + "epoch": 2.269107257546564, + "grad_norm": 0.956553350470016, + "learning_rate": 1.702392930926115e-06, + "loss": 0.0274, + "step": 3533 + }, + { + "epoch": 2.2697495183044314, + "grad_norm": 0.7170168399600096, + "learning_rate": 1.699584560790502e-06, + "loss": 0.0287, + "step": 3534 + }, + { + "epoch": 2.270391779062299, + "grad_norm": 0.5955881431151817, + "learning_rate": 1.6967780346066487e-06, + "loss": 0.0308, + "step": 3535 + }, + { + "epoch": 2.271034039820167, + "grad_norm": 0.48160043213329706, + "learning_rate": 1.6939733539425717e-06, + "loss": 0.0172, + "step": 3536 + }, + { + "epoch": 2.2716763005780347, + "grad_norm": 0.820943696393198, + "learning_rate": 1.6911705203652506e-06, + "loss": 0.0363, + "step": 3537 + }, + { + "epoch": 2.2723185613359025, + "grad_norm": 0.4415706131083725, + "learning_rate": 1.6883695354406387e-06, + "loss": 0.011, + "step": 3538 + }, + { + "epoch": 2.2729608220937703, + "grad_norm": 0.9395889198863401, + "learning_rate": 1.6855704007336544e-06, + "loss": 0.0502, + "step": 3539 + }, + { + "epoch": 2.2736030828516376, + "grad_norm": 1.027785370469706, + "learning_rate": 1.6827731178081825e-06, + "loss": 0.0513, + "step": 3540 + }, + { + "epoch": 2.2742453436095054, + "grad_norm": 0.7814923010956623, + "learning_rate": 1.6799776882270707e-06, + "loss": 0.0348, + "step": 3541 + }, + { + "epoch": 2.274887604367373, + "grad_norm": 0.6344631347650332, + "learning_rate": 1.6771841135521383e-06, + "loss": 0.0251, + "step": 3542 + }, + { + "epoch": 2.275529865125241, + "grad_norm": 0.5335871184690629, + "learning_rate": 1.6743923953441616e-06, + "loss": 0.0144, + "step": 3543 + }, + { + "epoch": 2.2761721258831087, + "grad_norm": 0.8043076928694868, + "learning_rate": 1.6716025351628827e-06, + "loss": 0.0343, + "step": 3544 + }, + { + "epoch": 2.276814386640976, + "grad_norm": 0.6015330171910458, + "learning_rate": 1.668814534567002e-06, + "loss": 0.0227, + "step": 3545 + }, + { + "epoch": 2.277456647398844, + "grad_norm": 0.894996541131454, + "learning_rate": 1.6660283951141847e-06, + "loss": 0.0495, + "step": 3546 + }, + { + "epoch": 2.2780989081567116, + "grad_norm": 0.46844910924394034, + "learning_rate": 1.663244118361056e-06, + "loss": 0.0129, + "step": 3547 + }, + { + "epoch": 2.2787411689145793, + "grad_norm": 0.47805861888940654, + "learning_rate": 1.6604617058631983e-06, + "loss": 0.0216, + "step": 3548 + }, + { + "epoch": 2.279383429672447, + "grad_norm": 0.9446146901541599, + "learning_rate": 1.6576811591751545e-06, + "loss": 0.0474, + "step": 3549 + }, + { + "epoch": 2.280025690430315, + "grad_norm": 0.5903376752635773, + "learning_rate": 1.6549024798504236e-06, + "loss": 0.0237, + "step": 3550 + }, + { + "epoch": 2.280667951188182, + "grad_norm": 0.6526490021245588, + "learning_rate": 1.6521256694414622e-06, + "loss": 0.0294, + "step": 3551 + }, + { + "epoch": 2.28131021194605, + "grad_norm": 0.9846479558996006, + "learning_rate": 1.6493507294996825e-06, + "loss": 0.035, + "step": 3552 + }, + { + "epoch": 2.2819524727039178, + "grad_norm": 0.6725292706623651, + "learning_rate": 1.646577661575451e-06, + "loss": 0.0315, + "step": 3553 + }, + { + "epoch": 2.2825947334617855, + "grad_norm": 0.47236929259142585, + "learning_rate": 1.6438064672180893e-06, + "loss": 0.0166, + "step": 3554 + }, + { + "epoch": 2.2832369942196533, + "grad_norm": 0.7035595104385141, + "learning_rate": 1.641037147975872e-06, + "loss": 0.0241, + "step": 3555 + }, + { + "epoch": 2.2838792549775206, + "grad_norm": 1.066952155374257, + "learning_rate": 1.6382697053960251e-06, + "loss": 0.0462, + "step": 3556 + }, + { + "epoch": 2.2845215157353884, + "grad_norm": 0.7633801293273276, + "learning_rate": 1.6355041410247274e-06, + "loss": 0.0262, + "step": 3557 + }, + { + "epoch": 2.285163776493256, + "grad_norm": 1.0028542106956315, + "learning_rate": 1.632740456407108e-06, + "loss": 0.0487, + "step": 3558 + }, + { + "epoch": 2.285806037251124, + "grad_norm": 0.7679907659537593, + "learning_rate": 1.6299786530872453e-06, + "loss": 0.0322, + "step": 3559 + }, + { + "epoch": 2.2864482980089917, + "grad_norm": 0.9482681377458007, + "learning_rate": 1.627218732608169e-06, + "loss": 0.0527, + "step": 3560 + }, + { + "epoch": 2.2870905587668595, + "grad_norm": 0.8518887782185082, + "learning_rate": 1.6244606965118504e-06, + "loss": 0.03, + "step": 3561 + }, + { + "epoch": 2.287732819524727, + "grad_norm": 0.5814597499259083, + "learning_rate": 1.6217045463392128e-06, + "loss": 0.0239, + "step": 3562 + }, + { + "epoch": 2.2883750802825946, + "grad_norm": 0.599105836632565, + "learning_rate": 1.618950283630129e-06, + "loss": 0.0215, + "step": 3563 + }, + { + "epoch": 2.2890173410404624, + "grad_norm": 0.7796000461578123, + "learning_rate": 1.616197909923412e-06, + "loss": 0.0282, + "step": 3564 + }, + { + "epoch": 2.28965960179833, + "grad_norm": 0.7275207322382874, + "learning_rate": 1.6134474267568207e-06, + "loss": 0.0336, + "step": 3565 + }, + { + "epoch": 2.290301862556198, + "grad_norm": 0.6392597391595739, + "learning_rate": 1.6106988356670584e-06, + "loss": 0.023, + "step": 3566 + }, + { + "epoch": 2.2909441233140653, + "grad_norm": 0.6181072081042894, + "learning_rate": 1.6079521381897707e-06, + "loss": 0.0256, + "step": 3567 + }, + { + "epoch": 2.291586384071933, + "grad_norm": 0.539694764751977, + "learning_rate": 1.6052073358595472e-06, + "loss": 0.0178, + "step": 3568 + }, + { + "epoch": 2.292228644829801, + "grad_norm": 0.8189738538849245, + "learning_rate": 1.602464430209914e-06, + "loss": 0.0327, + "step": 3569 + }, + { + "epoch": 2.2928709055876686, + "grad_norm": 0.9092697161268417, + "learning_rate": 1.5997234227733417e-06, + "loss": 0.0418, + "step": 3570 + }, + { + "epoch": 2.2935131663455364, + "grad_norm": 0.6696720563787807, + "learning_rate": 1.5969843150812392e-06, + "loss": 0.0163, + "step": 3571 + }, + { + "epoch": 2.294155427103404, + "grad_norm": 0.40693272658030666, + "learning_rate": 1.5942471086639543e-06, + "loss": 0.0134, + "step": 3572 + }, + { + "epoch": 2.294797687861272, + "grad_norm": 0.6199070207997924, + "learning_rate": 1.591511805050772e-06, + "loss": 0.0256, + "step": 3573 + }, + { + "epoch": 2.2954399486191392, + "grad_norm": 1.083303800840011, + "learning_rate": 1.588778405769915e-06, + "loss": 0.0615, + "step": 3574 + }, + { + "epoch": 2.296082209377007, + "grad_norm": 0.9679308542275294, + "learning_rate": 1.5860469123485389e-06, + "loss": 0.0462, + "step": 3575 + }, + { + "epoch": 2.296724470134875, + "grad_norm": 0.8554955504823838, + "learning_rate": 1.5833173263127427e-06, + "loss": 0.0307, + "step": 3576 + }, + { + "epoch": 2.2973667308927426, + "grad_norm": 0.7274771979882341, + "learning_rate": 1.580589649187549e-06, + "loss": 0.0245, + "step": 3577 + }, + { + "epoch": 2.29800899165061, + "grad_norm": 0.8666335433989327, + "learning_rate": 1.5778638824969206e-06, + "loss": 0.0385, + "step": 3578 + }, + { + "epoch": 2.2986512524084777, + "grad_norm": 0.825612874507597, + "learning_rate": 1.5751400277637514e-06, + "loss": 0.0363, + "step": 3579 + }, + { + "epoch": 2.2992935131663454, + "grad_norm": 0.6329164924743577, + "learning_rate": 1.5724180865098665e-06, + "loss": 0.0254, + "step": 3580 + }, + { + "epoch": 2.2999357739242132, + "grad_norm": 0.3909721338235195, + "learning_rate": 1.569698060256023e-06, + "loss": 0.0103, + "step": 3581 + }, + { + "epoch": 2.300578034682081, + "grad_norm": 1.1270038175306214, + "learning_rate": 1.5669799505219069e-06, + "loss": 0.0645, + "step": 3582 + }, + { + "epoch": 2.3012202954399488, + "grad_norm": 0.8638921948143339, + "learning_rate": 1.5642637588261367e-06, + "loss": 0.0455, + "step": 3583 + }, + { + "epoch": 2.3018625561978165, + "grad_norm": 0.8854266987497567, + "learning_rate": 1.5615494866862512e-06, + "loss": 0.0393, + "step": 3584 + }, + { + "epoch": 2.302504816955684, + "grad_norm": 0.6395881750442347, + "learning_rate": 1.558837135618727e-06, + "loss": 0.0328, + "step": 3585 + }, + { + "epoch": 2.3031470777135516, + "grad_norm": 0.76326036129717, + "learning_rate": 1.5561267071389606e-06, + "loss": 0.0338, + "step": 3586 + }, + { + "epoch": 2.3037893384714194, + "grad_norm": 1.0563398912643513, + "learning_rate": 1.5534182027612776e-06, + "loss": 0.0583, + "step": 3587 + }, + { + "epoch": 2.304431599229287, + "grad_norm": 0.751351294437116, + "learning_rate": 1.5507116239989262e-06, + "loss": 0.0179, + "step": 3588 + }, + { + "epoch": 2.305073859987155, + "grad_norm": 0.5917324372106778, + "learning_rate": 1.5480069723640795e-06, + "loss": 0.0179, + "step": 3589 + }, + { + "epoch": 2.3057161207450223, + "grad_norm": 0.8184951601476342, + "learning_rate": 1.5453042493678356e-06, + "loss": 0.0298, + "step": 3590 + }, + { + "epoch": 2.30635838150289, + "grad_norm": 0.5871341197591803, + "learning_rate": 1.542603456520214e-06, + "loss": 0.0195, + "step": 3591 + }, + { + "epoch": 2.307000642260758, + "grad_norm": 0.6484019922545525, + "learning_rate": 1.5399045953301539e-06, + "loss": 0.0252, + "step": 3592 + }, + { + "epoch": 2.3076429030186256, + "grad_norm": 0.8160901045064359, + "learning_rate": 1.537207667305517e-06, + "loss": 0.039, + "step": 3593 + }, + { + "epoch": 2.3082851637764934, + "grad_norm": 0.5761180941519981, + "learning_rate": 1.5345126739530857e-06, + "loss": 0.0208, + "step": 3594 + }, + { + "epoch": 2.308927424534361, + "grad_norm": 0.8030300624236311, + "learning_rate": 1.531819616778561e-06, + "loss": 0.0281, + "step": 3595 + }, + { + "epoch": 2.3095696852922285, + "grad_norm": 0.8026752262998608, + "learning_rate": 1.529128497286559e-06, + "loss": 0.0238, + "step": 3596 + }, + { + "epoch": 2.3102119460500963, + "grad_norm": 0.6805116866619696, + "learning_rate": 1.5264393169806202e-06, + "loss": 0.0194, + "step": 3597 + }, + { + "epoch": 2.310854206807964, + "grad_norm": 1.2412593191188284, + "learning_rate": 1.523752077363196e-06, + "loss": 0.0678, + "step": 3598 + }, + { + "epoch": 2.311496467565832, + "grad_norm": 0.6615861624386179, + "learning_rate": 1.521066779935656e-06, + "loss": 0.0244, + "step": 3599 + }, + { + "epoch": 2.3121387283236996, + "grad_norm": 0.8704800145755766, + "learning_rate": 1.5183834261982804e-06, + "loss": 0.0259, + "step": 3600 + }, + { + "epoch": 2.312780989081567, + "grad_norm": 0.7338237189632218, + "learning_rate": 1.5157020176502695e-06, + "loss": 0.0288, + "step": 3601 + }, + { + "epoch": 2.3134232498394347, + "grad_norm": 0.7646804674409474, + "learning_rate": 1.5130225557897332e-06, + "loss": 0.0307, + "step": 3602 + }, + { + "epoch": 2.3140655105973025, + "grad_norm": 0.6318278098679664, + "learning_rate": 1.510345042113695e-06, + "loss": 0.0213, + "step": 3603 + }, + { + "epoch": 2.3147077713551703, + "grad_norm": 0.5636429072877498, + "learning_rate": 1.5076694781180895e-06, + "loss": 0.0146, + "step": 3604 + }, + { + "epoch": 2.315350032113038, + "grad_norm": 0.9199535141888329, + "learning_rate": 1.504995865297762e-06, + "loss": 0.0454, + "step": 3605 + }, + { + "epoch": 2.315992292870906, + "grad_norm": 0.6407000254243712, + "learning_rate": 1.5023242051464677e-06, + "loss": 0.0257, + "step": 3606 + }, + { + "epoch": 2.316634553628773, + "grad_norm": 0.8166852551366244, + "learning_rate": 1.4996544991568718e-06, + "loss": 0.0351, + "step": 3607 + }, + { + "epoch": 2.317276814386641, + "grad_norm": 0.7293367775185455, + "learning_rate": 1.4969867488205458e-06, + "loss": 0.0414, + "step": 3608 + }, + { + "epoch": 2.3179190751445087, + "grad_norm": 0.6400037449100937, + "learning_rate": 1.4943209556279698e-06, + "loss": 0.0254, + "step": 3609 + }, + { + "epoch": 2.3185613359023765, + "grad_norm": 0.5947372066386746, + "learning_rate": 1.491657121068531e-06, + "loss": 0.0192, + "step": 3610 + }, + { + "epoch": 2.3192035966602442, + "grad_norm": 0.50635988952308, + "learning_rate": 1.4889952466305207e-06, + "loss": 0.0138, + "step": 3611 + }, + { + "epoch": 2.3198458574181116, + "grad_norm": 0.8227821108149866, + "learning_rate": 1.4863353338011367e-06, + "loss": 0.0263, + "step": 3612 + }, + { + "epoch": 2.3204881181759793, + "grad_norm": 0.7278361419500429, + "learning_rate": 1.4836773840664798e-06, + "loss": 0.0251, + "step": 3613 + }, + { + "epoch": 2.321130378933847, + "grad_norm": 0.7077285517353684, + "learning_rate": 1.4810213989115536e-06, + "loss": 0.0254, + "step": 3614 + }, + { + "epoch": 2.321772639691715, + "grad_norm": 0.835188504438919, + "learning_rate": 1.4783673798202674e-06, + "loss": 0.0198, + "step": 3615 + }, + { + "epoch": 2.3224149004495827, + "grad_norm": 0.8750792807662505, + "learning_rate": 1.4757153282754255e-06, + "loss": 0.0361, + "step": 3616 + }, + { + "epoch": 2.3230571612074504, + "grad_norm": 0.9144626527075307, + "learning_rate": 1.4730652457587386e-06, + "loss": 0.0321, + "step": 3617 + }, + { + "epoch": 2.3236994219653178, + "grad_norm": 0.6188718195637551, + "learning_rate": 1.4704171337508144e-06, + "loss": 0.0235, + "step": 3618 + }, + { + "epoch": 2.3243416827231855, + "grad_norm": 1.0882472970815384, + "learning_rate": 1.4677709937311636e-06, + "loss": 0.0529, + "step": 3619 + }, + { + "epoch": 2.3249839434810533, + "grad_norm": 0.6222226157584902, + "learning_rate": 1.465126827178191e-06, + "loss": 0.0178, + "step": 3620 + }, + { + "epoch": 2.325626204238921, + "grad_norm": 0.7654149008179505, + "learning_rate": 1.4624846355692002e-06, + "loss": 0.022, + "step": 3621 + }, + { + "epoch": 2.326268464996789, + "grad_norm": 0.9756973108525397, + "learning_rate": 1.4598444203803908e-06, + "loss": 0.0439, + "step": 3622 + }, + { + "epoch": 2.326910725754656, + "grad_norm": 0.7787746582722757, + "learning_rate": 1.4572061830868612e-06, + "loss": 0.0294, + "step": 3623 + }, + { + "epoch": 2.327552986512524, + "grad_norm": 0.9830355656561385, + "learning_rate": 1.4545699251625982e-06, + "loss": 0.0448, + "step": 3624 + }, + { + "epoch": 2.3281952472703917, + "grad_norm": 0.8308234787190815, + "learning_rate": 1.451935648080489e-06, + "loss": 0.0294, + "step": 3625 + }, + { + "epoch": 2.3288375080282595, + "grad_norm": 0.7188667837098013, + "learning_rate": 1.4493033533123125e-06, + "loss": 0.0343, + "step": 3626 + }, + { + "epoch": 2.3294797687861273, + "grad_norm": 0.5933480071823289, + "learning_rate": 1.4466730423287385e-06, + "loss": 0.0245, + "step": 3627 + }, + { + "epoch": 2.330122029543995, + "grad_norm": 0.6926615555331728, + "learning_rate": 1.4440447165993292e-06, + "loss": 0.024, + "step": 3628 + }, + { + "epoch": 2.3307642903018624, + "grad_norm": 1.1619441615958566, + "learning_rate": 1.4414183775925389e-06, + "loss": 0.07, + "step": 3629 + }, + { + "epoch": 2.33140655105973, + "grad_norm": 0.9592077221597384, + "learning_rate": 1.4387940267757084e-06, + "loss": 0.037, + "step": 3630 + }, + { + "epoch": 2.332048811817598, + "grad_norm": 0.847610813740467, + "learning_rate": 1.4361716656150759e-06, + "loss": 0.0436, + "step": 3631 + }, + { + "epoch": 2.3326910725754657, + "grad_norm": 0.6398715124706327, + "learning_rate": 1.4335512955757568e-06, + "loss": 0.0133, + "step": 3632 + }, + { + "epoch": 2.3333333333333335, + "grad_norm": 0.6277876594147598, + "learning_rate": 1.4309329181217607e-06, + "loss": 0.017, + "step": 3633 + }, + { + "epoch": 2.333975594091201, + "grad_norm": 0.653401513250579, + "learning_rate": 1.4283165347159838e-06, + "loss": 0.0261, + "step": 3634 + }, + { + "epoch": 2.3346178548490686, + "grad_norm": 0.5569124643064362, + "learning_rate": 1.4257021468202065e-06, + "loss": 0.021, + "step": 3635 + }, + { + "epoch": 2.3352601156069364, + "grad_norm": 0.7791432421430815, + "learning_rate": 1.423089755895095e-06, + "loss": 0.0269, + "step": 3636 + }, + { + "epoch": 2.335902376364804, + "grad_norm": 0.930879092217143, + "learning_rate": 1.4204793634002006e-06, + "loss": 0.0448, + "step": 3637 + }, + { + "epoch": 2.336544637122672, + "grad_norm": 0.878838464902805, + "learning_rate": 1.4178709707939576e-06, + "loss": 0.0434, + "step": 3638 + }, + { + "epoch": 2.3371868978805397, + "grad_norm": 0.734668348227472, + "learning_rate": 1.4152645795336777e-06, + "loss": 0.0272, + "step": 3639 + }, + { + "epoch": 2.337829158638407, + "grad_norm": 0.7899563881895629, + "learning_rate": 1.4126601910755655e-06, + "loss": 0.0327, + "step": 3640 + }, + { + "epoch": 2.338471419396275, + "grad_norm": 0.9462302067020157, + "learning_rate": 1.4100578068746974e-06, + "loss": 0.0288, + "step": 3641 + }, + { + "epoch": 2.3391136801541426, + "grad_norm": 0.4026728428529778, + "learning_rate": 1.4074574283850339e-06, + "loss": 0.0115, + "step": 3642 + }, + { + "epoch": 2.3397559409120103, + "grad_norm": 0.6329700557751123, + "learning_rate": 1.404859057059414e-06, + "loss": 0.0232, + "step": 3643 + }, + { + "epoch": 2.340398201669878, + "grad_norm": 0.6737718164973577, + "learning_rate": 1.4022626943495553e-06, + "loss": 0.0163, + "step": 3644 + }, + { + "epoch": 2.3410404624277454, + "grad_norm": 0.86461532304177, + "learning_rate": 1.399668341706053e-06, + "loss": 0.0255, + "step": 3645 + }, + { + "epoch": 2.341682723185613, + "grad_norm": 0.7626041264127075, + "learning_rate": 1.3970760005783813e-06, + "loss": 0.0359, + "step": 3646 + }, + { + "epoch": 2.342324983943481, + "grad_norm": 0.8913878554093544, + "learning_rate": 1.3944856724148853e-06, + "loss": 0.0304, + "step": 3647 + }, + { + "epoch": 2.3429672447013488, + "grad_norm": 0.9088154975378596, + "learning_rate": 1.391897358662791e-06, + "loss": 0.0324, + "step": 3648 + }, + { + "epoch": 2.3436095054592165, + "grad_norm": 0.770505613828181, + "learning_rate": 1.389311060768197e-06, + "loss": 0.0351, + "step": 3649 + }, + { + "epoch": 2.3442517662170843, + "grad_norm": 0.5958112466984912, + "learning_rate": 1.3867267801760748e-06, + "loss": 0.0189, + "step": 3650 + }, + { + "epoch": 2.3448940269749516, + "grad_norm": 1.0161310559382941, + "learning_rate": 1.3841445183302682e-06, + "loss": 0.0438, + "step": 3651 + }, + { + "epoch": 2.3455362877328194, + "grad_norm": 1.0900562331428258, + "learning_rate": 1.3815642766734978e-06, + "loss": 0.0479, + "step": 3652 + }, + { + "epoch": 2.346178548490687, + "grad_norm": 0.8935065147882977, + "learning_rate": 1.37898605664735e-06, + "loss": 0.0441, + "step": 3653 + }, + { + "epoch": 2.346820809248555, + "grad_norm": 0.7600600541549287, + "learning_rate": 1.3764098596922865e-06, + "loss": 0.0348, + "step": 3654 + }, + { + "epoch": 2.3474630700064227, + "grad_norm": 0.7675267027631106, + "learning_rate": 1.3738356872476322e-06, + "loss": 0.0298, + "step": 3655 + }, + { + "epoch": 2.34810533076429, + "grad_norm": 0.7443189016599652, + "learning_rate": 1.371263540751586e-06, + "loss": 0.0256, + "step": 3656 + }, + { + "epoch": 2.348747591522158, + "grad_norm": 0.8611636472047727, + "learning_rate": 1.368693421641215e-06, + "loss": 0.0421, + "step": 3657 + }, + { + "epoch": 2.3493898522800256, + "grad_norm": 0.7989533511846363, + "learning_rate": 1.3661253313524509e-06, + "loss": 0.019, + "step": 3658 + }, + { + "epoch": 2.3500321130378934, + "grad_norm": 0.7748553867408243, + "learning_rate": 1.363559271320094e-06, + "loss": 0.018, + "step": 3659 + }, + { + "epoch": 2.350674373795761, + "grad_norm": 0.6910663912747925, + "learning_rate": 1.3609952429778094e-06, + "loss": 0.0273, + "step": 3660 + }, + { + "epoch": 2.351316634553629, + "grad_norm": 0.8171816712774966, + "learning_rate": 1.3584332477581269e-06, + "loss": 0.0346, + "step": 3661 + }, + { + "epoch": 2.3519588953114967, + "grad_norm": 1.0957241607100145, + "learning_rate": 1.3558732870924414e-06, + "loss": 0.0285, + "step": 3662 + }, + { + "epoch": 2.352601156069364, + "grad_norm": 0.6514707172844506, + "learning_rate": 1.3533153624110097e-06, + "loss": 0.0212, + "step": 3663 + }, + { + "epoch": 2.353243416827232, + "grad_norm": 0.723589636024266, + "learning_rate": 1.3507594751429515e-06, + "loss": 0.0255, + "step": 3664 + }, + { + "epoch": 2.3538856775850996, + "grad_norm": 0.7099514077535453, + "learning_rate": 1.3482056267162497e-06, + "loss": 0.0292, + "step": 3665 + }, + { + "epoch": 2.3545279383429674, + "grad_norm": 0.627608388204718, + "learning_rate": 1.3456538185577467e-06, + "loss": 0.0258, + "step": 3666 + }, + { + "epoch": 2.3551701991008347, + "grad_norm": 0.9552336478496908, + "learning_rate": 1.3431040520931438e-06, + "loss": 0.0332, + "step": 3667 + }, + { + "epoch": 2.3558124598587025, + "grad_norm": 0.6566727810619335, + "learning_rate": 1.3405563287470046e-06, + "loss": 0.0328, + "step": 3668 + }, + { + "epoch": 2.3564547206165702, + "grad_norm": 0.9612787031931807, + "learning_rate": 1.3380106499427486e-06, + "loss": 0.0368, + "step": 3669 + }, + { + "epoch": 2.357096981374438, + "grad_norm": 0.8217062371806092, + "learning_rate": 1.3354670171026562e-06, + "loss": 0.0337, + "step": 3670 + }, + { + "epoch": 2.357739242132306, + "grad_norm": 0.7231730719003583, + "learning_rate": 1.332925431647859e-06, + "loss": 0.0358, + "step": 3671 + }, + { + "epoch": 2.3583815028901736, + "grad_norm": 0.6026200328125043, + "learning_rate": 1.3303858949983495e-06, + "loss": 0.0169, + "step": 3672 + }, + { + "epoch": 2.3590237636480413, + "grad_norm": 0.5446038613417785, + "learning_rate": 1.3278484085729736e-06, + "loss": 0.0231, + "step": 3673 + }, + { + "epoch": 2.3596660244059087, + "grad_norm": 0.9450233373053333, + "learning_rate": 1.325312973789435e-06, + "loss": 0.0408, + "step": 3674 + }, + { + "epoch": 2.3603082851637764, + "grad_norm": 0.9540983418398955, + "learning_rate": 1.3227795920642861e-06, + "loss": 0.0359, + "step": 3675 + }, + { + "epoch": 2.360950545921644, + "grad_norm": 0.8475711768764634, + "learning_rate": 1.320248264812936e-06, + "loss": 0.0347, + "step": 3676 + }, + { + "epoch": 2.361592806679512, + "grad_norm": 0.7505441192805897, + "learning_rate": 1.3177189934496443e-06, + "loss": 0.0167, + "step": 3677 + }, + { + "epoch": 2.3622350674373798, + "grad_norm": 0.9423177344997896, + "learning_rate": 1.3151917793875236e-06, + "loss": 0.0349, + "step": 3678 + }, + { + "epoch": 2.362877328195247, + "grad_norm": 0.744847152579368, + "learning_rate": 1.3126666240385328e-06, + "loss": 0.0412, + "step": 3679 + }, + { + "epoch": 2.363519588953115, + "grad_norm": 0.8186387701622767, + "learning_rate": 1.310143528813485e-06, + "loss": 0.0361, + "step": 3680 + }, + { + "epoch": 2.3641618497109826, + "grad_norm": 0.8881463631527913, + "learning_rate": 1.3076224951220413e-06, + "loss": 0.0356, + "step": 3681 + }, + { + "epoch": 2.3648041104688504, + "grad_norm": 0.7345496500783122, + "learning_rate": 1.30510352437271e-06, + "loss": 0.0239, + "step": 3682 + }, + { + "epoch": 2.365446371226718, + "grad_norm": 0.6999291433903357, + "learning_rate": 1.3025866179728481e-06, + "loss": 0.0241, + "step": 3683 + }, + { + "epoch": 2.366088631984586, + "grad_norm": 0.9991083102994782, + "learning_rate": 1.3000717773286582e-06, + "loss": 0.0418, + "step": 3684 + }, + { + "epoch": 2.3667308927424533, + "grad_norm": 0.6258859780400112, + "learning_rate": 1.2975590038451874e-06, + "loss": 0.0209, + "step": 3685 + }, + { + "epoch": 2.367373153500321, + "grad_norm": 1.1329719952459292, + "learning_rate": 1.295048298926334e-06, + "loss": 0.0543, + "step": 3686 + }, + { + "epoch": 2.368015414258189, + "grad_norm": 0.8072403647579103, + "learning_rate": 1.2925396639748321e-06, + "loss": 0.0388, + "step": 3687 + }, + { + "epoch": 2.3686576750160566, + "grad_norm": 0.7128825007313846, + "learning_rate": 1.2900331003922656e-06, + "loss": 0.0279, + "step": 3688 + }, + { + "epoch": 2.3692999357739244, + "grad_norm": 1.2042400876550716, + "learning_rate": 1.2875286095790574e-06, + "loss": 0.0532, + "step": 3689 + }, + { + "epoch": 2.3699421965317917, + "grad_norm": 0.5531783460716754, + "learning_rate": 1.2850261929344748e-06, + "loss": 0.017, + "step": 3690 + }, + { + "epoch": 2.3705844572896595, + "grad_norm": 0.5957770128436917, + "learning_rate": 1.2825258518566246e-06, + "loss": 0.0226, + "step": 3691 + }, + { + "epoch": 2.3712267180475273, + "grad_norm": 0.6546496131768716, + "learning_rate": 1.2800275877424555e-06, + "loss": 0.0256, + "step": 3692 + }, + { + "epoch": 2.371868978805395, + "grad_norm": 0.6977332062413512, + "learning_rate": 1.277531401987756e-06, + "loss": 0.0226, + "step": 3693 + }, + { + "epoch": 2.372511239563263, + "grad_norm": 0.9223815735010189, + "learning_rate": 1.2750372959871482e-06, + "loss": 0.0643, + "step": 3694 + }, + { + "epoch": 2.3731535003211306, + "grad_norm": 0.9450692590109001, + "learning_rate": 1.2725452711341002e-06, + "loss": 0.0286, + "step": 3695 + }, + { + "epoch": 2.373795761078998, + "grad_norm": 0.8074562354010631, + "learning_rate": 1.2700553288209129e-06, + "loss": 0.0382, + "step": 3696 + }, + { + "epoch": 2.3744380218368657, + "grad_norm": 0.87003617932423, + "learning_rate": 1.2675674704387242e-06, + "loss": 0.0409, + "step": 3697 + }, + { + "epoch": 2.3750802825947335, + "grad_norm": 0.8624548049607745, + "learning_rate": 1.265081697377507e-06, + "loss": 0.036, + "step": 3698 + }, + { + "epoch": 2.3757225433526012, + "grad_norm": 1.0523488923377293, + "learning_rate": 1.2625980110260711e-06, + "loss": 0.0349, + "step": 3699 + }, + { + "epoch": 2.376364804110469, + "grad_norm": 0.8935448332620956, + "learning_rate": 1.260116412772059e-06, + "loss": 0.0362, + "step": 3700 + }, + { + "epoch": 2.3770070648683364, + "grad_norm": 0.7284426564945694, + "learning_rate": 1.2576369040019483e-06, + "loss": 0.0274, + "step": 3701 + }, + { + "epoch": 2.377649325626204, + "grad_norm": 0.6352456184656788, + "learning_rate": 1.2551594861010441e-06, + "loss": 0.0204, + "step": 3702 + }, + { + "epoch": 2.378291586384072, + "grad_norm": 0.7311047328507669, + "learning_rate": 1.25268416045349e-06, + "loss": 0.0219, + "step": 3703 + }, + { + "epoch": 2.3789338471419397, + "grad_norm": 1.1683949517053662, + "learning_rate": 1.2502109284422564e-06, + "loss": 0.0603, + "step": 3704 + }, + { + "epoch": 2.3795761078998074, + "grad_norm": 0.5000465313337814, + "learning_rate": 1.2477397914491462e-06, + "loss": 0.0158, + "step": 3705 + }, + { + "epoch": 2.380218368657675, + "grad_norm": 0.8333605594132175, + "learning_rate": 1.24527075085479e-06, + "loss": 0.0443, + "step": 3706 + }, + { + "epoch": 2.3808606294155426, + "grad_norm": 0.626722918941087, + "learning_rate": 1.2428038080386474e-06, + "loss": 0.0252, + "step": 3707 + }, + { + "epoch": 2.3815028901734103, + "grad_norm": 0.7636023912190202, + "learning_rate": 1.2403389643790086e-06, + "loss": 0.0302, + "step": 3708 + }, + { + "epoch": 2.382145150931278, + "grad_norm": 0.7698421055800245, + "learning_rate": 1.2378762212529893e-06, + "loss": 0.0201, + "step": 3709 + }, + { + "epoch": 2.382787411689146, + "grad_norm": 0.6107227662180199, + "learning_rate": 1.2354155800365292e-06, + "loss": 0.0212, + "step": 3710 + }, + { + "epoch": 2.3834296724470136, + "grad_norm": 0.6685900492216744, + "learning_rate": 1.2329570421043957e-06, + "loss": 0.0339, + "step": 3711 + }, + { + "epoch": 2.384071933204881, + "grad_norm": 0.4568137180331092, + "learning_rate": 1.2305006088301825e-06, + "loss": 0.0117, + "step": 3712 + }, + { + "epoch": 2.3847141939627488, + "grad_norm": 1.1430847814320548, + "learning_rate": 1.2280462815863048e-06, + "loss": 0.0679, + "step": 3713 + }, + { + "epoch": 2.3853564547206165, + "grad_norm": 0.6999646021128558, + "learning_rate": 1.2255940617440038e-06, + "loss": 0.0314, + "step": 3714 + }, + { + "epoch": 2.3859987154784843, + "grad_norm": 1.0281794445785042, + "learning_rate": 1.2231439506733412e-06, + "loss": 0.043, + "step": 3715 + }, + { + "epoch": 2.386640976236352, + "grad_norm": 0.6931119006949129, + "learning_rate": 1.2206959497432007e-06, + "loss": 0.0164, + "step": 3716 + }, + { + "epoch": 2.38728323699422, + "grad_norm": 0.4288320019954368, + "learning_rate": 1.2182500603212882e-06, + "loss": 0.0148, + "step": 3717 + }, + { + "epoch": 2.387925497752087, + "grad_norm": 0.7143976399284087, + "learning_rate": 1.2158062837741291e-06, + "loss": 0.035, + "step": 3718 + }, + { + "epoch": 2.388567758509955, + "grad_norm": 0.8435584091872873, + "learning_rate": 1.2133646214670686e-06, + "loss": 0.0321, + "step": 3719 + }, + { + "epoch": 2.3892100192678227, + "grad_norm": 0.5298150406520792, + "learning_rate": 1.2109250747642709e-06, + "loss": 0.0115, + "step": 3720 + }, + { + "epoch": 2.3898522800256905, + "grad_norm": 0.946927079987968, + "learning_rate": 1.2084876450287175e-06, + "loss": 0.0491, + "step": 3721 + }, + { + "epoch": 2.3904945407835583, + "grad_norm": 0.9462610136849949, + "learning_rate": 1.2060523336222069e-06, + "loss": 0.0564, + "step": 3722 + }, + { + "epoch": 2.3911368015414256, + "grad_norm": 0.5058562965995588, + "learning_rate": 1.2036191419053561e-06, + "loss": 0.0162, + "step": 3723 + }, + { + "epoch": 2.3917790622992934, + "grad_norm": 0.7440665726641763, + "learning_rate": 1.2011880712375951e-06, + "loss": 0.0349, + "step": 3724 + }, + { + "epoch": 2.392421323057161, + "grad_norm": 1.047908761037079, + "learning_rate": 1.1987591229771728e-06, + "loss": 0.0403, + "step": 3725 + }, + { + "epoch": 2.393063583815029, + "grad_norm": 0.8979425529489828, + "learning_rate": 1.1963322984811454e-06, + "loss": 0.0368, + "step": 3726 + }, + { + "epoch": 2.3937058445728967, + "grad_norm": 0.6574800388767851, + "learning_rate": 1.19390759910539e-06, + "loss": 0.0152, + "step": 3727 + }, + { + "epoch": 2.3943481053307645, + "grad_norm": 0.859076422176542, + "learning_rate": 1.1914850262045903e-06, + "loss": 0.0453, + "step": 3728 + }, + { + "epoch": 2.394990366088632, + "grad_norm": 0.7318116892848443, + "learning_rate": 1.1890645811322487e-06, + "loss": 0.0266, + "step": 3729 + }, + { + "epoch": 2.3956326268464996, + "grad_norm": 0.883303916765513, + "learning_rate": 1.186646265240673e-06, + "loss": 0.0327, + "step": 3730 + }, + { + "epoch": 2.3962748876043674, + "grad_norm": 0.9391261835760442, + "learning_rate": 1.1842300798809837e-06, + "loss": 0.0453, + "step": 3731 + }, + { + "epoch": 2.396917148362235, + "grad_norm": 0.8388392344045262, + "learning_rate": 1.1818160264031097e-06, + "loss": 0.0385, + "step": 3732 + }, + { + "epoch": 2.397559409120103, + "grad_norm": 0.5399298094636439, + "learning_rate": 1.1794041061557926e-06, + "loss": 0.0139, + "step": 3733 + }, + { + "epoch": 2.3982016698779702, + "grad_norm": 0.44608846210564135, + "learning_rate": 1.1769943204865758e-06, + "loss": 0.0158, + "step": 3734 + }, + { + "epoch": 2.398843930635838, + "grad_norm": 0.9174116557708953, + "learning_rate": 1.1745866707418146e-06, + "loss": 0.0399, + "step": 3735 + }, + { + "epoch": 2.399486191393706, + "grad_norm": 0.6788753595452213, + "learning_rate": 1.172181158266671e-06, + "loss": 0.029, + "step": 3736 + }, + { + "epoch": 2.4001284521515736, + "grad_norm": 0.4145283685496298, + "learning_rate": 1.1697777844051105e-06, + "loss": 0.0122, + "step": 3737 + }, + { + "epoch": 2.4007707129094413, + "grad_norm": 0.86643890000884, + "learning_rate": 1.1673765504999062e-06, + "loss": 0.0331, + "step": 3738 + }, + { + "epoch": 2.401412973667309, + "grad_norm": 1.0752102859503734, + "learning_rate": 1.1649774578926337e-06, + "loss": 0.0444, + "step": 3739 + }, + { + "epoch": 2.4020552344251764, + "grad_norm": 0.6349503505076329, + "learning_rate": 1.162580507923674e-06, + "loss": 0.0265, + "step": 3740 + }, + { + "epoch": 2.402697495183044, + "grad_norm": 0.6645801739281773, + "learning_rate": 1.1601857019322093e-06, + "loss": 0.0225, + "step": 3741 + }, + { + "epoch": 2.403339755940912, + "grad_norm": 0.6778748049830092, + "learning_rate": 1.157793041256225e-06, + "loss": 0.0182, + "step": 3742 + }, + { + "epoch": 2.4039820166987798, + "grad_norm": 0.8254807437404275, + "learning_rate": 1.155402527232507e-06, + "loss": 0.0252, + "step": 3743 + }, + { + "epoch": 2.4046242774566475, + "grad_norm": 0.8948704368557262, + "learning_rate": 1.1530141611966438e-06, + "loss": 0.0446, + "step": 3744 + }, + { + "epoch": 2.405266538214515, + "grad_norm": 0.7707718927791404, + "learning_rate": 1.1506279444830216e-06, + "loss": 0.0216, + "step": 3745 + }, + { + "epoch": 2.4059087989723826, + "grad_norm": 0.6314810647184308, + "learning_rate": 1.1482438784248262e-06, + "loss": 0.0079, + "step": 3746 + }, + { + "epoch": 2.4065510597302504, + "grad_norm": 1.0106398653291409, + "learning_rate": 1.1458619643540437e-06, + "loss": 0.0433, + "step": 3747 + }, + { + "epoch": 2.407193320488118, + "grad_norm": 1.103895526279372, + "learning_rate": 1.1434822036014564e-06, + "loss": 0.0798, + "step": 3748 + }, + { + "epoch": 2.407835581245986, + "grad_norm": 1.0100834440008384, + "learning_rate": 1.1411045974966417e-06, + "loss": 0.0515, + "step": 3749 + }, + { + "epoch": 2.4084778420038537, + "grad_norm": 0.8208994186128443, + "learning_rate": 1.1387291473679746e-06, + "loss": 0.0402, + "step": 3750 + }, + { + "epoch": 2.4091201027617215, + "grad_norm": 0.8617191504544283, + "learning_rate": 1.136355854542629e-06, + "loss": 0.0364, + "step": 3751 + }, + { + "epoch": 2.409762363519589, + "grad_norm": 0.606359991514418, + "learning_rate": 1.1339847203465692e-06, + "loss": 0.0249, + "step": 3752 + }, + { + "epoch": 2.4104046242774566, + "grad_norm": 0.4848705124076297, + "learning_rate": 1.1316157461045553e-06, + "loss": 0.0175, + "step": 3753 + }, + { + "epoch": 2.4110468850353244, + "grad_norm": 0.812810217783158, + "learning_rate": 1.1292489331401391e-06, + "loss": 0.0303, + "step": 3754 + }, + { + "epoch": 2.411689145793192, + "grad_norm": 0.6029072785105867, + "learning_rate": 1.1268842827756665e-06, + "loss": 0.0134, + "step": 3755 + }, + { + "epoch": 2.4123314065510595, + "grad_norm": 0.7004905001882673, + "learning_rate": 1.1245217963322763e-06, + "loss": 0.0323, + "step": 3756 + }, + { + "epoch": 2.4129736673089273, + "grad_norm": 0.7479414509981465, + "learning_rate": 1.122161475129893e-06, + "loss": 0.0359, + "step": 3757 + }, + { + "epoch": 2.413615928066795, + "grad_norm": 0.633869333999919, + "learning_rate": 1.119803320487237e-06, + "loss": 0.0191, + "step": 3758 + }, + { + "epoch": 2.414258188824663, + "grad_norm": 0.6388783461299069, + "learning_rate": 1.1174473337218162e-06, + "loss": 0.0213, + "step": 3759 + }, + { + "epoch": 2.4149004495825306, + "grad_norm": 0.5996270021370579, + "learning_rate": 1.1150935161499266e-06, + "loss": 0.0221, + "step": 3760 + }, + { + "epoch": 2.4155427103403984, + "grad_norm": 1.028783463039958, + "learning_rate": 1.1127418690866537e-06, + "loss": 0.0429, + "step": 3761 + }, + { + "epoch": 2.416184971098266, + "grad_norm": 0.93571852248027, + "learning_rate": 1.1103923938458677e-06, + "loss": 0.0502, + "step": 3762 + }, + { + "epoch": 2.4168272318561335, + "grad_norm": 0.7666781805249379, + "learning_rate": 1.1080450917402308e-06, + "loss": 0.0295, + "step": 3763 + }, + { + "epoch": 2.4174694926140012, + "grad_norm": 1.29462701219975, + "learning_rate": 1.1056999640811866e-06, + "loss": 0.0872, + "step": 3764 + }, + { + "epoch": 2.418111753371869, + "grad_norm": 0.795726894652308, + "learning_rate": 1.1033570121789628e-06, + "loss": 0.0354, + "step": 3765 + }, + { + "epoch": 2.418754014129737, + "grad_norm": 0.7870764334671706, + "learning_rate": 1.1010162373425743e-06, + "loss": 0.0216, + "step": 3766 + }, + { + "epoch": 2.4193962748876046, + "grad_norm": 0.6499742531700223, + "learning_rate": 1.0986776408798188e-06, + "loss": 0.0245, + "step": 3767 + }, + { + "epoch": 2.420038535645472, + "grad_norm": 0.5534706666189165, + "learning_rate": 1.0963412240972775e-06, + "loss": 0.0204, + "step": 3768 + }, + { + "epoch": 2.4206807964033397, + "grad_norm": 0.6677002440042734, + "learning_rate": 1.0940069883003136e-06, + "loss": 0.0275, + "step": 3769 + }, + { + "epoch": 2.4213230571612074, + "grad_norm": 0.6132919794341785, + "learning_rate": 1.0916749347930706e-06, + "loss": 0.0111, + "step": 3770 + }, + { + "epoch": 2.421965317919075, + "grad_norm": 0.6412212877277955, + "learning_rate": 1.0893450648784736e-06, + "loss": 0.0204, + "step": 3771 + }, + { + "epoch": 2.422607578676943, + "grad_norm": 0.5302359782424138, + "learning_rate": 1.0870173798582284e-06, + "loss": 0.0162, + "step": 3772 + }, + { + "epoch": 2.4232498394348108, + "grad_norm": 0.8901331303379699, + "learning_rate": 1.0846918810328188e-06, + "loss": 0.0221, + "step": 3773 + }, + { + "epoch": 2.423892100192678, + "grad_norm": 0.5142656896077802, + "learning_rate": 1.0823685697015085e-06, + "loss": 0.0284, + "step": 3774 + }, + { + "epoch": 2.424534360950546, + "grad_norm": 0.8588265512802157, + "learning_rate": 1.080047447162338e-06, + "loss": 0.0254, + "step": 3775 + }, + { + "epoch": 2.4251766217084136, + "grad_norm": 0.664803647966512, + "learning_rate": 1.0777285147121247e-06, + "loss": 0.0214, + "step": 3776 + }, + { + "epoch": 2.4258188824662814, + "grad_norm": 0.8270614215536595, + "learning_rate": 1.0754117736464637e-06, + "loss": 0.0327, + "step": 3777 + }, + { + "epoch": 2.426461143224149, + "grad_norm": 0.7015726112651675, + "learning_rate": 1.073097225259725e-06, + "loss": 0.0208, + "step": 3778 + }, + { + "epoch": 2.4271034039820165, + "grad_norm": 0.5392505522683174, + "learning_rate": 1.070784870845053e-06, + "loss": 0.0165, + "step": 3779 + }, + { + "epoch": 2.4277456647398843, + "grad_norm": 1.1336013655772106, + "learning_rate": 1.0684747116943683e-06, + "loss": 0.0367, + "step": 3780 + }, + { + "epoch": 2.428387925497752, + "grad_norm": 0.4484351309106021, + "learning_rate": 1.0661667490983613e-06, + "loss": 0.0132, + "step": 3781 + }, + { + "epoch": 2.42903018625562, + "grad_norm": 0.9121689177314648, + "learning_rate": 1.0638609843464988e-06, + "loss": 0.0323, + "step": 3782 + }, + { + "epoch": 2.4296724470134876, + "grad_norm": 0.5640373589401946, + "learning_rate": 1.0615574187270167e-06, + "loss": 0.0176, + "step": 3783 + }, + { + "epoch": 2.4303147077713554, + "grad_norm": 0.9821140897906183, + "learning_rate": 1.0592560535269264e-06, + "loss": 0.0496, + "step": 3784 + }, + { + "epoch": 2.4309569685292227, + "grad_norm": 0.7309888628681344, + "learning_rate": 1.0569568900320065e-06, + "loss": 0.0233, + "step": 3785 + }, + { + "epoch": 2.4315992292870905, + "grad_norm": 0.7000364760322945, + "learning_rate": 1.0546599295268063e-06, + "loss": 0.0239, + "step": 3786 + }, + { + "epoch": 2.4322414900449583, + "grad_norm": 1.0856358609495795, + "learning_rate": 1.0523651732946439e-06, + "loss": 0.0545, + "step": 3787 + }, + { + "epoch": 2.432883750802826, + "grad_norm": 0.9717917044398995, + "learning_rate": 1.0500726226176078e-06, + "loss": 0.0409, + "step": 3788 + }, + { + "epoch": 2.433526011560694, + "grad_norm": 0.6760139626924317, + "learning_rate": 1.04778227877655e-06, + "loss": 0.0283, + "step": 3789 + }, + { + "epoch": 2.434168272318561, + "grad_norm": 0.9392693110616426, + "learning_rate": 1.0454941430510945e-06, + "loss": 0.0489, + "step": 3790 + }, + { + "epoch": 2.434810533076429, + "grad_norm": 0.9220283368875798, + "learning_rate": 1.0432082167196284e-06, + "loss": 0.0432, + "step": 3791 + }, + { + "epoch": 2.4354527938342967, + "grad_norm": 0.7067319626266474, + "learning_rate": 1.0409245010593055e-06, + "loss": 0.0222, + "step": 3792 + }, + { + "epoch": 2.4360950545921645, + "grad_norm": 1.093412192775075, + "learning_rate": 1.0386429973460455e-06, + "loss": 0.0626, + "step": 3793 + }, + { + "epoch": 2.4367373153500322, + "grad_norm": 0.8416694529528281, + "learning_rate": 1.0363637068545296e-06, + "loss": 0.0329, + "step": 3794 + }, + { + "epoch": 2.4373795761079, + "grad_norm": 1.1340312755653281, + "learning_rate": 1.0340866308582053e-06, + "loss": 0.0298, + "step": 3795 + }, + { + "epoch": 2.4380218368657673, + "grad_norm": 1.183548840738679, + "learning_rate": 1.0318117706292813e-06, + "loss": 0.0521, + "step": 3796 + }, + { + "epoch": 2.438664097623635, + "grad_norm": 0.9960187385715135, + "learning_rate": 1.0295391274387283e-06, + "loss": 0.0523, + "step": 3797 + }, + { + "epoch": 2.439306358381503, + "grad_norm": 0.8031156727241425, + "learning_rate": 1.0272687025562794e-06, + "loss": 0.0296, + "step": 3798 + }, + { + "epoch": 2.4399486191393707, + "grad_norm": 0.5896415226851589, + "learning_rate": 1.025000497250427e-06, + "loss": 0.0112, + "step": 3799 + }, + { + "epoch": 2.4405908798972384, + "grad_norm": 0.8643945834953848, + "learning_rate": 1.0227345127884242e-06, + "loss": 0.0228, + "step": 3800 + }, + { + "epoch": 2.4412331406551058, + "grad_norm": 0.946051229473567, + "learning_rate": 1.0204707504362825e-06, + "loss": 0.0467, + "step": 3801 + }, + { + "epoch": 2.4418754014129735, + "grad_norm": 0.7811756798556294, + "learning_rate": 1.0182092114587743e-06, + "loss": 0.0293, + "step": 3802 + }, + { + "epoch": 2.4425176621708413, + "grad_norm": 0.9396045834124178, + "learning_rate": 1.0159498971194276e-06, + "loss": 0.0498, + "step": 3803 + }, + { + "epoch": 2.443159922928709, + "grad_norm": 0.7226024370596802, + "learning_rate": 1.0136928086805265e-06, + "loss": 0.0284, + "step": 3804 + }, + { + "epoch": 2.443802183686577, + "grad_norm": 0.8326028656775725, + "learning_rate": 1.0114379474031122e-06, + "loss": 0.038, + "step": 3805 + }, + { + "epoch": 2.4444444444444446, + "grad_norm": 0.7076404790059722, + "learning_rate": 1.009185314546985e-06, + "loss": 0.0373, + "step": 3806 + }, + { + "epoch": 2.445086705202312, + "grad_norm": 0.8559765854810828, + "learning_rate": 1.006934911370696e-06, + "loss": 0.0288, + "step": 3807 + }, + { + "epoch": 2.4457289659601797, + "grad_norm": 0.8743091047486535, + "learning_rate": 1.0046867391315529e-06, + "loss": 0.0358, + "step": 3808 + }, + { + "epoch": 2.4463712267180475, + "grad_norm": 0.5839182275440208, + "learning_rate": 1.002440799085615e-06, + "loss": 0.0176, + "step": 3809 + }, + { + "epoch": 2.4470134874759153, + "grad_norm": 0.9717883891827093, + "learning_rate": 1.0001970924876964e-06, + "loss": 0.039, + "step": 3810 + }, + { + "epoch": 2.447655748233783, + "grad_norm": 0.5857764559140741, + "learning_rate": 9.979556205913632e-07, + "loss": 0.0175, + "step": 3811 + }, + { + "epoch": 2.4482980089916504, + "grad_norm": 0.7520427713614131, + "learning_rate": 9.957163846489303e-07, + "loss": 0.0211, + "step": 3812 + }, + { + "epoch": 2.448940269749518, + "grad_norm": 0.9136885950092879, + "learning_rate": 9.934793859114667e-07, + "loss": 0.0259, + "step": 3813 + }, + { + "epoch": 2.449582530507386, + "grad_norm": 1.194153768687843, + "learning_rate": 9.912446256287895e-07, + "loss": 0.0772, + "step": 3814 + }, + { + "epoch": 2.4502247912652537, + "grad_norm": 0.46872408323253684, + "learning_rate": 9.89012105049466e-07, + "loss": 0.0122, + "step": 3815 + }, + { + "epoch": 2.4508670520231215, + "grad_norm": 1.045528724759226, + "learning_rate": 9.867818254208122e-07, + "loss": 0.0315, + "step": 3816 + }, + { + "epoch": 2.4515093127809893, + "grad_norm": 0.7655231740849958, + "learning_rate": 9.845537879888905e-07, + "loss": 0.0298, + "step": 3817 + }, + { + "epoch": 2.4521515735388566, + "grad_norm": 1.0228360065576474, + "learning_rate": 9.823279939985147e-07, + "loss": 0.0546, + "step": 3818 + }, + { + "epoch": 2.4527938342967244, + "grad_norm": 0.6274285239961911, + "learning_rate": 9.801044446932418e-07, + "loss": 0.0326, + "step": 3819 + }, + { + "epoch": 2.453436095054592, + "grad_norm": 0.9043124501219193, + "learning_rate": 9.778831413153734e-07, + "loss": 0.0304, + "step": 3820 + }, + { + "epoch": 2.45407835581246, + "grad_norm": 0.9504428231228633, + "learning_rate": 9.756640851059595e-07, + "loss": 0.0439, + "step": 3821 + }, + { + "epoch": 2.4547206165703277, + "grad_norm": 1.0059762272419692, + "learning_rate": 9.734472773047927e-07, + "loss": 0.0407, + "step": 3822 + }, + { + "epoch": 2.455362877328195, + "grad_norm": 0.506788071442149, + "learning_rate": 9.7123271915041e-07, + "loss": 0.0174, + "step": 3823 + }, + { + "epoch": 2.456005138086063, + "grad_norm": 0.9553661246354959, + "learning_rate": 9.690204118800922e-07, + "loss": 0.0497, + "step": 3824 + }, + { + "epoch": 2.4566473988439306, + "grad_norm": 0.5646383387681341, + "learning_rate": 9.668103567298615e-07, + "loss": 0.0163, + "step": 3825 + }, + { + "epoch": 2.4572896596017983, + "grad_norm": 0.5961663911210201, + "learning_rate": 9.646025549344823e-07, + "loss": 0.0181, + "step": 3826 + }, + { + "epoch": 2.457931920359666, + "grad_norm": 0.9187781193255812, + "learning_rate": 9.623970077274598e-07, + "loss": 0.032, + "step": 3827 + }, + { + "epoch": 2.458574181117534, + "grad_norm": 1.7717053174279174, + "learning_rate": 9.601937163410392e-07, + "loss": 0.0994, + "step": 3828 + }, + { + "epoch": 2.4592164418754012, + "grad_norm": 0.7744644933926513, + "learning_rate": 9.579926820062063e-07, + "loss": 0.0231, + "step": 3829 + }, + { + "epoch": 2.459858702633269, + "grad_norm": 0.6069331571251056, + "learning_rate": 9.557939059526862e-07, + "loss": 0.0221, + "step": 3830 + }, + { + "epoch": 2.4605009633911368, + "grad_norm": 1.0213037188330132, + "learning_rate": 9.535973894089401e-07, + "loss": 0.0196, + "step": 3831 + }, + { + "epoch": 2.4611432241490045, + "grad_norm": 0.6830748589761737, + "learning_rate": 9.514031336021684e-07, + "loss": 0.0164, + "step": 3832 + }, + { + "epoch": 2.4617854849068723, + "grad_norm": 0.9028433658489811, + "learning_rate": 9.492111397583092e-07, + "loss": 0.0419, + "step": 3833 + }, + { + "epoch": 2.4624277456647397, + "grad_norm": 0.5588895108085611, + "learning_rate": 9.470214091020358e-07, + "loss": 0.02, + "step": 3834 + }, + { + "epoch": 2.4630700064226074, + "grad_norm": 0.5654916471297732, + "learning_rate": 9.448339428567571e-07, + "loss": 0.0245, + "step": 3835 + }, + { + "epoch": 2.463712267180475, + "grad_norm": 1.0422967376766172, + "learning_rate": 9.426487422446157e-07, + "loss": 0.0621, + "step": 3836 + }, + { + "epoch": 2.464354527938343, + "grad_norm": 0.9672970542083411, + "learning_rate": 9.404658084864904e-07, + "loss": 0.0424, + "step": 3837 + }, + { + "epoch": 2.4649967886962107, + "grad_norm": 0.5905306458802004, + "learning_rate": 9.382851428019924e-07, + "loss": 0.0195, + "step": 3838 + }, + { + "epoch": 2.4656390494540785, + "grad_norm": 0.726002314915197, + "learning_rate": 9.361067464094653e-07, + "loss": 0.0297, + "step": 3839 + }, + { + "epoch": 2.4662813102119463, + "grad_norm": 0.6203681222982708, + "learning_rate": 9.339306205259874e-07, + "loss": 0.0188, + "step": 3840 + }, + { + "epoch": 2.4669235709698136, + "grad_norm": 0.6513493630453163, + "learning_rate": 9.317567663673655e-07, + "loss": 0.0138, + "step": 3841 + }, + { + "epoch": 2.4675658317276814, + "grad_norm": 0.844310513085244, + "learning_rate": 9.295851851481385e-07, + "loss": 0.021, + "step": 3842 + }, + { + "epoch": 2.468208092485549, + "grad_norm": 0.7463458895080431, + "learning_rate": 9.274158780815767e-07, + "loss": 0.0244, + "step": 3843 + }, + { + "epoch": 2.468850353243417, + "grad_norm": 0.9409480422138495, + "learning_rate": 9.252488463796749e-07, + "loss": 0.0311, + "step": 3844 + }, + { + "epoch": 2.4694926140012843, + "grad_norm": 0.5510175936618781, + "learning_rate": 9.230840912531625e-07, + "loss": 0.0183, + "step": 3845 + }, + { + "epoch": 2.470134874759152, + "grad_norm": 0.7858689751147481, + "learning_rate": 9.209216139114935e-07, + "loss": 0.0409, + "step": 3846 + }, + { + "epoch": 2.47077713551702, + "grad_norm": 0.6525976856543766, + "learning_rate": 9.187614155628504e-07, + "loss": 0.0254, + "step": 3847 + }, + { + "epoch": 2.4714193962748876, + "grad_norm": 1.0942661483544132, + "learning_rate": 9.166034974141435e-07, + "loss": 0.052, + "step": 3848 + }, + { + "epoch": 2.4720616570327554, + "grad_norm": 0.9204909762475241, + "learning_rate": 9.144478606710066e-07, + "loss": 0.041, + "step": 3849 + }, + { + "epoch": 2.472703917790623, + "grad_norm": 0.8720007817657149, + "learning_rate": 9.122945065378008e-07, + "loss": 0.0448, + "step": 3850 + }, + { + "epoch": 2.473346178548491, + "grad_norm": 0.47919143257609054, + "learning_rate": 9.101434362176109e-07, + "loss": 0.0079, + "step": 3851 + }, + { + "epoch": 2.4739884393063583, + "grad_norm": 0.8470758209516683, + "learning_rate": 9.079946509122473e-07, + "loss": 0.021, + "step": 3852 + }, + { + "epoch": 2.474630700064226, + "grad_norm": 0.9451993625312474, + "learning_rate": 9.058481518222417e-07, + "loss": 0.0535, + "step": 3853 + }, + { + "epoch": 2.475272960822094, + "grad_norm": 0.8299364311209919, + "learning_rate": 9.037039401468495e-07, + "loss": 0.0339, + "step": 3854 + }, + { + "epoch": 2.4759152215799616, + "grad_norm": 0.7223073664396453, + "learning_rate": 9.015620170840484e-07, + "loss": 0.0243, + "step": 3855 + }, + { + "epoch": 2.4765574823378294, + "grad_norm": 0.7788323435184351, + "learning_rate": 8.994223838305366e-07, + "loss": 0.0347, + "step": 3856 + }, + { + "epoch": 2.4771997430956967, + "grad_norm": 0.5724023965366805, + "learning_rate": 8.972850415817341e-07, + "loss": 0.023, + "step": 3857 + }, + { + "epoch": 2.4778420038535645, + "grad_norm": 0.8159799442642927, + "learning_rate": 8.951499915317807e-07, + "loss": 0.0352, + "step": 3858 + }, + { + "epoch": 2.4784842646114322, + "grad_norm": 0.5148719178971775, + "learning_rate": 8.930172348735339e-07, + "loss": 0.0166, + "step": 3859 + }, + { + "epoch": 2.4791265253693, + "grad_norm": 1.0121526639097123, + "learning_rate": 8.908867727985693e-07, + "loss": 0.055, + "step": 3860 + }, + { + "epoch": 2.479768786127168, + "grad_norm": 0.7515812165462714, + "learning_rate": 8.887586064971859e-07, + "loss": 0.035, + "step": 3861 + }, + { + "epoch": 2.4804110468850356, + "grad_norm": 0.9088111095021681, + "learning_rate": 8.866327371583954e-07, + "loss": 0.0313, + "step": 3862 + }, + { + "epoch": 2.481053307642903, + "grad_norm": 0.9097463097383401, + "learning_rate": 8.845091659699268e-07, + "loss": 0.0608, + "step": 3863 + }, + { + "epoch": 2.4816955684007707, + "grad_norm": 0.7401983458992702, + "learning_rate": 8.823878941182251e-07, + "loss": 0.0212, + "step": 3864 + }, + { + "epoch": 2.4823378291586384, + "grad_norm": 1.0029898372392902, + "learning_rate": 8.80268922788452e-07, + "loss": 0.044, + "step": 3865 + }, + { + "epoch": 2.482980089916506, + "grad_norm": 0.6999475767833689, + "learning_rate": 8.781522531644837e-07, + "loss": 0.0246, + "step": 3866 + }, + { + "epoch": 2.483622350674374, + "grad_norm": 0.6305071138330283, + "learning_rate": 8.760378864289076e-07, + "loss": 0.0222, + "step": 3867 + }, + { + "epoch": 2.4842646114322413, + "grad_norm": 0.8783900098886019, + "learning_rate": 8.739258237630272e-07, + "loss": 0.0331, + "step": 3868 + }, + { + "epoch": 2.484906872190109, + "grad_norm": 0.6946252424318528, + "learning_rate": 8.718160663468583e-07, + "loss": 0.0213, + "step": 3869 + }, + { + "epoch": 2.485549132947977, + "grad_norm": 0.556658915584491, + "learning_rate": 8.697086153591289e-07, + "loss": 0.02, + "step": 3870 + }, + { + "epoch": 2.4861913937058446, + "grad_norm": 1.0125561845789843, + "learning_rate": 8.676034719772774e-07, + "loss": 0.0362, + "step": 3871 + }, + { + "epoch": 2.4868336544637124, + "grad_norm": 0.605614881951648, + "learning_rate": 8.655006373774532e-07, + "loss": 0.0271, + "step": 3872 + }, + { + "epoch": 2.48747591522158, + "grad_norm": 0.6908940581036448, + "learning_rate": 8.634001127345182e-07, + "loss": 0.0295, + "step": 3873 + }, + { + "epoch": 2.4881181759794475, + "grad_norm": 0.9632358184091603, + "learning_rate": 8.613018992220412e-07, + "loss": 0.047, + "step": 3874 + }, + { + "epoch": 2.4887604367373153, + "grad_norm": 0.9309911264186819, + "learning_rate": 8.592059980122985e-07, + "loss": 0.0452, + "step": 3875 + }, + { + "epoch": 2.489402697495183, + "grad_norm": 0.8162379716241678, + "learning_rate": 8.571124102762768e-07, + "loss": 0.035, + "step": 3876 + }, + { + "epoch": 2.490044958253051, + "grad_norm": 0.785030596682132, + "learning_rate": 8.550211371836703e-07, + "loss": 0.0281, + "step": 3877 + }, + { + "epoch": 2.4906872190109186, + "grad_norm": 0.47868951387881964, + "learning_rate": 8.52932179902879e-07, + "loss": 0.0098, + "step": 3878 + }, + { + "epoch": 2.491329479768786, + "grad_norm": 0.7210615878136903, + "learning_rate": 8.508455396010096e-07, + "loss": 0.0306, + "step": 3879 + }, + { + "epoch": 2.4919717405266537, + "grad_norm": 1.1214901295817727, + "learning_rate": 8.487612174438742e-07, + "loss": 0.0697, + "step": 3880 + }, + { + "epoch": 2.4926140012845215, + "grad_norm": 0.8648072916994604, + "learning_rate": 8.46679214595989e-07, + "loss": 0.022, + "step": 3881 + }, + { + "epoch": 2.4932562620423893, + "grad_norm": 1.0950654363623828, + "learning_rate": 8.445995322205763e-07, + "loss": 0.0347, + "step": 3882 + }, + { + "epoch": 2.493898522800257, + "grad_norm": 0.7222853097345007, + "learning_rate": 8.4252217147956e-07, + "loss": 0.0302, + "step": 3883 + }, + { + "epoch": 2.494540783558125, + "grad_norm": 0.696413326244712, + "learning_rate": 8.404471335335673e-07, + "loss": 0.0196, + "step": 3884 + }, + { + "epoch": 2.495183044315992, + "grad_norm": 0.5438978765136245, + "learning_rate": 8.383744195419286e-07, + "loss": 0.0265, + "step": 3885 + }, + { + "epoch": 2.49582530507386, + "grad_norm": 0.8516885419135416, + "learning_rate": 8.363040306626752e-07, + "loss": 0.0309, + "step": 3886 + }, + { + "epoch": 2.4964675658317277, + "grad_norm": 1.165624141454188, + "learning_rate": 8.342359680525391e-07, + "loss": 0.0601, + "step": 3887 + }, + { + "epoch": 2.4971098265895955, + "grad_norm": 0.6890979071168036, + "learning_rate": 8.321702328669534e-07, + "loss": 0.0252, + "step": 3888 + }, + { + "epoch": 2.4977520873474632, + "grad_norm": 0.7333056015839496, + "learning_rate": 8.301068262600498e-07, + "loss": 0.0322, + "step": 3889 + }, + { + "epoch": 2.4983943481053306, + "grad_norm": 0.9220429652107489, + "learning_rate": 8.280457493846617e-07, + "loss": 0.0393, + "step": 3890 + }, + { + "epoch": 2.4990366088631983, + "grad_norm": 0.6510156693512569, + "learning_rate": 8.25987003392315e-07, + "loss": 0.0292, + "step": 3891 + }, + { + "epoch": 2.499678869621066, + "grad_norm": 0.8613155644996855, + "learning_rate": 8.239305894332394e-07, + "loss": 0.0486, + "step": 3892 + }, + { + "epoch": 2.500321130378934, + "grad_norm": 0.7519692869460528, + "learning_rate": 8.218765086563591e-07, + "loss": 0.0274, + "step": 3893 + }, + { + "epoch": 2.5009633911368017, + "grad_norm": 0.8075310031022273, + "learning_rate": 8.198247622092942e-07, + "loss": 0.0229, + "step": 3894 + }, + { + "epoch": 2.5016056518946694, + "grad_norm": 0.8702808381093818, + "learning_rate": 8.177753512383629e-07, + "loss": 0.0219, + "step": 3895 + }, + { + "epoch": 2.5022479126525368, + "grad_norm": 0.5960304734511156, + "learning_rate": 8.157282768885766e-07, + "loss": 0.0236, + "step": 3896 + }, + { + "epoch": 2.5028901734104045, + "grad_norm": 0.6439200421236088, + "learning_rate": 8.136835403036413e-07, + "loss": 0.0169, + "step": 3897 + }, + { + "epoch": 2.5035324341682723, + "grad_norm": 0.7475615754275176, + "learning_rate": 8.116411426259596e-07, + "loss": 0.03, + "step": 3898 + }, + { + "epoch": 2.50417469492614, + "grad_norm": 0.7368351552759251, + "learning_rate": 8.096010849966213e-07, + "loss": 0.0197, + "step": 3899 + }, + { + "epoch": 2.504816955684008, + "grad_norm": 0.8385116560706322, + "learning_rate": 8.075633685554146e-07, + "loss": 0.0355, + "step": 3900 + }, + { + "epoch": 2.505459216441875, + "grad_norm": 0.6851332054307185, + "learning_rate": 8.055279944408173e-07, + "loss": 0.0174, + "step": 3901 + }, + { + "epoch": 2.506101477199743, + "grad_norm": 0.8207366943839076, + "learning_rate": 8.034949637899986e-07, + "loss": 0.0296, + "step": 3902 + }, + { + "epoch": 2.5067437379576107, + "grad_norm": 0.660483358166476, + "learning_rate": 8.014642777388193e-07, + "loss": 0.0368, + "step": 3903 + }, + { + "epoch": 2.5073859987154785, + "grad_norm": 0.8164591151781314, + "learning_rate": 7.994359374218297e-07, + "loss": 0.0393, + "step": 3904 + }, + { + "epoch": 2.5080282594733463, + "grad_norm": 0.8700790319439985, + "learning_rate": 7.974099439722682e-07, + "loss": 0.0267, + "step": 3905 + }, + { + "epoch": 2.508670520231214, + "grad_norm": 0.9041347972674316, + "learning_rate": 7.95386298522065e-07, + "loss": 0.0482, + "step": 3906 + }, + { + "epoch": 2.509312780989082, + "grad_norm": 0.36072368874498023, + "learning_rate": 7.933650022018358e-07, + "loss": 0.0102, + "step": 3907 + }, + { + "epoch": 2.509955041746949, + "grad_norm": 0.6651600562541381, + "learning_rate": 7.913460561408848e-07, + "loss": 0.0316, + "step": 3908 + }, + { + "epoch": 2.510597302504817, + "grad_norm": 0.9098888650236516, + "learning_rate": 7.893294614672026e-07, + "loss": 0.0334, + "step": 3909 + }, + { + "epoch": 2.5112395632626847, + "grad_norm": 0.9286496252565198, + "learning_rate": 7.873152193074679e-07, + "loss": 0.0316, + "step": 3910 + }, + { + "epoch": 2.5118818240205525, + "grad_norm": 0.8005757261045404, + "learning_rate": 7.853033307870422e-07, + "loss": 0.0414, + "step": 3911 + }, + { + "epoch": 2.51252408477842, + "grad_norm": 0.7619384545912258, + "learning_rate": 7.832937970299747e-07, + "loss": 0.0251, + "step": 3912 + }, + { + "epoch": 2.5131663455362876, + "grad_norm": 0.6258673797578443, + "learning_rate": 7.812866191589985e-07, + "loss": 0.0174, + "step": 3913 + }, + { + "epoch": 2.5138086062941554, + "grad_norm": 0.6911949635468784, + "learning_rate": 7.79281798295527e-07, + "loss": 0.0307, + "step": 3914 + }, + { + "epoch": 2.514450867052023, + "grad_norm": 0.9314555553441952, + "learning_rate": 7.772793355596597e-07, + "loss": 0.0361, + "step": 3915 + }, + { + "epoch": 2.515093127809891, + "grad_norm": 0.7245145363178029, + "learning_rate": 7.752792320701813e-07, + "loss": 0.0238, + "step": 3916 + }, + { + "epoch": 2.5157353885677587, + "grad_norm": 0.6773340924267017, + "learning_rate": 7.732814889445539e-07, + "loss": 0.0243, + "step": 3917 + }, + { + "epoch": 2.5163776493256265, + "grad_norm": 0.9769265417823313, + "learning_rate": 7.71286107298922e-07, + "loss": 0.035, + "step": 3918 + }, + { + "epoch": 2.517019910083494, + "grad_norm": 0.5452385755690551, + "learning_rate": 7.692930882481114e-07, + "loss": 0.0237, + "step": 3919 + }, + { + "epoch": 2.5176621708413616, + "grad_norm": 0.6981524808684901, + "learning_rate": 7.673024329056283e-07, + "loss": 0.0265, + "step": 3920 + }, + { + "epoch": 2.5183044315992293, + "grad_norm": 1.1712914344871752, + "learning_rate": 7.653141423836585e-07, + "loss": 0.0367, + "step": 3921 + }, + { + "epoch": 2.518946692357097, + "grad_norm": 1.1212789256693503, + "learning_rate": 7.633282177930629e-07, + "loss": 0.0559, + "step": 3922 + }, + { + "epoch": 2.5195889531149644, + "grad_norm": 0.7250576409594239, + "learning_rate": 7.613446602433849e-07, + "loss": 0.0388, + "step": 3923 + }, + { + "epoch": 2.520231213872832, + "grad_norm": 1.1240647691412329, + "learning_rate": 7.593634708428438e-07, + "loss": 0.0589, + "step": 3924 + }, + { + "epoch": 2.5208734746307, + "grad_norm": 0.504800511216996, + "learning_rate": 7.573846506983363e-07, + "loss": 0.0164, + "step": 3925 + }, + { + "epoch": 2.5215157353885678, + "grad_norm": 0.7362999148928022, + "learning_rate": 7.554082009154346e-07, + "loss": 0.0186, + "step": 3926 + }, + { + "epoch": 2.5221579961464355, + "grad_norm": 0.786100768311548, + "learning_rate": 7.534341225983866e-07, + "loss": 0.0403, + "step": 3927 + }, + { + "epoch": 2.5228002569043033, + "grad_norm": 1.083256373028466, + "learning_rate": 7.514624168501156e-07, + "loss": 0.0624, + "step": 3928 + }, + { + "epoch": 2.523442517662171, + "grad_norm": 0.8684936176814833, + "learning_rate": 7.49493084772222e-07, + "loss": 0.0358, + "step": 3929 + }, + { + "epoch": 2.5240847784200384, + "grad_norm": 1.1667958062325294, + "learning_rate": 7.475261274649742e-07, + "loss": 0.0604, + "step": 3930 + }, + { + "epoch": 2.524727039177906, + "grad_norm": 0.7381943738732123, + "learning_rate": 7.45561546027318e-07, + "loss": 0.037, + "step": 3931 + }, + { + "epoch": 2.525369299935774, + "grad_norm": 0.6575403059324405, + "learning_rate": 7.43599341556871e-07, + "loss": 0.0203, + "step": 3932 + }, + { + "epoch": 2.5260115606936417, + "grad_norm": 1.2835627184772833, + "learning_rate": 7.416395151499223e-07, + "loss": 0.0295, + "step": 3933 + }, + { + "epoch": 2.526653821451509, + "grad_norm": 0.5046921831218336, + "learning_rate": 7.396820679014327e-07, + "loss": 0.0119, + "step": 3934 + }, + { + "epoch": 2.527296082209377, + "grad_norm": 1.2040041434478082, + "learning_rate": 7.377270009050341e-07, + "loss": 0.0559, + "step": 3935 + }, + { + "epoch": 2.5279383429672446, + "grad_norm": 0.8591015402835814, + "learning_rate": 7.357743152530272e-07, + "loss": 0.0435, + "step": 3936 + }, + { + "epoch": 2.5285806037251124, + "grad_norm": 0.6972749030577613, + "learning_rate": 7.338240120363849e-07, + "loss": 0.025, + "step": 3937 + }, + { + "epoch": 2.52922286448298, + "grad_norm": 0.7341480507510137, + "learning_rate": 7.318760923447454e-07, + "loss": 0.0306, + "step": 3938 + }, + { + "epoch": 2.529865125240848, + "grad_norm": 0.48497377415162546, + "learning_rate": 7.299305572664184e-07, + "loss": 0.0139, + "step": 3939 + }, + { + "epoch": 2.5305073859987157, + "grad_norm": 1.1162718961036613, + "learning_rate": 7.279874078883792e-07, + "loss": 0.0665, + "step": 3940 + }, + { + "epoch": 2.531149646756583, + "grad_norm": 0.9900141806972275, + "learning_rate": 7.260466452962717e-07, + "loss": 0.06, + "step": 3941 + }, + { + "epoch": 2.531791907514451, + "grad_norm": 0.8632193067462417, + "learning_rate": 7.241082705744057e-07, + "loss": 0.0379, + "step": 3942 + }, + { + "epoch": 2.5324341682723186, + "grad_norm": 0.7463653608613952, + "learning_rate": 7.221722848057561e-07, + "loss": 0.0306, + "step": 3943 + }, + { + "epoch": 2.5330764290301864, + "grad_norm": 0.7786418613767467, + "learning_rate": 7.202386890719648e-07, + "loss": 0.0234, + "step": 3944 + }, + { + "epoch": 2.5337186897880537, + "grad_norm": 0.6223017503799365, + "learning_rate": 7.183074844533372e-07, + "loss": 0.0237, + "step": 3945 + }, + { + "epoch": 2.5343609505459215, + "grad_norm": 0.7294138272889396, + "learning_rate": 7.163786720288424e-07, + "loss": 0.0318, + "step": 3946 + }, + { + "epoch": 2.5350032113037893, + "grad_norm": 0.6064305865838318, + "learning_rate": 7.144522528761128e-07, + "loss": 0.0128, + "step": 3947 + }, + { + "epoch": 2.535645472061657, + "grad_norm": 0.8131817903205066, + "learning_rate": 7.125282280714457e-07, + "loss": 0.031, + "step": 3948 + }, + { + "epoch": 2.536287732819525, + "grad_norm": 0.9702214985450909, + "learning_rate": 7.10606598689798e-07, + "loss": 0.0471, + "step": 3949 + }, + { + "epoch": 2.5369299935773926, + "grad_norm": 0.8763122775744493, + "learning_rate": 7.086873658047922e-07, + "loss": 0.0433, + "step": 3950 + }, + { + "epoch": 2.5375722543352603, + "grad_norm": 0.7831423821767484, + "learning_rate": 7.067705304887074e-07, + "loss": 0.0303, + "step": 3951 + }, + { + "epoch": 2.5382145150931277, + "grad_norm": 0.8347143516075097, + "learning_rate": 7.048560938124854e-07, + "loss": 0.0416, + "step": 3952 + }, + { + "epoch": 2.5388567758509955, + "grad_norm": 0.694911927957093, + "learning_rate": 7.029440568457291e-07, + "loss": 0.0208, + "step": 3953 + }, + { + "epoch": 2.5394990366088632, + "grad_norm": 0.42117205216329084, + "learning_rate": 7.010344206566966e-07, + "loss": 0.0151, + "step": 3954 + }, + { + "epoch": 2.540141297366731, + "grad_norm": 0.6675625071464978, + "learning_rate": 6.991271863123084e-07, + "loss": 0.0276, + "step": 3955 + }, + { + "epoch": 2.5407835581245983, + "grad_norm": 0.7395394240759134, + "learning_rate": 6.972223548781415e-07, + "loss": 0.0229, + "step": 3956 + }, + { + "epoch": 2.541425818882466, + "grad_norm": 0.9891620431689161, + "learning_rate": 6.953199274184313e-07, + "loss": 0.0405, + "step": 3957 + }, + { + "epoch": 2.542068079640334, + "grad_norm": 0.5251886596662068, + "learning_rate": 6.9341990499607e-07, + "loss": 0.0145, + "step": 3958 + }, + { + "epoch": 2.5427103403982017, + "grad_norm": 1.0182930159350174, + "learning_rate": 6.915222886726048e-07, + "loss": 0.0468, + "step": 3959 + }, + { + "epoch": 2.5433526011560694, + "grad_norm": 0.8174265934118023, + "learning_rate": 6.896270795082394e-07, + "loss": 0.0363, + "step": 3960 + }, + { + "epoch": 2.543994861913937, + "grad_norm": 0.761666875284188, + "learning_rate": 6.877342785618335e-07, + "loss": 0.0347, + "step": 3961 + }, + { + "epoch": 2.544637122671805, + "grad_norm": 0.8945877666694587, + "learning_rate": 6.858438868909007e-07, + "loss": 0.0399, + "step": 3962 + }, + { + "epoch": 2.5452793834296723, + "grad_norm": 0.40757616303592736, + "learning_rate": 6.839559055516071e-07, + "loss": 0.0109, + "step": 3963 + }, + { + "epoch": 2.54592164418754, + "grad_norm": 0.8069410441408976, + "learning_rate": 6.82070335598774e-07, + "loss": 0.0342, + "step": 3964 + }, + { + "epoch": 2.546563904945408, + "grad_norm": 0.7365643525686467, + "learning_rate": 6.801871780858754e-07, + "loss": 0.0249, + "step": 3965 + }, + { + "epoch": 2.5472061657032756, + "grad_norm": 0.875567393433182, + "learning_rate": 6.783064340650363e-07, + "loss": 0.043, + "step": 3966 + }, + { + "epoch": 2.547848426461143, + "grad_norm": 0.9129010144140733, + "learning_rate": 6.764281045870342e-07, + "loss": 0.0309, + "step": 3967 + }, + { + "epoch": 2.5484906872190107, + "grad_norm": 0.6584586654489506, + "learning_rate": 6.74552190701298e-07, + "loss": 0.024, + "step": 3968 + }, + { + "epoch": 2.5491329479768785, + "grad_norm": 0.9503967950365911, + "learning_rate": 6.726786934559048e-07, + "loss": 0.0544, + "step": 3969 + }, + { + "epoch": 2.5497752087347463, + "grad_norm": 0.6874899034558323, + "learning_rate": 6.708076138975822e-07, + "loss": 0.0295, + "step": 3970 + }, + { + "epoch": 2.550417469492614, + "grad_norm": 0.6875172600059066, + "learning_rate": 6.689389530717106e-07, + "loss": 0.0272, + "step": 3971 + }, + { + "epoch": 2.551059730250482, + "grad_norm": 0.73175631612725, + "learning_rate": 6.670727120223142e-07, + "loss": 0.026, + "step": 3972 + }, + { + "epoch": 2.5517019910083496, + "grad_norm": 0.8751293605650804, + "learning_rate": 6.652088917920685e-07, + "loss": 0.0472, + "step": 3973 + }, + { + "epoch": 2.552344251766217, + "grad_norm": 0.8289636790794851, + "learning_rate": 6.63347493422295e-07, + "loss": 0.0406, + "step": 3974 + }, + { + "epoch": 2.5529865125240847, + "grad_norm": 0.829587242590393, + "learning_rate": 6.614885179529617e-07, + "loss": 0.0327, + "step": 3975 + }, + { + "epoch": 2.5536287732819525, + "grad_norm": 0.7828383458778616, + "learning_rate": 6.596319664226858e-07, + "loss": 0.0362, + "step": 3976 + }, + { + "epoch": 2.5542710340398203, + "grad_norm": 0.5128329446543411, + "learning_rate": 6.577778398687251e-07, + "loss": 0.0112, + "step": 3977 + }, + { + "epoch": 2.5549132947976876, + "grad_norm": 0.3958831785184798, + "learning_rate": 6.559261393269872e-07, + "loss": 0.0071, + "step": 3978 + }, + { + "epoch": 2.5555555555555554, + "grad_norm": 0.8276826322250922, + "learning_rate": 6.540768658320218e-07, + "loss": 0.029, + "step": 3979 + }, + { + "epoch": 2.556197816313423, + "grad_norm": 0.7895896382149905, + "learning_rate": 6.522300204170245e-07, + "loss": 0.0347, + "step": 3980 + }, + { + "epoch": 2.556840077071291, + "grad_norm": 0.963418633202187, + "learning_rate": 6.503856041138324e-07, + "loss": 0.0355, + "step": 3981 + }, + { + "epoch": 2.5574823378291587, + "grad_norm": 1.0043567733472376, + "learning_rate": 6.485436179529264e-07, + "loss": 0.0461, + "step": 3982 + }, + { + "epoch": 2.5581245985870265, + "grad_norm": 0.6241040329427907, + "learning_rate": 6.467040629634286e-07, + "loss": 0.0191, + "step": 3983 + }, + { + "epoch": 2.5587668593448942, + "grad_norm": 0.8883831529067565, + "learning_rate": 6.448669401731067e-07, + "loss": 0.0483, + "step": 3984 + }, + { + "epoch": 2.5594091201027616, + "grad_norm": 0.8364992988614757, + "learning_rate": 6.430322506083642e-07, + "loss": 0.0294, + "step": 3985 + }, + { + "epoch": 2.5600513808606293, + "grad_norm": 0.7745054419692446, + "learning_rate": 6.411999952942477e-07, + "loss": 0.0294, + "step": 3986 + }, + { + "epoch": 2.560693641618497, + "grad_norm": 0.9259352715642293, + "learning_rate": 6.39370175254444e-07, + "loss": 0.0348, + "step": 3987 + }, + { + "epoch": 2.561335902376365, + "grad_norm": 1.180990436731236, + "learning_rate": 6.375427915112792e-07, + "loss": 0.05, + "step": 3988 + }, + { + "epoch": 2.5619781631342327, + "grad_norm": 0.7224275153434414, + "learning_rate": 6.357178450857171e-07, + "loss": 0.0238, + "step": 3989 + }, + { + "epoch": 2.5626204238921, + "grad_norm": 0.6617337601488024, + "learning_rate": 6.338953369973616e-07, + "loss": 0.0239, + "step": 3990 + }, + { + "epoch": 2.5632626846499678, + "grad_norm": 0.6909472268157614, + "learning_rate": 6.320752682644526e-07, + "loss": 0.0216, + "step": 3991 + }, + { + "epoch": 2.5639049454078355, + "grad_norm": 0.707024729215464, + "learning_rate": 6.302576399038679e-07, + "loss": 0.0217, + "step": 3992 + }, + { + "epoch": 2.5645472061657033, + "grad_norm": 0.699246519908476, + "learning_rate": 6.284424529311218e-07, + "loss": 0.0238, + "step": 3993 + }, + { + "epoch": 2.565189466923571, + "grad_norm": 0.5680197131831083, + "learning_rate": 6.26629708360365e-07, + "loss": 0.0167, + "step": 3994 + }, + { + "epoch": 2.565831727681439, + "grad_norm": 1.1091758220235055, + "learning_rate": 6.248194072043828e-07, + "loss": 0.0216, + "step": 3995 + }, + { + "epoch": 2.5664739884393066, + "grad_norm": 1.0552237470136627, + "learning_rate": 6.230115504745954e-07, + "loss": 0.042, + "step": 3996 + }, + { + "epoch": 2.567116249197174, + "grad_norm": 0.7638453197073097, + "learning_rate": 6.212061391810587e-07, + "loss": 0.0194, + "step": 3997 + }, + { + "epoch": 2.5677585099550417, + "grad_norm": 0.730180728533275, + "learning_rate": 6.194031743324602e-07, + "loss": 0.0185, + "step": 3998 + }, + { + "epoch": 2.5684007707129095, + "grad_norm": 1.0491173037840396, + "learning_rate": 6.176026569361221e-07, + "loss": 0.0376, + "step": 3999 + }, + { + "epoch": 2.5690430314707773, + "grad_norm": 0.7896232322968908, + "learning_rate": 6.158045879980002e-07, + "loss": 0.0226, + "step": 4000 + }, + { + "epoch": 2.5696852922286446, + "grad_norm": 0.8022582183309002, + "learning_rate": 6.14008968522678e-07, + "loss": 0.0272, + "step": 4001 + }, + { + "epoch": 2.5703275529865124, + "grad_norm": 0.8133121792786148, + "learning_rate": 6.122157995133749e-07, + "loss": 0.0274, + "step": 4002 + }, + { + "epoch": 2.57096981374438, + "grad_norm": 0.975454500662237, + "learning_rate": 6.104250819719393e-07, + "loss": 0.0369, + "step": 4003 + }, + { + "epoch": 2.571612074502248, + "grad_norm": 0.8437825481820735, + "learning_rate": 6.086368168988494e-07, + "loss": 0.0342, + "step": 4004 + }, + { + "epoch": 2.5722543352601157, + "grad_norm": 0.8351758694707462, + "learning_rate": 6.06851005293217e-07, + "loss": 0.0372, + "step": 4005 + }, + { + "epoch": 2.5728965960179835, + "grad_norm": 0.9219659020789877, + "learning_rate": 6.050676481527778e-07, + "loss": 0.04, + "step": 4006 + }, + { + "epoch": 2.5735388567758513, + "grad_norm": 1.0145302257803488, + "learning_rate": 6.032867464738995e-07, + "loss": 0.0376, + "step": 4007 + }, + { + "epoch": 2.5741811175337186, + "grad_norm": 0.6697514953150899, + "learning_rate": 6.015083012515782e-07, + "loss": 0.0295, + "step": 4008 + }, + { + "epoch": 2.5748233782915864, + "grad_norm": 0.7632596450193864, + "learning_rate": 5.997323134794342e-07, + "loss": 0.0195, + "step": 4009 + }, + { + "epoch": 2.575465639049454, + "grad_norm": 0.9084501428542671, + "learning_rate": 5.979587841497181e-07, + "loss": 0.0426, + "step": 4010 + }, + { + "epoch": 2.576107899807322, + "grad_norm": 0.7144165256645854, + "learning_rate": 5.961877142533056e-07, + "loss": 0.0207, + "step": 4011 + }, + { + "epoch": 2.5767501605651892, + "grad_norm": 0.7171728507175528, + "learning_rate": 5.944191047796987e-07, + "loss": 0.0238, + "step": 4012 + }, + { + "epoch": 2.577392421323057, + "grad_norm": 0.853430220525147, + "learning_rate": 5.926529567170259e-07, + "loss": 0.0442, + "step": 4013 + }, + { + "epoch": 2.578034682080925, + "grad_norm": 0.6239430176260484, + "learning_rate": 5.908892710520375e-07, + "loss": 0.021, + "step": 4014 + }, + { + "epoch": 2.5786769428387926, + "grad_norm": 0.9302238434346087, + "learning_rate": 5.891280487701106e-07, + "loss": 0.0406, + "step": 4015 + }, + { + "epoch": 2.5793192035966603, + "grad_norm": 1.0662725968446405, + "learning_rate": 5.873692908552458e-07, + "loss": 0.0637, + "step": 4016 + }, + { + "epoch": 2.579961464354528, + "grad_norm": 0.8684332813387721, + "learning_rate": 5.856129982900655e-07, + "loss": 0.0349, + "step": 4017 + }, + { + "epoch": 2.580603725112396, + "grad_norm": 0.8399732249427093, + "learning_rate": 5.838591720558156e-07, + "loss": 0.036, + "step": 4018 + }, + { + "epoch": 2.581245985870263, + "grad_norm": 0.825765867026004, + "learning_rate": 5.821078131323643e-07, + "loss": 0.0474, + "step": 4019 + }, + { + "epoch": 2.581888246628131, + "grad_norm": 0.863043696845772, + "learning_rate": 5.803589224982009e-07, + "loss": 0.0286, + "step": 4020 + }, + { + "epoch": 2.5825305073859988, + "grad_norm": 1.104650486752359, + "learning_rate": 5.786125011304355e-07, + "loss": 0.0556, + "step": 4021 + }, + { + "epoch": 2.5831727681438665, + "grad_norm": 0.6867658236662502, + "learning_rate": 5.768685500047988e-07, + "loss": 0.0207, + "step": 4022 + }, + { + "epoch": 2.583815028901734, + "grad_norm": 0.9385879324751089, + "learning_rate": 5.75127070095643e-07, + "loss": 0.0572, + "step": 4023 + }, + { + "epoch": 2.5844572896596016, + "grad_norm": 0.5738095488374118, + "learning_rate": 5.733880623759352e-07, + "loss": 0.0214, + "step": 4024 + }, + { + "epoch": 2.5850995504174694, + "grad_norm": 0.778386485976196, + "learning_rate": 5.716515278172651e-07, + "loss": 0.0314, + "step": 4025 + }, + { + "epoch": 2.585741811175337, + "grad_norm": 0.9632476904840558, + "learning_rate": 5.699174673898395e-07, + "loss": 0.0417, + "step": 4026 + }, + { + "epoch": 2.586384071933205, + "grad_norm": 1.3092129838406619, + "learning_rate": 5.681858820624842e-07, + "loss": 0.0411, + "step": 4027 + }, + { + "epoch": 2.5870263326910727, + "grad_norm": 0.8701169875112895, + "learning_rate": 5.664567728026398e-07, + "loss": 0.0351, + "step": 4028 + }, + { + "epoch": 2.5876685934489405, + "grad_norm": 0.691370328087994, + "learning_rate": 5.647301405763661e-07, + "loss": 0.0229, + "step": 4029 + }, + { + "epoch": 2.588310854206808, + "grad_norm": 0.5221026512602008, + "learning_rate": 5.630059863483361e-07, + "loss": 0.0186, + "step": 4030 + }, + { + "epoch": 2.5889531149646756, + "grad_norm": 1.0300583875694358, + "learning_rate": 5.612843110818417e-07, + "loss": 0.0343, + "step": 4031 + }, + { + "epoch": 2.5895953757225434, + "grad_norm": 0.7628811850634054, + "learning_rate": 5.595651157387855e-07, + "loss": 0.0411, + "step": 4032 + }, + { + "epoch": 2.590237636480411, + "grad_norm": 1.0301310363243055, + "learning_rate": 5.578484012796875e-07, + "loss": 0.055, + "step": 4033 + }, + { + "epoch": 2.5908798972382785, + "grad_norm": 0.8049965075800796, + "learning_rate": 5.561341686636823e-07, + "loss": 0.0301, + "step": 4034 + }, + { + "epoch": 2.5915221579961463, + "grad_norm": 0.9376299708303631, + "learning_rate": 5.544224188485159e-07, + "loss": 0.0314, + "step": 4035 + }, + { + "epoch": 2.592164418754014, + "grad_norm": 0.8469822034051938, + "learning_rate": 5.527131527905477e-07, + "loss": 0.026, + "step": 4036 + }, + { + "epoch": 2.592806679511882, + "grad_norm": 0.5894666648863213, + "learning_rate": 5.510063714447506e-07, + "loss": 0.0204, + "step": 4037 + }, + { + "epoch": 2.5934489402697496, + "grad_norm": 0.8792067050188866, + "learning_rate": 5.493020757647061e-07, + "loss": 0.0322, + "step": 4038 + }, + { + "epoch": 2.5940912010276174, + "grad_norm": 0.8727400058753715, + "learning_rate": 5.476002667026132e-07, + "loss": 0.0436, + "step": 4039 + }, + { + "epoch": 2.594733461785485, + "grad_norm": 0.7640973180141492, + "learning_rate": 5.45900945209274e-07, + "loss": 0.0276, + "step": 4040 + }, + { + "epoch": 2.5953757225433525, + "grad_norm": 1.0789754485988445, + "learning_rate": 5.442041122341057e-07, + "loss": 0.0511, + "step": 4041 + }, + { + "epoch": 2.5960179833012202, + "grad_norm": 0.9066008964781929, + "learning_rate": 5.425097687251335e-07, + "loss": 0.0415, + "step": 4042 + }, + { + "epoch": 2.596660244059088, + "grad_norm": 0.8872730129903379, + "learning_rate": 5.408179156289927e-07, + "loss": 0.0358, + "step": 4043 + }, + { + "epoch": 2.597302504816956, + "grad_norm": 1.0114726055773537, + "learning_rate": 5.391285538909257e-07, + "loss": 0.0462, + "step": 4044 + }, + { + "epoch": 2.597944765574823, + "grad_norm": 0.9266248707228939, + "learning_rate": 5.374416844547847e-07, + "loss": 0.0398, + "step": 4045 + }, + { + "epoch": 2.598587026332691, + "grad_norm": 0.8904946068031891, + "learning_rate": 5.357573082630279e-07, + "loss": 0.0288, + "step": 4046 + }, + { + "epoch": 2.5992292870905587, + "grad_norm": 0.7461482512041265, + "learning_rate": 5.340754262567221e-07, + "loss": 0.034, + "step": 4047 + }, + { + "epoch": 2.5998715478484264, + "grad_norm": 0.6379581573136739, + "learning_rate": 5.323960393755389e-07, + "loss": 0.0206, + "step": 4048 + }, + { + "epoch": 2.600513808606294, + "grad_norm": 1.3590973650001719, + "learning_rate": 5.307191485577568e-07, + "loss": 0.0322, + "step": 4049 + }, + { + "epoch": 2.601156069364162, + "grad_norm": 0.6813911939500648, + "learning_rate": 5.290447547402594e-07, + "loss": 0.0344, + "step": 4050 + }, + { + "epoch": 2.6017983301220298, + "grad_norm": 0.8297372823127955, + "learning_rate": 5.273728588585358e-07, + "loss": 0.0295, + "step": 4051 + }, + { + "epoch": 2.602440590879897, + "grad_norm": 0.7227733728893754, + "learning_rate": 5.257034618466794e-07, + "loss": 0.0234, + "step": 4052 + }, + { + "epoch": 2.603082851637765, + "grad_norm": 0.7569064028991586, + "learning_rate": 5.240365646373868e-07, + "loss": 0.0351, + "step": 4053 + }, + { + "epoch": 2.6037251123956326, + "grad_norm": 0.4958605402886765, + "learning_rate": 5.223721681619581e-07, + "loss": 0.013, + "step": 4054 + }, + { + "epoch": 2.6043673731535004, + "grad_norm": 0.4973009535882642, + "learning_rate": 5.207102733502978e-07, + "loss": 0.0137, + "step": 4055 + }, + { + "epoch": 2.6050096339113678, + "grad_norm": 0.6328462804253879, + "learning_rate": 5.190508811309092e-07, + "loss": 0.022, + "step": 4056 + }, + { + "epoch": 2.6056518946692355, + "grad_norm": 0.6987664212050777, + "learning_rate": 5.173939924309013e-07, + "loss": 0.0285, + "step": 4057 + }, + { + "epoch": 2.6062941554271033, + "grad_norm": 0.8062509043512534, + "learning_rate": 5.157396081759814e-07, + "loss": 0.0358, + "step": 4058 + }, + { + "epoch": 2.606936416184971, + "grad_norm": 0.4474184832796496, + "learning_rate": 5.14087729290459e-07, + "loss": 0.0119, + "step": 4059 + }, + { + "epoch": 2.607578676942839, + "grad_norm": 0.8480790195562893, + "learning_rate": 5.124383566972452e-07, + "loss": 0.0411, + "step": 4060 + }, + { + "epoch": 2.6082209377007066, + "grad_norm": 0.6042680980580047, + "learning_rate": 5.10791491317848e-07, + "loss": 0.0172, + "step": 4061 + }, + { + "epoch": 2.6088631984585744, + "grad_norm": 0.8610177219673681, + "learning_rate": 5.091471340723763e-07, + "loss": 0.0243, + "step": 4062 + }, + { + "epoch": 2.6095054592164417, + "grad_norm": 0.6387498347393284, + "learning_rate": 5.075052858795382e-07, + "loss": 0.0179, + "step": 4063 + }, + { + "epoch": 2.6101477199743095, + "grad_norm": 0.7818832021841267, + "learning_rate": 5.058659476566369e-07, + "loss": 0.0404, + "step": 4064 + }, + { + "epoch": 2.6107899807321773, + "grad_norm": 1.031030973468938, + "learning_rate": 5.042291203195765e-07, + "loss": 0.0523, + "step": 4065 + }, + { + "epoch": 2.611432241490045, + "grad_norm": 1.02207567871468, + "learning_rate": 5.025948047828577e-07, + "loss": 0.0516, + "step": 4066 + }, + { + "epoch": 2.6120745022479124, + "grad_norm": 0.8311231049965282, + "learning_rate": 5.009630019595774e-07, + "loss": 0.0283, + "step": 4067 + }, + { + "epoch": 2.61271676300578, + "grad_norm": 0.5908372454812197, + "learning_rate": 4.993337127614273e-07, + "loss": 0.0167, + "step": 4068 + }, + { + "epoch": 2.613359023763648, + "grad_norm": 0.8620514859771894, + "learning_rate": 4.977069380986976e-07, + "loss": 0.036, + "step": 4069 + }, + { + "epoch": 2.6140012845215157, + "grad_norm": 0.5807283699295648, + "learning_rate": 4.960826788802714e-07, + "loss": 0.0211, + "step": 4070 + }, + { + "epoch": 2.6146435452793835, + "grad_norm": 0.6856402467735184, + "learning_rate": 4.944609360136271e-07, + "loss": 0.0321, + "step": 4071 + }, + { + "epoch": 2.6152858060372512, + "grad_norm": 0.48420218577980617, + "learning_rate": 4.928417104048371e-07, + "loss": 0.0162, + "step": 4072 + }, + { + "epoch": 2.615928066795119, + "grad_norm": 1.0293249591253435, + "learning_rate": 4.912250029585675e-07, + "loss": 0.0413, + "step": 4073 + }, + { + "epoch": 2.6165703275529864, + "grad_norm": 0.6509559990099293, + "learning_rate": 4.896108145780775e-07, + "loss": 0.0275, + "step": 4074 + }, + { + "epoch": 2.617212588310854, + "grad_norm": 0.6566322638667411, + "learning_rate": 4.879991461652195e-07, + "loss": 0.0247, + "step": 4075 + }, + { + "epoch": 2.617854849068722, + "grad_norm": 0.5072147995071363, + "learning_rate": 4.863899986204362e-07, + "loss": 0.0148, + "step": 4076 + }, + { + "epoch": 2.6184971098265897, + "grad_norm": 0.48657472420150116, + "learning_rate": 4.847833728427636e-07, + "loss": 0.0138, + "step": 4077 + }, + { + "epoch": 2.6191393705844574, + "grad_norm": 1.1250543660206551, + "learning_rate": 4.831792697298293e-07, + "loss": 0.046, + "step": 4078 + }, + { + "epoch": 2.619781631342325, + "grad_norm": 0.961696992366738, + "learning_rate": 4.815776901778485e-07, + "loss": 0.046, + "step": 4079 + }, + { + "epoch": 2.6204238921001926, + "grad_norm": 0.7516640067370822, + "learning_rate": 4.799786350816294e-07, + "loss": 0.026, + "step": 4080 + }, + { + "epoch": 2.6210661528580603, + "grad_norm": 1.0047659977948409, + "learning_rate": 4.783821053345678e-07, + "loss": 0.0321, + "step": 4081 + }, + { + "epoch": 2.621708413615928, + "grad_norm": 0.9744863287268092, + "learning_rate": 4.7678810182865076e-07, + "loss": 0.0454, + "step": 4082 + }, + { + "epoch": 2.622350674373796, + "grad_norm": 0.6979702029866883, + "learning_rate": 4.7519662545445266e-07, + "loss": 0.0243, + "step": 4083 + }, + { + "epoch": 2.6229929351316636, + "grad_norm": 1.0148479331300444, + "learning_rate": 4.73607677101135e-07, + "loss": 0.0445, + "step": 4084 + }, + { + "epoch": 2.6236351958895314, + "grad_norm": 0.7317779927116093, + "learning_rate": 4.7202125765644923e-07, + "loss": 0.0272, + "step": 4085 + }, + { + "epoch": 2.6242774566473988, + "grad_norm": 0.731200876061173, + "learning_rate": 4.7043736800673254e-07, + "loss": 0.0225, + "step": 4086 + }, + { + "epoch": 2.6249197174052665, + "grad_norm": 0.7639369637031376, + "learning_rate": 4.6885600903690655e-07, + "loss": 0.0242, + "step": 4087 + }, + { + "epoch": 2.6255619781631343, + "grad_norm": 0.8093345164056921, + "learning_rate": 4.672771816304833e-07, + "loss": 0.0254, + "step": 4088 + }, + { + "epoch": 2.626204238921002, + "grad_norm": 0.6462435879634216, + "learning_rate": 4.657008866695567e-07, + "loss": 0.0232, + "step": 4089 + }, + { + "epoch": 2.6268464996788694, + "grad_norm": 0.6673223904207207, + "learning_rate": 4.6412712503480894e-07, + "loss": 0.0292, + "step": 4090 + }, + { + "epoch": 2.627488760436737, + "grad_norm": 0.9087902924142007, + "learning_rate": 4.625558976055039e-07, + "loss": 0.0353, + "step": 4091 + }, + { + "epoch": 2.628131021194605, + "grad_norm": 0.6356535615689707, + "learning_rate": 4.6098720525949217e-07, + "loss": 0.0179, + "step": 4092 + }, + { + "epoch": 2.6287732819524727, + "grad_norm": 0.9168023183547914, + "learning_rate": 4.5942104887320485e-07, + "loss": 0.0415, + "step": 4093 + }, + { + "epoch": 2.6294155427103405, + "grad_norm": 0.8825159700850859, + "learning_rate": 4.5785742932166144e-07, + "loss": 0.0333, + "step": 4094 + }, + { + "epoch": 2.6300578034682083, + "grad_norm": 0.6520429794670413, + "learning_rate": 4.5629634747845764e-07, + "loss": 0.0217, + "step": 4095 + }, + { + "epoch": 2.630700064226076, + "grad_norm": 0.6089658726650409, + "learning_rate": 4.5473780421577517e-07, + "loss": 0.0199, + "step": 4096 + }, + { + "epoch": 2.6313423249839434, + "grad_norm": 0.6412640336633882, + "learning_rate": 4.5318180040437756e-07, + "loss": 0.0221, + "step": 4097 + }, + { + "epoch": 2.631984585741811, + "grad_norm": 0.9432079040123514, + "learning_rate": 4.5162833691360765e-07, + "loss": 0.0418, + "step": 4098 + }, + { + "epoch": 2.632626846499679, + "grad_norm": 0.5476324585080554, + "learning_rate": 4.5007741461139066e-07, + "loss": 0.0122, + "step": 4099 + }, + { + "epoch": 2.6332691072575467, + "grad_norm": 0.9614484402759206, + "learning_rate": 4.485290343642318e-07, + "loss": 0.036, + "step": 4100 + }, + { + "epoch": 2.633911368015414, + "grad_norm": 0.6635976463773263, + "learning_rate": 4.469831970372146e-07, + "loss": 0.0245, + "step": 4101 + }, + { + "epoch": 2.634553628773282, + "grad_norm": 0.9049313187849871, + "learning_rate": 4.4543990349400377e-07, + "loss": 0.0482, + "step": 4102 + }, + { + "epoch": 2.6351958895311496, + "grad_norm": 0.9315172257841404, + "learning_rate": 4.438991545968413e-07, + "loss": 0.0258, + "step": 4103 + }, + { + "epoch": 2.6358381502890174, + "grad_norm": 0.6705444873240045, + "learning_rate": 4.423609512065485e-07, + "loss": 0.0183, + "step": 4104 + }, + { + "epoch": 2.636480411046885, + "grad_norm": 0.5536147314941959, + "learning_rate": 4.4082529418252415e-07, + "loss": 0.013, + "step": 4105 + }, + { + "epoch": 2.637122671804753, + "grad_norm": 0.7434026224384882, + "learning_rate": 4.392921843827441e-07, + "loss": 0.0311, + "step": 4106 + }, + { + "epoch": 2.6377649325626207, + "grad_norm": 0.9384183126051935, + "learning_rate": 4.377616226637615e-07, + "loss": 0.0339, + "step": 4107 + }, + { + "epoch": 2.638407193320488, + "grad_norm": 0.7859399157796089, + "learning_rate": 4.3623360988070515e-07, + "loss": 0.0289, + "step": 4108 + }, + { + "epoch": 2.639049454078356, + "grad_norm": 0.5116854908415135, + "learning_rate": 4.3470814688728105e-07, + "loss": 0.0233, + "step": 4109 + }, + { + "epoch": 2.6396917148362236, + "grad_norm": 0.760463071280189, + "learning_rate": 4.331852345357701e-07, + "loss": 0.0297, + "step": 4110 + }, + { + "epoch": 2.6403339755940913, + "grad_norm": 0.5627742983998769, + "learning_rate": 4.3166487367702627e-07, + "loss": 0.0173, + "step": 4111 + }, + { + "epoch": 2.6409762363519587, + "grad_norm": 0.7563893357961711, + "learning_rate": 4.301470651604806e-07, + "loss": 0.0238, + "step": 4112 + }, + { + "epoch": 2.6416184971098264, + "grad_norm": 0.5546396797056319, + "learning_rate": 4.2863180983413744e-07, + "loss": 0.0176, + "step": 4113 + }, + { + "epoch": 2.642260757867694, + "grad_norm": 0.7534325659019363, + "learning_rate": 4.271191085445736e-07, + "loss": 0.0392, + "step": 4114 + }, + { + "epoch": 2.642903018625562, + "grad_norm": 1.0794902789473033, + "learning_rate": 4.2560896213693913e-07, + "loss": 0.065, + "step": 4115 + }, + { + "epoch": 2.6435452793834298, + "grad_norm": 0.9208253859388537, + "learning_rate": 4.2410137145495965e-07, + "loss": 0.0463, + "step": 4116 + }, + { + "epoch": 2.6441875401412975, + "grad_norm": 0.7310320363873973, + "learning_rate": 4.2259633734092863e-07, + "loss": 0.025, + "step": 4117 + }, + { + "epoch": 2.6448298008991653, + "grad_norm": 0.7438838185841202, + "learning_rate": 4.210938606357146e-07, + "loss": 0.0275, + "step": 4118 + }, + { + "epoch": 2.6454720616570326, + "grad_norm": 0.8261908431813841, + "learning_rate": 4.195939421787537e-07, + "loss": 0.022, + "step": 4119 + }, + { + "epoch": 2.6461143224149004, + "grad_norm": 1.0255496119045955, + "learning_rate": 4.180965828080558e-07, + "loss": 0.0369, + "step": 4120 + }, + { + "epoch": 2.646756583172768, + "grad_norm": 0.5593993722158408, + "learning_rate": 4.1660178336019983e-07, + "loss": 0.0183, + "step": 4121 + }, + { + "epoch": 2.647398843930636, + "grad_norm": 0.978382304553732, + "learning_rate": 4.1510954467033457e-07, + "loss": 0.0424, + "step": 4122 + }, + { + "epoch": 2.6480411046885033, + "grad_norm": 0.7187752535776888, + "learning_rate": 4.136198675721781e-07, + "loss": 0.0251, + "step": 4123 + }, + { + "epoch": 2.648683365446371, + "grad_norm": 0.6379937314173718, + "learning_rate": 4.1213275289801734e-07, + "loss": 0.0208, + "step": 4124 + }, + { + "epoch": 2.649325626204239, + "grad_norm": 0.9042867861816296, + "learning_rate": 4.106482014787078e-07, + "loss": 0.05, + "step": 4125 + }, + { + "epoch": 2.6499678869621066, + "grad_norm": 0.8548493951047983, + "learning_rate": 4.0916621414367255e-07, + "loss": 0.0431, + "step": 4126 + }, + { + "epoch": 2.6506101477199744, + "grad_norm": 0.7045564152013616, + "learning_rate": 4.076867917209021e-07, + "loss": 0.0263, + "step": 4127 + }, + { + "epoch": 2.651252408477842, + "grad_norm": 0.9929065898044164, + "learning_rate": 4.0620993503695474e-07, + "loss": 0.0494, + "step": 4128 + }, + { + "epoch": 2.65189466923571, + "grad_norm": 0.8854750143394315, + "learning_rate": 4.047356449169537e-07, + "loss": 0.0299, + "step": 4129 + }, + { + "epoch": 2.6525369299935773, + "grad_norm": 0.6612905326555844, + "learning_rate": 4.032639221845902e-07, + "loss": 0.0214, + "step": 4130 + }, + { + "epoch": 2.653179190751445, + "grad_norm": 0.7304673319053846, + "learning_rate": 4.0179476766211865e-07, + "loss": 0.0276, + "step": 4131 + }, + { + "epoch": 2.653821451509313, + "grad_norm": 0.8634497397223542, + "learning_rate": 4.003281821703614e-07, + "loss": 0.0376, + "step": 4132 + }, + { + "epoch": 2.6544637122671806, + "grad_norm": 0.6959482536536848, + "learning_rate": 3.988641665287046e-07, + "loss": 0.0209, + "step": 4133 + }, + { + "epoch": 2.655105973025048, + "grad_norm": 0.6988600293279661, + "learning_rate": 3.9740272155509576e-07, + "loss": 0.0173, + "step": 4134 + }, + { + "epoch": 2.6557482337829157, + "grad_norm": 1.031980124999972, + "learning_rate": 3.9594384806604937e-07, + "loss": 0.0588, + "step": 4135 + }, + { + "epoch": 2.6563904945407835, + "grad_norm": 0.6337472854077523, + "learning_rate": 3.944875468766424e-07, + "loss": 0.0252, + "step": 4136 + }, + { + "epoch": 2.6570327552986512, + "grad_norm": 1.249840093543661, + "learning_rate": 3.9303381880051506e-07, + "loss": 0.0624, + "step": 4137 + }, + { + "epoch": 2.657675016056519, + "grad_norm": 0.5858742078365281, + "learning_rate": 3.915826646498694e-07, + "loss": 0.0114, + "step": 4138 + }, + { + "epoch": 2.658317276814387, + "grad_norm": 1.125749656438825, + "learning_rate": 3.9013408523546903e-07, + "loss": 0.0381, + "step": 4139 + }, + { + "epoch": 2.6589595375722546, + "grad_norm": 1.1014072091076075, + "learning_rate": 3.8868808136663995e-07, + "loss": 0.048, + "step": 4140 + }, + { + "epoch": 2.659601798330122, + "grad_norm": 0.6630905930519685, + "learning_rate": 3.8724465385126976e-07, + "loss": 0.0268, + "step": 4141 + }, + { + "epoch": 2.6602440590879897, + "grad_norm": 0.9402621763292162, + "learning_rate": 3.858038034958034e-07, + "loss": 0.036, + "step": 4142 + }, + { + "epoch": 2.6608863198458574, + "grad_norm": 0.9632623601802703, + "learning_rate": 3.843655311052491e-07, + "loss": 0.0421, + "step": 4143 + }, + { + "epoch": 2.661528580603725, + "grad_norm": 0.8756589316650113, + "learning_rate": 3.829298374831736e-07, + "loss": 0.0238, + "step": 4144 + }, + { + "epoch": 2.6621708413615925, + "grad_norm": 0.9556344377010694, + "learning_rate": 3.814967234317041e-07, + "loss": 0.0369, + "step": 4145 + }, + { + "epoch": 2.6628131021194603, + "grad_norm": 0.7678831128699125, + "learning_rate": 3.8006618975152456e-07, + "loss": 0.0323, + "step": 4146 + }, + { + "epoch": 2.663455362877328, + "grad_norm": 0.7100139043193722, + "learning_rate": 3.7863823724187896e-07, + "loss": 0.0198, + "step": 4147 + }, + { + "epoch": 2.664097623635196, + "grad_norm": 0.9781718340581325, + "learning_rate": 3.7721286670056866e-07, + "loss": 0.0541, + "step": 4148 + }, + { + "epoch": 2.6647398843930636, + "grad_norm": 0.6200551088979969, + "learning_rate": 3.757900789239516e-07, + "loss": 0.0209, + "step": 4149 + }, + { + "epoch": 2.6653821451509314, + "grad_norm": 0.8564092170951058, + "learning_rate": 3.743698747069441e-07, + "loss": 0.0363, + "step": 4150 + }, + { + "epoch": 2.666024405908799, + "grad_norm": 1.0407469242444793, + "learning_rate": 3.729522548430181e-07, + "loss": 0.0414, + "step": 4151 + }, + { + "epoch": 2.6666666666666665, + "grad_norm": 0.4445165257638502, + "learning_rate": 3.715372201242023e-07, + "loss": 0.0165, + "step": 4152 + }, + { + "epoch": 2.6673089274245343, + "grad_norm": 0.7892086712655708, + "learning_rate": 3.701247713410805e-07, + "loss": 0.0246, + "step": 4153 + }, + { + "epoch": 2.667951188182402, + "grad_norm": 0.9822404508166617, + "learning_rate": 3.6871490928279184e-07, + "loss": 0.0424, + "step": 4154 + }, + { + "epoch": 2.66859344894027, + "grad_norm": 1.0461523258514824, + "learning_rate": 3.673076347370308e-07, + "loss": 0.0481, + "step": 4155 + }, + { + "epoch": 2.669235709698137, + "grad_norm": 0.7581090315930765, + "learning_rate": 3.659029484900456e-07, + "loss": 0.0291, + "step": 4156 + }, + { + "epoch": 2.669877970456005, + "grad_norm": 0.8165652720240002, + "learning_rate": 3.64500851326639e-07, + "loss": 0.0324, + "step": 4157 + }, + { + "epoch": 2.6705202312138727, + "grad_norm": 0.5786469722314718, + "learning_rate": 3.631013440301645e-07, + "loss": 0.02, + "step": 4158 + }, + { + "epoch": 2.6711624919717405, + "grad_norm": 0.8261999072918903, + "learning_rate": 3.6170442738253377e-07, + "loss": 0.0288, + "step": 4159 + }, + { + "epoch": 2.6718047527296083, + "grad_norm": 0.6874855439379316, + "learning_rate": 3.6031010216420695e-07, + "loss": 0.0236, + "step": 4160 + }, + { + "epoch": 2.672447013487476, + "grad_norm": 0.6993811266745155, + "learning_rate": 3.589183691541981e-07, + "loss": 0.0211, + "step": 4161 + }, + { + "epoch": 2.673089274245344, + "grad_norm": 0.5728955349922138, + "learning_rate": 3.575292291300719e-07, + "loss": 0.0164, + "step": 4162 + }, + { + "epoch": 2.673731535003211, + "grad_norm": 0.9140545482042238, + "learning_rate": 3.561426828679454e-07, + "loss": 0.0412, + "step": 4163 + }, + { + "epoch": 2.674373795761079, + "grad_norm": 0.8908299598316368, + "learning_rate": 3.5475873114248607e-07, + "loss": 0.0365, + "step": 4164 + }, + { + "epoch": 2.6750160565189467, + "grad_norm": 0.8752969558931644, + "learning_rate": 3.5337737472691257e-07, + "loss": 0.0431, + "step": 4165 + }, + { + "epoch": 2.6756583172768145, + "grad_norm": 0.6755535041919708, + "learning_rate": 3.5199861439299097e-07, + "loss": 0.0274, + "step": 4166 + }, + { + "epoch": 2.6763005780346822, + "grad_norm": 1.2016536253110919, + "learning_rate": 3.50622450911039e-07, + "loss": 0.0379, + "step": 4167 + }, + { + "epoch": 2.6769428387925496, + "grad_norm": 1.068963202171223, + "learning_rate": 3.492488850499243e-07, + "loss": 0.0385, + "step": 4168 + }, + { + "epoch": 2.6775850995504173, + "grad_norm": 0.5380113891536615, + "learning_rate": 3.478779175770619e-07, + "loss": 0.0118, + "step": 4169 + }, + { + "epoch": 2.678227360308285, + "grad_norm": 0.5675798064733943, + "learning_rate": 3.465095492584136e-07, + "loss": 0.0187, + "step": 4170 + }, + { + "epoch": 2.678869621066153, + "grad_norm": 0.6573951395670335, + "learning_rate": 3.4514378085849344e-07, + "loss": 0.0238, + "step": 4171 + }, + { + "epoch": 2.6795118818240207, + "grad_norm": 0.6989428349287051, + "learning_rate": 3.4378061314035926e-07, + "loss": 0.0283, + "step": 4172 + }, + { + "epoch": 2.6801541425818884, + "grad_norm": 0.6769298888357493, + "learning_rate": 3.4242004686561746e-07, + "loss": 0.019, + "step": 4173 + }, + { + "epoch": 2.680796403339756, + "grad_norm": 0.582396288105216, + "learning_rate": 3.410620827944183e-07, + "loss": 0.0174, + "step": 4174 + }, + { + "epoch": 2.6814386640976235, + "grad_norm": 0.7726163470567566, + "learning_rate": 3.397067216854616e-07, + "loss": 0.0248, + "step": 4175 + }, + { + "epoch": 2.6820809248554913, + "grad_norm": 0.731645892920012, + "learning_rate": 3.383539642959915e-07, + "loss": 0.0282, + "step": 4176 + }, + { + "epoch": 2.682723185613359, + "grad_norm": 0.8716734633517422, + "learning_rate": 3.3700381138179806e-07, + "loss": 0.0356, + "step": 4177 + }, + { + "epoch": 2.683365446371227, + "grad_norm": 0.9944909306203501, + "learning_rate": 3.356562636972149e-07, + "loss": 0.0409, + "step": 4178 + }, + { + "epoch": 2.684007707129094, + "grad_norm": 0.733793869715519, + "learning_rate": 3.343113219951205e-07, + "loss": 0.0272, + "step": 4179 + }, + { + "epoch": 2.684649967886962, + "grad_norm": 0.6756117010813023, + "learning_rate": 3.329689870269393e-07, + "loss": 0.0296, + "step": 4180 + }, + { + "epoch": 2.6852922286448297, + "grad_norm": 1.0792311934194845, + "learning_rate": 3.316292595426363e-07, + "loss": 0.0519, + "step": 4181 + }, + { + "epoch": 2.6859344894026975, + "grad_norm": 0.7342535648328176, + "learning_rate": 3.302921402907211e-07, + "loss": 0.0127, + "step": 4182 + }, + { + "epoch": 2.6865767501605653, + "grad_norm": 0.4871046348304316, + "learning_rate": 3.289576300182473e-07, + "loss": 0.0173, + "step": 4183 + }, + { + "epoch": 2.687219010918433, + "grad_norm": 0.8936098967558159, + "learning_rate": 3.2762572947080907e-07, + "loss": 0.031, + "step": 4184 + }, + { + "epoch": 2.687861271676301, + "grad_norm": 0.8311369690061637, + "learning_rate": 3.262964393925433e-07, + "loss": 0.0304, + "step": 4185 + }, + { + "epoch": 2.688503532434168, + "grad_norm": 0.7404782993812165, + "learning_rate": 3.249697605261276e-07, + "loss": 0.0309, + "step": 4186 + }, + { + "epoch": 2.689145793192036, + "grad_norm": 1.0082697620588557, + "learning_rate": 3.2364569361278187e-07, + "loss": 0.0476, + "step": 4187 + }, + { + "epoch": 2.6897880539499037, + "grad_norm": 0.76793487965566, + "learning_rate": 3.223242393922671e-07, + "loss": 0.0324, + "step": 4188 + }, + { + "epoch": 2.6904303147077715, + "grad_norm": 0.8915729799170616, + "learning_rate": 3.210053986028822e-07, + "loss": 0.026, + "step": 4189 + }, + { + "epoch": 2.691072575465639, + "grad_norm": 0.7036096532764122, + "learning_rate": 3.1968917198146765e-07, + "loss": 0.0173, + "step": 4190 + }, + { + "epoch": 2.6917148362235066, + "grad_norm": 0.7778833127572934, + "learning_rate": 3.18375560263402e-07, + "loss": 0.0159, + "step": 4191 + }, + { + "epoch": 2.6923570969813744, + "grad_norm": 0.7886204052025356, + "learning_rate": 3.1706456418260643e-07, + "loss": 0.0239, + "step": 4192 + }, + { + "epoch": 2.692999357739242, + "grad_norm": 0.9370865639494265, + "learning_rate": 3.1575618447153736e-07, + "loss": 0.0481, + "step": 4193 + }, + { + "epoch": 2.69364161849711, + "grad_norm": 0.9334238740414301, + "learning_rate": 3.144504218611899e-07, + "loss": 0.045, + "step": 4194 + }, + { + "epoch": 2.6942838792549777, + "grad_norm": 0.8030953386205648, + "learning_rate": 3.1314727708109725e-07, + "loss": 0.0336, + "step": 4195 + }, + { + "epoch": 2.6949261400128455, + "grad_norm": 0.8663820985948306, + "learning_rate": 3.11846750859332e-07, + "loss": 0.0207, + "step": 4196 + }, + { + "epoch": 2.695568400770713, + "grad_norm": 1.1679418568668427, + "learning_rate": 3.1054884392249964e-07, + "loss": 0.0624, + "step": 4197 + }, + { + "epoch": 2.6962106615285806, + "grad_norm": 0.7456956815365957, + "learning_rate": 3.0925355699574554e-07, + "loss": 0.0197, + "step": 4198 + }, + { + "epoch": 2.6968529222864484, + "grad_norm": 0.4921754609659596, + "learning_rate": 3.079608908027498e-07, + "loss": 0.0125, + "step": 4199 + }, + { + "epoch": 2.697495183044316, + "grad_norm": 0.5765972486015469, + "learning_rate": 3.0667084606572973e-07, + "loss": 0.0225, + "step": 4200 + }, + { + "epoch": 2.6981374438021835, + "grad_norm": 0.616296304935857, + "learning_rate": 3.053834235054365e-07, + "loss": 0.017, + "step": 4201 + }, + { + "epoch": 2.6987797045600512, + "grad_norm": 1.1785408624817195, + "learning_rate": 3.04098623841157e-07, + "loss": 0.0606, + "step": 4202 + }, + { + "epoch": 2.699421965317919, + "grad_norm": 0.7402472238127319, + "learning_rate": 3.028164477907125e-07, + "loss": 0.0217, + "step": 4203 + }, + { + "epoch": 2.7000642260757868, + "grad_norm": 0.9608214827316761, + "learning_rate": 3.015368960704584e-07, + "loss": 0.0442, + "step": 4204 + }, + { + "epoch": 2.7007064868336546, + "grad_norm": 0.8749775106486772, + "learning_rate": 3.0025996939528403e-07, + "loss": 0.0272, + "step": 4205 + }, + { + "epoch": 2.7013487475915223, + "grad_norm": 0.7068370775901651, + "learning_rate": 2.9898566847861243e-07, + "loss": 0.0217, + "step": 4206 + }, + { + "epoch": 2.70199100834939, + "grad_norm": 0.3751297915091791, + "learning_rate": 2.977139940323992e-07, + "loss": 0.0109, + "step": 4207 + }, + { + "epoch": 2.7026332691072574, + "grad_norm": 0.7628368049987121, + "learning_rate": 2.964449467671321e-07, + "loss": 0.0314, + "step": 4208 + }, + { + "epoch": 2.703275529865125, + "grad_norm": 0.9943109081256634, + "learning_rate": 2.951785273918323e-07, + "loss": 0.0614, + "step": 4209 + }, + { + "epoch": 2.703917790622993, + "grad_norm": 0.9180223828519397, + "learning_rate": 2.93914736614051e-07, + "loss": 0.0348, + "step": 4210 + }, + { + "epoch": 2.7045600513808608, + "grad_norm": 0.5682435379699399, + "learning_rate": 2.9265357513987237e-07, + "loss": 0.0136, + "step": 4211 + }, + { + "epoch": 2.705202312138728, + "grad_norm": 0.6159721553525317, + "learning_rate": 2.913950436739116e-07, + "loss": 0.0168, + "step": 4212 + }, + { + "epoch": 2.705844572896596, + "grad_norm": 0.9132965847789054, + "learning_rate": 2.901391429193123e-07, + "loss": 0.0311, + "step": 4213 + }, + { + "epoch": 2.7064868336544636, + "grad_norm": 0.8535836685935405, + "learning_rate": 2.888858735777511e-07, + "loss": 0.0301, + "step": 4214 + }, + { + "epoch": 2.7071290944123314, + "grad_norm": 0.5180485791494617, + "learning_rate": 2.8763523634943336e-07, + "loss": 0.0195, + "step": 4215 + }, + { + "epoch": 2.707771355170199, + "grad_norm": 0.6538220327578576, + "learning_rate": 2.8638723193309295e-07, + "loss": 0.0182, + "step": 4216 + }, + { + "epoch": 2.708413615928067, + "grad_norm": 0.8213911906712176, + "learning_rate": 2.851418610259943e-07, + "loss": 0.0313, + "step": 4217 + }, + { + "epoch": 2.7090558766859347, + "grad_norm": 0.6756270022896452, + "learning_rate": 2.8389912432392974e-07, + "loss": 0.0254, + "step": 4218 + }, + { + "epoch": 2.709698137443802, + "grad_norm": 0.5433216788032404, + "learning_rate": 2.8265902252121946e-07, + "loss": 0.0193, + "step": 4219 + }, + { + "epoch": 2.71034039820167, + "grad_norm": 0.861367979569251, + "learning_rate": 2.814215563107131e-07, + "loss": 0.0345, + "step": 4220 + }, + { + "epoch": 2.7109826589595376, + "grad_norm": 0.7292921957626912, + "learning_rate": 2.8018672638378486e-07, + "loss": 0.0262, + "step": 4221 + }, + { + "epoch": 2.7116249197174054, + "grad_norm": 0.8454650319262101, + "learning_rate": 2.78954533430339e-07, + "loss": 0.0291, + "step": 4222 + }, + { + "epoch": 2.7122671804752727, + "grad_norm": 0.6758674615592871, + "learning_rate": 2.777249781388047e-07, + "loss": 0.0294, + "step": 4223 + }, + { + "epoch": 2.7129094412331405, + "grad_norm": 0.7903304880883154, + "learning_rate": 2.7649806119613877e-07, + "loss": 0.0273, + "step": 4224 + }, + { + "epoch": 2.7135517019910083, + "grad_norm": 1.3305573408805933, + "learning_rate": 2.752737832878222e-07, + "loss": 0.0682, + "step": 4225 + }, + { + "epoch": 2.714193962748876, + "grad_norm": 0.7844924029355778, + "learning_rate": 2.740521450978639e-07, + "loss": 0.0388, + "step": 4226 + }, + { + "epoch": 2.714836223506744, + "grad_norm": 0.9267072546810017, + "learning_rate": 2.728331473087964e-07, + "loss": 0.038, + "step": 4227 + }, + { + "epoch": 2.7154784842646116, + "grad_norm": 0.456377895038467, + "learning_rate": 2.7161679060167757e-07, + "loss": 0.0171, + "step": 4228 + }, + { + "epoch": 2.7161207450224794, + "grad_norm": 0.6594530213958549, + "learning_rate": 2.704030756560888e-07, + "loss": 0.0209, + "step": 4229 + }, + { + "epoch": 2.7167630057803467, + "grad_norm": 0.7384805140701434, + "learning_rate": 2.6919200315013606e-07, + "loss": 0.0289, + "step": 4230 + }, + { + "epoch": 2.7174052665382145, + "grad_norm": 0.9466589288045165, + "learning_rate": 2.679835737604491e-07, + "loss": 0.0176, + "step": 4231 + }, + { + "epoch": 2.7180475272960822, + "grad_norm": 1.3473696677579294, + "learning_rate": 2.667777881621819e-07, + "loss": 0.0831, + "step": 4232 + }, + { + "epoch": 2.71868978805395, + "grad_norm": 0.5494734421482029, + "learning_rate": 2.655746470290099e-07, + "loss": 0.018, + "step": 4233 + }, + { + "epoch": 2.7193320488118173, + "grad_norm": 0.6246006863306953, + "learning_rate": 2.64374151033131e-07, + "loss": 0.0172, + "step": 4234 + }, + { + "epoch": 2.719974309569685, + "grad_norm": 1.0017306841557576, + "learning_rate": 2.631763008452665e-07, + "loss": 0.0389, + "step": 4235 + }, + { + "epoch": 2.720616570327553, + "grad_norm": 0.9499392195326825, + "learning_rate": 2.619810971346587e-07, + "loss": 0.0404, + "step": 4236 + }, + { + "epoch": 2.7212588310854207, + "grad_norm": 0.7898838352654115, + "learning_rate": 2.6078854056907135e-07, + "loss": 0.0294, + "step": 4237 + }, + { + "epoch": 2.7219010918432884, + "grad_norm": 0.9097700277565617, + "learning_rate": 2.595986318147892e-07, + "loss": 0.0333, + "step": 4238 + }, + { + "epoch": 2.722543352601156, + "grad_norm": 0.6388714635434116, + "learning_rate": 2.5841137153661765e-07, + "loss": 0.0168, + "step": 4239 + }, + { + "epoch": 2.723185613359024, + "grad_norm": 0.7644053841678912, + "learning_rate": 2.5722676039788286e-07, + "loss": 0.0415, + "step": 4240 + }, + { + "epoch": 2.7238278741168913, + "grad_norm": 0.5312988003083194, + "learning_rate": 2.5604479906043035e-07, + "loss": 0.0152, + "step": 4241 + }, + { + "epoch": 2.724470134874759, + "grad_norm": 0.949887154083337, + "learning_rate": 2.548654881846252e-07, + "loss": 0.0314, + "step": 4242 + }, + { + "epoch": 2.725112395632627, + "grad_norm": 0.5686119471226309, + "learning_rate": 2.536888284293526e-07, + "loss": 0.0171, + "step": 4243 + }, + { + "epoch": 2.7257546563904946, + "grad_norm": 0.6896336002718985, + "learning_rate": 2.5251482045201446e-07, + "loss": 0.0243, + "step": 4244 + }, + { + "epoch": 2.726396917148362, + "grad_norm": 0.8948654693565046, + "learning_rate": 2.513434649085339e-07, + "loss": 0.0466, + "step": 4245 + }, + { + "epoch": 2.7270391779062297, + "grad_norm": 0.783885037160787, + "learning_rate": 2.5017476245334907e-07, + "loss": 0.0271, + "step": 4246 + }, + { + "epoch": 2.7276814386640975, + "grad_norm": 1.028498726278524, + "learning_rate": 2.4900871373941817e-07, + "loss": 0.0448, + "step": 4247 + }, + { + "epoch": 2.7283236994219653, + "grad_norm": 0.49193511403481954, + "learning_rate": 2.4784531941821675e-07, + "loss": 0.0118, + "step": 4248 + }, + { + "epoch": 2.728965960179833, + "grad_norm": 0.7156711583340456, + "learning_rate": 2.4668458013973664e-07, + "loss": 0.0266, + "step": 4249 + }, + { + "epoch": 2.729608220937701, + "grad_norm": 0.8163447951396711, + "learning_rate": 2.4552649655248507e-07, + "loss": 0.0391, + "step": 4250 + }, + { + "epoch": 2.7302504816955686, + "grad_norm": 1.0085205162034983, + "learning_rate": 2.4437106930348843e-07, + "loss": 0.0362, + "step": 4251 + }, + { + "epoch": 2.730892742453436, + "grad_norm": 0.8242215307386704, + "learning_rate": 2.432182990382853e-07, + "loss": 0.0244, + "step": 4252 + }, + { + "epoch": 2.7315350032113037, + "grad_norm": 0.7330215586686243, + "learning_rate": 2.4206818640093334e-07, + "loss": 0.0216, + "step": 4253 + }, + { + "epoch": 2.7321772639691715, + "grad_norm": 0.8277109386267635, + "learning_rate": 2.409207320340029e-07, + "loss": 0.0322, + "step": 4254 + }, + { + "epoch": 2.7328195247270393, + "grad_norm": 0.7124067637053263, + "learning_rate": 2.397759365785801e-07, + "loss": 0.0294, + "step": 4255 + }, + { + "epoch": 2.733461785484907, + "grad_norm": 0.8856373614008274, + "learning_rate": 2.386338006742661e-07, + "loss": 0.0283, + "step": 4256 + }, + { + "epoch": 2.7341040462427744, + "grad_norm": 0.9313507238042118, + "learning_rate": 2.3749432495917546e-07, + "loss": 0.0423, + "step": 4257 + }, + { + "epoch": 2.734746307000642, + "grad_norm": 0.9737491462841238, + "learning_rate": 2.3635751006993668e-07, + "loss": 0.0434, + "step": 4258 + }, + { + "epoch": 2.73538856775851, + "grad_norm": 0.9869040951946115, + "learning_rate": 2.3522335664169115e-07, + "loss": 0.0501, + "step": 4259 + }, + { + "epoch": 2.7360308285163777, + "grad_norm": 1.0923354615230438, + "learning_rate": 2.3409186530809425e-07, + "loss": 0.0476, + "step": 4260 + }, + { + "epoch": 2.7366730892742455, + "grad_norm": 0.8847446490546809, + "learning_rate": 2.329630367013125e-07, + "loss": 0.0363, + "step": 4261 + }, + { + "epoch": 2.7373153500321132, + "grad_norm": 0.845931176758415, + "learning_rate": 2.3183687145202694e-07, + "loss": 0.0374, + "step": 4262 + }, + { + "epoch": 2.7379576107899806, + "grad_norm": 1.1237326984395428, + "learning_rate": 2.3071337018942876e-07, + "loss": 0.0393, + "step": 4263 + }, + { + "epoch": 2.7385998715478483, + "grad_norm": 0.7040967512580185, + "learning_rate": 2.2959253354122135e-07, + "loss": 0.0257, + "step": 4264 + }, + { + "epoch": 2.739242132305716, + "grad_norm": 0.5360518174738905, + "learning_rate": 2.284743621336194e-07, + "loss": 0.0093, + "step": 4265 + }, + { + "epoch": 2.739884393063584, + "grad_norm": 0.8635762832792444, + "learning_rate": 2.2735885659134927e-07, + "loss": 0.033, + "step": 4266 + }, + { + "epoch": 2.7405266538214517, + "grad_norm": 0.7234859781912728, + "learning_rate": 2.2624601753764687e-07, + "loss": 0.0183, + "step": 4267 + }, + { + "epoch": 2.741168914579319, + "grad_norm": 1.0167855523369975, + "learning_rate": 2.2513584559425704e-07, + "loss": 0.032, + "step": 4268 + }, + { + "epoch": 2.7418111753371868, + "grad_norm": 0.5015848547785581, + "learning_rate": 2.240283413814387e-07, + "loss": 0.0143, + "step": 4269 + }, + { + "epoch": 2.7424534360950545, + "grad_norm": 0.7527320694808963, + "learning_rate": 2.2292350551795572e-07, + "loss": 0.025, + "step": 4270 + }, + { + "epoch": 2.7430956968529223, + "grad_norm": 0.8156182875032748, + "learning_rate": 2.2182133862108436e-07, + "loss": 0.0386, + "step": 4271 + }, + { + "epoch": 2.74373795761079, + "grad_norm": 0.7468447049070837, + "learning_rate": 2.2072184130660757e-07, + "loss": 0.0383, + "step": 4272 + }, + { + "epoch": 2.744380218368658, + "grad_norm": 0.8898753200040153, + "learning_rate": 2.1962501418881842e-07, + "loss": 0.039, + "step": 4273 + }, + { + "epoch": 2.7450224791265256, + "grad_norm": 0.6539448767947087, + "learning_rate": 2.1853085788051677e-07, + "loss": 0.0172, + "step": 4274 + }, + { + "epoch": 2.745664739884393, + "grad_norm": 0.8464468855245363, + "learning_rate": 2.1743937299301242e-07, + "loss": 0.0256, + "step": 4275 + }, + { + "epoch": 2.7463070006422607, + "grad_norm": 0.9294226396070959, + "learning_rate": 2.1635056013611877e-07, + "loss": 0.0415, + "step": 4276 + }, + { + "epoch": 2.7469492614001285, + "grad_norm": 1.0071185097116007, + "learning_rate": 2.1526441991816082e-07, + "loss": 0.0463, + "step": 4277 + }, + { + "epoch": 2.7475915221579963, + "grad_norm": 0.9019443861628618, + "learning_rate": 2.1418095294596707e-07, + "loss": 0.0427, + "step": 4278 + }, + { + "epoch": 2.7482337829158636, + "grad_norm": 0.6700514171817267, + "learning_rate": 2.1310015982487387e-07, + "loss": 0.0195, + "step": 4279 + }, + { + "epoch": 2.7488760436737314, + "grad_norm": 0.7891652036576821, + "learning_rate": 2.1202204115872316e-07, + "loss": 0.0267, + "step": 4280 + }, + { + "epoch": 2.749518304431599, + "grad_norm": 0.8861027479547599, + "learning_rate": 2.109465975498648e-07, + "loss": 0.0444, + "step": 4281 + }, + { + "epoch": 2.750160565189467, + "grad_norm": 0.5177780342536413, + "learning_rate": 2.0987382959915038e-07, + "loss": 0.0245, + "step": 4282 + }, + { + "epoch": 2.7508028259473347, + "grad_norm": 0.5863492654445509, + "learning_rate": 2.0880373790594043e-07, + "loss": 0.0222, + "step": 4283 + }, + { + "epoch": 2.7514450867052025, + "grad_norm": 0.7398499699427324, + "learning_rate": 2.0773632306809622e-07, + "loss": 0.0297, + "step": 4284 + }, + { + "epoch": 2.7520873474630703, + "grad_norm": 1.053746228256452, + "learning_rate": 2.0667158568198618e-07, + "loss": 0.0463, + "step": 4285 + }, + { + "epoch": 2.7527296082209376, + "grad_norm": 0.5471964334305265, + "learning_rate": 2.0560952634248231e-07, + "loss": 0.0171, + "step": 4286 + }, + { + "epoch": 2.7533718689788054, + "grad_norm": 0.6710181547290893, + "learning_rate": 2.045501456429605e-07, + "loss": 0.0199, + "step": 4287 + }, + { + "epoch": 2.754014129736673, + "grad_norm": 0.6341295126303701, + "learning_rate": 2.0349344417530004e-07, + "loss": 0.0209, + "step": 4288 + }, + { + "epoch": 2.754656390494541, + "grad_norm": 0.8818422640371842, + "learning_rate": 2.0243942252988203e-07, + "loss": 0.0337, + "step": 4289 + }, + { + "epoch": 2.7552986512524082, + "grad_norm": 0.592805015256855, + "learning_rate": 2.0138808129559261e-07, + "loss": 0.0193, + "step": 4290 + }, + { + "epoch": 2.755940912010276, + "grad_norm": 0.6251955320935363, + "learning_rate": 2.0033942105981918e-07, + "loss": 0.0187, + "step": 4291 + }, + { + "epoch": 2.756583172768144, + "grad_norm": 0.9359930528487022, + "learning_rate": 1.9929344240845027e-07, + "loss": 0.0192, + "step": 4292 + }, + { + "epoch": 2.7572254335260116, + "grad_norm": 0.7048778865692568, + "learning_rate": 1.9825014592587844e-07, + "loss": 0.0272, + "step": 4293 + }, + { + "epoch": 2.7578676942838793, + "grad_norm": 0.6288655298568974, + "learning_rate": 1.9720953219499516e-07, + "loss": 0.021, + "step": 4294 + }, + { + "epoch": 2.758509955041747, + "grad_norm": 0.7796490317144793, + "learning_rate": 1.9617160179719596e-07, + "loss": 0.0345, + "step": 4295 + }, + { + "epoch": 2.759152215799615, + "grad_norm": 0.7578074860257463, + "learning_rate": 1.9513635531237418e-07, + "loss": 0.0307, + "step": 4296 + }, + { + "epoch": 2.7597944765574822, + "grad_norm": 0.7986166880547875, + "learning_rate": 1.9410379331892547e-07, + "loss": 0.0353, + "step": 4297 + }, + { + "epoch": 2.76043673731535, + "grad_norm": 0.9526752344431306, + "learning_rate": 1.9307391639374616e-07, + "loss": 0.0417, + "step": 4298 + }, + { + "epoch": 2.7610789980732178, + "grad_norm": 0.9392198959727363, + "learning_rate": 1.9204672511223044e-07, + "loss": 0.0459, + "step": 4299 + }, + { + "epoch": 2.7617212588310855, + "grad_norm": 0.7222506750547831, + "learning_rate": 1.910222200482731e-07, + "loss": 0.0278, + "step": 4300 + }, + { + "epoch": 2.762363519588953, + "grad_norm": 0.5275216144541648, + "learning_rate": 1.9000040177426793e-07, + "loss": 0.0157, + "step": 4301 + }, + { + "epoch": 2.7630057803468207, + "grad_norm": 0.9685246338510003, + "learning_rate": 1.889812708611083e-07, + "loss": 0.039, + "step": 4302 + }, + { + "epoch": 2.7636480411046884, + "grad_norm": 0.9141442237031532, + "learning_rate": 1.8796482787818593e-07, + "loss": 0.0433, + "step": 4303 + }, + { + "epoch": 2.764290301862556, + "grad_norm": 0.811351425041696, + "learning_rate": 1.869510733933899e-07, + "loss": 0.0266, + "step": 4304 + }, + { + "epoch": 2.764932562620424, + "grad_norm": 0.7475058630964877, + "learning_rate": 1.859400079731083e-07, + "loss": 0.0308, + "step": 4305 + }, + { + "epoch": 2.7655748233782917, + "grad_norm": 0.8815247943698141, + "learning_rate": 1.8493163218222643e-07, + "loss": 0.0451, + "step": 4306 + }, + { + "epoch": 2.7662170841361595, + "grad_norm": 0.502824107539923, + "learning_rate": 1.8392594658412588e-07, + "loss": 0.0082, + "step": 4307 + }, + { + "epoch": 2.766859344894027, + "grad_norm": 1.365679025039031, + "learning_rate": 1.829229517406872e-07, + "loss": 0.083, + "step": 4308 + }, + { + "epoch": 2.7675016056518946, + "grad_norm": 0.557804457727747, + "learning_rate": 1.8192264821228544e-07, + "loss": 0.0168, + "step": 4309 + }, + { + "epoch": 2.7681438664097624, + "grad_norm": 0.8662255770363547, + "learning_rate": 1.8092503655779357e-07, + "loss": 0.0217, + "step": 4310 + }, + { + "epoch": 2.76878612716763, + "grad_norm": 0.7417084851467367, + "learning_rate": 1.7993011733458077e-07, + "loss": 0.0284, + "step": 4311 + }, + { + "epoch": 2.7694283879254975, + "grad_norm": 1.1567740009304583, + "learning_rate": 1.7893789109851069e-07, + "loss": 0.0372, + "step": 4312 + }, + { + "epoch": 2.7700706486833653, + "grad_norm": 0.9040917694858741, + "learning_rate": 1.779483584039432e-07, + "loss": 0.039, + "step": 4313 + }, + { + "epoch": 2.770712909441233, + "grad_norm": 0.6024970727235665, + "learning_rate": 1.7696151980373278e-07, + "loss": 0.0184, + "step": 4314 + }, + { + "epoch": 2.771355170199101, + "grad_norm": 0.8989696977793638, + "learning_rate": 1.7597737584922946e-07, + "loss": 0.0563, + "step": 4315 + }, + { + "epoch": 2.7719974309569686, + "grad_norm": 0.7047203553504782, + "learning_rate": 1.7499592709027679e-07, + "loss": 0.0329, + "step": 4316 + }, + { + "epoch": 2.7726396917148364, + "grad_norm": 0.5807493693452105, + "learning_rate": 1.740171740752139e-07, + "loss": 0.0276, + "step": 4317 + }, + { + "epoch": 2.773281952472704, + "grad_norm": 0.6969698241929231, + "learning_rate": 1.730411173508728e-07, + "loss": 0.0283, + "step": 4318 + }, + { + "epoch": 2.7739242132305715, + "grad_norm": 0.7648476240595594, + "learning_rate": 1.7206775746257898e-07, + "loss": 0.0284, + "step": 4319 + }, + { + "epoch": 2.7745664739884393, + "grad_norm": 0.862295271934136, + "learning_rate": 1.7109709495415073e-07, + "loss": 0.0289, + "step": 4320 + }, + { + "epoch": 2.775208734746307, + "grad_norm": 0.4837343224923704, + "learning_rate": 1.7012913036790145e-07, + "loss": 0.0136, + "step": 4321 + }, + { + "epoch": 2.775850995504175, + "grad_norm": 0.4278618127482307, + "learning_rate": 1.6916386424463527e-07, + "loss": 0.0157, + "step": 4322 + }, + { + "epoch": 2.776493256262042, + "grad_norm": 0.9088724862484867, + "learning_rate": 1.6820129712364852e-07, + "loss": 0.0419, + "step": 4323 + }, + { + "epoch": 2.77713551701991, + "grad_norm": 0.5466568735676143, + "learning_rate": 1.672414295427305e-07, + "loss": 0.0078, + "step": 4324 + }, + { + "epoch": 2.7777777777777777, + "grad_norm": 0.7490810923204136, + "learning_rate": 1.6628426203816273e-07, + "loss": 0.0138, + "step": 4325 + }, + { + "epoch": 2.7784200385356455, + "grad_norm": 0.8693791446589033, + "learning_rate": 1.6532979514471747e-07, + "loss": 0.0377, + "step": 4326 + }, + { + "epoch": 2.7790622992935132, + "grad_norm": 0.7367293578831005, + "learning_rate": 1.6437802939565705e-07, + "loss": 0.0398, + "step": 4327 + }, + { + "epoch": 2.779704560051381, + "grad_norm": 0.921680402175106, + "learning_rate": 1.6342896532273666e-07, + "loss": 0.0446, + "step": 4328 + }, + { + "epoch": 2.7803468208092488, + "grad_norm": 1.1944840718547334, + "learning_rate": 1.624826034562016e-07, + "loss": 0.066, + "step": 4329 + }, + { + "epoch": 2.780989081567116, + "grad_norm": 0.6449464014590781, + "learning_rate": 1.615389443247861e-07, + "loss": 0.0196, + "step": 4330 + }, + { + "epoch": 2.781631342324984, + "grad_norm": 0.5211934484946867, + "learning_rate": 1.6059798845571572e-07, + "loss": 0.0136, + "step": 4331 + }, + { + "epoch": 2.7822736030828517, + "grad_norm": 0.7597193144826918, + "learning_rate": 1.5965973637470489e-07, + "loss": 0.0277, + "step": 4332 + }, + { + "epoch": 2.7829158638407194, + "grad_norm": 0.6622308046416377, + "learning_rate": 1.5872418860595817e-07, + "loss": 0.0202, + "step": 4333 + }, + { + "epoch": 2.7835581245985868, + "grad_norm": 0.5004068818298342, + "learning_rate": 1.5779134567216857e-07, + "loss": 0.0146, + "step": 4334 + }, + { + "epoch": 2.7842003853564545, + "grad_norm": 0.8493857212447, + "learning_rate": 1.568612080945181e-07, + "loss": 0.0404, + "step": 4335 + }, + { + "epoch": 2.7848426461143223, + "grad_norm": 0.9883069640188202, + "learning_rate": 1.5593377639267715e-07, + "loss": 0.0455, + "step": 4336 + }, + { + "epoch": 2.78548490687219, + "grad_norm": 0.6198426231539965, + "learning_rate": 1.550090510848057e-07, + "loss": 0.0256, + "step": 4337 + }, + { + "epoch": 2.786127167630058, + "grad_norm": 0.43699315513527287, + "learning_rate": 1.5408703268754988e-07, + "loss": 0.0067, + "step": 4338 + }, + { + "epoch": 2.7867694283879256, + "grad_norm": 0.8068654059408836, + "learning_rate": 1.5316772171604378e-07, + "loss": 0.0393, + "step": 4339 + }, + { + "epoch": 2.7874116891457934, + "grad_norm": 0.8101176483033024, + "learning_rate": 1.5225111868390873e-07, + "loss": 0.0256, + "step": 4340 + }, + { + "epoch": 2.7880539499036607, + "grad_norm": 0.5078229631042466, + "learning_rate": 1.5133722410325458e-07, + "loss": 0.0103, + "step": 4341 + }, + { + "epoch": 2.7886962106615285, + "grad_norm": 0.9462782760891336, + "learning_rate": 1.5042603848467562e-07, + "loss": 0.037, + "step": 4342 + }, + { + "epoch": 2.7893384714193963, + "grad_norm": 0.601906615018365, + "learning_rate": 1.4951756233725522e-07, + "loss": 0.0205, + "step": 4343 + }, + { + "epoch": 2.789980732177264, + "grad_norm": 0.8400887667032821, + "learning_rate": 1.4861179616856126e-07, + "loss": 0.036, + "step": 4344 + }, + { + "epoch": 2.790622992935132, + "grad_norm": 0.6153071540006827, + "learning_rate": 1.4770874048464724e-07, + "loss": 0.0208, + "step": 4345 + }, + { + "epoch": 2.791265253692999, + "grad_norm": 0.5292448017808393, + "learning_rate": 1.4680839579005402e-07, + "loss": 0.0142, + "step": 4346 + }, + { + "epoch": 2.791907514450867, + "grad_norm": 0.9883010744847164, + "learning_rate": 1.459107625878059e-07, + "loss": 0.0302, + "step": 4347 + }, + { + "epoch": 2.7925497752087347, + "grad_norm": 0.8150553773449214, + "learning_rate": 1.4501584137941394e-07, + "loss": 0.0386, + "step": 4348 + }, + { + "epoch": 2.7931920359666025, + "grad_norm": 0.7306588086219973, + "learning_rate": 1.4412363266487318e-07, + "loss": 0.0242, + "step": 4349 + }, + { + "epoch": 2.7938342967244703, + "grad_norm": 0.9330712631890875, + "learning_rate": 1.4323413694266264e-07, + "loss": 0.0353, + "step": 4350 + }, + { + "epoch": 2.794476557482338, + "grad_norm": 0.8741911928350197, + "learning_rate": 1.4234735470974647e-07, + "loss": 0.0437, + "step": 4351 + }, + { + "epoch": 2.7951188182402054, + "grad_norm": 0.681987034728829, + "learning_rate": 1.4146328646157282e-07, + "loss": 0.0209, + "step": 4352 + }, + { + "epoch": 2.795761078998073, + "grad_norm": 0.9575816095954408, + "learning_rate": 1.4058193269207322e-07, + "loss": 0.0392, + "step": 4353 + }, + { + "epoch": 2.796403339755941, + "grad_norm": 0.7108331244017226, + "learning_rate": 1.3970329389366155e-07, + "loss": 0.0202, + "step": 4354 + }, + { + "epoch": 2.7970456005138087, + "grad_norm": 1.1597832381161175, + "learning_rate": 1.3882737055723737e-07, + "loss": 0.0472, + "step": 4355 + }, + { + "epoch": 2.7976878612716765, + "grad_norm": 0.8419207738794311, + "learning_rate": 1.3795416317218036e-07, + "loss": 0.0318, + "step": 4356 + }, + { + "epoch": 2.798330122029544, + "grad_norm": 0.7302216660131823, + "learning_rate": 1.370836722263541e-07, + "loss": 0.0299, + "step": 4357 + }, + { + "epoch": 2.7989723827874116, + "grad_norm": 0.6903866699260395, + "learning_rate": 1.362158982061057e-07, + "loss": 0.03, + "step": 4358 + }, + { + "epoch": 2.7996146435452793, + "grad_norm": 0.6223770923511173, + "learning_rate": 1.3535084159626176e-07, + "loss": 0.0291, + "step": 4359 + }, + { + "epoch": 2.800256904303147, + "grad_norm": 0.619847592352592, + "learning_rate": 1.3448850288013293e-07, + "loss": 0.0177, + "step": 4360 + }, + { + "epoch": 2.800899165061015, + "grad_norm": 0.7195824077586761, + "learning_rate": 1.336288825395099e-07, + "loss": 0.0377, + "step": 4361 + }, + { + "epoch": 2.8015414258188827, + "grad_norm": 1.1186768811223433, + "learning_rate": 1.3277198105466515e-07, + "loss": 0.0487, + "step": 4362 + }, + { + "epoch": 2.8021836865767504, + "grad_norm": 0.5760062834790133, + "learning_rate": 1.319177989043513e-07, + "loss": 0.0154, + "step": 4363 + }, + { + "epoch": 2.8028259473346178, + "grad_norm": 0.9839624708335346, + "learning_rate": 1.3106633656580326e-07, + "loss": 0.037, + "step": 4364 + }, + { + "epoch": 2.8034682080924855, + "grad_norm": 0.6674917665984413, + "learning_rate": 1.3021759451473548e-07, + "loss": 0.0271, + "step": 4365 + }, + { + "epoch": 2.8041104688503533, + "grad_norm": 1.2444523541593824, + "learning_rate": 1.2937157322534254e-07, + "loss": 0.0293, + "step": 4366 + }, + { + "epoch": 2.804752729608221, + "grad_norm": 0.8623136167472075, + "learning_rate": 1.2852827317029858e-07, + "loss": 0.0308, + "step": 4367 + }, + { + "epoch": 2.8053949903660884, + "grad_norm": 1.1970566821870545, + "learning_rate": 1.2768769482075838e-07, + "loss": 0.0602, + "step": 4368 + }, + { + "epoch": 2.806037251123956, + "grad_norm": 0.8762414083737403, + "learning_rate": 1.2684983864635514e-07, + "loss": 0.0437, + "step": 4369 + }, + { + "epoch": 2.806679511881824, + "grad_norm": 0.883935580387172, + "learning_rate": 1.2601470511520164e-07, + "loss": 0.0466, + "step": 4370 + }, + { + "epoch": 2.8073217726396917, + "grad_norm": 0.6992289745336197, + "learning_rate": 1.2518229469388965e-07, + "loss": 0.0317, + "step": 4371 + }, + { + "epoch": 2.8079640333975595, + "grad_norm": 0.6974279333704582, + "learning_rate": 1.2435260784748882e-07, + "loss": 0.0287, + "step": 4372 + }, + { + "epoch": 2.8086062941554273, + "grad_norm": 0.926159843985108, + "learning_rate": 1.2352564503954834e-07, + "loss": 0.0401, + "step": 4373 + }, + { + "epoch": 2.809248554913295, + "grad_norm": 0.6434366845716842, + "learning_rate": 1.2270140673209473e-07, + "loss": 0.018, + "step": 4374 + }, + { + "epoch": 2.8098908156711624, + "grad_norm": 0.7899370709138447, + "learning_rate": 1.2187989338563132e-07, + "loss": 0.0329, + "step": 4375 + }, + { + "epoch": 2.81053307642903, + "grad_norm": 1.0516152262782834, + "learning_rate": 1.2106110545914097e-07, + "loss": 0.0269, + "step": 4376 + }, + { + "epoch": 2.811175337186898, + "grad_norm": 0.8289732716398105, + "learning_rate": 1.202450434100827e-07, + "loss": 0.0324, + "step": 4377 + }, + { + "epoch": 2.8118175979447657, + "grad_norm": 0.7194941989099195, + "learning_rate": 1.1943170769439237e-07, + "loss": 0.0231, + "step": 4378 + }, + { + "epoch": 2.812459858702633, + "grad_norm": 0.9558250096793804, + "learning_rate": 1.186210987664832e-07, + "loss": 0.0439, + "step": 4379 + }, + { + "epoch": 2.813102119460501, + "grad_norm": 0.9611439663267745, + "learning_rate": 1.1781321707924508e-07, + "loss": 0.0252, + "step": 4380 + }, + { + "epoch": 2.8137443802183686, + "grad_norm": 0.5231928067593267, + "learning_rate": 1.1700806308404367e-07, + "loss": 0.0184, + "step": 4381 + }, + { + "epoch": 2.8143866409762364, + "grad_norm": 0.6329236435452874, + "learning_rate": 1.1620563723072131e-07, + "loss": 0.0229, + "step": 4382 + }, + { + "epoch": 2.815028901734104, + "grad_norm": 0.8420406739774751, + "learning_rate": 1.1540593996759441e-07, + "loss": 0.0452, + "step": 4383 + }, + { + "epoch": 2.815671162491972, + "grad_norm": 0.632834618269384, + "learning_rate": 1.146089717414578e-07, + "loss": 0.0269, + "step": 4384 + }, + { + "epoch": 2.8163134232498397, + "grad_norm": 0.8567712330055673, + "learning_rate": 1.1381473299757917e-07, + "loss": 0.0381, + "step": 4385 + }, + { + "epoch": 2.816955684007707, + "grad_norm": 0.6965103472956744, + "learning_rate": 1.1302322417970135e-07, + "loss": 0.0223, + "step": 4386 + }, + { + "epoch": 2.817597944765575, + "grad_norm": 0.9230618064029708, + "learning_rate": 1.1223444573004339e-07, + "loss": 0.0465, + "step": 4387 + }, + { + "epoch": 2.8182402055234426, + "grad_norm": 0.9956785176263, + "learning_rate": 1.1144839808929831e-07, + "loss": 0.0419, + "step": 4388 + }, + { + "epoch": 2.8188824662813103, + "grad_norm": 0.7711968365131255, + "learning_rate": 1.1066508169663315e-07, + "loss": 0.0244, + "step": 4389 + }, + { + "epoch": 2.8195247270391777, + "grad_norm": 0.6060340386247052, + "learning_rate": 1.0988449698968839e-07, + "loss": 0.0136, + "step": 4390 + }, + { + "epoch": 2.8201669877970454, + "grad_norm": 0.8471435533680078, + "learning_rate": 1.0910664440457907e-07, + "loss": 0.0288, + "step": 4391 + }, + { + "epoch": 2.820809248554913, + "grad_norm": 0.9097821655334047, + "learning_rate": 1.0833152437589423e-07, + "loss": 0.0514, + "step": 4392 + }, + { + "epoch": 2.821451509312781, + "grad_norm": 0.475815375012486, + "learning_rate": 1.0755913733669632e-07, + "loss": 0.017, + "step": 4393 + }, + { + "epoch": 2.8220937700706488, + "grad_norm": 0.5898522314593166, + "learning_rate": 1.0678948371851905e-07, + "loss": 0.0241, + "step": 4394 + }, + { + "epoch": 2.8227360308285165, + "grad_norm": 0.8076852963749673, + "learning_rate": 1.0602256395137068e-07, + "loss": 0.0301, + "step": 4395 + }, + { + "epoch": 2.8233782915863843, + "grad_norm": 0.9970187157164868, + "learning_rate": 1.0525837846373121e-07, + "loss": 0.0247, + "step": 4396 + }, + { + "epoch": 2.8240205523442516, + "grad_norm": 0.9792313568782284, + "learning_rate": 1.044969276825536e-07, + "loss": 0.0489, + "step": 4397 + }, + { + "epoch": 2.8246628131021194, + "grad_norm": 0.7798478365066376, + "learning_rate": 1.0373821203326251e-07, + "loss": 0.0273, + "step": 4398 + }, + { + "epoch": 2.825305073859987, + "grad_norm": 0.8047115676209899, + "learning_rate": 1.0298223193975499e-07, + "loss": 0.0422, + "step": 4399 + }, + { + "epoch": 2.825947334617855, + "grad_norm": 0.40351580498250883, + "learning_rate": 1.0222898782439872e-07, + "loss": 0.013, + "step": 4400 + }, + { + "epoch": 2.8265895953757223, + "grad_norm": 0.6956223049169935, + "learning_rate": 1.0147848010803319e-07, + "loss": 0.0286, + "step": 4401 + }, + { + "epoch": 2.82723185613359, + "grad_norm": 0.9777304026507073, + "learning_rate": 1.0073070920997018e-07, + "loss": 0.0414, + "step": 4402 + }, + { + "epoch": 2.827874116891458, + "grad_norm": 0.7597117079609784, + "learning_rate": 9.998567554799111e-08, + "loss": 0.0286, + "step": 4403 + }, + { + "epoch": 2.8285163776493256, + "grad_norm": 0.8860474398673087, + "learning_rate": 9.924337953834795e-08, + "loss": 0.0423, + "step": 4404 + }, + { + "epoch": 2.8291586384071934, + "grad_norm": 0.5990758902406836, + "learning_rate": 9.850382159576455e-08, + "loss": 0.0153, + "step": 4405 + }, + { + "epoch": 2.829800899165061, + "grad_norm": 0.6840339669055346, + "learning_rate": 9.776700213343315e-08, + "loss": 0.0253, + "step": 4406 + }, + { + "epoch": 2.830443159922929, + "grad_norm": 0.9954746641998881, + "learning_rate": 9.70329215630178e-08, + "loss": 0.0477, + "step": 4407 + }, + { + "epoch": 2.8310854206807963, + "grad_norm": 1.0548191561673959, + "learning_rate": 9.630158029465098e-08, + "loss": 0.041, + "step": 4408 + }, + { + "epoch": 2.831727681438664, + "grad_norm": 0.9822182289564342, + "learning_rate": 9.557297873693528e-08, + "loss": 0.042, + "step": 4409 + }, + { + "epoch": 2.832369942196532, + "grad_norm": 0.7810182202175423, + "learning_rate": 9.484711729694229e-08, + "loss": 0.0253, + "step": 4410 + }, + { + "epoch": 2.8330122029543996, + "grad_norm": 0.8197459238577111, + "learning_rate": 9.412399638021319e-08, + "loss": 0.0309, + "step": 4411 + }, + { + "epoch": 2.833654463712267, + "grad_norm": 0.7489165281772537, + "learning_rate": 9.340361639075701e-08, + "loss": 0.0319, + "step": 4412 + }, + { + "epoch": 2.8342967244701347, + "grad_norm": 1.0838427485461062, + "learning_rate": 9.268597773105347e-08, + "loss": 0.037, + "step": 4413 + }, + { + "epoch": 2.8349389852280025, + "grad_norm": 0.8438368011951318, + "learning_rate": 9.197108080204909e-08, + "loss": 0.0334, + "step": 4414 + }, + { + "epoch": 2.8355812459858702, + "grad_norm": 1.2471885746895182, + "learning_rate": 9.125892600315822e-08, + "loss": 0.0506, + "step": 4415 + }, + { + "epoch": 2.836223506743738, + "grad_norm": 0.3549841302434878, + "learning_rate": 9.054951373226483e-08, + "loss": 0.0113, + "step": 4416 + }, + { + "epoch": 2.836865767501606, + "grad_norm": 0.5358714010200176, + "learning_rate": 8.984284438571855e-08, + "loss": 0.0198, + "step": 4417 + }, + { + "epoch": 2.8375080282594736, + "grad_norm": 0.6535018845829633, + "learning_rate": 8.913891835833743e-08, + "loss": 0.022, + "step": 4418 + }, + { + "epoch": 2.838150289017341, + "grad_norm": 1.234411525600897, + "learning_rate": 8.8437736043408e-08, + "loss": 0.0741, + "step": 4419 + }, + { + "epoch": 2.8387925497752087, + "grad_norm": 0.8182580344586678, + "learning_rate": 8.773929783268297e-08, + "loss": 0.0316, + "step": 4420 + }, + { + "epoch": 2.8394348105330764, + "grad_norm": 1.3538240989977046, + "learning_rate": 8.704360411638079e-08, + "loss": 0.0445, + "step": 4421 + }, + { + "epoch": 2.840077071290944, + "grad_norm": 0.8327552433441529, + "learning_rate": 8.635065528318775e-08, + "loss": 0.0282, + "step": 4422 + }, + { + "epoch": 2.8407193320488116, + "grad_norm": 0.8520980007758189, + "learning_rate": 8.56604517202575e-08, + "loss": 0.024, + "step": 4423 + }, + { + "epoch": 2.8413615928066793, + "grad_norm": 0.9484318775092252, + "learning_rate": 8.497299381320767e-08, + "loss": 0.0441, + "step": 4424 + }, + { + "epoch": 2.842003853564547, + "grad_norm": 0.5150472362548703, + "learning_rate": 8.428828194612326e-08, + "loss": 0.0147, + "step": 4425 + }, + { + "epoch": 2.842646114322415, + "grad_norm": 0.7387271920297889, + "learning_rate": 8.360631650155493e-08, + "loss": 0.0262, + "step": 4426 + }, + { + "epoch": 2.8432883750802826, + "grad_norm": 0.5706007318624632, + "learning_rate": 8.292709786051845e-08, + "loss": 0.0123, + "step": 4427 + }, + { + "epoch": 2.8439306358381504, + "grad_norm": 0.8902244207830864, + "learning_rate": 8.225062640249636e-08, + "loss": 0.0297, + "step": 4428 + }, + { + "epoch": 2.844572896596018, + "grad_norm": 0.6941435328278193, + "learning_rate": 8.157690250543359e-08, + "loss": 0.0264, + "step": 4429 + }, + { + "epoch": 2.8452151573538855, + "grad_norm": 0.8286721988233076, + "learning_rate": 8.090592654574292e-08, + "loss": 0.037, + "step": 4430 + }, + { + "epoch": 2.8458574181117533, + "grad_norm": 0.5129366711174229, + "learning_rate": 8.023769889830002e-08, + "loss": 0.0153, + "step": 4431 + }, + { + "epoch": 2.846499678869621, + "grad_norm": 1.0348740918145178, + "learning_rate": 7.957221993644625e-08, + "loss": 0.0458, + "step": 4432 + }, + { + "epoch": 2.847141939627489, + "grad_norm": 0.6727727166411258, + "learning_rate": 7.89094900319859e-08, + "loss": 0.0198, + "step": 4433 + }, + { + "epoch": 2.847784200385356, + "grad_norm": 0.945211607500618, + "learning_rate": 7.824950955518773e-08, + "loss": 0.0409, + "step": 4434 + }, + { + "epoch": 2.848426461143224, + "grad_norm": 0.922343142715418, + "learning_rate": 7.75922788747857e-08, + "loss": 0.0521, + "step": 4435 + }, + { + "epoch": 2.8490687219010917, + "grad_norm": 0.6464783996941262, + "learning_rate": 7.69377983579761e-08, + "loss": 0.0231, + "step": 4436 + }, + { + "epoch": 2.8497109826589595, + "grad_norm": 0.7525953128043613, + "learning_rate": 7.628606837041974e-08, + "loss": 0.0289, + "step": 4437 + }, + { + "epoch": 2.8503532434168273, + "grad_norm": 0.8273055932653545, + "learning_rate": 7.563708927623925e-08, + "loss": 0.0336, + "step": 4438 + }, + { + "epoch": 2.850995504174695, + "grad_norm": 0.5635529049936503, + "learning_rate": 7.499086143802126e-08, + "loss": 0.0125, + "step": 4439 + }, + { + "epoch": 2.851637764932563, + "grad_norm": 0.7820028090493949, + "learning_rate": 7.434738521681584e-08, + "loss": 0.0271, + "step": 4440 + }, + { + "epoch": 2.85228002569043, + "grad_norm": 0.8757273110299298, + "learning_rate": 7.37066609721343e-08, + "loss": 0.0395, + "step": 4441 + }, + { + "epoch": 2.852922286448298, + "grad_norm": 0.820920813580508, + "learning_rate": 7.306868906195141e-08, + "loss": 0.0283, + "step": 4442 + }, + { + "epoch": 2.8535645472061657, + "grad_norm": 0.5643818651734639, + "learning_rate": 7.243346984270371e-08, + "loss": 0.0208, + "step": 4443 + }, + { + "epoch": 2.8542068079640335, + "grad_norm": 0.805007525657461, + "learning_rate": 7.180100366929066e-08, + "loss": 0.0294, + "step": 4444 + }, + { + "epoch": 2.8548490687219013, + "grad_norm": 0.8250663814435507, + "learning_rate": 7.117129089507236e-08, + "loss": 0.032, + "step": 4445 + }, + { + "epoch": 2.8554913294797686, + "grad_norm": 0.9001628480127161, + "learning_rate": 7.05443318718707e-08, + "loss": 0.0485, + "step": 4446 + }, + { + "epoch": 2.8561335902376364, + "grad_norm": 0.7462959681908365, + "learning_rate": 6.992012694997108e-08, + "loss": 0.0217, + "step": 4447 + }, + { + "epoch": 2.856775850995504, + "grad_norm": 1.022562299412395, + "learning_rate": 6.929867647811784e-08, + "loss": 0.0469, + "step": 4448 + }, + { + "epoch": 2.857418111753372, + "grad_norm": 0.8704556153970883, + "learning_rate": 6.867998080351656e-08, + "loss": 0.0286, + "step": 4449 + }, + { + "epoch": 2.8580603725112397, + "grad_norm": 0.8189360810658204, + "learning_rate": 6.806404027183466e-08, + "loss": 0.034, + "step": 4450 + }, + { + "epoch": 2.8587026332691075, + "grad_norm": 0.8802198826897267, + "learning_rate": 6.745085522720074e-08, + "loss": 0.0317, + "step": 4451 + }, + { + "epoch": 2.8593448940269752, + "grad_norm": 0.6648831985410436, + "learning_rate": 6.684042601220186e-08, + "loss": 0.0192, + "step": 4452 + }, + { + "epoch": 2.8599871547848426, + "grad_norm": 0.8596143130501622, + "learning_rate": 6.623275296788744e-08, + "loss": 0.026, + "step": 4453 + }, + { + "epoch": 2.8606294155427103, + "grad_norm": 0.793642250051679, + "learning_rate": 6.562783643376536e-08, + "loss": 0.0283, + "step": 4454 + }, + { + "epoch": 2.861271676300578, + "grad_norm": 0.7285416517470763, + "learning_rate": 6.502567674780524e-08, + "loss": 0.0355, + "step": 4455 + }, + { + "epoch": 2.861913937058446, + "grad_norm": 0.6485319387984496, + "learning_rate": 6.442627424643522e-08, + "loss": 0.0301, + "step": 4456 + }, + { + "epoch": 2.862556197816313, + "grad_norm": 0.7731816579367461, + "learning_rate": 6.382962926454295e-08, + "loss": 0.0255, + "step": 4457 + }, + { + "epoch": 2.863198458574181, + "grad_norm": 0.634290448158889, + "learning_rate": 6.323574213547567e-08, + "loss": 0.0165, + "step": 4458 + }, + { + "epoch": 2.8638407193320488, + "grad_norm": 0.6038519696740825, + "learning_rate": 6.264461319104021e-08, + "loss": 0.0259, + "step": 4459 + }, + { + "epoch": 2.8644829800899165, + "grad_norm": 0.6168375254554791, + "learning_rate": 6.20562427615018e-08, + "loss": 0.0191, + "step": 4460 + }, + { + "epoch": 2.8651252408477843, + "grad_norm": 0.738270780367607, + "learning_rate": 6.147063117558472e-08, + "loss": 0.0216, + "step": 4461 + }, + { + "epoch": 2.865767501605652, + "grad_norm": 0.539729554396391, + "learning_rate": 6.08877787604717e-08, + "loss": 0.0238, + "step": 4462 + }, + { + "epoch": 2.86640976236352, + "grad_norm": 0.8155269519957764, + "learning_rate": 6.030768584180446e-08, + "loss": 0.0276, + "step": 4463 + }, + { + "epoch": 2.867052023121387, + "grad_norm": 0.8675527045565018, + "learning_rate": 5.973035274368266e-08, + "loss": 0.0349, + "step": 4464 + }, + { + "epoch": 2.867694283879255, + "grad_norm": 1.0084266543315716, + "learning_rate": 5.915577978866327e-08, + "loss": 0.0412, + "step": 4465 + }, + { + "epoch": 2.8683365446371227, + "grad_norm": 0.7528778973561164, + "learning_rate": 5.8583967297762835e-08, + "loss": 0.0319, + "step": 4466 + }, + { + "epoch": 2.8689788053949905, + "grad_norm": 0.5503179285566616, + "learning_rate": 5.801491559045358e-08, + "loss": 0.0167, + "step": 4467 + }, + { + "epoch": 2.869621066152858, + "grad_norm": 0.9299076557479748, + "learning_rate": 5.7448624984667853e-08, + "loss": 0.0387, + "step": 4468 + }, + { + "epoch": 2.8702633269107256, + "grad_norm": 0.6127875305421774, + "learning_rate": 5.6885095796792574e-08, + "loss": 0.0178, + "step": 4469 + }, + { + "epoch": 2.8709055876685934, + "grad_norm": 0.9019223801584363, + "learning_rate": 5.6324328341673676e-08, + "loss": 0.0359, + "step": 4470 + }, + { + "epoch": 2.871547848426461, + "grad_norm": 0.609965629917331, + "learning_rate": 5.576632293261386e-08, + "loss": 0.027, + "step": 4471 + }, + { + "epoch": 2.872190109184329, + "grad_norm": 0.862807444829033, + "learning_rate": 5.521107988137153e-08, + "loss": 0.0253, + "step": 4472 + }, + { + "epoch": 2.8728323699421967, + "grad_norm": 0.9383855573240536, + "learning_rate": 5.465859949816299e-08, + "loss": 0.0381, + "step": 4473 + }, + { + "epoch": 2.8734746307000645, + "grad_norm": 1.139035501851501, + "learning_rate": 5.410888209166132e-08, + "loss": 0.028, + "step": 4474 + }, + { + "epoch": 2.874116891457932, + "grad_norm": 0.9072247579624596, + "learning_rate": 5.356192796899418e-08, + "loss": 0.0422, + "step": 4475 + }, + { + "epoch": 2.8747591522157996, + "grad_norm": 0.7054296116160877, + "learning_rate": 5.301773743574712e-08, + "loss": 0.0221, + "step": 4476 + }, + { + "epoch": 2.8754014129736674, + "grad_norm": 1.0173112215980238, + "learning_rate": 5.247631079596083e-08, + "loss": 0.0472, + "step": 4477 + }, + { + "epoch": 2.876043673731535, + "grad_norm": 1.000418349545813, + "learning_rate": 5.1937648352131665e-08, + "loss": 0.0417, + "step": 4478 + }, + { + "epoch": 2.8766859344894025, + "grad_norm": 0.9270856587175371, + "learning_rate": 5.140175040521222e-08, + "loss": 0.0416, + "step": 4479 + }, + { + "epoch": 2.8773281952472702, + "grad_norm": 0.9411875668117935, + "learning_rate": 5.086861725461023e-08, + "loss": 0.053, + "step": 4480 + }, + { + "epoch": 2.877970456005138, + "grad_norm": 0.7139684941953323, + "learning_rate": 5.0338249198188525e-08, + "loss": 0.0259, + "step": 4481 + }, + { + "epoch": 2.878612716763006, + "grad_norm": 0.8281165353018706, + "learning_rate": 4.981064653226564e-08, + "loss": 0.0229, + "step": 4482 + }, + { + "epoch": 2.8792549775208736, + "grad_norm": 0.917092288631559, + "learning_rate": 4.928580955161466e-08, + "loss": 0.0386, + "step": 4483 + }, + { + "epoch": 2.8798972382787413, + "grad_norm": 0.6250877333919185, + "learning_rate": 4.876373854946326e-08, + "loss": 0.0118, + "step": 4484 + }, + { + "epoch": 2.880539499036609, + "grad_norm": 0.7639497000224732, + "learning_rate": 4.824443381749477e-08, + "loss": 0.0222, + "step": 4485 + }, + { + "epoch": 2.8811817597944764, + "grad_norm": 1.104369717709035, + "learning_rate": 4.77278956458449e-08, + "loss": 0.0604, + "step": 4486 + }, + { + "epoch": 2.881824020552344, + "grad_norm": 0.8009281921247153, + "learning_rate": 4.721412432310668e-08, + "loss": 0.0333, + "step": 4487 + }, + { + "epoch": 2.882466281310212, + "grad_norm": 0.8331967969702297, + "learning_rate": 4.6703120136324384e-08, + "loss": 0.0254, + "step": 4488 + }, + { + "epoch": 2.8831085420680798, + "grad_norm": 0.4871817105161949, + "learning_rate": 4.6194883370997423e-08, + "loss": 0.0112, + "step": 4489 + }, + { + "epoch": 2.883750802825947, + "grad_norm": 0.9149814430238759, + "learning_rate": 4.5689414311080314e-08, + "loss": 0.0261, + "step": 4490 + }, + { + "epoch": 2.884393063583815, + "grad_norm": 0.42946176620154103, + "learning_rate": 4.5186713238979385e-08, + "loss": 0.01, + "step": 4491 + }, + { + "epoch": 2.8850353243416826, + "grad_norm": 0.8349261745500517, + "learning_rate": 4.468678043555497e-08, + "loss": 0.0356, + "step": 4492 + }, + { + "epoch": 2.8856775850995504, + "grad_norm": 0.8252764312985316, + "learning_rate": 4.4189616180121964e-08, + "loss": 0.0211, + "step": 4493 + }, + { + "epoch": 2.886319845857418, + "grad_norm": 0.6807422776414146, + "learning_rate": 4.369522075044652e-08, + "loss": 0.0287, + "step": 4494 + }, + { + "epoch": 2.886962106615286, + "grad_norm": 0.9568544240204855, + "learning_rate": 4.320359442274935e-08, + "loss": 0.0363, + "step": 4495 + }, + { + "epoch": 2.8876043673731537, + "grad_norm": 0.9086826118006427, + "learning_rate": 4.271473747170296e-08, + "loss": 0.0233, + "step": 4496 + }, + { + "epoch": 2.888246628131021, + "grad_norm": 0.5293341100981568, + "learning_rate": 4.222865017043387e-08, + "loss": 0.0131, + "step": 4497 + }, + { + "epoch": 2.888888888888889, + "grad_norm": 0.8105672374738198, + "learning_rate": 4.174533279051984e-08, + "loss": 0.0258, + "step": 4498 + }, + { + "epoch": 2.8895311496467566, + "grad_norm": 0.9263711362080506, + "learning_rate": 4.12647856019921e-08, + "loss": 0.0405, + "step": 4499 + }, + { + "epoch": 2.8901734104046244, + "grad_norm": 0.9513401315427319, + "learning_rate": 4.078700887333365e-08, + "loss": 0.0406, + "step": 4500 + }, + { + "epoch": 2.8908156711624917, + "grad_norm": 0.5565969199584224, + "learning_rate": 4.031200287147985e-08, + "loss": 0.0173, + "step": 4501 + }, + { + "epoch": 2.8914579319203595, + "grad_norm": 0.9103296569746663, + "learning_rate": 3.983976786181787e-08, + "loss": 0.0312, + "step": 4502 + }, + { + "epoch": 2.8921001926782273, + "grad_norm": 0.6116931208383539, + "learning_rate": 3.9370304108186653e-08, + "loss": 0.0271, + "step": 4503 + }, + { + "epoch": 2.892742453436095, + "grad_norm": 0.8919057512686256, + "learning_rate": 3.890361187287694e-08, + "loss": 0.0354, + "step": 4504 + }, + { + "epoch": 2.893384714193963, + "grad_norm": 0.6908063608379423, + "learning_rate": 3.843969141663128e-08, + "loss": 0.0153, + "step": 4505 + }, + { + "epoch": 2.8940269749518306, + "grad_norm": 0.7388663016532157, + "learning_rate": 3.797854299864345e-08, + "loss": 0.0252, + "step": 4506 + }, + { + "epoch": 2.8946692357096984, + "grad_norm": 0.6268367751458329, + "learning_rate": 3.752016687655846e-08, + "loss": 0.024, + "step": 4507 + }, + { + "epoch": 2.8953114964675657, + "grad_norm": 0.7475968081360125, + "learning_rate": 3.706456330647201e-08, + "loss": 0.0282, + "step": 4508 + }, + { + "epoch": 2.8959537572254335, + "grad_norm": 1.002746772123556, + "learning_rate": 3.6611732542931044e-08, + "loss": 0.0445, + "step": 4509 + }, + { + "epoch": 2.8965960179833012, + "grad_norm": 0.8326140219530062, + "learning_rate": 3.6161674838933736e-08, + "loss": 0.0355, + "step": 4510 + }, + { + "epoch": 2.897238278741169, + "grad_norm": 0.6488544744870905, + "learning_rate": 3.571439044592895e-08, + "loss": 0.025, + "step": 4511 + }, + { + "epoch": 2.8978805394990363, + "grad_norm": 1.0372182076349925, + "learning_rate": 3.526987961381567e-08, + "loss": 0.0395, + "step": 4512 + }, + { + "epoch": 2.898522800256904, + "grad_norm": 0.5704108339206875, + "learning_rate": 3.482814259094247e-08, + "loss": 0.0165, + "step": 4513 + }, + { + "epoch": 2.899165061014772, + "grad_norm": 0.9020215289594655, + "learning_rate": 3.438917962411026e-08, + "loss": 0.0392, + "step": 4514 + }, + { + "epoch": 2.8998073217726397, + "grad_norm": 0.5231957799025642, + "learning_rate": 3.3952990958568965e-08, + "loss": 0.0211, + "step": 4515 + }, + { + "epoch": 2.9004495825305074, + "grad_norm": 0.6924379505010684, + "learning_rate": 3.351957683801754e-08, + "loss": 0.0197, + "step": 4516 + }, + { + "epoch": 2.901091843288375, + "grad_norm": 0.7410798470917984, + "learning_rate": 3.3088937504606176e-08, + "loss": 0.0298, + "step": 4517 + }, + { + "epoch": 2.901734104046243, + "grad_norm": 0.685924201880083, + "learning_rate": 3.266107319893463e-08, + "loss": 0.0198, + "step": 4518 + }, + { + "epoch": 2.9023763648041103, + "grad_norm": 1.1587659450518384, + "learning_rate": 3.223598416005169e-08, + "loss": 0.0559, + "step": 4519 + }, + { + "epoch": 2.903018625561978, + "grad_norm": 2.0652548809361586, + "learning_rate": 3.1813670625455154e-08, + "loss": 0.0472, + "step": 4520 + }, + { + "epoch": 2.903660886319846, + "grad_norm": 0.9638739528949445, + "learning_rate": 3.139413283109405e-08, + "loss": 0.036, + "step": 4521 + }, + { + "epoch": 2.9043031470777136, + "grad_norm": 0.7151453423389489, + "learning_rate": 3.0977371011364776e-08, + "loss": 0.0329, + "step": 4522 + }, + { + "epoch": 2.904945407835581, + "grad_norm": 0.8509686233274443, + "learning_rate": 3.0563385399113296e-08, + "loss": 0.0309, + "step": 4523 + }, + { + "epoch": 2.9055876685934487, + "grad_norm": 0.6000753756886741, + "learning_rate": 3.01521762256346e-08, + "loss": 0.012, + "step": 4524 + }, + { + "epoch": 2.9062299293513165, + "grad_norm": 0.5177948825396056, + "learning_rate": 2.974374372067268e-08, + "loss": 0.0165, + "step": 4525 + }, + { + "epoch": 2.9068721901091843, + "grad_norm": 0.7594968723959976, + "learning_rate": 2.933808811242056e-08, + "loss": 0.0304, + "step": 4526 + }, + { + "epoch": 2.907514450867052, + "grad_norm": 0.5510342655918317, + "learning_rate": 2.89352096275175e-08, + "loss": 0.0199, + "step": 4527 + }, + { + "epoch": 2.90815671162492, + "grad_norm": 0.8285384248454047, + "learning_rate": 2.8535108491054565e-08, + "loss": 0.0247, + "step": 4528 + }, + { + "epoch": 2.9087989723827876, + "grad_norm": 0.8583219654413182, + "learning_rate": 2.813778492656849e-08, + "loss": 0.0355, + "step": 4529 + }, + { + "epoch": 2.909441233140655, + "grad_norm": 0.781180349608101, + "learning_rate": 2.7743239156045044e-08, + "loss": 0.0455, + "step": 4530 + }, + { + "epoch": 2.9100834938985227, + "grad_norm": 0.720227524936552, + "learning_rate": 2.7351471399919005e-08, + "loss": 0.0225, + "step": 4531 + }, + { + "epoch": 2.9107257546563905, + "grad_norm": 0.8250886517678525, + "learning_rate": 2.69624818770714e-08, + "loss": 0.0216, + "step": 4532 + }, + { + "epoch": 2.9113680154142583, + "grad_norm": 0.700329577189696, + "learning_rate": 2.6576270804831717e-08, + "loss": 0.0294, + "step": 4533 + }, + { + "epoch": 2.912010276172126, + "grad_norm": 0.9821599003745642, + "learning_rate": 2.6192838398977906e-08, + "loss": 0.0419, + "step": 4534 + }, + { + "epoch": 2.9126525369299934, + "grad_norm": 0.6465129399272227, + "learning_rate": 2.5812184873734158e-08, + "loss": 0.0102, + "step": 4535 + }, + { + "epoch": 2.913294797687861, + "grad_norm": 0.7391777298619374, + "learning_rate": 2.5434310441773135e-08, + "loss": 0.0272, + "step": 4536 + }, + { + "epoch": 2.913937058445729, + "grad_norm": 0.6516141095377758, + "learning_rate": 2.5059215314214292e-08, + "loss": 0.0297, + "step": 4537 + }, + { + "epoch": 2.9145793192035967, + "grad_norm": 0.7591387767841643, + "learning_rate": 2.468689970062388e-08, + "loss": 0.0265, + "step": 4538 + }, + { + "epoch": 2.9152215799614645, + "grad_norm": 0.6755817003517601, + "learning_rate": 2.4317363809016615e-08, + "loss": 0.0166, + "step": 4539 + }, + { + "epoch": 2.9158638407193322, + "grad_norm": 0.8036978112414912, + "learning_rate": 2.3950607845852346e-08, + "loss": 0.0381, + "step": 4540 + }, + { + "epoch": 2.9165061014772, + "grad_norm": 0.8896591207510949, + "learning_rate": 2.358663201603939e-08, + "loss": 0.0369, + "step": 4541 + }, + { + "epoch": 2.9171483622350673, + "grad_norm": 1.1167490993329108, + "learning_rate": 2.3225436522931743e-08, + "loss": 0.0314, + "step": 4542 + }, + { + "epoch": 2.917790622992935, + "grad_norm": 0.5119263370139258, + "learning_rate": 2.2867021568330205e-08, + "loss": 0.011, + "step": 4543 + }, + { + "epoch": 2.918432883750803, + "grad_norm": 0.7203925565020394, + "learning_rate": 2.2511387352482927e-08, + "loss": 0.0236, + "step": 4544 + }, + { + "epoch": 2.9190751445086707, + "grad_norm": 0.9082952441297333, + "learning_rate": 2.2158534074083193e-08, + "loss": 0.0324, + "step": 4545 + }, + { + "epoch": 2.919717405266538, + "grad_norm": 0.9127789262900279, + "learning_rate": 2.180846193027164e-08, + "loss": 0.0411, + "step": 4546 + }, + { + "epoch": 2.9203596660244058, + "grad_norm": 0.767685562984976, + "learning_rate": 2.146117111663404e-08, + "loss": 0.0286, + "step": 4547 + }, + { + "epoch": 2.9210019267822736, + "grad_norm": 1.0874508861499814, + "learning_rate": 2.1116661827202954e-08, + "loss": 0.0535, + "step": 4548 + }, + { + "epoch": 2.9216441875401413, + "grad_norm": 0.9952748694927424, + "learning_rate": 2.0774934254457202e-08, + "loss": 0.0583, + "step": 4549 + }, + { + "epoch": 2.922286448298009, + "grad_norm": 0.8514551023322845, + "learning_rate": 2.0435988589320723e-08, + "loss": 0.0406, + "step": 4550 + }, + { + "epoch": 2.922928709055877, + "grad_norm": 0.9286368650355399, + "learning_rate": 2.0099825021163143e-08, + "loss": 0.0384, + "step": 4551 + }, + { + "epoch": 2.9235709698137446, + "grad_norm": 0.7423513933609762, + "learning_rate": 1.9766443737800345e-08, + "loss": 0.0287, + "step": 4552 + }, + { + "epoch": 2.924213230571612, + "grad_norm": 0.6837424388830318, + "learning_rate": 1.943584492549333e-08, + "loss": 0.0252, + "step": 4553 + }, + { + "epoch": 2.9248554913294798, + "grad_norm": 0.7390190328057265, + "learning_rate": 1.910802876894824e-08, + "loss": 0.0353, + "step": 4554 + }, + { + "epoch": 2.9254977520873475, + "grad_norm": 1.0379657981864787, + "learning_rate": 1.878299545131801e-08, + "loss": 0.0449, + "step": 4555 + }, + { + "epoch": 2.9261400128452153, + "grad_norm": 0.8394931851724802, + "learning_rate": 1.846074515419849e-08, + "loss": 0.0548, + "step": 4556 + }, + { + "epoch": 2.9267822736030826, + "grad_norm": 0.6619827121678459, + "learning_rate": 1.8141278057632326e-08, + "loss": 0.0251, + "step": 4557 + }, + { + "epoch": 2.9274245343609504, + "grad_norm": 1.3740152373923518, + "learning_rate": 1.782459434010675e-08, + "loss": 0.0676, + "step": 4558 + }, + { + "epoch": 2.928066795118818, + "grad_norm": 0.8850209197961649, + "learning_rate": 1.751069417855411e-08, + "loss": 0.0357, + "step": 4559 + }, + { + "epoch": 2.928709055876686, + "grad_norm": 0.7861356851412793, + "learning_rate": 1.719957774835135e-08, + "loss": 0.0193, + "step": 4560 + }, + { + "epoch": 2.9293513166345537, + "grad_norm": 0.772550575809411, + "learning_rate": 1.689124522331942e-08, + "loss": 0.0365, + "step": 4561 + }, + { + "epoch": 2.9299935773924215, + "grad_norm": 0.556258760277339, + "learning_rate": 1.6585696775725524e-08, + "loss": 0.0203, + "step": 4562 + }, + { + "epoch": 2.9306358381502893, + "grad_norm": 0.9817017236129836, + "learning_rate": 1.6282932576279775e-08, + "loss": 0.0484, + "step": 4563 + }, + { + "epoch": 2.9312780989081566, + "grad_norm": 0.8036221267835146, + "learning_rate": 1.5982952794137973e-08, + "loss": 0.0291, + "step": 4564 + }, + { + "epoch": 2.9319203596660244, + "grad_norm": 0.7512003281785444, + "learning_rate": 1.568575759689883e-08, + "loss": 0.0273, + "step": 4565 + }, + { + "epoch": 2.932562620423892, + "grad_norm": 0.5784948391308091, + "learning_rate": 1.5391347150607305e-08, + "loss": 0.0183, + "step": 4566 + }, + { + "epoch": 2.93320488118176, + "grad_norm": 0.5800398416013018, + "learning_rate": 1.5099721619750707e-08, + "loss": 0.018, + "step": 4567 + }, + { + "epoch": 2.9338471419396273, + "grad_norm": 0.6838744858778681, + "learning_rate": 1.4810881167260927e-08, + "loss": 0.0211, + "step": 4568 + }, + { + "epoch": 2.934489402697495, + "grad_norm": 0.7372750923847509, + "learning_rate": 1.4524825954514432e-08, + "loss": 0.02, + "step": 4569 + }, + { + "epoch": 2.935131663455363, + "grad_norm": 1.018136410972737, + "learning_rate": 1.4241556141330604e-08, + "loss": 0.0527, + "step": 4570 + }, + { + "epoch": 2.9357739242132306, + "grad_norm": 0.6099100644802293, + "learning_rate": 1.3961071885973953e-08, + "loss": 0.0168, + "step": 4571 + }, + { + "epoch": 2.9364161849710984, + "grad_norm": 0.6884153628363775, + "learning_rate": 1.3683373345150796e-08, + "loss": 0.0179, + "step": 4572 + }, + { + "epoch": 2.937058445728966, + "grad_norm": 0.9812604111529455, + "learning_rate": 1.3408460674013135e-08, + "loss": 0.0381, + "step": 4573 + }, + { + "epoch": 2.937700706486834, + "grad_norm": 0.5094440392857147, + "learning_rate": 1.3136334026154773e-08, + "loss": 0.0116, + "step": 4574 + }, + { + "epoch": 2.9383429672447012, + "grad_norm": 0.9634894284109505, + "learning_rate": 1.2866993553614093e-08, + "loss": 0.0527, + "step": 4575 + }, + { + "epoch": 2.938985228002569, + "grad_norm": 0.7319186762981205, + "learning_rate": 1.2600439406871834e-08, + "loss": 0.0297, + "step": 4576 + }, + { + "epoch": 2.939627488760437, + "grad_norm": 0.4802016726159003, + "learning_rate": 1.233667173485331e-08, + "loss": 0.0133, + "step": 4577 + }, + { + "epoch": 2.9402697495183046, + "grad_norm": 0.7691573288519298, + "learning_rate": 1.2075690684925645e-08, + "loss": 0.0239, + "step": 4578 + }, + { + "epoch": 2.940912010276172, + "grad_norm": 0.6542532877492101, + "learning_rate": 1.1817496402899975e-08, + "loss": 0.0228, + "step": 4579 + }, + { + "epoch": 2.9415542710340397, + "grad_norm": 0.6945011439788644, + "learning_rate": 1.15620890330298e-08, + "loss": 0.0211, + "step": 4580 + }, + { + "epoch": 2.9421965317919074, + "grad_norm": 0.9004884443259417, + "learning_rate": 1.1309468718013194e-08, + "loss": 0.0279, + "step": 4581 + }, + { + "epoch": 2.942838792549775, + "grad_norm": 1.070377404165307, + "learning_rate": 1.1059635598988926e-08, + "loss": 0.041, + "step": 4582 + }, + { + "epoch": 2.943481053307643, + "grad_norm": 0.5897826646391873, + "learning_rate": 1.0812589815539232e-08, + "loss": 0.0164, + "step": 4583 + }, + { + "epoch": 2.9441233140655108, + "grad_norm": 0.9495310825504201, + "learning_rate": 1.056833150568981e-08, + "loss": 0.0428, + "step": 4584 + }, + { + "epoch": 2.9447655748233785, + "grad_norm": 0.4937042132914126, + "learning_rate": 1.0326860805908722e-08, + "loss": 0.0086, + "step": 4585 + }, + { + "epoch": 2.945407835581246, + "grad_norm": 0.8748685481099417, + "learning_rate": 1.008817785110583e-08, + "loss": 0.0435, + "step": 4586 + }, + { + "epoch": 2.9460500963391136, + "grad_norm": 0.6276337422034337, + "learning_rate": 9.852282774633903e-09, + "loss": 0.0173, + "step": 4587 + }, + { + "epoch": 2.9466923570969814, + "grad_norm": 0.974733275007613, + "learning_rate": 9.61917570828863e-09, + "loss": 0.0441, + "step": 4588 + }, + { + "epoch": 2.947334617854849, + "grad_norm": 0.7730585591542839, + "learning_rate": 9.388856782307498e-09, + "loss": 0.0309, + "step": 4589 + }, + { + "epoch": 2.9479768786127165, + "grad_norm": 0.7951548760781209, + "learning_rate": 9.16132612537035e-09, + "loss": 0.0416, + "step": 4590 + }, + { + "epoch": 2.9486191393705843, + "grad_norm": 0.6379948343074581, + "learning_rate": 8.93658386459939e-09, + "loss": 0.016, + "step": 4591 + }, + { + "epoch": 2.949261400128452, + "grad_norm": 0.7448676898641171, + "learning_rate": 8.714630125558066e-09, + "loss": 0.0293, + "step": 4592 + }, + { + "epoch": 2.94990366088632, + "grad_norm": 0.9626591626788761, + "learning_rate": 8.495465032253847e-09, + "loss": 0.0574, + "step": 4593 + }, + { + "epoch": 2.9505459216441876, + "grad_norm": 0.9220913197188956, + "learning_rate": 8.279088707133787e-09, + "loss": 0.044, + "step": 4594 + }, + { + "epoch": 2.9511881824020554, + "grad_norm": 0.5598390885678785, + "learning_rate": 8.065501271088405e-09, + "loss": 0.0211, + "step": 4595 + }, + { + "epoch": 2.951830443159923, + "grad_norm": 0.6046589587659158, + "learning_rate": 7.85470284344947e-09, + "loss": 0.0155, + "step": 4596 + }, + { + "epoch": 2.9524727039177905, + "grad_norm": 0.7652271085074858, + "learning_rate": 7.646693541991101e-09, + "loss": 0.0302, + "step": 4597 + }, + { + "epoch": 2.9531149646756583, + "grad_norm": 1.1261171202902998, + "learning_rate": 7.441473482928119e-09, + "loss": 0.0629, + "step": 4598 + }, + { + "epoch": 2.953757225433526, + "grad_norm": 0.9211471484438102, + "learning_rate": 7.2390427809176934e-09, + "loss": 0.0202, + "step": 4599 + }, + { + "epoch": 2.954399486191394, + "grad_norm": 0.9993027602733768, + "learning_rate": 7.0394015490588e-09, + "loss": 0.0363, + "step": 4600 + }, + { + "epoch": 2.955041746949261, + "grad_norm": 0.6246631324593608, + "learning_rate": 6.842549898891659e-09, + "loss": 0.0231, + "step": 4601 + }, + { + "epoch": 2.955684007707129, + "grad_norm": 0.7366427990379022, + "learning_rate": 6.648487940397186e-09, + "loss": 0.0262, + "step": 4602 + }, + { + "epoch": 2.9563262684649967, + "grad_norm": 0.7882723827313035, + "learning_rate": 6.4572157819992045e-09, + "loss": 0.0302, + "step": 4603 + }, + { + "epoch": 2.9569685292228645, + "grad_norm": 0.8285572973005604, + "learning_rate": 6.268733530561122e-09, + "loss": 0.0266, + "step": 4604 + }, + { + "epoch": 2.9576107899807322, + "grad_norm": 0.9816183649194576, + "learning_rate": 6.083041291389258e-09, + "loss": 0.0418, + "step": 4605 + }, + { + "epoch": 2.9582530507386, + "grad_norm": 0.9685058044765548, + "learning_rate": 5.900139168230068e-09, + "loss": 0.0377, + "step": 4606 + }, + { + "epoch": 2.958895311496468, + "grad_norm": 0.9262874069195374, + "learning_rate": 5.72002726327181e-09, + "loss": 0.0403, + "step": 4607 + }, + { + "epoch": 2.959537572254335, + "grad_norm": 1.0280987150472665, + "learning_rate": 5.542705677143434e-09, + "loss": 0.0365, + "step": 4608 + }, + { + "epoch": 2.960179833012203, + "grad_norm": 0.9852772834424813, + "learning_rate": 5.368174508914581e-09, + "loss": 0.0392, + "step": 4609 + }, + { + "epoch": 2.9608220937700707, + "grad_norm": 0.7009363468507469, + "learning_rate": 5.196433856096694e-09, + "loss": 0.022, + "step": 4610 + }, + { + "epoch": 2.9614643545279384, + "grad_norm": 0.8356941413001007, + "learning_rate": 5.027483814641354e-09, + "loss": 0.0285, + "step": 4611 + }, + { + "epoch": 2.9621066152858058, + "grad_norm": 0.9009818788101589, + "learning_rate": 4.8613244789419425e-09, + "loss": 0.0327, + "step": 4612 + }, + { + "epoch": 2.9627488760436735, + "grad_norm": 1.0774415133090904, + "learning_rate": 4.697955941832533e-09, + "loss": 0.044, + "step": 4613 + }, + { + "epoch": 2.9633911368015413, + "grad_norm": 0.483662638782465, + "learning_rate": 4.5373782945867805e-09, + "loss": 0.0095, + "step": 4614 + }, + { + "epoch": 2.964033397559409, + "grad_norm": 0.848598289476961, + "learning_rate": 4.379591626920699e-09, + "loss": 0.0316, + "step": 4615 + }, + { + "epoch": 2.964675658317277, + "grad_norm": 1.1115343386577516, + "learning_rate": 4.22459602698877e-09, + "loss": 0.0749, + "step": 4616 + }, + { + "epoch": 2.9653179190751446, + "grad_norm": 0.8303864144846576, + "learning_rate": 4.072391581388946e-09, + "loss": 0.0253, + "step": 4617 + }, + { + "epoch": 2.9659601798330124, + "grad_norm": 0.9206960949555558, + "learning_rate": 3.922978375158204e-09, + "loss": 0.0387, + "step": 4618 + }, + { + "epoch": 2.9666024405908797, + "grad_norm": 1.2376695475408326, + "learning_rate": 3.776356491773103e-09, + "loss": 0.0595, + "step": 4619 + }, + { + "epoch": 2.9672447013487475, + "grad_norm": 0.6406857406491756, + "learning_rate": 3.6325260131525597e-09, + "loss": 0.022, + "step": 4620 + }, + { + "epoch": 2.9678869621066153, + "grad_norm": 0.9354686662549183, + "learning_rate": 3.491487019654516e-09, + "loss": 0.033, + "step": 4621 + }, + { + "epoch": 2.968529222864483, + "grad_norm": 1.1175656459154597, + "learning_rate": 3.353239590078161e-09, + "loss": 0.0519, + "step": 4622 + }, + { + "epoch": 2.969171483622351, + "grad_norm": 0.8054920437881932, + "learning_rate": 3.2177838016628216e-09, + "loss": 0.0311, + "step": 4623 + }, + { + "epoch": 2.969813744380218, + "grad_norm": 0.9871528391402251, + "learning_rate": 3.08511973008796e-09, + "loss": 0.0444, + "step": 4624 + }, + { + "epoch": 2.970456005138086, + "grad_norm": 0.6477476935017596, + "learning_rate": 2.9552474494737304e-09, + "loss": 0.0264, + "step": 4625 + }, + { + "epoch": 2.9710982658959537, + "grad_norm": 0.7804856843379173, + "learning_rate": 2.8281670323798693e-09, + "loss": 0.0373, + "step": 4626 + }, + { + "epoch": 2.9717405266538215, + "grad_norm": 0.808277431547023, + "learning_rate": 2.703878549806249e-09, + "loss": 0.0264, + "step": 4627 + }, + { + "epoch": 2.9723827874116893, + "grad_norm": 0.4267455928685298, + "learning_rate": 2.5823820711934345e-09, + "loss": 0.0138, + "step": 4628 + }, + { + "epoch": 2.973025048169557, + "grad_norm": 0.7553364220259384, + "learning_rate": 2.463677664422126e-09, + "loss": 0.0247, + "step": 4629 + }, + { + "epoch": 2.9736673089274244, + "grad_norm": 0.7507295971854526, + "learning_rate": 2.347765395812052e-09, + "loss": 0.0337, + "step": 4630 + }, + { + "epoch": 2.974309569685292, + "grad_norm": 0.7864811629424241, + "learning_rate": 2.2346453301247405e-09, + "loss": 0.0229, + "step": 4631 + }, + { + "epoch": 2.97495183044316, + "grad_norm": 0.890337786896748, + "learning_rate": 2.124317530560749e-09, + "loss": 0.0521, + "step": 4632 + }, + { + "epoch": 2.9755940912010277, + "grad_norm": 0.695933678806735, + "learning_rate": 2.0167820587591036e-09, + "loss": 0.0226, + "step": 4633 + }, + { + "epoch": 2.9762363519588955, + "grad_norm": 0.6646756206571062, + "learning_rate": 1.9120389748017443e-09, + "loss": 0.0245, + "step": 4634 + }, + { + "epoch": 2.976878612716763, + "grad_norm": 0.7120191153619057, + "learning_rate": 1.8100883372085266e-09, + "loss": 0.0246, + "step": 4635 + }, + { + "epoch": 2.9775208734746306, + "grad_norm": 0.9301483494311812, + "learning_rate": 1.7109302029388874e-09, + "loss": 0.0405, + "step": 4636 + }, + { + "epoch": 2.9781631342324983, + "grad_norm": 0.6718514438515829, + "learning_rate": 1.614564627393511e-09, + "loss": 0.0274, + "step": 4637 + }, + { + "epoch": 2.978805394990366, + "grad_norm": 0.5702040339991336, + "learning_rate": 1.520991664411553e-09, + "loss": 0.0194, + "step": 4638 + }, + { + "epoch": 2.979447655748234, + "grad_norm": 0.8983489052035839, + "learning_rate": 1.4302113662734151e-09, + "loss": 0.0255, + "step": 4639 + }, + { + "epoch": 2.9800899165061017, + "grad_norm": 0.5031969778215329, + "learning_rate": 1.342223783697416e-09, + "loss": 0.0217, + "step": 4640 + }, + { + "epoch": 2.9807321772639694, + "grad_norm": 0.9177321194793221, + "learning_rate": 1.2570289658425662e-09, + "loss": 0.0418, + "step": 4641 + }, + { + "epoch": 2.9813744380218368, + "grad_norm": 0.5325593671824016, + "learning_rate": 1.1746269603080119e-09, + "loss": 0.0183, + "step": 4642 + }, + { + "epoch": 2.9820166987797045, + "grad_norm": 0.8833259935397139, + "learning_rate": 1.0950178131313716e-09, + "loss": 0.0428, + "step": 4643 + }, + { + "epoch": 2.9826589595375723, + "grad_norm": 0.8376610920624159, + "learning_rate": 1.0182015687909552e-09, + "loss": 0.0298, + "step": 4644 + }, + { + "epoch": 2.98330122029544, + "grad_norm": 0.9025351995071831, + "learning_rate": 9.44178270204099e-10, + "loss": 0.0372, + "step": 4645 + }, + { + "epoch": 2.9839434810533074, + "grad_norm": 0.9922665404497855, + "learning_rate": 8.729479587277212e-10, + "loss": 0.0577, + "step": 4646 + }, + { + "epoch": 2.984585741811175, + "grad_norm": 0.7174340883086482, + "learning_rate": 8.045106741583208e-10, + "loss": 0.0223, + "step": 4647 + }, + { + "epoch": 2.985228002569043, + "grad_norm": 0.8503527947568462, + "learning_rate": 7.388664547319791e-10, + "loss": 0.0344, + "step": 4648 + }, + { + "epoch": 2.9858702633269107, + "grad_norm": 0.9514838925632806, + "learning_rate": 6.760153371249134e-10, + "loss": 0.0392, + "step": 4649 + }, + { + "epoch": 2.9865125240847785, + "grad_norm": 1.0090826939396804, + "learning_rate": 6.159573564518128e-10, + "loss": 0.0497, + "step": 4650 + }, + { + "epoch": 2.9871547848426463, + "grad_norm": 0.9197380798510467, + "learning_rate": 5.58692546267503e-10, + "loss": 0.0507, + "step": 4651 + }, + { + "epoch": 2.987797045600514, + "grad_norm": 0.6494786727596622, + "learning_rate": 5.042209385652808e-10, + "loss": 0.0244, + "step": 4652 + }, + { + "epoch": 2.9884393063583814, + "grad_norm": 0.6647118729650385, + "learning_rate": 4.5254256377913474e-10, + "loss": 0.0226, + "step": 4653 + }, + { + "epoch": 2.989081567116249, + "grad_norm": 0.8591335868535287, + "learning_rate": 4.036574507820801e-10, + "loss": 0.0391, + "step": 4654 + }, + { + "epoch": 2.989723827874117, + "grad_norm": 1.0382332553028562, + "learning_rate": 3.5756562688615826e-10, + "loss": 0.0483, + "step": 4655 + }, + { + "epoch": 2.9903660886319847, + "grad_norm": 0.9358914696637898, + "learning_rate": 3.142671178429924e-10, + "loss": 0.043, + "step": 4656 + }, + { + "epoch": 2.991008349389852, + "grad_norm": 0.9975493181457766, + "learning_rate": 2.7376194784323186e-10, + "loss": 0.0549, + "step": 4657 + }, + { + "epoch": 2.99165061014772, + "grad_norm": 0.9798070086784848, + "learning_rate": 2.3605013951766285e-10, + "loss": 0.0477, + "step": 4658 + }, + { + "epoch": 2.9922928709055876, + "grad_norm": 0.42297642567603017, + "learning_rate": 2.0113171393609797e-10, + "loss": 0.0154, + "step": 4659 + }, + { + "epoch": 2.9929351316634554, + "grad_norm": 0.7482323822794105, + "learning_rate": 1.6900669060737617e-10, + "loss": 0.0247, + "step": 4660 + }, + { + "epoch": 2.993577392421323, + "grad_norm": 0.746102415989889, + "learning_rate": 1.3967508747991798e-10, + "loss": 0.0284, + "step": 4661 + }, + { + "epoch": 2.994219653179191, + "grad_norm": 1.0795963121304182, + "learning_rate": 1.1313692094117034e-10, + "loss": 0.046, + "step": 4662 + }, + { + "epoch": 2.9948619139370587, + "grad_norm": 0.9849952467677396, + "learning_rate": 8.939220581816177e-11, + "loss": 0.0376, + "step": 4663 + }, + { + "epoch": 2.995504174694926, + "grad_norm": 0.6901374926861387, + "learning_rate": 6.844095537750228e-11, + "loss": 0.0257, + "step": 4664 + }, + { + "epoch": 2.996146435452794, + "grad_norm": 1.0381743251206095, + "learning_rate": 5.028318132371812e-11, + "loss": 0.0357, + "step": 4665 + }, + { + "epoch": 2.9967886962106616, + "grad_norm": 0.6697009516342041, + "learning_rate": 3.491889380313751e-11, + "loss": 0.0219, + "step": 4666 + }, + { + "epoch": 2.9974309569685293, + "grad_norm": 1.107259054324942, + "learning_rate": 2.234810139833954e-11, + "loss": 0.0469, + "step": 4667 + }, + { + "epoch": 2.9980732177263967, + "grad_norm": 0.8744135027734813, + "learning_rate": 1.2570811133705285e-11, + "loss": 0.0307, + "step": 4668 + }, + { + "epoch": 2.9987154784842645, + "grad_norm": 0.8164194619024757, + "learning_rate": 5.587028470976919e-12, + "loss": 0.0326, + "step": 4669 + }, + { + "epoch": 2.9993577392421322, + "grad_norm": 0.7901412248270551, + "learning_rate": 1.3967573125883705e-12, + "loss": 0.0243, + "step": 4670 + }, + { + "epoch": 3.0, + "grad_norm": 0.7280663086629195, + "learning_rate": 0.0, + "loss": 0.028, + "step": 4671 + }, + { + "epoch": 3.0, + "step": 4671, + "total_flos": 215019869847552.0, + "train_loss": 0.09441747322642279, + "train_runtime": 40412.3051, + "train_samples_per_second": 1.387, + "train_steps_per_second": 0.116 + } + ], + "logging_steps": 1, + "max_steps": 4671, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 215019869847552.0, + "train_batch_size": 2, + "trial_name": null, + "trial_params": null +}