{ "best_metric": null, "best_model_checkpoint": null, "epoch": 4.994377034625629, "eval_steps": 500, "global_step": 4220, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0011837821840781297, "grad_norm": 27.25, "learning_rate": 4.739336492890996e-08, "loss": 4.5742, "step": 1 }, { "epoch": 0.0023675643681562593, "grad_norm": 25.875, "learning_rate": 9.478672985781992e-08, "loss": 4.1168, "step": 2 }, { "epoch": 0.003551346552234389, "grad_norm": 25.5, "learning_rate": 1.4218009478672986e-07, "loss": 4.2987, "step": 3 }, { "epoch": 0.004735128736312519, "grad_norm": 30.25, "learning_rate": 1.8957345971563984e-07, "loss": 4.5405, "step": 4 }, { "epoch": 0.005918910920390648, "grad_norm": 27.125, "learning_rate": 2.3696682464454978e-07, "loss": 4.3069, "step": 5 }, { "epoch": 0.007102693104468778, "grad_norm": 26.0, "learning_rate": 2.843601895734597e-07, "loss": 4.552, "step": 6 }, { "epoch": 0.008286475288546907, "grad_norm": 31.125, "learning_rate": 3.317535545023697e-07, "loss": 4.3416, "step": 7 }, { "epoch": 0.009470257472625037, "grad_norm": 25.25, "learning_rate": 3.791469194312797e-07, "loss": 4.5185, "step": 8 }, { "epoch": 0.010654039656703166, "grad_norm": 28.375, "learning_rate": 4.265402843601896e-07, "loss": 4.6644, "step": 9 }, { "epoch": 0.011837821840781295, "grad_norm": 24.25, "learning_rate": 4.7393364928909956e-07, "loss": 4.0805, "step": 10 }, { "epoch": 0.013021604024859426, "grad_norm": 31.0, "learning_rate": 5.213270142180095e-07, "loss": 4.8569, "step": 11 }, { "epoch": 0.014205386208937555, "grad_norm": 27.75, "learning_rate": 5.687203791469194e-07, "loss": 4.2514, "step": 12 }, { "epoch": 0.015389168393015686, "grad_norm": 25.0, "learning_rate": 6.161137440758294e-07, "loss": 4.0034, "step": 13 }, { "epoch": 0.016572950577093813, "grad_norm": 25.0, "learning_rate": 6.635071090047394e-07, "loss": 4.568, "step": 14 }, { "epoch": 0.017756732761171946, "grad_norm": 27.875, "learning_rate": 7.109004739336493e-07, "loss": 4.7432, "step": 15 }, { "epoch": 0.018940514945250075, "grad_norm": 24.375, "learning_rate": 7.582938388625594e-07, "loss": 4.1086, "step": 16 }, { "epoch": 0.020124297129328204, "grad_norm": 32.0, "learning_rate": 8.056872037914692e-07, "loss": 5.0878, "step": 17 }, { "epoch": 0.021308079313406333, "grad_norm": 27.875, "learning_rate": 8.530805687203792e-07, "loss": 4.1515, "step": 18 }, { "epoch": 0.022491861497484462, "grad_norm": 31.875, "learning_rate": 9.004739336492892e-07, "loss": 4.4924, "step": 19 }, { "epoch": 0.02367564368156259, "grad_norm": 30.75, "learning_rate": 9.478672985781991e-07, "loss": 4.1903, "step": 20 }, { "epoch": 0.024859425865640723, "grad_norm": 24.25, "learning_rate": 9.95260663507109e-07, "loss": 4.1328, "step": 21 }, { "epoch": 0.026043208049718852, "grad_norm": 29.75, "learning_rate": 1.042654028436019e-06, "loss": 4.3652, "step": 22 }, { "epoch": 0.02722699023379698, "grad_norm": 31.5, "learning_rate": 1.090047393364929e-06, "loss": 4.5804, "step": 23 }, { "epoch": 0.02841077241787511, "grad_norm": 27.875, "learning_rate": 1.1374407582938388e-06, "loss": 4.2407, "step": 24 }, { "epoch": 0.02959455460195324, "grad_norm": 26.0, "learning_rate": 1.184834123222749e-06, "loss": 3.823, "step": 25 }, { "epoch": 0.030778336786031372, "grad_norm": 24.75, "learning_rate": 1.2322274881516587e-06, "loss": 4.5008, "step": 26 }, { "epoch": 0.0319621189701095, "grad_norm": 34.5, "learning_rate": 1.2796208530805687e-06, "loss": 4.1535, "step": 27 }, { "epoch": 0.033145901154187626, "grad_norm": 30.375, "learning_rate": 1.3270142180094788e-06, "loss": 4.6596, "step": 28 }, { "epoch": 0.03432968333826576, "grad_norm": 28.375, "learning_rate": 1.3744075829383887e-06, "loss": 4.2302, "step": 29 }, { "epoch": 0.03551346552234389, "grad_norm": 29.75, "learning_rate": 1.4218009478672987e-06, "loss": 3.9536, "step": 30 }, { "epoch": 0.03669724770642202, "grad_norm": 25.625, "learning_rate": 1.4691943127962086e-06, "loss": 4.0247, "step": 31 }, { "epoch": 0.03788102989050015, "grad_norm": 28.625, "learning_rate": 1.5165876777251187e-06, "loss": 4.2866, "step": 32 }, { "epoch": 0.03906481207457828, "grad_norm": 27.0, "learning_rate": 1.5639810426540287e-06, "loss": 4.3479, "step": 33 }, { "epoch": 0.04024859425865641, "grad_norm": 27.125, "learning_rate": 1.6113744075829384e-06, "loss": 4.1932, "step": 34 }, { "epoch": 0.041432376442734536, "grad_norm": 31.375, "learning_rate": 1.6587677725118483e-06, "loss": 4.5418, "step": 35 }, { "epoch": 0.042616158626812665, "grad_norm": 24.5, "learning_rate": 1.7061611374407585e-06, "loss": 3.9115, "step": 36 }, { "epoch": 0.043799940810890794, "grad_norm": 29.0, "learning_rate": 1.7535545023696684e-06, "loss": 4.4143, "step": 37 }, { "epoch": 0.044983722994968924, "grad_norm": 27.25, "learning_rate": 1.8009478672985784e-06, "loss": 4.256, "step": 38 }, { "epoch": 0.04616750517904705, "grad_norm": 25.375, "learning_rate": 1.8483412322274883e-06, "loss": 3.8378, "step": 39 }, { "epoch": 0.04735128736312518, "grad_norm": 27.125, "learning_rate": 1.8957345971563982e-06, "loss": 3.9533, "step": 40 }, { "epoch": 0.04853506954720332, "grad_norm": 28.375, "learning_rate": 1.943127962085308e-06, "loss": 4.006, "step": 41 }, { "epoch": 0.049718851731281447, "grad_norm": 25.875, "learning_rate": 1.990521327014218e-06, "loss": 4.3288, "step": 42 }, { "epoch": 0.050902633915359576, "grad_norm": 25.875, "learning_rate": 2.037914691943128e-06, "loss": 4.1281, "step": 43 }, { "epoch": 0.052086416099437705, "grad_norm": 26.0, "learning_rate": 2.085308056872038e-06, "loss": 4.2562, "step": 44 }, { "epoch": 0.053270198283515834, "grad_norm": 29.375, "learning_rate": 2.1327014218009483e-06, "loss": 4.144, "step": 45 }, { "epoch": 0.05445398046759396, "grad_norm": 28.75, "learning_rate": 2.180094786729858e-06, "loss": 4.2635, "step": 46 }, { "epoch": 0.05563776265167209, "grad_norm": 30.5, "learning_rate": 2.2274881516587678e-06, "loss": 4.1032, "step": 47 }, { "epoch": 0.05682154483575022, "grad_norm": 28.375, "learning_rate": 2.2748815165876777e-06, "loss": 4.0056, "step": 48 }, { "epoch": 0.05800532701982835, "grad_norm": 21.25, "learning_rate": 2.322274881516588e-06, "loss": 3.6885, "step": 49 }, { "epoch": 0.05918910920390648, "grad_norm": 38.75, "learning_rate": 2.369668246445498e-06, "loss": 4.8308, "step": 50 }, { "epoch": 0.06037289138798461, "grad_norm": 27.5, "learning_rate": 2.417061611374408e-06, "loss": 3.9768, "step": 51 }, { "epoch": 0.061556673572062744, "grad_norm": 24.0, "learning_rate": 2.4644549763033174e-06, "loss": 3.6757, "step": 52 }, { "epoch": 0.06274045575614087, "grad_norm": 27.25, "learning_rate": 2.5118483412322274e-06, "loss": 4.0805, "step": 53 }, { "epoch": 0.063924237940219, "grad_norm": 24.25, "learning_rate": 2.5592417061611373e-06, "loss": 3.896, "step": 54 }, { "epoch": 0.06510802012429713, "grad_norm": 26.0, "learning_rate": 2.606635071090048e-06, "loss": 3.9153, "step": 55 }, { "epoch": 0.06629180230837525, "grad_norm": 23.25, "learning_rate": 2.6540284360189576e-06, "loss": 3.9136, "step": 56 }, { "epoch": 0.06747558449245339, "grad_norm": 27.0, "learning_rate": 2.7014218009478675e-06, "loss": 4.0031, "step": 57 }, { "epoch": 0.06865936667653152, "grad_norm": 25.75, "learning_rate": 2.7488151658767775e-06, "loss": 3.8149, "step": 58 }, { "epoch": 0.06984314886060965, "grad_norm": 23.5, "learning_rate": 2.7962085308056874e-06, "loss": 3.6599, "step": 59 }, { "epoch": 0.07102693104468778, "grad_norm": 23.625, "learning_rate": 2.8436018957345973e-06, "loss": 3.9296, "step": 60 }, { "epoch": 0.0722107132287659, "grad_norm": 22.5, "learning_rate": 2.8909952606635073e-06, "loss": 3.8185, "step": 61 }, { "epoch": 0.07339449541284404, "grad_norm": 25.375, "learning_rate": 2.938388625592417e-06, "loss": 4.0957, "step": 62 }, { "epoch": 0.07457827759692216, "grad_norm": 19.625, "learning_rate": 2.985781990521327e-06, "loss": 3.8312, "step": 63 }, { "epoch": 0.0757620597810003, "grad_norm": 24.75, "learning_rate": 3.0331753554502375e-06, "loss": 3.9194, "step": 64 }, { "epoch": 0.07694584196507842, "grad_norm": 24.625, "learning_rate": 3.0805687203791474e-06, "loss": 3.8047, "step": 65 }, { "epoch": 0.07812962414915656, "grad_norm": 20.375, "learning_rate": 3.1279620853080574e-06, "loss": 4.0207, "step": 66 }, { "epoch": 0.07931340633323468, "grad_norm": 21.875, "learning_rate": 3.1753554502369673e-06, "loss": 3.6675, "step": 67 }, { "epoch": 0.08049718851731281, "grad_norm": 17.875, "learning_rate": 3.222748815165877e-06, "loss": 3.5946, "step": 68 }, { "epoch": 0.08168097070139095, "grad_norm": 17.75, "learning_rate": 3.2701421800947867e-06, "loss": 3.5922, "step": 69 }, { "epoch": 0.08286475288546907, "grad_norm": 18.375, "learning_rate": 3.3175355450236967e-06, "loss": 3.4621, "step": 70 }, { "epoch": 0.08404853506954721, "grad_norm": 16.625, "learning_rate": 3.3649289099526066e-06, "loss": 3.4249, "step": 71 }, { "epoch": 0.08523231725362533, "grad_norm": 17.0, "learning_rate": 3.412322274881517e-06, "loss": 3.1019, "step": 72 }, { "epoch": 0.08641609943770347, "grad_norm": 16.875, "learning_rate": 3.459715639810427e-06, "loss": 3.8747, "step": 73 }, { "epoch": 0.08759988162178159, "grad_norm": 17.25, "learning_rate": 3.507109004739337e-06, "loss": 3.2951, "step": 74 }, { "epoch": 0.08878366380585972, "grad_norm": 14.875, "learning_rate": 3.5545023696682468e-06, "loss": 3.6178, "step": 75 }, { "epoch": 0.08996744598993785, "grad_norm": 14.875, "learning_rate": 3.6018957345971567e-06, "loss": 3.6471, "step": 76 }, { "epoch": 0.09115122817401598, "grad_norm": 15.375, "learning_rate": 3.6492890995260666e-06, "loss": 3.6079, "step": 77 }, { "epoch": 0.0923350103580941, "grad_norm": 14.5625, "learning_rate": 3.6966824644549766e-06, "loss": 3.5527, "step": 78 }, { "epoch": 0.09351879254217224, "grad_norm": 13.875, "learning_rate": 3.7440758293838865e-06, "loss": 3.6016, "step": 79 }, { "epoch": 0.09470257472625036, "grad_norm": 13.125, "learning_rate": 3.7914691943127964e-06, "loss": 3.142, "step": 80 }, { "epoch": 0.0958863569103285, "grad_norm": 13.3125, "learning_rate": 3.838862559241707e-06, "loss": 3.6945, "step": 81 }, { "epoch": 0.09707013909440664, "grad_norm": 13.1875, "learning_rate": 3.886255924170616e-06, "loss": 3.3941, "step": 82 }, { "epoch": 0.09825392127848476, "grad_norm": 12.375, "learning_rate": 3.933649289099527e-06, "loss": 3.4295, "step": 83 }, { "epoch": 0.09943770346256289, "grad_norm": 11.9375, "learning_rate": 3.981042654028436e-06, "loss": 3.3487, "step": 84 }, { "epoch": 0.10062148564664102, "grad_norm": 12.25, "learning_rate": 4.0284360189573465e-06, "loss": 2.9911, "step": 85 }, { "epoch": 0.10180526783071915, "grad_norm": 13.1875, "learning_rate": 4.075829383886256e-06, "loss": 3.2688, "step": 86 }, { "epoch": 0.10298905001479727, "grad_norm": 13.875, "learning_rate": 4.123222748815166e-06, "loss": 3.3584, "step": 87 }, { "epoch": 0.10417283219887541, "grad_norm": 11.1875, "learning_rate": 4.170616113744076e-06, "loss": 3.3488, "step": 88 }, { "epoch": 0.10535661438295353, "grad_norm": 9.625, "learning_rate": 4.218009478672986e-06, "loss": 3.0363, "step": 89 }, { "epoch": 0.10654039656703167, "grad_norm": 11.25, "learning_rate": 4.265402843601897e-06, "loss": 3.0994, "step": 90 }, { "epoch": 0.10772417875110979, "grad_norm": 10.625, "learning_rate": 4.312796208530806e-06, "loss": 3.1566, "step": 91 }, { "epoch": 0.10890796093518793, "grad_norm": 10.25, "learning_rate": 4.360189573459716e-06, "loss": 3.1273, "step": 92 }, { "epoch": 0.11009174311926606, "grad_norm": 10.25, "learning_rate": 4.407582938388626e-06, "loss": 3.3201, "step": 93 }, { "epoch": 0.11127552530334418, "grad_norm": 10.8125, "learning_rate": 4.4549763033175355e-06, "loss": 3.4172, "step": 94 }, { "epoch": 0.11245930748742232, "grad_norm": 9.4375, "learning_rate": 4.502369668246446e-06, "loss": 2.954, "step": 95 }, { "epoch": 0.11364308967150044, "grad_norm": 10.25, "learning_rate": 4.549763033175355e-06, "loss": 3.1252, "step": 96 }, { "epoch": 0.11482687185557858, "grad_norm": 11.625, "learning_rate": 4.597156398104266e-06, "loss": 3.2224, "step": 97 }, { "epoch": 0.1160106540396567, "grad_norm": 9.0625, "learning_rate": 4.644549763033176e-06, "loss": 3.2088, "step": 98 }, { "epoch": 0.11719443622373484, "grad_norm": 10.375, "learning_rate": 4.691943127962086e-06, "loss": 3.2489, "step": 99 }, { "epoch": 0.11837821840781296, "grad_norm": 9.8125, "learning_rate": 4.739336492890996e-06, "loss": 3.2759, "step": 100 }, { "epoch": 0.1195620005918911, "grad_norm": 9.125, "learning_rate": 4.7867298578199055e-06, "loss": 3.4241, "step": 101 }, { "epoch": 0.12074578277596922, "grad_norm": 9.4375, "learning_rate": 4.834123222748816e-06, "loss": 3.4328, "step": 102 }, { "epoch": 0.12192956496004735, "grad_norm": 8.4375, "learning_rate": 4.881516587677725e-06, "loss": 3.2802, "step": 103 }, { "epoch": 0.12311334714412549, "grad_norm": 9.125, "learning_rate": 4.928909952606635e-06, "loss": 3.1197, "step": 104 }, { "epoch": 0.12429712932820361, "grad_norm": 9.625, "learning_rate": 4.976303317535545e-06, "loss": 3.1239, "step": 105 }, { "epoch": 0.12548091151228175, "grad_norm": 8.3125, "learning_rate": 5.023696682464455e-06, "loss": 3.4392, "step": 106 }, { "epoch": 0.12666469369635988, "grad_norm": 8.625, "learning_rate": 5.071090047393366e-06, "loss": 2.8405, "step": 107 }, { "epoch": 0.127848475880438, "grad_norm": 8.875, "learning_rate": 5.118483412322275e-06, "loss": 3.1389, "step": 108 }, { "epoch": 0.12903225806451613, "grad_norm": 8.1875, "learning_rate": 5.165876777251185e-06, "loss": 3.102, "step": 109 }, { "epoch": 0.13021604024859426, "grad_norm": 8.25, "learning_rate": 5.213270142180096e-06, "loss": 3.3589, "step": 110 }, { "epoch": 0.1313998224326724, "grad_norm": 7.6875, "learning_rate": 5.260663507109005e-06, "loss": 2.9471, "step": 111 }, { "epoch": 0.1325836046167505, "grad_norm": 8.4375, "learning_rate": 5.308056872037915e-06, "loss": 3.5415, "step": 112 }, { "epoch": 0.13376738680082864, "grad_norm": 8.0625, "learning_rate": 5.355450236966825e-06, "loss": 3.157, "step": 113 }, { "epoch": 0.13495116898490678, "grad_norm": 8.75, "learning_rate": 5.402843601895735e-06, "loss": 3.2572, "step": 114 }, { "epoch": 0.1361349511689849, "grad_norm": 8.25, "learning_rate": 5.4502369668246446e-06, "loss": 2.5215, "step": 115 }, { "epoch": 0.13731873335306305, "grad_norm": 8.0, "learning_rate": 5.497630331753555e-06, "loss": 3.0994, "step": 116 }, { "epoch": 0.13850251553714116, "grad_norm": 6.71875, "learning_rate": 5.5450236966824644e-06, "loss": 2.7125, "step": 117 }, { "epoch": 0.1396862977212193, "grad_norm": 7.96875, "learning_rate": 5.592417061611375e-06, "loss": 3.2456, "step": 118 }, { "epoch": 0.14087007990529743, "grad_norm": 7.28125, "learning_rate": 5.639810426540285e-06, "loss": 3.0475, "step": 119 }, { "epoch": 0.14205386208937557, "grad_norm": 7.1875, "learning_rate": 5.687203791469195e-06, "loss": 2.6618, "step": 120 }, { "epoch": 0.14323764427345367, "grad_norm": 7.0625, "learning_rate": 5.734597156398105e-06, "loss": 2.9899, "step": 121 }, { "epoch": 0.1444214264575318, "grad_norm": 7.6875, "learning_rate": 5.7819905213270145e-06, "loss": 2.954, "step": 122 }, { "epoch": 0.14560520864160995, "grad_norm": 8.0625, "learning_rate": 5.829383886255925e-06, "loss": 3.1818, "step": 123 }, { "epoch": 0.14678899082568808, "grad_norm": 6.59375, "learning_rate": 5.876777251184834e-06, "loss": 2.9092, "step": 124 }, { "epoch": 0.1479727730097662, "grad_norm": 7.34375, "learning_rate": 5.924170616113745e-06, "loss": 2.7546, "step": 125 }, { "epoch": 0.14915655519384433, "grad_norm": 6.6875, "learning_rate": 5.971563981042654e-06, "loss": 2.7714, "step": 126 }, { "epoch": 0.15034033737792246, "grad_norm": 7.65625, "learning_rate": 6.018957345971565e-06, "loss": 3.1014, "step": 127 }, { "epoch": 0.1515241195620006, "grad_norm": 6.40625, "learning_rate": 6.066350710900475e-06, "loss": 2.6143, "step": 128 }, { "epoch": 0.15270790174607873, "grad_norm": 7.0, "learning_rate": 6.1137440758293845e-06, "loss": 2.7326, "step": 129 }, { "epoch": 0.15389168393015684, "grad_norm": 6.625, "learning_rate": 6.161137440758295e-06, "loss": 3.0394, "step": 130 }, { "epoch": 0.15507546611423498, "grad_norm": 6.4375, "learning_rate": 6.208530805687204e-06, "loss": 2.8432, "step": 131 }, { "epoch": 0.1562592482983131, "grad_norm": 6.3125, "learning_rate": 6.255924170616115e-06, "loss": 3.0408, "step": 132 }, { "epoch": 0.15744303048239125, "grad_norm": 6.53125, "learning_rate": 6.303317535545023e-06, "loss": 3.3141, "step": 133 }, { "epoch": 0.15862681266646936, "grad_norm": 7.15625, "learning_rate": 6.350710900473935e-06, "loss": 2.8043, "step": 134 }, { "epoch": 0.1598105948505475, "grad_norm": 6.5625, "learning_rate": 6.398104265402843e-06, "loss": 2.9025, "step": 135 }, { "epoch": 0.16099437703462563, "grad_norm": 6.28125, "learning_rate": 6.445497630331754e-06, "loss": 2.7204, "step": 136 }, { "epoch": 0.16217815921870377, "grad_norm": 6.625, "learning_rate": 6.492890995260665e-06, "loss": 3.2232, "step": 137 }, { "epoch": 0.1633619414027819, "grad_norm": 6.125, "learning_rate": 6.5402843601895735e-06, "loss": 2.7932, "step": 138 }, { "epoch": 0.16454572358686, "grad_norm": 5.4375, "learning_rate": 6.587677725118484e-06, "loss": 2.9037, "step": 139 }, { "epoch": 0.16572950577093815, "grad_norm": 6.53125, "learning_rate": 6.635071090047393e-06, "loss": 2.7124, "step": 140 }, { "epoch": 0.16691328795501628, "grad_norm": 5.8125, "learning_rate": 6.682464454976304e-06, "loss": 2.8579, "step": 141 }, { "epoch": 0.16809707013909442, "grad_norm": 6.09375, "learning_rate": 6.729857819905213e-06, "loss": 2.7268, "step": 142 }, { "epoch": 0.16928085232317253, "grad_norm": 5.4375, "learning_rate": 6.777251184834124e-06, "loss": 2.8448, "step": 143 }, { "epoch": 0.17046463450725066, "grad_norm": 6.09375, "learning_rate": 6.824644549763034e-06, "loss": 2.9575, "step": 144 }, { "epoch": 0.1716484166913288, "grad_norm": 5.46875, "learning_rate": 6.8720379146919435e-06, "loss": 3.0127, "step": 145 }, { "epoch": 0.17283219887540693, "grad_norm": 5.96875, "learning_rate": 6.919431279620854e-06, "loss": 3.029, "step": 146 }, { "epoch": 0.17401598105948504, "grad_norm": 5.875, "learning_rate": 6.966824644549763e-06, "loss": 2.981, "step": 147 }, { "epoch": 0.17519976324356318, "grad_norm": 5.9375, "learning_rate": 7.014218009478674e-06, "loss": 2.7306, "step": 148 }, { "epoch": 0.17638354542764131, "grad_norm": 6.375, "learning_rate": 7.061611374407583e-06, "loss": 3.0311, "step": 149 }, { "epoch": 0.17756732761171945, "grad_norm": 6.1875, "learning_rate": 7.1090047393364935e-06, "loss": 3.01, "step": 150 }, { "epoch": 0.17875110979579759, "grad_norm": 6.40625, "learning_rate": 7.156398104265403e-06, "loss": 2.8043, "step": 151 }, { "epoch": 0.1799348919798757, "grad_norm": 6.09375, "learning_rate": 7.203791469194313e-06, "loss": 3.0152, "step": 152 }, { "epoch": 0.18111867416395383, "grad_norm": 7.03125, "learning_rate": 7.251184834123224e-06, "loss": 3.0736, "step": 153 }, { "epoch": 0.18230245634803197, "grad_norm": 6.3125, "learning_rate": 7.298578199052133e-06, "loss": 2.8339, "step": 154 }, { "epoch": 0.1834862385321101, "grad_norm": 6.9375, "learning_rate": 7.345971563981044e-06, "loss": 3.1329, "step": 155 }, { "epoch": 0.1846700207161882, "grad_norm": 6.03125, "learning_rate": 7.393364928909953e-06, "loss": 3.1381, "step": 156 }, { "epoch": 0.18585380290026635, "grad_norm": 5.625, "learning_rate": 7.4407582938388635e-06, "loss": 2.918, "step": 157 }, { "epoch": 0.18703758508434448, "grad_norm": 5.375, "learning_rate": 7.488151658767773e-06, "loss": 2.895, "step": 158 }, { "epoch": 0.18822136726842262, "grad_norm": 5.90625, "learning_rate": 7.535545023696683e-06, "loss": 2.8723, "step": 159 }, { "epoch": 0.18940514945250073, "grad_norm": 6.625, "learning_rate": 7.582938388625593e-06, "loss": 2.7011, "step": 160 }, { "epoch": 0.19058893163657886, "grad_norm": 5.6875, "learning_rate": 7.630331753554503e-06, "loss": 2.9097, "step": 161 }, { "epoch": 0.191772713820657, "grad_norm": 5.625, "learning_rate": 7.677725118483414e-06, "loss": 3.1994, "step": 162 }, { "epoch": 0.19295649600473513, "grad_norm": 5.65625, "learning_rate": 7.725118483412322e-06, "loss": 2.9914, "step": 163 }, { "epoch": 0.19414027818881327, "grad_norm": 5.4375, "learning_rate": 7.772511848341233e-06, "loss": 2.8243, "step": 164 }, { "epoch": 0.19532406037289138, "grad_norm": 5.71875, "learning_rate": 7.819905213270143e-06, "loss": 2.7478, "step": 165 }, { "epoch": 0.19650784255696951, "grad_norm": 5.34375, "learning_rate": 7.867298578199053e-06, "loss": 2.8231, "step": 166 }, { "epoch": 0.19769162474104765, "grad_norm": 5.09375, "learning_rate": 7.914691943127962e-06, "loss": 2.9692, "step": 167 }, { "epoch": 0.19887540692512579, "grad_norm": 6.78125, "learning_rate": 7.962085308056872e-06, "loss": 2.9149, "step": 168 }, { "epoch": 0.2000591891092039, "grad_norm": 5.5, "learning_rate": 8.009478672985783e-06, "loss": 3.0864, "step": 169 }, { "epoch": 0.20124297129328203, "grad_norm": 5.34375, "learning_rate": 8.056872037914693e-06, "loss": 2.8221, "step": 170 }, { "epoch": 0.20242675347736017, "grad_norm": 5.53125, "learning_rate": 8.104265402843603e-06, "loss": 2.8299, "step": 171 }, { "epoch": 0.2036105356614383, "grad_norm": 5.25, "learning_rate": 8.151658767772512e-06, "loss": 2.5595, "step": 172 }, { "epoch": 0.20479431784551644, "grad_norm": 6.0625, "learning_rate": 8.199052132701422e-06, "loss": 3.0452, "step": 173 }, { "epoch": 0.20597810002959455, "grad_norm": 5.40625, "learning_rate": 8.246445497630333e-06, "loss": 2.6605, "step": 174 }, { "epoch": 0.20716188221367268, "grad_norm": 6.03125, "learning_rate": 8.293838862559243e-06, "loss": 2.6988, "step": 175 }, { "epoch": 0.20834566439775082, "grad_norm": 5.46875, "learning_rate": 8.341232227488152e-06, "loss": 2.8868, "step": 176 }, { "epoch": 0.20952944658182895, "grad_norm": 5.53125, "learning_rate": 8.388625592417062e-06, "loss": 2.7421, "step": 177 }, { "epoch": 0.21071322876590706, "grad_norm": 5.9375, "learning_rate": 8.436018957345973e-06, "loss": 2.9454, "step": 178 }, { "epoch": 0.2118970109499852, "grad_norm": 5.59375, "learning_rate": 8.483412322274883e-06, "loss": 2.913, "step": 179 }, { "epoch": 0.21308079313406333, "grad_norm": 5.71875, "learning_rate": 8.530805687203793e-06, "loss": 2.9006, "step": 180 }, { "epoch": 0.21426457531814147, "grad_norm": 5.6875, "learning_rate": 8.578199052132702e-06, "loss": 2.8203, "step": 181 }, { "epoch": 0.21544835750221958, "grad_norm": 6.21875, "learning_rate": 8.625592417061612e-06, "loss": 3.0179, "step": 182 }, { "epoch": 0.21663213968629771, "grad_norm": 5.3125, "learning_rate": 8.672985781990521e-06, "loss": 2.9326, "step": 183 }, { "epoch": 0.21781592187037585, "grad_norm": 5.71875, "learning_rate": 8.720379146919431e-06, "loss": 2.8996, "step": 184 }, { "epoch": 0.218999704054454, "grad_norm": 4.875, "learning_rate": 8.767772511848342e-06, "loss": 2.7646, "step": 185 }, { "epoch": 0.22018348623853212, "grad_norm": 5.96875, "learning_rate": 8.815165876777252e-06, "loss": 3.1929, "step": 186 }, { "epoch": 0.22136726842261023, "grad_norm": 5.6875, "learning_rate": 8.862559241706162e-06, "loss": 2.8751, "step": 187 }, { "epoch": 0.22255105060668837, "grad_norm": 5.75, "learning_rate": 8.909952606635071e-06, "loss": 3.2585, "step": 188 }, { "epoch": 0.2237348327907665, "grad_norm": 6.625, "learning_rate": 8.957345971563981e-06, "loss": 3.6785, "step": 189 }, { "epoch": 0.22491861497484464, "grad_norm": 5.0625, "learning_rate": 9.004739336492892e-06, "loss": 2.5239, "step": 190 }, { "epoch": 0.22610239715892275, "grad_norm": 5.78125, "learning_rate": 9.052132701421802e-06, "loss": 2.8453, "step": 191 }, { "epoch": 0.22728617934300088, "grad_norm": 5.375, "learning_rate": 9.09952606635071e-06, "loss": 2.9407, "step": 192 }, { "epoch": 0.22846996152707902, "grad_norm": 5.40625, "learning_rate": 9.146919431279621e-06, "loss": 2.7989, "step": 193 }, { "epoch": 0.22965374371115715, "grad_norm": 5.875, "learning_rate": 9.194312796208532e-06, "loss": 3.2391, "step": 194 }, { "epoch": 0.2308375258952353, "grad_norm": 6.15625, "learning_rate": 9.241706161137442e-06, "loss": 2.8217, "step": 195 }, { "epoch": 0.2320213080793134, "grad_norm": 5.375, "learning_rate": 9.289099526066352e-06, "loss": 3.0008, "step": 196 }, { "epoch": 0.23320509026339153, "grad_norm": 4.75, "learning_rate": 9.336492890995261e-06, "loss": 2.6389, "step": 197 }, { "epoch": 0.23438887244746967, "grad_norm": 5.03125, "learning_rate": 9.383886255924171e-06, "loss": 2.7761, "step": 198 }, { "epoch": 0.2355726546315478, "grad_norm": 5.4375, "learning_rate": 9.431279620853082e-06, "loss": 2.6324, "step": 199 }, { "epoch": 0.23675643681562591, "grad_norm": 5.6875, "learning_rate": 9.478672985781992e-06, "loss": 2.7339, "step": 200 }, { "epoch": 0.23794021899970405, "grad_norm": 5.09375, "learning_rate": 9.5260663507109e-06, "loss": 3.0592, "step": 201 }, { "epoch": 0.2391240011837822, "grad_norm": 6.21875, "learning_rate": 9.573459715639811e-06, "loss": 2.7575, "step": 202 }, { "epoch": 0.24030778336786032, "grad_norm": 6.0625, "learning_rate": 9.620853080568721e-06, "loss": 3.4123, "step": 203 }, { "epoch": 0.24149156555193843, "grad_norm": 5.3125, "learning_rate": 9.668246445497632e-06, "loss": 2.4718, "step": 204 }, { "epoch": 0.24267534773601657, "grad_norm": 4.9375, "learning_rate": 9.715639810426542e-06, "loss": 2.8354, "step": 205 }, { "epoch": 0.2438591299200947, "grad_norm": 4.8125, "learning_rate": 9.76303317535545e-06, "loss": 2.9036, "step": 206 }, { "epoch": 0.24504291210417284, "grad_norm": 5.59375, "learning_rate": 9.810426540284361e-06, "loss": 2.9836, "step": 207 }, { "epoch": 0.24622669428825097, "grad_norm": 5.59375, "learning_rate": 9.85781990521327e-06, "loss": 2.9898, "step": 208 }, { "epoch": 0.24741047647232908, "grad_norm": 6.5, "learning_rate": 9.905213270142182e-06, "loss": 2.5649, "step": 209 }, { "epoch": 0.24859425865640722, "grad_norm": 4.9375, "learning_rate": 9.95260663507109e-06, "loss": 2.5127, "step": 210 }, { "epoch": 0.24977804084048535, "grad_norm": 4.75, "learning_rate": 1e-05, "loss": 2.7704, "step": 211 }, { "epoch": 0.2509618230245635, "grad_norm": 4.9375, "learning_rate": 1.004739336492891e-05, "loss": 2.6617, "step": 212 }, { "epoch": 0.2521456052086416, "grad_norm": 5.46875, "learning_rate": 1.0094786729857822e-05, "loss": 3.0098, "step": 213 }, { "epoch": 0.25332938739271976, "grad_norm": 4.8125, "learning_rate": 1.0142180094786732e-05, "loss": 2.8491, "step": 214 }, { "epoch": 0.25451316957679787, "grad_norm": 4.8125, "learning_rate": 1.018957345971564e-05, "loss": 2.9454, "step": 215 }, { "epoch": 0.255696951760876, "grad_norm": 5.1875, "learning_rate": 1.023696682464455e-05, "loss": 2.9804, "step": 216 }, { "epoch": 0.25688073394495414, "grad_norm": 5.21875, "learning_rate": 1.0284360189573461e-05, "loss": 2.5158, "step": 217 }, { "epoch": 0.25806451612903225, "grad_norm": 4.90625, "learning_rate": 1.033175355450237e-05, "loss": 2.4959, "step": 218 }, { "epoch": 0.2592482983131104, "grad_norm": 5.15625, "learning_rate": 1.037914691943128e-05, "loss": 2.76, "step": 219 }, { "epoch": 0.2604320804971885, "grad_norm": 5.375, "learning_rate": 1.0426540284360192e-05, "loss": 3.1584, "step": 220 }, { "epoch": 0.26161586268126663, "grad_norm": 5.46875, "learning_rate": 1.0473933649289101e-05, "loss": 2.9186, "step": 221 }, { "epoch": 0.2627996448653448, "grad_norm": 5.0, "learning_rate": 1.052132701421801e-05, "loss": 2.549, "step": 222 }, { "epoch": 0.2639834270494229, "grad_norm": 5.625, "learning_rate": 1.056872037914692e-05, "loss": 2.8233, "step": 223 }, { "epoch": 0.265167209233501, "grad_norm": 5.71875, "learning_rate": 1.061611374407583e-05, "loss": 2.8856, "step": 224 }, { "epoch": 0.2663509914175792, "grad_norm": 4.53125, "learning_rate": 1.066350710900474e-05, "loss": 2.695, "step": 225 }, { "epoch": 0.2675347736016573, "grad_norm": 5.3125, "learning_rate": 1.071090047393365e-05, "loss": 2.4322, "step": 226 }, { "epoch": 0.26871855578573545, "grad_norm": 5.46875, "learning_rate": 1.075829383886256e-05, "loss": 2.9831, "step": 227 }, { "epoch": 0.26990233796981355, "grad_norm": 5.3125, "learning_rate": 1.080568720379147e-05, "loss": 2.7865, "step": 228 }, { "epoch": 0.27108612015389166, "grad_norm": 5.3125, "learning_rate": 1.085308056872038e-05, "loss": 2.853, "step": 229 }, { "epoch": 0.2722699023379698, "grad_norm": 5.25, "learning_rate": 1.0900473933649289e-05, "loss": 2.5822, "step": 230 }, { "epoch": 0.27345368452204794, "grad_norm": 5.25, "learning_rate": 1.0947867298578201e-05, "loss": 2.8897, "step": 231 }, { "epoch": 0.2746374667061261, "grad_norm": 4.375, "learning_rate": 1.099526066350711e-05, "loss": 2.3975, "step": 232 }, { "epoch": 0.2758212488902042, "grad_norm": 5.03125, "learning_rate": 1.104265402843602e-05, "loss": 2.6928, "step": 233 }, { "epoch": 0.2770050310742823, "grad_norm": 5.9375, "learning_rate": 1.1090047393364929e-05, "loss": 3.2237, "step": 234 }, { "epoch": 0.2781888132583605, "grad_norm": 5.75, "learning_rate": 1.1137440758293841e-05, "loss": 3.2215, "step": 235 }, { "epoch": 0.2793725954424386, "grad_norm": 4.46875, "learning_rate": 1.118483412322275e-05, "loss": 2.8132, "step": 236 }, { "epoch": 0.2805563776265167, "grad_norm": 5.15625, "learning_rate": 1.1232227488151658e-05, "loss": 3.0152, "step": 237 }, { "epoch": 0.28174015981059486, "grad_norm": 5.25, "learning_rate": 1.127962085308057e-05, "loss": 3.1626, "step": 238 }, { "epoch": 0.28292394199467297, "grad_norm": 5.4375, "learning_rate": 1.132701421800948e-05, "loss": 2.6208, "step": 239 }, { "epoch": 0.28410772417875113, "grad_norm": 5.28125, "learning_rate": 1.137440758293839e-05, "loss": 2.4686, "step": 240 }, { "epoch": 0.28529150636282924, "grad_norm": 4.96875, "learning_rate": 1.1421800947867298e-05, "loss": 3.0794, "step": 241 }, { "epoch": 0.28647528854690735, "grad_norm": 4.78125, "learning_rate": 1.146919431279621e-05, "loss": 2.5172, "step": 242 }, { "epoch": 0.2876590707309855, "grad_norm": 4.84375, "learning_rate": 1.1516587677725119e-05, "loss": 2.6685, "step": 243 }, { "epoch": 0.2888428529150636, "grad_norm": 4.84375, "learning_rate": 1.1563981042654029e-05, "loss": 2.8994, "step": 244 }, { "epoch": 0.2900266350991418, "grad_norm": 4.8125, "learning_rate": 1.1611374407582941e-05, "loss": 3.1005, "step": 245 }, { "epoch": 0.2912104172832199, "grad_norm": 5.0625, "learning_rate": 1.165876777251185e-05, "loss": 2.8181, "step": 246 }, { "epoch": 0.292394199467298, "grad_norm": 5.09375, "learning_rate": 1.1706161137440758e-05, "loss": 2.7756, "step": 247 }, { "epoch": 0.29357798165137616, "grad_norm": 4.5, "learning_rate": 1.1753554502369669e-05, "loss": 2.5114, "step": 248 }, { "epoch": 0.29476176383545427, "grad_norm": 5.53125, "learning_rate": 1.180094786729858e-05, "loss": 2.8279, "step": 249 }, { "epoch": 0.2959455460195324, "grad_norm": 5.03125, "learning_rate": 1.184834123222749e-05, "loss": 2.6553, "step": 250 }, { "epoch": 0.29712932820361054, "grad_norm": 4.8125, "learning_rate": 1.1895734597156398e-05, "loss": 2.6956, "step": 251 }, { "epoch": 0.29831311038768865, "grad_norm": 5.125, "learning_rate": 1.1943127962085309e-05, "loss": 2.7031, "step": 252 }, { "epoch": 0.2994968925717668, "grad_norm": 4.65625, "learning_rate": 1.1990521327014219e-05, "loss": 2.8315, "step": 253 }, { "epoch": 0.3006806747558449, "grad_norm": 5.28125, "learning_rate": 1.203791469194313e-05, "loss": 2.8987, "step": 254 }, { "epoch": 0.30186445693992303, "grad_norm": 4.46875, "learning_rate": 1.2085308056872038e-05, "loss": 2.7542, "step": 255 }, { "epoch": 0.3030482391240012, "grad_norm": 4.625, "learning_rate": 1.213270142180095e-05, "loss": 2.8007, "step": 256 }, { "epoch": 0.3042320213080793, "grad_norm": 5.5625, "learning_rate": 1.2180094786729859e-05, "loss": 3.22, "step": 257 }, { "epoch": 0.30541580349215747, "grad_norm": 5.15625, "learning_rate": 1.2227488151658769e-05, "loss": 2.4949, "step": 258 }, { "epoch": 0.3065995856762356, "grad_norm": 6.3125, "learning_rate": 1.2274881516587678e-05, "loss": 3.0903, "step": 259 }, { "epoch": 0.3077833678603137, "grad_norm": 4.71875, "learning_rate": 1.232227488151659e-05, "loss": 3.0174, "step": 260 }, { "epoch": 0.30896715004439185, "grad_norm": 5.3125, "learning_rate": 1.2369668246445498e-05, "loss": 2.7283, "step": 261 }, { "epoch": 0.31015093222846996, "grad_norm": 5.71875, "learning_rate": 1.2417061611374409e-05, "loss": 3.1703, "step": 262 }, { "epoch": 0.31133471441254806, "grad_norm": 5.53125, "learning_rate": 1.2464454976303319e-05, "loss": 2.4287, "step": 263 }, { "epoch": 0.3125184965966262, "grad_norm": 4.9375, "learning_rate": 1.251184834123223e-05, "loss": 2.8703, "step": 264 }, { "epoch": 0.31370227878070434, "grad_norm": 4.40625, "learning_rate": 1.2559241706161138e-05, "loss": 2.5718, "step": 265 }, { "epoch": 0.3148860609647825, "grad_norm": 4.21875, "learning_rate": 1.2606635071090047e-05, "loss": 2.4072, "step": 266 }, { "epoch": 0.3160698431488606, "grad_norm": 4.4375, "learning_rate": 1.2654028436018959e-05, "loss": 3.0766, "step": 267 }, { "epoch": 0.3172536253329387, "grad_norm": 4.3125, "learning_rate": 1.270142180094787e-05, "loss": 2.9474, "step": 268 }, { "epoch": 0.3184374075170169, "grad_norm": 6.09375, "learning_rate": 1.2748815165876778e-05, "loss": 2.851, "step": 269 }, { "epoch": 0.319621189701095, "grad_norm": 5.53125, "learning_rate": 1.2796208530805687e-05, "loss": 3.2042, "step": 270 }, { "epoch": 0.32080497188517315, "grad_norm": 5.0, "learning_rate": 1.2843601895734599e-05, "loss": 3.0455, "step": 271 }, { "epoch": 0.32198875406925126, "grad_norm": 4.71875, "learning_rate": 1.2890995260663507e-05, "loss": 3.0739, "step": 272 }, { "epoch": 0.32317253625332937, "grad_norm": 5.21875, "learning_rate": 1.2938388625592418e-05, "loss": 2.7899, "step": 273 }, { "epoch": 0.32435631843740753, "grad_norm": 5.09375, "learning_rate": 1.298578199052133e-05, "loss": 2.5534, "step": 274 }, { "epoch": 0.32554010062148564, "grad_norm": 4.96875, "learning_rate": 1.3033175355450238e-05, "loss": 3.0155, "step": 275 }, { "epoch": 0.3267238828055638, "grad_norm": 4.8125, "learning_rate": 1.3080568720379147e-05, "loss": 2.2551, "step": 276 }, { "epoch": 0.3279076649896419, "grad_norm": 4.84375, "learning_rate": 1.3127962085308057e-05, "loss": 2.9656, "step": 277 }, { "epoch": 0.32909144717372, "grad_norm": 5.25, "learning_rate": 1.3175355450236968e-05, "loss": 2.714, "step": 278 }, { "epoch": 0.3302752293577982, "grad_norm": 5.40625, "learning_rate": 1.3222748815165878e-05, "loss": 3.3093, "step": 279 }, { "epoch": 0.3314590115418763, "grad_norm": 4.78125, "learning_rate": 1.3270142180094787e-05, "loss": 2.8451, "step": 280 }, { "epoch": 0.3326427937259544, "grad_norm": 5.59375, "learning_rate": 1.3317535545023699e-05, "loss": 2.639, "step": 281 }, { "epoch": 0.33382657591003256, "grad_norm": 5.1875, "learning_rate": 1.3364928909952607e-05, "loss": 2.4743, "step": 282 }, { "epoch": 0.33501035809411067, "grad_norm": 4.71875, "learning_rate": 1.3412322274881518e-05, "loss": 2.6672, "step": 283 }, { "epoch": 0.33619414027818884, "grad_norm": 6.84375, "learning_rate": 1.3459715639810426e-05, "loss": 2.0719, "step": 284 }, { "epoch": 0.33737792246226694, "grad_norm": 4.8125, "learning_rate": 1.3507109004739339e-05, "loss": 2.91, "step": 285 }, { "epoch": 0.33856170464634505, "grad_norm": 5.15625, "learning_rate": 1.3554502369668247e-05, "loss": 2.9267, "step": 286 }, { "epoch": 0.3397454868304232, "grad_norm": 5.4375, "learning_rate": 1.3601895734597158e-05, "loss": 2.8538, "step": 287 }, { "epoch": 0.3409292690145013, "grad_norm": 4.8125, "learning_rate": 1.3649289099526068e-05, "loss": 2.255, "step": 288 }, { "epoch": 0.3421130511985795, "grad_norm": 4.9375, "learning_rate": 1.3696682464454978e-05, "loss": 2.8949, "step": 289 }, { "epoch": 0.3432968333826576, "grad_norm": 4.75, "learning_rate": 1.3744075829383887e-05, "loss": 2.7718, "step": 290 }, { "epoch": 0.3444806155667357, "grad_norm": 4.4375, "learning_rate": 1.3791469194312797e-05, "loss": 2.879, "step": 291 }, { "epoch": 0.34566439775081387, "grad_norm": 5.28125, "learning_rate": 1.3838862559241708e-05, "loss": 2.9023, "step": 292 }, { "epoch": 0.346848179934892, "grad_norm": 4.65625, "learning_rate": 1.3886255924170618e-05, "loss": 2.5042, "step": 293 }, { "epoch": 0.3480319621189701, "grad_norm": 4.8125, "learning_rate": 1.3933649289099527e-05, "loss": 2.7081, "step": 294 }, { "epoch": 0.34921574430304825, "grad_norm": 4.53125, "learning_rate": 1.3981042654028435e-05, "loss": 2.6192, "step": 295 }, { "epoch": 0.35039952648712636, "grad_norm": 4.75, "learning_rate": 1.4028436018957347e-05, "loss": 2.6729, "step": 296 }, { "epoch": 0.3515833086712045, "grad_norm": 4.75, "learning_rate": 1.4075829383886258e-05, "loss": 2.7472, "step": 297 }, { "epoch": 0.35276709085528263, "grad_norm": 4.625, "learning_rate": 1.4123222748815166e-05, "loss": 2.6944, "step": 298 }, { "epoch": 0.35395087303936074, "grad_norm": 4.875, "learning_rate": 1.4170616113744078e-05, "loss": 2.8301, "step": 299 }, { "epoch": 0.3551346552234389, "grad_norm": 4.75, "learning_rate": 1.4218009478672987e-05, "loss": 2.5333, "step": 300 }, { "epoch": 0.356318437407517, "grad_norm": 4.625, "learning_rate": 1.4265402843601896e-05, "loss": 2.6664, "step": 301 }, { "epoch": 0.35750221959159517, "grad_norm": 4.875, "learning_rate": 1.4312796208530806e-05, "loss": 2.6661, "step": 302 }, { "epoch": 0.3586860017756733, "grad_norm": 5.15625, "learning_rate": 1.4360189573459718e-05, "loss": 2.8139, "step": 303 }, { "epoch": 0.3598697839597514, "grad_norm": 4.84375, "learning_rate": 1.4407582938388627e-05, "loss": 2.8003, "step": 304 }, { "epoch": 0.36105356614382955, "grad_norm": 5.25, "learning_rate": 1.4454976303317535e-05, "loss": 2.7978, "step": 305 }, { "epoch": 0.36223734832790766, "grad_norm": 5.75, "learning_rate": 1.4502369668246448e-05, "loss": 3.0917, "step": 306 }, { "epoch": 0.36342113051198577, "grad_norm": 4.875, "learning_rate": 1.4549763033175356e-05, "loss": 2.7671, "step": 307 }, { "epoch": 0.36460491269606393, "grad_norm": 4.8125, "learning_rate": 1.4597156398104267e-05, "loss": 2.5533, "step": 308 }, { "epoch": 0.36578869488014204, "grad_norm": 5.09375, "learning_rate": 1.4644549763033175e-05, "loss": 2.6768, "step": 309 }, { "epoch": 0.3669724770642202, "grad_norm": 5.4375, "learning_rate": 1.4691943127962087e-05, "loss": 2.8195, "step": 310 }, { "epoch": 0.3681562592482983, "grad_norm": 4.46875, "learning_rate": 1.4739336492890996e-05, "loss": 2.5481, "step": 311 }, { "epoch": 0.3693400414323764, "grad_norm": 5.125, "learning_rate": 1.4786729857819906e-05, "loss": 2.8285, "step": 312 }, { "epoch": 0.3705238236164546, "grad_norm": 5.0625, "learning_rate": 1.4834123222748817e-05, "loss": 2.6925, "step": 313 }, { "epoch": 0.3717076058005327, "grad_norm": 5.34375, "learning_rate": 1.4881516587677727e-05, "loss": 3.0766, "step": 314 }, { "epoch": 0.37289138798461086, "grad_norm": 5.09375, "learning_rate": 1.4928909952606636e-05, "loss": 3.1, "step": 315 }, { "epoch": 0.37407517016868896, "grad_norm": 4.90625, "learning_rate": 1.4976303317535546e-05, "loss": 2.9711, "step": 316 }, { "epoch": 0.3752589523527671, "grad_norm": 5.4375, "learning_rate": 1.5023696682464456e-05, "loss": 2.8579, "step": 317 }, { "epoch": 0.37644273453684524, "grad_norm": 4.5, "learning_rate": 1.5071090047393367e-05, "loss": 2.7418, "step": 318 }, { "epoch": 0.37762651672092334, "grad_norm": 4.625, "learning_rate": 1.5118483412322275e-05, "loss": 2.4723, "step": 319 }, { "epoch": 0.37881029890500145, "grad_norm": 4.96875, "learning_rate": 1.5165876777251186e-05, "loss": 2.635, "step": 320 }, { "epoch": 0.3799940810890796, "grad_norm": 4.78125, "learning_rate": 1.5213270142180096e-05, "loss": 3.0549, "step": 321 }, { "epoch": 0.3811778632731577, "grad_norm": 4.75, "learning_rate": 1.5260663507109007e-05, "loss": 2.9297, "step": 322 }, { "epoch": 0.3823616454572359, "grad_norm": 4.71875, "learning_rate": 1.5308056872037915e-05, "loss": 2.708, "step": 323 }, { "epoch": 0.383545427641314, "grad_norm": 4.65625, "learning_rate": 1.5355450236966827e-05, "loss": 2.77, "step": 324 }, { "epoch": 0.3847292098253921, "grad_norm": 5.0625, "learning_rate": 1.5402843601895736e-05, "loss": 2.819, "step": 325 }, { "epoch": 0.38591299200947027, "grad_norm": 4.875, "learning_rate": 1.5450236966824645e-05, "loss": 2.5151, "step": 326 }, { "epoch": 0.3870967741935484, "grad_norm": 4.78125, "learning_rate": 1.5497630331753553e-05, "loss": 2.5492, "step": 327 }, { "epoch": 0.38828055637762654, "grad_norm": 5.21875, "learning_rate": 1.5545023696682465e-05, "loss": 3.1629, "step": 328 }, { "epoch": 0.38946433856170465, "grad_norm": 5.21875, "learning_rate": 1.5592417061611377e-05, "loss": 2.7724, "step": 329 }, { "epoch": 0.39064812074578276, "grad_norm": 5.1875, "learning_rate": 1.5639810426540286e-05, "loss": 2.7371, "step": 330 }, { "epoch": 0.3918319029298609, "grad_norm": 5.0, "learning_rate": 1.5687203791469198e-05, "loss": 2.5081, "step": 331 }, { "epoch": 0.39301568511393903, "grad_norm": 4.9375, "learning_rate": 1.5734597156398107e-05, "loss": 2.8753, "step": 332 }, { "epoch": 0.3941994672980172, "grad_norm": 5.21875, "learning_rate": 1.5781990521327015e-05, "loss": 2.9915, "step": 333 }, { "epoch": 0.3953832494820953, "grad_norm": 4.53125, "learning_rate": 1.5829383886255924e-05, "loss": 2.6159, "step": 334 }, { "epoch": 0.3965670316661734, "grad_norm": 4.84375, "learning_rate": 1.5876777251184836e-05, "loss": 2.6686, "step": 335 }, { "epoch": 0.39775081385025157, "grad_norm": 4.375, "learning_rate": 1.5924170616113745e-05, "loss": 2.7223, "step": 336 }, { "epoch": 0.3989345960343297, "grad_norm": 5.8125, "learning_rate": 1.5971563981042653e-05, "loss": 3.0036, "step": 337 }, { "epoch": 0.4001183782184078, "grad_norm": 4.96875, "learning_rate": 1.6018957345971565e-05, "loss": 3.0889, "step": 338 }, { "epoch": 0.40130216040248595, "grad_norm": 4.75, "learning_rate": 1.6066350710900474e-05, "loss": 2.4813, "step": 339 }, { "epoch": 0.40248594258656406, "grad_norm": 5.0625, "learning_rate": 1.6113744075829386e-05, "loss": 2.5486, "step": 340 }, { "epoch": 0.4036697247706422, "grad_norm": 4.75, "learning_rate": 1.6161137440758295e-05, "loss": 2.6352, "step": 341 }, { "epoch": 0.40485350695472033, "grad_norm": 4.65625, "learning_rate": 1.6208530805687207e-05, "loss": 2.944, "step": 342 }, { "epoch": 0.40603728913879844, "grad_norm": 4.53125, "learning_rate": 1.6255924170616116e-05, "loss": 2.7707, "step": 343 }, { "epoch": 0.4072210713228766, "grad_norm": 4.75, "learning_rate": 1.6303317535545024e-05, "loss": 2.5338, "step": 344 }, { "epoch": 0.4084048535069547, "grad_norm": 4.875, "learning_rate": 1.6350710900473933e-05, "loss": 2.764, "step": 345 }, { "epoch": 0.4095886356910329, "grad_norm": 5.34375, "learning_rate": 1.6398104265402845e-05, "loss": 2.8279, "step": 346 }, { "epoch": 0.410772417875111, "grad_norm": 5.125, "learning_rate": 1.6445497630331754e-05, "loss": 2.8334, "step": 347 }, { "epoch": 0.4119562000591891, "grad_norm": 4.5625, "learning_rate": 1.6492890995260666e-05, "loss": 2.807, "step": 348 }, { "epoch": 0.41313998224326726, "grad_norm": 4.6875, "learning_rate": 1.6540284360189574e-05, "loss": 2.7753, "step": 349 }, { "epoch": 0.41432376442734536, "grad_norm": 4.5625, "learning_rate": 1.6587677725118486e-05, "loss": 2.7581, "step": 350 }, { "epoch": 0.4155075466114235, "grad_norm": 5.3125, "learning_rate": 1.6635071090047395e-05, "loss": 2.4765, "step": 351 }, { "epoch": 0.41669132879550164, "grad_norm": 4.78125, "learning_rate": 1.6682464454976304e-05, "loss": 2.5752, "step": 352 }, { "epoch": 0.41787511097957974, "grad_norm": 4.6875, "learning_rate": 1.6729857819905216e-05, "loss": 2.8467, "step": 353 }, { "epoch": 0.4190588931636579, "grad_norm": 5.03125, "learning_rate": 1.6777251184834124e-05, "loss": 2.4303, "step": 354 }, { "epoch": 0.420242675347736, "grad_norm": 4.65625, "learning_rate": 1.6824644549763033e-05, "loss": 2.7506, "step": 355 }, { "epoch": 0.4214264575318141, "grad_norm": 4.71875, "learning_rate": 1.6872037914691945e-05, "loss": 2.6954, "step": 356 }, { "epoch": 0.4226102397158923, "grad_norm": 5.0625, "learning_rate": 1.6919431279620854e-05, "loss": 2.6394, "step": 357 }, { "epoch": 0.4237940218999704, "grad_norm": 4.5, "learning_rate": 1.6966824644549766e-05, "loss": 2.6936, "step": 358 }, { "epoch": 0.42497780408404856, "grad_norm": 5.25, "learning_rate": 1.7014218009478674e-05, "loss": 2.8144, "step": 359 }, { "epoch": 0.42616158626812667, "grad_norm": 5.09375, "learning_rate": 1.7061611374407587e-05, "loss": 2.9599, "step": 360 }, { "epoch": 0.4273453684522048, "grad_norm": 4.8125, "learning_rate": 1.7109004739336495e-05, "loss": 2.6884, "step": 361 }, { "epoch": 0.42852915063628294, "grad_norm": 5.46875, "learning_rate": 1.7156398104265404e-05, "loss": 2.4406, "step": 362 }, { "epoch": 0.42971293282036105, "grad_norm": 4.875, "learning_rate": 1.7203791469194316e-05, "loss": 2.8712, "step": 363 }, { "epoch": 0.43089671500443916, "grad_norm": 4.59375, "learning_rate": 1.7251184834123225e-05, "loss": 2.6344, "step": 364 }, { "epoch": 0.4320804971885173, "grad_norm": 5.0625, "learning_rate": 1.7298578199052133e-05, "loss": 2.9339, "step": 365 }, { "epoch": 0.43326427937259543, "grad_norm": 5.6875, "learning_rate": 1.7345971563981042e-05, "loss": 3.0146, "step": 366 }, { "epoch": 0.4344480615566736, "grad_norm": 5.0625, "learning_rate": 1.7393364928909954e-05, "loss": 2.8382, "step": 367 }, { "epoch": 0.4356318437407517, "grad_norm": 5.15625, "learning_rate": 1.7440758293838863e-05, "loss": 3.0245, "step": 368 }, { "epoch": 0.4368156259248298, "grad_norm": 4.875, "learning_rate": 1.7488151658767775e-05, "loss": 2.8044, "step": 369 }, { "epoch": 0.437999408108908, "grad_norm": 4.84375, "learning_rate": 1.7535545023696683e-05, "loss": 2.9253, "step": 370 }, { "epoch": 0.4391831902929861, "grad_norm": 4.84375, "learning_rate": 1.7582938388625595e-05, "loss": 3.1525, "step": 371 }, { "epoch": 0.44036697247706424, "grad_norm": 4.875, "learning_rate": 1.7630331753554504e-05, "loss": 2.5946, "step": 372 }, { "epoch": 0.44155075466114235, "grad_norm": 5.28125, "learning_rate": 1.7677725118483413e-05, "loss": 2.5128, "step": 373 }, { "epoch": 0.44273453684522046, "grad_norm": 4.5625, "learning_rate": 1.7725118483412325e-05, "loss": 3.1451, "step": 374 }, { "epoch": 0.4439183190292986, "grad_norm": 4.6875, "learning_rate": 1.7772511848341233e-05, "loss": 2.855, "step": 375 }, { "epoch": 0.44510210121337673, "grad_norm": 4.9375, "learning_rate": 1.7819905213270142e-05, "loss": 2.9335, "step": 376 }, { "epoch": 0.44628588339745484, "grad_norm": 4.65625, "learning_rate": 1.7867298578199054e-05, "loss": 2.7664, "step": 377 }, { "epoch": 0.447469665581533, "grad_norm": 4.75, "learning_rate": 1.7914691943127963e-05, "loss": 2.309, "step": 378 }, { "epoch": 0.4486534477656111, "grad_norm": 4.40625, "learning_rate": 1.7962085308056875e-05, "loss": 3.0141, "step": 379 }, { "epoch": 0.4498372299496893, "grad_norm": 4.96875, "learning_rate": 1.8009478672985784e-05, "loss": 2.3059, "step": 380 }, { "epoch": 0.4510210121337674, "grad_norm": 4.90625, "learning_rate": 1.8056872037914696e-05, "loss": 3.252, "step": 381 }, { "epoch": 0.4522047943178455, "grad_norm": 5.34375, "learning_rate": 1.8104265402843604e-05, "loss": 3.0884, "step": 382 }, { "epoch": 0.45338857650192366, "grad_norm": 4.5625, "learning_rate": 1.8151658767772513e-05, "loss": 2.4152, "step": 383 }, { "epoch": 0.45457235868600177, "grad_norm": 4.6875, "learning_rate": 1.819905213270142e-05, "loss": 2.4958, "step": 384 }, { "epoch": 0.45575614087007993, "grad_norm": 4.375, "learning_rate": 1.8246445497630334e-05, "loss": 2.7179, "step": 385 }, { "epoch": 0.45693992305415804, "grad_norm": 4.40625, "learning_rate": 1.8293838862559242e-05, "loss": 2.9143, "step": 386 }, { "epoch": 0.45812370523823615, "grad_norm": 4.5, "learning_rate": 1.8341232227488154e-05, "loss": 2.4235, "step": 387 }, { "epoch": 0.4593074874223143, "grad_norm": 4.78125, "learning_rate": 1.8388625592417063e-05, "loss": 3.1037, "step": 388 }, { "epoch": 0.4604912696063924, "grad_norm": 4.46875, "learning_rate": 1.8436018957345975e-05, "loss": 2.7058, "step": 389 }, { "epoch": 0.4616750517904706, "grad_norm": 4.53125, "learning_rate": 1.8483412322274884e-05, "loss": 2.7252, "step": 390 }, { "epoch": 0.4628588339745487, "grad_norm": 4.84375, "learning_rate": 1.8530805687203792e-05, "loss": 3.0053, "step": 391 }, { "epoch": 0.4640426161586268, "grad_norm": 4.75, "learning_rate": 1.8578199052132704e-05, "loss": 2.5684, "step": 392 }, { "epoch": 0.46522639834270496, "grad_norm": 4.625, "learning_rate": 1.8625592417061613e-05, "loss": 2.5282, "step": 393 }, { "epoch": 0.46641018052678307, "grad_norm": 4.9375, "learning_rate": 1.8672985781990522e-05, "loss": 2.4713, "step": 394 }, { "epoch": 0.4675939627108612, "grad_norm": 4.84375, "learning_rate": 1.872037914691943e-05, "loss": 2.4372, "step": 395 }, { "epoch": 0.46877774489493934, "grad_norm": 4.4375, "learning_rate": 1.8767772511848342e-05, "loss": 2.9421, "step": 396 }, { "epoch": 0.46996152707901745, "grad_norm": 5.4375, "learning_rate": 1.881516587677725e-05, "loss": 3.1285, "step": 397 }, { "epoch": 0.4711453092630956, "grad_norm": 4.28125, "learning_rate": 1.8862559241706163e-05, "loss": 2.2699, "step": 398 }, { "epoch": 0.4723290914471737, "grad_norm": 4.46875, "learning_rate": 1.8909952606635075e-05, "loss": 2.2156, "step": 399 }, { "epoch": 0.47351287363125183, "grad_norm": 5.25, "learning_rate": 1.8957345971563984e-05, "loss": 3.0645, "step": 400 }, { "epoch": 0.47469665581533, "grad_norm": 4.4375, "learning_rate": 1.9004739336492893e-05, "loss": 2.972, "step": 401 }, { "epoch": 0.4758804379994081, "grad_norm": 4.6875, "learning_rate": 1.90521327014218e-05, "loss": 2.6146, "step": 402 }, { "epoch": 0.47706422018348627, "grad_norm": 4.6875, "learning_rate": 1.9099526066350713e-05, "loss": 3.0106, "step": 403 }, { "epoch": 0.4782480023675644, "grad_norm": 4.875, "learning_rate": 1.9146919431279622e-05, "loss": 2.7097, "step": 404 }, { "epoch": 0.4794317845516425, "grad_norm": 4.4375, "learning_rate": 1.919431279620853e-05, "loss": 2.8993, "step": 405 }, { "epoch": 0.48061556673572065, "grad_norm": 4.59375, "learning_rate": 1.9241706161137443e-05, "loss": 2.6624, "step": 406 }, { "epoch": 0.48179934891979875, "grad_norm": 5.1875, "learning_rate": 1.928909952606635e-05, "loss": 2.8009, "step": 407 }, { "epoch": 0.48298313110387686, "grad_norm": 4.78125, "learning_rate": 1.9336492890995263e-05, "loss": 3.189, "step": 408 }, { "epoch": 0.484166913287955, "grad_norm": 4.40625, "learning_rate": 1.9383886255924172e-05, "loss": 2.7469, "step": 409 }, { "epoch": 0.48535069547203313, "grad_norm": 4.875, "learning_rate": 1.9431279620853084e-05, "loss": 2.4736, "step": 410 }, { "epoch": 0.4865344776561113, "grad_norm": 4.6875, "learning_rate": 1.9478672985781993e-05, "loss": 2.8223, "step": 411 }, { "epoch": 0.4877182598401894, "grad_norm": 4.84375, "learning_rate": 1.95260663507109e-05, "loss": 2.9889, "step": 412 }, { "epoch": 0.4889020420242675, "grad_norm": 4.28125, "learning_rate": 1.957345971563981e-05, "loss": 2.9711, "step": 413 }, { "epoch": 0.4900858242083457, "grad_norm": 4.875, "learning_rate": 1.9620853080568722e-05, "loss": 2.7299, "step": 414 }, { "epoch": 0.4912696063924238, "grad_norm": 4.75, "learning_rate": 1.966824644549763e-05, "loss": 2.7704, "step": 415 }, { "epoch": 0.49245338857650195, "grad_norm": 4.84375, "learning_rate": 1.971563981042654e-05, "loss": 3.0171, "step": 416 }, { "epoch": 0.49363717076058006, "grad_norm": 4.34375, "learning_rate": 1.976303317535545e-05, "loss": 2.7554, "step": 417 }, { "epoch": 0.49482095294465817, "grad_norm": 4.46875, "learning_rate": 1.9810426540284364e-05, "loss": 2.7022, "step": 418 }, { "epoch": 0.49600473512873633, "grad_norm": 4.21875, "learning_rate": 1.9857819905213272e-05, "loss": 2.7364, "step": 419 }, { "epoch": 0.49718851731281444, "grad_norm": 4.46875, "learning_rate": 1.990521327014218e-05, "loss": 2.5352, "step": 420 }, { "epoch": 0.49837229949689255, "grad_norm": 4.65625, "learning_rate": 1.9952606635071093e-05, "loss": 2.6267, "step": 421 }, { "epoch": 0.4995560816809707, "grad_norm": 4.59375, "learning_rate": 2e-05, "loss": 2.9234, "step": 422 }, { "epoch": 0.5007398638650489, "grad_norm": 5.34375, "learning_rate": 1.9999996578946986e-05, "loss": 2.8771, "step": 423 }, { "epoch": 0.501923646049127, "grad_norm": 4.59375, "learning_rate": 1.9999986315790288e-05, "loss": 2.5393, "step": 424 }, { "epoch": 0.5031074282332051, "grad_norm": 4.46875, "learning_rate": 1.9999969210536924e-05, "loss": 2.4292, "step": 425 }, { "epoch": 0.5042912104172832, "grad_norm": 4.5625, "learning_rate": 1.99999452631986e-05, "loss": 2.9455, "step": 426 }, { "epoch": 0.5054749926013613, "grad_norm": 5.4375, "learning_rate": 1.9999914473791704e-05, "loss": 2.8429, "step": 427 }, { "epoch": 0.5066587747854395, "grad_norm": 4.90625, "learning_rate": 1.9999876842337294e-05, "loss": 2.9468, "step": 428 }, { "epoch": 0.5078425569695176, "grad_norm": 5.25, "learning_rate": 1.9999832368861126e-05, "loss": 2.9273, "step": 429 }, { "epoch": 0.5090263391535957, "grad_norm": 5.15625, "learning_rate": 1.9999781053393626e-05, "loss": 3.2937, "step": 430 }, { "epoch": 0.5102101213376738, "grad_norm": 4.9375, "learning_rate": 1.9999722895969904e-05, "loss": 2.6501, "step": 431 }, { "epoch": 0.511393903521752, "grad_norm": 4.34375, "learning_rate": 1.9999657896629752e-05, "loss": 2.7118, "step": 432 }, { "epoch": 0.5125776857058302, "grad_norm": 4.84375, "learning_rate": 1.9999586055417645e-05, "loss": 2.617, "step": 433 }, { "epoch": 0.5137614678899083, "grad_norm": 4.1875, "learning_rate": 1.9999507372382738e-05, "loss": 2.8919, "step": 434 }, { "epoch": 0.5149452500739864, "grad_norm": 4.8125, "learning_rate": 1.9999421847578864e-05, "loss": 2.7508, "step": 435 }, { "epoch": 0.5161290322580645, "grad_norm": 4.59375, "learning_rate": 1.999932948106454e-05, "loss": 2.5654, "step": 436 }, { "epoch": 0.5173128144421426, "grad_norm": 4.78125, "learning_rate": 1.999923027290297e-05, "loss": 2.8408, "step": 437 }, { "epoch": 0.5184965966262208, "grad_norm": 4.5, "learning_rate": 1.9999124223162023e-05, "loss": 2.8589, "step": 438 }, { "epoch": 0.5196803788102989, "grad_norm": 4.375, "learning_rate": 1.9999011331914264e-05, "loss": 2.866, "step": 439 }, { "epoch": 0.520864160994377, "grad_norm": 4.71875, "learning_rate": 1.999889159923694e-05, "loss": 2.6634, "step": 440 }, { "epoch": 0.5220479431784552, "grad_norm": 4.40625, "learning_rate": 1.999876502521197e-05, "loss": 3.1033, "step": 441 }, { "epoch": 0.5232317253625333, "grad_norm": 4.34375, "learning_rate": 1.999863160992595e-05, "loss": 2.5757, "step": 442 }, { "epoch": 0.5244155075466114, "grad_norm": 4.875, "learning_rate": 1.9998491353470176e-05, "loss": 2.9225, "step": 443 }, { "epoch": 0.5255992897306896, "grad_norm": 4.46875, "learning_rate": 1.9998344255940602e-05, "loss": 2.9235, "step": 444 }, { "epoch": 0.5267830719147677, "grad_norm": 4.90625, "learning_rate": 1.9998190317437883e-05, "loss": 2.7662, "step": 445 }, { "epoch": 0.5279668540988458, "grad_norm": 4.5, "learning_rate": 1.999802953806734e-05, "loss": 3.0117, "step": 446 }, { "epoch": 0.5291506362829239, "grad_norm": 4.34375, "learning_rate": 1.999786191793898e-05, "loss": 2.8025, "step": 447 }, { "epoch": 0.530334418467002, "grad_norm": 4.3125, "learning_rate": 1.9997687457167495e-05, "loss": 2.7641, "step": 448 }, { "epoch": 0.5315182006510802, "grad_norm": 4.9375, "learning_rate": 1.9997506155872246e-05, "loss": 2.648, "step": 449 }, { "epoch": 0.5327019828351583, "grad_norm": 4.125, "learning_rate": 1.9997318014177283e-05, "loss": 2.5183, "step": 450 }, { "epoch": 0.5338857650192365, "grad_norm": 4.75, "learning_rate": 1.9997123032211343e-05, "loss": 3.0498, "step": 451 }, { "epoch": 0.5350695472033146, "grad_norm": 4.625, "learning_rate": 1.9996921210107823e-05, "loss": 2.6986, "step": 452 }, { "epoch": 0.5362533293873927, "grad_norm": 4.625, "learning_rate": 1.999671254800482e-05, "loss": 3.1015, "step": 453 }, { "epoch": 0.5374371115714709, "grad_norm": 5.15625, "learning_rate": 1.9996497046045096e-05, "loss": 3.0756, "step": 454 }, { "epoch": 0.538620893755549, "grad_norm": 4.09375, "learning_rate": 1.9996274704376105e-05, "loss": 2.7572, "step": 455 }, { "epoch": 0.5398046759396271, "grad_norm": 4.625, "learning_rate": 1.9996045523149977e-05, "loss": 2.8732, "step": 456 }, { "epoch": 0.5409884581237052, "grad_norm": 4.25, "learning_rate": 1.9995809502523513e-05, "loss": 2.4689, "step": 457 }, { "epoch": 0.5421722403077833, "grad_norm": 4.5, "learning_rate": 1.9995566642658208e-05, "loss": 2.5635, "step": 458 }, { "epoch": 0.5433560224918615, "grad_norm": 4.8125, "learning_rate": 1.9995316943720225e-05, "loss": 2.8135, "step": 459 }, { "epoch": 0.5445398046759397, "grad_norm": 4.375, "learning_rate": 1.9995060405880412e-05, "loss": 2.5494, "step": 460 }, { "epoch": 0.5457235868600178, "grad_norm": 4.59375, "learning_rate": 1.999479702931429e-05, "loss": 3.0223, "step": 461 }, { "epoch": 0.5469073690440959, "grad_norm": 4.8125, "learning_rate": 1.9994526814202076e-05, "loss": 2.4756, "step": 462 }, { "epoch": 0.548091151228174, "grad_norm": 4.53125, "learning_rate": 1.9994249760728645e-05, "loss": 2.7521, "step": 463 }, { "epoch": 0.5492749334122522, "grad_norm": 3.859375, "learning_rate": 1.9993965869083562e-05, "loss": 2.4649, "step": 464 }, { "epoch": 0.5504587155963303, "grad_norm": 4.75, "learning_rate": 1.999367513946107e-05, "loss": 2.6839, "step": 465 }, { "epoch": 0.5516424977804084, "grad_norm": 5.3125, "learning_rate": 1.9993377572060084e-05, "loss": 2.7087, "step": 466 }, { "epoch": 0.5528262799644865, "grad_norm": 4.59375, "learning_rate": 1.999307316708421e-05, "loss": 2.7079, "step": 467 }, { "epoch": 0.5540100621485646, "grad_norm": 4.375, "learning_rate": 1.9992761924741718e-05, "loss": 2.7192, "step": 468 }, { "epoch": 0.5551938443326427, "grad_norm": 4.9375, "learning_rate": 1.9992443845245572e-05, "loss": 2.8079, "step": 469 }, { "epoch": 0.556377626516721, "grad_norm": 4.1875, "learning_rate": 1.99921189288134e-05, "loss": 2.934, "step": 470 }, { "epoch": 0.5575614087007991, "grad_norm": 4.625, "learning_rate": 1.999178717566751e-05, "loss": 2.8011, "step": 471 }, { "epoch": 0.5587451908848772, "grad_norm": 4.34375, "learning_rate": 1.9991448586034897e-05, "loss": 2.5154, "step": 472 }, { "epoch": 0.5599289730689553, "grad_norm": 4.875, "learning_rate": 1.9991103160147226e-05, "loss": 3.0889, "step": 473 }, { "epoch": 0.5611127552530334, "grad_norm": 4.46875, "learning_rate": 1.9990750898240837e-05, "loss": 2.468, "step": 474 }, { "epoch": 0.5622965374371116, "grad_norm": 4.21875, "learning_rate": 1.9990391800556757e-05, "loss": 2.7099, "step": 475 }, { "epoch": 0.5634803196211897, "grad_norm": 4.6875, "learning_rate": 1.9990025867340683e-05, "loss": 2.6567, "step": 476 }, { "epoch": 0.5646641018052678, "grad_norm": 4.46875, "learning_rate": 1.998965309884299e-05, "loss": 2.4946, "step": 477 }, { "epoch": 0.5658478839893459, "grad_norm": 4.9375, "learning_rate": 1.9989273495318726e-05, "loss": 2.9824, "step": 478 }, { "epoch": 0.567031666173424, "grad_norm": 4.5, "learning_rate": 1.9988887057027627e-05, "loss": 2.6195, "step": 479 }, { "epoch": 0.5682154483575023, "grad_norm": 4.625, "learning_rate": 1.998849378423409e-05, "loss": 2.561, "step": 480 }, { "epoch": 0.5693992305415804, "grad_norm": 5.59375, "learning_rate": 1.998809367720721e-05, "loss": 2.6358, "step": 481 }, { "epoch": 0.5705830127256585, "grad_norm": 4.8125, "learning_rate": 1.9987686736220726e-05, "loss": 2.9791, "step": 482 }, { "epoch": 0.5717667949097366, "grad_norm": 4.78125, "learning_rate": 1.9987272961553087e-05, "loss": 2.774, "step": 483 }, { "epoch": 0.5729505770938147, "grad_norm": 4.625, "learning_rate": 1.998685235348739e-05, "loss": 2.9152, "step": 484 }, { "epoch": 0.5741343592778929, "grad_norm": 4.53125, "learning_rate": 1.998642491231143e-05, "loss": 2.9774, "step": 485 }, { "epoch": 0.575318141461971, "grad_norm": 4.4375, "learning_rate": 1.9985990638317663e-05, "loss": 2.6951, "step": 486 }, { "epoch": 0.5765019236460491, "grad_norm": 4.90625, "learning_rate": 1.998554953180322e-05, "loss": 3.0327, "step": 487 }, { "epoch": 0.5776857058301272, "grad_norm": 4.375, "learning_rate": 1.9985101593069914e-05, "loss": 2.8345, "step": 488 }, { "epoch": 0.5788694880142053, "grad_norm": 4.5625, "learning_rate": 1.998464682242423e-05, "loss": 2.2809, "step": 489 }, { "epoch": 0.5800532701982836, "grad_norm": 4.71875, "learning_rate": 1.9984185220177327e-05, "loss": 2.5339, "step": 490 }, { "epoch": 0.5812370523823617, "grad_norm": 3.890625, "learning_rate": 1.9983716786645035e-05, "loss": 2.8075, "step": 491 }, { "epoch": 0.5824208345664398, "grad_norm": 4.65625, "learning_rate": 1.9983241522147864e-05, "loss": 2.6906, "step": 492 }, { "epoch": 0.5836046167505179, "grad_norm": 4.4375, "learning_rate": 1.9982759427011e-05, "loss": 2.3835, "step": 493 }, { "epoch": 0.584788398934596, "grad_norm": 4.40625, "learning_rate": 1.9982270501564286e-05, "loss": 3.085, "step": 494 }, { "epoch": 0.5859721811186742, "grad_norm": 4.53125, "learning_rate": 1.9981774746142254e-05, "loss": 2.7208, "step": 495 }, { "epoch": 0.5871559633027523, "grad_norm": 4.875, "learning_rate": 1.9981272161084113e-05, "loss": 2.6536, "step": 496 }, { "epoch": 0.5883397454868304, "grad_norm": 4.9375, "learning_rate": 1.998076274673373e-05, "loss": 2.7804, "step": 497 }, { "epoch": 0.5895235276709085, "grad_norm": 4.8125, "learning_rate": 1.998024650343965e-05, "loss": 2.9326, "step": 498 }, { "epoch": 0.5907073098549867, "grad_norm": 5.125, "learning_rate": 1.9979723431555098e-05, "loss": 2.9943, "step": 499 }, { "epoch": 0.5918910920390648, "grad_norm": 4.625, "learning_rate": 1.9979193531437962e-05, "loss": 2.9678, "step": 500 }, { "epoch": 0.593074874223143, "grad_norm": 4.125, "learning_rate": 1.9978656803450804e-05, "loss": 2.6534, "step": 501 }, { "epoch": 0.5942586564072211, "grad_norm": 4.65625, "learning_rate": 1.9978113247960862e-05, "loss": 2.627, "step": 502 }, { "epoch": 0.5954424385912992, "grad_norm": 4.59375, "learning_rate": 1.997756286534004e-05, "loss": 2.6195, "step": 503 }, { "epoch": 0.5966262207753773, "grad_norm": 4.34375, "learning_rate": 1.9977005655964916e-05, "loss": 2.635, "step": 504 }, { "epoch": 0.5978100029594554, "grad_norm": 4.5, "learning_rate": 1.997644162021674e-05, "loss": 3.0525, "step": 505 }, { "epoch": 0.5989937851435336, "grad_norm": 4.9375, "learning_rate": 1.997587075848143e-05, "loss": 2.5361, "step": 506 }, { "epoch": 0.6001775673276117, "grad_norm": 4.5625, "learning_rate": 1.9975293071149577e-05, "loss": 2.6306, "step": 507 }, { "epoch": 0.6013613495116898, "grad_norm": 5.21875, "learning_rate": 1.997470855861644e-05, "loss": 3.0632, "step": 508 }, { "epoch": 0.602545131695768, "grad_norm": 5.3125, "learning_rate": 1.997411722128195e-05, "loss": 3.1031, "step": 509 }, { "epoch": 0.6037289138798461, "grad_norm": 4.59375, "learning_rate": 1.99735190595507e-05, "loss": 2.7346, "step": 510 }, { "epoch": 0.6049126960639243, "grad_norm": 4.28125, "learning_rate": 1.9972914073831967e-05, "loss": 2.9642, "step": 511 }, { "epoch": 0.6060964782480024, "grad_norm": 4.46875, "learning_rate": 1.9972302264539686e-05, "loss": 2.8106, "step": 512 }, { "epoch": 0.6072802604320805, "grad_norm": 4.46875, "learning_rate": 1.997168363209246e-05, "loss": 2.7211, "step": 513 }, { "epoch": 0.6084640426161586, "grad_norm": 4.03125, "learning_rate": 1.9971058176913566e-05, "loss": 2.4382, "step": 514 }, { "epoch": 0.6096478248002367, "grad_norm": 5.53125, "learning_rate": 1.997042589943095e-05, "loss": 2.7241, "step": 515 }, { "epoch": 0.6108316069843149, "grad_norm": 4.34375, "learning_rate": 1.9969786800077217e-05, "loss": 2.7381, "step": 516 }, { "epoch": 0.612015389168393, "grad_norm": 5.0, "learning_rate": 1.9969140879289654e-05, "loss": 2.8834, "step": 517 }, { "epoch": 0.6131991713524712, "grad_norm": 4.59375, "learning_rate": 1.99684881375102e-05, "loss": 2.8296, "step": 518 }, { "epoch": 0.6143829535365493, "grad_norm": 4.28125, "learning_rate": 1.9967828575185467e-05, "loss": 2.8824, "step": 519 }, { "epoch": 0.6155667357206274, "grad_norm": 4.8125, "learning_rate": 1.996716219276674e-05, "loss": 2.7783, "step": 520 }, { "epoch": 0.6167505179047056, "grad_norm": 4.03125, "learning_rate": 1.996648899070996e-05, "loss": 2.5804, "step": 521 }, { "epoch": 0.6179343000887837, "grad_norm": 4.96875, "learning_rate": 1.996580896947574e-05, "loss": 2.6926, "step": 522 }, { "epoch": 0.6191180822728618, "grad_norm": 5.1875, "learning_rate": 1.9965122129529362e-05, "loss": 2.6601, "step": 523 }, { "epoch": 0.6203018644569399, "grad_norm": 4.15625, "learning_rate": 1.9964428471340765e-05, "loss": 2.7183, "step": 524 }, { "epoch": 0.621485646641018, "grad_norm": 4.34375, "learning_rate": 1.9963727995384558e-05, "loss": 2.6511, "step": 525 }, { "epoch": 0.6226694288250961, "grad_norm": 4.3125, "learning_rate": 1.9963020702140016e-05, "loss": 2.6517, "step": 526 }, { "epoch": 0.6238532110091743, "grad_norm": 4.5625, "learning_rate": 1.9962306592091073e-05, "loss": 2.9489, "step": 527 }, { "epoch": 0.6250369931932525, "grad_norm": 4.4375, "learning_rate": 1.9961585665726332e-05, "loss": 2.6911, "step": 528 }, { "epoch": 0.6262207753773306, "grad_norm": 5.03125, "learning_rate": 1.996085792353906e-05, "loss": 2.9397, "step": 529 }, { "epoch": 0.6274045575614087, "grad_norm": 4.5, "learning_rate": 1.9960123366027187e-05, "loss": 2.6896, "step": 530 }, { "epoch": 0.6285883397454868, "grad_norm": 4.875, "learning_rate": 1.99593819936933e-05, "loss": 2.3817, "step": 531 }, { "epoch": 0.629772121929565, "grad_norm": 4.21875, "learning_rate": 1.9958633807044657e-05, "loss": 3.116, "step": 532 }, { "epoch": 0.6309559041136431, "grad_norm": 4.53125, "learning_rate": 1.9957878806593176e-05, "loss": 2.5594, "step": 533 }, { "epoch": 0.6321396862977212, "grad_norm": 4.125, "learning_rate": 1.9957116992855435e-05, "loss": 2.7295, "step": 534 }, { "epoch": 0.6333234684817993, "grad_norm": 4.375, "learning_rate": 1.9956348366352675e-05, "loss": 2.7084, "step": 535 }, { "epoch": 0.6345072506658774, "grad_norm": 4.75, "learning_rate": 1.9955572927610798e-05, "loss": 2.7115, "step": 536 }, { "epoch": 0.6356910328499557, "grad_norm": 4.4375, "learning_rate": 1.9954790677160368e-05, "loss": 2.9742, "step": 537 }, { "epoch": 0.6368748150340338, "grad_norm": 4.09375, "learning_rate": 1.9954001615536612e-05, "loss": 3.1781, "step": 538 }, { "epoch": 0.6380585972181119, "grad_norm": 4.65625, "learning_rate": 1.995320574327941e-05, "loss": 2.4473, "step": 539 }, { "epoch": 0.63924237940219, "grad_norm": 4.65625, "learning_rate": 1.995240306093331e-05, "loss": 2.8824, "step": 540 }, { "epoch": 0.6404261615862681, "grad_norm": 4.78125, "learning_rate": 1.9951593569047513e-05, "loss": 2.7433, "step": 541 }, { "epoch": 0.6416099437703463, "grad_norm": 4.1875, "learning_rate": 1.995077726817588e-05, "loss": 2.563, "step": 542 }, { "epoch": 0.6427937259544244, "grad_norm": 4.875, "learning_rate": 1.9949954158876938e-05, "loss": 2.9397, "step": 543 }, { "epoch": 0.6439775081385025, "grad_norm": 4.71875, "learning_rate": 1.994912424171386e-05, "loss": 3.1348, "step": 544 }, { "epoch": 0.6451612903225806, "grad_norm": 4.5, "learning_rate": 1.994828751725449e-05, "loss": 2.5916, "step": 545 }, { "epoch": 0.6463450725066587, "grad_norm": 3.9375, "learning_rate": 1.9947443986071328e-05, "loss": 2.6637, "step": 546 }, { "epoch": 0.647528854690737, "grad_norm": 4.40625, "learning_rate": 1.9946593648741515e-05, "loss": 2.5269, "step": 547 }, { "epoch": 0.6487126368748151, "grad_norm": 5.34375, "learning_rate": 1.9945736505846866e-05, "loss": 2.7976, "step": 548 }, { "epoch": 0.6498964190588932, "grad_norm": 4.25, "learning_rate": 1.994487255797385e-05, "loss": 2.747, "step": 549 }, { "epoch": 0.6510802012429713, "grad_norm": 4.375, "learning_rate": 1.994400180571359e-05, "loss": 2.2085, "step": 550 }, { "epoch": 0.6522639834270494, "grad_norm": 4.09375, "learning_rate": 1.994312424966186e-05, "loss": 2.7235, "step": 551 }, { "epoch": 0.6534477656111276, "grad_norm": 4.375, "learning_rate": 1.9942239890419097e-05, "loss": 3.007, "step": 552 }, { "epoch": 0.6546315477952057, "grad_norm": 4.40625, "learning_rate": 1.9941348728590385e-05, "loss": 2.6234, "step": 553 }, { "epoch": 0.6558153299792838, "grad_norm": 4.59375, "learning_rate": 1.9940450764785465e-05, "loss": 2.7641, "step": 554 }, { "epoch": 0.6569991121633619, "grad_norm": 4.78125, "learning_rate": 1.9939545999618737e-05, "loss": 2.4349, "step": 555 }, { "epoch": 0.65818289434744, "grad_norm": 4.125, "learning_rate": 1.9938634433709254e-05, "loss": 2.6999, "step": 556 }, { "epoch": 0.6593666765315181, "grad_norm": 4.53125, "learning_rate": 1.9937716067680712e-05, "loss": 2.5148, "step": 557 }, { "epoch": 0.6605504587155964, "grad_norm": 3.953125, "learning_rate": 1.993679090216147e-05, "loss": 3.0745, "step": 558 }, { "epoch": 0.6617342408996745, "grad_norm": 4.8125, "learning_rate": 1.9935858937784537e-05, "loss": 2.9972, "step": 559 }, { "epoch": 0.6629180230837526, "grad_norm": 4.90625, "learning_rate": 1.993492017518757e-05, "loss": 2.7369, "step": 560 }, { "epoch": 0.6641018052678307, "grad_norm": 4.59375, "learning_rate": 1.9933974615012888e-05, "loss": 2.73, "step": 561 }, { "epoch": 0.6652855874519088, "grad_norm": 4.65625, "learning_rate": 1.9933022257907445e-05, "loss": 2.5335, "step": 562 }, { "epoch": 0.666469369635987, "grad_norm": 4.6875, "learning_rate": 1.9932063104522857e-05, "loss": 2.3654, "step": 563 }, { "epoch": 0.6676531518200651, "grad_norm": 3.90625, "learning_rate": 1.9931097155515383e-05, "loss": 2.4536, "step": 564 }, { "epoch": 0.6688369340041432, "grad_norm": 4.625, "learning_rate": 1.9930124411545943e-05, "loss": 3.0691, "step": 565 }, { "epoch": 0.6700207161882213, "grad_norm": 4.71875, "learning_rate": 1.9929144873280092e-05, "loss": 2.7916, "step": 566 }, { "epoch": 0.6712044983722995, "grad_norm": 4.46875, "learning_rate": 1.9928158541388046e-05, "loss": 2.5077, "step": 567 }, { "epoch": 0.6723882805563777, "grad_norm": 5.3125, "learning_rate": 1.9927165416544656e-05, "loss": 3.0197, "step": 568 }, { "epoch": 0.6735720627404558, "grad_norm": 4.4375, "learning_rate": 1.9926165499429437e-05, "loss": 3.0066, "step": 569 }, { "epoch": 0.6747558449245339, "grad_norm": 4.34375, "learning_rate": 1.992515879072654e-05, "loss": 2.5028, "step": 570 }, { "epoch": 0.675939627108612, "grad_norm": 4.46875, "learning_rate": 1.9924145291124764e-05, "loss": 2.853, "step": 571 }, { "epoch": 0.6771234092926901, "grad_norm": 4.5, "learning_rate": 1.9923125001317557e-05, "loss": 2.9009, "step": 572 }, { "epoch": 0.6783071914767683, "grad_norm": 4.4375, "learning_rate": 1.9922097922003016e-05, "loss": 2.6961, "step": 573 }, { "epoch": 0.6794909736608464, "grad_norm": 4.21875, "learning_rate": 1.992106405388387e-05, "loss": 2.7008, "step": 574 }, { "epoch": 0.6806747558449245, "grad_norm": 4.15625, "learning_rate": 1.992002339766751e-05, "loss": 2.6626, "step": 575 }, { "epoch": 0.6818585380290026, "grad_norm": 3.96875, "learning_rate": 1.9918975954065964e-05, "loss": 2.5642, "step": 576 }, { "epoch": 0.6830423202130808, "grad_norm": 5.0, "learning_rate": 1.9917921723795903e-05, "loss": 2.464, "step": 577 }, { "epoch": 0.684226102397159, "grad_norm": 4.40625, "learning_rate": 1.9916860707578642e-05, "loss": 2.5868, "step": 578 }, { "epoch": 0.6854098845812371, "grad_norm": 4.5, "learning_rate": 1.9915792906140138e-05, "loss": 2.9558, "step": 579 }, { "epoch": 0.6865936667653152, "grad_norm": 4.28125, "learning_rate": 1.9914718320210994e-05, "loss": 2.4406, "step": 580 }, { "epoch": 0.6877774489493933, "grad_norm": 4.59375, "learning_rate": 1.9913636950526455e-05, "loss": 2.6716, "step": 581 }, { "epoch": 0.6889612311334714, "grad_norm": 4.5, "learning_rate": 1.9912548797826402e-05, "loss": 2.5916, "step": 582 }, { "epoch": 0.6901450133175495, "grad_norm": 4.59375, "learning_rate": 1.9911453862855362e-05, "loss": 2.8234, "step": 583 }, { "epoch": 0.6913287955016277, "grad_norm": 4.4375, "learning_rate": 1.99103521463625e-05, "loss": 2.5454, "step": 584 }, { "epoch": 0.6925125776857058, "grad_norm": 4.8125, "learning_rate": 1.9909243649101625e-05, "loss": 2.728, "step": 585 }, { "epoch": 0.693696359869784, "grad_norm": 5.03125, "learning_rate": 1.9908128371831177e-05, "loss": 2.7296, "step": 586 }, { "epoch": 0.6948801420538621, "grad_norm": 4.71875, "learning_rate": 1.990700631531425e-05, "loss": 2.8561, "step": 587 }, { "epoch": 0.6960639242379402, "grad_norm": 4.71875, "learning_rate": 1.9905877480318556e-05, "loss": 2.8237, "step": 588 }, { "epoch": 0.6972477064220184, "grad_norm": 4.40625, "learning_rate": 1.9904741867616463e-05, "loss": 2.4267, "step": 589 }, { "epoch": 0.6984314886060965, "grad_norm": 4.46875, "learning_rate": 1.990359947798497e-05, "loss": 2.9332, "step": 590 }, { "epoch": 0.6996152707901746, "grad_norm": 3.890625, "learning_rate": 1.9902450312205707e-05, "loss": 2.8438, "step": 591 }, { "epoch": 0.7007990529742527, "grad_norm": 5.15625, "learning_rate": 1.9901294371064947e-05, "loss": 2.7563, "step": 592 }, { "epoch": 0.7019828351583308, "grad_norm": 4.5625, "learning_rate": 1.9900131655353597e-05, "loss": 2.9209, "step": 593 }, { "epoch": 0.703166617342409, "grad_norm": 5.0, "learning_rate": 1.9898962165867204e-05, "loss": 2.4509, "step": 594 }, { "epoch": 0.7043503995264871, "grad_norm": 4.25, "learning_rate": 1.989778590340594e-05, "loss": 2.6434, "step": 595 }, { "epoch": 0.7055341817105653, "grad_norm": 4.28125, "learning_rate": 1.9896602868774617e-05, "loss": 2.8239, "step": 596 }, { "epoch": 0.7067179638946434, "grad_norm": 4.59375, "learning_rate": 1.9895413062782683e-05, "loss": 3.076, "step": 597 }, { "epoch": 0.7079017460787215, "grad_norm": 4.9375, "learning_rate": 1.9894216486244214e-05, "loss": 2.6014, "step": 598 }, { "epoch": 0.7090855282627997, "grad_norm": 5.25, "learning_rate": 1.9893013139977917e-05, "loss": 2.6695, "step": 599 }, { "epoch": 0.7102693104468778, "grad_norm": 4.0625, "learning_rate": 1.989180302480714e-05, "loss": 2.803, "step": 600 }, { "epoch": 0.7114530926309559, "grad_norm": 4.28125, "learning_rate": 1.989058614155985e-05, "loss": 2.6803, "step": 601 }, { "epoch": 0.712636874815034, "grad_norm": 4.90625, "learning_rate": 1.9889362491068658e-05, "loss": 2.7176, "step": 602 }, { "epoch": 0.7138206569991121, "grad_norm": 5.28125, "learning_rate": 1.9888132074170797e-05, "loss": 3.049, "step": 603 }, { "epoch": 0.7150044391831903, "grad_norm": 4.90625, "learning_rate": 1.988689489170813e-05, "loss": 2.7108, "step": 604 }, { "epoch": 0.7161882213672685, "grad_norm": 4.90625, "learning_rate": 1.9885650944527146e-05, "loss": 2.6473, "step": 605 }, { "epoch": 0.7173720035513466, "grad_norm": 4.625, "learning_rate": 1.9884400233478976e-05, "loss": 2.9112, "step": 606 }, { "epoch": 0.7185557857354247, "grad_norm": 4.34375, "learning_rate": 1.9883142759419364e-05, "loss": 2.9322, "step": 607 }, { "epoch": 0.7197395679195028, "grad_norm": 4.28125, "learning_rate": 1.9881878523208687e-05, "loss": 2.9772, "step": 608 }, { "epoch": 0.720923350103581, "grad_norm": 4.8125, "learning_rate": 1.9880607525711954e-05, "loss": 2.6923, "step": 609 }, { "epoch": 0.7221071322876591, "grad_norm": 4.3125, "learning_rate": 1.987932976779879e-05, "loss": 2.5212, "step": 610 }, { "epoch": 0.7232909144717372, "grad_norm": 4.5625, "learning_rate": 1.9878045250343445e-05, "loss": 2.3864, "step": 611 }, { "epoch": 0.7244746966558153, "grad_norm": 4.96875, "learning_rate": 1.987675397422481e-05, "loss": 2.5733, "step": 612 }, { "epoch": 0.7256584788398934, "grad_norm": 4.78125, "learning_rate": 1.9875455940326386e-05, "loss": 2.7831, "step": 613 }, { "epoch": 0.7268422610239715, "grad_norm": 5.03125, "learning_rate": 1.9874151149536304e-05, "loss": 2.9132, "step": 614 }, { "epoch": 0.7280260432080498, "grad_norm": 4.28125, "learning_rate": 1.9872839602747308e-05, "loss": 2.8321, "step": 615 }, { "epoch": 0.7292098253921279, "grad_norm": 4.90625, "learning_rate": 1.987152130085678e-05, "loss": 3.0614, "step": 616 }, { "epoch": 0.730393607576206, "grad_norm": 4.65625, "learning_rate": 1.9870196244766715e-05, "loss": 2.556, "step": 617 }, { "epoch": 0.7315773897602841, "grad_norm": 4.40625, "learning_rate": 1.9868864435383724e-05, "loss": 2.837, "step": 618 }, { "epoch": 0.7327611719443622, "grad_norm": 5.0, "learning_rate": 1.9867525873619056e-05, "loss": 2.9625, "step": 619 }, { "epoch": 0.7339449541284404, "grad_norm": 4.28125, "learning_rate": 1.986618056038856e-05, "loss": 2.4284, "step": 620 }, { "epoch": 0.7351287363125185, "grad_norm": 4.4375, "learning_rate": 1.9864828496612717e-05, "loss": 2.5562, "step": 621 }, { "epoch": 0.7363125184965966, "grad_norm": 4.125, "learning_rate": 1.9863469683216622e-05, "loss": 2.821, "step": 622 }, { "epoch": 0.7374963006806747, "grad_norm": 4.15625, "learning_rate": 1.9862104121129992e-05, "loss": 2.3722, "step": 623 }, { "epoch": 0.7386800828647528, "grad_norm": 3.953125, "learning_rate": 1.9860731811287157e-05, "loss": 2.6551, "step": 624 }, { "epoch": 0.7398638650488311, "grad_norm": 3.984375, "learning_rate": 1.9859352754627067e-05, "loss": 2.692, "step": 625 }, { "epoch": 0.7410476472329092, "grad_norm": 4.21875, "learning_rate": 1.9857966952093285e-05, "loss": 2.6048, "step": 626 }, { "epoch": 0.7422314294169873, "grad_norm": 4.625, "learning_rate": 1.9856574404633996e-05, "loss": 2.7087, "step": 627 }, { "epoch": 0.7434152116010654, "grad_norm": 4.75, "learning_rate": 1.9855175113201994e-05, "loss": 2.7129, "step": 628 }, { "epoch": 0.7445989937851435, "grad_norm": 4.375, "learning_rate": 1.9853769078754685e-05, "loss": 2.6135, "step": 629 }, { "epoch": 0.7457827759692217, "grad_norm": 4.3125, "learning_rate": 1.9852356302254098e-05, "loss": 2.2924, "step": 630 }, { "epoch": 0.7469665581532998, "grad_norm": 4.25, "learning_rate": 1.9850936784666865e-05, "loss": 2.5269, "step": 631 }, { "epoch": 0.7481503403373779, "grad_norm": 4.09375, "learning_rate": 1.9849510526964243e-05, "loss": 2.5773, "step": 632 }, { "epoch": 0.749334122521456, "grad_norm": 4.4375, "learning_rate": 1.9848077530122083e-05, "loss": 2.526, "step": 633 }, { "epoch": 0.7505179047055341, "grad_norm": 5.15625, "learning_rate": 1.984663779512086e-05, "loss": 2.6791, "step": 634 }, { "epoch": 0.7517016868896124, "grad_norm": 4.25, "learning_rate": 1.984519132294566e-05, "loss": 2.5493, "step": 635 }, { "epoch": 0.7528854690736905, "grad_norm": 4.1875, "learning_rate": 1.984373811458617e-05, "loss": 2.8055, "step": 636 }, { "epoch": 0.7540692512577686, "grad_norm": 4.40625, "learning_rate": 1.9842278171036692e-05, "loss": 2.9258, "step": 637 }, { "epoch": 0.7552530334418467, "grad_norm": 4.53125, "learning_rate": 1.9840811493296134e-05, "loss": 2.5569, "step": 638 }, { "epoch": 0.7564368156259248, "grad_norm": 4.25, "learning_rate": 1.9839338082368017e-05, "loss": 2.3998, "step": 639 }, { "epoch": 0.7576205978100029, "grad_norm": 5.0625, "learning_rate": 1.9837857939260457e-05, "loss": 2.8316, "step": 640 }, { "epoch": 0.7588043799940811, "grad_norm": 4.28125, "learning_rate": 1.983637106498619e-05, "loss": 2.726, "step": 641 }, { "epoch": 0.7599881621781592, "grad_norm": 4.625, "learning_rate": 1.9834877460562546e-05, "loss": 2.9761, "step": 642 }, { "epoch": 0.7611719443622373, "grad_norm": 4.375, "learning_rate": 1.9833377127011468e-05, "loss": 3.0479, "step": 643 }, { "epoch": 0.7623557265463154, "grad_norm": 4.625, "learning_rate": 1.98318700653595e-05, "loss": 2.9874, "step": 644 }, { "epoch": 0.7635395087303936, "grad_norm": 4.5, "learning_rate": 1.9830356276637788e-05, "loss": 2.5282, "step": 645 }, { "epoch": 0.7647232909144718, "grad_norm": 4.125, "learning_rate": 1.9828835761882085e-05, "loss": 2.9227, "step": 646 }, { "epoch": 0.7659070730985499, "grad_norm": 4.75, "learning_rate": 1.982730852213274e-05, "loss": 2.9667, "step": 647 }, { "epoch": 0.767090855282628, "grad_norm": 4.46875, "learning_rate": 1.9825774558434712e-05, "loss": 2.5896, "step": 648 }, { "epoch": 0.7682746374667061, "grad_norm": 3.859375, "learning_rate": 1.9824233871837547e-05, "loss": 2.521, "step": 649 }, { "epoch": 0.7694584196507842, "grad_norm": 4.3125, "learning_rate": 1.9822686463395406e-05, "loss": 2.3522, "step": 650 }, { "epoch": 0.7706422018348624, "grad_norm": 4.3125, "learning_rate": 1.982113233416704e-05, "loss": 2.6371, "step": 651 }, { "epoch": 0.7718259840189405, "grad_norm": 4.4375, "learning_rate": 1.98195714852158e-05, "loss": 2.8795, "step": 652 }, { "epoch": 0.7730097662030186, "grad_norm": 4.75, "learning_rate": 1.9818003917609637e-05, "loss": 2.2281, "step": 653 }, { "epoch": 0.7741935483870968, "grad_norm": 4.15625, "learning_rate": 1.9816429632421095e-05, "loss": 2.5841, "step": 654 }, { "epoch": 0.7753773305711749, "grad_norm": 4.21875, "learning_rate": 1.9814848630727323e-05, "loss": 3.0166, "step": 655 }, { "epoch": 0.7765611127552531, "grad_norm": 4.90625, "learning_rate": 1.9813260913610048e-05, "loss": 2.5244, "step": 656 }, { "epoch": 0.7777448949393312, "grad_norm": 3.84375, "learning_rate": 1.9811666482155608e-05, "loss": 2.5887, "step": 657 }, { "epoch": 0.7789286771234093, "grad_norm": 5.0, "learning_rate": 1.9810065337454934e-05, "loss": 3.0088, "step": 658 }, { "epoch": 0.7801124593074874, "grad_norm": 4.65625, "learning_rate": 1.9808457480603547e-05, "loss": 2.8829, "step": 659 }, { "epoch": 0.7812962414915655, "grad_norm": 4.5625, "learning_rate": 1.980684291270155e-05, "loss": 3.0234, "step": 660 }, { "epoch": 0.7824800236756437, "grad_norm": 4.84375, "learning_rate": 1.9805221634853656e-05, "loss": 2.9785, "step": 661 }, { "epoch": 0.7836638058597218, "grad_norm": 4.125, "learning_rate": 1.980359364816916e-05, "loss": 2.8455, "step": 662 }, { "epoch": 0.7848475880438, "grad_norm": 4.5625, "learning_rate": 1.980195895376194e-05, "loss": 2.9199, "step": 663 }, { "epoch": 0.7860313702278781, "grad_norm": 4.28125, "learning_rate": 1.980031755275048e-05, "loss": 2.6206, "step": 664 }, { "epoch": 0.7872151524119562, "grad_norm": 4.15625, "learning_rate": 1.9798669446257844e-05, "loss": 2.4331, "step": 665 }, { "epoch": 0.7883989345960344, "grad_norm": 4.28125, "learning_rate": 1.9797014635411674e-05, "loss": 2.4075, "step": 666 }, { "epoch": 0.7895827167801125, "grad_norm": 3.8125, "learning_rate": 1.979535312134422e-05, "loss": 2.5911, "step": 667 }, { "epoch": 0.7907664989641906, "grad_norm": 4.34375, "learning_rate": 1.9793684905192302e-05, "loss": 2.7445, "step": 668 }, { "epoch": 0.7919502811482687, "grad_norm": 4.0625, "learning_rate": 1.9792009988097334e-05, "loss": 2.4411, "step": 669 }, { "epoch": 0.7931340633323468, "grad_norm": 4.75, "learning_rate": 1.979032837120531e-05, "loss": 2.7688, "step": 670 }, { "epoch": 0.7943178455164249, "grad_norm": 4.21875, "learning_rate": 1.978864005566681e-05, "loss": 2.1891, "step": 671 }, { "epoch": 0.7955016277005031, "grad_norm": 4.4375, "learning_rate": 1.9786945042636998e-05, "loss": 2.8605, "step": 672 }, { "epoch": 0.7966854098845813, "grad_norm": 4.5, "learning_rate": 1.9785243333275622e-05, "loss": 2.7968, "step": 673 }, { "epoch": 0.7978691920686594, "grad_norm": 4.46875, "learning_rate": 1.9783534928747006e-05, "loss": 2.4442, "step": 674 }, { "epoch": 0.7990529742527375, "grad_norm": 4.6875, "learning_rate": 1.9781819830220058e-05, "loss": 2.6155, "step": 675 }, { "epoch": 0.8002367564368156, "grad_norm": 4.75, "learning_rate": 1.978009803886827e-05, "loss": 2.4824, "step": 676 }, { "epoch": 0.8014205386208938, "grad_norm": 4.53125, "learning_rate": 1.9778369555869714e-05, "loss": 2.6693, "step": 677 }, { "epoch": 0.8026043208049719, "grad_norm": 4.5625, "learning_rate": 1.9776634382407027e-05, "loss": 2.9669, "step": 678 }, { "epoch": 0.80378810298905, "grad_norm": 4.3125, "learning_rate": 1.9774892519667433e-05, "loss": 2.8019, "step": 679 }, { "epoch": 0.8049718851731281, "grad_norm": 4.21875, "learning_rate": 1.9773143968842743e-05, "loss": 2.7486, "step": 680 }, { "epoch": 0.8061556673572062, "grad_norm": 4.46875, "learning_rate": 1.9771388731129325e-05, "loss": 2.7975, "step": 681 }, { "epoch": 0.8073394495412844, "grad_norm": 4.78125, "learning_rate": 1.9769626807728134e-05, "loss": 2.663, "step": 682 }, { "epoch": 0.8085232317253626, "grad_norm": 4.375, "learning_rate": 1.9767858199844697e-05, "loss": 2.777, "step": 683 }, { "epoch": 0.8097070139094407, "grad_norm": 4.84375, "learning_rate": 1.9766082908689117e-05, "loss": 2.8866, "step": 684 }, { "epoch": 0.8108907960935188, "grad_norm": 4.5, "learning_rate": 1.9764300935476065e-05, "loss": 2.8192, "step": 685 }, { "epoch": 0.8120745782775969, "grad_norm": 4.90625, "learning_rate": 1.9762512281424778e-05, "loss": 2.9202, "step": 686 }, { "epoch": 0.8132583604616751, "grad_norm": 4.0, "learning_rate": 1.9760716947759086e-05, "loss": 2.3563, "step": 687 }, { "epoch": 0.8144421426457532, "grad_norm": 5.09375, "learning_rate": 1.9758914935707366e-05, "loss": 2.4792, "step": 688 }, { "epoch": 0.8156259248298313, "grad_norm": 4.59375, "learning_rate": 1.9757106246502577e-05, "loss": 2.4805, "step": 689 }, { "epoch": 0.8168097070139094, "grad_norm": 4.71875, "learning_rate": 1.9755290881382244e-05, "loss": 2.6702, "step": 690 }, { "epoch": 0.8179934891979875, "grad_norm": 4.0, "learning_rate": 1.9753468841588456e-05, "loss": 2.5932, "step": 691 }, { "epoch": 0.8191772713820658, "grad_norm": 4.90625, "learning_rate": 1.9751640128367872e-05, "loss": 2.5145, "step": 692 }, { "epoch": 0.8203610535661439, "grad_norm": 4.125, "learning_rate": 1.974980474297172e-05, "loss": 2.6263, "step": 693 }, { "epoch": 0.821544835750222, "grad_norm": 4.84375, "learning_rate": 1.974796268665579e-05, "loss": 2.8267, "step": 694 }, { "epoch": 0.8227286179343001, "grad_norm": 4.15625, "learning_rate": 1.9746113960680435e-05, "loss": 2.4361, "step": 695 }, { "epoch": 0.8239124001183782, "grad_norm": 4.5, "learning_rate": 1.9744258566310574e-05, "loss": 2.4392, "step": 696 }, { "epoch": 0.8250961823024563, "grad_norm": 4.15625, "learning_rate": 1.9742396504815687e-05, "loss": 2.6312, "step": 697 }, { "epoch": 0.8262799644865345, "grad_norm": 4.375, "learning_rate": 1.974052777746981e-05, "loss": 2.8784, "step": 698 }, { "epoch": 0.8274637466706126, "grad_norm": 4.53125, "learning_rate": 1.973865238555156e-05, "loss": 2.8234, "step": 699 }, { "epoch": 0.8286475288546907, "grad_norm": 4.4375, "learning_rate": 1.9736770330344084e-05, "loss": 2.8006, "step": 700 }, { "epoch": 0.8298313110387688, "grad_norm": 4.3125, "learning_rate": 1.973488161313512e-05, "loss": 2.448, "step": 701 }, { "epoch": 0.831015093222847, "grad_norm": 4.5625, "learning_rate": 1.9732986235216937e-05, "loss": 2.846, "step": 702 }, { "epoch": 0.8321988754069252, "grad_norm": 5.34375, "learning_rate": 1.9731084197886376e-05, "loss": 2.8718, "step": 703 }, { "epoch": 0.8333826575910033, "grad_norm": 4.59375, "learning_rate": 1.9729175502444832e-05, "loss": 2.6046, "step": 704 }, { "epoch": 0.8345664397750814, "grad_norm": 4.375, "learning_rate": 1.9727260150198254e-05, "loss": 2.5858, "step": 705 }, { "epoch": 0.8357502219591595, "grad_norm": 4.1875, "learning_rate": 1.9725338142457145e-05, "loss": 2.6371, "step": 706 }, { "epoch": 0.8369340041432376, "grad_norm": 4.53125, "learning_rate": 1.9723409480536566e-05, "loss": 2.6958, "step": 707 }, { "epoch": 0.8381177863273158, "grad_norm": 4.9375, "learning_rate": 1.9721474165756124e-05, "loss": 2.6909, "step": 708 }, { "epoch": 0.8393015685113939, "grad_norm": 4.21875, "learning_rate": 1.9719532199439986e-05, "loss": 2.5668, "step": 709 }, { "epoch": 0.840485350695472, "grad_norm": 4.84375, "learning_rate": 1.9717583582916862e-05, "loss": 2.7497, "step": 710 }, { "epoch": 0.8416691328795501, "grad_norm": 4.0625, "learning_rate": 1.9715628317520022e-05, "loss": 2.8988, "step": 711 }, { "epoch": 0.8428529150636282, "grad_norm": 4.6875, "learning_rate": 1.9713666404587274e-05, "loss": 2.5417, "step": 712 }, { "epoch": 0.8440366972477065, "grad_norm": 4.53125, "learning_rate": 1.9711697845460978e-05, "loss": 2.8351, "step": 713 }, { "epoch": 0.8452204794317846, "grad_norm": 5.15625, "learning_rate": 1.970972264148805e-05, "loss": 2.6508, "step": 714 }, { "epoch": 0.8464042616158627, "grad_norm": 4.90625, "learning_rate": 1.9707740794019937e-05, "loss": 2.8921, "step": 715 }, { "epoch": 0.8475880437999408, "grad_norm": 4.75, "learning_rate": 1.970575230441265e-05, "loss": 2.6767, "step": 716 }, { "epoch": 0.8487718259840189, "grad_norm": 4.4375, "learning_rate": 1.9703757174026726e-05, "loss": 2.4478, "step": 717 }, { "epoch": 0.8499556081680971, "grad_norm": 4.59375, "learning_rate": 1.970175540422726e-05, "loss": 2.8067, "step": 718 }, { "epoch": 0.8511393903521752, "grad_norm": 4.71875, "learning_rate": 1.969974699638388e-05, "loss": 2.9803, "step": 719 }, { "epoch": 0.8523231725362533, "grad_norm": 4.5625, "learning_rate": 1.969773195187076e-05, "loss": 2.5863, "step": 720 }, { "epoch": 0.8535069547203314, "grad_norm": 4.53125, "learning_rate": 1.9695710272066623e-05, "loss": 2.5013, "step": 721 }, { "epoch": 0.8546907369044096, "grad_norm": 4.6875, "learning_rate": 1.969368195835471e-05, "loss": 2.5718, "step": 722 }, { "epoch": 0.8558745190884878, "grad_norm": 3.9375, "learning_rate": 1.9691647012122825e-05, "loss": 2.5935, "step": 723 }, { "epoch": 0.8570583012725659, "grad_norm": 3.9375, "learning_rate": 1.9689605434763298e-05, "loss": 2.7084, "step": 724 }, { "epoch": 0.858242083456644, "grad_norm": 4.40625, "learning_rate": 1.9687557227672997e-05, "loss": 2.8408, "step": 725 }, { "epoch": 0.8594258656407221, "grad_norm": 5.9375, "learning_rate": 1.9685502392253326e-05, "loss": 2.5317, "step": 726 }, { "epoch": 0.8606096478248002, "grad_norm": 4.3125, "learning_rate": 1.9683440929910223e-05, "loss": 2.3043, "step": 727 }, { "epoch": 0.8617934300088783, "grad_norm": 4.78125, "learning_rate": 1.968137284205417e-05, "loss": 2.5764, "step": 728 }, { "epoch": 0.8629772121929565, "grad_norm": 4.1875, "learning_rate": 1.9679298130100165e-05, "loss": 2.8978, "step": 729 }, { "epoch": 0.8641609943770346, "grad_norm": 5.0, "learning_rate": 1.967721679546776e-05, "loss": 2.3907, "step": 730 }, { "epoch": 0.8653447765611127, "grad_norm": 4.90625, "learning_rate": 1.9675128839581013e-05, "loss": 2.9282, "step": 731 }, { "epoch": 0.8665285587451909, "grad_norm": 4.28125, "learning_rate": 1.9673034263868533e-05, "loss": 2.5821, "step": 732 }, { "epoch": 0.867712340929269, "grad_norm": 4.5625, "learning_rate": 1.967093306976345e-05, "loss": 2.622, "step": 733 }, { "epoch": 0.8688961231133472, "grad_norm": 4.5625, "learning_rate": 1.9668825258703424e-05, "loss": 2.9779, "step": 734 }, { "epoch": 0.8700799052974253, "grad_norm": 4.9375, "learning_rate": 1.966671083213064e-05, "loss": 2.8442, "step": 735 }, { "epoch": 0.8712636874815034, "grad_norm": 4.96875, "learning_rate": 1.9664589791491812e-05, "loss": 2.9357, "step": 736 }, { "epoch": 0.8724474696655815, "grad_norm": 4.5625, "learning_rate": 1.966246213823818e-05, "loss": 2.4707, "step": 737 }, { "epoch": 0.8736312518496596, "grad_norm": 4.59375, "learning_rate": 1.9660327873825503e-05, "loss": 2.538, "step": 738 }, { "epoch": 0.8748150340337378, "grad_norm": 4.65625, "learning_rate": 1.9658186999714067e-05, "loss": 2.8768, "step": 739 }, { "epoch": 0.875998816217816, "grad_norm": 4.25, "learning_rate": 1.9656039517368686e-05, "loss": 2.6136, "step": 740 }, { "epoch": 0.877182598401894, "grad_norm": 5.0625, "learning_rate": 1.9653885428258686e-05, "loss": 2.737, "step": 741 }, { "epoch": 0.8783663805859722, "grad_norm": 3.90625, "learning_rate": 1.965172473385792e-05, "loss": 2.2199, "step": 742 }, { "epoch": 0.8795501627700503, "grad_norm": 4.5, "learning_rate": 1.9649557435644757e-05, "loss": 2.7758, "step": 743 }, { "epoch": 0.8807339449541285, "grad_norm": 4.15625, "learning_rate": 1.9647383535102085e-05, "loss": 2.8697, "step": 744 }, { "epoch": 0.8819177271382066, "grad_norm": 4.4375, "learning_rate": 1.964520303371731e-05, "loss": 2.8299, "step": 745 }, { "epoch": 0.8831015093222847, "grad_norm": 4.4375, "learning_rate": 1.9643015932982355e-05, "loss": 2.6092, "step": 746 }, { "epoch": 0.8842852915063628, "grad_norm": 4.25, "learning_rate": 1.9640822234393654e-05, "loss": 2.5451, "step": 747 }, { "epoch": 0.8854690736904409, "grad_norm": 4.40625, "learning_rate": 1.9638621939452166e-05, "loss": 2.6748, "step": 748 }, { "epoch": 0.8866528558745191, "grad_norm": 4.65625, "learning_rate": 1.963641504966335e-05, "loss": 2.7055, "step": 749 }, { "epoch": 0.8878366380585972, "grad_norm": 4.21875, "learning_rate": 1.9634201566537182e-05, "loss": 2.7432, "step": 750 }, { "epoch": 0.8890204202426754, "grad_norm": 4.78125, "learning_rate": 1.963198149158816e-05, "loss": 3.2116, "step": 751 }, { "epoch": 0.8902042024267535, "grad_norm": 4.625, "learning_rate": 1.962975482633527e-05, "loss": 2.6164, "step": 752 }, { "epoch": 0.8913879846108316, "grad_norm": 4.53125, "learning_rate": 1.9627521572302035e-05, "loss": 2.7443, "step": 753 }, { "epoch": 0.8925717667949097, "grad_norm": 4.59375, "learning_rate": 1.9625281731016455e-05, "loss": 2.6206, "step": 754 }, { "epoch": 0.8937555489789879, "grad_norm": 4.09375, "learning_rate": 1.9623035304011062e-05, "loss": 2.5069, "step": 755 }, { "epoch": 0.894939331163066, "grad_norm": 4.25, "learning_rate": 1.962078229282289e-05, "loss": 2.3173, "step": 756 }, { "epoch": 0.8961231133471441, "grad_norm": 4.21875, "learning_rate": 1.961852269899346e-05, "loss": 2.4459, "step": 757 }, { "epoch": 0.8973068955312222, "grad_norm": 4.65625, "learning_rate": 1.9616256524068822e-05, "loss": 2.3314, "step": 758 }, { "epoch": 0.8984906777153003, "grad_norm": 4.59375, "learning_rate": 1.961398376959951e-05, "loss": 2.6148, "step": 759 }, { "epoch": 0.8996744598993786, "grad_norm": 4.15625, "learning_rate": 1.961170443714057e-05, "loss": 2.5566, "step": 760 }, { "epoch": 0.9008582420834567, "grad_norm": 4.40625, "learning_rate": 1.9609418528251543e-05, "loss": 2.7433, "step": 761 }, { "epoch": 0.9020420242675348, "grad_norm": 4.34375, "learning_rate": 1.9607126044496476e-05, "loss": 2.8541, "step": 762 }, { "epoch": 0.9032258064516129, "grad_norm": 4.34375, "learning_rate": 1.9604826987443905e-05, "loss": 2.5727, "step": 763 }, { "epoch": 0.904409588635691, "grad_norm": 4.46875, "learning_rate": 1.960252135866687e-05, "loss": 2.5726, "step": 764 }, { "epoch": 0.9055933708197692, "grad_norm": 4.625, "learning_rate": 1.9600209159742913e-05, "loss": 2.8777, "step": 765 }, { "epoch": 0.9067771530038473, "grad_norm": 4.875, "learning_rate": 1.959789039225406e-05, "loss": 2.8927, "step": 766 }, { "epoch": 0.9079609351879254, "grad_norm": 4.6875, "learning_rate": 1.9595565057786838e-05, "loss": 2.7569, "step": 767 }, { "epoch": 0.9091447173720035, "grad_norm": 4.59375, "learning_rate": 1.9593233157932265e-05, "loss": 2.5841, "step": 768 }, { "epoch": 0.9103284995560816, "grad_norm": 4.71875, "learning_rate": 1.9590894694285852e-05, "loss": 2.488, "step": 769 }, { "epoch": 0.9115122817401599, "grad_norm": 4.25, "learning_rate": 1.9588549668447595e-05, "loss": 2.9169, "step": 770 }, { "epoch": 0.912696063924238, "grad_norm": 4.625, "learning_rate": 1.9586198082021995e-05, "loss": 2.6622, "step": 771 }, { "epoch": 0.9138798461083161, "grad_norm": 4.5, "learning_rate": 1.958383993661803e-05, "loss": 2.8777, "step": 772 }, { "epoch": 0.9150636282923942, "grad_norm": 5.0625, "learning_rate": 1.9581475233849165e-05, "loss": 2.9759, "step": 773 }, { "epoch": 0.9162474104764723, "grad_norm": 4.40625, "learning_rate": 1.957910397533335e-05, "loss": 2.4661, "step": 774 }, { "epoch": 0.9174311926605505, "grad_norm": 4.53125, "learning_rate": 1.9576726162693036e-05, "loss": 2.7026, "step": 775 }, { "epoch": 0.9186149748446286, "grad_norm": 3.8125, "learning_rate": 1.9574341797555143e-05, "loss": 2.42, "step": 776 }, { "epoch": 0.9197987570287067, "grad_norm": 4.09375, "learning_rate": 1.9571950881551073e-05, "loss": 2.7335, "step": 777 }, { "epoch": 0.9209825392127848, "grad_norm": 4.5625, "learning_rate": 1.9569553416316725e-05, "loss": 2.5518, "step": 778 }, { "epoch": 0.9221663213968629, "grad_norm": 4.5, "learning_rate": 1.9567149403492467e-05, "loss": 2.5461, "step": 779 }, { "epoch": 0.9233501035809412, "grad_norm": 4.09375, "learning_rate": 1.9564738844723153e-05, "loss": 2.5361, "step": 780 }, { "epoch": 0.9245338857650193, "grad_norm": 4.625, "learning_rate": 1.9562321741658106e-05, "loss": 3.1115, "step": 781 }, { "epoch": 0.9257176679490974, "grad_norm": 4.78125, "learning_rate": 1.9559898095951137e-05, "loss": 2.7055, "step": 782 }, { "epoch": 0.9269014501331755, "grad_norm": 4.65625, "learning_rate": 1.955746790926053e-05, "loss": 2.9354, "step": 783 }, { "epoch": 0.9280852323172536, "grad_norm": 4.40625, "learning_rate": 1.9555031183249046e-05, "loss": 2.3607, "step": 784 }, { "epoch": 0.9292690145013317, "grad_norm": 4.21875, "learning_rate": 1.9552587919583915e-05, "loss": 2.5434, "step": 785 }, { "epoch": 0.9304527966854099, "grad_norm": 5.0, "learning_rate": 1.955013811993685e-05, "loss": 3.1603, "step": 786 }, { "epoch": 0.931636578869488, "grad_norm": 4.8125, "learning_rate": 1.9547681785984024e-05, "loss": 2.6145, "step": 787 }, { "epoch": 0.9328203610535661, "grad_norm": 4.125, "learning_rate": 1.954521891940609e-05, "loss": 2.8456, "step": 788 }, { "epoch": 0.9340041432376442, "grad_norm": 4.09375, "learning_rate": 1.9542749521888165e-05, "loss": 2.6692, "step": 789 }, { "epoch": 0.9351879254217224, "grad_norm": 4.65625, "learning_rate": 1.9540273595119842e-05, "loss": 2.7225, "step": 790 }, { "epoch": 0.9363717076058006, "grad_norm": 3.984375, "learning_rate": 1.953779114079517e-05, "loss": 2.0607, "step": 791 }, { "epoch": 0.9375554897898787, "grad_norm": 4.6875, "learning_rate": 1.9535302160612674e-05, "loss": 2.6539, "step": 792 }, { "epoch": 0.9387392719739568, "grad_norm": 4.5625, "learning_rate": 1.953280665627534e-05, "loss": 2.8294, "step": 793 }, { "epoch": 0.9399230541580349, "grad_norm": 4.5625, "learning_rate": 1.953030462949062e-05, "loss": 2.7902, "step": 794 }, { "epoch": 0.941106836342113, "grad_norm": 4.71875, "learning_rate": 1.9527796081970423e-05, "loss": 2.7581, "step": 795 }, { "epoch": 0.9422906185261912, "grad_norm": 4.0625, "learning_rate": 1.9525281015431127e-05, "loss": 2.7359, "step": 796 }, { "epoch": 0.9434744007102693, "grad_norm": 3.65625, "learning_rate": 1.9522759431593566e-05, "loss": 2.4659, "step": 797 }, { "epoch": 0.9446581828943474, "grad_norm": 4.25, "learning_rate": 1.9520231332183037e-05, "loss": 2.6183, "step": 798 }, { "epoch": 0.9458419650784256, "grad_norm": 4.0625, "learning_rate": 1.951769671892929e-05, "loss": 2.2723, "step": 799 }, { "epoch": 0.9470257472625037, "grad_norm": 4.34375, "learning_rate": 1.9515155593566536e-05, "loss": 2.9629, "step": 800 }, { "epoch": 0.9482095294465819, "grad_norm": 5.28125, "learning_rate": 1.9512607957833438e-05, "loss": 2.9293, "step": 801 }, { "epoch": 0.94939331163066, "grad_norm": 4.46875, "learning_rate": 1.9510053813473115e-05, "loss": 3.1096, "step": 802 }, { "epoch": 0.9505770938147381, "grad_norm": 4.3125, "learning_rate": 1.950749316223314e-05, "loss": 2.597, "step": 803 }, { "epoch": 0.9517608759988162, "grad_norm": 4.375, "learning_rate": 1.9504926005865542e-05, "loss": 2.5539, "step": 804 }, { "epoch": 0.9529446581828943, "grad_norm": 4.96875, "learning_rate": 1.950235234612679e-05, "loss": 2.5333, "step": 805 }, { "epoch": 0.9541284403669725, "grad_norm": 4.75, "learning_rate": 1.9499772184777814e-05, "loss": 3.3359, "step": 806 }, { "epoch": 0.9553122225510506, "grad_norm": 4.28125, "learning_rate": 1.9497185523583986e-05, "loss": 2.7828, "step": 807 }, { "epoch": 0.9564960047351287, "grad_norm": 4.21875, "learning_rate": 1.9494592364315125e-05, "loss": 2.5705, "step": 808 }, { "epoch": 0.9576797869192069, "grad_norm": 3.84375, "learning_rate": 1.9491992708745502e-05, "loss": 2.6607, "step": 809 }, { "epoch": 0.958863569103285, "grad_norm": 4.15625, "learning_rate": 1.948938655865383e-05, "loss": 2.4996, "step": 810 }, { "epoch": 0.9600473512873631, "grad_norm": 4.84375, "learning_rate": 1.9486773915823258e-05, "loss": 2.4498, "step": 811 }, { "epoch": 0.9612311334714413, "grad_norm": 4.1875, "learning_rate": 1.948415478204139e-05, "loss": 2.4218, "step": 812 }, { "epoch": 0.9624149156555194, "grad_norm": 4.84375, "learning_rate": 1.9481529159100258e-05, "loss": 2.5013, "step": 813 }, { "epoch": 0.9635986978395975, "grad_norm": 4.90625, "learning_rate": 1.947889704879635e-05, "loss": 2.9975, "step": 814 }, { "epoch": 0.9647824800236756, "grad_norm": 4.75, "learning_rate": 1.9476258452930577e-05, "loss": 2.5171, "step": 815 }, { "epoch": 0.9659662622077537, "grad_norm": 4.40625, "learning_rate": 1.94736133733083e-05, "loss": 2.9159, "step": 816 }, { "epoch": 0.9671500443918319, "grad_norm": 4.21875, "learning_rate": 1.9470961811739304e-05, "loss": 2.6609, "step": 817 }, { "epoch": 0.96833382657591, "grad_norm": 4.78125, "learning_rate": 1.946830377003782e-05, "loss": 3.1055, "step": 818 }, { "epoch": 0.9695176087599882, "grad_norm": 3.984375, "learning_rate": 1.946563925002251e-05, "loss": 2.3084, "step": 819 }, { "epoch": 0.9707013909440663, "grad_norm": 4.09375, "learning_rate": 1.9462968253516462e-05, "loss": 2.5402, "step": 820 }, { "epoch": 0.9718851731281444, "grad_norm": 4.25, "learning_rate": 1.94602907823472e-05, "loss": 2.6274, "step": 821 }, { "epoch": 0.9730689553122226, "grad_norm": 4.8125, "learning_rate": 1.945760683834668e-05, "loss": 2.6139, "step": 822 }, { "epoch": 0.9742527374963007, "grad_norm": 4.5, "learning_rate": 1.9454916423351288e-05, "loss": 2.5494, "step": 823 }, { "epoch": 0.9754365196803788, "grad_norm": 4.75, "learning_rate": 1.945221953920183e-05, "loss": 2.7888, "step": 824 }, { "epoch": 0.9766203018644569, "grad_norm": 4.40625, "learning_rate": 1.9449516187743546e-05, "loss": 2.8302, "step": 825 }, { "epoch": 0.977804084048535, "grad_norm": 4.34375, "learning_rate": 1.94468063708261e-05, "loss": 2.8079, "step": 826 }, { "epoch": 0.9789878662326132, "grad_norm": 4.28125, "learning_rate": 1.9444090090303567e-05, "loss": 2.8151, "step": 827 }, { "epoch": 0.9801716484166914, "grad_norm": 4.125, "learning_rate": 1.9441367348034464e-05, "loss": 2.4723, "step": 828 }, { "epoch": 0.9813554306007695, "grad_norm": 4.9375, "learning_rate": 1.9438638145881718e-05, "loss": 3.2675, "step": 829 }, { "epoch": 0.9825392127848476, "grad_norm": 4.6875, "learning_rate": 1.9435902485712677e-05, "loss": 2.8028, "step": 830 }, { "epoch": 0.9837229949689257, "grad_norm": 4.40625, "learning_rate": 1.9433160369399108e-05, "loss": 2.7554, "step": 831 }, { "epoch": 0.9849067771530039, "grad_norm": 4.5625, "learning_rate": 1.94304117988172e-05, "loss": 2.524, "step": 832 }, { "epoch": 0.986090559337082, "grad_norm": 4.34375, "learning_rate": 1.9427656775847553e-05, "loss": 2.3596, "step": 833 }, { "epoch": 0.9872743415211601, "grad_norm": 4.5, "learning_rate": 1.9424895302375177e-05, "loss": 2.6464, "step": 834 }, { "epoch": 0.9884581237052382, "grad_norm": 4.75, "learning_rate": 1.9422127380289508e-05, "loss": 2.4242, "step": 835 }, { "epoch": 0.9896419058893163, "grad_norm": 4.4375, "learning_rate": 1.941935301148439e-05, "loss": 2.4942, "step": 836 }, { "epoch": 0.9908256880733946, "grad_norm": 4.46875, "learning_rate": 1.9416572197858063e-05, "loss": 2.8784, "step": 837 }, { "epoch": 0.9920094702574727, "grad_norm": 4.96875, "learning_rate": 1.9413784941313203e-05, "loss": 2.9636, "step": 838 }, { "epoch": 0.9931932524415508, "grad_norm": 4.09375, "learning_rate": 1.941099124375687e-05, "loss": 2.4056, "step": 839 }, { "epoch": 0.9943770346256289, "grad_norm": 5.0625, "learning_rate": 1.9408191107100554e-05, "loss": 2.5094, "step": 840 }, { "epoch": 0.995560816809707, "grad_norm": 5.0, "learning_rate": 1.9405384533260124e-05, "loss": 2.7804, "step": 841 }, { "epoch": 0.9967445989937851, "grad_norm": 4.34375, "learning_rate": 1.9402571524155877e-05, "loss": 2.6124, "step": 842 }, { "epoch": 0.9979283811778633, "grad_norm": 4.6875, "learning_rate": 1.9399752081712498e-05, "loss": 2.7607, "step": 843 }, { "epoch": 0.9991121633619414, "grad_norm": 4.0, "learning_rate": 1.9396926207859085e-05, "loss": 2.5465, "step": 844 }, { "epoch": 1.0, "grad_norm": 3.609375, "learning_rate": 1.939409390452913e-05, "loss": 1.9983, "step": 845 }, { "epoch": 1.001183782184078, "grad_norm": 4.0625, "learning_rate": 1.9391255173660516e-05, "loss": 2.1965, "step": 846 }, { "epoch": 1.0023675643681562, "grad_norm": 4.15625, "learning_rate": 1.9388410017195544e-05, "loss": 2.7391, "step": 847 }, { "epoch": 1.0035513465522343, "grad_norm": 3.8125, "learning_rate": 1.9385558437080898e-05, "loss": 2.1862, "step": 848 }, { "epoch": 1.0047351287363124, "grad_norm": 4.625, "learning_rate": 1.9382700435267653e-05, "loss": 2.847, "step": 849 }, { "epoch": 1.0059189109203905, "grad_norm": 4.21875, "learning_rate": 1.937983601371129e-05, "loss": 2.5664, "step": 850 }, { "epoch": 1.0071026931044689, "grad_norm": 4.15625, "learning_rate": 1.9376965174371678e-05, "loss": 2.3289, "step": 851 }, { "epoch": 1.008286475288547, "grad_norm": 4.375, "learning_rate": 1.937408791921307e-05, "loss": 2.4666, "step": 852 }, { "epoch": 1.009470257472625, "grad_norm": 3.9375, "learning_rate": 1.9371204250204116e-05, "loss": 2.2233, "step": 853 }, { "epoch": 1.0106540396567032, "grad_norm": 4.34375, "learning_rate": 1.9368314169317858e-05, "loss": 2.7672, "step": 854 }, { "epoch": 1.0118378218407813, "grad_norm": 4.28125, "learning_rate": 1.9365417678531713e-05, "loss": 2.8016, "step": 855 }, { "epoch": 1.0130216040248594, "grad_norm": 3.875, "learning_rate": 1.9362514779827494e-05, "loss": 2.571, "step": 856 }, { "epoch": 1.0142053862089375, "grad_norm": 4.5625, "learning_rate": 1.9359605475191396e-05, "loss": 2.462, "step": 857 }, { "epoch": 1.0153891683930156, "grad_norm": 4.21875, "learning_rate": 1.9356689766613993e-05, "loss": 2.5557, "step": 858 }, { "epoch": 1.0165729505770937, "grad_norm": 3.890625, "learning_rate": 1.935376765609025e-05, "loss": 2.6646, "step": 859 }, { "epoch": 1.0177567327611718, "grad_norm": 4.71875, "learning_rate": 1.9350839145619496e-05, "loss": 2.9739, "step": 860 }, { "epoch": 1.0189405149452502, "grad_norm": 4.46875, "learning_rate": 1.934790423720546e-05, "loss": 2.5945, "step": 861 }, { "epoch": 1.0201242971293283, "grad_norm": 4.65625, "learning_rate": 1.9344962932856228e-05, "loss": 2.9959, "step": 862 }, { "epoch": 1.0213080793134064, "grad_norm": 3.984375, "learning_rate": 1.9342015234584277e-05, "loss": 2.5056, "step": 863 }, { "epoch": 1.0224918614974845, "grad_norm": 3.65625, "learning_rate": 1.9339061144406455e-05, "loss": 2.1112, "step": 864 }, { "epoch": 1.0236756436815626, "grad_norm": 3.78125, "learning_rate": 1.9336100664343976e-05, "loss": 2.3835, "step": 865 }, { "epoch": 1.0248594258656407, "grad_norm": 4.4375, "learning_rate": 1.9333133796422437e-05, "loss": 2.783, "step": 866 }, { "epoch": 1.0260432080497188, "grad_norm": 4.21875, "learning_rate": 1.9330160542671794e-05, "loss": 2.712, "step": 867 }, { "epoch": 1.027226990233797, "grad_norm": 4.09375, "learning_rate": 1.9327180905126386e-05, "loss": 2.4345, "step": 868 }, { "epoch": 1.028410772417875, "grad_norm": 4.4375, "learning_rate": 1.932419488582491e-05, "loss": 2.4697, "step": 869 }, { "epoch": 1.0295945546019531, "grad_norm": 4.5625, "learning_rate": 1.9321202486810427e-05, "loss": 3.0157, "step": 870 }, { "epoch": 1.0307783367860315, "grad_norm": 4.0625, "learning_rate": 1.931820371013038e-05, "loss": 2.3198, "step": 871 }, { "epoch": 1.0319621189701096, "grad_norm": 4.1875, "learning_rate": 1.9315198557836555e-05, "loss": 2.527, "step": 872 }, { "epoch": 1.0331459011541877, "grad_norm": 4.71875, "learning_rate": 1.9312187031985105e-05, "loss": 2.7784, "step": 873 }, { "epoch": 1.0343296833382658, "grad_norm": 3.796875, "learning_rate": 1.9309169134636558e-05, "loss": 2.4139, "step": 874 }, { "epoch": 1.035513465522344, "grad_norm": 4.25, "learning_rate": 1.930614486785579e-05, "loss": 2.7683, "step": 875 }, { "epoch": 1.036697247706422, "grad_norm": 4.28125, "learning_rate": 1.930311423371203e-05, "loss": 2.7165, "step": 876 }, { "epoch": 1.0378810298905001, "grad_norm": 4.75, "learning_rate": 1.930007723427887e-05, "loss": 2.5122, "step": 877 }, { "epoch": 1.0390648120745782, "grad_norm": 4.28125, "learning_rate": 1.9297033871634266e-05, "loss": 2.7893, "step": 878 }, { "epoch": 1.0402485942586563, "grad_norm": 4.0625, "learning_rate": 1.929398414786051e-05, "loss": 2.7538, "step": 879 }, { "epoch": 1.0414323764427345, "grad_norm": 4.78125, "learning_rate": 1.929092806504426e-05, "loss": 2.3285, "step": 880 }, { "epoch": 1.0426161586268126, "grad_norm": 4.09375, "learning_rate": 1.928786562527652e-05, "loss": 2.6879, "step": 881 }, { "epoch": 1.043799940810891, "grad_norm": 4.9375, "learning_rate": 1.9284796830652642e-05, "loss": 2.6848, "step": 882 }, { "epoch": 1.044983722994969, "grad_norm": 4.4375, "learning_rate": 1.928172168327233e-05, "loss": 2.7197, "step": 883 }, { "epoch": 1.046167505179047, "grad_norm": 4.5625, "learning_rate": 1.927864018523963e-05, "loss": 2.7418, "step": 884 }, { "epoch": 1.0473512873631252, "grad_norm": 4.53125, "learning_rate": 1.927555233866294e-05, "loss": 2.2609, "step": 885 }, { "epoch": 1.0485350695472033, "grad_norm": 4.34375, "learning_rate": 1.927245814565499e-05, "loss": 2.4154, "step": 886 }, { "epoch": 1.0497188517312814, "grad_norm": 5.03125, "learning_rate": 1.9269357608332867e-05, "loss": 2.6026, "step": 887 }, { "epoch": 1.0509026339153595, "grad_norm": 4.09375, "learning_rate": 1.9266250728817985e-05, "loss": 2.7567, "step": 888 }, { "epoch": 1.0520864160994376, "grad_norm": 4.375, "learning_rate": 1.926313750923611e-05, "loss": 2.5849, "step": 889 }, { "epoch": 1.0532701982835158, "grad_norm": 4.75, "learning_rate": 1.9260017951717334e-05, "loss": 2.4421, "step": 890 }, { "epoch": 1.0544539804675939, "grad_norm": 4.65625, "learning_rate": 1.9256892058396098e-05, "loss": 2.7461, "step": 891 }, { "epoch": 1.0556377626516722, "grad_norm": 4.25, "learning_rate": 1.9253759831411166e-05, "loss": 2.633, "step": 892 }, { "epoch": 1.0568215448357503, "grad_norm": 4.375, "learning_rate": 1.9250621272905643e-05, "loss": 3.0122, "step": 893 }, { "epoch": 1.0580053270198284, "grad_norm": 4.375, "learning_rate": 1.924747638502696e-05, "loss": 2.5463, "step": 894 }, { "epoch": 1.0591891092039065, "grad_norm": 4.65625, "learning_rate": 1.924432516992689e-05, "loss": 2.4236, "step": 895 }, { "epoch": 1.0603728913879846, "grad_norm": 4.40625, "learning_rate": 1.9241167629761527e-05, "loss": 2.976, "step": 896 }, { "epoch": 1.0615566735720627, "grad_norm": 4.625, "learning_rate": 1.923800376669129e-05, "loss": 2.625, "step": 897 }, { "epoch": 1.0627404557561408, "grad_norm": 4.21875, "learning_rate": 1.9234833582880923e-05, "loss": 2.5064, "step": 898 }, { "epoch": 1.063924237940219, "grad_norm": 4.0, "learning_rate": 1.9231657080499507e-05, "loss": 2.2659, "step": 899 }, { "epoch": 1.065108020124297, "grad_norm": 4.09375, "learning_rate": 1.922847426172044e-05, "loss": 2.4006, "step": 900 }, { "epoch": 1.0662918023083752, "grad_norm": 4.1875, "learning_rate": 1.9225285128721435e-05, "loss": 2.6392, "step": 901 }, { "epoch": 1.0674755844924535, "grad_norm": 4.40625, "learning_rate": 1.9222089683684527e-05, "loss": 2.4323, "step": 902 }, { "epoch": 1.0686593666765316, "grad_norm": 4.6875, "learning_rate": 1.9218887928796083e-05, "loss": 2.7481, "step": 903 }, { "epoch": 1.0698431488606097, "grad_norm": 4.03125, "learning_rate": 1.9215679866246774e-05, "loss": 2.2299, "step": 904 }, { "epoch": 1.0710269310446878, "grad_norm": 4.03125, "learning_rate": 1.9212465498231587e-05, "loss": 2.7067, "step": 905 }, { "epoch": 1.072210713228766, "grad_norm": 5.53125, "learning_rate": 1.9209244826949833e-05, "loss": 2.6517, "step": 906 }, { "epoch": 1.073394495412844, "grad_norm": 5.0625, "learning_rate": 1.9206017854605122e-05, "loss": 2.6319, "step": 907 }, { "epoch": 1.0745782775969221, "grad_norm": 4.59375, "learning_rate": 1.9202784583405386e-05, "loss": 2.8945, "step": 908 }, { "epoch": 1.0757620597810003, "grad_norm": 4.59375, "learning_rate": 1.9199545015562866e-05, "loss": 2.7861, "step": 909 }, { "epoch": 1.0769458419650784, "grad_norm": 4.1875, "learning_rate": 1.9196299153294107e-05, "loss": 2.4771, "step": 910 }, { "epoch": 1.0781296241491565, "grad_norm": 3.9375, "learning_rate": 1.919304699881996e-05, "loss": 2.6498, "step": 911 }, { "epoch": 1.0793134063332346, "grad_norm": 4.5, "learning_rate": 1.9189788554365586e-05, "loss": 2.6156, "step": 912 }, { "epoch": 1.080497188517313, "grad_norm": 5.65625, "learning_rate": 1.918652382216045e-05, "loss": 2.7727, "step": 913 }, { "epoch": 1.081680970701391, "grad_norm": 4.53125, "learning_rate": 1.918325280443831e-05, "loss": 2.4498, "step": 914 }, { "epoch": 1.0828647528854691, "grad_norm": 4.1875, "learning_rate": 1.9179975503437235e-05, "loss": 2.3922, "step": 915 }, { "epoch": 1.0840485350695472, "grad_norm": 4.8125, "learning_rate": 1.9176691921399585e-05, "loss": 2.798, "step": 916 }, { "epoch": 1.0852323172536253, "grad_norm": 4.25, "learning_rate": 1.9173402060572028e-05, "loss": 2.6004, "step": 917 }, { "epoch": 1.0864160994377035, "grad_norm": 4.21875, "learning_rate": 1.9170105923205517e-05, "loss": 2.4725, "step": 918 }, { "epoch": 1.0875998816217816, "grad_norm": 4.5625, "learning_rate": 1.916680351155531e-05, "loss": 2.9994, "step": 919 }, { "epoch": 1.0887836638058597, "grad_norm": 4.5625, "learning_rate": 1.9163494827880943e-05, "loss": 2.7512, "step": 920 }, { "epoch": 1.0899674459899378, "grad_norm": 4.5625, "learning_rate": 1.916017987444626e-05, "loss": 2.2205, "step": 921 }, { "epoch": 1.0911512281740159, "grad_norm": 4.03125, "learning_rate": 1.915685865351938e-05, "loss": 2.2966, "step": 922 }, { "epoch": 1.0923350103580942, "grad_norm": 4.5625, "learning_rate": 1.9153531167372725e-05, "loss": 2.9036, "step": 923 }, { "epoch": 1.0935187925421723, "grad_norm": 4.5625, "learning_rate": 1.9150197418282993e-05, "loss": 2.6612, "step": 924 }, { "epoch": 1.0947025747262504, "grad_norm": 4.71875, "learning_rate": 1.9146857408531174e-05, "loss": 2.5755, "step": 925 }, { "epoch": 1.0958863569103285, "grad_norm": 3.90625, "learning_rate": 1.9143511140402532e-05, "loss": 2.4299, "step": 926 }, { "epoch": 1.0970701390944066, "grad_norm": 4.875, "learning_rate": 1.9140158616186625e-05, "loss": 2.8994, "step": 927 }, { "epoch": 1.0982539212784848, "grad_norm": 4.90625, "learning_rate": 1.913679983817728e-05, "loss": 2.7894, "step": 928 }, { "epoch": 1.0994377034625629, "grad_norm": 4.5625, "learning_rate": 1.913343480867261e-05, "loss": 2.6881, "step": 929 }, { "epoch": 1.100621485646641, "grad_norm": 4.46875, "learning_rate": 1.9130063529975004e-05, "loss": 2.4815, "step": 930 }, { "epoch": 1.101805267830719, "grad_norm": 4.3125, "learning_rate": 1.912668600439113e-05, "loss": 2.3617, "step": 931 }, { "epoch": 1.1029890500147972, "grad_norm": 4.5625, "learning_rate": 1.9123302234231924e-05, "loss": 2.5874, "step": 932 }, { "epoch": 1.1041728321988753, "grad_norm": 4.21875, "learning_rate": 1.91199122218126e-05, "loss": 3.0262, "step": 933 }, { "epoch": 1.1053566143829536, "grad_norm": 4.5625, "learning_rate": 1.9116515969452637e-05, "loss": 2.4456, "step": 934 }, { "epoch": 1.1065403965670317, "grad_norm": 4.4375, "learning_rate": 1.9113113479475784e-05, "loss": 2.6316, "step": 935 }, { "epoch": 1.1077241787511098, "grad_norm": 4.5625, "learning_rate": 1.910970475421007e-05, "loss": 2.3733, "step": 936 }, { "epoch": 1.108907960935188, "grad_norm": 3.953125, "learning_rate": 1.9106289795987774e-05, "loss": 2.563, "step": 937 }, { "epoch": 1.110091743119266, "grad_norm": 4.46875, "learning_rate": 1.910286860714545e-05, "loss": 2.6222, "step": 938 }, { "epoch": 1.1112755253033442, "grad_norm": 4.5625, "learning_rate": 1.909944119002391e-05, "loss": 2.3614, "step": 939 }, { "epoch": 1.1124593074874223, "grad_norm": 4.1875, "learning_rate": 1.909600754696823e-05, "loss": 2.225, "step": 940 }, { "epoch": 1.1136430896715004, "grad_norm": 4.90625, "learning_rate": 1.909256768032774e-05, "loss": 2.8916, "step": 941 }, { "epoch": 1.1148268718555785, "grad_norm": 4.46875, "learning_rate": 1.9089121592456044e-05, "loss": 2.5328, "step": 942 }, { "epoch": 1.1160106540396566, "grad_norm": 4.96875, "learning_rate": 1.908566928571098e-05, "loss": 2.8489, "step": 943 }, { "epoch": 1.117194436223735, "grad_norm": 4.5, "learning_rate": 1.908221076245466e-05, "loss": 2.4992, "step": 944 }, { "epoch": 1.118378218407813, "grad_norm": 5.5, "learning_rate": 1.9078746025053442e-05, "loss": 2.2961, "step": 945 }, { "epoch": 1.1195620005918911, "grad_norm": 4.46875, "learning_rate": 1.907527507587793e-05, "loss": 2.0889, "step": 946 }, { "epoch": 1.1207457827759693, "grad_norm": 4.3125, "learning_rate": 1.9071797917302995e-05, "loss": 2.9316, "step": 947 }, { "epoch": 1.1219295649600474, "grad_norm": 3.9375, "learning_rate": 1.9068314551707738e-05, "loss": 2.4495, "step": 948 }, { "epoch": 1.1231133471441255, "grad_norm": 4.0625, "learning_rate": 1.9064824981475514e-05, "loss": 2.4254, "step": 949 }, { "epoch": 1.1242971293282036, "grad_norm": 4.84375, "learning_rate": 1.9061329208993927e-05, "loss": 2.6113, "step": 950 }, { "epoch": 1.1254809115122817, "grad_norm": 4.4375, "learning_rate": 1.905782723665482e-05, "loss": 2.4485, "step": 951 }, { "epoch": 1.1266646936963598, "grad_norm": 4.53125, "learning_rate": 1.9054319066854285e-05, "loss": 2.6314, "step": 952 }, { "epoch": 1.127848475880438, "grad_norm": 4.46875, "learning_rate": 1.905080470199264e-05, "loss": 3.0611, "step": 953 }, { "epoch": 1.129032258064516, "grad_norm": 4.15625, "learning_rate": 1.904728414447446e-05, "loss": 2.4389, "step": 954 }, { "epoch": 1.1302160402485943, "grad_norm": 5.125, "learning_rate": 1.9043757396708536e-05, "loss": 2.9296, "step": 955 }, { "epoch": 1.1313998224326725, "grad_norm": 4.375, "learning_rate": 1.9040224461107918e-05, "loss": 2.9357, "step": 956 }, { "epoch": 1.1325836046167506, "grad_norm": 5.03125, "learning_rate": 1.903668534008987e-05, "loss": 2.8333, "step": 957 }, { "epoch": 1.1337673868008287, "grad_norm": 4.46875, "learning_rate": 1.90331400360759e-05, "loss": 2.5514, "step": 958 }, { "epoch": 1.1349511689849068, "grad_norm": 4.75, "learning_rate": 1.9029588551491743e-05, "loss": 2.699, "step": 959 }, { "epoch": 1.1361349511689849, "grad_norm": 4.53125, "learning_rate": 1.9026030888767364e-05, "loss": 2.4681, "step": 960 }, { "epoch": 1.137318733353063, "grad_norm": 4.1875, "learning_rate": 1.9022467050336945e-05, "loss": 2.7829, "step": 961 }, { "epoch": 1.138502515537141, "grad_norm": 4.53125, "learning_rate": 1.901889703863891e-05, "loss": 2.5516, "step": 962 }, { "epoch": 1.1396862977212192, "grad_norm": 4.0625, "learning_rate": 1.90153208561159e-05, "loss": 2.7109, "step": 963 }, { "epoch": 1.1408700799052975, "grad_norm": 5.09375, "learning_rate": 1.901173850521477e-05, "loss": 2.7871, "step": 964 }, { "epoch": 1.1420538620893756, "grad_norm": 4.6875, "learning_rate": 1.9008149988386608e-05, "loss": 2.4154, "step": 965 }, { "epoch": 1.1432376442734538, "grad_norm": 4.09375, "learning_rate": 1.900455530808671e-05, "loss": 2.7106, "step": 966 }, { "epoch": 1.1444214264575319, "grad_norm": 4.84375, "learning_rate": 1.9000954466774603e-05, "loss": 3.0353, "step": 967 }, { "epoch": 1.14560520864161, "grad_norm": 4.21875, "learning_rate": 1.8997347466914014e-05, "loss": 2.5992, "step": 968 }, { "epoch": 1.146788990825688, "grad_norm": 4.09375, "learning_rate": 1.899373431097289e-05, "loss": 2.4005, "step": 969 }, { "epoch": 1.1479727730097662, "grad_norm": 4.90625, "learning_rate": 1.8990115001423396e-05, "loss": 2.5662, "step": 970 }, { "epoch": 1.1491565551938443, "grad_norm": 4.40625, "learning_rate": 1.8986489540741895e-05, "loss": 2.5324, "step": 971 }, { "epoch": 1.1503403373779224, "grad_norm": 4.375, "learning_rate": 1.8982857931408972e-05, "loss": 2.4786, "step": 972 }, { "epoch": 1.1515241195620005, "grad_norm": 4.46875, "learning_rate": 1.8979220175909408e-05, "loss": 2.3976, "step": 973 }, { "epoch": 1.1527079017460786, "grad_norm": 4.03125, "learning_rate": 1.8975576276732196e-05, "loss": 2.8298, "step": 974 }, { "epoch": 1.1538916839301567, "grad_norm": 4.40625, "learning_rate": 1.8971926236370532e-05, "loss": 2.2758, "step": 975 }, { "epoch": 1.155075466114235, "grad_norm": 4.8125, "learning_rate": 1.896827005732181e-05, "loss": 2.6975, "step": 976 }, { "epoch": 1.1562592482983132, "grad_norm": 4.21875, "learning_rate": 1.8964607742087626e-05, "loss": 2.3422, "step": 977 }, { "epoch": 1.1574430304823913, "grad_norm": 4.875, "learning_rate": 1.8960939293173777e-05, "loss": 2.6862, "step": 978 }, { "epoch": 1.1586268126664694, "grad_norm": 4.46875, "learning_rate": 1.8957264713090256e-05, "loss": 2.7147, "step": 979 }, { "epoch": 1.1598105948505475, "grad_norm": 4.375, "learning_rate": 1.8953584004351243e-05, "loss": 2.5892, "step": 980 }, { "epoch": 1.1609943770346256, "grad_norm": 4.75, "learning_rate": 1.8949897169475126e-05, "loss": 2.5837, "step": 981 }, { "epoch": 1.1621781592187037, "grad_norm": 5.125, "learning_rate": 1.8946204210984468e-05, "loss": 2.5486, "step": 982 }, { "epoch": 1.1633619414027818, "grad_norm": 4.09375, "learning_rate": 1.894250513140604e-05, "loss": 2.914, "step": 983 }, { "epoch": 1.16454572358686, "grad_norm": 4.90625, "learning_rate": 1.8938799933270784e-05, "loss": 2.6898, "step": 984 }, { "epoch": 1.1657295057709383, "grad_norm": 4.96875, "learning_rate": 1.893508861911384e-05, "loss": 2.8653, "step": 985 }, { "epoch": 1.1669132879550164, "grad_norm": 4.28125, "learning_rate": 1.8931371191474526e-05, "loss": 2.5123, "step": 986 }, { "epoch": 1.1680970701390945, "grad_norm": 4.34375, "learning_rate": 1.8927647652896347e-05, "loss": 2.729, "step": 987 }, { "epoch": 1.1692808523231726, "grad_norm": 4.875, "learning_rate": 1.8923918005926984e-05, "loss": 2.4978, "step": 988 }, { "epoch": 1.1704646345072507, "grad_norm": 4.65625, "learning_rate": 1.892018225311831e-05, "loss": 2.8551, "step": 989 }, { "epoch": 1.1716484166913288, "grad_norm": 4.5, "learning_rate": 1.8916440397026355e-05, "loss": 2.5271, "step": 990 }, { "epoch": 1.172832198875407, "grad_norm": 3.90625, "learning_rate": 1.8912692440211345e-05, "loss": 2.3416, "step": 991 }, { "epoch": 1.174015981059485, "grad_norm": 4.40625, "learning_rate": 1.8908938385237667e-05, "loss": 2.5373, "step": 992 }, { "epoch": 1.1751997632435631, "grad_norm": 4.0625, "learning_rate": 1.8905178234673886e-05, "loss": 2.5744, "step": 993 }, { "epoch": 1.1763835454276412, "grad_norm": 4.40625, "learning_rate": 1.890141199109274e-05, "loss": 2.5072, "step": 994 }, { "epoch": 1.1775673276117193, "grad_norm": 4.375, "learning_rate": 1.8897639657071135e-05, "loss": 2.2451, "step": 995 }, { "epoch": 1.1787511097957977, "grad_norm": 4.0625, "learning_rate": 1.889386123519013e-05, "loss": 2.3775, "step": 996 }, { "epoch": 1.1799348919798758, "grad_norm": 4.875, "learning_rate": 1.8890076728034974e-05, "loss": 2.7903, "step": 997 }, { "epoch": 1.1811186741639539, "grad_norm": 4.6875, "learning_rate": 1.8886286138195063e-05, "loss": 2.6283, "step": 998 }, { "epoch": 1.182302456348032, "grad_norm": 4.78125, "learning_rate": 1.8882489468263954e-05, "loss": 2.5453, "step": 999 }, { "epoch": 1.18348623853211, "grad_norm": 4.65625, "learning_rate": 1.8878686720839378e-05, "loss": 2.5393, "step": 1000 }, { "epoch": 1.1846700207161882, "grad_norm": 4.6875, "learning_rate": 1.8874877898523204e-05, "loss": 2.8348, "step": 1001 }, { "epoch": 1.1858538029002663, "grad_norm": 4.15625, "learning_rate": 1.887106300392148e-05, "loss": 2.3043, "step": 1002 }, { "epoch": 1.1870375850843444, "grad_norm": 4.9375, "learning_rate": 1.8867242039644387e-05, "loss": 2.5853, "step": 1003 }, { "epoch": 1.1882213672684225, "grad_norm": 4.78125, "learning_rate": 1.8863415008306277e-05, "loss": 2.688, "step": 1004 }, { "epoch": 1.1894051494525006, "grad_norm": 4.15625, "learning_rate": 1.8859581912525636e-05, "loss": 2.5532, "step": 1005 }, { "epoch": 1.190588931636579, "grad_norm": 4.5, "learning_rate": 1.8855742754925122e-05, "loss": 2.5092, "step": 1006 }, { "epoch": 1.191772713820657, "grad_norm": 4.1875, "learning_rate": 1.885189753813152e-05, "loss": 1.9922, "step": 1007 }, { "epoch": 1.1929564960047352, "grad_norm": 4.09375, "learning_rate": 1.8848046264775766e-05, "loss": 2.553, "step": 1008 }, { "epoch": 1.1941402781888133, "grad_norm": 5.125, "learning_rate": 1.8844188937492948e-05, "loss": 2.6262, "step": 1009 }, { "epoch": 1.1953240603728914, "grad_norm": 4.21875, "learning_rate": 1.8840325558922285e-05, "loss": 2.2356, "step": 1010 }, { "epoch": 1.1965078425569695, "grad_norm": 4.96875, "learning_rate": 1.8836456131707143e-05, "loss": 2.0316, "step": 1011 }, { "epoch": 1.1976916247410476, "grad_norm": 4.71875, "learning_rate": 1.8832580658495026e-05, "loss": 2.5724, "step": 1012 }, { "epoch": 1.1988754069251257, "grad_norm": 4.71875, "learning_rate": 1.882869914193757e-05, "loss": 2.8693, "step": 1013 }, { "epoch": 1.2000591891092038, "grad_norm": 4.34375, "learning_rate": 1.8824811584690557e-05, "loss": 2.6483, "step": 1014 }, { "epoch": 1.201242971293282, "grad_norm": 4.78125, "learning_rate": 1.882091798941389e-05, "loss": 2.8219, "step": 1015 }, { "epoch": 1.20242675347736, "grad_norm": 4.5625, "learning_rate": 1.8817018358771612e-05, "loss": 2.7421, "step": 1016 }, { "epoch": 1.2036105356614384, "grad_norm": 4.21875, "learning_rate": 1.8813112695431885e-05, "loss": 2.3775, "step": 1017 }, { "epoch": 1.2047943178455165, "grad_norm": 3.921875, "learning_rate": 1.880920100206701e-05, "loss": 2.6713, "step": 1018 }, { "epoch": 1.2059781000295946, "grad_norm": 4.03125, "learning_rate": 1.8805283281353413e-05, "loss": 2.2959, "step": 1019 }, { "epoch": 1.2071618822136727, "grad_norm": 4.84375, "learning_rate": 1.8801359535971628e-05, "loss": 2.5452, "step": 1020 }, { "epoch": 1.2083456643977508, "grad_norm": 4.5625, "learning_rate": 1.8797429768606336e-05, "loss": 2.4283, "step": 1021 }, { "epoch": 1.209529446581829, "grad_norm": 4.0625, "learning_rate": 1.879349398194632e-05, "loss": 2.4009, "step": 1022 }, { "epoch": 1.210713228765907, "grad_norm": 4.84375, "learning_rate": 1.8789552178684483e-05, "loss": 2.8035, "step": 1023 }, { "epoch": 1.2118970109499851, "grad_norm": 4.625, "learning_rate": 1.8785604361517852e-05, "loss": 2.4967, "step": 1024 }, { "epoch": 1.2130807931340633, "grad_norm": 4.75, "learning_rate": 1.8781650533147572e-05, "loss": 2.4462, "step": 1025 }, { "epoch": 1.2142645753181416, "grad_norm": 4.90625, "learning_rate": 1.877769069627888e-05, "loss": 2.5485, "step": 1026 }, { "epoch": 1.2154483575022197, "grad_norm": 4.71875, "learning_rate": 1.8773724853621154e-05, "loss": 2.5827, "step": 1027 }, { "epoch": 1.2166321396862978, "grad_norm": 4.71875, "learning_rate": 1.8769753007887858e-05, "loss": 2.4994, "step": 1028 }, { "epoch": 1.217815921870376, "grad_norm": 4.53125, "learning_rate": 1.8765775161796565e-05, "loss": 2.6071, "step": 1029 }, { "epoch": 1.218999704054454, "grad_norm": 4.375, "learning_rate": 1.8761791318068972e-05, "loss": 2.5286, "step": 1030 }, { "epoch": 1.2201834862385321, "grad_norm": 5.21875, "learning_rate": 1.8757801479430858e-05, "loss": 2.6873, "step": 1031 }, { "epoch": 1.2213672684226102, "grad_norm": 4.59375, "learning_rate": 1.8753805648612115e-05, "loss": 2.639, "step": 1032 }, { "epoch": 1.2225510506066883, "grad_norm": 4.28125, "learning_rate": 1.8749803828346732e-05, "loss": 2.5839, "step": 1033 }, { "epoch": 1.2237348327907664, "grad_norm": 4.28125, "learning_rate": 1.87457960213728e-05, "loss": 2.0155, "step": 1034 }, { "epoch": 1.2249186149748446, "grad_norm": 4.71875, "learning_rate": 1.8741782230432503e-05, "loss": 2.6397, "step": 1035 }, { "epoch": 1.2261023971589227, "grad_norm": 4.75, "learning_rate": 1.8737762458272114e-05, "loss": 2.8569, "step": 1036 }, { "epoch": 1.2272861793430008, "grad_norm": 4.96875, "learning_rate": 1.8733736707642012e-05, "loss": 2.9823, "step": 1037 }, { "epoch": 1.228469961527079, "grad_norm": 4.40625, "learning_rate": 1.8729704981296654e-05, "loss": 2.5711, "step": 1038 }, { "epoch": 1.2296537437111572, "grad_norm": 4.75, "learning_rate": 1.8725667281994584e-05, "loss": 2.8559, "step": 1039 }, { "epoch": 1.2308375258952353, "grad_norm": 4.46875, "learning_rate": 1.872162361249845e-05, "loss": 2.7566, "step": 1040 }, { "epoch": 1.2320213080793134, "grad_norm": 3.96875, "learning_rate": 1.8717573975574962e-05, "loss": 2.6358, "step": 1041 }, { "epoch": 1.2332050902633915, "grad_norm": 4.1875, "learning_rate": 1.8713518373994932e-05, "loss": 2.6499, "step": 1042 }, { "epoch": 1.2343888724474696, "grad_norm": 4.84375, "learning_rate": 1.8709456810533248e-05, "loss": 2.4739, "step": 1043 }, { "epoch": 1.2355726546315478, "grad_norm": 4.21875, "learning_rate": 1.8705389287968866e-05, "loss": 2.8616, "step": 1044 }, { "epoch": 1.2367564368156259, "grad_norm": 4.09375, "learning_rate": 1.8701315809084835e-05, "loss": 2.2364, "step": 1045 }, { "epoch": 1.237940218999704, "grad_norm": 4.59375, "learning_rate": 1.869723637666827e-05, "loss": 2.8452, "step": 1046 }, { "epoch": 1.2391240011837823, "grad_norm": 4.34375, "learning_rate": 1.869315099351036e-05, "loss": 2.382, "step": 1047 }, { "epoch": 1.2403077833678604, "grad_norm": 4.28125, "learning_rate": 1.8689059662406373e-05, "loss": 2.9791, "step": 1048 }, { "epoch": 1.2414915655519385, "grad_norm": 4.53125, "learning_rate": 1.8684962386155633e-05, "loss": 2.2928, "step": 1049 }, { "epoch": 1.2426753477360166, "grad_norm": 5.28125, "learning_rate": 1.868085916756155e-05, "loss": 2.9982, "step": 1050 }, { "epoch": 1.2438591299200947, "grad_norm": 4.53125, "learning_rate": 1.867675000943158e-05, "loss": 2.4284, "step": 1051 }, { "epoch": 1.2450429121041728, "grad_norm": 4.625, "learning_rate": 1.867263491457726e-05, "loss": 2.5898, "step": 1052 }, { "epoch": 1.246226694288251, "grad_norm": 4.875, "learning_rate": 1.8668513885814176e-05, "loss": 2.9048, "step": 1053 }, { "epoch": 1.247410476472329, "grad_norm": 4.5625, "learning_rate": 1.866438692596198e-05, "loss": 2.8116, "step": 1054 }, { "epoch": 1.2485942586564072, "grad_norm": 5.125, "learning_rate": 1.866025403784439e-05, "loss": 2.9503, "step": 1055 }, { "epoch": 1.2497780408404853, "grad_norm": 4.6875, "learning_rate": 1.865611522428916e-05, "loss": 2.7029, "step": 1056 }, { "epoch": 1.2509618230245634, "grad_norm": 4.1875, "learning_rate": 1.8651970488128115e-05, "loss": 2.2373, "step": 1057 }, { "epoch": 1.2521456052086415, "grad_norm": 4.53125, "learning_rate": 1.864781983219713e-05, "loss": 2.8468, "step": 1058 }, { "epoch": 1.2533293873927198, "grad_norm": 4.625, "learning_rate": 1.8643663259336126e-05, "loss": 2.8442, "step": 1059 }, { "epoch": 1.254513169576798, "grad_norm": 4.34375, "learning_rate": 1.8639500772389073e-05, "loss": 2.2124, "step": 1060 }, { "epoch": 1.255696951760876, "grad_norm": 4.25, "learning_rate": 1.8635332374203993e-05, "loss": 2.3626, "step": 1061 }, { "epoch": 1.2568807339449541, "grad_norm": 4.5625, "learning_rate": 1.8631158067632942e-05, "loss": 2.7902, "step": 1062 }, { "epoch": 1.2580645161290323, "grad_norm": 5.34375, "learning_rate": 1.862697785553203e-05, "loss": 2.596, "step": 1063 }, { "epoch": 1.2592482983131104, "grad_norm": 3.921875, "learning_rate": 1.8622791740761398e-05, "loss": 2.5014, "step": 1064 }, { "epoch": 1.2604320804971885, "grad_norm": 4.21875, "learning_rate": 1.861859972618523e-05, "loss": 2.6159, "step": 1065 }, { "epoch": 1.2616158626812666, "grad_norm": 4.78125, "learning_rate": 1.861440181467175e-05, "loss": 2.585, "step": 1066 }, { "epoch": 1.262799644865345, "grad_norm": 4.5, "learning_rate": 1.861019800909322e-05, "loss": 2.4301, "step": 1067 }, { "epoch": 1.263983427049423, "grad_norm": 4.6875, "learning_rate": 1.8605988312325915e-05, "loss": 2.4612, "step": 1068 }, { "epoch": 1.2651672092335011, "grad_norm": 4.1875, "learning_rate": 1.8601772727250158e-05, "loss": 2.5472, "step": 1069 }, { "epoch": 1.2663509914175792, "grad_norm": 4.28125, "learning_rate": 1.85975512567503e-05, "loss": 2.236, "step": 1070 }, { "epoch": 1.2675347736016573, "grad_norm": 4.15625, "learning_rate": 1.8593323903714718e-05, "loss": 2.5559, "step": 1071 }, { "epoch": 1.2687185557857354, "grad_norm": 4.5, "learning_rate": 1.858909067103581e-05, "loss": 2.3871, "step": 1072 }, { "epoch": 1.2699023379698136, "grad_norm": 5.21875, "learning_rate": 1.8584851561609995e-05, "loss": 2.4459, "step": 1073 }, { "epoch": 1.2710861201538917, "grad_norm": 4.03125, "learning_rate": 1.8580606578337718e-05, "loss": 2.678, "step": 1074 }, { "epoch": 1.2722699023379698, "grad_norm": 4.9375, "learning_rate": 1.8576355724123446e-05, "loss": 2.6419, "step": 1075 }, { "epoch": 1.2734536845220479, "grad_norm": 4.65625, "learning_rate": 1.857209900187565e-05, "loss": 2.4828, "step": 1076 }, { "epoch": 1.274637466706126, "grad_norm": 5.21875, "learning_rate": 1.8567836414506835e-05, "loss": 2.7105, "step": 1077 }, { "epoch": 1.275821248890204, "grad_norm": 4.15625, "learning_rate": 1.85635679649335e-05, "loss": 2.5663, "step": 1078 }, { "epoch": 1.2770050310742822, "grad_norm": 4.375, "learning_rate": 1.8559293656076167e-05, "loss": 2.3501, "step": 1079 }, { "epoch": 1.2781888132583605, "grad_norm": 4.375, "learning_rate": 1.8555013490859367e-05, "loss": 2.4822, "step": 1080 }, { "epoch": 1.2793725954424386, "grad_norm": 4.21875, "learning_rate": 1.8550727472211624e-05, "loss": 2.352, "step": 1081 }, { "epoch": 1.2805563776265168, "grad_norm": 5.03125, "learning_rate": 1.8546435603065488e-05, "loss": 2.7915, "step": 1082 }, { "epoch": 1.2817401598105949, "grad_norm": 3.96875, "learning_rate": 1.85421378863575e-05, "loss": 2.3281, "step": 1083 }, { "epoch": 1.282923941994673, "grad_norm": 4.40625, "learning_rate": 1.8537834325028196e-05, "loss": 2.5629, "step": 1084 }, { "epoch": 1.284107724178751, "grad_norm": 5.09375, "learning_rate": 1.8533524922022123e-05, "loss": 2.7008, "step": 1085 }, { "epoch": 1.2852915063628292, "grad_norm": 4.75, "learning_rate": 1.852920968028782e-05, "loss": 2.6991, "step": 1086 }, { "epoch": 1.2864752885469073, "grad_norm": 4.15625, "learning_rate": 1.852488860277782e-05, "loss": 2.1747, "step": 1087 }, { "epoch": 1.2876590707309856, "grad_norm": 4.71875, "learning_rate": 1.8520561692448655e-05, "loss": 2.7122, "step": 1088 }, { "epoch": 1.2888428529150637, "grad_norm": 4.40625, "learning_rate": 1.851622895226084e-05, "loss": 2.7581, "step": 1089 }, { "epoch": 1.2900266350991418, "grad_norm": 4.5625, "learning_rate": 1.851189038517888e-05, "loss": 2.92, "step": 1090 }, { "epoch": 1.29121041728322, "grad_norm": 4.78125, "learning_rate": 1.850754599417127e-05, "loss": 2.6662, "step": 1091 }, { "epoch": 1.292394199467298, "grad_norm": 4.1875, "learning_rate": 1.8503195782210484e-05, "loss": 2.7061, "step": 1092 }, { "epoch": 1.2935779816513762, "grad_norm": 4.09375, "learning_rate": 1.849883975227299e-05, "loss": 2.242, "step": 1093 }, { "epoch": 1.2947617638354543, "grad_norm": 4.65625, "learning_rate": 1.8494477907339224e-05, "loss": 2.8971, "step": 1094 }, { "epoch": 1.2959455460195324, "grad_norm": 4.375, "learning_rate": 1.8490110250393617e-05, "loss": 2.415, "step": 1095 }, { "epoch": 1.2971293282036105, "grad_norm": 4.28125, "learning_rate": 1.8485736784424554e-05, "loss": 2.3724, "step": 1096 }, { "epoch": 1.2983131103876886, "grad_norm": 4.0, "learning_rate": 1.848135751242441e-05, "loss": 2.4682, "step": 1097 }, { "epoch": 1.2994968925717667, "grad_norm": 4.34375, "learning_rate": 1.8476972437389535e-05, "loss": 2.6683, "step": 1098 }, { "epoch": 1.3006806747558448, "grad_norm": 4.6875, "learning_rate": 1.8472581562320235e-05, "loss": 2.9325, "step": 1099 }, { "epoch": 1.301864456939923, "grad_norm": 4.15625, "learning_rate": 1.8468184890220803e-05, "loss": 2.3947, "step": 1100 }, { "epoch": 1.3030482391240013, "grad_norm": 4.0625, "learning_rate": 1.8463782424099478e-05, "loss": 2.6796, "step": 1101 }, { "epoch": 1.3042320213080794, "grad_norm": 4.1875, "learning_rate": 1.8459374166968483e-05, "loss": 2.4751, "step": 1102 }, { "epoch": 1.3054158034921575, "grad_norm": 4.1875, "learning_rate": 1.845496012184399e-05, "loss": 2.3151, "step": 1103 }, { "epoch": 1.3065995856762356, "grad_norm": 4.78125, "learning_rate": 1.8450540291746138e-05, "loss": 2.4611, "step": 1104 }, { "epoch": 1.3077833678603137, "grad_norm": 4.65625, "learning_rate": 1.844611467969902e-05, "loss": 2.424, "step": 1105 }, { "epoch": 1.3089671500443918, "grad_norm": 4.125, "learning_rate": 1.8441683288730686e-05, "loss": 2.2544, "step": 1106 }, { "epoch": 1.31015093222847, "grad_norm": 3.875, "learning_rate": 1.8437246121873143e-05, "loss": 2.0394, "step": 1107 }, { "epoch": 1.311334714412548, "grad_norm": 4.4375, "learning_rate": 1.8432803182162346e-05, "loss": 2.4734, "step": 1108 }, { "epoch": 1.3125184965966263, "grad_norm": 4.625, "learning_rate": 1.84283544726382e-05, "loss": 2.6486, "step": 1109 }, { "epoch": 1.3137022787807044, "grad_norm": 4.5, "learning_rate": 1.842389999634456e-05, "loss": 2.875, "step": 1110 }, { "epoch": 1.3148860609647826, "grad_norm": 4.65625, "learning_rate": 1.8419439756329226e-05, "loss": 2.2371, "step": 1111 }, { "epoch": 1.3160698431488607, "grad_norm": 4.71875, "learning_rate": 1.8414973755643942e-05, "loss": 2.5338, "step": 1112 }, { "epoch": 1.3172536253329388, "grad_norm": 4.15625, "learning_rate": 1.8410501997344394e-05, "loss": 2.7477, "step": 1113 }, { "epoch": 1.3184374075170169, "grad_norm": 4.46875, "learning_rate": 1.8406024484490207e-05, "loss": 2.6165, "step": 1114 }, { "epoch": 1.319621189701095, "grad_norm": 3.671875, "learning_rate": 1.840154122014494e-05, "loss": 2.2214, "step": 1115 }, { "epoch": 1.320804971885173, "grad_norm": 4.25, "learning_rate": 1.839705220737609e-05, "loss": 2.2074, "step": 1116 }, { "epoch": 1.3219887540692512, "grad_norm": 5.1875, "learning_rate": 1.8392557449255094e-05, "loss": 2.7187, "step": 1117 }, { "epoch": 1.3231725362533293, "grad_norm": 4.96875, "learning_rate": 1.8388056948857303e-05, "loss": 2.8812, "step": 1118 }, { "epoch": 1.3243563184374074, "grad_norm": 4.15625, "learning_rate": 1.838355070926201e-05, "loss": 2.3947, "step": 1119 }, { "epoch": 1.3255401006214855, "grad_norm": 4.40625, "learning_rate": 1.8379038733552435e-05, "loss": 2.422, "step": 1120 }, { "epoch": 1.3267238828055639, "grad_norm": 4.9375, "learning_rate": 1.837452102481572e-05, "loss": 2.3949, "step": 1121 }, { "epoch": 1.327907664989642, "grad_norm": 4.21875, "learning_rate": 1.836999758614293e-05, "loss": 2.5187, "step": 1122 }, { "epoch": 1.32909144717372, "grad_norm": 4.625, "learning_rate": 1.836546842062904e-05, "loss": 2.6798, "step": 1123 }, { "epoch": 1.3302752293577982, "grad_norm": 4.75, "learning_rate": 1.836093353137297e-05, "loss": 3.1093, "step": 1124 }, { "epoch": 1.3314590115418763, "grad_norm": 4.0, "learning_rate": 1.835639292147752e-05, "loss": 2.342, "step": 1125 }, { "epoch": 1.3326427937259544, "grad_norm": 5.40625, "learning_rate": 1.835184659404944e-05, "loss": 2.6164, "step": 1126 }, { "epoch": 1.3338265759100325, "grad_norm": 4.34375, "learning_rate": 1.8347294552199366e-05, "loss": 2.6084, "step": 1127 }, { "epoch": 1.3350103580941106, "grad_norm": 4.8125, "learning_rate": 1.834273679904185e-05, "loss": 2.6156, "step": 1128 }, { "epoch": 1.336194140278189, "grad_norm": 4.3125, "learning_rate": 1.833817333769537e-05, "loss": 2.4742, "step": 1129 }, { "epoch": 1.337377922462267, "grad_norm": 4.3125, "learning_rate": 1.833360417128228e-05, "loss": 2.6574, "step": 1130 }, { "epoch": 1.3385617046463452, "grad_norm": 4.125, "learning_rate": 1.8329029302928858e-05, "loss": 2.7547, "step": 1131 }, { "epoch": 1.3397454868304233, "grad_norm": 4.0, "learning_rate": 1.8324448735765278e-05, "loss": 2.4209, "step": 1132 }, { "epoch": 1.3409292690145014, "grad_norm": 5.09375, "learning_rate": 1.831986247292561e-05, "loss": 2.3412, "step": 1133 }, { "epoch": 1.3421130511985795, "grad_norm": 4.15625, "learning_rate": 1.831527051754783e-05, "loss": 2.254, "step": 1134 }, { "epoch": 1.3432968333826576, "grad_norm": 4.5, "learning_rate": 1.8310672872773788e-05, "loss": 2.6987, "step": 1135 }, { "epoch": 1.3444806155667357, "grad_norm": 4.5625, "learning_rate": 1.830606954174926e-05, "loss": 2.986, "step": 1136 }, { "epoch": 1.3456643977508138, "grad_norm": 4.59375, "learning_rate": 1.830146052762388e-05, "loss": 2.4428, "step": 1137 }, { "epoch": 1.346848179934892, "grad_norm": 5.0, "learning_rate": 1.8296845833551192e-05, "loss": 2.7552, "step": 1138 }, { "epoch": 1.34803196211897, "grad_norm": 4.0, "learning_rate": 1.8292225462688616e-05, "loss": 2.2925, "step": 1139 }, { "epoch": 1.3492157443030481, "grad_norm": 4.71875, "learning_rate": 1.8287599418197457e-05, "loss": 2.9589, "step": 1140 }, { "epoch": 1.3503995264871262, "grad_norm": 4.34375, "learning_rate": 1.8282967703242905e-05, "loss": 2.3513, "step": 1141 }, { "epoch": 1.3515833086712046, "grad_norm": 4.375, "learning_rate": 1.8278330320994035e-05, "loss": 2.2683, "step": 1142 }, { "epoch": 1.3527670908552827, "grad_norm": 4.34375, "learning_rate": 1.827368727462378e-05, "loss": 2.1976, "step": 1143 }, { "epoch": 1.3539508730393608, "grad_norm": 4.8125, "learning_rate": 1.8269038567308968e-05, "loss": 2.6852, "step": 1144 }, { "epoch": 1.355134655223439, "grad_norm": 4.625, "learning_rate": 1.82643842022303e-05, "loss": 2.718, "step": 1145 }, { "epoch": 1.356318437407517, "grad_norm": 4.84375, "learning_rate": 1.825972418257233e-05, "loss": 2.4865, "step": 1146 }, { "epoch": 1.3575022195915951, "grad_norm": 4.28125, "learning_rate": 1.82550585115235e-05, "loss": 2.3886, "step": 1147 }, { "epoch": 1.3586860017756732, "grad_norm": 4.9375, "learning_rate": 1.8250387192276115e-05, "loss": 2.6228, "step": 1148 }, { "epoch": 1.3598697839597513, "grad_norm": 4.65625, "learning_rate": 1.8245710228026336e-05, "loss": 2.3396, "step": 1149 }, { "epoch": 1.3610535661438297, "grad_norm": 4.53125, "learning_rate": 1.824102762197419e-05, "loss": 2.4541, "step": 1150 }, { "epoch": 1.3622373483279078, "grad_norm": 4.8125, "learning_rate": 1.823633937732357e-05, "loss": 2.457, "step": 1151 }, { "epoch": 1.3634211305119859, "grad_norm": 4.40625, "learning_rate": 1.8231645497282218e-05, "loss": 2.5605, "step": 1152 }, { "epoch": 1.364604912696064, "grad_norm": 4.59375, "learning_rate": 1.8226945985061743e-05, "loss": 2.5264, "step": 1153 }, { "epoch": 1.365788694880142, "grad_norm": 4.21875, "learning_rate": 1.8222240843877593e-05, "loss": 2.3909, "step": 1154 }, { "epoch": 1.3669724770642202, "grad_norm": 4.96875, "learning_rate": 1.8217530076949083e-05, "loss": 2.5238, "step": 1155 }, { "epoch": 1.3681562592482983, "grad_norm": 4.5625, "learning_rate": 1.8212813687499365e-05, "loss": 2.5568, "step": 1156 }, { "epoch": 1.3693400414323764, "grad_norm": 4.71875, "learning_rate": 1.8208091678755444e-05, "loss": 2.6436, "step": 1157 }, { "epoch": 1.3705238236164545, "grad_norm": 4.375, "learning_rate": 1.8203364053948173e-05, "loss": 2.4389, "step": 1158 }, { "epoch": 1.3717076058005326, "grad_norm": 4.625, "learning_rate": 1.8198630816312236e-05, "loss": 2.5581, "step": 1159 }, { "epoch": 1.3728913879846107, "grad_norm": 4.0625, "learning_rate": 1.8193891969086164e-05, "loss": 2.608, "step": 1160 }, { "epoch": 1.3740751701686889, "grad_norm": 4.4375, "learning_rate": 1.818914751551233e-05, "loss": 2.4879, "step": 1161 }, { "epoch": 1.375258952352767, "grad_norm": 5.03125, "learning_rate": 1.818439745883694e-05, "loss": 2.6737, "step": 1162 }, { "epoch": 1.3764427345368453, "grad_norm": 4.125, "learning_rate": 1.8179641802310035e-05, "loss": 2.4277, "step": 1163 }, { "epoch": 1.3776265167209234, "grad_norm": 3.921875, "learning_rate": 1.817488054918548e-05, "loss": 1.9466, "step": 1164 }, { "epoch": 1.3788102989050015, "grad_norm": 5.0, "learning_rate": 1.817011370272098e-05, "loss": 2.7201, "step": 1165 }, { "epoch": 1.3799940810890796, "grad_norm": 4.3125, "learning_rate": 1.8165341266178056e-05, "loss": 2.4141, "step": 1166 }, { "epoch": 1.3811778632731577, "grad_norm": 4.15625, "learning_rate": 1.8160563242822068e-05, "loss": 2.6925, "step": 1167 }, { "epoch": 1.3823616454572358, "grad_norm": 5.21875, "learning_rate": 1.8155779635922178e-05, "loss": 2.3468, "step": 1168 }, { "epoch": 1.383545427641314, "grad_norm": 4.65625, "learning_rate": 1.8150990448751393e-05, "loss": 2.5707, "step": 1169 }, { "epoch": 1.384729209825392, "grad_norm": 5.375, "learning_rate": 1.814619568458652e-05, "loss": 2.5062, "step": 1170 }, { "epoch": 1.3859129920094704, "grad_norm": 5.03125, "learning_rate": 1.8141395346708192e-05, "loss": 2.7683, "step": 1171 }, { "epoch": 1.3870967741935485, "grad_norm": 4.34375, "learning_rate": 1.8136589438400843e-05, "loss": 2.496, "step": 1172 }, { "epoch": 1.3882805563776266, "grad_norm": 4.46875, "learning_rate": 1.8131777962952732e-05, "loss": 3.0078, "step": 1173 }, { "epoch": 1.3894643385617047, "grad_norm": 4.65625, "learning_rate": 1.8126960923655915e-05, "loss": 2.5032, "step": 1174 }, { "epoch": 1.3906481207457828, "grad_norm": 4.71875, "learning_rate": 1.812213832380627e-05, "loss": 2.4877, "step": 1175 }, { "epoch": 1.391831902929861, "grad_norm": 3.84375, "learning_rate": 1.811731016670347e-05, "loss": 2.5772, "step": 1176 }, { "epoch": 1.393015685113939, "grad_norm": 4.5, "learning_rate": 1.8112476455650987e-05, "loss": 2.9662, "step": 1177 }, { "epoch": 1.3941994672980171, "grad_norm": 4.0, "learning_rate": 1.8107637193956102e-05, "loss": 2.5746, "step": 1178 }, { "epoch": 1.3953832494820952, "grad_norm": 4.15625, "learning_rate": 1.810279238492988e-05, "loss": 2.2699, "step": 1179 }, { "epoch": 1.3965670316661734, "grad_norm": 4.6875, "learning_rate": 1.80979420318872e-05, "loss": 2.9844, "step": 1180 }, { "epoch": 1.3977508138502515, "grad_norm": 4.4375, "learning_rate": 1.8093086138146723e-05, "loss": 2.2282, "step": 1181 }, { "epoch": 1.3989345960343296, "grad_norm": 4.09375, "learning_rate": 1.8088224707030903e-05, "loss": 2.6382, "step": 1182 }, { "epoch": 1.4001183782184077, "grad_norm": 4.03125, "learning_rate": 1.8083357741865976e-05, "loss": 2.2456, "step": 1183 }, { "epoch": 1.401302160402486, "grad_norm": 4.78125, "learning_rate": 1.8078485245981982e-05, "loss": 2.6471, "step": 1184 }, { "epoch": 1.4024859425865641, "grad_norm": 4.5625, "learning_rate": 1.807360722271273e-05, "loss": 2.6463, "step": 1185 }, { "epoch": 1.4036697247706422, "grad_norm": 4.71875, "learning_rate": 1.806872367539581e-05, "loss": 2.6843, "step": 1186 }, { "epoch": 1.4048535069547203, "grad_norm": 4.65625, "learning_rate": 1.8063834607372603e-05, "loss": 2.6204, "step": 1187 }, { "epoch": 1.4060372891387984, "grad_norm": 4.78125, "learning_rate": 1.8058940021988257e-05, "loss": 2.4325, "step": 1188 }, { "epoch": 1.4072210713228765, "grad_norm": 4.53125, "learning_rate": 1.8054039922591705e-05, "loss": 2.4552, "step": 1189 }, { "epoch": 1.4084048535069547, "grad_norm": 4.46875, "learning_rate": 1.804913431253564e-05, "loss": 2.5584, "step": 1190 }, { "epoch": 1.409588635691033, "grad_norm": 4.03125, "learning_rate": 1.804422319517654e-05, "loss": 2.6413, "step": 1191 }, { "epoch": 1.410772417875111, "grad_norm": 5.0, "learning_rate": 1.803930657387464e-05, "loss": 2.6358, "step": 1192 }, { "epoch": 1.4119562000591892, "grad_norm": 4.53125, "learning_rate": 1.803438445199394e-05, "loss": 2.4831, "step": 1193 }, { "epoch": 1.4131399822432673, "grad_norm": 4.5, "learning_rate": 1.8029456832902215e-05, "loss": 2.8009, "step": 1194 }, { "epoch": 1.4143237644273454, "grad_norm": 4.3125, "learning_rate": 1.802452371997099e-05, "loss": 2.2168, "step": 1195 }, { "epoch": 1.4155075466114235, "grad_norm": 5.09375, "learning_rate": 1.8019585116575554e-05, "loss": 2.5859, "step": 1196 }, { "epoch": 1.4166913287955016, "grad_norm": 4.5, "learning_rate": 1.801464102609495e-05, "loss": 2.7876, "step": 1197 }, { "epoch": 1.4178751109795797, "grad_norm": 4.21875, "learning_rate": 1.8009691451911986e-05, "loss": 2.3209, "step": 1198 }, { "epoch": 1.4190588931636579, "grad_norm": 4.5625, "learning_rate": 1.8004736397413204e-05, "loss": 2.6372, "step": 1199 }, { "epoch": 1.420242675347736, "grad_norm": 4.0625, "learning_rate": 1.7999775865988903e-05, "loss": 2.4994, "step": 1200 }, { "epoch": 1.421426457531814, "grad_norm": 4.9375, "learning_rate": 1.799480986103314e-05, "loss": 2.5162, "step": 1201 }, { "epoch": 1.4226102397158922, "grad_norm": 4.71875, "learning_rate": 1.7989838385943697e-05, "loss": 2.6173, "step": 1202 }, { "epoch": 1.4237940218999703, "grad_norm": 4.4375, "learning_rate": 1.7984861444122127e-05, "loss": 2.6266, "step": 1203 }, { "epoch": 1.4249778040840486, "grad_norm": 4.5, "learning_rate": 1.7979879038973686e-05, "loss": 2.6698, "step": 1204 }, { "epoch": 1.4261615862681267, "grad_norm": 4.53125, "learning_rate": 1.7974891173907406e-05, "loss": 2.2274, "step": 1205 }, { "epoch": 1.4273453684522048, "grad_norm": 4.625, "learning_rate": 1.7969897852336028e-05, "loss": 2.4529, "step": 1206 }, { "epoch": 1.428529150636283, "grad_norm": 4.59375, "learning_rate": 1.7964899077676034e-05, "loss": 2.7299, "step": 1207 }, { "epoch": 1.429712932820361, "grad_norm": 4.375, "learning_rate": 1.795989485334764e-05, "loss": 2.399, "step": 1208 }, { "epoch": 1.4308967150044392, "grad_norm": 4.46875, "learning_rate": 1.7954885182774798e-05, "loss": 2.79, "step": 1209 }, { "epoch": 1.4320804971885173, "grad_norm": 4.25, "learning_rate": 1.794987006938517e-05, "loss": 2.6937, "step": 1210 }, { "epoch": 1.4332642793725954, "grad_norm": 5.15625, "learning_rate": 1.7944849516610147e-05, "loss": 2.7369, "step": 1211 }, { "epoch": 1.4344480615566737, "grad_norm": 4.71875, "learning_rate": 1.7939823527884845e-05, "loss": 3.0087, "step": 1212 }, { "epoch": 1.4356318437407518, "grad_norm": 4.4375, "learning_rate": 1.793479210664811e-05, "loss": 2.6007, "step": 1213 }, { "epoch": 1.43681562592483, "grad_norm": 5.03125, "learning_rate": 1.792975525634248e-05, "loss": 2.2619, "step": 1214 }, { "epoch": 1.437999408108908, "grad_norm": 4.65625, "learning_rate": 1.7924712980414227e-05, "loss": 2.4106, "step": 1215 }, { "epoch": 1.4391831902929861, "grad_norm": 4.34375, "learning_rate": 1.7919665282313335e-05, "loss": 2.4133, "step": 1216 }, { "epoch": 1.4403669724770642, "grad_norm": 4.59375, "learning_rate": 1.7914612165493483e-05, "loss": 2.4946, "step": 1217 }, { "epoch": 1.4415507546611424, "grad_norm": 4.03125, "learning_rate": 1.7909553633412068e-05, "loss": 2.8181, "step": 1218 }, { "epoch": 1.4427345368452205, "grad_norm": 4.5, "learning_rate": 1.79044896895302e-05, "loss": 2.3086, "step": 1219 }, { "epoch": 1.4439183190292986, "grad_norm": 4.8125, "learning_rate": 1.7899420337312675e-05, "loss": 2.696, "step": 1220 }, { "epoch": 1.4451021012133767, "grad_norm": 3.96875, "learning_rate": 1.7894345580228e-05, "loss": 2.6789, "step": 1221 }, { "epoch": 1.4462858833974548, "grad_norm": 4.15625, "learning_rate": 1.788926542174838e-05, "loss": 2.6589, "step": 1222 }, { "epoch": 1.447469665581533, "grad_norm": 4.71875, "learning_rate": 1.7884179865349713e-05, "loss": 2.1945, "step": 1223 }, { "epoch": 1.448653447765611, "grad_norm": 4.375, "learning_rate": 1.7879088914511583e-05, "loss": 2.7278, "step": 1224 }, { "epoch": 1.4498372299496893, "grad_norm": 4.75, "learning_rate": 1.7873992572717282e-05, "loss": 2.5408, "step": 1225 }, { "epoch": 1.4510210121337674, "grad_norm": 4.84375, "learning_rate": 1.7868890843453776e-05, "loss": 2.6264, "step": 1226 }, { "epoch": 1.4522047943178455, "grad_norm": 5.03125, "learning_rate": 1.7863783730211722e-05, "loss": 2.4778, "step": 1227 }, { "epoch": 1.4533885765019237, "grad_norm": 4.625, "learning_rate": 1.7858671236485466e-05, "loss": 2.7322, "step": 1228 }, { "epoch": 1.4545723586860018, "grad_norm": 4.625, "learning_rate": 1.7853553365773027e-05, "loss": 2.7738, "step": 1229 }, { "epoch": 1.4557561408700799, "grad_norm": 4.0625, "learning_rate": 1.7848430121576105e-05, "loss": 2.3925, "step": 1230 }, { "epoch": 1.456939923054158, "grad_norm": 4.75, "learning_rate": 1.784330150740008e-05, "loss": 2.8966, "step": 1231 }, { "epoch": 1.458123705238236, "grad_norm": 4.65625, "learning_rate": 1.7838167526754002e-05, "loss": 2.507, "step": 1232 }, { "epoch": 1.4593074874223144, "grad_norm": 4.21875, "learning_rate": 1.7833028183150597e-05, "loss": 2.7007, "step": 1233 }, { "epoch": 1.4604912696063925, "grad_norm": 4.25, "learning_rate": 1.782788348010626e-05, "loss": 2.3969, "step": 1234 }, { "epoch": 1.4616750517904706, "grad_norm": 4.78125, "learning_rate": 1.7822733421141046e-05, "loss": 2.353, "step": 1235 }, { "epoch": 1.4628588339745487, "grad_norm": 4.96875, "learning_rate": 1.7817578009778687e-05, "loss": 2.9249, "step": 1236 }, { "epoch": 1.4640426161586269, "grad_norm": 4.3125, "learning_rate": 1.7812417249546564e-05, "loss": 2.6607, "step": 1237 }, { "epoch": 1.465226398342705, "grad_norm": 4.34375, "learning_rate": 1.7807251143975727e-05, "loss": 2.2167, "step": 1238 }, { "epoch": 1.466410180526783, "grad_norm": 4.8125, "learning_rate": 1.780207969660088e-05, "loss": 2.743, "step": 1239 }, { "epoch": 1.4675939627108612, "grad_norm": 4.1875, "learning_rate": 1.779690291096038e-05, "loss": 2.5529, "step": 1240 }, { "epoch": 1.4687777448949393, "grad_norm": 4.3125, "learning_rate": 1.7791720790596242e-05, "loss": 2.5904, "step": 1241 }, { "epoch": 1.4699615270790174, "grad_norm": 4.125, "learning_rate": 1.7786533339054126e-05, "loss": 2.3343, "step": 1242 }, { "epoch": 1.4711453092630955, "grad_norm": 4.34375, "learning_rate": 1.778134055988334e-05, "loss": 2.303, "step": 1243 }, { "epoch": 1.4723290914471736, "grad_norm": 4.5, "learning_rate": 1.7776142456636843e-05, "loss": 2.4325, "step": 1244 }, { "epoch": 1.4735128736312517, "grad_norm": 4.3125, "learning_rate": 1.7770939032871227e-05, "loss": 2.3065, "step": 1245 }, { "epoch": 1.47469665581533, "grad_norm": 4.65625, "learning_rate": 1.776573029214673e-05, "loss": 2.3058, "step": 1246 }, { "epoch": 1.4758804379994082, "grad_norm": 4.5, "learning_rate": 1.7760516238027233e-05, "loss": 2.6124, "step": 1247 }, { "epoch": 1.4770642201834863, "grad_norm": 4.03125, "learning_rate": 1.7755296874080242e-05, "loss": 2.2885, "step": 1248 }, { "epoch": 1.4782480023675644, "grad_norm": 4.6875, "learning_rate": 1.7750072203876905e-05, "loss": 2.3909, "step": 1249 }, { "epoch": 1.4794317845516425, "grad_norm": 4.9375, "learning_rate": 1.774484223099199e-05, "loss": 2.234, "step": 1250 }, { "epoch": 1.4806155667357206, "grad_norm": 4.75, "learning_rate": 1.773960695900391e-05, "loss": 2.6671, "step": 1251 }, { "epoch": 1.4817993489197987, "grad_norm": 4.90625, "learning_rate": 1.7734366391494686e-05, "loss": 2.2606, "step": 1252 }, { "epoch": 1.4829831311038768, "grad_norm": 4.28125, "learning_rate": 1.7729120532049967e-05, "loss": 2.8626, "step": 1253 }, { "epoch": 1.4841669132879551, "grad_norm": 4.5625, "learning_rate": 1.7723869384259038e-05, "loss": 2.6066, "step": 1254 }, { "epoch": 1.4853506954720332, "grad_norm": 4.25, "learning_rate": 1.771861295171478e-05, "loss": 2.5285, "step": 1255 }, { "epoch": 1.4865344776561114, "grad_norm": 4.59375, "learning_rate": 1.77133512380137e-05, "loss": 2.697, "step": 1256 }, { "epoch": 1.4877182598401895, "grad_norm": 3.828125, "learning_rate": 1.7708084246755924e-05, "loss": 2.1182, "step": 1257 }, { "epoch": 1.4889020420242676, "grad_norm": 4.28125, "learning_rate": 1.7702811981545177e-05, "loss": 2.6596, "step": 1258 }, { "epoch": 1.4900858242083457, "grad_norm": 4.46875, "learning_rate": 1.7697534445988804e-05, "loss": 2.3811, "step": 1259 }, { "epoch": 1.4912696063924238, "grad_norm": 4.78125, "learning_rate": 1.7692251643697748e-05, "loss": 2.7523, "step": 1260 }, { "epoch": 1.492453388576502, "grad_norm": 4.25, "learning_rate": 1.768696357828656e-05, "loss": 2.7208, "step": 1261 }, { "epoch": 1.49363717076058, "grad_norm": 4.34375, "learning_rate": 1.7681670253373387e-05, "loss": 2.4957, "step": 1262 }, { "epoch": 1.494820952944658, "grad_norm": 4.875, "learning_rate": 1.7676371672579983e-05, "loss": 2.1128, "step": 1263 }, { "epoch": 1.4960047351287362, "grad_norm": 4.34375, "learning_rate": 1.767106783953169e-05, "loss": 2.1075, "step": 1264 }, { "epoch": 1.4971885173128143, "grad_norm": 3.9375, "learning_rate": 1.7665758757857445e-05, "loss": 2.5593, "step": 1265 }, { "epoch": 1.4983722994968924, "grad_norm": 4.25, "learning_rate": 1.766044443118978e-05, "loss": 2.3336, "step": 1266 }, { "epoch": 1.4995560816809708, "grad_norm": 4.40625, "learning_rate": 1.7655124863164817e-05, "loss": 2.3089, "step": 1267 }, { "epoch": 1.5007398638650489, "grad_norm": 4.84375, "learning_rate": 1.7649800057422256e-05, "loss": 2.7454, "step": 1268 }, { "epoch": 1.501923646049127, "grad_norm": 4.53125, "learning_rate": 1.764447001760539e-05, "loss": 2.5279, "step": 1269 }, { "epoch": 1.503107428233205, "grad_norm": 4.1875, "learning_rate": 1.7639134747361084e-05, "loss": 2.5411, "step": 1270 }, { "epoch": 1.5042912104172832, "grad_norm": 4.5625, "learning_rate": 1.763379425033979e-05, "loss": 2.7088, "step": 1271 }, { "epoch": 1.5054749926013613, "grad_norm": 5.21875, "learning_rate": 1.7628448530195527e-05, "loss": 2.6978, "step": 1272 }, { "epoch": 1.5066587747854396, "grad_norm": 4.84375, "learning_rate": 1.7623097590585902e-05, "loss": 2.1795, "step": 1273 }, { "epoch": 1.5078425569695177, "grad_norm": 4.3125, "learning_rate": 1.761774143517208e-05, "loss": 2.9879, "step": 1274 }, { "epoch": 1.5090263391535959, "grad_norm": 4.3125, "learning_rate": 1.7612380067618796e-05, "loss": 2.6055, "step": 1275 }, { "epoch": 1.510210121337674, "grad_norm": 5.03125, "learning_rate": 1.7607013491594363e-05, "loss": 2.4955, "step": 1276 }, { "epoch": 1.511393903521752, "grad_norm": 4.75, "learning_rate": 1.760164171077064e-05, "loss": 2.7496, "step": 1277 }, { "epoch": 1.5125776857058302, "grad_norm": 4.8125, "learning_rate": 1.7596264728823063e-05, "loss": 2.5829, "step": 1278 }, { "epoch": 1.5137614678899083, "grad_norm": 5.5, "learning_rate": 1.7590882549430617e-05, "loss": 2.5579, "step": 1279 }, { "epoch": 1.5149452500739864, "grad_norm": 5.03125, "learning_rate": 1.758549517627585e-05, "loss": 2.9256, "step": 1280 }, { "epoch": 1.5161290322580645, "grad_norm": 4.15625, "learning_rate": 1.7580102613044853e-05, "loss": 2.251, "step": 1281 }, { "epoch": 1.5173128144421426, "grad_norm": 4.375, "learning_rate": 1.7574704863427277e-05, "loss": 2.7189, "step": 1282 }, { "epoch": 1.5184965966262207, "grad_norm": 4.34375, "learning_rate": 1.7569301931116326e-05, "loss": 2.2822, "step": 1283 }, { "epoch": 1.5196803788102988, "grad_norm": 4.84375, "learning_rate": 1.7563893819808736e-05, "loss": 2.7641, "step": 1284 }, { "epoch": 1.520864160994377, "grad_norm": 4.75, "learning_rate": 1.75584805332048e-05, "loss": 2.7623, "step": 1285 }, { "epoch": 1.522047943178455, "grad_norm": 4.1875, "learning_rate": 1.755306207500834e-05, "loss": 2.7521, "step": 1286 }, { "epoch": 1.5232317253625332, "grad_norm": 4.5, "learning_rate": 1.754763844892673e-05, "loss": 2.0891, "step": 1287 }, { "epoch": 1.5244155075466113, "grad_norm": 4.3125, "learning_rate": 1.7542209658670862e-05, "loss": 3.1161, "step": 1288 }, { "epoch": 1.5255992897306896, "grad_norm": 5.3125, "learning_rate": 1.7536775707955182e-05, "loss": 3.0585, "step": 1289 }, { "epoch": 1.5267830719147677, "grad_norm": 4.53125, "learning_rate": 1.753133660049765e-05, "loss": 2.476, "step": 1290 }, { "epoch": 1.5279668540988458, "grad_norm": 4.21875, "learning_rate": 1.7525892340019766e-05, "loss": 2.5468, "step": 1291 }, { "epoch": 1.529150636282924, "grad_norm": 4.8125, "learning_rate": 1.7520442930246546e-05, "loss": 2.4685, "step": 1292 }, { "epoch": 1.530334418467002, "grad_norm": 4.28125, "learning_rate": 1.7514988374906535e-05, "loss": 2.9253, "step": 1293 }, { "epoch": 1.5315182006510804, "grad_norm": 4.21875, "learning_rate": 1.7509528677731806e-05, "loss": 2.3209, "step": 1294 }, { "epoch": 1.5327019828351585, "grad_norm": 4.6875, "learning_rate": 1.750406384245793e-05, "loss": 2.2165, "step": 1295 }, { "epoch": 1.5338857650192366, "grad_norm": 4.75, "learning_rate": 1.749859387282401e-05, "loss": 2.74, "step": 1296 }, { "epoch": 1.5350695472033147, "grad_norm": 4.65625, "learning_rate": 1.7493118772572655e-05, "loss": 2.2641, "step": 1297 }, { "epoch": 1.5362533293873928, "grad_norm": 4.34375, "learning_rate": 1.7487638545449993e-05, "loss": 2.0723, "step": 1298 }, { "epoch": 1.537437111571471, "grad_norm": 4.3125, "learning_rate": 1.7482153195205646e-05, "loss": 2.5558, "step": 1299 }, { "epoch": 1.538620893755549, "grad_norm": 4.8125, "learning_rate": 1.747666272559275e-05, "loss": 2.548, "step": 1300 }, { "epoch": 1.539804675939627, "grad_norm": 4.34375, "learning_rate": 1.7471167140367948e-05, "loss": 2.6114, "step": 1301 }, { "epoch": 1.5409884581237052, "grad_norm": 4.40625, "learning_rate": 1.7465666443291374e-05, "loss": 2.4576, "step": 1302 }, { "epoch": 1.5421722403077833, "grad_norm": 4.59375, "learning_rate": 1.7460160638126663e-05, "loss": 2.5051, "step": 1303 }, { "epoch": 1.5433560224918614, "grad_norm": 4.25, "learning_rate": 1.7454649728640944e-05, "loss": 2.3127, "step": 1304 }, { "epoch": 1.5445398046759395, "grad_norm": 4.21875, "learning_rate": 1.7449133718604845e-05, "loss": 2.6161, "step": 1305 }, { "epoch": 1.5457235868600177, "grad_norm": 5.21875, "learning_rate": 1.7443612611792473e-05, "loss": 2.9661, "step": 1306 }, { "epoch": 1.5469073690440958, "grad_norm": 4.40625, "learning_rate": 1.743808641198143e-05, "loss": 2.5946, "step": 1307 }, { "epoch": 1.5480911512281739, "grad_norm": 4.5625, "learning_rate": 1.74325551229528e-05, "loss": 2.8905, "step": 1308 }, { "epoch": 1.5492749334122522, "grad_norm": 5.125, "learning_rate": 1.742701874849115e-05, "loss": 2.8611, "step": 1309 }, { "epoch": 1.5504587155963303, "grad_norm": 4.59375, "learning_rate": 1.7421477292384524e-05, "loss": 2.3031, "step": 1310 }, { "epoch": 1.5516424977804084, "grad_norm": 3.953125, "learning_rate": 1.741593075842445e-05, "loss": 1.9637, "step": 1311 }, { "epoch": 1.5528262799644865, "grad_norm": 4.5, "learning_rate": 1.7410379150405924e-05, "loss": 2.6594, "step": 1312 }, { "epoch": 1.5540100621485646, "grad_norm": 4.4375, "learning_rate": 1.7404822472127406e-05, "loss": 2.7401, "step": 1313 }, { "epoch": 1.5551938443326427, "grad_norm": 4.40625, "learning_rate": 1.7399260727390846e-05, "loss": 2.5288, "step": 1314 }, { "epoch": 1.556377626516721, "grad_norm": 4.1875, "learning_rate": 1.7393693920001647e-05, "loss": 2.5483, "step": 1315 }, { "epoch": 1.5575614087007992, "grad_norm": 4.5, "learning_rate": 1.7388122053768673e-05, "loss": 2.8354, "step": 1316 }, { "epoch": 1.5587451908848773, "grad_norm": 4.375, "learning_rate": 1.7382545132504255e-05, "loss": 2.1904, "step": 1317 }, { "epoch": 1.5599289730689554, "grad_norm": 4.0, "learning_rate": 1.7376963160024184e-05, "loss": 2.5104, "step": 1318 }, { "epoch": 1.5611127552530335, "grad_norm": 4.8125, "learning_rate": 1.7371376140147705e-05, "loss": 2.6556, "step": 1319 }, { "epoch": 1.5622965374371116, "grad_norm": 4.46875, "learning_rate": 1.736578407669751e-05, "loss": 2.4073, "step": 1320 }, { "epoch": 1.5634803196211897, "grad_norm": 4.71875, "learning_rate": 1.7360186973499752e-05, "loss": 2.5832, "step": 1321 }, { "epoch": 1.5646641018052678, "grad_norm": 4.875, "learning_rate": 1.7354584834384036e-05, "loss": 2.5554, "step": 1322 }, { "epoch": 1.565847883989346, "grad_norm": 4.625, "learning_rate": 1.7348977663183392e-05, "loss": 2.6885, "step": 1323 }, { "epoch": 1.567031666173424, "grad_norm": 4.75, "learning_rate": 1.7343365463734314e-05, "loss": 2.6513, "step": 1324 }, { "epoch": 1.5682154483575022, "grad_norm": 4.78125, "learning_rate": 1.733774823987673e-05, "loss": 2.6054, "step": 1325 }, { "epoch": 1.5693992305415803, "grad_norm": 4.125, "learning_rate": 1.7332125995453992e-05, "loss": 2.3713, "step": 1326 }, { "epoch": 1.5705830127256584, "grad_norm": 4.4375, "learning_rate": 1.732649873431291e-05, "loss": 2.3309, "step": 1327 }, { "epoch": 1.5717667949097365, "grad_norm": 5.03125, "learning_rate": 1.732086646030372e-05, "loss": 2.8258, "step": 1328 }, { "epoch": 1.5729505770938146, "grad_norm": 5.40625, "learning_rate": 1.7315229177280075e-05, "loss": 2.4212, "step": 1329 }, { "epoch": 1.574134359277893, "grad_norm": 4.0625, "learning_rate": 1.7309586889099063e-05, "loss": 2.7065, "step": 1330 }, { "epoch": 1.575318141461971, "grad_norm": 4.21875, "learning_rate": 1.73039395996212e-05, "loss": 2.5907, "step": 1331 }, { "epoch": 1.5765019236460491, "grad_norm": 4.75, "learning_rate": 1.7298287312710423e-05, "loss": 2.6052, "step": 1332 }, { "epoch": 1.5776857058301272, "grad_norm": 4.65625, "learning_rate": 1.7292630032234086e-05, "loss": 2.6578, "step": 1333 }, { "epoch": 1.5788694880142053, "grad_norm": 4.34375, "learning_rate": 1.7286967762062958e-05, "loss": 2.7185, "step": 1334 }, { "epoch": 1.5800532701982837, "grad_norm": 3.96875, "learning_rate": 1.728130050607123e-05, "loss": 2.6222, "step": 1335 }, { "epoch": 1.5812370523823618, "grad_norm": 4.40625, "learning_rate": 1.7275628268136487e-05, "loss": 2.7472, "step": 1336 }, { "epoch": 1.58242083456644, "grad_norm": 4.34375, "learning_rate": 1.7269951052139745e-05, "loss": 2.1986, "step": 1337 }, { "epoch": 1.583604616750518, "grad_norm": 4.25, "learning_rate": 1.7264268861965416e-05, "loss": 2.4003, "step": 1338 }, { "epoch": 1.584788398934596, "grad_norm": 4.25, "learning_rate": 1.7258581701501305e-05, "loss": 2.5563, "step": 1339 }, { "epoch": 1.5859721811186742, "grad_norm": 3.96875, "learning_rate": 1.725288957463864e-05, "loss": 2.3025, "step": 1340 }, { "epoch": 1.5871559633027523, "grad_norm": 4.5625, "learning_rate": 1.724719248527202e-05, "loss": 2.3185, "step": 1341 }, { "epoch": 1.5883397454868304, "grad_norm": 4.34375, "learning_rate": 1.724149043729947e-05, "loss": 2.711, "step": 1342 }, { "epoch": 1.5895235276709085, "grad_norm": 4.5625, "learning_rate": 1.723578343462238e-05, "loss": 2.7395, "step": 1343 }, { "epoch": 1.5907073098549867, "grad_norm": 4.8125, "learning_rate": 1.7230071481145546e-05, "loss": 2.6068, "step": 1344 }, { "epoch": 1.5918910920390648, "grad_norm": 4.4375, "learning_rate": 1.7224354580777144e-05, "loss": 2.5887, "step": 1345 }, { "epoch": 1.5930748742231429, "grad_norm": 4.5625, "learning_rate": 1.7218632737428743e-05, "loss": 2.6219, "step": 1346 }, { "epoch": 1.594258656407221, "grad_norm": 4.625, "learning_rate": 1.7212905955015285e-05, "loss": 2.5291, "step": 1347 }, { "epoch": 1.595442438591299, "grad_norm": 4.4375, "learning_rate": 1.7207174237455095e-05, "loss": 2.7055, "step": 1348 }, { "epoch": 1.5966262207753772, "grad_norm": 4.625, "learning_rate": 1.720143758866988e-05, "loss": 2.4857, "step": 1349 }, { "epoch": 1.5978100029594553, "grad_norm": 4.0625, "learning_rate": 1.719569601258471e-05, "loss": 2.4049, "step": 1350 }, { "epoch": 1.5989937851435336, "grad_norm": 4.75, "learning_rate": 1.7189949513128033e-05, "loss": 2.7603, "step": 1351 }, { "epoch": 1.6001775673276117, "grad_norm": 5.09375, "learning_rate": 1.718419809423167e-05, "loss": 2.7363, "step": 1352 }, { "epoch": 1.6013613495116898, "grad_norm": 4.28125, "learning_rate": 1.7178441759830797e-05, "loss": 2.6146, "step": 1353 }, { "epoch": 1.602545131695768, "grad_norm": 4.25, "learning_rate": 1.717268051386396e-05, "loss": 2.2028, "step": 1354 }, { "epoch": 1.603728913879846, "grad_norm": 4.375, "learning_rate": 1.716691436027307e-05, "loss": 2.4238, "step": 1355 }, { "epoch": 1.6049126960639244, "grad_norm": 4.15625, "learning_rate": 1.7161143303003382e-05, "loss": 2.3352, "step": 1356 }, { "epoch": 1.6060964782480025, "grad_norm": 4.65625, "learning_rate": 1.715536734600352e-05, "loss": 3.0964, "step": 1357 }, { "epoch": 1.6072802604320806, "grad_norm": 4.5625, "learning_rate": 1.7149586493225453e-05, "loss": 2.8319, "step": 1358 }, { "epoch": 1.6084640426161587, "grad_norm": 4.125, "learning_rate": 1.7143800748624507e-05, "loss": 2.3561, "step": 1359 }, { "epoch": 1.6096478248002368, "grad_norm": 4.96875, "learning_rate": 1.7138010116159342e-05, "loss": 2.8158, "step": 1360 }, { "epoch": 1.610831606984315, "grad_norm": 4.625, "learning_rate": 1.7132214599791975e-05, "loss": 2.4812, "step": 1361 }, { "epoch": 1.612015389168393, "grad_norm": 4.53125, "learning_rate": 1.7126414203487757e-05, "loss": 2.5256, "step": 1362 }, { "epoch": 1.6131991713524712, "grad_norm": 4.5625, "learning_rate": 1.712060893121538e-05, "loss": 2.9876, "step": 1363 }, { "epoch": 1.6143829535365493, "grad_norm": 4.25, "learning_rate": 1.7114798786946875e-05, "loss": 2.6882, "step": 1364 }, { "epoch": 1.6155667357206274, "grad_norm": 4.15625, "learning_rate": 1.710898377465761e-05, "loss": 2.1729, "step": 1365 }, { "epoch": 1.6167505179047055, "grad_norm": 5.1875, "learning_rate": 1.7103163898326265e-05, "loss": 3.0367, "step": 1366 }, { "epoch": 1.6179343000887836, "grad_norm": 4.40625, "learning_rate": 1.709733916193487e-05, "loss": 2.5724, "step": 1367 }, { "epoch": 1.6191180822728617, "grad_norm": 4.1875, "learning_rate": 1.709150956946877e-05, "loss": 2.6671, "step": 1368 }, { "epoch": 1.6203018644569398, "grad_norm": 4.40625, "learning_rate": 1.7085675124916638e-05, "loss": 2.4107, "step": 1369 }, { "epoch": 1.621485646641018, "grad_norm": 4.09375, "learning_rate": 1.7079835832270455e-05, "loss": 2.4961, "step": 1370 }, { "epoch": 1.622669428825096, "grad_norm": 4.6875, "learning_rate": 1.7073991695525528e-05, "loss": 2.6022, "step": 1371 }, { "epoch": 1.6238532110091743, "grad_norm": 4.375, "learning_rate": 1.706814271868048e-05, "loss": 2.2053, "step": 1372 }, { "epoch": 1.6250369931932525, "grad_norm": 4.5625, "learning_rate": 1.7062288905737245e-05, "loss": 2.566, "step": 1373 }, { "epoch": 1.6262207753773306, "grad_norm": 4.21875, "learning_rate": 1.705643026070106e-05, "loss": 2.535, "step": 1374 }, { "epoch": 1.6274045575614087, "grad_norm": 4.84375, "learning_rate": 1.7050566787580477e-05, "loss": 2.5855, "step": 1375 }, { "epoch": 1.6285883397454868, "grad_norm": 4.21875, "learning_rate": 1.704469849038734e-05, "loss": 2.3321, "step": 1376 }, { "epoch": 1.629772121929565, "grad_norm": 4.8125, "learning_rate": 1.7038825373136806e-05, "loss": 2.4789, "step": 1377 }, { "epoch": 1.6309559041136432, "grad_norm": 4.25, "learning_rate": 1.7032947439847316e-05, "loss": 2.6342, "step": 1378 }, { "epoch": 1.6321396862977213, "grad_norm": 4.125, "learning_rate": 1.7027064694540623e-05, "loss": 2.0086, "step": 1379 }, { "epoch": 1.6333234684817994, "grad_norm": 4.1875, "learning_rate": 1.7021177141241757e-05, "loss": 2.6677, "step": 1380 }, { "epoch": 1.6345072506658775, "grad_norm": 4.15625, "learning_rate": 1.701528478397905e-05, "loss": 2.4837, "step": 1381 }, { "epoch": 1.6356910328499557, "grad_norm": 4.0, "learning_rate": 1.7009387626784118e-05, "loss": 2.6369, "step": 1382 }, { "epoch": 1.6368748150340338, "grad_norm": 4.78125, "learning_rate": 1.7003485673691845e-05, "loss": 2.6081, "step": 1383 }, { "epoch": 1.6380585972181119, "grad_norm": 4.65625, "learning_rate": 1.6997578928740423e-05, "loss": 2.6229, "step": 1384 }, { "epoch": 1.63924237940219, "grad_norm": 4.53125, "learning_rate": 1.6991667395971306e-05, "loss": 2.5262, "step": 1385 }, { "epoch": 1.640426161586268, "grad_norm": 4.5625, "learning_rate": 1.6985751079429226e-05, "loss": 2.6387, "step": 1386 }, { "epoch": 1.6416099437703462, "grad_norm": 5.09375, "learning_rate": 1.697982998316219e-05, "loss": 2.3306, "step": 1387 }, { "epoch": 1.6427937259544243, "grad_norm": 4.84375, "learning_rate": 1.697390411122147e-05, "loss": 2.626, "step": 1388 }, { "epoch": 1.6439775081385024, "grad_norm": 4.71875, "learning_rate": 1.696797346766162e-05, "loss": 2.6455, "step": 1389 }, { "epoch": 1.6451612903225805, "grad_norm": 4.03125, "learning_rate": 1.696203805654044e-05, "loss": 2.4825, "step": 1390 }, { "epoch": 1.6463450725066586, "grad_norm": 4.875, "learning_rate": 1.6956097881919006e-05, "loss": 2.5461, "step": 1391 }, { "epoch": 1.647528854690737, "grad_norm": 4.6875, "learning_rate": 1.695015294786165e-05, "loss": 2.5078, "step": 1392 }, { "epoch": 1.648712636874815, "grad_norm": 4.59375, "learning_rate": 1.6944203258435957e-05, "loss": 2.8854, "step": 1393 }, { "epoch": 1.6498964190588932, "grad_norm": 4.75, "learning_rate": 1.6938248817712767e-05, "loss": 2.7106, "step": 1394 }, { "epoch": 1.6510802012429713, "grad_norm": 4.46875, "learning_rate": 1.693228962976617e-05, "loss": 2.6005, "step": 1395 }, { "epoch": 1.6522639834270494, "grad_norm": 4.40625, "learning_rate": 1.6926325698673513e-05, "loss": 2.5247, "step": 1396 }, { "epoch": 1.6534477656111277, "grad_norm": 3.984375, "learning_rate": 1.6920357028515368e-05, "loss": 2.2898, "step": 1397 }, { "epoch": 1.6546315477952058, "grad_norm": 4.5625, "learning_rate": 1.6914383623375577e-05, "loss": 2.3182, "step": 1398 }, { "epoch": 1.655815329979284, "grad_norm": 4.25, "learning_rate": 1.6908405487341195e-05, "loss": 2.8461, "step": 1399 }, { "epoch": 1.656999112163362, "grad_norm": 4.53125, "learning_rate": 1.690242262450253e-05, "loss": 2.6245, "step": 1400 }, { "epoch": 1.6581828943474402, "grad_norm": 4.4375, "learning_rate": 1.6896435038953126e-05, "loss": 2.3649, "step": 1401 }, { "epoch": 1.6593666765315183, "grad_norm": 4.0625, "learning_rate": 1.6890442734789743e-05, "loss": 2.3616, "step": 1402 }, { "epoch": 1.6605504587155964, "grad_norm": 4.875, "learning_rate": 1.6884445716112388e-05, "loss": 2.5875, "step": 1403 }, { "epoch": 1.6617342408996745, "grad_norm": 4.6875, "learning_rate": 1.6878443987024276e-05, "loss": 2.2728, "step": 1404 }, { "epoch": 1.6629180230837526, "grad_norm": 5.15625, "learning_rate": 1.6872437551631863e-05, "loss": 2.5128, "step": 1405 }, { "epoch": 1.6641018052678307, "grad_norm": 4.5, "learning_rate": 1.686642641404481e-05, "loss": 2.9068, "step": 1406 }, { "epoch": 1.6652855874519088, "grad_norm": 4.28125, "learning_rate": 1.6860410578375997e-05, "loss": 2.6206, "step": 1407 }, { "epoch": 1.666469369635987, "grad_norm": 4.28125, "learning_rate": 1.685439004874153e-05, "loss": 2.6417, "step": 1408 }, { "epoch": 1.667653151820065, "grad_norm": 4.84375, "learning_rate": 1.684836482926072e-05, "loss": 2.7346, "step": 1409 }, { "epoch": 1.6688369340041431, "grad_norm": 4.21875, "learning_rate": 1.6842334924056082e-05, "loss": 2.759, "step": 1410 }, { "epoch": 1.6700207161882212, "grad_norm": 4.34375, "learning_rate": 1.6836300337253337e-05, "loss": 2.7647, "step": 1411 }, { "epoch": 1.6712044983722993, "grad_norm": 4.84375, "learning_rate": 1.6830261072981423e-05, "loss": 2.4305, "step": 1412 }, { "epoch": 1.6723882805563777, "grad_norm": 4.78125, "learning_rate": 1.682421713537246e-05, "loss": 2.5846, "step": 1413 }, { "epoch": 1.6735720627404558, "grad_norm": 4.0, "learning_rate": 1.6818168528561783e-05, "loss": 2.393, "step": 1414 }, { "epoch": 1.6747558449245339, "grad_norm": 4.40625, "learning_rate": 1.6812115256687906e-05, "loss": 2.5185, "step": 1415 }, { "epoch": 1.675939627108612, "grad_norm": 4.1875, "learning_rate": 1.6806057323892542e-05, "loss": 2.1699, "step": 1416 }, { "epoch": 1.67712340929269, "grad_norm": 4.75, "learning_rate": 1.6799994734320596e-05, "loss": 3.1158, "step": 1417 }, { "epoch": 1.6783071914767684, "grad_norm": 4.53125, "learning_rate": 1.679392749212015e-05, "loss": 2.6015, "step": 1418 }, { "epoch": 1.6794909736608465, "grad_norm": 4.3125, "learning_rate": 1.678785560144249e-05, "loss": 2.4406, "step": 1419 }, { "epoch": 1.6806747558449247, "grad_norm": 4.15625, "learning_rate": 1.678177906644205e-05, "loss": 2.3714, "step": 1420 }, { "epoch": 1.6818585380290028, "grad_norm": 4.84375, "learning_rate": 1.677569789127647e-05, "loss": 2.9597, "step": 1421 }, { "epoch": 1.6830423202130809, "grad_norm": 4.6875, "learning_rate": 1.6769612080106554e-05, "loss": 2.8087, "step": 1422 }, { "epoch": 1.684226102397159, "grad_norm": 4.34375, "learning_rate": 1.6763521637096272e-05, "loss": 2.7023, "step": 1423 }, { "epoch": 1.685409884581237, "grad_norm": 4.40625, "learning_rate": 1.675742656641278e-05, "loss": 2.2902, "step": 1424 }, { "epoch": 1.6865936667653152, "grad_norm": 4.65625, "learning_rate": 1.675132687222638e-05, "loss": 2.454, "step": 1425 }, { "epoch": 1.6877774489493933, "grad_norm": 5.15625, "learning_rate": 1.6745222558710556e-05, "loss": 2.5885, "step": 1426 }, { "epoch": 1.6889612311334714, "grad_norm": 4.875, "learning_rate": 1.6739113630041937e-05, "loss": 2.4226, "step": 1427 }, { "epoch": 1.6901450133175495, "grad_norm": 5.03125, "learning_rate": 1.673300009040032e-05, "loss": 3.0137, "step": 1428 }, { "epoch": 1.6913287955016276, "grad_norm": 4.125, "learning_rate": 1.6726881943968653e-05, "loss": 2.2918, "step": 1429 }, { "epoch": 1.6925125776857057, "grad_norm": 4.875, "learning_rate": 1.6720759194933037e-05, "loss": 2.8138, "step": 1430 }, { "epoch": 1.6936963598697838, "grad_norm": 4.53125, "learning_rate": 1.6714631847482724e-05, "loss": 2.5452, "step": 1431 }, { "epoch": 1.694880142053862, "grad_norm": 4.75, "learning_rate": 1.6708499905810107e-05, "loss": 2.8078, "step": 1432 }, { "epoch": 1.69606392423794, "grad_norm": 5.46875, "learning_rate": 1.6702363374110726e-05, "loss": 2.7123, "step": 1433 }, { "epoch": 1.6972477064220184, "grad_norm": 4.0625, "learning_rate": 1.669622225658326e-05, "loss": 2.5211, "step": 1434 }, { "epoch": 1.6984314886060965, "grad_norm": 4.4375, "learning_rate": 1.669007655742953e-05, "loss": 2.3926, "step": 1435 }, { "epoch": 1.6996152707901746, "grad_norm": 4.625, "learning_rate": 1.6683926280854484e-05, "loss": 2.7259, "step": 1436 }, { "epoch": 1.7007990529742527, "grad_norm": 5.0, "learning_rate": 1.6677771431066213e-05, "loss": 2.5375, "step": 1437 }, { "epoch": 1.7019828351583308, "grad_norm": 4.25, "learning_rate": 1.6671612012275924e-05, "loss": 2.8644, "step": 1438 }, { "epoch": 1.7031666173424092, "grad_norm": 4.6875, "learning_rate": 1.666544802869796e-05, "loss": 2.4893, "step": 1439 }, { "epoch": 1.7043503995264873, "grad_norm": 4.84375, "learning_rate": 1.6659279484549782e-05, "loss": 2.5081, "step": 1440 }, { "epoch": 1.7055341817105654, "grad_norm": 4.3125, "learning_rate": 1.6653106384051977e-05, "loss": 2.3039, "step": 1441 }, { "epoch": 1.7067179638946435, "grad_norm": 4.53125, "learning_rate": 1.664692873142824e-05, "loss": 2.5191, "step": 1442 }, { "epoch": 1.7079017460787216, "grad_norm": 4.21875, "learning_rate": 1.6640746530905395e-05, "loss": 2.3776, "step": 1443 }, { "epoch": 1.7090855282627997, "grad_norm": 4.65625, "learning_rate": 1.663455978671336e-05, "loss": 2.4087, "step": 1444 }, { "epoch": 1.7102693104468778, "grad_norm": 4.96875, "learning_rate": 1.6628368503085174e-05, "loss": 2.4383, "step": 1445 }, { "epoch": 1.711453092630956, "grad_norm": 4.125, "learning_rate": 1.6622172684256984e-05, "loss": 2.6581, "step": 1446 }, { "epoch": 1.712636874815034, "grad_norm": 4.5, "learning_rate": 1.6615972334468023e-05, "loss": 2.4004, "step": 1447 }, { "epoch": 1.7138206569991121, "grad_norm": 4.46875, "learning_rate": 1.660976745796065e-05, "loss": 2.4652, "step": 1448 }, { "epoch": 1.7150044391831902, "grad_norm": 4.21875, "learning_rate": 1.6603558058980296e-05, "loss": 2.8164, "step": 1449 }, { "epoch": 1.7161882213672683, "grad_norm": 5.1875, "learning_rate": 1.659734414177551e-05, "loss": 2.3604, "step": 1450 }, { "epoch": 1.7173720035513464, "grad_norm": 4.28125, "learning_rate": 1.6591125710597908e-05, "loss": 2.5042, "step": 1451 }, { "epoch": 1.7185557857354246, "grad_norm": 4.15625, "learning_rate": 1.6584902769702214e-05, "loss": 2.4158, "step": 1452 }, { "epoch": 1.7197395679195027, "grad_norm": 5.40625, "learning_rate": 1.6578675323346224e-05, "loss": 2.9033, "step": 1453 }, { "epoch": 1.720923350103581, "grad_norm": 5.1875, "learning_rate": 1.6572443375790824e-05, "loss": 2.7059, "step": 1454 }, { "epoch": 1.722107132287659, "grad_norm": 4.96875, "learning_rate": 1.6566206931299987e-05, "loss": 2.5813, "step": 1455 }, { "epoch": 1.7232909144717372, "grad_norm": 3.90625, "learning_rate": 1.655996599414075e-05, "loss": 2.3797, "step": 1456 }, { "epoch": 1.7244746966558153, "grad_norm": 3.953125, "learning_rate": 1.655372056858322e-05, "loss": 2.5937, "step": 1457 }, { "epoch": 1.7256584788398934, "grad_norm": 4.25, "learning_rate": 1.6547470658900592e-05, "loss": 2.5907, "step": 1458 }, { "epoch": 1.7268422610239715, "grad_norm": 3.859375, "learning_rate": 1.654121626936912e-05, "loss": 2.5054, "step": 1459 }, { "epoch": 1.7280260432080499, "grad_norm": 4.5, "learning_rate": 1.653495740426812e-05, "loss": 2.542, "step": 1460 }, { "epoch": 1.729209825392128, "grad_norm": 4.59375, "learning_rate": 1.6528694067879977e-05, "loss": 2.4028, "step": 1461 }, { "epoch": 1.730393607576206, "grad_norm": 4.34375, "learning_rate": 1.652242626449013e-05, "loss": 3.0703, "step": 1462 }, { "epoch": 1.7315773897602842, "grad_norm": 5.0625, "learning_rate": 1.6516153998387077e-05, "loss": 2.7525, "step": 1463 }, { "epoch": 1.7327611719443623, "grad_norm": 4.875, "learning_rate": 1.650987727386237e-05, "loss": 2.6688, "step": 1464 }, { "epoch": 1.7339449541284404, "grad_norm": 4.40625, "learning_rate": 1.650359609521061e-05, "loss": 2.2844, "step": 1465 }, { "epoch": 1.7351287363125185, "grad_norm": 4.59375, "learning_rate": 1.6497310466729448e-05, "loss": 2.6341, "step": 1466 }, { "epoch": 1.7363125184965966, "grad_norm": 4.25, "learning_rate": 1.6491020392719575e-05, "loss": 2.2827, "step": 1467 }, { "epoch": 1.7374963006806747, "grad_norm": 4.3125, "learning_rate": 1.648472587748473e-05, "loss": 2.6333, "step": 1468 }, { "epoch": 1.7386800828647528, "grad_norm": 4.25, "learning_rate": 1.647842692533168e-05, "loss": 2.8514, "step": 1469 }, { "epoch": 1.739863865048831, "grad_norm": 4.6875, "learning_rate": 1.647212354057024e-05, "loss": 2.7596, "step": 1470 }, { "epoch": 1.741047647232909, "grad_norm": 4.46875, "learning_rate": 1.6465815727513253e-05, "loss": 2.7558, "step": 1471 }, { "epoch": 1.7422314294169872, "grad_norm": 5.03125, "learning_rate": 1.645950349047659e-05, "loss": 2.7593, "step": 1472 }, { "epoch": 1.7434152116010653, "grad_norm": 4.5625, "learning_rate": 1.645318683377915e-05, "loss": 2.556, "step": 1473 }, { "epoch": 1.7445989937851434, "grad_norm": 5.0, "learning_rate": 1.644686576174286e-05, "loss": 2.5426, "step": 1474 }, { "epoch": 1.7457827759692217, "grad_norm": 5.9375, "learning_rate": 1.6440540278692656e-05, "loss": 2.6117, "step": 1475 }, { "epoch": 1.7469665581532998, "grad_norm": 4.03125, "learning_rate": 1.643421038895651e-05, "loss": 2.339, "step": 1476 }, { "epoch": 1.748150340337378, "grad_norm": 4.34375, "learning_rate": 1.6427876096865394e-05, "loss": 2.4806, "step": 1477 }, { "epoch": 1.749334122521456, "grad_norm": 4.71875, "learning_rate": 1.6421537406753302e-05, "loss": 2.8306, "step": 1478 }, { "epoch": 1.7505179047055341, "grad_norm": 3.96875, "learning_rate": 1.6415194322957232e-05, "loss": 2.3558, "step": 1479 }, { "epoch": 1.7517016868896125, "grad_norm": 4.28125, "learning_rate": 1.6408846849817182e-05, "loss": 2.2791, "step": 1480 }, { "epoch": 1.7528854690736906, "grad_norm": 4.59375, "learning_rate": 1.6402494991676174e-05, "loss": 2.2555, "step": 1481 }, { "epoch": 1.7540692512577687, "grad_norm": 4.46875, "learning_rate": 1.6396138752880205e-05, "loss": 2.7667, "step": 1482 }, { "epoch": 1.7552530334418468, "grad_norm": 4.78125, "learning_rate": 1.6389778137778284e-05, "loss": 2.4026, "step": 1483 }, { "epoch": 1.756436815625925, "grad_norm": 4.40625, "learning_rate": 1.6383413150722417e-05, "loss": 2.4662, "step": 1484 }, { "epoch": 1.757620597810003, "grad_norm": 4.4375, "learning_rate": 1.6377043796067585e-05, "loss": 2.4811, "step": 1485 }, { "epoch": 1.7588043799940811, "grad_norm": 4.21875, "learning_rate": 1.637067007817178e-05, "loss": 2.5875, "step": 1486 }, { "epoch": 1.7599881621781592, "grad_norm": 5.0, "learning_rate": 1.6364292001395958e-05, "loss": 2.817, "step": 1487 }, { "epoch": 1.7611719443622373, "grad_norm": 4.78125, "learning_rate": 1.635790957010407e-05, "loss": 2.4671, "step": 1488 }, { "epoch": 1.7623557265463154, "grad_norm": 4.59375, "learning_rate": 1.6351522788663044e-05, "loss": 2.3021, "step": 1489 }, { "epoch": 1.7635395087303936, "grad_norm": 4.4375, "learning_rate": 1.6345131661442782e-05, "loss": 2.7499, "step": 1490 }, { "epoch": 1.7647232909144717, "grad_norm": 4.125, "learning_rate": 1.6338736192816163e-05, "loss": 2.6982, "step": 1491 }, { "epoch": 1.7659070730985498, "grad_norm": 4.5, "learning_rate": 1.6332336387159032e-05, "loss": 2.9568, "step": 1492 }, { "epoch": 1.7670908552826279, "grad_norm": 4.28125, "learning_rate": 1.6325932248850206e-05, "loss": 2.7163, "step": 1493 }, { "epoch": 1.768274637466706, "grad_norm": 4.5, "learning_rate": 1.6319523782271462e-05, "loss": 2.6274, "step": 1494 }, { "epoch": 1.769458419650784, "grad_norm": 4.96875, "learning_rate": 1.6313110991807546e-05, "loss": 2.6527, "step": 1495 }, { "epoch": 1.7706422018348624, "grad_norm": 5.0, "learning_rate": 1.6306693881846152e-05, "loss": 2.3608, "step": 1496 }, { "epoch": 1.7718259840189405, "grad_norm": 4.5625, "learning_rate": 1.6300272456777933e-05, "loss": 2.7968, "step": 1497 }, { "epoch": 1.7730097662030186, "grad_norm": 4.875, "learning_rate": 1.6293846720996504e-05, "loss": 2.8405, "step": 1498 }, { "epoch": 1.7741935483870968, "grad_norm": 4.6875, "learning_rate": 1.6287416678898415e-05, "loss": 2.6164, "step": 1499 }, { "epoch": 1.7753773305711749, "grad_norm": 4.625, "learning_rate": 1.628098233488317e-05, "loss": 2.667, "step": 1500 }, { "epoch": 1.7765611127552532, "grad_norm": 4.375, "learning_rate": 1.6274543693353218e-05, "loss": 2.8137, "step": 1501 }, { "epoch": 1.7777448949393313, "grad_norm": 4.46875, "learning_rate": 1.626810075871394e-05, "loss": 2.2615, "step": 1502 }, { "epoch": 1.7789286771234094, "grad_norm": 4.3125, "learning_rate": 1.6261653535373668e-05, "loss": 2.6864, "step": 1503 }, { "epoch": 1.7801124593074875, "grad_norm": 4.5625, "learning_rate": 1.6255202027743658e-05, "loss": 2.8991, "step": 1504 }, { "epoch": 1.7812962414915656, "grad_norm": 4.125, "learning_rate": 1.6248746240238095e-05, "loss": 2.5732, "step": 1505 }, { "epoch": 1.7824800236756437, "grad_norm": 4.8125, "learning_rate": 1.6242286177274103e-05, "loss": 2.7713, "step": 1506 }, { "epoch": 1.7836638058597218, "grad_norm": 5.125, "learning_rate": 1.6235821843271723e-05, "loss": 2.7348, "step": 1507 }, { "epoch": 1.7848475880438, "grad_norm": 4.96875, "learning_rate": 1.6229353242653922e-05, "loss": 2.5342, "step": 1508 }, { "epoch": 1.786031370227878, "grad_norm": 4.75, "learning_rate": 1.6222880379846584e-05, "loss": 2.3652, "step": 1509 }, { "epoch": 1.7872151524119562, "grad_norm": 4.71875, "learning_rate": 1.6216403259278514e-05, "loss": 2.6787, "step": 1510 }, { "epoch": 1.7883989345960343, "grad_norm": 4.53125, "learning_rate": 1.6209921885381418e-05, "loss": 2.5929, "step": 1511 }, { "epoch": 1.7895827167801124, "grad_norm": 4.875, "learning_rate": 1.6203436262589927e-05, "loss": 2.6339, "step": 1512 }, { "epoch": 1.7907664989641905, "grad_norm": 4.875, "learning_rate": 1.6196946395341576e-05, "loss": 2.6158, "step": 1513 }, { "epoch": 1.7919502811482686, "grad_norm": 4.34375, "learning_rate": 1.6190452288076795e-05, "loss": 2.3103, "step": 1514 }, { "epoch": 1.7931340633323467, "grad_norm": 4.9375, "learning_rate": 1.6183953945238924e-05, "loss": 2.3014, "step": 1515 }, { "epoch": 1.7943178455164248, "grad_norm": 4.25, "learning_rate": 1.6177451371274197e-05, "loss": 2.5633, "step": 1516 }, { "epoch": 1.7955016277005031, "grad_norm": 4.90625, "learning_rate": 1.6170944570631743e-05, "loss": 2.4785, "step": 1517 }, { "epoch": 1.7966854098845813, "grad_norm": 4.15625, "learning_rate": 1.6164433547763584e-05, "loss": 2.2275, "step": 1518 }, { "epoch": 1.7978691920686594, "grad_norm": 4.21875, "learning_rate": 1.6157918307124635e-05, "loss": 2.7579, "step": 1519 }, { "epoch": 1.7990529742527375, "grad_norm": 4.5625, "learning_rate": 1.615139885317269e-05, "loss": 2.6342, "step": 1520 }, { "epoch": 1.8002367564368156, "grad_norm": 4.625, "learning_rate": 1.6144875190368426e-05, "loss": 2.4538, "step": 1521 }, { "epoch": 1.801420538620894, "grad_norm": 4.8125, "learning_rate": 1.6138347323175402e-05, "loss": 2.7132, "step": 1522 }, { "epoch": 1.802604320804972, "grad_norm": 4.40625, "learning_rate": 1.6131815256060064e-05, "loss": 2.5983, "step": 1523 }, { "epoch": 1.8037881029890501, "grad_norm": 4.75, "learning_rate": 1.612527899349171e-05, "loss": 2.274, "step": 1524 }, { "epoch": 1.8049718851731282, "grad_norm": 4.25, "learning_rate": 1.6118738539942524e-05, "loss": 2.6839, "step": 1525 }, { "epoch": 1.8061556673572063, "grad_norm": 4.40625, "learning_rate": 1.6112193899887555e-05, "loss": 2.4272, "step": 1526 }, { "epoch": 1.8073394495412844, "grad_norm": 4.9375, "learning_rate": 1.6105645077804713e-05, "loss": 2.6545, "step": 1527 }, { "epoch": 1.8085232317253626, "grad_norm": 4.71875, "learning_rate": 1.6099092078174773e-05, "loss": 2.4089, "step": 1528 }, { "epoch": 1.8097070139094407, "grad_norm": 3.96875, "learning_rate": 1.6092534905481367e-05, "loss": 2.4973, "step": 1529 }, { "epoch": 1.8108907960935188, "grad_norm": 4.71875, "learning_rate": 1.608597356421098e-05, "loss": 2.3043, "step": 1530 }, { "epoch": 1.8120745782775969, "grad_norm": 4.46875, "learning_rate": 1.6079408058852956e-05, "loss": 2.5248, "step": 1531 }, { "epoch": 1.813258360461675, "grad_norm": 3.828125, "learning_rate": 1.6072838393899476e-05, "loss": 2.2844, "step": 1532 }, { "epoch": 1.814442142645753, "grad_norm": 4.71875, "learning_rate": 1.606626457384558e-05, "loss": 2.4149, "step": 1533 }, { "epoch": 1.8156259248298312, "grad_norm": 4.59375, "learning_rate": 1.6059686603189146e-05, "loss": 2.7061, "step": 1534 }, { "epoch": 1.8168097070139093, "grad_norm": 4.09375, "learning_rate": 1.6053104486430887e-05, "loss": 2.5458, "step": 1535 }, { "epoch": 1.8179934891979874, "grad_norm": 4.5, "learning_rate": 1.604651822807436e-05, "loss": 2.3286, "step": 1536 }, { "epoch": 1.8191772713820658, "grad_norm": 4.6875, "learning_rate": 1.6039927832625953e-05, "loss": 2.3905, "step": 1537 }, { "epoch": 1.8203610535661439, "grad_norm": 5.40625, "learning_rate": 1.6033333304594886e-05, "loss": 2.6047, "step": 1538 }, { "epoch": 1.821544835750222, "grad_norm": 4.71875, "learning_rate": 1.6026734648493195e-05, "loss": 2.6729, "step": 1539 }, { "epoch": 1.8227286179343, "grad_norm": 4.28125, "learning_rate": 1.6020131868835762e-05, "loss": 2.6592, "step": 1540 }, { "epoch": 1.8239124001183782, "grad_norm": 4.40625, "learning_rate": 1.6013524970140276e-05, "loss": 2.7581, "step": 1541 }, { "epoch": 1.8250961823024563, "grad_norm": 4.5, "learning_rate": 1.6006913956927242e-05, "loss": 2.8888, "step": 1542 }, { "epoch": 1.8262799644865346, "grad_norm": 4.5, "learning_rate": 1.6000298833719996e-05, "loss": 2.4704, "step": 1543 }, { "epoch": 1.8274637466706127, "grad_norm": 4.0, "learning_rate": 1.5993679605044662e-05, "loss": 2.1948, "step": 1544 }, { "epoch": 1.8286475288546908, "grad_norm": 4.625, "learning_rate": 1.5987056275430196e-05, "loss": 2.9459, "step": 1545 }, { "epoch": 1.829831311038769, "grad_norm": 4.3125, "learning_rate": 1.598042884940835e-05, "loss": 2.6676, "step": 1546 }, { "epoch": 1.831015093222847, "grad_norm": 4.0625, "learning_rate": 1.5973797331513674e-05, "loss": 2.2903, "step": 1547 }, { "epoch": 1.8321988754069252, "grad_norm": 4.28125, "learning_rate": 1.5967161726283527e-05, "loss": 2.6227, "step": 1548 }, { "epoch": 1.8333826575910033, "grad_norm": 4.3125, "learning_rate": 1.5960522038258057e-05, "loss": 2.4619, "step": 1549 }, { "epoch": 1.8345664397750814, "grad_norm": 4.40625, "learning_rate": 1.5953878271980212e-05, "loss": 2.0762, "step": 1550 }, { "epoch": 1.8357502219591595, "grad_norm": 4.78125, "learning_rate": 1.5947230431995725e-05, "loss": 2.6358, "step": 1551 }, { "epoch": 1.8369340041432376, "grad_norm": 4.28125, "learning_rate": 1.594057852285312e-05, "loss": 2.7416, "step": 1552 }, { "epoch": 1.8381177863273157, "grad_norm": 4.375, "learning_rate": 1.5933922549103708e-05, "loss": 2.6301, "step": 1553 }, { "epoch": 1.8393015685113938, "grad_norm": 4.15625, "learning_rate": 1.5927262515301565e-05, "loss": 2.8739, "step": 1554 }, { "epoch": 1.840485350695472, "grad_norm": 4.53125, "learning_rate": 1.592059842600357e-05, "loss": 2.765, "step": 1555 }, { "epoch": 1.84166913287955, "grad_norm": 4.84375, "learning_rate": 1.5913930285769356e-05, "loss": 2.4774, "step": 1556 }, { "epoch": 1.8428529150636281, "grad_norm": 4.78125, "learning_rate": 1.5907258099161335e-05, "loss": 2.7309, "step": 1557 }, { "epoch": 1.8440366972477065, "grad_norm": 4.9375, "learning_rate": 1.5900581870744692e-05, "loss": 2.7091, "step": 1558 }, { "epoch": 1.8452204794317846, "grad_norm": 4.375, "learning_rate": 1.589390160508737e-05, "loss": 2.4672, "step": 1559 }, { "epoch": 1.8464042616158627, "grad_norm": 4.34375, "learning_rate": 1.5887217306760082e-05, "loss": 2.462, "step": 1560 }, { "epoch": 1.8475880437999408, "grad_norm": 4.59375, "learning_rate": 1.588052898033629e-05, "loss": 2.4996, "step": 1561 }, { "epoch": 1.848771825984019, "grad_norm": 3.859375, "learning_rate": 1.5873836630392218e-05, "loss": 2.2212, "step": 1562 }, { "epoch": 1.8499556081680972, "grad_norm": 4.625, "learning_rate": 1.5867140261506845e-05, "loss": 2.4566, "step": 1563 }, { "epoch": 1.8511393903521753, "grad_norm": 4.3125, "learning_rate": 1.5860439878261905e-05, "loss": 2.2199, "step": 1564 }, { "epoch": 1.8523231725362534, "grad_norm": 3.6875, "learning_rate": 1.5853735485241858e-05, "loss": 2.3118, "step": 1565 }, { "epoch": 1.8535069547203316, "grad_norm": 4.46875, "learning_rate": 1.5847027087033926e-05, "loss": 2.4885, "step": 1566 }, { "epoch": 1.8546907369044097, "grad_norm": 4.5, "learning_rate": 1.584031468822807e-05, "loss": 2.6762, "step": 1567 }, { "epoch": 1.8558745190884878, "grad_norm": 4.0625, "learning_rate": 1.5833598293416978e-05, "loss": 2.4962, "step": 1568 }, { "epoch": 1.8570583012725659, "grad_norm": 4.3125, "learning_rate": 1.5826877907196084e-05, "loss": 2.5425, "step": 1569 }, { "epoch": 1.858242083456644, "grad_norm": 4.59375, "learning_rate": 1.5820153534163543e-05, "loss": 2.9535, "step": 1570 }, { "epoch": 1.859425865640722, "grad_norm": 3.734375, "learning_rate": 1.581342517892025e-05, "loss": 2.4433, "step": 1571 }, { "epoch": 1.8606096478248002, "grad_norm": 4.53125, "learning_rate": 1.5806692846069807e-05, "loss": 2.7046, "step": 1572 }, { "epoch": 1.8617934300088783, "grad_norm": 4.28125, "learning_rate": 1.5799956540218553e-05, "loss": 2.3486, "step": 1573 }, { "epoch": 1.8629772121929564, "grad_norm": 4.4375, "learning_rate": 1.579321626597554e-05, "loss": 2.7323, "step": 1574 }, { "epoch": 1.8641609943770345, "grad_norm": 5.46875, "learning_rate": 1.5786472027952533e-05, "loss": 2.4158, "step": 1575 }, { "epoch": 1.8653447765611126, "grad_norm": 4.3125, "learning_rate": 1.5779723830764013e-05, "loss": 2.47, "step": 1576 }, { "epoch": 1.8665285587451907, "grad_norm": 4.375, "learning_rate": 1.5772971679027168e-05, "loss": 2.7919, "step": 1577 }, { "epoch": 1.8677123409292689, "grad_norm": 5.3125, "learning_rate": 1.576621557736189e-05, "loss": 2.3671, "step": 1578 }, { "epoch": 1.8688961231133472, "grad_norm": 4.25, "learning_rate": 1.5759455530390782e-05, "loss": 2.2356, "step": 1579 }, { "epoch": 1.8700799052974253, "grad_norm": 4.34375, "learning_rate": 1.575269154273913e-05, "loss": 2.5139, "step": 1580 }, { "epoch": 1.8712636874815034, "grad_norm": 4.34375, "learning_rate": 1.5745923619034928e-05, "loss": 2.5149, "step": 1581 }, { "epoch": 1.8724474696655815, "grad_norm": 4.78125, "learning_rate": 1.5739151763908867e-05, "loss": 2.4043, "step": 1582 }, { "epoch": 1.8736312518496596, "grad_norm": 4.71875, "learning_rate": 1.573237598199432e-05, "loss": 2.5043, "step": 1583 }, { "epoch": 1.874815034033738, "grad_norm": 4.59375, "learning_rate": 1.5725596277927342e-05, "loss": 2.3628, "step": 1584 }, { "epoch": 1.875998816217816, "grad_norm": 4.53125, "learning_rate": 1.571881265634669e-05, "loss": 2.5394, "step": 1585 }, { "epoch": 1.8771825984018942, "grad_norm": 4.71875, "learning_rate": 1.571202512189378e-05, "loss": 2.901, "step": 1586 }, { "epoch": 1.8783663805859723, "grad_norm": 4.6875, "learning_rate": 1.570523367921272e-05, "loss": 2.722, "step": 1587 }, { "epoch": 1.8795501627700504, "grad_norm": 4.75, "learning_rate": 1.5698438332950287e-05, "loss": 2.5781, "step": 1588 }, { "epoch": 1.8807339449541285, "grad_norm": 4.25, "learning_rate": 1.5691639087755928e-05, "loss": 2.8246, "step": 1589 }, { "epoch": 1.8819177271382066, "grad_norm": 4.4375, "learning_rate": 1.568483594828176e-05, "loss": 2.2658, "step": 1590 }, { "epoch": 1.8831015093222847, "grad_norm": 4.96875, "learning_rate": 1.5678028919182558e-05, "loss": 2.5398, "step": 1591 }, { "epoch": 1.8842852915063628, "grad_norm": 4.375, "learning_rate": 1.5671218005115767e-05, "loss": 2.6135, "step": 1592 }, { "epoch": 1.885469073690441, "grad_norm": 4.90625, "learning_rate": 1.566440321074149e-05, "loss": 2.6339, "step": 1593 }, { "epoch": 1.886652855874519, "grad_norm": 4.65625, "learning_rate": 1.5657584540722476e-05, "loss": 2.5428, "step": 1594 }, { "epoch": 1.8878366380585971, "grad_norm": 3.890625, "learning_rate": 1.5650761999724134e-05, "loss": 2.1484, "step": 1595 }, { "epoch": 1.8890204202426752, "grad_norm": 4.65625, "learning_rate": 1.5643935592414518e-05, "loss": 2.5943, "step": 1596 }, { "epoch": 1.8902042024267534, "grad_norm": 4.78125, "learning_rate": 1.563710532346433e-05, "loss": 2.6228, "step": 1597 }, { "epoch": 1.8913879846108315, "grad_norm": 4.6875, "learning_rate": 1.563027119754691e-05, "loss": 2.2851, "step": 1598 }, { "epoch": 1.8925717667949096, "grad_norm": 4.65625, "learning_rate": 1.5623433219338244e-05, "loss": 2.6253, "step": 1599 }, { "epoch": 1.893755548978988, "grad_norm": 4.59375, "learning_rate": 1.5616591393516946e-05, "loss": 2.3895, "step": 1600 }, { "epoch": 1.894939331163066, "grad_norm": 4.625, "learning_rate": 1.5609745724764264e-05, "loss": 2.5855, "step": 1601 }, { "epoch": 1.8961231133471441, "grad_norm": 4.21875, "learning_rate": 1.5602896217764075e-05, "loss": 2.4339, "step": 1602 }, { "epoch": 1.8973068955312222, "grad_norm": 4.28125, "learning_rate": 1.559604287720289e-05, "loss": 2.2148, "step": 1603 }, { "epoch": 1.8984906777153003, "grad_norm": 4.65625, "learning_rate": 1.558918570776984e-05, "loss": 2.6404, "step": 1604 }, { "epoch": 1.8996744598993787, "grad_norm": 4.28125, "learning_rate": 1.5582324714156663e-05, "loss": 2.3721, "step": 1605 }, { "epoch": 1.9008582420834568, "grad_norm": 4.5625, "learning_rate": 1.557545990105773e-05, "loss": 2.5847, "step": 1606 }, { "epoch": 1.9020420242675349, "grad_norm": 4.09375, "learning_rate": 1.556859127317002e-05, "loss": 2.819, "step": 1607 }, { "epoch": 1.903225806451613, "grad_norm": 4.71875, "learning_rate": 1.5561718835193117e-05, "loss": 2.5872, "step": 1608 }, { "epoch": 1.904409588635691, "grad_norm": 4.1875, "learning_rate": 1.555484259182922e-05, "loss": 2.2552, "step": 1609 }, { "epoch": 1.9055933708197692, "grad_norm": 4.6875, "learning_rate": 1.5547962547783126e-05, "loss": 2.7986, "step": 1610 }, { "epoch": 1.9067771530038473, "grad_norm": 4.53125, "learning_rate": 1.554107870776223e-05, "loss": 2.6342, "step": 1611 }, { "epoch": 1.9079609351879254, "grad_norm": 3.890625, "learning_rate": 1.553419107647653e-05, "loss": 2.2921, "step": 1612 }, { "epoch": 1.9091447173720035, "grad_norm": 4.3125, "learning_rate": 1.5527299658638627e-05, "loss": 2.6741, "step": 1613 }, { "epoch": 1.9103284995560816, "grad_norm": 4.0, "learning_rate": 1.5520404458963683e-05, "loss": 2.5391, "step": 1614 }, { "epoch": 1.9115122817401597, "grad_norm": 4.78125, "learning_rate": 1.551350548216948e-05, "loss": 2.4837, "step": 1615 }, { "epoch": 1.9126960639242379, "grad_norm": 4.71875, "learning_rate": 1.5506602732976373e-05, "loss": 3.1054, "step": 1616 }, { "epoch": 1.913879846108316, "grad_norm": 4.84375, "learning_rate": 1.5499696216107293e-05, "loss": 2.8691, "step": 1617 }, { "epoch": 1.915063628292394, "grad_norm": 4.59375, "learning_rate": 1.5492785936287744e-05, "loss": 2.4272, "step": 1618 }, { "epoch": 1.9162474104764722, "grad_norm": 4.8125, "learning_rate": 1.5485871898245824e-05, "loss": 2.7555, "step": 1619 }, { "epoch": 1.9174311926605505, "grad_norm": 4.125, "learning_rate": 1.547895410671218e-05, "loss": 2.3232, "step": 1620 }, { "epoch": 1.9186149748446286, "grad_norm": 4.625, "learning_rate": 1.5472032566420047e-05, "loss": 2.666, "step": 1621 }, { "epoch": 1.9197987570287067, "grad_norm": 4.53125, "learning_rate": 1.546510728210522e-05, "loss": 2.4065, "step": 1622 }, { "epoch": 1.9209825392127848, "grad_norm": 4.59375, "learning_rate": 1.545817825850604e-05, "loss": 2.6675, "step": 1623 }, { "epoch": 1.922166321396863, "grad_norm": 5.375, "learning_rate": 1.5451245500363424e-05, "loss": 2.7244, "step": 1624 }, { "epoch": 1.9233501035809413, "grad_norm": 4.75, "learning_rate": 1.544430901242084e-05, "loss": 2.8175, "step": 1625 }, { "epoch": 1.9245338857650194, "grad_norm": 4.34375, "learning_rate": 1.5437368799424307e-05, "loss": 2.6767, "step": 1626 }, { "epoch": 1.9257176679490975, "grad_norm": 4.53125, "learning_rate": 1.5430424866122386e-05, "loss": 2.5192, "step": 1627 }, { "epoch": 1.9269014501331756, "grad_norm": 4.90625, "learning_rate": 1.54234772172662e-05, "loss": 2.7755, "step": 1628 }, { "epoch": 1.9280852323172537, "grad_norm": 4.90625, "learning_rate": 1.5416525857609396e-05, "loss": 2.8522, "step": 1629 }, { "epoch": 1.9292690145013318, "grad_norm": 5.1875, "learning_rate": 1.5409570791908172e-05, "loss": 2.8513, "step": 1630 }, { "epoch": 1.93045279668541, "grad_norm": 4.65625, "learning_rate": 1.5402612024921254e-05, "loss": 2.5188, "step": 1631 }, { "epoch": 1.931636578869488, "grad_norm": 4.59375, "learning_rate": 1.5395649561409903e-05, "loss": 2.2901, "step": 1632 }, { "epoch": 1.9328203610535661, "grad_norm": 4.3125, "learning_rate": 1.5388683406137923e-05, "loss": 2.8115, "step": 1633 }, { "epoch": 1.9340041432376442, "grad_norm": 5.0625, "learning_rate": 1.5381713563871616e-05, "loss": 2.6666, "step": 1634 }, { "epoch": 1.9351879254217224, "grad_norm": 4.46875, "learning_rate": 1.537474003937983e-05, "loss": 2.8578, "step": 1635 }, { "epoch": 1.9363717076058005, "grad_norm": 4.03125, "learning_rate": 1.536776283743392e-05, "loss": 2.3264, "step": 1636 }, { "epoch": 1.9375554897898786, "grad_norm": 4.40625, "learning_rate": 1.536078196280777e-05, "loss": 2.8186, "step": 1637 }, { "epoch": 1.9387392719739567, "grad_norm": 4.4375, "learning_rate": 1.5353797420277755e-05, "loss": 2.6879, "step": 1638 }, { "epoch": 1.9399230541580348, "grad_norm": 4.375, "learning_rate": 1.5346809214622785e-05, "loss": 2.8315, "step": 1639 }, { "epoch": 1.941106836342113, "grad_norm": 5.0625, "learning_rate": 1.533981735062426e-05, "loss": 2.5794, "step": 1640 }, { "epoch": 1.9422906185261912, "grad_norm": 4.9375, "learning_rate": 1.5332821833066083e-05, "loss": 2.6975, "step": 1641 }, { "epoch": 1.9434744007102693, "grad_norm": 4.5, "learning_rate": 1.532582266673467e-05, "loss": 2.3622, "step": 1642 }, { "epoch": 1.9446581828943474, "grad_norm": 4.3125, "learning_rate": 1.5318819856418918e-05, "loss": 2.6653, "step": 1643 }, { "epoch": 1.9458419650784256, "grad_norm": 4.53125, "learning_rate": 1.5311813406910227e-05, "loss": 2.5855, "step": 1644 }, { "epoch": 1.9470257472625037, "grad_norm": 4.375, "learning_rate": 1.5304803323002483e-05, "loss": 2.6988, "step": 1645 }, { "epoch": 1.948209529446582, "grad_norm": 3.703125, "learning_rate": 1.5297789609492062e-05, "loss": 2.4194, "step": 1646 }, { "epoch": 1.94939331163066, "grad_norm": 4.15625, "learning_rate": 1.5290772271177818e-05, "loss": 2.7034, "step": 1647 }, { "epoch": 1.9505770938147382, "grad_norm": 4.34375, "learning_rate": 1.5283751312861093e-05, "loss": 2.465, "step": 1648 }, { "epoch": 1.9517608759988163, "grad_norm": 4.21875, "learning_rate": 1.527672673934569e-05, "loss": 2.6465, "step": 1649 }, { "epoch": 1.9529446581828944, "grad_norm": 4.25, "learning_rate": 1.5269698555437913e-05, "loss": 2.1861, "step": 1650 }, { "epoch": 1.9541284403669725, "grad_norm": 4.21875, "learning_rate": 1.526266676594651e-05, "loss": 2.5179, "step": 1651 }, { "epoch": 1.9553122225510506, "grad_norm": 4.59375, "learning_rate": 1.5255631375682702e-05, "loss": 2.3044, "step": 1652 }, { "epoch": 1.9564960047351287, "grad_norm": 4.5, "learning_rate": 1.5248592389460189e-05, "loss": 2.5345, "step": 1653 }, { "epoch": 1.9576797869192069, "grad_norm": 4.1875, "learning_rate": 1.5241549812095114e-05, "loss": 2.4082, "step": 1654 }, { "epoch": 1.958863569103285, "grad_norm": 4.625, "learning_rate": 1.5234503648406075e-05, "loss": 2.5556, "step": 1655 }, { "epoch": 1.960047351287363, "grad_norm": 4.75, "learning_rate": 1.5227453903214147e-05, "loss": 2.4638, "step": 1656 }, { "epoch": 1.9612311334714412, "grad_norm": 4.8125, "learning_rate": 1.5220400581342833e-05, "loss": 3.154, "step": 1657 }, { "epoch": 1.9624149156555193, "grad_norm": 4.84375, "learning_rate": 1.521334368761809e-05, "loss": 2.4266, "step": 1658 }, { "epoch": 1.9635986978395974, "grad_norm": 4.46875, "learning_rate": 1.520628322686832e-05, "loss": 2.7446, "step": 1659 }, { "epoch": 1.9647824800236755, "grad_norm": 4.875, "learning_rate": 1.5199219203924367e-05, "loss": 2.7426, "step": 1660 }, { "epoch": 1.9659662622077536, "grad_norm": 4.15625, "learning_rate": 1.5192151623619506e-05, "loss": 2.512, "step": 1661 }, { "epoch": 1.967150044391832, "grad_norm": 5.25, "learning_rate": 1.5185080490789457e-05, "loss": 2.9938, "step": 1662 }, { "epoch": 1.96833382657591, "grad_norm": 4.8125, "learning_rate": 1.517800581027236e-05, "loss": 2.783, "step": 1663 }, { "epoch": 1.9695176087599882, "grad_norm": 4.5, "learning_rate": 1.5170927586908787e-05, "loss": 2.5316, "step": 1664 }, { "epoch": 1.9707013909440663, "grad_norm": 4.5, "learning_rate": 1.516384582554173e-05, "loss": 2.4079, "step": 1665 }, { "epoch": 1.9718851731281444, "grad_norm": 5.0, "learning_rate": 1.515676053101661e-05, "loss": 2.5656, "step": 1666 }, { "epoch": 1.9730689553122227, "grad_norm": 4.3125, "learning_rate": 1.514967170818126e-05, "loss": 2.4211, "step": 1667 }, { "epoch": 1.9742527374963008, "grad_norm": 4.53125, "learning_rate": 1.5142579361885927e-05, "loss": 2.738, "step": 1668 }, { "epoch": 1.975436519680379, "grad_norm": 4.03125, "learning_rate": 1.5135483496983273e-05, "loss": 2.3867, "step": 1669 }, { "epoch": 1.976620301864457, "grad_norm": 5.25, "learning_rate": 1.5128384118328355e-05, "loss": 3.0491, "step": 1670 }, { "epoch": 1.9778040840485351, "grad_norm": 4.59375, "learning_rate": 1.512128123077865e-05, "loss": 2.5732, "step": 1671 }, { "epoch": 1.9789878662326132, "grad_norm": 4.375, "learning_rate": 1.5114174839194029e-05, "loss": 2.5626, "step": 1672 }, { "epoch": 1.9801716484166914, "grad_norm": 4.8125, "learning_rate": 1.5107064948436758e-05, "loss": 2.5721, "step": 1673 }, { "epoch": 1.9813554306007695, "grad_norm": 4.90625, "learning_rate": 1.50999515633715e-05, "loss": 2.5355, "step": 1674 }, { "epoch": 1.9825392127848476, "grad_norm": 4.875, "learning_rate": 1.5092834688865311e-05, "loss": 2.8889, "step": 1675 }, { "epoch": 1.9837229949689257, "grad_norm": 4.46875, "learning_rate": 1.5085714329787631e-05, "loss": 2.5055, "step": 1676 }, { "epoch": 1.9849067771530038, "grad_norm": 4.15625, "learning_rate": 1.507859049101028e-05, "loss": 2.3902, "step": 1677 }, { "epoch": 1.986090559337082, "grad_norm": 4.96875, "learning_rate": 1.5071463177407473e-05, "loss": 2.5625, "step": 1678 }, { "epoch": 1.98727434152116, "grad_norm": 4.21875, "learning_rate": 1.5064332393855787e-05, "loss": 2.7436, "step": 1679 }, { "epoch": 1.9884581237052381, "grad_norm": 4.84375, "learning_rate": 1.5057198145234182e-05, "loss": 2.7899, "step": 1680 }, { "epoch": 1.9896419058893162, "grad_norm": 4.875, "learning_rate": 1.5050060436423986e-05, "loss": 2.6249, "step": 1681 }, { "epoch": 1.9908256880733946, "grad_norm": 4.96875, "learning_rate": 1.5042919272308895e-05, "loss": 2.5594, "step": 1682 }, { "epoch": 1.9920094702574727, "grad_norm": 4.03125, "learning_rate": 1.5035774657774972e-05, "loss": 2.3699, "step": 1683 }, { "epoch": 1.9931932524415508, "grad_norm": 4.125, "learning_rate": 1.5028626597710632e-05, "loss": 2.5744, "step": 1684 }, { "epoch": 1.9943770346256289, "grad_norm": 4.34375, "learning_rate": 1.5021475097006657e-05, "loss": 2.7558, "step": 1685 }, { "epoch": 1.995560816809707, "grad_norm": 4.75, "learning_rate": 1.5014320160556182e-05, "loss": 2.6761, "step": 1686 }, { "epoch": 1.996744598993785, "grad_norm": 4.65625, "learning_rate": 1.5007161793254686e-05, "loss": 2.5028, "step": 1687 }, { "epoch": 1.9979283811778634, "grad_norm": 4.46875, "learning_rate": 1.5000000000000002e-05, "loss": 2.7599, "step": 1688 }, { "epoch": 1.9991121633619415, "grad_norm": 4.59375, "learning_rate": 1.4992834785692303e-05, "loss": 2.4262, "step": 1689 }, { "epoch": 2.0, "grad_norm": 4.0625, "learning_rate": 1.4985666155234109e-05, "loss": 2.355, "step": 1690 }, { "epoch": 2.001183782184078, "grad_norm": 4.65625, "learning_rate": 1.4978494113530268e-05, "loss": 2.5362, "step": 1691 }, { "epoch": 2.002367564368156, "grad_norm": 4.3125, "learning_rate": 1.4971318665487974e-05, "loss": 2.3415, "step": 1692 }, { "epoch": 2.0035513465522343, "grad_norm": 3.9375, "learning_rate": 1.4964139816016737e-05, "loss": 2.5667, "step": 1693 }, { "epoch": 2.0047351287363124, "grad_norm": 4.25, "learning_rate": 1.4956957570028401e-05, "loss": 2.8624, "step": 1694 }, { "epoch": 2.0059189109203905, "grad_norm": 4.46875, "learning_rate": 1.494977193243714e-05, "loss": 2.8211, "step": 1695 }, { "epoch": 2.0071026931044686, "grad_norm": 4.21875, "learning_rate": 1.4942582908159446e-05, "loss": 2.3195, "step": 1696 }, { "epoch": 2.0082864752885468, "grad_norm": 4.40625, "learning_rate": 1.4935390502114118e-05, "loss": 2.4066, "step": 1697 }, { "epoch": 2.009470257472625, "grad_norm": 4.03125, "learning_rate": 1.492819471922228e-05, "loss": 2.7981, "step": 1698 }, { "epoch": 2.010654039656703, "grad_norm": 4.1875, "learning_rate": 1.4920995564407368e-05, "loss": 2.4407, "step": 1699 }, { "epoch": 2.011837821840781, "grad_norm": 5.09375, "learning_rate": 1.4913793042595109e-05, "loss": 2.9573, "step": 1700 }, { "epoch": 2.0130216040248596, "grad_norm": 4.21875, "learning_rate": 1.4906587158713552e-05, "loss": 2.3029, "step": 1701 }, { "epoch": 2.0142053862089377, "grad_norm": 4.8125, "learning_rate": 1.4899377917693041e-05, "loss": 2.3616, "step": 1702 }, { "epoch": 2.015389168393016, "grad_norm": 4.5625, "learning_rate": 1.489216532446621e-05, "loss": 2.2945, "step": 1703 }, { "epoch": 2.016572950577094, "grad_norm": 3.90625, "learning_rate": 1.4884949383967993e-05, "loss": 2.3251, "step": 1704 }, { "epoch": 2.017756732761172, "grad_norm": 3.84375, "learning_rate": 1.4877730101135619e-05, "loss": 2.1333, "step": 1705 }, { "epoch": 2.01894051494525, "grad_norm": 3.828125, "learning_rate": 1.4870507480908587e-05, "loss": 2.4205, "step": 1706 }, { "epoch": 2.0201242971293283, "grad_norm": 4.1875, "learning_rate": 1.4863281528228698e-05, "loss": 2.3213, "step": 1707 }, { "epoch": 2.0213080793134064, "grad_norm": 4.71875, "learning_rate": 1.4856052248040023e-05, "loss": 2.4637, "step": 1708 }, { "epoch": 2.0224918614974845, "grad_norm": 4.125, "learning_rate": 1.4848819645288915e-05, "loss": 2.5064, "step": 1709 }, { "epoch": 2.0236756436815626, "grad_norm": 4.6875, "learning_rate": 1.4841583724923994e-05, "loss": 2.5004, "step": 1710 }, { "epoch": 2.0248594258656407, "grad_norm": 4.6875, "learning_rate": 1.4834344491896156e-05, "loss": 2.332, "step": 1711 }, { "epoch": 2.026043208049719, "grad_norm": 4.3125, "learning_rate": 1.4827101951158555e-05, "loss": 2.6641, "step": 1712 }, { "epoch": 2.027226990233797, "grad_norm": 4.6875, "learning_rate": 1.4819856107666622e-05, "loss": 2.5881, "step": 1713 }, { "epoch": 2.028410772417875, "grad_norm": 4.1875, "learning_rate": 1.4812606966378039e-05, "loss": 2.5364, "step": 1714 }, { "epoch": 2.029594554601953, "grad_norm": 4.3125, "learning_rate": 1.480535453225274e-05, "loss": 2.8768, "step": 1715 }, { "epoch": 2.0307783367860313, "grad_norm": 4.21875, "learning_rate": 1.479809881025292e-05, "loss": 2.6438, "step": 1716 }, { "epoch": 2.0319621189701094, "grad_norm": 3.84375, "learning_rate": 1.4790839805343025e-05, "loss": 2.1244, "step": 1717 }, { "epoch": 2.0331459011541875, "grad_norm": 4.375, "learning_rate": 1.4783577522489733e-05, "loss": 2.3285, "step": 1718 }, { "epoch": 2.0343296833382656, "grad_norm": 4.4375, "learning_rate": 1.4776311966661987e-05, "loss": 2.2874, "step": 1719 }, { "epoch": 2.0355134655223437, "grad_norm": 4.46875, "learning_rate": 1.476904314283095e-05, "loss": 2.6197, "step": 1720 }, { "epoch": 2.036697247706422, "grad_norm": 4.03125, "learning_rate": 1.4761771055970032e-05, "loss": 2.6129, "step": 1721 }, { "epoch": 2.0378810298905004, "grad_norm": 4.59375, "learning_rate": 1.4754495711054866e-05, "loss": 2.3914, "step": 1722 }, { "epoch": 2.0390648120745785, "grad_norm": 4.40625, "learning_rate": 1.4747217113063328e-05, "loss": 2.4114, "step": 1723 }, { "epoch": 2.0402485942586566, "grad_norm": 4.65625, "learning_rate": 1.4739935266975503e-05, "loss": 2.5446, "step": 1724 }, { "epoch": 2.0414323764427347, "grad_norm": 4.40625, "learning_rate": 1.4732650177773717e-05, "loss": 2.3897, "step": 1725 }, { "epoch": 2.042616158626813, "grad_norm": 4.46875, "learning_rate": 1.4725361850442502e-05, "loss": 2.5595, "step": 1726 }, { "epoch": 2.043799940810891, "grad_norm": 4.75, "learning_rate": 1.4718070289968602e-05, "loss": 2.0317, "step": 1727 }, { "epoch": 2.044983722994969, "grad_norm": 4.125, "learning_rate": 1.4710775501340988e-05, "loss": 2.3788, "step": 1728 }, { "epoch": 2.046167505179047, "grad_norm": 4.53125, "learning_rate": 1.470347748955083e-05, "loss": 2.4617, "step": 1729 }, { "epoch": 2.047351287363125, "grad_norm": 4.78125, "learning_rate": 1.4696176259591501e-05, "loss": 2.3813, "step": 1730 }, { "epoch": 2.0485350695472033, "grad_norm": 4.3125, "learning_rate": 1.4688871816458585e-05, "loss": 2.5347, "step": 1731 }, { "epoch": 2.0497188517312814, "grad_norm": 4.875, "learning_rate": 1.4681564165149863e-05, "loss": 2.6147, "step": 1732 }, { "epoch": 2.0509026339153595, "grad_norm": 4.71875, "learning_rate": 1.4674253310665294e-05, "loss": 2.2024, "step": 1733 }, { "epoch": 2.0520864160994376, "grad_norm": 4.0625, "learning_rate": 1.4666939258007054e-05, "loss": 2.2973, "step": 1734 }, { "epoch": 2.0532701982835158, "grad_norm": 4.1875, "learning_rate": 1.4659622012179493e-05, "loss": 2.4476, "step": 1735 }, { "epoch": 2.054453980467594, "grad_norm": 4.5625, "learning_rate": 1.4652301578189141e-05, "loss": 2.2416, "step": 1736 }, { "epoch": 2.055637762651672, "grad_norm": 4.21875, "learning_rate": 1.4644977961044725e-05, "loss": 2.4832, "step": 1737 }, { "epoch": 2.05682154483575, "grad_norm": 4.75, "learning_rate": 1.4637651165757143e-05, "loss": 2.8891, "step": 1738 }, { "epoch": 2.058005327019828, "grad_norm": 5.125, "learning_rate": 1.463032119733946e-05, "loss": 2.6613, "step": 1739 }, { "epoch": 2.0591891092039063, "grad_norm": 4.625, "learning_rate": 1.4622988060806917e-05, "loss": 2.1469, "step": 1740 }, { "epoch": 2.0603728913879844, "grad_norm": 4.75, "learning_rate": 1.4615651761176928e-05, "loss": 2.2883, "step": 1741 }, { "epoch": 2.061556673572063, "grad_norm": 4.125, "learning_rate": 1.4608312303469067e-05, "loss": 1.9205, "step": 1742 }, { "epoch": 2.062740455756141, "grad_norm": 4.40625, "learning_rate": 1.4600969692705067e-05, "loss": 2.1989, "step": 1743 }, { "epoch": 2.063924237940219, "grad_norm": 4.4375, "learning_rate": 1.4593623933908822e-05, "loss": 2.3431, "step": 1744 }, { "epoch": 2.0651080201242973, "grad_norm": 4.34375, "learning_rate": 1.4586275032106373e-05, "loss": 2.6643, "step": 1745 }, { "epoch": 2.0662918023083754, "grad_norm": 4.375, "learning_rate": 1.4578922992325924e-05, "loss": 2.3968, "step": 1746 }, { "epoch": 2.0674755844924535, "grad_norm": 3.9375, "learning_rate": 1.4571567819597812e-05, "loss": 2.4407, "step": 1747 }, { "epoch": 2.0686593666765316, "grad_norm": 4.28125, "learning_rate": 1.4564209518954528e-05, "loss": 2.3108, "step": 1748 }, { "epoch": 2.0698431488606097, "grad_norm": 4.03125, "learning_rate": 1.4556848095430698e-05, "loss": 2.5595, "step": 1749 }, { "epoch": 2.071026931044688, "grad_norm": 5.1875, "learning_rate": 1.454948355406309e-05, "loss": 2.3537, "step": 1750 }, { "epoch": 2.072210713228766, "grad_norm": 4.53125, "learning_rate": 1.4542115899890594e-05, "loss": 2.3694, "step": 1751 }, { "epoch": 2.073394495412844, "grad_norm": 4.25, "learning_rate": 1.453474513795424e-05, "loss": 2.2887, "step": 1752 }, { "epoch": 2.074578277596922, "grad_norm": 4.21875, "learning_rate": 1.4527371273297184e-05, "loss": 2.2114, "step": 1753 }, { "epoch": 2.0757620597810003, "grad_norm": 4.9375, "learning_rate": 1.4519994310964697e-05, "loss": 2.4878, "step": 1754 }, { "epoch": 2.0769458419650784, "grad_norm": 4.875, "learning_rate": 1.4512614256004182e-05, "loss": 2.5361, "step": 1755 }, { "epoch": 2.0781296241491565, "grad_norm": 5.75, "learning_rate": 1.4505231113465148e-05, "loss": 2.2848, "step": 1756 }, { "epoch": 2.0793134063332346, "grad_norm": 4.53125, "learning_rate": 1.4497844888399216e-05, "loss": 2.8935, "step": 1757 }, { "epoch": 2.0804971885173127, "grad_norm": 3.921875, "learning_rate": 1.4490455585860124e-05, "loss": 2.2565, "step": 1758 }, { "epoch": 2.081680970701391, "grad_norm": 4.40625, "learning_rate": 1.448306321090371e-05, "loss": 2.1338, "step": 1759 }, { "epoch": 2.082864752885469, "grad_norm": 4.65625, "learning_rate": 1.447566776858791e-05, "loss": 2.3315, "step": 1760 }, { "epoch": 2.084048535069547, "grad_norm": 4.53125, "learning_rate": 1.4468269263972773e-05, "loss": 2.3271, "step": 1761 }, { "epoch": 2.085232317253625, "grad_norm": 4.40625, "learning_rate": 1.4460867702120432e-05, "loss": 2.463, "step": 1762 }, { "epoch": 2.0864160994377037, "grad_norm": 4.4375, "learning_rate": 1.4453463088095108e-05, "loss": 2.8607, "step": 1763 }, { "epoch": 2.087599881621782, "grad_norm": 4.875, "learning_rate": 1.4446055426963121e-05, "loss": 2.6397, "step": 1764 }, { "epoch": 2.08878366380586, "grad_norm": 4.46875, "learning_rate": 1.443864472379287e-05, "loss": 2.3896, "step": 1765 }, { "epoch": 2.089967445989938, "grad_norm": 4.28125, "learning_rate": 1.4431230983654839e-05, "loss": 2.6684, "step": 1766 }, { "epoch": 2.091151228174016, "grad_norm": 4.34375, "learning_rate": 1.4423814211621586e-05, "loss": 2.6584, "step": 1767 }, { "epoch": 2.092335010358094, "grad_norm": 4.78125, "learning_rate": 1.4416394412767747e-05, "loss": 2.6279, "step": 1768 }, { "epoch": 2.0935187925421723, "grad_norm": 4.40625, "learning_rate": 1.440897159217002e-05, "loss": 2.2147, "step": 1769 }, { "epoch": 2.0947025747262504, "grad_norm": 4.28125, "learning_rate": 1.4401545754907187e-05, "loss": 2.4643, "step": 1770 }, { "epoch": 2.0958863569103285, "grad_norm": 4.40625, "learning_rate": 1.4394116906060078e-05, "loss": 2.1649, "step": 1771 }, { "epoch": 2.0970701390944066, "grad_norm": 4.78125, "learning_rate": 1.4386685050711593e-05, "loss": 2.4068, "step": 1772 }, { "epoch": 2.0982539212784848, "grad_norm": 4.34375, "learning_rate": 1.4379250193946685e-05, "loss": 2.1727, "step": 1773 }, { "epoch": 2.099437703462563, "grad_norm": 4.78125, "learning_rate": 1.4371812340852366e-05, "loss": 2.3846, "step": 1774 }, { "epoch": 2.100621485646641, "grad_norm": 4.03125, "learning_rate": 1.4364371496517688e-05, "loss": 2.353, "step": 1775 }, { "epoch": 2.101805267830719, "grad_norm": 4.5625, "learning_rate": 1.435692766603376e-05, "loss": 2.3814, "step": 1776 }, { "epoch": 2.102989050014797, "grad_norm": 4.1875, "learning_rate": 1.4349480854493724e-05, "loss": 2.6886, "step": 1777 }, { "epoch": 2.1041728321988753, "grad_norm": 4.4375, "learning_rate": 1.4342031066992775e-05, "loss": 2.7724, "step": 1778 }, { "epoch": 2.1053566143829534, "grad_norm": 4.625, "learning_rate": 1.4334578308628131e-05, "loss": 2.4362, "step": 1779 }, { "epoch": 2.1065403965670315, "grad_norm": 4.75, "learning_rate": 1.4327122584499052e-05, "loss": 2.7077, "step": 1780 }, { "epoch": 2.1077241787511096, "grad_norm": 4.28125, "learning_rate": 1.4319663899706818e-05, "loss": 2.4197, "step": 1781 }, { "epoch": 2.1089079609351877, "grad_norm": 4.15625, "learning_rate": 1.4312202259354746e-05, "loss": 2.3042, "step": 1782 }, { "epoch": 2.1100917431192663, "grad_norm": 4.5, "learning_rate": 1.4304737668548165e-05, "loss": 2.4247, "step": 1783 }, { "epoch": 2.1112755253033444, "grad_norm": 4.40625, "learning_rate": 1.4297270132394432e-05, "loss": 2.4303, "step": 1784 }, { "epoch": 2.1124593074874225, "grad_norm": 5.0, "learning_rate": 1.4289799656002912e-05, "loss": 2.4613, "step": 1785 }, { "epoch": 2.1136430896715006, "grad_norm": 4.625, "learning_rate": 1.4282326244484983e-05, "loss": 2.6143, "step": 1786 }, { "epoch": 2.1148268718555787, "grad_norm": 4.5625, "learning_rate": 1.4274849902954034e-05, "loss": 2.3809, "step": 1787 }, { "epoch": 2.116010654039657, "grad_norm": 4.125, "learning_rate": 1.4267370636525457e-05, "loss": 2.375, "step": 1788 }, { "epoch": 2.117194436223735, "grad_norm": 4.78125, "learning_rate": 1.4259888450316641e-05, "loss": 2.8907, "step": 1789 }, { "epoch": 2.118378218407813, "grad_norm": 4.25, "learning_rate": 1.4252403349446986e-05, "loss": 2.6747, "step": 1790 }, { "epoch": 2.119562000591891, "grad_norm": 4.75, "learning_rate": 1.424491533903787e-05, "loss": 2.463, "step": 1791 }, { "epoch": 2.1207457827759693, "grad_norm": 4.875, "learning_rate": 1.4237424424212673e-05, "loss": 2.6864, "step": 1792 }, { "epoch": 2.1219295649600474, "grad_norm": 4.96875, "learning_rate": 1.4229930610096754e-05, "loss": 2.4775, "step": 1793 }, { "epoch": 2.1231133471441255, "grad_norm": 4.78125, "learning_rate": 1.4222433901817468e-05, "loss": 2.7405, "step": 1794 }, { "epoch": 2.1242971293282036, "grad_norm": 4.40625, "learning_rate": 1.4214934304504133e-05, "loss": 2.3605, "step": 1795 }, { "epoch": 2.1254809115122817, "grad_norm": 4.96875, "learning_rate": 1.4207431823288058e-05, "loss": 2.7435, "step": 1796 }, { "epoch": 2.12666469369636, "grad_norm": 4.46875, "learning_rate": 1.419992646330252e-05, "loss": 2.4695, "step": 1797 }, { "epoch": 2.127848475880438, "grad_norm": 4.4375, "learning_rate": 1.4192418229682768e-05, "loss": 2.6574, "step": 1798 }, { "epoch": 2.129032258064516, "grad_norm": 4.90625, "learning_rate": 1.4184907127566006e-05, "loss": 2.74, "step": 1799 }, { "epoch": 2.130216040248594, "grad_norm": 4.21875, "learning_rate": 1.4177393162091421e-05, "loss": 2.3665, "step": 1800 }, { "epoch": 2.1313998224326722, "grad_norm": 4.28125, "learning_rate": 1.4169876338400141e-05, "loss": 2.3922, "step": 1801 }, { "epoch": 2.1325836046167503, "grad_norm": 4.40625, "learning_rate": 1.4162356661635262e-05, "loss": 1.9996, "step": 1802 }, { "epoch": 2.1337673868008284, "grad_norm": 4.8125, "learning_rate": 1.4154834136941817e-05, "loss": 2.6442, "step": 1803 }, { "epoch": 2.134951168984907, "grad_norm": 4.9375, "learning_rate": 1.414730876946681e-05, "loss": 2.5194, "step": 1804 }, { "epoch": 2.136134951168985, "grad_norm": 4.28125, "learning_rate": 1.4139780564359162e-05, "loss": 2.1657, "step": 1805 }, { "epoch": 2.137318733353063, "grad_norm": 4.375, "learning_rate": 1.4132249526769765e-05, "loss": 2.3975, "step": 1806 }, { "epoch": 2.1385025155371413, "grad_norm": 4.5, "learning_rate": 1.412471566185143e-05, "loss": 2.4665, "step": 1807 }, { "epoch": 2.1396862977212194, "grad_norm": 4.5625, "learning_rate": 1.4117178974758903e-05, "loss": 2.69, "step": 1808 }, { "epoch": 2.1408700799052975, "grad_norm": 4.8125, "learning_rate": 1.4109639470648868e-05, "loss": 2.408, "step": 1809 }, { "epoch": 2.1420538620893756, "grad_norm": 4.84375, "learning_rate": 1.4102097154679938e-05, "loss": 2.2774, "step": 1810 }, { "epoch": 2.1432376442734538, "grad_norm": 4.375, "learning_rate": 1.4094552032012635e-05, "loss": 2.4685, "step": 1811 }, { "epoch": 2.144421426457532, "grad_norm": 4.40625, "learning_rate": 1.4087004107809424e-05, "loss": 2.3241, "step": 1812 }, { "epoch": 2.14560520864161, "grad_norm": 4.09375, "learning_rate": 1.4079453387234667e-05, "loss": 2.425, "step": 1813 }, { "epoch": 2.146788990825688, "grad_norm": 4.40625, "learning_rate": 1.407189987545465e-05, "loss": 2.4282, "step": 1814 }, { "epoch": 2.147972773009766, "grad_norm": 5.375, "learning_rate": 1.4064343577637565e-05, "loss": 2.6459, "step": 1815 }, { "epoch": 2.1491565551938443, "grad_norm": 4.3125, "learning_rate": 1.4056784498953511e-05, "loss": 2.3669, "step": 1816 }, { "epoch": 2.1503403373779224, "grad_norm": 4.71875, "learning_rate": 1.404922264457449e-05, "loss": 2.726, "step": 1817 }, { "epoch": 2.1515241195620005, "grad_norm": 4.53125, "learning_rate": 1.4041658019674405e-05, "loss": 2.4405, "step": 1818 }, { "epoch": 2.1527079017460786, "grad_norm": 4.03125, "learning_rate": 1.4034090629429047e-05, "loss": 2.1815, "step": 1819 }, { "epoch": 2.1538916839301567, "grad_norm": 4.3125, "learning_rate": 1.402652047901611e-05, "loss": 2.5336, "step": 1820 }, { "epoch": 2.155075466114235, "grad_norm": 3.96875, "learning_rate": 1.401894757361517e-05, "loss": 2.3335, "step": 1821 }, { "epoch": 2.156259248298313, "grad_norm": 4.5, "learning_rate": 1.4011371918407686e-05, "loss": 2.1404, "step": 1822 }, { "epoch": 2.157443030482391, "grad_norm": 5.25, "learning_rate": 1.4003793518577007e-05, "loss": 2.604, "step": 1823 }, { "epoch": 2.158626812666469, "grad_norm": 4.375, "learning_rate": 1.3996212379308351e-05, "loss": 2.2331, "step": 1824 }, { "epoch": 2.1598105948505477, "grad_norm": 4.5625, "learning_rate": 1.3988628505788814e-05, "loss": 2.381, "step": 1825 }, { "epoch": 2.160994377034626, "grad_norm": 4.65625, "learning_rate": 1.3981041903207364e-05, "loss": 2.264, "step": 1826 }, { "epoch": 2.162178159218704, "grad_norm": 5.6875, "learning_rate": 1.3973452576754833e-05, "loss": 2.6553, "step": 1827 }, { "epoch": 2.163361941402782, "grad_norm": 4.25, "learning_rate": 1.396586053162392e-05, "loss": 2.1041, "step": 1828 }, { "epoch": 2.16454572358686, "grad_norm": 5.0625, "learning_rate": 1.3958265773009184e-05, "loss": 2.206, "step": 1829 }, { "epoch": 2.1657295057709383, "grad_norm": 5.3125, "learning_rate": 1.3950668306107035e-05, "loss": 2.5869, "step": 1830 }, { "epoch": 2.1669132879550164, "grad_norm": 4.03125, "learning_rate": 1.3943068136115745e-05, "loss": 2.0516, "step": 1831 }, { "epoch": 2.1680970701390945, "grad_norm": 4.4375, "learning_rate": 1.3935465268235429e-05, "loss": 2.6177, "step": 1832 }, { "epoch": 2.1692808523231726, "grad_norm": 5.4375, "learning_rate": 1.392785970766805e-05, "loss": 2.4327, "step": 1833 }, { "epoch": 2.1704646345072507, "grad_norm": 4.875, "learning_rate": 1.3920251459617413e-05, "loss": 2.6276, "step": 1834 }, { "epoch": 2.171648416691329, "grad_norm": 4.53125, "learning_rate": 1.3912640529289163e-05, "loss": 2.5906, "step": 1835 }, { "epoch": 2.172832198875407, "grad_norm": 4.3125, "learning_rate": 1.390502692189078e-05, "loss": 2.1734, "step": 1836 }, { "epoch": 2.174015981059485, "grad_norm": 4.71875, "learning_rate": 1.389741064263157e-05, "loss": 2.3805, "step": 1837 }, { "epoch": 2.175199763243563, "grad_norm": 5.34375, "learning_rate": 1.3889791696722676e-05, "loss": 2.6029, "step": 1838 }, { "epoch": 2.1763835454276412, "grad_norm": 4.53125, "learning_rate": 1.388217008937706e-05, "loss": 2.1681, "step": 1839 }, { "epoch": 2.1775673276117193, "grad_norm": 4.59375, "learning_rate": 1.3874545825809509e-05, "loss": 2.4004, "step": 1840 }, { "epoch": 2.1787511097957974, "grad_norm": 4.53125, "learning_rate": 1.3866918911236624e-05, "loss": 2.6339, "step": 1841 }, { "epoch": 2.1799348919798756, "grad_norm": 4.4375, "learning_rate": 1.3859289350876822e-05, "loss": 2.427, "step": 1842 }, { "epoch": 2.1811186741639537, "grad_norm": 4.78125, "learning_rate": 1.3851657149950325e-05, "loss": 2.4561, "step": 1843 }, { "epoch": 2.1823024563480318, "grad_norm": 4.75, "learning_rate": 1.3844022313679167e-05, "loss": 2.5039, "step": 1844 }, { "epoch": 2.18348623853211, "grad_norm": 4.40625, "learning_rate": 1.3836384847287186e-05, "loss": 2.3525, "step": 1845 }, { "epoch": 2.1846700207161884, "grad_norm": 4.78125, "learning_rate": 1.3828744756000015e-05, "loss": 2.3911, "step": 1846 }, { "epoch": 2.1858538029002665, "grad_norm": 4.5625, "learning_rate": 1.3821102045045088e-05, "loss": 2.3933, "step": 1847 }, { "epoch": 2.1870375850843446, "grad_norm": 4.6875, "learning_rate": 1.381345671965163e-05, "loss": 2.2233, "step": 1848 }, { "epoch": 2.1882213672684228, "grad_norm": 4.8125, "learning_rate": 1.380580878505065e-05, "loss": 2.3741, "step": 1849 }, { "epoch": 2.189405149452501, "grad_norm": 4.28125, "learning_rate": 1.3798158246474946e-05, "loss": 2.6141, "step": 1850 }, { "epoch": 2.190588931636579, "grad_norm": 4.03125, "learning_rate": 1.3790505109159101e-05, "loss": 2.3754, "step": 1851 }, { "epoch": 2.191772713820657, "grad_norm": 4.625, "learning_rate": 1.3782849378339469e-05, "loss": 2.7928, "step": 1852 }, { "epoch": 2.192956496004735, "grad_norm": 4.78125, "learning_rate": 1.3775191059254185e-05, "loss": 2.28, "step": 1853 }, { "epoch": 2.1941402781888133, "grad_norm": 4.625, "learning_rate": 1.3767530157143154e-05, "loss": 2.0261, "step": 1854 }, { "epoch": 2.1953240603728914, "grad_norm": 4.4375, "learning_rate": 1.3759866677248041e-05, "loss": 2.6561, "step": 1855 }, { "epoch": 2.1965078425569695, "grad_norm": 4.34375, "learning_rate": 1.3752200624812282e-05, "loss": 2.1501, "step": 1856 }, { "epoch": 2.1976916247410476, "grad_norm": 5.0, "learning_rate": 1.3744532005081072e-05, "loss": 2.5123, "step": 1857 }, { "epoch": 2.1988754069251257, "grad_norm": 4.6875, "learning_rate": 1.3736860823301364e-05, "loss": 2.3715, "step": 1858 }, { "epoch": 2.200059189109204, "grad_norm": 5.0625, "learning_rate": 1.3729187084721858e-05, "loss": 2.9818, "step": 1859 }, { "epoch": 2.201242971293282, "grad_norm": 4.75, "learning_rate": 1.3721510794593012e-05, "loss": 2.4336, "step": 1860 }, { "epoch": 2.20242675347736, "grad_norm": 4.625, "learning_rate": 1.3713831958167018e-05, "loss": 2.7295, "step": 1861 }, { "epoch": 2.203610535661438, "grad_norm": 4.8125, "learning_rate": 1.3706150580697826e-05, "loss": 2.401, "step": 1862 }, { "epoch": 2.2047943178455163, "grad_norm": 5.09375, "learning_rate": 1.369846666744111e-05, "loss": 2.2277, "step": 1863 }, { "epoch": 2.2059781000295944, "grad_norm": 4.09375, "learning_rate": 1.3690780223654286e-05, "loss": 2.3052, "step": 1864 }, { "epoch": 2.2071618822136725, "grad_norm": 4.4375, "learning_rate": 1.36830912545965e-05, "loss": 2.7898, "step": 1865 }, { "epoch": 2.2083456643977506, "grad_norm": 4.96875, "learning_rate": 1.367539976552863e-05, "loss": 2.2699, "step": 1866 }, { "epoch": 2.209529446581829, "grad_norm": 4.40625, "learning_rate": 1.366770576171327e-05, "loss": 2.3287, "step": 1867 }, { "epoch": 2.2107132287659073, "grad_norm": 4.25, "learning_rate": 1.3660009248414735e-05, "loss": 2.2977, "step": 1868 }, { "epoch": 2.2118970109499854, "grad_norm": 4.46875, "learning_rate": 1.3652310230899074e-05, "loss": 2.2517, "step": 1869 }, { "epoch": 2.2130807931340635, "grad_norm": 5.09375, "learning_rate": 1.3644608714434027e-05, "loss": 2.6559, "step": 1870 }, { "epoch": 2.2142645753181416, "grad_norm": 4.09375, "learning_rate": 1.3636904704289053e-05, "loss": 2.5086, "step": 1871 }, { "epoch": 2.2154483575022197, "grad_norm": 5.0, "learning_rate": 1.362919820573532e-05, "loss": 2.3636, "step": 1872 }, { "epoch": 2.216632139686298, "grad_norm": 4.78125, "learning_rate": 1.3621489224045696e-05, "loss": 2.5281, "step": 1873 }, { "epoch": 2.217815921870376, "grad_norm": 4.75, "learning_rate": 1.3613777764494747e-05, "loss": 2.3574, "step": 1874 }, { "epoch": 2.218999704054454, "grad_norm": 4.875, "learning_rate": 1.3606063832358736e-05, "loss": 2.4623, "step": 1875 }, { "epoch": 2.220183486238532, "grad_norm": 4.75, "learning_rate": 1.3598347432915617e-05, "loss": 2.9833, "step": 1876 }, { "epoch": 2.2213672684226102, "grad_norm": 4.90625, "learning_rate": 1.3590628571445033e-05, "loss": 2.5842, "step": 1877 }, { "epoch": 2.2225510506066883, "grad_norm": 4.40625, "learning_rate": 1.358290725322831e-05, "loss": 2.648, "step": 1878 }, { "epoch": 2.2237348327907664, "grad_norm": 4.75, "learning_rate": 1.3575183483548453e-05, "loss": 2.1666, "step": 1879 }, { "epoch": 2.2249186149748446, "grad_norm": 4.53125, "learning_rate": 1.3567457267690152e-05, "loss": 2.6704, "step": 1880 }, { "epoch": 2.2261023971589227, "grad_norm": 4.71875, "learning_rate": 1.3559728610939767e-05, "loss": 2.427, "step": 1881 }, { "epoch": 2.2272861793430008, "grad_norm": 4.25, "learning_rate": 1.3551997518585317e-05, "loss": 2.4591, "step": 1882 }, { "epoch": 2.228469961527079, "grad_norm": 4.5625, "learning_rate": 1.3544263995916508e-05, "loss": 1.8919, "step": 1883 }, { "epoch": 2.229653743711157, "grad_norm": 5.0625, "learning_rate": 1.3536528048224698e-05, "loss": 2.9449, "step": 1884 }, { "epoch": 2.230837525895235, "grad_norm": 4.84375, "learning_rate": 1.3528789680802896e-05, "loss": 2.6015, "step": 1885 }, { "epoch": 2.232021308079313, "grad_norm": 4.71875, "learning_rate": 1.3521048898945778e-05, "loss": 2.3254, "step": 1886 }, { "epoch": 2.2332050902633913, "grad_norm": 4.90625, "learning_rate": 1.3513305707949675e-05, "loss": 2.6851, "step": 1887 }, { "epoch": 2.23438887244747, "grad_norm": 4.71875, "learning_rate": 1.3505560113112555e-05, "loss": 2.3163, "step": 1888 }, { "epoch": 2.235572654631548, "grad_norm": 5.5, "learning_rate": 1.3497812119734037e-05, "loss": 2.4845, "step": 1889 }, { "epoch": 2.236756436815626, "grad_norm": 4.53125, "learning_rate": 1.3490061733115383e-05, "loss": 2.4337, "step": 1890 }, { "epoch": 2.237940218999704, "grad_norm": 4.625, "learning_rate": 1.3482308958559483e-05, "loss": 2.5866, "step": 1891 }, { "epoch": 2.2391240011837823, "grad_norm": 4.375, "learning_rate": 1.3474553801370871e-05, "loss": 2.6681, "step": 1892 }, { "epoch": 2.2403077833678604, "grad_norm": 4.0, "learning_rate": 1.3466796266855713e-05, "loss": 2.2691, "step": 1893 }, { "epoch": 2.2414915655519385, "grad_norm": 4.625, "learning_rate": 1.345903636032179e-05, "loss": 2.5285, "step": 1894 }, { "epoch": 2.2426753477360166, "grad_norm": 4.03125, "learning_rate": 1.3451274087078514e-05, "loss": 2.3269, "step": 1895 }, { "epoch": 2.2438591299200947, "grad_norm": 3.6875, "learning_rate": 1.3443509452436917e-05, "loss": 1.8805, "step": 1896 }, { "epoch": 2.245042912104173, "grad_norm": 4.125, "learning_rate": 1.3435742461709635e-05, "loss": 2.2577, "step": 1897 }, { "epoch": 2.246226694288251, "grad_norm": 5.40625, "learning_rate": 1.342797312021094e-05, "loss": 2.7196, "step": 1898 }, { "epoch": 2.247410476472329, "grad_norm": 4.9375, "learning_rate": 1.342020143325669e-05, "loss": 2.5034, "step": 1899 }, { "epoch": 2.248594258656407, "grad_norm": 4.625, "learning_rate": 1.3412427406164353e-05, "loss": 2.4901, "step": 1900 }, { "epoch": 2.2497780408404853, "grad_norm": 4.03125, "learning_rate": 1.3404651044253006e-05, "loss": 2.3284, "step": 1901 }, { "epoch": 2.2509618230245634, "grad_norm": 4.4375, "learning_rate": 1.3396872352843318e-05, "loss": 2.255, "step": 1902 }, { "epoch": 2.2521456052086415, "grad_norm": 4.1875, "learning_rate": 1.3389091337257543e-05, "loss": 2.244, "step": 1903 }, { "epoch": 2.2533293873927196, "grad_norm": 4.875, "learning_rate": 1.3381308002819546e-05, "loss": 2.315, "step": 1904 }, { "epoch": 2.2545131695767977, "grad_norm": 3.953125, "learning_rate": 1.3373522354854765e-05, "loss": 2.1306, "step": 1905 }, { "epoch": 2.255696951760876, "grad_norm": 4.78125, "learning_rate": 1.3365734398690217e-05, "loss": 2.6124, "step": 1906 }, { "epoch": 2.2568807339449544, "grad_norm": 4.75, "learning_rate": 1.3357944139654508e-05, "loss": 2.3498, "step": 1907 }, { "epoch": 2.258064516129032, "grad_norm": 4.4375, "learning_rate": 1.335015158307782e-05, "loss": 2.1802, "step": 1908 }, { "epoch": 2.2592482983131106, "grad_norm": 4.5, "learning_rate": 1.334235673429189e-05, "loss": 2.787, "step": 1909 }, { "epoch": 2.2604320804971887, "grad_norm": 4.28125, "learning_rate": 1.333455959863005e-05, "loss": 2.1723, "step": 1910 }, { "epoch": 2.261615862681267, "grad_norm": 4.78125, "learning_rate": 1.332676018142718e-05, "loss": 2.3373, "step": 1911 }, { "epoch": 2.262799644865345, "grad_norm": 5.0625, "learning_rate": 1.3318958488019716e-05, "loss": 2.5705, "step": 1912 }, { "epoch": 2.263983427049423, "grad_norm": 4.4375, "learning_rate": 1.3311154523745663e-05, "loss": 2.2126, "step": 1913 }, { "epoch": 2.265167209233501, "grad_norm": 4.125, "learning_rate": 1.3303348293944585e-05, "loss": 2.1853, "step": 1914 }, { "epoch": 2.2663509914175792, "grad_norm": 4.625, "learning_rate": 1.3295539803957573e-05, "loss": 2.63, "step": 1915 }, { "epoch": 2.2675347736016573, "grad_norm": 5.03125, "learning_rate": 1.3287729059127288e-05, "loss": 2.9392, "step": 1916 }, { "epoch": 2.2687185557857354, "grad_norm": 4.375, "learning_rate": 1.3279916064797924e-05, "loss": 2.2827, "step": 1917 }, { "epoch": 2.2699023379698136, "grad_norm": 4.5625, "learning_rate": 1.327210082631521e-05, "loss": 2.5639, "step": 1918 }, { "epoch": 2.2710861201538917, "grad_norm": 4.34375, "learning_rate": 1.3264283349026419e-05, "loss": 2.5249, "step": 1919 }, { "epoch": 2.2722699023379698, "grad_norm": 3.875, "learning_rate": 1.325646363828035e-05, "loss": 2.4246, "step": 1920 }, { "epoch": 2.273453684522048, "grad_norm": 5.09375, "learning_rate": 1.3248641699427332e-05, "loss": 2.3138, "step": 1921 }, { "epoch": 2.274637466706126, "grad_norm": 4.65625, "learning_rate": 1.324081753781922e-05, "loss": 2.5512, "step": 1922 }, { "epoch": 2.275821248890204, "grad_norm": 5.09375, "learning_rate": 1.323299115880939e-05, "loss": 2.6956, "step": 1923 }, { "epoch": 2.277005031074282, "grad_norm": 4.9375, "learning_rate": 1.3225162567752725e-05, "loss": 2.4824, "step": 1924 }, { "epoch": 2.2781888132583603, "grad_norm": 4.78125, "learning_rate": 1.3217331770005639e-05, "loss": 2.5514, "step": 1925 }, { "epoch": 2.2793725954424384, "grad_norm": 4.28125, "learning_rate": 1.3209498770926043e-05, "loss": 2.2069, "step": 1926 }, { "epoch": 2.2805563776265165, "grad_norm": 4.25, "learning_rate": 1.3201663575873359e-05, "loss": 2.3894, "step": 1927 }, { "epoch": 2.281740159810595, "grad_norm": 4.21875, "learning_rate": 1.3193826190208507e-05, "loss": 2.3747, "step": 1928 }, { "epoch": 2.2829239419946727, "grad_norm": 5.0625, "learning_rate": 1.318598661929392e-05, "loss": 2.7194, "step": 1929 }, { "epoch": 2.2841077241787513, "grad_norm": 4.40625, "learning_rate": 1.3178144868493502e-05, "loss": 2.6552, "step": 1930 }, { "epoch": 2.2852915063628294, "grad_norm": 4.53125, "learning_rate": 1.3170300943172669e-05, "loss": 2.2784, "step": 1931 }, { "epoch": 2.2864752885469075, "grad_norm": 4.78125, "learning_rate": 1.3162454848698318e-05, "loss": 2.7236, "step": 1932 }, { "epoch": 2.2876590707309856, "grad_norm": 4.71875, "learning_rate": 1.3154606590438828e-05, "loss": 2.55, "step": 1933 }, { "epoch": 2.2888428529150637, "grad_norm": 4.59375, "learning_rate": 1.3146756173764061e-05, "loss": 2.5079, "step": 1934 }, { "epoch": 2.290026635099142, "grad_norm": 4.28125, "learning_rate": 1.3138903604045358e-05, "loss": 2.6022, "step": 1935 }, { "epoch": 2.29121041728322, "grad_norm": 4.09375, "learning_rate": 1.3131048886655529e-05, "loss": 2.0834, "step": 1936 }, { "epoch": 2.292394199467298, "grad_norm": 4.21875, "learning_rate": 1.3123192026968851e-05, "loss": 2.515, "step": 1937 }, { "epoch": 2.293577981651376, "grad_norm": 4.78125, "learning_rate": 1.3115333030361078e-05, "loss": 2.7158, "step": 1938 }, { "epoch": 2.2947617638354543, "grad_norm": 4.375, "learning_rate": 1.310747190220941e-05, "loss": 2.5531, "step": 1939 }, { "epoch": 2.2959455460195324, "grad_norm": 5.0625, "learning_rate": 1.309960864789252e-05, "loss": 2.6213, "step": 1940 }, { "epoch": 2.2971293282036105, "grad_norm": 4.46875, "learning_rate": 1.3091743272790533e-05, "loss": 2.5357, "step": 1941 }, { "epoch": 2.2983131103876886, "grad_norm": 4.625, "learning_rate": 1.3083875782285015e-05, "loss": 2.5705, "step": 1942 }, { "epoch": 2.2994968925717667, "grad_norm": 4.21875, "learning_rate": 1.3076006181758989e-05, "loss": 2.4425, "step": 1943 }, { "epoch": 2.300680674755845, "grad_norm": 5.1875, "learning_rate": 1.306813447659692e-05, "loss": 2.8868, "step": 1944 }, { "epoch": 2.301864456939923, "grad_norm": 4.75, "learning_rate": 1.3060260672184713e-05, "loss": 2.5714, "step": 1945 }, { "epoch": 2.303048239124001, "grad_norm": 4.78125, "learning_rate": 1.3052384773909706e-05, "loss": 2.7782, "step": 1946 }, { "epoch": 2.304232021308079, "grad_norm": 4.5, "learning_rate": 1.3044506787160676e-05, "loss": 2.4336, "step": 1947 }, { "epoch": 2.3054158034921572, "grad_norm": 4.34375, "learning_rate": 1.3036626717327818e-05, "loss": 2.2095, "step": 1948 }, { "epoch": 2.306599585676236, "grad_norm": 4.53125, "learning_rate": 1.3028744569802765e-05, "loss": 2.7133, "step": 1949 }, { "epoch": 2.3077833678603135, "grad_norm": 4.75, "learning_rate": 1.3020860349978564e-05, "loss": 2.5646, "step": 1950 }, { "epoch": 2.308967150044392, "grad_norm": 4.65625, "learning_rate": 1.3012974063249684e-05, "loss": 2.7099, "step": 1951 }, { "epoch": 2.31015093222847, "grad_norm": 4.71875, "learning_rate": 1.3005085715012003e-05, "loss": 2.522, "step": 1952 }, { "epoch": 2.3113347144125482, "grad_norm": 4.71875, "learning_rate": 1.2997195310662816e-05, "loss": 2.9279, "step": 1953 }, { "epoch": 2.3125184965966263, "grad_norm": 4.34375, "learning_rate": 1.2989302855600816e-05, "loss": 2.3937, "step": 1954 }, { "epoch": 2.3137022787807044, "grad_norm": 4.09375, "learning_rate": 1.2981408355226109e-05, "loss": 2.3069, "step": 1955 }, { "epoch": 2.3148860609647826, "grad_norm": 4.34375, "learning_rate": 1.2973511814940194e-05, "loss": 2.6464, "step": 1956 }, { "epoch": 2.3160698431488607, "grad_norm": 3.96875, "learning_rate": 1.2965613240145969e-05, "loss": 2.3584, "step": 1957 }, { "epoch": 2.3172536253329388, "grad_norm": 4.6875, "learning_rate": 1.2957712636247722e-05, "loss": 2.59, "step": 1958 }, { "epoch": 2.318437407517017, "grad_norm": 4.75, "learning_rate": 1.2949810008651129e-05, "loss": 2.4227, "step": 1959 }, { "epoch": 2.319621189701095, "grad_norm": 4.5625, "learning_rate": 1.2941905362763252e-05, "loss": 2.3548, "step": 1960 }, { "epoch": 2.320804971885173, "grad_norm": 4.40625, "learning_rate": 1.2933998703992531e-05, "loss": 2.2282, "step": 1961 }, { "epoch": 2.321988754069251, "grad_norm": 4.9375, "learning_rate": 1.2926090037748793e-05, "loss": 2.1482, "step": 1962 }, { "epoch": 2.3231725362533293, "grad_norm": 4.875, "learning_rate": 1.2918179369443224e-05, "loss": 2.6413, "step": 1963 }, { "epoch": 2.3243563184374074, "grad_norm": 4.34375, "learning_rate": 1.291026670448839e-05, "loss": 2.5337, "step": 1964 }, { "epoch": 2.3255401006214855, "grad_norm": 4.375, "learning_rate": 1.2902352048298224e-05, "loss": 2.4493, "step": 1965 }, { "epoch": 2.3267238828055636, "grad_norm": 4.21875, "learning_rate": 1.289443540628801e-05, "loss": 2.2321, "step": 1966 }, { "epoch": 2.3279076649896417, "grad_norm": 4.6875, "learning_rate": 1.2886516783874403e-05, "loss": 2.4129, "step": 1967 }, { "epoch": 2.32909144717372, "grad_norm": 4.8125, "learning_rate": 1.2878596186475408e-05, "loss": 2.7259, "step": 1968 }, { "epoch": 2.330275229357798, "grad_norm": 4.40625, "learning_rate": 1.2870673619510382e-05, "loss": 2.1951, "step": 1969 }, { "epoch": 2.3314590115418765, "grad_norm": 4.125, "learning_rate": 1.2862749088400026e-05, "loss": 2.3104, "step": 1970 }, { "epoch": 2.332642793725954, "grad_norm": 4.53125, "learning_rate": 1.2854822598566394e-05, "loss": 2.4097, "step": 1971 }, { "epoch": 2.3338265759100327, "grad_norm": 4.71875, "learning_rate": 1.2846894155432868e-05, "loss": 2.1739, "step": 1972 }, { "epoch": 2.335010358094111, "grad_norm": 4.4375, "learning_rate": 1.283896376442418e-05, "loss": 2.3009, "step": 1973 }, { "epoch": 2.336194140278189, "grad_norm": 4.6875, "learning_rate": 1.283103143096638e-05, "loss": 2.7456, "step": 1974 }, { "epoch": 2.337377922462267, "grad_norm": 4.6875, "learning_rate": 1.2823097160486861e-05, "loss": 2.5979, "step": 1975 }, { "epoch": 2.338561704646345, "grad_norm": 3.984375, "learning_rate": 1.2815160958414332e-05, "loss": 2.1394, "step": 1976 }, { "epoch": 2.3397454868304233, "grad_norm": 4.59375, "learning_rate": 1.280722283017883e-05, "loss": 2.9308, "step": 1977 }, { "epoch": 2.3409292690145014, "grad_norm": 4.40625, "learning_rate": 1.2799282781211698e-05, "loss": 2.3911, "step": 1978 }, { "epoch": 2.3421130511985795, "grad_norm": 4.4375, "learning_rate": 1.279134081694561e-05, "loss": 2.5997, "step": 1979 }, { "epoch": 2.3432968333826576, "grad_norm": 4.375, "learning_rate": 1.278339694281454e-05, "loss": 2.159, "step": 1980 }, { "epoch": 2.3444806155667357, "grad_norm": 4.375, "learning_rate": 1.2775451164253767e-05, "loss": 2.5387, "step": 1981 }, { "epoch": 2.345664397750814, "grad_norm": 4.8125, "learning_rate": 1.2767503486699884e-05, "loss": 2.4691, "step": 1982 }, { "epoch": 2.346848179934892, "grad_norm": 4.53125, "learning_rate": 1.275955391559077e-05, "loss": 2.6654, "step": 1983 }, { "epoch": 2.34803196211897, "grad_norm": 4.0, "learning_rate": 1.2751602456365608e-05, "loss": 2.3952, "step": 1984 }, { "epoch": 2.349215744303048, "grad_norm": 4.09375, "learning_rate": 1.2743649114464873e-05, "loss": 2.2212, "step": 1985 }, { "epoch": 2.3503995264871262, "grad_norm": 5.1875, "learning_rate": 1.2735693895330324e-05, "loss": 2.4196, "step": 1986 }, { "epoch": 2.3515833086712044, "grad_norm": 4.5625, "learning_rate": 1.2727736804405005e-05, "loss": 2.5507, "step": 1987 }, { "epoch": 2.3527670908552825, "grad_norm": 4.84375, "learning_rate": 1.2719777847133241e-05, "loss": 2.6894, "step": 1988 }, { "epoch": 2.3539508730393606, "grad_norm": 4.59375, "learning_rate": 1.2711817028960642e-05, "loss": 2.699, "step": 1989 }, { "epoch": 2.3551346552234387, "grad_norm": 4.75, "learning_rate": 1.2703854355334073e-05, "loss": 2.6527, "step": 1990 }, { "epoch": 2.3563184374075172, "grad_norm": 4.34375, "learning_rate": 1.2695889831701691e-05, "loss": 2.3664, "step": 1991 }, { "epoch": 2.3575022195915953, "grad_norm": 4.40625, "learning_rate": 1.26879234635129e-05, "loss": 2.8566, "step": 1992 }, { "epoch": 2.3586860017756734, "grad_norm": 4.59375, "learning_rate": 1.2679955256218379e-05, "loss": 2.5967, "step": 1993 }, { "epoch": 2.3598697839597516, "grad_norm": 4.84375, "learning_rate": 1.2671985215270055e-05, "loss": 2.4393, "step": 1994 }, { "epoch": 2.3610535661438297, "grad_norm": 4.625, "learning_rate": 1.2664013346121115e-05, "loss": 2.5967, "step": 1995 }, { "epoch": 2.3622373483279078, "grad_norm": 4.6875, "learning_rate": 1.2656039654226e-05, "loss": 2.2489, "step": 1996 }, { "epoch": 2.363421130511986, "grad_norm": 4.9375, "learning_rate": 1.2648064145040392e-05, "loss": 2.5152, "step": 1997 }, { "epoch": 2.364604912696064, "grad_norm": 4.03125, "learning_rate": 1.264008682402122e-05, "loss": 2.2452, "step": 1998 }, { "epoch": 2.365788694880142, "grad_norm": 4.53125, "learning_rate": 1.2632107696626652e-05, "loss": 2.8025, "step": 1999 }, { "epoch": 2.36697247706422, "grad_norm": 5.1875, "learning_rate": 1.2624126768316088e-05, "loss": 2.6085, "step": 2000 }, { "epoch": 2.3681562592482983, "grad_norm": 4.03125, "learning_rate": 1.2616144044550167e-05, "loss": 2.3143, "step": 2001 }, { "epoch": 2.3693400414323764, "grad_norm": 4.3125, "learning_rate": 1.260815953079075e-05, "loss": 2.5876, "step": 2002 }, { "epoch": 2.3705238236164545, "grad_norm": 4.71875, "learning_rate": 1.2600173232500935e-05, "loss": 2.44, "step": 2003 }, { "epoch": 2.3717076058005326, "grad_norm": 4.28125, "learning_rate": 1.2592185155145023e-05, "loss": 2.2949, "step": 2004 }, { "epoch": 2.3728913879846107, "grad_norm": 4.625, "learning_rate": 1.2584195304188542e-05, "loss": 2.2805, "step": 2005 }, { "epoch": 2.374075170168689, "grad_norm": 5.0625, "learning_rate": 1.2576203685098233e-05, "loss": 2.9191, "step": 2006 }, { "epoch": 2.375258952352767, "grad_norm": 4.90625, "learning_rate": 1.256821030334205e-05, "loss": 2.607, "step": 2007 }, { "epoch": 2.376442734536845, "grad_norm": 4.53125, "learning_rate": 1.2560215164389149e-05, "loss": 2.4083, "step": 2008 }, { "epoch": 2.377626516720923, "grad_norm": 4.6875, "learning_rate": 1.2552218273709889e-05, "loss": 2.9244, "step": 2009 }, { "epoch": 2.3788102989050013, "grad_norm": 4.9375, "learning_rate": 1.254421963677582e-05, "loss": 2.542, "step": 2010 }, { "epoch": 2.3799940810890794, "grad_norm": 4.9375, "learning_rate": 1.2536219259059703e-05, "loss": 2.5614, "step": 2011 }, { "epoch": 2.381177863273158, "grad_norm": 4.25, "learning_rate": 1.2528217146035478e-05, "loss": 2.4155, "step": 2012 }, { "epoch": 2.382361645457236, "grad_norm": 5.125, "learning_rate": 1.2520213303178274e-05, "loss": 2.576, "step": 2013 }, { "epoch": 2.383545427641314, "grad_norm": 4.78125, "learning_rate": 1.251220773596441e-05, "loss": 2.5152, "step": 2014 }, { "epoch": 2.3847292098253923, "grad_norm": 4.96875, "learning_rate": 1.2504200449871378e-05, "loss": 2.6738, "step": 2015 }, { "epoch": 2.3859129920094704, "grad_norm": 4.0625, "learning_rate": 1.2496191450377844e-05, "loss": 2.4966, "step": 2016 }, { "epoch": 2.3870967741935485, "grad_norm": 4.3125, "learning_rate": 1.2488180742963654e-05, "loss": 2.1964, "step": 2017 }, { "epoch": 2.3882805563776266, "grad_norm": 4.28125, "learning_rate": 1.2480168333109818e-05, "loss": 2.536, "step": 2018 }, { "epoch": 2.3894643385617047, "grad_norm": 4.625, "learning_rate": 1.2472154226298513e-05, "loss": 2.3762, "step": 2019 }, { "epoch": 2.390648120745783, "grad_norm": 4.75, "learning_rate": 1.2464138428013072e-05, "loss": 2.7676, "step": 2020 }, { "epoch": 2.391831902929861, "grad_norm": 4.75, "learning_rate": 1.2456120943737996e-05, "loss": 2.6322, "step": 2021 }, { "epoch": 2.393015685113939, "grad_norm": 4.8125, "learning_rate": 1.244810177895893e-05, "loss": 2.8814, "step": 2022 }, { "epoch": 2.394199467298017, "grad_norm": 4.5625, "learning_rate": 1.2440080939162666e-05, "loss": 2.0925, "step": 2023 }, { "epoch": 2.3953832494820952, "grad_norm": 5.8125, "learning_rate": 1.2432058429837153e-05, "loss": 2.7459, "step": 2024 }, { "epoch": 2.3965670316661734, "grad_norm": 4.5, "learning_rate": 1.2424034256471478e-05, "loss": 2.4914, "step": 2025 }, { "epoch": 2.3977508138502515, "grad_norm": 4.6875, "learning_rate": 1.241600842455586e-05, "loss": 2.5003, "step": 2026 }, { "epoch": 2.3989345960343296, "grad_norm": 4.65625, "learning_rate": 1.2407980939581665e-05, "loss": 2.2493, "step": 2027 }, { "epoch": 2.4001183782184077, "grad_norm": 5.25, "learning_rate": 1.239995180704138e-05, "loss": 2.4934, "step": 2028 }, { "epoch": 2.401302160402486, "grad_norm": 5.46875, "learning_rate": 1.2391921032428622e-05, "loss": 2.693, "step": 2029 }, { "epoch": 2.402485942586564, "grad_norm": 5.0, "learning_rate": 1.2383888621238132e-05, "loss": 2.526, "step": 2030 }, { "epoch": 2.4036697247706424, "grad_norm": 4.3125, "learning_rate": 1.2375854578965771e-05, "loss": 2.6468, "step": 2031 }, { "epoch": 2.40485350695472, "grad_norm": 4.46875, "learning_rate": 1.2367818911108518e-05, "loss": 2.6167, "step": 2032 }, { "epoch": 2.4060372891387987, "grad_norm": 4.3125, "learning_rate": 1.2359781623164465e-05, "loss": 2.207, "step": 2033 }, { "epoch": 2.4072210713228768, "grad_norm": 4.65625, "learning_rate": 1.2351742720632798e-05, "loss": 2.7003, "step": 2034 }, { "epoch": 2.408404853506955, "grad_norm": 4.46875, "learning_rate": 1.234370220901383e-05, "loss": 2.6636, "step": 2035 }, { "epoch": 2.409588635691033, "grad_norm": 4.625, "learning_rate": 1.2335660093808962e-05, "loss": 2.5797, "step": 2036 }, { "epoch": 2.410772417875111, "grad_norm": 4.78125, "learning_rate": 1.232761638052069e-05, "loss": 2.6054, "step": 2037 }, { "epoch": 2.411956200059189, "grad_norm": 4.65625, "learning_rate": 1.2319571074652614e-05, "loss": 2.5611, "step": 2038 }, { "epoch": 2.4131399822432673, "grad_norm": 4.65625, "learning_rate": 1.2311524181709417e-05, "loss": 2.397, "step": 2039 }, { "epoch": 2.4143237644273454, "grad_norm": 4.71875, "learning_rate": 1.2303475707196866e-05, "loss": 2.4198, "step": 2040 }, { "epoch": 2.4155075466114235, "grad_norm": 4.71875, "learning_rate": 1.2295425656621812e-05, "loss": 2.5809, "step": 2041 }, { "epoch": 2.4166913287955016, "grad_norm": 4.40625, "learning_rate": 1.2287374035492184e-05, "loss": 2.4757, "step": 2042 }, { "epoch": 2.4178751109795797, "grad_norm": 4.53125, "learning_rate": 1.2279320849316991e-05, "loss": 2.5796, "step": 2043 }, { "epoch": 2.419058893163658, "grad_norm": 4.09375, "learning_rate": 1.2271266103606305e-05, "loss": 2.2757, "step": 2044 }, { "epoch": 2.420242675347736, "grad_norm": 4.15625, "learning_rate": 1.2263209803871271e-05, "loss": 2.2948, "step": 2045 }, { "epoch": 2.421426457531814, "grad_norm": 4.125, "learning_rate": 1.2255151955624092e-05, "loss": 2.3288, "step": 2046 }, { "epoch": 2.422610239715892, "grad_norm": 4.8125, "learning_rate": 1.2247092564378032e-05, "loss": 2.9012, "step": 2047 }, { "epoch": 2.4237940218999703, "grad_norm": 4.71875, "learning_rate": 1.2239031635647417e-05, "loss": 3.0116, "step": 2048 }, { "epoch": 2.4249778040840484, "grad_norm": 4.4375, "learning_rate": 1.2230969174947619e-05, "loss": 2.5322, "step": 2049 }, { "epoch": 2.4261615862681265, "grad_norm": 5.03125, "learning_rate": 1.2222905187795053e-05, "loss": 2.474, "step": 2050 }, { "epoch": 2.4273453684522046, "grad_norm": 5.3125, "learning_rate": 1.2214839679707193e-05, "loss": 2.7569, "step": 2051 }, { "epoch": 2.428529150636283, "grad_norm": 4.59375, "learning_rate": 1.220677265620254e-05, "loss": 2.7163, "step": 2052 }, { "epoch": 2.429712932820361, "grad_norm": 3.890625, "learning_rate": 1.2198704122800633e-05, "loss": 2.1411, "step": 2053 }, { "epoch": 2.4308967150044394, "grad_norm": 4.4375, "learning_rate": 1.2190634085022056e-05, "loss": 2.4638, "step": 2054 }, { "epoch": 2.4320804971885175, "grad_norm": 4.78125, "learning_rate": 1.2182562548388414e-05, "loss": 2.7036, "step": 2055 }, { "epoch": 2.4332642793725956, "grad_norm": 4.34375, "learning_rate": 1.2174489518422333e-05, "loss": 2.4019, "step": 2056 }, { "epoch": 2.4344480615566737, "grad_norm": 4.3125, "learning_rate": 1.216641500064747e-05, "loss": 2.1875, "step": 2057 }, { "epoch": 2.435631843740752, "grad_norm": 4.25, "learning_rate": 1.2158339000588493e-05, "loss": 2.3812, "step": 2058 }, { "epoch": 2.43681562592483, "grad_norm": 5.0625, "learning_rate": 1.2150261523771083e-05, "loss": 2.7453, "step": 2059 }, { "epoch": 2.437999408108908, "grad_norm": 4.34375, "learning_rate": 1.2142182575721946e-05, "loss": 2.6306, "step": 2060 }, { "epoch": 2.439183190292986, "grad_norm": 4.25, "learning_rate": 1.2134102161968775e-05, "loss": 2.5069, "step": 2061 }, { "epoch": 2.4403669724770642, "grad_norm": 4.53125, "learning_rate": 1.212602028804028e-05, "loss": 2.4164, "step": 2062 }, { "epoch": 2.4415507546611424, "grad_norm": 4.78125, "learning_rate": 1.211793695946616e-05, "loss": 2.7737, "step": 2063 }, { "epoch": 2.4427345368452205, "grad_norm": 4.9375, "learning_rate": 1.2109852181777117e-05, "loss": 2.6702, "step": 2064 }, { "epoch": 2.4439183190292986, "grad_norm": 4.09375, "learning_rate": 1.210176596050484e-05, "loss": 2.4145, "step": 2065 }, { "epoch": 2.4451021012133767, "grad_norm": 4.25, "learning_rate": 1.2093678301182013e-05, "loss": 1.9222, "step": 2066 }, { "epoch": 2.446285883397455, "grad_norm": 4.46875, "learning_rate": 1.2085589209342287e-05, "loss": 2.617, "step": 2067 }, { "epoch": 2.447469665581533, "grad_norm": 4.5625, "learning_rate": 1.2077498690520316e-05, "loss": 1.9896, "step": 2068 }, { "epoch": 2.448653447765611, "grad_norm": 4.4375, "learning_rate": 1.2069406750251713e-05, "loss": 2.4341, "step": 2069 }, { "epoch": 2.449837229949689, "grad_norm": 4.75, "learning_rate": 1.206131339407307e-05, "loss": 2.2972, "step": 2070 }, { "epoch": 2.451021012133767, "grad_norm": 4.46875, "learning_rate": 1.2053218627521946e-05, "loss": 2.4708, "step": 2071 }, { "epoch": 2.4522047943178453, "grad_norm": 4.5625, "learning_rate": 1.204512245613687e-05, "loss": 2.3367, "step": 2072 }, { "epoch": 2.453388576501924, "grad_norm": 4.8125, "learning_rate": 1.2037024885457325e-05, "loss": 2.1515, "step": 2073 }, { "epoch": 2.4545723586860015, "grad_norm": 4.34375, "learning_rate": 1.2028925921023753e-05, "loss": 2.4118, "step": 2074 }, { "epoch": 2.45575614087008, "grad_norm": 4.78125, "learning_rate": 1.2020825568377556e-05, "loss": 2.5097, "step": 2075 }, { "epoch": 2.456939923054158, "grad_norm": 4.625, "learning_rate": 1.2012723833061077e-05, "loss": 2.0135, "step": 2076 }, { "epoch": 2.4581237052382363, "grad_norm": 4.6875, "learning_rate": 1.2004620720617613e-05, "loss": 2.5324, "step": 2077 }, { "epoch": 2.4593074874223144, "grad_norm": 4.65625, "learning_rate": 1.1996516236591398e-05, "loss": 2.4326, "step": 2078 }, { "epoch": 2.4604912696063925, "grad_norm": 4.90625, "learning_rate": 1.1988410386527605e-05, "loss": 2.422, "step": 2079 }, { "epoch": 2.4616750517904706, "grad_norm": 4.65625, "learning_rate": 1.1980303175972343e-05, "loss": 2.3809, "step": 2080 }, { "epoch": 2.4628588339745487, "grad_norm": 4.875, "learning_rate": 1.1972194610472655e-05, "loss": 2.388, "step": 2081 }, { "epoch": 2.464042616158627, "grad_norm": 4.4375, "learning_rate": 1.1964084695576496e-05, "loss": 2.3935, "step": 2082 }, { "epoch": 2.465226398342705, "grad_norm": 4.65625, "learning_rate": 1.1955973436832768e-05, "loss": 2.4469, "step": 2083 }, { "epoch": 2.466410180526783, "grad_norm": 4.28125, "learning_rate": 1.1947860839791276e-05, "loss": 2.2801, "step": 2084 }, { "epoch": 2.467593962710861, "grad_norm": 4.625, "learning_rate": 1.1939746910002745e-05, "loss": 2.5286, "step": 2085 }, { "epoch": 2.4687777448949393, "grad_norm": 4.5625, "learning_rate": 1.1931631653018812e-05, "loss": 2.404, "step": 2086 }, { "epoch": 2.4699615270790174, "grad_norm": 4.59375, "learning_rate": 1.1923515074392022e-05, "loss": 2.2529, "step": 2087 }, { "epoch": 2.4711453092630955, "grad_norm": 4.65625, "learning_rate": 1.191539717967582e-05, "loss": 2.7151, "step": 2088 }, { "epoch": 2.4723290914471736, "grad_norm": 3.875, "learning_rate": 1.1907277974424562e-05, "loss": 2.3582, "step": 2089 }, { "epoch": 2.4735128736312517, "grad_norm": 4.46875, "learning_rate": 1.1899157464193493e-05, "loss": 2.4708, "step": 2090 }, { "epoch": 2.47469665581533, "grad_norm": 4.46875, "learning_rate": 1.189103565453875e-05, "loss": 2.1957, "step": 2091 }, { "epoch": 2.475880437999408, "grad_norm": 4.65625, "learning_rate": 1.1882912551017362e-05, "loss": 2.5477, "step": 2092 }, { "epoch": 2.477064220183486, "grad_norm": 4.71875, "learning_rate": 1.1874788159187243e-05, "loss": 2.4484, "step": 2093 }, { "epoch": 2.4782480023675646, "grad_norm": 4.0625, "learning_rate": 1.1866662484607186e-05, "loss": 2.4199, "step": 2094 }, { "epoch": 2.4794317845516423, "grad_norm": 4.0, "learning_rate": 1.1858535532836866e-05, "loss": 2.058, "step": 2095 }, { "epoch": 2.480615566735721, "grad_norm": 4.78125, "learning_rate": 1.1850407309436831e-05, "loss": 2.4588, "step": 2096 }, { "epoch": 2.481799348919799, "grad_norm": 4.5, "learning_rate": 1.1842277819968493e-05, "loss": 2.4086, "step": 2097 }, { "epoch": 2.482983131103877, "grad_norm": 5.0625, "learning_rate": 1.183414706999414e-05, "loss": 2.5456, "step": 2098 }, { "epoch": 2.484166913287955, "grad_norm": 4.375, "learning_rate": 1.1826015065076912e-05, "loss": 2.4892, "step": 2099 }, { "epoch": 2.4853506954720332, "grad_norm": 4.71875, "learning_rate": 1.1817881810780815e-05, "loss": 2.8484, "step": 2100 }, { "epoch": 2.4865344776561114, "grad_norm": 4.90625, "learning_rate": 1.1809747312670711e-05, "loss": 2.474, "step": 2101 }, { "epoch": 2.4877182598401895, "grad_norm": 4.875, "learning_rate": 1.180161157631231e-05, "loss": 2.459, "step": 2102 }, { "epoch": 2.4889020420242676, "grad_norm": 4.5, "learning_rate": 1.1793474607272164e-05, "loss": 2.5983, "step": 2103 }, { "epoch": 2.4900858242083457, "grad_norm": 5.0625, "learning_rate": 1.1785336411117675e-05, "loss": 2.5771, "step": 2104 }, { "epoch": 2.491269606392424, "grad_norm": 5.0625, "learning_rate": 1.1777196993417087e-05, "loss": 2.4599, "step": 2105 }, { "epoch": 2.492453388576502, "grad_norm": 4.59375, "learning_rate": 1.176905635973947e-05, "loss": 2.2652, "step": 2106 }, { "epoch": 2.49363717076058, "grad_norm": 5.1875, "learning_rate": 1.1760914515654735e-05, "loss": 2.2073, "step": 2107 }, { "epoch": 2.494820952944658, "grad_norm": 4.46875, "learning_rate": 1.1752771466733622e-05, "loss": 2.2577, "step": 2108 }, { "epoch": 2.496004735128736, "grad_norm": 4.90625, "learning_rate": 1.1744627218547683e-05, "loss": 2.9351, "step": 2109 }, { "epoch": 2.4971885173128143, "grad_norm": 4.71875, "learning_rate": 1.1736481776669307e-05, "loss": 2.448, "step": 2110 }, { "epoch": 2.4983722994968924, "grad_norm": 4.9375, "learning_rate": 1.1728335146671683e-05, "loss": 2.4413, "step": 2111 }, { "epoch": 2.4995560816809705, "grad_norm": 4.65625, "learning_rate": 1.172018733412883e-05, "loss": 2.672, "step": 2112 }, { "epoch": 2.500739863865049, "grad_norm": 4.6875, "learning_rate": 1.1712038344615564e-05, "loss": 2.5672, "step": 2113 }, { "epoch": 2.5019236460491268, "grad_norm": 5.03125, "learning_rate": 1.1703888183707513e-05, "loss": 2.578, "step": 2114 }, { "epoch": 2.5031074282332053, "grad_norm": 4.59375, "learning_rate": 1.1695736856981097e-05, "loss": 2.6862, "step": 2115 }, { "epoch": 2.504291210417283, "grad_norm": 4.21875, "learning_rate": 1.1687584370013545e-05, "loss": 1.9629, "step": 2116 }, { "epoch": 2.5054749926013615, "grad_norm": 4.6875, "learning_rate": 1.1679430728382875e-05, "loss": 2.6615, "step": 2117 }, { "epoch": 2.5066587747854396, "grad_norm": 4.25, "learning_rate": 1.1671275937667895e-05, "loss": 2.3438, "step": 2118 }, { "epoch": 2.5078425569695177, "grad_norm": 4.28125, "learning_rate": 1.1663120003448195e-05, "loss": 2.0126, "step": 2119 }, { "epoch": 2.509026339153596, "grad_norm": 4.5, "learning_rate": 1.1654962931304159e-05, "loss": 2.4617, "step": 2120 }, { "epoch": 2.510210121337674, "grad_norm": 4.84375, "learning_rate": 1.1646804726816937e-05, "loss": 2.6137, "step": 2121 }, { "epoch": 2.511393903521752, "grad_norm": 4.625, "learning_rate": 1.1638645395568458e-05, "loss": 2.2657, "step": 2122 }, { "epoch": 2.51257768570583, "grad_norm": 5.1875, "learning_rate": 1.1630484943141428e-05, "loss": 2.3597, "step": 2123 }, { "epoch": 2.5137614678899083, "grad_norm": 4.3125, "learning_rate": 1.1622323375119312e-05, "loss": 2.3877, "step": 2124 }, { "epoch": 2.5149452500739864, "grad_norm": 4.875, "learning_rate": 1.161416069708634e-05, "loss": 1.9127, "step": 2125 }, { "epoch": 2.5161290322580645, "grad_norm": 4.6875, "learning_rate": 1.1605996914627509e-05, "loss": 2.5164, "step": 2126 }, { "epoch": 2.5173128144421426, "grad_norm": 5.0, "learning_rate": 1.1597832033328558e-05, "loss": 2.0658, "step": 2127 }, { "epoch": 2.5184965966262207, "grad_norm": 4.625, "learning_rate": 1.1589666058775986e-05, "loss": 2.5354, "step": 2128 }, { "epoch": 2.519680378810299, "grad_norm": 4.3125, "learning_rate": 1.1581498996557046e-05, "loss": 1.9649, "step": 2129 }, { "epoch": 2.520864160994377, "grad_norm": 4.4375, "learning_rate": 1.1573330852259724e-05, "loss": 2.5577, "step": 2130 }, { "epoch": 2.522047943178455, "grad_norm": 4.25, "learning_rate": 1.156516163147275e-05, "loss": 2.5916, "step": 2131 }, { "epoch": 2.523231725362533, "grad_norm": 4.40625, "learning_rate": 1.1556991339785595e-05, "loss": 2.3876, "step": 2132 }, { "epoch": 2.5244155075466113, "grad_norm": 4.75, "learning_rate": 1.1548819982788455e-05, "loss": 2.3475, "step": 2133 }, { "epoch": 2.52559928973069, "grad_norm": 5.25, "learning_rate": 1.1540647566072259e-05, "loss": 2.785, "step": 2134 }, { "epoch": 2.5267830719147675, "grad_norm": 5.09375, "learning_rate": 1.1532474095228667e-05, "loss": 2.51, "step": 2135 }, { "epoch": 2.527966854098846, "grad_norm": 4.6875, "learning_rate": 1.1524299575850048e-05, "loss": 2.3989, "step": 2136 }, { "epoch": 2.5291506362829237, "grad_norm": 4.5625, "learning_rate": 1.15161240135295e-05, "loss": 2.5095, "step": 2137 }, { "epoch": 2.5303344184670022, "grad_norm": 4.375, "learning_rate": 1.1507947413860827e-05, "loss": 2.6041, "step": 2138 }, { "epoch": 2.5315182006510804, "grad_norm": 5.0, "learning_rate": 1.1499769782438547e-05, "loss": 2.6412, "step": 2139 }, { "epoch": 2.5327019828351585, "grad_norm": 4.40625, "learning_rate": 1.1491591124857873e-05, "loss": 2.4505, "step": 2140 }, { "epoch": 2.5338857650192366, "grad_norm": 4.78125, "learning_rate": 1.1483411446714744e-05, "loss": 2.6604, "step": 2141 }, { "epoch": 2.5350695472033147, "grad_norm": 4.71875, "learning_rate": 1.147523075360577e-05, "loss": 2.4831, "step": 2142 }, { "epoch": 2.536253329387393, "grad_norm": 4.6875, "learning_rate": 1.1467049051128275e-05, "loss": 2.149, "step": 2143 }, { "epoch": 2.537437111571471, "grad_norm": 4.5, "learning_rate": 1.1458866344880266e-05, "loss": 2.4218, "step": 2144 }, { "epoch": 2.538620893755549, "grad_norm": 5.5, "learning_rate": 1.1450682640460434e-05, "loss": 2.4921, "step": 2145 }, { "epoch": 2.539804675939627, "grad_norm": 4.375, "learning_rate": 1.1442497943468158e-05, "loss": 2.4612, "step": 2146 }, { "epoch": 2.540988458123705, "grad_norm": 5.0625, "learning_rate": 1.1434312259503498e-05, "loss": 2.6346, "step": 2147 }, { "epoch": 2.5421722403077833, "grad_norm": 4.6875, "learning_rate": 1.1426125594167179e-05, "loss": 2.8272, "step": 2148 }, { "epoch": 2.5433560224918614, "grad_norm": 4.5, "learning_rate": 1.1417937953060607e-05, "loss": 2.2427, "step": 2149 }, { "epoch": 2.5445398046759395, "grad_norm": 5.0, "learning_rate": 1.1409749341785859e-05, "loss": 2.4477, "step": 2150 }, { "epoch": 2.5457235868600177, "grad_norm": 4.78125, "learning_rate": 1.1401559765945657e-05, "loss": 2.6471, "step": 2151 }, { "epoch": 2.5469073690440958, "grad_norm": 4.40625, "learning_rate": 1.1393369231143406e-05, "loss": 2.2792, "step": 2152 }, { "epoch": 2.548091151228174, "grad_norm": 4.78125, "learning_rate": 1.1385177742983155e-05, "loss": 2.6679, "step": 2153 }, { "epoch": 2.549274933412252, "grad_norm": 4.34375, "learning_rate": 1.1376985307069606e-05, "loss": 2.5152, "step": 2154 }, { "epoch": 2.5504587155963305, "grad_norm": 4.8125, "learning_rate": 1.1368791929008107e-05, "loss": 2.7993, "step": 2155 }, { "epoch": 2.551642497780408, "grad_norm": 4.75, "learning_rate": 1.1360597614404663e-05, "loss": 2.542, "step": 2156 }, { "epoch": 2.5528262799644867, "grad_norm": 5.375, "learning_rate": 1.1352402368865898e-05, "loss": 2.3704, "step": 2157 }, { "epoch": 2.5540100621485644, "grad_norm": 5.21875, "learning_rate": 1.1344206197999095e-05, "loss": 2.3734, "step": 2158 }, { "epoch": 2.555193844332643, "grad_norm": 4.625, "learning_rate": 1.1336009107412162e-05, "loss": 2.6098, "step": 2159 }, { "epoch": 2.556377626516721, "grad_norm": 4.5625, "learning_rate": 1.1327811102713633e-05, "loss": 2.572, "step": 2160 }, { "epoch": 2.557561408700799, "grad_norm": 4.4375, "learning_rate": 1.1319612189512665e-05, "loss": 2.703, "step": 2161 }, { "epoch": 2.5587451908848773, "grad_norm": 5.0625, "learning_rate": 1.1311412373419051e-05, "loss": 2.7641, "step": 2162 }, { "epoch": 2.5599289730689554, "grad_norm": 4.65625, "learning_rate": 1.1303211660043181e-05, "loss": 2.4891, "step": 2163 }, { "epoch": 2.5611127552530335, "grad_norm": 4.78125, "learning_rate": 1.1295010054996079e-05, "loss": 2.205, "step": 2164 }, { "epoch": 2.5622965374371116, "grad_norm": 4.84375, "learning_rate": 1.1286807563889366e-05, "loss": 2.8578, "step": 2165 }, { "epoch": 2.5634803196211897, "grad_norm": 4.34375, "learning_rate": 1.1278604192335273e-05, "loss": 2.2615, "step": 2166 }, { "epoch": 2.564664101805268, "grad_norm": 4.5, "learning_rate": 1.1270399945946638e-05, "loss": 2.4207, "step": 2167 }, { "epoch": 2.565847883989346, "grad_norm": 4.46875, "learning_rate": 1.1262194830336888e-05, "loss": 2.3258, "step": 2168 }, { "epoch": 2.567031666173424, "grad_norm": 5.3125, "learning_rate": 1.125398885112005e-05, "loss": 2.3561, "step": 2169 }, { "epoch": 2.568215448357502, "grad_norm": 4.875, "learning_rate": 1.1245782013910748e-05, "loss": 2.5897, "step": 2170 }, { "epoch": 2.5693992305415803, "grad_norm": 4.53125, "learning_rate": 1.1237574324324185e-05, "loss": 2.5087, "step": 2171 }, { "epoch": 2.5705830127256584, "grad_norm": 4.625, "learning_rate": 1.1229365787976145e-05, "loss": 2.483, "step": 2172 }, { "epoch": 2.5717667949097365, "grad_norm": 4.6875, "learning_rate": 1.1221156410482998e-05, "loss": 2.6365, "step": 2173 }, { "epoch": 2.5729505770938146, "grad_norm": 4.90625, "learning_rate": 1.1212946197461688e-05, "loss": 2.5512, "step": 2174 }, { "epoch": 2.5741343592778927, "grad_norm": 4.71875, "learning_rate": 1.1204735154529728e-05, "loss": 2.3252, "step": 2175 }, { "epoch": 2.5753181414619712, "grad_norm": 4.375, "learning_rate": 1.1196523287305204e-05, "loss": 2.1852, "step": 2176 }, { "epoch": 2.576501923646049, "grad_norm": 4.53125, "learning_rate": 1.118831060140676e-05, "loss": 2.6461, "step": 2177 }, { "epoch": 2.5776857058301275, "grad_norm": 4.65625, "learning_rate": 1.1180097102453604e-05, "loss": 2.5046, "step": 2178 }, { "epoch": 2.578869488014205, "grad_norm": 5.25, "learning_rate": 1.1171882796065496e-05, "loss": 2.5001, "step": 2179 }, { "epoch": 2.5800532701982837, "grad_norm": 4.34375, "learning_rate": 1.1163667687862755e-05, "loss": 2.366, "step": 2180 }, { "epoch": 2.581237052382362, "grad_norm": 4.875, "learning_rate": 1.1155451783466244e-05, "loss": 2.7373, "step": 2181 }, { "epoch": 2.58242083456644, "grad_norm": 4.96875, "learning_rate": 1.1147235088497372e-05, "loss": 2.7343, "step": 2182 }, { "epoch": 2.583604616750518, "grad_norm": 4.5, "learning_rate": 1.1139017608578088e-05, "loss": 2.623, "step": 2183 }, { "epoch": 2.584788398934596, "grad_norm": 4.84375, "learning_rate": 1.1130799349330881e-05, "loss": 2.5351, "step": 2184 }, { "epoch": 2.585972181118674, "grad_norm": 4.4375, "learning_rate": 1.1122580316378769e-05, "loss": 2.5278, "step": 2185 }, { "epoch": 2.5871559633027523, "grad_norm": 4.65625, "learning_rate": 1.1114360515345301e-05, "loss": 2.4142, "step": 2186 }, { "epoch": 2.5883397454868304, "grad_norm": 4.78125, "learning_rate": 1.1106139951854555e-05, "loss": 2.4415, "step": 2187 }, { "epoch": 2.5895235276709085, "grad_norm": 4.71875, "learning_rate": 1.1097918631531123e-05, "loss": 2.7461, "step": 2188 }, { "epoch": 2.5907073098549867, "grad_norm": 4.90625, "learning_rate": 1.1089696560000128e-05, "loss": 2.5119, "step": 2189 }, { "epoch": 2.5918910920390648, "grad_norm": 4.625, "learning_rate": 1.108147374288719e-05, "loss": 2.3121, "step": 2190 }, { "epoch": 2.593074874223143, "grad_norm": 4.53125, "learning_rate": 1.1073250185818449e-05, "loss": 2.6285, "step": 2191 }, { "epoch": 2.594258656407221, "grad_norm": 4.625, "learning_rate": 1.1065025894420551e-05, "loss": 2.6171, "step": 2192 }, { "epoch": 2.595442438591299, "grad_norm": 4.71875, "learning_rate": 1.1056800874320646e-05, "loss": 2.1678, "step": 2193 }, { "epoch": 2.596626220775377, "grad_norm": 4.625, "learning_rate": 1.1048575131146377e-05, "loss": 2.2262, "step": 2194 }, { "epoch": 2.5978100029594553, "grad_norm": 4.4375, "learning_rate": 1.1040348670525889e-05, "loss": 2.4138, "step": 2195 }, { "epoch": 2.5989937851435334, "grad_norm": 4.8125, "learning_rate": 1.1032121498087807e-05, "loss": 2.3612, "step": 2196 }, { "epoch": 2.600177567327612, "grad_norm": 4.125, "learning_rate": 1.1023893619461253e-05, "loss": 2.2043, "step": 2197 }, { "epoch": 2.6013613495116896, "grad_norm": 4.90625, "learning_rate": 1.1015665040275827e-05, "loss": 2.3735, "step": 2198 }, { "epoch": 2.602545131695768, "grad_norm": 4.625, "learning_rate": 1.1007435766161612e-05, "loss": 2.2058, "step": 2199 }, { "epoch": 2.603728913879846, "grad_norm": 4.375, "learning_rate": 1.0999205802749165e-05, "loss": 2.2881, "step": 2200 }, { "epoch": 2.6049126960639244, "grad_norm": 4.21875, "learning_rate": 1.0990975155669516e-05, "loss": 2.6506, "step": 2201 }, { "epoch": 2.6060964782480025, "grad_norm": 4.4375, "learning_rate": 1.0982743830554157e-05, "loss": 2.4836, "step": 2202 }, { "epoch": 2.6072802604320806, "grad_norm": 4.875, "learning_rate": 1.0974511833035048e-05, "loss": 2.6941, "step": 2203 }, { "epoch": 2.6084640426161587, "grad_norm": 5.15625, "learning_rate": 1.096627916874461e-05, "loss": 2.3966, "step": 2204 }, { "epoch": 2.609647824800237, "grad_norm": 4.28125, "learning_rate": 1.0958045843315722e-05, "loss": 2.5606, "step": 2205 }, { "epoch": 2.610831606984315, "grad_norm": 4.5625, "learning_rate": 1.0949811862381708e-05, "loss": 2.3439, "step": 2206 }, { "epoch": 2.612015389168393, "grad_norm": 4.4375, "learning_rate": 1.094157723157635e-05, "loss": 2.2478, "step": 2207 }, { "epoch": 2.613199171352471, "grad_norm": 4.84375, "learning_rate": 1.0933341956533865e-05, "loss": 3.0881, "step": 2208 }, { "epoch": 2.6143829535365493, "grad_norm": 4.53125, "learning_rate": 1.0925106042888915e-05, "loss": 2.4804, "step": 2209 }, { "epoch": 2.6155667357206274, "grad_norm": 4.5, "learning_rate": 1.0916869496276604e-05, "loss": 2.6077, "step": 2210 }, { "epoch": 2.6167505179047055, "grad_norm": 4.40625, "learning_rate": 1.0908632322332464e-05, "loss": 2.2322, "step": 2211 }, { "epoch": 2.6179343000887836, "grad_norm": 4.8125, "learning_rate": 1.0900394526692453e-05, "loss": 2.5108, "step": 2212 }, { "epoch": 2.6191180822728617, "grad_norm": 4.5, "learning_rate": 1.0892156114992963e-05, "loss": 2.548, "step": 2213 }, { "epoch": 2.62030186445694, "grad_norm": 4.875, "learning_rate": 1.0883917092870797e-05, "loss": 2.6502, "step": 2214 }, { "epoch": 2.621485646641018, "grad_norm": 4.53125, "learning_rate": 1.0875677465963186e-05, "loss": 2.1164, "step": 2215 }, { "epoch": 2.622669428825096, "grad_norm": 4.59375, "learning_rate": 1.0867437239907766e-05, "loss": 2.338, "step": 2216 }, { "epoch": 2.623853211009174, "grad_norm": 5.25, "learning_rate": 1.085919642034259e-05, "loss": 2.5516, "step": 2217 }, { "epoch": 2.6250369931932527, "grad_norm": 4.21875, "learning_rate": 1.0850955012906114e-05, "loss": 2.4052, "step": 2218 }, { "epoch": 2.6262207753773303, "grad_norm": 4.59375, "learning_rate": 1.08427130232372e-05, "loss": 2.5511, "step": 2219 }, { "epoch": 2.627404557561409, "grad_norm": 4.15625, "learning_rate": 1.0834470456975092e-05, "loss": 2.5787, "step": 2220 }, { "epoch": 2.6285883397454866, "grad_norm": 5.0625, "learning_rate": 1.0826227319759458e-05, "loss": 2.4521, "step": 2221 }, { "epoch": 2.629772121929565, "grad_norm": 4.71875, "learning_rate": 1.0817983617230326e-05, "loss": 2.555, "step": 2222 }, { "epoch": 2.630955904113643, "grad_norm": 4.4375, "learning_rate": 1.080973935502813e-05, "loss": 2.3562, "step": 2223 }, { "epoch": 2.6321396862977213, "grad_norm": 4.34375, "learning_rate": 1.0801494538793684e-05, "loss": 2.0583, "step": 2224 }, { "epoch": 2.6333234684817994, "grad_norm": 4.5, "learning_rate": 1.079324917416818e-05, "loss": 2.4229, "step": 2225 }, { "epoch": 2.6345072506658775, "grad_norm": 5.09375, "learning_rate": 1.0785003266793173e-05, "loss": 2.8452, "step": 2226 }, { "epoch": 2.6356910328499557, "grad_norm": 4.65625, "learning_rate": 1.0776756822310615e-05, "loss": 2.4981, "step": 2227 }, { "epoch": 2.6368748150340338, "grad_norm": 4.71875, "learning_rate": 1.0768509846362798e-05, "loss": 2.4662, "step": 2228 }, { "epoch": 2.638058597218112, "grad_norm": 4.3125, "learning_rate": 1.0760262344592399e-05, "loss": 2.7941, "step": 2229 }, { "epoch": 2.63924237940219, "grad_norm": 4.59375, "learning_rate": 1.075201432264244e-05, "loss": 2.3297, "step": 2230 }, { "epoch": 2.640426161586268, "grad_norm": 4.65625, "learning_rate": 1.0743765786156313e-05, "loss": 2.5938, "step": 2231 }, { "epoch": 2.641609943770346, "grad_norm": 4.9375, "learning_rate": 1.0735516740777742e-05, "loss": 2.6882, "step": 2232 }, { "epoch": 2.6427937259544243, "grad_norm": 4.53125, "learning_rate": 1.0727267192150825e-05, "loss": 2.2164, "step": 2233 }, { "epoch": 2.6439775081385024, "grad_norm": 4.59375, "learning_rate": 1.0719017145919984e-05, "loss": 2.4794, "step": 2234 }, { "epoch": 2.6451612903225805, "grad_norm": 4.75, "learning_rate": 1.0710766607729983e-05, "loss": 2.5018, "step": 2235 }, { "epoch": 2.6463450725066586, "grad_norm": 4.71875, "learning_rate": 1.0702515583225936e-05, "loss": 2.3054, "step": 2236 }, { "epoch": 2.647528854690737, "grad_norm": 5.21875, "learning_rate": 1.069426407805328e-05, "loss": 2.5062, "step": 2237 }, { "epoch": 2.648712636874815, "grad_norm": 5.5625, "learning_rate": 1.068601209785778e-05, "loss": 2.2485, "step": 2238 }, { "epoch": 2.6498964190588934, "grad_norm": 4.25, "learning_rate": 1.0677759648285529e-05, "loss": 2.5505, "step": 2239 }, { "epoch": 2.651080201242971, "grad_norm": 5.09375, "learning_rate": 1.066950673498294e-05, "loss": 2.5214, "step": 2240 }, { "epoch": 2.6522639834270496, "grad_norm": 4.4375, "learning_rate": 1.0661253363596742e-05, "loss": 2.3392, "step": 2241 }, { "epoch": 2.6534477656111277, "grad_norm": 4.15625, "learning_rate": 1.0652999539773985e-05, "loss": 2.3871, "step": 2242 }, { "epoch": 2.654631547795206, "grad_norm": 4.75, "learning_rate": 1.064474526916202e-05, "loss": 2.4524, "step": 2243 }, { "epoch": 2.655815329979284, "grad_norm": 4.59375, "learning_rate": 1.0636490557408501e-05, "loss": 2.2829, "step": 2244 }, { "epoch": 2.656999112163362, "grad_norm": 4.5625, "learning_rate": 1.0628235410161398e-05, "loss": 2.6885, "step": 2245 }, { "epoch": 2.65818289434744, "grad_norm": 5.0, "learning_rate": 1.0619979833068965e-05, "loss": 2.7395, "step": 2246 }, { "epoch": 2.6593666765315183, "grad_norm": 4.6875, "learning_rate": 1.0611723831779756e-05, "loss": 2.2646, "step": 2247 }, { "epoch": 2.6605504587155964, "grad_norm": 5.15625, "learning_rate": 1.0603467411942618e-05, "loss": 2.784, "step": 2248 }, { "epoch": 2.6617342408996745, "grad_norm": 4.3125, "learning_rate": 1.0595210579206676e-05, "loss": 2.1767, "step": 2249 }, { "epoch": 2.6629180230837526, "grad_norm": 5.34375, "learning_rate": 1.0586953339221346e-05, "loss": 2.7792, "step": 2250 }, { "epoch": 2.6641018052678307, "grad_norm": 4.71875, "learning_rate": 1.057869569763632e-05, "loss": 2.5714, "step": 2251 }, { "epoch": 2.665285587451909, "grad_norm": 4.875, "learning_rate": 1.057043766010156e-05, "loss": 2.3906, "step": 2252 }, { "epoch": 2.666469369635987, "grad_norm": 4.25, "learning_rate": 1.0562179232267304e-05, "loss": 2.6521, "step": 2253 }, { "epoch": 2.667653151820065, "grad_norm": 4.375, "learning_rate": 1.0553920419784056e-05, "loss": 2.6358, "step": 2254 }, { "epoch": 2.668836934004143, "grad_norm": 4.53125, "learning_rate": 1.0545661228302586e-05, "loss": 2.1, "step": 2255 }, { "epoch": 2.6700207161882212, "grad_norm": 3.859375, "learning_rate": 1.0537401663473916e-05, "loss": 2.2394, "step": 2256 }, { "epoch": 2.6712044983722993, "grad_norm": 4.53125, "learning_rate": 1.0529141730949334e-05, "loss": 2.0746, "step": 2257 }, { "epoch": 2.672388280556378, "grad_norm": 4.21875, "learning_rate": 1.0520881436380366e-05, "loss": 2.4515, "step": 2258 }, { "epoch": 2.6735720627404556, "grad_norm": 4.9375, "learning_rate": 1.0512620785418796e-05, "loss": 2.8442, "step": 2259 }, { "epoch": 2.674755844924534, "grad_norm": 4.4375, "learning_rate": 1.0504359783716652e-05, "loss": 2.3171, "step": 2260 }, { "epoch": 2.6759396271086118, "grad_norm": 4.03125, "learning_rate": 1.0496098436926195e-05, "loss": 2.5538, "step": 2261 }, { "epoch": 2.6771234092926903, "grad_norm": 4.71875, "learning_rate": 1.0487836750699924e-05, "loss": 2.6779, "step": 2262 }, { "epoch": 2.6783071914767684, "grad_norm": 4.90625, "learning_rate": 1.0479574730690583e-05, "loss": 2.3859, "step": 2263 }, { "epoch": 2.6794909736608465, "grad_norm": 4.75, "learning_rate": 1.0471312382551121e-05, "loss": 2.8107, "step": 2264 }, { "epoch": 2.6806747558449247, "grad_norm": 4.375, "learning_rate": 1.0463049711934731e-05, "loss": 2.3393, "step": 2265 }, { "epoch": 2.6818585380290028, "grad_norm": 4.34375, "learning_rate": 1.0454786724494819e-05, "loss": 2.1809, "step": 2266 }, { "epoch": 2.683042320213081, "grad_norm": 4.6875, "learning_rate": 1.0446523425885008e-05, "loss": 2.5745, "step": 2267 }, { "epoch": 2.684226102397159, "grad_norm": 4.6875, "learning_rate": 1.0438259821759133e-05, "loss": 2.6838, "step": 2268 }, { "epoch": 2.685409884581237, "grad_norm": 4.9375, "learning_rate": 1.0429995917771247e-05, "loss": 2.7722, "step": 2269 }, { "epoch": 2.686593666765315, "grad_norm": 4.5625, "learning_rate": 1.042173171957559e-05, "loss": 2.4263, "step": 2270 }, { "epoch": 2.6877774489493933, "grad_norm": 4.75, "learning_rate": 1.041346723282662e-05, "loss": 2.3291, "step": 2271 }, { "epoch": 2.6889612311334714, "grad_norm": 4.90625, "learning_rate": 1.0405202463178985e-05, "loss": 2.2318, "step": 2272 }, { "epoch": 2.6901450133175495, "grad_norm": 4.6875, "learning_rate": 1.0396937416287527e-05, "loss": 2.2387, "step": 2273 }, { "epoch": 2.6913287955016276, "grad_norm": 4.8125, "learning_rate": 1.0388672097807282e-05, "loss": 3.0395, "step": 2274 }, { "epoch": 2.6925125776857057, "grad_norm": 4.5, "learning_rate": 1.0380406513393469e-05, "loss": 2.431, "step": 2275 }, { "epoch": 2.693696359869784, "grad_norm": 4.90625, "learning_rate": 1.0372140668701483e-05, "loss": 2.4743, "step": 2276 }, { "epoch": 2.694880142053862, "grad_norm": 4.46875, "learning_rate": 1.0363874569386907e-05, "loss": 2.2859, "step": 2277 }, { "epoch": 2.69606392423794, "grad_norm": 4.1875, "learning_rate": 1.035560822110549e-05, "loss": 2.3715, "step": 2278 }, { "epoch": 2.6972477064220186, "grad_norm": 4.59375, "learning_rate": 1.034734162951316e-05, "loss": 2.434, "step": 2279 }, { "epoch": 2.6984314886060963, "grad_norm": 4.59375, "learning_rate": 1.0339074800266005e-05, "loss": 2.0274, "step": 2280 }, { "epoch": 2.699615270790175, "grad_norm": 4.28125, "learning_rate": 1.0330807739020278e-05, "loss": 2.2388, "step": 2281 }, { "epoch": 2.7007990529742525, "grad_norm": 4.75, "learning_rate": 1.0322540451432385e-05, "loss": 2.3483, "step": 2282 }, { "epoch": 2.701982835158331, "grad_norm": 4.15625, "learning_rate": 1.0314272943158899e-05, "loss": 2.136, "step": 2283 }, { "epoch": 2.703166617342409, "grad_norm": 4.28125, "learning_rate": 1.030600521985653e-05, "loss": 2.374, "step": 2284 }, { "epoch": 2.7043503995264873, "grad_norm": 4.71875, "learning_rate": 1.0297737287182144e-05, "loss": 2.2558, "step": 2285 }, { "epoch": 2.7055341817105654, "grad_norm": 4.375, "learning_rate": 1.0289469150792753e-05, "loss": 2.2155, "step": 2286 }, { "epoch": 2.7067179638946435, "grad_norm": 4.125, "learning_rate": 1.0281200816345498e-05, "loss": 2.4106, "step": 2287 }, { "epoch": 2.7079017460787216, "grad_norm": 5.25, "learning_rate": 1.0272932289497664e-05, "loss": 2.6136, "step": 2288 }, { "epoch": 2.7090855282627997, "grad_norm": 4.625, "learning_rate": 1.0264663575906661e-05, "loss": 2.266, "step": 2289 }, { "epoch": 2.710269310446878, "grad_norm": 4.46875, "learning_rate": 1.0256394681230035e-05, "loss": 2.3676, "step": 2290 }, { "epoch": 2.711453092630956, "grad_norm": 4.84375, "learning_rate": 1.024812561112545e-05, "loss": 2.549, "step": 2291 }, { "epoch": 2.712636874815034, "grad_norm": 5.1875, "learning_rate": 1.023985637125069e-05, "loss": 2.8416, "step": 2292 }, { "epoch": 2.713820656999112, "grad_norm": 4.71875, "learning_rate": 1.023158696726366e-05, "loss": 2.889, "step": 2293 }, { "epoch": 2.7150044391831902, "grad_norm": 4.6875, "learning_rate": 1.022331740482237e-05, "loss": 2.6952, "step": 2294 }, { "epoch": 2.7161882213672683, "grad_norm": 4.375, "learning_rate": 1.0215047689584943e-05, "loss": 2.2556, "step": 2295 }, { "epoch": 2.7173720035513464, "grad_norm": 4.78125, "learning_rate": 1.0206777827209609e-05, "loss": 2.6879, "step": 2296 }, { "epoch": 2.7185557857354246, "grad_norm": 4.875, "learning_rate": 1.0198507823354692e-05, "loss": 2.5995, "step": 2297 }, { "epoch": 2.7197395679195027, "grad_norm": 4.6875, "learning_rate": 1.0190237683678614e-05, "loss": 2.3361, "step": 2298 }, { "epoch": 2.7209233501035808, "grad_norm": 4.625, "learning_rate": 1.0181967413839901e-05, "loss": 2.2356, "step": 2299 }, { "epoch": 2.7221071322876593, "grad_norm": 4.6875, "learning_rate": 1.0173697019497152e-05, "loss": 2.5006, "step": 2300 }, { "epoch": 2.723290914471737, "grad_norm": 4.25, "learning_rate": 1.0165426506309059e-05, "loss": 2.482, "step": 2301 }, { "epoch": 2.7244746966558155, "grad_norm": 4.625, "learning_rate": 1.0157155879934399e-05, "loss": 2.4843, "step": 2302 }, { "epoch": 2.725658478839893, "grad_norm": 5.03125, "learning_rate": 1.014888514603202e-05, "loss": 2.2904, "step": 2303 }, { "epoch": 2.7268422610239718, "grad_norm": 4.90625, "learning_rate": 1.0140614310260844e-05, "loss": 2.807, "step": 2304 }, { "epoch": 2.72802604320805, "grad_norm": 4.71875, "learning_rate": 1.0132343378279868e-05, "loss": 2.5591, "step": 2305 }, { "epoch": 2.729209825392128, "grad_norm": 4.28125, "learning_rate": 1.012407235574815e-05, "loss": 2.3692, "step": 2306 }, { "epoch": 2.730393607576206, "grad_norm": 4.875, "learning_rate": 1.0115801248324803e-05, "loss": 2.5479, "step": 2307 }, { "epoch": 2.731577389760284, "grad_norm": 4.5625, "learning_rate": 1.0107530061669021e-05, "loss": 2.586, "step": 2308 }, { "epoch": 2.7327611719443623, "grad_norm": 5.15625, "learning_rate": 1.009925880144003e-05, "loss": 2.5653, "step": 2309 }, { "epoch": 2.7339449541284404, "grad_norm": 5.03125, "learning_rate": 1.0090987473297113e-05, "loss": 2.6549, "step": 2310 }, { "epoch": 2.7351287363125185, "grad_norm": 5.09375, "learning_rate": 1.0082716082899605e-05, "loss": 2.7236, "step": 2311 }, { "epoch": 2.7363125184965966, "grad_norm": 4.90625, "learning_rate": 1.0074444635906875e-05, "loss": 2.6439, "step": 2312 }, { "epoch": 2.7374963006806747, "grad_norm": 4.71875, "learning_rate": 1.0066173137978333e-05, "loss": 2.2053, "step": 2313 }, { "epoch": 2.738680082864753, "grad_norm": 4.09375, "learning_rate": 1.0057901594773432e-05, "loss": 2.1171, "step": 2314 }, { "epoch": 2.739863865048831, "grad_norm": 4.65625, "learning_rate": 1.0049630011951643e-05, "loss": 2.3331, "step": 2315 }, { "epoch": 2.741047647232909, "grad_norm": 4.5625, "learning_rate": 1.0041358395172473e-05, "loss": 2.2266, "step": 2316 }, { "epoch": 2.742231429416987, "grad_norm": 5.34375, "learning_rate": 1.0033086750095455e-05, "loss": 2.9613, "step": 2317 }, { "epoch": 2.7434152116010653, "grad_norm": 5.03125, "learning_rate": 1.0024815082380131e-05, "loss": 2.4671, "step": 2318 }, { "epoch": 2.7445989937851434, "grad_norm": 4.59375, "learning_rate": 1.0016543397686059e-05, "loss": 2.5058, "step": 2319 }, { "epoch": 2.7457827759692215, "grad_norm": 3.90625, "learning_rate": 1.0008271701672825e-05, "loss": 2.5076, "step": 2320 }, { "epoch": 2.7469665581533, "grad_norm": 4.59375, "learning_rate": 1e-05, "loss": 2.4264, "step": 2321 }, { "epoch": 2.7481503403373777, "grad_norm": 3.6875, "learning_rate": 9.991728298327179e-06, "loss": 2.2325, "step": 2322 }, { "epoch": 2.7493341225214563, "grad_norm": 5.0, "learning_rate": 9.983456602313944e-06, "loss": 2.4974, "step": 2323 }, { "epoch": 2.750517904705534, "grad_norm": 4.59375, "learning_rate": 9.975184917619874e-06, "loss": 2.3579, "step": 2324 }, { "epoch": 2.7517016868896125, "grad_norm": 4.5625, "learning_rate": 9.96691324990455e-06, "loss": 2.2826, "step": 2325 }, { "epoch": 2.7528854690736906, "grad_norm": 4.9375, "learning_rate": 9.958641604827529e-06, "loss": 2.0722, "step": 2326 }, { "epoch": 2.7540692512577687, "grad_norm": 4.3125, "learning_rate": 9.950369988048357e-06, "loss": 2.729, "step": 2327 }, { "epoch": 2.755253033441847, "grad_norm": 4.34375, "learning_rate": 9.942098405226571e-06, "loss": 2.1514, "step": 2328 }, { "epoch": 2.756436815625925, "grad_norm": 4.375, "learning_rate": 9.933826862021669e-06, "loss": 2.4662, "step": 2329 }, { "epoch": 2.757620597810003, "grad_norm": 4.6875, "learning_rate": 9.92555536409313e-06, "loss": 2.2715, "step": 2330 }, { "epoch": 2.758804379994081, "grad_norm": 4.75, "learning_rate": 9.917283917100396e-06, "loss": 2.6139, "step": 2331 }, { "epoch": 2.7599881621781592, "grad_norm": 4.34375, "learning_rate": 9.909012526702889e-06, "loss": 2.3516, "step": 2332 }, { "epoch": 2.7611719443622373, "grad_norm": 4.6875, "learning_rate": 9.900741198559971e-06, "loss": 2.2936, "step": 2333 }, { "epoch": 2.7623557265463154, "grad_norm": 4.25, "learning_rate": 9.89246993833098e-06, "loss": 2.4982, "step": 2334 }, { "epoch": 2.7635395087303936, "grad_norm": 4.4375, "learning_rate": 9.884198751675198e-06, "loss": 2.3823, "step": 2335 }, { "epoch": 2.7647232909144717, "grad_norm": 4.46875, "learning_rate": 9.875927644251856e-06, "loss": 2.1579, "step": 2336 }, { "epoch": 2.7659070730985498, "grad_norm": 5.0, "learning_rate": 9.867656621720137e-06, "loss": 2.5832, "step": 2337 }, { "epoch": 2.767090855282628, "grad_norm": 4.75, "learning_rate": 9.859385689739157e-06, "loss": 2.6344, "step": 2338 }, { "epoch": 2.768274637466706, "grad_norm": 4.875, "learning_rate": 9.85111485396798e-06, "loss": 2.3965, "step": 2339 }, { "epoch": 2.769458419650784, "grad_norm": 4.71875, "learning_rate": 9.842844120065603e-06, "loss": 2.205, "step": 2340 }, { "epoch": 2.770642201834862, "grad_norm": 4.75, "learning_rate": 9.834573493690941e-06, "loss": 2.3776, "step": 2341 }, { "epoch": 2.7718259840189408, "grad_norm": 4.46875, "learning_rate": 9.826302980502853e-06, "loss": 2.2229, "step": 2342 }, { "epoch": 2.7730097662030184, "grad_norm": 5.125, "learning_rate": 9.8180325861601e-06, "loss": 2.5135, "step": 2343 }, { "epoch": 2.774193548387097, "grad_norm": 4.875, "learning_rate": 9.809762316321388e-06, "loss": 2.4335, "step": 2344 }, { "epoch": 2.7753773305711746, "grad_norm": 4.3125, "learning_rate": 9.801492176645313e-06, "loss": 2.2806, "step": 2345 }, { "epoch": 2.776561112755253, "grad_norm": 4.625, "learning_rate": 9.793222172790396e-06, "loss": 2.5608, "step": 2346 }, { "epoch": 2.7777448949393313, "grad_norm": 5.25, "learning_rate": 9.784952310415062e-06, "loss": 2.5284, "step": 2347 }, { "epoch": 2.7789286771234094, "grad_norm": 4.46875, "learning_rate": 9.776682595177633e-06, "loss": 2.3928, "step": 2348 }, { "epoch": 2.7801124593074875, "grad_norm": 4.875, "learning_rate": 9.768413032736345e-06, "loss": 2.366, "step": 2349 }, { "epoch": 2.7812962414915656, "grad_norm": 5.21875, "learning_rate": 9.760143628749312e-06, "loss": 2.5614, "step": 2350 }, { "epoch": 2.7824800236756437, "grad_norm": 4.65625, "learning_rate": 9.75187438887455e-06, "loss": 2.6084, "step": 2351 }, { "epoch": 2.783663805859722, "grad_norm": 4.90625, "learning_rate": 9.743605318769967e-06, "loss": 2.3517, "step": 2352 }, { "epoch": 2.7848475880438, "grad_norm": 4.03125, "learning_rate": 9.735336424093342e-06, "loss": 2.2319, "step": 2353 }, { "epoch": 2.786031370227878, "grad_norm": 4.96875, "learning_rate": 9.727067710502341e-06, "loss": 2.4409, "step": 2354 }, { "epoch": 2.787215152411956, "grad_norm": 4.84375, "learning_rate": 9.718799183654505e-06, "loss": 2.7257, "step": 2355 }, { "epoch": 2.7883989345960343, "grad_norm": 4.21875, "learning_rate": 9.710530849207249e-06, "loss": 2.3073, "step": 2356 }, { "epoch": 2.7895827167801124, "grad_norm": 4.75, "learning_rate": 9.702262712817857e-06, "loss": 2.3967, "step": 2357 }, { "epoch": 2.7907664989641905, "grad_norm": 5.15625, "learning_rate": 9.693994780143475e-06, "loss": 2.6722, "step": 2358 }, { "epoch": 2.7919502811482686, "grad_norm": 4.1875, "learning_rate": 9.685727056841108e-06, "loss": 2.2356, "step": 2359 }, { "epoch": 2.7931340633323467, "grad_norm": 4.25, "learning_rate": 9.677459548567617e-06, "loss": 2.599, "step": 2360 }, { "epoch": 2.794317845516425, "grad_norm": 5.375, "learning_rate": 9.669192260979727e-06, "loss": 2.5658, "step": 2361 }, { "epoch": 2.795501627700503, "grad_norm": 4.875, "learning_rate": 9.660925199733997e-06, "loss": 1.869, "step": 2362 }, { "epoch": 2.7966854098845815, "grad_norm": 4.09375, "learning_rate": 9.65265837048684e-06, "loss": 2.1326, "step": 2363 }, { "epoch": 2.797869192068659, "grad_norm": 4.5625, "learning_rate": 9.64439177889451e-06, "loss": 2.654, "step": 2364 }, { "epoch": 2.7990529742527377, "grad_norm": 4.71875, "learning_rate": 9.636125430613096e-06, "loss": 2.6066, "step": 2365 }, { "epoch": 2.8002367564368154, "grad_norm": 4.875, "learning_rate": 9.627859331298522e-06, "loss": 2.3879, "step": 2366 }, { "epoch": 2.801420538620894, "grad_norm": 4.90625, "learning_rate": 9.619593486606533e-06, "loss": 2.5346, "step": 2367 }, { "epoch": 2.802604320804972, "grad_norm": 5.125, "learning_rate": 9.611327902192718e-06, "loss": 2.9095, "step": 2368 }, { "epoch": 2.80378810298905, "grad_norm": 4.375, "learning_rate": 9.603062583712475e-06, "loss": 2.3351, "step": 2369 }, { "epoch": 2.8049718851731282, "grad_norm": 4.5, "learning_rate": 9.594797536821018e-06, "loss": 2.3234, "step": 2370 }, { "epoch": 2.8061556673572063, "grad_norm": 5.375, "learning_rate": 9.586532767173385e-06, "loss": 2.1862, "step": 2371 }, { "epoch": 2.8073394495412844, "grad_norm": 4.5, "learning_rate": 9.578268280424413e-06, "loss": 2.3135, "step": 2372 }, { "epoch": 2.8085232317253626, "grad_norm": 4.5, "learning_rate": 9.570004082228757e-06, "loss": 2.5411, "step": 2373 }, { "epoch": 2.8097070139094407, "grad_norm": 4.625, "learning_rate": 9.561740178240868e-06, "loss": 2.6365, "step": 2374 }, { "epoch": 2.8108907960935188, "grad_norm": 4.78125, "learning_rate": 9.553476574114993e-06, "loss": 2.6803, "step": 2375 }, { "epoch": 2.812074578277597, "grad_norm": 4.9375, "learning_rate": 9.545213275505183e-06, "loss": 2.2167, "step": 2376 }, { "epoch": 2.813258360461675, "grad_norm": 5.0625, "learning_rate": 9.53695028806527e-06, "loss": 2.3731, "step": 2377 }, { "epoch": 2.814442142645753, "grad_norm": 5.03125, "learning_rate": 9.528687617448884e-06, "loss": 2.7816, "step": 2378 }, { "epoch": 2.815625924829831, "grad_norm": 4.4375, "learning_rate": 9.52042526930942e-06, "loss": 2.6167, "step": 2379 }, { "epoch": 2.8168097070139093, "grad_norm": 4.5625, "learning_rate": 9.512163249300074e-06, "loss": 2.5452, "step": 2380 }, { "epoch": 2.8179934891979874, "grad_norm": 5.34375, "learning_rate": 9.50390156307381e-06, "loss": 2.3962, "step": 2381 }, { "epoch": 2.819177271382066, "grad_norm": 4.875, "learning_rate": 9.495640216283352e-06, "loss": 2.8182, "step": 2382 }, { "epoch": 2.8203610535661436, "grad_norm": 5.1875, "learning_rate": 9.487379214581209e-06, "loss": 2.3253, "step": 2383 }, { "epoch": 2.821544835750222, "grad_norm": 4.34375, "learning_rate": 9.479118563619638e-06, "loss": 2.2548, "step": 2384 }, { "epoch": 2.8227286179343, "grad_norm": 4.34375, "learning_rate": 9.470858269050669e-06, "loss": 2.2546, "step": 2385 }, { "epoch": 2.8239124001183784, "grad_norm": 4.75, "learning_rate": 9.462598336526086e-06, "loss": 2.2748, "step": 2386 }, { "epoch": 2.825096182302456, "grad_norm": 4.40625, "learning_rate": 9.454338771697414e-06, "loss": 2.2588, "step": 2387 }, { "epoch": 2.8262799644865346, "grad_norm": 5.0, "learning_rate": 9.446079580215946e-06, "loss": 2.4415, "step": 2388 }, { "epoch": 2.8274637466706127, "grad_norm": 4.84375, "learning_rate": 9.437820767732699e-06, "loss": 2.6722, "step": 2389 }, { "epoch": 2.828647528854691, "grad_norm": 4.3125, "learning_rate": 9.429562339898446e-06, "loss": 2.5482, "step": 2390 }, { "epoch": 2.829831311038769, "grad_norm": 4.46875, "learning_rate": 9.421304302363683e-06, "loss": 2.4508, "step": 2391 }, { "epoch": 2.831015093222847, "grad_norm": 4.625, "learning_rate": 9.413046660778654e-06, "loss": 2.4059, "step": 2392 }, { "epoch": 2.832198875406925, "grad_norm": 4.65625, "learning_rate": 9.404789420793327e-06, "loss": 2.4456, "step": 2393 }, { "epoch": 2.8333826575910033, "grad_norm": 4.78125, "learning_rate": 9.396532588057384e-06, "loss": 2.6404, "step": 2394 }, { "epoch": 2.8345664397750814, "grad_norm": 4.84375, "learning_rate": 9.388276168220248e-06, "loss": 2.8367, "step": 2395 }, { "epoch": 2.8357502219591595, "grad_norm": 4.96875, "learning_rate": 9.380020166931037e-06, "loss": 2.6085, "step": 2396 }, { "epoch": 2.8369340041432376, "grad_norm": 4.5, "learning_rate": 9.371764589838605e-06, "loss": 2.7481, "step": 2397 }, { "epoch": 2.8381177863273157, "grad_norm": 4.25, "learning_rate": 9.363509442591502e-06, "loss": 2.3981, "step": 2398 }, { "epoch": 2.839301568511394, "grad_norm": 4.65625, "learning_rate": 9.355254730837983e-06, "loss": 2.7307, "step": 2399 }, { "epoch": 2.840485350695472, "grad_norm": 4.71875, "learning_rate": 9.34700046022602e-06, "loss": 2.2642, "step": 2400 }, { "epoch": 2.84166913287955, "grad_norm": 5.5625, "learning_rate": 9.33874663640326e-06, "loss": 2.9404, "step": 2401 }, { "epoch": 2.842852915063628, "grad_norm": 4.65625, "learning_rate": 9.330493265017062e-06, "loss": 2.73, "step": 2402 }, { "epoch": 2.8440366972477067, "grad_norm": 5.34375, "learning_rate": 9.322240351714475e-06, "loss": 2.2542, "step": 2403 }, { "epoch": 2.8452204794317844, "grad_norm": 4.78125, "learning_rate": 9.313987902142222e-06, "loss": 2.5241, "step": 2404 }, { "epoch": 2.846404261615863, "grad_norm": 4.40625, "learning_rate": 9.305735921946724e-06, "loss": 2.3155, "step": 2405 }, { "epoch": 2.8475880437999406, "grad_norm": 4.0, "learning_rate": 9.297484416774066e-06, "loss": 2.311, "step": 2406 }, { "epoch": 2.848771825984019, "grad_norm": 4.78125, "learning_rate": 9.28923339227002e-06, "loss": 2.8111, "step": 2407 }, { "epoch": 2.8499556081680972, "grad_norm": 4.5, "learning_rate": 9.280982854080021e-06, "loss": 2.4221, "step": 2408 }, { "epoch": 2.8511393903521753, "grad_norm": 4.5, "learning_rate": 9.272732807849177e-06, "loss": 2.4401, "step": 2409 }, { "epoch": 2.8523231725362534, "grad_norm": 4.65625, "learning_rate": 9.26448325922226e-06, "loss": 2.5704, "step": 2410 }, { "epoch": 2.8535069547203316, "grad_norm": 4.90625, "learning_rate": 9.25623421384369e-06, "loss": 2.5873, "step": 2411 }, { "epoch": 2.8546907369044097, "grad_norm": 4.71875, "learning_rate": 9.247985677357563e-06, "loss": 2.5537, "step": 2412 }, { "epoch": 2.8558745190884878, "grad_norm": 4.5625, "learning_rate": 9.239737655407605e-06, "loss": 2.3474, "step": 2413 }, { "epoch": 2.857058301272566, "grad_norm": 4.6875, "learning_rate": 9.231490153637203e-06, "loss": 2.7118, "step": 2414 }, { "epoch": 2.858242083456644, "grad_norm": 4.125, "learning_rate": 9.223243177689389e-06, "loss": 2.3488, "step": 2415 }, { "epoch": 2.859425865640722, "grad_norm": 4.84375, "learning_rate": 9.214996733206827e-06, "loss": 2.4184, "step": 2416 }, { "epoch": 2.8606096478248, "grad_norm": 4.21875, "learning_rate": 9.206750825831826e-06, "loss": 2.2223, "step": 2417 }, { "epoch": 2.8617934300088783, "grad_norm": 4.4375, "learning_rate": 9.198505461206318e-06, "loss": 2.2593, "step": 2418 }, { "epoch": 2.8629772121929564, "grad_norm": 4.375, "learning_rate": 9.190260644971874e-06, "loss": 2.5034, "step": 2419 }, { "epoch": 2.8641609943770345, "grad_norm": 4.96875, "learning_rate": 9.182016382769678e-06, "loss": 2.4333, "step": 2420 }, { "epoch": 2.8653447765611126, "grad_norm": 4.84375, "learning_rate": 9.173772680240547e-06, "loss": 2.446, "step": 2421 }, { "epoch": 2.8665285587451907, "grad_norm": 4.84375, "learning_rate": 9.165529543024911e-06, "loss": 2.6337, "step": 2422 }, { "epoch": 2.867712340929269, "grad_norm": 4.40625, "learning_rate": 9.157286976762805e-06, "loss": 2.3994, "step": 2423 }, { "epoch": 2.8688961231133474, "grad_norm": 4.28125, "learning_rate": 9.149044987093887e-06, "loss": 2.4575, "step": 2424 }, { "epoch": 2.870079905297425, "grad_norm": 4.1875, "learning_rate": 9.140803579657413e-06, "loss": 2.0493, "step": 2425 }, { "epoch": 2.8712636874815036, "grad_norm": 4.34375, "learning_rate": 9.132562760092234e-06, "loss": 2.0269, "step": 2426 }, { "epoch": 2.8724474696655813, "grad_norm": 4.25, "learning_rate": 9.124322534036816e-06, "loss": 2.4618, "step": 2427 }, { "epoch": 2.87363125184966, "grad_norm": 4.09375, "learning_rate": 9.116082907129204e-06, "loss": 2.3648, "step": 2428 }, { "epoch": 2.874815034033738, "grad_norm": 4.625, "learning_rate": 9.107843885007042e-06, "loss": 2.7653, "step": 2429 }, { "epoch": 2.875998816217816, "grad_norm": 4.78125, "learning_rate": 9.09960547330755e-06, "loss": 2.2666, "step": 2430 }, { "epoch": 2.877182598401894, "grad_norm": 4.75, "learning_rate": 9.091367677667538e-06, "loss": 2.535, "step": 2431 }, { "epoch": 2.8783663805859723, "grad_norm": 4.75, "learning_rate": 9.083130503723397e-06, "loss": 2.6712, "step": 2432 }, { "epoch": 2.8795501627700504, "grad_norm": 4.875, "learning_rate": 9.074893957111087e-06, "loss": 2.3286, "step": 2433 }, { "epoch": 2.8807339449541285, "grad_norm": 5.03125, "learning_rate": 9.06665804346614e-06, "loss": 2.6067, "step": 2434 }, { "epoch": 2.8819177271382066, "grad_norm": 4.90625, "learning_rate": 9.058422768423653e-06, "loss": 2.8425, "step": 2435 }, { "epoch": 2.8831015093222847, "grad_norm": 4.53125, "learning_rate": 9.050188137618296e-06, "loss": 2.3759, "step": 2436 }, { "epoch": 2.884285291506363, "grad_norm": 4.59375, "learning_rate": 9.041954156684282e-06, "loss": 2.5844, "step": 2437 }, { "epoch": 2.885469073690441, "grad_norm": 4.78125, "learning_rate": 9.033720831255391e-06, "loss": 2.4364, "step": 2438 }, { "epoch": 2.886652855874519, "grad_norm": 4.59375, "learning_rate": 9.025488166964955e-06, "loss": 2.1299, "step": 2439 }, { "epoch": 2.887836638058597, "grad_norm": 4.46875, "learning_rate": 9.017256169445847e-06, "loss": 2.2889, "step": 2440 }, { "epoch": 2.8890204202426752, "grad_norm": 4.78125, "learning_rate": 9.009024844330489e-06, "loss": 2.4842, "step": 2441 }, { "epoch": 2.8902042024267534, "grad_norm": 4.71875, "learning_rate": 9.000794197250836e-06, "loss": 2.6546, "step": 2442 }, { "epoch": 2.8913879846108315, "grad_norm": 4.34375, "learning_rate": 8.992564233838388e-06, "loss": 2.1698, "step": 2443 }, { "epoch": 2.8925717667949096, "grad_norm": 4.46875, "learning_rate": 8.984334959724177e-06, "loss": 2.3843, "step": 2444 }, { "epoch": 2.893755548978988, "grad_norm": 4.90625, "learning_rate": 8.976106380538752e-06, "loss": 2.5527, "step": 2445 }, { "epoch": 2.894939331163066, "grad_norm": 5.0, "learning_rate": 8.9678785019122e-06, "loss": 2.3577, "step": 2446 }, { "epoch": 2.8961231133471443, "grad_norm": 4.28125, "learning_rate": 8.959651329474115e-06, "loss": 2.2662, "step": 2447 }, { "epoch": 2.897306895531222, "grad_norm": 4.3125, "learning_rate": 8.951424868853623e-06, "loss": 1.9764, "step": 2448 }, { "epoch": 2.8984906777153006, "grad_norm": 4.1875, "learning_rate": 8.943199125679357e-06, "loss": 2.0989, "step": 2449 }, { "epoch": 2.8996744598993787, "grad_norm": 4.90625, "learning_rate": 8.934974105579449e-06, "loss": 2.576, "step": 2450 }, { "epoch": 2.9008582420834568, "grad_norm": 4.46875, "learning_rate": 8.926749814181555e-06, "loss": 2.5389, "step": 2451 }, { "epoch": 2.902042024267535, "grad_norm": 4.53125, "learning_rate": 8.918526257112815e-06, "loss": 2.6568, "step": 2452 }, { "epoch": 2.903225806451613, "grad_norm": 5.21875, "learning_rate": 8.910303439999877e-06, "loss": 2.5161, "step": 2453 }, { "epoch": 2.904409588635691, "grad_norm": 4.78125, "learning_rate": 8.902081368468879e-06, "loss": 2.6051, "step": 2454 }, { "epoch": 2.905593370819769, "grad_norm": 5.1875, "learning_rate": 8.893860048145446e-06, "loss": 2.7391, "step": 2455 }, { "epoch": 2.9067771530038473, "grad_norm": 4.65625, "learning_rate": 8.8856394846547e-06, "loss": 2.3827, "step": 2456 }, { "epoch": 2.9079609351879254, "grad_norm": 4.875, "learning_rate": 8.877419683621233e-06, "loss": 2.3782, "step": 2457 }, { "epoch": 2.9091447173720035, "grad_norm": 5.3125, "learning_rate": 8.869200650669122e-06, "loss": 2.6512, "step": 2458 }, { "epoch": 2.9103284995560816, "grad_norm": 4.6875, "learning_rate": 8.860982391421914e-06, "loss": 2.3, "step": 2459 }, { "epoch": 2.9115122817401597, "grad_norm": 4.8125, "learning_rate": 8.85276491150263e-06, "loss": 2.3814, "step": 2460 }, { "epoch": 2.912696063924238, "grad_norm": 5.0625, "learning_rate": 8.844548216533759e-06, "loss": 2.5416, "step": 2461 }, { "epoch": 2.913879846108316, "grad_norm": 4.5, "learning_rate": 8.836332312137246e-06, "loss": 2.7414, "step": 2462 }, { "epoch": 2.915063628292394, "grad_norm": 4.28125, "learning_rate": 8.828117203934506e-06, "loss": 2.47, "step": 2463 }, { "epoch": 2.916247410476472, "grad_norm": 3.9375, "learning_rate": 8.8199028975464e-06, "loss": 1.9698, "step": 2464 }, { "epoch": 2.9174311926605503, "grad_norm": 4.5625, "learning_rate": 8.811689398593245e-06, "loss": 2.6574, "step": 2465 }, { "epoch": 2.918614974844629, "grad_norm": 4.625, "learning_rate": 8.8034767126948e-06, "loss": 1.9778, "step": 2466 }, { "epoch": 2.9197987570287065, "grad_norm": 4.84375, "learning_rate": 8.795264845470272e-06, "loss": 2.6999, "step": 2467 }, { "epoch": 2.920982539212785, "grad_norm": 4.6875, "learning_rate": 8.787053802538316e-06, "loss": 2.622, "step": 2468 }, { "epoch": 2.9221663213968627, "grad_norm": 5.21875, "learning_rate": 8.778843589517005e-06, "loss": 2.3133, "step": 2469 }, { "epoch": 2.9233501035809413, "grad_norm": 4.5625, "learning_rate": 8.770634212023861e-06, "loss": 2.6155, "step": 2470 }, { "epoch": 2.9245338857650194, "grad_norm": 4.5, "learning_rate": 8.762425675675818e-06, "loss": 2.5978, "step": 2471 }, { "epoch": 2.9257176679490975, "grad_norm": 4.25, "learning_rate": 8.754217986089253e-06, "loss": 2.1365, "step": 2472 }, { "epoch": 2.9269014501331756, "grad_norm": 4.6875, "learning_rate": 8.746011148879951e-06, "loss": 2.6182, "step": 2473 }, { "epoch": 2.9280852323172537, "grad_norm": 4.71875, "learning_rate": 8.737805169663113e-06, "loss": 2.5527, "step": 2474 }, { "epoch": 2.929269014501332, "grad_norm": 4.8125, "learning_rate": 8.729600054053367e-06, "loss": 2.7936, "step": 2475 }, { "epoch": 2.93045279668541, "grad_norm": 4.71875, "learning_rate": 8.72139580766473e-06, "loss": 2.368, "step": 2476 }, { "epoch": 2.931636578869488, "grad_norm": 4.4375, "learning_rate": 8.713192436110635e-06, "loss": 2.0711, "step": 2477 }, { "epoch": 2.932820361053566, "grad_norm": 4.46875, "learning_rate": 8.704989945003926e-06, "loss": 2.7566, "step": 2478 }, { "epoch": 2.9340041432376442, "grad_norm": 4.65625, "learning_rate": 8.696788339956819e-06, "loss": 2.7842, "step": 2479 }, { "epoch": 2.9351879254217224, "grad_norm": 4.375, "learning_rate": 8.688587626580954e-06, "loss": 2.6173, "step": 2480 }, { "epoch": 2.9363717076058005, "grad_norm": 4.4375, "learning_rate": 8.680387810487337e-06, "loss": 2.6441, "step": 2481 }, { "epoch": 2.9375554897898786, "grad_norm": 5.03125, "learning_rate": 8.672188897286374e-06, "loss": 2.3929, "step": 2482 }, { "epoch": 2.9387392719739567, "grad_norm": 4.5, "learning_rate": 8.663990892587839e-06, "loss": 2.2711, "step": 2483 }, { "epoch": 2.939923054158035, "grad_norm": 4.84375, "learning_rate": 8.655793802000905e-06, "loss": 2.2069, "step": 2484 }, { "epoch": 2.941106836342113, "grad_norm": 4.59375, "learning_rate": 8.647597631134105e-06, "loss": 2.4882, "step": 2485 }, { "epoch": 2.942290618526191, "grad_norm": 4.25, "learning_rate": 8.639402385595342e-06, "loss": 2.359, "step": 2486 }, { "epoch": 2.9434744007102696, "grad_norm": 4.15625, "learning_rate": 8.631208070991896e-06, "loss": 2.403, "step": 2487 }, { "epoch": 2.944658182894347, "grad_norm": 4.71875, "learning_rate": 8.623014692930399e-06, "loss": 2.4515, "step": 2488 }, { "epoch": 2.9458419650784258, "grad_norm": 4.5, "learning_rate": 8.614822257016845e-06, "loss": 2.5954, "step": 2489 }, { "epoch": 2.9470257472625034, "grad_norm": 5.1875, "learning_rate": 8.606630768856596e-06, "loss": 2.3924, "step": 2490 }, { "epoch": 2.948209529446582, "grad_norm": 4.59375, "learning_rate": 8.598440234054345e-06, "loss": 2.3575, "step": 2491 }, { "epoch": 2.94939331163066, "grad_norm": 5.0625, "learning_rate": 8.590250658214148e-06, "loss": 2.2157, "step": 2492 }, { "epoch": 2.950577093814738, "grad_norm": 5.15625, "learning_rate": 8.582062046939396e-06, "loss": 2.4116, "step": 2493 }, { "epoch": 2.9517608759988163, "grad_norm": 5.25, "learning_rate": 8.573874405832828e-06, "loss": 2.7682, "step": 2494 }, { "epoch": 2.9529446581828944, "grad_norm": 4.6875, "learning_rate": 8.565687740496507e-06, "loss": 2.4964, "step": 2495 }, { "epoch": 2.9541284403669725, "grad_norm": 4.9375, "learning_rate": 8.557502056531844e-06, "loss": 2.4064, "step": 2496 }, { "epoch": 2.9553122225510506, "grad_norm": 4.9375, "learning_rate": 8.549317359539571e-06, "loss": 3.1008, "step": 2497 }, { "epoch": 2.9564960047351287, "grad_norm": 4.53125, "learning_rate": 8.541133655119736e-06, "loss": 2.2388, "step": 2498 }, { "epoch": 2.957679786919207, "grad_norm": 4.90625, "learning_rate": 8.532950948871728e-06, "loss": 2.6238, "step": 2499 }, { "epoch": 2.958863569103285, "grad_norm": 4.6875, "learning_rate": 8.524769246394233e-06, "loss": 2.7269, "step": 2500 }, { "epoch": 2.960047351287363, "grad_norm": 4.6875, "learning_rate": 8.516588553285258e-06, "loss": 2.1742, "step": 2501 }, { "epoch": 2.961231133471441, "grad_norm": 4.71875, "learning_rate": 8.50840887514213e-06, "loss": 2.2089, "step": 2502 }, { "epoch": 2.9624149156555193, "grad_norm": 4.21875, "learning_rate": 8.500230217561459e-06, "loss": 2.3663, "step": 2503 }, { "epoch": 2.9635986978395974, "grad_norm": 4.625, "learning_rate": 8.492052586139176e-06, "loss": 2.6739, "step": 2504 }, { "epoch": 2.9647824800236755, "grad_norm": 4.875, "learning_rate": 8.483875986470502e-06, "loss": 2.5822, "step": 2505 }, { "epoch": 2.9659662622077536, "grad_norm": 3.9375, "learning_rate": 8.47570042414995e-06, "loss": 2.0864, "step": 2506 }, { "epoch": 2.9671500443918317, "grad_norm": 4.125, "learning_rate": 8.467525904771336e-06, "loss": 2.1753, "step": 2507 }, { "epoch": 2.9683338265759103, "grad_norm": 5.1875, "learning_rate": 8.459352433927743e-06, "loss": 2.4663, "step": 2508 }, { "epoch": 2.969517608759988, "grad_norm": 4.5, "learning_rate": 8.45118001721155e-06, "loss": 2.3658, "step": 2509 }, { "epoch": 2.9707013909440665, "grad_norm": 4.90625, "learning_rate": 8.443008660214409e-06, "loss": 2.3725, "step": 2510 }, { "epoch": 2.971885173128144, "grad_norm": 4.625, "learning_rate": 8.434838368527254e-06, "loss": 2.6719, "step": 2511 }, { "epoch": 2.9730689553122227, "grad_norm": 4.78125, "learning_rate": 8.42666914774028e-06, "loss": 2.3905, "step": 2512 }, { "epoch": 2.974252737496301, "grad_norm": 4.09375, "learning_rate": 8.418501003442954e-06, "loss": 2.1407, "step": 2513 }, { "epoch": 2.975436519680379, "grad_norm": 5.3125, "learning_rate": 8.410333941224017e-06, "loss": 2.7638, "step": 2514 }, { "epoch": 2.976620301864457, "grad_norm": 4.71875, "learning_rate": 8.402167966671445e-06, "loss": 2.4441, "step": 2515 }, { "epoch": 2.977804084048535, "grad_norm": 5.09375, "learning_rate": 8.394003085372496e-06, "loss": 2.5726, "step": 2516 }, { "epoch": 2.9789878662326132, "grad_norm": 4.875, "learning_rate": 8.385839302913661e-06, "loss": 2.6332, "step": 2517 }, { "epoch": 2.9801716484166914, "grad_norm": 5.15625, "learning_rate": 8.377676624880688e-06, "loss": 2.6673, "step": 2518 }, { "epoch": 2.9813554306007695, "grad_norm": 4.40625, "learning_rate": 8.369515056858575e-06, "loss": 2.0887, "step": 2519 }, { "epoch": 2.9825392127848476, "grad_norm": 4.53125, "learning_rate": 8.361354604431544e-06, "loss": 2.0741, "step": 2520 }, { "epoch": 2.9837229949689257, "grad_norm": 4.28125, "learning_rate": 8.353195273183068e-06, "loss": 2.0276, "step": 2521 }, { "epoch": 2.984906777153004, "grad_norm": 4.53125, "learning_rate": 8.345037068695844e-06, "loss": 2.4392, "step": 2522 }, { "epoch": 2.986090559337082, "grad_norm": 4.21875, "learning_rate": 8.336879996551807e-06, "loss": 2.473, "step": 2523 }, { "epoch": 2.98727434152116, "grad_norm": 4.71875, "learning_rate": 8.32872406233211e-06, "loss": 2.3669, "step": 2524 }, { "epoch": 2.988458123705238, "grad_norm": 4.34375, "learning_rate": 8.320569271617127e-06, "loss": 2.3632, "step": 2525 }, { "epoch": 2.989641905889316, "grad_norm": 5.375, "learning_rate": 8.31241562998646e-06, "loss": 2.8585, "step": 2526 }, { "epoch": 2.9908256880733948, "grad_norm": 4.25, "learning_rate": 8.304263143018906e-06, "loss": 2.1514, "step": 2527 }, { "epoch": 2.9920094702574724, "grad_norm": 5.03125, "learning_rate": 8.296111816292494e-06, "loss": 2.7298, "step": 2528 }, { "epoch": 2.993193252441551, "grad_norm": 5.34375, "learning_rate": 8.287961655384439e-06, "loss": 2.4287, "step": 2529 }, { "epoch": 2.9943770346256287, "grad_norm": 4.71875, "learning_rate": 8.279812665871171e-06, "loss": 2.5324, "step": 2530 }, { "epoch": 2.995560816809707, "grad_norm": 5.28125, "learning_rate": 8.271664853328319e-06, "loss": 2.8279, "step": 2531 }, { "epoch": 2.996744598993785, "grad_norm": 4.75, "learning_rate": 8.263518223330698e-06, "loss": 2.1335, "step": 2532 }, { "epoch": 2.9979283811778634, "grad_norm": 4.875, "learning_rate": 8.255372781452322e-06, "loss": 2.2986, "step": 2533 }, { "epoch": 2.9991121633619415, "grad_norm": 4.5625, "learning_rate": 8.247228533266381e-06, "loss": 2.3969, "step": 2534 }, { "epoch": 3.0, "grad_norm": 4.09375, "learning_rate": 8.239085484345265e-06, "loss": 1.733, "step": 2535 }, { "epoch": 3.001183782184078, "grad_norm": 4.46875, "learning_rate": 8.230943640260533e-06, "loss": 2.6609, "step": 2536 }, { "epoch": 3.002367564368156, "grad_norm": 4.96875, "learning_rate": 8.222803006582915e-06, "loss": 2.1878, "step": 2537 }, { "epoch": 3.0035513465522343, "grad_norm": 4.84375, "learning_rate": 8.21466358888233e-06, "loss": 2.4261, "step": 2538 }, { "epoch": 3.0047351287363124, "grad_norm": 3.921875, "learning_rate": 8.20652539272784e-06, "loss": 2.086, "step": 2539 }, { "epoch": 3.0059189109203905, "grad_norm": 4.4375, "learning_rate": 8.198388423687694e-06, "loss": 2.1258, "step": 2540 }, { "epoch": 3.0071026931044686, "grad_norm": 4.375, "learning_rate": 8.19025268732929e-06, "loss": 2.1337, "step": 2541 }, { "epoch": 3.0082864752885468, "grad_norm": 4.46875, "learning_rate": 8.182118189219183e-06, "loss": 2.2343, "step": 2542 }, { "epoch": 3.009470257472625, "grad_norm": 4.59375, "learning_rate": 8.173984934923091e-06, "loss": 2.5537, "step": 2543 }, { "epoch": 3.010654039656703, "grad_norm": 4.0625, "learning_rate": 8.165852930005865e-06, "loss": 2.1428, "step": 2544 }, { "epoch": 3.011837821840781, "grad_norm": 4.40625, "learning_rate": 8.157722180031512e-06, "loss": 2.3848, "step": 2545 }, { "epoch": 3.0130216040248596, "grad_norm": 4.0625, "learning_rate": 8.149592690563172e-06, "loss": 2.2123, "step": 2546 }, { "epoch": 3.0142053862089377, "grad_norm": 4.71875, "learning_rate": 8.141464467163134e-06, "loss": 2.4875, "step": 2547 }, { "epoch": 3.015389168393016, "grad_norm": 4.4375, "learning_rate": 8.133337515392818e-06, "loss": 2.2035, "step": 2548 }, { "epoch": 3.016572950577094, "grad_norm": 4.53125, "learning_rate": 8.125211840812759e-06, "loss": 2.2769, "step": 2549 }, { "epoch": 3.017756732761172, "grad_norm": 4.28125, "learning_rate": 8.117087448982643e-06, "loss": 2.5116, "step": 2550 }, { "epoch": 3.01894051494525, "grad_norm": 4.34375, "learning_rate": 8.108964345461252e-06, "loss": 2.2511, "step": 2551 }, { "epoch": 3.0201242971293283, "grad_norm": 4.84375, "learning_rate": 8.100842535806509e-06, "loss": 2.5429, "step": 2552 }, { "epoch": 3.0213080793134064, "grad_norm": 4.46875, "learning_rate": 8.09272202557544e-06, "loss": 2.5193, "step": 2553 }, { "epoch": 3.0224918614974845, "grad_norm": 4.1875, "learning_rate": 8.08460282032418e-06, "loss": 2.2294, "step": 2554 }, { "epoch": 3.0236756436815626, "grad_norm": 4.8125, "learning_rate": 8.076484925607983e-06, "loss": 2.6043, "step": 2555 }, { "epoch": 3.0248594258656407, "grad_norm": 4.875, "learning_rate": 8.068368346981192e-06, "loss": 2.6556, "step": 2556 }, { "epoch": 3.026043208049719, "grad_norm": 5.0625, "learning_rate": 8.06025308999726e-06, "loss": 2.2034, "step": 2557 }, { "epoch": 3.027226990233797, "grad_norm": 4.875, "learning_rate": 8.052139160208725e-06, "loss": 2.3218, "step": 2558 }, { "epoch": 3.028410772417875, "grad_norm": 4.21875, "learning_rate": 8.044026563167233e-06, "loss": 2.2805, "step": 2559 }, { "epoch": 3.029594554601953, "grad_norm": 4.46875, "learning_rate": 8.035915304423507e-06, "loss": 2.7445, "step": 2560 }, { "epoch": 3.0307783367860313, "grad_norm": 4.75, "learning_rate": 8.02780538952735e-06, "loss": 2.1474, "step": 2561 }, { "epoch": 3.0319621189701094, "grad_norm": 4.6875, "learning_rate": 8.019696824027662e-06, "loss": 2.6946, "step": 2562 }, { "epoch": 3.0331459011541875, "grad_norm": 4.09375, "learning_rate": 8.011589613472397e-06, "loss": 2.3348, "step": 2563 }, { "epoch": 3.0343296833382656, "grad_norm": 4.34375, "learning_rate": 8.003483763408604e-06, "loss": 2.7248, "step": 2564 }, { "epoch": 3.0355134655223437, "grad_norm": 4.65625, "learning_rate": 7.99537927938239e-06, "loss": 2.6593, "step": 2565 }, { "epoch": 3.036697247706422, "grad_norm": 4.15625, "learning_rate": 7.987276166938923e-06, "loss": 2.1451, "step": 2566 }, { "epoch": 3.0378810298905004, "grad_norm": 4.5, "learning_rate": 7.979174431622447e-06, "loss": 2.2961, "step": 2567 }, { "epoch": 3.0390648120745785, "grad_norm": 4.8125, "learning_rate": 7.97107407897625e-06, "loss": 2.4264, "step": 2568 }, { "epoch": 3.0402485942586566, "grad_norm": 4.53125, "learning_rate": 7.962975114542681e-06, "loss": 2.3389, "step": 2569 }, { "epoch": 3.0414323764427347, "grad_norm": 4.34375, "learning_rate": 7.954877543863133e-06, "loss": 2.2481, "step": 2570 }, { "epoch": 3.042616158626813, "grad_norm": 4.59375, "learning_rate": 7.946781372478056e-06, "loss": 2.7133, "step": 2571 }, { "epoch": 3.043799940810891, "grad_norm": 4.1875, "learning_rate": 7.938686605926934e-06, "loss": 2.1327, "step": 2572 }, { "epoch": 3.044983722994969, "grad_norm": 4.65625, "learning_rate": 7.930593249748289e-06, "loss": 2.17, "step": 2573 }, { "epoch": 3.046167505179047, "grad_norm": 4.25, "learning_rate": 7.92250130947969e-06, "loss": 2.3086, "step": 2574 }, { "epoch": 3.047351287363125, "grad_norm": 4.59375, "learning_rate": 7.914410790657715e-06, "loss": 2.5572, "step": 2575 }, { "epoch": 3.0485350695472033, "grad_norm": 4.34375, "learning_rate": 7.906321698817992e-06, "loss": 2.4178, "step": 2576 }, { "epoch": 3.0497188517312814, "grad_norm": 5.25, "learning_rate": 7.898234039495162e-06, "loss": 2.2825, "step": 2577 }, { "epoch": 3.0509026339153595, "grad_norm": 4.15625, "learning_rate": 7.890147818222885e-06, "loss": 2.1797, "step": 2578 }, { "epoch": 3.0520864160994376, "grad_norm": 4.6875, "learning_rate": 7.882063040533844e-06, "loss": 2.6411, "step": 2579 }, { "epoch": 3.0532701982835158, "grad_norm": 4.84375, "learning_rate": 7.873979711959724e-06, "loss": 2.9264, "step": 2580 }, { "epoch": 3.054453980467594, "grad_norm": 4.8125, "learning_rate": 7.865897838031225e-06, "loss": 2.6477, "step": 2581 }, { "epoch": 3.055637762651672, "grad_norm": 4.53125, "learning_rate": 7.857817424278056e-06, "loss": 2.4148, "step": 2582 }, { "epoch": 3.05682154483575, "grad_norm": 4.78125, "learning_rate": 7.849738476228916e-06, "loss": 2.7095, "step": 2583 }, { "epoch": 3.058005327019828, "grad_norm": 4.4375, "learning_rate": 7.841660999411513e-06, "loss": 2.4231, "step": 2584 }, { "epoch": 3.0591891092039063, "grad_norm": 4.46875, "learning_rate": 7.833584999352533e-06, "loss": 2.536, "step": 2585 }, { "epoch": 3.0603728913879844, "grad_norm": 5.65625, "learning_rate": 7.825510481577672e-06, "loss": 2.2879, "step": 2586 }, { "epoch": 3.061556673572063, "grad_norm": 4.3125, "learning_rate": 7.81743745161159e-06, "loss": 2.1196, "step": 2587 }, { "epoch": 3.062740455756141, "grad_norm": 4.5625, "learning_rate": 7.809365914977945e-06, "loss": 2.4781, "step": 2588 }, { "epoch": 3.063924237940219, "grad_norm": 4.75, "learning_rate": 7.80129587719937e-06, "loss": 2.3809, "step": 2589 }, { "epoch": 3.0651080201242973, "grad_norm": 4.78125, "learning_rate": 7.793227343797465e-06, "loss": 2.3982, "step": 2590 }, { "epoch": 3.0662918023083754, "grad_norm": 4.34375, "learning_rate": 7.785160320292812e-06, "loss": 2.4975, "step": 2591 }, { "epoch": 3.0674755844924535, "grad_norm": 3.875, "learning_rate": 7.777094812204949e-06, "loss": 2.1415, "step": 2592 }, { "epoch": 3.0686593666765316, "grad_norm": 5.0, "learning_rate": 7.769030825052383e-06, "loss": 2.2712, "step": 2593 }, { "epoch": 3.0698431488606097, "grad_norm": 4.71875, "learning_rate": 7.760968364352584e-06, "loss": 2.7984, "step": 2594 }, { "epoch": 3.071026931044688, "grad_norm": 4.8125, "learning_rate": 7.752907435621968e-06, "loss": 2.4561, "step": 2595 }, { "epoch": 3.072210713228766, "grad_norm": 4.71875, "learning_rate": 7.744848044375913e-06, "loss": 2.3618, "step": 2596 }, { "epoch": 3.073394495412844, "grad_norm": 4.59375, "learning_rate": 7.736790196128732e-06, "loss": 2.4514, "step": 2597 }, { "epoch": 3.074578277596922, "grad_norm": 4.84375, "learning_rate": 7.7287338963937e-06, "loss": 2.4863, "step": 2598 }, { "epoch": 3.0757620597810003, "grad_norm": 4.34375, "learning_rate": 7.720679150683012e-06, "loss": 2.3315, "step": 2599 }, { "epoch": 3.0769458419650784, "grad_norm": 4.84375, "learning_rate": 7.712625964507818e-06, "loss": 3.0371, "step": 2600 }, { "epoch": 3.0781296241491565, "grad_norm": 4.125, "learning_rate": 7.704574343378193e-06, "loss": 2.0149, "step": 2601 }, { "epoch": 3.0793134063332346, "grad_norm": 4.6875, "learning_rate": 7.696524292803138e-06, "loss": 2.4756, "step": 2602 }, { "epoch": 3.0804971885173127, "grad_norm": 4.65625, "learning_rate": 7.688475818290588e-06, "loss": 2.2102, "step": 2603 }, { "epoch": 3.081680970701391, "grad_norm": 4.6875, "learning_rate": 7.680428925347388e-06, "loss": 2.6138, "step": 2604 }, { "epoch": 3.082864752885469, "grad_norm": 4.90625, "learning_rate": 7.67238361947931e-06, "loss": 2.5365, "step": 2605 }, { "epoch": 3.084048535069547, "grad_norm": 4.84375, "learning_rate": 7.664339906191042e-06, "loss": 2.694, "step": 2606 }, { "epoch": 3.085232317253625, "grad_norm": 4.09375, "learning_rate": 7.656297790986173e-06, "loss": 2.167, "step": 2607 }, { "epoch": 3.0864160994377037, "grad_norm": 4.65625, "learning_rate": 7.648257279367206e-06, "loss": 2.1874, "step": 2608 }, { "epoch": 3.087599881621782, "grad_norm": 4.125, "learning_rate": 7.64021837683554e-06, "loss": 2.5109, "step": 2609 }, { "epoch": 3.08878366380586, "grad_norm": 4.46875, "learning_rate": 7.632181088891482e-06, "loss": 2.4886, "step": 2610 }, { "epoch": 3.089967445989938, "grad_norm": 4.53125, "learning_rate": 7.624145421034231e-06, "loss": 2.5443, "step": 2611 }, { "epoch": 3.091151228174016, "grad_norm": 4.59375, "learning_rate": 7.616111378761872e-06, "loss": 2.4076, "step": 2612 }, { "epoch": 3.092335010358094, "grad_norm": 5.65625, "learning_rate": 7.608078967571384e-06, "loss": 2.6544, "step": 2613 }, { "epoch": 3.0935187925421723, "grad_norm": 5.0625, "learning_rate": 7.600048192958622e-06, "loss": 2.5363, "step": 2614 }, { "epoch": 3.0947025747262504, "grad_norm": 4.6875, "learning_rate": 7.592019060418339e-06, "loss": 2.173, "step": 2615 }, { "epoch": 3.0958863569103285, "grad_norm": 4.59375, "learning_rate": 7.583991575444143e-06, "loss": 2.4514, "step": 2616 }, { "epoch": 3.0970701390944066, "grad_norm": 5.21875, "learning_rate": 7.575965743528524e-06, "loss": 2.306, "step": 2617 }, { "epoch": 3.0982539212784848, "grad_norm": 4.875, "learning_rate": 7.567941570162849e-06, "loss": 2.3405, "step": 2618 }, { "epoch": 3.099437703462563, "grad_norm": 5.0, "learning_rate": 7.559919060837337e-06, "loss": 2.8327, "step": 2619 }, { "epoch": 3.100621485646641, "grad_norm": 4.46875, "learning_rate": 7.551898221041076e-06, "loss": 2.1692, "step": 2620 }, { "epoch": 3.101805267830719, "grad_norm": 4.46875, "learning_rate": 7.543879056262004e-06, "loss": 2.2766, "step": 2621 }, { "epoch": 3.102989050014797, "grad_norm": 4.875, "learning_rate": 7.535861571986927e-06, "loss": 2.5376, "step": 2622 }, { "epoch": 3.1041728321988753, "grad_norm": 4.375, "learning_rate": 7.52784577370149e-06, "loss": 2.2406, "step": 2623 }, { "epoch": 3.1053566143829534, "grad_norm": 4.625, "learning_rate": 7.519831666890185e-06, "loss": 2.4085, "step": 2624 }, { "epoch": 3.1065403965670315, "grad_norm": 4.65625, "learning_rate": 7.51181925703635e-06, "loss": 2.4102, "step": 2625 }, { "epoch": 3.1077241787511096, "grad_norm": 4.4375, "learning_rate": 7.503808549622158e-06, "loss": 2.3188, "step": 2626 }, { "epoch": 3.1089079609351877, "grad_norm": 5.5, "learning_rate": 7.495799550128625e-06, "loss": 2.596, "step": 2627 }, { "epoch": 3.1100917431192663, "grad_norm": 4.6875, "learning_rate": 7.487792264035593e-06, "loss": 2.6349, "step": 2628 }, { "epoch": 3.1112755253033444, "grad_norm": 4.90625, "learning_rate": 7.479786696821724e-06, "loss": 2.5923, "step": 2629 }, { "epoch": 3.1124593074874225, "grad_norm": 4.75, "learning_rate": 7.471782853964525e-06, "loss": 2.4004, "step": 2630 }, { "epoch": 3.1136430896715006, "grad_norm": 4.53125, "learning_rate": 7.463780740940299e-06, "loss": 2.3254, "step": 2631 }, { "epoch": 3.1148268718555787, "grad_norm": 4.5625, "learning_rate": 7.455780363224184e-06, "loss": 2.1148, "step": 2632 }, { "epoch": 3.116010654039657, "grad_norm": 4.75, "learning_rate": 7.447781726290116e-06, "loss": 2.572, "step": 2633 }, { "epoch": 3.117194436223735, "grad_norm": 4.75, "learning_rate": 7.439784835610853e-06, "loss": 2.518, "step": 2634 }, { "epoch": 3.118378218407813, "grad_norm": 4.53125, "learning_rate": 7.431789696657952e-06, "loss": 2.3588, "step": 2635 }, { "epoch": 3.119562000591891, "grad_norm": 4.34375, "learning_rate": 7.423796314901769e-06, "loss": 2.0895, "step": 2636 }, { "epoch": 3.1207457827759693, "grad_norm": 4.59375, "learning_rate": 7.4158046958114635e-06, "loss": 2.425, "step": 2637 }, { "epoch": 3.1219295649600474, "grad_norm": 4.4375, "learning_rate": 7.407814844854981e-06, "loss": 2.3045, "step": 2638 }, { "epoch": 3.1231133471441255, "grad_norm": 4.46875, "learning_rate": 7.399826767499068e-06, "loss": 2.4503, "step": 2639 }, { "epoch": 3.1242971293282036, "grad_norm": 4.8125, "learning_rate": 7.3918404692092504e-06, "loss": 2.6254, "step": 2640 }, { "epoch": 3.1254809115122817, "grad_norm": 4.5625, "learning_rate": 7.383855955449835e-06, "loss": 2.31, "step": 2641 }, { "epoch": 3.12666469369636, "grad_norm": 4.0625, "learning_rate": 7.3758732316839155e-06, "loss": 2.3466, "step": 2642 }, { "epoch": 3.127848475880438, "grad_norm": 4.875, "learning_rate": 7.367892303373352e-06, "loss": 2.471, "step": 2643 }, { "epoch": 3.129032258064516, "grad_norm": 4.53125, "learning_rate": 7.3599131759787835e-06, "loss": 2.655, "step": 2644 }, { "epoch": 3.130216040248594, "grad_norm": 4.125, "learning_rate": 7.351935854959608e-06, "loss": 2.5136, "step": 2645 }, { "epoch": 3.1313998224326722, "grad_norm": 5.15625, "learning_rate": 7.343960345774001e-06, "loss": 2.4041, "step": 2646 }, { "epoch": 3.1325836046167503, "grad_norm": 4.0625, "learning_rate": 7.335986653878887e-06, "loss": 2.1301, "step": 2647 }, { "epoch": 3.1337673868008284, "grad_norm": 5.25, "learning_rate": 7.328014784729949e-06, "loss": 2.4806, "step": 2648 }, { "epoch": 3.134951168984907, "grad_norm": 4.6875, "learning_rate": 7.320044743781627e-06, "loss": 2.5777, "step": 2649 }, { "epoch": 3.136134951168985, "grad_norm": 4.21875, "learning_rate": 7.3120765364871e-06, "loss": 2.3345, "step": 2650 }, { "epoch": 3.137318733353063, "grad_norm": 4.5625, "learning_rate": 7.30411016829831e-06, "loss": 2.3022, "step": 2651 }, { "epoch": 3.1385025155371413, "grad_norm": 4.5625, "learning_rate": 7.296145644665928e-06, "loss": 2.578, "step": 2652 }, { "epoch": 3.1396862977212194, "grad_norm": 4.59375, "learning_rate": 7.288182971039362e-06, "loss": 2.3453, "step": 2653 }, { "epoch": 3.1408700799052975, "grad_norm": 4.21875, "learning_rate": 7.2802221528667604e-06, "loss": 2.5698, "step": 2654 }, { "epoch": 3.1420538620893756, "grad_norm": 4.40625, "learning_rate": 7.272263195594999e-06, "loss": 2.5263, "step": 2655 }, { "epoch": 3.1432376442734538, "grad_norm": 4.53125, "learning_rate": 7.264306104669678e-06, "loss": 2.2147, "step": 2656 }, { "epoch": 3.144421426457532, "grad_norm": 4.53125, "learning_rate": 7.256350885535129e-06, "loss": 2.493, "step": 2657 }, { "epoch": 3.14560520864161, "grad_norm": 4.90625, "learning_rate": 7.248397543634394e-06, "loss": 2.133, "step": 2658 }, { "epoch": 3.146788990825688, "grad_norm": 4.21875, "learning_rate": 7.240446084409234e-06, "loss": 2.1151, "step": 2659 }, { "epoch": 3.147972773009766, "grad_norm": 4.75, "learning_rate": 7.23249651330012e-06, "loss": 2.2754, "step": 2660 }, { "epoch": 3.1491565551938443, "grad_norm": 4.625, "learning_rate": 7.224548835746238e-06, "loss": 2.35, "step": 2661 }, { "epoch": 3.1503403373779224, "grad_norm": 4.1875, "learning_rate": 7.216603057185465e-06, "loss": 2.6902, "step": 2662 }, { "epoch": 3.1515241195620005, "grad_norm": 4.8125, "learning_rate": 7.208659183054393e-06, "loss": 2.2819, "step": 2663 }, { "epoch": 3.1527079017460786, "grad_norm": 4.71875, "learning_rate": 7.200717218788306e-06, "loss": 2.164, "step": 2664 }, { "epoch": 3.1538916839301567, "grad_norm": 4.34375, "learning_rate": 7.192777169821174e-06, "loss": 2.4852, "step": 2665 }, { "epoch": 3.155075466114235, "grad_norm": 5.15625, "learning_rate": 7.184839041585671e-06, "loss": 2.7045, "step": 2666 }, { "epoch": 3.156259248298313, "grad_norm": 4.46875, "learning_rate": 7.1769028395131415e-06, "loss": 2.5077, "step": 2667 }, { "epoch": 3.157443030482391, "grad_norm": 4.34375, "learning_rate": 7.168968569033619e-06, "loss": 2.2478, "step": 2668 }, { "epoch": 3.158626812666469, "grad_norm": 4.21875, "learning_rate": 7.161036235575822e-06, "loss": 2.1159, "step": 2669 }, { "epoch": 3.1598105948505477, "grad_norm": 5.09375, "learning_rate": 7.153105844567133e-06, "loss": 2.398, "step": 2670 }, { "epoch": 3.160994377034626, "grad_norm": 4.59375, "learning_rate": 7.145177401433611e-06, "loss": 2.3717, "step": 2671 }, { "epoch": 3.162178159218704, "grad_norm": 5.09375, "learning_rate": 7.137250911599978e-06, "loss": 2.5663, "step": 2672 }, { "epoch": 3.163361941402782, "grad_norm": 4.40625, "learning_rate": 7.129326380489625e-06, "loss": 2.2969, "step": 2673 }, { "epoch": 3.16454572358686, "grad_norm": 4.84375, "learning_rate": 7.121403813524596e-06, "loss": 2.5767, "step": 2674 }, { "epoch": 3.1657295057709383, "grad_norm": 4.53125, "learning_rate": 7.1134832161256004e-06, "loss": 2.3355, "step": 2675 }, { "epoch": 3.1669132879550164, "grad_norm": 4.71875, "learning_rate": 7.1055645937119956e-06, "loss": 2.5551, "step": 2676 }, { "epoch": 3.1680970701390945, "grad_norm": 4.5625, "learning_rate": 7.097647951701779e-06, "loss": 2.3701, "step": 2677 }, { "epoch": 3.1692808523231726, "grad_norm": 4.625, "learning_rate": 7.0897332955116115e-06, "loss": 2.2337, "step": 2678 }, { "epoch": 3.1704646345072507, "grad_norm": 4.46875, "learning_rate": 7.0818206305567795e-06, "loss": 2.8086, "step": 2679 }, { "epoch": 3.171648416691329, "grad_norm": 5.5625, "learning_rate": 7.0739099622512085e-06, "loss": 2.7678, "step": 2680 }, { "epoch": 3.172832198875407, "grad_norm": 4.90625, "learning_rate": 7.066001296007469e-06, "loss": 2.9598, "step": 2681 }, { "epoch": 3.174015981059485, "grad_norm": 5.125, "learning_rate": 7.058094637236752e-06, "loss": 2.7456, "step": 2682 }, { "epoch": 3.175199763243563, "grad_norm": 4.1875, "learning_rate": 7.050189991348877e-06, "loss": 2.3107, "step": 2683 }, { "epoch": 3.1763835454276412, "grad_norm": 4.4375, "learning_rate": 7.0422873637522825e-06, "loss": 2.4162, "step": 2684 }, { "epoch": 3.1775673276117193, "grad_norm": 5.34375, "learning_rate": 7.034386759854033e-06, "loss": 2.5036, "step": 2685 }, { "epoch": 3.1787511097957974, "grad_norm": 4.4375, "learning_rate": 7.026488185059808e-06, "loss": 2.1769, "step": 2686 }, { "epoch": 3.1799348919798756, "grad_norm": 4.84375, "learning_rate": 7.018591644773893e-06, "loss": 2.2768, "step": 2687 }, { "epoch": 3.1811186741639537, "grad_norm": 4.625, "learning_rate": 7.0106971443991874e-06, "loss": 2.2892, "step": 2688 }, { "epoch": 3.1823024563480318, "grad_norm": 4.40625, "learning_rate": 7.002804689337188e-06, "loss": 2.4396, "step": 2689 }, { "epoch": 3.18348623853211, "grad_norm": 4.71875, "learning_rate": 6.9949142849880015e-06, "loss": 2.7118, "step": 2690 }, { "epoch": 3.1846700207161884, "grad_norm": 4.625, "learning_rate": 6.98702593675032e-06, "loss": 2.2922, "step": 2691 }, { "epoch": 3.1858538029002665, "grad_norm": 5.3125, "learning_rate": 6.979139650021436e-06, "loss": 2.5858, "step": 2692 }, { "epoch": 3.1870375850843446, "grad_norm": 4.71875, "learning_rate": 6.971255430197238e-06, "loss": 2.4864, "step": 2693 }, { "epoch": 3.1882213672684228, "grad_norm": 4.25, "learning_rate": 6.963373282672185e-06, "loss": 2.3389, "step": 2694 }, { "epoch": 3.189405149452501, "grad_norm": 4.875, "learning_rate": 6.95549321283933e-06, "loss": 2.3214, "step": 2695 }, { "epoch": 3.190588931636579, "grad_norm": 4.46875, "learning_rate": 6.947615226090297e-06, "loss": 2.5789, "step": 2696 }, { "epoch": 3.191772713820657, "grad_norm": 5.125, "learning_rate": 6.939739327815288e-06, "loss": 2.7423, "step": 2697 }, { "epoch": 3.192956496004735, "grad_norm": 4.28125, "learning_rate": 6.931865523403082e-06, "loss": 2.39, "step": 2698 }, { "epoch": 3.1941402781888133, "grad_norm": 4.375, "learning_rate": 6.9239938182410126e-06, "loss": 2.665, "step": 2699 }, { "epoch": 3.1953240603728914, "grad_norm": 5.34375, "learning_rate": 6.916124217714989e-06, "loss": 2.3589, "step": 2700 }, { "epoch": 3.1965078425569695, "grad_norm": 4.59375, "learning_rate": 6.908256727209469e-06, "loss": 2.2345, "step": 2701 }, { "epoch": 3.1976916247410476, "grad_norm": 4.59375, "learning_rate": 6.900391352107478e-06, "loss": 2.5295, "step": 2702 }, { "epoch": 3.1988754069251257, "grad_norm": 4.8125, "learning_rate": 6.892528097790592e-06, "loss": 2.8888, "step": 2703 }, { "epoch": 3.200059189109204, "grad_norm": 4.28125, "learning_rate": 6.884666969638925e-06, "loss": 2.5116, "step": 2704 }, { "epoch": 3.201242971293282, "grad_norm": 4.15625, "learning_rate": 6.87680797303115e-06, "loss": 2.6789, "step": 2705 }, { "epoch": 3.20242675347736, "grad_norm": 4.25, "learning_rate": 6.868951113344473e-06, "loss": 2.4385, "step": 2706 }, { "epoch": 3.203610535661438, "grad_norm": 4.46875, "learning_rate": 6.861096395954645e-06, "loss": 2.4669, "step": 2707 }, { "epoch": 3.2047943178455163, "grad_norm": 4.15625, "learning_rate": 6.8532438262359404e-06, "loss": 2.0326, "step": 2708 }, { "epoch": 3.2059781000295944, "grad_norm": 4.75, "learning_rate": 6.845393409561172e-06, "loss": 2.0611, "step": 2709 }, { "epoch": 3.2071618822136725, "grad_norm": 4.875, "learning_rate": 6.837545151301685e-06, "loss": 2.3783, "step": 2710 }, { "epoch": 3.2083456643977506, "grad_norm": 4.71875, "learning_rate": 6.8296990568273335e-06, "loss": 2.5802, "step": 2711 }, { "epoch": 3.209529446581829, "grad_norm": 5.46875, "learning_rate": 6.821855131506503e-06, "loss": 2.6158, "step": 2712 }, { "epoch": 3.2107132287659073, "grad_norm": 4.4375, "learning_rate": 6.8140133807060834e-06, "loss": 2.218, "step": 2713 }, { "epoch": 3.2118970109499854, "grad_norm": 4.65625, "learning_rate": 6.806173809791492e-06, "loss": 2.1826, "step": 2714 }, { "epoch": 3.2130807931340635, "grad_norm": 4.8125, "learning_rate": 6.798336424126645e-06, "loss": 2.3813, "step": 2715 }, { "epoch": 3.2142645753181416, "grad_norm": 4.25, "learning_rate": 6.790501229073958e-06, "loss": 1.8073, "step": 2716 }, { "epoch": 3.2154483575022197, "grad_norm": 4.53125, "learning_rate": 6.7826682299943635e-06, "loss": 2.429, "step": 2717 }, { "epoch": 3.216632139686298, "grad_norm": 4.3125, "learning_rate": 6.774837432247276e-06, "loss": 2.1236, "step": 2718 }, { "epoch": 3.217815921870376, "grad_norm": 4.96875, "learning_rate": 6.767008841190616e-06, "loss": 2.4085, "step": 2719 }, { "epoch": 3.218999704054454, "grad_norm": 4.1875, "learning_rate": 6.759182462180782e-06, "loss": 2.0795, "step": 2720 }, { "epoch": 3.220183486238532, "grad_norm": 5.03125, "learning_rate": 6.751358300572668e-06, "loss": 2.3171, "step": 2721 }, { "epoch": 3.2213672684226102, "grad_norm": 4.46875, "learning_rate": 6.743536361719651e-06, "loss": 2.2781, "step": 2722 }, { "epoch": 3.2225510506066883, "grad_norm": 4.75, "learning_rate": 6.735716650973583e-06, "loss": 2.6604, "step": 2723 }, { "epoch": 3.2237348327907664, "grad_norm": 4.59375, "learning_rate": 6.727899173684793e-06, "loss": 2.4896, "step": 2724 }, { "epoch": 3.2249186149748446, "grad_norm": 4.5, "learning_rate": 6.720083935202078e-06, "loss": 2.3391, "step": 2725 }, { "epoch": 3.2261023971589227, "grad_norm": 4.15625, "learning_rate": 6.712270940872713e-06, "loss": 2.7024, "step": 2726 }, { "epoch": 3.2272861793430008, "grad_norm": 4.75, "learning_rate": 6.70446019604243e-06, "loss": 2.2981, "step": 2727 }, { "epoch": 3.228469961527079, "grad_norm": 4.75, "learning_rate": 6.6966517060554185e-06, "loss": 2.6387, "step": 2728 }, { "epoch": 3.229653743711157, "grad_norm": 4.65625, "learning_rate": 6.688845476254337e-06, "loss": 2.2259, "step": 2729 }, { "epoch": 3.230837525895235, "grad_norm": 4.5625, "learning_rate": 6.681041511980289e-06, "loss": 2.1181, "step": 2730 }, { "epoch": 3.232021308079313, "grad_norm": 4.53125, "learning_rate": 6.673239818572826e-06, "loss": 2.3816, "step": 2731 }, { "epoch": 3.2332050902633913, "grad_norm": 4.625, "learning_rate": 6.665440401369954e-06, "loss": 2.3442, "step": 2732 }, { "epoch": 3.23438887244747, "grad_norm": 4.84375, "learning_rate": 6.6576432657081095e-06, "loss": 2.2406, "step": 2733 }, { "epoch": 3.235572654631548, "grad_norm": 4.53125, "learning_rate": 6.6498484169221865e-06, "loss": 2.4985, "step": 2734 }, { "epoch": 3.236756436815626, "grad_norm": 4.71875, "learning_rate": 6.642055860345494e-06, "loss": 2.4211, "step": 2735 }, { "epoch": 3.237940218999704, "grad_norm": 4.40625, "learning_rate": 6.634265601309787e-06, "loss": 2.3593, "step": 2736 }, { "epoch": 3.2391240011837823, "grad_norm": 5.25, "learning_rate": 6.626477645145238e-06, "loss": 2.4952, "step": 2737 }, { "epoch": 3.2403077833678604, "grad_norm": 4.5625, "learning_rate": 6.618691997180456e-06, "loss": 2.6839, "step": 2738 }, { "epoch": 3.2414915655519385, "grad_norm": 4.75, "learning_rate": 6.610908662742459e-06, "loss": 2.5885, "step": 2739 }, { "epoch": 3.2426753477360166, "grad_norm": 4.0625, "learning_rate": 6.603127647156686e-06, "loss": 1.8079, "step": 2740 }, { "epoch": 3.2438591299200947, "grad_norm": 4.59375, "learning_rate": 6.5953489557469975e-06, "loss": 2.5435, "step": 2741 }, { "epoch": 3.245042912104173, "grad_norm": 4.53125, "learning_rate": 6.587572593835649e-06, "loss": 2.1046, "step": 2742 }, { "epoch": 3.246226694288251, "grad_norm": 4.8125, "learning_rate": 6.579798566743314e-06, "loss": 2.386, "step": 2743 }, { "epoch": 3.247410476472329, "grad_norm": 5.03125, "learning_rate": 6.572026879789064e-06, "loss": 2.465, "step": 2744 }, { "epoch": 3.248594258656407, "grad_norm": 5.1875, "learning_rate": 6.564257538290364e-06, "loss": 2.5139, "step": 2745 }, { "epoch": 3.2497780408404853, "grad_norm": 4.5, "learning_rate": 6.556490547563089e-06, "loss": 2.4235, "step": 2746 }, { "epoch": 3.2509618230245634, "grad_norm": 4.75, "learning_rate": 6.548725912921489e-06, "loss": 2.4489, "step": 2747 }, { "epoch": 3.2521456052086415, "grad_norm": 4.625, "learning_rate": 6.540963639678215e-06, "loss": 2.0889, "step": 2748 }, { "epoch": 3.2533293873927196, "grad_norm": 4.625, "learning_rate": 6.53320373314429e-06, "loss": 2.3884, "step": 2749 }, { "epoch": 3.2545131695767977, "grad_norm": 4.15625, "learning_rate": 6.52544619862913e-06, "loss": 2.3286, "step": 2750 }, { "epoch": 3.255696951760876, "grad_norm": 4.53125, "learning_rate": 6.517691041440522e-06, "loss": 2.5938, "step": 2751 }, { "epoch": 3.2568807339449544, "grad_norm": 4.65625, "learning_rate": 6.5099382668846215e-06, "loss": 2.8, "step": 2752 }, { "epoch": 3.258064516129032, "grad_norm": 5.375, "learning_rate": 6.502187880265969e-06, "loss": 2.4133, "step": 2753 }, { "epoch": 3.2592482983131106, "grad_norm": 4.25, "learning_rate": 6.494439886887448e-06, "loss": 2.4492, "step": 2754 }, { "epoch": 3.2604320804971887, "grad_norm": 4.53125, "learning_rate": 6.4866942920503274e-06, "loss": 2.3945, "step": 2755 }, { "epoch": 3.261615862681267, "grad_norm": 4.09375, "learning_rate": 6.478951101054225e-06, "loss": 2.1815, "step": 2756 }, { "epoch": 3.262799644865345, "grad_norm": 5.28125, "learning_rate": 6.471210319197108e-06, "loss": 2.8635, "step": 2757 }, { "epoch": 3.263983427049423, "grad_norm": 5.25, "learning_rate": 6.4634719517753075e-06, "loss": 2.6244, "step": 2758 }, { "epoch": 3.265167209233501, "grad_norm": 4.6875, "learning_rate": 6.455736004083494e-06, "loss": 2.4256, "step": 2759 }, { "epoch": 3.2663509914175792, "grad_norm": 4.625, "learning_rate": 6.448002481414681e-06, "loss": 2.574, "step": 2760 }, { "epoch": 3.2675347736016573, "grad_norm": 4.90625, "learning_rate": 6.440271389060238e-06, "loss": 2.4306, "step": 2761 }, { "epoch": 3.2687185557857354, "grad_norm": 4.8125, "learning_rate": 6.43254273230985e-06, "loss": 2.5598, "step": 2762 }, { "epoch": 3.2699023379698136, "grad_norm": 4.84375, "learning_rate": 6.424816516451551e-06, "loss": 2.4086, "step": 2763 }, { "epoch": 3.2710861201538917, "grad_norm": 4.59375, "learning_rate": 6.417092746771693e-06, "loss": 2.3947, "step": 2764 }, { "epoch": 3.2722699023379698, "grad_norm": 4.8125, "learning_rate": 6.4093714285549736e-06, "loss": 2.1203, "step": 2765 }, { "epoch": 3.273453684522048, "grad_norm": 4.71875, "learning_rate": 6.401652567084386e-06, "loss": 2.7111, "step": 2766 }, { "epoch": 3.274637466706126, "grad_norm": 4.28125, "learning_rate": 6.393936167641266e-06, "loss": 2.2634, "step": 2767 }, { "epoch": 3.275821248890204, "grad_norm": 4.6875, "learning_rate": 6.386222235505257e-06, "loss": 2.4338, "step": 2768 }, { "epoch": 3.277005031074282, "grad_norm": 4.0625, "learning_rate": 6.378510775954305e-06, "loss": 2.2848, "step": 2769 }, { "epoch": 3.2781888132583603, "grad_norm": 4.75, "learning_rate": 6.370801794264683e-06, "loss": 2.2617, "step": 2770 }, { "epoch": 3.2793725954424384, "grad_norm": 4.3125, "learning_rate": 6.36309529571095e-06, "loss": 2.4984, "step": 2771 }, { "epoch": 3.2805563776265165, "grad_norm": 4.40625, "learning_rate": 6.355391285565975e-06, "loss": 2.2345, "step": 2772 }, { "epoch": 3.281740159810595, "grad_norm": 4.5625, "learning_rate": 6.347689769100929e-06, "loss": 2.6791, "step": 2773 }, { "epoch": 3.2829239419946727, "grad_norm": 5.0625, "learning_rate": 6.3399907515852645e-06, "loss": 2.741, "step": 2774 }, { "epoch": 3.2841077241787513, "grad_norm": 4.21875, "learning_rate": 6.332294238286736e-06, "loss": 1.9332, "step": 2775 }, { "epoch": 3.2852915063628294, "grad_norm": 4.28125, "learning_rate": 6.324600234471372e-06, "loss": 2.0935, "step": 2776 }, { "epoch": 3.2864752885469075, "grad_norm": 4.09375, "learning_rate": 6.3169087454035005e-06, "loss": 2.1015, "step": 2777 }, { "epoch": 3.2876590707309856, "grad_norm": 4.8125, "learning_rate": 6.309219776345718e-06, "loss": 2.4224, "step": 2778 }, { "epoch": 3.2888428529150637, "grad_norm": 4.4375, "learning_rate": 6.3015333325588934e-06, "loss": 2.6597, "step": 2779 }, { "epoch": 3.290026635099142, "grad_norm": 4.46875, "learning_rate": 6.293849419302179e-06, "loss": 2.2918, "step": 2780 }, { "epoch": 3.29121041728322, "grad_norm": 4.4375, "learning_rate": 6.286168041832983e-06, "loss": 2.6401, "step": 2781 }, { "epoch": 3.292394199467298, "grad_norm": 4.53125, "learning_rate": 6.278489205406992e-06, "loss": 2.5898, "step": 2782 }, { "epoch": 3.293577981651376, "grad_norm": 4.6875, "learning_rate": 6.270812915278145e-06, "loss": 2.1629, "step": 2783 }, { "epoch": 3.2947617638354543, "grad_norm": 4.65625, "learning_rate": 6.263139176698638e-06, "loss": 2.5164, "step": 2784 }, { "epoch": 3.2959455460195324, "grad_norm": 4.59375, "learning_rate": 6.25546799491893e-06, "loss": 2.2646, "step": 2785 }, { "epoch": 3.2971293282036105, "grad_norm": 5.0625, "learning_rate": 6.247799375187721e-06, "loss": 2.3319, "step": 2786 }, { "epoch": 3.2983131103876886, "grad_norm": 4.59375, "learning_rate": 6.240133322751965e-06, "loss": 2.3602, "step": 2787 }, { "epoch": 3.2994968925717667, "grad_norm": 4.78125, "learning_rate": 6.23246984285685e-06, "loss": 2.3991, "step": 2788 }, { "epoch": 3.300680674755845, "grad_norm": 3.90625, "learning_rate": 6.224808940745814e-06, "loss": 2.2652, "step": 2789 }, { "epoch": 3.301864456939923, "grad_norm": 4.8125, "learning_rate": 6.217150621660532e-06, "loss": 2.1657, "step": 2790 }, { "epoch": 3.303048239124001, "grad_norm": 4.875, "learning_rate": 6.2094948908409024e-06, "loss": 2.3951, "step": 2791 }, { "epoch": 3.304232021308079, "grad_norm": 4.5, "learning_rate": 6.201841753525059e-06, "loss": 2.4226, "step": 2792 }, { "epoch": 3.3054158034921572, "grad_norm": 5.0625, "learning_rate": 6.194191214949354e-06, "loss": 1.9181, "step": 2793 }, { "epoch": 3.306599585676236, "grad_norm": 5.03125, "learning_rate": 6.1865432803483765e-06, "loss": 2.5312, "step": 2794 }, { "epoch": 3.3077833678603135, "grad_norm": 4.3125, "learning_rate": 6.1788979549549155e-06, "loss": 2.4178, "step": 2795 }, { "epoch": 3.308967150044392, "grad_norm": 5.03125, "learning_rate": 6.171255243999988e-06, "loss": 2.5624, "step": 2796 }, { "epoch": 3.31015093222847, "grad_norm": 5.0625, "learning_rate": 6.1636151527128185e-06, "loss": 2.5753, "step": 2797 }, { "epoch": 3.3113347144125482, "grad_norm": 4.46875, "learning_rate": 6.155977686320837e-06, "loss": 2.7109, "step": 2798 }, { "epoch": 3.3125184965966263, "grad_norm": 4.875, "learning_rate": 6.148342850049682e-06, "loss": 2.3894, "step": 2799 }, { "epoch": 3.3137022787807044, "grad_norm": 4.96875, "learning_rate": 6.140710649123183e-06, "loss": 2.6395, "step": 2800 }, { "epoch": 3.3148860609647826, "grad_norm": 4.71875, "learning_rate": 6.133081088763377e-06, "loss": 1.9602, "step": 2801 }, { "epoch": 3.3160698431488607, "grad_norm": 4.53125, "learning_rate": 6.125454174190493e-06, "loss": 2.4568, "step": 2802 }, { "epoch": 3.3172536253329388, "grad_norm": 5.15625, "learning_rate": 6.117829910622942e-06, "loss": 2.5673, "step": 2803 }, { "epoch": 3.318437407517017, "grad_norm": 4.34375, "learning_rate": 6.110208303277329e-06, "loss": 2.4558, "step": 2804 }, { "epoch": 3.319621189701095, "grad_norm": 4.40625, "learning_rate": 6.102589357368434e-06, "loss": 2.1837, "step": 2805 }, { "epoch": 3.320804971885173, "grad_norm": 4.96875, "learning_rate": 6.094973078109223e-06, "loss": 2.2754, "step": 2806 }, { "epoch": 3.321988754069251, "grad_norm": 5.15625, "learning_rate": 6.087359470710841e-06, "loss": 2.8253, "step": 2807 }, { "epoch": 3.3231725362533293, "grad_norm": 5.0625, "learning_rate": 6.079748540382588e-06, "loss": 2.4083, "step": 2808 }, { "epoch": 3.3243563184374074, "grad_norm": 4.46875, "learning_rate": 6.072140292331951e-06, "loss": 1.9674, "step": 2809 }, { "epoch": 3.3255401006214855, "grad_norm": 4.21875, "learning_rate": 6.0645347317645735e-06, "loss": 2.3687, "step": 2810 }, { "epoch": 3.3267238828055636, "grad_norm": 4.5625, "learning_rate": 6.056931863884259e-06, "loss": 2.2946, "step": 2811 }, { "epoch": 3.3279076649896417, "grad_norm": 4.875, "learning_rate": 6.049331693892966e-06, "loss": 2.2534, "step": 2812 }, { "epoch": 3.32909144717372, "grad_norm": 4.75, "learning_rate": 6.041734226990819e-06, "loss": 2.3338, "step": 2813 }, { "epoch": 3.330275229357798, "grad_norm": 4.625, "learning_rate": 6.0341394683760835e-06, "loss": 2.5809, "step": 2814 }, { "epoch": 3.3314590115418765, "grad_norm": 4.46875, "learning_rate": 6.026547423245171e-06, "loss": 2.2934, "step": 2815 }, { "epoch": 3.332642793725954, "grad_norm": 4.625, "learning_rate": 6.018958096792642e-06, "loss": 2.6467, "step": 2816 }, { "epoch": 3.3338265759100327, "grad_norm": 4.5, "learning_rate": 6.011371494211189e-06, "loss": 2.3212, "step": 2817 }, { "epoch": 3.335010358094111, "grad_norm": 5.65625, "learning_rate": 6.003787620691651e-06, "loss": 2.4567, "step": 2818 }, { "epoch": 3.336194140278189, "grad_norm": 5.21875, "learning_rate": 5.996206481422997e-06, "loss": 2.5975, "step": 2819 }, { "epoch": 3.337377922462267, "grad_norm": 4.375, "learning_rate": 5.9886280815923135e-06, "loss": 2.4137, "step": 2820 }, { "epoch": 3.338561704646345, "grad_norm": 4.5625, "learning_rate": 5.9810524263848335e-06, "loss": 2.3617, "step": 2821 }, { "epoch": 3.3397454868304233, "grad_norm": 4.34375, "learning_rate": 5.973479520983892e-06, "loss": 2.5103, "step": 2822 }, { "epoch": 3.3409292690145014, "grad_norm": 4.75, "learning_rate": 5.965909370570957e-06, "loss": 2.2506, "step": 2823 }, { "epoch": 3.3421130511985795, "grad_norm": 4.4375, "learning_rate": 5.958341980325598e-06, "loss": 2.3668, "step": 2824 }, { "epoch": 3.3432968333826576, "grad_norm": 5.46875, "learning_rate": 5.950777355425511e-06, "loss": 2.4682, "step": 2825 }, { "epoch": 3.3444806155667357, "grad_norm": 4.53125, "learning_rate": 5.943215501046492e-06, "loss": 2.4182, "step": 2826 }, { "epoch": 3.345664397750814, "grad_norm": 5.0, "learning_rate": 5.935656422362438e-06, "loss": 2.2091, "step": 2827 }, { "epoch": 3.346848179934892, "grad_norm": 4.34375, "learning_rate": 5.9281001245453555e-06, "loss": 2.3592, "step": 2828 }, { "epoch": 3.34803196211897, "grad_norm": 5.1875, "learning_rate": 5.9205466127653345e-06, "loss": 2.811, "step": 2829 }, { "epoch": 3.349215744303048, "grad_norm": 5.15625, "learning_rate": 5.912995892190578e-06, "loss": 2.4968, "step": 2830 }, { "epoch": 3.3503995264871262, "grad_norm": 4.46875, "learning_rate": 5.905447967987368e-06, "loss": 2.0317, "step": 2831 }, { "epoch": 3.3515833086712044, "grad_norm": 4.71875, "learning_rate": 5.897902845320065e-06, "loss": 2.3952, "step": 2832 }, { "epoch": 3.3527670908552825, "grad_norm": 5.0, "learning_rate": 5.890360529351133e-06, "loss": 2.4324, "step": 2833 }, { "epoch": 3.3539508730393606, "grad_norm": 4.1875, "learning_rate": 5.8828210252411e-06, "loss": 2.4662, "step": 2834 }, { "epoch": 3.3551346552234387, "grad_norm": 4.21875, "learning_rate": 5.875284338148571e-06, "loss": 2.5252, "step": 2835 }, { "epoch": 3.3563184374075172, "grad_norm": 4.3125, "learning_rate": 5.867750473230236e-06, "loss": 2.1478, "step": 2836 }, { "epoch": 3.3575022195915953, "grad_norm": 4.15625, "learning_rate": 5.860219435640837e-06, "loss": 2.3234, "step": 2837 }, { "epoch": 3.3586860017756734, "grad_norm": 5.125, "learning_rate": 5.852691230533196e-06, "loss": 2.4082, "step": 2838 }, { "epoch": 3.3598697839597516, "grad_norm": 4.875, "learning_rate": 5.845165863058184e-06, "loss": 2.451, "step": 2839 }, { "epoch": 3.3610535661438297, "grad_norm": 4.75, "learning_rate": 5.837643338364744e-06, "loss": 2.5953, "step": 2840 }, { "epoch": 3.3622373483279078, "grad_norm": 4.03125, "learning_rate": 5.83012366159986e-06, "loss": 2.5093, "step": 2841 }, { "epoch": 3.363421130511986, "grad_norm": 4.375, "learning_rate": 5.8226068379085784e-06, "loss": 2.2887, "step": 2842 }, { "epoch": 3.364604912696064, "grad_norm": 4.1875, "learning_rate": 5.815092872433994e-06, "loss": 2.1759, "step": 2843 }, { "epoch": 3.365788694880142, "grad_norm": 4.8125, "learning_rate": 5.807581770317237e-06, "loss": 2.4927, "step": 2844 }, { "epoch": 3.36697247706422, "grad_norm": 4.46875, "learning_rate": 5.8000735366974845e-06, "loss": 2.1663, "step": 2845 }, { "epoch": 3.3681562592482983, "grad_norm": 4.5625, "learning_rate": 5.792568176711945e-06, "loss": 2.3966, "step": 2846 }, { "epoch": 3.3693400414323764, "grad_norm": 4.375, "learning_rate": 5.785065695495868e-06, "loss": 2.5399, "step": 2847 }, { "epoch": 3.3705238236164545, "grad_norm": 4.6875, "learning_rate": 5.7775660981825365e-06, "loss": 2.4754, "step": 2848 }, { "epoch": 3.3717076058005326, "grad_norm": 4.96875, "learning_rate": 5.7700693899032444e-06, "loss": 2.2857, "step": 2849 }, { "epoch": 3.3728913879846107, "grad_norm": 4.40625, "learning_rate": 5.762575575787333e-06, "loss": 2.3629, "step": 2850 }, { "epoch": 3.374075170168689, "grad_norm": 4.84375, "learning_rate": 5.755084660962134e-06, "loss": 2.1907, "step": 2851 }, { "epoch": 3.375258952352767, "grad_norm": 5.21875, "learning_rate": 5.74759665055302e-06, "loss": 2.706, "step": 2852 }, { "epoch": 3.376442734536845, "grad_norm": 4.59375, "learning_rate": 5.7401115496833605e-06, "loss": 2.6888, "step": 2853 }, { "epoch": 3.377626516720923, "grad_norm": 4.84375, "learning_rate": 5.732629363474544e-06, "loss": 2.7228, "step": 2854 }, { "epoch": 3.3788102989050013, "grad_norm": 4.25, "learning_rate": 5.725150097045968e-06, "loss": 2.4507, "step": 2855 }, { "epoch": 3.3799940810890794, "grad_norm": 4.25, "learning_rate": 5.717673755515021e-06, "loss": 2.4285, "step": 2856 }, { "epoch": 3.381177863273158, "grad_norm": 4.8125, "learning_rate": 5.710200343997094e-06, "loss": 2.3254, "step": 2857 }, { "epoch": 3.382361645457236, "grad_norm": 5.1875, "learning_rate": 5.702729867605571e-06, "loss": 2.9991, "step": 2858 }, { "epoch": 3.383545427641314, "grad_norm": 4.28125, "learning_rate": 5.695262331451834e-06, "loss": 2.2803, "step": 2859 }, { "epoch": 3.3847292098253923, "grad_norm": 4.59375, "learning_rate": 5.687797740645258e-06, "loss": 2.3677, "step": 2860 }, { "epoch": 3.3859129920094704, "grad_norm": 5.125, "learning_rate": 5.680336100293182e-06, "loss": 2.8972, "step": 2861 }, { "epoch": 3.3870967741935485, "grad_norm": 4.84375, "learning_rate": 5.672877415500957e-06, "loss": 2.6224, "step": 2862 }, { "epoch": 3.3882805563776266, "grad_norm": 4.71875, "learning_rate": 5.665421691371873e-06, "loss": 2.5579, "step": 2863 }, { "epoch": 3.3894643385617047, "grad_norm": 4.40625, "learning_rate": 5.6579689330072275e-06, "loss": 2.4563, "step": 2864 }, { "epoch": 3.390648120745783, "grad_norm": 4.59375, "learning_rate": 5.650519145506278e-06, "loss": 2.6941, "step": 2865 }, { "epoch": 3.391831902929861, "grad_norm": 4.375, "learning_rate": 5.6430723339662415e-06, "loss": 2.4429, "step": 2866 }, { "epoch": 3.393015685113939, "grad_norm": 4.78125, "learning_rate": 5.635628503482313e-06, "loss": 2.4156, "step": 2867 }, { "epoch": 3.394199467298017, "grad_norm": 4.5625, "learning_rate": 5.628187659147637e-06, "loss": 2.1198, "step": 2868 }, { "epoch": 3.3953832494820952, "grad_norm": 5.375, "learning_rate": 5.620749806053318e-06, "loss": 2.3903, "step": 2869 }, { "epoch": 3.3965670316661734, "grad_norm": 4.3125, "learning_rate": 5.613314949288409e-06, "loss": 2.0907, "step": 2870 }, { "epoch": 3.3977508138502515, "grad_norm": 4.65625, "learning_rate": 5.605883093939922e-06, "loss": 2.6693, "step": 2871 }, { "epoch": 3.3989345960343296, "grad_norm": 4.46875, "learning_rate": 5.598454245092816e-06, "loss": 2.6769, "step": 2872 }, { "epoch": 3.4001183782184077, "grad_norm": 4.4375, "learning_rate": 5.591028407829979e-06, "loss": 2.489, "step": 2873 }, { "epoch": 3.401302160402486, "grad_norm": 4.625, "learning_rate": 5.583605587232262e-06, "loss": 2.4129, "step": 2874 }, { "epoch": 3.402485942586564, "grad_norm": 4.78125, "learning_rate": 5.576185788378418e-06, "loss": 2.5651, "step": 2875 }, { "epoch": 3.4036697247706424, "grad_norm": 4.59375, "learning_rate": 5.568769016345163e-06, "loss": 2.0553, "step": 2876 }, { "epoch": 3.40485350695472, "grad_norm": 4.625, "learning_rate": 5.561355276207132e-06, "loss": 2.5987, "step": 2877 }, { "epoch": 3.4060372891387987, "grad_norm": 4.4375, "learning_rate": 5.55394457303688e-06, "loss": 2.5461, "step": 2878 }, { "epoch": 3.4072210713228768, "grad_norm": 4.34375, "learning_rate": 5.546536911904896e-06, "loss": 1.9614, "step": 2879 }, { "epoch": 3.408404853506955, "grad_norm": 4.75, "learning_rate": 5.539132297879575e-06, "loss": 2.1318, "step": 2880 }, { "epoch": 3.409588635691033, "grad_norm": 4.4375, "learning_rate": 5.5317307360272275e-06, "loss": 2.6735, "step": 2881 }, { "epoch": 3.410772417875111, "grad_norm": 4.8125, "learning_rate": 5.52433223141209e-06, "loss": 2.3154, "step": 2882 }, { "epoch": 3.411956200059189, "grad_norm": 4.59375, "learning_rate": 5.516936789096292e-06, "loss": 2.5809, "step": 2883 }, { "epoch": 3.4131399822432673, "grad_norm": 4.5, "learning_rate": 5.509544414139878e-06, "loss": 2.2727, "step": 2884 }, { "epoch": 3.4143237644273454, "grad_norm": 4.6875, "learning_rate": 5.5021551116007835e-06, "loss": 2.5856, "step": 2885 }, { "epoch": 3.4155075466114235, "grad_norm": 4.84375, "learning_rate": 5.494768886534858e-06, "loss": 2.3159, "step": 2886 }, { "epoch": 3.4166913287955016, "grad_norm": 5.25, "learning_rate": 5.487385743995821e-06, "loss": 2.3773, "step": 2887 }, { "epoch": 3.4178751109795797, "grad_norm": 4.03125, "learning_rate": 5.4800056890353025e-06, "loss": 2.1956, "step": 2888 }, { "epoch": 3.419058893163658, "grad_norm": 5.25, "learning_rate": 5.47262872670282e-06, "loss": 2.7977, "step": 2889 }, { "epoch": 3.420242675347736, "grad_norm": 4.84375, "learning_rate": 5.465254862045761e-06, "loss": 2.3877, "step": 2890 }, { "epoch": 3.421426457531814, "grad_norm": 4.84375, "learning_rate": 5.45788410010941e-06, "loss": 2.2725, "step": 2891 }, { "epoch": 3.422610239715892, "grad_norm": 4.25, "learning_rate": 5.450516445936915e-06, "loss": 2.6642, "step": 2892 }, { "epoch": 3.4237940218999703, "grad_norm": 4.6875, "learning_rate": 5.443151904569302e-06, "loss": 2.5676, "step": 2893 }, { "epoch": 3.4249778040840484, "grad_norm": 4.375, "learning_rate": 5.435790481045474e-06, "loss": 2.6762, "step": 2894 }, { "epoch": 3.4261615862681265, "grad_norm": 4.34375, "learning_rate": 5.428432180402188e-06, "loss": 2.618, "step": 2895 }, { "epoch": 3.4273453684522046, "grad_norm": 4.40625, "learning_rate": 5.42107700767408e-06, "loss": 2.5642, "step": 2896 }, { "epoch": 3.428529150636283, "grad_norm": 4.71875, "learning_rate": 5.4137249678936265e-06, "loss": 2.4313, "step": 2897 }, { "epoch": 3.429712932820361, "grad_norm": 4.625, "learning_rate": 5.4063760660911855e-06, "loss": 2.6091, "step": 2898 }, { "epoch": 3.4308967150044394, "grad_norm": 4.78125, "learning_rate": 5.399030307294937e-06, "loss": 2.4199, "step": 2899 }, { "epoch": 3.4320804971885175, "grad_norm": 4.5, "learning_rate": 5.391687696530934e-06, "loss": 2.5962, "step": 2900 }, { "epoch": 3.4332642793725956, "grad_norm": 5.125, "learning_rate": 5.384348238823074e-06, "loss": 2.5128, "step": 2901 }, { "epoch": 3.4344480615566737, "grad_norm": 5.25, "learning_rate": 5.377011939193084e-06, "loss": 2.4747, "step": 2902 }, { "epoch": 3.435631843740752, "grad_norm": 4.625, "learning_rate": 5.369678802660544e-06, "loss": 2.7092, "step": 2903 }, { "epoch": 3.43681562592483, "grad_norm": 4.34375, "learning_rate": 5.362348834242861e-06, "loss": 2.1631, "step": 2904 }, { "epoch": 3.437999408108908, "grad_norm": 5.25, "learning_rate": 5.3550220389552745e-06, "loss": 2.6741, "step": 2905 }, { "epoch": 3.439183190292986, "grad_norm": 4.4375, "learning_rate": 5.347698421810861e-06, "loss": 2.1295, "step": 2906 }, { "epoch": 3.4403669724770642, "grad_norm": 5.21875, "learning_rate": 5.340377987820511e-06, "loss": 2.9632, "step": 2907 }, { "epoch": 3.4415507546611424, "grad_norm": 4.84375, "learning_rate": 5.333060741992949e-06, "loss": 2.7046, "step": 2908 }, { "epoch": 3.4427345368452205, "grad_norm": 5.21875, "learning_rate": 5.325746689334707e-06, "loss": 2.4864, "step": 2909 }, { "epoch": 3.4439183190292986, "grad_norm": 4.46875, "learning_rate": 5.3184358348501415e-06, "loss": 2.3938, "step": 2910 }, { "epoch": 3.4451021012133767, "grad_norm": 4.75, "learning_rate": 5.311128183541416e-06, "loss": 2.1073, "step": 2911 }, { "epoch": 3.446285883397455, "grad_norm": 4.53125, "learning_rate": 5.3038237404084995e-06, "loss": 2.2855, "step": 2912 }, { "epoch": 3.447469665581533, "grad_norm": 4.46875, "learning_rate": 5.296522510449174e-06, "loss": 1.867, "step": 2913 }, { "epoch": 3.448653447765611, "grad_norm": 4.34375, "learning_rate": 5.2892244986590135e-06, "loss": 2.2762, "step": 2914 }, { "epoch": 3.449837229949689, "grad_norm": 4.5625, "learning_rate": 5.2819297100314e-06, "loss": 2.5611, "step": 2915 }, { "epoch": 3.451021012133767, "grad_norm": 4.8125, "learning_rate": 5.274638149557505e-06, "loss": 2.3458, "step": 2916 }, { "epoch": 3.4522047943178453, "grad_norm": 4.65625, "learning_rate": 5.267349822226284e-06, "loss": 2.4232, "step": 2917 }, { "epoch": 3.453388576501924, "grad_norm": 4.65625, "learning_rate": 5.260064733024499e-06, "loss": 2.4412, "step": 2918 }, { "epoch": 3.4545723586860015, "grad_norm": 4.71875, "learning_rate": 5.252782886936675e-06, "loss": 2.438, "step": 2919 }, { "epoch": 3.45575614087008, "grad_norm": 4.8125, "learning_rate": 5.245504288945137e-06, "loss": 2.417, "step": 2920 }, { "epoch": 3.456939923054158, "grad_norm": 4.90625, "learning_rate": 5.238228944029971e-06, "loss": 2.2227, "step": 2921 }, { "epoch": 3.4581237052382363, "grad_norm": 4.3125, "learning_rate": 5.230956857169052e-06, "loss": 2.5948, "step": 2922 }, { "epoch": 3.4593074874223144, "grad_norm": 5.28125, "learning_rate": 5.223688033338017e-06, "loss": 2.2622, "step": 2923 }, { "epoch": 3.4604912696063925, "grad_norm": 4.3125, "learning_rate": 5.216422477510267e-06, "loss": 2.123, "step": 2924 }, { "epoch": 3.4616750517904706, "grad_norm": 4.625, "learning_rate": 5.20916019465698e-06, "loss": 2.3063, "step": 2925 }, { "epoch": 3.4628588339745487, "grad_norm": 5.0, "learning_rate": 5.201901189747081e-06, "loss": 2.3203, "step": 2926 }, { "epoch": 3.464042616158627, "grad_norm": 4.78125, "learning_rate": 5.1946454677472635e-06, "loss": 2.3244, "step": 2927 }, { "epoch": 3.465226398342705, "grad_norm": 4.71875, "learning_rate": 5.187393033621967e-06, "loss": 2.461, "step": 2928 }, { "epoch": 3.466410180526783, "grad_norm": 4.875, "learning_rate": 5.1801438923333805e-06, "loss": 2.6674, "step": 2929 }, { "epoch": 3.467593962710861, "grad_norm": 4.28125, "learning_rate": 5.172898048841448e-06, "loss": 2.7249, "step": 2930 }, { "epoch": 3.4687777448949393, "grad_norm": 5.09375, "learning_rate": 5.165655508103848e-06, "loss": 2.691, "step": 2931 }, { "epoch": 3.4699615270790174, "grad_norm": 4.3125, "learning_rate": 5.1584162750760105e-06, "loss": 2.1242, "step": 2932 }, { "epoch": 3.4711453092630955, "grad_norm": 5.25, "learning_rate": 5.151180354711087e-06, "loss": 2.4948, "step": 2933 }, { "epoch": 3.4723290914471736, "grad_norm": 4.46875, "learning_rate": 5.143947751959979e-06, "loss": 2.3478, "step": 2934 }, { "epoch": 3.4735128736312517, "grad_norm": 4.46875, "learning_rate": 5.136718471771307e-06, "loss": 2.5255, "step": 2935 }, { "epoch": 3.47469665581533, "grad_norm": 4.90625, "learning_rate": 5.129492519091415e-06, "loss": 2.4584, "step": 2936 }, { "epoch": 3.475880437999408, "grad_norm": 4.1875, "learning_rate": 5.122269898864387e-06, "loss": 2.2794, "step": 2937 }, { "epoch": 3.477064220183486, "grad_norm": 4.34375, "learning_rate": 5.115050616032007e-06, "loss": 2.2711, "step": 2938 }, { "epoch": 3.4782480023675646, "grad_norm": 5.0, "learning_rate": 5.1078346755337935e-06, "loss": 2.7898, "step": 2939 }, { "epoch": 3.4794317845516423, "grad_norm": 5.0625, "learning_rate": 5.100622082306964e-06, "loss": 2.3706, "step": 2940 }, { "epoch": 3.480615566735721, "grad_norm": 4.34375, "learning_rate": 5.093412841286449e-06, "loss": 2.3373, "step": 2941 }, { "epoch": 3.481799348919799, "grad_norm": 4.78125, "learning_rate": 5.0862069574048956e-06, "loss": 2.3161, "step": 2942 }, { "epoch": 3.482983131103877, "grad_norm": 4.46875, "learning_rate": 5.079004435592636e-06, "loss": 2.4665, "step": 2943 }, { "epoch": 3.484166913287955, "grad_norm": 4.5, "learning_rate": 5.071805280777722e-06, "loss": 2.0956, "step": 2944 }, { "epoch": 3.4853506954720332, "grad_norm": 4.9375, "learning_rate": 5.064609497885883e-06, "loss": 2.3769, "step": 2945 }, { "epoch": 3.4865344776561114, "grad_norm": 5.4375, "learning_rate": 5.057417091840558e-06, "loss": 1.9634, "step": 2946 }, { "epoch": 3.4877182598401895, "grad_norm": 5.15625, "learning_rate": 5.0502280675628625e-06, "loss": 2.8807, "step": 2947 }, { "epoch": 3.4889020420242676, "grad_norm": 4.5625, "learning_rate": 5.0430424299716005e-06, "loss": 2.4776, "step": 2948 }, { "epoch": 3.4900858242083457, "grad_norm": 4.75, "learning_rate": 5.03586018398327e-06, "loss": 2.3372, "step": 2949 }, { "epoch": 3.491269606392424, "grad_norm": 4.65625, "learning_rate": 5.028681334512029e-06, "loss": 2.1777, "step": 2950 }, { "epoch": 3.492453388576502, "grad_norm": 4.625, "learning_rate": 5.021505886469733e-06, "loss": 2.4301, "step": 2951 }, { "epoch": 3.49363717076058, "grad_norm": 4.90625, "learning_rate": 5.014333844765895e-06, "loss": 2.3891, "step": 2952 }, { "epoch": 3.494820952944658, "grad_norm": 4.28125, "learning_rate": 5.007165214307699e-06, "loss": 2.1741, "step": 2953 }, { "epoch": 3.496004735128736, "grad_norm": 5.09375, "learning_rate": 5.000000000000003e-06, "loss": 2.8017, "step": 2954 }, { "epoch": 3.4971885173128143, "grad_norm": 4.84375, "learning_rate": 4.992838206745316e-06, "loss": 2.712, "step": 2955 }, { "epoch": 3.4983722994968924, "grad_norm": 4.90625, "learning_rate": 4.985679839443819e-06, "loss": 2.4151, "step": 2956 }, { "epoch": 3.4995560816809705, "grad_norm": 5.125, "learning_rate": 4.978524902993343e-06, "loss": 2.5709, "step": 2957 }, { "epoch": 3.500739863865049, "grad_norm": 4.3125, "learning_rate": 4.971373402289371e-06, "loss": 2.2906, "step": 2958 }, { "epoch": 3.5019236460491268, "grad_norm": 4.21875, "learning_rate": 4.964225342225034e-06, "loss": 2.2769, "step": 2959 }, { "epoch": 3.5031074282332053, "grad_norm": 4.46875, "learning_rate": 4.957080727691107e-06, "loss": 2.4017, "step": 2960 }, { "epoch": 3.504291210417283, "grad_norm": 5.125, "learning_rate": 4.9499395635760174e-06, "loss": 2.8595, "step": 2961 }, { "epoch": 3.5054749926013615, "grad_norm": 4.28125, "learning_rate": 4.94280185476582e-06, "loss": 2.3952, "step": 2962 }, { "epoch": 3.5066587747854396, "grad_norm": 4.46875, "learning_rate": 4.9356676061442154e-06, "loss": 2.1087, "step": 2963 }, { "epoch": 3.5078425569695177, "grad_norm": 4.84375, "learning_rate": 4.928536822592531e-06, "loss": 2.4911, "step": 2964 }, { "epoch": 3.509026339153596, "grad_norm": 4.6875, "learning_rate": 4.921409508989721e-06, "loss": 2.1223, "step": 2965 }, { "epoch": 3.510210121337674, "grad_norm": 4.6875, "learning_rate": 4.9142856702123745e-06, "loss": 2.3984, "step": 2966 }, { "epoch": 3.511393903521752, "grad_norm": 4.875, "learning_rate": 4.90716531113469e-06, "loss": 2.5082, "step": 2967 }, { "epoch": 3.51257768570583, "grad_norm": 4.09375, "learning_rate": 4.900048436628498e-06, "loss": 2.1517, "step": 2968 }, { "epoch": 3.5137614678899083, "grad_norm": 4.375, "learning_rate": 4.892935051563243e-06, "loss": 2.219, "step": 2969 }, { "epoch": 3.5149452500739864, "grad_norm": 5.4375, "learning_rate": 4.8858251608059735e-06, "loss": 2.7248, "step": 2970 }, { "epoch": 3.5161290322580645, "grad_norm": 4.78125, "learning_rate": 4.878718769221354e-06, "loss": 2.6666, "step": 2971 }, { "epoch": 3.5173128144421426, "grad_norm": 4.59375, "learning_rate": 4.8716158816716476e-06, "loss": 2.4636, "step": 2972 }, { "epoch": 3.5184965966262207, "grad_norm": 4.28125, "learning_rate": 4.864516503016733e-06, "loss": 2.2017, "step": 2973 }, { "epoch": 3.519680378810299, "grad_norm": 4.8125, "learning_rate": 4.857420638114073e-06, "loss": 2.4524, "step": 2974 }, { "epoch": 3.520864160994377, "grad_norm": 4.34375, "learning_rate": 4.850328291818742e-06, "loss": 2.3789, "step": 2975 }, { "epoch": 3.522047943178455, "grad_norm": 4.625, "learning_rate": 4.843239468983394e-06, "loss": 2.4771, "step": 2976 }, { "epoch": 3.523231725362533, "grad_norm": 4.96875, "learning_rate": 4.8361541744582715e-06, "loss": 2.2777, "step": 2977 }, { "epoch": 3.5244155075466113, "grad_norm": 4.53125, "learning_rate": 4.829072413091219e-06, "loss": 2.1586, "step": 2978 }, { "epoch": 3.52559928973069, "grad_norm": 5.0625, "learning_rate": 4.8219941897276425e-06, "loss": 2.703, "step": 2979 }, { "epoch": 3.5267830719147675, "grad_norm": 4.84375, "learning_rate": 4.8149195092105425e-06, "loss": 2.7423, "step": 2980 }, { "epoch": 3.527966854098846, "grad_norm": 5.15625, "learning_rate": 4.807848376380494e-06, "loss": 2.6297, "step": 2981 }, { "epoch": 3.5291506362829237, "grad_norm": 4.625, "learning_rate": 4.800780796075637e-06, "loss": 2.3929, "step": 2982 }, { "epoch": 3.5303344184670022, "grad_norm": 4.3125, "learning_rate": 4.793716773131685e-06, "loss": 2.3509, "step": 2983 }, { "epoch": 3.5315182006510804, "grad_norm": 4.75, "learning_rate": 4.786656312381913e-06, "loss": 2.923, "step": 2984 }, { "epoch": 3.5327019828351585, "grad_norm": 4.9375, "learning_rate": 4.779599418657168e-06, "loss": 2.2542, "step": 2985 }, { "epoch": 3.5338857650192366, "grad_norm": 5.0625, "learning_rate": 4.772546096785854e-06, "loss": 2.4777, "step": 2986 }, { "epoch": 3.5350695472033147, "grad_norm": 4.78125, "learning_rate": 4.765496351593927e-06, "loss": 2.3795, "step": 2987 }, { "epoch": 3.536253329387393, "grad_norm": 4.71875, "learning_rate": 4.758450187904895e-06, "loss": 2.4923, "step": 2988 }, { "epoch": 3.537437111571471, "grad_norm": 5.40625, "learning_rate": 4.751407610539815e-06, "loss": 2.6404, "step": 2989 }, { "epoch": 3.538620893755549, "grad_norm": 4.8125, "learning_rate": 4.7443686243173015e-06, "loss": 2.5426, "step": 2990 }, { "epoch": 3.539804675939627, "grad_norm": 4.78125, "learning_rate": 4.737333234053494e-06, "loss": 2.7189, "step": 2991 }, { "epoch": 3.540988458123705, "grad_norm": 4.78125, "learning_rate": 4.730301444562088e-06, "loss": 2.3333, "step": 2992 }, { "epoch": 3.5421722403077833, "grad_norm": 4.9375, "learning_rate": 4.7232732606543085e-06, "loss": 2.2335, "step": 2993 }, { "epoch": 3.5433560224918614, "grad_norm": 5.09375, "learning_rate": 4.7162486871389125e-06, "loss": 2.1636, "step": 2994 }, { "epoch": 3.5445398046759395, "grad_norm": 5.0625, "learning_rate": 4.7092277288221865e-06, "loss": 2.4844, "step": 2995 }, { "epoch": 3.5457235868600177, "grad_norm": 4.25, "learning_rate": 4.7022103905079405e-06, "loss": 2.2663, "step": 2996 }, { "epoch": 3.5469073690440958, "grad_norm": 4.65625, "learning_rate": 4.695196676997517e-06, "loss": 2.4574, "step": 2997 }, { "epoch": 3.548091151228174, "grad_norm": 4.65625, "learning_rate": 4.6881865930897756e-06, "loss": 2.6026, "step": 2998 }, { "epoch": 3.549274933412252, "grad_norm": 5.40625, "learning_rate": 4.681180143581086e-06, "loss": 2.4024, "step": 2999 }, { "epoch": 3.5504587155963305, "grad_norm": 4.875, "learning_rate": 4.674177333265336e-06, "loss": 2.4098, "step": 3000 }, { "epoch": 3.551642497780408, "grad_norm": 4.75, "learning_rate": 4.66717816693392e-06, "loss": 2.5078, "step": 3001 }, { "epoch": 3.5528262799644867, "grad_norm": 4.84375, "learning_rate": 4.660182649375747e-06, "loss": 2.3451, "step": 3002 }, { "epoch": 3.5540100621485644, "grad_norm": 5.09375, "learning_rate": 4.653190785377218e-06, "loss": 2.4108, "step": 3003 }, { "epoch": 3.555193844332643, "grad_norm": 5.28125, "learning_rate": 4.6462025797222445e-06, "loss": 2.2514, "step": 3004 }, { "epoch": 3.556377626516721, "grad_norm": 4.78125, "learning_rate": 4.639218037192235e-06, "loss": 2.245, "step": 3005 }, { "epoch": 3.557561408700799, "grad_norm": 4.53125, "learning_rate": 4.6322371625660825e-06, "loss": 2.3657, "step": 3006 }, { "epoch": 3.5587451908848773, "grad_norm": 4.4375, "learning_rate": 4.6252599606201755e-06, "loss": 2.2657, "step": 3007 }, { "epoch": 3.5599289730689554, "grad_norm": 4.09375, "learning_rate": 4.618286436128386e-06, "loss": 2.133, "step": 3008 }, { "epoch": 3.5611127552530335, "grad_norm": 4.53125, "learning_rate": 4.6113165938620785e-06, "loss": 2.5018, "step": 3009 }, { "epoch": 3.5622965374371116, "grad_norm": 4.34375, "learning_rate": 4.604350438590095e-06, "loss": 2.6418, "step": 3010 }, { "epoch": 3.5634803196211897, "grad_norm": 5.375, "learning_rate": 4.597387975078751e-06, "loss": 2.2786, "step": 3011 }, { "epoch": 3.564664101805268, "grad_norm": 4.75, "learning_rate": 4.5904292080918355e-06, "loss": 2.6621, "step": 3012 }, { "epoch": 3.565847883989346, "grad_norm": 4.28125, "learning_rate": 4.583474142390608e-06, "loss": 2.5398, "step": 3013 }, { "epoch": 3.567031666173424, "grad_norm": 4.53125, "learning_rate": 4.576522782733802e-06, "loss": 2.504, "step": 3014 }, { "epoch": 3.568215448357502, "grad_norm": 4.65625, "learning_rate": 4.569575133877615e-06, "loss": 2.5355, "step": 3015 }, { "epoch": 3.5693992305415803, "grad_norm": 5.0, "learning_rate": 4.562631200575696e-06, "loss": 2.9229, "step": 3016 }, { "epoch": 3.5705830127256584, "grad_norm": 4.59375, "learning_rate": 4.555690987579162e-06, "loss": 2.109, "step": 3017 }, { "epoch": 3.5717667949097365, "grad_norm": 4.65625, "learning_rate": 4.54875449963658e-06, "loss": 2.2589, "step": 3018 }, { "epoch": 3.5729505770938146, "grad_norm": 5.40625, "learning_rate": 4.541821741493967e-06, "loss": 2.5289, "step": 3019 }, { "epoch": 3.5741343592778927, "grad_norm": 4.84375, "learning_rate": 4.534892717894785e-06, "loss": 2.6407, "step": 3020 }, { "epoch": 3.5753181414619712, "grad_norm": 4.96875, "learning_rate": 4.527967433579952e-06, "loss": 2.6886, "step": 3021 }, { "epoch": 3.576501923646049, "grad_norm": 4.90625, "learning_rate": 4.521045893287821e-06, "loss": 2.5637, "step": 3022 }, { "epoch": 3.5776857058301275, "grad_norm": 4.71875, "learning_rate": 4.514128101754183e-06, "loss": 2.6671, "step": 3023 }, { "epoch": 3.578869488014205, "grad_norm": 5.0625, "learning_rate": 4.507214063712262e-06, "loss": 2.4949, "step": 3024 }, { "epoch": 3.5800532701982837, "grad_norm": 4.25, "learning_rate": 4.500303783892712e-06, "loss": 2.5029, "step": 3025 }, { "epoch": 3.581237052382362, "grad_norm": 4.6875, "learning_rate": 4.493397267023626e-06, "loss": 2.2327, "step": 3026 }, { "epoch": 3.58242083456644, "grad_norm": 4.78125, "learning_rate": 4.486494517830519e-06, "loss": 2.4028, "step": 3027 }, { "epoch": 3.583604616750518, "grad_norm": 5.03125, "learning_rate": 4.479595541036316e-06, "loss": 2.585, "step": 3028 }, { "epoch": 3.584788398934596, "grad_norm": 4.8125, "learning_rate": 4.472700341361378e-06, "loss": 2.3665, "step": 3029 }, { "epoch": 3.585972181118674, "grad_norm": 4.5, "learning_rate": 4.465808923523471e-06, "loss": 2.5724, "step": 3030 }, { "epoch": 3.5871559633027523, "grad_norm": 4.4375, "learning_rate": 4.458921292237773e-06, "loss": 2.3797, "step": 3031 }, { "epoch": 3.5883397454868304, "grad_norm": 4.75, "learning_rate": 4.45203745221688e-06, "loss": 2.4689, "step": 3032 }, { "epoch": 3.5895235276709085, "grad_norm": 4.9375, "learning_rate": 4.445157408170781e-06, "loss": 2.7495, "step": 3033 }, { "epoch": 3.5907073098549867, "grad_norm": 5.34375, "learning_rate": 4.438281164806885e-06, "loss": 2.5591, "step": 3034 }, { "epoch": 3.5918910920390648, "grad_norm": 4.59375, "learning_rate": 4.431408726829984e-06, "loss": 2.7009, "step": 3035 }, { "epoch": 3.593074874223143, "grad_norm": 4.375, "learning_rate": 4.424540098942275e-06, "loss": 2.127, "step": 3036 }, { "epoch": 3.594258656407221, "grad_norm": 5.0625, "learning_rate": 4.41767528584334e-06, "loss": 2.4841, "step": 3037 }, { "epoch": 3.595442438591299, "grad_norm": 4.78125, "learning_rate": 4.410814292230163e-06, "loss": 2.4974, "step": 3038 }, { "epoch": 3.596626220775377, "grad_norm": 4.65625, "learning_rate": 4.403957122797111e-06, "loss": 2.2256, "step": 3039 }, { "epoch": 3.5978100029594553, "grad_norm": 4.53125, "learning_rate": 4.397103782235925e-06, "loss": 2.6373, "step": 3040 }, { "epoch": 3.5989937851435334, "grad_norm": 4.625, "learning_rate": 4.3902542752357415e-06, "loss": 2.3206, "step": 3041 }, { "epoch": 3.600177567327612, "grad_norm": 4.59375, "learning_rate": 4.383408606483061e-06, "loss": 2.424, "step": 3042 }, { "epoch": 3.6013613495116896, "grad_norm": 4.53125, "learning_rate": 4.376566780661757e-06, "loss": 2.1114, "step": 3043 }, { "epoch": 3.602545131695768, "grad_norm": 4.46875, "learning_rate": 4.369728802453091e-06, "loss": 2.4699, "step": 3044 }, { "epoch": 3.603728913879846, "grad_norm": 4.3125, "learning_rate": 4.36289467653567e-06, "loss": 2.4146, "step": 3045 }, { "epoch": 3.6049126960639244, "grad_norm": 4.65625, "learning_rate": 4.356064407585484e-06, "loss": 2.3935, "step": 3046 }, { "epoch": 3.6060964782480025, "grad_norm": 4.46875, "learning_rate": 4.349238000275871e-06, "loss": 2.4792, "step": 3047 }, { "epoch": 3.6072802604320806, "grad_norm": 5.5625, "learning_rate": 4.34241545927753e-06, "loss": 2.4266, "step": 3048 }, { "epoch": 3.6084640426161587, "grad_norm": 5.34375, "learning_rate": 4.335596789258515e-06, "loss": 2.3415, "step": 3049 }, { "epoch": 3.609647824800237, "grad_norm": 4.96875, "learning_rate": 4.3287819948842334e-06, "loss": 2.6145, "step": 3050 }, { "epoch": 3.610831606984315, "grad_norm": 5.03125, "learning_rate": 4.3219710808174465e-06, "loss": 2.5088, "step": 3051 }, { "epoch": 3.612015389168393, "grad_norm": 5.21875, "learning_rate": 4.315164051718243e-06, "loss": 2.3685, "step": 3052 }, { "epoch": 3.613199171352471, "grad_norm": 5.28125, "learning_rate": 4.308360912244074e-06, "loss": 2.3952, "step": 3053 }, { "epoch": 3.6143829535365493, "grad_norm": 4.71875, "learning_rate": 4.301561667049716e-06, "loss": 2.7124, "step": 3054 }, { "epoch": 3.6155667357206274, "grad_norm": 5.125, "learning_rate": 4.2947663207872804e-06, "loss": 2.5739, "step": 3055 }, { "epoch": 3.6167505179047055, "grad_norm": 4.875, "learning_rate": 4.287974878106222e-06, "loss": 2.4789, "step": 3056 }, { "epoch": 3.6179343000887836, "grad_norm": 4.78125, "learning_rate": 4.2811873436533116e-06, "loss": 2.3881, "step": 3057 }, { "epoch": 3.6191180822728617, "grad_norm": 4.90625, "learning_rate": 4.274403722072658e-06, "loss": 2.2032, "step": 3058 }, { "epoch": 3.62030186445694, "grad_norm": 4.21875, "learning_rate": 4.267624018005686e-06, "loss": 1.9781, "step": 3059 }, { "epoch": 3.621485646641018, "grad_norm": 4.71875, "learning_rate": 4.260848236091135e-06, "loss": 2.3201, "step": 3060 }, { "epoch": 3.622669428825096, "grad_norm": 4.53125, "learning_rate": 4.2540763809650745e-06, "loss": 2.274, "step": 3061 }, { "epoch": 3.623853211009174, "grad_norm": 4.84375, "learning_rate": 4.247308457260874e-06, "loss": 2.3996, "step": 3062 }, { "epoch": 3.6250369931932527, "grad_norm": 4.59375, "learning_rate": 4.240544469609222e-06, "loss": 2.1842, "step": 3063 }, { "epoch": 3.6262207753773303, "grad_norm": 4.65625, "learning_rate": 4.2337844226381085e-06, "loss": 2.873, "step": 3064 }, { "epoch": 3.627404557561409, "grad_norm": 4.71875, "learning_rate": 4.227028320972832e-06, "loss": 2.4926, "step": 3065 }, { "epoch": 3.6285883397454866, "grad_norm": 5.09375, "learning_rate": 4.22027616923599e-06, "loss": 2.6304, "step": 3066 }, { "epoch": 3.629772121929565, "grad_norm": 4.875, "learning_rate": 4.2135279720474675e-06, "loss": 2.5721, "step": 3067 }, { "epoch": 3.630955904113643, "grad_norm": 4.5, "learning_rate": 4.206783734024463e-06, "loss": 2.5038, "step": 3068 }, { "epoch": 3.6321396862977213, "grad_norm": 4.4375, "learning_rate": 4.200043459781448e-06, "loss": 2.2993, "step": 3069 }, { "epoch": 3.6333234684817994, "grad_norm": 4.78125, "learning_rate": 4.1933071539301965e-06, "loss": 2.1101, "step": 3070 }, { "epoch": 3.6345072506658775, "grad_norm": 5.375, "learning_rate": 4.186574821079755e-06, "loss": 2.5541, "step": 3071 }, { "epoch": 3.6356910328499557, "grad_norm": 4.375, "learning_rate": 4.1798464658364566e-06, "loss": 2.0293, "step": 3072 }, { "epoch": 3.6368748150340338, "grad_norm": 3.875, "learning_rate": 4.17312209280392e-06, "loss": 2.3955, "step": 3073 }, { "epoch": 3.638058597218112, "grad_norm": 4.71875, "learning_rate": 4.1664017065830235e-06, "loss": 2.3839, "step": 3074 }, { "epoch": 3.63924237940219, "grad_norm": 4.8125, "learning_rate": 4.159685311771935e-06, "loss": 2.2192, "step": 3075 }, { "epoch": 3.640426161586268, "grad_norm": 4.6875, "learning_rate": 4.152972912966075e-06, "loss": 2.5038, "step": 3076 }, { "epoch": 3.641609943770346, "grad_norm": 4.78125, "learning_rate": 4.1462645147581456e-06, "loss": 2.6362, "step": 3077 }, { "epoch": 3.6427937259544243, "grad_norm": 4.40625, "learning_rate": 4.139560121738101e-06, "loss": 2.1143, "step": 3078 }, { "epoch": 3.6439775081385024, "grad_norm": 4.625, "learning_rate": 4.132859738493154e-06, "loss": 2.5035, "step": 3079 }, { "epoch": 3.6451612903225805, "grad_norm": 4.375, "learning_rate": 4.126163369607784e-06, "loss": 2.4594, "step": 3080 }, { "epoch": 3.6463450725066586, "grad_norm": 5.03125, "learning_rate": 4.119471019663712e-06, "loss": 2.4875, "step": 3081 }, { "epoch": 3.647528854690737, "grad_norm": 4.625, "learning_rate": 4.112782693239922e-06, "loss": 2.3913, "step": 3082 }, { "epoch": 3.648712636874815, "grad_norm": 4.59375, "learning_rate": 4.106098394912632e-06, "loss": 2.533, "step": 3083 }, { "epoch": 3.6498964190588934, "grad_norm": 4.71875, "learning_rate": 4.099418129255309e-06, "loss": 2.2082, "step": 3084 }, { "epoch": 3.651080201242971, "grad_norm": 5.59375, "learning_rate": 4.092741900838667e-06, "loss": 2.6034, "step": 3085 }, { "epoch": 3.6522639834270496, "grad_norm": 4.6875, "learning_rate": 4.086069714230646e-06, "loss": 2.5189, "step": 3086 }, { "epoch": 3.6534477656111277, "grad_norm": 4.65625, "learning_rate": 4.079401573996433e-06, "loss": 2.2693, "step": 3087 }, { "epoch": 3.654631547795206, "grad_norm": 4.375, "learning_rate": 4.072737484698435e-06, "loss": 2.5964, "step": 3088 }, { "epoch": 3.655815329979284, "grad_norm": 4.78125, "learning_rate": 4.0660774508962964e-06, "loss": 2.2052, "step": 3089 }, { "epoch": 3.656999112163362, "grad_norm": 4.75, "learning_rate": 4.059421477146882e-06, "loss": 2.2372, "step": 3090 }, { "epoch": 3.65818289434744, "grad_norm": 4.75, "learning_rate": 4.052769568004277e-06, "loss": 2.5152, "step": 3091 }, { "epoch": 3.6593666765315183, "grad_norm": 4.53125, "learning_rate": 4.046121728019792e-06, "loss": 2.2539, "step": 3092 }, { "epoch": 3.6605504587155964, "grad_norm": 4.71875, "learning_rate": 4.039477961741944e-06, "loss": 2.3604, "step": 3093 }, { "epoch": 3.6617342408996745, "grad_norm": 4.625, "learning_rate": 4.032838273716476e-06, "loss": 2.162, "step": 3094 }, { "epoch": 3.6629180230837526, "grad_norm": 4.375, "learning_rate": 4.0262026684863295e-06, "loss": 2.3328, "step": 3095 }, { "epoch": 3.6641018052678307, "grad_norm": 5.0625, "learning_rate": 4.019571150591653e-06, "loss": 2.5786, "step": 3096 }, { "epoch": 3.665285587451909, "grad_norm": 4.34375, "learning_rate": 4.012943724569806e-06, "loss": 2.3763, "step": 3097 }, { "epoch": 3.666469369635987, "grad_norm": 5.3125, "learning_rate": 4.006320394955337e-06, "loss": 2.4163, "step": 3098 }, { "epoch": 3.667653151820065, "grad_norm": 4.6875, "learning_rate": 3.999701166280008e-06, "loss": 2.4951, "step": 3099 }, { "epoch": 3.668836934004143, "grad_norm": 4.78125, "learning_rate": 3.993086043072756e-06, "loss": 2.3472, "step": 3100 }, { "epoch": 3.6700207161882212, "grad_norm": 4.15625, "learning_rate": 3.9864750298597255e-06, "loss": 2.2153, "step": 3101 }, { "epoch": 3.6712044983722993, "grad_norm": 4.75, "learning_rate": 3.979868131164241e-06, "loss": 2.5759, "step": 3102 }, { "epoch": 3.672388280556378, "grad_norm": 4.8125, "learning_rate": 3.973265351506805e-06, "loss": 2.436, "step": 3103 }, { "epoch": 3.6735720627404556, "grad_norm": 4.625, "learning_rate": 3.96666669540512e-06, "loss": 1.9787, "step": 3104 }, { "epoch": 3.674755844924534, "grad_norm": 4.375, "learning_rate": 3.960072167374047e-06, "loss": 2.3792, "step": 3105 }, { "epoch": 3.6759396271086118, "grad_norm": 4.1875, "learning_rate": 3.953481771925641e-06, "loss": 2.7783, "step": 3106 }, { "epoch": 3.6771234092926903, "grad_norm": 4.46875, "learning_rate": 3.946895513569116e-06, "loss": 2.2706, "step": 3107 }, { "epoch": 3.6783071914767684, "grad_norm": 5.0625, "learning_rate": 3.940313396810855e-06, "loss": 2.2945, "step": 3108 }, { "epoch": 3.6794909736608465, "grad_norm": 4.28125, "learning_rate": 3.933735426154421e-06, "loss": 2.5213, "step": 3109 }, { "epoch": 3.6806747558449247, "grad_norm": 4.625, "learning_rate": 3.9271616061005235e-06, "loss": 2.726, "step": 3110 }, { "epoch": 3.6818585380290028, "grad_norm": 4.15625, "learning_rate": 3.920591941147047e-06, "loss": 1.9649, "step": 3111 }, { "epoch": 3.683042320213081, "grad_norm": 4.28125, "learning_rate": 3.914026435789019e-06, "loss": 2.4495, "step": 3112 }, { "epoch": 3.684226102397159, "grad_norm": 4.625, "learning_rate": 3.907465094518636e-06, "loss": 2.7822, "step": 3113 }, { "epoch": 3.685409884581237, "grad_norm": 4.65625, "learning_rate": 3.90090792182523e-06, "loss": 2.5324, "step": 3114 }, { "epoch": 3.686593666765315, "grad_norm": 4.625, "learning_rate": 3.894354922195288e-06, "loss": 2.5463, "step": 3115 }, { "epoch": 3.6877774489493933, "grad_norm": 4.59375, "learning_rate": 3.887806100112449e-06, "loss": 2.3658, "step": 3116 }, { "epoch": 3.6889612311334714, "grad_norm": 4.6875, "learning_rate": 3.881261460057477e-06, "loss": 2.8263, "step": 3117 }, { "epoch": 3.6901450133175495, "grad_norm": 4.1875, "learning_rate": 3.874721006508293e-06, "loss": 2.138, "step": 3118 }, { "epoch": 3.6913287955016276, "grad_norm": 4.59375, "learning_rate": 3.868184743939941e-06, "loss": 2.7329, "step": 3119 }, { "epoch": 3.6925125776857057, "grad_norm": 4.5, "learning_rate": 3.861652676824597e-06, "loss": 2.361, "step": 3120 }, { "epoch": 3.693696359869784, "grad_norm": 4.8125, "learning_rate": 3.8551248096315785e-06, "loss": 2.4214, "step": 3121 }, { "epoch": 3.694880142053862, "grad_norm": 4.71875, "learning_rate": 3.8486011468273145e-06, "loss": 2.3995, "step": 3122 }, { "epoch": 3.69606392423794, "grad_norm": 4.65625, "learning_rate": 3.842081692875369e-06, "loss": 2.533, "step": 3123 }, { "epoch": 3.6972477064220186, "grad_norm": 4.1875, "learning_rate": 3.835566452236416e-06, "loss": 2.2243, "step": 3124 }, { "epoch": 3.6984314886060963, "grad_norm": 4.78125, "learning_rate": 3.8290554293682614e-06, "loss": 2.3228, "step": 3125 }, { "epoch": 3.699615270790175, "grad_norm": 4.78125, "learning_rate": 3.82254862872581e-06, "loss": 2.6037, "step": 3126 }, { "epoch": 3.7007990529742525, "grad_norm": 4.6875, "learning_rate": 3.8160460547610785e-06, "loss": 2.2385, "step": 3127 }, { "epoch": 3.701982835158331, "grad_norm": 4.75, "learning_rate": 3.8095477119232092e-06, "loss": 2.8516, "step": 3128 }, { "epoch": 3.703166617342409, "grad_norm": 3.90625, "learning_rate": 3.8030536046584253e-06, "loss": 2.3177, "step": 3129 }, { "epoch": 3.7043503995264873, "grad_norm": 4.59375, "learning_rate": 3.796563737410074e-06, "loss": 2.1202, "step": 3130 }, { "epoch": 3.7055341817105654, "grad_norm": 4.8125, "learning_rate": 3.790078114618586e-06, "loss": 2.7272, "step": 3131 }, { "epoch": 3.7067179638946435, "grad_norm": 4.1875, "learning_rate": 3.7835967407214913e-06, "loss": 2.1526, "step": 3132 }, { "epoch": 3.7079017460787216, "grad_norm": 4.4375, "learning_rate": 3.77711962015342e-06, "loss": 2.6296, "step": 3133 }, { "epoch": 3.7090855282627997, "grad_norm": 4.625, "learning_rate": 3.770646757346079e-06, "loss": 2.6426, "step": 3134 }, { "epoch": 3.710269310446878, "grad_norm": 4.6875, "learning_rate": 3.7641781567282767e-06, "loss": 2.3539, "step": 3135 }, { "epoch": 3.711453092630956, "grad_norm": 4.90625, "learning_rate": 3.757713822725898e-06, "loss": 2.6401, "step": 3136 }, { "epoch": 3.712636874815034, "grad_norm": 5.375, "learning_rate": 3.751253759761907e-06, "loss": 2.6785, "step": 3137 }, { "epoch": 3.713820656999112, "grad_norm": 4.4375, "learning_rate": 3.7447979722563464e-06, "loss": 2.5014, "step": 3138 }, { "epoch": 3.7150044391831902, "grad_norm": 4.71875, "learning_rate": 3.7383464646263324e-06, "loss": 2.5923, "step": 3139 }, { "epoch": 3.7161882213672683, "grad_norm": 4.9375, "learning_rate": 3.731899241286061e-06, "loss": 2.6988, "step": 3140 }, { "epoch": 3.7173720035513464, "grad_norm": 4.625, "learning_rate": 3.725456306646784e-06, "loss": 2.4091, "step": 3141 }, { "epoch": 3.7185557857354246, "grad_norm": 5.03125, "learning_rate": 3.719017665116833e-06, "loss": 2.5312, "step": 3142 }, { "epoch": 3.7197395679195027, "grad_norm": 4.71875, "learning_rate": 3.7125833211015895e-06, "loss": 2.4382, "step": 3143 }, { "epoch": 3.7209233501035808, "grad_norm": 4.84375, "learning_rate": 3.706153279003498e-06, "loss": 2.5747, "step": 3144 }, { "epoch": 3.7221071322876593, "grad_norm": 5.125, "learning_rate": 3.699727543222069e-06, "loss": 2.6457, "step": 3145 }, { "epoch": 3.723290914471737, "grad_norm": 4.34375, "learning_rate": 3.69330611815385e-06, "loss": 2.2734, "step": 3146 }, { "epoch": 3.7244746966558155, "grad_norm": 4.34375, "learning_rate": 3.6868890081924537e-06, "loss": 2.3492, "step": 3147 }, { "epoch": 3.725658478839893, "grad_norm": 4.75, "learning_rate": 3.680476217728537e-06, "loss": 2.1152, "step": 3148 }, { "epoch": 3.7268422610239718, "grad_norm": 4.5625, "learning_rate": 3.6740677511497958e-06, "loss": 2.3454, "step": 3149 }, { "epoch": 3.72802604320805, "grad_norm": 4.8125, "learning_rate": 3.6676636128409713e-06, "loss": 2.6799, "step": 3150 }, { "epoch": 3.729209825392128, "grad_norm": 4.65625, "learning_rate": 3.6612638071838393e-06, "loss": 2.3442, "step": 3151 }, { "epoch": 3.730393607576206, "grad_norm": 4.96875, "learning_rate": 3.6548683385572215e-06, "loss": 2.486, "step": 3152 }, { "epoch": 3.731577389760284, "grad_norm": 4.5625, "learning_rate": 3.648477211336958e-06, "loss": 2.4508, "step": 3153 }, { "epoch": 3.7327611719443623, "grad_norm": 4.5625, "learning_rate": 3.6420904298959336e-06, "loss": 2.2577, "step": 3154 }, { "epoch": 3.7339449541284404, "grad_norm": 4.40625, "learning_rate": 3.6357079986040465e-06, "loss": 1.9562, "step": 3155 }, { "epoch": 3.7351287363125185, "grad_norm": 4.8125, "learning_rate": 3.629329921828224e-06, "loss": 2.6723, "step": 3156 }, { "epoch": 3.7363125184965966, "grad_norm": 4.84375, "learning_rate": 3.622956203932417e-06, "loss": 2.7883, "step": 3157 }, { "epoch": 3.7374963006806747, "grad_norm": 4.21875, "learning_rate": 3.616586849277587e-06, "loss": 2.4617, "step": 3158 }, { "epoch": 3.738680082864753, "grad_norm": 5.375, "learning_rate": 3.6102218622217143e-06, "loss": 2.6011, "step": 3159 }, { "epoch": 3.739863865048831, "grad_norm": 4.90625, "learning_rate": 3.6038612471197963e-06, "loss": 2.5285, "step": 3160 }, { "epoch": 3.741047647232909, "grad_norm": 4.21875, "learning_rate": 3.5975050083238294e-06, "loss": 2.4135, "step": 3161 }, { "epoch": 3.742231429416987, "grad_norm": 5.0, "learning_rate": 3.59115315018282e-06, "loss": 2.7593, "step": 3162 }, { "epoch": 3.7434152116010653, "grad_norm": 4.5625, "learning_rate": 3.584805677042772e-06, "loss": 2.7156, "step": 3163 }, { "epoch": 3.7445989937851434, "grad_norm": 4.5625, "learning_rate": 3.5784625932466978e-06, "loss": 2.5849, "step": 3164 }, { "epoch": 3.7457827759692215, "grad_norm": 4.9375, "learning_rate": 3.5721239031346067e-06, "loss": 2.9121, "step": 3165 }, { "epoch": 3.7469665581533, "grad_norm": 4.875, "learning_rate": 3.5657896110434942e-06, "loss": 2.4037, "step": 3166 }, { "epoch": 3.7481503403373777, "grad_norm": 5.5, "learning_rate": 3.559459721307349e-06, "loss": 2.2254, "step": 3167 }, { "epoch": 3.7493341225214563, "grad_norm": 5.0625, "learning_rate": 3.5531342382571467e-06, "loss": 2.515, "step": 3168 }, { "epoch": 3.750517904705534, "grad_norm": 4.375, "learning_rate": 3.5468131662208547e-06, "loss": 2.6784, "step": 3169 }, { "epoch": 3.7517016868896125, "grad_norm": 4.53125, "learning_rate": 3.5404965095234136e-06, "loss": 2.615, "step": 3170 }, { "epoch": 3.7528854690736906, "grad_norm": 4.125, "learning_rate": 3.534184272486747e-06, "loss": 1.9687, "step": 3171 }, { "epoch": 3.7540692512577687, "grad_norm": 5.375, "learning_rate": 3.5278764594297608e-06, "loss": 2.4934, "step": 3172 }, { "epoch": 3.755253033441847, "grad_norm": 4.28125, "learning_rate": 3.521573074668323e-06, "loss": 2.3701, "step": 3173 }, { "epoch": 3.756436815625925, "grad_norm": 4.5625, "learning_rate": 3.5152741225152755e-06, "loss": 2.1919, "step": 3174 }, { "epoch": 3.757620597810003, "grad_norm": 4.4375, "learning_rate": 3.508979607280426e-06, "loss": 2.6189, "step": 3175 }, { "epoch": 3.758804379994081, "grad_norm": 4.625, "learning_rate": 3.5026895332705504e-06, "loss": 2.4911, "step": 3176 }, { "epoch": 3.7599881621781592, "grad_norm": 5.03125, "learning_rate": 3.4964039047893893e-06, "loss": 2.3854, "step": 3177 }, { "epoch": 3.7611719443622373, "grad_norm": 4.96875, "learning_rate": 3.4901227261376326e-06, "loss": 2.6171, "step": 3178 }, { "epoch": 3.7623557265463154, "grad_norm": 4.59375, "learning_rate": 3.4838460016129273e-06, "loss": 2.2677, "step": 3179 }, { "epoch": 3.7635395087303936, "grad_norm": 4.875, "learning_rate": 3.4775737355098737e-06, "loss": 2.4758, "step": 3180 }, { "epoch": 3.7647232909144717, "grad_norm": 4.5625, "learning_rate": 3.4713059321200293e-06, "loss": 2.1977, "step": 3181 }, { "epoch": 3.7659070730985498, "grad_norm": 5.03125, "learning_rate": 3.4650425957318846e-06, "loss": 2.2007, "step": 3182 }, { "epoch": 3.767090855282628, "grad_norm": 4.40625, "learning_rate": 3.4587837306308823e-06, "loss": 2.1708, "step": 3183 }, { "epoch": 3.768274637466706, "grad_norm": 5.0625, "learning_rate": 3.4525293410994108e-06, "loss": 2.5975, "step": 3184 }, { "epoch": 3.769458419650784, "grad_norm": 4.75, "learning_rate": 3.4462794314167846e-06, "loss": 2.2244, "step": 3185 }, { "epoch": 3.770642201834862, "grad_norm": 4.90625, "learning_rate": 3.4400340058592587e-06, "loss": 2.8176, "step": 3186 }, { "epoch": 3.7718259840189408, "grad_norm": 5.0, "learning_rate": 3.433793068700014e-06, "loss": 2.5333, "step": 3187 }, { "epoch": 3.7730097662030184, "grad_norm": 4.59375, "learning_rate": 3.427556624209173e-06, "loss": 2.359, "step": 3188 }, { "epoch": 3.774193548387097, "grad_norm": 5.09375, "learning_rate": 3.4213246766537776e-06, "loss": 2.5535, "step": 3189 }, { "epoch": 3.7753773305711746, "grad_norm": 4.65625, "learning_rate": 3.4150972302977914e-06, "loss": 2.3016, "step": 3190 }, { "epoch": 3.776561112755253, "grad_norm": 4.78125, "learning_rate": 3.4088742894020966e-06, "loss": 2.3911, "step": 3191 }, { "epoch": 3.7777448949393313, "grad_norm": 4.59375, "learning_rate": 3.4026558582244928e-06, "loss": 2.6216, "step": 3192 }, { "epoch": 3.7789286771234094, "grad_norm": 4.53125, "learning_rate": 3.3964419410197013e-06, "loss": 2.3724, "step": 3193 }, { "epoch": 3.7801124593074875, "grad_norm": 5.75, "learning_rate": 3.390232542039352e-06, "loss": 2.5684, "step": 3194 }, { "epoch": 3.7812962414915656, "grad_norm": 5.375, "learning_rate": 3.384027665531976e-06, "loss": 2.611, "step": 3195 }, { "epoch": 3.7824800236756437, "grad_norm": 4.65625, "learning_rate": 3.377827315743021e-06, "loss": 2.5498, "step": 3196 }, { "epoch": 3.783663805859722, "grad_norm": 4.5625, "learning_rate": 3.371631496914829e-06, "loss": 2.4785, "step": 3197 }, { "epoch": 3.7848475880438, "grad_norm": 5.1875, "learning_rate": 3.3654402132866458e-06, "loss": 2.3482, "step": 3198 }, { "epoch": 3.786031370227878, "grad_norm": 4.875, "learning_rate": 3.35925346909461e-06, "loss": 2.2719, "step": 3199 }, { "epoch": 3.787215152411956, "grad_norm": 4.71875, "learning_rate": 3.353071268571759e-06, "loss": 2.3063, "step": 3200 }, { "epoch": 3.7883989345960343, "grad_norm": 4.53125, "learning_rate": 3.346893615948026e-06, "loss": 2.2966, "step": 3201 }, { "epoch": 3.7895827167801124, "grad_norm": 4.21875, "learning_rate": 3.340720515450221e-06, "loss": 2.2279, "step": 3202 }, { "epoch": 3.7907664989641905, "grad_norm": 4.5, "learning_rate": 3.3345519713020445e-06, "loss": 2.5349, "step": 3203 }, { "epoch": 3.7919502811482686, "grad_norm": 4.71875, "learning_rate": 3.328387987724079e-06, "loss": 2.7566, "step": 3204 }, { "epoch": 3.7931340633323467, "grad_norm": 4.9375, "learning_rate": 3.3222285689337886e-06, "loss": 2.5685, "step": 3205 }, { "epoch": 3.794317845516425, "grad_norm": 4.53125, "learning_rate": 3.3160737191455173e-06, "loss": 2.444, "step": 3206 }, { "epoch": 3.795501627700503, "grad_norm": 4.75, "learning_rate": 3.309923442570472e-06, "loss": 2.4323, "step": 3207 }, { "epoch": 3.7966854098845815, "grad_norm": 5.46875, "learning_rate": 3.3037777434167418e-06, "loss": 2.4867, "step": 3208 }, { "epoch": 3.797869192068659, "grad_norm": 4.28125, "learning_rate": 3.297636625889279e-06, "loss": 2.2034, "step": 3209 }, { "epoch": 3.7990529742527377, "grad_norm": 4.46875, "learning_rate": 3.2915000941898958e-06, "loss": 2.3877, "step": 3210 }, { "epoch": 3.8002367564368154, "grad_norm": 4.15625, "learning_rate": 3.2853681525172796e-06, "loss": 2.1541, "step": 3211 }, { "epoch": 3.801420538620894, "grad_norm": 5.0, "learning_rate": 3.2792408050669634e-06, "loss": 2.6421, "step": 3212 }, { "epoch": 3.802604320804972, "grad_norm": 4.34375, "learning_rate": 3.273118056031349e-06, "loss": 2.1719, "step": 3213 }, { "epoch": 3.80378810298905, "grad_norm": 4.625, "learning_rate": 3.266999909599684e-06, "loss": 2.2887, "step": 3214 }, { "epoch": 3.8049718851731282, "grad_norm": 4.4375, "learning_rate": 3.2608863699580684e-06, "loss": 2.5385, "step": 3215 }, { "epoch": 3.8061556673572063, "grad_norm": 4.84375, "learning_rate": 3.2547774412894485e-06, "loss": 2.2414, "step": 3216 }, { "epoch": 3.8073394495412844, "grad_norm": 4.09375, "learning_rate": 3.2486731277736205e-06, "loss": 2.3534, "step": 3217 }, { "epoch": 3.8085232317253626, "grad_norm": 4.84375, "learning_rate": 3.242573433587224e-06, "loss": 2.2991, "step": 3218 }, { "epoch": 3.8097070139094407, "grad_norm": 4.96875, "learning_rate": 3.2364783629037277e-06, "loss": 2.732, "step": 3219 }, { "epoch": 3.8108907960935188, "grad_norm": 4.3125, "learning_rate": 3.2303879198934494e-06, "loss": 2.029, "step": 3220 }, { "epoch": 3.812074578277597, "grad_norm": 3.96875, "learning_rate": 3.2243021087235336e-06, "loss": 2.3385, "step": 3221 }, { "epoch": 3.813258360461675, "grad_norm": 4.25, "learning_rate": 3.2182209335579516e-06, "loss": 2.0225, "step": 3222 }, { "epoch": 3.814442142645753, "grad_norm": 5.09375, "learning_rate": 3.212144398557515e-06, "loss": 2.6652, "step": 3223 }, { "epoch": 3.815625924829831, "grad_norm": 4.8125, "learning_rate": 3.206072507879847e-06, "loss": 2.586, "step": 3224 }, { "epoch": 3.8168097070139093, "grad_norm": 4.6875, "learning_rate": 3.2000052656794066e-06, "loss": 2.5069, "step": 3225 }, { "epoch": 3.8179934891979874, "grad_norm": 4.6875, "learning_rate": 3.193942676107462e-06, "loss": 2.5185, "step": 3226 }, { "epoch": 3.819177271382066, "grad_norm": 4.625, "learning_rate": 3.1878847433121005e-06, "loss": 2.3581, "step": 3227 }, { "epoch": 3.8203610535661436, "grad_norm": 5.3125, "learning_rate": 3.1818314714382202e-06, "loss": 2.7109, "step": 3228 }, { "epoch": 3.821544835750222, "grad_norm": 5.3125, "learning_rate": 3.175782864627539e-06, "loss": 2.4238, "step": 3229 }, { "epoch": 3.8227286179343, "grad_norm": 4.75, "learning_rate": 3.16973892701858e-06, "loss": 2.8892, "step": 3230 }, { "epoch": 3.8239124001183784, "grad_norm": 5.75, "learning_rate": 3.1636996627466622e-06, "loss": 2.5289, "step": 3231 }, { "epoch": 3.825096182302456, "grad_norm": 4.375, "learning_rate": 3.157665075943922e-06, "loss": 2.2162, "step": 3232 }, { "epoch": 3.8262799644865346, "grad_norm": 4.6875, "learning_rate": 3.1516351707392843e-06, "loss": 2.4166, "step": 3233 }, { "epoch": 3.8274637466706127, "grad_norm": 4.15625, "learning_rate": 3.14560995125847e-06, "loss": 2.0285, "step": 3234 }, { "epoch": 3.828647528854691, "grad_norm": 4.375, "learning_rate": 3.1395894216240054e-06, "loss": 2.2721, "step": 3235 }, { "epoch": 3.829831311038769, "grad_norm": 4.59375, "learning_rate": 3.133573585955194e-06, "loss": 1.9762, "step": 3236 }, { "epoch": 3.831015093222847, "grad_norm": 4.3125, "learning_rate": 3.12756244836814e-06, "loss": 2.5094, "step": 3237 }, { "epoch": 3.832198875406925, "grad_norm": 4.4375, "learning_rate": 3.121556012975726e-06, "loss": 2.4413, "step": 3238 }, { "epoch": 3.8333826575910033, "grad_norm": 4.25, "learning_rate": 3.115554283887614e-06, "loss": 2.2776, "step": 3239 }, { "epoch": 3.8345664397750814, "grad_norm": 4.78125, "learning_rate": 3.109557265210259e-06, "loss": 2.0161, "step": 3240 }, { "epoch": 3.8357502219591595, "grad_norm": 4.34375, "learning_rate": 3.1035649610468753e-06, "loss": 2.293, "step": 3241 }, { "epoch": 3.8369340041432376, "grad_norm": 4.96875, "learning_rate": 3.0975773754974705e-06, "loss": 2.3178, "step": 3242 }, { "epoch": 3.8381177863273157, "grad_norm": 5.3125, "learning_rate": 3.0915945126588063e-06, "loss": 2.5618, "step": 3243 }, { "epoch": 3.839301568511394, "grad_norm": 4.9375, "learning_rate": 3.085616376624426e-06, "loss": 2.5077, "step": 3244 }, { "epoch": 3.840485350695472, "grad_norm": 4.28125, "learning_rate": 3.079642971484633e-06, "loss": 2.3493, "step": 3245 }, { "epoch": 3.84166913287955, "grad_norm": 4.96875, "learning_rate": 3.0736743013264903e-06, "loss": 2.6199, "step": 3246 }, { "epoch": 3.842852915063628, "grad_norm": 4.46875, "learning_rate": 3.0677103702338306e-06, "loss": 2.4093, "step": 3247 }, { "epoch": 3.8440366972477067, "grad_norm": 4.25, "learning_rate": 3.0617511822872337e-06, "loss": 2.5201, "step": 3248 }, { "epoch": 3.8452204794317844, "grad_norm": 5.15625, "learning_rate": 3.0557967415640456e-06, "loss": 2.3154, "step": 3249 }, { "epoch": 3.846404261615863, "grad_norm": 4.15625, "learning_rate": 3.0498470521383527e-06, "loss": 2.2999, "step": 3250 }, { "epoch": 3.8475880437999406, "grad_norm": 5.03125, "learning_rate": 3.0439021180809946e-06, "loss": 2.5099, "step": 3251 }, { "epoch": 3.848771825984019, "grad_norm": 5.96875, "learning_rate": 3.037961943459563e-06, "loss": 2.5034, "step": 3252 }, { "epoch": 3.8499556081680972, "grad_norm": 4.28125, "learning_rate": 3.0320265323383834e-06, "loss": 2.462, "step": 3253 }, { "epoch": 3.8511393903521753, "grad_norm": 4.1875, "learning_rate": 3.026095888778533e-06, "loss": 2.5404, "step": 3254 }, { "epoch": 3.8523231725362534, "grad_norm": 3.9375, "learning_rate": 3.0201700168378147e-06, "loss": 2.3641, "step": 3255 }, { "epoch": 3.8535069547203316, "grad_norm": 4.3125, "learning_rate": 3.014248920570778e-06, "loss": 2.4841, "step": 3256 }, { "epoch": 3.8546907369044097, "grad_norm": 5.40625, "learning_rate": 3.0083326040286977e-06, "loss": 2.4138, "step": 3257 }, { "epoch": 3.8558745190884878, "grad_norm": 4.5, "learning_rate": 3.0024210712595768e-06, "loss": 2.4017, "step": 3258 }, { "epoch": 3.857058301272566, "grad_norm": 4.09375, "learning_rate": 2.9965143263081554e-06, "loss": 2.2029, "step": 3259 }, { "epoch": 3.858242083456644, "grad_norm": 4.53125, "learning_rate": 2.9906123732158842e-06, "loss": 2.158, "step": 3260 }, { "epoch": 3.859425865640722, "grad_norm": 4.34375, "learning_rate": 2.9847152160209494e-06, "loss": 2.3191, "step": 3261 }, { "epoch": 3.8606096478248, "grad_norm": 4.9375, "learning_rate": 2.9788228587582447e-06, "loss": 2.4063, "step": 3262 }, { "epoch": 3.8617934300088783, "grad_norm": 5.1875, "learning_rate": 2.972935305459379e-06, "loss": 2.3238, "step": 3263 }, { "epoch": 3.8629772121929564, "grad_norm": 4.6875, "learning_rate": 2.967052560152687e-06, "loss": 2.3304, "step": 3264 }, { "epoch": 3.8641609943770345, "grad_norm": 5.03125, "learning_rate": 2.961174626863198e-06, "loss": 2.5136, "step": 3265 }, { "epoch": 3.8653447765611126, "grad_norm": 4.78125, "learning_rate": 2.9553015096126638e-06, "loss": 2.4552, "step": 3266 }, { "epoch": 3.8665285587451907, "grad_norm": 5.34375, "learning_rate": 2.9494332124195247e-06, "loss": 2.5955, "step": 3267 }, { "epoch": 3.867712340929269, "grad_norm": 4.6875, "learning_rate": 2.9435697392989405e-06, "loss": 2.5173, "step": 3268 }, { "epoch": 3.8688961231133474, "grad_norm": 4.65625, "learning_rate": 2.9377110942627573e-06, "loss": 2.4018, "step": 3269 }, { "epoch": 3.870079905297425, "grad_norm": 4.625, "learning_rate": 2.93185728131952e-06, "loss": 2.3725, "step": 3270 }, { "epoch": 3.8712636874815036, "grad_norm": 4.21875, "learning_rate": 2.926008304474475e-06, "loss": 2.1697, "step": 3271 }, { "epoch": 3.8724474696655813, "grad_norm": 4.75, "learning_rate": 2.920164167729548e-06, "loss": 2.4166, "step": 3272 }, { "epoch": 3.87363125184966, "grad_norm": 5.0625, "learning_rate": 2.914324875083366e-06, "loss": 2.2029, "step": 3273 }, { "epoch": 3.874815034033738, "grad_norm": 4.125, "learning_rate": 2.9084904305312324e-06, "loss": 2.3019, "step": 3274 }, { "epoch": 3.875998816217816, "grad_norm": 5.125, "learning_rate": 2.902660838065131e-06, "loss": 2.1578, "step": 3275 }, { "epoch": 3.877182598401894, "grad_norm": 4.375, "learning_rate": 2.896836101673738e-06, "loss": 1.9725, "step": 3276 }, { "epoch": 3.8783663805859723, "grad_norm": 4.3125, "learning_rate": 2.8910162253423947e-06, "loss": 2.2161, "step": 3277 }, { "epoch": 3.8795501627700504, "grad_norm": 4.5, "learning_rate": 2.885201213053126e-06, "loss": 2.4827, "step": 3278 }, { "epoch": 3.8807339449541285, "grad_norm": 5.0, "learning_rate": 2.8793910687846203e-06, "loss": 1.9791, "step": 3279 }, { "epoch": 3.8819177271382066, "grad_norm": 4.875, "learning_rate": 2.873585796512247e-06, "loss": 2.2997, "step": 3280 }, { "epoch": 3.8831015093222847, "grad_norm": 5.03125, "learning_rate": 2.8677854002080307e-06, "loss": 2.4671, "step": 3281 }, { "epoch": 3.884285291506363, "grad_norm": 4.6875, "learning_rate": 2.86198988384066e-06, "loss": 2.2963, "step": 3282 }, { "epoch": 3.885469073690441, "grad_norm": 4.34375, "learning_rate": 2.8561992513754967e-06, "loss": 2.3101, "step": 3283 }, { "epoch": 3.886652855874519, "grad_norm": 5.34375, "learning_rate": 2.8504135067745463e-06, "loss": 2.3675, "step": 3284 }, { "epoch": 3.887836638058597, "grad_norm": 5.0, "learning_rate": 2.8446326539964818e-06, "loss": 2.5904, "step": 3285 }, { "epoch": 3.8890204202426752, "grad_norm": 4.1875, "learning_rate": 2.838856696996621e-06, "loss": 2.4075, "step": 3286 }, { "epoch": 3.8902042024267534, "grad_norm": 4.53125, "learning_rate": 2.8330856397269336e-06, "loss": 2.4705, "step": 3287 }, { "epoch": 3.8913879846108315, "grad_norm": 4.6875, "learning_rate": 2.827319486136042e-06, "loss": 2.3184, "step": 3288 }, { "epoch": 3.8925717667949096, "grad_norm": 5.15625, "learning_rate": 2.8215582401692054e-06, "loss": 2.4858, "step": 3289 }, { "epoch": 3.893755548978988, "grad_norm": 4.78125, "learning_rate": 2.815801905768334e-06, "loss": 2.1901, "step": 3290 }, { "epoch": 3.894939331163066, "grad_norm": 5.0625, "learning_rate": 2.810050486871968e-06, "loss": 2.4507, "step": 3291 }, { "epoch": 3.8961231133471443, "grad_norm": 5.40625, "learning_rate": 2.8043039874152945e-06, "loss": 2.6147, "step": 3292 }, { "epoch": 3.897306895531222, "grad_norm": 4.5, "learning_rate": 2.798562411330126e-06, "loss": 2.2716, "step": 3293 }, { "epoch": 3.8984906777153006, "grad_norm": 5.125, "learning_rate": 2.792825762544907e-06, "loss": 2.5757, "step": 3294 }, { "epoch": 3.8996744598993787, "grad_norm": 4.71875, "learning_rate": 2.7870940449847194e-06, "loss": 2.3367, "step": 3295 }, { "epoch": 3.9008582420834568, "grad_norm": 4.625, "learning_rate": 2.781367262571261e-06, "loss": 2.3161, "step": 3296 }, { "epoch": 3.902042024267535, "grad_norm": 4.40625, "learning_rate": 2.7756454192228597e-06, "loss": 2.3002, "step": 3297 }, { "epoch": 3.903225806451613, "grad_norm": 4.71875, "learning_rate": 2.76992851885446e-06, "loss": 2.4987, "step": 3298 }, { "epoch": 3.904409588635691, "grad_norm": 4.59375, "learning_rate": 2.7642165653776242e-06, "loss": 2.3388, "step": 3299 }, { "epoch": 3.905593370819769, "grad_norm": 4.6875, "learning_rate": 2.7585095627005353e-06, "loss": 2.3585, "step": 3300 }, { "epoch": 3.9067771530038473, "grad_norm": 4.3125, "learning_rate": 2.75280751472798e-06, "loss": 2.5887, "step": 3301 }, { "epoch": 3.9079609351879254, "grad_norm": 4.6875, "learning_rate": 2.7471104253613645e-06, "loss": 2.306, "step": 3302 }, { "epoch": 3.9091447173720035, "grad_norm": 4.28125, "learning_rate": 2.7414182984986947e-06, "loss": 2.58, "step": 3303 }, { "epoch": 3.9103284995560816, "grad_norm": 4.5, "learning_rate": 2.7357311380345873e-06, "loss": 2.2202, "step": 3304 }, { "epoch": 3.9115122817401597, "grad_norm": 4.59375, "learning_rate": 2.7300489478602567e-06, "loss": 2.0338, "step": 3305 }, { "epoch": 3.912696063924238, "grad_norm": 4.96875, "learning_rate": 2.7243717318635143e-06, "loss": 2.4835, "step": 3306 }, { "epoch": 3.913879846108316, "grad_norm": 5.34375, "learning_rate": 2.718699493928776e-06, "loss": 2.608, "step": 3307 }, { "epoch": 3.915063628292394, "grad_norm": 4.75, "learning_rate": 2.7130322379370434e-06, "loss": 2.4327, "step": 3308 }, { "epoch": 3.916247410476472, "grad_norm": 4.84375, "learning_rate": 2.707369967765917e-06, "loss": 2.3825, "step": 3309 }, { "epoch": 3.9174311926605503, "grad_norm": 4.53125, "learning_rate": 2.7017126872895805e-06, "loss": 2.2145, "step": 3310 }, { "epoch": 3.918614974844629, "grad_norm": 4.3125, "learning_rate": 2.6960604003788014e-06, "loss": 2.4129, "step": 3311 }, { "epoch": 3.9197987570287065, "grad_norm": 4.875, "learning_rate": 2.690413110900941e-06, "loss": 2.2128, "step": 3312 }, { "epoch": 3.920982539212785, "grad_norm": 4.75, "learning_rate": 2.684770822719929e-06, "loss": 2.3734, "step": 3313 }, { "epoch": 3.9221663213968627, "grad_norm": 4.75, "learning_rate": 2.6791335396962794e-06, "loss": 2.4472, "step": 3314 }, { "epoch": 3.9233501035809413, "grad_norm": 4.75, "learning_rate": 2.6735012656870874e-06, "loss": 2.5779, "step": 3315 }, { "epoch": 3.9245338857650194, "grad_norm": 4.6875, "learning_rate": 2.6678740045460085e-06, "loss": 2.2683, "step": 3316 }, { "epoch": 3.9257176679490975, "grad_norm": 4.71875, "learning_rate": 2.6622517601232766e-06, "loss": 2.5038, "step": 3317 }, { "epoch": 3.9269014501331756, "grad_norm": 5.3125, "learning_rate": 2.6566345362656876e-06, "loss": 2.6, "step": 3318 }, { "epoch": 3.9280852323172537, "grad_norm": 4.6875, "learning_rate": 2.651022336816611e-06, "loss": 2.2728, "step": 3319 }, { "epoch": 3.929269014501332, "grad_norm": 4.53125, "learning_rate": 2.6454151656159666e-06, "loss": 2.3878, "step": 3320 }, { "epoch": 3.93045279668541, "grad_norm": 4.65625, "learning_rate": 2.6398130265002476e-06, "loss": 2.5279, "step": 3321 }, { "epoch": 3.931636578869488, "grad_norm": 4.59375, "learning_rate": 2.6342159233024945e-06, "loss": 2.4412, "step": 3322 }, { "epoch": 3.932820361053566, "grad_norm": 5.21875, "learning_rate": 2.6286238598522993e-06, "loss": 2.7072, "step": 3323 }, { "epoch": 3.9340041432376442, "grad_norm": 4.5, "learning_rate": 2.623036839975819e-06, "loss": 2.4326, "step": 3324 }, { "epoch": 3.9351879254217224, "grad_norm": 4.59375, "learning_rate": 2.6174548674957457e-06, "loss": 2.3908, "step": 3325 }, { "epoch": 3.9363717076058005, "grad_norm": 5.0, "learning_rate": 2.611877946231327e-06, "loss": 2.6698, "step": 3326 }, { "epoch": 3.9375554897898786, "grad_norm": 4.75, "learning_rate": 2.606306079998354e-06, "loss": 2.2741, "step": 3327 }, { "epoch": 3.9387392719739567, "grad_norm": 5.3125, "learning_rate": 2.600739272609154e-06, "loss": 2.4741, "step": 3328 }, { "epoch": 3.939923054158035, "grad_norm": 4.625, "learning_rate": 2.5951775278725956e-06, "loss": 2.4583, "step": 3329 }, { "epoch": 3.941106836342113, "grad_norm": 4.90625, "learning_rate": 2.58962084959408e-06, "loss": 2.5826, "step": 3330 }, { "epoch": 3.942290618526191, "grad_norm": 4.90625, "learning_rate": 2.584069241575553e-06, "loss": 2.3529, "step": 3331 }, { "epoch": 3.9434744007102696, "grad_norm": 4.25, "learning_rate": 2.578522707615476e-06, "loss": 2.3948, "step": 3332 }, { "epoch": 3.944658182894347, "grad_norm": 4.46875, "learning_rate": 2.572981251508854e-06, "loss": 2.6316, "step": 3333 }, { "epoch": 3.9458419650784258, "grad_norm": 4.78125, "learning_rate": 2.567444877047205e-06, "loss": 2.3834, "step": 3334 }, { "epoch": 3.9470257472625034, "grad_norm": 4.875, "learning_rate": 2.561913588018573e-06, "loss": 2.7805, "step": 3335 }, { "epoch": 3.948209529446582, "grad_norm": 4.46875, "learning_rate": 2.556387388207531e-06, "loss": 2.3945, "step": 3336 }, { "epoch": 3.94939331163066, "grad_norm": 4.8125, "learning_rate": 2.550866281395157e-06, "loss": 2.394, "step": 3337 }, { "epoch": 3.950577093814738, "grad_norm": 4.4375, "learning_rate": 2.545350271359055e-06, "loss": 2.5195, "step": 3338 }, { "epoch": 3.9517608759988163, "grad_norm": 4.40625, "learning_rate": 2.5398393618733386e-06, "loss": 2.2572, "step": 3339 }, { "epoch": 3.9529446581828944, "grad_norm": 4.6875, "learning_rate": 2.534333556708628e-06, "loss": 2.5222, "step": 3340 }, { "epoch": 3.9541284403669725, "grad_norm": 4.59375, "learning_rate": 2.5288328596320557e-06, "loss": 2.4527, "step": 3341 }, { "epoch": 3.9553122225510506, "grad_norm": 4.625, "learning_rate": 2.5233372744072504e-06, "loss": 2.4783, "step": 3342 }, { "epoch": 3.9564960047351287, "grad_norm": 4.84375, "learning_rate": 2.517846804794356e-06, "loss": 2.6602, "step": 3343 }, { "epoch": 3.957679786919207, "grad_norm": 4.5625, "learning_rate": 2.5123614545500107e-06, "loss": 2.1903, "step": 3344 }, { "epoch": 3.958863569103285, "grad_norm": 4.625, "learning_rate": 2.506881227427348e-06, "loss": 2.1514, "step": 3345 }, { "epoch": 3.960047351287363, "grad_norm": 4.3125, "learning_rate": 2.501406127175996e-06, "loss": 2.2576, "step": 3346 }, { "epoch": 3.961231133471441, "grad_norm": 4.625, "learning_rate": 2.495936157542074e-06, "loss": 2.2597, "step": 3347 }, { "epoch": 3.9624149156555193, "grad_norm": 4.46875, "learning_rate": 2.4904713222682e-06, "loss": 2.4155, "step": 3348 }, { "epoch": 3.9635986978395974, "grad_norm": 5.25, "learning_rate": 2.4850116250934642e-06, "loss": 2.5355, "step": 3349 }, { "epoch": 3.9647824800236755, "grad_norm": 4.75, "learning_rate": 2.4795570697534544e-06, "loss": 2.0899, "step": 3350 }, { "epoch": 3.9659662622077536, "grad_norm": 4.96875, "learning_rate": 2.474107659980236e-06, "loss": 2.1866, "step": 3351 }, { "epoch": 3.9671500443918317, "grad_norm": 4.875, "learning_rate": 2.4686633995023525e-06, "loss": 2.3906, "step": 3352 }, { "epoch": 3.9683338265759103, "grad_norm": 4.46875, "learning_rate": 2.4632242920448233e-06, "loss": 2.2085, "step": 3353 }, { "epoch": 3.969517608759988, "grad_norm": 4.5, "learning_rate": 2.4577903413291405e-06, "loss": 2.6053, "step": 3354 }, { "epoch": 3.9707013909440665, "grad_norm": 4.65625, "learning_rate": 2.4523615510732734e-06, "loss": 2.609, "step": 3355 }, { "epoch": 3.971885173128144, "grad_norm": 4.40625, "learning_rate": 2.4469379249916614e-06, "loss": 2.4562, "step": 3356 }, { "epoch": 3.9730689553122227, "grad_norm": 4.59375, "learning_rate": 2.4415194667952038e-06, "loss": 2.5545, "step": 3357 }, { "epoch": 3.974252737496301, "grad_norm": 4.9375, "learning_rate": 2.4361061801912666e-06, "loss": 2.237, "step": 3358 }, { "epoch": 3.975436519680379, "grad_norm": 4.78125, "learning_rate": 2.430698068883677e-06, "loss": 2.5205, "step": 3359 }, { "epoch": 3.976620301864457, "grad_norm": 4.34375, "learning_rate": 2.4252951365727218e-06, "loss": 2.2751, "step": 3360 }, { "epoch": 3.977804084048535, "grad_norm": 4.84375, "learning_rate": 2.41989738695515e-06, "loss": 2.4446, "step": 3361 }, { "epoch": 3.9789878662326132, "grad_norm": 4.71875, "learning_rate": 2.414504823724153e-06, "loss": 2.7056, "step": 3362 }, { "epoch": 3.9801716484166914, "grad_norm": 4.625, "learning_rate": 2.409117450569385e-06, "loss": 2.4967, "step": 3363 }, { "epoch": 3.9813554306007695, "grad_norm": 4.65625, "learning_rate": 2.40373527117694e-06, "loss": 2.4286, "step": 3364 }, { "epoch": 3.9825392127848476, "grad_norm": 4.78125, "learning_rate": 2.3983582892293642e-06, "loss": 2.2915, "step": 3365 }, { "epoch": 3.9837229949689257, "grad_norm": 4.5625, "learning_rate": 2.3929865084056415e-06, "loss": 2.3736, "step": 3366 }, { "epoch": 3.984906777153004, "grad_norm": 4.5, "learning_rate": 2.3876199323812043e-06, "loss": 2.2278, "step": 3367 }, { "epoch": 3.986090559337082, "grad_norm": 4.71875, "learning_rate": 2.382258564827924e-06, "loss": 2.255, "step": 3368 }, { "epoch": 3.98727434152116, "grad_norm": 4.5625, "learning_rate": 2.3769024094141015e-06, "loss": 2.1986, "step": 3369 }, { "epoch": 3.988458123705238, "grad_norm": 4.46875, "learning_rate": 2.3715514698044762e-06, "loss": 2.2535, "step": 3370 }, { "epoch": 3.989641905889316, "grad_norm": 4.9375, "learning_rate": 2.3662057496602142e-06, "loss": 2.2835, "step": 3371 }, { "epoch": 3.9908256880733948, "grad_norm": 4.5625, "learning_rate": 2.3608652526389177e-06, "loss": 1.8209, "step": 3372 }, { "epoch": 3.9920094702574724, "grad_norm": 4.9375, "learning_rate": 2.355529982394613e-06, "loss": 2.2474, "step": 3373 }, { "epoch": 3.993193252441551, "grad_norm": 4.8125, "learning_rate": 2.3501999425777433e-06, "loss": 2.6601, "step": 3374 }, { "epoch": 3.9943770346256287, "grad_norm": 4.875, "learning_rate": 2.344875136835184e-06, "loss": 2.7011, "step": 3375 }, { "epoch": 3.995560816809707, "grad_norm": 4.3125, "learning_rate": 2.339555568810221e-06, "loss": 2.1414, "step": 3376 }, { "epoch": 3.996744598993785, "grad_norm": 4.21875, "learning_rate": 2.3342412421425585e-06, "loss": 2.184, "step": 3377 }, { "epoch": 3.9979283811778634, "grad_norm": 4.8125, "learning_rate": 2.3289321604683135e-06, "loss": 2.5714, "step": 3378 }, { "epoch": 3.9991121633619415, "grad_norm": 4.21875, "learning_rate": 2.3236283274200187e-06, "loss": 2.3567, "step": 3379 }, { "epoch": 4.0, "grad_norm": 4.25, "learning_rate": 2.3183297466266142e-06, "loss": 1.7179, "step": 3380 }, { "epoch": 4.0011837821840786, "grad_norm": 4.5625, "learning_rate": 2.3130364217134436e-06, "loss": 2.6223, "step": 3381 }, { "epoch": 4.002367564368156, "grad_norm": 4.71875, "learning_rate": 2.307748356302256e-06, "loss": 2.9615, "step": 3382 }, { "epoch": 4.003551346552235, "grad_norm": 4.90625, "learning_rate": 2.3024655540111984e-06, "loss": 2.6711, "step": 3383 }, { "epoch": 4.004735128736312, "grad_norm": 4.96875, "learning_rate": 2.2971880184548234e-06, "loss": 2.4145, "step": 3384 }, { "epoch": 4.005918910920391, "grad_norm": 4.625, "learning_rate": 2.2919157532440796e-06, "loss": 2.682, "step": 3385 }, { "epoch": 4.007102693104469, "grad_norm": 4.3125, "learning_rate": 2.2866487619863e-06, "loss": 2.1903, "step": 3386 }, { "epoch": 4.008286475288547, "grad_norm": 5.375, "learning_rate": 2.281387048285223e-06, "loss": 2.4602, "step": 3387 }, { "epoch": 4.009470257472625, "grad_norm": 4.8125, "learning_rate": 2.2761306157409657e-06, "loss": 2.2213, "step": 3388 }, { "epoch": 4.010654039656703, "grad_norm": 4.28125, "learning_rate": 2.2708794679500325e-06, "loss": 2.5336, "step": 3389 }, { "epoch": 4.011837821840781, "grad_norm": 4.15625, "learning_rate": 2.2656336085053186e-06, "loss": 2.0241, "step": 3390 }, { "epoch": 4.01302160402486, "grad_norm": 4.125, "learning_rate": 2.2603930409960927e-06, "loss": 2.1632, "step": 3391 }, { "epoch": 4.014205386208937, "grad_norm": 4.65625, "learning_rate": 2.255157769008011e-06, "loss": 2.2096, "step": 3392 }, { "epoch": 4.015389168393016, "grad_norm": 4.6875, "learning_rate": 2.249927796123099e-06, "loss": 2.5702, "step": 3393 }, { "epoch": 4.0165729505770935, "grad_norm": 4.34375, "learning_rate": 2.2447031259197614e-06, "loss": 2.3624, "step": 3394 }, { "epoch": 4.017756732761172, "grad_norm": 4.40625, "learning_rate": 2.2394837619727692e-06, "loss": 2.4078, "step": 3395 }, { "epoch": 4.01894051494525, "grad_norm": 4.75, "learning_rate": 2.2342697078532695e-06, "loss": 2.6305, "step": 3396 }, { "epoch": 4.020124297129328, "grad_norm": 4.46875, "learning_rate": 2.2290609671287756e-06, "loss": 2.4112, "step": 3397 }, { "epoch": 4.021308079313406, "grad_norm": 4.84375, "learning_rate": 2.223857543363158e-06, "loss": 2.2295, "step": 3398 }, { "epoch": 4.0224918614974845, "grad_norm": 4.90625, "learning_rate": 2.2186594401166604e-06, "loss": 2.575, "step": 3399 }, { "epoch": 4.023675643681562, "grad_norm": 4.8125, "learning_rate": 2.2134666609458764e-06, "loss": 2.4993, "step": 3400 }, { "epoch": 4.024859425865641, "grad_norm": 5.0, "learning_rate": 2.2082792094037585e-06, "loss": 2.1731, "step": 3401 }, { "epoch": 4.026043208049719, "grad_norm": 4.78125, "learning_rate": 2.203097089039621e-06, "loss": 2.4464, "step": 3402 }, { "epoch": 4.027226990233797, "grad_norm": 4.71875, "learning_rate": 2.1979203033991205e-06, "loss": 2.6542, "step": 3403 }, { "epoch": 4.0284107724178755, "grad_norm": 4.5, "learning_rate": 2.192748856024275e-06, "loss": 2.6391, "step": 3404 }, { "epoch": 4.029594554601953, "grad_norm": 4.84375, "learning_rate": 2.1875827504534387e-06, "loss": 2.5975, "step": 3405 }, { "epoch": 4.030778336786032, "grad_norm": 5.03125, "learning_rate": 2.1824219902213183e-06, "loss": 2.4955, "step": 3406 }, { "epoch": 4.031962118970109, "grad_norm": 4.96875, "learning_rate": 2.1772665788589566e-06, "loss": 2.6245, "step": 3407 }, { "epoch": 4.033145901154188, "grad_norm": 4.3125, "learning_rate": 2.172116519893742e-06, "loss": 2.0977, "step": 3408 }, { "epoch": 4.034329683338266, "grad_norm": 4.75, "learning_rate": 2.166971816849406e-06, "loss": 2.2449, "step": 3409 }, { "epoch": 4.035513465522344, "grad_norm": 4.59375, "learning_rate": 2.1618324732459993e-06, "loss": 2.427, "step": 3410 }, { "epoch": 4.036697247706422, "grad_norm": 4.9375, "learning_rate": 2.1566984925999234e-06, "loss": 2.59, "step": 3411 }, { "epoch": 4.0378810298905, "grad_norm": 4.65625, "learning_rate": 2.151569878423899e-06, "loss": 2.4347, "step": 3412 }, { "epoch": 4.039064812074578, "grad_norm": 4.28125, "learning_rate": 2.1464466342269763e-06, "loss": 2.1405, "step": 3413 }, { "epoch": 4.040248594258657, "grad_norm": 4.4375, "learning_rate": 2.1413287635145365e-06, "loss": 2.3374, "step": 3414 }, { "epoch": 4.041432376442734, "grad_norm": 4.34375, "learning_rate": 2.1362162697882773e-06, "loss": 2.3332, "step": 3415 }, { "epoch": 4.042616158626813, "grad_norm": 4.65625, "learning_rate": 2.131109156546226e-06, "loss": 2.2647, "step": 3416 }, { "epoch": 4.0437999408108904, "grad_norm": 5.0, "learning_rate": 2.1260074272827227e-06, "loss": 2.3169, "step": 3417 }, { "epoch": 4.044983722994969, "grad_norm": 4.53125, "learning_rate": 2.1209110854884184e-06, "loss": 2.1336, "step": 3418 }, { "epoch": 4.046167505179047, "grad_norm": 4.9375, "learning_rate": 2.1158201346502927e-06, "loss": 2.3096, "step": 3419 }, { "epoch": 4.047351287363125, "grad_norm": 4.8125, "learning_rate": 2.1107345782516207e-06, "loss": 2.3208, "step": 3420 }, { "epoch": 4.048535069547203, "grad_norm": 4.59375, "learning_rate": 2.1056544197720006e-06, "loss": 2.6449, "step": 3421 }, { "epoch": 4.049718851731281, "grad_norm": 4.78125, "learning_rate": 2.1005796626873253e-06, "loss": 2.4548, "step": 3422 }, { "epoch": 4.05090263391536, "grad_norm": 4.3125, "learning_rate": 2.0955103104698026e-06, "loss": 2.3009, "step": 3423 }, { "epoch": 4.052086416099438, "grad_norm": 4.65625, "learning_rate": 2.090446366587934e-06, "loss": 2.5852, "step": 3424 }, { "epoch": 4.053270198283516, "grad_norm": 4.625, "learning_rate": 2.0853878345065203e-06, "loss": 2.4242, "step": 3425 }, { "epoch": 4.054453980467594, "grad_norm": 4.84375, "learning_rate": 2.080334717686671e-06, "loss": 2.2476, "step": 3426 }, { "epoch": 4.055637762651672, "grad_norm": 4.46875, "learning_rate": 2.075287019585773e-06, "loss": 2.3717, "step": 3427 }, { "epoch": 4.05682154483575, "grad_norm": 4.625, "learning_rate": 2.0702447436575223e-06, "loss": 2.461, "step": 3428 }, { "epoch": 4.058005327019829, "grad_norm": 4.625, "learning_rate": 2.065207893351895e-06, "loss": 2.4927, "step": 3429 }, { "epoch": 4.059189109203906, "grad_norm": 4.75, "learning_rate": 2.060176472115155e-06, "loss": 2.5923, "step": 3430 }, { "epoch": 4.060372891387985, "grad_norm": 4.46875, "learning_rate": 2.0551504833898582e-06, "loss": 2.232, "step": 3431 }, { "epoch": 4.0615566735720625, "grad_norm": 5.34375, "learning_rate": 2.050129930614835e-06, "loss": 2.8469, "step": 3432 }, { "epoch": 4.062740455756141, "grad_norm": 5.0, "learning_rate": 2.0451148172252057e-06, "loss": 2.5307, "step": 3433 }, { "epoch": 4.063924237940219, "grad_norm": 4.875, "learning_rate": 2.040105146652358e-06, "loss": 2.5913, "step": 3434 }, { "epoch": 4.065108020124297, "grad_norm": 4.28125, "learning_rate": 2.0351009223239693e-06, "loss": 2.207, "step": 3435 }, { "epoch": 4.066291802308375, "grad_norm": 4.625, "learning_rate": 2.0301021476639782e-06, "loss": 2.4291, "step": 3436 }, { "epoch": 4.0674755844924535, "grad_norm": 4.625, "learning_rate": 2.0251088260925967e-06, "loss": 2.4846, "step": 3437 }, { "epoch": 4.068659366676531, "grad_norm": 4.3125, "learning_rate": 2.0201209610263153e-06, "loss": 2.2322, "step": 3438 }, { "epoch": 4.06984314886061, "grad_norm": 4.5625, "learning_rate": 2.015138555877877e-06, "loss": 2.528, "step": 3439 }, { "epoch": 4.071026931044687, "grad_norm": 4.96875, "learning_rate": 2.010161614056302e-06, "loss": 2.0412, "step": 3440 }, { "epoch": 4.072210713228766, "grad_norm": 5.09375, "learning_rate": 2.0051901389668647e-06, "loss": 2.7417, "step": 3441 }, { "epoch": 4.073394495412844, "grad_norm": 4.34375, "learning_rate": 2.0002241340110983e-06, "loss": 2.0943, "step": 3442 }, { "epoch": 4.074578277596922, "grad_norm": 4.96875, "learning_rate": 1.9952636025868012e-06, "loss": 2.3619, "step": 3443 }, { "epoch": 4.075762059781001, "grad_norm": 5.125, "learning_rate": 1.990308548088017e-06, "loss": 2.5491, "step": 3444 }, { "epoch": 4.076945841965078, "grad_norm": 5.0, "learning_rate": 1.9853589739050494e-06, "loss": 2.7875, "step": 3445 }, { "epoch": 4.078129624149157, "grad_norm": 4.125, "learning_rate": 1.9804148834244465e-06, "loss": 2.1228, "step": 3446 }, { "epoch": 4.079313406333235, "grad_norm": 4.125, "learning_rate": 1.975476280029013e-06, "loss": 1.9528, "step": 3447 }, { "epoch": 4.080497188517313, "grad_norm": 4.21875, "learning_rate": 1.970543167097789e-06, "loss": 2.371, "step": 3448 }, { "epoch": 4.081680970701391, "grad_norm": 4.5, "learning_rate": 1.9656155480060613e-06, "loss": 2.271, "step": 3449 }, { "epoch": 4.082864752885469, "grad_norm": 4.28125, "learning_rate": 1.960693426125364e-06, "loss": 2.451, "step": 3450 }, { "epoch": 4.084048535069547, "grad_norm": 4.375, "learning_rate": 1.9557768048234606e-06, "loss": 2.3741, "step": 3451 }, { "epoch": 4.085232317253626, "grad_norm": 4.28125, "learning_rate": 1.9508656874643604e-06, "loss": 2.3331, "step": 3452 }, { "epoch": 4.086416099437703, "grad_norm": 4.96875, "learning_rate": 1.9459600774082987e-06, "loss": 2.5657, "step": 3453 }, { "epoch": 4.087599881621782, "grad_norm": 4.875, "learning_rate": 1.9410599780117443e-06, "loss": 2.5194, "step": 3454 }, { "epoch": 4.0887836638058594, "grad_norm": 4.8125, "learning_rate": 1.9361653926274016e-06, "loss": 2.2442, "step": 3455 }, { "epoch": 4.089967445989938, "grad_norm": 4.75, "learning_rate": 1.9312763246041934e-06, "loss": 2.2031, "step": 3456 }, { "epoch": 4.091151228174016, "grad_norm": 4.625, "learning_rate": 1.9263927772872757e-06, "loss": 2.5169, "step": 3457 }, { "epoch": 4.092335010358094, "grad_norm": 4.78125, "learning_rate": 1.9215147540180203e-06, "loss": 2.5472, "step": 3458 }, { "epoch": 4.093518792542172, "grad_norm": 4.78125, "learning_rate": 1.916642258134026e-06, "loss": 2.4903, "step": 3459 }, { "epoch": 4.09470257472625, "grad_norm": 5.15625, "learning_rate": 1.9117752929691037e-06, "loss": 2.5776, "step": 3460 }, { "epoch": 4.095886356910328, "grad_norm": 4.53125, "learning_rate": 1.906913861853279e-06, "loss": 2.5561, "step": 3461 }, { "epoch": 4.097070139094407, "grad_norm": 4.96875, "learning_rate": 1.9020579681128027e-06, "loss": 2.0345, "step": 3462 }, { "epoch": 4.098253921278484, "grad_norm": 4.5625, "learning_rate": 1.8972076150701212e-06, "loss": 2.3207, "step": 3463 }, { "epoch": 4.099437703462563, "grad_norm": 4.59375, "learning_rate": 1.8923628060439037e-06, "loss": 2.6066, "step": 3464 }, { "epoch": 4.100621485646641, "grad_norm": 5.1875, "learning_rate": 1.887523544349016e-06, "loss": 2.4123, "step": 3465 }, { "epoch": 4.101805267830719, "grad_norm": 4.625, "learning_rate": 1.8826898332965315e-06, "loss": 2.1441, "step": 3466 }, { "epoch": 4.102989050014798, "grad_norm": 4.53125, "learning_rate": 1.8778616761937308e-06, "loss": 2.2158, "step": 3467 }, { "epoch": 4.104172832198875, "grad_norm": 4.375, "learning_rate": 1.8730390763440853e-06, "loss": 2.5415, "step": 3468 }, { "epoch": 4.105356614382954, "grad_norm": 4.1875, "learning_rate": 1.8682220370472737e-06, "loss": 2.4004, "step": 3469 }, { "epoch": 4.1065403965670315, "grad_norm": 5.34375, "learning_rate": 1.8634105615991592e-06, "loss": 2.7506, "step": 3470 }, { "epoch": 4.10772417875111, "grad_norm": 5.125, "learning_rate": 1.8586046532918123e-06, "loss": 2.6869, "step": 3471 }, { "epoch": 4.108907960935188, "grad_norm": 4.1875, "learning_rate": 1.853804315413481e-06, "loss": 2.4065, "step": 3472 }, { "epoch": 4.110091743119266, "grad_norm": 4.3125, "learning_rate": 1.8490095512486072e-06, "loss": 2.3277, "step": 3473 }, { "epoch": 4.111275525303344, "grad_norm": 4.5625, "learning_rate": 1.8442203640778222e-06, "loss": 2.4248, "step": 3474 }, { "epoch": 4.1124593074874225, "grad_norm": 5.125, "learning_rate": 1.839436757177936e-06, "loss": 2.5416, "step": 3475 }, { "epoch": 4.1136430896715, "grad_norm": 4.53125, "learning_rate": 1.8346587338219456e-06, "loss": 2.4022, "step": 3476 }, { "epoch": 4.114826871855579, "grad_norm": 4.5625, "learning_rate": 1.8298862972790243e-06, "loss": 2.5828, "step": 3477 }, { "epoch": 4.116010654039656, "grad_norm": 4.65625, "learning_rate": 1.825119450814522e-06, "loss": 2.6597, "step": 3478 }, { "epoch": 4.117194436223735, "grad_norm": 4.46875, "learning_rate": 1.8203581976899687e-06, "loss": 2.3653, "step": 3479 }, { "epoch": 4.118378218407813, "grad_norm": 5.5, "learning_rate": 1.8156025411630596e-06, "loss": 2.793, "step": 3480 }, { "epoch": 4.119562000591891, "grad_norm": 4.75, "learning_rate": 1.810852484487672e-06, "loss": 2.2804, "step": 3481 }, { "epoch": 4.120745782775969, "grad_norm": 4.5625, "learning_rate": 1.8061080309138379e-06, "loss": 2.3428, "step": 3482 }, { "epoch": 4.121929564960047, "grad_norm": 4.75, "learning_rate": 1.80136918368777e-06, "loss": 2.7398, "step": 3483 }, { "epoch": 4.123113347144126, "grad_norm": 4.625, "learning_rate": 1.7966359460518323e-06, "loss": 2.3599, "step": 3484 }, { "epoch": 4.124297129328204, "grad_norm": 4.53125, "learning_rate": 1.7919083212445566e-06, "loss": 2.259, "step": 3485 }, { "epoch": 4.125480911512282, "grad_norm": 4.875, "learning_rate": 1.7871863125006383e-06, "loss": 2.1664, "step": 3486 }, { "epoch": 4.12666469369636, "grad_norm": 5.0, "learning_rate": 1.782469923050919e-06, "loss": 2.269, "step": 3487 }, { "epoch": 4.127848475880438, "grad_norm": 4.5625, "learning_rate": 1.7777591561224094e-06, "loss": 2.3823, "step": 3488 }, { "epoch": 4.129032258064516, "grad_norm": 4.84375, "learning_rate": 1.7730540149382625e-06, "loss": 2.4885, "step": 3489 }, { "epoch": 4.130216040248595, "grad_norm": 4.21875, "learning_rate": 1.768354502717784e-06, "loss": 2.4543, "step": 3490 }, { "epoch": 4.131399822432672, "grad_norm": 4.4375, "learning_rate": 1.7636606226764353e-06, "loss": 2.1715, "step": 3491 }, { "epoch": 4.132583604616751, "grad_norm": 4.40625, "learning_rate": 1.7589723780258127e-06, "loss": 2.1567, "step": 3492 }, { "epoch": 4.1337673868008284, "grad_norm": 4.46875, "learning_rate": 1.7542897719736662e-06, "loss": 2.6403, "step": 3493 }, { "epoch": 4.134951168984907, "grad_norm": 4.46875, "learning_rate": 1.7496128077238872e-06, "loss": 2.5365, "step": 3494 }, { "epoch": 4.136134951168985, "grad_norm": 4.65625, "learning_rate": 1.7449414884765005e-06, "loss": 2.5625, "step": 3495 }, { "epoch": 4.137318733353063, "grad_norm": 4.5625, "learning_rate": 1.7402758174276734e-06, "loss": 2.6356, "step": 3496 }, { "epoch": 4.138502515537141, "grad_norm": 5.125, "learning_rate": 1.7356157977697052e-06, "loss": 2.5931, "step": 3497 }, { "epoch": 4.139686297721219, "grad_norm": 4.40625, "learning_rate": 1.730961432691034e-06, "loss": 2.2707, "step": 3498 }, { "epoch": 4.140870079905297, "grad_norm": 4.6875, "learning_rate": 1.7263127253762234e-06, "loss": 2.2192, "step": 3499 }, { "epoch": 4.142053862089376, "grad_norm": 4.28125, "learning_rate": 1.7216696790059718e-06, "loss": 2.4697, "step": 3500 }, { "epoch": 4.143237644273453, "grad_norm": 4.59375, "learning_rate": 1.7170322967570974e-06, "loss": 1.9952, "step": 3501 }, { "epoch": 4.144421426457532, "grad_norm": 4.96875, "learning_rate": 1.7124005818025447e-06, "loss": 2.9509, "step": 3502 }, { "epoch": 4.1456052086416095, "grad_norm": 4.21875, "learning_rate": 1.7077745373113874e-06, "loss": 2.2231, "step": 3503 }, { "epoch": 4.146788990825688, "grad_norm": 5.15625, "learning_rate": 1.7031541664488093e-06, "loss": 2.6038, "step": 3504 }, { "epoch": 4.147972773009766, "grad_norm": 4.8125, "learning_rate": 1.6985394723761194e-06, "loss": 2.5208, "step": 3505 }, { "epoch": 4.149156555193844, "grad_norm": 4.75, "learning_rate": 1.693930458250742e-06, "loss": 2.2641, "step": 3506 }, { "epoch": 4.150340337377923, "grad_norm": 4.78125, "learning_rate": 1.6893271272262123e-06, "loss": 2.5372, "step": 3507 }, { "epoch": 4.1515241195620005, "grad_norm": 4.65625, "learning_rate": 1.6847294824521777e-06, "loss": 2.3975, "step": 3508 }, { "epoch": 4.152707901746079, "grad_norm": 4.5, "learning_rate": 1.6801375270743925e-06, "loss": 2.2023, "step": 3509 }, { "epoch": 4.153891683930157, "grad_norm": 4.65625, "learning_rate": 1.675551264234726e-06, "loss": 2.4854, "step": 3510 }, { "epoch": 4.155075466114235, "grad_norm": 4.1875, "learning_rate": 1.6709706970711447e-06, "loss": 2.2641, "step": 3511 }, { "epoch": 4.156259248298313, "grad_norm": 4.6875, "learning_rate": 1.666395828717724e-06, "loss": 2.2288, "step": 3512 }, { "epoch": 4.1574430304823915, "grad_norm": 5.03125, "learning_rate": 1.6618266623046363e-06, "loss": 2.3456, "step": 3513 }, { "epoch": 4.158626812666469, "grad_norm": 4.53125, "learning_rate": 1.6572632009581502e-06, "loss": 2.5578, "step": 3514 }, { "epoch": 4.159810594850548, "grad_norm": 4.8125, "learning_rate": 1.65270544780064e-06, "loss": 2.5841, "step": 3515 }, { "epoch": 4.160994377034625, "grad_norm": 4.84375, "learning_rate": 1.6481534059505643e-06, "loss": 2.3424, "step": 3516 }, { "epoch": 4.162178159218704, "grad_norm": 4.34375, "learning_rate": 1.64360707852248e-06, "loss": 2.458, "step": 3517 }, { "epoch": 4.163361941402782, "grad_norm": 5.0625, "learning_rate": 1.6390664686270342e-06, "loss": 2.3774, "step": 3518 }, { "epoch": 4.16454572358686, "grad_norm": 4.5625, "learning_rate": 1.6345315793709592e-06, "loss": 2.3538, "step": 3519 }, { "epoch": 4.165729505770938, "grad_norm": 4.46875, "learning_rate": 1.6300024138570748e-06, "loss": 2.3486, "step": 3520 }, { "epoch": 4.166913287955016, "grad_norm": 4.96875, "learning_rate": 1.6254789751842803e-06, "loss": 2.4667, "step": 3521 }, { "epoch": 4.168097070139094, "grad_norm": 4.84375, "learning_rate": 1.6209612664475639e-06, "loss": 2.3464, "step": 3522 }, { "epoch": 4.169280852323173, "grad_norm": 5.03125, "learning_rate": 1.6164492907379903e-06, "loss": 2.4641, "step": 3523 }, { "epoch": 4.17046463450725, "grad_norm": 4.5, "learning_rate": 1.6119430511427015e-06, "loss": 2.5351, "step": 3524 }, { "epoch": 4.171648416691329, "grad_norm": 4.9375, "learning_rate": 1.6074425507449122e-06, "loss": 2.8779, "step": 3525 }, { "epoch": 4.172832198875407, "grad_norm": 4.75, "learning_rate": 1.6029477926239102e-06, "loss": 2.3484, "step": 3526 }, { "epoch": 4.174015981059485, "grad_norm": 4.59375, "learning_rate": 1.5984587798550633e-06, "loss": 2.3748, "step": 3527 }, { "epoch": 4.175199763243564, "grad_norm": 4.6875, "learning_rate": 1.5939755155097947e-06, "loss": 2.6494, "step": 3528 }, { "epoch": 4.176383545427641, "grad_norm": 5.0, "learning_rate": 1.5894980026556062e-06, "loss": 2.7447, "step": 3529 }, { "epoch": 4.17756732761172, "grad_norm": 4.71875, "learning_rate": 1.5850262443560594e-06, "loss": 2.5287, "step": 3530 }, { "epoch": 4.1787511097957974, "grad_norm": 4.78125, "learning_rate": 1.580560243670778e-06, "loss": 2.2543, "step": 3531 }, { "epoch": 4.179934891979876, "grad_norm": 4.4375, "learning_rate": 1.5761000036554453e-06, "loss": 2.3911, "step": 3532 }, { "epoch": 4.181118674163954, "grad_norm": 4.875, "learning_rate": 1.5716455273618048e-06, "loss": 2.5282, "step": 3533 }, { "epoch": 4.182302456348032, "grad_norm": 4.5625, "learning_rate": 1.5671968178376574e-06, "loss": 2.4657, "step": 3534 }, { "epoch": 4.18348623853211, "grad_norm": 4.40625, "learning_rate": 1.5627538781268591e-06, "loss": 2.6592, "step": 3535 }, { "epoch": 4.184670020716188, "grad_norm": 5.03125, "learning_rate": 1.5583167112693153e-06, "loss": 2.297, "step": 3536 }, { "epoch": 4.185853802900266, "grad_norm": 4.8125, "learning_rate": 1.553885320300983e-06, "loss": 2.0553, "step": 3537 }, { "epoch": 4.187037585084345, "grad_norm": 4.40625, "learning_rate": 1.549459708253863e-06, "loss": 2.53, "step": 3538 }, { "epoch": 4.188221367268422, "grad_norm": 5.25, "learning_rate": 1.5450398781560083e-06, "loss": 2.7967, "step": 3539 }, { "epoch": 4.189405149452501, "grad_norm": 5.125, "learning_rate": 1.5406258330315172e-06, "loss": 2.125, "step": 3540 }, { "epoch": 4.1905889316365785, "grad_norm": 4.34375, "learning_rate": 1.5362175759005204e-06, "loss": 2.2947, "step": 3541 }, { "epoch": 4.191772713820657, "grad_norm": 4.625, "learning_rate": 1.5318151097792e-06, "loss": 2.6241, "step": 3542 }, { "epoch": 4.192956496004735, "grad_norm": 4.65625, "learning_rate": 1.5274184376797662e-06, "loss": 2.3478, "step": 3543 }, { "epoch": 4.194140278188813, "grad_norm": 4.84375, "learning_rate": 1.5230275626104707e-06, "loss": 2.6976, "step": 3544 }, { "epoch": 4.195324060372891, "grad_norm": 4.75, "learning_rate": 1.518642487575591e-06, "loss": 2.7696, "step": 3545 }, { "epoch": 4.1965078425569695, "grad_norm": 4.96875, "learning_rate": 1.5142632155754478e-06, "loss": 2.4877, "step": 3546 }, { "epoch": 4.197691624741048, "grad_norm": 4.71875, "learning_rate": 1.509889749606387e-06, "loss": 2.2869, "step": 3547 }, { "epoch": 4.198875406925126, "grad_norm": 4.625, "learning_rate": 1.505522092660776e-06, "loss": 2.3866, "step": 3548 }, { "epoch": 4.200059189109204, "grad_norm": 5.125, "learning_rate": 1.5011602477270137e-06, "loss": 2.3655, "step": 3549 }, { "epoch": 4.201242971293282, "grad_norm": 4.125, "learning_rate": 1.496804217789518e-06, "loss": 1.9877, "step": 3550 }, { "epoch": 4.2024267534773605, "grad_norm": 4.59375, "learning_rate": 1.492454005828734e-06, "loss": 2.3497, "step": 3551 }, { "epoch": 4.203610535661438, "grad_norm": 4.625, "learning_rate": 1.4881096148211238e-06, "loss": 2.2255, "step": 3552 }, { "epoch": 4.204794317845517, "grad_norm": 4.875, "learning_rate": 1.483771047739161e-06, "loss": 2.5424, "step": 3553 }, { "epoch": 4.205978100029594, "grad_norm": 4.84375, "learning_rate": 1.4794383075513453e-06, "loss": 2.5105, "step": 3554 }, { "epoch": 4.207161882213673, "grad_norm": 4.84375, "learning_rate": 1.4751113972221797e-06, "loss": 2.4849, "step": 3555 }, { "epoch": 4.208345664397751, "grad_norm": 5.5625, "learning_rate": 1.470790319712183e-06, "loss": 2.7546, "step": 3556 }, { "epoch": 4.209529446581829, "grad_norm": 4.96875, "learning_rate": 1.4664750779778802e-06, "loss": 2.2054, "step": 3557 }, { "epoch": 4.210713228765907, "grad_norm": 4.25, "learning_rate": 1.4621656749718072e-06, "loss": 2.1944, "step": 3558 }, { "epoch": 4.211897010949985, "grad_norm": 6.125, "learning_rate": 1.4578621136425053e-06, "loss": 2.5673, "step": 3559 }, { "epoch": 4.213080793134063, "grad_norm": 4.625, "learning_rate": 1.4535643969345149e-06, "loss": 2.7356, "step": 3560 }, { "epoch": 4.214264575318142, "grad_norm": 5.03125, "learning_rate": 1.4492725277883791e-06, "loss": 2.5599, "step": 3561 }, { "epoch": 4.215448357502219, "grad_norm": 4.5, "learning_rate": 1.444986509140638e-06, "loss": 2.529, "step": 3562 }, { "epoch": 4.216632139686298, "grad_norm": 4.59375, "learning_rate": 1.4407063439238333e-06, "loss": 2.3386, "step": 3563 }, { "epoch": 4.2178159218703755, "grad_norm": 4.96875, "learning_rate": 1.4364320350665018e-06, "loss": 2.6546, "step": 3564 }, { "epoch": 4.218999704054454, "grad_norm": 4.78125, "learning_rate": 1.4321635854931671e-06, "loss": 2.7149, "step": 3565 }, { "epoch": 4.220183486238533, "grad_norm": 4.90625, "learning_rate": 1.4279009981243507e-06, "loss": 2.2408, "step": 3566 }, { "epoch": 4.22136726842261, "grad_norm": 4.4375, "learning_rate": 1.423644275876559e-06, "loss": 2.4464, "step": 3567 }, { "epoch": 4.222551050606689, "grad_norm": 4.71875, "learning_rate": 1.419393421662284e-06, "loss": 2.4357, "step": 3568 }, { "epoch": 4.2237348327907664, "grad_norm": 4.625, "learning_rate": 1.4151484383900093e-06, "loss": 2.305, "step": 3569 }, { "epoch": 4.224918614974845, "grad_norm": 4.9375, "learning_rate": 1.410909328964193e-06, "loss": 2.5828, "step": 3570 }, { "epoch": 4.226102397158923, "grad_norm": 5.03125, "learning_rate": 1.4066760962852832e-06, "loss": 2.3859, "step": 3571 }, { "epoch": 4.227286179343001, "grad_norm": 4.5625, "learning_rate": 1.4024487432497013e-06, "loss": 2.5123, "step": 3572 }, { "epoch": 4.228469961527079, "grad_norm": 5.0, "learning_rate": 1.398227272749846e-06, "loss": 2.5821, "step": 3573 }, { "epoch": 4.229653743711157, "grad_norm": 4.4375, "learning_rate": 1.3940116876740905e-06, "loss": 2.4118, "step": 3574 }, { "epoch": 4.230837525895235, "grad_norm": 5.09375, "learning_rate": 1.389801990906784e-06, "loss": 2.5656, "step": 3575 }, { "epoch": 4.232021308079314, "grad_norm": 4.4375, "learning_rate": 1.3855981853282496e-06, "loss": 2.41, "step": 3576 }, { "epoch": 4.233205090263391, "grad_norm": 4.53125, "learning_rate": 1.3814002738147704e-06, "loss": 2.4043, "step": 3577 }, { "epoch": 4.23438887244747, "grad_norm": 4.46875, "learning_rate": 1.377208259238606e-06, "loss": 2.5606, "step": 3578 }, { "epoch": 4.2355726546315475, "grad_norm": 4.28125, "learning_rate": 1.3730221444679748e-06, "loss": 2.2143, "step": 3579 }, { "epoch": 4.236756436815626, "grad_norm": 4.9375, "learning_rate": 1.3688419323670598e-06, "loss": 2.2149, "step": 3580 }, { "epoch": 4.237940218999704, "grad_norm": 5.34375, "learning_rate": 1.36466762579601e-06, "loss": 2.297, "step": 3581 }, { "epoch": 4.239124001183782, "grad_norm": 4.90625, "learning_rate": 1.3604992276109263e-06, "loss": 2.8893, "step": 3582 }, { "epoch": 4.24030778336786, "grad_norm": 5.15625, "learning_rate": 1.356336740663875e-06, "loss": 2.1526, "step": 3583 }, { "epoch": 4.2414915655519385, "grad_norm": 4.8125, "learning_rate": 1.3521801678028713e-06, "loss": 2.4469, "step": 3584 }, { "epoch": 4.242675347736016, "grad_norm": 4.375, "learning_rate": 1.3480295118718877e-06, "loss": 2.2397, "step": 3585 }, { "epoch": 4.243859129920095, "grad_norm": 4.40625, "learning_rate": 1.343884775710843e-06, "loss": 2.2801, "step": 3586 }, { "epoch": 4.245042912104173, "grad_norm": 4.4375, "learning_rate": 1.339745962155613e-06, "loss": 2.3714, "step": 3587 }, { "epoch": 4.246226694288251, "grad_norm": 4.84375, "learning_rate": 1.3356130740380203e-06, "loss": 2.6625, "step": 3588 }, { "epoch": 4.2474104764723295, "grad_norm": 5.21875, "learning_rate": 1.331486114185826e-06, "loss": 2.65, "step": 3589 }, { "epoch": 4.248594258656407, "grad_norm": 4.34375, "learning_rate": 1.3273650854227438e-06, "loss": 2.4587, "step": 3590 }, { "epoch": 4.249778040840486, "grad_norm": 4.59375, "learning_rate": 1.3232499905684226e-06, "loss": 2.4318, "step": 3591 }, { "epoch": 4.250961823024563, "grad_norm": 4.21875, "learning_rate": 1.3191408324384525e-06, "loss": 2.1808, "step": 3592 }, { "epoch": 4.252145605208642, "grad_norm": 4.84375, "learning_rate": 1.3150376138443676e-06, "loss": 2.4056, "step": 3593 }, { "epoch": 4.25332938739272, "grad_norm": 4.59375, "learning_rate": 1.3109403375936302e-06, "loss": 2.7228, "step": 3594 }, { "epoch": 4.254513169576798, "grad_norm": 5.15625, "learning_rate": 1.3068490064896422e-06, "loss": 2.792, "step": 3595 }, { "epoch": 4.255696951760876, "grad_norm": 4.4375, "learning_rate": 1.3027636233317342e-06, "loss": 2.1474, "step": 3596 }, { "epoch": 4.256880733944954, "grad_norm": 4.6875, "learning_rate": 1.2986841909151683e-06, "loss": 2.5336, "step": 3597 }, { "epoch": 4.258064516129032, "grad_norm": 4.34375, "learning_rate": 1.2946107120311369e-06, "loss": 2.187, "step": 3598 }, { "epoch": 4.259248298313111, "grad_norm": 4.6875, "learning_rate": 1.2905431894667552e-06, "loss": 2.1697, "step": 3599 }, { "epoch": 4.260432080497188, "grad_norm": 4.8125, "learning_rate": 1.2864816260050694e-06, "loss": 2.4644, "step": 3600 }, { "epoch": 4.261615862681267, "grad_norm": 4.90625, "learning_rate": 1.2824260244250385e-06, "loss": 2.4258, "step": 3601 }, { "epoch": 4.2627996448653445, "grad_norm": 5.03125, "learning_rate": 1.2783763875015542e-06, "loss": 2.0854, "step": 3602 }, { "epoch": 4.263983427049423, "grad_norm": 4.4375, "learning_rate": 1.2743327180054188e-06, "loss": 2.4626, "step": 3603 }, { "epoch": 4.265167209233501, "grad_norm": 4.65625, "learning_rate": 1.2702950187033502e-06, "loss": 2.1628, "step": 3604 }, { "epoch": 4.266350991417579, "grad_norm": 4.4375, "learning_rate": 1.2662632923579899e-06, "loss": 2.4593, "step": 3605 }, { "epoch": 4.267534773601657, "grad_norm": 4.75, "learning_rate": 1.2622375417278843e-06, "loss": 2.246, "step": 3606 }, { "epoch": 4.2687185557857354, "grad_norm": 4.625, "learning_rate": 1.258217769567499e-06, "loss": 2.4345, "step": 3607 }, { "epoch": 4.269902337969814, "grad_norm": 4.34375, "learning_rate": 1.2542039786272008e-06, "loss": 2.2518, "step": 3608 }, { "epoch": 4.271086120153892, "grad_norm": 4.03125, "learning_rate": 1.2501961716532685e-06, "loss": 2.2569, "step": 3609 }, { "epoch": 4.27226990233797, "grad_norm": 4.125, "learning_rate": 1.2461943513878883e-06, "loss": 2.1885, "step": 3610 }, { "epoch": 4.273453684522048, "grad_norm": 5.46875, "learning_rate": 1.2421985205691444e-06, "loss": 2.5154, "step": 3611 }, { "epoch": 4.274637466706126, "grad_norm": 4.3125, "learning_rate": 1.2382086819310312e-06, "loss": 2.5555, "step": 3612 }, { "epoch": 4.275821248890204, "grad_norm": 4.9375, "learning_rate": 1.234224838203435e-06, "loss": 2.4683, "step": 3613 }, { "epoch": 4.277005031074283, "grad_norm": 4.53125, "learning_rate": 1.2302469921121462e-06, "loss": 2.1651, "step": 3614 }, { "epoch": 4.27818881325836, "grad_norm": 4.65625, "learning_rate": 1.226275146378848e-06, "loss": 2.2472, "step": 3615 }, { "epoch": 4.279372595442439, "grad_norm": 4.78125, "learning_rate": 1.2223093037211186e-06, "loss": 2.0002, "step": 3616 }, { "epoch": 4.2805563776265165, "grad_norm": 4.0625, "learning_rate": 1.218349466852432e-06, "loss": 2.1396, "step": 3617 }, { "epoch": 4.281740159810595, "grad_norm": 5.09375, "learning_rate": 1.2143956384821475e-06, "loss": 2.3543, "step": 3618 }, { "epoch": 4.282923941994673, "grad_norm": 4.5, "learning_rate": 1.2104478213155202e-06, "loss": 2.0911, "step": 3619 }, { "epoch": 4.284107724178751, "grad_norm": 4.65625, "learning_rate": 1.2065060180536858e-06, "loss": 2.4308, "step": 3620 }, { "epoch": 4.285291506362829, "grad_norm": 4.5625, "learning_rate": 1.202570231393666e-06, "loss": 2.3528, "step": 3621 }, { "epoch": 4.2864752885469075, "grad_norm": 4.71875, "learning_rate": 1.1986404640283734e-06, "loss": 2.5092, "step": 3622 }, { "epoch": 4.287659070730985, "grad_norm": 4.5, "learning_rate": 1.1947167186465903e-06, "loss": 2.1825, "step": 3623 }, { "epoch": 4.288842852915064, "grad_norm": 4.625, "learning_rate": 1.1907989979329904e-06, "loss": 2.4468, "step": 3624 }, { "epoch": 4.290026635099141, "grad_norm": 4.875, "learning_rate": 1.186887304568115e-06, "loss": 2.1283, "step": 3625 }, { "epoch": 4.29121041728322, "grad_norm": 4.5, "learning_rate": 1.1829816412283912e-06, "loss": 2.4531, "step": 3626 }, { "epoch": 4.292394199467298, "grad_norm": 4.8125, "learning_rate": 1.1790820105861123e-06, "loss": 2.2134, "step": 3627 }, { "epoch": 4.293577981651376, "grad_norm": 4.4375, "learning_rate": 1.175188415309444e-06, "loss": 2.3501, "step": 3628 }, { "epoch": 4.294761763835455, "grad_norm": 4.625, "learning_rate": 1.1713008580624319e-06, "loss": 2.6299, "step": 3629 }, { "epoch": 4.295945546019532, "grad_norm": 4.46875, "learning_rate": 1.1674193415049773e-06, "loss": 2.2, "step": 3630 }, { "epoch": 4.297129328203611, "grad_norm": 4.75, "learning_rate": 1.1635438682928613e-06, "loss": 2.304, "step": 3631 }, { "epoch": 4.298313110387689, "grad_norm": 5.5625, "learning_rate": 1.1596744410777206e-06, "loss": 2.7069, "step": 3632 }, { "epoch": 4.299496892571767, "grad_norm": 4.5625, "learning_rate": 1.155811062507055e-06, "loss": 2.1436, "step": 3633 }, { "epoch": 4.300680674755845, "grad_norm": 4.875, "learning_rate": 1.151953735224236e-06, "loss": 2.5101, "step": 3634 }, { "epoch": 4.301864456939923, "grad_norm": 3.75, "learning_rate": 1.1481024618684821e-06, "loss": 2.173, "step": 3635 }, { "epoch": 4.303048239124001, "grad_norm": 4.3125, "learning_rate": 1.14425724507488e-06, "loss": 2.2432, "step": 3636 }, { "epoch": 4.30423202130808, "grad_norm": 4.375, "learning_rate": 1.140418087474362e-06, "loss": 2.3933, "step": 3637 }, { "epoch": 4.305415803492157, "grad_norm": 5.25, "learning_rate": 1.1365849916937277e-06, "loss": 2.5649, "step": 3638 }, { "epoch": 4.306599585676236, "grad_norm": 5.4375, "learning_rate": 1.1327579603556172e-06, "loss": 2.2709, "step": 3639 }, { "epoch": 4.3077833678603135, "grad_norm": 5.03125, "learning_rate": 1.1289369960785234e-06, "loss": 2.5446, "step": 3640 }, { "epoch": 4.308967150044392, "grad_norm": 4.71875, "learning_rate": 1.1251221014767977e-06, "loss": 2.5356, "step": 3641 }, { "epoch": 4.31015093222847, "grad_norm": 4.65625, "learning_rate": 1.1213132791606251e-06, "loss": 2.2063, "step": 3642 }, { "epoch": 4.311334714412548, "grad_norm": 4.5, "learning_rate": 1.1175105317360446e-06, "loss": 2.4536, "step": 3643 }, { "epoch": 4.312518496596626, "grad_norm": 4.625, "learning_rate": 1.1137138618049403e-06, "loss": 2.4671, "step": 3644 }, { "epoch": 4.3137022787807044, "grad_norm": 4.875, "learning_rate": 1.1099232719650265e-06, "loss": 2.4258, "step": 3645 }, { "epoch": 4.314886060964782, "grad_norm": 5.25, "learning_rate": 1.106138764809871e-06, "loss": 2.4125, "step": 3646 }, { "epoch": 4.316069843148861, "grad_norm": 5.0, "learning_rate": 1.1023603429288688e-06, "loss": 2.6761, "step": 3647 }, { "epoch": 4.317253625332938, "grad_norm": 4.8125, "learning_rate": 1.0985880089072609e-06, "loss": 2.079, "step": 3648 }, { "epoch": 4.318437407517017, "grad_norm": 4.6875, "learning_rate": 1.094821765326114e-06, "loss": 2.8437, "step": 3649 }, { "epoch": 4.319621189701095, "grad_norm": 4.53125, "learning_rate": 1.0910616147623365e-06, "loss": 2.1528, "step": 3650 }, { "epoch": 4.320804971885173, "grad_norm": 4.625, "learning_rate": 1.0873075597886607e-06, "loss": 2.7085, "step": 3651 }, { "epoch": 4.321988754069252, "grad_norm": 5.0625, "learning_rate": 1.0835596029736484e-06, "loss": 2.6192, "step": 3652 }, { "epoch": 4.323172536253329, "grad_norm": 5.625, "learning_rate": 1.079817746881696e-06, "loss": 2.5744, "step": 3653 }, { "epoch": 4.324356318437408, "grad_norm": 5.1875, "learning_rate": 1.076081994073017e-06, "loss": 2.3858, "step": 3654 }, { "epoch": 4.3255401006214855, "grad_norm": 4.3125, "learning_rate": 1.0723523471036545e-06, "loss": 2.2712, "step": 3655 }, { "epoch": 4.326723882805564, "grad_norm": 4.625, "learning_rate": 1.0686288085254782e-06, "loss": 2.4658, "step": 3656 }, { "epoch": 4.327907664989642, "grad_norm": 4.5625, "learning_rate": 1.0649113808861633e-06, "loss": 2.2904, "step": 3657 }, { "epoch": 4.32909144717372, "grad_norm": 4.9375, "learning_rate": 1.0612000667292188e-06, "loss": 2.2173, "step": 3658 }, { "epoch": 4.330275229357798, "grad_norm": 4.40625, "learning_rate": 1.0574948685939624e-06, "loss": 2.5635, "step": 3659 }, { "epoch": 4.3314590115418765, "grad_norm": 4.25, "learning_rate": 1.0537957890155338e-06, "loss": 2.2352, "step": 3660 }, { "epoch": 4.332642793725954, "grad_norm": 4.4375, "learning_rate": 1.0501028305248773e-06, "loss": 2.5207, "step": 3661 }, { "epoch": 4.333826575910033, "grad_norm": 5.0, "learning_rate": 1.0464159956487596e-06, "loss": 2.4558, "step": 3662 }, { "epoch": 4.33501035809411, "grad_norm": 4.96875, "learning_rate": 1.0427352869097495e-06, "loss": 2.6259, "step": 3663 }, { "epoch": 4.336194140278189, "grad_norm": 4.78125, "learning_rate": 1.0390607068262248e-06, "loss": 2.6184, "step": 3664 }, { "epoch": 4.337377922462267, "grad_norm": 4.46875, "learning_rate": 1.0353922579123765e-06, "loss": 2.1921, "step": 3665 }, { "epoch": 4.338561704646345, "grad_norm": 4.78125, "learning_rate": 1.0317299426781924e-06, "loss": 2.4328, "step": 3666 }, { "epoch": 4.339745486830423, "grad_norm": 4.90625, "learning_rate": 1.028073763629469e-06, "loss": 2.2431, "step": 3667 }, { "epoch": 4.340929269014501, "grad_norm": 4.6875, "learning_rate": 1.0244237232678067e-06, "loss": 2.4765, "step": 3668 }, { "epoch": 4.342113051198579, "grad_norm": 5.0625, "learning_rate": 1.0207798240905954e-06, "loss": 2.4459, "step": 3669 }, { "epoch": 4.343296833382658, "grad_norm": 4.84375, "learning_rate": 1.0171420685910328e-06, "loss": 2.3271, "step": 3670 }, { "epoch": 4.344480615566736, "grad_norm": 4.3125, "learning_rate": 1.013510459258108e-06, "loss": 2.6613, "step": 3671 }, { "epoch": 4.345664397750814, "grad_norm": 4.5, "learning_rate": 1.0098849985766067e-06, "loss": 2.4187, "step": 3672 }, { "epoch": 4.346848179934892, "grad_norm": 5.09375, "learning_rate": 1.0062656890271116e-06, "loss": 2.2971, "step": 3673 }, { "epoch": 4.34803196211897, "grad_norm": 4.53125, "learning_rate": 1.0026525330859904e-06, "loss": 2.3503, "step": 3674 }, { "epoch": 4.349215744303049, "grad_norm": 5.21875, "learning_rate": 9.990455332254012e-07, "loss": 2.2491, "step": 3675 }, { "epoch": 4.350399526487126, "grad_norm": 4.75, "learning_rate": 9.954446919132898e-07, "loss": 2.3431, "step": 3676 }, { "epoch": 4.351583308671205, "grad_norm": 4.59375, "learning_rate": 9.918500116133957e-07, "loss": 2.3158, "step": 3677 }, { "epoch": 4.3527670908552825, "grad_norm": 5.125, "learning_rate": 9.88261494785232e-07, "loss": 2.4618, "step": 3678 }, { "epoch": 4.353950873039361, "grad_norm": 4.6875, "learning_rate": 9.84679143884102e-07, "loss": 2.4937, "step": 3679 }, { "epoch": 4.355134655223439, "grad_norm": 4.625, "learning_rate": 9.811029613610913e-07, "loss": 2.4987, "step": 3680 }, { "epoch": 4.356318437407517, "grad_norm": 4.09375, "learning_rate": 9.775329496630559e-07, "loss": 2.1916, "step": 3681 }, { "epoch": 4.357502219591595, "grad_norm": 4.59375, "learning_rate": 9.73969111232641e-07, "loss": 2.1315, "step": 3682 }, { "epoch": 4.3586860017756734, "grad_norm": 4.625, "learning_rate": 9.704114485082571e-07, "loss": 2.2933, "step": 3683 }, { "epoch": 4.359869783959751, "grad_norm": 4.125, "learning_rate": 9.668599639240993e-07, "loss": 2.2141, "step": 3684 }, { "epoch": 4.36105356614383, "grad_norm": 3.8125, "learning_rate": 9.633146599101306e-07, "loss": 2.1963, "step": 3685 }, { "epoch": 4.362237348327907, "grad_norm": 4.1875, "learning_rate": 9.597755388920848e-07, "loss": 2.1733, "step": 3686 }, { "epoch": 4.363421130511986, "grad_norm": 5.1875, "learning_rate": 9.562426032914663e-07, "loss": 2.4011, "step": 3687 }, { "epoch": 4.3646049126960635, "grad_norm": 4.65625, "learning_rate": 9.527158555255445e-07, "loss": 2.4802, "step": 3688 }, { "epoch": 4.365788694880142, "grad_norm": 4.84375, "learning_rate": 9.491952980073604e-07, "loss": 2.3232, "step": 3689 }, { "epoch": 4.36697247706422, "grad_norm": 4.40625, "learning_rate": 9.456809331457173e-07, "loss": 2.6518, "step": 3690 }, { "epoch": 4.368156259248298, "grad_norm": 4.4375, "learning_rate": 9.421727633451783e-07, "loss": 1.9556, "step": 3691 }, { "epoch": 4.369340041432377, "grad_norm": 4.65625, "learning_rate": 9.386707910060755e-07, "loss": 2.4678, "step": 3692 }, { "epoch": 4.3705238236164545, "grad_norm": 4.71875, "learning_rate": 9.351750185244879e-07, "loss": 2.7121, "step": 3693 }, { "epoch": 4.371707605800533, "grad_norm": 4.3125, "learning_rate": 9.316854482922655e-07, "loss": 2.1498, "step": 3694 }, { "epoch": 4.372891387984611, "grad_norm": 4.875, "learning_rate": 9.282020826970061e-07, "loss": 2.3308, "step": 3695 }, { "epoch": 4.374075170168689, "grad_norm": 4.65625, "learning_rate": 9.247249241220679e-07, "loss": 2.5338, "step": 3696 }, { "epoch": 4.375258952352767, "grad_norm": 4.90625, "learning_rate": 9.212539749465598e-07, "loss": 2.9066, "step": 3697 }, { "epoch": 4.3764427345368455, "grad_norm": 4.96875, "learning_rate": 9.177892375453413e-07, "loss": 2.3806, "step": 3698 }, { "epoch": 4.377626516720923, "grad_norm": 5.375, "learning_rate": 9.143307142890224e-07, "loss": 2.8194, "step": 3699 }, { "epoch": 4.378810298905002, "grad_norm": 4.8125, "learning_rate": 9.108784075439603e-07, "loss": 2.4913, "step": 3700 }, { "epoch": 4.379994081089079, "grad_norm": 5.0625, "learning_rate": 9.074323196722589e-07, "loss": 2.4638, "step": 3701 }, { "epoch": 4.381177863273158, "grad_norm": 4.25, "learning_rate": 9.039924530317734e-07, "loss": 2.3021, "step": 3702 }, { "epoch": 4.382361645457236, "grad_norm": 4.09375, "learning_rate": 9.005588099760909e-07, "loss": 2.642, "step": 3703 }, { "epoch": 4.383545427641314, "grad_norm": 4.8125, "learning_rate": 8.971313928545522e-07, "loss": 2.3139, "step": 3704 }, { "epoch": 4.384729209825392, "grad_norm": 4.78125, "learning_rate": 8.93710204012227e-07, "loss": 2.3511, "step": 3705 }, { "epoch": 4.38591299200947, "grad_norm": 4.75, "learning_rate": 8.902952457899317e-07, "loss": 2.5486, "step": 3706 }, { "epoch": 4.387096774193548, "grad_norm": 4.53125, "learning_rate": 8.86886520524216e-07, "loss": 2.3987, "step": 3707 }, { "epoch": 4.388280556377627, "grad_norm": 4.78125, "learning_rate": 8.834840305473658e-07, "loss": 2.149, "step": 3708 }, { "epoch": 4.389464338561704, "grad_norm": 3.859375, "learning_rate": 8.800877781874028e-07, "loss": 1.7665, "step": 3709 }, { "epoch": 4.390648120745783, "grad_norm": 4.34375, "learning_rate": 8.766977657680776e-07, "loss": 2.4213, "step": 3710 }, { "epoch": 4.3918319029298605, "grad_norm": 4.71875, "learning_rate": 8.733139956088732e-07, "loss": 2.6303, "step": 3711 }, { "epoch": 4.393015685113939, "grad_norm": 4.5625, "learning_rate": 8.699364700249979e-07, "loss": 2.2111, "step": 3712 }, { "epoch": 4.394199467298018, "grad_norm": 4.59375, "learning_rate": 8.665651913273931e-07, "loss": 2.4387, "step": 3713 }, { "epoch": 4.395383249482095, "grad_norm": 4.3125, "learning_rate": 8.632001618227248e-07, "loss": 2.409, "step": 3714 }, { "epoch": 4.396567031666174, "grad_norm": 4.875, "learning_rate": 8.59841383813379e-07, "loss": 2.5266, "step": 3715 }, { "epoch": 4.3977508138502515, "grad_norm": 4.1875, "learning_rate": 8.564888595974718e-07, "loss": 1.8161, "step": 3716 }, { "epoch": 4.39893459603433, "grad_norm": 4.40625, "learning_rate": 8.531425914688285e-07, "loss": 2.5286, "step": 3717 }, { "epoch": 4.400118378218408, "grad_norm": 5.34375, "learning_rate": 8.498025817170063e-07, "loss": 2.4194, "step": 3718 }, { "epoch": 4.401302160402486, "grad_norm": 4.1875, "learning_rate": 8.464688326272752e-07, "loss": 2.1851, "step": 3719 }, { "epoch": 4.402485942586564, "grad_norm": 4.6875, "learning_rate": 8.431413464806193e-07, "loss": 2.4494, "step": 3720 }, { "epoch": 4.4036697247706424, "grad_norm": 4.6875, "learning_rate": 8.398201255537431e-07, "loss": 2.7874, "step": 3721 }, { "epoch": 4.40485350695472, "grad_norm": 4.375, "learning_rate": 8.365051721190598e-07, "loss": 2.248, "step": 3722 }, { "epoch": 4.406037289138799, "grad_norm": 4.65625, "learning_rate": 8.331964884446953e-07, "loss": 2.374, "step": 3723 }, { "epoch": 4.407221071322876, "grad_norm": 4.28125, "learning_rate": 8.298940767944841e-07, "loss": 2.2773, "step": 3724 }, { "epoch": 4.408404853506955, "grad_norm": 5.1875, "learning_rate": 8.265979394279732e-07, "loss": 3.0118, "step": 3725 }, { "epoch": 4.4095886356910325, "grad_norm": 4.5625, "learning_rate": 8.233080786004167e-07, "loss": 2.5272, "step": 3726 }, { "epoch": 4.410772417875111, "grad_norm": 4.4375, "learning_rate": 8.20024496562768e-07, "loss": 2.4349, "step": 3727 }, { "epoch": 4.411956200059189, "grad_norm": 4.6875, "learning_rate": 8.167471955616946e-07, "loss": 2.3924, "step": 3728 }, { "epoch": 4.413139982243267, "grad_norm": 4.4375, "learning_rate": 8.134761778395539e-07, "loss": 2.5057, "step": 3729 }, { "epoch": 4.414323764427345, "grad_norm": 5.125, "learning_rate": 8.102114456344145e-07, "loss": 2.521, "step": 3730 }, { "epoch": 4.4155075466114235, "grad_norm": 4.9375, "learning_rate": 8.06953001180042e-07, "loss": 2.3293, "step": 3731 }, { "epoch": 4.416691328795501, "grad_norm": 4.84375, "learning_rate": 8.037008467058948e-07, "loss": 2.6979, "step": 3732 }, { "epoch": 4.41787511097958, "grad_norm": 5.375, "learning_rate": 8.004549844371357e-07, "loss": 2.3321, "step": 3733 }, { "epoch": 4.419058893163658, "grad_norm": 4.875, "learning_rate": 7.972154165946155e-07, "loss": 2.4076, "step": 3734 }, { "epoch": 4.420242675347736, "grad_norm": 4.84375, "learning_rate": 7.939821453948826e-07, "loss": 2.4313, "step": 3735 }, { "epoch": 4.4214264575318145, "grad_norm": 4.53125, "learning_rate": 7.907551730501717e-07, "loss": 2.7446, "step": 3736 }, { "epoch": 4.422610239715892, "grad_norm": 4.96875, "learning_rate": 7.875345017684144e-07, "loss": 2.3618, "step": 3737 }, { "epoch": 4.423794021899971, "grad_norm": 4.03125, "learning_rate": 7.843201337532292e-07, "loss": 1.9729, "step": 3738 }, { "epoch": 4.424977804084048, "grad_norm": 4.84375, "learning_rate": 7.811120712039178e-07, "loss": 2.4757, "step": 3739 }, { "epoch": 4.426161586268127, "grad_norm": 5.0, "learning_rate": 7.779103163154756e-07, "loss": 2.547, "step": 3740 }, { "epoch": 4.427345368452205, "grad_norm": 5.25, "learning_rate": 7.747148712785701e-07, "loss": 2.3456, "step": 3741 }, { "epoch": 4.428529150636283, "grad_norm": 4.28125, "learning_rate": 7.715257382795627e-07, "loss": 2.5021, "step": 3742 }, { "epoch": 4.429712932820361, "grad_norm": 5.15625, "learning_rate": 7.683429195004932e-07, "loss": 2.3407, "step": 3743 }, { "epoch": 4.430896715004439, "grad_norm": 4.1875, "learning_rate": 7.651664171190765e-07, "loss": 2.2293, "step": 3744 }, { "epoch": 4.432080497188517, "grad_norm": 4.5625, "learning_rate": 7.619962333087138e-07, "loss": 2.3084, "step": 3745 }, { "epoch": 4.433264279372596, "grad_norm": 5.09375, "learning_rate": 7.588323702384748e-07, "loss": 2.4068, "step": 3746 }, { "epoch": 4.434448061556673, "grad_norm": 4.6875, "learning_rate": 7.556748300731087e-07, "loss": 2.1159, "step": 3747 }, { "epoch": 4.435631843740752, "grad_norm": 4.3125, "learning_rate": 7.525236149730397e-07, "loss": 2.2697, "step": 3748 }, { "epoch": 4.4368156259248295, "grad_norm": 5.0, "learning_rate": 7.493787270943586e-07, "loss": 2.5805, "step": 3749 }, { "epoch": 4.437999408108908, "grad_norm": 5.15625, "learning_rate": 7.462401685888365e-07, "loss": 2.3954, "step": 3750 }, { "epoch": 4.439183190292986, "grad_norm": 4.9375, "learning_rate": 7.431079416039045e-07, "loss": 2.7598, "step": 3751 }, { "epoch": 4.440366972477064, "grad_norm": 3.890625, "learning_rate": 7.399820482826692e-07, "loss": 2.1313, "step": 3752 }, { "epoch": 4.441550754661142, "grad_norm": 4.5625, "learning_rate": 7.368624907638944e-07, "loss": 2.3439, "step": 3753 }, { "epoch": 4.4427345368452205, "grad_norm": 4.96875, "learning_rate": 7.337492711820171e-07, "loss": 2.5782, "step": 3754 }, { "epoch": 4.443918319029299, "grad_norm": 4.25, "learning_rate": 7.306423916671379e-07, "loss": 2.1385, "step": 3755 }, { "epoch": 4.445102101213377, "grad_norm": 4.53125, "learning_rate": 7.275418543450119e-07, "loss": 2.3959, "step": 3756 }, { "epoch": 4.446285883397455, "grad_norm": 4.8125, "learning_rate": 7.244476613370644e-07, "loss": 2.334, "step": 3757 }, { "epoch": 4.447469665581533, "grad_norm": 5.09375, "learning_rate": 7.213598147603717e-07, "loss": 2.4839, "step": 3758 }, { "epoch": 4.4486534477656114, "grad_norm": 4.3125, "learning_rate": 7.182783167276719e-07, "loss": 2.3583, "step": 3759 }, { "epoch": 4.449837229949689, "grad_norm": 4.71875, "learning_rate": 7.152031693473593e-07, "loss": 2.5888, "step": 3760 }, { "epoch": 4.451021012133768, "grad_norm": 4.5, "learning_rate": 7.12134374723481e-07, "loss": 2.4223, "step": 3761 }, { "epoch": 4.452204794317845, "grad_norm": 4.3125, "learning_rate": 7.090719349557407e-07, "loss": 2.6527, "step": 3762 }, { "epoch": 4.453388576501924, "grad_norm": 4.34375, "learning_rate": 7.0601585213949e-07, "loss": 2.4754, "step": 3763 }, { "epoch": 4.4545723586860015, "grad_norm": 4.78125, "learning_rate": 7.029661283657385e-07, "loss": 2.5143, "step": 3764 }, { "epoch": 4.45575614087008, "grad_norm": 5.34375, "learning_rate": 6.999227657211316e-07, "loss": 2.9617, "step": 3765 }, { "epoch": 4.456939923054158, "grad_norm": 4.1875, "learning_rate": 6.968857662879735e-07, "loss": 2.3163, "step": 3766 }, { "epoch": 4.458123705238236, "grad_norm": 4.84375, "learning_rate": 6.938551321442144e-07, "loss": 2.7237, "step": 3767 }, { "epoch": 4.459307487422314, "grad_norm": 4.03125, "learning_rate": 6.908308653634421e-07, "loss": 2.1324, "step": 3768 }, { "epoch": 4.4604912696063925, "grad_norm": 4.5, "learning_rate": 6.878129680148948e-07, "loss": 2.1603, "step": 3769 }, { "epoch": 4.46167505179047, "grad_norm": 4.9375, "learning_rate": 6.848014421634497e-07, "loss": 2.2724, "step": 3770 }, { "epoch": 4.462858833974549, "grad_norm": 4.3125, "learning_rate": 6.817962898696218e-07, "loss": 2.2474, "step": 3771 }, { "epoch": 4.464042616158626, "grad_norm": 4.625, "learning_rate": 6.787975131895718e-07, "loss": 2.2083, "step": 3772 }, { "epoch": 4.465226398342705, "grad_norm": 4.90625, "learning_rate": 6.758051141750921e-07, "loss": 1.9625, "step": 3773 }, { "epoch": 4.466410180526783, "grad_norm": 4.375, "learning_rate": 6.728190948736158e-07, "loss": 2.2312, "step": 3774 }, { "epoch": 4.467593962710861, "grad_norm": 4.84375, "learning_rate": 6.698394573282063e-07, "loss": 2.3374, "step": 3775 }, { "epoch": 4.46877774489494, "grad_norm": 4.4375, "learning_rate": 6.668662035775674e-07, "loss": 2.2264, "step": 3776 }, { "epoch": 4.469961527079017, "grad_norm": 4.625, "learning_rate": 6.638993356560275e-07, "loss": 2.4778, "step": 3777 }, { "epoch": 4.471145309263096, "grad_norm": 4.875, "learning_rate": 6.609388555935481e-07, "loss": 2.1753, "step": 3778 }, { "epoch": 4.472329091447174, "grad_norm": 5.0625, "learning_rate": 6.579847654157234e-07, "loss": 2.4476, "step": 3779 }, { "epoch": 4.473512873631252, "grad_norm": 3.96875, "learning_rate": 6.550370671437723e-07, "loss": 1.9698, "step": 3780 }, { "epoch": 4.47469665581533, "grad_norm": 4.4375, "learning_rate": 6.520957627945424e-07, "loss": 2.293, "step": 3781 }, { "epoch": 4.475880437999408, "grad_norm": 4.84375, "learning_rate": 6.49160854380505e-07, "loss": 2.3312, "step": 3782 }, { "epoch": 4.477064220183486, "grad_norm": 5.0625, "learning_rate": 6.462323439097528e-07, "loss": 2.5526, "step": 3783 }, { "epoch": 4.478248002367565, "grad_norm": 4.78125, "learning_rate": 6.433102333860075e-07, "loss": 2.5222, "step": 3784 }, { "epoch": 4.479431784551642, "grad_norm": 4.875, "learning_rate": 6.403945248086052e-07, "loss": 2.2898, "step": 3785 }, { "epoch": 4.480615566735721, "grad_norm": 5.1875, "learning_rate": 6.374852201725079e-07, "loss": 2.5589, "step": 3786 }, { "epoch": 4.4817993489197985, "grad_norm": 4.625, "learning_rate": 6.345823214682889e-07, "loss": 2.2657, "step": 3787 }, { "epoch": 4.482983131103877, "grad_norm": 5.40625, "learning_rate": 6.316858306821449e-07, "loss": 2.65, "step": 3788 }, { "epoch": 4.484166913287955, "grad_norm": 4.3125, "learning_rate": 6.287957497958852e-07, "loss": 1.9965, "step": 3789 }, { "epoch": 4.485350695472033, "grad_norm": 5.28125, "learning_rate": 6.259120807869324e-07, "loss": 2.3375, "step": 3790 }, { "epoch": 4.486534477656111, "grad_norm": 4.53125, "learning_rate": 6.230348256283247e-07, "loss": 2.2392, "step": 3791 }, { "epoch": 4.4877182598401895, "grad_norm": 4.0, "learning_rate": 6.201639862887099e-07, "loss": 2.1753, "step": 3792 }, { "epoch": 4.488902042024267, "grad_norm": 5.0, "learning_rate": 6.172995647323477e-07, "loss": 2.58, "step": 3793 }, { "epoch": 4.490085824208346, "grad_norm": 5.125, "learning_rate": 6.144415629191058e-07, "loss": 2.7015, "step": 3794 }, { "epoch": 4.491269606392423, "grad_norm": 4.90625, "learning_rate": 6.115899828044569e-07, "loss": 2.5566, "step": 3795 }, { "epoch": 4.492453388576502, "grad_norm": 4.59375, "learning_rate": 6.087448263394846e-07, "loss": 2.7246, "step": 3796 }, { "epoch": 4.4936371707605804, "grad_norm": 4.375, "learning_rate": 6.05906095470874e-07, "loss": 2.0082, "step": 3797 }, { "epoch": 4.494820952944658, "grad_norm": 4.28125, "learning_rate": 6.030737921409169e-07, "loss": 2.5085, "step": 3798 }, { "epoch": 4.496004735128737, "grad_norm": 4.53125, "learning_rate": 6.002479182875021e-07, "loss": 2.3416, "step": 3799 }, { "epoch": 4.497188517312814, "grad_norm": 4.90625, "learning_rate": 5.974284758441251e-07, "loss": 2.399, "step": 3800 }, { "epoch": 4.498372299496893, "grad_norm": 4.375, "learning_rate": 5.946154667398785e-07, "loss": 2.1852, "step": 3801 }, { "epoch": 4.4995560816809705, "grad_norm": 5.0, "learning_rate": 5.918088928994492e-07, "loss": 2.7057, "step": 3802 }, { "epoch": 4.500739863865049, "grad_norm": 4.5625, "learning_rate": 5.890087562431301e-07, "loss": 2.4569, "step": 3803 }, { "epoch": 4.501923646049127, "grad_norm": 4.90625, "learning_rate": 5.862150586867998e-07, "loss": 2.2805, "step": 3804 }, { "epoch": 4.503107428233205, "grad_norm": 5.03125, "learning_rate": 5.834278021419382e-07, "loss": 2.6523, "step": 3805 }, { "epoch": 4.504291210417283, "grad_norm": 5.46875, "learning_rate": 5.806469885156163e-07, "loss": 2.619, "step": 3806 }, { "epoch": 4.5054749926013615, "grad_norm": 5.03125, "learning_rate": 5.778726197104934e-07, "loss": 2.8172, "step": 3807 }, { "epoch": 4.506658774785439, "grad_norm": 4.21875, "learning_rate": 5.751046976248254e-07, "loss": 2.2752, "step": 3808 }, { "epoch": 4.507842556969518, "grad_norm": 4.6875, "learning_rate": 5.723432241524507e-07, "loss": 2.3532, "step": 3809 }, { "epoch": 4.509026339153595, "grad_norm": 4.5625, "learning_rate": 5.695882011828024e-07, "loss": 2.437, "step": 3810 }, { "epoch": 4.510210121337674, "grad_norm": 4.5, "learning_rate": 5.668396306008928e-07, "loss": 2.529, "step": 3811 }, { "epoch": 4.511393903521752, "grad_norm": 4.21875, "learning_rate": 5.640975142873262e-07, "loss": 2.395, "step": 3812 }, { "epoch": 4.51257768570583, "grad_norm": 4.5, "learning_rate": 5.613618541182852e-07, "loss": 2.1278, "step": 3813 }, { "epoch": 4.513761467889909, "grad_norm": 4.65625, "learning_rate": 5.586326519655383e-07, "loss": 2.4158, "step": 3814 }, { "epoch": 4.514945250073986, "grad_norm": 4.59375, "learning_rate": 5.55909909696436e-07, "loss": 2.4119, "step": 3815 }, { "epoch": 4.516129032258064, "grad_norm": 4.71875, "learning_rate": 5.531936291739037e-07, "loss": 2.16, "step": 3816 }, { "epoch": 4.517312814442143, "grad_norm": 4.71875, "learning_rate": 5.504838122564537e-07, "loss": 2.3642, "step": 3817 }, { "epoch": 4.518496596626221, "grad_norm": 4.96875, "learning_rate": 5.477804607981707e-07, "loss": 2.5445, "step": 3818 }, { "epoch": 4.519680378810299, "grad_norm": 4.65625, "learning_rate": 5.450835766487128e-07, "loss": 2.5617, "step": 3819 }, { "epoch": 4.520864160994377, "grad_norm": 4.53125, "learning_rate": 5.423931616533207e-07, "loss": 2.6179, "step": 3820 }, { "epoch": 4.522047943178455, "grad_norm": 4.625, "learning_rate": 5.397092176528018e-07, "loss": 2.4574, "step": 3821 }, { "epoch": 4.523231725362534, "grad_norm": 4.59375, "learning_rate": 5.370317464835406e-07, "loss": 1.9866, "step": 3822 }, { "epoch": 4.524415507546611, "grad_norm": 4.6875, "learning_rate": 5.343607499774928e-07, "loss": 2.344, "step": 3823 }, { "epoch": 4.52559928973069, "grad_norm": 4.4375, "learning_rate": 5.316962299621808e-07, "loss": 2.4111, "step": 3824 }, { "epoch": 4.5267830719147675, "grad_norm": 5.1875, "learning_rate": 5.290381882606977e-07, "loss": 2.5011, "step": 3825 }, { "epoch": 4.527966854098846, "grad_norm": 4.84375, "learning_rate": 5.263866266917028e-07, "loss": 2.0953, "step": 3826 }, { "epoch": 4.529150636282924, "grad_norm": 4.1875, "learning_rate": 5.23741547069424e-07, "loss": 2.3992, "step": 3827 }, { "epoch": 4.530334418467002, "grad_norm": 4.5625, "learning_rate": 5.211029512036514e-07, "loss": 2.5933, "step": 3828 }, { "epoch": 4.53151820065108, "grad_norm": 4.78125, "learning_rate": 5.184708408997429e-07, "loss": 2.456, "step": 3829 }, { "epoch": 4.5327019828351585, "grad_norm": 4.09375, "learning_rate": 5.158452179586148e-07, "loss": 2.2266, "step": 3830 }, { "epoch": 4.533885765019236, "grad_norm": 4.84375, "learning_rate": 5.132260841767444e-07, "loss": 2.3519, "step": 3831 }, { "epoch": 4.535069547203315, "grad_norm": 4.875, "learning_rate": 5.106134413461739e-07, "loss": 2.6917, "step": 3832 }, { "epoch": 4.536253329387392, "grad_norm": 4.75, "learning_rate": 5.080072912544987e-07, "loss": 2.3365, "step": 3833 }, { "epoch": 4.537437111571471, "grad_norm": 4.375, "learning_rate": 5.054076356848748e-07, "loss": 2.548, "step": 3834 }, { "epoch": 4.5386208937555494, "grad_norm": 5.3125, "learning_rate": 5.028144764160159e-07, "loss": 2.7304, "step": 3835 }, { "epoch": 4.539804675939627, "grad_norm": 4.9375, "learning_rate": 5.002278152221885e-07, "loss": 2.0731, "step": 3836 }, { "epoch": 4.540988458123705, "grad_norm": 4.84375, "learning_rate": 4.976476538732123e-07, "loss": 2.1406, "step": 3837 }, { "epoch": 4.542172240307783, "grad_norm": 4.875, "learning_rate": 4.950739941344607e-07, "loss": 1.9704, "step": 3838 }, { "epoch": 4.543356022491862, "grad_norm": 4.46875, "learning_rate": 4.925068377668618e-07, "loss": 2.523, "step": 3839 }, { "epoch": 4.5445398046759395, "grad_norm": 4.53125, "learning_rate": 4.899461865268873e-07, "loss": 2.511, "step": 3840 }, { "epoch": 4.545723586860018, "grad_norm": 4.71875, "learning_rate": 4.873920421665657e-07, "loss": 2.7051, "step": 3841 }, { "epoch": 4.546907369044096, "grad_norm": 4.375, "learning_rate": 4.848444064334679e-07, "loss": 2.3994, "step": 3842 }, { "epoch": 4.548091151228174, "grad_norm": 4.34375, "learning_rate": 4.823032810707118e-07, "loss": 2.2904, "step": 3843 }, { "epoch": 4.549274933412252, "grad_norm": 4.6875, "learning_rate": 4.797686678169655e-07, "loss": 2.3947, "step": 3844 }, { "epoch": 4.5504587155963305, "grad_norm": 5.09375, "learning_rate": 4.772405684064352e-07, "loss": 2.1723, "step": 3845 }, { "epoch": 4.551642497780408, "grad_norm": 4.78125, "learning_rate": 4.7471898456887487e-07, "loss": 2.3621, "step": 3846 }, { "epoch": 4.552826279964487, "grad_norm": 4.71875, "learning_rate": 4.72203918029579e-07, "loss": 2.4187, "step": 3847 }, { "epoch": 4.554010062148564, "grad_norm": 4.9375, "learning_rate": 4.696953705093843e-07, "loss": 2.5223, "step": 3848 }, { "epoch": 4.555193844332643, "grad_norm": 4.15625, "learning_rate": 4.671933437246623e-07, "loss": 2.1406, "step": 3849 }, { "epoch": 4.556377626516721, "grad_norm": 5.03125, "learning_rate": 4.64697839387328e-07, "loss": 2.3719, "step": 3850 }, { "epoch": 4.557561408700799, "grad_norm": 4.90625, "learning_rate": 4.6220885920483014e-07, "loss": 2.1041, "step": 3851 }, { "epoch": 4.558745190884877, "grad_norm": 4.6875, "learning_rate": 4.5972640488015975e-07, "loss": 2.1159, "step": 3852 }, { "epoch": 4.559928973068955, "grad_norm": 5.15625, "learning_rate": 4.5725047811183475e-07, "loss": 2.6296, "step": 3853 }, { "epoch": 4.561112755253033, "grad_norm": 4.59375, "learning_rate": 4.5478108059391123e-07, "loss": 2.2662, "step": 3854 }, { "epoch": 4.562296537437112, "grad_norm": 4.78125, "learning_rate": 4.523182140159765e-07, "loss": 2.2482, "step": 3855 }, { "epoch": 4.56348031962119, "grad_norm": 5.0625, "learning_rate": 4.4986188006315157e-07, "loss": 2.3635, "step": 3856 }, { "epoch": 4.564664101805268, "grad_norm": 4.53125, "learning_rate": 4.474120804160842e-07, "loss": 2.59, "step": 3857 }, { "epoch": 4.5658478839893455, "grad_norm": 4.96875, "learning_rate": 4.4496881675095473e-07, "loss": 2.3764, "step": 3858 }, { "epoch": 4.567031666173424, "grad_norm": 4.3125, "learning_rate": 4.4253209073947034e-07, "loss": 2.4881, "step": 3859 }, { "epoch": 4.568215448357503, "grad_norm": 4.21875, "learning_rate": 4.401019040488652e-07, "loss": 2.4549, "step": 3860 }, { "epoch": 4.56939923054158, "grad_norm": 4.84375, "learning_rate": 4.3767825834189703e-07, "loss": 3.0531, "step": 3861 }, { "epoch": 4.570583012725659, "grad_norm": 4.46875, "learning_rate": 4.3526115527684933e-07, "loss": 2.3551, "step": 3862 }, { "epoch": 4.5717667949097365, "grad_norm": 4.5, "learning_rate": 4.3285059650753135e-07, "loss": 2.5057, "step": 3863 }, { "epoch": 4.572950577093815, "grad_norm": 4.3125, "learning_rate": 4.3044658368327383e-07, "loss": 2.5718, "step": 3864 }, { "epoch": 4.574134359277893, "grad_norm": 4.71875, "learning_rate": 4.280491184489266e-07, "loss": 2.4572, "step": 3865 }, { "epoch": 4.575318141461971, "grad_norm": 4.03125, "learning_rate": 4.2565820244486075e-07, "loss": 2.3074, "step": 3866 }, { "epoch": 4.576501923646049, "grad_norm": 4.75, "learning_rate": 4.232738373069656e-07, "loss": 2.544, "step": 3867 }, { "epoch": 4.5776857058301275, "grad_norm": 4.1875, "learning_rate": 4.2089602466665045e-07, "loss": 2.465, "step": 3868 }, { "epoch": 4.578869488014205, "grad_norm": 4.4375, "learning_rate": 4.1852476615083957e-07, "loss": 2.5956, "step": 3869 }, { "epoch": 4.580053270198284, "grad_norm": 4.8125, "learning_rate": 4.161600633819718e-07, "loss": 2.3866, "step": 3870 }, { "epoch": 4.581237052382361, "grad_norm": 4.46875, "learning_rate": 4.138019179780062e-07, "loss": 2.1589, "step": 3871 }, { "epoch": 4.58242083456644, "grad_norm": 4.59375, "learning_rate": 4.114503315524043e-07, "loss": 2.3342, "step": 3872 }, { "epoch": 4.583604616750518, "grad_norm": 4.46875, "learning_rate": 4.0910530571415143e-07, "loss": 2.3824, "step": 3873 }, { "epoch": 4.584788398934596, "grad_norm": 4.21875, "learning_rate": 4.067668420677373e-07, "loss": 2.2996, "step": 3874 }, { "epoch": 4.585972181118674, "grad_norm": 4.59375, "learning_rate": 4.0443494221316215e-07, "loss": 2.7027, "step": 3875 }, { "epoch": 4.587155963302752, "grad_norm": 4.6875, "learning_rate": 4.021096077459408e-07, "loss": 2.5142, "step": 3876 }, { "epoch": 4.588339745486831, "grad_norm": 5.1875, "learning_rate": 3.9979084025708825e-07, "loss": 2.9205, "step": 3877 }, { "epoch": 4.5895235276709085, "grad_norm": 4.78125, "learning_rate": 3.974786413331311e-07, "loss": 2.368, "step": 3878 }, { "epoch": 4.590707309854986, "grad_norm": 5.125, "learning_rate": 3.951730125560993e-07, "loss": 2.7406, "step": 3879 }, { "epoch": 4.591891092039065, "grad_norm": 4.5625, "learning_rate": 3.928739555035288e-07, "loss": 2.3182, "step": 3880 }, { "epoch": 4.593074874223143, "grad_norm": 4.9375, "learning_rate": 3.9058147174846016e-07, "loss": 2.4209, "step": 3881 }, { "epoch": 4.594258656407221, "grad_norm": 4.5, "learning_rate": 3.8829556285943314e-07, "loss": 2.3156, "step": 3882 }, { "epoch": 4.5954424385912995, "grad_norm": 4.96875, "learning_rate": 3.860162304004933e-07, "loss": 2.5233, "step": 3883 }, { "epoch": 4.596626220775377, "grad_norm": 5.25, "learning_rate": 3.837434759311809e-07, "loss": 2.3196, "step": 3884 }, { "epoch": 4.597810002959456, "grad_norm": 4.78125, "learning_rate": 3.81477301006542e-07, "loss": 2.4575, "step": 3885 }, { "epoch": 4.598993785143533, "grad_norm": 5.40625, "learning_rate": 3.7921770717711414e-07, "loss": 2.4105, "step": 3886 }, { "epoch": 4.600177567327612, "grad_norm": 4.4375, "learning_rate": 3.7696469598893727e-07, "loss": 2.3868, "step": 3887 }, { "epoch": 4.60136134951169, "grad_norm": 5.0625, "learning_rate": 3.747182689835471e-07, "loss": 2.5327, "step": 3888 }, { "epoch": 4.602545131695768, "grad_norm": 4.84375, "learning_rate": 3.7247842769796983e-07, "loss": 2.6079, "step": 3889 }, { "epoch": 4.603728913879846, "grad_norm": 4.78125, "learning_rate": 3.702451736647306e-07, "loss": 2.3231, "step": 3890 }, { "epoch": 4.604912696063924, "grad_norm": 5.09375, "learning_rate": 3.6801850841184375e-07, "loss": 2.8621, "step": 3891 }, { "epoch": 4.606096478248002, "grad_norm": 4.40625, "learning_rate": 3.657984334628173e-07, "loss": 2.3995, "step": 3892 }, { "epoch": 4.607280260432081, "grad_norm": 4.9375, "learning_rate": 3.6358495033665286e-07, "loss": 2.245, "step": 3893 }, { "epoch": 4.608464042616158, "grad_norm": 4.8125, "learning_rate": 3.6137806054783674e-07, "loss": 2.7083, "step": 3894 }, { "epoch": 4.609647824800237, "grad_norm": 4.5, "learning_rate": 3.5917776560634775e-07, "loss": 2.3505, "step": 3895 }, { "epoch": 4.6108316069843145, "grad_norm": 4.34375, "learning_rate": 3.569840670176483e-07, "loss": 2.2618, "step": 3896 }, { "epoch": 4.612015389168393, "grad_norm": 4.28125, "learning_rate": 3.5479696628269225e-07, "loss": 2.2109, "step": 3897 }, { "epoch": 4.613199171352472, "grad_norm": 4.25, "learning_rate": 3.52616464897918e-07, "loss": 2.2142, "step": 3898 }, { "epoch": 4.614382953536549, "grad_norm": 4.375, "learning_rate": 3.504425643552445e-07, "loss": 2.5708, "step": 3899 }, { "epoch": 4.615566735720627, "grad_norm": 4.3125, "learning_rate": 3.4827526614208186e-07, "loss": 2.2557, "step": 3900 }, { "epoch": 4.6167505179047055, "grad_norm": 5.1875, "learning_rate": 3.461145717413161e-07, "loss": 2.5037, "step": 3901 }, { "epoch": 4.617934300088784, "grad_norm": 4.9375, "learning_rate": 3.4396048263131697e-07, "loss": 2.5533, "step": 3902 }, { "epoch": 4.619118082272862, "grad_norm": 5.125, "learning_rate": 3.4181300028593433e-07, "loss": 2.7843, "step": 3903 }, { "epoch": 4.62030186445694, "grad_norm": 4.625, "learning_rate": 3.3967212617449954e-07, "loss": 2.1382, "step": 3904 }, { "epoch": 4.621485646641018, "grad_norm": 5.09375, "learning_rate": 3.3753786176182303e-07, "loss": 2.3401, "step": 3905 }, { "epoch": 4.6226694288250965, "grad_norm": 4.3125, "learning_rate": 3.354102085081878e-07, "loss": 1.983, "step": 3906 }, { "epoch": 4.623853211009174, "grad_norm": 4.8125, "learning_rate": 3.332891678693617e-07, "loss": 2.5739, "step": 3907 }, { "epoch": 4.625036993193253, "grad_norm": 4.5, "learning_rate": 3.31174741296576e-07, "loss": 2.1022, "step": 3908 }, { "epoch": 4.62622077537733, "grad_norm": 4.96875, "learning_rate": 3.29066930236549e-07, "loss": 2.6861, "step": 3909 }, { "epoch": 4.627404557561409, "grad_norm": 4.125, "learning_rate": 3.269657361314671e-07, "loss": 2.2138, "step": 3910 }, { "epoch": 4.628588339745487, "grad_norm": 4.9375, "learning_rate": 3.248711604189869e-07, "loss": 2.5592, "step": 3911 }, { "epoch": 4.629772121929565, "grad_norm": 4.625, "learning_rate": 3.227832045322432e-07, "loss": 2.1991, "step": 3912 }, { "epoch": 4.630955904113643, "grad_norm": 4.9375, "learning_rate": 3.207018698998343e-07, "loss": 2.2566, "step": 3913 }, { "epoch": 4.632139686297721, "grad_norm": 5.25, "learning_rate": 3.186271579458333e-07, "loss": 2.2482, "step": 3914 }, { "epoch": 4.633323468481799, "grad_norm": 4.1875, "learning_rate": 3.1655907008977695e-07, "loss": 2.5244, "step": 3915 }, { "epoch": 4.6345072506658775, "grad_norm": 4.71875, "learning_rate": 3.144976077466766e-07, "loss": 2.2014, "step": 3916 }, { "epoch": 4.635691032849955, "grad_norm": 4.90625, "learning_rate": 3.1244277232700624e-07, "loss": 1.7296, "step": 3917 }, { "epoch": 4.636874815034034, "grad_norm": 4.21875, "learning_rate": 3.103945652367035e-07, "loss": 2.5925, "step": 3918 }, { "epoch": 4.638058597218112, "grad_norm": 5.40625, "learning_rate": 3.083529878771774e-07, "loss": 2.615, "step": 3919 }, { "epoch": 4.63924237940219, "grad_norm": 4.4375, "learning_rate": 3.0631804164529156e-07, "loss": 2.4052, "step": 3920 }, { "epoch": 4.640426161586268, "grad_norm": 4.8125, "learning_rate": 3.042897279333812e-07, "loss": 2.1863, "step": 3921 }, { "epoch": 4.641609943770346, "grad_norm": 4.65625, "learning_rate": 3.022680481292406e-07, "loss": 2.5596, "step": 3922 }, { "epoch": 4.642793725954425, "grad_norm": 5.40625, "learning_rate": 3.002530036161222e-07, "loss": 2.6483, "step": 3923 }, { "epoch": 4.643977508138502, "grad_norm": 4.3125, "learning_rate": 2.982445957727431e-07, "loss": 2.5868, "step": 3924 }, { "epoch": 4.645161290322581, "grad_norm": 4.96875, "learning_rate": 2.9624282597327637e-07, "loss": 2.3585, "step": 3925 }, { "epoch": 4.646345072506659, "grad_norm": 4.9375, "learning_rate": 2.94247695587353e-07, "loss": 2.5614, "step": 3926 }, { "epoch": 4.647528854690737, "grad_norm": 4.3125, "learning_rate": 2.922592059800644e-07, "loss": 2.5157, "step": 3927 }, { "epoch": 4.648712636874815, "grad_norm": 4.6875, "learning_rate": 2.9027735851195336e-07, "loss": 2.6024, "step": 3928 }, { "epoch": 4.649896419058893, "grad_norm": 4.3125, "learning_rate": 2.883021545390241e-07, "loss": 2.7689, "step": 3929 }, { "epoch": 4.651080201242971, "grad_norm": 4.84375, "learning_rate": 2.8633359541273e-07, "loss": 2.4299, "step": 3930 }, { "epoch": 4.65226398342705, "grad_norm": 4.625, "learning_rate": 2.8437168247998245e-07, "loss": 2.4795, "step": 3931 }, { "epoch": 4.653447765611127, "grad_norm": 4.8125, "learning_rate": 2.824164170831389e-07, "loss": 2.6401, "step": 3932 }, { "epoch": 4.654631547795206, "grad_norm": 4.84375, "learning_rate": 2.8046780056001587e-07, "loss": 2.6094, "step": 3933 }, { "epoch": 4.6558153299792835, "grad_norm": 4.4375, "learning_rate": 2.785258342438779e-07, "loss": 2.6112, "step": 3934 }, { "epoch": 4.656999112163362, "grad_norm": 4.625, "learning_rate": 2.765905194634366e-07, "loss": 2.796, "step": 3935 }, { "epoch": 4.65818289434744, "grad_norm": 4.1875, "learning_rate": 2.7466185754285723e-07, "loss": 2.1712, "step": 3936 }, { "epoch": 4.659366676531518, "grad_norm": 4.40625, "learning_rate": 2.7273984980174863e-07, "loss": 2.04, "step": 3937 }, { "epoch": 4.660550458715596, "grad_norm": 4.53125, "learning_rate": 2.7082449755516995e-07, "loss": 2.3565, "step": 3938 }, { "epoch": 4.6617342408996745, "grad_norm": 4.4375, "learning_rate": 2.689158021136251e-07, "loss": 2.0485, "step": 3939 }, { "epoch": 4.662918023083753, "grad_norm": 4.875, "learning_rate": 2.6701376478306396e-07, "loss": 2.5765, "step": 3940 }, { "epoch": 4.664101805267831, "grad_norm": 4.46875, "learning_rate": 2.651183868648821e-07, "loss": 2.2719, "step": 3941 }, { "epoch": 4.665285587451908, "grad_norm": 4.9375, "learning_rate": 2.6322966965591444e-07, "loss": 2.5673, "step": 3942 }, { "epoch": 4.666469369635987, "grad_norm": 4.71875, "learning_rate": 2.613476144484428e-07, "loss": 2.4457, "step": 3943 }, { "epoch": 4.6676531518200655, "grad_norm": 4.84375, "learning_rate": 2.594722225301893e-07, "loss": 2.8268, "step": 3944 }, { "epoch": 4.668836934004143, "grad_norm": 5.0625, "learning_rate": 2.576034951843165e-07, "loss": 2.5593, "step": 3945 }, { "epoch": 4.670020716188222, "grad_norm": 4.59375, "learning_rate": 2.5574143368942817e-07, "loss": 2.2602, "step": 3946 }, { "epoch": 4.671204498372299, "grad_norm": 4.4375, "learning_rate": 2.5388603931956635e-07, "loss": 2.6715, "step": 3947 }, { "epoch": 4.672388280556378, "grad_norm": 4.75, "learning_rate": 2.5203731334421113e-07, "loss": 1.952, "step": 3948 }, { "epoch": 4.673572062740456, "grad_norm": 4.59375, "learning_rate": 2.5019525702828063e-07, "loss": 2.6184, "step": 3949 }, { "epoch": 4.674755844924534, "grad_norm": 4.59375, "learning_rate": 2.483598716321289e-07, "loss": 2.3969, "step": 3950 }, { "epoch": 4.675939627108612, "grad_norm": 4.40625, "learning_rate": 2.4653115841154704e-07, "loss": 2.2455, "step": 3951 }, { "epoch": 4.67712340929269, "grad_norm": 4.125, "learning_rate": 2.447091186177586e-07, "loss": 2.4744, "step": 3952 }, { "epoch": 4.678307191476768, "grad_norm": 4.625, "learning_rate": 2.4289375349742516e-07, "loss": 2.2778, "step": 3953 }, { "epoch": 4.6794909736608465, "grad_norm": 4.96875, "learning_rate": 2.4108506429263547e-07, "loss": 2.3202, "step": 3954 }, { "epoch": 4.680674755844924, "grad_norm": 5.5625, "learning_rate": 2.392830522409162e-07, "loss": 2.8122, "step": 3955 }, { "epoch": 4.681858538029003, "grad_norm": 5.3125, "learning_rate": 2.3748771857522224e-07, "loss": 2.5375, "step": 3956 }, { "epoch": 4.68304232021308, "grad_norm": 4.3125, "learning_rate": 2.3569906452393876e-07, "loss": 2.5858, "step": 3957 }, { "epoch": 4.684226102397159, "grad_norm": 4.53125, "learning_rate": 2.3391709131088457e-07, "loss": 2.3169, "step": 3958 }, { "epoch": 4.685409884581237, "grad_norm": 4.5, "learning_rate": 2.321418001553022e-07, "loss": 2.7503, "step": 3959 }, { "epoch": 4.686593666765315, "grad_norm": 5.125, "learning_rate": 2.303731922718666e-07, "loss": 2.0484, "step": 3960 }, { "epoch": 4.687777448949394, "grad_norm": 4.59375, "learning_rate": 2.2861126887067652e-07, "loss": 2.6768, "step": 3961 }, { "epoch": 4.688961231133471, "grad_norm": 5.25, "learning_rate": 2.2685603115725874e-07, "loss": 2.2209, "step": 3962 }, { "epoch": 4.690145013317549, "grad_norm": 5.125, "learning_rate": 2.2510748033256703e-07, "loss": 2.5519, "step": 3963 }, { "epoch": 4.691328795501628, "grad_norm": 5.15625, "learning_rate": 2.233656175929766e-07, "loss": 2.2993, "step": 3964 }, { "epoch": 4.692512577685706, "grad_norm": 4.4375, "learning_rate": 2.2163044413028855e-07, "loss": 2.1855, "step": 3965 }, { "epoch": 4.693696359869784, "grad_norm": 4.09375, "learning_rate": 2.1990196113172768e-07, "loss": 2.2675, "step": 3966 }, { "epoch": 4.694880142053862, "grad_norm": 4.90625, "learning_rate": 2.1818016977994128e-07, "loss": 2.4831, "step": 3967 }, { "epoch": 4.69606392423794, "grad_norm": 4.84375, "learning_rate": 2.1646507125299587e-07, "loss": 2.6537, "step": 3968 }, { "epoch": 4.697247706422019, "grad_norm": 4.125, "learning_rate": 2.1475666672437945e-07, "loss": 2.2676, "step": 3969 }, { "epoch": 4.698431488606096, "grad_norm": 4.90625, "learning_rate": 2.1305495736300253e-07, "loss": 2.4288, "step": 3970 }, { "epoch": 4.699615270790175, "grad_norm": 5.09375, "learning_rate": 2.1135994433319152e-07, "loss": 2.5295, "step": 3971 }, { "epoch": 4.7007990529742525, "grad_norm": 4.8125, "learning_rate": 2.0967162879469206e-07, "loss": 2.1656, "step": 3972 }, { "epoch": 4.701982835158331, "grad_norm": 4.5, "learning_rate": 2.0799001190266788e-07, "loss": 2.1632, "step": 3973 }, { "epoch": 4.703166617342409, "grad_norm": 4.8125, "learning_rate": 2.0631509480769862e-07, "loss": 2.2333, "step": 3974 }, { "epoch": 4.704350399526487, "grad_norm": 4.96875, "learning_rate": 2.0464687865578204e-07, "loss": 2.6782, "step": 3975 }, { "epoch": 4.705534181710565, "grad_norm": 4.875, "learning_rate": 2.0298536458832617e-07, "loss": 1.9627, "step": 3976 }, { "epoch": 4.7067179638946435, "grad_norm": 5.34375, "learning_rate": 2.013305537421606e-07, "loss": 2.3298, "step": 3977 }, { "epoch": 4.707901746078721, "grad_norm": 5.03125, "learning_rate": 1.996824472495207e-07, "loss": 2.42, "step": 3978 }, { "epoch": 4.7090855282628, "grad_norm": 4.65625, "learning_rate": 1.9804104623806107e-07, "loss": 2.0163, "step": 3979 }, { "epoch": 4.710269310446877, "grad_norm": 4.25, "learning_rate": 1.9640635183084344e-07, "loss": 1.9777, "step": 3980 }, { "epoch": 4.711453092630956, "grad_norm": 4.65625, "learning_rate": 1.947783651463453e-07, "loss": 2.3322, "step": 3981 }, { "epoch": 4.7126368748150345, "grad_norm": 4.25, "learning_rate": 1.9315708729845117e-07, "loss": 2.3211, "step": 3982 }, { "epoch": 4.713820656999112, "grad_norm": 4.96875, "learning_rate": 1.9154251939645595e-07, "loss": 2.7035, "step": 3983 }, { "epoch": 4.715004439183191, "grad_norm": 4.53125, "learning_rate": 1.8993466254506489e-07, "loss": 2.0653, "step": 3984 }, { "epoch": 4.716188221367268, "grad_norm": 5.09375, "learning_rate": 1.8833351784439237e-07, "loss": 2.7036, "step": 3985 }, { "epoch": 4.717372003551347, "grad_norm": 4.9375, "learning_rate": 1.867390863899543e-07, "loss": 2.1492, "step": 3986 }, { "epoch": 4.718555785735425, "grad_norm": 4.59375, "learning_rate": 1.8515136927268251e-07, "loss": 2.0871, "step": 3987 }, { "epoch": 4.719739567919503, "grad_norm": 4.46875, "learning_rate": 1.8357036757890578e-07, "loss": 2.2835, "step": 3988 }, { "epoch": 4.720923350103581, "grad_norm": 5.09375, "learning_rate": 1.8199608239036547e-07, "loss": 2.6206, "step": 3989 }, { "epoch": 4.722107132287659, "grad_norm": 5.09375, "learning_rate": 1.804285147842011e-07, "loss": 2.632, "step": 3990 }, { "epoch": 4.723290914471737, "grad_norm": 4.15625, "learning_rate": 1.7886766583296245e-07, "loss": 2.3769, "step": 3991 }, { "epoch": 4.7244746966558155, "grad_norm": 4.59375, "learning_rate": 1.7731353660459637e-07, "loss": 1.9914, "step": 3992 }, { "epoch": 4.725658478839893, "grad_norm": 4.625, "learning_rate": 1.7576612816245454e-07, "loss": 2.1987, "step": 3993 }, { "epoch": 4.726842261023972, "grad_norm": 4.9375, "learning_rate": 1.7422544156529218e-07, "loss": 2.3797, "step": 3994 }, { "epoch": 4.728026043208049, "grad_norm": 4.3125, "learning_rate": 1.726914778672606e-07, "loss": 2.2927, "step": 3995 }, { "epoch": 4.729209825392128, "grad_norm": 4.40625, "learning_rate": 1.7116423811791794e-07, "loss": 1.9932, "step": 3996 }, { "epoch": 4.730393607576206, "grad_norm": 4.71875, "learning_rate": 1.696437233622139e-07, "loss": 2.2116, "step": 3997 }, { "epoch": 4.731577389760284, "grad_norm": 5.3125, "learning_rate": 1.6812993464050298e-07, "loss": 2.3422, "step": 3998 }, { "epoch": 4.732761171944362, "grad_norm": 4.4375, "learning_rate": 1.6662287298853553e-07, "loss": 2.446, "step": 3999 }, { "epoch": 4.73394495412844, "grad_norm": 4.375, "learning_rate": 1.6512253943745671e-07, "loss": 2.2277, "step": 4000 }, { "epoch": 4.735128736312518, "grad_norm": 4.09375, "learning_rate": 1.6362893501381317e-07, "loss": 2.083, "step": 4001 }, { "epoch": 4.736312518496597, "grad_norm": 4.71875, "learning_rate": 1.6214206073954518e-07, "loss": 2.3726, "step": 4002 }, { "epoch": 4.737496300680675, "grad_norm": 5.09375, "learning_rate": 1.606619176319868e-07, "loss": 2.6775, "step": 4003 }, { "epoch": 4.738680082864753, "grad_norm": 5.28125, "learning_rate": 1.5918850670386677e-07, "loss": 2.6289, "step": 4004 }, { "epoch": 4.739863865048831, "grad_norm": 5.03125, "learning_rate": 1.577218289633098e-07, "loss": 2.6886, "step": 4005 }, { "epoch": 4.741047647232909, "grad_norm": 4.5, "learning_rate": 1.5626188541383203e-07, "loss": 2.2139, "step": 4006 }, { "epoch": 4.742231429416988, "grad_norm": 4.03125, "learning_rate": 1.548086770543422e-07, "loss": 2.3218, "step": 4007 }, { "epoch": 4.743415211601065, "grad_norm": 4.15625, "learning_rate": 1.5336220487914054e-07, "loss": 2.4793, "step": 4008 }, { "epoch": 4.744598993785144, "grad_norm": 4.875, "learning_rate": 1.519224698779198e-07, "loss": 2.5437, "step": 4009 }, { "epoch": 4.7457827759692215, "grad_norm": 4.65625, "learning_rate": 1.504894730357609e-07, "loss": 2.4408, "step": 4010 }, { "epoch": 4.7469665581533, "grad_norm": 4.96875, "learning_rate": 1.4906321533313507e-07, "loss": 2.5632, "step": 4011 }, { "epoch": 4.748150340337378, "grad_norm": 4.5, "learning_rate": 1.4764369774590392e-07, "loss": 2.6173, "step": 4012 }, { "epoch": 4.749334122521456, "grad_norm": 5.125, "learning_rate": 1.4623092124531613e-07, "loss": 2.5786, "step": 4013 }, { "epoch": 4.750517904705534, "grad_norm": 4.5, "learning_rate": 1.4482488679800844e-07, "loss": 2.4917, "step": 4014 }, { "epoch": 4.7517016868896125, "grad_norm": 4.875, "learning_rate": 1.4342559536600576e-07, "loss": 2.7631, "step": 4015 }, { "epoch": 4.75288546907369, "grad_norm": 5.46875, "learning_rate": 1.4203304790671557e-07, "loss": 2.8213, "step": 4016 }, { "epoch": 4.754069251257769, "grad_norm": 4.96875, "learning_rate": 1.4064724537293462e-07, "loss": 2.3611, "step": 4017 }, { "epoch": 4.755253033441846, "grad_norm": 4.6875, "learning_rate": 1.392681887128433e-07, "loss": 2.456, "step": 4018 }, { "epoch": 4.756436815625925, "grad_norm": 4.90625, "learning_rate": 1.378958788700091e-07, "loss": 2.5932, "step": 4019 }, { "epoch": 4.757620597810003, "grad_norm": 4.84375, "learning_rate": 1.3653031678337869e-07, "loss": 2.1828, "step": 4020 }, { "epoch": 4.758804379994081, "grad_norm": 4.78125, "learning_rate": 1.3517150338728468e-07, "loss": 2.4946, "step": 4021 }, { "epoch": 4.759988162178159, "grad_norm": 4.84375, "learning_rate": 1.3381943961144118e-07, "loss": 2.6686, "step": 4022 }, { "epoch": 4.761171944362237, "grad_norm": 4.9375, "learning_rate": 1.3247412638094593e-07, "loss": 2.4309, "step": 4023 }, { "epoch": 4.762355726546316, "grad_norm": 4.4375, "learning_rate": 1.3113556461627486e-07, "loss": 2.631, "step": 4024 }, { "epoch": 4.763539508730394, "grad_norm": 4.375, "learning_rate": 1.2980375523328648e-07, "loss": 2.3769, "step": 4025 }, { "epoch": 4.764723290914472, "grad_norm": 4.78125, "learning_rate": 1.2847869914321965e-07, "loss": 2.5895, "step": 4026 }, { "epoch": 4.76590707309855, "grad_norm": 4.28125, "learning_rate": 1.2716039725269247e-07, "loss": 2.2648, "step": 4027 }, { "epoch": 4.767090855282628, "grad_norm": 4.5625, "learning_rate": 1.25848850463699e-07, "loss": 2.4345, "step": 4028 }, { "epoch": 4.768274637466706, "grad_norm": 4.15625, "learning_rate": 1.245440596736147e-07, "loss": 2.2166, "step": 4029 }, { "epoch": 4.7694584196507845, "grad_norm": 4.6875, "learning_rate": 1.2324602577518997e-07, "loss": 2.2707, "step": 4030 }, { "epoch": 4.770642201834862, "grad_norm": 4.6875, "learning_rate": 1.2195474965655652e-07, "loss": 2.4983, "step": 4031 }, { "epoch": 4.771825984018941, "grad_norm": 4.84375, "learning_rate": 1.2067023220121654e-07, "loss": 2.3548, "step": 4032 }, { "epoch": 4.773009766203018, "grad_norm": 4.71875, "learning_rate": 1.1939247428805035e-07, "loss": 2.3295, "step": 4033 }, { "epoch": 4.774193548387097, "grad_norm": 5.4375, "learning_rate": 1.1812147679131414e-07, "loss": 2.662, "step": 4034 }, { "epoch": 4.775377330571175, "grad_norm": 5.46875, "learning_rate": 1.1685724058063896e-07, "loss": 2.1145, "step": 4035 }, { "epoch": 4.776561112755253, "grad_norm": 4.5, "learning_rate": 1.1559976652102622e-07, "loss": 2.4294, "step": 4036 }, { "epoch": 4.777744894939331, "grad_norm": 4.8125, "learning_rate": 1.1434905547285436e-07, "loss": 2.451, "step": 4037 }, { "epoch": 4.778928677123409, "grad_norm": 4.25, "learning_rate": 1.1310510829187326e-07, "loss": 2.2127, "step": 4038 }, { "epoch": 4.780112459307487, "grad_norm": 5.1875, "learning_rate": 1.1186792582920547e-07, "loss": 2.6265, "step": 4039 }, { "epoch": 4.781296241491566, "grad_norm": 4.6875, "learning_rate": 1.1063750893134273e-07, "loss": 2.4518, "step": 4040 }, { "epoch": 4.782480023675644, "grad_norm": 4.75, "learning_rate": 1.0941385844015051e-07, "loss": 2.6943, "step": 4041 }, { "epoch": 4.783663805859722, "grad_norm": 4.71875, "learning_rate": 1.0819697519286243e-07, "loss": 2.817, "step": 4042 }, { "epoch": 4.7848475880437995, "grad_norm": 4.40625, "learning_rate": 1.0698686002208469e-07, "loss": 2.3126, "step": 4043 }, { "epoch": 4.786031370227878, "grad_norm": 4.40625, "learning_rate": 1.0578351375578943e-07, "loss": 2.462, "step": 4044 }, { "epoch": 4.787215152411957, "grad_norm": 5.25, "learning_rate": 1.0458693721731916e-07, "loss": 2.6405, "step": 4045 }, { "epoch": 4.788398934596034, "grad_norm": 5.0625, "learning_rate": 1.0339713122538342e-07, "loss": 2.5254, "step": 4046 }, { "epoch": 4.789582716780113, "grad_norm": 4.6875, "learning_rate": 1.0221409659406101e-07, "loss": 2.7405, "step": 4047 }, { "epoch": 4.7907664989641905, "grad_norm": 4.90625, "learning_rate": 1.0103783413279777e-07, "loss": 2.3573, "step": 4048 }, { "epoch": 4.791950281148269, "grad_norm": 4.84375, "learning_rate": 9.986834464640328e-08, "loss": 2.6665, "step": 4049 }, { "epoch": 4.793134063332347, "grad_norm": 5.40625, "learning_rate": 9.870562893505519e-08, "loss": 2.6999, "step": 4050 }, { "epoch": 4.794317845516425, "grad_norm": 4.625, "learning_rate": 9.754968779429719e-08, "loss": 2.5253, "step": 4051 }, { "epoch": 4.795501627700503, "grad_norm": 4.40625, "learning_rate": 9.640052201503436e-08, "loss": 2.306, "step": 4052 }, { "epoch": 4.7966854098845815, "grad_norm": 5.15625, "learning_rate": 9.525813238353887e-08, "loss": 2.4779, "step": 4053 }, { "epoch": 4.797869192068659, "grad_norm": 3.953125, "learning_rate": 9.412251968144547e-08, "loss": 2.1265, "step": 4054 }, { "epoch": 4.799052974252738, "grad_norm": 4.875, "learning_rate": 9.299368468575376e-08, "loss": 2.5625, "step": 4055 }, { "epoch": 4.800236756436815, "grad_norm": 4.28125, "learning_rate": 9.187162816882477e-08, "loss": 2.286, "step": 4056 }, { "epoch": 4.801420538620894, "grad_norm": 4.71875, "learning_rate": 9.075635089837886e-08, "loss": 2.2879, "step": 4057 }, { "epoch": 4.802604320804972, "grad_norm": 5.21875, "learning_rate": 8.964785363750228e-08, "loss": 2.6876, "step": 4058 }, { "epoch": 4.80378810298905, "grad_norm": 4.125, "learning_rate": 8.854613714464166e-08, "loss": 2.3731, "step": 4059 }, { "epoch": 4.804971885173128, "grad_norm": 4.625, "learning_rate": 8.74512021736007e-08, "loss": 2.3953, "step": 4060 }, { "epoch": 4.806155667357206, "grad_norm": 5.21875, "learning_rate": 8.636304947354679e-08, "loss": 2.5255, "step": 4061 }, { "epoch": 4.807339449541285, "grad_norm": 4.0625, "learning_rate": 8.52816797890066e-08, "loss": 2.4446, "step": 4062 }, { "epoch": 4.808523231725363, "grad_norm": 5.875, "learning_rate": 8.42070938598638e-08, "loss": 2.7121, "step": 4063 }, { "epoch": 4.80970701390944, "grad_norm": 4.9375, "learning_rate": 8.313929242136031e-08, "loss": 2.2732, "step": 4064 }, { "epoch": 4.810890796093519, "grad_norm": 4.84375, "learning_rate": 8.207827620409836e-08, "loss": 2.2986, "step": 4065 }, { "epoch": 4.812074578277597, "grad_norm": 4.40625, "learning_rate": 8.102404593403612e-08, "loss": 2.2756, "step": 4066 }, { "epoch": 4.813258360461675, "grad_norm": 4.75, "learning_rate": 7.997660233249105e-08, "loss": 2.4703, "step": 4067 }, { "epoch": 4.8144421426457535, "grad_norm": 4.40625, "learning_rate": 7.893594611613208e-08, "loss": 2.2035, "step": 4068 }, { "epoch": 4.815625924829831, "grad_norm": 5.09375, "learning_rate": 7.790207799698857e-08, "loss": 2.7657, "step": 4069 }, { "epoch": 4.81680970701391, "grad_norm": 4.625, "learning_rate": 7.687499868244464e-08, "loss": 2.276, "step": 4070 }, { "epoch": 4.817993489197987, "grad_norm": 4.59375, "learning_rate": 7.585470887523705e-08, "loss": 2.3746, "step": 4071 }, { "epoch": 4.819177271382066, "grad_norm": 4.9375, "learning_rate": 7.484120927346184e-08, "loss": 2.6318, "step": 4072 }, { "epoch": 4.820361053566144, "grad_norm": 4.8125, "learning_rate": 7.383450057056318e-08, "loss": 2.4165, "step": 4073 }, { "epoch": 4.821544835750222, "grad_norm": 4.1875, "learning_rate": 7.283458345534456e-08, "loss": 2.4921, "step": 4074 }, { "epoch": 4.8227286179343, "grad_norm": 4.28125, "learning_rate": 7.184145861195757e-08, "loss": 2.3043, "step": 4075 }, { "epoch": 4.823912400118378, "grad_norm": 4.5625, "learning_rate": 7.08551267199098e-08, "loss": 2.2546, "step": 4076 }, { "epoch": 4.825096182302456, "grad_norm": 5.0625, "learning_rate": 6.987558845405917e-08, "loss": 2.2677, "step": 4077 }, { "epoch": 4.826279964486535, "grad_norm": 4.21875, "learning_rate": 6.890284448461738e-08, "loss": 2.064, "step": 4078 }, { "epoch": 4.827463746670612, "grad_norm": 4.625, "learning_rate": 6.793689547714644e-08, "loss": 2.3217, "step": 4079 }, { "epoch": 4.828647528854691, "grad_norm": 4.78125, "learning_rate": 6.69777420925577e-08, "loss": 2.3077, "step": 4080 }, { "epoch": 4.8298313110387685, "grad_norm": 4.15625, "learning_rate": 6.602538498711397e-08, "loss": 2.2326, "step": 4081 }, { "epoch": 4.831015093222847, "grad_norm": 4.375, "learning_rate": 6.507982481242847e-08, "loss": 2.1803, "step": 4082 }, { "epoch": 4.832198875406926, "grad_norm": 5.59375, "learning_rate": 6.414106221546368e-08, "loss": 2.7034, "step": 4083 }, { "epoch": 4.833382657591003, "grad_norm": 5.03125, "learning_rate": 6.320909783853136e-08, "loss": 2.4346, "step": 4084 }, { "epoch": 4.834566439775081, "grad_norm": 5.0, "learning_rate": 6.22839323192892e-08, "loss": 2.4796, "step": 4085 }, { "epoch": 4.8357502219591595, "grad_norm": 4.46875, "learning_rate": 6.136556629074863e-08, "loss": 2.3844, "step": 4086 }, { "epoch": 4.836934004143238, "grad_norm": 5.15625, "learning_rate": 6.04540003812637e-08, "loss": 2.6362, "step": 4087 }, { "epoch": 4.838117786327316, "grad_norm": 4.5625, "learning_rate": 5.95492352145377e-08, "loss": 2.2129, "step": 4088 }, { "epoch": 4.839301568511394, "grad_norm": 4.90625, "learning_rate": 5.86512714096199e-08, "loss": 2.4203, "step": 4089 }, { "epoch": 4.840485350695472, "grad_norm": 4.96875, "learning_rate": 5.776010958090661e-08, "loss": 2.623, "step": 4090 }, { "epoch": 4.8416691328795505, "grad_norm": 5.0, "learning_rate": 5.687575033814119e-08, "loss": 2.4466, "step": 4091 }, { "epoch": 4.842852915063628, "grad_norm": 4.75, "learning_rate": 5.5998194286410736e-08, "loss": 2.4679, "step": 4092 }, { "epoch": 4.844036697247707, "grad_norm": 5.03125, "learning_rate": 5.512744202614939e-08, "loss": 2.3825, "step": 4093 }, { "epoch": 4.845220479431784, "grad_norm": 4.59375, "learning_rate": 5.426349415313503e-08, "loss": 2.6926, "step": 4094 }, { "epoch": 4.846404261615863, "grad_norm": 4.40625, "learning_rate": 5.340635125848814e-08, "loss": 2.4737, "step": 4095 }, { "epoch": 4.847588043799941, "grad_norm": 4.40625, "learning_rate": 5.255601392867626e-08, "loss": 2.2897, "step": 4096 }, { "epoch": 4.848771825984019, "grad_norm": 5.0, "learning_rate": 5.171248274550955e-08, "loss": 2.1911, "step": 4097 }, { "epoch": 4.849955608168097, "grad_norm": 4.375, "learning_rate": 5.087575828614077e-08, "loss": 2.5319, "step": 4098 }, { "epoch": 4.851139390352175, "grad_norm": 4.78125, "learning_rate": 5.004584112306643e-08, "loss": 2.6249, "step": 4099 }, { "epoch": 4.852323172536253, "grad_norm": 4.65625, "learning_rate": 4.9222731824122295e-08, "loss": 2.6968, "step": 4100 }, { "epoch": 4.853506954720332, "grad_norm": 4.65625, "learning_rate": 4.8406430952491204e-08, "loss": 2.2646, "step": 4101 }, { "epoch": 4.854690736904409, "grad_norm": 4.625, "learning_rate": 4.7596939066691936e-08, "loss": 2.3449, "step": 4102 }, { "epoch": 4.855874519088488, "grad_norm": 4.78125, "learning_rate": 4.679425672059035e-08, "loss": 2.3861, "step": 4103 }, { "epoch": 4.857058301272566, "grad_norm": 5.1875, "learning_rate": 4.5998384463389335e-08, "loss": 2.2868, "step": 4104 }, { "epoch": 4.858242083456644, "grad_norm": 4.78125, "learning_rate": 4.5209322839631085e-08, "loss": 2.3173, "step": 4105 }, { "epoch": 4.859425865640722, "grad_norm": 4.34375, "learning_rate": 4.442707238920263e-08, "loss": 2.0896, "step": 4106 }, { "epoch": 4.8606096478248, "grad_norm": 4.0625, "learning_rate": 4.365163364732694e-08, "loss": 2.3424, "step": 4107 }, { "epoch": 4.861793430008879, "grad_norm": 5.0625, "learning_rate": 4.288300714456739e-08, "loss": 2.2877, "step": 4108 }, { "epoch": 4.862977212192956, "grad_norm": 5.1875, "learning_rate": 4.212119340682663e-08, "loss": 2.3853, "step": 4109 }, { "epoch": 4.864160994377035, "grad_norm": 4.78125, "learning_rate": 4.1366192955345495e-08, "loss": 2.3195, "step": 4110 }, { "epoch": 4.865344776561113, "grad_norm": 5.28125, "learning_rate": 4.061800630670298e-08, "loss": 2.6329, "step": 4111 }, { "epoch": 4.866528558745191, "grad_norm": 4.53125, "learning_rate": 3.987663397281627e-08, "loss": 2.5052, "step": 4112 }, { "epoch": 4.867712340929269, "grad_norm": 5.09375, "learning_rate": 3.9142076460941816e-08, "loss": 2.2065, "step": 4113 }, { "epoch": 4.868896123113347, "grad_norm": 5.28125, "learning_rate": 3.8414334273669804e-08, "loss": 2.8571, "step": 4114 }, { "epoch": 4.870079905297425, "grad_norm": 4.65625, "learning_rate": 3.769340790892973e-08, "loss": 2.4676, "step": 4115 }, { "epoch": 4.871263687481504, "grad_norm": 4.6875, "learning_rate": 3.6979297859986995e-08, "loss": 2.5965, "step": 4116 }, { "epoch": 4.872447469665581, "grad_norm": 5.1875, "learning_rate": 3.6272004615443e-08, "loss": 2.5963, "step": 4117 }, { "epoch": 4.87363125184966, "grad_norm": 4.65625, "learning_rate": 3.5571528659236185e-08, "loss": 2.5293, "step": 4118 }, { "epoch": 4.8748150340337375, "grad_norm": 4.375, "learning_rate": 3.487787047063873e-08, "loss": 2.2661, "step": 4119 }, { "epoch": 4.875998816217816, "grad_norm": 4.9375, "learning_rate": 3.4191030524259874e-08, "loss": 2.7181, "step": 4120 }, { "epoch": 4.877182598401894, "grad_norm": 4.6875, "learning_rate": 3.3511009290042585e-08, "loss": 2.5247, "step": 4121 }, { "epoch": 4.878366380585972, "grad_norm": 4.625, "learning_rate": 3.283780723326358e-08, "loss": 2.3741, "step": 4122 }, { "epoch": 4.87955016277005, "grad_norm": 4.9375, "learning_rate": 3.21714248145355e-08, "loss": 2.5194, "step": 4123 }, { "epoch": 4.8807339449541285, "grad_norm": 5.53125, "learning_rate": 3.1511862489803644e-08, "loss": 2.4685, "step": 4124 }, { "epoch": 4.881917727138207, "grad_norm": 4.75, "learning_rate": 3.085912071034813e-08, "loss": 2.399, "step": 4125 }, { "epoch": 4.883101509322285, "grad_norm": 4.6875, "learning_rate": 3.021319992278282e-08, "loss": 2.6626, "step": 4126 }, { "epoch": 4.884285291506362, "grad_norm": 4.34375, "learning_rate": 2.9574100569051967e-08, "loss": 1.9464, "step": 4127 }, { "epoch": 4.885469073690441, "grad_norm": 4.40625, "learning_rate": 2.894182308643467e-08, "loss": 2.0755, "step": 4128 }, { "epoch": 4.8866528558745195, "grad_norm": 4.53125, "learning_rate": 2.8316367907542663e-08, "loss": 2.2576, "step": 4129 }, { "epoch": 4.887836638058597, "grad_norm": 5.21875, "learning_rate": 2.7697735460316954e-08, "loss": 2.464, "step": 4130 }, { "epoch": 4.889020420242676, "grad_norm": 4.3125, "learning_rate": 2.7085926168034517e-08, "loss": 2.2335, "step": 4131 }, { "epoch": 4.890204202426753, "grad_norm": 3.9375, "learning_rate": 2.6480940449301607e-08, "loss": 2.4245, "step": 4132 }, { "epoch": 4.891387984610832, "grad_norm": 4.4375, "learning_rate": 2.588277871805378e-08, "loss": 2.5716, "step": 4133 }, { "epoch": 4.89257176679491, "grad_norm": 4.15625, "learning_rate": 2.529144138356254e-08, "loss": 2.3707, "step": 4134 }, { "epoch": 4.893755548978988, "grad_norm": 5.09375, "learning_rate": 2.470692885042536e-08, "loss": 2.4816, "step": 4135 }, { "epoch": 4.894939331163066, "grad_norm": 4.59375, "learning_rate": 2.412924151857121e-08, "loss": 2.1755, "step": 4136 }, { "epoch": 4.896123113347144, "grad_norm": 4.96875, "learning_rate": 2.35583797832617e-08, "loss": 2.1764, "step": 4137 }, { "epoch": 4.897306895531222, "grad_norm": 4.53125, "learning_rate": 2.29943440350866e-08, "loss": 2.5664, "step": 4138 }, { "epoch": 4.898490677715301, "grad_norm": 4.53125, "learning_rate": 2.2437134659962777e-08, "loss": 2.6596, "step": 4139 }, { "epoch": 4.899674459899378, "grad_norm": 4.8125, "learning_rate": 2.188675203914192e-08, "loss": 2.3427, "step": 4140 }, { "epoch": 4.900858242083457, "grad_norm": 4.75, "learning_rate": 2.1343196549198363e-08, "loss": 2.5028, "step": 4141 }, { "epoch": 4.902042024267534, "grad_norm": 4.40625, "learning_rate": 2.0806468562041272e-08, "loss": 2.2326, "step": 4142 }, { "epoch": 4.903225806451613, "grad_norm": 4.53125, "learning_rate": 2.0276568444904666e-08, "loss": 2.398, "step": 4143 }, { "epoch": 4.904409588635691, "grad_norm": 4.84375, "learning_rate": 1.975349656035075e-08, "loss": 2.4517, "step": 4144 }, { "epoch": 4.905593370819769, "grad_norm": 4.28125, "learning_rate": 1.9237253266272126e-08, "loss": 2.3444, "step": 4145 }, { "epoch": 4.906777153003848, "grad_norm": 4.5, "learning_rate": 1.8727838915888476e-08, "loss": 2.1252, "step": 4146 }, { "epoch": 4.907960935187925, "grad_norm": 4.84375, "learning_rate": 1.822525385774543e-08, "loss": 2.4378, "step": 4147 }, { "epoch": 4.909144717372003, "grad_norm": 5.3125, "learning_rate": 1.7729498435716808e-08, "loss": 2.3854, "step": 4148 }, { "epoch": 4.910328499556082, "grad_norm": 4.6875, "learning_rate": 1.7240572989003502e-08, "loss": 2.3599, "step": 4149 }, { "epoch": 4.91151228174016, "grad_norm": 5.1875, "learning_rate": 1.6758477852135692e-08, "loss": 2.6746, "step": 4150 }, { "epoch": 4.912696063924238, "grad_norm": 4.53125, "learning_rate": 1.6283213354965077e-08, "loss": 2.3497, "step": 4151 }, { "epoch": 4.913879846108316, "grad_norm": 5.46875, "learning_rate": 1.5814779822674876e-08, "loss": 3.0008, "step": 4152 }, { "epoch": 4.915063628292394, "grad_norm": 4.875, "learning_rate": 1.535317757577093e-08, "loss": 2.5705, "step": 4153 }, { "epoch": 4.916247410476473, "grad_norm": 4.28125, "learning_rate": 1.4898406930087262e-08, "loss": 2.3327, "step": 4154 }, { "epoch": 4.91743119266055, "grad_norm": 4.40625, "learning_rate": 1.4450468196781641e-08, "loss": 2.4283, "step": 4155 }, { "epoch": 4.918614974844629, "grad_norm": 5.0, "learning_rate": 1.4009361682340018e-08, "loss": 2.8192, "step": 4156 }, { "epoch": 4.9197987570287065, "grad_norm": 4.875, "learning_rate": 1.3575087688570965e-08, "loss": 2.4475, "step": 4157 }, { "epoch": 4.920982539212785, "grad_norm": 4.4375, "learning_rate": 1.3147646512610135e-08, "loss": 2.3864, "step": 4158 }, { "epoch": 4.922166321396863, "grad_norm": 4.25, "learning_rate": 1.2727038446916918e-08, "loss": 2.0274, "step": 4159 }, { "epoch": 4.923350103580941, "grad_norm": 4.34375, "learning_rate": 1.2313263779275553e-08, "loss": 1.9728, "step": 4160 }, { "epoch": 4.924533885765019, "grad_norm": 4.34375, "learning_rate": 1.1906322792795132e-08, "loss": 1.991, "step": 4161 }, { "epoch": 4.9257176679490975, "grad_norm": 4.625, "learning_rate": 1.15062157659096e-08, "loss": 2.3236, "step": 4162 }, { "epoch": 4.926901450133175, "grad_norm": 4.34375, "learning_rate": 1.111294297237664e-08, "loss": 2.2241, "step": 4163 }, { "epoch": 4.928085232317254, "grad_norm": 4.96875, "learning_rate": 1.0726504681275452e-08, "loss": 2.7443, "step": 4164 }, { "epoch": 4.929269014501331, "grad_norm": 4.03125, "learning_rate": 1.0346901157014532e-08, "loss": 2.3371, "step": 4165 }, { "epoch": 4.93045279668541, "grad_norm": 4.28125, "learning_rate": 9.974132659319457e-09, "loss": 2.3553, "step": 4166 }, { "epoch": 4.9316365788694885, "grad_norm": 4.8125, "learning_rate": 9.608199443243981e-09, "loss": 2.3193, "step": 4167 }, { "epoch": 4.932820361053566, "grad_norm": 4.5625, "learning_rate": 9.249101759164492e-09, "loss": 2.324, "step": 4168 }, { "epoch": 4.934004143237644, "grad_norm": 4.46875, "learning_rate": 8.896839852777783e-09, "loss": 2.2032, "step": 4169 }, { "epoch": 4.935187925421722, "grad_norm": 4.71875, "learning_rate": 8.551413965105505e-09, "loss": 2.4461, "step": 4170 }, { "epoch": 4.936371707605801, "grad_norm": 5.0, "learning_rate": 8.212824332491931e-09, "loss": 2.0969, "step": 4171 }, { "epoch": 4.937555489789879, "grad_norm": 4.75, "learning_rate": 7.881071186602863e-09, "loss": 2.3606, "step": 4172 }, { "epoch": 4.938739271973957, "grad_norm": 4.75, "learning_rate": 7.556154754428946e-09, "loss": 2.5134, "step": 4173 }, { "epoch": 4.939923054158035, "grad_norm": 4.1875, "learning_rate": 7.238075258281241e-09, "loss": 1.8232, "step": 4174 }, { "epoch": 4.941106836342113, "grad_norm": 4.84375, "learning_rate": 6.926832915791215e-09, "loss": 2.6035, "step": 4175 }, { "epoch": 4.942290618526191, "grad_norm": 5.28125, "learning_rate": 6.622427939916298e-09, "loss": 2.5939, "step": 4176 }, { "epoch": 4.94347440071027, "grad_norm": 4.8125, "learning_rate": 6.32486053893211e-09, "loss": 2.4181, "step": 4177 }, { "epoch": 4.944658182894347, "grad_norm": 4.6875, "learning_rate": 6.034130916439118e-09, "loss": 2.4462, "step": 4178 }, { "epoch": 4.945841965078426, "grad_norm": 4.5625, "learning_rate": 5.750239271355984e-09, "loss": 2.5009, "step": 4179 }, { "epoch": 4.947025747262503, "grad_norm": 4.84375, "learning_rate": 5.473185797923997e-09, "loss": 2.3037, "step": 4180 }, { "epoch": 4.948209529446582, "grad_norm": 5.03125, "learning_rate": 5.202970685708186e-09, "loss": 2.3294, "step": 4181 }, { "epoch": 4.94939331163066, "grad_norm": 4.90625, "learning_rate": 4.939594119590663e-09, "loss": 2.7777, "step": 4182 }, { "epoch": 4.950577093814738, "grad_norm": 4.9375, "learning_rate": 4.6830562797783865e-09, "loss": 2.2669, "step": 4183 }, { "epoch": 4.951760875998816, "grad_norm": 4.71875, "learning_rate": 4.433357341795397e-09, "loss": 2.5013, "step": 4184 }, { "epoch": 4.952944658182894, "grad_norm": 5.4375, "learning_rate": 4.190497476488364e-09, "loss": 2.1376, "step": 4185 }, { "epoch": 4.954128440366972, "grad_norm": 4.53125, "learning_rate": 3.95447685002659e-09, "loss": 2.6578, "step": 4186 }, { "epoch": 4.955312222551051, "grad_norm": 4.21875, "learning_rate": 3.725295623896452e-09, "loss": 2.4628, "step": 4187 }, { "epoch": 4.956496004735129, "grad_norm": 5.125, "learning_rate": 3.502953954905852e-09, "loss": 2.4765, "step": 4188 }, { "epoch": 4.957679786919207, "grad_norm": 5.15625, "learning_rate": 3.2874519951830995e-09, "loss": 2.464, "step": 4189 }, { "epoch": 4.9588635691032845, "grad_norm": 5.15625, "learning_rate": 3.078789892179135e-09, "loss": 2.4101, "step": 4190 }, { "epoch": 4.960047351287363, "grad_norm": 4.84375, "learning_rate": 2.876967788659757e-09, "loss": 2.1874, "step": 4191 }, { "epoch": 4.961231133471442, "grad_norm": 4.3125, "learning_rate": 2.681985822716726e-09, "loss": 2.2637, "step": 4192 }, { "epoch": 4.962414915655519, "grad_norm": 5.4375, "learning_rate": 2.4938441277566615e-09, "loss": 2.4731, "step": 4193 }, { "epoch": 4.963598697839598, "grad_norm": 4.25, "learning_rate": 2.3125428325088127e-09, "loss": 2.2927, "step": 4194 }, { "epoch": 4.9647824800236755, "grad_norm": 5.0, "learning_rate": 2.138082061021729e-09, "loss": 2.4161, "step": 4195 }, { "epoch": 4.965966262207754, "grad_norm": 4.75, "learning_rate": 1.970461932662149e-09, "loss": 2.2863, "step": 4196 }, { "epoch": 4.967150044391832, "grad_norm": 4.90625, "learning_rate": 1.8096825621194414e-09, "loss": 2.6376, "step": 4197 }, { "epoch": 4.96833382657591, "grad_norm": 4.9375, "learning_rate": 1.655744059398945e-09, "loss": 2.5403, "step": 4198 }, { "epoch": 4.969517608759988, "grad_norm": 4.34375, "learning_rate": 1.5086465298275178e-09, "loss": 2.4763, "step": 4199 }, { "epoch": 4.9707013909440665, "grad_norm": 4.6875, "learning_rate": 1.3683900740513178e-09, "loss": 2.5293, "step": 4200 }, { "epoch": 4.971885173128144, "grad_norm": 4.5, "learning_rate": 1.2349747880335828e-09, "loss": 2.0236, "step": 4201 }, { "epoch": 4.973068955312223, "grad_norm": 4.84375, "learning_rate": 1.1084007630612903e-09, "loss": 2.4991, "step": 4202 }, { "epoch": 4.9742527374963, "grad_norm": 5.21875, "learning_rate": 9.886680857362773e-10, "loss": 2.6411, "step": 4203 }, { "epoch": 4.975436519680379, "grad_norm": 4.84375, "learning_rate": 8.757768379796805e-10, "loss": 2.3682, "step": 4204 }, { "epoch": 4.976620301864457, "grad_norm": 4.59375, "learning_rate": 7.697270970352666e-10, "loss": 2.4433, "step": 4205 }, { "epoch": 4.977804084048535, "grad_norm": 4.71875, "learning_rate": 6.705189354616615e-10, "loss": 2.5157, "step": 4206 }, { "epoch": 4.978987866232613, "grad_norm": 4.25, "learning_rate": 5.781524211379008e-10, "loss": 2.1076, "step": 4207 }, { "epoch": 4.980171648416691, "grad_norm": 4.15625, "learning_rate": 4.926276172645405e-10, "loss": 2.4651, "step": 4208 }, { "epoch": 4.98135543060077, "grad_norm": 5.03125, "learning_rate": 4.139445823558852e-10, "loss": 2.6818, "step": 4209 }, { "epoch": 4.982539212784848, "grad_norm": 4.78125, "learning_rate": 3.4210337024886965e-10, "loss": 2.5765, "step": 4210 }, { "epoch": 4.983722994968925, "grad_norm": 4.3125, "learning_rate": 2.7710403009750807e-10, "loss": 1.857, "step": 4211 }, { "epoch": 4.984906777153004, "grad_norm": 5.21875, "learning_rate": 2.1894660637622467e-10, "loss": 2.4707, "step": 4212 }, { "epoch": 4.986090559337082, "grad_norm": 4.5, "learning_rate": 1.676311388754126e-10, "loss": 2.3514, "step": 4213 }, { "epoch": 4.98727434152116, "grad_norm": 4.375, "learning_rate": 1.2315766270698527e-10, "loss": 2.1707, "step": 4214 }, { "epoch": 4.988458123705239, "grad_norm": 4.625, "learning_rate": 8.552620829882508e-11, "loss": 2.518, "step": 4215 }, { "epoch": 4.989641905889316, "grad_norm": 4.5625, "learning_rate": 5.473680140033466e-11, "loss": 2.6213, "step": 4216 }, { "epoch": 4.990825688073395, "grad_norm": 5.25, "learning_rate": 3.078946307577546e-11, "loss": 2.9001, "step": 4217 }, { "epoch": 4.992009470257472, "grad_norm": 5.28125, "learning_rate": 1.3684209713149543e-11, "loss": 2.4019, "step": 4218 }, { "epoch": 4.993193252441551, "grad_norm": 4.5625, "learning_rate": 3.421053013097364e-12, "loss": 2.5534, "step": 4219 }, { "epoch": 4.994377034625629, "grad_norm": 4.625, "learning_rate": 0.0, "loss": 2.3164, "step": 4220 } ], "logging_steps": 1.0, "max_steps": 4220, "num_input_tokens_seen": 0, "num_train_epochs": 5, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 6.550625427846758e+17, "train_batch_size": 12, "trial_name": null, "trial_params": null }