{ "best_metric": null, "best_model_checkpoint": null, "epoch": 2.0, "eval_steps": 500, "global_step": 2576, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0007763975155279503, "grad_norm": 1.206859980450724, "learning_rate": 9.999996281669553e-06, "loss": 0.2843, "step": 1 }, { "epoch": 0.0015527950310559005, "grad_norm": 1.292421926454779, "learning_rate": 9.99998512668374e-06, "loss": 0.3119, "step": 2 }, { "epoch": 0.002329192546583851, "grad_norm": 0.8896832351603652, "learning_rate": 9.999966535059155e-06, "loss": 0.2717, "step": 3 }, { "epoch": 0.003105590062111801, "grad_norm": 1.1583501521389068, "learning_rate": 9.999940506823446e-06, "loss": 0.2852, "step": 4 }, { "epoch": 0.0038819875776397515, "grad_norm": 0.7346881638183943, "learning_rate": 9.99990704201533e-06, "loss": 0.2517, "step": 5 }, { "epoch": 0.004658385093167702, "grad_norm": 0.6963754543114964, "learning_rate": 9.999866140684576e-06, "loss": 0.2601, "step": 6 }, { "epoch": 0.005434782608695652, "grad_norm": 0.843357718451674, "learning_rate": 9.999817802892021e-06, "loss": 0.2898, "step": 7 }, { "epoch": 0.006211180124223602, "grad_norm": 1.9510278816506652, "learning_rate": 9.999762028709558e-06, "loss": 0.9315, "step": 8 }, { "epoch": 0.006987577639751553, "grad_norm": 1.200291927876152, "learning_rate": 9.999698818220142e-06, "loss": 0.284, "step": 9 }, { "epoch": 0.007763975155279503, "grad_norm": 1.2960084683060504, "learning_rate": 9.999628171517788e-06, "loss": 0.282, "step": 10 }, { "epoch": 0.008540372670807454, "grad_norm": 0.8822430723783882, "learning_rate": 9.999550088707572e-06, "loss": 0.2123, "step": 11 }, { "epoch": 0.009316770186335404, "grad_norm": 0.7688657923244316, "learning_rate": 9.999464569905628e-06, "loss": 0.2007, "step": 12 }, { "epoch": 0.010093167701863354, "grad_norm": 0.9323325746827641, "learning_rate": 9.99937161523915e-06, "loss": 0.2819, "step": 13 }, { "epoch": 0.010869565217391304, "grad_norm": 0.7536959869429912, "learning_rate": 9.999271224846397e-06, "loss": 0.2335, "step": 14 }, { "epoch": 0.011645962732919254, "grad_norm": 0.8994931453802645, "learning_rate": 9.999163398876676e-06, "loss": 0.2211, "step": 15 }, { "epoch": 0.012422360248447204, "grad_norm": 0.8597322167830309, "learning_rate": 9.999048137490364e-06, "loss": 0.2655, "step": 16 }, { "epoch": 0.013198757763975156, "grad_norm": 0.878290522902155, "learning_rate": 9.998925440858894e-06, "loss": 0.2346, "step": 17 }, { "epoch": 0.013975155279503106, "grad_norm": 0.6893850427012543, "learning_rate": 9.998795309164754e-06, "loss": 0.2575, "step": 18 }, { "epoch": 0.014751552795031056, "grad_norm": 0.8848046587560067, "learning_rate": 9.998657742601495e-06, "loss": 0.2273, "step": 19 }, { "epoch": 0.015527950310559006, "grad_norm": 0.5938793788867393, "learning_rate": 9.99851274137372e-06, "loss": 0.2203, "step": 20 }, { "epoch": 0.016304347826086956, "grad_norm": 0.7774430939640068, "learning_rate": 9.9983603056971e-06, "loss": 0.2339, "step": 21 }, { "epoch": 0.017080745341614908, "grad_norm": 0.5452738860641565, "learning_rate": 9.998200435798355e-06, "loss": 0.2306, "step": 22 }, { "epoch": 0.017857142857142856, "grad_norm": 0.7352838892262695, "learning_rate": 9.998033131915266e-06, "loss": 0.33, "step": 23 }, { "epoch": 0.018633540372670808, "grad_norm": 0.9770502603420645, "learning_rate": 9.997858394296666e-06, "loss": 0.2988, "step": 24 }, { "epoch": 0.019409937888198756, "grad_norm": 0.8756580939739894, "learning_rate": 9.99767622320245e-06, "loss": 0.295, "step": 25 }, { "epoch": 0.020186335403726708, "grad_norm": 1.0332871704752271, "learning_rate": 9.997486618903566e-06, "loss": 0.2862, "step": 26 }, { "epoch": 0.02096273291925466, "grad_norm": 0.5573708554903846, "learning_rate": 9.99728958168202e-06, "loss": 0.2375, "step": 27 }, { "epoch": 0.021739130434782608, "grad_norm": 0.6498402799121863, "learning_rate": 9.99708511183087e-06, "loss": 0.2867, "step": 28 }, { "epoch": 0.02251552795031056, "grad_norm": 0.7092153015429724, "learning_rate": 9.996873209654234e-06, "loss": 0.2568, "step": 29 }, { "epoch": 0.023291925465838508, "grad_norm": 0.991926550991633, "learning_rate": 9.996653875467278e-06, "loss": 0.3604, "step": 30 }, { "epoch": 0.02406832298136646, "grad_norm": 0.8585078516367767, "learning_rate": 9.996427109596224e-06, "loss": 0.2447, "step": 31 }, { "epoch": 0.024844720496894408, "grad_norm": 0.7771406101066608, "learning_rate": 9.99619291237835e-06, "loss": 0.2828, "step": 32 }, { "epoch": 0.02562111801242236, "grad_norm": 0.9634554773438586, "learning_rate": 9.995951284161986e-06, "loss": 0.3146, "step": 33 }, { "epoch": 0.026397515527950312, "grad_norm": 0.7307166899347248, "learning_rate": 9.99570222530651e-06, "loss": 0.2435, "step": 34 }, { "epoch": 0.02717391304347826, "grad_norm": 1.0240760130367368, "learning_rate": 9.99544573618236e-06, "loss": 0.2884, "step": 35 }, { "epoch": 0.027950310559006212, "grad_norm": 1.0050510515719273, "learning_rate": 9.995181817171016e-06, "loss": 0.2451, "step": 36 }, { "epoch": 0.02872670807453416, "grad_norm": 0.724568953751987, "learning_rate": 9.994910468665014e-06, "loss": 0.2921, "step": 37 }, { "epoch": 0.029503105590062112, "grad_norm": 0.6587697452976861, "learning_rate": 9.994631691067941e-06, "loss": 0.2775, "step": 38 }, { "epoch": 0.030279503105590064, "grad_norm": 0.5970710472443918, "learning_rate": 9.994345484794433e-06, "loss": 0.2096, "step": 39 }, { "epoch": 0.031055900621118012, "grad_norm": 0.7163769456640317, "learning_rate": 9.994051850270172e-06, "loss": 0.2855, "step": 40 }, { "epoch": 0.03183229813664596, "grad_norm": 0.69746092270152, "learning_rate": 9.993750787931888e-06, "loss": 0.2506, "step": 41 }, { "epoch": 0.03260869565217391, "grad_norm": 0.5896052019215333, "learning_rate": 9.993442298227365e-06, "loss": 0.2155, "step": 42 }, { "epoch": 0.033385093167701864, "grad_norm": 0.958014963997802, "learning_rate": 9.993126381615428e-06, "loss": 0.2545, "step": 43 }, { "epoch": 0.034161490683229816, "grad_norm": 0.7277569776014874, "learning_rate": 9.992803038565947e-06, "loss": 0.2619, "step": 44 }, { "epoch": 0.03493788819875776, "grad_norm": 0.8519310625655102, "learning_rate": 9.992472269559844e-06, "loss": 0.265, "step": 45 }, { "epoch": 0.03571428571428571, "grad_norm": 0.8728399141339458, "learning_rate": 9.992134075089085e-06, "loss": 0.3042, "step": 46 }, { "epoch": 0.036490683229813664, "grad_norm": 0.67483647212221, "learning_rate": 9.99178845565667e-06, "loss": 0.2427, "step": 47 }, { "epoch": 0.037267080745341616, "grad_norm": 0.7360691801256862, "learning_rate": 9.991435411776654e-06, "loss": 0.3287, "step": 48 }, { "epoch": 0.03804347826086957, "grad_norm": 0.7026126473677711, "learning_rate": 9.991074943974132e-06, "loss": 0.2552, "step": 49 }, { "epoch": 0.03881987577639751, "grad_norm": 0.7041333908345635, "learning_rate": 9.990707052785236e-06, "loss": 0.2928, "step": 50 }, { "epoch": 0.039596273291925464, "grad_norm": 0.685360089579432, "learning_rate": 9.990331738757147e-06, "loss": 0.266, "step": 51 }, { "epoch": 0.040372670807453416, "grad_norm": 0.6093560430331627, "learning_rate": 9.989949002448076e-06, "loss": 0.2321, "step": 52 }, { "epoch": 0.04114906832298137, "grad_norm": 0.9568238207129881, "learning_rate": 9.989558844427283e-06, "loss": 0.3754, "step": 53 }, { "epoch": 0.04192546583850932, "grad_norm": 0.7005122862129407, "learning_rate": 9.98916126527506e-06, "loss": 0.219, "step": 54 }, { "epoch": 0.042701863354037264, "grad_norm": 1.2762041591575515, "learning_rate": 9.988756265582742e-06, "loss": 0.786, "step": 55 }, { "epoch": 0.043478260869565216, "grad_norm": 0.7995682467588096, "learning_rate": 9.988343845952697e-06, "loss": 0.2872, "step": 56 }, { "epoch": 0.04425465838509317, "grad_norm": 1.4213625748751193, "learning_rate": 9.987924006998329e-06, "loss": 0.2752, "step": 57 }, { "epoch": 0.04503105590062112, "grad_norm": 0.7077314382684141, "learning_rate": 9.98749674934408e-06, "loss": 0.2551, "step": 58 }, { "epoch": 0.04580745341614907, "grad_norm": 0.7289024388317277, "learning_rate": 9.98706207362542e-06, "loss": 0.2196, "step": 59 }, { "epoch": 0.046583850931677016, "grad_norm": 0.8835718960244464, "learning_rate": 9.98661998048886e-06, "loss": 0.1916, "step": 60 }, { "epoch": 0.04736024844720497, "grad_norm": 0.6048753777511681, "learning_rate": 9.98617047059194e-06, "loss": 0.2287, "step": 61 }, { "epoch": 0.04813664596273292, "grad_norm": 0.6789003516480578, "learning_rate": 9.985713544603229e-06, "loss": 0.1615, "step": 62 }, { "epoch": 0.04891304347826087, "grad_norm": 0.5856621619960953, "learning_rate": 9.985249203202324e-06, "loss": 0.2516, "step": 63 }, { "epoch": 0.049689440993788817, "grad_norm": 0.5888394459427023, "learning_rate": 9.984777447079861e-06, "loss": 0.2536, "step": 64 }, { "epoch": 0.05046583850931677, "grad_norm": 0.8459170368482436, "learning_rate": 9.984298276937496e-06, "loss": 0.2896, "step": 65 }, { "epoch": 0.05124223602484472, "grad_norm": 0.6772225316751757, "learning_rate": 9.983811693487913e-06, "loss": 0.2614, "step": 66 }, { "epoch": 0.05201863354037267, "grad_norm": 0.7071411954093008, "learning_rate": 9.983317697454824e-06, "loss": 0.2315, "step": 67 }, { "epoch": 0.052795031055900624, "grad_norm": 0.8334080186363372, "learning_rate": 9.982816289572966e-06, "loss": 0.286, "step": 68 }, { "epoch": 0.05357142857142857, "grad_norm": 0.6824647482319222, "learning_rate": 9.982307470588097e-06, "loss": 0.2131, "step": 69 }, { "epoch": 0.05434782608695652, "grad_norm": 0.8149734621411512, "learning_rate": 9.981791241257001e-06, "loss": 0.2418, "step": 70 }, { "epoch": 0.05512422360248447, "grad_norm": 0.75836007149244, "learning_rate": 9.981267602347482e-06, "loss": 0.2607, "step": 71 }, { "epoch": 0.055900621118012424, "grad_norm": 0.7972144674273145, "learning_rate": 9.980736554638367e-06, "loss": 0.2864, "step": 72 }, { "epoch": 0.056677018633540376, "grad_norm": 0.6715545096298097, "learning_rate": 9.980198098919499e-06, "loss": 0.2233, "step": 73 }, { "epoch": 0.05745341614906832, "grad_norm": 0.8080273094791947, "learning_rate": 9.979652235991738e-06, "loss": 0.2314, "step": 74 }, { "epoch": 0.05822981366459627, "grad_norm": 0.6956331862132645, "learning_rate": 9.97909896666697e-06, "loss": 0.2814, "step": 75 }, { "epoch": 0.059006211180124224, "grad_norm": 0.7856238403998391, "learning_rate": 9.978538291768083e-06, "loss": 0.2555, "step": 76 }, { "epoch": 0.059782608695652176, "grad_norm": 0.6371467504046289, "learning_rate": 9.97797021212899e-06, "loss": 0.2124, "step": 77 }, { "epoch": 0.06055900621118013, "grad_norm": 0.7121395368569311, "learning_rate": 9.977394728594614e-06, "loss": 0.2549, "step": 78 }, { "epoch": 0.06133540372670807, "grad_norm": 0.69975287787416, "learning_rate": 9.976811842020893e-06, "loss": 0.2404, "step": 79 }, { "epoch": 0.062111801242236024, "grad_norm": 0.7804515556875262, "learning_rate": 9.976221553274767e-06, "loss": 0.2454, "step": 80 }, { "epoch": 0.06288819875776397, "grad_norm": 0.5791580613098702, "learning_rate": 9.975623863234197e-06, "loss": 0.2467, "step": 81 }, { "epoch": 0.06366459627329192, "grad_norm": 0.6652146747320993, "learning_rate": 9.975018772788144e-06, "loss": 0.2181, "step": 82 }, { "epoch": 0.06444099378881987, "grad_norm": 0.8395788853733256, "learning_rate": 9.974406282836577e-06, "loss": 0.2961, "step": 83 }, { "epoch": 0.06521739130434782, "grad_norm": 0.6190961616815417, "learning_rate": 9.973786394290475e-06, "loss": 0.3059, "step": 84 }, { "epoch": 0.06599378881987578, "grad_norm": 0.759186058439185, "learning_rate": 9.973159108071817e-06, "loss": 0.3126, "step": 85 }, { "epoch": 0.06677018633540373, "grad_norm": 0.8061202202576687, "learning_rate": 9.972524425113585e-06, "loss": 0.2298, "step": 86 }, { "epoch": 0.06754658385093168, "grad_norm": 0.7177412856966549, "learning_rate": 9.971882346359767e-06, "loss": 0.2237, "step": 87 }, { "epoch": 0.06832298136645963, "grad_norm": 0.8259704516473783, "learning_rate": 9.971232872765344e-06, "loss": 0.3401, "step": 88 }, { "epoch": 0.06909937888198758, "grad_norm": 0.9478481926088907, "learning_rate": 9.970576005296297e-06, "loss": 0.2732, "step": 89 }, { "epoch": 0.06987577639751552, "grad_norm": 0.704321847081857, "learning_rate": 9.969911744929613e-06, "loss": 0.2385, "step": 90 }, { "epoch": 0.07065217391304347, "grad_norm": 0.6935102156074717, "learning_rate": 9.96924009265326e-06, "loss": 0.2301, "step": 91 }, { "epoch": 0.07142857142857142, "grad_norm": 1.0802362963447119, "learning_rate": 9.968561049466214e-06, "loss": 0.328, "step": 92 }, { "epoch": 0.07220496894409938, "grad_norm": 0.6775244014925002, "learning_rate": 9.967874616378434e-06, "loss": 0.2396, "step": 93 }, { "epoch": 0.07298136645962733, "grad_norm": 0.75388510467654, "learning_rate": 9.967180794410878e-06, "loss": 0.2436, "step": 94 }, { "epoch": 0.07375776397515528, "grad_norm": 0.6551533335508573, "learning_rate": 9.966479584595485e-06, "loss": 0.239, "step": 95 }, { "epoch": 0.07453416149068323, "grad_norm": 0.8482335772668737, "learning_rate": 9.96577098797519e-06, "loss": 0.2984, "step": 96 }, { "epoch": 0.07531055900621118, "grad_norm": 0.6663462496956936, "learning_rate": 9.96505500560391e-06, "loss": 0.2888, "step": 97 }, { "epoch": 0.07608695652173914, "grad_norm": 1.4796813212742437, "learning_rate": 9.96433163854655e-06, "loss": 0.915, "step": 98 }, { "epoch": 0.07686335403726709, "grad_norm": 0.8034395688143036, "learning_rate": 9.963600887878998e-06, "loss": 0.2135, "step": 99 }, { "epoch": 0.07763975155279502, "grad_norm": 0.7763163119480374, "learning_rate": 9.962862754688118e-06, "loss": 0.2511, "step": 100 }, { "epoch": 0.07841614906832298, "grad_norm": 0.8513116563830826, "learning_rate": 9.962117240071765e-06, "loss": 0.2932, "step": 101 }, { "epoch": 0.07919254658385093, "grad_norm": 1.2865966183728745, "learning_rate": 9.961364345138764e-06, "loss": 0.2623, "step": 102 }, { "epoch": 0.07996894409937888, "grad_norm": 0.8584652727563721, "learning_rate": 9.960604071008922e-06, "loss": 0.2391, "step": 103 }, { "epoch": 0.08074534161490683, "grad_norm": 0.8223927302482605, "learning_rate": 9.959836418813016e-06, "loss": 0.2842, "step": 104 }, { "epoch": 0.08152173913043478, "grad_norm": 0.7637213969822378, "learning_rate": 9.9590613896928e-06, "loss": 0.3256, "step": 105 }, { "epoch": 0.08229813664596274, "grad_norm": 0.866528328740758, "learning_rate": 9.958278984801003e-06, "loss": 0.2327, "step": 106 }, { "epoch": 0.08307453416149069, "grad_norm": 0.6906702260061015, "learning_rate": 9.95748920530132e-06, "loss": 0.2198, "step": 107 }, { "epoch": 0.08385093167701864, "grad_norm": 0.7914836020716904, "learning_rate": 9.956692052368416e-06, "loss": 0.3056, "step": 108 }, { "epoch": 0.08462732919254658, "grad_norm": 0.6025966871481879, "learning_rate": 9.955887527187918e-06, "loss": 0.2551, "step": 109 }, { "epoch": 0.08540372670807453, "grad_norm": 0.7056049429110242, "learning_rate": 9.955075630956425e-06, "loss": 0.2112, "step": 110 }, { "epoch": 0.08618012422360248, "grad_norm": 0.576508332016202, "learning_rate": 9.954256364881497e-06, "loss": 0.2275, "step": 111 }, { "epoch": 0.08695652173913043, "grad_norm": 0.6141069176401166, "learning_rate": 9.953429730181653e-06, "loss": 0.2606, "step": 112 }, { "epoch": 0.08773291925465838, "grad_norm": 0.7696596382475545, "learning_rate": 9.952595728086379e-06, "loss": 0.2372, "step": 113 }, { "epoch": 0.08850931677018634, "grad_norm": 0.5552571201453254, "learning_rate": 9.951754359836104e-06, "loss": 0.3003, "step": 114 }, { "epoch": 0.08928571428571429, "grad_norm": 0.9105176541436153, "learning_rate": 9.950905626682229e-06, "loss": 0.277, "step": 115 }, { "epoch": 0.09006211180124224, "grad_norm": 0.7101965450936458, "learning_rate": 9.950049529887098e-06, "loss": 0.3023, "step": 116 }, { "epoch": 0.09083850931677019, "grad_norm": 0.8751485174519528, "learning_rate": 9.949186070724017e-06, "loss": 0.1805, "step": 117 }, { "epoch": 0.09161490683229814, "grad_norm": 0.7416134548645074, "learning_rate": 9.94831525047723e-06, "loss": 0.2905, "step": 118 }, { "epoch": 0.09239130434782608, "grad_norm": 0.7885909920840118, "learning_rate": 9.947437070441938e-06, "loss": 0.2662, "step": 119 }, { "epoch": 0.09316770186335403, "grad_norm": 0.6989169484381573, "learning_rate": 9.94655153192429e-06, "loss": 0.2392, "step": 120 }, { "epoch": 0.09394409937888198, "grad_norm": 0.7976339120547224, "learning_rate": 9.945658636241372e-06, "loss": 0.2147, "step": 121 }, { "epoch": 0.09472049689440994, "grad_norm": 1.3806689540868466, "learning_rate": 9.944758384721218e-06, "loss": 0.7698, "step": 122 }, { "epoch": 0.09549689440993789, "grad_norm": 0.7522712754348362, "learning_rate": 9.943850778702799e-06, "loss": 0.3253, "step": 123 }, { "epoch": 0.09627329192546584, "grad_norm": 1.0415267119889349, "learning_rate": 9.94293581953603e-06, "loss": 0.3238, "step": 124 }, { "epoch": 0.09704968944099379, "grad_norm": 0.8366030518740681, "learning_rate": 9.942013508581758e-06, "loss": 0.2613, "step": 125 }, { "epoch": 0.09782608695652174, "grad_norm": 0.9181179252126481, "learning_rate": 9.941083847211765e-06, "loss": 0.3228, "step": 126 }, { "epoch": 0.0986024844720497, "grad_norm": 0.8777200579818952, "learning_rate": 9.940146836808766e-06, "loss": 0.2026, "step": 127 }, { "epoch": 0.09937888198757763, "grad_norm": 0.7368199257381471, "learning_rate": 9.939202478766408e-06, "loss": 0.244, "step": 128 }, { "epoch": 0.10015527950310558, "grad_norm": 0.6024306476282612, "learning_rate": 9.938250774489264e-06, "loss": 0.2491, "step": 129 }, { "epoch": 0.10093167701863354, "grad_norm": 3.1689114350552745, "learning_rate": 9.937291725392835e-06, "loss": 0.8197, "step": 130 }, { "epoch": 0.10170807453416149, "grad_norm": 0.7679828030406183, "learning_rate": 9.936325332903545e-06, "loss": 0.3607, "step": 131 }, { "epoch": 0.10248447204968944, "grad_norm": 0.5816563676739679, "learning_rate": 9.935351598458743e-06, "loss": 0.2326, "step": 132 }, { "epoch": 0.10326086956521739, "grad_norm": 0.9808756517180838, "learning_rate": 9.934370523506691e-06, "loss": 0.2871, "step": 133 }, { "epoch": 0.10403726708074534, "grad_norm": 0.7603588610189349, "learning_rate": 9.933382109506577e-06, "loss": 0.3315, "step": 134 }, { "epoch": 0.1048136645962733, "grad_norm": 0.8537706876053889, "learning_rate": 9.9323863579285e-06, "loss": 0.2804, "step": 135 }, { "epoch": 0.10559006211180125, "grad_norm": 1.36971146686854, "learning_rate": 9.931383270253475e-06, "loss": 0.8024, "step": 136 }, { "epoch": 0.1063664596273292, "grad_norm": 0.8274082041126947, "learning_rate": 9.930372847973424e-06, "loss": 0.3188, "step": 137 }, { "epoch": 0.10714285714285714, "grad_norm": 0.5962916756758992, "learning_rate": 9.92935509259118e-06, "loss": 0.1935, "step": 138 }, { "epoch": 0.10791925465838509, "grad_norm": 0.7304723971311897, "learning_rate": 9.928330005620485e-06, "loss": 0.3207, "step": 139 }, { "epoch": 0.10869565217391304, "grad_norm": 0.7780718162190278, "learning_rate": 9.927297588585984e-06, "loss": 0.258, "step": 140 }, { "epoch": 0.10947204968944099, "grad_norm": 0.7901233265188956, "learning_rate": 9.926257843023226e-06, "loss": 0.3386, "step": 141 }, { "epoch": 0.11024844720496894, "grad_norm": 0.8829651412725995, "learning_rate": 9.925210770478653e-06, "loss": 0.3456, "step": 142 }, { "epoch": 0.1110248447204969, "grad_norm": 0.9407026941257203, "learning_rate": 9.924156372509612e-06, "loss": 0.2913, "step": 143 }, { "epoch": 0.11180124223602485, "grad_norm": 0.6901996902748596, "learning_rate": 9.923094650684346e-06, "loss": 0.2483, "step": 144 }, { "epoch": 0.1125776397515528, "grad_norm": 0.7007100923779781, "learning_rate": 9.922025606581984e-06, "loss": 0.2769, "step": 145 }, { "epoch": 0.11335403726708075, "grad_norm": 0.826041754288556, "learning_rate": 9.92094924179255e-06, "loss": 0.2232, "step": 146 }, { "epoch": 0.11413043478260869, "grad_norm": 0.6487481024895015, "learning_rate": 9.91986555791696e-06, "loss": 0.2465, "step": 147 }, { "epoch": 0.11490683229813664, "grad_norm": 0.691178152053723, "learning_rate": 9.918774556567007e-06, "loss": 0.206, "step": 148 }, { "epoch": 0.11568322981366459, "grad_norm": 1.7193893421932167, "learning_rate": 9.917676239365374e-06, "loss": 0.8688, "step": 149 }, { "epoch": 0.11645962732919254, "grad_norm": 0.8236688168595007, "learning_rate": 9.916570607945627e-06, "loss": 0.3264, "step": 150 }, { "epoch": 0.1172360248447205, "grad_norm": 0.666810953033679, "learning_rate": 9.915457663952201e-06, "loss": 0.2779, "step": 151 }, { "epoch": 0.11801242236024845, "grad_norm": 0.6048761040503008, "learning_rate": 9.914337409040418e-06, "loss": 0.23, "step": 152 }, { "epoch": 0.1187888198757764, "grad_norm": 0.7579743408065878, "learning_rate": 9.913209844876467e-06, "loss": 0.2016, "step": 153 }, { "epoch": 0.11956521739130435, "grad_norm": 0.763917158683279, "learning_rate": 9.912074973137413e-06, "loss": 0.2477, "step": 154 }, { "epoch": 0.1203416149068323, "grad_norm": 0.6955052254991648, "learning_rate": 9.910932795511185e-06, "loss": 0.2235, "step": 155 }, { "epoch": 0.12111801242236025, "grad_norm": 0.8033780034314835, "learning_rate": 9.90978331369658e-06, "loss": 0.1724, "step": 156 }, { "epoch": 0.12189440993788819, "grad_norm": 0.89751404434573, "learning_rate": 9.908626529403263e-06, "loss": 0.2141, "step": 157 }, { "epoch": 0.12267080745341614, "grad_norm": 0.8268378933862544, "learning_rate": 9.90746244435175e-06, "loss": 0.3427, "step": 158 }, { "epoch": 0.1234472049689441, "grad_norm": 0.9227861739419518, "learning_rate": 9.906291060273427e-06, "loss": 0.2866, "step": 159 }, { "epoch": 0.12422360248447205, "grad_norm": 0.819361559580782, "learning_rate": 9.905112378910532e-06, "loss": 0.2595, "step": 160 }, { "epoch": 0.125, "grad_norm": 0.5472282621201959, "learning_rate": 9.903926402016153e-06, "loss": 0.2733, "step": 161 }, { "epoch": 0.12577639751552794, "grad_norm": 0.6404117516972506, "learning_rate": 9.902733131354232e-06, "loss": 0.2481, "step": 162 }, { "epoch": 0.1265527950310559, "grad_norm": 0.8840171403670056, "learning_rate": 9.901532568699562e-06, "loss": 0.293, "step": 163 }, { "epoch": 0.12732919254658384, "grad_norm": 0.8445946924864275, "learning_rate": 9.900324715837772e-06, "loss": 0.292, "step": 164 }, { "epoch": 0.1281055900621118, "grad_norm": 0.7976035932435953, "learning_rate": 9.899109574565346e-06, "loss": 0.215, "step": 165 }, { "epoch": 0.12888198757763975, "grad_norm": 0.6879216350571397, "learning_rate": 9.897887146689603e-06, "loss": 0.2784, "step": 166 }, { "epoch": 0.1296583850931677, "grad_norm": 0.780036920959478, "learning_rate": 9.896657434028693e-06, "loss": 0.3172, "step": 167 }, { "epoch": 0.13043478260869565, "grad_norm": 0.7234166486681037, "learning_rate": 9.895420438411616e-06, "loss": 0.2617, "step": 168 }, { "epoch": 0.13121118012422361, "grad_norm": 0.7574380276063051, "learning_rate": 9.894176161678188e-06, "loss": 0.3138, "step": 169 }, { "epoch": 0.13198757763975155, "grad_norm": 0.5945434647398978, "learning_rate": 9.892924605679066e-06, "loss": 0.2533, "step": 170 }, { "epoch": 0.1327639751552795, "grad_norm": 0.6545102261389113, "learning_rate": 9.891665772275728e-06, "loss": 0.3302, "step": 171 }, { "epoch": 0.13354037267080746, "grad_norm": 1.5073268439062604, "learning_rate": 9.890399663340478e-06, "loss": 0.318, "step": 172 }, { "epoch": 0.1343167701863354, "grad_norm": 0.6999690602046588, "learning_rate": 9.88912628075644e-06, "loss": 0.3086, "step": 173 }, { "epoch": 0.13509316770186336, "grad_norm": 0.956012257209688, "learning_rate": 9.887845626417556e-06, "loss": 0.292, "step": 174 }, { "epoch": 0.1358695652173913, "grad_norm": 0.7417182408407419, "learning_rate": 9.886557702228588e-06, "loss": 0.2791, "step": 175 }, { "epoch": 0.13664596273291926, "grad_norm": 0.7546824982027776, "learning_rate": 9.885262510105102e-06, "loss": 0.3447, "step": 176 }, { "epoch": 0.1374223602484472, "grad_norm": 1.0424664939496846, "learning_rate": 9.883960051973485e-06, "loss": 0.295, "step": 177 }, { "epoch": 0.13819875776397517, "grad_norm": 0.8928669496812321, "learning_rate": 9.882650329770919e-06, "loss": 0.3122, "step": 178 }, { "epoch": 0.1389751552795031, "grad_norm": 0.4786972064005782, "learning_rate": 9.881333345445398e-06, "loss": 0.1992, "step": 179 }, { "epoch": 0.13975155279503104, "grad_norm": 0.5701666810480438, "learning_rate": 9.880009100955716e-06, "loss": 0.1757, "step": 180 }, { "epoch": 0.140527950310559, "grad_norm": 0.6520021813555659, "learning_rate": 9.878677598271466e-06, "loss": 0.2692, "step": 181 }, { "epoch": 0.14130434782608695, "grad_norm": 0.8612416931375627, "learning_rate": 9.877338839373032e-06, "loss": 0.2616, "step": 182 }, { "epoch": 0.1420807453416149, "grad_norm": 0.5761426711562194, "learning_rate": 9.875992826251593e-06, "loss": 0.2238, "step": 183 }, { "epoch": 0.14285714285714285, "grad_norm": 0.6832269731230959, "learning_rate": 9.874639560909118e-06, "loss": 0.2691, "step": 184 }, { "epoch": 0.14363354037267081, "grad_norm": 0.7078257347882394, "learning_rate": 9.873279045358365e-06, "loss": 0.2887, "step": 185 }, { "epoch": 0.14440993788819875, "grad_norm": 0.5340192756574801, "learning_rate": 9.871911281622869e-06, "loss": 0.2906, "step": 186 }, { "epoch": 0.14518633540372672, "grad_norm": 0.754920500406312, "learning_rate": 9.87053627173695e-06, "loss": 0.2636, "step": 187 }, { "epoch": 0.14596273291925466, "grad_norm": 0.9585922064884287, "learning_rate": 9.869154017745706e-06, "loss": 0.3048, "step": 188 }, { "epoch": 0.14673913043478262, "grad_norm": 0.7574523018515066, "learning_rate": 9.867764521705006e-06, "loss": 0.3453, "step": 189 }, { "epoch": 0.14751552795031056, "grad_norm": 0.6625288834429959, "learning_rate": 9.866367785681493e-06, "loss": 0.2736, "step": 190 }, { "epoch": 0.1482919254658385, "grad_norm": 0.7955412025179897, "learning_rate": 9.864963811752577e-06, "loss": 0.2437, "step": 191 }, { "epoch": 0.14906832298136646, "grad_norm": 0.5310585223204625, "learning_rate": 9.863552602006435e-06, "loss": 0.185, "step": 192 }, { "epoch": 0.1498447204968944, "grad_norm": 0.717819837972584, "learning_rate": 9.862134158542e-06, "loss": 0.2361, "step": 193 }, { "epoch": 0.15062111801242237, "grad_norm": 0.7267931515698243, "learning_rate": 9.860708483468975e-06, "loss": 0.2747, "step": 194 }, { "epoch": 0.1513975155279503, "grad_norm": 0.536623600421279, "learning_rate": 9.85927557890781e-06, "loss": 0.2571, "step": 195 }, { "epoch": 0.15217391304347827, "grad_norm": 0.7902737065352243, "learning_rate": 9.857835446989708e-06, "loss": 0.2553, "step": 196 }, { "epoch": 0.1529503105590062, "grad_norm": 0.7835115195505311, "learning_rate": 9.856388089856626e-06, "loss": 0.3089, "step": 197 }, { "epoch": 0.15372670807453417, "grad_norm": 0.7595092439282967, "learning_rate": 9.854933509661264e-06, "loss": 0.2403, "step": 198 }, { "epoch": 0.1545031055900621, "grad_norm": 0.556334762072189, "learning_rate": 9.853471708567066e-06, "loss": 0.2799, "step": 199 }, { "epoch": 0.15527950310559005, "grad_norm": 0.7193706370628716, "learning_rate": 9.852002688748214e-06, "loss": 0.3244, "step": 200 }, { "epoch": 0.15605590062111802, "grad_norm": 0.9040407408224647, "learning_rate": 9.850526452389632e-06, "loss": 0.2209, "step": 201 }, { "epoch": 0.15683229813664595, "grad_norm": 0.7753662270165959, "learning_rate": 9.84904300168697e-06, "loss": 0.2456, "step": 202 }, { "epoch": 0.15760869565217392, "grad_norm": 1.1329134327357462, "learning_rate": 9.847552338846617e-06, "loss": 0.2107, "step": 203 }, { "epoch": 0.15838509316770186, "grad_norm": 0.7715663099414143, "learning_rate": 9.84605446608568e-06, "loss": 0.2534, "step": 204 }, { "epoch": 0.15916149068322982, "grad_norm": 0.7509807400887848, "learning_rate": 9.844549385631991e-06, "loss": 0.2725, "step": 205 }, { "epoch": 0.15993788819875776, "grad_norm": 0.8511096698197393, "learning_rate": 9.843037099724111e-06, "loss": 0.2826, "step": 206 }, { "epoch": 0.16071428571428573, "grad_norm": 0.705665771616296, "learning_rate": 9.841517610611309e-06, "loss": 0.1792, "step": 207 }, { "epoch": 0.16149068322981366, "grad_norm": 0.8613321704583725, "learning_rate": 9.839990920553566e-06, "loss": 0.297, "step": 208 }, { "epoch": 0.1622670807453416, "grad_norm": 0.8026698795685391, "learning_rate": 9.838457031821583e-06, "loss": 0.2085, "step": 209 }, { "epoch": 0.16304347826086957, "grad_norm": 0.6273441681162587, "learning_rate": 9.83691594669676e-06, "loss": 0.23, "step": 210 }, { "epoch": 0.1638198757763975, "grad_norm": 0.7402842378532869, "learning_rate": 9.835367667471199e-06, "loss": 0.2479, "step": 211 }, { "epoch": 0.16459627329192547, "grad_norm": 0.5263251198299485, "learning_rate": 9.833812196447712e-06, "loss": 0.257, "step": 212 }, { "epoch": 0.1653726708074534, "grad_norm": 0.7121458510664959, "learning_rate": 9.832249535939795e-06, "loss": 0.2339, "step": 213 }, { "epoch": 0.16614906832298137, "grad_norm": 0.9606531745140163, "learning_rate": 9.830679688271647e-06, "loss": 0.2449, "step": 214 }, { "epoch": 0.1669254658385093, "grad_norm": 0.9510256966163668, "learning_rate": 9.829102655778152e-06, "loss": 0.2797, "step": 215 }, { "epoch": 0.16770186335403728, "grad_norm": 0.7490362669223012, "learning_rate": 9.827518440804882e-06, "loss": 0.2415, "step": 216 }, { "epoch": 0.16847826086956522, "grad_norm": 0.7619764995051643, "learning_rate": 9.825927045708087e-06, "loss": 0.228, "step": 217 }, { "epoch": 0.16925465838509315, "grad_norm": 0.7544769310552952, "learning_rate": 9.824328472854706e-06, "loss": 0.195, "step": 218 }, { "epoch": 0.17003105590062112, "grad_norm": 0.7261149987531643, "learning_rate": 9.822722724622345e-06, "loss": 0.2371, "step": 219 }, { "epoch": 0.17080745341614906, "grad_norm": 0.6594260572599918, "learning_rate": 9.821109803399283e-06, "loss": 0.2602, "step": 220 }, { "epoch": 0.17158385093167702, "grad_norm": 0.8532464140989225, "learning_rate": 9.819489711584474e-06, "loss": 0.2628, "step": 221 }, { "epoch": 0.17236024844720496, "grad_norm": 0.6386035970236451, "learning_rate": 9.817862451587527e-06, "loss": 0.1962, "step": 222 }, { "epoch": 0.17313664596273293, "grad_norm": 0.9106691550162497, "learning_rate": 9.816228025828724e-06, "loss": 0.3053, "step": 223 }, { "epoch": 0.17391304347826086, "grad_norm": 0.9461963100003425, "learning_rate": 9.814586436738998e-06, "loss": 0.2868, "step": 224 }, { "epoch": 0.17468944099378883, "grad_norm": 0.6656660010080179, "learning_rate": 9.812937686759933e-06, "loss": 0.2151, "step": 225 }, { "epoch": 0.17546583850931677, "grad_norm": 0.7042524145615784, "learning_rate": 9.811281778343773e-06, "loss": 0.2277, "step": 226 }, { "epoch": 0.17624223602484473, "grad_norm": 0.6987617051974729, "learning_rate": 9.8096187139534e-06, "loss": 0.2368, "step": 227 }, { "epoch": 0.17701863354037267, "grad_norm": 0.6101845406070561, "learning_rate": 9.807948496062344e-06, "loss": 0.227, "step": 228 }, { "epoch": 0.1777950310559006, "grad_norm": 0.7916194557181534, "learning_rate": 9.806271127154775e-06, "loss": 0.2696, "step": 229 }, { "epoch": 0.17857142857142858, "grad_norm": 0.7035117017287085, "learning_rate": 9.804586609725499e-06, "loss": 0.2495, "step": 230 }, { "epoch": 0.1793478260869565, "grad_norm": 0.7400071560338868, "learning_rate": 9.802894946279951e-06, "loss": 0.3227, "step": 231 }, { "epoch": 0.18012422360248448, "grad_norm": 0.9011629998044673, "learning_rate": 9.801196139334195e-06, "loss": 0.3036, "step": 232 }, { "epoch": 0.18090062111801242, "grad_norm": 0.6554241700693474, "learning_rate": 9.799490191414924e-06, "loss": 0.2504, "step": 233 }, { "epoch": 0.18167701863354038, "grad_norm": 0.8647051948688668, "learning_rate": 9.797777105059448e-06, "loss": 0.3129, "step": 234 }, { "epoch": 0.18245341614906832, "grad_norm": 0.7522231539890677, "learning_rate": 9.796056882815695e-06, "loss": 0.2564, "step": 235 }, { "epoch": 0.18322981366459629, "grad_norm": 0.6455748611210668, "learning_rate": 9.794329527242209e-06, "loss": 0.2153, "step": 236 }, { "epoch": 0.18400621118012422, "grad_norm": 0.6497352675636411, "learning_rate": 9.792595040908139e-06, "loss": 0.2344, "step": 237 }, { "epoch": 0.18478260869565216, "grad_norm": 0.7537289955578826, "learning_rate": 9.790853426393246e-06, "loss": 0.2055, "step": 238 }, { "epoch": 0.18555900621118013, "grad_norm": 0.7088586848088342, "learning_rate": 9.789104686287883e-06, "loss": 0.2151, "step": 239 }, { "epoch": 0.18633540372670807, "grad_norm": 0.7086343259651879, "learning_rate": 9.787348823193013e-06, "loss": 0.2882, "step": 240 }, { "epoch": 0.18711180124223603, "grad_norm": 0.5710667750286661, "learning_rate": 9.785585839720187e-06, "loss": 0.2372, "step": 241 }, { "epoch": 0.18788819875776397, "grad_norm": 0.6036880405876073, "learning_rate": 9.783815738491544e-06, "loss": 0.221, "step": 242 }, { "epoch": 0.18866459627329193, "grad_norm": 0.6939623923416208, "learning_rate": 9.782038522139816e-06, "loss": 0.2571, "step": 243 }, { "epoch": 0.18944099378881987, "grad_norm": 0.5890865493340847, "learning_rate": 9.780254193308312e-06, "loss": 0.2532, "step": 244 }, { "epoch": 0.19021739130434784, "grad_norm": 0.7367902935865125, "learning_rate": 9.778462754650922e-06, "loss": 0.2624, "step": 245 }, { "epoch": 0.19099378881987578, "grad_norm": 0.7465401011107221, "learning_rate": 9.776664208832111e-06, "loss": 0.3428, "step": 246 }, { "epoch": 0.19177018633540371, "grad_norm": 0.51417163514693, "learning_rate": 9.774858558526914e-06, "loss": 0.2769, "step": 247 }, { "epoch": 0.19254658385093168, "grad_norm": 0.8362944848950593, "learning_rate": 9.77304580642093e-06, "loss": 0.2831, "step": 248 }, { "epoch": 0.19332298136645962, "grad_norm": 0.7562413344452334, "learning_rate": 9.771225955210328e-06, "loss": 0.2627, "step": 249 }, { "epoch": 0.19409937888198758, "grad_norm": 0.6998917576726986, "learning_rate": 9.76939900760183e-06, "loss": 0.3243, "step": 250 }, { "epoch": 0.19487577639751552, "grad_norm": 0.7057360262382673, "learning_rate": 9.767564966312712e-06, "loss": 0.2659, "step": 251 }, { "epoch": 0.1956521739130435, "grad_norm": 0.7528296965505865, "learning_rate": 9.765723834070805e-06, "loss": 0.2809, "step": 252 }, { "epoch": 0.19642857142857142, "grad_norm": 0.7947291116837203, "learning_rate": 9.763875613614482e-06, "loss": 0.3406, "step": 253 }, { "epoch": 0.1972049689440994, "grad_norm": 0.9083052432221724, "learning_rate": 9.762020307692662e-06, "loss": 0.3117, "step": 254 }, { "epoch": 0.19798136645962733, "grad_norm": 0.737852648007368, "learning_rate": 9.760157919064801e-06, "loss": 0.2841, "step": 255 }, { "epoch": 0.19875776397515527, "grad_norm": 0.7121008956916511, "learning_rate": 9.75828845050089e-06, "loss": 0.3033, "step": 256 }, { "epoch": 0.19953416149068323, "grad_norm": 0.981037829126822, "learning_rate": 9.75641190478145e-06, "loss": 0.7481, "step": 257 }, { "epoch": 0.20031055900621117, "grad_norm": 0.6491607692478617, "learning_rate": 9.754528284697526e-06, "loss": 0.2141, "step": 258 }, { "epoch": 0.20108695652173914, "grad_norm": 0.5773516922324985, "learning_rate": 9.752637593050689e-06, "loss": 0.275, "step": 259 }, { "epoch": 0.20186335403726707, "grad_norm": 0.706373076132815, "learning_rate": 9.750739832653023e-06, "loss": 0.253, "step": 260 }, { "epoch": 0.20263975155279504, "grad_norm": 0.7292540365146915, "learning_rate": 9.74883500632713e-06, "loss": 0.2495, "step": 261 }, { "epoch": 0.20341614906832298, "grad_norm": 0.7112194328115442, "learning_rate": 9.74692311690612e-06, "loss": 0.2278, "step": 262 }, { "epoch": 0.20419254658385094, "grad_norm": 0.768241259515224, "learning_rate": 9.745004167233606e-06, "loss": 0.4111, "step": 263 }, { "epoch": 0.20496894409937888, "grad_norm": 0.7459097946404543, "learning_rate": 9.743078160163703e-06, "loss": 0.248, "step": 264 }, { "epoch": 0.20574534161490685, "grad_norm": 0.5995140926432768, "learning_rate": 9.741145098561027e-06, "loss": 0.2463, "step": 265 }, { "epoch": 0.20652173913043478, "grad_norm": 0.5573520511272386, "learning_rate": 9.73920498530068e-06, "loss": 0.2615, "step": 266 }, { "epoch": 0.20729813664596272, "grad_norm": 1.1723963204785914, "learning_rate": 9.737257823268254e-06, "loss": 0.7321, "step": 267 }, { "epoch": 0.2080745341614907, "grad_norm": 0.7872304421507842, "learning_rate": 9.735303615359827e-06, "loss": 0.276, "step": 268 }, { "epoch": 0.20885093167701863, "grad_norm": 0.8669779947683175, "learning_rate": 9.733342364481957e-06, "loss": 0.2556, "step": 269 }, { "epoch": 0.2096273291925466, "grad_norm": 0.816351825252327, "learning_rate": 9.731374073551671e-06, "loss": 0.1744, "step": 270 }, { "epoch": 0.21040372670807453, "grad_norm": 0.7223666081966089, "learning_rate": 9.729398745496477e-06, "loss": 0.235, "step": 271 }, { "epoch": 0.2111801242236025, "grad_norm": 0.7133538718317837, "learning_rate": 9.72741638325434e-06, "loss": 0.2639, "step": 272 }, { "epoch": 0.21195652173913043, "grad_norm": 0.8555357926702251, "learning_rate": 9.725426989773692e-06, "loss": 0.2857, "step": 273 }, { "epoch": 0.2127329192546584, "grad_norm": 0.7674022860243805, "learning_rate": 9.723430568013423e-06, "loss": 0.3139, "step": 274 }, { "epoch": 0.21350931677018634, "grad_norm": 0.7731535716459734, "learning_rate": 9.721427120942875e-06, "loss": 0.3319, "step": 275 }, { "epoch": 0.21428571428571427, "grad_norm": 0.8025354601145694, "learning_rate": 9.719416651541839e-06, "loss": 0.2601, "step": 276 }, { "epoch": 0.21506211180124224, "grad_norm": 0.75882924484993, "learning_rate": 9.717399162800549e-06, "loss": 0.2229, "step": 277 }, { "epoch": 0.21583850931677018, "grad_norm": 0.5983772120581281, "learning_rate": 9.715374657719685e-06, "loss": 0.2101, "step": 278 }, { "epoch": 0.21661490683229814, "grad_norm": 0.6475373277048004, "learning_rate": 9.713343139310355e-06, "loss": 0.2442, "step": 279 }, { "epoch": 0.21739130434782608, "grad_norm": 0.6136493917546132, "learning_rate": 9.711304610594104e-06, "loss": 0.2165, "step": 280 }, { "epoch": 0.21816770186335405, "grad_norm": 0.6768969959703242, "learning_rate": 9.7092590746029e-06, "loss": 0.2359, "step": 281 }, { "epoch": 0.21894409937888198, "grad_norm": 0.7412130164685556, "learning_rate": 9.707206534379133e-06, "loss": 0.2921, "step": 282 }, { "epoch": 0.21972049689440995, "grad_norm": 0.7665835989043408, "learning_rate": 9.705146992975617e-06, "loss": 0.2713, "step": 283 }, { "epoch": 0.2204968944099379, "grad_norm": 0.6975813930551933, "learning_rate": 9.703080453455569e-06, "loss": 0.2381, "step": 284 }, { "epoch": 0.22127329192546583, "grad_norm": 0.8568589847899396, "learning_rate": 9.701006918892621e-06, "loss": 0.2159, "step": 285 }, { "epoch": 0.2220496894409938, "grad_norm": 0.7488130106069226, "learning_rate": 9.69892639237081e-06, "loss": 0.1943, "step": 286 }, { "epoch": 0.22282608695652173, "grad_norm": 0.7087175805299524, "learning_rate": 9.696838876984568e-06, "loss": 0.2729, "step": 287 }, { "epoch": 0.2236024844720497, "grad_norm": 0.7338682179474846, "learning_rate": 9.694744375838725e-06, "loss": 0.2883, "step": 288 }, { "epoch": 0.22437888198757763, "grad_norm": 1.546862600077784, "learning_rate": 9.6926428920485e-06, "loss": 0.8323, "step": 289 }, { "epoch": 0.2251552795031056, "grad_norm": 0.6856205796015825, "learning_rate": 9.690534428739497e-06, "loss": 0.2161, "step": 290 }, { "epoch": 0.22593167701863354, "grad_norm": 0.8288199755004688, "learning_rate": 9.6884189890477e-06, "loss": 0.3192, "step": 291 }, { "epoch": 0.2267080745341615, "grad_norm": 0.7673358054957363, "learning_rate": 9.68629657611947e-06, "loss": 0.3289, "step": 292 }, { "epoch": 0.22748447204968944, "grad_norm": 0.8131046883838445, "learning_rate": 9.684167193111546e-06, "loss": 0.2937, "step": 293 }, { "epoch": 0.22826086956521738, "grad_norm": 0.7195904675240062, "learning_rate": 9.682030843191021e-06, "loss": 0.245, "step": 294 }, { "epoch": 0.22903726708074534, "grad_norm": 0.8279592436205141, "learning_rate": 9.67988752953536e-06, "loss": 0.2779, "step": 295 }, { "epoch": 0.22981366459627328, "grad_norm": 0.5607521173070182, "learning_rate": 9.677737255332381e-06, "loss": 0.2511, "step": 296 }, { "epoch": 0.23059006211180125, "grad_norm": 0.6167422673738514, "learning_rate": 9.675580023780258e-06, "loss": 0.266, "step": 297 }, { "epoch": 0.23136645962732919, "grad_norm": 0.9308766214747854, "learning_rate": 9.67341583808751e-06, "loss": 0.2681, "step": 298 }, { "epoch": 0.23214285714285715, "grad_norm": 0.6011008797867972, "learning_rate": 9.671244701472999e-06, "loss": 0.2615, "step": 299 }, { "epoch": 0.2329192546583851, "grad_norm": 0.7426310667619165, "learning_rate": 9.66906661716593e-06, "loss": 0.3348, "step": 300 }, { "epoch": 0.23369565217391305, "grad_norm": 0.770930687798981, "learning_rate": 9.666881588405832e-06, "loss": 0.2434, "step": 301 }, { "epoch": 0.234472049689441, "grad_norm": 0.9113290392354924, "learning_rate": 9.664689618442572e-06, "loss": 0.2727, "step": 302 }, { "epoch": 0.23524844720496896, "grad_norm": 0.5953603346230537, "learning_rate": 9.66249071053634e-06, "loss": 0.2195, "step": 303 }, { "epoch": 0.2360248447204969, "grad_norm": 0.6668283267581371, "learning_rate": 9.660284867957637e-06, "loss": 0.2888, "step": 304 }, { "epoch": 0.23680124223602483, "grad_norm": 0.8154798511700109, "learning_rate": 9.658072093987286e-06, "loss": 0.2543, "step": 305 }, { "epoch": 0.2375776397515528, "grad_norm": 0.6614351785987349, "learning_rate": 9.655852391916418e-06, "loss": 0.21, "step": 306 }, { "epoch": 0.23835403726708074, "grad_norm": 0.7316615864641156, "learning_rate": 9.653625765046469e-06, "loss": 0.3121, "step": 307 }, { "epoch": 0.2391304347826087, "grad_norm": 0.7525658009539544, "learning_rate": 9.651392216689167e-06, "loss": 0.2754, "step": 308 }, { "epoch": 0.23990683229813664, "grad_norm": 0.7262007390661361, "learning_rate": 9.649151750166544e-06, "loss": 0.2806, "step": 309 }, { "epoch": 0.2406832298136646, "grad_norm": 0.6814881012567792, "learning_rate": 9.646904368810919e-06, "loss": 0.3108, "step": 310 }, { "epoch": 0.24145962732919254, "grad_norm": 0.6087443982914852, "learning_rate": 9.644650075964892e-06, "loss": 0.217, "step": 311 }, { "epoch": 0.2422360248447205, "grad_norm": 0.9174881380935349, "learning_rate": 9.642388874981348e-06, "loss": 0.3173, "step": 312 }, { "epoch": 0.24301242236024845, "grad_norm": 0.6391502710582059, "learning_rate": 9.640120769223443e-06, "loss": 0.2791, "step": 313 }, { "epoch": 0.24378881987577639, "grad_norm": 0.5871131124358696, "learning_rate": 9.637845762064602e-06, "loss": 0.2327, "step": 314 }, { "epoch": 0.24456521739130435, "grad_norm": 0.933266948623436, "learning_rate": 9.635563856888516e-06, "loss": 0.2497, "step": 315 }, { "epoch": 0.2453416149068323, "grad_norm": 0.8283710683070198, "learning_rate": 9.63327505708914e-06, "loss": 0.2808, "step": 316 }, { "epoch": 0.24611801242236025, "grad_norm": 0.702235378789715, "learning_rate": 9.630979366070676e-06, "loss": 0.2272, "step": 317 }, { "epoch": 0.2468944099378882, "grad_norm": 0.6796281305612405, "learning_rate": 9.628676787247582e-06, "loss": 0.3022, "step": 318 }, { "epoch": 0.24767080745341616, "grad_norm": 0.6712810245511208, "learning_rate": 9.626367324044554e-06, "loss": 0.2478, "step": 319 }, { "epoch": 0.2484472049689441, "grad_norm": 0.6031940034095149, "learning_rate": 9.624050979896533e-06, "loss": 0.1947, "step": 320 }, { "epoch": 0.24922360248447206, "grad_norm": 0.6276718061737261, "learning_rate": 9.621727758248694e-06, "loss": 0.2204, "step": 321 }, { "epoch": 0.25, "grad_norm": 0.69783872811453, "learning_rate": 9.619397662556434e-06, "loss": 0.2672, "step": 322 }, { "epoch": 0.25077639751552794, "grad_norm": 0.6385355371206729, "learning_rate": 9.617060696285384e-06, "loss": 0.263, "step": 323 }, { "epoch": 0.2515527950310559, "grad_norm": 0.7715814158836041, "learning_rate": 9.614716862911388e-06, "loss": 0.2919, "step": 324 }, { "epoch": 0.25232919254658387, "grad_norm": 0.6133448792075293, "learning_rate": 9.612366165920505e-06, "loss": 0.2042, "step": 325 }, { "epoch": 0.2531055900621118, "grad_norm": 0.609829293200316, "learning_rate": 9.610008608809003e-06, "loss": 0.3322, "step": 326 }, { "epoch": 0.25388198757763975, "grad_norm": 0.7157404447464578, "learning_rate": 9.60764419508335e-06, "loss": 0.2223, "step": 327 }, { "epoch": 0.2546583850931677, "grad_norm": 1.2013629845482579, "learning_rate": 9.605272928260215e-06, "loss": 0.7803, "step": 328 }, { "epoch": 0.2554347826086957, "grad_norm": 0.7880799686391092, "learning_rate": 9.602894811866461e-06, "loss": 0.2648, "step": 329 }, { "epoch": 0.2562111801242236, "grad_norm": 0.7039341960554237, "learning_rate": 9.600509849439136e-06, "loss": 0.2836, "step": 330 }, { "epoch": 0.25698757763975155, "grad_norm": 0.812468578872357, "learning_rate": 9.598118044525476e-06, "loss": 0.2868, "step": 331 }, { "epoch": 0.2577639751552795, "grad_norm": 0.5911858570558136, "learning_rate": 9.595719400682882e-06, "loss": 0.2263, "step": 332 }, { "epoch": 0.25854037267080743, "grad_norm": 0.7998821321657226, "learning_rate": 9.593313921478936e-06, "loss": 0.2793, "step": 333 }, { "epoch": 0.2593167701863354, "grad_norm": 0.623684243858524, "learning_rate": 9.590901610491388e-06, "loss": 0.2763, "step": 334 }, { "epoch": 0.26009316770186336, "grad_norm": 0.8099609734581914, "learning_rate": 9.588482471308143e-06, "loss": 0.3559, "step": 335 }, { "epoch": 0.2608695652173913, "grad_norm": 0.6594726126124933, "learning_rate": 9.586056507527266e-06, "loss": 0.2487, "step": 336 }, { "epoch": 0.26164596273291924, "grad_norm": 0.5484064345624527, "learning_rate": 9.58362372275697e-06, "loss": 0.2576, "step": 337 }, { "epoch": 0.26242236024844723, "grad_norm": 0.5835645063634163, "learning_rate": 9.581184120615615e-06, "loss": 0.251, "step": 338 }, { "epoch": 0.26319875776397517, "grad_norm": 0.6435537292124005, "learning_rate": 9.578737704731698e-06, "loss": 0.2504, "step": 339 }, { "epoch": 0.2639751552795031, "grad_norm": 0.7931132029412593, "learning_rate": 9.576284478743855e-06, "loss": 0.3117, "step": 340 }, { "epoch": 0.26475155279503104, "grad_norm": 0.6626900779554077, "learning_rate": 9.573824446300845e-06, "loss": 0.3158, "step": 341 }, { "epoch": 0.265527950310559, "grad_norm": 0.7631035332692743, "learning_rate": 9.571357611061556e-06, "loss": 0.2345, "step": 342 }, { "epoch": 0.266304347826087, "grad_norm": 0.8333978314116464, "learning_rate": 9.56888397669499e-06, "loss": 0.2608, "step": 343 }, { "epoch": 0.2670807453416149, "grad_norm": 0.6732763404168998, "learning_rate": 9.566403546880262e-06, "loss": 0.2757, "step": 344 }, { "epoch": 0.26785714285714285, "grad_norm": 0.8509096253670358, "learning_rate": 9.563916325306595e-06, "loss": 0.2262, "step": 345 }, { "epoch": 0.2686335403726708, "grad_norm": 0.6468519448636082, "learning_rate": 9.561422315673316e-06, "loss": 0.2548, "step": 346 }, { "epoch": 0.2694099378881988, "grad_norm": 0.6966455105997112, "learning_rate": 9.558921521689845e-06, "loss": 0.2196, "step": 347 }, { "epoch": 0.2701863354037267, "grad_norm": 0.8677449326290912, "learning_rate": 9.556413947075693e-06, "loss": 0.211, "step": 348 }, { "epoch": 0.27096273291925466, "grad_norm": 0.6989282674058415, "learning_rate": 9.553899595560457e-06, "loss": 0.2403, "step": 349 }, { "epoch": 0.2717391304347826, "grad_norm": 0.7846967542721013, "learning_rate": 9.551378470883813e-06, "loss": 0.2812, "step": 350 }, { "epoch": 0.27251552795031053, "grad_norm": 0.67091164111833, "learning_rate": 9.548850576795508e-06, "loss": 0.1987, "step": 351 }, { "epoch": 0.2732919254658385, "grad_norm": 0.5436522796250561, "learning_rate": 9.546315917055362e-06, "loss": 0.1806, "step": 352 }, { "epoch": 0.27406832298136646, "grad_norm": 0.6621273291250352, "learning_rate": 9.543774495433258e-06, "loss": 0.2172, "step": 353 }, { "epoch": 0.2748447204968944, "grad_norm": 0.6433850300048549, "learning_rate": 9.541226315709133e-06, "loss": 0.3217, "step": 354 }, { "epoch": 0.27562111801242234, "grad_norm": 0.6524400873228027, "learning_rate": 9.538671381672975e-06, "loss": 0.2196, "step": 355 }, { "epoch": 0.27639751552795033, "grad_norm": 0.8778038817359622, "learning_rate": 9.536109697124822e-06, "loss": 0.2043, "step": 356 }, { "epoch": 0.27717391304347827, "grad_norm": 0.7339167215371009, "learning_rate": 9.533541265874748e-06, "loss": 0.2922, "step": 357 }, { "epoch": 0.2779503105590062, "grad_norm": 0.8566186659526986, "learning_rate": 9.530966091742863e-06, "loss": 0.2834, "step": 358 }, { "epoch": 0.27872670807453415, "grad_norm": 0.7777846073007713, "learning_rate": 9.52838417855931e-06, "loss": 0.2235, "step": 359 }, { "epoch": 0.2795031055900621, "grad_norm": 0.759310464683535, "learning_rate": 9.525795530164248e-06, "loss": 0.2838, "step": 360 }, { "epoch": 0.2802795031055901, "grad_norm": 0.7419276628290562, "learning_rate": 9.523200150407858e-06, "loss": 0.2739, "step": 361 }, { "epoch": 0.281055900621118, "grad_norm": 1.0819566086677523, "learning_rate": 9.520598043150333e-06, "loss": 0.2186, "step": 362 }, { "epoch": 0.28183229813664595, "grad_norm": 0.7032161164690148, "learning_rate": 9.517989212261868e-06, "loss": 0.2051, "step": 363 }, { "epoch": 0.2826086956521739, "grad_norm": 0.7269169386396658, "learning_rate": 9.515373661622665e-06, "loss": 0.2623, "step": 364 }, { "epoch": 0.2833850931677019, "grad_norm": 0.688344878083794, "learning_rate": 9.512751395122912e-06, "loss": 0.233, "step": 365 }, { "epoch": 0.2841614906832298, "grad_norm": 0.8058677946510884, "learning_rate": 9.510122416662795e-06, "loss": 0.3603, "step": 366 }, { "epoch": 0.28493788819875776, "grad_norm": 0.5573150231093567, "learning_rate": 9.507486730152475e-06, "loss": 0.2358, "step": 367 }, { "epoch": 0.2857142857142857, "grad_norm": 0.7410636495869829, "learning_rate": 9.504844339512096e-06, "loss": 0.2737, "step": 368 }, { "epoch": 0.2864906832298137, "grad_norm": 0.8996763452829911, "learning_rate": 9.502195248671768e-06, "loss": 0.2759, "step": 369 }, { "epoch": 0.28726708074534163, "grad_norm": 0.9612347805650744, "learning_rate": 9.499539461571571e-06, "loss": 0.3099, "step": 370 }, { "epoch": 0.28804347826086957, "grad_norm": 0.7571088031672728, "learning_rate": 9.496876982161543e-06, "loss": 0.2504, "step": 371 }, { "epoch": 0.2888198757763975, "grad_norm": 1.4345115981696597, "learning_rate": 9.494207814401673e-06, "loss": 0.8713, "step": 372 }, { "epoch": 0.28959627329192544, "grad_norm": 0.6734582393152051, "learning_rate": 9.4915319622619e-06, "loss": 0.2649, "step": 373 }, { "epoch": 0.29037267080745344, "grad_norm": 0.6920413278197786, "learning_rate": 9.488849429722108e-06, "loss": 0.3161, "step": 374 }, { "epoch": 0.2911490683229814, "grad_norm": 0.7300384502981266, "learning_rate": 9.486160220772112e-06, "loss": 0.3284, "step": 375 }, { "epoch": 0.2919254658385093, "grad_norm": 0.8146780678483106, "learning_rate": 9.483464339411658e-06, "loss": 0.2829, "step": 376 }, { "epoch": 0.29270186335403725, "grad_norm": 0.6534999489514138, "learning_rate": 9.480761789650418e-06, "loss": 0.2851, "step": 377 }, { "epoch": 0.29347826086956524, "grad_norm": 0.8976526406928339, "learning_rate": 9.478052575507983e-06, "loss": 0.32, "step": 378 }, { "epoch": 0.2942546583850932, "grad_norm": 0.6966680846364296, "learning_rate": 9.475336701013853e-06, "loss": 0.3127, "step": 379 }, { "epoch": 0.2950310559006211, "grad_norm": 1.0377621843419973, "learning_rate": 9.472614170207435e-06, "loss": 0.7701, "step": 380 }, { "epoch": 0.29580745341614906, "grad_norm": 0.754339774299203, "learning_rate": 9.469884987138038e-06, "loss": 0.2813, "step": 381 }, { "epoch": 0.296583850931677, "grad_norm": 0.916270316512492, "learning_rate": 9.467149155864864e-06, "loss": 0.2666, "step": 382 }, { "epoch": 0.297360248447205, "grad_norm": 0.9377676835119274, "learning_rate": 9.464406680457e-06, "loss": 0.8459, "step": 383 }, { "epoch": 0.2981366459627329, "grad_norm": 0.6611864378902077, "learning_rate": 9.461657564993419e-06, "loss": 0.2333, "step": 384 }, { "epoch": 0.29891304347826086, "grad_norm": 0.611840498471507, "learning_rate": 9.45890181356297e-06, "loss": 0.2727, "step": 385 }, { "epoch": 0.2996894409937888, "grad_norm": 0.5086500019115702, "learning_rate": 9.456139430264371e-06, "loss": 0.2537, "step": 386 }, { "epoch": 0.3004658385093168, "grad_norm": 1.1657031952881638, "learning_rate": 9.453370419206202e-06, "loss": 0.7532, "step": 387 }, { "epoch": 0.30124223602484473, "grad_norm": 0.6619947305724861, "learning_rate": 9.450594784506906e-06, "loss": 0.1997, "step": 388 }, { "epoch": 0.30201863354037267, "grad_norm": 0.6636409456182927, "learning_rate": 9.447812530294769e-06, "loss": 0.3249, "step": 389 }, { "epoch": 0.3027950310559006, "grad_norm": 0.8345713085857899, "learning_rate": 9.445023660707928e-06, "loss": 0.2929, "step": 390 }, { "epoch": 0.30357142857142855, "grad_norm": 0.851417671378014, "learning_rate": 9.442228179894362e-06, "loss": 0.2532, "step": 391 }, { "epoch": 0.30434782608695654, "grad_norm": 0.6874743122968594, "learning_rate": 9.439426092011877e-06, "loss": 0.2638, "step": 392 }, { "epoch": 0.3051242236024845, "grad_norm": 0.7327933928399368, "learning_rate": 9.436617401228108e-06, "loss": 0.273, "step": 393 }, { "epoch": 0.3059006211180124, "grad_norm": 0.6886208820992911, "learning_rate": 9.433802111720513e-06, "loss": 0.2257, "step": 394 }, { "epoch": 0.30667701863354035, "grad_norm": 0.678885190284414, "learning_rate": 9.430980227676361e-06, "loss": 0.3454, "step": 395 }, { "epoch": 0.30745341614906835, "grad_norm": 0.7115808946817689, "learning_rate": 9.42815175329273e-06, "loss": 0.226, "step": 396 }, { "epoch": 0.3082298136645963, "grad_norm": 1.146297208790663, "learning_rate": 9.425316692776506e-06, "loss": 0.276, "step": 397 }, { "epoch": 0.3090062111801242, "grad_norm": 0.7758472190282243, "learning_rate": 9.42247505034436e-06, "loss": 0.2588, "step": 398 }, { "epoch": 0.30978260869565216, "grad_norm": 0.6603657534217925, "learning_rate": 9.419626830222762e-06, "loss": 0.2685, "step": 399 }, { "epoch": 0.3105590062111801, "grad_norm": 0.6909326365275471, "learning_rate": 9.416772036647959e-06, "loss": 0.2019, "step": 400 }, { "epoch": 0.3113354037267081, "grad_norm": 0.6831615516655736, "learning_rate": 9.413910673865978e-06, "loss": 0.2813, "step": 401 }, { "epoch": 0.31211180124223603, "grad_norm": 0.8088174218512328, "learning_rate": 9.411042746132617e-06, "loss": 0.2897, "step": 402 }, { "epoch": 0.31288819875776397, "grad_norm": 0.9459857436055191, "learning_rate": 9.408168257713437e-06, "loss": 0.2638, "step": 403 }, { "epoch": 0.3136645962732919, "grad_norm": 1.2458692405636418, "learning_rate": 9.405287212883756e-06, "loss": 0.6347, "step": 404 }, { "epoch": 0.3144409937888199, "grad_norm": 0.8803321358561785, "learning_rate": 9.402399615928646e-06, "loss": 0.203, "step": 405 }, { "epoch": 0.31521739130434784, "grad_norm": 0.7697255838869926, "learning_rate": 9.39950547114292e-06, "loss": 0.2818, "step": 406 }, { "epoch": 0.3159937888198758, "grad_norm": 0.9034593510000288, "learning_rate": 9.396604782831136e-06, "loss": 0.2326, "step": 407 }, { "epoch": 0.3167701863354037, "grad_norm": 0.5560648113805571, "learning_rate": 9.393697555307581e-06, "loss": 0.179, "step": 408 }, { "epoch": 0.31754658385093165, "grad_norm": 0.7498583512793877, "learning_rate": 9.390783792896265e-06, "loss": 0.2569, "step": 409 }, { "epoch": 0.31832298136645965, "grad_norm": 0.949575920118167, "learning_rate": 9.387863499930925e-06, "loss": 0.2695, "step": 410 }, { "epoch": 0.3190993788819876, "grad_norm": 0.8562815175629891, "learning_rate": 9.384936680755002e-06, "loss": 0.2202, "step": 411 }, { "epoch": 0.3198757763975155, "grad_norm": 0.78957621940109, "learning_rate": 9.382003339721653e-06, "loss": 0.257, "step": 412 }, { "epoch": 0.32065217391304346, "grad_norm": 0.6801983518367611, "learning_rate": 9.379063481193725e-06, "loss": 0.2722, "step": 413 }, { "epoch": 0.32142857142857145, "grad_norm": 0.7056064011925675, "learning_rate": 9.376117109543769e-06, "loss": 0.2251, "step": 414 }, { "epoch": 0.3222049689440994, "grad_norm": 0.7475976370790739, "learning_rate": 9.373164229154017e-06, "loss": 0.2554, "step": 415 }, { "epoch": 0.32298136645962733, "grad_norm": 0.5912628009748869, "learning_rate": 9.370204844416381e-06, "loss": 0.229, "step": 416 }, { "epoch": 0.32375776397515527, "grad_norm": 0.7427201154659653, "learning_rate": 9.367238959732453e-06, "loss": 0.2656, "step": 417 }, { "epoch": 0.3245341614906832, "grad_norm": 0.8540377968061292, "learning_rate": 9.364266579513486e-06, "loss": 0.2444, "step": 418 }, { "epoch": 0.3253105590062112, "grad_norm": 1.1420777397036996, "learning_rate": 9.361287708180397e-06, "loss": 0.3374, "step": 419 }, { "epoch": 0.32608695652173914, "grad_norm": 0.696744738994674, "learning_rate": 9.358302350163758e-06, "loss": 0.3232, "step": 420 }, { "epoch": 0.3268633540372671, "grad_norm": 0.6265940801741551, "learning_rate": 9.355310509903786e-06, "loss": 0.2272, "step": 421 }, { "epoch": 0.327639751552795, "grad_norm": 0.89363526878372, "learning_rate": 9.352312191850343e-06, "loss": 0.2578, "step": 422 }, { "epoch": 0.328416149068323, "grad_norm": 0.6640038240708326, "learning_rate": 9.349307400462927e-06, "loss": 0.2951, "step": 423 }, { "epoch": 0.32919254658385094, "grad_norm": 0.5322220103238925, "learning_rate": 9.346296140210653e-06, "loss": 0.2142, "step": 424 }, { "epoch": 0.3299689440993789, "grad_norm": 0.6852910887638334, "learning_rate": 9.343278415572273e-06, "loss": 0.2032, "step": 425 }, { "epoch": 0.3307453416149068, "grad_norm": 0.7421963985044882, "learning_rate": 9.340254231036141e-06, "loss": 0.2486, "step": 426 }, { "epoch": 0.33152173913043476, "grad_norm": 0.5596691441314783, "learning_rate": 9.337223591100228e-06, "loss": 0.2607, "step": 427 }, { "epoch": 0.33229813664596275, "grad_norm": 0.7925478326579457, "learning_rate": 9.334186500272098e-06, "loss": 0.2542, "step": 428 }, { "epoch": 0.3330745341614907, "grad_norm": 0.7874493614623873, "learning_rate": 9.331142963068918e-06, "loss": 0.4164, "step": 429 }, { "epoch": 0.3338509316770186, "grad_norm": 0.7823757030699767, "learning_rate": 9.328092984017436e-06, "loss": 0.2838, "step": 430 }, { "epoch": 0.33462732919254656, "grad_norm": 0.47752694881726465, "learning_rate": 9.325036567653986e-06, "loss": 0.1741, "step": 431 }, { "epoch": 0.33540372670807456, "grad_norm": 0.6360562819825298, "learning_rate": 9.321973718524472e-06, "loss": 0.2487, "step": 432 }, { "epoch": 0.3361801242236025, "grad_norm": 0.7985317558935908, "learning_rate": 9.31890444118437e-06, "loss": 0.3572, "step": 433 }, { "epoch": 0.33695652173913043, "grad_norm": 0.6127414467586937, "learning_rate": 9.315828740198714e-06, "loss": 0.2014, "step": 434 }, { "epoch": 0.33773291925465837, "grad_norm": 0.5573132802700778, "learning_rate": 9.312746620142095e-06, "loss": 0.2449, "step": 435 }, { "epoch": 0.3385093167701863, "grad_norm": 0.6535540667285964, "learning_rate": 9.309658085598646e-06, "loss": 0.2799, "step": 436 }, { "epoch": 0.3392857142857143, "grad_norm": 0.6904825705469757, "learning_rate": 9.306563141162046e-06, "loss": 0.2919, "step": 437 }, { "epoch": 0.34006211180124224, "grad_norm": 0.5608674924265615, "learning_rate": 9.303461791435504e-06, "loss": 0.2313, "step": 438 }, { "epoch": 0.3408385093167702, "grad_norm": 0.7599947893405529, "learning_rate": 9.300354041031759e-06, "loss": 0.365, "step": 439 }, { "epoch": 0.3416149068322981, "grad_norm": 0.6317159586529446, "learning_rate": 9.29723989457307e-06, "loss": 0.2444, "step": 440 }, { "epoch": 0.3423913043478261, "grad_norm": 0.686710926814513, "learning_rate": 9.2941193566912e-06, "loss": 0.2586, "step": 441 }, { "epoch": 0.34316770186335405, "grad_norm": 0.6438792748746542, "learning_rate": 9.290992432027433e-06, "loss": 0.3073, "step": 442 }, { "epoch": 0.343944099378882, "grad_norm": 0.7610609450413235, "learning_rate": 9.287859125232541e-06, "loss": 0.2189, "step": 443 }, { "epoch": 0.3447204968944099, "grad_norm": 0.6169161123733696, "learning_rate": 9.284719440966794e-06, "loss": 0.2313, "step": 444 }, { "epoch": 0.3454968944099379, "grad_norm": 0.7568131361257411, "learning_rate": 9.281573383899942e-06, "loss": 0.3038, "step": 445 }, { "epoch": 0.34627329192546585, "grad_norm": 0.7282146295285906, "learning_rate": 9.278420958711221e-06, "loss": 0.2573, "step": 446 }, { "epoch": 0.3470496894409938, "grad_norm": 0.6964080754715453, "learning_rate": 9.275262170089332e-06, "loss": 0.2782, "step": 447 }, { "epoch": 0.34782608695652173, "grad_norm": 1.1703297502004966, "learning_rate": 9.272097022732444e-06, "loss": 0.2958, "step": 448 }, { "epoch": 0.34860248447204967, "grad_norm": 0.7647545983246937, "learning_rate": 9.268925521348181e-06, "loss": 0.2337, "step": 449 }, { "epoch": 0.34937888198757766, "grad_norm": 0.7297079536015576, "learning_rate": 9.26574767065362e-06, "loss": 0.238, "step": 450 }, { "epoch": 0.3501552795031056, "grad_norm": 1.0400376571753553, "learning_rate": 9.262563475375282e-06, "loss": 0.715, "step": 451 }, { "epoch": 0.35093167701863354, "grad_norm": 0.6727130396139342, "learning_rate": 9.259372940249121e-06, "loss": 0.2434, "step": 452 }, { "epoch": 0.3517080745341615, "grad_norm": 0.7607920173532492, "learning_rate": 9.256176070020525e-06, "loss": 0.1873, "step": 453 }, { "epoch": 0.35248447204968947, "grad_norm": 0.5726544569327583, "learning_rate": 9.252972869444297e-06, "loss": 0.2713, "step": 454 }, { "epoch": 0.3532608695652174, "grad_norm": 0.7082936978394738, "learning_rate": 9.249763343284665e-06, "loss": 0.2831, "step": 455 }, { "epoch": 0.35403726708074534, "grad_norm": 0.5358382040432067, "learning_rate": 9.24654749631526e-06, "loss": 0.2679, "step": 456 }, { "epoch": 0.3548136645962733, "grad_norm": 0.5476906731404848, "learning_rate": 9.243325333319115e-06, "loss": 0.2091, "step": 457 }, { "epoch": 0.3555900621118012, "grad_norm": 0.973752780035315, "learning_rate": 9.240096859088653e-06, "loss": 0.2685, "step": 458 }, { "epoch": 0.3563664596273292, "grad_norm": 0.7691325095548355, "learning_rate": 9.236862078425692e-06, "loss": 0.2301, "step": 459 }, { "epoch": 0.35714285714285715, "grad_norm": 0.6112391890950866, "learning_rate": 9.233620996141421e-06, "loss": 0.2273, "step": 460 }, { "epoch": 0.3579192546583851, "grad_norm": 0.5963178078478285, "learning_rate": 9.23037361705641e-06, "loss": 0.302, "step": 461 }, { "epoch": 0.358695652173913, "grad_norm": 0.6992451907597728, "learning_rate": 9.22711994600059e-06, "loss": 0.2346, "step": 462 }, { "epoch": 0.359472049689441, "grad_norm": 0.7036713714264811, "learning_rate": 9.223859987813248e-06, "loss": 0.2485, "step": 463 }, { "epoch": 0.36024844720496896, "grad_norm": 0.6197034877139489, "learning_rate": 9.220593747343028e-06, "loss": 0.2183, "step": 464 }, { "epoch": 0.3610248447204969, "grad_norm": 0.7461546480476386, "learning_rate": 9.217321229447912e-06, "loss": 0.2474, "step": 465 }, { "epoch": 0.36180124223602483, "grad_norm": 0.7024056254368729, "learning_rate": 9.214042438995222e-06, "loss": 0.2816, "step": 466 }, { "epoch": 0.36257763975155277, "grad_norm": 0.7304869430091281, "learning_rate": 9.210757380861608e-06, "loss": 0.2337, "step": 467 }, { "epoch": 0.36335403726708076, "grad_norm": 1.4505076103082102, "learning_rate": 9.207466059933044e-06, "loss": 0.7207, "step": 468 }, { "epoch": 0.3641304347826087, "grad_norm": 0.6746715247996453, "learning_rate": 9.204168481104815e-06, "loss": 0.2162, "step": 469 }, { "epoch": 0.36490683229813664, "grad_norm": 0.9485076379377025, "learning_rate": 9.20086464928152e-06, "loss": 0.225, "step": 470 }, { "epoch": 0.3656832298136646, "grad_norm": 0.7957063408211883, "learning_rate": 9.197554569377052e-06, "loss": 0.2917, "step": 471 }, { "epoch": 0.36645962732919257, "grad_norm": 0.5502990107446403, "learning_rate": 9.194238246314599e-06, "loss": 0.2098, "step": 472 }, { "epoch": 0.3672360248447205, "grad_norm": 0.7791193954448994, "learning_rate": 9.190915685026635e-06, "loss": 0.3535, "step": 473 }, { "epoch": 0.36801242236024845, "grad_norm": 1.0865165418866805, "learning_rate": 9.187586890454916e-06, "loss": 0.3504, "step": 474 }, { "epoch": 0.3687888198757764, "grad_norm": 0.7147351343239978, "learning_rate": 9.18425186755046e-06, "loss": 0.3636, "step": 475 }, { "epoch": 0.3695652173913043, "grad_norm": 0.746025078842336, "learning_rate": 9.180910621273555e-06, "loss": 0.2689, "step": 476 }, { "epoch": 0.3703416149068323, "grad_norm": 0.7708479130290989, "learning_rate": 9.177563156593748e-06, "loss": 0.2176, "step": 477 }, { "epoch": 0.37111801242236025, "grad_norm": 0.687692232830949, "learning_rate": 9.174209478489826e-06, "loss": 0.2356, "step": 478 }, { "epoch": 0.3718944099378882, "grad_norm": 0.6137353256077296, "learning_rate": 9.170849591949825e-06, "loss": 0.2074, "step": 479 }, { "epoch": 0.37267080745341613, "grad_norm": 0.7090269179319088, "learning_rate": 9.16748350197101e-06, "loss": 0.2481, "step": 480 }, { "epoch": 0.3734472049689441, "grad_norm": 0.7058593273941547, "learning_rate": 9.16411121355988e-06, "loss": 0.279, "step": 481 }, { "epoch": 0.37422360248447206, "grad_norm": 0.6245071743796308, "learning_rate": 9.160732731732144e-06, "loss": 0.2692, "step": 482 }, { "epoch": 0.375, "grad_norm": 0.5453727319130969, "learning_rate": 9.157348061512728e-06, "loss": 0.2735, "step": 483 }, { "epoch": 0.37577639751552794, "grad_norm": 0.703535098446562, "learning_rate": 9.153957207935758e-06, "loss": 0.2958, "step": 484 }, { "epoch": 0.3765527950310559, "grad_norm": 0.723570756468883, "learning_rate": 9.150560176044565e-06, "loss": 0.2051, "step": 485 }, { "epoch": 0.37732919254658387, "grad_norm": 1.3224785923902083, "learning_rate": 9.14715697089166e-06, "loss": 0.8806, "step": 486 }, { "epoch": 0.3781055900621118, "grad_norm": 1.394040761632437, "learning_rate": 9.143747597538741e-06, "loss": 0.6155, "step": 487 }, { "epoch": 0.37888198757763975, "grad_norm": 0.7758014688536107, "learning_rate": 9.140332061056678e-06, "loss": 0.3247, "step": 488 }, { "epoch": 0.3796583850931677, "grad_norm": 0.9053230619414083, "learning_rate": 9.136910366525508e-06, "loss": 0.2599, "step": 489 }, { "epoch": 0.3804347826086957, "grad_norm": 0.6767489975303443, "learning_rate": 9.133482519034428e-06, "loss": 0.2758, "step": 490 }, { "epoch": 0.3812111801242236, "grad_norm": 0.8158589433226974, "learning_rate": 9.130048523681788e-06, "loss": 0.2225, "step": 491 }, { "epoch": 0.38198757763975155, "grad_norm": 0.7494135337456389, "learning_rate": 9.126608385575076e-06, "loss": 0.285, "step": 492 }, { "epoch": 0.3827639751552795, "grad_norm": 0.6901237056170184, "learning_rate": 9.123162109830924e-06, "loss": 0.3096, "step": 493 }, { "epoch": 0.38354037267080743, "grad_norm": 0.5510914348270773, "learning_rate": 9.119709701575084e-06, "loss": 0.231, "step": 494 }, { "epoch": 0.3843167701863354, "grad_norm": 0.9107514202923879, "learning_rate": 9.116251165942435e-06, "loss": 0.2811, "step": 495 }, { "epoch": 0.38509316770186336, "grad_norm": 0.984981388690512, "learning_rate": 9.112786508076972e-06, "loss": 0.3056, "step": 496 }, { "epoch": 0.3858695652173913, "grad_norm": 0.6373327538458631, "learning_rate": 9.109315733131792e-06, "loss": 0.27, "step": 497 }, { "epoch": 0.38664596273291924, "grad_norm": 0.8710268586365859, "learning_rate": 9.105838846269085e-06, "loss": 0.2023, "step": 498 }, { "epoch": 0.38742236024844723, "grad_norm": 0.575711713479674, "learning_rate": 9.102355852660141e-06, "loss": 0.2054, "step": 499 }, { "epoch": 0.38819875776397517, "grad_norm": 0.6082185377989394, "learning_rate": 9.098866757485328e-06, "loss": 0.3108, "step": 500 }, { "epoch": 0.38819875776397517, "eval_loss": 0.3134617209434509, "eval_runtime": 3.9706, "eval_samples_per_second": 13.348, "eval_steps_per_second": 3.526, "step": 500 }, { "epoch": 0.3889751552795031, "grad_norm": 0.7159421907226899, "learning_rate": 9.095371565934088e-06, "loss": 0.2153, "step": 501 }, { "epoch": 0.38975155279503104, "grad_norm": 0.4863435181887124, "learning_rate": 9.091870283204934e-06, "loss": 0.1893, "step": 502 }, { "epoch": 0.390527950310559, "grad_norm": 0.676029686428065, "learning_rate": 9.088362914505434e-06, "loss": 0.2068, "step": 503 }, { "epoch": 0.391304347826087, "grad_norm": 0.6544292301335922, "learning_rate": 9.08484946505221e-06, "loss": 0.2798, "step": 504 }, { "epoch": 0.3920807453416149, "grad_norm": 0.6379778163028933, "learning_rate": 9.081329940070933e-06, "loss": 0.2935, "step": 505 }, { "epoch": 0.39285714285714285, "grad_norm": 0.6472426204714379, "learning_rate": 9.077804344796302e-06, "loss": 0.3066, "step": 506 }, { "epoch": 0.3936335403726708, "grad_norm": 0.6387589351227656, "learning_rate": 9.074272684472048e-06, "loss": 0.2998, "step": 507 }, { "epoch": 0.3944099378881988, "grad_norm": 0.8464361830292904, "learning_rate": 9.070734964350926e-06, "loss": 0.2895, "step": 508 }, { "epoch": 0.3951863354037267, "grad_norm": 1.0479983224219307, "learning_rate": 9.067191189694698e-06, "loss": 0.3328, "step": 509 }, { "epoch": 0.39596273291925466, "grad_norm": 0.6821909147806972, "learning_rate": 9.063641365774136e-06, "loss": 0.2258, "step": 510 }, { "epoch": 0.3967391304347826, "grad_norm": 0.7990022691467012, "learning_rate": 9.060085497869004e-06, "loss": 0.2855, "step": 511 }, { "epoch": 0.39751552795031053, "grad_norm": 0.9673893804773842, "learning_rate": 9.056523591268064e-06, "loss": 0.2496, "step": 512 }, { "epoch": 0.3982919254658385, "grad_norm": 0.8077682038830548, "learning_rate": 9.05295565126905e-06, "loss": 0.1946, "step": 513 }, { "epoch": 0.39906832298136646, "grad_norm": 0.9022656996888044, "learning_rate": 9.049381683178677e-06, "loss": 0.316, "step": 514 }, { "epoch": 0.3998447204968944, "grad_norm": 0.9735809863977192, "learning_rate": 9.045801692312622e-06, "loss": 0.3347, "step": 515 }, { "epoch": 0.40062111801242234, "grad_norm": 0.7573652554341767, "learning_rate": 9.042215683995516e-06, "loss": 0.2456, "step": 516 }, { "epoch": 0.40139751552795033, "grad_norm": 0.7793643276129235, "learning_rate": 9.038623663560952e-06, "loss": 0.2103, "step": 517 }, { "epoch": 0.40217391304347827, "grad_norm": 0.64198693052374, "learning_rate": 9.035025636351453e-06, "loss": 0.2779, "step": 518 }, { "epoch": 0.4029503105590062, "grad_norm": 0.7294462698803592, "learning_rate": 9.031421607718483e-06, "loss": 0.2642, "step": 519 }, { "epoch": 0.40372670807453415, "grad_norm": 0.7414435544388838, "learning_rate": 9.027811583022427e-06, "loss": 0.2376, "step": 520 }, { "epoch": 0.4045031055900621, "grad_norm": 0.6781716220059273, "learning_rate": 9.024195567632593e-06, "loss": 0.3174, "step": 521 }, { "epoch": 0.4052795031055901, "grad_norm": 0.8435833640716248, "learning_rate": 9.020573566927196e-06, "loss": 0.2363, "step": 522 }, { "epoch": 0.406055900621118, "grad_norm": 0.8460762063797301, "learning_rate": 9.016945586293355e-06, "loss": 0.2779, "step": 523 }, { "epoch": 0.40683229813664595, "grad_norm": 0.7233147190385311, "learning_rate": 9.013311631127082e-06, "loss": 0.2684, "step": 524 }, { "epoch": 0.4076086956521739, "grad_norm": 0.7856374205942598, "learning_rate": 9.009671706833275e-06, "loss": 0.1922, "step": 525 }, { "epoch": 0.4083850931677019, "grad_norm": 0.7400687216230374, "learning_rate": 9.006025818825712e-06, "loss": 0.2352, "step": 526 }, { "epoch": 0.4091614906832298, "grad_norm": 0.8479428701578311, "learning_rate": 9.002373972527038e-06, "loss": 0.3405, "step": 527 }, { "epoch": 0.40993788819875776, "grad_norm": 0.6802030908572582, "learning_rate": 8.998716173368762e-06, "loss": 0.2748, "step": 528 }, { "epoch": 0.4107142857142857, "grad_norm": 0.675808392701892, "learning_rate": 8.995052426791247e-06, "loss": 0.2262, "step": 529 }, { "epoch": 0.4114906832298137, "grad_norm": 0.7649619586283201, "learning_rate": 8.991382738243699e-06, "loss": 0.3148, "step": 530 }, { "epoch": 0.41226708074534163, "grad_norm": 0.6179000273491322, "learning_rate": 8.987707113184167e-06, "loss": 0.2779, "step": 531 }, { "epoch": 0.41304347826086957, "grad_norm": 0.6293139674737878, "learning_rate": 8.984025557079523e-06, "loss": 0.2223, "step": 532 }, { "epoch": 0.4138198757763975, "grad_norm": 0.6910121455254707, "learning_rate": 8.980338075405468e-06, "loss": 0.2527, "step": 533 }, { "epoch": 0.41459627329192544, "grad_norm": 1.0606353749159212, "learning_rate": 8.97664467364651e-06, "loss": 0.2819, "step": 534 }, { "epoch": 0.41537267080745344, "grad_norm": 0.7001382260069741, "learning_rate": 8.972945357295963e-06, "loss": 0.299, "step": 535 }, { "epoch": 0.4161490683229814, "grad_norm": 0.7580590084650086, "learning_rate": 8.96924013185594e-06, "loss": 0.2309, "step": 536 }, { "epoch": 0.4169254658385093, "grad_norm": 0.5751904515316945, "learning_rate": 8.965529002837344e-06, "loss": 0.2899, "step": 537 }, { "epoch": 0.41770186335403725, "grad_norm": 1.0887845546468633, "learning_rate": 8.961811975759854e-06, "loss": 0.304, "step": 538 }, { "epoch": 0.41847826086956524, "grad_norm": 0.580187197394633, "learning_rate": 8.958089056151925e-06, "loss": 0.2611, "step": 539 }, { "epoch": 0.4192546583850932, "grad_norm": 0.7139135468051988, "learning_rate": 8.954360249550776e-06, "loss": 0.2524, "step": 540 }, { "epoch": 0.4200310559006211, "grad_norm": 0.7788593050578159, "learning_rate": 8.95062556150238e-06, "loss": 0.2841, "step": 541 }, { "epoch": 0.42080745341614906, "grad_norm": 0.8314900881826957, "learning_rate": 8.946884997561458e-06, "loss": 0.2532, "step": 542 }, { "epoch": 0.421583850931677, "grad_norm": 0.7515111882967425, "learning_rate": 8.943138563291474e-06, "loss": 0.2138, "step": 543 }, { "epoch": 0.422360248447205, "grad_norm": 0.7072708201252831, "learning_rate": 8.939386264264616e-06, "loss": 0.2613, "step": 544 }, { "epoch": 0.4231366459627329, "grad_norm": 0.5401350751929722, "learning_rate": 8.935628106061803e-06, "loss": 0.2591, "step": 545 }, { "epoch": 0.42391304347826086, "grad_norm": 0.8771202569131178, "learning_rate": 8.931864094272663e-06, "loss": 0.297, "step": 546 }, { "epoch": 0.4246894409937888, "grad_norm": 0.8366917765314889, "learning_rate": 8.928094234495532e-06, "loss": 0.3132, "step": 547 }, { "epoch": 0.4254658385093168, "grad_norm": 0.7935622690074949, "learning_rate": 8.924318532337444e-06, "loss": 0.235, "step": 548 }, { "epoch": 0.42624223602484473, "grad_norm": 0.5994621078813062, "learning_rate": 8.92053699341412e-06, "loss": 0.2164, "step": 549 }, { "epoch": 0.42701863354037267, "grad_norm": 0.7397869788924669, "learning_rate": 8.91674962334997e-06, "loss": 0.2453, "step": 550 }, { "epoch": 0.4277950310559006, "grad_norm": 1.0129872307625516, "learning_rate": 8.912956427778065e-06, "loss": 0.3202, "step": 551 }, { "epoch": 0.42857142857142855, "grad_norm": 0.7058580534780542, "learning_rate": 8.90915741234015e-06, "loss": 0.2635, "step": 552 }, { "epoch": 0.42934782608695654, "grad_norm": 0.5767473424227427, "learning_rate": 8.905352582686622e-06, "loss": 0.2507, "step": 553 }, { "epoch": 0.4301242236024845, "grad_norm": 1.1547216985481703, "learning_rate": 8.901541944476529e-06, "loss": 0.2714, "step": 554 }, { "epoch": 0.4309006211180124, "grad_norm": 0.8659092676200819, "learning_rate": 8.897725503377555e-06, "loss": 0.2692, "step": 555 }, { "epoch": 0.43167701863354035, "grad_norm": 0.6477252557874283, "learning_rate": 8.893903265066011e-06, "loss": 0.2125, "step": 556 }, { "epoch": 0.43245341614906835, "grad_norm": 0.9109923794705705, "learning_rate": 8.890075235226841e-06, "loss": 0.2978, "step": 557 }, { "epoch": 0.4332298136645963, "grad_norm": 0.9523914931572308, "learning_rate": 8.886241419553595e-06, "loss": 0.2851, "step": 558 }, { "epoch": 0.4340062111801242, "grad_norm": 0.8968282807622944, "learning_rate": 8.88240182374843e-06, "loss": 0.2445, "step": 559 }, { "epoch": 0.43478260869565216, "grad_norm": 0.7586542734852664, "learning_rate": 8.8785564535221e-06, "loss": 0.3175, "step": 560 }, { "epoch": 0.4355590062111801, "grad_norm": 0.7272760468942658, "learning_rate": 8.87470531459395e-06, "loss": 0.2664, "step": 561 }, { "epoch": 0.4363354037267081, "grad_norm": 0.9351035658746765, "learning_rate": 8.870848412691899e-06, "loss": 0.2872, "step": 562 }, { "epoch": 0.43711180124223603, "grad_norm": 0.7512978190221707, "learning_rate": 8.866985753552444e-06, "loss": 0.3248, "step": 563 }, { "epoch": 0.43788819875776397, "grad_norm": 0.6368450570931754, "learning_rate": 8.863117342920641e-06, "loss": 0.2275, "step": 564 }, { "epoch": 0.4386645962732919, "grad_norm": 0.6970575626755525, "learning_rate": 8.859243186550103e-06, "loss": 0.2542, "step": 565 }, { "epoch": 0.4394409937888199, "grad_norm": 0.6211345879542, "learning_rate": 8.855363290202988e-06, "loss": 0.2562, "step": 566 }, { "epoch": 0.44021739130434784, "grad_norm": 0.7180253103180386, "learning_rate": 8.85147765964999e-06, "loss": 0.2435, "step": 567 }, { "epoch": 0.4409937888198758, "grad_norm": 0.6924115639261369, "learning_rate": 8.84758630067033e-06, "loss": 0.2563, "step": 568 }, { "epoch": 0.4417701863354037, "grad_norm": 0.6581183747496749, "learning_rate": 8.843689219051754e-06, "loss": 0.3166, "step": 569 }, { "epoch": 0.44254658385093165, "grad_norm": 0.6229826659856843, "learning_rate": 8.839786420590517e-06, "loss": 0.2448, "step": 570 }, { "epoch": 0.44332298136645965, "grad_norm": 0.670975293815163, "learning_rate": 8.835877911091376e-06, "loss": 0.2569, "step": 571 }, { "epoch": 0.4440993788819876, "grad_norm": 0.6775914436665526, "learning_rate": 8.83196369636758e-06, "loss": 0.2463, "step": 572 }, { "epoch": 0.4448757763975155, "grad_norm": 0.772656389184756, "learning_rate": 8.828043782240873e-06, "loss": 0.2944, "step": 573 }, { "epoch": 0.44565217391304346, "grad_norm": 1.4550591312350571, "learning_rate": 8.824118174541464e-06, "loss": 0.3023, "step": 574 }, { "epoch": 0.44642857142857145, "grad_norm": 0.5711732568255075, "learning_rate": 8.820186879108038e-06, "loss": 0.2854, "step": 575 }, { "epoch": 0.4472049689440994, "grad_norm": 0.6484000827547776, "learning_rate": 8.816249901787736e-06, "loss": 0.2982, "step": 576 }, { "epoch": 0.44798136645962733, "grad_norm": 0.7076521053525475, "learning_rate": 8.812307248436152e-06, "loss": 0.3255, "step": 577 }, { "epoch": 0.44875776397515527, "grad_norm": 0.9171804161244709, "learning_rate": 8.808358924917322e-06, "loss": 0.3212, "step": 578 }, { "epoch": 0.4495341614906832, "grad_norm": 0.6340238752175346, "learning_rate": 8.804404937103713e-06, "loss": 0.3043, "step": 579 }, { "epoch": 0.4503105590062112, "grad_norm": 0.6773268171216221, "learning_rate": 8.800445290876218e-06, "loss": 0.2051, "step": 580 }, { "epoch": 0.45108695652173914, "grad_norm": 0.5820818295463773, "learning_rate": 8.79647999212415e-06, "loss": 0.2715, "step": 581 }, { "epoch": 0.4518633540372671, "grad_norm": 0.7008159478478186, "learning_rate": 8.79250904674522e-06, "loss": 0.2839, "step": 582 }, { "epoch": 0.452639751552795, "grad_norm": 0.5980591574839044, "learning_rate": 8.788532460645546e-06, "loss": 0.3324, "step": 583 }, { "epoch": 0.453416149068323, "grad_norm": 0.7369085913496215, "learning_rate": 8.78455023973963e-06, "loss": 0.2942, "step": 584 }, { "epoch": 0.45419254658385094, "grad_norm": 0.7138465518125519, "learning_rate": 8.780562389950362e-06, "loss": 0.3326, "step": 585 }, { "epoch": 0.4549689440993789, "grad_norm": 0.7158720647830331, "learning_rate": 8.776568917208996e-06, "loss": 0.2757, "step": 586 }, { "epoch": 0.4557453416149068, "grad_norm": 1.334397984124625, "learning_rate": 8.772569827455151e-06, "loss": 0.8561, "step": 587 }, { "epoch": 0.45652173913043476, "grad_norm": 0.6883291675757118, "learning_rate": 8.768565126636806e-06, "loss": 0.2441, "step": 588 }, { "epoch": 0.45729813664596275, "grad_norm": 0.5944329884172267, "learning_rate": 8.764554820710279e-06, "loss": 0.2856, "step": 589 }, { "epoch": 0.4580745341614907, "grad_norm": 0.7519427767612432, "learning_rate": 8.760538915640227e-06, "loss": 0.233, "step": 590 }, { "epoch": 0.4588509316770186, "grad_norm": 0.6085926746100909, "learning_rate": 8.756517417399635e-06, "loss": 0.277, "step": 591 }, { "epoch": 0.45962732919254656, "grad_norm": 0.904560629460832, "learning_rate": 8.752490331969807e-06, "loss": 0.3185, "step": 592 }, { "epoch": 0.46040372670807456, "grad_norm": 0.7412448236629708, "learning_rate": 8.748457665340357e-06, "loss": 0.3054, "step": 593 }, { "epoch": 0.4611801242236025, "grad_norm": 0.7199255507176379, "learning_rate": 8.744419423509198e-06, "loss": 0.1707, "step": 594 }, { "epoch": 0.46195652173913043, "grad_norm": 0.6343922296491266, "learning_rate": 8.740375612482541e-06, "loss": 0.2744, "step": 595 }, { "epoch": 0.46273291925465837, "grad_norm": 0.7709421176824103, "learning_rate": 8.736326238274873e-06, "loss": 0.2451, "step": 596 }, { "epoch": 0.4635093167701863, "grad_norm": 0.691371124530588, "learning_rate": 8.73227130690896e-06, "loss": 0.2792, "step": 597 }, { "epoch": 0.4642857142857143, "grad_norm": 0.8014629690384926, "learning_rate": 8.728210824415829e-06, "loss": 0.2041, "step": 598 }, { "epoch": 0.46506211180124224, "grad_norm": 0.7755351352984338, "learning_rate": 8.72414479683477e-06, "loss": 0.3768, "step": 599 }, { "epoch": 0.4658385093167702, "grad_norm": 0.7244101352596073, "learning_rate": 8.720073230213315e-06, "loss": 0.2487, "step": 600 }, { "epoch": 0.4666149068322981, "grad_norm": 0.750491851726308, "learning_rate": 8.715996130607236e-06, "loss": 0.2845, "step": 601 }, { "epoch": 0.4673913043478261, "grad_norm": 0.8794389941435429, "learning_rate": 8.711913504080534e-06, "loss": 0.2878, "step": 602 }, { "epoch": 0.46816770186335405, "grad_norm": 0.6960575033140103, "learning_rate": 8.707825356705434e-06, "loss": 0.2755, "step": 603 }, { "epoch": 0.468944099378882, "grad_norm": 0.7234991799129945, "learning_rate": 8.703731694562364e-06, "loss": 0.3717, "step": 604 }, { "epoch": 0.4697204968944099, "grad_norm": 0.6232696088154713, "learning_rate": 8.699632523739963e-06, "loss": 0.278, "step": 605 }, { "epoch": 0.4704968944099379, "grad_norm": 0.7980886205482958, "learning_rate": 8.69552785033506e-06, "loss": 0.2521, "step": 606 }, { "epoch": 0.47127329192546585, "grad_norm": 0.6299660549119234, "learning_rate": 8.691417680452666e-06, "loss": 0.3195, "step": 607 }, { "epoch": 0.4720496894409938, "grad_norm": 1.8596557254791795, "learning_rate": 8.687302020205967e-06, "loss": 0.8985, "step": 608 }, { "epoch": 0.47282608695652173, "grad_norm": 0.9290588110731003, "learning_rate": 8.683180875716323e-06, "loss": 0.2849, "step": 609 }, { "epoch": 0.47360248447204967, "grad_norm": 0.6642085910077938, "learning_rate": 8.679054253113239e-06, "loss": 0.2537, "step": 610 }, { "epoch": 0.47437888198757766, "grad_norm": 0.867708410214105, "learning_rate": 8.674922158534377e-06, "loss": 0.2804, "step": 611 }, { "epoch": 0.4751552795031056, "grad_norm": 0.7789259795891917, "learning_rate": 8.670784598125532e-06, "loss": 0.3287, "step": 612 }, { "epoch": 0.47593167701863354, "grad_norm": 0.6635625780259038, "learning_rate": 8.666641578040633e-06, "loss": 0.2657, "step": 613 }, { "epoch": 0.4767080745341615, "grad_norm": 0.6133940035968016, "learning_rate": 8.662493104441725e-06, "loss": 0.2837, "step": 614 }, { "epoch": 0.47748447204968947, "grad_norm": 0.7037970885733695, "learning_rate": 8.658339183498968e-06, "loss": 0.2385, "step": 615 }, { "epoch": 0.4782608695652174, "grad_norm": 0.7657496951893137, "learning_rate": 8.65417982139062e-06, "loss": 0.2623, "step": 616 }, { "epoch": 0.47903726708074534, "grad_norm": 0.615056887482628, "learning_rate": 8.650015024303038e-06, "loss": 0.2368, "step": 617 }, { "epoch": 0.4798136645962733, "grad_norm": 0.6121213244664102, "learning_rate": 8.645844798430656e-06, "loss": 0.2025, "step": 618 }, { "epoch": 0.4805900621118012, "grad_norm": 0.6248663878231849, "learning_rate": 8.641669149975985e-06, "loss": 0.2747, "step": 619 }, { "epoch": 0.4813664596273292, "grad_norm": 1.1216166823298932, "learning_rate": 8.637488085149598e-06, "loss": 0.7603, "step": 620 }, { "epoch": 0.48214285714285715, "grad_norm": 0.6683903100490735, "learning_rate": 8.633301610170136e-06, "loss": 0.2579, "step": 621 }, { "epoch": 0.4829192546583851, "grad_norm": 0.6347400187451929, "learning_rate": 8.629109731264272e-06, "loss": 0.2228, "step": 622 }, { "epoch": 0.483695652173913, "grad_norm": 0.5893680124139343, "learning_rate": 8.624912454666722e-06, "loss": 0.3133, "step": 623 }, { "epoch": 0.484472049689441, "grad_norm": 0.6769986840126655, "learning_rate": 8.620709786620231e-06, "loss": 0.29, "step": 624 }, { "epoch": 0.48524844720496896, "grad_norm": 0.6965282518168245, "learning_rate": 8.616501733375567e-06, "loss": 0.239, "step": 625 }, { "epoch": 0.4860248447204969, "grad_norm": 0.7491600195012381, "learning_rate": 8.612288301191497e-06, "loss": 0.2786, "step": 626 }, { "epoch": 0.48680124223602483, "grad_norm": 0.5379148182447155, "learning_rate": 8.608069496334797e-06, "loss": 0.263, "step": 627 }, { "epoch": 0.48757763975155277, "grad_norm": 0.6761050769595672, "learning_rate": 8.60384532508023e-06, "loss": 0.2719, "step": 628 }, { "epoch": 0.48835403726708076, "grad_norm": 0.6514046500963037, "learning_rate": 8.599615793710544e-06, "loss": 0.2981, "step": 629 }, { "epoch": 0.4891304347826087, "grad_norm": 0.7496609163541108, "learning_rate": 8.595380908516454e-06, "loss": 0.2643, "step": 630 }, { "epoch": 0.48990683229813664, "grad_norm": 1.084929697602944, "learning_rate": 8.591140675796644e-06, "loss": 0.3279, "step": 631 }, { "epoch": 0.4906832298136646, "grad_norm": 0.5486454587888825, "learning_rate": 8.586895101857747e-06, "loss": 0.201, "step": 632 }, { "epoch": 0.49145962732919257, "grad_norm": 0.5589845698102587, "learning_rate": 8.582644193014341e-06, "loss": 0.1994, "step": 633 }, { "epoch": 0.4922360248447205, "grad_norm": 0.5220107234548629, "learning_rate": 8.578387955588944e-06, "loss": 0.225, "step": 634 }, { "epoch": 0.49301242236024845, "grad_norm": 0.6405959511512125, "learning_rate": 8.57412639591199e-06, "loss": 0.2591, "step": 635 }, { "epoch": 0.4937888198757764, "grad_norm": 0.658341236957448, "learning_rate": 8.569859520321835e-06, "loss": 0.2066, "step": 636 }, { "epoch": 0.4945652173913043, "grad_norm": 0.8762828178844598, "learning_rate": 8.56558733516474e-06, "loss": 0.237, "step": 637 }, { "epoch": 0.4953416149068323, "grad_norm": 0.7333270786029018, "learning_rate": 8.561309846794865e-06, "loss": 0.3056, "step": 638 }, { "epoch": 0.49611801242236025, "grad_norm": 0.5943259669859408, "learning_rate": 8.557027061574254e-06, "loss": 0.2919, "step": 639 }, { "epoch": 0.4968944099378882, "grad_norm": 0.7476812533028646, "learning_rate": 8.552738985872834e-06, "loss": 0.2428, "step": 640 }, { "epoch": 0.49767080745341613, "grad_norm": 0.6419636036912502, "learning_rate": 8.548445626068395e-06, "loss": 0.2162, "step": 641 }, { "epoch": 0.4984472049689441, "grad_norm": 0.7990003969947816, "learning_rate": 8.544146988546594e-06, "loss": 0.3118, "step": 642 }, { "epoch": 0.49922360248447206, "grad_norm": 0.823517440110996, "learning_rate": 8.539843079700927e-06, "loss": 0.3049, "step": 643 }, { "epoch": 0.5, "grad_norm": 0.7925290311147454, "learning_rate": 8.535533905932739e-06, "loss": 0.2917, "step": 644 }, { "epoch": 0.5007763975155279, "grad_norm": 0.7229896652495374, "learning_rate": 8.531219473651203e-06, "loss": 0.3016, "step": 645 }, { "epoch": 0.5015527950310559, "grad_norm": 0.8000997956343845, "learning_rate": 8.526899789273312e-06, "loss": 0.2466, "step": 646 }, { "epoch": 0.5023291925465838, "grad_norm": 0.5260941497023572, "learning_rate": 8.522574859223872e-06, "loss": 0.2948, "step": 647 }, { "epoch": 0.5031055900621118, "grad_norm": 0.7045816695568168, "learning_rate": 8.518244689935491e-06, "loss": 0.2419, "step": 648 }, { "epoch": 0.5038819875776398, "grad_norm": 0.7265570401596317, "learning_rate": 8.513909287848568e-06, "loss": 0.3543, "step": 649 }, { "epoch": 0.5046583850931677, "grad_norm": 0.7957691820136679, "learning_rate": 8.509568659411288e-06, "loss": 0.2981, "step": 650 }, { "epoch": 0.5054347826086957, "grad_norm": 0.666976574072143, "learning_rate": 8.505222811079607e-06, "loss": 0.259, "step": 651 }, { "epoch": 0.5062111801242236, "grad_norm": 0.6686622871770296, "learning_rate": 8.500871749317244e-06, "loss": 0.2738, "step": 652 }, { "epoch": 0.5069875776397516, "grad_norm": 0.5340010729189706, "learning_rate": 8.496515480595673e-06, "loss": 0.2259, "step": 653 }, { "epoch": 0.5077639751552795, "grad_norm": 0.6673097089912332, "learning_rate": 8.492154011394113e-06, "loss": 0.2914, "step": 654 }, { "epoch": 0.5085403726708074, "grad_norm": 0.8402178803148846, "learning_rate": 8.487787348199517e-06, "loss": 0.3182, "step": 655 }, { "epoch": 0.5093167701863354, "grad_norm": 0.5876492136305395, "learning_rate": 8.483415497506567e-06, "loss": 0.2534, "step": 656 }, { "epoch": 0.5100931677018633, "grad_norm": 0.5950668090198291, "learning_rate": 8.47903846581765e-06, "loss": 0.2614, "step": 657 }, { "epoch": 0.5108695652173914, "grad_norm": 0.6994653491199905, "learning_rate": 8.474656259642874e-06, "loss": 0.2661, "step": 658 }, { "epoch": 0.5116459627329193, "grad_norm": 0.8747386232836167, "learning_rate": 8.47026888550003e-06, "loss": 0.3205, "step": 659 }, { "epoch": 0.5124223602484472, "grad_norm": 0.7181481520576358, "learning_rate": 8.465876349914605e-06, "loss": 0.2319, "step": 660 }, { "epoch": 0.5131987577639752, "grad_norm": 0.7352069249529691, "learning_rate": 8.461478659419756e-06, "loss": 0.2879, "step": 661 }, { "epoch": 0.5139751552795031, "grad_norm": 1.0538058792096248, "learning_rate": 8.457075820556306e-06, "loss": 0.2902, "step": 662 }, { "epoch": 0.514751552795031, "grad_norm": 0.6861027643598765, "learning_rate": 8.452667839872745e-06, "loss": 0.3291, "step": 663 }, { "epoch": 0.515527950310559, "grad_norm": 0.826642765182863, "learning_rate": 8.448254723925205e-06, "loss": 0.3639, "step": 664 }, { "epoch": 0.5163043478260869, "grad_norm": 0.787403919734455, "learning_rate": 8.44383647927745e-06, "loss": 0.2681, "step": 665 }, { "epoch": 0.5170807453416149, "grad_norm": 0.884369100305334, "learning_rate": 8.439413112500878e-06, "loss": 0.242, "step": 666 }, { "epoch": 0.5178571428571429, "grad_norm": 0.6009740427231636, "learning_rate": 8.43498463017451e-06, "loss": 0.2257, "step": 667 }, { "epoch": 0.5186335403726708, "grad_norm": 0.8780300167062591, "learning_rate": 8.430551038884965e-06, "loss": 0.2345, "step": 668 }, { "epoch": 0.5194099378881988, "grad_norm": 0.6180432582398789, "learning_rate": 8.426112345226469e-06, "loss": 0.2445, "step": 669 }, { "epoch": 0.5201863354037267, "grad_norm": 0.512668103169893, "learning_rate": 8.421668555800831e-06, "loss": 0.2339, "step": 670 }, { "epoch": 0.5209627329192547, "grad_norm": 0.6671589888035822, "learning_rate": 8.417219677217446e-06, "loss": 0.241, "step": 671 }, { "epoch": 0.5217391304347826, "grad_norm": 0.9646435747001375, "learning_rate": 8.412765716093273e-06, "loss": 0.2352, "step": 672 }, { "epoch": 0.5225155279503105, "grad_norm": 0.6684936612692239, "learning_rate": 8.408306679052828e-06, "loss": 0.262, "step": 673 }, { "epoch": 0.5232919254658385, "grad_norm": 0.6719966752248916, "learning_rate": 8.403842572728186e-06, "loss": 0.2219, "step": 674 }, { "epoch": 0.5240683229813664, "grad_norm": 0.9312453212114247, "learning_rate": 8.39937340375895e-06, "loss": 0.3101, "step": 675 }, { "epoch": 0.5248447204968945, "grad_norm": 0.7514112596602864, "learning_rate": 8.394899178792262e-06, "loss": 0.3602, "step": 676 }, { "epoch": 0.5256211180124224, "grad_norm": 0.8538687670264569, "learning_rate": 8.390419904482783e-06, "loss": 0.2, "step": 677 }, { "epoch": 0.5263975155279503, "grad_norm": 0.6955948699120955, "learning_rate": 8.38593558749268e-06, "loss": 0.2444, "step": 678 }, { "epoch": 0.5271739130434783, "grad_norm": 2.17832255170732, "learning_rate": 8.381446234491618e-06, "loss": 0.7441, "step": 679 }, { "epoch": 0.5279503105590062, "grad_norm": 0.8436818818057459, "learning_rate": 8.376951852156764e-06, "loss": 0.2869, "step": 680 }, { "epoch": 0.5287267080745341, "grad_norm": 0.8611608713490491, "learning_rate": 8.372452447172752e-06, "loss": 0.2235, "step": 681 }, { "epoch": 0.5295031055900621, "grad_norm": 0.7482795062077086, "learning_rate": 8.367948026231692e-06, "loss": 0.2265, "step": 682 }, { "epoch": 0.53027950310559, "grad_norm": 0.699204599517569, "learning_rate": 8.363438596033155e-06, "loss": 0.3022, "step": 683 }, { "epoch": 0.531055900621118, "grad_norm": 1.1972203957328034, "learning_rate": 8.358924163284164e-06, "loss": 0.7471, "step": 684 }, { "epoch": 0.531832298136646, "grad_norm": 0.6781547315216909, "learning_rate": 8.354404734699174e-06, "loss": 0.2247, "step": 685 }, { "epoch": 0.532608695652174, "grad_norm": 0.6902310237239045, "learning_rate": 8.349880317000083e-06, "loss": 0.1964, "step": 686 }, { "epoch": 0.5333850931677019, "grad_norm": 0.8163114357263984, "learning_rate": 8.345350916916199e-06, "loss": 0.277, "step": 687 }, { "epoch": 0.5341614906832298, "grad_norm": 0.5597730914129729, "learning_rate": 8.34081654118425e-06, "loss": 0.2937, "step": 688 }, { "epoch": 0.5349378881987578, "grad_norm": 0.7635367347393934, "learning_rate": 8.33627719654835e-06, "loss": 0.2877, "step": 689 }, { "epoch": 0.5357142857142857, "grad_norm": 0.6701014039836696, "learning_rate": 8.331732889760021e-06, "loss": 0.242, "step": 690 }, { "epoch": 0.5364906832298136, "grad_norm": 0.8844546119947191, "learning_rate": 8.327183627578151e-06, "loss": 0.26, "step": 691 }, { "epoch": 0.5372670807453416, "grad_norm": 0.7451434094010564, "learning_rate": 8.322629416769007e-06, "loss": 0.1808, "step": 692 }, { "epoch": 0.5380434782608695, "grad_norm": 0.8036798182478685, "learning_rate": 8.318070264106213e-06, "loss": 0.2756, "step": 693 }, { "epoch": 0.5388198757763976, "grad_norm": 0.6472067154581117, "learning_rate": 8.313506176370741e-06, "loss": 0.2369, "step": 694 }, { "epoch": 0.5395962732919255, "grad_norm": 0.5725707513905985, "learning_rate": 8.308937160350907e-06, "loss": 0.1751, "step": 695 }, { "epoch": 0.5403726708074534, "grad_norm": 0.6628323826244552, "learning_rate": 8.304363222842358e-06, "loss": 0.2053, "step": 696 }, { "epoch": 0.5411490683229814, "grad_norm": 0.733845848510582, "learning_rate": 8.299784370648056e-06, "loss": 0.2505, "step": 697 }, { "epoch": 0.5419254658385093, "grad_norm": 0.5722445979726504, "learning_rate": 8.295200610578275e-06, "loss": 0.2162, "step": 698 }, { "epoch": 0.5427018633540373, "grad_norm": 1.6792735431362154, "learning_rate": 8.29061194945059e-06, "loss": 0.8181, "step": 699 }, { "epoch": 0.5434782608695652, "grad_norm": 0.5970160837385581, "learning_rate": 8.286018394089864e-06, "loss": 0.2722, "step": 700 }, { "epoch": 0.5442546583850931, "grad_norm": 0.6351510881135748, "learning_rate": 8.281419951328238e-06, "loss": 0.2655, "step": 701 }, { "epoch": 0.5450310559006211, "grad_norm": 0.5461928630701677, "learning_rate": 8.276816628005127e-06, "loss": 0.2147, "step": 702 }, { "epoch": 0.5458074534161491, "grad_norm": 0.7662472686319565, "learning_rate": 8.2722084309672e-06, "loss": 0.2115, "step": 703 }, { "epoch": 0.546583850931677, "grad_norm": 0.6775452202731999, "learning_rate": 8.267595367068375e-06, "loss": 0.3439, "step": 704 }, { "epoch": 0.547360248447205, "grad_norm": 0.931882792850587, "learning_rate": 8.262977443169816e-06, "loss": 0.7199, "step": 705 }, { "epoch": 0.5481366459627329, "grad_norm": 0.621749520825153, "learning_rate": 8.258354666139903e-06, "loss": 0.2219, "step": 706 }, { "epoch": 0.5489130434782609, "grad_norm": 0.8030819787653105, "learning_rate": 8.253727042854245e-06, "loss": 0.2186, "step": 707 }, { "epoch": 0.5496894409937888, "grad_norm": 1.2512210983953191, "learning_rate": 8.249094580195655e-06, "loss": 0.2742, "step": 708 }, { "epoch": 0.5504658385093167, "grad_norm": 0.6489932240639311, "learning_rate": 8.244457285054144e-06, "loss": 0.244, "step": 709 }, { "epoch": 0.5512422360248447, "grad_norm": 0.6406274495765292, "learning_rate": 8.239815164326907e-06, "loss": 0.2342, "step": 710 }, { "epoch": 0.5520186335403726, "grad_norm": 0.6496338226115757, "learning_rate": 8.235168224918325e-06, "loss": 0.2269, "step": 711 }, { "epoch": 0.5527950310559007, "grad_norm": 0.8154219349416311, "learning_rate": 8.230516473739934e-06, "loss": 0.2729, "step": 712 }, { "epoch": 0.5535714285714286, "grad_norm": 0.916528302897263, "learning_rate": 8.22585991771044e-06, "loss": 0.3466, "step": 713 }, { "epoch": 0.5543478260869565, "grad_norm": 0.8032351078131162, "learning_rate": 8.221198563755683e-06, "loss": 0.2037, "step": 714 }, { "epoch": 0.5551242236024845, "grad_norm": 0.563932086747544, "learning_rate": 8.216532418808647e-06, "loss": 0.2325, "step": 715 }, { "epoch": 0.5559006211180124, "grad_norm": 0.7725840263098726, "learning_rate": 8.211861489809441e-06, "loss": 0.2947, "step": 716 }, { "epoch": 0.5566770186335404, "grad_norm": 0.8927965281807622, "learning_rate": 8.207185783705285e-06, "loss": 0.2218, "step": 717 }, { "epoch": 0.5574534161490683, "grad_norm": 0.7120371919124281, "learning_rate": 8.202505307450509e-06, "loss": 0.273, "step": 718 }, { "epoch": 0.5582298136645962, "grad_norm": 0.5383119061434889, "learning_rate": 8.197820068006537e-06, "loss": 0.235, "step": 719 }, { "epoch": 0.5590062111801242, "grad_norm": 0.538244049355254, "learning_rate": 8.193130072341872e-06, "loss": 0.2073, "step": 720 }, { "epoch": 0.5597826086956522, "grad_norm": 0.9814462939789473, "learning_rate": 8.1884353274321e-06, "loss": 0.3163, "step": 721 }, { "epoch": 0.5605590062111802, "grad_norm": 0.6533966146351818, "learning_rate": 8.183735840259867e-06, "loss": 0.2648, "step": 722 }, { "epoch": 0.5613354037267081, "grad_norm": 0.7116080990442452, "learning_rate": 8.179031617814867e-06, "loss": 0.2685, "step": 723 }, { "epoch": 0.562111801242236, "grad_norm": 0.7913991644170923, "learning_rate": 8.174322667093843e-06, "loss": 0.2378, "step": 724 }, { "epoch": 0.562888198757764, "grad_norm": 1.3321321771679742, "learning_rate": 8.16960899510057e-06, "loss": 0.8284, "step": 725 }, { "epoch": 0.5636645962732919, "grad_norm": 0.634270811595439, "learning_rate": 8.164890608845846e-06, "loss": 0.235, "step": 726 }, { "epoch": 0.5644409937888198, "grad_norm": 0.5298928679154982, "learning_rate": 8.160167515347474e-06, "loss": 0.2758, "step": 727 }, { "epoch": 0.5652173913043478, "grad_norm": 0.5698938638051309, "learning_rate": 8.155439721630265e-06, "loss": 0.2617, "step": 728 }, { "epoch": 0.5659937888198758, "grad_norm": 0.5859218975812209, "learning_rate": 8.150707234726019e-06, "loss": 0.2605, "step": 729 }, { "epoch": 0.5667701863354038, "grad_norm": 0.8146799652273451, "learning_rate": 8.145970061673517e-06, "loss": 0.2674, "step": 730 }, { "epoch": 0.5675465838509317, "grad_norm": 0.6066290222527029, "learning_rate": 8.141228209518506e-06, "loss": 0.3021, "step": 731 }, { "epoch": 0.5683229813664596, "grad_norm": 0.6210973348278607, "learning_rate": 8.136481685313699e-06, "loss": 0.2334, "step": 732 }, { "epoch": 0.5690993788819876, "grad_norm": 0.883257044874417, "learning_rate": 8.131730496118753e-06, "loss": 0.3133, "step": 733 }, { "epoch": 0.5698757763975155, "grad_norm": 0.823970910566647, "learning_rate": 8.126974649000264e-06, "loss": 0.2338, "step": 734 }, { "epoch": 0.5706521739130435, "grad_norm": 0.8919596203142681, "learning_rate": 8.122214151031753e-06, "loss": 0.2951, "step": 735 }, { "epoch": 0.5714285714285714, "grad_norm": 0.7559317752215345, "learning_rate": 8.117449009293668e-06, "loss": 0.231, "step": 736 }, { "epoch": 0.5722049689440993, "grad_norm": 0.7803068601810641, "learning_rate": 8.112679230873355e-06, "loss": 0.2518, "step": 737 }, { "epoch": 0.5729813664596274, "grad_norm": 0.6372822631466126, "learning_rate": 8.107904822865058e-06, "loss": 0.2665, "step": 738 }, { "epoch": 0.5737577639751553, "grad_norm": 0.8389751404267385, "learning_rate": 8.103125792369909e-06, "loss": 0.3964, "step": 739 }, { "epoch": 0.5745341614906833, "grad_norm": 0.5910016441340465, "learning_rate": 8.098342146495913e-06, "loss": 0.3018, "step": 740 }, { "epoch": 0.5753105590062112, "grad_norm": 0.8850220684276294, "learning_rate": 8.09355389235794e-06, "loss": 0.3392, "step": 741 }, { "epoch": 0.5760869565217391, "grad_norm": 0.712469243801041, "learning_rate": 8.088761037077718e-06, "loss": 0.2295, "step": 742 }, { "epoch": 0.5768633540372671, "grad_norm": 0.621921704044139, "learning_rate": 8.08396358778381e-06, "loss": 0.2609, "step": 743 }, { "epoch": 0.577639751552795, "grad_norm": 0.7301147027421087, "learning_rate": 8.07916155161162e-06, "loss": 0.1803, "step": 744 }, { "epoch": 0.578416149068323, "grad_norm": 1.4662030191296285, "learning_rate": 8.07435493570337e-06, "loss": 0.8354, "step": 745 }, { "epoch": 0.5791925465838509, "grad_norm": 0.6489654895041818, "learning_rate": 8.069543747208092e-06, "loss": 0.295, "step": 746 }, { "epoch": 0.5799689440993789, "grad_norm": 1.249540381008089, "learning_rate": 8.064727993281625e-06, "loss": 0.732, "step": 747 }, { "epoch": 0.5807453416149069, "grad_norm": 0.5424265298024268, "learning_rate": 8.059907681086594e-06, "loss": 0.2332, "step": 748 }, { "epoch": 0.5815217391304348, "grad_norm": 0.6739343115647475, "learning_rate": 8.055082817792404e-06, "loss": 0.2434, "step": 749 }, { "epoch": 0.5822981366459627, "grad_norm": 0.7252026493936019, "learning_rate": 8.050253410575228e-06, "loss": 0.2843, "step": 750 }, { "epoch": 0.5830745341614907, "grad_norm": 0.6111500833152194, "learning_rate": 8.045419466618002e-06, "loss": 0.3139, "step": 751 }, { "epoch": 0.5838509316770186, "grad_norm": 0.7180990624455077, "learning_rate": 8.040580993110404e-06, "loss": 0.3116, "step": 752 }, { "epoch": 0.5846273291925466, "grad_norm": 0.5757302260711658, "learning_rate": 8.035737997248851e-06, "loss": 0.2645, "step": 753 }, { "epoch": 0.5854037267080745, "grad_norm": 0.6233360625708801, "learning_rate": 8.030890486236489e-06, "loss": 0.2309, "step": 754 }, { "epoch": 0.5861801242236024, "grad_norm": 0.5810748381493868, "learning_rate": 8.026038467283176e-06, "loss": 0.1597, "step": 755 }, { "epoch": 0.5869565217391305, "grad_norm": 0.719851824140483, "learning_rate": 8.021181947605474e-06, "loss": 0.2671, "step": 756 }, { "epoch": 0.5877329192546584, "grad_norm": 0.7399267113621002, "learning_rate": 8.016320934426642e-06, "loss": 0.31, "step": 757 }, { "epoch": 0.5885093167701864, "grad_norm": 0.8226365657976165, "learning_rate": 8.011455434976622e-06, "loss": 0.2335, "step": 758 }, { "epoch": 0.5892857142857143, "grad_norm": 0.5771880162985638, "learning_rate": 8.00658545649203e-06, "loss": 0.2624, "step": 759 }, { "epoch": 0.5900621118012422, "grad_norm": 0.8108655415216031, "learning_rate": 8.001711006216138e-06, "loss": 0.3109, "step": 760 }, { "epoch": 0.5908385093167702, "grad_norm": 0.7891858959985019, "learning_rate": 7.996832091398874e-06, "loss": 0.2492, "step": 761 }, { "epoch": 0.5916149068322981, "grad_norm": 0.6479600181401016, "learning_rate": 7.991948719296806e-06, "loss": 0.2494, "step": 762 }, { "epoch": 0.592391304347826, "grad_norm": 0.7968292418092983, "learning_rate": 7.98706089717313e-06, "loss": 0.2807, "step": 763 }, { "epoch": 0.593167701863354, "grad_norm": 0.7530713140491678, "learning_rate": 7.98216863229766e-06, "loss": 0.2976, "step": 764 }, { "epoch": 0.593944099378882, "grad_norm": 0.5536675648310918, "learning_rate": 7.977271931946819e-06, "loss": 0.2537, "step": 765 }, { "epoch": 0.59472049689441, "grad_norm": 0.6176001866191071, "learning_rate": 7.972370803403628e-06, "loss": 0.2596, "step": 766 }, { "epoch": 0.5954968944099379, "grad_norm": 0.804507807342622, "learning_rate": 7.967465253957693e-06, "loss": 0.2675, "step": 767 }, { "epoch": 0.5962732919254659, "grad_norm": 0.8546304222787072, "learning_rate": 7.962555290905198e-06, "loss": 0.2606, "step": 768 }, { "epoch": 0.5970496894409938, "grad_norm": 0.7747606699007534, "learning_rate": 7.957640921548884e-06, "loss": 0.3648, "step": 769 }, { "epoch": 0.5978260869565217, "grad_norm": 0.9081291386649974, "learning_rate": 7.952722153198054e-06, "loss": 0.352, "step": 770 }, { "epoch": 0.5986024844720497, "grad_norm": 1.473184472990211, "learning_rate": 7.94779899316855e-06, "loss": 0.7509, "step": 771 }, { "epoch": 0.5993788819875776, "grad_norm": 0.6930270538901716, "learning_rate": 7.942871448782747e-06, "loss": 0.3446, "step": 772 }, { "epoch": 0.6001552795031055, "grad_norm": 0.7597873869021534, "learning_rate": 7.937939527369537e-06, "loss": 0.2969, "step": 773 }, { "epoch": 0.6009316770186336, "grad_norm": 0.8398513134900033, "learning_rate": 7.93300323626433e-06, "loss": 0.303, "step": 774 }, { "epoch": 0.6017080745341615, "grad_norm": 0.7448716286441663, "learning_rate": 7.928062582809028e-06, "loss": 0.3399, "step": 775 }, { "epoch": 0.6024844720496895, "grad_norm": 0.7809819753858184, "learning_rate": 7.923117574352024e-06, "loss": 0.3179, "step": 776 }, { "epoch": 0.6032608695652174, "grad_norm": 0.575259101277246, "learning_rate": 7.918168218248189e-06, "loss": 0.2004, "step": 777 }, { "epoch": 0.6040372670807453, "grad_norm": 0.6500375999804231, "learning_rate": 7.913214521858858e-06, "loss": 0.2461, "step": 778 }, { "epoch": 0.6048136645962733, "grad_norm": 0.7115553553203081, "learning_rate": 7.908256492551825e-06, "loss": 0.3227, "step": 779 }, { "epoch": 0.6055900621118012, "grad_norm": 0.7953203156311331, "learning_rate": 7.903294137701324e-06, "loss": 0.2413, "step": 780 }, { "epoch": 0.6063664596273292, "grad_norm": 0.7008906467083916, "learning_rate": 7.898327464688029e-06, "loss": 0.2643, "step": 781 }, { "epoch": 0.6071428571428571, "grad_norm": 0.6530552449087438, "learning_rate": 7.89335648089903e-06, "loss": 0.2152, "step": 782 }, { "epoch": 0.6079192546583851, "grad_norm": 0.7103956587681162, "learning_rate": 7.88838119372783e-06, "loss": 0.2926, "step": 783 }, { "epoch": 0.6086956521739131, "grad_norm": 0.7559980661665352, "learning_rate": 7.883401610574338e-06, "loss": 0.3304, "step": 784 }, { "epoch": 0.609472049689441, "grad_norm": 0.7461067421518572, "learning_rate": 7.878417738844841e-06, "loss": 0.2502, "step": 785 }, { "epoch": 0.610248447204969, "grad_norm": 0.6240633340549151, "learning_rate": 7.87342958595202e-06, "loss": 0.2381, "step": 786 }, { "epoch": 0.6110248447204969, "grad_norm": 0.5492794987993357, "learning_rate": 7.86843715931491e-06, "loss": 0.229, "step": 787 }, { "epoch": 0.6118012422360248, "grad_norm": 0.5920105112363264, "learning_rate": 7.86344046635891e-06, "loss": 0.1957, "step": 788 }, { "epoch": 0.6125776397515528, "grad_norm": 0.5809900385481488, "learning_rate": 7.858439514515762e-06, "loss": 0.2548, "step": 789 }, { "epoch": 0.6133540372670807, "grad_norm": 0.5963258804668068, "learning_rate": 7.853434311223542e-06, "loss": 0.2584, "step": 790 }, { "epoch": 0.6141304347826086, "grad_norm": 0.7889676244290511, "learning_rate": 7.848424863926648e-06, "loss": 0.7887, "step": 791 }, { "epoch": 0.6149068322981367, "grad_norm": 0.8371644145772603, "learning_rate": 7.843411180075795e-06, "loss": 0.3652, "step": 792 }, { "epoch": 0.6156832298136646, "grad_norm": 0.679873963272173, "learning_rate": 7.838393267127996e-06, "loss": 0.2097, "step": 793 }, { "epoch": 0.6164596273291926, "grad_norm": 0.6904462587835543, "learning_rate": 7.833371132546556e-06, "loss": 0.2607, "step": 794 }, { "epoch": 0.6172360248447205, "grad_norm": 0.7156380162898401, "learning_rate": 7.828344783801055e-06, "loss": 0.288, "step": 795 }, { "epoch": 0.6180124223602484, "grad_norm": 0.8752807210924154, "learning_rate": 7.82331422836734e-06, "loss": 0.3069, "step": 796 }, { "epoch": 0.6187888198757764, "grad_norm": 0.6731871318789172, "learning_rate": 7.818279473727525e-06, "loss": 0.3043, "step": 797 }, { "epoch": 0.6195652173913043, "grad_norm": 0.6015336695665592, "learning_rate": 7.813240527369958e-06, "loss": 0.2657, "step": 798 }, { "epoch": 0.6203416149068323, "grad_norm": 0.9745557700885299, "learning_rate": 7.808197396789228e-06, "loss": 0.2952, "step": 799 }, { "epoch": 0.6211180124223602, "grad_norm": 1.3429364698997956, "learning_rate": 7.803150089486144e-06, "loss": 0.8094, "step": 800 }, { "epoch": 0.6218944099378882, "grad_norm": 0.768258179019345, "learning_rate": 7.798098612967729e-06, "loss": 0.243, "step": 801 }, { "epoch": 0.6226708074534162, "grad_norm": 0.5812747560031819, "learning_rate": 7.793042974747204e-06, "loss": 0.2332, "step": 802 }, { "epoch": 0.6234472049689441, "grad_norm": 0.7769631196821635, "learning_rate": 7.78798318234399e-06, "loss": 0.2846, "step": 803 }, { "epoch": 0.6242236024844721, "grad_norm": 0.936057396843279, "learning_rate": 7.782919243283671e-06, "loss": 0.2928, "step": 804 }, { "epoch": 0.625, "grad_norm": 0.47313038427494636, "learning_rate": 7.777851165098012e-06, "loss": 0.218, "step": 805 }, { "epoch": 0.6257763975155279, "grad_norm": 0.8005080316166346, "learning_rate": 7.772778955324925e-06, "loss": 0.2359, "step": 806 }, { "epoch": 0.6265527950310559, "grad_norm": 0.6243723154334405, "learning_rate": 7.767702621508472e-06, "loss": 0.3162, "step": 807 }, { "epoch": 0.6273291925465838, "grad_norm": 0.6220272622900332, "learning_rate": 7.76262217119885e-06, "loss": 0.2412, "step": 808 }, { "epoch": 0.6281055900621118, "grad_norm": 0.655149511946975, "learning_rate": 7.757537611952373e-06, "loss": 0.2894, "step": 809 }, { "epoch": 0.6288819875776398, "grad_norm": 0.6078112955915205, "learning_rate": 7.75244895133147e-06, "loss": 0.2242, "step": 810 }, { "epoch": 0.6296583850931677, "grad_norm": 0.9411396438345203, "learning_rate": 7.747356196904671e-06, "loss": 0.2816, "step": 811 }, { "epoch": 0.6304347826086957, "grad_norm": 0.7031399511332489, "learning_rate": 7.742259356246594e-06, "loss": 0.2619, "step": 812 }, { "epoch": 0.6312111801242236, "grad_norm": 0.6714746118038507, "learning_rate": 7.737158436937932e-06, "loss": 0.2509, "step": 813 }, { "epoch": 0.6319875776397516, "grad_norm": 0.6763378219419454, "learning_rate": 7.732053446565448e-06, "loss": 0.2877, "step": 814 }, { "epoch": 0.6327639751552795, "grad_norm": 0.8317182153945251, "learning_rate": 7.726944392721958e-06, "loss": 0.2363, "step": 815 }, { "epoch": 0.6335403726708074, "grad_norm": 0.8033609264385366, "learning_rate": 7.721831283006323e-06, "loss": 0.2811, "step": 816 }, { "epoch": 0.6343167701863354, "grad_norm": 1.0125538484584589, "learning_rate": 7.716714125023435e-06, "loss": 0.7466, "step": 817 }, { "epoch": 0.6350931677018633, "grad_norm": 0.8937285978754824, "learning_rate": 7.711592926384206e-06, "loss": 0.2901, "step": 818 }, { "epoch": 0.6358695652173914, "grad_norm": 0.8069792167921794, "learning_rate": 7.706467694705562e-06, "loss": 0.3175, "step": 819 }, { "epoch": 0.6366459627329193, "grad_norm": 0.8268145169002962, "learning_rate": 7.701338437610423e-06, "loss": 0.2323, "step": 820 }, { "epoch": 0.6374223602484472, "grad_norm": 0.9007732140329642, "learning_rate": 7.696205162727699e-06, "loss": 0.2551, "step": 821 }, { "epoch": 0.6381987577639752, "grad_norm": 0.6401371667793756, "learning_rate": 7.691067877692277e-06, "loss": 0.2961, "step": 822 }, { "epoch": 0.6389751552795031, "grad_norm": 0.8667435274273917, "learning_rate": 7.685926590145005e-06, "loss": 0.2928, "step": 823 }, { "epoch": 0.639751552795031, "grad_norm": 0.6125562949328575, "learning_rate": 7.680781307732683e-06, "loss": 0.2656, "step": 824 }, { "epoch": 0.640527950310559, "grad_norm": 1.1634873334507156, "learning_rate": 7.675632038108057e-06, "loss": 0.2569, "step": 825 }, { "epoch": 0.6413043478260869, "grad_norm": 0.8351122564494515, "learning_rate": 7.670478788929803e-06, "loss": 0.3241, "step": 826 }, { "epoch": 0.6420807453416149, "grad_norm": 0.7204121155526788, "learning_rate": 7.665321567862511e-06, "loss": 0.3063, "step": 827 }, { "epoch": 0.6428571428571429, "grad_norm": 0.7259579136339235, "learning_rate": 7.660160382576683e-06, "loss": 0.3291, "step": 828 }, { "epoch": 0.6436335403726708, "grad_norm": 0.8547738056973722, "learning_rate": 7.654995240748719e-06, "loss": 0.2569, "step": 829 }, { "epoch": 0.6444099378881988, "grad_norm": 0.6981114776848291, "learning_rate": 7.649826150060896e-06, "loss": 0.2463, "step": 830 }, { "epoch": 0.6451863354037267, "grad_norm": 0.6815430879905778, "learning_rate": 7.644653118201372e-06, "loss": 0.2885, "step": 831 }, { "epoch": 0.6459627329192547, "grad_norm": 0.9205799105610926, "learning_rate": 7.639476152864163e-06, "loss": 0.2747, "step": 832 }, { "epoch": 0.6467391304347826, "grad_norm": 0.7839991112263055, "learning_rate": 7.634295261749136e-06, "loss": 0.2253, "step": 833 }, { "epoch": 0.6475155279503105, "grad_norm": 0.639101709558868, "learning_rate": 7.6291104525619985e-06, "loss": 0.1997, "step": 834 }, { "epoch": 0.6482919254658385, "grad_norm": 0.600156359407937, "learning_rate": 7.62392173301428e-06, "loss": 0.2268, "step": 835 }, { "epoch": 0.6490683229813664, "grad_norm": 0.5759058640819728, "learning_rate": 7.618729110823334e-06, "loss": 0.3313, "step": 836 }, { "epoch": 0.6498447204968945, "grad_norm": 1.259944028452704, "learning_rate": 7.6135325937123115e-06, "loss": 0.7996, "step": 837 }, { "epoch": 0.6506211180124224, "grad_norm": 1.1583255368995333, "learning_rate": 7.608332189410163e-06, "loss": 0.3292, "step": 838 }, { "epoch": 0.6513975155279503, "grad_norm": 0.5993259323049365, "learning_rate": 7.603127905651617e-06, "loss": 0.276, "step": 839 }, { "epoch": 0.6521739130434783, "grad_norm": 0.685283385780455, "learning_rate": 7.597919750177168e-06, "loss": 0.2565, "step": 840 }, { "epoch": 0.6529503105590062, "grad_norm": 0.6438373326868927, "learning_rate": 7.592707730733077e-06, "loss": 0.2755, "step": 841 }, { "epoch": 0.6537267080745341, "grad_norm": 0.7541491000249798, "learning_rate": 7.587491855071347e-06, "loss": 0.3203, "step": 842 }, { "epoch": 0.6545031055900621, "grad_norm": 0.6074983999628224, "learning_rate": 7.582272130949719e-06, "loss": 0.2731, "step": 843 }, { "epoch": 0.65527950310559, "grad_norm": 0.5306553534999173, "learning_rate": 7.577048566131655e-06, "loss": 0.1974, "step": 844 }, { "epoch": 0.656055900621118, "grad_norm": 0.5125572048573105, "learning_rate": 7.5718211683863315e-06, "loss": 0.2146, "step": 845 }, { "epoch": 0.656832298136646, "grad_norm": 0.6320423787861578, "learning_rate": 7.566589945488625e-06, "loss": 0.2258, "step": 846 }, { "epoch": 0.657608695652174, "grad_norm": 0.8123561993915259, "learning_rate": 7.561354905219102e-06, "loss": 0.2607, "step": 847 }, { "epoch": 0.6583850931677019, "grad_norm": 0.8230806451748813, "learning_rate": 7.556116055364008e-06, "loss": 0.2146, "step": 848 }, { "epoch": 0.6591614906832298, "grad_norm": 0.8026197613822755, "learning_rate": 7.550873403715251e-06, "loss": 0.2991, "step": 849 }, { "epoch": 0.6599378881987578, "grad_norm": 0.7442294199007335, "learning_rate": 7.545626958070394e-06, "loss": 0.2625, "step": 850 }, { "epoch": 0.6607142857142857, "grad_norm": 0.8491113976360859, "learning_rate": 7.540376726232648e-06, "loss": 0.2865, "step": 851 }, { "epoch": 0.6614906832298136, "grad_norm": 0.652200157522898, "learning_rate": 7.5351227160108495e-06, "loss": 0.1772, "step": 852 }, { "epoch": 0.6622670807453416, "grad_norm": 0.8392601834451524, "learning_rate": 7.529864935219459e-06, "loss": 0.2536, "step": 853 }, { "epoch": 0.6630434782608695, "grad_norm": 0.6333477317743353, "learning_rate": 7.524603391678541e-06, "loss": 0.2859, "step": 854 }, { "epoch": 0.6638198757763976, "grad_norm": 0.717657931190261, "learning_rate": 7.519338093213757e-06, "loss": 0.2868, "step": 855 }, { "epoch": 0.6645962732919255, "grad_norm": 0.555781835394527, "learning_rate": 7.51406904765636e-06, "loss": 0.2285, "step": 856 }, { "epoch": 0.6653726708074534, "grad_norm": 0.8049345945018093, "learning_rate": 7.5087962628431665e-06, "loss": 0.2798, "step": 857 }, { "epoch": 0.6661490683229814, "grad_norm": 0.7372283985735612, "learning_rate": 7.503519746616562e-06, "loss": 0.2996, "step": 858 }, { "epoch": 0.6669254658385093, "grad_norm": 0.8462936957276879, "learning_rate": 7.498239506824476e-06, "loss": 0.3346, "step": 859 }, { "epoch": 0.6677018633540373, "grad_norm": 0.8474026211768049, "learning_rate": 7.492955551320381e-06, "loss": 0.2864, "step": 860 }, { "epoch": 0.6684782608695652, "grad_norm": 0.7426579768153135, "learning_rate": 7.487667887963273e-06, "loss": 0.2741, "step": 861 }, { "epoch": 0.6692546583850931, "grad_norm": 0.7058538416048865, "learning_rate": 7.482376524617666e-06, "loss": 0.1926, "step": 862 }, { "epoch": 0.6700310559006211, "grad_norm": 0.8518068072244538, "learning_rate": 7.477081469153572e-06, "loss": 0.2534, "step": 863 }, { "epoch": 0.6708074534161491, "grad_norm": 0.6078541656170362, "learning_rate": 7.4717827294464996e-06, "loss": 0.2672, "step": 864 }, { "epoch": 0.671583850931677, "grad_norm": 0.8460032034810582, "learning_rate": 7.466480313377433e-06, "loss": 0.3411, "step": 865 }, { "epoch": 0.672360248447205, "grad_norm": 0.5381498430717333, "learning_rate": 7.461174228832828e-06, "loss": 0.3271, "step": 866 }, { "epoch": 0.6731366459627329, "grad_norm": 0.8022837102860096, "learning_rate": 7.455864483704594e-06, "loss": 0.2676, "step": 867 }, { "epoch": 0.6739130434782609, "grad_norm": 0.5336629405654473, "learning_rate": 7.450551085890087e-06, "loss": 0.2427, "step": 868 }, { "epoch": 0.6746894409937888, "grad_norm": 0.7469690139363668, "learning_rate": 7.445234043292094e-06, "loss": 0.2759, "step": 869 }, { "epoch": 0.6754658385093167, "grad_norm": 0.6685213951170458, "learning_rate": 7.43991336381882e-06, "loss": 0.268, "step": 870 }, { "epoch": 0.6762422360248447, "grad_norm": 0.8454453054299584, "learning_rate": 7.434589055383889e-06, "loss": 0.2507, "step": 871 }, { "epoch": 0.6770186335403726, "grad_norm": 0.5662460887694007, "learning_rate": 7.4292611259063105e-06, "loss": 0.2184, "step": 872 }, { "epoch": 0.6777950310559007, "grad_norm": 0.8390285101032285, "learning_rate": 7.42392958331049e-06, "loss": 0.2718, "step": 873 }, { "epoch": 0.6785714285714286, "grad_norm": 0.719046128349501, "learning_rate": 7.4185944355261996e-06, "loss": 0.2998, "step": 874 }, { "epoch": 0.6793478260869565, "grad_norm": 0.6220319771997246, "learning_rate": 7.413255690488578e-06, "loss": 0.2843, "step": 875 }, { "epoch": 0.6801242236024845, "grad_norm": 0.6132840327755115, "learning_rate": 7.40791335613811e-06, "loss": 0.2069, "step": 876 }, { "epoch": 0.6809006211180124, "grad_norm": 0.6686521098199149, "learning_rate": 7.402567440420625e-06, "loss": 0.2909, "step": 877 }, { "epoch": 0.6816770186335404, "grad_norm": 0.6866179936254955, "learning_rate": 7.397217951287272e-06, "loss": 0.2457, "step": 878 }, { "epoch": 0.6824534161490683, "grad_norm": 0.7561963072872507, "learning_rate": 7.391864896694521e-06, "loss": 0.2293, "step": 879 }, { "epoch": 0.6832298136645962, "grad_norm": 0.7486975267053527, "learning_rate": 7.3865082846041415e-06, "loss": 0.2604, "step": 880 }, { "epoch": 0.6840062111801242, "grad_norm": 0.9464026767732586, "learning_rate": 7.381148122983194e-06, "loss": 0.2309, "step": 881 }, { "epoch": 0.6847826086956522, "grad_norm": 0.5765204634339792, "learning_rate": 7.375784419804018e-06, "loss": 0.255, "step": 882 }, { "epoch": 0.6855590062111802, "grad_norm": 0.834003649910673, "learning_rate": 7.370417183044227e-06, "loss": 0.3398, "step": 883 }, { "epoch": 0.6863354037267081, "grad_norm": 0.7852251464178506, "learning_rate": 7.365046420686681e-06, "loss": 0.2562, "step": 884 }, { "epoch": 0.687111801242236, "grad_norm": 0.9539600812090855, "learning_rate": 7.359672140719489e-06, "loss": 0.2434, "step": 885 }, { "epoch": 0.687888198757764, "grad_norm": 0.7997012170306546, "learning_rate": 7.354294351135989e-06, "loss": 0.2411, "step": 886 }, { "epoch": 0.6886645962732919, "grad_norm": 0.7612592610254438, "learning_rate": 7.348913059934742e-06, "loss": 0.2285, "step": 887 }, { "epoch": 0.6894409937888198, "grad_norm": 0.7356603003171129, "learning_rate": 7.343528275119515e-06, "loss": 0.3087, "step": 888 }, { "epoch": 0.6902173913043478, "grad_norm": 0.6992926120822256, "learning_rate": 7.338140004699272e-06, "loss": 0.2784, "step": 889 }, { "epoch": 0.6909937888198758, "grad_norm": 0.7229546588226432, "learning_rate": 7.33274825668816e-06, "loss": 0.2479, "step": 890 }, { "epoch": 0.6917701863354038, "grad_norm": 0.8068357117119884, "learning_rate": 7.327353039105501e-06, "loss": 0.2351, "step": 891 }, { "epoch": 0.6925465838509317, "grad_norm": 0.5787456468869818, "learning_rate": 7.321954359975776e-06, "loss": 0.1887, "step": 892 }, { "epoch": 0.6933229813664596, "grad_norm": 0.7022474409684327, "learning_rate": 7.316552227328612e-06, "loss": 0.2102, "step": 893 }, { "epoch": 0.6940993788819876, "grad_norm": 0.8104972908196963, "learning_rate": 7.311146649198777e-06, "loss": 0.2174, "step": 894 }, { "epoch": 0.6948757763975155, "grad_norm": 0.7075124683103852, "learning_rate": 7.305737633626159e-06, "loss": 0.2713, "step": 895 }, { "epoch": 0.6956521739130435, "grad_norm": 0.6793756288139935, "learning_rate": 7.300325188655762e-06, "loss": 0.2013, "step": 896 }, { "epoch": 0.6964285714285714, "grad_norm": 0.6318294278037246, "learning_rate": 7.294909322337689e-06, "loss": 0.3001, "step": 897 }, { "epoch": 0.6972049689440993, "grad_norm": 0.6443312748230396, "learning_rate": 7.289490042727134e-06, "loss": 0.2579, "step": 898 }, { "epoch": 0.6979813664596274, "grad_norm": 0.5527155833807017, "learning_rate": 7.284067357884364e-06, "loss": 0.2707, "step": 899 }, { "epoch": 0.6987577639751553, "grad_norm": 0.6939204259446968, "learning_rate": 7.278641275874714e-06, "loss": 0.3006, "step": 900 }, { "epoch": 0.6995341614906833, "grad_norm": 0.7463655845440587, "learning_rate": 7.27321180476857e-06, "loss": 0.2627, "step": 901 }, { "epoch": 0.7003105590062112, "grad_norm": 0.7776266263641919, "learning_rate": 7.267778952641358e-06, "loss": 0.3175, "step": 902 }, { "epoch": 0.7010869565217391, "grad_norm": 0.738223327729549, "learning_rate": 7.262342727573536e-06, "loss": 0.3226, "step": 903 }, { "epoch": 0.7018633540372671, "grad_norm": 0.7379259096735798, "learning_rate": 7.256903137650575e-06, "loss": 0.2401, "step": 904 }, { "epoch": 0.702639751552795, "grad_norm": 0.6412176743025649, "learning_rate": 7.2514601909629525e-06, "loss": 0.2637, "step": 905 }, { "epoch": 0.703416149068323, "grad_norm": 0.782059661304405, "learning_rate": 7.2460138956061375e-06, "loss": 0.2103, "step": 906 }, { "epoch": 0.7041925465838509, "grad_norm": 0.7824763987178509, "learning_rate": 7.2405642596805815e-06, "loss": 0.3104, "step": 907 }, { "epoch": 0.7049689440993789, "grad_norm": 0.5894288462245859, "learning_rate": 7.235111291291702e-06, "loss": 0.2409, "step": 908 }, { "epoch": 0.7057453416149069, "grad_norm": 0.7145969027280433, "learning_rate": 7.229654998549875e-06, "loss": 0.2589, "step": 909 }, { "epoch": 0.7065217391304348, "grad_norm": 0.8147502420845835, "learning_rate": 7.224195389570422e-06, "loss": 0.2804, "step": 910 }, { "epoch": 0.7072981366459627, "grad_norm": 0.5525347511392322, "learning_rate": 7.218732472473593e-06, "loss": 0.2353, "step": 911 }, { "epoch": 0.7080745341614907, "grad_norm": 0.9987564415869262, "learning_rate": 7.21326625538456e-06, "loss": 0.2953, "step": 912 }, { "epoch": 0.7088509316770186, "grad_norm": 0.6128133169459449, "learning_rate": 7.207796746433405e-06, "loss": 0.2938, "step": 913 }, { "epoch": 0.7096273291925466, "grad_norm": 1.1093498522183207, "learning_rate": 7.202323953755106e-06, "loss": 0.8007, "step": 914 }, { "epoch": 0.7104037267080745, "grad_norm": 0.6927099339754995, "learning_rate": 7.196847885489519e-06, "loss": 0.2482, "step": 915 }, { "epoch": 0.7111801242236024, "grad_norm": 0.7554905743266254, "learning_rate": 7.191368549781382e-06, "loss": 0.3079, "step": 916 }, { "epoch": 0.7119565217391305, "grad_norm": 0.9156735625757821, "learning_rate": 7.185885954780283e-06, "loss": 0.2613, "step": 917 }, { "epoch": 0.7127329192546584, "grad_norm": 0.8244903980565818, "learning_rate": 7.180400108640665e-06, "loss": 0.2568, "step": 918 }, { "epoch": 0.7135093167701864, "grad_norm": 0.918344204837153, "learning_rate": 7.174911019521801e-06, "loss": 0.2815, "step": 919 }, { "epoch": 0.7142857142857143, "grad_norm": 0.6519622623997611, "learning_rate": 7.169418695587791e-06, "loss": 0.2449, "step": 920 }, { "epoch": 0.7150621118012422, "grad_norm": 0.6629184154068417, "learning_rate": 7.163923145007547e-06, "loss": 0.2059, "step": 921 }, { "epoch": 0.7158385093167702, "grad_norm": 0.7082930343196882, "learning_rate": 7.158424375954775e-06, "loss": 0.2495, "step": 922 }, { "epoch": 0.7166149068322981, "grad_norm": 0.9235853971667726, "learning_rate": 7.1529223966079715e-06, "loss": 0.3452, "step": 923 }, { "epoch": 0.717391304347826, "grad_norm": 0.6990413887608579, "learning_rate": 7.147417215150411e-06, "loss": 0.2448, "step": 924 }, { "epoch": 0.718167701863354, "grad_norm": 0.516881957424605, "learning_rate": 7.141908839770123e-06, "loss": 0.2536, "step": 925 }, { "epoch": 0.718944099378882, "grad_norm": 0.7116993195404256, "learning_rate": 7.1363972786598925e-06, "loss": 0.3011, "step": 926 }, { "epoch": 0.71972049689441, "grad_norm": 0.7439062828304455, "learning_rate": 7.1308825400172434e-06, "loss": 0.2514, "step": 927 }, { "epoch": 0.7204968944099379, "grad_norm": 1.233166255566092, "learning_rate": 7.125364632044423e-06, "loss": 0.2668, "step": 928 }, { "epoch": 0.7212732919254659, "grad_norm": 0.6712130126778718, "learning_rate": 7.119843562948392e-06, "loss": 0.2594, "step": 929 }, { "epoch": 0.7220496894409938, "grad_norm": 0.7979508209572215, "learning_rate": 7.1143193409408165e-06, "loss": 0.3168, "step": 930 }, { "epoch": 0.7228260869565217, "grad_norm": 0.7062940029368757, "learning_rate": 7.108791974238047e-06, "loss": 0.2758, "step": 931 }, { "epoch": 0.7236024844720497, "grad_norm": 0.9272061901157742, "learning_rate": 7.103261471061116e-06, "loss": 0.255, "step": 932 }, { "epoch": 0.7243788819875776, "grad_norm": 0.6724549530187803, "learning_rate": 7.097727839635717e-06, "loss": 0.2224, "step": 933 }, { "epoch": 0.7251552795031055, "grad_norm": 0.8889475047699679, "learning_rate": 7.092191088192201e-06, "loss": 0.2453, "step": 934 }, { "epoch": 0.7259316770186336, "grad_norm": 0.8192107459956408, "learning_rate": 7.0866512249655535e-06, "loss": 0.3083, "step": 935 }, { "epoch": 0.7267080745341615, "grad_norm": 0.8137458693318159, "learning_rate": 7.0811082581953935e-06, "loss": 0.2562, "step": 936 }, { "epoch": 0.7274844720496895, "grad_norm": 0.7369036865128763, "learning_rate": 7.075562196125952e-06, "loss": 0.2567, "step": 937 }, { "epoch": 0.7282608695652174, "grad_norm": 0.7103666792391728, "learning_rate": 7.070013047006068e-06, "loss": 0.2575, "step": 938 }, { "epoch": 0.7290372670807453, "grad_norm": 0.6640586035494042, "learning_rate": 7.064460819089168e-06, "loss": 0.2266, "step": 939 }, { "epoch": 0.7298136645962733, "grad_norm": 0.7451604667403403, "learning_rate": 7.058905520633259e-06, "loss": 0.2835, "step": 940 }, { "epoch": 0.7305900621118012, "grad_norm": 0.7958827084275032, "learning_rate": 7.053347159900915e-06, "loss": 0.2873, "step": 941 }, { "epoch": 0.7313664596273292, "grad_norm": 0.7347941185746376, "learning_rate": 7.047785745159266e-06, "loss": 0.276, "step": 942 }, { "epoch": 0.7321428571428571, "grad_norm": 0.7823560954953805, "learning_rate": 7.042221284679982e-06, "loss": 0.2762, "step": 943 }, { "epoch": 0.7329192546583851, "grad_norm": 0.8100577840679859, "learning_rate": 7.036653786739264e-06, "loss": 0.2383, "step": 944 }, { "epoch": 0.7336956521739131, "grad_norm": 0.7197736305765283, "learning_rate": 7.0310832596178325e-06, "loss": 0.2688, "step": 945 }, { "epoch": 0.734472049689441, "grad_norm": 0.6623297317459618, "learning_rate": 7.02550971160091e-06, "loss": 0.3053, "step": 946 }, { "epoch": 0.735248447204969, "grad_norm": 0.7518829377474519, "learning_rate": 7.0199331509782135e-06, "loss": 0.2277, "step": 947 }, { "epoch": 0.7360248447204969, "grad_norm": 0.8998588284893294, "learning_rate": 7.0143535860439425e-06, "loss": 0.2774, "step": 948 }, { "epoch": 0.7368012422360248, "grad_norm": 0.7520318021583489, "learning_rate": 7.008771025096766e-06, "loss": 0.2859, "step": 949 }, { "epoch": 0.7375776397515528, "grad_norm": 0.5031125105715735, "learning_rate": 7.0031854764398e-06, "loss": 0.2455, "step": 950 }, { "epoch": 0.7383540372670807, "grad_norm": 0.7423930982817352, "learning_rate": 6.997596948380615e-06, "loss": 0.3193, "step": 951 }, { "epoch": 0.7391304347826086, "grad_norm": 0.9181012342859477, "learning_rate": 6.9920054492312086e-06, "loss": 0.2487, "step": 952 }, { "epoch": 0.7399068322981367, "grad_norm": 0.5240163960368798, "learning_rate": 6.9864109873079955e-06, "loss": 0.2207, "step": 953 }, { "epoch": 0.7406832298136646, "grad_norm": 0.630336881847136, "learning_rate": 6.9808135709318015e-06, "loss": 0.2082, "step": 954 }, { "epoch": 0.7414596273291926, "grad_norm": 0.8336357732139943, "learning_rate": 6.9752132084278425e-06, "loss": 0.2885, "step": 955 }, { "epoch": 0.7422360248447205, "grad_norm": 0.7644159255171679, "learning_rate": 6.9696099081257176e-06, "loss": 0.2837, "step": 956 }, { "epoch": 0.7430124223602484, "grad_norm": 0.7293198193330719, "learning_rate": 6.964003678359396e-06, "loss": 0.2715, "step": 957 }, { "epoch": 0.7437888198757764, "grad_norm": 0.48712701530186675, "learning_rate": 6.958394527467204e-06, "loss": 0.1836, "step": 958 }, { "epoch": 0.7445652173913043, "grad_norm": 0.837658012693797, "learning_rate": 6.9527824637918125e-06, "loss": 0.3034, "step": 959 }, { "epoch": 0.7453416149068323, "grad_norm": 0.9349536049767194, "learning_rate": 6.947167495680224e-06, "loss": 0.2834, "step": 960 }, { "epoch": 0.7461180124223602, "grad_norm": 0.7244637117503869, "learning_rate": 6.9415496314837605e-06, "loss": 0.2397, "step": 961 }, { "epoch": 0.7468944099378882, "grad_norm": 0.9235627659910602, "learning_rate": 6.935928879558052e-06, "loss": 0.28, "step": 962 }, { "epoch": 0.7476708074534162, "grad_norm": 0.8076990766008035, "learning_rate": 6.930305248263025e-06, "loss": 0.2306, "step": 963 }, { "epoch": 0.7484472049689441, "grad_norm": 0.7215782546224564, "learning_rate": 6.9246787459628895e-06, "loss": 0.3317, "step": 964 }, { "epoch": 0.7492236024844721, "grad_norm": 0.686891447886399, "learning_rate": 6.919049381026119e-06, "loss": 0.233, "step": 965 }, { "epoch": 0.75, "grad_norm": 0.7962332812740949, "learning_rate": 6.913417161825449e-06, "loss": 0.2453, "step": 966 }, { "epoch": 0.7507763975155279, "grad_norm": 0.6162259331622334, "learning_rate": 6.907782096737865e-06, "loss": 0.188, "step": 967 }, { "epoch": 0.7515527950310559, "grad_norm": 0.5767541587779196, "learning_rate": 6.902144194144576e-06, "loss": 0.2557, "step": 968 }, { "epoch": 0.7523291925465838, "grad_norm": 0.5392215532146856, "learning_rate": 6.89650346243102e-06, "loss": 0.2258, "step": 969 }, { "epoch": 0.7531055900621118, "grad_norm": 0.7298588146003028, "learning_rate": 6.890859909986835e-06, "loss": 0.2502, "step": 970 }, { "epoch": 0.7538819875776398, "grad_norm": 0.7661018449033149, "learning_rate": 6.885213545205859e-06, "loss": 0.2761, "step": 971 }, { "epoch": 0.7546583850931677, "grad_norm": 0.7467911264556004, "learning_rate": 6.8795643764861144e-06, "loss": 0.2155, "step": 972 }, { "epoch": 0.7554347826086957, "grad_norm": 0.8153392977971265, "learning_rate": 6.873912412229788e-06, "loss": 0.2919, "step": 973 }, { "epoch": 0.7562111801242236, "grad_norm": 0.5493008005105624, "learning_rate": 6.868257660843234e-06, "loss": 0.1843, "step": 974 }, { "epoch": 0.7569875776397516, "grad_norm": 0.8559647068486819, "learning_rate": 6.862600130736939e-06, "loss": 0.3011, "step": 975 }, { "epoch": 0.7577639751552795, "grad_norm": 0.495067223236983, "learning_rate": 6.8569398303255345e-06, "loss": 0.2254, "step": 976 }, { "epoch": 0.7585403726708074, "grad_norm": 0.5629601024587605, "learning_rate": 6.851276768027764e-06, "loss": 0.1709, "step": 977 }, { "epoch": 0.7593167701863354, "grad_norm": 0.8064966351816911, "learning_rate": 6.845610952266486e-06, "loss": 0.3637, "step": 978 }, { "epoch": 0.7600931677018633, "grad_norm": 0.808367456476702, "learning_rate": 6.839942391468648e-06, "loss": 0.2691, "step": 979 }, { "epoch": 0.7608695652173914, "grad_norm": 0.8378295066968838, "learning_rate": 6.834271094065284e-06, "loss": 0.2904, "step": 980 }, { "epoch": 0.7616459627329193, "grad_norm": 0.8329500473816263, "learning_rate": 6.828597068491495e-06, "loss": 0.3288, "step": 981 }, { "epoch": 0.7624223602484472, "grad_norm": 0.6627798673180002, "learning_rate": 6.822920323186445e-06, "loss": 0.2766, "step": 982 }, { "epoch": 0.7631987577639752, "grad_norm": 0.7134376173320647, "learning_rate": 6.817240866593338e-06, "loss": 0.2671, "step": 983 }, { "epoch": 0.7639751552795031, "grad_norm": 0.8832547342295445, "learning_rate": 6.811558707159414e-06, "loss": 0.2656, "step": 984 }, { "epoch": 0.764751552795031, "grad_norm": 0.7741634090503544, "learning_rate": 6.805873853335929e-06, "loss": 0.2779, "step": 985 }, { "epoch": 0.765527950310559, "grad_norm": 0.5872764506539936, "learning_rate": 6.800186313578152e-06, "loss": 0.2281, "step": 986 }, { "epoch": 0.7663043478260869, "grad_norm": 0.7605636586036619, "learning_rate": 6.794496096345341e-06, "loss": 0.3025, "step": 987 }, { "epoch": 0.7670807453416149, "grad_norm": 0.7469233210800624, "learning_rate": 6.7888032101007416e-06, "loss": 0.2848, "step": 988 }, { "epoch": 0.7678571428571429, "grad_norm": 0.6963736434753417, "learning_rate": 6.783107663311566e-06, "loss": 0.2232, "step": 989 }, { "epoch": 0.7686335403726708, "grad_norm": 0.7271578972906806, "learning_rate": 6.777409464448983e-06, "loss": 0.3011, "step": 990 }, { "epoch": 0.7694099378881988, "grad_norm": 0.9300331815701023, "learning_rate": 6.7717086219881085e-06, "loss": 0.3306, "step": 991 }, { "epoch": 0.7701863354037267, "grad_norm": 1.0986943123360733, "learning_rate": 6.76600514440799e-06, "loss": 0.3144, "step": 992 }, { "epoch": 0.7709627329192547, "grad_norm": 0.6724054176333542, "learning_rate": 6.7602990401915914e-06, "loss": 0.2331, "step": 993 }, { "epoch": 0.7717391304347826, "grad_norm": 0.6335666249410896, "learning_rate": 6.754590317825785e-06, "loss": 0.2345, "step": 994 }, { "epoch": 0.7725155279503105, "grad_norm": 0.8183251222293396, "learning_rate": 6.748878985801338e-06, "loss": 0.2303, "step": 995 }, { "epoch": 0.7732919254658385, "grad_norm": 1.0738182591214844, "learning_rate": 6.743165052612899e-06, "loss": 0.7789, "step": 996 }, { "epoch": 0.7740683229813664, "grad_norm": 0.5824225694232393, "learning_rate": 6.7374485267589815e-06, "loss": 0.1415, "step": 997 }, { "epoch": 0.7748447204968945, "grad_norm": 0.6716211374039279, "learning_rate": 6.731729416741961e-06, "loss": 0.2657, "step": 998 }, { "epoch": 0.7756211180124224, "grad_norm": 0.9822836750996523, "learning_rate": 6.726007731068056e-06, "loss": 0.3393, "step": 999 }, { "epoch": 0.7763975155279503, "grad_norm": 0.7155859834052613, "learning_rate": 6.72028347824731e-06, "loss": 0.2439, "step": 1000 }, { "epoch": 0.7763975155279503, "eval_loss": 0.3099202513694763, "eval_runtime": 3.9751, "eval_samples_per_second": 13.333, "eval_steps_per_second": 3.522, "step": 1000 }, { "epoch": 0.7771739130434783, "grad_norm": 0.6698645805789248, "learning_rate": 6.7145566667935904e-06, "loss": 0.2608, "step": 1001 }, { "epoch": 0.7779503105590062, "grad_norm": 0.7691282699341503, "learning_rate": 6.708827305224566e-06, "loss": 0.2911, "step": 1002 }, { "epoch": 0.7787267080745341, "grad_norm": 1.0129856184427477, "learning_rate": 6.703095402061702e-06, "loss": 0.275, "step": 1003 }, { "epoch": 0.7795031055900621, "grad_norm": 0.5353861100912237, "learning_rate": 6.697360965830244e-06, "loss": 0.1794, "step": 1004 }, { "epoch": 0.78027950310559, "grad_norm": 0.8221454190599102, "learning_rate": 6.691624005059202e-06, "loss": 0.229, "step": 1005 }, { "epoch": 0.781055900621118, "grad_norm": 0.6147311124427502, "learning_rate": 6.6858845282813416e-06, "loss": 0.2594, "step": 1006 }, { "epoch": 0.781832298136646, "grad_norm": 0.9095140210996556, "learning_rate": 6.6801425440331715e-06, "loss": 0.3235, "step": 1007 }, { "epoch": 0.782608695652174, "grad_norm": 0.7785439363040415, "learning_rate": 6.674398060854931e-06, "loss": 0.3264, "step": 1008 }, { "epoch": 0.7833850931677019, "grad_norm": 0.7103585971423244, "learning_rate": 6.668651087290575e-06, "loss": 0.2623, "step": 1009 }, { "epoch": 0.7841614906832298, "grad_norm": 0.6065732004369001, "learning_rate": 6.662901631887761e-06, "loss": 0.2224, "step": 1010 }, { "epoch": 0.7849378881987578, "grad_norm": 1.055082872628148, "learning_rate": 6.657149703197839e-06, "loss": 0.2446, "step": 1011 }, { "epoch": 0.7857142857142857, "grad_norm": 0.8900123110888646, "learning_rate": 6.651395309775837e-06, "loss": 0.2934, "step": 1012 }, { "epoch": 0.7864906832298136, "grad_norm": 0.8417198573532227, "learning_rate": 6.64563846018045e-06, "loss": 0.3044, "step": 1013 }, { "epoch": 0.7872670807453416, "grad_norm": 0.7181115311345895, "learning_rate": 6.639879162974027e-06, "loss": 0.274, "step": 1014 }, { "epoch": 0.7880434782608695, "grad_norm": 0.6287439696789474, "learning_rate": 6.634117426722556e-06, "loss": 0.1928, "step": 1015 }, { "epoch": 0.7888198757763976, "grad_norm": 0.7510185509750477, "learning_rate": 6.62835325999565e-06, "loss": 0.2806, "step": 1016 }, { "epoch": 0.7895962732919255, "grad_norm": 0.6582490591218132, "learning_rate": 6.622586671366542e-06, "loss": 0.2536, "step": 1017 }, { "epoch": 0.7903726708074534, "grad_norm": 0.7588304511178368, "learning_rate": 6.616817669412066e-06, "loss": 0.2764, "step": 1018 }, { "epoch": 0.7911490683229814, "grad_norm": 0.6891268956813951, "learning_rate": 6.611046262712643e-06, "loss": 0.2423, "step": 1019 }, { "epoch": 0.7919254658385093, "grad_norm": 0.7824683623299926, "learning_rate": 6.6052724598522696e-06, "loss": 0.2798, "step": 1020 }, { "epoch": 0.7927018633540373, "grad_norm": 0.9655752586955633, "learning_rate": 6.5994962694185105e-06, "loss": 0.2823, "step": 1021 }, { "epoch": 0.7934782608695652, "grad_norm": 0.8427700994190502, "learning_rate": 6.59371770000248e-06, "loss": 0.304, "step": 1022 }, { "epoch": 0.7942546583850931, "grad_norm": 1.2168891565027191, "learning_rate": 6.5879367601988296e-06, "loss": 0.8959, "step": 1023 }, { "epoch": 0.7950310559006211, "grad_norm": 0.7854694124785142, "learning_rate": 6.582153458605738e-06, "loss": 0.254, "step": 1024 }, { "epoch": 0.7958074534161491, "grad_norm": 0.637738544260162, "learning_rate": 6.576367803824895e-06, "loss": 0.2551, "step": 1025 }, { "epoch": 0.796583850931677, "grad_norm": 1.1517257496793396, "learning_rate": 6.570579804461492e-06, "loss": 0.8139, "step": 1026 }, { "epoch": 0.797360248447205, "grad_norm": 0.6607782989565202, "learning_rate": 6.5647894691242056e-06, "loss": 0.2548, "step": 1027 }, { "epoch": 0.7981366459627329, "grad_norm": 0.8151294310628485, "learning_rate": 6.558996806425188e-06, "loss": 0.2817, "step": 1028 }, { "epoch": 0.7989130434782609, "grad_norm": 0.8292028746061968, "learning_rate": 6.5532018249800544e-06, "loss": 0.2232, "step": 1029 }, { "epoch": 0.7996894409937888, "grad_norm": 0.5885232254402075, "learning_rate": 6.5474045334078654e-06, "loss": 0.2451, "step": 1030 }, { "epoch": 0.8004658385093167, "grad_norm": 0.7189756815071324, "learning_rate": 6.541604940331119e-06, "loss": 0.2613, "step": 1031 }, { "epoch": 0.8012422360248447, "grad_norm": 0.6135247725841431, "learning_rate": 6.5358030543757375e-06, "loss": 0.2776, "step": 1032 }, { "epoch": 0.8020186335403726, "grad_norm": 0.9369401376700838, "learning_rate": 6.529998884171054e-06, "loss": 0.3514, "step": 1033 }, { "epoch": 0.8027950310559007, "grad_norm": 0.9067673661692961, "learning_rate": 6.524192438349794e-06, "loss": 0.2652, "step": 1034 }, { "epoch": 0.8035714285714286, "grad_norm": 0.6295986621261835, "learning_rate": 6.518383725548074e-06, "loss": 0.2743, "step": 1035 }, { "epoch": 0.8043478260869565, "grad_norm": 0.9008152518728403, "learning_rate": 6.51257275440538e-06, "loss": 0.6901, "step": 1036 }, { "epoch": 0.8051242236024845, "grad_norm": 0.5709392492667682, "learning_rate": 6.506759533564554e-06, "loss": 0.2555, "step": 1037 }, { "epoch": 0.8059006211180124, "grad_norm": 0.9752423372679726, "learning_rate": 6.500944071671789e-06, "loss": 0.2704, "step": 1038 }, { "epoch": 0.8066770186335404, "grad_norm": 0.5919929135172691, "learning_rate": 6.495126377376606e-06, "loss": 0.2322, "step": 1039 }, { "epoch": 0.8074534161490683, "grad_norm": 0.6605793272627524, "learning_rate": 6.489306459331851e-06, "loss": 0.2238, "step": 1040 }, { "epoch": 0.8082298136645962, "grad_norm": 0.7548145495344198, "learning_rate": 6.483484326193674e-06, "loss": 0.2628, "step": 1041 }, { "epoch": 0.8090062111801242, "grad_norm": 0.6800497626787422, "learning_rate": 6.477659986621522e-06, "loss": 0.2753, "step": 1042 }, { "epoch": 0.8097826086956522, "grad_norm": 0.8731586460124315, "learning_rate": 6.47183344927812e-06, "loss": 0.392, "step": 1043 }, { "epoch": 0.8105590062111802, "grad_norm": 0.6804067077261498, "learning_rate": 6.46600472282947e-06, "loss": 0.2753, "step": 1044 }, { "epoch": 0.8113354037267081, "grad_norm": 0.7065009537662466, "learning_rate": 6.460173815944819e-06, "loss": 0.2947, "step": 1045 }, { "epoch": 0.812111801242236, "grad_norm": 0.7409917256603874, "learning_rate": 6.454340737296665e-06, "loss": 0.3096, "step": 1046 }, { "epoch": 0.812888198757764, "grad_norm": 0.8129356993945802, "learning_rate": 6.448505495560733e-06, "loss": 0.2741, "step": 1047 }, { "epoch": 0.8136645962732919, "grad_norm": 0.875072225441524, "learning_rate": 6.442668099415967e-06, "loss": 0.424, "step": 1048 }, { "epoch": 0.8144409937888198, "grad_norm": 0.7235067560080553, "learning_rate": 6.436828557544511e-06, "loss": 0.3089, "step": 1049 }, { "epoch": 0.8152173913043478, "grad_norm": 0.5826583559381654, "learning_rate": 6.430986878631708e-06, "loss": 0.1895, "step": 1050 }, { "epoch": 0.8159937888198758, "grad_norm": 0.7259891203543739, "learning_rate": 6.4251430713660705e-06, "loss": 0.341, "step": 1051 }, { "epoch": 0.8167701863354038, "grad_norm": 0.803112667450657, "learning_rate": 6.4192971444392835e-06, "loss": 0.3973, "step": 1052 }, { "epoch": 0.8175465838509317, "grad_norm": 0.8436269505376903, "learning_rate": 6.4134491065461824e-06, "loss": 0.2774, "step": 1053 }, { "epoch": 0.8183229813664596, "grad_norm": 1.0252727629680451, "learning_rate": 6.407598966384743e-06, "loss": 0.242, "step": 1054 }, { "epoch": 0.8190993788819876, "grad_norm": 0.7510651446670793, "learning_rate": 6.401746732656065e-06, "loss": 0.2755, "step": 1055 }, { "epoch": 0.8198757763975155, "grad_norm": 0.6341981182172612, "learning_rate": 6.395892414064363e-06, "loss": 0.2489, "step": 1056 }, { "epoch": 0.8206521739130435, "grad_norm": 0.5748941653037674, "learning_rate": 6.3900360193169565e-06, "loss": 0.1987, "step": 1057 }, { "epoch": 0.8214285714285714, "grad_norm": 0.5685151476061286, "learning_rate": 6.384177557124247e-06, "loss": 0.2066, "step": 1058 }, { "epoch": 0.8222049689440993, "grad_norm": 0.5840450853502518, "learning_rate": 6.378317036199717e-06, "loss": 0.2828, "step": 1059 }, { "epoch": 0.8229813664596274, "grad_norm": 0.8400038966456961, "learning_rate": 6.372454465259905e-06, "loss": 0.1993, "step": 1060 }, { "epoch": 0.8237577639751553, "grad_norm": 0.7698977607886737, "learning_rate": 6.366589853024403e-06, "loss": 0.2475, "step": 1061 }, { "epoch": 0.8245341614906833, "grad_norm": 0.5789182965615954, "learning_rate": 6.360723208215838e-06, "loss": 0.2947, "step": 1062 }, { "epoch": 0.8253105590062112, "grad_norm": 0.783336221861108, "learning_rate": 6.354854539559856e-06, "loss": 0.7543, "step": 1063 }, { "epoch": 0.8260869565217391, "grad_norm": 0.8205937422323907, "learning_rate": 6.348983855785122e-06, "loss": 0.2081, "step": 1064 }, { "epoch": 0.8268633540372671, "grad_norm": 0.7306048326711619, "learning_rate": 6.34311116562329e-06, "loss": 0.3278, "step": 1065 }, { "epoch": 0.827639751552795, "grad_norm": 0.7861792684982135, "learning_rate": 6.3372364778090005e-06, "loss": 0.3032, "step": 1066 }, { "epoch": 0.828416149068323, "grad_norm": 0.5826415383225536, "learning_rate": 6.331359801079866e-06, "loss": 0.2336, "step": 1067 }, { "epoch": 0.8291925465838509, "grad_norm": 0.9026753518946457, "learning_rate": 6.325481144176458e-06, "loss": 0.2662, "step": 1068 }, { "epoch": 0.8299689440993789, "grad_norm": 0.7321918482732019, "learning_rate": 6.319600515842292e-06, "loss": 0.2536, "step": 1069 }, { "epoch": 0.8307453416149069, "grad_norm": 0.5212651507633511, "learning_rate": 6.3137179248238145e-06, "loss": 0.2519, "step": 1070 }, { "epoch": 0.8315217391304348, "grad_norm": 0.7846100336453622, "learning_rate": 6.307833379870394e-06, "loss": 0.3285, "step": 1071 }, { "epoch": 0.8322981366459627, "grad_norm": 0.6903631480446897, "learning_rate": 6.301946889734302e-06, "loss": 0.3406, "step": 1072 }, { "epoch": 0.8330745341614907, "grad_norm": 0.9456413594494363, "learning_rate": 6.2960584631707055e-06, "loss": 0.305, "step": 1073 }, { "epoch": 0.8338509316770186, "grad_norm": 0.6128613059958373, "learning_rate": 6.290168108937651e-06, "loss": 0.2737, "step": 1074 }, { "epoch": 0.8346273291925466, "grad_norm": 0.6303962048738531, "learning_rate": 6.284275835796053e-06, "loss": 0.2354, "step": 1075 }, { "epoch": 0.8354037267080745, "grad_norm": 0.694968205167666, "learning_rate": 6.2783816525096765e-06, "loss": 0.2358, "step": 1076 }, { "epoch": 0.8361801242236024, "grad_norm": 0.8312490763387801, "learning_rate": 6.272485567845132e-06, "loss": 0.2405, "step": 1077 }, { "epoch": 0.8369565217391305, "grad_norm": 0.7260791456625563, "learning_rate": 6.266587590571852e-06, "loss": 0.2457, "step": 1078 }, { "epoch": 0.8377329192546584, "grad_norm": 0.787341551688716, "learning_rate": 6.260687729462094e-06, "loss": 0.3135, "step": 1079 }, { "epoch": 0.8385093167701864, "grad_norm": 0.7619532999483746, "learning_rate": 6.254785993290907e-06, "loss": 0.2678, "step": 1080 }, { "epoch": 0.8392857142857143, "grad_norm": 0.6175591418568888, "learning_rate": 6.248882390836135e-06, "loss": 0.3481, "step": 1081 }, { "epoch": 0.8400621118012422, "grad_norm": 0.6415965178371213, "learning_rate": 6.242976930878395e-06, "loss": 0.2492, "step": 1082 }, { "epoch": 0.8408385093167702, "grad_norm": 0.5635918474306785, "learning_rate": 6.237069622201067e-06, "loss": 0.3112, "step": 1083 }, { "epoch": 0.8416149068322981, "grad_norm": 0.8015385917063408, "learning_rate": 6.231160473590283e-06, "loss": 0.2951, "step": 1084 }, { "epoch": 0.842391304347826, "grad_norm": 0.5949288626999686, "learning_rate": 6.22524949383491e-06, "loss": 0.2419, "step": 1085 }, { "epoch": 0.843167701863354, "grad_norm": 1.193481253782612, "learning_rate": 6.219336691726537e-06, "loss": 0.4992, "step": 1086 }, { "epoch": 0.843944099378882, "grad_norm": 0.5934843836847864, "learning_rate": 6.213422076059463e-06, "loss": 0.2695, "step": 1087 }, { "epoch": 0.84472049689441, "grad_norm": 0.5621098332455039, "learning_rate": 6.20750565563069e-06, "loss": 0.2111, "step": 1088 }, { "epoch": 0.8454968944099379, "grad_norm": 0.8312867659452824, "learning_rate": 6.201587439239901e-06, "loss": 0.3158, "step": 1089 }, { "epoch": 0.8462732919254659, "grad_norm": 0.5989856763638888, "learning_rate": 6.195667435689445e-06, "loss": 0.2177, "step": 1090 }, { "epoch": 0.8470496894409938, "grad_norm": 0.6028639695650494, "learning_rate": 6.189745653784338e-06, "loss": 0.2779, "step": 1091 }, { "epoch": 0.8478260869565217, "grad_norm": 0.7508558842482044, "learning_rate": 6.183822102332234e-06, "loss": 0.2574, "step": 1092 }, { "epoch": 0.8486024844720497, "grad_norm": 0.6479589978886029, "learning_rate": 6.177896790143424e-06, "loss": 0.2513, "step": 1093 }, { "epoch": 0.8493788819875776, "grad_norm": 0.7035931246240551, "learning_rate": 6.171969726030814e-06, "loss": 0.3408, "step": 1094 }, { "epoch": 0.8501552795031055, "grad_norm": 0.6794006253328824, "learning_rate": 6.166040918809918e-06, "loss": 0.2628, "step": 1095 }, { "epoch": 0.8509316770186336, "grad_norm": 0.7211051558720706, "learning_rate": 6.16011037729884e-06, "loss": 0.2455, "step": 1096 }, { "epoch": 0.8517080745341615, "grad_norm": 0.6535279531636462, "learning_rate": 6.1541781103182674e-06, "loss": 0.2456, "step": 1097 }, { "epoch": 0.8524844720496895, "grad_norm": 0.719851042321642, "learning_rate": 6.148244126691451e-06, "loss": 0.2579, "step": 1098 }, { "epoch": 0.8532608695652174, "grad_norm": 0.8876018714266221, "learning_rate": 6.142308435244195e-06, "loss": 0.3563, "step": 1099 }, { "epoch": 0.8540372670807453, "grad_norm": 0.7840240221531412, "learning_rate": 6.136371044804847e-06, "loss": 0.3148, "step": 1100 }, { "epoch": 0.8548136645962733, "grad_norm": 0.8352665660679446, "learning_rate": 6.130431964204276e-06, "loss": 0.2032, "step": 1101 }, { "epoch": 0.8555900621118012, "grad_norm": 0.579488904767999, "learning_rate": 6.124491202275866e-06, "loss": 0.2396, "step": 1102 }, { "epoch": 0.8563664596273292, "grad_norm": 0.6481556604558226, "learning_rate": 6.118548767855508e-06, "loss": 0.3397, "step": 1103 }, { "epoch": 0.8571428571428571, "grad_norm": 0.5394635056622308, "learning_rate": 6.112604669781572e-06, "loss": 0.3005, "step": 1104 }, { "epoch": 0.8579192546583851, "grad_norm": 1.0367652964173981, "learning_rate": 6.10665891689491e-06, "loss": 0.7442, "step": 1105 }, { "epoch": 0.8586956521739131, "grad_norm": 0.7479496563524808, "learning_rate": 6.100711518038828e-06, "loss": 0.2731, "step": 1106 }, { "epoch": 0.859472049689441, "grad_norm": 0.8247250417233352, "learning_rate": 6.094762482059087e-06, "loss": 0.2317, "step": 1107 }, { "epoch": 0.860248447204969, "grad_norm": 0.7718402556198664, "learning_rate": 6.088811817803877e-06, "loss": 0.2543, "step": 1108 }, { "epoch": 0.8610248447204969, "grad_norm": 0.6764649960899797, "learning_rate": 6.082859534123813e-06, "loss": 0.2223, "step": 1109 }, { "epoch": 0.8618012422360248, "grad_norm": 0.549144082421052, "learning_rate": 6.0769056398719205e-06, "loss": 0.2429, "step": 1110 }, { "epoch": 0.8625776397515528, "grad_norm": 0.5783258235644949, "learning_rate": 6.070950143903615e-06, "loss": 0.2328, "step": 1111 }, { "epoch": 0.8633540372670807, "grad_norm": 0.7895514551501311, "learning_rate": 6.064993055076697e-06, "loss": 0.1981, "step": 1112 }, { "epoch": 0.8641304347826086, "grad_norm": 0.9092526083069462, "learning_rate": 6.059034382251339e-06, "loss": 0.302, "step": 1113 }, { "epoch": 0.8649068322981367, "grad_norm": 0.5623252490148752, "learning_rate": 6.053074134290065e-06, "loss": 0.266, "step": 1114 }, { "epoch": 0.8656832298136646, "grad_norm": 0.5001507266614104, "learning_rate": 6.047112320057744e-06, "loss": 0.1935, "step": 1115 }, { "epoch": 0.8664596273291926, "grad_norm": 0.8509105209877769, "learning_rate": 6.041148948421575e-06, "loss": 0.2178, "step": 1116 }, { "epoch": 0.8672360248447205, "grad_norm": 1.0253806836670072, "learning_rate": 6.035184028251071e-06, "loss": 0.4725, "step": 1117 }, { "epoch": 0.8680124223602484, "grad_norm": 0.7962880068204174, "learning_rate": 6.0292175684180506e-06, "loss": 0.2681, "step": 1118 }, { "epoch": 0.8687888198757764, "grad_norm": 0.7470525352402699, "learning_rate": 6.0232495777966226e-06, "loss": 0.2971, "step": 1119 }, { "epoch": 0.8695652173913043, "grad_norm": 0.7436400971837855, "learning_rate": 6.0172800652631706e-06, "loss": 0.2459, "step": 1120 }, { "epoch": 0.8703416149068323, "grad_norm": 0.8889505257560508, "learning_rate": 6.011309039696341e-06, "loss": 0.2841, "step": 1121 }, { "epoch": 0.8711180124223602, "grad_norm": 0.8170973378520037, "learning_rate": 6.005336509977035e-06, "loss": 0.2526, "step": 1122 }, { "epoch": 0.8718944099378882, "grad_norm": 0.7031076275104816, "learning_rate": 5.999362484988384e-06, "loss": 0.2476, "step": 1123 }, { "epoch": 0.8726708074534162, "grad_norm": 1.1046738834459182, "learning_rate": 5.9933869736157524e-06, "loss": 0.2155, "step": 1124 }, { "epoch": 0.8734472049689441, "grad_norm": 0.871331748154851, "learning_rate": 5.987409984746711e-06, "loss": 0.2718, "step": 1125 }, { "epoch": 0.8742236024844721, "grad_norm": 0.5638697652701167, "learning_rate": 5.981431527271022e-06, "loss": 0.2405, "step": 1126 }, { "epoch": 0.875, "grad_norm": 0.8679623571399537, "learning_rate": 5.975451610080643e-06, "loss": 0.3045, "step": 1127 }, { "epoch": 0.8757763975155279, "grad_norm": 0.7912077979905656, "learning_rate": 5.9694702420696935e-06, "loss": 0.3254, "step": 1128 }, { "epoch": 0.8765527950310559, "grad_norm": 0.7594158779854742, "learning_rate": 5.963487432134457e-06, "loss": 0.2714, "step": 1129 }, { "epoch": 0.8773291925465838, "grad_norm": 0.679182487623916, "learning_rate": 5.9575031891733594e-06, "loss": 0.2739, "step": 1130 }, { "epoch": 0.8781055900621118, "grad_norm": 0.6380635794920811, "learning_rate": 5.951517522086957e-06, "loss": 0.2139, "step": 1131 }, { "epoch": 0.8788819875776398, "grad_norm": 0.8226151281494567, "learning_rate": 5.945530439777924e-06, "loss": 0.2162, "step": 1132 }, { "epoch": 0.8796583850931677, "grad_norm": 0.6640985053972502, "learning_rate": 5.939541951151041e-06, "loss": 0.2084, "step": 1133 }, { "epoch": 0.8804347826086957, "grad_norm": 0.6412405423413835, "learning_rate": 5.933552065113182e-06, "loss": 0.2407, "step": 1134 }, { "epoch": 0.8812111801242236, "grad_norm": 0.8558764607275015, "learning_rate": 5.927560790573296e-06, "loss": 0.2568, "step": 1135 }, { "epoch": 0.8819875776397516, "grad_norm": 1.508621571808603, "learning_rate": 5.9215681364423975e-06, "loss": 0.8311, "step": 1136 }, { "epoch": 0.8827639751552795, "grad_norm": 0.6444255437921915, "learning_rate": 5.915574111633554e-06, "loss": 0.2934, "step": 1137 }, { "epoch": 0.8835403726708074, "grad_norm": 0.7088683320297179, "learning_rate": 5.909578725061872e-06, "loss": 0.3062, "step": 1138 }, { "epoch": 0.8843167701863354, "grad_norm": 0.9895823902999982, "learning_rate": 5.9035819856444835e-06, "loss": 0.3245, "step": 1139 }, { "epoch": 0.8850931677018633, "grad_norm": 0.7032508634006223, "learning_rate": 5.897583902300531e-06, "loss": 0.2849, "step": 1140 }, { "epoch": 0.8858695652173914, "grad_norm": 0.6299890894539765, "learning_rate": 5.891584483951157e-06, "loss": 0.2983, "step": 1141 }, { "epoch": 0.8866459627329193, "grad_norm": 0.8788212131873125, "learning_rate": 5.885583739519489e-06, "loss": 0.319, "step": 1142 }, { "epoch": 0.8874223602484472, "grad_norm": 0.8155882365395793, "learning_rate": 5.879581677930629e-06, "loss": 0.2271, "step": 1143 }, { "epoch": 0.8881987577639752, "grad_norm": 0.52581135121979, "learning_rate": 5.873578308111636e-06, "loss": 0.2199, "step": 1144 }, { "epoch": 0.8889751552795031, "grad_norm": 0.7787698465180992, "learning_rate": 5.867573638991514e-06, "loss": 0.256, "step": 1145 }, { "epoch": 0.889751552795031, "grad_norm": 0.7580505733490497, "learning_rate": 5.861567679501201e-06, "loss": 0.2138, "step": 1146 }, { "epoch": 0.890527950310559, "grad_norm": 1.9943317859393914, "learning_rate": 5.855560438573553e-06, "loss": 0.9319, "step": 1147 }, { "epoch": 0.8913043478260869, "grad_norm": 0.6216753024158402, "learning_rate": 5.849551925143334e-06, "loss": 0.2609, "step": 1148 }, { "epoch": 0.8920807453416149, "grad_norm": 0.8181775826047529, "learning_rate": 5.843542148147199e-06, "loss": 0.2703, "step": 1149 }, { "epoch": 0.8928571428571429, "grad_norm": 0.7525196435766732, "learning_rate": 5.837531116523683e-06, "loss": 0.2848, "step": 1150 }, { "epoch": 0.8936335403726708, "grad_norm": 1.028646180620769, "learning_rate": 5.831518839213186e-06, "loss": 0.828, "step": 1151 }, { "epoch": 0.8944099378881988, "grad_norm": 0.6734822838484019, "learning_rate": 5.825505325157962e-06, "loss": 0.3463, "step": 1152 }, { "epoch": 0.8951863354037267, "grad_norm": 0.8483953460478394, "learning_rate": 5.819490583302103e-06, "loss": 0.2658, "step": 1153 }, { "epoch": 0.8959627329192547, "grad_norm": 0.876526724662098, "learning_rate": 5.8134746225915305e-06, "loss": 0.2018, "step": 1154 }, { "epoch": 0.8967391304347826, "grad_norm": 0.7309323976146965, "learning_rate": 5.807457451973976e-06, "loss": 0.2643, "step": 1155 }, { "epoch": 0.8975155279503105, "grad_norm": 0.9356942791420667, "learning_rate": 5.801439080398968e-06, "loss": 0.2832, "step": 1156 }, { "epoch": 0.8982919254658385, "grad_norm": 0.5739052194831438, "learning_rate": 5.795419516817826e-06, "loss": 0.2495, "step": 1157 }, { "epoch": 0.8990683229813664, "grad_norm": 0.8671102612819551, "learning_rate": 5.789398770183642e-06, "loss": 0.2322, "step": 1158 }, { "epoch": 0.8998447204968945, "grad_norm": 0.7956384960138583, "learning_rate": 5.783376849451263e-06, "loss": 0.2661, "step": 1159 }, { "epoch": 0.9006211180124224, "grad_norm": 0.8407481116465576, "learning_rate": 5.77735376357729e-06, "loss": 0.3255, "step": 1160 }, { "epoch": 0.9013975155279503, "grad_norm": 0.6816842691541677, "learning_rate": 5.7713295215200485e-06, "loss": 0.2471, "step": 1161 }, { "epoch": 0.9021739130434783, "grad_norm": 0.8252812440988943, "learning_rate": 5.76530413223959e-06, "loss": 0.8636, "step": 1162 }, { "epoch": 0.9029503105590062, "grad_norm": 0.8615788831580373, "learning_rate": 5.759277604697667e-06, "loss": 0.2675, "step": 1163 }, { "epoch": 0.9037267080745341, "grad_norm": 0.7637146940378313, "learning_rate": 5.753249947857731e-06, "loss": 0.2459, "step": 1164 }, { "epoch": 0.9045031055900621, "grad_norm": 0.815949061135136, "learning_rate": 5.747221170684909e-06, "loss": 0.289, "step": 1165 }, { "epoch": 0.90527950310559, "grad_norm": 0.7288703922838393, "learning_rate": 5.741191282145995e-06, "loss": 0.2986, "step": 1166 }, { "epoch": 0.906055900621118, "grad_norm": 0.6281786675270065, "learning_rate": 5.735160291209436e-06, "loss": 0.2457, "step": 1167 }, { "epoch": 0.906832298136646, "grad_norm": 0.800013064239338, "learning_rate": 5.729128206845317e-06, "loss": 0.84, "step": 1168 }, { "epoch": 0.907608695652174, "grad_norm": 0.5269476732928574, "learning_rate": 5.723095038025355e-06, "loss": 0.2232, "step": 1169 }, { "epoch": 0.9083850931677019, "grad_norm": 0.6609771525397284, "learning_rate": 5.717060793722876e-06, "loss": 0.2324, "step": 1170 }, { "epoch": 0.9091614906832298, "grad_norm": 0.7098565664782565, "learning_rate": 5.711025482912802e-06, "loss": 0.3569, "step": 1171 }, { "epoch": 0.9099378881987578, "grad_norm": 0.6926367341354882, "learning_rate": 5.704989114571649e-06, "loss": 0.334, "step": 1172 }, { "epoch": 0.9107142857142857, "grad_norm": 0.8947569677479935, "learning_rate": 5.698951697677498e-06, "loss": 0.2898, "step": 1173 }, { "epoch": 0.9114906832298136, "grad_norm": 0.6486955348285195, "learning_rate": 5.692913241209997e-06, "loss": 0.2802, "step": 1174 }, { "epoch": 0.9122670807453416, "grad_norm": 0.8038592139271136, "learning_rate": 5.686873754150334e-06, "loss": 0.2578, "step": 1175 }, { "epoch": 0.9130434782608695, "grad_norm": 0.7665290942665837, "learning_rate": 5.680833245481234e-06, "loss": 0.2606, "step": 1176 }, { "epoch": 0.9138198757763976, "grad_norm": 0.5837572971098813, "learning_rate": 5.674791724186939e-06, "loss": 0.2525, "step": 1177 }, { "epoch": 0.9145962732919255, "grad_norm": 0.8551822262660714, "learning_rate": 5.668749199253198e-06, "loss": 0.2652, "step": 1178 }, { "epoch": 0.9153726708074534, "grad_norm": 0.7925547380297197, "learning_rate": 5.662705679667254e-06, "loss": 0.354, "step": 1179 }, { "epoch": 0.9161490683229814, "grad_norm": 0.5302330950869637, "learning_rate": 5.656661174417829e-06, "loss": 0.2425, "step": 1180 }, { "epoch": 0.9169254658385093, "grad_norm": 0.7490908991702302, "learning_rate": 5.650615692495106e-06, "loss": 0.3294, "step": 1181 }, { "epoch": 0.9177018633540373, "grad_norm": 0.70087106679741, "learning_rate": 5.644569242890728e-06, "loss": 0.2496, "step": 1182 }, { "epoch": 0.9184782608695652, "grad_norm": 1.0038725160872672, "learning_rate": 5.6385218345977745e-06, "loss": 0.2568, "step": 1183 }, { "epoch": 0.9192546583850931, "grad_norm": 0.8258994738371787, "learning_rate": 5.632473476610748e-06, "loss": 0.3419, "step": 1184 }, { "epoch": 0.9200310559006211, "grad_norm": 0.6021914316338413, "learning_rate": 5.626424177925568e-06, "loss": 0.274, "step": 1185 }, { "epoch": 0.9208074534161491, "grad_norm": 0.6949225363815668, "learning_rate": 5.62037394753955e-06, "loss": 0.2628, "step": 1186 }, { "epoch": 0.921583850931677, "grad_norm": 0.7776332160316921, "learning_rate": 5.614322794451397e-06, "loss": 0.4728, "step": 1187 }, { "epoch": 0.922360248447205, "grad_norm": 0.725520272359154, "learning_rate": 5.608270727661183e-06, "loss": 0.2394, "step": 1188 }, { "epoch": 0.9231366459627329, "grad_norm": 0.5886120019071476, "learning_rate": 5.602217756170344e-06, "loss": 0.2415, "step": 1189 }, { "epoch": 0.9239130434782609, "grad_norm": 0.7570245814440495, "learning_rate": 5.596163888981656e-06, "loss": 0.2473, "step": 1190 }, { "epoch": 0.9246894409937888, "grad_norm": 0.752799950828631, "learning_rate": 5.590109135099233e-06, "loss": 0.2695, "step": 1191 }, { "epoch": 0.9254658385093167, "grad_norm": 0.9820602880716167, "learning_rate": 5.584053503528503e-06, "loss": 0.2577, "step": 1192 }, { "epoch": 0.9262422360248447, "grad_norm": 0.750130167149084, "learning_rate": 5.577997003276203e-06, "loss": 0.2336, "step": 1193 }, { "epoch": 0.9270186335403726, "grad_norm": 0.6161102534300852, "learning_rate": 5.5719396433503615e-06, "loss": 0.2842, "step": 1194 }, { "epoch": 0.9277950310559007, "grad_norm": 0.7042284647373019, "learning_rate": 5.565881432760283e-06, "loss": 0.2011, "step": 1195 }, { "epoch": 0.9285714285714286, "grad_norm": 0.6407622509997999, "learning_rate": 5.559822380516539e-06, "loss": 0.2113, "step": 1196 }, { "epoch": 0.9293478260869565, "grad_norm": 0.8798892875070309, "learning_rate": 5.553762495630957e-06, "loss": 0.2182, "step": 1197 }, { "epoch": 0.9301242236024845, "grad_norm": 0.7711017355252828, "learning_rate": 5.5477017871165925e-06, "loss": 0.2813, "step": 1198 }, { "epoch": 0.9309006211180124, "grad_norm": 0.8829616953134621, "learning_rate": 5.5416402639877375e-06, "loss": 0.7657, "step": 1199 }, { "epoch": 0.9316770186335404, "grad_norm": 0.9096826897973084, "learning_rate": 5.5355779352598884e-06, "loss": 0.2678, "step": 1200 }, { "epoch": 0.9324534161490683, "grad_norm": 0.7099814193385338, "learning_rate": 5.529514809949742e-06, "loss": 0.3112, "step": 1201 }, { "epoch": 0.9332298136645962, "grad_norm": 0.8043644235133904, "learning_rate": 5.523450897075179e-06, "loss": 0.212, "step": 1202 }, { "epoch": 0.9340062111801242, "grad_norm": 0.8314796016368029, "learning_rate": 5.517386205655252e-06, "loss": 0.3066, "step": 1203 }, { "epoch": 0.9347826086956522, "grad_norm": 0.7454245593358939, "learning_rate": 5.511320744710171e-06, "loss": 0.2107, "step": 1204 }, { "epoch": 0.9355590062111802, "grad_norm": 0.8414302380691713, "learning_rate": 5.505254523261295e-06, "loss": 0.7958, "step": 1205 }, { "epoch": 0.9363354037267081, "grad_norm": 0.5106116824180411, "learning_rate": 5.4991875503311075e-06, "loss": 0.3124, "step": 1206 }, { "epoch": 0.937111801242236, "grad_norm": 0.832531739622361, "learning_rate": 5.493119834943212e-06, "loss": 0.3428, "step": 1207 }, { "epoch": 0.937888198757764, "grad_norm": 0.6299443680489833, "learning_rate": 5.487051386122319e-06, "loss": 0.2813, "step": 1208 }, { "epoch": 0.9386645962732919, "grad_norm": 0.5640881626115966, "learning_rate": 5.480982212894226e-06, "loss": 0.2166, "step": 1209 }, { "epoch": 0.9394409937888198, "grad_norm": 0.7939921535028214, "learning_rate": 5.47491232428581e-06, "loss": 0.2922, "step": 1210 }, { "epoch": 0.9402173913043478, "grad_norm": 0.7425666590595001, "learning_rate": 5.468841729325014e-06, "loss": 0.7222, "step": 1211 }, { "epoch": 0.9409937888198758, "grad_norm": 0.7043225615970836, "learning_rate": 5.462770437040824e-06, "loss": 0.208, "step": 1212 }, { "epoch": 0.9417701863354038, "grad_norm": 0.9680219438015032, "learning_rate": 5.456698456463273e-06, "loss": 0.3637, "step": 1213 }, { "epoch": 0.9425465838509317, "grad_norm": 0.7918278435095899, "learning_rate": 5.450625796623411e-06, "loss": 0.2783, "step": 1214 }, { "epoch": 0.9433229813664596, "grad_norm": 0.718677769684927, "learning_rate": 5.444552466553302e-06, "loss": 0.2731, "step": 1215 }, { "epoch": 0.9440993788819876, "grad_norm": 0.6220371249705906, "learning_rate": 5.438478475286003e-06, "loss": 0.2354, "step": 1216 }, { "epoch": 0.9448757763975155, "grad_norm": 0.795850654675338, "learning_rate": 5.432403831855559e-06, "loss": 0.2941, "step": 1217 }, { "epoch": 0.9456521739130435, "grad_norm": 0.6595008295775245, "learning_rate": 5.4263285452969805e-06, "loss": 0.2853, "step": 1218 }, { "epoch": 0.9464285714285714, "grad_norm": 0.9280120511934674, "learning_rate": 5.420252624646238e-06, "loss": 0.2343, "step": 1219 }, { "epoch": 0.9472049689440993, "grad_norm": 0.533746261853156, "learning_rate": 5.414176078940244e-06, "loss": 0.2406, "step": 1220 }, { "epoch": 0.9479813664596274, "grad_norm": 0.7215084549862628, "learning_rate": 5.40809891721684e-06, "loss": 0.2726, "step": 1221 }, { "epoch": 0.9487577639751553, "grad_norm": 0.5429656818952893, "learning_rate": 5.402021148514784e-06, "loss": 0.2378, "step": 1222 }, { "epoch": 0.9495341614906833, "grad_norm": 0.694260139231815, "learning_rate": 5.395942781873736e-06, "loss": 0.2979, "step": 1223 }, { "epoch": 0.9503105590062112, "grad_norm": 0.48080412056652694, "learning_rate": 5.389863826334248e-06, "loss": 0.293, "step": 1224 }, { "epoch": 0.9510869565217391, "grad_norm": 0.6298002968801176, "learning_rate": 5.383784290937747e-06, "loss": 0.2777, "step": 1225 }, { "epoch": 0.9518633540372671, "grad_norm": 0.7812350958388091, "learning_rate": 5.37770418472652e-06, "loss": 0.2338, "step": 1226 }, { "epoch": 0.952639751552795, "grad_norm": 0.711781565509409, "learning_rate": 5.371623516743705e-06, "loss": 0.2533, "step": 1227 }, { "epoch": 0.953416149068323, "grad_norm": 0.5626142245617642, "learning_rate": 5.365542296033274e-06, "loss": 0.2171, "step": 1228 }, { "epoch": 0.9541925465838509, "grad_norm": 0.7595440683913137, "learning_rate": 5.359460531640024e-06, "loss": 0.7248, "step": 1229 }, { "epoch": 0.9549689440993789, "grad_norm": 0.7918030021293927, "learning_rate": 5.353378232609557e-06, "loss": 0.2942, "step": 1230 }, { "epoch": 0.9557453416149069, "grad_norm": 0.6840449813439891, "learning_rate": 5.347295407988274e-06, "loss": 0.2101, "step": 1231 }, { "epoch": 0.9565217391304348, "grad_norm": 0.7227660887244846, "learning_rate": 5.341212066823356e-06, "loss": 0.2618, "step": 1232 }, { "epoch": 0.9572981366459627, "grad_norm": 0.6052750946153505, "learning_rate": 5.335128218162749e-06, "loss": 0.2447, "step": 1233 }, { "epoch": 0.9580745341614907, "grad_norm": 0.7667104969304076, "learning_rate": 5.329043871055161e-06, "loss": 0.2432, "step": 1234 }, { "epoch": 0.9588509316770186, "grad_norm": 0.6976008098179689, "learning_rate": 5.322959034550035e-06, "loss": 0.2146, "step": 1235 }, { "epoch": 0.9596273291925466, "grad_norm": 0.8476312427145755, "learning_rate": 5.316873717697545e-06, "loss": 0.2771, "step": 1236 }, { "epoch": 0.9604037267080745, "grad_norm": 0.6844024359334461, "learning_rate": 5.3107879295485764e-06, "loss": 0.2746, "step": 1237 }, { "epoch": 0.9611801242236024, "grad_norm": 0.5039783362689135, "learning_rate": 5.30470167915472e-06, "loss": 0.2743, "step": 1238 }, { "epoch": 0.9619565217391305, "grad_norm": 0.8585386885537081, "learning_rate": 5.298614975568249e-06, "loss": 0.228, "step": 1239 }, { "epoch": 0.9627329192546584, "grad_norm": 0.5802171654068117, "learning_rate": 5.29252782784212e-06, "loss": 0.17, "step": 1240 }, { "epoch": 0.9635093167701864, "grad_norm": 0.6208751688881062, "learning_rate": 5.286440245029937e-06, "loss": 0.2738, "step": 1241 }, { "epoch": 0.9642857142857143, "grad_norm": 0.758463841649618, "learning_rate": 5.2803522361859596e-06, "loss": 0.2148, "step": 1242 }, { "epoch": 0.9650621118012422, "grad_norm": 0.8404994095635101, "learning_rate": 5.27426381036508e-06, "loss": 0.2748, "step": 1243 }, { "epoch": 0.9658385093167702, "grad_norm": 0.7095569189189892, "learning_rate": 5.268174976622811e-06, "loss": 0.2581, "step": 1244 }, { "epoch": 0.9666149068322981, "grad_norm": 0.6240030113268366, "learning_rate": 5.2620857440152686e-06, "loss": 0.3224, "step": 1245 }, { "epoch": 0.967391304347826, "grad_norm": 0.6935097782383427, "learning_rate": 5.255996121599167e-06, "loss": 0.2776, "step": 1246 }, { "epoch": 0.968167701863354, "grad_norm": 0.7053047685470675, "learning_rate": 5.249906118431794e-06, "loss": 0.3109, "step": 1247 }, { "epoch": 0.968944099378882, "grad_norm": 0.909947905445919, "learning_rate": 5.243815743571009e-06, "loss": 0.2538, "step": 1248 }, { "epoch": 0.96972049689441, "grad_norm": 0.8253393714241705, "learning_rate": 5.2377250060752234e-06, "loss": 0.2477, "step": 1249 }, { "epoch": 0.9704968944099379, "grad_norm": 0.5130473355971035, "learning_rate": 5.231633915003389e-06, "loss": 0.2264, "step": 1250 }, { "epoch": 0.9712732919254659, "grad_norm": 0.5910388776712077, "learning_rate": 5.2255424794149765e-06, "loss": 0.2797, "step": 1251 }, { "epoch": 0.9720496894409938, "grad_norm": 0.6049263042315812, "learning_rate": 5.219450708369977e-06, "loss": 0.2845, "step": 1252 }, { "epoch": 0.9728260869565217, "grad_norm": 0.7769778846630881, "learning_rate": 5.213358610928878e-06, "loss": 0.3424, "step": 1253 }, { "epoch": 0.9736024844720497, "grad_norm": 0.618811370797138, "learning_rate": 5.2072661961526505e-06, "loss": 0.2211, "step": 1254 }, { "epoch": 0.9743788819875776, "grad_norm": 0.6459709857517596, "learning_rate": 5.20117347310274e-06, "loss": 0.2637, "step": 1255 }, { "epoch": 0.9751552795031055, "grad_norm": 0.5955868024988595, "learning_rate": 5.19508045084105e-06, "loss": 0.2671, "step": 1256 }, { "epoch": 0.9759316770186336, "grad_norm": 0.7010570993087017, "learning_rate": 5.188987138429926e-06, "loss": 0.2349, "step": 1257 }, { "epoch": 0.9767080745341615, "grad_norm": 0.6425007996975166, "learning_rate": 5.1828935449321515e-06, "loss": 0.2767, "step": 1258 }, { "epoch": 0.9774844720496895, "grad_norm": 0.772894347462212, "learning_rate": 5.176799679410922e-06, "loss": 0.461, "step": 1259 }, { "epoch": 0.9782608695652174, "grad_norm": 0.7348098114656484, "learning_rate": 5.17070555092984e-06, "loss": 0.2384, "step": 1260 }, { "epoch": 0.9790372670807453, "grad_norm": 0.6813433532054999, "learning_rate": 5.1646111685529e-06, "loss": 0.2914, "step": 1261 }, { "epoch": 0.9798136645962733, "grad_norm": 0.7640968833955561, "learning_rate": 5.15851654134447e-06, "loss": 0.2593, "step": 1262 }, { "epoch": 0.9805900621118012, "grad_norm": 0.9583425961057376, "learning_rate": 5.152421678369288e-06, "loss": 0.3525, "step": 1263 }, { "epoch": 0.9813664596273292, "grad_norm": 0.6160546199841387, "learning_rate": 5.146326588692439e-06, "loss": 0.2191, "step": 1264 }, { "epoch": 0.9821428571428571, "grad_norm": 0.7277196769847523, "learning_rate": 5.140231281379345e-06, "loss": 0.2624, "step": 1265 }, { "epoch": 0.9829192546583851, "grad_norm": 0.6161059474592493, "learning_rate": 5.1341357654957546e-06, "loss": 0.2409, "step": 1266 }, { "epoch": 0.9836956521739131, "grad_norm": 0.8291746741227934, "learning_rate": 5.128040050107724e-06, "loss": 0.3136, "step": 1267 }, { "epoch": 0.984472049689441, "grad_norm": 0.6393363709896841, "learning_rate": 5.121944144281606e-06, "loss": 0.1837, "step": 1268 }, { "epoch": 0.985248447204969, "grad_norm": 0.6867658777984186, "learning_rate": 5.115848057084039e-06, "loss": 0.2422, "step": 1269 }, { "epoch": 0.9860248447204969, "grad_norm": 0.7681509613996212, "learning_rate": 5.109751797581929e-06, "loss": 0.3076, "step": 1270 }, { "epoch": 0.9868012422360248, "grad_norm": 0.5782258089160104, "learning_rate": 5.103655374842439e-06, "loss": 0.22, "step": 1271 }, { "epoch": 0.9875776397515528, "grad_norm": 0.7386818265356727, "learning_rate": 5.097558797932973e-06, "loss": 0.2761, "step": 1272 }, { "epoch": 0.9883540372670807, "grad_norm": 0.638581672234056, "learning_rate": 5.091462075921169e-06, "loss": 0.2562, "step": 1273 }, { "epoch": 0.9891304347826086, "grad_norm": 0.8529940259484317, "learning_rate": 5.085365217874875e-06, "loss": 0.2496, "step": 1274 }, { "epoch": 0.9899068322981367, "grad_norm": 0.7803049200859326, "learning_rate": 5.0792682328621455e-06, "loss": 0.2603, "step": 1275 }, { "epoch": 0.9906832298136646, "grad_norm": 0.9198687104835407, "learning_rate": 5.073171129951223e-06, "loss": 0.3992, "step": 1276 }, { "epoch": 0.9914596273291926, "grad_norm": 0.9369356446945384, "learning_rate": 5.067073918210525e-06, "loss": 0.244, "step": 1277 }, { "epoch": 0.9922360248447205, "grad_norm": 0.6344136725248555, "learning_rate": 5.060976606708628e-06, "loss": 0.2283, "step": 1278 }, { "epoch": 0.9930124223602484, "grad_norm": 0.8940885283851997, "learning_rate": 5.054879204514263e-06, "loss": 0.2607, "step": 1279 }, { "epoch": 0.9937888198757764, "grad_norm": 0.8927870513461171, "learning_rate": 5.048781720696291e-06, "loss": 0.2985, "step": 1280 }, { "epoch": 0.9945652173913043, "grad_norm": 0.6408374891890262, "learning_rate": 5.042684164323698e-06, "loss": 0.29, "step": 1281 }, { "epoch": 0.9953416149068323, "grad_norm": 0.7343078086233087, "learning_rate": 5.036586544465571e-06, "loss": 0.2489, "step": 1282 }, { "epoch": 0.9961180124223602, "grad_norm": 0.6222807503831567, "learning_rate": 5.0304888701911e-06, "loss": 0.2435, "step": 1283 }, { "epoch": 0.9968944099378882, "grad_norm": 0.8461029007658184, "learning_rate": 5.024391150569551e-06, "loss": 0.2793, "step": 1284 }, { "epoch": 0.9976708074534162, "grad_norm": 0.8717963343881938, "learning_rate": 5.018293394670259e-06, "loss": 0.2754, "step": 1285 }, { "epoch": 0.9984472049689441, "grad_norm": 0.6208403255641849, "learning_rate": 5.012195611562613e-06, "loss": 0.2827, "step": 1286 }, { "epoch": 0.9992236024844721, "grad_norm": 0.8122646269152067, "learning_rate": 5.006097810316042e-06, "loss": 0.3414, "step": 1287 }, { "epoch": 1.0, "grad_norm": 0.6346256823193085, "learning_rate": 5e-06, "loss": 0.1695, "step": 1288 }, { "epoch": 1.000776397515528, "grad_norm": 0.5328421174718584, "learning_rate": 4.9939021896839596e-06, "loss": 0.2593, "step": 1289 }, { "epoch": 1.0015527950310559, "grad_norm": 0.5510739879796169, "learning_rate": 4.987804388437388e-06, "loss": 0.2403, "step": 1290 }, { "epoch": 1.0023291925465838, "grad_norm": 0.8843076178274797, "learning_rate": 4.9817066053297425e-06, "loss": 0.1517, "step": 1291 }, { "epoch": 1.0031055900621118, "grad_norm": 0.555920237392263, "learning_rate": 4.975608849430451e-06, "loss": 0.1414, "step": 1292 }, { "epoch": 1.0038819875776397, "grad_norm": 0.3998243419475203, "learning_rate": 4.9695111298089025e-06, "loss": 0.1723, "step": 1293 }, { "epoch": 1.0046583850931676, "grad_norm": 0.8198918134292718, "learning_rate": 4.963413455534431e-06, "loss": 0.17, "step": 1294 }, { "epoch": 1.0054347826086956, "grad_norm": 0.6707544382805484, "learning_rate": 4.957315835676305e-06, "loss": 0.2645, "step": 1295 }, { "epoch": 1.0062111801242235, "grad_norm": 0.6071108442444486, "learning_rate": 4.95121827930371e-06, "loss": 0.2638, "step": 1296 }, { "epoch": 1.0069875776397517, "grad_norm": 0.4309474018082429, "learning_rate": 4.9451207954857374e-06, "loss": 0.1921, "step": 1297 }, { "epoch": 1.0077639751552796, "grad_norm": 0.6454019770456677, "learning_rate": 4.939023393291373e-06, "loss": 0.1986, "step": 1298 }, { "epoch": 1.0085403726708075, "grad_norm": 0.6307230142504124, "learning_rate": 4.932926081789476e-06, "loss": 0.3136, "step": 1299 }, { "epoch": 1.0093167701863355, "grad_norm": 0.5940229078135638, "learning_rate": 4.926828870048779e-06, "loss": 0.1795, "step": 1300 }, { "epoch": 1.0100931677018634, "grad_norm": 0.5529180046376888, "learning_rate": 4.920731767137856e-06, "loss": 0.1923, "step": 1301 }, { "epoch": 1.0108695652173914, "grad_norm": 0.5048210596426576, "learning_rate": 4.914634782125127e-06, "loss": 0.1889, "step": 1302 }, { "epoch": 1.0116459627329193, "grad_norm": 0.5367046324047513, "learning_rate": 4.908537924078833e-06, "loss": 0.1746, "step": 1303 }, { "epoch": 1.0124223602484472, "grad_norm": 0.6091364432373333, "learning_rate": 4.902441202067028e-06, "loss": 0.1939, "step": 1304 }, { "epoch": 1.0131987577639752, "grad_norm": 0.6226409827054811, "learning_rate": 4.896344625157563e-06, "loss": 0.1954, "step": 1305 }, { "epoch": 1.013975155279503, "grad_norm": 0.6082684735603986, "learning_rate": 4.890248202418073e-06, "loss": 0.2391, "step": 1306 }, { "epoch": 1.014751552795031, "grad_norm": 0.5967547875985121, "learning_rate": 4.884151942915962e-06, "loss": 0.2092, "step": 1307 }, { "epoch": 1.015527950310559, "grad_norm": 0.5344160765340419, "learning_rate": 4.878055855718395e-06, "loss": 0.2409, "step": 1308 }, { "epoch": 1.016304347826087, "grad_norm": 0.6157656537935127, "learning_rate": 4.871959949892278e-06, "loss": 0.1654, "step": 1309 }, { "epoch": 1.0170807453416149, "grad_norm": 0.6035904532371178, "learning_rate": 4.865864234504246e-06, "loss": 0.2288, "step": 1310 }, { "epoch": 1.0178571428571428, "grad_norm": 0.5058358693243095, "learning_rate": 4.859768718620656e-06, "loss": 0.1726, "step": 1311 }, { "epoch": 1.0186335403726707, "grad_norm": 0.5864465235850802, "learning_rate": 4.853673411307564e-06, "loss": 0.2164, "step": 1312 }, { "epoch": 1.0194099378881987, "grad_norm": 0.5710851056031502, "learning_rate": 4.847578321630715e-06, "loss": 0.2692, "step": 1313 }, { "epoch": 1.0201863354037266, "grad_norm": 0.5874749107242496, "learning_rate": 4.841483458655532e-06, "loss": 0.203, "step": 1314 }, { "epoch": 1.0209627329192548, "grad_norm": 0.7115529242037953, "learning_rate": 4.835388831447102e-06, "loss": 0.2228, "step": 1315 }, { "epoch": 1.0217391304347827, "grad_norm": 0.7347856038185626, "learning_rate": 4.829294449070161e-06, "loss": 0.2181, "step": 1316 }, { "epoch": 1.0225155279503106, "grad_norm": 0.6186615462920271, "learning_rate": 4.8232003205890785e-06, "loss": 0.1503, "step": 1317 }, { "epoch": 1.0232919254658386, "grad_norm": 0.7828659236902598, "learning_rate": 4.817106455067848e-06, "loss": 0.2449, "step": 1318 }, { "epoch": 1.0240683229813665, "grad_norm": 0.7390691507052978, "learning_rate": 4.811012861570074e-06, "loss": 0.1719, "step": 1319 }, { "epoch": 1.0248447204968945, "grad_norm": 0.5382438482341255, "learning_rate": 4.804919549158951e-06, "loss": 0.2439, "step": 1320 }, { "epoch": 1.0256211180124224, "grad_norm": 0.7106300397333185, "learning_rate": 4.798826526897261e-06, "loss": 0.2436, "step": 1321 }, { "epoch": 1.0263975155279503, "grad_norm": 0.5277708821267684, "learning_rate": 4.792733803847351e-06, "loss": 0.2362, "step": 1322 }, { "epoch": 1.0271739130434783, "grad_norm": 0.6002656808278589, "learning_rate": 4.786641389071123e-06, "loss": 0.1731, "step": 1323 }, { "epoch": 1.0279503105590062, "grad_norm": 0.6043359837779335, "learning_rate": 4.7805492916300235e-06, "loss": 0.2308, "step": 1324 }, { "epoch": 1.0287267080745341, "grad_norm": 0.650071613156722, "learning_rate": 4.774457520585024e-06, "loss": 0.2117, "step": 1325 }, { "epoch": 1.029503105590062, "grad_norm": 0.8453365425096577, "learning_rate": 4.768366084996612e-06, "loss": 0.2174, "step": 1326 }, { "epoch": 1.03027950310559, "grad_norm": 0.7301632530965457, "learning_rate": 4.762274993924777e-06, "loss": 0.2455, "step": 1327 }, { "epoch": 1.031055900621118, "grad_norm": 0.6579491281046702, "learning_rate": 4.756184256428992e-06, "loss": 0.2121, "step": 1328 }, { "epoch": 1.031832298136646, "grad_norm": 0.6301137370027096, "learning_rate": 4.750093881568208e-06, "loss": 0.1662, "step": 1329 }, { "epoch": 1.0326086956521738, "grad_norm": 0.6024165716862413, "learning_rate": 4.744003878400836e-06, "loss": 0.1993, "step": 1330 }, { "epoch": 1.0333850931677018, "grad_norm": 0.6117172975035697, "learning_rate": 4.737914255984733e-06, "loss": 0.2771, "step": 1331 }, { "epoch": 1.0341614906832297, "grad_norm": 0.6910931415720766, "learning_rate": 4.731825023377191e-06, "loss": 0.1727, "step": 1332 }, { "epoch": 1.0349378881987579, "grad_norm": 0.6171169399428571, "learning_rate": 4.72573618963492e-06, "loss": 0.2525, "step": 1333 }, { "epoch": 1.0357142857142858, "grad_norm": 0.6131828615348762, "learning_rate": 4.719647763814041e-06, "loss": 0.271, "step": 1334 }, { "epoch": 1.0364906832298137, "grad_norm": 0.7072005823882377, "learning_rate": 4.713559754970065e-06, "loss": 0.2436, "step": 1335 }, { "epoch": 1.0372670807453417, "grad_norm": 0.5331636002747678, "learning_rate": 4.707472172157882e-06, "loss": 0.2376, "step": 1336 }, { "epoch": 1.0380434782608696, "grad_norm": 0.5984788523463818, "learning_rate": 4.701385024431751e-06, "loss": 0.1882, "step": 1337 }, { "epoch": 1.0388198757763976, "grad_norm": 0.6204024941661178, "learning_rate": 4.695298320845282e-06, "loss": 0.2687, "step": 1338 }, { "epoch": 1.0395962732919255, "grad_norm": 0.6685064867964873, "learning_rate": 4.689212070451425e-06, "loss": 0.1821, "step": 1339 }, { "epoch": 1.0403726708074534, "grad_norm": 0.6338006704034125, "learning_rate": 4.683126282302457e-06, "loss": 0.1629, "step": 1340 }, { "epoch": 1.0411490683229814, "grad_norm": 0.7154452828140689, "learning_rate": 4.677040965449967e-06, "loss": 0.2505, "step": 1341 }, { "epoch": 1.0419254658385093, "grad_norm": 0.6221840409828215, "learning_rate": 4.67095612894484e-06, "loss": 0.2469, "step": 1342 }, { "epoch": 1.0427018633540373, "grad_norm": 0.550878318416789, "learning_rate": 4.664871781837252e-06, "loss": 0.1757, "step": 1343 }, { "epoch": 1.0434782608695652, "grad_norm": 0.6637937134264269, "learning_rate": 4.6587879331766465e-06, "loss": 0.2186, "step": 1344 }, { "epoch": 1.0442546583850931, "grad_norm": 0.7099289630476762, "learning_rate": 4.652704592011727e-06, "loss": 0.1595, "step": 1345 }, { "epoch": 1.045031055900621, "grad_norm": 0.9632802179606169, "learning_rate": 4.646621767390444e-06, "loss": 0.8282, "step": 1346 }, { "epoch": 1.045807453416149, "grad_norm": 0.7918201753337603, "learning_rate": 4.640539468359979e-06, "loss": 0.1649, "step": 1347 }, { "epoch": 1.046583850931677, "grad_norm": 0.6152003610401385, "learning_rate": 4.634457703966729e-06, "loss": 0.2176, "step": 1348 }, { "epoch": 1.0473602484472049, "grad_norm": 0.5543141376718322, "learning_rate": 4.628376483256298e-06, "loss": 0.2046, "step": 1349 }, { "epoch": 1.0481366459627328, "grad_norm": 0.47331053834280645, "learning_rate": 4.622295815273482e-06, "loss": 0.219, "step": 1350 }, { "epoch": 1.048913043478261, "grad_norm": 0.5564576160913547, "learning_rate": 4.616215709062254e-06, "loss": 0.2728, "step": 1351 }, { "epoch": 1.049689440993789, "grad_norm": 0.6888152341518974, "learning_rate": 4.610136173665751e-06, "loss": 0.157, "step": 1352 }, { "epoch": 1.0504658385093169, "grad_norm": 0.5023945303679258, "learning_rate": 4.604057218126264e-06, "loss": 0.1982, "step": 1353 }, { "epoch": 1.0512422360248448, "grad_norm": 0.6476852064519764, "learning_rate": 4.597978851485217e-06, "loss": 0.1689, "step": 1354 }, { "epoch": 1.0520186335403727, "grad_norm": 0.6384672147482549, "learning_rate": 4.59190108278316e-06, "loss": 0.2656, "step": 1355 }, { "epoch": 1.0527950310559007, "grad_norm": 0.598681626558817, "learning_rate": 4.585823921059757e-06, "loss": 0.1981, "step": 1356 }, { "epoch": 1.0535714285714286, "grad_norm": 0.6954025737372213, "learning_rate": 4.579747375353763e-06, "loss": 0.1717, "step": 1357 }, { "epoch": 1.0543478260869565, "grad_norm": 0.8382466722902082, "learning_rate": 4.57367145470302e-06, "loss": 0.28, "step": 1358 }, { "epoch": 1.0551242236024845, "grad_norm": 0.6394858159534675, "learning_rate": 4.567596168144443e-06, "loss": 0.2109, "step": 1359 }, { "epoch": 1.0559006211180124, "grad_norm": 0.6536991556661943, "learning_rate": 4.561521524713998e-06, "loss": 0.176, "step": 1360 }, { "epoch": 1.0566770186335404, "grad_norm": 0.6124151753066694, "learning_rate": 4.555447533446699e-06, "loss": 0.2901, "step": 1361 }, { "epoch": 1.0574534161490683, "grad_norm": 0.5621191956535986, "learning_rate": 4.5493742033765906e-06, "loss": 0.2144, "step": 1362 }, { "epoch": 1.0582298136645962, "grad_norm": 0.8743823456512968, "learning_rate": 4.543301543536729e-06, "loss": 0.1828, "step": 1363 }, { "epoch": 1.0590062111801242, "grad_norm": 0.5239066635303923, "learning_rate": 4.537229562959178e-06, "loss": 0.2159, "step": 1364 }, { "epoch": 1.059782608695652, "grad_norm": 0.7153115833691454, "learning_rate": 4.531158270674989e-06, "loss": 0.204, "step": 1365 }, { "epoch": 1.06055900621118, "grad_norm": 0.6093004493418376, "learning_rate": 4.525087675714191e-06, "loss": 0.2028, "step": 1366 }, { "epoch": 1.061335403726708, "grad_norm": 0.5460391970634304, "learning_rate": 4.519017787105776e-06, "loss": 0.2316, "step": 1367 }, { "epoch": 1.062111801242236, "grad_norm": 0.5739922090044687, "learning_rate": 4.512948613877683e-06, "loss": 0.1595, "step": 1368 }, { "epoch": 1.062888198757764, "grad_norm": 0.6553584142289404, "learning_rate": 4.5068801650567876e-06, "loss": 0.2311, "step": 1369 }, { "epoch": 1.063664596273292, "grad_norm": 0.6006200750889886, "learning_rate": 4.5008124496688925e-06, "loss": 0.2226, "step": 1370 }, { "epoch": 1.06444099378882, "grad_norm": 0.7266001937754981, "learning_rate": 4.4947454767387065e-06, "loss": 0.272, "step": 1371 }, { "epoch": 1.065217391304348, "grad_norm": 0.7142626969697591, "learning_rate": 4.488679255289829e-06, "loss": 0.1299, "step": 1372 }, { "epoch": 1.0659937888198758, "grad_norm": 0.8539908565861214, "learning_rate": 4.48261379434475e-06, "loss": 0.2402, "step": 1373 }, { "epoch": 1.0667701863354038, "grad_norm": 0.5934000465522966, "learning_rate": 4.4765491029248235e-06, "loss": 0.1669, "step": 1374 }, { "epoch": 1.0675465838509317, "grad_norm": 0.6021981278179978, "learning_rate": 4.47048519005026e-06, "loss": 0.2331, "step": 1375 }, { "epoch": 1.0683229813664596, "grad_norm": 0.6332613825420248, "learning_rate": 4.464422064740114e-06, "loss": 0.2169, "step": 1376 }, { "epoch": 1.0690993788819876, "grad_norm": 0.6228965329697439, "learning_rate": 4.458359736012264e-06, "loss": 0.1753, "step": 1377 }, { "epoch": 1.0698757763975155, "grad_norm": 0.5958723683198901, "learning_rate": 4.452298212883408e-06, "loss": 0.181, "step": 1378 }, { "epoch": 1.0706521739130435, "grad_norm": 0.7630976613801054, "learning_rate": 4.446237504369045e-06, "loss": 0.201, "step": 1379 }, { "epoch": 1.0714285714285714, "grad_norm": 0.5823298514963497, "learning_rate": 4.4401776194834615e-06, "loss": 0.2383, "step": 1380 }, { "epoch": 1.0722049689440993, "grad_norm": 0.6657831657637293, "learning_rate": 4.4341185672397185e-06, "loss": 0.2188, "step": 1381 }, { "epoch": 1.0729813664596273, "grad_norm": 0.6566622152974803, "learning_rate": 4.428060356649642e-06, "loss": 0.153, "step": 1382 }, { "epoch": 1.0737577639751552, "grad_norm": 0.9977667602036284, "learning_rate": 4.4220029967237994e-06, "loss": 0.254, "step": 1383 }, { "epoch": 1.0745341614906831, "grad_norm": 0.47584469765205023, "learning_rate": 4.4159464964715e-06, "loss": 0.1364, "step": 1384 }, { "epoch": 1.075310559006211, "grad_norm": 0.6541753700722142, "learning_rate": 4.40989086490077e-06, "loss": 0.1787, "step": 1385 }, { "epoch": 1.0760869565217392, "grad_norm": 0.6512833367840122, "learning_rate": 4.403836111018346e-06, "loss": 0.1809, "step": 1386 }, { "epoch": 1.0768633540372672, "grad_norm": 0.6498541399657408, "learning_rate": 4.397782243829657e-06, "loss": 0.2099, "step": 1387 }, { "epoch": 1.0776397515527951, "grad_norm": 0.721828581990816, "learning_rate": 4.391729272338817e-06, "loss": 0.2417, "step": 1388 }, { "epoch": 1.078416149068323, "grad_norm": 0.4801461340470756, "learning_rate": 4.385677205548603e-06, "loss": 0.1929, "step": 1389 }, { "epoch": 1.079192546583851, "grad_norm": 0.5948055101485626, "learning_rate": 4.37962605246045e-06, "loss": 0.2393, "step": 1390 }, { "epoch": 1.079968944099379, "grad_norm": 0.5799845252576484, "learning_rate": 4.3735758220744336e-06, "loss": 0.2079, "step": 1391 }, { "epoch": 1.0807453416149069, "grad_norm": 0.6593872647998508, "learning_rate": 4.367526523389253e-06, "loss": 0.2191, "step": 1392 }, { "epoch": 1.0815217391304348, "grad_norm": 0.5851244882183909, "learning_rate": 4.361478165402228e-06, "loss": 0.2101, "step": 1393 }, { "epoch": 1.0822981366459627, "grad_norm": 0.633621930184431, "learning_rate": 4.355430757109273e-06, "loss": 0.2501, "step": 1394 }, { "epoch": 1.0830745341614907, "grad_norm": 0.5770828144263015, "learning_rate": 4.3493843075048945e-06, "loss": 0.2618, "step": 1395 }, { "epoch": 1.0838509316770186, "grad_norm": 0.8577169740737467, "learning_rate": 4.343338825582173e-06, "loss": 0.7627, "step": 1396 }, { "epoch": 1.0846273291925466, "grad_norm": 0.7370567481549112, "learning_rate": 4.337294320332747e-06, "loss": 0.1778, "step": 1397 }, { "epoch": 1.0854037267080745, "grad_norm": 0.5342729182165655, "learning_rate": 4.331250800746803e-06, "loss": 0.2144, "step": 1398 }, { "epoch": 1.0861801242236024, "grad_norm": 0.5219282696318805, "learning_rate": 4.325208275813063e-06, "loss": 0.2105, "step": 1399 }, { "epoch": 1.0869565217391304, "grad_norm": 0.6825221866802847, "learning_rate": 4.319166754518768e-06, "loss": 0.1533, "step": 1400 }, { "epoch": 1.0877329192546583, "grad_norm": 0.6288047810200107, "learning_rate": 4.313126245849668e-06, "loss": 0.2306, "step": 1401 }, { "epoch": 1.0885093167701863, "grad_norm": 0.6240490964521019, "learning_rate": 4.307086758790005e-06, "loss": 0.2242, "step": 1402 }, { "epoch": 1.0892857142857142, "grad_norm": 0.6800849495737042, "learning_rate": 4.3010483023225045e-06, "loss": 0.2303, "step": 1403 }, { "epoch": 1.0900621118012421, "grad_norm": 0.8822615501043439, "learning_rate": 4.295010885428354e-06, "loss": 0.256, "step": 1404 }, { "epoch": 1.0908385093167703, "grad_norm": 0.7864374623845388, "learning_rate": 4.2889745170871975e-06, "loss": 0.1557, "step": 1405 }, { "epoch": 1.0916149068322982, "grad_norm": 0.6353304571375447, "learning_rate": 4.282939206277126e-06, "loss": 0.1467, "step": 1406 }, { "epoch": 1.0923913043478262, "grad_norm": 0.5816585491401296, "learning_rate": 4.276904961974645e-06, "loss": 0.2405, "step": 1407 }, { "epoch": 1.093167701863354, "grad_norm": 0.570133141588144, "learning_rate": 4.270871793154683e-06, "loss": 0.1474, "step": 1408 }, { "epoch": 1.093944099378882, "grad_norm": 0.7778127387223461, "learning_rate": 4.264839708790565e-06, "loss": 0.2233, "step": 1409 }, { "epoch": 1.09472049689441, "grad_norm": 0.6434127981851007, "learning_rate": 4.258808717854006e-06, "loss": 0.1535, "step": 1410 }, { "epoch": 1.095496894409938, "grad_norm": 0.6547061387797319, "learning_rate": 4.252778829315093e-06, "loss": 0.1865, "step": 1411 }, { "epoch": 1.0962732919254659, "grad_norm": 0.6239117669963112, "learning_rate": 4.2467500521422696e-06, "loss": 0.2214, "step": 1412 }, { "epoch": 1.0970496894409938, "grad_norm": 0.46189400967730915, "learning_rate": 4.240722395302334e-06, "loss": 0.1289, "step": 1413 }, { "epoch": 1.0978260869565217, "grad_norm": 0.5751684001478858, "learning_rate": 4.234695867760412e-06, "loss": 0.2338, "step": 1414 }, { "epoch": 1.0986024844720497, "grad_norm": 0.7219659475749641, "learning_rate": 4.228670478479952e-06, "loss": 0.2192, "step": 1415 }, { "epoch": 1.0993788819875776, "grad_norm": 0.6566128431574525, "learning_rate": 4.222646236422711e-06, "loss": 0.2778, "step": 1416 }, { "epoch": 1.1001552795031055, "grad_norm": 0.529298389173228, "learning_rate": 4.216623150548738e-06, "loss": 0.2142, "step": 1417 }, { "epoch": 1.1009316770186335, "grad_norm": 0.6783091114062795, "learning_rate": 4.21060122981636e-06, "loss": 0.2021, "step": 1418 }, { "epoch": 1.1017080745341614, "grad_norm": 0.7598876751355712, "learning_rate": 4.204580483182176e-06, "loss": 0.1844, "step": 1419 }, { "epoch": 1.1024844720496894, "grad_norm": 0.7310722978207778, "learning_rate": 4.198560919601034e-06, "loss": 0.215, "step": 1420 }, { "epoch": 1.1032608695652173, "grad_norm": 0.49748162812499297, "learning_rate": 4.192542548026027e-06, "loss": 0.1618, "step": 1421 }, { "epoch": 1.1040372670807455, "grad_norm": 0.6420607784471071, "learning_rate": 4.186525377408471e-06, "loss": 0.2126, "step": 1422 }, { "epoch": 1.1048136645962734, "grad_norm": 0.5530904909565056, "learning_rate": 4.180509416697896e-06, "loss": 0.178, "step": 1423 }, { "epoch": 1.1055900621118013, "grad_norm": 0.6789304953028978, "learning_rate": 4.174494674842038e-06, "loss": 0.2088, "step": 1424 }, { "epoch": 1.1063664596273293, "grad_norm": 0.6577440011201212, "learning_rate": 4.168481160786815e-06, "loss": 0.1534, "step": 1425 }, { "epoch": 1.1071428571428572, "grad_norm": 0.6658287750995232, "learning_rate": 4.162468883476319e-06, "loss": 0.2284, "step": 1426 }, { "epoch": 1.1079192546583851, "grad_norm": 0.651632830689297, "learning_rate": 4.156457851852802e-06, "loss": 0.1643, "step": 1427 }, { "epoch": 1.108695652173913, "grad_norm": 0.6126864038321018, "learning_rate": 4.150448074856667e-06, "loss": 0.1937, "step": 1428 }, { "epoch": 1.109472049689441, "grad_norm": 0.5766679985354151, "learning_rate": 4.144439561426448e-06, "loss": 0.1962, "step": 1429 }, { "epoch": 1.110248447204969, "grad_norm": 0.6019446125484941, "learning_rate": 4.138432320498801e-06, "loss": 0.2093, "step": 1430 }, { "epoch": 1.111024844720497, "grad_norm": 0.7451023124894438, "learning_rate": 4.132426361008487e-06, "loss": 0.1919, "step": 1431 }, { "epoch": 1.1118012422360248, "grad_norm": 0.69991578283543, "learning_rate": 4.126421691888366e-06, "loss": 0.1804, "step": 1432 }, { "epoch": 1.1125776397515528, "grad_norm": 0.6207767776555319, "learning_rate": 4.120418322069372e-06, "loss": 0.2036, "step": 1433 }, { "epoch": 1.1133540372670807, "grad_norm": 0.5260862456627268, "learning_rate": 4.114416260480512e-06, "loss": 0.1897, "step": 1434 }, { "epoch": 1.1141304347826086, "grad_norm": 0.7053982915261502, "learning_rate": 4.1084155160488445e-06, "loss": 0.2055, "step": 1435 }, { "epoch": 1.1149068322981366, "grad_norm": 0.5427356907460346, "learning_rate": 4.102416097699471e-06, "loss": 0.2283, "step": 1436 }, { "epoch": 1.1156832298136645, "grad_norm": 0.8472316436739602, "learning_rate": 4.096418014355518e-06, "loss": 0.8291, "step": 1437 }, { "epoch": 1.1164596273291925, "grad_norm": 0.7988813741901943, "learning_rate": 4.09042127493813e-06, "loss": 0.2334, "step": 1438 }, { "epoch": 1.1172360248447204, "grad_norm": 0.6380724830749165, "learning_rate": 4.084425888366449e-06, "loss": 0.2347, "step": 1439 }, { "epoch": 1.1180124223602483, "grad_norm": 0.714070655547183, "learning_rate": 4.078431863557605e-06, "loss": 0.2124, "step": 1440 }, { "epoch": 1.1187888198757765, "grad_norm": 0.7011165294373944, "learning_rate": 4.072439209426706e-06, "loss": 0.1896, "step": 1441 }, { "epoch": 1.1195652173913044, "grad_norm": 0.6307158111742209, "learning_rate": 4.066447934886819e-06, "loss": 0.1753, "step": 1442 }, { "epoch": 1.1203416149068324, "grad_norm": 0.715134453108934, "learning_rate": 4.060458048848959e-06, "loss": 0.1366, "step": 1443 }, { "epoch": 1.1211180124223603, "grad_norm": 0.6278973701596383, "learning_rate": 4.054469560222077e-06, "loss": 0.2389, "step": 1444 }, { "epoch": 1.1218944099378882, "grad_norm": 0.5760750733884931, "learning_rate": 4.048482477913045e-06, "loss": 0.2215, "step": 1445 }, { "epoch": 1.1226708074534162, "grad_norm": 0.6264306814609263, "learning_rate": 4.042496810826641e-06, "loss": 0.1339, "step": 1446 }, { "epoch": 1.1234472049689441, "grad_norm": 0.5921253887666209, "learning_rate": 4.036512567865545e-06, "loss": 0.1873, "step": 1447 }, { "epoch": 1.124223602484472, "grad_norm": 0.5841688136613704, "learning_rate": 4.030529757930308e-06, "loss": 0.2793, "step": 1448 }, { "epoch": 1.125, "grad_norm": 0.6376351987674348, "learning_rate": 4.02454838991936e-06, "loss": 0.1977, "step": 1449 }, { "epoch": 1.125776397515528, "grad_norm": 0.7351400765681381, "learning_rate": 4.018568472728979e-06, "loss": 0.2393, "step": 1450 }, { "epoch": 1.1265527950310559, "grad_norm": 0.7963200702721851, "learning_rate": 4.012590015253292e-06, "loss": 0.2929, "step": 1451 }, { "epoch": 1.1273291925465838, "grad_norm": 0.6986618536383545, "learning_rate": 4.006613026384249e-06, "loss": 0.2005, "step": 1452 }, { "epoch": 1.1281055900621118, "grad_norm": 1.005106518899402, "learning_rate": 4.000637515011617e-06, "loss": 0.7565, "step": 1453 }, { "epoch": 1.1288819875776397, "grad_norm": 0.7913981049200507, "learning_rate": 3.9946634900229685e-06, "loss": 0.2108, "step": 1454 }, { "epoch": 1.1296583850931676, "grad_norm": 0.5469640118796705, "learning_rate": 3.988690960303662e-06, "loss": 0.1957, "step": 1455 }, { "epoch": 1.1304347826086956, "grad_norm": 0.6981375494045127, "learning_rate": 3.982719934736832e-06, "loss": 0.1794, "step": 1456 }, { "epoch": 1.1312111801242235, "grad_norm": 0.6446656924570747, "learning_rate": 3.976750422203379e-06, "loss": 0.1669, "step": 1457 }, { "epoch": 1.1319875776397517, "grad_norm": 0.5895275792483008, "learning_rate": 3.970782431581949e-06, "loss": 0.2534, "step": 1458 }, { "epoch": 1.1327639751552794, "grad_norm": 0.7326399216131159, "learning_rate": 3.964815971748929e-06, "loss": 0.159, "step": 1459 }, { "epoch": 1.1335403726708075, "grad_norm": 0.7682528821396283, "learning_rate": 3.958851051578425e-06, "loss": 0.2167, "step": 1460 }, { "epoch": 1.1343167701863355, "grad_norm": 2.3461028035474065, "learning_rate": 3.952887679942257e-06, "loss": 0.4855, "step": 1461 }, { "epoch": 1.1350931677018634, "grad_norm": 0.5785945544716371, "learning_rate": 3.946925865709936e-06, "loss": 0.2082, "step": 1462 }, { "epoch": 1.1358695652173914, "grad_norm": 0.5738387520838932, "learning_rate": 3.940965617748662e-06, "loss": 0.2295, "step": 1463 }, { "epoch": 1.1366459627329193, "grad_norm": 0.5338070030486437, "learning_rate": 3.935006944923304e-06, "loss": 0.2235, "step": 1464 }, { "epoch": 1.1374223602484472, "grad_norm": 0.643734026553885, "learning_rate": 3.929049856096387e-06, "loss": 0.168, "step": 1465 }, { "epoch": 1.1381987577639752, "grad_norm": 0.904902019730716, "learning_rate": 3.923094360128081e-06, "loss": 0.7234, "step": 1466 }, { "epoch": 1.138975155279503, "grad_norm": 0.576924310153994, "learning_rate": 3.917140465876188e-06, "loss": 0.2358, "step": 1467 }, { "epoch": 1.139751552795031, "grad_norm": 0.614539680708074, "learning_rate": 3.911188182196124e-06, "loss": 0.2248, "step": 1468 }, { "epoch": 1.140527950310559, "grad_norm": 0.6906632973289419, "learning_rate": 3.905237517940915e-06, "loss": 0.1929, "step": 1469 }, { "epoch": 1.141304347826087, "grad_norm": 0.8529810559606029, "learning_rate": 3.899288481961173e-06, "loss": 0.212, "step": 1470 }, { "epoch": 1.1420807453416149, "grad_norm": 0.8706035310724406, "learning_rate": 3.893341083105092e-06, "loss": 0.2354, "step": 1471 }, { "epoch": 1.1428571428571428, "grad_norm": 0.7360410091726284, "learning_rate": 3.887395330218429e-06, "loss": 0.2113, "step": 1472 }, { "epoch": 1.1436335403726707, "grad_norm": 0.6322352821428924, "learning_rate": 3.881451232144495e-06, "loss": 0.1412, "step": 1473 }, { "epoch": 1.1444099378881987, "grad_norm": 0.7289087014921535, "learning_rate": 3.875508797724135e-06, "loss": 0.1586, "step": 1474 }, { "epoch": 1.1451863354037268, "grad_norm": 0.5763308674133177, "learning_rate": 3.869568035795728e-06, "loss": 0.2058, "step": 1475 }, { "epoch": 1.1459627329192545, "grad_norm": 0.6386277725838595, "learning_rate": 3.8636289551951545e-06, "loss": 0.221, "step": 1476 }, { "epoch": 1.1467391304347827, "grad_norm": 0.6472174900338679, "learning_rate": 3.857691564755805e-06, "loss": 0.1932, "step": 1477 }, { "epoch": 1.1475155279503106, "grad_norm": 0.6419091590287029, "learning_rate": 3.851755873308549e-06, "loss": 0.1787, "step": 1478 }, { "epoch": 1.1482919254658386, "grad_norm": 0.724870331796381, "learning_rate": 3.8458218896817325e-06, "loss": 0.1686, "step": 1479 }, { "epoch": 1.1490683229813665, "grad_norm": 0.5466432702982377, "learning_rate": 3.8398896227011604e-06, "loss": 0.2292, "step": 1480 }, { "epoch": 1.1498447204968945, "grad_norm": 0.634014687319818, "learning_rate": 3.833959081190083e-06, "loss": 0.1188, "step": 1481 }, { "epoch": 1.1506211180124224, "grad_norm": 0.6328825062912516, "learning_rate": 3.8280302739691874e-06, "loss": 0.173, "step": 1482 }, { "epoch": 1.1513975155279503, "grad_norm": 0.6964200342867122, "learning_rate": 3.822103209856578e-06, "loss": 0.2009, "step": 1483 }, { "epoch": 1.1521739130434783, "grad_norm": 0.49551077969980595, "learning_rate": 3.816177897667767e-06, "loss": 0.2113, "step": 1484 }, { "epoch": 1.1529503105590062, "grad_norm": 0.5591375820826269, "learning_rate": 3.8102543462156634e-06, "loss": 0.1325, "step": 1485 }, { "epoch": 1.1537267080745341, "grad_norm": 0.6567518871323117, "learning_rate": 3.8043325643105554e-06, "loss": 0.1474, "step": 1486 }, { "epoch": 1.154503105590062, "grad_norm": 0.6132458001449259, "learning_rate": 3.798412560760101e-06, "loss": 0.169, "step": 1487 }, { "epoch": 1.15527950310559, "grad_norm": 0.6710897877326639, "learning_rate": 3.792494344369311e-06, "loss": 0.1659, "step": 1488 }, { "epoch": 1.156055900621118, "grad_norm": 0.5863501728272343, "learning_rate": 3.786577923940538e-06, "loss": 0.1759, "step": 1489 }, { "epoch": 1.156832298136646, "grad_norm": 0.6455293411581026, "learning_rate": 3.780663308273466e-06, "loss": 0.2101, "step": 1490 }, { "epoch": 1.1576086956521738, "grad_norm": 0.5684645192398646, "learning_rate": 3.774750506165092e-06, "loss": 0.1576, "step": 1491 }, { "epoch": 1.1583850931677018, "grad_norm": 0.570464193582543, "learning_rate": 3.7688395264097177e-06, "loss": 0.2401, "step": 1492 }, { "epoch": 1.1591614906832297, "grad_norm": 0.6525937779922868, "learning_rate": 3.762930377798935e-06, "loss": 0.2337, "step": 1493 }, { "epoch": 1.1599378881987579, "grad_norm": 0.5090909272179148, "learning_rate": 3.7570230691216055e-06, "loss": 0.2434, "step": 1494 }, { "epoch": 1.1607142857142858, "grad_norm": 0.6901437746018023, "learning_rate": 3.751117609163865e-06, "loss": 0.1531, "step": 1495 }, { "epoch": 1.1614906832298137, "grad_norm": 0.5282781690488967, "learning_rate": 3.7452140067090936e-06, "loss": 0.1649, "step": 1496 }, { "epoch": 1.1622670807453417, "grad_norm": 0.6803129353498367, "learning_rate": 3.739312270537907e-06, "loss": 0.2035, "step": 1497 }, { "epoch": 1.1630434782608696, "grad_norm": 0.798670049110019, "learning_rate": 3.733412409428148e-06, "loss": 0.1819, "step": 1498 }, { "epoch": 1.1638198757763976, "grad_norm": 0.6274198966978077, "learning_rate": 3.727514432154871e-06, "loss": 0.2288, "step": 1499 }, { "epoch": 1.1645962732919255, "grad_norm": 0.756938505511367, "learning_rate": 3.7216183474903243e-06, "loss": 0.2351, "step": 1500 }, { "epoch": 1.1645962732919255, "eval_loss": 0.31522291898727417, "eval_runtime": 3.9862, "eval_samples_per_second": 13.296, "eval_steps_per_second": 3.512, "step": 1500 }, { "epoch": 1.1653726708074534, "grad_norm": 0.6481971774363438, "learning_rate": 3.715724164203948e-06, "loss": 0.232, "step": 1501 }, { "epoch": 1.1661490683229814, "grad_norm": 0.7132356501385287, "learning_rate": 3.70983189106235e-06, "loss": 0.1078, "step": 1502 }, { "epoch": 1.1669254658385093, "grad_norm": 1.0445142590001233, "learning_rate": 3.7039415368292954e-06, "loss": 0.7165, "step": 1503 }, { "epoch": 1.1677018633540373, "grad_norm": 0.7517750314219066, "learning_rate": 3.698053110265699e-06, "loss": 0.2132, "step": 1504 }, { "epoch": 1.1684782608695652, "grad_norm": 0.6538742839807546, "learning_rate": 3.692166620129608e-06, "loss": 0.1749, "step": 1505 }, { "epoch": 1.1692546583850931, "grad_norm": 0.7940639096392335, "learning_rate": 3.6862820751761863e-06, "loss": 0.2153, "step": 1506 }, { "epoch": 1.170031055900621, "grad_norm": 0.7432961396988097, "learning_rate": 3.6803994841577097e-06, "loss": 0.2686, "step": 1507 }, { "epoch": 1.170807453416149, "grad_norm": 0.7013683088644364, "learning_rate": 3.6745188558235443e-06, "loss": 0.2399, "step": 1508 }, { "epoch": 1.171583850931677, "grad_norm": 0.5026323448112697, "learning_rate": 3.668640198920136e-06, "loss": 0.1646, "step": 1509 }, { "epoch": 1.1723602484472049, "grad_norm": 0.6701887253080663, "learning_rate": 3.662763522191002e-06, "loss": 0.1646, "step": 1510 }, { "epoch": 1.173136645962733, "grad_norm": 0.5439118008943431, "learning_rate": 3.656888834376713e-06, "loss": 0.2155, "step": 1511 }, { "epoch": 1.1739130434782608, "grad_norm": 0.5853198929796543, "learning_rate": 3.6510161442148783e-06, "loss": 0.2141, "step": 1512 }, { "epoch": 1.174689440993789, "grad_norm": 0.5884476543839349, "learning_rate": 3.6451454604401435e-06, "loss": 0.1947, "step": 1513 }, { "epoch": 1.1754658385093169, "grad_norm": 0.9222024643720622, "learning_rate": 3.639276791784163e-06, "loss": 0.6862, "step": 1514 }, { "epoch": 1.1762422360248448, "grad_norm": 0.7306441922385069, "learning_rate": 3.6334101469755973e-06, "loss": 0.2366, "step": 1515 }, { "epoch": 1.1770186335403727, "grad_norm": 0.5666352368436661, "learning_rate": 3.6275455347400952e-06, "loss": 0.1971, "step": 1516 }, { "epoch": 1.1777950310559007, "grad_norm": 0.5517702010828616, "learning_rate": 3.621682963800285e-06, "loss": 0.1962, "step": 1517 }, { "epoch": 1.1785714285714286, "grad_norm": 0.6791807937794904, "learning_rate": 3.6158224428757538e-06, "loss": 0.2148, "step": 1518 }, { "epoch": 1.1793478260869565, "grad_norm": 0.5082677646321153, "learning_rate": 3.6099639806830456e-06, "loss": 0.2035, "step": 1519 }, { "epoch": 1.1801242236024845, "grad_norm": 0.6127484028172472, "learning_rate": 3.6041075859356383e-06, "loss": 0.2307, "step": 1520 }, { "epoch": 1.1809006211180124, "grad_norm": 0.5616065307358075, "learning_rate": 3.5982532673439376e-06, "loss": 0.1568, "step": 1521 }, { "epoch": 1.1816770186335404, "grad_norm": 0.5506500485740522, "learning_rate": 3.592401033615259e-06, "loss": 0.1427, "step": 1522 }, { "epoch": 1.1824534161490683, "grad_norm": 0.8318116322362841, "learning_rate": 3.586550893453819e-06, "loss": 0.1855, "step": 1523 }, { "epoch": 1.1832298136645962, "grad_norm": 0.6547177856297937, "learning_rate": 3.580702855560718e-06, "loss": 0.2217, "step": 1524 }, { "epoch": 1.1840062111801242, "grad_norm": 0.5757545875179495, "learning_rate": 3.574856928633931e-06, "loss": 0.2357, "step": 1525 }, { "epoch": 1.184782608695652, "grad_norm": 0.6071284807904894, "learning_rate": 3.5690131213682943e-06, "loss": 0.2149, "step": 1526 }, { "epoch": 1.18555900621118, "grad_norm": 0.6375633600651719, "learning_rate": 3.5631714424554903e-06, "loss": 0.2047, "step": 1527 }, { "epoch": 1.186335403726708, "grad_norm": 0.8289580419772019, "learning_rate": 3.5573319005840363e-06, "loss": 0.2621, "step": 1528 }, { "epoch": 1.187111801242236, "grad_norm": 0.47716844148367155, "learning_rate": 3.551494504439269e-06, "loss": 0.17, "step": 1529 }, { "epoch": 1.187888198757764, "grad_norm": 0.6540102803458754, "learning_rate": 3.5456592627033355e-06, "loss": 0.293, "step": 1530 }, { "epoch": 1.188664596273292, "grad_norm": 0.8327085588993902, "learning_rate": 3.5398261840551817e-06, "loss": 0.2313, "step": 1531 }, { "epoch": 1.18944099378882, "grad_norm": 0.5858223883784479, "learning_rate": 3.533995277170532e-06, "loss": 0.2377, "step": 1532 }, { "epoch": 1.190217391304348, "grad_norm": 0.5516404957399876, "learning_rate": 3.5281665507218804e-06, "loss": 0.1936, "step": 1533 }, { "epoch": 1.1909937888198758, "grad_norm": 0.5309769559068138, "learning_rate": 3.52234001337848e-06, "loss": 0.2312, "step": 1534 }, { "epoch": 1.1917701863354038, "grad_norm": 0.7854504347642042, "learning_rate": 3.5165156738063276e-06, "loss": 0.212, "step": 1535 }, { "epoch": 1.1925465838509317, "grad_norm": 0.8342146621899854, "learning_rate": 3.510693540668151e-06, "loss": 0.1636, "step": 1536 }, { "epoch": 1.1933229813664596, "grad_norm": 0.6555698848662207, "learning_rate": 3.504873622623396e-06, "loss": 0.2283, "step": 1537 }, { "epoch": 1.1940993788819876, "grad_norm": 0.5588015144989423, "learning_rate": 3.4990559283282133e-06, "loss": 0.2455, "step": 1538 }, { "epoch": 1.1948757763975155, "grad_norm": 0.6398105776428774, "learning_rate": 3.4932404664354473e-06, "loss": 0.1948, "step": 1539 }, { "epoch": 1.1956521739130435, "grad_norm": 0.6262031485162681, "learning_rate": 3.4874272455946217e-06, "loss": 0.2237, "step": 1540 }, { "epoch": 1.1964285714285714, "grad_norm": 0.660857455387905, "learning_rate": 3.4816162744519266e-06, "loss": 0.1577, "step": 1541 }, { "epoch": 1.1972049689440993, "grad_norm": 0.7472114156052007, "learning_rate": 3.4758075616502075e-06, "loss": 0.2131, "step": 1542 }, { "epoch": 1.1979813664596273, "grad_norm": 0.5853262683907036, "learning_rate": 3.4700011158289494e-06, "loss": 0.2722, "step": 1543 }, { "epoch": 1.1987577639751552, "grad_norm": 0.6234735569564842, "learning_rate": 3.4641969456242646e-06, "loss": 0.2243, "step": 1544 }, { "epoch": 1.1995341614906831, "grad_norm": 0.6982156840349234, "learning_rate": 3.458395059668883e-06, "loss": 0.2795, "step": 1545 }, { "epoch": 1.200310559006211, "grad_norm": 0.6760051221275307, "learning_rate": 3.4525954665921375e-06, "loss": 0.2089, "step": 1546 }, { "epoch": 1.2010869565217392, "grad_norm": 0.6551104516988796, "learning_rate": 3.446798175019948e-06, "loss": 0.2654, "step": 1547 }, { "epoch": 1.201863354037267, "grad_norm": 0.6348238677931224, "learning_rate": 3.4410031935748124e-06, "loss": 0.1969, "step": 1548 }, { "epoch": 1.2026397515527951, "grad_norm": 0.6464025046923142, "learning_rate": 3.435210530875795e-06, "loss": 0.2137, "step": 1549 }, { "epoch": 1.203416149068323, "grad_norm": 0.6949978138645053, "learning_rate": 3.4294201955385086e-06, "loss": 0.2381, "step": 1550 }, { "epoch": 1.204192546583851, "grad_norm": 0.5078885522305696, "learning_rate": 3.423632196175105e-06, "loss": 0.2188, "step": 1551 }, { "epoch": 1.204968944099379, "grad_norm": 1.0724629949158453, "learning_rate": 3.4178465413942628e-06, "loss": 0.7602, "step": 1552 }, { "epoch": 1.2057453416149069, "grad_norm": 0.5623349573772817, "learning_rate": 3.412063239801171e-06, "loss": 0.1809, "step": 1553 }, { "epoch": 1.2065217391304348, "grad_norm": 0.5921753421417463, "learning_rate": 3.406282299997521e-06, "loss": 0.2678, "step": 1554 }, { "epoch": 1.2072981366459627, "grad_norm": 0.5777245444404734, "learning_rate": 3.4005037305814907e-06, "loss": 0.237, "step": 1555 }, { "epoch": 1.2080745341614907, "grad_norm": 0.5701609724995311, "learning_rate": 3.394727540147732e-06, "loss": 0.1411, "step": 1556 }, { "epoch": 1.2088509316770186, "grad_norm": 0.768770090049981, "learning_rate": 3.388953737287359e-06, "loss": 0.2077, "step": 1557 }, { "epoch": 1.2096273291925466, "grad_norm": 0.5712413553047341, "learning_rate": 3.383182330587936e-06, "loss": 0.1901, "step": 1558 }, { "epoch": 1.2104037267080745, "grad_norm": 0.7431373461964966, "learning_rate": 3.377413328633459e-06, "loss": 0.1623, "step": 1559 }, { "epoch": 1.2111801242236024, "grad_norm": 0.6502304883310449, "learning_rate": 3.3716467400043514e-06, "loss": 0.1425, "step": 1560 }, { "epoch": 1.2119565217391304, "grad_norm": 0.6873713990121193, "learning_rate": 3.3658825732774457e-06, "loss": 0.2634, "step": 1561 }, { "epoch": 1.2127329192546583, "grad_norm": 1.1074364550397804, "learning_rate": 3.360120837025974e-06, "loss": 0.7283, "step": 1562 }, { "epoch": 1.2135093167701863, "grad_norm": 0.7324483352590833, "learning_rate": 3.3543615398195518e-06, "loss": 0.2314, "step": 1563 }, { "epoch": 1.2142857142857142, "grad_norm": 0.647422222934075, "learning_rate": 3.3486046902241663e-06, "loss": 0.2244, "step": 1564 }, { "epoch": 1.2150621118012421, "grad_norm": 0.5697274420691223, "learning_rate": 3.3428502968021647e-06, "loss": 0.1732, "step": 1565 }, { "epoch": 1.2158385093167703, "grad_norm": 0.6474355763037578, "learning_rate": 3.3370983681122405e-06, "loss": 0.2227, "step": 1566 }, { "epoch": 1.2166149068322982, "grad_norm": 0.5681646907482077, "learning_rate": 3.3313489127094267e-06, "loss": 0.1356, "step": 1567 }, { "epoch": 1.2173913043478262, "grad_norm": 0.6429035182400586, "learning_rate": 3.3256019391450696e-06, "loss": 0.1773, "step": 1568 }, { "epoch": 1.218167701863354, "grad_norm": 0.7328625197048463, "learning_rate": 3.319857455966829e-06, "loss": 0.1673, "step": 1569 }, { "epoch": 1.218944099378882, "grad_norm": 0.6743977037428598, "learning_rate": 3.3141154717186597e-06, "loss": 0.2031, "step": 1570 }, { "epoch": 1.21972049689441, "grad_norm": 0.6299900448539044, "learning_rate": 3.308375994940799e-06, "loss": 0.2324, "step": 1571 }, { "epoch": 1.220496894409938, "grad_norm": 0.6956933381662846, "learning_rate": 3.302639034169758e-06, "loss": 0.1263, "step": 1572 }, { "epoch": 1.2212732919254659, "grad_norm": 0.6688336583126787, "learning_rate": 3.2969045979382984e-06, "loss": 0.2085, "step": 1573 }, { "epoch": 1.2220496894409938, "grad_norm": 0.7139491101561722, "learning_rate": 3.291172694775435e-06, "loss": 0.2175, "step": 1574 }, { "epoch": 1.2228260869565217, "grad_norm": 0.7525151862313993, "learning_rate": 3.2854433332064116e-06, "loss": 0.2152, "step": 1575 }, { "epoch": 1.2236024844720497, "grad_norm": 0.7644012880557404, "learning_rate": 3.2797165217526915e-06, "loss": 0.2696, "step": 1576 }, { "epoch": 1.2243788819875776, "grad_norm": 0.7371848596701032, "learning_rate": 3.2739922689319453e-06, "loss": 0.1806, "step": 1577 }, { "epoch": 1.2251552795031055, "grad_norm": 0.7833895371772298, "learning_rate": 3.2682705832580395e-06, "loss": 0.1504, "step": 1578 }, { "epoch": 1.2259316770186335, "grad_norm": 0.5379207487083036, "learning_rate": 3.26255147324102e-06, "loss": 0.2097, "step": 1579 }, { "epoch": 1.2267080745341614, "grad_norm": 0.5533378811987465, "learning_rate": 3.2568349473871044e-06, "loss": 0.1616, "step": 1580 }, { "epoch": 1.2274844720496894, "grad_norm": 0.6824906540627445, "learning_rate": 3.251121014198664e-06, "loss": 0.2058, "step": 1581 }, { "epoch": 1.2282608695652173, "grad_norm": 0.7262089611152323, "learning_rate": 3.245409682174217e-06, "loss": 0.2666, "step": 1582 }, { "epoch": 1.2290372670807455, "grad_norm": 0.7117272483074153, "learning_rate": 3.2397009598084106e-06, "loss": 0.2149, "step": 1583 }, { "epoch": 1.2298136645962732, "grad_norm": 0.6075745898837905, "learning_rate": 3.2339948555920103e-06, "loss": 0.2259, "step": 1584 }, { "epoch": 1.2305900621118013, "grad_norm": 0.5936949355304739, "learning_rate": 3.2282913780118906e-06, "loss": 0.2985, "step": 1585 }, { "epoch": 1.2313664596273293, "grad_norm": 0.5458240927150682, "learning_rate": 3.222590535551017e-06, "loss": 0.235, "step": 1586 }, { "epoch": 1.2321428571428572, "grad_norm": 0.8058226756294579, "learning_rate": 3.216892336688435e-06, "loss": 0.2471, "step": 1587 }, { "epoch": 1.2329192546583851, "grad_norm": 0.7467521488043476, "learning_rate": 3.2111967898992597e-06, "loss": 0.1912, "step": 1588 }, { "epoch": 1.233695652173913, "grad_norm": 0.5757441938726231, "learning_rate": 3.20550390365466e-06, "loss": 0.1915, "step": 1589 }, { "epoch": 1.234472049689441, "grad_norm": 0.6538288982991125, "learning_rate": 3.1998136864218497e-06, "loss": 0.2476, "step": 1590 }, { "epoch": 1.235248447204969, "grad_norm": 0.7365181714551577, "learning_rate": 3.1941261466640715e-06, "loss": 0.1837, "step": 1591 }, { "epoch": 1.236024844720497, "grad_norm": 0.5661557256541736, "learning_rate": 3.188441292840587e-06, "loss": 0.2549, "step": 1592 }, { "epoch": 1.2368012422360248, "grad_norm": 0.7713049327574409, "learning_rate": 3.1827591334066636e-06, "loss": 0.2683, "step": 1593 }, { "epoch": 1.2375776397515528, "grad_norm": 0.7290962641661689, "learning_rate": 3.177079676813557e-06, "loss": 0.1637, "step": 1594 }, { "epoch": 1.2383540372670807, "grad_norm": 0.7442456517582374, "learning_rate": 3.171402931508506e-06, "loss": 0.1887, "step": 1595 }, { "epoch": 1.2391304347826086, "grad_norm": 0.654794730935946, "learning_rate": 3.1657289059347184e-06, "loss": 0.1591, "step": 1596 }, { "epoch": 1.2399068322981366, "grad_norm": 0.7105392007706035, "learning_rate": 3.160057608531354e-06, "loss": 0.2102, "step": 1597 }, { "epoch": 1.2406832298136645, "grad_norm": 0.6583344998630074, "learning_rate": 3.1543890477335153e-06, "loss": 0.1687, "step": 1598 }, { "epoch": 1.2414596273291925, "grad_norm": 0.6354233810408771, "learning_rate": 3.1487232319722372e-06, "loss": 0.1454, "step": 1599 }, { "epoch": 1.2422360248447206, "grad_norm": 0.6101099634521301, "learning_rate": 3.1430601696744676e-06, "loss": 0.1675, "step": 1600 }, { "epoch": 1.2430124223602483, "grad_norm": 0.6288306757324621, "learning_rate": 3.1373998692630627e-06, "loss": 0.1729, "step": 1601 }, { "epoch": 1.2437888198757765, "grad_norm": 0.7635760620038479, "learning_rate": 3.131742339156768e-06, "loss": 0.2125, "step": 1602 }, { "epoch": 1.2445652173913044, "grad_norm": 0.5283108435010211, "learning_rate": 3.1260875877702113e-06, "loss": 0.2044, "step": 1603 }, { "epoch": 1.2453416149068324, "grad_norm": 1.3701807126885388, "learning_rate": 3.120435623513887e-06, "loss": 0.7719, "step": 1604 }, { "epoch": 1.2461180124223603, "grad_norm": 0.5897065814918718, "learning_rate": 3.114786454794141e-06, "loss": 0.1999, "step": 1605 }, { "epoch": 1.2468944099378882, "grad_norm": 0.6422717557427315, "learning_rate": 3.1091400900131662e-06, "loss": 0.1932, "step": 1606 }, { "epoch": 1.2476708074534162, "grad_norm": 0.5122349487965251, "learning_rate": 3.1034965375689817e-06, "loss": 0.2015, "step": 1607 }, { "epoch": 1.2484472049689441, "grad_norm": 0.6096190260562351, "learning_rate": 3.0978558058554255e-06, "loss": 0.2788, "step": 1608 }, { "epoch": 1.249223602484472, "grad_norm": 0.6887785031099344, "learning_rate": 3.092217903262137e-06, "loss": 0.2034, "step": 1609 }, { "epoch": 1.25, "grad_norm": 0.579631441238835, "learning_rate": 3.0865828381745515e-06, "loss": 0.2189, "step": 1610 }, { "epoch": 1.250776397515528, "grad_norm": 0.6758191425814526, "learning_rate": 3.080950618973884e-06, "loss": 0.178, "step": 1611 }, { "epoch": 1.2515527950310559, "grad_norm": 0.6746771974165483, "learning_rate": 3.0753212540371126e-06, "loss": 0.2278, "step": 1612 }, { "epoch": 1.2523291925465838, "grad_norm": 0.625255096538184, "learning_rate": 3.0696947517369756e-06, "loss": 0.1806, "step": 1613 }, { "epoch": 1.2531055900621118, "grad_norm": 0.6667335567569667, "learning_rate": 3.0640711204419495e-06, "loss": 0.2111, "step": 1614 }, { "epoch": 1.2538819875776397, "grad_norm": 0.581272726775766, "learning_rate": 3.058450368516242e-06, "loss": 0.1832, "step": 1615 }, { "epoch": 1.2546583850931676, "grad_norm": 0.7167905124485063, "learning_rate": 3.0528325043197786e-06, "loss": 0.1715, "step": 1616 }, { "epoch": 1.2554347826086958, "grad_norm": 0.6810625000376773, "learning_rate": 3.0472175362081896e-06, "loss": 0.227, "step": 1617 }, { "epoch": 1.2562111801242235, "grad_norm": 0.826113020614311, "learning_rate": 3.0416054725327975e-06, "loss": 0.691, "step": 1618 }, { "epoch": 1.2569875776397517, "grad_norm": 0.8320596087861548, "learning_rate": 3.0359963216406046e-06, "loss": 0.2908, "step": 1619 }, { "epoch": 1.2577639751552794, "grad_norm": 0.6799648418195353, "learning_rate": 3.0303900918742832e-06, "loss": 0.1728, "step": 1620 }, { "epoch": 1.2585403726708075, "grad_norm": 0.6139278438302413, "learning_rate": 3.0247867915721583e-06, "loss": 0.1673, "step": 1621 }, { "epoch": 1.2593167701863355, "grad_norm": 0.5859631359329678, "learning_rate": 3.0191864290682006e-06, "loss": 0.1287, "step": 1622 }, { "epoch": 1.2600931677018634, "grad_norm": 0.6898549256695403, "learning_rate": 3.0135890126920054e-06, "loss": 0.2021, "step": 1623 }, { "epoch": 1.2608695652173914, "grad_norm": 0.6518956428467594, "learning_rate": 3.007994550768793e-06, "loss": 0.2373, "step": 1624 }, { "epoch": 1.2616459627329193, "grad_norm": 0.5758537542964383, "learning_rate": 3.0024030516193856e-06, "loss": 0.1445, "step": 1625 }, { "epoch": 1.2624223602484472, "grad_norm": 0.6223864944321574, "learning_rate": 2.9968145235602014e-06, "loss": 0.1821, "step": 1626 }, { "epoch": 1.2631987577639752, "grad_norm": 0.9602650729191626, "learning_rate": 2.9912289749032362e-06, "loss": 0.7685, "step": 1627 }, { "epoch": 1.263975155279503, "grad_norm": 0.7268602377822969, "learning_rate": 2.9856464139560583e-06, "loss": 0.2323, "step": 1628 }, { "epoch": 1.264751552795031, "grad_norm": 0.7652996384356141, "learning_rate": 2.9800668490217878e-06, "loss": 0.2004, "step": 1629 }, { "epoch": 1.265527950310559, "grad_norm": 0.515078410416559, "learning_rate": 2.9744902883990923e-06, "loss": 0.2046, "step": 1630 }, { "epoch": 1.266304347826087, "grad_norm": 0.6011466037055163, "learning_rate": 2.9689167403821696e-06, "loss": 0.1707, "step": 1631 }, { "epoch": 1.2670807453416149, "grad_norm": 0.7295785281900856, "learning_rate": 2.963346213260737e-06, "loss": 0.2055, "step": 1632 }, { "epoch": 1.2678571428571428, "grad_norm": 0.7067827186315742, "learning_rate": 2.95777871532002e-06, "loss": 0.1782, "step": 1633 }, { "epoch": 1.2686335403726707, "grad_norm": 0.7119930158824129, "learning_rate": 2.9522142548407366e-06, "loss": 0.1903, "step": 1634 }, { "epoch": 1.2694099378881987, "grad_norm": 0.7489514651442392, "learning_rate": 2.9466528400990873e-06, "loss": 0.2525, "step": 1635 }, { "epoch": 1.2701863354037268, "grad_norm": 0.7351756013924604, "learning_rate": 2.941094479366744e-06, "loss": 0.2564, "step": 1636 }, { "epoch": 1.2709627329192545, "grad_norm": 0.897988815650448, "learning_rate": 2.9355391809108335e-06, "loss": 0.1688, "step": 1637 }, { "epoch": 1.2717391304347827, "grad_norm": 0.7380923017934373, "learning_rate": 2.929986952993933e-06, "loss": 0.2056, "step": 1638 }, { "epoch": 1.2725155279503104, "grad_norm": 0.7537368352044772, "learning_rate": 2.9244378038740477e-06, "loss": 0.1862, "step": 1639 }, { "epoch": 1.2732919254658386, "grad_norm": 0.7108431100365376, "learning_rate": 2.918891741804607e-06, "loss": 0.2032, "step": 1640 }, { "epoch": 1.2740683229813665, "grad_norm": 0.7231100945306405, "learning_rate": 2.9133487750344473e-06, "loss": 0.2346, "step": 1641 }, { "epoch": 1.2748447204968945, "grad_norm": 0.5046755643982631, "learning_rate": 2.9078089118077994e-06, "loss": 0.1943, "step": 1642 }, { "epoch": 1.2756211180124224, "grad_norm": 0.8065669775096975, "learning_rate": 2.9022721603642834e-06, "loss": 0.1717, "step": 1643 }, { "epoch": 1.2763975155279503, "grad_norm": 0.6175213735606777, "learning_rate": 2.896738528938885e-06, "loss": 0.2438, "step": 1644 }, { "epoch": 1.2771739130434783, "grad_norm": 0.650473670805497, "learning_rate": 2.8912080257619546e-06, "loss": 0.2167, "step": 1645 }, { "epoch": 1.2779503105590062, "grad_norm": 0.7453067615339689, "learning_rate": 2.885680659059187e-06, "loss": 0.1695, "step": 1646 }, { "epoch": 1.2787267080745341, "grad_norm": 0.6329889416680305, "learning_rate": 2.8801564370516095e-06, "loss": 0.1972, "step": 1647 }, { "epoch": 1.279503105590062, "grad_norm": 0.5574337724238981, "learning_rate": 2.8746353679555794e-06, "loss": 0.1393, "step": 1648 }, { "epoch": 1.28027950310559, "grad_norm": 0.7048873135471649, "learning_rate": 2.869117459982758e-06, "loss": 0.1737, "step": 1649 }, { "epoch": 1.281055900621118, "grad_norm": 0.6093619066858105, "learning_rate": 2.863602721340109e-06, "loss": 0.182, "step": 1650 }, { "epoch": 1.281832298136646, "grad_norm": 0.6757342473389127, "learning_rate": 2.85809116022988e-06, "loss": 0.228, "step": 1651 }, { "epoch": 1.2826086956521738, "grad_norm": 0.5898231884457856, "learning_rate": 2.8525827848495912e-06, "loss": 0.212, "step": 1652 }, { "epoch": 1.283385093167702, "grad_norm": 0.6766699076729125, "learning_rate": 2.8470776033920306e-06, "loss": 0.1149, "step": 1653 }, { "epoch": 1.2841614906832297, "grad_norm": 0.6990995859261374, "learning_rate": 2.8415756240452274e-06, "loss": 0.2406, "step": 1654 }, { "epoch": 1.2849378881987579, "grad_norm": 0.6262776315546574, "learning_rate": 2.8360768549924545e-06, "loss": 0.1591, "step": 1655 }, { "epoch": 1.2857142857142856, "grad_norm": 0.6292834243419796, "learning_rate": 2.83058130441221e-06, "loss": 0.1709, "step": 1656 }, { "epoch": 1.2864906832298137, "grad_norm": 0.6361804333505288, "learning_rate": 2.8250889804781993e-06, "loss": 0.2171, "step": 1657 }, { "epoch": 1.2872670807453417, "grad_norm": 0.6533246362216089, "learning_rate": 2.819599891359337e-06, "loss": 0.1677, "step": 1658 }, { "epoch": 1.2880434782608696, "grad_norm": 0.6897922412964063, "learning_rate": 2.814114045219717e-06, "loss": 0.1747, "step": 1659 }, { "epoch": 1.2888198757763976, "grad_norm": 0.5787093060540083, "learning_rate": 2.8086314502186197e-06, "loss": 0.2135, "step": 1660 }, { "epoch": 1.2895962732919255, "grad_norm": 0.828550516334709, "learning_rate": 2.8031521145104825e-06, "loss": 0.2209, "step": 1661 }, { "epoch": 1.2903726708074534, "grad_norm": 0.5323283325107762, "learning_rate": 2.7976760462448955e-06, "loss": 0.2985, "step": 1662 }, { "epoch": 1.2911490683229814, "grad_norm": 0.7428508105454147, "learning_rate": 2.7922032535665956e-06, "loss": 0.2023, "step": 1663 }, { "epoch": 1.2919254658385093, "grad_norm": 0.6719440458923299, "learning_rate": 2.7867337446154397e-06, "loss": 0.2211, "step": 1664 }, { "epoch": 1.2927018633540373, "grad_norm": 0.5984357828603092, "learning_rate": 2.781267527526409e-06, "loss": 0.1623, "step": 1665 }, { "epoch": 1.2934782608695652, "grad_norm": 0.6197801566247532, "learning_rate": 2.77580461042958e-06, "loss": 0.1921, "step": 1666 }, { "epoch": 1.2942546583850931, "grad_norm": 0.6297846733635665, "learning_rate": 2.7703450014501252e-06, "loss": 0.2537, "step": 1667 }, { "epoch": 1.295031055900621, "grad_norm": 0.6626905166343645, "learning_rate": 2.7648887087082996e-06, "loss": 0.1451, "step": 1668 }, { "epoch": 1.295807453416149, "grad_norm": 1.0398474044374706, "learning_rate": 2.7594357403194198e-06, "loss": 0.6714, "step": 1669 }, { "epoch": 1.296583850931677, "grad_norm": 0.6823908091683872, "learning_rate": 2.753986104393864e-06, "loss": 0.2295, "step": 1670 }, { "epoch": 1.2973602484472049, "grad_norm": 0.663873888514176, "learning_rate": 2.7485398090370487e-06, "loss": 0.2312, "step": 1671 }, { "epoch": 1.298136645962733, "grad_norm": 0.7635011272034118, "learning_rate": 2.743096862349427e-06, "loss": 0.1572, "step": 1672 }, { "epoch": 1.2989130434782608, "grad_norm": 0.5989191480221435, "learning_rate": 2.7376572724264642e-06, "loss": 0.1854, "step": 1673 }, { "epoch": 1.299689440993789, "grad_norm": 0.6145288454263249, "learning_rate": 2.7322210473586406e-06, "loss": 0.2383, "step": 1674 }, { "epoch": 1.3004658385093169, "grad_norm": 0.567457511925161, "learning_rate": 2.7267881952314305e-06, "loss": 0.183, "step": 1675 }, { "epoch": 1.3012422360248448, "grad_norm": 0.6238352495641353, "learning_rate": 2.721358724125287e-06, "loss": 0.2022, "step": 1676 }, { "epoch": 1.3020186335403727, "grad_norm": 0.6820979247675104, "learning_rate": 2.7159326421156362e-06, "loss": 0.224, "step": 1677 }, { "epoch": 1.3027950310559007, "grad_norm": 0.571552060553543, "learning_rate": 2.7105099572728676e-06, "loss": 0.2284, "step": 1678 }, { "epoch": 1.3035714285714286, "grad_norm": 0.5232359380598259, "learning_rate": 2.705090677662311e-06, "loss": 0.1654, "step": 1679 }, { "epoch": 1.3043478260869565, "grad_norm": 0.6466175953534916, "learning_rate": 2.6996748113442397e-06, "loss": 0.2362, "step": 1680 }, { "epoch": 1.3051242236024845, "grad_norm": 0.7799879998825444, "learning_rate": 2.694262366373844e-06, "loss": 0.2225, "step": 1681 }, { "epoch": 1.3059006211180124, "grad_norm": 0.8119867129648916, "learning_rate": 2.6888533508012247e-06, "loss": 0.1747, "step": 1682 }, { "epoch": 1.3066770186335404, "grad_norm": 0.6939049754796298, "learning_rate": 2.6834477726713897e-06, "loss": 0.2015, "step": 1683 }, { "epoch": 1.3074534161490683, "grad_norm": 0.49473488047788283, "learning_rate": 2.678045640024225e-06, "loss": 0.2102, "step": 1684 }, { "epoch": 1.3082298136645962, "grad_norm": 0.8096894108368577, "learning_rate": 2.6726469608945e-06, "loss": 0.1886, "step": 1685 }, { "epoch": 1.3090062111801242, "grad_norm": 0.6556803036414336, "learning_rate": 2.66725174331184e-06, "loss": 0.28, "step": 1686 }, { "epoch": 1.309782608695652, "grad_norm": 0.6275370478958799, "learning_rate": 2.6618599953007294e-06, "loss": 0.2126, "step": 1687 }, { "epoch": 1.31055900621118, "grad_norm": 0.6735256747179654, "learning_rate": 2.6564717248804876e-06, "loss": 0.1884, "step": 1688 }, { "epoch": 1.3113354037267082, "grad_norm": 0.6210327087275301, "learning_rate": 2.6510869400652596e-06, "loss": 0.2144, "step": 1689 }, { "epoch": 1.312111801242236, "grad_norm": 0.5658829382661712, "learning_rate": 2.6457056488640135e-06, "loss": 0.1607, "step": 1690 }, { "epoch": 1.312888198757764, "grad_norm": 0.6438512570632733, "learning_rate": 2.640327859280513e-06, "loss": 0.7259, "step": 1691 }, { "epoch": 1.3136645962732918, "grad_norm": 0.679508731246534, "learning_rate": 2.6349535793133196e-06, "loss": 0.246, "step": 1692 }, { "epoch": 1.31444099378882, "grad_norm": 0.8317066638456185, "learning_rate": 2.6295828169557743e-06, "loss": 0.3049, "step": 1693 }, { "epoch": 1.315217391304348, "grad_norm": 0.6690082802263326, "learning_rate": 2.6242155801959814e-06, "loss": 0.1639, "step": 1694 }, { "epoch": 1.3159937888198758, "grad_norm": 0.6941022741348313, "learning_rate": 2.618851877016808e-06, "loss": 0.2632, "step": 1695 }, { "epoch": 1.3167701863354038, "grad_norm": 0.7014836527777776, "learning_rate": 2.613491715395861e-06, "loss": 0.3175, "step": 1696 }, { "epoch": 1.3175465838509317, "grad_norm": 0.7199664938250353, "learning_rate": 2.6081351033054795e-06, "loss": 0.1901, "step": 1697 }, { "epoch": 1.3183229813664596, "grad_norm": 0.637861392593009, "learning_rate": 2.602782048712729e-06, "loss": 0.1734, "step": 1698 }, { "epoch": 1.3190993788819876, "grad_norm": 0.6819648450400675, "learning_rate": 2.597432559579376e-06, "loss": 0.2391, "step": 1699 }, { "epoch": 1.3198757763975155, "grad_norm": 0.8904024124974081, "learning_rate": 2.592086643861891e-06, "loss": 0.2855, "step": 1700 }, { "epoch": 1.3206521739130435, "grad_norm": 0.48745896848011455, "learning_rate": 2.5867443095114248e-06, "loss": 0.1897, "step": 1701 }, { "epoch": 1.3214285714285714, "grad_norm": 0.8460537091503546, "learning_rate": 2.5814055644738013e-06, "loss": 0.2509, "step": 1702 }, { "epoch": 1.3222049689440993, "grad_norm": 0.6683381603059002, "learning_rate": 2.5760704166895124e-06, "loss": 0.2903, "step": 1703 }, { "epoch": 1.3229813664596273, "grad_norm": 0.782094876471117, "learning_rate": 2.57073887409369e-06, "loss": 0.205, "step": 1704 }, { "epoch": 1.3237577639751552, "grad_norm": 0.7173554779741868, "learning_rate": 2.565410944616114e-06, "loss": 0.241, "step": 1705 }, { "epoch": 1.3245341614906831, "grad_norm": 0.636346680071142, "learning_rate": 2.5600866361811804e-06, "loss": 0.276, "step": 1706 }, { "epoch": 1.325310559006211, "grad_norm": 0.5606963529674622, "learning_rate": 2.5547659567079093e-06, "loss": 0.1481, "step": 1707 }, { "epoch": 1.3260869565217392, "grad_norm": 0.6557900759186601, "learning_rate": 2.5494489141099155e-06, "loss": 0.2145, "step": 1708 }, { "epoch": 1.326863354037267, "grad_norm": 0.6125589358387794, "learning_rate": 2.5441355162954053e-06, "loss": 0.2378, "step": 1709 }, { "epoch": 1.3276397515527951, "grad_norm": 0.6247549740451861, "learning_rate": 2.5388257711671723e-06, "loss": 0.2458, "step": 1710 }, { "epoch": 1.328416149068323, "grad_norm": 0.752161339583613, "learning_rate": 2.533519686622568e-06, "loss": 0.2114, "step": 1711 }, { "epoch": 1.329192546583851, "grad_norm": 0.7433630382005589, "learning_rate": 2.5282172705535013e-06, "loss": 0.2281, "step": 1712 }, { "epoch": 1.329968944099379, "grad_norm": 0.6813971417170122, "learning_rate": 2.5229185308464287e-06, "loss": 0.2035, "step": 1713 }, { "epoch": 1.3307453416149069, "grad_norm": 0.5103737813692204, "learning_rate": 2.517623475382335e-06, "loss": 0.2106, "step": 1714 }, { "epoch": 1.3315217391304348, "grad_norm": 0.5962820416502506, "learning_rate": 2.512332112036727e-06, "loss": 0.1782, "step": 1715 }, { "epoch": 1.3322981366459627, "grad_norm": 0.6914273258414828, "learning_rate": 2.507044448679621e-06, "loss": 0.2278, "step": 1716 }, { "epoch": 1.3330745341614907, "grad_norm": 0.7128151007058063, "learning_rate": 2.5017604931755248e-06, "loss": 0.1864, "step": 1717 }, { "epoch": 1.3338509316770186, "grad_norm": 0.558659133458875, "learning_rate": 2.4964802533834404e-06, "loss": 0.1904, "step": 1718 }, { "epoch": 1.3346273291925466, "grad_norm": 0.7603631590706666, "learning_rate": 2.4912037371568335e-06, "loss": 0.1555, "step": 1719 }, { "epoch": 1.3354037267080745, "grad_norm": 0.5445807272979503, "learning_rate": 2.4859309523436415e-06, "loss": 0.1778, "step": 1720 }, { "epoch": 1.3361801242236024, "grad_norm": 0.7033033541395278, "learning_rate": 2.4806619067862426e-06, "loss": 0.1779, "step": 1721 }, { "epoch": 1.3369565217391304, "grad_norm": 0.642320085112207, "learning_rate": 2.4753966083214613e-06, "loss": 0.204, "step": 1722 }, { "epoch": 1.3377329192546583, "grad_norm": 0.6372350294799459, "learning_rate": 2.470135064780544e-06, "loss": 0.2151, "step": 1723 }, { "epoch": 1.3385093167701863, "grad_norm": 0.7505876072154745, "learning_rate": 2.4648772839891517e-06, "loss": 0.1626, "step": 1724 }, { "epoch": 1.3392857142857144, "grad_norm": 0.9557088297278173, "learning_rate": 2.4596232737673544e-06, "loss": 0.1619, "step": 1725 }, { "epoch": 1.3400621118012421, "grad_norm": 0.5395302013858202, "learning_rate": 2.454373041929607e-06, "loss": 0.1786, "step": 1726 }, { "epoch": 1.3408385093167703, "grad_norm": 0.6502702375847003, "learning_rate": 2.4491265962847503e-06, "loss": 0.2514, "step": 1727 }, { "epoch": 1.341614906832298, "grad_norm": 0.653431667366223, "learning_rate": 2.4438839446359936e-06, "loss": 0.1475, "step": 1728 }, { "epoch": 1.3423913043478262, "grad_norm": 0.6255450867838394, "learning_rate": 2.4386450947808976e-06, "loss": 0.2353, "step": 1729 }, { "epoch": 1.343167701863354, "grad_norm": 0.672235293993061, "learning_rate": 2.433410054511376e-06, "loss": 0.2024, "step": 1730 }, { "epoch": 1.343944099378882, "grad_norm": 0.6113321731370445, "learning_rate": 2.4281788316136706e-06, "loss": 0.21, "step": 1731 }, { "epoch": 1.34472049689441, "grad_norm": 0.6282685415937647, "learning_rate": 2.422951433868346e-06, "loss": 0.1987, "step": 1732 }, { "epoch": 1.345496894409938, "grad_norm": 0.6111327487282882, "learning_rate": 2.4177278690502828e-06, "loss": 0.1828, "step": 1733 }, { "epoch": 1.3462732919254659, "grad_norm": 0.7048672873399557, "learning_rate": 2.4125081449286523e-06, "loss": 0.2033, "step": 1734 }, { "epoch": 1.3470496894409938, "grad_norm": 0.7425369287807069, "learning_rate": 2.407292269266924e-06, "loss": 0.2534, "step": 1735 }, { "epoch": 1.3478260869565217, "grad_norm": 0.551486152344872, "learning_rate": 2.4020802498228333e-06, "loss": 0.2016, "step": 1736 }, { "epoch": 1.3486024844720497, "grad_norm": 0.9498212467532069, "learning_rate": 2.3968720943483847e-06, "loss": 0.7871, "step": 1737 }, { "epoch": 1.3493788819875776, "grad_norm": 0.7453172688564581, "learning_rate": 2.3916678105898376e-06, "loss": 0.1431, "step": 1738 }, { "epoch": 1.3501552795031055, "grad_norm": 0.6373443951154786, "learning_rate": 2.3864674062876885e-06, "loss": 0.2077, "step": 1739 }, { "epoch": 1.3509316770186335, "grad_norm": 0.7694710727993578, "learning_rate": 2.3812708891766683e-06, "loss": 0.1939, "step": 1740 }, { "epoch": 1.3517080745341614, "grad_norm": 0.538584886646891, "learning_rate": 2.3760782669857208e-06, "loss": 0.2379, "step": 1741 }, { "epoch": 1.3524844720496896, "grad_norm": 0.7753882469053355, "learning_rate": 2.3708895474380045e-06, "loss": 0.1866, "step": 1742 }, { "epoch": 1.3532608695652173, "grad_norm": 0.509861914064838, "learning_rate": 2.3657047382508663e-06, "loss": 0.2091, "step": 1743 }, { "epoch": 1.3540372670807455, "grad_norm": 0.8224134040391431, "learning_rate": 2.360523847135838e-06, "loss": 0.1965, "step": 1744 }, { "epoch": 1.3548136645962732, "grad_norm": 0.8611601155813887, "learning_rate": 2.3553468817986282e-06, "loss": 0.2736, "step": 1745 }, { "epoch": 1.3555900621118013, "grad_norm": 0.7390022259992656, "learning_rate": 2.3501738499391053e-06, "loss": 0.1813, "step": 1746 }, { "epoch": 1.3563664596273293, "grad_norm": 0.7895079708533315, "learning_rate": 2.345004759251282e-06, "loss": 0.2003, "step": 1747 }, { "epoch": 1.3571428571428572, "grad_norm": 0.7540264986103109, "learning_rate": 2.339839617423318e-06, "loss": 0.1858, "step": 1748 }, { "epoch": 1.3579192546583851, "grad_norm": 0.7187120770854308, "learning_rate": 2.33467843213749e-06, "loss": 0.1644, "step": 1749 }, { "epoch": 1.358695652173913, "grad_norm": 0.571861843828106, "learning_rate": 2.3295212110701994e-06, "loss": 0.1662, "step": 1750 }, { "epoch": 1.359472049689441, "grad_norm": 0.7973528881829164, "learning_rate": 2.324367961891945e-06, "loss": 0.6632, "step": 1751 }, { "epoch": 1.360248447204969, "grad_norm": 0.6015874040620137, "learning_rate": 2.3192186922673187e-06, "loss": 0.2744, "step": 1752 }, { "epoch": 1.361024844720497, "grad_norm": 0.6953803763467398, "learning_rate": 2.3140734098549984e-06, "loss": 0.2209, "step": 1753 }, { "epoch": 1.3618012422360248, "grad_norm": 0.6455320127259806, "learning_rate": 2.308932122307724e-06, "loss": 0.2344, "step": 1754 }, { "epoch": 1.3625776397515528, "grad_norm": 0.503312686738009, "learning_rate": 2.3037948372723017e-06, "loss": 0.1848, "step": 1755 }, { "epoch": 1.3633540372670807, "grad_norm": 0.6571236210637411, "learning_rate": 2.2986615623895774e-06, "loss": 0.2171, "step": 1756 }, { "epoch": 1.3641304347826086, "grad_norm": 0.6833336596794264, "learning_rate": 2.2935323052944404e-06, "loss": 0.1982, "step": 1757 }, { "epoch": 1.3649068322981366, "grad_norm": 0.6576339383426487, "learning_rate": 2.288407073615797e-06, "loss": 0.2888, "step": 1758 }, { "epoch": 1.3656832298136645, "grad_norm": 0.6285514572368271, "learning_rate": 2.2832858749765676e-06, "loss": 0.17, "step": 1759 }, { "epoch": 1.3664596273291925, "grad_norm": 0.6559569703157409, "learning_rate": 2.2781687169936794e-06, "loss": 0.1569, "step": 1760 }, { "epoch": 1.3672360248447206, "grad_norm": 0.6692412146871892, "learning_rate": 2.2730556072780423e-06, "loss": 0.1773, "step": 1761 }, { "epoch": 1.3680124223602483, "grad_norm": 0.6173777310908342, "learning_rate": 2.2679465534345534e-06, "loss": 0.2063, "step": 1762 }, { "epoch": 1.3687888198757765, "grad_norm": 0.6604681380703692, "learning_rate": 2.2628415630620683e-06, "loss": 0.2567, "step": 1763 }, { "epoch": 1.3695652173913042, "grad_norm": 0.7938356530371221, "learning_rate": 2.2577406437534055e-06, "loss": 0.1783, "step": 1764 }, { "epoch": 1.3703416149068324, "grad_norm": 0.7181379092120469, "learning_rate": 2.252643803095329e-06, "loss": 0.79, "step": 1765 }, { "epoch": 1.3711180124223603, "grad_norm": 0.5918090664848134, "learning_rate": 2.247551048668531e-06, "loss": 0.1626, "step": 1766 }, { "epoch": 1.3718944099378882, "grad_norm": 0.6867867367088795, "learning_rate": 2.2424623880476275e-06, "loss": 0.1785, "step": 1767 }, { "epoch": 1.3726708074534162, "grad_norm": 0.6429010490424727, "learning_rate": 2.2373778288011517e-06, "loss": 0.1864, "step": 1768 }, { "epoch": 1.3734472049689441, "grad_norm": 0.7255081883677849, "learning_rate": 2.2322973784915275e-06, "loss": 0.2406, "step": 1769 }, { "epoch": 1.374223602484472, "grad_norm": 0.6349555805203102, "learning_rate": 2.2272210446750757e-06, "loss": 0.1758, "step": 1770 }, { "epoch": 1.375, "grad_norm": 0.6659699204496362, "learning_rate": 2.2221488349019903e-06, "loss": 0.1886, "step": 1771 }, { "epoch": 1.375776397515528, "grad_norm": 0.8248524993899632, "learning_rate": 2.2170807567163294e-06, "loss": 1.1341, "step": 1772 }, { "epoch": 1.3765527950310559, "grad_norm": 0.7259564758074841, "learning_rate": 2.2120168176560127e-06, "loss": 0.2169, "step": 1773 }, { "epoch": 1.3773291925465838, "grad_norm": 0.5801717448832732, "learning_rate": 2.2069570252527954e-06, "loss": 0.1955, "step": 1774 }, { "epoch": 1.3781055900621118, "grad_norm": 0.7868853527721734, "learning_rate": 2.201901387032274e-06, "loss": 0.2142, "step": 1775 }, { "epoch": 1.3788819875776397, "grad_norm": 0.7231136808263647, "learning_rate": 2.196849910513858e-06, "loss": 0.2129, "step": 1776 }, { "epoch": 1.3796583850931676, "grad_norm": 0.81071959289729, "learning_rate": 2.1918026032107746e-06, "loss": 0.2072, "step": 1777 }, { "epoch": 1.3804347826086958, "grad_norm": 0.8508111916377957, "learning_rate": 2.186759472630045e-06, "loss": 0.2134, "step": 1778 }, { "epoch": 1.3812111801242235, "grad_norm": 0.6327199061385926, "learning_rate": 2.1817205262724765e-06, "loss": 0.1631, "step": 1779 }, { "epoch": 1.3819875776397517, "grad_norm": 0.6851532352261612, "learning_rate": 2.17668577163266e-06, "loss": 0.1632, "step": 1780 }, { "epoch": 1.3827639751552794, "grad_norm": 0.7477936776803643, "learning_rate": 2.1716552161989478e-06, "loss": 0.1848, "step": 1781 }, { "epoch": 1.3835403726708075, "grad_norm": 0.6689055989674673, "learning_rate": 2.1666288674534446e-06, "loss": 0.2362, "step": 1782 }, { "epoch": 1.3843167701863355, "grad_norm": 0.5422326028913114, "learning_rate": 2.1616067328720035e-06, "loss": 0.1689, "step": 1783 }, { "epoch": 1.3850931677018634, "grad_norm": 0.7319980871603197, "learning_rate": 2.156588819924204e-06, "loss": 0.219, "step": 1784 }, { "epoch": 1.3858695652173914, "grad_norm": 0.5352676917052108, "learning_rate": 2.151575136073353e-06, "loss": 0.233, "step": 1785 }, { "epoch": 1.3866459627329193, "grad_norm": 0.6223555556220352, "learning_rate": 2.1465656887764615e-06, "loss": 0.2073, "step": 1786 }, { "epoch": 1.3874223602484472, "grad_norm": 0.7419288756490057, "learning_rate": 2.14156048548424e-06, "loss": 0.2657, "step": 1787 }, { "epoch": 1.3881987577639752, "grad_norm": 0.5427061636797029, "learning_rate": 2.136559533641092e-06, "loss": 0.1586, "step": 1788 }, { "epoch": 1.388975155279503, "grad_norm": 0.5641880176001941, "learning_rate": 2.1315628406850908e-06, "loss": 0.1729, "step": 1789 }, { "epoch": 1.389751552795031, "grad_norm": 0.8318192085995413, "learning_rate": 2.126570414047982e-06, "loss": 0.1806, "step": 1790 }, { "epoch": 1.390527950310559, "grad_norm": 0.6184700356795247, "learning_rate": 2.1215822611551583e-06, "loss": 0.2296, "step": 1791 }, { "epoch": 1.391304347826087, "grad_norm": 0.603639456615519, "learning_rate": 2.1165983894256647e-06, "loss": 0.2944, "step": 1792 }, { "epoch": 1.3920807453416149, "grad_norm": 0.6921626959249401, "learning_rate": 2.111618806272172e-06, "loss": 0.1843, "step": 1793 }, { "epoch": 1.3928571428571428, "grad_norm": 0.7131879785219588, "learning_rate": 2.1066435191009717e-06, "loss": 0.1973, "step": 1794 }, { "epoch": 1.3936335403726707, "grad_norm": 0.6061238520608145, "learning_rate": 2.1016725353119734e-06, "loss": 0.1901, "step": 1795 }, { "epoch": 1.3944099378881987, "grad_norm": 0.5514288544427138, "learning_rate": 2.096705862298676e-06, "loss": 0.2271, "step": 1796 }, { "epoch": 1.3951863354037268, "grad_norm": 0.5516163241505421, "learning_rate": 2.091743507448177e-06, "loss": 0.1533, "step": 1797 }, { "epoch": 1.3959627329192545, "grad_norm": 0.63588613874994, "learning_rate": 2.086785478141143e-06, "loss": 0.2545, "step": 1798 }, { "epoch": 1.3967391304347827, "grad_norm": 0.5743804565741555, "learning_rate": 2.0818317817518117e-06, "loss": 0.2307, "step": 1799 }, { "epoch": 1.3975155279503104, "grad_norm": 0.6950547604983858, "learning_rate": 2.076882425647977e-06, "loss": 0.2025, "step": 1800 }, { "epoch": 1.3982919254658386, "grad_norm": 0.9131555222262, "learning_rate": 2.0719374171909734e-06, "loss": 0.7944, "step": 1801 }, { "epoch": 1.3990683229813665, "grad_norm": 0.8301228682888449, "learning_rate": 2.0669967637356702e-06, "loss": 0.1973, "step": 1802 }, { "epoch": 1.3998447204968945, "grad_norm": 0.6235655847803269, "learning_rate": 2.0620604726304635e-06, "loss": 0.21, "step": 1803 }, { "epoch": 1.4006211180124224, "grad_norm": 0.5617501609824901, "learning_rate": 2.057128551217254e-06, "loss": 0.2035, "step": 1804 }, { "epoch": 1.4013975155279503, "grad_norm": 0.6501164251533084, "learning_rate": 2.0522010068314506e-06, "loss": 0.1852, "step": 1805 }, { "epoch": 1.4021739130434783, "grad_norm": 0.6794378931338078, "learning_rate": 2.0472778468019456e-06, "loss": 0.1581, "step": 1806 }, { "epoch": 1.4029503105590062, "grad_norm": 0.6346759155856735, "learning_rate": 2.0423590784511167e-06, "loss": 0.1854, "step": 1807 }, { "epoch": 1.4037267080745341, "grad_norm": 0.5386939136925871, "learning_rate": 2.037444709094804e-06, "loss": 0.2594, "step": 1808 }, { "epoch": 1.404503105590062, "grad_norm": 0.6262587830010402, "learning_rate": 2.032534746042307e-06, "loss": 0.2219, "step": 1809 }, { "epoch": 1.40527950310559, "grad_norm": 0.7070829179299605, "learning_rate": 2.027629196596373e-06, "loss": 0.1788, "step": 1810 }, { "epoch": 1.406055900621118, "grad_norm": 0.5872629170349625, "learning_rate": 2.0227280680531817e-06, "loss": 0.2305, "step": 1811 }, { "epoch": 1.406832298136646, "grad_norm": 0.6279807234455336, "learning_rate": 2.0178313677023427e-06, "loss": 0.1836, "step": 1812 }, { "epoch": 1.4076086956521738, "grad_norm": 0.5664937756439307, "learning_rate": 2.0129391028268736e-06, "loss": 0.2005, "step": 1813 }, { "epoch": 1.408385093167702, "grad_norm": 0.767830794550858, "learning_rate": 2.008051280703196e-06, "loss": 0.202, "step": 1814 }, { "epoch": 1.4091614906832297, "grad_norm": 0.6998868863613296, "learning_rate": 2.003167908601128e-06, "loss": 0.1669, "step": 1815 }, { "epoch": 1.4099378881987579, "grad_norm": 0.6722902670978219, "learning_rate": 1.9982889937838635e-06, "loss": 0.2155, "step": 1816 }, { "epoch": 1.4107142857142856, "grad_norm": 0.5313532612341066, "learning_rate": 1.9934145435079705e-06, "loss": 0.1453, "step": 1817 }, { "epoch": 1.4114906832298137, "grad_norm": 0.5594204659576875, "learning_rate": 1.988544565023378e-06, "loss": 0.2438, "step": 1818 }, { "epoch": 1.4122670807453417, "grad_norm": 0.6298259995789122, "learning_rate": 1.983679065573358e-06, "loss": 0.1685, "step": 1819 }, { "epoch": 1.4130434782608696, "grad_norm": 0.6974565225341264, "learning_rate": 1.978818052394528e-06, "loss": 0.214, "step": 1820 }, { "epoch": 1.4138198757763976, "grad_norm": 0.6965364975136294, "learning_rate": 1.973961532716827e-06, "loss": 0.2675, "step": 1821 }, { "epoch": 1.4145962732919255, "grad_norm": 0.6011089493095797, "learning_rate": 1.9691095137635117e-06, "loss": 0.1872, "step": 1822 }, { "epoch": 1.4153726708074534, "grad_norm": 0.6105927158487074, "learning_rate": 1.96426200275115e-06, "loss": 0.2052, "step": 1823 }, { "epoch": 1.4161490683229814, "grad_norm": 0.6991825545928124, "learning_rate": 1.9594190068895967e-06, "loss": 0.2674, "step": 1824 }, { "epoch": 1.4169254658385093, "grad_norm": 0.6051474562202951, "learning_rate": 1.9545805333819995e-06, "loss": 0.1838, "step": 1825 }, { "epoch": 1.4177018633540373, "grad_norm": 0.626244006107462, "learning_rate": 1.949746589424772e-06, "loss": 0.2149, "step": 1826 }, { "epoch": 1.4184782608695652, "grad_norm": 0.7100884579550758, "learning_rate": 1.9449171822075973e-06, "loss": 0.2575, "step": 1827 }, { "epoch": 1.4192546583850931, "grad_norm": 0.817940896504196, "learning_rate": 1.940092318913408e-06, "loss": 0.2025, "step": 1828 }, { "epoch": 1.420031055900621, "grad_norm": 0.6327927926712754, "learning_rate": 1.9352720067183756e-06, "loss": 0.1679, "step": 1829 }, { "epoch": 1.420807453416149, "grad_norm": 0.7791646072224102, "learning_rate": 1.9304562527919094e-06, "loss": 0.1835, "step": 1830 }, { "epoch": 1.421583850931677, "grad_norm": 0.7070615794886972, "learning_rate": 1.9256450642966324e-06, "loss": 0.2488, "step": 1831 }, { "epoch": 1.4223602484472049, "grad_norm": 0.7487136867305914, "learning_rate": 1.920838448388382e-06, "loss": 0.6974, "step": 1832 }, { "epoch": 1.423136645962733, "grad_norm": 0.7862649013814144, "learning_rate": 1.9160364122161924e-06, "loss": 0.2098, "step": 1833 }, { "epoch": 1.4239130434782608, "grad_norm": 0.6850315098828365, "learning_rate": 1.911238962922282e-06, "loss": 0.1856, "step": 1834 }, { "epoch": 1.424689440993789, "grad_norm": 0.7878323929076622, "learning_rate": 1.9064461076420594e-06, "loss": 0.2097, "step": 1835 }, { "epoch": 1.4254658385093169, "grad_norm": 0.8197085093587646, "learning_rate": 1.901657853504088e-06, "loss": 0.2147, "step": 1836 }, { "epoch": 1.4262422360248448, "grad_norm": 0.7222217406312015, "learning_rate": 1.8968742076300912e-06, "loss": 0.227, "step": 1837 }, { "epoch": 1.4270186335403727, "grad_norm": 0.821148119896323, "learning_rate": 1.8920951771349428e-06, "loss": 0.2035, "step": 1838 }, { "epoch": 1.4277950310559007, "grad_norm": 0.5775582898931353, "learning_rate": 1.8873207691266454e-06, "loss": 0.173, "step": 1839 }, { "epoch": 1.4285714285714286, "grad_norm": 0.6088761238602526, "learning_rate": 1.8825509907063328e-06, "loss": 0.1837, "step": 1840 }, { "epoch": 1.4293478260869565, "grad_norm": 0.6690261195179364, "learning_rate": 1.8777858489682465e-06, "loss": 0.1826, "step": 1841 }, { "epoch": 1.4301242236024845, "grad_norm": 0.5585302603994581, "learning_rate": 1.8730253509997388e-06, "loss": 0.2238, "step": 1842 }, { "epoch": 1.4309006211180124, "grad_norm": 0.6366129201829734, "learning_rate": 1.8682695038812493e-06, "loss": 0.2647, "step": 1843 }, { "epoch": 1.4316770186335404, "grad_norm": 0.6430699408941061, "learning_rate": 1.8635183146863016e-06, "loss": 0.1482, "step": 1844 }, { "epoch": 1.4324534161490683, "grad_norm": 0.5919266035034595, "learning_rate": 1.858771790481495e-06, "loss": 0.1991, "step": 1845 }, { "epoch": 1.4332298136645962, "grad_norm": 0.5859605965958484, "learning_rate": 1.8540299383264843e-06, "loss": 0.2208, "step": 1846 }, { "epoch": 1.4340062111801242, "grad_norm": 0.8385644590202775, "learning_rate": 1.8492927652739823e-06, "loss": 0.235, "step": 1847 }, { "epoch": 1.434782608695652, "grad_norm": 0.6607434933814742, "learning_rate": 1.8445602783697375e-06, "loss": 0.18, "step": 1848 }, { "epoch": 1.43555900621118, "grad_norm": 0.7028198087690847, "learning_rate": 1.8398324846525284e-06, "loss": 0.1961, "step": 1849 }, { "epoch": 1.4363354037267082, "grad_norm": 0.5447249801816522, "learning_rate": 1.8351093911541573e-06, "loss": 0.1948, "step": 1850 }, { "epoch": 1.437111801242236, "grad_norm": 0.6205247766600416, "learning_rate": 1.83039100489943e-06, "loss": 0.1991, "step": 1851 }, { "epoch": 1.437888198757764, "grad_norm": 0.5711800053241141, "learning_rate": 1.8256773329061566e-06, "loss": 0.199, "step": 1852 }, { "epoch": 1.4386645962732918, "grad_norm": 0.626683730585946, "learning_rate": 1.8209683821851342e-06, "loss": 0.248, "step": 1853 }, { "epoch": 1.43944099378882, "grad_norm": 0.6333763210436228, "learning_rate": 1.8162641597401338e-06, "loss": 0.2078, "step": 1854 }, { "epoch": 1.440217391304348, "grad_norm": 0.8693988905792619, "learning_rate": 1.8115646725678997e-06, "loss": 0.1933, "step": 1855 }, { "epoch": 1.4409937888198758, "grad_norm": 0.7075394909024574, "learning_rate": 1.8068699276581286e-06, "loss": 0.1414, "step": 1856 }, { "epoch": 1.4417701863354038, "grad_norm": 0.6030566658024629, "learning_rate": 1.8021799319934646e-06, "loss": 0.184, "step": 1857 }, { "epoch": 1.4425465838509317, "grad_norm": 0.74309802146998, "learning_rate": 1.7974946925494925e-06, "loss": 0.1745, "step": 1858 }, { "epoch": 1.4433229813664596, "grad_norm": 0.6034558561337676, "learning_rate": 1.792814216294716e-06, "loss": 0.1918, "step": 1859 }, { "epoch": 1.4440993788819876, "grad_norm": 0.567782159658038, "learning_rate": 1.7881385101905613e-06, "loss": 0.1575, "step": 1860 }, { "epoch": 1.4448757763975155, "grad_norm": 0.6642356347094795, "learning_rate": 1.7834675811913538e-06, "loss": 0.2574, "step": 1861 }, { "epoch": 1.4456521739130435, "grad_norm": 0.661648209363816, "learning_rate": 1.778801436244319e-06, "loss": 0.2, "step": 1862 }, { "epoch": 1.4464285714285714, "grad_norm": 0.6567226064531684, "learning_rate": 1.7741400822895633e-06, "loss": 0.2129, "step": 1863 }, { "epoch": 1.4472049689440993, "grad_norm": 0.6711653774080248, "learning_rate": 1.7694835262600668e-06, "loss": 0.2052, "step": 1864 }, { "epoch": 1.4479813664596273, "grad_norm": 0.679617906863488, "learning_rate": 1.7648317750816779e-06, "loss": 0.1509, "step": 1865 }, { "epoch": 1.4487577639751552, "grad_norm": 0.6715228357806005, "learning_rate": 1.7601848356730933e-06, "loss": 0.1883, "step": 1866 }, { "epoch": 1.4495341614906831, "grad_norm": 0.7726102636282149, "learning_rate": 1.755542714945858e-06, "loss": 0.1872, "step": 1867 }, { "epoch": 1.450310559006211, "grad_norm": 0.5507356644800848, "learning_rate": 1.7509054198043473e-06, "loss": 0.1003, "step": 1868 }, { "epoch": 1.4510869565217392, "grad_norm": 0.6146435919236274, "learning_rate": 1.746272957145756e-06, "loss": 0.1999, "step": 1869 }, { "epoch": 1.451863354037267, "grad_norm": 0.7120022766439346, "learning_rate": 1.741645333860098e-06, "loss": 0.224, "step": 1870 }, { "epoch": 1.4526397515527951, "grad_norm": 0.6059001307168056, "learning_rate": 1.7370225568301869e-06, "loss": 0.1865, "step": 1871 }, { "epoch": 1.453416149068323, "grad_norm": 0.7187024108717517, "learning_rate": 1.7324046329316253e-06, "loss": 0.1629, "step": 1872 }, { "epoch": 1.454192546583851, "grad_norm": 0.4696169157975532, "learning_rate": 1.7277915690328022e-06, "loss": 0.1573, "step": 1873 }, { "epoch": 1.454968944099379, "grad_norm": 0.6743725831488606, "learning_rate": 1.7231833719948743e-06, "loss": 0.1367, "step": 1874 }, { "epoch": 1.4557453416149069, "grad_norm": 0.6060166347146454, "learning_rate": 1.718580048671763e-06, "loss": 0.1913, "step": 1875 }, { "epoch": 1.4565217391304348, "grad_norm": 0.6725437912456225, "learning_rate": 1.7139816059101372e-06, "loss": 0.2568, "step": 1876 }, { "epoch": 1.4572981366459627, "grad_norm": 0.6175644351838124, "learning_rate": 1.7093880505494116e-06, "loss": 0.2023, "step": 1877 }, { "epoch": 1.4580745341614907, "grad_norm": 0.5271806907288218, "learning_rate": 1.7047993894217269e-06, "loss": 0.2394, "step": 1878 }, { "epoch": 1.4588509316770186, "grad_norm": 0.5928494250887386, "learning_rate": 1.700215629351945e-06, "loss": 0.1855, "step": 1879 }, { "epoch": 1.4596273291925466, "grad_norm": 0.6911279643252211, "learning_rate": 1.6956367771576432e-06, "loss": 0.1657, "step": 1880 }, { "epoch": 1.4604037267080745, "grad_norm": 0.5155784121430018, "learning_rate": 1.6910628396490925e-06, "loss": 0.2102, "step": 1881 }, { "epoch": 1.4611801242236024, "grad_norm": 0.5324277827463, "learning_rate": 1.6864938236292606e-06, "loss": 0.1965, "step": 1882 }, { "epoch": 1.4619565217391304, "grad_norm": 0.6185215248883015, "learning_rate": 1.6819297358937903e-06, "loss": 0.1792, "step": 1883 }, { "epoch": 1.4627329192546583, "grad_norm": 0.5569943109286264, "learning_rate": 1.6773705832309945e-06, "loss": 0.1891, "step": 1884 }, { "epoch": 1.4635093167701863, "grad_norm": 0.7970766734619165, "learning_rate": 1.672816372421851e-06, "loss": 0.1914, "step": 1885 }, { "epoch": 1.4642857142857144, "grad_norm": 0.5601887358962904, "learning_rate": 1.6682671102399806e-06, "loss": 0.1891, "step": 1886 }, { "epoch": 1.4650621118012421, "grad_norm": 0.6530418703177325, "learning_rate": 1.6637228034516512e-06, "loss": 0.1429, "step": 1887 }, { "epoch": 1.4658385093167703, "grad_norm": 0.675614976918855, "learning_rate": 1.6591834588157523e-06, "loss": 0.1814, "step": 1888 }, { "epoch": 1.466614906832298, "grad_norm": 0.6209555221712149, "learning_rate": 1.6546490830837996e-06, "loss": 0.1705, "step": 1889 }, { "epoch": 1.4673913043478262, "grad_norm": 0.6538820843544405, "learning_rate": 1.6501196829999179e-06, "loss": 0.2311, "step": 1890 }, { "epoch": 1.468167701863354, "grad_norm": 0.6421705170268723, "learning_rate": 1.6455952653008255e-06, "loss": 0.2183, "step": 1891 }, { "epoch": 1.468944099378882, "grad_norm": 0.7155714028941796, "learning_rate": 1.6410758367158386e-06, "loss": 0.2228, "step": 1892 }, { "epoch": 1.46972049689441, "grad_norm": 0.714495043604776, "learning_rate": 1.6365614039668464e-06, "loss": 0.2444, "step": 1893 }, { "epoch": 1.470496894409938, "grad_norm": 0.6123317125962588, "learning_rate": 1.6320519737683095e-06, "loss": 0.2061, "step": 1894 }, { "epoch": 1.4712732919254659, "grad_norm": 0.8231675991258433, "learning_rate": 1.62754755282725e-06, "loss": 0.2179, "step": 1895 }, { "epoch": 1.4720496894409938, "grad_norm": 0.5891736586120095, "learning_rate": 1.6230481478432364e-06, "loss": 0.1946, "step": 1896 }, { "epoch": 1.4728260869565217, "grad_norm": 0.5630575826397033, "learning_rate": 1.6185537655083817e-06, "loss": 0.276, "step": 1897 }, { "epoch": 1.4736024844720497, "grad_norm": 0.6844425684996079, "learning_rate": 1.6140644125073223e-06, "loss": 0.1868, "step": 1898 }, { "epoch": 1.4743788819875776, "grad_norm": 0.6620508779592771, "learning_rate": 1.6095800955172175e-06, "loss": 0.2345, "step": 1899 }, { "epoch": 1.4751552795031055, "grad_norm": 0.6484097317611907, "learning_rate": 1.6051008212077384e-06, "loss": 0.2607, "step": 1900 }, { "epoch": 1.4759316770186335, "grad_norm": 0.720269962303897, "learning_rate": 1.600626596241051e-06, "loss": 0.2122, "step": 1901 }, { "epoch": 1.4767080745341614, "grad_norm": 0.7154189743904121, "learning_rate": 1.5961574272718179e-06, "loss": 0.2395, "step": 1902 }, { "epoch": 1.4774844720496896, "grad_norm": 0.8043840453966414, "learning_rate": 1.5916933209471752e-06, "loss": 0.2631, "step": 1903 }, { "epoch": 1.4782608695652173, "grad_norm": 0.8787144013996321, "learning_rate": 1.5872342839067305e-06, "loss": 0.7161, "step": 1904 }, { "epoch": 1.4790372670807455, "grad_norm": 0.6854496332063493, "learning_rate": 1.5827803227825567e-06, "loss": 0.2282, "step": 1905 }, { "epoch": 1.4798136645962732, "grad_norm": 0.5596232072155545, "learning_rate": 1.5783314441991698e-06, "loss": 0.2277, "step": 1906 }, { "epoch": 1.4805900621118013, "grad_norm": 0.5597967347628113, "learning_rate": 1.5738876547735327e-06, "loss": 0.2189, "step": 1907 }, { "epoch": 1.4813664596273293, "grad_norm": 0.6641006838191906, "learning_rate": 1.569448961115037e-06, "loss": 0.146, "step": 1908 }, { "epoch": 1.4821428571428572, "grad_norm": 0.7102992736170242, "learning_rate": 1.5650153698254916e-06, "loss": 0.1942, "step": 1909 }, { "epoch": 1.4829192546583851, "grad_norm": 0.6107173403673253, "learning_rate": 1.5605868874991225e-06, "loss": 0.1831, "step": 1910 }, { "epoch": 1.483695652173913, "grad_norm": 0.6708042729029113, "learning_rate": 1.5561635207225518e-06, "loss": 0.1808, "step": 1911 }, { "epoch": 1.484472049689441, "grad_norm": 0.4494650532431989, "learning_rate": 1.5517452760747975e-06, "loss": 0.1679, "step": 1912 }, { "epoch": 1.485248447204969, "grad_norm": 0.6480985119540503, "learning_rate": 1.5473321601272557e-06, "loss": 0.1787, "step": 1913 }, { "epoch": 1.486024844720497, "grad_norm": 0.7539793473153977, "learning_rate": 1.5429241794436939e-06, "loss": 0.1961, "step": 1914 }, { "epoch": 1.4868012422360248, "grad_norm": 0.7203087334438455, "learning_rate": 1.5385213405802473e-06, "loss": 0.1813, "step": 1915 }, { "epoch": 1.4875776397515528, "grad_norm": 0.7050598979157746, "learning_rate": 1.5341236500853963e-06, "loss": 0.2256, "step": 1916 }, { "epoch": 1.4883540372670807, "grad_norm": 0.6309849755487895, "learning_rate": 1.5297311144999705e-06, "loss": 0.1871, "step": 1917 }, { "epoch": 1.4891304347826086, "grad_norm": 0.76967903294531, "learning_rate": 1.525343740357128e-06, "loss": 0.5842, "step": 1918 }, { "epoch": 1.4899068322981366, "grad_norm": 0.6502698434129698, "learning_rate": 1.5209615341823509e-06, "loss": 0.2098, "step": 1919 }, { "epoch": 1.4906832298136645, "grad_norm": 0.7328639999269213, "learning_rate": 1.5165845024934366e-06, "loss": 0.2118, "step": 1920 }, { "epoch": 1.4914596273291925, "grad_norm": 0.7421662055881708, "learning_rate": 1.5122126518004842e-06, "loss": 0.1903, "step": 1921 }, { "epoch": 1.4922360248447206, "grad_norm": 0.6663678001175555, "learning_rate": 1.5078459886058894e-06, "loss": 0.6924, "step": 1922 }, { "epoch": 1.4930124223602483, "grad_norm": 0.6343518606784154, "learning_rate": 1.503484519404329e-06, "loss": 0.1829, "step": 1923 }, { "epoch": 1.4937888198757765, "grad_norm": 0.6611487561020187, "learning_rate": 1.499128250682757e-06, "loss": 0.2668, "step": 1924 }, { "epoch": 1.4945652173913042, "grad_norm": 0.5459760097974846, "learning_rate": 1.4947771889203937e-06, "loss": 0.2272, "step": 1925 }, { "epoch": 1.4953416149068324, "grad_norm": 0.6637152397722487, "learning_rate": 1.4904313405887111e-06, "loss": 0.1759, "step": 1926 }, { "epoch": 1.4961180124223603, "grad_norm": 0.7182340559337645, "learning_rate": 1.4860907121514317e-06, "loss": 0.1786, "step": 1927 }, { "epoch": 1.4968944099378882, "grad_norm": 0.6871590878317885, "learning_rate": 1.4817553100645105e-06, "loss": 0.256, "step": 1928 }, { "epoch": 1.4976708074534162, "grad_norm": 0.685388807280918, "learning_rate": 1.4774251407761286e-06, "loss": 0.2012, "step": 1929 }, { "epoch": 1.4984472049689441, "grad_norm": 0.5535172376661915, "learning_rate": 1.4731002107266902e-06, "loss": 0.2005, "step": 1930 }, { "epoch": 1.499223602484472, "grad_norm": 0.5758793700442794, "learning_rate": 1.4687805263487985e-06, "loss": 0.2221, "step": 1931 }, { "epoch": 1.5, "grad_norm": 0.8781900407015714, "learning_rate": 1.4644660940672628e-06, "loss": 0.2236, "step": 1932 }, { "epoch": 1.500776397515528, "grad_norm": 0.6634036720509223, "learning_rate": 1.4601569202990757e-06, "loss": 0.1603, "step": 1933 }, { "epoch": 1.5015527950310559, "grad_norm": 0.7026439682707553, "learning_rate": 1.455853011453408e-06, "loss": 0.1905, "step": 1934 }, { "epoch": 1.5023291925465838, "grad_norm": 0.7315055920095311, "learning_rate": 1.4515543739316058e-06, "loss": 0.2007, "step": 1935 }, { "epoch": 1.5031055900621118, "grad_norm": 0.5974042979315395, "learning_rate": 1.447261014127167e-06, "loss": 0.2147, "step": 1936 }, { "epoch": 1.50388198757764, "grad_norm": 0.5302798838741785, "learning_rate": 1.4429729384257474e-06, "loss": 0.1509, "step": 1937 }, { "epoch": 1.5046583850931676, "grad_norm": 0.4930635925375197, "learning_rate": 1.4386901532051379e-06, "loss": 0.1972, "step": 1938 }, { "epoch": 1.5054347826086958, "grad_norm": 0.6359700168769267, "learning_rate": 1.434412664835262e-06, "loss": 0.2324, "step": 1939 }, { "epoch": 1.5062111801242235, "grad_norm": 0.7389846608062239, "learning_rate": 1.430140479678168e-06, "loss": 0.217, "step": 1940 }, { "epoch": 1.5069875776397517, "grad_norm": 0.6040819251512451, "learning_rate": 1.4258736040880122e-06, "loss": 0.1524, "step": 1941 }, { "epoch": 1.5077639751552794, "grad_norm": 0.6623583233070779, "learning_rate": 1.4216120444110566e-06, "loss": 0.2269, "step": 1942 }, { "epoch": 1.5085403726708075, "grad_norm": 0.5521253539274268, "learning_rate": 1.4173558069856585e-06, "loss": 0.1613, "step": 1943 }, { "epoch": 1.5093167701863353, "grad_norm": 0.5704842739670795, "learning_rate": 1.4131048981422529e-06, "loss": 0.2425, "step": 1944 }, { "epoch": 1.5100931677018634, "grad_norm": 0.42994552817371695, "learning_rate": 1.4088593242033572e-06, "loss": 0.2332, "step": 1945 }, { "epoch": 1.5108695652173914, "grad_norm": 0.6454043537842014, "learning_rate": 1.404619091483546e-06, "loss": 0.1901, "step": 1946 }, { "epoch": 1.5116459627329193, "grad_norm": 0.7996487735364943, "learning_rate": 1.400384206289458e-06, "loss": 0.2368, "step": 1947 }, { "epoch": 1.5124223602484472, "grad_norm": 0.5931376193756718, "learning_rate": 1.3961546749197719e-06, "loss": 0.2515, "step": 1948 }, { "epoch": 1.5131987577639752, "grad_norm": 0.5682434186747588, "learning_rate": 1.3919305036652048e-06, "loss": 0.1892, "step": 1949 }, { "epoch": 1.513975155279503, "grad_norm": 0.6730397377257166, "learning_rate": 1.387711698808505e-06, "loss": 0.2212, "step": 1950 }, { "epoch": 1.514751552795031, "grad_norm": 0.5962471928850358, "learning_rate": 1.3834982666244346e-06, "loss": 0.1762, "step": 1951 }, { "epoch": 1.515527950310559, "grad_norm": 0.6209659660314643, "learning_rate": 1.3792902133797692e-06, "loss": 0.2827, "step": 1952 }, { "epoch": 1.516304347826087, "grad_norm": 0.6080901364174677, "learning_rate": 1.3750875453332801e-06, "loss": 0.2703, "step": 1953 }, { "epoch": 1.5170807453416149, "grad_norm": 0.6765947719205124, "learning_rate": 1.37089026873573e-06, "loss": 0.1517, "step": 1954 }, { "epoch": 1.5178571428571428, "grad_norm": 0.5856848664227117, "learning_rate": 1.3666983898298659e-06, "loss": 0.1992, "step": 1955 }, { "epoch": 1.518633540372671, "grad_norm": 0.5536268836330699, "learning_rate": 1.3625119148504013e-06, "loss": 0.1893, "step": 1956 }, { "epoch": 1.5194099378881987, "grad_norm": 0.7258874772249648, "learning_rate": 1.3583308500240183e-06, "loss": 0.199, "step": 1957 }, { "epoch": 1.5201863354037268, "grad_norm": 0.6243161809571292, "learning_rate": 1.354155201569346e-06, "loss": 0.1685, "step": 1958 }, { "epoch": 1.5209627329192545, "grad_norm": 0.9692943891720399, "learning_rate": 1.3499849756969624e-06, "loss": 0.2506, "step": 1959 }, { "epoch": 1.5217391304347827, "grad_norm": 0.5322450916633802, "learning_rate": 1.3458201786093795e-06, "loss": 0.2496, "step": 1960 }, { "epoch": 1.5225155279503104, "grad_norm": 0.6074627456705953, "learning_rate": 1.341660816501032e-06, "loss": 0.1898, "step": 1961 }, { "epoch": 1.5232919254658386, "grad_norm": 0.639085164897093, "learning_rate": 1.3375068955582754e-06, "loss": 0.2143, "step": 1962 }, { "epoch": 1.5240683229813663, "grad_norm": 0.7987797410549221, "learning_rate": 1.3333584219593682e-06, "loss": 0.2494, "step": 1963 }, { "epoch": 1.5248447204968945, "grad_norm": 0.655521113519075, "learning_rate": 1.329215401874468e-06, "loss": 0.1408, "step": 1964 }, { "epoch": 1.5256211180124224, "grad_norm": 0.7293047901240286, "learning_rate": 1.3250778414656246e-06, "loss": 0.2064, "step": 1965 }, { "epoch": 1.5263975155279503, "grad_norm": 0.595132570744481, "learning_rate": 1.3209457468867614e-06, "loss": 0.2515, "step": 1966 }, { "epoch": 1.5271739130434783, "grad_norm": 0.6234036029781592, "learning_rate": 1.3168191242836787e-06, "loss": 0.1779, "step": 1967 }, { "epoch": 1.5279503105590062, "grad_norm": 0.7231176667772927, "learning_rate": 1.3126979797940336e-06, "loss": 0.723, "step": 1968 }, { "epoch": 1.5287267080745341, "grad_norm": 0.504485564886107, "learning_rate": 1.3085823195473364e-06, "loss": 0.2227, "step": 1969 }, { "epoch": 1.529503105590062, "grad_norm": 0.6989974518895856, "learning_rate": 1.3044721496649427e-06, "loss": 0.2544, "step": 1970 }, { "epoch": 1.53027950310559, "grad_norm": 0.776239230733502, "learning_rate": 1.3003674762600377e-06, "loss": 0.1877, "step": 1971 }, { "epoch": 1.531055900621118, "grad_norm": 0.7000859318051769, "learning_rate": 1.2962683054376375e-06, "loss": 0.2161, "step": 1972 }, { "epoch": 1.5318322981366461, "grad_norm": 0.5691730971216581, "learning_rate": 1.2921746432945692e-06, "loss": 0.2084, "step": 1973 }, { "epoch": 1.5326086956521738, "grad_norm": 0.6646698026308883, "learning_rate": 1.2880864959194666e-06, "loss": 0.1334, "step": 1974 }, { "epoch": 1.533385093167702, "grad_norm": 0.5689158951486097, "learning_rate": 1.2840038693927665e-06, "loss": 0.1989, "step": 1975 }, { "epoch": 1.5341614906832297, "grad_norm": 0.6323490300825377, "learning_rate": 1.2799267697866868e-06, "loss": 0.2036, "step": 1976 }, { "epoch": 1.5349378881987579, "grad_norm": 0.5291111346206925, "learning_rate": 1.275855203165231e-06, "loss": 0.1709, "step": 1977 }, { "epoch": 1.5357142857142856, "grad_norm": 0.6105516135878298, "learning_rate": 1.2717891755841722e-06, "loss": 0.1603, "step": 1978 }, { "epoch": 1.5364906832298137, "grad_norm": 0.7045645988364566, "learning_rate": 1.2677286930910416e-06, "loss": 0.1477, "step": 1979 }, { "epoch": 1.5372670807453415, "grad_norm": 0.6452963618131091, "learning_rate": 1.2636737617251283e-06, "loss": 0.2303, "step": 1980 }, { "epoch": 1.5380434782608696, "grad_norm": 0.6290393445609296, "learning_rate": 1.259624387517459e-06, "loss": 0.2464, "step": 1981 }, { "epoch": 1.5388198757763976, "grad_norm": 0.6296040502033894, "learning_rate": 1.255580576490802e-06, "loss": 0.1639, "step": 1982 }, { "epoch": 1.5395962732919255, "grad_norm": 0.5864644810198766, "learning_rate": 1.2515423346596445e-06, "loss": 0.2145, "step": 1983 }, { "epoch": 1.5403726708074534, "grad_norm": 0.5175947394566879, "learning_rate": 1.2475096680301934e-06, "loss": 0.1769, "step": 1984 }, { "epoch": 1.5411490683229814, "grad_norm": 0.5336331045906748, "learning_rate": 1.2434825826003665e-06, "loss": 0.2174, "step": 1985 }, { "epoch": 1.5419254658385093, "grad_norm": 0.6650844710940713, "learning_rate": 1.2394610843597737e-06, "loss": 0.2337, "step": 1986 }, { "epoch": 1.5427018633540373, "grad_norm": 0.7720074758759522, "learning_rate": 1.235445179289722e-06, "loss": 0.2074, "step": 1987 }, { "epoch": 1.5434782608695652, "grad_norm": 0.6334487966208829, "learning_rate": 1.2314348733631958e-06, "loss": 0.1587, "step": 1988 }, { "epoch": 1.5442546583850931, "grad_norm": 1.1065281197492085, "learning_rate": 1.227430172544849e-06, "loss": 0.7817, "step": 1989 }, { "epoch": 1.545031055900621, "grad_norm": 0.839710670895085, "learning_rate": 1.2234310827910063e-06, "loss": 0.1989, "step": 1990 }, { "epoch": 1.545807453416149, "grad_norm": 0.5124811920079922, "learning_rate": 1.219437610049639e-06, "loss": 0.1716, "step": 1991 }, { "epoch": 1.5465838509316772, "grad_norm": 0.471911189392708, "learning_rate": 1.2154497602603704e-06, "loss": 0.1829, "step": 1992 }, { "epoch": 1.5473602484472049, "grad_norm": 0.6620201576793278, "learning_rate": 1.2114675393544556e-06, "loss": 0.3263, "step": 1993 }, { "epoch": 1.548136645962733, "grad_norm": 0.6400166318583723, "learning_rate": 1.2074909532547824e-06, "loss": 0.282, "step": 1994 }, { "epoch": 1.5489130434782608, "grad_norm": 0.5923559562234854, "learning_rate": 1.2035200078758518e-06, "loss": 0.203, "step": 1995 }, { "epoch": 1.549689440993789, "grad_norm": 0.5689453149364595, "learning_rate": 1.1995547091237814e-06, "loss": 0.2186, "step": 1996 }, { "epoch": 1.5504658385093166, "grad_norm": 0.7793810793889036, "learning_rate": 1.1955950628962876e-06, "loss": 0.2082, "step": 1997 }, { "epoch": 1.5512422360248448, "grad_norm": 0.7503075147533679, "learning_rate": 1.191641075082679e-06, "loss": 0.1699, "step": 1998 }, { "epoch": 1.5520186335403725, "grad_norm": 0.6388069779999476, "learning_rate": 1.1876927515638475e-06, "loss": 0.1531, "step": 1999 }, { "epoch": 1.5527950310559007, "grad_norm": 0.7243243276073082, "learning_rate": 1.1837500982122646e-06, "loss": 0.1904, "step": 2000 }, { "epoch": 1.5527950310559007, "eval_loss": 0.3134809136390686, "eval_runtime": 3.982, "eval_samples_per_second": 13.31, "eval_steps_per_second": 3.516, "step": 2000 }, { "epoch": 1.5535714285714286, "grad_norm": 0.6243469201897738, "learning_rate": 1.1798131208919628e-06, "loss": 0.1273, "step": 2001 }, { "epoch": 1.5543478260869565, "grad_norm": 1.005252430592908, "learning_rate": 1.175881825458537e-06, "loss": 0.2473, "step": 2002 }, { "epoch": 1.5551242236024845, "grad_norm": 0.5676415472250038, "learning_rate": 1.1719562177591292e-06, "loss": 0.2182, "step": 2003 }, { "epoch": 1.5559006211180124, "grad_norm": 0.6865430538317854, "learning_rate": 1.16803630363242e-06, "loss": 0.2137, "step": 2004 }, { "epoch": 1.5566770186335404, "grad_norm": 0.6347985378406492, "learning_rate": 1.164122088908627e-06, "loss": 0.7785, "step": 2005 }, { "epoch": 1.5574534161490683, "grad_norm": 0.5570803613999937, "learning_rate": 1.1602135794094843e-06, "loss": 0.1642, "step": 2006 }, { "epoch": 1.5582298136645962, "grad_norm": 0.6779500239215897, "learning_rate": 1.156310780948247e-06, "loss": 0.18, "step": 2007 }, { "epoch": 1.5590062111801242, "grad_norm": 0.6278505706573736, "learning_rate": 1.152413699329671e-06, "loss": 0.2102, "step": 2008 }, { "epoch": 1.5597826086956523, "grad_norm": 0.5878300992893405, "learning_rate": 1.1485223403500118e-06, "loss": 0.1829, "step": 2009 }, { "epoch": 1.56055900621118, "grad_norm": 0.6647689518439409, "learning_rate": 1.1446367097970134e-06, "loss": 0.1731, "step": 2010 }, { "epoch": 1.5613354037267082, "grad_norm": 0.7085725081928542, "learning_rate": 1.140756813449897e-06, "loss": 0.2326, "step": 2011 }, { "epoch": 1.562111801242236, "grad_norm": 0.7158347991555991, "learning_rate": 1.1368826570793602e-06, "loss": 0.1943, "step": 2012 }, { "epoch": 1.562888198757764, "grad_norm": 0.6869147654029412, "learning_rate": 1.1330142464475574e-06, "loss": 0.2203, "step": 2013 }, { "epoch": 1.5636645962732918, "grad_norm": 0.6119913872122231, "learning_rate": 1.129151587308102e-06, "loss": 0.2228, "step": 2014 }, { "epoch": 1.56444099378882, "grad_norm": 0.6545691918584836, "learning_rate": 1.1252946854060525e-06, "loss": 0.188, "step": 2015 }, { "epoch": 1.5652173913043477, "grad_norm": 0.6844813465886946, "learning_rate": 1.1214435464779006e-06, "loss": 0.1981, "step": 2016 }, { "epoch": 1.5659937888198758, "grad_norm": 0.6434566064571111, "learning_rate": 1.1175981762515714e-06, "loss": 0.2232, "step": 2017 }, { "epoch": 1.5667701863354038, "grad_norm": 0.6446005723471936, "learning_rate": 1.1137585804464062e-06, "loss": 0.1779, "step": 2018 }, { "epoch": 1.5675465838509317, "grad_norm": 0.5950925220787231, "learning_rate": 1.1099247647731592e-06, "loss": 0.2138, "step": 2019 }, { "epoch": 1.5683229813664596, "grad_norm": 0.6136065840094133, "learning_rate": 1.1060967349339897e-06, "loss": 0.3112, "step": 2020 }, { "epoch": 1.5690993788819876, "grad_norm": 0.6347913051575715, "learning_rate": 1.1022744966224469e-06, "loss": 0.1552, "step": 2021 }, { "epoch": 1.5698757763975155, "grad_norm": 0.5543019693538003, "learning_rate": 1.098458055523472e-06, "loss": 0.2234, "step": 2022 }, { "epoch": 1.5706521739130435, "grad_norm": 0.5849301699058854, "learning_rate": 1.0946474173133791e-06, "loss": 0.2353, "step": 2023 }, { "epoch": 1.5714285714285714, "grad_norm": 0.5838751233913393, "learning_rate": 1.0908425876598512e-06, "loss": 0.1282, "step": 2024 }, { "epoch": 1.5722049689440993, "grad_norm": 0.714374206234519, "learning_rate": 1.0870435722219375e-06, "loss": 0.199, "step": 2025 }, { "epoch": 1.5729813664596275, "grad_norm": 0.5221042133606175, "learning_rate": 1.0832503766500324e-06, "loss": 0.1581, "step": 2026 }, { "epoch": 1.5737577639751552, "grad_norm": 0.6295683645861719, "learning_rate": 1.0794630065858807e-06, "loss": 0.23, "step": 2027 }, { "epoch": 1.5745341614906834, "grad_norm": 0.7490401752332905, "learning_rate": 1.0756814676625576e-06, "loss": 0.2299, "step": 2028 }, { "epoch": 1.575310559006211, "grad_norm": 0.7624721314516063, "learning_rate": 1.0719057655044696e-06, "loss": 0.2136, "step": 2029 }, { "epoch": 1.5760869565217392, "grad_norm": 0.6867201756212027, "learning_rate": 1.0681359057273388e-06, "loss": 0.2004, "step": 2030 }, { "epoch": 1.576863354037267, "grad_norm": 0.6121502959954718, "learning_rate": 1.0643718939381965e-06, "loss": 0.1814, "step": 2031 }, { "epoch": 1.5776397515527951, "grad_norm": 0.6865687933788768, "learning_rate": 1.060613735735384e-06, "loss": 0.2224, "step": 2032 }, { "epoch": 1.5784161490683228, "grad_norm": 0.6635178756400925, "learning_rate": 1.0568614367085279e-06, "loss": 0.1334, "step": 2033 }, { "epoch": 1.579192546583851, "grad_norm": 0.6065742474376788, "learning_rate": 1.0531150024385423e-06, "loss": 0.1911, "step": 2034 }, { "epoch": 1.579968944099379, "grad_norm": 0.6916913542248722, "learning_rate": 1.0493744384976218e-06, "loss": 0.2696, "step": 2035 }, { "epoch": 1.5807453416149069, "grad_norm": 0.5678319051945901, "learning_rate": 1.045639750449225e-06, "loss": 0.1598, "step": 2036 }, { "epoch": 1.5815217391304348, "grad_norm": 0.765776087767289, "learning_rate": 1.0419109438480762e-06, "loss": 0.1669, "step": 2037 }, { "epoch": 1.5822981366459627, "grad_norm": 0.5880990258121002, "learning_rate": 1.0381880242401483e-06, "loss": 0.2281, "step": 2038 }, { "epoch": 1.5830745341614907, "grad_norm": 0.7253029649791327, "learning_rate": 1.0344709971626576e-06, "loss": 0.1689, "step": 2039 }, { "epoch": 1.5838509316770186, "grad_norm": 0.5416004994155745, "learning_rate": 1.0307598681440617e-06, "loss": 0.3089, "step": 2040 }, { "epoch": 1.5846273291925466, "grad_norm": 0.6229523251938697, "learning_rate": 1.0270546427040384e-06, "loss": 0.1773, "step": 2041 }, { "epoch": 1.5854037267080745, "grad_norm": 0.6331951519512938, "learning_rate": 1.0233553263534924e-06, "loss": 0.2042, "step": 2042 }, { "epoch": 1.5861801242236024, "grad_norm": 0.7780280293886345, "learning_rate": 1.0196619245945327e-06, "loss": 0.1849, "step": 2043 }, { "epoch": 1.5869565217391304, "grad_norm": 0.6907656684536578, "learning_rate": 1.0159744429204776e-06, "loss": 0.2084, "step": 2044 }, { "epoch": 1.5877329192546585, "grad_norm": 0.6748474895148372, "learning_rate": 1.0122928868158355e-06, "loss": 0.1989, "step": 2045 }, { "epoch": 1.5885093167701863, "grad_norm": 0.6025631840965298, "learning_rate": 1.0086172617563022e-06, "loss": 0.2116, "step": 2046 }, { "epoch": 1.5892857142857144, "grad_norm": 0.8337816157882725, "learning_rate": 1.004947573208756e-06, "loss": 0.2109, "step": 2047 }, { "epoch": 1.5900621118012421, "grad_norm": 0.56048821918126, "learning_rate": 1.0012838266312397e-06, "loss": 0.1472, "step": 2048 }, { "epoch": 1.5908385093167703, "grad_norm": 0.6524286813947466, "learning_rate": 9.976260274729627e-07, "loss": 0.2034, "step": 2049 }, { "epoch": 1.591614906832298, "grad_norm": 0.6045561897365807, "learning_rate": 9.939741811742893e-07, "loss": 0.2322, "step": 2050 }, { "epoch": 1.5923913043478262, "grad_norm": 0.7276698361583366, "learning_rate": 9.903282931667246e-07, "loss": 0.2435, "step": 2051 }, { "epoch": 1.5931677018633539, "grad_norm": 0.6185350484978911, "learning_rate": 9.866883688729189e-07, "loss": 0.2498, "step": 2052 }, { "epoch": 1.593944099378882, "grad_norm": 0.8027730959477859, "learning_rate": 9.830544137066462e-07, "loss": 0.2076, "step": 2053 }, { "epoch": 1.59472049689441, "grad_norm": 0.612856023339252, "learning_rate": 9.794264330728043e-07, "loss": 0.2158, "step": 2054 }, { "epoch": 1.595496894409938, "grad_norm": 0.7588043925170859, "learning_rate": 9.758044323674082e-07, "loss": 0.1781, "step": 2055 }, { "epoch": 1.5962732919254659, "grad_norm": 0.6756541344434406, "learning_rate": 9.721884169775735e-07, "loss": 0.2288, "step": 2056 }, { "epoch": 1.5970496894409938, "grad_norm": 0.6776169235067183, "learning_rate": 9.685783922815185e-07, "loss": 0.2015, "step": 2057 }, { "epoch": 1.5978260869565217, "grad_norm": 0.6961987544764325, "learning_rate": 9.64974363648548e-07, "loss": 0.2135, "step": 2058 }, { "epoch": 1.5986024844720497, "grad_norm": 0.7985114917708772, "learning_rate": 9.613763364390489e-07, "loss": 0.2477, "step": 2059 }, { "epoch": 1.5993788819875776, "grad_norm": 0.610097273715727, "learning_rate": 9.577843160044842e-07, "loss": 0.2057, "step": 2060 }, { "epoch": 1.6001552795031055, "grad_norm": 0.7504388179840579, "learning_rate": 9.541983076873802e-07, "loss": 0.1904, "step": 2061 }, { "epoch": 1.6009316770186337, "grad_norm": 0.634198733030679, "learning_rate": 9.506183168213245e-07, "loss": 0.2423, "step": 2062 }, { "epoch": 1.6017080745341614, "grad_norm": 0.9201024119552077, "learning_rate": 9.4704434873095e-07, "loss": 0.1313, "step": 2063 }, { "epoch": 1.6024844720496896, "grad_norm": 0.769165582380895, "learning_rate": 9.434764087319376e-07, "loss": 0.2302, "step": 2064 }, { "epoch": 1.6032608695652173, "grad_norm": 0.5478742797977199, "learning_rate": 9.399145021309974e-07, "loss": 0.2168, "step": 2065 }, { "epoch": 1.6040372670807455, "grad_norm": 0.8178006425968316, "learning_rate": 9.363586342258662e-07, "loss": 0.2237, "step": 2066 }, { "epoch": 1.6048136645962732, "grad_norm": 0.7696054954813781, "learning_rate": 9.328088103053029e-07, "loss": 0.8015, "step": 2067 }, { "epoch": 1.6055900621118013, "grad_norm": 0.6624741757295508, "learning_rate": 9.292650356490757e-07, "loss": 0.1918, "step": 2068 }, { "epoch": 1.606366459627329, "grad_norm": 0.6629161997621554, "learning_rate": 9.257273155279522e-07, "loss": 0.1604, "step": 2069 }, { "epoch": 1.6071428571428572, "grad_norm": 0.678961473119724, "learning_rate": 9.221956552036992e-07, "loss": 0.169, "step": 2070 }, { "epoch": 1.6079192546583851, "grad_norm": 0.6537158680796149, "learning_rate": 9.186700599290677e-07, "loss": 0.2112, "step": 2071 }, { "epoch": 1.608695652173913, "grad_norm": 0.792668913956635, "learning_rate": 9.151505349477901e-07, "loss": 0.1996, "step": 2072 }, { "epoch": 1.609472049689441, "grad_norm": 0.7425512628403044, "learning_rate": 9.116370854945688e-07, "loss": 0.2105, "step": 2073 }, { "epoch": 1.610248447204969, "grad_norm": 0.6046735418485633, "learning_rate": 9.081297167950682e-07, "loss": 0.2065, "step": 2074 }, { "epoch": 1.611024844720497, "grad_norm": 0.5152595691799187, "learning_rate": 9.04628434065914e-07, "loss": 0.2132, "step": 2075 }, { "epoch": 1.6118012422360248, "grad_norm": 0.7150431701339327, "learning_rate": 9.011332425146734e-07, "loss": 0.2043, "step": 2076 }, { "epoch": 1.6125776397515528, "grad_norm": 0.626211435671538, "learning_rate": 8.976441473398606e-07, "loss": 0.1332, "step": 2077 }, { "epoch": 1.6133540372670807, "grad_norm": 0.6812688888340848, "learning_rate": 8.941611537309158e-07, "loss": 0.2264, "step": 2078 }, { "epoch": 1.6141304347826086, "grad_norm": 0.6593281616776219, "learning_rate": 8.906842668682102e-07, "loss": 0.167, "step": 2079 }, { "epoch": 1.6149068322981366, "grad_norm": 0.7453640560106071, "learning_rate": 8.87213491923029e-07, "loss": 0.191, "step": 2080 }, { "epoch": 1.6156832298136647, "grad_norm": 0.6310209849244803, "learning_rate": 8.837488340575651e-07, "loss": 0.1879, "step": 2081 }, { "epoch": 1.6164596273291925, "grad_norm": 0.5732144180060315, "learning_rate": 8.80290298424919e-07, "loss": 0.1671, "step": 2082 }, { "epoch": 1.6172360248447206, "grad_norm": 0.6769788833260927, "learning_rate": 8.768378901690788e-07, "loss": 0.1659, "step": 2083 }, { "epoch": 1.6180124223602483, "grad_norm": 0.712268812333683, "learning_rate": 8.733916144249255e-07, "loss": 0.2158, "step": 2084 }, { "epoch": 1.6187888198757765, "grad_norm": 0.6901442459219139, "learning_rate": 8.699514763182127e-07, "loss": 0.2342, "step": 2085 }, { "epoch": 1.6195652173913042, "grad_norm": 0.6072959565785248, "learning_rate": 8.665174809655707e-07, "loss": 0.1886, "step": 2086 }, { "epoch": 1.6203416149068324, "grad_norm": 0.6068619714848527, "learning_rate": 8.63089633474492e-07, "loss": 0.257, "step": 2087 }, { "epoch": 1.62111801242236, "grad_norm": 0.5914474204008046, "learning_rate": 8.596679389433232e-07, "loss": 0.1825, "step": 2088 }, { "epoch": 1.6218944099378882, "grad_norm": 0.5846681110773513, "learning_rate": 8.562524024612595e-07, "loss": 0.1752, "step": 2089 }, { "epoch": 1.6226708074534162, "grad_norm": 0.6876228966101708, "learning_rate": 8.528430291083412e-07, "loss": 0.1787, "step": 2090 }, { "epoch": 1.6234472049689441, "grad_norm": 0.6975560610561341, "learning_rate": 8.494398239554358e-07, "loss": 0.2571, "step": 2091 }, { "epoch": 1.624223602484472, "grad_norm": 0.87171620953947, "learning_rate": 8.460427920642422e-07, "loss": 0.2142, "step": 2092 }, { "epoch": 1.625, "grad_norm": 0.5429633744967545, "learning_rate": 8.426519384872733e-07, "loss": 0.2357, "step": 2093 }, { "epoch": 1.625776397515528, "grad_norm": 0.5763974600690878, "learning_rate": 8.392672682678577e-07, "loss": 0.2057, "step": 2094 }, { "epoch": 1.6265527950310559, "grad_norm": 0.5575161616573114, "learning_rate": 8.358887864401216e-07, "loss": 0.1454, "step": 2095 }, { "epoch": 1.6273291925465838, "grad_norm": 0.8750998359078028, "learning_rate": 8.325164980289896e-07, "loss": 0.8414, "step": 2096 }, { "epoch": 1.6281055900621118, "grad_norm": 0.6274574606398781, "learning_rate": 8.291504080501772e-07, "loss": 0.1812, "step": 2097 }, { "epoch": 1.62888198757764, "grad_norm": 0.5643209591949653, "learning_rate": 8.257905215101758e-07, "loss": 0.2229, "step": 2098 }, { "epoch": 1.6296583850931676, "grad_norm": 0.5527997089061604, "learning_rate": 8.224368434062547e-07, "loss": 0.2195, "step": 2099 }, { "epoch": 1.6304347826086958, "grad_norm": 0.6412411203467223, "learning_rate": 8.19089378726447e-07, "loss": 0.2244, "step": 2100 }, { "epoch": 1.6312111801242235, "grad_norm": 0.6207831483261946, "learning_rate": 8.157481324495425e-07, "loss": 0.2079, "step": 2101 }, { "epoch": 1.6319875776397517, "grad_norm": 0.5670704365921443, "learning_rate": 8.12413109545086e-07, "loss": 0.2153, "step": 2102 }, { "epoch": 1.6327639751552794, "grad_norm": 0.7745931384064199, "learning_rate": 8.090843149733651e-07, "loss": 0.7329, "step": 2103 }, { "epoch": 1.6335403726708075, "grad_norm": 0.6959448764171771, "learning_rate": 8.057617536854012e-07, "loss": 0.2264, "step": 2104 }, { "epoch": 1.6343167701863353, "grad_norm": 0.5780964658766106, "learning_rate": 8.02445430622949e-07, "loss": 0.1613, "step": 2105 }, { "epoch": 1.6350931677018634, "grad_norm": 0.6092273222059209, "learning_rate": 7.991353507184801e-07, "loss": 0.1748, "step": 2106 }, { "epoch": 1.6358695652173914, "grad_norm": 0.6699989146500065, "learning_rate": 7.958315188951848e-07, "loss": 0.2482, "step": 2107 }, { "epoch": 1.6366459627329193, "grad_norm": 0.7647249901360793, "learning_rate": 7.92533940066958e-07, "loss": 0.1781, "step": 2108 }, { "epoch": 1.6374223602484472, "grad_norm": 0.6786392813006273, "learning_rate": 7.892426191383928e-07, "loss": 0.1921, "step": 2109 }, { "epoch": 1.6381987577639752, "grad_norm": 0.6887930059339987, "learning_rate": 7.859575610047798e-07, "loss": 0.2545, "step": 2110 }, { "epoch": 1.638975155279503, "grad_norm": 0.9942992834769765, "learning_rate": 7.826787705520894e-07, "loss": 0.7236, "step": 2111 }, { "epoch": 1.639751552795031, "grad_norm": 0.72429537745208, "learning_rate": 7.794062526569735e-07, "loss": 0.1871, "step": 2112 }, { "epoch": 1.640527950310559, "grad_norm": 0.6255433961617222, "learning_rate": 7.761400121867519e-07, "loss": 0.2383, "step": 2113 }, { "epoch": 1.641304347826087, "grad_norm": 0.6920864279867797, "learning_rate": 7.728800539994113e-07, "loss": 0.1642, "step": 2114 }, { "epoch": 1.6420807453416149, "grad_norm": 0.7479447662433372, "learning_rate": 7.696263829435907e-07, "loss": 0.2038, "step": 2115 }, { "epoch": 1.6428571428571428, "grad_norm": 0.716100230190535, "learning_rate": 7.663790038585794e-07, "loss": 0.1408, "step": 2116 }, { "epoch": 1.643633540372671, "grad_norm": 0.6201401561980051, "learning_rate": 7.631379215743106e-07, "loss": 0.1868, "step": 2117 }, { "epoch": 1.6444099378881987, "grad_norm": 0.6844866839964218, "learning_rate": 7.599031409113483e-07, "loss": 0.1809, "step": 2118 }, { "epoch": 1.6451863354037268, "grad_norm": 0.6255836714442998, "learning_rate": 7.566746666808877e-07, "loss": 0.1734, "step": 2119 }, { "epoch": 1.6459627329192545, "grad_norm": 0.5601186145652669, "learning_rate": 7.534525036847406e-07, "loss": 0.1797, "step": 2120 }, { "epoch": 1.6467391304347827, "grad_norm": 0.6678561409733669, "learning_rate": 7.502366567153346e-07, "loss": 0.1689, "step": 2121 }, { "epoch": 1.6475155279503104, "grad_norm": 0.6490121222921542, "learning_rate": 7.470271305557036e-07, "loss": 0.2367, "step": 2122 }, { "epoch": 1.6482919254658386, "grad_norm": 0.7249577786801418, "learning_rate": 7.43823929979478e-07, "loss": 0.1911, "step": 2123 }, { "epoch": 1.6490683229813663, "grad_norm": 0.5614298223845781, "learning_rate": 7.406270597508796e-07, "loss": 0.1852, "step": 2124 }, { "epoch": 1.6498447204968945, "grad_norm": 0.7159435293892111, "learning_rate": 7.374365246247195e-07, "loss": 0.1393, "step": 2125 }, { "epoch": 1.6506211180124224, "grad_norm": 0.7185320151577201, "learning_rate": 7.342523293463799e-07, "loss": 0.212, "step": 2126 }, { "epoch": 1.6513975155279503, "grad_norm": 0.8225083227280462, "learning_rate": 7.310744786518204e-07, "loss": 0.2036, "step": 2127 }, { "epoch": 1.6521739130434783, "grad_norm": 0.5198762613350739, "learning_rate": 7.279029772675572e-07, "loss": 0.1675, "step": 2128 }, { "epoch": 1.6529503105590062, "grad_norm": 0.6836242857604933, "learning_rate": 7.247378299106689e-07, "loss": 0.1751, "step": 2129 }, { "epoch": 1.6537267080745341, "grad_norm": 0.6042749886256453, "learning_rate": 7.215790412887802e-07, "loss": 0.219, "step": 2130 }, { "epoch": 1.654503105590062, "grad_norm": 0.5389759802652087, "learning_rate": 7.184266161000581e-07, "loss": 0.2015, "step": 2131 }, { "epoch": 1.65527950310559, "grad_norm": 0.6292653177717036, "learning_rate": 7.152805590332079e-07, "loss": 0.2489, "step": 2132 }, { "epoch": 1.656055900621118, "grad_norm": 0.5082296249660022, "learning_rate": 7.12140874767459e-07, "loss": 0.2368, "step": 2133 }, { "epoch": 1.6568322981366461, "grad_norm": 0.8310618855856859, "learning_rate": 7.090075679725683e-07, "loss": 0.164, "step": 2134 }, { "epoch": 1.6576086956521738, "grad_norm": 0.5839729170501283, "learning_rate": 7.058806433088011e-07, "loss": 0.1601, "step": 2135 }, { "epoch": 1.658385093167702, "grad_norm": 0.6050580724069987, "learning_rate": 7.027601054269329e-07, "loss": 0.1871, "step": 2136 }, { "epoch": 1.6591614906832297, "grad_norm": 0.7213032520471859, "learning_rate": 6.99645958968242e-07, "loss": 0.2545, "step": 2137 }, { "epoch": 1.6599378881987579, "grad_norm": 0.9635490579518035, "learning_rate": 6.965382085644968e-07, "loss": 0.129, "step": 2138 }, { "epoch": 1.6607142857142856, "grad_norm": 0.7301959296312487, "learning_rate": 6.934368588379553e-07, "loss": 0.2349, "step": 2139 }, { "epoch": 1.6614906832298137, "grad_norm": 0.6549885522627227, "learning_rate": 6.903419144013556e-07, "loss": 0.182, "step": 2140 }, { "epoch": 1.6622670807453415, "grad_norm": 0.9745751768486177, "learning_rate": 6.872533798579061e-07, "loss": 0.6803, "step": 2141 }, { "epoch": 1.6630434782608696, "grad_norm": 0.699594204144472, "learning_rate": 6.841712598012867e-07, "loss": 0.1615, "step": 2142 }, { "epoch": 1.6638198757763976, "grad_norm": 0.6923240096508857, "learning_rate": 6.810955588156315e-07, "loss": 0.1416, "step": 2143 }, { "epoch": 1.6645962732919255, "grad_norm": 0.7888277556254891, "learning_rate": 6.780262814755284e-07, "loss": 0.2588, "step": 2144 }, { "epoch": 1.6653726708074534, "grad_norm": 0.6833884596777732, "learning_rate": 6.749634323460158e-07, "loss": 0.2775, "step": 2145 }, { "epoch": 1.6661490683229814, "grad_norm": 0.7253747570481633, "learning_rate": 6.719070159825642e-07, "loss": 0.1479, "step": 2146 }, { "epoch": 1.6669254658385093, "grad_norm": 0.6687219246469147, "learning_rate": 6.688570369310831e-07, "loss": 0.1959, "step": 2147 }, { "epoch": 1.6677018633540373, "grad_norm": 0.6146584988663688, "learning_rate": 6.65813499727902e-07, "loss": 0.2041, "step": 2148 }, { "epoch": 1.6684782608695652, "grad_norm": 0.5272185503806076, "learning_rate": 6.627764088997735e-07, "loss": 0.2197, "step": 2149 }, { "epoch": 1.6692546583850931, "grad_norm": 0.6969531052324376, "learning_rate": 6.597457689638598e-07, "loss": 0.1871, "step": 2150 }, { "epoch": 1.670031055900621, "grad_norm": 0.5915226046697082, "learning_rate": 6.567215844277286e-07, "loss": 0.213, "step": 2151 }, { "epoch": 1.670807453416149, "grad_norm": 0.7009300799843693, "learning_rate": 6.53703859789348e-07, "loss": 0.7353, "step": 2152 }, { "epoch": 1.6715838509316772, "grad_norm": 0.7761820432550739, "learning_rate": 6.506925995370756e-07, "loss": 0.1948, "step": 2153 }, { "epoch": 1.6723602484472049, "grad_norm": 0.6158495146766632, "learning_rate": 6.476878081496579e-07, "loss": 0.2188, "step": 2154 }, { "epoch": 1.673136645962733, "grad_norm": 0.5529235092984697, "learning_rate": 6.446894900962164e-07, "loss": 0.2093, "step": 2155 }, { "epoch": 1.6739130434782608, "grad_norm": 0.6051489889928287, "learning_rate": 6.416976498362432e-07, "loss": 0.1945, "step": 2156 }, { "epoch": 1.674689440993789, "grad_norm": 0.6583333428830438, "learning_rate": 6.387122918196043e-07, "loss": 0.2136, "step": 2157 }, { "epoch": 1.6754658385093166, "grad_norm": 0.6334778872210811, "learning_rate": 6.357334204865156e-07, "loss": 0.1396, "step": 2158 }, { "epoch": 1.6762422360248448, "grad_norm": 0.645915841382617, "learning_rate": 6.327610402675477e-07, "loss": 0.2173, "step": 2159 }, { "epoch": 1.6770186335403725, "grad_norm": 0.6274881894581916, "learning_rate": 6.29795155583619e-07, "loss": 0.165, "step": 2160 }, { "epoch": 1.6777950310559007, "grad_norm": 0.624827347420838, "learning_rate": 6.268357708459838e-07, "loss": 0.1668, "step": 2161 }, { "epoch": 1.6785714285714286, "grad_norm": 0.5455425757622606, "learning_rate": 6.238828904562316e-07, "loss": 0.2025, "step": 2162 }, { "epoch": 1.6793478260869565, "grad_norm": 0.6936358157299921, "learning_rate": 6.20936518806275e-07, "loss": 0.1786, "step": 2163 }, { "epoch": 1.6801242236024845, "grad_norm": 0.8327923088732178, "learning_rate": 6.179966602783493e-07, "loss": 0.1822, "step": 2164 }, { "epoch": 1.6809006211180124, "grad_norm": 0.8139978032058921, "learning_rate": 6.15063319244999e-07, "loss": 0.2363, "step": 2165 }, { "epoch": 1.6816770186335404, "grad_norm": 0.8100524909539383, "learning_rate": 6.121365000690766e-07, "loss": 0.2732, "step": 2166 }, { "epoch": 1.6824534161490683, "grad_norm": 0.5640888886166102, "learning_rate": 6.092162071037355e-07, "loss": 0.2193, "step": 2167 }, { "epoch": 1.6832298136645962, "grad_norm": 0.7918272380439385, "learning_rate": 6.063024446924198e-07, "loss": 0.1738, "step": 2168 }, { "epoch": 1.6840062111801242, "grad_norm": 0.6523094305103532, "learning_rate": 6.033952171688645e-07, "loss": 0.156, "step": 2169 }, { "epoch": 1.6847826086956523, "grad_norm": 0.660742921829772, "learning_rate": 6.004945288570813e-07, "loss": 0.1879, "step": 2170 }, { "epoch": 1.68555900621118, "grad_norm": 0.5520758714195751, "learning_rate": 5.976003840713562e-07, "loss": 0.247, "step": 2171 }, { "epoch": 1.6863354037267082, "grad_norm": 0.6805988486319322, "learning_rate": 5.947127871162455e-07, "loss": 0.2568, "step": 2172 }, { "epoch": 1.687111801242236, "grad_norm": 0.668939302906692, "learning_rate": 5.918317422865644e-07, "loss": 0.2583, "step": 2173 }, { "epoch": 1.687888198757764, "grad_norm": 0.4901701763544616, "learning_rate": 5.889572538673826e-07, "loss": 0.1972, "step": 2174 }, { "epoch": 1.6886645962732918, "grad_norm": 0.6423368480344289, "learning_rate": 5.860893261340222e-07, "loss": 0.1516, "step": 2175 }, { "epoch": 1.68944099378882, "grad_norm": 0.8106446302009761, "learning_rate": 5.832279633520411e-07, "loss": 0.2315, "step": 2176 }, { "epoch": 1.6902173913043477, "grad_norm": 0.5293092372927625, "learning_rate": 5.803731697772391e-07, "loss": 0.2044, "step": 2177 }, { "epoch": 1.6909937888198758, "grad_norm": 0.6317876762075053, "learning_rate": 5.775249496556406e-07, "loss": 0.1128, "step": 2178 }, { "epoch": 1.6917701863354038, "grad_norm": 0.7083033831611477, "learning_rate": 5.746833072234947e-07, "loss": 0.2125, "step": 2179 }, { "epoch": 1.6925465838509317, "grad_norm": 0.881233418356288, "learning_rate": 5.718482467072695e-07, "loss": 0.1712, "step": 2180 }, { "epoch": 1.6933229813664596, "grad_norm": 0.6047455767688756, "learning_rate": 5.690197723236396e-07, "loss": 0.225, "step": 2181 }, { "epoch": 1.6940993788819876, "grad_norm": 0.6607398150362706, "learning_rate": 5.661978882794883e-07, "loss": 0.1514, "step": 2182 }, { "epoch": 1.6948757763975155, "grad_norm": 0.6377751210887682, "learning_rate": 5.633825987718922e-07, "loss": 0.7323, "step": 2183 }, { "epoch": 1.6956521739130435, "grad_norm": 0.641683472711413, "learning_rate": 5.60573907988124e-07, "loss": 0.1832, "step": 2184 }, { "epoch": 1.6964285714285714, "grad_norm": 0.5641021927489731, "learning_rate": 5.577718201056392e-07, "loss": 0.2345, "step": 2185 }, { "epoch": 1.6972049689440993, "grad_norm": 0.6320665483232464, "learning_rate": 5.54976339292072e-07, "loss": 0.1566, "step": 2186 }, { "epoch": 1.6979813664596275, "grad_norm": 0.7422952167278752, "learning_rate": 5.521874697052332e-07, "loss": 0.7212, "step": 2187 }, { "epoch": 1.6987577639751552, "grad_norm": 0.8672803362148844, "learning_rate": 5.494052154930956e-07, "loss": 0.7863, "step": 2188 }, { "epoch": 1.6995341614906834, "grad_norm": 0.6420371111892704, "learning_rate": 5.466295807937988e-07, "loss": 0.1643, "step": 2189 }, { "epoch": 1.700310559006211, "grad_norm": 0.6794646765653142, "learning_rate": 5.438605697356309e-07, "loss": 0.108, "step": 2190 }, { "epoch": 1.7010869565217392, "grad_norm": 0.5305381844334287, "learning_rate": 5.410981864370313e-07, "loss": 0.2589, "step": 2191 }, { "epoch": 1.701863354037267, "grad_norm": 0.7135206518815389, "learning_rate": 5.383424350065825e-07, "loss": 0.1829, "step": 2192 }, { "epoch": 1.7026397515527951, "grad_norm": 0.7006735375964502, "learning_rate": 5.355933195430025e-07, "loss": 0.2402, "step": 2193 }, { "epoch": 1.7034161490683228, "grad_norm": 1.0967405815444695, "learning_rate": 5.328508441351382e-07, "loss": 0.1744, "step": 2194 }, { "epoch": 1.704192546583851, "grad_norm": 0.6693605010430574, "learning_rate": 5.301150128619631e-07, "loss": 0.2729, "step": 2195 }, { "epoch": 1.704968944099379, "grad_norm": 0.6613744010365211, "learning_rate": 5.273858297925649e-07, "loss": 0.3309, "step": 2196 }, { "epoch": 1.7057453416149069, "grad_norm": 0.6829891271152455, "learning_rate": 5.246632989861472e-07, "loss": 0.2717, "step": 2197 }, { "epoch": 1.7065217391304348, "grad_norm": 0.5407987574456078, "learning_rate": 5.219474244920164e-07, "loss": 0.1436, "step": 2198 }, { "epoch": 1.7072981366459627, "grad_norm": 0.5920323249526167, "learning_rate": 5.192382103495819e-07, "loss": 0.2283, "step": 2199 }, { "epoch": 1.7080745341614907, "grad_norm": 0.6007896356165294, "learning_rate": 5.165356605883432e-07, "loss": 0.2673, "step": 2200 }, { "epoch": 1.7088509316770186, "grad_norm": 0.7437750204446826, "learning_rate": 5.138397792278898e-07, "loss": 0.1856, "step": 2201 }, { "epoch": 1.7096273291925466, "grad_norm": 0.7288036008399335, "learning_rate": 5.111505702778935e-07, "loss": 0.2215, "step": 2202 }, { "epoch": 1.7104037267080745, "grad_norm": 0.5696169871530257, "learning_rate": 5.084680377381007e-07, "loss": 0.2127, "step": 2203 }, { "epoch": 1.7111801242236024, "grad_norm": 0.7538168557320032, "learning_rate": 5.057921855983288e-07, "loss": 0.2073, "step": 2204 }, { "epoch": 1.7119565217391304, "grad_norm": 0.5723366223429379, "learning_rate": 5.031230178384594e-07, "loss": 0.1371, "step": 2205 }, { "epoch": 1.7127329192546585, "grad_norm": 0.6440355763933053, "learning_rate": 5.004605384284295e-07, "loss": 0.2335, "step": 2206 }, { "epoch": 1.7135093167701863, "grad_norm": 0.6170213706307033, "learning_rate": 4.978047513282325e-07, "loss": 0.2609, "step": 2207 }, { "epoch": 1.7142857142857144, "grad_norm": 0.7073227162206148, "learning_rate": 4.951556604879049e-07, "loss": 0.7518, "step": 2208 }, { "epoch": 1.7150621118012421, "grad_norm": 0.7412644737879149, "learning_rate": 4.925132698475255e-07, "loss": 0.215, "step": 2209 }, { "epoch": 1.7158385093167703, "grad_norm": 0.7859108547722722, "learning_rate": 4.898775833372055e-07, "loss": 0.1444, "step": 2210 }, { "epoch": 1.716614906832298, "grad_norm": 0.5136352003228063, "learning_rate": 4.872486048770875e-07, "loss": 0.1591, "step": 2211 }, { "epoch": 1.7173913043478262, "grad_norm": 0.702562522987539, "learning_rate": 4.846263383773364e-07, "loss": 0.1698, "step": 2212 }, { "epoch": 1.7181677018633539, "grad_norm": 0.686147994906913, "learning_rate": 4.820107877381319e-07, "loss": 0.1875, "step": 2213 }, { "epoch": 1.718944099378882, "grad_norm": 0.6737259604765029, "learning_rate": 4.794019568496683e-07, "loss": 0.1868, "step": 2214 }, { "epoch": 1.71972049689441, "grad_norm": 0.6191120126856626, "learning_rate": 4.7679984959214275e-07, "loss": 0.2225, "step": 2215 }, { "epoch": 1.720496894409938, "grad_norm": 0.8153195783039732, "learning_rate": 4.7420446983575286e-07, "loss": 0.2284, "step": 2216 }, { "epoch": 1.7212732919254659, "grad_norm": 0.5491628692125796, "learning_rate": 4.716158214406913e-07, "loss": 0.1627, "step": 2217 }, { "epoch": 1.7220496894409938, "grad_norm": 0.747447249272742, "learning_rate": 4.690339082571366e-07, "loss": 0.2166, "step": 2218 }, { "epoch": 1.7228260869565217, "grad_norm": 0.665034877116582, "learning_rate": 4.66458734125253e-07, "loss": 0.2239, "step": 2219 }, { "epoch": 1.7236024844720497, "grad_norm": 0.8012777195090058, "learning_rate": 4.638903028751796e-07, "loss": 0.1879, "step": 2220 }, { "epoch": 1.7243788819875776, "grad_norm": 0.624424671916942, "learning_rate": 4.613286183270255e-07, "loss": 0.25, "step": 2221 }, { "epoch": 1.7251552795031055, "grad_norm": 0.6654196678598928, "learning_rate": 4.5877368429086857e-07, "loss": 0.2275, "step": 2222 }, { "epoch": 1.7259316770186337, "grad_norm": 0.6442391029568874, "learning_rate": 4.5622550456674244e-07, "loss": 0.1599, "step": 2223 }, { "epoch": 1.7267080745341614, "grad_norm": 0.666520706427069, "learning_rate": 4.5368408294463873e-07, "loss": 0.1991, "step": 2224 }, { "epoch": 1.7274844720496896, "grad_norm": 0.7313083873330023, "learning_rate": 4.511494232044944e-07, "loss": 0.1975, "step": 2225 }, { "epoch": 1.7282608695652173, "grad_norm": 0.8480767253300301, "learning_rate": 4.486215291161894e-07, "loss": 0.1898, "step": 2226 }, { "epoch": 1.7290372670807455, "grad_norm": 0.9504328651106198, "learning_rate": 4.461004044395445e-07, "loss": 0.7373, "step": 2227 }, { "epoch": 1.7298136645962732, "grad_norm": 0.7519979290598707, "learning_rate": 4.4358605292430743e-07, "loss": 0.2251, "step": 2228 }, { "epoch": 1.7305900621118013, "grad_norm": 0.4993189195063605, "learning_rate": 4.4107847831015516e-07, "loss": 0.1833, "step": 2229 }, { "epoch": 1.731366459627329, "grad_norm": 0.6768878854028756, "learning_rate": 4.385776843266842e-07, "loss": 0.273, "step": 2230 }, { "epoch": 1.7321428571428572, "grad_norm": 0.7056048273278474, "learning_rate": 4.3608367469340553e-07, "loss": 0.1816, "step": 2231 }, { "epoch": 1.7329192546583851, "grad_norm": 0.7354841031779735, "learning_rate": 4.335964531197401e-07, "loss": 0.1669, "step": 2232 }, { "epoch": 1.733695652173913, "grad_norm": 0.6584853832580064, "learning_rate": 4.311160233050121e-07, "loss": 0.1799, "step": 2233 }, { "epoch": 1.734472049689441, "grad_norm": 0.6003045029459562, "learning_rate": 4.286423889384456e-07, "loss": 0.1597, "step": 2234 }, { "epoch": 1.735248447204969, "grad_norm": 0.7379016191425144, "learning_rate": 4.2617555369915633e-07, "loss": 0.2381, "step": 2235 }, { "epoch": 1.736024844720497, "grad_norm": 0.7422256207793142, "learning_rate": 4.2371552125614614e-07, "loss": 0.2101, "step": 2236 }, { "epoch": 1.7368012422360248, "grad_norm": 0.5201458402406661, "learning_rate": 4.212622952683026e-07, "loss": 0.2269, "step": 2237 }, { "epoch": 1.7375776397515528, "grad_norm": 0.6967821739871902, "learning_rate": 4.1881587938438606e-07, "loss": 0.2724, "step": 2238 }, { "epoch": 1.7383540372670807, "grad_norm": 3.4437977788508913, "learning_rate": 4.163762772430313e-07, "loss": 0.4898, "step": 2239 }, { "epoch": 1.7391304347826086, "grad_norm": 0.6704453694285265, "learning_rate": 4.139434924727359e-07, "loss": 0.1528, "step": 2240 }, { "epoch": 1.7399068322981366, "grad_norm": 0.6703721628364068, "learning_rate": 4.1151752869185826e-07, "loss": 0.2057, "step": 2241 }, { "epoch": 1.7406832298136647, "grad_norm": 0.5625814740484841, "learning_rate": 4.090983895086137e-07, "loss": 0.1244, "step": 2242 }, { "epoch": 1.7414596273291925, "grad_norm": 0.5969396798470712, "learning_rate": 4.0668607852106525e-07, "loss": 0.1501, "step": 2243 }, { "epoch": 1.7422360248447206, "grad_norm": 0.7504514965885776, "learning_rate": 4.0428059931712116e-07, "loss": 0.1704, "step": 2244 }, { "epoch": 1.7430124223602483, "grad_norm": 0.60941368319685, "learning_rate": 4.018819554745262e-07, "loss": 0.2257, "step": 2245 }, { "epoch": 1.7437888198757765, "grad_norm": 0.7172771056540518, "learning_rate": 3.994901505608628e-07, "loss": 0.1487, "step": 2246 }, { "epoch": 1.7445652173913042, "grad_norm": 0.6376831081950538, "learning_rate": 3.9710518813353914e-07, "loss": 0.1918, "step": 2247 }, { "epoch": 1.7453416149068324, "grad_norm": 0.6186409450553324, "learning_rate": 3.94727071739785e-07, "loss": 0.1845, "step": 2248 }, { "epoch": 1.74611801242236, "grad_norm": 0.7473359333966668, "learning_rate": 3.9235580491665146e-07, "loss": 0.2069, "step": 2249 }, { "epoch": 1.7468944099378882, "grad_norm": 0.718078165397013, "learning_rate": 3.899913911909986e-07, "loss": 0.2432, "step": 2250 }, { "epoch": 1.7476708074534162, "grad_norm": 0.6145698257554367, "learning_rate": 3.8763383407949496e-07, "loss": 0.1897, "step": 2251 }, { "epoch": 1.7484472049689441, "grad_norm": 0.81745490690453, "learning_rate": 3.8528313708861173e-07, "loss": 0.2208, "step": 2252 }, { "epoch": 1.749223602484472, "grad_norm": 0.5921968015240198, "learning_rate": 3.829393037146156e-07, "loss": 0.1729, "step": 2253 }, { "epoch": 1.75, "grad_norm": 0.7339162132570188, "learning_rate": 3.8060233744356634e-07, "loss": 0.2192, "step": 2254 }, { "epoch": 1.750776397515528, "grad_norm": 0.6887451309405772, "learning_rate": 3.782722417513085e-07, "loss": 0.1989, "step": 2255 }, { "epoch": 1.7515527950310559, "grad_norm": 0.6649874173520368, "learning_rate": 3.759490201034677e-07, "loss": 0.2154, "step": 2256 }, { "epoch": 1.7523291925465838, "grad_norm": 0.5846371774457635, "learning_rate": 3.736326759554476e-07, "loss": 0.2638, "step": 2257 }, { "epoch": 1.7531055900621118, "grad_norm": 0.5491831368246992, "learning_rate": 3.7132321275241966e-07, "loss": 0.2397, "step": 2258 }, { "epoch": 1.75388198757764, "grad_norm": 0.5505025473615571, "learning_rate": 3.690206339293245e-07, "loss": 0.2088, "step": 2259 }, { "epoch": 1.7546583850931676, "grad_norm": 0.690297356604943, "learning_rate": 3.6672494291086156e-07, "loss": 0.213, "step": 2260 }, { "epoch": 1.7554347826086958, "grad_norm": 0.7531256571281513, "learning_rate": 3.6443614311148456e-07, "loss": 0.2529, "step": 2261 }, { "epoch": 1.7562111801242235, "grad_norm": 0.5599586451752269, "learning_rate": 3.6215423793540036e-07, "loss": 0.1763, "step": 2262 }, { "epoch": 1.7569875776397517, "grad_norm": 0.5633755923323782, "learning_rate": 3.598792307765592e-07, "loss": 0.1993, "step": 2263 }, { "epoch": 1.7577639751552794, "grad_norm": 0.5818715486643902, "learning_rate": 3.576111250186526e-07, "loss": 0.2449, "step": 2264 }, { "epoch": 1.7585403726708075, "grad_norm": 0.7038993397855446, "learning_rate": 3.5534992403510825e-07, "loss": 0.2323, "step": 2265 }, { "epoch": 1.7593167701863353, "grad_norm": 0.6830476547994361, "learning_rate": 3.530956311890815e-07, "loss": 0.3585, "step": 2266 }, { "epoch": 1.7600931677018634, "grad_norm": 0.6676487518654834, "learning_rate": 3.5084824983345643e-07, "loss": 0.202, "step": 2267 }, { "epoch": 1.7608695652173914, "grad_norm": 0.5259783570799547, "learning_rate": 3.486077833108342e-07, "loss": 0.2111, "step": 2268 }, { "epoch": 1.7616459627329193, "grad_norm": 0.8929659454978657, "learning_rate": 3.4637423495353374e-07, "loss": 0.1974, "step": 2269 }, { "epoch": 1.7624223602484472, "grad_norm": 0.8129995899329389, "learning_rate": 3.441476080835826e-07, "loss": 0.7971, "step": 2270 }, { "epoch": 1.7631987577639752, "grad_norm": 0.6565161618474293, "learning_rate": 3.419279060127145e-07, "loss": 0.6604, "step": 2271 }, { "epoch": 1.763975155279503, "grad_norm": 0.800538546799533, "learning_rate": 3.397151320423647e-07, "loss": 0.1972, "step": 2272 }, { "epoch": 1.764751552795031, "grad_norm": 0.6461909344241639, "learning_rate": 3.375092894636617e-07, "loss": 0.2596, "step": 2273 }, { "epoch": 1.765527950310559, "grad_norm": 0.5518162092350322, "learning_rate": 3.353103815574282e-07, "loss": 0.2126, "step": 2274 }, { "epoch": 1.766304347826087, "grad_norm": 0.6530227634384478, "learning_rate": 3.331184115941694e-07, "loss": 0.1546, "step": 2275 }, { "epoch": 1.7670807453416149, "grad_norm": 0.7134030594927142, "learning_rate": 3.309333828340722e-07, "loss": 0.2304, "step": 2276 }, { "epoch": 1.7678571428571428, "grad_norm": 0.6115579196764442, "learning_rate": 3.287552985270015e-07, "loss": 0.2084, "step": 2277 }, { "epoch": 1.768633540372671, "grad_norm": 0.5729612692651621, "learning_rate": 3.2658416191249045e-07, "loss": 0.1518, "step": 2278 }, { "epoch": 1.7694099378881987, "grad_norm": 0.7390225801521411, "learning_rate": 3.2441997621974254e-07, "loss": 0.2209, "step": 2279 }, { "epoch": 1.7701863354037268, "grad_norm": 0.5649368892086003, "learning_rate": 3.2226274466761906e-07, "loss": 0.2904, "step": 2280 }, { "epoch": 1.7709627329192545, "grad_norm": 0.750673854717588, "learning_rate": 3.2011247046464076e-07, "loss": 0.2062, "step": 2281 }, { "epoch": 1.7717391304347827, "grad_norm": 0.6441453746703912, "learning_rate": 3.179691568089799e-07, "loss": 0.1841, "step": 2282 }, { "epoch": 1.7725155279503104, "grad_norm": 0.6221262568664719, "learning_rate": 3.1583280688845477e-07, "loss": 0.1909, "step": 2283 }, { "epoch": 1.7732919254658386, "grad_norm": 0.610305231298291, "learning_rate": 3.137034238805292e-07, "loss": 0.2578, "step": 2284 }, { "epoch": 1.7740683229813663, "grad_norm": 0.6075231603594251, "learning_rate": 3.115810109523015e-07, "loss": 0.1541, "step": 2285 }, { "epoch": 1.7748447204968945, "grad_norm": 0.7706060124359744, "learning_rate": 3.0946557126050435e-07, "loss": 0.1808, "step": 2286 }, { "epoch": 1.7756211180124224, "grad_norm": 0.619503702946644, "learning_rate": 3.0735710795150086e-07, "loss": 0.2869, "step": 2287 }, { "epoch": 1.7763975155279503, "grad_norm": 0.6755168476439153, "learning_rate": 3.0525562416127473e-07, "loss": 0.1943, "step": 2288 }, { "epoch": 1.7771739130434783, "grad_norm": 0.6339611426325776, "learning_rate": 3.0316112301543243e-07, "loss": 0.2313, "step": 2289 }, { "epoch": 1.7779503105590062, "grad_norm": 0.7027413890971433, "learning_rate": 3.010736076291909e-07, "loss": 0.198, "step": 2290 }, { "epoch": 1.7787267080745341, "grad_norm": 0.6417501330309919, "learning_rate": 2.9899308110737933e-07, "loss": 0.1928, "step": 2291 }, { "epoch": 1.779503105590062, "grad_norm": 0.6237663165707273, "learning_rate": 2.9691954654443353e-07, "loss": 0.1591, "step": 2292 }, { "epoch": 1.78027950310559, "grad_norm": 0.5673410647196676, "learning_rate": 2.9485300702438547e-07, "loss": 0.2109, "step": 2293 }, { "epoch": 1.781055900621118, "grad_norm": 0.8399254983943069, "learning_rate": 2.927934656208681e-07, "loss": 0.2542, "step": 2294 }, { "epoch": 1.7818322981366461, "grad_norm": 0.6544719008138792, "learning_rate": 2.9074092539710275e-07, "loss": 0.1797, "step": 2295 }, { "epoch": 1.7826086956521738, "grad_norm": 0.5779761497137368, "learning_rate": 2.88695389405898e-07, "loss": 0.2609, "step": 2296 }, { "epoch": 1.783385093167702, "grad_norm": 0.7375434958769704, "learning_rate": 2.866568606896464e-07, "loss": 0.1575, "step": 2297 }, { "epoch": 1.7841614906832297, "grad_norm": 0.578328863841798, "learning_rate": 2.8462534228031636e-07, "loss": 0.2455, "step": 2298 }, { "epoch": 1.7849378881987579, "grad_norm": 0.5474434888154038, "learning_rate": 2.82600837199451e-07, "loss": 0.2005, "step": 2299 }, { "epoch": 1.7857142857142856, "grad_norm": 0.6458628971644762, "learning_rate": 2.8058334845816214e-07, "loss": 0.2023, "step": 2300 }, { "epoch": 1.7864906832298137, "grad_norm": 0.6718719574618979, "learning_rate": 2.785728790571257e-07, "loss": 0.1654, "step": 2301 }, { "epoch": 1.7872670807453415, "grad_norm": 0.7287781253657167, "learning_rate": 2.765694319865775e-07, "loss": 0.2713, "step": 2302 }, { "epoch": 1.7880434782608696, "grad_norm": 0.8193887886796778, "learning_rate": 2.745730102263078e-07, "loss": 0.1499, "step": 2303 }, { "epoch": 1.7888198757763976, "grad_norm": 0.6352883713738162, "learning_rate": 2.725836167456608e-07, "loss": 0.1452, "step": 2304 }, { "epoch": 1.7895962732919255, "grad_norm": 0.529132978109686, "learning_rate": 2.7060125450352413e-07, "loss": 0.1711, "step": 2305 }, { "epoch": 1.7903726708074534, "grad_norm": 0.6830796784091436, "learning_rate": 2.6862592644832875e-07, "loss": 0.2263, "step": 2306 }, { "epoch": 1.7911490683229814, "grad_norm": 0.5727094004063175, "learning_rate": 2.666576355180445e-07, "loss": 0.2249, "step": 2307 }, { "epoch": 1.7919254658385093, "grad_norm": 0.7203346360158414, "learning_rate": 2.646963846401729e-07, "loss": 0.1911, "step": 2308 }, { "epoch": 1.7927018633540373, "grad_norm": 0.6107095868350646, "learning_rate": 2.627421767317467e-07, "loss": 0.1905, "step": 2309 }, { "epoch": 1.7934782608695652, "grad_norm": 0.5372429532094284, "learning_rate": 2.6079501469932154e-07, "loss": 0.2058, "step": 2310 }, { "epoch": 1.7942546583850931, "grad_norm": 0.5444659357981373, "learning_rate": 2.588549014389735e-07, "loss": 0.161, "step": 2311 }, { "epoch": 1.795031055900621, "grad_norm": 0.6531218078489792, "learning_rate": 2.569218398362971e-07, "loss": 0.2035, "step": 2312 }, { "epoch": 1.795807453416149, "grad_norm": 0.6449306566912736, "learning_rate": 2.5499583276639537e-07, "loss": 0.1628, "step": 2313 }, { "epoch": 1.7965838509316772, "grad_norm": 0.7417530191958079, "learning_rate": 2.530768830938818e-07, "loss": 0.174, "step": 2314 }, { "epoch": 1.7973602484472049, "grad_norm": 0.7125948849242233, "learning_rate": 2.511649936728705e-07, "loss": 0.2407, "step": 2315 }, { "epoch": 1.798136645962733, "grad_norm": 0.695861562966091, "learning_rate": 2.49260167346978e-07, "loss": 0.2236, "step": 2316 }, { "epoch": 1.7989130434782608, "grad_norm": 0.6202782359910222, "learning_rate": 2.4736240694931245e-07, "loss": 0.2067, "step": 2317 }, { "epoch": 1.799689440993789, "grad_norm": 0.7222503705584108, "learning_rate": 2.4547171530247437e-07, "loss": 0.238, "step": 2318 }, { "epoch": 1.8004658385093166, "grad_norm": 0.729752150180235, "learning_rate": 2.4358809521855096e-07, "loss": 0.2125, "step": 2319 }, { "epoch": 1.8012422360248448, "grad_norm": 0.6115805068401482, "learning_rate": 2.417115494991107e-07, "loss": 0.1554, "step": 2320 }, { "epoch": 1.8020186335403725, "grad_norm": 0.5743668032100842, "learning_rate": 2.398420809351992e-07, "loss": 0.144, "step": 2321 }, { "epoch": 1.8027950310559007, "grad_norm": 0.6227579781375686, "learning_rate": 2.3797969230733853e-07, "loss": 0.1774, "step": 2322 }, { "epoch": 1.8035714285714286, "grad_norm": 0.5270395751050219, "learning_rate": 2.3612438638551837e-07, "loss": 0.1907, "step": 2323 }, { "epoch": 1.8043478260869565, "grad_norm": 0.812375998226221, "learning_rate": 2.3427616592919587e-07, "loss": 0.2226, "step": 2324 }, { "epoch": 1.8051242236024845, "grad_norm": 0.7584646436549877, "learning_rate": 2.3243503368728882e-07, "loss": 0.2013, "step": 2325 }, { "epoch": 1.8059006211180124, "grad_norm": 0.48305525093207097, "learning_rate": 2.3060099239817059e-07, "loss": 0.1679, "step": 2326 }, { "epoch": 1.8066770186335404, "grad_norm": 0.7073268897618278, "learning_rate": 2.287740447896719e-07, "loss": 0.149, "step": 2327 }, { "epoch": 1.8074534161490683, "grad_norm": 0.5592207238242515, "learning_rate": 2.2695419357906978e-07, "loss": 0.251, "step": 2328 }, { "epoch": 1.8082298136645962, "grad_norm": 0.6160041542305085, "learning_rate": 2.2514144147308791e-07, "loss": 0.161, "step": 2329 }, { "epoch": 1.8090062111801242, "grad_norm": 0.8112211597896865, "learning_rate": 2.2333579116788962e-07, "loss": 0.1511, "step": 2330 }, { "epoch": 1.8097826086956523, "grad_norm": 0.5455747839176426, "learning_rate": 2.2153724534907884e-07, "loss": 0.2198, "step": 2331 }, { "epoch": 1.81055900621118, "grad_norm": 0.6632160097902242, "learning_rate": 2.197458066916891e-07, "loss": 0.2278, "step": 2332 }, { "epoch": 1.8113354037267082, "grad_norm": 0.699055684894906, "learning_rate": 2.1796147786018518e-07, "loss": 0.2338, "step": 2333 }, { "epoch": 1.812111801242236, "grad_norm": 0.6680650250904743, "learning_rate": 2.161842615084564e-07, "loss": 0.1617, "step": 2334 }, { "epoch": 1.812888198757764, "grad_norm": 0.721632665409549, "learning_rate": 2.1441416027981442e-07, "loss": 0.2207, "step": 2335 }, { "epoch": 1.8136645962732918, "grad_norm": 0.6433118742891467, "learning_rate": 2.1265117680698665e-07, "loss": 0.226, "step": 2336 }, { "epoch": 1.81444099378882, "grad_norm": 0.61398221513298, "learning_rate": 2.1089531371211726e-07, "loss": 0.1867, "step": 2337 }, { "epoch": 1.8152173913043477, "grad_norm": 0.5515455109402342, "learning_rate": 2.0914657360675606e-07, "loss": 0.2175, "step": 2338 }, { "epoch": 1.8159937888198758, "grad_norm": 0.7696139685467325, "learning_rate": 2.0740495909186142e-07, "loss": 0.1842, "step": 2339 }, { "epoch": 1.8167701863354038, "grad_norm": 0.5590219394257142, "learning_rate": 2.0567047275779184e-07, "loss": 0.2282, "step": 2340 }, { "epoch": 1.8175465838509317, "grad_norm": 0.5196350635019215, "learning_rate": 2.0394311718430539e-07, "loss": 0.2871, "step": 2341 }, { "epoch": 1.8183229813664596, "grad_norm": 0.5725502770771436, "learning_rate": 2.0222289494055357e-07, "loss": 0.3157, "step": 2342 }, { "epoch": 1.8190993788819876, "grad_norm": 0.7301712020393325, "learning_rate": 2.0050980858507708e-07, "loss": 0.1302, "step": 2343 }, { "epoch": 1.8198757763975155, "grad_norm": 0.5783442984270823, "learning_rate": 1.988038606658066e-07, "loss": 0.1907, "step": 2344 }, { "epoch": 1.8206521739130435, "grad_norm": 0.6318382277737654, "learning_rate": 1.971050537200514e-07, "loss": 0.2826, "step": 2345 }, { "epoch": 1.8214285714285714, "grad_norm": 1.0022984308292442, "learning_rate": 1.9541339027450256e-07, "loss": 0.2886, "step": 2346 }, { "epoch": 1.8222049689440993, "grad_norm": 0.6412544414180604, "learning_rate": 1.937288728452258e-07, "loss": 0.201, "step": 2347 }, { "epoch": 1.8229813664596275, "grad_norm": 0.6707586877420321, "learning_rate": 1.9205150393765692e-07, "loss": 0.202, "step": 2348 }, { "epoch": 1.8237577639751552, "grad_norm": 0.8548144216408746, "learning_rate": 1.90381286046602e-07, "loss": 0.185, "step": 2349 }, { "epoch": 1.8245341614906834, "grad_norm": 0.5939348131754414, "learning_rate": 1.887182216562289e-07, "loss": 0.2117, "step": 2350 }, { "epoch": 1.825310559006211, "grad_norm": 0.7095672526549811, "learning_rate": 1.8706231324006786e-07, "loss": 0.189, "step": 2351 }, { "epoch": 1.8260869565217392, "grad_norm": 0.6608616118261222, "learning_rate": 1.8541356326100436e-07, "loss": 0.2476, "step": 2352 }, { "epoch": 1.826863354037267, "grad_norm": 0.5623707509486817, "learning_rate": 1.8377197417127512e-07, "loss": 0.1711, "step": 2353 }, { "epoch": 1.8276397515527951, "grad_norm": 0.6248706328184572, "learning_rate": 1.821375484124721e-07, "loss": 0.2072, "step": 2354 }, { "epoch": 1.8284161490683228, "grad_norm": 0.5646956707319339, "learning_rate": 1.8051028841552798e-07, "loss": 0.171, "step": 2355 }, { "epoch": 1.829192546583851, "grad_norm": 0.6008008222068656, "learning_rate": 1.7889019660071728e-07, "loss": 0.1265, "step": 2356 }, { "epoch": 1.829968944099379, "grad_norm": 0.618877225396034, "learning_rate": 1.7727727537765694e-07, "loss": 0.1824, "step": 2357 }, { "epoch": 1.8307453416149069, "grad_norm": 0.7514405789307902, "learning_rate": 1.7567152714529468e-07, "loss": 0.1678, "step": 2358 }, { "epoch": 1.8315217391304348, "grad_norm": 0.6694830566098754, "learning_rate": 1.740729542919134e-07, "loss": 0.1624, "step": 2359 }, { "epoch": 1.8322981366459627, "grad_norm": 0.5839867608738196, "learning_rate": 1.7248155919512067e-07, "loss": 0.2368, "step": 2360 }, { "epoch": 1.8330745341614907, "grad_norm": 0.5357005370009952, "learning_rate": 1.7089734422184923e-07, "loss": 0.1084, "step": 2361 }, { "epoch": 1.8338509316770186, "grad_norm": 0.6070607211075664, "learning_rate": 1.693203117283543e-07, "loss": 0.2145, "step": 2362 }, { "epoch": 1.8346273291925466, "grad_norm": 0.6090348782448376, "learning_rate": 1.6775046406020622e-07, "loss": 0.2254, "step": 2363 }, { "epoch": 1.8354037267080745, "grad_norm": 0.8107167060438266, "learning_rate": 1.6618780355229069e-07, "loss": 0.2616, "step": 2364 }, { "epoch": 1.8361801242236024, "grad_norm": 0.5818224488831619, "learning_rate": 1.6463233252880185e-07, "loss": 0.2163, "step": 2365 }, { "epoch": 1.8369565217391304, "grad_norm": 0.6493284238802738, "learning_rate": 1.6308405330324294e-07, "loss": 0.1842, "step": 2366 }, { "epoch": 1.8377329192546585, "grad_norm": 0.5850915192728534, "learning_rate": 1.6154296817841865e-07, "loss": 0.2442, "step": 2367 }, { "epoch": 1.8385093167701863, "grad_norm": 0.8104735575242498, "learning_rate": 1.6000907944643428e-07, "loss": 0.7196, "step": 2368 }, { "epoch": 1.8392857142857144, "grad_norm": 0.5790328564422283, "learning_rate": 1.5848238938869332e-07, "loss": 0.2602, "step": 2369 }, { "epoch": 1.8400621118012421, "grad_norm": 0.5791426370271611, "learning_rate": 1.5696290027588933e-07, "loss": 0.2255, "step": 2370 }, { "epoch": 1.8408385093167703, "grad_norm": 0.6820418036247005, "learning_rate": 1.554506143680079e-07, "loss": 0.1926, "step": 2371 }, { "epoch": 1.841614906832298, "grad_norm": 0.6198558534547038, "learning_rate": 1.5394553391432142e-07, "loss": 0.233, "step": 2372 }, { "epoch": 1.8423913043478262, "grad_norm": 0.628174365470761, "learning_rate": 1.5244766115338372e-07, "loss": 0.1363, "step": 2373 }, { "epoch": 1.8431677018633539, "grad_norm": 1.018858179615367, "learning_rate": 1.5095699831302935e-07, "loss": 0.1951, "step": 2374 }, { "epoch": 1.843944099378882, "grad_norm": 0.7048767456163497, "learning_rate": 1.4947354761036925e-07, "loss": 0.1874, "step": 2375 }, { "epoch": 1.84472049689441, "grad_norm": 0.6365732016906298, "learning_rate": 1.479973112517863e-07, "loss": 0.1478, "step": 2376 }, { "epoch": 1.845496894409938, "grad_norm": 0.7979525122360168, "learning_rate": 1.465282914329358e-07, "loss": 0.3325, "step": 2377 }, { "epoch": 1.8462732919254659, "grad_norm": 0.6308765165275677, "learning_rate": 1.450664903387372e-07, "loss": 0.1858, "step": 2378 }, { "epoch": 1.8470496894409938, "grad_norm": 0.6653049738729776, "learning_rate": 1.4361191014337528e-07, "loss": 0.1742, "step": 2379 }, { "epoch": 1.8478260869565217, "grad_norm": 0.7339746479001282, "learning_rate": 1.4216455301029274e-07, "loss": 0.7757, "step": 2380 }, { "epoch": 1.8486024844720497, "grad_norm": 0.7158459623228457, "learning_rate": 1.4072442109219154e-07, "loss": 0.1756, "step": 2381 }, { "epoch": 1.8493788819875776, "grad_norm": 0.5860165470937984, "learning_rate": 1.3929151653102546e-07, "loss": 0.164, "step": 2382 }, { "epoch": 1.8501552795031055, "grad_norm": 0.6067598026620309, "learning_rate": 1.3786584145799974e-07, "loss": 0.1897, "step": 2383 }, { "epoch": 1.8509316770186337, "grad_norm": 0.5198508492860111, "learning_rate": 1.3644739799356765e-07, "loss": 0.24, "step": 2384 }, { "epoch": 1.8517080745341614, "grad_norm": 1.1622993230544258, "learning_rate": 1.3503618824742382e-07, "loss": 0.2, "step": 2385 }, { "epoch": 1.8524844720496896, "grad_norm": 0.7144059336069338, "learning_rate": 1.3363221431850816e-07, "loss": 0.1893, "step": 2386 }, { "epoch": 1.8532608695652173, "grad_norm": 0.5954071023561154, "learning_rate": 1.3223547829499527e-07, "loss": 0.1846, "step": 2387 }, { "epoch": 1.8540372670807455, "grad_norm": 0.4959351172404745, "learning_rate": 1.308459822542951e-07, "loss": 0.1569, "step": 2388 }, { "epoch": 1.8548136645962732, "grad_norm": 0.6648875949174101, "learning_rate": 1.294637282630501e-07, "loss": 0.2424, "step": 2389 }, { "epoch": 1.8555900621118013, "grad_norm": 0.7540795885881229, "learning_rate": 1.2808871837713177e-07, "loss": 0.202, "step": 2390 }, { "epoch": 1.856366459627329, "grad_norm": 0.6429047644286038, "learning_rate": 1.267209546416359e-07, "loss": 0.2466, "step": 2391 }, { "epoch": 1.8571428571428572, "grad_norm": 0.6917441482270147, "learning_rate": 1.253604390908819e-07, "loss": 0.2388, "step": 2392 }, { "epoch": 1.8579192546583851, "grad_norm": 0.637902267930199, "learning_rate": 1.2400717374840776e-07, "loss": 0.1246, "step": 2393 }, { "epoch": 1.858695652173913, "grad_norm": 0.6037619673197702, "learning_rate": 1.2266116062696954e-07, "loss": 0.2275, "step": 2394 }, { "epoch": 1.859472049689441, "grad_norm": 0.7714643918818767, "learning_rate": 1.2132240172853527e-07, "loss": 0.214, "step": 2395 }, { "epoch": 1.860248447204969, "grad_norm": 0.6139404773844165, "learning_rate": 1.1999089904428385e-07, "loss": 0.1591, "step": 2396 }, { "epoch": 1.861024844720497, "grad_norm": 0.6360246198248629, "learning_rate": 1.186666545546028e-07, "loss": 0.1282, "step": 2397 }, { "epoch": 1.8618012422360248, "grad_norm": 0.6516611991232458, "learning_rate": 1.1734967022908217e-07, "loss": 0.2311, "step": 2398 }, { "epoch": 1.8625776397515528, "grad_norm": 0.707884005958673, "learning_rate": 1.1603994802651675e-07, "loss": 0.2391, "step": 2399 }, { "epoch": 1.8633540372670807, "grad_norm": 0.6804890021925889, "learning_rate": 1.1473748989489775e-07, "loss": 0.1822, "step": 2400 }, { "epoch": 1.8641304347826086, "grad_norm": 0.5902732694484942, "learning_rate": 1.1344229777141336e-07, "loss": 0.2458, "step": 2401 }, { "epoch": 1.8649068322981366, "grad_norm": 0.6177758334267591, "learning_rate": 1.121543735824443e-07, "loss": 0.2105, "step": 2402 }, { "epoch": 1.8656832298136647, "grad_norm": 0.7674310364013758, "learning_rate": 1.1087371924356105e-07, "loss": 0.2123, "step": 2403 }, { "epoch": 1.8664596273291925, "grad_norm": 0.6356945613770523, "learning_rate": 1.096003366595233e-07, "loss": 0.209, "step": 2404 }, { "epoch": 1.8672360248447206, "grad_norm": 0.6783191721217887, "learning_rate": 1.0833422772427271e-07, "loss": 0.2645, "step": 2405 }, { "epoch": 1.8680124223602483, "grad_norm": 0.6455953667795277, "learning_rate": 1.0707539432093517e-07, "loss": 0.2384, "step": 2406 }, { "epoch": 1.8687888198757765, "grad_norm": 0.6165551043299845, "learning_rate": 1.0582383832181242e-07, "loss": 0.1896, "step": 2407 }, { "epoch": 1.8695652173913042, "grad_norm": 0.4657359714186312, "learning_rate": 1.0457956158838545e-07, "loss": 0.1715, "step": 2408 }, { "epoch": 1.8703416149068324, "grad_norm": 0.6392720899112643, "learning_rate": 1.0334256597130665e-07, "loss": 0.2359, "step": 2409 }, { "epoch": 1.87111801242236, "grad_norm": 0.510793106434991, "learning_rate": 1.0211285331039933e-07, "loss": 0.2402, "step": 2410 }, { "epoch": 1.8718944099378882, "grad_norm": 0.5133377232096518, "learning_rate": 1.0089042543465433e-07, "loss": 0.1826, "step": 2411 }, { "epoch": 1.8726708074534162, "grad_norm": 0.8393231052256136, "learning_rate": 9.967528416222838e-08, "loss": 0.7704, "step": 2412 }, { "epoch": 1.8734472049689441, "grad_norm": 0.6229636872553718, "learning_rate": 9.846743130044024e-08, "loss": 0.16, "step": 2413 }, { "epoch": 1.874223602484472, "grad_norm": 0.7109037227178647, "learning_rate": 9.72668686457684e-08, "loss": 0.1657, "step": 2414 }, { "epoch": 1.875, "grad_norm": 0.6784594934034299, "learning_rate": 9.607359798384785e-08, "loss": 0.1612, "step": 2415 }, { "epoch": 1.875776397515528, "grad_norm": 0.6805609245358737, "learning_rate": 9.488762108946891e-08, "loss": 0.1919, "step": 2416 }, { "epoch": 1.8765527950310559, "grad_norm": 0.7030579472397169, "learning_rate": 9.37089397265728e-08, "loss": 0.2443, "step": 2417 }, { "epoch": 1.8773291925465838, "grad_norm": 0.573771241450167, "learning_rate": 9.25375556482505e-08, "loss": 0.2179, "step": 2418 }, { "epoch": 1.8781055900621118, "grad_norm": 0.6795770589784744, "learning_rate": 9.137347059673951e-08, "loss": 0.2169, "step": 2419 }, { "epoch": 1.87888198757764, "grad_norm": 0.7472966807143107, "learning_rate": 9.02166863034204e-08, "loss": 0.261, "step": 2420 }, { "epoch": 1.8796583850931676, "grad_norm": 0.5087896732314761, "learning_rate": 8.906720448881579e-08, "loss": 0.1959, "step": 2421 }, { "epoch": 1.8804347826086958, "grad_norm": 0.6660348981576361, "learning_rate": 8.792502686258752e-08, "loss": 0.2272, "step": 2422 }, { "epoch": 1.8812111801242235, "grad_norm": 0.6430459341056183, "learning_rate": 8.679015512353228e-08, "loss": 0.2347, "step": 2423 }, { "epoch": 1.8819875776397517, "grad_norm": 0.8155587608273217, "learning_rate": 8.566259095958207e-08, "loss": 0.2321, "step": 2424 }, { "epoch": 1.8827639751552794, "grad_norm": 0.489853008343923, "learning_rate": 8.454233604779927e-08, "loss": 0.1492, "step": 2425 }, { "epoch": 1.8835403726708075, "grad_norm": 0.8599338308286891, "learning_rate": 8.342939205437384e-08, "loss": 0.6831, "step": 2426 }, { "epoch": 1.8843167701863353, "grad_norm": 0.6664929989176589, "learning_rate": 8.232376063462555e-08, "loss": 0.2382, "step": 2427 }, { "epoch": 1.8850931677018634, "grad_norm": 0.6089550636391452, "learning_rate": 8.122544343299343e-08, "loss": 0.2188, "step": 2428 }, { "epoch": 1.8858695652173914, "grad_norm": 0.5609982258343735, "learning_rate": 8.013444208304133e-08, "loss": 0.1724, "step": 2429 }, { "epoch": 1.8866459627329193, "grad_norm": 0.6608835926580511, "learning_rate": 7.905075820744956e-08, "loss": 0.2184, "step": 2430 }, { "epoch": 1.8874223602484472, "grad_norm": 0.5175177126515256, "learning_rate": 7.797439341801716e-08, "loss": 0.2239, "step": 2431 }, { "epoch": 1.8881987577639752, "grad_norm": 0.5262921140262644, "learning_rate": 7.690534931565519e-08, "loss": 0.2423, "step": 2432 }, { "epoch": 1.888975155279503, "grad_norm": 0.702773737166945, "learning_rate": 7.584362749038787e-08, "loss": 0.1773, "step": 2433 }, { "epoch": 1.889751552795031, "grad_norm": 0.6630717983225015, "learning_rate": 7.478922952134815e-08, "loss": 0.1863, "step": 2434 }, { "epoch": 1.890527950310559, "grad_norm": 0.6851403319492946, "learning_rate": 7.374215697677544e-08, "loss": 0.2122, "step": 2435 }, { "epoch": 1.891304347826087, "grad_norm": 0.6044881918421585, "learning_rate": 7.270241141401568e-08, "loss": 0.2246, "step": 2436 }, { "epoch": 1.8920807453416149, "grad_norm": 0.6822877057746702, "learning_rate": 7.166999437951516e-08, "loss": 0.6711, "step": 2437 }, { "epoch": 1.8928571428571428, "grad_norm": 0.7347932362600578, "learning_rate": 7.064490740882057e-08, "loss": 0.1248, "step": 2438 }, { "epoch": 1.893633540372671, "grad_norm": 0.839042772890988, "learning_rate": 6.962715202657788e-08, "loss": 0.2038, "step": 2439 }, { "epoch": 1.8944099378881987, "grad_norm": 0.7958718971308407, "learning_rate": 6.861672974652622e-08, "loss": 0.1889, "step": 2440 }, { "epoch": 1.8951863354037268, "grad_norm": 0.6060911871256407, "learning_rate": 6.761364207149956e-08, "loss": 0.1631, "step": 2441 }, { "epoch": 1.8959627329192545, "grad_norm": 0.7586257627271918, "learning_rate": 6.661789049342282e-08, "loss": 0.2237, "step": 2442 }, { "epoch": 1.8967391304347827, "grad_norm": 0.6321382872435289, "learning_rate": 6.562947649330853e-08, "loss": 0.1902, "step": 2443 }, { "epoch": 1.8975155279503104, "grad_norm": 0.6858604134250831, "learning_rate": 6.464840154125795e-08, "loss": 0.2092, "step": 2444 }, { "epoch": 1.8982919254658386, "grad_norm": 0.7701450177504258, "learning_rate": 6.367466709645498e-08, "loss": 0.8117, "step": 2445 }, { "epoch": 1.8990683229813663, "grad_norm": 0.6973878490101212, "learning_rate": 6.270827460716499e-08, "loss": 0.1767, "step": 2446 }, { "epoch": 1.8998447204968945, "grad_norm": 0.5775729476270149, "learning_rate": 6.174922551073658e-08, "loss": 0.1616, "step": 2447 }, { "epoch": 1.9006211180124224, "grad_norm": 0.6637586968615783, "learning_rate": 6.079752123359261e-08, "loss": 0.1689, "step": 2448 }, { "epoch": 1.9013975155279503, "grad_norm": 0.715922653228285, "learning_rate": 5.985316319123413e-08, "loss": 0.2977, "step": 2449 }, { "epoch": 1.9021739130434783, "grad_norm": 0.7053835440529207, "learning_rate": 5.891615278823537e-08, "loss": 0.1757, "step": 2450 }, { "epoch": 1.9029503105590062, "grad_norm": 0.7908951874441863, "learning_rate": 5.798649141824264e-08, "loss": 0.2056, "step": 2451 }, { "epoch": 1.9037267080745341, "grad_norm": 0.7074681201546265, "learning_rate": 5.7064180463969886e-08, "loss": 0.2192, "step": 2452 }, { "epoch": 1.904503105590062, "grad_norm": 0.6695479164264474, "learning_rate": 5.614922129720091e-08, "loss": 0.1976, "step": 2453 }, { "epoch": 1.90527950310559, "grad_norm": 0.6460839604640133, "learning_rate": 5.5241615278783245e-08, "loss": 0.1737, "step": 2454 }, { "epoch": 1.906055900621118, "grad_norm": 0.6534124287585097, "learning_rate": 5.434136375862875e-08, "loss": 0.2057, "step": 2455 }, { "epoch": 1.9068322981366461, "grad_norm": 0.5523134072327986, "learning_rate": 5.34484680757108e-08, "loss": 0.2531, "step": 2456 }, { "epoch": 1.9076086956521738, "grad_norm": 0.6296118913810554, "learning_rate": 5.256292955806208e-08, "loss": 0.1786, "step": 2457 }, { "epoch": 1.908385093167702, "grad_norm": 0.49290171864687005, "learning_rate": 5.168474952277125e-08, "loss": 0.2084, "step": 2458 }, { "epoch": 1.9091614906832297, "grad_norm": 0.651461881625399, "learning_rate": 5.081392927598516e-08, "loss": 0.1731, "step": 2459 }, { "epoch": 1.9099378881987579, "grad_norm": 0.4412803540261127, "learning_rate": 4.995047011290166e-08, "loss": 0.1603, "step": 2460 }, { "epoch": 1.9107142857142856, "grad_norm": 0.7604379082474207, "learning_rate": 4.909437331777178e-08, "loss": 0.1191, "step": 2461 }, { "epoch": 1.9114906832298137, "grad_norm": 0.8592694945370408, "learning_rate": 4.8245640163896433e-08, "loss": 0.2726, "step": 2462 }, { "epoch": 1.9122670807453415, "grad_norm": 0.5815693532200321, "learning_rate": 4.740427191362307e-08, "loss": 0.248, "step": 2463 }, { "epoch": 1.9130434782608696, "grad_norm": 0.8395681848294368, "learning_rate": 4.657026981834623e-08, "loss": 0.2393, "step": 2464 }, { "epoch": 1.9138198757763976, "grad_norm": 0.4536241777172203, "learning_rate": 4.574363511850366e-08, "loss": 0.1626, "step": 2465 }, { "epoch": 1.9145962732919255, "grad_norm": 0.758191664390571, "learning_rate": 4.4924369043575776e-08, "loss": 0.2335, "step": 2466 }, { "epoch": 1.9153726708074534, "grad_norm": 0.620858346338644, "learning_rate": 4.411247281208342e-08, "loss": 0.2123, "step": 2467 }, { "epoch": 1.9161490683229814, "grad_norm": 0.8191607348457283, "learning_rate": 4.330794763158619e-08, "loss": 0.2169, "step": 2468 }, { "epoch": 1.9169254658385093, "grad_norm": 0.5906991611927213, "learning_rate": 4.2510794698680245e-08, "loss": 0.1564, "step": 2469 }, { "epoch": 1.9177018633540373, "grad_norm": 0.6493122264766409, "learning_rate": 4.172101519899607e-08, "loss": 0.2059, "step": 2470 }, { "epoch": 1.9184782608695652, "grad_norm": 0.669319050475821, "learning_rate": 4.093861030719903e-08, "loss": 0.1837, "step": 2471 }, { "epoch": 1.9192546583850931, "grad_norm": 0.7533838177314006, "learning_rate": 4.0163581186984935e-08, "loss": 0.1702, "step": 2472 }, { "epoch": 1.920031055900621, "grad_norm": 0.547927202555868, "learning_rate": 3.939592899107947e-08, "loss": 0.2274, "step": 2473 }, { "epoch": 1.920807453416149, "grad_norm": 0.717973126628545, "learning_rate": 3.8635654861235994e-08, "loss": 0.2776, "step": 2474 }, { "epoch": 1.9215838509316772, "grad_norm": 0.686382087208196, "learning_rate": 3.7882759928235536e-08, "loss": 0.1495, "step": 2475 }, { "epoch": 1.9223602484472049, "grad_norm": 0.6139170737116706, "learning_rate": 3.713724531188234e-08, "loss": 0.1753, "step": 2476 }, { "epoch": 1.923136645962733, "grad_norm": 0.648580559663829, "learning_rate": 3.6399112121004443e-08, "loss": 0.2103, "step": 2477 }, { "epoch": 1.9239130434782608, "grad_norm": 0.5110674483272314, "learning_rate": 3.566836145345032e-08, "loss": 0.209, "step": 2478 }, { "epoch": 1.924689440993789, "grad_norm": 0.6548515434409434, "learning_rate": 3.494499439609056e-08, "loss": 0.1646, "step": 2479 }, { "epoch": 1.9254658385093166, "grad_norm": 0.764880142258276, "learning_rate": 3.4229012024811215e-08, "loss": 0.2561, "step": 2480 }, { "epoch": 1.9262422360248448, "grad_norm": 0.5541491661660127, "learning_rate": 3.3520415404516004e-08, "loss": 0.2229, "step": 2481 }, { "epoch": 1.9270186335403725, "grad_norm": 0.6611822817503102, "learning_rate": 3.28192055891241e-08, "loss": 0.2061, "step": 2482 }, { "epoch": 1.9277950310559007, "grad_norm": 0.5362300535518394, "learning_rate": 3.2125383621566254e-08, "loss": 0.1535, "step": 2483 }, { "epoch": 1.9285714285714286, "grad_norm": 0.5912517069614756, "learning_rate": 3.143895053378698e-08, "loss": 0.2464, "step": 2484 }, { "epoch": 1.9293478260869565, "grad_norm": 0.6819527711840282, "learning_rate": 3.075990734674017e-08, "loss": 0.2564, "step": 2485 }, { "epoch": 1.9301242236024845, "grad_norm": 0.7547025911246026, "learning_rate": 3.0088255070389037e-08, "loss": 0.2237, "step": 2486 }, { "epoch": 1.9309006211180124, "grad_norm": 0.5401321027253259, "learning_rate": 2.942399470370283e-08, "loss": 0.1853, "step": 2487 }, { "epoch": 1.9316770186335404, "grad_norm": 0.6351311234790262, "learning_rate": 2.876712723465791e-08, "loss": 0.1457, "step": 2488 }, { "epoch": 1.9324534161490683, "grad_norm": 0.6664685251477799, "learning_rate": 2.8117653640234445e-08, "loss": 0.219, "step": 2489 }, { "epoch": 1.9332298136645962, "grad_norm": 0.7101099460751318, "learning_rate": 2.7475574886414726e-08, "loss": 0.2035, "step": 2490 }, { "epoch": 1.9340062111801242, "grad_norm": 0.5116327429316853, "learning_rate": 2.6840891928183733e-08, "loss": 0.2429, "step": 2491 }, { "epoch": 1.9347826086956523, "grad_norm": 0.5547294345901339, "learning_rate": 2.6213605709525803e-08, "loss": 0.1959, "step": 2492 }, { "epoch": 1.93555900621118, "grad_norm": 0.6516990038667757, "learning_rate": 2.559371716342407e-08, "loss": 0.2007, "step": 2493 }, { "epoch": 1.9363354037267082, "grad_norm": 0.8524541272868353, "learning_rate": 2.498122721185825e-08, "loss": 0.7735, "step": 2494 }, { "epoch": 1.937111801242236, "grad_norm": 0.7239651500697576, "learning_rate": 2.4376136765804638e-08, "loss": 0.2364, "step": 2495 }, { "epoch": 1.937888198757764, "grad_norm": 0.7955879839380682, "learning_rate": 2.3778446725233328e-08, "loss": 0.2249, "step": 2496 }, { "epoch": 1.9386645962732918, "grad_norm": 0.8526543619517943, "learning_rate": 2.3188157979109338e-08, "loss": 0.1333, "step": 2497 }, { "epoch": 1.93944099378882, "grad_norm": 0.6185927053772423, "learning_rate": 2.2605271405385932e-08, "loss": 0.2259, "step": 2498 }, { "epoch": 1.9402173913043477, "grad_norm": 0.5769772239849578, "learning_rate": 2.2029787871010734e-08, "loss": 0.1581, "step": 2499 }, { "epoch": 1.9409937888198758, "grad_norm": 0.6814720701392353, "learning_rate": 2.1461708231919064e-08, "loss": 0.2399, "step": 2500 }, { "epoch": 1.9409937888198758, "eval_loss": 0.3133056163787842, "eval_runtime": 3.9781, "eval_samples_per_second": 13.323, "eval_steps_per_second": 3.519, "step": 2500 }, { "epoch": 1.9417701863354038, "grad_norm": 0.6226159965810021, "learning_rate": 2.0901033333032282e-08, "loss": 0.2077, "step": 2501 }, { "epoch": 1.9425465838509317, "grad_norm": 0.6364319875095231, "learning_rate": 2.0347764008262216e-08, "loss": 0.2165, "step": 2502 }, { "epoch": 1.9433229813664596, "grad_norm": 0.5758131093449095, "learning_rate": 1.9801901080502838e-08, "loss": 0.1867, "step": 2503 }, { "epoch": 1.9440993788819876, "grad_norm": 0.6590493288800638, "learning_rate": 1.926344536163416e-08, "loss": 0.2068, "step": 2504 }, { "epoch": 1.9448757763975155, "grad_norm": 0.6273280261159312, "learning_rate": 1.8732397652518332e-08, "loss": 0.2684, "step": 2505 }, { "epoch": 1.9456521739130435, "grad_norm": 0.5396710256960094, "learning_rate": 1.820875874300021e-08, "loss": 0.2128, "step": 2506 }, { "epoch": 1.9464285714285714, "grad_norm": 0.5833428834755244, "learning_rate": 1.769252941190458e-08, "loss": 0.2631, "step": 2507 }, { "epoch": 1.9472049689440993, "grad_norm": 0.7595803546005336, "learning_rate": 1.7183710427035595e-08, "loss": 0.702, "step": 2508 }, { "epoch": 1.9479813664596275, "grad_norm": 0.6194383634691917, "learning_rate": 1.6682302545176776e-08, "loss": 0.2645, "step": 2509 }, { "epoch": 1.9487577639751552, "grad_norm": 0.8299123480180727, "learning_rate": 1.6188306512087692e-08, "loss": 0.1213, "step": 2510 }, { "epoch": 1.9495341614906834, "grad_norm": 0.46296627926604345, "learning_rate": 1.5701723062504505e-08, "loss": 0.1613, "step": 2511 }, { "epoch": 1.950310559006211, "grad_norm": 0.6751933431583884, "learning_rate": 1.5222552920138855e-08, "loss": 0.1768, "step": 2512 }, { "epoch": 1.9510869565217392, "grad_norm": 0.5753746334357088, "learning_rate": 1.475079679767566e-08, "loss": 0.3129, "step": 2513 }, { "epoch": 1.951863354037267, "grad_norm": 0.5933392681011879, "learning_rate": 1.4286455396773092e-08, "loss": 0.1916, "step": 2514 }, { "epoch": 1.9526397515527951, "grad_norm": 0.4819696614536231, "learning_rate": 1.3829529408060927e-08, "loss": 0.255, "step": 2515 }, { "epoch": 1.9534161490683228, "grad_norm": 0.6082557062880922, "learning_rate": 1.338001951113943e-08, "loss": 0.2504, "step": 2516 }, { "epoch": 1.954192546583851, "grad_norm": 0.5640071579783711, "learning_rate": 1.2937926374580468e-08, "loss": 0.2346, "step": 2517 }, { "epoch": 1.954968944099379, "grad_norm": 0.5543406941315527, "learning_rate": 1.250325065592195e-08, "loss": 0.166, "step": 2518 }, { "epoch": 1.9557453416149069, "grad_norm": 0.5020628335545096, "learning_rate": 1.2075993001671727e-08, "loss": 0.2224, "step": 2519 }, { "epoch": 1.9565217391304348, "grad_norm": 0.6060081970248722, "learning_rate": 1.1656154047303691e-08, "loss": 0.2333, "step": 2520 }, { "epoch": 1.9572981366459627, "grad_norm": 0.7240307273523194, "learning_rate": 1.1243734417258345e-08, "loss": 0.1944, "step": 2521 }, { "epoch": 1.9580745341614907, "grad_norm": 0.5143661654845493, "learning_rate": 1.083873472494057e-08, "loss": 0.2541, "step": 2522 }, { "epoch": 1.9588509316770186, "grad_norm": 0.7808982722119052, "learning_rate": 1.044115557271852e-08, "loss": 0.2536, "step": 2523 }, { "epoch": 1.9596273291925466, "grad_norm": 0.6615225479240548, "learning_rate": 1.005099755192529e-08, "loss": 0.2569, "step": 2524 }, { "epoch": 1.9604037267080745, "grad_norm": 0.6076670694889196, "learning_rate": 9.668261242855026e-09, "loss": 0.214, "step": 2525 }, { "epoch": 1.9611801242236024, "grad_norm": 0.6505037676279427, "learning_rate": 9.292947214764036e-09, "loss": 0.2348, "step": 2526 }, { "epoch": 1.9619565217391304, "grad_norm": 0.5846447981997565, "learning_rate": 8.925056025869128e-09, "loss": 0.224, "step": 2527 }, { "epoch": 1.9627329192546585, "grad_norm": 0.8450717967531856, "learning_rate": 8.564588223346493e-09, "loss": 0.167, "step": 2528 }, { "epoch": 1.9635093167701863, "grad_norm": 0.8510535211087851, "learning_rate": 8.211544343331156e-09, "loss": 0.6971, "step": 2529 }, { "epoch": 1.9642857142857144, "grad_norm": 0.7152226215786991, "learning_rate": 7.865924910916977e-09, "loss": 0.1595, "step": 2530 }, { "epoch": 1.9650621118012421, "grad_norm": 0.5919021500938425, "learning_rate": 7.527730440155534e-09, "loss": 0.1627, "step": 2531 }, { "epoch": 1.9658385093167703, "grad_norm": 0.686205962982409, "learning_rate": 7.196961434052796e-09, "loss": 0.1581, "step": 2532 }, { "epoch": 1.966614906832298, "grad_norm": 0.6460508254133783, "learning_rate": 6.873618384573566e-09, "loss": 0.695, "step": 2533 }, { "epoch": 1.9673913043478262, "grad_norm": 0.7248582051483685, "learning_rate": 6.557701772635372e-09, "loss": 0.2001, "step": 2534 }, { "epoch": 1.9681677018633539, "grad_norm": 0.5246847071209239, "learning_rate": 6.249212068111799e-09, "loss": 0.1905, "step": 2535 }, { "epoch": 1.968944099378882, "grad_norm": 0.6254722922801192, "learning_rate": 5.9481497298291556e-09, "loss": 0.1726, "step": 2536 }, { "epoch": 1.96972049689441, "grad_norm": 0.6644454615171993, "learning_rate": 5.65451520556759e-09, "loss": 0.2009, "step": 2537 }, { "epoch": 1.970496894409938, "grad_norm": 0.51998482299615, "learning_rate": 5.368308932058863e-09, "loss": 0.1607, "step": 2538 }, { "epoch": 1.9712732919254659, "grad_norm": 0.5981344038306725, "learning_rate": 5.0895313349863525e-09, "loss": 0.2574, "step": 2539 }, { "epoch": 1.9720496894409938, "grad_norm": 0.6548258845235168, "learning_rate": 4.818182828985607e-09, "loss": 0.1785, "step": 2540 }, { "epoch": 1.9728260869565217, "grad_norm": 0.760998836104482, "learning_rate": 4.554263817642124e-09, "loss": 0.2093, "step": 2541 }, { "epoch": 1.9736024844720497, "grad_norm": 1.758426577954001, "learning_rate": 4.2977746934907975e-09, "loss": 0.5592, "step": 2542 }, { "epoch": 1.9743788819875776, "grad_norm": 0.594312650527668, "learning_rate": 4.04871583801536e-09, "loss": 0.1288, "step": 2543 }, { "epoch": 1.9751552795031055, "grad_norm": 0.4711051284327973, "learning_rate": 3.807087621650607e-09, "loss": 0.1708, "step": 2544 }, { "epoch": 1.9759316770186337, "grad_norm": 0.47390060680958934, "learning_rate": 3.5728904037768407e-09, "loss": 0.1616, "step": 2545 }, { "epoch": 1.9767080745341614, "grad_norm": 0.8300246453393274, "learning_rate": 3.3461245327237603e-09, "loss": 0.2197, "step": 2546 }, { "epoch": 1.9774844720496896, "grad_norm": 0.5271648084569576, "learning_rate": 3.1267903457671276e-09, "loss": 0.24, "step": 2547 }, { "epoch": 1.9782608695652173, "grad_norm": 0.5755239951105046, "learning_rate": 2.9148881691298812e-09, "loss": 0.1973, "step": 2548 }, { "epoch": 1.9790372670807455, "grad_norm": 0.5859550168011981, "learning_rate": 2.7104183179810227e-09, "loss": 0.1531, "step": 2549 }, { "epoch": 1.9798136645962732, "grad_norm": 0.7650908518037307, "learning_rate": 2.5133810964350635e-09, "loss": 0.1953, "step": 2550 }, { "epoch": 1.9805900621118013, "grad_norm": 0.7494903601504802, "learning_rate": 2.323776797551469e-09, "loss": 0.1596, "step": 2551 }, { "epoch": 1.981366459627329, "grad_norm": 0.5932984119296748, "learning_rate": 2.1416057033352144e-09, "loss": 0.1864, "step": 2552 }, { "epoch": 1.9821428571428572, "grad_norm": 0.7019262802112171, "learning_rate": 1.9668680847356735e-09, "loss": 0.2686, "step": 2553 }, { "epoch": 1.9829192546583851, "grad_norm": 0.5251876130175284, "learning_rate": 1.7995642016449544e-09, "loss": 0.1656, "step": 2554 }, { "epoch": 1.983695652173913, "grad_norm": 0.5705282113865494, "learning_rate": 1.6396943028995638e-09, "loss": 0.2626, "step": 2555 }, { "epoch": 1.984472049689441, "grad_norm": 0.7283599388891592, "learning_rate": 1.4872586262792975e-09, "loss": 0.2441, "step": 2556 }, { "epoch": 1.985248447204969, "grad_norm": 0.7877292617955438, "learning_rate": 1.3422573985072408e-09, "loss": 0.1436, "step": 2557 }, { "epoch": 1.986024844720497, "grad_norm": 0.6098652782302874, "learning_rate": 1.2046908352475462e-09, "loss": 0.2125, "step": 2558 }, { "epoch": 1.9868012422360248, "grad_norm": 0.6450065505055326, "learning_rate": 1.0745591411076562e-09, "loss": 0.2679, "step": 2559 }, { "epoch": 1.9875776397515528, "grad_norm": 0.6169417695049447, "learning_rate": 9.518625096366364e-10, "loss": 0.2397, "step": 2560 }, { "epoch": 1.9883540372670807, "grad_norm": 0.6728377968965827, "learning_rate": 8.366011233251758e-10, "loss": 0.1627, "step": 2561 }, { "epoch": 1.9891304347826086, "grad_norm": 0.6194127785647411, "learning_rate": 7.287751536050324e-10, "loss": 0.2468, "step": 2562 }, { "epoch": 1.9899068322981366, "grad_norm": 0.7520997597464877, "learning_rate": 6.283847608495874e-10, "loss": 0.2882, "step": 2563 }, { "epoch": 1.9906832298136647, "grad_norm": 0.6379474927038744, "learning_rate": 5.354300943727353e-10, "loss": 0.2218, "step": 2564 }, { "epoch": 1.9914596273291925, "grad_norm": 0.5624422358486367, "learning_rate": 4.4991129242832885e-10, "loss": 0.1594, "step": 2565 }, { "epoch": 1.9922360248447206, "grad_norm": 0.5922644414484881, "learning_rate": 3.718284822118445e-10, "loss": 0.2497, "step": 2566 }, { "epoch": 1.9930124223602483, "grad_norm": 0.8586011374224444, "learning_rate": 3.011817798581618e-10, "loss": 0.1844, "step": 2567 }, { "epoch": 1.9937888198757765, "grad_norm": 0.7301410262096385, "learning_rate": 2.379712904426734e-10, "loss": 0.142, "step": 2568 }, { "epoch": 1.9945652173913042, "grad_norm": 0.7405713314500527, "learning_rate": 1.8219710797962032e-10, "loss": 0.2041, "step": 2569 }, { "epoch": 1.9953416149068324, "grad_norm": 0.7058796055887808, "learning_rate": 1.3385931542486686e-10, "loss": 0.4683, "step": 2570 }, { "epoch": 1.99611801242236, "grad_norm": 0.8121225438022246, "learning_rate": 9.295798467146011e-11, "loss": 0.2907, "step": 2571 }, { "epoch": 1.9968944099378882, "grad_norm": 0.7991039180304573, "learning_rate": 5.949317655462583e-11, "loss": 0.2008, "step": 2572 }, { "epoch": 1.9976708074534162, "grad_norm": 0.5565123179355396, "learning_rate": 3.3464940846217317e-11, "loss": 0.1666, "step": 2573 }, { "epoch": 1.9984472049689441, "grad_norm": 0.5125257275434825, "learning_rate": 1.4873316260266558e-11, "loss": 0.1665, "step": 2574 }, { "epoch": 1.999223602484472, "grad_norm": 0.6376007172915024, "learning_rate": 3.718330447433083e-12, "loss": 0.216, "step": 2575 }, { "epoch": 2.0, "grad_norm": 0.6952079112982549, "learning_rate": 0.0, "loss": 0.1995, "step": 2576 }, { "epoch": 2.0, "step": 2576, "total_flos": 31843636273152.0, "train_loss": 0.256138851449634, "train_runtime": 2689.4563, "train_samples_per_second": 3.83, "train_steps_per_second": 0.958 } ], "logging_steps": 1, "max_steps": 2576, "num_input_tokens_seen": 0, "num_train_epochs": 2, "save_steps": 70000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 31843636273152.0, "train_batch_size": 1, "trial_name": null, "trial_params": null }