{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 3.0, "eval_steps": 500, "global_step": 9222, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.00032530904359141186, "grad_norm": 12.347868240518158, "learning_rate": 0.0, "loss": 1.1987, "step": 1 }, { "epoch": 0.0006506180871828237, "grad_norm": 13.0008291383857, "learning_rate": 1.0834236186348863e-08, "loss": 1.2647, "step": 2 }, { "epoch": 0.0009759271307742355, "grad_norm": 14.130450159426484, "learning_rate": 2.1668472372697727e-08, "loss": 1.3393, "step": 3 }, { "epoch": 0.0013012361743656475, "grad_norm": 12.671307148633076, "learning_rate": 3.250270855904659e-08, "loss": 1.3307, "step": 4 }, { "epoch": 0.0016265452179570592, "grad_norm": 12.23336332295595, "learning_rate": 4.3336944745395454e-08, "loss": 1.17, "step": 5 }, { "epoch": 0.001951854261548471, "grad_norm": 17.036967375139287, "learning_rate": 5.4171180931744316e-08, "loss": 1.4122, "step": 6 }, { "epoch": 0.002277163305139883, "grad_norm": 11.95173494743226, "learning_rate": 6.500541711809318e-08, "loss": 1.2114, "step": 7 }, { "epoch": 0.002602472348731295, "grad_norm": 14.32259810509405, "learning_rate": 7.583965330444205e-08, "loss": 1.3589, "step": 8 }, { "epoch": 0.0029277813923227064, "grad_norm": 12.143408672419524, "learning_rate": 8.667388949079091e-08, "loss": 1.214, "step": 9 }, { "epoch": 0.0032530904359141183, "grad_norm": 13.002088536001375, "learning_rate": 9.750812567713977e-08, "loss": 1.1958, "step": 10 }, { "epoch": 0.0035783994795055302, "grad_norm": 13.571939922367806, "learning_rate": 1.0834236186348863e-07, "loss": 1.1854, "step": 11 }, { "epoch": 0.003903708523096942, "grad_norm": 17.021800863865803, "learning_rate": 1.1917659804983749e-07, "loss": 1.4601, "step": 12 }, { "epoch": 0.004229017566688354, "grad_norm": 13.113313304728948, "learning_rate": 1.3001083423618637e-07, "loss": 1.2893, "step": 13 }, { "epoch": 0.004554326610279766, "grad_norm": 11.185257005666685, "learning_rate": 1.4084507042253522e-07, "loss": 1.085, "step": 14 }, { "epoch": 0.004879635653871178, "grad_norm": 15.605034399460783, "learning_rate": 1.516793066088841e-07, "loss": 1.4587, "step": 15 }, { "epoch": 0.00520494469746259, "grad_norm": 9.643417160883487, "learning_rate": 1.6251354279523294e-07, "loss": 1.0561, "step": 16 }, { "epoch": 0.005530253741054001, "grad_norm": 10.054098046986542, "learning_rate": 1.7334777898158182e-07, "loss": 1.1807, "step": 17 }, { "epoch": 0.005855562784645413, "grad_norm": 13.054340231846659, "learning_rate": 1.8418201516793066e-07, "loss": 1.2409, "step": 18 }, { "epoch": 0.006180871828236825, "grad_norm": 12.259238804165165, "learning_rate": 1.9501625135427954e-07, "loss": 1.1754, "step": 19 }, { "epoch": 0.006506180871828237, "grad_norm": 12.008103890394214, "learning_rate": 2.058504875406284e-07, "loss": 1.1086, "step": 20 }, { "epoch": 0.0068314899154196486, "grad_norm": 11.507790105330304, "learning_rate": 2.1668472372697726e-07, "loss": 1.2473, "step": 21 }, { "epoch": 0.0071567989590110605, "grad_norm": 12.433129159655142, "learning_rate": 2.2751895991332614e-07, "loss": 1.3283, "step": 22 }, { "epoch": 0.007482108002602472, "grad_norm": 12.994143445234336, "learning_rate": 2.3835319609967499e-07, "loss": 1.2455, "step": 23 }, { "epoch": 0.007807417046193884, "grad_norm": 9.694768048208534, "learning_rate": 2.4918743228602383e-07, "loss": 1.1298, "step": 24 }, { "epoch": 0.008132726089785295, "grad_norm": 11.4109155296846, "learning_rate": 2.6002166847237274e-07, "loss": 1.2697, "step": 25 }, { "epoch": 0.008458035133376708, "grad_norm": 13.753485758092939, "learning_rate": 2.708559046587216e-07, "loss": 1.3188, "step": 26 }, { "epoch": 0.00878334417696812, "grad_norm": 10.143106438845308, "learning_rate": 2.8169014084507043e-07, "loss": 1.0821, "step": 27 }, { "epoch": 0.009108653220559532, "grad_norm": 11.374688973213996, "learning_rate": 2.9252437703141933e-07, "loss": 1.2706, "step": 28 }, { "epoch": 0.009433962264150943, "grad_norm": 12.304279542743947, "learning_rate": 3.033586132177682e-07, "loss": 1.2467, "step": 29 }, { "epoch": 0.009759271307742356, "grad_norm": 10.635423375280512, "learning_rate": 3.1419284940411703e-07, "loss": 1.1016, "step": 30 }, { "epoch": 0.010084580351333767, "grad_norm": 10.289102026458458, "learning_rate": 3.250270855904659e-07, "loss": 1.1783, "step": 31 }, { "epoch": 0.01040988939492518, "grad_norm": 9.797216120878767, "learning_rate": 3.3586132177681473e-07, "loss": 1.0147, "step": 32 }, { "epoch": 0.01073519843851659, "grad_norm": 8.204501471786864, "learning_rate": 3.4669555796316363e-07, "loss": 1.0385, "step": 33 }, { "epoch": 0.011060507482108002, "grad_norm": 8.946324357485569, "learning_rate": 3.575297941495125e-07, "loss": 1.0317, "step": 34 }, { "epoch": 0.011385816525699415, "grad_norm": 9.66573742392951, "learning_rate": 3.6836403033586133e-07, "loss": 1.0279, "step": 35 }, { "epoch": 0.011711125569290826, "grad_norm": 9.04347938449296, "learning_rate": 3.791982665222102e-07, "loss": 1.1047, "step": 36 }, { "epoch": 0.012036434612882238, "grad_norm": 9.591041376397563, "learning_rate": 3.900325027085591e-07, "loss": 1.149, "step": 37 }, { "epoch": 0.01236174365647365, "grad_norm": 7.001482194520532, "learning_rate": 4.008667388949079e-07, "loss": 0.8643, "step": 38 }, { "epoch": 0.012687052700065062, "grad_norm": 8.049259540154383, "learning_rate": 4.117009750812568e-07, "loss": 0.8865, "step": 39 }, { "epoch": 0.013012361743656473, "grad_norm": 7.4102598530112465, "learning_rate": 4.225352112676057e-07, "loss": 0.9337, "step": 40 }, { "epoch": 0.013337670787247886, "grad_norm": 7.8974420110393275, "learning_rate": 4.333694474539545e-07, "loss": 0.9432, "step": 41 }, { "epoch": 0.013662979830839297, "grad_norm": 6.940060327503368, "learning_rate": 4.4420368364030337e-07, "loss": 0.8757, "step": 42 }, { "epoch": 0.01398828887443071, "grad_norm": 5.950837265557857, "learning_rate": 4.550379198266523e-07, "loss": 0.8656, "step": 43 }, { "epoch": 0.014313597918022121, "grad_norm": 6.365939446877231, "learning_rate": 4.658721560130011e-07, "loss": 0.9847, "step": 44 }, { "epoch": 0.014638906961613532, "grad_norm": 7.323187082840691, "learning_rate": 4.7670639219934997e-07, "loss": 0.9573, "step": 45 }, { "epoch": 0.014964216005204945, "grad_norm": 6.563015556578668, "learning_rate": 4.875406283856989e-07, "loss": 0.9142, "step": 46 }, { "epoch": 0.015289525048796356, "grad_norm": 7.536928569923618, "learning_rate": 4.983748645720477e-07, "loss": 0.9562, "step": 47 }, { "epoch": 0.015614834092387769, "grad_norm": 6.2625857740708994, "learning_rate": 5.092091007583966e-07, "loss": 0.9323, "step": 48 }, { "epoch": 0.01594014313597918, "grad_norm": 7.348070813891148, "learning_rate": 5.200433369447455e-07, "loss": 0.8791, "step": 49 }, { "epoch": 0.01626545217957059, "grad_norm": 4.968638450263368, "learning_rate": 5.308775731310943e-07, "loss": 0.7636, "step": 50 }, { "epoch": 0.016590761223162005, "grad_norm": 7.142476479854788, "learning_rate": 5.417118093174432e-07, "loss": 1.0474, "step": 51 }, { "epoch": 0.016916070266753416, "grad_norm": 6.306699755858201, "learning_rate": 5.525460455037921e-07, "loss": 1.018, "step": 52 }, { "epoch": 0.017241379310344827, "grad_norm": 6.849753502148755, "learning_rate": 5.633802816901409e-07, "loss": 0.9776, "step": 53 }, { "epoch": 0.01756668835393624, "grad_norm": 5.006932844259666, "learning_rate": 5.742145178764898e-07, "loss": 0.909, "step": 54 }, { "epoch": 0.017891997397527653, "grad_norm": 5.097012190638636, "learning_rate": 5.850487540628387e-07, "loss": 0.8549, "step": 55 }, { "epoch": 0.018217306441119064, "grad_norm": 4.941539692482677, "learning_rate": 5.958829902491875e-07, "loss": 0.819, "step": 56 }, { "epoch": 0.018542615484710475, "grad_norm": 4.3744440909529745, "learning_rate": 6.067172264355364e-07, "loss": 0.7341, "step": 57 }, { "epoch": 0.018867924528301886, "grad_norm": 4.109079199138532, "learning_rate": 6.175514626218853e-07, "loss": 0.7145, "step": 58 }, { "epoch": 0.019193233571893297, "grad_norm": 4.416097256828061, "learning_rate": 6.283856988082341e-07, "loss": 0.7452, "step": 59 }, { "epoch": 0.01951854261548471, "grad_norm": 5.672634145836419, "learning_rate": 6.39219934994583e-07, "loss": 0.772, "step": 60 }, { "epoch": 0.019843851659076123, "grad_norm": 5.229571267239312, "learning_rate": 6.500541711809318e-07, "loss": 0.816, "step": 61 }, { "epoch": 0.020169160702667534, "grad_norm": 4.2533020486422215, "learning_rate": 6.608884073672807e-07, "loss": 0.6972, "step": 62 }, { "epoch": 0.020494469746258945, "grad_norm": 3.912304871573295, "learning_rate": 6.717226435536295e-07, "loss": 0.6951, "step": 63 }, { "epoch": 0.02081977878985036, "grad_norm": 3.9756253200568756, "learning_rate": 6.825568797399785e-07, "loss": 0.7544, "step": 64 }, { "epoch": 0.02114508783344177, "grad_norm": 3.9285862578896116, "learning_rate": 6.933911159263273e-07, "loss": 0.7607, "step": 65 }, { "epoch": 0.02147039687703318, "grad_norm": 3.9976581361209207, "learning_rate": 7.042253521126762e-07, "loss": 0.7001, "step": 66 }, { "epoch": 0.021795705920624592, "grad_norm": 4.197491316083219, "learning_rate": 7.15059588299025e-07, "loss": 0.7434, "step": 67 }, { "epoch": 0.022121014964216004, "grad_norm": 4.222536654966533, "learning_rate": 7.258938244853739e-07, "loss": 0.7817, "step": 68 }, { "epoch": 0.022446324007807418, "grad_norm": 3.24580887727528, "learning_rate": 7.367280606717227e-07, "loss": 0.7773, "step": 69 }, { "epoch": 0.02277163305139883, "grad_norm": 3.663970665208711, "learning_rate": 7.475622968580716e-07, "loss": 0.7102, "step": 70 }, { "epoch": 0.02309694209499024, "grad_norm": 3.688411079421714, "learning_rate": 7.583965330444204e-07, "loss": 0.7578, "step": 71 }, { "epoch": 0.02342225113858165, "grad_norm": 3.3125606147082918, "learning_rate": 7.692307692307694e-07, "loss": 0.6542, "step": 72 }, { "epoch": 0.023747560182173066, "grad_norm": 4.19208543017989, "learning_rate": 7.800650054171182e-07, "loss": 0.8974, "step": 73 }, { "epoch": 0.024072869225764477, "grad_norm": 3.401890586919847, "learning_rate": 7.908992416034671e-07, "loss": 0.7021, "step": 74 }, { "epoch": 0.024398178269355888, "grad_norm": 2.9343286538400637, "learning_rate": 8.017334777898159e-07, "loss": 0.68, "step": 75 }, { "epoch": 0.0247234873129473, "grad_norm": 3.6496246418969918, "learning_rate": 8.125677139761648e-07, "loss": 0.7647, "step": 76 }, { "epoch": 0.025048796356538713, "grad_norm": 3.4099513866850573, "learning_rate": 8.234019501625135e-07, "loss": 0.6845, "step": 77 }, { "epoch": 0.025374105400130124, "grad_norm": 2.99873280374095, "learning_rate": 8.342361863488626e-07, "loss": 0.6982, "step": 78 }, { "epoch": 0.025699414443721535, "grad_norm": 3.3748955218841146, "learning_rate": 8.450704225352114e-07, "loss": 0.8748, "step": 79 }, { "epoch": 0.026024723487312947, "grad_norm": 3.1316788795771378, "learning_rate": 8.559046587215603e-07, "loss": 0.7936, "step": 80 }, { "epoch": 0.026350032530904358, "grad_norm": 3.4734857028387265, "learning_rate": 8.66738894907909e-07, "loss": 0.7961, "step": 81 }, { "epoch": 0.026675341574495772, "grad_norm": 3.5254066723981805, "learning_rate": 8.77573131094258e-07, "loss": 0.7543, "step": 82 }, { "epoch": 0.027000650618087183, "grad_norm": 3.0255748936550573, "learning_rate": 8.884073672806067e-07, "loss": 0.7818, "step": 83 }, { "epoch": 0.027325959661678594, "grad_norm": 2.931848373912857, "learning_rate": 8.992416034669558e-07, "loss": 0.7041, "step": 84 }, { "epoch": 0.027651268705270005, "grad_norm": 2.870168474460606, "learning_rate": 9.100758396533045e-07, "loss": 0.6775, "step": 85 }, { "epoch": 0.02797657774886142, "grad_norm": 3.1488802064850265, "learning_rate": 9.209100758396535e-07, "loss": 0.642, "step": 86 }, { "epoch": 0.02830188679245283, "grad_norm": 3.0829770671875285, "learning_rate": 9.317443120260022e-07, "loss": 0.7574, "step": 87 }, { "epoch": 0.028627195836044242, "grad_norm": 3.0857763223788006, "learning_rate": 9.425785482123511e-07, "loss": 0.6858, "step": 88 }, { "epoch": 0.028952504879635653, "grad_norm": 2.9368293787113284, "learning_rate": 9.534127843986999e-07, "loss": 0.7177, "step": 89 }, { "epoch": 0.029277813923227064, "grad_norm": 3.0119276854034163, "learning_rate": 9.642470205850487e-07, "loss": 0.6997, "step": 90 }, { "epoch": 0.02960312296681848, "grad_norm": 3.149583085399912, "learning_rate": 9.750812567713977e-07, "loss": 0.791, "step": 91 }, { "epoch": 0.02992843201040989, "grad_norm": 3.649584696682866, "learning_rate": 9.859154929577465e-07, "loss": 0.7659, "step": 92 }, { "epoch": 0.0302537410540013, "grad_norm": 3.3135340143933214, "learning_rate": 9.967497291440953e-07, "loss": 0.7486, "step": 93 }, { "epoch": 0.03057905009759271, "grad_norm": 3.4750143270474916, "learning_rate": 1.0075839653304441e-06, "loss": 0.854, "step": 94 }, { "epoch": 0.030904359141184126, "grad_norm": 3.0549276940748498, "learning_rate": 1.0184182015167931e-06, "loss": 0.7636, "step": 95 }, { "epoch": 0.031229668184775537, "grad_norm": 3.21491548678472, "learning_rate": 1.029252437703142e-06, "loss": 0.7689, "step": 96 }, { "epoch": 0.03155497722836695, "grad_norm": 2.877064993025877, "learning_rate": 1.040086673889491e-06, "loss": 0.6448, "step": 97 }, { "epoch": 0.03188028627195836, "grad_norm": 3.633887781873793, "learning_rate": 1.0509209100758397e-06, "loss": 0.7887, "step": 98 }, { "epoch": 0.03220559531554977, "grad_norm": 3.2045489908821483, "learning_rate": 1.0617551462621885e-06, "loss": 0.7362, "step": 99 }, { "epoch": 0.03253090435914118, "grad_norm": 3.010871615134529, "learning_rate": 1.0725893824485373e-06, "loss": 0.7538, "step": 100 }, { "epoch": 0.03285621340273259, "grad_norm": 3.2236100901731763, "learning_rate": 1.0834236186348863e-06, "loss": 0.7734, "step": 101 }, { "epoch": 0.03318152244632401, "grad_norm": 3.098897768989396, "learning_rate": 1.0942578548212351e-06, "loss": 0.7599, "step": 102 }, { "epoch": 0.03350683148991542, "grad_norm": 3.841171892122135, "learning_rate": 1.1050920910075841e-06, "loss": 0.8911, "step": 103 }, { "epoch": 0.03383214053350683, "grad_norm": 3.221173299198862, "learning_rate": 1.115926327193933e-06, "loss": 0.726, "step": 104 }, { "epoch": 0.034157449577098244, "grad_norm": 2.69352451999635, "learning_rate": 1.1267605633802817e-06, "loss": 0.6047, "step": 105 }, { "epoch": 0.034482758620689655, "grad_norm": 2.938294026466022, "learning_rate": 1.1375947995666305e-06, "loss": 0.7512, "step": 106 }, { "epoch": 0.034808067664281066, "grad_norm": 3.0499409041561494, "learning_rate": 1.1484290357529795e-06, "loss": 0.6816, "step": 107 }, { "epoch": 0.03513337670787248, "grad_norm": 3.0883735123675655, "learning_rate": 1.1592632719393283e-06, "loss": 0.752, "step": 108 }, { "epoch": 0.03545868575146389, "grad_norm": 2.887065332473537, "learning_rate": 1.1700975081256773e-06, "loss": 0.6411, "step": 109 }, { "epoch": 0.035783994795055306, "grad_norm": 3.3701993578296685, "learning_rate": 1.1809317443120261e-06, "loss": 0.7086, "step": 110 }, { "epoch": 0.03610930383864672, "grad_norm": 2.834233514879352, "learning_rate": 1.191765980498375e-06, "loss": 0.6741, "step": 111 }, { "epoch": 0.03643461288223813, "grad_norm": 2.888688751398159, "learning_rate": 1.2026002166847237e-06, "loss": 0.7255, "step": 112 }, { "epoch": 0.03675992192582954, "grad_norm": 3.088833619391446, "learning_rate": 1.2134344528710727e-06, "loss": 0.8049, "step": 113 }, { "epoch": 0.03708523096942095, "grad_norm": 3.352887348944667, "learning_rate": 1.2242686890574215e-06, "loss": 0.8609, "step": 114 }, { "epoch": 0.03741054001301236, "grad_norm": 2.9652612315293143, "learning_rate": 1.2351029252437705e-06, "loss": 0.7705, "step": 115 }, { "epoch": 0.03773584905660377, "grad_norm": 2.9490167437914154, "learning_rate": 1.2459371614301193e-06, "loss": 0.7622, "step": 116 }, { "epoch": 0.03806115810019518, "grad_norm": 3.1427258709810166, "learning_rate": 1.2567713976164681e-06, "loss": 0.8207, "step": 117 }, { "epoch": 0.038386467143786594, "grad_norm": 2.669593248303349, "learning_rate": 1.267605633802817e-06, "loss": 0.6356, "step": 118 }, { "epoch": 0.03871177618737801, "grad_norm": 3.773369767629891, "learning_rate": 1.278439869989166e-06, "loss": 0.7567, "step": 119 }, { "epoch": 0.03903708523096942, "grad_norm": 2.8614051865820755, "learning_rate": 1.289274106175515e-06, "loss": 0.7611, "step": 120 }, { "epoch": 0.039362394274560834, "grad_norm": 3.016212226193705, "learning_rate": 1.3001083423618635e-06, "loss": 0.5816, "step": 121 }, { "epoch": 0.039687703318152245, "grad_norm": 3.279896379814327, "learning_rate": 1.3109425785482125e-06, "loss": 0.7599, "step": 122 }, { "epoch": 0.040013012361743656, "grad_norm": 3.6504258147793185, "learning_rate": 1.3217768147345613e-06, "loss": 0.843, "step": 123 }, { "epoch": 0.04033832140533507, "grad_norm": 3.207374459902964, "learning_rate": 1.3326110509209103e-06, "loss": 0.7309, "step": 124 }, { "epoch": 0.04066363044892648, "grad_norm": 2.7181468221213914, "learning_rate": 1.343445287107259e-06, "loss": 0.6862, "step": 125 }, { "epoch": 0.04098893949251789, "grad_norm": 3.05741774174071, "learning_rate": 1.354279523293608e-06, "loss": 0.6725, "step": 126 }, { "epoch": 0.0413142485361093, "grad_norm": 2.759916410530077, "learning_rate": 1.365113759479957e-06, "loss": 0.6653, "step": 127 }, { "epoch": 0.04163955757970072, "grad_norm": 2.6350064921276575, "learning_rate": 1.3759479956663057e-06, "loss": 0.7122, "step": 128 }, { "epoch": 0.04196486662329213, "grad_norm": 3.5885858074063317, "learning_rate": 1.3867822318526545e-06, "loss": 0.7322, "step": 129 }, { "epoch": 0.04229017566688354, "grad_norm": 3.110883020681901, "learning_rate": 1.3976164680390033e-06, "loss": 0.6832, "step": 130 }, { "epoch": 0.04261548471047495, "grad_norm": 2.571982995083505, "learning_rate": 1.4084507042253523e-06, "loss": 0.6491, "step": 131 }, { "epoch": 0.04294079375406636, "grad_norm": 3.0120051043950684, "learning_rate": 1.4192849404117013e-06, "loss": 0.8324, "step": 132 }, { "epoch": 0.043266102797657774, "grad_norm": 3.0501735628416795, "learning_rate": 1.43011917659805e-06, "loss": 0.7885, "step": 133 }, { "epoch": 0.043591411841249185, "grad_norm": 3.038874631891749, "learning_rate": 1.440953412784399e-06, "loss": 0.7171, "step": 134 }, { "epoch": 0.043916720884840596, "grad_norm": 2.511274076269136, "learning_rate": 1.4517876489707477e-06, "loss": 0.619, "step": 135 }, { "epoch": 0.04424202992843201, "grad_norm": 2.857688649727413, "learning_rate": 1.4626218851570965e-06, "loss": 0.6939, "step": 136 }, { "epoch": 0.044567338972023425, "grad_norm": 3.091170911204102, "learning_rate": 1.4734561213434453e-06, "loss": 0.6803, "step": 137 }, { "epoch": 0.044892648015614836, "grad_norm": 3.3478206778369963, "learning_rate": 1.4842903575297943e-06, "loss": 0.7956, "step": 138 }, { "epoch": 0.04521795705920625, "grad_norm": 3.3127496404867127, "learning_rate": 1.4951245937161431e-06, "loss": 0.7084, "step": 139 }, { "epoch": 0.04554326610279766, "grad_norm": 3.0908027080061564, "learning_rate": 1.505958829902492e-06, "loss": 0.6525, "step": 140 }, { "epoch": 0.04586857514638907, "grad_norm": 2.8831624297182783, "learning_rate": 1.5167930660888407e-06, "loss": 0.6355, "step": 141 }, { "epoch": 0.04619388418998048, "grad_norm": 3.1140597297917556, "learning_rate": 1.5276273022751897e-06, "loss": 0.7353, "step": 142 }, { "epoch": 0.04651919323357189, "grad_norm": 3.2701954421917523, "learning_rate": 1.5384615384615387e-06, "loss": 0.7456, "step": 143 }, { "epoch": 0.0468445022771633, "grad_norm": 3.2880887769141665, "learning_rate": 1.5492957746478873e-06, "loss": 0.8149, "step": 144 }, { "epoch": 0.04716981132075472, "grad_norm": 2.767942240385027, "learning_rate": 1.5601300108342363e-06, "loss": 0.6131, "step": 145 }, { "epoch": 0.04749512036434613, "grad_norm": 2.9098751316446823, "learning_rate": 1.570964247020585e-06, "loss": 0.7719, "step": 146 }, { "epoch": 0.04782042940793754, "grad_norm": 2.665688586222419, "learning_rate": 1.5817984832069341e-06, "loss": 0.583, "step": 147 }, { "epoch": 0.048145738451528954, "grad_norm": 3.0195353121960746, "learning_rate": 1.5926327193932827e-06, "loss": 0.7328, "step": 148 }, { "epoch": 0.048471047495120365, "grad_norm": 3.2270344384842096, "learning_rate": 1.6034669555796317e-06, "loss": 0.6846, "step": 149 }, { "epoch": 0.048796356538711776, "grad_norm": 3.3514457677635305, "learning_rate": 1.6143011917659807e-06, "loss": 0.7448, "step": 150 }, { "epoch": 0.04912166558230319, "grad_norm": 2.8654250868734352, "learning_rate": 1.6251354279523295e-06, "loss": 0.6631, "step": 151 }, { "epoch": 0.0494469746258946, "grad_norm": 2.6396820227756104, "learning_rate": 1.6359696641386783e-06, "loss": 0.6204, "step": 152 }, { "epoch": 0.04977228366948601, "grad_norm": 3.0810975612178058, "learning_rate": 1.646803900325027e-06, "loss": 0.6357, "step": 153 }, { "epoch": 0.05009759271307743, "grad_norm": 3.0201933207652694, "learning_rate": 1.657638136511376e-06, "loss": 0.723, "step": 154 }, { "epoch": 0.05042290175666884, "grad_norm": 3.2314245426880013, "learning_rate": 1.6684723726977251e-06, "loss": 0.7844, "step": 155 }, { "epoch": 0.05074821080026025, "grad_norm": 2.673242878285273, "learning_rate": 1.6793066088840737e-06, "loss": 0.582, "step": 156 }, { "epoch": 0.05107351984385166, "grad_norm": 3.1247824611152404, "learning_rate": 1.6901408450704227e-06, "loss": 0.6965, "step": 157 }, { "epoch": 0.05139882888744307, "grad_norm": 2.5307837609521098, "learning_rate": 1.7009750812567715e-06, "loss": 0.5803, "step": 158 }, { "epoch": 0.05172413793103448, "grad_norm": 2.6954845577102073, "learning_rate": 1.7118093174431205e-06, "loss": 0.6381, "step": 159 }, { "epoch": 0.05204944697462589, "grad_norm": 2.565872741274244, "learning_rate": 1.722643553629469e-06, "loss": 0.6117, "step": 160 }, { "epoch": 0.052374756018217304, "grad_norm": 3.022412131947003, "learning_rate": 1.733477789815818e-06, "loss": 0.7482, "step": 161 }, { "epoch": 0.052700065061808715, "grad_norm": 2.834545681673406, "learning_rate": 1.744312026002167e-06, "loss": 0.6744, "step": 162 }, { "epoch": 0.05302537410540013, "grad_norm": 2.473821915240064, "learning_rate": 1.755146262188516e-06, "loss": 0.647, "step": 163 }, { "epoch": 0.053350683148991544, "grad_norm": 2.829040447509092, "learning_rate": 1.7659804983748647e-06, "loss": 0.6549, "step": 164 }, { "epoch": 0.053675992192582955, "grad_norm": 3.14851222709044, "learning_rate": 1.7768147345612135e-06, "loss": 0.7358, "step": 165 }, { "epoch": 0.054001301236174366, "grad_norm": 2.8502550232430717, "learning_rate": 1.7876489707475625e-06, "loss": 0.7493, "step": 166 }, { "epoch": 0.05432661027976578, "grad_norm": 3.810110566895711, "learning_rate": 1.7984832069339115e-06, "loss": 0.8368, "step": 167 }, { "epoch": 0.05465191932335719, "grad_norm": 2.7095608948460224, "learning_rate": 1.80931744312026e-06, "loss": 0.6786, "step": 168 }, { "epoch": 0.0549772283669486, "grad_norm": 2.6940229794262898, "learning_rate": 1.820151679306609e-06, "loss": 0.6694, "step": 169 }, { "epoch": 0.05530253741054001, "grad_norm": 2.4705641672327623, "learning_rate": 1.8309859154929579e-06, "loss": 0.6589, "step": 170 }, { "epoch": 0.05562784645413142, "grad_norm": 3.471412410355648, "learning_rate": 1.841820151679307e-06, "loss": 0.9281, "step": 171 }, { "epoch": 0.05595315549772284, "grad_norm": 3.330257089112079, "learning_rate": 1.8526543878656555e-06, "loss": 0.7403, "step": 172 }, { "epoch": 0.05627846454131425, "grad_norm": 3.6108443897278857, "learning_rate": 1.8634886240520045e-06, "loss": 0.7626, "step": 173 }, { "epoch": 0.05660377358490566, "grad_norm": 2.6007437679542855, "learning_rate": 1.8743228602383535e-06, "loss": 0.7509, "step": 174 }, { "epoch": 0.05692908262849707, "grad_norm": 2.6267204280113803, "learning_rate": 1.8851570964247023e-06, "loss": 0.5935, "step": 175 }, { "epoch": 0.057254391672088484, "grad_norm": 3.140179292957315, "learning_rate": 1.895991332611051e-06, "loss": 0.8687, "step": 176 }, { "epoch": 0.057579700715679895, "grad_norm": 2.641479483273352, "learning_rate": 1.9068255687973999e-06, "loss": 0.6495, "step": 177 }, { "epoch": 0.057905009759271306, "grad_norm": 2.9469480239439307, "learning_rate": 1.917659804983749e-06, "loss": 0.7402, "step": 178 }, { "epoch": 0.05823031880286272, "grad_norm": 3.2413294307422618, "learning_rate": 1.9284940411700975e-06, "loss": 0.632, "step": 179 }, { "epoch": 0.05855562784645413, "grad_norm": 2.9008175032053978, "learning_rate": 1.9393282773564465e-06, "loss": 0.6851, "step": 180 }, { "epoch": 0.058880936890045546, "grad_norm": 2.883104514751169, "learning_rate": 1.9501625135427955e-06, "loss": 0.6946, "step": 181 }, { "epoch": 0.05920624593363696, "grad_norm": 3.363824908589225, "learning_rate": 1.9609967497291445e-06, "loss": 0.784, "step": 182 }, { "epoch": 0.05953155497722837, "grad_norm": 2.998043207462315, "learning_rate": 1.971830985915493e-06, "loss": 0.7023, "step": 183 }, { "epoch": 0.05985686402081978, "grad_norm": 3.038033505912177, "learning_rate": 1.982665222101842e-06, "loss": 0.7035, "step": 184 }, { "epoch": 0.06018217306441119, "grad_norm": 3.44847986162391, "learning_rate": 1.9934994582881907e-06, "loss": 0.7628, "step": 185 }, { "epoch": 0.0605074821080026, "grad_norm": 2.7778122937042777, "learning_rate": 2.0043336944745397e-06, "loss": 0.6221, "step": 186 }, { "epoch": 0.06083279115159401, "grad_norm": 2.99412309737159, "learning_rate": 2.0151679306608883e-06, "loss": 0.6398, "step": 187 }, { "epoch": 0.06115810019518542, "grad_norm": 2.958867051575464, "learning_rate": 2.0260021668472373e-06, "loss": 0.6648, "step": 188 }, { "epoch": 0.06148340923877684, "grad_norm": 3.1575099091803587, "learning_rate": 2.0368364030335863e-06, "loss": 0.7302, "step": 189 }, { "epoch": 0.06180871828236825, "grad_norm": 3.335365459532145, "learning_rate": 2.0476706392199353e-06, "loss": 0.7994, "step": 190 }, { "epoch": 0.06213402732595966, "grad_norm": 2.983903182645578, "learning_rate": 2.058504875406284e-06, "loss": 0.6399, "step": 191 }, { "epoch": 0.062459336369551074, "grad_norm": 2.773575073553165, "learning_rate": 2.069339111592633e-06, "loss": 0.6708, "step": 192 }, { "epoch": 0.06278464541314248, "grad_norm": 2.7428068048395136, "learning_rate": 2.080173347778982e-06, "loss": 0.593, "step": 193 }, { "epoch": 0.0631099544567339, "grad_norm": 2.5097821689496875, "learning_rate": 2.091007583965331e-06, "loss": 0.5689, "step": 194 }, { "epoch": 0.06343526350032531, "grad_norm": 2.8612019171483585, "learning_rate": 2.1018418201516795e-06, "loss": 0.6169, "step": 195 }, { "epoch": 0.06376057254391672, "grad_norm": 2.910053130330112, "learning_rate": 2.1126760563380285e-06, "loss": 0.7627, "step": 196 }, { "epoch": 0.06408588158750814, "grad_norm": 2.998959681990686, "learning_rate": 2.123510292524377e-06, "loss": 0.7095, "step": 197 }, { "epoch": 0.06441119063109954, "grad_norm": 3.3682741403289627, "learning_rate": 2.134344528710726e-06, "loss": 0.7592, "step": 198 }, { "epoch": 0.06473649967469096, "grad_norm": 2.9079834422602455, "learning_rate": 2.1451787648970747e-06, "loss": 0.688, "step": 199 }, { "epoch": 0.06506180871828236, "grad_norm": 2.4270062081017323, "learning_rate": 2.1560130010834237e-06, "loss": 0.5191, "step": 200 }, { "epoch": 0.06538711776187378, "grad_norm": 3.189993087961325, "learning_rate": 2.1668472372697727e-06, "loss": 0.8135, "step": 201 }, { "epoch": 0.06571242680546518, "grad_norm": 2.813916817222436, "learning_rate": 2.1776814734561217e-06, "loss": 0.6263, "step": 202 }, { "epoch": 0.0660377358490566, "grad_norm": 3.0850356045473872, "learning_rate": 2.1885157096424703e-06, "loss": 0.7555, "step": 203 }, { "epoch": 0.06636304489264802, "grad_norm": 2.869912849466986, "learning_rate": 2.1993499458288193e-06, "loss": 0.6618, "step": 204 }, { "epoch": 0.06668835393623943, "grad_norm": 2.65886989411455, "learning_rate": 2.2101841820151683e-06, "loss": 0.629, "step": 205 }, { "epoch": 0.06701366297983084, "grad_norm": 2.9725884889916303, "learning_rate": 2.2210184182015173e-06, "loss": 0.6813, "step": 206 }, { "epoch": 0.06733897202342225, "grad_norm": 2.8841250747726432, "learning_rate": 2.231852654387866e-06, "loss": 0.7393, "step": 207 }, { "epoch": 0.06766428106701367, "grad_norm": 2.8816072116392637, "learning_rate": 2.242686890574215e-06, "loss": 0.6132, "step": 208 }, { "epoch": 0.06798959011060507, "grad_norm": 2.4055259800984277, "learning_rate": 2.2535211267605635e-06, "loss": 0.5631, "step": 209 }, { "epoch": 0.06831489915419649, "grad_norm": 3.4065863873209583, "learning_rate": 2.2643553629469125e-06, "loss": 0.7836, "step": 210 }, { "epoch": 0.0686402081977879, "grad_norm": 2.828877658460593, "learning_rate": 2.275189599133261e-06, "loss": 0.6399, "step": 211 }, { "epoch": 0.06896551724137931, "grad_norm": 2.897989545796447, "learning_rate": 2.28602383531961e-06, "loss": 0.7698, "step": 212 }, { "epoch": 0.06929082628497073, "grad_norm": 3.14486508948451, "learning_rate": 2.296858071505959e-06, "loss": 0.7888, "step": 213 }, { "epoch": 0.06961613532856213, "grad_norm": 2.5606294180164273, "learning_rate": 2.307692307692308e-06, "loss": 0.626, "step": 214 }, { "epoch": 0.06994144437215355, "grad_norm": 3.0628132202388567, "learning_rate": 2.3185265438786567e-06, "loss": 0.6416, "step": 215 }, { "epoch": 0.07026675341574495, "grad_norm": 2.5732009838675736, "learning_rate": 2.3293607800650057e-06, "loss": 0.6251, "step": 216 }, { "epoch": 0.07059206245933637, "grad_norm": 3.3223595022877257, "learning_rate": 2.3401950162513547e-06, "loss": 0.727, "step": 217 }, { "epoch": 0.07091737150292778, "grad_norm": 3.3398326442871262, "learning_rate": 2.3510292524377033e-06, "loss": 0.7155, "step": 218 }, { "epoch": 0.0712426805465192, "grad_norm": 2.9683885514972035, "learning_rate": 2.3618634886240523e-06, "loss": 0.7394, "step": 219 }, { "epoch": 0.07156798959011061, "grad_norm": 2.672880522298248, "learning_rate": 2.372697724810401e-06, "loss": 0.6943, "step": 220 }, { "epoch": 0.07189329863370202, "grad_norm": 2.583860701893445, "learning_rate": 2.38353196099675e-06, "loss": 0.6875, "step": 221 }, { "epoch": 0.07221860767729343, "grad_norm": 2.7018580465091078, "learning_rate": 2.3943661971830984e-06, "loss": 0.7188, "step": 222 }, { "epoch": 0.07254391672088484, "grad_norm": 2.786651412046087, "learning_rate": 2.4052004333694474e-06, "loss": 0.6627, "step": 223 }, { "epoch": 0.07286922576447626, "grad_norm": 3.0554672203487407, "learning_rate": 2.4160346695557965e-06, "loss": 0.7281, "step": 224 }, { "epoch": 0.07319453480806766, "grad_norm": 2.7743954390463323, "learning_rate": 2.4268689057421455e-06, "loss": 0.6498, "step": 225 }, { "epoch": 0.07351984385165908, "grad_norm": 3.18017217300826, "learning_rate": 2.437703141928494e-06, "loss": 0.7575, "step": 226 }, { "epoch": 0.07384515289525048, "grad_norm": 2.5971038138592992, "learning_rate": 2.448537378114843e-06, "loss": 0.5396, "step": 227 }, { "epoch": 0.0741704619388419, "grad_norm": 3.1161725103770577, "learning_rate": 2.459371614301192e-06, "loss": 0.7092, "step": 228 }, { "epoch": 0.07449577098243332, "grad_norm": 3.1674288410977494, "learning_rate": 2.470205850487541e-06, "loss": 0.7068, "step": 229 }, { "epoch": 0.07482108002602472, "grad_norm": 2.560398374379332, "learning_rate": 2.4810400866738897e-06, "loss": 0.6937, "step": 230 }, { "epoch": 0.07514638906961614, "grad_norm": 2.9865443880806817, "learning_rate": 2.4918743228602387e-06, "loss": 0.6891, "step": 231 }, { "epoch": 0.07547169811320754, "grad_norm": 3.0918780551692078, "learning_rate": 2.5027085590465872e-06, "loss": 0.6585, "step": 232 }, { "epoch": 0.07579700715679896, "grad_norm": 3.3847898272332695, "learning_rate": 2.5135427952329363e-06, "loss": 0.8672, "step": 233 }, { "epoch": 0.07612231620039037, "grad_norm": 2.846633529360505, "learning_rate": 2.524377031419285e-06, "loss": 0.5986, "step": 234 }, { "epoch": 0.07644762524398178, "grad_norm": 3.1283163477332843, "learning_rate": 2.535211267605634e-06, "loss": 0.6392, "step": 235 }, { "epoch": 0.07677293428757319, "grad_norm": 3.3472678566070844, "learning_rate": 2.546045503791983e-06, "loss": 0.7515, "step": 236 }, { "epoch": 0.0770982433311646, "grad_norm": 3.0640925404658614, "learning_rate": 2.556879739978332e-06, "loss": 0.7108, "step": 237 }, { "epoch": 0.07742355237475602, "grad_norm": 3.637932938533045, "learning_rate": 2.567713976164681e-06, "loss": 0.6901, "step": 238 }, { "epoch": 0.07774886141834743, "grad_norm": 2.8991883310340634, "learning_rate": 2.57854821235103e-06, "loss": 0.7503, "step": 239 }, { "epoch": 0.07807417046193885, "grad_norm": 3.317415351673713, "learning_rate": 2.589382448537378e-06, "loss": 0.7642, "step": 240 }, { "epoch": 0.07839947950553025, "grad_norm": 3.1794322273426348, "learning_rate": 2.600216684723727e-06, "loss": 0.6967, "step": 241 }, { "epoch": 0.07872478854912167, "grad_norm": 3.265331748518726, "learning_rate": 2.611050920910076e-06, "loss": 0.7904, "step": 242 }, { "epoch": 0.07905009759271307, "grad_norm": 3.186641059960696, "learning_rate": 2.621885157096425e-06, "loss": 0.7468, "step": 243 }, { "epoch": 0.07937540663630449, "grad_norm": 3.093664520927547, "learning_rate": 2.6327193932827736e-06, "loss": 0.807, "step": 244 }, { "epoch": 0.0797007156798959, "grad_norm": 2.9976920960810842, "learning_rate": 2.6435536294691226e-06, "loss": 0.7539, "step": 245 }, { "epoch": 0.08002602472348731, "grad_norm": 2.455606554334428, "learning_rate": 2.6543878656554717e-06, "loss": 0.5422, "step": 246 }, { "epoch": 0.08035133376707873, "grad_norm": 2.816116919324623, "learning_rate": 2.6652221018418207e-06, "loss": 0.6762, "step": 247 }, { "epoch": 0.08067664281067013, "grad_norm": 2.618335467261087, "learning_rate": 2.676056338028169e-06, "loss": 0.598, "step": 248 }, { "epoch": 0.08100195185426155, "grad_norm": 2.5118059279036555, "learning_rate": 2.686890574214518e-06, "loss": 0.5809, "step": 249 }, { "epoch": 0.08132726089785296, "grad_norm": 2.8747455635371257, "learning_rate": 2.697724810400867e-06, "loss": 0.8251, "step": 250 }, { "epoch": 0.08165256994144438, "grad_norm": 2.57235754681046, "learning_rate": 2.708559046587216e-06, "loss": 0.6375, "step": 251 }, { "epoch": 0.08197787898503578, "grad_norm": 3.2055525463898364, "learning_rate": 2.719393282773565e-06, "loss": 0.7859, "step": 252 }, { "epoch": 0.0823031880286272, "grad_norm": 3.1018207004042364, "learning_rate": 2.730227518959914e-06, "loss": 0.7347, "step": 253 }, { "epoch": 0.0826284970722186, "grad_norm": 3.338859047671947, "learning_rate": 2.7410617551462624e-06, "loss": 0.8323, "step": 254 }, { "epoch": 0.08295380611581002, "grad_norm": 2.7016316951871135, "learning_rate": 2.7518959913326114e-06, "loss": 0.6881, "step": 255 }, { "epoch": 0.08327911515940144, "grad_norm": 3.1082835477861592, "learning_rate": 2.76273022751896e-06, "loss": 0.717, "step": 256 }, { "epoch": 0.08360442420299284, "grad_norm": 2.6657165337200386, "learning_rate": 2.773564463705309e-06, "loss": 0.657, "step": 257 }, { "epoch": 0.08392973324658426, "grad_norm": 2.7311088506603376, "learning_rate": 2.7843986998916576e-06, "loss": 0.6442, "step": 258 }, { "epoch": 0.08425504229017566, "grad_norm": 3.2141752389582767, "learning_rate": 2.7952329360780066e-06, "loss": 0.7192, "step": 259 }, { "epoch": 0.08458035133376708, "grad_norm": 3.2567017443517465, "learning_rate": 2.8060671722643556e-06, "loss": 0.7327, "step": 260 }, { "epoch": 0.08490566037735849, "grad_norm": 2.8491809443972373, "learning_rate": 2.8169014084507046e-06, "loss": 0.713, "step": 261 }, { "epoch": 0.0852309694209499, "grad_norm": 2.67587199951439, "learning_rate": 2.8277356446370537e-06, "loss": 0.6858, "step": 262 }, { "epoch": 0.08555627846454131, "grad_norm": 2.5777073927040575, "learning_rate": 2.8385698808234027e-06, "loss": 0.6599, "step": 263 }, { "epoch": 0.08588158750813273, "grad_norm": 2.824349961965847, "learning_rate": 2.849404117009751e-06, "loss": 0.6846, "step": 264 }, { "epoch": 0.08620689655172414, "grad_norm": 3.1453931297018025, "learning_rate": 2.8602383531961e-06, "loss": 0.7698, "step": 265 }, { "epoch": 0.08653220559531555, "grad_norm": 2.30771938332001, "learning_rate": 2.871072589382449e-06, "loss": 0.6636, "step": 266 }, { "epoch": 0.08685751463890697, "grad_norm": 2.885097345910378, "learning_rate": 2.881906825568798e-06, "loss": 0.6477, "step": 267 }, { "epoch": 0.08718282368249837, "grad_norm": 2.6072343143417323, "learning_rate": 2.8927410617551464e-06, "loss": 0.6522, "step": 268 }, { "epoch": 0.08750813272608979, "grad_norm": 2.8620794758627683, "learning_rate": 2.9035752979414954e-06, "loss": 0.5939, "step": 269 }, { "epoch": 0.08783344176968119, "grad_norm": 3.018523052371896, "learning_rate": 2.9144095341278444e-06, "loss": 0.6896, "step": 270 }, { "epoch": 0.08815875081327261, "grad_norm": 3.2141441399885817, "learning_rate": 2.925243770314193e-06, "loss": 0.7563, "step": 271 }, { "epoch": 0.08848405985686401, "grad_norm": 2.979759626062556, "learning_rate": 2.9360780065005416e-06, "loss": 0.6766, "step": 272 }, { "epoch": 0.08880936890045543, "grad_norm": 2.871964923588251, "learning_rate": 2.9469122426868906e-06, "loss": 0.6749, "step": 273 }, { "epoch": 0.08913467794404685, "grad_norm": 3.1092418047945247, "learning_rate": 2.9577464788732396e-06, "loss": 0.6403, "step": 274 }, { "epoch": 0.08945998698763825, "grad_norm": 3.3768678880414296, "learning_rate": 2.9685807150595886e-06, "loss": 0.7808, "step": 275 }, { "epoch": 0.08978529603122967, "grad_norm": 3.1326430777632375, "learning_rate": 2.9794149512459376e-06, "loss": 0.7119, "step": 276 }, { "epoch": 0.09011060507482108, "grad_norm": 2.776836813304402, "learning_rate": 2.9902491874322862e-06, "loss": 0.5551, "step": 277 }, { "epoch": 0.0904359141184125, "grad_norm": 2.956571326942952, "learning_rate": 3.0010834236186352e-06, "loss": 0.6693, "step": 278 }, { "epoch": 0.0907612231620039, "grad_norm": 2.865659459070978, "learning_rate": 3.011917659804984e-06, "loss": 0.643, "step": 279 }, { "epoch": 0.09108653220559532, "grad_norm": 2.8270881940363988, "learning_rate": 3.022751895991333e-06, "loss": 0.7557, "step": 280 }, { "epoch": 0.09141184124918672, "grad_norm": 3.0093442691732863, "learning_rate": 3.0335861321776814e-06, "loss": 0.6525, "step": 281 }, { "epoch": 0.09173715029277814, "grad_norm": 2.7590712078764112, "learning_rate": 3.0444203683640304e-06, "loss": 0.7085, "step": 282 }, { "epoch": 0.09206245933636956, "grad_norm": 2.574215491559002, "learning_rate": 3.0552546045503794e-06, "loss": 0.6907, "step": 283 }, { "epoch": 0.09238776837996096, "grad_norm": 2.579721448995013, "learning_rate": 3.0660888407367284e-06, "loss": 0.6815, "step": 284 }, { "epoch": 0.09271307742355238, "grad_norm": 2.682190032727447, "learning_rate": 3.0769230769230774e-06, "loss": 0.6226, "step": 285 }, { "epoch": 0.09303838646714378, "grad_norm": 2.6343505999121954, "learning_rate": 3.0877573131094264e-06, "loss": 0.5642, "step": 286 }, { "epoch": 0.0933636955107352, "grad_norm": 2.2571729901209148, "learning_rate": 3.0985915492957746e-06, "loss": 0.5227, "step": 287 }, { "epoch": 0.0936890045543266, "grad_norm": 2.96018784071169, "learning_rate": 3.1094257854821236e-06, "loss": 0.671, "step": 288 }, { "epoch": 0.09401431359791802, "grad_norm": 3.0959466318895346, "learning_rate": 3.1202600216684726e-06, "loss": 0.643, "step": 289 }, { "epoch": 0.09433962264150944, "grad_norm": 2.8918873796976716, "learning_rate": 3.1310942578548216e-06, "loss": 0.6576, "step": 290 }, { "epoch": 0.09466493168510084, "grad_norm": 3.1294246005359927, "learning_rate": 3.14192849404117e-06, "loss": 0.7362, "step": 291 }, { "epoch": 0.09499024072869226, "grad_norm": 2.5771814447856354, "learning_rate": 3.1527627302275192e-06, "loss": 0.5737, "step": 292 }, { "epoch": 0.09531554977228367, "grad_norm": 3.1623727533266615, "learning_rate": 3.1635969664138682e-06, "loss": 0.6763, "step": 293 }, { "epoch": 0.09564085881587508, "grad_norm": 2.5956382087053647, "learning_rate": 3.1744312026002172e-06, "loss": 0.6974, "step": 294 }, { "epoch": 0.09596616785946649, "grad_norm": 2.9558132978682865, "learning_rate": 3.1852654387865654e-06, "loss": 0.6903, "step": 295 }, { "epoch": 0.09629147690305791, "grad_norm": 3.177428823820122, "learning_rate": 3.1960996749729144e-06, "loss": 0.7665, "step": 296 }, { "epoch": 0.09661678594664931, "grad_norm": 2.8108681034916128, "learning_rate": 3.2069339111592634e-06, "loss": 0.6709, "step": 297 }, { "epoch": 0.09694209499024073, "grad_norm": 2.82376983678017, "learning_rate": 3.2177681473456124e-06, "loss": 0.6584, "step": 298 }, { "epoch": 0.09726740403383215, "grad_norm": 2.615653713048519, "learning_rate": 3.2286023835319614e-06, "loss": 0.5343, "step": 299 }, { "epoch": 0.09759271307742355, "grad_norm": 3.2736322690781243, "learning_rate": 3.2394366197183104e-06, "loss": 0.6233, "step": 300 }, { "epoch": 0.09791802212101497, "grad_norm": 3.06681646699081, "learning_rate": 3.250270855904659e-06, "loss": 0.69, "step": 301 }, { "epoch": 0.09824333116460637, "grad_norm": 2.9922103806011195, "learning_rate": 3.261105092091008e-06, "loss": 0.6929, "step": 302 }, { "epoch": 0.09856864020819779, "grad_norm": 2.8639537761854434, "learning_rate": 3.2719393282773566e-06, "loss": 0.7043, "step": 303 }, { "epoch": 0.0988939492517892, "grad_norm": 3.290738023885257, "learning_rate": 3.2827735644637056e-06, "loss": 0.8249, "step": 304 }, { "epoch": 0.09921925829538061, "grad_norm": 3.0236808184041735, "learning_rate": 3.293607800650054e-06, "loss": 0.6775, "step": 305 }, { "epoch": 0.09954456733897202, "grad_norm": 2.7382355807342598, "learning_rate": 3.304442036836403e-06, "loss": 0.6176, "step": 306 }, { "epoch": 0.09986987638256344, "grad_norm": 2.978179721716736, "learning_rate": 3.315276273022752e-06, "loss": 0.6556, "step": 307 }, { "epoch": 0.10019518542615485, "grad_norm": 2.5433468096562564, "learning_rate": 3.3261105092091012e-06, "loss": 0.6865, "step": 308 }, { "epoch": 0.10052049446974626, "grad_norm": 2.922734618904153, "learning_rate": 3.3369447453954502e-06, "loss": 0.6286, "step": 309 }, { "epoch": 0.10084580351333768, "grad_norm": 2.7600140362741072, "learning_rate": 3.3477789815817984e-06, "loss": 0.6391, "step": 310 }, { "epoch": 0.10117111255692908, "grad_norm": 2.7979651093389286, "learning_rate": 3.3586132177681474e-06, "loss": 0.6993, "step": 311 }, { "epoch": 0.1014964216005205, "grad_norm": 2.840021740525229, "learning_rate": 3.3694474539544964e-06, "loss": 0.6993, "step": 312 }, { "epoch": 0.1018217306441119, "grad_norm": 2.7535406271923843, "learning_rate": 3.3802816901408454e-06, "loss": 0.6635, "step": 313 }, { "epoch": 0.10214703968770332, "grad_norm": 2.9208438385672824, "learning_rate": 3.3911159263271944e-06, "loss": 0.6775, "step": 314 }, { "epoch": 0.10247234873129472, "grad_norm": 2.461476633640722, "learning_rate": 3.401950162513543e-06, "loss": 0.6984, "step": 315 }, { "epoch": 0.10279765777488614, "grad_norm": 3.270760465308238, "learning_rate": 3.412784398699892e-06, "loss": 0.5453, "step": 316 }, { "epoch": 0.10312296681847756, "grad_norm": 3.8222456574660915, "learning_rate": 3.423618634886241e-06, "loss": 0.7464, "step": 317 }, { "epoch": 0.10344827586206896, "grad_norm": 3.210490185498244, "learning_rate": 3.4344528710725896e-06, "loss": 0.7181, "step": 318 }, { "epoch": 0.10377358490566038, "grad_norm": 2.6738327144875216, "learning_rate": 3.445287107258938e-06, "loss": 0.6427, "step": 319 }, { "epoch": 0.10409889394925179, "grad_norm": 2.875568974394758, "learning_rate": 3.456121343445287e-06, "loss": 0.6879, "step": 320 }, { "epoch": 0.1044242029928432, "grad_norm": 3.2603070678149133, "learning_rate": 3.466955579631636e-06, "loss": 0.7799, "step": 321 }, { "epoch": 0.10474951203643461, "grad_norm": 2.3394389168672447, "learning_rate": 3.477789815817985e-06, "loss": 0.5542, "step": 322 }, { "epoch": 0.10507482108002603, "grad_norm": 3.310365650428119, "learning_rate": 3.488624052004334e-06, "loss": 0.7519, "step": 323 }, { "epoch": 0.10540013012361743, "grad_norm": 2.8657935603644797, "learning_rate": 3.4994582881906832e-06, "loss": 0.6695, "step": 324 }, { "epoch": 0.10572543916720885, "grad_norm": 3.1087446004764043, "learning_rate": 3.510292524377032e-06, "loss": 0.7392, "step": 325 }, { "epoch": 0.10605074821080027, "grad_norm": 3.226483771905756, "learning_rate": 3.5211267605633804e-06, "loss": 0.7058, "step": 326 }, { "epoch": 0.10637605725439167, "grad_norm": 2.557421979845275, "learning_rate": 3.5319609967497294e-06, "loss": 0.6665, "step": 327 }, { "epoch": 0.10670136629798309, "grad_norm": 2.839101782903436, "learning_rate": 3.5427952329360784e-06, "loss": 0.7587, "step": 328 }, { "epoch": 0.10702667534157449, "grad_norm": 3.0271040678639216, "learning_rate": 3.553629469122427e-06, "loss": 0.6641, "step": 329 }, { "epoch": 0.10735198438516591, "grad_norm": 2.4140623701712016, "learning_rate": 3.564463705308776e-06, "loss": 0.5364, "step": 330 }, { "epoch": 0.10767729342875731, "grad_norm": 3.1798931790938125, "learning_rate": 3.575297941495125e-06, "loss": 0.7686, "step": 331 }, { "epoch": 0.10800260247234873, "grad_norm": 2.6381112711535484, "learning_rate": 3.586132177681474e-06, "loss": 0.5739, "step": 332 }, { "epoch": 0.10832791151594014, "grad_norm": 2.5891399075109405, "learning_rate": 3.596966413867823e-06, "loss": 0.587, "step": 333 }, { "epoch": 0.10865322055953155, "grad_norm": 2.974487515854205, "learning_rate": 3.607800650054171e-06, "loss": 0.5733, "step": 334 }, { "epoch": 0.10897852960312297, "grad_norm": 3.310183251067066, "learning_rate": 3.61863488624052e-06, "loss": 0.6865, "step": 335 }, { "epoch": 0.10930383864671438, "grad_norm": 2.4939756025967204, "learning_rate": 3.629469122426869e-06, "loss": 0.5987, "step": 336 }, { "epoch": 0.1096291476903058, "grad_norm": 3.2787349215451362, "learning_rate": 3.640303358613218e-06, "loss": 0.7123, "step": 337 }, { "epoch": 0.1099544567338972, "grad_norm": 2.485314189282654, "learning_rate": 3.6511375947995668e-06, "loss": 0.5554, "step": 338 }, { "epoch": 0.11027976577748862, "grad_norm": 3.1388302832612567, "learning_rate": 3.6619718309859158e-06, "loss": 0.7497, "step": 339 }, { "epoch": 0.11060507482108002, "grad_norm": 3.137598287271287, "learning_rate": 3.672806067172265e-06, "loss": 0.6958, "step": 340 }, { "epoch": 0.11093038386467144, "grad_norm": 2.9624466988764624, "learning_rate": 3.683640303358614e-06, "loss": 0.7259, "step": 341 }, { "epoch": 0.11125569290826284, "grad_norm": 3.1187964288799597, "learning_rate": 3.694474539544962e-06, "loss": 0.6819, "step": 342 }, { "epoch": 0.11158100195185426, "grad_norm": 3.341013340181796, "learning_rate": 3.705308775731311e-06, "loss": 0.6518, "step": 343 }, { "epoch": 0.11190631099544568, "grad_norm": 2.9240997951250955, "learning_rate": 3.71614301191766e-06, "loss": 0.5345, "step": 344 }, { "epoch": 0.11223162003903708, "grad_norm": 3.09126500322825, "learning_rate": 3.726977248104009e-06, "loss": 0.748, "step": 345 }, { "epoch": 0.1125569290826285, "grad_norm": 2.9562012944308664, "learning_rate": 3.737811484290358e-06, "loss": 0.7106, "step": 346 }, { "epoch": 0.1128822381262199, "grad_norm": 3.2425993652236493, "learning_rate": 3.748645720476707e-06, "loss": 0.6646, "step": 347 }, { "epoch": 0.11320754716981132, "grad_norm": 2.9593134517283453, "learning_rate": 3.7594799566630556e-06, "loss": 0.6466, "step": 348 }, { "epoch": 0.11353285621340273, "grad_norm": 3.217355824371547, "learning_rate": 3.7703141928494046e-06, "loss": 0.7789, "step": 349 }, { "epoch": 0.11385816525699415, "grad_norm": 2.7804894428967364, "learning_rate": 3.781148429035753e-06, "loss": 0.7091, "step": 350 }, { "epoch": 0.11418347430058555, "grad_norm": 2.73124348853071, "learning_rate": 3.791982665222102e-06, "loss": 0.6626, "step": 351 }, { "epoch": 0.11450878334417697, "grad_norm": 3.1511202972938377, "learning_rate": 3.8028169014084508e-06, "loss": 0.6676, "step": 352 }, { "epoch": 0.11483409238776839, "grad_norm": 2.954819964482009, "learning_rate": 3.8136511375947998e-06, "loss": 0.7255, "step": 353 }, { "epoch": 0.11515940143135979, "grad_norm": 3.0716313774161, "learning_rate": 3.824485373781149e-06, "loss": 0.6155, "step": 354 }, { "epoch": 0.11548471047495121, "grad_norm": 2.803735817807721, "learning_rate": 3.835319609967498e-06, "loss": 0.73, "step": 355 }, { "epoch": 0.11581001951854261, "grad_norm": 2.9885920484629014, "learning_rate": 3.846153846153847e-06, "loss": 0.6946, "step": 356 }, { "epoch": 0.11613532856213403, "grad_norm": 2.9339385789945256, "learning_rate": 3.856988082340195e-06, "loss": 0.773, "step": 357 }, { "epoch": 0.11646063760572543, "grad_norm": 2.8027269096651706, "learning_rate": 3.867822318526544e-06, "loss": 0.6572, "step": 358 }, { "epoch": 0.11678594664931685, "grad_norm": 2.7533369536810093, "learning_rate": 3.878656554712893e-06, "loss": 0.7153, "step": 359 }, { "epoch": 0.11711125569290826, "grad_norm": 2.5072348146963654, "learning_rate": 3.889490790899242e-06, "loss": 0.5614, "step": 360 }, { "epoch": 0.11743656473649967, "grad_norm": 2.8739307726717245, "learning_rate": 3.900325027085591e-06, "loss": 0.725, "step": 361 }, { "epoch": 0.11776187378009109, "grad_norm": 2.707458909983197, "learning_rate": 3.91115926327194e-06, "loss": 0.6568, "step": 362 }, { "epoch": 0.1180871828236825, "grad_norm": 3.3354775836625965, "learning_rate": 3.921993499458289e-06, "loss": 0.7502, "step": 363 }, { "epoch": 0.11841249186727391, "grad_norm": 2.9289590122799325, "learning_rate": 3.932827735644638e-06, "loss": 0.6953, "step": 364 }, { "epoch": 0.11873780091086532, "grad_norm": 3.342407611204002, "learning_rate": 3.943661971830986e-06, "loss": 0.723, "step": 365 }, { "epoch": 0.11906310995445674, "grad_norm": 3.2557408848333993, "learning_rate": 3.954496208017335e-06, "loss": 0.6738, "step": 366 }, { "epoch": 0.11938841899804814, "grad_norm": 3.1908824998254697, "learning_rate": 3.965330444203684e-06, "loss": 0.6342, "step": 367 }, { "epoch": 0.11971372804163956, "grad_norm": 3.086739951274934, "learning_rate": 3.976164680390033e-06, "loss": 0.7245, "step": 368 }, { "epoch": 0.12003903708523098, "grad_norm": 2.777813086357138, "learning_rate": 3.986998916576381e-06, "loss": 0.7463, "step": 369 }, { "epoch": 0.12036434612882238, "grad_norm": 2.567386455966998, "learning_rate": 3.99783315276273e-06, "loss": 0.5132, "step": 370 }, { "epoch": 0.1206896551724138, "grad_norm": 3.133706123308997, "learning_rate": 4.008667388949079e-06, "loss": 0.7951, "step": 371 }, { "epoch": 0.1210149642160052, "grad_norm": 2.8082223702865643, "learning_rate": 4.019501625135428e-06, "loss": 0.6064, "step": 372 }, { "epoch": 0.12134027325959662, "grad_norm": 3.180052737170842, "learning_rate": 4.0303358613217765e-06, "loss": 0.6457, "step": 373 }, { "epoch": 0.12166558230318802, "grad_norm": 3.0295373305520146, "learning_rate": 4.0411700975081255e-06, "loss": 0.6819, "step": 374 }, { "epoch": 0.12199089134677944, "grad_norm": 3.1226866475804, "learning_rate": 4.0520043336944745e-06, "loss": 0.6555, "step": 375 }, { "epoch": 0.12231620039037085, "grad_norm": 2.6297510636365313, "learning_rate": 4.0628385698808236e-06, "loss": 0.6116, "step": 376 }, { "epoch": 0.12264150943396226, "grad_norm": 2.682361658964605, "learning_rate": 4.0736728060671726e-06, "loss": 0.701, "step": 377 }, { "epoch": 0.12296681847755368, "grad_norm": 2.5890980448606573, "learning_rate": 4.0845070422535216e-06, "loss": 0.5727, "step": 378 }, { "epoch": 0.12329212752114509, "grad_norm": 3.051861823958655, "learning_rate": 4.0953412784398706e-06, "loss": 0.6103, "step": 379 }, { "epoch": 0.1236174365647365, "grad_norm": 2.5146082706963684, "learning_rate": 4.10617551462622e-06, "loss": 0.5475, "step": 380 }, { "epoch": 0.12394274560832791, "grad_norm": 2.4604187409702, "learning_rate": 4.117009750812568e-06, "loss": 0.6081, "step": 381 }, { "epoch": 0.12426805465191933, "grad_norm": 2.6473747899629645, "learning_rate": 4.127843986998917e-06, "loss": 0.5713, "step": 382 }, { "epoch": 0.12459336369551073, "grad_norm": 3.4619745592869737, "learning_rate": 4.138678223185266e-06, "loss": 0.6608, "step": 383 }, { "epoch": 0.12491867273910215, "grad_norm": 3.671249461511834, "learning_rate": 4.149512459371615e-06, "loss": 0.6469, "step": 384 }, { "epoch": 0.12524398178269355, "grad_norm": 3.050443573468867, "learning_rate": 4.160346695557964e-06, "loss": 0.6736, "step": 385 }, { "epoch": 0.12556929082628496, "grad_norm": 2.766664138266501, "learning_rate": 4.171180931744313e-06, "loss": 0.7507, "step": 386 }, { "epoch": 0.1258945998698764, "grad_norm": 2.964646842336732, "learning_rate": 4.182015167930662e-06, "loss": 0.6832, "step": 387 }, { "epoch": 0.1262199089134678, "grad_norm": 3.017710659850756, "learning_rate": 4.19284940411701e-06, "loss": 0.6428, "step": 388 }, { "epoch": 0.1265452179570592, "grad_norm": 2.8115015785513315, "learning_rate": 4.203683640303359e-06, "loss": 0.7027, "step": 389 }, { "epoch": 0.12687052700065063, "grad_norm": 2.8402749985229607, "learning_rate": 4.214517876489708e-06, "loss": 0.6502, "step": 390 }, { "epoch": 0.12719583604424203, "grad_norm": 2.866220500395111, "learning_rate": 4.225352112676057e-06, "loss": 0.6465, "step": 391 }, { "epoch": 0.12752114508783344, "grad_norm": 3.0042900487149344, "learning_rate": 4.236186348862405e-06, "loss": 0.6857, "step": 392 }, { "epoch": 0.12784645413142484, "grad_norm": 2.96492967829707, "learning_rate": 4.247020585048754e-06, "loss": 0.6885, "step": 393 }, { "epoch": 0.12817176317501627, "grad_norm": 2.452555821425341, "learning_rate": 4.257854821235103e-06, "loss": 0.5627, "step": 394 }, { "epoch": 0.12849707221860768, "grad_norm": 2.9400025356041417, "learning_rate": 4.268689057421452e-06, "loss": 0.6954, "step": 395 }, { "epoch": 0.12882238126219908, "grad_norm": 2.199348672996651, "learning_rate": 4.279523293607801e-06, "loss": 0.5407, "step": 396 }, { "epoch": 0.1291476903057905, "grad_norm": 2.846577884142826, "learning_rate": 4.290357529794149e-06, "loss": 0.773, "step": 397 }, { "epoch": 0.12947299934938192, "grad_norm": 2.687367511147579, "learning_rate": 4.301191765980498e-06, "loss": 0.7072, "step": 398 }, { "epoch": 0.12979830839297332, "grad_norm": 2.9398410619555064, "learning_rate": 4.312026002166847e-06, "loss": 0.7395, "step": 399 }, { "epoch": 0.13012361743656473, "grad_norm": 3.39521760980338, "learning_rate": 4.322860238353196e-06, "loss": 0.7052, "step": 400 }, { "epoch": 0.13044892648015616, "grad_norm": 3.054988787897669, "learning_rate": 4.333694474539545e-06, "loss": 0.6515, "step": 401 }, { "epoch": 0.13077423552374756, "grad_norm": 2.412466373492731, "learning_rate": 4.344528710725894e-06, "loss": 0.6486, "step": 402 }, { "epoch": 0.13109954456733897, "grad_norm": 2.63028028714477, "learning_rate": 4.355362946912243e-06, "loss": 0.5941, "step": 403 }, { "epoch": 0.13142485361093037, "grad_norm": 2.5408824746691345, "learning_rate": 4.3661971830985915e-06, "loss": 0.5522, "step": 404 }, { "epoch": 0.1317501626545218, "grad_norm": 2.6756088975137944, "learning_rate": 4.3770314192849405e-06, "loss": 0.6885, "step": 405 }, { "epoch": 0.1320754716981132, "grad_norm": 2.5561067732133647, "learning_rate": 4.3878656554712895e-06, "loss": 0.7024, "step": 406 }, { "epoch": 0.1324007807417046, "grad_norm": 3.131742161717271, "learning_rate": 4.3986998916576385e-06, "loss": 0.6764, "step": 407 }, { "epoch": 0.13272608978529604, "grad_norm": 2.722264995132695, "learning_rate": 4.4095341278439876e-06, "loss": 0.6938, "step": 408 }, { "epoch": 0.13305139882888745, "grad_norm": 2.6601887857358673, "learning_rate": 4.4203683640303366e-06, "loss": 0.5446, "step": 409 }, { "epoch": 0.13337670787247885, "grad_norm": 2.409863871967864, "learning_rate": 4.4312026002166856e-06, "loss": 0.541, "step": 410 }, { "epoch": 0.13370201691607025, "grad_norm": 2.7424947758377365, "learning_rate": 4.4420368364030346e-06, "loss": 0.628, "step": 411 }, { "epoch": 0.1340273259596617, "grad_norm": 2.525110964619196, "learning_rate": 4.452871072589383e-06, "loss": 0.6832, "step": 412 }, { "epoch": 0.1343526350032531, "grad_norm": 2.761426118923358, "learning_rate": 4.463705308775732e-06, "loss": 0.6075, "step": 413 }, { "epoch": 0.1346779440468445, "grad_norm": 2.868026254931754, "learning_rate": 4.474539544962081e-06, "loss": 0.5232, "step": 414 }, { "epoch": 0.13500325309043593, "grad_norm": 3.0308005905066024, "learning_rate": 4.48537378114843e-06, "loss": 0.7003, "step": 415 }, { "epoch": 0.13532856213402733, "grad_norm": 2.610391522606638, "learning_rate": 4.496208017334778e-06, "loss": 0.6621, "step": 416 }, { "epoch": 0.13565387117761873, "grad_norm": 2.871007044467075, "learning_rate": 4.507042253521127e-06, "loss": 0.6394, "step": 417 }, { "epoch": 0.13597918022121014, "grad_norm": 3.0136971801228163, "learning_rate": 4.517876489707476e-06, "loss": 0.6572, "step": 418 }, { "epoch": 0.13630448926480157, "grad_norm": 2.8390866153897316, "learning_rate": 4.528710725893825e-06, "loss": 0.6747, "step": 419 }, { "epoch": 0.13662979830839297, "grad_norm": 2.5345469165866827, "learning_rate": 4.539544962080173e-06, "loss": 0.6133, "step": 420 }, { "epoch": 0.13695510735198438, "grad_norm": 2.6509596156948927, "learning_rate": 4.550379198266522e-06, "loss": 0.5938, "step": 421 }, { "epoch": 0.1372804163955758, "grad_norm": 3.271069484904463, "learning_rate": 4.561213434452871e-06, "loss": 0.6711, "step": 422 }, { "epoch": 0.13760572543916721, "grad_norm": 2.6828599008550316, "learning_rate": 4.57204767063922e-06, "loss": 0.6717, "step": 423 }, { "epoch": 0.13793103448275862, "grad_norm": 2.6708974828613834, "learning_rate": 4.582881906825569e-06, "loss": 0.5734, "step": 424 }, { "epoch": 0.13825634352635002, "grad_norm": 3.4543840504163628, "learning_rate": 4.593716143011918e-06, "loss": 0.7372, "step": 425 }, { "epoch": 0.13858165256994145, "grad_norm": 2.6874264167388278, "learning_rate": 4.604550379198267e-06, "loss": 0.6677, "step": 426 }, { "epoch": 0.13890696161353286, "grad_norm": 2.7343575636159434, "learning_rate": 4.615384615384616e-06, "loss": 0.6074, "step": 427 }, { "epoch": 0.13923227065712426, "grad_norm": 3.0085222250679684, "learning_rate": 4.626218851570964e-06, "loss": 0.7234, "step": 428 }, { "epoch": 0.13955757970071567, "grad_norm": 2.993901788741483, "learning_rate": 4.637053087757313e-06, "loss": 0.7356, "step": 429 }, { "epoch": 0.1398828887443071, "grad_norm": 2.778022151402203, "learning_rate": 4.647887323943662e-06, "loss": 0.7073, "step": 430 }, { "epoch": 0.1402081977878985, "grad_norm": 2.6189838472492797, "learning_rate": 4.658721560130011e-06, "loss": 0.6349, "step": 431 }, { "epoch": 0.1405335068314899, "grad_norm": 2.971512334711116, "learning_rate": 4.66955579631636e-06, "loss": 0.7047, "step": 432 }, { "epoch": 0.14085881587508134, "grad_norm": 2.7843175995259863, "learning_rate": 4.680390032502709e-06, "loss": 0.7596, "step": 433 }, { "epoch": 0.14118412491867274, "grad_norm": 2.6877601889177374, "learning_rate": 4.691224268689058e-06, "loss": 0.6252, "step": 434 }, { "epoch": 0.14150943396226415, "grad_norm": 2.465484020644823, "learning_rate": 4.7020585048754065e-06, "loss": 0.5594, "step": 435 }, { "epoch": 0.14183474300585555, "grad_norm": 3.1348151352642963, "learning_rate": 4.7128927410617555e-06, "loss": 0.7282, "step": 436 }, { "epoch": 0.14216005204944698, "grad_norm": 2.964567690541006, "learning_rate": 4.7237269772481045e-06, "loss": 0.6004, "step": 437 }, { "epoch": 0.1424853610930384, "grad_norm": 2.9986354469004217, "learning_rate": 4.7345612134344535e-06, "loss": 0.6987, "step": 438 }, { "epoch": 0.1428106701366298, "grad_norm": 2.6247611961195623, "learning_rate": 4.745395449620802e-06, "loss": 0.6336, "step": 439 }, { "epoch": 0.14313597918022122, "grad_norm": 3.282536115712243, "learning_rate": 4.756229685807151e-06, "loss": 0.681, "step": 440 }, { "epoch": 0.14346128822381263, "grad_norm": 2.6231602140903045, "learning_rate": 4.7670639219935e-06, "loss": 0.5809, "step": 441 }, { "epoch": 0.14378659726740403, "grad_norm": 2.717083634633252, "learning_rate": 4.777898158179849e-06, "loss": 0.6919, "step": 442 }, { "epoch": 0.14411190631099544, "grad_norm": 2.805040159405104, "learning_rate": 4.788732394366197e-06, "loss": 0.6885, "step": 443 }, { "epoch": 0.14443721535458687, "grad_norm": 2.7387127046463147, "learning_rate": 4.799566630552546e-06, "loss": 0.6992, "step": 444 }, { "epoch": 0.14476252439817827, "grad_norm": 3.245419310865835, "learning_rate": 4.810400866738895e-06, "loss": 0.7614, "step": 445 }, { "epoch": 0.14508783344176968, "grad_norm": 2.6582220396150653, "learning_rate": 4.821235102925244e-06, "loss": 0.5895, "step": 446 }, { "epoch": 0.14541314248536108, "grad_norm": 2.6010190434637606, "learning_rate": 4.832069339111593e-06, "loss": 0.5862, "step": 447 }, { "epoch": 0.1457384515289525, "grad_norm": 2.6192745750874833, "learning_rate": 4.842903575297942e-06, "loss": 0.6172, "step": 448 }, { "epoch": 0.14606376057254392, "grad_norm": 2.4692190902466353, "learning_rate": 4.853737811484291e-06, "loss": 0.597, "step": 449 }, { "epoch": 0.14638906961613532, "grad_norm": 2.5312719364682796, "learning_rate": 4.86457204767064e-06, "loss": 0.5865, "step": 450 }, { "epoch": 0.14671437865972675, "grad_norm": 2.9309698347740194, "learning_rate": 4.875406283856988e-06, "loss": 0.6929, "step": 451 }, { "epoch": 0.14703968770331816, "grad_norm": 3.1496886205402532, "learning_rate": 4.886240520043337e-06, "loss": 0.7454, "step": 452 }, { "epoch": 0.14736499674690956, "grad_norm": 2.903769638064444, "learning_rate": 4.897074756229686e-06, "loss": 0.6716, "step": 453 }, { "epoch": 0.14769030579050096, "grad_norm": 2.692606594399888, "learning_rate": 4.907908992416035e-06, "loss": 0.671, "step": 454 }, { "epoch": 0.1480156148340924, "grad_norm": 2.5642324965419188, "learning_rate": 4.918743228602384e-06, "loss": 0.696, "step": 455 }, { "epoch": 0.1483409238776838, "grad_norm": 3.5024025646953754, "learning_rate": 4.929577464788733e-06, "loss": 0.6961, "step": 456 }, { "epoch": 0.1486662329212752, "grad_norm": 2.6024778812601386, "learning_rate": 4.940411700975082e-06, "loss": 0.5662, "step": 457 }, { "epoch": 0.14899154196486664, "grad_norm": 3.0631790191267396, "learning_rate": 4.951245937161431e-06, "loss": 0.6963, "step": 458 }, { "epoch": 0.14931685100845804, "grad_norm": 2.63986646772824, "learning_rate": 4.962080173347779e-06, "loss": 0.6426, "step": 459 }, { "epoch": 0.14964216005204944, "grad_norm": 2.6582066890980043, "learning_rate": 4.972914409534128e-06, "loss": 0.6157, "step": 460 }, { "epoch": 0.14996746909564085, "grad_norm": 3.1463779582929607, "learning_rate": 4.983748645720477e-06, "loss": 0.7021, "step": 461 }, { "epoch": 0.15029277813923228, "grad_norm": 2.8426794994305475, "learning_rate": 4.994582881906826e-06, "loss": 0.7204, "step": 462 }, { "epoch": 0.15061808718282368, "grad_norm": 2.5908147281663227, "learning_rate": 5.0054171180931745e-06, "loss": 0.6449, "step": 463 }, { "epoch": 0.1509433962264151, "grad_norm": 3.0809380732300062, "learning_rate": 5.0162513542795235e-06, "loss": 0.6752, "step": 464 }, { "epoch": 0.1512687052700065, "grad_norm": 2.448289798571044, "learning_rate": 5.0270855904658725e-06, "loss": 0.6325, "step": 465 }, { "epoch": 0.15159401431359792, "grad_norm": 3.016850560875038, "learning_rate": 5.0379198266522215e-06, "loss": 0.7653, "step": 466 }, { "epoch": 0.15191932335718933, "grad_norm": 2.903850772601131, "learning_rate": 5.04875406283857e-06, "loss": 0.6665, "step": 467 }, { "epoch": 0.15224463240078073, "grad_norm": 3.1562619466746864, "learning_rate": 5.059588299024919e-06, "loss": 0.6587, "step": 468 }, { "epoch": 0.15256994144437216, "grad_norm": 3.8391731406929805, "learning_rate": 5.070422535211268e-06, "loss": 0.7186, "step": 469 }, { "epoch": 0.15289525048796357, "grad_norm": 3.161229655849869, "learning_rate": 5.081256771397617e-06, "loss": 0.7018, "step": 470 }, { "epoch": 0.15322055953155497, "grad_norm": 2.541634154918415, "learning_rate": 5.092091007583966e-06, "loss": 0.6321, "step": 471 }, { "epoch": 0.15354586857514638, "grad_norm": 2.487664646032365, "learning_rate": 5.102925243770315e-06, "loss": 0.6535, "step": 472 }, { "epoch": 0.1538711776187378, "grad_norm": 2.5800417661777537, "learning_rate": 5.113759479956664e-06, "loss": 0.5839, "step": 473 }, { "epoch": 0.1541964866623292, "grad_norm": 2.594213443649038, "learning_rate": 5.124593716143013e-06, "loss": 0.6257, "step": 474 }, { "epoch": 0.15452179570592062, "grad_norm": 2.7640547230479653, "learning_rate": 5.135427952329362e-06, "loss": 0.6819, "step": 475 }, { "epoch": 0.15484710474951205, "grad_norm": 2.633595926217027, "learning_rate": 5.146262188515711e-06, "loss": 0.7543, "step": 476 }, { "epoch": 0.15517241379310345, "grad_norm": 2.8871453219216905, "learning_rate": 5.15709642470206e-06, "loss": 0.7326, "step": 477 }, { "epoch": 0.15549772283669486, "grad_norm": 2.714418932167984, "learning_rate": 5.167930660888409e-06, "loss": 0.5275, "step": 478 }, { "epoch": 0.15582303188028626, "grad_norm": 2.664480499393509, "learning_rate": 5.178764897074756e-06, "loss": 0.6337, "step": 479 }, { "epoch": 0.1561483409238777, "grad_norm": 3.0153973678397614, "learning_rate": 5.189599133261105e-06, "loss": 0.5316, "step": 480 }, { "epoch": 0.1564736499674691, "grad_norm": 2.9440271921022045, "learning_rate": 5.200433369447454e-06, "loss": 0.7265, "step": 481 }, { "epoch": 0.1567989590110605, "grad_norm": 2.7511394967954415, "learning_rate": 5.211267605633803e-06, "loss": 0.7233, "step": 482 }, { "epoch": 0.1571242680546519, "grad_norm": 3.131183224953084, "learning_rate": 5.222101841820152e-06, "loss": 0.6672, "step": 483 }, { "epoch": 0.15744957709824334, "grad_norm": 3.265172156443722, "learning_rate": 5.232936078006501e-06, "loss": 0.6649, "step": 484 }, { "epoch": 0.15777488614183474, "grad_norm": 2.8038501954265422, "learning_rate": 5.24377031419285e-06, "loss": 0.5995, "step": 485 }, { "epoch": 0.15810019518542615, "grad_norm": 2.927202264705359, "learning_rate": 5.254604550379199e-06, "loss": 0.7347, "step": 486 }, { "epoch": 0.15842550422901758, "grad_norm": 2.6005188554949603, "learning_rate": 5.265438786565547e-06, "loss": 0.648, "step": 487 }, { "epoch": 0.15875081327260898, "grad_norm": 3.1274630473029355, "learning_rate": 5.276273022751896e-06, "loss": 0.728, "step": 488 }, { "epoch": 0.15907612231620039, "grad_norm": 2.395019214820471, "learning_rate": 5.287107258938245e-06, "loss": 0.5862, "step": 489 }, { "epoch": 0.1594014313597918, "grad_norm": 2.9731901851685336, "learning_rate": 5.297941495124594e-06, "loss": 0.7225, "step": 490 }, { "epoch": 0.15972674040338322, "grad_norm": 3.1396696458461135, "learning_rate": 5.308775731310943e-06, "loss": 0.6347, "step": 491 }, { "epoch": 0.16005204944697463, "grad_norm": 3.0538061717415728, "learning_rate": 5.319609967497292e-06, "loss": 0.6213, "step": 492 }, { "epoch": 0.16037735849056603, "grad_norm": 3.1083279465354132, "learning_rate": 5.330444203683641e-06, "loss": 0.7423, "step": 493 }, { "epoch": 0.16070266753415746, "grad_norm": 2.481229170187959, "learning_rate": 5.3412784398699895e-06, "loss": 0.5625, "step": 494 }, { "epoch": 0.16102797657774887, "grad_norm": 2.9779496407968713, "learning_rate": 5.352112676056338e-06, "loss": 0.6181, "step": 495 }, { "epoch": 0.16135328562134027, "grad_norm": 2.613272513010822, "learning_rate": 5.362946912242687e-06, "loss": 0.731, "step": 496 }, { "epoch": 0.16167859466493167, "grad_norm": 2.556787507546498, "learning_rate": 5.373781148429036e-06, "loss": 0.5646, "step": 497 }, { "epoch": 0.1620039037085231, "grad_norm": 2.6788644163128823, "learning_rate": 5.384615384615385e-06, "loss": 0.6911, "step": 498 }, { "epoch": 0.1623292127521145, "grad_norm": 2.9642285729086746, "learning_rate": 5.395449620801734e-06, "loss": 0.5513, "step": 499 }, { "epoch": 0.16265452179570591, "grad_norm": 3.095715204236702, "learning_rate": 5.406283856988083e-06, "loss": 0.7765, "step": 500 }, { "epoch": 0.16297983083929735, "grad_norm": 2.8954593484271425, "learning_rate": 5.417118093174432e-06, "loss": 0.6104, "step": 501 }, { "epoch": 0.16330513988288875, "grad_norm": 2.8868934443853833, "learning_rate": 5.427952329360781e-06, "loss": 0.7293, "step": 502 }, { "epoch": 0.16363044892648015, "grad_norm": 2.7379882074957123, "learning_rate": 5.43878656554713e-06, "loss": 0.6979, "step": 503 }, { "epoch": 0.16395575797007156, "grad_norm": 2.8457971022006996, "learning_rate": 5.449620801733479e-06, "loss": 0.7636, "step": 504 }, { "epoch": 0.164281067013663, "grad_norm": 2.223417677584685, "learning_rate": 5.460455037919828e-06, "loss": 0.5798, "step": 505 }, { "epoch": 0.1646063760572544, "grad_norm": 3.086532333292132, "learning_rate": 5.471289274106176e-06, "loss": 0.6812, "step": 506 }, { "epoch": 0.1649316851008458, "grad_norm": 3.1271742736506045, "learning_rate": 5.482123510292525e-06, "loss": 0.7565, "step": 507 }, { "epoch": 0.1652569941444372, "grad_norm": 2.732415813196074, "learning_rate": 5.492957746478874e-06, "loss": 0.7103, "step": 508 }, { "epoch": 0.16558230318802863, "grad_norm": 3.0481089500199006, "learning_rate": 5.503791982665223e-06, "loss": 0.6845, "step": 509 }, { "epoch": 0.16590761223162004, "grad_norm": 2.7867579376211293, "learning_rate": 5.514626218851571e-06, "loss": 0.5729, "step": 510 }, { "epoch": 0.16623292127521144, "grad_norm": 2.5658961026223417, "learning_rate": 5.52546045503792e-06, "loss": 0.696, "step": 511 }, { "epoch": 0.16655823031880287, "grad_norm": 2.8954803899545793, "learning_rate": 5.536294691224269e-06, "loss": 0.7197, "step": 512 }, { "epoch": 0.16688353936239428, "grad_norm": 2.7360460038054333, "learning_rate": 5.547128927410618e-06, "loss": 0.5936, "step": 513 }, { "epoch": 0.16720884840598568, "grad_norm": 3.0637038773467538, "learning_rate": 5.557963163596966e-06, "loss": 0.674, "step": 514 }, { "epoch": 0.1675341574495771, "grad_norm": 2.5342496478986183, "learning_rate": 5.568797399783315e-06, "loss": 0.6422, "step": 515 }, { "epoch": 0.16785946649316852, "grad_norm": 2.7925956950040454, "learning_rate": 5.579631635969664e-06, "loss": 0.6255, "step": 516 }, { "epoch": 0.16818477553675992, "grad_norm": 2.933953270076944, "learning_rate": 5.590465872156013e-06, "loss": 0.552, "step": 517 }, { "epoch": 0.16851008458035133, "grad_norm": 2.634472875762277, "learning_rate": 5.601300108342362e-06, "loss": 0.6294, "step": 518 }, { "epoch": 0.16883539362394276, "grad_norm": 2.5086930871150876, "learning_rate": 5.612134344528711e-06, "loss": 0.6567, "step": 519 }, { "epoch": 0.16916070266753416, "grad_norm": 2.7848853993969023, "learning_rate": 5.62296858071506e-06, "loss": 0.6727, "step": 520 }, { "epoch": 0.16948601171112557, "grad_norm": 3.0226022968660637, "learning_rate": 5.633802816901409e-06, "loss": 0.7137, "step": 521 }, { "epoch": 0.16981132075471697, "grad_norm": 2.677334467186192, "learning_rate": 5.644637053087758e-06, "loss": 0.7308, "step": 522 }, { "epoch": 0.1701366297983084, "grad_norm": 2.758753806830887, "learning_rate": 5.655471289274107e-06, "loss": 0.7249, "step": 523 }, { "epoch": 0.1704619388418998, "grad_norm": 2.9829418048447227, "learning_rate": 5.666305525460456e-06, "loss": 0.7376, "step": 524 }, { "epoch": 0.1707872478854912, "grad_norm": 2.97839033558053, "learning_rate": 5.677139761646805e-06, "loss": 0.7269, "step": 525 }, { "epoch": 0.17111255692908262, "grad_norm": 2.672473081603002, "learning_rate": 5.687973997833153e-06, "loss": 0.6024, "step": 526 }, { "epoch": 0.17143786597267405, "grad_norm": 2.2123886048917076, "learning_rate": 5.698808234019502e-06, "loss": 0.5383, "step": 527 }, { "epoch": 0.17176317501626545, "grad_norm": 3.2778579149989238, "learning_rate": 5.709642470205851e-06, "loss": 0.7356, "step": 528 }, { "epoch": 0.17208848405985686, "grad_norm": 2.374217220867115, "learning_rate": 5.7204767063922e-06, "loss": 0.5576, "step": 529 }, { "epoch": 0.1724137931034483, "grad_norm": 3.2726002532147827, "learning_rate": 5.731310942578549e-06, "loss": 0.6462, "step": 530 }, { "epoch": 0.1727391021470397, "grad_norm": 2.970248395738558, "learning_rate": 5.742145178764898e-06, "loss": 0.7413, "step": 531 }, { "epoch": 0.1730644111906311, "grad_norm": 2.4880642731334666, "learning_rate": 5.752979414951247e-06, "loss": 0.6776, "step": 532 }, { "epoch": 0.1733897202342225, "grad_norm": 3.0298631170635386, "learning_rate": 5.763813651137596e-06, "loss": 0.7729, "step": 533 }, { "epoch": 0.17371502927781393, "grad_norm": 2.682613626161563, "learning_rate": 5.774647887323944e-06, "loss": 0.6935, "step": 534 }, { "epoch": 0.17404033832140534, "grad_norm": 2.210237222564042, "learning_rate": 5.785482123510293e-06, "loss": 0.526, "step": 535 }, { "epoch": 0.17436564736499674, "grad_norm": 2.7263420479961846, "learning_rate": 5.796316359696642e-06, "loss": 0.5929, "step": 536 }, { "epoch": 0.17469095640858817, "grad_norm": 2.547737436289673, "learning_rate": 5.807150595882991e-06, "loss": 0.5553, "step": 537 }, { "epoch": 0.17501626545217958, "grad_norm": 2.633872250363435, "learning_rate": 5.81798483206934e-06, "loss": 0.6816, "step": 538 }, { "epoch": 0.17534157449577098, "grad_norm": 2.9970271030356916, "learning_rate": 5.828819068255689e-06, "loss": 0.7934, "step": 539 }, { "epoch": 0.17566688353936238, "grad_norm": 2.2560248840499106, "learning_rate": 5.839653304442038e-06, "loss": 0.5506, "step": 540 }, { "epoch": 0.17599219258295382, "grad_norm": 2.9086813390178783, "learning_rate": 5.850487540628386e-06, "loss": 0.6311, "step": 541 }, { "epoch": 0.17631750162654522, "grad_norm": 2.397844647312947, "learning_rate": 5.861321776814734e-06, "loss": 0.5729, "step": 542 }, { "epoch": 0.17664281067013662, "grad_norm": 2.819082948859934, "learning_rate": 5.872156013001083e-06, "loss": 0.5983, "step": 543 }, { "epoch": 0.17696811971372803, "grad_norm": 2.921900314371784, "learning_rate": 5.882990249187432e-06, "loss": 0.6759, "step": 544 }, { "epoch": 0.17729342875731946, "grad_norm": 3.2381585469346015, "learning_rate": 5.893824485373781e-06, "loss": 0.7718, "step": 545 }, { "epoch": 0.17761873780091086, "grad_norm": 2.9987430668149084, "learning_rate": 5.90465872156013e-06, "loss": 0.6157, "step": 546 }, { "epoch": 0.17794404684450227, "grad_norm": 2.5556859040621824, "learning_rate": 5.915492957746479e-06, "loss": 0.6578, "step": 547 }, { "epoch": 0.1782693558880937, "grad_norm": 2.6767129994848666, "learning_rate": 5.926327193932828e-06, "loss": 0.659, "step": 548 }, { "epoch": 0.1785946649316851, "grad_norm": 2.741333553141404, "learning_rate": 5.937161430119177e-06, "loss": 0.7137, "step": 549 }, { "epoch": 0.1789199739752765, "grad_norm": 3.0727780502806117, "learning_rate": 5.947995666305526e-06, "loss": 0.7219, "step": 550 }, { "epoch": 0.1792452830188679, "grad_norm": 2.482938652636518, "learning_rate": 5.958829902491875e-06, "loss": 0.6375, "step": 551 }, { "epoch": 0.17957059206245934, "grad_norm": 2.4065883412337574, "learning_rate": 5.969664138678224e-06, "loss": 0.5751, "step": 552 }, { "epoch": 0.17989590110605075, "grad_norm": 2.477034794742042, "learning_rate": 5.9804983748645724e-06, "loss": 0.5866, "step": 553 }, { "epoch": 0.18022121014964215, "grad_norm": 2.69184587131351, "learning_rate": 5.9913326110509215e-06, "loss": 0.6769, "step": 554 }, { "epoch": 0.18054651919323358, "grad_norm": 2.5222488143928308, "learning_rate": 6.0021668472372705e-06, "loss": 0.6056, "step": 555 }, { "epoch": 0.180871828236825, "grad_norm": 2.832633436154769, "learning_rate": 6.0130010834236195e-06, "loss": 0.6275, "step": 556 }, { "epoch": 0.1811971372804164, "grad_norm": 3.024272985690148, "learning_rate": 6.023835319609968e-06, "loss": 0.6534, "step": 557 }, { "epoch": 0.1815224463240078, "grad_norm": 2.725047741725292, "learning_rate": 6.034669555796317e-06, "loss": 0.7309, "step": 558 }, { "epoch": 0.18184775536759923, "grad_norm": 2.580347167704942, "learning_rate": 6.045503791982666e-06, "loss": 0.7118, "step": 559 }, { "epoch": 0.18217306441119063, "grad_norm": 2.6540132392030547, "learning_rate": 6.056338028169015e-06, "loss": 0.7435, "step": 560 }, { "epoch": 0.18249837345478204, "grad_norm": 2.872773447081678, "learning_rate": 6.067172264355363e-06, "loss": 0.7017, "step": 561 }, { "epoch": 0.18282368249837344, "grad_norm": 2.6881099144346585, "learning_rate": 6.078006500541712e-06, "loss": 0.67, "step": 562 }, { "epoch": 0.18314899154196487, "grad_norm": 2.6398295789934254, "learning_rate": 6.088840736728061e-06, "loss": 0.7135, "step": 563 }, { "epoch": 0.18347430058555628, "grad_norm": 2.7384489278257917, "learning_rate": 6.09967497291441e-06, "loss": 0.5554, "step": 564 }, { "epoch": 0.18379960962914768, "grad_norm": 2.315325928348771, "learning_rate": 6.110509209100759e-06, "loss": 0.5894, "step": 565 }, { "epoch": 0.1841249186727391, "grad_norm": 3.1681197244915538, "learning_rate": 6.121343445287108e-06, "loss": 0.6703, "step": 566 }, { "epoch": 0.18445022771633052, "grad_norm": 2.676153147726583, "learning_rate": 6.132177681473457e-06, "loss": 0.649, "step": 567 }, { "epoch": 0.18477553675992192, "grad_norm": 2.7343120789019846, "learning_rate": 6.143011917659806e-06, "loss": 0.6505, "step": 568 }, { "epoch": 0.18510084580351333, "grad_norm": 2.4276008975888295, "learning_rate": 6.153846153846155e-06, "loss": 0.6119, "step": 569 }, { "epoch": 0.18542615484710476, "grad_norm": 2.7729901745214205, "learning_rate": 6.164680390032504e-06, "loss": 0.6713, "step": 570 }, { "epoch": 0.18575146389069616, "grad_norm": 2.9568238077901956, "learning_rate": 6.175514626218853e-06, "loss": 0.659, "step": 571 }, { "epoch": 0.18607677293428757, "grad_norm": 2.4012771447792822, "learning_rate": 6.186348862405202e-06, "loss": 0.5971, "step": 572 }, { "epoch": 0.186402081977879, "grad_norm": 2.4870489701033804, "learning_rate": 6.197183098591549e-06, "loss": 0.69, "step": 573 }, { "epoch": 0.1867273910214704, "grad_norm": 2.931405785591823, "learning_rate": 6.208017334777898e-06, "loss": 0.7104, "step": 574 }, { "epoch": 0.1870527000650618, "grad_norm": 2.698728528972767, "learning_rate": 6.218851570964247e-06, "loss": 0.6708, "step": 575 }, { "epoch": 0.1873780091086532, "grad_norm": 2.304654328306907, "learning_rate": 6.229685807150596e-06, "loss": 0.4915, "step": 576 }, { "epoch": 0.18770331815224464, "grad_norm": 2.8623336789340033, "learning_rate": 6.240520043336945e-06, "loss": 0.667, "step": 577 }, { "epoch": 0.18802862719583605, "grad_norm": 2.7282270641037916, "learning_rate": 6.251354279523294e-06, "loss": 0.6403, "step": 578 }, { "epoch": 0.18835393623942745, "grad_norm": 2.263574933770867, "learning_rate": 6.262188515709643e-06, "loss": 0.5343, "step": 579 }, { "epoch": 0.18867924528301888, "grad_norm": 2.7039599979924933, "learning_rate": 6.273022751895992e-06, "loss": 0.6878, "step": 580 }, { "epoch": 0.18900455432661029, "grad_norm": 2.819724576865102, "learning_rate": 6.28385698808234e-06, "loss": 0.7144, "step": 581 }, { "epoch": 0.1893298633702017, "grad_norm": 3.0861699400387836, "learning_rate": 6.2946912242686894e-06, "loss": 0.6049, "step": 582 }, { "epoch": 0.1896551724137931, "grad_norm": 2.666395957213144, "learning_rate": 6.3055254604550384e-06, "loss": 0.748, "step": 583 }, { "epoch": 0.18998048145738453, "grad_norm": 2.6690024306506666, "learning_rate": 6.3163596966413874e-06, "loss": 0.6118, "step": 584 }, { "epoch": 0.19030579050097593, "grad_norm": 2.637732242769014, "learning_rate": 6.3271939328277364e-06, "loss": 0.7412, "step": 585 }, { "epoch": 0.19063109954456733, "grad_norm": 3.2573962033182866, "learning_rate": 6.3380281690140855e-06, "loss": 0.7116, "step": 586 }, { "epoch": 0.19095640858815874, "grad_norm": 2.9535955377796044, "learning_rate": 6.3488624052004345e-06, "loss": 0.7459, "step": 587 }, { "epoch": 0.19128171763175017, "grad_norm": 2.614485992383488, "learning_rate": 6.359696641386783e-06, "loss": 0.6224, "step": 588 }, { "epoch": 0.19160702667534157, "grad_norm": 3.443605892583483, "learning_rate": 6.370530877573131e-06, "loss": 0.8184, "step": 589 }, { "epoch": 0.19193233571893298, "grad_norm": 2.534478738841137, "learning_rate": 6.38136511375948e-06, "loss": 0.5814, "step": 590 }, { "epoch": 0.1922576447625244, "grad_norm": 2.7672618110132663, "learning_rate": 6.392199349945829e-06, "loss": 0.7321, "step": 591 }, { "epoch": 0.19258295380611581, "grad_norm": 2.7590738908882395, "learning_rate": 6.403033586132178e-06, "loss": 0.7415, "step": 592 }, { "epoch": 0.19290826284970722, "grad_norm": 3.291553823097532, "learning_rate": 6.413867822318527e-06, "loss": 0.6465, "step": 593 }, { "epoch": 0.19323357189329862, "grad_norm": 9.00289744629127, "learning_rate": 6.424702058504876e-06, "loss": 0.7493, "step": 594 }, { "epoch": 0.19355888093689005, "grad_norm": 2.8152568288170743, "learning_rate": 6.435536294691225e-06, "loss": 0.6314, "step": 595 }, { "epoch": 0.19388418998048146, "grad_norm": 42.040722795420926, "learning_rate": 6.446370530877574e-06, "loss": 0.5702, "step": 596 }, { "epoch": 0.19420949902407286, "grad_norm": 2.3605893787796295, "learning_rate": 6.457204767063923e-06, "loss": 0.6358, "step": 597 }, { "epoch": 0.1945348080676643, "grad_norm": 15.875320945439341, "learning_rate": 6.468039003250272e-06, "loss": 0.6668, "step": 598 }, { "epoch": 0.1948601171112557, "grad_norm": 2.8332586668035717, "learning_rate": 6.478873239436621e-06, "loss": 0.6385, "step": 599 }, { "epoch": 0.1951854261548471, "grad_norm": 6.668002661267808, "learning_rate": 6.489707475622969e-06, "loss": 0.7117, "step": 600 }, { "epoch": 0.1955107351984385, "grad_norm": 6.634103959566555, "learning_rate": 6.500541711809318e-06, "loss": 0.606, "step": 601 }, { "epoch": 0.19583604424202994, "grad_norm": 3.376317596730842, "learning_rate": 6.511375947995667e-06, "loss": 0.6487, "step": 602 }, { "epoch": 0.19616135328562134, "grad_norm": 3.6450903670835, "learning_rate": 6.522210184182016e-06, "loss": 0.6269, "step": 603 }, { "epoch": 0.19648666232921275, "grad_norm": 7.677220775546482, "learning_rate": 6.533044420368364e-06, "loss": 0.7519, "step": 604 }, { "epoch": 0.19681197137280415, "grad_norm": 2.46112695637979, "learning_rate": 6.543878656554713e-06, "loss": 0.6496, "step": 605 }, { "epoch": 0.19713728041639558, "grad_norm": 2.7671680014275273, "learning_rate": 6.554712892741062e-06, "loss": 0.5975, "step": 606 }, { "epoch": 0.197462589459987, "grad_norm": 3.1014780860118765, "learning_rate": 6.565547128927411e-06, "loss": 0.7179, "step": 607 }, { "epoch": 0.1977878985035784, "grad_norm": 2.7488605705206446, "learning_rate": 6.576381365113759e-06, "loss": 0.6321, "step": 608 }, { "epoch": 0.19811320754716982, "grad_norm": 2.451268830896274, "learning_rate": 6.587215601300108e-06, "loss": 0.5963, "step": 609 }, { "epoch": 0.19843851659076123, "grad_norm": 2.591150734929631, "learning_rate": 6.598049837486457e-06, "loss": 0.5981, "step": 610 }, { "epoch": 0.19876382563435263, "grad_norm": 2.6049849005354404, "learning_rate": 6.608884073672806e-06, "loss": 0.6027, "step": 611 }, { "epoch": 0.19908913467794404, "grad_norm": 2.5712955928483816, "learning_rate": 6.619718309859155e-06, "loss": 0.6539, "step": 612 }, { "epoch": 0.19941444372153547, "grad_norm": 2.9417485943077115, "learning_rate": 6.630552546045504e-06, "loss": 0.6919, "step": 613 }, { "epoch": 0.19973975276512687, "grad_norm": 2.437329830094086, "learning_rate": 6.6413867822318534e-06, "loss": 0.7095, "step": 614 }, { "epoch": 0.20006506180871828, "grad_norm": 2.6040059488907716, "learning_rate": 6.6522210184182024e-06, "loss": 0.6412, "step": 615 }, { "epoch": 0.2003903708523097, "grad_norm": 2.7625142938518876, "learning_rate": 6.6630552546045514e-06, "loss": 0.6979, "step": 616 }, { "epoch": 0.2007156798959011, "grad_norm": 2.7601649845643244, "learning_rate": 6.6738894907909004e-06, "loss": 0.6916, "step": 617 }, { "epoch": 0.20104098893949252, "grad_norm": 2.583555206399283, "learning_rate": 6.6847237269772495e-06, "loss": 0.5775, "step": 618 }, { "epoch": 0.20136629798308392, "grad_norm": 2.7753834703261795, "learning_rate": 6.695557963163597e-06, "loss": 0.6559, "step": 619 }, { "epoch": 0.20169160702667535, "grad_norm": 2.4580670524238397, "learning_rate": 6.706392199349946e-06, "loss": 0.7582, "step": 620 }, { "epoch": 0.20201691607026676, "grad_norm": 2.904216215025015, "learning_rate": 6.717226435536295e-06, "loss": 0.6803, "step": 621 }, { "epoch": 0.20234222511385816, "grad_norm": 2.773884725820699, "learning_rate": 6.728060671722644e-06, "loss": 0.6945, "step": 622 }, { "epoch": 0.20266753415744956, "grad_norm": 2.70426845117727, "learning_rate": 6.738894907908993e-06, "loss": 0.7081, "step": 623 }, { "epoch": 0.202992843201041, "grad_norm": 3.1264959392698186, "learning_rate": 6.749729144095342e-06, "loss": 0.6018, "step": 624 }, { "epoch": 0.2033181522446324, "grad_norm": 2.5713097286412396, "learning_rate": 6.760563380281691e-06, "loss": 0.5185, "step": 625 }, { "epoch": 0.2036434612882238, "grad_norm": 2.456736525537519, "learning_rate": 6.77139761646804e-06, "loss": 0.6, "step": 626 }, { "epoch": 0.20396877033181524, "grad_norm": 2.8316961398121907, "learning_rate": 6.782231852654389e-06, "loss": 0.6831, "step": 627 }, { "epoch": 0.20429407937540664, "grad_norm": 2.706535593709507, "learning_rate": 6.793066088840737e-06, "loss": 0.6136, "step": 628 }, { "epoch": 0.20461938841899804, "grad_norm": 2.2554685815036426, "learning_rate": 6.803900325027086e-06, "loss": 0.6876, "step": 629 }, { "epoch": 0.20494469746258945, "grad_norm": 2.900454802534485, "learning_rate": 6.814734561213435e-06, "loss": 0.7544, "step": 630 }, { "epoch": 0.20527000650618088, "grad_norm": 2.7326190009963316, "learning_rate": 6.825568797399784e-06, "loss": 0.6589, "step": 631 }, { "epoch": 0.20559531554977228, "grad_norm": 2.7211541332954696, "learning_rate": 6.836403033586133e-06, "loss": 0.5755, "step": 632 }, { "epoch": 0.2059206245933637, "grad_norm": 2.351840915884736, "learning_rate": 6.847237269772482e-06, "loss": 0.5658, "step": 633 }, { "epoch": 0.20624593363695512, "grad_norm": 2.918287994778528, "learning_rate": 6.858071505958831e-06, "loss": 0.6958, "step": 634 }, { "epoch": 0.20657124268054652, "grad_norm": 2.48692070466879, "learning_rate": 6.868905742145179e-06, "loss": 0.5124, "step": 635 }, { "epoch": 0.20689655172413793, "grad_norm": 2.8580014695120304, "learning_rate": 6.879739978331527e-06, "loss": 0.7099, "step": 636 }, { "epoch": 0.20722186076772933, "grad_norm": 2.1994119889622405, "learning_rate": 6.890574214517876e-06, "loss": 0.5417, "step": 637 }, { "epoch": 0.20754716981132076, "grad_norm": 2.775620777702323, "learning_rate": 6.901408450704225e-06, "loss": 0.6994, "step": 638 }, { "epoch": 0.20787247885491217, "grad_norm": 2.3990328684790563, "learning_rate": 6.912242686890574e-06, "loss": 0.6709, "step": 639 }, { "epoch": 0.20819778789850357, "grad_norm": 2.482968754199733, "learning_rate": 6.923076923076923e-06, "loss": 0.5963, "step": 640 }, { "epoch": 0.20852309694209498, "grad_norm": 2.561707610866889, "learning_rate": 6.933911159263272e-06, "loss": 0.5401, "step": 641 }, { "epoch": 0.2088484059856864, "grad_norm": 2.435068446289159, "learning_rate": 6.944745395449621e-06, "loss": 0.5905, "step": 642 }, { "epoch": 0.2091737150292778, "grad_norm": 2.665581653830356, "learning_rate": 6.95557963163597e-06, "loss": 0.6397, "step": 643 }, { "epoch": 0.20949902407286922, "grad_norm": 3.1155191278982355, "learning_rate": 6.966413867822319e-06, "loss": 0.7429, "step": 644 }, { "epoch": 0.20982433311646065, "grad_norm": 2.251952842444163, "learning_rate": 6.977248104008668e-06, "loss": 0.5951, "step": 645 }, { "epoch": 0.21014964216005205, "grad_norm": 2.878095863777894, "learning_rate": 6.9880823401950174e-06, "loss": 0.6882, "step": 646 }, { "epoch": 0.21047495120364346, "grad_norm": 2.787959288471512, "learning_rate": 6.9989165763813664e-06, "loss": 0.6702, "step": 647 }, { "epoch": 0.21080026024723486, "grad_norm": 2.3876927276421505, "learning_rate": 7.009750812567715e-06, "loss": 0.6647, "step": 648 }, { "epoch": 0.2111255692908263, "grad_norm": 2.262573584992177, "learning_rate": 7.020585048754064e-06, "loss": 0.6319, "step": 649 }, { "epoch": 0.2114508783344177, "grad_norm": 2.4939576479802645, "learning_rate": 7.031419284940413e-06, "loss": 0.6979, "step": 650 }, { "epoch": 0.2117761873780091, "grad_norm": 3.292579360832095, "learning_rate": 7.042253521126761e-06, "loss": 0.7328, "step": 651 }, { "epoch": 0.21210149642160053, "grad_norm": 2.9395502196923733, "learning_rate": 7.05308775731311e-06, "loss": 0.6516, "step": 652 }, { "epoch": 0.21242680546519194, "grad_norm": 2.4199964907780607, "learning_rate": 7.063921993499459e-06, "loss": 0.5747, "step": 653 }, { "epoch": 0.21275211450878334, "grad_norm": 2.76357237874787, "learning_rate": 7.074756229685808e-06, "loss": 0.6945, "step": 654 }, { "epoch": 0.21307742355237475, "grad_norm": 2.6150720540255796, "learning_rate": 7.085590465872157e-06, "loss": 0.6169, "step": 655 }, { "epoch": 0.21340273259596618, "grad_norm": 2.7360784150262636, "learning_rate": 7.096424702058505e-06, "loss": 0.7032, "step": 656 }, { "epoch": 0.21372804163955758, "grad_norm": 2.3030824842647792, "learning_rate": 7.107258938244854e-06, "loss": 0.5456, "step": 657 }, { "epoch": 0.21405335068314899, "grad_norm": 2.7233924452013225, "learning_rate": 7.118093174431203e-06, "loss": 0.7207, "step": 658 }, { "epoch": 0.21437865972674042, "grad_norm": 2.6557446711968913, "learning_rate": 7.128927410617552e-06, "loss": 0.7036, "step": 659 }, { "epoch": 0.21470396877033182, "grad_norm": 2.9537279031015164, "learning_rate": 7.139761646803901e-06, "loss": 0.7554, "step": 660 }, { "epoch": 0.21502927781392323, "grad_norm": 2.425737721457921, "learning_rate": 7.15059588299025e-06, "loss": 0.5938, "step": 661 }, { "epoch": 0.21535458685751463, "grad_norm": 2.5424744049607777, "learning_rate": 7.161430119176599e-06, "loss": 0.6651, "step": 662 }, { "epoch": 0.21567989590110606, "grad_norm": 2.567009782570523, "learning_rate": 7.172264355362948e-06, "loss": 0.6038, "step": 663 }, { "epoch": 0.21600520494469747, "grad_norm": 2.7722394386968525, "learning_rate": 7.183098591549297e-06, "loss": 0.6876, "step": 664 }, { "epoch": 0.21633051398828887, "grad_norm": 2.46688822335434, "learning_rate": 7.193932827735646e-06, "loss": 0.6508, "step": 665 }, { "epoch": 0.21665582303188027, "grad_norm": 2.8423346344141875, "learning_rate": 7.204767063921993e-06, "loss": 0.73, "step": 666 }, { "epoch": 0.2169811320754717, "grad_norm": 2.7371771490751557, "learning_rate": 7.215601300108342e-06, "loss": 0.6734, "step": 667 }, { "epoch": 0.2173064411190631, "grad_norm": 2.7307528603747144, "learning_rate": 7.226435536294691e-06, "loss": 0.6957, "step": 668 }, { "epoch": 0.2176317501626545, "grad_norm": 2.6793010318583828, "learning_rate": 7.23726977248104e-06, "loss": 0.578, "step": 669 }, { "epoch": 0.21795705920624595, "grad_norm": 2.5024349665548966, "learning_rate": 7.248104008667389e-06, "loss": 0.5467, "step": 670 }, { "epoch": 0.21828236824983735, "grad_norm": 2.6442372392446036, "learning_rate": 7.258938244853738e-06, "loss": 0.5896, "step": 671 }, { "epoch": 0.21860767729342875, "grad_norm": 2.524199752717034, "learning_rate": 7.269772481040087e-06, "loss": 0.5817, "step": 672 }, { "epoch": 0.21893298633702016, "grad_norm": 2.6956181665986367, "learning_rate": 7.280606717226436e-06, "loss": 0.6354, "step": 673 }, { "epoch": 0.2192582953806116, "grad_norm": 2.444315544080005, "learning_rate": 7.291440953412785e-06, "loss": 0.6851, "step": 674 }, { "epoch": 0.219583604424203, "grad_norm": 2.489325518135392, "learning_rate": 7.3022751895991336e-06, "loss": 0.6151, "step": 675 }, { "epoch": 0.2199089134677944, "grad_norm": 2.7391701136800615, "learning_rate": 7.3131094257854826e-06, "loss": 0.7575, "step": 676 }, { "epoch": 0.22023422251138583, "grad_norm": 2.648114642481834, "learning_rate": 7.3239436619718316e-06, "loss": 0.5878, "step": 677 }, { "epoch": 0.22055953155497723, "grad_norm": 2.6212342505913298, "learning_rate": 7.334777898158181e-06, "loss": 0.6593, "step": 678 }, { "epoch": 0.22088484059856864, "grad_norm": 2.688634206444137, "learning_rate": 7.34561213434453e-06, "loss": 0.7417, "step": 679 }, { "epoch": 0.22121014964216004, "grad_norm": 2.3355408559483126, "learning_rate": 7.356446370530879e-06, "loss": 0.6324, "step": 680 }, { "epoch": 0.22153545868575147, "grad_norm": 2.307095642265365, "learning_rate": 7.367280606717228e-06, "loss": 0.5805, "step": 681 }, { "epoch": 0.22186076772934288, "grad_norm": 2.345966121634411, "learning_rate": 7.378114842903576e-06, "loss": 0.6141, "step": 682 }, { "epoch": 0.22218607677293428, "grad_norm": 2.3193752248236446, "learning_rate": 7.388949079089924e-06, "loss": 0.5292, "step": 683 }, { "epoch": 0.2225113858165257, "grad_norm": 2.940327821287861, "learning_rate": 7.399783315276273e-06, "loss": 0.7481, "step": 684 }, { "epoch": 0.22283669486011712, "grad_norm": 2.523516939393145, "learning_rate": 7.410617551462622e-06, "loss": 0.6343, "step": 685 }, { "epoch": 0.22316200390370852, "grad_norm": 2.382421340054898, "learning_rate": 7.421451787648971e-06, "loss": 0.6954, "step": 686 }, { "epoch": 0.22348731294729993, "grad_norm": 2.5638515396758943, "learning_rate": 7.43228602383532e-06, "loss": 0.723, "step": 687 }, { "epoch": 0.22381262199089136, "grad_norm": 2.705073880896878, "learning_rate": 7.443120260021669e-06, "loss": 0.5606, "step": 688 }, { "epoch": 0.22413793103448276, "grad_norm": 2.4109821773164333, "learning_rate": 7.453954496208018e-06, "loss": 0.7251, "step": 689 }, { "epoch": 0.22446324007807417, "grad_norm": 2.3750349376780155, "learning_rate": 7.464788732394367e-06, "loss": 0.6173, "step": 690 }, { "epoch": 0.22478854912166557, "grad_norm": 2.3036885749147635, "learning_rate": 7.475622968580716e-06, "loss": 0.6294, "step": 691 }, { "epoch": 0.225113858165257, "grad_norm": 2.8682831258332167, "learning_rate": 7.486457204767065e-06, "loss": 0.7507, "step": 692 }, { "epoch": 0.2254391672088484, "grad_norm": 2.5797876612506894, "learning_rate": 7.497291440953414e-06, "loss": 0.7367, "step": 693 }, { "epoch": 0.2257644762524398, "grad_norm": 2.4365694639767628, "learning_rate": 7.508125677139763e-06, "loss": 0.674, "step": 694 }, { "epoch": 0.22608978529603124, "grad_norm": 2.444094456663834, "learning_rate": 7.518959913326111e-06, "loss": 0.6617, "step": 695 }, { "epoch": 0.22641509433962265, "grad_norm": 2.232186428077931, "learning_rate": 7.52979414951246e-06, "loss": 0.6483, "step": 696 }, { "epoch": 0.22674040338321405, "grad_norm": 2.5672276818176187, "learning_rate": 7.540628385698809e-06, "loss": 0.6089, "step": 697 }, { "epoch": 0.22706571242680545, "grad_norm": 2.608125052660408, "learning_rate": 7.551462621885157e-06, "loss": 0.6423, "step": 698 }, { "epoch": 0.2273910214703969, "grad_norm": 2.1759749884888473, "learning_rate": 7.562296858071506e-06, "loss": 0.5969, "step": 699 }, { "epoch": 0.2277163305139883, "grad_norm": 2.5898052160879055, "learning_rate": 7.573131094257855e-06, "loss": 0.6873, "step": 700 }, { "epoch": 0.2280416395575797, "grad_norm": 2.826372811748989, "learning_rate": 7.583965330444204e-06, "loss": 0.7238, "step": 701 }, { "epoch": 0.2283669486011711, "grad_norm": 2.12818802183708, "learning_rate": 7.594799566630553e-06, "loss": 0.6105, "step": 702 }, { "epoch": 0.22869225764476253, "grad_norm": 2.515153309593915, "learning_rate": 7.6056338028169015e-06, "loss": 0.6934, "step": 703 }, { "epoch": 0.22901756668835394, "grad_norm": 2.359478503448573, "learning_rate": 7.6164680390032505e-06, "loss": 0.5977, "step": 704 }, { "epoch": 0.22934287573194534, "grad_norm": 3.365100610542566, "learning_rate": 7.6273022751895995e-06, "loss": 0.77, "step": 705 }, { "epoch": 0.22966818477553677, "grad_norm": 2.8266728559619727, "learning_rate": 7.63813651137595e-06, "loss": 0.7893, "step": 706 }, { "epoch": 0.22999349381912818, "grad_norm": 2.4919823537625967, "learning_rate": 7.648970747562298e-06, "loss": 0.6077, "step": 707 }, { "epoch": 0.23031880286271958, "grad_norm": 2.5560208350901075, "learning_rate": 7.659804983748646e-06, "loss": 0.6889, "step": 708 }, { "epoch": 0.23064411190631098, "grad_norm": 2.6604543017060664, "learning_rate": 7.670639219934996e-06, "loss": 0.7006, "step": 709 }, { "epoch": 0.23096942094990242, "grad_norm": 2.736636216905109, "learning_rate": 7.681473456121344e-06, "loss": 0.5292, "step": 710 }, { "epoch": 0.23129472999349382, "grad_norm": 2.056319084026975, "learning_rate": 7.692307692307694e-06, "loss": 0.582, "step": 711 }, { "epoch": 0.23162003903708522, "grad_norm": 2.5880835037486594, "learning_rate": 7.703141928494042e-06, "loss": 0.6547, "step": 712 }, { "epoch": 0.23194534808067666, "grad_norm": 2.8458025323273404, "learning_rate": 7.71397616468039e-06, "loss": 0.7424, "step": 713 }, { "epoch": 0.23227065712426806, "grad_norm": 2.6838700432058773, "learning_rate": 7.72481040086674e-06, "loss": 0.8168, "step": 714 }, { "epoch": 0.23259596616785946, "grad_norm": 2.8055730690770493, "learning_rate": 7.735644637053088e-06, "loss": 0.7614, "step": 715 }, { "epoch": 0.23292127521145087, "grad_norm": 2.7618689658427913, "learning_rate": 7.746478873239436e-06, "loss": 0.7278, "step": 716 }, { "epoch": 0.2332465842550423, "grad_norm": 2.033732600362779, "learning_rate": 7.757313109425786e-06, "loss": 0.5421, "step": 717 }, { "epoch": 0.2335718932986337, "grad_norm": 2.6086284095311507, "learning_rate": 7.768147345612134e-06, "loss": 0.671, "step": 718 }, { "epoch": 0.2338972023422251, "grad_norm": 2.4024210452668084, "learning_rate": 7.778981581798484e-06, "loss": 0.5779, "step": 719 }, { "epoch": 0.2342225113858165, "grad_norm": 2.614756610444594, "learning_rate": 7.789815817984832e-06, "loss": 0.6387, "step": 720 }, { "epoch": 0.23454782042940794, "grad_norm": 2.716822759782943, "learning_rate": 7.800650054171182e-06, "loss": 0.7076, "step": 721 }, { "epoch": 0.23487312947299935, "grad_norm": 2.786834196062593, "learning_rate": 7.81148429035753e-06, "loss": 0.707, "step": 722 }, { "epoch": 0.23519843851659075, "grad_norm": 2.2636932588620944, "learning_rate": 7.82231852654388e-06, "loss": 0.4957, "step": 723 }, { "epoch": 0.23552374756018218, "grad_norm": 2.25048737917112, "learning_rate": 7.833152762730228e-06, "loss": 0.5981, "step": 724 }, { "epoch": 0.2358490566037736, "grad_norm": 2.4639597253006515, "learning_rate": 7.843986998916578e-06, "loss": 0.6417, "step": 725 }, { "epoch": 0.236174365647365, "grad_norm": 2.48431387085183, "learning_rate": 7.854821235102926e-06, "loss": 0.6963, "step": 726 }, { "epoch": 0.2364996746909564, "grad_norm": 2.4087593765912985, "learning_rate": 7.865655471289276e-06, "loss": 0.5715, "step": 727 }, { "epoch": 0.23682498373454783, "grad_norm": 2.7554291576327055, "learning_rate": 7.876489707475624e-06, "loss": 0.7645, "step": 728 }, { "epoch": 0.23715029277813923, "grad_norm": 2.3473739880559195, "learning_rate": 7.887323943661972e-06, "loss": 0.6289, "step": 729 }, { "epoch": 0.23747560182173064, "grad_norm": 2.456094220682668, "learning_rate": 7.89815817984832e-06, "loss": 0.6004, "step": 730 }, { "epoch": 0.23780091086532207, "grad_norm": 2.680714084690913, "learning_rate": 7.90899241603467e-06, "loss": 0.6245, "step": 731 }, { "epoch": 0.23812621990891347, "grad_norm": 2.1787573965556337, "learning_rate": 7.919826652221019e-06, "loss": 0.547, "step": 732 }, { "epoch": 0.23845152895250488, "grad_norm": 2.751859652015132, "learning_rate": 7.930660888407368e-06, "loss": 0.6711, "step": 733 }, { "epoch": 0.23877683799609628, "grad_norm": 2.603929979470435, "learning_rate": 7.941495124593717e-06, "loss": 0.7559, "step": 734 }, { "epoch": 0.2391021470396877, "grad_norm": 2.4379248395577022, "learning_rate": 7.952329360780066e-06, "loss": 0.6303, "step": 735 }, { "epoch": 0.23942745608327912, "grad_norm": 2.5222703628634022, "learning_rate": 7.963163596966415e-06, "loss": 0.6846, "step": 736 }, { "epoch": 0.23975276512687052, "grad_norm": 2.153676506415249, "learning_rate": 7.973997833152763e-06, "loss": 0.6436, "step": 737 }, { "epoch": 0.24007807417046195, "grad_norm": 2.1849800868556146, "learning_rate": 7.984832069339113e-06, "loss": 0.5793, "step": 738 }, { "epoch": 0.24040338321405336, "grad_norm": 2.4345379090056394, "learning_rate": 7.99566630552546e-06, "loss": 0.6752, "step": 739 }, { "epoch": 0.24072869225764476, "grad_norm": 2.364942896552245, "learning_rate": 8.00650054171181e-06, "loss": 0.5775, "step": 740 }, { "epoch": 0.24105400130123616, "grad_norm": 2.4515616845984605, "learning_rate": 8.017334777898159e-06, "loss": 0.6613, "step": 741 }, { "epoch": 0.2413793103448276, "grad_norm": 2.339867542116927, "learning_rate": 8.028169014084509e-06, "loss": 0.5693, "step": 742 }, { "epoch": 0.241704619388419, "grad_norm": 2.749273401048426, "learning_rate": 8.039003250270857e-06, "loss": 0.7405, "step": 743 }, { "epoch": 0.2420299284320104, "grad_norm": 2.797514004875084, "learning_rate": 8.049837486457207e-06, "loss": 0.7594, "step": 744 }, { "epoch": 0.2423552374756018, "grad_norm": 2.451145676282246, "learning_rate": 8.060671722643553e-06, "loss": 0.6571, "step": 745 }, { "epoch": 0.24268054651919324, "grad_norm": 2.297998783048407, "learning_rate": 8.071505958829903e-06, "loss": 0.587, "step": 746 }, { "epoch": 0.24300585556278465, "grad_norm": 2.160897373367224, "learning_rate": 8.082340195016251e-06, "loss": 0.6159, "step": 747 }, { "epoch": 0.24333116460637605, "grad_norm": 2.581870807448273, "learning_rate": 8.093174431202601e-06, "loss": 0.7457, "step": 748 }, { "epoch": 0.24365647364996748, "grad_norm": 2.66655644993239, "learning_rate": 8.104008667388949e-06, "loss": 0.6282, "step": 749 }, { "epoch": 0.24398178269355889, "grad_norm": 2.3911122770625437, "learning_rate": 8.114842903575299e-06, "loss": 0.6806, "step": 750 }, { "epoch": 0.2443070917371503, "grad_norm": 2.6875473577849336, "learning_rate": 8.125677139761647e-06, "loss": 0.8029, "step": 751 }, { "epoch": 0.2446324007807417, "grad_norm": 2.1176344752419247, "learning_rate": 8.136511375947997e-06, "loss": 0.5603, "step": 752 }, { "epoch": 0.24495770982433313, "grad_norm": 2.519479264515592, "learning_rate": 8.147345612134345e-06, "loss": 0.6373, "step": 753 }, { "epoch": 0.24528301886792453, "grad_norm": 2.410302233130028, "learning_rate": 8.158179848320695e-06, "loss": 0.6577, "step": 754 }, { "epoch": 0.24560832791151593, "grad_norm": 2.508499345427746, "learning_rate": 8.169014084507043e-06, "loss": 0.6411, "step": 755 }, { "epoch": 0.24593363695510737, "grad_norm": 2.1604356302766683, "learning_rate": 8.179848320693391e-06, "loss": 0.6267, "step": 756 }, { "epoch": 0.24625894599869877, "grad_norm": 2.467424895693171, "learning_rate": 8.190682556879741e-06, "loss": 0.6511, "step": 757 }, { "epoch": 0.24658425504229017, "grad_norm": 2.1042610901072507, "learning_rate": 8.20151679306609e-06, "loss": 0.5034, "step": 758 }, { "epoch": 0.24690956408588158, "grad_norm": 2.472390132452989, "learning_rate": 8.21235102925244e-06, "loss": 0.6037, "step": 759 }, { "epoch": 0.247234873129473, "grad_norm": 2.2006282110701, "learning_rate": 8.223185265438787e-06, "loss": 0.6654, "step": 760 }, { "epoch": 0.2475601821730644, "grad_norm": 2.3607362170204667, "learning_rate": 8.234019501625135e-06, "loss": 0.6282, "step": 761 }, { "epoch": 0.24788549121665582, "grad_norm": 2.2816275122142433, "learning_rate": 8.244853737811485e-06, "loss": 0.7001, "step": 762 }, { "epoch": 0.24821080026024722, "grad_norm": 2.2124030920567845, "learning_rate": 8.255687973997834e-06, "loss": 0.5724, "step": 763 }, { "epoch": 0.24853610930383865, "grad_norm": 2.8040581557286797, "learning_rate": 8.266522210184182e-06, "loss": 0.7296, "step": 764 }, { "epoch": 0.24886141834743006, "grad_norm": 2.6147628257844775, "learning_rate": 8.277356446370532e-06, "loss": 0.7028, "step": 765 }, { "epoch": 0.24918672739102146, "grad_norm": 2.730928425943692, "learning_rate": 8.28819068255688e-06, "loss": 0.7207, "step": 766 }, { "epoch": 0.2495120364346129, "grad_norm": 2.453087017327771, "learning_rate": 8.29902491874323e-06, "loss": 0.7526, "step": 767 }, { "epoch": 0.2498373454782043, "grad_norm": 2.391589099205925, "learning_rate": 8.309859154929578e-06, "loss": 0.6654, "step": 768 }, { "epoch": 0.2501626545217957, "grad_norm": 2.280158785817263, "learning_rate": 8.320693391115928e-06, "loss": 0.5717, "step": 769 }, { "epoch": 0.2504879635653871, "grad_norm": 2.287642054376832, "learning_rate": 8.331527627302276e-06, "loss": 0.5484, "step": 770 }, { "epoch": 0.2508132726089785, "grad_norm": 2.3425172921228956, "learning_rate": 8.342361863488626e-06, "loss": 0.5741, "step": 771 }, { "epoch": 0.2511385816525699, "grad_norm": 2.41923439664663, "learning_rate": 8.353196099674974e-06, "loss": 0.6437, "step": 772 }, { "epoch": 0.2514638906961614, "grad_norm": 2.5314010982108726, "learning_rate": 8.364030335861324e-06, "loss": 0.6814, "step": 773 }, { "epoch": 0.2517891997397528, "grad_norm": 2.563678598511223, "learning_rate": 8.374864572047672e-06, "loss": 0.6635, "step": 774 }, { "epoch": 0.2521145087833442, "grad_norm": 2.2410741722115732, "learning_rate": 8.38569880823402e-06, "loss": 0.6863, "step": 775 }, { "epoch": 0.2524398178269356, "grad_norm": 2.410402699827324, "learning_rate": 8.396533044420368e-06, "loss": 0.611, "step": 776 }, { "epoch": 0.252765126870527, "grad_norm": 2.8106127781186694, "learning_rate": 8.407367280606718e-06, "loss": 0.7718, "step": 777 }, { "epoch": 0.2530904359141184, "grad_norm": 2.1061039550714242, "learning_rate": 8.418201516793066e-06, "loss": 0.6556, "step": 778 }, { "epoch": 0.2534157449577098, "grad_norm": 2.5521947602804396, "learning_rate": 8.429035752979416e-06, "loss": 0.7162, "step": 779 }, { "epoch": 0.25374105400130126, "grad_norm": 2.4435046215168397, "learning_rate": 8.439869989165764e-06, "loss": 0.5977, "step": 780 }, { "epoch": 0.25406636304489266, "grad_norm": 2.4390193766363257, "learning_rate": 8.450704225352114e-06, "loss": 0.7114, "step": 781 }, { "epoch": 0.25439167208848407, "grad_norm": 2.6587914289950474, "learning_rate": 8.461538461538462e-06, "loss": 0.6841, "step": 782 }, { "epoch": 0.25471698113207547, "grad_norm": 2.263262356466253, "learning_rate": 8.47237269772481e-06, "loss": 0.5396, "step": 783 }, { "epoch": 0.2550422901756669, "grad_norm": 2.617663518786959, "learning_rate": 8.48320693391116e-06, "loss": 0.6707, "step": 784 }, { "epoch": 0.2553675992192583, "grad_norm": 2.6725442989906685, "learning_rate": 8.494041170097508e-06, "loss": 0.6177, "step": 785 }, { "epoch": 0.2556929082628497, "grad_norm": 2.667962203641771, "learning_rate": 8.504875406283858e-06, "loss": 0.6355, "step": 786 }, { "epoch": 0.25601821730644114, "grad_norm": 2.4490039707498905, "learning_rate": 8.515709642470206e-06, "loss": 0.6383, "step": 787 }, { "epoch": 0.25634352635003255, "grad_norm": 2.4782475075935952, "learning_rate": 8.526543878656556e-06, "loss": 0.6946, "step": 788 }, { "epoch": 0.25666883539362395, "grad_norm": 2.297788410305421, "learning_rate": 8.537378114842904e-06, "loss": 0.5808, "step": 789 }, { "epoch": 0.25699414443721535, "grad_norm": 2.218373166588296, "learning_rate": 8.548212351029254e-06, "loss": 0.5492, "step": 790 }, { "epoch": 0.25731945348080676, "grad_norm": 2.314269477863587, "learning_rate": 8.559046587215602e-06, "loss": 0.5501, "step": 791 }, { "epoch": 0.25764476252439816, "grad_norm": 2.814518581698955, "learning_rate": 8.56988082340195e-06, "loss": 0.7794, "step": 792 }, { "epoch": 0.25797007156798957, "grad_norm": 2.5646949895725393, "learning_rate": 8.580715059588299e-06, "loss": 0.6542, "step": 793 }, { "epoch": 0.258295380611581, "grad_norm": 2.6213018601045093, "learning_rate": 8.591549295774648e-06, "loss": 0.5606, "step": 794 }, { "epoch": 0.25862068965517243, "grad_norm": 2.706106796534924, "learning_rate": 8.602383531960997e-06, "loss": 0.6563, "step": 795 }, { "epoch": 0.25894599869876384, "grad_norm": 2.1041994248389355, "learning_rate": 8.613217768147347e-06, "loss": 0.5098, "step": 796 }, { "epoch": 0.25927130774235524, "grad_norm": 2.598600302983177, "learning_rate": 8.624052004333695e-06, "loss": 0.5634, "step": 797 }, { "epoch": 0.25959661678594664, "grad_norm": 2.2797719725828918, "learning_rate": 8.634886240520045e-06, "loss": 0.631, "step": 798 }, { "epoch": 0.25992192582953805, "grad_norm": 2.4002180418487216, "learning_rate": 8.645720476706393e-06, "loss": 0.7156, "step": 799 }, { "epoch": 0.26024723487312945, "grad_norm": 2.350124917749766, "learning_rate": 8.656554712892743e-06, "loss": 0.69, "step": 800 }, { "epoch": 0.2605725439167209, "grad_norm": 2.500485522700592, "learning_rate": 8.66738894907909e-06, "loss": 0.5953, "step": 801 }, { "epoch": 0.2608978529603123, "grad_norm": 2.7905183618317912, "learning_rate": 8.67822318526544e-06, "loss": 0.7746, "step": 802 }, { "epoch": 0.2612231620039037, "grad_norm": 2.2867080387664234, "learning_rate": 8.689057421451789e-06, "loss": 0.5772, "step": 803 }, { "epoch": 0.2615484710474951, "grad_norm": 2.4273936161971372, "learning_rate": 8.699891657638137e-06, "loss": 0.686, "step": 804 }, { "epoch": 0.2618737800910865, "grad_norm": 2.0437646728606698, "learning_rate": 8.710725893824487e-06, "loss": 0.5692, "step": 805 }, { "epoch": 0.26219908913467793, "grad_norm": 2.0145837068801655, "learning_rate": 8.721560130010835e-06, "loss": 0.5461, "step": 806 }, { "epoch": 0.26252439817826934, "grad_norm": 2.4217569033271444, "learning_rate": 8.732394366197183e-06, "loss": 0.5807, "step": 807 }, { "epoch": 0.26284970722186074, "grad_norm": 2.2794494102699945, "learning_rate": 8.743228602383533e-06, "loss": 0.6321, "step": 808 }, { "epoch": 0.2631750162654522, "grad_norm": 2.8451812195139037, "learning_rate": 8.754062838569881e-06, "loss": 0.7629, "step": 809 }, { "epoch": 0.2635003253090436, "grad_norm": 2.0551393958347006, "learning_rate": 8.764897074756231e-06, "loss": 0.6101, "step": 810 }, { "epoch": 0.263825634352635, "grad_norm": 2.519595864999017, "learning_rate": 8.775731310942579e-06, "loss": 0.7312, "step": 811 }, { "epoch": 0.2641509433962264, "grad_norm": 2.6055704789222656, "learning_rate": 8.786565547128927e-06, "loss": 0.6118, "step": 812 }, { "epoch": 0.2644762524398178, "grad_norm": 2.5508188108330523, "learning_rate": 8.797399783315277e-06, "loss": 0.7438, "step": 813 }, { "epoch": 0.2648015614834092, "grad_norm": 2.17318494033093, "learning_rate": 8.808234019501625e-06, "loss": 0.5712, "step": 814 }, { "epoch": 0.2651268705270006, "grad_norm": 2.370285103558436, "learning_rate": 8.819068255687975e-06, "loss": 0.6038, "step": 815 }, { "epoch": 0.2654521795705921, "grad_norm": 2.769122671789359, "learning_rate": 8.829902491874323e-06, "loss": 0.6796, "step": 816 }, { "epoch": 0.2657774886141835, "grad_norm": 1.978875391183421, "learning_rate": 8.840736728060673e-06, "loss": 0.5509, "step": 817 }, { "epoch": 0.2661027976577749, "grad_norm": 2.244579128364061, "learning_rate": 8.851570964247021e-06, "loss": 0.5638, "step": 818 }, { "epoch": 0.2664281067013663, "grad_norm": 2.6484803625768, "learning_rate": 8.862405200433371e-06, "loss": 0.6786, "step": 819 }, { "epoch": 0.2667534157449577, "grad_norm": 2.3129707826255497, "learning_rate": 8.87323943661972e-06, "loss": 0.5414, "step": 820 }, { "epoch": 0.2670787247885491, "grad_norm": 2.1103410963897637, "learning_rate": 8.884073672806069e-06, "loss": 0.539, "step": 821 }, { "epoch": 0.2674040338321405, "grad_norm": 2.5708122020525113, "learning_rate": 8.894907908992417e-06, "loss": 0.6097, "step": 822 }, { "epoch": 0.26772934287573197, "grad_norm": 2.2323799487157627, "learning_rate": 8.905742145178765e-06, "loss": 0.6186, "step": 823 }, { "epoch": 0.2680546519193234, "grad_norm": 2.3325573286304855, "learning_rate": 8.916576381365114e-06, "loss": 0.5896, "step": 824 }, { "epoch": 0.2683799609629148, "grad_norm": 2.300672681969749, "learning_rate": 8.927410617551463e-06, "loss": 0.5791, "step": 825 }, { "epoch": 0.2687052700065062, "grad_norm": 2.5037817838093033, "learning_rate": 8.938244853737812e-06, "loss": 0.6714, "step": 826 }, { "epoch": 0.2690305790500976, "grad_norm": 2.080715623391905, "learning_rate": 8.949079089924162e-06, "loss": 0.6328, "step": 827 }, { "epoch": 0.269355888093689, "grad_norm": 2.2002716386036467, "learning_rate": 8.95991332611051e-06, "loss": 0.6281, "step": 828 }, { "epoch": 0.2696811971372804, "grad_norm": 2.3593351869803048, "learning_rate": 8.97074756229686e-06, "loss": 0.6288, "step": 829 }, { "epoch": 0.27000650618087185, "grad_norm": 2.433672848051986, "learning_rate": 8.981581798483208e-06, "loss": 0.6311, "step": 830 }, { "epoch": 0.27033181522446326, "grad_norm": 2.819334106375019, "learning_rate": 8.992416034669556e-06, "loss": 0.6913, "step": 831 }, { "epoch": 0.27065712426805466, "grad_norm": 2.4795926079481077, "learning_rate": 9.003250270855906e-06, "loss": 0.7228, "step": 832 }, { "epoch": 0.27098243331164606, "grad_norm": 2.7843025493380353, "learning_rate": 9.014084507042254e-06, "loss": 0.7253, "step": 833 }, { "epoch": 0.27130774235523747, "grad_norm": 2.3256222733132517, "learning_rate": 9.024918743228604e-06, "loss": 0.6563, "step": 834 }, { "epoch": 0.2716330513988289, "grad_norm": 2.4583091685374505, "learning_rate": 9.035752979414952e-06, "loss": 0.6228, "step": 835 }, { "epoch": 0.2719583604424203, "grad_norm": 2.750732953663972, "learning_rate": 9.046587215601302e-06, "loss": 0.6836, "step": 836 }, { "epoch": 0.27228366948601174, "grad_norm": 2.386577873338172, "learning_rate": 9.05742145178765e-06, "loss": 0.6022, "step": 837 }, { "epoch": 0.27260897852960314, "grad_norm": 2.4416609830567104, "learning_rate": 9.068255687973998e-06, "loss": 0.686, "step": 838 }, { "epoch": 0.27293428757319455, "grad_norm": 2.144219279570106, "learning_rate": 9.079089924160346e-06, "loss": 0.5446, "step": 839 }, { "epoch": 0.27325959661678595, "grad_norm": 2.6328092611769893, "learning_rate": 9.089924160346696e-06, "loss": 0.6216, "step": 840 }, { "epoch": 0.27358490566037735, "grad_norm": 2.400756477648594, "learning_rate": 9.100758396533044e-06, "loss": 0.7876, "step": 841 }, { "epoch": 0.27391021470396876, "grad_norm": 2.659111025402615, "learning_rate": 9.111592632719394e-06, "loss": 0.6882, "step": 842 }, { "epoch": 0.27423552374756016, "grad_norm": 2.1268868414730413, "learning_rate": 9.122426868905742e-06, "loss": 0.6282, "step": 843 }, { "epoch": 0.2745608327911516, "grad_norm": 2.2591503240714523, "learning_rate": 9.133261105092092e-06, "loss": 0.6762, "step": 844 }, { "epoch": 0.274886141834743, "grad_norm": 2.3273913984124643, "learning_rate": 9.14409534127844e-06, "loss": 0.6603, "step": 845 }, { "epoch": 0.27521145087833443, "grad_norm": 2.2308565044241626, "learning_rate": 9.15492957746479e-06, "loss": 0.5864, "step": 846 }, { "epoch": 0.27553675992192583, "grad_norm": 2.213701617729169, "learning_rate": 9.165763813651138e-06, "loss": 0.7194, "step": 847 }, { "epoch": 0.27586206896551724, "grad_norm": 2.55276873085806, "learning_rate": 9.176598049837488e-06, "loss": 0.6883, "step": 848 }, { "epoch": 0.27618737800910864, "grad_norm": 2.569182956904159, "learning_rate": 9.187432286023836e-06, "loss": 0.6642, "step": 849 }, { "epoch": 0.27651268705270005, "grad_norm": 2.430027560883123, "learning_rate": 9.198266522210184e-06, "loss": 0.7474, "step": 850 }, { "epoch": 0.27683799609629145, "grad_norm": 2.7922004122349486, "learning_rate": 9.209100758396534e-06, "loss": 0.7811, "step": 851 }, { "epoch": 0.2771633051398829, "grad_norm": 2.0661466977311957, "learning_rate": 9.219934994582882e-06, "loss": 0.6273, "step": 852 }, { "epoch": 0.2774886141834743, "grad_norm": 2.0830440283750438, "learning_rate": 9.230769230769232e-06, "loss": 0.6288, "step": 853 }, { "epoch": 0.2778139232270657, "grad_norm": 2.3051754644520166, "learning_rate": 9.24160346695558e-06, "loss": 0.6093, "step": 854 }, { "epoch": 0.2781392322706571, "grad_norm": 2.43341642565417, "learning_rate": 9.252437703141929e-06, "loss": 0.747, "step": 855 }, { "epoch": 0.2784645413142485, "grad_norm": 2.2853148838378705, "learning_rate": 9.263271939328278e-06, "loss": 0.592, "step": 856 }, { "epoch": 0.27878985035783993, "grad_norm": 2.9639234419367586, "learning_rate": 9.274106175514627e-06, "loss": 0.7928, "step": 857 }, { "epoch": 0.27911515940143133, "grad_norm": 2.353751220764572, "learning_rate": 9.284940411700975e-06, "loss": 0.7452, "step": 858 }, { "epoch": 0.2794404684450228, "grad_norm": 2.452414633150602, "learning_rate": 9.295774647887325e-06, "loss": 0.6747, "step": 859 }, { "epoch": 0.2797657774886142, "grad_norm": 2.5089261285224507, "learning_rate": 9.306608884073673e-06, "loss": 0.6136, "step": 860 }, { "epoch": 0.2800910865322056, "grad_norm": 2.4097779472071106, "learning_rate": 9.317443120260023e-06, "loss": 0.6558, "step": 861 }, { "epoch": 0.280416395575797, "grad_norm": 2.2308183787239892, "learning_rate": 9.32827735644637e-06, "loss": 0.6577, "step": 862 }, { "epoch": 0.2807417046193884, "grad_norm": 2.4424381272394613, "learning_rate": 9.33911159263272e-06, "loss": 0.6009, "step": 863 }, { "epoch": 0.2810670136629798, "grad_norm": 2.1530529837411745, "learning_rate": 9.349945828819069e-06, "loss": 0.6199, "step": 864 }, { "epoch": 0.2813923227065712, "grad_norm": 2.5138685475932787, "learning_rate": 9.360780065005419e-06, "loss": 0.6357, "step": 865 }, { "epoch": 0.2817176317501627, "grad_norm": 2.389666024477209, "learning_rate": 9.371614301191767e-06, "loss": 0.5958, "step": 866 }, { "epoch": 0.2820429407937541, "grad_norm": 2.144524467949945, "learning_rate": 9.382448537378117e-06, "loss": 0.7442, "step": 867 }, { "epoch": 0.2823682498373455, "grad_norm": 2.27380812270397, "learning_rate": 9.393282773564465e-06, "loss": 0.6356, "step": 868 }, { "epoch": 0.2826935588809369, "grad_norm": 2.7078658095057646, "learning_rate": 9.404117009750813e-06, "loss": 0.5656, "step": 869 }, { "epoch": 0.2830188679245283, "grad_norm": 2.375743444944045, "learning_rate": 9.414951245937161e-06, "loss": 0.6453, "step": 870 }, { "epoch": 0.2833441769681197, "grad_norm": 2.2235400832744308, "learning_rate": 9.425785482123511e-06, "loss": 0.681, "step": 871 }, { "epoch": 0.2836694860117111, "grad_norm": 2.4387408830726818, "learning_rate": 9.43661971830986e-06, "loss": 0.7131, "step": 872 }, { "epoch": 0.28399479505530256, "grad_norm": 2.1972480735865503, "learning_rate": 9.447453954496209e-06, "loss": 0.6642, "step": 873 }, { "epoch": 0.28432010409889397, "grad_norm": 2.6114406710434013, "learning_rate": 9.458288190682557e-06, "loss": 0.6737, "step": 874 }, { "epoch": 0.28464541314248537, "grad_norm": 2.3873320035985923, "learning_rate": 9.469122426868907e-06, "loss": 0.7032, "step": 875 }, { "epoch": 0.2849707221860768, "grad_norm": 2.1695735141948465, "learning_rate": 9.479956663055255e-06, "loss": 0.5481, "step": 876 }, { "epoch": 0.2852960312296682, "grad_norm": 2.192383800767499, "learning_rate": 9.490790899241603e-06, "loss": 0.6011, "step": 877 }, { "epoch": 0.2856213402732596, "grad_norm": 2.077385050995418, "learning_rate": 9.501625135427953e-06, "loss": 0.5532, "step": 878 }, { "epoch": 0.285946649316851, "grad_norm": 2.7430402560463905, "learning_rate": 9.512459371614301e-06, "loss": 0.7014, "step": 879 }, { "epoch": 0.28627195836044245, "grad_norm": 2.6716212757791853, "learning_rate": 9.523293607800651e-06, "loss": 0.8128, "step": 880 }, { "epoch": 0.28659726740403385, "grad_norm": 2.255904351075396, "learning_rate": 9.534127843987e-06, "loss": 0.6327, "step": 881 }, { "epoch": 0.28692257644762525, "grad_norm": 2.264869682083267, "learning_rate": 9.54496208017335e-06, "loss": 0.5571, "step": 882 }, { "epoch": 0.28724788549121666, "grad_norm": 2.392585301211353, "learning_rate": 9.555796316359697e-06, "loss": 0.762, "step": 883 }, { "epoch": 0.28757319453480806, "grad_norm": 2.3938985926355114, "learning_rate": 9.566630552546047e-06, "loss": 0.6538, "step": 884 }, { "epoch": 0.28789850357839947, "grad_norm": 2.19896131107596, "learning_rate": 9.577464788732394e-06, "loss": 0.6395, "step": 885 }, { "epoch": 0.28822381262199087, "grad_norm": 1.919850634290604, "learning_rate": 9.588299024918744e-06, "loss": 0.5829, "step": 886 }, { "epoch": 0.2885491216655823, "grad_norm": 2.3026824346521133, "learning_rate": 9.599133261105092e-06, "loss": 0.6868, "step": 887 }, { "epoch": 0.28887443070917374, "grad_norm": 2.241882646146785, "learning_rate": 9.609967497291442e-06, "loss": 0.6203, "step": 888 }, { "epoch": 0.28919973975276514, "grad_norm": 2.4023032543475207, "learning_rate": 9.62080173347779e-06, "loss": 0.6638, "step": 889 }, { "epoch": 0.28952504879635654, "grad_norm": 2.4724540749344097, "learning_rate": 9.63163596966414e-06, "loss": 0.6776, "step": 890 }, { "epoch": 0.28985035783994795, "grad_norm": 1.8946707331747181, "learning_rate": 9.642470205850488e-06, "loss": 0.6159, "step": 891 }, { "epoch": 0.29017566688353935, "grad_norm": 2.7326705583652533, "learning_rate": 9.653304442036838e-06, "loss": 0.7035, "step": 892 }, { "epoch": 0.29050097592713076, "grad_norm": 2.330543834334248, "learning_rate": 9.664138678223186e-06, "loss": 0.7737, "step": 893 }, { "epoch": 0.29082628497072216, "grad_norm": 2.1105343131061427, "learning_rate": 9.674972914409536e-06, "loss": 0.6446, "step": 894 }, { "epoch": 0.2911515940143136, "grad_norm": 2.000266521647179, "learning_rate": 9.685807150595884e-06, "loss": 0.6039, "step": 895 }, { "epoch": 0.291476903057905, "grad_norm": 2.4932394073593165, "learning_rate": 9.696641386782234e-06, "loss": 0.6956, "step": 896 }, { "epoch": 0.2918022121014964, "grad_norm": 2.3228564918577304, "learning_rate": 9.707475622968582e-06, "loss": 0.5897, "step": 897 }, { "epoch": 0.29212752114508783, "grad_norm": 2.355421384799299, "learning_rate": 9.71830985915493e-06, "loss": 0.6429, "step": 898 }, { "epoch": 0.29245283018867924, "grad_norm": 2.187956514463671, "learning_rate": 9.72914409534128e-06, "loss": 0.6202, "step": 899 }, { "epoch": 0.29277813923227064, "grad_norm": 2.2824406250544373, "learning_rate": 9.739978331527628e-06, "loss": 0.5987, "step": 900 }, { "epoch": 0.29310344827586204, "grad_norm": 2.126593877521654, "learning_rate": 9.750812567713976e-06, "loss": 0.6584, "step": 901 }, { "epoch": 0.2934287573194535, "grad_norm": 1.90519410048207, "learning_rate": 9.761646803900326e-06, "loss": 0.5712, "step": 902 }, { "epoch": 0.2937540663630449, "grad_norm": 2.5400529396410514, "learning_rate": 9.772481040086674e-06, "loss": 0.7419, "step": 903 }, { "epoch": 0.2940793754066363, "grad_norm": 2.259971031254078, "learning_rate": 9.783315276273024e-06, "loss": 0.6289, "step": 904 }, { "epoch": 0.2944046844502277, "grad_norm": 1.927085375853492, "learning_rate": 9.794149512459372e-06, "loss": 0.5037, "step": 905 }, { "epoch": 0.2947299934938191, "grad_norm": 2.141381219466804, "learning_rate": 9.80498374864572e-06, "loss": 0.5925, "step": 906 }, { "epoch": 0.2950553025374105, "grad_norm": 2.125764517729045, "learning_rate": 9.81581798483207e-06, "loss": 0.5536, "step": 907 }, { "epoch": 0.29538061158100193, "grad_norm": 2.7247311567542774, "learning_rate": 9.826652221018418e-06, "loss": 0.7024, "step": 908 }, { "epoch": 0.2957059206245934, "grad_norm": 2.1294422150655676, "learning_rate": 9.837486457204768e-06, "loss": 0.6717, "step": 909 }, { "epoch": 0.2960312296681848, "grad_norm": 2.1006036548108264, "learning_rate": 9.848320693391116e-06, "loss": 0.5377, "step": 910 }, { "epoch": 0.2963565387117762, "grad_norm": 2.351339562054951, "learning_rate": 9.859154929577466e-06, "loss": 0.7224, "step": 911 }, { "epoch": 0.2966818477553676, "grad_norm": 2.417177248891981, "learning_rate": 9.869989165763814e-06, "loss": 0.6516, "step": 912 }, { "epoch": 0.297007156798959, "grad_norm": 2.631489974615871, "learning_rate": 9.880823401950164e-06, "loss": 0.7769, "step": 913 }, { "epoch": 0.2973324658425504, "grad_norm": 2.2576633333185074, "learning_rate": 9.891657638136512e-06, "loss": 0.5833, "step": 914 }, { "epoch": 0.2976577748861418, "grad_norm": 2.47699491000542, "learning_rate": 9.902491874322862e-06, "loss": 0.6764, "step": 915 }, { "epoch": 0.2979830839297333, "grad_norm": 2.3564042835134282, "learning_rate": 9.91332611050921e-06, "loss": 0.623, "step": 916 }, { "epoch": 0.2983083929733247, "grad_norm": 1.6950416309389367, "learning_rate": 9.924160346695559e-06, "loss": 0.5262, "step": 917 }, { "epoch": 0.2986337020169161, "grad_norm": 2.2927742216132683, "learning_rate": 9.934994582881907e-06, "loss": 0.6741, "step": 918 }, { "epoch": 0.2989590110605075, "grad_norm": 2.2942904144199385, "learning_rate": 9.945828819068257e-06, "loss": 0.6326, "step": 919 }, { "epoch": 0.2992843201040989, "grad_norm": 2.0120321327987827, "learning_rate": 9.956663055254605e-06, "loss": 0.5536, "step": 920 }, { "epoch": 0.2996096291476903, "grad_norm": 2.410017879092579, "learning_rate": 9.967497291440955e-06, "loss": 0.7352, "step": 921 }, { "epoch": 0.2999349381912817, "grad_norm": 2.3588019832323774, "learning_rate": 9.978331527627303e-06, "loss": 0.7181, "step": 922 }, { "epoch": 0.30026024723487316, "grad_norm": 2.2130300916878767, "learning_rate": 9.989165763813653e-06, "loss": 0.6017, "step": 923 }, { "epoch": 0.30058555627846456, "grad_norm": 2.1943357861210084, "learning_rate": 1e-05, "loss": 0.5684, "step": 924 }, { "epoch": 0.30091086532205596, "grad_norm": 2.165943299362978, "learning_rate": 9.999999641748334e-06, "loss": 0.5762, "step": 925 }, { "epoch": 0.30123617436564737, "grad_norm": 2.3971301244356633, "learning_rate": 9.999998566993386e-06, "loss": 0.6621, "step": 926 }, { "epoch": 0.3015614834092388, "grad_norm": 2.248476139690311, "learning_rate": 9.999996775735307e-06, "loss": 0.6029, "step": 927 }, { "epoch": 0.3018867924528302, "grad_norm": 2.361691696669373, "learning_rate": 9.999994267974358e-06, "loss": 0.647, "step": 928 }, { "epoch": 0.3022121014964216, "grad_norm": 2.253485389097502, "learning_rate": 9.999991043710898e-06, "loss": 0.6573, "step": 929 }, { "epoch": 0.302537410540013, "grad_norm": 2.3099644029316098, "learning_rate": 9.999987102945386e-06, "loss": 0.679, "step": 930 }, { "epoch": 0.30286271958360445, "grad_norm": 2.41965188321896, "learning_rate": 9.99998244567839e-06, "loss": 0.5872, "step": 931 }, { "epoch": 0.30318802862719585, "grad_norm": 2.1307355613081875, "learning_rate": 9.999977071910574e-06, "loss": 0.6418, "step": 932 }, { "epoch": 0.30351333767078725, "grad_norm": 2.4562458343206957, "learning_rate": 9.99997098164271e-06, "loss": 0.6908, "step": 933 }, { "epoch": 0.30383864671437866, "grad_norm": 2.0677801992309752, "learning_rate": 9.999964174875672e-06, "loss": 0.6687, "step": 934 }, { "epoch": 0.30416395575797006, "grad_norm": 2.356347368284543, "learning_rate": 9.999956651610434e-06, "loss": 0.6189, "step": 935 }, { "epoch": 0.30448926480156147, "grad_norm": 2.570763530134155, "learning_rate": 9.999948411848073e-06, "loss": 0.7202, "step": 936 }, { "epoch": 0.30481457384515287, "grad_norm": 2.5370781150203543, "learning_rate": 9.999939455589773e-06, "loss": 0.6947, "step": 937 }, { "epoch": 0.30513988288874433, "grad_norm": 2.3694340070553737, "learning_rate": 9.999929782836813e-06, "loss": 0.6601, "step": 938 }, { "epoch": 0.30546519193233573, "grad_norm": 2.1728267089326154, "learning_rate": 9.999919393590584e-06, "loss": 0.6736, "step": 939 }, { "epoch": 0.30579050097592714, "grad_norm": 2.327080092220392, "learning_rate": 9.99990828785257e-06, "loss": 0.6491, "step": 940 }, { "epoch": 0.30611581001951854, "grad_norm": 2.183037321530034, "learning_rate": 9.999896465624367e-06, "loss": 0.6013, "step": 941 }, { "epoch": 0.30644111906310995, "grad_norm": 2.4241684659023246, "learning_rate": 9.999883926907665e-06, "loss": 0.5565, "step": 942 }, { "epoch": 0.30676642810670135, "grad_norm": 1.9647658212088608, "learning_rate": 9.999870671704264e-06, "loss": 0.5179, "step": 943 }, { "epoch": 0.30709173715029275, "grad_norm": 2.505776026127724, "learning_rate": 9.999856700016061e-06, "loss": 0.6816, "step": 944 }, { "epoch": 0.3074170461938842, "grad_norm": 2.0595832616459844, "learning_rate": 9.99984201184506e-06, "loss": 0.5872, "step": 945 }, { "epoch": 0.3077423552374756, "grad_norm": 2.1261082553425217, "learning_rate": 9.999826607193366e-06, "loss": 0.6061, "step": 946 }, { "epoch": 0.308067664281067, "grad_norm": 2.394510000413653, "learning_rate": 9.999810486063183e-06, "loss": 0.7009, "step": 947 }, { "epoch": 0.3083929733246584, "grad_norm": 2.3444964718047565, "learning_rate": 9.999793648456826e-06, "loss": 0.6242, "step": 948 }, { "epoch": 0.30871828236824983, "grad_norm": 2.5484677788642682, "learning_rate": 9.999776094376707e-06, "loss": 0.5942, "step": 949 }, { "epoch": 0.30904359141184123, "grad_norm": 2.468695939451968, "learning_rate": 9.999757823825338e-06, "loss": 0.6934, "step": 950 }, { "epoch": 0.30936890045543264, "grad_norm": 2.2435320836300257, "learning_rate": 9.999738836805339e-06, "loss": 0.6187, "step": 951 }, { "epoch": 0.3096942094990241, "grad_norm": 2.256204548731802, "learning_rate": 9.999719133319432e-06, "loss": 0.5387, "step": 952 }, { "epoch": 0.3100195185426155, "grad_norm": 2.2372152188926657, "learning_rate": 9.99969871337044e-06, "loss": 0.6794, "step": 953 }, { "epoch": 0.3103448275862069, "grad_norm": 2.039032830830498, "learning_rate": 9.999677576961289e-06, "loss": 0.5925, "step": 954 }, { "epoch": 0.3106701366297983, "grad_norm": 2.204660939604768, "learning_rate": 9.999655724095008e-06, "loss": 0.5573, "step": 955 }, { "epoch": 0.3109954456733897, "grad_norm": 2.27130614011351, "learning_rate": 9.999633154774729e-06, "loss": 0.5631, "step": 956 }, { "epoch": 0.3113207547169811, "grad_norm": 2.069044508633492, "learning_rate": 9.999609869003684e-06, "loss": 0.5635, "step": 957 }, { "epoch": 0.3116460637605725, "grad_norm": 2.189264292638539, "learning_rate": 9.99958586678521e-06, "loss": 0.6566, "step": 958 }, { "epoch": 0.311971372804164, "grad_norm": 2.025211712260269, "learning_rate": 9.999561148122749e-06, "loss": 0.6854, "step": 959 }, { "epoch": 0.3122966818477554, "grad_norm": 2.36285675030267, "learning_rate": 9.999535713019843e-06, "loss": 0.691, "step": 960 }, { "epoch": 0.3126219908913468, "grad_norm": 2.472633511495545, "learning_rate": 9.999509561480136e-06, "loss": 0.6641, "step": 961 }, { "epoch": 0.3129472999349382, "grad_norm": 2.163843777092068, "learning_rate": 9.999482693507376e-06, "loss": 0.6106, "step": 962 }, { "epoch": 0.3132726089785296, "grad_norm": 2.135824992988056, "learning_rate": 9.999455109105411e-06, "loss": 0.6097, "step": 963 }, { "epoch": 0.313597918022121, "grad_norm": 2.3210923546651845, "learning_rate": 9.999426808278197e-06, "loss": 0.7426, "step": 964 }, { "epoch": 0.3139232270657124, "grad_norm": 2.25311086530556, "learning_rate": 9.999397791029788e-06, "loss": 0.5948, "step": 965 }, { "epoch": 0.3142485361093038, "grad_norm": 2.266734201729345, "learning_rate": 9.999368057364343e-06, "loss": 0.5819, "step": 966 }, { "epoch": 0.31457384515289527, "grad_norm": 1.975910213860415, "learning_rate": 9.99933760728612e-06, "loss": 0.5951, "step": 967 }, { "epoch": 0.3148991541964867, "grad_norm": 2.3549754523504447, "learning_rate": 9.999306440799485e-06, "loss": 0.5966, "step": 968 }, { "epoch": 0.3152244632400781, "grad_norm": 2.1545521591486314, "learning_rate": 9.999274557908906e-06, "loss": 0.6717, "step": 969 }, { "epoch": 0.3155497722836695, "grad_norm": 2.4632632699935137, "learning_rate": 9.999241958618948e-06, "loss": 0.7002, "step": 970 }, { "epoch": 0.3158750813272609, "grad_norm": 2.2982006023151595, "learning_rate": 9.999208642934285e-06, "loss": 0.6591, "step": 971 }, { "epoch": 0.3162003903708523, "grad_norm": 2.3358193491630024, "learning_rate": 9.99917461085969e-06, "loss": 0.6203, "step": 972 }, { "epoch": 0.3165256994144437, "grad_norm": 2.743650069791567, "learning_rate": 9.99913986240004e-06, "loss": 0.8541, "step": 973 }, { "epoch": 0.31685100845803515, "grad_norm": 2.1872181082615865, "learning_rate": 9.999104397560314e-06, "loss": 0.544, "step": 974 }, { "epoch": 0.31717631750162656, "grad_norm": 2.150008369359386, "learning_rate": 9.999068216345596e-06, "loss": 0.6181, "step": 975 }, { "epoch": 0.31750162654521796, "grad_norm": 2.4350302644372017, "learning_rate": 9.999031318761068e-06, "loss": 0.7207, "step": 976 }, { "epoch": 0.31782693558880937, "grad_norm": 2.6093444265288945, "learning_rate": 9.99899370481202e-06, "loss": 0.7943, "step": 977 }, { "epoch": 0.31815224463240077, "grad_norm": 2.1264837258000666, "learning_rate": 9.998955374503843e-06, "loss": 0.6892, "step": 978 }, { "epoch": 0.3184775536759922, "grad_norm": 1.8423403424775926, "learning_rate": 9.998916327842026e-06, "loss": 0.5522, "step": 979 }, { "epoch": 0.3188028627195836, "grad_norm": 2.0046350299736297, "learning_rate": 9.998876564832165e-06, "loss": 0.7113, "step": 980 }, { "epoch": 0.31912817176317504, "grad_norm": 2.291977136795208, "learning_rate": 9.998836085479961e-06, "loss": 0.6679, "step": 981 }, { "epoch": 0.31945348080676644, "grad_norm": 2.3915172333978245, "learning_rate": 9.998794889791213e-06, "loss": 0.7009, "step": 982 }, { "epoch": 0.31977878985035785, "grad_norm": 2.055429748016032, "learning_rate": 9.998752977771824e-06, "loss": 0.6175, "step": 983 }, { "epoch": 0.32010409889394925, "grad_norm": 2.4302071190153054, "learning_rate": 9.9987103494278e-06, "loss": 0.6954, "step": 984 }, { "epoch": 0.32042940793754066, "grad_norm": 2.0948945288392755, "learning_rate": 9.998667004765253e-06, "loss": 0.5983, "step": 985 }, { "epoch": 0.32075471698113206, "grad_norm": 2.238912150055718, "learning_rate": 9.998622943790388e-06, "loss": 0.5643, "step": 986 }, { "epoch": 0.32108002602472346, "grad_norm": 2.22833314493787, "learning_rate": 9.998578166509524e-06, "loss": 0.5753, "step": 987 }, { "epoch": 0.3214053350683149, "grad_norm": 1.7538774816843892, "learning_rate": 9.998532672929076e-06, "loss": 0.5872, "step": 988 }, { "epoch": 0.3217306441119063, "grad_norm": 2.091316344646663, "learning_rate": 9.998486463055563e-06, "loss": 0.5958, "step": 989 }, { "epoch": 0.32205595315549773, "grad_norm": 2.261877042916997, "learning_rate": 9.998439536895607e-06, "loss": 0.7028, "step": 990 }, { "epoch": 0.32238126219908914, "grad_norm": 2.028062295493333, "learning_rate": 9.998391894455933e-06, "loss": 0.6272, "step": 991 }, { "epoch": 0.32270657124268054, "grad_norm": 2.4209040121032697, "learning_rate": 9.998343535743369e-06, "loss": 0.6595, "step": 992 }, { "epoch": 0.32303188028627194, "grad_norm": 2.130778815141387, "learning_rate": 9.998294460764841e-06, "loss": 0.6414, "step": 993 }, { "epoch": 0.32335718932986335, "grad_norm": 2.084165451109852, "learning_rate": 9.998244669527387e-06, "loss": 0.6189, "step": 994 }, { "epoch": 0.3236824983734548, "grad_norm": 2.274634783735324, "learning_rate": 9.998194162038137e-06, "loss": 0.715, "step": 995 }, { "epoch": 0.3240078074170462, "grad_norm": 2.463906239941042, "learning_rate": 9.998142938304334e-06, "loss": 0.708, "step": 996 }, { "epoch": 0.3243331164606376, "grad_norm": 2.1761666558543222, "learning_rate": 9.998090998333314e-06, "loss": 0.6161, "step": 997 }, { "epoch": 0.324658425504229, "grad_norm": 2.5366097747831238, "learning_rate": 9.998038342132522e-06, "loss": 0.7974, "step": 998 }, { "epoch": 0.3249837345478204, "grad_norm": 2.0313812272128504, "learning_rate": 9.997984969709503e-06, "loss": 0.6266, "step": 999 }, { "epoch": 0.32530904359141183, "grad_norm": 2.295314551319269, "learning_rate": 9.997930881071905e-06, "loss": 0.6774, "step": 1000 }, { "epoch": 0.32563435263500323, "grad_norm": 1.902035074156549, "learning_rate": 9.997876076227479e-06, "loss": 0.5951, "step": 1001 }, { "epoch": 0.3259596616785947, "grad_norm": 2.488966206705957, "learning_rate": 9.99782055518408e-06, "loss": 0.752, "step": 1002 }, { "epoch": 0.3262849707221861, "grad_norm": 2.4537955143274113, "learning_rate": 9.997764317949662e-06, "loss": 0.6205, "step": 1003 }, { "epoch": 0.3266102797657775, "grad_norm": 2.274099338067784, "learning_rate": 9.997707364532287e-06, "loss": 0.6473, "step": 1004 }, { "epoch": 0.3269355888093689, "grad_norm": 2.098380052943401, "learning_rate": 9.997649694940114e-06, "loss": 0.6002, "step": 1005 }, { "epoch": 0.3272608978529603, "grad_norm": 2.1863424120093167, "learning_rate": 9.997591309181408e-06, "loss": 0.6566, "step": 1006 }, { "epoch": 0.3275862068965517, "grad_norm": 1.9835036995388817, "learning_rate": 9.997532207264535e-06, "loss": 0.5509, "step": 1007 }, { "epoch": 0.3279115159401431, "grad_norm": 2.1908375073476205, "learning_rate": 9.997472389197964e-06, "loss": 0.6434, "step": 1008 }, { "epoch": 0.3282368249837345, "grad_norm": 2.086778673375399, "learning_rate": 9.997411854990268e-06, "loss": 0.643, "step": 1009 }, { "epoch": 0.328562134027326, "grad_norm": 1.7936113942673866, "learning_rate": 9.997350604650123e-06, "loss": 0.561, "step": 1010 }, { "epoch": 0.3288874430709174, "grad_norm": 1.9259217245098066, "learning_rate": 9.997288638186302e-06, "loss": 0.6134, "step": 1011 }, { "epoch": 0.3292127521145088, "grad_norm": 2.059231131181774, "learning_rate": 9.997225955607689e-06, "loss": 0.6043, "step": 1012 }, { "epoch": 0.3295380611581002, "grad_norm": 2.144631987097279, "learning_rate": 9.997162556923264e-06, "loss": 0.6833, "step": 1013 }, { "epoch": 0.3298633702016916, "grad_norm": 2.1626081715253775, "learning_rate": 9.99709844214211e-06, "loss": 0.6155, "step": 1014 }, { "epoch": 0.330188679245283, "grad_norm": 2.219299657422334, "learning_rate": 9.997033611273422e-06, "loss": 0.6897, "step": 1015 }, { "epoch": 0.3305139882888744, "grad_norm": 2.044690144740596, "learning_rate": 9.996968064326484e-06, "loss": 0.6752, "step": 1016 }, { "epoch": 0.33083929733246586, "grad_norm": 2.0876137387077236, "learning_rate": 9.996901801310692e-06, "loss": 0.7345, "step": 1017 }, { "epoch": 0.33116460637605727, "grad_norm": 2.035177135205357, "learning_rate": 9.996834822235538e-06, "loss": 0.5986, "step": 1018 }, { "epoch": 0.3314899154196487, "grad_norm": 1.9830543880566753, "learning_rate": 9.996767127110623e-06, "loss": 0.5817, "step": 1019 }, { "epoch": 0.3318152244632401, "grad_norm": 2.108613120594639, "learning_rate": 9.996698715945649e-06, "loss": 0.6395, "step": 1020 }, { "epoch": 0.3321405335068315, "grad_norm": 2.0456446902786842, "learning_rate": 9.996629588750416e-06, "loss": 0.6657, "step": 1021 }, { "epoch": 0.3324658425504229, "grad_norm": 1.8096423235096388, "learning_rate": 9.996559745534832e-06, "loss": 0.5485, "step": 1022 }, { "epoch": 0.3327911515940143, "grad_norm": 2.197846982597601, "learning_rate": 9.996489186308905e-06, "loss": 0.6271, "step": 1023 }, { "epoch": 0.33311646063760575, "grad_norm": 2.513503452892213, "learning_rate": 9.996417911082746e-06, "loss": 0.7825, "step": 1024 }, { "epoch": 0.33344176968119715, "grad_norm": 2.420513564451033, "learning_rate": 9.99634591986657e-06, "loss": 0.6996, "step": 1025 }, { "epoch": 0.33376707872478856, "grad_norm": 2.088157069980611, "learning_rate": 9.99627321267069e-06, "loss": 0.646, "step": 1026 }, { "epoch": 0.33409238776837996, "grad_norm": 2.3362141311962965, "learning_rate": 9.99619978950553e-06, "loss": 0.6953, "step": 1027 }, { "epoch": 0.33441769681197137, "grad_norm": 2.3692127549553983, "learning_rate": 9.99612565038161e-06, "loss": 0.7262, "step": 1028 }, { "epoch": 0.33474300585556277, "grad_norm": 2.1168114463439966, "learning_rate": 9.99605079530955e-06, "loss": 0.6091, "step": 1029 }, { "epoch": 0.3350683148991542, "grad_norm": 2.142336828550073, "learning_rate": 9.995975224300083e-06, "loss": 0.6316, "step": 1030 }, { "epoch": 0.33539362394274563, "grad_norm": 1.8480089992925266, "learning_rate": 9.995898937364034e-06, "loss": 0.494, "step": 1031 }, { "epoch": 0.33571893298633704, "grad_norm": 1.9923267916129805, "learning_rate": 9.995821934512336e-06, "loss": 0.5594, "step": 1032 }, { "epoch": 0.33604424202992844, "grad_norm": 2.0325285703998786, "learning_rate": 9.995744215756024e-06, "loss": 0.6499, "step": 1033 }, { "epoch": 0.33636955107351985, "grad_norm": 2.1947503700359436, "learning_rate": 9.995665781106236e-06, "loss": 0.7095, "step": 1034 }, { "epoch": 0.33669486011711125, "grad_norm": 1.953356927161464, "learning_rate": 9.99558663057421e-06, "loss": 0.5726, "step": 1035 }, { "epoch": 0.33702016916070265, "grad_norm": 2.138988837987795, "learning_rate": 9.99550676417129e-06, "loss": 0.6814, "step": 1036 }, { "epoch": 0.33734547820429406, "grad_norm": 2.2186185828364695, "learning_rate": 9.995426181908918e-06, "loss": 0.6185, "step": 1037 }, { "epoch": 0.3376707872478855, "grad_norm": 1.9409285709183448, "learning_rate": 9.995344883798645e-06, "loss": 0.657, "step": 1038 }, { "epoch": 0.3379960962914769, "grad_norm": 2.327852608742101, "learning_rate": 9.99526286985212e-06, "loss": 0.6563, "step": 1039 }, { "epoch": 0.3383214053350683, "grad_norm": 2.127671658818656, "learning_rate": 9.995180140081095e-06, "loss": 0.6646, "step": 1040 }, { "epoch": 0.33864671437865973, "grad_norm": 2.5816600477131133, "learning_rate": 9.995096694497425e-06, "loss": 0.72, "step": 1041 }, { "epoch": 0.33897202342225113, "grad_norm": 2.392076180871714, "learning_rate": 9.995012533113068e-06, "loss": 0.6186, "step": 1042 }, { "epoch": 0.33929733246584254, "grad_norm": 2.042313303707614, "learning_rate": 9.994927655940085e-06, "loss": 0.5931, "step": 1043 }, { "epoch": 0.33962264150943394, "grad_norm": 2.6143931236295783, "learning_rate": 9.99484206299064e-06, "loss": 0.6801, "step": 1044 }, { "epoch": 0.33994795055302535, "grad_norm": 2.1729827825588313, "learning_rate": 9.994755754276997e-06, "loss": 0.6663, "step": 1045 }, { "epoch": 0.3402732595966168, "grad_norm": 2.3562664281509442, "learning_rate": 9.994668729811523e-06, "loss": 0.6895, "step": 1046 }, { "epoch": 0.3405985686402082, "grad_norm": 2.624237310553947, "learning_rate": 9.99458098960669e-06, "loss": 0.699, "step": 1047 }, { "epoch": 0.3409238776837996, "grad_norm": 2.2821418092753856, "learning_rate": 9.99449253367507e-06, "loss": 0.5872, "step": 1048 }, { "epoch": 0.341249186727391, "grad_norm": 2.091048520930278, "learning_rate": 9.994403362029341e-06, "loss": 0.608, "step": 1049 }, { "epoch": 0.3415744957709824, "grad_norm": 2.3335116359848262, "learning_rate": 9.994313474682282e-06, "loss": 0.711, "step": 1050 }, { "epoch": 0.3418998048145738, "grad_norm": 2.1916098349713122, "learning_rate": 9.994222871646771e-06, "loss": 0.717, "step": 1051 }, { "epoch": 0.34222511385816523, "grad_norm": 2.1637006865736392, "learning_rate": 9.994131552935793e-06, "loss": 0.7096, "step": 1052 }, { "epoch": 0.3425504229017567, "grad_norm": 1.9390173061795433, "learning_rate": 9.994039518562433e-06, "loss": 0.4698, "step": 1053 }, { "epoch": 0.3428757319453481, "grad_norm": 2.2405133952223646, "learning_rate": 9.99394676853988e-06, "loss": 0.6616, "step": 1054 }, { "epoch": 0.3432010409889395, "grad_norm": 1.9176107920527838, "learning_rate": 9.993853302881427e-06, "loss": 0.5635, "step": 1055 }, { "epoch": 0.3435263500325309, "grad_norm": 2.264928683790364, "learning_rate": 9.993759121600464e-06, "loss": 0.6534, "step": 1056 }, { "epoch": 0.3438516590761223, "grad_norm": 2.1834794383205085, "learning_rate": 9.99366422471049e-06, "loss": 0.6723, "step": 1057 }, { "epoch": 0.3441769681197137, "grad_norm": 2.2719308673851306, "learning_rate": 9.993568612225104e-06, "loss": 0.7689, "step": 1058 }, { "epoch": 0.3445022771633051, "grad_norm": 2.623380798807266, "learning_rate": 9.993472284158007e-06, "loss": 0.6411, "step": 1059 }, { "epoch": 0.3448275862068966, "grad_norm": 2.1089964429409074, "learning_rate": 9.993375240523003e-06, "loss": 0.5493, "step": 1060 }, { "epoch": 0.345152895250488, "grad_norm": 1.962521672763483, "learning_rate": 9.993277481333996e-06, "loss": 0.6171, "step": 1061 }, { "epoch": 0.3454782042940794, "grad_norm": 2.0440065654480355, "learning_rate": 9.993179006604996e-06, "loss": 0.5901, "step": 1062 }, { "epoch": 0.3458035133376708, "grad_norm": 2.310201670758111, "learning_rate": 9.993079816350117e-06, "loss": 0.643, "step": 1063 }, { "epoch": 0.3461288223812622, "grad_norm": 2.1658205185028234, "learning_rate": 9.992979910583571e-06, "loss": 0.7551, "step": 1064 }, { "epoch": 0.3464541314248536, "grad_norm": 2.640464051782944, "learning_rate": 9.992879289319675e-06, "loss": 0.7271, "step": 1065 }, { "epoch": 0.346779440468445, "grad_norm": 1.923634863709433, "learning_rate": 9.992777952572847e-06, "loss": 0.6146, "step": 1066 }, { "epoch": 0.34710474951203646, "grad_norm": 2.277436982213218, "learning_rate": 9.99267590035761e-06, "loss": 0.6523, "step": 1067 }, { "epoch": 0.34743005855562786, "grad_norm": 2.1820904270570773, "learning_rate": 9.992573132688588e-06, "loss": 0.6033, "step": 1068 }, { "epoch": 0.34775536759921927, "grad_norm": 2.001468938948523, "learning_rate": 9.992469649580507e-06, "loss": 0.5528, "step": 1069 }, { "epoch": 0.34808067664281067, "grad_norm": 1.8308858565049386, "learning_rate": 9.992365451048197e-06, "loss": 0.5515, "step": 1070 }, { "epoch": 0.3484059856864021, "grad_norm": 2.4174295166476245, "learning_rate": 9.992260537106588e-06, "loss": 0.6613, "step": 1071 }, { "epoch": 0.3487312947299935, "grad_norm": 2.6551175032261516, "learning_rate": 9.992154907770717e-06, "loss": 0.771, "step": 1072 }, { "epoch": 0.3490566037735849, "grad_norm": 1.9470951054698868, "learning_rate": 9.992048563055719e-06, "loss": 0.5966, "step": 1073 }, { "epoch": 0.34938191281717634, "grad_norm": 2.2915431782882405, "learning_rate": 9.991941502976835e-06, "loss": 0.6224, "step": 1074 }, { "epoch": 0.34970722186076775, "grad_norm": 2.26095667665187, "learning_rate": 9.991833727549402e-06, "loss": 0.7078, "step": 1075 }, { "epoch": 0.35003253090435915, "grad_norm": 1.8512798469454623, "learning_rate": 9.991725236788871e-06, "loss": 0.5513, "step": 1076 }, { "epoch": 0.35035783994795056, "grad_norm": 2.020914582107366, "learning_rate": 9.991616030710784e-06, "loss": 0.5842, "step": 1077 }, { "epoch": 0.35068314899154196, "grad_norm": 2.6820497575181896, "learning_rate": 9.99150610933079e-06, "loss": 0.7215, "step": 1078 }, { "epoch": 0.35100845803513336, "grad_norm": 2.1180497848774134, "learning_rate": 9.991395472664646e-06, "loss": 0.6115, "step": 1079 }, { "epoch": 0.35133376707872477, "grad_norm": 2.3860091227791727, "learning_rate": 9.9912841207282e-06, "loss": 0.756, "step": 1080 }, { "epoch": 0.3516590761223162, "grad_norm": 2.0233791141912207, "learning_rate": 9.991172053537414e-06, "loss": 0.6071, "step": 1081 }, { "epoch": 0.35198438516590763, "grad_norm": 2.2217851557825363, "learning_rate": 9.991059271108342e-06, "loss": 0.6415, "step": 1082 }, { "epoch": 0.35230969420949904, "grad_norm": 2.182353182432987, "learning_rate": 9.990945773457152e-06, "loss": 0.6097, "step": 1083 }, { "epoch": 0.35263500325309044, "grad_norm": 2.335593809519379, "learning_rate": 9.990831560600102e-06, "loss": 0.6438, "step": 1084 }, { "epoch": 0.35296031229668184, "grad_norm": 2.0030310506223543, "learning_rate": 9.990716632553566e-06, "loss": 0.6279, "step": 1085 }, { "epoch": 0.35328562134027325, "grad_norm": 2.293651788430977, "learning_rate": 9.990600989334005e-06, "loss": 0.7248, "step": 1086 }, { "epoch": 0.35361093038386465, "grad_norm": 1.8942637365118526, "learning_rate": 9.990484630957997e-06, "loss": 0.5966, "step": 1087 }, { "epoch": 0.35393623942745606, "grad_norm": 1.9311098104517699, "learning_rate": 9.990367557442213e-06, "loss": 0.5816, "step": 1088 }, { "epoch": 0.3542615484710475, "grad_norm": 1.9622747644919438, "learning_rate": 9.990249768803431e-06, "loss": 0.5723, "step": 1089 }, { "epoch": 0.3545868575146389, "grad_norm": 1.8406576242914938, "learning_rate": 9.99013126505853e-06, "loss": 0.5313, "step": 1090 }, { "epoch": 0.3549121665582303, "grad_norm": 2.046777196272402, "learning_rate": 9.990012046224492e-06, "loss": 0.6487, "step": 1091 }, { "epoch": 0.35523747560182173, "grad_norm": 2.1498959757756637, "learning_rate": 9.9898921123184e-06, "loss": 0.6534, "step": 1092 }, { "epoch": 0.35556278464541313, "grad_norm": 2.0994935449100653, "learning_rate": 9.989771463357442e-06, "loss": 0.6153, "step": 1093 }, { "epoch": 0.35588809368900454, "grad_norm": 2.232978702760988, "learning_rate": 9.989650099358908e-06, "loss": 0.6986, "step": 1094 }, { "epoch": 0.35621340273259594, "grad_norm": 2.1697244008634446, "learning_rate": 9.989528020340186e-06, "loss": 0.6325, "step": 1095 }, { "epoch": 0.3565387117761874, "grad_norm": 2.116654424841932, "learning_rate": 9.989405226318772e-06, "loss": 0.64, "step": 1096 }, { "epoch": 0.3568640208197788, "grad_norm": 1.9863667988453146, "learning_rate": 9.989281717312263e-06, "loss": 0.7083, "step": 1097 }, { "epoch": 0.3571893298633702, "grad_norm": 2.212682193826147, "learning_rate": 9.989157493338356e-06, "loss": 0.6919, "step": 1098 }, { "epoch": 0.3575146389069616, "grad_norm": 1.8035951032094144, "learning_rate": 9.989032554414858e-06, "loss": 0.567, "step": 1099 }, { "epoch": 0.357839947950553, "grad_norm": 2.1608374013557703, "learning_rate": 9.988906900559665e-06, "loss": 0.6316, "step": 1100 }, { "epoch": 0.3581652569941444, "grad_norm": 1.8702391887219345, "learning_rate": 9.988780531790786e-06, "loss": 0.5938, "step": 1101 }, { "epoch": 0.3584905660377358, "grad_norm": 2.581173854371041, "learning_rate": 9.988653448126333e-06, "loss": 0.7573, "step": 1102 }, { "epoch": 0.3588158750813273, "grad_norm": 2.376560805670364, "learning_rate": 9.988525649584515e-06, "loss": 0.5928, "step": 1103 }, { "epoch": 0.3591411841249187, "grad_norm": 2.240830723370489, "learning_rate": 9.988397136183646e-06, "loss": 0.6237, "step": 1104 }, { "epoch": 0.3594664931685101, "grad_norm": 2.332416402466616, "learning_rate": 9.988267907942138e-06, "loss": 0.835, "step": 1105 }, { "epoch": 0.3597918022121015, "grad_norm": 2.5747006774747474, "learning_rate": 9.988137964878517e-06, "loss": 0.7075, "step": 1106 }, { "epoch": 0.3601171112556929, "grad_norm": 2.1175832797563636, "learning_rate": 9.988007307011395e-06, "loss": 0.5859, "step": 1107 }, { "epoch": 0.3604424202992843, "grad_norm": 2.376741665738186, "learning_rate": 9.987875934359506e-06, "loss": 0.748, "step": 1108 }, { "epoch": 0.3607677293428757, "grad_norm": 2.105549599837769, "learning_rate": 9.987743846941666e-06, "loss": 0.6806, "step": 1109 }, { "epoch": 0.36109303838646717, "grad_norm": 2.237413385207724, "learning_rate": 9.987611044776809e-06, "loss": 0.5266, "step": 1110 }, { "epoch": 0.3614183474300586, "grad_norm": 2.1791273692633157, "learning_rate": 9.987477527883963e-06, "loss": 0.6022, "step": 1111 }, { "epoch": 0.36174365647365, "grad_norm": 2.3546753642244815, "learning_rate": 9.987343296282261e-06, "loss": 0.6765, "step": 1112 }, { "epoch": 0.3620689655172414, "grad_norm": 2.3116184403624453, "learning_rate": 9.987208349990943e-06, "loss": 0.6556, "step": 1113 }, { "epoch": 0.3623942745608328, "grad_norm": 2.077646570537694, "learning_rate": 9.98707268902934e-06, "loss": 0.6966, "step": 1114 }, { "epoch": 0.3627195836044242, "grad_norm": 1.994065733855817, "learning_rate": 9.986936313416898e-06, "loss": 0.7037, "step": 1115 }, { "epoch": 0.3630448926480156, "grad_norm": 2.078329418954455, "learning_rate": 9.986799223173154e-06, "loss": 0.5693, "step": 1116 }, { "epoch": 0.36337020169160705, "grad_norm": 1.8402108924471003, "learning_rate": 9.986661418317759e-06, "loss": 0.5478, "step": 1117 }, { "epoch": 0.36369551073519846, "grad_norm": 2.1284042738121145, "learning_rate": 9.986522898870458e-06, "loss": 0.6733, "step": 1118 }, { "epoch": 0.36402081977878986, "grad_norm": 1.9799751252831503, "learning_rate": 9.9863836648511e-06, "loss": 0.6203, "step": 1119 }, { "epoch": 0.36434612882238127, "grad_norm": 2.276875476814628, "learning_rate": 9.98624371627964e-06, "loss": 0.7482, "step": 1120 }, { "epoch": 0.36467143786597267, "grad_norm": 2.036357457019278, "learning_rate": 9.986103053176128e-06, "loss": 0.6396, "step": 1121 }, { "epoch": 0.3649967469095641, "grad_norm": 2.3939163908528887, "learning_rate": 9.985961675560726e-06, "loss": 0.6329, "step": 1122 }, { "epoch": 0.3653220559531555, "grad_norm": 1.9416536433274498, "learning_rate": 9.985819583453693e-06, "loss": 0.596, "step": 1123 }, { "epoch": 0.3656473649967469, "grad_norm": 1.8372565753732466, "learning_rate": 9.985676776875386e-06, "loss": 0.5222, "step": 1124 }, { "epoch": 0.36597267404033834, "grad_norm": 2.235766562188749, "learning_rate": 9.985533255846276e-06, "loss": 0.7089, "step": 1125 }, { "epoch": 0.36629798308392975, "grad_norm": 2.3410323529591603, "learning_rate": 9.985389020386924e-06, "loss": 0.7174, "step": 1126 }, { "epoch": 0.36662329212752115, "grad_norm": 2.2801794648697933, "learning_rate": 9.985244070518003e-06, "loss": 0.7619, "step": 1127 }, { "epoch": 0.36694860117111255, "grad_norm": 2.0589662359532195, "learning_rate": 9.985098406260283e-06, "loss": 0.6177, "step": 1128 }, { "epoch": 0.36727391021470396, "grad_norm": 2.016444707803789, "learning_rate": 9.984952027634637e-06, "loss": 0.6194, "step": 1129 }, { "epoch": 0.36759921925829536, "grad_norm": 2.052348930846168, "learning_rate": 9.984804934662042e-06, "loss": 0.5608, "step": 1130 }, { "epoch": 0.36792452830188677, "grad_norm": 1.9493246271063085, "learning_rate": 9.984657127363576e-06, "loss": 0.6295, "step": 1131 }, { "epoch": 0.3682498373454782, "grad_norm": 2.0876484107001647, "learning_rate": 9.984508605760421e-06, "loss": 0.6659, "step": 1132 }, { "epoch": 0.36857514638906963, "grad_norm": 1.901638410670997, "learning_rate": 9.984359369873858e-06, "loss": 0.5713, "step": 1133 }, { "epoch": 0.36890045543266103, "grad_norm": 1.9827075578202178, "learning_rate": 9.984209419725274e-06, "loss": 0.5827, "step": 1134 }, { "epoch": 0.36922576447625244, "grad_norm": 2.1237259278844527, "learning_rate": 9.984058755336159e-06, "loss": 0.6769, "step": 1135 }, { "epoch": 0.36955107351984384, "grad_norm": 1.9602743764588693, "learning_rate": 9.983907376728102e-06, "loss": 0.5494, "step": 1136 }, { "epoch": 0.36987638256343525, "grad_norm": 2.4231698012863667, "learning_rate": 9.983755283922792e-06, "loss": 0.7405, "step": 1137 }, { "epoch": 0.37020169160702665, "grad_norm": 2.1592410414541092, "learning_rate": 9.983602476942029e-06, "loss": 0.5497, "step": 1138 }, { "epoch": 0.3705270006506181, "grad_norm": 2.0968433444228207, "learning_rate": 9.983448955807708e-06, "loss": 0.6133, "step": 1139 }, { "epoch": 0.3708523096942095, "grad_norm": 2.05993167773475, "learning_rate": 9.98329472054183e-06, "loss": 0.4846, "step": 1140 }, { "epoch": 0.3711776187378009, "grad_norm": 2.2328025461853103, "learning_rate": 9.983139771166494e-06, "loss": 0.7274, "step": 1141 }, { "epoch": 0.3715029277813923, "grad_norm": 2.055094992298178, "learning_rate": 9.982984107703907e-06, "loss": 0.6065, "step": 1142 }, { "epoch": 0.3718282368249837, "grad_norm": 2.187618405310842, "learning_rate": 9.982827730176374e-06, "loss": 0.6563, "step": 1143 }, { "epoch": 0.37215354586857513, "grad_norm": 2.2887443339339466, "learning_rate": 9.982670638606308e-06, "loss": 0.7599, "step": 1144 }, { "epoch": 0.37247885491216653, "grad_norm": 1.9540209050040795, "learning_rate": 9.982512833016216e-06, "loss": 0.6567, "step": 1145 }, { "epoch": 0.372804163955758, "grad_norm": 2.0525104527932485, "learning_rate": 9.982354313428713e-06, "loss": 0.5845, "step": 1146 }, { "epoch": 0.3731294729993494, "grad_norm": 2.140692858758069, "learning_rate": 9.982195079866515e-06, "loss": 0.7393, "step": 1147 }, { "epoch": 0.3734547820429408, "grad_norm": 2.208064418732068, "learning_rate": 9.98203513235244e-06, "loss": 0.6915, "step": 1148 }, { "epoch": 0.3737800910865322, "grad_norm": 1.9486931774305551, "learning_rate": 9.981874470909408e-06, "loss": 0.6182, "step": 1149 }, { "epoch": 0.3741054001301236, "grad_norm": 2.187969552246943, "learning_rate": 9.981713095560443e-06, "loss": 0.7614, "step": 1150 }, { "epoch": 0.374430709173715, "grad_norm": 1.9480296518134428, "learning_rate": 9.981551006328671e-06, "loss": 0.5636, "step": 1151 }, { "epoch": 0.3747560182173064, "grad_norm": 2.054711688883067, "learning_rate": 9.981388203237319e-06, "loss": 0.636, "step": 1152 }, { "epoch": 0.3750813272608979, "grad_norm": 2.1941357326783004, "learning_rate": 9.981224686309714e-06, "loss": 0.6397, "step": 1153 }, { "epoch": 0.3754066363044893, "grad_norm": 2.1954571417466204, "learning_rate": 9.981060455569292e-06, "loss": 0.6345, "step": 1154 }, { "epoch": 0.3757319453480807, "grad_norm": 1.8806759401911322, "learning_rate": 9.980895511039586e-06, "loss": 0.6122, "step": 1155 }, { "epoch": 0.3760572543916721, "grad_norm": 2.036140179743051, "learning_rate": 9.980729852744231e-06, "loss": 0.7364, "step": 1156 }, { "epoch": 0.3763825634352635, "grad_norm": 1.7662881421698788, "learning_rate": 9.980563480706969e-06, "loss": 0.5204, "step": 1157 }, { "epoch": 0.3767078724788549, "grad_norm": 1.9969196614396876, "learning_rate": 9.980396394951639e-06, "loss": 0.6533, "step": 1158 }, { "epoch": 0.3770331815224463, "grad_norm": 1.8564537996988628, "learning_rate": 9.980228595502183e-06, "loss": 0.6087, "step": 1159 }, { "epoch": 0.37735849056603776, "grad_norm": 2.441984368831454, "learning_rate": 9.980060082382652e-06, "loss": 0.7466, "step": 1160 }, { "epoch": 0.37768379960962917, "grad_norm": 2.0839415110998747, "learning_rate": 9.979890855617188e-06, "loss": 0.6811, "step": 1161 }, { "epoch": 0.37800910865322057, "grad_norm": 1.9159030616287933, "learning_rate": 9.979720915230045e-06, "loss": 0.5332, "step": 1162 }, { "epoch": 0.378334417696812, "grad_norm": 2.282120798754863, "learning_rate": 9.979550261245576e-06, "loss": 0.552, "step": 1163 }, { "epoch": 0.3786597267404034, "grad_norm": 2.119271549361588, "learning_rate": 9.979378893688235e-06, "loss": 0.6236, "step": 1164 }, { "epoch": 0.3789850357839948, "grad_norm": 2.1254642275071163, "learning_rate": 9.979206812582576e-06, "loss": 0.5318, "step": 1165 }, { "epoch": 0.3793103448275862, "grad_norm": 2.228884889434015, "learning_rate": 9.979034017953262e-06, "loss": 0.6557, "step": 1166 }, { "epoch": 0.3796356538711776, "grad_norm": 2.065973018235971, "learning_rate": 9.978860509825054e-06, "loss": 0.6657, "step": 1167 }, { "epoch": 0.37996096291476905, "grad_norm": 1.7576496949212035, "learning_rate": 9.978686288222815e-06, "loss": 0.591, "step": 1168 }, { "epoch": 0.38028627195836046, "grad_norm": 2.1834790757612743, "learning_rate": 9.978511353171512e-06, "loss": 0.5778, "step": 1169 }, { "epoch": 0.38061158100195186, "grad_norm": 2.285696587552193, "learning_rate": 9.978335704696211e-06, "loss": 0.6698, "step": 1170 }, { "epoch": 0.38093689004554326, "grad_norm": 2.0030044069224084, "learning_rate": 9.978159342822088e-06, "loss": 0.6335, "step": 1171 }, { "epoch": 0.38126219908913467, "grad_norm": 2.4597832205446313, "learning_rate": 9.977982267574409e-06, "loss": 0.6798, "step": 1172 }, { "epoch": 0.38158750813272607, "grad_norm": 2.1294870323999464, "learning_rate": 9.977804478978552e-06, "loss": 0.601, "step": 1173 }, { "epoch": 0.3819128171763175, "grad_norm": 2.1655372768770444, "learning_rate": 9.977625977059995e-06, "loss": 0.6777, "step": 1174 }, { "epoch": 0.38223812621990894, "grad_norm": 1.8565427794623173, "learning_rate": 9.977446761844317e-06, "loss": 0.5655, "step": 1175 }, { "epoch": 0.38256343526350034, "grad_norm": 2.2540169774533316, "learning_rate": 9.9772668333572e-06, "loss": 0.6535, "step": 1176 }, { "epoch": 0.38288874430709174, "grad_norm": 1.8172735203994443, "learning_rate": 9.977086191624426e-06, "loss": 0.5865, "step": 1177 }, { "epoch": 0.38321405335068315, "grad_norm": 2.1898437010051333, "learning_rate": 9.976904836671883e-06, "loss": 0.6823, "step": 1178 }, { "epoch": 0.38353936239427455, "grad_norm": 2.4201855236792333, "learning_rate": 9.97672276852556e-06, "loss": 0.7871, "step": 1179 }, { "epoch": 0.38386467143786596, "grad_norm": 1.9700807181425484, "learning_rate": 9.976539987211544e-06, "loss": 0.514, "step": 1180 }, { "epoch": 0.38418998048145736, "grad_norm": 2.2814393746046226, "learning_rate": 9.976356492756031e-06, "loss": 0.6994, "step": 1181 }, { "epoch": 0.3845152895250488, "grad_norm": 2.2958488721530133, "learning_rate": 9.976172285185315e-06, "loss": 0.6615, "step": 1182 }, { "epoch": 0.3848405985686402, "grad_norm": 2.286408089393525, "learning_rate": 9.975987364525793e-06, "loss": 0.7124, "step": 1183 }, { "epoch": 0.38516590761223163, "grad_norm": 2.0339158162343316, "learning_rate": 9.975801730803962e-06, "loss": 0.5778, "step": 1184 }, { "epoch": 0.38549121665582303, "grad_norm": 2.276593333667505, "learning_rate": 9.975615384046426e-06, "loss": 0.6322, "step": 1185 }, { "epoch": 0.38581652569941444, "grad_norm": 1.9572831727873077, "learning_rate": 9.97542832427989e-06, "loss": 0.6503, "step": 1186 }, { "epoch": 0.38614183474300584, "grad_norm": 2.2284259739980663, "learning_rate": 9.975240551531155e-06, "loss": 0.6117, "step": 1187 }, { "epoch": 0.38646714378659724, "grad_norm": 1.9032910010057196, "learning_rate": 9.975052065827134e-06, "loss": 0.6358, "step": 1188 }, { "epoch": 0.3867924528301887, "grad_norm": 2.3587380741633805, "learning_rate": 9.974862867194835e-06, "loss": 0.5233, "step": 1189 }, { "epoch": 0.3871177618737801, "grad_norm": 1.6943216250104685, "learning_rate": 9.974672955661368e-06, "loss": 0.5458, "step": 1190 }, { "epoch": 0.3874430709173715, "grad_norm": 2.162782439531007, "learning_rate": 9.974482331253952e-06, "loss": 0.5848, "step": 1191 }, { "epoch": 0.3877683799609629, "grad_norm": 1.9116029725018557, "learning_rate": 9.9742909939999e-06, "loss": 0.5882, "step": 1192 }, { "epoch": 0.3880936890045543, "grad_norm": 1.9049709412307885, "learning_rate": 9.974098943926633e-06, "loss": 0.5777, "step": 1193 }, { "epoch": 0.3884189980481457, "grad_norm": 2.268407118582987, "learning_rate": 9.97390618106167e-06, "loss": 0.7054, "step": 1194 }, { "epoch": 0.38874430709173713, "grad_norm": 1.928057789869351, "learning_rate": 9.973712705432637e-06, "loss": 0.6427, "step": 1195 }, { "epoch": 0.3890696161353286, "grad_norm": 2.2706932172593195, "learning_rate": 9.973518517067256e-06, "loss": 0.8194, "step": 1196 }, { "epoch": 0.38939492517892, "grad_norm": 2.089089866343163, "learning_rate": 9.973323615993355e-06, "loss": 0.6276, "step": 1197 }, { "epoch": 0.3897202342225114, "grad_norm": 2.1260645692083306, "learning_rate": 9.973128002238865e-06, "loss": 0.629, "step": 1198 }, { "epoch": 0.3900455432661028, "grad_norm": 2.3613509876097853, "learning_rate": 9.972931675831816e-06, "loss": 0.6704, "step": 1199 }, { "epoch": 0.3903708523096942, "grad_norm": 2.056783287129113, "learning_rate": 9.97273463680034e-06, "loss": 0.5885, "step": 1200 }, { "epoch": 0.3906961613532856, "grad_norm": 1.9563699510750752, "learning_rate": 9.97253688517268e-06, "loss": 0.6699, "step": 1201 }, { "epoch": 0.391021470396877, "grad_norm": 2.326328212232434, "learning_rate": 9.972338420977164e-06, "loss": 0.7026, "step": 1202 }, { "epoch": 0.3913467794404684, "grad_norm": 2.285119753933177, "learning_rate": 9.97213924424224e-06, "loss": 0.6523, "step": 1203 }, { "epoch": 0.3916720884840599, "grad_norm": 2.1897100914405296, "learning_rate": 9.971939354996447e-06, "loss": 0.609, "step": 1204 }, { "epoch": 0.3919973975276513, "grad_norm": 2.0439964476841808, "learning_rate": 9.971738753268428e-06, "loss": 0.6372, "step": 1205 }, { "epoch": 0.3923227065712427, "grad_norm": 2.202992429084596, "learning_rate": 9.971537439086932e-06, "loss": 0.7743, "step": 1206 }, { "epoch": 0.3926480156148341, "grad_norm": 1.9442072808849995, "learning_rate": 9.971335412480806e-06, "loss": 0.5994, "step": 1207 }, { "epoch": 0.3929733246584255, "grad_norm": 1.9512214042953886, "learning_rate": 9.971132673479001e-06, "loss": 0.6182, "step": 1208 }, { "epoch": 0.3932986337020169, "grad_norm": 1.84786992002797, "learning_rate": 9.970929222110571e-06, "loss": 0.6001, "step": 1209 }, { "epoch": 0.3936239427456083, "grad_norm": 2.0025004675029505, "learning_rate": 9.970725058404666e-06, "loss": 0.6544, "step": 1210 }, { "epoch": 0.39394925178919976, "grad_norm": 1.9575222154487144, "learning_rate": 9.970520182390548e-06, "loss": 0.5629, "step": 1211 }, { "epoch": 0.39427456083279117, "grad_norm": 1.879360985902343, "learning_rate": 9.970314594097574e-06, "loss": 0.6204, "step": 1212 }, { "epoch": 0.39459986987638257, "grad_norm": 1.9033877630961498, "learning_rate": 9.970108293555204e-06, "loss": 0.5353, "step": 1213 }, { "epoch": 0.394925178919974, "grad_norm": 2.3112637026282807, "learning_rate": 9.969901280793001e-06, "loss": 0.6795, "step": 1214 }, { "epoch": 0.3952504879635654, "grad_norm": 2.211617577147442, "learning_rate": 9.969693555840635e-06, "loss": 0.7536, "step": 1215 }, { "epoch": 0.3955757970071568, "grad_norm": 1.9728080285214016, "learning_rate": 9.969485118727863e-06, "loss": 0.6064, "step": 1216 }, { "epoch": 0.3959011060507482, "grad_norm": 2.2623157452458624, "learning_rate": 9.969275969484565e-06, "loss": 0.687, "step": 1217 }, { "epoch": 0.39622641509433965, "grad_norm": 2.0644732902779133, "learning_rate": 9.969066108140706e-06, "loss": 0.7163, "step": 1218 }, { "epoch": 0.39655172413793105, "grad_norm": 1.9373377482034324, "learning_rate": 9.968855534726359e-06, "loss": 0.6103, "step": 1219 }, { "epoch": 0.39687703318152245, "grad_norm": 2.0765587302601873, "learning_rate": 9.968644249271703e-06, "loss": 0.6791, "step": 1220 }, { "epoch": 0.39720234222511386, "grad_norm": 2.2781656553082987, "learning_rate": 9.968432251807012e-06, "loss": 0.6636, "step": 1221 }, { "epoch": 0.39752765126870526, "grad_norm": 2.2641785880131122, "learning_rate": 9.968219542362668e-06, "loss": 0.6612, "step": 1222 }, { "epoch": 0.39785296031229667, "grad_norm": 1.9866366064218992, "learning_rate": 9.96800612096915e-06, "loss": 0.6938, "step": 1223 }, { "epoch": 0.39817826935588807, "grad_norm": 1.8901453310457932, "learning_rate": 9.967791987657042e-06, "loss": 0.642, "step": 1224 }, { "epoch": 0.39850357839947953, "grad_norm": 2.3960464403517117, "learning_rate": 9.967577142457031e-06, "loss": 0.6805, "step": 1225 }, { "epoch": 0.39882888744307093, "grad_norm": 2.210007846115566, "learning_rate": 9.967361585399905e-06, "loss": 0.7566, "step": 1226 }, { "epoch": 0.39915419648666234, "grad_norm": 2.077392571984756, "learning_rate": 9.96714531651655e-06, "loss": 0.5839, "step": 1227 }, { "epoch": 0.39947950553025374, "grad_norm": 1.839009853475656, "learning_rate": 9.96692833583796e-06, "loss": 0.6344, "step": 1228 }, { "epoch": 0.39980481457384515, "grad_norm": 2.211406618881842, "learning_rate": 9.966710643395227e-06, "loss": 0.813, "step": 1229 }, { "epoch": 0.40013012361743655, "grad_norm": 2.0582328013036393, "learning_rate": 9.96649223921955e-06, "loss": 0.6447, "step": 1230 }, { "epoch": 0.40045543266102795, "grad_norm": 1.7227483779346788, "learning_rate": 9.96627312334222e-06, "loss": 0.5419, "step": 1231 }, { "epoch": 0.4007807417046194, "grad_norm": 1.9894288444421457, "learning_rate": 9.966053295794643e-06, "loss": 0.7317, "step": 1232 }, { "epoch": 0.4011060507482108, "grad_norm": 2.1331113627800953, "learning_rate": 9.965832756608317e-06, "loss": 0.696, "step": 1233 }, { "epoch": 0.4014313597918022, "grad_norm": 2.1032776186572164, "learning_rate": 9.965611505814845e-06, "loss": 0.6494, "step": 1234 }, { "epoch": 0.4017566688353936, "grad_norm": 1.7196803234095546, "learning_rate": 9.965389543445935e-06, "loss": 0.5726, "step": 1235 }, { "epoch": 0.40208197787898503, "grad_norm": 2.113804389505844, "learning_rate": 9.965166869533393e-06, "loss": 0.6608, "step": 1236 }, { "epoch": 0.40240728692257643, "grad_norm": 2.0824557529910197, "learning_rate": 9.964943484109126e-06, "loss": 0.6534, "step": 1237 }, { "epoch": 0.40273259596616784, "grad_norm": 1.9745297875718337, "learning_rate": 9.96471938720515e-06, "loss": 0.5882, "step": 1238 }, { "epoch": 0.4030579050097593, "grad_norm": 2.0655975365539425, "learning_rate": 9.964494578853574e-06, "loss": 0.7145, "step": 1239 }, { "epoch": 0.4033832140533507, "grad_norm": 2.0980927021985925, "learning_rate": 9.964269059086615e-06, "loss": 0.6525, "step": 1240 }, { "epoch": 0.4037085230969421, "grad_norm": 2.234538485623777, "learning_rate": 9.964042827936591e-06, "loss": 0.6176, "step": 1241 }, { "epoch": 0.4040338321405335, "grad_norm": 2.0635657293753966, "learning_rate": 9.96381588543592e-06, "loss": 0.5997, "step": 1242 }, { "epoch": 0.4043591411841249, "grad_norm": 2.1051873030493815, "learning_rate": 9.963588231617124e-06, "loss": 0.4827, "step": 1243 }, { "epoch": 0.4046844502277163, "grad_norm": 2.8257529188852417, "learning_rate": 9.963359866512823e-06, "loss": 0.5108, "step": 1244 }, { "epoch": 0.4050097592713077, "grad_norm": 2.0121602164401042, "learning_rate": 9.963130790155744e-06, "loss": 0.6742, "step": 1245 }, { "epoch": 0.4053350683148991, "grad_norm": 2.2187357717693574, "learning_rate": 9.962901002578714e-06, "loss": 0.6511, "step": 1246 }, { "epoch": 0.4056603773584906, "grad_norm": 2.0498144155848985, "learning_rate": 9.962670503814662e-06, "loss": 0.574, "step": 1247 }, { "epoch": 0.405985686402082, "grad_norm": 2.0497083026688196, "learning_rate": 9.962439293896617e-06, "loss": 0.682, "step": 1248 }, { "epoch": 0.4063109954456734, "grad_norm": 2.1861499975252365, "learning_rate": 9.962207372857715e-06, "loss": 0.6331, "step": 1249 }, { "epoch": 0.4066363044892648, "grad_norm": 2.1343435988084463, "learning_rate": 9.961974740731186e-06, "loss": 0.5605, "step": 1250 }, { "epoch": 0.4069616135328562, "grad_norm": 2.4118505687063707, "learning_rate": 9.961741397550369e-06, "loss": 0.6557, "step": 1251 }, { "epoch": 0.4072869225764476, "grad_norm": 1.9021493609461215, "learning_rate": 9.9615073433487e-06, "loss": 0.5395, "step": 1252 }, { "epoch": 0.407612231620039, "grad_norm": 2.0298604799985878, "learning_rate": 9.961272578159723e-06, "loss": 0.642, "step": 1253 }, { "epoch": 0.40793754066363047, "grad_norm": 2.141435704699988, "learning_rate": 9.961037102017076e-06, "loss": 0.6539, "step": 1254 }, { "epoch": 0.4082628497072219, "grad_norm": 1.7797936163927282, "learning_rate": 9.960800914954506e-06, "loss": 0.51, "step": 1255 }, { "epoch": 0.4085881587508133, "grad_norm": 1.9768642817579674, "learning_rate": 9.960564017005857e-06, "loss": 0.4723, "step": 1256 }, { "epoch": 0.4089134677944047, "grad_norm": 2.1269453505003333, "learning_rate": 9.960326408205077e-06, "loss": 0.6274, "step": 1257 }, { "epoch": 0.4092387768379961, "grad_norm": 2.275042256208475, "learning_rate": 9.960088088586214e-06, "loss": 0.6886, "step": 1258 }, { "epoch": 0.4095640858815875, "grad_norm": 1.986839316807896, "learning_rate": 9.959849058183423e-06, "loss": 0.5755, "step": 1259 }, { "epoch": 0.4098893949251789, "grad_norm": 2.2225345770102165, "learning_rate": 9.959609317030954e-06, "loss": 0.6988, "step": 1260 }, { "epoch": 0.41021470396877036, "grad_norm": 2.0687991237255337, "learning_rate": 9.959368865163163e-06, "loss": 0.6873, "step": 1261 }, { "epoch": 0.41054001301236176, "grad_norm": 1.941282640870157, "learning_rate": 9.959127702614509e-06, "loss": 0.5477, "step": 1262 }, { "epoch": 0.41086532205595316, "grad_norm": 2.07086443750922, "learning_rate": 9.958885829419547e-06, "loss": 0.639, "step": 1263 }, { "epoch": 0.41119063109954457, "grad_norm": 2.6047487143407797, "learning_rate": 9.958643245612938e-06, "loss": 0.694, "step": 1264 }, { "epoch": 0.41151594014313597, "grad_norm": 2.1505367859406834, "learning_rate": 9.958399951229448e-06, "loss": 0.6573, "step": 1265 }, { "epoch": 0.4118412491867274, "grad_norm": 2.4148534090579346, "learning_rate": 9.958155946303939e-06, "loss": 0.673, "step": 1266 }, { "epoch": 0.4121665582303188, "grad_norm": 2.169763665036965, "learning_rate": 9.957911230871376e-06, "loss": 0.6626, "step": 1267 }, { "epoch": 0.41249186727391024, "grad_norm": 2.028340420396592, "learning_rate": 9.95766580496683e-06, "loss": 0.6673, "step": 1268 }, { "epoch": 0.41281717631750164, "grad_norm": 2.192849350394137, "learning_rate": 9.957419668625468e-06, "loss": 0.6992, "step": 1269 }, { "epoch": 0.41314248536109305, "grad_norm": 2.1313997298987277, "learning_rate": 9.95717282188256e-06, "loss": 0.7026, "step": 1270 }, { "epoch": 0.41346779440468445, "grad_norm": 2.170264945289364, "learning_rate": 9.956925264773483e-06, "loss": 0.7108, "step": 1271 }, { "epoch": 0.41379310344827586, "grad_norm": 2.0622510938141967, "learning_rate": 9.956676997333713e-06, "loss": 0.5644, "step": 1272 }, { "epoch": 0.41411841249186726, "grad_norm": 2.218016636428406, "learning_rate": 9.956428019598821e-06, "loss": 0.6512, "step": 1273 }, { "epoch": 0.41444372153545866, "grad_norm": 2.100865240335942, "learning_rate": 9.956178331604491e-06, "loss": 0.7017, "step": 1274 }, { "epoch": 0.4147690305790501, "grad_norm": 2.137921249213932, "learning_rate": 9.9559279333865e-06, "loss": 0.5602, "step": 1275 }, { "epoch": 0.41509433962264153, "grad_norm": 1.8844872574092075, "learning_rate": 9.955676824980734e-06, "loss": 0.6373, "step": 1276 }, { "epoch": 0.41541964866623293, "grad_norm": 2.4985422011531324, "learning_rate": 9.955425006423174e-06, "loss": 0.5568, "step": 1277 }, { "epoch": 0.41574495770982434, "grad_norm": 2.4837668829168136, "learning_rate": 9.955172477749907e-06, "loss": 0.7062, "step": 1278 }, { "epoch": 0.41607026675341574, "grad_norm": 2.0036727256175, "learning_rate": 9.95491923899712e-06, "loss": 0.6302, "step": 1279 }, { "epoch": 0.41639557579700714, "grad_norm": 2.378544619927388, "learning_rate": 9.954665290201102e-06, "loss": 0.6287, "step": 1280 }, { "epoch": 0.41672088484059855, "grad_norm": 2.24274027327028, "learning_rate": 9.954410631398245e-06, "loss": 0.6537, "step": 1281 }, { "epoch": 0.41704619388418995, "grad_norm": 2.067553659580072, "learning_rate": 9.954155262625041e-06, "loss": 0.6462, "step": 1282 }, { "epoch": 0.4173715029277814, "grad_norm": 2.1181456675235784, "learning_rate": 9.953899183918085e-06, "loss": 0.7092, "step": 1283 }, { "epoch": 0.4176968119713728, "grad_norm": 2.1226032165912634, "learning_rate": 9.953642395314072e-06, "loss": 0.6196, "step": 1284 }, { "epoch": 0.4180221210149642, "grad_norm": 1.6572455090818132, "learning_rate": 9.953384896849802e-06, "loss": 0.5604, "step": 1285 }, { "epoch": 0.4183474300585556, "grad_norm": 2.0258246093079784, "learning_rate": 9.953126688562174e-06, "loss": 0.6223, "step": 1286 }, { "epoch": 0.41867273910214703, "grad_norm": 2.3893412145797277, "learning_rate": 9.952867770488188e-06, "loss": 0.6048, "step": 1287 }, { "epoch": 0.41899804814573843, "grad_norm": 2.0603024966678682, "learning_rate": 9.95260814266495e-06, "loss": 0.7094, "step": 1288 }, { "epoch": 0.41932335718932984, "grad_norm": 2.0703624538108487, "learning_rate": 9.952347805129661e-06, "loss": 0.6088, "step": 1289 }, { "epoch": 0.4196486662329213, "grad_norm": 2.0182405766881253, "learning_rate": 9.952086757919632e-06, "loss": 0.5937, "step": 1290 }, { "epoch": 0.4199739752765127, "grad_norm": 1.908336136882204, "learning_rate": 9.951825001072268e-06, "loss": 0.5776, "step": 1291 }, { "epoch": 0.4202992843201041, "grad_norm": 2.1980666971244807, "learning_rate": 9.95156253462508e-06, "loss": 0.6127, "step": 1292 }, { "epoch": 0.4206245933636955, "grad_norm": 1.8435938267828849, "learning_rate": 9.951299358615677e-06, "loss": 0.6231, "step": 1293 }, { "epoch": 0.4209499024072869, "grad_norm": 2.0065720589712717, "learning_rate": 9.951035473081776e-06, "loss": 0.6738, "step": 1294 }, { "epoch": 0.4212752114508783, "grad_norm": 1.8725088627644497, "learning_rate": 9.950770878061191e-06, "loss": 0.6225, "step": 1295 }, { "epoch": 0.4216005204944697, "grad_norm": 1.9199005867819452, "learning_rate": 9.95050557359184e-06, "loss": 0.642, "step": 1296 }, { "epoch": 0.4219258295380612, "grad_norm": 2.356138474171413, "learning_rate": 9.950239559711738e-06, "loss": 0.5373, "step": 1297 }, { "epoch": 0.4222511385816526, "grad_norm": 1.9186240349203505, "learning_rate": 9.949972836459008e-06, "loss": 0.6693, "step": 1298 }, { "epoch": 0.422576447625244, "grad_norm": 2.2737825304990182, "learning_rate": 9.949705403871868e-06, "loss": 0.6394, "step": 1299 }, { "epoch": 0.4229017566688354, "grad_norm": 2.0643797392634533, "learning_rate": 9.949437261988645e-06, "loss": 0.5644, "step": 1300 }, { "epoch": 0.4232270657124268, "grad_norm": 2.3022758940755206, "learning_rate": 9.949168410847763e-06, "loss": 0.7853, "step": 1301 }, { "epoch": 0.4235523747560182, "grad_norm": 2.086325280276489, "learning_rate": 9.948898850487747e-06, "loss": 0.6889, "step": 1302 }, { "epoch": 0.4238776837996096, "grad_norm": 1.8503090067127999, "learning_rate": 9.948628580947227e-06, "loss": 0.605, "step": 1303 }, { "epoch": 0.42420299284320107, "grad_norm": 1.9877164580319742, "learning_rate": 9.94835760226493e-06, "loss": 0.592, "step": 1304 }, { "epoch": 0.42452830188679247, "grad_norm": 1.99199181252276, "learning_rate": 9.94808591447969e-06, "loss": 0.6, "step": 1305 }, { "epoch": 0.4248536109303839, "grad_norm": 2.3175576361115686, "learning_rate": 9.94781351763044e-06, "loss": 0.6556, "step": 1306 }, { "epoch": 0.4251789199739753, "grad_norm": 2.0311572854070246, "learning_rate": 9.947540411756214e-06, "loss": 0.5765, "step": 1307 }, { "epoch": 0.4255042290175667, "grad_norm": 2.2315724269367183, "learning_rate": 9.94726659689615e-06, "loss": 0.6654, "step": 1308 }, { "epoch": 0.4258295380611581, "grad_norm": 2.4341251265660224, "learning_rate": 9.946992073089482e-06, "loss": 0.7783, "step": 1309 }, { "epoch": 0.4261548471047495, "grad_norm": 2.0430014625951824, "learning_rate": 9.946716840375552e-06, "loss": 0.6484, "step": 1310 }, { "epoch": 0.42648015614834095, "grad_norm": 1.8675442213623386, "learning_rate": 9.9464408987938e-06, "loss": 0.5764, "step": 1311 }, { "epoch": 0.42680546519193235, "grad_norm": 1.6778964335644277, "learning_rate": 9.94616424838377e-06, "loss": 0.5533, "step": 1312 }, { "epoch": 0.42713077423552376, "grad_norm": 2.328158830471749, "learning_rate": 9.945886889185106e-06, "loss": 0.6716, "step": 1313 }, { "epoch": 0.42745608327911516, "grad_norm": 2.215847089105337, "learning_rate": 9.945608821237554e-06, "loss": 0.6866, "step": 1314 }, { "epoch": 0.42778139232270657, "grad_norm": 2.120057526986393, "learning_rate": 9.945330044580958e-06, "loss": 0.6986, "step": 1315 }, { "epoch": 0.42810670136629797, "grad_norm": 2.0871158363760567, "learning_rate": 9.945050559255271e-06, "loss": 0.5748, "step": 1316 }, { "epoch": 0.4284320104098894, "grad_norm": 1.8822250852109628, "learning_rate": 9.944770365300543e-06, "loss": 0.6218, "step": 1317 }, { "epoch": 0.42875731945348083, "grad_norm": 2.0508788122207586, "learning_rate": 9.944489462756923e-06, "loss": 0.6419, "step": 1318 }, { "epoch": 0.42908262849707224, "grad_norm": 1.9350589270110057, "learning_rate": 9.944207851664665e-06, "loss": 0.5308, "step": 1319 }, { "epoch": 0.42940793754066364, "grad_norm": 1.9149683761048233, "learning_rate": 9.943925532064129e-06, "loss": 0.5243, "step": 1320 }, { "epoch": 0.42973324658425505, "grad_norm": 1.8147052215475525, "learning_rate": 9.943642503995765e-06, "loss": 0.5862, "step": 1321 }, { "epoch": 0.43005855562784645, "grad_norm": 1.8060613123265328, "learning_rate": 9.943358767500135e-06, "loss": 0.5712, "step": 1322 }, { "epoch": 0.43038386467143785, "grad_norm": 2.0102703626519025, "learning_rate": 9.943074322617899e-06, "loss": 0.648, "step": 1323 }, { "epoch": 0.43070917371502926, "grad_norm": 2.0195042312768354, "learning_rate": 9.942789169389815e-06, "loss": 0.564, "step": 1324 }, { "epoch": 0.43103448275862066, "grad_norm": 1.9162150340833652, "learning_rate": 9.942503307856747e-06, "loss": 0.6311, "step": 1325 }, { "epoch": 0.4313597918022121, "grad_norm": 2.0599205986177416, "learning_rate": 9.942216738059663e-06, "loss": 0.5538, "step": 1326 }, { "epoch": 0.4316851008458035, "grad_norm": 1.913189341614243, "learning_rate": 9.941929460039623e-06, "loss": 0.6454, "step": 1327 }, { "epoch": 0.43201040988939493, "grad_norm": 2.34603614775514, "learning_rate": 9.941641473837797e-06, "loss": 0.7268, "step": 1328 }, { "epoch": 0.43233571893298633, "grad_norm": 2.1935269248655187, "learning_rate": 9.941352779495453e-06, "loss": 0.6362, "step": 1329 }, { "epoch": 0.43266102797657774, "grad_norm": 2.1745283183507755, "learning_rate": 9.941063377053961e-06, "loss": 0.6123, "step": 1330 }, { "epoch": 0.43298633702016914, "grad_norm": 1.9199764436683875, "learning_rate": 9.940773266554793e-06, "loss": 0.5928, "step": 1331 }, { "epoch": 0.43331164606376055, "grad_norm": 1.9977621531567262, "learning_rate": 9.94048244803952e-06, "loss": 0.6182, "step": 1332 }, { "epoch": 0.433636955107352, "grad_norm": 2.0355367954944015, "learning_rate": 9.94019092154982e-06, "loss": 0.568, "step": 1333 }, { "epoch": 0.4339622641509434, "grad_norm": 1.9734120069598353, "learning_rate": 9.939898687127468e-06, "loss": 0.696, "step": 1334 }, { "epoch": 0.4342875731945348, "grad_norm": 2.323652327488641, "learning_rate": 9.939605744814338e-06, "loss": 0.6728, "step": 1335 }, { "epoch": 0.4346128822381262, "grad_norm": 2.238666878060987, "learning_rate": 9.939312094652414e-06, "loss": 0.6638, "step": 1336 }, { "epoch": 0.4349381912817176, "grad_norm": 1.780779456656095, "learning_rate": 9.939017736683772e-06, "loss": 0.6581, "step": 1337 }, { "epoch": 0.435263500325309, "grad_norm": 1.9864887884478188, "learning_rate": 9.938722670950595e-06, "loss": 0.5642, "step": 1338 }, { "epoch": 0.43558880936890043, "grad_norm": 1.9639108874907873, "learning_rate": 9.938426897495168e-06, "loss": 0.6389, "step": 1339 }, { "epoch": 0.4359141184124919, "grad_norm": 2.1040455121777346, "learning_rate": 9.938130416359873e-06, "loss": 0.6403, "step": 1340 }, { "epoch": 0.4362394274560833, "grad_norm": 2.081261853128373, "learning_rate": 9.937833227587197e-06, "loss": 0.6031, "step": 1341 }, { "epoch": 0.4365647364996747, "grad_norm": 1.7573885084324465, "learning_rate": 9.937535331219728e-06, "loss": 0.4575, "step": 1342 }, { "epoch": 0.4368900455432661, "grad_norm": 1.7661724695551158, "learning_rate": 9.937236727300155e-06, "loss": 0.5432, "step": 1343 }, { "epoch": 0.4372153545868575, "grad_norm": 2.003808334199563, "learning_rate": 9.936937415871265e-06, "loss": 0.6631, "step": 1344 }, { "epoch": 0.4375406636304489, "grad_norm": 2.2309826591719775, "learning_rate": 9.936637396975952e-06, "loss": 0.6207, "step": 1345 }, { "epoch": 0.4378659726740403, "grad_norm": 2.234333611298523, "learning_rate": 9.93633667065721e-06, "loss": 0.6883, "step": 1346 }, { "epoch": 0.4381912817176318, "grad_norm": 2.0990324230680173, "learning_rate": 9.93603523695813e-06, "loss": 0.6475, "step": 1347 }, { "epoch": 0.4385165907612232, "grad_norm": 2.169926532480583, "learning_rate": 9.935733095921911e-06, "loss": 0.7275, "step": 1348 }, { "epoch": 0.4388418998048146, "grad_norm": 2.2143606997204826, "learning_rate": 9.935430247591849e-06, "loss": 0.717, "step": 1349 }, { "epoch": 0.439167208848406, "grad_norm": 2.146660923547246, "learning_rate": 9.935126692011339e-06, "loss": 0.6231, "step": 1350 }, { "epoch": 0.4394925178919974, "grad_norm": 2.067184214023533, "learning_rate": 9.934822429223887e-06, "loss": 0.6012, "step": 1351 }, { "epoch": 0.4398178269355888, "grad_norm": 2.235798552637545, "learning_rate": 9.934517459273088e-06, "loss": 0.7181, "step": 1352 }, { "epoch": 0.4401431359791802, "grad_norm": 2.1697505025439305, "learning_rate": 9.93421178220265e-06, "loss": 0.6779, "step": 1353 }, { "epoch": 0.44046844502277166, "grad_norm": 1.9896852259053837, "learning_rate": 9.933905398056371e-06, "loss": 0.5938, "step": 1354 }, { "epoch": 0.44079375406636306, "grad_norm": 1.7981848668724882, "learning_rate": 9.93359830687816e-06, "loss": 0.5826, "step": 1355 }, { "epoch": 0.44111906310995447, "grad_norm": 1.835896484022937, "learning_rate": 9.933290508712023e-06, "loss": 0.5456, "step": 1356 }, { "epoch": 0.44144437215354587, "grad_norm": 2.0513239692674814, "learning_rate": 9.932982003602066e-06, "loss": 0.5941, "step": 1357 }, { "epoch": 0.4417696811971373, "grad_norm": 2.276655035158557, "learning_rate": 9.932672791592499e-06, "loss": 0.7165, "step": 1358 }, { "epoch": 0.4420949902407287, "grad_norm": 2.000460339078954, "learning_rate": 9.932362872727632e-06, "loss": 0.6135, "step": 1359 }, { "epoch": 0.4424202992843201, "grad_norm": 2.184796668516001, "learning_rate": 9.932052247051878e-06, "loss": 0.6451, "step": 1360 }, { "epoch": 0.4427456083279115, "grad_norm": 2.1854444112086013, "learning_rate": 9.931740914609747e-06, "loss": 0.6766, "step": 1361 }, { "epoch": 0.44307091737150295, "grad_norm": 1.7274740001176092, "learning_rate": 9.931428875445857e-06, "loss": 0.4936, "step": 1362 }, { "epoch": 0.44339622641509435, "grad_norm": 1.9279048226219815, "learning_rate": 9.931116129604919e-06, "loss": 0.5673, "step": 1363 }, { "epoch": 0.44372153545868576, "grad_norm": 1.9640478186399652, "learning_rate": 9.930802677131752e-06, "loss": 0.7493, "step": 1364 }, { "epoch": 0.44404684450227716, "grad_norm": 2.0922350904017777, "learning_rate": 9.930488518071275e-06, "loss": 0.628, "step": 1365 }, { "epoch": 0.44437215354586856, "grad_norm": 1.9292166129957733, "learning_rate": 9.930173652468507e-06, "loss": 0.6188, "step": 1366 }, { "epoch": 0.44469746258945997, "grad_norm": 1.7557731182683387, "learning_rate": 9.929858080368564e-06, "loss": 0.5349, "step": 1367 }, { "epoch": 0.4450227716330514, "grad_norm": 1.9796887857298082, "learning_rate": 9.929541801816674e-06, "loss": 0.6217, "step": 1368 }, { "epoch": 0.44534808067664283, "grad_norm": 1.9775728596320863, "learning_rate": 9.929224816858158e-06, "loss": 0.6276, "step": 1369 }, { "epoch": 0.44567338972023424, "grad_norm": 1.7828012705953071, "learning_rate": 9.928907125538436e-06, "loss": 0.5515, "step": 1370 }, { "epoch": 0.44599869876382564, "grad_norm": 2.415047232188689, "learning_rate": 9.92858872790304e-06, "loss": 0.6263, "step": 1371 }, { "epoch": 0.44632400780741704, "grad_norm": 2.423431274074802, "learning_rate": 9.928269623997591e-06, "loss": 0.787, "step": 1372 }, { "epoch": 0.44664931685100845, "grad_norm": 2.2471973678190946, "learning_rate": 9.92794981386782e-06, "loss": 0.6264, "step": 1373 }, { "epoch": 0.44697462589459985, "grad_norm": 2.0476483359286237, "learning_rate": 9.927629297559555e-06, "loss": 0.7567, "step": 1374 }, { "epoch": 0.44729993493819126, "grad_norm": 2.0587793276653565, "learning_rate": 9.927308075118726e-06, "loss": 0.46, "step": 1375 }, { "epoch": 0.4476252439817827, "grad_norm": 1.985650733142002, "learning_rate": 9.926986146591364e-06, "loss": 0.6274, "step": 1376 }, { "epoch": 0.4479505530253741, "grad_norm": 2.113709637993379, "learning_rate": 9.926663512023605e-06, "loss": 0.7281, "step": 1377 }, { "epoch": 0.4482758620689655, "grad_norm": 1.7066341446959505, "learning_rate": 9.926340171461677e-06, "loss": 0.5079, "step": 1378 }, { "epoch": 0.44860117111255693, "grad_norm": 2.035226921400331, "learning_rate": 9.926016124951918e-06, "loss": 0.6408, "step": 1379 }, { "epoch": 0.44892648015614833, "grad_norm": 1.962140943557285, "learning_rate": 9.925691372540765e-06, "loss": 0.5935, "step": 1380 }, { "epoch": 0.44925178919973974, "grad_norm": 2.0138505632089703, "learning_rate": 9.925365914274753e-06, "loss": 0.6601, "step": 1381 }, { "epoch": 0.44957709824333114, "grad_norm": 2.1389915604558905, "learning_rate": 9.925039750200524e-06, "loss": 0.5639, "step": 1382 }, { "epoch": 0.4499024072869226, "grad_norm": 2.1342164242070023, "learning_rate": 9.924712880364813e-06, "loss": 0.5501, "step": 1383 }, { "epoch": 0.450227716330514, "grad_norm": 2.104525069768656, "learning_rate": 9.924385304814464e-06, "loss": 0.7143, "step": 1384 }, { "epoch": 0.4505530253741054, "grad_norm": 2.182548890607546, "learning_rate": 9.924057023596418e-06, "loss": 0.5804, "step": 1385 }, { "epoch": 0.4508783344176968, "grad_norm": 2.0118587637055856, "learning_rate": 9.923728036757716e-06, "loss": 0.7049, "step": 1386 }, { "epoch": 0.4512036434612882, "grad_norm": 2.300172425657615, "learning_rate": 9.923398344345504e-06, "loss": 0.6837, "step": 1387 }, { "epoch": 0.4515289525048796, "grad_norm": 1.7519632232332185, "learning_rate": 9.923067946407028e-06, "loss": 0.5649, "step": 1388 }, { "epoch": 0.451854261548471, "grad_norm": 2.040867523021838, "learning_rate": 9.922736842989631e-06, "loss": 0.6297, "step": 1389 }, { "epoch": 0.4521795705920625, "grad_norm": 2.316773456818668, "learning_rate": 9.922405034140763e-06, "loss": 0.6803, "step": 1390 }, { "epoch": 0.4525048796356539, "grad_norm": 2.1865117885937186, "learning_rate": 9.922072519907972e-06, "loss": 0.6315, "step": 1391 }, { "epoch": 0.4528301886792453, "grad_norm": 1.8927763860849292, "learning_rate": 9.921739300338906e-06, "loss": 0.5775, "step": 1392 }, { "epoch": 0.4531554977228367, "grad_norm": 2.3076703917111514, "learning_rate": 9.921405375481318e-06, "loss": 0.6817, "step": 1393 }, { "epoch": 0.4534808067664281, "grad_norm": 2.0816313811053955, "learning_rate": 9.92107074538306e-06, "loss": 0.6452, "step": 1394 }, { "epoch": 0.4538061158100195, "grad_norm": 2.3163409677816396, "learning_rate": 9.92073541009208e-06, "loss": 0.6975, "step": 1395 }, { "epoch": 0.4541314248536109, "grad_norm": 1.8851606980364193, "learning_rate": 9.920399369656437e-06, "loss": 0.576, "step": 1396 }, { "epoch": 0.45445673389720237, "grad_norm": 2.029998505446986, "learning_rate": 9.920062624124282e-06, "loss": 0.5727, "step": 1397 }, { "epoch": 0.4547820429407938, "grad_norm": 2.1771863839809438, "learning_rate": 9.919725173543874e-06, "loss": 0.744, "step": 1398 }, { "epoch": 0.4551073519843852, "grad_norm": 2.155064996536362, "learning_rate": 9.91938701796357e-06, "loss": 0.6761, "step": 1399 }, { "epoch": 0.4554326610279766, "grad_norm": 1.9176494958918213, "learning_rate": 9.919048157431823e-06, "loss": 0.5913, "step": 1400 }, { "epoch": 0.455757970071568, "grad_norm": 2.269301963785391, "learning_rate": 9.918708591997197e-06, "loss": 0.7043, "step": 1401 }, { "epoch": 0.4560832791151594, "grad_norm": 2.3622547013820245, "learning_rate": 9.918368321708349e-06, "loss": 0.6837, "step": 1402 }, { "epoch": 0.4564085881587508, "grad_norm": 1.8917944294318043, "learning_rate": 9.918027346614043e-06, "loss": 0.5957, "step": 1403 }, { "epoch": 0.4567338972023422, "grad_norm": 2.176098117588663, "learning_rate": 9.917685666763139e-06, "loss": 0.6552, "step": 1404 }, { "epoch": 0.45705920624593366, "grad_norm": 1.8035041902255435, "learning_rate": 9.9173432822046e-06, "loss": 0.5563, "step": 1405 }, { "epoch": 0.45738451528952506, "grad_norm": 2.168739582392316, "learning_rate": 9.91700019298749e-06, "loss": 0.6563, "step": 1406 }, { "epoch": 0.45770982433311647, "grad_norm": 1.97828968624238, "learning_rate": 9.916656399160974e-06, "loss": 0.5075, "step": 1407 }, { "epoch": 0.45803513337670787, "grad_norm": 2.537839303327791, "learning_rate": 9.916311900774318e-06, "loss": 0.7253, "step": 1408 }, { "epoch": 0.4583604424202993, "grad_norm": 2.101308418529594, "learning_rate": 9.91596669787689e-06, "loss": 0.6397, "step": 1409 }, { "epoch": 0.4586857514638907, "grad_norm": 1.9787740015741424, "learning_rate": 9.915620790518155e-06, "loss": 0.6853, "step": 1410 }, { "epoch": 0.4590110605074821, "grad_norm": 2.3438630407606347, "learning_rate": 9.915274178747684e-06, "loss": 0.6872, "step": 1411 }, { "epoch": 0.45933636955107354, "grad_norm": 2.0885060768649617, "learning_rate": 9.914926862615145e-06, "loss": 0.6233, "step": 1412 }, { "epoch": 0.45966167859466495, "grad_norm": 1.907198837977907, "learning_rate": 9.91457884217031e-06, "loss": 0.5395, "step": 1413 }, { "epoch": 0.45998698763825635, "grad_norm": 2.033356771742178, "learning_rate": 9.91423011746305e-06, "loss": 0.6923, "step": 1414 }, { "epoch": 0.46031229668184775, "grad_norm": 2.1747745438606954, "learning_rate": 9.91388068854334e-06, "loss": 0.6165, "step": 1415 }, { "epoch": 0.46063760572543916, "grad_norm": 2.186960526730423, "learning_rate": 9.91353055546125e-06, "loss": 0.6306, "step": 1416 }, { "epoch": 0.46096291476903056, "grad_norm": 1.678579189212653, "learning_rate": 9.913179718266954e-06, "loss": 0.5452, "step": 1417 }, { "epoch": 0.46128822381262197, "grad_norm": 1.9856660492709435, "learning_rate": 9.912828177010729e-06, "loss": 0.6371, "step": 1418 }, { "epoch": 0.4616135328562134, "grad_norm": 1.716636308939893, "learning_rate": 9.91247593174295e-06, "loss": 0.5693, "step": 1419 }, { "epoch": 0.46193884189980483, "grad_norm": 2.431673367475699, "learning_rate": 9.912122982514098e-06, "loss": 0.7023, "step": 1420 }, { "epoch": 0.46226415094339623, "grad_norm": 1.8572250116604996, "learning_rate": 9.911769329374746e-06, "loss": 0.5051, "step": 1421 }, { "epoch": 0.46258945998698764, "grad_norm": 2.873027914974647, "learning_rate": 9.911414972375573e-06, "loss": 0.715, "step": 1422 }, { "epoch": 0.46291476903057904, "grad_norm": 1.901196343404097, "learning_rate": 9.91105991156736e-06, "loss": 0.5402, "step": 1423 }, { "epoch": 0.46324007807417045, "grad_norm": 1.8888472156648903, "learning_rate": 9.910704147000988e-06, "loss": 0.6196, "step": 1424 }, { "epoch": 0.46356538711776185, "grad_norm": 1.9721982884339149, "learning_rate": 9.910347678727437e-06, "loss": 0.691, "step": 1425 }, { "epoch": 0.4638906961613533, "grad_norm": 2.25763483107995, "learning_rate": 9.90999050679779e-06, "loss": 0.6202, "step": 1426 }, { "epoch": 0.4642160052049447, "grad_norm": 1.966030474796569, "learning_rate": 9.90963263126323e-06, "loss": 0.5725, "step": 1427 }, { "epoch": 0.4645413142485361, "grad_norm": 2.063381633134983, "learning_rate": 9.909274052175042e-06, "loss": 0.6477, "step": 1428 }, { "epoch": 0.4648666232921275, "grad_norm": 2.1999599332498554, "learning_rate": 9.908914769584608e-06, "loss": 0.5999, "step": 1429 }, { "epoch": 0.4651919323357189, "grad_norm": 1.7527323107075785, "learning_rate": 9.908554783543414e-06, "loss": 0.4714, "step": 1430 }, { "epoch": 0.46551724137931033, "grad_norm": 2.1636419229460486, "learning_rate": 9.908194094103046e-06, "loss": 0.6941, "step": 1431 }, { "epoch": 0.46584255042290174, "grad_norm": 1.8940900008043922, "learning_rate": 9.907832701315194e-06, "loss": 0.6594, "step": 1432 }, { "epoch": 0.4661678594664932, "grad_norm": 1.9207574440524335, "learning_rate": 9.907470605231642e-06, "loss": 0.6206, "step": 1433 }, { "epoch": 0.4664931685100846, "grad_norm": 1.869676911336321, "learning_rate": 9.907107805904281e-06, "loss": 0.5797, "step": 1434 }, { "epoch": 0.466818477553676, "grad_norm": 1.9805641090694892, "learning_rate": 9.9067443033851e-06, "loss": 0.5464, "step": 1435 }, { "epoch": 0.4671437865972674, "grad_norm": 1.8377367575113819, "learning_rate": 9.90638009772619e-06, "loss": 0.5423, "step": 1436 }, { "epoch": 0.4674690956408588, "grad_norm": 1.661193141345869, "learning_rate": 9.906015188979738e-06, "loss": 0.4983, "step": 1437 }, { "epoch": 0.4677944046844502, "grad_norm": 2.0290979266247873, "learning_rate": 9.90564957719804e-06, "loss": 0.6105, "step": 1438 }, { "epoch": 0.4681197137280416, "grad_norm": 2.0988117101714527, "learning_rate": 9.905283262433487e-06, "loss": 0.6167, "step": 1439 }, { "epoch": 0.468445022771633, "grad_norm": 1.81319174861374, "learning_rate": 9.904916244738572e-06, "loss": 0.552, "step": 1440 }, { "epoch": 0.4687703318152245, "grad_norm": 2.277104433541024, "learning_rate": 9.904548524165888e-06, "loss": 0.7637, "step": 1441 }, { "epoch": 0.4690956408588159, "grad_norm": 2.1430889663059314, "learning_rate": 9.904180100768131e-06, "loss": 0.5807, "step": 1442 }, { "epoch": 0.4694209499024073, "grad_norm": 2.3603136010152634, "learning_rate": 9.903810974598095e-06, "loss": 0.7714, "step": 1443 }, { "epoch": 0.4697462589459987, "grad_norm": 1.8508072257611765, "learning_rate": 9.903441145708679e-06, "loss": 0.5735, "step": 1444 }, { "epoch": 0.4700715679895901, "grad_norm": 2.0766148336734442, "learning_rate": 9.903070614152875e-06, "loss": 0.5663, "step": 1445 }, { "epoch": 0.4703968770331815, "grad_norm": 2.0144916434867897, "learning_rate": 9.902699379983784e-06, "loss": 0.6345, "step": 1446 }, { "epoch": 0.4707221860767729, "grad_norm": 1.8556049993866386, "learning_rate": 9.902327443254603e-06, "loss": 0.5627, "step": 1447 }, { "epoch": 0.47104749512036437, "grad_norm": 1.914822888056776, "learning_rate": 9.901954804018632e-06, "loss": 0.6501, "step": 1448 }, { "epoch": 0.47137280416395577, "grad_norm": 1.7908925666768656, "learning_rate": 9.901581462329268e-06, "loss": 0.6056, "step": 1449 }, { "epoch": 0.4716981132075472, "grad_norm": 1.8690899531539706, "learning_rate": 9.901207418240012e-06, "loss": 0.5381, "step": 1450 }, { "epoch": 0.4720234222511386, "grad_norm": 2.2107462453198634, "learning_rate": 9.900832671804466e-06, "loss": 0.6686, "step": 1451 }, { "epoch": 0.47234873129473, "grad_norm": 2.2702176460882697, "learning_rate": 9.900457223076332e-06, "loss": 0.681, "step": 1452 }, { "epoch": 0.4726740403383214, "grad_norm": 1.8239989222789603, "learning_rate": 9.900081072109409e-06, "loss": 0.6178, "step": 1453 }, { "epoch": 0.4729993493819128, "grad_norm": 2.237543897988283, "learning_rate": 9.899704218957602e-06, "loss": 0.6777, "step": 1454 }, { "epoch": 0.47332465842550425, "grad_norm": 1.8843156618779222, "learning_rate": 9.899326663674913e-06, "loss": 0.6497, "step": 1455 }, { "epoch": 0.47364996746909566, "grad_norm": 2.0224225895709678, "learning_rate": 9.898948406315446e-06, "loss": 0.7362, "step": 1456 }, { "epoch": 0.47397527651268706, "grad_norm": 1.8609872593341552, "learning_rate": 9.898569446933406e-06, "loss": 0.6504, "step": 1457 }, { "epoch": 0.47430058555627846, "grad_norm": 2.0062574449415815, "learning_rate": 9.898189785583102e-06, "loss": 0.6894, "step": 1458 }, { "epoch": 0.47462589459986987, "grad_norm": 1.8101761354875732, "learning_rate": 9.897809422318932e-06, "loss": 0.592, "step": 1459 }, { "epoch": 0.4749512036434613, "grad_norm": 1.726417278438717, "learning_rate": 9.897428357195406e-06, "loss": 0.4995, "step": 1460 }, { "epoch": 0.4752765126870527, "grad_norm": 1.6872149616603656, "learning_rate": 9.897046590267135e-06, "loss": 0.5133, "step": 1461 }, { "epoch": 0.47560182173064414, "grad_norm": 1.9420749190243247, "learning_rate": 9.89666412158882e-06, "loss": 0.647, "step": 1462 }, { "epoch": 0.47592713077423554, "grad_norm": 2.0305888845814586, "learning_rate": 9.89628095121527e-06, "loss": 0.6698, "step": 1463 }, { "epoch": 0.47625243981782694, "grad_norm": 2.032362086144091, "learning_rate": 9.895897079201397e-06, "loss": 0.5974, "step": 1464 }, { "epoch": 0.47657774886141835, "grad_norm": 1.955638839393503, "learning_rate": 9.895512505602208e-06, "loss": 0.6157, "step": 1465 }, { "epoch": 0.47690305790500975, "grad_norm": 1.812593446434037, "learning_rate": 9.895127230472815e-06, "loss": 0.6944, "step": 1466 }, { "epoch": 0.47722836694860116, "grad_norm": 2.2756059745486397, "learning_rate": 9.894741253868423e-06, "loss": 0.7132, "step": 1467 }, { "epoch": 0.47755367599219256, "grad_norm": 1.763013752145134, "learning_rate": 9.894354575844347e-06, "loss": 0.5245, "step": 1468 }, { "epoch": 0.477878985035784, "grad_norm": 1.9525610125199717, "learning_rate": 9.893967196455997e-06, "loss": 0.5617, "step": 1469 }, { "epoch": 0.4782042940793754, "grad_norm": 1.872420759824631, "learning_rate": 9.893579115758886e-06, "loss": 0.5978, "step": 1470 }, { "epoch": 0.47852960312296683, "grad_norm": 1.8340117815204893, "learning_rate": 9.893190333808625e-06, "loss": 0.5388, "step": 1471 }, { "epoch": 0.47885491216655823, "grad_norm": 1.8980455324297176, "learning_rate": 9.892800850660924e-06, "loss": 0.6201, "step": 1472 }, { "epoch": 0.47918022121014964, "grad_norm": 1.8762272161841904, "learning_rate": 9.892410666371602e-06, "loss": 0.5815, "step": 1473 }, { "epoch": 0.47950553025374104, "grad_norm": 1.727799801394269, "learning_rate": 9.892019780996568e-06, "loss": 0.497, "step": 1474 }, { "epoch": 0.47983083929733245, "grad_norm": 1.934236816904039, "learning_rate": 9.891628194591839e-06, "loss": 0.622, "step": 1475 }, { "epoch": 0.4801561483409239, "grad_norm": 2.239745439774118, "learning_rate": 9.891235907213525e-06, "loss": 0.6641, "step": 1476 }, { "epoch": 0.4804814573845153, "grad_norm": 1.96107222974959, "learning_rate": 9.890842918917849e-06, "loss": 0.6623, "step": 1477 }, { "epoch": 0.4808067664281067, "grad_norm": 2.033772556224288, "learning_rate": 9.89044922976112e-06, "loss": 0.7959, "step": 1478 }, { "epoch": 0.4811320754716981, "grad_norm": 1.9068031109454664, "learning_rate": 9.890054839799755e-06, "loss": 0.616, "step": 1479 }, { "epoch": 0.4814573845152895, "grad_norm": 2.024575692673324, "learning_rate": 9.88965974909027e-06, "loss": 0.6023, "step": 1480 }, { "epoch": 0.4817826935588809, "grad_norm": 1.9732085911701591, "learning_rate": 9.889263957689287e-06, "loss": 0.6112, "step": 1481 }, { "epoch": 0.48210800260247233, "grad_norm": 1.8092557391440054, "learning_rate": 9.888867465653514e-06, "loss": 0.5821, "step": 1482 }, { "epoch": 0.48243331164606373, "grad_norm": 2.117699515074846, "learning_rate": 9.888470273039776e-06, "loss": 0.6316, "step": 1483 }, { "epoch": 0.4827586206896552, "grad_norm": 2.130752708301083, "learning_rate": 9.88807237990499e-06, "loss": 0.6498, "step": 1484 }, { "epoch": 0.4830839297332466, "grad_norm": 2.2957521029452, "learning_rate": 9.887673786306172e-06, "loss": 0.7445, "step": 1485 }, { "epoch": 0.483409238776838, "grad_norm": 2.280282465992969, "learning_rate": 9.887274492300441e-06, "loss": 0.7142, "step": 1486 }, { "epoch": 0.4837345478204294, "grad_norm": 1.9764872656181838, "learning_rate": 9.886874497945015e-06, "loss": 0.6875, "step": 1487 }, { "epoch": 0.4840598568640208, "grad_norm": 2.053980528799275, "learning_rate": 9.886473803297218e-06, "loss": 0.6511, "step": 1488 }, { "epoch": 0.4843851659076122, "grad_norm": 2.10675292181098, "learning_rate": 9.886072408414466e-06, "loss": 0.6924, "step": 1489 }, { "epoch": 0.4847104749512036, "grad_norm": 2.172978917019874, "learning_rate": 9.885670313354281e-06, "loss": 0.6271, "step": 1490 }, { "epoch": 0.4850357839947951, "grad_norm": 1.8156704175194462, "learning_rate": 9.885267518174282e-06, "loss": 0.5931, "step": 1491 }, { "epoch": 0.4853610930383865, "grad_norm": 1.835983979464552, "learning_rate": 9.88486402293219e-06, "loss": 0.6284, "step": 1492 }, { "epoch": 0.4856864020819779, "grad_norm": 2.157652711383864, "learning_rate": 9.884459827685828e-06, "loss": 0.6625, "step": 1493 }, { "epoch": 0.4860117111255693, "grad_norm": 2.083558539639187, "learning_rate": 9.884054932493115e-06, "loss": 0.6915, "step": 1494 }, { "epoch": 0.4863370201691607, "grad_norm": 1.8798304786211604, "learning_rate": 9.883649337412073e-06, "loss": 0.6093, "step": 1495 }, { "epoch": 0.4866623292127521, "grad_norm": 1.9554234353244018, "learning_rate": 9.883243042500827e-06, "loss": 0.5646, "step": 1496 }, { "epoch": 0.4869876382563435, "grad_norm": 2.2621309648270214, "learning_rate": 9.882836047817596e-06, "loss": 0.7259, "step": 1497 }, { "epoch": 0.48731294729993496, "grad_norm": 1.8330696167281533, "learning_rate": 9.882428353420705e-06, "loss": 0.6443, "step": 1498 }, { "epoch": 0.48763825634352637, "grad_norm": 1.8827164096270794, "learning_rate": 9.882019959368574e-06, "loss": 0.5881, "step": 1499 }, { "epoch": 0.48796356538711777, "grad_norm": 1.9059398889586185, "learning_rate": 9.881610865719728e-06, "loss": 0.4923, "step": 1500 }, { "epoch": 0.4882888744307092, "grad_norm": 2.1291660966487154, "learning_rate": 9.88120107253279e-06, "loss": 0.7134, "step": 1501 }, { "epoch": 0.4886141834743006, "grad_norm": 2.044522136815203, "learning_rate": 9.880790579866485e-06, "loss": 0.5821, "step": 1502 }, { "epoch": 0.488939492517892, "grad_norm": 2.283293196652707, "learning_rate": 9.880379387779637e-06, "loss": 0.6351, "step": 1503 }, { "epoch": 0.4892648015614834, "grad_norm": 1.8090970965872175, "learning_rate": 9.879967496331166e-06, "loss": 0.6382, "step": 1504 }, { "epoch": 0.48959011060507485, "grad_norm": 2.137372804320905, "learning_rate": 9.8795549055801e-06, "loss": 0.73, "step": 1505 }, { "epoch": 0.48991541964866625, "grad_norm": 2.047385521792071, "learning_rate": 9.879141615585564e-06, "loss": 0.5745, "step": 1506 }, { "epoch": 0.49024072869225765, "grad_norm": 2.329667879618475, "learning_rate": 9.878727626406782e-06, "loss": 0.8245, "step": 1507 }, { "epoch": 0.49056603773584906, "grad_norm": 1.953468882236508, "learning_rate": 9.878312938103076e-06, "loss": 0.6454, "step": 1508 }, { "epoch": 0.49089134677944046, "grad_norm": 1.8930542289650385, "learning_rate": 9.877897550733874e-06, "loss": 0.6667, "step": 1509 }, { "epoch": 0.49121665582303187, "grad_norm": 2.025831673682426, "learning_rate": 9.877481464358701e-06, "loss": 0.5926, "step": 1510 }, { "epoch": 0.49154196486662327, "grad_norm": 2.073558537695087, "learning_rate": 9.877064679037183e-06, "loss": 0.6816, "step": 1511 }, { "epoch": 0.49186727391021473, "grad_norm": 2.0918806058983597, "learning_rate": 9.876647194829043e-06, "loss": 0.7307, "step": 1512 }, { "epoch": 0.49219258295380613, "grad_norm": 2.1101464361323092, "learning_rate": 9.87622901179411e-06, "loss": 0.6884, "step": 1513 }, { "epoch": 0.49251789199739754, "grad_norm": 2.167975850905635, "learning_rate": 9.875810129992309e-06, "loss": 0.6051, "step": 1514 }, { "epoch": 0.49284320104098894, "grad_norm": 2.0366672793634066, "learning_rate": 9.875390549483662e-06, "loss": 0.5873, "step": 1515 }, { "epoch": 0.49316851008458035, "grad_norm": 2.1596572581523437, "learning_rate": 9.874970270328304e-06, "loss": 0.6465, "step": 1516 }, { "epoch": 0.49349381912817175, "grad_norm": 2.095864701906522, "learning_rate": 9.874549292586453e-06, "loss": 0.625, "step": 1517 }, { "epoch": 0.49381912817176316, "grad_norm": 1.7478445485918006, "learning_rate": 9.874127616318438e-06, "loss": 0.5368, "step": 1518 }, { "epoch": 0.49414443721535456, "grad_norm": 2.0422745451242252, "learning_rate": 9.873705241584686e-06, "loss": 0.6173, "step": 1519 }, { "epoch": 0.494469746258946, "grad_norm": 1.9598622885886043, "learning_rate": 9.873282168445725e-06, "loss": 0.559, "step": 1520 }, { "epoch": 0.4947950553025374, "grad_norm": 1.9049998700684347, "learning_rate": 9.872858396962179e-06, "loss": 0.6314, "step": 1521 }, { "epoch": 0.4951203643461288, "grad_norm": 1.869958790267797, "learning_rate": 9.872433927194776e-06, "loss": 0.6466, "step": 1522 }, { "epoch": 0.49544567338972023, "grad_norm": 1.8682084187542862, "learning_rate": 9.872008759204343e-06, "loss": 0.5564, "step": 1523 }, { "epoch": 0.49577098243331164, "grad_norm": 2.0325867229456884, "learning_rate": 9.871582893051805e-06, "loss": 0.6497, "step": 1524 }, { "epoch": 0.49609629147690304, "grad_norm": 1.903288279957025, "learning_rate": 9.871156328798191e-06, "loss": 0.6285, "step": 1525 }, { "epoch": 0.49642160052049444, "grad_norm": 2.08341116446343, "learning_rate": 9.870729066504629e-06, "loss": 0.658, "step": 1526 }, { "epoch": 0.4967469095640859, "grad_norm": 2.021373716777203, "learning_rate": 9.870301106232343e-06, "loss": 0.5991, "step": 1527 }, { "epoch": 0.4970722186076773, "grad_norm": 1.969991375065408, "learning_rate": 9.869872448042662e-06, "loss": 0.6734, "step": 1528 }, { "epoch": 0.4973975276512687, "grad_norm": 1.9818761313322666, "learning_rate": 9.86944309199701e-06, "loss": 0.6055, "step": 1529 }, { "epoch": 0.4977228366948601, "grad_norm": 1.6784391431846328, "learning_rate": 9.869013038156919e-06, "loss": 0.5299, "step": 1530 }, { "epoch": 0.4980481457384515, "grad_norm": 1.9770322991083018, "learning_rate": 9.868582286584012e-06, "loss": 0.548, "step": 1531 }, { "epoch": 0.4983734547820429, "grad_norm": 2.1468175648033188, "learning_rate": 9.868150837340019e-06, "loss": 0.6178, "step": 1532 }, { "epoch": 0.49869876382563433, "grad_norm": 1.7593389032354887, "learning_rate": 9.867718690486762e-06, "loss": 0.5127, "step": 1533 }, { "epoch": 0.4990240728692258, "grad_norm": 1.8869772606963309, "learning_rate": 9.867285846086175e-06, "loss": 0.6634, "step": 1534 }, { "epoch": 0.4993493819128172, "grad_norm": 2.0577027074397445, "learning_rate": 9.866852304200278e-06, "loss": 0.6464, "step": 1535 }, { "epoch": 0.4996746909564086, "grad_norm": 1.902813482327519, "learning_rate": 9.866418064891201e-06, "loss": 0.627, "step": 1536 }, { "epoch": 0.5, "grad_norm": 1.94069185464925, "learning_rate": 9.865983128221173e-06, "loss": 0.4935, "step": 1537 }, { "epoch": 0.5003253090435914, "grad_norm": 2.088922211974032, "learning_rate": 9.865547494252516e-06, "loss": 0.618, "step": 1538 }, { "epoch": 0.5006506180871828, "grad_norm": 1.9210211206756673, "learning_rate": 9.865111163047662e-06, "loss": 0.5589, "step": 1539 }, { "epoch": 0.5009759271307742, "grad_norm": 2.06970511683147, "learning_rate": 9.864674134669132e-06, "loss": 0.7083, "step": 1540 }, { "epoch": 0.5013012361743656, "grad_norm": 1.927210593459205, "learning_rate": 9.864236409179556e-06, "loss": 0.576, "step": 1541 }, { "epoch": 0.501626545217957, "grad_norm": 2.1022356068615724, "learning_rate": 9.86379798664166e-06, "loss": 0.732, "step": 1542 }, { "epoch": 0.5019518542615484, "grad_norm": 2.1643500128168704, "learning_rate": 9.86335886711827e-06, "loss": 0.6508, "step": 1543 }, { "epoch": 0.5022771633051398, "grad_norm": 2.1314461840781718, "learning_rate": 9.86291905067231e-06, "loss": 0.5677, "step": 1544 }, { "epoch": 0.5026024723487313, "grad_norm": 2.0711089777493443, "learning_rate": 9.862478537366808e-06, "loss": 0.5314, "step": 1545 }, { "epoch": 0.5029277813923227, "grad_norm": 1.9548339308843925, "learning_rate": 9.86203732726489e-06, "loss": 0.5935, "step": 1546 }, { "epoch": 0.5032530904359142, "grad_norm": 2.076976285170037, "learning_rate": 9.861595420429784e-06, "loss": 0.6624, "step": 1547 }, { "epoch": 0.5035783994795056, "grad_norm": 1.9625080505492372, "learning_rate": 9.861152816924809e-06, "loss": 0.6336, "step": 1548 }, { "epoch": 0.503903708523097, "grad_norm": 1.8044920482888955, "learning_rate": 9.860709516813395e-06, "loss": 0.6452, "step": 1549 }, { "epoch": 0.5042290175666884, "grad_norm": 1.7379175373573574, "learning_rate": 9.860265520159069e-06, "loss": 0.6046, "step": 1550 }, { "epoch": 0.5045543266102798, "grad_norm": 1.8468327154049695, "learning_rate": 9.859820827025451e-06, "loss": 0.5679, "step": 1551 }, { "epoch": 0.5048796356538712, "grad_norm": 1.918877702461706, "learning_rate": 9.85937543747627e-06, "loss": 0.5784, "step": 1552 }, { "epoch": 0.5052049446974626, "grad_norm": 1.979031235244771, "learning_rate": 9.858929351575349e-06, "loss": 0.5697, "step": 1553 }, { "epoch": 0.505530253741054, "grad_norm": 2.108038515884951, "learning_rate": 9.858482569386609e-06, "loss": 0.5843, "step": 1554 }, { "epoch": 0.5058555627846454, "grad_norm": 2.0709629492663884, "learning_rate": 9.858035090974081e-06, "loss": 0.6096, "step": 1555 }, { "epoch": 0.5061808718282368, "grad_norm": 1.8506682636748493, "learning_rate": 9.857586916401885e-06, "loss": 0.637, "step": 1556 }, { "epoch": 0.5065061808718282, "grad_norm": 2.172026790907268, "learning_rate": 9.857138045734243e-06, "loss": 0.7149, "step": 1557 }, { "epoch": 0.5068314899154196, "grad_norm": 1.6983506536657853, "learning_rate": 9.856688479035484e-06, "loss": 0.5211, "step": 1558 }, { "epoch": 0.5071567989590111, "grad_norm": 2.178302478456175, "learning_rate": 9.856238216370024e-06, "loss": 0.7187, "step": 1559 }, { "epoch": 0.5074821080026025, "grad_norm": 1.9139723756799774, "learning_rate": 9.855787257802393e-06, "loss": 0.5502, "step": 1560 }, { "epoch": 0.5078074170461939, "grad_norm": 1.8314477054004454, "learning_rate": 9.85533560339721e-06, "loss": 0.6427, "step": 1561 }, { "epoch": 0.5081327260897853, "grad_norm": 2.077813365204314, "learning_rate": 9.854883253219196e-06, "loss": 0.7129, "step": 1562 }, { "epoch": 0.5084580351333767, "grad_norm": 1.777405819707308, "learning_rate": 9.854430207333176e-06, "loss": 0.5208, "step": 1563 }, { "epoch": 0.5087833441769681, "grad_norm": 2.1993411101372033, "learning_rate": 9.853976465804072e-06, "loss": 0.6015, "step": 1564 }, { "epoch": 0.5091086532205595, "grad_norm": 1.9251116847380554, "learning_rate": 9.853522028696902e-06, "loss": 0.5876, "step": 1565 }, { "epoch": 0.5094339622641509, "grad_norm": 1.8473898762522927, "learning_rate": 9.853066896076791e-06, "loss": 0.6644, "step": 1566 }, { "epoch": 0.5097592713077423, "grad_norm": 1.9096998131298608, "learning_rate": 9.852611068008958e-06, "loss": 0.6147, "step": 1567 }, { "epoch": 0.5100845803513337, "grad_norm": 1.864870019452556, "learning_rate": 9.852154544558721e-06, "loss": 0.5907, "step": 1568 }, { "epoch": 0.5104098893949252, "grad_norm": 1.8638315672765273, "learning_rate": 9.851697325791505e-06, "loss": 0.606, "step": 1569 }, { "epoch": 0.5107351984385166, "grad_norm": 2.113120569614777, "learning_rate": 9.851239411772828e-06, "loss": 0.7331, "step": 1570 }, { "epoch": 0.511060507482108, "grad_norm": 2.006356484444145, "learning_rate": 9.850780802568308e-06, "loss": 0.6383, "step": 1571 }, { "epoch": 0.5113858165256994, "grad_norm": 1.6379513174572133, "learning_rate": 9.850321498243664e-06, "loss": 0.517, "step": 1572 }, { "epoch": 0.5117111255692909, "grad_norm": 1.7450338150533178, "learning_rate": 9.849861498864715e-06, "loss": 0.4819, "step": 1573 }, { "epoch": 0.5120364346128823, "grad_norm": 1.867486143760923, "learning_rate": 9.849400804497383e-06, "loss": 0.5964, "step": 1574 }, { "epoch": 0.5123617436564737, "grad_norm": 1.9141512428062362, "learning_rate": 9.848939415207678e-06, "loss": 0.6235, "step": 1575 }, { "epoch": 0.5126870527000651, "grad_norm": 1.9762632171562509, "learning_rate": 9.848477331061726e-06, "loss": 0.5922, "step": 1576 }, { "epoch": 0.5130123617436565, "grad_norm": 1.9534920416771429, "learning_rate": 9.848014552125738e-06, "loss": 0.5756, "step": 1577 }, { "epoch": 0.5133376707872479, "grad_norm": 2.098092516523483, "learning_rate": 9.847551078466032e-06, "loss": 0.6987, "step": 1578 }, { "epoch": 0.5136629798308393, "grad_norm": 1.8229917646534681, "learning_rate": 9.847086910149024e-06, "loss": 0.5949, "step": 1579 }, { "epoch": 0.5139882888744307, "grad_norm": 2.365067303135434, "learning_rate": 9.846622047241232e-06, "loss": 0.6845, "step": 1580 }, { "epoch": 0.5143135979180221, "grad_norm": 2.2458036331244253, "learning_rate": 9.846156489809267e-06, "loss": 0.6742, "step": 1581 }, { "epoch": 0.5146389069616135, "grad_norm": 1.992344868558257, "learning_rate": 9.845690237919848e-06, "loss": 0.5906, "step": 1582 }, { "epoch": 0.5149642160052049, "grad_norm": 2.371292935149425, "learning_rate": 9.845223291639784e-06, "loss": 0.8141, "step": 1583 }, { "epoch": 0.5152895250487963, "grad_norm": 1.9651740891885743, "learning_rate": 9.844755651035996e-06, "loss": 0.6583, "step": 1584 }, { "epoch": 0.5156148340923877, "grad_norm": 1.6961563032174547, "learning_rate": 9.844287316175491e-06, "loss": 0.5311, "step": 1585 }, { "epoch": 0.5159401431359791, "grad_norm": 2.016879370119817, "learning_rate": 9.843818287125383e-06, "loss": 0.555, "step": 1586 }, { "epoch": 0.5162654521795705, "grad_norm": 1.9006732985281514, "learning_rate": 9.843348563952885e-06, "loss": 0.6369, "step": 1587 }, { "epoch": 0.516590761223162, "grad_norm": 2.0086704933699684, "learning_rate": 9.84287814672531e-06, "loss": 0.5608, "step": 1588 }, { "epoch": 0.5169160702667535, "grad_norm": 2.111646156350438, "learning_rate": 9.842407035510067e-06, "loss": 0.6426, "step": 1589 }, { "epoch": 0.5172413793103449, "grad_norm": 1.8785250324897003, "learning_rate": 9.841935230374668e-06, "loss": 0.5903, "step": 1590 }, { "epoch": 0.5175666883539363, "grad_norm": 1.732533474351391, "learning_rate": 9.841462731386722e-06, "loss": 0.5696, "step": 1591 }, { "epoch": 0.5178919973975277, "grad_norm": 1.9499458790109931, "learning_rate": 9.840989538613937e-06, "loss": 0.5955, "step": 1592 }, { "epoch": 0.5182173064411191, "grad_norm": 1.8930691643009883, "learning_rate": 9.840515652124125e-06, "loss": 0.62, "step": 1593 }, { "epoch": 0.5185426154847105, "grad_norm": 1.9921871960506699, "learning_rate": 9.840041071985193e-06, "loss": 0.5615, "step": 1594 }, { "epoch": 0.5188679245283019, "grad_norm": 2.002345739446018, "learning_rate": 9.839565798265146e-06, "loss": 0.5826, "step": 1595 }, { "epoch": 0.5191932335718933, "grad_norm": 2.0918876758483447, "learning_rate": 9.839089831032095e-06, "loss": 0.6528, "step": 1596 }, { "epoch": 0.5195185426154847, "grad_norm": 1.8841557212964302, "learning_rate": 9.838613170354248e-06, "loss": 0.5809, "step": 1597 }, { "epoch": 0.5198438516590761, "grad_norm": 1.9532562333208598, "learning_rate": 9.838135816299902e-06, "loss": 0.6657, "step": 1598 }, { "epoch": 0.5201691607026675, "grad_norm": 2.1747268934832396, "learning_rate": 9.837657768937471e-06, "loss": 0.7517, "step": 1599 }, { "epoch": 0.5204944697462589, "grad_norm": 1.8085255982706472, "learning_rate": 9.837179028335457e-06, "loss": 0.5604, "step": 1600 }, { "epoch": 0.5208197787898503, "grad_norm": 2.0923474765252275, "learning_rate": 9.836699594562462e-06, "loss": 0.6062, "step": 1601 }, { "epoch": 0.5211450878334418, "grad_norm": 1.9564850799015883, "learning_rate": 9.83621946768719e-06, "loss": 0.6333, "step": 1602 }, { "epoch": 0.5214703968770332, "grad_norm": 2.0057892982621865, "learning_rate": 9.835738647778445e-06, "loss": 0.5728, "step": 1603 }, { "epoch": 0.5217957059206246, "grad_norm": 1.9421597080882476, "learning_rate": 9.835257134905126e-06, "loss": 0.642, "step": 1604 }, { "epoch": 0.522121014964216, "grad_norm": 2.2307785602531434, "learning_rate": 9.834774929136237e-06, "loss": 0.7195, "step": 1605 }, { "epoch": 0.5224463240078074, "grad_norm": 2.4836102339043062, "learning_rate": 9.834292030540876e-06, "loss": 0.6841, "step": 1606 }, { "epoch": 0.5227716330513988, "grad_norm": 1.9003288066210897, "learning_rate": 9.833808439188244e-06, "loss": 0.5921, "step": 1607 }, { "epoch": 0.5230969420949902, "grad_norm": 1.9769939551257434, "learning_rate": 9.83332415514764e-06, "loss": 0.6587, "step": 1608 }, { "epoch": 0.5234222511385817, "grad_norm": 1.8800204517780954, "learning_rate": 9.832839178488463e-06, "loss": 0.6589, "step": 1609 }, { "epoch": 0.523747560182173, "grad_norm": 2.0870488048054985, "learning_rate": 9.83235350928021e-06, "loss": 0.5723, "step": 1610 }, { "epoch": 0.5240728692257645, "grad_norm": 2.2230935129371545, "learning_rate": 9.831867147592475e-06, "loss": 0.7312, "step": 1611 }, { "epoch": 0.5243981782693559, "grad_norm": 2.105832971598692, "learning_rate": 9.831380093494957e-06, "loss": 0.6768, "step": 1612 }, { "epoch": 0.5247234873129473, "grad_norm": 2.2527550230459825, "learning_rate": 9.830892347057453e-06, "loss": 0.7113, "step": 1613 }, { "epoch": 0.5250487963565387, "grad_norm": 2.4010170349637843, "learning_rate": 9.830403908349852e-06, "loss": 0.6838, "step": 1614 }, { "epoch": 0.5253741054001301, "grad_norm": 2.07235211438816, "learning_rate": 9.82991477744215e-06, "loss": 0.6708, "step": 1615 }, { "epoch": 0.5256994144437215, "grad_norm": 1.872397512994976, "learning_rate": 9.829424954404441e-06, "loss": 0.6277, "step": 1616 }, { "epoch": 0.526024723487313, "grad_norm": 1.9506666483448978, "learning_rate": 9.828934439306916e-06, "loss": 0.5089, "step": 1617 }, { "epoch": 0.5263500325309044, "grad_norm": 2.341314491118769, "learning_rate": 9.828443232219866e-06, "loss": 0.7279, "step": 1618 }, { "epoch": 0.5266753415744958, "grad_norm": 2.3005878780651856, "learning_rate": 9.827951333213684e-06, "loss": 0.6088, "step": 1619 }, { "epoch": 0.5270006506180872, "grad_norm": 1.82436359585127, "learning_rate": 9.827458742358854e-06, "loss": 0.5826, "step": 1620 }, { "epoch": 0.5273259596616786, "grad_norm": 2.157205857886732, "learning_rate": 9.826965459725968e-06, "loss": 0.5661, "step": 1621 }, { "epoch": 0.52765126870527, "grad_norm": 1.9519915079805288, "learning_rate": 9.826471485385714e-06, "loss": 0.5954, "step": 1622 }, { "epoch": 0.5279765777488614, "grad_norm": 2.0896980836065873, "learning_rate": 9.825976819408878e-06, "loss": 0.7118, "step": 1623 }, { "epoch": 0.5283018867924528, "grad_norm": 2.360526278249054, "learning_rate": 9.825481461866347e-06, "loss": 0.6977, "step": 1624 }, { "epoch": 0.5286271958360442, "grad_norm": 1.8437203165503857, "learning_rate": 9.824985412829104e-06, "loss": 0.5961, "step": 1625 }, { "epoch": 0.5289525048796356, "grad_norm": 2.047480303520928, "learning_rate": 9.824488672368236e-06, "loss": 0.6883, "step": 1626 }, { "epoch": 0.529277813923227, "grad_norm": 1.994176725971946, "learning_rate": 9.823991240554922e-06, "loss": 0.6041, "step": 1627 }, { "epoch": 0.5296031229668184, "grad_norm": 2.04462328448344, "learning_rate": 9.82349311746045e-06, "loss": 0.6449, "step": 1628 }, { "epoch": 0.5299284320104098, "grad_norm": 2.227664705762709, "learning_rate": 9.822994303156196e-06, "loss": 0.6519, "step": 1629 }, { "epoch": 0.5302537410540012, "grad_norm": 2.2244887089878, "learning_rate": 9.822494797713644e-06, "loss": 0.7297, "step": 1630 }, { "epoch": 0.5305790500975928, "grad_norm": 1.8368155223128908, "learning_rate": 9.821994601204372e-06, "loss": 0.6216, "step": 1631 }, { "epoch": 0.5309043591411842, "grad_norm": 2.032141111145951, "learning_rate": 9.821493713700058e-06, "loss": 0.6138, "step": 1632 }, { "epoch": 0.5312296681847756, "grad_norm": 1.8446231639872155, "learning_rate": 9.82099213527248e-06, "loss": 0.6047, "step": 1633 }, { "epoch": 0.531554977228367, "grad_norm": 2.203153637242501, "learning_rate": 9.820489865993516e-06, "loss": 0.5773, "step": 1634 }, { "epoch": 0.5318802862719584, "grad_norm": 2.0412831378973406, "learning_rate": 9.81998690593514e-06, "loss": 0.5767, "step": 1635 }, { "epoch": 0.5322055953155498, "grad_norm": 2.424371142991608, "learning_rate": 9.819483255169426e-06, "loss": 0.6855, "step": 1636 }, { "epoch": 0.5325309043591412, "grad_norm": 2.021047229769061, "learning_rate": 9.818978913768547e-06, "loss": 0.5943, "step": 1637 }, { "epoch": 0.5328562134027326, "grad_norm": 1.9024517429762526, "learning_rate": 9.81847388180478e-06, "loss": 0.6025, "step": 1638 }, { "epoch": 0.533181522446324, "grad_norm": 1.947260054289786, "learning_rate": 9.817968159350491e-06, "loss": 0.6422, "step": 1639 }, { "epoch": 0.5335068314899154, "grad_norm": 1.959494724506767, "learning_rate": 9.817461746478151e-06, "loss": 0.6724, "step": 1640 }, { "epoch": 0.5338321405335068, "grad_norm": 1.8375044380968284, "learning_rate": 9.816954643260334e-06, "loss": 0.6201, "step": 1641 }, { "epoch": 0.5341574495770982, "grad_norm": 1.851144908054521, "learning_rate": 9.816446849769703e-06, "loss": 0.5422, "step": 1642 }, { "epoch": 0.5344827586206896, "grad_norm": 2.062138524796001, "learning_rate": 9.815938366079028e-06, "loss": 0.6811, "step": 1643 }, { "epoch": 0.534808067664281, "grad_norm": 1.968938370519247, "learning_rate": 9.815429192261175e-06, "loss": 0.612, "step": 1644 }, { "epoch": 0.5351333767078725, "grad_norm": 1.9012433600075607, "learning_rate": 9.814919328389107e-06, "loss": 0.5216, "step": 1645 }, { "epoch": 0.5354586857514639, "grad_norm": 1.7788406502949063, "learning_rate": 9.81440877453589e-06, "loss": 0.5171, "step": 1646 }, { "epoch": 0.5357839947950553, "grad_norm": 1.7510496714608852, "learning_rate": 9.813897530774683e-06, "loss": 0.574, "step": 1647 }, { "epoch": 0.5361093038386467, "grad_norm": 2.0585179966182428, "learning_rate": 9.813385597178752e-06, "loss": 0.7117, "step": 1648 }, { "epoch": 0.5364346128822381, "grad_norm": 2.077149992103565, "learning_rate": 9.812872973821457e-06, "loss": 0.7384, "step": 1649 }, { "epoch": 0.5367599219258296, "grad_norm": 2.0109511295144546, "learning_rate": 9.812359660776254e-06, "loss": 0.5535, "step": 1650 }, { "epoch": 0.537085230969421, "grad_norm": 2.075706172889847, "learning_rate": 9.811845658116703e-06, "loss": 0.6346, "step": 1651 }, { "epoch": 0.5374105400130124, "grad_norm": 1.8353883604287764, "learning_rate": 9.811330965916462e-06, "loss": 0.6232, "step": 1652 }, { "epoch": 0.5377358490566038, "grad_norm": 1.9613763045214538, "learning_rate": 9.810815584249285e-06, "loss": 0.699, "step": 1653 }, { "epoch": 0.5380611581001952, "grad_norm": 2.1879937738499073, "learning_rate": 9.810299513189028e-06, "loss": 0.6717, "step": 1654 }, { "epoch": 0.5383864671437866, "grad_norm": 1.7154016919007056, "learning_rate": 9.809782752809644e-06, "loss": 0.5605, "step": 1655 }, { "epoch": 0.538711776187378, "grad_norm": 1.9690014787623948, "learning_rate": 9.809265303185185e-06, "loss": 0.6255, "step": 1656 }, { "epoch": 0.5390370852309694, "grad_norm": 1.7893641469136732, "learning_rate": 9.8087471643898e-06, "loss": 0.5781, "step": 1657 }, { "epoch": 0.5393623942745608, "grad_norm": 1.7083921799375905, "learning_rate": 9.80822833649774e-06, "loss": 0.5802, "step": 1658 }, { "epoch": 0.5396877033181522, "grad_norm": 2.1539310134968854, "learning_rate": 9.807708819583355e-06, "loss": 0.739, "step": 1659 }, { "epoch": 0.5400130123617437, "grad_norm": 2.257226860769454, "learning_rate": 9.80718861372109e-06, "loss": 0.6651, "step": 1660 }, { "epoch": 0.5403383214053351, "grad_norm": 1.7102691073374063, "learning_rate": 9.806667718985491e-06, "loss": 0.5157, "step": 1661 }, { "epoch": 0.5406636304489265, "grad_norm": 2.0294946543559287, "learning_rate": 9.806146135451204e-06, "loss": 0.6701, "step": 1662 }, { "epoch": 0.5409889394925179, "grad_norm": 1.722032528156316, "learning_rate": 9.805623863192972e-06, "loss": 0.5381, "step": 1663 }, { "epoch": 0.5413142485361093, "grad_norm": 2.1324680864990673, "learning_rate": 9.805100902285635e-06, "loss": 0.6326, "step": 1664 }, { "epoch": 0.5416395575797007, "grad_norm": 1.7252167522019892, "learning_rate": 9.804577252804137e-06, "loss": 0.5424, "step": 1665 }, { "epoch": 0.5419648666232921, "grad_norm": 1.986269249529255, "learning_rate": 9.804052914823513e-06, "loss": 0.6067, "step": 1666 }, { "epoch": 0.5422901756668835, "grad_norm": 1.9630536766186282, "learning_rate": 9.803527888418905e-06, "loss": 0.4999, "step": 1667 }, { "epoch": 0.5426154847104749, "grad_norm": 2.058753827346341, "learning_rate": 9.803002173665549e-06, "loss": 0.7709, "step": 1668 }, { "epoch": 0.5429407937540663, "grad_norm": 1.9832673781571986, "learning_rate": 9.802475770638774e-06, "loss": 0.6746, "step": 1669 }, { "epoch": 0.5432661027976577, "grad_norm": 1.9839304862182179, "learning_rate": 9.801948679414024e-06, "loss": 0.7149, "step": 1670 }, { "epoch": 0.5435914118412492, "grad_norm": 2.0968122450820292, "learning_rate": 9.801420900066824e-06, "loss": 0.6585, "step": 1671 }, { "epoch": 0.5439167208848406, "grad_norm": 1.9863829888702655, "learning_rate": 9.800892432672811e-06, "loss": 0.5957, "step": 1672 }, { "epoch": 0.544242029928432, "grad_norm": 2.0039043796476523, "learning_rate": 9.800363277307708e-06, "loss": 0.5799, "step": 1673 }, { "epoch": 0.5445673389720235, "grad_norm": 2.049564284950728, "learning_rate": 9.799833434047349e-06, "loss": 0.6072, "step": 1674 }, { "epoch": 0.5448926480156149, "grad_norm": 2.270767608963717, "learning_rate": 9.799302902967657e-06, "loss": 0.6124, "step": 1675 }, { "epoch": 0.5452179570592063, "grad_norm": 2.087374389805558, "learning_rate": 9.798771684144662e-06, "loss": 0.6538, "step": 1676 }, { "epoch": 0.5455432661027977, "grad_norm": 1.9226135666891713, "learning_rate": 9.798239777654481e-06, "loss": 0.653, "step": 1677 }, { "epoch": 0.5458685751463891, "grad_norm": 2.12394516726047, "learning_rate": 9.797707183573344e-06, "loss": 0.6605, "step": 1678 }, { "epoch": 0.5461938841899805, "grad_norm": 1.809381131101202, "learning_rate": 9.797173901977568e-06, "loss": 0.6256, "step": 1679 }, { "epoch": 0.5465191932335719, "grad_norm": 1.8838168874668415, "learning_rate": 9.796639932943574e-06, "loss": 0.558, "step": 1680 }, { "epoch": 0.5468445022771633, "grad_norm": 1.972298122818047, "learning_rate": 9.796105276547878e-06, "loss": 0.5805, "step": 1681 }, { "epoch": 0.5471698113207547, "grad_norm": 2.1866830170353384, "learning_rate": 9.795569932867099e-06, "loss": 0.6892, "step": 1682 }, { "epoch": 0.5474951203643461, "grad_norm": 1.805249523548903, "learning_rate": 9.79503390197795e-06, "loss": 0.626, "step": 1683 }, { "epoch": 0.5478204294079375, "grad_norm": 1.912495772825779, "learning_rate": 9.794497183957248e-06, "loss": 0.6238, "step": 1684 }, { "epoch": 0.5481457384515289, "grad_norm": 1.9226481213707147, "learning_rate": 9.793959778881904e-06, "loss": 0.6317, "step": 1685 }, { "epoch": 0.5484710474951203, "grad_norm": 1.8915820874182392, "learning_rate": 9.793421686828923e-06, "loss": 0.5836, "step": 1686 }, { "epoch": 0.5487963565387117, "grad_norm": 1.8947357507008553, "learning_rate": 9.792882907875422e-06, "loss": 0.5008, "step": 1687 }, { "epoch": 0.5491216655823032, "grad_norm": 2.4179527565630186, "learning_rate": 9.792343442098603e-06, "loss": 0.7561, "step": 1688 }, { "epoch": 0.5494469746258946, "grad_norm": 1.8464487961302247, "learning_rate": 9.791803289575774e-06, "loss": 0.6045, "step": 1689 }, { "epoch": 0.549772283669486, "grad_norm": 1.9006794464842045, "learning_rate": 9.791262450384338e-06, "loss": 0.5961, "step": 1690 }, { "epoch": 0.5500975927130775, "grad_norm": 2.03571514726117, "learning_rate": 9.7907209246018e-06, "loss": 0.5866, "step": 1691 }, { "epoch": 0.5504229017566689, "grad_norm": 1.7050321896919505, "learning_rate": 9.790178712305758e-06, "loss": 0.5435, "step": 1692 }, { "epoch": 0.5507482108002603, "grad_norm": 2.027893737591295, "learning_rate": 9.789635813573915e-06, "loss": 0.6746, "step": 1693 }, { "epoch": 0.5510735198438517, "grad_norm": 1.7533293218227326, "learning_rate": 9.789092228484065e-06, "loss": 0.5484, "step": 1694 }, { "epoch": 0.5513988288874431, "grad_norm": 2.2792298826240383, "learning_rate": 9.788547957114104e-06, "loss": 0.8352, "step": 1695 }, { "epoch": 0.5517241379310345, "grad_norm": 1.8406873316387393, "learning_rate": 9.78800299954203e-06, "loss": 0.5322, "step": 1696 }, { "epoch": 0.5520494469746259, "grad_norm": 1.7979082943220193, "learning_rate": 9.787457355845934e-06, "loss": 0.6349, "step": 1697 }, { "epoch": 0.5523747560182173, "grad_norm": 2.124776902855471, "learning_rate": 9.786911026104007e-06, "loss": 0.6055, "step": 1698 }, { "epoch": 0.5527000650618087, "grad_norm": 1.9057844536240423, "learning_rate": 9.786364010394537e-06, "loss": 0.4767, "step": 1699 }, { "epoch": 0.5530253741054001, "grad_norm": 2.0873877507953407, "learning_rate": 9.785816308795914e-06, "loss": 0.7194, "step": 1700 }, { "epoch": 0.5533506831489915, "grad_norm": 2.159432543513589, "learning_rate": 9.785267921386623e-06, "loss": 0.6787, "step": 1701 }, { "epoch": 0.5536759921925829, "grad_norm": 2.134603318231602, "learning_rate": 9.784718848245249e-06, "loss": 0.7091, "step": 1702 }, { "epoch": 0.5540013012361744, "grad_norm": 2.171459609243806, "learning_rate": 9.784169089450473e-06, "loss": 0.7092, "step": 1703 }, { "epoch": 0.5543266102797658, "grad_norm": 1.9786974574616638, "learning_rate": 9.783618645081077e-06, "loss": 0.6716, "step": 1704 }, { "epoch": 0.5546519193233572, "grad_norm": 1.9864262764427296, "learning_rate": 9.78306751521594e-06, "loss": 0.6088, "step": 1705 }, { "epoch": 0.5549772283669486, "grad_norm": 2.118381776518212, "learning_rate": 9.782515699934039e-06, "loss": 0.5234, "step": 1706 }, { "epoch": 0.55530253741054, "grad_norm": 1.779236727065701, "learning_rate": 9.78196319931445e-06, "loss": 0.535, "step": 1707 }, { "epoch": 0.5556278464541314, "grad_norm": 2.153645153561451, "learning_rate": 9.781410013436345e-06, "loss": 0.7861, "step": 1708 }, { "epoch": 0.5559531554977228, "grad_norm": 2.2769246829429592, "learning_rate": 9.780856142378997e-06, "loss": 0.7251, "step": 1709 }, { "epoch": 0.5562784645413142, "grad_norm": 2.0364599559983954, "learning_rate": 9.780301586221778e-06, "loss": 0.6325, "step": 1710 }, { "epoch": 0.5566037735849056, "grad_norm": 2.0977125436905877, "learning_rate": 9.779746345044155e-06, "loss": 0.7762, "step": 1711 }, { "epoch": 0.556929082628497, "grad_norm": 1.7420708004805119, "learning_rate": 9.779190418925693e-06, "loss": 0.6484, "step": 1712 }, { "epoch": 0.5572543916720885, "grad_norm": 2.23677943298724, "learning_rate": 9.778633807946056e-06, "loss": 0.7824, "step": 1713 }, { "epoch": 0.5575797007156799, "grad_norm": 1.8640917726796387, "learning_rate": 9.778076512185009e-06, "loss": 0.5445, "step": 1714 }, { "epoch": 0.5579050097592713, "grad_norm": 2.0530849928705788, "learning_rate": 9.777518531722414e-06, "loss": 0.6793, "step": 1715 }, { "epoch": 0.5582303188028627, "grad_norm": 2.0547900771786067, "learning_rate": 9.776959866638225e-06, "loss": 0.7451, "step": 1716 }, { "epoch": 0.5585556278464542, "grad_norm": 1.6871972400072635, "learning_rate": 9.776400517012506e-06, "loss": 0.5524, "step": 1717 }, { "epoch": 0.5588809368900456, "grad_norm": 1.8978590845697114, "learning_rate": 9.775840482925404e-06, "loss": 0.6454, "step": 1718 }, { "epoch": 0.559206245933637, "grad_norm": 1.6389236664120905, "learning_rate": 9.77527976445718e-06, "loss": 0.588, "step": 1719 }, { "epoch": 0.5595315549772284, "grad_norm": 1.5738069794276826, "learning_rate": 9.77471836168818e-06, "loss": 0.5163, "step": 1720 }, { "epoch": 0.5598568640208198, "grad_norm": 1.6879971832616192, "learning_rate": 9.774156274698856e-06, "loss": 0.6063, "step": 1721 }, { "epoch": 0.5601821730644112, "grad_norm": 2.044615818486332, "learning_rate": 9.773593503569756e-06, "loss": 0.6326, "step": 1722 }, { "epoch": 0.5605074821080026, "grad_norm": 1.9794913944641481, "learning_rate": 9.773030048381522e-06, "loss": 0.5659, "step": 1723 }, { "epoch": 0.560832791151594, "grad_norm": 1.9221145494005238, "learning_rate": 9.7724659092149e-06, "loss": 0.6076, "step": 1724 }, { "epoch": 0.5611581001951854, "grad_norm": 2.0289985030382316, "learning_rate": 9.771901086150733e-06, "loss": 0.7039, "step": 1725 }, { "epoch": 0.5614834092387768, "grad_norm": 1.8277934353007648, "learning_rate": 9.771335579269959e-06, "loss": 0.6195, "step": 1726 }, { "epoch": 0.5618087182823682, "grad_norm": 2.3262176917510513, "learning_rate": 9.770769388653614e-06, "loss": 0.7625, "step": 1727 }, { "epoch": 0.5621340273259596, "grad_norm": 1.8563311051410654, "learning_rate": 9.770202514382833e-06, "loss": 0.5834, "step": 1728 }, { "epoch": 0.562459336369551, "grad_norm": 1.9115057473901738, "learning_rate": 9.769634956538854e-06, "loss": 0.5665, "step": 1729 }, { "epoch": 0.5627846454131424, "grad_norm": 1.7832568900441015, "learning_rate": 9.769066715203006e-06, "loss": 0.5482, "step": 1730 }, { "epoch": 0.563109954456734, "grad_norm": 1.7712143678307788, "learning_rate": 9.768497790456716e-06, "loss": 0.6213, "step": 1731 }, { "epoch": 0.5634352635003254, "grad_norm": 2.0684814782425565, "learning_rate": 9.767928182381515e-06, "loss": 0.5771, "step": 1732 }, { "epoch": 0.5637605725439168, "grad_norm": 2.417056403497373, "learning_rate": 9.767357891059026e-06, "loss": 0.692, "step": 1733 }, { "epoch": 0.5640858815875082, "grad_norm": 1.5980268650398035, "learning_rate": 9.766786916570972e-06, "loss": 0.5613, "step": 1734 }, { "epoch": 0.5644111906310996, "grad_norm": 1.9448023287520402, "learning_rate": 9.766215258999176e-06, "loss": 0.5313, "step": 1735 }, { "epoch": 0.564736499674691, "grad_norm": 1.998601600810624, "learning_rate": 9.765642918425555e-06, "loss": 0.6001, "step": 1736 }, { "epoch": 0.5650618087182824, "grad_norm": 1.7977815040760636, "learning_rate": 9.765069894932126e-06, "loss": 0.7369, "step": 1737 }, { "epoch": 0.5653871177618738, "grad_norm": 1.9853228921307686, "learning_rate": 9.764496188601003e-06, "loss": 0.6149, "step": 1738 }, { "epoch": 0.5657124268054652, "grad_norm": 1.9225778406067053, "learning_rate": 9.763921799514401e-06, "loss": 0.6629, "step": 1739 }, { "epoch": 0.5660377358490566, "grad_norm": 1.6810263827819736, "learning_rate": 9.76334672775463e-06, "loss": 0.6264, "step": 1740 }, { "epoch": 0.566363044892648, "grad_norm": 1.9963908574972486, "learning_rate": 9.762770973404094e-06, "loss": 0.6713, "step": 1741 }, { "epoch": 0.5666883539362394, "grad_norm": 2.0770379927597697, "learning_rate": 9.762194536545304e-06, "loss": 0.6343, "step": 1742 }, { "epoch": 0.5670136629798308, "grad_norm": 1.9433761423408076, "learning_rate": 9.761617417260862e-06, "loss": 0.6084, "step": 1743 }, { "epoch": 0.5673389720234222, "grad_norm": 2.0348209159994712, "learning_rate": 9.761039615633468e-06, "loss": 0.6057, "step": 1744 }, { "epoch": 0.5676642810670136, "grad_norm": 1.6093305400381746, "learning_rate": 9.760461131745925e-06, "loss": 0.5064, "step": 1745 }, { "epoch": 0.5679895901106051, "grad_norm": 1.702634809174086, "learning_rate": 9.759881965681126e-06, "loss": 0.5607, "step": 1746 }, { "epoch": 0.5683148991541965, "grad_norm": 1.7457313761184743, "learning_rate": 9.75930211752207e-06, "loss": 0.581, "step": 1747 }, { "epoch": 0.5686402081977879, "grad_norm": 1.7484817529695753, "learning_rate": 9.758721587351846e-06, "loss": 0.5404, "step": 1748 }, { "epoch": 0.5689655172413793, "grad_norm": 1.906499399331572, "learning_rate": 9.758140375253644e-06, "loss": 0.5874, "step": 1749 }, { "epoch": 0.5692908262849707, "grad_norm": 2.1346269357961836, "learning_rate": 9.757558481310754e-06, "loss": 0.5971, "step": 1750 }, { "epoch": 0.5696161353285621, "grad_norm": 2.104032207148445, "learning_rate": 9.756975905606565e-06, "loss": 0.6146, "step": 1751 }, { "epoch": 0.5699414443721535, "grad_norm": 1.7912528399293963, "learning_rate": 9.756392648224553e-06, "loss": 0.5662, "step": 1752 }, { "epoch": 0.570266753415745, "grad_norm": 2.012569726056378, "learning_rate": 9.755808709248306e-06, "loss": 0.5987, "step": 1753 }, { "epoch": 0.5705920624593364, "grad_norm": 1.9496965837620466, "learning_rate": 9.755224088761499e-06, "loss": 0.6033, "step": 1754 }, { "epoch": 0.5709173715029278, "grad_norm": 1.679091442271442, "learning_rate": 9.754638786847911e-06, "loss": 0.6143, "step": 1755 }, { "epoch": 0.5712426805465192, "grad_norm": 1.7687607640444238, "learning_rate": 9.754052803591413e-06, "loss": 0.6102, "step": 1756 }, { "epoch": 0.5715679895901106, "grad_norm": 2.022987567048365, "learning_rate": 9.75346613907598e-06, "loss": 0.7289, "step": 1757 }, { "epoch": 0.571893298633702, "grad_norm": 1.7963794743656534, "learning_rate": 9.75287879338568e-06, "loss": 0.5729, "step": 1758 }, { "epoch": 0.5722186076772934, "grad_norm": 1.9534756913452094, "learning_rate": 9.75229076660468e-06, "loss": 0.5421, "step": 1759 }, { "epoch": 0.5725439167208849, "grad_norm": 1.7446551702706714, "learning_rate": 9.751702058817243e-06, "loss": 0.5428, "step": 1760 }, { "epoch": 0.5728692257644763, "grad_norm": 1.7445363044674163, "learning_rate": 9.751112670107734e-06, "loss": 0.618, "step": 1761 }, { "epoch": 0.5731945348080677, "grad_norm": 1.758058734280378, "learning_rate": 9.750522600560613e-06, "loss": 0.5552, "step": 1762 }, { "epoch": 0.5735198438516591, "grad_norm": 1.7854736561553934, "learning_rate": 9.749931850260434e-06, "loss": 0.5934, "step": 1763 }, { "epoch": 0.5738451528952505, "grad_norm": 1.891523758052588, "learning_rate": 9.749340419291856e-06, "loss": 0.5834, "step": 1764 }, { "epoch": 0.5741704619388419, "grad_norm": 1.6793705110986812, "learning_rate": 9.748748307739626e-06, "loss": 0.6046, "step": 1765 }, { "epoch": 0.5744957709824333, "grad_norm": 1.7326568077656066, "learning_rate": 9.7481555156886e-06, "loss": 0.6005, "step": 1766 }, { "epoch": 0.5748210800260247, "grad_norm": 1.7492557792940067, "learning_rate": 9.747562043223721e-06, "loss": 0.5278, "step": 1767 }, { "epoch": 0.5751463890696161, "grad_norm": 2.1120701911141944, "learning_rate": 9.746967890430038e-06, "loss": 0.717, "step": 1768 }, { "epoch": 0.5754716981132075, "grad_norm": 2.0793634342830605, "learning_rate": 9.74637305739269e-06, "loss": 0.5727, "step": 1769 }, { "epoch": 0.5757970071567989, "grad_norm": 1.7307043939274496, "learning_rate": 9.74577754419692e-06, "loss": 0.5846, "step": 1770 }, { "epoch": 0.5761223162003903, "grad_norm": 2.124138190604576, "learning_rate": 9.745181350928062e-06, "loss": 0.7123, "step": 1771 }, { "epoch": 0.5764476252439817, "grad_norm": 1.988683204096316, "learning_rate": 9.744584477671553e-06, "loss": 0.6803, "step": 1772 }, { "epoch": 0.5767729342875731, "grad_norm": 2.109608369749957, "learning_rate": 9.743986924512925e-06, "loss": 0.5888, "step": 1773 }, { "epoch": 0.5770982433311646, "grad_norm": 1.8688898887566088, "learning_rate": 9.743388691537808e-06, "loss": 0.6336, "step": 1774 }, { "epoch": 0.5774235523747561, "grad_norm": 1.8112684698494397, "learning_rate": 9.742789778831929e-06, "loss": 0.5729, "step": 1775 }, { "epoch": 0.5777488614183475, "grad_norm": 1.7767411502568278, "learning_rate": 9.742190186481112e-06, "loss": 0.6157, "step": 1776 }, { "epoch": 0.5780741704619389, "grad_norm": 2.0715959251048712, "learning_rate": 9.741589914571282e-06, "loss": 0.7301, "step": 1777 }, { "epoch": 0.5783994795055303, "grad_norm": 2.050555028391527, "learning_rate": 9.740988963188453e-06, "loss": 0.6868, "step": 1778 }, { "epoch": 0.5787247885491217, "grad_norm": 1.5668922870371844, "learning_rate": 9.740387332418746e-06, "loss": 0.4736, "step": 1779 }, { "epoch": 0.5790500975927131, "grad_norm": 1.6325272018714427, "learning_rate": 9.739785022348373e-06, "loss": 0.4634, "step": 1780 }, { "epoch": 0.5793754066363045, "grad_norm": 1.9710105004539582, "learning_rate": 9.739182033063645e-06, "loss": 0.7595, "step": 1781 }, { "epoch": 0.5797007156798959, "grad_norm": 1.9796593680844252, "learning_rate": 9.738578364650975e-06, "loss": 0.6091, "step": 1782 }, { "epoch": 0.5800260247234873, "grad_norm": 2.0494606663439305, "learning_rate": 9.737974017196862e-06, "loss": 0.6146, "step": 1783 }, { "epoch": 0.5803513337670787, "grad_norm": 1.8486804543994644, "learning_rate": 9.737368990787917e-06, "loss": 0.6233, "step": 1784 }, { "epoch": 0.5806766428106701, "grad_norm": 1.995028798606057, "learning_rate": 9.736763285510836e-06, "loss": 0.6432, "step": 1785 }, { "epoch": 0.5810019518542615, "grad_norm": 1.6131638635980279, "learning_rate": 9.736156901452417e-06, "loss": 0.6568, "step": 1786 }, { "epoch": 0.5813272608978529, "grad_norm": 1.8580723469021738, "learning_rate": 9.735549838699557e-06, "loss": 0.6136, "step": 1787 }, { "epoch": 0.5816525699414443, "grad_norm": 2.1934712939463274, "learning_rate": 9.734942097339248e-06, "loss": 0.7256, "step": 1788 }, { "epoch": 0.5819778789850358, "grad_norm": 1.8891515577107016, "learning_rate": 9.734333677458579e-06, "loss": 0.6843, "step": 1789 }, { "epoch": 0.5823031880286272, "grad_norm": 1.829541175666889, "learning_rate": 9.73372457914474e-06, "loss": 0.5947, "step": 1790 }, { "epoch": 0.5826284970722186, "grad_norm": 2.0495508023714075, "learning_rate": 9.733114802485008e-06, "loss": 0.6872, "step": 1791 }, { "epoch": 0.58295380611581, "grad_norm": 1.9359407972467513, "learning_rate": 9.732504347566773e-06, "loss": 0.6832, "step": 1792 }, { "epoch": 0.5832791151594015, "grad_norm": 1.8814705973113912, "learning_rate": 9.731893214477507e-06, "loss": 0.6168, "step": 1793 }, { "epoch": 0.5836044242029929, "grad_norm": 1.8993117931713293, "learning_rate": 9.731281403304791e-06, "loss": 0.5388, "step": 1794 }, { "epoch": 0.5839297332465843, "grad_norm": 2.0911568218314613, "learning_rate": 9.730668914136294e-06, "loss": 0.6242, "step": 1795 }, { "epoch": 0.5842550422901757, "grad_norm": 1.923328378945653, "learning_rate": 9.730055747059789e-06, "loss": 0.5771, "step": 1796 }, { "epoch": 0.5845803513337671, "grad_norm": 1.81370370836487, "learning_rate": 9.72944190216314e-06, "loss": 0.6651, "step": 1797 }, { "epoch": 0.5849056603773585, "grad_norm": 1.6983467260917207, "learning_rate": 9.728827379534313e-06, "loss": 0.5099, "step": 1798 }, { "epoch": 0.5852309694209499, "grad_norm": 2.0195493208511306, "learning_rate": 9.728212179261371e-06, "loss": 0.5378, "step": 1799 }, { "epoch": 0.5855562784645413, "grad_norm": 1.9442226155233007, "learning_rate": 9.727596301432469e-06, "loss": 0.5269, "step": 1800 }, { "epoch": 0.5858815875081327, "grad_norm": 1.8972525721876234, "learning_rate": 9.726979746135868e-06, "loss": 0.5373, "step": 1801 }, { "epoch": 0.5862068965517241, "grad_norm": 2.506414963853503, "learning_rate": 9.726362513459916e-06, "loss": 0.6804, "step": 1802 }, { "epoch": 0.5865322055953156, "grad_norm": 1.9357559984130654, "learning_rate": 9.725744603493065e-06, "loss": 0.5438, "step": 1803 }, { "epoch": 0.586857514638907, "grad_norm": 1.9189480350391683, "learning_rate": 9.725126016323862e-06, "loss": 0.6129, "step": 1804 }, { "epoch": 0.5871828236824984, "grad_norm": 1.5896189224735224, "learning_rate": 9.72450675204095e-06, "loss": 0.5411, "step": 1805 }, { "epoch": 0.5875081327260898, "grad_norm": 1.7870976991948442, "learning_rate": 9.72388681073307e-06, "loss": 0.6842, "step": 1806 }, { "epoch": 0.5878334417696812, "grad_norm": 1.8091509340102527, "learning_rate": 9.72326619248906e-06, "loss": 0.5962, "step": 1807 }, { "epoch": 0.5881587508132726, "grad_norm": 1.726603765487954, "learning_rate": 9.722644897397858e-06, "loss": 0.5536, "step": 1808 }, { "epoch": 0.588484059856864, "grad_norm": 2.0555653234809705, "learning_rate": 9.722022925548491e-06, "loss": 0.6016, "step": 1809 }, { "epoch": 0.5888093689004554, "grad_norm": 2.00642605808431, "learning_rate": 9.72140027703009e-06, "loss": 0.6809, "step": 1810 }, { "epoch": 0.5891346779440468, "grad_norm": 2.023627193015717, "learning_rate": 9.720776951931885e-06, "loss": 0.7241, "step": 1811 }, { "epoch": 0.5894599869876382, "grad_norm": 2.1535498584114894, "learning_rate": 9.720152950343192e-06, "loss": 0.6047, "step": 1812 }, { "epoch": 0.5897852960312296, "grad_norm": 1.8285291680075433, "learning_rate": 9.719528272353436e-06, "loss": 0.6025, "step": 1813 }, { "epoch": 0.590110605074821, "grad_norm": 2.0298307637960704, "learning_rate": 9.718902918052132e-06, "loss": 0.6214, "step": 1814 }, { "epoch": 0.5904359141184125, "grad_norm": 2.16038320155592, "learning_rate": 9.718276887528892e-06, "loss": 0.6169, "step": 1815 }, { "epoch": 0.5907612231620039, "grad_norm": 1.8132891780304465, "learning_rate": 9.717650180873428e-06, "loss": 0.6179, "step": 1816 }, { "epoch": 0.5910865322055953, "grad_norm": 1.9286099463274395, "learning_rate": 9.717022798175549e-06, "loss": 0.7176, "step": 1817 }, { "epoch": 0.5914118412491868, "grad_norm": 1.760886631928857, "learning_rate": 9.716394739525158e-06, "loss": 0.5546, "step": 1818 }, { "epoch": 0.5917371502927782, "grad_norm": 1.9413610556647258, "learning_rate": 9.715766005012254e-06, "loss": 0.5621, "step": 1819 }, { "epoch": 0.5920624593363696, "grad_norm": 1.8971392647774026, "learning_rate": 9.715136594726941e-06, "loss": 0.7147, "step": 1820 }, { "epoch": 0.592387768379961, "grad_norm": 1.8471607085821828, "learning_rate": 9.714506508759408e-06, "loss": 0.7078, "step": 1821 }, { "epoch": 0.5927130774235524, "grad_norm": 2.0203835909045305, "learning_rate": 9.71387574719995e-06, "loss": 0.6072, "step": 1822 }, { "epoch": 0.5930383864671438, "grad_norm": 1.6491184504936625, "learning_rate": 9.713244310138955e-06, "loss": 0.545, "step": 1823 }, { "epoch": 0.5933636955107352, "grad_norm": 1.6663170235776452, "learning_rate": 9.712612197666906e-06, "loss": 0.5879, "step": 1824 }, { "epoch": 0.5936890045543266, "grad_norm": 1.6075456055817525, "learning_rate": 9.711979409874388e-06, "loss": 0.4527, "step": 1825 }, { "epoch": 0.594014313597918, "grad_norm": 1.8595172868839318, "learning_rate": 9.711345946852078e-06, "loss": 0.6133, "step": 1826 }, { "epoch": 0.5943396226415094, "grad_norm": 1.7598505422815272, "learning_rate": 9.710711808690754e-06, "loss": 0.6022, "step": 1827 }, { "epoch": 0.5946649316851008, "grad_norm": 1.7996066863049298, "learning_rate": 9.710076995481288e-06, "loss": 0.5663, "step": 1828 }, { "epoch": 0.5949902407286922, "grad_norm": 1.955950804713439, "learning_rate": 9.709441507314645e-06, "loss": 0.6974, "step": 1829 }, { "epoch": 0.5953155497722836, "grad_norm": 1.9332687976165301, "learning_rate": 9.708805344281898e-06, "loss": 0.5731, "step": 1830 }, { "epoch": 0.595640858815875, "grad_norm": 1.9734834563697943, "learning_rate": 9.708168506474203e-06, "loss": 0.6444, "step": 1831 }, { "epoch": 0.5959661678594665, "grad_norm": 1.7445267936774145, "learning_rate": 9.707530993982822e-06, "loss": 0.5861, "step": 1832 }, { "epoch": 0.596291476903058, "grad_norm": 1.7132301670530545, "learning_rate": 9.706892806899112e-06, "loss": 0.5195, "step": 1833 }, { "epoch": 0.5966167859466494, "grad_norm": 1.8546503382554, "learning_rate": 9.706253945314523e-06, "loss": 0.7328, "step": 1834 }, { "epoch": 0.5969420949902408, "grad_norm": 1.7152071201790662, "learning_rate": 9.705614409320607e-06, "loss": 0.6771, "step": 1835 }, { "epoch": 0.5972674040338322, "grad_norm": 1.9462689480486943, "learning_rate": 9.704974199009009e-06, "loss": 0.6771, "step": 1836 }, { "epoch": 0.5975927130774236, "grad_norm": 1.616488178551497, "learning_rate": 9.70433331447147e-06, "loss": 0.6153, "step": 1837 }, { "epoch": 0.597918022121015, "grad_norm": 1.6362488730600062, "learning_rate": 9.703691755799832e-06, "loss": 0.5797, "step": 1838 }, { "epoch": 0.5982433311646064, "grad_norm": 1.914822332052564, "learning_rate": 9.70304952308603e-06, "loss": 0.6849, "step": 1839 }, { "epoch": 0.5985686402081978, "grad_norm": 1.8152255669041122, "learning_rate": 9.702406616422093e-06, "loss": 0.5739, "step": 1840 }, { "epoch": 0.5988939492517892, "grad_norm": 2.2285786852753637, "learning_rate": 9.701763035900153e-06, "loss": 0.7375, "step": 1841 }, { "epoch": 0.5992192582953806, "grad_norm": 2.2963254169825933, "learning_rate": 9.701118781612437e-06, "loss": 0.7001, "step": 1842 }, { "epoch": 0.599544567338972, "grad_norm": 1.6015150894213626, "learning_rate": 9.700473853651264e-06, "loss": 0.515, "step": 1843 }, { "epoch": 0.5998698763825634, "grad_norm": 2.0101503264883527, "learning_rate": 9.699828252109057e-06, "loss": 0.5566, "step": 1844 }, { "epoch": 0.6001951854261548, "grad_norm": 1.6825334750902883, "learning_rate": 9.699181977078324e-06, "loss": 0.4967, "step": 1845 }, { "epoch": 0.6005204944697463, "grad_norm": 1.6375202761821503, "learning_rate": 9.698535028651682e-06, "loss": 0.4415, "step": 1846 }, { "epoch": 0.6008458035133377, "grad_norm": 1.8306101571583284, "learning_rate": 9.697887406921838e-06, "loss": 0.5714, "step": 1847 }, { "epoch": 0.6011711125569291, "grad_norm": 2.064370911027012, "learning_rate": 9.697239111981595e-06, "loss": 0.6523, "step": 1848 }, { "epoch": 0.6014964216005205, "grad_norm": 1.8002135572346951, "learning_rate": 9.696590143923856e-06, "loss": 0.594, "step": 1849 }, { "epoch": 0.6018217306441119, "grad_norm": 2.1942028764164414, "learning_rate": 9.695940502841618e-06, "loss": 0.6918, "step": 1850 }, { "epoch": 0.6021470396877033, "grad_norm": 1.9048201527220079, "learning_rate": 9.695290188827975e-06, "loss": 0.6364, "step": 1851 }, { "epoch": 0.6024723487312947, "grad_norm": 1.9821153051176685, "learning_rate": 9.69463920197612e-06, "loss": 0.679, "step": 1852 }, { "epoch": 0.6027976577748861, "grad_norm": 1.9065821155399052, "learning_rate": 9.693987542379333e-06, "loss": 0.656, "step": 1853 }, { "epoch": 0.6031229668184775, "grad_norm": 2.1735808928959646, "learning_rate": 9.693335210131005e-06, "loss": 0.6794, "step": 1854 }, { "epoch": 0.603448275862069, "grad_norm": 1.9395745861172407, "learning_rate": 9.69268220532461e-06, "loss": 0.5906, "step": 1855 }, { "epoch": 0.6037735849056604, "grad_norm": 1.721334839898162, "learning_rate": 9.692028528053727e-06, "loss": 0.539, "step": 1856 }, { "epoch": 0.6040988939492518, "grad_norm": 1.799606236401057, "learning_rate": 9.69137417841203e-06, "loss": 0.6506, "step": 1857 }, { "epoch": 0.6044242029928432, "grad_norm": 2.160773491454479, "learning_rate": 9.690719156493283e-06, "loss": 0.7463, "step": 1858 }, { "epoch": 0.6047495120364346, "grad_norm": 1.917003276622555, "learning_rate": 9.690063462391353e-06, "loss": 0.717, "step": 1859 }, { "epoch": 0.605074821080026, "grad_norm": 1.5623326778300746, "learning_rate": 9.689407096200202e-06, "loss": 0.5025, "step": 1860 }, { "epoch": 0.6054001301236175, "grad_norm": 1.8811796373127438, "learning_rate": 9.688750058013887e-06, "loss": 0.6533, "step": 1861 }, { "epoch": 0.6057254391672089, "grad_norm": 1.8351009226749795, "learning_rate": 9.688092347926565e-06, "loss": 0.4681, "step": 1862 }, { "epoch": 0.6060507482108003, "grad_norm": 1.936187707964093, "learning_rate": 9.687433966032482e-06, "loss": 0.5689, "step": 1863 }, { "epoch": 0.6063760572543917, "grad_norm": 1.8026855420724097, "learning_rate": 9.686774912425987e-06, "loss": 0.6749, "step": 1864 }, { "epoch": 0.6067013662979831, "grad_norm": 2.0097691221333673, "learning_rate": 9.686115187201521e-06, "loss": 0.5693, "step": 1865 }, { "epoch": 0.6070266753415745, "grad_norm": 1.9893857776351058, "learning_rate": 9.685454790453626e-06, "loss": 0.7125, "step": 1866 }, { "epoch": 0.6073519843851659, "grad_norm": 1.972253058818607, "learning_rate": 9.684793722276936e-06, "loss": 0.6081, "step": 1867 }, { "epoch": 0.6076772934287573, "grad_norm": 1.77282891717712, "learning_rate": 9.68413198276618e-06, "loss": 0.5132, "step": 1868 }, { "epoch": 0.6080026024723487, "grad_norm": 1.7113717628535272, "learning_rate": 9.68346957201619e-06, "loss": 0.5846, "step": 1869 }, { "epoch": 0.6083279115159401, "grad_norm": 2.020716013804498, "learning_rate": 9.682806490121886e-06, "loss": 0.6828, "step": 1870 }, { "epoch": 0.6086532205595315, "grad_norm": 1.8465183157040497, "learning_rate": 9.68214273717829e-06, "loss": 0.6872, "step": 1871 }, { "epoch": 0.6089785296031229, "grad_norm": 1.7131582555804978, "learning_rate": 9.681478313280518e-06, "loss": 0.5665, "step": 1872 }, { "epoch": 0.6093038386467143, "grad_norm": 1.6987302758637448, "learning_rate": 9.680813218523783e-06, "loss": 0.5506, "step": 1873 }, { "epoch": 0.6096291476903057, "grad_norm": 1.9395273858427662, "learning_rate": 9.680147453003392e-06, "loss": 0.6791, "step": 1874 }, { "epoch": 0.6099544567338973, "grad_norm": 1.9844599183940597, "learning_rate": 9.679481016814752e-06, "loss": 0.6755, "step": 1875 }, { "epoch": 0.6102797657774887, "grad_norm": 1.9554764104902995, "learning_rate": 9.678813910053362e-06, "loss": 0.6715, "step": 1876 }, { "epoch": 0.6106050748210801, "grad_norm": 1.9079873351469077, "learning_rate": 9.67814613281482e-06, "loss": 0.5251, "step": 1877 }, { "epoch": 0.6109303838646715, "grad_norm": 1.790347479920413, "learning_rate": 9.677477685194816e-06, "loss": 0.562, "step": 1878 }, { "epoch": 0.6112556929082629, "grad_norm": 1.8215497040765507, "learning_rate": 9.676808567289143e-06, "loss": 0.5618, "step": 1879 }, { "epoch": 0.6115810019518543, "grad_norm": 1.7776665540964887, "learning_rate": 9.676138779193681e-06, "loss": 0.6535, "step": 1880 }, { "epoch": 0.6119063109954457, "grad_norm": 1.8819398710434059, "learning_rate": 9.675468321004418e-06, "loss": 0.6227, "step": 1881 }, { "epoch": 0.6122316200390371, "grad_norm": 2.15739041341808, "learning_rate": 9.674797192817424e-06, "loss": 0.6537, "step": 1882 }, { "epoch": 0.6125569290826285, "grad_norm": 1.7942209350324747, "learning_rate": 9.674125394728878e-06, "loss": 0.5406, "step": 1883 }, { "epoch": 0.6128822381262199, "grad_norm": 1.9303584497387876, "learning_rate": 9.673452926835047e-06, "loss": 0.6909, "step": 1884 }, { "epoch": 0.6132075471698113, "grad_norm": 1.8418079588832947, "learning_rate": 9.672779789232294e-06, "loss": 0.6309, "step": 1885 }, { "epoch": 0.6135328562134027, "grad_norm": 1.6613067986343482, "learning_rate": 9.672105982017082e-06, "loss": 0.5447, "step": 1886 }, { "epoch": 0.6138581652569941, "grad_norm": 1.7695898481133305, "learning_rate": 9.671431505285968e-06, "loss": 0.653, "step": 1887 }, { "epoch": 0.6141834743005855, "grad_norm": 1.6792530168714015, "learning_rate": 9.670756359135605e-06, "loss": 0.5515, "step": 1888 }, { "epoch": 0.614508783344177, "grad_norm": 1.7876563271565777, "learning_rate": 9.670080543662742e-06, "loss": 0.6115, "step": 1889 }, { "epoch": 0.6148340923877684, "grad_norm": 1.931853624113656, "learning_rate": 9.669404058964222e-06, "loss": 0.6033, "step": 1890 }, { "epoch": 0.6151594014313598, "grad_norm": 1.9830048887501426, "learning_rate": 9.668726905136987e-06, "loss": 0.6681, "step": 1891 }, { "epoch": 0.6154847104749512, "grad_norm": 2.037214693346545, "learning_rate": 9.668049082278075e-06, "loss": 0.7096, "step": 1892 }, { "epoch": 0.6158100195185426, "grad_norm": 1.877582345996286, "learning_rate": 9.667370590484617e-06, "loss": 0.5482, "step": 1893 }, { "epoch": 0.616135328562134, "grad_norm": 1.9620519016335503, "learning_rate": 9.66669142985384e-06, "loss": 0.656, "step": 1894 }, { "epoch": 0.6164606376057254, "grad_norm": 1.927848141947784, "learning_rate": 9.666011600483071e-06, "loss": 0.669, "step": 1895 }, { "epoch": 0.6167859466493169, "grad_norm": 1.7867685173108434, "learning_rate": 9.665331102469728e-06, "loss": 0.5133, "step": 1896 }, { "epoch": 0.6171112556929083, "grad_norm": 1.9642118329530283, "learning_rate": 9.66464993591133e-06, "loss": 0.7259, "step": 1897 }, { "epoch": 0.6174365647364997, "grad_norm": 1.9643255000924116, "learning_rate": 9.663968100905483e-06, "loss": 0.7133, "step": 1898 }, { "epoch": 0.6177618737800911, "grad_norm": 1.6769831001528102, "learning_rate": 9.663285597549898e-06, "loss": 0.5547, "step": 1899 }, { "epoch": 0.6180871828236825, "grad_norm": 1.9218743623576489, "learning_rate": 9.66260242594238e-06, "loss": 0.6024, "step": 1900 }, { "epoch": 0.6184124918672739, "grad_norm": 1.9935012627489963, "learning_rate": 9.661918586180822e-06, "loss": 0.7069, "step": 1901 }, { "epoch": 0.6187378009108653, "grad_norm": 2.013616348690308, "learning_rate": 9.661234078363226e-06, "loss": 0.7495, "step": 1902 }, { "epoch": 0.6190631099544567, "grad_norm": 2.100790629270762, "learning_rate": 9.660548902587678e-06, "loss": 0.5994, "step": 1903 }, { "epoch": 0.6193884189980482, "grad_norm": 1.7195687888273838, "learning_rate": 9.659863058952364e-06, "loss": 0.558, "step": 1904 }, { "epoch": 0.6197137280416396, "grad_norm": 1.8231418790765572, "learning_rate": 9.659176547555567e-06, "loss": 0.5548, "step": 1905 }, { "epoch": 0.620039037085231, "grad_norm": 2.009788010743706, "learning_rate": 9.658489368495664e-06, "loss": 0.7493, "step": 1906 }, { "epoch": 0.6203643461288224, "grad_norm": 1.9500438242652554, "learning_rate": 9.65780152187113e-06, "loss": 0.5643, "step": 1907 }, { "epoch": 0.6206896551724138, "grad_norm": 1.6734614090073496, "learning_rate": 9.65711300778053e-06, "loss": 0.4895, "step": 1908 }, { "epoch": 0.6210149642160052, "grad_norm": 1.9050122299340333, "learning_rate": 9.656423826322532e-06, "loss": 0.6362, "step": 1909 }, { "epoch": 0.6213402732595966, "grad_norm": 2.147284218825127, "learning_rate": 9.655733977595897e-06, "loss": 0.688, "step": 1910 }, { "epoch": 0.621665582303188, "grad_norm": 1.7980554192239235, "learning_rate": 9.655043461699476e-06, "loss": 0.5359, "step": 1911 }, { "epoch": 0.6219908913467794, "grad_norm": 1.7489325140720862, "learning_rate": 9.654352278732225e-06, "loss": 0.6174, "step": 1912 }, { "epoch": 0.6223162003903708, "grad_norm": 1.7839697174020601, "learning_rate": 9.653660428793188e-06, "loss": 0.6177, "step": 1913 }, { "epoch": 0.6226415094339622, "grad_norm": 2.067819492930412, "learning_rate": 9.65296791198151e-06, "loss": 0.6633, "step": 1914 }, { "epoch": 0.6229668184775536, "grad_norm": 1.9368683522232293, "learning_rate": 9.652274728396426e-06, "loss": 0.7538, "step": 1915 }, { "epoch": 0.623292127521145, "grad_norm": 1.7914392324371837, "learning_rate": 9.651580878137274e-06, "loss": 0.5967, "step": 1916 }, { "epoch": 0.6236174365647364, "grad_norm": 2.292046864618702, "learning_rate": 9.65088636130348e-06, "loss": 0.6707, "step": 1917 }, { "epoch": 0.623942745608328, "grad_norm": 1.7163129583077013, "learning_rate": 9.65019117799457e-06, "loss": 0.5004, "step": 1918 }, { "epoch": 0.6242680546519194, "grad_norm": 1.97381750088539, "learning_rate": 9.649495328310163e-06, "loss": 0.7161, "step": 1919 }, { "epoch": 0.6245933636955108, "grad_norm": 1.9167849241610977, "learning_rate": 9.648798812349975e-06, "loss": 0.5848, "step": 1920 }, { "epoch": 0.6249186727391022, "grad_norm": 1.8601268667476163, "learning_rate": 9.64810163021382e-06, "loss": 0.5494, "step": 1921 }, { "epoch": 0.6252439817826936, "grad_norm": 2.0696194514505555, "learning_rate": 9.647403782001601e-06, "loss": 0.4975, "step": 1922 }, { "epoch": 0.625569290826285, "grad_norm": 2.0500555130522082, "learning_rate": 9.646705267813321e-06, "loss": 0.7037, "step": 1923 }, { "epoch": 0.6258945998698764, "grad_norm": 2.189614492490939, "learning_rate": 9.646006087749078e-06, "loss": 0.7414, "step": 1924 }, { "epoch": 0.6262199089134678, "grad_norm": 2.0841825398092344, "learning_rate": 9.645306241909064e-06, "loss": 0.7117, "step": 1925 }, { "epoch": 0.6265452179570592, "grad_norm": 2.1992630908603985, "learning_rate": 9.64460573039357e-06, "loss": 0.6552, "step": 1926 }, { "epoch": 0.6268705270006506, "grad_norm": 1.7094525395283766, "learning_rate": 9.643904553302979e-06, "loss": 0.5201, "step": 1927 }, { "epoch": 0.627195836044242, "grad_norm": 1.9919511370283431, "learning_rate": 9.643202710737768e-06, "loss": 0.6217, "step": 1928 }, { "epoch": 0.6275211450878334, "grad_norm": 1.7287832383726247, "learning_rate": 9.642500202798513e-06, "loss": 0.5714, "step": 1929 }, { "epoch": 0.6278464541314248, "grad_norm": 2.072879557264934, "learning_rate": 9.641797029585883e-06, "loss": 0.7445, "step": 1930 }, { "epoch": 0.6281717631750162, "grad_norm": 1.9133398392474055, "learning_rate": 9.641093191200646e-06, "loss": 0.6827, "step": 1931 }, { "epoch": 0.6284970722186076, "grad_norm": 1.918568731133241, "learning_rate": 9.640388687743659e-06, "loss": 0.584, "step": 1932 }, { "epoch": 0.6288223812621991, "grad_norm": 2.0050840322553576, "learning_rate": 9.639683519315877e-06, "loss": 0.7179, "step": 1933 }, { "epoch": 0.6291476903057905, "grad_norm": 1.8370253735325404, "learning_rate": 9.638977686018356e-06, "loss": 0.5883, "step": 1934 }, { "epoch": 0.629472999349382, "grad_norm": 1.8888013073454875, "learning_rate": 9.638271187952238e-06, "loss": 0.6685, "step": 1935 }, { "epoch": 0.6297983083929733, "grad_norm": 2.1375817151797305, "learning_rate": 9.637564025218767e-06, "loss": 0.7221, "step": 1936 }, { "epoch": 0.6301236174365648, "grad_norm": 2.007313127807263, "learning_rate": 9.63685619791928e-06, "loss": 0.6289, "step": 1937 }, { "epoch": 0.6304489264801562, "grad_norm": 1.992312348899614, "learning_rate": 9.636147706155208e-06, "loss": 0.6269, "step": 1938 }, { "epoch": 0.6307742355237476, "grad_norm": 1.8915569899417592, "learning_rate": 9.635438550028078e-06, "loss": 0.7392, "step": 1939 }, { "epoch": 0.631099544567339, "grad_norm": 1.8753163051784343, "learning_rate": 9.634728729639513e-06, "loss": 0.6488, "step": 1940 }, { "epoch": 0.6314248536109304, "grad_norm": 1.68157223601388, "learning_rate": 9.634018245091232e-06, "loss": 0.6095, "step": 1941 }, { "epoch": 0.6317501626545218, "grad_norm": 1.7851719054234043, "learning_rate": 9.633307096485046e-06, "loss": 0.5296, "step": 1942 }, { "epoch": 0.6320754716981132, "grad_norm": 1.404109467140235, "learning_rate": 9.632595283922863e-06, "loss": 0.4524, "step": 1943 }, { "epoch": 0.6324007807417046, "grad_norm": 1.8680839038627695, "learning_rate": 9.63188280750669e-06, "loss": 0.7417, "step": 1944 }, { "epoch": 0.632726089785296, "grad_norm": 1.960077077616839, "learning_rate": 9.63116966733862e-06, "loss": 0.6107, "step": 1945 }, { "epoch": 0.6330513988288874, "grad_norm": 1.823628281083946, "learning_rate": 9.63045586352085e-06, "loss": 0.5423, "step": 1946 }, { "epoch": 0.6333767078724789, "grad_norm": 1.9892503628574607, "learning_rate": 9.629741396155667e-06, "loss": 0.6127, "step": 1947 }, { "epoch": 0.6337020169160703, "grad_norm": 2.1321735894333136, "learning_rate": 9.629026265345456e-06, "loss": 0.7083, "step": 1948 }, { "epoch": 0.6340273259596617, "grad_norm": 1.822129258822322, "learning_rate": 9.628310471192693e-06, "loss": 0.5975, "step": 1949 }, { "epoch": 0.6343526350032531, "grad_norm": 2.001026578561931, "learning_rate": 9.627594013799955e-06, "loss": 0.6095, "step": 1950 }, { "epoch": 0.6346779440468445, "grad_norm": 1.8209408520328783, "learning_rate": 9.62687689326991e-06, "loss": 0.5668, "step": 1951 }, { "epoch": 0.6350032530904359, "grad_norm": 1.940437489763354, "learning_rate": 9.62615910970532e-06, "loss": 0.7674, "step": 1952 }, { "epoch": 0.6353285621340273, "grad_norm": 2.209758256502193, "learning_rate": 9.625440663209048e-06, "loss": 0.676, "step": 1953 }, { "epoch": 0.6356538711776187, "grad_norm": 1.890018659322208, "learning_rate": 9.624721553884042e-06, "loss": 0.6498, "step": 1954 }, { "epoch": 0.6359791802212101, "grad_norm": 1.9629218261513601, "learning_rate": 9.624001781833354e-06, "loss": 0.6129, "step": 1955 }, { "epoch": 0.6363044892648015, "grad_norm": 1.833841995298733, "learning_rate": 9.623281347160129e-06, "loss": 0.5562, "step": 1956 }, { "epoch": 0.636629798308393, "grad_norm": 1.9889319238047152, "learning_rate": 9.622560249967602e-06, "loss": 0.6528, "step": 1957 }, { "epoch": 0.6369551073519844, "grad_norm": 1.6555540989755444, "learning_rate": 9.621838490359112e-06, "loss": 0.583, "step": 1958 }, { "epoch": 0.6372804163955758, "grad_norm": 2.154466931495036, "learning_rate": 9.621116068438084e-06, "loss": 0.727, "step": 1959 }, { "epoch": 0.6376057254391672, "grad_norm": 1.8555830355110257, "learning_rate": 9.620392984308042e-06, "loss": 0.55, "step": 1960 }, { "epoch": 0.6379310344827587, "grad_norm": 1.835120098804383, "learning_rate": 9.619669238072607e-06, "loss": 0.6202, "step": 1961 }, { "epoch": 0.6382563435263501, "grad_norm": 1.8926841132714913, "learning_rate": 9.618944829835487e-06, "loss": 0.601, "step": 1962 }, { "epoch": 0.6385816525699415, "grad_norm": 1.8678269387265776, "learning_rate": 9.618219759700495e-06, "loss": 0.6004, "step": 1963 }, { "epoch": 0.6389069616135329, "grad_norm": 1.7724738628456367, "learning_rate": 9.617494027771531e-06, "loss": 0.5625, "step": 1964 }, { "epoch": 0.6392322706571243, "grad_norm": 2.0680125662786697, "learning_rate": 9.616767634152596e-06, "loss": 0.6474, "step": 1965 }, { "epoch": 0.6395575797007157, "grad_norm": 1.906701928004161, "learning_rate": 9.616040578947779e-06, "loss": 0.6528, "step": 1966 }, { "epoch": 0.6398828887443071, "grad_norm": 1.7125968719764193, "learning_rate": 9.61531286226127e-06, "loss": 0.5611, "step": 1967 }, { "epoch": 0.6402081977878985, "grad_norm": 1.9171624510695657, "learning_rate": 9.61458448419735e-06, "loss": 0.6199, "step": 1968 }, { "epoch": 0.6405335068314899, "grad_norm": 2.0266712618480422, "learning_rate": 9.6138554448604e-06, "loss": 0.665, "step": 1969 }, { "epoch": 0.6408588158750813, "grad_norm": 1.8575409333330224, "learning_rate": 9.613125744354884e-06, "loss": 0.5204, "step": 1970 }, { "epoch": 0.6411841249186727, "grad_norm": 2.38682806996742, "learning_rate": 9.612395382785377e-06, "loss": 0.6665, "step": 1971 }, { "epoch": 0.6415094339622641, "grad_norm": 1.917162562809391, "learning_rate": 9.611664360256534e-06, "loss": 0.6899, "step": 1972 }, { "epoch": 0.6418347430058555, "grad_norm": 2.00743138978012, "learning_rate": 9.610932676873114e-06, "loss": 0.6374, "step": 1973 }, { "epoch": 0.6421600520494469, "grad_norm": 2.0500967685263984, "learning_rate": 9.610200332739968e-06, "loss": 0.5649, "step": 1974 }, { "epoch": 0.6424853610930383, "grad_norm": 2.1168694151785736, "learning_rate": 9.609467327962041e-06, "loss": 0.645, "step": 1975 }, { "epoch": 0.6428106701366298, "grad_norm": 1.8808818298924574, "learning_rate": 9.608733662644371e-06, "loss": 0.5186, "step": 1976 }, { "epoch": 0.6431359791802213, "grad_norm": 1.9639156537898717, "learning_rate": 9.607999336892095e-06, "loss": 0.5898, "step": 1977 }, { "epoch": 0.6434612882238127, "grad_norm": 1.8394749631838974, "learning_rate": 9.607264350810442e-06, "loss": 0.7294, "step": 1978 }, { "epoch": 0.6437865972674041, "grad_norm": 2.006205926300794, "learning_rate": 9.606528704504734e-06, "loss": 0.5927, "step": 1979 }, { "epoch": 0.6441119063109955, "grad_norm": 1.880360130563734, "learning_rate": 9.605792398080393e-06, "loss": 0.6052, "step": 1980 }, { "epoch": 0.6444372153545869, "grad_norm": 1.784307158411635, "learning_rate": 9.60505543164293e-06, "loss": 0.5325, "step": 1981 }, { "epoch": 0.6447625243981783, "grad_norm": 1.9341361542864237, "learning_rate": 9.604317805297955e-06, "loss": 0.6968, "step": 1982 }, { "epoch": 0.6450878334417697, "grad_norm": 1.873346416105067, "learning_rate": 9.603579519151167e-06, "loss": 0.5434, "step": 1983 }, { "epoch": 0.6454131424853611, "grad_norm": 1.8756749539754514, "learning_rate": 9.602840573308365e-06, "loss": 0.6612, "step": 1984 }, { "epoch": 0.6457384515289525, "grad_norm": 1.5078911486092377, "learning_rate": 9.60210096787544e-06, "loss": 0.4233, "step": 1985 }, { "epoch": 0.6460637605725439, "grad_norm": 2.2741835751796606, "learning_rate": 9.601360702958377e-06, "loss": 0.7635, "step": 1986 }, { "epoch": 0.6463890696161353, "grad_norm": 1.8161343629824893, "learning_rate": 9.600619778663258e-06, "loss": 0.6089, "step": 1987 }, { "epoch": 0.6467143786597267, "grad_norm": 2.041677589921477, "learning_rate": 9.599878195096258e-06, "loss": 0.5319, "step": 1988 }, { "epoch": 0.6470396877033181, "grad_norm": 1.8608881724209454, "learning_rate": 9.599135952363646e-06, "loss": 0.6385, "step": 1989 }, { "epoch": 0.6473649967469096, "grad_norm": 1.8544384025616456, "learning_rate": 9.598393050571786e-06, "loss": 0.6024, "step": 1990 }, { "epoch": 0.647690305790501, "grad_norm": 1.9173943951988537, "learning_rate": 9.597649489827134e-06, "loss": 0.5967, "step": 1991 }, { "epoch": 0.6480156148340924, "grad_norm": 2.1344105797678528, "learning_rate": 9.596905270236245e-06, "loss": 0.6195, "step": 1992 }, { "epoch": 0.6483409238776838, "grad_norm": 1.9240769831822047, "learning_rate": 9.596160391905767e-06, "loss": 0.6886, "step": 1993 }, { "epoch": 0.6486662329212752, "grad_norm": 2.1567725782555853, "learning_rate": 9.59541485494244e-06, "loss": 0.6305, "step": 1994 }, { "epoch": 0.6489915419648666, "grad_norm": 2.0446826202832784, "learning_rate": 9.5946686594531e-06, "loss": 0.6811, "step": 1995 }, { "epoch": 0.649316851008458, "grad_norm": 1.927177037941682, "learning_rate": 9.59392180554468e-06, "loss": 0.6171, "step": 1996 }, { "epoch": 0.6496421600520494, "grad_norm": 1.9774258105951859, "learning_rate": 9.593174293324198e-06, "loss": 0.6745, "step": 1997 }, { "epoch": 0.6499674690956408, "grad_norm": 1.7984641176397451, "learning_rate": 9.592426122898781e-06, "loss": 0.5876, "step": 1998 }, { "epoch": 0.6502927781392323, "grad_norm": 1.7263913061508858, "learning_rate": 9.591677294375637e-06, "loss": 0.5673, "step": 1999 }, { "epoch": 0.6506180871828237, "grad_norm": 1.755515372755005, "learning_rate": 9.590927807862077e-06, "loss": 0.5799, "step": 2000 }, { "epoch": 0.6509433962264151, "grad_norm": 1.8169161061751724, "learning_rate": 9.590177663465499e-06, "loss": 0.5459, "step": 2001 }, { "epoch": 0.6512687052700065, "grad_norm": 1.6749311124803623, "learning_rate": 9.589426861293402e-06, "loss": 0.5357, "step": 2002 }, { "epoch": 0.6515940143135979, "grad_norm": 1.8300222350322015, "learning_rate": 9.588675401453376e-06, "loss": 0.614, "step": 2003 }, { "epoch": 0.6519193233571894, "grad_norm": 1.7066480028342201, "learning_rate": 9.587923284053107e-06, "loss": 0.616, "step": 2004 }, { "epoch": 0.6522446324007808, "grad_norm": 1.9635226624416016, "learning_rate": 9.587170509200372e-06, "loss": 0.6639, "step": 2005 }, { "epoch": 0.6525699414443722, "grad_norm": 1.7521438749389229, "learning_rate": 9.586417077003044e-06, "loss": 0.6402, "step": 2006 }, { "epoch": 0.6528952504879636, "grad_norm": 1.800874565934784, "learning_rate": 9.58566298756909e-06, "loss": 0.5265, "step": 2007 }, { "epoch": 0.653220559531555, "grad_norm": 1.7278848641777707, "learning_rate": 9.584908241006575e-06, "loss": 0.5506, "step": 2008 }, { "epoch": 0.6535458685751464, "grad_norm": 1.7675189465921068, "learning_rate": 9.584152837423649e-06, "loss": 0.5733, "step": 2009 }, { "epoch": 0.6538711776187378, "grad_norm": 1.892638779023361, "learning_rate": 9.583396776928566e-06, "loss": 0.5463, "step": 2010 }, { "epoch": 0.6541964866623292, "grad_norm": 1.9369915775632294, "learning_rate": 9.58264005962967e-06, "loss": 0.6555, "step": 2011 }, { "epoch": 0.6545217957059206, "grad_norm": 1.8599543925609519, "learning_rate": 9.581882685635398e-06, "loss": 0.6252, "step": 2012 }, { "epoch": 0.654847104749512, "grad_norm": 1.8936616989749948, "learning_rate": 9.581124655054281e-06, "loss": 0.4967, "step": 2013 }, { "epoch": 0.6551724137931034, "grad_norm": 1.9420497406410104, "learning_rate": 9.580365967994949e-06, "loss": 0.6387, "step": 2014 }, { "epoch": 0.6554977228366948, "grad_norm": 1.6265989714728974, "learning_rate": 9.579606624566117e-06, "loss": 0.4503, "step": 2015 }, { "epoch": 0.6558230318802862, "grad_norm": 1.898973132046907, "learning_rate": 9.578846624876605e-06, "loss": 0.5224, "step": 2016 }, { "epoch": 0.6561483409238776, "grad_norm": 1.7857690569689935, "learning_rate": 9.578085969035317e-06, "loss": 0.5827, "step": 2017 }, { "epoch": 0.656473649967469, "grad_norm": 1.9565933385354122, "learning_rate": 9.577324657151259e-06, "loss": 0.7712, "step": 2018 }, { "epoch": 0.6567989590110606, "grad_norm": 1.8468704202097526, "learning_rate": 9.576562689333524e-06, "loss": 0.7159, "step": 2019 }, { "epoch": 0.657124268054652, "grad_norm": 1.889456635072558, "learning_rate": 9.575800065691306e-06, "loss": 0.5793, "step": 2020 }, { "epoch": 0.6574495770982434, "grad_norm": 1.9492293089050223, "learning_rate": 9.575036786333886e-06, "loss": 0.6624, "step": 2021 }, { "epoch": 0.6577748861418348, "grad_norm": 1.8499708031652282, "learning_rate": 9.574272851370645e-06, "loss": 0.5692, "step": 2022 }, { "epoch": 0.6581001951854262, "grad_norm": 1.7863743331272044, "learning_rate": 9.573508260911053e-06, "loss": 0.5927, "step": 2023 }, { "epoch": 0.6584255042290176, "grad_norm": 1.6104961821916681, "learning_rate": 9.57274301506468e-06, "loss": 0.5784, "step": 2024 }, { "epoch": 0.658750813272609, "grad_norm": 1.7103648868926804, "learning_rate": 9.571977113941184e-06, "loss": 0.4772, "step": 2025 }, { "epoch": 0.6590761223162004, "grad_norm": 1.81935205998446, "learning_rate": 9.571210557650319e-06, "loss": 0.7315, "step": 2026 }, { "epoch": 0.6594014313597918, "grad_norm": 1.8925218327890014, "learning_rate": 9.570443346301933e-06, "loss": 0.4118, "step": 2027 }, { "epoch": 0.6597267404033832, "grad_norm": 2.2070486930767403, "learning_rate": 9.569675480005969e-06, "loss": 0.6385, "step": 2028 }, { "epoch": 0.6600520494469746, "grad_norm": 1.970156157614698, "learning_rate": 9.568906958872461e-06, "loss": 0.7273, "step": 2029 }, { "epoch": 0.660377358490566, "grad_norm": 1.8927267553665519, "learning_rate": 9.56813778301154e-06, "loss": 0.6126, "step": 2030 }, { "epoch": 0.6607026675341574, "grad_norm": 2.0430292643661048, "learning_rate": 9.567367952533429e-06, "loss": 0.6202, "step": 2031 }, { "epoch": 0.6610279765777488, "grad_norm": 1.8271279534396037, "learning_rate": 9.566597467548444e-06, "loss": 0.6634, "step": 2032 }, { "epoch": 0.6613532856213403, "grad_norm": 1.7735036512985467, "learning_rate": 9.565826328167e-06, "loss": 0.6529, "step": 2033 }, { "epoch": 0.6616785946649317, "grad_norm": 1.6993207986985344, "learning_rate": 9.565054534499596e-06, "loss": 0.4546, "step": 2034 }, { "epoch": 0.6620039037085231, "grad_norm": 1.8545976978913812, "learning_rate": 9.564282086656836e-06, "loss": 0.7039, "step": 2035 }, { "epoch": 0.6623292127521145, "grad_norm": 1.800517563869697, "learning_rate": 9.563508984749408e-06, "loss": 0.5514, "step": 2036 }, { "epoch": 0.6626545217957059, "grad_norm": 1.9101675248663532, "learning_rate": 9.5627352288881e-06, "loss": 0.6779, "step": 2037 }, { "epoch": 0.6629798308392973, "grad_norm": 1.8899166120338406, "learning_rate": 9.561960819183793e-06, "loss": 0.6456, "step": 2038 }, { "epoch": 0.6633051398828888, "grad_norm": 1.973282053999799, "learning_rate": 9.56118575574746e-06, "loss": 0.706, "step": 2039 }, { "epoch": 0.6636304489264802, "grad_norm": 1.7338515961850078, "learning_rate": 9.560410038690164e-06, "loss": 0.5622, "step": 2040 }, { "epoch": 0.6639557579700716, "grad_norm": 2.114642684032658, "learning_rate": 9.55963366812307e-06, "loss": 0.6754, "step": 2041 }, { "epoch": 0.664281067013663, "grad_norm": 1.7964406061314555, "learning_rate": 9.558856644157432e-06, "loss": 0.61, "step": 2042 }, { "epoch": 0.6646063760572544, "grad_norm": 2.027657035769995, "learning_rate": 9.558078966904597e-06, "loss": 0.7399, "step": 2043 }, { "epoch": 0.6649316851008458, "grad_norm": 2.023515032806664, "learning_rate": 9.557300636476007e-06, "loss": 0.6373, "step": 2044 }, { "epoch": 0.6652569941444372, "grad_norm": 1.9617414104585724, "learning_rate": 9.556521652983198e-06, "loss": 0.6129, "step": 2045 }, { "epoch": 0.6655823031880286, "grad_norm": 1.8089892210688137, "learning_rate": 9.555742016537797e-06, "loss": 0.6546, "step": 2046 }, { "epoch": 0.6659076122316201, "grad_norm": 1.656048703111828, "learning_rate": 9.554961727251529e-06, "loss": 0.5154, "step": 2047 }, { "epoch": 0.6662329212752115, "grad_norm": 1.7267542699905778, "learning_rate": 9.554180785236206e-06, "loss": 0.4825, "step": 2048 }, { "epoch": 0.6665582303188029, "grad_norm": 1.8136537041263523, "learning_rate": 9.553399190603743e-06, "loss": 0.4912, "step": 2049 }, { "epoch": 0.6668835393623943, "grad_norm": 1.9609453500759817, "learning_rate": 9.552616943466138e-06, "loss": 0.6405, "step": 2050 }, { "epoch": 0.6672088484059857, "grad_norm": 2.0120821067970627, "learning_rate": 9.55183404393549e-06, "loss": 0.6698, "step": 2051 }, { "epoch": 0.6675341574495771, "grad_norm": 1.7900674205720835, "learning_rate": 9.55105049212399e-06, "loss": 0.5639, "step": 2052 }, { "epoch": 0.6678594664931685, "grad_norm": 1.998987991383015, "learning_rate": 9.55026628814392e-06, "loss": 0.6771, "step": 2053 }, { "epoch": 0.6681847755367599, "grad_norm": 1.7903691502093122, "learning_rate": 9.549481432107655e-06, "loss": 0.6454, "step": 2054 }, { "epoch": 0.6685100845803513, "grad_norm": 1.9216518487853724, "learning_rate": 9.548695924127669e-06, "loss": 0.5888, "step": 2055 }, { "epoch": 0.6688353936239427, "grad_norm": 1.657110623849892, "learning_rate": 9.547909764316523e-06, "loss": 0.5253, "step": 2056 }, { "epoch": 0.6691607026675341, "grad_norm": 1.8844705785122102, "learning_rate": 9.547122952786878e-06, "loss": 0.5867, "step": 2057 }, { "epoch": 0.6694860117111255, "grad_norm": 1.9310323341677513, "learning_rate": 9.546335489651481e-06, "loss": 0.6621, "step": 2058 }, { "epoch": 0.6698113207547169, "grad_norm": 1.8068811731699144, "learning_rate": 9.545547375023176e-06, "loss": 0.623, "step": 2059 }, { "epoch": 0.6701366297983083, "grad_norm": 1.7684953544090978, "learning_rate": 9.544758609014904e-06, "loss": 0.588, "step": 2060 }, { "epoch": 0.6704619388418998, "grad_norm": 2.1531500846866125, "learning_rate": 9.543969191739691e-06, "loss": 0.6932, "step": 2061 }, { "epoch": 0.6707872478854913, "grad_norm": 1.916112426322765, "learning_rate": 9.543179123310665e-06, "loss": 0.5884, "step": 2062 }, { "epoch": 0.6711125569290827, "grad_norm": 1.928301784098597, "learning_rate": 9.54238840384104e-06, "loss": 0.6938, "step": 2063 }, { "epoch": 0.6714378659726741, "grad_norm": 1.8035609520806135, "learning_rate": 9.541597033444127e-06, "loss": 0.499, "step": 2064 }, { "epoch": 0.6717631750162655, "grad_norm": 2.24049779731818, "learning_rate": 9.540805012233333e-06, "loss": 0.7174, "step": 2065 }, { "epoch": 0.6720884840598569, "grad_norm": 1.5839057747726635, "learning_rate": 9.540012340322154e-06, "loss": 0.5744, "step": 2066 }, { "epoch": 0.6724137931034483, "grad_norm": 1.8828921422869642, "learning_rate": 9.539219017824179e-06, "loss": 0.654, "step": 2067 }, { "epoch": 0.6727391021470397, "grad_norm": 1.908383130367219, "learning_rate": 9.538425044853092e-06, "loss": 0.6981, "step": 2068 }, { "epoch": 0.6730644111906311, "grad_norm": 1.9437230118011375, "learning_rate": 9.53763042152267e-06, "loss": 0.6553, "step": 2069 }, { "epoch": 0.6733897202342225, "grad_norm": 1.7700888727894415, "learning_rate": 9.536835147946783e-06, "loss": 0.6259, "step": 2070 }, { "epoch": 0.6737150292778139, "grad_norm": 1.9828049898385252, "learning_rate": 9.536039224239396e-06, "loss": 0.6123, "step": 2071 }, { "epoch": 0.6740403383214053, "grad_norm": 2.0565710795973695, "learning_rate": 9.535242650514562e-06, "loss": 0.6108, "step": 2072 }, { "epoch": 0.6743656473649967, "grad_norm": 2.114357342149304, "learning_rate": 9.534445426886434e-06, "loss": 0.7531, "step": 2073 }, { "epoch": 0.6746909564085881, "grad_norm": 1.8249810449303616, "learning_rate": 9.533647553469253e-06, "loss": 0.6243, "step": 2074 }, { "epoch": 0.6750162654521795, "grad_norm": 1.9805637787223138, "learning_rate": 9.532849030377353e-06, "loss": 0.7098, "step": 2075 }, { "epoch": 0.675341574495771, "grad_norm": 1.6568321469074678, "learning_rate": 9.532049857725165e-06, "loss": 0.5511, "step": 2076 }, { "epoch": 0.6756668835393624, "grad_norm": 2.00258412466319, "learning_rate": 9.531250035627212e-06, "loss": 0.6333, "step": 2077 }, { "epoch": 0.6759921925829538, "grad_norm": 1.600431169628804, "learning_rate": 9.530449564198108e-06, "loss": 0.6272, "step": 2078 }, { "epoch": 0.6763175016265452, "grad_norm": 1.6319871864072741, "learning_rate": 9.52964844355256e-06, "loss": 0.5908, "step": 2079 }, { "epoch": 0.6766428106701367, "grad_norm": 1.7804254508152517, "learning_rate": 9.52884667380537e-06, "loss": 0.5649, "step": 2080 }, { "epoch": 0.6769681197137281, "grad_norm": 1.979453585189807, "learning_rate": 9.528044255071432e-06, "loss": 0.6047, "step": 2081 }, { "epoch": 0.6772934287573195, "grad_norm": 1.9549388802775347, "learning_rate": 9.527241187465735e-06, "loss": 0.5737, "step": 2082 }, { "epoch": 0.6776187378009109, "grad_norm": 2.0874589311494054, "learning_rate": 9.526437471103354e-06, "loss": 0.7053, "step": 2083 }, { "epoch": 0.6779440468445023, "grad_norm": 1.7693889389993076, "learning_rate": 9.525633106099468e-06, "loss": 0.5559, "step": 2084 }, { "epoch": 0.6782693558880937, "grad_norm": 1.9107043942178992, "learning_rate": 9.52482809256934e-06, "loss": 0.6331, "step": 2085 }, { "epoch": 0.6785946649316851, "grad_norm": 2.077155248290129, "learning_rate": 9.52402243062833e-06, "loss": 0.6466, "step": 2086 }, { "epoch": 0.6789199739752765, "grad_norm": 1.684147913582535, "learning_rate": 9.523216120391889e-06, "loss": 0.513, "step": 2087 }, { "epoch": 0.6792452830188679, "grad_norm": 1.6875648224304776, "learning_rate": 9.522409161975563e-06, "loss": 0.4717, "step": 2088 }, { "epoch": 0.6795705920624593, "grad_norm": 1.8771259945259287, "learning_rate": 9.521601555494988e-06, "loss": 0.639, "step": 2089 }, { "epoch": 0.6798959011060507, "grad_norm": 1.860872857842261, "learning_rate": 9.520793301065894e-06, "loss": 0.6701, "step": 2090 }, { "epoch": 0.6802212101496422, "grad_norm": 1.8271663275943182, "learning_rate": 9.519984398804108e-06, "loss": 0.6053, "step": 2091 }, { "epoch": 0.6805465191932336, "grad_norm": 1.7731635640894883, "learning_rate": 9.519174848825543e-06, "loss": 0.5834, "step": 2092 }, { "epoch": 0.680871828236825, "grad_norm": 2.1869027599021287, "learning_rate": 9.51836465124621e-06, "loss": 0.7158, "step": 2093 }, { "epoch": 0.6811971372804164, "grad_norm": 1.864333381786004, "learning_rate": 9.51755380618221e-06, "loss": 0.6153, "step": 2094 }, { "epoch": 0.6815224463240078, "grad_norm": 1.731605861544367, "learning_rate": 9.516742313749736e-06, "loss": 0.5956, "step": 2095 }, { "epoch": 0.6818477553675992, "grad_norm": 2.259333961686709, "learning_rate": 9.515930174065078e-06, "loss": 0.7413, "step": 2096 }, { "epoch": 0.6821730644111906, "grad_norm": 2.0149097344368965, "learning_rate": 9.515117387244615e-06, "loss": 0.6374, "step": 2097 }, { "epoch": 0.682498373454782, "grad_norm": 1.4807643384907072, "learning_rate": 9.51430395340482e-06, "loss": 0.4461, "step": 2098 }, { "epoch": 0.6828236824983734, "grad_norm": 1.914921364268363, "learning_rate": 9.513489872662258e-06, "loss": 0.549, "step": 2099 }, { "epoch": 0.6831489915419648, "grad_norm": 1.667504483922668, "learning_rate": 9.51267514513359e-06, "loss": 0.564, "step": 2100 }, { "epoch": 0.6834743005855562, "grad_norm": 1.847416698209194, "learning_rate": 9.511859770935564e-06, "loss": 0.5502, "step": 2101 }, { "epoch": 0.6837996096291477, "grad_norm": 1.7192323776653613, "learning_rate": 9.511043750185023e-06, "loss": 0.5835, "step": 2102 }, { "epoch": 0.6841249186727391, "grad_norm": 2.1738865211722307, "learning_rate": 9.510227082998907e-06, "loss": 0.6223, "step": 2103 }, { "epoch": 0.6844502277163305, "grad_norm": 1.8931480047452534, "learning_rate": 9.50940976949424e-06, "loss": 0.5926, "step": 2104 }, { "epoch": 0.684775536759922, "grad_norm": 1.7947502758391163, "learning_rate": 9.50859180978815e-06, "loss": 0.6038, "step": 2105 }, { "epoch": 0.6851008458035134, "grad_norm": 1.739784019812482, "learning_rate": 9.507773203997845e-06, "loss": 0.6355, "step": 2106 }, { "epoch": 0.6854261548471048, "grad_norm": 1.7122655300730856, "learning_rate": 9.506953952240637e-06, "loss": 0.6218, "step": 2107 }, { "epoch": 0.6857514638906962, "grad_norm": 1.6116285768641825, "learning_rate": 9.506134054633922e-06, "loss": 0.4648, "step": 2108 }, { "epoch": 0.6860767729342876, "grad_norm": 1.8172322991701095, "learning_rate": 9.505313511295192e-06, "loss": 0.6472, "step": 2109 }, { "epoch": 0.686402081977879, "grad_norm": 1.9389439375450557, "learning_rate": 9.504492322342032e-06, "loss": 0.6211, "step": 2110 }, { "epoch": 0.6867273910214704, "grad_norm": 1.773483596075516, "learning_rate": 9.503670487892117e-06, "loss": 0.523, "step": 2111 }, { "epoch": 0.6870527000650618, "grad_norm": 1.941006466107922, "learning_rate": 9.502848008063221e-06, "loss": 0.5716, "step": 2112 }, { "epoch": 0.6873780091086532, "grad_norm": 2.284468191042164, "learning_rate": 9.502024882973203e-06, "loss": 0.6764, "step": 2113 }, { "epoch": 0.6877033181522446, "grad_norm": 1.8825237053596995, "learning_rate": 9.501201112740017e-06, "loss": 0.5594, "step": 2114 }, { "epoch": 0.688028627195836, "grad_norm": 1.814501997887259, "learning_rate": 9.50037669748171e-06, "loss": 0.6237, "step": 2115 }, { "epoch": 0.6883539362394274, "grad_norm": 1.812631924985818, "learning_rate": 9.499551637316423e-06, "loss": 0.6066, "step": 2116 }, { "epoch": 0.6886792452830188, "grad_norm": 1.6992447513519966, "learning_rate": 9.498725932362385e-06, "loss": 0.6113, "step": 2117 }, { "epoch": 0.6890045543266102, "grad_norm": 1.7273729003682812, "learning_rate": 9.497899582737924e-06, "loss": 0.5668, "step": 2118 }, { "epoch": 0.6893298633702017, "grad_norm": 2.0242965484916793, "learning_rate": 9.497072588561451e-06, "loss": 0.632, "step": 2119 }, { "epoch": 0.6896551724137931, "grad_norm": 1.9061152556151295, "learning_rate": 9.496244949951478e-06, "loss": 0.6912, "step": 2120 }, { "epoch": 0.6899804814573846, "grad_norm": 1.706334160732356, "learning_rate": 9.495416667026607e-06, "loss": 0.5788, "step": 2121 }, { "epoch": 0.690305790500976, "grad_norm": 2.025692628447176, "learning_rate": 9.494587739905531e-06, "loss": 0.6477, "step": 2122 }, { "epoch": 0.6906310995445674, "grad_norm": 2.362308875304334, "learning_rate": 9.493758168707034e-06, "loss": 0.7458, "step": 2123 }, { "epoch": 0.6909564085881588, "grad_norm": 1.9704995018998475, "learning_rate": 9.492927953549996e-06, "loss": 0.6975, "step": 2124 }, { "epoch": 0.6912817176317502, "grad_norm": 2.074300971253896, "learning_rate": 9.492097094553387e-06, "loss": 0.6975, "step": 2125 }, { "epoch": 0.6916070266753416, "grad_norm": 1.741407900420256, "learning_rate": 9.491265591836267e-06, "loss": 0.537, "step": 2126 }, { "epoch": 0.691932335718933, "grad_norm": 1.8382588321878233, "learning_rate": 9.490433445517795e-06, "loss": 0.6737, "step": 2127 }, { "epoch": 0.6922576447625244, "grad_norm": 1.754301149506055, "learning_rate": 9.489600655717217e-06, "loss": 0.5862, "step": 2128 }, { "epoch": 0.6925829538061158, "grad_norm": 1.7142827713018294, "learning_rate": 9.488767222553871e-06, "loss": 0.5306, "step": 2129 }, { "epoch": 0.6929082628497072, "grad_norm": 2.0842943922982617, "learning_rate": 9.48793314614719e-06, "loss": 0.7079, "step": 2130 }, { "epoch": 0.6932335718932986, "grad_norm": 1.8768718144351988, "learning_rate": 9.487098426616696e-06, "loss": 0.6405, "step": 2131 }, { "epoch": 0.69355888093689, "grad_norm": 1.984821035688983, "learning_rate": 9.486263064082005e-06, "loss": 0.5798, "step": 2132 }, { "epoch": 0.6938841899804814, "grad_norm": 1.900803139368456, "learning_rate": 9.485427058662827e-06, "loss": 0.6254, "step": 2133 }, { "epoch": 0.6942094990240729, "grad_norm": 1.9402330534300556, "learning_rate": 9.48459041047896e-06, "loss": 0.5501, "step": 2134 }, { "epoch": 0.6945348080676643, "grad_norm": 1.8519675592371643, "learning_rate": 9.483753119650299e-06, "loss": 0.5827, "step": 2135 }, { "epoch": 0.6948601171112557, "grad_norm": 1.8808967257588836, "learning_rate": 9.482915186296824e-06, "loss": 0.5775, "step": 2136 }, { "epoch": 0.6951854261548471, "grad_norm": 2.064449235002876, "learning_rate": 9.482076610538615e-06, "loss": 0.7277, "step": 2137 }, { "epoch": 0.6955107351984385, "grad_norm": 2.12796038858214, "learning_rate": 9.48123739249584e-06, "loss": 0.7036, "step": 2138 }, { "epoch": 0.6958360442420299, "grad_norm": 1.6834477214151546, "learning_rate": 9.480397532288758e-06, "loss": 0.5407, "step": 2139 }, { "epoch": 0.6961613532856213, "grad_norm": 1.765029143319988, "learning_rate": 9.479557030037723e-06, "loss": 0.5327, "step": 2140 }, { "epoch": 0.6964866623292127, "grad_norm": 1.8528387041819252, "learning_rate": 9.47871588586318e-06, "loss": 0.6288, "step": 2141 }, { "epoch": 0.6968119713728042, "grad_norm": 1.8619355295768407, "learning_rate": 9.477874099885662e-06, "loss": 0.542, "step": 2142 }, { "epoch": 0.6971372804163956, "grad_norm": 1.9443736579894113, "learning_rate": 9.477031672225802e-06, "loss": 0.7065, "step": 2143 }, { "epoch": 0.697462589459987, "grad_norm": 1.7072368312669923, "learning_rate": 9.476188603004317e-06, "loss": 0.5612, "step": 2144 }, { "epoch": 0.6977878985035784, "grad_norm": 1.9672147673863796, "learning_rate": 9.475344892342021e-06, "loss": 0.7571, "step": 2145 }, { "epoch": 0.6981132075471698, "grad_norm": 2.133272161625641, "learning_rate": 9.474500540359819e-06, "loss": 0.5709, "step": 2146 }, { "epoch": 0.6984385165907612, "grad_norm": 1.9899137306037555, "learning_rate": 9.473655547178706e-06, "loss": 0.5651, "step": 2147 }, { "epoch": 0.6987638256343527, "grad_norm": 2.0482362084466588, "learning_rate": 9.472809912919769e-06, "loss": 0.766, "step": 2148 }, { "epoch": 0.6990891346779441, "grad_norm": 1.9596207009797073, "learning_rate": 9.471963637704191e-06, "loss": 0.6759, "step": 2149 }, { "epoch": 0.6994144437215355, "grad_norm": 1.8686428791877114, "learning_rate": 9.471116721653242e-06, "loss": 0.668, "step": 2150 }, { "epoch": 0.6997397527651269, "grad_norm": 2.004817672491072, "learning_rate": 9.470269164888285e-06, "loss": 0.6394, "step": 2151 }, { "epoch": 0.7000650618087183, "grad_norm": 1.8458583718526644, "learning_rate": 9.469420967530776e-06, "loss": 0.6831, "step": 2152 }, { "epoch": 0.7003903708523097, "grad_norm": 1.7298488044295117, "learning_rate": 9.468572129702263e-06, "loss": 0.6268, "step": 2153 }, { "epoch": 0.7007156798959011, "grad_norm": 2.065809434589859, "learning_rate": 9.467722651524385e-06, "loss": 0.6041, "step": 2154 }, { "epoch": 0.7010409889394925, "grad_norm": 1.800365933371558, "learning_rate": 9.466872533118872e-06, "loss": 0.6006, "step": 2155 }, { "epoch": 0.7013662979830839, "grad_norm": 1.9739672853249461, "learning_rate": 9.466021774607545e-06, "loss": 0.6046, "step": 2156 }, { "epoch": 0.7016916070266753, "grad_norm": 1.646854558991183, "learning_rate": 9.465170376112322e-06, "loss": 0.5761, "step": 2157 }, { "epoch": 0.7020169160702667, "grad_norm": 1.7179502762735777, "learning_rate": 9.464318337755207e-06, "loss": 0.5739, "step": 2158 }, { "epoch": 0.7023422251138581, "grad_norm": 2.0121098788914806, "learning_rate": 9.463465659658298e-06, "loss": 0.7272, "step": 2159 }, { "epoch": 0.7026675341574495, "grad_norm": 1.738595307337853, "learning_rate": 9.462612341943783e-06, "loss": 0.614, "step": 2160 }, { "epoch": 0.7029928432010409, "grad_norm": 1.9416149984055526, "learning_rate": 9.461758384733945e-06, "loss": 0.5794, "step": 2161 }, { "epoch": 0.7033181522446325, "grad_norm": 2.0132046319998818, "learning_rate": 9.460903788151155e-06, "loss": 0.7291, "step": 2162 }, { "epoch": 0.7036434612882239, "grad_norm": 1.8969185133684634, "learning_rate": 9.460048552317878e-06, "loss": 0.507, "step": 2163 }, { "epoch": 0.7039687703318153, "grad_norm": 2.04088298145175, "learning_rate": 9.45919267735667e-06, "loss": 0.6758, "step": 2164 }, { "epoch": 0.7042940793754067, "grad_norm": 1.7153694222784976, "learning_rate": 9.458336163390178e-06, "loss": 0.5134, "step": 2165 }, { "epoch": 0.7046193884189981, "grad_norm": 1.7715480969799362, "learning_rate": 9.457479010541142e-06, "loss": 0.6216, "step": 2166 }, { "epoch": 0.7049446974625895, "grad_norm": 1.780029714408088, "learning_rate": 9.456621218932392e-06, "loss": 0.4923, "step": 2167 }, { "epoch": 0.7052700065061809, "grad_norm": 2.121645991283781, "learning_rate": 9.455762788686849e-06, "loss": 0.6766, "step": 2168 }, { "epoch": 0.7055953155497723, "grad_norm": 1.827650258233489, "learning_rate": 9.454903719927527e-06, "loss": 0.5622, "step": 2169 }, { "epoch": 0.7059206245933637, "grad_norm": 2.0841706822027657, "learning_rate": 9.454044012777533e-06, "loss": 0.662, "step": 2170 }, { "epoch": 0.7062459336369551, "grad_norm": 1.9313606479370575, "learning_rate": 9.453183667360062e-06, "loss": 0.558, "step": 2171 }, { "epoch": 0.7065712426805465, "grad_norm": 1.9011803223839654, "learning_rate": 9.452322683798402e-06, "loss": 0.593, "step": 2172 }, { "epoch": 0.7068965517241379, "grad_norm": 1.7278338485164921, "learning_rate": 9.451461062215933e-06, "loss": 0.5622, "step": 2173 }, { "epoch": 0.7072218607677293, "grad_norm": 1.764947779580279, "learning_rate": 9.450598802736126e-06, "loss": 0.6294, "step": 2174 }, { "epoch": 0.7075471698113207, "grad_norm": 1.861359316903831, "learning_rate": 9.449735905482543e-06, "loss": 0.6652, "step": 2175 }, { "epoch": 0.7078724788549121, "grad_norm": 1.6163202537099435, "learning_rate": 9.44887237057884e-06, "loss": 0.5376, "step": 2176 }, { "epoch": 0.7081977878985036, "grad_norm": 1.5961066228148857, "learning_rate": 9.448008198148757e-06, "loss": 0.5464, "step": 2177 }, { "epoch": 0.708523096942095, "grad_norm": 1.755878485217664, "learning_rate": 9.447143388316136e-06, "loss": 0.6478, "step": 2178 }, { "epoch": 0.7088484059856864, "grad_norm": 1.7352799873519738, "learning_rate": 9.4462779412049e-06, "loss": 0.5665, "step": 2179 }, { "epoch": 0.7091737150292778, "grad_norm": 1.8169656725947445, "learning_rate": 9.445411856939073e-06, "loss": 0.5734, "step": 2180 }, { "epoch": 0.7094990240728692, "grad_norm": 1.6332482503092838, "learning_rate": 9.444545135642762e-06, "loss": 0.5057, "step": 2181 }, { "epoch": 0.7098243331164606, "grad_norm": 1.6699963957043096, "learning_rate": 9.44367777744017e-06, "loss": 0.5578, "step": 2182 }, { "epoch": 0.710149642160052, "grad_norm": 2.1201574571560124, "learning_rate": 9.442809782455589e-06, "loss": 0.745, "step": 2183 }, { "epoch": 0.7104749512036435, "grad_norm": 1.921446470891562, "learning_rate": 9.441941150813404e-06, "loss": 0.6833, "step": 2184 }, { "epoch": 0.7108002602472349, "grad_norm": 1.655847401247011, "learning_rate": 9.441071882638091e-06, "loss": 0.5633, "step": 2185 }, { "epoch": 0.7111255692908263, "grad_norm": 2.039399213017121, "learning_rate": 9.440201978054217e-06, "loss": 0.71, "step": 2186 }, { "epoch": 0.7114508783344177, "grad_norm": 1.7132959318780443, "learning_rate": 9.439331437186438e-06, "loss": 0.5668, "step": 2187 }, { "epoch": 0.7117761873780091, "grad_norm": 1.735292428504644, "learning_rate": 9.438460260159504e-06, "loss": 0.6179, "step": 2188 }, { "epoch": 0.7121014964216005, "grad_norm": 1.9305592283707484, "learning_rate": 9.437588447098255e-06, "loss": 0.5902, "step": 2189 }, { "epoch": 0.7124268054651919, "grad_norm": 1.700675325442621, "learning_rate": 9.436715998127626e-06, "loss": 0.6129, "step": 2190 }, { "epoch": 0.7127521145087834, "grad_norm": 1.8786592638776318, "learning_rate": 9.435842913372632e-06, "loss": 0.563, "step": 2191 }, { "epoch": 0.7130774235523748, "grad_norm": 1.9469856213004606, "learning_rate": 9.434969192958394e-06, "loss": 0.607, "step": 2192 }, { "epoch": 0.7134027325959662, "grad_norm": 1.8063362647497143, "learning_rate": 9.434094837010113e-06, "loss": 0.5649, "step": 2193 }, { "epoch": 0.7137280416395576, "grad_norm": 1.7310902004006858, "learning_rate": 9.433219845653085e-06, "loss": 0.5089, "step": 2194 }, { "epoch": 0.714053350683149, "grad_norm": 2.0468100590592377, "learning_rate": 9.4323442190127e-06, "loss": 0.7449, "step": 2195 }, { "epoch": 0.7143786597267404, "grad_norm": 1.7747657400342074, "learning_rate": 9.431467957214431e-06, "loss": 0.5544, "step": 2196 }, { "epoch": 0.7147039687703318, "grad_norm": 1.820828527905242, "learning_rate": 9.430591060383848e-06, "loss": 0.6007, "step": 2197 }, { "epoch": 0.7150292778139232, "grad_norm": 1.934098773394653, "learning_rate": 9.429713528646613e-06, "loss": 0.7091, "step": 2198 }, { "epoch": 0.7153545868575146, "grad_norm": 1.8308882324288196, "learning_rate": 9.428835362128477e-06, "loss": 0.5209, "step": 2199 }, { "epoch": 0.715679895901106, "grad_norm": 1.961549209618246, "learning_rate": 9.42795656095528e-06, "loss": 0.7147, "step": 2200 }, { "epoch": 0.7160052049446974, "grad_norm": 1.7061360375370815, "learning_rate": 9.427077125252956e-06, "loss": 0.5092, "step": 2201 }, { "epoch": 0.7163305139882888, "grad_norm": 2.066605967928497, "learning_rate": 9.426197055147528e-06, "loss": 0.6973, "step": 2202 }, { "epoch": 0.7166558230318802, "grad_norm": 2.0798710769747775, "learning_rate": 9.42531635076511e-06, "loss": 0.636, "step": 2203 }, { "epoch": 0.7169811320754716, "grad_norm": 1.9120523205766253, "learning_rate": 9.42443501223191e-06, "loss": 0.6658, "step": 2204 }, { "epoch": 0.7173064411190632, "grad_norm": 1.9786234197086863, "learning_rate": 9.423553039674222e-06, "loss": 0.5767, "step": 2205 }, { "epoch": 0.7176317501626546, "grad_norm": 1.9714080547231265, "learning_rate": 9.422670433218433e-06, "loss": 0.5963, "step": 2206 }, { "epoch": 0.717957059206246, "grad_norm": 1.7393932287730063, "learning_rate": 9.421787192991024e-06, "loss": 0.5312, "step": 2207 }, { "epoch": 0.7182823682498374, "grad_norm": 1.8410112346955272, "learning_rate": 9.420903319118561e-06, "loss": 0.5986, "step": 2208 }, { "epoch": 0.7186076772934288, "grad_norm": 1.7305547591974138, "learning_rate": 9.420018811727707e-06, "loss": 0.5188, "step": 2209 }, { "epoch": 0.7189329863370202, "grad_norm": 1.865109893189064, "learning_rate": 9.419133670945207e-06, "loss": 0.5604, "step": 2210 }, { "epoch": 0.7192582953806116, "grad_norm": 1.7912697702197238, "learning_rate": 9.418247896897907e-06, "loss": 0.5715, "step": 2211 }, { "epoch": 0.719583604424203, "grad_norm": 17.516484490190553, "learning_rate": 9.417361489712738e-06, "loss": 0.6442, "step": 2212 }, { "epoch": 0.7199089134677944, "grad_norm": 2.249524550666106, "learning_rate": 9.416474449516722e-06, "loss": 0.7345, "step": 2213 }, { "epoch": 0.7202342225113858, "grad_norm": 1.8496760209128766, "learning_rate": 9.415586776436973e-06, "loss": 0.6918, "step": 2214 }, { "epoch": 0.7205595315549772, "grad_norm": 1.8250675013055704, "learning_rate": 9.414698470600697e-06, "loss": 0.6103, "step": 2215 }, { "epoch": 0.7208848405985686, "grad_norm": 2.164084393600635, "learning_rate": 9.413809532135184e-06, "loss": 0.7051, "step": 2216 }, { "epoch": 0.72121014964216, "grad_norm": 1.7229417771895486, "learning_rate": 9.412919961167821e-06, "loss": 0.545, "step": 2217 }, { "epoch": 0.7215354586857514, "grad_norm": 1.9192447447146948, "learning_rate": 9.412029757826088e-06, "loss": 0.5944, "step": 2218 }, { "epoch": 0.7218607677293428, "grad_norm": 1.9672139199868364, "learning_rate": 9.411138922237548e-06, "loss": 0.6461, "step": 2219 }, { "epoch": 0.7221860767729343, "grad_norm": 1.7647778141186958, "learning_rate": 9.410247454529859e-06, "loss": 0.5521, "step": 2220 }, { "epoch": 0.7225113858165257, "grad_norm": 1.860748323479554, "learning_rate": 9.409355354830772e-06, "loss": 0.4967, "step": 2221 }, { "epoch": 0.7228366948601171, "grad_norm": 1.8840022724973295, "learning_rate": 9.40846262326812e-06, "loss": 0.5412, "step": 2222 }, { "epoch": 0.7231620039037086, "grad_norm": 1.9400114885160982, "learning_rate": 9.407569259969836e-06, "loss": 0.7324, "step": 2223 }, { "epoch": 0.7234873129473, "grad_norm": 1.886861309525122, "learning_rate": 9.406675265063939e-06, "loss": 0.5271, "step": 2224 }, { "epoch": 0.7238126219908914, "grad_norm": 1.9918909762486525, "learning_rate": 9.405780638678537e-06, "loss": 0.5459, "step": 2225 }, { "epoch": 0.7241379310344828, "grad_norm": 1.7795367814818392, "learning_rate": 9.404885380941832e-06, "loss": 0.513, "step": 2226 }, { "epoch": 0.7244632400780742, "grad_norm": 2.4656427009647315, "learning_rate": 9.403989491982115e-06, "loss": 0.8259, "step": 2227 }, { "epoch": 0.7247885491216656, "grad_norm": 1.7667208340988878, "learning_rate": 9.403092971927767e-06, "loss": 0.6538, "step": 2228 }, { "epoch": 0.725113858165257, "grad_norm": 1.8022160705172265, "learning_rate": 9.402195820907262e-06, "loss": 0.6437, "step": 2229 }, { "epoch": 0.7254391672088484, "grad_norm": 2.069140343815953, "learning_rate": 9.40129803904916e-06, "loss": 0.6589, "step": 2230 }, { "epoch": 0.7257644762524398, "grad_norm": 1.776662556615743, "learning_rate": 9.400399626482117e-06, "loss": 0.5559, "step": 2231 }, { "epoch": 0.7260897852960312, "grad_norm": 1.7320764177768957, "learning_rate": 9.39950058333487e-06, "loss": 0.5405, "step": 2232 }, { "epoch": 0.7264150943396226, "grad_norm": 1.8387131536882526, "learning_rate": 9.398600909736257e-06, "loss": 0.6644, "step": 2233 }, { "epoch": 0.7267404033832141, "grad_norm": 1.7109851101089193, "learning_rate": 9.397700605815203e-06, "loss": 0.5835, "step": 2234 }, { "epoch": 0.7270657124268055, "grad_norm": 1.6356910034463046, "learning_rate": 9.396799671700719e-06, "loss": 0.4781, "step": 2235 }, { "epoch": 0.7273910214703969, "grad_norm": 1.8702368936075462, "learning_rate": 9.39589810752191e-06, "loss": 0.6083, "step": 2236 }, { "epoch": 0.7277163305139883, "grad_norm": 1.7055275246337391, "learning_rate": 9.394995913407972e-06, "loss": 0.4914, "step": 2237 }, { "epoch": 0.7280416395575797, "grad_norm": 1.5779945095288792, "learning_rate": 9.394093089488189e-06, "loss": 0.5597, "step": 2238 }, { "epoch": 0.7283669486011711, "grad_norm": 1.9503922444563602, "learning_rate": 9.393189635891935e-06, "loss": 0.6419, "step": 2239 }, { "epoch": 0.7286922576447625, "grad_norm": 1.9553853269958692, "learning_rate": 9.392285552748679e-06, "loss": 0.6444, "step": 2240 }, { "epoch": 0.7290175666883539, "grad_norm": 1.753390525916602, "learning_rate": 9.391380840187974e-06, "loss": 0.5766, "step": 2241 }, { "epoch": 0.7293428757319453, "grad_norm": 1.9235077644905019, "learning_rate": 9.390475498339469e-06, "loss": 0.7196, "step": 2242 }, { "epoch": 0.7296681847755367, "grad_norm": 1.760646781725512, "learning_rate": 9.389569527332895e-06, "loss": 0.6072, "step": 2243 }, { "epoch": 0.7299934938191281, "grad_norm": 1.860980552033928, "learning_rate": 9.388662927298082e-06, "loss": 0.7342, "step": 2244 }, { "epoch": 0.7303188028627196, "grad_norm": 1.8740708308168503, "learning_rate": 9.387755698364945e-06, "loss": 0.5853, "step": 2245 }, { "epoch": 0.730644111906311, "grad_norm": 2.0968986624838832, "learning_rate": 9.386847840663491e-06, "loss": 0.6162, "step": 2246 }, { "epoch": 0.7309694209499024, "grad_norm": 1.675868081682591, "learning_rate": 9.385939354323819e-06, "loss": 0.4803, "step": 2247 }, { "epoch": 0.7312947299934938, "grad_norm": 1.9462898017670152, "learning_rate": 9.38503023947611e-06, "loss": 0.632, "step": 2248 }, { "epoch": 0.7316200390370853, "grad_norm": 1.8516374554193185, "learning_rate": 9.384120496250646e-06, "loss": 0.5803, "step": 2249 }, { "epoch": 0.7319453480806767, "grad_norm": 1.9687850106812987, "learning_rate": 9.383210124777791e-06, "loss": 0.6308, "step": 2250 }, { "epoch": 0.7322706571242681, "grad_norm": 1.7862020985229505, "learning_rate": 9.382299125188004e-06, "loss": 0.6153, "step": 2251 }, { "epoch": 0.7325959661678595, "grad_norm": 1.7827398310633713, "learning_rate": 9.381387497611828e-06, "loss": 0.5321, "step": 2252 }, { "epoch": 0.7329212752114509, "grad_norm": 2.3528131496200007, "learning_rate": 9.380475242179906e-06, "loss": 0.6522, "step": 2253 }, { "epoch": 0.7332465842550423, "grad_norm": 2.067112276401984, "learning_rate": 9.379562359022958e-06, "loss": 0.7013, "step": 2254 }, { "epoch": 0.7335718932986337, "grad_norm": 1.8481643206653235, "learning_rate": 9.378648848271806e-06, "loss": 0.624, "step": 2255 }, { "epoch": 0.7338972023422251, "grad_norm": 1.856896700380652, "learning_rate": 9.377734710057353e-06, "loss": 0.5902, "step": 2256 }, { "epoch": 0.7342225113858165, "grad_norm": 2.040473533623585, "learning_rate": 9.376819944510598e-06, "loss": 0.5986, "step": 2257 }, { "epoch": 0.7345478204294079, "grad_norm": 1.9673409340113943, "learning_rate": 9.375904551762627e-06, "loss": 0.6533, "step": 2258 }, { "epoch": 0.7348731294729993, "grad_norm": 1.7127117386421198, "learning_rate": 9.374988531944617e-06, "loss": 0.6454, "step": 2259 }, { "epoch": 0.7351984385165907, "grad_norm": 1.9633002939202966, "learning_rate": 9.374071885187832e-06, "loss": 0.6458, "step": 2260 }, { "epoch": 0.7355237475601821, "grad_norm": 1.6535280616479924, "learning_rate": 9.373154611623629e-06, "loss": 0.622, "step": 2261 }, { "epoch": 0.7358490566037735, "grad_norm": 2.0208035526853445, "learning_rate": 9.372236711383457e-06, "loss": 0.6513, "step": 2262 }, { "epoch": 0.736174365647365, "grad_norm": 1.9502742447598658, "learning_rate": 9.371318184598847e-06, "loss": 0.6753, "step": 2263 }, { "epoch": 0.7364996746909565, "grad_norm": 1.7416829845794468, "learning_rate": 9.370399031401427e-06, "loss": 0.5395, "step": 2264 }, { "epoch": 0.7368249837345479, "grad_norm": 1.6608483734739634, "learning_rate": 9.369479251922913e-06, "loss": 0.5488, "step": 2265 }, { "epoch": 0.7371502927781393, "grad_norm": 1.8545448527973287, "learning_rate": 9.368558846295108e-06, "loss": 0.5842, "step": 2266 }, { "epoch": 0.7374756018217307, "grad_norm": 1.9603554858792256, "learning_rate": 9.367637814649908e-06, "loss": 0.6398, "step": 2267 }, { "epoch": 0.7378009108653221, "grad_norm": 1.8630295210186651, "learning_rate": 9.366716157119296e-06, "loss": 0.6403, "step": 2268 }, { "epoch": 0.7381262199089135, "grad_norm": 1.8575136274820707, "learning_rate": 9.36579387383535e-06, "loss": 0.6231, "step": 2269 }, { "epoch": 0.7384515289525049, "grad_norm": 1.915398622962438, "learning_rate": 9.364870964930229e-06, "loss": 0.6126, "step": 2270 }, { "epoch": 0.7387768379960963, "grad_norm": 1.8412218800197455, "learning_rate": 9.363947430536189e-06, "loss": 0.5684, "step": 2271 }, { "epoch": 0.7391021470396877, "grad_norm": 1.823719411878022, "learning_rate": 9.363023270785572e-06, "loss": 0.5816, "step": 2272 }, { "epoch": 0.7394274560832791, "grad_norm": 1.8079700249703812, "learning_rate": 9.362098485810811e-06, "loss": 0.6895, "step": 2273 }, { "epoch": 0.7397527651268705, "grad_norm": 1.9261641542420835, "learning_rate": 9.36117307574443e-06, "loss": 0.5718, "step": 2274 }, { "epoch": 0.7400780741704619, "grad_norm": 2.1606784496051383, "learning_rate": 9.36024704071904e-06, "loss": 0.6268, "step": 2275 }, { "epoch": 0.7404033832140533, "grad_norm": 1.5950647097115145, "learning_rate": 9.35932038086734e-06, "loss": 0.4936, "step": 2276 }, { "epoch": 0.7407286922576448, "grad_norm": 1.8109152317168842, "learning_rate": 9.358393096322125e-06, "loss": 0.5573, "step": 2277 }, { "epoch": 0.7410540013012362, "grad_norm": 2.215786769392306, "learning_rate": 9.357465187216271e-06, "loss": 0.6366, "step": 2278 }, { "epoch": 0.7413793103448276, "grad_norm": 1.6885411669461392, "learning_rate": 9.356536653682752e-06, "loss": 0.5569, "step": 2279 }, { "epoch": 0.741704619388419, "grad_norm": 1.7641060004847677, "learning_rate": 9.355607495854625e-06, "loss": 0.5432, "step": 2280 }, { "epoch": 0.7420299284320104, "grad_norm": 1.7683903840122916, "learning_rate": 9.354677713865042e-06, "loss": 0.5665, "step": 2281 }, { "epoch": 0.7423552374756018, "grad_norm": 1.6665027442692928, "learning_rate": 9.353747307847237e-06, "loss": 0.5506, "step": 2282 }, { "epoch": 0.7426805465191932, "grad_norm": 2.0025668831072267, "learning_rate": 9.35281627793454e-06, "loss": 0.524, "step": 2283 }, { "epoch": 0.7430058555627846, "grad_norm": 1.937034281050326, "learning_rate": 9.351884624260369e-06, "loss": 0.6595, "step": 2284 }, { "epoch": 0.743331164606376, "grad_norm": 1.69901111971438, "learning_rate": 9.35095234695823e-06, "loss": 0.5175, "step": 2285 }, { "epoch": 0.7436564736499675, "grad_norm": 1.853752198236785, "learning_rate": 9.350019446161717e-06, "loss": 0.5944, "step": 2286 }, { "epoch": 0.7439817826935589, "grad_norm": 1.749762480309813, "learning_rate": 9.34908592200452e-06, "loss": 0.6336, "step": 2287 }, { "epoch": 0.7443070917371503, "grad_norm": 1.8159318161564502, "learning_rate": 9.348151774620408e-06, "loss": 0.643, "step": 2288 }, { "epoch": 0.7446324007807417, "grad_norm": 1.9555007096786208, "learning_rate": 9.34721700414325e-06, "loss": 0.5364, "step": 2289 }, { "epoch": 0.7449577098243331, "grad_norm": 2.0281393933350502, "learning_rate": 9.346281610706996e-06, "loss": 0.5978, "step": 2290 }, { "epoch": 0.7452830188679245, "grad_norm": 1.7705098657024323, "learning_rate": 9.34534559444569e-06, "loss": 0.6406, "step": 2291 }, { "epoch": 0.745608327911516, "grad_norm": 1.8092437519131908, "learning_rate": 9.344408955493463e-06, "loss": 0.613, "step": 2292 }, { "epoch": 0.7459336369551074, "grad_norm": 1.9938975962451504, "learning_rate": 9.343471693984534e-06, "loss": 0.6181, "step": 2293 }, { "epoch": 0.7462589459986988, "grad_norm": 1.497056200124492, "learning_rate": 9.342533810053218e-06, "loss": 0.4608, "step": 2294 }, { "epoch": 0.7465842550422902, "grad_norm": 1.716638226791072, "learning_rate": 9.34159530383391e-06, "loss": 0.5174, "step": 2295 }, { "epoch": 0.7469095640858816, "grad_norm": 1.811578117311291, "learning_rate": 9.3406561754611e-06, "loss": 0.6842, "step": 2296 }, { "epoch": 0.747234873129473, "grad_norm": 1.7505224165802042, "learning_rate": 9.339716425069368e-06, "loss": 0.6809, "step": 2297 }, { "epoch": 0.7475601821730644, "grad_norm": 1.7279037647425706, "learning_rate": 9.338776052793377e-06, "loss": 0.5576, "step": 2298 }, { "epoch": 0.7478854912166558, "grad_norm": 1.6368002444123595, "learning_rate": 9.337835058767886e-06, "loss": 0.5734, "step": 2299 }, { "epoch": 0.7482108002602472, "grad_norm": 2.2281956157428073, "learning_rate": 9.336893443127739e-06, "loss": 0.8141, "step": 2300 }, { "epoch": 0.7485361093038386, "grad_norm": 1.7844274645604277, "learning_rate": 9.33595120600787e-06, "loss": 0.5883, "step": 2301 }, { "epoch": 0.74886141834743, "grad_norm": 1.9320802985098224, "learning_rate": 9.3350083475433e-06, "loss": 0.5984, "step": 2302 }, { "epoch": 0.7491867273910214, "grad_norm": 2.11793074071067, "learning_rate": 9.334064867869145e-06, "loss": 0.6373, "step": 2303 }, { "epoch": 0.7495120364346128, "grad_norm": 1.9183381608169068, "learning_rate": 9.333120767120607e-06, "loss": 0.6639, "step": 2304 }, { "epoch": 0.7498373454782042, "grad_norm": 1.9393709272321964, "learning_rate": 9.332176045432971e-06, "loss": 0.6507, "step": 2305 }, { "epoch": 0.7501626545217958, "grad_norm": 1.9848318178285336, "learning_rate": 9.33123070294162e-06, "loss": 0.6633, "step": 2306 }, { "epoch": 0.7504879635653872, "grad_norm": 1.9695893016586623, "learning_rate": 9.330284739782022e-06, "loss": 0.5958, "step": 2307 }, { "epoch": 0.7508132726089786, "grad_norm": 1.806505017605035, "learning_rate": 9.329338156089735e-06, "loss": 0.6206, "step": 2308 }, { "epoch": 0.75113858165257, "grad_norm": 1.920651585089439, "learning_rate": 9.328390952000402e-06, "loss": 0.6815, "step": 2309 }, { "epoch": 0.7514638906961614, "grad_norm": 1.7495337685459127, "learning_rate": 9.32744312764976e-06, "loss": 0.5375, "step": 2310 }, { "epoch": 0.7517891997397528, "grad_norm": 1.721307659485904, "learning_rate": 9.326494683173632e-06, "loss": 0.4905, "step": 2311 }, { "epoch": 0.7521145087833442, "grad_norm": 1.9390740001191586, "learning_rate": 9.325545618707932e-06, "loss": 0.5985, "step": 2312 }, { "epoch": 0.7524398178269356, "grad_norm": 1.8867535896980794, "learning_rate": 9.324595934388661e-06, "loss": 0.6477, "step": 2313 }, { "epoch": 0.752765126870527, "grad_norm": 1.691898538715496, "learning_rate": 9.323645630351909e-06, "loss": 0.5711, "step": 2314 }, { "epoch": 0.7530904359141184, "grad_norm": 1.9943317242329694, "learning_rate": 9.322694706733857e-06, "loss": 0.6967, "step": 2315 }, { "epoch": 0.7534157449577098, "grad_norm": 1.941370596524822, "learning_rate": 9.321743163670769e-06, "loss": 0.5191, "step": 2316 }, { "epoch": 0.7537410540013012, "grad_norm": 1.9073057858666855, "learning_rate": 9.320791001299006e-06, "loss": 0.5413, "step": 2317 }, { "epoch": 0.7540663630448926, "grad_norm": 1.656344023022331, "learning_rate": 9.319838219755011e-06, "loss": 0.5407, "step": 2318 }, { "epoch": 0.754391672088484, "grad_norm": 1.5740469977621838, "learning_rate": 9.318884819175317e-06, "loss": 0.4749, "step": 2319 }, { "epoch": 0.7547169811320755, "grad_norm": 1.9614936102095528, "learning_rate": 9.317930799696553e-06, "loss": 0.7328, "step": 2320 }, { "epoch": 0.7550422901756669, "grad_norm": 1.7508345729834063, "learning_rate": 9.316976161455425e-06, "loss": 0.5582, "step": 2321 }, { "epoch": 0.7553675992192583, "grad_norm": 1.6925308916998734, "learning_rate": 9.316020904588733e-06, "loss": 0.6052, "step": 2322 }, { "epoch": 0.7556929082628497, "grad_norm": 1.6767344409846923, "learning_rate": 9.31506502923337e-06, "loss": 0.5623, "step": 2323 }, { "epoch": 0.7560182173064411, "grad_norm": 1.674306143448844, "learning_rate": 9.31410853552631e-06, "loss": 0.5465, "step": 2324 }, { "epoch": 0.7563435263500325, "grad_norm": 2.053130093491741, "learning_rate": 9.313151423604621e-06, "loss": 0.6533, "step": 2325 }, { "epoch": 0.756668835393624, "grad_norm": 1.8192055778916678, "learning_rate": 9.312193693605456e-06, "loss": 0.5794, "step": 2326 }, { "epoch": 0.7569941444372154, "grad_norm": 1.8304128476976327, "learning_rate": 9.31123534566606e-06, "loss": 0.5438, "step": 2327 }, { "epoch": 0.7573194534808068, "grad_norm": 1.775398050843437, "learning_rate": 9.310276379923766e-06, "loss": 0.6042, "step": 2328 }, { "epoch": 0.7576447625243982, "grad_norm": 1.9575110464880405, "learning_rate": 9.309316796515993e-06, "loss": 0.6668, "step": 2329 }, { "epoch": 0.7579700715679896, "grad_norm": 1.5954884715098703, "learning_rate": 9.308356595580248e-06, "loss": 0.5746, "step": 2330 }, { "epoch": 0.758295380611581, "grad_norm": 1.8145156875804043, "learning_rate": 9.30739577725413e-06, "loss": 0.5996, "step": 2331 }, { "epoch": 0.7586206896551724, "grad_norm": 1.52320517635665, "learning_rate": 9.306434341675328e-06, "loss": 0.5129, "step": 2332 }, { "epoch": 0.7589459986987638, "grad_norm": 1.7718609944174901, "learning_rate": 9.305472288981612e-06, "loss": 0.5745, "step": 2333 }, { "epoch": 0.7592713077423552, "grad_norm": 1.9502076692261499, "learning_rate": 9.304509619310847e-06, "loss": 0.6859, "step": 2334 }, { "epoch": 0.7595966167859467, "grad_norm": 1.7701887125008462, "learning_rate": 9.303546332800984e-06, "loss": 0.59, "step": 2335 }, { "epoch": 0.7599219258295381, "grad_norm": 1.9469014925279078, "learning_rate": 9.302582429590061e-06, "loss": 0.6035, "step": 2336 }, { "epoch": 0.7602472348731295, "grad_norm": 1.7373208901299033, "learning_rate": 9.301617909816207e-06, "loss": 0.4852, "step": 2337 }, { "epoch": 0.7605725439167209, "grad_norm": 1.9705264855624327, "learning_rate": 9.30065277361764e-06, "loss": 0.6816, "step": 2338 }, { "epoch": 0.7608978529603123, "grad_norm": 1.7633607026546612, "learning_rate": 9.299687021132662e-06, "loss": 0.619, "step": 2339 }, { "epoch": 0.7612231620039037, "grad_norm": 2.004421857577905, "learning_rate": 9.298720652499667e-06, "loss": 0.5836, "step": 2340 }, { "epoch": 0.7615484710474951, "grad_norm": 1.9225465879523618, "learning_rate": 9.297753667857138e-06, "loss": 0.648, "step": 2341 }, { "epoch": 0.7618737800910865, "grad_norm": 1.8873454325230619, "learning_rate": 9.296786067343641e-06, "loss": 0.575, "step": 2342 }, { "epoch": 0.7621990891346779, "grad_norm": 1.6607036651325509, "learning_rate": 9.295817851097836e-06, "loss": 0.5761, "step": 2343 }, { "epoch": 0.7625243981782693, "grad_norm": 1.9328876864544844, "learning_rate": 9.29484901925847e-06, "loss": 0.5636, "step": 2344 }, { "epoch": 0.7628497072218607, "grad_norm": 1.9514703111655725, "learning_rate": 9.293879571964376e-06, "loss": 0.5646, "step": 2345 }, { "epoch": 0.7631750162654521, "grad_norm": 1.6228398755335471, "learning_rate": 9.292909509354476e-06, "loss": 0.5049, "step": 2346 }, { "epoch": 0.7635003253090435, "grad_norm": 1.7150283920561877, "learning_rate": 9.291938831567781e-06, "loss": 0.5764, "step": 2347 }, { "epoch": 0.763825634352635, "grad_norm": 1.9059564566716143, "learning_rate": 9.29096753874339e-06, "loss": 0.6074, "step": 2348 }, { "epoch": 0.7641509433962265, "grad_norm": 1.950276362024176, "learning_rate": 9.28999563102049e-06, "loss": 0.6669, "step": 2349 }, { "epoch": 0.7644762524398179, "grad_norm": 1.8242869728004445, "learning_rate": 9.289023108538355e-06, "loss": 0.6779, "step": 2350 }, { "epoch": 0.7648015614834093, "grad_norm": 1.863053032264282, "learning_rate": 9.28804997143635e-06, "loss": 0.5966, "step": 2351 }, { "epoch": 0.7651268705270007, "grad_norm": 1.7606343966225126, "learning_rate": 9.287076219853923e-06, "loss": 0.664, "step": 2352 }, { "epoch": 0.7654521795705921, "grad_norm": 1.841105147806749, "learning_rate": 9.286101853930618e-06, "loss": 0.6791, "step": 2353 }, { "epoch": 0.7657774886141835, "grad_norm": 1.6392939679901908, "learning_rate": 9.28512687380606e-06, "loss": 0.6326, "step": 2354 }, { "epoch": 0.7661027976577749, "grad_norm": 1.9411633868133917, "learning_rate": 9.284151279619961e-06, "loss": 0.5748, "step": 2355 }, { "epoch": 0.7664281067013663, "grad_norm": 1.786509485935223, "learning_rate": 9.28317507151213e-06, "loss": 0.5681, "step": 2356 }, { "epoch": 0.7667534157449577, "grad_norm": 1.9134182225277567, "learning_rate": 9.282198249622455e-06, "loss": 0.7257, "step": 2357 }, { "epoch": 0.7670787247885491, "grad_norm": 2.0163471370874206, "learning_rate": 9.281220814090913e-06, "loss": 0.7162, "step": 2358 }, { "epoch": 0.7674040338321405, "grad_norm": 1.7903524319229964, "learning_rate": 9.280242765057576e-06, "loss": 0.6431, "step": 2359 }, { "epoch": 0.7677293428757319, "grad_norm": 1.7939383186968392, "learning_rate": 9.279264102662598e-06, "loss": 0.6125, "step": 2360 }, { "epoch": 0.7680546519193233, "grad_norm": 1.8645338557518107, "learning_rate": 9.27828482704622e-06, "loss": 0.5146, "step": 2361 }, { "epoch": 0.7683799609629147, "grad_norm": 1.6342994293198154, "learning_rate": 9.277304938348773e-06, "loss": 0.6003, "step": 2362 }, { "epoch": 0.7687052700065062, "grad_norm": 1.7318275907038911, "learning_rate": 9.276324436710677e-06, "loss": 0.5298, "step": 2363 }, { "epoch": 0.7690305790500976, "grad_norm": 1.8855345962644172, "learning_rate": 9.275343322272438e-06, "loss": 0.5209, "step": 2364 }, { "epoch": 0.769355888093689, "grad_norm": 1.9053395795282477, "learning_rate": 9.274361595174652e-06, "loss": 0.5935, "step": 2365 }, { "epoch": 0.7696811971372804, "grad_norm": 2.1857444449741505, "learning_rate": 9.273379255557998e-06, "loss": 0.5899, "step": 2366 }, { "epoch": 0.7700065061808719, "grad_norm": 1.6859816920139936, "learning_rate": 9.272396303563249e-06, "loss": 0.5863, "step": 2367 }, { "epoch": 0.7703318152244633, "grad_norm": 1.9282141548139853, "learning_rate": 9.27141273933126e-06, "loss": 0.5094, "step": 2368 }, { "epoch": 0.7706571242680547, "grad_norm": 1.7879208909008892, "learning_rate": 9.270428563002978e-06, "loss": 0.5847, "step": 2369 }, { "epoch": 0.7709824333116461, "grad_norm": 1.843160119685152, "learning_rate": 9.269443774719435e-06, "loss": 0.5369, "step": 2370 }, { "epoch": 0.7713077423552375, "grad_norm": 1.7792015188476027, "learning_rate": 9.268458374621752e-06, "loss": 0.5416, "step": 2371 }, { "epoch": 0.7716330513988289, "grad_norm": 2.27288213893532, "learning_rate": 9.26747236285114e-06, "loss": 0.7164, "step": 2372 }, { "epoch": 0.7719583604424203, "grad_norm": 1.7361372404126303, "learning_rate": 9.266485739548892e-06, "loss": 0.5351, "step": 2373 }, { "epoch": 0.7722836694860117, "grad_norm": 1.5211407940457364, "learning_rate": 9.265498504856391e-06, "loss": 0.4788, "step": 2374 }, { "epoch": 0.7726089785296031, "grad_norm": 1.759302044143912, "learning_rate": 9.264510658915114e-06, "loss": 0.5721, "step": 2375 }, { "epoch": 0.7729342875731945, "grad_norm": 1.9588036507629927, "learning_rate": 9.263522201866614e-06, "loss": 0.6309, "step": 2376 }, { "epoch": 0.7732595966167859, "grad_norm": 1.8769928179990805, "learning_rate": 9.26253313385254e-06, "loss": 0.6326, "step": 2377 }, { "epoch": 0.7735849056603774, "grad_norm": 1.6199647039220553, "learning_rate": 9.261543455014627e-06, "loss": 0.5572, "step": 2378 }, { "epoch": 0.7739102147039688, "grad_norm": 1.6982608633164573, "learning_rate": 9.260553165494695e-06, "loss": 0.4289, "step": 2379 }, { "epoch": 0.7742355237475602, "grad_norm": 1.7425471833730815, "learning_rate": 9.259562265434653e-06, "loss": 0.5019, "step": 2380 }, { "epoch": 0.7745608327911516, "grad_norm": 1.9503936541386906, "learning_rate": 9.2585707549765e-06, "loss": 0.5782, "step": 2381 }, { "epoch": 0.774886141834743, "grad_norm": 1.5446937604955673, "learning_rate": 9.257578634262317e-06, "loss": 0.5342, "step": 2382 }, { "epoch": 0.7752114508783344, "grad_norm": 1.7802693501656126, "learning_rate": 9.256585903434279e-06, "loss": 0.6658, "step": 2383 }, { "epoch": 0.7755367599219258, "grad_norm": 1.9933380557641196, "learning_rate": 9.255592562634642e-06, "loss": 0.6627, "step": 2384 }, { "epoch": 0.7758620689655172, "grad_norm": 1.9634102373595335, "learning_rate": 9.254598612005752e-06, "loss": 0.6109, "step": 2385 }, { "epoch": 0.7761873780091086, "grad_norm": 1.6991980975517096, "learning_rate": 9.253604051690047e-06, "loss": 0.6137, "step": 2386 }, { "epoch": 0.7765126870527, "grad_norm": 1.7984773789894135, "learning_rate": 9.252608881830044e-06, "loss": 0.5531, "step": 2387 }, { "epoch": 0.7768379960962914, "grad_norm": 1.6624984301461598, "learning_rate": 9.251613102568351e-06, "loss": 0.5407, "step": 2388 }, { "epoch": 0.7771633051398829, "grad_norm": 1.9936646974018573, "learning_rate": 9.250616714047668e-06, "loss": 0.6514, "step": 2389 }, { "epoch": 0.7774886141834743, "grad_norm": 1.9724512940215273, "learning_rate": 9.249619716410774e-06, "loss": 0.559, "step": 2390 }, { "epoch": 0.7778139232270657, "grad_norm": 1.7944988277128255, "learning_rate": 9.248622109800544e-06, "loss": 0.5216, "step": 2391 }, { "epoch": 0.7781392322706572, "grad_norm": 1.7490175417373304, "learning_rate": 9.247623894359931e-06, "loss": 0.5809, "step": 2392 }, { "epoch": 0.7784645413142486, "grad_norm": 1.7625791845988803, "learning_rate": 9.246625070231983e-06, "loss": 0.6464, "step": 2393 }, { "epoch": 0.77878985035784, "grad_norm": 1.9610534374887925, "learning_rate": 9.24562563755983e-06, "loss": 0.5765, "step": 2394 }, { "epoch": 0.7791151594014314, "grad_norm": 1.7831904966288235, "learning_rate": 9.244625596486693e-06, "loss": 0.6245, "step": 2395 }, { "epoch": 0.7794404684450228, "grad_norm": 2.025550766045844, "learning_rate": 9.243624947155878e-06, "loss": 0.7487, "step": 2396 }, { "epoch": 0.7797657774886142, "grad_norm": 2.0468303339335994, "learning_rate": 9.242623689710778e-06, "loss": 0.6384, "step": 2397 }, { "epoch": 0.7800910865322056, "grad_norm": 1.9618924245733855, "learning_rate": 9.241621824294877e-06, "loss": 0.6316, "step": 2398 }, { "epoch": 0.780416395575797, "grad_norm": 1.9607055084931928, "learning_rate": 9.240619351051739e-06, "loss": 0.5718, "step": 2399 }, { "epoch": 0.7807417046193884, "grad_norm": 1.7565590430083229, "learning_rate": 9.23961627012502e-06, "loss": 0.5583, "step": 2400 }, { "epoch": 0.7810670136629798, "grad_norm": 1.6201396977011506, "learning_rate": 9.238612581658465e-06, "loss": 0.4957, "step": 2401 }, { "epoch": 0.7813923227065712, "grad_norm": 1.6710746462522639, "learning_rate": 9.237608285795899e-06, "loss": 0.5404, "step": 2402 }, { "epoch": 0.7817176317501626, "grad_norm": 1.6358706712548359, "learning_rate": 9.23660338268124e-06, "loss": 0.6176, "step": 2403 }, { "epoch": 0.782042940793754, "grad_norm": 1.8045358323311915, "learning_rate": 9.235597872458491e-06, "loss": 0.5929, "step": 2404 }, { "epoch": 0.7823682498373454, "grad_norm": 1.7056969878870514, "learning_rate": 9.234591755271744e-06, "loss": 0.5557, "step": 2405 }, { "epoch": 0.7826935588809368, "grad_norm": 1.7664912498931733, "learning_rate": 9.233585031265175e-06, "loss": 0.6265, "step": 2406 }, { "epoch": 0.7830188679245284, "grad_norm": 1.9466087858076513, "learning_rate": 9.23257770058305e-06, "loss": 0.6219, "step": 2407 }, { "epoch": 0.7833441769681198, "grad_norm": 1.8594796710142336, "learning_rate": 9.231569763369716e-06, "loss": 0.763, "step": 2408 }, { "epoch": 0.7836694860117112, "grad_norm": 1.807901929646028, "learning_rate": 9.230561219769616e-06, "loss": 0.5766, "step": 2409 }, { "epoch": 0.7839947950553026, "grad_norm": 1.8752961400703172, "learning_rate": 9.229552069927272e-06, "loss": 0.648, "step": 2410 }, { "epoch": 0.784320104098894, "grad_norm": 1.6735137604038546, "learning_rate": 9.228542313987293e-06, "loss": 0.5473, "step": 2411 }, { "epoch": 0.7846454131424854, "grad_norm": 2.1934304612844846, "learning_rate": 9.227531952094387e-06, "loss": 0.5953, "step": 2412 }, { "epoch": 0.7849707221860768, "grad_norm": 1.8014318312031268, "learning_rate": 9.22652098439333e-06, "loss": 0.6517, "step": 2413 }, { "epoch": 0.7852960312296682, "grad_norm": 1.7705444049826085, "learning_rate": 9.225509411028999e-06, "loss": 0.599, "step": 2414 }, { "epoch": 0.7856213402732596, "grad_norm": 1.7869900517590085, "learning_rate": 9.22449723214635e-06, "loss": 0.5107, "step": 2415 }, { "epoch": 0.785946649316851, "grad_norm": 1.9607870522254887, "learning_rate": 9.223484447890435e-06, "loss": 0.6389, "step": 2416 }, { "epoch": 0.7862719583604424, "grad_norm": 1.9003605712231109, "learning_rate": 9.222471058406379e-06, "loss": 0.6423, "step": 2417 }, { "epoch": 0.7865972674040338, "grad_norm": 1.9650993887846353, "learning_rate": 9.221457063839407e-06, "loss": 0.6974, "step": 2418 }, { "epoch": 0.7869225764476252, "grad_norm": 1.8037683748503484, "learning_rate": 9.220442464334823e-06, "loss": 0.6453, "step": 2419 }, { "epoch": 0.7872478854912166, "grad_norm": 1.7839880536127204, "learning_rate": 9.219427260038018e-06, "loss": 0.5337, "step": 2420 }, { "epoch": 0.7875731945348081, "grad_norm": 1.799926373957024, "learning_rate": 9.218411451094475e-06, "loss": 0.6005, "step": 2421 }, { "epoch": 0.7878985035783995, "grad_norm": 1.649702529823099, "learning_rate": 9.21739503764976e-06, "loss": 0.5435, "step": 2422 }, { "epoch": 0.7882238126219909, "grad_norm": 2.0156293746440026, "learning_rate": 9.216378019849523e-06, "loss": 0.6439, "step": 2423 }, { "epoch": 0.7885491216655823, "grad_norm": 1.7924094817678735, "learning_rate": 9.215360397839504e-06, "loss": 0.5519, "step": 2424 }, { "epoch": 0.7888744307091737, "grad_norm": 1.8037290480991113, "learning_rate": 9.21434217176553e-06, "loss": 0.5186, "step": 2425 }, { "epoch": 0.7891997397527651, "grad_norm": 1.833535452444719, "learning_rate": 9.213323341773512e-06, "loss": 0.546, "step": 2426 }, { "epoch": 0.7895250487963565, "grad_norm": 1.787405311800099, "learning_rate": 9.212303908009451e-06, "loss": 0.5562, "step": 2427 }, { "epoch": 0.789850357839948, "grad_norm": 1.6822642857900187, "learning_rate": 9.211283870619431e-06, "loss": 0.6161, "step": 2428 }, { "epoch": 0.7901756668835394, "grad_norm": 1.78752376709451, "learning_rate": 9.210263229749626e-06, "loss": 0.5647, "step": 2429 }, { "epoch": 0.7905009759271308, "grad_norm": 1.8275096033368017, "learning_rate": 9.20924198554629e-06, "loss": 0.4978, "step": 2430 }, { "epoch": 0.7908262849707222, "grad_norm": 2.081363895371776, "learning_rate": 9.208220138155774e-06, "loss": 0.6592, "step": 2431 }, { "epoch": 0.7911515940143136, "grad_norm": 1.8525318214606747, "learning_rate": 9.207197687724506e-06, "loss": 0.583, "step": 2432 }, { "epoch": 0.791476903057905, "grad_norm": 1.727510865009268, "learning_rate": 9.206174634399003e-06, "loss": 0.4479, "step": 2433 }, { "epoch": 0.7918022121014964, "grad_norm": 1.7911244781344644, "learning_rate": 9.205150978325872e-06, "loss": 0.594, "step": 2434 }, { "epoch": 0.7921275211450879, "grad_norm": 1.7677809899506842, "learning_rate": 9.204126719651801e-06, "loss": 0.6776, "step": 2435 }, { "epoch": 0.7924528301886793, "grad_norm": 1.787154805483964, "learning_rate": 9.203101858523569e-06, "loss": 0.6267, "step": 2436 }, { "epoch": 0.7927781392322707, "grad_norm": 1.7682952325634136, "learning_rate": 9.202076395088038e-06, "loss": 0.5706, "step": 2437 }, { "epoch": 0.7931034482758621, "grad_norm": 1.8336884740555337, "learning_rate": 9.201050329492156e-06, "loss": 0.5693, "step": 2438 }, { "epoch": 0.7934287573194535, "grad_norm": 1.9089297362904387, "learning_rate": 9.200023661882963e-06, "loss": 0.5873, "step": 2439 }, { "epoch": 0.7937540663630449, "grad_norm": 1.777387035485139, "learning_rate": 9.19899639240758e-06, "loss": 0.6123, "step": 2440 }, { "epoch": 0.7940793754066363, "grad_norm": 1.845933311279525, "learning_rate": 9.197968521213211e-06, "loss": 0.6237, "step": 2441 }, { "epoch": 0.7944046844502277, "grad_norm": 1.696608747602296, "learning_rate": 9.196940048447157e-06, "loss": 0.5815, "step": 2442 }, { "epoch": 0.7947299934938191, "grad_norm": 1.7689860334544827, "learning_rate": 9.195910974256795e-06, "loss": 0.5754, "step": 2443 }, { "epoch": 0.7950553025374105, "grad_norm": 1.5419795912470466, "learning_rate": 9.194881298789592e-06, "loss": 0.458, "step": 2444 }, { "epoch": 0.7953806115810019, "grad_norm": 2.2019403382016094, "learning_rate": 9.193851022193103e-06, "loss": 0.7237, "step": 2445 }, { "epoch": 0.7957059206245933, "grad_norm": 1.9245279689212562, "learning_rate": 9.192820144614964e-06, "loss": 0.5581, "step": 2446 }, { "epoch": 0.7960312296681847, "grad_norm": 1.9680700056403766, "learning_rate": 9.191788666202905e-06, "loss": 0.6002, "step": 2447 }, { "epoch": 0.7963565387117761, "grad_norm": 2.149885632905272, "learning_rate": 9.190756587104736e-06, "loss": 0.6467, "step": 2448 }, { "epoch": 0.7966818477553675, "grad_norm": 1.6963929335303352, "learning_rate": 9.189723907468352e-06, "loss": 0.5384, "step": 2449 }, { "epoch": 0.7970071567989591, "grad_norm": 1.791920690260513, "learning_rate": 9.188690627441741e-06, "loss": 0.5408, "step": 2450 }, { "epoch": 0.7973324658425505, "grad_norm": 1.8874654010153658, "learning_rate": 9.187656747172968e-06, "loss": 0.6675, "step": 2451 }, { "epoch": 0.7976577748861419, "grad_norm": 1.7326931793298372, "learning_rate": 9.186622266810192e-06, "loss": 0.5947, "step": 2452 }, { "epoch": 0.7979830839297333, "grad_norm": 2.1159232410635216, "learning_rate": 9.185587186501654e-06, "loss": 0.6288, "step": 2453 }, { "epoch": 0.7983083929733247, "grad_norm": 1.5935030683170386, "learning_rate": 9.184551506395682e-06, "loss": 0.5996, "step": 2454 }, { "epoch": 0.7986337020169161, "grad_norm": 1.8401398874404642, "learning_rate": 9.18351522664069e-06, "loss": 0.6556, "step": 2455 }, { "epoch": 0.7989590110605075, "grad_norm": 1.751982333102711, "learning_rate": 9.182478347385175e-06, "loss": 0.6517, "step": 2456 }, { "epoch": 0.7992843201040989, "grad_norm": 1.8064381906420748, "learning_rate": 9.181440868777724e-06, "loss": 0.5348, "step": 2457 }, { "epoch": 0.7996096291476903, "grad_norm": 1.585031872091866, "learning_rate": 9.18040279096701e-06, "loss": 0.6019, "step": 2458 }, { "epoch": 0.7999349381912817, "grad_norm": 1.8578975149680221, "learning_rate": 9.179364114101788e-06, "loss": 0.6371, "step": 2459 }, { "epoch": 0.8002602472348731, "grad_norm": 1.8942814612209824, "learning_rate": 9.1783248383309e-06, "loss": 0.7003, "step": 2460 }, { "epoch": 0.8005855562784645, "grad_norm": 1.8049566760534956, "learning_rate": 9.177284963803282e-06, "loss": 0.6297, "step": 2461 }, { "epoch": 0.8009108653220559, "grad_norm": 1.7227981776282961, "learning_rate": 9.176244490667939e-06, "loss": 0.5898, "step": 2462 }, { "epoch": 0.8012361743656473, "grad_norm": 1.697089134057486, "learning_rate": 9.175203419073977e-06, "loss": 0.5752, "step": 2463 }, { "epoch": 0.8015614834092388, "grad_norm": 1.7209906887695403, "learning_rate": 9.174161749170583e-06, "loss": 0.5917, "step": 2464 }, { "epoch": 0.8018867924528302, "grad_norm": 1.6972744980064987, "learning_rate": 9.173119481107025e-06, "loss": 0.5611, "step": 2465 }, { "epoch": 0.8022121014964216, "grad_norm": 2.013259496527344, "learning_rate": 9.172076615032665e-06, "loss": 0.7377, "step": 2466 }, { "epoch": 0.802537410540013, "grad_norm": 1.8278078265606579, "learning_rate": 9.171033151096943e-06, "loss": 0.6537, "step": 2467 }, { "epoch": 0.8028627195836044, "grad_norm": 1.9879720974072976, "learning_rate": 9.16998908944939e-06, "loss": 0.6064, "step": 2468 }, { "epoch": 0.8031880286271958, "grad_norm": 1.7703110863860432, "learning_rate": 9.168944430239621e-06, "loss": 0.6939, "step": 2469 }, { "epoch": 0.8035133376707873, "grad_norm": 1.8231734421545092, "learning_rate": 9.167899173617336e-06, "loss": 0.4965, "step": 2470 }, { "epoch": 0.8038386467143787, "grad_norm": 1.6424213471456834, "learning_rate": 9.166853319732318e-06, "loss": 0.5984, "step": 2471 }, { "epoch": 0.8041639557579701, "grad_norm": 1.8271600594095772, "learning_rate": 9.165806868734444e-06, "loss": 0.4914, "step": 2472 }, { "epoch": 0.8044892648015615, "grad_norm": 1.7380111073759477, "learning_rate": 9.164759820773666e-06, "loss": 0.5344, "step": 2473 }, { "epoch": 0.8048145738451529, "grad_norm": 2.0421891938435843, "learning_rate": 9.163712176000029e-06, "loss": 0.7889, "step": 2474 }, { "epoch": 0.8051398828887443, "grad_norm": 2.1810798985965003, "learning_rate": 9.162663934563663e-06, "loss": 0.6204, "step": 2475 }, { "epoch": 0.8054651919323357, "grad_norm": 1.9290768010249488, "learning_rate": 9.161615096614778e-06, "loss": 0.7357, "step": 2476 }, { "epoch": 0.8057905009759271, "grad_norm": 1.7769786385696638, "learning_rate": 9.160565662303675e-06, "loss": 0.5664, "step": 2477 }, { "epoch": 0.8061158100195186, "grad_norm": 1.8224129394152826, "learning_rate": 9.15951563178074e-06, "loss": 0.6433, "step": 2478 }, { "epoch": 0.80644111906311, "grad_norm": 1.6498032293967386, "learning_rate": 9.15846500519644e-06, "loss": 0.573, "step": 2479 }, { "epoch": 0.8067664281067014, "grad_norm": 1.9113619122809502, "learning_rate": 9.157413782701334e-06, "loss": 0.5951, "step": 2480 }, { "epoch": 0.8070917371502928, "grad_norm": 1.9386552616494324, "learning_rate": 9.15636196444606e-06, "loss": 0.6469, "step": 2481 }, { "epoch": 0.8074170461938842, "grad_norm": 1.4890790649767511, "learning_rate": 9.155309550581346e-06, "loss": 0.4862, "step": 2482 }, { "epoch": 0.8077423552374756, "grad_norm": 1.6979096440640826, "learning_rate": 9.154256541258002e-06, "loss": 0.421, "step": 2483 }, { "epoch": 0.808067664281067, "grad_norm": 1.6945563239207864, "learning_rate": 9.153202936626927e-06, "loss": 0.5598, "step": 2484 }, { "epoch": 0.8083929733246584, "grad_norm": 1.8273799118898957, "learning_rate": 9.152148736839102e-06, "loss": 0.6698, "step": 2485 }, { "epoch": 0.8087182823682498, "grad_norm": 1.7503996715263637, "learning_rate": 9.151093942045594e-06, "loss": 0.5892, "step": 2486 }, { "epoch": 0.8090435914118412, "grad_norm": 1.741685022905209, "learning_rate": 9.150038552397555e-06, "loss": 0.5811, "step": 2487 }, { "epoch": 0.8093689004554326, "grad_norm": 2.1095505948664814, "learning_rate": 9.148982568046228e-06, "loss": 0.6467, "step": 2488 }, { "epoch": 0.809694209499024, "grad_norm": 1.8406885750576616, "learning_rate": 9.14792598914293e-06, "loss": 0.5229, "step": 2489 }, { "epoch": 0.8100195185426154, "grad_norm": 1.973642537971173, "learning_rate": 9.146868815839073e-06, "loss": 0.6326, "step": 2490 }, { "epoch": 0.8103448275862069, "grad_norm": 1.892860047349794, "learning_rate": 9.14581104828615e-06, "loss": 0.6034, "step": 2491 }, { "epoch": 0.8106701366297983, "grad_norm": 1.9501153015590977, "learning_rate": 9.144752686635739e-06, "loss": 0.618, "step": 2492 }, { "epoch": 0.8109954456733898, "grad_norm": 2.0134262227026625, "learning_rate": 9.143693731039504e-06, "loss": 0.6167, "step": 2493 }, { "epoch": 0.8113207547169812, "grad_norm": 1.7951526272281173, "learning_rate": 9.142634181649196e-06, "loss": 0.5692, "step": 2494 }, { "epoch": 0.8116460637605726, "grad_norm": 1.8485708996497134, "learning_rate": 9.141574038616648e-06, "loss": 0.6208, "step": 2495 }, { "epoch": 0.811971372804164, "grad_norm": 1.9231039162936447, "learning_rate": 9.14051330209378e-06, "loss": 0.6576, "step": 2496 }, { "epoch": 0.8122966818477554, "grad_norm": 2.0377752859864935, "learning_rate": 9.139451972232593e-06, "loss": 0.7048, "step": 2497 }, { "epoch": 0.8126219908913468, "grad_norm": 1.849383429170146, "learning_rate": 9.138390049185179e-06, "loss": 0.6148, "step": 2498 }, { "epoch": 0.8129472999349382, "grad_norm": 1.8705236485282122, "learning_rate": 9.137327533103711e-06, "loss": 0.6007, "step": 2499 }, { "epoch": 0.8132726089785296, "grad_norm": 1.6985738870429656, "learning_rate": 9.136264424140451e-06, "loss": 0.537, "step": 2500 }, { "epoch": 0.813597918022121, "grad_norm": 1.891725052614439, "learning_rate": 9.135200722447738e-06, "loss": 0.7046, "step": 2501 }, { "epoch": 0.8139232270657124, "grad_norm": 1.9632742528843483, "learning_rate": 9.134136428178007e-06, "loss": 0.6302, "step": 2502 }, { "epoch": 0.8142485361093038, "grad_norm": 1.976255519346558, "learning_rate": 9.13307154148377e-06, "loss": 0.6469, "step": 2503 }, { "epoch": 0.8145738451528952, "grad_norm": 1.5364256585746678, "learning_rate": 9.132006062517625e-06, "loss": 0.5756, "step": 2504 }, { "epoch": 0.8148991541964866, "grad_norm": 1.739478956662833, "learning_rate": 9.130939991432255e-06, "loss": 0.5695, "step": 2505 }, { "epoch": 0.815224463240078, "grad_norm": 1.7649525844432512, "learning_rate": 9.129873328380432e-06, "loss": 0.5302, "step": 2506 }, { "epoch": 0.8155497722836695, "grad_norm": 1.7683566110114126, "learning_rate": 9.128806073515005e-06, "loss": 0.5404, "step": 2507 }, { "epoch": 0.8158750813272609, "grad_norm": 1.7964848202368888, "learning_rate": 9.127738226988917e-06, "loss": 0.6749, "step": 2508 }, { "epoch": 0.8162003903708523, "grad_norm": 1.9201650223101363, "learning_rate": 9.126669788955188e-06, "loss": 0.5995, "step": 2509 }, { "epoch": 0.8165256994144438, "grad_norm": 1.5381958220885965, "learning_rate": 9.125600759566927e-06, "loss": 0.4901, "step": 2510 }, { "epoch": 0.8168510084580352, "grad_norm": 1.8065835644121193, "learning_rate": 9.124531138977326e-06, "loss": 0.6507, "step": 2511 }, { "epoch": 0.8171763175016266, "grad_norm": 1.7122174467070845, "learning_rate": 9.123460927339664e-06, "loss": 0.5004, "step": 2512 }, { "epoch": 0.817501626545218, "grad_norm": 1.7083356405626642, "learning_rate": 9.122390124807301e-06, "loss": 0.6289, "step": 2513 }, { "epoch": 0.8178269355888094, "grad_norm": 2.171299169807552, "learning_rate": 9.121318731533685e-06, "loss": 0.6262, "step": 2514 }, { "epoch": 0.8181522446324008, "grad_norm": 1.7528542981842856, "learning_rate": 9.120246747672347e-06, "loss": 0.5495, "step": 2515 }, { "epoch": 0.8184775536759922, "grad_norm": 1.6660916129342933, "learning_rate": 9.119174173376903e-06, "loss": 0.585, "step": 2516 }, { "epoch": 0.8188028627195836, "grad_norm": 1.9270765059609674, "learning_rate": 9.118101008801053e-06, "loss": 0.6034, "step": 2517 }, { "epoch": 0.819128171763175, "grad_norm": 1.7521641020833676, "learning_rate": 9.117027254098585e-06, "loss": 0.5088, "step": 2518 }, { "epoch": 0.8194534808067664, "grad_norm": 1.8295779756368202, "learning_rate": 9.115952909423364e-06, "loss": 0.5458, "step": 2519 }, { "epoch": 0.8197787898503578, "grad_norm": 1.8897108578777153, "learning_rate": 9.114877974929347e-06, "loss": 0.5671, "step": 2520 }, { "epoch": 0.8201040988939493, "grad_norm": 1.63189089240289, "learning_rate": 9.113802450770573e-06, "loss": 0.5281, "step": 2521 }, { "epoch": 0.8204294079375407, "grad_norm": 1.625120892989853, "learning_rate": 9.112726337101166e-06, "loss": 0.5151, "step": 2522 }, { "epoch": 0.8207547169811321, "grad_norm": 2.0072411684035787, "learning_rate": 9.111649634075332e-06, "loss": 0.7215, "step": 2523 }, { "epoch": 0.8210800260247235, "grad_norm": 1.9429608098410311, "learning_rate": 9.110572341847363e-06, "loss": 0.5383, "step": 2524 }, { "epoch": 0.8214053350683149, "grad_norm": 1.8166690117683162, "learning_rate": 9.10949446057164e-06, "loss": 0.6323, "step": 2525 }, { "epoch": 0.8217306441119063, "grad_norm": 1.7516378793719238, "learning_rate": 9.108415990402616e-06, "loss": 0.5477, "step": 2526 }, { "epoch": 0.8220559531554977, "grad_norm": 1.7277154100068817, "learning_rate": 9.107336931494846e-06, "loss": 0.5554, "step": 2527 }, { "epoch": 0.8223812621990891, "grad_norm": 1.7118583682026023, "learning_rate": 9.106257284002953e-06, "loss": 0.6449, "step": 2528 }, { "epoch": 0.8227065712426805, "grad_norm": 1.666369805888282, "learning_rate": 9.105177048081655e-06, "loss": 0.5017, "step": 2529 }, { "epoch": 0.8230318802862719, "grad_norm": 1.8820917146995755, "learning_rate": 9.104096223885748e-06, "loss": 0.6527, "step": 2530 }, { "epoch": 0.8233571893298633, "grad_norm": 1.8471918748471605, "learning_rate": 9.103014811570115e-06, "loss": 0.6262, "step": 2531 }, { "epoch": 0.8236824983734548, "grad_norm": 1.714972376854325, "learning_rate": 9.101932811289725e-06, "loss": 0.5626, "step": 2532 }, { "epoch": 0.8240078074170462, "grad_norm": 1.7960511270887904, "learning_rate": 9.100850223199627e-06, "loss": 0.6785, "step": 2533 }, { "epoch": 0.8243331164606376, "grad_norm": 1.6805835335277215, "learning_rate": 9.09976704745496e-06, "loss": 0.5742, "step": 2534 }, { "epoch": 0.824658425504229, "grad_norm": 1.9599083799936197, "learning_rate": 9.098683284210938e-06, "loss": 0.6396, "step": 2535 }, { "epoch": 0.8249837345478205, "grad_norm": 1.8455630784549288, "learning_rate": 9.097598933622873e-06, "loss": 0.6141, "step": 2536 }, { "epoch": 0.8253090435914119, "grad_norm": 1.9549213742054075, "learning_rate": 9.096513995846145e-06, "loss": 0.6343, "step": 2537 }, { "epoch": 0.8256343526350033, "grad_norm": 2.1203912734215002, "learning_rate": 9.095428471036232e-06, "loss": 0.5767, "step": 2538 }, { "epoch": 0.8259596616785947, "grad_norm": 1.6395544121274666, "learning_rate": 9.094342359348688e-06, "loss": 0.4993, "step": 2539 }, { "epoch": 0.8262849707221861, "grad_norm": 1.5680957690800006, "learning_rate": 9.093255660939155e-06, "loss": 0.5797, "step": 2540 }, { "epoch": 0.8266102797657775, "grad_norm": 1.6562813484586938, "learning_rate": 9.092168375963358e-06, "loss": 0.521, "step": 2541 }, { "epoch": 0.8269355888093689, "grad_norm": 1.81697506467263, "learning_rate": 9.091080504577102e-06, "loss": 0.6247, "step": 2542 }, { "epoch": 0.8272608978529603, "grad_norm": 1.6743487700338593, "learning_rate": 9.089992046936282e-06, "loss": 0.5894, "step": 2543 }, { "epoch": 0.8275862068965517, "grad_norm": 2.0349002224041515, "learning_rate": 9.088903003196875e-06, "loss": 0.6104, "step": 2544 }, { "epoch": 0.8279115159401431, "grad_norm": 1.564857010154496, "learning_rate": 9.087813373514942e-06, "loss": 0.5242, "step": 2545 }, { "epoch": 0.8282368249837345, "grad_norm": 1.984235959200739, "learning_rate": 9.08672315804663e-06, "loss": 0.6622, "step": 2546 }, { "epoch": 0.8285621340273259, "grad_norm": 1.649698365092323, "learning_rate": 9.08563235694816e-06, "loss": 0.5428, "step": 2547 }, { "epoch": 0.8288874430709173, "grad_norm": 1.8900489134201708, "learning_rate": 9.08454097037585e-06, "loss": 0.6452, "step": 2548 }, { "epoch": 0.8292127521145087, "grad_norm": 2.0850671650822865, "learning_rate": 9.083448998486097e-06, "loss": 0.5619, "step": 2549 }, { "epoch": 0.8295380611581002, "grad_norm": 1.7489436829690115, "learning_rate": 9.082356441435379e-06, "loss": 0.5917, "step": 2550 }, { "epoch": 0.8298633702016917, "grad_norm": 1.7175177467126634, "learning_rate": 9.081263299380262e-06, "loss": 0.5397, "step": 2551 }, { "epoch": 0.8301886792452831, "grad_norm": 2.1234745020823658, "learning_rate": 9.080169572477394e-06, "loss": 0.6503, "step": 2552 }, { "epoch": 0.8305139882888745, "grad_norm": 2.298544681299416, "learning_rate": 9.079075260883502e-06, "loss": 0.672, "step": 2553 }, { "epoch": 0.8308392973324659, "grad_norm": 1.8803304584057026, "learning_rate": 9.07798036475541e-06, "loss": 0.5162, "step": 2554 }, { "epoch": 0.8311646063760573, "grad_norm": 1.7512162889773688, "learning_rate": 9.07688488425001e-06, "loss": 0.5755, "step": 2555 }, { "epoch": 0.8314899154196487, "grad_norm": 1.7660544989363498, "learning_rate": 9.07578881952429e-06, "loss": 0.6583, "step": 2556 }, { "epoch": 0.8318152244632401, "grad_norm": 1.8846492986206065, "learning_rate": 9.074692170735313e-06, "loss": 0.5075, "step": 2557 }, { "epoch": 0.8321405335068315, "grad_norm": 1.8201766529940482, "learning_rate": 9.073594938040231e-06, "loss": 0.6326, "step": 2558 }, { "epoch": 0.8324658425504229, "grad_norm": 1.834009446739423, "learning_rate": 9.072497121596279e-06, "loss": 0.5515, "step": 2559 }, { "epoch": 0.8327911515940143, "grad_norm": 1.988702088495948, "learning_rate": 9.071398721560776e-06, "loss": 0.5466, "step": 2560 }, { "epoch": 0.8331164606376057, "grad_norm": 1.888898369268901, "learning_rate": 9.07029973809112e-06, "loss": 0.5582, "step": 2561 }, { "epoch": 0.8334417696811971, "grad_norm": 2.01920598305574, "learning_rate": 9.069200171344799e-06, "loss": 0.5675, "step": 2562 }, { "epoch": 0.8337670787247885, "grad_norm": 1.8413912196489481, "learning_rate": 9.068100021479379e-06, "loss": 0.6354, "step": 2563 }, { "epoch": 0.8340923877683799, "grad_norm": 1.9639049532903343, "learning_rate": 9.066999288652513e-06, "loss": 0.5745, "step": 2564 }, { "epoch": 0.8344176968119714, "grad_norm": 1.6531082259913232, "learning_rate": 9.065897973021938e-06, "loss": 0.4715, "step": 2565 }, { "epoch": 0.8347430058555628, "grad_norm": 1.8796968814746904, "learning_rate": 9.064796074745473e-06, "loss": 0.6765, "step": 2566 }, { "epoch": 0.8350683148991542, "grad_norm": 2.0266002462720074, "learning_rate": 9.06369359398102e-06, "loss": 0.5601, "step": 2567 }, { "epoch": 0.8353936239427456, "grad_norm": 1.6110581623338078, "learning_rate": 9.062590530886566e-06, "loss": 0.5531, "step": 2568 }, { "epoch": 0.835718932986337, "grad_norm": 1.5320707342871869, "learning_rate": 9.06148688562018e-06, "loss": 0.5147, "step": 2569 }, { "epoch": 0.8360442420299284, "grad_norm": 2.098358865335974, "learning_rate": 9.060382658340015e-06, "loss": 0.6357, "step": 2570 }, { "epoch": 0.8363695510735198, "grad_norm": 1.877230854139004, "learning_rate": 9.059277849204308e-06, "loss": 0.623, "step": 2571 }, { "epoch": 0.8366948601171112, "grad_norm": 1.8373876833896254, "learning_rate": 9.058172458371379e-06, "loss": 0.5606, "step": 2572 }, { "epoch": 0.8370201691607027, "grad_norm": 2.049994477185124, "learning_rate": 9.057066485999633e-06, "loss": 0.6778, "step": 2573 }, { "epoch": 0.8373454782042941, "grad_norm": 1.5502800790106055, "learning_rate": 9.05595993224755e-06, "loss": 0.5028, "step": 2574 }, { "epoch": 0.8376707872478855, "grad_norm": 2.0768153327636503, "learning_rate": 9.05485279727371e-06, "loss": 0.6274, "step": 2575 }, { "epoch": 0.8379960962914769, "grad_norm": 1.9415122965823726, "learning_rate": 9.053745081236756e-06, "loss": 0.6815, "step": 2576 }, { "epoch": 0.8383214053350683, "grad_norm": 1.9170833853396687, "learning_rate": 9.052636784295432e-06, "loss": 0.6369, "step": 2577 }, { "epoch": 0.8386467143786597, "grad_norm": 1.7310504830154012, "learning_rate": 9.051527906608552e-06, "loss": 0.5976, "step": 2578 }, { "epoch": 0.8389720234222512, "grad_norm": 1.554465596234883, "learning_rate": 9.050418448335025e-06, "loss": 0.5495, "step": 2579 }, { "epoch": 0.8392973324658426, "grad_norm": 1.4914403074899154, "learning_rate": 9.049308409633832e-06, "loss": 0.4324, "step": 2580 }, { "epoch": 0.839622641509434, "grad_norm": 1.4802509697865633, "learning_rate": 9.048197790664044e-06, "loss": 0.4918, "step": 2581 }, { "epoch": 0.8399479505530254, "grad_norm": 1.8314532412630382, "learning_rate": 9.047086591584814e-06, "loss": 0.5477, "step": 2582 }, { "epoch": 0.8402732595966168, "grad_norm": 1.8129118715528192, "learning_rate": 9.045974812555377e-06, "loss": 0.6479, "step": 2583 }, { "epoch": 0.8405985686402082, "grad_norm": 1.8154376235557945, "learning_rate": 9.044862453735051e-06, "loss": 0.5247, "step": 2584 }, { "epoch": 0.8409238776837996, "grad_norm": 1.712573079370915, "learning_rate": 9.04374951528324e-06, "loss": 0.5363, "step": 2585 }, { "epoch": 0.841249186727391, "grad_norm": 1.9727333366400839, "learning_rate": 9.042635997359428e-06, "loss": 0.5809, "step": 2586 }, { "epoch": 0.8415744957709824, "grad_norm": 1.8165699040153735, "learning_rate": 9.04152190012318e-06, "loss": 0.6516, "step": 2587 }, { "epoch": 0.8418998048145738, "grad_norm": 1.9688075029505476, "learning_rate": 9.04040722373415e-06, "loss": 0.5875, "step": 2588 }, { "epoch": 0.8422251138581652, "grad_norm": 1.9160094250275572, "learning_rate": 9.03929196835207e-06, "loss": 0.5804, "step": 2589 }, { "epoch": 0.8425504229017566, "grad_norm": 1.9460461574104426, "learning_rate": 9.03817613413676e-06, "loss": 0.6142, "step": 2590 }, { "epoch": 0.842875731945348, "grad_norm": 1.8450135374589482, "learning_rate": 9.037059721248116e-06, "loss": 0.5771, "step": 2591 }, { "epoch": 0.8432010409889394, "grad_norm": 1.7158920562217668, "learning_rate": 9.035942729846123e-06, "loss": 0.4991, "step": 2592 }, { "epoch": 0.843526350032531, "grad_norm": 1.5743561944263353, "learning_rate": 9.034825160090845e-06, "loss": 0.4673, "step": 2593 }, { "epoch": 0.8438516590761224, "grad_norm": 1.9109045589868012, "learning_rate": 9.033707012142433e-06, "loss": 0.5419, "step": 2594 }, { "epoch": 0.8441769681197138, "grad_norm": 1.8635598738059955, "learning_rate": 9.032588286161116e-06, "loss": 0.6098, "step": 2595 }, { "epoch": 0.8445022771633052, "grad_norm": 1.9180232312854184, "learning_rate": 9.031468982307208e-06, "loss": 0.5972, "step": 2596 }, { "epoch": 0.8448275862068966, "grad_norm": 1.8421987049285202, "learning_rate": 9.03034910074111e-06, "loss": 0.6252, "step": 2597 }, { "epoch": 0.845152895250488, "grad_norm": 1.4880888778242936, "learning_rate": 9.029228641623295e-06, "loss": 0.5223, "step": 2598 }, { "epoch": 0.8454782042940794, "grad_norm": 2.03415245499072, "learning_rate": 9.028107605114332e-06, "loss": 0.6722, "step": 2599 }, { "epoch": 0.8458035133376708, "grad_norm": 1.8836438472388246, "learning_rate": 9.026985991374862e-06, "loss": 0.5313, "step": 2600 }, { "epoch": 0.8461288223812622, "grad_norm": 1.7632532303614747, "learning_rate": 9.025863800565614e-06, "loss": 0.6642, "step": 2601 }, { "epoch": 0.8464541314248536, "grad_norm": 1.7918944248331472, "learning_rate": 9.0247410328474e-06, "loss": 0.576, "step": 2602 }, { "epoch": 0.846779440468445, "grad_norm": 2.128261420670822, "learning_rate": 9.023617688381112e-06, "loss": 0.666, "step": 2603 }, { "epoch": 0.8471047495120364, "grad_norm": 1.6528179808610912, "learning_rate": 9.022493767327728e-06, "loss": 0.5332, "step": 2604 }, { "epoch": 0.8474300585556278, "grad_norm": 1.824314420948906, "learning_rate": 9.021369269848303e-06, "loss": 0.5948, "step": 2605 }, { "epoch": 0.8477553675992192, "grad_norm": 2.2195425823248653, "learning_rate": 9.020244196103983e-06, "loss": 0.7196, "step": 2606 }, { "epoch": 0.8480806766428106, "grad_norm": 1.8609720824471752, "learning_rate": 9.01911854625599e-06, "loss": 0.5726, "step": 2607 }, { "epoch": 0.8484059856864021, "grad_norm": 1.938581035470728, "learning_rate": 9.017992320465626e-06, "loss": 0.5933, "step": 2608 }, { "epoch": 0.8487312947299935, "grad_norm": 1.8431360679475206, "learning_rate": 9.016865518894287e-06, "loss": 0.5669, "step": 2609 }, { "epoch": 0.8490566037735849, "grad_norm": 1.7914886511451127, "learning_rate": 9.015738141703438e-06, "loss": 0.5729, "step": 2610 }, { "epoch": 0.8493819128171763, "grad_norm": 1.778257891062959, "learning_rate": 9.014610189054637e-06, "loss": 0.5901, "step": 2611 }, { "epoch": 0.8497072218607677, "grad_norm": 1.9029171705587662, "learning_rate": 9.01348166110952e-06, "loss": 0.5741, "step": 2612 }, { "epoch": 0.8500325309043592, "grad_norm": 1.9214202314181705, "learning_rate": 9.012352558029804e-06, "loss": 0.5403, "step": 2613 }, { "epoch": 0.8503578399479506, "grad_norm": 1.9407356809168723, "learning_rate": 9.011222879977292e-06, "loss": 0.5401, "step": 2614 }, { "epoch": 0.850683148991542, "grad_norm": 1.870709889072192, "learning_rate": 9.010092627113867e-06, "loss": 0.6034, "step": 2615 }, { "epoch": 0.8510084580351334, "grad_norm": 1.794938405739105, "learning_rate": 9.008961799601494e-06, "loss": 0.5912, "step": 2616 }, { "epoch": 0.8513337670787248, "grad_norm": 1.829235704374066, "learning_rate": 9.007830397602225e-06, "loss": 0.7549, "step": 2617 }, { "epoch": 0.8516590761223162, "grad_norm": 2.011700980092887, "learning_rate": 9.006698421278183e-06, "loss": 0.6618, "step": 2618 }, { "epoch": 0.8519843851659076, "grad_norm": 1.7822338741275046, "learning_rate": 9.005565870791591e-06, "loss": 0.5807, "step": 2619 }, { "epoch": 0.852309694209499, "grad_norm": 1.5967744125272691, "learning_rate": 9.004432746304737e-06, "loss": 0.4565, "step": 2620 }, { "epoch": 0.8526350032530904, "grad_norm": 1.6359663318312394, "learning_rate": 9.003299047979998e-06, "loss": 0.5693, "step": 2621 }, { "epoch": 0.8529603122966819, "grad_norm": 1.6670692108358376, "learning_rate": 9.00216477597984e-06, "loss": 0.6309, "step": 2622 }, { "epoch": 0.8532856213402733, "grad_norm": 1.8875892464937696, "learning_rate": 9.001029930466799e-06, "loss": 0.5231, "step": 2623 }, { "epoch": 0.8536109303838647, "grad_norm": 1.9339554687467264, "learning_rate": 8.999894511603503e-06, "loss": 0.5036, "step": 2624 }, { "epoch": 0.8539362394274561, "grad_norm": 1.7401275382855108, "learning_rate": 8.998758519552655e-06, "loss": 0.6701, "step": 2625 }, { "epoch": 0.8542615484710475, "grad_norm": 1.8730085017768474, "learning_rate": 8.997621954477046e-06, "loss": 0.6502, "step": 2626 }, { "epoch": 0.8545868575146389, "grad_norm": 1.8891743175341353, "learning_rate": 8.996484816539547e-06, "loss": 0.6573, "step": 2627 }, { "epoch": 0.8549121665582303, "grad_norm": 1.7914014140811052, "learning_rate": 8.995347105903107e-06, "loss": 0.5336, "step": 2628 }, { "epoch": 0.8552374756018217, "grad_norm": 1.9619316128977697, "learning_rate": 8.994208822730764e-06, "loss": 0.644, "step": 2629 }, { "epoch": 0.8555627846454131, "grad_norm": 1.8947525246436014, "learning_rate": 8.993069967185632e-06, "loss": 0.5766, "step": 2630 }, { "epoch": 0.8558880936890045, "grad_norm": 1.6840892764080115, "learning_rate": 8.991930539430913e-06, "loss": 0.5469, "step": 2631 }, { "epoch": 0.8562134027325959, "grad_norm": 2.103363886182933, "learning_rate": 8.990790539629886e-06, "loss": 0.6221, "step": 2632 }, { "epoch": 0.8565387117761873, "grad_norm": 1.8849605011853927, "learning_rate": 8.989649967945913e-06, "loss": 0.6409, "step": 2633 }, { "epoch": 0.8568640208197787, "grad_norm": 1.778886399086077, "learning_rate": 8.98850882454244e-06, "loss": 0.6392, "step": 2634 }, { "epoch": 0.8571893298633702, "grad_norm": 2.1664600803372833, "learning_rate": 8.987367109582995e-06, "loss": 0.7166, "step": 2635 }, { "epoch": 0.8575146389069617, "grad_norm": 1.9111765977976027, "learning_rate": 8.986224823231182e-06, "loss": 0.6127, "step": 2636 }, { "epoch": 0.8578399479505531, "grad_norm": 2.0234486927403244, "learning_rate": 8.985081965650695e-06, "loss": 0.7595, "step": 2637 }, { "epoch": 0.8581652569941445, "grad_norm": 1.8251010365289475, "learning_rate": 8.983938537005305e-06, "loss": 0.6307, "step": 2638 }, { "epoch": 0.8584905660377359, "grad_norm": 1.727111715235697, "learning_rate": 8.982794537458868e-06, "loss": 0.5228, "step": 2639 }, { "epoch": 0.8588158750813273, "grad_norm": 1.6513364104995796, "learning_rate": 8.981649967175317e-06, "loss": 0.6134, "step": 2640 }, { "epoch": 0.8591411841249187, "grad_norm": 1.6583546814799042, "learning_rate": 8.980504826318673e-06, "loss": 0.609, "step": 2641 }, { "epoch": 0.8594664931685101, "grad_norm": 2.0445531961435846, "learning_rate": 8.97935911505303e-06, "loss": 0.739, "step": 2642 }, { "epoch": 0.8597918022121015, "grad_norm": 1.6521651134797586, "learning_rate": 8.978212833542575e-06, "loss": 0.5504, "step": 2643 }, { "epoch": 0.8601171112556929, "grad_norm": 1.5945211291667563, "learning_rate": 8.977065981951567e-06, "loss": 0.5035, "step": 2644 }, { "epoch": 0.8604424202992843, "grad_norm": 1.7290743352948086, "learning_rate": 8.975918560444355e-06, "loss": 0.6125, "step": 2645 }, { "epoch": 0.8607677293428757, "grad_norm": 1.8058212534811842, "learning_rate": 8.97477056918536e-06, "loss": 0.7312, "step": 2646 }, { "epoch": 0.8610930383864671, "grad_norm": 1.7372545328776658, "learning_rate": 8.973622008339094e-06, "loss": 0.5824, "step": 2647 }, { "epoch": 0.8614183474300585, "grad_norm": 1.976991868483562, "learning_rate": 8.972472878070141e-06, "loss": 0.6722, "step": 2648 }, { "epoch": 0.8617436564736499, "grad_norm": 1.79359526662008, "learning_rate": 8.971323178543181e-06, "loss": 0.6028, "step": 2649 }, { "epoch": 0.8620689655172413, "grad_norm": 1.6413072590769215, "learning_rate": 8.970172909922958e-06, "loss": 0.5432, "step": 2650 }, { "epoch": 0.8623942745608328, "grad_norm": 1.6879166230542206, "learning_rate": 8.969022072374312e-06, "loss": 0.5306, "step": 2651 }, { "epoch": 0.8627195836044242, "grad_norm": 1.9634703732331824, "learning_rate": 8.967870666062158e-06, "loss": 0.5554, "step": 2652 }, { "epoch": 0.8630448926480156, "grad_norm": 1.727843637903657, "learning_rate": 8.96671869115149e-06, "loss": 0.4954, "step": 2653 }, { "epoch": 0.863370201691607, "grad_norm": 1.8476261500861022, "learning_rate": 8.965566147807388e-06, "loss": 0.5911, "step": 2654 }, { "epoch": 0.8636955107351985, "grad_norm": 1.7976902221597428, "learning_rate": 8.964413036195014e-06, "loss": 0.5617, "step": 2655 }, { "epoch": 0.8640208197787899, "grad_norm": 1.858159096756848, "learning_rate": 8.96325935647961e-06, "loss": 0.6964, "step": 2656 }, { "epoch": 0.8643461288223813, "grad_norm": 1.9086853919161968, "learning_rate": 8.962105108826497e-06, "loss": 0.6363, "step": 2657 }, { "epoch": 0.8646714378659727, "grad_norm": 1.9085205545950994, "learning_rate": 8.960950293401079e-06, "loss": 0.5662, "step": 2658 }, { "epoch": 0.8649967469095641, "grad_norm": 1.5948004694473543, "learning_rate": 8.959794910368844e-06, "loss": 0.5323, "step": 2659 }, { "epoch": 0.8653220559531555, "grad_norm": 1.9940702241573323, "learning_rate": 8.95863895989536e-06, "loss": 0.6117, "step": 2660 }, { "epoch": 0.8656473649967469, "grad_norm": 1.779471381155761, "learning_rate": 8.957482442146271e-06, "loss": 0.5091, "step": 2661 }, { "epoch": 0.8659726740403383, "grad_norm": 1.9101170350409757, "learning_rate": 8.95632535728731e-06, "loss": 0.6342, "step": 2662 }, { "epoch": 0.8662979830839297, "grad_norm": 1.9017940730781315, "learning_rate": 8.955167705484289e-06, "loss": 0.5671, "step": 2663 }, { "epoch": 0.8666232921275211, "grad_norm": 1.7562805087355222, "learning_rate": 8.954009486903097e-06, "loss": 0.5191, "step": 2664 }, { "epoch": 0.8669486011711126, "grad_norm": 2.018462550509543, "learning_rate": 8.95285070170971e-06, "loss": 0.7382, "step": 2665 }, { "epoch": 0.867273910214704, "grad_norm": 1.642650202802664, "learning_rate": 8.951691350070182e-06, "loss": 0.5299, "step": 2666 }, { "epoch": 0.8675992192582954, "grad_norm": 1.814383592556127, "learning_rate": 8.95053143215065e-06, "loss": 0.653, "step": 2667 }, { "epoch": 0.8679245283018868, "grad_norm": 1.8237278607262004, "learning_rate": 8.949370948117328e-06, "loss": 0.6189, "step": 2668 }, { "epoch": 0.8682498373454782, "grad_norm": 1.643348686475121, "learning_rate": 8.948209898136517e-06, "loss": 0.5693, "step": 2669 }, { "epoch": 0.8685751463890696, "grad_norm": 1.7892743044145372, "learning_rate": 8.947048282374595e-06, "loss": 0.5577, "step": 2670 }, { "epoch": 0.868900455432661, "grad_norm": 1.7247295363208737, "learning_rate": 8.945886100998022e-06, "loss": 0.5823, "step": 2671 }, { "epoch": 0.8692257644762524, "grad_norm": 1.897339110013074, "learning_rate": 8.944723354173341e-06, "loss": 0.521, "step": 2672 }, { "epoch": 0.8695510735198438, "grad_norm": 1.964216611881327, "learning_rate": 8.943560042067172e-06, "loss": 0.5666, "step": 2673 }, { "epoch": 0.8698763825634352, "grad_norm": 1.7032812403416984, "learning_rate": 8.942396164846222e-06, "loss": 0.4844, "step": 2674 }, { "epoch": 0.8702016916070267, "grad_norm": 1.902929351462325, "learning_rate": 8.94123172267727e-06, "loss": 0.6586, "step": 2675 }, { "epoch": 0.870527000650618, "grad_norm": 1.8258010705190988, "learning_rate": 8.940066715727185e-06, "loss": 0.6415, "step": 2676 }, { "epoch": 0.8708523096942095, "grad_norm": 1.8353346486816924, "learning_rate": 8.938901144162913e-06, "loss": 0.5807, "step": 2677 }, { "epoch": 0.8711776187378009, "grad_norm": 1.7373402697579885, "learning_rate": 8.937735008151482e-06, "loss": 0.5913, "step": 2678 }, { "epoch": 0.8715029277813924, "grad_norm": 1.769499055417518, "learning_rate": 8.936568307859999e-06, "loss": 0.6165, "step": 2679 }, { "epoch": 0.8718282368249838, "grad_norm": 1.9436745724824513, "learning_rate": 8.935401043455652e-06, "loss": 0.7117, "step": 2680 }, { "epoch": 0.8721535458685752, "grad_norm": 1.86414680608341, "learning_rate": 8.934233215105711e-06, "loss": 0.6511, "step": 2681 }, { "epoch": 0.8724788549121666, "grad_norm": 2.054839033031337, "learning_rate": 8.933064822977529e-06, "loss": 0.6901, "step": 2682 }, { "epoch": 0.872804163955758, "grad_norm": 1.8169435445255304, "learning_rate": 8.931895867238533e-06, "loss": 0.6566, "step": 2683 }, { "epoch": 0.8731294729993494, "grad_norm": 1.7820232360083277, "learning_rate": 8.93072634805624e-06, "loss": 0.602, "step": 2684 }, { "epoch": 0.8734547820429408, "grad_norm": 1.9635715574290147, "learning_rate": 8.929556265598239e-06, "loss": 0.6271, "step": 2685 }, { "epoch": 0.8737800910865322, "grad_norm": 1.6065835824264603, "learning_rate": 8.928385620032207e-06, "loss": 0.5433, "step": 2686 }, { "epoch": 0.8741054001301236, "grad_norm": 1.6041411089951132, "learning_rate": 8.927214411525895e-06, "loss": 0.5182, "step": 2687 }, { "epoch": 0.874430709173715, "grad_norm": 1.4647648144984737, "learning_rate": 8.92604264024714e-06, "loss": 0.5088, "step": 2688 }, { "epoch": 0.8747560182173064, "grad_norm": 1.7762349550223866, "learning_rate": 8.924870306363856e-06, "loss": 0.6146, "step": 2689 }, { "epoch": 0.8750813272608978, "grad_norm": 1.6350844326250236, "learning_rate": 8.923697410044043e-06, "loss": 0.5624, "step": 2690 }, { "epoch": 0.8754066363044892, "grad_norm": 2.027690327857465, "learning_rate": 8.922523951455773e-06, "loss": 0.6247, "step": 2691 }, { "epoch": 0.8757319453480806, "grad_norm": 1.8162255688955278, "learning_rate": 8.921349930767205e-06, "loss": 0.6941, "step": 2692 }, { "epoch": 0.876057254391672, "grad_norm": 2.048045897271327, "learning_rate": 8.920175348146579e-06, "loss": 0.5529, "step": 2693 }, { "epoch": 0.8763825634352636, "grad_norm": 1.8858183191341273, "learning_rate": 8.91900020376221e-06, "loss": 0.6209, "step": 2694 }, { "epoch": 0.876707872478855, "grad_norm": 1.8877790684652802, "learning_rate": 8.9178244977825e-06, "loss": 0.6354, "step": 2695 }, { "epoch": 0.8770331815224464, "grad_norm": 1.6268083827061832, "learning_rate": 8.916648230375928e-06, "loss": 0.5374, "step": 2696 }, { "epoch": 0.8773584905660378, "grad_norm": 2.0250039742500126, "learning_rate": 8.915471401711052e-06, "loss": 0.5692, "step": 2697 }, { "epoch": 0.8776837996096292, "grad_norm": 1.7064404172221632, "learning_rate": 8.914294011956514e-06, "loss": 0.5981, "step": 2698 }, { "epoch": 0.8780091086532206, "grad_norm": 2.04419951355597, "learning_rate": 8.913116061281034e-06, "loss": 0.7038, "step": 2699 }, { "epoch": 0.878334417696812, "grad_norm": 1.7326187267118067, "learning_rate": 8.911937549853416e-06, "loss": 0.5318, "step": 2700 }, { "epoch": 0.8786597267404034, "grad_norm": 1.6656676073799095, "learning_rate": 8.910758477842536e-06, "loss": 0.5961, "step": 2701 }, { "epoch": 0.8789850357839948, "grad_norm": 1.9602516663007001, "learning_rate": 8.909578845417359e-06, "loss": 0.6771, "step": 2702 }, { "epoch": 0.8793103448275862, "grad_norm": 1.8647785052727586, "learning_rate": 8.908398652746929e-06, "loss": 0.5206, "step": 2703 }, { "epoch": 0.8796356538711776, "grad_norm": 1.9217529667295963, "learning_rate": 8.907217900000364e-06, "loss": 0.6926, "step": 2704 }, { "epoch": 0.879960962914769, "grad_norm": 1.8140304658353195, "learning_rate": 8.906036587346868e-06, "loss": 0.6002, "step": 2705 }, { "epoch": 0.8802862719583604, "grad_norm": 1.7544075090177835, "learning_rate": 8.904854714955726e-06, "loss": 0.5853, "step": 2706 }, { "epoch": 0.8806115810019518, "grad_norm": 2.000520734897661, "learning_rate": 8.9036722829963e-06, "loss": 0.6738, "step": 2707 }, { "epoch": 0.8809368900455433, "grad_norm": 1.665204099284549, "learning_rate": 8.902489291638034e-06, "loss": 0.562, "step": 2708 }, { "epoch": 0.8812621990891347, "grad_norm": 1.7266245188710918, "learning_rate": 8.90130574105045e-06, "loss": 0.5531, "step": 2709 }, { "epoch": 0.8815875081327261, "grad_norm": 1.769609321373508, "learning_rate": 8.900121631403152e-06, "loss": 0.6857, "step": 2710 }, { "epoch": 0.8819128171763175, "grad_norm": 1.7973522574610055, "learning_rate": 8.898936962865823e-06, "loss": 0.6255, "step": 2711 }, { "epoch": 0.8822381262199089, "grad_norm": 1.7378826146207758, "learning_rate": 8.89775173560823e-06, "loss": 0.5912, "step": 2712 }, { "epoch": 0.8825634352635003, "grad_norm": 1.846465280655399, "learning_rate": 8.896565949800214e-06, "loss": 0.5709, "step": 2713 }, { "epoch": 0.8828887443070917, "grad_norm": 1.8936909447972887, "learning_rate": 8.895379605611698e-06, "loss": 0.6293, "step": 2714 }, { "epoch": 0.8832140533506831, "grad_norm": 1.9344233526811714, "learning_rate": 8.894192703212687e-06, "loss": 0.5381, "step": 2715 }, { "epoch": 0.8835393623942746, "grad_norm": 1.7119865021313283, "learning_rate": 8.893005242773268e-06, "loss": 0.6272, "step": 2716 }, { "epoch": 0.883864671437866, "grad_norm": 1.7735652105924122, "learning_rate": 8.891817224463599e-06, "loss": 0.5491, "step": 2717 }, { "epoch": 0.8841899804814574, "grad_norm": 1.8131063028013517, "learning_rate": 8.890628648453927e-06, "loss": 0.5904, "step": 2718 }, { "epoch": 0.8845152895250488, "grad_norm": 1.7233809409892673, "learning_rate": 8.889439514914578e-06, "loss": 0.5547, "step": 2719 }, { "epoch": 0.8848405985686402, "grad_norm": 2.155066173600221, "learning_rate": 8.888249824015952e-06, "loss": 0.631, "step": 2720 }, { "epoch": 0.8851659076122316, "grad_norm": 1.934206989755426, "learning_rate": 8.887059575928534e-06, "loss": 0.6408, "step": 2721 }, { "epoch": 0.885491216655823, "grad_norm": 1.873135611884096, "learning_rate": 8.885868770822887e-06, "loss": 0.6807, "step": 2722 }, { "epoch": 0.8858165256994145, "grad_norm": 1.948613056808307, "learning_rate": 8.884677408869655e-06, "loss": 0.6764, "step": 2723 }, { "epoch": 0.8861418347430059, "grad_norm": 2.011996139999059, "learning_rate": 8.883485490239559e-06, "loss": 0.6671, "step": 2724 }, { "epoch": 0.8864671437865973, "grad_norm": 1.732374330345784, "learning_rate": 8.882293015103404e-06, "loss": 0.6039, "step": 2725 }, { "epoch": 0.8867924528301887, "grad_norm": 1.7038713174402114, "learning_rate": 8.881099983632071e-06, "loss": 0.6916, "step": 2726 }, { "epoch": 0.8871177618737801, "grad_norm": 1.9923607076203058, "learning_rate": 8.879906395996523e-06, "loss": 0.7289, "step": 2727 }, { "epoch": 0.8874430709173715, "grad_norm": 1.8569067317190828, "learning_rate": 8.878712252367803e-06, "loss": 0.6422, "step": 2728 }, { "epoch": 0.8877683799609629, "grad_norm": 1.6425105869619627, "learning_rate": 8.87751755291703e-06, "loss": 0.5426, "step": 2729 }, { "epoch": 0.8880936890045543, "grad_norm": 1.7052731128410985, "learning_rate": 8.876322297815406e-06, "loss": 0.5588, "step": 2730 }, { "epoch": 0.8884189980481457, "grad_norm": 1.7925596467549088, "learning_rate": 8.87512648723421e-06, "loss": 0.6411, "step": 2731 }, { "epoch": 0.8887443070917371, "grad_norm": 1.8393809856708336, "learning_rate": 8.873930121344809e-06, "loss": 0.6176, "step": 2732 }, { "epoch": 0.8890696161353285, "grad_norm": 1.7091558390860095, "learning_rate": 8.872733200318638e-06, "loss": 0.5776, "step": 2733 }, { "epoch": 0.8893949251789199, "grad_norm": 1.84175560888096, "learning_rate": 8.871535724327217e-06, "loss": 0.6524, "step": 2734 }, { "epoch": 0.8897202342225113, "grad_norm": 1.6888499633833158, "learning_rate": 8.870337693542144e-06, "loss": 0.544, "step": 2735 }, { "epoch": 0.8900455432661027, "grad_norm": 1.6992087194234333, "learning_rate": 8.869139108135099e-06, "loss": 0.5973, "step": 2736 }, { "epoch": 0.8903708523096943, "grad_norm": 1.8246138152074058, "learning_rate": 8.867939968277842e-06, "loss": 0.702, "step": 2737 }, { "epoch": 0.8906961613532857, "grad_norm": 1.682296270165466, "learning_rate": 8.866740274142207e-06, "loss": 0.6098, "step": 2738 }, { "epoch": 0.8910214703968771, "grad_norm": 1.695890992703635, "learning_rate": 8.865540025900112e-06, "loss": 0.5242, "step": 2739 }, { "epoch": 0.8913467794404685, "grad_norm": 1.6779406002905615, "learning_rate": 8.864339223723556e-06, "loss": 0.52, "step": 2740 }, { "epoch": 0.8916720884840599, "grad_norm": 1.9904893117532634, "learning_rate": 8.863137867784612e-06, "loss": 0.5859, "step": 2741 }, { "epoch": 0.8919973975276513, "grad_norm": 1.8999369376980997, "learning_rate": 8.861935958255436e-06, "loss": 0.615, "step": 2742 }, { "epoch": 0.8923227065712427, "grad_norm": 1.7903744707847498, "learning_rate": 8.860733495308262e-06, "loss": 0.6057, "step": 2743 }, { "epoch": 0.8926480156148341, "grad_norm": 1.6238237517942669, "learning_rate": 8.859530479115404e-06, "loss": 0.559, "step": 2744 }, { "epoch": 0.8929733246584255, "grad_norm": 1.812155577246766, "learning_rate": 8.858326909849255e-06, "loss": 0.5713, "step": 2745 }, { "epoch": 0.8932986337020169, "grad_norm": 1.9740353283243275, "learning_rate": 8.85712278768229e-06, "loss": 0.6667, "step": 2746 }, { "epoch": 0.8936239427456083, "grad_norm": 1.8133667164771745, "learning_rate": 8.855918112787054e-06, "loss": 0.6216, "step": 2747 }, { "epoch": 0.8939492517891997, "grad_norm": 1.720310243079733, "learning_rate": 8.854712885336183e-06, "loss": 0.5459, "step": 2748 }, { "epoch": 0.8942745608327911, "grad_norm": 1.8103244088864845, "learning_rate": 8.853507105502385e-06, "loss": 0.619, "step": 2749 }, { "epoch": 0.8945998698763825, "grad_norm": 1.9297629171124921, "learning_rate": 8.85230077345845e-06, "loss": 0.6586, "step": 2750 }, { "epoch": 0.894925178919974, "grad_norm": 1.4607642850688167, "learning_rate": 8.851093889377247e-06, "loss": 0.4757, "step": 2751 }, { "epoch": 0.8952504879635654, "grad_norm": 2.1075546533458978, "learning_rate": 8.84988645343172e-06, "loss": 0.6242, "step": 2752 }, { "epoch": 0.8955757970071568, "grad_norm": 1.836542846643849, "learning_rate": 8.848678465794899e-06, "loss": 0.5716, "step": 2753 }, { "epoch": 0.8959011060507482, "grad_norm": 2.207894307900039, "learning_rate": 8.847469926639887e-06, "loss": 0.6125, "step": 2754 }, { "epoch": 0.8962264150943396, "grad_norm": 1.5896338252386129, "learning_rate": 8.846260836139869e-06, "loss": 0.6039, "step": 2755 }, { "epoch": 0.896551724137931, "grad_norm": 1.9492735499441811, "learning_rate": 8.84505119446811e-06, "loss": 0.572, "step": 2756 }, { "epoch": 0.8968770331815225, "grad_norm": 1.678326558790322, "learning_rate": 8.84384100179795e-06, "loss": 0.5877, "step": 2757 }, { "epoch": 0.8972023422251139, "grad_norm": 1.8026508886515844, "learning_rate": 8.842630258302812e-06, "loss": 0.6878, "step": 2758 }, { "epoch": 0.8975276512687053, "grad_norm": 1.8742287258305883, "learning_rate": 8.841418964156197e-06, "loss": 0.555, "step": 2759 }, { "epoch": 0.8978529603122967, "grad_norm": 1.7139664176996916, "learning_rate": 8.840207119531681e-06, "loss": 0.6465, "step": 2760 }, { "epoch": 0.8981782693558881, "grad_norm": 1.8538362361678968, "learning_rate": 8.838994724602925e-06, "loss": 0.5886, "step": 2761 }, { "epoch": 0.8985035783994795, "grad_norm": 1.878913740510167, "learning_rate": 8.837781779543668e-06, "loss": 0.7336, "step": 2762 }, { "epoch": 0.8988288874430709, "grad_norm": 1.9758632938193543, "learning_rate": 8.83656828452772e-06, "loss": 0.642, "step": 2763 }, { "epoch": 0.8991541964866623, "grad_norm": 1.578025699231738, "learning_rate": 8.83535423972898e-06, "loss": 0.5699, "step": 2764 }, { "epoch": 0.8994795055302537, "grad_norm": 1.7461935664780566, "learning_rate": 8.83413964532142e-06, "loss": 0.6158, "step": 2765 }, { "epoch": 0.8998048145738452, "grad_norm": 1.7048849454235, "learning_rate": 8.832924501479094e-06, "loss": 0.5111, "step": 2766 }, { "epoch": 0.9001301236174366, "grad_norm": 1.7027742558450547, "learning_rate": 8.831708808376131e-06, "loss": 0.4781, "step": 2767 }, { "epoch": 0.900455432661028, "grad_norm": 1.897356069145964, "learning_rate": 8.830492566186738e-06, "loss": 0.6063, "step": 2768 }, { "epoch": 0.9007807417046194, "grad_norm": 1.6838831470184128, "learning_rate": 8.829275775085212e-06, "loss": 0.5398, "step": 2769 }, { "epoch": 0.9011060507482108, "grad_norm": 1.5266700580561505, "learning_rate": 8.82805843524591e-06, "loss": 0.5638, "step": 2770 }, { "epoch": 0.9014313597918022, "grad_norm": 2.0041049615070463, "learning_rate": 8.826840546843283e-06, "loss": 0.627, "step": 2771 }, { "epoch": 0.9017566688353936, "grad_norm": 1.6545590953633842, "learning_rate": 8.825622110051854e-06, "loss": 0.5421, "step": 2772 }, { "epoch": 0.902081977878985, "grad_norm": 1.7459865745204033, "learning_rate": 8.824403125046225e-06, "loss": 0.6299, "step": 2773 }, { "epoch": 0.9024072869225764, "grad_norm": 1.9564018913684735, "learning_rate": 8.823183592001082e-06, "loss": 0.6324, "step": 2774 }, { "epoch": 0.9027325959661678, "grad_norm": 1.6822025498910242, "learning_rate": 8.821963511091178e-06, "loss": 0.596, "step": 2775 }, { "epoch": 0.9030579050097592, "grad_norm": 1.852749580309484, "learning_rate": 8.820742882491354e-06, "loss": 0.6498, "step": 2776 }, { "epoch": 0.9033832140533506, "grad_norm": 1.635351423592114, "learning_rate": 8.81952170637653e-06, "loss": 0.6516, "step": 2777 }, { "epoch": 0.903708523096942, "grad_norm": 1.6661986614780508, "learning_rate": 8.818299982921698e-06, "loss": 0.5523, "step": 2778 }, { "epoch": 0.9040338321405335, "grad_norm": 2.069607686620754, "learning_rate": 8.817077712301931e-06, "loss": 0.6374, "step": 2779 }, { "epoch": 0.904359141184125, "grad_norm": 1.828216419766549, "learning_rate": 8.815854894692385e-06, "loss": 0.5641, "step": 2780 }, { "epoch": 0.9046844502277164, "grad_norm": 1.9249336220796907, "learning_rate": 8.814631530268288e-06, "loss": 0.6986, "step": 2781 }, { "epoch": 0.9050097592713078, "grad_norm": 1.7322850912702283, "learning_rate": 8.81340761920495e-06, "loss": 0.5876, "step": 2782 }, { "epoch": 0.9053350683148992, "grad_norm": 1.7330618492977212, "learning_rate": 8.812183161677756e-06, "loss": 0.5513, "step": 2783 }, { "epoch": 0.9056603773584906, "grad_norm": 1.6950219615716537, "learning_rate": 8.810958157862173e-06, "loss": 0.5634, "step": 2784 }, { "epoch": 0.905985686402082, "grad_norm": 1.6164180439428912, "learning_rate": 8.809732607933748e-06, "loss": 0.4866, "step": 2785 }, { "epoch": 0.9063109954456734, "grad_norm": 1.7474801654017136, "learning_rate": 8.808506512068098e-06, "loss": 0.6066, "step": 2786 }, { "epoch": 0.9066363044892648, "grad_norm": 1.9615587012881979, "learning_rate": 8.807279870440925e-06, "loss": 0.6497, "step": 2787 }, { "epoch": 0.9069616135328562, "grad_norm": 1.7671247699368642, "learning_rate": 8.806052683228011e-06, "loss": 0.5795, "step": 2788 }, { "epoch": 0.9072869225764476, "grad_norm": 1.7484316519572796, "learning_rate": 8.804824950605207e-06, "loss": 0.5925, "step": 2789 }, { "epoch": 0.907612231620039, "grad_norm": 1.6927325056649594, "learning_rate": 8.803596672748452e-06, "loss": 0.5261, "step": 2790 }, { "epoch": 0.9079375406636304, "grad_norm": 1.8932451839099966, "learning_rate": 8.802367849833757e-06, "loss": 0.7143, "step": 2791 }, { "epoch": 0.9082628497072218, "grad_norm": 1.8689004090037262, "learning_rate": 8.801138482037214e-06, "loss": 0.668, "step": 2792 }, { "epoch": 0.9085881587508132, "grad_norm": 1.7530371848725648, "learning_rate": 8.799908569534995e-06, "loss": 0.5703, "step": 2793 }, { "epoch": 0.9089134677944047, "grad_norm": 2.075598263881386, "learning_rate": 8.798678112503342e-06, "loss": 0.7005, "step": 2794 }, { "epoch": 0.9092387768379961, "grad_norm": 2.143292561025281, "learning_rate": 8.797447111118581e-06, "loss": 0.6834, "step": 2795 }, { "epoch": 0.9095640858815875, "grad_norm": 1.5949095909678885, "learning_rate": 8.796215565557121e-06, "loss": 0.5415, "step": 2796 }, { "epoch": 0.909889394925179, "grad_norm": 1.9327303446002664, "learning_rate": 8.794983475995438e-06, "loss": 0.6631, "step": 2797 }, { "epoch": 0.9102147039687704, "grad_norm": 1.715821463139042, "learning_rate": 8.793750842610092e-06, "loss": 0.6642, "step": 2798 }, { "epoch": 0.9105400130123618, "grad_norm": 1.6569715359981654, "learning_rate": 8.79251766557772e-06, "loss": 0.5885, "step": 2799 }, { "epoch": 0.9108653220559532, "grad_norm": 1.612317845775809, "learning_rate": 8.79128394507504e-06, "loss": 0.6568, "step": 2800 }, { "epoch": 0.9111906310995446, "grad_norm": 1.643401088836453, "learning_rate": 8.790049681278841e-06, "loss": 0.5663, "step": 2801 }, { "epoch": 0.911515940143136, "grad_norm": 1.7415082898651735, "learning_rate": 8.788814874365997e-06, "loss": 0.4919, "step": 2802 }, { "epoch": 0.9118412491867274, "grad_norm": 1.6956923014622811, "learning_rate": 8.787579524513455e-06, "loss": 0.5437, "step": 2803 }, { "epoch": 0.9121665582303188, "grad_norm": 1.843552306361835, "learning_rate": 8.786343631898242e-06, "loss": 0.5555, "step": 2804 }, { "epoch": 0.9124918672739102, "grad_norm": 2.053877611397669, "learning_rate": 8.785107196697462e-06, "loss": 0.7008, "step": 2805 }, { "epoch": 0.9128171763175016, "grad_norm": 1.7914821228719102, "learning_rate": 8.783870219088298e-06, "loss": 0.5161, "step": 2806 }, { "epoch": 0.913142485361093, "grad_norm": 1.8473700806131552, "learning_rate": 8.782632699248008e-06, "loss": 0.6196, "step": 2807 }, { "epoch": 0.9134677944046844, "grad_norm": 1.6189923418082488, "learning_rate": 8.78139463735393e-06, "loss": 0.5476, "step": 2808 }, { "epoch": 0.9137931034482759, "grad_norm": 1.6329538211270354, "learning_rate": 8.78015603358348e-06, "loss": 0.5436, "step": 2809 }, { "epoch": 0.9141184124918673, "grad_norm": 1.7921426983882855, "learning_rate": 8.778916888114149e-06, "loss": 0.6003, "step": 2810 }, { "epoch": 0.9144437215354587, "grad_norm": 1.7972158685632589, "learning_rate": 8.777677201123507e-06, "loss": 0.5178, "step": 2811 }, { "epoch": 0.9147690305790501, "grad_norm": 1.6856213518577747, "learning_rate": 8.776436972789206e-06, "loss": 0.651, "step": 2812 }, { "epoch": 0.9150943396226415, "grad_norm": 1.55663444363394, "learning_rate": 8.775196203288969e-06, "loss": 0.4274, "step": 2813 }, { "epoch": 0.9154196486662329, "grad_norm": 2.114327865865613, "learning_rate": 8.773954892800597e-06, "loss": 0.7139, "step": 2814 }, { "epoch": 0.9157449577098243, "grad_norm": 1.8074934642311518, "learning_rate": 8.772713041501974e-06, "loss": 0.5812, "step": 2815 }, { "epoch": 0.9160702667534157, "grad_norm": 2.033983001252358, "learning_rate": 8.771470649571056e-06, "loss": 0.6671, "step": 2816 }, { "epoch": 0.9163955757970071, "grad_norm": 1.8633353464781484, "learning_rate": 8.770227717185881e-06, "loss": 0.7022, "step": 2817 }, { "epoch": 0.9167208848405985, "grad_norm": 1.6066796668570211, "learning_rate": 8.76898424452456e-06, "loss": 0.5371, "step": 2818 }, { "epoch": 0.91704619388419, "grad_norm": 1.853627993127823, "learning_rate": 8.767740231765282e-06, "loss": 0.5435, "step": 2819 }, { "epoch": 0.9173715029277814, "grad_norm": 1.7903800023234595, "learning_rate": 8.766495679086319e-06, "loss": 0.4888, "step": 2820 }, { "epoch": 0.9176968119713728, "grad_norm": 1.8074913442068041, "learning_rate": 8.765250586666013e-06, "loss": 0.6284, "step": 2821 }, { "epoch": 0.9180221210149642, "grad_norm": 1.6970910447018508, "learning_rate": 8.764004954682788e-06, "loss": 0.5437, "step": 2822 }, { "epoch": 0.9183474300585557, "grad_norm": 1.702068549717593, "learning_rate": 8.762758783315141e-06, "loss": 0.6391, "step": 2823 }, { "epoch": 0.9186727391021471, "grad_norm": 1.877692529611802, "learning_rate": 8.761512072741655e-06, "loss": 0.5773, "step": 2824 }, { "epoch": 0.9189980481457385, "grad_norm": 1.7366750552327062, "learning_rate": 8.76026482314098e-06, "loss": 0.6249, "step": 2825 }, { "epoch": 0.9193233571893299, "grad_norm": 1.6279437030531159, "learning_rate": 8.759017034691847e-06, "loss": 0.6028, "step": 2826 }, { "epoch": 0.9196486662329213, "grad_norm": 1.7541710339886294, "learning_rate": 8.757768707573069e-06, "loss": 0.6258, "step": 2827 }, { "epoch": 0.9199739752765127, "grad_norm": 1.7138690669519523, "learning_rate": 8.756519841963528e-06, "loss": 0.5188, "step": 2828 }, { "epoch": 0.9202992843201041, "grad_norm": 1.8948918259982173, "learning_rate": 8.75527043804219e-06, "loss": 0.5861, "step": 2829 }, { "epoch": 0.9206245933636955, "grad_norm": 1.772248918550815, "learning_rate": 8.754020495988093e-06, "loss": 0.5802, "step": 2830 }, { "epoch": 0.9209499024072869, "grad_norm": 1.9278559284667682, "learning_rate": 8.752770015980359e-06, "loss": 0.7071, "step": 2831 }, { "epoch": 0.9212752114508783, "grad_norm": 1.9883111674081362, "learning_rate": 8.751518998198179e-06, "loss": 0.6859, "step": 2832 }, { "epoch": 0.9216005204944697, "grad_norm": 1.5217494356940204, "learning_rate": 8.750267442820823e-06, "loss": 0.4929, "step": 2833 }, { "epoch": 0.9219258295380611, "grad_norm": 1.5650903809265044, "learning_rate": 8.749015350027642e-06, "loss": 0.4421, "step": 2834 }, { "epoch": 0.9222511385816525, "grad_norm": 2.216776659358019, "learning_rate": 8.747762719998065e-06, "loss": 0.7505, "step": 2835 }, { "epoch": 0.9225764476252439, "grad_norm": 1.766065833280233, "learning_rate": 8.746509552911588e-06, "loss": 0.5489, "step": 2836 }, { "epoch": 0.9229017566688354, "grad_norm": 1.7843997281043726, "learning_rate": 8.745255848947797e-06, "loss": 0.5337, "step": 2837 }, { "epoch": 0.9232270657124269, "grad_norm": 1.440655547307111, "learning_rate": 8.744001608286344e-06, "loss": 0.5844, "step": 2838 }, { "epoch": 0.9235523747560183, "grad_norm": 1.9562046993230362, "learning_rate": 8.742746831106965e-06, "loss": 0.67, "step": 2839 }, { "epoch": 0.9238776837996097, "grad_norm": 1.7130132777907294, "learning_rate": 8.741491517589469e-06, "loss": 0.5821, "step": 2840 }, { "epoch": 0.9242029928432011, "grad_norm": 1.6449545986200662, "learning_rate": 8.740235667913745e-06, "loss": 0.5261, "step": 2841 }, { "epoch": 0.9245283018867925, "grad_norm": 1.765783228459685, "learning_rate": 8.738979282259755e-06, "loss": 0.5173, "step": 2842 }, { "epoch": 0.9248536109303839, "grad_norm": 1.9156245303712236, "learning_rate": 8.737722360807541e-06, "loss": 0.5635, "step": 2843 }, { "epoch": 0.9251789199739753, "grad_norm": 1.8545219369281905, "learning_rate": 8.736464903737222e-06, "loss": 0.593, "step": 2844 }, { "epoch": 0.9255042290175667, "grad_norm": 2.0650202014018784, "learning_rate": 8.73520691122899e-06, "loss": 0.6544, "step": 2845 }, { "epoch": 0.9258295380611581, "grad_norm": 1.5813333682527468, "learning_rate": 8.733948383463117e-06, "loss": 0.5318, "step": 2846 }, { "epoch": 0.9261548471047495, "grad_norm": 1.6937733581778038, "learning_rate": 8.732689320619952e-06, "loss": 0.5637, "step": 2847 }, { "epoch": 0.9264801561483409, "grad_norm": 1.7327819593518692, "learning_rate": 8.731429722879918e-06, "loss": 0.5887, "step": 2848 }, { "epoch": 0.9268054651919323, "grad_norm": 1.9704242192017802, "learning_rate": 8.730169590423518e-06, "loss": 0.6367, "step": 2849 }, { "epoch": 0.9271307742355237, "grad_norm": 1.8209902033915908, "learning_rate": 8.728908923431327e-06, "loss": 0.6446, "step": 2850 }, { "epoch": 0.9274560832791151, "grad_norm": 1.7565103433761167, "learning_rate": 8.727647722084003e-06, "loss": 0.576, "step": 2851 }, { "epoch": 0.9277813923227066, "grad_norm": 1.5437661201016544, "learning_rate": 8.726385986562273e-06, "loss": 0.5212, "step": 2852 }, { "epoch": 0.928106701366298, "grad_norm": 1.7770642697288435, "learning_rate": 8.725123717046948e-06, "loss": 0.5242, "step": 2853 }, { "epoch": 0.9284320104098894, "grad_norm": 1.8638981968024702, "learning_rate": 8.72386091371891e-06, "loss": 0.5958, "step": 2854 }, { "epoch": 0.9287573194534808, "grad_norm": 1.9160560164275664, "learning_rate": 8.72259757675912e-06, "loss": 0.6287, "step": 2855 }, { "epoch": 0.9290826284970722, "grad_norm": 1.8578019641606651, "learning_rate": 8.721333706348616e-06, "loss": 0.6304, "step": 2856 }, { "epoch": 0.9294079375406636, "grad_norm": 1.9009871406202394, "learning_rate": 8.720069302668511e-06, "loss": 0.6409, "step": 2857 }, { "epoch": 0.929733246584255, "grad_norm": 2.014398378843864, "learning_rate": 8.718804365899994e-06, "loss": 0.5835, "step": 2858 }, { "epoch": 0.9300585556278465, "grad_norm": 2.009924101480953, "learning_rate": 8.717538896224333e-06, "loss": 0.6475, "step": 2859 }, { "epoch": 0.9303838646714379, "grad_norm": 1.9797343565381589, "learning_rate": 8.716272893822868e-06, "loss": 0.6671, "step": 2860 }, { "epoch": 0.9307091737150293, "grad_norm": 1.5820157599544311, "learning_rate": 8.715006358877021e-06, "loss": 0.5225, "step": 2861 }, { "epoch": 0.9310344827586207, "grad_norm": 1.7016398803282131, "learning_rate": 8.713739291568286e-06, "loss": 0.6295, "step": 2862 }, { "epoch": 0.9313597918022121, "grad_norm": 1.9407806095145257, "learning_rate": 8.712471692078233e-06, "loss": 0.5949, "step": 2863 }, { "epoch": 0.9316851008458035, "grad_norm": 1.753017926474801, "learning_rate": 8.71120356058851e-06, "loss": 0.5901, "step": 2864 }, { "epoch": 0.9320104098893949, "grad_norm": 1.641584904561287, "learning_rate": 8.709934897280845e-06, "loss": 0.5619, "step": 2865 }, { "epoch": 0.9323357189329864, "grad_norm": 1.9739614671009693, "learning_rate": 8.708665702337036e-06, "loss": 0.6179, "step": 2866 }, { "epoch": 0.9326610279765778, "grad_norm": 1.890399071581635, "learning_rate": 8.707395975938958e-06, "loss": 0.6689, "step": 2867 }, { "epoch": 0.9329863370201692, "grad_norm": 1.715144815394492, "learning_rate": 8.706125718268565e-06, "loss": 0.6424, "step": 2868 }, { "epoch": 0.9333116460637606, "grad_norm": 2.0149014215225813, "learning_rate": 8.704854929507886e-06, "loss": 0.6638, "step": 2869 }, { "epoch": 0.933636955107352, "grad_norm": 1.6831246852220945, "learning_rate": 8.703583609839025e-06, "loss": 0.5758, "step": 2870 }, { "epoch": 0.9339622641509434, "grad_norm": 1.8198082228528103, "learning_rate": 8.702311759444163e-06, "loss": 0.5823, "step": 2871 }, { "epoch": 0.9342875731945348, "grad_norm": 1.851379547188979, "learning_rate": 8.701039378505558e-06, "loss": 0.4676, "step": 2872 }, { "epoch": 0.9346128822381262, "grad_norm": 1.8634579855440434, "learning_rate": 8.699766467205543e-06, "loss": 0.5057, "step": 2873 }, { "epoch": 0.9349381912817176, "grad_norm": 1.6803838392009078, "learning_rate": 8.698493025726526e-06, "loss": 0.5444, "step": 2874 }, { "epoch": 0.935263500325309, "grad_norm": 1.5384321944393675, "learning_rate": 8.697219054250992e-06, "loss": 0.4709, "step": 2875 }, { "epoch": 0.9355888093689004, "grad_norm": 2.113276616739381, "learning_rate": 8.695944552961503e-06, "loss": 0.628, "step": 2876 }, { "epoch": 0.9359141184124918, "grad_norm": 1.6622934052503298, "learning_rate": 8.694669522040695e-06, "loss": 0.5261, "step": 2877 }, { "epoch": 0.9362394274560832, "grad_norm": 1.9183213272591266, "learning_rate": 8.693393961671282e-06, "loss": 0.5698, "step": 2878 }, { "epoch": 0.9365647364996746, "grad_norm": 1.814010326105444, "learning_rate": 8.692117872036051e-06, "loss": 0.6366, "step": 2879 }, { "epoch": 0.936890045543266, "grad_norm": 2.049657858438379, "learning_rate": 8.690841253317868e-06, "loss": 0.6355, "step": 2880 }, { "epoch": 0.9372153545868576, "grad_norm": 1.9095906380897987, "learning_rate": 8.689564105699672e-06, "loss": 0.6805, "step": 2881 }, { "epoch": 0.937540663630449, "grad_norm": 1.934503021124236, "learning_rate": 8.688286429364478e-06, "loss": 0.6905, "step": 2882 }, { "epoch": 0.9378659726740404, "grad_norm": 1.7503386574625357, "learning_rate": 8.687008224495383e-06, "loss": 0.5721, "step": 2883 }, { "epoch": 0.9381912817176318, "grad_norm": 1.7463739246909302, "learning_rate": 8.68572949127555e-06, "loss": 0.5677, "step": 2884 }, { "epoch": 0.9385165907612232, "grad_norm": 1.8579918055675362, "learning_rate": 8.684450229888223e-06, "loss": 0.6564, "step": 2885 }, { "epoch": 0.9388418998048146, "grad_norm": 1.9601294168075827, "learning_rate": 8.683170440516721e-06, "loss": 0.6485, "step": 2886 }, { "epoch": 0.939167208848406, "grad_norm": 1.9685278998493152, "learning_rate": 8.681890123344441e-06, "loss": 0.5907, "step": 2887 }, { "epoch": 0.9394925178919974, "grad_norm": 2.1458535232752944, "learning_rate": 8.680609278554851e-06, "loss": 0.7258, "step": 2888 }, { "epoch": 0.9398178269355888, "grad_norm": 1.7416341171630576, "learning_rate": 8.679327906331495e-06, "loss": 0.5728, "step": 2889 }, { "epoch": 0.9401431359791802, "grad_norm": 1.7839646916449983, "learning_rate": 8.678046006857999e-06, "loss": 0.6094, "step": 2890 }, { "epoch": 0.9404684450227716, "grad_norm": 1.5985744494906875, "learning_rate": 8.676763580318058e-06, "loss": 0.4659, "step": 2891 }, { "epoch": 0.940793754066363, "grad_norm": 1.8201354198837163, "learning_rate": 8.675480626895443e-06, "loss": 0.5954, "step": 2892 }, { "epoch": 0.9411190631099544, "grad_norm": 1.5206653428783234, "learning_rate": 8.674197146774005e-06, "loss": 0.4482, "step": 2893 }, { "epoch": 0.9414443721535458, "grad_norm": 1.634783684931469, "learning_rate": 8.672913140137666e-06, "loss": 0.5758, "step": 2894 }, { "epoch": 0.9417696811971373, "grad_norm": 1.7094643909861313, "learning_rate": 8.671628607170427e-06, "loss": 0.5884, "step": 2895 }, { "epoch": 0.9420949902407287, "grad_norm": 1.5136405380948863, "learning_rate": 8.670343548056356e-06, "loss": 0.4612, "step": 2896 }, { "epoch": 0.9424202992843201, "grad_norm": 1.7670142722810949, "learning_rate": 8.669057962979612e-06, "loss": 0.5794, "step": 2897 }, { "epoch": 0.9427456083279115, "grad_norm": 1.9137436403160337, "learning_rate": 8.667771852124414e-06, "loss": 0.6507, "step": 2898 }, { "epoch": 0.943070917371503, "grad_norm": 1.8772795113271736, "learning_rate": 8.666485215675063e-06, "loss": 0.5515, "step": 2899 }, { "epoch": 0.9433962264150944, "grad_norm": 1.6797100377340821, "learning_rate": 8.665198053815938e-06, "loss": 0.5284, "step": 2900 }, { "epoch": 0.9437215354586858, "grad_norm": 1.5032451048239959, "learning_rate": 8.663910366731487e-06, "loss": 0.5343, "step": 2901 }, { "epoch": 0.9440468445022772, "grad_norm": 1.663407108007874, "learning_rate": 8.662622154606238e-06, "loss": 0.5457, "step": 2902 }, { "epoch": 0.9443721535458686, "grad_norm": 1.7932011556953755, "learning_rate": 8.661333417624792e-06, "loss": 0.6286, "step": 2903 }, { "epoch": 0.94469746258946, "grad_norm": 1.7962388564336964, "learning_rate": 8.660044155971826e-06, "loss": 0.4741, "step": 2904 }, { "epoch": 0.9450227716330514, "grad_norm": 1.748937375672057, "learning_rate": 8.658754369832091e-06, "loss": 0.5718, "step": 2905 }, { "epoch": 0.9453480806766428, "grad_norm": 2.141710872951059, "learning_rate": 8.657464059390418e-06, "loss": 0.6184, "step": 2906 }, { "epoch": 0.9456733897202342, "grad_norm": 1.7412936359141682, "learning_rate": 8.656173224831706e-06, "loss": 0.5387, "step": 2907 }, { "epoch": 0.9459986987638256, "grad_norm": 1.8371298777064813, "learning_rate": 8.654881866340932e-06, "loss": 0.6274, "step": 2908 }, { "epoch": 0.9463240078074171, "grad_norm": 1.712817282221003, "learning_rate": 8.65358998410315e-06, "loss": 0.5618, "step": 2909 }, { "epoch": 0.9466493168510085, "grad_norm": 1.6029294674496337, "learning_rate": 8.652297578303488e-06, "loss": 0.5882, "step": 2910 }, { "epoch": 0.9469746258945999, "grad_norm": 2.044325014982439, "learning_rate": 8.651004649127148e-06, "loss": 0.5457, "step": 2911 }, { "epoch": 0.9472999349381913, "grad_norm": 1.802124681370089, "learning_rate": 8.649711196759407e-06, "loss": 0.6274, "step": 2912 }, { "epoch": 0.9476252439817827, "grad_norm": 1.5921883540349198, "learning_rate": 8.648417221385618e-06, "loss": 0.545, "step": 2913 }, { "epoch": 0.9479505530253741, "grad_norm": 1.6432412681720843, "learning_rate": 8.64712272319121e-06, "loss": 0.5452, "step": 2914 }, { "epoch": 0.9482758620689655, "grad_norm": 1.818167707991131, "learning_rate": 8.645827702361683e-06, "loss": 0.5372, "step": 2915 }, { "epoch": 0.9486011711125569, "grad_norm": 1.8676685097510843, "learning_rate": 8.644532159082616e-06, "loss": 0.7134, "step": 2916 }, { "epoch": 0.9489264801561483, "grad_norm": 1.9800505043543037, "learning_rate": 8.64323609353966e-06, "loss": 0.6024, "step": 2917 }, { "epoch": 0.9492517891997397, "grad_norm": 1.8856192385165094, "learning_rate": 8.641939505918544e-06, "loss": 0.5725, "step": 2918 }, { "epoch": 0.9495770982433311, "grad_norm": 1.6898291831612107, "learning_rate": 8.640642396405068e-06, "loss": 0.631, "step": 2919 }, { "epoch": 0.9499024072869225, "grad_norm": 1.9425794956051803, "learning_rate": 8.639344765185108e-06, "loss": 0.6683, "step": 2920 }, { "epoch": 0.950227716330514, "grad_norm": 2.037294653308386, "learning_rate": 8.638046612444618e-06, "loss": 0.6775, "step": 2921 }, { "epoch": 0.9505530253741054, "grad_norm": 1.8837769043300434, "learning_rate": 8.636747938369623e-06, "loss": 0.5695, "step": 2922 }, { "epoch": 0.9508783344176968, "grad_norm": 1.7696542179896493, "learning_rate": 8.635448743146223e-06, "loss": 0.6057, "step": 2923 }, { "epoch": 0.9512036434612883, "grad_norm": 1.7674918250458092, "learning_rate": 8.634149026960594e-06, "loss": 0.6677, "step": 2924 }, { "epoch": 0.9515289525048797, "grad_norm": 2.114581277505755, "learning_rate": 8.632848789998986e-06, "loss": 0.687, "step": 2925 }, { "epoch": 0.9518542615484711, "grad_norm": 2.1700883330878047, "learning_rate": 8.631548032447726e-06, "loss": 0.5934, "step": 2926 }, { "epoch": 0.9521795705920625, "grad_norm": 2.020045832680568, "learning_rate": 8.63024675449321e-06, "loss": 0.5552, "step": 2927 }, { "epoch": 0.9525048796356539, "grad_norm": 1.8704999326211824, "learning_rate": 8.628944956321917e-06, "loss": 0.5536, "step": 2928 }, { "epoch": 0.9528301886792453, "grad_norm": 1.8961540424090157, "learning_rate": 8.627642638120389e-06, "loss": 0.5652, "step": 2929 }, { "epoch": 0.9531554977228367, "grad_norm": 2.0030520959230818, "learning_rate": 8.626339800075252e-06, "loss": 0.6232, "step": 2930 }, { "epoch": 0.9534808067664281, "grad_norm": 1.7141322897696851, "learning_rate": 8.625036442373207e-06, "loss": 0.5097, "step": 2931 }, { "epoch": 0.9538061158100195, "grad_norm": 1.8342916216206415, "learning_rate": 8.623732565201019e-06, "loss": 0.5822, "step": 2932 }, { "epoch": 0.9541314248536109, "grad_norm": 1.7432016746769559, "learning_rate": 8.622428168745542e-06, "loss": 0.5313, "step": 2933 }, { "epoch": 0.9544567338972023, "grad_norm": 2.02371267782925, "learning_rate": 8.62112325319369e-06, "loss": 0.6227, "step": 2934 }, { "epoch": 0.9547820429407937, "grad_norm": 1.6292002036030195, "learning_rate": 8.619817818732463e-06, "loss": 0.5133, "step": 2935 }, { "epoch": 0.9551073519843851, "grad_norm": 1.700795910343844, "learning_rate": 8.618511865548929e-06, "loss": 0.5343, "step": 2936 }, { "epoch": 0.9554326610279765, "grad_norm": 1.8974820202702707, "learning_rate": 8.617205393830232e-06, "loss": 0.5342, "step": 2937 }, { "epoch": 0.955757970071568, "grad_norm": 1.7360704510751424, "learning_rate": 8.615898403763592e-06, "loss": 0.65, "step": 2938 }, { "epoch": 0.9560832791151594, "grad_norm": 1.9715614001229644, "learning_rate": 8.614590895536298e-06, "loss": 0.61, "step": 2939 }, { "epoch": 0.9564085881587508, "grad_norm": 1.7611328349612108, "learning_rate": 8.61328286933572e-06, "loss": 0.5667, "step": 2940 }, { "epoch": 0.9567338972023423, "grad_norm": 1.8677392717276677, "learning_rate": 8.611974325349298e-06, "loss": 0.6568, "step": 2941 }, { "epoch": 0.9570592062459337, "grad_norm": 1.760901791236257, "learning_rate": 8.610665263764543e-06, "loss": 0.6085, "step": 2942 }, { "epoch": 0.9573845152895251, "grad_norm": 1.7746077391219497, "learning_rate": 8.609355684769052e-06, "loss": 0.5053, "step": 2943 }, { "epoch": 0.9577098243331165, "grad_norm": 1.7231704602082458, "learning_rate": 8.608045588550484e-06, "loss": 0.6045, "step": 2944 }, { "epoch": 0.9580351333767079, "grad_norm": 1.8937667752389666, "learning_rate": 8.606734975296578e-06, "loss": 0.6115, "step": 2945 }, { "epoch": 0.9583604424202993, "grad_norm": 1.742985359151245, "learning_rate": 8.605423845195145e-06, "loss": 0.6764, "step": 2946 }, { "epoch": 0.9586857514638907, "grad_norm": 1.6694986785681931, "learning_rate": 8.604112198434072e-06, "loss": 0.4732, "step": 2947 }, { "epoch": 0.9590110605074821, "grad_norm": 1.8968302680293685, "learning_rate": 8.602800035201317e-06, "loss": 0.7208, "step": 2948 }, { "epoch": 0.9593363695510735, "grad_norm": 1.6580345770188352, "learning_rate": 8.601487355684915e-06, "loss": 0.613, "step": 2949 }, { "epoch": 0.9596616785946649, "grad_norm": 1.9453290829379712, "learning_rate": 8.600174160072974e-06, "loss": 0.6302, "step": 2950 }, { "epoch": 0.9599869876382563, "grad_norm": 1.8411781766963011, "learning_rate": 8.598860448553675e-06, "loss": 0.6424, "step": 2951 }, { "epoch": 0.9603122966818478, "grad_norm": 1.8291594830385782, "learning_rate": 8.597546221315275e-06, "loss": 0.6413, "step": 2952 }, { "epoch": 0.9606376057254392, "grad_norm": 1.848378243238424, "learning_rate": 8.596231478546102e-06, "loss": 0.6395, "step": 2953 }, { "epoch": 0.9609629147690306, "grad_norm": 1.7983765571440213, "learning_rate": 8.594916220434561e-06, "loss": 0.6798, "step": 2954 }, { "epoch": 0.961288223812622, "grad_norm": 1.7542084923500172, "learning_rate": 8.593600447169129e-06, "loss": 0.6538, "step": 2955 }, { "epoch": 0.9616135328562134, "grad_norm": 1.8824411048390426, "learning_rate": 8.592284158938357e-06, "loss": 0.6844, "step": 2956 }, { "epoch": 0.9619388418998048, "grad_norm": 1.55920993554036, "learning_rate": 8.590967355930869e-06, "loss": 0.5214, "step": 2957 }, { "epoch": 0.9622641509433962, "grad_norm": 1.6795403701502698, "learning_rate": 8.589650038335366e-06, "loss": 0.5722, "step": 2958 }, { "epoch": 0.9625894599869876, "grad_norm": 1.827812313527842, "learning_rate": 8.588332206340617e-06, "loss": 0.497, "step": 2959 }, { "epoch": 0.962914769030579, "grad_norm": 1.6675457519282049, "learning_rate": 8.587013860135473e-06, "loss": 0.5794, "step": 2960 }, { "epoch": 0.9632400780741704, "grad_norm": 1.6850821234670896, "learning_rate": 8.585694999908849e-06, "loss": 0.5463, "step": 2961 }, { "epoch": 0.9635653871177619, "grad_norm": 2.074226019643338, "learning_rate": 8.584375625849743e-06, "loss": 0.7915, "step": 2962 }, { "epoch": 0.9638906961613533, "grad_norm": 1.9314610471134246, "learning_rate": 8.583055738147219e-06, "loss": 0.6008, "step": 2963 }, { "epoch": 0.9642160052049447, "grad_norm": 1.6851047697105797, "learning_rate": 8.581735336990418e-06, "loss": 0.5978, "step": 2964 }, { "epoch": 0.9645413142485361, "grad_norm": 1.925035899754262, "learning_rate": 8.580414422568557e-06, "loss": 0.4972, "step": 2965 }, { "epoch": 0.9648666232921275, "grad_norm": 1.8949222683427527, "learning_rate": 8.57909299507092e-06, "loss": 0.5884, "step": 2966 }, { "epoch": 0.965191932335719, "grad_norm": 1.939078758744815, "learning_rate": 8.577771054686871e-06, "loss": 0.6356, "step": 2967 }, { "epoch": 0.9655172413793104, "grad_norm": 1.7997116767314987, "learning_rate": 8.576448601605845e-06, "loss": 0.5701, "step": 2968 }, { "epoch": 0.9658425504229018, "grad_norm": 1.881525247459023, "learning_rate": 8.57512563601735e-06, "loss": 0.6186, "step": 2969 }, { "epoch": 0.9661678594664932, "grad_norm": 1.9363748353712242, "learning_rate": 8.57380215811097e-06, "loss": 0.6256, "step": 2970 }, { "epoch": 0.9664931685100846, "grad_norm": 1.4880158332180933, "learning_rate": 8.572478168076352e-06, "loss": 0.4671, "step": 2971 }, { "epoch": 0.966818477553676, "grad_norm": 1.8956531910259888, "learning_rate": 8.571153666103235e-06, "loss": 0.6463, "step": 2972 }, { "epoch": 0.9671437865972674, "grad_norm": 1.640496515388062, "learning_rate": 8.569828652381417e-06, "loss": 0.6166, "step": 2973 }, { "epoch": 0.9674690956408588, "grad_norm": 1.6033001797306108, "learning_rate": 8.568503127100771e-06, "loss": 0.609, "step": 2974 }, { "epoch": 0.9677944046844502, "grad_norm": 1.6623987162028901, "learning_rate": 8.567177090451249e-06, "loss": 0.533, "step": 2975 }, { "epoch": 0.9681197137280416, "grad_norm": 2.033443352310173, "learning_rate": 8.56585054262287e-06, "loss": 0.7716, "step": 2976 }, { "epoch": 0.968445022771633, "grad_norm": 1.8056711840016273, "learning_rate": 8.564523483805733e-06, "loss": 0.6017, "step": 2977 }, { "epoch": 0.9687703318152244, "grad_norm": 1.7751662317986214, "learning_rate": 8.563195914190003e-06, "loss": 0.6379, "step": 2978 }, { "epoch": 0.9690956408588158, "grad_norm": 1.7530098206799676, "learning_rate": 8.561867833965922e-06, "loss": 0.6208, "step": 2979 }, { "epoch": 0.9694209499024072, "grad_norm": 1.908027234086113, "learning_rate": 8.560539243323806e-06, "loss": 0.5903, "step": 2980 }, { "epoch": 0.9697462589459988, "grad_norm": 2.0165532090968696, "learning_rate": 8.559210142454043e-06, "loss": 0.6286, "step": 2981 }, { "epoch": 0.9700715679895902, "grad_norm": 1.840124113894837, "learning_rate": 8.557880531547093e-06, "loss": 0.5213, "step": 2982 }, { "epoch": 0.9703968770331816, "grad_norm": 1.599479167410141, "learning_rate": 8.556550410793494e-06, "loss": 0.5728, "step": 2983 }, { "epoch": 0.970722186076773, "grad_norm": 1.747079389842805, "learning_rate": 8.555219780383846e-06, "loss": 0.4911, "step": 2984 }, { "epoch": 0.9710474951203644, "grad_norm": 1.857100801793416, "learning_rate": 8.553888640508834e-06, "loss": 0.6925, "step": 2985 }, { "epoch": 0.9713728041639558, "grad_norm": 1.7407694108664333, "learning_rate": 8.552556991359211e-06, "loss": 0.5951, "step": 2986 }, { "epoch": 0.9716981132075472, "grad_norm": 1.9686811008691414, "learning_rate": 8.551224833125802e-06, "loss": 0.5987, "step": 2987 }, { "epoch": 0.9720234222511386, "grad_norm": 1.8270801312941092, "learning_rate": 8.549892165999505e-06, "loss": 0.5677, "step": 2988 }, { "epoch": 0.97234873129473, "grad_norm": 1.8430162624505118, "learning_rate": 8.548558990171297e-06, "loss": 0.6607, "step": 2989 }, { "epoch": 0.9726740403383214, "grad_norm": 1.834802237531956, "learning_rate": 8.547225305832218e-06, "loss": 0.6832, "step": 2990 }, { "epoch": 0.9729993493819128, "grad_norm": 1.7471215152707904, "learning_rate": 8.545891113173388e-06, "loss": 0.5309, "step": 2991 }, { "epoch": 0.9733246584255042, "grad_norm": 1.8417304287043945, "learning_rate": 8.544556412385997e-06, "loss": 0.4706, "step": 2992 }, { "epoch": 0.9736499674690956, "grad_norm": 2.021094218392835, "learning_rate": 8.543221203661309e-06, "loss": 0.5896, "step": 2993 }, { "epoch": 0.973975276512687, "grad_norm": 2.305766023280336, "learning_rate": 8.54188548719066e-06, "loss": 0.7166, "step": 2994 }, { "epoch": 0.9743005855562785, "grad_norm": 2.0146037612410397, "learning_rate": 8.54054926316546e-06, "loss": 0.588, "step": 2995 }, { "epoch": 0.9746258945998699, "grad_norm": 1.8844968127664656, "learning_rate": 8.53921253177719e-06, "loss": 0.5611, "step": 2996 }, { "epoch": 0.9749512036434613, "grad_norm": 1.6925503462437457, "learning_rate": 8.537875293217404e-06, "loss": 0.5468, "step": 2997 }, { "epoch": 0.9752765126870527, "grad_norm": 1.6473991277304847, "learning_rate": 8.536537547677729e-06, "loss": 0.5615, "step": 2998 }, { "epoch": 0.9756018217306441, "grad_norm": 1.7202318157259768, "learning_rate": 8.535199295349867e-06, "loss": 0.5475, "step": 2999 }, { "epoch": 0.9759271307742355, "grad_norm": 2.0728061724646807, "learning_rate": 8.533860536425586e-06, "loss": 0.5731, "step": 3000 }, { "epoch": 0.976252439817827, "grad_norm": 1.8495931012000397, "learning_rate": 8.532521271096738e-06, "loss": 0.715, "step": 3001 }, { "epoch": 0.9765777488614183, "grad_norm": 1.733647648209118, "learning_rate": 8.531181499555235e-06, "loss": 0.5473, "step": 3002 }, { "epoch": 0.9769030579050098, "grad_norm": 2.1071257786915893, "learning_rate": 8.529841221993068e-06, "loss": 0.7536, "step": 3003 }, { "epoch": 0.9772283669486012, "grad_norm": 2.1408763893586875, "learning_rate": 8.528500438602301e-06, "loss": 0.6105, "step": 3004 }, { "epoch": 0.9775536759921926, "grad_norm": 1.5917067211529339, "learning_rate": 8.527159149575068e-06, "loss": 0.5531, "step": 3005 }, { "epoch": 0.977878985035784, "grad_norm": 1.6423372027851617, "learning_rate": 8.525817355103578e-06, "loss": 0.5084, "step": 3006 }, { "epoch": 0.9782042940793754, "grad_norm": 1.7817880593964635, "learning_rate": 8.52447505538011e-06, "loss": 0.5689, "step": 3007 }, { "epoch": 0.9785296031229668, "grad_norm": 1.7438249698507289, "learning_rate": 8.523132250597017e-06, "loss": 0.5388, "step": 3008 }, { "epoch": 0.9788549121665582, "grad_norm": 1.6980121333793117, "learning_rate": 8.521788940946723e-06, "loss": 0.4648, "step": 3009 }, { "epoch": 0.9791802212101497, "grad_norm": 1.6941279425110367, "learning_rate": 8.520445126621726e-06, "loss": 0.5069, "step": 3010 }, { "epoch": 0.9795055302537411, "grad_norm": 1.7588634773999672, "learning_rate": 8.519100807814596e-06, "loss": 0.5544, "step": 3011 }, { "epoch": 0.9798308392973325, "grad_norm": 1.9001871797769265, "learning_rate": 8.517755984717972e-06, "loss": 0.6425, "step": 3012 }, { "epoch": 0.9801561483409239, "grad_norm": 1.7982205486699685, "learning_rate": 8.51641065752457e-06, "loss": 0.4802, "step": 3013 }, { "epoch": 0.9804814573845153, "grad_norm": 1.9409458087262184, "learning_rate": 8.515064826427179e-06, "loss": 0.7498, "step": 3014 }, { "epoch": 0.9808067664281067, "grad_norm": 2.179721771005282, "learning_rate": 8.513718491618651e-06, "loss": 0.7291, "step": 3015 }, { "epoch": 0.9811320754716981, "grad_norm": 1.8470215615490313, "learning_rate": 8.512371653291924e-06, "loss": 0.5616, "step": 3016 }, { "epoch": 0.9814573845152895, "grad_norm": 1.7103815800823032, "learning_rate": 8.511024311639995e-06, "loss": 0.4784, "step": 3017 }, { "epoch": 0.9817826935588809, "grad_norm": 1.644737313618674, "learning_rate": 8.509676466855943e-06, "loss": 0.5722, "step": 3018 }, { "epoch": 0.9821080026024723, "grad_norm": 1.5977355450530981, "learning_rate": 8.508328119132912e-06, "loss": 0.4816, "step": 3019 }, { "epoch": 0.9824333116460637, "grad_norm": 1.9543994715790296, "learning_rate": 8.506979268664124e-06, "loss": 0.6455, "step": 3020 }, { "epoch": 0.9827586206896551, "grad_norm": 1.8052902583779502, "learning_rate": 8.505629915642866e-06, "loss": 0.5707, "step": 3021 }, { "epoch": 0.9830839297332465, "grad_norm": 1.7955331193296087, "learning_rate": 8.504280060262506e-06, "loss": 0.5072, "step": 3022 }, { "epoch": 0.983409238776838, "grad_norm": 1.7792548850640764, "learning_rate": 8.502929702716475e-06, "loss": 0.6633, "step": 3023 }, { "epoch": 0.9837345478204295, "grad_norm": 1.8328640735949784, "learning_rate": 8.501578843198283e-06, "loss": 0.5598, "step": 3024 }, { "epoch": 0.9840598568640209, "grad_norm": 1.6877563219433296, "learning_rate": 8.500227481901509e-06, "loss": 0.5353, "step": 3025 }, { "epoch": 0.9843851659076123, "grad_norm": 1.6919733099148908, "learning_rate": 8.498875619019802e-06, "loss": 0.7095, "step": 3026 }, { "epoch": 0.9847104749512037, "grad_norm": 1.6665362876361571, "learning_rate": 8.497523254746886e-06, "loss": 0.5994, "step": 3027 }, { "epoch": 0.9850357839947951, "grad_norm": 2.0379348406260465, "learning_rate": 8.496170389276557e-06, "loss": 0.76, "step": 3028 }, { "epoch": 0.9853610930383865, "grad_norm": 1.9054252517687291, "learning_rate": 8.49481702280268e-06, "loss": 0.6091, "step": 3029 }, { "epoch": 0.9856864020819779, "grad_norm": 1.9137050217961449, "learning_rate": 8.493463155519193e-06, "loss": 0.6371, "step": 3030 }, { "epoch": 0.9860117111255693, "grad_norm": 1.7107683522877575, "learning_rate": 8.492108787620106e-06, "loss": 0.5029, "step": 3031 }, { "epoch": 0.9863370201691607, "grad_norm": 2.0570254107955703, "learning_rate": 8.490753919299501e-06, "loss": 0.783, "step": 3032 }, { "epoch": 0.9866623292127521, "grad_norm": 1.6623561303144336, "learning_rate": 8.489398550751534e-06, "loss": 0.5225, "step": 3033 }, { "epoch": 0.9869876382563435, "grad_norm": 1.8278083895447823, "learning_rate": 8.488042682170428e-06, "loss": 0.4884, "step": 3034 }, { "epoch": 0.9873129472999349, "grad_norm": 1.7689447616026912, "learning_rate": 8.486686313750478e-06, "loss": 0.5031, "step": 3035 }, { "epoch": 0.9876382563435263, "grad_norm": 2.0413591576623107, "learning_rate": 8.485329445686056e-06, "loss": 0.6243, "step": 3036 }, { "epoch": 0.9879635653871177, "grad_norm": 2.0486873808950645, "learning_rate": 8.4839720781716e-06, "loss": 0.6398, "step": 3037 }, { "epoch": 0.9882888744307091, "grad_norm": 1.8263059477149266, "learning_rate": 8.482614211401623e-06, "loss": 0.4989, "step": 3038 }, { "epoch": 0.9886141834743006, "grad_norm": 1.960376655612137, "learning_rate": 8.481255845570707e-06, "loss": 0.7228, "step": 3039 }, { "epoch": 0.988939492517892, "grad_norm": 1.9656190332668289, "learning_rate": 8.479896980873508e-06, "loss": 0.6125, "step": 3040 }, { "epoch": 0.9892648015614834, "grad_norm": 1.6782726035160043, "learning_rate": 8.47853761750475e-06, "loss": 0.5811, "step": 3041 }, { "epoch": 0.9895901106050748, "grad_norm": 1.8972852932822173, "learning_rate": 8.477177755659234e-06, "loss": 0.6464, "step": 3042 }, { "epoch": 0.9899154196486663, "grad_norm": 1.8831086952566418, "learning_rate": 8.475817395531827e-06, "loss": 0.6909, "step": 3043 }, { "epoch": 0.9902407286922577, "grad_norm": 1.5703958341555488, "learning_rate": 8.474456537317469e-06, "loss": 0.595, "step": 3044 }, { "epoch": 0.9905660377358491, "grad_norm": 1.8507538137522919, "learning_rate": 8.473095181211173e-06, "loss": 0.564, "step": 3045 }, { "epoch": 0.9908913467794405, "grad_norm": 1.6881433787450055, "learning_rate": 8.471733327408023e-06, "loss": 0.446, "step": 3046 }, { "epoch": 0.9912166558230319, "grad_norm": 1.647431347184413, "learning_rate": 8.470370976103171e-06, "loss": 0.5781, "step": 3047 }, { "epoch": 0.9915419648666233, "grad_norm": 2.108069124851279, "learning_rate": 8.469008127491845e-06, "loss": 0.681, "step": 3048 }, { "epoch": 0.9918672739102147, "grad_norm": 1.9171955262519262, "learning_rate": 8.46764478176934e-06, "loss": 0.5456, "step": 3049 }, { "epoch": 0.9921925829538061, "grad_norm": 1.797620259332808, "learning_rate": 8.466280939131027e-06, "loss": 0.6889, "step": 3050 }, { "epoch": 0.9925178919973975, "grad_norm": 1.7356173044250953, "learning_rate": 8.464916599772344e-06, "loss": 0.6462, "step": 3051 }, { "epoch": 0.9928432010409889, "grad_norm": 1.6262257382775054, "learning_rate": 8.463551763888802e-06, "loss": 0.5071, "step": 3052 }, { "epoch": 0.9931685100845804, "grad_norm": 1.6850537217825512, "learning_rate": 8.462186431675984e-06, "loss": 0.6157, "step": 3053 }, { "epoch": 0.9934938191281718, "grad_norm": 1.8781442276827884, "learning_rate": 8.460820603329542e-06, "loss": 0.6307, "step": 3054 }, { "epoch": 0.9938191281717632, "grad_norm": 1.8231370691287812, "learning_rate": 8.459454279045198e-06, "loss": 0.6377, "step": 3055 }, { "epoch": 0.9941444372153546, "grad_norm": 1.8398842522989962, "learning_rate": 8.45808745901875e-06, "loss": 0.5366, "step": 3056 }, { "epoch": 0.994469746258946, "grad_norm": 1.8955595909898393, "learning_rate": 8.456720143446066e-06, "loss": 0.7084, "step": 3057 }, { "epoch": 0.9947950553025374, "grad_norm": 2.0093541583218424, "learning_rate": 8.455352332523079e-06, "loss": 0.6145, "step": 3058 }, { "epoch": 0.9951203643461288, "grad_norm": 1.9446657987070537, "learning_rate": 8.453984026445799e-06, "loss": 0.6739, "step": 3059 }, { "epoch": 0.9954456733897202, "grad_norm": 1.9715731138430779, "learning_rate": 8.452615225410306e-06, "loss": 0.6558, "step": 3060 }, { "epoch": 0.9957709824333116, "grad_norm": 1.9002527132105276, "learning_rate": 8.451245929612747e-06, "loss": 0.5923, "step": 3061 }, { "epoch": 0.996096291476903, "grad_norm": 1.8102520467828276, "learning_rate": 8.449876139249346e-06, "loss": 0.5798, "step": 3062 }, { "epoch": 0.9964216005204944, "grad_norm": 1.8655259119333116, "learning_rate": 8.448505854516396e-06, "loss": 0.6331, "step": 3063 }, { "epoch": 0.9967469095640858, "grad_norm": 1.5325109459807902, "learning_rate": 8.447135075610257e-06, "loss": 0.4643, "step": 3064 }, { "epoch": 0.9970722186076773, "grad_norm": 1.7583827706791535, "learning_rate": 8.445763802727363e-06, "loss": 0.6271, "step": 3065 }, { "epoch": 0.9973975276512687, "grad_norm": 1.7145316389471672, "learning_rate": 8.444392036064218e-06, "loss": 0.5375, "step": 3066 }, { "epoch": 0.9977228366948602, "grad_norm": 1.8744375781589881, "learning_rate": 8.443019775817398e-06, "loss": 0.7159, "step": 3067 }, { "epoch": 0.9980481457384516, "grad_norm": 1.87314001750543, "learning_rate": 8.441647022183548e-06, "loss": 0.6252, "step": 3068 }, { "epoch": 0.998373454782043, "grad_norm": 1.8034321487376022, "learning_rate": 8.440273775359386e-06, "loss": 0.6551, "step": 3069 }, { "epoch": 0.9986987638256344, "grad_norm": 1.7098825785775165, "learning_rate": 8.438900035541699e-06, "loss": 0.5143, "step": 3070 }, { "epoch": 0.9990240728692258, "grad_norm": 1.7000594295575797, "learning_rate": 8.437525802927344e-06, "loss": 0.5772, "step": 3071 }, { "epoch": 0.9993493819128172, "grad_norm": 1.7676575788532702, "learning_rate": 8.436151077713248e-06, "loss": 0.606, "step": 3072 }, { "epoch": 0.9996746909564086, "grad_norm": 1.57712783522087, "learning_rate": 8.434775860096414e-06, "loss": 0.5097, "step": 3073 }, { "epoch": 1.0, "grad_norm": 1.748258732180043, "learning_rate": 8.433400150273907e-06, "loss": 0.6634, "step": 3074 }, { "epoch": 1.0003253090435915, "grad_norm": 1.6489836462057683, "learning_rate": 8.432023948442869e-06, "loss": 0.3984, "step": 3075 }, { "epoch": 1.0006506180871828, "grad_norm": 1.6592418652412437, "learning_rate": 8.430647254800513e-06, "loss": 0.4462, "step": 3076 }, { "epoch": 1.0009759271307743, "grad_norm": 1.8124735440032436, "learning_rate": 8.429270069544116e-06, "loss": 0.4899, "step": 3077 }, { "epoch": 1.0013012361743656, "grad_norm": 2.0032284998966996, "learning_rate": 8.427892392871033e-06, "loss": 0.5191, "step": 3078 }, { "epoch": 1.0016265452179571, "grad_norm": 1.5664755519450622, "learning_rate": 8.426514224978684e-06, "loss": 0.4016, "step": 3079 }, { "epoch": 1.0019518542615484, "grad_norm": 1.9078624543267761, "learning_rate": 8.425135566064562e-06, "loss": 0.3931, "step": 3080 }, { "epoch": 1.00227716330514, "grad_norm": 1.7192907036522127, "learning_rate": 8.42375641632623e-06, "loss": 0.4602, "step": 3081 }, { "epoch": 1.0026024723487312, "grad_norm": 2.0541255287440423, "learning_rate": 8.42237677596132e-06, "loss": 0.4351, "step": 3082 }, { "epoch": 1.0029277813923227, "grad_norm": 1.7547446332805428, "learning_rate": 8.420996645167539e-06, "loss": 0.3463, "step": 3083 }, { "epoch": 1.003253090435914, "grad_norm": 2.3388798774888744, "learning_rate": 8.419616024142655e-06, "loss": 0.5383, "step": 3084 }, { "epoch": 1.0035783994795056, "grad_norm": 2.0956788272956786, "learning_rate": 8.418234913084515e-06, "loss": 0.4319, "step": 3085 }, { "epoch": 1.0039037085230968, "grad_norm": 2.124690621652687, "learning_rate": 8.416853312191034e-06, "loss": 0.4736, "step": 3086 }, { "epoch": 1.0042290175666884, "grad_norm": 1.8768355114662771, "learning_rate": 8.415471221660194e-06, "loss": 0.3878, "step": 3087 }, { "epoch": 1.0045543266102797, "grad_norm": 1.8909226408013782, "learning_rate": 8.414088641690052e-06, "loss": 0.4472, "step": 3088 }, { "epoch": 1.0048796356538712, "grad_norm": 1.5809853229182345, "learning_rate": 8.41270557247873e-06, "loss": 0.3636, "step": 3089 }, { "epoch": 1.0052049446974627, "grad_norm": 1.6119601655908598, "learning_rate": 8.411322014224426e-06, "loss": 0.3439, "step": 3090 }, { "epoch": 1.005530253741054, "grad_norm": 1.6485514708009856, "learning_rate": 8.409937967125402e-06, "loss": 0.373, "step": 3091 }, { "epoch": 1.0058555627846455, "grad_norm": 1.6862684295796286, "learning_rate": 8.408553431379995e-06, "loss": 0.4033, "step": 3092 }, { "epoch": 1.0061808718282368, "grad_norm": 1.8015615105855793, "learning_rate": 8.407168407186607e-06, "loss": 0.3934, "step": 3093 }, { "epoch": 1.0065061808718283, "grad_norm": 1.9160203486108625, "learning_rate": 8.405782894743715e-06, "loss": 0.3742, "step": 3094 }, { "epoch": 1.0068314899154196, "grad_norm": 1.723914887118715, "learning_rate": 8.404396894249863e-06, "loss": 0.3609, "step": 3095 }, { "epoch": 1.0071567989590111, "grad_norm": 1.9400258231547702, "learning_rate": 8.403010405903667e-06, "loss": 0.3746, "step": 3096 }, { "epoch": 1.0074821080026024, "grad_norm": 1.8570045896600644, "learning_rate": 8.40162342990381e-06, "loss": 0.4066, "step": 3097 }, { "epoch": 1.007807417046194, "grad_norm": 2.0018213022202405, "learning_rate": 8.400235966449048e-06, "loss": 0.4055, "step": 3098 }, { "epoch": 1.0081327260897852, "grad_norm": 1.8927305803489507, "learning_rate": 8.398848015738205e-06, "loss": 0.4591, "step": 3099 }, { "epoch": 1.0084580351333767, "grad_norm": 1.8071329376887733, "learning_rate": 8.397459577970175e-06, "loss": 0.3762, "step": 3100 }, { "epoch": 1.008783344176968, "grad_norm": 1.7277303221190117, "learning_rate": 8.396070653343924e-06, "loss": 0.3557, "step": 3101 }, { "epoch": 1.0091086532205595, "grad_norm": 1.7305963371871684, "learning_rate": 8.394681242058481e-06, "loss": 0.3633, "step": 3102 }, { "epoch": 1.009433962264151, "grad_norm": 1.7143036627739328, "learning_rate": 8.393291344312957e-06, "loss": 0.3382, "step": 3103 }, { "epoch": 1.0097592713077423, "grad_norm": 1.740186888316746, "learning_rate": 8.391900960306518e-06, "loss": 0.3456, "step": 3104 }, { "epoch": 1.0100845803513339, "grad_norm": 2.232745038455019, "learning_rate": 8.39051009023841e-06, "loss": 0.47, "step": 3105 }, { "epoch": 1.0104098893949252, "grad_norm": 1.8227814317021014, "learning_rate": 8.389118734307946e-06, "loss": 0.4326, "step": 3106 }, { "epoch": 1.0107351984385167, "grad_norm": 1.879335776378683, "learning_rate": 8.387726892714507e-06, "loss": 0.4082, "step": 3107 }, { "epoch": 1.011060507482108, "grad_norm": 1.872082235480213, "learning_rate": 8.386334565657546e-06, "loss": 0.4606, "step": 3108 }, { "epoch": 1.0113858165256995, "grad_norm": 1.5919989432604393, "learning_rate": 8.384941753336586e-06, "loss": 0.3682, "step": 3109 }, { "epoch": 1.0117111255692908, "grad_norm": 1.965390690221655, "learning_rate": 8.383548455951213e-06, "loss": 0.4354, "step": 3110 }, { "epoch": 1.0120364346128823, "grad_norm": 1.8847462132947579, "learning_rate": 8.382154673701092e-06, "loss": 0.4517, "step": 3111 }, { "epoch": 1.0123617436564736, "grad_norm": 1.6840689023652025, "learning_rate": 8.38076040678595e-06, "loss": 0.3454, "step": 3112 }, { "epoch": 1.012687052700065, "grad_norm": 1.875945591465394, "learning_rate": 8.37936565540559e-06, "loss": 0.4125, "step": 3113 }, { "epoch": 1.0130123617436564, "grad_norm": 1.770877563706732, "learning_rate": 8.377970419759875e-06, "loss": 0.3781, "step": 3114 }, { "epoch": 1.013337670787248, "grad_norm": 2.183250549313216, "learning_rate": 8.376574700048749e-06, "loss": 0.3771, "step": 3115 }, { "epoch": 1.0136629798308392, "grad_norm": 1.80361928183861, "learning_rate": 8.375178496472216e-06, "loss": 0.3866, "step": 3116 }, { "epoch": 1.0139882888744307, "grad_norm": 1.660654403879758, "learning_rate": 8.373781809230355e-06, "loss": 0.316, "step": 3117 }, { "epoch": 1.0143135979180222, "grad_norm": 1.7455809127797182, "learning_rate": 8.372384638523311e-06, "loss": 0.3456, "step": 3118 }, { "epoch": 1.0146389069616135, "grad_norm": 1.816736773395219, "learning_rate": 8.370986984551301e-06, "loss": 0.3749, "step": 3119 }, { "epoch": 1.014964216005205, "grad_norm": 1.8810055384442554, "learning_rate": 8.369588847514607e-06, "loss": 0.4173, "step": 3120 }, { "epoch": 1.0152895250487963, "grad_norm": 1.833480738777907, "learning_rate": 8.368190227613586e-06, "loss": 0.4123, "step": 3121 }, { "epoch": 1.0156148340923878, "grad_norm": 1.7488806970453863, "learning_rate": 8.36679112504866e-06, "loss": 0.358, "step": 3122 }, { "epoch": 1.0159401431359791, "grad_norm": 2.005785885475764, "learning_rate": 8.365391540020318e-06, "loss": 0.3913, "step": 3123 }, { "epoch": 1.0162654521795706, "grad_norm": 1.9717712642695695, "learning_rate": 8.363991472729127e-06, "loss": 0.4416, "step": 3124 }, { "epoch": 1.016590761223162, "grad_norm": 1.644319074902569, "learning_rate": 8.362590923375715e-06, "loss": 0.3836, "step": 3125 }, { "epoch": 1.0169160702667535, "grad_norm": 1.987155520790856, "learning_rate": 8.361189892160781e-06, "loss": 0.4267, "step": 3126 }, { "epoch": 1.0172413793103448, "grad_norm": 1.8098661473599038, "learning_rate": 8.359788379285094e-06, "loss": 0.3592, "step": 3127 }, { "epoch": 1.0175666883539363, "grad_norm": 1.7241499924177002, "learning_rate": 8.358386384949493e-06, "loss": 0.3576, "step": 3128 }, { "epoch": 1.0178919973975276, "grad_norm": 1.886593187119879, "learning_rate": 8.356983909354884e-06, "loss": 0.4042, "step": 3129 }, { "epoch": 1.018217306441119, "grad_norm": 1.7860280942670421, "learning_rate": 8.355580952702242e-06, "loss": 0.409, "step": 3130 }, { "epoch": 1.0185426154847104, "grad_norm": 2.1016140468282525, "learning_rate": 8.354177515192612e-06, "loss": 0.484, "step": 3131 }, { "epoch": 1.0188679245283019, "grad_norm": 1.8841880505353232, "learning_rate": 8.352773597027109e-06, "loss": 0.3999, "step": 3132 }, { "epoch": 1.0191932335718934, "grad_norm": 1.9845699364021028, "learning_rate": 8.351369198406913e-06, "loss": 0.4914, "step": 3133 }, { "epoch": 1.0195185426154847, "grad_norm": 1.9973988944288756, "learning_rate": 8.349964319533277e-06, "loss": 0.4566, "step": 3134 }, { "epoch": 1.0198438516590762, "grad_norm": 1.7625847026152526, "learning_rate": 8.34855896060752e-06, "loss": 0.4305, "step": 3135 }, { "epoch": 1.0201691607026675, "grad_norm": 1.7442618603645548, "learning_rate": 8.347153121831032e-06, "loss": 0.3798, "step": 3136 }, { "epoch": 1.020494469746259, "grad_norm": 1.7127695024780507, "learning_rate": 8.34574680340527e-06, "loss": 0.3568, "step": 3137 }, { "epoch": 1.0208197787898503, "grad_norm": 1.9049412714784477, "learning_rate": 8.344340005531759e-06, "loss": 0.4347, "step": 3138 }, { "epoch": 1.0211450878334418, "grad_norm": 1.8418235059365742, "learning_rate": 8.342932728412097e-06, "loss": 0.4717, "step": 3139 }, { "epoch": 1.0214703968770331, "grad_norm": 1.9905217632420857, "learning_rate": 8.341524972247944e-06, "loss": 0.4694, "step": 3140 }, { "epoch": 1.0217957059206246, "grad_norm": 1.9462348373246565, "learning_rate": 8.34011673724104e-06, "loss": 0.3781, "step": 3141 }, { "epoch": 1.022121014964216, "grad_norm": 1.6495051602856772, "learning_rate": 8.338708023593175e-06, "loss": 0.4084, "step": 3142 }, { "epoch": 1.0224463240078074, "grad_norm": 1.751169799504824, "learning_rate": 8.337298831506227e-06, "loss": 0.3677, "step": 3143 }, { "epoch": 1.0227716330513987, "grad_norm": 1.7399245320934043, "learning_rate": 8.335889161182129e-06, "loss": 0.4027, "step": 3144 }, { "epoch": 1.0230969420949902, "grad_norm": 1.806906228871202, "learning_rate": 8.334479012822892e-06, "loss": 0.3671, "step": 3145 }, { "epoch": 1.0234222511385815, "grad_norm": 1.683892380726615, "learning_rate": 8.33306838663059e-06, "loss": 0.4046, "step": 3146 }, { "epoch": 1.023747560182173, "grad_norm": 1.9984198398160578, "learning_rate": 8.331657282807363e-06, "loss": 0.492, "step": 3147 }, { "epoch": 1.0240728692257646, "grad_norm": 2.246095803146014, "learning_rate": 8.330245701555428e-06, "loss": 0.5251, "step": 3148 }, { "epoch": 1.0243981782693559, "grad_norm": 2.0365148048541943, "learning_rate": 8.328833643077064e-06, "loss": 0.3597, "step": 3149 }, { "epoch": 1.0247234873129474, "grad_norm": 1.5313616312832923, "learning_rate": 8.327421107574618e-06, "loss": 0.3097, "step": 3150 }, { "epoch": 1.0250487963565387, "grad_norm": 1.9115400705007004, "learning_rate": 8.326008095250512e-06, "loss": 0.3772, "step": 3151 }, { "epoch": 1.0253741054001302, "grad_norm": 2.070497379290309, "learning_rate": 8.324594606307224e-06, "loss": 0.4092, "step": 3152 }, { "epoch": 1.0256994144437215, "grad_norm": 2.116167724525505, "learning_rate": 8.323180640947314e-06, "loss": 0.4313, "step": 3153 }, { "epoch": 1.026024723487313, "grad_norm": 2.2394114962225045, "learning_rate": 8.321766199373405e-06, "loss": 0.5563, "step": 3154 }, { "epoch": 1.0263500325309043, "grad_norm": 1.9799117485416566, "learning_rate": 8.320351281788182e-06, "loss": 0.3996, "step": 3155 }, { "epoch": 1.0266753415744958, "grad_norm": 2.1492194227960684, "learning_rate": 8.318935888394406e-06, "loss": 0.4587, "step": 3156 }, { "epoch": 1.027000650618087, "grad_norm": 1.6524375709103776, "learning_rate": 8.317520019394907e-06, "loss": 0.3903, "step": 3157 }, { "epoch": 1.0273259596616786, "grad_norm": 1.8576074905477935, "learning_rate": 8.316103674992575e-06, "loss": 0.3959, "step": 3158 }, { "epoch": 1.02765126870527, "grad_norm": 1.7798340749964696, "learning_rate": 8.314686855390378e-06, "loss": 0.3561, "step": 3159 }, { "epoch": 1.0279765777488614, "grad_norm": 1.7536131135224502, "learning_rate": 8.313269560791343e-06, "loss": 0.4605, "step": 3160 }, { "epoch": 1.028301886792453, "grad_norm": 2.0816217811416458, "learning_rate": 8.311851791398571e-06, "loss": 0.4365, "step": 3161 }, { "epoch": 1.0286271958360442, "grad_norm": 1.6893331672174268, "learning_rate": 8.310433547415228e-06, "loss": 0.383, "step": 3162 }, { "epoch": 1.0289525048796357, "grad_norm": 1.5383387714906218, "learning_rate": 8.309014829044552e-06, "loss": 0.3656, "step": 3163 }, { "epoch": 1.029277813923227, "grad_norm": 1.8177276495355834, "learning_rate": 8.307595636489845e-06, "loss": 0.3628, "step": 3164 }, { "epoch": 1.0296031229668186, "grad_norm": 2.057742623033041, "learning_rate": 8.306175969954477e-06, "loss": 0.4854, "step": 3165 }, { "epoch": 1.0299284320104098, "grad_norm": 2.258106476382871, "learning_rate": 8.304755829641888e-06, "loss": 0.4864, "step": 3166 }, { "epoch": 1.0302537410540014, "grad_norm": 1.567880052232512, "learning_rate": 8.303335215755587e-06, "loss": 0.3564, "step": 3167 }, { "epoch": 1.0305790500975927, "grad_norm": 1.5860011320932144, "learning_rate": 8.301914128499146e-06, "loss": 0.3602, "step": 3168 }, { "epoch": 1.0309043591411842, "grad_norm": 1.695342107655222, "learning_rate": 8.300492568076208e-06, "loss": 0.3537, "step": 3169 }, { "epoch": 1.0312296681847755, "grad_norm": 2.0362879212802634, "learning_rate": 8.299070534690486e-06, "loss": 0.4124, "step": 3170 }, { "epoch": 1.031554977228367, "grad_norm": 1.7172156470106668, "learning_rate": 8.297648028545755e-06, "loss": 0.3771, "step": 3171 }, { "epoch": 1.0318802862719583, "grad_norm": 2.314243776709592, "learning_rate": 8.296225049845865e-06, "loss": 0.5032, "step": 3172 }, { "epoch": 1.0322055953155498, "grad_norm": 1.7909831020498352, "learning_rate": 8.294801598794728e-06, "loss": 0.359, "step": 3173 }, { "epoch": 1.032530904359141, "grad_norm": 1.6190956416035331, "learning_rate": 8.293377675596323e-06, "loss": 0.3461, "step": 3174 }, { "epoch": 1.0328562134027326, "grad_norm": 1.771073447714695, "learning_rate": 8.291953280454702e-06, "loss": 0.3931, "step": 3175 }, { "epoch": 1.033181522446324, "grad_norm": 1.5961137466176565, "learning_rate": 8.290528413573982e-06, "loss": 0.3459, "step": 3176 }, { "epoch": 1.0335068314899154, "grad_norm": 1.6393627931132109, "learning_rate": 8.289103075158345e-06, "loss": 0.3933, "step": 3177 }, { "epoch": 1.033832140533507, "grad_norm": 2.131921876662202, "learning_rate": 8.287677265412044e-06, "loss": 0.421, "step": 3178 }, { "epoch": 1.0341574495770982, "grad_norm": 1.8815367521039106, "learning_rate": 8.2862509845394e-06, "loss": 0.4012, "step": 3179 }, { "epoch": 1.0344827586206897, "grad_norm": 1.6368689221809072, "learning_rate": 8.284824232744799e-06, "loss": 0.3099, "step": 3180 }, { "epoch": 1.034808067664281, "grad_norm": 2.090438574178612, "learning_rate": 8.283397010232696e-06, "loss": 0.4593, "step": 3181 }, { "epoch": 1.0351333767078725, "grad_norm": 1.9752357939949734, "learning_rate": 8.281969317207611e-06, "loss": 0.4345, "step": 3182 }, { "epoch": 1.0354586857514638, "grad_norm": 1.8963493999277168, "learning_rate": 8.280541153874135e-06, "loss": 0.4518, "step": 3183 }, { "epoch": 1.0357839947950553, "grad_norm": 1.7001388309793157, "learning_rate": 8.279112520436926e-06, "loss": 0.4061, "step": 3184 }, { "epoch": 1.0361093038386466, "grad_norm": 2.015314058367259, "learning_rate": 8.277683417100704e-06, "loss": 0.4412, "step": 3185 }, { "epoch": 1.0364346128822381, "grad_norm": 1.703407306171444, "learning_rate": 8.276253844070264e-06, "loss": 0.3596, "step": 3186 }, { "epoch": 1.0367599219258294, "grad_norm": 1.9949383919111265, "learning_rate": 8.274823801550465e-06, "loss": 0.4043, "step": 3187 }, { "epoch": 1.037085230969421, "grad_norm": 2.160013079653597, "learning_rate": 8.27339328974623e-06, "loss": 0.5016, "step": 3188 }, { "epoch": 1.0374105400130125, "grad_norm": 1.7658579687295513, "learning_rate": 8.271962308862554e-06, "loss": 0.4015, "step": 3189 }, { "epoch": 1.0377358490566038, "grad_norm": 1.661956076557708, "learning_rate": 8.270530859104499e-06, "loss": 0.3457, "step": 3190 }, { "epoch": 1.0380611581001953, "grad_norm": 1.75568902118036, "learning_rate": 8.269098940677191e-06, "loss": 0.3979, "step": 3191 }, { "epoch": 1.0383864671437866, "grad_norm": 2.001619116671795, "learning_rate": 8.267666553785825e-06, "loss": 0.38, "step": 3192 }, { "epoch": 1.038711776187378, "grad_norm": 1.6507660824148453, "learning_rate": 8.266233698635662e-06, "loss": 0.3429, "step": 3193 }, { "epoch": 1.0390370852309694, "grad_norm": 1.6372616754678317, "learning_rate": 8.264800375432034e-06, "loss": 0.2913, "step": 3194 }, { "epoch": 1.039362394274561, "grad_norm": 2.08749720902689, "learning_rate": 8.263366584380334e-06, "loss": 0.4698, "step": 3195 }, { "epoch": 1.0396877033181522, "grad_norm": 1.7318510821633941, "learning_rate": 8.261932325686029e-06, "loss": 0.3882, "step": 3196 }, { "epoch": 1.0400130123617437, "grad_norm": 1.7996367900010841, "learning_rate": 8.260497599554645e-06, "loss": 0.4219, "step": 3197 }, { "epoch": 1.040338321405335, "grad_norm": 1.7806214383161105, "learning_rate": 8.259062406191781e-06, "loss": 0.4011, "step": 3198 }, { "epoch": 1.0406636304489265, "grad_norm": 1.6274187720106048, "learning_rate": 8.257626745803103e-06, "loss": 0.2607, "step": 3199 }, { "epoch": 1.0409889394925178, "grad_norm": 1.7458171405568212, "learning_rate": 8.256190618594338e-06, "loss": 0.399, "step": 3200 }, { "epoch": 1.0413142485361093, "grad_norm": 2.139564398784938, "learning_rate": 8.254754024771287e-06, "loss": 0.4937, "step": 3201 }, { "epoch": 1.0416395575797006, "grad_norm": 1.7695581606891628, "learning_rate": 8.253316964539815e-06, "loss": 0.3301, "step": 3202 }, { "epoch": 1.0419648666232921, "grad_norm": 1.8949341678825427, "learning_rate": 8.251879438105854e-06, "loss": 0.3719, "step": 3203 }, { "epoch": 1.0422901756668836, "grad_norm": 1.4179774037719146, "learning_rate": 8.2504414456754e-06, "loss": 0.3088, "step": 3204 }, { "epoch": 1.042615484710475, "grad_norm": 1.8117667616416082, "learning_rate": 8.24900298745452e-06, "loss": 0.3634, "step": 3205 }, { "epoch": 1.0429407937540665, "grad_norm": 1.985340878966728, "learning_rate": 8.247564063649344e-06, "loss": 0.3973, "step": 3206 }, { "epoch": 1.0432661027976577, "grad_norm": 1.8962271410152962, "learning_rate": 8.246124674466074e-06, "loss": 0.3833, "step": 3207 }, { "epoch": 1.0435914118412493, "grad_norm": 1.836700354380386, "learning_rate": 8.244684820110974e-06, "loss": 0.4239, "step": 3208 }, { "epoch": 1.0439167208848406, "grad_norm": 1.8127794321579735, "learning_rate": 8.243244500790376e-06, "loss": 0.467, "step": 3209 }, { "epoch": 1.044242029928432, "grad_norm": 2.1212727703761933, "learning_rate": 8.241803716710678e-06, "loss": 0.3961, "step": 3210 }, { "epoch": 1.0445673389720234, "grad_norm": 1.993604537806533, "learning_rate": 8.240362468078346e-06, "loss": 0.3635, "step": 3211 }, { "epoch": 1.0448926480156149, "grad_norm": 1.8919151841723463, "learning_rate": 8.238920755099913e-06, "loss": 0.405, "step": 3212 }, { "epoch": 1.0452179570592062, "grad_norm": 1.987133571465202, "learning_rate": 8.237478577981976e-06, "loss": 0.4681, "step": 3213 }, { "epoch": 1.0455432661027977, "grad_norm": 1.6150033474269094, "learning_rate": 8.2360359369312e-06, "loss": 0.3294, "step": 3214 }, { "epoch": 1.045868575146389, "grad_norm": 1.9242075501594313, "learning_rate": 8.234592832154316e-06, "loss": 0.4351, "step": 3215 }, { "epoch": 1.0461938841899805, "grad_norm": 2.025840100535751, "learning_rate": 8.233149263858122e-06, "loss": 0.3618, "step": 3216 }, { "epoch": 1.0465191932335718, "grad_norm": 1.8134770935552884, "learning_rate": 8.231705232249485e-06, "loss": 0.4181, "step": 3217 }, { "epoch": 1.0468445022771633, "grad_norm": 2.01488838935879, "learning_rate": 8.230260737535332e-06, "loss": 0.4122, "step": 3218 }, { "epoch": 1.0471698113207548, "grad_norm": 1.876805450719626, "learning_rate": 8.228815779922662e-06, "loss": 0.419, "step": 3219 }, { "epoch": 1.047495120364346, "grad_norm": 1.8279837637107461, "learning_rate": 8.227370359618539e-06, "loss": 0.5285, "step": 3220 }, { "epoch": 1.0478204294079376, "grad_norm": 2.1727660359139804, "learning_rate": 8.225924476830089e-06, "loss": 0.4788, "step": 3221 }, { "epoch": 1.048145738451529, "grad_norm": 1.8371058600693249, "learning_rate": 8.224478131764513e-06, "loss": 0.417, "step": 3222 }, { "epoch": 1.0484710474951204, "grad_norm": 1.9304285447087792, "learning_rate": 8.223031324629069e-06, "loss": 0.4077, "step": 3223 }, { "epoch": 1.0487963565387117, "grad_norm": 1.5305356194390036, "learning_rate": 8.221584055631088e-06, "loss": 0.3334, "step": 3224 }, { "epoch": 1.0491216655823032, "grad_norm": 1.4360643242793347, "learning_rate": 8.220136324977966e-06, "loss": 0.3336, "step": 3225 }, { "epoch": 1.0494469746258945, "grad_norm": 2.011785574693979, "learning_rate": 8.218688132877159e-06, "loss": 0.3861, "step": 3226 }, { "epoch": 1.049772283669486, "grad_norm": 1.971781683345634, "learning_rate": 8.217239479536199e-06, "loss": 0.4407, "step": 3227 }, { "epoch": 1.0500975927130773, "grad_norm": 2.0045530305259924, "learning_rate": 8.215790365162674e-06, "loss": 0.4062, "step": 3228 }, { "epoch": 1.0504229017566689, "grad_norm": 2.0840411157069347, "learning_rate": 8.214340789964246e-06, "loss": 0.4548, "step": 3229 }, { "epoch": 1.0507482108002602, "grad_norm": 1.8702934536439089, "learning_rate": 8.21289075414864e-06, "loss": 0.4079, "step": 3230 }, { "epoch": 1.0510735198438517, "grad_norm": 2.0765260426371412, "learning_rate": 8.211440257923648e-06, "loss": 0.4077, "step": 3231 }, { "epoch": 1.051398828887443, "grad_norm": 2.039098128607094, "learning_rate": 8.209989301497125e-06, "loss": 0.3739, "step": 3232 }, { "epoch": 1.0517241379310345, "grad_norm": 1.4458418535562256, "learning_rate": 8.208537885076996e-06, "loss": 0.2956, "step": 3233 }, { "epoch": 1.052049446974626, "grad_norm": 1.9251671714416452, "learning_rate": 8.207086008871249e-06, "loss": 0.5157, "step": 3234 }, { "epoch": 1.0523747560182173, "grad_norm": 1.7514728589004143, "learning_rate": 8.205633673087937e-06, "loss": 0.3273, "step": 3235 }, { "epoch": 1.0527000650618088, "grad_norm": 1.7999579714117726, "learning_rate": 8.204180877935183e-06, "loss": 0.4414, "step": 3236 }, { "epoch": 1.0530253741054, "grad_norm": 1.652814377493338, "learning_rate": 8.202727623621171e-06, "loss": 0.2945, "step": 3237 }, { "epoch": 1.0533506831489916, "grad_norm": 1.9683115142291672, "learning_rate": 8.201273910354159e-06, "loss": 0.4399, "step": 3238 }, { "epoch": 1.053675992192583, "grad_norm": 1.7854817009138075, "learning_rate": 8.199819738342458e-06, "loss": 0.4464, "step": 3239 }, { "epoch": 1.0540013012361744, "grad_norm": 1.8369354669217153, "learning_rate": 8.198365107794457e-06, "loss": 0.4358, "step": 3240 }, { "epoch": 1.0543266102797657, "grad_norm": 1.9962997420133344, "learning_rate": 8.196910018918603e-06, "loss": 0.4447, "step": 3241 }, { "epoch": 1.0546519193233572, "grad_norm": 2.1697071552472034, "learning_rate": 8.195454471923413e-06, "loss": 0.4808, "step": 3242 }, { "epoch": 1.0549772283669485, "grad_norm": 1.6080492737099767, "learning_rate": 8.193998467017463e-06, "loss": 0.3143, "step": 3243 }, { "epoch": 1.05530253741054, "grad_norm": 1.7189163390722224, "learning_rate": 8.192542004409408e-06, "loss": 0.3897, "step": 3244 }, { "epoch": 1.0556278464541313, "grad_norm": 2.248437425441394, "learning_rate": 8.191085084307951e-06, "loss": 0.5203, "step": 3245 }, { "epoch": 1.0559531554977228, "grad_norm": 1.9842803659977628, "learning_rate": 8.189627706921876e-06, "loss": 0.4312, "step": 3246 }, { "epoch": 1.0562784645413144, "grad_norm": 1.8644091068544528, "learning_rate": 8.188169872460023e-06, "loss": 0.4618, "step": 3247 }, { "epoch": 1.0566037735849056, "grad_norm": 1.6733978672636507, "learning_rate": 8.186711581131303e-06, "loss": 0.3608, "step": 3248 }, { "epoch": 1.0569290826284972, "grad_norm": 1.6722205781260655, "learning_rate": 8.185252833144688e-06, "loss": 0.3429, "step": 3249 }, { "epoch": 1.0572543916720885, "grad_norm": 1.7491834873954561, "learning_rate": 8.183793628709218e-06, "loss": 0.368, "step": 3250 }, { "epoch": 1.05757970071568, "grad_norm": 1.6428898859458714, "learning_rate": 8.182333968033999e-06, "loss": 0.352, "step": 3251 }, { "epoch": 1.0579050097592713, "grad_norm": 1.8707934984618175, "learning_rate": 8.1808738513282e-06, "loss": 0.4353, "step": 3252 }, { "epoch": 1.0582303188028628, "grad_norm": 1.7496229756265722, "learning_rate": 8.179413278801058e-06, "loss": 0.3898, "step": 3253 }, { "epoch": 1.058555627846454, "grad_norm": 1.7395467399084592, "learning_rate": 8.177952250661873e-06, "loss": 0.3958, "step": 3254 }, { "epoch": 1.0588809368900456, "grad_norm": 1.7702197486026165, "learning_rate": 8.176490767120011e-06, "loss": 0.3231, "step": 3255 }, { "epoch": 1.0592062459336369, "grad_norm": 2.079342657550188, "learning_rate": 8.175028828384904e-06, "loss": 0.3603, "step": 3256 }, { "epoch": 1.0595315549772284, "grad_norm": 2.321111008849438, "learning_rate": 8.173566434666049e-06, "loss": 0.4066, "step": 3257 }, { "epoch": 1.0598568640208197, "grad_norm": 1.9224256296420177, "learning_rate": 8.172103586173011e-06, "loss": 0.4004, "step": 3258 }, { "epoch": 1.0601821730644112, "grad_norm": 1.9725834532336641, "learning_rate": 8.17064028311541e-06, "loss": 0.3785, "step": 3259 }, { "epoch": 1.0605074821080025, "grad_norm": 2.174579999000839, "learning_rate": 8.169176525702944e-06, "loss": 0.4767, "step": 3260 }, { "epoch": 1.060832791151594, "grad_norm": 1.8464321751052466, "learning_rate": 8.167712314145371e-06, "loss": 0.3785, "step": 3261 }, { "epoch": 1.0611581001951855, "grad_norm": 2.1352496981089337, "learning_rate": 8.166247648652508e-06, "loss": 0.5117, "step": 3262 }, { "epoch": 1.0614834092387768, "grad_norm": 1.6173492474958673, "learning_rate": 8.164782529434248e-06, "loss": 0.3555, "step": 3263 }, { "epoch": 1.0618087182823683, "grad_norm": 2.0045173945835337, "learning_rate": 8.163316956700541e-06, "loss": 0.4657, "step": 3264 }, { "epoch": 1.0621340273259596, "grad_norm": 2.049963086065168, "learning_rate": 8.161850930661405e-06, "loss": 0.431, "step": 3265 }, { "epoch": 1.0624593363695511, "grad_norm": 1.8093117519863247, "learning_rate": 8.160384451526923e-06, "loss": 0.3758, "step": 3266 }, { "epoch": 1.0627846454131424, "grad_norm": 2.1197721858235616, "learning_rate": 8.158917519507242e-06, "loss": 0.4631, "step": 3267 }, { "epoch": 1.063109954456734, "grad_norm": 1.8198394452499809, "learning_rate": 8.157450134812574e-06, "loss": 0.347, "step": 3268 }, { "epoch": 1.0634352635003252, "grad_norm": 2.027980295113629, "learning_rate": 8.155982297653195e-06, "loss": 0.4482, "step": 3269 }, { "epoch": 1.0637605725439168, "grad_norm": 1.8239526263729147, "learning_rate": 8.154514008239452e-06, "loss": 0.4379, "step": 3270 }, { "epoch": 1.064085881587508, "grad_norm": 1.584723360024325, "learning_rate": 8.153045266781746e-06, "loss": 0.3023, "step": 3271 }, { "epoch": 1.0644111906310996, "grad_norm": 1.713419268391497, "learning_rate": 8.15157607349055e-06, "loss": 0.3392, "step": 3272 }, { "epoch": 1.0647364996746909, "grad_norm": 1.76141529429696, "learning_rate": 8.150106428576403e-06, "loss": 0.3777, "step": 3273 }, { "epoch": 1.0650618087182824, "grad_norm": 1.8692037424708137, "learning_rate": 8.148636332249904e-06, "loss": 0.3938, "step": 3274 }, { "epoch": 1.065387117761874, "grad_norm": 2.0066765995073106, "learning_rate": 8.14716578472172e-06, "loss": 0.502, "step": 3275 }, { "epoch": 1.0657124268054652, "grad_norm": 2.302937686624616, "learning_rate": 8.14569478620258e-06, "loss": 0.3659, "step": 3276 }, { "epoch": 1.0660377358490567, "grad_norm": 1.7087633038251875, "learning_rate": 8.144223336903281e-06, "loss": 0.3705, "step": 3277 }, { "epoch": 1.066363044892648, "grad_norm": 1.722109477008629, "learning_rate": 8.142751437034681e-06, "loss": 0.4271, "step": 3278 }, { "epoch": 1.0666883539362395, "grad_norm": 1.652961608602287, "learning_rate": 8.141279086807703e-06, "loss": 0.3523, "step": 3279 }, { "epoch": 1.0670136629798308, "grad_norm": 1.8124805548118603, "learning_rate": 8.13980628643334e-06, "loss": 0.4119, "step": 3280 }, { "epoch": 1.0673389720234223, "grad_norm": 1.6931904551405068, "learning_rate": 8.138333036122642e-06, "loss": 0.374, "step": 3281 }, { "epoch": 1.0676642810670136, "grad_norm": 1.7353300440619972, "learning_rate": 8.136859336086727e-06, "loss": 0.4732, "step": 3282 }, { "epoch": 1.0679895901106051, "grad_norm": 1.8669155183621222, "learning_rate": 8.135385186536778e-06, "loss": 0.4037, "step": 3283 }, { "epoch": 1.0683148991541964, "grad_norm": 1.72804880258719, "learning_rate": 8.13391058768404e-06, "loss": 0.3861, "step": 3284 }, { "epoch": 1.068640208197788, "grad_norm": 1.9686692219795416, "learning_rate": 8.132435539739825e-06, "loss": 0.4363, "step": 3285 }, { "epoch": 1.0689655172413792, "grad_norm": 1.550039799986217, "learning_rate": 8.130960042915512e-06, "loss": 0.361, "step": 3286 }, { "epoch": 1.0692908262849707, "grad_norm": 2.1110796908251044, "learning_rate": 8.129484097422535e-06, "loss": 0.4616, "step": 3287 }, { "epoch": 1.069616135328562, "grad_norm": 1.8946524690648952, "learning_rate": 8.128007703472401e-06, "loss": 0.3336, "step": 3288 }, { "epoch": 1.0699414443721535, "grad_norm": 2.037045298503493, "learning_rate": 8.126530861276677e-06, "loss": 0.4224, "step": 3289 }, { "epoch": 1.070266753415745, "grad_norm": 1.6941765483914508, "learning_rate": 8.125053571046997e-06, "loss": 0.3162, "step": 3290 }, { "epoch": 1.0705920624593364, "grad_norm": 1.9279463888942108, "learning_rate": 8.123575832995055e-06, "loss": 0.4304, "step": 3291 }, { "epoch": 1.0709173715029279, "grad_norm": 1.9890053373544898, "learning_rate": 8.122097647332614e-06, "loss": 0.3814, "step": 3292 }, { "epoch": 1.0712426805465192, "grad_norm": 1.956586950940869, "learning_rate": 8.1206190142715e-06, "loss": 0.4238, "step": 3293 }, { "epoch": 1.0715679895901107, "grad_norm": 1.8760604493428124, "learning_rate": 8.119139934023598e-06, "loss": 0.3917, "step": 3294 }, { "epoch": 1.071893298633702, "grad_norm": 1.9643912606361504, "learning_rate": 8.117660406800867e-06, "loss": 0.4841, "step": 3295 }, { "epoch": 1.0722186076772935, "grad_norm": 1.9646854953756039, "learning_rate": 8.116180432815319e-06, "loss": 0.4417, "step": 3296 }, { "epoch": 1.0725439167208848, "grad_norm": 1.7924219188236652, "learning_rate": 8.114700012279039e-06, "loss": 0.3407, "step": 3297 }, { "epoch": 1.0728692257644763, "grad_norm": 2.1992697113483084, "learning_rate": 8.11321914540417e-06, "loss": 0.3722, "step": 3298 }, { "epoch": 1.0731945348080676, "grad_norm": 1.8680249949857415, "learning_rate": 8.11173783240292e-06, "loss": 0.4738, "step": 3299 }, { "epoch": 1.073519843851659, "grad_norm": 1.9076312251388394, "learning_rate": 8.110256073487566e-06, "loss": 0.4173, "step": 3300 }, { "epoch": 1.0738451528952504, "grad_norm": 1.9022524579491187, "learning_rate": 8.108773868870441e-06, "loss": 0.4657, "step": 3301 }, { "epoch": 1.074170461938842, "grad_norm": 2.0453275607504047, "learning_rate": 8.107291218763951e-06, "loss": 0.4499, "step": 3302 }, { "epoch": 1.0744957709824332, "grad_norm": 2.3044527074682857, "learning_rate": 8.105808123380554e-06, "loss": 0.4818, "step": 3303 }, { "epoch": 1.0748210800260247, "grad_norm": 1.9220577654234239, "learning_rate": 8.104324582932786e-06, "loss": 0.3768, "step": 3304 }, { "epoch": 1.0751463890696162, "grad_norm": 2.0874815290081754, "learning_rate": 8.102840597633233e-06, "loss": 0.4436, "step": 3305 }, { "epoch": 1.0754716981132075, "grad_norm": 1.8838804826871693, "learning_rate": 8.101356167694554e-06, "loss": 0.3452, "step": 3306 }, { "epoch": 1.075797007156799, "grad_norm": 1.977412681951198, "learning_rate": 8.099871293329469e-06, "loss": 0.4301, "step": 3307 }, { "epoch": 1.0761223162003903, "grad_norm": 2.1456205443655887, "learning_rate": 8.09838597475076e-06, "loss": 0.4507, "step": 3308 }, { "epoch": 1.0764476252439819, "grad_norm": 2.203344897021463, "learning_rate": 8.096900212171274e-06, "loss": 0.3913, "step": 3309 }, { "epoch": 1.0767729342875731, "grad_norm": 1.7038879330401226, "learning_rate": 8.095414005803926e-06, "loss": 0.3997, "step": 3310 }, { "epoch": 1.0770982433311647, "grad_norm": 1.6274979391214746, "learning_rate": 8.093927355861683e-06, "loss": 0.315, "step": 3311 }, { "epoch": 1.077423552374756, "grad_norm": 1.9356581728292364, "learning_rate": 8.092440262557591e-06, "loss": 0.3984, "step": 3312 }, { "epoch": 1.0777488614183475, "grad_norm": 2.191428830420603, "learning_rate": 8.090952726104747e-06, "loss": 0.4589, "step": 3313 }, { "epoch": 1.0780741704619388, "grad_norm": 1.755256763972341, "learning_rate": 8.089464746716314e-06, "loss": 0.3644, "step": 3314 }, { "epoch": 1.0783994795055303, "grad_norm": 1.8599751644965523, "learning_rate": 8.087976324605525e-06, "loss": 0.3967, "step": 3315 }, { "epoch": 1.0787247885491216, "grad_norm": 1.6600743138098097, "learning_rate": 8.086487459985671e-06, "loss": 0.379, "step": 3316 }, { "epoch": 1.079050097592713, "grad_norm": 2.1746809713759836, "learning_rate": 8.084998153070103e-06, "loss": 0.5098, "step": 3317 }, { "epoch": 1.0793754066363044, "grad_norm": 2.0057222382187816, "learning_rate": 8.083508404072244e-06, "loss": 0.3604, "step": 3318 }, { "epoch": 1.079700715679896, "grad_norm": 1.815558975741808, "learning_rate": 8.082018213205576e-06, "loss": 0.4004, "step": 3319 }, { "epoch": 1.0800260247234874, "grad_norm": 1.7758816333856673, "learning_rate": 8.08052758068364e-06, "loss": 0.4393, "step": 3320 }, { "epoch": 1.0803513337670787, "grad_norm": 1.9478466037287043, "learning_rate": 8.079036506720051e-06, "loss": 0.4455, "step": 3321 }, { "epoch": 1.0806766428106702, "grad_norm": 2.261401781165103, "learning_rate": 8.077544991528478e-06, "loss": 0.5478, "step": 3322 }, { "epoch": 1.0810019518542615, "grad_norm": 1.6865152760886575, "learning_rate": 8.076053035322655e-06, "loss": 0.277, "step": 3323 }, { "epoch": 1.081327260897853, "grad_norm": 1.9399035173297066, "learning_rate": 8.07456063831638e-06, "loss": 0.4209, "step": 3324 }, { "epoch": 1.0816525699414443, "grad_norm": 1.8721621914607565, "learning_rate": 8.073067800723515e-06, "loss": 0.3748, "step": 3325 }, { "epoch": 1.0819778789850358, "grad_norm": 2.101525755560038, "learning_rate": 8.071574522757987e-06, "loss": 0.4823, "step": 3326 }, { "epoch": 1.0823031880286271, "grad_norm": 2.0630137741343124, "learning_rate": 8.070080804633783e-06, "loss": 0.4277, "step": 3327 }, { "epoch": 1.0826284970722186, "grad_norm": 1.7532450077572144, "learning_rate": 8.06858664656495e-06, "loss": 0.3631, "step": 3328 }, { "epoch": 1.08295380611581, "grad_norm": 1.845519384419718, "learning_rate": 8.067092048765607e-06, "loss": 0.3838, "step": 3329 }, { "epoch": 1.0832791151594015, "grad_norm": 1.758293682863978, "learning_rate": 8.065597011449928e-06, "loss": 0.3975, "step": 3330 }, { "epoch": 1.0836044242029927, "grad_norm": 1.7237476928492188, "learning_rate": 8.064101534832152e-06, "loss": 0.4217, "step": 3331 }, { "epoch": 1.0839297332465843, "grad_norm": 1.9289296099246491, "learning_rate": 8.062605619126585e-06, "loss": 0.4231, "step": 3332 }, { "epoch": 1.0842550422901756, "grad_norm": 1.9695058413867126, "learning_rate": 8.06110926454759e-06, "loss": 0.4671, "step": 3333 }, { "epoch": 1.084580351333767, "grad_norm": 1.8552896040825002, "learning_rate": 8.059612471309594e-06, "loss": 0.3339, "step": 3334 }, { "epoch": 1.0849056603773586, "grad_norm": 1.8042258404568223, "learning_rate": 8.058115239627094e-06, "loss": 0.3608, "step": 3335 }, { "epoch": 1.0852309694209499, "grad_norm": 2.1176464761724945, "learning_rate": 8.056617569714641e-06, "loss": 0.4239, "step": 3336 }, { "epoch": 1.0855562784645414, "grad_norm": 1.9103907206411355, "learning_rate": 8.05511946178685e-06, "loss": 0.2865, "step": 3337 }, { "epoch": 1.0858815875081327, "grad_norm": 1.8932740214295525, "learning_rate": 8.053620916058407e-06, "loss": 0.4107, "step": 3338 }, { "epoch": 1.0862068965517242, "grad_norm": 2.256984656803612, "learning_rate": 8.052121932744048e-06, "loss": 0.4984, "step": 3339 }, { "epoch": 1.0865322055953155, "grad_norm": 2.0409307116929982, "learning_rate": 8.050622512058582e-06, "loss": 0.4375, "step": 3340 }, { "epoch": 1.086857514638907, "grad_norm": 1.9617262880587192, "learning_rate": 8.049122654216877e-06, "loss": 0.3857, "step": 3341 }, { "epoch": 1.0871828236824983, "grad_norm": 1.9390891309682485, "learning_rate": 8.047622359433861e-06, "loss": 0.3984, "step": 3342 }, { "epoch": 1.0875081327260898, "grad_norm": 2.0659485559167745, "learning_rate": 8.04612162792453e-06, "loss": 0.423, "step": 3343 }, { "epoch": 1.087833441769681, "grad_norm": 1.7602634507653208, "learning_rate": 8.044620459903937e-06, "loss": 0.4494, "step": 3344 }, { "epoch": 1.0881587508132726, "grad_norm": 1.8160930949254124, "learning_rate": 8.043118855587204e-06, "loss": 0.3546, "step": 3345 }, { "epoch": 1.088484059856864, "grad_norm": 1.9857945853849484, "learning_rate": 8.04161681518951e-06, "loss": 0.4124, "step": 3346 }, { "epoch": 1.0888093689004554, "grad_norm": 1.6937234086927635, "learning_rate": 8.040114338926098e-06, "loss": 0.3455, "step": 3347 }, { "epoch": 1.089134677944047, "grad_norm": 2.0850748418233027, "learning_rate": 8.038611427012275e-06, "loss": 0.4651, "step": 3348 }, { "epoch": 1.0894599869876382, "grad_norm": 1.961639025234395, "learning_rate": 8.037108079663408e-06, "loss": 0.339, "step": 3349 }, { "epoch": 1.0897852960312298, "grad_norm": 2.2204922368205833, "learning_rate": 8.035604297094928e-06, "loss": 0.4931, "step": 3350 }, { "epoch": 1.090110605074821, "grad_norm": 1.8726549103536068, "learning_rate": 8.034100079522328e-06, "loss": 0.3671, "step": 3351 }, { "epoch": 1.0904359141184126, "grad_norm": 2.1724208595105687, "learning_rate": 8.032595427161165e-06, "loss": 0.408, "step": 3352 }, { "epoch": 1.0907612231620039, "grad_norm": 2.093074276699235, "learning_rate": 8.031090340227054e-06, "loss": 0.4618, "step": 3353 }, { "epoch": 1.0910865322055954, "grad_norm": 2.015375432784331, "learning_rate": 8.029584818935677e-06, "loss": 0.4036, "step": 3354 }, { "epoch": 1.0914118412491867, "grad_norm": 1.8269006625069366, "learning_rate": 8.028078863502775e-06, "loss": 0.3399, "step": 3355 }, { "epoch": 1.0917371502927782, "grad_norm": 1.5079730898449426, "learning_rate": 8.026572474144152e-06, "loss": 0.3421, "step": 3356 }, { "epoch": 1.0920624593363695, "grad_norm": 1.6499609322453586, "learning_rate": 8.025065651075678e-06, "loss": 0.3281, "step": 3357 }, { "epoch": 1.092387768379961, "grad_norm": 2.0624776705264702, "learning_rate": 8.023558394513278e-06, "loss": 0.4452, "step": 3358 }, { "epoch": 1.0927130774235523, "grad_norm": 2.0209950657221003, "learning_rate": 8.022050704672945e-06, "loss": 0.3477, "step": 3359 }, { "epoch": 1.0930383864671438, "grad_norm": 1.9065065834191082, "learning_rate": 8.02054258177073e-06, "loss": 0.3966, "step": 3360 }, { "epoch": 1.0933636955107353, "grad_norm": 1.9575220227716557, "learning_rate": 8.01903402602275e-06, "loss": 0.4221, "step": 3361 }, { "epoch": 1.0936890045543266, "grad_norm": 1.7987538998435817, "learning_rate": 8.017525037645179e-06, "loss": 0.3239, "step": 3362 }, { "epoch": 1.0940143135979181, "grad_norm": 1.9275365350130496, "learning_rate": 8.016015616854261e-06, "loss": 0.4269, "step": 3363 }, { "epoch": 1.0943396226415094, "grad_norm": 1.9493043646322072, "learning_rate": 8.014505763866294e-06, "loss": 0.4404, "step": 3364 }, { "epoch": 1.094664931685101, "grad_norm": 1.8854885497409695, "learning_rate": 8.01299547889764e-06, "loss": 0.4526, "step": 3365 }, { "epoch": 1.0949902407286922, "grad_norm": 1.6260324624775384, "learning_rate": 8.011484762164726e-06, "loss": 0.317, "step": 3366 }, { "epoch": 1.0953155497722837, "grad_norm": 2.1078775010569077, "learning_rate": 8.009973613884038e-06, "loss": 0.4504, "step": 3367 }, { "epoch": 1.095640858815875, "grad_norm": 1.99688318676658, "learning_rate": 8.008462034272123e-06, "loss": 0.4354, "step": 3368 }, { "epoch": 1.0959661678594665, "grad_norm": 1.8901115948912457, "learning_rate": 8.006950023545593e-06, "loss": 0.4238, "step": 3369 }, { "epoch": 1.0962914769030578, "grad_norm": 1.586963945206445, "learning_rate": 8.005437581921119e-06, "loss": 0.3361, "step": 3370 }, { "epoch": 1.0966167859466494, "grad_norm": 1.7653965518428427, "learning_rate": 8.003924709615437e-06, "loss": 0.2618, "step": 3371 }, { "epoch": 1.0969420949902406, "grad_norm": 1.844347330659265, "learning_rate": 8.00241140684534e-06, "loss": 0.2906, "step": 3372 }, { "epoch": 1.0972674040338322, "grad_norm": 1.9239988253172278, "learning_rate": 8.00089767382769e-06, "loss": 0.3902, "step": 3373 }, { "epoch": 1.0975927130774235, "grad_norm": 1.7772016401772615, "learning_rate": 7.999383510779399e-06, "loss": 0.3456, "step": 3374 }, { "epoch": 1.097918022121015, "grad_norm": 1.7902658480915907, "learning_rate": 7.997868917917453e-06, "loss": 0.3545, "step": 3375 }, { "epoch": 1.0982433311646065, "grad_norm": 1.954473683336408, "learning_rate": 7.99635389545889e-06, "loss": 0.3912, "step": 3376 }, { "epoch": 1.0985686402081978, "grad_norm": 1.9946839423671734, "learning_rate": 7.994838443620818e-06, "loss": 0.4155, "step": 3377 }, { "epoch": 1.0988939492517893, "grad_norm": 2.133237547128156, "learning_rate": 7.9933225626204e-06, "loss": 0.4629, "step": 3378 }, { "epoch": 1.0992192582953806, "grad_norm": 2.0081608524986585, "learning_rate": 7.991806252674863e-06, "loss": 0.3963, "step": 3379 }, { "epoch": 1.099544567338972, "grad_norm": 1.8393980451411176, "learning_rate": 7.990289514001495e-06, "loss": 0.4039, "step": 3380 }, { "epoch": 1.0998698763825634, "grad_norm": 1.9650847190246679, "learning_rate": 7.988772346817646e-06, "loss": 0.3606, "step": 3381 }, { "epoch": 1.100195185426155, "grad_norm": 1.9310575945397166, "learning_rate": 7.987254751340728e-06, "loss": 0.3851, "step": 3382 }, { "epoch": 1.1005204944697462, "grad_norm": 1.8219225555511331, "learning_rate": 7.985736727788212e-06, "loss": 0.3944, "step": 3383 }, { "epoch": 1.1008458035133377, "grad_norm": 1.7172315161801786, "learning_rate": 7.984218276377631e-06, "loss": 0.3344, "step": 3384 }, { "epoch": 1.101171112556929, "grad_norm": 1.8244961590000721, "learning_rate": 7.982699397326582e-06, "loss": 0.3484, "step": 3385 }, { "epoch": 1.1014964216005205, "grad_norm": 1.8188437991304405, "learning_rate": 7.981180090852721e-06, "loss": 0.4115, "step": 3386 }, { "epoch": 1.1018217306441118, "grad_norm": 1.9538210287185678, "learning_rate": 7.979660357173765e-06, "loss": 0.3973, "step": 3387 }, { "epoch": 1.1021470396877033, "grad_norm": 1.7446668486711787, "learning_rate": 7.978140196507493e-06, "loss": 0.3447, "step": 3388 }, { "epoch": 1.1024723487312946, "grad_norm": 1.817776307537243, "learning_rate": 7.976619609071746e-06, "loss": 0.3906, "step": 3389 }, { "epoch": 1.1027976577748861, "grad_norm": 2.0862935769954714, "learning_rate": 7.975098595084422e-06, "loss": 0.5227, "step": 3390 }, { "epoch": 1.1031229668184777, "grad_norm": 1.954593425677414, "learning_rate": 7.97357715476349e-06, "loss": 0.4267, "step": 3391 }, { "epoch": 1.103448275862069, "grad_norm": 1.6270955615522478, "learning_rate": 7.972055288326966e-06, "loss": 0.2779, "step": 3392 }, { "epoch": 1.1037735849056605, "grad_norm": 1.9853748236437396, "learning_rate": 7.970532995992939e-06, "loss": 0.4717, "step": 3393 }, { "epoch": 1.1040988939492518, "grad_norm": 1.7124940477714945, "learning_rate": 7.969010277979552e-06, "loss": 0.397, "step": 3394 }, { "epoch": 1.1044242029928433, "grad_norm": 1.939564788964016, "learning_rate": 7.967487134505014e-06, "loss": 0.4054, "step": 3395 }, { "epoch": 1.1047495120364346, "grad_norm": 1.68652040219451, "learning_rate": 7.965963565787589e-06, "loss": 0.36, "step": 3396 }, { "epoch": 1.105074821080026, "grad_norm": 1.830920806569264, "learning_rate": 7.96443957204561e-06, "loss": 0.4073, "step": 3397 }, { "epoch": 1.1054001301236174, "grad_norm": 2.120340527372947, "learning_rate": 7.962915153497461e-06, "loss": 0.4558, "step": 3398 }, { "epoch": 1.105725439167209, "grad_norm": 2.0499883766250546, "learning_rate": 7.961390310361599e-06, "loss": 0.5215, "step": 3399 }, { "epoch": 1.1060507482108002, "grad_norm": 2.1007514536446714, "learning_rate": 7.959865042856527e-06, "loss": 0.4055, "step": 3400 }, { "epoch": 1.1063760572543917, "grad_norm": 1.826873847586324, "learning_rate": 7.958339351200824e-06, "loss": 0.3788, "step": 3401 }, { "epoch": 1.106701366297983, "grad_norm": 1.661780544213811, "learning_rate": 7.95681323561312e-06, "loss": 0.3622, "step": 3402 }, { "epoch": 1.1070266753415745, "grad_norm": 1.9021216957758593, "learning_rate": 7.955286696312104e-06, "loss": 0.4175, "step": 3403 }, { "epoch": 1.1073519843851658, "grad_norm": 1.7297150174454812, "learning_rate": 7.953759733516537e-06, "loss": 0.3646, "step": 3404 }, { "epoch": 1.1076772934287573, "grad_norm": 1.9911769648447888, "learning_rate": 7.95223234744523e-06, "loss": 0.3863, "step": 3405 }, { "epoch": 1.1080026024723488, "grad_norm": 1.6560373878618182, "learning_rate": 7.95070453831706e-06, "loss": 0.3648, "step": 3406 }, { "epoch": 1.1083279115159401, "grad_norm": 1.9961926640209602, "learning_rate": 7.949176306350962e-06, "loss": 0.4053, "step": 3407 }, { "epoch": 1.1086532205595316, "grad_norm": 1.7883019448541335, "learning_rate": 7.947647651765932e-06, "loss": 0.3647, "step": 3408 }, { "epoch": 1.108978529603123, "grad_norm": 2.002141840019012, "learning_rate": 7.94611857478103e-06, "loss": 0.333, "step": 3409 }, { "epoch": 1.1093038386467144, "grad_norm": 1.9667309007309035, "learning_rate": 7.94458907561537e-06, "loss": 0.3633, "step": 3410 }, { "epoch": 1.1096291476903057, "grad_norm": 2.172003621753082, "learning_rate": 7.943059154488133e-06, "loss": 0.3312, "step": 3411 }, { "epoch": 1.1099544567338973, "grad_norm": 2.282012818315479, "learning_rate": 7.941528811618557e-06, "loss": 0.4421, "step": 3412 }, { "epoch": 1.1102797657774885, "grad_norm": 2.29324992857728, "learning_rate": 7.939998047225942e-06, "loss": 0.4293, "step": 3413 }, { "epoch": 1.11060507482108, "grad_norm": 1.9381125268424595, "learning_rate": 7.938466861529646e-06, "loss": 0.3037, "step": 3414 }, { "epoch": 1.1109303838646714, "grad_norm": 1.8653193854523094, "learning_rate": 7.93693525474909e-06, "loss": 0.4831, "step": 3415 }, { "epoch": 1.1112556929082629, "grad_norm": 2.03432461733242, "learning_rate": 7.935403227103752e-06, "loss": 0.3855, "step": 3416 }, { "epoch": 1.1115810019518542, "grad_norm": 2.2697633429266233, "learning_rate": 7.933870778813176e-06, "loss": 0.4265, "step": 3417 }, { "epoch": 1.1119063109954457, "grad_norm": 1.8192172748439064, "learning_rate": 7.93233791009696e-06, "loss": 0.3764, "step": 3418 }, { "epoch": 1.112231620039037, "grad_norm": 1.955867277610787, "learning_rate": 7.930804621174768e-06, "loss": 0.4201, "step": 3419 }, { "epoch": 1.1125569290826285, "grad_norm": 1.8175266173157385, "learning_rate": 7.92927091226632e-06, "loss": 0.3873, "step": 3420 }, { "epoch": 1.11288223812622, "grad_norm": 2.052993683924772, "learning_rate": 7.927736783591395e-06, "loss": 0.4939, "step": 3421 }, { "epoch": 1.1132075471698113, "grad_norm": 1.905733633024591, "learning_rate": 7.92620223536984e-06, "loss": 0.3836, "step": 3422 }, { "epoch": 1.1135328562134028, "grad_norm": 1.9928518059653453, "learning_rate": 7.92466726782155e-06, "loss": 0.4084, "step": 3423 }, { "epoch": 1.113858165256994, "grad_norm": 1.7716595240680104, "learning_rate": 7.923131881166493e-06, "loss": 0.4093, "step": 3424 }, { "epoch": 1.1141834743005856, "grad_norm": 1.8829331114009562, "learning_rate": 7.921596075624688e-06, "loss": 0.371, "step": 3425 }, { "epoch": 1.114508783344177, "grad_norm": 1.9076269437346371, "learning_rate": 7.92005985141622e-06, "loss": 0.4115, "step": 3426 }, { "epoch": 1.1148340923877684, "grad_norm": 2.4501243826228354, "learning_rate": 7.918523208761224e-06, "loss": 0.431, "step": 3427 }, { "epoch": 1.1151594014313597, "grad_norm": 2.2285188251661996, "learning_rate": 7.916986147879908e-06, "loss": 0.5069, "step": 3428 }, { "epoch": 1.1154847104749512, "grad_norm": 1.4188477604685599, "learning_rate": 7.915448668992534e-06, "loss": 0.2543, "step": 3429 }, { "epoch": 1.1158100195185425, "grad_norm": 1.7413480459937194, "learning_rate": 7.91391077231942e-06, "loss": 0.4035, "step": 3430 }, { "epoch": 1.116135328562134, "grad_norm": 2.1026944676601107, "learning_rate": 7.91237245808095e-06, "loss": 0.4357, "step": 3431 }, { "epoch": 1.1164606376057253, "grad_norm": 1.8717492723001063, "learning_rate": 7.910833726497564e-06, "loss": 0.4273, "step": 3432 }, { "epoch": 1.1167859466493169, "grad_norm": 1.790883097777093, "learning_rate": 7.909294577789765e-06, "loss": 0.4216, "step": 3433 }, { "epoch": 1.1171112556929081, "grad_norm": 1.8211761052176378, "learning_rate": 7.907755012178114e-06, "loss": 0.3842, "step": 3434 }, { "epoch": 1.1174365647364997, "grad_norm": 1.7639707029389895, "learning_rate": 7.906215029883231e-06, "loss": 0.3925, "step": 3435 }, { "epoch": 1.1177618737800912, "grad_norm": 2.3804983239217457, "learning_rate": 7.904674631125795e-06, "loss": 0.4566, "step": 3436 }, { "epoch": 1.1180871828236825, "grad_norm": 2.246957060348719, "learning_rate": 7.903133816126549e-06, "loss": 0.5069, "step": 3437 }, { "epoch": 1.118412491867274, "grad_norm": 2.074379564222522, "learning_rate": 7.90159258510629e-06, "loss": 0.5077, "step": 3438 }, { "epoch": 1.1187378009108653, "grad_norm": 1.9054915496384572, "learning_rate": 7.900050938285881e-06, "loss": 0.419, "step": 3439 }, { "epoch": 1.1190631099544568, "grad_norm": 1.9181843246132182, "learning_rate": 7.898508875886238e-06, "loss": 0.4846, "step": 3440 }, { "epoch": 1.119388418998048, "grad_norm": 1.8326477616070072, "learning_rate": 7.896966398128341e-06, "loss": 0.4096, "step": 3441 }, { "epoch": 1.1197137280416396, "grad_norm": 1.7267870228596105, "learning_rate": 7.895423505233227e-06, "loss": 0.4005, "step": 3442 }, { "epoch": 1.120039037085231, "grad_norm": 1.8348344665665253, "learning_rate": 7.893880197421995e-06, "loss": 0.4072, "step": 3443 }, { "epoch": 1.1203643461288224, "grad_norm": 1.835700374263099, "learning_rate": 7.8923364749158e-06, "loss": 0.3937, "step": 3444 }, { "epoch": 1.1206896551724137, "grad_norm": 1.7143100423921696, "learning_rate": 7.890792337935862e-06, "loss": 0.3649, "step": 3445 }, { "epoch": 1.1210149642160052, "grad_norm": 1.8459904213812424, "learning_rate": 7.889247786703453e-06, "loss": 0.4763, "step": 3446 }, { "epoch": 1.1213402732595967, "grad_norm": 2.0624135626520768, "learning_rate": 7.88770282143991e-06, "loss": 0.3961, "step": 3447 }, { "epoch": 1.121665582303188, "grad_norm": 1.833302774083473, "learning_rate": 7.886157442366628e-06, "loss": 0.3384, "step": 3448 }, { "epoch": 1.1219908913467795, "grad_norm": 1.7870440377709897, "learning_rate": 7.884611649705061e-06, "loss": 0.4067, "step": 3449 }, { "epoch": 1.1223162003903708, "grad_norm": 1.667714052346025, "learning_rate": 7.883065443676718e-06, "loss": 0.3412, "step": 3450 }, { "epoch": 1.1226415094339623, "grad_norm": 2.3256737930860467, "learning_rate": 7.881518824503179e-06, "loss": 0.4491, "step": 3451 }, { "epoch": 1.1229668184775536, "grad_norm": 1.8984658794747338, "learning_rate": 7.879971792406068e-06, "loss": 0.4633, "step": 3452 }, { "epoch": 1.1232921275211452, "grad_norm": 1.7728688285057437, "learning_rate": 7.87842434760708e-06, "loss": 0.3511, "step": 3453 }, { "epoch": 1.1236174365647364, "grad_norm": 1.9086815519461509, "learning_rate": 7.876876490327963e-06, "loss": 0.3579, "step": 3454 }, { "epoch": 1.123942745608328, "grad_norm": 2.0959106135810384, "learning_rate": 7.875328220790527e-06, "loss": 0.498, "step": 3455 }, { "epoch": 1.1242680546519193, "grad_norm": 1.7944949277097824, "learning_rate": 7.873779539216641e-06, "loss": 0.3568, "step": 3456 }, { "epoch": 1.1245933636955108, "grad_norm": 1.9486049846776474, "learning_rate": 7.87223044582823e-06, "loss": 0.4231, "step": 3457 }, { "epoch": 1.124918672739102, "grad_norm": 1.7997630669418199, "learning_rate": 7.870680940847283e-06, "loss": 0.3971, "step": 3458 }, { "epoch": 1.1252439817826936, "grad_norm": 1.9758698287792182, "learning_rate": 7.869131024495843e-06, "loss": 0.3796, "step": 3459 }, { "epoch": 1.1255692908262849, "grad_norm": 1.7955165437384244, "learning_rate": 7.867580696996012e-06, "loss": 0.3876, "step": 3460 }, { "epoch": 1.1258945998698764, "grad_norm": 1.958523328813209, "learning_rate": 7.866029958569956e-06, "loss": 0.4387, "step": 3461 }, { "epoch": 1.126219908913468, "grad_norm": 2.357323016005702, "learning_rate": 7.864478809439898e-06, "loss": 0.5239, "step": 3462 }, { "epoch": 1.1265452179570592, "grad_norm": 2.187184619144449, "learning_rate": 7.862927249828113e-06, "loss": 0.4688, "step": 3463 }, { "epoch": 1.1268705270006507, "grad_norm": 2.1134008836609777, "learning_rate": 7.861375279956948e-06, "loss": 0.4294, "step": 3464 }, { "epoch": 1.127195836044242, "grad_norm": 1.7871430034418132, "learning_rate": 7.859822900048796e-06, "loss": 0.4242, "step": 3465 }, { "epoch": 1.1275211450878335, "grad_norm": 1.838636724675516, "learning_rate": 7.858270110326116e-06, "loss": 0.3751, "step": 3466 }, { "epoch": 1.1278464541314248, "grad_norm": 1.8765111030080597, "learning_rate": 7.856716911011424e-06, "loss": 0.3906, "step": 3467 }, { "epoch": 1.1281717631750163, "grad_norm": 1.822448916219003, "learning_rate": 7.855163302327292e-06, "loss": 0.4349, "step": 3468 }, { "epoch": 1.1284970722186076, "grad_norm": 1.7728196211200318, "learning_rate": 7.853609284496358e-06, "loss": 0.3928, "step": 3469 }, { "epoch": 1.1288223812621991, "grad_norm": 2.0833610174465838, "learning_rate": 7.852054857741309e-06, "loss": 0.4094, "step": 3470 }, { "epoch": 1.1291476903057904, "grad_norm": 1.9325746159168307, "learning_rate": 7.850500022284897e-06, "loss": 0.3597, "step": 3471 }, { "epoch": 1.129472999349382, "grad_norm": 1.613618414680551, "learning_rate": 7.848944778349931e-06, "loss": 0.3336, "step": 3472 }, { "epoch": 1.1297983083929732, "grad_norm": 2.1050424341987775, "learning_rate": 7.847389126159281e-06, "loss": 0.4491, "step": 3473 }, { "epoch": 1.1301236174365648, "grad_norm": 1.9317068548071754, "learning_rate": 7.84583306593587e-06, "loss": 0.402, "step": 3474 }, { "epoch": 1.130448926480156, "grad_norm": 1.974100619272965, "learning_rate": 7.844276597902683e-06, "loss": 0.4358, "step": 3475 }, { "epoch": 1.1307742355237476, "grad_norm": 1.95238346386447, "learning_rate": 7.842719722282762e-06, "loss": 0.377, "step": 3476 }, { "epoch": 1.131099544567339, "grad_norm": 2.131355004730424, "learning_rate": 7.841162439299211e-06, "loss": 0.414, "step": 3477 }, { "epoch": 1.1314248536109304, "grad_norm": 1.8625059106754476, "learning_rate": 7.839604749175188e-06, "loss": 0.4626, "step": 3478 }, { "epoch": 1.1317501626545219, "grad_norm": 1.6356925959946509, "learning_rate": 7.838046652133911e-06, "loss": 0.3549, "step": 3479 }, { "epoch": 1.1320754716981132, "grad_norm": 1.9138829473967653, "learning_rate": 7.836488148398655e-06, "loss": 0.3867, "step": 3480 }, { "epoch": 1.1324007807417047, "grad_norm": 2.3587041258035715, "learning_rate": 7.834929238192759e-06, "loss": 0.5421, "step": 3481 }, { "epoch": 1.132726089785296, "grad_norm": 2.0564256615158856, "learning_rate": 7.83336992173961e-06, "loss": 0.4784, "step": 3482 }, { "epoch": 1.1330513988288875, "grad_norm": 1.878681694339782, "learning_rate": 7.831810199262664e-06, "loss": 0.3996, "step": 3483 }, { "epoch": 1.1333767078724788, "grad_norm": 1.853094648338448, "learning_rate": 7.830250070985429e-06, "loss": 0.3757, "step": 3484 }, { "epoch": 1.1337020169160703, "grad_norm": 1.9078894516038292, "learning_rate": 7.828689537131471e-06, "loss": 0.4104, "step": 3485 }, { "epoch": 1.1340273259596616, "grad_norm": 1.9384494729384865, "learning_rate": 7.827128597924416e-06, "loss": 0.3248, "step": 3486 }, { "epoch": 1.1343526350032531, "grad_norm": 2.1771005468878455, "learning_rate": 7.825567253587947e-06, "loss": 0.461, "step": 3487 }, { "epoch": 1.1346779440468444, "grad_norm": 1.7620716477081657, "learning_rate": 7.82400550434581e-06, "loss": 0.3749, "step": 3488 }, { "epoch": 1.135003253090436, "grad_norm": 1.8531934225803344, "learning_rate": 7.822443350421798e-06, "loss": 0.4387, "step": 3489 }, { "epoch": 1.1353285621340272, "grad_norm": 1.8656739381558771, "learning_rate": 7.820880792039774e-06, "loss": 0.4336, "step": 3490 }, { "epoch": 1.1356538711776187, "grad_norm": 2.0570984837014104, "learning_rate": 7.819317829423652e-06, "loss": 0.4614, "step": 3491 }, { "epoch": 1.1359791802212102, "grad_norm": 1.989562787880136, "learning_rate": 7.817754462797404e-06, "loss": 0.3865, "step": 3492 }, { "epoch": 1.1363044892648015, "grad_norm": 2.217178799042679, "learning_rate": 7.816190692385062e-06, "loss": 0.4798, "step": 3493 }, { "epoch": 1.136629798308393, "grad_norm": 1.8400717296207432, "learning_rate": 7.814626518410719e-06, "loss": 0.3294, "step": 3494 }, { "epoch": 1.1369551073519844, "grad_norm": 1.8113736716454347, "learning_rate": 7.813061941098517e-06, "loss": 0.3406, "step": 3495 }, { "epoch": 1.1372804163955759, "grad_norm": 2.0935193193152286, "learning_rate": 7.811496960672665e-06, "loss": 0.5008, "step": 3496 }, { "epoch": 1.1376057254391672, "grad_norm": 1.7820264768718308, "learning_rate": 7.809931577357423e-06, "loss": 0.3645, "step": 3497 }, { "epoch": 1.1379310344827587, "grad_norm": 2.0535400245366606, "learning_rate": 7.808365791377113e-06, "loss": 0.4229, "step": 3498 }, { "epoch": 1.13825634352635, "grad_norm": 1.9464365619935249, "learning_rate": 7.806799602956111e-06, "loss": 0.3648, "step": 3499 }, { "epoch": 1.1385816525699415, "grad_norm": 1.9413820881267667, "learning_rate": 7.805233012318856e-06, "loss": 0.3667, "step": 3500 }, { "epoch": 1.1389069616135328, "grad_norm": 2.0233362124346344, "learning_rate": 7.80366601968984e-06, "loss": 0.4363, "step": 3501 }, { "epoch": 1.1392322706571243, "grad_norm": 1.956521754896893, "learning_rate": 7.802098625293614e-06, "loss": 0.4052, "step": 3502 }, { "epoch": 1.1395575797007156, "grad_norm": 1.8697359246402732, "learning_rate": 7.800530829354787e-06, "loss": 0.3846, "step": 3503 }, { "epoch": 1.139882888744307, "grad_norm": 1.9756545783260424, "learning_rate": 7.798962632098024e-06, "loss": 0.3968, "step": 3504 }, { "epoch": 1.1402081977878984, "grad_norm": 1.6376742899626704, "learning_rate": 7.797394033748049e-06, "loss": 0.2809, "step": 3505 }, { "epoch": 1.14053350683149, "grad_norm": 1.9147778096914496, "learning_rate": 7.795825034529646e-06, "loss": 0.4187, "step": 3506 }, { "epoch": 1.1408588158750814, "grad_norm": 1.9143537595271605, "learning_rate": 7.794255634667652e-06, "loss": 0.418, "step": 3507 }, { "epoch": 1.1411841249186727, "grad_norm": 2.050414462591418, "learning_rate": 7.792685834386959e-06, "loss": 0.4161, "step": 3508 }, { "epoch": 1.1415094339622642, "grad_norm": 1.7573302815508305, "learning_rate": 7.791115633912527e-06, "loss": 0.4047, "step": 3509 }, { "epoch": 1.1418347430058555, "grad_norm": 2.1324237108045514, "learning_rate": 7.789545033469364e-06, "loss": 0.4004, "step": 3510 }, { "epoch": 1.142160052049447, "grad_norm": 1.89921126204729, "learning_rate": 7.787974033282535e-06, "loss": 0.4038, "step": 3511 }, { "epoch": 1.1424853610930383, "grad_norm": 2.217720649720604, "learning_rate": 7.786402633577171e-06, "loss": 0.4744, "step": 3512 }, { "epoch": 1.1428106701366298, "grad_norm": 1.9233396825709625, "learning_rate": 7.78483083457845e-06, "loss": 0.3153, "step": 3513 }, { "epoch": 1.1431359791802211, "grad_norm": 2.2933115386241916, "learning_rate": 7.783258636511615e-06, "loss": 0.4516, "step": 3514 }, { "epoch": 1.1434612882238127, "grad_norm": 1.9025076163552288, "learning_rate": 7.78168603960196e-06, "loss": 0.399, "step": 3515 }, { "epoch": 1.143786597267404, "grad_norm": 1.985056253567049, "learning_rate": 7.780113044074842e-06, "loss": 0.4228, "step": 3516 }, { "epoch": 1.1441119063109955, "grad_norm": 1.9269859036648376, "learning_rate": 7.778539650155672e-06, "loss": 0.4134, "step": 3517 }, { "epoch": 1.144437215354587, "grad_norm": 2.340641305213996, "learning_rate": 7.776965858069916e-06, "loss": 0.5214, "step": 3518 }, { "epoch": 1.1447625243981783, "grad_norm": 1.9165337303488121, "learning_rate": 7.775391668043102e-06, "loss": 0.442, "step": 3519 }, { "epoch": 1.1450878334417696, "grad_norm": 1.7680023438665835, "learning_rate": 7.773817080300812e-06, "loss": 0.3075, "step": 3520 }, { "epoch": 1.145413142485361, "grad_norm": 2.094557192580737, "learning_rate": 7.772242095068682e-06, "loss": 0.4219, "step": 3521 }, { "epoch": 1.1457384515289526, "grad_norm": 2.010193662088263, "learning_rate": 7.770666712572417e-06, "loss": 0.4936, "step": 3522 }, { "epoch": 1.1460637605725439, "grad_norm": 1.7741428671818313, "learning_rate": 7.76909093303776e-06, "loss": 0.4016, "step": 3523 }, { "epoch": 1.1463890696161354, "grad_norm": 2.033883282781993, "learning_rate": 7.767514756690529e-06, "loss": 0.427, "step": 3524 }, { "epoch": 1.1467143786597267, "grad_norm": 1.819989455339348, "learning_rate": 7.765938183756585e-06, "loss": 0.4137, "step": 3525 }, { "epoch": 1.1470396877033182, "grad_norm": 2.0098461435308135, "learning_rate": 7.764361214461857e-06, "loss": 0.4838, "step": 3526 }, { "epoch": 1.1473649967469095, "grad_norm": 2.071043212348639, "learning_rate": 7.762783849032324e-06, "loss": 0.4307, "step": 3527 }, { "epoch": 1.147690305790501, "grad_norm": 1.7726032003033134, "learning_rate": 7.761206087694025e-06, "loss": 0.3971, "step": 3528 }, { "epoch": 1.1480156148340923, "grad_norm": 1.700240472717814, "learning_rate": 7.75962793067305e-06, "loss": 0.3285, "step": 3529 }, { "epoch": 1.1483409238776838, "grad_norm": 2.043309548486661, "learning_rate": 7.758049378195554e-06, "loss": 0.4066, "step": 3530 }, { "epoch": 1.1486662329212751, "grad_norm": 1.9490480402223622, "learning_rate": 7.756470430487742e-06, "loss": 0.4075, "step": 3531 }, { "epoch": 1.1489915419648666, "grad_norm": 1.6183890910565444, "learning_rate": 7.75489108777588e-06, "loss": 0.3599, "step": 3532 }, { "epoch": 1.1493168510084582, "grad_norm": 1.7665492545128847, "learning_rate": 7.753311350286288e-06, "loss": 0.4243, "step": 3533 }, { "epoch": 1.1496421600520494, "grad_norm": 2.24726315783371, "learning_rate": 7.751731218245346e-06, "loss": 0.4575, "step": 3534 }, { "epoch": 1.1499674690956407, "grad_norm": 1.5568140768172116, "learning_rate": 7.750150691879484e-06, "loss": 0.2957, "step": 3535 }, { "epoch": 1.1502927781392323, "grad_norm": 1.7895240449951937, "learning_rate": 7.748569771415195e-06, "loss": 0.4184, "step": 3536 }, { "epoch": 1.1506180871828238, "grad_norm": 2.191010661901101, "learning_rate": 7.746988457079024e-06, "loss": 0.4393, "step": 3537 }, { "epoch": 1.150943396226415, "grad_norm": 1.8398385163513744, "learning_rate": 7.745406749097577e-06, "loss": 0.4004, "step": 3538 }, { "epoch": 1.1512687052700066, "grad_norm": 1.625133986384516, "learning_rate": 7.743824647697512e-06, "loss": 0.3282, "step": 3539 }, { "epoch": 1.1515940143135979, "grad_norm": 2.3784580111520572, "learning_rate": 7.742242153105544e-06, "loss": 0.4696, "step": 3540 }, { "epoch": 1.1519193233571894, "grad_norm": 2.1158052350853107, "learning_rate": 7.740659265548447e-06, "loss": 0.4746, "step": 3541 }, { "epoch": 1.1522446324007807, "grad_norm": 2.1058051186755717, "learning_rate": 7.739075985253053e-06, "loss": 0.4113, "step": 3542 }, { "epoch": 1.1525699414443722, "grad_norm": 1.9513552868336932, "learning_rate": 7.737492312446242e-06, "loss": 0.4129, "step": 3543 }, { "epoch": 1.1528952504879635, "grad_norm": 2.005898827088665, "learning_rate": 7.735908247354957e-06, "loss": 0.4353, "step": 3544 }, { "epoch": 1.153220559531555, "grad_norm": 2.197063239434779, "learning_rate": 7.734323790206197e-06, "loss": 0.4287, "step": 3545 }, { "epoch": 1.1535458685751463, "grad_norm": 1.785308483447223, "learning_rate": 7.732738941227016e-06, "loss": 0.3569, "step": 3546 }, { "epoch": 1.1538711776187378, "grad_norm": 2.0383170191874944, "learning_rate": 7.73115370064452e-06, "loss": 0.4762, "step": 3547 }, { "epoch": 1.1541964866623293, "grad_norm": 2.016141451101769, "learning_rate": 7.72956806868588e-06, "loss": 0.3485, "step": 3548 }, { "epoch": 1.1545217957059206, "grad_norm": 1.8567144691909818, "learning_rate": 7.727982045578315e-06, "loss": 0.4167, "step": 3549 }, { "epoch": 1.1548471047495121, "grad_norm": 1.906471364760139, "learning_rate": 7.726395631549105e-06, "loss": 0.3978, "step": 3550 }, { "epoch": 1.1551724137931034, "grad_norm": 1.9222327135561483, "learning_rate": 7.724808826825582e-06, "loss": 0.4067, "step": 3551 }, { "epoch": 1.155497722836695, "grad_norm": 1.6012146081288487, "learning_rate": 7.723221631635139e-06, "loss": 0.3915, "step": 3552 }, { "epoch": 1.1558230318802862, "grad_norm": 1.622025360805375, "learning_rate": 7.721634046205217e-06, "loss": 0.3885, "step": 3553 }, { "epoch": 1.1561483409238777, "grad_norm": 1.7952491575830423, "learning_rate": 7.720046070763324e-06, "loss": 0.3776, "step": 3554 }, { "epoch": 1.156473649967469, "grad_norm": 1.683860672629995, "learning_rate": 7.718457705537016e-06, "loss": 0.4327, "step": 3555 }, { "epoch": 1.1567989590110606, "grad_norm": 2.132457938401384, "learning_rate": 7.716868950753907e-06, "loss": 0.4062, "step": 3556 }, { "epoch": 1.1571242680546518, "grad_norm": 1.8299614780839626, "learning_rate": 7.715279806641664e-06, "loss": 0.4097, "step": 3557 }, { "epoch": 1.1574495770982434, "grad_norm": 2.309263518844273, "learning_rate": 7.713690273428013e-06, "loss": 0.5246, "step": 3558 }, { "epoch": 1.1577748861418347, "grad_norm": 1.8225285978225703, "learning_rate": 7.71210035134074e-06, "loss": 0.3086, "step": 3559 }, { "epoch": 1.1581001951854262, "grad_norm": 1.6838664155633138, "learning_rate": 7.710510040607677e-06, "loss": 0.3158, "step": 3560 }, { "epoch": 1.1584255042290175, "grad_norm": 2.0285906417593202, "learning_rate": 7.708919341456718e-06, "loss": 0.4333, "step": 3561 }, { "epoch": 1.158750813272609, "grad_norm": 2.032197406710385, "learning_rate": 7.707328254115812e-06, "loss": 0.3998, "step": 3562 }, { "epoch": 1.1590761223162005, "grad_norm": 1.8797739746188147, "learning_rate": 7.70573677881296e-06, "loss": 0.3866, "step": 3563 }, { "epoch": 1.1594014313597918, "grad_norm": 1.8820520473566285, "learning_rate": 7.704144915776225e-06, "loss": 0.3768, "step": 3564 }, { "epoch": 1.1597267404033833, "grad_norm": 2.0916899069326504, "learning_rate": 7.70255266523372e-06, "loss": 0.3593, "step": 3565 }, { "epoch": 1.1600520494469746, "grad_norm": 2.0313818680034537, "learning_rate": 7.700960027413617e-06, "loss": 0.4376, "step": 3566 }, { "epoch": 1.1603773584905661, "grad_norm": 2.097504691692854, "learning_rate": 7.69936700254414e-06, "loss": 0.4094, "step": 3567 }, { "epoch": 1.1607026675341574, "grad_norm": 1.9312371757727862, "learning_rate": 7.697773590853576e-06, "loss": 0.3635, "step": 3568 }, { "epoch": 1.161027976577749, "grad_norm": 2.2835519424913295, "learning_rate": 7.696179792570254e-06, "loss": 0.4636, "step": 3569 }, { "epoch": 1.1613532856213402, "grad_norm": 1.6219203031084106, "learning_rate": 7.69458560792257e-06, "loss": 0.2826, "step": 3570 }, { "epoch": 1.1616785946649317, "grad_norm": 2.280978508434454, "learning_rate": 7.692991037138972e-06, "loss": 0.4933, "step": 3571 }, { "epoch": 1.162003903708523, "grad_norm": 1.7766919696535992, "learning_rate": 7.691396080447963e-06, "loss": 0.367, "step": 3572 }, { "epoch": 1.1623292127521145, "grad_norm": 1.8596177067048165, "learning_rate": 7.689800738078102e-06, "loss": 0.3815, "step": 3573 }, { "epoch": 1.1626545217957058, "grad_norm": 2.029577917532511, "learning_rate": 7.688205010258003e-06, "loss": 0.4856, "step": 3574 }, { "epoch": 1.1629798308392973, "grad_norm": 1.7010253407545468, "learning_rate": 7.686608897216331e-06, "loss": 0.3526, "step": 3575 }, { "epoch": 1.1633051398828886, "grad_norm": 1.7668339835276767, "learning_rate": 7.685012399181815e-06, "loss": 0.3343, "step": 3576 }, { "epoch": 1.1636304489264802, "grad_norm": 1.7803871932570166, "learning_rate": 7.68341551638323e-06, "loss": 0.3976, "step": 3577 }, { "epoch": 1.1639557579700717, "grad_norm": 1.987002708882182, "learning_rate": 7.681818249049414e-06, "loss": 0.4392, "step": 3578 }, { "epoch": 1.164281067013663, "grad_norm": 1.7796252097373455, "learning_rate": 7.680220597409254e-06, "loss": 0.3054, "step": 3579 }, { "epoch": 1.1646063760572545, "grad_norm": 2.1353824581312497, "learning_rate": 7.678622561691696e-06, "loss": 0.3874, "step": 3580 }, { "epoch": 1.1649316851008458, "grad_norm": 1.7847151854962606, "learning_rate": 7.677024142125739e-06, "loss": 0.4495, "step": 3581 }, { "epoch": 1.1652569941444373, "grad_norm": 1.9113650332515089, "learning_rate": 7.675425338940437e-06, "loss": 0.408, "step": 3582 }, { "epoch": 1.1655823031880286, "grad_norm": 2.0931465608970807, "learning_rate": 7.673826152364899e-06, "loss": 0.4274, "step": 3583 }, { "epoch": 1.16590761223162, "grad_norm": 1.7281226691516174, "learning_rate": 7.672226582628294e-06, "loss": 0.3718, "step": 3584 }, { "epoch": 1.1662329212752114, "grad_norm": 1.7657180859659922, "learning_rate": 7.670626629959835e-06, "loss": 0.3366, "step": 3585 }, { "epoch": 1.166558230318803, "grad_norm": 1.8756754670785611, "learning_rate": 7.669026294588801e-06, "loss": 0.392, "step": 3586 }, { "epoch": 1.1668835393623942, "grad_norm": 1.8270315587756423, "learning_rate": 7.667425576744517e-06, "loss": 0.3496, "step": 3587 }, { "epoch": 1.1672088484059857, "grad_norm": 1.6893858579024625, "learning_rate": 7.665824476656374e-06, "loss": 0.3216, "step": 3588 }, { "epoch": 1.167534157449577, "grad_norm": 2.179535319518572, "learning_rate": 7.664222994553803e-06, "loss": 0.4406, "step": 3589 }, { "epoch": 1.1678594664931685, "grad_norm": 1.6802305017924624, "learning_rate": 7.6626211306663e-06, "loss": 0.2894, "step": 3590 }, { "epoch": 1.1681847755367598, "grad_norm": 1.6762415610015562, "learning_rate": 7.661018885223414e-06, "loss": 0.3436, "step": 3591 }, { "epoch": 1.1685100845803513, "grad_norm": 1.6485924469239113, "learning_rate": 7.65941625845475e-06, "loss": 0.3644, "step": 3592 }, { "epoch": 1.1688353936239428, "grad_norm": 1.9873357212599179, "learning_rate": 7.657813250589961e-06, "loss": 0.4668, "step": 3593 }, { "epoch": 1.1691607026675341, "grad_norm": 1.8890099368630993, "learning_rate": 7.65620986185876e-06, "loss": 0.3581, "step": 3594 }, { "epoch": 1.1694860117111257, "grad_norm": 2.2108421102419547, "learning_rate": 7.654606092490917e-06, "loss": 0.3864, "step": 3595 }, { "epoch": 1.169811320754717, "grad_norm": 1.7489259519296296, "learning_rate": 7.653001942716251e-06, "loss": 0.3415, "step": 3596 }, { "epoch": 1.1701366297983085, "grad_norm": 2.1507696445940447, "learning_rate": 7.651397412764637e-06, "loss": 0.5147, "step": 3597 }, { "epoch": 1.1704619388418998, "grad_norm": 1.9408895853001127, "learning_rate": 7.649792502866007e-06, "loss": 0.4232, "step": 3598 }, { "epoch": 1.1707872478854913, "grad_norm": 1.6988087863583783, "learning_rate": 7.648187213250341e-06, "loss": 0.3734, "step": 3599 }, { "epoch": 1.1711125569290826, "grad_norm": 1.969947551561966, "learning_rate": 7.646581544147686e-06, "loss": 0.3992, "step": 3600 }, { "epoch": 1.171437865972674, "grad_norm": 1.9287955532000614, "learning_rate": 7.644975495788129e-06, "loss": 0.3722, "step": 3601 }, { "epoch": 1.1717631750162654, "grad_norm": 1.755823980556907, "learning_rate": 7.643369068401823e-06, "loss": 0.3834, "step": 3602 }, { "epoch": 1.1720884840598569, "grad_norm": 1.7939065273662902, "learning_rate": 7.641762262218964e-06, "loss": 0.3358, "step": 3603 }, { "epoch": 1.1724137931034484, "grad_norm": 1.956025509652689, "learning_rate": 7.640155077469812e-06, "loss": 0.4028, "step": 3604 }, { "epoch": 1.1727391021470397, "grad_norm": 1.8003010450553303, "learning_rate": 7.638547514384676e-06, "loss": 0.31, "step": 3605 }, { "epoch": 1.173064411190631, "grad_norm": 1.6526424318908821, "learning_rate": 7.636939573193924e-06, "loss": 0.3124, "step": 3606 }, { "epoch": 1.1733897202342225, "grad_norm": 1.6512615931312211, "learning_rate": 7.635331254127973e-06, "loss": 0.4215, "step": 3607 }, { "epoch": 1.173715029277814, "grad_norm": 1.9134639091664083, "learning_rate": 7.633722557417293e-06, "loss": 0.389, "step": 3608 }, { "epoch": 1.1740403383214053, "grad_norm": 1.8799767594657342, "learning_rate": 7.632113483292419e-06, "loss": 0.4058, "step": 3609 }, { "epoch": 1.1743656473649968, "grad_norm": 1.860238790368424, "learning_rate": 7.630504031983925e-06, "loss": 0.3961, "step": 3610 }, { "epoch": 1.1746909564085881, "grad_norm": 1.636187362992296, "learning_rate": 7.6288942037224515e-06, "loss": 0.318, "step": 3611 }, { "epoch": 1.1750162654521796, "grad_norm": 1.6914378146876212, "learning_rate": 7.6272839987386834e-06, "loss": 0.3894, "step": 3612 }, { "epoch": 1.175341574495771, "grad_norm": 2.252367758392996, "learning_rate": 7.62567341726337e-06, "loss": 0.4137, "step": 3613 }, { "epoch": 1.1756668835393624, "grad_norm": 1.8153457241787827, "learning_rate": 7.624062459527302e-06, "loss": 0.362, "step": 3614 }, { "epoch": 1.1759921925829537, "grad_norm": 1.834417625403336, "learning_rate": 7.622451125761335e-06, "loss": 0.407, "step": 3615 }, { "epoch": 1.1763175016265452, "grad_norm": 1.8748912738023285, "learning_rate": 7.620839416196375e-06, "loss": 0.4273, "step": 3616 }, { "epoch": 1.1766428106701365, "grad_norm": 1.9032449380434944, "learning_rate": 7.619227331063377e-06, "loss": 0.3787, "step": 3617 }, { "epoch": 1.176968119713728, "grad_norm": 1.9389415308476026, "learning_rate": 7.617614870593356e-06, "loss": 0.4475, "step": 3618 }, { "epoch": 1.1772934287573196, "grad_norm": 2.1555494848077696, "learning_rate": 7.616002035017381e-06, "loss": 0.3626, "step": 3619 }, { "epoch": 1.1776187378009109, "grad_norm": 1.7941807351234111, "learning_rate": 7.614388824566567e-06, "loss": 0.3944, "step": 3620 }, { "epoch": 1.1779440468445022, "grad_norm": 1.9290086966701692, "learning_rate": 7.6127752394720945e-06, "loss": 0.3844, "step": 3621 }, { "epoch": 1.1782693558880937, "grad_norm": 1.7749936050560218, "learning_rate": 7.611161279965187e-06, "loss": 0.3575, "step": 3622 }, { "epoch": 1.1785946649316852, "grad_norm": 2.1195112453693072, "learning_rate": 7.6095469462771276e-06, "loss": 0.4774, "step": 3623 }, { "epoch": 1.1789199739752765, "grad_norm": 1.7408793052791962, "learning_rate": 7.60793223863925e-06, "loss": 0.4136, "step": 3624 }, { "epoch": 1.179245283018868, "grad_norm": 2.254083106106941, "learning_rate": 7.606317157282944e-06, "loss": 0.4514, "step": 3625 }, { "epoch": 1.1795705920624593, "grad_norm": 1.8533142222565118, "learning_rate": 7.604701702439652e-06, "loss": 0.4256, "step": 3626 }, { "epoch": 1.1798959011060508, "grad_norm": 2.216731228547199, "learning_rate": 7.60308587434087e-06, "loss": 0.4781, "step": 3627 }, { "epoch": 1.180221210149642, "grad_norm": 1.9594501807095204, "learning_rate": 7.601469673218144e-06, "loss": 0.4452, "step": 3628 }, { "epoch": 1.1805465191932336, "grad_norm": 1.8068655366961552, "learning_rate": 7.599853099303083e-06, "loss": 0.3565, "step": 3629 }, { "epoch": 1.180871828236825, "grad_norm": 1.933500948016494, "learning_rate": 7.598236152827337e-06, "loss": 0.4022, "step": 3630 }, { "epoch": 1.1811971372804164, "grad_norm": 1.7498778457828201, "learning_rate": 7.59661883402262e-06, "loss": 0.342, "step": 3631 }, { "epoch": 1.1815224463240077, "grad_norm": 1.8658279226112808, "learning_rate": 7.59500114312069e-06, "loss": 0.4451, "step": 3632 }, { "epoch": 1.1818477553675992, "grad_norm": 2.2940342159061773, "learning_rate": 7.593383080353369e-06, "loss": 0.4874, "step": 3633 }, { "epoch": 1.1821730644111907, "grad_norm": 1.884143621950697, "learning_rate": 7.591764645952522e-06, "loss": 0.3958, "step": 3634 }, { "epoch": 1.182498373454782, "grad_norm": 1.9469048354887726, "learning_rate": 7.590145840150073e-06, "loss": 0.327, "step": 3635 }, { "epoch": 1.1828236824983733, "grad_norm": 1.7739607691497472, "learning_rate": 7.588526663177997e-06, "loss": 0.3457, "step": 3636 }, { "epoch": 1.1831489915419648, "grad_norm": 1.818971009099171, "learning_rate": 7.586907115268327e-06, "loss": 0.3223, "step": 3637 }, { "epoch": 1.1834743005855564, "grad_norm": 1.891364147111136, "learning_rate": 7.585287196653142e-06, "loss": 0.4239, "step": 3638 }, { "epoch": 1.1837996096291477, "grad_norm": 1.7148386346340982, "learning_rate": 7.583666907564577e-06, "loss": 0.3375, "step": 3639 }, { "epoch": 1.1841249186727392, "grad_norm": 2.8398089484576166, "learning_rate": 7.5820462482348226e-06, "loss": 0.4166, "step": 3640 }, { "epoch": 1.1844502277163305, "grad_norm": 1.8909281753362301, "learning_rate": 7.580425218896118e-06, "loss": 0.4182, "step": 3641 }, { "epoch": 1.184775536759922, "grad_norm": 1.766923450677292, "learning_rate": 7.57880381978076e-06, "loss": 0.3754, "step": 3642 }, { "epoch": 1.1851008458035133, "grad_norm": 2.164632288773321, "learning_rate": 7.577182051121095e-06, "loss": 0.3826, "step": 3643 }, { "epoch": 1.1854261548471048, "grad_norm": 1.903241884825783, "learning_rate": 7.575559913149524e-06, "loss": 0.4356, "step": 3644 }, { "epoch": 1.185751463890696, "grad_norm": 2.020743704526156, "learning_rate": 7.573937406098501e-06, "loss": 0.4377, "step": 3645 }, { "epoch": 1.1860767729342876, "grad_norm": 1.9081294396272124, "learning_rate": 7.5723145302005294e-06, "loss": 0.3976, "step": 3646 }, { "epoch": 1.1864020819778789, "grad_norm": 1.8224798479578102, "learning_rate": 7.570691285688172e-06, "loss": 0.3565, "step": 3647 }, { "epoch": 1.1867273910214704, "grad_norm": 1.9217491121561812, "learning_rate": 7.56906767279404e-06, "loss": 0.4239, "step": 3648 }, { "epoch": 1.187052700065062, "grad_norm": 1.9937215994309656, "learning_rate": 7.567443691750797e-06, "loss": 0.4338, "step": 3649 }, { "epoch": 1.1873780091086532, "grad_norm": 1.7591016430739166, "learning_rate": 7.565819342791161e-06, "loss": 0.3372, "step": 3650 }, { "epoch": 1.1877033181522447, "grad_norm": 1.821445821740526, "learning_rate": 7.564194626147903e-06, "loss": 0.3664, "step": 3651 }, { "epoch": 1.188028627195836, "grad_norm": 1.8930533388360218, "learning_rate": 7.562569542053843e-06, "loss": 0.4326, "step": 3652 }, { "epoch": 1.1883539362394275, "grad_norm": 1.8292370596710776, "learning_rate": 7.5609440907418615e-06, "loss": 0.3369, "step": 3653 }, { "epoch": 1.1886792452830188, "grad_norm": 1.8210001270224963, "learning_rate": 7.5593182724448846e-06, "loss": 0.3671, "step": 3654 }, { "epoch": 1.1890045543266103, "grad_norm": 1.8898128158141392, "learning_rate": 7.5576920873958925e-06, "loss": 0.4707, "step": 3655 }, { "epoch": 1.1893298633702016, "grad_norm": 1.769063039959445, "learning_rate": 7.556065535827919e-06, "loss": 0.3189, "step": 3656 }, { "epoch": 1.1896551724137931, "grad_norm": 1.6971938808826952, "learning_rate": 7.55443861797405e-06, "loss": 0.385, "step": 3657 }, { "epoch": 1.1899804814573844, "grad_norm": 2.103123512424674, "learning_rate": 7.552811334067423e-06, "loss": 0.4688, "step": 3658 }, { "epoch": 1.190305790500976, "grad_norm": 1.5576526313240724, "learning_rate": 7.55118368434123e-06, "loss": 0.2879, "step": 3659 }, { "epoch": 1.1906310995445673, "grad_norm": 1.995140292968665, "learning_rate": 7.549555669028714e-06, "loss": 0.4494, "step": 3660 }, { "epoch": 1.1909564085881588, "grad_norm": 1.5862565344535442, "learning_rate": 7.547927288363173e-06, "loss": 0.349, "step": 3661 }, { "epoch": 1.19128171763175, "grad_norm": 1.9884280731355057, "learning_rate": 7.54629854257795e-06, "loss": 0.3647, "step": 3662 }, { "epoch": 1.1916070266753416, "grad_norm": 2.2462858141456175, "learning_rate": 7.5446694319064495e-06, "loss": 0.4173, "step": 3663 }, { "epoch": 1.191932335718933, "grad_norm": 1.7256833979738937, "learning_rate": 7.5430399565821235e-06, "loss": 0.3832, "step": 3664 }, { "epoch": 1.1922576447625244, "grad_norm": 2.121548640879621, "learning_rate": 7.541410116838475e-06, "loss": 0.4343, "step": 3665 }, { "epoch": 1.192582953806116, "grad_norm": 1.8329410785895, "learning_rate": 7.539779912909062e-06, "loss": 0.3657, "step": 3666 }, { "epoch": 1.1929082628497072, "grad_norm": 1.6168603922041769, "learning_rate": 7.538149345027494e-06, "loss": 0.371, "step": 3667 }, { "epoch": 1.1932335718932987, "grad_norm": 2.05742677258118, "learning_rate": 7.5365184134274335e-06, "loss": 0.3993, "step": 3668 }, { "epoch": 1.19355888093689, "grad_norm": 2.119132564593787, "learning_rate": 7.534887118342593e-06, "loss": 0.4065, "step": 3669 }, { "epoch": 1.1938841899804815, "grad_norm": 2.040615794201393, "learning_rate": 7.533255460006738e-06, "loss": 0.4082, "step": 3670 }, { "epoch": 1.1942094990240728, "grad_norm": 1.7980514064271613, "learning_rate": 7.531623438653688e-06, "loss": 0.3656, "step": 3671 }, { "epoch": 1.1945348080676643, "grad_norm": 1.9035714504017966, "learning_rate": 7.529991054517309e-06, "loss": 0.4359, "step": 3672 }, { "epoch": 1.1948601171112556, "grad_norm": 2.0770415120936274, "learning_rate": 7.528358307831527e-06, "loss": 0.586, "step": 3673 }, { "epoch": 1.1951854261548471, "grad_norm": 1.97874486375728, "learning_rate": 7.526725198830314e-06, "loss": 0.4136, "step": 3674 }, { "epoch": 1.1955107351984384, "grad_norm": 1.7023213197611422, "learning_rate": 7.525091727747693e-06, "loss": 0.3603, "step": 3675 }, { "epoch": 1.19583604424203, "grad_norm": 2.0765069592676384, "learning_rate": 7.523457894817745e-06, "loss": 0.4065, "step": 3676 }, { "epoch": 1.1961613532856212, "grad_norm": 1.8308182216674465, "learning_rate": 7.521823700274599e-06, "loss": 0.3542, "step": 3677 }, { "epoch": 1.1964866623292127, "grad_norm": 1.8929330564887616, "learning_rate": 7.520189144352434e-06, "loss": 0.4501, "step": 3678 }, { "epoch": 1.1968119713728043, "grad_norm": 1.5991918487431314, "learning_rate": 7.518554227285486e-06, "loss": 0.3282, "step": 3679 }, { "epoch": 1.1971372804163956, "grad_norm": 2.1195601942767066, "learning_rate": 7.516918949308039e-06, "loss": 0.4715, "step": 3680 }, { "epoch": 1.197462589459987, "grad_norm": 1.6464226059977427, "learning_rate": 7.515283310654427e-06, "loss": 0.3405, "step": 3681 }, { "epoch": 1.1977878985035784, "grad_norm": 2.0860405250458265, "learning_rate": 7.513647311559038e-06, "loss": 0.4601, "step": 3682 }, { "epoch": 1.1981132075471699, "grad_norm": 1.558169720202972, "learning_rate": 7.512010952256315e-06, "loss": 0.3341, "step": 3683 }, { "epoch": 1.1984385165907612, "grad_norm": 2.224231259520042, "learning_rate": 7.510374232980746e-06, "loss": 0.3746, "step": 3684 }, { "epoch": 1.1987638256343527, "grad_norm": 1.6664498349994017, "learning_rate": 7.508737153966878e-06, "loss": 0.3661, "step": 3685 }, { "epoch": 1.199089134677944, "grad_norm": 2.0118240517845494, "learning_rate": 7.507099715449302e-06, "loss": 0.3781, "step": 3686 }, { "epoch": 1.1994144437215355, "grad_norm": 2.097983137486417, "learning_rate": 7.505461917662666e-06, "loss": 0.3163, "step": 3687 }, { "epoch": 1.1997397527651268, "grad_norm": 1.8105619708215563, "learning_rate": 7.503823760841665e-06, "loss": 0.358, "step": 3688 }, { "epoch": 1.2000650618087183, "grad_norm": 2.1962993246467004, "learning_rate": 7.5021852452210515e-06, "loss": 0.4264, "step": 3689 }, { "epoch": 1.2003903708523098, "grad_norm": 1.9976446752960233, "learning_rate": 7.500546371035624e-06, "loss": 0.4363, "step": 3690 }, { "epoch": 1.2007156798959011, "grad_norm": 2.0704373484852625, "learning_rate": 7.498907138520233e-06, "loss": 0.494, "step": 3691 }, { "epoch": 1.2010409889394924, "grad_norm": 1.5841275859326407, "learning_rate": 7.497267547909783e-06, "loss": 0.3463, "step": 3692 }, { "epoch": 1.201366297983084, "grad_norm": 1.7661822368626665, "learning_rate": 7.4956275994392295e-06, "loss": 0.3989, "step": 3693 }, { "epoch": 1.2016916070266754, "grad_norm": 1.8058547587050235, "learning_rate": 7.493987293343576e-06, "loss": 0.3244, "step": 3694 }, { "epoch": 1.2020169160702667, "grad_norm": 2.0991971233480746, "learning_rate": 7.4923466298578825e-06, "loss": 0.4374, "step": 3695 }, { "epoch": 1.2023422251138582, "grad_norm": 2.139075266229832, "learning_rate": 7.4907056092172526e-06, "loss": 0.3924, "step": 3696 }, { "epoch": 1.2026675341574495, "grad_norm": 1.772729702282939, "learning_rate": 7.489064231656851e-06, "loss": 0.3537, "step": 3697 }, { "epoch": 1.202992843201041, "grad_norm": 1.5911910478525533, "learning_rate": 7.487422497411884e-06, "loss": 0.3209, "step": 3698 }, { "epoch": 1.2033181522446323, "grad_norm": 1.9618087064093979, "learning_rate": 7.485780406717615e-06, "loss": 0.3489, "step": 3699 }, { "epoch": 1.2036434612882239, "grad_norm": 1.7947749203799988, "learning_rate": 7.4841379598093554e-06, "loss": 0.3647, "step": 3700 }, { "epoch": 1.2039687703318152, "grad_norm": 1.6919538357465682, "learning_rate": 7.482495156922471e-06, "loss": 0.3322, "step": 3701 }, { "epoch": 1.2042940793754067, "grad_norm": 2.0956189240311884, "learning_rate": 7.480851998292374e-06, "loss": 0.4669, "step": 3702 }, { "epoch": 1.204619388418998, "grad_norm": 1.4989243175156406, "learning_rate": 7.4792084841545346e-06, "loss": 0.3347, "step": 3703 }, { "epoch": 1.2049446974625895, "grad_norm": 1.732919294782211, "learning_rate": 7.477564614744464e-06, "loss": 0.3679, "step": 3704 }, { "epoch": 1.205270006506181, "grad_norm": 1.700510493123436, "learning_rate": 7.475920390297733e-06, "loss": 0.3418, "step": 3705 }, { "epoch": 1.2055953155497723, "grad_norm": 2.2969385751309996, "learning_rate": 7.474275811049959e-06, "loss": 0.5195, "step": 3706 }, { "epoch": 1.2059206245933636, "grad_norm": 1.9739736521407085, "learning_rate": 7.472630877236811e-06, "loss": 0.3743, "step": 3707 }, { "epoch": 1.206245933636955, "grad_norm": 1.7005956827170225, "learning_rate": 7.470985589094011e-06, "loss": 0.3694, "step": 3708 }, { "epoch": 1.2065712426805466, "grad_norm": 1.8415022522587776, "learning_rate": 7.469339946857328e-06, "loss": 0.3656, "step": 3709 }, { "epoch": 1.206896551724138, "grad_norm": 1.873085959448262, "learning_rate": 7.4676939507625846e-06, "loss": 0.3956, "step": 3710 }, { "epoch": 1.2072218607677294, "grad_norm": 1.8671766951459403, "learning_rate": 7.466047601045652e-06, "loss": 0.3286, "step": 3711 }, { "epoch": 1.2075471698113207, "grad_norm": 1.7972460701440187, "learning_rate": 7.464400897942454e-06, "loss": 0.3682, "step": 3712 }, { "epoch": 1.2078724788549122, "grad_norm": 1.9431009475295637, "learning_rate": 7.462753841688965e-06, "loss": 0.4343, "step": 3713 }, { "epoch": 1.2081977878985035, "grad_norm": 2.0417332799529695, "learning_rate": 7.461106432521208e-06, "loss": 0.4465, "step": 3714 }, { "epoch": 1.208523096942095, "grad_norm": 1.6752505706226481, "learning_rate": 7.4594586706752594e-06, "loss": 0.3285, "step": 3715 }, { "epoch": 1.2088484059856863, "grad_norm": 1.6384344175750287, "learning_rate": 7.457810556387242e-06, "loss": 0.3584, "step": 3716 }, { "epoch": 1.2091737150292778, "grad_norm": 1.6858850060529544, "learning_rate": 7.456162089893334e-06, "loss": 0.2849, "step": 3717 }, { "epoch": 1.2094990240728691, "grad_norm": 2.1525592239591664, "learning_rate": 7.45451327142976e-06, "loss": 0.445, "step": 3718 }, { "epoch": 1.2098243331164606, "grad_norm": 1.8888087788394639, "learning_rate": 7.452864101232798e-06, "loss": 0.3356, "step": 3719 }, { "epoch": 1.2101496421600522, "grad_norm": 1.8858710479204257, "learning_rate": 7.451214579538776e-06, "loss": 0.3765, "step": 3720 }, { "epoch": 1.2104749512036435, "grad_norm": 1.831970141212628, "learning_rate": 7.449564706584069e-06, "loss": 0.3902, "step": 3721 }, { "epoch": 1.2108002602472347, "grad_norm": 1.7901650154852604, "learning_rate": 7.447914482605108e-06, "loss": 0.4102, "step": 3722 }, { "epoch": 1.2111255692908263, "grad_norm": 2.079611592033664, "learning_rate": 7.446263907838369e-06, "loss": 0.3516, "step": 3723 }, { "epoch": 1.2114508783344178, "grad_norm": 1.905646343876103, "learning_rate": 7.4446129825203795e-06, "loss": 0.4148, "step": 3724 }, { "epoch": 1.211776187378009, "grad_norm": 2.0043380521549232, "learning_rate": 7.442961706887721e-06, "loss": 0.4583, "step": 3725 }, { "epoch": 1.2121014964216006, "grad_norm": 1.624158505836114, "learning_rate": 7.441310081177021e-06, "loss": 0.3043, "step": 3726 }, { "epoch": 1.2124268054651919, "grad_norm": 1.9557519667463492, "learning_rate": 7.43965810562496e-06, "loss": 0.3911, "step": 3727 }, { "epoch": 1.2127521145087834, "grad_norm": 1.7804158540492665, "learning_rate": 7.438005780468263e-06, "loss": 0.3877, "step": 3728 }, { "epoch": 1.2130774235523747, "grad_norm": 1.5843382249560514, "learning_rate": 7.436353105943714e-06, "loss": 0.3429, "step": 3729 }, { "epoch": 1.2134027325959662, "grad_norm": 1.9570495092820677, "learning_rate": 7.43470008228814e-06, "loss": 0.4246, "step": 3730 }, { "epoch": 1.2137280416395575, "grad_norm": 2.3702565486387845, "learning_rate": 7.433046709738422e-06, "loss": 0.4501, "step": 3731 }, { "epoch": 1.214053350683149, "grad_norm": 1.9887693411973575, "learning_rate": 7.431392988531487e-06, "loss": 0.4371, "step": 3732 }, { "epoch": 1.2143786597267403, "grad_norm": 1.7934268667750723, "learning_rate": 7.429738918904315e-06, "loss": 0.3762, "step": 3733 }, { "epoch": 1.2147039687703318, "grad_norm": 1.7856873013820442, "learning_rate": 7.428084501093937e-06, "loss": 0.3891, "step": 3734 }, { "epoch": 1.2150292778139233, "grad_norm": 1.6263608824652738, "learning_rate": 7.426429735337432e-06, "loss": 0.381, "step": 3735 }, { "epoch": 1.2153545868575146, "grad_norm": 1.8402629177827332, "learning_rate": 7.424774621871927e-06, "loss": 0.3486, "step": 3736 }, { "epoch": 1.2156798959011061, "grad_norm": 1.751843543026903, "learning_rate": 7.423119160934602e-06, "loss": 0.3591, "step": 3737 }, { "epoch": 1.2160052049446974, "grad_norm": 1.8591809226233482, "learning_rate": 7.421463352762686e-06, "loss": 0.366, "step": 3738 }, { "epoch": 1.216330513988289, "grad_norm": 2.2365508105069765, "learning_rate": 7.419807197593459e-06, "loss": 0.4785, "step": 3739 }, { "epoch": 1.2166558230318802, "grad_norm": 1.5274872503700967, "learning_rate": 7.418150695664245e-06, "loss": 0.2992, "step": 3740 }, { "epoch": 1.2169811320754718, "grad_norm": 1.879759986761919, "learning_rate": 7.416493847212423e-06, "loss": 0.4253, "step": 3741 }, { "epoch": 1.217306441119063, "grad_norm": 2.2322402293244066, "learning_rate": 7.414836652475424e-06, "loss": 0.4846, "step": 3742 }, { "epoch": 1.2176317501626546, "grad_norm": 1.8981243585033492, "learning_rate": 7.41317911169072e-06, "loss": 0.4508, "step": 3743 }, { "epoch": 1.2179570592062459, "grad_norm": 2.0723280278470044, "learning_rate": 7.411521225095844e-06, "loss": 0.3848, "step": 3744 }, { "epoch": 1.2182823682498374, "grad_norm": 1.9978972978319955, "learning_rate": 7.409862992928365e-06, "loss": 0.4838, "step": 3745 }, { "epoch": 1.2186076772934287, "grad_norm": 2.063124945638725, "learning_rate": 7.4082044154259145e-06, "loss": 0.3362, "step": 3746 }, { "epoch": 1.2189329863370202, "grad_norm": 1.7155936093578301, "learning_rate": 7.406545492826165e-06, "loss": 0.3468, "step": 3747 }, { "epoch": 1.2192582953806115, "grad_norm": 1.9405617493675578, "learning_rate": 7.404886225366843e-06, "loss": 0.3792, "step": 3748 }, { "epoch": 1.219583604424203, "grad_norm": 1.7671232316748104, "learning_rate": 7.403226613285719e-06, "loss": 0.3523, "step": 3749 }, { "epoch": 1.2199089134677945, "grad_norm": 1.725270388764906, "learning_rate": 7.401566656820621e-06, "loss": 0.3747, "step": 3750 }, { "epoch": 1.2202342225113858, "grad_norm": 1.8629751717842284, "learning_rate": 7.399906356209419e-06, "loss": 0.4086, "step": 3751 }, { "epoch": 1.2205595315549773, "grad_norm": 1.7012669542751848, "learning_rate": 7.398245711690037e-06, "loss": 0.335, "step": 3752 }, { "epoch": 1.2208848405985686, "grad_norm": 2.2006842290168174, "learning_rate": 7.396584723500444e-06, "loss": 0.4267, "step": 3753 }, { "epoch": 1.2212101496421601, "grad_norm": 1.8398160334882379, "learning_rate": 7.394923391878664e-06, "loss": 0.4788, "step": 3754 }, { "epoch": 1.2215354586857514, "grad_norm": 1.7786845482918576, "learning_rate": 7.393261717062764e-06, "loss": 0.3861, "step": 3755 }, { "epoch": 1.221860767729343, "grad_norm": 1.6198667810452692, "learning_rate": 7.3915996992908656e-06, "loss": 0.3468, "step": 3756 }, { "epoch": 1.2221860767729342, "grad_norm": 1.7159557272042596, "learning_rate": 7.389937338801133e-06, "loss": 0.3881, "step": 3757 }, { "epoch": 1.2225113858165257, "grad_norm": 1.9504259419114929, "learning_rate": 7.388274635831789e-06, "loss": 0.4494, "step": 3758 }, { "epoch": 1.222836694860117, "grad_norm": 2.1007700867968184, "learning_rate": 7.386611590621097e-06, "loss": 0.453, "step": 3759 }, { "epoch": 1.2231620039037086, "grad_norm": 1.870131645794676, "learning_rate": 7.384948203407372e-06, "loss": 0.4313, "step": 3760 }, { "epoch": 1.2234873129472998, "grad_norm": 1.9210943010397754, "learning_rate": 7.38328447442898e-06, "loss": 0.4945, "step": 3761 }, { "epoch": 1.2238126219908914, "grad_norm": 1.7627012603245107, "learning_rate": 7.381620403924333e-06, "loss": 0.4145, "step": 3762 }, { "epoch": 1.2241379310344827, "grad_norm": 1.875239197532312, "learning_rate": 7.379955992131894e-06, "loss": 0.4308, "step": 3763 }, { "epoch": 1.2244632400780742, "grad_norm": 1.8675305524626873, "learning_rate": 7.378291239290176e-06, "loss": 0.3549, "step": 3764 }, { "epoch": 1.2247885491216657, "grad_norm": 2.144201496677445, "learning_rate": 7.376626145637736e-06, "loss": 0.4408, "step": 3765 }, { "epoch": 1.225113858165257, "grad_norm": 1.9255212634557053, "learning_rate": 7.374960711413186e-06, "loss": 0.3925, "step": 3766 }, { "epoch": 1.2254391672088485, "grad_norm": 1.7507457175884276, "learning_rate": 7.3732949368551836e-06, "loss": 0.4223, "step": 3767 }, { "epoch": 1.2257644762524398, "grad_norm": 1.8805563621481116, "learning_rate": 7.371628822202432e-06, "loss": 0.4282, "step": 3768 }, { "epoch": 1.2260897852960313, "grad_norm": 1.9109564496082034, "learning_rate": 7.3699623676936905e-06, "loss": 0.447, "step": 3769 }, { "epoch": 1.2264150943396226, "grad_norm": 1.92063243237351, "learning_rate": 7.3682955735677604e-06, "loss": 0.4087, "step": 3770 }, { "epoch": 1.226740403383214, "grad_norm": 1.9325430718375407, "learning_rate": 7.366628440063497e-06, "loss": 0.4218, "step": 3771 }, { "epoch": 1.2270657124268054, "grad_norm": 1.990601882539685, "learning_rate": 7.3649609674198e-06, "loss": 0.4623, "step": 3772 }, { "epoch": 1.227391021470397, "grad_norm": 1.7059022411520854, "learning_rate": 7.363293155875618e-06, "loss": 0.3505, "step": 3773 }, { "epoch": 1.2277163305139882, "grad_norm": 2.141973393058294, "learning_rate": 7.361625005669953e-06, "loss": 0.3635, "step": 3774 }, { "epoch": 1.2280416395575797, "grad_norm": 1.8099560383527282, "learning_rate": 7.359956517041849e-06, "loss": 0.3662, "step": 3775 }, { "epoch": 1.228366948601171, "grad_norm": 1.6005020629507603, "learning_rate": 7.3582876902304035e-06, "loss": 0.323, "step": 3776 }, { "epoch": 1.2286922576447625, "grad_norm": 1.985835340034494, "learning_rate": 7.356618525474759e-06, "loss": 0.3526, "step": 3777 }, { "epoch": 1.2290175666883538, "grad_norm": 1.8501074982597694, "learning_rate": 7.354949023014108e-06, "loss": 0.4428, "step": 3778 }, { "epoch": 1.2293428757319453, "grad_norm": 2.0362132670791735, "learning_rate": 7.353279183087692e-06, "loss": 0.3901, "step": 3779 }, { "epoch": 1.2296681847755369, "grad_norm": 1.914746653408976, "learning_rate": 7.351609005934801e-06, "loss": 0.4012, "step": 3780 }, { "epoch": 1.2299934938191281, "grad_norm": 1.9135100997158518, "learning_rate": 7.349938491794772e-06, "loss": 0.3401, "step": 3781 }, { "epoch": 1.2303188028627197, "grad_norm": 2.174867668114791, "learning_rate": 7.34826764090699e-06, "loss": 0.4013, "step": 3782 }, { "epoch": 1.230644111906311, "grad_norm": 1.971476472942138, "learning_rate": 7.346596453510888e-06, "loss": 0.3663, "step": 3783 }, { "epoch": 1.2309694209499025, "grad_norm": 1.9842004450063313, "learning_rate": 7.344924929845951e-06, "loss": 0.4141, "step": 3784 }, { "epoch": 1.2312947299934938, "grad_norm": 1.88849501131239, "learning_rate": 7.343253070151708e-06, "loss": 0.4634, "step": 3785 }, { "epoch": 1.2316200390370853, "grad_norm": 1.7270629080010949, "learning_rate": 7.341580874667737e-06, "loss": 0.3857, "step": 3786 }, { "epoch": 1.2319453480806766, "grad_norm": 1.8433279473462776, "learning_rate": 7.339908343633667e-06, "loss": 0.3798, "step": 3787 }, { "epoch": 1.232270657124268, "grad_norm": 1.9171857929528695, "learning_rate": 7.3382354772891704e-06, "loss": 0.4429, "step": 3788 }, { "epoch": 1.2325959661678594, "grad_norm": 1.8089160156201274, "learning_rate": 7.336562275873969e-06, "loss": 0.3331, "step": 3789 }, { "epoch": 1.232921275211451, "grad_norm": 1.7355907789438898, "learning_rate": 7.334888739627839e-06, "loss": 0.4504, "step": 3790 }, { "epoch": 1.2332465842550424, "grad_norm": 2.037324259578292, "learning_rate": 7.333214868790593e-06, "loss": 0.4565, "step": 3791 }, { "epoch": 1.2335718932986337, "grad_norm": 1.9638427791916748, "learning_rate": 7.331540663602104e-06, "loss": 0.4263, "step": 3792 }, { "epoch": 1.233897202342225, "grad_norm": 2.154854697323872, "learning_rate": 7.329866124302281e-06, "loss": 0.4241, "step": 3793 }, { "epoch": 1.2342225113858165, "grad_norm": 1.817867609777191, "learning_rate": 7.3281912511310895e-06, "loss": 0.3936, "step": 3794 }, { "epoch": 1.234547820429408, "grad_norm": 1.9753284973753495, "learning_rate": 7.326516044328538e-06, "loss": 0.3806, "step": 3795 }, { "epoch": 1.2348731294729993, "grad_norm": 1.8394151119259514, "learning_rate": 7.324840504134687e-06, "loss": 0.3876, "step": 3796 }, { "epoch": 1.2351984385165908, "grad_norm": 1.9519806572701437, "learning_rate": 7.323164630789643e-06, "loss": 0.3993, "step": 3797 }, { "epoch": 1.2355237475601821, "grad_norm": 1.8272438247804932, "learning_rate": 7.321488424533558e-06, "loss": 0.3354, "step": 3798 }, { "epoch": 1.2358490566037736, "grad_norm": 1.6766795382950854, "learning_rate": 7.319811885606633e-06, "loss": 0.3478, "step": 3799 }, { "epoch": 1.236174365647365, "grad_norm": 1.892920881011204, "learning_rate": 7.31813501424912e-06, "loss": 0.3604, "step": 3800 }, { "epoch": 1.2364996746909565, "grad_norm": 1.7821618869335965, "learning_rate": 7.316457810701312e-06, "loss": 0.3889, "step": 3801 }, { "epoch": 1.2368249837345477, "grad_norm": 1.568379941564889, "learning_rate": 7.314780275203556e-06, "loss": 0.2594, "step": 3802 }, { "epoch": 1.2371502927781393, "grad_norm": 1.6507704624549158, "learning_rate": 7.313102407996243e-06, "loss": 0.3892, "step": 3803 }, { "epoch": 1.2374756018217306, "grad_norm": 1.9031209266019948, "learning_rate": 7.311424209319813e-06, "loss": 0.3599, "step": 3804 }, { "epoch": 1.237800910865322, "grad_norm": 1.848353472917856, "learning_rate": 7.309745679414751e-06, "loss": 0.3596, "step": 3805 }, { "epoch": 1.2381262199089136, "grad_norm": 1.8531076463952125, "learning_rate": 7.308066818521595e-06, "loss": 0.474, "step": 3806 }, { "epoch": 1.2384515289525049, "grad_norm": 1.7920608355708403, "learning_rate": 7.306387626880925e-06, "loss": 0.4149, "step": 3807 }, { "epoch": 1.2387768379960962, "grad_norm": 1.879087293081075, "learning_rate": 7.304708104733372e-06, "loss": 0.4397, "step": 3808 }, { "epoch": 1.2391021470396877, "grad_norm": 2.1269205300166223, "learning_rate": 7.303028252319608e-06, "loss": 0.4076, "step": 3809 }, { "epoch": 1.2394274560832792, "grad_norm": 1.5245571387856756, "learning_rate": 7.301348069880361e-06, "loss": 0.3117, "step": 3810 }, { "epoch": 1.2397527651268705, "grad_norm": 1.8672192827671, "learning_rate": 7.2996675576564e-06, "loss": 0.433, "step": 3811 }, { "epoch": 1.240078074170462, "grad_norm": 1.8455122260577617, "learning_rate": 7.297986715888547e-06, "loss": 0.4166, "step": 3812 }, { "epoch": 1.2404033832140533, "grad_norm": 1.9693105498009318, "learning_rate": 7.296305544817664e-06, "loss": 0.4548, "step": 3813 }, { "epoch": 1.2407286922576448, "grad_norm": 1.527927782169179, "learning_rate": 7.294624044684666e-06, "loss": 0.3337, "step": 3814 }, { "epoch": 1.241054001301236, "grad_norm": 1.8240297604282445, "learning_rate": 7.292942215730512e-06, "loss": 0.4379, "step": 3815 }, { "epoch": 1.2413793103448276, "grad_norm": 1.8615875418648302, "learning_rate": 7.2912600581962115e-06, "loss": 0.4034, "step": 3816 }, { "epoch": 1.241704619388419, "grad_norm": 2.2321957198677036, "learning_rate": 7.2895775723228144e-06, "loss": 0.5374, "step": 3817 }, { "epoch": 1.2420299284320104, "grad_norm": 1.9453775530449955, "learning_rate": 7.287894758351426e-06, "loss": 0.4318, "step": 3818 }, { "epoch": 1.2423552374756017, "grad_norm": 1.747930156752555, "learning_rate": 7.286211616523193e-06, "loss": 0.3528, "step": 3819 }, { "epoch": 1.2426805465191932, "grad_norm": 1.9355114976889938, "learning_rate": 7.284528147079313e-06, "loss": 0.4756, "step": 3820 }, { "epoch": 1.2430058555627848, "grad_norm": 1.9997708183068175, "learning_rate": 7.282844350261023e-06, "loss": 0.3604, "step": 3821 }, { "epoch": 1.243331164606376, "grad_norm": 2.1337609593043245, "learning_rate": 7.281160226309617e-06, "loss": 0.506, "step": 3822 }, { "epoch": 1.2436564736499676, "grad_norm": 1.9071738314052835, "learning_rate": 7.27947577546643e-06, "loss": 0.4268, "step": 3823 }, { "epoch": 1.2439817826935589, "grad_norm": 1.7047930332044587, "learning_rate": 7.277790997972845e-06, "loss": 0.4064, "step": 3824 }, { "epoch": 1.2443070917371504, "grad_norm": 1.964520536308463, "learning_rate": 7.27610589407029e-06, "loss": 0.3793, "step": 3825 }, { "epoch": 1.2446324007807417, "grad_norm": 1.654054738062978, "learning_rate": 7.274420464000245e-06, "loss": 0.3041, "step": 3826 }, { "epoch": 1.2449577098243332, "grad_norm": 1.901056984393162, "learning_rate": 7.2727347080042275e-06, "loss": 0.3986, "step": 3827 }, { "epoch": 1.2452830188679245, "grad_norm": 2.1211002623917987, "learning_rate": 7.271048626323814e-06, "loss": 0.4115, "step": 3828 }, { "epoch": 1.245608327911516, "grad_norm": 1.8216275025640392, "learning_rate": 7.269362219200618e-06, "loss": 0.3519, "step": 3829 }, { "epoch": 1.2459336369551073, "grad_norm": 1.9061377458012219, "learning_rate": 7.267675486876302e-06, "loss": 0.3568, "step": 3830 }, { "epoch": 1.2462589459986988, "grad_norm": 2.2104196972126, "learning_rate": 7.2659884295925764e-06, "loss": 0.4696, "step": 3831 }, { "epoch": 1.24658425504229, "grad_norm": 1.835414320916087, "learning_rate": 7.2643010475912e-06, "loss": 0.3658, "step": 3832 }, { "epoch": 1.2469095640858816, "grad_norm": 1.9066728876168029, "learning_rate": 7.262613341113972e-06, "loss": 0.387, "step": 3833 }, { "epoch": 1.247234873129473, "grad_norm": 1.7624786730747657, "learning_rate": 7.260925310402745e-06, "loss": 0.3345, "step": 3834 }, { "epoch": 1.2475601821730644, "grad_norm": 1.9653303934600506, "learning_rate": 7.259236955699412e-06, "loss": 0.4292, "step": 3835 }, { "epoch": 1.247885491216656, "grad_norm": 1.9265294139872993, "learning_rate": 7.257548277245918e-06, "loss": 0.4243, "step": 3836 }, { "epoch": 1.2482108002602472, "grad_norm": 1.8694804139887136, "learning_rate": 7.255859275284249e-06, "loss": 0.3973, "step": 3837 }, { "epoch": 1.2485361093038387, "grad_norm": 1.746257271565063, "learning_rate": 7.254169950056444e-06, "loss": 0.3495, "step": 3838 }, { "epoch": 1.24886141834743, "grad_norm": 2.107709358664245, "learning_rate": 7.252480301804582e-06, "loss": 0.4028, "step": 3839 }, { "epoch": 1.2491867273910215, "grad_norm": 1.9387456349903875, "learning_rate": 7.25079033077079e-06, "loss": 0.4217, "step": 3840 }, { "epoch": 1.2495120364346128, "grad_norm": 2.1346645266031166, "learning_rate": 7.249100037197243e-06, "loss": 0.429, "step": 3841 }, { "epoch": 1.2498373454782044, "grad_norm": 1.84161659419398, "learning_rate": 7.247409421326162e-06, "loss": 0.46, "step": 3842 }, { "epoch": 1.2501626545217956, "grad_norm": 1.7273051291571109, "learning_rate": 7.245718483399812e-06, "loss": 0.3429, "step": 3843 }, { "epoch": 1.2504879635653872, "grad_norm": 2.3423352660747145, "learning_rate": 7.244027223660506e-06, "loss": 0.4737, "step": 3844 }, { "epoch": 1.2508132726089785, "grad_norm": 1.666263153539991, "learning_rate": 7.2423356423506035e-06, "loss": 0.3253, "step": 3845 }, { "epoch": 1.25113858165257, "grad_norm": 2.091692606270953, "learning_rate": 7.240643739712507e-06, "loss": 0.3763, "step": 3846 }, { "epoch": 1.2514638906961615, "grad_norm": 1.823929567230261, "learning_rate": 7.238951515988669e-06, "loss": 0.4398, "step": 3847 }, { "epoch": 1.2517891997397528, "grad_norm": 1.992934483977828, "learning_rate": 7.237258971421587e-06, "loss": 0.3761, "step": 3848 }, { "epoch": 1.252114508783344, "grad_norm": 2.0100852119920463, "learning_rate": 7.235566106253802e-06, "loss": 0.4804, "step": 3849 }, { "epoch": 1.2524398178269356, "grad_norm": 2.0106894391022054, "learning_rate": 7.233872920727903e-06, "loss": 0.4338, "step": 3850 }, { "epoch": 1.252765126870527, "grad_norm": 1.6691111006936021, "learning_rate": 7.232179415086526e-06, "loss": 0.3966, "step": 3851 }, { "epoch": 1.2530904359141184, "grad_norm": 1.809567298516728, "learning_rate": 7.23048558957235e-06, "loss": 0.3573, "step": 3852 }, { "epoch": 1.2534157449577097, "grad_norm": 1.8804351526775707, "learning_rate": 7.2287914444281015e-06, "loss": 0.3852, "step": 3853 }, { "epoch": 1.2537410540013012, "grad_norm": 1.6645727196720455, "learning_rate": 7.227096979896554e-06, "loss": 0.3514, "step": 3854 }, { "epoch": 1.2540663630448927, "grad_norm": 1.7447147543123809, "learning_rate": 7.225402196220524e-06, "loss": 0.3106, "step": 3855 }, { "epoch": 1.254391672088484, "grad_norm": 2.0418611748438193, "learning_rate": 7.223707093642875e-06, "loss": 0.5134, "step": 3856 }, { "epoch": 1.2547169811320755, "grad_norm": 1.6856934966710106, "learning_rate": 7.222011672406517e-06, "loss": 0.369, "step": 3857 }, { "epoch": 1.2550422901756668, "grad_norm": 1.931450116478559, "learning_rate": 7.220315932754407e-06, "loss": 0.3702, "step": 3858 }, { "epoch": 1.2553675992192583, "grad_norm": 1.5201507686071178, "learning_rate": 7.218619874929541e-06, "loss": 0.3137, "step": 3859 }, { "epoch": 1.2556929082628496, "grad_norm": 1.967166093689716, "learning_rate": 7.216923499174968e-06, "loss": 0.4894, "step": 3860 }, { "epoch": 1.2560182173064411, "grad_norm": 2.015856604998651, "learning_rate": 7.21522680573378e-06, "loss": 0.4089, "step": 3861 }, { "epoch": 1.2563435263500327, "grad_norm": 1.852058183507189, "learning_rate": 7.213529794849113e-06, "loss": 0.3989, "step": 3862 }, { "epoch": 1.256668835393624, "grad_norm": 1.785518759711623, "learning_rate": 7.2118324667641524e-06, "loss": 0.4442, "step": 3863 }, { "epoch": 1.2569941444372152, "grad_norm": 1.757898117283754, "learning_rate": 7.210134821722123e-06, "loss": 0.3815, "step": 3864 }, { "epoch": 1.2573194534808068, "grad_norm": 1.9513829621476133, "learning_rate": 7.208436859966301e-06, "loss": 0.3893, "step": 3865 }, { "epoch": 1.2576447625243983, "grad_norm": 1.9364353816236537, "learning_rate": 7.206738581740004e-06, "loss": 0.4519, "step": 3866 }, { "epoch": 1.2579700715679896, "grad_norm": 2.1729453320862975, "learning_rate": 7.205039987286597e-06, "loss": 0.5337, "step": 3867 }, { "epoch": 1.258295380611581, "grad_norm": 1.804462311443845, "learning_rate": 7.203341076849489e-06, "loss": 0.3713, "step": 3868 }, { "epoch": 1.2586206896551724, "grad_norm": 2.079051757549074, "learning_rate": 7.201641850672137e-06, "loss": 0.3615, "step": 3869 }, { "epoch": 1.258945998698764, "grad_norm": 1.860003245657039, "learning_rate": 7.199942308998038e-06, "loss": 0.399, "step": 3870 }, { "epoch": 1.2592713077423552, "grad_norm": 1.866248609878607, "learning_rate": 7.198242452070743e-06, "loss": 0.3641, "step": 3871 }, { "epoch": 1.2595966167859467, "grad_norm": 1.717061317093126, "learning_rate": 7.196542280133836e-06, "loss": 0.3564, "step": 3872 }, { "epoch": 1.259921925829538, "grad_norm": 1.9346047139725155, "learning_rate": 7.1948417934309556e-06, "loss": 0.5016, "step": 3873 }, { "epoch": 1.2602472348731295, "grad_norm": 1.638951820826241, "learning_rate": 7.1931409922057835e-06, "loss": 0.2957, "step": 3874 }, { "epoch": 1.2605725439167208, "grad_norm": 2.0805767657977654, "learning_rate": 7.191439876702045e-06, "loss": 0.4935, "step": 3875 }, { "epoch": 1.2608978529603123, "grad_norm": 2.139155179336006, "learning_rate": 7.1897384471635114e-06, "loss": 0.4423, "step": 3876 }, { "epoch": 1.2612231620039038, "grad_norm": 2.1920994011987758, "learning_rate": 7.188036703833999e-06, "loss": 0.4844, "step": 3877 }, { "epoch": 1.2615484710474951, "grad_norm": 1.6306489752286861, "learning_rate": 7.186334646957367e-06, "loss": 0.3913, "step": 3878 }, { "epoch": 1.2618737800910864, "grad_norm": 1.716804039244586, "learning_rate": 7.1846322767775224e-06, "loss": 0.3635, "step": 3879 }, { "epoch": 1.262199089134678, "grad_norm": 1.9052717967341974, "learning_rate": 7.182929593538416e-06, "loss": 0.4165, "step": 3880 }, { "epoch": 1.2625243981782694, "grad_norm": 1.9085226214855717, "learning_rate": 7.1812265974840435e-06, "loss": 0.4116, "step": 3881 }, { "epoch": 1.2628497072218607, "grad_norm": 1.682481613731416, "learning_rate": 7.179523288858445e-06, "loss": 0.3184, "step": 3882 }, { "epoch": 1.2631750162654523, "grad_norm": 2.0260984217880966, "learning_rate": 7.1778196679057065e-06, "loss": 0.3715, "step": 3883 }, { "epoch": 1.2635003253090435, "grad_norm": 1.8943641083757667, "learning_rate": 7.176115734869957e-06, "loss": 0.3751, "step": 3884 }, { "epoch": 1.263825634352635, "grad_norm": 2.198680355361005, "learning_rate": 7.174411489995373e-06, "loss": 0.4426, "step": 3885 }, { "epoch": 1.2641509433962264, "grad_norm": 2.0989859028887707, "learning_rate": 7.172706933526172e-06, "loss": 0.4578, "step": 3886 }, { "epoch": 1.2644762524398179, "grad_norm": 1.598070329186216, "learning_rate": 7.17100206570662e-06, "loss": 0.3655, "step": 3887 }, { "epoch": 1.2648015614834092, "grad_norm": 1.8058023133733527, "learning_rate": 7.169296886781023e-06, "loss": 0.4346, "step": 3888 }, { "epoch": 1.2651268705270007, "grad_norm": 1.7812389420119221, "learning_rate": 7.167591396993736e-06, "loss": 0.3746, "step": 3889 }, { "epoch": 1.265452179570592, "grad_norm": 1.8047887449298334, "learning_rate": 7.165885596589157e-06, "loss": 0.363, "step": 3890 }, { "epoch": 1.2657774886141835, "grad_norm": 1.9136014275744393, "learning_rate": 7.164179485811728e-06, "loss": 0.3971, "step": 3891 }, { "epoch": 1.266102797657775, "grad_norm": 1.7195291199405456, "learning_rate": 7.162473064905935e-06, "loss": 0.3404, "step": 3892 }, { "epoch": 1.2664281067013663, "grad_norm": 2.3863898666842713, "learning_rate": 7.1607663341163105e-06, "loss": 0.483, "step": 3893 }, { "epoch": 1.2667534157449576, "grad_norm": 2.0152818822141305, "learning_rate": 7.159059293687429e-06, "loss": 0.4464, "step": 3894 }, { "epoch": 1.267078724788549, "grad_norm": 1.9316023270322824, "learning_rate": 7.157351943863913e-06, "loss": 0.376, "step": 3895 }, { "epoch": 1.2674040338321406, "grad_norm": 1.9392162021331505, "learning_rate": 7.155644284890424e-06, "loss": 0.4103, "step": 3896 }, { "epoch": 1.267729342875732, "grad_norm": 2.2200627539915763, "learning_rate": 7.153936317011672e-06, "loss": 0.3567, "step": 3897 }, { "epoch": 1.2680546519193234, "grad_norm": 1.7154885756236766, "learning_rate": 7.152228040472409e-06, "loss": 0.2838, "step": 3898 }, { "epoch": 1.2683799609629147, "grad_norm": 1.768482264240266, "learning_rate": 7.1505194555174325e-06, "loss": 0.4326, "step": 3899 }, { "epoch": 1.2687052700065062, "grad_norm": 2.2618562439256027, "learning_rate": 7.148810562391586e-06, "loss": 0.3761, "step": 3900 }, { "epoch": 1.2690305790500975, "grad_norm": 1.9488032871068468, "learning_rate": 7.1471013613397524e-06, "loss": 0.4071, "step": 3901 }, { "epoch": 1.269355888093689, "grad_norm": 1.8518328507174935, "learning_rate": 7.145391852606861e-06, "loss": 0.3865, "step": 3902 }, { "epoch": 1.2696811971372803, "grad_norm": 2.0473688016014298, "learning_rate": 7.143682036437889e-06, "loss": 0.3556, "step": 3903 }, { "epoch": 1.2700065061808719, "grad_norm": 2.376066590771817, "learning_rate": 7.14197191307785e-06, "loss": 0.5391, "step": 3904 }, { "epoch": 1.2703318152244631, "grad_norm": 1.905850400787604, "learning_rate": 7.140261482771809e-06, "loss": 0.4288, "step": 3905 }, { "epoch": 1.2706571242680547, "grad_norm": 2.107109751891516, "learning_rate": 7.138550745764869e-06, "loss": 0.3685, "step": 3906 }, { "epoch": 1.2709824333116462, "grad_norm": 1.850840591955562, "learning_rate": 7.136839702302183e-06, "loss": 0.4538, "step": 3907 }, { "epoch": 1.2713077423552375, "grad_norm": 2.003482095922194, "learning_rate": 7.1351283526289415e-06, "loss": 0.3662, "step": 3908 }, { "epoch": 1.2716330513988288, "grad_norm": 1.8579547503492406, "learning_rate": 7.133416696990384e-06, "loss": 0.3618, "step": 3909 }, { "epoch": 1.2719583604424203, "grad_norm": 2.061520563972565, "learning_rate": 7.131704735631792e-06, "loss": 0.4567, "step": 3910 }, { "epoch": 1.2722836694860118, "grad_norm": 1.574509676043371, "learning_rate": 7.12999246879849e-06, "loss": 0.3325, "step": 3911 }, { "epoch": 1.272608978529603, "grad_norm": 2.278114225097416, "learning_rate": 7.128279896735845e-06, "loss": 0.4227, "step": 3912 }, { "epoch": 1.2729342875731946, "grad_norm": 2.13008754588245, "learning_rate": 7.126567019689275e-06, "loss": 0.4224, "step": 3913 }, { "epoch": 1.273259596616786, "grad_norm": 1.6462766983463286, "learning_rate": 7.124853837904232e-06, "loss": 0.3335, "step": 3914 }, { "epoch": 1.2735849056603774, "grad_norm": 1.8787952310966765, "learning_rate": 7.1231403516262165e-06, "loss": 0.3737, "step": 3915 }, { "epoch": 1.2739102147039687, "grad_norm": 1.7999362265141878, "learning_rate": 7.121426561100775e-06, "loss": 0.4152, "step": 3916 }, { "epoch": 1.2742355237475602, "grad_norm": 2.033952812065207, "learning_rate": 7.119712466573492e-06, "loss": 0.4302, "step": 3917 }, { "epoch": 1.2745608327911517, "grad_norm": 2.030169057146954, "learning_rate": 7.1179980682899975e-06, "loss": 0.4511, "step": 3918 }, { "epoch": 1.274886141834743, "grad_norm": 1.8168749909439932, "learning_rate": 7.116283366495971e-06, "loss": 0.411, "step": 3919 }, { "epoch": 1.2752114508783343, "grad_norm": 2.0140323425010984, "learning_rate": 7.114568361437124e-06, "loss": 0.4133, "step": 3920 }, { "epoch": 1.2755367599219258, "grad_norm": 2.3071329013894784, "learning_rate": 7.112853053359222e-06, "loss": 0.454, "step": 3921 }, { "epoch": 1.2758620689655173, "grad_norm": 1.8292628223349234, "learning_rate": 7.1111374425080695e-06, "loss": 0.3577, "step": 3922 }, { "epoch": 1.2761873780091086, "grad_norm": 1.862240843997679, "learning_rate": 7.109421529129514e-06, "loss": 0.3576, "step": 3923 }, { "epoch": 1.2765126870527, "grad_norm": 1.774452646611934, "learning_rate": 7.107705313469444e-06, "loss": 0.3822, "step": 3924 }, { "epoch": 1.2768379960962914, "grad_norm": 1.9427507300335505, "learning_rate": 7.1059887957738005e-06, "loss": 0.4271, "step": 3925 }, { "epoch": 1.277163305139883, "grad_norm": 2.297929543428564, "learning_rate": 7.104271976288557e-06, "loss": 0.4073, "step": 3926 }, { "epoch": 1.2774886141834743, "grad_norm": 1.798553608543225, "learning_rate": 7.102554855259738e-06, "loss": 0.3464, "step": 3927 }, { "epoch": 1.2778139232270658, "grad_norm": 2.027231821034803, "learning_rate": 7.100837432933404e-06, "loss": 0.4167, "step": 3928 }, { "epoch": 1.278139232270657, "grad_norm": 2.1586072871894504, "learning_rate": 7.099119709555668e-06, "loss": 0.4262, "step": 3929 }, { "epoch": 1.2784645413142486, "grad_norm": 1.7993436234005638, "learning_rate": 7.097401685372676e-06, "loss": 0.3537, "step": 3930 }, { "epoch": 1.2787898503578399, "grad_norm": 1.8926575827798815, "learning_rate": 7.095683360630624e-06, "loss": 0.3555, "step": 3931 }, { "epoch": 1.2791151594014314, "grad_norm": 2.2340707184737045, "learning_rate": 7.09396473557575e-06, "loss": 0.4113, "step": 3932 }, { "epoch": 1.279440468445023, "grad_norm": 1.9816404745685694, "learning_rate": 7.092245810454334e-06, "loss": 0.4094, "step": 3933 }, { "epoch": 1.2797657774886142, "grad_norm": 2.016825284491454, "learning_rate": 7.090526585512696e-06, "loss": 0.4736, "step": 3934 }, { "epoch": 1.2800910865322055, "grad_norm": 1.9750705654354554, "learning_rate": 7.0888070609972055e-06, "loss": 0.4844, "step": 3935 }, { "epoch": 1.280416395575797, "grad_norm": 2.249024476555899, "learning_rate": 7.087087237154271e-06, "loss": 0.4202, "step": 3936 }, { "epoch": 1.2807417046193885, "grad_norm": 1.8868435564971828, "learning_rate": 7.085367114230344e-06, "loss": 0.3846, "step": 3937 }, { "epoch": 1.2810670136629798, "grad_norm": 1.6001433945633696, "learning_rate": 7.083646692471918e-06, "loss": 0.3107, "step": 3938 }, { "epoch": 1.281392322706571, "grad_norm": 1.7845509357679914, "learning_rate": 7.081925972125534e-06, "loss": 0.3525, "step": 3939 }, { "epoch": 1.2817176317501626, "grad_norm": 1.6782183022473667, "learning_rate": 7.080204953437768e-06, "loss": 0.3544, "step": 3940 }, { "epoch": 1.2820429407937541, "grad_norm": 1.8832253053724872, "learning_rate": 7.0784836366552465e-06, "loss": 0.4243, "step": 3941 }, { "epoch": 1.2823682498373454, "grad_norm": 2.009384738707545, "learning_rate": 7.076762022024634e-06, "loss": 0.399, "step": 3942 }, { "epoch": 1.282693558880937, "grad_norm": 1.932285580007033, "learning_rate": 7.075040109792639e-06, "loss": 0.4451, "step": 3943 }, { "epoch": 1.2830188679245282, "grad_norm": 1.8468983104490178, "learning_rate": 7.073317900206012e-06, "loss": 0.389, "step": 3944 }, { "epoch": 1.2833441769681198, "grad_norm": 1.7624497931219176, "learning_rate": 7.0715953935115485e-06, "loss": 0.3443, "step": 3945 }, { "epoch": 1.283669486011711, "grad_norm": 2.0595173053513056, "learning_rate": 7.069872589956083e-06, "loss": 0.4316, "step": 3946 }, { "epoch": 1.2839947950553026, "grad_norm": 1.8315340150408776, "learning_rate": 7.068149489786496e-06, "loss": 0.3673, "step": 3947 }, { "epoch": 1.284320104098894, "grad_norm": 1.9139259664022072, "learning_rate": 7.066426093249707e-06, "loss": 0.4252, "step": 3948 }, { "epoch": 1.2846454131424854, "grad_norm": 1.897220192801261, "learning_rate": 7.064702400592683e-06, "loss": 0.3775, "step": 3949 }, { "epoch": 1.2849707221860767, "grad_norm": 2.196964552543337, "learning_rate": 7.062978412062426e-06, "loss": 0.5251, "step": 3950 }, { "epoch": 1.2852960312296682, "grad_norm": 1.8590579161638323, "learning_rate": 7.06125412790599e-06, "loss": 0.3682, "step": 3951 }, { "epoch": 1.2856213402732597, "grad_norm": 2.1500957341668934, "learning_rate": 7.059529548370462e-06, "loss": 0.4781, "step": 3952 }, { "epoch": 1.285946649316851, "grad_norm": 2.116001458587348, "learning_rate": 7.057804673702977e-06, "loss": 0.4356, "step": 3953 }, { "epoch": 1.2862719583604425, "grad_norm": 2.017953615316193, "learning_rate": 7.056079504150708e-06, "loss": 0.4429, "step": 3954 }, { "epoch": 1.2865972674040338, "grad_norm": 1.82897766330252, "learning_rate": 7.054354039960878e-06, "loss": 0.3916, "step": 3955 }, { "epoch": 1.2869225764476253, "grad_norm": 1.6097577432328358, "learning_rate": 7.052628281380743e-06, "loss": 0.3188, "step": 3956 }, { "epoch": 1.2872478854912166, "grad_norm": 1.994440080665025, "learning_rate": 7.0509022286576065e-06, "loss": 0.4393, "step": 3957 }, { "epoch": 1.2875731945348081, "grad_norm": 1.8551656756920798, "learning_rate": 7.049175882038814e-06, "loss": 0.4817, "step": 3958 }, { "epoch": 1.2878985035783994, "grad_norm": 1.67211841108335, "learning_rate": 7.047449241771751e-06, "loss": 0.3902, "step": 3959 }, { "epoch": 1.288223812621991, "grad_norm": 2.0191561574537396, "learning_rate": 7.045722308103846e-06, "loss": 0.4913, "step": 3960 }, { "epoch": 1.2885491216655822, "grad_norm": 1.554722454567616, "learning_rate": 7.043995081282571e-06, "loss": 0.3118, "step": 3961 }, { "epoch": 1.2888744307091737, "grad_norm": 2.1821910651513625, "learning_rate": 7.042267561555438e-06, "loss": 0.3775, "step": 3962 }, { "epoch": 1.2891997397527653, "grad_norm": 1.9098562069984946, "learning_rate": 7.040539749170002e-06, "loss": 0.3679, "step": 3963 }, { "epoch": 1.2895250487963565, "grad_norm": 1.9424659165215405, "learning_rate": 7.03881164437386e-06, "loss": 0.3651, "step": 3964 }, { "epoch": 1.2898503578399478, "grad_norm": 1.8855539213576755, "learning_rate": 7.037083247414649e-06, "loss": 0.4306, "step": 3965 }, { "epoch": 1.2901756668835394, "grad_norm": 2.20799464915258, "learning_rate": 7.03535455854005e-06, "loss": 0.4632, "step": 3966 }, { "epoch": 1.2905009759271309, "grad_norm": 1.8919221927884988, "learning_rate": 7.033625577997787e-06, "loss": 0.4352, "step": 3967 }, { "epoch": 1.2908262849707222, "grad_norm": 1.8668303467254226, "learning_rate": 7.0318963060356215e-06, "loss": 0.3924, "step": 3968 }, { "epoch": 1.2911515940143137, "grad_norm": 1.9334890140995329, "learning_rate": 7.030166742901361e-06, "loss": 0.4149, "step": 3969 }, { "epoch": 1.291476903057905, "grad_norm": 2.051341402310685, "learning_rate": 7.028436888842853e-06, "loss": 0.3238, "step": 3970 }, { "epoch": 1.2918022121014965, "grad_norm": 1.8634712583384325, "learning_rate": 7.026706744107987e-06, "loss": 0.4379, "step": 3971 }, { "epoch": 1.2921275211450878, "grad_norm": 2.033767657160282, "learning_rate": 7.024976308944692e-06, "loss": 0.4436, "step": 3972 }, { "epoch": 1.2924528301886793, "grad_norm": 2.015986840236182, "learning_rate": 7.023245583600943e-06, "loss": 0.4114, "step": 3973 }, { "epoch": 1.2927781392322706, "grad_norm": 1.7699359662368173, "learning_rate": 7.021514568324752e-06, "loss": 0.3669, "step": 3974 }, { "epoch": 1.293103448275862, "grad_norm": 1.9149068892093253, "learning_rate": 7.019783263364177e-06, "loss": 0.3729, "step": 3975 }, { "epoch": 1.2934287573194534, "grad_norm": 1.7017131835963761, "learning_rate": 7.018051668967312e-06, "loss": 0.3604, "step": 3976 }, { "epoch": 1.293754066363045, "grad_norm": 2.186807562841153, "learning_rate": 7.0163197853822975e-06, "loss": 0.4518, "step": 3977 }, { "epoch": 1.2940793754066364, "grad_norm": 1.9565963213450557, "learning_rate": 7.014587612857314e-06, "loss": 0.3809, "step": 3978 }, { "epoch": 1.2944046844502277, "grad_norm": 1.6914580419026888, "learning_rate": 7.012855151640583e-06, "loss": 0.3512, "step": 3979 }, { "epoch": 1.294729993493819, "grad_norm": 2.1045611744830985, "learning_rate": 7.011122401980366e-06, "loss": 0.4666, "step": 3980 }, { "epoch": 1.2950553025374105, "grad_norm": 2.070754649561451, "learning_rate": 7.009389364124968e-06, "loss": 0.3684, "step": 3981 }, { "epoch": 1.295380611581002, "grad_norm": 1.8684199113936852, "learning_rate": 7.007656038322735e-06, "loss": 0.3401, "step": 3982 }, { "epoch": 1.2957059206245933, "grad_norm": 1.8830162965161166, "learning_rate": 7.0059224248220516e-06, "loss": 0.4086, "step": 3983 }, { "epoch": 1.2960312296681848, "grad_norm": 2.0850633926312003, "learning_rate": 7.004188523871349e-06, "loss": 0.3352, "step": 3984 }, { "epoch": 1.2963565387117761, "grad_norm": 1.7356950698734357, "learning_rate": 7.002454335719093e-06, "loss": 0.3703, "step": 3985 }, { "epoch": 1.2966818477553677, "grad_norm": 1.8503611079246556, "learning_rate": 7.000719860613796e-06, "loss": 0.3698, "step": 3986 }, { "epoch": 1.297007156798959, "grad_norm": 2.056235460575432, "learning_rate": 6.998985098804011e-06, "loss": 0.3948, "step": 3987 }, { "epoch": 1.2973324658425505, "grad_norm": 1.5974276801447123, "learning_rate": 6.9972500505383265e-06, "loss": 0.3567, "step": 3988 }, { "epoch": 1.2976577748861418, "grad_norm": 2.1963220198972238, "learning_rate": 6.995514716065378e-06, "loss": 0.4485, "step": 3989 }, { "epoch": 1.2979830839297333, "grad_norm": 1.7096183359929755, "learning_rate": 6.993779095633842e-06, "loss": 0.3462, "step": 3990 }, { "epoch": 1.2983083929733246, "grad_norm": 1.797334493943504, "learning_rate": 6.99204318949243e-06, "loss": 0.3268, "step": 3991 }, { "epoch": 1.298633702016916, "grad_norm": 2.1446298509946433, "learning_rate": 6.990306997889903e-06, "loss": 0.4649, "step": 3992 }, { "epoch": 1.2989590110605076, "grad_norm": 1.8636885290471763, "learning_rate": 6.988570521075055e-06, "loss": 0.3978, "step": 3993 }, { "epoch": 1.2992843201040989, "grad_norm": 1.8895849507050673, "learning_rate": 6.986833759296726e-06, "loss": 0.3532, "step": 3994 }, { "epoch": 1.2996096291476902, "grad_norm": 1.7755537893455884, "learning_rate": 6.985096712803794e-06, "loss": 0.3876, "step": 3995 }, { "epoch": 1.2999349381912817, "grad_norm": 1.8497846720032238, "learning_rate": 6.9833593818451795e-06, "loss": 0.3884, "step": 3996 }, { "epoch": 1.3002602472348732, "grad_norm": 1.7219346079299789, "learning_rate": 6.9816217666698436e-06, "loss": 0.3966, "step": 3997 }, { "epoch": 1.3005855562784645, "grad_norm": 1.8327411925521908, "learning_rate": 6.9798838675267875e-06, "loss": 0.3808, "step": 3998 }, { "epoch": 1.300910865322056, "grad_norm": 1.7788734032495512, "learning_rate": 6.978145684665053e-06, "loss": 0.3236, "step": 3999 }, { "epoch": 1.3012361743656473, "grad_norm": 1.881844237927837, "learning_rate": 6.976407218333724e-06, "loss": 0.3743, "step": 4000 }, { "epoch": 1.3015614834092388, "grad_norm": 1.9940591019100526, "learning_rate": 6.974668468781921e-06, "loss": 0.3771, "step": 4001 }, { "epoch": 1.3018867924528301, "grad_norm": 1.9284091963640067, "learning_rate": 6.97292943625881e-06, "loss": 0.4043, "step": 4002 }, { "epoch": 1.3022121014964216, "grad_norm": 2.0540640115078745, "learning_rate": 6.971190121013597e-06, "loss": 0.3945, "step": 4003 }, { "epoch": 1.302537410540013, "grad_norm": 1.7977031086108357, "learning_rate": 6.969450523295525e-06, "loss": 0.3338, "step": 4004 }, { "epoch": 1.3028627195836044, "grad_norm": 1.8300221151627831, "learning_rate": 6.9677106433538785e-06, "loss": 0.306, "step": 4005 }, { "epoch": 1.3031880286271957, "grad_norm": 2.2605374063757493, "learning_rate": 6.965970481437987e-06, "loss": 0.4324, "step": 4006 }, { "epoch": 1.3035133376707873, "grad_norm": 1.9608712666168266, "learning_rate": 6.964230037797213e-06, "loss": 0.3471, "step": 4007 }, { "epoch": 1.3038386467143788, "grad_norm": 2.256664777473886, "learning_rate": 6.962489312680965e-06, "loss": 0.3901, "step": 4008 }, { "epoch": 1.30416395575797, "grad_norm": 1.7897357650667354, "learning_rate": 6.960748306338692e-06, "loss": 0.3672, "step": 4009 }, { "epoch": 1.3044892648015614, "grad_norm": 2.293633369037443, "learning_rate": 6.959007019019876e-06, "loss": 0.3854, "step": 4010 }, { "epoch": 1.3048145738451529, "grad_norm": 2.042128090340105, "learning_rate": 6.95726545097405e-06, "loss": 0.4267, "step": 4011 }, { "epoch": 1.3051398828887444, "grad_norm": 1.6517145277251595, "learning_rate": 6.95552360245078e-06, "loss": 0.2962, "step": 4012 }, { "epoch": 1.3054651919323357, "grad_norm": 1.9065528919842785, "learning_rate": 6.953781473699675e-06, "loss": 0.3953, "step": 4013 }, { "epoch": 1.3057905009759272, "grad_norm": 1.9954520787112924, "learning_rate": 6.95203906497038e-06, "loss": 0.3848, "step": 4014 }, { "epoch": 1.3061158100195185, "grad_norm": 2.1787782332466463, "learning_rate": 6.950296376512586e-06, "loss": 0.4386, "step": 4015 }, { "epoch": 1.30644111906311, "grad_norm": 2.2121152604076086, "learning_rate": 6.948553408576023e-06, "loss": 0.4648, "step": 4016 }, { "epoch": 1.3067664281067013, "grad_norm": 1.697417977269865, "learning_rate": 6.946810161410455e-06, "loss": 0.36, "step": 4017 }, { "epoch": 1.3070917371502928, "grad_norm": 2.272211723394754, "learning_rate": 6.945066635265694e-06, "loss": 0.4555, "step": 4018 }, { "epoch": 1.3074170461938843, "grad_norm": 1.936632784713171, "learning_rate": 6.943322830391588e-06, "loss": 0.4297, "step": 4019 }, { "epoch": 1.3077423552374756, "grad_norm": 1.8959977774626393, "learning_rate": 6.941578747038024e-06, "loss": 0.402, "step": 4020 }, { "epoch": 1.308067664281067, "grad_norm": 1.8158607042665975, "learning_rate": 6.93983438545493e-06, "loss": 0.3513, "step": 4021 }, { "epoch": 1.3083929733246584, "grad_norm": 1.682477117679161, "learning_rate": 6.938089745892276e-06, "loss": 0.3511, "step": 4022 }, { "epoch": 1.30871828236825, "grad_norm": 1.9737201609541075, "learning_rate": 6.93634482860007e-06, "loss": 0.391, "step": 4023 }, { "epoch": 1.3090435914118412, "grad_norm": 1.863058112639666, "learning_rate": 6.934599633828359e-06, "loss": 0.401, "step": 4024 }, { "epoch": 1.3093689004554325, "grad_norm": 1.825980761741704, "learning_rate": 6.9328541618272295e-06, "loss": 0.3763, "step": 4025 }, { "epoch": 1.309694209499024, "grad_norm": 1.7520920175069754, "learning_rate": 6.93110841284681e-06, "loss": 0.3376, "step": 4026 }, { "epoch": 1.3100195185426156, "grad_norm": 1.794747177663179, "learning_rate": 6.929362387137267e-06, "loss": 0.3348, "step": 4027 }, { "epoch": 1.3103448275862069, "grad_norm": 2.1087296426694375, "learning_rate": 6.92761608494881e-06, "loss": 0.4555, "step": 4028 }, { "epoch": 1.3106701366297984, "grad_norm": 2.142296764416225, "learning_rate": 6.92586950653168e-06, "loss": 0.3571, "step": 4029 }, { "epoch": 1.3109954456733897, "grad_norm": 1.9986204037265454, "learning_rate": 6.924122652136167e-06, "loss": 0.3995, "step": 4030 }, { "epoch": 1.3113207547169812, "grad_norm": 1.9109263154045604, "learning_rate": 6.9223755220125945e-06, "loss": 0.376, "step": 4031 }, { "epoch": 1.3116460637605725, "grad_norm": 1.9397229948342676, "learning_rate": 6.9206281164113296e-06, "loss": 0.4376, "step": 4032 }, { "epoch": 1.311971372804164, "grad_norm": 1.785875482108035, "learning_rate": 6.918880435582773e-06, "loss": 0.3752, "step": 4033 }, { "epoch": 1.3122966818477555, "grad_norm": 1.7112363980850094, "learning_rate": 6.917132479777372e-06, "loss": 0.2949, "step": 4034 }, { "epoch": 1.3126219908913468, "grad_norm": 1.7099297472399828, "learning_rate": 6.915384249245608e-06, "loss": 0.301, "step": 4035 }, { "epoch": 1.312947299934938, "grad_norm": 1.6453707393152186, "learning_rate": 6.913635744238004e-06, "loss": 0.356, "step": 4036 }, { "epoch": 1.3132726089785296, "grad_norm": 1.8948159848651893, "learning_rate": 6.911886965005121e-06, "loss": 0.4336, "step": 4037 }, { "epoch": 1.3135979180221211, "grad_norm": 1.7425589037780758, "learning_rate": 6.910137911797563e-06, "loss": 0.355, "step": 4038 }, { "epoch": 1.3139232270657124, "grad_norm": 1.9629826193521254, "learning_rate": 6.908388584865968e-06, "loss": 0.4218, "step": 4039 }, { "epoch": 1.3142485361093037, "grad_norm": 1.6881896005008408, "learning_rate": 6.906638984461016e-06, "loss": 0.3312, "step": 4040 }, { "epoch": 1.3145738451528952, "grad_norm": 1.8682548278605848, "learning_rate": 6.904889110833426e-06, "loss": 0.4148, "step": 4041 }, { "epoch": 1.3148991541964867, "grad_norm": 2.112842467922743, "learning_rate": 6.9031389642339585e-06, "loss": 0.4314, "step": 4042 }, { "epoch": 1.315224463240078, "grad_norm": 1.8628789228264142, "learning_rate": 6.901388544913405e-06, "loss": 0.3558, "step": 4043 }, { "epoch": 1.3155497722836695, "grad_norm": 1.9407187102335428, "learning_rate": 6.899637853122608e-06, "loss": 0.3516, "step": 4044 }, { "epoch": 1.3158750813272608, "grad_norm": 2.1148143147550496, "learning_rate": 6.897886889112439e-06, "loss": 0.5576, "step": 4045 }, { "epoch": 1.3162003903708523, "grad_norm": 1.8800821945108568, "learning_rate": 6.8961356531338155e-06, "loss": 0.3379, "step": 4046 }, { "epoch": 1.3165256994144436, "grad_norm": 1.629643860300768, "learning_rate": 6.894384145437686e-06, "loss": 0.288, "step": 4047 }, { "epoch": 1.3168510084580352, "grad_norm": 1.762122048486663, "learning_rate": 6.8926323662750474e-06, "loss": 0.3597, "step": 4048 }, { "epoch": 1.3171763175016267, "grad_norm": 2.340887376650927, "learning_rate": 6.890880315896928e-06, "loss": 0.4657, "step": 4049 }, { "epoch": 1.317501626545218, "grad_norm": 1.693615493698415, "learning_rate": 6.889127994554399e-06, "loss": 0.3828, "step": 4050 }, { "epoch": 1.3178269355888093, "grad_norm": 1.8778687960666032, "learning_rate": 6.887375402498568e-06, "loss": 0.4145, "step": 4051 }, { "epoch": 1.3181522446324008, "grad_norm": 2.2453351430670274, "learning_rate": 6.885622539980583e-06, "loss": 0.4809, "step": 4052 }, { "epoch": 1.3184775536759923, "grad_norm": 1.9178858983469464, "learning_rate": 6.883869407251631e-06, "loss": 0.3577, "step": 4053 }, { "epoch": 1.3188028627195836, "grad_norm": 2.480262866478394, "learning_rate": 6.882116004562938e-06, "loss": 0.4486, "step": 4054 }, { "epoch": 1.319128171763175, "grad_norm": 1.8780409351962073, "learning_rate": 6.8803623321657664e-06, "loss": 0.3766, "step": 4055 }, { "epoch": 1.3194534808067664, "grad_norm": 1.816498312935138, "learning_rate": 6.878608390311419e-06, "loss": 0.4439, "step": 4056 }, { "epoch": 1.319778789850358, "grad_norm": 1.912518565730827, "learning_rate": 6.876854179251235e-06, "loss": 0.4126, "step": 4057 }, { "epoch": 1.3201040988939492, "grad_norm": 1.7244514817750423, "learning_rate": 6.8750996992365995e-06, "loss": 0.3139, "step": 4058 }, { "epoch": 1.3204294079375407, "grad_norm": 1.7746724691352462, "learning_rate": 6.873344950518923e-06, "loss": 0.3455, "step": 4059 }, { "epoch": 1.320754716981132, "grad_norm": 2.0578757225236703, "learning_rate": 6.871589933349668e-06, "loss": 0.4104, "step": 4060 }, { "epoch": 1.3210800260247235, "grad_norm": 1.9195869135312804, "learning_rate": 6.869834647980328e-06, "loss": 0.4638, "step": 4061 }, { "epoch": 1.3214053350683148, "grad_norm": 1.7033948901874183, "learning_rate": 6.868079094662437e-06, "loss": 0.3751, "step": 4062 }, { "epoch": 1.3217306441119063, "grad_norm": 2.0462581055181692, "learning_rate": 6.866323273647564e-06, "loss": 0.3729, "step": 4063 }, { "epoch": 1.3220559531554978, "grad_norm": 2.1662308199983733, "learning_rate": 6.8645671851873215e-06, "loss": 0.4085, "step": 4064 }, { "epoch": 1.3223812621990891, "grad_norm": 1.9732242479968092, "learning_rate": 6.86281082953336e-06, "loss": 0.4066, "step": 4065 }, { "epoch": 1.3227065712426804, "grad_norm": 1.9106066857704374, "learning_rate": 6.861054206937364e-06, "loss": 0.3485, "step": 4066 }, { "epoch": 1.323031880286272, "grad_norm": 1.8664111277932622, "learning_rate": 6.859297317651058e-06, "loss": 0.3983, "step": 4067 }, { "epoch": 1.3233571893298635, "grad_norm": 1.626549465798688, "learning_rate": 6.85754016192621e-06, "loss": 0.3101, "step": 4068 }, { "epoch": 1.3236824983734548, "grad_norm": 1.836193066657755, "learning_rate": 6.855782740014615e-06, "loss": 0.3804, "step": 4069 }, { "epoch": 1.3240078074170463, "grad_norm": 1.8997424672094225, "learning_rate": 6.854025052168117e-06, "loss": 0.364, "step": 4070 }, { "epoch": 1.3243331164606376, "grad_norm": 1.735225609935469, "learning_rate": 6.852267098638593e-06, "loss": 0.3394, "step": 4071 }, { "epoch": 1.324658425504229, "grad_norm": 1.9336012549341535, "learning_rate": 6.85050887967796e-06, "loss": 0.4135, "step": 4072 }, { "epoch": 1.3249837345478204, "grad_norm": 2.0755829356692197, "learning_rate": 6.8487503955381684e-06, "loss": 0.3595, "step": 4073 }, { "epoch": 1.3253090435914119, "grad_norm": 1.8996201368602625, "learning_rate": 6.846991646471215e-06, "loss": 0.4468, "step": 4074 }, { "epoch": 1.3256343526350032, "grad_norm": 1.7304127781888121, "learning_rate": 6.845232632729125e-06, "loss": 0.3946, "step": 4075 }, { "epoch": 1.3259596616785947, "grad_norm": 1.9474625880509486, "learning_rate": 6.84347335456397e-06, "loss": 0.4015, "step": 4076 }, { "epoch": 1.326284970722186, "grad_norm": 1.9349048955717143, "learning_rate": 6.841713812227853e-06, "loss": 0.4776, "step": 4077 }, { "epoch": 1.3266102797657775, "grad_norm": 2.2035010112627176, "learning_rate": 6.839954005972921e-06, "loss": 0.5393, "step": 4078 }, { "epoch": 1.326935588809369, "grad_norm": 1.7689902063182958, "learning_rate": 6.8381939360513506e-06, "loss": 0.3241, "step": 4079 }, { "epoch": 1.3272608978529603, "grad_norm": 1.9273530647228292, "learning_rate": 6.836433602715364e-06, "loss": 0.3718, "step": 4080 }, { "epoch": 1.3275862068965516, "grad_norm": 1.7667256914985887, "learning_rate": 6.8346730062172196e-06, "loss": 0.2641, "step": 4081 }, { "epoch": 1.3279115159401431, "grad_norm": 1.9689033015509605, "learning_rate": 6.832912146809209e-06, "loss": 0.3859, "step": 4082 }, { "epoch": 1.3282368249837346, "grad_norm": 1.9161050288312327, "learning_rate": 6.831151024743665e-06, "loss": 0.4502, "step": 4083 }, { "epoch": 1.328562134027326, "grad_norm": 1.7375439499914287, "learning_rate": 6.82938964027296e-06, "loss": 0.3636, "step": 4084 }, { "epoch": 1.3288874430709174, "grad_norm": 1.888869072328084, "learning_rate": 6.827627993649497e-06, "loss": 0.3807, "step": 4085 }, { "epoch": 1.3292127521145087, "grad_norm": 1.7112853647647257, "learning_rate": 6.825866085125727e-06, "loss": 0.3931, "step": 4086 }, { "epoch": 1.3295380611581002, "grad_norm": 2.1714736565818584, "learning_rate": 6.824103914954128e-06, "loss": 0.35, "step": 4087 }, { "epoch": 1.3298633702016915, "grad_norm": 2.06123245836041, "learning_rate": 6.8223414833872225e-06, "loss": 0.4423, "step": 4088 }, { "epoch": 1.330188679245283, "grad_norm": 2.073322523916377, "learning_rate": 6.820578790677567e-06, "loss": 0.4633, "step": 4089 }, { "epoch": 1.3305139882888743, "grad_norm": 1.9574723391870326, "learning_rate": 6.818815837077757e-06, "loss": 0.3816, "step": 4090 }, { "epoch": 1.3308392973324659, "grad_norm": 1.8958106468336662, "learning_rate": 6.817052622840426e-06, "loss": 0.4055, "step": 4091 }, { "epoch": 1.3311646063760572, "grad_norm": 2.3095918895132703, "learning_rate": 6.81528914821824e-06, "loss": 0.4014, "step": 4092 }, { "epoch": 1.3314899154196487, "grad_norm": 2.2653521280195488, "learning_rate": 6.8135254134639105e-06, "loss": 0.4789, "step": 4093 }, { "epoch": 1.3318152244632402, "grad_norm": 1.7554839913140345, "learning_rate": 6.8117614188301806e-06, "loss": 0.3989, "step": 4094 }, { "epoch": 1.3321405335068315, "grad_norm": 2.039138912742803, "learning_rate": 6.809997164569831e-06, "loss": 0.4822, "step": 4095 }, { "epoch": 1.3324658425504228, "grad_norm": 1.7319100539914416, "learning_rate": 6.80823265093568e-06, "loss": 0.3868, "step": 4096 }, { "epoch": 1.3327911515940143, "grad_norm": 1.7768110918028786, "learning_rate": 6.806467878180586e-06, "loss": 0.4164, "step": 4097 }, { "epoch": 1.3331164606376058, "grad_norm": 1.4664368278174238, "learning_rate": 6.804702846557441e-06, "loss": 0.2808, "step": 4098 }, { "epoch": 1.333441769681197, "grad_norm": 2.1100547452041996, "learning_rate": 6.802937556319171e-06, "loss": 0.4692, "step": 4099 }, { "epoch": 1.3337670787247886, "grad_norm": 1.8120013304587042, "learning_rate": 6.8011720077187515e-06, "loss": 0.3832, "step": 4100 }, { "epoch": 1.33409238776838, "grad_norm": 1.8068090826667902, "learning_rate": 6.79940620100918e-06, "loss": 0.3536, "step": 4101 }, { "epoch": 1.3344176968119714, "grad_norm": 2.070642852408146, "learning_rate": 6.797640136443502e-06, "loss": 0.3703, "step": 4102 }, { "epoch": 1.3347430058555627, "grad_norm": 2.0476553575823404, "learning_rate": 6.795873814274792e-06, "loss": 0.3311, "step": 4103 }, { "epoch": 1.3350683148991542, "grad_norm": 1.9217626010587634, "learning_rate": 6.794107234756169e-06, "loss": 0.3782, "step": 4104 }, { "epoch": 1.3353936239427457, "grad_norm": 1.8170027081559033, "learning_rate": 6.792340398140781e-06, "loss": 0.3718, "step": 4105 }, { "epoch": 1.335718932986337, "grad_norm": 1.981576427243361, "learning_rate": 6.79057330468182e-06, "loss": 0.4163, "step": 4106 }, { "epoch": 1.3360442420299283, "grad_norm": 2.3020686849803362, "learning_rate": 6.78880595463251e-06, "loss": 0.4585, "step": 4107 }, { "epoch": 1.3363695510735198, "grad_norm": 1.898771145520356, "learning_rate": 6.787038348246116e-06, "loss": 0.4031, "step": 4108 }, { "epoch": 1.3366948601171114, "grad_norm": 1.9090141267156924, "learning_rate": 6.785270485775932e-06, "loss": 0.4117, "step": 4109 }, { "epoch": 1.3370201691607027, "grad_norm": 1.6899515669689855, "learning_rate": 6.783502367475301e-06, "loss": 0.3599, "step": 4110 }, { "epoch": 1.337345478204294, "grad_norm": 1.7428387021675773, "learning_rate": 6.781733993597588e-06, "loss": 0.361, "step": 4111 }, { "epoch": 1.3376707872478855, "grad_norm": 1.9231054758743824, "learning_rate": 6.779965364396209e-06, "loss": 0.3881, "step": 4112 }, { "epoch": 1.337996096291477, "grad_norm": 1.9138302107985175, "learning_rate": 6.778196480124603e-06, "loss": 0.3549, "step": 4113 }, { "epoch": 1.3383214053350683, "grad_norm": 1.951700456438994, "learning_rate": 6.776427341036258e-06, "loss": 0.3742, "step": 4114 }, { "epoch": 1.3386467143786598, "grad_norm": 1.7197138854583853, "learning_rate": 6.77465794738469e-06, "loss": 0.3552, "step": 4115 }, { "epoch": 1.338972023422251, "grad_norm": 1.8826863904491193, "learning_rate": 6.7728882994234555e-06, "loss": 0.3115, "step": 4116 }, { "epoch": 1.3392973324658426, "grad_norm": 1.8402703645777798, "learning_rate": 6.7711183974061455e-06, "loss": 0.3061, "step": 4117 }, { "epoch": 1.3396226415094339, "grad_norm": 2.070715249665671, "learning_rate": 6.769348241586388e-06, "loss": 0.4118, "step": 4118 }, { "epoch": 1.3399479505530254, "grad_norm": 2.3456433232201497, "learning_rate": 6.767577832217849e-06, "loss": 0.4649, "step": 4119 }, { "epoch": 1.340273259596617, "grad_norm": 2.0768057719685653, "learning_rate": 6.765807169554226e-06, "loss": 0.4403, "step": 4120 }, { "epoch": 1.3405985686402082, "grad_norm": 2.0395313398271444, "learning_rate": 6.76403625384926e-06, "loss": 0.3712, "step": 4121 }, { "epoch": 1.3409238776837995, "grad_norm": 2.180195535193108, "learning_rate": 6.762265085356722e-06, "loss": 0.3928, "step": 4122 }, { "epoch": 1.341249186727391, "grad_norm": 2.1217138368679227, "learning_rate": 6.760493664330422e-06, "loss": 0.4168, "step": 4123 }, { "epoch": 1.3415744957709825, "grad_norm": 2.018339973607485, "learning_rate": 6.7587219910242065e-06, "loss": 0.4395, "step": 4124 }, { "epoch": 1.3418998048145738, "grad_norm": 1.8980160727289574, "learning_rate": 6.756950065691956e-06, "loss": 0.3378, "step": 4125 }, { "epoch": 1.3422251138581651, "grad_norm": 1.9890092911371697, "learning_rate": 6.755177888587592e-06, "loss": 0.3224, "step": 4126 }, { "epoch": 1.3425504229017566, "grad_norm": 1.9326110298786479, "learning_rate": 6.753405459965065e-06, "loss": 0.3499, "step": 4127 }, { "epoch": 1.3428757319453482, "grad_norm": 2.454696290040916, "learning_rate": 6.7516327800783656e-06, "loss": 0.5192, "step": 4128 }, { "epoch": 1.3432010409889394, "grad_norm": 1.8550182575933318, "learning_rate": 6.749859849181523e-06, "loss": 0.4105, "step": 4129 }, { "epoch": 1.343526350032531, "grad_norm": 2.1322295398617936, "learning_rate": 6.748086667528596e-06, "loss": 0.5055, "step": 4130 }, { "epoch": 1.3438516590761223, "grad_norm": 1.908126078498255, "learning_rate": 6.746313235373685e-06, "loss": 0.3819, "step": 4131 }, { "epoch": 1.3441769681197138, "grad_norm": 1.8610942770199568, "learning_rate": 6.744539552970922e-06, "loss": 0.3539, "step": 4132 }, { "epoch": 1.344502277163305, "grad_norm": 2.570713423151804, "learning_rate": 6.742765620574478e-06, "loss": 0.4854, "step": 4133 }, { "epoch": 1.3448275862068966, "grad_norm": 1.8625155185890772, "learning_rate": 6.740991438438559e-06, "loss": 0.3566, "step": 4134 }, { "epoch": 1.345152895250488, "grad_norm": 2.1525699407124703, "learning_rate": 6.739217006817408e-06, "loss": 0.3886, "step": 4135 }, { "epoch": 1.3454782042940794, "grad_norm": 2.288521787793921, "learning_rate": 6.7374423259652976e-06, "loss": 0.4227, "step": 4136 }, { "epoch": 1.3458035133376707, "grad_norm": 1.9596305153397868, "learning_rate": 6.735667396136544e-06, "loss": 0.4556, "step": 4137 }, { "epoch": 1.3461288223812622, "grad_norm": 1.868223503553155, "learning_rate": 6.733892217585496e-06, "loss": 0.3197, "step": 4138 }, { "epoch": 1.3464541314248537, "grad_norm": 2.047863564303728, "learning_rate": 6.732116790566537e-06, "loss": 0.4093, "step": 4139 }, { "epoch": 1.346779440468445, "grad_norm": 1.7490090338169018, "learning_rate": 6.730341115334086e-06, "loss": 0.3925, "step": 4140 }, { "epoch": 1.3471047495120365, "grad_norm": 1.7749691773278604, "learning_rate": 6.7285651921426e-06, "loss": 0.3489, "step": 4141 }, { "epoch": 1.3474300585556278, "grad_norm": 1.968473257027544, "learning_rate": 6.72678902124657e-06, "loss": 0.4116, "step": 4142 }, { "epoch": 1.3477553675992193, "grad_norm": 1.7984787888161076, "learning_rate": 6.725012602900523e-06, "loss": 0.359, "step": 4143 }, { "epoch": 1.3480806766428106, "grad_norm": 1.8515911237001261, "learning_rate": 6.723235937359017e-06, "loss": 0.3405, "step": 4144 }, { "epoch": 1.3484059856864021, "grad_norm": 2.325539238502515, "learning_rate": 6.721459024876656e-06, "loss": 0.4556, "step": 4145 }, { "epoch": 1.3487312947299934, "grad_norm": 1.9199884299051464, "learning_rate": 6.719681865708065e-06, "loss": 0.3828, "step": 4146 }, { "epoch": 1.349056603773585, "grad_norm": 2.0169477479610682, "learning_rate": 6.717904460107919e-06, "loss": 0.4433, "step": 4147 }, { "epoch": 1.3493819128171762, "grad_norm": 1.831164164563647, "learning_rate": 6.716126808330917e-06, "loss": 0.3177, "step": 4148 }, { "epoch": 1.3497072218607677, "grad_norm": 1.721003045306402, "learning_rate": 6.7143489106318e-06, "loss": 0.3579, "step": 4149 }, { "epoch": 1.3500325309043593, "grad_norm": 1.8758834692775583, "learning_rate": 6.712570767265341e-06, "loss": 0.387, "step": 4150 }, { "epoch": 1.3503578399479506, "grad_norm": 1.7461870136370095, "learning_rate": 6.710792378486348e-06, "loss": 0.3585, "step": 4151 }, { "epoch": 1.3506831489915418, "grad_norm": 2.2138173532594903, "learning_rate": 6.709013744549668e-06, "loss": 0.3969, "step": 4152 }, { "epoch": 1.3510084580351334, "grad_norm": 2.016297546525397, "learning_rate": 6.707234865710179e-06, "loss": 0.4958, "step": 4153 }, { "epoch": 1.3513337670787249, "grad_norm": 1.8112722477315832, "learning_rate": 6.7054557422227936e-06, "loss": 0.4045, "step": 4154 }, { "epoch": 1.3516590761223162, "grad_norm": 1.7783068820158154, "learning_rate": 6.703676374342467e-06, "loss": 0.3663, "step": 4155 }, { "epoch": 1.3519843851659077, "grad_norm": 1.540275764272273, "learning_rate": 6.701896762324176e-06, "loss": 0.2806, "step": 4156 }, { "epoch": 1.352309694209499, "grad_norm": 1.8332980286689375, "learning_rate": 6.700116906422946e-06, "loss": 0.4195, "step": 4157 }, { "epoch": 1.3526350032530905, "grad_norm": 1.6899251311187924, "learning_rate": 6.6983368068938305e-06, "loss": 0.3074, "step": 4158 }, { "epoch": 1.3529603122966818, "grad_norm": 1.9181028910489153, "learning_rate": 6.696556463991918e-06, "loss": 0.4394, "step": 4159 }, { "epoch": 1.3532856213402733, "grad_norm": 2.1211402401673176, "learning_rate": 6.6947758779723315e-06, "loss": 0.4025, "step": 4160 }, { "epoch": 1.3536109303838646, "grad_norm": 1.8454007767584626, "learning_rate": 6.692995049090234e-06, "loss": 0.3676, "step": 4161 }, { "epoch": 1.3539362394274561, "grad_norm": 1.5249549383290533, "learning_rate": 6.691213977600815e-06, "loss": 0.3114, "step": 4162 }, { "epoch": 1.3542615484710474, "grad_norm": 1.8532222791435091, "learning_rate": 6.689432663759307e-06, "loss": 0.4122, "step": 4163 }, { "epoch": 1.354586857514639, "grad_norm": 1.6844869685315333, "learning_rate": 6.6876511078209716e-06, "loss": 0.3314, "step": 4164 }, { "epoch": 1.3549121665582304, "grad_norm": 1.7271153828847212, "learning_rate": 6.685869310041106e-06, "loss": 0.362, "step": 4165 }, { "epoch": 1.3552374756018217, "grad_norm": 1.8827281882590856, "learning_rate": 6.6840872706750445e-06, "loss": 0.4579, "step": 4166 }, { "epoch": 1.355562784645413, "grad_norm": 2.278363552574502, "learning_rate": 6.682304989978156e-06, "loss": 0.4499, "step": 4167 }, { "epoch": 1.3558880936890045, "grad_norm": 1.9105338839223165, "learning_rate": 6.680522468205839e-06, "loss": 0.3917, "step": 4168 }, { "epoch": 1.356213402732596, "grad_norm": 1.824295151049543, "learning_rate": 6.678739705613533e-06, "loss": 0.344, "step": 4169 }, { "epoch": 1.3565387117761873, "grad_norm": 1.7853529283593734, "learning_rate": 6.676956702456707e-06, "loss": 0.3626, "step": 4170 }, { "epoch": 1.3568640208197789, "grad_norm": 2.082816160554715, "learning_rate": 6.675173458990869e-06, "loss": 0.3682, "step": 4171 }, { "epoch": 1.3571893298633702, "grad_norm": 2.1673124936300825, "learning_rate": 6.673389975471556e-06, "loss": 0.4405, "step": 4172 }, { "epoch": 1.3575146389069617, "grad_norm": 1.669675577822694, "learning_rate": 6.671606252154345e-06, "loss": 0.333, "step": 4173 }, { "epoch": 1.357839947950553, "grad_norm": 1.8104436612585906, "learning_rate": 6.669822289294843e-06, "loss": 0.3188, "step": 4174 }, { "epoch": 1.3581652569941445, "grad_norm": 2.014228079507381, "learning_rate": 6.6680380871486925e-06, "loss": 0.4126, "step": 4175 }, { "epoch": 1.3584905660377358, "grad_norm": 2.173721530545266, "learning_rate": 6.666253645971572e-06, "loss": 0.4539, "step": 4176 }, { "epoch": 1.3588158750813273, "grad_norm": 1.5934734431001325, "learning_rate": 6.664468966019195e-06, "loss": 0.3306, "step": 4177 }, { "epoch": 1.3591411841249186, "grad_norm": 1.855258852796417, "learning_rate": 6.662684047547304e-06, "loss": 0.3296, "step": 4178 }, { "epoch": 1.35946649316851, "grad_norm": 1.7078239131217663, "learning_rate": 6.660898890811681e-06, "loss": 0.3278, "step": 4179 }, { "epoch": 1.3597918022121016, "grad_norm": 1.8352921411085816, "learning_rate": 6.6591134960681406e-06, "loss": 0.3837, "step": 4180 }, { "epoch": 1.360117111255693, "grad_norm": 1.670292083476253, "learning_rate": 6.657327863572528e-06, "loss": 0.3217, "step": 4181 }, { "epoch": 1.3604424202992842, "grad_norm": 2.124176667978397, "learning_rate": 6.655541993580728e-06, "loss": 0.4377, "step": 4182 }, { "epoch": 1.3607677293428757, "grad_norm": 2.191202605174572, "learning_rate": 6.653755886348657e-06, "loss": 0.3231, "step": 4183 }, { "epoch": 1.3610930383864672, "grad_norm": 1.9795712516785489, "learning_rate": 6.651969542132265e-06, "loss": 0.3921, "step": 4184 }, { "epoch": 1.3614183474300585, "grad_norm": 1.8928944973586679, "learning_rate": 6.650182961187537e-06, "loss": 0.4354, "step": 4185 }, { "epoch": 1.36174365647365, "grad_norm": 2.1716345674124162, "learning_rate": 6.648396143770488e-06, "loss": 0.4634, "step": 4186 }, { "epoch": 1.3620689655172413, "grad_norm": 2.12099578651264, "learning_rate": 6.646609090137175e-06, "loss": 0.3769, "step": 4187 }, { "epoch": 1.3623942745608328, "grad_norm": 2.1083937461476374, "learning_rate": 6.64482180054368e-06, "loss": 0.4846, "step": 4188 }, { "epoch": 1.3627195836044241, "grad_norm": 1.9999307850682946, "learning_rate": 6.6430342752461254e-06, "loss": 0.43, "step": 4189 }, { "epoch": 1.3630448926480156, "grad_norm": 1.790707410054789, "learning_rate": 6.641246514500663e-06, "loss": 0.346, "step": 4190 }, { "epoch": 1.3633702016916072, "grad_norm": 2.117707833060098, "learning_rate": 6.63945851856348e-06, "loss": 0.4513, "step": 4191 }, { "epoch": 1.3636955107351985, "grad_norm": 1.976492068691864, "learning_rate": 6.6376702876908e-06, "loss": 0.4314, "step": 4192 }, { "epoch": 1.3640208197787898, "grad_norm": 1.5814343375270756, "learning_rate": 6.635881822138874e-06, "loss": 0.3355, "step": 4193 }, { "epoch": 1.3643461288223813, "grad_norm": 1.8572474661669656, "learning_rate": 6.634093122163994e-06, "loss": 0.4465, "step": 4194 }, { "epoch": 1.3646714378659728, "grad_norm": 1.747480250504679, "learning_rate": 6.632304188022481e-06, "loss": 0.3597, "step": 4195 }, { "epoch": 1.364996746909564, "grad_norm": 2.0975554311572884, "learning_rate": 6.630515019970686e-06, "loss": 0.4237, "step": 4196 }, { "epoch": 1.3653220559531554, "grad_norm": 1.953362674191788, "learning_rate": 6.628725618265004e-06, "loss": 0.3828, "step": 4197 }, { "epoch": 1.3656473649967469, "grad_norm": 2.2509936797961143, "learning_rate": 6.626935983161854e-06, "loss": 0.4374, "step": 4198 }, { "epoch": 1.3659726740403384, "grad_norm": 1.638075573076931, "learning_rate": 6.625146114917693e-06, "loss": 0.3183, "step": 4199 }, { "epoch": 1.3662979830839297, "grad_norm": 1.8478135006821055, "learning_rate": 6.623356013789012e-06, "loss": 0.3348, "step": 4200 }, { "epoch": 1.3666232921275212, "grad_norm": 1.9608877390886947, "learning_rate": 6.62156568003233e-06, "loss": 0.4496, "step": 4201 }, { "epoch": 1.3669486011711125, "grad_norm": 2.0898703966672203, "learning_rate": 6.619775113904205e-06, "loss": 0.4007, "step": 4202 }, { "epoch": 1.367273910214704, "grad_norm": 1.7823701194363475, "learning_rate": 6.617984315661228e-06, "loss": 0.407, "step": 4203 }, { "epoch": 1.3675992192582953, "grad_norm": 1.8521104057274018, "learning_rate": 6.616193285560018e-06, "loss": 0.3717, "step": 4204 }, { "epoch": 1.3679245283018868, "grad_norm": 1.805836948962434, "learning_rate": 6.614402023857231e-06, "loss": 0.3956, "step": 4205 }, { "epoch": 1.3682498373454783, "grad_norm": 1.8012448037860038, "learning_rate": 6.612610530809561e-06, "loss": 0.3531, "step": 4206 }, { "epoch": 1.3685751463890696, "grad_norm": 1.8916441143756473, "learning_rate": 6.610818806673726e-06, "loss": 0.397, "step": 4207 }, { "epoch": 1.368900455432661, "grad_norm": 2.170174749500781, "learning_rate": 6.6090268517064795e-06, "loss": 0.3998, "step": 4208 }, { "epoch": 1.3692257644762524, "grad_norm": 1.9986434927350292, "learning_rate": 6.607234666164615e-06, "loss": 0.3541, "step": 4209 }, { "epoch": 1.369551073519844, "grad_norm": 1.7580369384242591, "learning_rate": 6.60544225030495e-06, "loss": 0.3819, "step": 4210 }, { "epoch": 1.3698763825634352, "grad_norm": 2.1460232879137604, "learning_rate": 6.603649604384341e-06, "loss": 0.3821, "step": 4211 }, { "epoch": 1.3702016916070265, "grad_norm": 1.9014227991858317, "learning_rate": 6.6018567286596735e-06, "loss": 0.385, "step": 4212 }, { "epoch": 1.370527000650618, "grad_norm": 2.1652288188034605, "learning_rate": 6.600063623387871e-06, "loss": 0.4976, "step": 4213 }, { "epoch": 1.3708523096942096, "grad_norm": 1.6956972674674753, "learning_rate": 6.598270288825881e-06, "loss": 0.3029, "step": 4214 }, { "epoch": 1.3711776187378009, "grad_norm": 1.77868774387542, "learning_rate": 6.596476725230696e-06, "loss": 0.3978, "step": 4215 }, { "epoch": 1.3715029277813924, "grad_norm": 2.1567324408795927, "learning_rate": 6.594682932859331e-06, "loss": 0.3852, "step": 4216 }, { "epoch": 1.3718282368249837, "grad_norm": 1.9747498639229562, "learning_rate": 6.592888911968838e-06, "loss": 0.3569, "step": 4217 }, { "epoch": 1.3721535458685752, "grad_norm": 1.8773626252349975, "learning_rate": 6.591094662816301e-06, "loss": 0.3627, "step": 4218 }, { "epoch": 1.3724788549121665, "grad_norm": 1.8508533912123006, "learning_rate": 6.589300185658839e-06, "loss": 0.3753, "step": 4219 }, { "epoch": 1.372804163955758, "grad_norm": 1.7254991502640704, "learning_rate": 6.587505480753602e-06, "loss": 0.3619, "step": 4220 }, { "epoch": 1.3731294729993495, "grad_norm": 1.6769767844806824, "learning_rate": 6.58571054835777e-06, "loss": 0.3644, "step": 4221 }, { "epoch": 1.3734547820429408, "grad_norm": 1.8457966561501904, "learning_rate": 6.5839153887285586e-06, "loss": 0.3522, "step": 4222 }, { "epoch": 1.373780091086532, "grad_norm": 1.9332001173299311, "learning_rate": 6.582120002123217e-06, "loss": 0.3675, "step": 4223 }, { "epoch": 1.3741054001301236, "grad_norm": 2.0336907147822214, "learning_rate": 6.580324388799023e-06, "loss": 0.3812, "step": 4224 }, { "epoch": 1.3744307091737151, "grad_norm": 2.1259496178515467, "learning_rate": 6.578528549013292e-06, "loss": 0.4133, "step": 4225 }, { "epoch": 1.3747560182173064, "grad_norm": 1.774011020793395, "learning_rate": 6.576732483023367e-06, "loss": 0.3915, "step": 4226 }, { "epoch": 1.375081327260898, "grad_norm": 1.6801088401354507, "learning_rate": 6.574936191086625e-06, "loss": 0.3356, "step": 4227 }, { "epoch": 1.3754066363044892, "grad_norm": 2.0213277095971645, "learning_rate": 6.573139673460477e-06, "loss": 0.3901, "step": 4228 }, { "epoch": 1.3757319453480807, "grad_norm": 2.082099008716667, "learning_rate": 6.571342930402367e-06, "loss": 0.3881, "step": 4229 }, { "epoch": 1.376057254391672, "grad_norm": 2.027101109949548, "learning_rate": 6.569545962169764e-06, "loss": 0.4842, "step": 4230 }, { "epoch": 1.3763825634352636, "grad_norm": 1.9037284172381006, "learning_rate": 6.5677487690201806e-06, "loss": 0.4226, "step": 4231 }, { "epoch": 1.3767078724788548, "grad_norm": 1.7981131999878823, "learning_rate": 6.565951351211153e-06, "loss": 0.3971, "step": 4232 }, { "epoch": 1.3770331815224464, "grad_norm": 1.8324373289397682, "learning_rate": 6.564153709000252e-06, "loss": 0.3692, "step": 4233 }, { "epoch": 1.3773584905660377, "grad_norm": 2.00234244852875, "learning_rate": 6.5623558426450806e-06, "loss": 0.4102, "step": 4234 }, { "epoch": 1.3776837996096292, "grad_norm": 1.8817145852834987, "learning_rate": 6.560557752403277e-06, "loss": 0.3323, "step": 4235 }, { "epoch": 1.3780091086532207, "grad_norm": 1.9111546119015914, "learning_rate": 6.558759438532507e-06, "loss": 0.3623, "step": 4236 }, { "epoch": 1.378334417696812, "grad_norm": 2.1755279801041025, "learning_rate": 6.55696090129047e-06, "loss": 0.4442, "step": 4237 }, { "epoch": 1.3786597267404033, "grad_norm": 1.8833055351195191, "learning_rate": 6.555162140934896e-06, "loss": 0.3213, "step": 4238 }, { "epoch": 1.3789850357839948, "grad_norm": 1.9615008938992715, "learning_rate": 6.553363157723555e-06, "loss": 0.4858, "step": 4239 }, { "epoch": 1.3793103448275863, "grad_norm": 1.8741329028554221, "learning_rate": 6.551563951914233e-06, "loss": 0.3598, "step": 4240 }, { "epoch": 1.3796356538711776, "grad_norm": 1.9318609545232004, "learning_rate": 6.5497645237647646e-06, "loss": 0.3678, "step": 4241 }, { "epoch": 1.379960962914769, "grad_norm": 1.90580907405733, "learning_rate": 6.547964873533006e-06, "loss": 0.4077, "step": 4242 }, { "epoch": 1.3802862719583604, "grad_norm": 2.18243792022587, "learning_rate": 6.5461650014768496e-06, "loss": 0.3849, "step": 4243 }, { "epoch": 1.380611581001952, "grad_norm": 1.9163710405361207, "learning_rate": 6.544364907854216e-06, "loss": 0.4013, "step": 4244 }, { "epoch": 1.3809368900455432, "grad_norm": 1.929528667980327, "learning_rate": 6.542564592923062e-06, "loss": 0.407, "step": 4245 }, { "epoch": 1.3812621990891347, "grad_norm": 1.969268406429242, "learning_rate": 6.540764056941375e-06, "loss": 0.4161, "step": 4246 }, { "epoch": 1.381587508132726, "grad_norm": 1.606116308254548, "learning_rate": 6.538963300167169e-06, "loss": 0.3444, "step": 4247 }, { "epoch": 1.3819128171763175, "grad_norm": 2.229958038782214, "learning_rate": 6.537162322858498e-06, "loss": 0.4526, "step": 4248 }, { "epoch": 1.3822381262199088, "grad_norm": 1.7392045200264878, "learning_rate": 6.5353611252734405e-06, "loss": 0.3396, "step": 4249 }, { "epoch": 1.3825634352635003, "grad_norm": 2.0427245956760385, "learning_rate": 6.53355970767011e-06, "loss": 0.4463, "step": 4250 }, { "epoch": 1.3828887443070919, "grad_norm": 1.9724415680730258, "learning_rate": 6.531758070306651e-06, "loss": 0.4083, "step": 4251 }, { "epoch": 1.3832140533506831, "grad_norm": 1.7292821963759935, "learning_rate": 6.529956213441239e-06, "loss": 0.3036, "step": 4252 }, { "epoch": 1.3835393623942744, "grad_norm": 1.6711252905702512, "learning_rate": 6.528154137332081e-06, "loss": 0.3491, "step": 4253 }, { "epoch": 1.383864671437866, "grad_norm": 1.6033849142560614, "learning_rate": 6.526351842237417e-06, "loss": 0.3652, "step": 4254 }, { "epoch": 1.3841899804814575, "grad_norm": 1.744013276508531, "learning_rate": 6.524549328415517e-06, "loss": 0.296, "step": 4255 }, { "epoch": 1.3845152895250488, "grad_norm": 1.711736712051342, "learning_rate": 6.52274659612468e-06, "loss": 0.3189, "step": 4256 }, { "epoch": 1.3848405985686403, "grad_norm": 1.8304863102577698, "learning_rate": 6.520943645623242e-06, "loss": 0.3998, "step": 4257 }, { "epoch": 1.3851659076122316, "grad_norm": 1.6794570943224658, "learning_rate": 6.519140477169567e-06, "loss": 0.3742, "step": 4258 }, { "epoch": 1.385491216655823, "grad_norm": 1.9269636249185553, "learning_rate": 6.517337091022046e-06, "loss": 0.3043, "step": 4259 }, { "epoch": 1.3858165256994144, "grad_norm": 1.6976784475833462, "learning_rate": 6.515533487439108e-06, "loss": 0.284, "step": 4260 }, { "epoch": 1.386141834743006, "grad_norm": 2.1560273334644453, "learning_rate": 6.513729666679213e-06, "loss": 0.4043, "step": 4261 }, { "epoch": 1.3864671437865972, "grad_norm": 1.6294623006315756, "learning_rate": 6.511925629000847e-06, "loss": 0.3794, "step": 4262 }, { "epoch": 1.3867924528301887, "grad_norm": 1.767675437353628, "learning_rate": 6.510121374662528e-06, "loss": 0.2939, "step": 4263 }, { "epoch": 1.38711776187378, "grad_norm": 1.9705421228565467, "learning_rate": 6.508316903922814e-06, "loss": 0.4021, "step": 4264 }, { "epoch": 1.3874430709173715, "grad_norm": 1.9684918466059484, "learning_rate": 6.506512217040277e-06, "loss": 0.3765, "step": 4265 }, { "epoch": 1.387768379960963, "grad_norm": 1.8881211337710229, "learning_rate": 6.504707314273537e-06, "loss": 0.3731, "step": 4266 }, { "epoch": 1.3880936890045543, "grad_norm": 1.8963006296170157, "learning_rate": 6.502902195881236e-06, "loss": 0.4135, "step": 4267 }, { "epoch": 1.3884189980481456, "grad_norm": 2.1494035520387924, "learning_rate": 6.5010968621220474e-06, "loss": 0.5302, "step": 4268 }, { "epoch": 1.3887443070917371, "grad_norm": 2.6766284315294726, "learning_rate": 6.499291313254677e-06, "loss": 0.4272, "step": 4269 }, { "epoch": 1.3890696161353286, "grad_norm": 1.8205700196119776, "learning_rate": 6.497485549537863e-06, "loss": 0.3778, "step": 4270 }, { "epoch": 1.38939492517892, "grad_norm": 2.462669223121987, "learning_rate": 6.495679571230371e-06, "loss": 0.501, "step": 4271 }, { "epoch": 1.3897202342225115, "grad_norm": 1.766490263698123, "learning_rate": 6.493873378591e-06, "loss": 0.3502, "step": 4272 }, { "epoch": 1.3900455432661027, "grad_norm": 1.8644168278497615, "learning_rate": 6.492066971878577e-06, "loss": 0.3449, "step": 4273 }, { "epoch": 1.3903708523096943, "grad_norm": 2.3014304338082625, "learning_rate": 6.4902603513519636e-06, "loss": 0.4625, "step": 4274 }, { "epoch": 1.3906961613532856, "grad_norm": 1.9499329095437798, "learning_rate": 6.488453517270046e-06, "loss": 0.423, "step": 4275 }, { "epoch": 1.391021470396877, "grad_norm": 1.836650024895301, "learning_rate": 6.486646469891748e-06, "loss": 0.4588, "step": 4276 }, { "epoch": 1.3913467794404684, "grad_norm": 1.9495568047796774, "learning_rate": 6.484839209476019e-06, "loss": 0.4263, "step": 4277 }, { "epoch": 1.3916720884840599, "grad_norm": 1.4838512482644237, "learning_rate": 6.483031736281843e-06, "loss": 0.2896, "step": 4278 }, { "epoch": 1.3919973975276512, "grad_norm": 1.8827884172030716, "learning_rate": 6.481224050568229e-06, "loss": 0.3594, "step": 4279 }, { "epoch": 1.3923227065712427, "grad_norm": 1.9656318126201329, "learning_rate": 6.47941615259422e-06, "loss": 0.4179, "step": 4280 }, { "epoch": 1.3926480156148342, "grad_norm": 2.0451472487812827, "learning_rate": 6.477608042618893e-06, "loss": 0.4146, "step": 4281 }, { "epoch": 1.3929733246584255, "grad_norm": 2.063250552661618, "learning_rate": 6.475799720901345e-06, "loss": 0.4138, "step": 4282 }, { "epoch": 1.3932986337020168, "grad_norm": 1.4347148401739933, "learning_rate": 6.4739911877007144e-06, "loss": 0.2737, "step": 4283 }, { "epoch": 1.3936239427456083, "grad_norm": 2.04764150451574, "learning_rate": 6.472182443276166e-06, "loss": 0.3431, "step": 4284 }, { "epoch": 1.3939492517891998, "grad_norm": 2.034791413559339, "learning_rate": 6.470373487886888e-06, "loss": 0.4677, "step": 4285 }, { "epoch": 1.394274560832791, "grad_norm": 1.8510438265354099, "learning_rate": 6.468564321792112e-06, "loss": 0.3844, "step": 4286 }, { "epoch": 1.3945998698763826, "grad_norm": 2.0179091019612057, "learning_rate": 6.466754945251088e-06, "loss": 0.4485, "step": 4287 }, { "epoch": 1.394925178919974, "grad_norm": 2.0240224209638895, "learning_rate": 6.464945358523104e-06, "loss": 0.4001, "step": 4288 }, { "epoch": 1.3952504879635654, "grad_norm": 2.1391261889006237, "learning_rate": 6.463135561867471e-06, "loss": 0.4307, "step": 4289 }, { "epoch": 1.3955757970071567, "grad_norm": 1.8192655455958697, "learning_rate": 6.4613255555435384e-06, "loss": 0.3827, "step": 4290 }, { "epoch": 1.3959011060507482, "grad_norm": 2.1332984129119756, "learning_rate": 6.459515339810678e-06, "loss": 0.3982, "step": 4291 }, { "epoch": 1.3962264150943398, "grad_norm": 1.787706067241576, "learning_rate": 6.457704914928298e-06, "loss": 0.3577, "step": 4292 }, { "epoch": 1.396551724137931, "grad_norm": 2.0755525589909603, "learning_rate": 6.45589428115583e-06, "loss": 0.4283, "step": 4293 }, { "epoch": 1.3968770331815223, "grad_norm": 2.040813471638451, "learning_rate": 6.454083438752743e-06, "loss": 0.4625, "step": 4294 }, { "epoch": 1.3972023422251139, "grad_norm": 1.9492391457921683, "learning_rate": 6.452272387978526e-06, "loss": 0.3751, "step": 4295 }, { "epoch": 1.3975276512687054, "grad_norm": 2.042674067654646, "learning_rate": 6.450461129092711e-06, "loss": 0.4248, "step": 4296 }, { "epoch": 1.3978529603122967, "grad_norm": 2.1149046316148583, "learning_rate": 6.448649662354848e-06, "loss": 0.4736, "step": 4297 }, { "epoch": 1.398178269355888, "grad_norm": 1.9764838218527494, "learning_rate": 6.4468379880245224e-06, "loss": 0.358, "step": 4298 }, { "epoch": 1.3985035783994795, "grad_norm": 1.595712896134815, "learning_rate": 6.445026106361348e-06, "loss": 0.3592, "step": 4299 }, { "epoch": 1.398828887443071, "grad_norm": 1.7854956007742895, "learning_rate": 6.443214017624971e-06, "loss": 0.3406, "step": 4300 }, { "epoch": 1.3991541964866623, "grad_norm": 1.6848154141231377, "learning_rate": 6.4414017220750615e-06, "loss": 0.333, "step": 4301 }, { "epoch": 1.3994795055302538, "grad_norm": 1.8545013361952702, "learning_rate": 6.439589219971324e-06, "loss": 0.3246, "step": 4302 }, { "epoch": 1.399804814573845, "grad_norm": 2.064427843204056, "learning_rate": 6.437776511573493e-06, "loss": 0.419, "step": 4303 }, { "epoch": 1.4001301236174366, "grad_norm": 1.9623723226130354, "learning_rate": 6.435963597141328e-06, "loss": 0.3048, "step": 4304 }, { "epoch": 1.400455432661028, "grad_norm": 1.7993571141408602, "learning_rate": 6.434150476934621e-06, "loss": 0.3272, "step": 4305 }, { "epoch": 1.4007807417046194, "grad_norm": 1.922441982415984, "learning_rate": 6.432337151213196e-06, "loss": 0.3873, "step": 4306 }, { "epoch": 1.401106050748211, "grad_norm": 1.592199932949857, "learning_rate": 6.430523620236903e-06, "loss": 0.3374, "step": 4307 }, { "epoch": 1.4014313597918022, "grad_norm": 1.932438427472514, "learning_rate": 6.42870988426562e-06, "loss": 0.3908, "step": 4308 }, { "epoch": 1.4017566688353935, "grad_norm": 2.015897280921482, "learning_rate": 6.426895943559259e-06, "loss": 0.3932, "step": 4309 }, { "epoch": 1.402081977878985, "grad_norm": 1.8625614301342668, "learning_rate": 6.425081798377757e-06, "loss": 0.3716, "step": 4310 }, { "epoch": 1.4024072869225765, "grad_norm": 1.8687429135515705, "learning_rate": 6.423267448981083e-06, "loss": 0.4239, "step": 4311 }, { "epoch": 1.4027325959661678, "grad_norm": 1.8785349838287393, "learning_rate": 6.4214528956292355e-06, "loss": 0.4029, "step": 4312 }, { "epoch": 1.4030579050097594, "grad_norm": 1.858021151945395, "learning_rate": 6.419638138582241e-06, "loss": 0.3647, "step": 4313 }, { "epoch": 1.4033832140533506, "grad_norm": 2.0401258342550355, "learning_rate": 6.417823178100153e-06, "loss": 0.4022, "step": 4314 }, { "epoch": 1.4037085230969422, "grad_norm": 2.290323131432851, "learning_rate": 6.416008014443058e-06, "loss": 0.4257, "step": 4315 }, { "epoch": 1.4040338321405335, "grad_norm": 1.85407771454322, "learning_rate": 6.414192647871074e-06, "loss": 0.3678, "step": 4316 }, { "epoch": 1.404359141184125, "grad_norm": 1.959040062608495, "learning_rate": 6.412377078644336e-06, "loss": 0.4478, "step": 4317 }, { "epoch": 1.4046844502277163, "grad_norm": 1.8506173213781985, "learning_rate": 6.4105613070230245e-06, "loss": 0.412, "step": 4318 }, { "epoch": 1.4050097592713078, "grad_norm": 2.290049180671799, "learning_rate": 6.408745333267336e-06, "loss": 0.4572, "step": 4319 }, { "epoch": 1.405335068314899, "grad_norm": 1.989049633632767, "learning_rate": 6.406929157637502e-06, "loss": 0.3561, "step": 4320 }, { "epoch": 1.4056603773584906, "grad_norm": 1.8462178883702265, "learning_rate": 6.405112780393781e-06, "loss": 0.3603, "step": 4321 }, { "epoch": 1.405985686402082, "grad_norm": 1.6243550311614172, "learning_rate": 6.403296201796462e-06, "loss": 0.3679, "step": 4322 }, { "epoch": 1.4063109954456734, "grad_norm": 1.853201959224126, "learning_rate": 6.401479422105862e-06, "loss": 0.3742, "step": 4323 }, { "epoch": 1.4066363044892647, "grad_norm": 2.285416963684348, "learning_rate": 6.399662441582326e-06, "loss": 0.5174, "step": 4324 }, { "epoch": 1.4069616135328562, "grad_norm": 2.025552497730295, "learning_rate": 6.3978452604862285e-06, "loss": 0.4181, "step": 4325 }, { "epoch": 1.4072869225764477, "grad_norm": 2.3408695615587716, "learning_rate": 6.396027879077975e-06, "loss": 0.4445, "step": 4326 }, { "epoch": 1.407612231620039, "grad_norm": 2.0245120308732814, "learning_rate": 6.3942102976179945e-06, "loss": 0.4186, "step": 4327 }, { "epoch": 1.4079375406636305, "grad_norm": 1.9175780903693822, "learning_rate": 6.392392516366749e-06, "loss": 0.4501, "step": 4328 }, { "epoch": 1.4082628497072218, "grad_norm": 1.8658522919019684, "learning_rate": 6.390574535584728e-06, "loss": 0.3561, "step": 4329 }, { "epoch": 1.4085881587508133, "grad_norm": 2.0114172051932497, "learning_rate": 6.388756355532448e-06, "loss": 0.4213, "step": 4330 }, { "epoch": 1.4089134677944046, "grad_norm": 1.7582285672841647, "learning_rate": 6.386937976470457e-06, "loss": 0.3665, "step": 4331 }, { "epoch": 1.4092387768379961, "grad_norm": 2.090563877216578, "learning_rate": 6.38511939865933e-06, "loss": 0.4231, "step": 4332 }, { "epoch": 1.4095640858815874, "grad_norm": 2.0853831997344985, "learning_rate": 6.383300622359667e-06, "loss": 0.407, "step": 4333 }, { "epoch": 1.409889394925179, "grad_norm": 2.279490088763339, "learning_rate": 6.381481647832104e-06, "loss": 0.5173, "step": 4334 }, { "epoch": 1.4102147039687702, "grad_norm": 2.136883402993382, "learning_rate": 6.379662475337301e-06, "loss": 0.4099, "step": 4335 }, { "epoch": 1.4105400130123618, "grad_norm": 1.6921736713167435, "learning_rate": 6.377843105135944e-06, "loss": 0.376, "step": 4336 }, { "epoch": 1.4108653220559533, "grad_norm": 1.7609203425511708, "learning_rate": 6.376023537488751e-06, "loss": 0.3608, "step": 4337 }, { "epoch": 1.4111906310995446, "grad_norm": 1.6613851803382351, "learning_rate": 6.374203772656469e-06, "loss": 0.2938, "step": 4338 }, { "epoch": 1.4115159401431359, "grad_norm": 1.5939506175395035, "learning_rate": 6.372383810899869e-06, "loss": 0.3227, "step": 4339 }, { "epoch": 1.4118412491867274, "grad_norm": 1.9504678198195384, "learning_rate": 6.3705636524797556e-06, "loss": 0.4354, "step": 4340 }, { "epoch": 1.412166558230319, "grad_norm": 1.7930118056016393, "learning_rate": 6.368743297656954e-06, "loss": 0.3984, "step": 4341 }, { "epoch": 1.4124918672739102, "grad_norm": 1.7348761412675768, "learning_rate": 6.366922746692329e-06, "loss": 0.3924, "step": 4342 }, { "epoch": 1.4128171763175017, "grad_norm": 1.9614249010194815, "learning_rate": 6.36510199984676e-06, "loss": 0.3889, "step": 4343 }, { "epoch": 1.413142485361093, "grad_norm": 1.9035222190778691, "learning_rate": 6.363281057381166e-06, "loss": 0.3673, "step": 4344 }, { "epoch": 1.4134677944046845, "grad_norm": 2.08222606214327, "learning_rate": 6.361459919556487e-06, "loss": 0.4217, "step": 4345 }, { "epoch": 1.4137931034482758, "grad_norm": 1.801668885574436, "learning_rate": 6.359638586633694e-06, "loss": 0.4233, "step": 4346 }, { "epoch": 1.4141184124918673, "grad_norm": 2.1292738715795387, "learning_rate": 6.357817058873783e-06, "loss": 0.486, "step": 4347 }, { "epoch": 1.4144437215354586, "grad_norm": 2.003832188148827, "learning_rate": 6.355995336537784e-06, "loss": 0.4364, "step": 4348 }, { "epoch": 1.4147690305790501, "grad_norm": 1.8968239646645435, "learning_rate": 6.35417341988675e-06, "loss": 0.3833, "step": 4349 }, { "epoch": 1.4150943396226414, "grad_norm": 1.794959566258927, "learning_rate": 6.352351309181761e-06, "loss": 0.3149, "step": 4350 }, { "epoch": 1.415419648666233, "grad_norm": 2.1008090440865104, "learning_rate": 6.3505290046839285e-06, "loss": 0.4993, "step": 4351 }, { "epoch": 1.4157449577098244, "grad_norm": 1.8682129916975136, "learning_rate": 6.348706506654389e-06, "loss": 0.3454, "step": 4352 }, { "epoch": 1.4160702667534157, "grad_norm": 1.876472295436608, "learning_rate": 6.346883815354307e-06, "loss": 0.3995, "step": 4353 }, { "epoch": 1.416395575797007, "grad_norm": 2.2598145480756533, "learning_rate": 6.345060931044876e-06, "loss": 0.4202, "step": 4354 }, { "epoch": 1.4167208848405985, "grad_norm": 1.867249165917756, "learning_rate": 6.343237853987318e-06, "loss": 0.4054, "step": 4355 }, { "epoch": 1.41704619388419, "grad_norm": 1.6465580823173973, "learning_rate": 6.34141458444288e-06, "loss": 0.3254, "step": 4356 }, { "epoch": 1.4173715029277814, "grad_norm": 1.9088813930695163, "learning_rate": 6.339591122672836e-06, "loss": 0.3619, "step": 4357 }, { "epoch": 1.4176968119713729, "grad_norm": 1.8896770348804224, "learning_rate": 6.337767468938494e-06, "loss": 0.4545, "step": 4358 }, { "epoch": 1.4180221210149642, "grad_norm": 2.0209951992221273, "learning_rate": 6.3359436235011785e-06, "loss": 0.5054, "step": 4359 }, { "epoch": 1.4183474300585557, "grad_norm": 1.933276443394172, "learning_rate": 6.334119586622253e-06, "loss": 0.3716, "step": 4360 }, { "epoch": 1.418672739102147, "grad_norm": 1.9088275997201098, "learning_rate": 6.3322953585631005e-06, "loss": 0.4605, "step": 4361 }, { "epoch": 1.4189980481457385, "grad_norm": 1.7784257980091787, "learning_rate": 6.330470939585136e-06, "loss": 0.3533, "step": 4362 }, { "epoch": 1.4193233571893298, "grad_norm": 2.2708631139183053, "learning_rate": 6.328646329949797e-06, "loss": 0.4514, "step": 4363 }, { "epoch": 1.4196486662329213, "grad_norm": 1.9332006937743758, "learning_rate": 6.3268215299185545e-06, "loss": 0.377, "step": 4364 }, { "epoch": 1.4199739752765126, "grad_norm": 1.8276295595527792, "learning_rate": 6.324996539752901e-06, "loss": 0.308, "step": 4365 }, { "epoch": 1.420299284320104, "grad_norm": 1.7079222487028114, "learning_rate": 6.32317135971436e-06, "loss": 0.3427, "step": 4366 }, { "epoch": 1.4206245933636956, "grad_norm": 2.2428823137893232, "learning_rate": 6.321345990064481e-06, "loss": 0.407, "step": 4367 }, { "epoch": 1.420949902407287, "grad_norm": 2.177023987823137, "learning_rate": 6.319520431064842e-06, "loss": 0.4836, "step": 4368 }, { "epoch": 1.4212752114508782, "grad_norm": 1.87052960867931, "learning_rate": 6.3176946829770445e-06, "loss": 0.432, "step": 4369 }, { "epoch": 1.4216005204944697, "grad_norm": 1.5687375835048951, "learning_rate": 6.315868746062721e-06, "loss": 0.2995, "step": 4370 }, { "epoch": 1.4219258295380612, "grad_norm": 2.263992241952478, "learning_rate": 6.314042620583528e-06, "loss": 0.5045, "step": 4371 }, { "epoch": 1.4222511385816525, "grad_norm": 1.5550373745974713, "learning_rate": 6.312216306801151e-06, "loss": 0.338, "step": 4372 }, { "epoch": 1.422576447625244, "grad_norm": 1.9832628960038505, "learning_rate": 6.310389804977303e-06, "loss": 0.4437, "step": 4373 }, { "epoch": 1.4229017566688353, "grad_norm": 1.8161440553534995, "learning_rate": 6.308563115373723e-06, "loss": 0.3775, "step": 4374 }, { "epoch": 1.4232270657124269, "grad_norm": 1.9382513886539068, "learning_rate": 6.306736238252175e-06, "loss": 0.3899, "step": 4375 }, { "epoch": 1.4235523747560181, "grad_norm": 1.868668380491538, "learning_rate": 6.304909173874452e-06, "loss": 0.393, "step": 4376 }, { "epoch": 1.4238776837996097, "grad_norm": 1.9838759079494785, "learning_rate": 6.303081922502376e-06, "loss": 0.371, "step": 4377 }, { "epoch": 1.4242029928432012, "grad_norm": 1.9198840682075256, "learning_rate": 6.301254484397792e-06, "loss": 0.3806, "step": 4378 }, { "epoch": 1.4245283018867925, "grad_norm": 1.8304923558629391, "learning_rate": 6.299426859822572e-06, "loss": 0.3347, "step": 4379 }, { "epoch": 1.4248536109303838, "grad_norm": 1.8131044359491304, "learning_rate": 6.297599049038616e-06, "loss": 0.3365, "step": 4380 }, { "epoch": 1.4251789199739753, "grad_norm": 1.9894193092878327, "learning_rate": 6.295771052307853e-06, "loss": 0.3277, "step": 4381 }, { "epoch": 1.4255042290175668, "grad_norm": 1.5839340678928815, "learning_rate": 6.293942869892231e-06, "loss": 0.2913, "step": 4382 }, { "epoch": 1.425829538061158, "grad_norm": 2.153988057900784, "learning_rate": 6.292114502053736e-06, "loss": 0.5567, "step": 4383 }, { "epoch": 1.4261548471047494, "grad_norm": 1.9956032509700832, "learning_rate": 6.290285949054371e-06, "loss": 0.4379, "step": 4384 }, { "epoch": 1.426480156148341, "grad_norm": 1.742980305256766, "learning_rate": 6.288457211156169e-06, "loss": 0.3605, "step": 4385 }, { "epoch": 1.4268054651919324, "grad_norm": 1.8175346731569022, "learning_rate": 6.286628288621188e-06, "loss": 0.3508, "step": 4386 }, { "epoch": 1.4271307742355237, "grad_norm": 2.1744997002363204, "learning_rate": 6.2847991817115185e-06, "loss": 0.4041, "step": 4387 }, { "epoch": 1.4274560832791152, "grad_norm": 2.0868622454002983, "learning_rate": 6.282969890689267e-06, "loss": 0.3633, "step": 4388 }, { "epoch": 1.4277813923227065, "grad_norm": 2.127209593513163, "learning_rate": 6.281140415816576e-06, "loss": 0.4506, "step": 4389 }, { "epoch": 1.428106701366298, "grad_norm": 1.9497639656566566, "learning_rate": 6.2793107573556074e-06, "loss": 0.364, "step": 4390 }, { "epoch": 1.4284320104098893, "grad_norm": 2.1360905601541433, "learning_rate": 6.277480915568556e-06, "loss": 0.5042, "step": 4391 }, { "epoch": 1.4287573194534808, "grad_norm": 1.9366217092819573, "learning_rate": 6.275650890717637e-06, "loss": 0.4151, "step": 4392 }, { "epoch": 1.4290826284970723, "grad_norm": 1.959962651863937, "learning_rate": 6.273820683065098e-06, "loss": 0.3154, "step": 4393 }, { "epoch": 1.4294079375406636, "grad_norm": 2.01943529508687, "learning_rate": 6.271990292873201e-06, "loss": 0.358, "step": 4394 }, { "epoch": 1.429733246584255, "grad_norm": 2.0530607746623084, "learning_rate": 6.27015972040425e-06, "loss": 0.3862, "step": 4395 }, { "epoch": 1.4300585556278465, "grad_norm": 1.6859641548222262, "learning_rate": 6.268328965920563e-06, "loss": 0.3744, "step": 4396 }, { "epoch": 1.430383864671438, "grad_norm": 1.8551449764545698, "learning_rate": 6.26649802968449e-06, "loss": 0.3674, "step": 4397 }, { "epoch": 1.4307091737150293, "grad_norm": 1.7561583470722426, "learning_rate": 6.264666911958404e-06, "loss": 0.3987, "step": 4398 }, { "epoch": 1.4310344827586206, "grad_norm": 1.9086876495550147, "learning_rate": 6.262835613004708e-06, "loss": 0.3712, "step": 4399 }, { "epoch": 1.431359791802212, "grad_norm": 1.6613896343453987, "learning_rate": 6.261004133085825e-06, "loss": 0.3705, "step": 4400 }, { "epoch": 1.4316851008458036, "grad_norm": 2.0312383291325657, "learning_rate": 6.25917247246421e-06, "loss": 0.3845, "step": 4401 }, { "epoch": 1.4320104098893949, "grad_norm": 2.074871780528014, "learning_rate": 6.25734063140234e-06, "loss": 0.471, "step": 4402 }, { "epoch": 1.4323357189329864, "grad_norm": 1.818026535993756, "learning_rate": 6.2555086101627194e-06, "loss": 0.3637, "step": 4403 }, { "epoch": 1.4326610279765777, "grad_norm": 1.6946551049120253, "learning_rate": 6.253676409007877e-06, "loss": 0.3384, "step": 4404 }, { "epoch": 1.4329863370201692, "grad_norm": 2.1199507414518006, "learning_rate": 6.25184402820037e-06, "loss": 0.4173, "step": 4405 }, { "epoch": 1.4333116460637605, "grad_norm": 1.9289835447451416, "learning_rate": 6.250011468002779e-06, "loss": 0.3491, "step": 4406 }, { "epoch": 1.433636955107352, "grad_norm": 1.9427505781570573, "learning_rate": 6.2481787286777116e-06, "loss": 0.3079, "step": 4407 }, { "epoch": 1.4339622641509435, "grad_norm": 1.6882527866265058, "learning_rate": 6.246345810487799e-06, "loss": 0.3224, "step": 4408 }, { "epoch": 1.4342875731945348, "grad_norm": 1.6769408331140068, "learning_rate": 6.244512713695702e-06, "loss": 0.3072, "step": 4409 }, { "epoch": 1.434612882238126, "grad_norm": 2.0309581167477306, "learning_rate": 6.242679438564103e-06, "loss": 0.3256, "step": 4410 }, { "epoch": 1.4349381912817176, "grad_norm": 2.001341549626523, "learning_rate": 6.240845985355712e-06, "loss": 0.4088, "step": 4411 }, { "epoch": 1.4352635003253091, "grad_norm": 2.0293770648925618, "learning_rate": 6.239012354333264e-06, "loss": 0.4524, "step": 4412 }, { "epoch": 1.4355888093689004, "grad_norm": 2.0881637256284926, "learning_rate": 6.23717854575952e-06, "loss": 0.4035, "step": 4413 }, { "epoch": 1.435914118412492, "grad_norm": 1.9808039280863645, "learning_rate": 6.235344559897265e-06, "loss": 0.3812, "step": 4414 }, { "epoch": 1.4362394274560832, "grad_norm": 2.1009102646624895, "learning_rate": 6.233510397009311e-06, "loss": 0.4061, "step": 4415 }, { "epoch": 1.4365647364996748, "grad_norm": 2.0942558350527904, "learning_rate": 6.231676057358495e-06, "loss": 0.447, "step": 4416 }, { "epoch": 1.436890045543266, "grad_norm": 1.8318989467967828, "learning_rate": 6.229841541207679e-06, "loss": 0.3671, "step": 4417 }, { "epoch": 1.4372153545868576, "grad_norm": 2.1929747282560412, "learning_rate": 6.228006848819747e-06, "loss": 0.4743, "step": 4418 }, { "epoch": 1.4375406636304489, "grad_norm": 2.1695620369977324, "learning_rate": 6.226171980457619e-06, "loss": 0.4364, "step": 4419 }, { "epoch": 1.4378659726740404, "grad_norm": 1.9391148190646412, "learning_rate": 6.224336936384227e-06, "loss": 0.3695, "step": 4420 }, { "epoch": 1.4381912817176317, "grad_norm": 1.670901913274823, "learning_rate": 6.222501716862535e-06, "loss": 0.3164, "step": 4421 }, { "epoch": 1.4385165907612232, "grad_norm": 2.0226713280771422, "learning_rate": 6.220666322155533e-06, "loss": 0.4126, "step": 4422 }, { "epoch": 1.4388418998048147, "grad_norm": 1.5770709562879943, "learning_rate": 6.218830752526234e-06, "loss": 0.3128, "step": 4423 }, { "epoch": 1.439167208848406, "grad_norm": 1.9364149750782058, "learning_rate": 6.216995008237673e-06, "loss": 0.3719, "step": 4424 }, { "epoch": 1.4394925178919973, "grad_norm": 1.9370714403662974, "learning_rate": 6.215159089552918e-06, "loss": 0.4667, "step": 4425 }, { "epoch": 1.4398178269355888, "grad_norm": 1.7796919038827779, "learning_rate": 6.213322996735054e-06, "loss": 0.3816, "step": 4426 }, { "epoch": 1.4401431359791803, "grad_norm": 2.1307848167070054, "learning_rate": 6.211486730047197e-06, "loss": 0.4046, "step": 4427 }, { "epoch": 1.4404684450227716, "grad_norm": 2.2522424383083357, "learning_rate": 6.2096502897524825e-06, "loss": 0.5777, "step": 4428 }, { "epoch": 1.4407937540663631, "grad_norm": 1.7161927654347258, "learning_rate": 6.207813676114076e-06, "loss": 0.2827, "step": 4429 }, { "epoch": 1.4411190631099544, "grad_norm": 2.071318914903698, "learning_rate": 6.205976889395164e-06, "loss": 0.3988, "step": 4430 }, { "epoch": 1.441444372153546, "grad_norm": 1.823907450621164, "learning_rate": 6.204139929858959e-06, "loss": 0.426, "step": 4431 }, { "epoch": 1.4417696811971372, "grad_norm": 1.6936778515452593, "learning_rate": 6.2023027977687e-06, "loss": 0.343, "step": 4432 }, { "epoch": 1.4420949902407287, "grad_norm": 1.7857748044902437, "learning_rate": 6.200465493387648e-06, "loss": 0.3346, "step": 4433 }, { "epoch": 1.44242029928432, "grad_norm": 2.2786312032786595, "learning_rate": 6.19862801697909e-06, "loss": 0.3912, "step": 4434 }, { "epoch": 1.4427456083279115, "grad_norm": 1.875579178167743, "learning_rate": 6.196790368806339e-06, "loss": 0.3257, "step": 4435 }, { "epoch": 1.4430709173715028, "grad_norm": 1.8168857811095487, "learning_rate": 6.1949525491327314e-06, "loss": 0.3603, "step": 4436 }, { "epoch": 1.4433962264150944, "grad_norm": 1.7652455278028074, "learning_rate": 6.193114558221624e-06, "loss": 0.3495, "step": 4437 }, { "epoch": 1.4437215354586859, "grad_norm": 1.6495377276741727, "learning_rate": 6.1912763963364074e-06, "loss": 0.4065, "step": 4438 }, { "epoch": 1.4440468445022772, "grad_norm": 1.9248638236005733, "learning_rate": 6.189438063740487e-06, "loss": 0.5015, "step": 4439 }, { "epoch": 1.4443721535458685, "grad_norm": 1.860770028927872, "learning_rate": 6.1875995606972995e-06, "loss": 0.4542, "step": 4440 }, { "epoch": 1.44469746258946, "grad_norm": 1.9149868378007038, "learning_rate": 6.185760887470304e-06, "loss": 0.4, "step": 4441 }, { "epoch": 1.4450227716330515, "grad_norm": 1.9779378022915834, "learning_rate": 6.183922044322982e-06, "loss": 0.4692, "step": 4442 }, { "epoch": 1.4453480806766428, "grad_norm": 1.7084854588445602, "learning_rate": 6.182083031518843e-06, "loss": 0.3377, "step": 4443 }, { "epoch": 1.4456733897202343, "grad_norm": 1.7993777920075078, "learning_rate": 6.180243849321414e-06, "loss": 0.4349, "step": 4444 }, { "epoch": 1.4459986987638256, "grad_norm": 1.9698601487225733, "learning_rate": 6.178404497994259e-06, "loss": 0.4153, "step": 4445 }, { "epoch": 1.446324007807417, "grad_norm": 1.8341697412408695, "learning_rate": 6.176564977800951e-06, "loss": 0.3545, "step": 4446 }, { "epoch": 1.4466493168510084, "grad_norm": 2.443174986441989, "learning_rate": 6.1747252890050975e-06, "loss": 0.3869, "step": 4447 }, { "epoch": 1.4469746258946, "grad_norm": 1.7569915397467062, "learning_rate": 6.172885431870328e-06, "loss": 0.3345, "step": 4448 }, { "epoch": 1.4472999349381912, "grad_norm": 1.7541152277672998, "learning_rate": 6.171045406660293e-06, "loss": 0.2685, "step": 4449 }, { "epoch": 1.4476252439817827, "grad_norm": 2.013274397074882, "learning_rate": 6.169205213638671e-06, "loss": 0.3723, "step": 4450 }, { "epoch": 1.447950553025374, "grad_norm": 1.8145575639883675, "learning_rate": 6.167364853069162e-06, "loss": 0.4079, "step": 4451 }, { "epoch": 1.4482758620689655, "grad_norm": 1.9921731200323762, "learning_rate": 6.165524325215492e-06, "loss": 0.4075, "step": 4452 }, { "epoch": 1.448601171112557, "grad_norm": 1.6900767454267178, "learning_rate": 6.1636836303414086e-06, "loss": 0.3612, "step": 4453 }, { "epoch": 1.4489264801561483, "grad_norm": 2.403555049076995, "learning_rate": 6.1618427687106845e-06, "loss": 0.3734, "step": 4454 }, { "epoch": 1.4492517891997396, "grad_norm": 2.1469509056408684, "learning_rate": 6.160001740587118e-06, "loss": 0.3191, "step": 4455 }, { "epoch": 1.4495770982433311, "grad_norm": 1.8420204398818927, "learning_rate": 6.158160546234527e-06, "loss": 0.3546, "step": 4456 }, { "epoch": 1.4499024072869227, "grad_norm": 1.6814648849193237, "learning_rate": 6.156319185916759e-06, "loss": 0.3458, "step": 4457 }, { "epoch": 1.450227716330514, "grad_norm": 1.8390387299890545, "learning_rate": 6.154477659897681e-06, "loss": 0.3927, "step": 4458 }, { "epoch": 1.4505530253741055, "grad_norm": 2.078515174545004, "learning_rate": 6.152635968441184e-06, "loss": 0.3641, "step": 4459 }, { "epoch": 1.4508783344176968, "grad_norm": 1.81009397072984, "learning_rate": 6.150794111811184e-06, "loss": 0.4275, "step": 4460 }, { "epoch": 1.4512036434612883, "grad_norm": 1.9569471004544783, "learning_rate": 6.148952090271622e-06, "loss": 0.3971, "step": 4461 }, { "epoch": 1.4515289525048796, "grad_norm": 1.8274450421825585, "learning_rate": 6.147109904086457e-06, "loss": 0.3538, "step": 4462 }, { "epoch": 1.451854261548471, "grad_norm": 1.7649524136573087, "learning_rate": 6.145267553519679e-06, "loss": 0.4012, "step": 4463 }, { "epoch": 1.4521795705920626, "grad_norm": 1.9131100610371503, "learning_rate": 6.143425038835298e-06, "loss": 0.4815, "step": 4464 }, { "epoch": 1.452504879635654, "grad_norm": 1.946906182817819, "learning_rate": 6.141582360297344e-06, "loss": 0.4035, "step": 4465 }, { "epoch": 1.4528301886792452, "grad_norm": 1.8901540214703159, "learning_rate": 6.139739518169879e-06, "loss": 0.3567, "step": 4466 }, { "epoch": 1.4531554977228367, "grad_norm": 1.8429149158087839, "learning_rate": 6.13789651271698e-06, "loss": 0.338, "step": 4467 }, { "epoch": 1.4534808067664282, "grad_norm": 2.235681196567662, "learning_rate": 6.136053344202752e-06, "loss": 0.4272, "step": 4468 }, { "epoch": 1.4538061158100195, "grad_norm": 2.1103681450157095, "learning_rate": 6.134210012891324e-06, "loss": 0.4243, "step": 4469 }, { "epoch": 1.4541314248536108, "grad_norm": 1.8066803780713214, "learning_rate": 6.132366519046843e-06, "loss": 0.4081, "step": 4470 }, { "epoch": 1.4544567338972023, "grad_norm": 1.8033472449095094, "learning_rate": 6.130522862933486e-06, "loss": 0.334, "step": 4471 }, { "epoch": 1.4547820429407938, "grad_norm": 1.8956170654900066, "learning_rate": 6.128679044815449e-06, "loss": 0.3679, "step": 4472 }, { "epoch": 1.4551073519843851, "grad_norm": 2.263123688368698, "learning_rate": 6.126835064956952e-06, "loss": 0.4119, "step": 4473 }, { "epoch": 1.4554326610279766, "grad_norm": 1.769874691771892, "learning_rate": 6.12499092362224e-06, "loss": 0.338, "step": 4474 }, { "epoch": 1.455757970071568, "grad_norm": 1.8644462438044134, "learning_rate": 6.123146621075576e-06, "loss": 0.3788, "step": 4475 }, { "epoch": 1.4560832791151594, "grad_norm": 2.002374829491037, "learning_rate": 6.121302157581254e-06, "loss": 0.4414, "step": 4476 }, { "epoch": 1.4564085881587507, "grad_norm": 1.6936400527887752, "learning_rate": 6.119457533403585e-06, "loss": 0.3404, "step": 4477 }, { "epoch": 1.4567338972023423, "grad_norm": 2.1306544304045607, "learning_rate": 6.1176127488069045e-06, "loss": 0.3732, "step": 4478 }, { "epoch": 1.4570592062459338, "grad_norm": 2.029609948534565, "learning_rate": 6.1157678040555735e-06, "loss": 0.4095, "step": 4479 }, { "epoch": 1.457384515289525, "grad_norm": 1.9427459291626759, "learning_rate": 6.113922699413972e-06, "loss": 0.3733, "step": 4480 }, { "epoch": 1.4577098243331164, "grad_norm": 1.6971341391455816, "learning_rate": 6.112077435146504e-06, "loss": 0.3745, "step": 4481 }, { "epoch": 1.4580351333767079, "grad_norm": 1.9958391589117512, "learning_rate": 6.110232011517599e-06, "loss": 0.3793, "step": 4482 }, { "epoch": 1.4583604424202994, "grad_norm": 2.125520646279647, "learning_rate": 6.108386428791705e-06, "loss": 0.4748, "step": 4483 }, { "epoch": 1.4586857514638907, "grad_norm": 1.938359638958231, "learning_rate": 6.106540687233298e-06, "loss": 0.4601, "step": 4484 }, { "epoch": 1.459011060507482, "grad_norm": 1.9974403469950521, "learning_rate": 6.104694787106873e-06, "loss": 0.3835, "step": 4485 }, { "epoch": 1.4593363695510735, "grad_norm": 1.8286744520594864, "learning_rate": 6.102848728676947e-06, "loss": 0.4416, "step": 4486 }, { "epoch": 1.459661678594665, "grad_norm": 1.8769085484903036, "learning_rate": 6.1010025122080655e-06, "loss": 0.3946, "step": 4487 }, { "epoch": 1.4599869876382563, "grad_norm": 1.8441362505841432, "learning_rate": 6.099156137964787e-06, "loss": 0.378, "step": 4488 }, { "epoch": 1.4603122966818478, "grad_norm": 1.8457308380605078, "learning_rate": 6.097309606211702e-06, "loss": 0.4386, "step": 4489 }, { "epoch": 1.460637605725439, "grad_norm": 1.8080238378202023, "learning_rate": 6.09546291721342e-06, "loss": 0.3224, "step": 4490 }, { "epoch": 1.4609629147690306, "grad_norm": 1.829098528704883, "learning_rate": 6.0936160712345705e-06, "loss": 0.3822, "step": 4491 }, { "epoch": 1.461288223812622, "grad_norm": 1.5424626225110283, "learning_rate": 6.091769068539808e-06, "loss": 0.31, "step": 4492 }, { "epoch": 1.4616135328562134, "grad_norm": 1.8145163466026122, "learning_rate": 6.089921909393812e-06, "loss": 0.3455, "step": 4493 }, { "epoch": 1.461938841899805, "grad_norm": 1.6407424687563124, "learning_rate": 6.088074594061279e-06, "loss": 0.3157, "step": 4494 }, { "epoch": 1.4622641509433962, "grad_norm": 1.7698289137137422, "learning_rate": 6.086227122806931e-06, "loss": 0.2963, "step": 4495 }, { "epoch": 1.4625894599869875, "grad_norm": 2.0752664989312346, "learning_rate": 6.084379495895512e-06, "loss": 0.4189, "step": 4496 }, { "epoch": 1.462914769030579, "grad_norm": 1.762436046498322, "learning_rate": 6.08253171359179e-06, "loss": 0.3062, "step": 4497 }, { "epoch": 1.4632400780741706, "grad_norm": 2.107462765853073, "learning_rate": 6.080683776160549e-06, "loss": 0.4459, "step": 4498 }, { "epoch": 1.4635653871177619, "grad_norm": 1.6655194460977578, "learning_rate": 6.078835683866602e-06, "loss": 0.3459, "step": 4499 }, { "epoch": 1.4638906961613534, "grad_norm": 1.8553539868893711, "learning_rate": 6.076987436974785e-06, "loss": 0.3989, "step": 4500 }, { "epoch": 1.4642160052049447, "grad_norm": 2.2661512471191303, "learning_rate": 6.0751390357499485e-06, "loss": 0.415, "step": 4501 }, { "epoch": 1.4645413142485362, "grad_norm": 2.2162447591099568, "learning_rate": 6.073290480456972e-06, "loss": 0.4063, "step": 4502 }, { "epoch": 1.4648666232921275, "grad_norm": 2.2513083107259573, "learning_rate": 6.071441771360753e-06, "loss": 0.4571, "step": 4503 }, { "epoch": 1.465191932335719, "grad_norm": 2.0175433649730152, "learning_rate": 6.069592908726215e-06, "loss": 0.3633, "step": 4504 }, { "epoch": 1.4655172413793103, "grad_norm": 1.9874495694551857, "learning_rate": 6.067743892818299e-06, "loss": 0.3547, "step": 4505 }, { "epoch": 1.4658425504229018, "grad_norm": 1.774364575512831, "learning_rate": 6.065894723901973e-06, "loss": 0.3764, "step": 4506 }, { "epoch": 1.466167859466493, "grad_norm": 2.1578666349257016, "learning_rate": 6.06404540224222e-06, "loss": 0.4156, "step": 4507 }, { "epoch": 1.4664931685100846, "grad_norm": 2.180040253927554, "learning_rate": 6.0621959281040536e-06, "loss": 0.4696, "step": 4508 }, { "epoch": 1.4668184775536761, "grad_norm": 1.7463821336386067, "learning_rate": 6.060346301752502e-06, "loss": 0.3459, "step": 4509 }, { "epoch": 1.4671437865972674, "grad_norm": 1.7719228625048857, "learning_rate": 6.058496523452619e-06, "loss": 0.3447, "step": 4510 }, { "epoch": 1.4674690956408587, "grad_norm": 1.957550502982453, "learning_rate": 6.056646593469475e-06, "loss": 0.4161, "step": 4511 }, { "epoch": 1.4677944046844502, "grad_norm": 1.862774170663318, "learning_rate": 6.054796512068173e-06, "loss": 0.3936, "step": 4512 }, { "epoch": 1.4681197137280417, "grad_norm": 1.7044244853543917, "learning_rate": 6.052946279513827e-06, "loss": 0.3345, "step": 4513 }, { "epoch": 1.468445022771633, "grad_norm": 1.7350844382687929, "learning_rate": 6.051095896071577e-06, "loss": 0.3394, "step": 4514 }, { "epoch": 1.4687703318152245, "grad_norm": 1.8449048034686435, "learning_rate": 6.049245362006584e-06, "loss": 0.4015, "step": 4515 }, { "epoch": 1.4690956408588158, "grad_norm": 1.7961780135683427, "learning_rate": 6.047394677584032e-06, "loss": 0.3629, "step": 4516 }, { "epoch": 1.4694209499024073, "grad_norm": 1.8457425438100616, "learning_rate": 6.045543843069123e-06, "loss": 0.3611, "step": 4517 }, { "epoch": 1.4697462589459986, "grad_norm": 1.9990524532515446, "learning_rate": 6.043692858727085e-06, "loss": 0.3838, "step": 4518 }, { "epoch": 1.4700715679895902, "grad_norm": 1.8663767826699422, "learning_rate": 6.041841724823164e-06, "loss": 0.397, "step": 4519 }, { "epoch": 1.4703968770331814, "grad_norm": 1.7473416689620072, "learning_rate": 6.0399904416226295e-06, "loss": 0.4298, "step": 4520 }, { "epoch": 1.470722186076773, "grad_norm": 2.2175828853663075, "learning_rate": 6.038139009390771e-06, "loss": 0.5705, "step": 4521 }, { "epoch": 1.4710474951203643, "grad_norm": 1.9614776613116067, "learning_rate": 6.0362874283928995e-06, "loss": 0.3733, "step": 4522 }, { "epoch": 1.4713728041639558, "grad_norm": 2.4365276266302747, "learning_rate": 6.034435698894351e-06, "loss": 0.4821, "step": 4523 }, { "epoch": 1.4716981132075473, "grad_norm": 1.857715245882551, "learning_rate": 6.032583821160476e-06, "loss": 0.3914, "step": 4524 }, { "epoch": 1.4720234222511386, "grad_norm": 1.7837541356230422, "learning_rate": 6.03073179545665e-06, "loss": 0.3343, "step": 4525 }, { "epoch": 1.4723487312947299, "grad_norm": 2.193440690310382, "learning_rate": 6.0288796220482715e-06, "loss": 0.4517, "step": 4526 }, { "epoch": 1.4726740403383214, "grad_norm": 1.7513481894194678, "learning_rate": 6.027027301200756e-06, "loss": 0.3926, "step": 4527 }, { "epoch": 1.472999349381913, "grad_norm": 1.7757867676343642, "learning_rate": 6.025174833179546e-06, "loss": 0.3922, "step": 4528 }, { "epoch": 1.4733246584255042, "grad_norm": 1.8253001416347685, "learning_rate": 6.023322218250097e-06, "loss": 0.3513, "step": 4529 }, { "epoch": 1.4736499674690957, "grad_norm": 2.371170575095469, "learning_rate": 6.021469456677892e-06, "loss": 0.4026, "step": 4530 }, { "epoch": 1.473975276512687, "grad_norm": 2.038807077688439, "learning_rate": 6.019616548728432e-06, "loss": 0.4238, "step": 4531 }, { "epoch": 1.4743005855562785, "grad_norm": 1.9734377702617434, "learning_rate": 6.0177634946672436e-06, "loss": 0.3263, "step": 4532 }, { "epoch": 1.4746258945998698, "grad_norm": 1.9012918519670579, "learning_rate": 6.0159102947598645e-06, "loss": 0.3824, "step": 4533 }, { "epoch": 1.4749512036434613, "grad_norm": 2.3214066650226246, "learning_rate": 6.014056949271865e-06, "loss": 0.441, "step": 4534 }, { "epoch": 1.4752765126870526, "grad_norm": 2.345739438994774, "learning_rate": 6.012203458468828e-06, "loss": 0.4719, "step": 4535 }, { "epoch": 1.4756018217306441, "grad_norm": 1.8927479308071429, "learning_rate": 6.01034982261636e-06, "loss": 0.4026, "step": 4536 }, { "epoch": 1.4759271307742354, "grad_norm": 2.0822904102579267, "learning_rate": 6.008496041980089e-06, "loss": 0.3678, "step": 4537 }, { "epoch": 1.476252439817827, "grad_norm": 1.919282533568131, "learning_rate": 6.006642116825662e-06, "loss": 0.3883, "step": 4538 }, { "epoch": 1.4765777488614185, "grad_norm": 2.0472416347771083, "learning_rate": 6.00478804741875e-06, "loss": 0.3927, "step": 4539 }, { "epoch": 1.4769030579050098, "grad_norm": 1.8579179285689662, "learning_rate": 6.002933834025042e-06, "loss": 0.3067, "step": 4540 }, { "epoch": 1.477228366948601, "grad_norm": 1.8915202576706291, "learning_rate": 6.001079476910244e-06, "loss": 0.3354, "step": 4541 }, { "epoch": 1.4775536759921926, "grad_norm": 2.062933453427332, "learning_rate": 5.999224976340092e-06, "loss": 0.3988, "step": 4542 }, { "epoch": 1.477878985035784, "grad_norm": 1.7371291842769485, "learning_rate": 5.997370332580333e-06, "loss": 0.3426, "step": 4543 }, { "epoch": 1.4782042940793754, "grad_norm": 1.6368590462289891, "learning_rate": 5.995515545896741e-06, "loss": 0.3842, "step": 4544 }, { "epoch": 1.4785296031229669, "grad_norm": 1.931721796981134, "learning_rate": 5.993660616555107e-06, "loss": 0.3758, "step": 4545 }, { "epoch": 1.4788549121665582, "grad_norm": 2.085092281794047, "learning_rate": 5.991805544821246e-06, "loss": 0.4205, "step": 4546 }, { "epoch": 1.4791802212101497, "grad_norm": 1.997445293287473, "learning_rate": 5.989950330960985e-06, "loss": 0.4001, "step": 4547 }, { "epoch": 1.479505530253741, "grad_norm": 1.7556394264323645, "learning_rate": 5.988094975240186e-06, "loss": 0.341, "step": 4548 }, { "epoch": 1.4798308392973325, "grad_norm": 1.8970192416079756, "learning_rate": 5.9862394779247155e-06, "loss": 0.3251, "step": 4549 }, { "epoch": 1.480156148340924, "grad_norm": 2.1892918391270753, "learning_rate": 5.984383839280472e-06, "loss": 0.4726, "step": 4550 }, { "epoch": 1.4804814573845153, "grad_norm": 1.876303580117379, "learning_rate": 5.982528059573367e-06, "loss": 0.4619, "step": 4551 }, { "epoch": 1.4808067664281066, "grad_norm": 2.0079013648201895, "learning_rate": 5.9806721390693364e-06, "loss": 0.3881, "step": 4552 }, { "epoch": 1.4811320754716981, "grad_norm": 2.2208192448106217, "learning_rate": 5.978816078034334e-06, "loss": 0.4176, "step": 4553 }, { "epoch": 1.4814573845152896, "grad_norm": 1.8900823566101514, "learning_rate": 5.9769598767343355e-06, "loss": 0.3607, "step": 4554 }, { "epoch": 1.481782693558881, "grad_norm": 2.233236944210306, "learning_rate": 5.975103535435335e-06, "loss": 0.4613, "step": 4555 }, { "epoch": 1.4821080026024722, "grad_norm": 1.9475303353787083, "learning_rate": 5.973247054403348e-06, "loss": 0.3497, "step": 4556 }, { "epoch": 1.4824333116460637, "grad_norm": 1.8415935095220124, "learning_rate": 5.971390433904408e-06, "loss": 0.3976, "step": 4557 }, { "epoch": 1.4827586206896552, "grad_norm": 1.9847022093217306, "learning_rate": 5.969533674204574e-06, "loss": 0.3872, "step": 4558 }, { "epoch": 1.4830839297332465, "grad_norm": 1.9044734296330053, "learning_rate": 5.967676775569915e-06, "loss": 0.416, "step": 4559 }, { "epoch": 1.483409238776838, "grad_norm": 1.8388929442140334, "learning_rate": 5.965819738266531e-06, "loss": 0.3701, "step": 4560 }, { "epoch": 1.4837345478204294, "grad_norm": 2.228069530813059, "learning_rate": 5.963962562560534e-06, "loss": 0.393, "step": 4561 }, { "epoch": 1.4840598568640209, "grad_norm": 1.776338557308041, "learning_rate": 5.962105248718058e-06, "loss": 0.3034, "step": 4562 }, { "epoch": 1.4843851659076122, "grad_norm": 1.760178791456149, "learning_rate": 5.960247797005259e-06, "loss": 0.3491, "step": 4563 }, { "epoch": 1.4847104749512037, "grad_norm": 1.9444361234608023, "learning_rate": 5.95839020768831e-06, "loss": 0.451, "step": 4564 }, { "epoch": 1.4850357839947952, "grad_norm": 2.13467419199994, "learning_rate": 5.956532481033407e-06, "loss": 0.4881, "step": 4565 }, { "epoch": 1.4853610930383865, "grad_norm": 1.9752947979814712, "learning_rate": 5.95467461730676e-06, "loss": 0.3866, "step": 4566 }, { "epoch": 1.4856864020819778, "grad_norm": 1.7590105188231837, "learning_rate": 5.952816616774605e-06, "loss": 0.3618, "step": 4567 }, { "epoch": 1.4860117111255693, "grad_norm": 1.8911969637815278, "learning_rate": 5.9509584797031925e-06, "loss": 0.409, "step": 4568 }, { "epoch": 1.4863370201691608, "grad_norm": 1.7994204490242387, "learning_rate": 5.949100206358795e-06, "loss": 0.3409, "step": 4569 }, { "epoch": 1.486662329212752, "grad_norm": 1.639232305393448, "learning_rate": 5.9472417970077075e-06, "loss": 0.3446, "step": 4570 }, { "epoch": 1.4869876382563434, "grad_norm": 1.878628037160236, "learning_rate": 5.945383251916238e-06, "loss": 0.396, "step": 4571 }, { "epoch": 1.487312947299935, "grad_norm": 1.7406240006541562, "learning_rate": 5.9435245713507185e-06, "loss": 0.361, "step": 4572 }, { "epoch": 1.4876382563435264, "grad_norm": 1.879996734099377, "learning_rate": 5.941665755577497e-06, "loss": 0.4422, "step": 4573 }, { "epoch": 1.4879635653871177, "grad_norm": 2.1300067219342584, "learning_rate": 5.939806804862949e-06, "loss": 0.3964, "step": 4574 }, { "epoch": 1.4882888744307092, "grad_norm": 1.795599114756051, "learning_rate": 5.9379477194734545e-06, "loss": 0.4282, "step": 4575 }, { "epoch": 1.4886141834743005, "grad_norm": 1.912840560669269, "learning_rate": 5.93608849967543e-06, "loss": 0.3395, "step": 4576 }, { "epoch": 1.488939492517892, "grad_norm": 1.965760310862704, "learning_rate": 5.9342291457352975e-06, "loss": 0.3285, "step": 4577 }, { "epoch": 1.4892648015614833, "grad_norm": 2.1886139494487127, "learning_rate": 5.932369657919506e-06, "loss": 0.3959, "step": 4578 }, { "epoch": 1.4895901106050748, "grad_norm": 2.108802456510415, "learning_rate": 5.93051003649452e-06, "loss": 0.3796, "step": 4579 }, { "epoch": 1.4899154196486664, "grad_norm": 1.8918177222533825, "learning_rate": 5.928650281726826e-06, "loss": 0.3348, "step": 4580 }, { "epoch": 1.4902407286922577, "grad_norm": 2.023073568982563, "learning_rate": 5.9267903938829284e-06, "loss": 0.4214, "step": 4581 }, { "epoch": 1.490566037735849, "grad_norm": 1.716717627524723, "learning_rate": 5.924930373229348e-06, "loss": 0.3731, "step": 4582 }, { "epoch": 1.4908913467794405, "grad_norm": 2.0370468746734027, "learning_rate": 5.923070220032629e-06, "loss": 0.3417, "step": 4583 }, { "epoch": 1.491216655823032, "grad_norm": 2.0597759290045348, "learning_rate": 5.921209934559333e-06, "loss": 0.4232, "step": 4584 }, { "epoch": 1.4915419648666233, "grad_norm": 1.8317335641292933, "learning_rate": 5.9193495170760375e-06, "loss": 0.3724, "step": 4585 }, { "epoch": 1.4918672739102148, "grad_norm": 2.430030379673864, "learning_rate": 5.917488967849345e-06, "loss": 0.4734, "step": 4586 }, { "epoch": 1.492192582953806, "grad_norm": 2.031041095431902, "learning_rate": 5.9156282871458705e-06, "loss": 0.4176, "step": 4587 }, { "epoch": 1.4925178919973976, "grad_norm": 2.0824493906717754, "learning_rate": 5.913767475232253e-06, "loss": 0.4344, "step": 4588 }, { "epoch": 1.4928432010409889, "grad_norm": 1.507209794365775, "learning_rate": 5.911906532375145e-06, "loss": 0.2843, "step": 4589 }, { "epoch": 1.4931685100845804, "grad_norm": 1.6113968754737429, "learning_rate": 5.910045458841226e-06, "loss": 0.2776, "step": 4590 }, { "epoch": 1.4934938191281717, "grad_norm": 1.7595978730260193, "learning_rate": 5.908184254897183e-06, "loss": 0.3766, "step": 4591 }, { "epoch": 1.4938191281717632, "grad_norm": 1.6923423533324795, "learning_rate": 5.906322920809732e-06, "loss": 0.4088, "step": 4592 }, { "epoch": 1.4941444372153545, "grad_norm": 1.911279587557715, "learning_rate": 5.904461456845603e-06, "loss": 0.4, "step": 4593 }, { "epoch": 1.494469746258946, "grad_norm": 2.028883890099962, "learning_rate": 5.902599863271543e-06, "loss": 0.3917, "step": 4594 }, { "epoch": 1.4947950553025375, "grad_norm": 1.8182100061068267, "learning_rate": 5.900738140354321e-06, "loss": 0.3673, "step": 4595 }, { "epoch": 1.4951203643461288, "grad_norm": 1.9005035551348481, "learning_rate": 5.898876288360723e-06, "loss": 0.4122, "step": 4596 }, { "epoch": 1.4954456733897201, "grad_norm": 2.287773459439513, "learning_rate": 5.8970143075575535e-06, "loss": 0.3792, "step": 4597 }, { "epoch": 1.4957709824333116, "grad_norm": 1.9541482672964394, "learning_rate": 5.895152198211636e-06, "loss": 0.3946, "step": 4598 }, { "epoch": 1.4960962914769032, "grad_norm": 1.835772723075395, "learning_rate": 5.893289960589811e-06, "loss": 0.3284, "step": 4599 }, { "epoch": 1.4964216005204944, "grad_norm": 1.5718759838419016, "learning_rate": 5.89142759495894e-06, "loss": 0.2969, "step": 4600 }, { "epoch": 1.496746909564086, "grad_norm": 1.9549060993041192, "learning_rate": 5.889565101585899e-06, "loss": 0.3901, "step": 4601 }, { "epoch": 1.4970722186076773, "grad_norm": 1.7617515805621078, "learning_rate": 5.887702480737587e-06, "loss": 0.3764, "step": 4602 }, { "epoch": 1.4973975276512688, "grad_norm": 1.97201230617549, "learning_rate": 5.885839732680918e-06, "loss": 0.344, "step": 4603 }, { "epoch": 1.49772283669486, "grad_norm": 2.1313892533488534, "learning_rate": 5.883976857682825e-06, "loss": 0.423, "step": 4604 }, { "epoch": 1.4980481457384516, "grad_norm": 1.8660228311295968, "learning_rate": 5.882113856010256e-06, "loss": 0.3203, "step": 4605 }, { "epoch": 1.4983734547820429, "grad_norm": 1.4965037048578407, "learning_rate": 5.880250727930186e-06, "loss": 0.2541, "step": 4606 }, { "epoch": 1.4986987638256344, "grad_norm": 1.8542854185453752, "learning_rate": 5.878387473709599e-06, "loss": 0.3689, "step": 4607 }, { "epoch": 1.4990240728692257, "grad_norm": 2.05821625007477, "learning_rate": 5.876524093615502e-06, "loss": 0.3541, "step": 4608 }, { "epoch": 1.4993493819128172, "grad_norm": 1.8341250723210099, "learning_rate": 5.874660587914918e-06, "loss": 0.3692, "step": 4609 }, { "epoch": 1.4996746909564087, "grad_norm": 1.9416792677913177, "learning_rate": 5.872796956874888e-06, "loss": 0.4246, "step": 4610 }, { "epoch": 1.5, "grad_norm": 1.7061275459581224, "learning_rate": 5.870933200762471e-06, "loss": 0.3525, "step": 4611 }, { "epoch": 1.5003253090435913, "grad_norm": 1.8423879516824113, "learning_rate": 5.8690693198447464e-06, "loss": 0.3557, "step": 4612 }, { "epoch": 1.5006506180871828, "grad_norm": 1.8103118515626975, "learning_rate": 5.867205314388811e-06, "loss": 0.3588, "step": 4613 }, { "epoch": 1.5009759271307743, "grad_norm": 1.8310264798919904, "learning_rate": 5.865341184661774e-06, "loss": 0.3806, "step": 4614 }, { "epoch": 1.5013012361743656, "grad_norm": 1.9569198314363705, "learning_rate": 5.863476930930767e-06, "loss": 0.4072, "step": 4615 }, { "epoch": 1.501626545217957, "grad_norm": 1.9033832519590215, "learning_rate": 5.861612553462943e-06, "loss": 0.4223, "step": 4616 }, { "epoch": 1.5019518542615484, "grad_norm": 1.6991713543788431, "learning_rate": 5.859748052525464e-06, "loss": 0.3135, "step": 4617 }, { "epoch": 1.50227716330514, "grad_norm": 1.601413387311972, "learning_rate": 5.857883428385516e-06, "loss": 0.3074, "step": 4618 }, { "epoch": 1.5026024723487312, "grad_norm": 1.7643155642022978, "learning_rate": 5.8560186813103e-06, "loss": 0.3591, "step": 4619 }, { "epoch": 1.5029277813923227, "grad_norm": 1.6295212729941817, "learning_rate": 5.854153811567039e-06, "loss": 0.3352, "step": 4620 }, { "epoch": 1.5032530904359143, "grad_norm": 1.8455245567319152, "learning_rate": 5.852288819422963e-06, "loss": 0.3891, "step": 4621 }, { "epoch": 1.5035783994795056, "grad_norm": 1.9108539674800455, "learning_rate": 5.850423705145334e-06, "loss": 0.38, "step": 4622 }, { "epoch": 1.5039037085230968, "grad_norm": 1.5237103969960841, "learning_rate": 5.84855846900142e-06, "loss": 0.3307, "step": 4623 }, { "epoch": 1.5042290175666884, "grad_norm": 1.903276542349119, "learning_rate": 5.846693111258512e-06, "loss": 0.3909, "step": 4624 }, { "epoch": 1.5045543266102799, "grad_norm": 2.0293485094940715, "learning_rate": 5.844827632183916e-06, "loss": 0.3587, "step": 4625 }, { "epoch": 1.5048796356538712, "grad_norm": 1.914850222453087, "learning_rate": 5.842962032044959e-06, "loss": 0.375, "step": 4626 }, { "epoch": 1.5052049446974625, "grad_norm": 1.9175019923852217, "learning_rate": 5.841096311108978e-06, "loss": 0.4061, "step": 4627 }, { "epoch": 1.505530253741054, "grad_norm": 1.9015779628775389, "learning_rate": 5.839230469643335e-06, "loss": 0.3582, "step": 4628 }, { "epoch": 1.5058555627846455, "grad_norm": 1.8128208868809463, "learning_rate": 5.837364507915408e-06, "loss": 0.3476, "step": 4629 }, { "epoch": 1.5061808718282368, "grad_norm": 2.056264247868984, "learning_rate": 5.8354984261925885e-06, "loss": 0.4707, "step": 4630 }, { "epoch": 1.506506180871828, "grad_norm": 1.9354841983657767, "learning_rate": 5.8336322247422875e-06, "loss": 0.4259, "step": 4631 }, { "epoch": 1.5068314899154196, "grad_norm": 1.9379810223605634, "learning_rate": 5.8317659038319315e-06, "loss": 0.3503, "step": 4632 }, { "epoch": 1.5071567989590111, "grad_norm": 1.954964736428264, "learning_rate": 5.8298994637289686e-06, "loss": 0.3979, "step": 4633 }, { "epoch": 1.5074821080026024, "grad_norm": 2.0206581540160813, "learning_rate": 5.828032904700858e-06, "loss": 0.3547, "step": 4634 }, { "epoch": 1.507807417046194, "grad_norm": 2.190814336838948, "learning_rate": 5.826166227015082e-06, "loss": 0.4691, "step": 4635 }, { "epoch": 1.5081327260897854, "grad_norm": 1.969708448824403, "learning_rate": 5.824299430939132e-06, "loss": 0.4067, "step": 4636 }, { "epoch": 1.5084580351333767, "grad_norm": 2.2309496243657407, "learning_rate": 5.822432516740526e-06, "loss": 0.5118, "step": 4637 }, { "epoch": 1.508783344176968, "grad_norm": 1.6778653771998728, "learning_rate": 5.820565484686791e-06, "loss": 0.3001, "step": 4638 }, { "epoch": 1.5091086532205595, "grad_norm": 2.0131777997410834, "learning_rate": 5.818698335045476e-06, "loss": 0.4242, "step": 4639 }, { "epoch": 1.509433962264151, "grad_norm": 1.9520319790932457, "learning_rate": 5.816831068084142e-06, "loss": 0.4382, "step": 4640 }, { "epoch": 1.5097592713077423, "grad_norm": 2.0237957350847737, "learning_rate": 5.8149636840703726e-06, "loss": 0.4178, "step": 4641 }, { "epoch": 1.5100845803513336, "grad_norm": 1.7816484042667375, "learning_rate": 5.8130961832717634e-06, "loss": 0.3208, "step": 4642 }, { "epoch": 1.5104098893949252, "grad_norm": 2.046882420194204, "learning_rate": 5.81122856595593e-06, "loss": 0.4092, "step": 4643 }, { "epoch": 1.5107351984385167, "grad_norm": 1.9270021327379032, "learning_rate": 5.8093608323905e-06, "loss": 0.4929, "step": 4644 }, { "epoch": 1.511060507482108, "grad_norm": 1.925623356039329, "learning_rate": 5.807492982843126e-06, "loss": 0.4182, "step": 4645 }, { "epoch": 1.5113858165256993, "grad_norm": 1.8095338506955454, "learning_rate": 5.805625017581467e-06, "loss": 0.4225, "step": 4646 }, { "epoch": 1.511711125569291, "grad_norm": 2.0457456652858617, "learning_rate": 5.803756936873207e-06, "loss": 0.4017, "step": 4647 }, { "epoch": 1.5120364346128823, "grad_norm": 1.9701404414671606, "learning_rate": 5.801888740986043e-06, "loss": 0.3483, "step": 4648 }, { "epoch": 1.5123617436564736, "grad_norm": 2.0525843561757613, "learning_rate": 5.800020430187686e-06, "loss": 0.3834, "step": 4649 }, { "epoch": 1.512687052700065, "grad_norm": 1.7644627826419061, "learning_rate": 5.798152004745867e-06, "loss": 0.3521, "step": 4650 }, { "epoch": 1.5130123617436566, "grad_norm": 2.097961304381957, "learning_rate": 5.796283464928336e-06, "loss": 0.3931, "step": 4651 }, { "epoch": 1.513337670787248, "grad_norm": 1.9437011383313136, "learning_rate": 5.794414811002853e-06, "loss": 0.3653, "step": 4652 }, { "epoch": 1.5136629798308392, "grad_norm": 1.999422762923862, "learning_rate": 5.7925460432371975e-06, "loss": 0.4336, "step": 4653 }, { "epoch": 1.5139882888744307, "grad_norm": 1.7530151070506927, "learning_rate": 5.790677161899165e-06, "loss": 0.3774, "step": 4654 }, { "epoch": 1.5143135979180222, "grad_norm": 1.8651880915439243, "learning_rate": 5.78880816725657e-06, "loss": 0.3258, "step": 4655 }, { "epoch": 1.5146389069616135, "grad_norm": 1.8630399671636881, "learning_rate": 5.786939059577236e-06, "loss": 0.3361, "step": 4656 }, { "epoch": 1.5149642160052048, "grad_norm": 1.8849104833902757, "learning_rate": 5.785069839129012e-06, "loss": 0.3675, "step": 4657 }, { "epoch": 1.5152895250487963, "grad_norm": 1.7911223700627281, "learning_rate": 5.7832005061797555e-06, "loss": 0.3229, "step": 4658 }, { "epoch": 1.5156148340923878, "grad_norm": 2.029988761365657, "learning_rate": 5.781331060997346e-06, "loss": 0.4118, "step": 4659 }, { "epoch": 1.5159401431359791, "grad_norm": 2.0468964554466687, "learning_rate": 5.779461503849672e-06, "loss": 0.4078, "step": 4660 }, { "epoch": 1.5162654521795704, "grad_norm": 1.949850578169451, "learning_rate": 5.777591835004648e-06, "loss": 0.4121, "step": 4661 }, { "epoch": 1.5165907612231622, "grad_norm": 2.394691084660908, "learning_rate": 5.7757220547301926e-06, "loss": 0.414, "step": 4662 }, { "epoch": 1.5169160702667535, "grad_norm": 1.7128302288931805, "learning_rate": 5.773852163294251e-06, "loss": 0.3287, "step": 4663 }, { "epoch": 1.5172413793103448, "grad_norm": 1.778958734917958, "learning_rate": 5.771982160964779e-06, "loss": 0.3534, "step": 4664 }, { "epoch": 1.5175666883539363, "grad_norm": 1.7251925059046453, "learning_rate": 5.770112048009747e-06, "loss": 0.373, "step": 4665 }, { "epoch": 1.5178919973975278, "grad_norm": 1.5876338057443036, "learning_rate": 5.768241824697145e-06, "loss": 0.3842, "step": 4666 }, { "epoch": 1.518217306441119, "grad_norm": 2.1246322433603, "learning_rate": 5.766371491294979e-06, "loss": 0.4134, "step": 4667 }, { "epoch": 1.5185426154847104, "grad_norm": 1.6869821617447207, "learning_rate": 5.764501048071266e-06, "loss": 0.3182, "step": 4668 }, { "epoch": 1.5188679245283019, "grad_norm": 2.0930144182078396, "learning_rate": 5.762630495294044e-06, "loss": 0.4907, "step": 4669 }, { "epoch": 1.5191932335718934, "grad_norm": 1.761813789740847, "learning_rate": 5.760759833231363e-06, "loss": 0.3934, "step": 4670 }, { "epoch": 1.5195185426154847, "grad_norm": 2.112165411827879, "learning_rate": 5.7588890621512915e-06, "loss": 0.4751, "step": 4671 }, { "epoch": 1.519843851659076, "grad_norm": 2.012464317618185, "learning_rate": 5.757018182321909e-06, "loss": 0.3834, "step": 4672 }, { "epoch": 1.5201691607026675, "grad_norm": 2.0824075315681023, "learning_rate": 5.755147194011319e-06, "loss": 0.4398, "step": 4673 }, { "epoch": 1.520494469746259, "grad_norm": 1.807632018011583, "learning_rate": 5.753276097487632e-06, "loss": 0.4032, "step": 4674 }, { "epoch": 1.5208197787898503, "grad_norm": 1.8443094714012276, "learning_rate": 5.751404893018978e-06, "loss": 0.4126, "step": 4675 }, { "epoch": 1.5211450878334418, "grad_norm": 2.06104755097073, "learning_rate": 5.749533580873501e-06, "loss": 0.3923, "step": 4676 }, { "epoch": 1.5214703968770333, "grad_norm": 2.128830116144821, "learning_rate": 5.747662161319364e-06, "loss": 0.4387, "step": 4677 }, { "epoch": 1.5217957059206246, "grad_norm": 1.7626642438229903, "learning_rate": 5.745790634624738e-06, "loss": 0.4173, "step": 4678 }, { "epoch": 1.522121014964216, "grad_norm": 1.9059614582239872, "learning_rate": 5.7439190010578184e-06, "loss": 0.3534, "step": 4679 }, { "epoch": 1.5224463240078074, "grad_norm": 2.1812885513402898, "learning_rate": 5.742047260886811e-06, "loss": 0.4352, "step": 4680 }, { "epoch": 1.522771633051399, "grad_norm": 1.7279240151513247, "learning_rate": 5.740175414379935e-06, "loss": 0.326, "step": 4681 }, { "epoch": 1.5230969420949902, "grad_norm": 1.7176367031140773, "learning_rate": 5.738303461805428e-06, "loss": 0.4326, "step": 4682 }, { "epoch": 1.5234222511385815, "grad_norm": 1.9081908045343916, "learning_rate": 5.7364314034315425e-06, "loss": 0.3296, "step": 4683 }, { "epoch": 1.523747560182173, "grad_norm": 1.696645922617255, "learning_rate": 5.734559239526548e-06, "loss": 0.3363, "step": 4684 }, { "epoch": 1.5240728692257646, "grad_norm": 2.1344830763793703, "learning_rate": 5.732686970358723e-06, "loss": 0.5108, "step": 4685 }, { "epoch": 1.5243981782693559, "grad_norm": 1.79377028662924, "learning_rate": 5.730814596196367e-06, "loss": 0.3785, "step": 4686 }, { "epoch": 1.5247234873129472, "grad_norm": 1.7867948165322094, "learning_rate": 5.728942117307793e-06, "loss": 0.3896, "step": 4687 }, { "epoch": 1.5250487963565387, "grad_norm": 1.9313815202423494, "learning_rate": 5.727069533961327e-06, "loss": 0.3847, "step": 4688 }, { "epoch": 1.5253741054001302, "grad_norm": 2.0881774683668364, "learning_rate": 5.725196846425312e-06, "loss": 0.4341, "step": 4689 }, { "epoch": 1.5256994144437215, "grad_norm": 1.8804807099279839, "learning_rate": 5.7233240549681046e-06, "loss": 0.3459, "step": 4690 }, { "epoch": 1.526024723487313, "grad_norm": 1.6935039714194657, "learning_rate": 5.721451159858079e-06, "loss": 0.3469, "step": 4691 }, { "epoch": 1.5263500325309045, "grad_norm": 1.849795620552769, "learning_rate": 5.71957816136362e-06, "loss": 0.3124, "step": 4692 }, { "epoch": 1.5266753415744958, "grad_norm": 1.9510159412276082, "learning_rate": 5.717705059753131e-06, "loss": 0.424, "step": 4693 }, { "epoch": 1.527000650618087, "grad_norm": 2.1870742400866394, "learning_rate": 5.71583185529503e-06, "loss": 0.518, "step": 4694 }, { "epoch": 1.5273259596616786, "grad_norm": 2.196997746078138, "learning_rate": 5.713958548257745e-06, "loss": 0.3719, "step": 4695 }, { "epoch": 1.5276512687052701, "grad_norm": 1.8151059428976868, "learning_rate": 5.712085138909726e-06, "loss": 0.3617, "step": 4696 }, { "epoch": 1.5279765777488614, "grad_norm": 1.8334046213962971, "learning_rate": 5.710211627519431e-06, "loss": 0.4146, "step": 4697 }, { "epoch": 1.5283018867924527, "grad_norm": 1.869364416888324, "learning_rate": 5.708338014355335e-06, "loss": 0.3672, "step": 4698 }, { "epoch": 1.5286271958360442, "grad_norm": 1.7258263319417875, "learning_rate": 5.7064642996859295e-06, "loss": 0.3184, "step": 4699 }, { "epoch": 1.5289525048796357, "grad_norm": 2.200729379053643, "learning_rate": 5.704590483779721e-06, "loss": 0.4811, "step": 4700 }, { "epoch": 1.529277813923227, "grad_norm": 1.825157281823837, "learning_rate": 5.702716566905224e-06, "loss": 0.3447, "step": 4701 }, { "epoch": 1.5296031229668183, "grad_norm": 2.06938210808675, "learning_rate": 5.7008425493309745e-06, "loss": 0.3989, "step": 4702 }, { "epoch": 1.5299284320104098, "grad_norm": 2.0071442699208464, "learning_rate": 5.698968431325522e-06, "loss": 0.3906, "step": 4703 }, { "epoch": 1.5302537410540014, "grad_norm": 2.178843896354467, "learning_rate": 5.697094213157427e-06, "loss": 0.3985, "step": 4704 }, { "epoch": 1.5305790500975927, "grad_norm": 1.820187121497636, "learning_rate": 5.695219895095265e-06, "loss": 0.3502, "step": 4705 }, { "epoch": 1.5309043591411842, "grad_norm": 1.9175993564905347, "learning_rate": 5.69334547740763e-06, "loss": 0.413, "step": 4706 }, { "epoch": 1.5312296681847757, "grad_norm": 1.8064866468272085, "learning_rate": 5.691470960363126e-06, "loss": 0.3621, "step": 4707 }, { "epoch": 1.531554977228367, "grad_norm": 1.953004776879151, "learning_rate": 5.68959634423037e-06, "loss": 0.4452, "step": 4708 }, { "epoch": 1.5318802862719583, "grad_norm": 1.9986306610066376, "learning_rate": 5.687721629278e-06, "loss": 0.3982, "step": 4709 }, { "epoch": 1.5322055953155498, "grad_norm": 1.8251744010486297, "learning_rate": 5.685846815774662e-06, "loss": 0.3783, "step": 4710 }, { "epoch": 1.5325309043591413, "grad_norm": 1.8862215687225556, "learning_rate": 5.683971903989018e-06, "loss": 0.3203, "step": 4711 }, { "epoch": 1.5328562134027326, "grad_norm": 1.6756558378435977, "learning_rate": 5.682096894189744e-06, "loss": 0.3834, "step": 4712 }, { "epoch": 1.5331815224463239, "grad_norm": 1.7611198315602101, "learning_rate": 5.680221786645531e-06, "loss": 0.3044, "step": 4713 }, { "epoch": 1.5335068314899154, "grad_norm": 1.7992788443223047, "learning_rate": 5.6783465816250805e-06, "loss": 0.3327, "step": 4714 }, { "epoch": 1.533832140533507, "grad_norm": 1.9669705240780446, "learning_rate": 5.676471279397115e-06, "loss": 0.3143, "step": 4715 }, { "epoch": 1.5341574495770982, "grad_norm": 1.8932756657841945, "learning_rate": 5.674595880230363e-06, "loss": 0.4198, "step": 4716 }, { "epoch": 1.5344827586206895, "grad_norm": 2.3561046751107426, "learning_rate": 5.672720384393571e-06, "loss": 0.486, "step": 4717 }, { "epoch": 1.534808067664281, "grad_norm": 1.9879932843602626, "learning_rate": 5.670844792155499e-06, "loss": 0.4709, "step": 4718 }, { "epoch": 1.5351333767078725, "grad_norm": 1.8238324410068467, "learning_rate": 5.668969103784922e-06, "loss": 0.3407, "step": 4719 }, { "epoch": 1.5354586857514638, "grad_norm": 2.016651956543467, "learning_rate": 5.667093319550626e-06, "loss": 0.47, "step": 4720 }, { "epoch": 1.5357839947950553, "grad_norm": 1.8376855845015772, "learning_rate": 5.6652174397214136e-06, "loss": 0.3415, "step": 4721 }, { "epoch": 1.5361093038386469, "grad_norm": 2.0596605074069965, "learning_rate": 5.663341464566098e-06, "loss": 0.4359, "step": 4722 }, { "epoch": 1.5364346128822381, "grad_norm": 1.7859485484044824, "learning_rate": 5.661465394353508e-06, "loss": 0.3413, "step": 4723 }, { "epoch": 1.5367599219258294, "grad_norm": 1.8893860073265418, "learning_rate": 5.659589229352486e-06, "loss": 0.3284, "step": 4724 }, { "epoch": 1.537085230969421, "grad_norm": 2.148331209228141, "learning_rate": 5.657712969831888e-06, "loss": 0.4437, "step": 4725 }, { "epoch": 1.5374105400130125, "grad_norm": 1.9677879954019493, "learning_rate": 5.6558366160605815e-06, "loss": 0.3413, "step": 4726 }, { "epoch": 1.5377358490566038, "grad_norm": 1.7333746227140203, "learning_rate": 5.653960168307453e-06, "loss": 0.3247, "step": 4727 }, { "epoch": 1.538061158100195, "grad_norm": 2.0429260259838906, "learning_rate": 5.6520836268413936e-06, "loss": 0.4234, "step": 4728 }, { "epoch": 1.5383864671437866, "grad_norm": 1.933098924831115, "learning_rate": 5.650206991931319e-06, "loss": 0.3409, "step": 4729 }, { "epoch": 1.538711776187378, "grad_norm": 1.935174888693084, "learning_rate": 5.648330263846146e-06, "loss": 0.4291, "step": 4730 }, { "epoch": 1.5390370852309694, "grad_norm": 2.126440145210191, "learning_rate": 5.646453442854816e-06, "loss": 0.4439, "step": 4731 }, { "epoch": 1.5393623942745607, "grad_norm": 1.684929553721669, "learning_rate": 5.6445765292262755e-06, "loss": 0.2867, "step": 4732 }, { "epoch": 1.5396877033181522, "grad_norm": 2.0226837311513903, "learning_rate": 5.642699523229489e-06, "loss": 0.4352, "step": 4733 }, { "epoch": 1.5400130123617437, "grad_norm": 1.6823704333326228, "learning_rate": 5.640822425133431e-06, "loss": 0.3229, "step": 4734 }, { "epoch": 1.540338321405335, "grad_norm": 2.0350914623071747, "learning_rate": 5.638945235207094e-06, "loss": 0.4382, "step": 4735 }, { "epoch": 1.5406636304489265, "grad_norm": 1.6040577902588538, "learning_rate": 5.637067953719477e-06, "loss": 0.397, "step": 4736 }, { "epoch": 1.540988939492518, "grad_norm": 1.9282576688744972, "learning_rate": 5.6351905809396e-06, "loss": 0.4156, "step": 4737 }, { "epoch": 1.5413142485361093, "grad_norm": 1.7627463929512197, "learning_rate": 5.633313117136487e-06, "loss": 0.4048, "step": 4738 }, { "epoch": 1.5416395575797006, "grad_norm": 2.136462570370218, "learning_rate": 5.631435562579184e-06, "loss": 0.4268, "step": 4739 }, { "epoch": 1.5419648666232921, "grad_norm": 1.801278604818353, "learning_rate": 5.629557917536742e-06, "loss": 0.3758, "step": 4740 }, { "epoch": 1.5422901756668836, "grad_norm": 1.8681678859952544, "learning_rate": 5.6276801822782315e-06, "loss": 0.3903, "step": 4741 }, { "epoch": 1.542615484710475, "grad_norm": 1.8200588728397553, "learning_rate": 5.6258023570727326e-06, "loss": 0.343, "step": 4742 }, { "epoch": 1.5429407937540662, "grad_norm": 1.9985313930665292, "learning_rate": 5.623924442189339e-06, "loss": 0.5077, "step": 4743 }, { "epoch": 1.5432661027976577, "grad_norm": 2.0700201111776955, "learning_rate": 5.622046437897155e-06, "loss": 0.53, "step": 4744 }, { "epoch": 1.5435914118412493, "grad_norm": 1.6294666935531628, "learning_rate": 5.620168344465304e-06, "loss": 0.2625, "step": 4745 }, { "epoch": 1.5439167208848406, "grad_norm": 1.808095869096039, "learning_rate": 5.618290162162915e-06, "loss": 0.4276, "step": 4746 }, { "epoch": 1.5442420299284318, "grad_norm": 2.0402093868369864, "learning_rate": 5.616411891259133e-06, "loss": 0.4777, "step": 4747 }, { "epoch": 1.5445673389720236, "grad_norm": 1.7733830945774884, "learning_rate": 5.614533532023117e-06, "loss": 0.3819, "step": 4748 }, { "epoch": 1.5448926480156149, "grad_norm": 1.9343145412223413, "learning_rate": 5.612655084724035e-06, "loss": 0.4958, "step": 4749 }, { "epoch": 1.5452179570592062, "grad_norm": 1.7342850360683977, "learning_rate": 5.610776549631071e-06, "loss": 0.3135, "step": 4750 }, { "epoch": 1.5455432661027977, "grad_norm": 1.824478182753616, "learning_rate": 5.60889792701342e-06, "loss": 0.4229, "step": 4751 }, { "epoch": 1.5458685751463892, "grad_norm": 1.9640083741409322, "learning_rate": 5.607019217140291e-06, "loss": 0.4392, "step": 4752 }, { "epoch": 1.5461938841899805, "grad_norm": 2.1644943243924897, "learning_rate": 5.605140420280904e-06, "loss": 0.4177, "step": 4753 }, { "epoch": 1.5465191932335718, "grad_norm": 1.8988384626493415, "learning_rate": 5.60326153670449e-06, "loss": 0.3184, "step": 4754 }, { "epoch": 1.5468445022771633, "grad_norm": 1.7200090588970203, "learning_rate": 5.601382566680295e-06, "loss": 0.3355, "step": 4755 }, { "epoch": 1.5471698113207548, "grad_norm": 2.091308000928067, "learning_rate": 5.599503510477579e-06, "loss": 0.4447, "step": 4756 }, { "epoch": 1.547495120364346, "grad_norm": 1.8217504142729246, "learning_rate": 5.597624368365608e-06, "loss": 0.3071, "step": 4757 }, { "epoch": 1.5478204294079374, "grad_norm": 1.835977522699692, "learning_rate": 5.59574514061367e-06, "loss": 0.3888, "step": 4758 }, { "epoch": 1.548145738451529, "grad_norm": 1.6570995777847919, "learning_rate": 5.593865827491053e-06, "loss": 0.3552, "step": 4759 }, { "epoch": 1.5484710474951204, "grad_norm": 1.9482182004934814, "learning_rate": 5.591986429267069e-06, "loss": 0.4003, "step": 4760 }, { "epoch": 1.5487963565387117, "grad_norm": 1.8810161843867126, "learning_rate": 5.590106946211034e-06, "loss": 0.3209, "step": 4761 }, { "epoch": 1.5491216655823032, "grad_norm": 2.287793990110565, "learning_rate": 5.58822737859228e-06, "loss": 0.4576, "step": 4762 }, { "epoch": 1.5494469746258948, "grad_norm": 2.09840603823355, "learning_rate": 5.58634772668015e-06, "loss": 0.44, "step": 4763 }, { "epoch": 1.549772283669486, "grad_norm": 2.1687045114724928, "learning_rate": 5.584467990744002e-06, "loss": 0.3635, "step": 4764 }, { "epoch": 1.5500975927130773, "grad_norm": 1.8517935773366734, "learning_rate": 5.582588171053199e-06, "loss": 0.3653, "step": 4765 }, { "epoch": 1.5504229017566689, "grad_norm": 1.8383879101552076, "learning_rate": 5.580708267877124e-06, "loss": 0.3992, "step": 4766 }, { "epoch": 1.5507482108002604, "grad_norm": 1.8804371501334733, "learning_rate": 5.578828281485167e-06, "loss": 0.4159, "step": 4767 }, { "epoch": 1.5510735198438517, "grad_norm": 1.7257491897381536, "learning_rate": 5.576948212146731e-06, "loss": 0.3501, "step": 4768 }, { "epoch": 1.551398828887443, "grad_norm": 1.4892459458386074, "learning_rate": 5.57506806013123e-06, "loss": 0.2781, "step": 4769 }, { "epoch": 1.5517241379310345, "grad_norm": 2.1519358989274506, "learning_rate": 5.573187825708094e-06, "loss": 0.4111, "step": 4770 }, { "epoch": 1.552049446974626, "grad_norm": 1.934758646670982, "learning_rate": 5.571307509146761e-06, "loss": 0.3841, "step": 4771 }, { "epoch": 1.5523747560182173, "grad_norm": 1.7590279479665438, "learning_rate": 5.569427110716681e-06, "loss": 0.4322, "step": 4772 }, { "epoch": 1.5527000650618086, "grad_norm": 2.304469066523032, "learning_rate": 5.567546630687315e-06, "loss": 0.4575, "step": 4773 }, { "epoch": 1.5530253741054, "grad_norm": 1.8760433485650894, "learning_rate": 5.565666069328139e-06, "loss": 0.316, "step": 4774 }, { "epoch": 1.5533506831489916, "grad_norm": 1.8039538072392494, "learning_rate": 5.563785426908639e-06, "loss": 0.4363, "step": 4775 }, { "epoch": 1.553675992192583, "grad_norm": 2.199041730804751, "learning_rate": 5.561904703698311e-06, "loss": 0.4285, "step": 4776 }, { "epoch": 1.5540013012361744, "grad_norm": 1.9782589102120092, "learning_rate": 5.560023899966663e-06, "loss": 0.4046, "step": 4777 }, { "epoch": 1.554326610279766, "grad_norm": 1.676557358081998, "learning_rate": 5.558143015983218e-06, "loss": 0.3061, "step": 4778 }, { "epoch": 1.5546519193233572, "grad_norm": 2.2163844796436565, "learning_rate": 5.556262052017507e-06, "loss": 0.3894, "step": 4779 }, { "epoch": 1.5549772283669485, "grad_norm": 1.875706114945357, "learning_rate": 5.554381008339073e-06, "loss": 0.3783, "step": 4780 }, { "epoch": 1.55530253741054, "grad_norm": 2.121066824176875, "learning_rate": 5.552499885217471e-06, "loss": 0.4079, "step": 4781 }, { "epoch": 1.5556278464541315, "grad_norm": 1.9773145922403392, "learning_rate": 5.550618682922266e-06, "loss": 0.3488, "step": 4782 }, { "epoch": 1.5559531554977228, "grad_norm": 2.051156541372392, "learning_rate": 5.548737401723037e-06, "loss": 0.3966, "step": 4783 }, { "epoch": 1.5562784645413141, "grad_norm": 2.004617523471978, "learning_rate": 5.546856041889374e-06, "loss": 0.488, "step": 4784 }, { "epoch": 1.5566037735849056, "grad_norm": 1.9164176505424133, "learning_rate": 5.544974603690873e-06, "loss": 0.361, "step": 4785 }, { "epoch": 1.5569290826284972, "grad_norm": 1.60507923818067, "learning_rate": 5.543093087397149e-06, "loss": 0.3107, "step": 4786 }, { "epoch": 1.5572543916720885, "grad_norm": 1.7569745529409801, "learning_rate": 5.541211493277824e-06, "loss": 0.3526, "step": 4787 }, { "epoch": 1.5575797007156797, "grad_norm": 1.6791310385483316, "learning_rate": 5.53932982160253e-06, "loss": 0.3158, "step": 4788 }, { "epoch": 1.5579050097592713, "grad_norm": 2.632626781790961, "learning_rate": 5.537448072640914e-06, "loss": 0.4259, "step": 4789 }, { "epoch": 1.5582303188028628, "grad_norm": 2.2703477983316325, "learning_rate": 5.535566246662631e-06, "loss": 0.4495, "step": 4790 }, { "epoch": 1.558555627846454, "grad_norm": 1.9432972771024628, "learning_rate": 5.5336843439373475e-06, "loss": 0.4288, "step": 4791 }, { "epoch": 1.5588809368900456, "grad_norm": 2.1441685742393597, "learning_rate": 5.531802364734742e-06, "loss": 0.416, "step": 4792 }, { "epoch": 1.559206245933637, "grad_norm": 2.0021758849635782, "learning_rate": 5.529920309324503e-06, "loss": 0.3713, "step": 4793 }, { "epoch": 1.5595315549772284, "grad_norm": 1.9895153574579532, "learning_rate": 5.52803817797633e-06, "loss": 0.3872, "step": 4794 }, { "epoch": 1.5598568640208197, "grad_norm": 2.032196476956217, "learning_rate": 5.526155970959933e-06, "loss": 0.3957, "step": 4795 }, { "epoch": 1.5601821730644112, "grad_norm": 1.7201718837462396, "learning_rate": 5.524273688545037e-06, "loss": 0.3302, "step": 4796 }, { "epoch": 1.5605074821080027, "grad_norm": 2.0940195257887835, "learning_rate": 5.52239133100137e-06, "loss": 0.4323, "step": 4797 }, { "epoch": 1.560832791151594, "grad_norm": 2.112143904983999, "learning_rate": 5.520508898598678e-06, "loss": 0.4303, "step": 4798 }, { "epoch": 1.5611581001951853, "grad_norm": 1.940741508976379, "learning_rate": 5.518626391606712e-06, "loss": 0.3651, "step": 4799 }, { "epoch": 1.5614834092387768, "grad_norm": 1.804572181147511, "learning_rate": 5.5167438102952405e-06, "loss": 0.3309, "step": 4800 }, { "epoch": 1.5618087182823683, "grad_norm": 1.890668815748716, "learning_rate": 5.514861154934035e-06, "loss": 0.3767, "step": 4801 }, { "epoch": 1.5621340273259596, "grad_norm": 1.8820031204050212, "learning_rate": 5.512978425792883e-06, "loss": 0.3605, "step": 4802 }, { "epoch": 1.562459336369551, "grad_norm": 1.8144271957118445, "learning_rate": 5.511095623141581e-06, "loss": 0.3224, "step": 4803 }, { "epoch": 1.5627846454131424, "grad_norm": 1.8405429013759322, "learning_rate": 5.509212747249936e-06, "loss": 0.397, "step": 4804 }, { "epoch": 1.563109954456734, "grad_norm": 1.8415511838729928, "learning_rate": 5.507329798387764e-06, "loss": 0.3251, "step": 4805 }, { "epoch": 1.5634352635003252, "grad_norm": 1.682172498358812, "learning_rate": 5.505446776824894e-06, "loss": 0.3417, "step": 4806 }, { "epoch": 1.5637605725439168, "grad_norm": 1.6185655290866279, "learning_rate": 5.503563682831163e-06, "loss": 0.3375, "step": 4807 }, { "epoch": 1.5640858815875083, "grad_norm": 1.9578211308239288, "learning_rate": 5.501680516676422e-06, "loss": 0.3586, "step": 4808 }, { "epoch": 1.5644111906310996, "grad_norm": 1.8690180760430297, "learning_rate": 5.499797278630528e-06, "loss": 0.3181, "step": 4809 }, { "epoch": 1.5647364996746909, "grad_norm": 2.0014132322610423, "learning_rate": 5.49791396896335e-06, "loss": 0.4494, "step": 4810 }, { "epoch": 1.5650618087182824, "grad_norm": 2.0294451490950616, "learning_rate": 5.4960305879447675e-06, "loss": 0.4261, "step": 4811 }, { "epoch": 1.565387117761874, "grad_norm": 1.8669559287026969, "learning_rate": 5.494147135844673e-06, "loss": 0.3911, "step": 4812 }, { "epoch": 1.5657124268054652, "grad_norm": 1.9415086405792634, "learning_rate": 5.492263612932963e-06, "loss": 0.3831, "step": 4813 }, { "epoch": 1.5660377358490565, "grad_norm": 1.7712695821639217, "learning_rate": 5.490380019479551e-06, "loss": 0.4184, "step": 4814 }, { "epoch": 1.566363044892648, "grad_norm": 2.214382605030725, "learning_rate": 5.488496355754353e-06, "loss": 0.422, "step": 4815 }, { "epoch": 1.5666883539362395, "grad_norm": 1.6623183513202275, "learning_rate": 5.486612622027303e-06, "loss": 0.4071, "step": 4816 }, { "epoch": 1.5670136629798308, "grad_norm": 1.674453307144788, "learning_rate": 5.484728818568339e-06, "loss": 0.2725, "step": 4817 }, { "epoch": 1.567338972023422, "grad_norm": 1.7596372162343046, "learning_rate": 5.482844945647412e-06, "loss": 0.3758, "step": 4818 }, { "epoch": 1.5676642810670136, "grad_norm": 1.9347338183845328, "learning_rate": 5.480961003534483e-06, "loss": 0.3308, "step": 4819 }, { "epoch": 1.5679895901106051, "grad_norm": 2.106410567183768, "learning_rate": 5.479076992499523e-06, "loss": 0.3852, "step": 4820 }, { "epoch": 1.5683148991541964, "grad_norm": 2.381222088528441, "learning_rate": 5.4771929128125066e-06, "loss": 0.4302, "step": 4821 }, { "epoch": 1.568640208197788, "grad_norm": 1.9942011665610695, "learning_rate": 5.47530876474343e-06, "loss": 0.4085, "step": 4822 }, { "epoch": 1.5689655172413794, "grad_norm": 2.3191081020278714, "learning_rate": 5.473424548562291e-06, "loss": 0.4856, "step": 4823 }, { "epoch": 1.5692908262849707, "grad_norm": 1.7894558236948244, "learning_rate": 5.471540264539098e-06, "loss": 0.3498, "step": 4824 }, { "epoch": 1.569616135328562, "grad_norm": 1.6683738086366517, "learning_rate": 5.4696559129438705e-06, "loss": 0.326, "step": 4825 }, { "epoch": 1.5699414443721535, "grad_norm": 2.212500899619917, "learning_rate": 5.467771494046639e-06, "loss": 0.4118, "step": 4826 }, { "epoch": 1.570266753415745, "grad_norm": 1.5645657478670467, "learning_rate": 5.465887008117438e-06, "loss": 0.3275, "step": 4827 }, { "epoch": 1.5705920624593364, "grad_norm": 2.0637785344697788, "learning_rate": 5.464002455426319e-06, "loss": 0.412, "step": 4828 }, { "epoch": 1.5709173715029277, "grad_norm": 1.9059663290700677, "learning_rate": 5.46211783624334e-06, "loss": 0.3376, "step": 4829 }, { "epoch": 1.5712426805465192, "grad_norm": 2.22836862556621, "learning_rate": 5.460233150838566e-06, "loss": 0.5099, "step": 4830 }, { "epoch": 1.5715679895901107, "grad_norm": 1.8511573319848633, "learning_rate": 5.458348399482074e-06, "loss": 0.3851, "step": 4831 }, { "epoch": 1.571893298633702, "grad_norm": 1.7516411091145356, "learning_rate": 5.456463582443952e-06, "loss": 0.3527, "step": 4832 }, { "epoch": 1.5722186076772933, "grad_norm": 2.0297344736725096, "learning_rate": 5.454578699994294e-06, "loss": 0.373, "step": 4833 }, { "epoch": 1.572543916720885, "grad_norm": 1.7368117426416154, "learning_rate": 5.452693752403204e-06, "loss": 0.3546, "step": 4834 }, { "epoch": 1.5728692257644763, "grad_norm": 1.8613837140943208, "learning_rate": 5.4508087399407995e-06, "loss": 0.4108, "step": 4835 }, { "epoch": 1.5731945348080676, "grad_norm": 1.9648009614746902, "learning_rate": 5.448923662877201e-06, "loss": 0.422, "step": 4836 }, { "epoch": 1.573519843851659, "grad_norm": 1.9921794565010746, "learning_rate": 5.447038521482542e-06, "loss": 0.4445, "step": 4837 }, { "epoch": 1.5738451528952506, "grad_norm": 1.7346053159629786, "learning_rate": 5.445153316026964e-06, "loss": 0.3355, "step": 4838 }, { "epoch": 1.574170461938842, "grad_norm": 1.9244990156831032, "learning_rate": 5.443268046780621e-06, "loss": 0.3574, "step": 4839 }, { "epoch": 1.5744957709824332, "grad_norm": 1.8951865820287368, "learning_rate": 5.441382714013671e-06, "loss": 0.4366, "step": 4840 }, { "epoch": 1.5748210800260247, "grad_norm": 1.6143926499997678, "learning_rate": 5.4394973179962815e-06, "loss": 0.3371, "step": 4841 }, { "epoch": 1.5751463890696162, "grad_norm": 1.726547337673839, "learning_rate": 5.437611858998637e-06, "loss": 0.3464, "step": 4842 }, { "epoch": 1.5754716981132075, "grad_norm": 1.9055518612829507, "learning_rate": 5.435726337290919e-06, "loss": 0.4506, "step": 4843 }, { "epoch": 1.5757970071567988, "grad_norm": 1.8194417111507444, "learning_rate": 5.433840753143326e-06, "loss": 0.4321, "step": 4844 }, { "epoch": 1.5761223162003903, "grad_norm": 1.7189120741723236, "learning_rate": 5.431955106826065e-06, "loss": 0.39, "step": 4845 }, { "epoch": 1.5764476252439819, "grad_norm": 1.8204293217882777, "learning_rate": 5.430069398609349e-06, "loss": 0.3834, "step": 4846 }, { "epoch": 1.5767729342875731, "grad_norm": 1.9722327005216589, "learning_rate": 5.428183628763401e-06, "loss": 0.4062, "step": 4847 }, { "epoch": 1.5770982433311644, "grad_norm": 1.7749266988467458, "learning_rate": 5.426297797558454e-06, "loss": 0.3322, "step": 4848 }, { "epoch": 1.5774235523747562, "grad_norm": 1.920193283850729, "learning_rate": 5.424411905264747e-06, "loss": 0.4015, "step": 4849 }, { "epoch": 1.5777488614183475, "grad_norm": 1.931698113363212, "learning_rate": 5.422525952152533e-06, "loss": 0.379, "step": 4850 }, { "epoch": 1.5780741704619388, "grad_norm": 1.7734256555448316, "learning_rate": 5.420639938492067e-06, "loss": 0.3138, "step": 4851 }, { "epoch": 1.5783994795055303, "grad_norm": 1.9192667637650933, "learning_rate": 5.418753864553618e-06, "loss": 0.3826, "step": 4852 }, { "epoch": 1.5787247885491218, "grad_norm": 1.743229920861452, "learning_rate": 5.41686773060746e-06, "loss": 0.3791, "step": 4853 }, { "epoch": 1.579050097592713, "grad_norm": 1.834428260957788, "learning_rate": 5.41498153692388e-06, "loss": 0.3905, "step": 4854 }, { "epoch": 1.5793754066363044, "grad_norm": 2.1317109167829447, "learning_rate": 5.413095283773167e-06, "loss": 0.4583, "step": 4855 }, { "epoch": 1.579700715679896, "grad_norm": 1.9544896507366742, "learning_rate": 5.411208971425627e-06, "loss": 0.4271, "step": 4856 }, { "epoch": 1.5800260247234874, "grad_norm": 1.5669395736732683, "learning_rate": 5.4093226001515645e-06, "loss": 0.3301, "step": 4857 }, { "epoch": 1.5803513337670787, "grad_norm": 1.9794444470277628, "learning_rate": 5.407436170221303e-06, "loss": 0.3503, "step": 4858 }, { "epoch": 1.58067664281067, "grad_norm": 1.6241876938744193, "learning_rate": 5.405549681905164e-06, "loss": 0.3226, "step": 4859 }, { "epoch": 1.5810019518542615, "grad_norm": 2.104691313404, "learning_rate": 5.403663135473487e-06, "loss": 0.4626, "step": 4860 }, { "epoch": 1.581327260897853, "grad_norm": 1.8691582495787107, "learning_rate": 5.401776531196614e-06, "loss": 0.4036, "step": 4861 }, { "epoch": 1.5816525699414443, "grad_norm": 2.1004231736656322, "learning_rate": 5.399889869344895e-06, "loss": 0.424, "step": 4862 }, { "epoch": 1.5819778789850358, "grad_norm": 1.8685999132298157, "learning_rate": 5.398003150188691e-06, "loss": 0.3882, "step": 4863 }, { "epoch": 1.5823031880286273, "grad_norm": 1.902161195123946, "learning_rate": 5.396116373998371e-06, "loss": 0.3885, "step": 4864 }, { "epoch": 1.5826284970722186, "grad_norm": 2.015882713350689, "learning_rate": 5.39422954104431e-06, "loss": 0.3766, "step": 4865 }, { "epoch": 1.58295380611581, "grad_norm": 2.154295952681017, "learning_rate": 5.392342651596894e-06, "loss": 0.4524, "step": 4866 }, { "epoch": 1.5832791151594015, "grad_norm": 2.104871643255363, "learning_rate": 5.390455705926513e-06, "loss": 0.3904, "step": 4867 }, { "epoch": 1.583604424202993, "grad_norm": 1.873710957005527, "learning_rate": 5.38856870430357e-06, "loss": 0.3641, "step": 4868 }, { "epoch": 1.5839297332465843, "grad_norm": 2.0930144640648622, "learning_rate": 5.386681646998472e-06, "loss": 0.3746, "step": 4869 }, { "epoch": 1.5842550422901756, "grad_norm": 1.8108333053775238, "learning_rate": 5.384794534281636e-06, "loss": 0.4238, "step": 4870 }, { "epoch": 1.584580351333767, "grad_norm": 1.4709400351708115, "learning_rate": 5.382907366423487e-06, "loss": 0.2809, "step": 4871 }, { "epoch": 1.5849056603773586, "grad_norm": 1.737912865220009, "learning_rate": 5.3810201436944575e-06, "loss": 0.3918, "step": 4872 }, { "epoch": 1.5852309694209499, "grad_norm": 2.1091733283694554, "learning_rate": 5.379132866364987e-06, "loss": 0.4671, "step": 4873 }, { "epoch": 1.5855562784645412, "grad_norm": 1.906037817710211, "learning_rate": 5.377245534705525e-06, "loss": 0.3843, "step": 4874 }, { "epoch": 1.5858815875081327, "grad_norm": 1.898175017511401, "learning_rate": 5.3753581489865245e-06, "loss": 0.459, "step": 4875 }, { "epoch": 1.5862068965517242, "grad_norm": 2.046515145135511, "learning_rate": 5.373470709478452e-06, "loss": 0.4204, "step": 4876 }, { "epoch": 1.5865322055953155, "grad_norm": 1.75627406451247, "learning_rate": 5.371583216451778e-06, "loss": 0.3463, "step": 4877 }, { "epoch": 1.586857514638907, "grad_norm": 1.8394718661343559, "learning_rate": 5.369695670176981e-06, "loss": 0.4193, "step": 4878 }, { "epoch": 1.5871828236824985, "grad_norm": 1.8852393997308867, "learning_rate": 5.367808070924547e-06, "loss": 0.3947, "step": 4879 }, { "epoch": 1.5875081327260898, "grad_norm": 1.8030512410780637, "learning_rate": 5.365920418964973e-06, "loss": 0.3635, "step": 4880 }, { "epoch": 1.587833441769681, "grad_norm": 1.9570708188179973, "learning_rate": 5.364032714568758e-06, "loss": 0.4058, "step": 4881 }, { "epoch": 1.5881587508132726, "grad_norm": 2.0371097568291567, "learning_rate": 5.362144958006411e-06, "loss": 0.347, "step": 4882 }, { "epoch": 1.5884840598568641, "grad_norm": 1.9060167546032347, "learning_rate": 5.360257149548453e-06, "loss": 0.4156, "step": 4883 }, { "epoch": 1.5888093689004554, "grad_norm": 2.1624212874310804, "learning_rate": 5.358369289465402e-06, "loss": 0.4743, "step": 4884 }, { "epoch": 1.5891346779440467, "grad_norm": 1.764198505320464, "learning_rate": 5.356481378027796e-06, "loss": 0.3102, "step": 4885 }, { "epoch": 1.5894599869876382, "grad_norm": 1.982919253925493, "learning_rate": 5.354593415506168e-06, "loss": 0.3328, "step": 4886 }, { "epoch": 1.5897852960312298, "grad_norm": 2.0187403152447962, "learning_rate": 5.352705402171071e-06, "loss": 0.389, "step": 4887 }, { "epoch": 1.590110605074821, "grad_norm": 2.286634629345919, "learning_rate": 5.35081733829305e-06, "loss": 0.4434, "step": 4888 }, { "epoch": 1.5904359141184123, "grad_norm": 2.194289142351243, "learning_rate": 5.348929224142674e-06, "loss": 0.4791, "step": 4889 }, { "epoch": 1.5907612231620039, "grad_norm": 1.8630822735129873, "learning_rate": 5.3470410599905065e-06, "loss": 0.358, "step": 4890 }, { "epoch": 1.5910865322055954, "grad_norm": 2.125836931718017, "learning_rate": 5.345152846107125e-06, "loss": 0.4041, "step": 4891 }, { "epoch": 1.5914118412491867, "grad_norm": 1.9415175389804695, "learning_rate": 5.343264582763108e-06, "loss": 0.4721, "step": 4892 }, { "epoch": 1.5917371502927782, "grad_norm": 2.2709609609707004, "learning_rate": 5.3413762702290495e-06, "loss": 0.3889, "step": 4893 }, { "epoch": 1.5920624593363697, "grad_norm": 1.6782178537944807, "learning_rate": 5.3394879087755426e-06, "loss": 0.3231, "step": 4894 }, { "epoch": 1.592387768379961, "grad_norm": 1.9343917868990763, "learning_rate": 5.337599498673193e-06, "loss": 0.4171, "step": 4895 }, { "epoch": 1.5927130774235523, "grad_norm": 1.8687000692329117, "learning_rate": 5.335711040192609e-06, "loss": 0.3698, "step": 4896 }, { "epoch": 1.5930383864671438, "grad_norm": 1.741303477604468, "learning_rate": 5.3338225336044105e-06, "loss": 0.371, "step": 4897 }, { "epoch": 1.5933636955107353, "grad_norm": 1.7671509512665995, "learning_rate": 5.331933979179218e-06, "loss": 0.3985, "step": 4898 }, { "epoch": 1.5936890045543266, "grad_norm": 2.1701565569832306, "learning_rate": 5.330045377187667e-06, "loss": 0.4375, "step": 4899 }, { "epoch": 1.594014313597918, "grad_norm": 1.8645153734828495, "learning_rate": 5.328156727900393e-06, "loss": 0.3654, "step": 4900 }, { "epoch": 1.5943396226415094, "grad_norm": 1.7853358983067777, "learning_rate": 5.326268031588041e-06, "loss": 0.2705, "step": 4901 }, { "epoch": 1.594664931685101, "grad_norm": 2.2462918527670497, "learning_rate": 5.324379288521262e-06, "loss": 0.521, "step": 4902 }, { "epoch": 1.5949902407286922, "grad_norm": 1.7549567290325865, "learning_rate": 5.322490498970716e-06, "loss": 0.3503, "step": 4903 }, { "epoch": 1.5953155497722835, "grad_norm": 1.8058776068904827, "learning_rate": 5.320601663207066e-06, "loss": 0.4883, "step": 4904 }, { "epoch": 1.595640858815875, "grad_norm": 2.037715443150099, "learning_rate": 5.318712781500983e-06, "loss": 0.4211, "step": 4905 }, { "epoch": 1.5959661678594665, "grad_norm": 1.7069257608522834, "learning_rate": 5.316823854123147e-06, "loss": 0.3509, "step": 4906 }, { "epoch": 1.5962914769030578, "grad_norm": 1.9052613958377933, "learning_rate": 5.314934881344241e-06, "loss": 0.385, "step": 4907 }, { "epoch": 1.5966167859466494, "grad_norm": 1.779143963749479, "learning_rate": 5.3130458634349555e-06, "loss": 0.2801, "step": 4908 }, { "epoch": 1.5969420949902409, "grad_norm": 2.1941706161877677, "learning_rate": 5.311156800665991e-06, "loss": 0.4627, "step": 4909 }, { "epoch": 1.5972674040338322, "grad_norm": 1.9279528133723733, "learning_rate": 5.309267693308049e-06, "loss": 0.39, "step": 4910 }, { "epoch": 1.5975927130774235, "grad_norm": 1.8437723099200878, "learning_rate": 5.307378541631839e-06, "loss": 0.3701, "step": 4911 }, { "epoch": 1.597918022121015, "grad_norm": 1.9103406314476064, "learning_rate": 5.30548934590808e-06, "loss": 0.3951, "step": 4912 }, { "epoch": 1.5982433311646065, "grad_norm": 2.0300816167911946, "learning_rate": 5.3036001064074945e-06, "loss": 0.4172, "step": 4913 }, { "epoch": 1.5985686402081978, "grad_norm": 1.7172950310213622, "learning_rate": 5.3017108234008105e-06, "loss": 0.3277, "step": 4914 }, { "epoch": 1.598893949251789, "grad_norm": 1.910009929830712, "learning_rate": 5.2998214971587645e-06, "loss": 0.4149, "step": 4915 }, { "epoch": 1.5992192582953806, "grad_norm": 1.9163491253173501, "learning_rate": 5.297932127952099e-06, "loss": 0.3259, "step": 4916 }, { "epoch": 1.599544567338972, "grad_norm": 2.054996821076831, "learning_rate": 5.296042716051561e-06, "loss": 0.4532, "step": 4917 }, { "epoch": 1.5998698763825634, "grad_norm": 2.0587484020132583, "learning_rate": 5.294153261727903e-06, "loss": 0.4591, "step": 4918 }, { "epoch": 1.6001951854261547, "grad_norm": 1.8372499471233774, "learning_rate": 5.292263765251887e-06, "loss": 0.3302, "step": 4919 }, { "epoch": 1.6005204944697464, "grad_norm": 1.834044731266551, "learning_rate": 5.290374226894278e-06, "loss": 0.3835, "step": 4920 }, { "epoch": 1.6008458035133377, "grad_norm": 1.8598863456922141, "learning_rate": 5.28848464692585e-06, "loss": 0.362, "step": 4921 }, { "epoch": 1.601171112556929, "grad_norm": 2.190969794970417, "learning_rate": 5.286595025617379e-06, "loss": 0.5144, "step": 4922 }, { "epoch": 1.6014964216005205, "grad_norm": 1.6642939734141422, "learning_rate": 5.284705363239651e-06, "loss": 0.3184, "step": 4923 }, { "epoch": 1.601821730644112, "grad_norm": 1.820838471748514, "learning_rate": 5.282815660063451e-06, "loss": 0.3921, "step": 4924 }, { "epoch": 1.6021470396877033, "grad_norm": 1.8061868489920672, "learning_rate": 5.280925916359582e-06, "loss": 0.3513, "step": 4925 }, { "epoch": 1.6024723487312946, "grad_norm": 2.444083165145758, "learning_rate": 5.2790361323988415e-06, "loss": 0.5083, "step": 4926 }, { "epoch": 1.6027976577748861, "grad_norm": 1.9010329530266887, "learning_rate": 5.277146308452036e-06, "loss": 0.4015, "step": 4927 }, { "epoch": 1.6031229668184777, "grad_norm": 2.2786798075568937, "learning_rate": 5.275256444789979e-06, "loss": 0.4769, "step": 4928 }, { "epoch": 1.603448275862069, "grad_norm": 1.9263388684451246, "learning_rate": 5.273366541683492e-06, "loss": 0.3691, "step": 4929 }, { "epoch": 1.6037735849056602, "grad_norm": 1.767136212823214, "learning_rate": 5.271476599403397e-06, "loss": 0.364, "step": 4930 }, { "epoch": 1.6040988939492518, "grad_norm": 2.0319673298328107, "learning_rate": 5.269586618220523e-06, "loss": 0.3765, "step": 4931 }, { "epoch": 1.6044242029928433, "grad_norm": 1.814998460116486, "learning_rate": 5.267696598405709e-06, "loss": 0.4006, "step": 4932 }, { "epoch": 1.6047495120364346, "grad_norm": 1.6148114605275987, "learning_rate": 5.265806540229792e-06, "loss": 0.3518, "step": 4933 }, { "epoch": 1.6050748210800259, "grad_norm": 1.9067982230271707, "learning_rate": 5.26391644396362e-06, "loss": 0.4054, "step": 4934 }, { "epoch": 1.6054001301236176, "grad_norm": 2.047694919398706, "learning_rate": 5.262026309878048e-06, "loss": 0.4402, "step": 4935 }, { "epoch": 1.605725439167209, "grad_norm": 2.0194404251794267, "learning_rate": 5.260136138243929e-06, "loss": 0.3753, "step": 4936 }, { "epoch": 1.6060507482108002, "grad_norm": 1.9443005441272674, "learning_rate": 5.258245929332128e-06, "loss": 0.3842, "step": 4937 }, { "epoch": 1.6063760572543917, "grad_norm": 1.6994090847692118, "learning_rate": 5.256355683413515e-06, "loss": 0.3139, "step": 4938 }, { "epoch": 1.6067013662979832, "grad_norm": 1.8142081569892512, "learning_rate": 5.25446540075896e-06, "loss": 0.4212, "step": 4939 }, { "epoch": 1.6070266753415745, "grad_norm": 2.143528319521485, "learning_rate": 5.252575081639344e-06, "loss": 0.423, "step": 4940 }, { "epoch": 1.6073519843851658, "grad_norm": 2.4768170256265623, "learning_rate": 5.25068472632555e-06, "loss": 0.44, "step": 4941 }, { "epoch": 1.6076772934287573, "grad_norm": 1.7891607855251965, "learning_rate": 5.248794335088468e-06, "loss": 0.4012, "step": 4942 }, { "epoch": 1.6080026024723488, "grad_norm": 2.0582011287780713, "learning_rate": 5.2469039081989925e-06, "loss": 0.3379, "step": 4943 }, { "epoch": 1.6083279115159401, "grad_norm": 2.040328876487194, "learning_rate": 5.245013445928021e-06, "loss": 0.3772, "step": 4944 }, { "epoch": 1.6086532205595314, "grad_norm": 1.873251746183112, "learning_rate": 5.24312294854646e-06, "loss": 0.3822, "step": 4945 }, { "epoch": 1.608978529603123, "grad_norm": 1.919795130481349, "learning_rate": 5.241232416325218e-06, "loss": 0.4127, "step": 4946 }, { "epoch": 1.6093038386467144, "grad_norm": 1.7562642642768795, "learning_rate": 5.239341849535211e-06, "loss": 0.3293, "step": 4947 }, { "epoch": 1.6096291476903057, "grad_norm": 1.9941886306796242, "learning_rate": 5.2374512484473575e-06, "loss": 0.3917, "step": 4948 }, { "epoch": 1.6099544567338973, "grad_norm": 1.9532276292786461, "learning_rate": 5.235560613332581e-06, "loss": 0.437, "step": 4949 }, { "epoch": 1.6102797657774888, "grad_norm": 1.8871380930156632, "learning_rate": 5.23366994446181e-06, "loss": 0.4122, "step": 4950 }, { "epoch": 1.61060507482108, "grad_norm": 1.6820505569142101, "learning_rate": 5.231779242105982e-06, "loss": 0.389, "step": 4951 }, { "epoch": 1.6109303838646714, "grad_norm": 1.8900813987649752, "learning_rate": 5.229888506536034e-06, "loss": 0.3628, "step": 4952 }, { "epoch": 1.6112556929082629, "grad_norm": 2.006147332070203, "learning_rate": 5.2279977380229095e-06, "loss": 0.4517, "step": 4953 }, { "epoch": 1.6115810019518544, "grad_norm": 1.7704249199608952, "learning_rate": 5.226106936837558e-06, "loss": 0.3202, "step": 4954 }, { "epoch": 1.6119063109954457, "grad_norm": 1.954188820806164, "learning_rate": 5.2242161032509305e-06, "loss": 0.447, "step": 4955 }, { "epoch": 1.612231620039037, "grad_norm": 2.303571375966424, "learning_rate": 5.2223252375339864e-06, "loss": 0.4492, "step": 4956 }, { "epoch": 1.6125569290826285, "grad_norm": 1.8732152554247743, "learning_rate": 5.220434339957687e-06, "loss": 0.3961, "step": 4957 }, { "epoch": 1.61288223812622, "grad_norm": 2.097536142505791, "learning_rate": 5.218543410793e-06, "loss": 0.4677, "step": 4958 }, { "epoch": 1.6132075471698113, "grad_norm": 1.4714382509242825, "learning_rate": 5.216652450310897e-06, "loss": 0.3546, "step": 4959 }, { "epoch": 1.6135328562134026, "grad_norm": 1.6504722249495947, "learning_rate": 5.214761458782353e-06, "loss": 0.3585, "step": 4960 }, { "epoch": 1.613858165256994, "grad_norm": 2.0950809351018274, "learning_rate": 5.212870436478349e-06, "loss": 0.3757, "step": 4961 }, { "epoch": 1.6141834743005856, "grad_norm": 2.3296338382566093, "learning_rate": 5.210979383669868e-06, "loss": 0.429, "step": 4962 }, { "epoch": 1.614508783344177, "grad_norm": 1.709185100540275, "learning_rate": 5.209088300627902e-06, "loss": 0.3148, "step": 4963 }, { "epoch": 1.6148340923877684, "grad_norm": 1.825848683150527, "learning_rate": 5.207197187623444e-06, "loss": 0.3584, "step": 4964 }, { "epoch": 1.61515940143136, "grad_norm": 2.1474613302064673, "learning_rate": 5.20530604492749e-06, "loss": 0.4594, "step": 4965 }, { "epoch": 1.6154847104749512, "grad_norm": 1.8616856183977355, "learning_rate": 5.203414872811042e-06, "loss": 0.3636, "step": 4966 }, { "epoch": 1.6158100195185425, "grad_norm": 2.0972514050872983, "learning_rate": 5.201523671545109e-06, "loss": 0.4084, "step": 4967 }, { "epoch": 1.616135328562134, "grad_norm": 2.013404211535329, "learning_rate": 5.199632441400698e-06, "loss": 0.4508, "step": 4968 }, { "epoch": 1.6164606376057256, "grad_norm": 2.055656023895125, "learning_rate": 5.1977411826488254e-06, "loss": 0.3786, "step": 4969 }, { "epoch": 1.6167859466493169, "grad_norm": 1.9958840980506634, "learning_rate": 5.1958498955605085e-06, "loss": 0.4507, "step": 4970 }, { "epoch": 1.6171112556929081, "grad_norm": 2.0606691809361846, "learning_rate": 5.193958580406774e-06, "loss": 0.4264, "step": 4971 }, { "epoch": 1.6174365647364997, "grad_norm": 2.0179244649042682, "learning_rate": 5.192067237458642e-06, "loss": 0.402, "step": 4972 }, { "epoch": 1.6177618737800912, "grad_norm": 1.744312054779638, "learning_rate": 5.190175866987149e-06, "loss": 0.3504, "step": 4973 }, { "epoch": 1.6180871828236825, "grad_norm": 1.7454960489546436, "learning_rate": 5.1882844692633275e-06, "loss": 0.3469, "step": 4974 }, { "epoch": 1.6184124918672738, "grad_norm": 1.7703011050892172, "learning_rate": 5.186393044558216e-06, "loss": 0.3623, "step": 4975 }, { "epoch": 1.6187378009108653, "grad_norm": 2.0310284138915953, "learning_rate": 5.184501593142856e-06, "loss": 0.3559, "step": 4976 }, { "epoch": 1.6190631099544568, "grad_norm": 1.8781209092271918, "learning_rate": 5.182610115288296e-06, "loss": 0.3796, "step": 4977 }, { "epoch": 1.619388418998048, "grad_norm": 1.822339245692941, "learning_rate": 5.180718611265583e-06, "loss": 0.3953, "step": 4978 }, { "epoch": 1.6197137280416396, "grad_norm": 1.7974645510607423, "learning_rate": 5.178827081345773e-06, "loss": 0.4046, "step": 4979 }, { "epoch": 1.6200390370852311, "grad_norm": 2.0264020601017014, "learning_rate": 5.176935525799925e-06, "loss": 0.3624, "step": 4980 }, { "epoch": 1.6203643461288224, "grad_norm": 1.9533872920364166, "learning_rate": 5.1750439448990956e-06, "loss": 0.4467, "step": 4981 }, { "epoch": 1.6206896551724137, "grad_norm": 1.7587471701474162, "learning_rate": 5.173152338914352e-06, "loss": 0.3291, "step": 4982 }, { "epoch": 1.6210149642160052, "grad_norm": 1.7686791327718918, "learning_rate": 5.1712607081167645e-06, "loss": 0.3869, "step": 4983 }, { "epoch": 1.6213402732595967, "grad_norm": 1.9815712201913713, "learning_rate": 5.1693690527774034e-06, "loss": 0.4259, "step": 4984 }, { "epoch": 1.621665582303188, "grad_norm": 2.018729791211919, "learning_rate": 5.167477373167343e-06, "loss": 0.4417, "step": 4985 }, { "epoch": 1.6219908913467793, "grad_norm": 1.8889791336594757, "learning_rate": 5.165585669557663e-06, "loss": 0.3217, "step": 4986 }, { "epoch": 1.6223162003903708, "grad_norm": 1.7890361514175186, "learning_rate": 5.163693942219449e-06, "loss": 0.3617, "step": 4987 }, { "epoch": 1.6226415094339623, "grad_norm": 2.2254743267686896, "learning_rate": 5.16180219142378e-06, "loss": 0.4289, "step": 4988 }, { "epoch": 1.6229668184775536, "grad_norm": 1.8866010982219277, "learning_rate": 5.159910417441752e-06, "loss": 0.3856, "step": 4989 }, { "epoch": 1.623292127521145, "grad_norm": 1.8121027374442826, "learning_rate": 5.158018620544454e-06, "loss": 0.3691, "step": 4990 }, { "epoch": 1.6236174365647364, "grad_norm": 1.6305763084741685, "learning_rate": 5.156126801002982e-06, "loss": 0.3385, "step": 4991 }, { "epoch": 1.623942745608328, "grad_norm": 1.8729012834910181, "learning_rate": 5.1542349590884345e-06, "loss": 0.413, "step": 4992 }, { "epoch": 1.6242680546519193, "grad_norm": 1.6807403645064696, "learning_rate": 5.152343095071915e-06, "loss": 0.3047, "step": 4993 }, { "epoch": 1.6245933636955108, "grad_norm": 1.770939703438272, "learning_rate": 5.1504512092245285e-06, "loss": 0.3277, "step": 4994 }, { "epoch": 1.6249186727391023, "grad_norm": 1.564700357546072, "learning_rate": 5.148559301817384e-06, "loss": 0.3355, "step": 4995 }, { "epoch": 1.6252439817826936, "grad_norm": 2.0893789670118825, "learning_rate": 5.146667373121591e-06, "loss": 0.4666, "step": 4996 }, { "epoch": 1.6255692908262849, "grad_norm": 1.9589619077300189, "learning_rate": 5.144775423408268e-06, "loss": 0.3767, "step": 4997 }, { "epoch": 1.6258945998698764, "grad_norm": 1.884755963639399, "learning_rate": 5.142883452948529e-06, "loss": 0.3861, "step": 4998 }, { "epoch": 1.626219908913468, "grad_norm": 2.0966756882515063, "learning_rate": 5.140991462013496e-06, "loss": 0.4592, "step": 4999 }, { "epoch": 1.6265452179570592, "grad_norm": 2.3217834121281675, "learning_rate": 5.139099450874292e-06, "loss": 0.404, "step": 5000 }, { "epoch": 1.6268705270006505, "grad_norm": 2.0880304301069894, "learning_rate": 5.137207419802045e-06, "loss": 0.4839, "step": 5001 }, { "epoch": 1.627195836044242, "grad_norm": 2.1953230203989924, "learning_rate": 5.135315369067881e-06, "loss": 0.4483, "step": 5002 }, { "epoch": 1.6275211450878335, "grad_norm": 1.8638253534290938, "learning_rate": 5.133423298942938e-06, "loss": 0.3879, "step": 5003 }, { "epoch": 1.6278464541314248, "grad_norm": 1.8683098399303146, "learning_rate": 5.131531209698345e-06, "loss": 0.4516, "step": 5004 }, { "epoch": 1.628171763175016, "grad_norm": 1.9838769166836456, "learning_rate": 5.129639101605242e-06, "loss": 0.3404, "step": 5005 }, { "epoch": 1.6284970722186076, "grad_norm": 1.7637829476738347, "learning_rate": 5.12774697493477e-06, "loss": 0.352, "step": 5006 }, { "epoch": 1.6288223812621991, "grad_norm": 2.012395741640791, "learning_rate": 5.125854829958071e-06, "loss": 0.3739, "step": 5007 }, { "epoch": 1.6291476903057904, "grad_norm": 2.1016157469574317, "learning_rate": 5.123962666946289e-06, "loss": 0.3808, "step": 5008 }, { "epoch": 1.629472999349382, "grad_norm": 1.9762532544285578, "learning_rate": 5.1220704861705775e-06, "loss": 0.4347, "step": 5009 }, { "epoch": 1.6297983083929735, "grad_norm": 2.006564090066713, "learning_rate": 5.120178287902082e-06, "loss": 0.3365, "step": 5010 }, { "epoch": 1.6301236174365648, "grad_norm": 1.9312932215649312, "learning_rate": 5.118286072411958e-06, "loss": 0.359, "step": 5011 }, { "epoch": 1.630448926480156, "grad_norm": 2.0121323085362977, "learning_rate": 5.1163938399713605e-06, "loss": 0.4095, "step": 5012 }, { "epoch": 1.6307742355237476, "grad_norm": 1.8161613140141077, "learning_rate": 5.114501590851449e-06, "loss": 0.4068, "step": 5013 }, { "epoch": 1.631099544567339, "grad_norm": 1.8927058075162064, "learning_rate": 5.112609325323384e-06, "loss": 0.3832, "step": 5014 }, { "epoch": 1.6314248536109304, "grad_norm": 1.8870900785132896, "learning_rate": 5.110717043658325e-06, "loss": 0.3779, "step": 5015 }, { "epoch": 1.6317501626545217, "grad_norm": 1.709667457714813, "learning_rate": 5.108824746127443e-06, "loss": 0.3592, "step": 5016 }, { "epoch": 1.6320754716981132, "grad_norm": 1.8656604967073647, "learning_rate": 5.1069324330019e-06, "loss": 0.3946, "step": 5017 }, { "epoch": 1.6324007807417047, "grad_norm": 1.6877511363622202, "learning_rate": 5.1050401045528695e-06, "loss": 0.3452, "step": 5018 }, { "epoch": 1.632726089785296, "grad_norm": 2.1106076811743617, "learning_rate": 5.103147761051521e-06, "loss": 0.3993, "step": 5019 }, { "epoch": 1.6330513988288873, "grad_norm": 1.8629666069996356, "learning_rate": 5.10125540276903e-06, "loss": 0.3143, "step": 5020 }, { "epoch": 1.633376707872479, "grad_norm": 1.7854001893318387, "learning_rate": 5.099363029976572e-06, "loss": 0.3864, "step": 5021 }, { "epoch": 1.6337020169160703, "grad_norm": 1.7830495637737014, "learning_rate": 5.097470642945327e-06, "loss": 0.3999, "step": 5022 }, { "epoch": 1.6340273259596616, "grad_norm": 2.034439349441734, "learning_rate": 5.095578241946471e-06, "loss": 0.3941, "step": 5023 }, { "epoch": 1.6343526350032531, "grad_norm": 1.9782273870840397, "learning_rate": 5.093685827251191e-06, "loss": 0.4102, "step": 5024 }, { "epoch": 1.6346779440468446, "grad_norm": 2.0266962339474244, "learning_rate": 5.09179339913067e-06, "loss": 0.4062, "step": 5025 }, { "epoch": 1.635003253090436, "grad_norm": 2.154495960243838, "learning_rate": 5.089900957856093e-06, "loss": 0.4287, "step": 5026 }, { "epoch": 1.6353285621340272, "grad_norm": 1.962067288565051, "learning_rate": 5.088008503698648e-06, "loss": 0.357, "step": 5027 }, { "epoch": 1.6356538711776187, "grad_norm": 1.7389057265584111, "learning_rate": 5.086116036929526e-06, "loss": 0.4388, "step": 5028 }, { "epoch": 1.6359791802212102, "grad_norm": 1.93932100115712, "learning_rate": 5.08422355781992e-06, "loss": 0.3842, "step": 5029 }, { "epoch": 1.6363044892648015, "grad_norm": 1.5385745126236938, "learning_rate": 5.08233106664102e-06, "loss": 0.3106, "step": 5030 }, { "epoch": 1.6366297983083928, "grad_norm": 1.8160169891678348, "learning_rate": 5.080438563664024e-06, "loss": 0.2935, "step": 5031 }, { "epoch": 1.6369551073519844, "grad_norm": 2.0303123016766578, "learning_rate": 5.07854604916013e-06, "loss": 0.3805, "step": 5032 }, { "epoch": 1.6372804163955759, "grad_norm": 2.08147730902445, "learning_rate": 5.076653523400532e-06, "loss": 0.4089, "step": 5033 }, { "epoch": 1.6376057254391672, "grad_norm": 1.8578349965405596, "learning_rate": 5.074760986656435e-06, "loss": 0.3245, "step": 5034 }, { "epoch": 1.6379310344827587, "grad_norm": 1.9468390232732333, "learning_rate": 5.072868439199038e-06, "loss": 0.3037, "step": 5035 }, { "epoch": 1.6382563435263502, "grad_norm": 2.1525399214983083, "learning_rate": 5.070975881299546e-06, "loss": 0.4362, "step": 5036 }, { "epoch": 1.6385816525699415, "grad_norm": 1.9894318744431523, "learning_rate": 5.069083313229161e-06, "loss": 0.3897, "step": 5037 }, { "epoch": 1.6389069616135328, "grad_norm": 2.598443247326733, "learning_rate": 5.0671907352590925e-06, "loss": 0.4469, "step": 5038 }, { "epoch": 1.6392322706571243, "grad_norm": 1.9793145711041458, "learning_rate": 5.065298147660546e-06, "loss": 0.3181, "step": 5039 }, { "epoch": 1.6395575797007158, "grad_norm": 1.8036844821725966, "learning_rate": 5.063405550704732e-06, "loss": 0.3539, "step": 5040 }, { "epoch": 1.639882888744307, "grad_norm": 1.7441031360974868, "learning_rate": 5.061512944662861e-06, "loss": 0.3089, "step": 5041 }, { "epoch": 1.6402081977878984, "grad_norm": 2.2193721698747515, "learning_rate": 5.0596203298061435e-06, "loss": 0.4112, "step": 5042 }, { "epoch": 1.64053350683149, "grad_norm": 1.969183228036988, "learning_rate": 5.057727706405791e-06, "loss": 0.3972, "step": 5043 }, { "epoch": 1.6408588158750814, "grad_norm": 1.6408915875035925, "learning_rate": 5.055835074733022e-06, "loss": 0.3309, "step": 5044 }, { "epoch": 1.6411841249186727, "grad_norm": 2.1277638809372403, "learning_rate": 5.053942435059049e-06, "loss": 0.4027, "step": 5045 }, { "epoch": 1.641509433962264, "grad_norm": 1.5391249562372158, "learning_rate": 5.052049787655089e-06, "loss": 0.3548, "step": 5046 }, { "epoch": 1.6418347430058555, "grad_norm": 1.9541253557176745, "learning_rate": 5.050157132792359e-06, "loss": 0.384, "step": 5047 }, { "epoch": 1.642160052049447, "grad_norm": 1.9952452038990698, "learning_rate": 5.04826447074208e-06, "loss": 0.4115, "step": 5048 }, { "epoch": 1.6424853610930383, "grad_norm": 1.7258311629517629, "learning_rate": 5.046371801775469e-06, "loss": 0.3149, "step": 5049 }, { "epoch": 1.6428106701366298, "grad_norm": 1.9725638046541154, "learning_rate": 5.044479126163748e-06, "loss": 0.3858, "step": 5050 }, { "epoch": 1.6431359791802214, "grad_norm": 2.278967117665045, "learning_rate": 5.042586444178139e-06, "loss": 0.4252, "step": 5051 }, { "epoch": 1.6434612882238127, "grad_norm": 1.8622286569308557, "learning_rate": 5.040693756089865e-06, "loss": 0.3688, "step": 5052 }, { "epoch": 1.643786597267404, "grad_norm": 1.7849997852452342, "learning_rate": 5.038801062170147e-06, "loss": 0.3589, "step": 5053 }, { "epoch": 1.6441119063109955, "grad_norm": 1.8626925731156694, "learning_rate": 5.03690836269021e-06, "loss": 0.3231, "step": 5054 }, { "epoch": 1.644437215354587, "grad_norm": 2.1501172725584863, "learning_rate": 5.035015657921283e-06, "loss": 0.484, "step": 5055 }, { "epoch": 1.6447625243981783, "grad_norm": 1.787783783879044, "learning_rate": 5.033122948134587e-06, "loss": 0.4389, "step": 5056 }, { "epoch": 1.6450878334417696, "grad_norm": 1.8403512628778231, "learning_rate": 5.0312302336013494e-06, "loss": 0.3545, "step": 5057 }, { "epoch": 1.645413142485361, "grad_norm": 2.162461148138729, "learning_rate": 5.0293375145928026e-06, "loss": 0.3513, "step": 5058 }, { "epoch": 1.6457384515289526, "grad_norm": 1.9463248381022575, "learning_rate": 5.027444791380166e-06, "loss": 0.4312, "step": 5059 }, { "epoch": 1.6460637605725439, "grad_norm": 1.9091966022952331, "learning_rate": 5.025552064234674e-06, "loss": 0.3413, "step": 5060 }, { "epoch": 1.6463890696161352, "grad_norm": 1.9685391161677135, "learning_rate": 5.023659333427556e-06, "loss": 0.3881, "step": 5061 }, { "epoch": 1.6467143786597267, "grad_norm": 1.766523750255741, "learning_rate": 5.021766599230038e-06, "loss": 0.3545, "step": 5062 }, { "epoch": 1.6470396877033182, "grad_norm": 1.722556380705883, "learning_rate": 5.0198738619133505e-06, "loss": 0.3453, "step": 5063 }, { "epoch": 1.6473649967469095, "grad_norm": 1.8804156568538657, "learning_rate": 5.0179811217487285e-06, "loss": 0.3973, "step": 5064 }, { "epoch": 1.647690305790501, "grad_norm": 2.187044153117335, "learning_rate": 5.016088379007397e-06, "loss": 0.3928, "step": 5065 }, { "epoch": 1.6480156148340925, "grad_norm": 2.0419202073604668, "learning_rate": 5.014195633960591e-06, "loss": 0.427, "step": 5066 }, { "epoch": 1.6483409238776838, "grad_norm": 1.7256446888260857, "learning_rate": 5.012302886879542e-06, "loss": 0.3896, "step": 5067 }, { "epoch": 1.6486662329212751, "grad_norm": 2.219627114771072, "learning_rate": 5.010410138035481e-06, "loss": 0.4817, "step": 5068 }, { "epoch": 1.6489915419648666, "grad_norm": 2.001909821514017, "learning_rate": 5.008517387699638e-06, "loss": 0.3605, "step": 5069 }, { "epoch": 1.6493168510084582, "grad_norm": 1.8259178627876345, "learning_rate": 5.00662463614325e-06, "loss": 0.2993, "step": 5070 }, { "epoch": 1.6496421600520494, "grad_norm": 1.9214364798811976, "learning_rate": 5.004731883637547e-06, "loss": 0.356, "step": 5071 }, { "epoch": 1.6499674690956407, "grad_norm": 1.9094809080347448, "learning_rate": 5.0028391304537615e-06, "loss": 0.396, "step": 5072 }, { "epoch": 1.6502927781392323, "grad_norm": 2.0654061219355744, "learning_rate": 5.000946376863126e-06, "loss": 0.4632, "step": 5073 }, { "epoch": 1.6506180871828238, "grad_norm": 2.125847373602413, "learning_rate": 4.999053623136875e-06, "loss": 0.4319, "step": 5074 }, { "epoch": 1.650943396226415, "grad_norm": 2.0702928222521972, "learning_rate": 4.99716086954624e-06, "loss": 0.4319, "step": 5075 }, { "epoch": 1.6512687052700064, "grad_norm": 1.8121993857361363, "learning_rate": 4.995268116362455e-06, "loss": 0.3666, "step": 5076 }, { "epoch": 1.6515940143135979, "grad_norm": 1.8942899748274762, "learning_rate": 4.993375363856752e-06, "loss": 0.4156, "step": 5077 }, { "epoch": 1.6519193233571894, "grad_norm": 1.870835851322801, "learning_rate": 4.991482612300362e-06, "loss": 0.3654, "step": 5078 }, { "epoch": 1.6522446324007807, "grad_norm": 1.9598327091511205, "learning_rate": 4.9895898619645215e-06, "loss": 0.3741, "step": 5079 }, { "epoch": 1.6525699414443722, "grad_norm": 1.7056599769383982, "learning_rate": 4.98769711312046e-06, "loss": 0.4269, "step": 5080 }, { "epoch": 1.6528952504879637, "grad_norm": 1.9079687677542805, "learning_rate": 4.985804366039411e-06, "loss": 0.4205, "step": 5081 }, { "epoch": 1.653220559531555, "grad_norm": 2.1821873907137683, "learning_rate": 4.983911620992603e-06, "loss": 0.3886, "step": 5082 }, { "epoch": 1.6535458685751463, "grad_norm": 1.8680920854807677, "learning_rate": 4.982018878251273e-06, "loss": 0.3666, "step": 5083 }, { "epoch": 1.6538711776187378, "grad_norm": 1.725841679095186, "learning_rate": 4.98012613808665e-06, "loss": 0.2656, "step": 5084 }, { "epoch": 1.6541964866623293, "grad_norm": 1.9861841565421798, "learning_rate": 4.978233400769965e-06, "loss": 0.3807, "step": 5085 }, { "epoch": 1.6545217957059206, "grad_norm": 2.189568371652666, "learning_rate": 4.976340666572446e-06, "loss": 0.4005, "step": 5086 }, { "epoch": 1.654847104749512, "grad_norm": 1.891872479215115, "learning_rate": 4.9744479357653266e-06, "loss": 0.3764, "step": 5087 }, { "epoch": 1.6551724137931034, "grad_norm": 1.9256804666909337, "learning_rate": 4.972555208619836e-06, "loss": 0.4166, "step": 5088 }, { "epoch": 1.655497722836695, "grad_norm": 1.9599995408953914, "learning_rate": 4.970662485407201e-06, "loss": 0.3848, "step": 5089 }, { "epoch": 1.6558230318802862, "grad_norm": 1.9788739307731205, "learning_rate": 4.96876976639865e-06, "loss": 0.4663, "step": 5090 }, { "epoch": 1.6561483409238775, "grad_norm": 2.0471675505645934, "learning_rate": 4.966877051865414e-06, "loss": 0.4494, "step": 5091 }, { "epoch": 1.656473649967469, "grad_norm": 1.6264015440736526, "learning_rate": 4.964984342078718e-06, "loss": 0.2764, "step": 5092 }, { "epoch": 1.6567989590110606, "grad_norm": 1.9131661795852575, "learning_rate": 4.9630916373097905e-06, "loss": 0.4389, "step": 5093 }, { "epoch": 1.6571242680546518, "grad_norm": 1.9060113099898184, "learning_rate": 4.961198937829854e-06, "loss": 0.3685, "step": 5094 }, { "epoch": 1.6574495770982434, "grad_norm": 1.7575235847092907, "learning_rate": 4.959306243910137e-06, "loss": 0.3219, "step": 5095 }, { "epoch": 1.6577748861418349, "grad_norm": 1.8231424840129755, "learning_rate": 4.957413555821862e-06, "loss": 0.3575, "step": 5096 }, { "epoch": 1.6581001951854262, "grad_norm": 1.89069878862586, "learning_rate": 4.9555208738362545e-06, "loss": 0.3898, "step": 5097 }, { "epoch": 1.6584255042290175, "grad_norm": 1.737094093756191, "learning_rate": 4.953628198224532e-06, "loss": 0.2801, "step": 5098 }, { "epoch": 1.658750813272609, "grad_norm": 1.8362204218224139, "learning_rate": 4.9517355292579215e-06, "loss": 0.332, "step": 5099 }, { "epoch": 1.6590761223162005, "grad_norm": 1.8327077680013586, "learning_rate": 4.949842867207643e-06, "loss": 0.3201, "step": 5100 }, { "epoch": 1.6594014313597918, "grad_norm": 1.7704263578547494, "learning_rate": 4.947950212344914e-06, "loss": 0.3168, "step": 5101 }, { "epoch": 1.659726740403383, "grad_norm": 2.0093313661586127, "learning_rate": 4.946057564940952e-06, "loss": 0.3694, "step": 5102 }, { "epoch": 1.6600520494469746, "grad_norm": 2.002630688037598, "learning_rate": 4.94416492526698e-06, "loss": 0.3725, "step": 5103 }, { "epoch": 1.6603773584905661, "grad_norm": 2.1001796652315123, "learning_rate": 4.94227229359421e-06, "loss": 0.3821, "step": 5104 }, { "epoch": 1.6607026675341574, "grad_norm": 1.80874537884998, "learning_rate": 4.940379670193859e-06, "loss": 0.3737, "step": 5105 }, { "epoch": 1.6610279765777487, "grad_norm": 1.8908297065992525, "learning_rate": 4.9384870553371415e-06, "loss": 0.3114, "step": 5106 }, { "epoch": 1.6613532856213404, "grad_norm": 1.7377291708735871, "learning_rate": 4.93659444929527e-06, "loss": 0.2921, "step": 5107 }, { "epoch": 1.6616785946649317, "grad_norm": 1.987307100620611, "learning_rate": 4.934701852339456e-06, "loss": 0.4091, "step": 5108 }, { "epoch": 1.662003903708523, "grad_norm": 1.9940553820584077, "learning_rate": 4.93280926474091e-06, "loss": 0.3597, "step": 5109 }, { "epoch": 1.6623292127521145, "grad_norm": 1.9414861854632588, "learning_rate": 4.93091668677084e-06, "loss": 0.333, "step": 5110 }, { "epoch": 1.662654521795706, "grad_norm": 2.106165678477128, "learning_rate": 4.929024118700456e-06, "loss": 0.4293, "step": 5111 }, { "epoch": 1.6629798308392973, "grad_norm": 1.8835258419423497, "learning_rate": 4.927131560800964e-06, "loss": 0.3497, "step": 5112 }, { "epoch": 1.6633051398828886, "grad_norm": 1.6523802095242683, "learning_rate": 4.925239013343568e-06, "loss": 0.3258, "step": 5113 }, { "epoch": 1.6636304489264802, "grad_norm": 2.197316879419648, "learning_rate": 4.9233464765994685e-06, "loss": 0.4505, "step": 5114 }, { "epoch": 1.6639557579700717, "grad_norm": 1.6530729575139291, "learning_rate": 4.9214539508398715e-06, "loss": 0.323, "step": 5115 }, { "epoch": 1.664281067013663, "grad_norm": 1.8558624568271211, "learning_rate": 4.919561436335977e-06, "loss": 0.3782, "step": 5116 }, { "epoch": 1.6646063760572543, "grad_norm": 1.8037573867395478, "learning_rate": 4.917668933358981e-06, "loss": 0.3413, "step": 5117 }, { "epoch": 1.6649316851008458, "grad_norm": 2.1066539911769264, "learning_rate": 4.91577644218008e-06, "loss": 0.4197, "step": 5118 }, { "epoch": 1.6652569941444373, "grad_norm": 1.8562563439995352, "learning_rate": 4.913883963070475e-06, "loss": 0.3718, "step": 5119 }, { "epoch": 1.6655823031880286, "grad_norm": 1.8719566213855714, "learning_rate": 4.911991496301353e-06, "loss": 0.3659, "step": 5120 }, { "epoch": 1.66590761223162, "grad_norm": 2.0149646451538974, "learning_rate": 4.9100990421439095e-06, "loss": 0.4067, "step": 5121 }, { "epoch": 1.6662329212752116, "grad_norm": 1.801535119679771, "learning_rate": 4.908206600869332e-06, "loss": 0.4226, "step": 5122 }, { "epoch": 1.666558230318803, "grad_norm": 1.6263517763213764, "learning_rate": 4.90631417274881e-06, "loss": 0.2839, "step": 5123 }, { "epoch": 1.6668835393623942, "grad_norm": 2.063487394270264, "learning_rate": 4.90442175805353e-06, "loss": 0.3762, "step": 5124 }, { "epoch": 1.6672088484059857, "grad_norm": 1.8509071532280883, "learning_rate": 4.9025293570546766e-06, "loss": 0.4083, "step": 5125 }, { "epoch": 1.6675341574495772, "grad_norm": 2.1352798425556636, "learning_rate": 4.900636970023428e-06, "loss": 0.4519, "step": 5126 }, { "epoch": 1.6678594664931685, "grad_norm": 1.8520168728822186, "learning_rate": 4.898744597230971e-06, "loss": 0.3158, "step": 5127 }, { "epoch": 1.6681847755367598, "grad_norm": 1.926127398198977, "learning_rate": 4.89685223894848e-06, "loss": 0.366, "step": 5128 }, { "epoch": 1.6685100845803513, "grad_norm": 1.9528770614715982, "learning_rate": 4.894959895447132e-06, "loss": 0.4114, "step": 5129 }, { "epoch": 1.6688353936239428, "grad_norm": 1.6715798549054033, "learning_rate": 4.8930675669981005e-06, "loss": 0.3058, "step": 5130 }, { "epoch": 1.6691607026675341, "grad_norm": 2.214822253204632, "learning_rate": 4.891175253872558e-06, "loss": 0.4215, "step": 5131 }, { "epoch": 1.6694860117111254, "grad_norm": 1.6984642968275026, "learning_rate": 4.889282956341676e-06, "loss": 0.3033, "step": 5132 }, { "epoch": 1.669811320754717, "grad_norm": 1.9747667369978892, "learning_rate": 4.887390674676619e-06, "loss": 0.3784, "step": 5133 }, { "epoch": 1.6701366297983085, "grad_norm": 1.9125121403582805, "learning_rate": 4.885498409148551e-06, "loss": 0.3906, "step": 5134 }, { "epoch": 1.6704619388418998, "grad_norm": 2.0680660051439097, "learning_rate": 4.88360616002864e-06, "loss": 0.4876, "step": 5135 }, { "epoch": 1.6707872478854913, "grad_norm": 1.9403155910572416, "learning_rate": 4.881713927588044e-06, "loss": 0.3596, "step": 5136 }, { "epoch": 1.6711125569290828, "grad_norm": 1.7258863945905791, "learning_rate": 4.87982171209792e-06, "loss": 0.3174, "step": 5137 }, { "epoch": 1.671437865972674, "grad_norm": 2.301561838761158, "learning_rate": 4.877929513829424e-06, "loss": 0.346, "step": 5138 }, { "epoch": 1.6717631750162654, "grad_norm": 2.0548901689139916, "learning_rate": 4.876037333053711e-06, "loss": 0.4425, "step": 5139 }, { "epoch": 1.6720884840598569, "grad_norm": 1.6072335284778896, "learning_rate": 4.874145170041932e-06, "loss": 0.3592, "step": 5140 }, { "epoch": 1.6724137931034484, "grad_norm": 2.037477737468502, "learning_rate": 4.872253025065232e-06, "loss": 0.4621, "step": 5141 }, { "epoch": 1.6727391021470397, "grad_norm": 1.743953597246846, "learning_rate": 4.870360898394759e-06, "loss": 0.3703, "step": 5142 }, { "epoch": 1.673064411190631, "grad_norm": 2.02439714441892, "learning_rate": 4.868468790301657e-06, "loss": 0.4378, "step": 5143 }, { "epoch": 1.6733897202342225, "grad_norm": 1.9641661076943682, "learning_rate": 4.866576701057065e-06, "loss": 0.395, "step": 5144 }, { "epoch": 1.673715029277814, "grad_norm": 1.8634901164569582, "learning_rate": 4.864684630932119e-06, "loss": 0.3913, "step": 5145 }, { "epoch": 1.6740403383214053, "grad_norm": 1.9772922570083322, "learning_rate": 4.862792580197957e-06, "loss": 0.3821, "step": 5146 }, { "epoch": 1.6743656473649966, "grad_norm": 1.756944670311949, "learning_rate": 4.860900549125709e-06, "loss": 0.3154, "step": 5147 }, { "epoch": 1.6746909564085881, "grad_norm": 1.8285616776488502, "learning_rate": 4.859008537986506e-06, "loss": 0.3529, "step": 5148 }, { "epoch": 1.6750162654521796, "grad_norm": 1.795128299180378, "learning_rate": 4.857116547051472e-06, "loss": 0.3443, "step": 5149 }, { "epoch": 1.675341574495771, "grad_norm": 1.9030132840005236, "learning_rate": 4.855224576591733e-06, "loss": 0.3247, "step": 5150 }, { "epoch": 1.6756668835393624, "grad_norm": 1.9531110168503452, "learning_rate": 4.853332626878411e-06, "loss": 0.3869, "step": 5151 }, { "epoch": 1.675992192582954, "grad_norm": 2.0072550226915613, "learning_rate": 4.8514406981826185e-06, "loss": 0.3687, "step": 5152 }, { "epoch": 1.6763175016265452, "grad_norm": 2.373637958208087, "learning_rate": 4.8495487907754714e-06, "loss": 0.4571, "step": 5153 }, { "epoch": 1.6766428106701365, "grad_norm": 1.7987153857632905, "learning_rate": 4.847656904928086e-06, "loss": 0.3727, "step": 5154 }, { "epoch": 1.676968119713728, "grad_norm": 2.114781659831791, "learning_rate": 4.845765040911567e-06, "loss": 0.4785, "step": 5155 }, { "epoch": 1.6772934287573196, "grad_norm": 1.7108944913242377, "learning_rate": 4.843873198997021e-06, "loss": 0.3827, "step": 5156 }, { "epoch": 1.6776187378009109, "grad_norm": 1.8160181530976995, "learning_rate": 4.841981379455548e-06, "loss": 0.3097, "step": 5157 }, { "epoch": 1.6779440468445022, "grad_norm": 1.9800889486632791, "learning_rate": 4.840089582558249e-06, "loss": 0.4142, "step": 5158 }, { "epoch": 1.6782693558880937, "grad_norm": 2.2676165628010363, "learning_rate": 4.838197808576221e-06, "loss": 0.5018, "step": 5159 }, { "epoch": 1.6785946649316852, "grad_norm": 1.6650897369993072, "learning_rate": 4.8363060577805544e-06, "loss": 0.3663, "step": 5160 }, { "epoch": 1.6789199739752765, "grad_norm": 1.8318342481056318, "learning_rate": 4.8344143304423375e-06, "loss": 0.3757, "step": 5161 }, { "epoch": 1.6792452830188678, "grad_norm": 1.648212865288036, "learning_rate": 4.832522626832658e-06, "loss": 0.3538, "step": 5162 }, { "epoch": 1.6795705920624593, "grad_norm": 1.878705788583328, "learning_rate": 4.830630947222599e-06, "loss": 0.4538, "step": 5163 }, { "epoch": 1.6798959011060508, "grad_norm": 2.0286063121390665, "learning_rate": 4.828739291883237e-06, "loss": 0.3993, "step": 5164 }, { "epoch": 1.680221210149642, "grad_norm": 2.021628671454035, "learning_rate": 4.826847661085648e-06, "loss": 0.4106, "step": 5165 }, { "epoch": 1.6805465191932336, "grad_norm": 1.6113059791077666, "learning_rate": 4.824956055100905e-06, "loss": 0.2931, "step": 5166 }, { "epoch": 1.6808718282368251, "grad_norm": 1.9748335204418461, "learning_rate": 4.823064474200078e-06, "loss": 0.384, "step": 5167 }, { "epoch": 1.6811971372804164, "grad_norm": 1.6809037018031687, "learning_rate": 4.8211729186542295e-06, "loss": 0.3243, "step": 5168 }, { "epoch": 1.6815224463240077, "grad_norm": 2.1493521128716355, "learning_rate": 4.819281388734418e-06, "loss": 0.3826, "step": 5169 }, { "epoch": 1.6818477553675992, "grad_norm": 2.0076638644392983, "learning_rate": 4.817389884711706e-06, "loss": 0.4733, "step": 5170 }, { "epoch": 1.6821730644111907, "grad_norm": 2.0400461981315035, "learning_rate": 4.815498406857146e-06, "loss": 0.3664, "step": 5171 }, { "epoch": 1.682498373454782, "grad_norm": 1.8634606152154816, "learning_rate": 4.813606955441787e-06, "loss": 0.3573, "step": 5172 }, { "epoch": 1.6828236824983733, "grad_norm": 1.771619732206903, "learning_rate": 4.811715530736673e-06, "loss": 0.3578, "step": 5173 }, { "epoch": 1.6831489915419648, "grad_norm": 1.800396414588169, "learning_rate": 4.809824133012852e-06, "loss": 0.3199, "step": 5174 }, { "epoch": 1.6834743005855564, "grad_norm": 2.2456605810318786, "learning_rate": 4.8079327625413585e-06, "loss": 0.4468, "step": 5175 }, { "epoch": 1.6837996096291477, "grad_norm": 1.785915426791318, "learning_rate": 4.806041419593229e-06, "loss": 0.3251, "step": 5176 }, { "epoch": 1.684124918672739, "grad_norm": 1.9005231324167933, "learning_rate": 4.8041501044394915e-06, "loss": 0.3774, "step": 5177 }, { "epoch": 1.6844502277163305, "grad_norm": 1.85170501941623, "learning_rate": 4.802258817351176e-06, "loss": 0.3486, "step": 5178 }, { "epoch": 1.684775536759922, "grad_norm": 2.3556956275430987, "learning_rate": 4.800367558599303e-06, "loss": 0.4236, "step": 5179 }, { "epoch": 1.6851008458035133, "grad_norm": 2.041991866856133, "learning_rate": 4.798476328454893e-06, "loss": 0.4292, "step": 5180 }, { "epoch": 1.6854261548471048, "grad_norm": 1.7909499440802774, "learning_rate": 4.796585127188958e-06, "loss": 0.3388, "step": 5181 }, { "epoch": 1.6857514638906963, "grad_norm": 2.024687441017129, "learning_rate": 4.794693955072511e-06, "loss": 0.3702, "step": 5182 }, { "epoch": 1.6860767729342876, "grad_norm": 1.8157098873575506, "learning_rate": 4.7928028123765575e-06, "loss": 0.4125, "step": 5183 }, { "epoch": 1.6864020819778789, "grad_norm": 1.971208614295115, "learning_rate": 4.790911699372099e-06, "loss": 0.3842, "step": 5184 }, { "epoch": 1.6867273910214704, "grad_norm": 1.8251935472699659, "learning_rate": 4.7890206163301316e-06, "loss": 0.3448, "step": 5185 }, { "epoch": 1.687052700065062, "grad_norm": 1.8306373126907225, "learning_rate": 4.787129563521652e-06, "loss": 0.4177, "step": 5186 }, { "epoch": 1.6873780091086532, "grad_norm": 1.658230727001696, "learning_rate": 4.78523854121765e-06, "loss": 0.325, "step": 5187 }, { "epoch": 1.6877033181522445, "grad_norm": 2.158808113213444, "learning_rate": 4.783347549689105e-06, "loss": 0.3687, "step": 5188 }, { "epoch": 1.688028627195836, "grad_norm": 1.6115639688164323, "learning_rate": 4.781456589207e-06, "loss": 0.2714, "step": 5189 }, { "epoch": 1.6883539362394275, "grad_norm": 2.0231194938898533, "learning_rate": 4.779565660042315e-06, "loss": 0.4448, "step": 5190 }, { "epoch": 1.6886792452830188, "grad_norm": 1.9073094316633201, "learning_rate": 4.777674762466016e-06, "loss": 0.3541, "step": 5191 }, { "epoch": 1.6890045543266101, "grad_norm": 1.8790793552060812, "learning_rate": 4.775783896749072e-06, "loss": 0.3591, "step": 5192 }, { "epoch": 1.6893298633702019, "grad_norm": 2.132638147126592, "learning_rate": 4.773893063162443e-06, "loss": 0.4087, "step": 5193 }, { "epoch": 1.6896551724137931, "grad_norm": 2.0325047053474465, "learning_rate": 4.772002261977091e-06, "loss": 0.3606, "step": 5194 }, { "epoch": 1.6899804814573844, "grad_norm": 2.0836910335311214, "learning_rate": 4.770111493463967e-06, "loss": 0.4743, "step": 5195 }, { "epoch": 1.690305790500976, "grad_norm": 2.0372611226568855, "learning_rate": 4.768220757894019e-06, "loss": 0.4159, "step": 5196 }, { "epoch": 1.6906310995445675, "grad_norm": 1.672149961663288, "learning_rate": 4.766330055538189e-06, "loss": 0.3369, "step": 5197 }, { "epoch": 1.6909564085881588, "grad_norm": 2.152766752700387, "learning_rate": 4.7644393866674205e-06, "loss": 0.4384, "step": 5198 }, { "epoch": 1.69128171763175, "grad_norm": 1.8566717451303087, "learning_rate": 4.762548751552644e-06, "loss": 0.3773, "step": 5199 }, { "epoch": 1.6916070266753416, "grad_norm": 1.7415202156907623, "learning_rate": 4.7606581504647916e-06, "loss": 0.3549, "step": 5200 }, { "epoch": 1.691932335718933, "grad_norm": 2.1567767768029507, "learning_rate": 4.758767583674782e-06, "loss": 0.4408, "step": 5201 }, { "epoch": 1.6922576447625244, "grad_norm": 1.8617394219132635, "learning_rate": 4.75687705145354e-06, "loss": 0.3427, "step": 5202 }, { "epoch": 1.6925829538061157, "grad_norm": 2.307395249963546, "learning_rate": 4.754986554071981e-06, "loss": 0.551, "step": 5203 }, { "epoch": 1.6929082628497072, "grad_norm": 2.1727656410179454, "learning_rate": 4.75309609180101e-06, "loss": 0.4815, "step": 5204 }, { "epoch": 1.6932335718932987, "grad_norm": 2.0411103709719827, "learning_rate": 4.751205664911532e-06, "loss": 0.4055, "step": 5205 }, { "epoch": 1.69355888093689, "grad_norm": 2.0541095653294876, "learning_rate": 4.749315273674451e-06, "loss": 0.4232, "step": 5206 }, { "epoch": 1.6938841899804813, "grad_norm": 2.272611965265972, "learning_rate": 4.747424918360658e-06, "loss": 0.4067, "step": 5207 }, { "epoch": 1.694209499024073, "grad_norm": 1.649178530183466, "learning_rate": 4.745534599241041e-06, "loss": 0.3398, "step": 5208 }, { "epoch": 1.6945348080676643, "grad_norm": 2.093263832908556, "learning_rate": 4.7436443165864865e-06, "loss": 0.3293, "step": 5209 }, { "epoch": 1.6948601171112556, "grad_norm": 1.918579295254898, "learning_rate": 4.7417540706678725e-06, "loss": 0.4384, "step": 5210 }, { "epoch": 1.6951854261548471, "grad_norm": 2.0165227893156206, "learning_rate": 4.739863861756073e-06, "loss": 0.436, "step": 5211 }, { "epoch": 1.6955107351984386, "grad_norm": 2.08325004984653, "learning_rate": 4.737973690121954e-06, "loss": 0.4572, "step": 5212 }, { "epoch": 1.69583604424203, "grad_norm": 1.9350177367587922, "learning_rate": 4.73608355603638e-06, "loss": 0.4471, "step": 5213 }, { "epoch": 1.6961613532856212, "grad_norm": 1.9852135849842467, "learning_rate": 4.73419345977021e-06, "loss": 0.4788, "step": 5214 }, { "epoch": 1.6964866623292127, "grad_norm": 1.8475272738070085, "learning_rate": 4.732303401594294e-06, "loss": 0.3642, "step": 5215 }, { "epoch": 1.6968119713728043, "grad_norm": 2.0379575379604504, "learning_rate": 4.730413381779479e-06, "loss": 0.4046, "step": 5216 }, { "epoch": 1.6971372804163956, "grad_norm": 1.9805735817668233, "learning_rate": 4.728523400596604e-06, "loss": 0.3912, "step": 5217 }, { "epoch": 1.6974625894599868, "grad_norm": 1.9488911448278357, "learning_rate": 4.726633458316509e-06, "loss": 0.4424, "step": 5218 }, { "epoch": 1.6977878985035784, "grad_norm": 1.7774015743202256, "learning_rate": 4.724743555210022e-06, "loss": 0.3688, "step": 5219 }, { "epoch": 1.6981132075471699, "grad_norm": 1.8580234061773842, "learning_rate": 4.722853691547966e-06, "loss": 0.3714, "step": 5220 }, { "epoch": 1.6984385165907612, "grad_norm": 1.9611811714985088, "learning_rate": 4.720963867601159e-06, "loss": 0.3969, "step": 5221 }, { "epoch": 1.6987638256343527, "grad_norm": 2.126332031577764, "learning_rate": 4.719074083640419e-06, "loss": 0.4147, "step": 5222 }, { "epoch": 1.6990891346779442, "grad_norm": 1.6692457419311537, "learning_rate": 4.7171843399365495e-06, "loss": 0.3812, "step": 5223 }, { "epoch": 1.6994144437215355, "grad_norm": 1.8447308394532194, "learning_rate": 4.715294636760352e-06, "loss": 0.4116, "step": 5224 }, { "epoch": 1.6997397527651268, "grad_norm": 1.9304822241448807, "learning_rate": 4.713404974382622e-06, "loss": 0.3916, "step": 5225 }, { "epoch": 1.7000650618087183, "grad_norm": 1.9777502331695997, "learning_rate": 4.711515353074152e-06, "loss": 0.4369, "step": 5226 }, { "epoch": 1.7003903708523098, "grad_norm": 2.195890500198371, "learning_rate": 4.709625773105723e-06, "loss": 0.3706, "step": 5227 }, { "epoch": 1.7007156798959011, "grad_norm": 1.9047854229701031, "learning_rate": 4.7077362347481145e-06, "loss": 0.3984, "step": 5228 }, { "epoch": 1.7010409889394924, "grad_norm": 1.770541505631111, "learning_rate": 4.705846738272098e-06, "loss": 0.3151, "step": 5229 }, { "epoch": 1.701366297983084, "grad_norm": 2.143722392208159, "learning_rate": 4.703957283948441e-06, "loss": 0.3731, "step": 5230 }, { "epoch": 1.7016916070266754, "grad_norm": 2.007279875955092, "learning_rate": 4.702067872047902e-06, "loss": 0.3854, "step": 5231 }, { "epoch": 1.7020169160702667, "grad_norm": 1.9708295508435751, "learning_rate": 4.700178502841236e-06, "loss": 0.3491, "step": 5232 }, { "epoch": 1.702342225113858, "grad_norm": 1.882867112754897, "learning_rate": 4.6982891765991895e-06, "loss": 0.3083, "step": 5233 }, { "epoch": 1.7026675341574495, "grad_norm": 1.5597056988696534, "learning_rate": 4.696399893592506e-06, "loss": 0.2919, "step": 5234 }, { "epoch": 1.702992843201041, "grad_norm": 1.9717681768896116, "learning_rate": 4.6945106540919204e-06, "loss": 0.3693, "step": 5235 }, { "epoch": 1.7033181522446323, "grad_norm": 2.465472665249711, "learning_rate": 4.692621458368163e-06, "loss": 0.4539, "step": 5236 }, { "epoch": 1.7036434612882239, "grad_norm": 1.717697179293206, "learning_rate": 4.690732306691952e-06, "loss": 0.343, "step": 5237 }, { "epoch": 1.7039687703318154, "grad_norm": 1.7037526890686279, "learning_rate": 4.688843199334011e-06, "loss": 0.3122, "step": 5238 }, { "epoch": 1.7042940793754067, "grad_norm": 2.077730500175737, "learning_rate": 4.686954136565045e-06, "loss": 0.4004, "step": 5239 }, { "epoch": 1.704619388418998, "grad_norm": 1.887117486184635, "learning_rate": 4.685065118655762e-06, "loss": 0.3474, "step": 5240 }, { "epoch": 1.7049446974625895, "grad_norm": 2.0225367516982584, "learning_rate": 4.683176145876854e-06, "loss": 0.3979, "step": 5241 }, { "epoch": 1.705270006506181, "grad_norm": 1.9120180959791306, "learning_rate": 4.681287218499018e-06, "loss": 0.3639, "step": 5242 }, { "epoch": 1.7055953155497723, "grad_norm": 2.199697803559264, "learning_rate": 4.679398336792937e-06, "loss": 0.4529, "step": 5243 }, { "epoch": 1.7059206245933636, "grad_norm": 1.8934209837038072, "learning_rate": 4.677509501029286e-06, "loss": 0.4288, "step": 5244 }, { "epoch": 1.706245933636955, "grad_norm": 1.8739300439459543, "learning_rate": 4.675620711478738e-06, "loss": 0.3398, "step": 5245 }, { "epoch": 1.7065712426805466, "grad_norm": 1.9363874702485497, "learning_rate": 4.67373196841196e-06, "loss": 0.3602, "step": 5246 }, { "epoch": 1.706896551724138, "grad_norm": 1.7859739646663955, "learning_rate": 4.671843272099609e-06, "loss": 0.3529, "step": 5247 }, { "epoch": 1.7072218607677292, "grad_norm": 1.8574265032192854, "learning_rate": 4.669954622812335e-06, "loss": 0.3377, "step": 5248 }, { "epoch": 1.7075471698113207, "grad_norm": 1.8876583630978445, "learning_rate": 4.668066020820782e-06, "loss": 0.4015, "step": 5249 }, { "epoch": 1.7078724788549122, "grad_norm": 2.1815069073471953, "learning_rate": 4.666177466395592e-06, "loss": 0.4557, "step": 5250 }, { "epoch": 1.7081977878985035, "grad_norm": 1.7885908770521888, "learning_rate": 4.6642889598073925e-06, "loss": 0.4229, "step": 5251 }, { "epoch": 1.708523096942095, "grad_norm": 2.1118299930850983, "learning_rate": 4.66240050132681e-06, "loss": 0.4897, "step": 5252 }, { "epoch": 1.7088484059856865, "grad_norm": 1.6592256019972043, "learning_rate": 4.660512091224458e-06, "loss": 0.3401, "step": 5253 }, { "epoch": 1.7091737150292778, "grad_norm": 1.9544033350604348, "learning_rate": 4.658623729770952e-06, "loss": 0.3564, "step": 5254 }, { "epoch": 1.7094990240728691, "grad_norm": 1.6305034494227928, "learning_rate": 4.656735417236894e-06, "loss": 0.3127, "step": 5255 }, { "epoch": 1.7098243331164606, "grad_norm": 2.2597692569326946, "learning_rate": 4.654847153892878e-06, "loss": 0.5013, "step": 5256 }, { "epoch": 1.7101496421600522, "grad_norm": 1.6687840523087811, "learning_rate": 4.652958940009494e-06, "loss": 0.2688, "step": 5257 }, { "epoch": 1.7104749512036435, "grad_norm": 1.845364256750676, "learning_rate": 4.651070775857327e-06, "loss": 0.3434, "step": 5258 }, { "epoch": 1.7108002602472347, "grad_norm": 1.705356752263405, "learning_rate": 4.649182661706951e-06, "loss": 0.3053, "step": 5259 }, { "epoch": 1.7111255692908263, "grad_norm": 2.0447951832115296, "learning_rate": 4.647294597828932e-06, "loss": 0.3503, "step": 5260 }, { "epoch": 1.7114508783344178, "grad_norm": 1.6704483189364054, "learning_rate": 4.645406584493832e-06, "loss": 0.3075, "step": 5261 }, { "epoch": 1.711776187378009, "grad_norm": 2.022009880391087, "learning_rate": 4.643518621972206e-06, "loss": 0.3556, "step": 5262 }, { "epoch": 1.7121014964216004, "grad_norm": 2.10439992765317, "learning_rate": 4.6416307105345985e-06, "loss": 0.4325, "step": 5263 }, { "epoch": 1.7124268054651919, "grad_norm": 1.7285483968321735, "learning_rate": 4.63974285045155e-06, "loss": 0.3343, "step": 5264 }, { "epoch": 1.7127521145087834, "grad_norm": 1.9506598148111467, "learning_rate": 4.637855041993589e-06, "loss": 0.398, "step": 5265 }, { "epoch": 1.7130774235523747, "grad_norm": 2.057763518190211, "learning_rate": 4.635967285431244e-06, "loss": 0.41, "step": 5266 }, { "epoch": 1.7134027325959662, "grad_norm": 2.0014267785220703, "learning_rate": 4.634079581035029e-06, "loss": 0.4485, "step": 5267 }, { "epoch": 1.7137280416395577, "grad_norm": 1.9452964475794583, "learning_rate": 4.632191929075454e-06, "loss": 0.433, "step": 5268 }, { "epoch": 1.714053350683149, "grad_norm": 1.814828459687946, "learning_rate": 4.630304329823021e-06, "loss": 0.4089, "step": 5269 }, { "epoch": 1.7143786597267403, "grad_norm": 1.804604091246641, "learning_rate": 4.6284167835482235e-06, "loss": 0.3491, "step": 5270 }, { "epoch": 1.7147039687703318, "grad_norm": 1.9907872232623103, "learning_rate": 4.626529290521551e-06, "loss": 0.4032, "step": 5271 }, { "epoch": 1.7150292778139233, "grad_norm": 1.8112059610494922, "learning_rate": 4.624641851013476e-06, "loss": 0.3503, "step": 5272 }, { "epoch": 1.7153545868575146, "grad_norm": 2.0193459067916804, "learning_rate": 4.622754465294476e-06, "loss": 0.4553, "step": 5273 }, { "epoch": 1.715679895901106, "grad_norm": 1.8881100835945595, "learning_rate": 4.620867133635015e-06, "loss": 0.4485, "step": 5274 }, { "epoch": 1.7160052049446974, "grad_norm": 1.7373722168966228, "learning_rate": 4.618979856305544e-06, "loss": 0.3424, "step": 5275 }, { "epoch": 1.716330513988289, "grad_norm": 1.784749466845568, "learning_rate": 4.617092633576513e-06, "loss": 0.3578, "step": 5276 }, { "epoch": 1.7166558230318802, "grad_norm": 1.5875480395829749, "learning_rate": 4.615205465718365e-06, "loss": 0.2774, "step": 5277 }, { "epoch": 1.7169811320754715, "grad_norm": 1.8258372138459027, "learning_rate": 4.613318353001529e-06, "loss": 0.3665, "step": 5278 }, { "epoch": 1.7173064411190633, "grad_norm": 2.030202816636801, "learning_rate": 4.6114312956964315e-06, "loss": 0.3844, "step": 5279 }, { "epoch": 1.7176317501626546, "grad_norm": 2.079394943093674, "learning_rate": 4.609544294073488e-06, "loss": 0.4261, "step": 5280 }, { "epoch": 1.7179570592062459, "grad_norm": 2.1672817513158122, "learning_rate": 4.607657348403107e-06, "loss": 0.3793, "step": 5281 }, { "epoch": 1.7182823682498374, "grad_norm": 2.0421742896759754, "learning_rate": 4.6057704589556905e-06, "loss": 0.4585, "step": 5282 }, { "epoch": 1.718607677293429, "grad_norm": 2.0223177094350775, "learning_rate": 4.60388362600163e-06, "loss": 0.3572, "step": 5283 }, { "epoch": 1.7189329863370202, "grad_norm": 2.0165943786066545, "learning_rate": 4.601996849811309e-06, "loss": 0.3818, "step": 5284 }, { "epoch": 1.7192582953806115, "grad_norm": 1.9521691145793547, "learning_rate": 4.600110130655106e-06, "loss": 0.3728, "step": 5285 }, { "epoch": 1.719583604424203, "grad_norm": 1.96778532069796, "learning_rate": 4.598223468803388e-06, "loss": 0.3837, "step": 5286 }, { "epoch": 1.7199089134677945, "grad_norm": 2.180254545545451, "learning_rate": 4.596336864526515e-06, "loss": 0.4848, "step": 5287 }, { "epoch": 1.7202342225113858, "grad_norm": 2.2810810858331863, "learning_rate": 4.594450318094836e-06, "loss": 0.4149, "step": 5288 }, { "epoch": 1.720559531554977, "grad_norm": 1.7012808871096317, "learning_rate": 4.592563829778698e-06, "loss": 0.3347, "step": 5289 }, { "epoch": 1.7208848405985686, "grad_norm": 1.909246693505344, "learning_rate": 4.590677399848436e-06, "loss": 0.3881, "step": 5290 }, { "epoch": 1.7212101496421601, "grad_norm": 1.7239975867716242, "learning_rate": 4.588791028574376e-06, "loss": 0.347, "step": 5291 }, { "epoch": 1.7215354586857514, "grad_norm": 1.6561347020320245, "learning_rate": 4.586904716226833e-06, "loss": 0.3235, "step": 5292 }, { "epoch": 1.7218607677293427, "grad_norm": 1.8725594608793785, "learning_rate": 4.585018463076122e-06, "loss": 0.319, "step": 5293 }, { "epoch": 1.7221860767729344, "grad_norm": 1.7467081888598235, "learning_rate": 4.583132269392542e-06, "loss": 0.3409, "step": 5294 }, { "epoch": 1.7225113858165257, "grad_norm": 1.9801223115517845, "learning_rate": 4.581246135446384e-06, "loss": 0.38, "step": 5295 }, { "epoch": 1.722836694860117, "grad_norm": 1.6262488496364431, "learning_rate": 4.579360061507934e-06, "loss": 0.3094, "step": 5296 }, { "epoch": 1.7231620039037086, "grad_norm": 2.2337259080062237, "learning_rate": 4.5774740478474685e-06, "loss": 0.5391, "step": 5297 }, { "epoch": 1.7234873129473, "grad_norm": 2.0485171544770604, "learning_rate": 4.5755880947352535e-06, "loss": 0.4975, "step": 5298 }, { "epoch": 1.7238126219908914, "grad_norm": 1.892467092045332, "learning_rate": 4.573702202441548e-06, "loss": 0.2829, "step": 5299 }, { "epoch": 1.7241379310344827, "grad_norm": 1.9375068981149077, "learning_rate": 4.5718163712366e-06, "loss": 0.4176, "step": 5300 }, { "epoch": 1.7244632400780742, "grad_norm": 2.308515238663329, "learning_rate": 4.569930601390653e-06, "loss": 0.3639, "step": 5301 }, { "epoch": 1.7247885491216657, "grad_norm": 1.9612665418724013, "learning_rate": 4.568044893173936e-06, "loss": 0.3993, "step": 5302 }, { "epoch": 1.725113858165257, "grad_norm": 2.100950614476816, "learning_rate": 4.566159246856677e-06, "loss": 0.3967, "step": 5303 }, { "epoch": 1.7254391672088483, "grad_norm": 2.042884042965862, "learning_rate": 4.564273662709082e-06, "loss": 0.3992, "step": 5304 }, { "epoch": 1.7257644762524398, "grad_norm": 2.32260431113838, "learning_rate": 4.562388141001365e-06, "loss": 0.4031, "step": 5305 }, { "epoch": 1.7260897852960313, "grad_norm": 1.7166173647164775, "learning_rate": 4.560502682003719e-06, "loss": 0.3763, "step": 5306 }, { "epoch": 1.7264150943396226, "grad_norm": 1.9635689795403695, "learning_rate": 4.558617285986332e-06, "loss": 0.3544, "step": 5307 }, { "epoch": 1.726740403383214, "grad_norm": 1.6308073325132608, "learning_rate": 4.556731953219379e-06, "loss": 0.4151, "step": 5308 }, { "epoch": 1.7270657124268056, "grad_norm": 2.1130150935791137, "learning_rate": 4.554846683973037e-06, "loss": 0.4239, "step": 5309 }, { "epoch": 1.727391021470397, "grad_norm": 1.7640830911959373, "learning_rate": 4.5529614785174606e-06, "loss": 0.3473, "step": 5310 }, { "epoch": 1.7277163305139882, "grad_norm": 1.9932992804939298, "learning_rate": 4.5510763371228015e-06, "loss": 0.3712, "step": 5311 }, { "epoch": 1.7280416395575797, "grad_norm": 1.807066571840341, "learning_rate": 4.549191260059202e-06, "loss": 0.3867, "step": 5312 }, { "epoch": 1.7283669486011712, "grad_norm": 1.796381353033323, "learning_rate": 4.5473062475967965e-06, "loss": 0.3964, "step": 5313 }, { "epoch": 1.7286922576447625, "grad_norm": 2.061623581820646, "learning_rate": 4.545421300005708e-06, "loss": 0.4277, "step": 5314 }, { "epoch": 1.7290175666883538, "grad_norm": 2.2762975771776923, "learning_rate": 4.54353641755605e-06, "loss": 0.5375, "step": 5315 }, { "epoch": 1.7293428757319453, "grad_norm": 1.7776058985532548, "learning_rate": 4.541651600517927e-06, "loss": 0.3333, "step": 5316 }, { "epoch": 1.7296681847755369, "grad_norm": 1.83502621732875, "learning_rate": 4.539766849161436e-06, "loss": 0.3627, "step": 5317 }, { "epoch": 1.7299934938191281, "grad_norm": 1.9257112796526192, "learning_rate": 4.537882163756661e-06, "loss": 0.3771, "step": 5318 }, { "epoch": 1.7303188028627194, "grad_norm": 2.15021801057927, "learning_rate": 4.535997544573682e-06, "loss": 0.3999, "step": 5319 }, { "epoch": 1.730644111906311, "grad_norm": 2.001301677478061, "learning_rate": 4.534112991882562e-06, "loss": 0.3885, "step": 5320 }, { "epoch": 1.7309694209499025, "grad_norm": 1.9934762924122331, "learning_rate": 4.532228505953363e-06, "loss": 0.3804, "step": 5321 }, { "epoch": 1.7312947299934938, "grad_norm": 2.037544250099966, "learning_rate": 4.53034408705613e-06, "loss": 0.3703, "step": 5322 }, { "epoch": 1.7316200390370853, "grad_norm": 2.249674680717884, "learning_rate": 4.528459735460905e-06, "loss": 0.4423, "step": 5323 }, { "epoch": 1.7319453480806768, "grad_norm": 1.715806422477115, "learning_rate": 4.526575451437709e-06, "loss": 0.3384, "step": 5324 }, { "epoch": 1.732270657124268, "grad_norm": 1.8397324373867443, "learning_rate": 4.524691235256571e-06, "loss": 0.3644, "step": 5325 }, { "epoch": 1.7325959661678594, "grad_norm": 2.0500502775217555, "learning_rate": 4.522807087187494e-06, "loss": 0.3953, "step": 5326 }, { "epoch": 1.732921275211451, "grad_norm": 1.792020968088264, "learning_rate": 4.520923007500481e-06, "loss": 0.3747, "step": 5327 }, { "epoch": 1.7332465842550424, "grad_norm": 1.6334497427967685, "learning_rate": 4.519038996465518e-06, "loss": 0.3616, "step": 5328 }, { "epoch": 1.7335718932986337, "grad_norm": 1.6205238061802574, "learning_rate": 4.517155054352589e-06, "loss": 0.3356, "step": 5329 }, { "epoch": 1.733897202342225, "grad_norm": 2.2883364688915475, "learning_rate": 4.515271181431663e-06, "loss": 0.4347, "step": 5330 }, { "epoch": 1.7342225113858165, "grad_norm": 2.1042169749639115, "learning_rate": 4.5133873779727e-06, "loss": 0.4423, "step": 5331 }, { "epoch": 1.734547820429408, "grad_norm": 2.0740260809796203, "learning_rate": 4.511503644245648e-06, "loss": 0.365, "step": 5332 }, { "epoch": 1.7348731294729993, "grad_norm": 2.2174494813286665, "learning_rate": 4.509619980520451e-06, "loss": 0.3222, "step": 5333 }, { "epoch": 1.7351984385165906, "grad_norm": 2.185612869646451, "learning_rate": 4.507736387067038e-06, "loss": 0.4901, "step": 5334 }, { "epoch": 1.7355237475601821, "grad_norm": 1.8783905240864116, "learning_rate": 4.505852864155329e-06, "loss": 0.3883, "step": 5335 }, { "epoch": 1.7358490566037736, "grad_norm": 1.596548606668603, "learning_rate": 4.5039694120552325e-06, "loss": 0.272, "step": 5336 }, { "epoch": 1.736174365647365, "grad_norm": 2.0557343777942485, "learning_rate": 4.5020860310366514e-06, "loss": 0.4304, "step": 5337 }, { "epoch": 1.7364996746909565, "grad_norm": 2.1494832591485227, "learning_rate": 4.500202721369474e-06, "loss": 0.4385, "step": 5338 }, { "epoch": 1.736824983734548, "grad_norm": 1.7389152835167119, "learning_rate": 4.498319483323581e-06, "loss": 0.3587, "step": 5339 }, { "epoch": 1.7371502927781393, "grad_norm": 2.24954176356486, "learning_rate": 4.496436317168837e-06, "loss": 0.4349, "step": 5340 }, { "epoch": 1.7374756018217306, "grad_norm": 1.8612071828951162, "learning_rate": 4.4945532231751075e-06, "loss": 0.3067, "step": 5341 }, { "epoch": 1.737800910865322, "grad_norm": 2.1013393877922266, "learning_rate": 4.492670201612238e-06, "loss": 0.3966, "step": 5342 }, { "epoch": 1.7381262199089136, "grad_norm": 2.1218661241745056, "learning_rate": 4.4907872527500665e-06, "loss": 0.4633, "step": 5343 }, { "epoch": 1.7384515289525049, "grad_norm": 1.9445505375774652, "learning_rate": 4.48890437685842e-06, "loss": 0.4098, "step": 5344 }, { "epoch": 1.7387768379960962, "grad_norm": 1.551524867934054, "learning_rate": 4.4870215742071174e-06, "loss": 0.2605, "step": 5345 }, { "epoch": 1.7391021470396877, "grad_norm": 1.9885441727071826, "learning_rate": 4.485138845065966e-06, "loss": 0.4801, "step": 5346 }, { "epoch": 1.7394274560832792, "grad_norm": 1.778365982962061, "learning_rate": 4.483256189704762e-06, "loss": 0.2994, "step": 5347 }, { "epoch": 1.7397527651268705, "grad_norm": 1.7885643458227882, "learning_rate": 4.481373608393288e-06, "loss": 0.3739, "step": 5348 }, { "epoch": 1.7400780741704618, "grad_norm": 1.9275804243504189, "learning_rate": 4.479491101401324e-06, "loss": 0.3562, "step": 5349 }, { "epoch": 1.7404033832140533, "grad_norm": 1.5342220574496135, "learning_rate": 4.477608668998631e-06, "loss": 0.298, "step": 5350 }, { "epoch": 1.7407286922576448, "grad_norm": 1.7416244414595887, "learning_rate": 4.4757263114549656e-06, "loss": 0.316, "step": 5351 }, { "epoch": 1.741054001301236, "grad_norm": 1.8163334661517598, "learning_rate": 4.473844029040067e-06, "loss": 0.4525, "step": 5352 }, { "epoch": 1.7413793103448276, "grad_norm": 2.0450141761025917, "learning_rate": 4.4719618220236715e-06, "loss": 0.44, "step": 5353 }, { "epoch": 1.7417046193884191, "grad_norm": 2.2035979858268697, "learning_rate": 4.470079690675499e-06, "loss": 0.4767, "step": 5354 }, { "epoch": 1.7420299284320104, "grad_norm": 2.1962912368135585, "learning_rate": 4.468197635265261e-06, "loss": 0.5193, "step": 5355 }, { "epoch": 1.7423552374756017, "grad_norm": 1.8503056661252113, "learning_rate": 4.466315656062653e-06, "loss": 0.3778, "step": 5356 }, { "epoch": 1.7426805465191932, "grad_norm": 1.825862010710886, "learning_rate": 4.46443375333737e-06, "loss": 0.3833, "step": 5357 }, { "epoch": 1.7430058555627848, "grad_norm": 1.754163223856075, "learning_rate": 4.462551927359088e-06, "loss": 0.3438, "step": 5358 }, { "epoch": 1.743331164606376, "grad_norm": 1.795969502604162, "learning_rate": 4.4606701783974714e-06, "loss": 0.2945, "step": 5359 }, { "epoch": 1.7436564736499673, "grad_norm": 1.795115050054522, "learning_rate": 4.458788506722176e-06, "loss": 0.3419, "step": 5360 }, { "epoch": 1.7439817826935589, "grad_norm": 2.0422771319857387, "learning_rate": 4.456906912602852e-06, "loss": 0.3629, "step": 5361 }, { "epoch": 1.7443070917371504, "grad_norm": 1.8916872181114273, "learning_rate": 4.455025396309129e-06, "loss": 0.442, "step": 5362 }, { "epoch": 1.7446324007807417, "grad_norm": 2.139124347981232, "learning_rate": 4.4531439581106295e-06, "loss": 0.375, "step": 5363 }, { "epoch": 1.744957709824333, "grad_norm": 2.0471713986035294, "learning_rate": 4.451262598276964e-06, "loss": 0.4337, "step": 5364 }, { "epoch": 1.7452830188679245, "grad_norm": 1.786692380419444, "learning_rate": 4.449381317077736e-06, "loss": 0.3562, "step": 5365 }, { "epoch": 1.745608327911516, "grad_norm": 2.0751787130990826, "learning_rate": 4.447500114782532e-06, "loss": 0.4557, "step": 5366 }, { "epoch": 1.7459336369551073, "grad_norm": 1.8298637221548117, "learning_rate": 4.445618991660929e-06, "loss": 0.3203, "step": 5367 }, { "epoch": 1.7462589459986988, "grad_norm": 2.1301716911697324, "learning_rate": 4.443737947982494e-06, "loss": 0.4246, "step": 5368 }, { "epoch": 1.7465842550422903, "grad_norm": 1.8486931112337006, "learning_rate": 4.441856984016783e-06, "loss": 0.406, "step": 5369 }, { "epoch": 1.7469095640858816, "grad_norm": 1.885803193658349, "learning_rate": 4.4399761000333375e-06, "loss": 0.418, "step": 5370 }, { "epoch": 1.747234873129473, "grad_norm": 1.811654891282541, "learning_rate": 4.438095296301692e-06, "loss": 0.3392, "step": 5371 }, { "epoch": 1.7475601821730644, "grad_norm": 1.7156949571056228, "learning_rate": 4.436214573091362e-06, "loss": 0.3255, "step": 5372 }, { "epoch": 1.747885491216656, "grad_norm": 1.9023459525100943, "learning_rate": 4.4343339306718615e-06, "loss": 0.3189, "step": 5373 }, { "epoch": 1.7482108002602472, "grad_norm": 2.019235773135864, "learning_rate": 4.432453369312687e-06, "loss": 0.3797, "step": 5374 }, { "epoch": 1.7485361093038385, "grad_norm": 1.944409908375529, "learning_rate": 4.430572889283322e-06, "loss": 0.4147, "step": 5375 }, { "epoch": 1.74886141834743, "grad_norm": 2.056806188827285, "learning_rate": 4.428692490853239e-06, "loss": 0.4226, "step": 5376 }, { "epoch": 1.7491867273910215, "grad_norm": 2.0984056566232767, "learning_rate": 4.426812174291907e-06, "loss": 0.4118, "step": 5377 }, { "epoch": 1.7495120364346128, "grad_norm": 1.71825181328482, "learning_rate": 4.424931939868771e-06, "loss": 0.325, "step": 5378 }, { "epoch": 1.7498373454782041, "grad_norm": 2.0541308018205604, "learning_rate": 4.423051787853272e-06, "loss": 0.3839, "step": 5379 }, { "epoch": 1.7501626545217959, "grad_norm": 1.6992484411836273, "learning_rate": 4.421171718514835e-06, "loss": 0.3387, "step": 5380 }, { "epoch": 1.7504879635653872, "grad_norm": 1.874331500156052, "learning_rate": 4.419291732122878e-06, "loss": 0.3995, "step": 5381 }, { "epoch": 1.7508132726089785, "grad_norm": 2.087039450489764, "learning_rate": 4.417411828946803e-06, "loss": 0.4152, "step": 5382 }, { "epoch": 1.75113858165257, "grad_norm": 2.183547898452342, "learning_rate": 4.415532009256e-06, "loss": 0.4382, "step": 5383 }, { "epoch": 1.7514638906961615, "grad_norm": 1.8236850716049267, "learning_rate": 4.41365227331985e-06, "loss": 0.3903, "step": 5384 }, { "epoch": 1.7517891997397528, "grad_norm": 1.780654761230999, "learning_rate": 4.411772621407721e-06, "loss": 0.4033, "step": 5385 }, { "epoch": 1.752114508783344, "grad_norm": 1.8006320976760257, "learning_rate": 4.409893053788968e-06, "loss": 0.362, "step": 5386 }, { "epoch": 1.7524398178269356, "grad_norm": 1.832521036052629, "learning_rate": 4.408013570732933e-06, "loss": 0.2573, "step": 5387 }, { "epoch": 1.752765126870527, "grad_norm": 2.081470584201108, "learning_rate": 4.406134172508947e-06, "loss": 0.3108, "step": 5388 }, { "epoch": 1.7530904359141184, "grad_norm": 1.8873894456371973, "learning_rate": 4.404254859386332e-06, "loss": 0.404, "step": 5389 }, { "epoch": 1.7534157449577097, "grad_norm": 2.1984893624386186, "learning_rate": 4.402375631634393e-06, "loss": 0.3769, "step": 5390 }, { "epoch": 1.7537410540013012, "grad_norm": 2.098954450774147, "learning_rate": 4.400496489522424e-06, "loss": 0.4479, "step": 5391 }, { "epoch": 1.7540663630448927, "grad_norm": 1.9972078413740892, "learning_rate": 4.3986174333197055e-06, "loss": 0.4574, "step": 5392 }, { "epoch": 1.754391672088484, "grad_norm": 2.051643851942404, "learning_rate": 4.396738463295512e-06, "loss": 0.3997, "step": 5393 }, { "epoch": 1.7547169811320755, "grad_norm": 1.825851238368021, "learning_rate": 4.394859579719099e-06, "loss": 0.345, "step": 5394 }, { "epoch": 1.755042290175667, "grad_norm": 2.284651716157283, "learning_rate": 4.39298078285971e-06, "loss": 0.4151, "step": 5395 }, { "epoch": 1.7553675992192583, "grad_norm": 2.3315904865728987, "learning_rate": 4.391102072986581e-06, "loss": 0.4705, "step": 5396 }, { "epoch": 1.7556929082628496, "grad_norm": 1.8177384904304903, "learning_rate": 4.389223450368931e-06, "loss": 0.3539, "step": 5397 }, { "epoch": 1.7560182173064411, "grad_norm": 2.2112085813326354, "learning_rate": 4.3873449152759675e-06, "loss": 0.3943, "step": 5398 }, { "epoch": 1.7563435263500327, "grad_norm": 2.120561429402129, "learning_rate": 4.385466467976885e-06, "loss": 0.4047, "step": 5399 }, { "epoch": 1.756668835393624, "grad_norm": 1.814357560509441, "learning_rate": 4.383588108740868e-06, "loss": 0.331, "step": 5400 }, { "epoch": 1.7569941444372152, "grad_norm": 1.9165416027692217, "learning_rate": 4.381709837837087e-06, "loss": 0.3607, "step": 5401 }, { "epoch": 1.7573194534808068, "grad_norm": 1.946930786474713, "learning_rate": 4.379831655534698e-06, "loss": 0.4108, "step": 5402 }, { "epoch": 1.7576447625243983, "grad_norm": 1.9716260591821055, "learning_rate": 4.377953562102845e-06, "loss": 0.3617, "step": 5403 }, { "epoch": 1.7579700715679896, "grad_norm": 2.2594922479026107, "learning_rate": 4.376075557810663e-06, "loss": 0.4509, "step": 5404 }, { "epoch": 1.7582953806115809, "grad_norm": 1.8940907830247087, "learning_rate": 4.374197642927268e-06, "loss": 0.4435, "step": 5405 }, { "epoch": 1.7586206896551724, "grad_norm": 1.8513589790700473, "learning_rate": 4.372319817721769e-06, "loss": 0.3615, "step": 5406 }, { "epoch": 1.758945998698764, "grad_norm": 1.8155003819100515, "learning_rate": 4.370442082463259e-06, "loss": 0.3133, "step": 5407 }, { "epoch": 1.7592713077423552, "grad_norm": 2.0417077665497767, "learning_rate": 4.3685644374208174e-06, "loss": 0.384, "step": 5408 }, { "epoch": 1.7595966167859467, "grad_norm": 2.215734412918619, "learning_rate": 4.366686882863514e-06, "loss": 0.3888, "step": 5409 }, { "epoch": 1.7599219258295382, "grad_norm": 1.8361688122117605, "learning_rate": 4.364809419060403e-06, "loss": 0.3245, "step": 5410 }, { "epoch": 1.7602472348731295, "grad_norm": 1.7480933412650879, "learning_rate": 4.362932046280523e-06, "loss": 0.3586, "step": 5411 }, { "epoch": 1.7605725439167208, "grad_norm": 1.9965056052272845, "learning_rate": 4.361054764792908e-06, "loss": 0.5235, "step": 5412 }, { "epoch": 1.7608978529603123, "grad_norm": 1.9068242343785757, "learning_rate": 4.3591775748665705e-06, "loss": 0.345, "step": 5413 }, { "epoch": 1.7612231620039038, "grad_norm": 1.8381990648067057, "learning_rate": 4.357300476770514e-06, "loss": 0.3366, "step": 5414 }, { "epoch": 1.7615484710474951, "grad_norm": 1.963328771753703, "learning_rate": 4.355423470773726e-06, "loss": 0.3836, "step": 5415 }, { "epoch": 1.7618737800910864, "grad_norm": 1.9320871233416288, "learning_rate": 4.353546557145187e-06, "loss": 0.3365, "step": 5416 }, { "epoch": 1.762199089134678, "grad_norm": 1.7833630450067792, "learning_rate": 4.351669736153856e-06, "loss": 0.341, "step": 5417 }, { "epoch": 1.7625243981782694, "grad_norm": 2.2031715097271, "learning_rate": 4.349793008068684e-06, "loss": 0.4483, "step": 5418 }, { "epoch": 1.7628497072218607, "grad_norm": 1.6628863073268985, "learning_rate": 4.347916373158606e-06, "loss": 0.341, "step": 5419 }, { "epoch": 1.763175016265452, "grad_norm": 1.7034230667378458, "learning_rate": 4.346039831692549e-06, "loss": 0.3458, "step": 5420 }, { "epoch": 1.7635003253090435, "grad_norm": 1.8407509044233867, "learning_rate": 4.344163383939419e-06, "loss": 0.3099, "step": 5421 }, { "epoch": 1.763825634352635, "grad_norm": 2.2230182730159944, "learning_rate": 4.3422870301681145e-06, "loss": 0.4694, "step": 5422 }, { "epoch": 1.7641509433962264, "grad_norm": 1.9529133348354153, "learning_rate": 4.340410770647515e-06, "loss": 0.4063, "step": 5423 }, { "epoch": 1.7644762524398179, "grad_norm": 2.0564645910987007, "learning_rate": 4.338534605646493e-06, "loss": 0.4543, "step": 5424 }, { "epoch": 1.7648015614834094, "grad_norm": 1.7694185231061952, "learning_rate": 4.336658535433903e-06, "loss": 0.3595, "step": 5425 }, { "epoch": 1.7651268705270007, "grad_norm": 1.8690114627634766, "learning_rate": 4.334782560278589e-06, "loss": 0.3864, "step": 5426 }, { "epoch": 1.765452179570592, "grad_norm": 2.2228615947210217, "learning_rate": 4.332906680449374e-06, "loss": 0.4278, "step": 5427 }, { "epoch": 1.7657774886141835, "grad_norm": 1.8084297287401578, "learning_rate": 4.3310308962150786e-06, "loss": 0.3513, "step": 5428 }, { "epoch": 1.766102797657775, "grad_norm": 1.7781458796405452, "learning_rate": 4.329155207844502e-06, "loss": 0.3306, "step": 5429 }, { "epoch": 1.7664281067013663, "grad_norm": 1.6997321111141674, "learning_rate": 4.327279615606431e-06, "loss": 0.2748, "step": 5430 }, { "epoch": 1.7667534157449576, "grad_norm": 2.1293601813538197, "learning_rate": 4.325404119769639e-06, "loss": 0.3983, "step": 5431 }, { "epoch": 1.767078724788549, "grad_norm": 1.7678604727419154, "learning_rate": 4.323528720602887e-06, "loss": 0.3209, "step": 5432 }, { "epoch": 1.7674040338321406, "grad_norm": 1.9123830357004685, "learning_rate": 4.32165341837492e-06, "loss": 0.3522, "step": 5433 }, { "epoch": 1.767729342875732, "grad_norm": 1.97129067504362, "learning_rate": 4.319778213354472e-06, "loss": 0.3737, "step": 5434 }, { "epoch": 1.7680546519193232, "grad_norm": 1.7777069895800652, "learning_rate": 4.317903105810257e-06, "loss": 0.3376, "step": 5435 }, { "epoch": 1.7683799609629147, "grad_norm": 1.785910706378806, "learning_rate": 4.316028096010983e-06, "loss": 0.315, "step": 5436 }, { "epoch": 1.7687052700065062, "grad_norm": 2.1021473706375016, "learning_rate": 4.3141531842253384e-06, "loss": 0.3679, "step": 5437 }, { "epoch": 1.7690305790500975, "grad_norm": 2.3238588370168562, "learning_rate": 4.312278370722001e-06, "loss": 0.3717, "step": 5438 }, { "epoch": 1.769355888093689, "grad_norm": 1.9522846103850449, "learning_rate": 4.310403655769629e-06, "loss": 0.4312, "step": 5439 }, { "epoch": 1.7696811971372806, "grad_norm": 2.0607873321726746, "learning_rate": 4.308529039636876e-06, "loss": 0.4615, "step": 5440 }, { "epoch": 1.7700065061808719, "grad_norm": 1.904197121627503, "learning_rate": 4.306654522592371e-06, "loss": 0.4118, "step": 5441 }, { "epoch": 1.7703318152244631, "grad_norm": 2.1184758398614654, "learning_rate": 4.304780104904737e-06, "loss": 0.3386, "step": 5442 }, { "epoch": 1.7706571242680547, "grad_norm": 2.041805961444166, "learning_rate": 4.302905786842574e-06, "loss": 0.4725, "step": 5443 }, { "epoch": 1.7709824333116462, "grad_norm": 1.97774369202805, "learning_rate": 4.301031568674479e-06, "loss": 0.3278, "step": 5444 }, { "epoch": 1.7713077423552375, "grad_norm": 2.0040651304828954, "learning_rate": 4.299157450669026e-06, "loss": 0.4485, "step": 5445 }, { "epoch": 1.7716330513988288, "grad_norm": 1.7475196441173446, "learning_rate": 4.297283433094779e-06, "loss": 0.3201, "step": 5446 }, { "epoch": 1.7719583604424203, "grad_norm": 1.9883124918677781, "learning_rate": 4.29540951622028e-06, "loss": 0.4478, "step": 5447 }, { "epoch": 1.7722836694860118, "grad_norm": 2.004577316748654, "learning_rate": 4.293535700314071e-06, "loss": 0.3464, "step": 5448 }, { "epoch": 1.772608978529603, "grad_norm": 1.8236128057680128, "learning_rate": 4.291661985644667e-06, "loss": 0.2578, "step": 5449 }, { "epoch": 1.7729342875731944, "grad_norm": 1.8780550802230562, "learning_rate": 4.2897883724805715e-06, "loss": 0.3804, "step": 5450 }, { "epoch": 1.773259596616786, "grad_norm": 2.13025631068327, "learning_rate": 4.287914861090276e-06, "loss": 0.4125, "step": 5451 }, { "epoch": 1.7735849056603774, "grad_norm": 2.037383350850934, "learning_rate": 4.286041451742256e-06, "loss": 0.4036, "step": 5452 }, { "epoch": 1.7739102147039687, "grad_norm": 2.1545834139637803, "learning_rate": 4.284168144704971e-06, "loss": 0.4063, "step": 5453 }, { "epoch": 1.7742355237475602, "grad_norm": 1.7763931894639484, "learning_rate": 4.28229494024687e-06, "loss": 0.3635, "step": 5454 }, { "epoch": 1.7745608327911517, "grad_norm": 1.776359599399186, "learning_rate": 4.2804218386363805e-06, "loss": 0.3738, "step": 5455 }, { "epoch": 1.774886141834743, "grad_norm": 1.942159678712029, "learning_rate": 4.278548840141922e-06, "loss": 0.3355, "step": 5456 }, { "epoch": 1.7752114508783343, "grad_norm": 1.970261897091448, "learning_rate": 4.276675945031896e-06, "loss": 0.3262, "step": 5457 }, { "epoch": 1.7755367599219258, "grad_norm": 2.1767991480557525, "learning_rate": 4.274803153574691e-06, "loss": 0.4966, "step": 5458 }, { "epoch": 1.7758620689655173, "grad_norm": 2.0759862853781903, "learning_rate": 4.272930466038674e-06, "loss": 0.3838, "step": 5459 }, { "epoch": 1.7761873780091086, "grad_norm": 1.9845645474437084, "learning_rate": 4.271057882692208e-06, "loss": 0.4076, "step": 5460 }, { "epoch": 1.7765126870527, "grad_norm": 1.9769163085621735, "learning_rate": 4.269185403803635e-06, "loss": 0.402, "step": 5461 }, { "epoch": 1.7768379960962914, "grad_norm": 2.229116894245326, "learning_rate": 4.267313029641279e-06, "loss": 0.3908, "step": 5462 }, { "epoch": 1.777163305139883, "grad_norm": 1.9543969056736492, "learning_rate": 4.265440760473453e-06, "loss": 0.381, "step": 5463 }, { "epoch": 1.7774886141834743, "grad_norm": 2.0490357971021607, "learning_rate": 4.263568596568458e-06, "loss": 0.3757, "step": 5464 }, { "epoch": 1.7778139232270656, "grad_norm": 2.0381308920091694, "learning_rate": 4.261696538194574e-06, "loss": 0.4002, "step": 5465 }, { "epoch": 1.7781392322706573, "grad_norm": 1.994422049719301, "learning_rate": 4.259824585620068e-06, "loss": 0.3935, "step": 5466 }, { "epoch": 1.7784645413142486, "grad_norm": 2.3571148914849265, "learning_rate": 4.257952739113191e-06, "loss": 0.373, "step": 5467 }, { "epoch": 1.7787898503578399, "grad_norm": 2.1979024981157362, "learning_rate": 4.256080998942182e-06, "loss": 0.4218, "step": 5468 }, { "epoch": 1.7791151594014314, "grad_norm": 2.0412559793695264, "learning_rate": 4.254209365375263e-06, "loss": 0.4626, "step": 5469 }, { "epoch": 1.779440468445023, "grad_norm": 2.1971258171618566, "learning_rate": 4.2523378386806394e-06, "loss": 0.4211, "step": 5470 }, { "epoch": 1.7797657774886142, "grad_norm": 2.2473412899775527, "learning_rate": 4.2504664191265e-06, "loss": 0.4417, "step": 5471 }, { "epoch": 1.7800910865322055, "grad_norm": 2.3482889224458696, "learning_rate": 4.2485951069810235e-06, "loss": 0.4498, "step": 5472 }, { "epoch": 1.780416395575797, "grad_norm": 1.8091728433542775, "learning_rate": 4.246723902512369e-06, "loss": 0.3165, "step": 5473 }, { "epoch": 1.7807417046193885, "grad_norm": 1.9340854832572454, "learning_rate": 4.244852805988682e-06, "loss": 0.3245, "step": 5474 }, { "epoch": 1.7810670136629798, "grad_norm": 2.088397965321318, "learning_rate": 4.24298181767809e-06, "loss": 0.3365, "step": 5475 }, { "epoch": 1.781392322706571, "grad_norm": 2.025630281548363, "learning_rate": 4.241110937848709e-06, "loss": 0.4117, "step": 5476 }, { "epoch": 1.7817176317501626, "grad_norm": 2.102013871073822, "learning_rate": 4.239240166768639e-06, "loss": 0.4551, "step": 5477 }, { "epoch": 1.7820429407937541, "grad_norm": 1.7173004161662286, "learning_rate": 4.2373695047059585e-06, "loss": 0.3558, "step": 5478 }, { "epoch": 1.7823682498373454, "grad_norm": 2.1645351474022747, "learning_rate": 4.235498951928734e-06, "loss": 0.4605, "step": 5479 }, { "epoch": 1.7826935588809367, "grad_norm": 1.9959899899809408, "learning_rate": 4.233628508705023e-06, "loss": 0.4916, "step": 5480 }, { "epoch": 1.7830188679245285, "grad_norm": 1.9670434771683505, "learning_rate": 4.231758175302856e-06, "loss": 0.3173, "step": 5481 }, { "epoch": 1.7833441769681198, "grad_norm": 1.4961396534283042, "learning_rate": 4.229887951990255e-06, "loss": 0.2673, "step": 5482 }, { "epoch": 1.783669486011711, "grad_norm": 2.0582651055929477, "learning_rate": 4.228017839035223e-06, "loss": 0.3718, "step": 5483 }, { "epoch": 1.7839947950553026, "grad_norm": 2.143372815171192, "learning_rate": 4.226147836705751e-06, "loss": 0.3817, "step": 5484 }, { "epoch": 1.784320104098894, "grad_norm": 1.734409293932248, "learning_rate": 4.224277945269809e-06, "loss": 0.3701, "step": 5485 }, { "epoch": 1.7846454131424854, "grad_norm": 2.0801369836881487, "learning_rate": 4.222408164995355e-06, "loss": 0.4867, "step": 5486 }, { "epoch": 1.7849707221860767, "grad_norm": 1.8083245439828564, "learning_rate": 4.220538496150328e-06, "loss": 0.3506, "step": 5487 }, { "epoch": 1.7852960312296682, "grad_norm": 2.2067647798363126, "learning_rate": 4.2186689390026555e-06, "loss": 0.3634, "step": 5488 }, { "epoch": 1.7856213402732597, "grad_norm": 1.7137688279050391, "learning_rate": 4.216799493820245e-06, "loss": 0.279, "step": 5489 }, { "epoch": 1.785946649316851, "grad_norm": 2.050146124227589, "learning_rate": 4.21493016087099e-06, "loss": 0.4645, "step": 5490 }, { "epoch": 1.7862719583604423, "grad_norm": 1.7841220467078214, "learning_rate": 4.213060940422764e-06, "loss": 0.4439, "step": 5491 }, { "epoch": 1.7865972674040338, "grad_norm": 1.7097641165362985, "learning_rate": 4.211191832743432e-06, "loss": 0.38, "step": 5492 }, { "epoch": 1.7869225764476253, "grad_norm": 1.6542140823809866, "learning_rate": 4.209322838100836e-06, "loss": 0.3251, "step": 5493 }, { "epoch": 1.7872478854912166, "grad_norm": 1.8240270848433267, "learning_rate": 4.207453956762805e-06, "loss": 0.4107, "step": 5494 }, { "epoch": 1.7875731945348081, "grad_norm": 2.047972057300521, "learning_rate": 4.205585188997148e-06, "loss": 0.3768, "step": 5495 }, { "epoch": 1.7878985035783996, "grad_norm": 1.8195405705753498, "learning_rate": 4.203716535071665e-06, "loss": 0.3356, "step": 5496 }, { "epoch": 1.788223812621991, "grad_norm": 1.7621968248955542, "learning_rate": 4.2018479952541335e-06, "loss": 0.4088, "step": 5497 }, { "epoch": 1.7885491216655822, "grad_norm": 2.0122940292451967, "learning_rate": 4.199979569812316e-06, "loss": 0.5085, "step": 5498 }, { "epoch": 1.7888744307091737, "grad_norm": 1.812517613903262, "learning_rate": 4.198111259013959e-06, "loss": 0.4408, "step": 5499 }, { "epoch": 1.7891997397527653, "grad_norm": 1.966285694831251, "learning_rate": 4.196243063126794e-06, "loss": 0.3647, "step": 5500 }, { "epoch": 1.7895250487963565, "grad_norm": 1.9195027088691374, "learning_rate": 4.194374982418534e-06, "loss": 0.3413, "step": 5501 }, { "epoch": 1.7898503578399478, "grad_norm": 1.8990000379486747, "learning_rate": 4.192507017156876e-06, "loss": 0.3197, "step": 5502 }, { "epoch": 1.7901756668835394, "grad_norm": 1.7048986109690292, "learning_rate": 4.1906391676095e-06, "loss": 0.3382, "step": 5503 }, { "epoch": 1.7905009759271309, "grad_norm": 1.7806555324170426, "learning_rate": 4.188771434044072e-06, "loss": 0.2909, "step": 5504 }, { "epoch": 1.7908262849707222, "grad_norm": 1.8364716420006986, "learning_rate": 4.186903816728237e-06, "loss": 0.3532, "step": 5505 }, { "epoch": 1.7911515940143135, "grad_norm": 2.0876127096214723, "learning_rate": 4.185036315929629e-06, "loss": 0.3933, "step": 5506 }, { "epoch": 1.791476903057905, "grad_norm": 2.1877875861427913, "learning_rate": 4.183168931915859e-06, "loss": 0.3697, "step": 5507 }, { "epoch": 1.7918022121014965, "grad_norm": 1.755073046718192, "learning_rate": 4.181301664954526e-06, "loss": 0.3557, "step": 5508 }, { "epoch": 1.7921275211450878, "grad_norm": 1.8207509714272698, "learning_rate": 4.17943451531321e-06, "loss": 0.2887, "step": 5509 }, { "epoch": 1.7924528301886793, "grad_norm": 1.8977174435973612, "learning_rate": 4.177567483259477e-06, "loss": 0.3817, "step": 5510 }, { "epoch": 1.7927781392322708, "grad_norm": 1.7698879462236188, "learning_rate": 4.175700569060868e-06, "loss": 0.3323, "step": 5511 }, { "epoch": 1.793103448275862, "grad_norm": 1.932538141199549, "learning_rate": 4.17383377298492e-06, "loss": 0.3784, "step": 5512 }, { "epoch": 1.7934287573194534, "grad_norm": 2.1368002725800928, "learning_rate": 4.171967095299144e-06, "loss": 0.3292, "step": 5513 }, { "epoch": 1.793754066363045, "grad_norm": 1.9811328140848954, "learning_rate": 4.170100536271034e-06, "loss": 0.4551, "step": 5514 }, { "epoch": 1.7940793754066364, "grad_norm": 1.9949505569520545, "learning_rate": 4.1682340961680685e-06, "loss": 0.3725, "step": 5515 }, { "epoch": 1.7944046844502277, "grad_norm": 1.8542049692296152, "learning_rate": 4.166367775257715e-06, "loss": 0.2977, "step": 5516 }, { "epoch": 1.794729993493819, "grad_norm": 1.8239225257110445, "learning_rate": 4.164501573807413e-06, "loss": 0.3951, "step": 5517 }, { "epoch": 1.7950553025374105, "grad_norm": 1.8806860095168039, "learning_rate": 4.162635492084593e-06, "loss": 0.4192, "step": 5518 }, { "epoch": 1.795380611581002, "grad_norm": 2.0433390902183524, "learning_rate": 4.1607695303566655e-06, "loss": 0.3482, "step": 5519 }, { "epoch": 1.7957059206245933, "grad_norm": 1.8369110495137935, "learning_rate": 4.1589036888910235e-06, "loss": 0.3914, "step": 5520 }, { "epoch": 1.7960312296681846, "grad_norm": 1.8997050650031357, "learning_rate": 4.1570379679550436e-06, "loss": 0.3506, "step": 5521 }, { "epoch": 1.7963565387117761, "grad_norm": 1.9689315960603408, "learning_rate": 4.155172367816085e-06, "loss": 0.4483, "step": 5522 }, { "epoch": 1.7966818477553677, "grad_norm": 2.0530548253060257, "learning_rate": 4.153306888741489e-06, "loss": 0.4067, "step": 5523 }, { "epoch": 1.797007156798959, "grad_norm": 1.880648308581128, "learning_rate": 4.151441530998581e-06, "loss": 0.3497, "step": 5524 }, { "epoch": 1.7973324658425505, "grad_norm": 2.0853337847941567, "learning_rate": 4.149576294854668e-06, "loss": 0.421, "step": 5525 }, { "epoch": 1.797657774886142, "grad_norm": 1.992558345247525, "learning_rate": 4.147711180577037e-06, "loss": 0.4718, "step": 5526 }, { "epoch": 1.7979830839297333, "grad_norm": 1.8532458707452573, "learning_rate": 4.145846188432964e-06, "loss": 0.3858, "step": 5527 }, { "epoch": 1.7983083929733246, "grad_norm": 2.0557433839076125, "learning_rate": 4.143981318689701e-06, "loss": 0.4779, "step": 5528 }, { "epoch": 1.798633702016916, "grad_norm": 1.9932909674705737, "learning_rate": 4.1421165716144874e-06, "loss": 0.3466, "step": 5529 }, { "epoch": 1.7989590110605076, "grad_norm": 1.987180296841394, "learning_rate": 4.140251947474537e-06, "loss": 0.4021, "step": 5530 }, { "epoch": 1.7992843201040989, "grad_norm": 2.127406500437127, "learning_rate": 4.138387446537058e-06, "loss": 0.4096, "step": 5531 }, { "epoch": 1.7996096291476902, "grad_norm": 2.075917299117505, "learning_rate": 4.136523069069235e-06, "loss": 0.3878, "step": 5532 }, { "epoch": 1.7999349381912817, "grad_norm": 1.6210215991897983, "learning_rate": 4.1346588153382285e-06, "loss": 0.3236, "step": 5533 }, { "epoch": 1.8002602472348732, "grad_norm": 1.7427939861179826, "learning_rate": 4.13279468561119e-06, "loss": 0.3827, "step": 5534 }, { "epoch": 1.8005855562784645, "grad_norm": 2.272353081409726, "learning_rate": 4.130930680155254e-06, "loss": 0.4723, "step": 5535 }, { "epoch": 1.8009108653220558, "grad_norm": 1.8695306052800638, "learning_rate": 4.1290667992375296e-06, "loss": 0.3639, "step": 5536 }, { "epoch": 1.8012361743656473, "grad_norm": 2.125670478862597, "learning_rate": 4.127203043125115e-06, "loss": 0.3575, "step": 5537 }, { "epoch": 1.8015614834092388, "grad_norm": 2.1125397620294715, "learning_rate": 4.125339412085084e-06, "loss": 0.4225, "step": 5538 }, { "epoch": 1.8018867924528301, "grad_norm": 1.8631599704351345, "learning_rate": 4.1234759063845e-06, "loss": 0.4411, "step": 5539 }, { "epoch": 1.8022121014964216, "grad_norm": 2.1580387246652504, "learning_rate": 4.121612526290402e-06, "loss": 0.4346, "step": 5540 }, { "epoch": 1.8025374105400132, "grad_norm": 1.9935283791232588, "learning_rate": 4.119749272069815e-06, "loss": 0.322, "step": 5541 }, { "epoch": 1.8028627195836044, "grad_norm": 1.852680388544757, "learning_rate": 4.117886143989744e-06, "loss": 0.3104, "step": 5542 }, { "epoch": 1.8031880286271957, "grad_norm": 2.0522225959584013, "learning_rate": 4.116023142317177e-06, "loss": 0.3862, "step": 5543 }, { "epoch": 1.8035133376707873, "grad_norm": 1.732936629677962, "learning_rate": 4.1141602673190836e-06, "loss": 0.3716, "step": 5544 }, { "epoch": 1.8038386467143788, "grad_norm": 1.9820574406868934, "learning_rate": 4.112297519262415e-06, "loss": 0.3658, "step": 5545 }, { "epoch": 1.80416395575797, "grad_norm": 1.986629513464296, "learning_rate": 4.110434898414101e-06, "loss": 0.4113, "step": 5546 }, { "epoch": 1.8044892648015614, "grad_norm": 2.117438202853207, "learning_rate": 4.1085724050410604e-06, "loss": 0.3673, "step": 5547 }, { "epoch": 1.8048145738451529, "grad_norm": 2.2062415080420044, "learning_rate": 4.106710039410191e-06, "loss": 0.4872, "step": 5548 }, { "epoch": 1.8051398828887444, "grad_norm": 1.8173360182515406, "learning_rate": 4.104847801788367e-06, "loss": 0.3843, "step": 5549 }, { "epoch": 1.8054651919323357, "grad_norm": 1.8227819931566935, "learning_rate": 4.1029856924424465e-06, "loss": 0.352, "step": 5550 }, { "epoch": 1.805790500975927, "grad_norm": 1.7832905110257646, "learning_rate": 4.101123711639279e-06, "loss": 0.3441, "step": 5551 }, { "epoch": 1.8061158100195187, "grad_norm": 2.063259506588398, "learning_rate": 4.099261859645681e-06, "loss": 0.4973, "step": 5552 }, { "epoch": 1.80644111906311, "grad_norm": 1.9069012850950797, "learning_rate": 4.09740013672846e-06, "loss": 0.4837, "step": 5553 }, { "epoch": 1.8067664281067013, "grad_norm": 1.779355519584027, "learning_rate": 4.0955385431543994e-06, "loss": 0.3018, "step": 5554 }, { "epoch": 1.8070917371502928, "grad_norm": 1.8477407895232074, "learning_rate": 4.093677079190269e-06, "loss": 0.359, "step": 5555 }, { "epoch": 1.8074170461938843, "grad_norm": 1.6457965668629735, "learning_rate": 4.091815745102818e-06, "loss": 0.3149, "step": 5556 }, { "epoch": 1.8077423552374756, "grad_norm": 2.296718644157661, "learning_rate": 4.089954541158777e-06, "loss": 0.4316, "step": 5557 }, { "epoch": 1.808067664281067, "grad_norm": 2.0231772552880343, "learning_rate": 4.088093467624855e-06, "loss": 0.4289, "step": 5558 }, { "epoch": 1.8083929733246584, "grad_norm": 1.5060455315540913, "learning_rate": 4.086232524767749e-06, "loss": 0.2429, "step": 5559 }, { "epoch": 1.80871828236825, "grad_norm": 1.7577445761486201, "learning_rate": 4.08437171285413e-06, "loss": 0.3154, "step": 5560 }, { "epoch": 1.8090435914118412, "grad_norm": 1.7779245190781274, "learning_rate": 4.082511032150658e-06, "loss": 0.3411, "step": 5561 }, { "epoch": 1.8093689004554325, "grad_norm": 1.9798397349443189, "learning_rate": 4.0806504829239625e-06, "loss": 0.3908, "step": 5562 }, { "epoch": 1.809694209499024, "grad_norm": 2.035709078695766, "learning_rate": 4.0787900654406686e-06, "loss": 0.3797, "step": 5563 }, { "epoch": 1.8100195185426156, "grad_norm": 1.9745198486755982, "learning_rate": 4.076929779967372e-06, "loss": 0.373, "step": 5564 }, { "epoch": 1.8103448275862069, "grad_norm": 2.1686214929015413, "learning_rate": 4.075069626770654e-06, "loss": 0.4327, "step": 5565 }, { "epoch": 1.8106701366297981, "grad_norm": 2.0792529743297434, "learning_rate": 4.073209606117072e-06, "loss": 0.4259, "step": 5566 }, { "epoch": 1.8109954456733899, "grad_norm": 1.688307444986641, "learning_rate": 4.071349718273175e-06, "loss": 0.2498, "step": 5567 }, { "epoch": 1.8113207547169812, "grad_norm": 1.7065163566657617, "learning_rate": 4.069489963505482e-06, "loss": 0.3273, "step": 5568 }, { "epoch": 1.8116460637605725, "grad_norm": 1.8020232860209071, "learning_rate": 4.067630342080496e-06, "loss": 0.4006, "step": 5569 }, { "epoch": 1.811971372804164, "grad_norm": 2.0349802453501087, "learning_rate": 4.065770854264704e-06, "loss": 0.3862, "step": 5570 }, { "epoch": 1.8122966818477555, "grad_norm": 1.7600175926535169, "learning_rate": 4.063911500324572e-06, "loss": 0.3716, "step": 5571 }, { "epoch": 1.8126219908913468, "grad_norm": 1.8435166892528487, "learning_rate": 4.062052280526546e-06, "loss": 0.3768, "step": 5572 }, { "epoch": 1.812947299934938, "grad_norm": 1.9614292127710555, "learning_rate": 4.060193195137055e-06, "loss": 0.3384, "step": 5573 }, { "epoch": 1.8132726089785296, "grad_norm": 2.123997112146137, "learning_rate": 4.0583342444225025e-06, "loss": 0.3892, "step": 5574 }, { "epoch": 1.8135979180221211, "grad_norm": 1.983322840972921, "learning_rate": 4.056475428649283e-06, "loss": 0.3553, "step": 5575 }, { "epoch": 1.8139232270657124, "grad_norm": 2.216648060635333, "learning_rate": 4.0546167480837625e-06, "loss": 0.4444, "step": 5576 }, { "epoch": 1.8142485361093037, "grad_norm": 2.1584793275568894, "learning_rate": 4.052758202992294e-06, "loss": 0.4359, "step": 5577 }, { "epoch": 1.8145738451528952, "grad_norm": 2.043784778584342, "learning_rate": 4.050899793641204e-06, "loss": 0.3728, "step": 5578 }, { "epoch": 1.8148991541964867, "grad_norm": 1.934383838753633, "learning_rate": 4.049041520296808e-06, "loss": 0.3716, "step": 5579 }, { "epoch": 1.815224463240078, "grad_norm": 1.888920514178865, "learning_rate": 4.047183383225397e-06, "loss": 0.3602, "step": 5580 }, { "epoch": 1.8155497722836695, "grad_norm": 1.7907108206504954, "learning_rate": 4.045325382693242e-06, "loss": 0.3204, "step": 5581 }, { "epoch": 1.815875081327261, "grad_norm": 1.738949825623453, "learning_rate": 4.043467518966594e-06, "loss": 0.3766, "step": 5582 }, { "epoch": 1.8162003903708523, "grad_norm": 1.878862276133087, "learning_rate": 4.041609792311691e-06, "loss": 0.3896, "step": 5583 }, { "epoch": 1.8165256994144436, "grad_norm": 1.8177893481501506, "learning_rate": 4.039752202994743e-06, "loss": 0.3838, "step": 5584 }, { "epoch": 1.8168510084580352, "grad_norm": 1.821424277895658, "learning_rate": 4.037894751281944e-06, "loss": 0.3932, "step": 5585 }, { "epoch": 1.8171763175016267, "grad_norm": 1.9387255167023856, "learning_rate": 4.036037437439469e-06, "loss": 0.4456, "step": 5586 }, { "epoch": 1.817501626545218, "grad_norm": 1.9872213632685463, "learning_rate": 4.034180261733471e-06, "loss": 0.3389, "step": 5587 }, { "epoch": 1.8178269355888093, "grad_norm": 1.6191967040325126, "learning_rate": 4.032323224430087e-06, "loss": 0.2998, "step": 5588 }, { "epoch": 1.8181522446324008, "grad_norm": 2.040084942781481, "learning_rate": 4.030466325795429e-06, "loss": 0.437, "step": 5589 }, { "epoch": 1.8184775536759923, "grad_norm": 1.8253687072576257, "learning_rate": 4.028609566095593e-06, "loss": 0.3541, "step": 5590 }, { "epoch": 1.8188028627195836, "grad_norm": 2.0478151596823273, "learning_rate": 4.026752945596653e-06, "loss": 0.4145, "step": 5591 }, { "epoch": 1.8191281717631749, "grad_norm": 1.8750499874870354, "learning_rate": 4.024896464564666e-06, "loss": 0.376, "step": 5592 }, { "epoch": 1.8194534808067664, "grad_norm": 1.998039985042082, "learning_rate": 4.023040123265666e-06, "loss": 0.3875, "step": 5593 }, { "epoch": 1.819778789850358, "grad_norm": 1.693929045063291, "learning_rate": 4.021183921965667e-06, "loss": 0.3266, "step": 5594 }, { "epoch": 1.8201040988939492, "grad_norm": 2.1082313251590925, "learning_rate": 4.019327860930664e-06, "loss": 0.4198, "step": 5595 }, { "epoch": 1.8204294079375407, "grad_norm": 2.20053532007628, "learning_rate": 4.017471940426635e-06, "loss": 0.3889, "step": 5596 }, { "epoch": 1.8207547169811322, "grad_norm": 1.620703934402932, "learning_rate": 4.015616160719531e-06, "loss": 0.3183, "step": 5597 }, { "epoch": 1.8210800260247235, "grad_norm": 2.3540906665742067, "learning_rate": 4.0137605220752845e-06, "loss": 0.4817, "step": 5598 }, { "epoch": 1.8214053350683148, "grad_norm": 1.700249601593804, "learning_rate": 4.011905024759815e-06, "loss": 0.2956, "step": 5599 }, { "epoch": 1.8217306441119063, "grad_norm": 2.237885195050128, "learning_rate": 4.010049669039015e-06, "loss": 0.4455, "step": 5600 }, { "epoch": 1.8220559531554978, "grad_norm": 1.960905656454024, "learning_rate": 4.008194455178758e-06, "loss": 0.3991, "step": 5601 }, { "epoch": 1.8223812621990891, "grad_norm": 2.1119673543668505, "learning_rate": 4.006339383444894e-06, "loss": 0.3936, "step": 5602 }, { "epoch": 1.8227065712426804, "grad_norm": 2.1509220306755936, "learning_rate": 4.004484454103261e-06, "loss": 0.4058, "step": 5603 }, { "epoch": 1.823031880286272, "grad_norm": 1.7002797314225822, "learning_rate": 4.002629667419668e-06, "loss": 0.3088, "step": 5604 }, { "epoch": 1.8233571893298635, "grad_norm": 1.912125488158941, "learning_rate": 4.000775023659911e-06, "loss": 0.3333, "step": 5605 }, { "epoch": 1.8236824983734548, "grad_norm": 2.0647696176545356, "learning_rate": 3.998920523089757e-06, "loss": 0.3058, "step": 5606 }, { "epoch": 1.824007807417046, "grad_norm": 2.0520354122790394, "learning_rate": 3.99706616597496e-06, "loss": 0.3759, "step": 5607 }, { "epoch": 1.8243331164606376, "grad_norm": 2.0284881090083893, "learning_rate": 3.9952119525812506e-06, "loss": 0.3299, "step": 5608 }, { "epoch": 1.824658425504229, "grad_norm": 1.9715826471486348, "learning_rate": 3.993357883174338e-06, "loss": 0.3868, "step": 5609 }, { "epoch": 1.8249837345478204, "grad_norm": 1.554942258872616, "learning_rate": 3.991503958019911e-06, "loss": 0.2974, "step": 5610 }, { "epoch": 1.8253090435914119, "grad_norm": 1.7529049037785158, "learning_rate": 3.989650177383641e-06, "loss": 0.3426, "step": 5611 }, { "epoch": 1.8256343526350034, "grad_norm": 1.7701990424180016, "learning_rate": 3.987796541531174e-06, "loss": 0.3586, "step": 5612 }, { "epoch": 1.8259596616785947, "grad_norm": 1.945575197514644, "learning_rate": 3.985943050728138e-06, "loss": 0.3917, "step": 5613 }, { "epoch": 1.826284970722186, "grad_norm": 2.0822496552253047, "learning_rate": 3.9840897052401355e-06, "loss": 0.3841, "step": 5614 }, { "epoch": 1.8266102797657775, "grad_norm": 2.297198199090189, "learning_rate": 3.982236505332758e-06, "loss": 0.5, "step": 5615 }, { "epoch": 1.826935588809369, "grad_norm": 1.748045855972482, "learning_rate": 3.980383451271569e-06, "loss": 0.344, "step": 5616 }, { "epoch": 1.8272608978529603, "grad_norm": 1.900265003962877, "learning_rate": 3.97853054332211e-06, "loss": 0.4183, "step": 5617 }, { "epoch": 1.8275862068965516, "grad_norm": 1.6702069641268724, "learning_rate": 3.976677781749904e-06, "loss": 0.2661, "step": 5618 }, { "epoch": 1.8279115159401431, "grad_norm": 1.930601325717023, "learning_rate": 3.974825166820456e-06, "loss": 0.323, "step": 5619 }, { "epoch": 1.8282368249837346, "grad_norm": 2.121867899546261, "learning_rate": 3.972972698799245e-06, "loss": 0.385, "step": 5620 }, { "epoch": 1.828562134027326, "grad_norm": 1.903951469494103, "learning_rate": 3.97112037795173e-06, "loss": 0.403, "step": 5621 }, { "epoch": 1.8288874430709172, "grad_norm": 2.0238985608799447, "learning_rate": 3.969268204543351e-06, "loss": 0.3288, "step": 5622 }, { "epoch": 1.8292127521145087, "grad_norm": 2.0972450732876258, "learning_rate": 3.967416178839526e-06, "loss": 0.3504, "step": 5623 }, { "epoch": 1.8295380611581002, "grad_norm": 1.9078781817858819, "learning_rate": 3.965564301105651e-06, "loss": 0.3751, "step": 5624 }, { "epoch": 1.8298633702016915, "grad_norm": 2.068679898145419, "learning_rate": 3.963712571607101e-06, "loss": 0.4243, "step": 5625 }, { "epoch": 1.830188679245283, "grad_norm": 2.237177583739506, "learning_rate": 3.96186099060923e-06, "loss": 0.4569, "step": 5626 }, { "epoch": 1.8305139882888746, "grad_norm": 2.1841571228514836, "learning_rate": 3.960009558377372e-06, "loss": 0.4775, "step": 5627 }, { "epoch": 1.8308392973324659, "grad_norm": 1.8603032336911856, "learning_rate": 3.958158275176838e-06, "loss": 0.3208, "step": 5628 }, { "epoch": 1.8311646063760572, "grad_norm": 2.067662899604649, "learning_rate": 3.956307141272918e-06, "loss": 0.3735, "step": 5629 }, { "epoch": 1.8314899154196487, "grad_norm": 2.109499229756992, "learning_rate": 3.954456156930878e-06, "loss": 0.4517, "step": 5630 }, { "epoch": 1.8318152244632402, "grad_norm": 1.963659167295568, "learning_rate": 3.95260532241597e-06, "loss": 0.433, "step": 5631 }, { "epoch": 1.8321405335068315, "grad_norm": 1.8282232190674048, "learning_rate": 3.950754637993418e-06, "loss": 0.3016, "step": 5632 }, { "epoch": 1.8324658425504228, "grad_norm": 1.8012239801011494, "learning_rate": 3.948904103928426e-06, "loss": 0.3824, "step": 5633 }, { "epoch": 1.8327911515940143, "grad_norm": 1.92448407386118, "learning_rate": 3.947053720486174e-06, "loss": 0.4834, "step": 5634 }, { "epoch": 1.8331164606376058, "grad_norm": 1.8511702751871555, "learning_rate": 3.945203487931828e-06, "loss": 0.3238, "step": 5635 }, { "epoch": 1.833441769681197, "grad_norm": 1.8317180442874725, "learning_rate": 3.9433534065305255e-06, "loss": 0.295, "step": 5636 }, { "epoch": 1.8337670787247884, "grad_norm": 1.8044653749753832, "learning_rate": 3.941503476547385e-06, "loss": 0.2717, "step": 5637 }, { "epoch": 1.83409238776838, "grad_norm": 1.8815043144014791, "learning_rate": 3.9396536982475e-06, "loss": 0.3942, "step": 5638 }, { "epoch": 1.8344176968119714, "grad_norm": 1.9083818899650158, "learning_rate": 3.937804071895947e-06, "loss": 0.3697, "step": 5639 }, { "epoch": 1.8347430058555627, "grad_norm": 2.207464223472285, "learning_rate": 3.935954597757781e-06, "loss": 0.3693, "step": 5640 }, { "epoch": 1.8350683148991542, "grad_norm": 1.5443369908601248, "learning_rate": 3.934105276098029e-06, "loss": 0.28, "step": 5641 }, { "epoch": 1.8353936239427457, "grad_norm": 2.0811841621725877, "learning_rate": 3.932256107181701e-06, "loss": 0.3447, "step": 5642 }, { "epoch": 1.835718932986337, "grad_norm": 2.1842929526902037, "learning_rate": 3.9304070912737865e-06, "loss": 0.4469, "step": 5643 }, { "epoch": 1.8360442420299283, "grad_norm": 1.7426761815354908, "learning_rate": 3.928558228639248e-06, "loss": 0.3796, "step": 5644 }, { "epoch": 1.8363695510735198, "grad_norm": 2.0151170111155143, "learning_rate": 3.92670951954303e-06, "loss": 0.4713, "step": 5645 }, { "epoch": 1.8366948601171114, "grad_norm": 2.007314050161128, "learning_rate": 3.924860964250052e-06, "loss": 0.3468, "step": 5646 }, { "epoch": 1.8370201691607027, "grad_norm": 1.9561140018622503, "learning_rate": 3.9230125630252165e-06, "loss": 0.2899, "step": 5647 }, { "epoch": 1.837345478204294, "grad_norm": 1.9853344729759097, "learning_rate": 3.9211643161334e-06, "loss": 0.347, "step": 5648 }, { "epoch": 1.8376707872478855, "grad_norm": 2.0230012214915365, "learning_rate": 3.919316223839452e-06, "loss": 0.4543, "step": 5649 }, { "epoch": 1.837996096291477, "grad_norm": 1.8079352194833749, "learning_rate": 3.917468286408213e-06, "loss": 0.3773, "step": 5650 }, { "epoch": 1.8383214053350683, "grad_norm": 2.229264908687081, "learning_rate": 3.91562050410449e-06, "loss": 0.3848, "step": 5651 }, { "epoch": 1.8386467143786596, "grad_norm": 1.8506079417184673, "learning_rate": 3.913772877193071e-06, "loss": 0.3494, "step": 5652 }, { "epoch": 1.8389720234222513, "grad_norm": 1.9426095879317784, "learning_rate": 3.911925405938721e-06, "loss": 0.4143, "step": 5653 }, { "epoch": 1.8392973324658426, "grad_norm": 1.849477639420354, "learning_rate": 3.91007809060619e-06, "loss": 0.3168, "step": 5654 }, { "epoch": 1.8396226415094339, "grad_norm": 2.2861519986634917, "learning_rate": 3.908230931460193e-06, "loss": 0.429, "step": 5655 }, { "epoch": 1.8399479505530254, "grad_norm": 1.8415741130701269, "learning_rate": 3.906383928765432e-06, "loss": 0.2953, "step": 5656 }, { "epoch": 1.840273259596617, "grad_norm": 1.7502294680503905, "learning_rate": 3.904537082786581e-06, "loss": 0.298, "step": 5657 }, { "epoch": 1.8405985686402082, "grad_norm": 1.9316620626557341, "learning_rate": 3.902690393788299e-06, "loss": 0.3765, "step": 5658 }, { "epoch": 1.8409238776837995, "grad_norm": 1.773432621053558, "learning_rate": 3.900843862035215e-06, "loss": 0.288, "step": 5659 }, { "epoch": 1.841249186727391, "grad_norm": 1.9468372195287915, "learning_rate": 3.898997487791938e-06, "loss": 0.4331, "step": 5660 }, { "epoch": 1.8415744957709825, "grad_norm": 1.9954127347242079, "learning_rate": 3.897151271323054e-06, "loss": 0.3111, "step": 5661 }, { "epoch": 1.8418998048145738, "grad_norm": 1.8783925508068238, "learning_rate": 3.895305212893129e-06, "loss": 0.4215, "step": 5662 }, { "epoch": 1.8422251138581651, "grad_norm": 1.8838660969626684, "learning_rate": 3.893459312766704e-06, "loss": 0.4479, "step": 5663 }, { "epoch": 1.8425504229017566, "grad_norm": 2.159517674327206, "learning_rate": 3.891613571208297e-06, "loss": 0.3271, "step": 5664 }, { "epoch": 1.8428757319453482, "grad_norm": 1.9127651877899672, "learning_rate": 3.889767988482402e-06, "loss": 0.3673, "step": 5665 }, { "epoch": 1.8432010409889394, "grad_norm": 1.7714934701951903, "learning_rate": 3.887922564853496e-06, "loss": 0.3241, "step": 5666 }, { "epoch": 1.843526350032531, "grad_norm": 2.0556563710146265, "learning_rate": 3.88607730058603e-06, "loss": 0.4209, "step": 5667 }, { "epoch": 1.8438516590761225, "grad_norm": 1.8732170047315846, "learning_rate": 3.884232195944429e-06, "loss": 0.3059, "step": 5668 }, { "epoch": 1.8441769681197138, "grad_norm": 1.8611502088773082, "learning_rate": 3.882387251193095e-06, "loss": 0.3435, "step": 5669 }, { "epoch": 1.844502277163305, "grad_norm": 1.8053817310148714, "learning_rate": 3.880542466596416e-06, "loss": 0.3678, "step": 5670 }, { "epoch": 1.8448275862068966, "grad_norm": 2.2908727389671943, "learning_rate": 3.878697842418748e-06, "loss": 0.4781, "step": 5671 }, { "epoch": 1.845152895250488, "grad_norm": 1.7205540669115296, "learning_rate": 3.876853378924426e-06, "loss": 0.2855, "step": 5672 }, { "epoch": 1.8454782042940794, "grad_norm": 1.8146886169500487, "learning_rate": 3.875009076377763e-06, "loss": 0.3685, "step": 5673 }, { "epoch": 1.8458035133376707, "grad_norm": 2.049726936671566, "learning_rate": 3.87316493504305e-06, "loss": 0.458, "step": 5674 }, { "epoch": 1.8461288223812622, "grad_norm": 2.113900683731171, "learning_rate": 3.871320955184553e-06, "loss": 0.469, "step": 5675 }, { "epoch": 1.8464541314248537, "grad_norm": 1.9842800522976092, "learning_rate": 3.869477137066515e-06, "loss": 0.4538, "step": 5676 }, { "epoch": 1.846779440468445, "grad_norm": 1.7731695106354899, "learning_rate": 3.867633480953157e-06, "loss": 0.3628, "step": 5677 }, { "epoch": 1.8471047495120363, "grad_norm": 2.0893550822413265, "learning_rate": 3.865789987108677e-06, "loss": 0.4298, "step": 5678 }, { "epoch": 1.8474300585556278, "grad_norm": 2.360052092785287, "learning_rate": 3.8639466557972485e-06, "loss": 0.4191, "step": 5679 }, { "epoch": 1.8477553675992193, "grad_norm": 1.7437465453971777, "learning_rate": 3.862103487283021e-06, "loss": 0.3633, "step": 5680 }, { "epoch": 1.8480806766428106, "grad_norm": 1.8077310553061736, "learning_rate": 3.860260481830121e-06, "loss": 0.2885, "step": 5681 }, { "epoch": 1.8484059856864021, "grad_norm": 2.2886901662752153, "learning_rate": 3.8584176397026565e-06, "loss": 0.5134, "step": 5682 }, { "epoch": 1.8487312947299936, "grad_norm": 1.8596348703261893, "learning_rate": 3.8565749611647045e-06, "loss": 0.389, "step": 5683 }, { "epoch": 1.849056603773585, "grad_norm": 2.0652268935846676, "learning_rate": 3.8547324464803235e-06, "loss": 0.362, "step": 5684 }, { "epoch": 1.8493819128171762, "grad_norm": 2.0836709972187797, "learning_rate": 3.852890095913544e-06, "loss": 0.4245, "step": 5685 }, { "epoch": 1.8497072218607677, "grad_norm": 1.7191666519579318, "learning_rate": 3.851047909728381e-06, "loss": 0.3948, "step": 5686 }, { "epoch": 1.8500325309043593, "grad_norm": 1.9094814962243893, "learning_rate": 3.8492058881888186e-06, "loss": 0.4095, "step": 5687 }, { "epoch": 1.8503578399479506, "grad_norm": 1.89827377608854, "learning_rate": 3.847364031558819e-06, "loss": 0.2897, "step": 5688 }, { "epoch": 1.8506831489915418, "grad_norm": 1.6948109574427759, "learning_rate": 3.845522340102321e-06, "loss": 0.2654, "step": 5689 }, { "epoch": 1.8510084580351334, "grad_norm": 1.8963789193821188, "learning_rate": 3.843680814083242e-06, "loss": 0.4083, "step": 5690 }, { "epoch": 1.8513337670787249, "grad_norm": 2.0828948154223275, "learning_rate": 3.841839453765475e-06, "loss": 0.3868, "step": 5691 }, { "epoch": 1.8516590761223162, "grad_norm": 1.8700390456111182, "learning_rate": 3.839998259412885e-06, "loss": 0.4082, "step": 5692 }, { "epoch": 1.8519843851659075, "grad_norm": 2.0504801005626216, "learning_rate": 3.838157231289316e-06, "loss": 0.4105, "step": 5693 }, { "epoch": 1.852309694209499, "grad_norm": 1.94547558752628, "learning_rate": 3.836316369658594e-06, "loss": 0.3659, "step": 5694 }, { "epoch": 1.8526350032530905, "grad_norm": 1.7440853224493746, "learning_rate": 3.83447567478451e-06, "loss": 0.3882, "step": 5695 }, { "epoch": 1.8529603122966818, "grad_norm": 2.0090703332332596, "learning_rate": 3.83263514693084e-06, "loss": 0.3941, "step": 5696 }, { "epoch": 1.8532856213402733, "grad_norm": 2.0984055888382422, "learning_rate": 3.83079478636133e-06, "loss": 0.3624, "step": 5697 }, { "epoch": 1.8536109303838648, "grad_norm": 2.08154108193629, "learning_rate": 3.828954593339708e-06, "loss": 0.3769, "step": 5698 }, { "epoch": 1.8539362394274561, "grad_norm": 1.9864082627659516, "learning_rate": 3.827114568129674e-06, "loss": 0.3448, "step": 5699 }, { "epoch": 1.8542615484710474, "grad_norm": 1.9761602253671289, "learning_rate": 3.825274710994904e-06, "loss": 0.4212, "step": 5700 }, { "epoch": 1.854586857514639, "grad_norm": 1.9553004010058985, "learning_rate": 3.823435022199049e-06, "loss": 0.4071, "step": 5701 }, { "epoch": 1.8549121665582304, "grad_norm": 1.7856926447768038, "learning_rate": 3.821595502005742e-06, "loss": 0.3483, "step": 5702 }, { "epoch": 1.8552374756018217, "grad_norm": 1.926493832961955, "learning_rate": 3.819756150678587e-06, "loss": 0.3762, "step": 5703 }, { "epoch": 1.855562784645413, "grad_norm": 2.1874642992352022, "learning_rate": 3.8179169684811605e-06, "loss": 0.3389, "step": 5704 }, { "epoch": 1.8558880936890045, "grad_norm": 1.6949962568899333, "learning_rate": 3.816077955677018e-06, "loss": 0.3213, "step": 5705 }, { "epoch": 1.856213402732596, "grad_norm": 2.198656439007266, "learning_rate": 3.8142391125296974e-06, "loss": 0.4263, "step": 5706 }, { "epoch": 1.8565387117761873, "grad_norm": 1.846954259144254, "learning_rate": 3.8124004393027013e-06, "loss": 0.3786, "step": 5707 }, { "epoch": 1.8568640208197786, "grad_norm": 2.1177425092901156, "learning_rate": 3.8105619362595148e-06, "loss": 0.4065, "step": 5708 }, { "epoch": 1.8571893298633702, "grad_norm": 2.0956003814806174, "learning_rate": 3.8087236036635942e-06, "loss": 0.4561, "step": 5709 }, { "epoch": 1.8575146389069617, "grad_norm": 2.062591464052356, "learning_rate": 3.8068854417783766e-06, "loss": 0.3958, "step": 5710 }, { "epoch": 1.857839947950553, "grad_norm": 1.9002004650560536, "learning_rate": 3.8050474508672706e-06, "loss": 0.3759, "step": 5711 }, { "epoch": 1.8581652569941445, "grad_norm": 2.0402091321184064, "learning_rate": 3.8032096311936615e-06, "loss": 0.3725, "step": 5712 }, { "epoch": 1.858490566037736, "grad_norm": 2.0076369150178697, "learning_rate": 3.80137198302091e-06, "loss": 0.3649, "step": 5713 }, { "epoch": 1.8588158750813273, "grad_norm": 2.004045439093732, "learning_rate": 3.799534506612353e-06, "loss": 0.3994, "step": 5714 }, { "epoch": 1.8591411841249186, "grad_norm": 2.070940354205855, "learning_rate": 3.7976972022313013e-06, "loss": 0.4244, "step": 5715 }, { "epoch": 1.85946649316851, "grad_norm": 2.0791231911794115, "learning_rate": 3.7958600701410434e-06, "loss": 0.3711, "step": 5716 }, { "epoch": 1.8597918022121016, "grad_norm": 2.0213139884439544, "learning_rate": 3.7940231106048368e-06, "loss": 0.3921, "step": 5717 }, { "epoch": 1.860117111255693, "grad_norm": 1.7903753042946995, "learning_rate": 3.7921863238859252e-06, "loss": 0.3455, "step": 5718 }, { "epoch": 1.8604424202992842, "grad_norm": 2.0174107666641428, "learning_rate": 3.7903497102475196e-06, "loss": 0.3871, "step": 5719 }, { "epoch": 1.8607677293428757, "grad_norm": 1.8334693411211627, "learning_rate": 3.788513269952806e-06, "loss": 0.4062, "step": 5720 }, { "epoch": 1.8610930383864672, "grad_norm": 1.885569116117417, "learning_rate": 3.7866770032649458e-06, "loss": 0.3348, "step": 5721 }, { "epoch": 1.8614183474300585, "grad_norm": 1.9787557406098335, "learning_rate": 3.784840910447084e-06, "loss": 0.4892, "step": 5722 }, { "epoch": 1.8617436564736498, "grad_norm": 2.028071937183886, "learning_rate": 3.7830049917623284e-06, "loss": 0.3812, "step": 5723 }, { "epoch": 1.8620689655172413, "grad_norm": 1.8215987805511566, "learning_rate": 3.7811692474737695e-06, "loss": 0.3174, "step": 5724 }, { "epoch": 1.8623942745608328, "grad_norm": 1.7405763439901523, "learning_rate": 3.7793336778444678e-06, "loss": 0.2939, "step": 5725 }, { "epoch": 1.8627195836044241, "grad_norm": 1.7576206071583513, "learning_rate": 3.777498283137466e-06, "loss": 0.2773, "step": 5726 }, { "epoch": 1.8630448926480156, "grad_norm": 1.8823534301178146, "learning_rate": 3.775663063615775e-06, "loss": 0.3805, "step": 5727 }, { "epoch": 1.8633702016916072, "grad_norm": 1.7457228437154149, "learning_rate": 3.773828019542383e-06, "loss": 0.3701, "step": 5728 }, { "epoch": 1.8636955107351985, "grad_norm": 1.8952874627775842, "learning_rate": 3.7719931511802527e-06, "loss": 0.3731, "step": 5729 }, { "epoch": 1.8640208197787898, "grad_norm": 1.9641076618356632, "learning_rate": 3.7701584587923233e-06, "loss": 0.3438, "step": 5730 }, { "epoch": 1.8643461288223813, "grad_norm": 1.9296072127634816, "learning_rate": 3.768323942641507e-06, "loss": 0.4037, "step": 5731 }, { "epoch": 1.8646714378659728, "grad_norm": 2.1918375223663924, "learning_rate": 3.7664896029906906e-06, "loss": 0.4356, "step": 5732 }, { "epoch": 1.864996746909564, "grad_norm": 1.9993095732545114, "learning_rate": 3.7646554401027356e-06, "loss": 0.3485, "step": 5733 }, { "epoch": 1.8653220559531554, "grad_norm": 1.7413281511988488, "learning_rate": 3.762821454240481e-06, "loss": 0.2763, "step": 5734 }, { "epoch": 1.8656473649967469, "grad_norm": 1.999676919412539, "learning_rate": 3.760987645666738e-06, "loss": 0.3941, "step": 5735 }, { "epoch": 1.8659726740403384, "grad_norm": 2.3175145147381118, "learning_rate": 3.75915401464429e-06, "loss": 0.4375, "step": 5736 }, { "epoch": 1.8662979830839297, "grad_norm": 1.8734104882045324, "learning_rate": 3.757320561435897e-06, "loss": 0.3426, "step": 5737 }, { "epoch": 1.866623292127521, "grad_norm": 2.0604977220656244, "learning_rate": 3.755487286304299e-06, "loss": 0.3872, "step": 5738 }, { "epoch": 1.8669486011711127, "grad_norm": 1.946225273525884, "learning_rate": 3.753654189512202e-06, "loss": 0.4157, "step": 5739 }, { "epoch": 1.867273910214704, "grad_norm": 1.6664344876554662, "learning_rate": 3.7518212713222905e-06, "loss": 0.3313, "step": 5740 }, { "epoch": 1.8675992192582953, "grad_norm": 2.276870111159591, "learning_rate": 3.7499885319972217e-06, "loss": 0.3684, "step": 5741 }, { "epoch": 1.8679245283018868, "grad_norm": 1.6921825561450836, "learning_rate": 3.7481559717996313e-06, "loss": 0.3211, "step": 5742 }, { "epoch": 1.8682498373454783, "grad_norm": 2.150400610630673, "learning_rate": 3.746323590992125e-06, "loss": 0.4222, "step": 5743 }, { "epoch": 1.8685751463890696, "grad_norm": 1.96478314321902, "learning_rate": 3.7444913898372826e-06, "loss": 0.3355, "step": 5744 }, { "epoch": 1.868900455432661, "grad_norm": 1.8783908207913824, "learning_rate": 3.7426593685976607e-06, "loss": 0.3877, "step": 5745 }, { "epoch": 1.8692257644762524, "grad_norm": 1.9260898132788178, "learning_rate": 3.740827527535791e-06, "loss": 0.3833, "step": 5746 }, { "epoch": 1.869551073519844, "grad_norm": 1.942752068042013, "learning_rate": 3.738995866914176e-06, "loss": 0.3479, "step": 5747 }, { "epoch": 1.8698763825634352, "grad_norm": 1.973287958008146, "learning_rate": 3.7371643869952944e-06, "loss": 0.4127, "step": 5748 }, { "epoch": 1.8702016916070265, "grad_norm": 2.3126801943031015, "learning_rate": 3.7353330880415963e-06, "loss": 0.4266, "step": 5749 }, { "epoch": 1.870527000650618, "grad_norm": 1.9116675350417367, "learning_rate": 3.7335019703155114e-06, "loss": 0.3845, "step": 5750 }, { "epoch": 1.8708523096942096, "grad_norm": 2.05690506190175, "learning_rate": 3.731671034079438e-06, "loss": 0.353, "step": 5751 }, { "epoch": 1.8711776187378009, "grad_norm": 1.717014168396074, "learning_rate": 3.729840279595753e-06, "loss": 0.3327, "step": 5752 }, { "epoch": 1.8715029277813924, "grad_norm": 1.5386461547653187, "learning_rate": 3.728009707126799e-06, "loss": 0.292, "step": 5753 }, { "epoch": 1.871828236824984, "grad_norm": 1.9723800582826605, "learning_rate": 3.726179316934904e-06, "loss": 0.355, "step": 5754 }, { "epoch": 1.8721535458685752, "grad_norm": 2.1673237787909696, "learning_rate": 3.7243491092823637e-06, "loss": 0.4162, "step": 5755 }, { "epoch": 1.8724788549121665, "grad_norm": 1.8743359424141133, "learning_rate": 3.7225190844314453e-06, "loss": 0.3996, "step": 5756 }, { "epoch": 1.872804163955758, "grad_norm": 1.7186259672852326, "learning_rate": 3.720689242644393e-06, "loss": 0.3363, "step": 5757 }, { "epoch": 1.8731294729993495, "grad_norm": 1.9108256545317497, "learning_rate": 3.7188595841834263e-06, "loss": 0.3784, "step": 5758 }, { "epoch": 1.8734547820429408, "grad_norm": 1.928297961543499, "learning_rate": 3.7170301093107353e-06, "loss": 0.4148, "step": 5759 }, { "epoch": 1.873780091086532, "grad_norm": 1.9765138384808445, "learning_rate": 3.715200818288485e-06, "loss": 0.3939, "step": 5760 }, { "epoch": 1.8741054001301236, "grad_norm": 1.837162109504649, "learning_rate": 3.7133717113788125e-06, "loss": 0.3298, "step": 5761 }, { "epoch": 1.8744307091737151, "grad_norm": 1.6017105256938626, "learning_rate": 3.7115427888438326e-06, "loss": 0.3081, "step": 5762 }, { "epoch": 1.8747560182173064, "grad_norm": 1.8658891245385716, "learning_rate": 3.7097140509456307e-06, "loss": 0.386, "step": 5763 }, { "epoch": 1.8750813272608977, "grad_norm": 1.937267316792055, "learning_rate": 3.7078854979462653e-06, "loss": 0.3153, "step": 5764 }, { "epoch": 1.8754066363044892, "grad_norm": 2.019884596964925, "learning_rate": 3.7060571301077685e-06, "loss": 0.3694, "step": 5765 }, { "epoch": 1.8757319453480807, "grad_norm": 1.8944873967386615, "learning_rate": 3.704228947692149e-06, "loss": 0.3748, "step": 5766 }, { "epoch": 1.876057254391672, "grad_norm": 1.7592815889002604, "learning_rate": 3.7024009509613847e-06, "loss": 0.3746, "step": 5767 }, { "epoch": 1.8763825634352636, "grad_norm": 1.8495544756948479, "learning_rate": 3.7005731401774307e-06, "loss": 0.3982, "step": 5768 }, { "epoch": 1.876707872478855, "grad_norm": 1.8662127226656267, "learning_rate": 3.698745515602209e-06, "loss": 0.3796, "step": 5769 }, { "epoch": 1.8770331815224464, "grad_norm": 1.8513610735381343, "learning_rate": 3.696918077497624e-06, "loss": 0.4064, "step": 5770 }, { "epoch": 1.8773584905660377, "grad_norm": 1.9475304872872365, "learning_rate": 3.6950908261255493e-06, "loss": 0.2827, "step": 5771 }, { "epoch": 1.8776837996096292, "grad_norm": 1.9453875017380433, "learning_rate": 3.693263761747827e-06, "loss": 0.3894, "step": 5772 }, { "epoch": 1.8780091086532207, "grad_norm": 2.0231259731346083, "learning_rate": 3.6914368846262783e-06, "loss": 0.3987, "step": 5773 }, { "epoch": 1.878334417696812, "grad_norm": 1.6421332087266045, "learning_rate": 3.689610195022699e-06, "loss": 0.3202, "step": 5774 }, { "epoch": 1.8786597267404033, "grad_norm": 1.8207619756274813, "learning_rate": 3.687783693198851e-06, "loss": 0.3852, "step": 5775 }, { "epoch": 1.8789850357839948, "grad_norm": 1.789427265577003, "learning_rate": 3.685957379416474e-06, "loss": 0.3982, "step": 5776 }, { "epoch": 1.8793103448275863, "grad_norm": 2.1240610263197213, "learning_rate": 3.6841312539372807e-06, "loss": 0.3777, "step": 5777 }, { "epoch": 1.8796356538711776, "grad_norm": 1.725056391255185, "learning_rate": 3.682305317022957e-06, "loss": 0.2782, "step": 5778 }, { "epoch": 1.8799609629147689, "grad_norm": 2.076148097488284, "learning_rate": 3.68047956893516e-06, "loss": 0.4552, "step": 5779 }, { "epoch": 1.8802862719583604, "grad_norm": 1.8196334114749018, "learning_rate": 3.6786540099355195e-06, "loss": 0.3589, "step": 5780 }, { "epoch": 1.880611581001952, "grad_norm": 2.074097897490855, "learning_rate": 3.6768286402856405e-06, "loss": 0.4117, "step": 5781 }, { "epoch": 1.8809368900455432, "grad_norm": 2.069163791529246, "learning_rate": 3.675003460247101e-06, "loss": 0.4666, "step": 5782 }, { "epoch": 1.8812621990891347, "grad_norm": 1.9177098524418763, "learning_rate": 3.6731784700814476e-06, "loss": 0.2997, "step": 5783 }, { "epoch": 1.8815875081327262, "grad_norm": 1.9363129769402285, "learning_rate": 3.671353670050204e-06, "loss": 0.4333, "step": 5784 }, { "epoch": 1.8819128171763175, "grad_norm": 2.186862666633713, "learning_rate": 3.669529060414866e-06, "loss": 0.4482, "step": 5785 }, { "epoch": 1.8822381262199088, "grad_norm": 2.0126791076543, "learning_rate": 3.6677046414369007e-06, "loss": 0.4124, "step": 5786 }, { "epoch": 1.8825634352635003, "grad_norm": 2.0616560677077875, "learning_rate": 3.665880413377749e-06, "loss": 0.4445, "step": 5787 }, { "epoch": 1.8828887443070919, "grad_norm": 1.8398842795157195, "learning_rate": 3.6640563764988214e-06, "loss": 0.3222, "step": 5788 }, { "epoch": 1.8832140533506831, "grad_norm": 1.7999380308214759, "learning_rate": 3.6622325310615075e-06, "loss": 0.3393, "step": 5789 }, { "epoch": 1.8835393623942744, "grad_norm": 2.0172677757465336, "learning_rate": 3.6604088773271653e-06, "loss": 0.395, "step": 5790 }, { "epoch": 1.883864671437866, "grad_norm": 2.0103437509141293, "learning_rate": 3.658585415557122e-06, "loss": 0.3482, "step": 5791 }, { "epoch": 1.8841899804814575, "grad_norm": 1.888546955878043, "learning_rate": 3.656762146012682e-06, "loss": 0.3534, "step": 5792 }, { "epoch": 1.8845152895250488, "grad_norm": 2.781949624473673, "learning_rate": 3.6549390689551247e-06, "loss": 0.4403, "step": 5793 }, { "epoch": 1.88484059856864, "grad_norm": 1.7081160243895217, "learning_rate": 3.6531161846456946e-06, "loss": 0.2736, "step": 5794 }, { "epoch": 1.8851659076122316, "grad_norm": 2.007066902825207, "learning_rate": 3.6512934933456133e-06, "loss": 0.3736, "step": 5795 }, { "epoch": 1.885491216655823, "grad_norm": 1.7939305640820895, "learning_rate": 3.6494709953160728e-06, "loss": 0.3278, "step": 5796 }, { "epoch": 1.8858165256994144, "grad_norm": 2.0557609790300972, "learning_rate": 3.6476486908182396e-06, "loss": 0.342, "step": 5797 }, { "epoch": 1.886141834743006, "grad_norm": 2.0628698078339087, "learning_rate": 3.6458265801132514e-06, "loss": 0.3363, "step": 5798 }, { "epoch": 1.8864671437865974, "grad_norm": 2.073168676513103, "learning_rate": 3.644004663462217e-06, "loss": 0.4126, "step": 5799 }, { "epoch": 1.8867924528301887, "grad_norm": 1.9330744777324105, "learning_rate": 3.6421829411262167e-06, "loss": 0.3916, "step": 5800 }, { "epoch": 1.88711776187378, "grad_norm": 1.9460235679589508, "learning_rate": 3.640361413366308e-06, "loss": 0.3566, "step": 5801 }, { "epoch": 1.8874430709173715, "grad_norm": 2.0366010841844693, "learning_rate": 3.6385400804435146e-06, "loss": 0.3298, "step": 5802 }, { "epoch": 1.887768379960963, "grad_norm": 2.1258410421034437, "learning_rate": 3.6367189426188365e-06, "loss": 0.4333, "step": 5803 }, { "epoch": 1.8880936890045543, "grad_norm": 1.943850682181012, "learning_rate": 3.63489800015324e-06, "loss": 0.317, "step": 5804 }, { "epoch": 1.8884189980481456, "grad_norm": 2.0700130906131418, "learning_rate": 3.633077253307673e-06, "loss": 0.4086, "step": 5805 }, { "epoch": 1.8887443070917371, "grad_norm": 1.8585940181020013, "learning_rate": 3.6312567023430468e-06, "loss": 0.3887, "step": 5806 }, { "epoch": 1.8890696161353286, "grad_norm": 1.8658276250402719, "learning_rate": 3.629436347520248e-06, "loss": 0.3952, "step": 5807 }, { "epoch": 1.88939492517892, "grad_norm": 1.8573142708133608, "learning_rate": 3.627616189100131e-06, "loss": 0.3255, "step": 5808 }, { "epoch": 1.8897202342225112, "grad_norm": 1.7070296910228426, "learning_rate": 3.6257962273435333e-06, "loss": 0.3538, "step": 5809 }, { "epoch": 1.8900455432661027, "grad_norm": 1.7376836685974864, "learning_rate": 3.6239764625112508e-06, "loss": 0.3227, "step": 5810 }, { "epoch": 1.8903708523096943, "grad_norm": 2.409995416054302, "learning_rate": 3.6221568948640582e-06, "loss": 0.4235, "step": 5811 }, { "epoch": 1.8906961613532856, "grad_norm": 2.100257919162898, "learning_rate": 3.620337524662701e-06, "loss": 0.4436, "step": 5812 }, { "epoch": 1.891021470396877, "grad_norm": 1.8555042204029137, "learning_rate": 3.618518352167897e-06, "loss": 0.3909, "step": 5813 }, { "epoch": 1.8913467794404686, "grad_norm": 2.0355148661476963, "learning_rate": 3.6166993776403347e-06, "loss": 0.3276, "step": 5814 }, { "epoch": 1.8916720884840599, "grad_norm": 1.57087943558461, "learning_rate": 3.6148806013406734e-06, "loss": 0.2278, "step": 5815 }, { "epoch": 1.8919973975276512, "grad_norm": 1.5434031384382096, "learning_rate": 3.6130620235295444e-06, "loss": 0.3088, "step": 5816 }, { "epoch": 1.8923227065712427, "grad_norm": 2.1928612225925543, "learning_rate": 3.6112436444675535e-06, "loss": 0.4648, "step": 5817 }, { "epoch": 1.8926480156148342, "grad_norm": 1.7767344248698687, "learning_rate": 3.6094254644152737e-06, "loss": 0.3962, "step": 5818 }, { "epoch": 1.8929733246584255, "grad_norm": 1.8796915712855387, "learning_rate": 3.6076074836332525e-06, "loss": 0.4072, "step": 5819 }, { "epoch": 1.8932986337020168, "grad_norm": 2.000748805271628, "learning_rate": 3.605789702382006e-06, "loss": 0.3549, "step": 5820 }, { "epoch": 1.8936239427456083, "grad_norm": 1.7184235444291385, "learning_rate": 3.6039721209220257e-06, "loss": 0.2789, "step": 5821 }, { "epoch": 1.8939492517891998, "grad_norm": 1.688639563302373, "learning_rate": 3.6021547395137723e-06, "loss": 0.3226, "step": 5822 }, { "epoch": 1.894274560832791, "grad_norm": 1.7033506331102355, "learning_rate": 3.600337558417676e-06, "loss": 0.3048, "step": 5823 }, { "epoch": 1.8945998698763824, "grad_norm": 1.9462428852892393, "learning_rate": 3.5985205778941385e-06, "loss": 0.404, "step": 5824 }, { "epoch": 1.8949251789199741, "grad_norm": 2.0200098929854833, "learning_rate": 3.5967037982035397e-06, "loss": 0.4581, "step": 5825 }, { "epoch": 1.8952504879635654, "grad_norm": 2.1718182422991217, "learning_rate": 3.594887219606221e-06, "loss": 0.3794, "step": 5826 }, { "epoch": 1.8955757970071567, "grad_norm": 1.8718759756392922, "learning_rate": 3.5930708423625003e-06, "loss": 0.4196, "step": 5827 }, { "epoch": 1.8959011060507482, "grad_norm": 2.082655392569147, "learning_rate": 3.591254666732666e-06, "loss": 0.5074, "step": 5828 }, { "epoch": 1.8962264150943398, "grad_norm": 1.9204414668801506, "learning_rate": 3.5894386929769776e-06, "loss": 0.4233, "step": 5829 }, { "epoch": 1.896551724137931, "grad_norm": 2.041110201987104, "learning_rate": 3.5876229213556647e-06, "loss": 0.363, "step": 5830 }, { "epoch": 1.8968770331815223, "grad_norm": 1.7591342606053602, "learning_rate": 3.5858073521289295e-06, "loss": 0.3955, "step": 5831 }, { "epoch": 1.8972023422251139, "grad_norm": 1.917768503235422, "learning_rate": 3.5839919855569415e-06, "loss": 0.3831, "step": 5832 }, { "epoch": 1.8975276512687054, "grad_norm": 1.723405164414437, "learning_rate": 3.5821768218998474e-06, "loss": 0.3482, "step": 5833 }, { "epoch": 1.8978529603122967, "grad_norm": 1.8576698114331376, "learning_rate": 3.580361861417761e-06, "loss": 0.3497, "step": 5834 }, { "epoch": 1.898178269355888, "grad_norm": 2.275631239906097, "learning_rate": 3.5785471043707653e-06, "loss": 0.4884, "step": 5835 }, { "epoch": 1.8985035783994795, "grad_norm": 2.2091157047973597, "learning_rate": 3.5767325510189172e-06, "loss": 0.396, "step": 5836 }, { "epoch": 1.898828887443071, "grad_norm": 2.309428262497906, "learning_rate": 3.574918201622244e-06, "loss": 0.4119, "step": 5837 }, { "epoch": 1.8991541964866623, "grad_norm": 1.7607970227477674, "learning_rate": 3.5731040564407426e-06, "loss": 0.3403, "step": 5838 }, { "epoch": 1.8994795055302536, "grad_norm": 1.6459836498033298, "learning_rate": 3.5712901157343825e-06, "loss": 0.2991, "step": 5839 }, { "epoch": 1.8998048145738453, "grad_norm": 1.9960682943520145, "learning_rate": 3.569476379763098e-06, "loss": 0.3753, "step": 5840 }, { "epoch": 1.9001301236174366, "grad_norm": 1.6200764771737068, "learning_rate": 3.567662848786805e-06, "loss": 0.297, "step": 5841 }, { "epoch": 1.900455432661028, "grad_norm": 1.8492578341516948, "learning_rate": 3.56584952306538e-06, "loss": 0.3427, "step": 5842 }, { "epoch": 1.9007807417046194, "grad_norm": 2.1471218940360663, "learning_rate": 3.5640364028586748e-06, "loss": 0.4273, "step": 5843 }, { "epoch": 1.901106050748211, "grad_norm": 1.8764290109655088, "learning_rate": 3.5622234884265095e-06, "loss": 0.3721, "step": 5844 }, { "epoch": 1.9014313597918022, "grad_norm": 2.097510547318705, "learning_rate": 3.560410780028677e-06, "loss": 0.3891, "step": 5845 }, { "epoch": 1.9017566688353935, "grad_norm": 1.932645697927045, "learning_rate": 3.5585982779249406e-06, "loss": 0.3857, "step": 5846 }, { "epoch": 1.902081977878985, "grad_norm": 2.2863278275331465, "learning_rate": 3.5567859823750316e-06, "loss": 0.3922, "step": 5847 }, { "epoch": 1.9024072869225765, "grad_norm": 2.1426654013860724, "learning_rate": 3.5549738936386526e-06, "loss": 0.4787, "step": 5848 }, { "epoch": 1.9027325959661678, "grad_norm": 2.145445721806196, "learning_rate": 3.5531620119754784e-06, "loss": 0.466, "step": 5849 }, { "epoch": 1.9030579050097591, "grad_norm": 2.048061581531979, "learning_rate": 3.551350337645153e-06, "loss": 0.3947, "step": 5850 }, { "epoch": 1.9033832140533506, "grad_norm": 1.8534927159075072, "learning_rate": 3.5495388709072907e-06, "loss": 0.4263, "step": 5851 }, { "epoch": 1.9037085230969422, "grad_norm": 1.9991878458955947, "learning_rate": 3.5477276120214737e-06, "loss": 0.3634, "step": 5852 }, { "epoch": 1.9040338321405335, "grad_norm": 1.747232056141017, "learning_rate": 3.545916561247259e-06, "loss": 0.3253, "step": 5853 }, { "epoch": 1.904359141184125, "grad_norm": 1.6758655289571516, "learning_rate": 3.5441057188441707e-06, "loss": 0.3159, "step": 5854 }, { "epoch": 1.9046844502277165, "grad_norm": 2.026048441738098, "learning_rate": 3.5422950850717054e-06, "loss": 0.3677, "step": 5855 }, { "epoch": 1.9050097592713078, "grad_norm": 1.8811608423348578, "learning_rate": 3.5404846601893217e-06, "loss": 0.3308, "step": 5856 }, { "epoch": 1.905335068314899, "grad_norm": 1.7615567200429099, "learning_rate": 3.5386744444564624e-06, "loss": 0.341, "step": 5857 }, { "epoch": 1.9056603773584906, "grad_norm": 1.6425990848137164, "learning_rate": 3.53686443813253e-06, "loss": 0.3389, "step": 5858 }, { "epoch": 1.905985686402082, "grad_norm": 1.9700568858131984, "learning_rate": 3.535054641476899e-06, "loss": 0.3966, "step": 5859 }, { "epoch": 1.9063109954456734, "grad_norm": 1.9600663499056987, "learning_rate": 3.533245054748912e-06, "loss": 0.4187, "step": 5860 }, { "epoch": 1.9066363044892647, "grad_norm": 2.0616484562438675, "learning_rate": 3.5314356782078896e-06, "loss": 0.4653, "step": 5861 }, { "epoch": 1.9069616135328562, "grad_norm": 1.9921122277372465, "learning_rate": 3.5296265121131124e-06, "loss": 0.3325, "step": 5862 }, { "epoch": 1.9072869225764477, "grad_norm": 1.8551032892810437, "learning_rate": 3.527817556723837e-06, "loss": 0.3304, "step": 5863 }, { "epoch": 1.907612231620039, "grad_norm": 2.104115901860568, "learning_rate": 3.526008812299285e-06, "loss": 0.3969, "step": 5864 }, { "epoch": 1.9079375406636303, "grad_norm": 2.0900789573568197, "learning_rate": 3.524200279098655e-06, "loss": 0.4301, "step": 5865 }, { "epoch": 1.9082628497072218, "grad_norm": 2.238448925364313, "learning_rate": 3.5223919573811092e-06, "loss": 0.5693, "step": 5866 }, { "epoch": 1.9085881587508133, "grad_norm": 1.8653395947336022, "learning_rate": 3.5205838474057806e-06, "loss": 0.369, "step": 5867 }, { "epoch": 1.9089134677944046, "grad_norm": 2.1696478751871453, "learning_rate": 3.518775949431772e-06, "loss": 0.4538, "step": 5868 }, { "epoch": 1.9092387768379961, "grad_norm": 1.755631274627111, "learning_rate": 3.516968263718159e-06, "loss": 0.3777, "step": 5869 }, { "epoch": 1.9095640858815877, "grad_norm": 2.0462890014446518, "learning_rate": 3.5151607905239815e-06, "loss": 0.3304, "step": 5870 }, { "epoch": 1.909889394925179, "grad_norm": 1.8237469781592486, "learning_rate": 3.5133535301082545e-06, "loss": 0.2774, "step": 5871 }, { "epoch": 1.9102147039687702, "grad_norm": 1.8937530553878255, "learning_rate": 3.5115464827299552e-06, "loss": 0.3769, "step": 5872 }, { "epoch": 1.9105400130123618, "grad_norm": 1.9914652959118464, "learning_rate": 3.5097396486480385e-06, "loss": 0.3106, "step": 5873 }, { "epoch": 1.9108653220559533, "grad_norm": 1.9366066293694988, "learning_rate": 3.5079330281214254e-06, "loss": 0.3915, "step": 5874 }, { "epoch": 1.9111906310995446, "grad_norm": 1.9662008699956384, "learning_rate": 3.5061266214090022e-06, "loss": 0.3405, "step": 5875 }, { "epoch": 1.9115159401431359, "grad_norm": 1.9469954866667962, "learning_rate": 3.5043204287696286e-06, "loss": 0.3978, "step": 5876 }, { "epoch": 1.9118412491867274, "grad_norm": 1.927902761286604, "learning_rate": 3.502514450462138e-06, "loss": 0.362, "step": 5877 }, { "epoch": 1.912166558230319, "grad_norm": 2.0259127343317567, "learning_rate": 3.5007086867453244e-06, "loss": 0.308, "step": 5878 }, { "epoch": 1.9124918672739102, "grad_norm": 2.1150028317614695, "learning_rate": 3.4989031378779546e-06, "loss": 0.4401, "step": 5879 }, { "epoch": 1.9128171763175015, "grad_norm": 2.153337078817791, "learning_rate": 3.497097804118766e-06, "loss": 0.4616, "step": 5880 }, { "epoch": 1.913142485361093, "grad_norm": 1.8755078768838607, "learning_rate": 3.4952926857264644e-06, "loss": 0.3944, "step": 5881 }, { "epoch": 1.9134677944046845, "grad_norm": 1.9989650001284693, "learning_rate": 3.4934877829597248e-06, "loss": 0.3992, "step": 5882 }, { "epoch": 1.9137931034482758, "grad_norm": 1.9028884226594196, "learning_rate": 3.4916830960771897e-06, "loss": 0.4162, "step": 5883 }, { "epoch": 1.9141184124918673, "grad_norm": 2.0397066042840675, "learning_rate": 3.4898786253374716e-06, "loss": 0.3561, "step": 5884 }, { "epoch": 1.9144437215354588, "grad_norm": 1.9200813848129192, "learning_rate": 3.488074370999155e-06, "loss": 0.2674, "step": 5885 }, { "epoch": 1.9147690305790501, "grad_norm": 1.8468103499836754, "learning_rate": 3.486270333320788e-06, "loss": 0.3497, "step": 5886 }, { "epoch": 1.9150943396226414, "grad_norm": 1.817279910990594, "learning_rate": 3.4844665125608933e-06, "loss": 0.3866, "step": 5887 }, { "epoch": 1.915419648666233, "grad_norm": 2.0554453453812798, "learning_rate": 3.4826629089779553e-06, "loss": 0.3607, "step": 5888 }, { "epoch": 1.9157449577098244, "grad_norm": 2.105746158713741, "learning_rate": 3.480859522830435e-06, "loss": 0.4157, "step": 5889 }, { "epoch": 1.9160702667534157, "grad_norm": 2.5687130107659044, "learning_rate": 3.4790563543767595e-06, "loss": 0.5276, "step": 5890 }, { "epoch": 1.916395575797007, "grad_norm": 1.7421475519879297, "learning_rate": 3.4772534038753215e-06, "loss": 0.3633, "step": 5891 }, { "epoch": 1.9167208848405985, "grad_norm": 1.8474303431630794, "learning_rate": 3.475450671584484e-06, "loss": 0.2974, "step": 5892 }, { "epoch": 1.91704619388419, "grad_norm": 1.9346907693623157, "learning_rate": 3.473648157762584e-06, "loss": 0.3282, "step": 5893 }, { "epoch": 1.9173715029277814, "grad_norm": 1.9275394327834117, "learning_rate": 3.4718458626679203e-06, "loss": 0.3553, "step": 5894 }, { "epoch": 1.9176968119713727, "grad_norm": 1.757710296122235, "learning_rate": 3.470043786558763e-06, "loss": 0.3197, "step": 5895 }, { "epoch": 1.9180221210149642, "grad_norm": 1.786783265051212, "learning_rate": 3.4682419296933507e-06, "loss": 0.3094, "step": 5896 }, { "epoch": 1.9183474300585557, "grad_norm": 1.7235960689109808, "learning_rate": 3.466440292329892e-06, "loss": 0.2833, "step": 5897 }, { "epoch": 1.918672739102147, "grad_norm": 1.8658616287534968, "learning_rate": 3.464638874726561e-06, "loss": 0.3862, "step": 5898 }, { "epoch": 1.9189980481457385, "grad_norm": 1.5688283860220205, "learning_rate": 3.462837677141504e-06, "loss": 0.3225, "step": 5899 }, { "epoch": 1.91932335718933, "grad_norm": 2.06562062334134, "learning_rate": 3.4610366998328314e-06, "loss": 0.3793, "step": 5900 }, { "epoch": 1.9196486662329213, "grad_norm": 2.698379269510953, "learning_rate": 3.459235943058627e-06, "loss": 0.4398, "step": 5901 }, { "epoch": 1.9199739752765126, "grad_norm": 1.6500265048442437, "learning_rate": 3.457435407076939e-06, "loss": 0.3051, "step": 5902 }, { "epoch": 1.920299284320104, "grad_norm": 1.6179312036731643, "learning_rate": 3.455635092145786e-06, "loss": 0.3403, "step": 5903 }, { "epoch": 1.9206245933636956, "grad_norm": 1.806735967925334, "learning_rate": 3.453834998523152e-06, "loss": 0.4127, "step": 5904 }, { "epoch": 1.920949902407287, "grad_norm": 1.7943825348242548, "learning_rate": 3.4520351264669954e-06, "loss": 0.3085, "step": 5905 }, { "epoch": 1.9212752114508782, "grad_norm": 2.029622246398904, "learning_rate": 3.450235476235237e-06, "loss": 0.3775, "step": 5906 }, { "epoch": 1.9216005204944697, "grad_norm": 2.0123726435581903, "learning_rate": 3.448436048085767e-06, "loss": 0.3827, "step": 5907 }, { "epoch": 1.9219258295380612, "grad_norm": 2.086856583155522, "learning_rate": 3.446636842276447e-06, "loss": 0.3885, "step": 5908 }, { "epoch": 1.9222511385816525, "grad_norm": 1.975636817869186, "learning_rate": 3.4448378590651045e-06, "loss": 0.3306, "step": 5909 }, { "epoch": 1.9225764476252438, "grad_norm": 1.8141551152896735, "learning_rate": 3.443039098709532e-06, "loss": 0.3305, "step": 5910 }, { "epoch": 1.9229017566688356, "grad_norm": 1.842144549706399, "learning_rate": 3.441240561467493e-06, "loss": 0.3911, "step": 5911 }, { "epoch": 1.9232270657124269, "grad_norm": 2.14588967994984, "learning_rate": 3.439442247596724e-06, "loss": 0.355, "step": 5912 }, { "epoch": 1.9235523747560181, "grad_norm": 1.743979071304815, "learning_rate": 3.4376441573549203e-06, "loss": 0.3489, "step": 5913 }, { "epoch": 1.9238776837996097, "grad_norm": 1.8097595210341426, "learning_rate": 3.4358462909997502e-06, "loss": 0.3957, "step": 5914 }, { "epoch": 1.9242029928432012, "grad_norm": 1.8827745875036281, "learning_rate": 3.4340486487888493e-06, "loss": 0.3998, "step": 5915 }, { "epoch": 1.9245283018867925, "grad_norm": 1.901671233856633, "learning_rate": 3.432251230979821e-06, "loss": 0.3723, "step": 5916 }, { "epoch": 1.9248536109303838, "grad_norm": 2.2906278788645427, "learning_rate": 3.4304540378302376e-06, "loss": 0.3672, "step": 5917 }, { "epoch": 1.9251789199739753, "grad_norm": 2.1946261075476836, "learning_rate": 3.4286570695976363e-06, "loss": 0.3661, "step": 5918 }, { "epoch": 1.9255042290175668, "grad_norm": 1.9875692498872846, "learning_rate": 3.426860326539524e-06, "loss": 0.3941, "step": 5919 }, { "epoch": 1.925829538061158, "grad_norm": 2.0500044119615657, "learning_rate": 3.425063808913376e-06, "loss": 0.3609, "step": 5920 }, { "epoch": 1.9261548471047494, "grad_norm": 2.272282540048997, "learning_rate": 3.423267516976635e-06, "loss": 0.4286, "step": 5921 }, { "epoch": 1.926480156148341, "grad_norm": 1.9754505795323982, "learning_rate": 3.4214714509867097e-06, "loss": 0.3698, "step": 5922 }, { "epoch": 1.9268054651919324, "grad_norm": 1.805376926456429, "learning_rate": 3.419675611200977e-06, "loss": 0.3407, "step": 5923 }, { "epoch": 1.9271307742355237, "grad_norm": 1.9269510863890025, "learning_rate": 3.4178799978767842e-06, "loss": 0.4329, "step": 5924 }, { "epoch": 1.927456083279115, "grad_norm": 1.9433506836975005, "learning_rate": 3.4160846112714423e-06, "loss": 0.3565, "step": 5925 }, { "epoch": 1.9277813923227067, "grad_norm": 2.034748310578355, "learning_rate": 3.414289451642233e-06, "loss": 0.3252, "step": 5926 }, { "epoch": 1.928106701366298, "grad_norm": 1.8819889242969008, "learning_rate": 3.4124945192463986e-06, "loss": 0.3932, "step": 5927 }, { "epoch": 1.9284320104098893, "grad_norm": 2.07582336970223, "learning_rate": 3.4106998143411618e-06, "loss": 0.4431, "step": 5928 }, { "epoch": 1.9287573194534808, "grad_norm": 2.1787333353751226, "learning_rate": 3.4089053371837e-06, "loss": 0.4617, "step": 5929 }, { "epoch": 1.9290826284970723, "grad_norm": 1.8046581532800217, "learning_rate": 3.4071110880311644e-06, "loss": 0.3626, "step": 5930 }, { "epoch": 1.9294079375406636, "grad_norm": 2.018227749225444, "learning_rate": 3.405317067140671e-06, "loss": 0.4623, "step": 5931 }, { "epoch": 1.929733246584255, "grad_norm": 2.5163501282258713, "learning_rate": 3.403523274769306e-06, "loss": 0.4708, "step": 5932 }, { "epoch": 1.9300585556278465, "grad_norm": 1.9244651049451509, "learning_rate": 3.4017297111741198e-06, "loss": 0.4105, "step": 5933 }, { "epoch": 1.930383864671438, "grad_norm": 2.02115964560441, "learning_rate": 3.3999363766121317e-06, "loss": 0.3338, "step": 5934 }, { "epoch": 1.9307091737150293, "grad_norm": 2.1609917027728347, "learning_rate": 3.3981432713403273e-06, "loss": 0.4081, "step": 5935 }, { "epoch": 1.9310344827586206, "grad_norm": 1.6979159236036658, "learning_rate": 3.3963503956156603e-06, "loss": 0.2894, "step": 5936 }, { "epoch": 1.931359791802212, "grad_norm": 1.8581011031065946, "learning_rate": 3.394557749695051e-06, "loss": 0.3367, "step": 5937 }, { "epoch": 1.9316851008458036, "grad_norm": 2.0423577350170805, "learning_rate": 3.3927653338353865e-06, "loss": 0.3954, "step": 5938 }, { "epoch": 1.9320104098893949, "grad_norm": 1.894093324276823, "learning_rate": 3.3909731482935204e-06, "loss": 0.3435, "step": 5939 }, { "epoch": 1.9323357189329864, "grad_norm": 1.882493785807442, "learning_rate": 3.389181193326276e-06, "loss": 0.3817, "step": 5940 }, { "epoch": 1.932661027976578, "grad_norm": 1.749041668814243, "learning_rate": 3.3873894691904407e-06, "loss": 0.3806, "step": 5941 }, { "epoch": 1.9329863370201692, "grad_norm": 1.884162320541837, "learning_rate": 3.3855979761427705e-06, "loss": 0.352, "step": 5942 }, { "epoch": 1.9333116460637605, "grad_norm": 1.7898886109408851, "learning_rate": 3.3838067144399833e-06, "loss": 0.361, "step": 5943 }, { "epoch": 1.933636955107352, "grad_norm": 2.191452024098375, "learning_rate": 3.3820156843387743e-06, "loss": 0.4033, "step": 5944 }, { "epoch": 1.9339622641509435, "grad_norm": 1.9472081073965046, "learning_rate": 3.380224886095797e-06, "loss": 0.3285, "step": 5945 }, { "epoch": 1.9342875731945348, "grad_norm": 2.0639242609013957, "learning_rate": 3.3784343199676727e-06, "loss": 0.3764, "step": 5946 }, { "epoch": 1.934612882238126, "grad_norm": 1.8868402638315573, "learning_rate": 3.376643986210989e-06, "loss": 0.3634, "step": 5947 }, { "epoch": 1.9349381912817176, "grad_norm": 1.73592214628294, "learning_rate": 3.3748538850823074e-06, "loss": 0.3073, "step": 5948 }, { "epoch": 1.9352635003253091, "grad_norm": 1.7359749999670706, "learning_rate": 3.3730640168381474e-06, "loss": 0.3288, "step": 5949 }, { "epoch": 1.9355888093689004, "grad_norm": 1.7711142464166465, "learning_rate": 3.371274381734998e-06, "loss": 0.3448, "step": 5950 }, { "epoch": 1.9359141184124917, "grad_norm": 2.059311101288671, "learning_rate": 3.3694849800293144e-06, "loss": 0.3912, "step": 5951 }, { "epoch": 1.9362394274560832, "grad_norm": 2.1346423495677795, "learning_rate": 3.3676958119775215e-06, "loss": 0.4011, "step": 5952 }, { "epoch": 1.9365647364996748, "grad_norm": 1.9036886623464835, "learning_rate": 3.3659068778360072e-06, "loss": 0.3117, "step": 5953 }, { "epoch": 1.936890045543266, "grad_norm": 1.9790161328512141, "learning_rate": 3.3641181778611266e-06, "loss": 0.3775, "step": 5954 }, { "epoch": 1.9372153545868576, "grad_norm": 1.8103112902687843, "learning_rate": 3.3623297123092007e-06, "loss": 0.4089, "step": 5955 }, { "epoch": 1.937540663630449, "grad_norm": 1.9299221716762351, "learning_rate": 3.36054148143652e-06, "loss": 0.3627, "step": 5956 }, { "epoch": 1.9378659726740404, "grad_norm": 2.287615287938936, "learning_rate": 3.3587534854993387e-06, "loss": 0.3966, "step": 5957 }, { "epoch": 1.9381912817176317, "grad_norm": 1.906007060461711, "learning_rate": 3.356965724753878e-06, "loss": 0.3543, "step": 5958 }, { "epoch": 1.9385165907612232, "grad_norm": 1.9361940141095877, "learning_rate": 3.3551781994563202e-06, "loss": 0.389, "step": 5959 }, { "epoch": 1.9388418998048147, "grad_norm": 1.869707909676793, "learning_rate": 3.3533909098628258e-06, "loss": 0.3569, "step": 5960 }, { "epoch": 1.939167208848406, "grad_norm": 1.7005339423357833, "learning_rate": 3.351603856229514e-06, "loss": 0.3353, "step": 5961 }, { "epoch": 1.9394925178919973, "grad_norm": 1.6596761557318847, "learning_rate": 3.3498170388124662e-06, "loss": 0.3197, "step": 5962 }, { "epoch": 1.9398178269355888, "grad_norm": 2.1188539973101097, "learning_rate": 3.348030457867735e-06, "loss": 0.3758, "step": 5963 }, { "epoch": 1.9401431359791803, "grad_norm": 1.9174884943161752, "learning_rate": 3.3462441136513447e-06, "loss": 0.3405, "step": 5964 }, { "epoch": 1.9404684450227716, "grad_norm": 2.09262787519966, "learning_rate": 3.3444580064192735e-06, "loss": 0.3499, "step": 5965 }, { "epoch": 1.940793754066363, "grad_norm": 1.8293717970485068, "learning_rate": 3.3426721364274737e-06, "loss": 0.418, "step": 5966 }, { "epoch": 1.9411190631099544, "grad_norm": 1.9162976395411084, "learning_rate": 3.3408865039318615e-06, "loss": 0.3411, "step": 5967 }, { "epoch": 1.941444372153546, "grad_norm": 2.028561505797444, "learning_rate": 3.33910110918832e-06, "loss": 0.3725, "step": 5968 }, { "epoch": 1.9417696811971372, "grad_norm": 1.8844472237340835, "learning_rate": 3.3373159524526972e-06, "loss": 0.3979, "step": 5969 }, { "epoch": 1.9420949902407287, "grad_norm": 2.108284347627033, "learning_rate": 3.335531033980807e-06, "loss": 0.4238, "step": 5970 }, { "epoch": 1.9424202992843203, "grad_norm": 1.8654116943716885, "learning_rate": 3.3337463540284274e-06, "loss": 0.3541, "step": 5971 }, { "epoch": 1.9427456083279115, "grad_norm": 2.065631585735479, "learning_rate": 3.3319619128513087e-06, "loss": 0.4013, "step": 5972 }, { "epoch": 1.9430709173715028, "grad_norm": 2.041041204142405, "learning_rate": 3.3301777107051593e-06, "loss": 0.4203, "step": 5973 }, { "epoch": 1.9433962264150944, "grad_norm": 1.9248987484636004, "learning_rate": 3.3283937478456587e-06, "loss": 0.3563, "step": 5974 }, { "epoch": 1.9437215354586859, "grad_norm": 1.9203599141237169, "learning_rate": 3.3266100245284448e-06, "loss": 0.4702, "step": 5975 }, { "epoch": 1.9440468445022772, "grad_norm": 1.8570289894004575, "learning_rate": 3.324826541009132e-06, "loss": 0.3765, "step": 5976 }, { "epoch": 1.9443721535458685, "grad_norm": 1.6730850867552947, "learning_rate": 3.3230432975432946e-06, "loss": 0.3392, "step": 5977 }, { "epoch": 1.94469746258946, "grad_norm": 2.228638339247448, "learning_rate": 3.321260294386469e-06, "loss": 0.4197, "step": 5978 }, { "epoch": 1.9450227716330515, "grad_norm": 2.265194524682165, "learning_rate": 3.319477531794161e-06, "loss": 0.4714, "step": 5979 }, { "epoch": 1.9453480806766428, "grad_norm": 1.6548813572743275, "learning_rate": 3.3176950100218457e-06, "loss": 0.3137, "step": 5980 }, { "epoch": 1.945673389720234, "grad_norm": 1.892847260378785, "learning_rate": 3.3159127293249564e-06, "loss": 0.4012, "step": 5981 }, { "epoch": 1.9459986987638256, "grad_norm": 1.6817158029583272, "learning_rate": 3.314130689958896e-06, "loss": 0.3371, "step": 5982 }, { "epoch": 1.946324007807417, "grad_norm": 1.9551176044373662, "learning_rate": 3.3123488921790305e-06, "loss": 0.427, "step": 5983 }, { "epoch": 1.9466493168510084, "grad_norm": 1.6862955161861841, "learning_rate": 3.310567336240695e-06, "loss": 0.3556, "step": 5984 }, { "epoch": 1.9469746258946, "grad_norm": 1.704223232180204, "learning_rate": 3.308786022399186e-06, "loss": 0.3118, "step": 5985 }, { "epoch": 1.9472999349381914, "grad_norm": 2.111469065941944, "learning_rate": 3.307004950909769e-06, "loss": 0.4533, "step": 5986 }, { "epoch": 1.9476252439817827, "grad_norm": 1.7684378860074619, "learning_rate": 3.3052241220276694e-06, "loss": 0.3277, "step": 5987 }, { "epoch": 1.947950553025374, "grad_norm": 1.98199100084571, "learning_rate": 3.303443536008084e-06, "loss": 0.3952, "step": 5988 }, { "epoch": 1.9482758620689655, "grad_norm": 1.8106166875556267, "learning_rate": 3.3016631931061712e-06, "loss": 0.3676, "step": 5989 }, { "epoch": 1.948601171112557, "grad_norm": 1.9511092314422298, "learning_rate": 3.2998830935770555e-06, "loss": 0.419, "step": 5990 }, { "epoch": 1.9489264801561483, "grad_norm": 2.1245070477614294, "learning_rate": 3.298103237675824e-06, "loss": 0.3785, "step": 5991 }, { "epoch": 1.9492517891997396, "grad_norm": 1.67469083045322, "learning_rate": 3.296323625657535e-06, "loss": 0.2937, "step": 5992 }, { "epoch": 1.9495770982433311, "grad_norm": 1.9380586110955667, "learning_rate": 3.294544257777207e-06, "loss": 0.3638, "step": 5993 }, { "epoch": 1.9499024072869227, "grad_norm": 2.0269054506040303, "learning_rate": 3.2927651342898238e-06, "loss": 0.3772, "step": 5994 }, { "epoch": 1.950227716330514, "grad_norm": 1.9336633613895737, "learning_rate": 3.290986255450332e-06, "loss": 0.3565, "step": 5995 }, { "epoch": 1.9505530253741052, "grad_norm": 1.9258701521331736, "learning_rate": 3.2892076215136524e-06, "loss": 0.3728, "step": 5996 }, { "epoch": 1.9508783344176968, "grad_norm": 2.0487186377095354, "learning_rate": 3.287429232734661e-06, "loss": 0.4094, "step": 5997 }, { "epoch": 1.9512036434612883, "grad_norm": 2.017028545295591, "learning_rate": 3.285651089368202e-06, "loss": 0.389, "step": 5998 }, { "epoch": 1.9515289525048796, "grad_norm": 1.8483548593753885, "learning_rate": 3.2838731916690835e-06, "loss": 0.3386, "step": 5999 }, { "epoch": 1.951854261548471, "grad_norm": 2.271013348725127, "learning_rate": 3.2820955398920827e-06, "loss": 0.4561, "step": 6000 }, { "epoch": 1.9521795705920626, "grad_norm": 1.9889325606652954, "learning_rate": 3.2803181342919355e-06, "loss": 0.3392, "step": 6001 }, { "epoch": 1.952504879635654, "grad_norm": 1.7352851431146012, "learning_rate": 3.2785409751233468e-06, "loss": 0.2706, "step": 6002 }, { "epoch": 1.9528301886792452, "grad_norm": 1.9782069872862225, "learning_rate": 3.276764062640983e-06, "loss": 0.393, "step": 6003 }, { "epoch": 1.9531554977228367, "grad_norm": 2.039243540271715, "learning_rate": 3.274987397099479e-06, "loss": 0.4192, "step": 6004 }, { "epoch": 1.9534808067664282, "grad_norm": 1.8399834045654657, "learning_rate": 3.2732109787534304e-06, "loss": 0.4081, "step": 6005 }, { "epoch": 1.9538061158100195, "grad_norm": 1.668619518882704, "learning_rate": 3.271434807857401e-06, "loss": 0.3286, "step": 6006 }, { "epoch": 1.9541314248536108, "grad_norm": 1.947951946531579, "learning_rate": 3.2696588846659146e-06, "loss": 0.3943, "step": 6007 }, { "epoch": 1.9544567338972023, "grad_norm": 1.9013348731693498, "learning_rate": 3.2678832094334645e-06, "loss": 0.414, "step": 6008 }, { "epoch": 1.9547820429407938, "grad_norm": 1.6430362781279284, "learning_rate": 3.2661077824145053e-06, "loss": 0.3402, "step": 6009 }, { "epoch": 1.9551073519843851, "grad_norm": 1.7371182190982088, "learning_rate": 3.2643326038634582e-06, "loss": 0.3565, "step": 6010 }, { "epoch": 1.9554326610279764, "grad_norm": 1.873743613388776, "learning_rate": 3.2625576740347033e-06, "loss": 0.4034, "step": 6011 }, { "epoch": 1.9557579700715682, "grad_norm": 1.8752811125255522, "learning_rate": 3.2607829931825943e-06, "loss": 0.3191, "step": 6012 }, { "epoch": 1.9560832791151594, "grad_norm": 1.8843503970662692, "learning_rate": 3.259008561561442e-06, "loss": 0.3153, "step": 6013 }, { "epoch": 1.9564085881587507, "grad_norm": 1.517772339273344, "learning_rate": 3.2572343794255233e-06, "loss": 0.2882, "step": 6014 }, { "epoch": 1.9567338972023423, "grad_norm": 1.6888980908219966, "learning_rate": 3.255460447029079e-06, "loss": 0.3781, "step": 6015 }, { "epoch": 1.9570592062459338, "grad_norm": 1.9672610390071095, "learning_rate": 3.2536867646263166e-06, "loss": 0.3648, "step": 6016 }, { "epoch": 1.957384515289525, "grad_norm": 1.920688525736914, "learning_rate": 3.2519133324714054e-06, "loss": 0.3666, "step": 6017 }, { "epoch": 1.9577098243331164, "grad_norm": 1.6558941411196957, "learning_rate": 3.250140150818479e-06, "loss": 0.2784, "step": 6018 }, { "epoch": 1.9580351333767079, "grad_norm": 1.7914352228514656, "learning_rate": 3.2483672199216344e-06, "loss": 0.3128, "step": 6019 }, { "epoch": 1.9583604424202994, "grad_norm": 2.037178253298656, "learning_rate": 3.2465945400349364e-06, "loss": 0.3307, "step": 6020 }, { "epoch": 1.9586857514638907, "grad_norm": 2.2581748566712694, "learning_rate": 3.2448221114124094e-06, "loss": 0.4078, "step": 6021 }, { "epoch": 1.959011060507482, "grad_norm": 1.9623416542507792, "learning_rate": 3.2430499343080445e-06, "loss": 0.359, "step": 6022 }, { "epoch": 1.9593363695510735, "grad_norm": 1.7861960513177655, "learning_rate": 3.2412780089757943e-06, "loss": 0.3695, "step": 6023 }, { "epoch": 1.959661678594665, "grad_norm": 1.7290688954741416, "learning_rate": 3.239506335669579e-06, "loss": 0.3316, "step": 6024 }, { "epoch": 1.9599869876382563, "grad_norm": 1.9205849275950242, "learning_rate": 3.2377349146432798e-06, "loss": 0.3354, "step": 6025 }, { "epoch": 1.9603122966818478, "grad_norm": 2.119933926557903, "learning_rate": 3.235963746150743e-06, "loss": 0.3777, "step": 6026 }, { "epoch": 1.9606376057254393, "grad_norm": 1.7372343813073376, "learning_rate": 3.2341928304457747e-06, "loss": 0.3881, "step": 6027 }, { "epoch": 1.9609629147690306, "grad_norm": 1.8462639625298192, "learning_rate": 3.2324221677821534e-06, "loss": 0.3647, "step": 6028 }, { "epoch": 1.961288223812622, "grad_norm": 2.1022292670352525, "learning_rate": 3.230651758413614e-06, "loss": 0.4524, "step": 6029 }, { "epoch": 1.9616135328562134, "grad_norm": 2.0046886863729334, "learning_rate": 3.228881602593857e-06, "loss": 0.4218, "step": 6030 }, { "epoch": 1.961938841899805, "grad_norm": 1.934554856796069, "learning_rate": 3.227111700576545e-06, "loss": 0.3374, "step": 6031 }, { "epoch": 1.9622641509433962, "grad_norm": 2.118073304711936, "learning_rate": 3.2253420526153113e-06, "loss": 0.3867, "step": 6032 }, { "epoch": 1.9625894599869875, "grad_norm": 1.7305446593771727, "learning_rate": 3.223572658963744e-06, "loss": 0.2963, "step": 6033 }, { "epoch": 1.962914769030579, "grad_norm": 2.0292584306479173, "learning_rate": 3.221803519875398e-06, "loss": 0.4252, "step": 6034 }, { "epoch": 1.9632400780741706, "grad_norm": 1.8331573363776088, "learning_rate": 3.2200346356037935e-06, "loss": 0.3769, "step": 6035 }, { "epoch": 1.9635653871177619, "grad_norm": 1.7883761096116855, "learning_rate": 3.218266006402413e-06, "loss": 0.3466, "step": 6036 }, { "epoch": 1.9638906961613531, "grad_norm": 1.6535345973200297, "learning_rate": 3.216497632524702e-06, "loss": 0.2884, "step": 6037 }, { "epoch": 1.9642160052049447, "grad_norm": 2.0461095779645917, "learning_rate": 3.214729514224068e-06, "loss": 0.4531, "step": 6038 }, { "epoch": 1.9645413142485362, "grad_norm": 1.6747275708912444, "learning_rate": 3.2129616517538863e-06, "loss": 0.3157, "step": 6039 }, { "epoch": 1.9648666232921275, "grad_norm": 2.336622764119388, "learning_rate": 3.2111940453674905e-06, "loss": 0.4045, "step": 6040 }, { "epoch": 1.965191932335719, "grad_norm": 2.010961272904839, "learning_rate": 3.2094266953181817e-06, "loss": 0.4086, "step": 6041 }, { "epoch": 1.9655172413793105, "grad_norm": 2.4720084933301454, "learning_rate": 3.2076596018592197e-06, "loss": 0.4977, "step": 6042 }, { "epoch": 1.9658425504229018, "grad_norm": 2.089330573554802, "learning_rate": 3.205892765243833e-06, "loss": 0.4339, "step": 6043 }, { "epoch": 1.966167859466493, "grad_norm": 1.6635692666139688, "learning_rate": 3.204126185725209e-06, "loss": 0.3205, "step": 6044 }, { "epoch": 1.9664931685100846, "grad_norm": 2.0359508660831414, "learning_rate": 3.2023598635565013e-06, "loss": 0.4691, "step": 6045 }, { "epoch": 1.9668184775536761, "grad_norm": 1.8655516261558456, "learning_rate": 3.20059379899082e-06, "loss": 0.3254, "step": 6046 }, { "epoch": 1.9671437865972674, "grad_norm": 2.0065844795979766, "learning_rate": 3.1988279922812493e-06, "loss": 0.3969, "step": 6047 }, { "epoch": 1.9674690956408587, "grad_norm": 1.663723872172351, "learning_rate": 3.1970624436808295e-06, "loss": 0.296, "step": 6048 }, { "epoch": 1.9677944046844502, "grad_norm": 1.6104091993472753, "learning_rate": 3.1952971534425627e-06, "loss": 0.336, "step": 6049 }, { "epoch": 1.9681197137280417, "grad_norm": 1.900702464831428, "learning_rate": 3.193532121819415e-06, "loss": 0.3998, "step": 6050 }, { "epoch": 1.968445022771633, "grad_norm": 1.693286249138902, "learning_rate": 3.191767349064321e-06, "loss": 0.3536, "step": 6051 }, { "epoch": 1.9687703318152243, "grad_norm": 1.97896000787288, "learning_rate": 3.190002835430171e-06, "loss": 0.3837, "step": 6052 }, { "epoch": 1.9690956408588158, "grad_norm": 1.9746596636025517, "learning_rate": 3.188238581169821e-06, "loss": 0.3971, "step": 6053 }, { "epoch": 1.9694209499024073, "grad_norm": 2.0376594978319154, "learning_rate": 3.1864745865360904e-06, "loss": 0.4171, "step": 6054 }, { "epoch": 1.9697462589459986, "grad_norm": 1.754030115587257, "learning_rate": 3.1847108517817604e-06, "loss": 0.4047, "step": 6055 }, { "epoch": 1.9700715679895902, "grad_norm": 1.8420293914195718, "learning_rate": 3.1829473771595766e-06, "loss": 0.3514, "step": 6056 }, { "epoch": 1.9703968770331817, "grad_norm": 1.7621698568088366, "learning_rate": 3.181184162922245e-06, "loss": 0.343, "step": 6057 }, { "epoch": 1.970722186076773, "grad_norm": 1.853095750961822, "learning_rate": 3.1794212093224343e-06, "loss": 0.3875, "step": 6058 }, { "epoch": 1.9710474951203643, "grad_norm": 1.977901393978721, "learning_rate": 3.1776585166127787e-06, "loss": 0.4006, "step": 6059 }, { "epoch": 1.9713728041639558, "grad_norm": 1.9161464835886899, "learning_rate": 3.1758960850458732e-06, "loss": 0.4512, "step": 6060 }, { "epoch": 1.9716981132075473, "grad_norm": 1.661539721892034, "learning_rate": 3.1741339148742755e-06, "loss": 0.2727, "step": 6061 }, { "epoch": 1.9720234222511386, "grad_norm": 1.8405284736620615, "learning_rate": 3.1723720063505024e-06, "loss": 0.3922, "step": 6062 }, { "epoch": 1.9723487312947299, "grad_norm": 2.153778367354869, "learning_rate": 3.1706103597270416e-06, "loss": 0.454, "step": 6063 }, { "epoch": 1.9726740403383214, "grad_norm": 2.0078172419185374, "learning_rate": 3.1688489752563367e-06, "loss": 0.4196, "step": 6064 }, { "epoch": 1.972999349381913, "grad_norm": 1.8213067922722281, "learning_rate": 3.1670878531907933e-06, "loss": 0.3436, "step": 6065 }, { "epoch": 1.9733246584255042, "grad_norm": 1.5766679165169595, "learning_rate": 3.1653269937827812e-06, "loss": 0.3087, "step": 6066 }, { "epoch": 1.9736499674690955, "grad_norm": 1.9609451070824193, "learning_rate": 3.1635663972846366e-06, "loss": 0.3763, "step": 6067 }, { "epoch": 1.973975276512687, "grad_norm": 1.9376229922101198, "learning_rate": 3.1618060639486507e-06, "loss": 0.3814, "step": 6068 }, { "epoch": 1.9743005855562785, "grad_norm": 2.275432948045713, "learning_rate": 3.160045994027082e-06, "loss": 0.4302, "step": 6069 }, { "epoch": 1.9746258945998698, "grad_norm": 1.9350218637953227, "learning_rate": 3.158286187772147e-06, "loss": 0.4084, "step": 6070 }, { "epoch": 1.9749512036434613, "grad_norm": 2.161251247910475, "learning_rate": 3.156526645436031e-06, "loss": 0.3519, "step": 6071 }, { "epoch": 1.9752765126870528, "grad_norm": 2.0517882807251713, "learning_rate": 3.154767367270876e-06, "loss": 0.4147, "step": 6072 }, { "epoch": 1.9756018217306441, "grad_norm": 2.0267058719353575, "learning_rate": 3.153008353528787e-06, "loss": 0.3784, "step": 6073 }, { "epoch": 1.9759271307742354, "grad_norm": 1.7172660834388231, "learning_rate": 3.151249604461832e-06, "loss": 0.3443, "step": 6074 }, { "epoch": 1.976252439817827, "grad_norm": 2.097783161076698, "learning_rate": 3.149491120322042e-06, "loss": 0.3064, "step": 6075 }, { "epoch": 1.9765777488614185, "grad_norm": 2.0979403665852994, "learning_rate": 3.1477329013614075e-06, "loss": 0.4268, "step": 6076 }, { "epoch": 1.9769030579050098, "grad_norm": 1.7043896315129496, "learning_rate": 3.1459749478318845e-06, "loss": 0.374, "step": 6077 }, { "epoch": 1.977228366948601, "grad_norm": 2.129878118976226, "learning_rate": 3.144217259985386e-06, "loss": 0.3826, "step": 6078 }, { "epoch": 1.9775536759921926, "grad_norm": 1.827832602404497, "learning_rate": 3.142459838073792e-06, "loss": 0.3655, "step": 6079 }, { "epoch": 1.977878985035784, "grad_norm": 2.0096986898466156, "learning_rate": 3.140702682348943e-06, "loss": 0.3725, "step": 6080 }, { "epoch": 1.9782042940793754, "grad_norm": 1.9163693862238498, "learning_rate": 3.1389457930626386e-06, "loss": 0.4431, "step": 6081 }, { "epoch": 1.9785296031229667, "grad_norm": 2.006038866129718, "learning_rate": 3.1371891704666403e-06, "loss": 0.366, "step": 6082 }, { "epoch": 1.9788549121665582, "grad_norm": 2.1320100187903877, "learning_rate": 3.1354328148126793e-06, "loss": 0.4726, "step": 6083 }, { "epoch": 1.9791802212101497, "grad_norm": 2.145387575383267, "learning_rate": 3.133676726352438e-06, "loss": 0.3747, "step": 6084 }, { "epoch": 1.979505530253741, "grad_norm": 1.8409157984272502, "learning_rate": 3.1319209053375667e-06, "loss": 0.3892, "step": 6085 }, { "epoch": 1.9798308392973325, "grad_norm": 2.4736607591283715, "learning_rate": 3.130165352019673e-06, "loss": 0.4352, "step": 6086 }, { "epoch": 1.980156148340924, "grad_norm": 1.9439613444822272, "learning_rate": 3.1284100666503327e-06, "loss": 0.4235, "step": 6087 }, { "epoch": 1.9804814573845153, "grad_norm": 1.7778186882768154, "learning_rate": 3.126655049481078e-06, "loss": 0.3489, "step": 6088 }, { "epoch": 1.9808067664281066, "grad_norm": 1.7806153509007638, "learning_rate": 3.1249003007634034e-06, "loss": 0.3721, "step": 6089 }, { "epoch": 1.9811320754716981, "grad_norm": 1.652250043982195, "learning_rate": 3.123145820748764e-06, "loss": 0.3703, "step": 6090 }, { "epoch": 1.9814573845152896, "grad_norm": 1.7987214283006554, "learning_rate": 3.121391609688582e-06, "loss": 0.3427, "step": 6091 }, { "epoch": 1.981782693558881, "grad_norm": 2.2083464207069388, "learning_rate": 3.1196376678342344e-06, "loss": 0.3806, "step": 6092 }, { "epoch": 1.9821080026024722, "grad_norm": 1.9499203447906521, "learning_rate": 3.1178839954370634e-06, "loss": 0.426, "step": 6093 }, { "epoch": 1.9824333116460637, "grad_norm": 2.07550165619031, "learning_rate": 3.1161305927483686e-06, "loss": 0.4531, "step": 6094 }, { "epoch": 1.9827586206896552, "grad_norm": 1.9445705814269822, "learning_rate": 3.1143774600194176e-06, "loss": 0.4283, "step": 6095 }, { "epoch": 1.9830839297332465, "grad_norm": 2.1496064366829284, "learning_rate": 3.1126245975014336e-06, "loss": 0.4723, "step": 6096 }, { "epoch": 1.9834092387768378, "grad_norm": 1.9892443757925833, "learning_rate": 3.1108720054456043e-06, "loss": 0.3712, "step": 6097 }, { "epoch": 1.9837345478204296, "grad_norm": 2.345965181881103, "learning_rate": 3.109119684103073e-06, "loss": 0.5091, "step": 6098 }, { "epoch": 1.9840598568640209, "grad_norm": 1.706552276597287, "learning_rate": 3.107367633724954e-06, "loss": 0.393, "step": 6099 }, { "epoch": 1.9843851659076122, "grad_norm": 1.4246689700175845, "learning_rate": 3.1056158545623156e-06, "loss": 0.2625, "step": 6100 }, { "epoch": 1.9847104749512037, "grad_norm": 1.7749415143904343, "learning_rate": 3.103864346866187e-06, "loss": 0.4307, "step": 6101 }, { "epoch": 1.9850357839947952, "grad_norm": 1.7541694310047493, "learning_rate": 3.102113110887561e-06, "loss": 0.2847, "step": 6102 }, { "epoch": 1.9853610930383865, "grad_norm": 2.041780822194924, "learning_rate": 3.1003621468773924e-06, "loss": 0.4166, "step": 6103 }, { "epoch": 1.9856864020819778, "grad_norm": 1.624759018371765, "learning_rate": 3.098611455086595e-06, "loss": 0.3222, "step": 6104 }, { "epoch": 1.9860117111255693, "grad_norm": 1.6686937728040085, "learning_rate": 3.096861035766044e-06, "loss": 0.3434, "step": 6105 }, { "epoch": 1.9863370201691608, "grad_norm": 1.8907132953927381, "learning_rate": 3.0951108891665744e-06, "loss": 0.3887, "step": 6106 }, { "epoch": 1.986662329212752, "grad_norm": 1.7737131307824898, "learning_rate": 3.093361015538985e-06, "loss": 0.3771, "step": 6107 }, { "epoch": 1.9869876382563434, "grad_norm": 2.012806757641333, "learning_rate": 3.091611415134034e-06, "loss": 0.3663, "step": 6108 }, { "epoch": 1.987312947299935, "grad_norm": 2.332752335363794, "learning_rate": 3.0898620882024387e-06, "loss": 0.4276, "step": 6109 }, { "epoch": 1.9876382563435264, "grad_norm": 1.8723378187933437, "learning_rate": 3.088113034994879e-06, "loss": 0.3899, "step": 6110 }, { "epoch": 1.9879635653871177, "grad_norm": 1.9813031067714788, "learning_rate": 3.086364255761997e-06, "loss": 0.3436, "step": 6111 }, { "epoch": 1.988288874430709, "grad_norm": 1.7450723673228594, "learning_rate": 3.084615750754393e-06, "loss": 0.3578, "step": 6112 }, { "epoch": 1.9886141834743007, "grad_norm": 2.419846300122123, "learning_rate": 3.0828675202226304e-06, "loss": 0.4563, "step": 6113 }, { "epoch": 1.988939492517892, "grad_norm": 1.978606704264608, "learning_rate": 3.0811195644172275e-06, "loss": 0.4401, "step": 6114 }, { "epoch": 1.9892648015614833, "grad_norm": 2.409064989743217, "learning_rate": 3.079371883588672e-06, "loss": 0.4139, "step": 6115 }, { "epoch": 1.9895901106050748, "grad_norm": 2.543274080406597, "learning_rate": 3.077624477987407e-06, "loss": 0.4405, "step": 6116 }, { "epoch": 1.9899154196486664, "grad_norm": 1.8448766211441183, "learning_rate": 3.075877347863835e-06, "loss": 0.3775, "step": 6117 }, { "epoch": 1.9902407286922577, "grad_norm": 1.834054687205183, "learning_rate": 3.07413049346832e-06, "loss": 0.3499, "step": 6118 }, { "epoch": 1.990566037735849, "grad_norm": 1.6839822795832131, "learning_rate": 3.0723839150511923e-06, "loss": 0.2849, "step": 6119 }, { "epoch": 1.9908913467794405, "grad_norm": 1.8498839586502336, "learning_rate": 3.070637612862734e-06, "loss": 0.3748, "step": 6120 }, { "epoch": 1.991216655823032, "grad_norm": 1.9501759321185788, "learning_rate": 3.0688915871531923e-06, "loss": 0.4058, "step": 6121 }, { "epoch": 1.9915419648666233, "grad_norm": 2.115831085506233, "learning_rate": 3.0671458381727726e-06, "loss": 0.3655, "step": 6122 }, { "epoch": 1.9918672739102146, "grad_norm": 1.8556944547617629, "learning_rate": 3.0654003661716438e-06, "loss": 0.3904, "step": 6123 }, { "epoch": 1.992192582953806, "grad_norm": 1.8339153109255801, "learning_rate": 3.063655171399932e-06, "loss": 0.4051, "step": 6124 }, { "epoch": 1.9925178919973976, "grad_norm": 1.8036260327271818, "learning_rate": 3.0619102541077252e-06, "loss": 0.3137, "step": 6125 }, { "epoch": 1.9928432010409889, "grad_norm": 1.8017548953029496, "learning_rate": 3.0601656145450707e-06, "loss": 0.3768, "step": 6126 }, { "epoch": 1.9931685100845804, "grad_norm": 2.192733356629307, "learning_rate": 3.0584212529619777e-06, "loss": 0.438, "step": 6127 }, { "epoch": 1.993493819128172, "grad_norm": 2.051324498968965, "learning_rate": 3.056677169608414e-06, "loss": 0.489, "step": 6128 }, { "epoch": 1.9938191281717632, "grad_norm": 1.863794405793962, "learning_rate": 3.054933364734308e-06, "loss": 0.3358, "step": 6129 }, { "epoch": 1.9941444372153545, "grad_norm": 1.8125845669265621, "learning_rate": 3.053189838589545e-06, "loss": 0.3738, "step": 6130 }, { "epoch": 1.994469746258946, "grad_norm": 2.132117233058164, "learning_rate": 3.0514465914239775e-06, "loss": 0.344, "step": 6131 }, { "epoch": 1.9947950553025375, "grad_norm": 1.7408368822415752, "learning_rate": 3.0497036234874144e-06, "loss": 0.3783, "step": 6132 }, { "epoch": 1.9951203643461288, "grad_norm": 1.9321369783394446, "learning_rate": 3.0479609350296213e-06, "loss": 0.3566, "step": 6133 }, { "epoch": 1.9954456733897201, "grad_norm": 1.8958750707803418, "learning_rate": 3.046218526300326e-06, "loss": 0.3619, "step": 6134 }, { "epoch": 1.9957709824333116, "grad_norm": 1.7487227936207361, "learning_rate": 3.044476397549221e-06, "loss": 0.2768, "step": 6135 }, { "epoch": 1.9960962914769032, "grad_norm": 1.7571806229406817, "learning_rate": 3.042734549025951e-06, "loss": 0.3405, "step": 6136 }, { "epoch": 1.9964216005204944, "grad_norm": 2.02736717780459, "learning_rate": 3.0409929809801253e-06, "loss": 0.4656, "step": 6137 }, { "epoch": 1.9967469095640857, "grad_norm": 1.804531877890102, "learning_rate": 3.0392516936613102e-06, "loss": 0.3574, "step": 6138 }, { "epoch": 1.9970722186076773, "grad_norm": 1.9101318450537497, "learning_rate": 3.0375106873190363e-06, "loss": 0.4098, "step": 6139 }, { "epoch": 1.9973975276512688, "grad_norm": 1.8727343759553576, "learning_rate": 3.0357699622027885e-06, "loss": 0.3783, "step": 6140 }, { "epoch": 1.99772283669486, "grad_norm": 1.9194047543421167, "learning_rate": 3.0340295185620155e-06, "loss": 0.4001, "step": 6141 }, { "epoch": 1.9980481457384516, "grad_norm": 1.9585959818742968, "learning_rate": 3.032289356646122e-06, "loss": 0.3148, "step": 6142 }, { "epoch": 1.998373454782043, "grad_norm": 1.8370759560996073, "learning_rate": 3.0305494767044764e-06, "loss": 0.4096, "step": 6143 }, { "epoch": 1.9986987638256344, "grad_norm": 2.037567055404491, "learning_rate": 3.028809878986404e-06, "loss": 0.4053, "step": 6144 }, { "epoch": 1.9990240728692257, "grad_norm": 2.283250441124736, "learning_rate": 3.0270705637411914e-06, "loss": 0.4734, "step": 6145 }, { "epoch": 1.9993493819128172, "grad_norm": 2.0581910046503933, "learning_rate": 3.02533153121808e-06, "loss": 0.4443, "step": 6146 }, { "epoch": 1.9996746909564087, "grad_norm": 1.9548680042659001, "learning_rate": 3.0235927816662775e-06, "loss": 0.3725, "step": 6147 }, { "epoch": 2.0, "grad_norm": 1.9651046851364535, "learning_rate": 3.021854315334949e-06, "loss": 0.3856, "step": 6148 }, { "epoch": 2.0003253090435913, "grad_norm": 1.4231609014835733, "learning_rate": 3.020116132473214e-06, "loss": 0.1734, "step": 6149 }, { "epoch": 2.000650618087183, "grad_norm": 1.7979923344057835, "learning_rate": 3.018378233330156e-06, "loss": 0.2131, "step": 6150 }, { "epoch": 2.0009759271307743, "grad_norm": 1.5863579844997255, "learning_rate": 3.0166406181548213e-06, "loss": 0.2281, "step": 6151 }, { "epoch": 2.0013012361743656, "grad_norm": 1.507294514521167, "learning_rate": 3.014903287196208e-06, "loss": 0.1755, "step": 6152 }, { "epoch": 2.001626545217957, "grad_norm": 1.462658811876351, "learning_rate": 3.0131662407032764e-06, "loss": 0.1138, "step": 6153 }, { "epoch": 2.0019518542615486, "grad_norm": 1.5187910325780982, "learning_rate": 3.0114294789249465e-06, "loss": 0.1771, "step": 6154 }, { "epoch": 2.00227716330514, "grad_norm": 1.4615970210126188, "learning_rate": 3.009693002110099e-06, "loss": 0.1975, "step": 6155 }, { "epoch": 2.0026024723487312, "grad_norm": 1.712822821093461, "learning_rate": 3.0079568105075706e-06, "loss": 0.1785, "step": 6156 }, { "epoch": 2.0029277813923225, "grad_norm": 1.782502544536015, "learning_rate": 3.0062209043661606e-06, "loss": 0.2379, "step": 6157 }, { "epoch": 2.0032530904359143, "grad_norm": 2.1020909743768867, "learning_rate": 3.004485283934623e-06, "loss": 0.1703, "step": 6158 }, { "epoch": 2.0035783994795056, "grad_norm": 1.9180008854892032, "learning_rate": 3.002749949461675e-06, "loss": 0.1971, "step": 6159 }, { "epoch": 2.003903708523097, "grad_norm": 1.991769105293887, "learning_rate": 3.001014901195991e-06, "loss": 0.1863, "step": 6160 }, { "epoch": 2.004229017566688, "grad_norm": 2.0209070437187573, "learning_rate": 2.9992801393862037e-06, "loss": 0.1902, "step": 6161 }, { "epoch": 2.00455432661028, "grad_norm": 2.425484060977465, "learning_rate": 2.9975456642809076e-06, "loss": 0.2276, "step": 6162 }, { "epoch": 2.004879635653871, "grad_norm": 1.93751627457909, "learning_rate": 2.9958114761286528e-06, "loss": 0.1768, "step": 6163 }, { "epoch": 2.0052049446974625, "grad_norm": 2.0236006225132708, "learning_rate": 2.9940775751779493e-06, "loss": 0.1944, "step": 6164 }, { "epoch": 2.005530253741054, "grad_norm": 2.122470213972408, "learning_rate": 2.9923439616772666e-06, "loss": 0.1577, "step": 6165 }, { "epoch": 2.0058555627846455, "grad_norm": 2.2741302599082345, "learning_rate": 2.9906106358750333e-06, "loss": 0.1878, "step": 6166 }, { "epoch": 2.006180871828237, "grad_norm": 2.1501313206710995, "learning_rate": 2.988877598019636e-06, "loss": 0.1605, "step": 6167 }, { "epoch": 2.006506180871828, "grad_norm": 2.1474747430391172, "learning_rate": 2.9871448483594196e-06, "loss": 0.1588, "step": 6168 }, { "epoch": 2.00683148991542, "grad_norm": 2.2249861130420867, "learning_rate": 2.985412387142686e-06, "loss": 0.1672, "step": 6169 }, { "epoch": 2.007156798959011, "grad_norm": 2.349632712666689, "learning_rate": 2.9836802146177034e-06, "loss": 0.1233, "step": 6170 }, { "epoch": 2.0074821080026024, "grad_norm": 2.0203640619819097, "learning_rate": 2.9819483310326903e-06, "loss": 0.1477, "step": 6171 }, { "epoch": 2.0078074170461937, "grad_norm": 2.1801355757649916, "learning_rate": 2.9802167366358257e-06, "loss": 0.1785, "step": 6172 }, { "epoch": 2.0081327260897854, "grad_norm": 2.0914323551783114, "learning_rate": 2.978485431675249e-06, "loss": 0.1998, "step": 6173 }, { "epoch": 2.0084580351333767, "grad_norm": 2.3708879357073336, "learning_rate": 2.9767544163990587e-06, "loss": 0.2031, "step": 6174 }, { "epoch": 2.008783344176968, "grad_norm": 2.3573612622413256, "learning_rate": 2.975023691055309e-06, "loss": 0.2463, "step": 6175 }, { "epoch": 2.0091086532205593, "grad_norm": 1.7784667787369877, "learning_rate": 2.973293255892015e-06, "loss": 0.1246, "step": 6176 }, { "epoch": 2.009433962264151, "grad_norm": 1.7814904257339936, "learning_rate": 2.9715631111571477e-06, "loss": 0.1503, "step": 6177 }, { "epoch": 2.0097592713077423, "grad_norm": 2.139373390725075, "learning_rate": 2.9698332570986397e-06, "loss": 0.1542, "step": 6178 }, { "epoch": 2.0100845803513336, "grad_norm": 2.259734344833114, "learning_rate": 2.96810369396438e-06, "loss": 0.1915, "step": 6179 }, { "epoch": 2.0104098893949254, "grad_norm": 1.8295550005414374, "learning_rate": 2.966374422002215e-06, "loss": 0.1562, "step": 6180 }, { "epoch": 2.0107351984385167, "grad_norm": 2.2641047075525162, "learning_rate": 2.9646454414599507e-06, "loss": 0.1502, "step": 6181 }, { "epoch": 2.011060507482108, "grad_norm": 1.9932122372968235, "learning_rate": 2.9629167525853525e-06, "loss": 0.1418, "step": 6182 }, { "epoch": 2.0113858165256993, "grad_norm": 2.2037026802553563, "learning_rate": 2.9611883556261423e-06, "loss": 0.1695, "step": 6183 }, { "epoch": 2.011711125569291, "grad_norm": 2.009559703589048, "learning_rate": 2.9594602508300006e-06, "loss": 0.2057, "step": 6184 }, { "epoch": 2.0120364346128823, "grad_norm": 2.2446327807306874, "learning_rate": 2.957732438444562e-06, "loss": 0.1855, "step": 6185 }, { "epoch": 2.0123617436564736, "grad_norm": 2.196879023319338, "learning_rate": 2.95600491871743e-06, "loss": 0.198, "step": 6186 }, { "epoch": 2.012687052700065, "grad_norm": 1.9674734011113866, "learning_rate": 2.9542776918961557e-06, "loss": 0.1624, "step": 6187 }, { "epoch": 2.0130123617436566, "grad_norm": 1.8957635708262643, "learning_rate": 2.952550758228251e-06, "loss": 0.1371, "step": 6188 }, { "epoch": 2.013337670787248, "grad_norm": 2.048202838233245, "learning_rate": 2.9508241179611875e-06, "loss": 0.1718, "step": 6189 }, { "epoch": 2.013662979830839, "grad_norm": 1.7467492786236472, "learning_rate": 2.9490977713423948e-06, "loss": 0.1276, "step": 6190 }, { "epoch": 2.0139882888744305, "grad_norm": 2.3138662097961014, "learning_rate": 2.9473717186192586e-06, "loss": 0.1662, "step": 6191 }, { "epoch": 2.0143135979180222, "grad_norm": 2.1504620201705245, "learning_rate": 2.945645960039124e-06, "loss": 0.1644, "step": 6192 }, { "epoch": 2.0146389069616135, "grad_norm": 2.3300749733688204, "learning_rate": 2.943920495849292e-06, "loss": 0.1648, "step": 6193 }, { "epoch": 2.014964216005205, "grad_norm": 2.2337327204785025, "learning_rate": 2.942195326297025e-06, "loss": 0.197, "step": 6194 }, { "epoch": 2.0152895250487965, "grad_norm": 2.0606186541932776, "learning_rate": 2.9404704516295396e-06, "loss": 0.1826, "step": 6195 }, { "epoch": 2.015614834092388, "grad_norm": 2.465867128717332, "learning_rate": 2.9387458720940116e-06, "loss": 0.1547, "step": 6196 }, { "epoch": 2.015940143135979, "grad_norm": 1.9813545421163037, "learning_rate": 2.9370215879375728e-06, "loss": 0.1891, "step": 6197 }, { "epoch": 2.0162654521795704, "grad_norm": 2.130393823589203, "learning_rate": 2.9352975994073178e-06, "loss": 0.1973, "step": 6198 }, { "epoch": 2.016590761223162, "grad_norm": 2.2128804265019264, "learning_rate": 2.9335739067502935e-06, "loss": 0.17, "step": 6199 }, { "epoch": 2.0169160702667535, "grad_norm": 2.03683221355432, "learning_rate": 2.9318505102135066e-06, "loss": 0.213, "step": 6200 }, { "epoch": 2.0172413793103448, "grad_norm": 1.581385963598581, "learning_rate": 2.930127410043917e-06, "loss": 0.1456, "step": 6201 }, { "epoch": 2.017566688353936, "grad_norm": 2.251146716155198, "learning_rate": 2.9284046064884523e-06, "loss": 0.1934, "step": 6202 }, { "epoch": 2.017891997397528, "grad_norm": 1.9663780646458944, "learning_rate": 2.92668209979399e-06, "loss": 0.1462, "step": 6203 }, { "epoch": 2.018217306441119, "grad_norm": 2.2069485287372754, "learning_rate": 2.9249598902073633e-06, "loss": 0.1915, "step": 6204 }, { "epoch": 2.0185426154847104, "grad_norm": 1.9475417332161684, "learning_rate": 2.9232379779753663e-06, "loss": 0.1723, "step": 6205 }, { "epoch": 2.018867924528302, "grad_norm": 1.9788993271166784, "learning_rate": 2.9215163633447548e-06, "loss": 0.2063, "step": 6206 }, { "epoch": 2.0191932335718934, "grad_norm": 2.0024436284424256, "learning_rate": 2.9197950465622325e-06, "loss": 0.1511, "step": 6207 }, { "epoch": 2.0195185426154847, "grad_norm": 1.7760556133632055, "learning_rate": 2.918074027874468e-06, "loss": 0.1352, "step": 6208 }, { "epoch": 2.019843851659076, "grad_norm": 1.5030937545423089, "learning_rate": 2.9163533075280813e-06, "loss": 0.12, "step": 6209 }, { "epoch": 2.0201691607026677, "grad_norm": 1.8628096859807104, "learning_rate": 2.914632885769656e-06, "loss": 0.155, "step": 6210 }, { "epoch": 2.020494469746259, "grad_norm": 2.080267856752296, "learning_rate": 2.91291276284573e-06, "loss": 0.1628, "step": 6211 }, { "epoch": 2.0208197787898503, "grad_norm": 2.013069476743297, "learning_rate": 2.9111929390027966e-06, "loss": 0.1751, "step": 6212 }, { "epoch": 2.0211450878334416, "grad_norm": 2.0836553720981823, "learning_rate": 2.9094734144873037e-06, "loss": 0.1701, "step": 6213 }, { "epoch": 2.0214703968770333, "grad_norm": 1.6333399087380842, "learning_rate": 2.9077541895456684e-06, "loss": 0.1578, "step": 6214 }, { "epoch": 2.0217957059206246, "grad_norm": 1.9485507977845222, "learning_rate": 2.9060352644242528e-06, "loss": 0.1647, "step": 6215 }, { "epoch": 2.022121014964216, "grad_norm": 2.2395097656597196, "learning_rate": 2.9043166393693777e-06, "loss": 0.3087, "step": 6216 }, { "epoch": 2.022446324007807, "grad_norm": 2.0515527677286833, "learning_rate": 2.9025983146273253e-06, "loss": 0.1569, "step": 6217 }, { "epoch": 2.022771633051399, "grad_norm": 1.9569454182495518, "learning_rate": 2.9008802904443343e-06, "loss": 0.1629, "step": 6218 }, { "epoch": 2.0230969420949902, "grad_norm": 1.832896827583433, "learning_rate": 2.8991625670665967e-06, "loss": 0.154, "step": 6219 }, { "epoch": 2.0234222511385815, "grad_norm": 2.3559352638347395, "learning_rate": 2.897445144740265e-06, "loss": 0.182, "step": 6220 }, { "epoch": 2.0237475601821733, "grad_norm": 1.8905755204716468, "learning_rate": 2.895728023711444e-06, "loss": 0.1482, "step": 6221 }, { "epoch": 2.0240728692257646, "grad_norm": 1.819242805901791, "learning_rate": 2.8940112042262003e-06, "loss": 0.1511, "step": 6222 }, { "epoch": 2.024398178269356, "grad_norm": 2.1923835455178318, "learning_rate": 2.8922946865305567e-06, "loss": 0.1785, "step": 6223 }, { "epoch": 2.024723487312947, "grad_norm": 2.280554958834565, "learning_rate": 2.89057847087049e-06, "loss": 0.1637, "step": 6224 }, { "epoch": 2.025048796356539, "grad_norm": 1.9228977222791879, "learning_rate": 2.8888625574919318e-06, "loss": 0.1285, "step": 6225 }, { "epoch": 2.02537410540013, "grad_norm": 1.9117938109814432, "learning_rate": 2.8871469466407797e-06, "loss": 0.1763, "step": 6226 }, { "epoch": 2.0256994144437215, "grad_norm": 1.8509649298577684, "learning_rate": 2.885431638562879e-06, "loss": 0.1364, "step": 6227 }, { "epoch": 2.0260247234873128, "grad_norm": 2.1826287517318104, "learning_rate": 2.883716633504032e-06, "loss": 0.1806, "step": 6228 }, { "epoch": 2.0263500325309045, "grad_norm": 1.9859468551736106, "learning_rate": 2.882001931710003e-06, "loss": 0.1738, "step": 6229 }, { "epoch": 2.026675341574496, "grad_norm": 1.9038773588091173, "learning_rate": 2.8802875334265113e-06, "loss": 0.1294, "step": 6230 }, { "epoch": 2.027000650618087, "grad_norm": 2.0687534310725946, "learning_rate": 2.878573438899227e-06, "loss": 0.1815, "step": 6231 }, { "epoch": 2.0273259596616784, "grad_norm": 2.1890555321445957, "learning_rate": 2.8768596483737856e-06, "loss": 0.1705, "step": 6232 }, { "epoch": 2.02765126870527, "grad_norm": 1.878069487602967, "learning_rate": 2.87514616209577e-06, "loss": 0.184, "step": 6233 }, { "epoch": 2.0279765777488614, "grad_norm": 2.170338861537782, "learning_rate": 2.8734329803107252e-06, "loss": 0.184, "step": 6234 }, { "epoch": 2.0283018867924527, "grad_norm": 2.294216476226862, "learning_rate": 2.871720103264155e-06, "loss": 0.1924, "step": 6235 }, { "epoch": 2.0286271958360445, "grad_norm": 1.8108083218943614, "learning_rate": 2.870007531201512e-06, "loss": 0.1671, "step": 6236 }, { "epoch": 2.0289525048796357, "grad_norm": 1.9136377582399648, "learning_rate": 2.868295264368208e-06, "loss": 0.1228, "step": 6237 }, { "epoch": 2.029277813923227, "grad_norm": 1.9851751322062943, "learning_rate": 2.8665833030096167e-06, "loss": 0.1586, "step": 6238 }, { "epoch": 2.0296031229668183, "grad_norm": 2.0123794910963366, "learning_rate": 2.8648716473710593e-06, "loss": 0.1974, "step": 6239 }, { "epoch": 2.02992843201041, "grad_norm": 2.0913531004161086, "learning_rate": 2.8631602976978195e-06, "loss": 0.1928, "step": 6240 }, { "epoch": 2.0302537410540014, "grad_norm": 2.2076380499872, "learning_rate": 2.8614492542351317e-06, "loss": 0.155, "step": 6241 }, { "epoch": 2.0305790500975927, "grad_norm": 2.1276301974736045, "learning_rate": 2.859738517228192e-06, "loss": 0.1632, "step": 6242 }, { "epoch": 2.030904359141184, "grad_norm": 2.4205997999996973, "learning_rate": 2.8580280869221517e-06, "loss": 0.1959, "step": 6243 }, { "epoch": 2.0312296681847757, "grad_norm": 2.144777099784484, "learning_rate": 2.856317963562114e-06, "loss": 0.1784, "step": 6244 }, { "epoch": 2.031554977228367, "grad_norm": 2.313638477674105, "learning_rate": 2.8546081473931387e-06, "loss": 0.1902, "step": 6245 }, { "epoch": 2.0318802862719583, "grad_norm": 2.5912322286705898, "learning_rate": 2.8528986386602492e-06, "loss": 0.211, "step": 6246 }, { "epoch": 2.0322055953155496, "grad_norm": 2.3025473915535555, "learning_rate": 2.8511894376084166e-06, "loss": 0.1799, "step": 6247 }, { "epoch": 2.0325309043591413, "grad_norm": 2.3535170140454054, "learning_rate": 2.849480544482568e-06, "loss": 0.2236, "step": 6248 }, { "epoch": 2.0328562134027326, "grad_norm": 2.1144618219623665, "learning_rate": 2.847771959527592e-06, "loss": 0.1594, "step": 6249 }, { "epoch": 2.033181522446324, "grad_norm": 1.9127064391386581, "learning_rate": 2.8460636829883302e-06, "loss": 0.1531, "step": 6250 }, { "epoch": 2.0335068314899156, "grad_norm": 1.5927442721842655, "learning_rate": 2.8443557151095775e-06, "loss": 0.1477, "step": 6251 }, { "epoch": 2.033832140533507, "grad_norm": 1.9109660849820556, "learning_rate": 2.8426480561360893e-06, "loss": 0.1969, "step": 6252 }, { "epoch": 2.034157449577098, "grad_norm": 2.0828797699681623, "learning_rate": 2.8409407063125715e-06, "loss": 0.1889, "step": 6253 }, { "epoch": 2.0344827586206895, "grad_norm": 1.9409116086677916, "learning_rate": 2.83923366588369e-06, "loss": 0.1557, "step": 6254 }, { "epoch": 2.0348080676642812, "grad_norm": 2.2405855069653224, "learning_rate": 2.837526935094067e-06, "loss": 0.1715, "step": 6255 }, { "epoch": 2.0351333767078725, "grad_norm": 2.3836823489908276, "learning_rate": 2.8358205141882735e-06, "loss": 0.185, "step": 6256 }, { "epoch": 2.035458685751464, "grad_norm": 2.0344041638330306, "learning_rate": 2.8341144034108435e-06, "loss": 0.1568, "step": 6257 }, { "epoch": 2.035783994795055, "grad_norm": 2.292909918296878, "learning_rate": 2.8324086030062657e-06, "loss": 0.188, "step": 6258 }, { "epoch": 2.036109303838647, "grad_norm": 2.140472160021825, "learning_rate": 2.8307031132189775e-06, "loss": 0.1852, "step": 6259 }, { "epoch": 2.036434612882238, "grad_norm": 2.4403771176141764, "learning_rate": 2.828997934293383e-06, "loss": 0.1843, "step": 6260 }, { "epoch": 2.0367599219258294, "grad_norm": 2.2518548213053844, "learning_rate": 2.8272930664738286e-06, "loss": 0.241, "step": 6261 }, { "epoch": 2.0370852309694207, "grad_norm": 2.1131798817710306, "learning_rate": 2.8255885100046266e-06, "loss": 0.1572, "step": 6262 }, { "epoch": 2.0374105400130125, "grad_norm": 2.002745673483671, "learning_rate": 2.8238842651300435e-06, "loss": 0.1651, "step": 6263 }, { "epoch": 2.0377358490566038, "grad_norm": 2.2031175239857625, "learning_rate": 2.822180332094295e-06, "loss": 0.1805, "step": 6264 }, { "epoch": 2.038061158100195, "grad_norm": 2.191092735242067, "learning_rate": 2.8204767111415564e-06, "loss": 0.1357, "step": 6265 }, { "epoch": 2.038386467143787, "grad_norm": 2.4865335436679565, "learning_rate": 2.8187734025159573e-06, "loss": 0.225, "step": 6266 }, { "epoch": 2.038711776187378, "grad_norm": 2.0773423128726565, "learning_rate": 2.817070406461586e-06, "loss": 0.1729, "step": 6267 }, { "epoch": 2.0390370852309694, "grad_norm": 2.2245284196460844, "learning_rate": 2.8153677232224792e-06, "loss": 0.1746, "step": 6268 }, { "epoch": 2.0393623942745607, "grad_norm": 1.8221291087861877, "learning_rate": 2.813665353042634e-06, "loss": 0.1378, "step": 6269 }, { "epoch": 2.0396877033181524, "grad_norm": 2.0673491533199058, "learning_rate": 2.811963296166004e-06, "loss": 0.177, "step": 6270 }, { "epoch": 2.0400130123617437, "grad_norm": 1.6879862329346516, "learning_rate": 2.81026155283649e-06, "loss": 0.1404, "step": 6271 }, { "epoch": 2.040338321405335, "grad_norm": 2.158497587682642, "learning_rate": 2.8085601232979565e-06, "loss": 0.1978, "step": 6272 }, { "epoch": 2.0406636304489263, "grad_norm": 1.876814899914989, "learning_rate": 2.806859007794217e-06, "loss": 0.1504, "step": 6273 }, { "epoch": 2.040988939492518, "grad_norm": 1.8819345424173046, "learning_rate": 2.805158206569045e-06, "loss": 0.1972, "step": 6274 }, { "epoch": 2.0413142485361093, "grad_norm": 2.2247148378259016, "learning_rate": 2.803457719866166e-06, "loss": 0.2079, "step": 6275 }, { "epoch": 2.0416395575797006, "grad_norm": 2.286033242335039, "learning_rate": 2.8017575479292615e-06, "loss": 0.1878, "step": 6276 }, { "epoch": 2.0419648666232924, "grad_norm": 2.823092661791541, "learning_rate": 2.8000576910019605e-06, "loss": 0.1953, "step": 6277 }, { "epoch": 2.0422901756668836, "grad_norm": 2.249692289966719, "learning_rate": 2.7983581493278643e-06, "loss": 0.2356, "step": 6278 }, { "epoch": 2.042615484710475, "grad_norm": 2.1879290050844835, "learning_rate": 2.7966589231505127e-06, "loss": 0.1374, "step": 6279 }, { "epoch": 2.0429407937540662, "grad_norm": 1.830766172350758, "learning_rate": 2.7949600127134047e-06, "loss": 0.1386, "step": 6280 }, { "epoch": 2.043266102797658, "grad_norm": 2.5168655218256557, "learning_rate": 2.793261418259997e-06, "loss": 0.1799, "step": 6281 }, { "epoch": 2.0435914118412493, "grad_norm": 2.2432444220591656, "learning_rate": 2.791563140033701e-06, "loss": 0.2184, "step": 6282 }, { "epoch": 2.0439167208848406, "grad_norm": 2.2373492913184174, "learning_rate": 2.789865178277878e-06, "loss": 0.1803, "step": 6283 }, { "epoch": 2.044242029928432, "grad_norm": 2.36333327892882, "learning_rate": 2.7881675332358505e-06, "loss": 0.1633, "step": 6284 }, { "epoch": 2.0445673389720236, "grad_norm": 2.2499023532530025, "learning_rate": 2.7864702051508872e-06, "loss": 0.1669, "step": 6285 }, { "epoch": 2.044892648015615, "grad_norm": 1.9678577303108618, "learning_rate": 2.7847731942662203e-06, "loss": 0.1798, "step": 6286 }, { "epoch": 2.045217957059206, "grad_norm": 2.077888005370918, "learning_rate": 2.7830765008250338e-06, "loss": 0.1561, "step": 6287 }, { "epoch": 2.0455432661027975, "grad_norm": 2.260872340655569, "learning_rate": 2.7813801250704604e-06, "loss": 0.1734, "step": 6288 }, { "epoch": 2.045868575146389, "grad_norm": 2.068556979775902, "learning_rate": 2.7796840672455944e-06, "loss": 0.1728, "step": 6289 }, { "epoch": 2.0461938841899805, "grad_norm": 1.9759663171137636, "learning_rate": 2.777988327593484e-06, "loss": 0.1712, "step": 6290 }, { "epoch": 2.046519193233572, "grad_norm": 2.18229865406413, "learning_rate": 2.7762929063571255e-06, "loss": 0.1736, "step": 6291 }, { "epoch": 2.046844502277163, "grad_norm": 1.9261755713626316, "learning_rate": 2.7745978037794764e-06, "loss": 0.1405, "step": 6292 }, { "epoch": 2.047169811320755, "grad_norm": 2.025316550816365, "learning_rate": 2.7729030201034474e-06, "loss": 0.171, "step": 6293 }, { "epoch": 2.047495120364346, "grad_norm": 2.0119920035248913, "learning_rate": 2.771208555571898e-06, "loss": 0.1734, "step": 6294 }, { "epoch": 2.0478204294079374, "grad_norm": 1.9074815605243214, "learning_rate": 2.7695144104276516e-06, "loss": 0.1573, "step": 6295 }, { "epoch": 2.048145738451529, "grad_norm": 2.2105077902138297, "learning_rate": 2.767820584913475e-06, "loss": 0.1771, "step": 6296 }, { "epoch": 2.0484710474951204, "grad_norm": 2.4177719346361726, "learning_rate": 2.7661270792720968e-06, "loss": 0.1828, "step": 6297 }, { "epoch": 2.0487963565387117, "grad_norm": 2.0387136648836686, "learning_rate": 2.7644338937461994e-06, "loss": 0.16, "step": 6298 }, { "epoch": 2.049121665582303, "grad_norm": 1.9954708593677095, "learning_rate": 2.7627410285784164e-06, "loss": 0.1491, "step": 6299 }, { "epoch": 2.0494469746258948, "grad_norm": 1.9527587560500967, "learning_rate": 2.7610484840113304e-06, "loss": 0.1549, "step": 6300 }, { "epoch": 2.049772283669486, "grad_norm": 2.3197046338265523, "learning_rate": 2.759356260287494e-06, "loss": 0.1744, "step": 6301 }, { "epoch": 2.0500975927130773, "grad_norm": 2.33539386398301, "learning_rate": 2.7576643576493995e-06, "loss": 0.278, "step": 6302 }, { "epoch": 2.0504229017566686, "grad_norm": 2.2442637845312503, "learning_rate": 2.755972776339495e-06, "loss": 0.1856, "step": 6303 }, { "epoch": 2.0507482108002604, "grad_norm": 1.9384157325391203, "learning_rate": 2.7542815166001884e-06, "loss": 0.1548, "step": 6304 }, { "epoch": 2.0510735198438517, "grad_norm": 1.9463997721649882, "learning_rate": 2.7525905786738393e-06, "loss": 0.1889, "step": 6305 }, { "epoch": 2.051398828887443, "grad_norm": 1.867956275245701, "learning_rate": 2.7508999628027575e-06, "loss": 0.1646, "step": 6306 }, { "epoch": 2.0517241379310347, "grad_norm": 2.229223648715864, "learning_rate": 2.749209669229212e-06, "loss": 0.1565, "step": 6307 }, { "epoch": 2.052049446974626, "grad_norm": 1.9854382019552208, "learning_rate": 2.74751969819542e-06, "loss": 0.1683, "step": 6308 }, { "epoch": 2.0523747560182173, "grad_norm": 1.957423290168689, "learning_rate": 2.7458300499435563e-06, "loss": 0.2118, "step": 6309 }, { "epoch": 2.0527000650618086, "grad_norm": 2.3079151229739505, "learning_rate": 2.7441407247157514e-06, "loss": 0.189, "step": 6310 }, { "epoch": 2.0530253741054003, "grad_norm": 1.8758482042912068, "learning_rate": 2.7424517227540855e-06, "loss": 0.16, "step": 6311 }, { "epoch": 2.0533506831489916, "grad_norm": 2.234403915557838, "learning_rate": 2.7407630443005886e-06, "loss": 0.1444, "step": 6312 }, { "epoch": 2.053675992192583, "grad_norm": 2.3626971288211482, "learning_rate": 2.7390746895972575e-06, "loss": 0.1636, "step": 6313 }, { "epoch": 2.054001301236174, "grad_norm": 2.16152734450669, "learning_rate": 2.737386658886031e-06, "loss": 0.2273, "step": 6314 }, { "epoch": 2.054326610279766, "grad_norm": 1.9912330286502473, "learning_rate": 2.735698952408802e-06, "loss": 0.1559, "step": 6315 }, { "epoch": 2.0546519193233572, "grad_norm": 2.2579889568212366, "learning_rate": 2.734011570407424e-06, "loss": 0.1507, "step": 6316 }, { "epoch": 2.0549772283669485, "grad_norm": 1.9894863067015554, "learning_rate": 2.7323245131237002e-06, "loss": 0.1272, "step": 6317 }, { "epoch": 2.05530253741054, "grad_norm": 2.034054659049985, "learning_rate": 2.7306377807993835e-06, "loss": 0.1306, "step": 6318 }, { "epoch": 2.0556278464541315, "grad_norm": 1.9835921784732686, "learning_rate": 2.7289513736761887e-06, "loss": 0.2114, "step": 6319 }, { "epoch": 2.055953155497723, "grad_norm": 2.2978819725694573, "learning_rate": 2.7272652919957733e-06, "loss": 0.2189, "step": 6320 }, { "epoch": 2.056278464541314, "grad_norm": 1.7616690762177172, "learning_rate": 2.7255795359997574e-06, "loss": 0.1198, "step": 6321 }, { "epoch": 2.056603773584906, "grad_norm": 1.954986474847796, "learning_rate": 2.723894105929712e-06, "loss": 0.14, "step": 6322 }, { "epoch": 2.056929082628497, "grad_norm": 2.05486669985315, "learning_rate": 2.7222090020271565e-06, "loss": 0.1391, "step": 6323 }, { "epoch": 2.0572543916720885, "grad_norm": 2.311858287981562, "learning_rate": 2.72052422453357e-06, "loss": 0.1735, "step": 6324 }, { "epoch": 2.0575797007156797, "grad_norm": 2.1418490151396323, "learning_rate": 2.718839773690384e-06, "loss": 0.1556, "step": 6325 }, { "epoch": 2.0579050097592715, "grad_norm": 2.345714689225114, "learning_rate": 2.7171556497389773e-06, "loss": 0.1757, "step": 6326 }, { "epoch": 2.058230318802863, "grad_norm": 2.696847378520471, "learning_rate": 2.71547185292069e-06, "loss": 0.2135, "step": 6327 }, { "epoch": 2.058555627846454, "grad_norm": 1.8941362517869993, "learning_rate": 2.7137883834768076e-06, "loss": 0.1571, "step": 6328 }, { "epoch": 2.0588809368900454, "grad_norm": 2.0733184790738917, "learning_rate": 2.7121052416485737e-06, "loss": 0.1703, "step": 6329 }, { "epoch": 2.059206245933637, "grad_norm": 2.0268276795190525, "learning_rate": 2.710422427677187e-06, "loss": 0.1966, "step": 6330 }, { "epoch": 2.0595315549772284, "grad_norm": 2.0237395970415073, "learning_rate": 2.708739941803792e-06, "loss": 0.1819, "step": 6331 }, { "epoch": 2.0598568640208197, "grad_norm": 2.201637514855914, "learning_rate": 2.7070577842694874e-06, "loss": 0.158, "step": 6332 }, { "epoch": 2.060182173064411, "grad_norm": 2.2883557032075994, "learning_rate": 2.705375955315335e-06, "loss": 0.188, "step": 6333 }, { "epoch": 2.0605074821080027, "grad_norm": 2.0527250232804857, "learning_rate": 2.7036944551823373e-06, "loss": 0.1496, "step": 6334 }, { "epoch": 2.060832791151594, "grad_norm": 2.1980098143955695, "learning_rate": 2.7020132841114543e-06, "loss": 0.1827, "step": 6335 }, { "epoch": 2.0611581001951853, "grad_norm": 2.0947712825387033, "learning_rate": 2.7003324423435993e-06, "loss": 0.1609, "step": 6336 }, { "epoch": 2.061483409238777, "grad_norm": 2.5108854852754234, "learning_rate": 2.6986519301196413e-06, "loss": 0.2162, "step": 6337 }, { "epoch": 2.0618087182823683, "grad_norm": 2.0482656317993557, "learning_rate": 2.6969717476803935e-06, "loss": 0.1899, "step": 6338 }, { "epoch": 2.0621340273259596, "grad_norm": 2.0906449575505546, "learning_rate": 2.6952918952666318e-06, "loss": 0.21, "step": 6339 }, { "epoch": 2.062459336369551, "grad_norm": 1.9827305705616254, "learning_rate": 2.6936123731190757e-06, "loss": 0.1694, "step": 6340 }, { "epoch": 2.0627846454131427, "grad_norm": 1.9363431578267782, "learning_rate": 2.6919331814784046e-06, "loss": 0.1521, "step": 6341 }, { "epoch": 2.063109954456734, "grad_norm": 1.8795155087797597, "learning_rate": 2.6902543205852496e-06, "loss": 0.1531, "step": 6342 }, { "epoch": 2.0634352635003252, "grad_norm": 2.154949615772764, "learning_rate": 2.6885757906801887e-06, "loss": 0.1312, "step": 6343 }, { "epoch": 2.0637605725439165, "grad_norm": 1.9410257228184553, "learning_rate": 2.6868975920037577e-06, "loss": 0.1696, "step": 6344 }, { "epoch": 2.0640858815875083, "grad_norm": 2.1154961661230063, "learning_rate": 2.685219724796446e-06, "loss": 0.1531, "step": 6345 }, { "epoch": 2.0644111906310996, "grad_norm": 2.3324677320584803, "learning_rate": 2.683542189298689e-06, "loss": 0.2135, "step": 6346 }, { "epoch": 2.064736499674691, "grad_norm": 2.2619300608181026, "learning_rate": 2.681864985750883e-06, "loss": 0.2231, "step": 6347 }, { "epoch": 2.065061808718282, "grad_norm": 1.8790343127979758, "learning_rate": 2.6801881143933677e-06, "loss": 0.1737, "step": 6348 }, { "epoch": 2.065387117761874, "grad_norm": 2.1012351435315137, "learning_rate": 2.678511575466445e-06, "loss": 0.1863, "step": 6349 }, { "epoch": 2.065712426805465, "grad_norm": 1.9125395299264958, "learning_rate": 2.6768353692103583e-06, "loss": 0.1313, "step": 6350 }, { "epoch": 2.0660377358490565, "grad_norm": 2.05777047307999, "learning_rate": 2.675159495865315e-06, "loss": 0.1569, "step": 6351 }, { "epoch": 2.066363044892648, "grad_norm": 2.0337576802006296, "learning_rate": 2.673483955671463e-06, "loss": 0.1749, "step": 6352 }, { "epoch": 2.0666883539362395, "grad_norm": 2.324132031713658, "learning_rate": 2.6718087488689126e-06, "loss": 0.171, "step": 6353 }, { "epoch": 2.067013662979831, "grad_norm": 2.0445545127297837, "learning_rate": 2.670133875697722e-06, "loss": 0.1499, "step": 6354 }, { "epoch": 2.067338972023422, "grad_norm": 1.9156592959051992, "learning_rate": 2.668459336397898e-06, "loss": 0.1474, "step": 6355 }, { "epoch": 2.067664281067014, "grad_norm": 2.207255889664797, "learning_rate": 2.666785131209406e-06, "loss": 0.144, "step": 6356 }, { "epoch": 2.067989590110605, "grad_norm": 2.2589868633449792, "learning_rate": 2.665111260372163e-06, "loss": 0.1865, "step": 6357 }, { "epoch": 2.0683148991541964, "grad_norm": 2.092326098788519, "learning_rate": 2.6634377241260307e-06, "loss": 0.1835, "step": 6358 }, { "epoch": 2.0686402081977877, "grad_norm": 2.111013743188773, "learning_rate": 2.6617645227108325e-06, "loss": 0.1489, "step": 6359 }, { "epoch": 2.0689655172413794, "grad_norm": 1.8824223767504855, "learning_rate": 2.660091656366335e-06, "loss": 0.1656, "step": 6360 }, { "epoch": 2.0692908262849707, "grad_norm": 2.10443707568768, "learning_rate": 2.6584191253322627e-06, "loss": 0.1927, "step": 6361 }, { "epoch": 2.069616135328562, "grad_norm": 2.5735174910029994, "learning_rate": 2.6567469298482935e-06, "loss": 0.2125, "step": 6362 }, { "epoch": 2.0699414443721533, "grad_norm": 2.4522677849808066, "learning_rate": 2.6550750701540518e-06, "loss": 0.1916, "step": 6363 }, { "epoch": 2.070266753415745, "grad_norm": 1.9708955915351176, "learning_rate": 2.653403546489112e-06, "loss": 0.1613, "step": 6364 }, { "epoch": 2.0705920624593364, "grad_norm": 2.0316945271360374, "learning_rate": 2.6517323590930115e-06, "loss": 0.1811, "step": 6365 }, { "epoch": 2.0709173715029277, "grad_norm": 1.6668552621107462, "learning_rate": 2.6500615082052305e-06, "loss": 0.1151, "step": 6366 }, { "epoch": 2.0712426805465194, "grad_norm": 1.9091535551061793, "learning_rate": 2.6483909940651996e-06, "loss": 0.1211, "step": 6367 }, { "epoch": 2.0715679895901107, "grad_norm": 2.7473304515388333, "learning_rate": 2.6467208169123076e-06, "loss": 0.1762, "step": 6368 }, { "epoch": 2.071893298633702, "grad_norm": 2.1223247073126945, "learning_rate": 2.6450509769858933e-06, "loss": 0.1899, "step": 6369 }, { "epoch": 2.0722186076772933, "grad_norm": 1.973046625190744, "learning_rate": 2.6433814745252425e-06, "loss": 0.1913, "step": 6370 }, { "epoch": 2.072543916720885, "grad_norm": 1.9854257715437142, "learning_rate": 2.641712309769599e-06, "loss": 0.1601, "step": 6371 }, { "epoch": 2.0728692257644763, "grad_norm": 1.9070394608455656, "learning_rate": 2.6400434829581523e-06, "loss": 0.143, "step": 6372 }, { "epoch": 2.0731945348080676, "grad_norm": 2.2166447113440975, "learning_rate": 2.638374994330048e-06, "loss": 0.1839, "step": 6373 }, { "epoch": 2.073519843851659, "grad_norm": 2.0043336610175038, "learning_rate": 2.6367068441243838e-06, "loss": 0.1608, "step": 6374 }, { "epoch": 2.0738451528952506, "grad_norm": 1.8769969758767044, "learning_rate": 2.635039032580202e-06, "loss": 0.1726, "step": 6375 }, { "epoch": 2.074170461938842, "grad_norm": 2.311628959758838, "learning_rate": 2.6333715599365033e-06, "loss": 0.1991, "step": 6376 }, { "epoch": 2.074495770982433, "grad_norm": 2.234716282604315, "learning_rate": 2.6317044264322404e-06, "loss": 0.1931, "step": 6377 }, { "epoch": 2.074821080026025, "grad_norm": 2.3912338949796212, "learning_rate": 2.6300376323063104e-06, "loss": 0.1773, "step": 6378 }, { "epoch": 2.0751463890696162, "grad_norm": 1.8142734999218024, "learning_rate": 2.6283711777975697e-06, "loss": 0.1744, "step": 6379 }, { "epoch": 2.0754716981132075, "grad_norm": 2.0018656583705003, "learning_rate": 2.626705063144818e-06, "loss": 0.131, "step": 6380 }, { "epoch": 2.075797007156799, "grad_norm": 2.227539036300819, "learning_rate": 2.6250392885868135e-06, "loss": 0.176, "step": 6381 }, { "epoch": 2.0761223162003906, "grad_norm": 2.469536991465773, "learning_rate": 2.6233738543622645e-06, "loss": 0.1924, "step": 6382 }, { "epoch": 2.076447625243982, "grad_norm": 2.2738088014993396, "learning_rate": 2.621708760709827e-06, "loss": 0.2442, "step": 6383 }, { "epoch": 2.076772934287573, "grad_norm": 1.9204345293000582, "learning_rate": 2.6200440078681054e-06, "loss": 0.1601, "step": 6384 }, { "epoch": 2.0770982433311644, "grad_norm": 2.0949322807666175, "learning_rate": 2.618379596075668e-06, "loss": 0.1495, "step": 6385 }, { "epoch": 2.077423552374756, "grad_norm": 1.8758842642005025, "learning_rate": 2.616715525571023e-06, "loss": 0.1526, "step": 6386 }, { "epoch": 2.0777488614183475, "grad_norm": 1.8686016944227657, "learning_rate": 2.6150517965926294e-06, "loss": 0.1828, "step": 6387 }, { "epoch": 2.0780741704619388, "grad_norm": 2.1973934092092624, "learning_rate": 2.613388409378904e-06, "loss": 0.194, "step": 6388 }, { "epoch": 2.07839947950553, "grad_norm": 1.843076103426068, "learning_rate": 2.6117253641682127e-06, "loss": 0.159, "step": 6389 }, { "epoch": 2.078724788549122, "grad_norm": 2.466775730190607, "learning_rate": 2.6100626611988666e-06, "loss": 0.1397, "step": 6390 }, { "epoch": 2.079050097592713, "grad_norm": 2.1589561858737936, "learning_rate": 2.6084003007091374e-06, "loss": 0.1591, "step": 6391 }, { "epoch": 2.0793754066363044, "grad_norm": 1.8607412746567369, "learning_rate": 2.6067382829372374e-06, "loss": 0.1482, "step": 6392 }, { "epoch": 2.0797007156798957, "grad_norm": 2.225185512152892, "learning_rate": 2.6050766081213367e-06, "loss": 0.1603, "step": 6393 }, { "epoch": 2.0800260247234874, "grad_norm": 1.9161612939116706, "learning_rate": 2.603415276499557e-06, "loss": 0.1612, "step": 6394 }, { "epoch": 2.0803513337670787, "grad_norm": 2.0674668980381528, "learning_rate": 2.601754288309966e-06, "loss": 0.1704, "step": 6395 }, { "epoch": 2.08067664281067, "grad_norm": 1.6644655859439055, "learning_rate": 2.600093643790581e-06, "loss": 0.1392, "step": 6396 }, { "epoch": 2.0810019518542617, "grad_norm": 2.029785187955629, "learning_rate": 2.5984333431793807e-06, "loss": 0.1788, "step": 6397 }, { "epoch": 2.081327260897853, "grad_norm": 2.6779570052283765, "learning_rate": 2.596773386714283e-06, "loss": 0.1561, "step": 6398 }, { "epoch": 2.0816525699414443, "grad_norm": 2.0195601258308575, "learning_rate": 2.5951137746331602e-06, "loss": 0.1811, "step": 6399 }, { "epoch": 2.0819778789850356, "grad_norm": 2.1146655960303526, "learning_rate": 2.593454507173836e-06, "loss": 0.2233, "step": 6400 }, { "epoch": 2.0823031880286273, "grad_norm": 2.2140181735492273, "learning_rate": 2.5917955845740876e-06, "loss": 0.1668, "step": 6401 }, { "epoch": 2.0826284970722186, "grad_norm": 2.3124024841359185, "learning_rate": 2.5901370070716358e-06, "loss": 0.1835, "step": 6402 }, { "epoch": 2.08295380611581, "grad_norm": 2.046527280753928, "learning_rate": 2.5884787749041596e-06, "loss": 0.1492, "step": 6403 }, { "epoch": 2.0832791151594012, "grad_norm": 2.0638971926212886, "learning_rate": 2.5868208883092804e-06, "loss": 0.2027, "step": 6404 }, { "epoch": 2.083604424202993, "grad_norm": 2.1654528899493437, "learning_rate": 2.5851633475245776e-06, "loss": 0.1579, "step": 6405 }, { "epoch": 2.0839297332465843, "grad_norm": 2.138563131609804, "learning_rate": 2.5835061527875782e-06, "loss": 0.2098, "step": 6406 }, { "epoch": 2.0842550422901756, "grad_norm": 2.2791725600570953, "learning_rate": 2.581849304335757e-06, "loss": 0.1845, "step": 6407 }, { "epoch": 2.0845803513337673, "grad_norm": 2.1906565052475493, "learning_rate": 2.5801928024065426e-06, "loss": 0.1059, "step": 6408 }, { "epoch": 2.0849056603773586, "grad_norm": 2.1006486489178795, "learning_rate": 2.5785366472373154e-06, "loss": 0.1908, "step": 6409 }, { "epoch": 2.08523096942095, "grad_norm": 2.018714274992726, "learning_rate": 2.576880839065398e-06, "loss": 0.1253, "step": 6410 }, { "epoch": 2.085556278464541, "grad_norm": 2.4261247952139033, "learning_rate": 2.5752253781280746e-06, "loss": 0.1593, "step": 6411 }, { "epoch": 2.085881587508133, "grad_norm": 1.6814423759994088, "learning_rate": 2.5735702646625694e-06, "loss": 0.1322, "step": 6412 }, { "epoch": 2.086206896551724, "grad_norm": 2.025658178128804, "learning_rate": 2.5719154989060622e-06, "loss": 0.1662, "step": 6413 }, { "epoch": 2.0865322055953155, "grad_norm": 1.8913049108941382, "learning_rate": 2.570261081095686e-06, "loss": 0.1422, "step": 6414 }, { "epoch": 2.086857514638907, "grad_norm": 2.404243546019954, "learning_rate": 2.5686070114685142e-06, "loss": 0.1626, "step": 6415 }, { "epoch": 2.0871828236824985, "grad_norm": 2.198248228306795, "learning_rate": 2.566953290261579e-06, "loss": 0.1751, "step": 6416 }, { "epoch": 2.08750813272609, "grad_norm": 2.3796261636650407, "learning_rate": 2.5652999177118607e-06, "loss": 0.1539, "step": 6417 }, { "epoch": 2.087833441769681, "grad_norm": 1.8036451106115605, "learning_rate": 2.5636468940562887e-06, "loss": 0.1352, "step": 6418 }, { "epoch": 2.0881587508132724, "grad_norm": 2.241884961717518, "learning_rate": 2.561994219531736e-06, "loss": 0.1638, "step": 6419 }, { "epoch": 2.088484059856864, "grad_norm": 2.4795626057738183, "learning_rate": 2.560341894375042e-06, "loss": 0.1726, "step": 6420 }, { "epoch": 2.0888093689004554, "grad_norm": 1.9338532500374837, "learning_rate": 2.558689918822981e-06, "loss": 0.1085, "step": 6421 }, { "epoch": 2.0891346779440467, "grad_norm": 1.8993686463845016, "learning_rate": 2.55703829311228e-06, "loss": 0.1157, "step": 6422 }, { "epoch": 2.0894599869876385, "grad_norm": 2.3328217292986966, "learning_rate": 2.55538701747962e-06, "loss": 0.2082, "step": 6423 }, { "epoch": 2.0897852960312298, "grad_norm": 2.4259905903502585, "learning_rate": 2.5537360921616332e-06, "loss": 0.2097, "step": 6424 }, { "epoch": 2.090110605074821, "grad_norm": 2.2362226434253256, "learning_rate": 2.5520855173948935e-06, "loss": 0.2053, "step": 6425 }, { "epoch": 2.0904359141184123, "grad_norm": 2.007745634390011, "learning_rate": 2.5504352934159328e-06, "loss": 0.1702, "step": 6426 }, { "epoch": 2.090761223162004, "grad_norm": 2.418023616779451, "learning_rate": 2.5487854204612252e-06, "loss": 0.1706, "step": 6427 }, { "epoch": 2.0910865322055954, "grad_norm": 1.9701593460279716, "learning_rate": 2.547135898767202e-06, "loss": 0.1606, "step": 6428 }, { "epoch": 2.0914118412491867, "grad_norm": 2.498656537387375, "learning_rate": 2.545486728570241e-06, "loss": 0.2202, "step": 6429 }, { "epoch": 2.091737150292778, "grad_norm": 2.4994257737410854, "learning_rate": 2.5438379101066675e-06, "loss": 0.2078, "step": 6430 }, { "epoch": 2.0920624593363697, "grad_norm": 2.4120546353171006, "learning_rate": 2.542189443612758e-06, "loss": 0.2017, "step": 6431 }, { "epoch": 2.092387768379961, "grad_norm": 1.6610760107202334, "learning_rate": 2.5405413293247427e-06, "loss": 0.1094, "step": 6432 }, { "epoch": 2.0927130774235523, "grad_norm": 2.010770964554526, "learning_rate": 2.538893567478792e-06, "loss": 0.1563, "step": 6433 }, { "epoch": 2.0930383864671436, "grad_norm": 1.8696447658327786, "learning_rate": 2.5372461583110364e-06, "loss": 0.1188, "step": 6434 }, { "epoch": 2.0933636955107353, "grad_norm": 2.1536752214160275, "learning_rate": 2.5355991020575464e-06, "loss": 0.1694, "step": 6435 }, { "epoch": 2.0936890045543266, "grad_norm": 2.186627820892039, "learning_rate": 2.53395239895435e-06, "loss": 0.1919, "step": 6436 }, { "epoch": 2.094014313597918, "grad_norm": 2.106262640375767, "learning_rate": 2.5323060492374175e-06, "loss": 0.1878, "step": 6437 }, { "epoch": 2.0943396226415096, "grad_norm": 2.1184652595963724, "learning_rate": 2.530660053142675e-06, "loss": 0.1541, "step": 6438 }, { "epoch": 2.094664931685101, "grad_norm": 2.112463040673879, "learning_rate": 2.529014410905991e-06, "loss": 0.1736, "step": 6439 }, { "epoch": 2.094990240728692, "grad_norm": 2.130770359748616, "learning_rate": 2.5273691227631895e-06, "loss": 0.1658, "step": 6440 }, { "epoch": 2.0953155497722835, "grad_norm": 2.0427579056167247, "learning_rate": 2.5257241889500435e-06, "loss": 0.1717, "step": 6441 }, { "epoch": 2.0956408588158753, "grad_norm": 1.9812588889957896, "learning_rate": 2.5240796097022692e-06, "loss": 0.1585, "step": 6442 }, { "epoch": 2.0959661678594665, "grad_norm": 2.3364523927585603, "learning_rate": 2.522435385255537e-06, "loss": 0.2156, "step": 6443 }, { "epoch": 2.096291476903058, "grad_norm": 2.4735927612248028, "learning_rate": 2.5207915158454675e-06, "loss": 0.1946, "step": 6444 }, { "epoch": 2.096616785946649, "grad_norm": 2.469035511941869, "learning_rate": 2.519148001707626e-06, "loss": 0.1823, "step": 6445 }, { "epoch": 2.096942094990241, "grad_norm": 1.8390726302322187, "learning_rate": 2.5175048430775314e-06, "loss": 0.1656, "step": 6446 }, { "epoch": 2.097267404033832, "grad_norm": 2.2770608528073746, "learning_rate": 2.5158620401906454e-06, "loss": 0.188, "step": 6447 }, { "epoch": 2.0975927130774235, "grad_norm": 2.1592954195592915, "learning_rate": 2.5142195932823858e-06, "loss": 0.1854, "step": 6448 }, { "epoch": 2.097918022121015, "grad_norm": 2.2615960436207403, "learning_rate": 2.5125775025881185e-06, "loss": 0.1851, "step": 6449 }, { "epoch": 2.0982433311646065, "grad_norm": 2.1403314611019773, "learning_rate": 2.5109357683431522e-06, "loss": 0.1329, "step": 6450 }, { "epoch": 2.0985686402081978, "grad_norm": 1.7695119696220984, "learning_rate": 2.5092943907827466e-06, "loss": 0.1542, "step": 6451 }, { "epoch": 2.098893949251789, "grad_norm": 1.699436429095452, "learning_rate": 2.507653370142119e-06, "loss": 0.156, "step": 6452 }, { "epoch": 2.099219258295381, "grad_norm": 2.1492698696121177, "learning_rate": 2.5060127066564255e-06, "loss": 0.1639, "step": 6453 }, { "epoch": 2.099544567338972, "grad_norm": 1.945983761341031, "learning_rate": 2.5043724005607718e-06, "loss": 0.169, "step": 6454 }, { "epoch": 2.0998698763825634, "grad_norm": 2.356493334062848, "learning_rate": 2.5027324520902172e-06, "loss": 0.1983, "step": 6455 }, { "epoch": 2.1001951854261547, "grad_norm": 1.773094591477855, "learning_rate": 2.501092861479769e-06, "loss": 0.1294, "step": 6456 }, { "epoch": 2.1005204944697464, "grad_norm": 1.9434013685305322, "learning_rate": 2.499453628964378e-06, "loss": 0.13, "step": 6457 }, { "epoch": 2.1008458035133377, "grad_norm": 1.934313337650035, "learning_rate": 2.497814754778951e-06, "loss": 0.1402, "step": 6458 }, { "epoch": 2.101171112556929, "grad_norm": 2.4937086290461674, "learning_rate": 2.496176239158336e-06, "loss": 0.1805, "step": 6459 }, { "epoch": 2.1014964216005203, "grad_norm": 2.7451512814160295, "learning_rate": 2.4945380823373354e-06, "loss": 0.2094, "step": 6460 }, { "epoch": 2.101821730644112, "grad_norm": 1.922017519643114, "learning_rate": 2.4929002845507e-06, "loss": 0.1137, "step": 6461 }, { "epoch": 2.1021470396877033, "grad_norm": 2.25209629351929, "learning_rate": 2.4912628460331235e-06, "loss": 0.2117, "step": 6462 }, { "epoch": 2.1024723487312946, "grad_norm": 2.392169695497371, "learning_rate": 2.4896257670192537e-06, "loss": 0.1978, "step": 6463 }, { "epoch": 2.102797657774886, "grad_norm": 2.249863112431728, "learning_rate": 2.487989047743687e-06, "loss": 0.193, "step": 6464 }, { "epoch": 2.1031229668184777, "grad_norm": 2.232892705913359, "learning_rate": 2.4863526884409627e-06, "loss": 0.1433, "step": 6465 }, { "epoch": 2.103448275862069, "grad_norm": 1.838784314605168, "learning_rate": 2.4847166893455767e-06, "loss": 0.1637, "step": 6466 }, { "epoch": 2.1037735849056602, "grad_norm": 2.341135372146886, "learning_rate": 2.4830810506919633e-06, "loss": 0.1511, "step": 6467 }, { "epoch": 2.104098893949252, "grad_norm": 2.359754649829374, "learning_rate": 2.4814457727145135e-06, "loss": 0.1982, "step": 6468 }, { "epoch": 2.1044242029928433, "grad_norm": 2.1432958330686755, "learning_rate": 2.4798108556475665e-06, "loss": 0.1816, "step": 6469 }, { "epoch": 2.1047495120364346, "grad_norm": 1.8187168117608974, "learning_rate": 2.478176299725403e-06, "loss": 0.1383, "step": 6470 }, { "epoch": 2.105074821080026, "grad_norm": 2.15177028105842, "learning_rate": 2.476542105182254e-06, "loss": 0.1754, "step": 6471 }, { "epoch": 2.1054001301236176, "grad_norm": 2.0832876269319147, "learning_rate": 2.4749082722523076e-06, "loss": 0.1339, "step": 6472 }, { "epoch": 2.105725439167209, "grad_norm": 2.35681648855081, "learning_rate": 2.4732748011696887e-06, "loss": 0.1967, "step": 6473 }, { "epoch": 2.1060507482108, "grad_norm": 2.041662309803424, "learning_rate": 2.4716416921684742e-06, "loss": 0.1938, "step": 6474 }, { "epoch": 2.1063760572543915, "grad_norm": 2.23004105125236, "learning_rate": 2.470008945482691e-06, "loss": 0.1852, "step": 6475 }, { "epoch": 2.106701366297983, "grad_norm": 2.1604702535561433, "learning_rate": 2.4683765613463144e-06, "loss": 0.1959, "step": 6476 }, { "epoch": 2.1070266753415745, "grad_norm": 2.3968084489166066, "learning_rate": 2.4667445399932625e-06, "loss": 0.1569, "step": 6477 }, { "epoch": 2.107351984385166, "grad_norm": 2.146858195488358, "learning_rate": 2.465112881657409e-06, "loss": 0.1336, "step": 6478 }, { "epoch": 2.1076772934287575, "grad_norm": 2.154490336120725, "learning_rate": 2.4634815865725677e-06, "loss": 0.1698, "step": 6479 }, { "epoch": 2.108002602472349, "grad_norm": 2.2308096316142616, "learning_rate": 2.4618506549725065e-06, "loss": 0.1628, "step": 6480 }, { "epoch": 2.10832791151594, "grad_norm": 2.323022585590141, "learning_rate": 2.46022008709094e-06, "loss": 0.1686, "step": 6481 }, { "epoch": 2.1086532205595314, "grad_norm": 2.1914830345751124, "learning_rate": 2.4585898831615286e-06, "loss": 0.184, "step": 6482 }, { "epoch": 2.108978529603123, "grad_norm": 2.0467368584816246, "learning_rate": 2.4569600434178777e-06, "loss": 0.1384, "step": 6483 }, { "epoch": 2.1093038386467144, "grad_norm": 1.9468641243766704, "learning_rate": 2.4553305680935518e-06, "loss": 0.1328, "step": 6484 }, { "epoch": 2.1096291476903057, "grad_norm": 2.3596587858957423, "learning_rate": 2.4537014574220524e-06, "loss": 0.1671, "step": 6485 }, { "epoch": 2.109954456733897, "grad_norm": 2.158156243490179, "learning_rate": 2.45207271163683e-06, "loss": 0.1696, "step": 6486 }, { "epoch": 2.1102797657774888, "grad_norm": 1.6244058151705967, "learning_rate": 2.4504443309712863e-06, "loss": 0.1518, "step": 6487 }, { "epoch": 2.11060507482108, "grad_norm": 1.8603130205479244, "learning_rate": 2.448816315658772e-06, "loss": 0.1413, "step": 6488 }, { "epoch": 2.1109303838646714, "grad_norm": 2.1154986557928126, "learning_rate": 2.447188665932578e-06, "loss": 0.1247, "step": 6489 }, { "epoch": 2.1112556929082626, "grad_norm": 1.9235332072173636, "learning_rate": 2.4455613820259533e-06, "loss": 0.1478, "step": 6490 }, { "epoch": 2.1115810019518544, "grad_norm": 2.015595996270591, "learning_rate": 2.443934464172083e-06, "loss": 0.1494, "step": 6491 }, { "epoch": 2.1119063109954457, "grad_norm": 2.075878799785841, "learning_rate": 2.4423079126041084e-06, "loss": 0.1447, "step": 6492 }, { "epoch": 2.112231620039037, "grad_norm": 1.9100850310569264, "learning_rate": 2.4406817275551175e-06, "loss": 0.1331, "step": 6493 }, { "epoch": 2.1125569290826287, "grad_norm": 2.2545297826328725, "learning_rate": 2.439055909258139e-06, "loss": 0.2164, "step": 6494 }, { "epoch": 2.11288223812622, "grad_norm": 2.3499270670536507, "learning_rate": 2.437430457946156e-06, "loss": 0.1713, "step": 6495 }, { "epoch": 2.1132075471698113, "grad_norm": 2.0116643347445757, "learning_rate": 2.4358053738520994e-06, "loss": 0.1234, "step": 6496 }, { "epoch": 2.1135328562134026, "grad_norm": 2.601631458325508, "learning_rate": 2.43418065720884e-06, "loss": 0.2173, "step": 6497 }, { "epoch": 2.1138581652569943, "grad_norm": 1.8530378910834358, "learning_rate": 2.4325563082492053e-06, "loss": 0.1415, "step": 6498 }, { "epoch": 2.1141834743005856, "grad_norm": 2.392792222811623, "learning_rate": 2.4309323272059615e-06, "loss": 0.1989, "step": 6499 }, { "epoch": 2.114508783344177, "grad_norm": 2.530933324354723, "learning_rate": 2.4293087143118282e-06, "loss": 0.1328, "step": 6500 }, { "epoch": 2.114834092387768, "grad_norm": 2.117933659128786, "learning_rate": 2.427685469799472e-06, "loss": 0.1356, "step": 6501 }, { "epoch": 2.11515940143136, "grad_norm": 2.1901379768404854, "learning_rate": 2.4260625939015028e-06, "loss": 0.1701, "step": 6502 }, { "epoch": 2.1154847104749512, "grad_norm": 1.9617398813036642, "learning_rate": 2.4244400868504757e-06, "loss": 0.1508, "step": 6503 }, { "epoch": 2.1158100195185425, "grad_norm": 2.089520958524423, "learning_rate": 2.422817948878906e-06, "loss": 0.135, "step": 6504 }, { "epoch": 2.116135328562134, "grad_norm": 1.9171482211020006, "learning_rate": 2.4211961802192425e-06, "loss": 0.1806, "step": 6505 }, { "epoch": 2.1164606376057256, "grad_norm": 1.9012541605630486, "learning_rate": 2.4195747811038835e-06, "loss": 0.1429, "step": 6506 }, { "epoch": 2.116785946649317, "grad_norm": 2.5202458327308266, "learning_rate": 2.4179537517651787e-06, "loss": 0.1609, "step": 6507 }, { "epoch": 2.117111255692908, "grad_norm": 1.8725374701341202, "learning_rate": 2.4163330924354243e-06, "loss": 0.1465, "step": 6508 }, { "epoch": 2.1174365647365, "grad_norm": 2.161441337772438, "learning_rate": 2.414712803346859e-06, "loss": 0.2394, "step": 6509 }, { "epoch": 2.117761873780091, "grad_norm": 2.3866001503063363, "learning_rate": 2.413092884731675e-06, "loss": 0.2104, "step": 6510 }, { "epoch": 2.1180871828236825, "grad_norm": 2.271870117540331, "learning_rate": 2.4114733368220035e-06, "loss": 0.2032, "step": 6511 }, { "epoch": 2.1184124918672738, "grad_norm": 1.959042133234405, "learning_rate": 2.409854159849928e-06, "loss": 0.1297, "step": 6512 }, { "epoch": 2.1187378009108655, "grad_norm": 2.2099802279811493, "learning_rate": 2.4082353540474802e-06, "loss": 0.2062, "step": 6513 }, { "epoch": 2.119063109954457, "grad_norm": 1.8331312285399957, "learning_rate": 2.4066169196466326e-06, "loss": 0.1502, "step": 6514 }, { "epoch": 2.119388418998048, "grad_norm": 2.1402535810195134, "learning_rate": 2.40499885687931e-06, "loss": 0.1435, "step": 6515 }, { "epoch": 2.1197137280416394, "grad_norm": 2.0304989440851315, "learning_rate": 2.403381165977382e-06, "loss": 0.1882, "step": 6516 }, { "epoch": 2.120039037085231, "grad_norm": 1.997012106832591, "learning_rate": 2.4017638471726633e-06, "loss": 0.1596, "step": 6517 }, { "epoch": 2.1203643461288224, "grad_norm": 1.878114238414347, "learning_rate": 2.4001469006969192e-06, "loss": 0.1279, "step": 6518 }, { "epoch": 2.1206896551724137, "grad_norm": 2.1095787012761327, "learning_rate": 2.3985303267818556e-06, "loss": 0.1582, "step": 6519 }, { "epoch": 2.121014964216005, "grad_norm": 2.6196897845269125, "learning_rate": 2.396914125659131e-06, "loss": 0.197, "step": 6520 }, { "epoch": 2.1213402732595967, "grad_norm": 2.0564602414115836, "learning_rate": 2.3952982975603494e-06, "loss": 0.1433, "step": 6521 }, { "epoch": 2.121665582303188, "grad_norm": 2.122642608889823, "learning_rate": 2.393682842717058e-06, "loss": 0.198, "step": 6522 }, { "epoch": 2.1219908913467793, "grad_norm": 2.0724883016765214, "learning_rate": 2.392067761360752e-06, "loss": 0.1584, "step": 6523 }, { "epoch": 2.122316200390371, "grad_norm": 1.7333518200723783, "learning_rate": 2.3904530537228737e-06, "loss": 0.1418, "step": 6524 }, { "epoch": 2.1226415094339623, "grad_norm": 1.950814953018288, "learning_rate": 2.388838720034815e-06, "loss": 0.1657, "step": 6525 }, { "epoch": 2.1229668184775536, "grad_norm": 1.938059613906395, "learning_rate": 2.3872247605279063e-06, "loss": 0.1592, "step": 6526 }, { "epoch": 2.123292127521145, "grad_norm": 2.263337522041446, "learning_rate": 2.3856111754334323e-06, "loss": 0.1878, "step": 6527 }, { "epoch": 2.1236174365647367, "grad_norm": 2.282062491257879, "learning_rate": 2.383997964982621e-06, "loss": 0.1468, "step": 6528 }, { "epoch": 2.123942745608328, "grad_norm": 1.8635517660806473, "learning_rate": 2.382385129406644e-06, "loss": 0.1526, "step": 6529 }, { "epoch": 2.1242680546519193, "grad_norm": 2.120754749102418, "learning_rate": 2.380772668936625e-06, "loss": 0.2123, "step": 6530 }, { "epoch": 2.1245933636955106, "grad_norm": 2.2440154773974927, "learning_rate": 2.379160583803627e-06, "loss": 0.2025, "step": 6531 }, { "epoch": 2.1249186727391023, "grad_norm": 2.1750850461406666, "learning_rate": 2.377548874238665e-06, "loss": 0.1501, "step": 6532 }, { "epoch": 2.1252439817826936, "grad_norm": 1.8892956197801356, "learning_rate": 2.3759375404726992e-06, "loss": 0.1438, "step": 6533 }, { "epoch": 2.125569290826285, "grad_norm": 2.2238641491548146, "learning_rate": 2.3743265827366336e-06, "loss": 0.1842, "step": 6534 }, { "epoch": 2.125894599869876, "grad_norm": 2.380026995757895, "learning_rate": 2.3727160012613153e-06, "loss": 0.1936, "step": 6535 }, { "epoch": 2.126219908913468, "grad_norm": 1.9729535622587255, "learning_rate": 2.3711057962775497e-06, "loss": 0.2117, "step": 6536 }, { "epoch": 2.126545217957059, "grad_norm": 1.9109561033539124, "learning_rate": 2.369495968016077e-06, "loss": 0.1713, "step": 6537 }, { "epoch": 2.1268705270006505, "grad_norm": 2.477292901497453, "learning_rate": 2.3678865167075825e-06, "loss": 0.2007, "step": 6538 }, { "epoch": 2.1271958360442422, "grad_norm": 2.49505774495564, "learning_rate": 2.3662774425827064e-06, "loss": 0.179, "step": 6539 }, { "epoch": 2.1275211450878335, "grad_norm": 1.7513525990248404, "learning_rate": 2.3646687458720297e-06, "loss": 0.1462, "step": 6540 }, { "epoch": 2.127846454131425, "grad_norm": 2.4302734763710783, "learning_rate": 2.3630604268060773e-06, "loss": 0.1742, "step": 6541 }, { "epoch": 2.128171763175016, "grad_norm": 1.7279655236211726, "learning_rate": 2.3614524856153238e-06, "loss": 0.1599, "step": 6542 }, { "epoch": 2.128497072218608, "grad_norm": 2.3424974304021493, "learning_rate": 2.3598449225301905e-06, "loss": 0.2036, "step": 6543 }, { "epoch": 2.128822381262199, "grad_norm": 2.0107825355645503, "learning_rate": 2.3582377377810378e-06, "loss": 0.1535, "step": 6544 }, { "epoch": 2.1291476903057904, "grad_norm": 2.1035981926113134, "learning_rate": 2.3566309315981807e-06, "loss": 0.1557, "step": 6545 }, { "epoch": 2.1294729993493817, "grad_norm": 2.1655627362924355, "learning_rate": 2.355024504211871e-06, "loss": 0.1932, "step": 6546 }, { "epoch": 2.1297983083929735, "grad_norm": 2.0894921365830967, "learning_rate": 2.353418455852314e-06, "loss": 0.2202, "step": 6547 }, { "epoch": 2.1301236174365648, "grad_norm": 2.1133859898267966, "learning_rate": 2.3518127867496586e-06, "loss": 0.1751, "step": 6548 }, { "epoch": 2.130448926480156, "grad_norm": 2.3922364796949473, "learning_rate": 2.350207497133995e-06, "loss": 0.2188, "step": 6549 }, { "epoch": 2.130774235523748, "grad_norm": 2.04453548323874, "learning_rate": 2.3486025872353634e-06, "loss": 0.1728, "step": 6550 }, { "epoch": 2.131099544567339, "grad_norm": 1.7269732390561532, "learning_rate": 2.3469980572837503e-06, "loss": 0.1373, "step": 6551 }, { "epoch": 2.1314248536109304, "grad_norm": 2.3094729107619676, "learning_rate": 2.345393907509083e-06, "loss": 0.1978, "step": 6552 }, { "epoch": 2.1317501626545217, "grad_norm": 1.9203309624149774, "learning_rate": 2.3437901381412405e-06, "loss": 0.1357, "step": 6553 }, { "epoch": 2.1320754716981134, "grad_norm": 1.9172772413646233, "learning_rate": 2.3421867494100405e-06, "loss": 0.1838, "step": 6554 }, { "epoch": 2.1324007807417047, "grad_norm": 2.127444060839045, "learning_rate": 2.340583741545251e-06, "loss": 0.1989, "step": 6555 }, { "epoch": 2.132726089785296, "grad_norm": 1.8546212790911376, "learning_rate": 2.3389811147765863e-06, "loss": 0.1269, "step": 6556 }, { "epoch": 2.1330513988288873, "grad_norm": 1.9990773853387662, "learning_rate": 2.3373788693337024e-06, "loss": 0.1353, "step": 6557 }, { "epoch": 2.133376707872479, "grad_norm": 2.1414659353421666, "learning_rate": 2.3357770054461975e-06, "loss": 0.2133, "step": 6558 }, { "epoch": 2.1337020169160703, "grad_norm": 2.280634266663062, "learning_rate": 2.334175523343628e-06, "loss": 0.1579, "step": 6559 }, { "epoch": 2.1340273259596616, "grad_norm": 1.914943003965037, "learning_rate": 2.3325744232554838e-06, "loss": 0.1421, "step": 6560 }, { "epoch": 2.134352635003253, "grad_norm": 2.258199716353669, "learning_rate": 2.3309737054112006e-06, "loss": 0.1543, "step": 6561 }, { "epoch": 2.1346779440468446, "grad_norm": 2.512189702769908, "learning_rate": 2.3293733700401653e-06, "loss": 0.1733, "step": 6562 }, { "epoch": 2.135003253090436, "grad_norm": 2.100506697176171, "learning_rate": 2.3277734173717085e-06, "loss": 0.157, "step": 6563 }, { "epoch": 2.135328562134027, "grad_norm": 2.067599733508948, "learning_rate": 2.326173847635101e-06, "loss": 0.1347, "step": 6564 }, { "epoch": 2.1356538711776185, "grad_norm": 1.7979493553820556, "learning_rate": 2.3245746610595655e-06, "loss": 0.189, "step": 6565 }, { "epoch": 2.1359791802212102, "grad_norm": 2.338660575458038, "learning_rate": 2.322975857874263e-06, "loss": 0.1583, "step": 6566 }, { "epoch": 2.1363044892648015, "grad_norm": 2.3533158682685356, "learning_rate": 2.3213774383083048e-06, "loss": 0.1688, "step": 6567 }, { "epoch": 2.136629798308393, "grad_norm": 2.7693882379070462, "learning_rate": 2.3197794025907477e-06, "loss": 0.179, "step": 6568 }, { "epoch": 2.1369551073519846, "grad_norm": 2.1876698951470157, "learning_rate": 2.318181750950589e-06, "loss": 0.22, "step": 6569 }, { "epoch": 2.137280416395576, "grad_norm": 1.8678467901451004, "learning_rate": 2.31658448361677e-06, "loss": 0.1719, "step": 6570 }, { "epoch": 2.137605725439167, "grad_norm": 2.0321059825479875, "learning_rate": 2.314987600818187e-06, "loss": 0.1871, "step": 6571 }, { "epoch": 2.1379310344827585, "grad_norm": 1.8106000659991837, "learning_rate": 2.313391102783671e-06, "loss": 0.1335, "step": 6572 }, { "epoch": 2.13825634352635, "grad_norm": 2.035271737103898, "learning_rate": 2.3117949897419996e-06, "loss": 0.193, "step": 6573 }, { "epoch": 2.1385816525699415, "grad_norm": 2.1262541585768124, "learning_rate": 2.310199261921898e-06, "loss": 0.1693, "step": 6574 }, { "epoch": 2.1389069616135328, "grad_norm": 1.89592148344303, "learning_rate": 2.308603919552038e-06, "loss": 0.1326, "step": 6575 }, { "epoch": 2.139232270657124, "grad_norm": 2.344315711421125, "learning_rate": 2.3070089628610287e-06, "loss": 0.1931, "step": 6576 }, { "epoch": 2.139557579700716, "grad_norm": 2.4604445608090524, "learning_rate": 2.3054143920774325e-06, "loss": 0.208, "step": 6577 }, { "epoch": 2.139882888744307, "grad_norm": 2.124949627120654, "learning_rate": 2.3038202074297485e-06, "loss": 0.1745, "step": 6578 }, { "epoch": 2.1402081977878984, "grad_norm": 2.1146326250349854, "learning_rate": 2.302226409146426e-06, "loss": 0.1693, "step": 6579 }, { "epoch": 2.14053350683149, "grad_norm": 2.271758448006512, "learning_rate": 2.3006329974558598e-06, "loss": 0.1889, "step": 6580 }, { "epoch": 2.1408588158750814, "grad_norm": 2.224197056684788, "learning_rate": 2.299039972586383e-06, "loss": 0.1711, "step": 6581 }, { "epoch": 2.1411841249186727, "grad_norm": 1.9371524489380236, "learning_rate": 2.297447334766279e-06, "loss": 0.1216, "step": 6582 }, { "epoch": 2.141509433962264, "grad_norm": 2.1192121838297124, "learning_rate": 2.295855084223776e-06, "loss": 0.1875, "step": 6583 }, { "epoch": 2.1418347430058557, "grad_norm": 2.375308502857867, "learning_rate": 2.2942632211870402e-06, "loss": 0.1815, "step": 6584 }, { "epoch": 2.142160052049447, "grad_norm": 1.9564125772271612, "learning_rate": 2.2926717458841906e-06, "loss": 0.1508, "step": 6585 }, { "epoch": 2.1424853610930383, "grad_norm": 2.1418453925778325, "learning_rate": 2.291080658543283e-06, "loss": 0.1547, "step": 6586 }, { "epoch": 2.1428106701366296, "grad_norm": 2.2080116609016747, "learning_rate": 2.289489959392323e-06, "loss": 0.1685, "step": 6587 }, { "epoch": 2.1431359791802214, "grad_norm": 2.624808759951995, "learning_rate": 2.2878996486592616e-06, "loss": 0.2026, "step": 6588 }, { "epoch": 2.1434612882238127, "grad_norm": 2.6802454058265117, "learning_rate": 2.286309726571988e-06, "loss": 0.1907, "step": 6589 }, { "epoch": 2.143786597267404, "grad_norm": 2.014025628841823, "learning_rate": 2.2847201933583362e-06, "loss": 0.1804, "step": 6590 }, { "epoch": 2.1441119063109952, "grad_norm": 2.3089662092487537, "learning_rate": 2.283131049246095e-06, "loss": 0.211, "step": 6591 }, { "epoch": 2.144437215354587, "grad_norm": 2.144127535267405, "learning_rate": 2.281542294462986e-06, "loss": 0.1766, "step": 6592 }, { "epoch": 2.1447625243981783, "grad_norm": 2.159369683153739, "learning_rate": 2.279953929236677e-06, "loss": 0.165, "step": 6593 }, { "epoch": 2.1450878334417696, "grad_norm": 1.9380021538477983, "learning_rate": 2.2783659537947826e-06, "loss": 0.1568, "step": 6594 }, { "epoch": 2.145413142485361, "grad_norm": 1.6090151966950748, "learning_rate": 2.2767783683648644e-06, "loss": 0.1441, "step": 6595 }, { "epoch": 2.1457384515289526, "grad_norm": 2.2260587146401325, "learning_rate": 2.2751911731744192e-06, "loss": 0.2046, "step": 6596 }, { "epoch": 2.146063760572544, "grad_norm": 2.0524450154102243, "learning_rate": 2.273604368450898e-06, "loss": 0.1456, "step": 6597 }, { "epoch": 2.146389069616135, "grad_norm": 2.085351840527987, "learning_rate": 2.272017954421686e-06, "loss": 0.1333, "step": 6598 }, { "epoch": 2.146714378659727, "grad_norm": 1.9862525026187465, "learning_rate": 2.2704319313141203e-06, "loss": 0.1849, "step": 6599 }, { "epoch": 2.147039687703318, "grad_norm": 1.8986128037296641, "learning_rate": 2.268846299355481e-06, "loss": 0.1581, "step": 6600 }, { "epoch": 2.1473649967469095, "grad_norm": 2.5446233660317747, "learning_rate": 2.2672610587729856e-06, "loss": 0.1718, "step": 6601 }, { "epoch": 2.147690305790501, "grad_norm": 1.7502726732456204, "learning_rate": 2.265676209793802e-06, "loss": 0.1158, "step": 6602 }, { "epoch": 2.1480156148340925, "grad_norm": 2.2598370214991133, "learning_rate": 2.2640917526450433e-06, "loss": 0.1645, "step": 6603 }, { "epoch": 2.148340923877684, "grad_norm": 2.0384977501657904, "learning_rate": 2.2625076875537588e-06, "loss": 0.1756, "step": 6604 }, { "epoch": 2.148666232921275, "grad_norm": 2.655704815862129, "learning_rate": 2.260924014746949e-06, "loss": 0.187, "step": 6605 }, { "epoch": 2.1489915419648664, "grad_norm": 2.146287234455426, "learning_rate": 2.2593407344515527e-06, "loss": 0.1416, "step": 6606 }, { "epoch": 2.149316851008458, "grad_norm": 2.32761642415574, "learning_rate": 2.2577578468944566e-06, "loss": 0.2121, "step": 6607 }, { "epoch": 2.1496421600520494, "grad_norm": 2.3299000231151843, "learning_rate": 2.2561753523024908e-06, "loss": 0.188, "step": 6608 }, { "epoch": 2.1499674690956407, "grad_norm": 2.218751940336341, "learning_rate": 2.2545932509024266e-06, "loss": 0.1403, "step": 6609 }, { "epoch": 2.1502927781392325, "grad_norm": 2.4682887720322904, "learning_rate": 2.253011542920978e-06, "loss": 0.1765, "step": 6610 }, { "epoch": 2.1506180871828238, "grad_norm": 2.461087545078668, "learning_rate": 2.2514302285848065e-06, "loss": 0.1948, "step": 6611 }, { "epoch": 2.150943396226415, "grad_norm": 2.101032246981965, "learning_rate": 2.2498493081205186e-06, "loss": 0.2097, "step": 6612 }, { "epoch": 2.1512687052700064, "grad_norm": 2.255757338316502, "learning_rate": 2.2482687817546554e-06, "loss": 0.1793, "step": 6613 }, { "epoch": 2.151594014313598, "grad_norm": 2.2682190802486266, "learning_rate": 2.2466886497137115e-06, "loss": 0.1711, "step": 6614 }, { "epoch": 2.1519193233571894, "grad_norm": 2.0971567147052235, "learning_rate": 2.245108912224121e-06, "loss": 0.1711, "step": 6615 }, { "epoch": 2.1522446324007807, "grad_norm": 2.1102131724051123, "learning_rate": 2.243529569512259e-06, "loss": 0.1729, "step": 6616 }, { "epoch": 2.152569941444372, "grad_norm": 2.141489028105564, "learning_rate": 2.2419506218044486e-06, "loss": 0.1794, "step": 6617 }, { "epoch": 2.1528952504879637, "grad_norm": 2.221801126301808, "learning_rate": 2.2403720693269513e-06, "loss": 0.1999, "step": 6618 }, { "epoch": 2.153220559531555, "grad_norm": 2.28797010164301, "learning_rate": 2.2387939123059767e-06, "loss": 0.1979, "step": 6619 }, { "epoch": 2.1535458685751463, "grad_norm": 2.0742951847003845, "learning_rate": 2.237216150967677e-06, "loss": 0.1518, "step": 6620 }, { "epoch": 2.153871177618738, "grad_norm": 1.8740490785311377, "learning_rate": 2.2356387855381452e-06, "loss": 0.1445, "step": 6621 }, { "epoch": 2.1541964866623293, "grad_norm": 2.2294475664437186, "learning_rate": 2.2340618162434143e-06, "loss": 0.1579, "step": 6622 }, { "epoch": 2.1545217957059206, "grad_norm": 2.3434400937341078, "learning_rate": 2.2324852433094734e-06, "loss": 0.1499, "step": 6623 }, { "epoch": 2.154847104749512, "grad_norm": 2.0469112456155343, "learning_rate": 2.230909066962242e-06, "loss": 0.1246, "step": 6624 }, { "epoch": 2.1551724137931036, "grad_norm": 1.921617776311707, "learning_rate": 2.2293332874275858e-06, "loss": 0.139, "step": 6625 }, { "epoch": 2.155497722836695, "grad_norm": 2.067375178881831, "learning_rate": 2.227757904931317e-06, "loss": 0.1508, "step": 6626 }, { "epoch": 2.1558230318802862, "grad_norm": 2.4085093164111835, "learning_rate": 2.22618291969919e-06, "loss": 0.1664, "step": 6627 }, { "epoch": 2.1561483409238775, "grad_norm": 1.9509694715538024, "learning_rate": 2.224608331956899e-06, "loss": 0.1548, "step": 6628 }, { "epoch": 2.1564736499674693, "grad_norm": 2.0928916210871695, "learning_rate": 2.223034141930086e-06, "loss": 0.1548, "step": 6629 }, { "epoch": 2.1567989590110606, "grad_norm": 1.8479643106812598, "learning_rate": 2.22146034984433e-06, "loss": 0.12, "step": 6630 }, { "epoch": 2.157124268054652, "grad_norm": 2.173841812090867, "learning_rate": 2.2198869559251585e-06, "loss": 0.1496, "step": 6631 }, { "epoch": 2.157449577098243, "grad_norm": 2.2856023971076613, "learning_rate": 2.2183139603980414e-06, "loss": 0.1957, "step": 6632 }, { "epoch": 2.157774886141835, "grad_norm": 2.026936449164263, "learning_rate": 2.2167413634883866e-06, "loss": 0.1619, "step": 6633 }, { "epoch": 2.158100195185426, "grad_norm": 2.0246785818907793, "learning_rate": 2.2151691654215508e-06, "loss": 0.1277, "step": 6634 }, { "epoch": 2.1584255042290175, "grad_norm": 2.2751774295362766, "learning_rate": 2.213597366422831e-06, "loss": 0.1499, "step": 6635 }, { "epoch": 2.1587508132726088, "grad_norm": 2.3411449812042333, "learning_rate": 2.2120259667174654e-06, "loss": 0.1785, "step": 6636 }, { "epoch": 2.1590761223162005, "grad_norm": 2.092177010948941, "learning_rate": 2.210454966530639e-06, "loss": 0.1491, "step": 6637 }, { "epoch": 2.159401431359792, "grad_norm": 1.8624686535026265, "learning_rate": 2.2088843660874743e-06, "loss": 0.1553, "step": 6638 }, { "epoch": 2.159726740403383, "grad_norm": 2.1381128338250712, "learning_rate": 2.20731416561304e-06, "loss": 0.1615, "step": 6639 }, { "epoch": 2.160052049446975, "grad_norm": 2.1507856504235847, "learning_rate": 2.2057443653323505e-06, "loss": 0.1811, "step": 6640 }, { "epoch": 2.160377358490566, "grad_norm": 2.3073915419958633, "learning_rate": 2.2041749654703563e-06, "loss": 0.1819, "step": 6641 }, { "epoch": 2.1607026675341574, "grad_norm": 2.1716750270406338, "learning_rate": 2.20260596625195e-06, "loss": 0.1419, "step": 6642 }, { "epoch": 2.1610279765777487, "grad_norm": 2.1704178145299307, "learning_rate": 2.2010373679019773e-06, "loss": 0.1585, "step": 6643 }, { "epoch": 2.1613532856213404, "grad_norm": 2.3404235600737198, "learning_rate": 2.199469170645216e-06, "loss": 0.1463, "step": 6644 }, { "epoch": 2.1616785946649317, "grad_norm": 2.30796403445445, "learning_rate": 2.1979013747063876e-06, "loss": 0.2156, "step": 6645 }, { "epoch": 2.162003903708523, "grad_norm": 1.8878374964217735, "learning_rate": 2.1963339803101607e-06, "loss": 0.1202, "step": 6646 }, { "epoch": 2.1623292127521143, "grad_norm": 2.151533621362546, "learning_rate": 2.1947669876811457e-06, "loss": 0.1715, "step": 6647 }, { "epoch": 2.162654521795706, "grad_norm": 2.1116496043121558, "learning_rate": 2.19320039704389e-06, "loss": 0.1642, "step": 6648 }, { "epoch": 2.1629798308392973, "grad_norm": 2.55308602006256, "learning_rate": 2.19163420862289e-06, "loss": 0.1636, "step": 6649 }, { "epoch": 2.1633051398828886, "grad_norm": 2.039625463000093, "learning_rate": 2.190068422642579e-06, "loss": 0.1713, "step": 6650 }, { "epoch": 2.1636304489264804, "grad_norm": 2.5289913851770813, "learning_rate": 2.1885030393273367e-06, "loss": 0.2103, "step": 6651 }, { "epoch": 2.1639557579700717, "grad_norm": 2.2454831532387542, "learning_rate": 2.1869380589014844e-06, "loss": 0.1728, "step": 6652 }, { "epoch": 2.164281067013663, "grad_norm": 2.093074038738334, "learning_rate": 2.185373481589284e-06, "loss": 0.1346, "step": 6653 }, { "epoch": 2.1646063760572543, "grad_norm": 2.5793090593567825, "learning_rate": 2.1838093076149368e-06, "loss": 0.2039, "step": 6654 }, { "epoch": 2.164931685100846, "grad_norm": 1.8439817697417862, "learning_rate": 2.1822455372025975e-06, "loss": 0.167, "step": 6655 }, { "epoch": 2.1652569941444373, "grad_norm": 2.3966289843898942, "learning_rate": 2.180682170576351e-06, "loss": 0.1829, "step": 6656 }, { "epoch": 2.1655823031880286, "grad_norm": 2.193505025497714, "learning_rate": 2.1791192079602274e-06, "loss": 0.1926, "step": 6657 }, { "epoch": 2.16590761223162, "grad_norm": 2.4631726910531997, "learning_rate": 2.177556649578202e-06, "loss": 0.2057, "step": 6658 }, { "epoch": 2.1662329212752116, "grad_norm": 2.006998003428443, "learning_rate": 2.175994495654193e-06, "loss": 0.1775, "step": 6659 }, { "epoch": 2.166558230318803, "grad_norm": 2.388304220947068, "learning_rate": 2.174432746412053e-06, "loss": 0.1874, "step": 6660 }, { "epoch": 2.166883539362394, "grad_norm": 2.3763326192650758, "learning_rate": 2.1728714020755865e-06, "loss": 0.2315, "step": 6661 }, { "epoch": 2.1672088484059855, "grad_norm": 2.111228184117871, "learning_rate": 2.1713104628685306e-06, "loss": 0.1522, "step": 6662 }, { "epoch": 2.1675341574495772, "grad_norm": 2.200149177205887, "learning_rate": 2.169749929014572e-06, "loss": 0.1635, "step": 6663 }, { "epoch": 2.1678594664931685, "grad_norm": 1.8568890076326392, "learning_rate": 2.1681898007373368e-06, "loss": 0.1517, "step": 6664 }, { "epoch": 2.16818477553676, "grad_norm": 2.2199206461036454, "learning_rate": 2.16663007826039e-06, "loss": 0.1447, "step": 6665 }, { "epoch": 2.168510084580351, "grad_norm": 2.760560700490789, "learning_rate": 2.165070761807242e-06, "loss": 0.2261, "step": 6666 }, { "epoch": 2.168835393623943, "grad_norm": 2.1069593398571063, "learning_rate": 2.163511851601346e-06, "loss": 0.1512, "step": 6667 }, { "epoch": 2.169160702667534, "grad_norm": 1.9669414452470484, "learning_rate": 2.1619533478660904e-06, "loss": 0.1617, "step": 6668 }, { "epoch": 2.1694860117111254, "grad_norm": 2.306144829419474, "learning_rate": 2.1603952508248144e-06, "loss": 0.1573, "step": 6669 }, { "epoch": 2.169811320754717, "grad_norm": 2.572188368801531, "learning_rate": 2.1588375607007904e-06, "loss": 0.1695, "step": 6670 }, { "epoch": 2.1701366297983085, "grad_norm": 2.073627097214207, "learning_rate": 2.157280277717238e-06, "loss": 0.1667, "step": 6671 }, { "epoch": 2.1704619388418998, "grad_norm": 2.233952850293862, "learning_rate": 2.1557234020973194e-06, "loss": 0.1662, "step": 6672 }, { "epoch": 2.170787247885491, "grad_norm": 2.0387962626773493, "learning_rate": 2.154166934064131e-06, "loss": 0.1603, "step": 6673 }, { "epoch": 2.171112556929083, "grad_norm": 2.130897967219633, "learning_rate": 2.152610873840719e-06, "loss": 0.1684, "step": 6674 }, { "epoch": 2.171437865972674, "grad_norm": 2.488106613464872, "learning_rate": 2.151055221650069e-06, "loss": 0.1595, "step": 6675 }, { "epoch": 2.1717631750162654, "grad_norm": 2.1372739631243536, "learning_rate": 2.1494999777151056e-06, "loss": 0.1811, "step": 6676 }, { "epoch": 2.1720884840598567, "grad_norm": 2.18899722568951, "learning_rate": 2.1479451422586918e-06, "loss": 0.1638, "step": 6677 }, { "epoch": 2.1724137931034484, "grad_norm": 2.075238099960896, "learning_rate": 2.146390715503644e-06, "loss": 0.1929, "step": 6678 }, { "epoch": 2.1727391021470397, "grad_norm": 1.722357027144325, "learning_rate": 2.1448366976727096e-06, "loss": 0.1273, "step": 6679 }, { "epoch": 2.173064411190631, "grad_norm": 2.1134810288516004, "learning_rate": 2.143283088988578e-06, "loss": 0.1673, "step": 6680 }, { "epoch": 2.1733897202342227, "grad_norm": 2.586097798191449, "learning_rate": 2.1417298896738847e-06, "loss": 0.1724, "step": 6681 }, { "epoch": 2.173715029277814, "grad_norm": 2.015481084142851, "learning_rate": 2.140177099951206e-06, "loss": 0.1515, "step": 6682 }, { "epoch": 2.1740403383214053, "grad_norm": 2.0193276133994913, "learning_rate": 2.138624720043053e-06, "loss": 0.1595, "step": 6683 }, { "epoch": 2.1743656473649966, "grad_norm": 2.2994369445936065, "learning_rate": 2.137072750171888e-06, "loss": 0.1811, "step": 6684 }, { "epoch": 2.1746909564085883, "grad_norm": 2.4478154260346434, "learning_rate": 2.135521190560104e-06, "loss": 0.177, "step": 6685 }, { "epoch": 2.1750162654521796, "grad_norm": 1.922711375914413, "learning_rate": 2.133970041430044e-06, "loss": 0.1515, "step": 6686 }, { "epoch": 2.175341574495771, "grad_norm": 1.7236694794222196, "learning_rate": 2.13241930300399e-06, "loss": 0.1144, "step": 6687 }, { "epoch": 2.175666883539362, "grad_norm": 2.1424139121275707, "learning_rate": 2.130868975504159e-06, "loss": 0.1812, "step": 6688 }, { "epoch": 2.175992192582954, "grad_norm": 2.13244214964986, "learning_rate": 2.1293190591527174e-06, "loss": 0.1812, "step": 6689 }, { "epoch": 2.1763175016265452, "grad_norm": 2.0114973032139294, "learning_rate": 2.1277695541717703e-06, "loss": 0.1841, "step": 6690 }, { "epoch": 2.1766428106701365, "grad_norm": 2.062687946412776, "learning_rate": 2.1262204607833587e-06, "loss": 0.1565, "step": 6691 }, { "epoch": 2.176968119713728, "grad_norm": 1.8876310495887978, "learning_rate": 2.1246717792094733e-06, "loss": 0.1466, "step": 6692 }, { "epoch": 2.1772934287573196, "grad_norm": 2.627642554351057, "learning_rate": 2.1231235096720377e-06, "loss": 0.2006, "step": 6693 }, { "epoch": 2.177618737800911, "grad_norm": 1.9254336851863951, "learning_rate": 2.121575652392923e-06, "loss": 0.1376, "step": 6694 }, { "epoch": 2.177944046844502, "grad_norm": 2.229819140712092, "learning_rate": 2.120028207593934e-06, "loss": 0.2361, "step": 6695 }, { "epoch": 2.178269355888094, "grad_norm": 2.101084015741735, "learning_rate": 2.1184811754968248e-06, "loss": 0.2022, "step": 6696 }, { "epoch": 2.178594664931685, "grad_norm": 1.9520407750295163, "learning_rate": 2.116934556323283e-06, "loss": 0.1622, "step": 6697 }, { "epoch": 2.1789199739752765, "grad_norm": 2.5621562238577518, "learning_rate": 2.115388350294941e-06, "loss": 0.2223, "step": 6698 }, { "epoch": 2.1792452830188678, "grad_norm": 1.9065402162720315, "learning_rate": 2.1138425576333736e-06, "loss": 0.161, "step": 6699 }, { "epoch": 2.1795705920624595, "grad_norm": 2.2491170296880005, "learning_rate": 2.112297178560091e-06, "loss": 0.1917, "step": 6700 }, { "epoch": 2.179895901106051, "grad_norm": 2.456910719844886, "learning_rate": 2.110752213296547e-06, "loss": 0.2025, "step": 6701 }, { "epoch": 2.180221210149642, "grad_norm": 2.3643960001391564, "learning_rate": 2.1092076620641396e-06, "loss": 0.194, "step": 6702 }, { "epoch": 2.1805465191932334, "grad_norm": 1.8813375503439531, "learning_rate": 2.107663525084199e-06, "loss": 0.1273, "step": 6703 }, { "epoch": 2.180871828236825, "grad_norm": 2.069327218321541, "learning_rate": 2.1061198025780065e-06, "loss": 0.1279, "step": 6704 }, { "epoch": 2.1811971372804164, "grad_norm": 2.4770338803456333, "learning_rate": 2.1045764947667734e-06, "loss": 0.2201, "step": 6705 }, { "epoch": 2.1815224463240077, "grad_norm": 2.00279603851345, "learning_rate": 2.103033601871659e-06, "loss": 0.1634, "step": 6706 }, { "epoch": 2.181847755367599, "grad_norm": 2.3106410017564887, "learning_rate": 2.101491124113763e-06, "loss": 0.2056, "step": 6707 }, { "epoch": 2.1821730644111907, "grad_norm": 2.017791771482004, "learning_rate": 2.0999490617141216e-06, "loss": 0.1595, "step": 6708 }, { "epoch": 2.182498373454782, "grad_norm": 2.021766432626424, "learning_rate": 2.0984074148937096e-06, "loss": 0.1965, "step": 6709 }, { "epoch": 2.1828236824983733, "grad_norm": 1.8748040744794696, "learning_rate": 2.0968661838734528e-06, "loss": 0.1305, "step": 6710 }, { "epoch": 2.183148991541965, "grad_norm": 1.945117091499822, "learning_rate": 2.0953253688742077e-06, "loss": 0.2027, "step": 6711 }, { "epoch": 2.1834743005855564, "grad_norm": 2.31172012047728, "learning_rate": 2.0937849701167716e-06, "loss": 0.2213, "step": 6712 }, { "epoch": 2.1837996096291477, "grad_norm": 1.9708684863100858, "learning_rate": 2.0922449878218874e-06, "loss": 0.1724, "step": 6713 }, { "epoch": 2.184124918672739, "grad_norm": 2.2434021848108507, "learning_rate": 2.0907054222102367e-06, "loss": 0.1683, "step": 6714 }, { "epoch": 2.1844502277163307, "grad_norm": 2.034486690974538, "learning_rate": 2.089166273502437e-06, "loss": 0.2004, "step": 6715 }, { "epoch": 2.184775536759922, "grad_norm": 2.076669580032004, "learning_rate": 2.087627541919053e-06, "loss": 0.2241, "step": 6716 }, { "epoch": 2.1851008458035133, "grad_norm": 1.9453997033346513, "learning_rate": 2.0860892276805823e-06, "loss": 0.1207, "step": 6717 }, { "epoch": 2.1854261548471046, "grad_norm": 1.785777895207517, "learning_rate": 2.084551331007468e-06, "loss": 0.1371, "step": 6718 }, { "epoch": 2.1857514638906963, "grad_norm": 1.948666475160894, "learning_rate": 2.0830138521200932e-06, "loss": 0.1466, "step": 6719 }, { "epoch": 2.1860767729342876, "grad_norm": 2.184268534901204, "learning_rate": 2.081476791238777e-06, "loss": 0.1792, "step": 6720 }, { "epoch": 2.186402081977879, "grad_norm": 1.9933211207237755, "learning_rate": 2.0799401485837816e-06, "loss": 0.1563, "step": 6721 }, { "epoch": 2.1867273910214706, "grad_norm": 1.6403982886736008, "learning_rate": 2.0784039243753124e-06, "loss": 0.1054, "step": 6722 }, { "epoch": 2.187052700065062, "grad_norm": 1.8011163906001662, "learning_rate": 2.0768681188335064e-06, "loss": 0.138, "step": 6723 }, { "epoch": 2.187378009108653, "grad_norm": 2.515791489314276, "learning_rate": 2.0753327321784504e-06, "loss": 0.1465, "step": 6724 }, { "epoch": 2.1877033181522445, "grad_norm": 2.5076042173573514, "learning_rate": 2.0737977646301615e-06, "loss": 0.1967, "step": 6725 }, { "epoch": 2.1880286271958362, "grad_norm": 2.1131536363284606, "learning_rate": 2.0722632164086045e-06, "loss": 0.1194, "step": 6726 }, { "epoch": 2.1883539362394275, "grad_norm": 2.0316610666959782, "learning_rate": 2.070729087733682e-06, "loss": 0.1597, "step": 6727 }, { "epoch": 2.188679245283019, "grad_norm": 2.106895619697814, "learning_rate": 2.069195378825234e-06, "loss": 0.1779, "step": 6728 }, { "epoch": 2.18900455432661, "grad_norm": 2.2119698009847863, "learning_rate": 2.0676620899030393e-06, "loss": 0.174, "step": 6729 }, { "epoch": 2.189329863370202, "grad_norm": 2.142975766327191, "learning_rate": 2.066129221186825e-06, "loss": 0.1938, "step": 6730 }, { "epoch": 2.189655172413793, "grad_norm": 1.8327936280110246, "learning_rate": 2.0645967728962496e-06, "loss": 0.1105, "step": 6731 }, { "epoch": 2.1899804814573844, "grad_norm": 2.1528357357129044, "learning_rate": 2.0630647452509124e-06, "loss": 0.2472, "step": 6732 }, { "epoch": 2.1903057905009757, "grad_norm": 2.569658878523697, "learning_rate": 2.061533138470355e-06, "loss": 0.22, "step": 6733 }, { "epoch": 2.1906310995445675, "grad_norm": 1.8994826306701198, "learning_rate": 2.06000195277406e-06, "loss": 0.1611, "step": 6734 }, { "epoch": 2.1909564085881588, "grad_norm": 2.1200552416553164, "learning_rate": 2.0584711883814435e-06, "loss": 0.1841, "step": 6735 }, { "epoch": 2.19128171763175, "grad_norm": 2.1522494704890436, "learning_rate": 2.0569408455118688e-06, "loss": 0.1622, "step": 6736 }, { "epoch": 2.1916070266753414, "grad_norm": 2.186372452378554, "learning_rate": 2.055410924384631e-06, "loss": 0.1899, "step": 6737 }, { "epoch": 2.191932335718933, "grad_norm": 1.9179583005098038, "learning_rate": 2.0538814252189714e-06, "loss": 0.1315, "step": 6738 }, { "epoch": 2.1922576447625244, "grad_norm": 2.2950865091780823, "learning_rate": 2.0523523482340695e-06, "loss": 0.1499, "step": 6739 }, { "epoch": 2.1925829538061157, "grad_norm": 1.9843295851567966, "learning_rate": 2.0508236936490417e-06, "loss": 0.1772, "step": 6740 }, { "epoch": 2.1929082628497074, "grad_norm": 1.7674890320829426, "learning_rate": 2.0492954616829406e-06, "loss": 0.1287, "step": 6741 }, { "epoch": 2.1932335718932987, "grad_norm": 2.0596324584863344, "learning_rate": 2.047767652554771e-06, "loss": 0.1797, "step": 6742 }, { "epoch": 2.19355888093689, "grad_norm": 2.2047507850509076, "learning_rate": 2.0462402664834654e-06, "loss": 0.2668, "step": 6743 }, { "epoch": 2.1938841899804813, "grad_norm": 2.543219882380594, "learning_rate": 2.044713303687897e-06, "loss": 0.1784, "step": 6744 }, { "epoch": 2.194209499024073, "grad_norm": 2.1642155061219954, "learning_rate": 2.0431867643868823e-06, "loss": 0.208, "step": 6745 }, { "epoch": 2.1945348080676643, "grad_norm": 1.9224719000612176, "learning_rate": 2.041660648799178e-06, "loss": 0.1641, "step": 6746 }, { "epoch": 2.1948601171112556, "grad_norm": 2.4177059832648977, "learning_rate": 2.0401349571434732e-06, "loss": 0.1961, "step": 6747 }, { "epoch": 2.195185426154847, "grad_norm": 2.241708844301298, "learning_rate": 2.038609689638404e-06, "loss": 0.1925, "step": 6748 }, { "epoch": 2.1955107351984386, "grad_norm": 2.3296630295842173, "learning_rate": 2.037084846502539e-06, "loss": 0.1928, "step": 6749 }, { "epoch": 2.19583604424203, "grad_norm": 2.1863819985241393, "learning_rate": 2.035560427954391e-06, "loss": 0.1749, "step": 6750 }, { "epoch": 2.1961613532856212, "grad_norm": 2.1480192918350185, "learning_rate": 2.034036434212412e-06, "loss": 0.1702, "step": 6751 }, { "epoch": 2.196486662329213, "grad_norm": 2.0129704776828823, "learning_rate": 2.032512865494988e-06, "loss": 0.1621, "step": 6752 }, { "epoch": 2.1968119713728043, "grad_norm": 2.166227169404926, "learning_rate": 2.030989722020448e-06, "loss": 0.1502, "step": 6753 }, { "epoch": 2.1971372804163956, "grad_norm": 2.3047132386521905, "learning_rate": 2.029467004007063e-06, "loss": 0.1992, "step": 6754 }, { "epoch": 2.197462589459987, "grad_norm": 2.055476633432784, "learning_rate": 2.0279447116730348e-06, "loss": 0.1493, "step": 6755 }, { "epoch": 2.1977878985035786, "grad_norm": 2.340418726379503, "learning_rate": 2.0264228452365127e-06, "loss": 0.1339, "step": 6756 }, { "epoch": 2.19811320754717, "grad_norm": 2.212403191916322, "learning_rate": 2.024901404915578e-06, "loss": 0.1515, "step": 6757 }, { "epoch": 2.198438516590761, "grad_norm": 2.101108402024072, "learning_rate": 2.023380390928255e-06, "loss": 0.1604, "step": 6758 }, { "epoch": 2.1987638256343525, "grad_norm": 2.078658023447989, "learning_rate": 2.0218598034925086e-06, "loss": 0.154, "step": 6759 }, { "epoch": 2.199089134677944, "grad_norm": 2.1775121494962884, "learning_rate": 2.020339642826238e-06, "loss": 0.1427, "step": 6760 }, { "epoch": 2.1994144437215355, "grad_norm": 2.222019816526071, "learning_rate": 2.0188199091472797e-06, "loss": 0.2054, "step": 6761 }, { "epoch": 2.199739752765127, "grad_norm": 2.013116754801156, "learning_rate": 2.017300602673419e-06, "loss": 0.1672, "step": 6762 }, { "epoch": 2.200065061808718, "grad_norm": 2.0838301737062936, "learning_rate": 2.0157817236223716e-06, "loss": 0.1254, "step": 6763 }, { "epoch": 2.20039037085231, "grad_norm": 2.61976659547388, "learning_rate": 2.0142632722117905e-06, "loss": 0.2582, "step": 6764 }, { "epoch": 2.200715679895901, "grad_norm": 2.3691020565200938, "learning_rate": 2.0127452486592737e-06, "loss": 0.1937, "step": 6765 }, { "epoch": 2.2010409889394924, "grad_norm": 2.419237210507148, "learning_rate": 2.0112276531823553e-06, "loss": 0.1909, "step": 6766 }, { "epoch": 2.2013662979830837, "grad_norm": 2.2183414471338088, "learning_rate": 2.009710485998506e-06, "loss": 0.2086, "step": 6767 }, { "epoch": 2.2016916070266754, "grad_norm": 2.0839669234611544, "learning_rate": 2.0081937473251395e-06, "loss": 0.1362, "step": 6768 }, { "epoch": 2.2020169160702667, "grad_norm": 2.0482063689404493, "learning_rate": 2.0066774373796016e-06, "loss": 0.1385, "step": 6769 }, { "epoch": 2.202342225113858, "grad_norm": 2.3757825827909684, "learning_rate": 2.0051615563791825e-06, "loss": 0.1713, "step": 6770 }, { "epoch": 2.2026675341574498, "grad_norm": 2.323504146592231, "learning_rate": 2.0036461045411115e-06, "loss": 0.1998, "step": 6771 }, { "epoch": 2.202992843201041, "grad_norm": 2.067058167779317, "learning_rate": 2.002131082082549e-06, "loss": 0.1968, "step": 6772 }, { "epoch": 2.2033181522446323, "grad_norm": 2.020489109989551, "learning_rate": 2.0006164892206016e-06, "loss": 0.1526, "step": 6773 }, { "epoch": 2.2036434612882236, "grad_norm": 2.5788031668373, "learning_rate": 1.999102326172313e-06, "loss": 0.1814, "step": 6774 }, { "epoch": 2.2039687703318154, "grad_norm": 2.1162486638194578, "learning_rate": 1.9975885931546595e-06, "loss": 0.1449, "step": 6775 }, { "epoch": 2.2042940793754067, "grad_norm": 1.9734921310121905, "learning_rate": 1.996075290384565e-06, "loss": 0.1373, "step": 6776 }, { "epoch": 2.204619388418998, "grad_norm": 2.0440525654866812, "learning_rate": 1.9945624180788814e-06, "loss": 0.1497, "step": 6777 }, { "epoch": 2.2049446974625893, "grad_norm": 1.8038209594045709, "learning_rate": 1.993049976454408e-06, "loss": 0.1253, "step": 6778 }, { "epoch": 2.205270006506181, "grad_norm": 2.2971351829667643, "learning_rate": 1.9915379657278794e-06, "loss": 0.2184, "step": 6779 }, { "epoch": 2.2055953155497723, "grad_norm": 2.0720112074784782, "learning_rate": 1.990026386115966e-06, "loss": 0.1777, "step": 6780 }, { "epoch": 2.2059206245933636, "grad_norm": 2.024100659283531, "learning_rate": 1.9885152378352763e-06, "loss": 0.183, "step": 6781 }, { "epoch": 2.2062459336369553, "grad_norm": 2.356136091976491, "learning_rate": 1.987004521102361e-06, "loss": 0.1794, "step": 6782 }, { "epoch": 2.2065712426805466, "grad_norm": 1.9381289151917414, "learning_rate": 1.9854942361337085e-06, "loss": 0.1721, "step": 6783 }, { "epoch": 2.206896551724138, "grad_norm": 2.0023474430336456, "learning_rate": 1.9839843831457402e-06, "loss": 0.1712, "step": 6784 }, { "epoch": 2.207221860767729, "grad_norm": 2.0275239437091512, "learning_rate": 1.9824749623548205e-06, "loss": 0.1388, "step": 6785 }, { "epoch": 2.207547169811321, "grad_norm": 2.204288495380004, "learning_rate": 1.9809659739772523e-06, "loss": 0.1699, "step": 6786 }, { "epoch": 2.2078724788549122, "grad_norm": 1.8813249168721937, "learning_rate": 1.9794574182292712e-06, "loss": 0.1328, "step": 6787 }, { "epoch": 2.2081977878985035, "grad_norm": 2.1172952871240036, "learning_rate": 1.977949295327058e-06, "loss": 0.1529, "step": 6788 }, { "epoch": 2.208523096942095, "grad_norm": 2.2117055535204404, "learning_rate": 1.9764416054867236e-06, "loss": 0.1576, "step": 6789 }, { "epoch": 2.2088484059856865, "grad_norm": 2.4391195665432974, "learning_rate": 1.9749343489243226e-06, "loss": 0.1743, "step": 6790 }, { "epoch": 2.209173715029278, "grad_norm": 1.6138230634803685, "learning_rate": 1.9734275258558487e-06, "loss": 0.1128, "step": 6791 }, { "epoch": 2.209499024072869, "grad_norm": 2.100618179071839, "learning_rate": 1.971921136497228e-06, "loss": 0.1564, "step": 6792 }, { "epoch": 2.209824333116461, "grad_norm": 1.858638317929531, "learning_rate": 1.9704151810643242e-06, "loss": 0.1582, "step": 6793 }, { "epoch": 2.210149642160052, "grad_norm": 2.0156749963070393, "learning_rate": 1.968909659772948e-06, "loss": 0.1426, "step": 6794 }, { "epoch": 2.2104749512036435, "grad_norm": 2.266679173938735, "learning_rate": 1.9674045728388384e-06, "loss": 0.1562, "step": 6795 }, { "epoch": 2.2108002602472347, "grad_norm": 2.089307564733391, "learning_rate": 1.965899920477674e-06, "loss": 0.1742, "step": 6796 }, { "epoch": 2.2111255692908265, "grad_norm": 2.1657853558448688, "learning_rate": 1.964395702905074e-06, "loss": 0.1949, "step": 6797 }, { "epoch": 2.211450878334418, "grad_norm": 2.2883938492126656, "learning_rate": 1.962891920336595e-06, "loss": 0.1435, "step": 6798 }, { "epoch": 2.211776187378009, "grad_norm": 1.878844760287899, "learning_rate": 1.961388572987727e-06, "loss": 0.1253, "step": 6799 }, { "epoch": 2.2121014964216004, "grad_norm": 1.7499246576190315, "learning_rate": 1.9598856610739024e-06, "loss": 0.1542, "step": 6800 }, { "epoch": 2.212426805465192, "grad_norm": 2.7112107105307066, "learning_rate": 1.9583831848104917e-06, "loss": 0.1994, "step": 6801 }, { "epoch": 2.2127521145087834, "grad_norm": 2.032089001182072, "learning_rate": 1.9568811444127965e-06, "loss": 0.1601, "step": 6802 }, { "epoch": 2.2130774235523747, "grad_norm": 2.1763960604724555, "learning_rate": 1.955379540096064e-06, "loss": 0.181, "step": 6803 }, { "epoch": 2.213402732595966, "grad_norm": 2.4991955510290613, "learning_rate": 1.953878372075472e-06, "loss": 0.1928, "step": 6804 }, { "epoch": 2.2137280416395577, "grad_norm": 1.8097500034045346, "learning_rate": 1.95237764056614e-06, "loss": 0.145, "step": 6805 }, { "epoch": 2.214053350683149, "grad_norm": 2.2858609247667387, "learning_rate": 1.9508773457831253e-06, "loss": 0.1952, "step": 6806 }, { "epoch": 2.2143786597267403, "grad_norm": 1.7948376304377147, "learning_rate": 1.9493774879414184e-06, "loss": 0.1609, "step": 6807 }, { "epoch": 2.2147039687703316, "grad_norm": 2.673414476881882, "learning_rate": 1.9478780672559515e-06, "loss": 0.2373, "step": 6808 }, { "epoch": 2.2150292778139233, "grad_norm": 2.0790106351217084, "learning_rate": 1.9463790839415946e-06, "loss": 0.1743, "step": 6809 }, { "epoch": 2.2153545868575146, "grad_norm": 1.8487129931324187, "learning_rate": 1.9448805382131488e-06, "loss": 0.129, "step": 6810 }, { "epoch": 2.215679895901106, "grad_norm": 1.864706518318956, "learning_rate": 1.9433824302853606e-06, "loss": 0.1379, "step": 6811 }, { "epoch": 2.2160052049446977, "grad_norm": 2.1641600104569934, "learning_rate": 1.9418847603729064e-06, "loss": 0.1693, "step": 6812 }, { "epoch": 2.216330513988289, "grad_norm": 2.3112324084412013, "learning_rate": 1.940387528690405e-06, "loss": 0.1939, "step": 6813 }, { "epoch": 2.2166558230318802, "grad_norm": 2.141334472606565, "learning_rate": 1.9388907354524123e-06, "loss": 0.1597, "step": 6814 }, { "epoch": 2.2169811320754715, "grad_norm": 2.388286512964058, "learning_rate": 1.937394380873418e-06, "loss": 0.1835, "step": 6815 }, { "epoch": 2.2173064411190633, "grad_norm": 2.0707048429305734, "learning_rate": 1.9358984651678474e-06, "loss": 0.2081, "step": 6816 }, { "epoch": 2.2176317501626546, "grad_norm": 2.0533820683012136, "learning_rate": 1.9344029885500737e-06, "loss": 0.155, "step": 6817 }, { "epoch": 2.217957059206246, "grad_norm": 1.6224410696223444, "learning_rate": 1.932907951234395e-06, "loss": 0.1028, "step": 6818 }, { "epoch": 2.218282368249837, "grad_norm": 2.052927145154526, "learning_rate": 1.9314133534350504e-06, "loss": 0.1432, "step": 6819 }, { "epoch": 2.218607677293429, "grad_norm": 1.6869501460936054, "learning_rate": 1.9299191953662183e-06, "loss": 0.1321, "step": 6820 }, { "epoch": 2.21893298633702, "grad_norm": 2.1267323054988827, "learning_rate": 1.9284254772420143e-06, "loss": 0.2062, "step": 6821 }, { "epoch": 2.2192582953806115, "grad_norm": 2.0403974246092855, "learning_rate": 1.926932199276485e-06, "loss": 0.1507, "step": 6822 }, { "epoch": 2.219583604424203, "grad_norm": 1.7082915730590549, "learning_rate": 1.925439361683622e-06, "loss": 0.1316, "step": 6823 }, { "epoch": 2.2199089134677945, "grad_norm": 2.1745216972766412, "learning_rate": 1.9239469646773473e-06, "loss": 0.1247, "step": 6824 }, { "epoch": 2.220234222511386, "grad_norm": 2.3782131694689403, "learning_rate": 1.9224550084715234e-06, "loss": 0.1625, "step": 6825 }, { "epoch": 2.220559531554977, "grad_norm": 1.8498348554182942, "learning_rate": 1.92096349327995e-06, "loss": 0.1347, "step": 6826 }, { "epoch": 2.220884840598569, "grad_norm": 1.8682526458454356, "learning_rate": 1.919472419316361e-06, "loss": 0.155, "step": 6827 }, { "epoch": 2.22121014964216, "grad_norm": 2.223118452396102, "learning_rate": 1.917981786794425e-06, "loss": 0.1931, "step": 6828 }, { "epoch": 2.2215354586857514, "grad_norm": 2.380977671838034, "learning_rate": 1.9164915959277568e-06, "loss": 0.2172, "step": 6829 }, { "epoch": 2.2218607677293427, "grad_norm": 1.9594090467211347, "learning_rate": 1.915001846929899e-06, "loss": 0.1747, "step": 6830 }, { "epoch": 2.2221860767729344, "grad_norm": 2.2401845915743297, "learning_rate": 1.9135125400143324e-06, "loss": 0.1951, "step": 6831 }, { "epoch": 2.2225113858165257, "grad_norm": 2.3889809481769344, "learning_rate": 1.912023675394476e-06, "loss": 0.1856, "step": 6832 }, { "epoch": 2.222836694860117, "grad_norm": 2.1803690976532253, "learning_rate": 1.9105352532836875e-06, "loss": 0.1724, "step": 6833 }, { "epoch": 2.2231620039037083, "grad_norm": 1.7690422439518019, "learning_rate": 1.9090472738952555e-06, "loss": 0.15, "step": 6834 }, { "epoch": 2.2234873129473, "grad_norm": 1.7943754844773752, "learning_rate": 1.9075597374424114e-06, "loss": 0.1548, "step": 6835 }, { "epoch": 2.2238126219908914, "grad_norm": 2.0833930698987597, "learning_rate": 1.906072644138317e-06, "loss": 0.1955, "step": 6836 }, { "epoch": 2.2241379310344827, "grad_norm": 2.290665182368085, "learning_rate": 1.904585994196076e-06, "loss": 0.2071, "step": 6837 }, { "epoch": 2.224463240078074, "grad_norm": 2.1267628420766442, "learning_rate": 1.903099787828727e-06, "loss": 0.1795, "step": 6838 }, { "epoch": 2.2247885491216657, "grad_norm": 2.2311166334221864, "learning_rate": 1.9016140252492417e-06, "loss": 0.1477, "step": 6839 }, { "epoch": 2.225113858165257, "grad_norm": 2.7566350653324894, "learning_rate": 1.900128706670532e-06, "loss": 0.1831, "step": 6840 }, { "epoch": 2.2254391672088483, "grad_norm": 1.8613582404151638, "learning_rate": 1.898643832305448e-06, "loss": 0.1418, "step": 6841 }, { "epoch": 2.22576447625244, "grad_norm": 2.232903726065992, "learning_rate": 1.8971594023667684e-06, "loss": 0.1997, "step": 6842 }, { "epoch": 2.2260897852960313, "grad_norm": 2.078363381817112, "learning_rate": 1.8956754170672165e-06, "loss": 0.1428, "step": 6843 }, { "epoch": 2.2264150943396226, "grad_norm": 1.8381570913903877, "learning_rate": 1.8941918766194462e-06, "loss": 0.1285, "step": 6844 }, { "epoch": 2.226740403383214, "grad_norm": 2.4876017300253492, "learning_rate": 1.89270878123605e-06, "loss": 0.1468, "step": 6845 }, { "epoch": 2.2270657124268056, "grad_norm": 2.3028493498345814, "learning_rate": 1.8912261311295594e-06, "loss": 0.2235, "step": 6846 }, { "epoch": 2.227391021470397, "grad_norm": 2.24600844312277, "learning_rate": 1.8897439265124367e-06, "loss": 0.1593, "step": 6847 }, { "epoch": 2.227716330513988, "grad_norm": 2.4133974820902826, "learning_rate": 1.888262167597079e-06, "loss": 0.181, "step": 6848 }, { "epoch": 2.2280416395575795, "grad_norm": 2.2529592400881406, "learning_rate": 1.8867808545958317e-06, "loss": 0.1809, "step": 6849 }, { "epoch": 2.2283669486011712, "grad_norm": 2.4429261117955017, "learning_rate": 1.8852999877209631e-06, "loss": 0.2014, "step": 6850 }, { "epoch": 2.2286922576447625, "grad_norm": 1.6352717044502247, "learning_rate": 1.8838195671846816e-06, "loss": 0.1196, "step": 6851 }, { "epoch": 2.229017566688354, "grad_norm": 2.1408757150512367, "learning_rate": 1.8823395931991333e-06, "loss": 0.1696, "step": 6852 }, { "epoch": 2.2293428757319456, "grad_norm": 2.1227082004378723, "learning_rate": 1.880860065976402e-06, "loss": 0.1846, "step": 6853 }, { "epoch": 2.229668184775537, "grad_norm": 2.2187496839654712, "learning_rate": 1.8793809857285012e-06, "loss": 0.1416, "step": 6854 }, { "epoch": 2.229993493819128, "grad_norm": 2.3039581350344247, "learning_rate": 1.8779023526673873e-06, "loss": 0.2009, "step": 6855 }, { "epoch": 2.2303188028627194, "grad_norm": 1.9226029384814967, "learning_rate": 1.8764241670049465e-06, "loss": 0.136, "step": 6856 }, { "epoch": 2.230644111906311, "grad_norm": 1.9354321106019798, "learning_rate": 1.8749464289530045e-06, "loss": 0.1967, "step": 6857 }, { "epoch": 2.2309694209499025, "grad_norm": 2.334876277864743, "learning_rate": 1.873469138723325e-06, "loss": 0.2223, "step": 6858 }, { "epoch": 2.2312947299934938, "grad_norm": 2.165567983059593, "learning_rate": 1.8719922965276006e-06, "loss": 0.1644, "step": 6859 }, { "epoch": 2.231620039037085, "grad_norm": 2.1321214492074474, "learning_rate": 1.8705159025774656e-06, "loss": 0.2102, "step": 6860 }, { "epoch": 2.231945348080677, "grad_norm": 2.5351455661713658, "learning_rate": 1.8690399570844903e-06, "loss": 0.1667, "step": 6861 }, { "epoch": 2.232270657124268, "grad_norm": 2.0423015947129417, "learning_rate": 1.8675644602601744e-06, "loss": 0.1428, "step": 6862 }, { "epoch": 2.2325959661678594, "grad_norm": 2.2938407284146183, "learning_rate": 1.8660894123159616e-06, "loss": 0.2009, "step": 6863 }, { "epoch": 2.2329212752114507, "grad_norm": 2.049093357731745, "learning_rate": 1.864614813463224e-06, "loss": 0.1689, "step": 6864 }, { "epoch": 2.2332465842550424, "grad_norm": 2.4556228303976395, "learning_rate": 1.8631406639132743e-06, "loss": 0.1696, "step": 6865 }, { "epoch": 2.2335718932986337, "grad_norm": 2.1223686137743925, "learning_rate": 1.8616669638773606e-06, "loss": 0.1538, "step": 6866 }, { "epoch": 2.233897202342225, "grad_norm": 1.7226433058749193, "learning_rate": 1.860193713566663e-06, "loss": 0.1532, "step": 6867 }, { "epoch": 2.2342225113858163, "grad_norm": 2.0878825423387357, "learning_rate": 1.8587209131922983e-06, "loss": 0.1723, "step": 6868 }, { "epoch": 2.234547820429408, "grad_norm": 2.350758589324767, "learning_rate": 1.8572485629653208e-06, "loss": 0.1733, "step": 6869 }, { "epoch": 2.2348731294729993, "grad_norm": 1.8800975047096327, "learning_rate": 1.8557766630967217e-06, "loss": 0.1389, "step": 6870 }, { "epoch": 2.2351984385165906, "grad_norm": 2.2751028427184155, "learning_rate": 1.854305213797421e-06, "loss": 0.2001, "step": 6871 }, { "epoch": 2.2355237475601824, "grad_norm": 2.225295153780684, "learning_rate": 1.8528342152782808e-06, "loss": 0.205, "step": 6872 }, { "epoch": 2.2358490566037736, "grad_norm": 1.8948487206180724, "learning_rate": 1.8513636677500974e-06, "loss": 0.1796, "step": 6873 }, { "epoch": 2.236174365647365, "grad_norm": 2.0114494512987964, "learning_rate": 1.8498935714235977e-06, "loss": 0.1556, "step": 6874 }, { "epoch": 2.2364996746909562, "grad_norm": 2.1696333060136963, "learning_rate": 1.8484239265094517e-06, "loss": 0.159, "step": 6875 }, { "epoch": 2.236824983734548, "grad_norm": 2.0743482873034336, "learning_rate": 1.8469547332182563e-06, "loss": 0.1682, "step": 6876 }, { "epoch": 2.2371502927781393, "grad_norm": 2.447301360233504, "learning_rate": 1.8454859917605499e-06, "loss": 0.1401, "step": 6877 }, { "epoch": 2.2374756018217306, "grad_norm": 2.5484693349598344, "learning_rate": 1.8440177023468059e-06, "loss": 0.1537, "step": 6878 }, { "epoch": 2.237800910865322, "grad_norm": 2.3677319041359706, "learning_rate": 1.8425498651874296e-06, "loss": 0.2102, "step": 6879 }, { "epoch": 2.2381262199089136, "grad_norm": 1.928888541628724, "learning_rate": 1.8410824804927585e-06, "loss": 0.1471, "step": 6880 }, { "epoch": 2.238451528952505, "grad_norm": 2.039695432735263, "learning_rate": 1.8396155484730783e-06, "loss": 0.1304, "step": 6881 }, { "epoch": 2.238776837996096, "grad_norm": 2.02000658899163, "learning_rate": 1.8381490693385968e-06, "loss": 0.1868, "step": 6882 }, { "epoch": 2.239102147039688, "grad_norm": 2.488528866570601, "learning_rate": 1.8366830432994604e-06, "loss": 0.2269, "step": 6883 }, { "epoch": 2.239427456083279, "grad_norm": 1.9624396360615124, "learning_rate": 1.8352174705657527e-06, "loss": 0.1762, "step": 6884 }, { "epoch": 2.2397527651268705, "grad_norm": 2.097576668025825, "learning_rate": 1.8337523513474931e-06, "loss": 0.1642, "step": 6885 }, { "epoch": 2.240078074170462, "grad_norm": 1.9331544626366377, "learning_rate": 1.8322876858546308e-06, "loss": 0.1327, "step": 6886 }, { "epoch": 2.2404033832140535, "grad_norm": 2.2084526002030986, "learning_rate": 1.8308234742970571e-06, "loss": 0.1724, "step": 6887 }, { "epoch": 2.240728692257645, "grad_norm": 1.9831649847867239, "learning_rate": 1.8293597168845906e-06, "loss": 0.1597, "step": 6888 }, { "epoch": 2.241054001301236, "grad_norm": 1.9601796232212194, "learning_rate": 1.8278964138269912e-06, "loss": 0.115, "step": 6889 }, { "epoch": 2.2413793103448274, "grad_norm": 1.938303881080863, "learning_rate": 1.8264335653339515e-06, "loss": 0.1839, "step": 6890 }, { "epoch": 2.241704619388419, "grad_norm": 2.1734331575574344, "learning_rate": 1.824971171615097e-06, "loss": 0.2201, "step": 6891 }, { "epoch": 2.2420299284320104, "grad_norm": 1.8479424932362665, "learning_rate": 1.8235092328799897e-06, "loss": 0.1301, "step": 6892 }, { "epoch": 2.2423552374756017, "grad_norm": 2.3244740892484157, "learning_rate": 1.8220477493381294e-06, "loss": 0.1991, "step": 6893 }, { "epoch": 2.2426805465191935, "grad_norm": 2.1656027051616324, "learning_rate": 1.8205867211989437e-06, "loss": 0.1421, "step": 6894 }, { "epoch": 2.2430058555627848, "grad_norm": 2.3771283946115322, "learning_rate": 1.819126148671802e-06, "loss": 0.2019, "step": 6895 }, { "epoch": 2.243331164606376, "grad_norm": 2.1547441680968986, "learning_rate": 1.8176660319660022e-06, "loss": 0.1682, "step": 6896 }, { "epoch": 2.2436564736499673, "grad_norm": 2.3241624042948628, "learning_rate": 1.8162063712907824e-06, "loss": 0.1869, "step": 6897 }, { "epoch": 2.243981782693559, "grad_norm": 2.0396068827563307, "learning_rate": 1.8147471668553135e-06, "loss": 0.1706, "step": 6898 }, { "epoch": 2.2443070917371504, "grad_norm": 2.0154508846253196, "learning_rate": 1.8132884188686994e-06, "loss": 0.1185, "step": 6899 }, { "epoch": 2.2446324007807417, "grad_norm": 2.0102244012785406, "learning_rate": 1.8118301275399764e-06, "loss": 0.1529, "step": 6900 }, { "epoch": 2.244957709824333, "grad_norm": 1.9345958548238613, "learning_rate": 1.8103722930781249e-06, "loss": 0.1518, "step": 6901 }, { "epoch": 2.2452830188679247, "grad_norm": 1.9675981113922134, "learning_rate": 1.8089149156920504e-06, "loss": 0.1235, "step": 6902 }, { "epoch": 2.245608327911516, "grad_norm": 2.089909032312556, "learning_rate": 1.8074579955905946e-06, "loss": 0.2121, "step": 6903 }, { "epoch": 2.2459336369551073, "grad_norm": 2.03377901747013, "learning_rate": 1.8060015329825364e-06, "loss": 0.1445, "step": 6904 }, { "epoch": 2.2462589459986986, "grad_norm": 1.8321204971252143, "learning_rate": 1.80454552807659e-06, "loss": 0.1275, "step": 6905 }, { "epoch": 2.2465842550422903, "grad_norm": 2.539350688330766, "learning_rate": 1.8030899810813979e-06, "loss": 0.1888, "step": 6906 }, { "epoch": 2.2469095640858816, "grad_norm": 2.38410464719978, "learning_rate": 1.8016348922055448e-06, "loss": 0.2245, "step": 6907 }, { "epoch": 2.247234873129473, "grad_norm": 2.1673303120713543, "learning_rate": 1.8001802616575426e-06, "loss": 0.159, "step": 6908 }, { "epoch": 2.247560182173064, "grad_norm": 2.4326859685229034, "learning_rate": 1.798726089645842e-06, "loss": 0.1693, "step": 6909 }, { "epoch": 2.247885491216656, "grad_norm": 2.3070510298443465, "learning_rate": 1.797272376378829e-06, "loss": 0.1702, "step": 6910 }, { "epoch": 2.248210800260247, "grad_norm": 2.057538214793607, "learning_rate": 1.795819122064818e-06, "loss": 0.1665, "step": 6911 }, { "epoch": 2.2485361093038385, "grad_norm": 2.190561197908345, "learning_rate": 1.7943663269120636e-06, "loss": 0.1536, "step": 6912 }, { "epoch": 2.2488614183474303, "grad_norm": 1.8546231892220382, "learning_rate": 1.7929139911287529e-06, "loss": 0.1322, "step": 6913 }, { "epoch": 2.2491867273910215, "grad_norm": 2.1671969177389516, "learning_rate": 1.791462114923006e-06, "loss": 0.1897, "step": 6914 }, { "epoch": 2.249512036434613, "grad_norm": 2.2555535943756193, "learning_rate": 1.7900106985028753e-06, "loss": 0.171, "step": 6915 }, { "epoch": 2.249837345478204, "grad_norm": 2.055222023270222, "learning_rate": 1.7885597420763517e-06, "loss": 0.167, "step": 6916 }, { "epoch": 2.250162654521796, "grad_norm": 2.0602823591352006, "learning_rate": 1.78710924585136e-06, "loss": 0.1921, "step": 6917 }, { "epoch": 2.250487963565387, "grad_norm": 1.6076211814143344, "learning_rate": 1.7856592100357539e-06, "loss": 0.1251, "step": 6918 }, { "epoch": 2.2508132726089785, "grad_norm": 2.433130095252691, "learning_rate": 1.7842096348373279e-06, "loss": 0.2007, "step": 6919 }, { "epoch": 2.2511385816525697, "grad_norm": 2.521518424294966, "learning_rate": 1.7827605204638033e-06, "loss": 0.1892, "step": 6920 }, { "epoch": 2.2514638906961615, "grad_norm": 2.0828069457867007, "learning_rate": 1.7813118671228413e-06, "loss": 0.1976, "step": 6921 }, { "epoch": 2.251789199739753, "grad_norm": 2.1783361026423336, "learning_rate": 1.7798636750220361e-06, "loss": 0.2029, "step": 6922 }, { "epoch": 2.252114508783344, "grad_norm": 2.169800129425055, "learning_rate": 1.7784159443689119e-06, "loss": 0.2036, "step": 6923 }, { "epoch": 2.252439817826936, "grad_norm": 2.1081474362713792, "learning_rate": 1.7769686753709308e-06, "loss": 0.158, "step": 6924 }, { "epoch": 2.252765126870527, "grad_norm": 1.9350517792170217, "learning_rate": 1.775521868235489e-06, "loss": 0.1268, "step": 6925 }, { "epoch": 2.2530904359141184, "grad_norm": 2.306623840293062, "learning_rate": 1.774075523169912e-06, "loss": 0.1871, "step": 6926 }, { "epoch": 2.2534157449577097, "grad_norm": 2.35466977964194, "learning_rate": 1.7726296403814625e-06, "loss": 0.1741, "step": 6927 }, { "epoch": 2.2537410540013014, "grad_norm": 1.9502648571306394, "learning_rate": 1.7711842200773398e-06, "loss": 0.141, "step": 6928 }, { "epoch": 2.2540663630448927, "grad_norm": 2.1637890260086086, "learning_rate": 1.7697392624646692e-06, "loss": 0.1605, "step": 6929 }, { "epoch": 2.254391672088484, "grad_norm": 2.001651523863071, "learning_rate": 1.7682947677505174e-06, "loss": 0.1419, "step": 6930 }, { "epoch": 2.2547169811320753, "grad_norm": 1.8531987164907904, "learning_rate": 1.7668507361418784e-06, "loss": 0.1482, "step": 6931 }, { "epoch": 2.255042290175667, "grad_norm": 2.00960368489261, "learning_rate": 1.7654071678456846e-06, "loss": 0.1639, "step": 6932 }, { "epoch": 2.2553675992192583, "grad_norm": 2.1760153662324915, "learning_rate": 1.7639640630688027e-06, "loss": 0.1673, "step": 6933 }, { "epoch": 2.2556929082628496, "grad_norm": 2.255798707975389, "learning_rate": 1.762521422018027e-06, "loss": 0.1927, "step": 6934 }, { "epoch": 2.2560182173064414, "grad_norm": 2.041402925415155, "learning_rate": 1.7610792449000873e-06, "loss": 0.1399, "step": 6935 }, { "epoch": 2.2563435263500327, "grad_norm": 2.482634028239722, "learning_rate": 1.7596375319216546e-06, "loss": 0.1735, "step": 6936 }, { "epoch": 2.256668835393624, "grad_norm": 2.155017653632303, "learning_rate": 1.758196283289324e-06, "loss": 0.1581, "step": 6937 }, { "epoch": 2.2569941444372152, "grad_norm": 2.2632275343347916, "learning_rate": 1.7567554992096253e-06, "loss": 0.1751, "step": 6938 }, { "epoch": 2.2573194534808065, "grad_norm": 2.2522122389632555, "learning_rate": 1.7553151798890266e-06, "loss": 0.1841, "step": 6939 }, { "epoch": 2.2576447625243983, "grad_norm": 2.066341075466503, "learning_rate": 1.7538753255339274e-06, "loss": 0.1632, "step": 6940 }, { "epoch": 2.2579700715679896, "grad_norm": 1.9377602995009087, "learning_rate": 1.7524359363506565e-06, "loss": 0.1407, "step": 6941 }, { "epoch": 2.258295380611581, "grad_norm": 2.1800468679201965, "learning_rate": 1.7509970125454829e-06, "loss": 0.1733, "step": 6942 }, { "epoch": 2.2586206896551726, "grad_norm": 2.112968928064591, "learning_rate": 1.7495585543246018e-06, "loss": 0.1553, "step": 6943 }, { "epoch": 2.258945998698764, "grad_norm": 2.272718935440992, "learning_rate": 1.7481205618941472e-06, "loss": 0.1866, "step": 6944 }, { "epoch": 2.259271307742355, "grad_norm": 2.240579694272307, "learning_rate": 1.746683035460186e-06, "loss": 0.1641, "step": 6945 }, { "epoch": 2.2595966167859465, "grad_norm": 1.992802314482237, "learning_rate": 1.7452459752287131e-06, "loss": 0.1483, "step": 6946 }, { "epoch": 2.259921925829538, "grad_norm": 1.9677855138430163, "learning_rate": 1.7438093814056623e-06, "loss": 0.1797, "step": 6947 }, { "epoch": 2.2602472348731295, "grad_norm": 1.9706394582177855, "learning_rate": 1.7423732541968992e-06, "loss": 0.1428, "step": 6948 }, { "epoch": 2.260572543916721, "grad_norm": 1.998059240385895, "learning_rate": 1.7409375938082196e-06, "loss": 0.135, "step": 6949 }, { "epoch": 2.260897852960312, "grad_norm": 1.972682741110319, "learning_rate": 1.739502400445357e-06, "loss": 0.1645, "step": 6950 }, { "epoch": 2.261223162003904, "grad_norm": 1.9769251243229116, "learning_rate": 1.7380676743139729e-06, "loss": 0.1904, "step": 6951 }, { "epoch": 2.261548471047495, "grad_norm": 2.1714270566714777, "learning_rate": 1.7366334156196656e-06, "loss": 0.1774, "step": 6952 }, { "epoch": 2.2618737800910864, "grad_norm": 2.5837740406741285, "learning_rate": 1.7351996245679676e-06, "loss": 0.1938, "step": 6953 }, { "epoch": 2.262199089134678, "grad_norm": 1.9736344191400428, "learning_rate": 1.7337663013643397e-06, "loss": 0.1577, "step": 6954 }, { "epoch": 2.2625243981782694, "grad_norm": 1.9276443934668306, "learning_rate": 1.7323334462141773e-06, "loss": 0.1089, "step": 6955 }, { "epoch": 2.2628497072218607, "grad_norm": 2.256687091638295, "learning_rate": 1.7309010593228103e-06, "loss": 0.1485, "step": 6956 }, { "epoch": 2.263175016265452, "grad_norm": 2.071609809160053, "learning_rate": 1.7294691408955033e-06, "loss": 0.1808, "step": 6957 }, { "epoch": 2.2635003253090438, "grad_norm": 2.257542461807334, "learning_rate": 1.7280376911374468e-06, "loss": 0.1499, "step": 6958 }, { "epoch": 2.263825634352635, "grad_norm": 2.064062031583758, "learning_rate": 1.726606710253771e-06, "loss": 0.1213, "step": 6959 }, { "epoch": 2.2641509433962264, "grad_norm": 2.1476812140576382, "learning_rate": 1.7251761984495374e-06, "loss": 0.1641, "step": 6960 }, { "epoch": 2.2644762524398176, "grad_norm": 2.143587574181965, "learning_rate": 1.723746155929737e-06, "loss": 0.1593, "step": 6961 }, { "epoch": 2.2648015614834094, "grad_norm": 1.9782177613854801, "learning_rate": 1.7223165828992977e-06, "loss": 0.1382, "step": 6962 }, { "epoch": 2.2651268705270007, "grad_norm": 2.1163396206953284, "learning_rate": 1.7208874795630765e-06, "loss": 0.1458, "step": 6963 }, { "epoch": 2.265452179570592, "grad_norm": 1.9352645016139376, "learning_rate": 1.7194588461258649e-06, "loss": 0.1454, "step": 6964 }, { "epoch": 2.2657774886141837, "grad_norm": 2.5040641120473484, "learning_rate": 1.7180306827923905e-06, "loss": 0.1537, "step": 6965 }, { "epoch": 2.266102797657775, "grad_norm": 1.953457471594786, "learning_rate": 1.716602989767307e-06, "loss": 0.1455, "step": 6966 }, { "epoch": 2.2664281067013663, "grad_norm": 2.104512264153626, "learning_rate": 1.7151757672552004e-06, "loss": 0.1501, "step": 6967 }, { "epoch": 2.2667534157449576, "grad_norm": 2.2065241884194267, "learning_rate": 1.7137490154606e-06, "loss": 0.2081, "step": 6968 }, { "epoch": 2.267078724788549, "grad_norm": 2.463912043637303, "learning_rate": 1.7123227345879573e-06, "loss": 0.2545, "step": 6969 }, { "epoch": 2.2674040338321406, "grad_norm": 2.0727082154466516, "learning_rate": 1.710896924841657e-06, "loss": 0.112, "step": 6970 }, { "epoch": 2.267729342875732, "grad_norm": 2.278243713737677, "learning_rate": 1.7094715864260198e-06, "loss": 0.1799, "step": 6971 }, { "epoch": 2.268054651919323, "grad_norm": 2.0860972199752137, "learning_rate": 1.7080467195452999e-06, "loss": 0.1741, "step": 6972 }, { "epoch": 2.268379960962915, "grad_norm": 1.6958663029144754, "learning_rate": 1.7066223244036783e-06, "loss": 0.1395, "step": 6973 }, { "epoch": 2.2687052700065062, "grad_norm": 2.0939878305742763, "learning_rate": 1.7051984012052753e-06, "loss": 0.1397, "step": 6974 }, { "epoch": 2.2690305790500975, "grad_norm": 1.8615811461987117, "learning_rate": 1.7037749501541361e-06, "loss": 0.1267, "step": 6975 }, { "epoch": 2.269355888093689, "grad_norm": 2.114696510962371, "learning_rate": 1.7023519714542447e-06, "loss": 0.1596, "step": 6976 }, { "epoch": 2.2696811971372806, "grad_norm": 2.114875457574273, "learning_rate": 1.700929465309516e-06, "loss": 0.1944, "step": 6977 }, { "epoch": 2.270006506180872, "grad_norm": 1.959079898239379, "learning_rate": 1.6995074319237925e-06, "loss": 0.1453, "step": 6978 }, { "epoch": 2.270331815224463, "grad_norm": 2.015180623824408, "learning_rate": 1.6980858715008546e-06, "loss": 0.1589, "step": 6979 }, { "epoch": 2.2706571242680544, "grad_norm": 2.262416753152574, "learning_rate": 1.6966647842444145e-06, "loss": 0.1771, "step": 6980 }, { "epoch": 2.270982433311646, "grad_norm": 2.1642715340305023, "learning_rate": 1.6952441703581118e-06, "loss": 0.1393, "step": 6981 }, { "epoch": 2.2713077423552375, "grad_norm": 1.782913389242029, "learning_rate": 1.6938240300455243e-06, "loss": 0.1215, "step": 6982 }, { "epoch": 2.2716330513988288, "grad_norm": 2.0758565352180907, "learning_rate": 1.6924043635101562e-06, "loss": 0.1464, "step": 6983 }, { "epoch": 2.2719583604424205, "grad_norm": 2.237845632226076, "learning_rate": 1.690985170955448e-06, "loss": 0.1605, "step": 6984 }, { "epoch": 2.272283669486012, "grad_norm": 2.058721095901164, "learning_rate": 1.6895664525847726e-06, "loss": 0.1696, "step": 6985 }, { "epoch": 2.272608978529603, "grad_norm": 2.1292571307803776, "learning_rate": 1.6881482086014317e-06, "loss": 0.128, "step": 6986 }, { "epoch": 2.2729342875731944, "grad_norm": 2.1939386297092343, "learning_rate": 1.6867304392086575e-06, "loss": 0.1542, "step": 6987 }, { "epoch": 2.273259596616786, "grad_norm": 2.1959091872782057, "learning_rate": 1.6853131446096238e-06, "loss": 0.1571, "step": 6988 }, { "epoch": 2.2735849056603774, "grad_norm": 2.1593662111422924, "learning_rate": 1.683896325007426e-06, "loss": 0.1407, "step": 6989 }, { "epoch": 2.2739102147039687, "grad_norm": 2.110272030122259, "learning_rate": 1.6824799806050945e-06, "loss": 0.1881, "step": 6990 }, { "epoch": 2.27423552374756, "grad_norm": 2.4460705769152997, "learning_rate": 1.6810641116055932e-06, "loss": 0.1848, "step": 6991 }, { "epoch": 2.2745608327911517, "grad_norm": 2.1660648942801646, "learning_rate": 1.67964871821182e-06, "loss": 0.1621, "step": 6992 }, { "epoch": 2.274886141834743, "grad_norm": 2.215560854424452, "learning_rate": 1.678233800626597e-06, "loss": 0.167, "step": 6993 }, { "epoch": 2.2752114508783343, "grad_norm": 2.3484608149566433, "learning_rate": 1.6768193590526865e-06, "loss": 0.1902, "step": 6994 }, { "epoch": 2.275536759921926, "grad_norm": 2.03304489185137, "learning_rate": 1.6754053936927767e-06, "loss": 0.1954, "step": 6995 }, { "epoch": 2.2758620689655173, "grad_norm": 2.1276712191973135, "learning_rate": 1.6739919047494902e-06, "loss": 0.1914, "step": 6996 }, { "epoch": 2.2761873780091086, "grad_norm": 2.1841636716396815, "learning_rate": 1.6725788924253827e-06, "loss": 0.1759, "step": 6997 }, { "epoch": 2.2765126870527, "grad_norm": 2.0343664658510123, "learning_rate": 1.671166356922937e-06, "loss": 0.1412, "step": 6998 }, { "epoch": 2.2768379960962912, "grad_norm": 1.9955107206981815, "learning_rate": 1.6697542984445713e-06, "loss": 0.1798, "step": 6999 }, { "epoch": 2.277163305139883, "grad_norm": 2.324116706379114, "learning_rate": 1.6683427171926376e-06, "loss": 0.1491, "step": 7000 }, { "epoch": 2.2774886141834743, "grad_norm": 1.9547770064671606, "learning_rate": 1.666931613369413e-06, "loss": 0.1686, "step": 7001 }, { "epoch": 2.2778139232270656, "grad_norm": 2.136337763257382, "learning_rate": 1.6655209871771089e-06, "loss": 0.1559, "step": 7002 }, { "epoch": 2.2781392322706573, "grad_norm": 2.2947085252251003, "learning_rate": 1.664110838817871e-06, "loss": 0.2198, "step": 7003 }, { "epoch": 2.2784645413142486, "grad_norm": 2.3902298412479195, "learning_rate": 1.6627011684937755e-06, "loss": 0.1494, "step": 7004 }, { "epoch": 2.27878985035784, "grad_norm": 1.9937358244454881, "learning_rate": 1.6612919764068259e-06, "loss": 0.16, "step": 7005 }, { "epoch": 2.279115159401431, "grad_norm": 2.1986418400603953, "learning_rate": 1.659883262758964e-06, "loss": 0.1646, "step": 7006 }, { "epoch": 2.279440468445023, "grad_norm": 2.099935931087243, "learning_rate": 1.658475027752055e-06, "loss": 0.2017, "step": 7007 }, { "epoch": 2.279765777488614, "grad_norm": 1.9964740879663871, "learning_rate": 1.6570672715879033e-06, "loss": 0.1237, "step": 7008 }, { "epoch": 2.2800910865322055, "grad_norm": 2.024714674874983, "learning_rate": 1.6556599944682417e-06, "loss": 0.13, "step": 7009 }, { "epoch": 2.280416395575797, "grad_norm": 2.5990574545364815, "learning_rate": 1.654253196594731e-06, "loss": 0.187, "step": 7010 }, { "epoch": 2.2807417046193885, "grad_norm": 1.9438828294039723, "learning_rate": 1.6528468781689683e-06, "loss": 0.1673, "step": 7011 }, { "epoch": 2.28106701366298, "grad_norm": 2.055894662652117, "learning_rate": 1.6514410393924806e-06, "loss": 0.1741, "step": 7012 }, { "epoch": 2.281392322706571, "grad_norm": 2.4876776302027612, "learning_rate": 1.650035680466724e-06, "loss": 0.1794, "step": 7013 }, { "epoch": 2.281717631750163, "grad_norm": 2.077612748961533, "learning_rate": 1.6486308015930885e-06, "loss": 0.1477, "step": 7014 }, { "epoch": 2.282042940793754, "grad_norm": 1.738926972906467, "learning_rate": 1.6472264029728924e-06, "loss": 0.1302, "step": 7015 }, { "epoch": 2.2823682498373454, "grad_norm": 2.165466874182432, "learning_rate": 1.6458224848073878e-06, "loss": 0.1651, "step": 7016 }, { "epoch": 2.2826935588809367, "grad_norm": 2.140977557357444, "learning_rate": 1.6444190472977594e-06, "loss": 0.171, "step": 7017 }, { "epoch": 2.2830188679245285, "grad_norm": 2.2311223174607036, "learning_rate": 1.6430160906451186e-06, "loss": 0.1415, "step": 7018 }, { "epoch": 2.2833441769681198, "grad_norm": 2.2876054314855034, "learning_rate": 1.6416136150505073e-06, "loss": 0.2339, "step": 7019 }, { "epoch": 2.283669486011711, "grad_norm": 1.9258691091534734, "learning_rate": 1.6402116207149072e-06, "loss": 0.135, "step": 7020 }, { "epoch": 2.2839947950553023, "grad_norm": 2.426080151394368, "learning_rate": 1.6388101078392216e-06, "loss": 0.2268, "step": 7021 }, { "epoch": 2.284320104098894, "grad_norm": 2.3298337328999605, "learning_rate": 1.6374090766242873e-06, "loss": 0.152, "step": 7022 }, { "epoch": 2.2846454131424854, "grad_norm": 2.080742023809348, "learning_rate": 1.6360085272708741e-06, "loss": 0.158, "step": 7023 }, { "epoch": 2.2849707221860767, "grad_norm": 2.0839417671914338, "learning_rate": 1.6346084599796835e-06, "loss": 0.1763, "step": 7024 }, { "epoch": 2.2852960312296684, "grad_norm": 1.8938552089561977, "learning_rate": 1.6332088749513426e-06, "loss": 0.2036, "step": 7025 }, { "epoch": 2.2856213402732597, "grad_norm": 1.6770726334895563, "learning_rate": 1.6318097723864163e-06, "loss": 0.1193, "step": 7026 }, { "epoch": 2.285946649316851, "grad_norm": 2.0598286916990314, "learning_rate": 1.6304111524853938e-06, "loss": 0.1359, "step": 7027 }, { "epoch": 2.2862719583604423, "grad_norm": 1.9734186038400972, "learning_rate": 1.6290130154487e-06, "loss": 0.1286, "step": 7028 }, { "epoch": 2.286597267404034, "grad_norm": 1.965267015719461, "learning_rate": 1.6276153614766898e-06, "loss": 0.162, "step": 7029 }, { "epoch": 2.2869225764476253, "grad_norm": 1.9477131629044648, "learning_rate": 1.6262181907696456e-06, "loss": 0.1509, "step": 7030 }, { "epoch": 2.2872478854912166, "grad_norm": 2.208461901424453, "learning_rate": 1.6248215035277838e-06, "loss": 0.1784, "step": 7031 }, { "epoch": 2.287573194534808, "grad_norm": 2.4618551493851277, "learning_rate": 1.6234252999512523e-06, "loss": 0.221, "step": 7032 }, { "epoch": 2.2878985035783996, "grad_norm": 1.9910571397268637, "learning_rate": 1.6220295802401248e-06, "loss": 0.181, "step": 7033 }, { "epoch": 2.288223812621991, "grad_norm": 1.9482457602854186, "learning_rate": 1.6206343445944128e-06, "loss": 0.1508, "step": 7034 }, { "epoch": 2.288549121665582, "grad_norm": 2.4508014892257717, "learning_rate": 1.6192395932140504e-06, "loss": 0.1487, "step": 7035 }, { "epoch": 2.288874430709174, "grad_norm": 2.06192680775385, "learning_rate": 1.6178453262989085e-06, "loss": 0.1892, "step": 7036 }, { "epoch": 2.2891997397527653, "grad_norm": 2.4117081645053213, "learning_rate": 1.6164515440487881e-06, "loss": 0.1866, "step": 7037 }, { "epoch": 2.2895250487963565, "grad_norm": 1.905580532183812, "learning_rate": 1.6150582466634173e-06, "loss": 0.1627, "step": 7038 }, { "epoch": 2.289850357839948, "grad_norm": 2.2321833258524584, "learning_rate": 1.613665434342453e-06, "loss": 0.1922, "step": 7039 }, { "epoch": 2.290175666883539, "grad_norm": 2.027720530476869, "learning_rate": 1.612273107285493e-06, "loss": 0.1297, "step": 7040 }, { "epoch": 2.290500975927131, "grad_norm": 2.1263537033216804, "learning_rate": 1.6108812656920558e-06, "loss": 0.1727, "step": 7041 }, { "epoch": 2.290826284970722, "grad_norm": 1.9368908297853904, "learning_rate": 1.6094899097615912e-06, "loss": 0.1247, "step": 7042 }, { "epoch": 2.2911515940143135, "grad_norm": 2.2847981105160198, "learning_rate": 1.6080990396934825e-06, "loss": 0.1673, "step": 7043 }, { "epoch": 2.291476903057905, "grad_norm": 2.1299340960348276, "learning_rate": 1.6067086556870453e-06, "loss": 0.1609, "step": 7044 }, { "epoch": 2.2918022121014965, "grad_norm": 2.3475914104675857, "learning_rate": 1.605318757941518e-06, "loss": 0.2098, "step": 7045 }, { "epoch": 2.2921275211450878, "grad_norm": 1.7085452284511504, "learning_rate": 1.603929346656078e-06, "loss": 0.1277, "step": 7046 }, { "epoch": 2.292452830188679, "grad_norm": 2.1724522401862787, "learning_rate": 1.602540422029825e-06, "loss": 0.1536, "step": 7047 }, { "epoch": 2.292778139232271, "grad_norm": 2.0227830090177275, "learning_rate": 1.6011519842617945e-06, "loss": 0.1269, "step": 7048 }, { "epoch": 2.293103448275862, "grad_norm": 1.987085289866579, "learning_rate": 1.5997640335509528e-06, "loss": 0.1616, "step": 7049 }, { "epoch": 2.2934287573194534, "grad_norm": 2.3675064353597657, "learning_rate": 1.5983765700961924e-06, "loss": 0.1962, "step": 7050 }, { "epoch": 2.2937540663630447, "grad_norm": 2.04449525331596, "learning_rate": 1.5969895940963338e-06, "loss": 0.1811, "step": 7051 }, { "epoch": 2.2940793754066364, "grad_norm": 2.1172986495141575, "learning_rate": 1.595603105750138e-06, "loss": 0.1573, "step": 7052 }, { "epoch": 2.2944046844502277, "grad_norm": 2.234413066000629, "learning_rate": 1.5942171052562876e-06, "loss": 0.1555, "step": 7053 }, { "epoch": 2.294729993493819, "grad_norm": 2.243494700345058, "learning_rate": 1.5928315928133935e-06, "loss": 0.1853, "step": 7054 }, { "epoch": 2.2950553025374107, "grad_norm": 2.1433976970033326, "learning_rate": 1.591446568620007e-06, "loss": 0.1234, "step": 7055 }, { "epoch": 2.295380611581002, "grad_norm": 2.3061978796045115, "learning_rate": 1.5900620328745998e-06, "loss": 0.173, "step": 7056 }, { "epoch": 2.2957059206245933, "grad_norm": 1.97794185313991, "learning_rate": 1.5886779857755751e-06, "loss": 0.1692, "step": 7057 }, { "epoch": 2.2960312296681846, "grad_norm": 2.1226295887976336, "learning_rate": 1.5872944275212692e-06, "loss": 0.1971, "step": 7058 }, { "epoch": 2.2963565387117764, "grad_norm": 2.1302091776190246, "learning_rate": 1.5859113583099495e-06, "loss": 0.1479, "step": 7059 }, { "epoch": 2.2966818477553677, "grad_norm": 1.6145186115192693, "learning_rate": 1.5845287783398061e-06, "loss": 0.1141, "step": 7060 }, { "epoch": 2.297007156798959, "grad_norm": 1.8876406859899906, "learning_rate": 1.583146687808968e-06, "loss": 0.1275, "step": 7061 }, { "epoch": 2.2973324658425502, "grad_norm": 2.3263707595104086, "learning_rate": 1.581765086915486e-06, "loss": 0.1861, "step": 7062 }, { "epoch": 2.297657774886142, "grad_norm": 2.276259343376312, "learning_rate": 1.5803839758573457e-06, "loss": 0.1911, "step": 7063 }, { "epoch": 2.2979830839297333, "grad_norm": 1.9964139839716275, "learning_rate": 1.5790033548324635e-06, "loss": 0.1548, "step": 7064 }, { "epoch": 2.2983083929733246, "grad_norm": 1.8635105004956163, "learning_rate": 1.57762322403868e-06, "loss": 0.1704, "step": 7065 }, { "epoch": 2.2986337020169163, "grad_norm": 2.260745841411198, "learning_rate": 1.5762435836737699e-06, "loss": 0.1342, "step": 7066 }, { "epoch": 2.2989590110605076, "grad_norm": 2.06198924850915, "learning_rate": 1.574864433935439e-06, "loss": 0.112, "step": 7067 }, { "epoch": 2.299284320104099, "grad_norm": 1.9845185247137607, "learning_rate": 1.5734857750213167e-06, "loss": 0.1839, "step": 7068 }, { "epoch": 2.29960962914769, "grad_norm": 2.191512342294194, "learning_rate": 1.5721076071289687e-06, "loss": 0.1804, "step": 7069 }, { "epoch": 2.2999349381912815, "grad_norm": 2.2056865496073157, "learning_rate": 1.5707299304558848e-06, "loss": 0.1827, "step": 7070 }, { "epoch": 2.300260247234873, "grad_norm": 2.090391815338713, "learning_rate": 1.5693527451994883e-06, "loss": 0.1323, "step": 7071 }, { "epoch": 2.3005855562784645, "grad_norm": 2.566513923164187, "learning_rate": 1.5679760515571318e-06, "loss": 0.2253, "step": 7072 }, { "epoch": 2.300910865322056, "grad_norm": 1.9294350666363862, "learning_rate": 1.5665998497260959e-06, "loss": 0.1599, "step": 7073 }, { "epoch": 2.3012361743656475, "grad_norm": 2.0719863279968953, "learning_rate": 1.5652241399035873e-06, "loss": 0.1447, "step": 7074 }, { "epoch": 2.301561483409239, "grad_norm": 2.5261284623750444, "learning_rate": 1.5638489222867526e-06, "loss": 0.1879, "step": 7075 }, { "epoch": 2.30188679245283, "grad_norm": 2.113563739078284, "learning_rate": 1.5624741970726587e-06, "loss": 0.1559, "step": 7076 }, { "epoch": 2.3022121014964214, "grad_norm": 2.0336946206871853, "learning_rate": 1.5610999644583018e-06, "loss": 0.1439, "step": 7077 }, { "epoch": 2.302537410540013, "grad_norm": 2.9125719101732312, "learning_rate": 1.5597262246406137e-06, "loss": 0.2262, "step": 7078 }, { "epoch": 2.3028627195836044, "grad_norm": 2.063119451456724, "learning_rate": 1.5583529778164525e-06, "loss": 0.1964, "step": 7079 }, { "epoch": 2.3031880286271957, "grad_norm": 2.112981023999656, "learning_rate": 1.5569802241826026e-06, "loss": 0.1558, "step": 7080 }, { "epoch": 2.303513337670787, "grad_norm": 2.4059786595447634, "learning_rate": 1.555607963935784e-06, "loss": 0.1897, "step": 7081 }, { "epoch": 2.3038386467143788, "grad_norm": 2.5263997708543644, "learning_rate": 1.554236197272639e-06, "loss": 0.1621, "step": 7082 }, { "epoch": 2.30416395575797, "grad_norm": 2.4061929489252982, "learning_rate": 1.5528649243897442e-06, "loss": 0.1755, "step": 7083 }, { "epoch": 2.3044892648015614, "grad_norm": 2.0436216465563657, "learning_rate": 1.5514941454836058e-06, "loss": 0.1668, "step": 7084 }, { "epoch": 2.304814573845153, "grad_norm": 2.113860630553734, "learning_rate": 1.5501238607506558e-06, "loss": 0.1607, "step": 7085 }, { "epoch": 2.3051398828887444, "grad_norm": 2.283134029785756, "learning_rate": 1.5487540703872528e-06, "loss": 0.2021, "step": 7086 }, { "epoch": 2.3054651919323357, "grad_norm": 2.5290398867635817, "learning_rate": 1.5473847745896963e-06, "loss": 0.2422, "step": 7087 }, { "epoch": 2.305790500975927, "grad_norm": 2.249918955725639, "learning_rate": 1.5460159735542036e-06, "loss": 0.1339, "step": 7088 }, { "epoch": 2.3061158100195187, "grad_norm": 2.0592900118868296, "learning_rate": 1.5446476674769228e-06, "loss": 0.1147, "step": 7089 }, { "epoch": 2.30644111906311, "grad_norm": 1.9764863837207531, "learning_rate": 1.5432798565539347e-06, "loss": 0.1624, "step": 7090 }, { "epoch": 2.3067664281067013, "grad_norm": 2.138925153341874, "learning_rate": 1.5419125409812502e-06, "loss": 0.1709, "step": 7091 }, { "epoch": 2.3070917371502926, "grad_norm": 1.946974980708494, "learning_rate": 1.5405457209548024e-06, "loss": 0.1308, "step": 7092 }, { "epoch": 2.3074170461938843, "grad_norm": 2.24914478763305, "learning_rate": 1.539179396670461e-06, "loss": 0.1561, "step": 7093 }, { "epoch": 2.3077423552374756, "grad_norm": 2.464316655699562, "learning_rate": 1.5378135683240174e-06, "loss": 0.1972, "step": 7094 }, { "epoch": 2.308067664281067, "grad_norm": 2.1616045263303945, "learning_rate": 1.5364482361111981e-06, "loss": 0.1777, "step": 7095 }, { "epoch": 2.3083929733246586, "grad_norm": 2.6603113629789816, "learning_rate": 1.5350834002276577e-06, "loss": 0.1021, "step": 7096 }, { "epoch": 2.30871828236825, "grad_norm": 2.360106994769398, "learning_rate": 1.533719060868974e-06, "loss": 0.1969, "step": 7097 }, { "epoch": 2.3090435914118412, "grad_norm": 2.072146082231644, "learning_rate": 1.5323552182306605e-06, "loss": 0.1717, "step": 7098 }, { "epoch": 2.3093689004554325, "grad_norm": 2.0108052440545117, "learning_rate": 1.5309918725081575e-06, "loss": 0.1515, "step": 7099 }, { "epoch": 2.3096942094990243, "grad_norm": 2.4168114388137023, "learning_rate": 1.5296290238968303e-06, "loss": 0.1994, "step": 7100 }, { "epoch": 2.3100195185426156, "grad_norm": 2.0348634450756515, "learning_rate": 1.5282666725919797e-06, "loss": 0.137, "step": 7101 }, { "epoch": 2.310344827586207, "grad_norm": 1.8808808859096038, "learning_rate": 1.526904818788828e-06, "loss": 0.1443, "step": 7102 }, { "epoch": 2.310670136629798, "grad_norm": 1.9876796262550718, "learning_rate": 1.5255434626825317e-06, "loss": 0.1546, "step": 7103 }, { "epoch": 2.31099544567339, "grad_norm": 2.4799712354847654, "learning_rate": 1.5241826044681745e-06, "loss": 0.1404, "step": 7104 }, { "epoch": 2.311320754716981, "grad_norm": 2.286165847573998, "learning_rate": 1.522822244340768e-06, "loss": 0.2012, "step": 7105 }, { "epoch": 2.3116460637605725, "grad_norm": 2.123724454600912, "learning_rate": 1.521462382495249e-06, "loss": 0.1348, "step": 7106 }, { "epoch": 2.311971372804164, "grad_norm": 2.252022760523588, "learning_rate": 1.5201030191264933e-06, "loss": 0.1695, "step": 7107 }, { "epoch": 2.3122966818477555, "grad_norm": 2.2020903893021737, "learning_rate": 1.5187441544292948e-06, "loss": 0.2088, "step": 7108 }, { "epoch": 2.312621990891347, "grad_norm": 2.1554108438556474, "learning_rate": 1.5173857885983784e-06, "loss": 0.1836, "step": 7109 }, { "epoch": 2.312947299934938, "grad_norm": 2.428382023078613, "learning_rate": 1.5160279218284007e-06, "loss": 0.1639, "step": 7110 }, { "epoch": 2.3132726089785294, "grad_norm": 2.3720681669267916, "learning_rate": 1.5146705543139462e-06, "loss": 0.1423, "step": 7111 }, { "epoch": 2.313597918022121, "grad_norm": 2.0827572982571816, "learning_rate": 1.5133136862495229e-06, "loss": 0.2073, "step": 7112 }, { "epoch": 2.3139232270657124, "grad_norm": 2.154193393967696, "learning_rate": 1.5119573178295755e-06, "loss": 0.1727, "step": 7113 }, { "epoch": 2.3142485361093037, "grad_norm": 2.3501929270144215, "learning_rate": 1.5106014492484676e-06, "loss": 0.2202, "step": 7114 }, { "epoch": 2.3145738451528954, "grad_norm": 2.2612753755116675, "learning_rate": 1.5092460807004987e-06, "loss": 0.1752, "step": 7115 }, { "epoch": 2.3148991541964867, "grad_norm": 2.152679114365783, "learning_rate": 1.507891212379896e-06, "loss": 0.2045, "step": 7116 }, { "epoch": 2.315224463240078, "grad_norm": 1.8834505146900558, "learning_rate": 1.5065368444808086e-06, "loss": 0.1435, "step": 7117 }, { "epoch": 2.3155497722836693, "grad_norm": 1.8336583773443862, "learning_rate": 1.505182977197321e-06, "loss": 0.1502, "step": 7118 }, { "epoch": 2.315875081327261, "grad_norm": 1.981608849664676, "learning_rate": 1.5038296107234446e-06, "loss": 0.1388, "step": 7119 }, { "epoch": 2.3162003903708523, "grad_norm": 2.09213368396302, "learning_rate": 1.5024767452531142e-06, "loss": 0.1861, "step": 7120 }, { "epoch": 2.3165256994144436, "grad_norm": 2.1522997256854732, "learning_rate": 1.5011243809801995e-06, "loss": 0.1737, "step": 7121 }, { "epoch": 2.316851008458035, "grad_norm": 1.9628123295651165, "learning_rate": 1.4997725180984924e-06, "loss": 0.1506, "step": 7122 }, { "epoch": 2.3171763175016267, "grad_norm": 2.1894493598363867, "learning_rate": 1.4984211568017171e-06, "loss": 0.1543, "step": 7123 }, { "epoch": 2.317501626545218, "grad_norm": 1.9434753546823211, "learning_rate": 1.497070297283526e-06, "loss": 0.147, "step": 7124 }, { "epoch": 2.3178269355888093, "grad_norm": 1.8056138809617557, "learning_rate": 1.4957199397374971e-06, "loss": 0.1408, "step": 7125 }, { "epoch": 2.318152244632401, "grad_norm": 2.2066212556738316, "learning_rate": 1.4943700843571357e-06, "loss": 0.1566, "step": 7126 }, { "epoch": 2.3184775536759923, "grad_norm": 1.8653558064159133, "learning_rate": 1.4930207313358785e-06, "loss": 0.1516, "step": 7127 }, { "epoch": 2.3188028627195836, "grad_norm": 2.4656589318605606, "learning_rate": 1.49167188086709e-06, "loss": 0.1964, "step": 7128 }, { "epoch": 2.319128171763175, "grad_norm": 2.5682247992684326, "learning_rate": 1.4903235331440586e-06, "loss": 0.1966, "step": 7129 }, { "epoch": 2.3194534808067666, "grad_norm": 2.0681000208166953, "learning_rate": 1.4889756883600048e-06, "loss": 0.1345, "step": 7130 }, { "epoch": 2.319778789850358, "grad_norm": 1.7079822420922923, "learning_rate": 1.4876283467080777e-06, "loss": 0.1408, "step": 7131 }, { "epoch": 2.320104098893949, "grad_norm": 1.9406616570807942, "learning_rate": 1.4862815083813487e-06, "loss": 0.181, "step": 7132 }, { "epoch": 2.3204294079375405, "grad_norm": 2.209338299478757, "learning_rate": 1.4849351735728234e-06, "loss": 0.1498, "step": 7133 }, { "epoch": 2.3207547169811322, "grad_norm": 2.217133356310873, "learning_rate": 1.4835893424754304e-06, "loss": 0.2168, "step": 7134 }, { "epoch": 2.3210800260247235, "grad_norm": 2.2680643624677383, "learning_rate": 1.4822440152820288e-06, "loss": 0.2117, "step": 7135 }, { "epoch": 2.321405335068315, "grad_norm": 2.2229860832700252, "learning_rate": 1.4808991921854065e-06, "loss": 0.1779, "step": 7136 }, { "epoch": 2.3217306441119065, "grad_norm": 1.7058296994839375, "learning_rate": 1.4795548733782767e-06, "loss": 0.1166, "step": 7137 }, { "epoch": 2.322055953155498, "grad_norm": 1.9085179051370496, "learning_rate": 1.4782110590532772e-06, "loss": 0.1501, "step": 7138 }, { "epoch": 2.322381262199089, "grad_norm": 2.184397366262956, "learning_rate": 1.4768677494029842e-06, "loss": 0.169, "step": 7139 }, { "epoch": 2.3227065712426804, "grad_norm": 2.0112931046417377, "learning_rate": 1.4755249446198921e-06, "loss": 0.1591, "step": 7140 }, { "epoch": 2.3230318802862717, "grad_norm": 2.0276415833124646, "learning_rate": 1.4741826448964236e-06, "loss": 0.1809, "step": 7141 }, { "epoch": 2.3233571893298635, "grad_norm": 2.180384667093216, "learning_rate": 1.4728408504249326e-06, "loss": 0.1599, "step": 7142 }, { "epoch": 2.3236824983734548, "grad_norm": 2.035111849289773, "learning_rate": 1.4714995613977012e-06, "loss": 0.1505, "step": 7143 }, { "epoch": 2.324007807417046, "grad_norm": 2.088487082917026, "learning_rate": 1.4701587780069333e-06, "loss": 0.1798, "step": 7144 }, { "epoch": 2.324333116460638, "grad_norm": 2.510750006090566, "learning_rate": 1.4688185004447681e-06, "loss": 0.2078, "step": 7145 }, { "epoch": 2.324658425504229, "grad_norm": 2.0264838128063105, "learning_rate": 1.467478728903264e-06, "loss": 0.1356, "step": 7146 }, { "epoch": 2.3249837345478204, "grad_norm": 2.1004027012732545, "learning_rate": 1.4661394635744136e-06, "loss": 0.2192, "step": 7147 }, { "epoch": 2.3253090435914117, "grad_norm": 2.341021092526263, "learning_rate": 1.4648007046501356e-06, "loss": 0.1804, "step": 7148 }, { "epoch": 2.3256343526350034, "grad_norm": 1.4823512078007304, "learning_rate": 1.463462452322272e-06, "loss": 0.0903, "step": 7149 }, { "epoch": 2.3259596616785947, "grad_norm": 2.1319671747270035, "learning_rate": 1.4621247067825972e-06, "loss": 0.1293, "step": 7150 }, { "epoch": 2.326284970722186, "grad_norm": 2.0624236588374205, "learning_rate": 1.4607874682228117e-06, "loss": 0.1616, "step": 7151 }, { "epoch": 2.3266102797657773, "grad_norm": 2.799911994943437, "learning_rate": 1.4594507368345407e-06, "loss": 0.2035, "step": 7152 }, { "epoch": 2.326935588809369, "grad_norm": 2.0878373950679947, "learning_rate": 1.458114512809341e-06, "loss": 0.1768, "step": 7153 }, { "epoch": 2.3272608978529603, "grad_norm": 2.686520590313222, "learning_rate": 1.456778796338692e-06, "loss": 0.2021, "step": 7154 }, { "epoch": 2.3275862068965516, "grad_norm": 2.4170772675050753, "learning_rate": 1.4554435876140033e-06, "loss": 0.175, "step": 7155 }, { "epoch": 2.3279115159401433, "grad_norm": 2.282505941050174, "learning_rate": 1.4541088868266133e-06, "loss": 0.1644, "step": 7156 }, { "epoch": 2.3282368249837346, "grad_norm": 2.1615620219130074, "learning_rate": 1.4527746941677845e-06, "loss": 0.1786, "step": 7157 }, { "epoch": 2.328562134027326, "grad_norm": 2.143764768619112, "learning_rate": 1.4514410098287036e-06, "loss": 0.144, "step": 7158 }, { "epoch": 2.328887443070917, "grad_norm": 2.1102453465886835, "learning_rate": 1.4501078340004954e-06, "loss": 0.1504, "step": 7159 }, { "epoch": 2.329212752114509, "grad_norm": 1.8772505791601632, "learning_rate": 1.4487751668742006e-06, "loss": 0.1426, "step": 7160 }, { "epoch": 2.3295380611581002, "grad_norm": 2.2718773134178347, "learning_rate": 1.4474430086407915e-06, "loss": 0.1751, "step": 7161 }, { "epoch": 2.3298633702016915, "grad_norm": 2.2320206589442115, "learning_rate": 1.446111359491167e-06, "loss": 0.1533, "step": 7162 }, { "epoch": 2.330188679245283, "grad_norm": 2.311179661329624, "learning_rate": 1.4447802196161564e-06, "loss": 0.1841, "step": 7163 }, { "epoch": 2.3305139882888746, "grad_norm": 1.8635599407602306, "learning_rate": 1.4434495892065088e-06, "loss": 0.162, "step": 7164 }, { "epoch": 2.330839297332466, "grad_norm": 2.066636430894424, "learning_rate": 1.4421194684529077e-06, "loss": 0.1553, "step": 7165 }, { "epoch": 2.331164606376057, "grad_norm": 1.9704939182723857, "learning_rate": 1.4407898575459578e-06, "loss": 0.153, "step": 7166 }, { "epoch": 2.331489915419649, "grad_norm": 1.8136955621653155, "learning_rate": 1.4394607566761941e-06, "loss": 0.1282, "step": 7167 }, { "epoch": 2.33181522446324, "grad_norm": 2.319584202132812, "learning_rate": 1.4381321660340791e-06, "loss": 0.2293, "step": 7168 }, { "epoch": 2.3321405335068315, "grad_norm": 2.011731413234165, "learning_rate": 1.4368040858099986e-06, "loss": 0.1397, "step": 7169 }, { "epoch": 2.3324658425504228, "grad_norm": 2.2627402091463624, "learning_rate": 1.435476516194268e-06, "loss": 0.1571, "step": 7170 }, { "epoch": 2.332791151594014, "grad_norm": 2.101024258351642, "learning_rate": 1.434149457377131e-06, "loss": 0.1331, "step": 7171 }, { "epoch": 2.333116460637606, "grad_norm": 2.044278913912239, "learning_rate": 1.432822909548754e-06, "loss": 0.1371, "step": 7172 }, { "epoch": 2.333441769681197, "grad_norm": 2.134089752814458, "learning_rate": 1.4314968728992306e-06, "loss": 0.1794, "step": 7173 }, { "epoch": 2.3337670787247884, "grad_norm": 2.170627617299551, "learning_rate": 1.4301713476185846e-06, "loss": 0.157, "step": 7174 }, { "epoch": 2.33409238776838, "grad_norm": 2.275278231152789, "learning_rate": 1.428846333896766e-06, "loss": 0.191, "step": 7175 }, { "epoch": 2.3344176968119714, "grad_norm": 2.2953271196497465, "learning_rate": 1.4275218319236477e-06, "loss": 0.155, "step": 7176 }, { "epoch": 2.3347430058555627, "grad_norm": 2.041589208475201, "learning_rate": 1.4261978418890342e-06, "loss": 0.1581, "step": 7177 }, { "epoch": 2.335068314899154, "grad_norm": 2.5474508173662787, "learning_rate": 1.4248743639826512e-06, "loss": 0.2172, "step": 7178 }, { "epoch": 2.3353936239427457, "grad_norm": 1.9488304772877563, "learning_rate": 1.4235513983941557e-06, "loss": 0.1565, "step": 7179 }, { "epoch": 2.335718932986337, "grad_norm": 2.508230410548805, "learning_rate": 1.4222289453131305e-06, "loss": 0.1894, "step": 7180 }, { "epoch": 2.3360442420299283, "grad_norm": 2.160669729474816, "learning_rate": 1.4209070049290812e-06, "loss": 0.1722, "step": 7181 }, { "epoch": 2.3363695510735196, "grad_norm": 2.002506298452354, "learning_rate": 1.4195855774314443e-06, "loss": 0.1399, "step": 7182 }, { "epoch": 2.3366948601171114, "grad_norm": 2.360110155505994, "learning_rate": 1.4182646630095832e-06, "loss": 0.1197, "step": 7183 }, { "epoch": 2.3370201691607027, "grad_norm": 2.280679287228364, "learning_rate": 1.4169442618527823e-06, "loss": 0.1905, "step": 7184 }, { "epoch": 2.337345478204294, "grad_norm": 1.9934807702606232, "learning_rate": 1.4156243741502579e-06, "loss": 0.1306, "step": 7185 }, { "epoch": 2.3376707872478857, "grad_norm": 2.048237327795241, "learning_rate": 1.4143050000911518e-06, "loss": 0.1459, "step": 7186 }, { "epoch": 2.337996096291477, "grad_norm": 2.4075924746019868, "learning_rate": 1.4129861398645279e-06, "loss": 0.2233, "step": 7187 }, { "epoch": 2.3383214053350683, "grad_norm": 2.1573755852189374, "learning_rate": 1.4116677936593837e-06, "loss": 0.1633, "step": 7188 }, { "epoch": 2.3386467143786596, "grad_norm": 2.022859238680536, "learning_rate": 1.4103499616646354e-06, "loss": 0.1492, "step": 7189 }, { "epoch": 2.3389720234222513, "grad_norm": 1.9187744361026744, "learning_rate": 1.4090326440691315e-06, "loss": 0.1362, "step": 7190 }, { "epoch": 2.3392973324658426, "grad_norm": 1.9772609796814276, "learning_rate": 1.4077158410616448e-06, "loss": 0.1358, "step": 7191 }, { "epoch": 2.339622641509434, "grad_norm": 2.394531077669703, "learning_rate": 1.4063995528308733e-06, "loss": 0.1439, "step": 7192 }, { "epoch": 2.339947950553025, "grad_norm": 2.068328297869881, "learning_rate": 1.4050837795654388e-06, "loss": 0.1897, "step": 7193 }, { "epoch": 2.340273259596617, "grad_norm": 2.0083588925242832, "learning_rate": 1.4037685214538988e-06, "loss": 0.1711, "step": 7194 }, { "epoch": 2.340598568640208, "grad_norm": 2.063778985314688, "learning_rate": 1.402453778684727e-06, "loss": 0.1395, "step": 7195 }, { "epoch": 2.3409238776837995, "grad_norm": 2.0917935016284885, "learning_rate": 1.401139551446326e-06, "loss": 0.1477, "step": 7196 }, { "epoch": 2.3412491867273912, "grad_norm": 2.131795788515752, "learning_rate": 1.3998258399270265e-06, "loss": 0.1349, "step": 7197 }, { "epoch": 2.3415744957709825, "grad_norm": 2.5124816068131257, "learning_rate": 1.3985126443150865e-06, "loss": 0.2064, "step": 7198 }, { "epoch": 2.341899804814574, "grad_norm": 1.9047201971413779, "learning_rate": 1.397199964798684e-06, "loss": 0.1678, "step": 7199 }, { "epoch": 2.342225113858165, "grad_norm": 2.295499758114355, "learning_rate": 1.39588780156593e-06, "loss": 0.1536, "step": 7200 }, { "epoch": 2.342550422901757, "grad_norm": 1.825193243685768, "learning_rate": 1.3945761548048559e-06, "loss": 0.153, "step": 7201 }, { "epoch": 2.342875731945348, "grad_norm": 1.9530254003929177, "learning_rate": 1.393265024703422e-06, "loss": 0.1253, "step": 7202 }, { "epoch": 2.3432010409889394, "grad_norm": 2.036221288917724, "learning_rate": 1.391954411449517e-06, "loss": 0.1334, "step": 7203 }, { "epoch": 2.3435263500325307, "grad_norm": 2.124763116147446, "learning_rate": 1.3906443152309484e-06, "loss": 0.1724, "step": 7204 }, { "epoch": 2.3438516590761225, "grad_norm": 2.1093536281126317, "learning_rate": 1.3893347362354559e-06, "loss": 0.111, "step": 7205 }, { "epoch": 2.3441769681197138, "grad_norm": 2.0196720501937335, "learning_rate": 1.3880256746507047e-06, "loss": 0.144, "step": 7206 }, { "epoch": 2.344502277163305, "grad_norm": 2.3160039513548925, "learning_rate": 1.3867171306642812e-06, "loss": 0.1678, "step": 7207 }, { "epoch": 2.344827586206897, "grad_norm": 2.15542462919147, "learning_rate": 1.3854091044637036e-06, "loss": 0.1754, "step": 7208 }, { "epoch": 2.345152895250488, "grad_norm": 2.2036950203648717, "learning_rate": 1.384101596236409e-06, "loss": 0.1253, "step": 7209 }, { "epoch": 2.3454782042940794, "grad_norm": 1.8122102776117295, "learning_rate": 1.3827946061697673e-06, "loss": 0.1458, "step": 7210 }, { "epoch": 2.3458035133376707, "grad_norm": 1.7841938069528291, "learning_rate": 1.381488134451071e-06, "loss": 0.1372, "step": 7211 }, { "epoch": 2.346128822381262, "grad_norm": 2.063087228821939, "learning_rate": 1.3801821812675387e-06, "loss": 0.1787, "step": 7212 }, { "epoch": 2.3464541314248537, "grad_norm": 1.8364413614741533, "learning_rate": 1.3788767468063108e-06, "loss": 0.1217, "step": 7213 }, { "epoch": 2.346779440468445, "grad_norm": 2.109088173460549, "learning_rate": 1.3775718312544594e-06, "loss": 0.1795, "step": 7214 }, { "epoch": 2.3471047495120363, "grad_norm": 1.946542087350828, "learning_rate": 1.376267434798982e-06, "loss": 0.1425, "step": 7215 }, { "epoch": 2.347430058555628, "grad_norm": 2.5111080007558066, "learning_rate": 1.374963557626795e-06, "loss": 0.1687, "step": 7216 }, { "epoch": 2.3477553675992193, "grad_norm": 1.8276275542488858, "learning_rate": 1.3736601999247474e-06, "loss": 0.1188, "step": 7217 }, { "epoch": 2.3480806766428106, "grad_norm": 2.4656844742558865, "learning_rate": 1.3723573618796126e-06, "loss": 0.1856, "step": 7218 }, { "epoch": 2.348405985686402, "grad_norm": 2.379122744971223, "learning_rate": 1.3710550436780846e-06, "loss": 0.2053, "step": 7219 }, { "epoch": 2.3487312947299936, "grad_norm": 1.9513490936254458, "learning_rate": 1.3697532455067903e-06, "loss": 0.1455, "step": 7220 }, { "epoch": 2.349056603773585, "grad_norm": 1.9088918392597747, "learning_rate": 1.3684519675522744e-06, "loss": 0.1311, "step": 7221 }, { "epoch": 2.3493819128171762, "grad_norm": 2.286852794633292, "learning_rate": 1.3671512100010131e-06, "loss": 0.1919, "step": 7222 }, { "epoch": 2.3497072218607675, "grad_norm": 2.3175632599096088, "learning_rate": 1.365850973039407e-06, "loss": 0.1483, "step": 7223 }, { "epoch": 2.3500325309043593, "grad_norm": 1.9386172914797217, "learning_rate": 1.3645512568537794e-06, "loss": 0.1378, "step": 7224 }, { "epoch": 2.3503578399479506, "grad_norm": 2.605414912410357, "learning_rate": 1.3632520616303774e-06, "loss": 0.2279, "step": 7225 }, { "epoch": 2.350683148991542, "grad_norm": 2.3280138280007545, "learning_rate": 1.361953387555383e-06, "loss": 0.194, "step": 7226 }, { "epoch": 2.3510084580351336, "grad_norm": 2.867009303880074, "learning_rate": 1.3606552348148937e-06, "loss": 0.2142, "step": 7227 }, { "epoch": 2.351333767078725, "grad_norm": 2.5052204651949377, "learning_rate": 1.359357603594934e-06, "loss": 0.1944, "step": 7228 }, { "epoch": 2.351659076122316, "grad_norm": 2.1660523223401174, "learning_rate": 1.3580604940814573e-06, "loss": 0.1725, "step": 7229 }, { "epoch": 2.3519843851659075, "grad_norm": 2.1468876104624313, "learning_rate": 1.356763906460341e-06, "loss": 0.1432, "step": 7230 }, { "epoch": 2.352309694209499, "grad_norm": 2.707016547708863, "learning_rate": 1.355467840917385e-06, "loss": 0.2422, "step": 7231 }, { "epoch": 2.3526350032530905, "grad_norm": 1.9554158311113463, "learning_rate": 1.3541722976383186e-06, "loss": 0.1639, "step": 7232 }, { "epoch": 2.352960312296682, "grad_norm": 1.628884739542974, "learning_rate": 1.352877276808791e-06, "loss": 0.0886, "step": 7233 }, { "epoch": 2.353285621340273, "grad_norm": 2.003172552265689, "learning_rate": 1.3515827786143821e-06, "loss": 0.1559, "step": 7234 }, { "epoch": 2.353610930383865, "grad_norm": 2.009520724452624, "learning_rate": 1.3502888032405942e-06, "loss": 0.1162, "step": 7235 }, { "epoch": 2.353936239427456, "grad_norm": 2.153335631599917, "learning_rate": 1.3489953508728526e-06, "loss": 0.1574, "step": 7236 }, { "epoch": 2.3542615484710474, "grad_norm": 2.0817536803137298, "learning_rate": 1.3477024216965118e-06, "loss": 0.1522, "step": 7237 }, { "epoch": 2.354586857514639, "grad_norm": 1.950465080993899, "learning_rate": 1.3464100158968507e-06, "loss": 0.1648, "step": 7238 }, { "epoch": 2.3549121665582304, "grad_norm": 2.00951367532617, "learning_rate": 1.3451181336590686e-06, "loss": 0.1437, "step": 7239 }, { "epoch": 2.3552374756018217, "grad_norm": 1.9192367770656675, "learning_rate": 1.3438267751682965e-06, "loss": 0.1252, "step": 7240 }, { "epoch": 2.355562784645413, "grad_norm": 2.172301118328846, "learning_rate": 1.3425359406095833e-06, "loss": 0.1873, "step": 7241 }, { "epoch": 2.3558880936890043, "grad_norm": 2.37873338391827, "learning_rate": 1.3412456301679083e-06, "loss": 0.1421, "step": 7242 }, { "epoch": 2.356213402732596, "grad_norm": 2.1895951861155063, "learning_rate": 1.3399558440281757e-06, "loss": 0.2488, "step": 7243 }, { "epoch": 2.3565387117761873, "grad_norm": 1.8700582883929582, "learning_rate": 1.338666582375211e-06, "loss": 0.1978, "step": 7244 }, { "epoch": 2.3568640208197786, "grad_norm": 2.885451042326359, "learning_rate": 1.337377845393763e-06, "loss": 0.2127, "step": 7245 }, { "epoch": 2.3571893298633704, "grad_norm": 2.2085554692393985, "learning_rate": 1.3360896332685147e-06, "loss": 0.2063, "step": 7246 }, { "epoch": 2.3575146389069617, "grad_norm": 1.8363929885539416, "learning_rate": 1.334801946184065e-06, "loss": 0.1047, "step": 7247 }, { "epoch": 2.357839947950553, "grad_norm": 2.0814893494604982, "learning_rate": 1.333514784324938e-06, "loss": 0.1547, "step": 7248 }, { "epoch": 2.3581652569941443, "grad_norm": 2.372330611532632, "learning_rate": 1.3322281478755878e-06, "loss": 0.1859, "step": 7249 }, { "epoch": 2.358490566037736, "grad_norm": 2.243269783568535, "learning_rate": 1.3309420370203902e-06, "loss": 0.1502, "step": 7250 }, { "epoch": 2.3588158750813273, "grad_norm": 1.8910000860425094, "learning_rate": 1.3296564519436439e-06, "loss": 0.1241, "step": 7251 }, { "epoch": 2.3591411841249186, "grad_norm": 2.4489717974799596, "learning_rate": 1.3283713928295767e-06, "loss": 0.1605, "step": 7252 }, { "epoch": 2.35946649316851, "grad_norm": 2.121680328610866, "learning_rate": 1.3270868598623349e-06, "loss": 0.1641, "step": 7253 }, { "epoch": 2.3597918022121016, "grad_norm": 1.7424074638425713, "learning_rate": 1.3258028532259953e-06, "loss": 0.1422, "step": 7254 }, { "epoch": 2.360117111255693, "grad_norm": 2.536258330264916, "learning_rate": 1.324519373104558e-06, "loss": 0.1792, "step": 7255 }, { "epoch": 2.360442420299284, "grad_norm": 2.144481138898033, "learning_rate": 1.3232364196819436e-06, "loss": 0.1476, "step": 7256 }, { "epoch": 2.360767729342876, "grad_norm": 2.413475727614872, "learning_rate": 1.3219539931420018e-06, "loss": 0.2092, "step": 7257 }, { "epoch": 2.3610930383864672, "grad_norm": 2.1622909334722653, "learning_rate": 1.3206720936685064e-06, "loss": 0.1945, "step": 7258 }, { "epoch": 2.3614183474300585, "grad_norm": 2.2591650793920373, "learning_rate": 1.3193907214451528e-06, "loss": 0.1761, "step": 7259 }, { "epoch": 2.36174365647365, "grad_norm": 2.203179877702702, "learning_rate": 1.3181098766555617e-06, "loss": 0.1543, "step": 7260 }, { "epoch": 2.3620689655172415, "grad_norm": 2.401897858541244, "learning_rate": 1.3168295594832797e-06, "loss": 0.179, "step": 7261 }, { "epoch": 2.362394274560833, "grad_norm": 2.208391715163099, "learning_rate": 1.3155497701117792e-06, "loss": 0.1635, "step": 7262 }, { "epoch": 2.362719583604424, "grad_norm": 2.206320411712165, "learning_rate": 1.3142705087244516e-06, "loss": 0.1435, "step": 7263 }, { "epoch": 2.3630448926480154, "grad_norm": 2.079906587102943, "learning_rate": 1.3129917755046195e-06, "loss": 0.1555, "step": 7264 }, { "epoch": 2.363370201691607, "grad_norm": 2.1847490710244992, "learning_rate": 1.3117135706355222e-06, "loss": 0.176, "step": 7265 }, { "epoch": 2.3636955107351985, "grad_norm": 2.352804093538025, "learning_rate": 1.3104358943003293e-06, "loss": 0.1454, "step": 7266 }, { "epoch": 2.3640208197787898, "grad_norm": 2.3888289777790614, "learning_rate": 1.3091587466821342e-06, "loss": 0.1642, "step": 7267 }, { "epoch": 2.3643461288223815, "grad_norm": 2.45075797235017, "learning_rate": 1.3078821279639503e-06, "loss": 0.155, "step": 7268 }, { "epoch": 2.364671437865973, "grad_norm": 1.7822238994483632, "learning_rate": 1.3066060383287189e-06, "loss": 0.1319, "step": 7269 }, { "epoch": 2.364996746909564, "grad_norm": 2.094103018516433, "learning_rate": 1.3053304779593063e-06, "loss": 0.1674, "step": 7270 }, { "epoch": 2.3653220559531554, "grad_norm": 2.04013766729533, "learning_rate": 1.304055447038498e-06, "loss": 0.135, "step": 7271 }, { "epoch": 2.3656473649967467, "grad_norm": 2.3764558887034406, "learning_rate": 1.30278094574901e-06, "loss": 0.2052, "step": 7272 }, { "epoch": 2.3659726740403384, "grad_norm": 2.1016285219674353, "learning_rate": 1.3015069742734759e-06, "loss": 0.152, "step": 7273 }, { "epoch": 2.3662979830839297, "grad_norm": 1.977434800766202, "learning_rate": 1.3002335327944582e-06, "loss": 0.1528, "step": 7274 }, { "epoch": 2.366623292127521, "grad_norm": 2.231939094498477, "learning_rate": 1.2989606214944434e-06, "loss": 0.1667, "step": 7275 }, { "epoch": 2.3669486011711127, "grad_norm": 2.3065523237440133, "learning_rate": 1.2976882405558393e-06, "loss": 0.1415, "step": 7276 }, { "epoch": 2.367273910214704, "grad_norm": 2.914141020820922, "learning_rate": 1.2964163901609756e-06, "loss": 0.1759, "step": 7277 }, { "epoch": 2.3675992192582953, "grad_norm": 2.064396866570569, "learning_rate": 1.2951450704921152e-06, "loss": 0.1648, "step": 7278 }, { "epoch": 2.3679245283018866, "grad_norm": 2.53222123795466, "learning_rate": 1.293874281731437e-06, "loss": 0.1954, "step": 7279 }, { "epoch": 2.3682498373454783, "grad_norm": 2.1499042597401976, "learning_rate": 1.2926040240610439e-06, "loss": 0.1581, "step": 7280 }, { "epoch": 2.3685751463890696, "grad_norm": 1.9280601167135887, "learning_rate": 1.291334297662965e-06, "loss": 0.1368, "step": 7281 }, { "epoch": 2.368900455432661, "grad_norm": 2.160972706513598, "learning_rate": 1.2900651027191564e-06, "loss": 0.1482, "step": 7282 }, { "epoch": 2.369225764476252, "grad_norm": 2.470276860185186, "learning_rate": 1.2887964394114898e-06, "loss": 0.2097, "step": 7283 }, { "epoch": 2.369551073519844, "grad_norm": 2.2366053284989857, "learning_rate": 1.2875283079217699e-06, "loss": 0.1898, "step": 7284 }, { "epoch": 2.3698763825634352, "grad_norm": 2.3885851694611238, "learning_rate": 1.2862607084317169e-06, "loss": 0.1968, "step": 7285 }, { "epoch": 2.3702016916070265, "grad_norm": 2.12722068660621, "learning_rate": 1.2849936411229802e-06, "loss": 0.1701, "step": 7286 }, { "epoch": 2.3705270006506183, "grad_norm": 2.172080250061427, "learning_rate": 1.2837271061771334e-06, "loss": 0.1607, "step": 7287 }, { "epoch": 2.3708523096942096, "grad_norm": 2.0458750306482134, "learning_rate": 1.2824611037756686e-06, "loss": 0.1485, "step": 7288 }, { "epoch": 2.371177618737801, "grad_norm": 1.9934262134357075, "learning_rate": 1.2811956341000064e-06, "loss": 0.1878, "step": 7289 }, { "epoch": 2.371502927781392, "grad_norm": 1.8410781096909, "learning_rate": 1.27993069733149e-06, "loss": 0.1622, "step": 7290 }, { "epoch": 2.371828236824984, "grad_norm": 2.384165029659318, "learning_rate": 1.2786662936513844e-06, "loss": 0.2043, "step": 7291 }, { "epoch": 2.372153545868575, "grad_norm": 2.4890788739497274, "learning_rate": 1.277402423240881e-06, "loss": 0.2104, "step": 7292 }, { "epoch": 2.3724788549121665, "grad_norm": 2.211320168452419, "learning_rate": 1.2761390862810907e-06, "loss": 0.1734, "step": 7293 }, { "epoch": 2.3728041639557578, "grad_norm": 1.9772838960014556, "learning_rate": 1.2748762829530525e-06, "loss": 0.1603, "step": 7294 }, { "epoch": 2.3731294729993495, "grad_norm": 1.7363120883435577, "learning_rate": 1.2736140134377283e-06, "loss": 0.1032, "step": 7295 }, { "epoch": 2.373454782042941, "grad_norm": 2.2136097420887766, "learning_rate": 1.2723522779160002e-06, "loss": 0.1591, "step": 7296 }, { "epoch": 2.373780091086532, "grad_norm": 2.141135918788523, "learning_rate": 1.271091076568673e-06, "loss": 0.1833, "step": 7297 }, { "epoch": 2.374105400130124, "grad_norm": 2.1766095389506415, "learning_rate": 1.2698304095764834e-06, "loss": 0.2176, "step": 7298 }, { "epoch": 2.374430709173715, "grad_norm": 2.0936096347040754, "learning_rate": 1.2685702771200835e-06, "loss": 0.1801, "step": 7299 }, { "epoch": 2.3747560182173064, "grad_norm": 2.1772663305868725, "learning_rate": 1.2673106793800489e-06, "loss": 0.1906, "step": 7300 }, { "epoch": 2.3750813272608977, "grad_norm": 2.324974449082723, "learning_rate": 1.266051616536883e-06, "loss": 0.1806, "step": 7301 }, { "epoch": 2.3754066363044894, "grad_norm": 1.8508822605047832, "learning_rate": 1.2647930887710114e-06, "loss": 0.1398, "step": 7302 }, { "epoch": 2.3757319453480807, "grad_norm": 1.8984732394201775, "learning_rate": 1.263535096262779e-06, "loss": 0.1542, "step": 7303 }, { "epoch": 2.376057254391672, "grad_norm": 2.223747725369333, "learning_rate": 1.2622776391924601e-06, "loss": 0.1749, "step": 7304 }, { "epoch": 2.3763825634352633, "grad_norm": 2.202585941346691, "learning_rate": 1.2610207177402462e-06, "loss": 0.1339, "step": 7305 }, { "epoch": 2.376707872478855, "grad_norm": 2.429826850500537, "learning_rate": 1.2597643320862557e-06, "loss": 0.1514, "step": 7306 }, { "epoch": 2.3770331815224464, "grad_norm": 2.258875658587564, "learning_rate": 1.258508482410532e-06, "loss": 0.1786, "step": 7307 }, { "epoch": 2.3773584905660377, "grad_norm": 1.9506314254645019, "learning_rate": 1.2572531688930361e-06, "loss": 0.2074, "step": 7308 }, { "epoch": 2.3776837996096294, "grad_norm": 2.1207417417169614, "learning_rate": 1.2559983917136565e-06, "loss": 0.169, "step": 7309 }, { "epoch": 2.3780091086532207, "grad_norm": 2.3501617955517817, "learning_rate": 1.2547441510522045e-06, "loss": 0.1721, "step": 7310 }, { "epoch": 2.378334417696812, "grad_norm": 2.479427135285777, "learning_rate": 1.253490447088413e-06, "loss": 0.1989, "step": 7311 }, { "epoch": 2.3786597267404033, "grad_norm": 1.918702741435637, "learning_rate": 1.2522372800019357e-06, "loss": 0.1991, "step": 7312 }, { "epoch": 2.3789850357839946, "grad_norm": 2.103920295270736, "learning_rate": 1.2509846499723577e-06, "loss": 0.1389, "step": 7313 }, { "epoch": 2.3793103448275863, "grad_norm": 2.298736546295991, "learning_rate": 1.2497325571791791e-06, "loss": 0.1991, "step": 7314 }, { "epoch": 2.3796356538711776, "grad_norm": 2.1331833586654096, "learning_rate": 1.2484810018018234e-06, "loss": 0.2022, "step": 7315 }, { "epoch": 2.379960962914769, "grad_norm": 1.742770948053193, "learning_rate": 1.2472299840196417e-06, "loss": 0.1092, "step": 7316 }, { "epoch": 2.3802862719583606, "grad_norm": 2.5103453610114994, "learning_rate": 1.2459795040119072e-06, "loss": 0.1901, "step": 7317 }, { "epoch": 2.380611581001952, "grad_norm": 2.1741093812881425, "learning_rate": 1.244729561957811e-06, "loss": 0.1424, "step": 7318 }, { "epoch": 2.380936890045543, "grad_norm": 2.2018139036133944, "learning_rate": 1.2434801580364735e-06, "loss": 0.229, "step": 7319 }, { "epoch": 2.3812621990891345, "grad_norm": 2.2082746914718197, "learning_rate": 1.2422312924269325e-06, "loss": 0.1573, "step": 7320 }, { "epoch": 2.3815875081327262, "grad_norm": 2.0014649907631243, "learning_rate": 1.2409829653081533e-06, "loss": 0.1521, "step": 7321 }, { "epoch": 2.3819128171763175, "grad_norm": 1.934780550390055, "learning_rate": 1.2397351768590226e-06, "loss": 0.1367, "step": 7322 }, { "epoch": 2.382238126219909, "grad_norm": 1.9071207875285952, "learning_rate": 1.2384879272583468e-06, "loss": 0.1628, "step": 7323 }, { "epoch": 2.3825634352635, "grad_norm": 1.9364569923611232, "learning_rate": 1.2372412166848585e-06, "loss": 0.1266, "step": 7324 }, { "epoch": 2.382888744307092, "grad_norm": 2.1989031489237765, "learning_rate": 1.2359950453172143e-06, "loss": 0.1893, "step": 7325 }, { "epoch": 2.383214053350683, "grad_norm": 2.012805432770116, "learning_rate": 1.2347494133339876e-06, "loss": 0.1465, "step": 7326 }, { "epoch": 2.3835393623942744, "grad_norm": 2.3638144790340623, "learning_rate": 1.2335043209136827e-06, "loss": 0.1909, "step": 7327 }, { "epoch": 2.383864671437866, "grad_norm": 1.8592702786884494, "learning_rate": 1.2322597682347181e-06, "loss": 0.1288, "step": 7328 }, { "epoch": 2.3841899804814575, "grad_norm": 2.4121714191190287, "learning_rate": 1.2310157554754409e-06, "loss": 0.1523, "step": 7329 }, { "epoch": 2.3845152895250488, "grad_norm": 2.205414558111629, "learning_rate": 1.22977228281412e-06, "loss": 0.1316, "step": 7330 }, { "epoch": 2.38484059856864, "grad_norm": 2.2318057362382486, "learning_rate": 1.2285293504289448e-06, "loss": 0.1555, "step": 7331 }, { "epoch": 2.385165907612232, "grad_norm": 1.9493250731873553, "learning_rate": 1.2272869584980257e-06, "loss": 0.1531, "step": 7332 }, { "epoch": 2.385491216655823, "grad_norm": 2.4347248459972466, "learning_rate": 1.2260451071994034e-06, "loss": 0.1837, "step": 7333 }, { "epoch": 2.3858165256994144, "grad_norm": 1.9752045766831403, "learning_rate": 1.2248037967110332e-06, "loss": 0.1179, "step": 7334 }, { "epoch": 2.3861418347430057, "grad_norm": 2.244468747228285, "learning_rate": 1.2235630272107952e-06, "loss": 0.1614, "step": 7335 }, { "epoch": 2.3864671437865974, "grad_norm": 2.004055690188518, "learning_rate": 1.2223227988764924e-06, "loss": 0.1712, "step": 7336 }, { "epoch": 2.3867924528301887, "grad_norm": 2.273444560944717, "learning_rate": 1.221083111885853e-06, "loss": 0.1744, "step": 7337 }, { "epoch": 2.38711776187378, "grad_norm": 1.9922491551064698, "learning_rate": 1.2198439664165219e-06, "loss": 0.1832, "step": 7338 }, { "epoch": 2.3874430709173717, "grad_norm": 2.252189885600813, "learning_rate": 1.218605362646072e-06, "loss": 0.1978, "step": 7339 }, { "epoch": 2.387768379960963, "grad_norm": 2.2706989370183934, "learning_rate": 1.2173673007519937e-06, "loss": 0.16, "step": 7340 }, { "epoch": 2.3880936890045543, "grad_norm": 2.1739998856511287, "learning_rate": 1.2161297809117028e-06, "loss": 0.1661, "step": 7341 }, { "epoch": 2.3884189980481456, "grad_norm": 2.1816853968974605, "learning_rate": 1.2148928033025393e-06, "loss": 0.1511, "step": 7342 }, { "epoch": 2.388744307091737, "grad_norm": 2.0403491431404497, "learning_rate": 1.2136563681017583e-06, "loss": 0.149, "step": 7343 }, { "epoch": 2.3890696161353286, "grad_norm": 2.3863202864910975, "learning_rate": 1.212420475486545e-06, "loss": 0.1813, "step": 7344 }, { "epoch": 2.38939492517892, "grad_norm": 1.846012612677589, "learning_rate": 1.2111851256340034e-06, "loss": 0.1437, "step": 7345 }, { "epoch": 2.3897202342225112, "grad_norm": 2.104404581271797, "learning_rate": 1.2099503187211603e-06, "loss": 0.1939, "step": 7346 }, { "epoch": 2.390045543266103, "grad_norm": 2.618921199766235, "learning_rate": 1.2087160549249616e-06, "loss": 0.2181, "step": 7347 }, { "epoch": 2.3903708523096943, "grad_norm": 2.276490693733249, "learning_rate": 1.2074823344222802e-06, "loss": 0.1932, "step": 7348 }, { "epoch": 2.3906961613532856, "grad_norm": 2.4373268184163366, "learning_rate": 1.20624915738991e-06, "loss": 0.1761, "step": 7349 }, { "epoch": 2.391021470396877, "grad_norm": 2.2094878105446054, "learning_rate": 1.2050165240045635e-06, "loss": 0.1983, "step": 7350 }, { "epoch": 2.3913467794404686, "grad_norm": 1.663223718682642, "learning_rate": 1.2037844344428812e-06, "loss": 0.1378, "step": 7351 }, { "epoch": 2.39167208848406, "grad_norm": 2.2135872779965564, "learning_rate": 1.2025528888814186e-06, "loss": 0.1718, "step": 7352 }, { "epoch": 2.391997397527651, "grad_norm": 2.0938550508776066, "learning_rate": 1.2013218874966592e-06, "loss": 0.1363, "step": 7353 }, { "epoch": 2.3923227065712425, "grad_norm": 2.184217422363047, "learning_rate": 1.2000914304650073e-06, "loss": 0.1714, "step": 7354 }, { "epoch": 2.392648015614834, "grad_norm": 2.126320557218107, "learning_rate": 1.1988615179627854e-06, "loss": 0.1315, "step": 7355 }, { "epoch": 2.3929733246584255, "grad_norm": 2.0320424203473872, "learning_rate": 1.1976321501662424e-06, "loss": 0.1509, "step": 7356 }, { "epoch": 2.393298633702017, "grad_norm": 1.9456460699622404, "learning_rate": 1.196403327251549e-06, "loss": 0.1381, "step": 7357 }, { "epoch": 2.3936239427456085, "grad_norm": 2.663187973290885, "learning_rate": 1.1951750493947934e-06, "loss": 0.1814, "step": 7358 }, { "epoch": 2.3939492517892, "grad_norm": 2.202197014417907, "learning_rate": 1.1939473167719911e-06, "loss": 0.1403, "step": 7359 }, { "epoch": 2.394274560832791, "grad_norm": 2.3388424573294713, "learning_rate": 1.192720129559075e-06, "loss": 0.1848, "step": 7360 }, { "epoch": 2.3945998698763824, "grad_norm": 2.098081124621249, "learning_rate": 1.1914934879319024e-06, "loss": 0.1705, "step": 7361 }, { "epoch": 2.394925178919974, "grad_norm": 2.2269712316926333, "learning_rate": 1.1902673920662538e-06, "loss": 0.1493, "step": 7362 }, { "epoch": 2.3952504879635654, "grad_norm": 2.208441191323644, "learning_rate": 1.189041842137828e-06, "loss": 0.1493, "step": 7363 }, { "epoch": 2.3955757970071567, "grad_norm": 1.91512047972787, "learning_rate": 1.187816838322244e-06, "loss": 0.1762, "step": 7364 }, { "epoch": 2.395901106050748, "grad_norm": 2.0293786969533376, "learning_rate": 1.186592380795052e-06, "loss": 0.1538, "step": 7365 }, { "epoch": 2.3962264150943398, "grad_norm": 2.1278157109581866, "learning_rate": 1.1853684697317136e-06, "loss": 0.16, "step": 7366 }, { "epoch": 2.396551724137931, "grad_norm": 2.0005348902510764, "learning_rate": 1.1841451053076157e-06, "loss": 0.1357, "step": 7367 }, { "epoch": 2.3968770331815223, "grad_norm": 1.9820191427627918, "learning_rate": 1.1829222876980689e-06, "loss": 0.1783, "step": 7368 }, { "epoch": 2.397202342225114, "grad_norm": 2.1302368658688153, "learning_rate": 1.1817000170783044e-06, "loss": 0.1844, "step": 7369 }, { "epoch": 2.3975276512687054, "grad_norm": 2.7320856007978973, "learning_rate": 1.1804782936234716e-06, "loss": 0.1605, "step": 7370 }, { "epoch": 2.3978529603122967, "grad_norm": 2.250651219952, "learning_rate": 1.179257117508647e-06, "loss": 0.1553, "step": 7371 }, { "epoch": 2.398178269355888, "grad_norm": 2.0089017906000266, "learning_rate": 1.1780364889088242e-06, "loss": 0.1711, "step": 7372 }, { "epoch": 2.3985035783994797, "grad_norm": 2.019799104608031, "learning_rate": 1.1768164079989204e-06, "loss": 0.1322, "step": 7373 }, { "epoch": 2.398828887443071, "grad_norm": 1.8040979387021052, "learning_rate": 1.1755968749537755e-06, "loss": 0.1244, "step": 7374 }, { "epoch": 2.3991541964866623, "grad_norm": 2.326045706415214, "learning_rate": 1.1743778899481474e-06, "loss": 0.2005, "step": 7375 }, { "epoch": 2.3994795055302536, "grad_norm": 1.9497676931857606, "learning_rate": 1.173159453156718e-06, "loss": 0.1585, "step": 7376 }, { "epoch": 2.3998048145738453, "grad_norm": 2.249487245240734, "learning_rate": 1.1719415647540922e-06, "loss": 0.1694, "step": 7377 }, { "epoch": 2.4001301236174366, "grad_norm": 2.2634379260496362, "learning_rate": 1.1707242249147904e-06, "loss": 0.1998, "step": 7378 }, { "epoch": 2.400455432661028, "grad_norm": 2.0201041563526165, "learning_rate": 1.1695074338132623e-06, "loss": 0.1488, "step": 7379 }, { "epoch": 2.4007807417046196, "grad_norm": 2.0922945584191055, "learning_rate": 1.168291191623871e-06, "loss": 0.1681, "step": 7380 }, { "epoch": 2.401106050748211, "grad_norm": 2.0648688619953712, "learning_rate": 1.1670754985209065e-06, "loss": 0.1489, "step": 7381 }, { "epoch": 2.4014313597918022, "grad_norm": 2.16574058121095, "learning_rate": 1.1658603546785806e-06, "loss": 0.1579, "step": 7382 }, { "epoch": 2.4017566688353935, "grad_norm": 2.040827946054032, "learning_rate": 1.164645760271022e-06, "loss": 0.2059, "step": 7383 }, { "epoch": 2.402081977878985, "grad_norm": 2.0567696458117433, "learning_rate": 1.1634317154722803e-06, "loss": 0.1585, "step": 7384 }, { "epoch": 2.4024072869225765, "grad_norm": 2.484324126651779, "learning_rate": 1.1622182204563342e-06, "loss": 0.2016, "step": 7385 }, { "epoch": 2.402732595966168, "grad_norm": 2.289725477177339, "learning_rate": 1.1610052753970763e-06, "loss": 0.1657, "step": 7386 }, { "epoch": 2.403057905009759, "grad_norm": 1.739135983499831, "learning_rate": 1.15979288046832e-06, "loss": 0.1203, "step": 7387 }, { "epoch": 2.403383214053351, "grad_norm": 1.8255282736251994, "learning_rate": 1.1585810358438048e-06, "loss": 0.1256, "step": 7388 }, { "epoch": 2.403708523096942, "grad_norm": 1.8989987347906911, "learning_rate": 1.1573697416971896e-06, "loss": 0.149, "step": 7389 }, { "epoch": 2.4040338321405335, "grad_norm": 2.3844942458354277, "learning_rate": 1.156158998202051e-06, "loss": 0.1669, "step": 7390 }, { "epoch": 2.4043591411841247, "grad_norm": 2.118994554155352, "learning_rate": 1.1549488055318925e-06, "loss": 0.1554, "step": 7391 }, { "epoch": 2.4046844502277165, "grad_norm": 2.265796466537307, "learning_rate": 1.153739163860132e-06, "loss": 0.2191, "step": 7392 }, { "epoch": 2.405009759271308, "grad_norm": 1.711093819915555, "learning_rate": 1.1525300733601136e-06, "loss": 0.0934, "step": 7393 }, { "epoch": 2.405335068314899, "grad_norm": 2.2764220475040773, "learning_rate": 1.1513215342051025e-06, "loss": 0.2251, "step": 7394 }, { "epoch": 2.4056603773584904, "grad_norm": 2.128855614163287, "learning_rate": 1.1501135465682816e-06, "loss": 0.2205, "step": 7395 }, { "epoch": 2.405985686402082, "grad_norm": 1.9952399634074793, "learning_rate": 1.148906110622754e-06, "loss": 0.151, "step": 7396 }, { "epoch": 2.4063109954456734, "grad_norm": 1.722216483793254, "learning_rate": 1.1476992265415505e-06, "loss": 0.1175, "step": 7397 }, { "epoch": 2.4066363044892647, "grad_norm": 2.0877127252877905, "learning_rate": 1.1464928944976166e-06, "loss": 0.1741, "step": 7398 }, { "epoch": 2.4069616135328564, "grad_norm": 2.0100886126136475, "learning_rate": 1.1452871146638184e-06, "loss": 0.15, "step": 7399 }, { "epoch": 2.4072869225764477, "grad_norm": 2.1836387447711196, "learning_rate": 1.1440818872129473e-06, "loss": 0.1491, "step": 7400 }, { "epoch": 2.407612231620039, "grad_norm": 1.9005635855807372, "learning_rate": 1.1428772123177134e-06, "loss": 0.1715, "step": 7401 }, { "epoch": 2.4079375406636303, "grad_norm": 2.0656503024586623, "learning_rate": 1.1416730901507456e-06, "loss": 0.2022, "step": 7402 }, { "epoch": 2.408262849707222, "grad_norm": 2.248235686902984, "learning_rate": 1.140469520884598e-06, "loss": 0.1472, "step": 7403 }, { "epoch": 2.4085881587508133, "grad_norm": 1.9713525791421929, "learning_rate": 1.1392665046917395e-06, "loss": 0.1268, "step": 7404 }, { "epoch": 2.4089134677944046, "grad_norm": 1.9928830274521923, "learning_rate": 1.1380640417445653e-06, "loss": 0.1504, "step": 7405 }, { "epoch": 2.409238776837996, "grad_norm": 1.7848936980653816, "learning_rate": 1.13686213221539e-06, "loss": 0.1342, "step": 7406 }, { "epoch": 2.4095640858815877, "grad_norm": 1.6184025398907145, "learning_rate": 1.1356607762764455e-06, "loss": 0.1336, "step": 7407 }, { "epoch": 2.409889394925179, "grad_norm": 1.9151617099666531, "learning_rate": 1.134459974099888e-06, "loss": 0.1373, "step": 7408 }, { "epoch": 2.4102147039687702, "grad_norm": 2.0196606201700456, "learning_rate": 1.1332597258577949e-06, "loss": 0.1212, "step": 7409 }, { "epoch": 2.410540013012362, "grad_norm": 1.9612879347293781, "learning_rate": 1.1320600317221603e-06, "loss": 0.1228, "step": 7410 }, { "epoch": 2.4108653220559533, "grad_norm": 2.3666717163814543, "learning_rate": 1.1308608918649028e-06, "loss": 0.1964, "step": 7411 }, { "epoch": 2.4111906310995446, "grad_norm": 2.438207047466255, "learning_rate": 1.1296623064578582e-06, "loss": 0.2002, "step": 7412 }, { "epoch": 2.411515940143136, "grad_norm": 2.2802273882440818, "learning_rate": 1.1284642756727853e-06, "loss": 0.1572, "step": 7413 }, { "epoch": 2.411841249186727, "grad_norm": 1.9341041829040138, "learning_rate": 1.127266799681364e-06, "loss": 0.1372, "step": 7414 }, { "epoch": 2.412166558230319, "grad_norm": 2.3785975393968153, "learning_rate": 1.1260698786551932e-06, "loss": 0.2209, "step": 7415 }, { "epoch": 2.41249186727391, "grad_norm": 2.0832908310580898, "learning_rate": 1.1248735127657884e-06, "loss": 0.1543, "step": 7416 }, { "epoch": 2.4128171763175015, "grad_norm": 1.872826450148111, "learning_rate": 1.1236777021845957e-06, "loss": 0.1399, "step": 7417 }, { "epoch": 2.413142485361093, "grad_norm": 2.262037734976551, "learning_rate": 1.1224824470829726e-06, "loss": 0.1511, "step": 7418 }, { "epoch": 2.4134677944046845, "grad_norm": 1.5934528077627828, "learning_rate": 1.1212877476321988e-06, "loss": 0.0974, "step": 7419 }, { "epoch": 2.413793103448276, "grad_norm": 2.1230421196013323, "learning_rate": 1.120093604003477e-06, "loss": 0.1551, "step": 7420 }, { "epoch": 2.414118412491867, "grad_norm": 2.034191348465686, "learning_rate": 1.1189000163679298e-06, "loss": 0.1893, "step": 7421 }, { "epoch": 2.414443721535459, "grad_norm": 2.074763784620178, "learning_rate": 1.1177069848965965e-06, "loss": 0.1347, "step": 7422 }, { "epoch": 2.41476903057905, "grad_norm": 1.8507422806517546, "learning_rate": 1.1165145097604425e-06, "loss": 0.1152, "step": 7423 }, { "epoch": 2.4150943396226414, "grad_norm": 2.224092121369031, "learning_rate": 1.1153225911303468e-06, "loss": 0.1662, "step": 7424 }, { "epoch": 2.4154196486662327, "grad_norm": 2.111193271446848, "learning_rate": 1.1141312291771133e-06, "loss": 0.1685, "step": 7425 }, { "epoch": 2.4157449577098244, "grad_norm": 1.8978857746673465, "learning_rate": 1.1129404240714674e-06, "loss": 0.1219, "step": 7426 }, { "epoch": 2.4160702667534157, "grad_norm": 2.6032970651433485, "learning_rate": 1.1117501759840489e-06, "loss": 0.1769, "step": 7427 }, { "epoch": 2.416395575797007, "grad_norm": 2.303449890261258, "learning_rate": 1.1105604850854223e-06, "loss": 0.155, "step": 7428 }, { "epoch": 2.4167208848405988, "grad_norm": 2.3650243802285082, "learning_rate": 1.1093713515460735e-06, "loss": 0.1498, "step": 7429 }, { "epoch": 2.41704619388419, "grad_norm": 2.229378563987307, "learning_rate": 1.1081827755364032e-06, "loss": 0.162, "step": 7430 }, { "epoch": 2.4173715029277814, "grad_norm": 2.1242523077525224, "learning_rate": 1.1069947572267354e-06, "loss": 0.1426, "step": 7431 }, { "epoch": 2.4176968119713727, "grad_norm": 2.1592124962454036, "learning_rate": 1.105807296787314e-06, "loss": 0.1952, "step": 7432 }, { "epoch": 2.4180221210149644, "grad_norm": 2.352378859388984, "learning_rate": 1.1046203943883043e-06, "loss": 0.1639, "step": 7433 }, { "epoch": 2.4183474300585557, "grad_norm": 2.781235552719833, "learning_rate": 1.1034340501997888e-06, "loss": 0.2177, "step": 7434 }, { "epoch": 2.418672739102147, "grad_norm": 2.889441083223554, "learning_rate": 1.1022482643917726e-06, "loss": 0.2103, "step": 7435 }, { "epoch": 2.4189980481457383, "grad_norm": 2.2648189047534153, "learning_rate": 1.1010630371341774e-06, "loss": 0.202, "step": 7436 }, { "epoch": 2.41932335718933, "grad_norm": 2.1685692041850415, "learning_rate": 1.099878368596849e-06, "loss": 0.178, "step": 7437 }, { "epoch": 2.4196486662329213, "grad_norm": 2.0030279787067777, "learning_rate": 1.098694258949552e-06, "loss": 0.1284, "step": 7438 }, { "epoch": 2.4199739752765126, "grad_norm": 2.4163239323898487, "learning_rate": 1.0975107083619668e-06, "loss": 0.1781, "step": 7439 }, { "epoch": 2.4202992843201043, "grad_norm": 2.037200964525797, "learning_rate": 1.0963277170036995e-06, "loss": 0.1444, "step": 7440 }, { "epoch": 2.4206245933636956, "grad_norm": 2.119061626718135, "learning_rate": 1.0951452850442746e-06, "loss": 0.1626, "step": 7441 }, { "epoch": 2.420949902407287, "grad_norm": 1.9516821396934716, "learning_rate": 1.0939634126531324e-06, "loss": 0.139, "step": 7442 }, { "epoch": 2.421275211450878, "grad_norm": 1.9821437896638625, "learning_rate": 1.0927820999996369e-06, "loss": 0.1348, "step": 7443 }, { "epoch": 2.4216005204944695, "grad_norm": 1.9767237264966828, "learning_rate": 1.0916013472530735e-06, "loss": 0.1465, "step": 7444 }, { "epoch": 2.4219258295380612, "grad_norm": 2.0133770115896046, "learning_rate": 1.0904211545826416e-06, "loss": 0.1466, "step": 7445 }, { "epoch": 2.4222511385816525, "grad_norm": 1.8811847341993475, "learning_rate": 1.089241522157466e-06, "loss": 0.1306, "step": 7446 }, { "epoch": 2.422576447625244, "grad_norm": 2.5846249881500536, "learning_rate": 1.0880624501465858e-06, "loss": 0.2408, "step": 7447 }, { "epoch": 2.4229017566688356, "grad_norm": 2.326982527954904, "learning_rate": 1.0868839387189655e-06, "loss": 0.1847, "step": 7448 }, { "epoch": 2.423227065712427, "grad_norm": 2.3291717325904573, "learning_rate": 1.0857059880434866e-06, "loss": 0.1534, "step": 7449 }, { "epoch": 2.423552374756018, "grad_norm": 2.3319128556929356, "learning_rate": 1.0845285982889498e-06, "loss": 0.1845, "step": 7450 }, { "epoch": 2.4238776837996094, "grad_norm": 2.4860355646483443, "learning_rate": 1.0833517696240724e-06, "loss": 0.1843, "step": 7451 }, { "epoch": 2.424202992843201, "grad_norm": 2.2213708182534146, "learning_rate": 1.0821755022175008e-06, "loss": 0.1516, "step": 7452 }, { "epoch": 2.4245283018867925, "grad_norm": 2.0513312197476825, "learning_rate": 1.0809997962377916e-06, "loss": 0.1542, "step": 7453 }, { "epoch": 2.4248536109303838, "grad_norm": 2.071392924131864, "learning_rate": 1.0798246518534227e-06, "loss": 0.1723, "step": 7454 }, { "epoch": 2.425178919973975, "grad_norm": 2.166811262982092, "learning_rate": 1.0786500692327955e-06, "loss": 0.1836, "step": 7455 }, { "epoch": 2.425504229017567, "grad_norm": 1.4658648031185824, "learning_rate": 1.0774760485442293e-06, "loss": 0.1104, "step": 7456 }, { "epoch": 2.425829538061158, "grad_norm": 2.072129221525896, "learning_rate": 1.0763025899559588e-06, "loss": 0.2017, "step": 7457 }, { "epoch": 2.4261548471047494, "grad_norm": 2.1230738596732825, "learning_rate": 1.075129693636145e-06, "loss": 0.1769, "step": 7458 }, { "epoch": 2.426480156148341, "grad_norm": 2.01412219232222, "learning_rate": 1.0739573597528613e-06, "loss": 0.1846, "step": 7459 }, { "epoch": 2.4268054651919324, "grad_norm": 1.8043787585859083, "learning_rate": 1.0727855884741057e-06, "loss": 0.1369, "step": 7460 }, { "epoch": 2.4271307742355237, "grad_norm": 2.603738415078829, "learning_rate": 1.0716143799677946e-06, "loss": 0.1763, "step": 7461 }, { "epoch": 2.427456083279115, "grad_norm": 2.4264088844241494, "learning_rate": 1.0704437344017615e-06, "loss": 0.1627, "step": 7462 }, { "epoch": 2.4277813923227067, "grad_norm": 1.8470340734199535, "learning_rate": 1.0692736519437608e-06, "loss": 0.1544, "step": 7463 }, { "epoch": 2.428106701366298, "grad_norm": 2.5109870598619466, "learning_rate": 1.0681041327614677e-06, "loss": 0.1657, "step": 7464 }, { "epoch": 2.4284320104098893, "grad_norm": 2.1938635423872612, "learning_rate": 1.0669351770224723e-06, "loss": 0.1718, "step": 7465 }, { "epoch": 2.4287573194534806, "grad_norm": 2.3605163004766787, "learning_rate": 1.0657667848942904e-06, "loss": 0.1765, "step": 7466 }, { "epoch": 2.4290826284970723, "grad_norm": 1.9423686988142892, "learning_rate": 1.0645989565443493e-06, "loss": 0.145, "step": 7467 }, { "epoch": 2.4294079375406636, "grad_norm": 2.3072531306240616, "learning_rate": 1.0634316921400017e-06, "loss": 0.1483, "step": 7468 }, { "epoch": 2.429733246584255, "grad_norm": 2.192648701360994, "learning_rate": 1.062264991848519e-06, "loss": 0.1778, "step": 7469 }, { "epoch": 2.4300585556278467, "grad_norm": 2.1278764480819667, "learning_rate": 1.0610988558370878e-06, "loss": 0.1868, "step": 7470 }, { "epoch": 2.430383864671438, "grad_norm": 2.335983449017227, "learning_rate": 1.0599332842728143e-06, "loss": 0.167, "step": 7471 }, { "epoch": 2.4307091737150293, "grad_norm": 1.9488943947947064, "learning_rate": 1.058768277322731e-06, "loss": 0.135, "step": 7472 }, { "epoch": 2.4310344827586206, "grad_norm": 1.9900192681348272, "learning_rate": 1.057603835153781e-06, "loss": 0.1827, "step": 7473 }, { "epoch": 2.4313597918022123, "grad_norm": 2.2532675993868567, "learning_rate": 1.0564399579328289e-06, "loss": 0.1942, "step": 7474 }, { "epoch": 2.4316851008458036, "grad_norm": 2.0211092271993345, "learning_rate": 1.05527664582666e-06, "loss": 0.2017, "step": 7475 }, { "epoch": 2.432010409889395, "grad_norm": 1.8593627134150645, "learning_rate": 1.054113899001979e-06, "loss": 0.1689, "step": 7476 }, { "epoch": 2.432335718932986, "grad_norm": 2.076747847055224, "learning_rate": 1.0529517176254055e-06, "loss": 0.1812, "step": 7477 }, { "epoch": 2.432661027976578, "grad_norm": 2.28114248831065, "learning_rate": 1.0517901018634847e-06, "loss": 0.19, "step": 7478 }, { "epoch": 2.432986337020169, "grad_norm": 1.9683338364471346, "learning_rate": 1.050629051882673e-06, "loss": 0.156, "step": 7479 }, { "epoch": 2.4333116460637605, "grad_norm": 2.236600145446325, "learning_rate": 1.049468567849351e-06, "loss": 0.1766, "step": 7480 }, { "epoch": 2.4336369551073522, "grad_norm": 1.9362999392815683, "learning_rate": 1.0483086499298185e-06, "loss": 0.1564, "step": 7481 }, { "epoch": 2.4339622641509435, "grad_norm": 2.099446432930369, "learning_rate": 1.0471492982902915e-06, "loss": 0.1691, "step": 7482 }, { "epoch": 2.434287573194535, "grad_norm": 2.1459518508522493, "learning_rate": 1.0459905130969027e-06, "loss": 0.1464, "step": 7483 }, { "epoch": 2.434612882238126, "grad_norm": 2.0451816591070147, "learning_rate": 1.0448322945157124e-06, "loss": 0.1403, "step": 7484 }, { "epoch": 2.4349381912817174, "grad_norm": 1.9913974953699676, "learning_rate": 1.043674642712691e-06, "loss": 0.1439, "step": 7485 }, { "epoch": 2.435263500325309, "grad_norm": 2.1748697171838, "learning_rate": 1.04251755785373e-06, "loss": 0.1551, "step": 7486 }, { "epoch": 2.4355888093689004, "grad_norm": 2.3299279054619433, "learning_rate": 1.0413610401046414e-06, "loss": 0.1708, "step": 7487 }, { "epoch": 2.4359141184124917, "grad_norm": 2.006449990996429, "learning_rate": 1.0402050896311566e-06, "loss": 0.175, "step": 7488 }, { "epoch": 2.4362394274560835, "grad_norm": 2.234577514280101, "learning_rate": 1.0390497065989213e-06, "loss": 0.173, "step": 7489 }, { "epoch": 2.4365647364996748, "grad_norm": 1.9286977949799557, "learning_rate": 1.0378948911735054e-06, "loss": 0.1519, "step": 7490 }, { "epoch": 2.436890045543266, "grad_norm": 1.8923213476166043, "learning_rate": 1.0367406435203913e-06, "loss": 0.1409, "step": 7491 }, { "epoch": 2.4372153545868573, "grad_norm": 2.1513096110323304, "learning_rate": 1.035586963804986e-06, "loss": 0.2029, "step": 7492 }, { "epoch": 2.437540663630449, "grad_norm": 2.086411291557411, "learning_rate": 1.0344338521926127e-06, "loss": 0.1348, "step": 7493 }, { "epoch": 2.4378659726740404, "grad_norm": 2.1552164988737745, "learning_rate": 1.0332813088485115e-06, "loss": 0.1485, "step": 7494 }, { "epoch": 2.4381912817176317, "grad_norm": 2.3892436983175425, "learning_rate": 1.0321293339378436e-06, "loss": 0.1489, "step": 7495 }, { "epoch": 2.438516590761223, "grad_norm": 2.400330851845699, "learning_rate": 1.0309779276256887e-06, "loss": 0.1662, "step": 7496 }, { "epoch": 2.4388418998048147, "grad_norm": 2.064566742076942, "learning_rate": 1.0298270900770418e-06, "loss": 0.1855, "step": 7497 }, { "epoch": 2.439167208848406, "grad_norm": 2.091812176487481, "learning_rate": 1.028676821456821e-06, "loss": 0.1844, "step": 7498 }, { "epoch": 2.4394925178919973, "grad_norm": 2.0270045454156547, "learning_rate": 1.0275271219298583e-06, "loss": 0.135, "step": 7499 }, { "epoch": 2.439817826935589, "grad_norm": 1.9711609272957344, "learning_rate": 1.026377991660908e-06, "loss": 0.1687, "step": 7500 }, { "epoch": 2.4401431359791803, "grad_norm": 2.078136007243635, "learning_rate": 1.0252294308146415e-06, "loss": 0.1832, "step": 7501 }, { "epoch": 2.4404684450227716, "grad_norm": 2.4777631663396407, "learning_rate": 1.024081439555648e-06, "loss": 0.1883, "step": 7502 }, { "epoch": 2.440793754066363, "grad_norm": 2.074597875734221, "learning_rate": 1.022934018048432e-06, "loss": 0.1434, "step": 7503 }, { "epoch": 2.4411190631099546, "grad_norm": 2.274192442896423, "learning_rate": 1.0217871664574263e-06, "loss": 0.1879, "step": 7504 }, { "epoch": 2.441444372153546, "grad_norm": 2.143455781375106, "learning_rate": 1.0206408849469713e-06, "loss": 0.1745, "step": 7505 }, { "epoch": 2.441769681197137, "grad_norm": 2.34252356683169, "learning_rate": 1.0194951736813297e-06, "loss": 0.211, "step": 7506 }, { "epoch": 2.4420949902407285, "grad_norm": 2.199155042894362, "learning_rate": 1.0183500328246836e-06, "loss": 0.1414, "step": 7507 }, { "epoch": 2.4424202992843203, "grad_norm": 2.096977682946435, "learning_rate": 1.0172054625411337e-06, "loss": 0.1614, "step": 7508 }, { "epoch": 2.4427456083279115, "grad_norm": 2.121640860934745, "learning_rate": 1.0160614629946952e-06, "loss": 0.1909, "step": 7509 }, { "epoch": 2.443070917371503, "grad_norm": 2.081938473175329, "learning_rate": 1.0149180343493064e-06, "loss": 0.0991, "step": 7510 }, { "epoch": 2.4433962264150946, "grad_norm": 2.7972808994253144, "learning_rate": 1.0137751767688193e-06, "loss": 0.2405, "step": 7511 }, { "epoch": 2.443721535458686, "grad_norm": 2.0015903777462936, "learning_rate": 1.012632890417007e-06, "loss": 0.1506, "step": 7512 }, { "epoch": 2.444046844502277, "grad_norm": 1.9855899325349966, "learning_rate": 1.0114911754575613e-06, "loss": 0.1452, "step": 7513 }, { "epoch": 2.4443721535458685, "grad_norm": 2.4998723265954412, "learning_rate": 1.0103500320540882e-06, "loss": 0.1621, "step": 7514 }, { "epoch": 2.4446974625894597, "grad_norm": 2.046854053049679, "learning_rate": 1.0092094603701153e-06, "loss": 0.1733, "step": 7515 }, { "epoch": 2.4450227716330515, "grad_norm": 2.4346323400380094, "learning_rate": 1.008069460569089e-06, "loss": 0.1538, "step": 7516 }, { "epoch": 2.4453480806766428, "grad_norm": 2.1304367860067615, "learning_rate": 1.00693003281437e-06, "loss": 0.1463, "step": 7517 }, { "epoch": 2.445673389720234, "grad_norm": 2.461432315011337, "learning_rate": 1.0057911772692386e-06, "loss": 0.1407, "step": 7518 }, { "epoch": 2.445998698763826, "grad_norm": 2.199047092899734, "learning_rate": 1.0046528940968947e-06, "loss": 0.1803, "step": 7519 }, { "epoch": 2.446324007807417, "grad_norm": 2.037403074937914, "learning_rate": 1.0035151834604561e-06, "loss": 0.2075, "step": 7520 }, { "epoch": 2.4466493168510084, "grad_norm": 2.1011419335332637, "learning_rate": 1.0023780455229554e-06, "loss": 0.1843, "step": 7521 }, { "epoch": 2.4469746258945997, "grad_norm": 2.2144383962025977, "learning_rate": 1.0012414804473469e-06, "loss": 0.1687, "step": 7522 }, { "epoch": 2.4472999349381914, "grad_norm": 2.189613144894975, "learning_rate": 1.0001054883964988e-06, "loss": 0.1617, "step": 7523 }, { "epoch": 2.4476252439817827, "grad_norm": 2.3605201277063292, "learning_rate": 9.989700695332016e-07, "loss": 0.1601, "step": 7524 }, { "epoch": 2.447950553025374, "grad_norm": 2.461835405141605, "learning_rate": 9.978352240201622e-07, "loss": 0.1623, "step": 7525 }, { "epoch": 2.4482758620689653, "grad_norm": 2.3717349127395537, "learning_rate": 9.967009520200022e-07, "loss": 0.1733, "step": 7526 }, { "epoch": 2.448601171112557, "grad_norm": 2.4192510430931895, "learning_rate": 9.95567253695265e-07, "loss": 0.1923, "step": 7527 }, { "epoch": 2.4489264801561483, "grad_norm": 2.1229269113009095, "learning_rate": 9.944341292084114e-07, "loss": 0.176, "step": 7528 }, { "epoch": 2.4492517891997396, "grad_norm": 1.8410080391442152, "learning_rate": 9.933015787218164e-07, "loss": 0.1387, "step": 7529 }, { "epoch": 2.4495770982433314, "grad_norm": 2.104830870231153, "learning_rate": 9.921696023977784e-07, "loss": 0.1538, "step": 7530 }, { "epoch": 2.4499024072869227, "grad_norm": 2.497962238242707, "learning_rate": 9.910382003985063e-07, "loss": 0.1736, "step": 7531 }, { "epoch": 2.450227716330514, "grad_norm": 1.8959933409224372, "learning_rate": 9.899073728861336e-07, "loss": 0.1702, "step": 7532 }, { "epoch": 2.4505530253741052, "grad_norm": 2.4682274363335486, "learning_rate": 9.887771200227091e-07, "loss": 0.1844, "step": 7533 }, { "epoch": 2.450878334417697, "grad_norm": 2.389814538085884, "learning_rate": 9.876474419701976e-07, "loss": 0.169, "step": 7534 }, { "epoch": 2.4512036434612883, "grad_norm": 2.1820922116293833, "learning_rate": 9.865183388904803e-07, "loss": 0.1947, "step": 7535 }, { "epoch": 2.4515289525048796, "grad_norm": 1.9669892087472542, "learning_rate": 9.853898109453635e-07, "loss": 0.1678, "step": 7536 }, { "epoch": 2.451854261548471, "grad_norm": 2.025803271450766, "learning_rate": 9.84261858296564e-07, "loss": 0.142, "step": 7537 }, { "epoch": 2.4521795705920626, "grad_norm": 1.907754402263447, "learning_rate": 9.831344811057158e-07, "loss": 0.1264, "step": 7538 }, { "epoch": 2.452504879635654, "grad_norm": 2.331434541417698, "learning_rate": 9.820076795343748e-07, "loss": 0.1752, "step": 7539 }, { "epoch": 2.452830188679245, "grad_norm": 1.9746998820154837, "learning_rate": 9.808814537440132e-07, "loss": 0.1258, "step": 7540 }, { "epoch": 2.453155497722837, "grad_norm": 1.814885413657009, "learning_rate": 9.797558038960175e-07, "loss": 0.1504, "step": 7541 }, { "epoch": 2.453480806766428, "grad_norm": 1.8379588364778228, "learning_rate": 9.786307301516974e-07, "loss": 0.1206, "step": 7542 }, { "epoch": 2.4538061158100195, "grad_norm": 2.2649057854437533, "learning_rate": 9.775062326722734e-07, "loss": 0.1603, "step": 7543 }, { "epoch": 2.454131424853611, "grad_norm": 1.5553158568670993, "learning_rate": 9.76382311618888e-07, "loss": 0.1104, "step": 7544 }, { "epoch": 2.4544567338972025, "grad_norm": 1.9307851400873388, "learning_rate": 9.752589671526014e-07, "loss": 0.1626, "step": 7545 }, { "epoch": 2.454782042940794, "grad_norm": 2.0450789749266494, "learning_rate": 9.741361994343867e-07, "loss": 0.1353, "step": 7546 }, { "epoch": 2.455107351984385, "grad_norm": 2.01124989407184, "learning_rate": 9.730140086251394e-07, "loss": 0.1567, "step": 7547 }, { "epoch": 2.4554326610279764, "grad_norm": 2.332115975570245, "learning_rate": 9.7189239488567e-07, "loss": 0.1555, "step": 7548 }, { "epoch": 2.455757970071568, "grad_norm": 2.263586450258382, "learning_rate": 9.707713583767054e-07, "loss": 0.182, "step": 7549 }, { "epoch": 2.4560832791151594, "grad_norm": 2.164572278847636, "learning_rate": 9.696508992588927e-07, "loss": 0.1691, "step": 7550 }, { "epoch": 2.4564085881587507, "grad_norm": 1.769685507199163, "learning_rate": 9.68531017692792e-07, "loss": 0.12, "step": 7551 }, { "epoch": 2.456733897202342, "grad_norm": 1.940579658559141, "learning_rate": 9.674117138388844e-07, "loss": 0.159, "step": 7552 }, { "epoch": 2.4570592062459338, "grad_norm": 1.7611629230074133, "learning_rate": 9.662929878575684e-07, "loss": 0.1296, "step": 7553 }, { "epoch": 2.457384515289525, "grad_norm": 2.2829730167221154, "learning_rate": 9.65174839909156e-07, "loss": 0.1646, "step": 7554 }, { "epoch": 2.4577098243331164, "grad_norm": 1.8236228830090606, "learning_rate": 9.640572701538769e-07, "loss": 0.0952, "step": 7555 }, { "epoch": 2.4580351333767076, "grad_norm": 2.1125209374582274, "learning_rate": 9.629402787518844e-07, "loss": 0.1347, "step": 7556 }, { "epoch": 2.4583604424202994, "grad_norm": 1.9132097936804255, "learning_rate": 9.618238658632417e-07, "loss": 0.1597, "step": 7557 }, { "epoch": 2.4586857514638907, "grad_norm": 2.1765222679552005, "learning_rate": 9.6070803164793e-07, "loss": 0.1377, "step": 7558 }, { "epoch": 2.459011060507482, "grad_norm": 2.010319094051518, "learning_rate": 9.595927762658508e-07, "loss": 0.1599, "step": 7559 }, { "epoch": 2.4593363695510737, "grad_norm": 2.0306799265204303, "learning_rate": 9.584780998768217e-07, "loss": 0.156, "step": 7560 }, { "epoch": 2.459661678594665, "grad_norm": 1.8421920874284683, "learning_rate": 9.57364002640574e-07, "loss": 0.1633, "step": 7561 }, { "epoch": 2.4599869876382563, "grad_norm": 1.57916751091789, "learning_rate": 9.562504847167614e-07, "loss": 0.1249, "step": 7562 }, { "epoch": 2.4603122966818476, "grad_norm": 2.0126256956213378, "learning_rate": 9.551375462649492e-07, "loss": 0.1675, "step": 7563 }, { "epoch": 2.4606376057254393, "grad_norm": 2.2173987980083774, "learning_rate": 9.540251874446238e-07, "loss": 0.1589, "step": 7564 }, { "epoch": 2.4609629147690306, "grad_norm": 1.8000978214830423, "learning_rate": 9.529134084151875e-07, "loss": 0.1767, "step": 7565 }, { "epoch": 2.461288223812622, "grad_norm": 1.986605304984212, "learning_rate": 9.518022093359569e-07, "loss": 0.1587, "step": 7566 }, { "epoch": 2.461613532856213, "grad_norm": 2.7863006343549697, "learning_rate": 9.506915903661685e-07, "loss": 0.1746, "step": 7567 }, { "epoch": 2.461938841899805, "grad_norm": 2.1677450823029125, "learning_rate": 9.495815516649764e-07, "loss": 0.1313, "step": 7568 }, { "epoch": 2.4622641509433962, "grad_norm": 2.3820205077940555, "learning_rate": 9.484720933914488e-07, "loss": 0.1653, "step": 7569 }, { "epoch": 2.4625894599869875, "grad_norm": 1.8456785657531172, "learning_rate": 9.473632157045687e-07, "loss": 0.1694, "step": 7570 }, { "epoch": 2.4629147690305793, "grad_norm": 2.4752056284787054, "learning_rate": 9.462549187632448e-07, "loss": 0.1799, "step": 7571 }, { "epoch": 2.4632400780741706, "grad_norm": 2.0559075366632675, "learning_rate": 9.451472027262931e-07, "loss": 0.1487, "step": 7572 }, { "epoch": 2.463565387117762, "grad_norm": 1.9029436745569257, "learning_rate": 9.440400677524497e-07, "loss": 0.152, "step": 7573 }, { "epoch": 2.463890696161353, "grad_norm": 2.132286233455895, "learning_rate": 9.429335140003687e-07, "loss": 0.1944, "step": 7574 }, { "epoch": 2.464216005204945, "grad_norm": 2.1414776448456774, "learning_rate": 9.418275416286216e-07, "loss": 0.1544, "step": 7575 }, { "epoch": 2.464541314248536, "grad_norm": 2.5047631068197274, "learning_rate": 9.407221507956921e-07, "loss": 0.1674, "step": 7576 }, { "epoch": 2.4648666232921275, "grad_norm": 2.083022162807841, "learning_rate": 9.396173416599863e-07, "loss": 0.1788, "step": 7577 }, { "epoch": 2.4651919323357188, "grad_norm": 1.9756654000769125, "learning_rate": 9.385131143798209e-07, "loss": 0.124, "step": 7578 }, { "epoch": 2.4655172413793105, "grad_norm": 2.0254800022726163, "learning_rate": 9.37409469113435e-07, "loss": 0.1547, "step": 7579 }, { "epoch": 2.465842550422902, "grad_norm": 2.234571202432988, "learning_rate": 9.363064060189814e-07, "loss": 0.1651, "step": 7580 }, { "epoch": 2.466167859466493, "grad_norm": 2.150767537245108, "learning_rate": 9.352039252545281e-07, "loss": 0.135, "step": 7581 }, { "epoch": 2.466493168510085, "grad_norm": 1.8797821131697723, "learning_rate": 9.341020269780621e-07, "loss": 0.1597, "step": 7582 }, { "epoch": 2.466818477553676, "grad_norm": 1.7709652669458362, "learning_rate": 9.330007113474881e-07, "loss": 0.1434, "step": 7583 }, { "epoch": 2.4671437865972674, "grad_norm": 1.958205173143474, "learning_rate": 9.318999785206223e-07, "loss": 0.144, "step": 7584 }, { "epoch": 2.4674690956408587, "grad_norm": 2.3044529308970496, "learning_rate": 9.307998286552033e-07, "loss": 0.2022, "step": 7585 }, { "epoch": 2.46779440468445, "grad_norm": 1.8740569814114296, "learning_rate": 9.297002619088807e-07, "loss": 0.1386, "step": 7586 }, { "epoch": 2.4681197137280417, "grad_norm": 1.7489802515402255, "learning_rate": 9.286012784392246e-07, "loss": 0.1404, "step": 7587 }, { "epoch": 2.468445022771633, "grad_norm": 2.2232637665734956, "learning_rate": 9.275028784037204e-07, "loss": 0.1575, "step": 7588 }, { "epoch": 2.4687703318152243, "grad_norm": 1.8348944454537361, "learning_rate": 9.264050619597697e-07, "loss": 0.1251, "step": 7589 }, { "epoch": 2.469095640858816, "grad_norm": 1.8023607441760114, "learning_rate": 9.253078292646867e-07, "loss": 0.1275, "step": 7590 }, { "epoch": 2.4694209499024073, "grad_norm": 1.784880329148447, "learning_rate": 9.242111804757109e-07, "loss": 0.1115, "step": 7591 }, { "epoch": 2.4697462589459986, "grad_norm": 2.0589970200346164, "learning_rate": 9.231151157499907e-07, "loss": 0.1499, "step": 7592 }, { "epoch": 2.47007156798959, "grad_norm": 2.0957308909657613, "learning_rate": 9.220196352445909e-07, "loss": 0.1774, "step": 7593 }, { "epoch": 2.4703968770331817, "grad_norm": 1.821645285492309, "learning_rate": 9.209247391164966e-07, "loss": 0.1407, "step": 7594 }, { "epoch": 2.470722186076773, "grad_norm": 2.2775027433605204, "learning_rate": 9.198304275226083e-07, "loss": 0.1795, "step": 7595 }, { "epoch": 2.4710474951203643, "grad_norm": 2.370722117483339, "learning_rate": 9.187367006197385e-07, "loss": 0.157, "step": 7596 }, { "epoch": 2.4713728041639555, "grad_norm": 1.8455888932627558, "learning_rate": 9.17643558564622e-07, "loss": 0.1307, "step": 7597 }, { "epoch": 2.4716981132075473, "grad_norm": 2.2032281115779258, "learning_rate": 9.16551001513904e-07, "loss": 0.1548, "step": 7598 }, { "epoch": 2.4720234222511386, "grad_norm": 1.8319869170071774, "learning_rate": 9.1545902962415e-07, "loss": 0.1364, "step": 7599 }, { "epoch": 2.47234873129473, "grad_norm": 2.5096912407128986, "learning_rate": 9.143676430518416e-07, "loss": 0.1683, "step": 7600 }, { "epoch": 2.4726740403383216, "grad_norm": 2.3504508207594133, "learning_rate": 9.132768419533727e-07, "loss": 0.1986, "step": 7601 }, { "epoch": 2.472999349381913, "grad_norm": 2.165336688210656, "learning_rate": 9.121866264850577e-07, "loss": 0.2425, "step": 7602 }, { "epoch": 2.473324658425504, "grad_norm": 2.0628861053412177, "learning_rate": 9.110969968031253e-07, "loss": 0.1169, "step": 7603 }, { "epoch": 2.4736499674690955, "grad_norm": 2.2301719599716785, "learning_rate": 9.100079530637196e-07, "loss": 0.1852, "step": 7604 }, { "epoch": 2.4739752765126872, "grad_norm": 2.531647460377429, "learning_rate": 9.089194954228997e-07, "loss": 0.2175, "step": 7605 }, { "epoch": 2.4743005855562785, "grad_norm": 2.131940752643397, "learning_rate": 9.078316240366436e-07, "loss": 0.1583, "step": 7606 }, { "epoch": 2.47462589459987, "grad_norm": 2.312539701509215, "learning_rate": 9.067443390608455e-07, "loss": 0.1526, "step": 7607 }, { "epoch": 2.474951203643461, "grad_norm": 2.458371488093377, "learning_rate": 9.056576406513118e-07, "loss": 0.2062, "step": 7608 }, { "epoch": 2.475276512687053, "grad_norm": 1.8452543827600143, "learning_rate": 9.045715289637691e-07, "loss": 0.1482, "step": 7609 }, { "epoch": 2.475601821730644, "grad_norm": 2.6222334176145483, "learning_rate": 9.034860041538557e-07, "loss": 0.2178, "step": 7610 }, { "epoch": 2.4759271307742354, "grad_norm": 2.1773754949932904, "learning_rate": 9.024010663771288e-07, "loss": 0.1346, "step": 7611 }, { "epoch": 2.476252439817827, "grad_norm": 2.359509130333215, "learning_rate": 9.013167157890624e-07, "loss": 0.1585, "step": 7612 }, { "epoch": 2.4765777488614185, "grad_norm": 1.9307833536733798, "learning_rate": 9.002329525450421e-07, "loss": 0.1327, "step": 7613 }, { "epoch": 2.4769030579050098, "grad_norm": 2.2515543843528483, "learning_rate": 8.991497768003737e-07, "loss": 0.1933, "step": 7614 }, { "epoch": 2.477228366948601, "grad_norm": 2.1370441665863567, "learning_rate": 8.980671887102771e-07, "loss": 0.1674, "step": 7615 }, { "epoch": 2.4775536759921923, "grad_norm": 1.9325282127992818, "learning_rate": 8.969851884298858e-07, "loss": 0.1555, "step": 7616 }, { "epoch": 2.477878985035784, "grad_norm": 2.3747447837188322, "learning_rate": 8.959037761142547e-07, "loss": 0.1529, "step": 7617 }, { "epoch": 2.4782042940793754, "grad_norm": 2.3053425468307682, "learning_rate": 8.948229519183466e-07, "loss": 0.1971, "step": 7618 }, { "epoch": 2.4785296031229667, "grad_norm": 2.3278939398985465, "learning_rate": 8.937427159970474e-07, "loss": 0.189, "step": 7619 }, { "epoch": 2.4788549121665584, "grad_norm": 2.1754544000429483, "learning_rate": 8.926630685051551e-07, "loss": 0.1622, "step": 7620 }, { "epoch": 2.4791802212101497, "grad_norm": 2.5041038261152284, "learning_rate": 8.915840095973844e-07, "loss": 0.1843, "step": 7621 }, { "epoch": 2.479505530253741, "grad_norm": 1.9814444950759216, "learning_rate": 8.905055394283613e-07, "loss": 0.1589, "step": 7622 }, { "epoch": 2.4798308392973323, "grad_norm": 2.285925893628546, "learning_rate": 8.894276581526373e-07, "loss": 0.1493, "step": 7623 }, { "epoch": 2.480156148340924, "grad_norm": 2.458417116174423, "learning_rate": 8.883503659246701e-07, "loss": 0.2047, "step": 7624 }, { "epoch": 2.4804814573845153, "grad_norm": 1.9853794670688236, "learning_rate": 8.872736628988354e-07, "loss": 0.163, "step": 7625 }, { "epoch": 2.4808067664281066, "grad_norm": 2.0246985992653426, "learning_rate": 8.861975492294272e-07, "loss": 0.154, "step": 7626 }, { "epoch": 2.481132075471698, "grad_norm": 2.0107085804310327, "learning_rate": 8.851220250706544e-07, "loss": 0.1492, "step": 7627 }, { "epoch": 2.4814573845152896, "grad_norm": 1.9986122231058154, "learning_rate": 8.840470905766374e-07, "loss": 0.1521, "step": 7628 }, { "epoch": 2.481782693558881, "grad_norm": 2.4178587693652567, "learning_rate": 8.829727459014182e-07, "loss": 0.2173, "step": 7629 }, { "epoch": 2.482108002602472, "grad_norm": 1.9005946997933338, "learning_rate": 8.818989911989478e-07, "loss": 0.13, "step": 7630 }, { "epoch": 2.482433311646064, "grad_norm": 2.3880857344636683, "learning_rate": 8.808258266230979e-07, "loss": 0.2112, "step": 7631 }, { "epoch": 2.4827586206896552, "grad_norm": 2.4381433274113777, "learning_rate": 8.797532523276542e-07, "loss": 0.2255, "step": 7632 }, { "epoch": 2.4830839297332465, "grad_norm": 2.3569349254090546, "learning_rate": 8.786812684663154e-07, "loss": 0.1617, "step": 7633 }, { "epoch": 2.483409238776838, "grad_norm": 2.211336879584138, "learning_rate": 8.776098751926993e-07, "loss": 0.2062, "step": 7634 }, { "epoch": 2.4837345478204296, "grad_norm": 2.101219488381871, "learning_rate": 8.765390726603373e-07, "loss": 0.1712, "step": 7635 }, { "epoch": 2.484059856864021, "grad_norm": 1.8031432881104261, "learning_rate": 8.754688610226742e-07, "loss": 0.1724, "step": 7636 }, { "epoch": 2.484385165907612, "grad_norm": 1.9450294170943978, "learning_rate": 8.743992404330742e-07, "loss": 0.168, "step": 7637 }, { "epoch": 2.4847104749512035, "grad_norm": 2.5144166090071414, "learning_rate": 8.733302110448128e-07, "loss": 0.1586, "step": 7638 }, { "epoch": 2.485035783994795, "grad_norm": 2.400606345951721, "learning_rate": 8.722617730110839e-07, "loss": 0.1595, "step": 7639 }, { "epoch": 2.4853610930383865, "grad_norm": 1.7124014114690531, "learning_rate": 8.711939264849956e-07, "loss": 0.1221, "step": 7640 }, { "epoch": 2.4856864020819778, "grad_norm": 2.2277329908142818, "learning_rate": 8.701266716195711e-07, "loss": 0.1336, "step": 7641 }, { "epoch": 2.4860117111255695, "grad_norm": 2.352801004068851, "learning_rate": 8.690600085677447e-07, "loss": 0.2146, "step": 7642 }, { "epoch": 2.486337020169161, "grad_norm": 2.9433324534940524, "learning_rate": 8.679939374823765e-07, "loss": 0.1938, "step": 7643 }, { "epoch": 2.486662329212752, "grad_norm": 2.161149173668036, "learning_rate": 8.669284585162319e-07, "loss": 0.1458, "step": 7644 }, { "epoch": 2.4869876382563434, "grad_norm": 2.0062821500979777, "learning_rate": 8.658635718219932e-07, "loss": 0.1296, "step": 7645 }, { "epoch": 2.487312947299935, "grad_norm": 2.190693542819501, "learning_rate": 8.647992775522612e-07, "loss": 0.1582, "step": 7646 }, { "epoch": 2.4876382563435264, "grad_norm": 2.580577284233086, "learning_rate": 8.637355758595511e-07, "loss": 0.1898, "step": 7647 }, { "epoch": 2.4879635653871177, "grad_norm": 1.6737398965473391, "learning_rate": 8.626724668962899e-07, "loss": 0.0909, "step": 7648 }, { "epoch": 2.488288874430709, "grad_norm": 1.8782169168779106, "learning_rate": 8.616099508148235e-07, "loss": 0.1478, "step": 7649 }, { "epoch": 2.4886141834743007, "grad_norm": 2.6246054206112155, "learning_rate": 8.605480277674089e-07, "loss": 0.1991, "step": 7650 }, { "epoch": 2.488939492517892, "grad_norm": 2.031506766251682, "learning_rate": 8.594866979062221e-07, "loss": 0.1198, "step": 7651 }, { "epoch": 2.4892648015614833, "grad_norm": 2.1029534424284204, "learning_rate": 8.584259613833534e-07, "loss": 0.1551, "step": 7652 }, { "epoch": 2.489590110605075, "grad_norm": 1.9390553380264797, "learning_rate": 8.573658183508055e-07, "loss": 0.1392, "step": 7653 }, { "epoch": 2.4899154196486664, "grad_norm": 2.112222393233758, "learning_rate": 8.563062689604956e-07, "loss": 0.1702, "step": 7654 }, { "epoch": 2.4902407286922577, "grad_norm": 2.260172008095237, "learning_rate": 8.552473133642619e-07, "loss": 0.1656, "step": 7655 }, { "epoch": 2.490566037735849, "grad_norm": 1.6447587956696847, "learning_rate": 8.541889517138524e-07, "loss": 0.139, "step": 7656 }, { "epoch": 2.4908913467794402, "grad_norm": 1.8562515343685813, "learning_rate": 8.531311841609285e-07, "loss": 0.125, "step": 7657 }, { "epoch": 2.491216655823032, "grad_norm": 2.524852763781444, "learning_rate": 8.520740108570713e-07, "loss": 0.1518, "step": 7658 }, { "epoch": 2.4915419648666233, "grad_norm": 2.5652220441806217, "learning_rate": 8.510174319537751e-07, "loss": 0.1686, "step": 7659 }, { "epoch": 2.4918672739102146, "grad_norm": 2.361311306795407, "learning_rate": 8.499614476024453e-07, "loss": 0.154, "step": 7660 }, { "epoch": 2.4921925829538063, "grad_norm": 2.158502294054691, "learning_rate": 8.48906057954409e-07, "loss": 0.218, "step": 7661 }, { "epoch": 2.4925178919973976, "grad_norm": 2.1006027132560874, "learning_rate": 8.478512631609004e-07, "loss": 0.1631, "step": 7662 }, { "epoch": 2.492843201040989, "grad_norm": 2.025228313721281, "learning_rate": 8.467970633730743e-07, "loss": 0.1277, "step": 7663 }, { "epoch": 2.49316851008458, "grad_norm": 2.2970771317985195, "learning_rate": 8.457434587419994e-07, "loss": 0.1717, "step": 7664 }, { "epoch": 2.493493819128172, "grad_norm": 1.9956877072660444, "learning_rate": 8.446904494186553e-07, "loss": 0.191, "step": 7665 }, { "epoch": 2.493819128171763, "grad_norm": 2.449544180415096, "learning_rate": 8.436380355539404e-07, "loss": 0.1914, "step": 7666 }, { "epoch": 2.4941444372153545, "grad_norm": 2.254442349869205, "learning_rate": 8.425862172986676e-07, "loss": 0.1517, "step": 7667 }, { "epoch": 2.494469746258946, "grad_norm": 2.279947713144762, "learning_rate": 8.4153499480356e-07, "loss": 0.1964, "step": 7668 }, { "epoch": 2.4947950553025375, "grad_norm": 2.0625637838402917, "learning_rate": 8.404843682192614e-07, "loss": 0.1348, "step": 7669 }, { "epoch": 2.495120364346129, "grad_norm": 2.5523352974941442, "learning_rate": 8.394343376963255e-07, "loss": 0.1895, "step": 7670 }, { "epoch": 2.49544567338972, "grad_norm": 2.131790948835324, "learning_rate": 8.383849033852226e-07, "loss": 0.1774, "step": 7671 }, { "epoch": 2.495770982433312, "grad_norm": 1.9319708841870755, "learning_rate": 8.373360654363388e-07, "loss": 0.1609, "step": 7672 }, { "epoch": 2.496096291476903, "grad_norm": 2.0417079136095553, "learning_rate": 8.362878239999728e-07, "loss": 0.1619, "step": 7673 }, { "epoch": 2.4964216005204944, "grad_norm": 1.839664286129007, "learning_rate": 8.352401792263348e-07, "loss": 0.1593, "step": 7674 }, { "epoch": 2.4967469095640857, "grad_norm": 1.8634559024230533, "learning_rate": 8.341931312655582e-07, "loss": 0.1391, "step": 7675 }, { "epoch": 2.4970722186076775, "grad_norm": 2.293474875092144, "learning_rate": 8.331466802676835e-07, "loss": 0.1854, "step": 7676 }, { "epoch": 2.4973975276512688, "grad_norm": 1.9698427893023667, "learning_rate": 8.321008263826663e-07, "loss": 0.1386, "step": 7677 }, { "epoch": 2.49772283669486, "grad_norm": 2.405408521305933, "learning_rate": 8.3105556976038e-07, "loss": 0.1463, "step": 7678 }, { "epoch": 2.4980481457384514, "grad_norm": 2.2460888818590465, "learning_rate": 8.30010910550611e-07, "loss": 0.2171, "step": 7679 }, { "epoch": 2.498373454782043, "grad_norm": 1.8965398062882501, "learning_rate": 8.289668489030578e-07, "loss": 0.1535, "step": 7680 }, { "epoch": 2.4986987638256344, "grad_norm": 2.1041604875035937, "learning_rate": 8.279233849673368e-07, "loss": 0.148, "step": 7681 }, { "epoch": 2.4990240728692257, "grad_norm": 2.3491782403015, "learning_rate": 8.268805188929757e-07, "loss": 0.1741, "step": 7682 }, { "epoch": 2.4993493819128174, "grad_norm": 1.9785440156307816, "learning_rate": 8.258382508294183e-07, "loss": 0.1332, "step": 7683 }, { "epoch": 2.4996746909564087, "grad_norm": 2.2844282320494878, "learning_rate": 8.247965809260239e-07, "loss": 0.1587, "step": 7684 }, { "epoch": 2.5, "grad_norm": 1.6796792570862038, "learning_rate": 8.23755509332062e-07, "loss": 0.1166, "step": 7685 }, { "epoch": 2.5003253090435913, "grad_norm": 2.1711430379230126, "learning_rate": 8.227150361967201e-07, "loss": 0.1437, "step": 7686 }, { "epoch": 2.5006506180871826, "grad_norm": 2.033549979730979, "learning_rate": 8.216751616690999e-07, "loss": 0.168, "step": 7687 }, { "epoch": 2.5009759271307743, "grad_norm": 2.1510662045527114, "learning_rate": 8.206358858982133e-07, "loss": 0.1722, "step": 7688 }, { "epoch": 2.5013012361743656, "grad_norm": 2.27905926644488, "learning_rate": 8.195972090329918e-07, "loss": 0.2138, "step": 7689 }, { "epoch": 2.501626545217957, "grad_norm": 2.0957740508475338, "learning_rate": 8.185591312222769e-07, "loss": 0.1321, "step": 7690 }, { "epoch": 2.5019518542615486, "grad_norm": 2.0049616744297563, "learning_rate": 8.175216526148272e-07, "loss": 0.1561, "step": 7691 }, { "epoch": 2.50227716330514, "grad_norm": 2.437936290827264, "learning_rate": 8.164847733593123e-07, "loss": 0.174, "step": 7692 }, { "epoch": 2.5026024723487312, "grad_norm": 1.9176105639102616, "learning_rate": 8.154484936043183e-07, "loss": 0.1452, "step": 7693 }, { "epoch": 2.502927781392323, "grad_norm": 1.9770240998572493, "learning_rate": 8.144128134983464e-07, "loss": 0.0935, "step": 7694 }, { "epoch": 2.5032530904359143, "grad_norm": 2.1882309545424636, "learning_rate": 8.133777331898079e-07, "loss": 0.1429, "step": 7695 }, { "epoch": 2.5035783994795056, "grad_norm": 2.3314148354911097, "learning_rate": 8.12343252827033e-07, "loss": 0.1792, "step": 7696 }, { "epoch": 2.503903708523097, "grad_norm": 2.008952303591209, "learning_rate": 8.113093725582604e-07, "loss": 0.1626, "step": 7697 }, { "epoch": 2.504229017566688, "grad_norm": 1.9091914706027053, "learning_rate": 8.102760925316477e-07, "loss": 0.185, "step": 7698 }, { "epoch": 2.50455432661028, "grad_norm": 2.1074071862006494, "learning_rate": 8.092434128952653e-07, "loss": 0.176, "step": 7699 }, { "epoch": 2.504879635653871, "grad_norm": 1.967519082730132, "learning_rate": 8.082113337970948e-07, "loss": 0.1667, "step": 7700 }, { "epoch": 2.5052049446974625, "grad_norm": 2.0112227870421733, "learning_rate": 8.071798553850352e-07, "loss": 0.1606, "step": 7701 }, { "epoch": 2.505530253741054, "grad_norm": 2.146165502619617, "learning_rate": 8.061489778068987e-07, "loss": 0.1704, "step": 7702 }, { "epoch": 2.5058555627846455, "grad_norm": 1.7236735334717892, "learning_rate": 8.051187012104089e-07, "loss": 0.1279, "step": 7703 }, { "epoch": 2.506180871828237, "grad_norm": 1.867294442609878, "learning_rate": 8.040890257432071e-07, "loss": 0.1503, "step": 7704 }, { "epoch": 2.506506180871828, "grad_norm": 2.4207409865770786, "learning_rate": 8.030599515528442e-07, "loss": 0.1829, "step": 7705 }, { "epoch": 2.5068314899154194, "grad_norm": 2.2619005172275597, "learning_rate": 8.020314787867888e-07, "loss": 0.2136, "step": 7706 }, { "epoch": 2.507156798959011, "grad_norm": 1.8675294766719213, "learning_rate": 8.010036075924221e-07, "loss": 0.1252, "step": 7707 }, { "epoch": 2.5074821080026024, "grad_norm": 2.1462008097582244, "learning_rate": 7.999763381170384e-07, "loss": 0.1708, "step": 7708 }, { "epoch": 2.5078074170461937, "grad_norm": 2.0961431989066877, "learning_rate": 7.989496705078436e-07, "loss": 0.1411, "step": 7709 }, { "epoch": 2.5081327260897854, "grad_norm": 1.9776915346426571, "learning_rate": 7.979236049119637e-07, "loss": 0.1552, "step": 7710 }, { "epoch": 2.5084580351333767, "grad_norm": 2.5959505981381965, "learning_rate": 7.968981414764331e-07, "loss": 0.1649, "step": 7711 }, { "epoch": 2.508783344176968, "grad_norm": 1.7883206744212619, "learning_rate": 7.958732803482e-07, "loss": 0.1414, "step": 7712 }, { "epoch": 2.5091086532205598, "grad_norm": 1.7995129103992968, "learning_rate": 7.948490216741289e-07, "loss": 0.1117, "step": 7713 }, { "epoch": 2.509433962264151, "grad_norm": 2.0502142792079048, "learning_rate": 7.938253656009981e-07, "loss": 0.1692, "step": 7714 }, { "epoch": 2.5097592713077423, "grad_norm": 2.227738834781452, "learning_rate": 7.928023122754952e-07, "loss": 0.1723, "step": 7715 }, { "epoch": 2.5100845803513336, "grad_norm": 1.8004605105673706, "learning_rate": 7.917798618442273e-07, "loss": 0.1143, "step": 7716 }, { "epoch": 2.510409889394925, "grad_norm": 2.12715393282277, "learning_rate": 7.907580144537103e-07, "loss": 0.1826, "step": 7717 }, { "epoch": 2.5107351984385167, "grad_norm": 2.085093855818402, "learning_rate": 7.897367702503755e-07, "loss": 0.1441, "step": 7718 }, { "epoch": 2.511060507482108, "grad_norm": 2.4104361699456924, "learning_rate": 7.8871612938057e-07, "loss": 0.202, "step": 7719 }, { "epoch": 2.5113858165256993, "grad_norm": 1.8391174193506468, "learning_rate": 7.876960919905497e-07, "loss": 0.1752, "step": 7720 }, { "epoch": 2.511711125569291, "grad_norm": 2.2896698429017457, "learning_rate": 7.866766582264884e-07, "loss": 0.1958, "step": 7721 }, { "epoch": 2.5120364346128823, "grad_norm": 1.9868599167466423, "learning_rate": 7.856578282344718e-07, "loss": 0.1325, "step": 7722 }, { "epoch": 2.5123617436564736, "grad_norm": 2.0370061948846723, "learning_rate": 7.846396021604973e-07, "loss": 0.1301, "step": 7723 }, { "epoch": 2.5126870527000653, "grad_norm": 2.13209454663558, "learning_rate": 7.836219801504791e-07, "loss": 0.1616, "step": 7724 }, { "epoch": 2.5130123617436566, "grad_norm": 2.175332084732153, "learning_rate": 7.826049623502419e-07, "loss": 0.173, "step": 7725 }, { "epoch": 2.513337670787248, "grad_norm": 2.152504288747891, "learning_rate": 7.815885489055247e-07, "loss": 0.173, "step": 7726 }, { "epoch": 2.513662979830839, "grad_norm": 2.059127461469397, "learning_rate": 7.805727399619822e-07, "loss": 0.1441, "step": 7727 }, { "epoch": 2.5139882888744305, "grad_norm": 2.1594079985702233, "learning_rate": 7.795575356651797e-07, "loss": 0.1658, "step": 7728 }, { "epoch": 2.5143135979180222, "grad_norm": 1.8025080443999904, "learning_rate": 7.78542936160594e-07, "loss": 0.1695, "step": 7729 }, { "epoch": 2.5146389069616135, "grad_norm": 2.09171129859514, "learning_rate": 7.775289415936221e-07, "loss": 0.1327, "step": 7730 }, { "epoch": 2.514964216005205, "grad_norm": 1.9702731922766175, "learning_rate": 7.765155521095679e-07, "loss": 0.1538, "step": 7731 }, { "epoch": 2.5152895250487965, "grad_norm": 2.0399581776317928, "learning_rate": 7.755027678536503e-07, "loss": 0.1566, "step": 7732 }, { "epoch": 2.515614834092388, "grad_norm": 2.1824249207478057, "learning_rate": 7.744905889710025e-07, "loss": 0.1647, "step": 7733 }, { "epoch": 2.515940143135979, "grad_norm": 2.375403988404548, "learning_rate": 7.734790156066718e-07, "loss": 0.1708, "step": 7734 }, { "epoch": 2.5162654521795704, "grad_norm": 2.3779792858893973, "learning_rate": 7.724680479056152e-07, "loss": 0.1598, "step": 7735 }, { "epoch": 2.516590761223162, "grad_norm": 2.2860339785048556, "learning_rate": 7.714576860127065e-07, "loss": 0.1681, "step": 7736 }, { "epoch": 2.5169160702667535, "grad_norm": 2.301946089474037, "learning_rate": 7.704479300727297e-07, "loss": 0.1972, "step": 7737 }, { "epoch": 2.5172413793103448, "grad_norm": 2.077861001398205, "learning_rate": 7.694387802303849e-07, "loss": 0.1491, "step": 7738 }, { "epoch": 2.517566688353936, "grad_norm": 2.2836063293251874, "learning_rate": 7.684302366302842e-07, "loss": 0.1832, "step": 7739 }, { "epoch": 2.517891997397528, "grad_norm": 1.7774608960230835, "learning_rate": 7.674222994169522e-07, "loss": 0.1168, "step": 7740 }, { "epoch": 2.518217306441119, "grad_norm": 2.199021476963734, "learning_rate": 7.664149687348243e-07, "loss": 0.1756, "step": 7741 }, { "epoch": 2.5185426154847104, "grad_norm": 2.124206445596249, "learning_rate": 7.654082447282562e-07, "loss": 0.1949, "step": 7742 }, { "epoch": 2.518867924528302, "grad_norm": 2.35513036096407, "learning_rate": 7.644021275415103e-07, "loss": 0.1411, "step": 7743 }, { "epoch": 2.5191932335718934, "grad_norm": 2.215340414865728, "learning_rate": 7.633966173187618e-07, "loss": 0.1849, "step": 7744 }, { "epoch": 2.5195185426154847, "grad_norm": 2.3058400298910056, "learning_rate": 7.623917142041032e-07, "loss": 0.1831, "step": 7745 }, { "epoch": 2.519843851659076, "grad_norm": 1.9859815639985838, "learning_rate": 7.613874183415382e-07, "loss": 0.1382, "step": 7746 }, { "epoch": 2.5201691607026673, "grad_norm": 2.3356434594927253, "learning_rate": 7.603837298749811e-07, "loss": 0.1861, "step": 7747 }, { "epoch": 2.520494469746259, "grad_norm": 2.376265258653805, "learning_rate": 7.593806489482635e-07, "loss": 0.1044, "step": 7748 }, { "epoch": 2.5208197787898503, "grad_norm": 2.449795759807282, "learning_rate": 7.583781757051251e-07, "loss": 0.1878, "step": 7749 }, { "epoch": 2.5211450878334416, "grad_norm": 1.8128543217962325, "learning_rate": 7.573763102892218e-07, "loss": 0.1087, "step": 7750 }, { "epoch": 2.5214703968770333, "grad_norm": 2.2246342006221704, "learning_rate": 7.563750528441233e-07, "loss": 0.1746, "step": 7751 }, { "epoch": 2.5217957059206246, "grad_norm": 2.13213950127519, "learning_rate": 7.553744035133081e-07, "loss": 0.1702, "step": 7752 }, { "epoch": 2.522121014964216, "grad_norm": 2.327075469981794, "learning_rate": 7.543743624401706e-07, "loss": 0.2042, "step": 7753 }, { "epoch": 2.5224463240078077, "grad_norm": 2.148236119416503, "learning_rate": 7.533749297680187e-07, "loss": 0.1887, "step": 7754 }, { "epoch": 2.522771633051399, "grad_norm": 2.1841004401287183, "learning_rate": 7.523761056400697e-07, "loss": 0.1771, "step": 7755 }, { "epoch": 2.5230969420949902, "grad_norm": 2.2551841483924036, "learning_rate": 7.513778901994578e-07, "loss": 0.1718, "step": 7756 }, { "epoch": 2.5234222511385815, "grad_norm": 2.4537339774835263, "learning_rate": 7.50380283589226e-07, "loss": 0.1579, "step": 7757 }, { "epoch": 2.523747560182173, "grad_norm": 2.0950744508616492, "learning_rate": 7.493832859523331e-07, "loss": 0.1559, "step": 7758 }, { "epoch": 2.5240728692257646, "grad_norm": 1.9809121689569935, "learning_rate": 7.483868974316499e-07, "loss": 0.1451, "step": 7759 }, { "epoch": 2.524398178269356, "grad_norm": 2.048307162483649, "learning_rate": 7.473911181699595e-07, "loss": 0.1547, "step": 7760 }, { "epoch": 2.524723487312947, "grad_norm": 2.148139726698375, "learning_rate": 7.463959483099547e-07, "loss": 0.1689, "step": 7761 }, { "epoch": 2.525048796356539, "grad_norm": 1.9597763663727699, "learning_rate": 7.45401387994249e-07, "loss": 0.1401, "step": 7762 }, { "epoch": 2.52537410540013, "grad_norm": 2.112540991231108, "learning_rate": 7.444074373653609e-07, "loss": 0.1587, "step": 7763 }, { "epoch": 2.5256994144437215, "grad_norm": 2.2168549754747113, "learning_rate": 7.434140965657233e-07, "loss": 0.1701, "step": 7764 }, { "epoch": 2.526024723487313, "grad_norm": 1.7538686325734498, "learning_rate": 7.424213657376833e-07, "loss": 0.1412, "step": 7765 }, { "epoch": 2.5263500325309045, "grad_norm": 1.8990720416586637, "learning_rate": 7.414292450235017e-07, "loss": 0.1118, "step": 7766 }, { "epoch": 2.526675341574496, "grad_norm": 2.1888854495323495, "learning_rate": 7.404377345653474e-07, "loss": 0.1595, "step": 7767 }, { "epoch": 2.527000650618087, "grad_norm": 2.237415689159962, "learning_rate": 7.394468345053068e-07, "loss": 0.1719, "step": 7768 }, { "epoch": 2.5273259596616784, "grad_norm": 1.7642985903357353, "learning_rate": 7.384565449853742e-07, "loss": 0.1275, "step": 7769 }, { "epoch": 2.52765126870527, "grad_norm": 2.1380314332722477, "learning_rate": 7.374668661474604e-07, "loss": 0.1538, "step": 7770 }, { "epoch": 2.5279765777488614, "grad_norm": 2.104881921685971, "learning_rate": 7.364777981333876e-07, "loss": 0.1582, "step": 7771 }, { "epoch": 2.5283018867924527, "grad_norm": 1.9851645310945323, "learning_rate": 7.354893410848874e-07, "loss": 0.1907, "step": 7772 }, { "epoch": 2.5286271958360445, "grad_norm": 2.1857648607477387, "learning_rate": 7.345014951436085e-07, "loss": 0.1944, "step": 7773 }, { "epoch": 2.5289525048796357, "grad_norm": 2.592610506008236, "learning_rate": 7.335142604511103e-07, "loss": 0.1986, "step": 7774 }, { "epoch": 2.529277813923227, "grad_norm": 1.9894465044858303, "learning_rate": 7.325276371488616e-07, "loss": 0.1826, "step": 7775 }, { "epoch": 2.5296031229668183, "grad_norm": 1.9023402900262116, "learning_rate": 7.315416253782492e-07, "loss": 0.1766, "step": 7776 }, { "epoch": 2.5299284320104096, "grad_norm": 2.34944898220808, "learning_rate": 7.305562252805665e-07, "loss": 0.1688, "step": 7777 }, { "epoch": 2.5302537410540014, "grad_norm": 2.201764845054548, "learning_rate": 7.295714369970247e-07, "loss": 0.1504, "step": 7778 }, { "epoch": 2.5305790500975927, "grad_norm": 2.1738100680857944, "learning_rate": 7.285872606687417e-07, "loss": 0.171, "step": 7779 }, { "epoch": 2.530904359141184, "grad_norm": 1.8983561393636645, "learning_rate": 7.276036964367539e-07, "loss": 0.1392, "step": 7780 }, { "epoch": 2.5312296681847757, "grad_norm": 2.156555844803226, "learning_rate": 7.266207444420031e-07, "loss": 0.1448, "step": 7781 }, { "epoch": 2.531554977228367, "grad_norm": 2.3525065272936048, "learning_rate": 7.25638404825349e-07, "loss": 0.1942, "step": 7782 }, { "epoch": 2.5318802862719583, "grad_norm": 2.023708678582079, "learning_rate": 7.246566777275627e-07, "loss": 0.1619, "step": 7783 }, { "epoch": 2.53220559531555, "grad_norm": 2.0818048992437155, "learning_rate": 7.236755632893233e-07, "loss": 0.1713, "step": 7784 }, { "epoch": 2.5325309043591413, "grad_norm": 2.24103436810185, "learning_rate": 7.226950616512274e-07, "loss": 0.1847, "step": 7785 }, { "epoch": 2.5328562134027326, "grad_norm": 2.4522261970188577, "learning_rate": 7.21715172953782e-07, "loss": 0.2284, "step": 7786 }, { "epoch": 2.533181522446324, "grad_norm": 2.2318600090791803, "learning_rate": 7.207358973374034e-07, "loss": 0.1747, "step": 7787 }, { "epoch": 2.533506831489915, "grad_norm": 2.5631151230999083, "learning_rate": 7.197572349424253e-07, "loss": 0.2233, "step": 7788 }, { "epoch": 2.533832140533507, "grad_norm": 2.1242815788538176, "learning_rate": 7.187791859090875e-07, "loss": 0.2296, "step": 7789 }, { "epoch": 2.534157449577098, "grad_norm": 2.29065372330826, "learning_rate": 7.178017503775475e-07, "loss": 0.1708, "step": 7790 }, { "epoch": 2.5344827586206895, "grad_norm": 1.8482363012128327, "learning_rate": 7.168249284878725e-07, "loss": 0.1383, "step": 7791 }, { "epoch": 2.5348080676642812, "grad_norm": 2.2284463573307915, "learning_rate": 7.158487203800412e-07, "loss": 0.1492, "step": 7792 }, { "epoch": 2.5351333767078725, "grad_norm": 2.1014923351226282, "learning_rate": 7.148731261939418e-07, "loss": 0.1793, "step": 7793 }, { "epoch": 2.535458685751464, "grad_norm": 2.1494107717985598, "learning_rate": 7.138981460693833e-07, "loss": 0.2027, "step": 7794 }, { "epoch": 2.5357839947950556, "grad_norm": 2.1509336447424428, "learning_rate": 7.129237801460781e-07, "loss": 0.1648, "step": 7795 }, { "epoch": 2.536109303838647, "grad_norm": 2.590561218542589, "learning_rate": 7.119500285636522e-07, "loss": 0.1511, "step": 7796 }, { "epoch": 2.536434612882238, "grad_norm": 2.0596857450656834, "learning_rate": 7.109768914616461e-07, "loss": 0.1762, "step": 7797 }, { "epoch": 2.5367599219258294, "grad_norm": 2.482514131491687, "learning_rate": 7.100043689795122e-07, "loss": 0.1496, "step": 7798 }, { "epoch": 2.5370852309694207, "grad_norm": 1.890555506294311, "learning_rate": 7.090324612566113e-07, "loss": 0.1448, "step": 7799 }, { "epoch": 2.5374105400130125, "grad_norm": 2.0922944069305185, "learning_rate": 7.080611684322208e-07, "loss": 0.1558, "step": 7800 }, { "epoch": 2.5377358490566038, "grad_norm": 1.998694698885831, "learning_rate": 7.070904906455256e-07, "loss": 0.1532, "step": 7801 }, { "epoch": 2.538061158100195, "grad_norm": 2.086245818365961, "learning_rate": 7.061204280356248e-07, "loss": 0.1641, "step": 7802 }, { "epoch": 2.538386467143787, "grad_norm": 2.0638378236837696, "learning_rate": 7.051509807415313e-07, "loss": 0.1484, "step": 7803 }, { "epoch": 2.538711776187378, "grad_norm": 2.1336143250758943, "learning_rate": 7.041821489021639e-07, "loss": 0.1837, "step": 7804 }, { "epoch": 2.5390370852309694, "grad_norm": 2.001193520417189, "learning_rate": 7.032139326563592e-07, "loss": 0.1234, "step": 7805 }, { "epoch": 2.5393623942745607, "grad_norm": 2.465103972009537, "learning_rate": 7.02246332142864e-07, "loss": 0.2347, "step": 7806 }, { "epoch": 2.539687703318152, "grad_norm": 1.978427752843486, "learning_rate": 7.012793475003333e-07, "loss": 0.1382, "step": 7807 }, { "epoch": 2.5400130123617437, "grad_norm": 1.9176631126232397, "learning_rate": 7.003129788673397e-07, "loss": 0.141, "step": 7808 }, { "epoch": 2.540338321405335, "grad_norm": 1.7979861667488282, "learning_rate": 6.993472263823614e-07, "loss": 0.1192, "step": 7809 }, { "epoch": 2.5406636304489263, "grad_norm": 1.8126069814688148, "learning_rate": 6.983820901837929e-07, "loss": 0.1024, "step": 7810 }, { "epoch": 2.540988939492518, "grad_norm": 1.9120747156166382, "learning_rate": 6.974175704099406e-07, "loss": 0.14, "step": 7811 }, { "epoch": 2.5413142485361093, "grad_norm": 2.1300021646007394, "learning_rate": 6.964536671990185e-07, "loss": 0.153, "step": 7812 }, { "epoch": 2.5416395575797006, "grad_norm": 2.4967868711215435, "learning_rate": 6.954903806891533e-07, "loss": 0.1346, "step": 7813 }, { "epoch": 2.5419648666232924, "grad_norm": 2.0417311097521122, "learning_rate": 6.94527711018389e-07, "loss": 0.1775, "step": 7814 }, { "epoch": 2.5422901756668836, "grad_norm": 2.6220673717828564, "learning_rate": 6.935656583246736e-07, "loss": 0.1842, "step": 7815 }, { "epoch": 2.542615484710475, "grad_norm": 2.530681647919051, "learning_rate": 6.926042227458701e-07, "loss": 0.1356, "step": 7816 }, { "epoch": 2.5429407937540662, "grad_norm": 2.4064026476012645, "learning_rate": 6.916434044197529e-07, "loss": 0.1815, "step": 7817 }, { "epoch": 2.5432661027976575, "grad_norm": 2.062116105722315, "learning_rate": 6.906832034840094e-07, "loss": 0.1536, "step": 7818 }, { "epoch": 2.5435914118412493, "grad_norm": 2.405492259406485, "learning_rate": 6.897236200762353e-07, "loss": 0.1928, "step": 7819 }, { "epoch": 2.5439167208848406, "grad_norm": 2.0432601253579863, "learning_rate": 6.887646543339399e-07, "loss": 0.1352, "step": 7820 }, { "epoch": 2.544242029928432, "grad_norm": 2.1337957009480197, "learning_rate": 6.878063063945456e-07, "loss": 0.1745, "step": 7821 }, { "epoch": 2.5445673389720236, "grad_norm": 2.157127850655407, "learning_rate": 6.868485763953809e-07, "loss": 0.1488, "step": 7822 }, { "epoch": 2.544892648015615, "grad_norm": 2.243707851110689, "learning_rate": 6.858914644736925e-07, "loss": 0.1438, "step": 7823 }, { "epoch": 2.545217957059206, "grad_norm": 2.1119214715580554, "learning_rate": 6.84934970766632e-07, "loss": 0.167, "step": 7824 }, { "epoch": 2.545543266102798, "grad_norm": 1.9313071625704488, "learning_rate": 6.839790954112674e-07, "loss": 0.1543, "step": 7825 }, { "epoch": 2.545868575146389, "grad_norm": 2.180279603681604, "learning_rate": 6.830238385445775e-07, "loss": 0.1643, "step": 7826 }, { "epoch": 2.5461938841899805, "grad_norm": 2.3767000586635496, "learning_rate": 6.820692003034496e-07, "loss": 0.1603, "step": 7827 }, { "epoch": 2.546519193233572, "grad_norm": 2.172626379097019, "learning_rate": 6.811151808246819e-07, "loss": 0.1842, "step": 7828 }, { "epoch": 2.546844502277163, "grad_norm": 2.818380797997404, "learning_rate": 6.8016178024499e-07, "loss": 0.2459, "step": 7829 }, { "epoch": 2.547169811320755, "grad_norm": 2.391975546427485, "learning_rate": 6.792089987009959e-07, "loss": 0.2215, "step": 7830 }, { "epoch": 2.547495120364346, "grad_norm": 2.0990533253657633, "learning_rate": 6.782568363292319e-07, "loss": 0.1451, "step": 7831 }, { "epoch": 2.5478204294079374, "grad_norm": 2.36377367674058, "learning_rate": 6.773052932661445e-07, "loss": 0.2077, "step": 7832 }, { "epoch": 2.548145738451529, "grad_norm": 1.9378394609583576, "learning_rate": 6.763543696480918e-07, "loss": 0.1074, "step": 7833 }, { "epoch": 2.5484710474951204, "grad_norm": 2.510013385630531, "learning_rate": 6.754040656113392e-07, "loss": 0.181, "step": 7834 }, { "epoch": 2.5487963565387117, "grad_norm": 2.393173869051581, "learning_rate": 6.74454381292069e-07, "loss": 0.1608, "step": 7835 }, { "epoch": 2.5491216655823035, "grad_norm": 2.118427556353261, "learning_rate": 6.735053168263683e-07, "loss": 0.1594, "step": 7836 }, { "epoch": 2.5494469746258948, "grad_norm": 2.203782714378607, "learning_rate": 6.725568723502407e-07, "loss": 0.1887, "step": 7837 }, { "epoch": 2.549772283669486, "grad_norm": 2.079227189989334, "learning_rate": 6.71609047999599e-07, "loss": 0.147, "step": 7838 }, { "epoch": 2.5500975927130773, "grad_norm": 1.9539188913087404, "learning_rate": 6.706618439102664e-07, "loss": 0.1505, "step": 7839 }, { "epoch": 2.5504229017566686, "grad_norm": 2.695066871514317, "learning_rate": 6.697152602179774e-07, "loss": 0.2073, "step": 7840 }, { "epoch": 2.5507482108002604, "grad_norm": 2.174673344323295, "learning_rate": 6.687692970583803e-07, "loss": 0.1697, "step": 7841 }, { "epoch": 2.5510735198438517, "grad_norm": 2.3313151288011946, "learning_rate": 6.67823954567029e-07, "loss": 0.201, "step": 7842 }, { "epoch": 2.551398828887443, "grad_norm": 1.7644939385246048, "learning_rate": 6.66879232879395e-07, "loss": 0.1027, "step": 7843 }, { "epoch": 2.5517241379310347, "grad_norm": 2.076122158276892, "learning_rate": 6.65935132130855e-07, "loss": 0.1779, "step": 7844 }, { "epoch": 2.552049446974626, "grad_norm": 1.9585944093567664, "learning_rate": 6.649916524566996e-07, "loss": 0.1884, "step": 7845 }, { "epoch": 2.5523747560182173, "grad_norm": 2.307075500632241, "learning_rate": 6.64048793992132e-07, "loss": 0.2092, "step": 7846 }, { "epoch": 2.5527000650618086, "grad_norm": 2.264083267187968, "learning_rate": 6.631065568722633e-07, "loss": 0.2157, "step": 7847 }, { "epoch": 2.5530253741054, "grad_norm": 2.2066001746162986, "learning_rate": 6.621649412321146e-07, "loss": 0.1905, "step": 7848 }, { "epoch": 2.5533506831489916, "grad_norm": 1.8245343817427455, "learning_rate": 6.612239472066234e-07, "loss": 0.1325, "step": 7849 }, { "epoch": 2.553675992192583, "grad_norm": 2.5580703085954233, "learning_rate": 6.602835749306341e-07, "loss": 0.2082, "step": 7850 }, { "epoch": 2.554001301236174, "grad_norm": 2.136540789497675, "learning_rate": 6.593438245389006e-07, "loss": 0.1773, "step": 7851 }, { "epoch": 2.554326610279766, "grad_norm": 1.7242231810132675, "learning_rate": 6.584046961660911e-07, "loss": 0.1211, "step": 7852 }, { "epoch": 2.5546519193233572, "grad_norm": 2.136406295262086, "learning_rate": 6.574661899467843e-07, "loss": 0.188, "step": 7853 }, { "epoch": 2.5549772283669485, "grad_norm": 1.8639456456306174, "learning_rate": 6.565283060154665e-07, "loss": 0.1469, "step": 7854 }, { "epoch": 2.5553025374105403, "grad_norm": 2.279560421685517, "learning_rate": 6.555910445065394e-07, "loss": 0.1863, "step": 7855 }, { "epoch": 2.5556278464541315, "grad_norm": 1.9906553524499861, "learning_rate": 6.546544055543113e-07, "loss": 0.1504, "step": 7856 }, { "epoch": 2.555953155497723, "grad_norm": 2.0640475713600455, "learning_rate": 6.537183892930044e-07, "loss": 0.1739, "step": 7857 }, { "epoch": 2.556278464541314, "grad_norm": 1.9351872211225185, "learning_rate": 6.527829958567511e-07, "loss": 0.1211, "step": 7858 }, { "epoch": 2.5566037735849054, "grad_norm": 1.9512459189794855, "learning_rate": 6.518482253795915e-07, "loss": 0.15, "step": 7859 }, { "epoch": 2.556929082628497, "grad_norm": 1.9149368937060955, "learning_rate": 6.509140779954804e-07, "loss": 0.1455, "step": 7860 }, { "epoch": 2.5572543916720885, "grad_norm": 1.9201331652297609, "learning_rate": 6.499805538382826e-07, "loss": 0.1408, "step": 7861 }, { "epoch": 2.5575797007156797, "grad_norm": 2.478814259795441, "learning_rate": 6.490476530417717e-07, "loss": 0.1955, "step": 7862 }, { "epoch": 2.5579050097592715, "grad_norm": 2.1508862627394123, "learning_rate": 6.481153757396319e-07, "loss": 0.1838, "step": 7863 }, { "epoch": 2.558230318802863, "grad_norm": 1.9747505970082624, "learning_rate": 6.471837220654604e-07, "loss": 0.1696, "step": 7864 }, { "epoch": 2.558555627846454, "grad_norm": 2.1669161925148805, "learning_rate": 6.462526921527651e-07, "loss": 0.1544, "step": 7865 }, { "epoch": 2.558880936890046, "grad_norm": 2.0945141402508294, "learning_rate": 6.4532228613496e-07, "loss": 0.1721, "step": 7866 }, { "epoch": 2.559206245933637, "grad_norm": 2.005028288324547, "learning_rate": 6.443925041453763e-07, "loss": 0.1233, "step": 7867 }, { "epoch": 2.5595315549772284, "grad_norm": 2.413714662587258, "learning_rate": 6.434633463172496e-07, "loss": 0.1733, "step": 7868 }, { "epoch": 2.5598568640208197, "grad_norm": 2.1565722965346423, "learning_rate": 6.4253481278373e-07, "loss": 0.1737, "step": 7869 }, { "epoch": 2.560182173064411, "grad_norm": 2.056764834859718, "learning_rate": 6.416069036778777e-07, "loss": 0.1626, "step": 7870 }, { "epoch": 2.5605074821080027, "grad_norm": 1.9357950488446072, "learning_rate": 6.406796191326614e-07, "loss": 0.1507, "step": 7871 }, { "epoch": 2.560832791151594, "grad_norm": 2.152415296673976, "learning_rate": 6.397529592809615e-07, "loss": 0.1376, "step": 7872 }, { "epoch": 2.5611581001951853, "grad_norm": 2.0642880494177462, "learning_rate": 6.388269242555717e-07, "loss": 0.1532, "step": 7873 }, { "epoch": 2.561483409238777, "grad_norm": 2.1861235590874646, "learning_rate": 6.379015141891892e-07, "loss": 0.1302, "step": 7874 }, { "epoch": 2.5618087182823683, "grad_norm": 2.162564170759231, "learning_rate": 6.369767292144302e-07, "loss": 0.1566, "step": 7875 }, { "epoch": 2.5621340273259596, "grad_norm": 2.241745053857409, "learning_rate": 6.360525694638131e-07, "loss": 0.1303, "step": 7876 }, { "epoch": 2.562459336369551, "grad_norm": 2.336894064305248, "learning_rate": 6.351290350697725e-07, "loss": 0.1922, "step": 7877 }, { "epoch": 2.562784645413142, "grad_norm": 2.0842336561210115, "learning_rate": 6.342061261646526e-07, "loss": 0.1452, "step": 7878 }, { "epoch": 2.563109954456734, "grad_norm": 2.140871748193791, "learning_rate": 6.332838428807053e-07, "loss": 0.1601, "step": 7879 }, { "epoch": 2.5634352635003252, "grad_norm": 1.7887114239297401, "learning_rate": 6.323621853500927e-07, "loss": 0.1338, "step": 7880 }, { "epoch": 2.5637605725439165, "grad_norm": 2.29511776483298, "learning_rate": 6.314411537048931e-07, "loss": 0.1573, "step": 7881 }, { "epoch": 2.5640858815875083, "grad_norm": 2.0457227248978143, "learning_rate": 6.305207480770892e-07, "loss": 0.1595, "step": 7882 }, { "epoch": 2.5644111906310996, "grad_norm": 1.9702293617863615, "learning_rate": 6.296009685985738e-07, "loss": 0.1454, "step": 7883 }, { "epoch": 2.564736499674691, "grad_norm": 1.9375513046546309, "learning_rate": 6.286818154011536e-07, "loss": 0.1661, "step": 7884 }, { "epoch": 2.5650618087182826, "grad_norm": 2.53255456680854, "learning_rate": 6.277632886165452e-07, "loss": 0.1418, "step": 7885 }, { "epoch": 2.565387117761874, "grad_norm": 2.219071129581694, "learning_rate": 6.268453883763714e-07, "loss": 0.1904, "step": 7886 }, { "epoch": 2.565712426805465, "grad_norm": 2.30195116634075, "learning_rate": 6.259281148121698e-07, "loss": 0.1687, "step": 7887 }, { "epoch": 2.5660377358490565, "grad_norm": 2.145555430939364, "learning_rate": 6.250114680553848e-07, "loss": 0.1809, "step": 7888 }, { "epoch": 2.5663630448926478, "grad_norm": 2.4640915138556756, "learning_rate": 6.240954482373735e-07, "loss": 0.2016, "step": 7889 }, { "epoch": 2.5666883539362395, "grad_norm": 2.4388384524830076, "learning_rate": 6.231800554894029e-07, "loss": 0.1757, "step": 7890 }, { "epoch": 2.567013662979831, "grad_norm": 2.011205480560793, "learning_rate": 6.222652899426479e-07, "loss": 0.1114, "step": 7891 }, { "epoch": 2.567338972023422, "grad_norm": 2.002486586851063, "learning_rate": 6.21351151728195e-07, "loss": 0.1334, "step": 7892 }, { "epoch": 2.567664281067014, "grad_norm": 2.1082730652064727, "learning_rate": 6.20437640977043e-07, "loss": 0.1819, "step": 7893 }, { "epoch": 2.567989590110605, "grad_norm": 2.22999003067929, "learning_rate": 6.195247578200958e-07, "loss": 0.2163, "step": 7894 }, { "epoch": 2.5683148991541964, "grad_norm": 1.889824724892978, "learning_rate": 6.186125023881723e-07, "loss": 0.1501, "step": 7895 }, { "epoch": 2.568640208197788, "grad_norm": 2.0889918398097946, "learning_rate": 6.177008748119978e-07, "loss": 0.1325, "step": 7896 }, { "epoch": 2.5689655172413794, "grad_norm": 1.8721381425827295, "learning_rate": 6.16789875222209e-07, "loss": 0.1369, "step": 7897 }, { "epoch": 2.5692908262849707, "grad_norm": 2.1684253046404156, "learning_rate": 6.158795037493553e-07, "loss": 0.1139, "step": 7898 }, { "epoch": 2.569616135328562, "grad_norm": 1.7322295963578513, "learning_rate": 6.149697605238914e-07, "loss": 0.1183, "step": 7899 }, { "epoch": 2.5699414443721533, "grad_norm": 2.2091058744428484, "learning_rate": 6.140606456761822e-07, "loss": 0.1486, "step": 7900 }, { "epoch": 2.570266753415745, "grad_norm": 2.340153255803041, "learning_rate": 6.131521593365091e-07, "loss": 0.1851, "step": 7901 }, { "epoch": 2.5705920624593364, "grad_norm": 1.7453959825279455, "learning_rate": 6.12244301635056e-07, "loss": 0.0969, "step": 7902 }, { "epoch": 2.5709173715029277, "grad_norm": 2.184645601752594, "learning_rate": 6.113370727019191e-07, "loss": 0.1462, "step": 7903 }, { "epoch": 2.5712426805465194, "grad_norm": 2.3667503895245714, "learning_rate": 6.104304726671062e-07, "loss": 0.1897, "step": 7904 }, { "epoch": 2.5715679895901107, "grad_norm": 2.60276143962546, "learning_rate": 6.095245016605333e-07, "loss": 0.1808, "step": 7905 }, { "epoch": 2.571893298633702, "grad_norm": 2.0569731768704784, "learning_rate": 6.086191598120262e-07, "loss": 0.1265, "step": 7906 }, { "epoch": 2.5722186076772933, "grad_norm": 2.172236694039555, "learning_rate": 6.077144472513224e-07, "loss": 0.1826, "step": 7907 }, { "epoch": 2.572543916720885, "grad_norm": 2.1452391741133723, "learning_rate": 6.068103641080658e-07, "loss": 0.1638, "step": 7908 }, { "epoch": 2.5728692257644763, "grad_norm": 2.259053347323247, "learning_rate": 6.059069105118126e-07, "loss": 0.1768, "step": 7909 }, { "epoch": 2.5731945348080676, "grad_norm": 2.4776269540020532, "learning_rate": 6.050040865920303e-07, "loss": 0.1802, "step": 7910 }, { "epoch": 2.573519843851659, "grad_norm": 2.4898954052148716, "learning_rate": 6.041018924780922e-07, "loss": 0.2072, "step": 7911 }, { "epoch": 2.5738451528952506, "grad_norm": 2.0753884213546745, "learning_rate": 6.032003282992821e-07, "loss": 0.1077, "step": 7912 }, { "epoch": 2.574170461938842, "grad_norm": 2.1548049143152808, "learning_rate": 6.022993941847987e-07, "loss": 0.1753, "step": 7913 }, { "epoch": 2.574495770982433, "grad_norm": 2.1169884218231183, "learning_rate": 6.01399090263744e-07, "loss": 0.1635, "step": 7914 }, { "epoch": 2.574821080026025, "grad_norm": 2.4756655075977196, "learning_rate": 6.004994166651307e-07, "loss": 0.1464, "step": 7915 }, { "epoch": 2.5751463890696162, "grad_norm": 2.097544862355446, "learning_rate": 5.996003735178852e-07, "loss": 0.1754, "step": 7916 }, { "epoch": 2.5754716981132075, "grad_norm": 1.863903610592742, "learning_rate": 5.987019609508404e-07, "loss": 0.1395, "step": 7917 }, { "epoch": 2.575797007156799, "grad_norm": 1.8490801289675713, "learning_rate": 5.978041790927386e-07, "loss": 0.1376, "step": 7918 }, { "epoch": 2.57612231620039, "grad_norm": 2.362560825997169, "learning_rate": 5.969070280722333e-07, "loss": 0.1889, "step": 7919 }, { "epoch": 2.576447625243982, "grad_norm": 2.300641757687597, "learning_rate": 5.960105080178863e-07, "loss": 0.1472, "step": 7920 }, { "epoch": 2.576772934287573, "grad_norm": 2.176636479762652, "learning_rate": 5.951146190581692e-07, "loss": 0.1366, "step": 7921 }, { "epoch": 2.5770982433311644, "grad_norm": 2.512226806549765, "learning_rate": 5.942193613214653e-07, "loss": 0.1703, "step": 7922 }, { "epoch": 2.577423552374756, "grad_norm": 1.8865617022929722, "learning_rate": 5.933247349360632e-07, "loss": 0.1106, "step": 7923 }, { "epoch": 2.5777488614183475, "grad_norm": 2.23325468358459, "learning_rate": 5.924307400301649e-07, "loss": 0.1719, "step": 7924 }, { "epoch": 2.5780741704619388, "grad_norm": 2.1303340732433536, "learning_rate": 5.915373767318811e-07, "loss": 0.1901, "step": 7925 }, { "epoch": 2.5783994795055305, "grad_norm": 2.042065534299938, "learning_rate": 5.906446451692299e-07, "loss": 0.1943, "step": 7926 }, { "epoch": 2.578724788549122, "grad_norm": 1.7711535947893535, "learning_rate": 5.897525454701414e-07, "loss": 0.1335, "step": 7927 }, { "epoch": 2.579050097592713, "grad_norm": 2.148616096768365, "learning_rate": 5.888610777624526e-07, "loss": 0.1378, "step": 7928 }, { "epoch": 2.5793754066363044, "grad_norm": 2.38889965931273, "learning_rate": 5.879702421739126e-07, "loss": 0.2024, "step": 7929 }, { "epoch": 2.5797007156798957, "grad_norm": 2.375333684328983, "learning_rate": 5.87080038832179e-07, "loss": 0.2026, "step": 7930 }, { "epoch": 2.5800260247234874, "grad_norm": 1.8048406699223911, "learning_rate": 5.861904678648189e-07, "loss": 0.1551, "step": 7931 }, { "epoch": 2.5803513337670787, "grad_norm": 1.8674110923452967, "learning_rate": 5.853015293993047e-07, "loss": 0.1199, "step": 7932 }, { "epoch": 2.58067664281067, "grad_norm": 1.9455893942336888, "learning_rate": 5.844132235630273e-07, "loss": 0.1683, "step": 7933 }, { "epoch": 2.5810019518542617, "grad_norm": 2.1850467727957583, "learning_rate": 5.835255504832787e-07, "loss": 0.1851, "step": 7934 }, { "epoch": 2.581327260897853, "grad_norm": 2.262912459877781, "learning_rate": 5.826385102872628e-07, "loss": 0.1537, "step": 7935 }, { "epoch": 2.5816525699414443, "grad_norm": 2.161080329496043, "learning_rate": 5.817521031020929e-07, "loss": 0.1765, "step": 7936 }, { "epoch": 2.581977878985036, "grad_norm": 2.3908866551975305, "learning_rate": 5.808663290547939e-07, "loss": 0.1906, "step": 7937 }, { "epoch": 2.5823031880286273, "grad_norm": 1.9504361377077277, "learning_rate": 5.799811882722949e-07, "loss": 0.1621, "step": 7938 }, { "epoch": 2.5826284970722186, "grad_norm": 2.1872225653768536, "learning_rate": 5.790966808814402e-07, "loss": 0.1888, "step": 7939 }, { "epoch": 2.58295380611581, "grad_norm": 1.8794240467975645, "learning_rate": 5.782128070089771e-07, "loss": 0.1439, "step": 7940 }, { "epoch": 2.5832791151594012, "grad_norm": 2.9571201752991265, "learning_rate": 5.77329566781567e-07, "loss": 0.1801, "step": 7941 }, { "epoch": 2.583604424202993, "grad_norm": 1.9891743845057563, "learning_rate": 5.7644696032578e-07, "loss": 0.1503, "step": 7942 }, { "epoch": 2.5839297332465843, "grad_norm": 1.875588263824523, "learning_rate": 5.755649877680914e-07, "loss": 0.1508, "step": 7943 }, { "epoch": 2.5842550422901756, "grad_norm": 2.015357263660724, "learning_rate": 5.7468364923489e-07, "loss": 0.1743, "step": 7944 }, { "epoch": 2.5845803513337673, "grad_norm": 2.4743091998557993, "learning_rate": 5.738029448524734e-07, "loss": 0.1643, "step": 7945 }, { "epoch": 2.5849056603773586, "grad_norm": 2.0939227642889198, "learning_rate": 5.72922874747045e-07, "loss": 0.1646, "step": 7946 }, { "epoch": 2.58523096942095, "grad_norm": 2.55811426600267, "learning_rate": 5.720434390447205e-07, "loss": 0.2605, "step": 7947 }, { "epoch": 2.585556278464541, "grad_norm": 2.1337512826743423, "learning_rate": 5.711646378715241e-07, "loss": 0.1294, "step": 7948 }, { "epoch": 2.5858815875081325, "grad_norm": 2.020862955174422, "learning_rate": 5.702864713533879e-07, "loss": 0.1448, "step": 7949 }, { "epoch": 2.586206896551724, "grad_norm": 2.289452371976719, "learning_rate": 5.694089396161529e-07, "loss": 0.1893, "step": 7950 }, { "epoch": 2.5865322055953155, "grad_norm": 2.3451121050393886, "learning_rate": 5.685320427855707e-07, "loss": 0.1809, "step": 7951 }, { "epoch": 2.586857514638907, "grad_norm": 2.4139209041889265, "learning_rate": 5.676557809873023e-07, "loss": 0.1888, "step": 7952 }, { "epoch": 2.5871828236824985, "grad_norm": 2.156378702857055, "learning_rate": 5.667801543469148e-07, "loss": 0.1226, "step": 7953 }, { "epoch": 2.58750813272609, "grad_norm": 2.6735922297140746, "learning_rate": 5.659051629898876e-07, "loss": 0.1598, "step": 7954 }, { "epoch": 2.587833441769681, "grad_norm": 1.9784665783315918, "learning_rate": 5.650308070416066e-07, "loss": 0.1564, "step": 7955 }, { "epoch": 2.588158750813273, "grad_norm": 1.900575953737884, "learning_rate": 5.641570866273671e-07, "loss": 0.1386, "step": 7956 }, { "epoch": 2.588484059856864, "grad_norm": 2.3020314787392464, "learning_rate": 5.632840018723762e-07, "loss": 0.1584, "step": 7957 }, { "epoch": 2.5888093689004554, "grad_norm": 2.425888904814928, "learning_rate": 5.624115529017449e-07, "loss": 0.1931, "step": 7958 }, { "epoch": 2.5891346779440467, "grad_norm": 2.4193921105753016, "learning_rate": 5.615397398404965e-07, "loss": 0.1919, "step": 7959 }, { "epoch": 2.589459986987638, "grad_norm": 2.4579511058219854, "learning_rate": 5.606685628135633e-07, "loss": 0.1453, "step": 7960 }, { "epoch": 2.5897852960312298, "grad_norm": 1.8592166476237875, "learning_rate": 5.597980219457843e-07, "loss": 0.1295, "step": 7961 }, { "epoch": 2.590110605074821, "grad_norm": 1.8463926476027688, "learning_rate": 5.589281173619105e-07, "loss": 0.1073, "step": 7962 }, { "epoch": 2.5904359141184123, "grad_norm": 1.9423492324057157, "learning_rate": 5.580588491865968e-07, "loss": 0.1549, "step": 7963 }, { "epoch": 2.590761223162004, "grad_norm": 2.163117467154332, "learning_rate": 5.571902175444121e-07, "loss": 0.1542, "step": 7964 }, { "epoch": 2.5910865322055954, "grad_norm": 1.8687474808025635, "learning_rate": 5.563222225598319e-07, "loss": 0.1361, "step": 7965 }, { "epoch": 2.5914118412491867, "grad_norm": 2.3931490870766394, "learning_rate": 5.554548643572405e-07, "loss": 0.1741, "step": 7966 }, { "epoch": 2.5917371502927784, "grad_norm": 2.6594522628956816, "learning_rate": 5.545881430609279e-07, "loss": 0.2181, "step": 7967 }, { "epoch": 2.5920624593363697, "grad_norm": 2.264731118955641, "learning_rate": 5.537220587951004e-07, "loss": 0.2069, "step": 7968 }, { "epoch": 2.592387768379961, "grad_norm": 2.3669276145468583, "learning_rate": 5.528566116838663e-07, "loss": 0.1974, "step": 7969 }, { "epoch": 2.5927130774235523, "grad_norm": 2.370362667166522, "learning_rate": 5.519918018512438e-07, "loss": 0.2237, "step": 7970 }, { "epoch": 2.5930383864671436, "grad_norm": 2.240786753985235, "learning_rate": 5.51127629421162e-07, "loss": 0.2029, "step": 7971 }, { "epoch": 2.5933636955107353, "grad_norm": 1.9986033479051506, "learning_rate": 5.502640945174581e-07, "loss": 0.1467, "step": 7972 }, { "epoch": 2.5936890045543266, "grad_norm": 1.9556548310075026, "learning_rate": 5.494011972638747e-07, "loss": 0.1549, "step": 7973 }, { "epoch": 2.594014313597918, "grad_norm": 1.9799071231289183, "learning_rate": 5.485389377840684e-07, "loss": 0.1396, "step": 7974 }, { "epoch": 2.5943396226415096, "grad_norm": 2.2646246610643566, "learning_rate": 5.476773162015992e-07, "loss": 0.1624, "step": 7975 }, { "epoch": 2.594664931685101, "grad_norm": 1.789669016148795, "learning_rate": 5.46816332639939e-07, "loss": 0.1394, "step": 7976 }, { "epoch": 2.594990240728692, "grad_norm": 1.8246273989594497, "learning_rate": 5.459559872224685e-07, "loss": 0.1237, "step": 7977 }, { "epoch": 2.5953155497722835, "grad_norm": 1.9790277303213473, "learning_rate": 5.450962800724741e-07, "loss": 0.16, "step": 7978 }, { "epoch": 2.595640858815875, "grad_norm": 2.2639795609192106, "learning_rate": 5.442372113131522e-07, "loss": 0.171, "step": 7979 }, { "epoch": 2.5959661678594665, "grad_norm": 2.8426949589234125, "learning_rate": 5.4337878106761e-07, "loss": 0.2143, "step": 7980 }, { "epoch": 2.596291476903058, "grad_norm": 2.386095296654944, "learning_rate": 5.42520989458859e-07, "loss": 0.1785, "step": 7981 }, { "epoch": 2.596616785946649, "grad_norm": 2.076537166740399, "learning_rate": 5.416638366098231e-07, "loss": 0.1523, "step": 7982 }, { "epoch": 2.596942094990241, "grad_norm": 2.3760301439353224, "learning_rate": 5.408073226433308e-07, "loss": 0.1536, "step": 7983 }, { "epoch": 2.597267404033832, "grad_norm": 2.6064278901822346, "learning_rate": 5.399514476821222e-07, "loss": 0.2618, "step": 7984 }, { "epoch": 2.5975927130774235, "grad_norm": 2.00957251043165, "learning_rate": 5.390962118488458e-07, "loss": 0.1201, "step": 7985 }, { "epoch": 2.597918022121015, "grad_norm": 2.057392257369431, "learning_rate": 5.382416152660569e-07, "loss": 0.1645, "step": 7986 }, { "epoch": 2.5982433311646065, "grad_norm": 2.1217556990761715, "learning_rate": 5.373876580562171e-07, "loss": 0.1445, "step": 7987 }, { "epoch": 2.5985686402081978, "grad_norm": 2.4765490506244676, "learning_rate": 5.365343403417029e-07, "loss": 0.1747, "step": 7988 }, { "epoch": 2.598893949251789, "grad_norm": 1.930558800449189, "learning_rate": 5.356816622447936e-07, "loss": 0.1614, "step": 7989 }, { "epoch": 2.5992192582953804, "grad_norm": 2.0866383116699736, "learning_rate": 5.348296238876782e-07, "loss": 0.1223, "step": 7990 }, { "epoch": 2.599544567338972, "grad_norm": 2.3227734713036883, "learning_rate": 5.339782253924542e-07, "loss": 0.228, "step": 7991 }, { "epoch": 2.5998698763825634, "grad_norm": 2.1554960318283944, "learning_rate": 5.331274668811292e-07, "loss": 0.1508, "step": 7992 }, { "epoch": 2.6001951854261547, "grad_norm": 2.058899169126284, "learning_rate": 5.322773484756161e-07, "loss": 0.1339, "step": 7993 }, { "epoch": 2.6005204944697464, "grad_norm": 1.987754865055522, "learning_rate": 5.314278702977377e-07, "loss": 0.1505, "step": 7994 }, { "epoch": 2.6008458035133377, "grad_norm": 1.8623691138604699, "learning_rate": 5.305790324692245e-07, "loss": 0.1694, "step": 7995 }, { "epoch": 2.601171112556929, "grad_norm": 2.0501602241223758, "learning_rate": 5.297308351117158e-07, "loss": 0.1274, "step": 7996 }, { "epoch": 2.6014964216005207, "grad_norm": 2.375331083599557, "learning_rate": 5.288832783467596e-07, "loss": 0.1883, "step": 7997 }, { "epoch": 2.601821730644112, "grad_norm": 2.1397787516197404, "learning_rate": 5.280363622958106e-07, "loss": 0.1843, "step": 7998 }, { "epoch": 2.6021470396877033, "grad_norm": 1.9942055774379985, "learning_rate": 5.271900870802305e-07, "loss": 0.1129, "step": 7999 }, { "epoch": 2.6024723487312946, "grad_norm": 1.930858530117592, "learning_rate": 5.263444528212952e-07, "loss": 0.1793, "step": 8000 }, { "epoch": 2.602797657774886, "grad_norm": 2.3743702476131543, "learning_rate": 5.25499459640183e-07, "loss": 0.1834, "step": 8001 }, { "epoch": 2.6031229668184777, "grad_norm": 2.148104634577576, "learning_rate": 5.246551076579798e-07, "loss": 0.1315, "step": 8002 }, { "epoch": 2.603448275862069, "grad_norm": 2.2829453719925343, "learning_rate": 5.238113969956837e-07, "loss": 0.1944, "step": 8003 }, { "epoch": 2.6037735849056602, "grad_norm": 2.0584950209249753, "learning_rate": 5.229683277741999e-07, "loss": 0.1496, "step": 8004 }, { "epoch": 2.604098893949252, "grad_norm": 2.05262303254131, "learning_rate": 5.221259001143386e-07, "loss": 0.1836, "step": 8005 }, { "epoch": 2.6044242029928433, "grad_norm": 2.067762935961767, "learning_rate": 5.212841141368225e-07, "loss": 0.1469, "step": 8006 }, { "epoch": 2.6047495120364346, "grad_norm": 1.8651543926040208, "learning_rate": 5.204429699622782e-07, "loss": 0.1243, "step": 8007 }, { "epoch": 2.605074821080026, "grad_norm": 1.7987618966862886, "learning_rate": 5.196024677112422e-07, "loss": 0.1553, "step": 8008 }, { "epoch": 2.6054001301236176, "grad_norm": 2.0150022466900253, "learning_rate": 5.187626075041613e-07, "loss": 0.148, "step": 8009 }, { "epoch": 2.605725439167209, "grad_norm": 2.116142266172272, "learning_rate": 5.179233894613855e-07, "loss": 0.2165, "step": 8010 }, { "epoch": 2.6060507482108, "grad_norm": 2.273804132615002, "learning_rate": 5.170848137031764e-07, "loss": 0.183, "step": 8011 }, { "epoch": 2.6063760572543915, "grad_norm": 1.8666707243251401, "learning_rate": 5.162468803497034e-07, "loss": 0.1222, "step": 8012 }, { "epoch": 2.606701366297983, "grad_norm": 1.843528236497049, "learning_rate": 5.154095895210404e-07, "loss": 0.1164, "step": 8013 }, { "epoch": 2.6070266753415745, "grad_norm": 2.057162436625234, "learning_rate": 5.145729413371747e-07, "loss": 0.1622, "step": 8014 }, { "epoch": 2.607351984385166, "grad_norm": 1.8318654143395459, "learning_rate": 5.137369359179961e-07, "loss": 0.1606, "step": 8015 }, { "epoch": 2.6076772934287575, "grad_norm": 2.1303142850385073, "learning_rate": 5.129015733833054e-07, "loss": 0.1903, "step": 8016 }, { "epoch": 2.608002602472349, "grad_norm": 2.1060585250800328, "learning_rate": 5.120668538528117e-07, "loss": 0.1697, "step": 8017 }, { "epoch": 2.60832791151594, "grad_norm": 1.834634978809684, "learning_rate": 5.112327774461306e-07, "loss": 0.1655, "step": 8018 }, { "epoch": 2.6086532205595314, "grad_norm": 1.9392521706084314, "learning_rate": 5.103993442827832e-07, "loss": 0.1362, "step": 8019 }, { "epoch": 2.6089785296031227, "grad_norm": 1.7226975598927807, "learning_rate": 5.095665544822053e-07, "loss": 0.1175, "step": 8020 }, { "epoch": 2.6093038386467144, "grad_norm": 1.7901286172201327, "learning_rate": 5.087344081637336e-07, "loss": 0.1676, "step": 8021 }, { "epoch": 2.6096291476903057, "grad_norm": 2.396789721459113, "learning_rate": 5.079029054466151e-07, "loss": 0.1754, "step": 8022 }, { "epoch": 2.609954456733897, "grad_norm": 2.2010589887383922, "learning_rate": 5.070720464500051e-07, "loss": 0.161, "step": 8023 }, { "epoch": 2.6102797657774888, "grad_norm": 2.1351281036159735, "learning_rate": 5.062418312929668e-07, "loss": 0.148, "step": 8024 }, { "epoch": 2.61060507482108, "grad_norm": 1.763855600083059, "learning_rate": 5.0541226009447e-07, "loss": 0.1062, "step": 8025 }, { "epoch": 2.6109303838646714, "grad_norm": 1.9057396339723254, "learning_rate": 5.045833329733934e-07, "loss": 0.1279, "step": 8026 }, { "epoch": 2.611255692908263, "grad_norm": 2.1401697789876506, "learning_rate": 5.037550500485222e-07, "loss": 0.169, "step": 8027 }, { "epoch": 2.6115810019518544, "grad_norm": 2.154586689112051, "learning_rate": 5.029274114385496e-07, "loss": 0.1307, "step": 8028 }, { "epoch": 2.6119063109954457, "grad_norm": 2.2812420190530314, "learning_rate": 5.02100417262078e-07, "loss": 0.1803, "step": 8029 }, { "epoch": 2.612231620039037, "grad_norm": 2.0358116792706684, "learning_rate": 5.012740676376149e-07, "loss": 0.1476, "step": 8030 }, { "epoch": 2.6125569290826283, "grad_norm": 2.2200435749739715, "learning_rate": 5.004483626835776e-07, "loss": 0.1479, "step": 8031 }, { "epoch": 2.61288223812622, "grad_norm": 2.552689213288422, "learning_rate": 4.996233025182907e-07, "loss": 0.1816, "step": 8032 }, { "epoch": 2.6132075471698113, "grad_norm": 2.179298455657177, "learning_rate": 4.987988872599836e-07, "loss": 0.1553, "step": 8033 }, { "epoch": 2.6135328562134026, "grad_norm": 2.268486702292326, "learning_rate": 4.979751170267982e-07, "loss": 0.167, "step": 8034 }, { "epoch": 2.6138581652569943, "grad_norm": 2.021953428662871, "learning_rate": 4.971519919367796e-07, "loss": 0.1454, "step": 8035 }, { "epoch": 2.6141834743005856, "grad_norm": 1.6814747254921745, "learning_rate": 4.963295121078838e-07, "loss": 0.1137, "step": 8036 }, { "epoch": 2.614508783344177, "grad_norm": 2.0297415631665325, "learning_rate": 4.955076776579699e-07, "loss": 0.1468, "step": 8037 }, { "epoch": 2.6148340923877686, "grad_norm": 2.209207706826224, "learning_rate": 4.946864887048108e-07, "loss": 0.2028, "step": 8038 }, { "epoch": 2.61515940143136, "grad_norm": 2.365846553194101, "learning_rate": 4.938659453660804e-07, "loss": 0.2398, "step": 8039 }, { "epoch": 2.6154847104749512, "grad_norm": 1.9889734288161915, "learning_rate": 4.93046047759364e-07, "loss": 0.134, "step": 8040 }, { "epoch": 2.6158100195185425, "grad_norm": 2.145500630395569, "learning_rate": 4.922267960021554e-07, "loss": 0.1626, "step": 8041 }, { "epoch": 2.616135328562134, "grad_norm": 2.130994363716853, "learning_rate": 4.914081902118511e-07, "loss": 0.1385, "step": 8042 }, { "epoch": 2.6164606376057256, "grad_norm": 2.16511826864423, "learning_rate": 4.905902305057597e-07, "loss": 0.1413, "step": 8043 }, { "epoch": 2.616785946649317, "grad_norm": 1.6590353920682073, "learning_rate": 4.897729170010951e-07, "loss": 0.1178, "step": 8044 }, { "epoch": 2.617111255692908, "grad_norm": 2.3514771479349412, "learning_rate": 4.88956249814978e-07, "loss": 0.1897, "step": 8045 }, { "epoch": 2.6174365647365, "grad_norm": 1.9398130339164366, "learning_rate": 4.881402290644388e-07, "loss": 0.13, "step": 8046 }, { "epoch": 2.617761873780091, "grad_norm": 2.237174556114247, "learning_rate": 4.873248548664122e-07, "loss": 0.2067, "step": 8047 }, { "epoch": 2.6180871828236825, "grad_norm": 2.310606054080708, "learning_rate": 4.865101273377427e-07, "loss": 0.1796, "step": 8048 }, { "epoch": 2.6184124918672738, "grad_norm": 2.2187190278724835, "learning_rate": 4.856960465951821e-07, "loss": 0.2058, "step": 8049 }, { "epoch": 2.618737800910865, "grad_norm": 2.1042979599532354, "learning_rate": 4.848826127553874e-07, "loss": 0.1559, "step": 8050 }, { "epoch": 2.619063109954457, "grad_norm": 1.7080169673797552, "learning_rate": 4.840698259349231e-07, "loss": 0.1302, "step": 8051 }, { "epoch": 2.619388418998048, "grad_norm": 2.1226975236183523, "learning_rate": 4.832576862502659e-07, "loss": 0.147, "step": 8052 }, { "epoch": 2.6197137280416394, "grad_norm": 1.6396721588832794, "learning_rate": 4.824461938177933e-07, "loss": 0.1089, "step": 8053 }, { "epoch": 2.620039037085231, "grad_norm": 2.188063398136882, "learning_rate": 4.816353487537923e-07, "loss": 0.1433, "step": 8054 }, { "epoch": 2.6203643461288224, "grad_norm": 2.314565155354435, "learning_rate": 4.808251511744583e-07, "loss": 0.1701, "step": 8055 }, { "epoch": 2.6206896551724137, "grad_norm": 1.9761651479947364, "learning_rate": 4.80015601195894e-07, "loss": 0.123, "step": 8056 }, { "epoch": 2.6210149642160054, "grad_norm": 2.0347096825827626, "learning_rate": 4.792066989341065e-07, "loss": 0.1254, "step": 8057 }, { "epoch": 2.6213402732595967, "grad_norm": 2.002618667594055, "learning_rate": 4.783984445050144e-07, "loss": 0.1385, "step": 8058 }, { "epoch": 2.621665582303188, "grad_norm": 2.1730789205115233, "learning_rate": 4.775908380244393e-07, "loss": 0.1191, "step": 8059 }, { "epoch": 2.6219908913467793, "grad_norm": 2.1636737894571296, "learning_rate": 4.76783879608112e-07, "loss": 0.1568, "step": 8060 }, { "epoch": 2.6223162003903706, "grad_norm": 2.2887548792103685, "learning_rate": 4.7597756937167115e-07, "loss": 0.1247, "step": 8061 }, { "epoch": 2.6226415094339623, "grad_norm": 1.8086979711650464, "learning_rate": 4.751719074306604e-07, "loss": 0.1169, "step": 8062 }, { "epoch": 2.6229668184775536, "grad_norm": 2.350073797728511, "learning_rate": 4.743668939005325e-07, "loss": 0.1474, "step": 8063 }, { "epoch": 2.623292127521145, "grad_norm": 2.1853521191340475, "learning_rate": 4.7356252889664613e-07, "loss": 0.1704, "step": 8064 }, { "epoch": 2.6236174365647367, "grad_norm": 1.8145643500639916, "learning_rate": 4.727588125342669e-07, "loss": 0.119, "step": 8065 }, { "epoch": 2.623942745608328, "grad_norm": 2.2308675441136603, "learning_rate": 4.719557449285689e-07, "loss": 0.1705, "step": 8066 }, { "epoch": 2.6242680546519193, "grad_norm": 1.9939097129802086, "learning_rate": 4.7115332619463105e-07, "loss": 0.1067, "step": 8067 }, { "epoch": 2.624593363695511, "grad_norm": 2.3107898326305962, "learning_rate": 4.703515564474409e-07, "loss": 0.1754, "step": 8068 }, { "epoch": 2.6249186727391023, "grad_norm": 1.8889290285313305, "learning_rate": 4.695504358018937e-07, "loss": 0.1404, "step": 8069 }, { "epoch": 2.6252439817826936, "grad_norm": 2.2273769621244313, "learning_rate": 4.6874996437279e-07, "loss": 0.1497, "step": 8070 }, { "epoch": 2.625569290826285, "grad_norm": 2.118520835878063, "learning_rate": 4.67950142274835e-07, "loss": 0.18, "step": 8071 }, { "epoch": 2.625894599869876, "grad_norm": 2.323460570011682, "learning_rate": 4.671509696226484e-07, "loss": 0.1865, "step": 8072 }, { "epoch": 2.626219908913468, "grad_norm": 2.3637030594874315, "learning_rate": 4.6635244653075007e-07, "loss": 0.1591, "step": 8073 }, { "epoch": 2.626545217957059, "grad_norm": 2.089092869889654, "learning_rate": 4.655545731135669e-07, "loss": 0.1249, "step": 8074 }, { "epoch": 2.6268705270006505, "grad_norm": 1.9845553644634593, "learning_rate": 4.6475734948543903e-07, "loss": 0.1288, "step": 8075 }, { "epoch": 2.6271958360442422, "grad_norm": 2.1962659532773077, "learning_rate": 4.639607757606063e-07, "loss": 0.1674, "step": 8076 }, { "epoch": 2.6275211450878335, "grad_norm": 2.882160556159379, "learning_rate": 4.631648520532178e-07, "loss": 0.2323, "step": 8077 }, { "epoch": 2.627846454131425, "grad_norm": 1.6346416466754885, "learning_rate": 4.623695784773308e-07, "loss": 0.1499, "step": 8078 }, { "epoch": 2.628171763175016, "grad_norm": 2.2233329782505686, "learning_rate": 4.6157495514690955e-07, "loss": 0.1578, "step": 8079 }, { "epoch": 2.6284970722186074, "grad_norm": 2.095306737626757, "learning_rate": 4.6078098217582243e-07, "loss": 0.169, "step": 8080 }, { "epoch": 2.628822381262199, "grad_norm": 2.2630177086634937, "learning_rate": 4.5998765967784744e-07, "loss": 0.1537, "step": 8081 }, { "epoch": 2.6291476903057904, "grad_norm": 1.9442058847033585, "learning_rate": 4.5919498776666737e-07, "loss": 0.1302, "step": 8082 }, { "epoch": 2.6294729993493817, "grad_norm": 1.9515102219745786, "learning_rate": 4.5840296655587256e-07, "loss": 0.1472, "step": 8083 }, { "epoch": 2.6297983083929735, "grad_norm": 1.7148968975292518, "learning_rate": 4.5761159615896167e-07, "loss": 0.1167, "step": 8084 }, { "epoch": 2.6301236174365648, "grad_norm": 2.535575251866531, "learning_rate": 4.5682087668933784e-07, "loss": 0.1842, "step": 8085 }, { "epoch": 2.630448926480156, "grad_norm": 1.9190423688131928, "learning_rate": 4.5603080826030986e-07, "loss": 0.1322, "step": 8086 }, { "epoch": 2.630774235523748, "grad_norm": 1.9129173179016388, "learning_rate": 4.552413909850978e-07, "loss": 0.1345, "step": 8087 }, { "epoch": 2.631099544567339, "grad_norm": 1.9215136824459604, "learning_rate": 4.5445262497682496e-07, "loss": 0.1678, "step": 8088 }, { "epoch": 2.6314248536109304, "grad_norm": 1.66549736752408, "learning_rate": 4.5366451034852097e-07, "loss": 0.1057, "step": 8089 }, { "epoch": 2.6317501626545217, "grad_norm": 1.8451092625894636, "learning_rate": 4.5287704721312333e-07, "loss": 0.1191, "step": 8090 }, { "epoch": 2.632075471698113, "grad_norm": 1.9510179946762058, "learning_rate": 4.5209023568347774e-07, "loss": 0.138, "step": 8091 }, { "epoch": 2.6324007807417047, "grad_norm": 2.000726734642639, "learning_rate": 4.513040758723325e-07, "loss": 0.1409, "step": 8092 }, { "epoch": 2.632726089785296, "grad_norm": 2.53863442737284, "learning_rate": 4.505185678923468e-07, "loss": 0.1253, "step": 8093 }, { "epoch": 2.6330513988288873, "grad_norm": 2.3713395294679827, "learning_rate": 4.497337118560829e-07, "loss": 0.2065, "step": 8094 }, { "epoch": 2.633376707872479, "grad_norm": 2.0006948377348395, "learning_rate": 4.4894950787601144e-07, "loss": 0.1292, "step": 8095 }, { "epoch": 2.6337020169160703, "grad_norm": 2.3988945567816335, "learning_rate": 4.481659560645113e-07, "loss": 0.157, "step": 8096 }, { "epoch": 2.6340273259596616, "grad_norm": 2.4829495497241023, "learning_rate": 4.4738305653386326e-07, "loss": 0.1835, "step": 8097 }, { "epoch": 2.6343526350032533, "grad_norm": 2.244960469649787, "learning_rate": 4.466008093962587e-07, "loss": 0.2048, "step": 8098 }, { "epoch": 2.6346779440468446, "grad_norm": 2.044864555378213, "learning_rate": 4.4581921476379466e-07, "loss": 0.1483, "step": 8099 }, { "epoch": 2.635003253090436, "grad_norm": 2.349837267353878, "learning_rate": 4.450382727484731e-07, "loss": 0.1636, "step": 8100 }, { "epoch": 2.635328562134027, "grad_norm": 1.9398263182836057, "learning_rate": 4.442579834622046e-07, "loss": 0.1655, "step": 8101 }, { "epoch": 2.6356538711776185, "grad_norm": 1.9181173565125604, "learning_rate": 4.434783470168036e-07, "loss": 0.1328, "step": 8102 }, { "epoch": 2.6359791802212102, "grad_norm": 1.868655427333823, "learning_rate": 4.4269936352399346e-07, "loss": 0.137, "step": 8103 }, { "epoch": 2.6363044892648015, "grad_norm": 2.096350378195093, "learning_rate": 4.4192103309540436e-07, "loss": 0.1395, "step": 8104 }, { "epoch": 2.636629798308393, "grad_norm": 2.20124244239614, "learning_rate": 4.4114335584256986e-07, "loss": 0.1585, "step": 8105 }, { "epoch": 2.6369551073519846, "grad_norm": 2.0299561086321813, "learning_rate": 4.4036633187693024e-07, "loss": 0.1705, "step": 8106 }, { "epoch": 2.637280416395576, "grad_norm": 2.198671156638624, "learning_rate": 4.39589961309837e-07, "loss": 0.1756, "step": 8107 }, { "epoch": 2.637605725439167, "grad_norm": 2.099243316252365, "learning_rate": 4.388142442525428e-07, "loss": 0.2001, "step": 8108 }, { "epoch": 2.637931034482759, "grad_norm": 2.084782210096592, "learning_rate": 4.380391808162077e-07, "loss": 0.1271, "step": 8109 }, { "epoch": 2.63825634352635, "grad_norm": 2.0645370010348283, "learning_rate": 4.3726477111189933e-07, "loss": 0.1697, "step": 8110 }, { "epoch": 2.6385816525699415, "grad_norm": 2.2690971609700976, "learning_rate": 4.3649101525059246e-07, "loss": 0.2216, "step": 8111 }, { "epoch": 2.6389069616135328, "grad_norm": 2.1231247823086563, "learning_rate": 4.3571791334316496e-07, "loss": 0.1529, "step": 8112 }, { "epoch": 2.639232270657124, "grad_norm": 2.1061146638659634, "learning_rate": 4.3494546550040486e-07, "loss": 0.1819, "step": 8113 }, { "epoch": 2.639557579700716, "grad_norm": 2.604503889319397, "learning_rate": 4.3417367183300143e-07, "loss": 0.2148, "step": 8114 }, { "epoch": 2.639882888744307, "grad_norm": 2.0222554131041677, "learning_rate": 4.334025324515556e-07, "loss": 0.1496, "step": 8115 }, { "epoch": 2.6402081977878984, "grad_norm": 2.236913480949328, "learning_rate": 4.3263204746657237e-07, "loss": 0.176, "step": 8116 }, { "epoch": 2.64053350683149, "grad_norm": 1.8544559184651817, "learning_rate": 4.318622169884612e-07, "loss": 0.1525, "step": 8117 }, { "epoch": 2.6408588158750814, "grad_norm": 2.076449659838032, "learning_rate": 4.310930411275399e-07, "loss": 0.1802, "step": 8118 }, { "epoch": 2.6411841249186727, "grad_norm": 2.203254186669197, "learning_rate": 4.303245199940326e-07, "loss": 0.1652, "step": 8119 }, { "epoch": 2.641509433962264, "grad_norm": 2.4543287915707945, "learning_rate": 4.295566536980678e-07, "loss": 0.1947, "step": 8120 }, { "epoch": 2.6418347430058553, "grad_norm": 1.9127663995160855, "learning_rate": 4.2878944234968247e-07, "loss": 0.1076, "step": 8121 }, { "epoch": 2.642160052049447, "grad_norm": 2.315213361093685, "learning_rate": 4.2802288605881705e-07, "loss": 0.1886, "step": 8122 }, { "epoch": 2.6424853610930383, "grad_norm": 2.1280806736623137, "learning_rate": 4.272569849353214e-07, "loss": 0.1461, "step": 8123 }, { "epoch": 2.6428106701366296, "grad_norm": 1.8426591868947382, "learning_rate": 4.2649173908894726e-07, "loss": 0.1465, "step": 8124 }, { "epoch": 2.6431359791802214, "grad_norm": 3.0571740057523686, "learning_rate": 4.2572714862935736e-07, "loss": 0.165, "step": 8125 }, { "epoch": 2.6434612882238127, "grad_norm": 2.3129372036289033, "learning_rate": 4.2496321366611584e-07, "loss": 0.1155, "step": 8126 }, { "epoch": 2.643786597267404, "grad_norm": 2.459677568025989, "learning_rate": 4.241999343086961e-07, "loss": 0.24, "step": 8127 }, { "epoch": 2.6441119063109957, "grad_norm": 2.068818992153399, "learning_rate": 4.234373106664774e-07, "loss": 0.1643, "step": 8128 }, { "epoch": 2.644437215354587, "grad_norm": 1.9027582367533327, "learning_rate": 4.226753428487429e-07, "loss": 0.1713, "step": 8129 }, { "epoch": 2.6447625243981783, "grad_norm": 2.28638252953794, "learning_rate": 4.2191403096468355e-07, "loss": 0.1497, "step": 8130 }, { "epoch": 2.6450878334417696, "grad_norm": 2.293151884517853, "learning_rate": 4.211533751233965e-07, "loss": 0.1562, "step": 8131 }, { "epoch": 2.645413142485361, "grad_norm": 2.1723944781749194, "learning_rate": 4.203933754338829e-07, "loss": 0.1437, "step": 8132 }, { "epoch": 2.6457384515289526, "grad_norm": 2.2158696690680455, "learning_rate": 4.1963403200505294e-07, "loss": 0.1759, "step": 8133 }, { "epoch": 2.646063760572544, "grad_norm": 1.6648327151377353, "learning_rate": 4.188753449457189e-07, "loss": 0.1337, "step": 8134 }, { "epoch": 2.646389069616135, "grad_norm": 2.0859320883588537, "learning_rate": 4.181173143646028e-07, "loss": 0.1675, "step": 8135 }, { "epoch": 2.646714378659727, "grad_norm": 2.4156065887874942, "learning_rate": 4.1735994037033103e-07, "loss": 0.1728, "step": 8136 }, { "epoch": 2.647039687703318, "grad_norm": 1.9139418544626807, "learning_rate": 4.1660322307143506e-07, "loss": 0.1308, "step": 8137 }, { "epoch": 2.6473649967469095, "grad_norm": 2.3237282972050917, "learning_rate": 4.158471625763516e-07, "loss": 0.1768, "step": 8138 }, { "epoch": 2.6476903057905012, "grad_norm": 1.7968671974316286, "learning_rate": 4.150917589934278e-07, "loss": 0.1211, "step": 8139 }, { "epoch": 2.6480156148340925, "grad_norm": 2.130954035193826, "learning_rate": 4.143370124309115e-07, "loss": 0.1656, "step": 8140 }, { "epoch": 2.648340923877684, "grad_norm": 2.894576172545542, "learning_rate": 4.1358292299695845e-07, "loss": 0.1708, "step": 8141 }, { "epoch": 2.648666232921275, "grad_norm": 2.6444695478569202, "learning_rate": 4.1282949079963e-07, "loss": 0.1974, "step": 8142 }, { "epoch": 2.6489915419648664, "grad_norm": 1.973622633765414, "learning_rate": 4.120767159468947e-07, "loss": 0.1829, "step": 8143 }, { "epoch": 2.649316851008458, "grad_norm": 2.0995351059088994, "learning_rate": 4.113245985466241e-07, "loss": 0.1653, "step": 8144 }, { "epoch": 2.6496421600520494, "grad_norm": 2.927618282985289, "learning_rate": 4.1057313870659923e-07, "loss": 0.2074, "step": 8145 }, { "epoch": 2.6499674690956407, "grad_norm": 1.7042713182692695, "learning_rate": 4.098223365345022e-07, "loss": 0.098, "step": 8146 }, { "epoch": 2.6502927781392325, "grad_norm": 2.265123392155693, "learning_rate": 4.090721921379248e-07, "loss": 0.1289, "step": 8147 }, { "epoch": 2.6506180871828238, "grad_norm": 1.9107035332147384, "learning_rate": 4.0832270562436436e-07, "loss": 0.1556, "step": 8148 }, { "epoch": 2.650943396226415, "grad_norm": 2.0145847116205546, "learning_rate": 4.075738771012205e-07, "loss": 0.1355, "step": 8149 }, { "epoch": 2.6512687052700064, "grad_norm": 2.1693193502702433, "learning_rate": 4.068257066758019e-07, "loss": 0.1365, "step": 8150 }, { "epoch": 2.6515940143135976, "grad_norm": 2.427958375926905, "learning_rate": 4.0607819445532273e-07, "loss": 0.2332, "step": 8151 }, { "epoch": 2.6519193233571894, "grad_norm": 1.9382434118552807, "learning_rate": 4.053313405469006e-07, "loss": 0.1441, "step": 8152 }, { "epoch": 2.6522446324007807, "grad_norm": 2.4416764836405696, "learning_rate": 4.045851450575616e-07, "loss": 0.1674, "step": 8153 }, { "epoch": 2.652569941444372, "grad_norm": 2.0348859350880133, "learning_rate": 4.0383960809423416e-07, "loss": 0.1419, "step": 8154 }, { "epoch": 2.6528952504879637, "grad_norm": 2.386994209293993, "learning_rate": 4.03094729763755e-07, "loss": 0.1945, "step": 8155 }, { "epoch": 2.653220559531555, "grad_norm": 2.112334263725838, "learning_rate": 4.023505101728675e-07, "loss": 0.1425, "step": 8156 }, { "epoch": 2.6535458685751463, "grad_norm": 2.4078588980125617, "learning_rate": 4.016069494282171e-07, "loss": 0.1587, "step": 8157 }, { "epoch": 2.653871177618738, "grad_norm": 1.7532146951539884, "learning_rate": 4.0086404763635454e-07, "loss": 0.1536, "step": 8158 }, { "epoch": 2.6541964866623293, "grad_norm": 2.0328611143051267, "learning_rate": 4.0012180490374254e-07, "loss": 0.1271, "step": 8159 }, { "epoch": 2.6545217957059206, "grad_norm": 2.2843689731913233, "learning_rate": 3.993802213367426e-07, "loss": 0.147, "step": 8160 }, { "epoch": 2.654847104749512, "grad_norm": 2.6806290364898073, "learning_rate": 3.9863929704162364e-07, "loss": 0.2197, "step": 8161 }, { "epoch": 2.655172413793103, "grad_norm": 1.9063569525751443, "learning_rate": 3.978990321245607e-07, "loss": 0.1565, "step": 8162 }, { "epoch": 2.655497722836695, "grad_norm": 2.064196320174675, "learning_rate": 3.971594266916362e-07, "loss": 0.147, "step": 8163 }, { "epoch": 2.6558230318802862, "grad_norm": 2.033160484779725, "learning_rate": 3.964204808488337e-07, "loss": 0.1561, "step": 8164 }, { "epoch": 2.6561483409238775, "grad_norm": 2.0467639348042854, "learning_rate": 3.956821947020467e-07, "loss": 0.1716, "step": 8165 }, { "epoch": 2.6564736499674693, "grad_norm": 2.1112560366734394, "learning_rate": 3.949445683570702e-07, "loss": 0.1847, "step": 8166 }, { "epoch": 2.6567989590110606, "grad_norm": 1.9236185908146288, "learning_rate": 3.9420760191960674e-07, "loss": 0.1551, "step": 8167 }, { "epoch": 2.657124268054652, "grad_norm": 2.116617589576679, "learning_rate": 3.934712954952663e-07, "loss": 0.1521, "step": 8168 }, { "epoch": 2.6574495770982436, "grad_norm": 2.5967872391071616, "learning_rate": 3.927356491895601e-07, "loss": 0.1997, "step": 8169 }, { "epoch": 2.657774886141835, "grad_norm": 2.368024164432911, "learning_rate": 3.9200066310790595e-07, "loss": 0.1948, "step": 8170 }, { "epoch": 2.658100195185426, "grad_norm": 2.3806548747219676, "learning_rate": 3.9126633735563026e-07, "loss": 0.1551, "step": 8171 }, { "epoch": 2.6584255042290175, "grad_norm": 2.2922630160044486, "learning_rate": 3.905326720379615e-07, "loss": 0.2191, "step": 8172 }, { "epoch": 2.6587508132726088, "grad_norm": 1.941892084670325, "learning_rate": 3.8979966726003294e-07, "loss": 0.1184, "step": 8173 }, { "epoch": 2.6590761223162005, "grad_norm": 2.096194729156089, "learning_rate": 3.89067323126886e-07, "loss": 0.1934, "step": 8174 }, { "epoch": 2.659401431359792, "grad_norm": 2.263272795692934, "learning_rate": 3.8833563974346676e-07, "loss": 0.1846, "step": 8175 }, { "epoch": 2.659726740403383, "grad_norm": 1.8344649218434985, "learning_rate": 3.8760461721462414e-07, "loss": 0.1323, "step": 8176 }, { "epoch": 2.660052049446975, "grad_norm": 1.9116188442695894, "learning_rate": 3.8687425564511607e-07, "loss": 0.1188, "step": 8177 }, { "epoch": 2.660377358490566, "grad_norm": 1.9802183366349824, "learning_rate": 3.8614455513960204e-07, "loss": 0.1627, "step": 8178 }, { "epoch": 2.6607026675341574, "grad_norm": 2.1441754981567356, "learning_rate": 3.854155158026496e-07, "loss": 0.2137, "step": 8179 }, { "epoch": 2.6610279765777487, "grad_norm": 2.0914821131503563, "learning_rate": 3.8468713773873124e-07, "loss": 0.1391, "step": 8180 }, { "epoch": 2.6613532856213404, "grad_norm": 2.653777290483448, "learning_rate": 3.8395942105222183e-07, "loss": 0.1694, "step": 8181 }, { "epoch": 2.6616785946649317, "grad_norm": 2.600381803462138, "learning_rate": 3.832323658474058e-07, "loss": 0.2387, "step": 8182 }, { "epoch": 2.662003903708523, "grad_norm": 2.044892270855358, "learning_rate": 3.825059722284702e-07, "loss": 0.1249, "step": 8183 }, { "epoch": 2.6623292127521143, "grad_norm": 2.139569308701271, "learning_rate": 3.8178024029950634e-07, "loss": 0.128, "step": 8184 }, { "epoch": 2.662654521795706, "grad_norm": 2.0069181054918284, "learning_rate": 3.8105517016451434e-07, "loss": 0.1428, "step": 8185 }, { "epoch": 2.6629798308392973, "grad_norm": 2.0911648874600504, "learning_rate": 3.803307619273955e-07, "loss": 0.1443, "step": 8186 }, { "epoch": 2.6633051398828886, "grad_norm": 2.6764047649177107, "learning_rate": 3.796070156919579e-07, "loss": 0.2396, "step": 8187 }, { "epoch": 2.6636304489264804, "grad_norm": 1.9309667064194656, "learning_rate": 3.7888393156191693e-07, "loss": 0.1314, "step": 8188 }, { "epoch": 2.6639557579700717, "grad_norm": 2.1958799185098434, "learning_rate": 3.781615096408892e-07, "loss": 0.1708, "step": 8189 }, { "epoch": 2.664281067013663, "grad_norm": 2.375739301356652, "learning_rate": 3.774397500323973e-07, "loss": 0.1756, "step": 8190 }, { "epoch": 2.6646063760572543, "grad_norm": 2.0910236283473194, "learning_rate": 3.7671865283987254e-07, "loss": 0.1574, "step": 8191 }, { "epoch": 2.6649316851008455, "grad_norm": 2.1454454564478014, "learning_rate": 3.759982181666477e-07, "loss": 0.1479, "step": 8192 }, { "epoch": 2.6652569941444373, "grad_norm": 1.6363415940674775, "learning_rate": 3.7527844611596033e-07, "loss": 0.103, "step": 8193 }, { "epoch": 2.6655823031880286, "grad_norm": 2.2312014365702852, "learning_rate": 3.7455933679095446e-07, "loss": 0.2009, "step": 8194 }, { "epoch": 2.66590761223162, "grad_norm": 1.8966875594087071, "learning_rate": 3.738408902946811e-07, "loss": 0.1513, "step": 8195 }, { "epoch": 2.6662329212752116, "grad_norm": 1.8656005957503377, "learning_rate": 3.731231067300911e-07, "loss": 0.1434, "step": 8196 }, { "epoch": 2.666558230318803, "grad_norm": 2.124992465081633, "learning_rate": 3.724059862000462e-07, "loss": 0.1451, "step": 8197 }, { "epoch": 2.666883539362394, "grad_norm": 2.223048799854828, "learning_rate": 3.7168952880730747e-07, "loss": 0.1534, "step": 8198 }, { "epoch": 2.667208848405986, "grad_norm": 2.12009416279753, "learning_rate": 3.7097373465454554e-07, "loss": 0.1999, "step": 8199 }, { "epoch": 2.6675341574495772, "grad_norm": 2.0266802562043855, "learning_rate": 3.7025860384433456e-07, "loss": 0.1687, "step": 8200 }, { "epoch": 2.6678594664931685, "grad_norm": 2.4508546778133358, "learning_rate": 3.695441364791513e-07, "loss": 0.1931, "step": 8201 }, { "epoch": 2.66818477553676, "grad_norm": 2.2591624489384983, "learning_rate": 3.688303326613807e-07, "loss": 0.1799, "step": 8202 }, { "epoch": 2.668510084580351, "grad_norm": 2.109280954379718, "learning_rate": 3.681171924933119e-07, "loss": 0.1201, "step": 8203 }, { "epoch": 2.668835393623943, "grad_norm": 2.215407261535292, "learning_rate": 3.6740471607713657e-07, "loss": 0.1914, "step": 8204 }, { "epoch": 2.669160702667534, "grad_norm": 2.1069673825315975, "learning_rate": 3.666929035149547e-07, "loss": 0.1585, "step": 8205 }, { "epoch": 2.6694860117111254, "grad_norm": 1.9148565485825717, "learning_rate": 3.6598175490876906e-07, "loss": 0.1265, "step": 8206 }, { "epoch": 2.669811320754717, "grad_norm": 2.150173869530609, "learning_rate": 3.6527127036048715e-07, "loss": 0.147, "step": 8207 }, { "epoch": 2.6701366297983085, "grad_norm": 1.9713048053494466, "learning_rate": 3.6456144997192355e-07, "loss": 0.1654, "step": 8208 }, { "epoch": 2.6704619388418998, "grad_norm": 2.040387624083913, "learning_rate": 3.6385229384479305e-07, "loss": 0.1434, "step": 8209 }, { "epoch": 2.6707872478854915, "grad_norm": 2.08607813327326, "learning_rate": 3.6314380208072153e-07, "loss": 0.1725, "step": 8210 }, { "epoch": 2.671112556929083, "grad_norm": 2.1979699267738333, "learning_rate": 3.624359747812334e-07, "loss": 0.19, "step": 8211 }, { "epoch": 2.671437865972674, "grad_norm": 2.118100100105911, "learning_rate": 3.61728812047763e-07, "loss": 0.132, "step": 8212 }, { "epoch": 2.6717631750162654, "grad_norm": 2.330898390828315, "learning_rate": 3.6102231398164545e-07, "loss": 0.1542, "step": 8213 }, { "epoch": 2.6720884840598567, "grad_norm": 2.308599729620566, "learning_rate": 3.603164806841236e-07, "loss": 0.1639, "step": 8214 }, { "epoch": 2.6724137931034484, "grad_norm": 2.1756136846523795, "learning_rate": 3.5961131225634385e-07, "loss": 0.1561, "step": 8215 }, { "epoch": 2.6727391021470397, "grad_norm": 2.4531383129058275, "learning_rate": 3.5890680879935637e-07, "loss": 0.1722, "step": 8216 }, { "epoch": 2.673064411190631, "grad_norm": 2.0386549140495616, "learning_rate": 3.582029704141171e-07, "loss": 0.1618, "step": 8217 }, { "epoch": 2.6733897202342227, "grad_norm": 2.3052627756897532, "learning_rate": 3.574997972014882e-07, "loss": 0.2095, "step": 8218 }, { "epoch": 2.673715029277814, "grad_norm": 1.9845044855300737, "learning_rate": 3.567972892622329e-07, "loss": 0.1683, "step": 8219 }, { "epoch": 2.6740403383214053, "grad_norm": 2.1002534507449147, "learning_rate": 3.560954466970223e-07, "loss": 0.1659, "step": 8220 }, { "epoch": 2.6743656473649966, "grad_norm": 2.10158650028064, "learning_rate": 3.5539426960643e-07, "loss": 0.1333, "step": 8221 }, { "epoch": 2.674690956408588, "grad_norm": 2.1487709696369137, "learning_rate": 3.5469375809093543e-07, "loss": 0.1727, "step": 8222 }, { "epoch": 2.6750162654521796, "grad_norm": 2.44551590616627, "learning_rate": 3.539939122509234e-07, "loss": 0.1758, "step": 8223 }, { "epoch": 2.675341574495771, "grad_norm": 2.215400670143872, "learning_rate": 3.5329473218668084e-07, "loss": 0.2004, "step": 8224 }, { "epoch": 2.675666883539362, "grad_norm": 2.240373324509367, "learning_rate": 3.525962179984005e-07, "loss": 0.2063, "step": 8225 }, { "epoch": 2.675992192582954, "grad_norm": 2.1259767245972667, "learning_rate": 3.5189836978618166e-07, "loss": 0.1381, "step": 8226 }, { "epoch": 2.6763175016265452, "grad_norm": 2.1205188266153305, "learning_rate": 3.512011876500254e-07, "loss": 0.1347, "step": 8227 }, { "epoch": 2.6766428106701365, "grad_norm": 1.9399301071672932, "learning_rate": 3.5050467168983806e-07, "loss": 0.1191, "step": 8228 }, { "epoch": 2.6769681197137283, "grad_norm": 2.157007271139625, "learning_rate": 3.4980882200543075e-07, "loss": 0.1592, "step": 8229 }, { "epoch": 2.6772934287573196, "grad_norm": 2.0451396782462554, "learning_rate": 3.4911363869652106e-07, "loss": 0.1172, "step": 8230 }, { "epoch": 2.677618737800911, "grad_norm": 2.018019869257194, "learning_rate": 3.4841912186272644e-07, "loss": 0.1772, "step": 8231 }, { "epoch": 2.677944046844502, "grad_norm": 2.1758480048706175, "learning_rate": 3.477252716035745e-07, "loss": 0.1691, "step": 8232 }, { "epoch": 2.6782693558880935, "grad_norm": 2.2384457025141544, "learning_rate": 3.470320880184913e-07, "loss": 0.1455, "step": 8233 }, { "epoch": 2.678594664931685, "grad_norm": 2.5243141664622994, "learning_rate": 3.4633957120681294e-07, "loss": 0.2012, "step": 8234 }, { "epoch": 2.6789199739752765, "grad_norm": 1.5706373193471714, "learning_rate": 3.4564772126777726e-07, "loss": 0.1116, "step": 8235 }, { "epoch": 2.6792452830188678, "grad_norm": 1.9288818206409917, "learning_rate": 3.449565383005249e-07, "loss": 0.1224, "step": 8236 }, { "epoch": 2.6795705920624595, "grad_norm": 1.805995459139844, "learning_rate": 3.44266022404105e-07, "loss": 0.1564, "step": 8237 }, { "epoch": 2.679895901106051, "grad_norm": 2.128120853947072, "learning_rate": 3.4357617367746897e-07, "loss": 0.1888, "step": 8238 }, { "epoch": 2.680221210149642, "grad_norm": 1.8930583696887835, "learning_rate": 3.4288699221947054e-07, "loss": 0.1366, "step": 8239 }, { "epoch": 2.680546519193234, "grad_norm": 2.4067351858336505, "learning_rate": 3.421984781288723e-07, "loss": 0.1339, "step": 8240 }, { "epoch": 2.680871828236825, "grad_norm": 2.136771232109204, "learning_rate": 3.415106315043371e-07, "loss": 0.1471, "step": 8241 }, { "epoch": 2.6811971372804164, "grad_norm": 1.9308726654672628, "learning_rate": 3.4082345244443425e-07, "loss": 0.1557, "step": 8242 }, { "epoch": 2.6815224463240077, "grad_norm": 2.079104792784434, "learning_rate": 3.401369410476374e-07, "loss": 0.1818, "step": 8243 }, { "epoch": 2.681847755367599, "grad_norm": 1.9515188078217542, "learning_rate": 3.394510974123244e-07, "loss": 0.1787, "step": 8244 }, { "epoch": 2.6821730644111907, "grad_norm": 2.2271415981580445, "learning_rate": 3.3876592163677447e-07, "loss": 0.143, "step": 8245 }, { "epoch": 2.682498373454782, "grad_norm": 2.1043710600859065, "learning_rate": 3.380814138191774e-07, "loss": 0.1543, "step": 8246 }, { "epoch": 2.6828236824983733, "grad_norm": 1.8906351916717687, "learning_rate": 3.373975740576219e-07, "loss": 0.1242, "step": 8247 }, { "epoch": 2.683148991541965, "grad_norm": 2.618976935581454, "learning_rate": 3.367144024501018e-07, "loss": 0.2546, "step": 8248 }, { "epoch": 2.6834743005855564, "grad_norm": 1.9308915437066065, "learning_rate": 3.3603189909451716e-07, "loss": 0.1427, "step": 8249 }, { "epoch": 2.6837996096291477, "grad_norm": 2.0133264495764935, "learning_rate": 3.35350064088672e-07, "loss": 0.1645, "step": 8250 }, { "epoch": 2.684124918672739, "grad_norm": 1.904162639207959, "learning_rate": 3.34668897530272e-07, "loss": 0.1522, "step": 8251 }, { "epoch": 2.6844502277163302, "grad_norm": 2.157640699267619, "learning_rate": 3.3398839951692965e-07, "loss": 0.1468, "step": 8252 }, { "epoch": 2.684775536759922, "grad_norm": 2.1347864935972254, "learning_rate": 3.3330857014616026e-07, "loss": 0.1745, "step": 8253 }, { "epoch": 2.6851008458035133, "grad_norm": 1.655478410005227, "learning_rate": 3.3262940951538425e-07, "loss": 0.1195, "step": 8254 }, { "epoch": 2.6854261548471046, "grad_norm": 2.217012522711263, "learning_rate": 3.3195091772192655e-07, "loss": 0.171, "step": 8255 }, { "epoch": 2.6857514638906963, "grad_norm": 2.763604925407769, "learning_rate": 3.312730948630144e-07, "loss": 0.2056, "step": 8256 }, { "epoch": 2.6860767729342876, "grad_norm": 2.6901466830068967, "learning_rate": 3.30595941035779e-07, "loss": 0.2275, "step": 8257 }, { "epoch": 2.686402081977879, "grad_norm": 2.4207771548194583, "learning_rate": 3.299194563372604e-07, "loss": 0.1721, "step": 8258 }, { "epoch": 2.6867273910214706, "grad_norm": 2.1188231066613508, "learning_rate": 3.292436408643968e-07, "loss": 0.1784, "step": 8259 }, { "epoch": 2.687052700065062, "grad_norm": 2.2050876865903644, "learning_rate": 3.285684947140333e-07, "loss": 0.132, "step": 8260 }, { "epoch": 2.687378009108653, "grad_norm": 1.961220158634259, "learning_rate": 3.2789401798291876e-07, "loss": 0.1286, "step": 8261 }, { "epoch": 2.6877033181522445, "grad_norm": 1.8612223229556804, "learning_rate": 3.2722021076770747e-07, "loss": 0.1429, "step": 8262 }, { "epoch": 2.688028627195836, "grad_norm": 2.5030088176688667, "learning_rate": 3.2654707316495447e-07, "loss": 0.1544, "step": 8263 }, { "epoch": 2.6883539362394275, "grad_norm": 1.9525761967472408, "learning_rate": 3.2587460527112314e-07, "loss": 0.1595, "step": 8264 }, { "epoch": 2.688679245283019, "grad_norm": 2.336910278703182, "learning_rate": 3.2520280718257577e-07, "loss": 0.1581, "step": 8265 }, { "epoch": 2.68900455432661, "grad_norm": 1.9629960572741731, "learning_rate": 3.2453167899558326e-07, "loss": 0.1578, "step": 8266 }, { "epoch": 2.689329863370202, "grad_norm": 1.9655547448231347, "learning_rate": 3.238612208063191e-07, "loss": 0.1546, "step": 8267 }, { "epoch": 2.689655172413793, "grad_norm": 2.0986503740231806, "learning_rate": 3.231914327108593e-07, "loss": 0.1266, "step": 8268 }, { "epoch": 2.6899804814573844, "grad_norm": 2.052867954490976, "learning_rate": 3.2252231480518537e-07, "loss": 0.1689, "step": 8269 }, { "epoch": 2.690305790500976, "grad_norm": 2.124690281047115, "learning_rate": 3.218538671851829e-07, "loss": 0.17, "step": 8270 }, { "epoch": 2.6906310995445675, "grad_norm": 2.1248049915870335, "learning_rate": 3.21186089946639e-07, "loss": 0.1265, "step": 8271 }, { "epoch": 2.6909564085881588, "grad_norm": 2.4805812453711398, "learning_rate": 3.205189831852495e-07, "loss": 0.2134, "step": 8272 }, { "epoch": 2.69128171763175, "grad_norm": 1.8675004661658743, "learning_rate": 3.198525469966085e-07, "loss": 0.1207, "step": 8273 }, { "epoch": 2.6916070266753414, "grad_norm": 2.0697982086747038, "learning_rate": 3.1918678147621787e-07, "loss": 0.2183, "step": 8274 }, { "epoch": 2.691932335718933, "grad_norm": 2.3226189063730023, "learning_rate": 3.185216867194835e-07, "loss": 0.2438, "step": 8275 }, { "epoch": 2.6922576447625244, "grad_norm": 1.891079979149548, "learning_rate": 3.1785726282171194e-07, "loss": 0.1249, "step": 8276 }, { "epoch": 2.6925829538061157, "grad_norm": 1.9660690131980425, "learning_rate": 3.1719350987811537e-07, "loss": 0.1323, "step": 8277 }, { "epoch": 2.6929082628497074, "grad_norm": 2.0601383689150965, "learning_rate": 3.1653042798381273e-07, "loss": 0.1203, "step": 8278 }, { "epoch": 2.6932335718932987, "grad_norm": 2.0341048587267294, "learning_rate": 3.158680172338219e-07, "loss": 0.1675, "step": 8279 }, { "epoch": 2.69355888093689, "grad_norm": 2.6023122853272858, "learning_rate": 3.152062777230663e-07, "loss": 0.1792, "step": 8280 }, { "epoch": 2.6938841899804813, "grad_norm": 2.259846219497933, "learning_rate": 3.145452095463752e-07, "loss": 0.1478, "step": 8281 }, { "epoch": 2.694209499024073, "grad_norm": 2.3146304805049156, "learning_rate": 3.1388481279848005e-07, "loss": 0.1853, "step": 8282 }, { "epoch": 2.6945348080676643, "grad_norm": 1.8037663635855468, "learning_rate": 3.132250875740145e-07, "loss": 0.1727, "step": 8283 }, { "epoch": 2.6948601171112556, "grad_norm": 2.2257330453545334, "learning_rate": 3.1256603396751963e-07, "loss": 0.223, "step": 8284 }, { "epoch": 2.695185426154847, "grad_norm": 2.189128533889663, "learning_rate": 3.119076520734371e-07, "loss": 0.171, "step": 8285 }, { "epoch": 2.6955107351984386, "grad_norm": 2.3606655138760932, "learning_rate": 3.1124994198611315e-07, "loss": 0.1797, "step": 8286 }, { "epoch": 2.69583604424203, "grad_norm": 1.7402739288725915, "learning_rate": 3.1059290379979956e-07, "loss": 0.1277, "step": 8287 }, { "epoch": 2.6961613532856212, "grad_norm": 2.769974278176389, "learning_rate": 3.0993653760864883e-07, "loss": 0.1608, "step": 8288 }, { "epoch": 2.696486662329213, "grad_norm": 2.0364246567104414, "learning_rate": 3.0928084350671905e-07, "loss": 0.1674, "step": 8289 }, { "epoch": 2.6968119713728043, "grad_norm": 2.039280154942017, "learning_rate": 3.086258215879723e-07, "loss": 0.1437, "step": 8290 }, { "epoch": 2.6971372804163956, "grad_norm": 2.294173165938486, "learning_rate": 3.079714719462729e-07, "loss": 0.1751, "step": 8291 }, { "epoch": 2.697462589459987, "grad_norm": 2.1410195762311672, "learning_rate": 3.073177946753908e-07, "loss": 0.1431, "step": 8292 }, { "epoch": 2.697787898503578, "grad_norm": 2.0585470884350374, "learning_rate": 3.0666478986899617e-07, "loss": 0.1501, "step": 8293 }, { "epoch": 2.69811320754717, "grad_norm": 1.9107242715695978, "learning_rate": 3.0601245762066746e-07, "loss": 0.1328, "step": 8294 }, { "epoch": 2.698438516590761, "grad_norm": 2.1483995746353415, "learning_rate": 3.0536079802388205e-07, "loss": 0.165, "step": 8295 }, { "epoch": 2.6987638256343525, "grad_norm": 2.147796954067559, "learning_rate": 3.047098111720259e-07, "loss": 0.1539, "step": 8296 }, { "epoch": 2.699089134677944, "grad_norm": 1.7826667443834243, "learning_rate": 3.040594971583827e-07, "loss": 0.1215, "step": 8297 }, { "epoch": 2.6994144437215355, "grad_norm": 2.0224645944353865, "learning_rate": 3.0340985607614503e-07, "loss": 0.1209, "step": 8298 }, { "epoch": 2.699739752765127, "grad_norm": 1.9144835654626509, "learning_rate": 3.0276088801840686e-07, "loss": 0.1715, "step": 8299 }, { "epoch": 2.7000650618087185, "grad_norm": 2.2457181337194005, "learning_rate": 3.0211259307816434e-07, "loss": 0.2003, "step": 8300 }, { "epoch": 2.70039037085231, "grad_norm": 2.2412569213480897, "learning_rate": 3.014649713483198e-07, "loss": 0.2071, "step": 8301 }, { "epoch": 2.700715679895901, "grad_norm": 1.8937140634555407, "learning_rate": 3.008180229216778e-07, "loss": 0.135, "step": 8302 }, { "epoch": 2.7010409889394924, "grad_norm": 2.0773572995596603, "learning_rate": 3.001717478909455e-07, "loss": 0.149, "step": 8303 }, { "epoch": 2.7013662979830837, "grad_norm": 2.125805559060463, "learning_rate": 2.9952614634873647e-07, "loss": 0.2265, "step": 8304 }, { "epoch": 2.7016916070266754, "grad_norm": 2.1496391430309836, "learning_rate": 2.988812183875633e-07, "loss": 0.1453, "step": 8305 }, { "epoch": 2.7020169160702667, "grad_norm": 1.7391984299118406, "learning_rate": 2.982369640998467e-07, "loss": 0.1359, "step": 8306 }, { "epoch": 2.702342225113858, "grad_norm": 2.5090563664237675, "learning_rate": 2.9759338357790824e-07, "loss": 0.1863, "step": 8307 }, { "epoch": 2.7026675341574498, "grad_norm": 1.985928131257586, "learning_rate": 2.9695047691397306e-07, "loss": 0.1599, "step": 8308 }, { "epoch": 2.702992843201041, "grad_norm": 2.244767592578776, "learning_rate": 2.9630824420016857e-07, "loss": 0.1371, "step": 8309 }, { "epoch": 2.7033181522446323, "grad_norm": 2.0356738313287304, "learning_rate": 2.956666855285306e-07, "loss": 0.1784, "step": 8310 }, { "epoch": 2.703643461288224, "grad_norm": 2.0135793478915494, "learning_rate": 2.950258009909929e-07, "loss": 0.1691, "step": 8311 }, { "epoch": 2.7039687703318154, "grad_norm": 2.2856788544071716, "learning_rate": 2.943855906793941e-07, "loss": 0.1982, "step": 8312 }, { "epoch": 2.7042940793754067, "grad_norm": 2.036033796190062, "learning_rate": 2.937460546854776e-07, "loss": 0.1812, "step": 8313 }, { "epoch": 2.704619388418998, "grad_norm": 2.270186227992726, "learning_rate": 2.9310719310089e-07, "loss": 0.1969, "step": 8314 }, { "epoch": 2.7049446974625893, "grad_norm": 2.20541384661913, "learning_rate": 2.924690060171792e-07, "loss": 0.1704, "step": 8315 }, { "epoch": 2.705270006506181, "grad_norm": 1.8867151580637203, "learning_rate": 2.918314935257993e-07, "loss": 0.1254, "step": 8316 }, { "epoch": 2.7055953155497723, "grad_norm": 2.1010165127042404, "learning_rate": 2.911946557181045e-07, "loss": 0.1485, "step": 8317 }, { "epoch": 2.7059206245933636, "grad_norm": 2.608379106730277, "learning_rate": 2.9055849268535496e-07, "loss": 0.1775, "step": 8318 }, { "epoch": 2.7062459336369553, "grad_norm": 2.3964770966755395, "learning_rate": 2.89923004518714e-07, "loss": 0.1866, "step": 8319 }, { "epoch": 2.7065712426805466, "grad_norm": 2.0478859098938877, "learning_rate": 2.8928819130924656e-07, "loss": 0.1739, "step": 8320 }, { "epoch": 2.706896551724138, "grad_norm": 1.7548538660079522, "learning_rate": 2.8865405314792204e-07, "loss": 0.1528, "step": 8321 }, { "epoch": 2.707221860767729, "grad_norm": 2.1606714352666088, "learning_rate": 2.880205901256133e-07, "loss": 0.1333, "step": 8322 }, { "epoch": 2.7075471698113205, "grad_norm": 2.1235634479547145, "learning_rate": 2.87387802333095e-07, "loss": 0.1432, "step": 8323 }, { "epoch": 2.7078724788549122, "grad_norm": 2.173944357032359, "learning_rate": 2.867556898610474e-07, "loss": 0.1833, "step": 8324 }, { "epoch": 2.7081977878985035, "grad_norm": 2.4882562296639756, "learning_rate": 2.861242528000513e-07, "loss": 0.1733, "step": 8325 }, { "epoch": 2.708523096942095, "grad_norm": 2.2528320352510174, "learning_rate": 2.854934912405921e-07, "loss": 0.1745, "step": 8326 }, { "epoch": 2.7088484059856865, "grad_norm": 2.4756280547563296, "learning_rate": 2.8486340527306033e-07, "loss": 0.1924, "step": 8327 }, { "epoch": 2.709173715029278, "grad_norm": 1.8779266375098862, "learning_rate": 2.8423399498774596e-07, "loss": 0.1589, "step": 8328 }, { "epoch": 2.709499024072869, "grad_norm": 2.3580935510189485, "learning_rate": 2.8360526047484296e-07, "loss": 0.1871, "step": 8329 }, { "epoch": 2.709824333116461, "grad_norm": 1.803754537407936, "learning_rate": 2.8297720182445197e-07, "loss": 0.138, "step": 8330 }, { "epoch": 2.710149642160052, "grad_norm": 2.6982150183774025, "learning_rate": 2.823498191265733e-07, "loss": 0.2069, "step": 8331 }, { "epoch": 2.7104749512036435, "grad_norm": 2.034023007908292, "learning_rate": 2.8172311247110895e-07, "loss": 0.1594, "step": 8332 }, { "epoch": 2.7108002602472347, "grad_norm": 2.1776811827222473, "learning_rate": 2.810970819478703e-07, "loss": 0.1619, "step": 8333 }, { "epoch": 2.711125569290826, "grad_norm": 2.021722141509892, "learning_rate": 2.804717276465657e-07, "loss": 0.1464, "step": 8334 }, { "epoch": 2.711450878334418, "grad_norm": 2.081336195549663, "learning_rate": 2.798470496568084e-07, "loss": 0.1435, "step": 8335 }, { "epoch": 2.711776187378009, "grad_norm": 2.3727877518106926, "learning_rate": 2.792230480681163e-07, "loss": 0.1772, "step": 8336 }, { "epoch": 2.7121014964216004, "grad_norm": 2.0502046771244795, "learning_rate": 2.7859972296990944e-07, "loss": 0.1505, "step": 8337 }, { "epoch": 2.712426805465192, "grad_norm": 1.807329423258449, "learning_rate": 2.779770744515098e-07, "loss": 0.1246, "step": 8338 }, { "epoch": 2.7127521145087834, "grad_norm": 2.113343340386397, "learning_rate": 2.773551026021437e-07, "loss": 0.1616, "step": 8339 }, { "epoch": 2.7130774235523747, "grad_norm": 2.1132911319898775, "learning_rate": 2.767338075109399e-07, "loss": 0.173, "step": 8340 }, { "epoch": 2.7134027325959664, "grad_norm": 1.8610613688988942, "learning_rate": 2.7611318926693055e-07, "loss": 0.1393, "step": 8341 }, { "epoch": 2.7137280416395577, "grad_norm": 2.6885263807840256, "learning_rate": 2.7549324795905164e-07, "loss": 0.1729, "step": 8342 }, { "epoch": 2.714053350683149, "grad_norm": 1.92000235646848, "learning_rate": 2.7487398367613994e-07, "loss": 0.1424, "step": 8343 }, { "epoch": 2.7143786597267403, "grad_norm": 1.9866354073028962, "learning_rate": 2.7425539650693556e-07, "loss": 0.1585, "step": 8344 }, { "epoch": 2.7147039687703316, "grad_norm": 2.3067428364077913, "learning_rate": 2.736374865400848e-07, "loss": 0.154, "step": 8345 }, { "epoch": 2.7150292778139233, "grad_norm": 2.063510320218858, "learning_rate": 2.7302025386413355e-07, "loss": 0.1452, "step": 8346 }, { "epoch": 2.7153545868575146, "grad_norm": 1.8569121777849422, "learning_rate": 2.72403698567531e-07, "loss": 0.136, "step": 8347 }, { "epoch": 2.715679895901106, "grad_norm": 2.5392800704174388, "learning_rate": 2.717878207386304e-07, "loss": 0.2181, "step": 8348 }, { "epoch": 2.7160052049446977, "grad_norm": 2.2322498307817753, "learning_rate": 2.711726204656884e-07, "loss": 0.1675, "step": 8349 }, { "epoch": 2.716330513988289, "grad_norm": 2.2638078054644546, "learning_rate": 2.705580978368616e-07, "loss": 0.1743, "step": 8350 }, { "epoch": 2.7166558230318802, "grad_norm": 2.250405277477784, "learning_rate": 2.699442529402135e-07, "loss": 0.1413, "step": 8351 }, { "epoch": 2.7169811320754715, "grad_norm": 1.8920602092874947, "learning_rate": 2.6933108586370717e-07, "loss": 0.14, "step": 8352 }, { "epoch": 2.7173064411190633, "grad_norm": 2.20953237294721, "learning_rate": 2.6871859669520994e-07, "loss": 0.1609, "step": 8353 }, { "epoch": 2.7176317501626546, "grad_norm": 1.9820188383277348, "learning_rate": 2.6810678552249335e-07, "loss": 0.1419, "step": 8354 }, { "epoch": 2.717957059206246, "grad_norm": 2.026823567303673, "learning_rate": 2.674956524332284e-07, "loss": 0.178, "step": 8355 }, { "epoch": 2.718282368249837, "grad_norm": 2.5942074064551277, "learning_rate": 2.6688519751499166e-07, "loss": 0.2218, "step": 8356 }, { "epoch": 2.718607677293429, "grad_norm": 2.1473307851803924, "learning_rate": 2.662754208552626e-07, "loss": 0.1286, "step": 8357 }, { "epoch": 2.71893298633702, "grad_norm": 1.976037940883902, "learning_rate": 2.656663225414213e-07, "loss": 0.139, "step": 8358 }, { "epoch": 2.7192582953806115, "grad_norm": 1.6523127493048178, "learning_rate": 2.65057902660753e-07, "loss": 0.1112, "step": 8359 }, { "epoch": 2.719583604424203, "grad_norm": 3.04958116352241, "learning_rate": 2.6445016130044356e-07, "loss": 0.1349, "step": 8360 }, { "epoch": 2.7199089134677945, "grad_norm": 2.147251458900358, "learning_rate": 2.6384309854758316e-07, "loss": 0.1709, "step": 8361 }, { "epoch": 2.720234222511386, "grad_norm": 2.0409479589649884, "learning_rate": 2.6323671448916567e-07, "loss": 0.1861, "step": 8362 }, { "epoch": 2.720559531554977, "grad_norm": 1.8408922244081114, "learning_rate": 2.6263100921208484e-07, "loss": 0.1132, "step": 8363 }, { "epoch": 2.7208848405985684, "grad_norm": 2.4360332301079106, "learning_rate": 2.620259828031374e-07, "loss": 0.1802, "step": 8364 }, { "epoch": 2.72121014964216, "grad_norm": 2.1649968808770064, "learning_rate": 2.614216353490268e-07, "loss": 0.134, "step": 8365 }, { "epoch": 2.7215354586857514, "grad_norm": 2.1199424723067803, "learning_rate": 2.6081796693635586e-07, "loss": 0.1677, "step": 8366 }, { "epoch": 2.7218607677293427, "grad_norm": 1.8785074996613604, "learning_rate": 2.602149776516288e-07, "loss": 0.1108, "step": 8367 }, { "epoch": 2.7221860767729344, "grad_norm": 1.9952786008481647, "learning_rate": 2.5961266758125547e-07, "loss": 0.1533, "step": 8368 }, { "epoch": 2.7225113858165257, "grad_norm": 1.8573684170504035, "learning_rate": 2.590110368115484e-07, "loss": 0.162, "step": 8369 }, { "epoch": 2.722836694860117, "grad_norm": 2.182202853328973, "learning_rate": 2.5841008542872026e-07, "loss": 0.2148, "step": 8370 }, { "epoch": 2.7231620039037088, "grad_norm": 2.1545743659346113, "learning_rate": 2.578098135188883e-07, "loss": 0.1707, "step": 8371 }, { "epoch": 2.7234873129473, "grad_norm": 1.9755210892292576, "learning_rate": 2.572102211680716e-07, "loss": 0.1304, "step": 8372 }, { "epoch": 2.7238126219908914, "grad_norm": 1.9257129192623506, "learning_rate": 2.566113084621924e-07, "loss": 0.1444, "step": 8373 }, { "epoch": 2.7241379310344827, "grad_norm": 2.651115741675179, "learning_rate": 2.56013075487076e-07, "loss": 0.2343, "step": 8374 }, { "epoch": 2.724463240078074, "grad_norm": 1.8194737960857075, "learning_rate": 2.5541552232844767e-07, "loss": 0.1548, "step": 8375 }, { "epoch": 2.7247885491216657, "grad_norm": 1.9042095160886825, "learning_rate": 2.548186490719384e-07, "loss": 0.1213, "step": 8376 }, { "epoch": 2.725113858165257, "grad_norm": 2.4005396673400488, "learning_rate": 2.542224558030815e-07, "loss": 0.1862, "step": 8377 }, { "epoch": 2.7254391672088483, "grad_norm": 1.883903123379051, "learning_rate": 2.5362694260731013e-07, "loss": 0.1492, "step": 8378 }, { "epoch": 2.72576447625244, "grad_norm": 2.1526048646464675, "learning_rate": 2.530321095699634e-07, "loss": 0.1611, "step": 8379 }, { "epoch": 2.7260897852960313, "grad_norm": 2.291626774009391, "learning_rate": 2.524379567762791e-07, "loss": 0.1671, "step": 8380 }, { "epoch": 2.7264150943396226, "grad_norm": 1.9145983993770825, "learning_rate": 2.518444843114015e-07, "loss": 0.137, "step": 8381 }, { "epoch": 2.7267404033832143, "grad_norm": 2.2189604511858363, "learning_rate": 2.512516922603747e-07, "loss": 0.1969, "step": 8382 }, { "epoch": 2.7270657124268056, "grad_norm": 2.111760344813331, "learning_rate": 2.5065958070814745e-07, "loss": 0.1403, "step": 8383 }, { "epoch": 2.727391021470397, "grad_norm": 1.987842182910681, "learning_rate": 2.500681497395674e-07, "loss": 0.1084, "step": 8384 }, { "epoch": 2.727716330513988, "grad_norm": 1.9046572545805438, "learning_rate": 2.49477399439389e-07, "loss": 0.1404, "step": 8385 }, { "epoch": 2.7280416395575795, "grad_norm": 2.3162064086019223, "learning_rate": 2.4888732989226726e-07, "loss": 0.1714, "step": 8386 }, { "epoch": 2.7283669486011712, "grad_norm": 2.3168917495459853, "learning_rate": 2.4829794118275796e-07, "loss": 0.1852, "step": 8387 }, { "epoch": 2.7286922576447625, "grad_norm": 2.2995660519597574, "learning_rate": 2.4770923339532183e-07, "loss": 0.2278, "step": 8388 }, { "epoch": 2.729017566688354, "grad_norm": 1.7185682529193786, "learning_rate": 2.471212066143214e-07, "loss": 0.0885, "step": 8389 }, { "epoch": 2.7293428757319456, "grad_norm": 1.760194896515885, "learning_rate": 2.465338609240209e-07, "loss": 0.1461, "step": 8390 }, { "epoch": 2.729668184775537, "grad_norm": 2.090097742528743, "learning_rate": 2.4594719640858746e-07, "loss": 0.1551, "step": 8391 }, { "epoch": 2.729993493819128, "grad_norm": 1.943901141678525, "learning_rate": 2.4536121315208993e-07, "loss": 0.1576, "step": 8392 }, { "epoch": 2.7303188028627194, "grad_norm": 1.9940922439644282, "learning_rate": 2.447759112385012e-07, "loss": 0.1507, "step": 8393 }, { "epoch": 2.7306441119063107, "grad_norm": 2.3011519790845525, "learning_rate": 2.4419129075169466e-07, "loss": 0.1756, "step": 8394 }, { "epoch": 2.7309694209499025, "grad_norm": 2.231665997011832, "learning_rate": 2.4360735177544783e-07, "loss": 0.1878, "step": 8395 }, { "epoch": 2.7312947299934938, "grad_norm": 2.2897105271181677, "learning_rate": 2.430240943934364e-07, "loss": 0.1558, "step": 8396 }, { "epoch": 2.731620039037085, "grad_norm": 1.8977421327815254, "learning_rate": 2.4244151868924595e-07, "loss": 0.1478, "step": 8397 }, { "epoch": 2.731945348080677, "grad_norm": 2.3171537802668833, "learning_rate": 2.418596247463578e-07, "loss": 0.1843, "step": 8398 }, { "epoch": 2.732270657124268, "grad_norm": 2.104654927091156, "learning_rate": 2.412784126481571e-07, "loss": 0.111, "step": 8399 }, { "epoch": 2.7325959661678594, "grad_norm": 2.294872601532799, "learning_rate": 2.406978824779321e-07, "loss": 0.2072, "step": 8400 }, { "epoch": 2.732921275211451, "grad_norm": 1.8327350093642136, "learning_rate": 2.4011803431887526e-07, "loss": 0.1743, "step": 8401 }, { "epoch": 2.7332465842550424, "grad_norm": 2.3743691991574067, "learning_rate": 2.395388682540767e-07, "loss": 0.1358, "step": 8402 }, { "epoch": 2.7335718932986337, "grad_norm": 1.915935161461607, "learning_rate": 2.389603843665328e-07, "loss": 0.1535, "step": 8403 }, { "epoch": 2.733897202342225, "grad_norm": 2.1244053262322447, "learning_rate": 2.3838258273913894e-07, "loss": 0.1494, "step": 8404 }, { "epoch": 2.7342225113858163, "grad_norm": 2.3012396174567824, "learning_rate": 2.3780546345469667e-07, "loss": 0.1693, "step": 8405 }, { "epoch": 2.734547820429408, "grad_norm": 2.3121135118228437, "learning_rate": 2.3722902659590653e-07, "loss": 0.2076, "step": 8406 }, { "epoch": 2.7348731294729993, "grad_norm": 2.3278577751598375, "learning_rate": 2.3665327224537182e-07, "loss": 0.1863, "step": 8407 }, { "epoch": 2.7351984385165906, "grad_norm": 2.0489756896936653, "learning_rate": 2.3607820048559937e-07, "loss": 0.1581, "step": 8408 }, { "epoch": 2.7355237475601824, "grad_norm": 2.0841146367297902, "learning_rate": 2.355038113989977e-07, "loss": 0.13, "step": 8409 }, { "epoch": 2.7358490566037736, "grad_norm": 2.6458322065132585, "learning_rate": 2.3493010506787538e-07, "loss": 0.1726, "step": 8410 }, { "epoch": 2.736174365647365, "grad_norm": 2.0837038713065326, "learning_rate": 2.3435708157444725e-07, "loss": 0.1842, "step": 8411 }, { "epoch": 2.7364996746909567, "grad_norm": 2.1159853986912704, "learning_rate": 2.3378474100082537e-07, "loss": 0.1477, "step": 8412 }, { "epoch": 2.736824983734548, "grad_norm": 2.173056417331016, "learning_rate": 2.332130834290286e-07, "loss": 0.1454, "step": 8413 }, { "epoch": 2.7371502927781393, "grad_norm": 2.3457859851072356, "learning_rate": 2.3264210894097527e-07, "loss": 0.1962, "step": 8414 }, { "epoch": 2.7374756018217306, "grad_norm": 2.3573407369604014, "learning_rate": 2.3207181761848664e-07, "loss": 0.1374, "step": 8415 }, { "epoch": 2.737800910865322, "grad_norm": 2.3553639675456015, "learning_rate": 2.3150220954328396e-07, "loss": 0.1678, "step": 8416 }, { "epoch": 2.7381262199089136, "grad_norm": 2.2217170275780753, "learning_rate": 2.309332847969953e-07, "loss": 0.1807, "step": 8417 }, { "epoch": 2.738451528952505, "grad_norm": 2.106349545536516, "learning_rate": 2.3036504346114653e-07, "loss": 0.1788, "step": 8418 }, { "epoch": 2.738776837996096, "grad_norm": 2.2558614375255224, "learning_rate": 2.297974856171664e-07, "loss": 0.1646, "step": 8419 }, { "epoch": 2.739102147039688, "grad_norm": 1.896869344400003, "learning_rate": 2.2923061134638712e-07, "loss": 0.1405, "step": 8420 }, { "epoch": 2.739427456083279, "grad_norm": 2.18751027146885, "learning_rate": 2.2866442073004259e-07, "loss": 0.1971, "step": 8421 }, { "epoch": 2.7397527651268705, "grad_norm": 2.0914279422952284, "learning_rate": 2.280989138492673e-07, "loss": 0.1452, "step": 8422 }, { "epoch": 2.740078074170462, "grad_norm": 2.5606165530565557, "learning_rate": 2.275340907850998e-07, "loss": 0.2204, "step": 8423 }, { "epoch": 2.740403383214053, "grad_norm": 2.237831718626352, "learning_rate": 2.269699516184787e-07, "loss": 0.1406, "step": 8424 }, { "epoch": 2.740728692257645, "grad_norm": 2.05268924297231, "learning_rate": 2.2640649643024538e-07, "loss": 0.1272, "step": 8425 }, { "epoch": 2.741054001301236, "grad_norm": 2.0433479508866843, "learning_rate": 2.2584372530114473e-07, "loss": 0.1022, "step": 8426 }, { "epoch": 2.7413793103448274, "grad_norm": 2.5006860740510732, "learning_rate": 2.2528163831182115e-07, "loss": 0.2176, "step": 8427 }, { "epoch": 2.741704619388419, "grad_norm": 1.9387567269980348, "learning_rate": 2.2472023554282074e-07, "loss": 0.1269, "step": 8428 }, { "epoch": 2.7420299284320104, "grad_norm": 2.0790205695749817, "learning_rate": 2.2415951707459582e-07, "loss": 0.1625, "step": 8429 }, { "epoch": 2.7423552374756017, "grad_norm": 2.210581152945364, "learning_rate": 2.2359948298749657e-07, "loss": 0.1953, "step": 8430 }, { "epoch": 2.7426805465191935, "grad_norm": 2.0556881331338035, "learning_rate": 2.2304013336177544e-07, "loss": 0.1356, "step": 8431 }, { "epoch": 2.7430058555627848, "grad_norm": 1.7176862002589577, "learning_rate": 2.2248146827758775e-07, "loss": 0.1082, "step": 8432 }, { "epoch": 2.743331164606376, "grad_norm": 2.064199853404822, "learning_rate": 2.2192348781499174e-07, "loss": 0.1684, "step": 8433 }, { "epoch": 2.7436564736499673, "grad_norm": 2.0204638508220407, "learning_rate": 2.2136619205394505e-07, "loss": 0.1781, "step": 8434 }, { "epoch": 2.7439817826935586, "grad_norm": 2.1295612459607893, "learning_rate": 2.208095810743094e-07, "loss": 0.188, "step": 8435 }, { "epoch": 2.7443070917371504, "grad_norm": 2.0784049444459445, "learning_rate": 2.2025365495584706e-07, "loss": 0.1621, "step": 8436 }, { "epoch": 2.7446324007807417, "grad_norm": 2.6131961501603174, "learning_rate": 2.1969841377822208e-07, "loss": 0.1533, "step": 8437 }, { "epoch": 2.744957709824333, "grad_norm": 2.486844102307299, "learning_rate": 2.191438576210031e-07, "loss": 0.1959, "step": 8438 }, { "epoch": 2.7452830188679247, "grad_norm": 2.0753889721815395, "learning_rate": 2.1858998656365538e-07, "loss": 0.1354, "step": 8439 }, { "epoch": 2.745608327911516, "grad_norm": 2.2661112203368634, "learning_rate": 2.18036800685551e-07, "loss": 0.1348, "step": 8440 }, { "epoch": 2.7459336369551073, "grad_norm": 1.9916091534887128, "learning_rate": 2.174843000659621e-07, "loss": 0.1578, "step": 8441 }, { "epoch": 2.746258945998699, "grad_norm": 2.00500234031165, "learning_rate": 2.1693248478406037e-07, "loss": 0.1235, "step": 8442 }, { "epoch": 2.7465842550422903, "grad_norm": 1.9587338926156284, "learning_rate": 2.1638135491892365e-07, "loss": 0.1387, "step": 8443 }, { "epoch": 2.7469095640858816, "grad_norm": 2.220464383302859, "learning_rate": 2.1583091054952765e-07, "loss": 0.196, "step": 8444 }, { "epoch": 2.747234873129473, "grad_norm": 1.9938773468531126, "learning_rate": 2.1528115175475207e-07, "loss": 0.127, "step": 8445 }, { "epoch": 2.747560182173064, "grad_norm": 2.2611822649083706, "learning_rate": 2.1473207861337776e-07, "loss": 0.1624, "step": 8446 }, { "epoch": 2.747885491216656, "grad_norm": 2.3530618145552324, "learning_rate": 2.1418369120408732e-07, "loss": 0.2065, "step": 8447 }, { "epoch": 2.748210800260247, "grad_norm": 1.8988793691620565, "learning_rate": 2.1363598960546294e-07, "loss": 0.1447, "step": 8448 }, { "epoch": 2.7485361093038385, "grad_norm": 2.019272613996408, "learning_rate": 2.130889738959946e-07, "loss": 0.1679, "step": 8449 }, { "epoch": 2.7488614183474303, "grad_norm": 1.9868557576861134, "learning_rate": 2.1254264415406733e-07, "loss": 0.1548, "step": 8450 }, { "epoch": 2.7491867273910215, "grad_norm": 2.1591135601944056, "learning_rate": 2.1199700045797077e-07, "loss": 0.1647, "step": 8451 }, { "epoch": 2.749512036434613, "grad_norm": 2.1230328556258278, "learning_rate": 2.1145204288589572e-07, "loss": 0.1487, "step": 8452 }, { "epoch": 2.749837345478204, "grad_norm": 2.9582216839482744, "learning_rate": 2.109077715159369e-07, "loss": 0.2039, "step": 8453 }, { "epoch": 2.750162654521796, "grad_norm": 2.363819457190731, "learning_rate": 2.1036418642608635e-07, "loss": 0.2009, "step": 8454 }, { "epoch": 2.750487963565387, "grad_norm": 2.2094331671146694, "learning_rate": 2.0982128769424238e-07, "loss": 0.1323, "step": 8455 }, { "epoch": 2.7508132726089785, "grad_norm": 1.7873446559647126, "learning_rate": 2.0927907539820047e-07, "loss": 0.1111, "step": 8456 }, { "epoch": 2.7511385816525697, "grad_norm": 1.9696379797878387, "learning_rate": 2.0873754961566183e-07, "loss": 0.1574, "step": 8457 }, { "epoch": 2.7514638906961615, "grad_norm": 2.3052813786135506, "learning_rate": 2.0819671042422663e-07, "loss": 0.1797, "step": 8458 }, { "epoch": 2.751789199739753, "grad_norm": 1.9180261687135132, "learning_rate": 2.0765655790139782e-07, "loss": 0.1765, "step": 8459 }, { "epoch": 2.752114508783344, "grad_norm": 2.923919365650144, "learning_rate": 2.071170921245791e-07, "loss": 0.1534, "step": 8460 }, { "epoch": 2.752439817826936, "grad_norm": 2.2342721896949436, "learning_rate": 2.0657831317107745e-07, "loss": 0.1797, "step": 8461 }, { "epoch": 2.752765126870527, "grad_norm": 2.1599525411315854, "learning_rate": 2.060402211180984e-07, "loss": 0.1533, "step": 8462 }, { "epoch": 2.7530904359141184, "grad_norm": 2.3140648512441477, "learning_rate": 2.055028160427519e-07, "loss": 0.1267, "step": 8463 }, { "epoch": 2.7534157449577097, "grad_norm": 2.0403198171279873, "learning_rate": 2.049660980220497e-07, "loss": 0.1607, "step": 8464 }, { "epoch": 2.753741054001301, "grad_norm": 2.428674327953179, "learning_rate": 2.0443006713290137e-07, "loss": 0.1915, "step": 8465 }, { "epoch": 2.7540663630448927, "grad_norm": 2.06953165510426, "learning_rate": 2.0389472345212325e-07, "loss": 0.1524, "step": 8466 }, { "epoch": 2.754391672088484, "grad_norm": 1.7930853850643196, "learning_rate": 2.033600670564273e-07, "loss": 0.1371, "step": 8467 }, { "epoch": 2.7547169811320753, "grad_norm": 2.381924351347268, "learning_rate": 2.0282609802243335e-07, "loss": 0.2147, "step": 8468 }, { "epoch": 2.755042290175667, "grad_norm": 1.925523724245848, "learning_rate": 2.022928164266569e-07, "loss": 0.1655, "step": 8469 }, { "epoch": 2.7553675992192583, "grad_norm": 1.7058007656711358, "learning_rate": 2.0176022234551894e-07, "loss": 0.0982, "step": 8470 }, { "epoch": 2.7556929082628496, "grad_norm": 1.698840412312072, "learning_rate": 2.012283158553402e-07, "loss": 0.1243, "step": 8471 }, { "epoch": 2.7560182173064414, "grad_norm": 2.3134338079342105, "learning_rate": 2.0069709703234297e-07, "loss": 0.176, "step": 8472 }, { "epoch": 2.7563435263500327, "grad_norm": 1.9345917797337318, "learning_rate": 2.0016656595265248e-07, "loss": 0.1519, "step": 8473 }, { "epoch": 2.756668835393624, "grad_norm": 2.0562227724523097, "learning_rate": 1.996367226922924e-07, "loss": 0.1571, "step": 8474 }, { "epoch": 2.7569941444372152, "grad_norm": 2.003608876718619, "learning_rate": 1.9910756732719083e-07, "loss": 0.1357, "step": 8475 }, { "epoch": 2.7573194534808065, "grad_norm": 2.4492853973279014, "learning_rate": 1.9857909993317602e-07, "loss": 0.1897, "step": 8476 }, { "epoch": 2.7576447625243983, "grad_norm": 1.856376887181477, "learning_rate": 1.9805132058597686e-07, "loss": 0.1288, "step": 8477 }, { "epoch": 2.7579700715679896, "grad_norm": 2.039755767658607, "learning_rate": 1.9752422936122562e-07, "loss": 0.1235, "step": 8478 }, { "epoch": 2.758295380611581, "grad_norm": 2.094907348448045, "learning_rate": 1.9699782633445353e-07, "loss": 0.1721, "step": 8479 }, { "epoch": 2.7586206896551726, "grad_norm": 2.041484124084684, "learning_rate": 1.964721115810958e-07, "loss": 0.127, "step": 8480 }, { "epoch": 2.758945998698764, "grad_norm": 2.186364435976736, "learning_rate": 1.959470851764872e-07, "loss": 0.1697, "step": 8481 }, { "epoch": 2.759271307742355, "grad_norm": 2.1451488101880045, "learning_rate": 1.9542274719586475e-07, "loss": 0.171, "step": 8482 }, { "epoch": 2.759596616785947, "grad_norm": 2.2598467822595905, "learning_rate": 1.9489909771436444e-07, "loss": 0.1718, "step": 8483 }, { "epoch": 2.759921925829538, "grad_norm": 2.6698729996978545, "learning_rate": 1.943761368070285e-07, "loss": 0.2172, "step": 8484 }, { "epoch": 2.7602472348731295, "grad_norm": 2.3383896941463806, "learning_rate": 1.9385386454879584e-07, "loss": 0.2016, "step": 8485 }, { "epoch": 2.760572543916721, "grad_norm": 2.0913602287938677, "learning_rate": 1.9333228101450884e-07, "loss": 0.1124, "step": 8486 }, { "epoch": 2.760897852960312, "grad_norm": 1.7478287521763602, "learning_rate": 1.9281138627890994e-07, "loss": 0.14, "step": 8487 }, { "epoch": 2.761223162003904, "grad_norm": 2.0071523517356638, "learning_rate": 1.9229118041664553e-07, "loss": 0.176, "step": 8488 }, { "epoch": 2.761548471047495, "grad_norm": 2.0326584854946272, "learning_rate": 1.9177166350225983e-07, "loss": 0.1569, "step": 8489 }, { "epoch": 2.7618737800910864, "grad_norm": 2.320500870653142, "learning_rate": 1.9125283561020114e-07, "loss": 0.1771, "step": 8490 }, { "epoch": 2.762199089134678, "grad_norm": 1.9918051208802878, "learning_rate": 1.907346968148166e-07, "loss": 0.1565, "step": 8491 }, { "epoch": 2.7625243981782694, "grad_norm": 2.0953061355661036, "learning_rate": 1.902172471903563e-07, "loss": 0.1481, "step": 8492 }, { "epoch": 2.7628497072218607, "grad_norm": 2.2838479598579484, "learning_rate": 1.8970048681097253e-07, "loss": 0.1657, "step": 8493 }, { "epoch": 2.763175016265452, "grad_norm": 2.580072600673427, "learning_rate": 1.89184415750715e-07, "loss": 0.2222, "step": 8494 }, { "epoch": 2.7635003253090433, "grad_norm": 2.38068523413836, "learning_rate": 1.8866903408353844e-07, "loss": 0.1673, "step": 8495 }, { "epoch": 2.763825634352635, "grad_norm": 2.2614813298266916, "learning_rate": 1.8815434188329707e-07, "loss": 0.1471, "step": 8496 }, { "epoch": 2.7641509433962264, "grad_norm": 1.8343260266005221, "learning_rate": 1.8764033922374692e-07, "loss": 0.1317, "step": 8497 }, { "epoch": 2.7644762524398176, "grad_norm": 1.9972769056079518, "learning_rate": 1.8712702617854517e-07, "loss": 0.1461, "step": 8498 }, { "epoch": 2.7648015614834094, "grad_norm": 1.9187274801022893, "learning_rate": 1.8661440282124853e-07, "loss": 0.1201, "step": 8499 }, { "epoch": 2.7651268705270007, "grad_norm": 2.580071559950045, "learning_rate": 1.8610246922531771e-07, "loss": 0.2322, "step": 8500 }, { "epoch": 2.765452179570592, "grad_norm": 1.993223960524627, "learning_rate": 1.855912254641129e-07, "loss": 0.1506, "step": 8501 }, { "epoch": 2.7657774886141837, "grad_norm": 1.9040114688511993, "learning_rate": 1.8508067161089494e-07, "loss": 0.1533, "step": 8502 }, { "epoch": 2.766102797657775, "grad_norm": 2.5709717606112807, "learning_rate": 1.845708077388264e-07, "loss": 0.1861, "step": 8503 }, { "epoch": 2.7664281067013663, "grad_norm": 2.4930144906202507, "learning_rate": 1.8406163392097277e-07, "loss": 0.1759, "step": 8504 }, { "epoch": 2.7667534157449576, "grad_norm": 2.046713276633908, "learning_rate": 1.8355315023029785e-07, "loss": 0.1151, "step": 8505 }, { "epoch": 2.767078724788549, "grad_norm": 2.0775465008444423, "learning_rate": 1.830453567396673e-07, "loss": 0.1641, "step": 8506 }, { "epoch": 2.7674040338321406, "grad_norm": 2.239773952553511, "learning_rate": 1.825382535218484e-07, "loss": 0.1836, "step": 8507 }, { "epoch": 2.767729342875732, "grad_norm": 2.129690228170761, "learning_rate": 1.8203184064951085e-07, "loss": 0.154, "step": 8508 }, { "epoch": 2.768054651919323, "grad_norm": 1.8926387496647503, "learning_rate": 1.815261181952216e-07, "loss": 0.1289, "step": 8509 }, { "epoch": 2.768379960962915, "grad_norm": 1.9395447668533152, "learning_rate": 1.810210862314532e-07, "loss": 0.1543, "step": 8510 }, { "epoch": 2.7687052700065062, "grad_norm": 2.1567124622795064, "learning_rate": 1.8051674483057557e-07, "loss": 0.1953, "step": 8511 }, { "epoch": 2.7690305790500975, "grad_norm": 2.343719097127925, "learning_rate": 1.8001309406486145e-07, "loss": 0.1608, "step": 8512 }, { "epoch": 2.7693558880936893, "grad_norm": 2.3715751795757622, "learning_rate": 1.7951013400648531e-07, "loss": 0.1511, "step": 8513 }, { "epoch": 2.7696811971372806, "grad_norm": 2.1487626364751238, "learning_rate": 1.790078647275212e-07, "loss": 0.1855, "step": 8514 }, { "epoch": 2.770006506180872, "grad_norm": 2.17244113543436, "learning_rate": 1.785062862999426e-07, "loss": 0.1819, "step": 8515 }, { "epoch": 2.770331815224463, "grad_norm": 2.0165769203974695, "learning_rate": 1.7800539879562927e-07, "loss": 0.1372, "step": 8516 }, { "epoch": 2.7706571242680544, "grad_norm": 2.1730584860485487, "learning_rate": 1.775052022863577e-07, "loss": 0.1437, "step": 8517 }, { "epoch": 2.770982433311646, "grad_norm": 2.0727726342856325, "learning_rate": 1.7700569684380498e-07, "loss": 0.1651, "step": 8518 }, { "epoch": 2.7713077423552375, "grad_norm": 2.0126410894762774, "learning_rate": 1.7650688253955107e-07, "loss": 0.1277, "step": 8519 }, { "epoch": 2.7716330513988288, "grad_norm": 2.077045440021775, "learning_rate": 1.7600875944507822e-07, "loss": 0.1571, "step": 8520 }, { "epoch": 2.7719583604424205, "grad_norm": 2.185186229270963, "learning_rate": 1.755113276317655e-07, "loss": 0.1621, "step": 8521 }, { "epoch": 2.772283669486012, "grad_norm": 2.262838581221027, "learning_rate": 1.750145871708969e-07, "loss": 0.1633, "step": 8522 }, { "epoch": 2.772608978529603, "grad_norm": 2.051416611078408, "learning_rate": 1.7451853813365393e-07, "loss": 0.1383, "step": 8523 }, { "epoch": 2.7729342875731944, "grad_norm": 2.121402973948653, "learning_rate": 1.740231805911219e-07, "loss": 0.1486, "step": 8524 }, { "epoch": 2.7732595966167857, "grad_norm": 2.0871570773333916, "learning_rate": 1.7352851461428622e-07, "loss": 0.191, "step": 8525 }, { "epoch": 2.7735849056603774, "grad_norm": 1.9480633904144926, "learning_rate": 1.7303454027403243e-07, "loss": 0.158, "step": 8526 }, { "epoch": 2.7739102147039687, "grad_norm": 2.075348060390974, "learning_rate": 1.725412576411467e-07, "loss": 0.153, "step": 8527 }, { "epoch": 2.77423552374756, "grad_norm": 1.8621093919687473, "learning_rate": 1.7204866678631804e-07, "loss": 0.1344, "step": 8528 }, { "epoch": 2.7745608327911517, "grad_norm": 2.313209670067461, "learning_rate": 1.7155676778013387e-07, "loss": 0.1962, "step": 8529 }, { "epoch": 2.774886141834743, "grad_norm": 2.0636360418822863, "learning_rate": 1.7106556069308444e-07, "loss": 0.1773, "step": 8530 }, { "epoch": 2.7752114508783343, "grad_norm": 2.336479733973356, "learning_rate": 1.705750455955596e-07, "loss": 0.1837, "step": 8531 }, { "epoch": 2.775536759921926, "grad_norm": 2.141209058650013, "learning_rate": 1.700852225578503e-07, "loss": 0.1563, "step": 8532 }, { "epoch": 2.7758620689655173, "grad_norm": 1.9406243313223352, "learning_rate": 1.6959609165014934e-07, "loss": 0.1488, "step": 8533 }, { "epoch": 2.7761873780091086, "grad_norm": 2.201748601660185, "learning_rate": 1.691076529425495e-07, "loss": 0.1501, "step": 8534 }, { "epoch": 2.7765126870527, "grad_norm": 2.0743330067141903, "learning_rate": 1.6861990650504256e-07, "loss": 0.158, "step": 8535 }, { "epoch": 2.7768379960962912, "grad_norm": 2.0980477720382416, "learning_rate": 1.6813285240752487e-07, "loss": 0.1594, "step": 8536 }, { "epoch": 2.777163305139883, "grad_norm": 2.243308616800356, "learning_rate": 1.6764649071979165e-07, "loss": 0.1372, "step": 8537 }, { "epoch": 2.7774886141834743, "grad_norm": 2.1196127329143923, "learning_rate": 1.6716082151153724e-07, "loss": 0.1403, "step": 8538 }, { "epoch": 2.7778139232270656, "grad_norm": 1.8518794247007486, "learning_rate": 1.6667584485235977e-07, "loss": 0.1469, "step": 8539 }, { "epoch": 2.7781392322706573, "grad_norm": 2.0690424708137134, "learning_rate": 1.6619156081175646e-07, "loss": 0.1679, "step": 8540 }, { "epoch": 2.7784645413142486, "grad_norm": 2.069286861000066, "learning_rate": 1.6570796945912458e-07, "loss": 0.1491, "step": 8541 }, { "epoch": 2.77878985035784, "grad_norm": 2.258051480088088, "learning_rate": 1.6522507086376483e-07, "loss": 0.1592, "step": 8542 }, { "epoch": 2.7791151594014316, "grad_norm": 1.7428690071437518, "learning_rate": 1.6474286509487515e-07, "loss": 0.1197, "step": 8543 }, { "epoch": 2.779440468445023, "grad_norm": 2.402935164505176, "learning_rate": 1.6426135222155637e-07, "loss": 0.1557, "step": 8544 }, { "epoch": 2.779765777488614, "grad_norm": 1.9682207643539948, "learning_rate": 1.637805323128111e-07, "loss": 0.1721, "step": 8545 }, { "epoch": 2.7800910865322055, "grad_norm": 2.1101866251381933, "learning_rate": 1.633004054375392e-07, "loss": 0.1566, "step": 8546 }, { "epoch": 2.780416395575797, "grad_norm": 1.9249028933117527, "learning_rate": 1.6282097166454346e-07, "loss": 0.1829, "step": 8547 }, { "epoch": 2.7807417046193885, "grad_norm": 2.316902008237845, "learning_rate": 1.6234223106252888e-07, "loss": 0.206, "step": 8548 }, { "epoch": 2.78106701366298, "grad_norm": 2.394329311102216, "learning_rate": 1.6186418370009727e-07, "loss": 0.1697, "step": 8549 }, { "epoch": 2.781392322706571, "grad_norm": 2.0862316053182406, "learning_rate": 1.6138682964575436e-07, "loss": 0.1807, "step": 8550 }, { "epoch": 2.781717631750163, "grad_norm": 2.1037687317952645, "learning_rate": 1.6091016896790434e-07, "loss": 0.1966, "step": 8551 }, { "epoch": 2.782042940793754, "grad_norm": 1.9082447474841455, "learning_rate": 1.6043420173485313e-07, "loss": 0.1454, "step": 8552 }, { "epoch": 2.7823682498373454, "grad_norm": 2.0768088115292427, "learning_rate": 1.599589280148084e-07, "loss": 0.1432, "step": 8553 }, { "epoch": 2.7826935588809367, "grad_norm": 2.2289729879114013, "learning_rate": 1.5948434787587618e-07, "loss": 0.1629, "step": 8554 }, { "epoch": 2.7830188679245285, "grad_norm": 2.254672891148195, "learning_rate": 1.5901046138606324e-07, "loss": 0.1489, "step": 8555 }, { "epoch": 2.7833441769681198, "grad_norm": 2.0954691771316316, "learning_rate": 1.5853726861327968e-07, "loss": 0.1347, "step": 8556 }, { "epoch": 2.783669486011711, "grad_norm": 2.1541289732231417, "learning_rate": 1.5806476962533347e-07, "loss": 0.1394, "step": 8557 }, { "epoch": 2.7839947950553023, "grad_norm": 2.2898272454347777, "learning_rate": 1.5759296448993323e-07, "loss": 0.1763, "step": 8558 }, { "epoch": 2.784320104098894, "grad_norm": 1.7496420055279815, "learning_rate": 1.5712185327469042e-07, "loss": 0.1352, "step": 8559 }, { "epoch": 2.7846454131424854, "grad_norm": 1.6400375432607566, "learning_rate": 1.5665143604711497e-07, "loss": 0.135, "step": 8560 }, { "epoch": 2.7849707221860767, "grad_norm": 1.8614736730612682, "learning_rate": 1.5618171287461737e-07, "loss": 0.1325, "step": 8561 }, { "epoch": 2.7852960312296684, "grad_norm": 1.5079154865870428, "learning_rate": 1.5571268382451098e-07, "loss": 0.1135, "step": 8562 }, { "epoch": 2.7856213402732597, "grad_norm": 2.1737610049776106, "learning_rate": 1.5524434896400542e-07, "loss": 0.1589, "step": 8563 }, { "epoch": 2.785946649316851, "grad_norm": 1.8996352127313239, "learning_rate": 1.547767083602153e-07, "loss": 0.1541, "step": 8564 }, { "epoch": 2.7862719583604423, "grad_norm": 1.6217533465052067, "learning_rate": 1.5430976208015425e-07, "loss": 0.114, "step": 8565 }, { "epoch": 2.7865972674040336, "grad_norm": 2.388331362077297, "learning_rate": 1.538435101907343e-07, "loss": 0.2365, "step": 8566 }, { "epoch": 2.7869225764476253, "grad_norm": 1.945119110758784, "learning_rate": 1.5337795275876975e-07, "loss": 0.1249, "step": 8567 }, { "epoch": 2.7872478854912166, "grad_norm": 2.0262732888086274, "learning_rate": 1.5291308985097675e-07, "loss": 0.1593, "step": 8568 }, { "epoch": 2.787573194534808, "grad_norm": 1.8709962526016095, "learning_rate": 1.5244892153397028e-07, "loss": 0.1518, "step": 8569 }, { "epoch": 2.7878985035783996, "grad_norm": 2.0782425258604964, "learning_rate": 1.519854478742644e-07, "loss": 0.1802, "step": 8570 }, { "epoch": 2.788223812621991, "grad_norm": 2.0008154491149823, "learning_rate": 1.5152266893827593e-07, "loss": 0.1598, "step": 8571 }, { "epoch": 2.788549121665582, "grad_norm": 1.889279356118514, "learning_rate": 1.5106058479232243e-07, "loss": 0.1779, "step": 8572 }, { "epoch": 2.788874430709174, "grad_norm": 2.0989119755259056, "learning_rate": 1.505991955026198e-07, "loss": 0.163, "step": 8573 }, { "epoch": 2.7891997397527653, "grad_norm": 1.8647716188925778, "learning_rate": 1.5013850113528573e-07, "loss": 0.15, "step": 8574 }, { "epoch": 2.7895250487963565, "grad_norm": 2.11278588196714, "learning_rate": 1.4967850175633736e-07, "loss": 0.1948, "step": 8575 }, { "epoch": 2.789850357839948, "grad_norm": 2.058425464984165, "learning_rate": 1.4921919743169421e-07, "loss": 0.1589, "step": 8576 }, { "epoch": 2.790175666883539, "grad_norm": 2.5715396100334047, "learning_rate": 1.4876058822717422e-07, "loss": 0.1933, "step": 8577 }, { "epoch": 2.790500975927131, "grad_norm": 1.9363250713636666, "learning_rate": 1.4830267420849587e-07, "loss": 0.1422, "step": 8578 }, { "epoch": 2.790826284970722, "grad_norm": 2.070216327352956, "learning_rate": 1.478454554412795e-07, "loss": 0.1584, "step": 8579 }, { "epoch": 2.7911515940143135, "grad_norm": 2.072663056156279, "learning_rate": 1.4738893199104487e-07, "loss": 0.1279, "step": 8580 }, { "epoch": 2.791476903057905, "grad_norm": 2.109787999129389, "learning_rate": 1.4693310392321081e-07, "loss": 0.1352, "step": 8581 }, { "epoch": 2.7918022121014965, "grad_norm": 2.241815736422823, "learning_rate": 1.464779713030995e-07, "loss": 0.2052, "step": 8582 }, { "epoch": 2.7921275211450878, "grad_norm": 1.996420052113909, "learning_rate": 1.4602353419592985e-07, "loss": 0.12, "step": 8583 }, { "epoch": 2.7924528301886795, "grad_norm": 2.025104557339622, "learning_rate": 1.4556979266682424e-07, "loss": 0.1437, "step": 8584 }, { "epoch": 2.792778139232271, "grad_norm": 1.877852017746581, "learning_rate": 1.4511674678080457e-07, "loss": 0.1116, "step": 8585 }, { "epoch": 2.793103448275862, "grad_norm": 1.9264190309185327, "learning_rate": 1.4466439660279164e-07, "loss": 0.1537, "step": 8586 }, { "epoch": 2.7934287573194534, "grad_norm": 2.1503122572185505, "learning_rate": 1.4421274219760806e-07, "loss": 0.1379, "step": 8587 }, { "epoch": 2.7937540663630447, "grad_norm": 2.3915556401904134, "learning_rate": 1.4376178362997596e-07, "loss": 0.1442, "step": 8588 }, { "epoch": 2.7940793754066364, "grad_norm": 2.0674310863594494, "learning_rate": 1.4331152096451807e-07, "loss": 0.1511, "step": 8589 }, { "epoch": 2.7944046844502277, "grad_norm": 2.2666440184581838, "learning_rate": 1.4286195426575666e-07, "loss": 0.2056, "step": 8590 }, { "epoch": 2.794729993493819, "grad_norm": 2.3699428992214, "learning_rate": 1.4241308359811633e-07, "loss": 0.1891, "step": 8591 }, { "epoch": 2.7950553025374107, "grad_norm": 1.9522639974774463, "learning_rate": 1.4196490902592008e-07, "loss": 0.1363, "step": 8592 }, { "epoch": 2.795380611581002, "grad_norm": 1.883399014920811, "learning_rate": 1.4151743061339096e-07, "loss": 0.1263, "step": 8593 }, { "epoch": 2.7957059206245933, "grad_norm": 1.8660412812489344, "learning_rate": 1.4107064842465324e-07, "loss": 0.199, "step": 8594 }, { "epoch": 2.7960312296681846, "grad_norm": 1.9073505992777084, "learning_rate": 1.4062456252373125e-07, "loss": 0.136, "step": 8595 }, { "epoch": 2.796356538711776, "grad_norm": 2.1745532363389453, "learning_rate": 1.401791729745494e-07, "loss": 0.1616, "step": 8596 }, { "epoch": 2.7966818477553677, "grad_norm": 2.1195175270273516, "learning_rate": 1.3973447984093224e-07, "loss": 0.1476, "step": 8597 }, { "epoch": 2.797007156798959, "grad_norm": 2.0743439699456836, "learning_rate": 1.3929048318660433e-07, "loss": 0.1214, "step": 8598 }, { "epoch": 2.7973324658425502, "grad_norm": 2.1473931387931917, "learning_rate": 1.388471830751914e-07, "loss": 0.2272, "step": 8599 }, { "epoch": 2.797657774886142, "grad_norm": 1.8785076726829242, "learning_rate": 1.3840457957021825e-07, "loss": 0.1866, "step": 8600 }, { "epoch": 2.7979830839297333, "grad_norm": 2.044612971139091, "learning_rate": 1.3796267273511023e-07, "loss": 0.1481, "step": 8601 }, { "epoch": 2.7983083929733246, "grad_norm": 2.595006523168898, "learning_rate": 1.3752146263319166e-07, "loss": 0.2162, "step": 8602 }, { "epoch": 2.7986337020169163, "grad_norm": 2.1984141828233286, "learning_rate": 1.370809493276909e-07, "loss": 0.1578, "step": 8603 }, { "epoch": 2.7989590110605076, "grad_norm": 1.799844040802961, "learning_rate": 1.3664113288173185e-07, "loss": 0.1349, "step": 8604 }, { "epoch": 2.799284320104099, "grad_norm": 1.829497372652135, "learning_rate": 1.3620201335834082e-07, "loss": 0.1379, "step": 8605 }, { "epoch": 2.79960962914769, "grad_norm": 2.099673958199561, "learning_rate": 1.357635908204441e-07, "loss": 0.1465, "step": 8606 }, { "epoch": 2.7999349381912815, "grad_norm": 2.2746132551149882, "learning_rate": 1.3532586533086866e-07, "loss": 0.1483, "step": 8607 }, { "epoch": 2.800260247234873, "grad_norm": 1.975355488575297, "learning_rate": 1.3488883695233934e-07, "loss": 0.1673, "step": 8608 }, { "epoch": 2.8005855562784645, "grad_norm": 2.034046126937067, "learning_rate": 1.3445250574748437e-07, "loss": 0.1235, "step": 8609 }, { "epoch": 2.800910865322056, "grad_norm": 2.3205239736046215, "learning_rate": 1.3401687177882817e-07, "loss": 0.1543, "step": 8610 }, { "epoch": 2.8012361743656475, "grad_norm": 2.3623535190325367, "learning_rate": 1.3358193510879912e-07, "loss": 0.1555, "step": 8611 }, { "epoch": 2.801561483409239, "grad_norm": 2.1387103570422847, "learning_rate": 1.3314769579972354e-07, "loss": 0.1631, "step": 8612 }, { "epoch": 2.80188679245283, "grad_norm": 2.162141605142109, "learning_rate": 1.3271415391382714e-07, "loss": 0.1443, "step": 8613 }, { "epoch": 2.802212101496422, "grad_norm": 1.9932162410274008, "learning_rate": 1.3228130951323804e-07, "loss": 0.1205, "step": 8614 }, { "epoch": 2.802537410540013, "grad_norm": 2.1726890016281404, "learning_rate": 1.3184916265998325e-07, "loss": 0.1513, "step": 8615 }, { "epoch": 2.8028627195836044, "grad_norm": 2.092080712037024, "learning_rate": 1.3141771341598885e-07, "loss": 0.1149, "step": 8616 }, { "epoch": 2.8031880286271957, "grad_norm": 2.1065265268153843, "learning_rate": 1.3098696184308257e-07, "loss": 0.1512, "step": 8617 }, { "epoch": 2.803513337670787, "grad_norm": 1.9458612547184018, "learning_rate": 1.305569080029906e-07, "loss": 0.153, "step": 8618 }, { "epoch": 2.8038386467143788, "grad_norm": 2.0801052900237504, "learning_rate": 1.3012755195733972e-07, "loss": 0.1789, "step": 8619 }, { "epoch": 2.80416395575797, "grad_norm": 2.2891847564850494, "learning_rate": 1.2969889376765853e-07, "loss": 0.1866, "step": 8620 }, { "epoch": 2.8044892648015614, "grad_norm": 2.2035264178861453, "learning_rate": 1.292709334953729e-07, "loss": 0.1621, "step": 8621 }, { "epoch": 2.804814573845153, "grad_norm": 2.043417511193732, "learning_rate": 1.2884367120180873e-07, "loss": 0.1429, "step": 8622 }, { "epoch": 2.8051398828887444, "grad_norm": 2.1274961736440403, "learning_rate": 1.28417106948196e-07, "loss": 0.1519, "step": 8623 }, { "epoch": 2.8054651919323357, "grad_norm": 2.252526298490976, "learning_rate": 1.2799124079565904e-07, "loss": 0.1664, "step": 8624 }, { "epoch": 2.805790500975927, "grad_norm": 1.936958539646664, "learning_rate": 1.2756607280522527e-07, "loss": 0.1327, "step": 8625 }, { "epoch": 2.8061158100195187, "grad_norm": 2.253510146293505, "learning_rate": 1.2714160303782196e-07, "loss": 0.1284, "step": 8626 }, { "epoch": 2.80644111906311, "grad_norm": 1.9196179733700027, "learning_rate": 1.2671783155427663e-07, "loss": 0.1299, "step": 8627 }, { "epoch": 2.8067664281067013, "grad_norm": 1.6660062794602681, "learning_rate": 1.2629475841531402e-07, "loss": 0.1101, "step": 8628 }, { "epoch": 2.8070917371502926, "grad_norm": 2.3750679369656447, "learning_rate": 1.2587238368156285e-07, "loss": 0.1781, "step": 8629 }, { "epoch": 2.8074170461938843, "grad_norm": 2.5978123358327876, "learning_rate": 1.254507074135486e-07, "loss": 0.1664, "step": 8630 }, { "epoch": 2.8077423552374756, "grad_norm": 1.901566027440011, "learning_rate": 1.2502972967169792e-07, "loss": 0.1233, "step": 8631 }, { "epoch": 2.808067664281067, "grad_norm": 1.9820319040412315, "learning_rate": 1.2460945051633754e-07, "loss": 0.144, "step": 8632 }, { "epoch": 2.8083929733246586, "grad_norm": 2.1614144207174806, "learning_rate": 1.2418987000769268e-07, "loss": 0.1628, "step": 8633 }, { "epoch": 2.80871828236825, "grad_norm": 2.014607621255343, "learning_rate": 1.2377098820589072e-07, "loss": 0.1455, "step": 8634 }, { "epoch": 2.8090435914118412, "grad_norm": 2.4634552465299278, "learning_rate": 1.233528051709576e-07, "loss": 0.2485, "step": 8635 }, { "epoch": 2.8093689004554325, "grad_norm": 2.013524147498112, "learning_rate": 1.2293532096281868e-07, "loss": 0.1463, "step": 8636 }, { "epoch": 2.809694209499024, "grad_norm": 2.170089912199244, "learning_rate": 1.225185356413e-07, "loss": 0.1665, "step": 8637 }, { "epoch": 2.8100195185426156, "grad_norm": 1.6777576185489727, "learning_rate": 1.2210244926612659e-07, "loss": 0.1363, "step": 8638 }, { "epoch": 2.810344827586207, "grad_norm": 2.1930608678675134, "learning_rate": 1.2168706189692515e-07, "loss": 0.1792, "step": 8639 }, { "epoch": 2.810670136629798, "grad_norm": 2.1896836428639816, "learning_rate": 1.212723735932203e-07, "loss": 0.1768, "step": 8640 }, { "epoch": 2.81099544567339, "grad_norm": 2.2106330056046826, "learning_rate": 1.2085838441443676e-07, "loss": 0.1907, "step": 8641 }, { "epoch": 2.811320754716981, "grad_norm": 2.156824087792616, "learning_rate": 1.2044509441989982e-07, "loss": 0.1495, "step": 8642 }, { "epoch": 2.8116460637605725, "grad_norm": 2.2727637624270174, "learning_rate": 1.2003250366883378e-07, "loss": 0.1973, "step": 8643 }, { "epoch": 2.811971372804164, "grad_norm": 2.0076256652026077, "learning_rate": 1.196206122203647e-07, "loss": 0.1461, "step": 8644 }, { "epoch": 2.8122966818477555, "grad_norm": 2.163099527038737, "learning_rate": 1.1920942013351477e-07, "loss": 0.1536, "step": 8645 }, { "epoch": 2.812621990891347, "grad_norm": 1.8954429216433484, "learning_rate": 1.1879892746720911e-07, "loss": 0.1489, "step": 8646 }, { "epoch": 2.812947299934938, "grad_norm": 2.017621305564225, "learning_rate": 1.1838913428027232e-07, "loss": 0.1464, "step": 8647 }, { "epoch": 2.8132726089785294, "grad_norm": 2.1260105339390694, "learning_rate": 1.1798004063142688e-07, "loss": 0.1476, "step": 8648 }, { "epoch": 2.813597918022121, "grad_norm": 2.2115994926020854, "learning_rate": 1.1757164657929698e-07, "loss": 0.1666, "step": 8649 }, { "epoch": 2.8139232270657124, "grad_norm": 2.0523520257796704, "learning_rate": 1.1716395218240473e-07, "loss": 0.1861, "step": 8650 }, { "epoch": 2.8142485361093037, "grad_norm": 2.316737166808619, "learning_rate": 1.1675695749917337e-07, "loss": 0.2095, "step": 8651 }, { "epoch": 2.8145738451528954, "grad_norm": 1.7348567108817432, "learning_rate": 1.1635066258792682e-07, "loss": 0.1688, "step": 8652 }, { "epoch": 2.8148991541964867, "grad_norm": 2.204950286510617, "learning_rate": 1.1594506750688628e-07, "loss": 0.1764, "step": 8653 }, { "epoch": 2.815224463240078, "grad_norm": 2.298312565333849, "learning_rate": 1.1554017231417303e-07, "loss": 0.1653, "step": 8654 }, { "epoch": 2.8155497722836698, "grad_norm": 2.121266631066779, "learning_rate": 1.1513597706781065e-07, "loss": 0.1608, "step": 8655 }, { "epoch": 2.815875081327261, "grad_norm": 2.117690614083964, "learning_rate": 1.147324818257195e-07, "loss": 0.139, "step": 8656 }, { "epoch": 2.8162003903708523, "grad_norm": 1.8221387765594943, "learning_rate": 1.1432968664572052e-07, "loss": 0.1475, "step": 8657 }, { "epoch": 2.8165256994144436, "grad_norm": 1.7655836466533625, "learning_rate": 1.1392759158553479e-07, "loss": 0.1516, "step": 8658 }, { "epoch": 2.816851008458035, "grad_norm": 2.1148959439842163, "learning_rate": 1.1352619670278287e-07, "loss": 0.1365, "step": 8659 }, { "epoch": 2.8171763175016267, "grad_norm": 2.0415213226740607, "learning_rate": 1.1312550205498485e-07, "loss": 0.1362, "step": 8660 }, { "epoch": 2.817501626545218, "grad_norm": 2.3890734390636332, "learning_rate": 1.1272550769956148e-07, "loss": 0.1679, "step": 8661 }, { "epoch": 2.8178269355888093, "grad_norm": 2.002923772165936, "learning_rate": 1.1232621369383023e-07, "loss": 0.1702, "step": 8662 }, { "epoch": 2.818152244632401, "grad_norm": 2.1047501126743406, "learning_rate": 1.1192762009501201e-07, "loss": 0.1726, "step": 8663 }, { "epoch": 2.8184775536759923, "grad_norm": 2.261102284892582, "learning_rate": 1.1152972696022447e-07, "loss": 0.176, "step": 8664 }, { "epoch": 2.8188028627195836, "grad_norm": 2.330956664130786, "learning_rate": 1.1113253434648641e-07, "loss": 0.1544, "step": 8665 }, { "epoch": 2.819128171763175, "grad_norm": 1.97926206587396, "learning_rate": 1.1073604231071567e-07, "loss": 0.1702, "step": 8666 }, { "epoch": 2.819453480806766, "grad_norm": 2.2837914374866113, "learning_rate": 1.1034025090973066e-07, "loss": 0.1529, "step": 8667 }, { "epoch": 2.819778789850358, "grad_norm": 2.006571577823865, "learning_rate": 1.0994516020024659e-07, "loss": 0.1448, "step": 8668 }, { "epoch": 2.820104098893949, "grad_norm": 2.558910753116456, "learning_rate": 1.0955077023888261e-07, "loss": 0.1965, "step": 8669 }, { "epoch": 2.8204294079375405, "grad_norm": 2.2567431932753523, "learning_rate": 1.0915708108215295e-07, "loss": 0.1952, "step": 8670 }, { "epoch": 2.8207547169811322, "grad_norm": 1.8971562884872433, "learning_rate": 1.0876409278647471e-07, "loss": 0.1087, "step": 8671 }, { "epoch": 2.8210800260247235, "grad_norm": 2.1035086125128455, "learning_rate": 1.0837180540816339e-07, "loss": 0.123, "step": 8672 }, { "epoch": 2.821405335068315, "grad_norm": 2.0727491293934412, "learning_rate": 1.0798021900343348e-07, "loss": 0.1543, "step": 8673 }, { "epoch": 2.8217306441119065, "grad_norm": 1.9379497072269924, "learning_rate": 1.0758933362839896e-07, "loss": 0.1564, "step": 8674 }, { "epoch": 2.822055953155498, "grad_norm": 1.956821709726736, "learning_rate": 1.0719914933907616e-07, "loss": 0.1359, "step": 8675 }, { "epoch": 2.822381262199089, "grad_norm": 2.3037894309179254, "learning_rate": 1.0680966619137756e-07, "loss": 0.2167, "step": 8676 }, { "epoch": 2.8227065712426804, "grad_norm": 2.23732222947245, "learning_rate": 1.0642088424111518e-07, "loss": 0.1669, "step": 8677 }, { "epoch": 2.8230318802862717, "grad_norm": 2.3216497261151448, "learning_rate": 1.0603280354400336e-07, "loss": 0.1436, "step": 8678 }, { "epoch": 2.8233571893298635, "grad_norm": 2.0585439123714226, "learning_rate": 1.056454241556537e-07, "loss": 0.1769, "step": 8679 }, { "epoch": 2.8236824983734548, "grad_norm": 1.9567392140058408, "learning_rate": 1.0525874613157793e-07, "loss": 0.1788, "step": 8680 }, { "epoch": 2.824007807417046, "grad_norm": 2.031660139617234, "learning_rate": 1.0487276952718728e-07, "loss": 0.1472, "step": 8681 }, { "epoch": 2.824333116460638, "grad_norm": 2.108510785241888, "learning_rate": 1.0448749439779249e-07, "loss": 0.1725, "step": 8682 }, { "epoch": 2.824658425504229, "grad_norm": 2.1779491326517877, "learning_rate": 1.0410292079860329e-07, "loss": 0.2245, "step": 8683 }, { "epoch": 2.8249837345478204, "grad_norm": 1.9543158589345047, "learning_rate": 1.037190487847306e-07, "loss": 0.147, "step": 8684 }, { "epoch": 2.825309043591412, "grad_norm": 2.444699865978484, "learning_rate": 1.0333587841118264e-07, "loss": 0.1471, "step": 8685 }, { "epoch": 2.8256343526350034, "grad_norm": 2.426131522664067, "learning_rate": 1.0295340973286716e-07, "loss": 0.1825, "step": 8686 }, { "epoch": 2.8259596616785947, "grad_norm": 2.1683883242347903, "learning_rate": 1.025716428045942e-07, "loss": 0.1553, "step": 8687 }, { "epoch": 2.826284970722186, "grad_norm": 1.9467015940043455, "learning_rate": 1.0219057768106943e-07, "loss": 0.1308, "step": 8688 }, { "epoch": 2.8266102797657773, "grad_norm": 2.141457472791119, "learning_rate": 1.0181021441690086e-07, "loss": 0.1445, "step": 8689 }, { "epoch": 2.826935588809369, "grad_norm": 1.890422902567157, "learning_rate": 1.0143055306659377e-07, "loss": 0.1376, "step": 8690 }, { "epoch": 2.8272608978529603, "grad_norm": 2.011646413820814, "learning_rate": 1.0105159368455464e-07, "loss": 0.1722, "step": 8691 }, { "epoch": 2.8275862068965516, "grad_norm": 1.8605939151164899, "learning_rate": 1.0067333632508835e-07, "loss": 0.1173, "step": 8692 }, { "epoch": 2.8279115159401433, "grad_norm": 2.1258207007789567, "learning_rate": 1.002957810423999e-07, "loss": 0.1255, "step": 8693 }, { "epoch": 2.8282368249837346, "grad_norm": 2.209270057994175, "learning_rate": 9.99189278905921e-08, "loss": 0.1494, "step": 8694 }, { "epoch": 2.828562134027326, "grad_norm": 2.111617091892645, "learning_rate": 9.954277692366954e-08, "loss": 0.1569, "step": 8695 }, { "epoch": 2.828887443070917, "grad_norm": 1.949373162121009, "learning_rate": 9.916732819553409e-08, "loss": 0.1822, "step": 8696 }, { "epoch": 2.8292127521145085, "grad_norm": 2.1471836293148434, "learning_rate": 9.879258175998774e-08, "loss": 0.1736, "step": 8697 }, { "epoch": 2.8295380611581002, "grad_norm": 2.653840438138391, "learning_rate": 9.84185376707325e-08, "loss": 0.1495, "step": 8698 }, { "epoch": 2.8298633702016915, "grad_norm": 1.901798968177735, "learning_rate": 9.804519598136942e-08, "loss": 0.1781, "step": 8699 }, { "epoch": 2.830188679245283, "grad_norm": 2.266170203367923, "learning_rate": 9.767255674539733e-08, "loss": 0.1941, "step": 8700 }, { "epoch": 2.8305139882888746, "grad_norm": 2.543074929531283, "learning_rate": 9.730062001621687e-08, "loss": 0.2016, "step": 8701 }, { "epoch": 2.830839297332466, "grad_norm": 2.2574822519144413, "learning_rate": 9.692938584712541e-08, "loss": 0.199, "step": 8702 }, { "epoch": 2.831164606376057, "grad_norm": 2.402913887762469, "learning_rate": 9.655885429132261e-08, "loss": 0.1573, "step": 8703 }, { "epoch": 2.831489915419649, "grad_norm": 2.110363885888704, "learning_rate": 9.618902540190545e-08, "loss": 0.1841, "step": 8704 }, { "epoch": 2.83181522446324, "grad_norm": 2.054057445229567, "learning_rate": 9.581989923187041e-08, "loss": 0.1173, "step": 8705 }, { "epoch": 2.8321405335068315, "grad_norm": 2.3839627079662566, "learning_rate": 9.545147583411296e-08, "loss": 0.18, "step": 8706 }, { "epoch": 2.8324658425504228, "grad_norm": 2.101298483791772, "learning_rate": 9.508375526142976e-08, "loss": 0.1307, "step": 8707 }, { "epoch": 2.832791151594014, "grad_norm": 1.9914202563666477, "learning_rate": 9.471673756651479e-08, "loss": 0.1754, "step": 8708 }, { "epoch": 2.833116460637606, "grad_norm": 2.267609416508195, "learning_rate": 9.435042280196094e-08, "loss": 0.172, "step": 8709 }, { "epoch": 2.833441769681197, "grad_norm": 2.28706226574497, "learning_rate": 9.398481102026291e-08, "loss": 0.131, "step": 8710 }, { "epoch": 2.8337670787247884, "grad_norm": 1.7677340593450248, "learning_rate": 9.361990227381268e-08, "loss": 0.1499, "step": 8711 }, { "epoch": 2.83409238776838, "grad_norm": 2.4555200439639804, "learning_rate": 9.325569661490064e-08, "loss": 0.2175, "step": 8712 }, { "epoch": 2.8344176968119714, "grad_norm": 1.9278832435909152, "learning_rate": 9.289219409571949e-08, "loss": 0.1648, "step": 8713 }, { "epoch": 2.8347430058555627, "grad_norm": 2.6776613810564696, "learning_rate": 9.252939476835866e-08, "loss": 0.1565, "step": 8714 }, { "epoch": 2.8350683148991545, "grad_norm": 1.8760577564412881, "learning_rate": 9.216729868480712e-08, "loss": 0.1146, "step": 8715 }, { "epoch": 2.8353936239427457, "grad_norm": 2.5724609775193166, "learning_rate": 9.180590589695448e-08, "loss": 0.1837, "step": 8716 }, { "epoch": 2.835718932986337, "grad_norm": 2.062473223122275, "learning_rate": 9.144521645658766e-08, "loss": 0.1347, "step": 8717 }, { "epoch": 2.8360442420299283, "grad_norm": 2.045236607033943, "learning_rate": 9.108523041539364e-08, "loss": 0.1268, "step": 8718 }, { "epoch": 2.8363695510735196, "grad_norm": 1.9710244919839566, "learning_rate": 9.072594782495947e-08, "loss": 0.165, "step": 8719 }, { "epoch": 2.8366948601171114, "grad_norm": 1.9984403711779983, "learning_rate": 9.03673687367701e-08, "loss": 0.1332, "step": 8720 }, { "epoch": 2.8370201691607027, "grad_norm": 2.1673233228714386, "learning_rate": 9.000949320220998e-08, "loss": 0.1582, "step": 8721 }, { "epoch": 2.837345478204294, "grad_norm": 2.3517374257768595, "learning_rate": 8.965232127256362e-08, "loss": 0.1896, "step": 8722 }, { "epoch": 2.8376707872478857, "grad_norm": 1.9881084343890756, "learning_rate": 8.929585299901289e-08, "loss": 0.1518, "step": 8723 }, { "epoch": 2.837996096291477, "grad_norm": 2.4746887702330223, "learning_rate": 8.89400884326408e-08, "loss": 0.2047, "step": 8724 }, { "epoch": 2.8383214053350683, "grad_norm": 2.084908912690008, "learning_rate": 8.858502762442822e-08, "loss": 0.1308, "step": 8725 }, { "epoch": 2.8386467143786596, "grad_norm": 1.991674473881176, "learning_rate": 8.823067062525615e-08, "loss": 0.1273, "step": 8726 }, { "epoch": 2.8389720234222513, "grad_norm": 1.8909307100500847, "learning_rate": 8.787701748590338e-08, "loss": 0.1431, "step": 8727 }, { "epoch": 2.8392973324658426, "grad_norm": 2.2858906635521543, "learning_rate": 8.75240682570494e-08, "loss": 0.1939, "step": 8728 }, { "epoch": 2.839622641509434, "grad_norm": 2.3748038172215757, "learning_rate": 8.717182298927152e-08, "loss": 0.1784, "step": 8729 }, { "epoch": 2.839947950553025, "grad_norm": 1.9652929408121298, "learning_rate": 8.682028173304713e-08, "loss": 0.1565, "step": 8730 }, { "epoch": 2.840273259596617, "grad_norm": 2.0909338391178296, "learning_rate": 8.646944453875205e-08, "loss": 0.1604, "step": 8731 }, { "epoch": 2.840598568640208, "grad_norm": 1.9782200015248708, "learning_rate": 8.611931145666108e-08, "loss": 0.1648, "step": 8732 }, { "epoch": 2.8409238776837995, "grad_norm": 2.481228799323598, "learning_rate": 8.576988253694962e-08, "loss": 0.1483, "step": 8733 }, { "epoch": 2.8412491867273912, "grad_norm": 1.8159005696428794, "learning_rate": 8.54211578296904e-08, "loss": 0.1214, "step": 8734 }, { "epoch": 2.8415744957709825, "grad_norm": 2.3623606413443716, "learning_rate": 8.507313738485567e-08, "loss": 0.1944, "step": 8735 }, { "epoch": 2.841899804814574, "grad_norm": 2.351593210508499, "learning_rate": 8.472582125231776e-08, "loss": 0.1668, "step": 8736 }, { "epoch": 2.842225113858165, "grad_norm": 2.161259079268066, "learning_rate": 8.437920948184631e-08, "loss": 0.1944, "step": 8737 }, { "epoch": 2.8425504229017564, "grad_norm": 1.954361414850813, "learning_rate": 8.403330212311156e-08, "loss": 0.1622, "step": 8738 }, { "epoch": 2.842875731945348, "grad_norm": 1.891205340831177, "learning_rate": 8.368809922568222e-08, "loss": 0.1365, "step": 8739 }, { "epoch": 2.8432010409889394, "grad_norm": 2.4835143138383096, "learning_rate": 8.334360083902649e-08, "loss": 0.1826, "step": 8740 }, { "epoch": 2.8435263500325307, "grad_norm": 2.2023146185753193, "learning_rate": 8.299980701251043e-08, "loss": 0.1814, "step": 8741 }, { "epoch": 2.8438516590761225, "grad_norm": 2.3215049715767844, "learning_rate": 8.265671779540075e-08, "loss": 0.1422, "step": 8742 }, { "epoch": 2.8441769681197138, "grad_norm": 2.189734153580393, "learning_rate": 8.2314333236862e-08, "loss": 0.1562, "step": 8743 }, { "epoch": 2.844502277163305, "grad_norm": 2.262549442138824, "learning_rate": 8.197265338595772e-08, "loss": 0.1863, "step": 8744 }, { "epoch": 2.844827586206897, "grad_norm": 2.144947345098191, "learning_rate": 8.163167829165098e-08, "loss": 0.1849, "step": 8745 }, { "epoch": 2.845152895250488, "grad_norm": 2.3198813636644884, "learning_rate": 8.129140800280434e-08, "loss": 0.1745, "step": 8746 }, { "epoch": 2.8454782042940794, "grad_norm": 2.027065721446264, "learning_rate": 8.095184256817823e-08, "loss": 0.1679, "step": 8747 }, { "epoch": 2.8458035133376707, "grad_norm": 1.932764491666238, "learning_rate": 8.06129820364332e-08, "loss": 0.1188, "step": 8748 }, { "epoch": 2.846128822381262, "grad_norm": 1.8391064871810179, "learning_rate": 8.027482645612706e-08, "loss": 0.1385, "step": 8749 }, { "epoch": 2.8464541314248537, "grad_norm": 2.5234087761620216, "learning_rate": 7.993737587571825e-08, "loss": 0.1783, "step": 8750 }, { "epoch": 2.846779440468445, "grad_norm": 2.213132432663554, "learning_rate": 7.96006303435648e-08, "loss": 0.1666, "step": 8751 }, { "epoch": 2.8471047495120363, "grad_norm": 2.11952266269772, "learning_rate": 7.926458990792086e-08, "loss": 0.1682, "step": 8752 }, { "epoch": 2.847430058555628, "grad_norm": 1.6943932564029807, "learning_rate": 7.892925461694234e-08, "loss": 0.1345, "step": 8753 }, { "epoch": 2.8477553675992193, "grad_norm": 2.0825860450539277, "learning_rate": 7.85946245186825e-08, "loss": 0.1576, "step": 8754 }, { "epoch": 2.8480806766428106, "grad_norm": 2.0273397323643505, "learning_rate": 7.826069966109407e-08, "loss": 0.1599, "step": 8755 }, { "epoch": 2.8484059856864024, "grad_norm": 2.323236391092954, "learning_rate": 7.792748009202932e-08, "loss": 0.1423, "step": 8756 }, { "epoch": 2.8487312947299936, "grad_norm": 1.9584118538226651, "learning_rate": 7.759496585923786e-08, "loss": 0.1417, "step": 8757 }, { "epoch": 2.849056603773585, "grad_norm": 1.9978226503146745, "learning_rate": 7.72631570103699e-08, "loss": 0.1583, "step": 8758 }, { "epoch": 2.8493819128171762, "grad_norm": 2.238194597867557, "learning_rate": 7.693205359297407e-08, "loss": 0.1711, "step": 8759 }, { "epoch": 2.8497072218607675, "grad_norm": 2.334590681775535, "learning_rate": 7.660165565449685e-08, "loss": 0.2133, "step": 8760 }, { "epoch": 2.8500325309043593, "grad_norm": 1.8887338960202973, "learning_rate": 7.627196324228481e-08, "loss": 0.1221, "step": 8761 }, { "epoch": 2.8503578399479506, "grad_norm": 2.0253449592444, "learning_rate": 7.59429764035835e-08, "loss": 0.1495, "step": 8762 }, { "epoch": 2.850683148991542, "grad_norm": 2.0596752070464914, "learning_rate": 7.561469518553687e-08, "loss": 0.1635, "step": 8763 }, { "epoch": 2.8510084580351336, "grad_norm": 1.7755272144836611, "learning_rate": 7.52871196351873e-08, "loss": 0.1656, "step": 8764 }, { "epoch": 2.851333767078725, "grad_norm": 2.477444038404139, "learning_rate": 7.496024979947669e-08, "loss": 0.1953, "step": 8765 }, { "epoch": 2.851659076122316, "grad_norm": 2.0226021033626935, "learning_rate": 7.4634085725247e-08, "loss": 0.1752, "step": 8766 }, { "epoch": 2.8519843851659075, "grad_norm": 2.216490587711545, "learning_rate": 7.430862745923584e-08, "loss": 0.224, "step": 8767 }, { "epoch": 2.8523096942094988, "grad_norm": 2.298764962391716, "learning_rate": 7.398387504808257e-08, "loss": 0.1933, "step": 8768 }, { "epoch": 2.8526350032530905, "grad_norm": 2.1032148321648605, "learning_rate": 7.365982853832443e-08, "loss": 0.1554, "step": 8769 }, { "epoch": 2.852960312296682, "grad_norm": 2.277117823484346, "learning_rate": 7.333648797639703e-08, "loss": 0.162, "step": 8770 }, { "epoch": 2.853285621340273, "grad_norm": 2.8269950674733715, "learning_rate": 7.301385340863665e-08, "loss": 0.1794, "step": 8771 }, { "epoch": 2.853610930383865, "grad_norm": 2.4901872690336933, "learning_rate": 7.269192488127518e-08, "loss": 0.1553, "step": 8772 }, { "epoch": 2.853936239427456, "grad_norm": 1.9439589233504737, "learning_rate": 7.237070244044575e-08, "loss": 0.1397, "step": 8773 }, { "epoch": 2.8542615484710474, "grad_norm": 2.0901032793388463, "learning_rate": 7.205018613218096e-08, "loss": 0.1492, "step": 8774 }, { "epoch": 2.854586857514639, "grad_norm": 2.434398399930339, "learning_rate": 7.173037600241017e-08, "loss": 0.186, "step": 8775 }, { "epoch": 2.8549121665582304, "grad_norm": 2.0783365456821215, "learning_rate": 7.141127209696175e-08, "loss": 0.1075, "step": 8776 }, { "epoch": 2.8552374756018217, "grad_norm": 1.7770848150247873, "learning_rate": 7.10928744615641e-08, "loss": 0.1112, "step": 8777 }, { "epoch": 2.855562784645413, "grad_norm": 2.3699462970786684, "learning_rate": 7.077518314184406e-08, "loss": 0.2496, "step": 8778 }, { "epoch": 2.8558880936890043, "grad_norm": 2.0800793360617513, "learning_rate": 7.045819818332633e-08, "loss": 0.1429, "step": 8779 }, { "epoch": 2.856213402732596, "grad_norm": 2.1867318580434043, "learning_rate": 7.014191963143568e-08, "loss": 0.1771, "step": 8780 }, { "epoch": 2.8565387117761873, "grad_norm": 2.0401311506381465, "learning_rate": 6.982634753149476e-08, "loss": 0.1413, "step": 8781 }, { "epoch": 2.8568640208197786, "grad_norm": 2.0347676767632716, "learning_rate": 6.95114819287257e-08, "loss": 0.1728, "step": 8782 }, { "epoch": 2.8571893298633704, "grad_norm": 1.9965563197276148, "learning_rate": 6.919732286824854e-08, "loss": 0.1385, "step": 8783 }, { "epoch": 2.8575146389069617, "grad_norm": 1.9459330087807638, "learning_rate": 6.888387039508171e-08, "loss": 0.1536, "step": 8784 }, { "epoch": 2.857839947950553, "grad_norm": 2.537474451368129, "learning_rate": 6.857112455414483e-08, "loss": 0.1687, "step": 8785 }, { "epoch": 2.8581652569941447, "grad_norm": 2.1967191674467434, "learning_rate": 6.825908539025372e-08, "loss": 0.1327, "step": 8786 }, { "epoch": 2.858490566037736, "grad_norm": 1.9082639581162117, "learning_rate": 6.794775294812317e-08, "loss": 0.1684, "step": 8787 }, { "epoch": 2.8588158750813273, "grad_norm": 2.2277687938862023, "learning_rate": 6.76371272723686e-08, "loss": 0.1921, "step": 8788 }, { "epoch": 2.8591411841249186, "grad_norm": 2.2049056291456357, "learning_rate": 6.732720840750163e-08, "loss": 0.1482, "step": 8789 }, { "epoch": 2.85946649316851, "grad_norm": 2.2551062279218335, "learning_rate": 6.701799639793504e-08, "loss": 0.1673, "step": 8790 }, { "epoch": 2.8597918022121016, "grad_norm": 2.2191616772057325, "learning_rate": 6.670949128797843e-08, "loss": 0.1714, "step": 8791 }, { "epoch": 2.860117111255693, "grad_norm": 1.8572780527302397, "learning_rate": 6.640169312184086e-08, "loss": 0.1445, "step": 8792 }, { "epoch": 2.860442420299284, "grad_norm": 2.1459158095012962, "learning_rate": 6.609460194362927e-08, "loss": 0.1516, "step": 8793 }, { "epoch": 2.860767729342876, "grad_norm": 1.7770016776919078, "learning_rate": 6.578821779735178e-08, "loss": 0.1222, "step": 8794 }, { "epoch": 2.8610930383864672, "grad_norm": 1.929513314707058, "learning_rate": 6.548254072691273e-08, "loss": 0.1432, "step": 8795 }, { "epoch": 2.8614183474300585, "grad_norm": 2.40797111756337, "learning_rate": 6.517757077611487e-08, "loss": 0.196, "step": 8796 }, { "epoch": 2.86174365647365, "grad_norm": 2.177468243677331, "learning_rate": 6.487330798866099e-08, "loss": 0.1449, "step": 8797 }, { "epoch": 2.862068965517241, "grad_norm": 2.1094443463489094, "learning_rate": 6.456975240815344e-08, "loss": 0.1638, "step": 8798 }, { "epoch": 2.862394274560833, "grad_norm": 1.8181779995588305, "learning_rate": 6.426690407809022e-08, "loss": 0.1269, "step": 8799 }, { "epoch": 2.862719583604424, "grad_norm": 2.380173557296112, "learning_rate": 6.396476304187105e-08, "loss": 0.1957, "step": 8800 }, { "epoch": 2.8630448926480154, "grad_norm": 1.9486472126356058, "learning_rate": 6.366332934279185e-08, "loss": 0.126, "step": 8801 }, { "epoch": 2.863370201691607, "grad_norm": 2.3445077395347584, "learning_rate": 6.336260302404861e-08, "loss": 0.1833, "step": 8802 }, { "epoch": 2.8636955107351985, "grad_norm": 2.2656987617996602, "learning_rate": 6.306258412873634e-08, "loss": 0.1897, "step": 8803 }, { "epoch": 2.8640208197787898, "grad_norm": 1.9606050862223605, "learning_rate": 6.276327269984672e-08, "loss": 0.1108, "step": 8804 }, { "epoch": 2.8643461288223815, "grad_norm": 2.2369184894428584, "learning_rate": 6.246466878027214e-08, "loss": 0.1612, "step": 8805 }, { "epoch": 2.864671437865973, "grad_norm": 2.0482706256860483, "learning_rate": 6.216677241280277e-08, "loss": 0.1543, "step": 8806 }, { "epoch": 2.864996746909564, "grad_norm": 1.8480290299975897, "learning_rate": 6.186958364012729e-08, "loss": 0.118, "step": 8807 }, { "epoch": 2.8653220559531554, "grad_norm": 1.8735536248970683, "learning_rate": 6.157310250483273e-08, "loss": 0.1642, "step": 8808 }, { "epoch": 2.8656473649967467, "grad_norm": 1.999085457051932, "learning_rate": 6.127732904940454e-08, "loss": 0.1747, "step": 8809 }, { "epoch": 2.8659726740403384, "grad_norm": 2.4333555918848386, "learning_rate": 6.098226331622825e-08, "loss": 0.1978, "step": 8810 }, { "epoch": 2.8662979830839297, "grad_norm": 1.8795608830593473, "learning_rate": 6.068790534758728e-08, "loss": 0.1587, "step": 8811 }, { "epoch": 2.866623292127521, "grad_norm": 1.8036828149651967, "learning_rate": 6.039425518566233e-08, "loss": 0.135, "step": 8812 }, { "epoch": 2.8669486011711127, "grad_norm": 2.253221278595955, "learning_rate": 6.010131287253363e-08, "loss": 0.1487, "step": 8813 }, { "epoch": 2.867273910214704, "grad_norm": 1.7878444718943247, "learning_rate": 5.980907845018036e-08, "loss": 0.152, "step": 8814 }, { "epoch": 2.8675992192582953, "grad_norm": 2.4507238838105443, "learning_rate": 5.951755196048015e-08, "loss": 0.2204, "step": 8815 }, { "epoch": 2.867924528301887, "grad_norm": 1.9040568729303236, "learning_rate": 5.922673344520846e-08, "loss": 0.1364, "step": 8816 }, { "epoch": 2.8682498373454783, "grad_norm": 2.234688441426556, "learning_rate": 5.89366229460403e-08, "loss": 0.1351, "step": 8817 }, { "epoch": 2.8685751463890696, "grad_norm": 2.244156086509107, "learning_rate": 5.864722050454852e-08, "loss": 0.1543, "step": 8818 }, { "epoch": 2.868900455432661, "grad_norm": 2.136635165258744, "learning_rate": 5.835852616220383e-08, "loss": 0.1643, "step": 8819 }, { "epoch": 2.869225764476252, "grad_norm": 2.091582036777126, "learning_rate": 5.8070539960378145e-08, "loss": 0.1558, "step": 8820 }, { "epoch": 2.869551073519844, "grad_norm": 2.098754299099209, "learning_rate": 5.778326194033845e-08, "loss": 0.1335, "step": 8821 }, { "epoch": 2.8698763825634352, "grad_norm": 1.9402150422254074, "learning_rate": 5.749669214325237e-08, "loss": 0.1536, "step": 8822 }, { "epoch": 2.8702016916070265, "grad_norm": 2.059582004873588, "learning_rate": 5.721083061018595e-08, "loss": 0.1198, "step": 8823 }, { "epoch": 2.8705270006506183, "grad_norm": 2.333307420736731, "learning_rate": 5.692567738210308e-08, "loss": 0.1554, "step": 8824 }, { "epoch": 2.8708523096942096, "grad_norm": 2.0217730116291928, "learning_rate": 5.664123249986553e-08, "loss": 0.1678, "step": 8825 }, { "epoch": 2.871177618737801, "grad_norm": 2.291928629699784, "learning_rate": 5.635749600423568e-08, "loss": 0.1688, "step": 8826 }, { "epoch": 2.8715029277813926, "grad_norm": 2.242347220328174, "learning_rate": 5.607446793587323e-08, "loss": 0.1394, "step": 8827 }, { "epoch": 2.871828236824984, "grad_norm": 2.067177269100132, "learning_rate": 5.57921483353352e-08, "loss": 0.1691, "step": 8828 }, { "epoch": 2.872153545868575, "grad_norm": 2.3806024813967874, "learning_rate": 5.5510537243078644e-08, "loss": 0.2153, "step": 8829 }, { "epoch": 2.8724788549121665, "grad_norm": 2.4049874840504692, "learning_rate": 5.522963469945908e-08, "loss": 0.1982, "step": 8830 }, { "epoch": 2.8728041639557578, "grad_norm": 1.8946694862962115, "learning_rate": 5.4949440744729856e-08, "loss": 0.1274, "step": 8831 }, { "epoch": 2.8731294729993495, "grad_norm": 2.3977902249900875, "learning_rate": 5.4669955419042745e-08, "loss": 0.1705, "step": 8832 }, { "epoch": 2.873454782042941, "grad_norm": 2.1248942264793262, "learning_rate": 5.4391178762447924e-08, "loss": 0.1356, "step": 8833 }, { "epoch": 2.873780091086532, "grad_norm": 2.6924048659294924, "learning_rate": 5.4113110814894563e-08, "loss": 0.1848, "step": 8834 }, { "epoch": 2.874105400130124, "grad_norm": 2.255620081128364, "learning_rate": 5.3835751616230223e-08, "loss": 0.1215, "step": 8835 }, { "epoch": 2.874430709173715, "grad_norm": 2.177467414748923, "learning_rate": 5.3559101206200337e-08, "loss": 0.1471, "step": 8836 }, { "epoch": 2.8747560182173064, "grad_norm": 2.3704129391550395, "learning_rate": 5.3283159624448745e-08, "loss": 0.165, "step": 8837 }, { "epoch": 2.8750813272608977, "grad_norm": 2.1227303022076205, "learning_rate": 5.300792691051937e-08, "loss": 0.1495, "step": 8838 }, { "epoch": 2.875406636304489, "grad_norm": 2.150394277223737, "learning_rate": 5.273340310385178e-08, "loss": 0.157, "step": 8839 }, { "epoch": 2.8757319453480807, "grad_norm": 2.439905254466263, "learning_rate": 5.245958824378616e-08, "loss": 0.1791, "step": 8840 }, { "epoch": 2.876057254391672, "grad_norm": 2.5052437591168544, "learning_rate": 5.218648236956059e-08, "loss": 0.2168, "step": 8841 }, { "epoch": 2.8763825634352633, "grad_norm": 2.2879100044619705, "learning_rate": 5.191408552031041e-08, "loss": 0.1876, "step": 8842 }, { "epoch": 2.876707872478855, "grad_norm": 2.105791051883426, "learning_rate": 5.164239773507107e-08, "loss": 0.1487, "step": 8843 }, { "epoch": 2.8770331815224464, "grad_norm": 2.1014352906883245, "learning_rate": 5.137141905277476e-08, "loss": 0.1296, "step": 8844 }, { "epoch": 2.8773584905660377, "grad_norm": 2.123030826021411, "learning_rate": 5.110114951225431e-08, "loss": 0.1455, "step": 8845 }, { "epoch": 2.8776837996096294, "grad_norm": 2.086556491124475, "learning_rate": 5.0831589152238184e-08, "loss": 0.1224, "step": 8846 }, { "epoch": 2.8780091086532207, "grad_norm": 2.4931583322527686, "learning_rate": 5.056273801135603e-08, "loss": 0.1877, "step": 8847 }, { "epoch": 2.878334417696812, "grad_norm": 1.799116243142679, "learning_rate": 5.0294596128132035e-08, "loss": 0.1324, "step": 8848 }, { "epoch": 2.8786597267404033, "grad_norm": 2.02093011751522, "learning_rate": 5.002716354099324e-08, "loss": 0.1454, "step": 8849 }, { "epoch": 2.8789850357839946, "grad_norm": 2.3912815796706473, "learning_rate": 4.9760440288262324e-08, "loss": 0.2335, "step": 8850 }, { "epoch": 2.8793103448275863, "grad_norm": 2.0440892081279984, "learning_rate": 4.9494426408160934e-08, "loss": 0.1334, "step": 8851 }, { "epoch": 2.8796356538711776, "grad_norm": 2.36023046360822, "learning_rate": 4.9229121938808576e-08, "loss": 0.1914, "step": 8852 }, { "epoch": 2.879960962914769, "grad_norm": 2.1303742072823018, "learning_rate": 4.8964526918223734e-08, "loss": 0.1666, "step": 8853 }, { "epoch": 2.8802862719583606, "grad_norm": 2.150536513352751, "learning_rate": 4.870064138432329e-08, "loss": 0.1308, "step": 8854 }, { "epoch": 2.880611581001952, "grad_norm": 2.1522642078422196, "learning_rate": 4.8437465374922556e-08, "loss": 0.1462, "step": 8855 }, { "epoch": 2.880936890045543, "grad_norm": 2.227860546538621, "learning_rate": 4.817499892773359e-08, "loss": 0.1303, "step": 8856 }, { "epoch": 2.881262199089135, "grad_norm": 2.258375125661458, "learning_rate": 4.791324208036907e-08, "loss": 0.1547, "step": 8857 }, { "epoch": 2.8815875081327262, "grad_norm": 2.0657410492318777, "learning_rate": 4.765219487033901e-08, "loss": 0.1451, "step": 8858 }, { "epoch": 2.8819128171763175, "grad_norm": 1.706358763276959, "learning_rate": 4.739185733505125e-08, "loss": 0.1074, "step": 8859 }, { "epoch": 2.882238126219909, "grad_norm": 1.8409332678229897, "learning_rate": 4.7132229511812067e-08, "loss": 0.1341, "step": 8860 }, { "epoch": 2.8825634352635, "grad_norm": 2.3447293826036124, "learning_rate": 4.6873311437826694e-08, "loss": 0.1907, "step": 8861 }, { "epoch": 2.882888744307092, "grad_norm": 1.828698743038496, "learning_rate": 4.6615103150198794e-08, "loss": 0.1753, "step": 8862 }, { "epoch": 2.883214053350683, "grad_norm": 2.2570200098424995, "learning_rate": 4.6357604685928204e-08, "loss": 0.2175, "step": 8863 }, { "epoch": 2.8835393623942744, "grad_norm": 1.9791900742410131, "learning_rate": 4.6100816081915965e-08, "loss": 0.1332, "step": 8864 }, { "epoch": 2.883864671437866, "grad_norm": 2.046198781909832, "learning_rate": 4.5844737374960426e-08, "loss": 0.1684, "step": 8865 }, { "epoch": 2.8841899804814575, "grad_norm": 2.1792931406247624, "learning_rate": 4.558936860175611e-08, "loss": 0.1657, "step": 8866 }, { "epoch": 2.8845152895250488, "grad_norm": 1.8337186303009383, "learning_rate": 4.533470979889931e-08, "loss": 0.126, "step": 8867 }, { "epoch": 2.88484059856864, "grad_norm": 2.0056998662631518, "learning_rate": 4.508076100288139e-08, "loss": 0.167, "step": 8868 }, { "epoch": 2.8851659076122314, "grad_norm": 2.006576592955633, "learning_rate": 4.482752225009379e-08, "loss": 0.1677, "step": 8869 }, { "epoch": 2.885491216655823, "grad_norm": 2.1591553918572446, "learning_rate": 4.457499357682693e-08, "loss": 0.1339, "step": 8870 }, { "epoch": 2.8858165256994144, "grad_norm": 2.28897344146665, "learning_rate": 4.432317501926686e-08, "loss": 0.153, "step": 8871 }, { "epoch": 2.8861418347430057, "grad_norm": 2.4363225192451115, "learning_rate": 4.407206661349972e-08, "loss": 0.1685, "step": 8872 }, { "epoch": 2.8864671437865974, "grad_norm": 2.130074417044703, "learning_rate": 4.382166839551005e-08, "loss": 0.1935, "step": 8873 }, { "epoch": 2.8867924528301887, "grad_norm": 2.133078393306113, "learning_rate": 4.3571980401179713e-08, "loss": 0.1299, "step": 8874 }, { "epoch": 2.88711776187378, "grad_norm": 1.825565010339022, "learning_rate": 4.332300266628897e-08, "loss": 0.1139, "step": 8875 }, { "epoch": 2.8874430709173717, "grad_norm": 1.6126803170918154, "learning_rate": 4.307473522651706e-08, "loss": 0.1159, "step": 8876 }, { "epoch": 2.887768379960963, "grad_norm": 2.450684269959642, "learning_rate": 4.2827178117439974e-08, "loss": 0.1809, "step": 8877 }, { "epoch": 2.8880936890045543, "grad_norm": 2.1007186782890623, "learning_rate": 4.258033137453377e-08, "loss": 0.1513, "step": 8878 }, { "epoch": 2.8884189980481456, "grad_norm": 2.1780113231323766, "learning_rate": 4.233419503317182e-08, "loss": 0.1401, "step": 8879 }, { "epoch": 2.888744307091737, "grad_norm": 2.146660727188287, "learning_rate": 4.2088769128623696e-08, "loss": 0.1614, "step": 8880 }, { "epoch": 2.8890696161353286, "grad_norm": 1.8638776302660927, "learning_rate": 4.184405369606181e-08, "loss": 0.1212, "step": 8881 }, { "epoch": 2.88939492517892, "grad_norm": 2.0017716735748508, "learning_rate": 4.160004877055257e-08, "loss": 0.191, "step": 8882 }, { "epoch": 2.8897202342225112, "grad_norm": 2.284618776736, "learning_rate": 4.135675438706188e-08, "loss": 0.1654, "step": 8883 }, { "epoch": 2.890045543266103, "grad_norm": 2.321685247768339, "learning_rate": 4.111417058045464e-08, "loss": 0.2227, "step": 8884 }, { "epoch": 2.8903708523096943, "grad_norm": 2.6909488559460404, "learning_rate": 4.0872297385493035e-08, "loss": 0.1858, "step": 8885 }, { "epoch": 2.8906961613532856, "grad_norm": 2.1421406831783827, "learning_rate": 4.063113483683712e-08, "loss": 0.1429, "step": 8886 }, { "epoch": 2.8910214703968773, "grad_norm": 1.8869808947659736, "learning_rate": 4.039068296904702e-08, "loss": 0.1202, "step": 8887 }, { "epoch": 2.8913467794404686, "grad_norm": 2.0147046770487846, "learning_rate": 4.015094181657797e-08, "loss": 0.1806, "step": 8888 }, { "epoch": 2.89167208848406, "grad_norm": 2.1119690301007674, "learning_rate": 3.991191141378581e-08, "loss": 0.1654, "step": 8889 }, { "epoch": 2.891997397527651, "grad_norm": 2.225091033413124, "learning_rate": 3.967359179492425e-08, "loss": 0.1512, "step": 8890 }, { "epoch": 2.8923227065712425, "grad_norm": 2.5321132198444958, "learning_rate": 3.9435982994144304e-08, "loss": 0.1574, "step": 8891 }, { "epoch": 2.892648015614834, "grad_norm": 2.1170715910232936, "learning_rate": 3.919908504549486e-08, "loss": 0.1586, "step": 8892 }, { "epoch": 2.8929733246584255, "grad_norm": 2.3040372084855374, "learning_rate": 3.8962897982924296e-08, "loss": 0.1419, "step": 8893 }, { "epoch": 2.893298633702017, "grad_norm": 2.306145580381871, "learning_rate": 3.872742184027778e-08, "loss": 0.1947, "step": 8894 }, { "epoch": 2.8936239427456085, "grad_norm": 1.9247416928807013, "learning_rate": 3.8492656651299974e-08, "loss": 0.1199, "step": 8895 }, { "epoch": 2.8939492517892, "grad_norm": 1.8967032338405627, "learning_rate": 3.825860244963231e-08, "loss": 0.1551, "step": 8896 }, { "epoch": 2.894274560832791, "grad_norm": 1.8768935170933303, "learning_rate": 3.8025259268815175e-08, "loss": 0.1084, "step": 8897 }, { "epoch": 2.8945998698763824, "grad_norm": 2.245477343176595, "learning_rate": 3.7792627142286265e-08, "loss": 0.1844, "step": 8898 }, { "epoch": 2.894925178919974, "grad_norm": 2.277132072531488, "learning_rate": 3.7560706103382807e-08, "loss": 0.226, "step": 8899 }, { "epoch": 2.8952504879635654, "grad_norm": 1.9570420697228879, "learning_rate": 3.7329496185338765e-08, "loss": 0.1796, "step": 8900 }, { "epoch": 2.8955757970071567, "grad_norm": 2.1640956808034435, "learning_rate": 3.709899742128653e-08, "loss": 0.1847, "step": 8901 }, { "epoch": 2.895901106050748, "grad_norm": 2.016429872911656, "learning_rate": 3.6869209844256904e-08, "loss": 0.1598, "step": 8902 }, { "epoch": 2.8962264150943398, "grad_norm": 1.944161116425936, "learning_rate": 3.664013348717854e-08, "loss": 0.1444, "step": 8903 }, { "epoch": 2.896551724137931, "grad_norm": 2.331181398215903, "learning_rate": 3.6411768382877965e-08, "loss": 0.1797, "step": 8904 }, { "epoch": 2.8968770331815223, "grad_norm": 2.4060976064869948, "learning_rate": 3.618411456408066e-08, "loss": 0.1848, "step": 8905 }, { "epoch": 2.897202342225114, "grad_norm": 2.418646424976829, "learning_rate": 3.5957172063409427e-08, "loss": 0.2214, "step": 8906 }, { "epoch": 2.8975276512687054, "grad_norm": 2.1933195654114974, "learning_rate": 3.5730940913384914e-08, "loss": 0.1735, "step": 8907 }, { "epoch": 2.8978529603122967, "grad_norm": 2.3973188745560785, "learning_rate": 3.5505421146426186e-08, "loss": 0.1973, "step": 8908 }, { "epoch": 2.898178269355888, "grad_norm": 2.078792493562616, "learning_rate": 3.5280612794850734e-08, "loss": 0.1259, "step": 8909 }, { "epoch": 2.8985035783994793, "grad_norm": 1.951201548563015, "learning_rate": 3.5056515890873885e-08, "loss": 0.1421, "step": 8910 }, { "epoch": 2.898828887443071, "grad_norm": 2.000652971361705, "learning_rate": 3.48331304666083e-08, "loss": 0.1431, "step": 8911 }, { "epoch": 2.8991541964866623, "grad_norm": 2.310341430574312, "learning_rate": 3.461045655406559e-08, "loss": 0.2379, "step": 8912 }, { "epoch": 2.8994795055302536, "grad_norm": 2.448648625871919, "learning_rate": 3.438849418515522e-08, "loss": 0.1544, "step": 8913 }, { "epoch": 2.8998048145738453, "grad_norm": 1.984038627040291, "learning_rate": 3.416724339168398e-08, "loss": 0.1588, "step": 8914 }, { "epoch": 2.9001301236174366, "grad_norm": 2.32988426568367, "learning_rate": 3.394670420535762e-08, "loss": 0.17, "step": 8915 }, { "epoch": 2.900455432661028, "grad_norm": 2.5070131493820007, "learning_rate": 3.372687665777974e-08, "loss": 0.1582, "step": 8916 }, { "epoch": 2.9007807417046196, "grad_norm": 2.1167656753705897, "learning_rate": 3.350776078045182e-08, "loss": 0.1626, "step": 8917 }, { "epoch": 2.901106050748211, "grad_norm": 2.420849783590967, "learning_rate": 3.328935660477261e-08, "loss": 0.1491, "step": 8918 }, { "epoch": 2.9014313597918022, "grad_norm": 2.313960049947075, "learning_rate": 3.307166416204044e-08, "loss": 0.2298, "step": 8919 }, { "epoch": 2.9017566688353935, "grad_norm": 1.996356699853501, "learning_rate": 3.2854683483450335e-08, "loss": 0.1539, "step": 8920 }, { "epoch": 2.902081977878985, "grad_norm": 2.515814793618078, "learning_rate": 3.2638414600095755e-08, "loss": 0.2251, "step": 8921 }, { "epoch": 2.9024072869225765, "grad_norm": 2.333016125748625, "learning_rate": 3.242285754296859e-08, "loss": 0.1462, "step": 8922 }, { "epoch": 2.902732595966168, "grad_norm": 2.223973319490312, "learning_rate": 3.220801234295745e-08, "loss": 0.1735, "step": 8923 }, { "epoch": 2.903057905009759, "grad_norm": 2.310759189815836, "learning_rate": 3.1993879030850495e-08, "loss": 0.2009, "step": 8924 }, { "epoch": 2.903383214053351, "grad_norm": 2.2464721387159288, "learning_rate": 3.1780457637333173e-08, "loss": 0.1499, "step": 8925 }, { "epoch": 2.903708523096942, "grad_norm": 2.1663457540312177, "learning_rate": 3.1567748192988246e-08, "loss": 0.1737, "step": 8926 }, { "epoch": 2.9040338321405335, "grad_norm": 2.39211546860808, "learning_rate": 3.135575072829799e-08, "loss": 0.1703, "step": 8927 }, { "epoch": 2.904359141184125, "grad_norm": 1.9296653816798524, "learning_rate": 3.1144465273641455e-08, "loss": 0.1571, "step": 8928 }, { "epoch": 2.9046844502277165, "grad_norm": 1.888202802428756, "learning_rate": 3.0933891859295515e-08, "loss": 0.1341, "step": 8929 }, { "epoch": 2.905009759271308, "grad_norm": 1.9575416989482084, "learning_rate": 3.0724030515436596e-08, "loss": 0.1363, "step": 8930 }, { "epoch": 2.905335068314899, "grad_norm": 1.7648396324680895, "learning_rate": 3.0514881272136754e-08, "loss": 0.146, "step": 8931 }, { "epoch": 2.9056603773584904, "grad_norm": 2.0951769498816333, "learning_rate": 3.0306444159367564e-08, "loss": 0.1259, "step": 8932 }, { "epoch": 2.905985686402082, "grad_norm": 2.0644979328546564, "learning_rate": 3.009871920699847e-08, "loss": 0.2031, "step": 8933 }, { "epoch": 2.9063109954456734, "grad_norm": 2.1891611021350035, "learning_rate": 2.989170644479677e-08, "loss": 0.1417, "step": 8934 }, { "epoch": 2.9066363044892647, "grad_norm": 2.2920804347412322, "learning_rate": 2.9685405902427076e-08, "loss": 0.1912, "step": 8935 }, { "epoch": 2.9069616135328564, "grad_norm": 2.1885984809224794, "learning_rate": 2.9479817609452955e-08, "loss": 0.1542, "step": 8936 }, { "epoch": 2.9072869225764477, "grad_norm": 2.0310253628774215, "learning_rate": 2.9274941595334728e-08, "loss": 0.1674, "step": 8937 }, { "epoch": 2.907612231620039, "grad_norm": 2.1238775126538694, "learning_rate": 2.9070777889431133e-08, "loss": 0.1381, "step": 8938 }, { "epoch": 2.9079375406636303, "grad_norm": 2.3808812544559226, "learning_rate": 2.8867326520999882e-08, "loss": 0.1616, "step": 8939 }, { "epoch": 2.9082628497072216, "grad_norm": 2.237036637281483, "learning_rate": 2.8664587519194876e-08, "loss": 0.2029, "step": 8940 }, { "epoch": 2.9085881587508133, "grad_norm": 2.177379308765913, "learning_rate": 2.8462560913068427e-08, "loss": 0.2316, "step": 8941 }, { "epoch": 2.9089134677944046, "grad_norm": 2.2728471964910315, "learning_rate": 2.8261246731572933e-08, "loss": 0.1766, "step": 8942 }, { "epoch": 2.909238776837996, "grad_norm": 2.0127101195214356, "learning_rate": 2.8060645003554764e-08, "loss": 0.1419, "step": 8943 }, { "epoch": 2.9095640858815877, "grad_norm": 2.1164847791788493, "learning_rate": 2.7860755757760927e-08, "loss": 0.2049, "step": 8944 }, { "epoch": 2.909889394925179, "grad_norm": 1.9839112184820091, "learning_rate": 2.766157902283628e-08, "loss": 0.1016, "step": 8945 }, { "epoch": 2.9102147039687702, "grad_norm": 2.3988187249702944, "learning_rate": 2.7463114827322447e-08, "loss": 0.1542, "step": 8946 }, { "epoch": 2.910540013012362, "grad_norm": 2.382074429800079, "learning_rate": 2.7265363199659446e-08, "loss": 0.1965, "step": 8947 }, { "epoch": 2.9108653220559533, "grad_norm": 2.105607692345549, "learning_rate": 2.7068324168185722e-08, "loss": 0.1553, "step": 8948 }, { "epoch": 2.9111906310995446, "grad_norm": 1.6174643204427586, "learning_rate": 2.687199776113647e-08, "loss": 0.1031, "step": 8949 }, { "epoch": 2.911515940143136, "grad_norm": 2.6963640082251428, "learning_rate": 2.667638400664585e-08, "loss": 0.1702, "step": 8950 }, { "epoch": 2.911841249186727, "grad_norm": 1.9440002473010771, "learning_rate": 2.648148293274533e-08, "loss": 0.1595, "step": 8951 }, { "epoch": 2.912166558230319, "grad_norm": 2.311972683564187, "learning_rate": 2.628729456736423e-08, "loss": 0.2001, "step": 8952 }, { "epoch": 2.91249186727391, "grad_norm": 2.254130324661361, "learning_rate": 2.6093818938329742e-08, "loss": 0.1754, "step": 8953 }, { "epoch": 2.9128171763175015, "grad_norm": 1.9396722368543848, "learning_rate": 2.590105607336746e-08, "loss": 0.165, "step": 8954 }, { "epoch": 2.913142485361093, "grad_norm": 2.056525603988778, "learning_rate": 2.5709006000099736e-08, "loss": 0.1636, "step": 8955 }, { "epoch": 2.9134677944046845, "grad_norm": 2.4497692956261106, "learning_rate": 2.551766874604844e-08, "loss": 0.1956, "step": 8956 }, { "epoch": 2.913793103448276, "grad_norm": 2.142866078952475, "learning_rate": 2.5327044338631646e-08, "loss": 0.1395, "step": 8957 }, { "epoch": 2.9141184124918675, "grad_norm": 1.8507108963875687, "learning_rate": 2.5137132805166386e-08, "loss": 0.1108, "step": 8958 }, { "epoch": 2.914443721535459, "grad_norm": 1.8315340542641014, "learning_rate": 2.4947934172867005e-08, "loss": 0.132, "step": 8959 }, { "epoch": 2.91476903057905, "grad_norm": 1.919788072864226, "learning_rate": 2.4759448468845148e-08, "loss": 0.122, "step": 8960 }, { "epoch": 2.9150943396226414, "grad_norm": 2.240330766835081, "learning_rate": 2.457167572011143e-08, "loss": 0.1501, "step": 8961 }, { "epoch": 2.9154196486662327, "grad_norm": 2.0705734074817257, "learning_rate": 2.4384615953574332e-08, "loss": 0.1796, "step": 8962 }, { "epoch": 2.9157449577098244, "grad_norm": 1.9818791867263301, "learning_rate": 2.419826919603907e-08, "loss": 0.1361, "step": 8963 }, { "epoch": 2.9160702667534157, "grad_norm": 1.8606189590680065, "learning_rate": 2.401263547420929e-08, "loss": 0.1451, "step": 8964 }, { "epoch": 2.916395575797007, "grad_norm": 1.9777719607178221, "learning_rate": 2.3827714814686488e-08, "loss": 0.1108, "step": 8965 }, { "epoch": 2.9167208848405988, "grad_norm": 2.184064723812681, "learning_rate": 2.364350724397002e-08, "loss": 0.1838, "step": 8966 }, { "epoch": 2.91704619388419, "grad_norm": 1.8815598495676138, "learning_rate": 2.3460012788456553e-08, "loss": 0.1473, "step": 8967 }, { "epoch": 2.9173715029277814, "grad_norm": 2.175460752835359, "learning_rate": 2.3277231474441165e-08, "loss": 0.1928, "step": 8968 }, { "epoch": 2.9176968119713727, "grad_norm": 2.2086587285721353, "learning_rate": 2.309516332811734e-08, "loss": 0.2193, "step": 8969 }, { "epoch": 2.918022121014964, "grad_norm": 2.043395714419583, "learning_rate": 2.291380837557422e-08, "loss": 0.1685, "step": 8970 }, { "epoch": 2.9183474300585557, "grad_norm": 2.24597192614568, "learning_rate": 2.2733166642801008e-08, "loss": 0.1832, "step": 8971 }, { "epoch": 2.918672739102147, "grad_norm": 2.038150770917819, "learning_rate": 2.255323815568311e-08, "loss": 0.1479, "step": 8972 }, { "epoch": 2.9189980481457383, "grad_norm": 2.0509950351682433, "learning_rate": 2.2374022940005457e-08, "loss": 0.1763, "step": 8973 }, { "epoch": 2.91932335718933, "grad_norm": 2.322519667166926, "learning_rate": 2.219552102144862e-08, "loss": 0.1645, "step": 8974 }, { "epoch": 2.9196486662329213, "grad_norm": 1.7930082410033448, "learning_rate": 2.201773242559213e-08, "loss": 0.1778, "step": 8975 }, { "epoch": 2.9199739752765126, "grad_norm": 2.473980958390003, "learning_rate": 2.184065717791395e-08, "loss": 0.1785, "step": 8976 }, { "epoch": 2.9202992843201043, "grad_norm": 2.4307665060443346, "learning_rate": 2.1664295303788773e-08, "loss": 0.1782, "step": 8977 }, { "epoch": 2.9206245933636956, "grad_norm": 2.0472561327991725, "learning_rate": 2.1488646828489167e-08, "loss": 0.1605, "step": 8978 }, { "epoch": 2.920949902407287, "grad_norm": 2.5267734644007636, "learning_rate": 2.131371177718611e-08, "loss": 0.194, "step": 8979 }, { "epoch": 2.921275211450878, "grad_norm": 2.040951718401503, "learning_rate": 2.113949017494732e-08, "loss": 0.1351, "step": 8980 }, { "epoch": 2.9216005204944695, "grad_norm": 2.295258225269121, "learning_rate": 2.0965982046738942e-08, "loss": 0.1564, "step": 8981 }, { "epoch": 2.9219258295380612, "grad_norm": 2.7506018144348547, "learning_rate": 2.0793187417425532e-08, "loss": 0.2121, "step": 8982 }, { "epoch": 2.9222511385816525, "grad_norm": 1.9444382697465836, "learning_rate": 2.0621106311767282e-08, "loss": 0.1487, "step": 8983 }, { "epoch": 2.922576447625244, "grad_norm": 2.41216126837854, "learning_rate": 2.044973875442502e-08, "loss": 0.1353, "step": 8984 }, { "epoch": 2.9229017566688356, "grad_norm": 2.1155931377756843, "learning_rate": 2.0279084769955215e-08, "loss": 0.1282, "step": 8985 }, { "epoch": 2.923227065712427, "grad_norm": 2.1040284719604667, "learning_rate": 2.010914438281275e-08, "loss": 0.1602, "step": 8986 }, { "epoch": 2.923552374756018, "grad_norm": 2.1231308566078266, "learning_rate": 1.9939917617349814e-08, "loss": 0.1642, "step": 8987 }, { "epoch": 2.92387768379961, "grad_norm": 2.3785119869632045, "learning_rate": 1.9771404497817005e-08, "loss": 0.159, "step": 8988 }, { "epoch": 2.924202992843201, "grad_norm": 1.7866341321397257, "learning_rate": 1.9603605048362783e-08, "loss": 0.1203, "step": 8989 }, { "epoch": 2.9245283018867925, "grad_norm": 2.1118407930582426, "learning_rate": 1.9436519293032364e-08, "loss": 0.1589, "step": 8990 }, { "epoch": 2.9248536109303838, "grad_norm": 2.10149630007463, "learning_rate": 1.9270147255769368e-08, "loss": 0.1889, "step": 8991 }, { "epoch": 2.925178919973975, "grad_norm": 1.914874547337455, "learning_rate": 1.910448896041528e-08, "loss": 0.1706, "step": 8992 }, { "epoch": 2.925504229017567, "grad_norm": 2.1051325369075524, "learning_rate": 1.8939544430708336e-08, "loss": 0.1498, "step": 8993 }, { "epoch": 2.925829538061158, "grad_norm": 2.0055173320679915, "learning_rate": 1.877531369028629e-08, "loss": 0.1344, "step": 8994 }, { "epoch": 2.9261548471047494, "grad_norm": 2.2201737450288275, "learning_rate": 1.8611796762682544e-08, "loss": 0.2015, "step": 8995 }, { "epoch": 2.926480156148341, "grad_norm": 1.9770088670077177, "learning_rate": 1.844899367132946e-08, "loss": 0.1465, "step": 8996 }, { "epoch": 2.9268054651919324, "grad_norm": 1.908198889166606, "learning_rate": 1.8286904439557275e-08, "loss": 0.1383, "step": 8997 }, { "epoch": 2.9271307742355237, "grad_norm": 2.7549661260020284, "learning_rate": 1.812552909059295e-08, "loss": 0.1636, "step": 8998 }, { "epoch": 2.927456083279115, "grad_norm": 2.038226840611822, "learning_rate": 1.796486764756189e-08, "loss": 0.143, "step": 8999 }, { "epoch": 2.9277813923227067, "grad_norm": 2.0569194044886725, "learning_rate": 1.780492013348678e-08, "loss": 0.1496, "step": 9000 }, { "epoch": 2.928106701366298, "grad_norm": 2.4375527792403124, "learning_rate": 1.7645686571288734e-08, "loss": 0.1858, "step": 9001 }, { "epoch": 2.9284320104098893, "grad_norm": 1.82533972049376, "learning_rate": 1.748716698378561e-08, "loss": 0.1438, "step": 9002 }, { "epoch": 2.9287573194534806, "grad_norm": 2.564556430292477, "learning_rate": 1.7329361393693124e-08, "loss": 0.1995, "step": 9003 }, { "epoch": 2.9290826284970723, "grad_norm": 2.109811788009235, "learning_rate": 1.7172269823625966e-08, "loss": 0.157, "step": 9004 }, { "epoch": 2.9294079375406636, "grad_norm": 1.9773786686181634, "learning_rate": 1.7015892296093906e-08, "loss": 0.1423, "step": 9005 }, { "epoch": 2.929733246584255, "grad_norm": 2.338753599758111, "learning_rate": 1.6860228833507353e-08, "loss": 0.1895, "step": 9006 }, { "epoch": 2.9300585556278467, "grad_norm": 2.400647625139459, "learning_rate": 1.6705279458172353e-08, "loss": 0.1664, "step": 9007 }, { "epoch": 2.930383864671438, "grad_norm": 2.2039595776490235, "learning_rate": 1.655104419229281e-08, "loss": 0.1019, "step": 9008 }, { "epoch": 2.9307091737150293, "grad_norm": 1.8179327287208074, "learning_rate": 1.63975230579716e-08, "loss": 0.1536, "step": 9009 }, { "epoch": 2.9310344827586206, "grad_norm": 2.1997921717504263, "learning_rate": 1.6244716077208343e-08, "loss": 0.1599, "step": 9010 }, { "epoch": 2.931359791802212, "grad_norm": 1.9278945282904538, "learning_rate": 1.6092623271899422e-08, "loss": 0.1339, "step": 9011 }, { "epoch": 2.9316851008458036, "grad_norm": 2.05881224709159, "learning_rate": 1.5941244663841282e-08, "loss": 0.1742, "step": 9012 }, { "epoch": 2.932010409889395, "grad_norm": 2.199047167081465, "learning_rate": 1.5790580274725466e-08, "loss": 0.1499, "step": 9013 }, { "epoch": 2.932335718932986, "grad_norm": 2.067897743959286, "learning_rate": 1.5640630126142475e-08, "loss": 0.1686, "step": 9014 }, { "epoch": 2.932661027976578, "grad_norm": 2.281777514519812, "learning_rate": 1.5491394239580125e-08, "loss": 0.1869, "step": 9015 }, { "epoch": 2.932986337020169, "grad_norm": 1.889314526137744, "learning_rate": 1.5342872636424643e-08, "loss": 0.1367, "step": 9016 }, { "epoch": 2.9333116460637605, "grad_norm": 2.288177334600268, "learning_rate": 1.5195065337959e-08, "loss": 0.1742, "step": 9017 }, { "epoch": 2.9336369551073522, "grad_norm": 2.2201702426255046, "learning_rate": 1.5047972365364592e-08, "loss": 0.1987, "step": 9018 }, { "epoch": 2.9339622641509435, "grad_norm": 2.1464216672894243, "learning_rate": 1.4901593739718445e-08, "loss": 0.1489, "step": 9019 }, { "epoch": 2.934287573194535, "grad_norm": 2.0106077307242765, "learning_rate": 1.4755929481997667e-08, "loss": 0.1842, "step": 9020 }, { "epoch": 2.934612882238126, "grad_norm": 2.045185604934483, "learning_rate": 1.461097961307667e-08, "loss": 0.1477, "step": 9021 }, { "epoch": 2.9349381912817174, "grad_norm": 2.231667016845439, "learning_rate": 1.4466744153725509e-08, "loss": 0.1523, "step": 9022 }, { "epoch": 2.935263500325309, "grad_norm": 1.920698727709525, "learning_rate": 1.4323223124614316e-08, "loss": 0.1655, "step": 9023 }, { "epoch": 2.9355888093689004, "grad_norm": 1.8320995584343158, "learning_rate": 1.4180416546308862e-08, "loss": 0.1167, "step": 9024 }, { "epoch": 2.9359141184124917, "grad_norm": 2.190909164323083, "learning_rate": 1.4038324439274442e-08, "loss": 0.1881, "step": 9025 }, { "epoch": 2.9362394274560835, "grad_norm": 2.623047021744601, "learning_rate": 1.3896946823872548e-08, "loss": 0.1845, "step": 9026 }, { "epoch": 2.9365647364996748, "grad_norm": 2.1247930959347436, "learning_rate": 1.3756283720361974e-08, "loss": 0.1419, "step": 9027 }, { "epoch": 2.936890045543266, "grad_norm": 1.5618704669829464, "learning_rate": 1.3616335148900483e-08, "loss": 0.1104, "step": 9028 }, { "epoch": 2.937215354586858, "grad_norm": 2.003254220769076, "learning_rate": 1.3477101129543146e-08, "loss": 0.1538, "step": 9029 }, { "epoch": 2.937540663630449, "grad_norm": 1.6090594870317747, "learning_rate": 1.333858168224178e-08, "loss": 0.101, "step": 9030 }, { "epoch": 2.9378659726740404, "grad_norm": 2.0917345381829913, "learning_rate": 1.3200776826846063e-08, "loss": 0.155, "step": 9031 }, { "epoch": 2.9381912817176317, "grad_norm": 2.0784256549072615, "learning_rate": 1.3063686583104085e-08, "loss": 0.1663, "step": 9032 }, { "epoch": 2.938516590761223, "grad_norm": 1.6910597129932043, "learning_rate": 1.2927310970661244e-08, "loss": 0.1094, "step": 9033 }, { "epoch": 2.9388418998048147, "grad_norm": 2.5240376857465576, "learning_rate": 1.279165000905913e-08, "loss": 0.1505, "step": 9034 }, { "epoch": 2.939167208848406, "grad_norm": 1.9806189400907148, "learning_rate": 1.265670371773886e-08, "loss": 0.1502, "step": 9035 }, { "epoch": 2.9394925178919973, "grad_norm": 2.0899419165675783, "learning_rate": 1.2522472116038297e-08, "loss": 0.175, "step": 9036 }, { "epoch": 2.939817826935589, "grad_norm": 1.933438921767652, "learning_rate": 1.238895522319261e-08, "loss": 0.1538, "step": 9037 }, { "epoch": 2.9401431359791803, "grad_norm": 2.195688599991983, "learning_rate": 1.225615305833483e-08, "loss": 0.1265, "step": 9038 }, { "epoch": 2.9404684450227716, "grad_norm": 2.121766639187889, "learning_rate": 1.2124065640496396e-08, "loss": 0.198, "step": 9039 }, { "epoch": 2.940793754066363, "grad_norm": 2.0036552307153896, "learning_rate": 1.1992692988604393e-08, "loss": 0.1357, "step": 9040 }, { "epoch": 2.941119063109954, "grad_norm": 2.0291015909063455, "learning_rate": 1.1862035121485426e-08, "loss": 0.1305, "step": 9041 }, { "epoch": 2.941444372153546, "grad_norm": 2.365408229622232, "learning_rate": 1.1732092057862298e-08, "loss": 0.1526, "step": 9042 }, { "epoch": 2.941769681197137, "grad_norm": 2.089265288904804, "learning_rate": 1.1602863816356225e-08, "loss": 0.1427, "step": 9043 }, { "epoch": 2.9420949902407285, "grad_norm": 2.138250934145395, "learning_rate": 1.1474350415485724e-08, "loss": 0.1879, "step": 9044 }, { "epoch": 2.9424202992843203, "grad_norm": 1.8647047048229877, "learning_rate": 1.1346551873667178e-08, "loss": 0.1563, "step": 9045 }, { "epoch": 2.9427456083279115, "grad_norm": 1.7469271515320868, "learning_rate": 1.1219468209213157e-08, "loss": 0.0983, "step": 9046 }, { "epoch": 2.943070917371503, "grad_norm": 1.9103173821042458, "learning_rate": 1.1093099440335764e-08, "loss": 0.1357, "step": 9047 }, { "epoch": 2.9433962264150946, "grad_norm": 2.2407147487189754, "learning_rate": 1.0967445585143842e-08, "loss": 0.1397, "step": 9048 }, { "epoch": 2.943721535458686, "grad_norm": 2.5149587994216325, "learning_rate": 1.0842506661642993e-08, "loss": 0.1683, "step": 9049 }, { "epoch": 2.944046844502277, "grad_norm": 2.816100638770473, "learning_rate": 1.071828268773778e-08, "loss": 0.2264, "step": 9050 }, { "epoch": 2.9443721535458685, "grad_norm": 2.069792570053991, "learning_rate": 1.059477368122841e-08, "loss": 0.1438, "step": 9051 }, { "epoch": 2.9446974625894597, "grad_norm": 2.1009131896687543, "learning_rate": 1.0471979659815168e-08, "loss": 0.1642, "step": 9052 }, { "epoch": 2.9450227716330515, "grad_norm": 2.0204550229298825, "learning_rate": 1.0349900641093424e-08, "loss": 0.1547, "step": 9053 }, { "epoch": 2.9453480806766428, "grad_norm": 2.3483769002571337, "learning_rate": 1.0228536642558073e-08, "loss": 0.2084, "step": 9054 }, { "epoch": 2.945673389720234, "grad_norm": 2.652889185885779, "learning_rate": 1.0107887681600204e-08, "loss": 0.1934, "step": 9055 }, { "epoch": 2.945998698763826, "grad_norm": 2.6608664072503396, "learning_rate": 9.987953775508764e-09, "loss": 0.1387, "step": 9056 }, { "epoch": 2.946324007807417, "grad_norm": 2.006525584157406, "learning_rate": 9.86873494147056e-09, "loss": 0.1827, "step": 9057 }, { "epoch": 2.9466493168510084, "grad_norm": 2.2643223832381953, "learning_rate": 9.750231196569704e-09, "loss": 0.1332, "step": 9058 }, { "epoch": 2.9469746258946, "grad_norm": 2.1113527578791733, "learning_rate": 9.632442557788168e-09, "loss": 0.1362, "step": 9059 }, { "epoch": 2.9472999349381914, "grad_norm": 2.1376247282094054, "learning_rate": 9.515369042004119e-09, "loss": 0.1059, "step": 9060 }, { "epoch": 2.9476252439817827, "grad_norm": 1.9899274687106263, "learning_rate": 9.3990106659958e-09, "loss": 0.1373, "step": 9061 }, { "epoch": 2.947950553025374, "grad_norm": 2.528480536352175, "learning_rate": 9.283367446435986e-09, "loss": 0.1897, "step": 9062 }, { "epoch": 2.9482758620689653, "grad_norm": 2.0419291893991858, "learning_rate": 9.168439399897533e-09, "loss": 0.1411, "step": 9063 }, { "epoch": 2.948601171112557, "grad_norm": 2.0253484956893972, "learning_rate": 9.054226542848932e-09, "loss": 0.1484, "step": 9064 }, { "epoch": 2.9489264801561483, "grad_norm": 2.2533064703389063, "learning_rate": 8.940728891657647e-09, "loss": 0.1685, "step": 9065 }, { "epoch": 2.9492517891997396, "grad_norm": 2.4434004718533378, "learning_rate": 8.827946462587333e-09, "loss": 0.2181, "step": 9066 }, { "epoch": 2.9495770982433314, "grad_norm": 1.8241914555277041, "learning_rate": 8.715879271800064e-09, "loss": 0.1391, "step": 9067 }, { "epoch": 2.9499024072869227, "grad_norm": 2.623793926519133, "learning_rate": 8.604527335355218e-09, "loss": 0.1706, "step": 9068 }, { "epoch": 2.950227716330514, "grad_norm": 2.337144981724471, "learning_rate": 8.49389066920947e-09, "loss": 0.1999, "step": 9069 }, { "epoch": 2.9505530253741052, "grad_norm": 2.3872448554605126, "learning_rate": 8.383969289217364e-09, "loss": 0.2282, "step": 9070 }, { "epoch": 2.9508783344176965, "grad_norm": 1.9348387890860752, "learning_rate": 8.274763211130189e-09, "loss": 0.1261, "step": 9071 }, { "epoch": 2.9512036434612883, "grad_norm": 2.263264851440659, "learning_rate": 8.166272450597645e-09, "loss": 0.1731, "step": 9072 }, { "epoch": 2.9515289525048796, "grad_norm": 1.9571126861436112, "learning_rate": 8.058497023166745e-09, "loss": 0.1685, "step": 9073 }, { "epoch": 2.951854261548471, "grad_norm": 2.141965876317623, "learning_rate": 7.951436944281243e-09, "loss": 0.1855, "step": 9074 }, { "epoch": 2.9521795705920626, "grad_norm": 2.079058226278506, "learning_rate": 7.845092229283313e-09, "loss": 0.1425, "step": 9075 }, { "epoch": 2.952504879635654, "grad_norm": 2.030437637473696, "learning_rate": 7.739462893411875e-09, "loss": 0.138, "step": 9076 }, { "epoch": 2.952830188679245, "grad_norm": 1.963956030138211, "learning_rate": 7.634548951803712e-09, "loss": 0.1689, "step": 9077 }, { "epoch": 2.953155497722837, "grad_norm": 2.012790393622604, "learning_rate": 7.530350419494014e-09, "loss": 0.133, "step": 9078 }, { "epoch": 2.953480806766428, "grad_norm": 2.301697462049199, "learning_rate": 7.426867311413067e-09, "loss": 0.2018, "step": 9079 }, { "epoch": 2.9538061158100195, "grad_norm": 2.2262605137516305, "learning_rate": 7.32409964239067e-09, "loss": 0.1874, "step": 9080 }, { "epoch": 2.954131424853611, "grad_norm": 1.974122222860128, "learning_rate": 7.2220474271539335e-09, "loss": 0.1542, "step": 9081 }, { "epoch": 2.954456733897202, "grad_norm": 2.3445699085688645, "learning_rate": 7.1207106803267145e-09, "loss": 0.1584, "step": 9082 }, { "epoch": 2.954782042940794, "grad_norm": 2.267299787746493, "learning_rate": 7.020089416430176e-09, "loss": 0.1478, "step": 9083 }, { "epoch": 2.955107351984385, "grad_norm": 2.1496491442408217, "learning_rate": 6.9201836498838935e-09, "loss": 0.1818, "step": 9084 }, { "epoch": 2.9554326610279764, "grad_norm": 2.5000678210627734, "learning_rate": 6.820993395004194e-09, "loss": 0.2129, "step": 9085 }, { "epoch": 2.955757970071568, "grad_norm": 2.142629189483976, "learning_rate": 6.722518666005817e-09, "loss": 0.1641, "step": 9086 }, { "epoch": 2.9560832791151594, "grad_norm": 2.038322023960769, "learning_rate": 6.624759476999143e-09, "loss": 0.1681, "step": 9087 }, { "epoch": 2.9564085881587507, "grad_norm": 2.0242073255151283, "learning_rate": 6.5277158419940756e-09, "loss": 0.1434, "step": 9088 }, { "epoch": 2.9567338972023425, "grad_norm": 2.214917875007446, "learning_rate": 6.431387774896158e-09, "loss": 0.1809, "step": 9089 }, { "epoch": 2.9570592062459338, "grad_norm": 2.1406304173124187, "learning_rate": 6.335775289509905e-09, "loss": 0.1599, "step": 9090 }, { "epoch": 2.957384515289525, "grad_norm": 2.4446934626871224, "learning_rate": 6.2408783995365766e-09, "loss": 0.2001, "step": 9091 }, { "epoch": 2.9577098243331164, "grad_norm": 2.14440861747328, "learning_rate": 6.146697118574741e-09, "loss": 0.1839, "step": 9092 }, { "epoch": 2.9580351333767076, "grad_norm": 2.073027775100094, "learning_rate": 6.0532314601208235e-09, "loss": 0.1296, "step": 9093 }, { "epoch": 2.9583604424202994, "grad_norm": 2.091468415000422, "learning_rate": 5.9604814375685546e-09, "loss": 0.1795, "step": 9094 }, { "epoch": 2.9586857514638907, "grad_norm": 2.047164067482997, "learning_rate": 5.8684470642084154e-09, "loss": 0.2027, "step": 9095 }, { "epoch": 2.959011060507482, "grad_norm": 1.9691345716088122, "learning_rate": 5.77712835323041e-09, "loss": 0.1151, "step": 9096 }, { "epoch": 2.9593363695510737, "grad_norm": 2.22567216680152, "learning_rate": 5.6865253177190714e-09, "loss": 0.177, "step": 9097 }, { "epoch": 2.959661678594665, "grad_norm": 2.5460921798339764, "learning_rate": 5.596637970658459e-09, "loss": 0.186, "step": 9098 }, { "epoch": 2.9599869876382563, "grad_norm": 2.1177033765399127, "learning_rate": 5.507466324929933e-09, "loss": 0.1648, "step": 9099 }, { "epoch": 2.960312296681848, "grad_norm": 2.323989169264866, "learning_rate": 5.419010393311608e-09, "loss": 0.1597, "step": 9100 }, { "epoch": 2.9606376057254393, "grad_norm": 2.155914966862075, "learning_rate": 5.3312701884788986e-09, "loss": 0.1376, "step": 9101 }, { "epoch": 2.9609629147690306, "grad_norm": 2.375220399173371, "learning_rate": 5.244245723005082e-09, "loss": 0.1869, "step": 9102 }, { "epoch": 2.961288223812622, "grad_norm": 2.248471412352345, "learning_rate": 5.157937009361291e-09, "loss": 0.187, "step": 9103 }, { "epoch": 2.961613532856213, "grad_norm": 1.9632208835211138, "learning_rate": 5.072344059914858e-09, "loss": 0.1286, "step": 9104 }, { "epoch": 2.961938841899805, "grad_norm": 2.1538017179444076, "learning_rate": 4.987466886932635e-09, "loss": 0.1487, "step": 9105 }, { "epoch": 2.9622641509433962, "grad_norm": 1.7297757928973514, "learning_rate": 4.903305502576006e-09, "loss": 0.1259, "step": 9106 }, { "epoch": 2.9625894599869875, "grad_norm": 2.1233799273083496, "learning_rate": 4.819859918905878e-09, "loss": 0.1648, "step": 9107 }, { "epoch": 2.9629147690305793, "grad_norm": 2.2199299090538642, "learning_rate": 4.7371301478810195e-09, "loss": 0.1507, "step": 9108 }, { "epoch": 2.9632400780741706, "grad_norm": 2.1340211823883295, "learning_rate": 4.655116201355281e-09, "loss": 0.1853, "step": 9109 }, { "epoch": 2.963565387117762, "grad_norm": 2.182165510170061, "learning_rate": 4.5738180910820385e-09, "loss": 0.1916, "step": 9110 }, { "epoch": 2.963890696161353, "grad_norm": 1.877670940597524, "learning_rate": 4.4932358287119726e-09, "loss": 0.1342, "step": 9111 }, { "epoch": 2.9642160052049444, "grad_norm": 1.949740755259622, "learning_rate": 4.413369425790848e-09, "loss": 0.1567, "step": 9112 }, { "epoch": 2.964541314248536, "grad_norm": 2.4165256570198834, "learning_rate": 4.334218893765063e-09, "loss": 0.1643, "step": 9113 }, { "epoch": 2.9648666232921275, "grad_norm": 1.955851532387126, "learning_rate": 4.255784243976657e-09, "loss": 0.1519, "step": 9114 }, { "epoch": 2.9651919323357188, "grad_norm": 2.13401981022908, "learning_rate": 4.178065487664973e-09, "loss": 0.1719, "step": 9115 }, { "epoch": 2.9655172413793105, "grad_norm": 1.7902611265289603, "learning_rate": 4.1010626359672125e-09, "loss": 0.1294, "step": 9116 }, { "epoch": 2.965842550422902, "grad_norm": 2.0294414216473764, "learning_rate": 4.024775699918438e-09, "loss": 0.159, "step": 9117 }, { "epoch": 2.966167859466493, "grad_norm": 2.035462001112047, "learning_rate": 3.949204690449904e-09, "loss": 0.159, "step": 9118 }, { "epoch": 2.966493168510085, "grad_norm": 2.122612262195843, "learning_rate": 3.8743496183918376e-09, "loss": 0.1575, "step": 9119 }, { "epoch": 2.966818477553676, "grad_norm": 2.1165840016391066, "learning_rate": 3.800210494470103e-09, "loss": 0.1317, "step": 9120 }, { "epoch": 2.9671437865972674, "grad_norm": 1.9263882199827802, "learning_rate": 3.726787329309534e-09, "loss": 0.1128, "step": 9121 }, { "epoch": 2.9674690956408587, "grad_norm": 2.2595322856514506, "learning_rate": 3.6540801334311595e-09, "loss": 0.1499, "step": 9122 }, { "epoch": 2.96779440468445, "grad_norm": 2.3317779854388485, "learning_rate": 3.5820889172544227e-09, "loss": 0.2008, "step": 9123 }, { "epoch": 2.9681197137280417, "grad_norm": 2.0213602976468374, "learning_rate": 3.510813691096071e-09, "loss": 0.1394, "step": 9124 }, { "epoch": 2.968445022771633, "grad_norm": 2.4300965477073437, "learning_rate": 3.4402544651690463e-09, "loss": 0.2052, "step": 9125 }, { "epoch": 2.9687703318152243, "grad_norm": 1.9697165182328493, "learning_rate": 3.370411249584704e-09, "loss": 0.144, "step": 9126 }, { "epoch": 2.969095640858816, "grad_norm": 1.9498271905947981, "learning_rate": 3.3012840543522605e-09, "loss": 0.1581, "step": 9127 }, { "epoch": 2.9694209499024073, "grad_norm": 2.3014945990680475, "learning_rate": 3.2328728893771256e-09, "loss": 0.159, "step": 9128 }, { "epoch": 2.9697462589459986, "grad_norm": 2.0430984927020335, "learning_rate": 3.1651777644625682e-09, "loss": 0.1136, "step": 9129 }, { "epoch": 2.9700715679895904, "grad_norm": 1.9665287654288093, "learning_rate": 3.098198689309717e-09, "loss": 0.1669, "step": 9130 }, { "epoch": 2.9703968770331817, "grad_norm": 2.2686323102201946, "learning_rate": 3.031935673517006e-09, "loss": 0.2005, "step": 9131 }, { "epoch": 2.970722186076773, "grad_norm": 2.141165446194031, "learning_rate": 2.9663887265790615e-09, "loss": 0.1554, "step": 9132 }, { "epoch": 2.9710474951203643, "grad_norm": 2.322581700242517, "learning_rate": 2.901557857889481e-09, "loss": 0.1926, "step": 9133 }, { "epoch": 2.9713728041639555, "grad_norm": 2.108828441854059, "learning_rate": 2.8374430767380556e-09, "loss": 0.16, "step": 9134 }, { "epoch": 2.9716981132075473, "grad_norm": 2.4050670660248015, "learning_rate": 2.7740443923129914e-09, "loss": 0.2271, "step": 9135 }, { "epoch": 2.9720234222511386, "grad_norm": 1.9405765014792107, "learning_rate": 2.7113618136992427e-09, "loss": 0.1671, "step": 9136 }, { "epoch": 2.97234873129473, "grad_norm": 2.117781152739391, "learning_rate": 2.6493953498790692e-09, "loss": 0.1906, "step": 9137 }, { "epoch": 2.9726740403383216, "grad_norm": 2.15430158502146, "learning_rate": 2.5881450097325898e-09, "loss": 0.1829, "step": 9138 }, { "epoch": 2.972999349381913, "grad_norm": 2.4510893988489313, "learning_rate": 2.527610802036673e-09, "loss": 0.181, "step": 9139 }, { "epoch": 2.973324658425504, "grad_norm": 2.373486451799443, "learning_rate": 2.4677927354660457e-09, "loss": 0.1841, "step": 9140 }, { "epoch": 2.9736499674690955, "grad_norm": 2.3145868497613473, "learning_rate": 2.4086908185927406e-09, "loss": 0.1485, "step": 9141 }, { "epoch": 2.973975276512687, "grad_norm": 2.304634086044517, "learning_rate": 2.3503050598860933e-09, "loss": 0.1637, "step": 9142 }, { "epoch": 2.9743005855562785, "grad_norm": 2.0862745635864925, "learning_rate": 2.2926354677133e-09, "loss": 0.182, "step": 9143 }, { "epoch": 2.97462589459987, "grad_norm": 2.2030262484202456, "learning_rate": 2.2356820503371957e-09, "loss": 0.1823, "step": 9144 }, { "epoch": 2.974951203643461, "grad_norm": 2.5396971176831977, "learning_rate": 2.1794448159206947e-09, "loss": 0.154, "step": 9145 }, { "epoch": 2.975276512687053, "grad_norm": 2.3552895497655895, "learning_rate": 2.1239237725217965e-09, "loss": 0.1616, "step": 9146 }, { "epoch": 2.975601821730644, "grad_norm": 2.219940532258486, "learning_rate": 2.0691189280963588e-09, "loss": 0.181, "step": 9147 }, { "epoch": 2.9759271307742354, "grad_norm": 2.0035483580562023, "learning_rate": 2.015030290498654e-09, "loss": 0.165, "step": 9148 }, { "epoch": 2.976252439817827, "grad_norm": 2.130817847966549, "learning_rate": 1.961657867479705e-09, "loss": 0.1626, "step": 9149 }, { "epoch": 2.9765777488614185, "grad_norm": 2.0416280229605794, "learning_rate": 1.9090016666872824e-09, "loss": 0.1178, "step": 9150 }, { "epoch": 2.9769030579050098, "grad_norm": 1.7796529593193402, "learning_rate": 1.8570616956675724e-09, "loss": 0.1343, "step": 9151 }, { "epoch": 2.977228366948601, "grad_norm": 2.1434944916896637, "learning_rate": 1.8058379618629551e-09, "loss": 0.1742, "step": 9152 }, { "epoch": 2.9775536759921923, "grad_norm": 2.1372129918387386, "learning_rate": 1.7553304726142251e-09, "loss": 0.215, "step": 9153 }, { "epoch": 2.977878985035784, "grad_norm": 1.9563427638840811, "learning_rate": 1.7055392351594813e-09, "loss": 0.1392, "step": 9154 }, { "epoch": 2.9782042940793754, "grad_norm": 2.4572727772616454, "learning_rate": 1.656464256633017e-09, "loss": 0.163, "step": 9155 }, { "epoch": 2.9785296031229667, "grad_norm": 2.0662376786937786, "learning_rate": 1.6081055440680949e-09, "loss": 0.1849, "step": 9156 }, { "epoch": 2.9788549121665584, "grad_norm": 1.990387914227451, "learning_rate": 1.560463104394172e-09, "loss": 0.1148, "step": 9157 }, { "epoch": 2.9791802212101497, "grad_norm": 2.130426941844139, "learning_rate": 1.513536944438565e-09, "loss": 0.1419, "step": 9158 }, { "epoch": 2.979505530253741, "grad_norm": 1.9698083340645436, "learning_rate": 1.4673270709253396e-09, "loss": 0.1891, "step": 9159 }, { "epoch": 2.9798308392973327, "grad_norm": 2.636272923626391, "learning_rate": 1.421833490477531e-09, "loss": 0.2072, "step": 9160 }, { "epoch": 2.980156148340924, "grad_norm": 2.481854157447376, "learning_rate": 1.377056209613259e-09, "loss": 0.1959, "step": 9161 }, { "epoch": 2.9804814573845153, "grad_norm": 1.9555890549970176, "learning_rate": 1.3329952347490572e-09, "loss": 0.1292, "step": 9162 }, { "epoch": 2.9808067664281066, "grad_norm": 1.7059194664808157, "learning_rate": 1.2896505721998743e-09, "loss": 0.1136, "step": 9163 }, { "epoch": 2.981132075471698, "grad_norm": 2.3486595838412914, "learning_rate": 1.2470222281768529e-09, "loss": 0.1743, "step": 9164 }, { "epoch": 2.9814573845152896, "grad_norm": 2.413828560165803, "learning_rate": 1.2051102087878852e-09, "loss": 0.2296, "step": 9165 }, { "epoch": 2.981782693558881, "grad_norm": 2.088956372730522, "learning_rate": 1.1639145200398327e-09, "loss": 0.1851, "step": 9166 }, { "epoch": 2.982108002602472, "grad_norm": 2.289800719867303, "learning_rate": 1.1234351678357514e-09, "loss": 0.157, "step": 9167 }, { "epoch": 2.982433311646064, "grad_norm": 2.388771389990641, "learning_rate": 1.0836721579760013e-09, "loss": 0.1937, "step": 9168 }, { "epoch": 2.9827586206896552, "grad_norm": 2.15429717883948, "learning_rate": 1.0446254961588021e-09, "loss": 0.1642, "step": 9169 }, { "epoch": 2.9830839297332465, "grad_norm": 2.964498279720191, "learning_rate": 1.006295187980233e-09, "loss": 0.1923, "step": 9170 }, { "epoch": 2.983409238776838, "grad_norm": 1.7738873254449483, "learning_rate": 9.686812389320122e-10, "loss": 0.1188, "step": 9171 }, { "epoch": 2.9837345478204296, "grad_norm": 2.127395469125764, "learning_rate": 9.317836544053826e-10, "loss": 0.1928, "step": 9172 }, { "epoch": 2.984059856864021, "grad_norm": 2.5739023324742907, "learning_rate": 8.956024396866713e-10, "loss": 0.1726, "step": 9173 }, { "epoch": 2.984385165907612, "grad_norm": 2.0229932223329183, "learning_rate": 8.60137599961175e-10, "loss": 0.1482, "step": 9174 }, { "epoch": 2.9847104749512035, "grad_norm": 2.1725475300947625, "learning_rate": 8.253891403114944e-10, "loss": 0.133, "step": 9175 }, { "epoch": 2.985035783994795, "grad_norm": 2.0713214381873173, "learning_rate": 7.913570657164249e-10, "loss": 0.1443, "step": 9176 }, { "epoch": 2.9853610930383865, "grad_norm": 1.840487294187702, "learning_rate": 7.580413810526211e-10, "loss": 0.1551, "step": 9177 }, { "epoch": 2.9856864020819778, "grad_norm": 2.204241013765666, "learning_rate": 7.254420910951521e-10, "loss": 0.1691, "step": 9178 }, { "epoch": 2.9860117111255695, "grad_norm": 2.324365792795966, "learning_rate": 6.935592005147263e-10, "loss": 0.1751, "step": 9179 }, { "epoch": 2.986337020169161, "grad_norm": 2.0371552322732582, "learning_rate": 6.623927138804665e-10, "loss": 0.1546, "step": 9180 }, { "epoch": 2.986662329212752, "grad_norm": 1.9672487313390603, "learning_rate": 6.319426356587999e-10, "loss": 0.1443, "step": 9181 }, { "epoch": 2.9869876382563434, "grad_norm": 2.081813679896192, "learning_rate": 6.022089702123479e-10, "loss": 0.1743, "step": 9182 }, { "epoch": 2.9873129472999347, "grad_norm": 2.1953150551396807, "learning_rate": 5.731917218032568e-10, "loss": 0.1383, "step": 9183 }, { "epoch": 2.9876382563435264, "grad_norm": 2.1231390754077806, "learning_rate": 5.448908945887565e-10, "loss": 0.1359, "step": 9184 }, { "epoch": 2.9879635653871177, "grad_norm": 2.1237022055796317, "learning_rate": 5.17306492625047e-10, "loss": 0.1509, "step": 9185 }, { "epoch": 2.988288874430709, "grad_norm": 2.0251458986192916, "learning_rate": 4.904385198639672e-10, "loss": 0.1358, "step": 9186 }, { "epoch": 2.9886141834743007, "grad_norm": 1.8481069919727948, "learning_rate": 4.6428698015688054e-10, "loss": 0.1424, "step": 9187 }, { "epoch": 2.988939492517892, "grad_norm": 2.0178179641088034, "learning_rate": 4.3885187725078995e-10, "loss": 0.1191, "step": 9188 }, { "epoch": 2.9892648015614833, "grad_norm": 2.0243791248652965, "learning_rate": 4.1413321479000236e-10, "loss": 0.1538, "step": 9189 }, { "epoch": 2.989590110605075, "grad_norm": 1.9541344828852738, "learning_rate": 3.9013099631779463e-10, "loss": 0.1611, "step": 9190 }, { "epoch": 2.9899154196486664, "grad_norm": 2.2922330491169336, "learning_rate": 3.668452252730825e-10, "loss": 0.1818, "step": 9191 }, { "epoch": 2.9902407286922577, "grad_norm": 1.9815495468363256, "learning_rate": 3.442759049931965e-10, "loss": 0.1299, "step": 9192 }, { "epoch": 2.990566037735849, "grad_norm": 2.3221454241350687, "learning_rate": 3.2242303871166114e-10, "loss": 0.1677, "step": 9193 }, { "epoch": 2.9908913467794402, "grad_norm": 2.3422079785933825, "learning_rate": 3.012866295604155e-10, "loss": 0.1812, "step": 9194 }, { "epoch": 2.991216655823032, "grad_norm": 1.8323466918086275, "learning_rate": 2.8086668056814816e-10, "loss": 0.1145, "step": 9195 }, { "epoch": 2.9915419648666233, "grad_norm": 1.7822526108466061, "learning_rate": 2.611631946614068e-10, "loss": 0.1165, "step": 9196 }, { "epoch": 2.9918672739102146, "grad_norm": 1.8737313865491392, "learning_rate": 2.4217617466293363e-10, "loss": 0.1723, "step": 9197 }, { "epoch": 2.9921925829538063, "grad_norm": 1.7806659380835341, "learning_rate": 2.2390562329444033e-10, "loss": 0.1487, "step": 9198 }, { "epoch": 2.9925178919973976, "grad_norm": 2.0738683152358584, "learning_rate": 2.0635154317383276e-10, "loss": 0.1826, "step": 9199 }, { "epoch": 2.992843201040989, "grad_norm": 2.167128826371228, "learning_rate": 1.8951393681632125e-10, "loss": 0.1233, "step": 9200 }, { "epoch": 2.9931685100845806, "grad_norm": 2.400724485193199, "learning_rate": 1.733928066349755e-10, "loss": 0.2159, "step": 9201 }, { "epoch": 2.993493819128172, "grad_norm": 1.9891841199337124, "learning_rate": 1.579881549401696e-10, "loss": 0.115, "step": 9202 }, { "epoch": 2.993819128171763, "grad_norm": 2.179980113953543, "learning_rate": 1.432999839395821e-10, "loss": 0.138, "step": 9203 }, { "epoch": 2.9941444372153545, "grad_norm": 1.9404810987934322, "learning_rate": 1.293282957370856e-10, "loss": 0.1413, "step": 9204 }, { "epoch": 2.994469746258946, "grad_norm": 2.300174015246515, "learning_rate": 1.1607309233552244e-10, "loss": 0.2031, "step": 9205 }, { "epoch": 2.9947950553025375, "grad_norm": 2.336164079188448, "learning_rate": 1.0353437563448421e-10, "loss": 0.1333, "step": 9206 }, { "epoch": 2.995120364346129, "grad_norm": 2.4438118533673228, "learning_rate": 9.171214743031175e-11, "loss": 0.1764, "step": 9207 }, { "epoch": 2.99544567338972, "grad_norm": 1.9321551193974287, "learning_rate": 8.060640941720543e-11, "loss": 0.1628, "step": 9208 }, { "epoch": 2.995770982433312, "grad_norm": 2.1576227156072827, "learning_rate": 7.021716318722504e-11, "loss": 0.1519, "step": 9209 }, { "epoch": 2.996096291476903, "grad_norm": 2.210144222959531, "learning_rate": 6.054441022806945e-11, "loss": 0.2024, "step": 9210 }, { "epoch": 2.9964216005204944, "grad_norm": 2.0094138657324745, "learning_rate": 5.1588151926962316e-11, "loss": 0.1446, "step": 9211 }, { "epoch": 2.9967469095640857, "grad_norm": 1.9217657527847374, "learning_rate": 4.334838956732146e-11, "loss": 0.1369, "step": 9212 }, { "epoch": 2.997072218607677, "grad_norm": 2.274364927667433, "learning_rate": 3.582512432875884e-11, "loss": 0.1512, "step": 9213 }, { "epoch": 2.9973975276512688, "grad_norm": 2.401943368402577, "learning_rate": 2.9018357290411247e-11, "loss": 0.2337, "step": 9214 }, { "epoch": 2.99772283669486, "grad_norm": 2.295934794676234, "learning_rate": 2.292808942705449e-11, "loss": 0.193, "step": 9215 }, { "epoch": 2.9980481457384514, "grad_norm": 2.262508577200657, "learning_rate": 1.755432161243409e-11, "loss": 0.1586, "step": 9216 }, { "epoch": 2.998373454782043, "grad_norm": 1.7360759710147908, "learning_rate": 1.289705461537949e-11, "loss": 0.1165, "step": 9217 }, { "epoch": 2.9986987638256344, "grad_norm": 2.294260424669441, "learning_rate": 8.956289103689842e-12, "loss": 0.1649, "step": 9218 }, { "epoch": 2.9990240728692257, "grad_norm": 2.010138406676108, "learning_rate": 5.732025642468664e-12, "loss": 0.1441, "step": 9219 }, { "epoch": 2.9993493819128174, "grad_norm": 1.9143980072335924, "learning_rate": 3.224264693568735e-12, "loss": 0.1451, "step": 9220 }, { "epoch": 2.9996746909564087, "grad_norm": 2.67584620311104, "learning_rate": 1.4330066161472033e-12, "loss": 0.259, "step": 9221 }, { "epoch": 3.0, "grad_norm": 2.4837642723562325, "learning_rate": 3.5825166666558777e-13, "loss": 0.1955, "step": 9222 }, { "epoch": 3.0, "step": 9222, "total_flos": 154685539491840.0, "train_loss": 0.3971013587030393, "train_runtime": 54304.2861, "train_samples_per_second": 4.076, "train_steps_per_second": 0.17 } ], "logging_steps": 1, "max_steps": 9222, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 4000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 154685539491840.0, "train_batch_size": 1, "trial_name": null, "trial_params": null }