{ "best_metric": null, "best_model_checkpoint": null, "epoch": 10.0, "eval_steps": 500, "global_step": 15630, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0006397952655150352, "grad_norm": 28.230282244523245, "learning_rate": 4.264392324093817e-08, "loss": 2.228, "step": 1 }, { "epoch": 0.0012795905310300703, "grad_norm": 28.779932942332085, "learning_rate": 8.528784648187634e-08, "loss": 2.2493, "step": 2 }, { "epoch": 0.0019193857965451055, "grad_norm": 27.627790624512762, "learning_rate": 1.279317697228145e-07, "loss": 2.1782, "step": 3 }, { "epoch": 0.0025591810620601407, "grad_norm": 29.458570217228345, "learning_rate": 1.7057569296375268e-07, "loss": 2.2629, "step": 4 }, { "epoch": 0.003198976327575176, "grad_norm": 28.539043984363687, "learning_rate": 2.1321961620469084e-07, "loss": 2.2335, "step": 5 }, { "epoch": 0.003838771593090211, "grad_norm": 28.393186411651307, "learning_rate": 2.55863539445629e-07, "loss": 2.2484, "step": 6 }, { "epoch": 0.004478566858605247, "grad_norm": 27.42742490718569, "learning_rate": 2.9850746268656716e-07, "loss": 2.2517, "step": 7 }, { "epoch": 0.005118362124120281, "grad_norm": 27.212240529726795, "learning_rate": 3.4115138592750537e-07, "loss": 2.1907, "step": 8 }, { "epoch": 0.005758157389635317, "grad_norm": 25.343488000474597, "learning_rate": 3.837953091684436e-07, "loss": 2.1539, "step": 9 }, { "epoch": 0.006397952655150352, "grad_norm": 25.500591604538503, "learning_rate": 4.264392324093817e-07, "loss": 2.1852, "step": 10 }, { "epoch": 0.007037747920665387, "grad_norm": 23.059809508295, "learning_rate": 4.690831556503199e-07, "loss": 2.1737, "step": 11 }, { "epoch": 0.007677543186180422, "grad_norm": 20.216887024842492, "learning_rate": 5.11727078891258e-07, "loss": 2.1319, "step": 12 }, { "epoch": 0.008317338451695458, "grad_norm": 19.572609912369835, "learning_rate": 5.543710021321962e-07, "loss": 2.1445, "step": 13 }, { "epoch": 0.008957133717210493, "grad_norm": 18.151237961010118, "learning_rate": 5.970149253731343e-07, "loss": 2.0963, "step": 14 }, { "epoch": 0.009596928982725527, "grad_norm": 16.288380651565078, "learning_rate": 6.396588486140725e-07, "loss": 2.0475, "step": 15 }, { "epoch": 0.010236724248240563, "grad_norm": 11.567397858925595, "learning_rate": 6.823027718550107e-07, "loss": 2.0006, "step": 16 }, { "epoch": 0.010876519513755598, "grad_norm": 11.45254483269606, "learning_rate": 7.249466950959488e-07, "loss": 1.9721, "step": 17 }, { "epoch": 0.011516314779270634, "grad_norm": 11.039873166692644, "learning_rate": 7.675906183368872e-07, "loss": 1.9778, "step": 18 }, { "epoch": 0.01215611004478567, "grad_norm": 10.986931822993936, "learning_rate": 8.102345415778253e-07, "loss": 1.9656, "step": 19 }, { "epoch": 0.012795905310300703, "grad_norm": 10.713179444284664, "learning_rate": 8.528784648187634e-07, "loss": 1.9253, "step": 20 }, { "epoch": 0.013435700575815739, "grad_norm": 9.043082395494617, "learning_rate": 8.955223880597015e-07, "loss": 1.8651, "step": 21 }, { "epoch": 0.014075495841330775, "grad_norm": 8.292588314596529, "learning_rate": 9.381663113006398e-07, "loss": 1.7508, "step": 22 }, { "epoch": 0.01471529110684581, "grad_norm": 8.556094907425562, "learning_rate": 9.808102345415779e-07, "loss": 1.7667, "step": 23 }, { "epoch": 0.015355086372360844, "grad_norm": 8.469722767708081, "learning_rate": 1.023454157782516e-06, "loss": 1.7727, "step": 24 }, { "epoch": 0.01599488163787588, "grad_norm": 8.241171463186367, "learning_rate": 1.066098081023454e-06, "loss": 1.7874, "step": 25 }, { "epoch": 0.016634676903390915, "grad_norm": 7.650549255659919, "learning_rate": 1.1087420042643924e-06, "loss": 1.8013, "step": 26 }, { "epoch": 0.01727447216890595, "grad_norm": 8.071139325445692, "learning_rate": 1.1513859275053305e-06, "loss": 1.7578, "step": 27 }, { "epoch": 0.017914267434420986, "grad_norm": 7.316423040772117, "learning_rate": 1.1940298507462686e-06, "loss": 1.6905, "step": 28 }, { "epoch": 0.018554062699936022, "grad_norm": 7.396708392444454, "learning_rate": 1.236673773987207e-06, "loss": 1.6141, "step": 29 }, { "epoch": 0.019193857965451054, "grad_norm": 7.331102353090355, "learning_rate": 1.279317697228145e-06, "loss": 1.6743, "step": 30 }, { "epoch": 0.01983365323096609, "grad_norm": 7.727502503383438, "learning_rate": 1.3219616204690834e-06, "loss": 1.7246, "step": 31 }, { "epoch": 0.020473448496481125, "grad_norm": 8.015060115381797, "learning_rate": 1.3646055437100215e-06, "loss": 1.7254, "step": 32 }, { "epoch": 0.02111324376199616, "grad_norm": 7.3733868681138715, "learning_rate": 1.4072494669509596e-06, "loss": 1.6786, "step": 33 }, { "epoch": 0.021753039027511197, "grad_norm": 6.977327042585721, "learning_rate": 1.4498933901918977e-06, "loss": 1.59, "step": 34 }, { "epoch": 0.022392834293026232, "grad_norm": 5.679040674087713, "learning_rate": 1.4925373134328358e-06, "loss": 1.5994, "step": 35 }, { "epoch": 0.023032629558541268, "grad_norm": 6.084846207417707, "learning_rate": 1.5351812366737743e-06, "loss": 1.547, "step": 36 }, { "epoch": 0.023672424824056303, "grad_norm": 5.1069352303559805, "learning_rate": 1.5778251599147124e-06, "loss": 1.5763, "step": 37 }, { "epoch": 0.02431222008957134, "grad_norm": 4.326057880644041, "learning_rate": 1.6204690831556505e-06, "loss": 1.6331, "step": 38 }, { "epoch": 0.02495201535508637, "grad_norm": 5.957922348312644, "learning_rate": 1.6631130063965886e-06, "loss": 1.6008, "step": 39 }, { "epoch": 0.025591810620601407, "grad_norm": 6.2519636833654335, "learning_rate": 1.7057569296375267e-06, "loss": 1.5461, "step": 40 }, { "epoch": 0.026231605886116442, "grad_norm": 6.676091063484218, "learning_rate": 1.7484008528784648e-06, "loss": 1.5988, "step": 41 }, { "epoch": 0.026871401151631478, "grad_norm": 7.344115668058566, "learning_rate": 1.791044776119403e-06, "loss": 1.4664, "step": 42 }, { "epoch": 0.027511196417146513, "grad_norm": 7.394107720970402, "learning_rate": 1.8336886993603415e-06, "loss": 1.5124, "step": 43 }, { "epoch": 0.02815099168266155, "grad_norm": 6.1105771344867446, "learning_rate": 1.8763326226012796e-06, "loss": 1.5049, "step": 44 }, { "epoch": 0.028790786948176585, "grad_norm": 6.203066406412429, "learning_rate": 1.9189765458422177e-06, "loss": 1.5393, "step": 45 }, { "epoch": 0.02943058221369162, "grad_norm": 6.156884269993957, "learning_rate": 1.9616204690831558e-06, "loss": 1.5739, "step": 46 }, { "epoch": 0.030070377479206652, "grad_norm": 4.963929532779532, "learning_rate": 2.004264392324094e-06, "loss": 1.4564, "step": 47 }, { "epoch": 0.030710172744721688, "grad_norm": 4.20597689713592, "learning_rate": 2.046908315565032e-06, "loss": 1.4053, "step": 48 }, { "epoch": 0.03134996801023673, "grad_norm": 3.6864821160119168, "learning_rate": 2.08955223880597e-06, "loss": 1.4483, "step": 49 }, { "epoch": 0.03198976327575176, "grad_norm": 3.585609883732322, "learning_rate": 2.132196162046908e-06, "loss": 1.4829, "step": 50 }, { "epoch": 0.03262955854126679, "grad_norm": 3.7333920378927354, "learning_rate": 2.1748400852878467e-06, "loss": 1.4463, "step": 51 }, { "epoch": 0.03326935380678183, "grad_norm": 3.9624041220914097, "learning_rate": 2.217484008528785e-06, "loss": 1.5092, "step": 52 }, { "epoch": 0.03390914907229686, "grad_norm": 3.98345084094273, "learning_rate": 2.260127931769723e-06, "loss": 1.4592, "step": 53 }, { "epoch": 0.0345489443378119, "grad_norm": 4.478493074114455, "learning_rate": 2.302771855010661e-06, "loss": 1.4184, "step": 54 }, { "epoch": 0.035188739603326934, "grad_norm": 5.195639051999769, "learning_rate": 2.345415778251599e-06, "loss": 1.4216, "step": 55 }, { "epoch": 0.03582853486884197, "grad_norm": 4.265247600640221, "learning_rate": 2.3880597014925373e-06, "loss": 1.382, "step": 56 }, { "epoch": 0.036468330134357005, "grad_norm": 3.9136381832311007, "learning_rate": 2.4307036247334758e-06, "loss": 1.4392, "step": 57 }, { "epoch": 0.037108125399872044, "grad_norm": 3.869474788766536, "learning_rate": 2.473347547974414e-06, "loss": 1.4185, "step": 58 }, { "epoch": 0.037747920665387076, "grad_norm": 3.4468536269240757, "learning_rate": 2.515991471215352e-06, "loss": 1.4331, "step": 59 }, { "epoch": 0.03838771593090211, "grad_norm": 3.1854951373551113, "learning_rate": 2.55863539445629e-06, "loss": 1.3636, "step": 60 }, { "epoch": 0.03902751119641715, "grad_norm": 3.6482579295188526, "learning_rate": 2.601279317697228e-06, "loss": 1.3728, "step": 61 }, { "epoch": 0.03966730646193218, "grad_norm": 3.1562286090188305, "learning_rate": 2.6439232409381667e-06, "loss": 1.406, "step": 62 }, { "epoch": 0.04030710172744722, "grad_norm": 3.673108964272774, "learning_rate": 2.686567164179105e-06, "loss": 1.3953, "step": 63 }, { "epoch": 0.04094689699296225, "grad_norm": 3.181114056097021, "learning_rate": 2.729211087420043e-06, "loss": 1.3943, "step": 64 }, { "epoch": 0.04158669225847729, "grad_norm": 3.0081996373121713, "learning_rate": 2.771855010660981e-06, "loss": 1.3782, "step": 65 }, { "epoch": 0.04222648752399232, "grad_norm": 3.3511848186757227, "learning_rate": 2.814498933901919e-06, "loss": 1.3791, "step": 66 }, { "epoch": 0.04286628278950736, "grad_norm": 3.6062171208897293, "learning_rate": 2.8571428571428573e-06, "loss": 1.3928, "step": 67 }, { "epoch": 0.04350607805502239, "grad_norm": 3.6829254765419988, "learning_rate": 2.8997867803837954e-06, "loss": 1.3514, "step": 68 }, { "epoch": 0.044145873320537425, "grad_norm": 3.342303004843198, "learning_rate": 2.9424307036247335e-06, "loss": 1.2891, "step": 69 }, { "epoch": 0.044785668586052464, "grad_norm": 2.9891266743989826, "learning_rate": 2.9850746268656716e-06, "loss": 1.3516, "step": 70 }, { "epoch": 0.045425463851567496, "grad_norm": 2.972348104100991, "learning_rate": 3.0277185501066105e-06, "loss": 1.3734, "step": 71 }, { "epoch": 0.046065259117082535, "grad_norm": 3.0621189771357478, "learning_rate": 3.0703624733475486e-06, "loss": 1.3632, "step": 72 }, { "epoch": 0.04670505438259757, "grad_norm": 3.0415055022225777, "learning_rate": 3.1130063965884867e-06, "loss": 1.3493, "step": 73 }, { "epoch": 0.04734484964811261, "grad_norm": 3.3569361807441336, "learning_rate": 3.155650319829425e-06, "loss": 1.2973, "step": 74 }, { "epoch": 0.04798464491362764, "grad_norm": 2.9671759958101367, "learning_rate": 3.198294243070363e-06, "loss": 1.3186, "step": 75 }, { "epoch": 0.04862444017914268, "grad_norm": 2.960009656652622, "learning_rate": 3.240938166311301e-06, "loss": 1.3275, "step": 76 }, { "epoch": 0.04926423544465771, "grad_norm": 3.0848005579425957, "learning_rate": 3.283582089552239e-06, "loss": 1.3399, "step": 77 }, { "epoch": 0.04990403071017274, "grad_norm": 3.1164930213288775, "learning_rate": 3.3262260127931773e-06, "loss": 1.2766, "step": 78 }, { "epoch": 0.05054382597568778, "grad_norm": 3.332402317298742, "learning_rate": 3.3688699360341154e-06, "loss": 1.2933, "step": 79 }, { "epoch": 0.05118362124120281, "grad_norm": 3.3430242101950154, "learning_rate": 3.4115138592750535e-06, "loss": 1.3156, "step": 80 }, { "epoch": 0.05182341650671785, "grad_norm": 3.1392639440004855, "learning_rate": 3.4541577825159916e-06, "loss": 1.3977, "step": 81 }, { "epoch": 0.052463211772232884, "grad_norm": 2.981423034192431, "learning_rate": 3.4968017057569297e-06, "loss": 1.2853, "step": 82 }, { "epoch": 0.053103007037747924, "grad_norm": 3.032017326680704, "learning_rate": 3.5394456289978678e-06, "loss": 1.3368, "step": 83 }, { "epoch": 0.053742802303262956, "grad_norm": 3.065142162098067, "learning_rate": 3.582089552238806e-06, "loss": 1.2942, "step": 84 }, { "epoch": 0.05438259756877799, "grad_norm": 3.0181798287898283, "learning_rate": 3.624733475479744e-06, "loss": 1.3675, "step": 85 }, { "epoch": 0.05502239283429303, "grad_norm": 3.3529300714180312, "learning_rate": 3.667377398720683e-06, "loss": 1.2556, "step": 86 }, { "epoch": 0.05566218809980806, "grad_norm": 3.1662470022589804, "learning_rate": 3.710021321961621e-06, "loss": 1.3344, "step": 87 }, { "epoch": 0.0563019833653231, "grad_norm": 2.685163311156544, "learning_rate": 3.752665245202559e-06, "loss": 1.307, "step": 88 }, { "epoch": 0.05694177863083813, "grad_norm": 2.641701628320942, "learning_rate": 3.7953091684434973e-06, "loss": 1.2878, "step": 89 }, { "epoch": 0.05758157389635317, "grad_norm": 3.0878704303758466, "learning_rate": 3.837953091684435e-06, "loss": 1.2409, "step": 90 }, { "epoch": 0.0582213691618682, "grad_norm": 2.8806761908635132, "learning_rate": 3.8805970149253735e-06, "loss": 1.3399, "step": 91 }, { "epoch": 0.05886116442738324, "grad_norm": 3.2835807116286895, "learning_rate": 3.9232409381663116e-06, "loss": 1.2997, "step": 92 }, { "epoch": 0.05950095969289827, "grad_norm": 2.856992797794588, "learning_rate": 3.96588486140725e-06, "loss": 1.2843, "step": 93 }, { "epoch": 0.060140754958413305, "grad_norm": 3.1927088727692308, "learning_rate": 4.008528784648188e-06, "loss": 1.2481, "step": 94 }, { "epoch": 0.060780550223928344, "grad_norm": 2.8258689207792496, "learning_rate": 4.051172707889126e-06, "loss": 1.2446, "step": 95 }, { "epoch": 0.061420345489443376, "grad_norm": 3.159461351742752, "learning_rate": 4.093816631130064e-06, "loss": 1.2056, "step": 96 }, { "epoch": 0.062060140754958415, "grad_norm": 3.0188925550295567, "learning_rate": 4.136460554371002e-06, "loss": 1.2609, "step": 97 }, { "epoch": 0.06269993602047345, "grad_norm": 3.1161169260935604, "learning_rate": 4.17910447761194e-06, "loss": 1.2137, "step": 98 }, { "epoch": 0.06333973128598848, "grad_norm": 3.099478986749759, "learning_rate": 4.221748400852878e-06, "loss": 1.2432, "step": 99 }, { "epoch": 0.06397952655150352, "grad_norm": 2.9983993380954854, "learning_rate": 4.264392324093816e-06, "loss": 1.3345, "step": 100 }, { "epoch": 0.06461932181701856, "grad_norm": 2.7906120823044978, "learning_rate": 4.3070362473347545e-06, "loss": 1.2469, "step": 101 }, { "epoch": 0.06525911708253358, "grad_norm": 2.9207578516104427, "learning_rate": 4.3496801705756935e-06, "loss": 1.258, "step": 102 }, { "epoch": 0.06589891234804862, "grad_norm": 3.138673322047274, "learning_rate": 4.3923240938166316e-06, "loss": 1.2403, "step": 103 }, { "epoch": 0.06653870761356366, "grad_norm": 2.8595411478101522, "learning_rate": 4.43496801705757e-06, "loss": 1.2444, "step": 104 }, { "epoch": 0.0671785028790787, "grad_norm": 3.336902329957571, "learning_rate": 4.477611940298508e-06, "loss": 1.2151, "step": 105 }, { "epoch": 0.06781829814459372, "grad_norm": 2.8270064728951367, "learning_rate": 4.520255863539446e-06, "loss": 1.2886, "step": 106 }, { "epoch": 0.06845809341010876, "grad_norm": 2.8604971999488504, "learning_rate": 4.562899786780384e-06, "loss": 1.2853, "step": 107 }, { "epoch": 0.0690978886756238, "grad_norm": 2.968060726520348, "learning_rate": 4.605543710021322e-06, "loss": 1.2481, "step": 108 }, { "epoch": 0.06973768394113884, "grad_norm": 3.3531467019727774, "learning_rate": 4.64818763326226e-06, "loss": 1.2271, "step": 109 }, { "epoch": 0.07037747920665387, "grad_norm": 3.3467942158455557, "learning_rate": 4.690831556503198e-06, "loss": 1.2348, "step": 110 }, { "epoch": 0.0710172744721689, "grad_norm": 3.2283478144322264, "learning_rate": 4.733475479744136e-06, "loss": 1.2885, "step": 111 }, { "epoch": 0.07165706973768395, "grad_norm": 3.139750250990695, "learning_rate": 4.7761194029850745e-06, "loss": 1.2052, "step": 112 }, { "epoch": 0.07229686500319897, "grad_norm": 2.830276410846628, "learning_rate": 4.8187633262260135e-06, "loss": 1.227, "step": 113 }, { "epoch": 0.07293666026871401, "grad_norm": 3.0544821048367905, "learning_rate": 4.8614072494669516e-06, "loss": 1.2118, "step": 114 }, { "epoch": 0.07357645553422905, "grad_norm": 3.094769553180934, "learning_rate": 4.90405117270789e-06, "loss": 1.2459, "step": 115 }, { "epoch": 0.07421625079974409, "grad_norm": 3.0605265089293896, "learning_rate": 4.946695095948828e-06, "loss": 1.2002, "step": 116 }, { "epoch": 0.07485604606525911, "grad_norm": 3.3704733425069797, "learning_rate": 4.989339019189766e-06, "loss": 1.2387, "step": 117 }, { "epoch": 0.07549584133077415, "grad_norm": 2.9151373582593667, "learning_rate": 5.031982942430704e-06, "loss": 1.2082, "step": 118 }, { "epoch": 0.07613563659628919, "grad_norm": 3.6427501374624294, "learning_rate": 5.074626865671642e-06, "loss": 1.2422, "step": 119 }, { "epoch": 0.07677543186180422, "grad_norm": 3.5668169265785528, "learning_rate": 5.11727078891258e-06, "loss": 1.2511, "step": 120 }, { "epoch": 0.07741522712731926, "grad_norm": 3.520397378253227, "learning_rate": 5.159914712153518e-06, "loss": 1.228, "step": 121 }, { "epoch": 0.0780550223928343, "grad_norm": 3.167689789742573, "learning_rate": 5.202558635394456e-06, "loss": 1.2432, "step": 122 }, { "epoch": 0.07869481765834933, "grad_norm": 3.007071871402184, "learning_rate": 5.245202558635395e-06, "loss": 1.209, "step": 123 }, { "epoch": 0.07933461292386436, "grad_norm": 2.8731611907299124, "learning_rate": 5.2878464818763335e-06, "loss": 1.2273, "step": 124 }, { "epoch": 0.0799744081893794, "grad_norm": 3.384669834985903, "learning_rate": 5.3304904051172716e-06, "loss": 1.2378, "step": 125 }, { "epoch": 0.08061420345489444, "grad_norm": 3.2852074041656474, "learning_rate": 5.37313432835821e-06, "loss": 1.2108, "step": 126 }, { "epoch": 0.08125399872040946, "grad_norm": 2.7484562866857676, "learning_rate": 5.415778251599148e-06, "loss": 1.2716, "step": 127 }, { "epoch": 0.0818937939859245, "grad_norm": 2.9741227736785754, "learning_rate": 5.458422174840086e-06, "loss": 1.1934, "step": 128 }, { "epoch": 0.08253358925143954, "grad_norm": 3.049070014127921, "learning_rate": 5.501066098081024e-06, "loss": 1.1768, "step": 129 }, { "epoch": 0.08317338451695458, "grad_norm": 2.735755172160257, "learning_rate": 5.543710021321962e-06, "loss": 1.2313, "step": 130 }, { "epoch": 0.0838131797824696, "grad_norm": 3.089750414346629, "learning_rate": 5.5863539445629e-06, "loss": 1.2293, "step": 131 }, { "epoch": 0.08445297504798464, "grad_norm": 3.0396627691229843, "learning_rate": 5.628997867803838e-06, "loss": 1.2633, "step": 132 }, { "epoch": 0.08509277031349968, "grad_norm": 3.199361848516151, "learning_rate": 5.671641791044776e-06, "loss": 1.2219, "step": 133 }, { "epoch": 0.08573256557901472, "grad_norm": 3.287473368802135, "learning_rate": 5.7142857142857145e-06, "loss": 1.2321, "step": 134 }, { "epoch": 0.08637236084452975, "grad_norm": 3.1340710107301493, "learning_rate": 5.756929637526653e-06, "loss": 1.142, "step": 135 }, { "epoch": 0.08701215611004479, "grad_norm": 3.0075392967337287, "learning_rate": 5.799573560767591e-06, "loss": 1.2074, "step": 136 }, { "epoch": 0.08765195137555983, "grad_norm": 3.383747556160692, "learning_rate": 5.842217484008529e-06, "loss": 1.1806, "step": 137 }, { "epoch": 0.08829174664107485, "grad_norm": 3.382924161065641, "learning_rate": 5.884861407249467e-06, "loss": 1.233, "step": 138 }, { "epoch": 0.08893154190658989, "grad_norm": 3.1899668834387276, "learning_rate": 5.927505330490405e-06, "loss": 1.2575, "step": 139 }, { "epoch": 0.08957133717210493, "grad_norm": 2.8793092577328094, "learning_rate": 5.970149253731343e-06, "loss": 1.2028, "step": 140 }, { "epoch": 0.09021113243761997, "grad_norm": 3.6436788772883175, "learning_rate": 6.012793176972282e-06, "loss": 1.2507, "step": 141 }, { "epoch": 0.09085092770313499, "grad_norm": 3.6262588921567236, "learning_rate": 6.055437100213221e-06, "loss": 1.2108, "step": 142 }, { "epoch": 0.09149072296865003, "grad_norm": 3.1473517799897763, "learning_rate": 6.098081023454159e-06, "loss": 1.2449, "step": 143 }, { "epoch": 0.09213051823416507, "grad_norm": 3.307945007363661, "learning_rate": 6.140724946695097e-06, "loss": 1.2516, "step": 144 }, { "epoch": 0.0927703134996801, "grad_norm": 3.0289539775685372, "learning_rate": 6.183368869936035e-06, "loss": 1.2053, "step": 145 }, { "epoch": 0.09341010876519514, "grad_norm": 3.276246965621132, "learning_rate": 6.2260127931769735e-06, "loss": 1.1388, "step": 146 }, { "epoch": 0.09404990403071017, "grad_norm": 3.3310477237439, "learning_rate": 6.2686567164179116e-06, "loss": 1.1301, "step": 147 }, { "epoch": 0.09468969929622521, "grad_norm": 3.1778585251272373, "learning_rate": 6.31130063965885e-06, "loss": 1.2527, "step": 148 }, { "epoch": 0.09532949456174024, "grad_norm": 3.0216287771014225, "learning_rate": 6.353944562899788e-06, "loss": 1.1331, "step": 149 }, { "epoch": 0.09596928982725528, "grad_norm": 3.049847133116838, "learning_rate": 6.396588486140726e-06, "loss": 1.2204, "step": 150 }, { "epoch": 0.09660908509277032, "grad_norm": 3.1623557988782736, "learning_rate": 6.439232409381664e-06, "loss": 1.2022, "step": 151 }, { "epoch": 0.09724888035828536, "grad_norm": 3.2774929620557436, "learning_rate": 6.481876332622602e-06, "loss": 1.197, "step": 152 }, { "epoch": 0.09788867562380038, "grad_norm": 3.344458343364583, "learning_rate": 6.52452025586354e-06, "loss": 1.2169, "step": 153 }, { "epoch": 0.09852847088931542, "grad_norm": 3.4024549133121855, "learning_rate": 6.567164179104478e-06, "loss": 1.1229, "step": 154 }, { "epoch": 0.09916826615483046, "grad_norm": 3.1855233274893084, "learning_rate": 6.609808102345416e-06, "loss": 1.1983, "step": 155 }, { "epoch": 0.09980806142034548, "grad_norm": 3.4427309543239653, "learning_rate": 6.6524520255863545e-06, "loss": 1.2292, "step": 156 }, { "epoch": 0.10044785668586052, "grad_norm": 3.411140909129944, "learning_rate": 6.695095948827293e-06, "loss": 1.2395, "step": 157 }, { "epoch": 0.10108765195137556, "grad_norm": 3.0239546515975286, "learning_rate": 6.737739872068231e-06, "loss": 1.1764, "step": 158 }, { "epoch": 0.1017274472168906, "grad_norm": 3.2350834521679777, "learning_rate": 6.780383795309169e-06, "loss": 1.2065, "step": 159 }, { "epoch": 0.10236724248240563, "grad_norm": 3.1036162695648963, "learning_rate": 6.823027718550107e-06, "loss": 1.2553, "step": 160 }, { "epoch": 0.10300703774792067, "grad_norm": 3.152283577583286, "learning_rate": 6.865671641791045e-06, "loss": 1.1697, "step": 161 }, { "epoch": 0.1036468330134357, "grad_norm": 3.7079305269817153, "learning_rate": 6.908315565031983e-06, "loss": 1.1167, "step": 162 }, { "epoch": 0.10428662827895073, "grad_norm": 3.5547152318454467, "learning_rate": 6.950959488272921e-06, "loss": 1.2456, "step": 163 }, { "epoch": 0.10492642354446577, "grad_norm": 3.06898549707911, "learning_rate": 6.993603411513859e-06, "loss": 1.2426, "step": 164 }, { "epoch": 0.10556621880998081, "grad_norm": 3.366438562667226, "learning_rate": 7.0362473347547975e-06, "loss": 1.1509, "step": 165 }, { "epoch": 0.10620601407549585, "grad_norm": 3.458152715220668, "learning_rate": 7.0788912579957356e-06, "loss": 1.1782, "step": 166 }, { "epoch": 0.10684580934101087, "grad_norm": 3.209922314341663, "learning_rate": 7.121535181236674e-06, "loss": 1.1544, "step": 167 }, { "epoch": 0.10748560460652591, "grad_norm": 3.0009439973338234, "learning_rate": 7.164179104477612e-06, "loss": 1.2137, "step": 168 }, { "epoch": 0.10812539987204095, "grad_norm": 3.117627609561351, "learning_rate": 7.20682302771855e-06, "loss": 1.2331, "step": 169 }, { "epoch": 0.10876519513755598, "grad_norm": 3.577808454265829, "learning_rate": 7.249466950959488e-06, "loss": 1.2267, "step": 170 }, { "epoch": 0.10940499040307101, "grad_norm": 3.289445435594237, "learning_rate": 7.292110874200427e-06, "loss": 1.184, "step": 171 }, { "epoch": 0.11004478566858605, "grad_norm": 2.907662264327517, "learning_rate": 7.334754797441366e-06, "loss": 1.1779, "step": 172 }, { "epoch": 0.11068458093410109, "grad_norm": 3.2026279149094905, "learning_rate": 7.377398720682304e-06, "loss": 1.1466, "step": 173 }, { "epoch": 0.11132437619961612, "grad_norm": 3.4369780763523274, "learning_rate": 7.420042643923242e-06, "loss": 1.1802, "step": 174 }, { "epoch": 0.11196417146513116, "grad_norm": 3.0796710965513463, "learning_rate": 7.46268656716418e-06, "loss": 1.21, "step": 175 }, { "epoch": 0.1126039667306462, "grad_norm": 3.2207395864573893, "learning_rate": 7.505330490405118e-06, "loss": 1.1377, "step": 176 }, { "epoch": 0.11324376199616124, "grad_norm": 3.3855131909475684, "learning_rate": 7.547974413646056e-06, "loss": 1.1903, "step": 177 }, { "epoch": 0.11388355726167626, "grad_norm": 3.687952656213279, "learning_rate": 7.5906183368869945e-06, "loss": 1.1581, "step": 178 }, { "epoch": 0.1145233525271913, "grad_norm": 3.226192347950398, "learning_rate": 7.633262260127933e-06, "loss": 1.1882, "step": 179 }, { "epoch": 0.11516314779270634, "grad_norm": 3.874537453978932, "learning_rate": 7.67590618336887e-06, "loss": 1.26, "step": 180 }, { "epoch": 0.11580294305822136, "grad_norm": 3.6364508047900075, "learning_rate": 7.718550106609809e-06, "loss": 1.1847, "step": 181 }, { "epoch": 0.1164427383237364, "grad_norm": 2.9359648459220713, "learning_rate": 7.761194029850747e-06, "loss": 1.1745, "step": 182 }, { "epoch": 0.11708253358925144, "grad_norm": 3.004400666652676, "learning_rate": 7.803837953091685e-06, "loss": 1.0837, "step": 183 }, { "epoch": 0.11772232885476648, "grad_norm": 3.216152901800306, "learning_rate": 7.846481876332623e-06, "loss": 1.1507, "step": 184 }, { "epoch": 0.1183621241202815, "grad_norm": 3.897438969572001, "learning_rate": 7.889125799573561e-06, "loss": 1.2199, "step": 185 }, { "epoch": 0.11900191938579655, "grad_norm": 3.7109098300971257, "learning_rate": 7.9317697228145e-06, "loss": 1.208, "step": 186 }, { "epoch": 0.11964171465131158, "grad_norm": 3.5499654457127696, "learning_rate": 7.974413646055437e-06, "loss": 1.1782, "step": 187 }, { "epoch": 0.12028150991682661, "grad_norm": 3.112590263830222, "learning_rate": 8.017057569296376e-06, "loss": 1.0883, "step": 188 }, { "epoch": 0.12092130518234165, "grad_norm": 3.4995291458352233, "learning_rate": 8.059701492537314e-06, "loss": 1.1382, "step": 189 }, { "epoch": 0.12156110044785669, "grad_norm": 3.4410116383291953, "learning_rate": 8.102345415778252e-06, "loss": 1.1404, "step": 190 }, { "epoch": 0.12220089571337173, "grad_norm": 3.3399313785719893, "learning_rate": 8.14498933901919e-06, "loss": 1.2012, "step": 191 }, { "epoch": 0.12284069097888675, "grad_norm": 3.4938618991620505, "learning_rate": 8.187633262260128e-06, "loss": 1.1947, "step": 192 }, { "epoch": 0.12348048624440179, "grad_norm": 3.515366301522175, "learning_rate": 8.230277185501066e-06, "loss": 1.1649, "step": 193 }, { "epoch": 0.12412028150991683, "grad_norm": 3.2486390606211315, "learning_rate": 8.272921108742004e-06, "loss": 1.1674, "step": 194 }, { "epoch": 0.12476007677543186, "grad_norm": 2.975608455592541, "learning_rate": 8.315565031982942e-06, "loss": 1.1467, "step": 195 }, { "epoch": 0.1253998720409469, "grad_norm": 3.3867784482661976, "learning_rate": 8.35820895522388e-06, "loss": 1.1859, "step": 196 }, { "epoch": 0.12603966730646193, "grad_norm": 3.9907038455951542, "learning_rate": 8.400852878464819e-06, "loss": 1.2112, "step": 197 }, { "epoch": 0.12667946257197696, "grad_norm": 3.494174807892128, "learning_rate": 8.443496801705757e-06, "loss": 1.1741, "step": 198 }, { "epoch": 0.127319257837492, "grad_norm": 3.2466592251263737, "learning_rate": 8.486140724946695e-06, "loss": 1.1583, "step": 199 }, { "epoch": 0.12795905310300704, "grad_norm": 3.228816203255189, "learning_rate": 8.528784648187633e-06, "loss": 1.174, "step": 200 }, { "epoch": 0.12859884836852206, "grad_norm": 3.2808316664056876, "learning_rate": 8.571428571428571e-06, "loss": 1.2359, "step": 201 }, { "epoch": 0.12923864363403711, "grad_norm": 3.22074317172062, "learning_rate": 8.614072494669509e-06, "loss": 1.1554, "step": 202 }, { "epoch": 0.12987843889955214, "grad_norm": 3.469604446716559, "learning_rate": 8.656716417910447e-06, "loss": 1.1333, "step": 203 }, { "epoch": 0.13051823416506717, "grad_norm": 3.1317347439579555, "learning_rate": 8.699360341151387e-06, "loss": 1.1801, "step": 204 }, { "epoch": 0.13115802943058222, "grad_norm": 3.49525882287199, "learning_rate": 8.742004264392325e-06, "loss": 1.1778, "step": 205 }, { "epoch": 0.13179782469609724, "grad_norm": 3.0169300883209127, "learning_rate": 8.784648187633263e-06, "loss": 1.14, "step": 206 }, { "epoch": 0.1324376199616123, "grad_norm": 3.070015193563139, "learning_rate": 8.827292110874201e-06, "loss": 1.1265, "step": 207 }, { "epoch": 0.13307741522712732, "grad_norm": 3.4537670458825853, "learning_rate": 8.86993603411514e-06, "loss": 1.146, "step": 208 }, { "epoch": 0.13371721049264235, "grad_norm": 3.243975152164699, "learning_rate": 8.912579957356077e-06, "loss": 1.1931, "step": 209 }, { "epoch": 0.1343570057581574, "grad_norm": 2.948391601026555, "learning_rate": 8.955223880597016e-06, "loss": 1.1237, "step": 210 }, { "epoch": 0.13499680102367242, "grad_norm": 3.201989245378149, "learning_rate": 8.997867803837954e-06, "loss": 1.1766, "step": 211 }, { "epoch": 0.13563659628918745, "grad_norm": 3.565345683744908, "learning_rate": 9.040511727078892e-06, "loss": 1.1638, "step": 212 }, { "epoch": 0.1362763915547025, "grad_norm": 3.3441250990643603, "learning_rate": 9.08315565031983e-06, "loss": 1.1455, "step": 213 }, { "epoch": 0.13691618682021753, "grad_norm": 3.2527573219557335, "learning_rate": 9.125799573560768e-06, "loss": 1.1335, "step": 214 }, { "epoch": 0.13755598208573255, "grad_norm": 3.5015582743322384, "learning_rate": 9.168443496801706e-06, "loss": 1.1654, "step": 215 }, { "epoch": 0.1381957773512476, "grad_norm": 3.8978020869768826, "learning_rate": 9.211087420042644e-06, "loss": 1.1488, "step": 216 }, { "epoch": 0.13883557261676263, "grad_norm": 3.6513509263672677, "learning_rate": 9.253731343283582e-06, "loss": 1.1896, "step": 217 }, { "epoch": 0.13947536788227768, "grad_norm": 3.3557121247336195, "learning_rate": 9.29637526652452e-06, "loss": 1.1219, "step": 218 }, { "epoch": 0.1401151631477927, "grad_norm": 3.4218814710196406, "learning_rate": 9.339019189765458e-06, "loss": 1.1598, "step": 219 }, { "epoch": 0.14075495841330773, "grad_norm": 4.255586952073893, "learning_rate": 9.381663113006397e-06, "loss": 1.2008, "step": 220 }, { "epoch": 0.1413947536788228, "grad_norm": 3.399292873068693, "learning_rate": 9.424307036247335e-06, "loss": 1.0875, "step": 221 }, { "epoch": 0.1420345489443378, "grad_norm": 3.3555221135687874, "learning_rate": 9.466950959488273e-06, "loss": 1.1562, "step": 222 }, { "epoch": 0.14267434420985284, "grad_norm": 3.750531546882094, "learning_rate": 9.509594882729211e-06, "loss": 1.1301, "step": 223 }, { "epoch": 0.1433141394753679, "grad_norm": 3.3836390695054446, "learning_rate": 9.552238805970149e-06, "loss": 1.1341, "step": 224 }, { "epoch": 0.14395393474088292, "grad_norm": 3.361743628118061, "learning_rate": 9.594882729211089e-06, "loss": 1.162, "step": 225 }, { "epoch": 0.14459373000639794, "grad_norm": 3.2773130680078397, "learning_rate": 9.637526652452027e-06, "loss": 1.1445, "step": 226 }, { "epoch": 0.145233525271913, "grad_norm": 3.4004701512238356, "learning_rate": 9.680170575692965e-06, "loss": 1.1369, "step": 227 }, { "epoch": 0.14587332053742802, "grad_norm": 3.4124939594051606, "learning_rate": 9.722814498933903e-06, "loss": 1.1899, "step": 228 }, { "epoch": 0.14651311580294304, "grad_norm": 3.1450518311723856, "learning_rate": 9.765458422174841e-06, "loss": 1.0955, "step": 229 }, { "epoch": 0.1471529110684581, "grad_norm": 2.9539706856332058, "learning_rate": 9.80810234541578e-06, "loss": 1.1282, "step": 230 }, { "epoch": 0.14779270633397312, "grad_norm": 3.378036572769568, "learning_rate": 9.850746268656717e-06, "loss": 1.1692, "step": 231 }, { "epoch": 0.14843250159948818, "grad_norm": 3.75232015061377, "learning_rate": 9.893390191897656e-06, "loss": 1.1829, "step": 232 }, { "epoch": 0.1490722968650032, "grad_norm": 4.13000196836793, "learning_rate": 9.936034115138594e-06, "loss": 1.1824, "step": 233 }, { "epoch": 0.14971209213051823, "grad_norm": 2.9979050753855017, "learning_rate": 9.978678038379532e-06, "loss": 1.1952, "step": 234 }, { "epoch": 0.15035188739603328, "grad_norm": 3.0402242966990682, "learning_rate": 1.002132196162047e-05, "loss": 1.1112, "step": 235 }, { "epoch": 0.1509916826615483, "grad_norm": 3.3554215931005453, "learning_rate": 1.0063965884861408e-05, "loss": 1.124, "step": 236 }, { "epoch": 0.15163147792706333, "grad_norm": 3.098944395460217, "learning_rate": 1.0106609808102346e-05, "loss": 1.108, "step": 237 }, { "epoch": 0.15227127319257838, "grad_norm": 3.4577555819340433, "learning_rate": 1.0149253731343284e-05, "loss": 1.092, "step": 238 }, { "epoch": 0.1529110684580934, "grad_norm": 3.6093027868989154, "learning_rate": 1.0191897654584222e-05, "loss": 1.2091, "step": 239 }, { "epoch": 0.15355086372360843, "grad_norm": 3.473987080129039, "learning_rate": 1.023454157782516e-05, "loss": 1.1471, "step": 240 }, { "epoch": 0.15419065898912349, "grad_norm": 3.1786761940721573, "learning_rate": 1.0277185501066098e-05, "loss": 1.142, "step": 241 }, { "epoch": 0.1548304542546385, "grad_norm": 3.3630786617374056, "learning_rate": 1.0319829424307037e-05, "loss": 1.1541, "step": 242 }, { "epoch": 0.15547024952015356, "grad_norm": 3.2246483852970105, "learning_rate": 1.0362473347547975e-05, "loss": 1.1866, "step": 243 }, { "epoch": 0.1561100447856686, "grad_norm": 3.9389824824648474, "learning_rate": 1.0405117270788913e-05, "loss": 1.0936, "step": 244 }, { "epoch": 0.15674984005118361, "grad_norm": 3.5372292681021364, "learning_rate": 1.0447761194029851e-05, "loss": 1.1256, "step": 245 }, { "epoch": 0.15738963531669867, "grad_norm": 3.134098913123201, "learning_rate": 1.049040511727079e-05, "loss": 1.1211, "step": 246 }, { "epoch": 0.1580294305822137, "grad_norm": 3.969673029607486, "learning_rate": 1.0533049040511727e-05, "loss": 1.1204, "step": 247 }, { "epoch": 0.15866922584772872, "grad_norm": 3.235193308305361, "learning_rate": 1.0575692963752667e-05, "loss": 1.1994, "step": 248 }, { "epoch": 0.15930902111324377, "grad_norm": 3.2214228183738673, "learning_rate": 1.0618336886993603e-05, "loss": 1.116, "step": 249 }, { "epoch": 0.1599488163787588, "grad_norm": 3.5506951489873964, "learning_rate": 1.0660980810234543e-05, "loss": 1.1205, "step": 250 }, { "epoch": 0.16058861164427382, "grad_norm": 3.044258266435914, "learning_rate": 1.0703624733475481e-05, "loss": 1.1024, "step": 251 }, { "epoch": 0.16122840690978887, "grad_norm": 3.62150448088908, "learning_rate": 1.074626865671642e-05, "loss": 1.1469, "step": 252 }, { "epoch": 0.1618682021753039, "grad_norm": 3.462536056930194, "learning_rate": 1.0788912579957357e-05, "loss": 1.1324, "step": 253 }, { "epoch": 0.16250799744081892, "grad_norm": 3.1905916671905095, "learning_rate": 1.0831556503198296e-05, "loss": 1.1709, "step": 254 }, { "epoch": 0.16314779270633398, "grad_norm": 3.5495436634553843, "learning_rate": 1.0874200426439234e-05, "loss": 1.1495, "step": 255 }, { "epoch": 0.163787587971849, "grad_norm": 3.5320417938535558, "learning_rate": 1.0916844349680172e-05, "loss": 1.0863, "step": 256 }, { "epoch": 0.16442738323736406, "grad_norm": 3.202251391490625, "learning_rate": 1.095948827292111e-05, "loss": 1.0929, "step": 257 }, { "epoch": 0.16506717850287908, "grad_norm": 3.6319638848728397, "learning_rate": 1.1002132196162048e-05, "loss": 1.1195, "step": 258 }, { "epoch": 0.1657069737683941, "grad_norm": 3.3691240062466266, "learning_rate": 1.1044776119402986e-05, "loss": 1.1666, "step": 259 }, { "epoch": 0.16634676903390916, "grad_norm": 3.4052611286585766, "learning_rate": 1.1087420042643924e-05, "loss": 1.1552, "step": 260 }, { "epoch": 0.16698656429942418, "grad_norm": 3.8948594840122492, "learning_rate": 1.1130063965884862e-05, "loss": 1.14, "step": 261 }, { "epoch": 0.1676263595649392, "grad_norm": 3.7004943759347007, "learning_rate": 1.11727078891258e-05, "loss": 1.1417, "step": 262 }, { "epoch": 0.16826615483045426, "grad_norm": 3.864493508944828, "learning_rate": 1.1215351812366738e-05, "loss": 1.1607, "step": 263 }, { "epoch": 0.1689059500959693, "grad_norm": 3.2029923083330476, "learning_rate": 1.1257995735607677e-05, "loss": 1.1472, "step": 264 }, { "epoch": 0.1695457453614843, "grad_norm": 3.38052183196855, "learning_rate": 1.1300639658848616e-05, "loss": 1.1364, "step": 265 }, { "epoch": 0.17018554062699937, "grad_norm": 3.4198475836246276, "learning_rate": 1.1343283582089553e-05, "loss": 1.1273, "step": 266 }, { "epoch": 0.1708253358925144, "grad_norm": 3.920591096108978, "learning_rate": 1.1385927505330493e-05, "loss": 1.1334, "step": 267 }, { "epoch": 0.17146513115802944, "grad_norm": 3.5075374198752605, "learning_rate": 1.1428571428571429e-05, "loss": 1.1632, "step": 268 }, { "epoch": 0.17210492642354447, "grad_norm": 3.245247394360144, "learning_rate": 1.1471215351812369e-05, "loss": 1.066, "step": 269 }, { "epoch": 0.1727447216890595, "grad_norm": 3.198559564540295, "learning_rate": 1.1513859275053305e-05, "loss": 1.1747, "step": 270 }, { "epoch": 0.17338451695457455, "grad_norm": 3.94161828008153, "learning_rate": 1.1556503198294245e-05, "loss": 1.1323, "step": 271 }, { "epoch": 0.17402431222008957, "grad_norm": 3.4867399509611627, "learning_rate": 1.1599147121535181e-05, "loss": 1.1582, "step": 272 }, { "epoch": 0.1746641074856046, "grad_norm": 3.374918469429571, "learning_rate": 1.1641791044776121e-05, "loss": 1.1533, "step": 273 }, { "epoch": 0.17530390275111965, "grad_norm": 2.930319630176569, "learning_rate": 1.1684434968017058e-05, "loss": 1.1047, "step": 274 }, { "epoch": 0.17594369801663468, "grad_norm": 3.5203056709071054, "learning_rate": 1.1727078891257997e-05, "loss": 1.1107, "step": 275 }, { "epoch": 0.1765834932821497, "grad_norm": 3.3210753229777956, "learning_rate": 1.1769722814498934e-05, "loss": 1.1882, "step": 276 }, { "epoch": 0.17722328854766475, "grad_norm": 3.270488293659176, "learning_rate": 1.1812366737739874e-05, "loss": 1.131, "step": 277 }, { "epoch": 0.17786308381317978, "grad_norm": 3.6766266838489816, "learning_rate": 1.185501066098081e-05, "loss": 1.1368, "step": 278 }, { "epoch": 0.1785028790786948, "grad_norm": 3.8658752675323216, "learning_rate": 1.189765458422175e-05, "loss": 1.1847, "step": 279 }, { "epoch": 0.17914267434420986, "grad_norm": 3.5898685278538465, "learning_rate": 1.1940298507462686e-05, "loss": 1.142, "step": 280 }, { "epoch": 0.17978246960972488, "grad_norm": 3.751748990336646, "learning_rate": 1.1982942430703626e-05, "loss": 1.1584, "step": 281 }, { "epoch": 0.18042226487523993, "grad_norm": 3.2536836935177678, "learning_rate": 1.2025586353944564e-05, "loss": 1.1599, "step": 282 }, { "epoch": 0.18106206014075496, "grad_norm": 3.478668416963261, "learning_rate": 1.2068230277185502e-05, "loss": 1.1023, "step": 283 }, { "epoch": 0.18170185540626999, "grad_norm": 4.013504847177782, "learning_rate": 1.2110874200426442e-05, "loss": 1.1231, "step": 284 }, { "epoch": 0.18234165067178504, "grad_norm": 3.513146143546194, "learning_rate": 1.2153518123667378e-05, "loss": 1.1747, "step": 285 }, { "epoch": 0.18298144593730006, "grad_norm": 3.8839132264766945, "learning_rate": 1.2196162046908318e-05, "loss": 1.0953, "step": 286 }, { "epoch": 0.1836212412028151, "grad_norm": 3.75524064515929, "learning_rate": 1.2238805970149255e-05, "loss": 1.1342, "step": 287 }, { "epoch": 0.18426103646833014, "grad_norm": 3.1691249041817, "learning_rate": 1.2281449893390195e-05, "loss": 1.0879, "step": 288 }, { "epoch": 0.18490083173384517, "grad_norm": 3.441953615492417, "learning_rate": 1.2324093816631131e-05, "loss": 1.1927, "step": 289 }, { "epoch": 0.1855406269993602, "grad_norm": 3.4245722797382125, "learning_rate": 1.236673773987207e-05, "loss": 1.1309, "step": 290 }, { "epoch": 0.18618042226487524, "grad_norm": 3.214204936626932, "learning_rate": 1.2409381663113007e-05, "loss": 1.0994, "step": 291 }, { "epoch": 0.18682021753039027, "grad_norm": 3.311386746776984, "learning_rate": 1.2452025586353947e-05, "loss": 1.12, "step": 292 }, { "epoch": 0.18746001279590532, "grad_norm": 3.790031158556305, "learning_rate": 1.2494669509594883e-05, "loss": 1.1245, "step": 293 }, { "epoch": 0.18809980806142035, "grad_norm": 3.67743215760485, "learning_rate": 1.2537313432835823e-05, "loss": 1.1748, "step": 294 }, { "epoch": 0.18873960332693537, "grad_norm": 2.8431640077752203, "learning_rate": 1.257995735607676e-05, "loss": 1.2132, "step": 295 }, { "epoch": 0.18937939859245043, "grad_norm": 3.366519681011942, "learning_rate": 1.26226012793177e-05, "loss": 1.1303, "step": 296 }, { "epoch": 0.19001919385796545, "grad_norm": 3.621480173034483, "learning_rate": 1.2665245202558636e-05, "loss": 1.1675, "step": 297 }, { "epoch": 0.19065898912348048, "grad_norm": 3.502066970413437, "learning_rate": 1.2707889125799576e-05, "loss": 1.0989, "step": 298 }, { "epoch": 0.19129878438899553, "grad_norm": 2.981354897594102, "learning_rate": 1.2750533049040512e-05, "loss": 1.1419, "step": 299 }, { "epoch": 0.19193857965451055, "grad_norm": 3.103101919872764, "learning_rate": 1.2793176972281452e-05, "loss": 1.1191, "step": 300 }, { "epoch": 0.19257837492002558, "grad_norm": 3.1406007400611133, "learning_rate": 1.2835820895522388e-05, "loss": 1.1374, "step": 301 }, { "epoch": 0.19321817018554063, "grad_norm": 3.4113030463850995, "learning_rate": 1.2878464818763328e-05, "loss": 1.1494, "step": 302 }, { "epoch": 0.19385796545105566, "grad_norm": 3.0386072265052584, "learning_rate": 1.2921108742004264e-05, "loss": 1.123, "step": 303 }, { "epoch": 0.1944977607165707, "grad_norm": 3.8951278353705927, "learning_rate": 1.2963752665245204e-05, "loss": 1.117, "step": 304 }, { "epoch": 0.19513755598208574, "grad_norm": 3.8291689780693314, "learning_rate": 1.300639658848614e-05, "loss": 1.1511, "step": 305 }, { "epoch": 0.19577735124760076, "grad_norm": 3.226573117655904, "learning_rate": 1.304904051172708e-05, "loss": 1.0921, "step": 306 }, { "epoch": 0.19641714651311581, "grad_norm": 3.326444796494388, "learning_rate": 1.3091684434968017e-05, "loss": 1.1396, "step": 307 }, { "epoch": 0.19705694177863084, "grad_norm": 3.7672151424686597, "learning_rate": 1.3134328358208957e-05, "loss": 1.1123, "step": 308 }, { "epoch": 0.19769673704414586, "grad_norm": 3.155611358399564, "learning_rate": 1.3176972281449893e-05, "loss": 1.1714, "step": 309 }, { "epoch": 0.19833653230966092, "grad_norm": 3.2677687933846995, "learning_rate": 1.3219616204690833e-05, "loss": 1.1301, "step": 310 }, { "epoch": 0.19897632757517594, "grad_norm": 3.285934027804973, "learning_rate": 1.326226012793177e-05, "loss": 1.0861, "step": 311 }, { "epoch": 0.19961612284069097, "grad_norm": 3.3043413673449913, "learning_rate": 1.3304904051172709e-05, "loss": 1.1083, "step": 312 }, { "epoch": 0.20025591810620602, "grad_norm": 3.13246207350718, "learning_rate": 1.3347547974413649e-05, "loss": 1.1212, "step": 313 }, { "epoch": 0.20089571337172105, "grad_norm": 3.2989598467531245, "learning_rate": 1.3390191897654585e-05, "loss": 1.1106, "step": 314 }, { "epoch": 0.20153550863723607, "grad_norm": 3.5364877581872225, "learning_rate": 1.3432835820895525e-05, "loss": 1.0669, "step": 315 }, { "epoch": 0.20217530390275112, "grad_norm": 3.150101097023388, "learning_rate": 1.3475479744136461e-05, "loss": 1.162, "step": 316 }, { "epoch": 0.20281509916826615, "grad_norm": 3.9017729850525193, "learning_rate": 1.3518123667377401e-05, "loss": 1.1929, "step": 317 }, { "epoch": 0.2034548944337812, "grad_norm": 3.2814391527669913, "learning_rate": 1.3560767590618338e-05, "loss": 1.118, "step": 318 }, { "epoch": 0.20409468969929623, "grad_norm": 3.086075075631408, "learning_rate": 1.3603411513859277e-05, "loss": 1.1069, "step": 319 }, { "epoch": 0.20473448496481125, "grad_norm": 3.5518070729354605, "learning_rate": 1.3646055437100214e-05, "loss": 1.1054, "step": 320 }, { "epoch": 0.2053742802303263, "grad_norm": 3.4192331782393945, "learning_rate": 1.3688699360341154e-05, "loss": 1.0983, "step": 321 }, { "epoch": 0.20601407549584133, "grad_norm": 3.1007670638578917, "learning_rate": 1.373134328358209e-05, "loss": 1.1491, "step": 322 }, { "epoch": 0.20665387076135636, "grad_norm": 3.4078753679610467, "learning_rate": 1.377398720682303e-05, "loss": 1.1498, "step": 323 }, { "epoch": 0.2072936660268714, "grad_norm": 3.2003923309063915, "learning_rate": 1.3816631130063966e-05, "loss": 1.1577, "step": 324 }, { "epoch": 0.20793346129238643, "grad_norm": 3.4434652151856184, "learning_rate": 1.3859275053304906e-05, "loss": 1.0865, "step": 325 }, { "epoch": 0.20857325655790146, "grad_norm": 3.347409021303636, "learning_rate": 1.3901918976545842e-05, "loss": 1.1672, "step": 326 }, { "epoch": 0.2092130518234165, "grad_norm": 3.1948645608418147, "learning_rate": 1.3944562899786782e-05, "loss": 1.1234, "step": 327 }, { "epoch": 0.20985284708893154, "grad_norm": 3.654146855088307, "learning_rate": 1.3987206823027719e-05, "loss": 1.0759, "step": 328 }, { "epoch": 0.2104926423544466, "grad_norm": 3.3316264755694287, "learning_rate": 1.4029850746268658e-05, "loss": 1.0735, "step": 329 }, { "epoch": 0.21113243761996162, "grad_norm": 3.3241224057955896, "learning_rate": 1.4072494669509595e-05, "loss": 1.0858, "step": 330 }, { "epoch": 0.21177223288547664, "grad_norm": 3.788944946188715, "learning_rate": 1.4115138592750535e-05, "loss": 1.1437, "step": 331 }, { "epoch": 0.2124120281509917, "grad_norm": 3.3117045389674926, "learning_rate": 1.4157782515991471e-05, "loss": 1.1452, "step": 332 }, { "epoch": 0.21305182341650672, "grad_norm": 3.7409136668915552, "learning_rate": 1.4200426439232411e-05, "loss": 1.1329, "step": 333 }, { "epoch": 0.21369161868202174, "grad_norm": 3.755990354934821, "learning_rate": 1.4243070362473347e-05, "loss": 1.1207, "step": 334 }, { "epoch": 0.2143314139475368, "grad_norm": 3.315759482977141, "learning_rate": 1.4285714285714287e-05, "loss": 1.1466, "step": 335 }, { "epoch": 0.21497120921305182, "grad_norm": 3.5938540629197426, "learning_rate": 1.4328358208955224e-05, "loss": 1.0831, "step": 336 }, { "epoch": 0.21561100447856685, "grad_norm": 3.3035758512011197, "learning_rate": 1.4371002132196163e-05, "loss": 1.1274, "step": 337 }, { "epoch": 0.2162507997440819, "grad_norm": 3.1717649565492323, "learning_rate": 1.44136460554371e-05, "loss": 1.1145, "step": 338 }, { "epoch": 0.21689059500959693, "grad_norm": 3.300432775278111, "learning_rate": 1.445628997867804e-05, "loss": 1.2125, "step": 339 }, { "epoch": 0.21753039027511195, "grad_norm": 3.2959037664472697, "learning_rate": 1.4498933901918976e-05, "loss": 1.1513, "step": 340 }, { "epoch": 0.218170185540627, "grad_norm": 3.271113090942919, "learning_rate": 1.4541577825159916e-05, "loss": 1.1154, "step": 341 }, { "epoch": 0.21880998080614203, "grad_norm": 2.9100169421646847, "learning_rate": 1.4584221748400854e-05, "loss": 1.0943, "step": 342 }, { "epoch": 0.21944977607165708, "grad_norm": 3.3046850614369307, "learning_rate": 1.4626865671641792e-05, "loss": 1.098, "step": 343 }, { "epoch": 0.2200895713371721, "grad_norm": 3.562610580502571, "learning_rate": 1.4669509594882732e-05, "loss": 1.1251, "step": 344 }, { "epoch": 0.22072936660268713, "grad_norm": 3.8215032725799194, "learning_rate": 1.4712153518123668e-05, "loss": 1.1107, "step": 345 }, { "epoch": 0.22136916186820219, "grad_norm": 3.5311278681415676, "learning_rate": 1.4754797441364608e-05, "loss": 1.068, "step": 346 }, { "epoch": 0.2220089571337172, "grad_norm": 3.6084567207881517, "learning_rate": 1.4797441364605544e-05, "loss": 1.1337, "step": 347 }, { "epoch": 0.22264875239923224, "grad_norm": 3.5468135519828303, "learning_rate": 1.4840085287846484e-05, "loss": 1.115, "step": 348 }, { "epoch": 0.2232885476647473, "grad_norm": 3.640646045647293, "learning_rate": 1.488272921108742e-05, "loss": 1.1177, "step": 349 }, { "epoch": 0.22392834293026231, "grad_norm": 3.6981889529975556, "learning_rate": 1.492537313432836e-05, "loss": 1.137, "step": 350 }, { "epoch": 0.22456813819577734, "grad_norm": 3.8257118022094345, "learning_rate": 1.4968017057569297e-05, "loss": 1.0381, "step": 351 }, { "epoch": 0.2252079334612924, "grad_norm": 3.212425862953948, "learning_rate": 1.5010660980810237e-05, "loss": 1.1149, "step": 352 }, { "epoch": 0.22584772872680742, "grad_norm": 3.3915476343753386, "learning_rate": 1.5053304904051173e-05, "loss": 1.0963, "step": 353 }, { "epoch": 0.22648752399232247, "grad_norm": 3.7823835469411873, "learning_rate": 1.5095948827292113e-05, "loss": 1.1237, "step": 354 }, { "epoch": 0.2271273192578375, "grad_norm": 2.9328449512354644, "learning_rate": 1.513859275053305e-05, "loss": 1.0751, "step": 355 }, { "epoch": 0.22776711452335252, "grad_norm": 3.4094496982041482, "learning_rate": 1.5181236673773989e-05, "loss": 1.1245, "step": 356 }, { "epoch": 0.22840690978886757, "grad_norm": 3.5882184927305767, "learning_rate": 1.5223880597014925e-05, "loss": 1.0629, "step": 357 }, { "epoch": 0.2290467050543826, "grad_norm": 3.4371354045727522, "learning_rate": 1.5266524520255865e-05, "loss": 1.137, "step": 358 }, { "epoch": 0.22968650031989762, "grad_norm": 3.50578455491051, "learning_rate": 1.5309168443496803e-05, "loss": 1.1492, "step": 359 }, { "epoch": 0.23032629558541268, "grad_norm": 3.5956303882820886, "learning_rate": 1.535181236673774e-05, "loss": 1.1345, "step": 360 }, { "epoch": 0.2309660908509277, "grad_norm": 3.34874096838116, "learning_rate": 1.539445628997868e-05, "loss": 1.1275, "step": 361 }, { "epoch": 0.23160588611644273, "grad_norm": 4.050759377027618, "learning_rate": 1.5437100213219618e-05, "loss": 1.0778, "step": 362 }, { "epoch": 0.23224568138195778, "grad_norm": 4.028521205843613, "learning_rate": 1.5479744136460556e-05, "loss": 1.1837, "step": 363 }, { "epoch": 0.2328854766474728, "grad_norm": 3.915606084726931, "learning_rate": 1.5522388059701494e-05, "loss": 1.106, "step": 364 }, { "epoch": 0.23352527191298783, "grad_norm": 3.6437754577328785, "learning_rate": 1.5565031982942432e-05, "loss": 1.1088, "step": 365 }, { "epoch": 0.23416506717850288, "grad_norm": 3.9718713296374624, "learning_rate": 1.560767590618337e-05, "loss": 1.0591, "step": 366 }, { "epoch": 0.2348048624440179, "grad_norm": 3.1215917184327915, "learning_rate": 1.5650319829424308e-05, "loss": 1.1569, "step": 367 }, { "epoch": 0.23544465770953296, "grad_norm": 3.755806028753162, "learning_rate": 1.5692963752665246e-05, "loss": 1.1547, "step": 368 }, { "epoch": 0.236084452975048, "grad_norm": 3.1829232662809224, "learning_rate": 1.5735607675906184e-05, "loss": 1.1579, "step": 369 }, { "epoch": 0.236724248240563, "grad_norm": 3.4573676196424237, "learning_rate": 1.5778251599147122e-05, "loss": 1.1296, "step": 370 }, { "epoch": 0.23736404350607807, "grad_norm": 3.872774470340054, "learning_rate": 1.582089552238806e-05, "loss": 1.0738, "step": 371 }, { "epoch": 0.2380038387715931, "grad_norm": 3.4595117076444395, "learning_rate": 1.5863539445629e-05, "loss": 1.095, "step": 372 }, { "epoch": 0.23864363403710812, "grad_norm": 3.392480279159697, "learning_rate": 1.5906183368869937e-05, "loss": 1.0898, "step": 373 }, { "epoch": 0.23928342930262317, "grad_norm": 3.156089368110268, "learning_rate": 1.5948827292110875e-05, "loss": 1.1528, "step": 374 }, { "epoch": 0.2399232245681382, "grad_norm": 3.5037860400691714, "learning_rate": 1.5991471215351813e-05, "loss": 1.1478, "step": 375 }, { "epoch": 0.24056301983365322, "grad_norm": 3.162636605487485, "learning_rate": 1.603411513859275e-05, "loss": 1.0863, "step": 376 }, { "epoch": 0.24120281509916827, "grad_norm": 3.0824723347636565, "learning_rate": 1.6076759061833693e-05, "loss": 1.0899, "step": 377 }, { "epoch": 0.2418426103646833, "grad_norm": 3.1622800248824983, "learning_rate": 1.6119402985074627e-05, "loss": 1.0963, "step": 378 }, { "epoch": 0.24248240563019835, "grad_norm": 3.4212761139463463, "learning_rate": 1.616204690831557e-05, "loss": 1.094, "step": 379 }, { "epoch": 0.24312220089571338, "grad_norm": 2.9686440229370015, "learning_rate": 1.6204690831556504e-05, "loss": 1.0773, "step": 380 }, { "epoch": 0.2437619961612284, "grad_norm": 3.3739857664228876, "learning_rate": 1.6247334754797445e-05, "loss": 1.1225, "step": 381 }, { "epoch": 0.24440179142674345, "grad_norm": 3.1216026162193846, "learning_rate": 1.628997867803838e-05, "loss": 1.1174, "step": 382 }, { "epoch": 0.24504158669225848, "grad_norm": 3.21718553609252, "learning_rate": 1.633262260127932e-05, "loss": 1.0997, "step": 383 }, { "epoch": 0.2456813819577735, "grad_norm": 3.33146036995984, "learning_rate": 1.6375266524520256e-05, "loss": 1.1103, "step": 384 }, { "epoch": 0.24632117722328856, "grad_norm": 3.523011752316942, "learning_rate": 1.6417910447761197e-05, "loss": 1.1102, "step": 385 }, { "epoch": 0.24696097248880358, "grad_norm": 3.5215095739197833, "learning_rate": 1.6460554371002132e-05, "loss": 1.1313, "step": 386 }, { "epoch": 0.2476007677543186, "grad_norm": 3.4903815883124887, "learning_rate": 1.6503198294243074e-05, "loss": 1.0896, "step": 387 }, { "epoch": 0.24824056301983366, "grad_norm": 3.30227077138365, "learning_rate": 1.654584221748401e-05, "loss": 1.133, "step": 388 }, { "epoch": 0.24888035828534869, "grad_norm": 3.5273582833690282, "learning_rate": 1.658848614072495e-05, "loss": 1.1317, "step": 389 }, { "epoch": 0.2495201535508637, "grad_norm": 2.966398186124767, "learning_rate": 1.6631130063965885e-05, "loss": 1.1113, "step": 390 }, { "epoch": 0.25015994881637876, "grad_norm": 3.514322120798775, "learning_rate": 1.6673773987206826e-05, "loss": 1.1101, "step": 391 }, { "epoch": 0.2507997440818938, "grad_norm": 3.5181993829244034, "learning_rate": 1.671641791044776e-05, "loss": 1.1249, "step": 392 }, { "epoch": 0.2514395393474088, "grad_norm": 3.355699064582302, "learning_rate": 1.6759061833688702e-05, "loss": 1.1654, "step": 393 }, { "epoch": 0.25207933461292387, "grad_norm": 3.501785968145833, "learning_rate": 1.6801705756929637e-05, "loss": 1.1961, "step": 394 }, { "epoch": 0.2527191298784389, "grad_norm": 4.216289545554502, "learning_rate": 1.684434968017058e-05, "loss": 1.0824, "step": 395 }, { "epoch": 0.2533589251439539, "grad_norm": 3.9052119145071136, "learning_rate": 1.6886993603411513e-05, "loss": 1.0889, "step": 396 }, { "epoch": 0.25399872040946897, "grad_norm": 3.294792222090486, "learning_rate": 1.6929637526652455e-05, "loss": 1.0946, "step": 397 }, { "epoch": 0.254638515674984, "grad_norm": 3.722272341509806, "learning_rate": 1.697228144989339e-05, "loss": 1.1115, "step": 398 }, { "epoch": 0.255278310940499, "grad_norm": 3.384677575790275, "learning_rate": 1.701492537313433e-05, "loss": 1.1778, "step": 399 }, { "epoch": 0.2559181062060141, "grad_norm": 3.065550235410346, "learning_rate": 1.7057569296375266e-05, "loss": 1.0979, "step": 400 }, { "epoch": 0.2565579014715291, "grad_norm": 3.293965021892677, "learning_rate": 1.7100213219616207e-05, "loss": 1.1457, "step": 401 }, { "epoch": 0.2571976967370441, "grad_norm": 3.398058797780089, "learning_rate": 1.7142857142857142e-05, "loss": 1.119, "step": 402 }, { "epoch": 0.2578374920025592, "grad_norm": 2.871555795954757, "learning_rate": 1.7185501066098083e-05, "loss": 1.1589, "step": 403 }, { "epoch": 0.25847728726807423, "grad_norm": 3.3149621959666593, "learning_rate": 1.7228144989339018e-05, "loss": 1.049, "step": 404 }, { "epoch": 0.2591170825335892, "grad_norm": 3.7884480387246295, "learning_rate": 1.727078891257996e-05, "loss": 1.1689, "step": 405 }, { "epoch": 0.2597568777991043, "grad_norm": 3.6395420055837713, "learning_rate": 1.7313432835820894e-05, "loss": 1.1644, "step": 406 }, { "epoch": 0.26039667306461933, "grad_norm": 3.0480050299747243, "learning_rate": 1.7356076759061836e-05, "loss": 1.1177, "step": 407 }, { "epoch": 0.26103646833013433, "grad_norm": 3.34001292083092, "learning_rate": 1.7398720682302774e-05, "loss": 1.0468, "step": 408 }, { "epoch": 0.2616762635956494, "grad_norm": 3.665032717767305, "learning_rate": 1.7441364605543712e-05, "loss": 1.1105, "step": 409 }, { "epoch": 0.26231605886116444, "grad_norm": 3.4804737347291534, "learning_rate": 1.748400852878465e-05, "loss": 1.0863, "step": 410 }, { "epoch": 0.2629558541266795, "grad_norm": 3.4065093254278764, "learning_rate": 1.7526652452025588e-05, "loss": 1.112, "step": 411 }, { "epoch": 0.2635956493921945, "grad_norm": 3.779749324256479, "learning_rate": 1.7569296375266526e-05, "loss": 1.05, "step": 412 }, { "epoch": 0.26423544465770954, "grad_norm": 3.3499142656146934, "learning_rate": 1.7611940298507464e-05, "loss": 1.1212, "step": 413 }, { "epoch": 0.2648752399232246, "grad_norm": 3.728304302380523, "learning_rate": 1.7654584221748402e-05, "loss": 1.1363, "step": 414 }, { "epoch": 0.2655150351887396, "grad_norm": 3.203130324538498, "learning_rate": 1.769722814498934e-05, "loss": 1.136, "step": 415 }, { "epoch": 0.26615483045425464, "grad_norm": 3.004725251746009, "learning_rate": 1.773987206823028e-05, "loss": 1.0928, "step": 416 }, { "epoch": 0.2667946257197697, "grad_norm": 3.5473308452647507, "learning_rate": 1.7782515991471217e-05, "loss": 1.1194, "step": 417 }, { "epoch": 0.2674344209852847, "grad_norm": 3.631466018925729, "learning_rate": 1.7825159914712155e-05, "loss": 1.0848, "step": 418 }, { "epoch": 0.26807421625079975, "grad_norm": 3.5705723170550523, "learning_rate": 1.7867803837953093e-05, "loss": 1.0857, "step": 419 }, { "epoch": 0.2687140115163148, "grad_norm": 3.3852889956071035, "learning_rate": 1.791044776119403e-05, "loss": 1.0602, "step": 420 }, { "epoch": 0.2693538067818298, "grad_norm": 3.5535436698909653, "learning_rate": 1.795309168443497e-05, "loss": 1.1164, "step": 421 }, { "epoch": 0.26999360204734485, "grad_norm": 3.791337661470951, "learning_rate": 1.7995735607675907e-05, "loss": 1.1135, "step": 422 }, { "epoch": 0.2706333973128599, "grad_norm": 3.4471880594514444, "learning_rate": 1.8038379530916845e-05, "loss": 1.0355, "step": 423 }, { "epoch": 0.2712731925783749, "grad_norm": 4.6556016368406805, "learning_rate": 1.8081023454157784e-05, "loss": 1.1389, "step": 424 }, { "epoch": 0.27191298784388995, "grad_norm": 3.1769437904673357, "learning_rate": 1.812366737739872e-05, "loss": 1.1148, "step": 425 }, { "epoch": 0.272552783109405, "grad_norm": 3.5384940948094954, "learning_rate": 1.816631130063966e-05, "loss": 1.118, "step": 426 }, { "epoch": 0.27319257837492, "grad_norm": 4.4182541996280404, "learning_rate": 1.8208955223880598e-05, "loss": 1.1078, "step": 427 }, { "epoch": 0.27383237364043506, "grad_norm": 4.268056737112456, "learning_rate": 1.8251599147121536e-05, "loss": 1.1644, "step": 428 }, { "epoch": 0.2744721689059501, "grad_norm": 4.187778056942183, "learning_rate": 1.8294243070362474e-05, "loss": 1.0874, "step": 429 }, { "epoch": 0.2751119641714651, "grad_norm": 4.004787737090405, "learning_rate": 1.8336886993603412e-05, "loss": 1.0889, "step": 430 }, { "epoch": 0.27575175943698016, "grad_norm": 3.240683608661553, "learning_rate": 1.837953091684435e-05, "loss": 1.1309, "step": 431 }, { "epoch": 0.2763915547024952, "grad_norm": 4.934566819334725, "learning_rate": 1.842217484008529e-05, "loss": 1.122, "step": 432 }, { "epoch": 0.2770313499680102, "grad_norm": 4.450298414038784, "learning_rate": 1.8464818763326226e-05, "loss": 1.0835, "step": 433 }, { "epoch": 0.27767114523352526, "grad_norm": 4.186565002192997, "learning_rate": 1.8507462686567165e-05, "loss": 1.1034, "step": 434 }, { "epoch": 0.2783109404990403, "grad_norm": 2.9625047479125834, "learning_rate": 1.8550106609808103e-05, "loss": 1.0776, "step": 435 }, { "epoch": 0.27895073576455537, "grad_norm": 3.7718395371995355, "learning_rate": 1.859275053304904e-05, "loss": 1.0645, "step": 436 }, { "epoch": 0.27959053103007037, "grad_norm": 4.31480130647734, "learning_rate": 1.863539445628998e-05, "loss": 1.1102, "step": 437 }, { "epoch": 0.2802303262955854, "grad_norm": 3.396901602615414, "learning_rate": 1.8678038379530917e-05, "loss": 1.0878, "step": 438 }, { "epoch": 0.28087012156110047, "grad_norm": 3.5394569583318862, "learning_rate": 1.872068230277186e-05, "loss": 1.1311, "step": 439 }, { "epoch": 0.28150991682661547, "grad_norm": 3.350861601369052, "learning_rate": 1.8763326226012793e-05, "loss": 1.1356, "step": 440 }, { "epoch": 0.2821497120921305, "grad_norm": 3.925367230264354, "learning_rate": 1.8805970149253735e-05, "loss": 1.1113, "step": 441 }, { "epoch": 0.2827895073576456, "grad_norm": 3.5577257126524624, "learning_rate": 1.884861407249467e-05, "loss": 1.1035, "step": 442 }, { "epoch": 0.2834293026231606, "grad_norm": 3.1498376906496564, "learning_rate": 1.889125799573561e-05, "loss": 1.1418, "step": 443 }, { "epoch": 0.2840690978886756, "grad_norm": 3.2944968417262896, "learning_rate": 1.8933901918976546e-05, "loss": 1.062, "step": 444 }, { "epoch": 0.2847088931541907, "grad_norm": 3.499066413110087, "learning_rate": 1.8976545842217487e-05, "loss": 1.059, "step": 445 }, { "epoch": 0.2853486884197057, "grad_norm": 4.969420957479504, "learning_rate": 1.9019189765458422e-05, "loss": 1.0358, "step": 446 }, { "epoch": 0.28598848368522073, "grad_norm": 3.4030711033475214, "learning_rate": 1.9061833688699363e-05, "loss": 1.1223, "step": 447 }, { "epoch": 0.2866282789507358, "grad_norm": 2.963574536416267, "learning_rate": 1.9104477611940298e-05, "loss": 1.0613, "step": 448 }, { "epoch": 0.2872680742162508, "grad_norm": 3.9078188095400193, "learning_rate": 1.914712153518124e-05, "loss": 1.0999, "step": 449 }, { "epoch": 0.28790786948176583, "grad_norm": 3.8404547054851395, "learning_rate": 1.9189765458422178e-05, "loss": 1.1377, "step": 450 }, { "epoch": 0.2885476647472809, "grad_norm": 3.1467251931103033, "learning_rate": 1.9232409381663116e-05, "loss": 1.0454, "step": 451 }, { "epoch": 0.2891874600127959, "grad_norm": 3.1704431861221734, "learning_rate": 1.9275053304904054e-05, "loss": 1.0988, "step": 452 }, { "epoch": 0.28982725527831094, "grad_norm": 3.0691213140893026, "learning_rate": 1.9317697228144992e-05, "loss": 1.089, "step": 453 }, { "epoch": 0.290467050543826, "grad_norm": 3.0517145958857577, "learning_rate": 1.936034115138593e-05, "loss": 1.0743, "step": 454 }, { "epoch": 0.291106845809341, "grad_norm": 3.491802749451594, "learning_rate": 1.9402985074626868e-05, "loss": 1.0588, "step": 455 }, { "epoch": 0.29174664107485604, "grad_norm": 3.360029133842905, "learning_rate": 1.9445628997867806e-05, "loss": 1.0271, "step": 456 }, { "epoch": 0.2923864363403711, "grad_norm": 3.0795669827300887, "learning_rate": 1.9488272921108744e-05, "loss": 1.1038, "step": 457 }, { "epoch": 0.2930262316058861, "grad_norm": 2.7977143071786816, "learning_rate": 1.9530916844349682e-05, "loss": 1.1204, "step": 458 }, { "epoch": 0.29366602687140114, "grad_norm": 3.5358532656178507, "learning_rate": 1.957356076759062e-05, "loss": 1.0955, "step": 459 }, { "epoch": 0.2943058221369162, "grad_norm": 3.335364684769508, "learning_rate": 1.961620469083156e-05, "loss": 1.0978, "step": 460 }, { "epoch": 0.29494561740243125, "grad_norm": 3.077985504661556, "learning_rate": 1.9658848614072497e-05, "loss": 1.1014, "step": 461 }, { "epoch": 0.29558541266794625, "grad_norm": 3.243831694652957, "learning_rate": 1.9701492537313435e-05, "loss": 1.1, "step": 462 }, { "epoch": 0.2962252079334613, "grad_norm": 3.3863163091548043, "learning_rate": 1.9744136460554373e-05, "loss": 1.1238, "step": 463 }, { "epoch": 0.29686500319897635, "grad_norm": 3.6558891957149173, "learning_rate": 1.978678038379531e-05, "loss": 1.0727, "step": 464 }, { "epoch": 0.29750479846449135, "grad_norm": 3.5774495595870053, "learning_rate": 1.982942430703625e-05, "loss": 1.1074, "step": 465 }, { "epoch": 0.2981445937300064, "grad_norm": 3.085419045447863, "learning_rate": 1.9872068230277187e-05, "loss": 1.1477, "step": 466 }, { "epoch": 0.29878438899552145, "grad_norm": 3.166471188443695, "learning_rate": 1.9914712153518125e-05, "loss": 1.1111, "step": 467 }, { "epoch": 0.29942418426103645, "grad_norm": 3.6314258579643557, "learning_rate": 1.9957356076759064e-05, "loss": 1.0576, "step": 468 }, { "epoch": 0.3000639795265515, "grad_norm": 3.2487461632822683, "learning_rate": 2e-05, "loss": 1.1221, "step": 469 }, { "epoch": 0.30070377479206656, "grad_norm": 3.0018672269838738, "learning_rate": 1.9999999785308896e-05, "loss": 1.0846, "step": 470 }, { "epoch": 0.30134357005758156, "grad_norm": 3.6316063218344294, "learning_rate": 1.9999999141235587e-05, "loss": 1.1056, "step": 471 }, { "epoch": 0.3019833653230966, "grad_norm": 3.251665926725323, "learning_rate": 1.9999998067780108e-05, "loss": 1.1286, "step": 472 }, { "epoch": 0.30262316058861166, "grad_norm": 3.6049822620206577, "learning_rate": 1.9999996564942496e-05, "loss": 1.1492, "step": 473 }, { "epoch": 0.30326295585412666, "grad_norm": 3.5742044738778445, "learning_rate": 1.999999463272282e-05, "loss": 1.079, "step": 474 }, { "epoch": 0.3039027511196417, "grad_norm": 3.313461976890784, "learning_rate": 1.9999992271121165e-05, "loss": 1.0981, "step": 475 }, { "epoch": 0.30454254638515676, "grad_norm": 3.1148444887086604, "learning_rate": 1.999998948013763e-05, "loss": 1.0301, "step": 476 }, { "epoch": 0.30518234165067176, "grad_norm": 3.53821338987834, "learning_rate": 1.9999986259772333e-05, "loss": 1.0858, "step": 477 }, { "epoch": 0.3058221369161868, "grad_norm": 3.124269157426491, "learning_rate": 1.999998261002542e-05, "loss": 1.1128, "step": 478 }, { "epoch": 0.30646193218170187, "grad_norm": 3.0713619729625385, "learning_rate": 1.9999978530897037e-05, "loss": 1.0302, "step": 479 }, { "epoch": 0.30710172744721687, "grad_norm": 3.2524897409295597, "learning_rate": 1.9999974022387366e-05, "loss": 1.1509, "step": 480 }, { "epoch": 0.3077415227127319, "grad_norm": 3.6824143190389647, "learning_rate": 1.99999690844966e-05, "loss": 1.1087, "step": 481 }, { "epoch": 0.30838131797824697, "grad_norm": 3.218618783720022, "learning_rate": 1.9999963717224952e-05, "loss": 1.1094, "step": 482 }, { "epoch": 0.30902111324376197, "grad_norm": 2.905164543516948, "learning_rate": 1.9999957920572647e-05, "loss": 1.1457, "step": 483 }, { "epoch": 0.309660908509277, "grad_norm": 3.63660545606228, "learning_rate": 1.999995169453994e-05, "loss": 1.0648, "step": 484 }, { "epoch": 0.3103007037747921, "grad_norm": 3.2022487969801823, "learning_rate": 1.9999945039127092e-05, "loss": 1.0697, "step": 485 }, { "epoch": 0.31094049904030713, "grad_norm": 3.2952068666611414, "learning_rate": 1.9999937954334396e-05, "loss": 1.1163, "step": 486 }, { "epoch": 0.3115802943058221, "grad_norm": 3.228159088785838, "learning_rate": 1.9999930440162152e-05, "loss": 1.1258, "step": 487 }, { "epoch": 0.3122200895713372, "grad_norm": 3.3522914630049536, "learning_rate": 1.9999922496610682e-05, "loss": 1.1242, "step": 488 }, { "epoch": 0.31285988483685223, "grad_norm": 3.8783317477742547, "learning_rate": 1.9999914123680332e-05, "loss": 1.1291, "step": 489 }, { "epoch": 0.31349968010236723, "grad_norm": 3.829466724596865, "learning_rate": 1.9999905321371454e-05, "loss": 1.0903, "step": 490 }, { "epoch": 0.3141394753678823, "grad_norm": 3.3184420834806936, "learning_rate": 1.999989608968443e-05, "loss": 1.1264, "step": 491 }, { "epoch": 0.31477927063339733, "grad_norm": 2.7827646214433326, "learning_rate": 1.999988642861966e-05, "loss": 1.0651, "step": 492 }, { "epoch": 0.31541906589891233, "grad_norm": 2.8315265048595655, "learning_rate": 1.9999876338177554e-05, "loss": 1.0389, "step": 493 }, { "epoch": 0.3160588611644274, "grad_norm": 3.004422929063441, "learning_rate": 1.9999865818358547e-05, "loss": 1.073, "step": 494 }, { "epoch": 0.31669865642994244, "grad_norm": 3.0907387551228775, "learning_rate": 1.9999854869163088e-05, "loss": 1.007, "step": 495 }, { "epoch": 0.31733845169545744, "grad_norm": 3.2753939808713994, "learning_rate": 1.999984349059165e-05, "loss": 1.1082, "step": 496 }, { "epoch": 0.3179782469609725, "grad_norm": 3.3387856550983424, "learning_rate": 1.999983168264472e-05, "loss": 1.0862, "step": 497 }, { "epoch": 0.31861804222648754, "grad_norm": 2.7397527986631363, "learning_rate": 1.9999819445322805e-05, "loss": 1.0922, "step": 498 }, { "epoch": 0.31925783749200254, "grad_norm": 2.9966139367827767, "learning_rate": 1.9999806778626435e-05, "loss": 1.0635, "step": 499 }, { "epoch": 0.3198976327575176, "grad_norm": 3.0785268352785087, "learning_rate": 1.999979368255615e-05, "loss": 1.0761, "step": 500 }, { "epoch": 0.32053742802303264, "grad_norm": 3.26018012715311, "learning_rate": 1.9999780157112506e-05, "loss": 1.1011, "step": 501 }, { "epoch": 0.32117722328854764, "grad_norm": 2.9033169020711775, "learning_rate": 1.9999766202296098e-05, "loss": 1.0408, "step": 502 }, { "epoch": 0.3218170185540627, "grad_norm": 3.0019815296919323, "learning_rate": 1.9999751818107515e-05, "loss": 1.0459, "step": 503 }, { "epoch": 0.32245681381957775, "grad_norm": 3.887160482706229, "learning_rate": 1.9999737004547377e-05, "loss": 1.0521, "step": 504 }, { "epoch": 0.32309660908509275, "grad_norm": 3.0924384939020646, "learning_rate": 1.9999721761616317e-05, "loss": 1.0722, "step": 505 }, { "epoch": 0.3237364043506078, "grad_norm": 3.0821727559701038, "learning_rate": 1.9999706089314997e-05, "loss": 1.0346, "step": 506 }, { "epoch": 0.32437619961612285, "grad_norm": 3.4732522739059877, "learning_rate": 1.9999689987644083e-05, "loss": 1.0795, "step": 507 }, { "epoch": 0.32501599488163785, "grad_norm": 3.811848032888385, "learning_rate": 1.9999673456604268e-05, "loss": 1.0811, "step": 508 }, { "epoch": 0.3256557901471529, "grad_norm": 3.600472572281304, "learning_rate": 1.999965649619627e-05, "loss": 1.0567, "step": 509 }, { "epoch": 0.32629558541266795, "grad_norm": 2.6887940464408238, "learning_rate": 1.9999639106420804e-05, "loss": 1.1344, "step": 510 }, { "epoch": 0.326935380678183, "grad_norm": 3.1155707608799568, "learning_rate": 1.9999621287278622e-05, "loss": 1.032, "step": 511 }, { "epoch": 0.327575175943698, "grad_norm": 3.388988476025516, "learning_rate": 1.9999603038770493e-05, "loss": 1.0006, "step": 512 }, { "epoch": 0.32821497120921306, "grad_norm": 3.893544519812779, "learning_rate": 1.9999584360897195e-05, "loss": 1.1261, "step": 513 }, { "epoch": 0.3288547664747281, "grad_norm": 2.7079103898830033, "learning_rate": 1.9999565253659532e-05, "loss": 1.0462, "step": 514 }, { "epoch": 0.3294945617402431, "grad_norm": 2.858013813216633, "learning_rate": 1.9999545717058327e-05, "loss": 1.0216, "step": 515 }, { "epoch": 0.33013435700575816, "grad_norm": 2.809686134043979, "learning_rate": 1.9999525751094417e-05, "loss": 1.0665, "step": 516 }, { "epoch": 0.3307741522712732, "grad_norm": 3.4667257629756376, "learning_rate": 1.999950535576866e-05, "loss": 1.1299, "step": 517 }, { "epoch": 0.3314139475367882, "grad_norm": 2.9508712754823425, "learning_rate": 1.9999484531081926e-05, "loss": 1.086, "step": 518 }, { "epoch": 0.33205374280230326, "grad_norm": 2.8310919920699136, "learning_rate": 1.9999463277035115e-05, "loss": 1.1277, "step": 519 }, { "epoch": 0.3326935380678183, "grad_norm": 2.8332478879666314, "learning_rate": 1.9999441593629142e-05, "loss": 1.0811, "step": 520 }, { "epoch": 0.3333333333333333, "grad_norm": 3.2999848186470597, "learning_rate": 1.9999419480864933e-05, "loss": 1.0808, "step": 521 }, { "epoch": 0.33397312859884837, "grad_norm": 3.3506496215267867, "learning_rate": 1.999939693874344e-05, "loss": 1.1202, "step": 522 }, { "epoch": 0.3346129238643634, "grad_norm": 2.739078769938623, "learning_rate": 1.9999373967265628e-05, "loss": 1.0874, "step": 523 }, { "epoch": 0.3352527191298784, "grad_norm": 3.393760011636325, "learning_rate": 1.9999350566432487e-05, "loss": 1.1263, "step": 524 }, { "epoch": 0.33589251439539347, "grad_norm": 3.0950713225255257, "learning_rate": 1.999932673624502e-05, "loss": 1.0852, "step": 525 }, { "epoch": 0.3365323096609085, "grad_norm": 3.2326298796783934, "learning_rate": 1.9999302476704246e-05, "loss": 1.0544, "step": 526 }, { "epoch": 0.3371721049264235, "grad_norm": 2.454801559721618, "learning_rate": 1.9999277787811216e-05, "loss": 1.072, "step": 527 }, { "epoch": 0.3378119001919386, "grad_norm": 3.1175227547336357, "learning_rate": 1.9999252669566987e-05, "loss": 1.0384, "step": 528 }, { "epoch": 0.3384516954574536, "grad_norm": 3.216495927675624, "learning_rate": 1.9999227121972633e-05, "loss": 1.1072, "step": 529 }, { "epoch": 0.3390914907229686, "grad_norm": 3.2362745009066667, "learning_rate": 1.999920114502925e-05, "loss": 1.0733, "step": 530 }, { "epoch": 0.3397312859884837, "grad_norm": 3.0325411131227877, "learning_rate": 1.9999174738737964e-05, "loss": 1.1038, "step": 531 }, { "epoch": 0.34037108125399873, "grad_norm": 3.0893536450731793, "learning_rate": 1.9999147903099896e-05, "loss": 1.1175, "step": 532 }, { "epoch": 0.34101087651951373, "grad_norm": 2.7015630970103857, "learning_rate": 1.999912063811621e-05, "loss": 1.1195, "step": 533 }, { "epoch": 0.3416506717850288, "grad_norm": 3.16687701549675, "learning_rate": 1.9999092943788065e-05, "loss": 1.074, "step": 534 }, { "epoch": 0.34229046705054383, "grad_norm": 3.104714218957641, "learning_rate": 1.9999064820116657e-05, "loss": 1.0307, "step": 535 }, { "epoch": 0.3429302623160589, "grad_norm": 2.791575625648502, "learning_rate": 1.9999036267103195e-05, "loss": 1.0743, "step": 536 }, { "epoch": 0.3435700575815739, "grad_norm": 2.8760273925804025, "learning_rate": 1.99990072847489e-05, "loss": 1.0636, "step": 537 }, { "epoch": 0.34420985284708894, "grad_norm": 2.766666614016878, "learning_rate": 1.9998977873055022e-05, "loss": 1.0845, "step": 538 }, { "epoch": 0.344849648112604, "grad_norm": 2.971854258924147, "learning_rate": 1.9998948032022816e-05, "loss": 1.0426, "step": 539 }, { "epoch": 0.345489443378119, "grad_norm": 2.677518391897263, "learning_rate": 1.9998917761653574e-05, "loss": 1.0605, "step": 540 }, { "epoch": 0.34612923864363404, "grad_norm": 3.797487571603181, "learning_rate": 1.9998887061948585e-05, "loss": 1.047, "step": 541 }, { "epoch": 0.3467690339091491, "grad_norm": 3.444725709488839, "learning_rate": 1.999885593290918e-05, "loss": 1.0907, "step": 542 }, { "epoch": 0.3474088291746641, "grad_norm": 2.8742642259251863, "learning_rate": 1.999882437453668e-05, "loss": 1.0199, "step": 543 }, { "epoch": 0.34804862444017914, "grad_norm": 3.719996531929092, "learning_rate": 1.999879238683245e-05, "loss": 1.0857, "step": 544 }, { "epoch": 0.3486884197056942, "grad_norm": 3.4402443696091325, "learning_rate": 1.9998759969797864e-05, "loss": 1.0262, "step": 545 }, { "epoch": 0.3493282149712092, "grad_norm": 2.974475921825772, "learning_rate": 1.9998727123434306e-05, "loss": 1.0708, "step": 546 }, { "epoch": 0.34996801023672425, "grad_norm": 3.2065260259655877, "learning_rate": 1.9998693847743194e-05, "loss": 1.0086, "step": 547 }, { "epoch": 0.3506078055022393, "grad_norm": 3.539292334786551, "learning_rate": 1.9998660142725956e-05, "loss": 1.0421, "step": 548 }, { "epoch": 0.3512476007677543, "grad_norm": 2.9957005872728, "learning_rate": 1.999862600838404e-05, "loss": 1.0511, "step": 549 }, { "epoch": 0.35188739603326935, "grad_norm": 3.282297188190174, "learning_rate": 1.9998591444718905e-05, "loss": 1.0692, "step": 550 }, { "epoch": 0.3525271912987844, "grad_norm": 3.044380817938427, "learning_rate": 1.999855645173204e-05, "loss": 1.1023, "step": 551 }, { "epoch": 0.3531669865642994, "grad_norm": 3.044225295273326, "learning_rate": 1.9998521029424948e-05, "loss": 1.1536, "step": 552 }, { "epoch": 0.35380678182981445, "grad_norm": 2.8303284618884774, "learning_rate": 1.999848517779915e-05, "loss": 1.1138, "step": 553 }, { "epoch": 0.3544465770953295, "grad_norm": 3.221542808664982, "learning_rate": 1.999844889685618e-05, "loss": 1.0545, "step": 554 }, { "epoch": 0.3550863723608445, "grad_norm": 2.9807737936315455, "learning_rate": 1.9998412186597605e-05, "loss": 1.1235, "step": 555 }, { "epoch": 0.35572616762635956, "grad_norm": 3.0459820559425568, "learning_rate": 1.9998375047024996e-05, "loss": 1.0687, "step": 556 }, { "epoch": 0.3563659628918746, "grad_norm": 2.9387847819443453, "learning_rate": 1.9998337478139947e-05, "loss": 1.1087, "step": 557 }, { "epoch": 0.3570057581573896, "grad_norm": 2.637212972334978, "learning_rate": 1.9998299479944072e-05, "loss": 1.1012, "step": 558 }, { "epoch": 0.35764555342290466, "grad_norm": 2.832599435623506, "learning_rate": 1.9998261052439005e-05, "loss": 1.0837, "step": 559 }, { "epoch": 0.3582853486884197, "grad_norm": 3.0347387934608316, "learning_rate": 1.9998222195626394e-05, "loss": 1.0919, "step": 560 }, { "epoch": 0.35892514395393477, "grad_norm": 2.737454236888958, "learning_rate": 1.9998182909507904e-05, "loss": 1.0447, "step": 561 }, { "epoch": 0.35956493921944976, "grad_norm": 2.7543372350087387, "learning_rate": 1.9998143194085228e-05, "loss": 1.0831, "step": 562 }, { "epoch": 0.3602047344849648, "grad_norm": 2.9941670218561263, "learning_rate": 1.9998103049360068e-05, "loss": 1.1098, "step": 563 }, { "epoch": 0.36084452975047987, "grad_norm": 2.8744844316476015, "learning_rate": 1.9998062475334148e-05, "loss": 1.0968, "step": 564 }, { "epoch": 0.36148432501599487, "grad_norm": 2.6871501095234076, "learning_rate": 1.999802147200921e-05, "loss": 1.109, "step": 565 }, { "epoch": 0.3621241202815099, "grad_norm": 3.1285340215876167, "learning_rate": 1.9997980039387014e-05, "loss": 1.0401, "step": 566 }, { "epoch": 0.362763915547025, "grad_norm": 2.960968285233088, "learning_rate": 1.999793817746934e-05, "loss": 1.0972, "step": 567 }, { "epoch": 0.36340371081253997, "grad_norm": 3.1431943082189515, "learning_rate": 1.999789588625799e-05, "loss": 1.0567, "step": 568 }, { "epoch": 0.364043506078055, "grad_norm": 3.0731904075570067, "learning_rate": 1.999785316575477e-05, "loss": 1.0543, "step": 569 }, { "epoch": 0.3646833013435701, "grad_norm": 3.2124901008772713, "learning_rate": 1.9997810015961524e-05, "loss": 1.0938, "step": 570 }, { "epoch": 0.3653230966090851, "grad_norm": 3.2071072906414613, "learning_rate": 1.99977664368801e-05, "loss": 1.0384, "step": 571 }, { "epoch": 0.3659628918746001, "grad_norm": 3.2729137954121477, "learning_rate": 1.999772242851237e-05, "loss": 1.0432, "step": 572 }, { "epoch": 0.3666026871401152, "grad_norm": 3.599910608412652, "learning_rate": 1.9997677990860223e-05, "loss": 1.0654, "step": 573 }, { "epoch": 0.3672424824056302, "grad_norm": 3.3176461469844583, "learning_rate": 1.9997633123925565e-05, "loss": 1.0928, "step": 574 }, { "epoch": 0.36788227767114523, "grad_norm": 3.4307100280772422, "learning_rate": 1.999758782771033e-05, "loss": 1.0914, "step": 575 }, { "epoch": 0.3685220729366603, "grad_norm": 3.869864666630401, "learning_rate": 1.9997542102216454e-05, "loss": 1.0343, "step": 576 }, { "epoch": 0.3691618682021753, "grad_norm": 3.136746648855464, "learning_rate": 1.9997495947445907e-05, "loss": 1.0496, "step": 577 }, { "epoch": 0.36980166346769033, "grad_norm": 3.180391924475081, "learning_rate": 1.999744936340067e-05, "loss": 1.0451, "step": 578 }, { "epoch": 0.3704414587332054, "grad_norm": 2.793918584556651, "learning_rate": 1.9997402350082737e-05, "loss": 1.1111, "step": 579 }, { "epoch": 0.3710812539987204, "grad_norm": 3.1252609044750357, "learning_rate": 1.9997354907494132e-05, "loss": 1.0804, "step": 580 }, { "epoch": 0.37172104926423544, "grad_norm": 3.2484445477520736, "learning_rate": 1.9997307035636887e-05, "loss": 1.0774, "step": 581 }, { "epoch": 0.3723608445297505, "grad_norm": 3.2297512371881822, "learning_rate": 1.9997258734513068e-05, "loss": 1.0411, "step": 582 }, { "epoch": 0.3730006397952655, "grad_norm": 3.3821342966001957, "learning_rate": 1.9997210004124742e-05, "loss": 1.0623, "step": 583 }, { "epoch": 0.37364043506078054, "grad_norm": 2.922996917017965, "learning_rate": 1.9997160844474e-05, "loss": 1.1023, "step": 584 }, { "epoch": 0.3742802303262956, "grad_norm": 3.546590976614763, "learning_rate": 1.9997111255562958e-05, "loss": 1.0636, "step": 585 }, { "epoch": 0.37492002559181065, "grad_norm": 3.54874893675424, "learning_rate": 1.999706123739374e-05, "loss": 1.0963, "step": 586 }, { "epoch": 0.37555982085732564, "grad_norm": 2.773217717425695, "learning_rate": 1.9997010789968493e-05, "loss": 1.107, "step": 587 }, { "epoch": 0.3761996161228407, "grad_norm": 2.9205180782051574, "learning_rate": 1.9996959913289392e-05, "loss": 1.0372, "step": 588 }, { "epoch": 0.37683941138835575, "grad_norm": 3.3101803394757923, "learning_rate": 1.9996908607358612e-05, "loss": 1.0725, "step": 589 }, { "epoch": 0.37747920665387075, "grad_norm": 2.6707534111777087, "learning_rate": 1.9996856872178358e-05, "loss": 1.1046, "step": 590 }, { "epoch": 0.3781190019193858, "grad_norm": 2.661585918183967, "learning_rate": 1.9996804707750853e-05, "loss": 1.1295, "step": 591 }, { "epoch": 0.37875879718490085, "grad_norm": 2.688960548470395, "learning_rate": 1.9996752114078343e-05, "loss": 1.0595, "step": 592 }, { "epoch": 0.37939859245041585, "grad_norm": 2.877952183997722, "learning_rate": 1.9996699091163075e-05, "loss": 1.044, "step": 593 }, { "epoch": 0.3800383877159309, "grad_norm": 2.636119388808952, "learning_rate": 1.999664563900733e-05, "loss": 1.0998, "step": 594 }, { "epoch": 0.38067818298144596, "grad_norm": 2.654250114875707, "learning_rate": 1.9996591757613407e-05, "loss": 1.0729, "step": 595 }, { "epoch": 0.38131797824696095, "grad_norm": 2.603180867944804, "learning_rate": 1.999653744698361e-05, "loss": 1.0639, "step": 596 }, { "epoch": 0.381957773512476, "grad_norm": 2.766706144230805, "learning_rate": 1.9996482707120287e-05, "loss": 1.1186, "step": 597 }, { "epoch": 0.38259756877799106, "grad_norm": 2.867905581526844, "learning_rate": 1.9996427538025775e-05, "loss": 1.0787, "step": 598 }, { "epoch": 0.38323736404350606, "grad_norm": 2.763135783484398, "learning_rate": 1.9996371939702442e-05, "loss": 1.0631, "step": 599 }, { "epoch": 0.3838771593090211, "grad_norm": 2.395526090482126, "learning_rate": 1.999631591215269e-05, "loss": 1.0962, "step": 600 }, { "epoch": 0.38451695457453616, "grad_norm": 3.5259194192787082, "learning_rate": 1.9996259455378907e-05, "loss": 1.0564, "step": 601 }, { "epoch": 0.38515674984005116, "grad_norm": 2.8996466784270014, "learning_rate": 1.999620256938353e-05, "loss": 1.0133, "step": 602 }, { "epoch": 0.3857965451055662, "grad_norm": 2.5950322273903343, "learning_rate": 1.999614525416899e-05, "loss": 1.0589, "step": 603 }, { "epoch": 0.38643634037108127, "grad_norm": 2.9014764562850797, "learning_rate": 1.9996087509737757e-05, "loss": 1.0923, "step": 604 }, { "epoch": 0.38707613563659626, "grad_norm": 2.7969257229230653, "learning_rate": 1.9996029336092308e-05, "loss": 1.0218, "step": 605 }, { "epoch": 0.3877159309021113, "grad_norm": 2.376464316810787, "learning_rate": 1.9995970733235142e-05, "loss": 1.1024, "step": 606 }, { "epoch": 0.38835572616762637, "grad_norm": 2.8736960487142666, "learning_rate": 1.9995911701168774e-05, "loss": 1.0734, "step": 607 }, { "epoch": 0.3889955214331414, "grad_norm": 3.10090595311659, "learning_rate": 1.999585223989574e-05, "loss": 1.0637, "step": 608 }, { "epoch": 0.3896353166986564, "grad_norm": 3.7041337840488064, "learning_rate": 1.999579234941859e-05, "loss": 1.048, "step": 609 }, { "epoch": 0.3902751119641715, "grad_norm": 2.8183189357213814, "learning_rate": 1.99957320297399e-05, "loss": 1.0882, "step": 610 }, { "epoch": 0.3909149072296865, "grad_norm": 2.6717605631889016, "learning_rate": 1.9995671280862257e-05, "loss": 1.1295, "step": 611 }, { "epoch": 0.3915547024952015, "grad_norm": 3.8004164715012085, "learning_rate": 1.9995610102788268e-05, "loss": 1.0419, "step": 612 }, { "epoch": 0.3921944977607166, "grad_norm": 3.1603944912669433, "learning_rate": 1.9995548495520566e-05, "loss": 1.0922, "step": 613 }, { "epoch": 0.39283429302623163, "grad_norm": 3.2107595942928793, "learning_rate": 1.9995486459061788e-05, "loss": 1.1025, "step": 614 }, { "epoch": 0.3934740882917466, "grad_norm": 2.8909536839266043, "learning_rate": 1.9995423993414604e-05, "loss": 1.064, "step": 615 }, { "epoch": 0.3941138835572617, "grad_norm": 3.186793515208253, "learning_rate": 1.9995361098581694e-05, "loss": 1.1262, "step": 616 }, { "epoch": 0.39475367882277673, "grad_norm": 2.6156385685529644, "learning_rate": 1.999529777456576e-05, "loss": 1.0186, "step": 617 }, { "epoch": 0.39539347408829173, "grad_norm": 2.874935738457726, "learning_rate": 1.9995234021369517e-05, "loss": 1.0218, "step": 618 }, { "epoch": 0.3960332693538068, "grad_norm": 3.176324426147613, "learning_rate": 1.9995169838995706e-05, "loss": 1.0705, "step": 619 }, { "epoch": 0.39667306461932184, "grad_norm": 3.0975431483592657, "learning_rate": 1.9995105227447084e-05, "loss": 1.0619, "step": 620 }, { "epoch": 0.39731285988483683, "grad_norm": 3.002081493128016, "learning_rate": 1.999504018672642e-05, "loss": 1.1211, "step": 621 }, { "epoch": 0.3979526551503519, "grad_norm": 2.8283886850065607, "learning_rate": 1.9994974716836516e-05, "loss": 1.0281, "step": 622 }, { "epoch": 0.39859245041586694, "grad_norm": 3.4361911230808277, "learning_rate": 1.9994908817780176e-05, "loss": 1.0255, "step": 623 }, { "epoch": 0.39923224568138194, "grad_norm": 3.1891555957121493, "learning_rate": 1.9994842489560227e-05, "loss": 1.0038, "step": 624 }, { "epoch": 0.399872040946897, "grad_norm": 2.544316491695822, "learning_rate": 1.999477573217952e-05, "loss": 1.0692, "step": 625 }, { "epoch": 0.40051183621241204, "grad_norm": 2.6508019289455094, "learning_rate": 1.9994708545640927e-05, "loss": 1.001, "step": 626 }, { "epoch": 0.40115163147792704, "grad_norm": 3.139246982959234, "learning_rate": 1.9994640929947328e-05, "loss": 1.0812, "step": 627 }, { "epoch": 0.4017914267434421, "grad_norm": 2.3512794418072414, "learning_rate": 1.9994572885101623e-05, "loss": 1.0075, "step": 628 }, { "epoch": 0.40243122200895715, "grad_norm": 2.826910137976899, "learning_rate": 1.999450441110674e-05, "loss": 1.0935, "step": 629 }, { "epoch": 0.40307101727447214, "grad_norm": 3.437743813786135, "learning_rate": 1.9994435507965616e-05, "loss": 1.0916, "step": 630 }, { "epoch": 0.4037108125399872, "grad_norm": 2.911319865371834, "learning_rate": 1.999436617568121e-05, "loss": 0.9759, "step": 631 }, { "epoch": 0.40435060780550225, "grad_norm": 3.66181295983702, "learning_rate": 1.9994296414256498e-05, "loss": 1.0645, "step": 632 }, { "epoch": 0.4049904030710173, "grad_norm": 3.0786093282667832, "learning_rate": 1.9994226223694476e-05, "loss": 1.1239, "step": 633 }, { "epoch": 0.4056301983365323, "grad_norm": 3.681817429989944, "learning_rate": 1.999415560399816e-05, "loss": 1.0592, "step": 634 }, { "epoch": 0.40626999360204735, "grad_norm": 3.185901348890919, "learning_rate": 1.9994084555170583e-05, "loss": 1.0357, "step": 635 }, { "epoch": 0.4069097888675624, "grad_norm": 2.852465646577552, "learning_rate": 1.999401307721479e-05, "loss": 1.0866, "step": 636 }, { "epoch": 0.4075495841330774, "grad_norm": 3.409832878645687, "learning_rate": 1.9993941170133856e-05, "loss": 1.0672, "step": 637 }, { "epoch": 0.40818937939859246, "grad_norm": 3.5332970107144046, "learning_rate": 1.9993868833930863e-05, "loss": 1.0622, "step": 638 }, { "epoch": 0.4088291746641075, "grad_norm": 2.8138866402285805, "learning_rate": 1.999379606860892e-05, "loss": 1.0945, "step": 639 }, { "epoch": 0.4094689699296225, "grad_norm": 2.9552917457752383, "learning_rate": 1.9993722874171153e-05, "loss": 1.0661, "step": 640 }, { "epoch": 0.41010876519513756, "grad_norm": 2.9654874668492845, "learning_rate": 1.99936492506207e-05, "loss": 1.0439, "step": 641 }, { "epoch": 0.4107485604606526, "grad_norm": 3.725199001217622, "learning_rate": 1.9993575197960727e-05, "loss": 1.0863, "step": 642 }, { "epoch": 0.4113883557261676, "grad_norm": 2.9163865140274763, "learning_rate": 1.9993500716194415e-05, "loss": 0.9716, "step": 643 }, { "epoch": 0.41202815099168266, "grad_norm": 3.2112571335338727, "learning_rate": 1.9993425805324956e-05, "loss": 1.1298, "step": 644 }, { "epoch": 0.4126679462571977, "grad_norm": 3.508565674873123, "learning_rate": 1.9993350465355572e-05, "loss": 1.051, "step": 645 }, { "epoch": 0.4133077415227127, "grad_norm": 4.134190703417477, "learning_rate": 1.9993274696289498e-05, "loss": 1.028, "step": 646 }, { "epoch": 0.41394753678822777, "grad_norm": 3.5474481671175524, "learning_rate": 1.999319849812998e-05, "loss": 0.9675, "step": 647 }, { "epoch": 0.4145873320537428, "grad_norm": 2.925270170451526, "learning_rate": 1.99931218708803e-05, "loss": 1.0643, "step": 648 }, { "epoch": 0.4152271273192578, "grad_norm": 3.8054888929888477, "learning_rate": 1.999304481454374e-05, "loss": 1.0511, "step": 649 }, { "epoch": 0.41586692258477287, "grad_norm": 3.5476023635828917, "learning_rate": 1.9992967329123613e-05, "loss": 1.0666, "step": 650 }, { "epoch": 0.4165067178502879, "grad_norm": 2.975997349224896, "learning_rate": 1.9992889414623244e-05, "loss": 1.0457, "step": 651 }, { "epoch": 0.4171465131158029, "grad_norm": 2.477210543190192, "learning_rate": 1.999281107104598e-05, "loss": 1.014, "step": 652 }, { "epoch": 0.417786308381318, "grad_norm": 3.535358249191018, "learning_rate": 1.9992732298395182e-05, "loss": 1.0632, "step": 653 }, { "epoch": 0.418426103646833, "grad_norm": 3.035052743703618, "learning_rate": 1.9992653096674236e-05, "loss": 1.0235, "step": 654 }, { "epoch": 0.419065898912348, "grad_norm": 3.0803903209257926, "learning_rate": 1.9992573465886544e-05, "loss": 1.0324, "step": 655 }, { "epoch": 0.4197056941778631, "grad_norm": 2.6432491928128186, "learning_rate": 1.9992493406035522e-05, "loss": 1.0769, "step": 656 }, { "epoch": 0.42034548944337813, "grad_norm": 2.9526580286857023, "learning_rate": 1.999241291712461e-05, "loss": 1.0386, "step": 657 }, { "epoch": 0.4209852847088932, "grad_norm": 2.8788645253465104, "learning_rate": 1.999233199915726e-05, "loss": 1.1034, "step": 658 }, { "epoch": 0.4216250799744082, "grad_norm": 3.1583313589940687, "learning_rate": 1.999225065213695e-05, "loss": 1.02, "step": 659 }, { "epoch": 0.42226487523992323, "grad_norm": 2.7084246844936724, "learning_rate": 1.9992168876067167e-05, "loss": 1.0525, "step": 660 }, { "epoch": 0.4229046705054383, "grad_norm": 3.206250810566487, "learning_rate": 1.9992086670951436e-05, "loss": 1.0143, "step": 661 }, { "epoch": 0.4235444657709533, "grad_norm": 2.775480687743589, "learning_rate": 1.999200403679327e-05, "loss": 1.0833, "step": 662 }, { "epoch": 0.42418426103646834, "grad_norm": 2.723314464622258, "learning_rate": 1.999192097359623e-05, "loss": 1.0755, "step": 663 }, { "epoch": 0.4248240563019834, "grad_norm": 2.7777611788166383, "learning_rate": 1.9991837481363876e-05, "loss": 1.0414, "step": 664 }, { "epoch": 0.4254638515674984, "grad_norm": 3.2156892547070037, "learning_rate": 1.9991753560099796e-05, "loss": 1.0317, "step": 665 }, { "epoch": 0.42610364683301344, "grad_norm": 3.028339175991124, "learning_rate": 1.999166920980759e-05, "loss": 1.0722, "step": 666 }, { "epoch": 0.4267434420985285, "grad_norm": 3.1010843827951033, "learning_rate": 1.999158443049088e-05, "loss": 1.0401, "step": 667 }, { "epoch": 0.4273832373640435, "grad_norm": 2.9863328380153558, "learning_rate": 1.9991499222153314e-05, "loss": 1.0842, "step": 668 }, { "epoch": 0.42802303262955854, "grad_norm": 2.7686945547305686, "learning_rate": 1.9991413584798542e-05, "loss": 1.0806, "step": 669 }, { "epoch": 0.4286628278950736, "grad_norm": 2.942252574169897, "learning_rate": 1.9991327518430242e-05, "loss": 1.0782, "step": 670 }, { "epoch": 0.4293026231605886, "grad_norm": 3.1312471385138116, "learning_rate": 1.999124102305211e-05, "loss": 1.0548, "step": 671 }, { "epoch": 0.42994241842610365, "grad_norm": 2.6736874132809345, "learning_rate": 1.999115409866787e-05, "loss": 1.1374, "step": 672 }, { "epoch": 0.4305822136916187, "grad_norm": 2.692726587293744, "learning_rate": 1.999106674528124e-05, "loss": 1.0522, "step": 673 }, { "epoch": 0.4312220089571337, "grad_norm": 3.3160035275230415, "learning_rate": 1.9990978962895977e-05, "loss": 1.0186, "step": 674 }, { "epoch": 0.43186180422264875, "grad_norm": 3.492085880933579, "learning_rate": 1.999089075151585e-05, "loss": 1.0775, "step": 675 }, { "epoch": 0.4325015994881638, "grad_norm": 3.1656690315163925, "learning_rate": 1.999080211114465e-05, "loss": 1.0622, "step": 676 }, { "epoch": 0.4331413947536788, "grad_norm": 2.9650975113773117, "learning_rate": 1.9990713041786175e-05, "loss": 1.0615, "step": 677 }, { "epoch": 0.43378119001919385, "grad_norm": 3.1291121902500874, "learning_rate": 1.9990623543444254e-05, "loss": 1.0215, "step": 678 }, { "epoch": 0.4344209852847089, "grad_norm": 2.9010810817365713, "learning_rate": 1.999053361612273e-05, "loss": 1.0323, "step": 679 }, { "epoch": 0.4350607805502239, "grad_norm": 2.9636667205315295, "learning_rate": 1.9990443259825468e-05, "loss": 1.0216, "step": 680 }, { "epoch": 0.43570057581573896, "grad_norm": 2.827828105856502, "learning_rate": 1.999035247455634e-05, "loss": 1.1184, "step": 681 }, { "epoch": 0.436340371081254, "grad_norm": 2.9763670118384757, "learning_rate": 1.9990261260319253e-05, "loss": 1.0607, "step": 682 }, { "epoch": 0.43698016634676906, "grad_norm": 3.481262079508208, "learning_rate": 1.9990169617118114e-05, "loss": 1.0662, "step": 683 }, { "epoch": 0.43761996161228406, "grad_norm": 2.5076353223261565, "learning_rate": 1.9990077544956865e-05, "loss": 1.0778, "step": 684 }, { "epoch": 0.4382597568777991, "grad_norm": 3.792423328279091, "learning_rate": 1.9989985043839456e-05, "loss": 1.0865, "step": 685 }, { "epoch": 0.43889955214331416, "grad_norm": 3.1210798568667153, "learning_rate": 1.9989892113769862e-05, "loss": 0.9935, "step": 686 }, { "epoch": 0.43953934740882916, "grad_norm": 3.1004106002519793, "learning_rate": 1.998979875475207e-05, "loss": 1.075, "step": 687 }, { "epoch": 0.4401791426743442, "grad_norm": 3.0671442473405413, "learning_rate": 1.998970496679009e-05, "loss": 1.0127, "step": 688 }, { "epoch": 0.44081893793985927, "grad_norm": 3.690652961450435, "learning_rate": 1.9989610749887948e-05, "loss": 1.077, "step": 689 }, { "epoch": 0.44145873320537427, "grad_norm": 3.676102441639255, "learning_rate": 1.9989516104049693e-05, "loss": 1.0805, "step": 690 }, { "epoch": 0.4420985284708893, "grad_norm": 2.9178422382267097, "learning_rate": 1.9989421029279384e-05, "loss": 1.0309, "step": 691 }, { "epoch": 0.44273832373640437, "grad_norm": 2.989573942477295, "learning_rate": 1.9989325525581108e-05, "loss": 1.0436, "step": 692 }, { "epoch": 0.44337811900191937, "grad_norm": 2.931227501946583, "learning_rate": 1.9989229592958965e-05, "loss": 1.0704, "step": 693 }, { "epoch": 0.4440179142674344, "grad_norm": 3.6671928758473875, "learning_rate": 1.9989133231417067e-05, "loss": 1.0722, "step": 694 }, { "epoch": 0.4446577095329495, "grad_norm": 2.9863495600535233, "learning_rate": 1.9989036440959562e-05, "loss": 1.0122, "step": 695 }, { "epoch": 0.44529750479846447, "grad_norm": 2.9509270248607153, "learning_rate": 1.99889392215906e-05, "loss": 1.0673, "step": 696 }, { "epoch": 0.4459373000639795, "grad_norm": 2.618156405202882, "learning_rate": 1.9988841573314355e-05, "loss": 1.0506, "step": 697 }, { "epoch": 0.4465770953294946, "grad_norm": 3.1421082472116386, "learning_rate": 1.9988743496135025e-05, "loss": 1.0124, "step": 698 }, { "epoch": 0.4472168905950096, "grad_norm": 2.6673216506211026, "learning_rate": 1.9988644990056816e-05, "loss": 1.0437, "step": 699 }, { "epoch": 0.44785668586052463, "grad_norm": 2.6064725366336554, "learning_rate": 1.9988546055083958e-05, "loss": 1.0326, "step": 700 }, { "epoch": 0.4484964811260397, "grad_norm": 3.0118880618832002, "learning_rate": 1.9988446691220704e-05, "loss": 1.0575, "step": 701 }, { "epoch": 0.4491362763915547, "grad_norm": 2.727147888177908, "learning_rate": 1.9988346898471316e-05, "loss": 1.0819, "step": 702 }, { "epoch": 0.44977607165706973, "grad_norm": 2.99291782097155, "learning_rate": 1.9988246676840076e-05, "loss": 1.0471, "step": 703 }, { "epoch": 0.4504158669225848, "grad_norm": 2.664660654754203, "learning_rate": 1.9988146026331295e-05, "loss": 1.0424, "step": 704 }, { "epoch": 0.4510556621880998, "grad_norm": 2.5630374089519226, "learning_rate": 1.9988044946949295e-05, "loss": 1.0698, "step": 705 }, { "epoch": 0.45169545745361483, "grad_norm": 3.5436941857839894, "learning_rate": 1.9987943438698407e-05, "loss": 1.016, "step": 706 }, { "epoch": 0.4523352527191299, "grad_norm": 3.3252472144247234, "learning_rate": 1.9987841501582995e-05, "loss": 1.0204, "step": 707 }, { "epoch": 0.45297504798464494, "grad_norm": 2.5494804886168136, "learning_rate": 1.998773913560744e-05, "loss": 1.0742, "step": 708 }, { "epoch": 0.45361484325015994, "grad_norm": 3.03362913884488, "learning_rate": 1.998763634077613e-05, "loss": 1.0705, "step": 709 }, { "epoch": 0.454254638515675, "grad_norm": 2.735137761263512, "learning_rate": 1.9987533117093484e-05, "loss": 1.0094, "step": 710 }, { "epoch": 0.45489443378119004, "grad_norm": 2.7216374039406945, "learning_rate": 1.998742946456393e-05, "loss": 1.0701, "step": 711 }, { "epoch": 0.45553422904670504, "grad_norm": 2.5851751196433344, "learning_rate": 1.998732538319192e-05, "loss": 1.0381, "step": 712 }, { "epoch": 0.4561740243122201, "grad_norm": 3.236566801543021, "learning_rate": 1.9987220872981926e-05, "loss": 1.0071, "step": 713 }, { "epoch": 0.45681381957773515, "grad_norm": 2.5742566338752857, "learning_rate": 1.9987115933938438e-05, "loss": 1.0881, "step": 714 }, { "epoch": 0.45745361484325014, "grad_norm": 2.7347399887727613, "learning_rate": 1.9987010566065953e-05, "loss": 1.0684, "step": 715 }, { "epoch": 0.4580934101087652, "grad_norm": 2.8332299917766384, "learning_rate": 1.9986904769369e-05, "loss": 1.0817, "step": 716 }, { "epoch": 0.45873320537428025, "grad_norm": 2.9038467313517664, "learning_rate": 1.998679854385212e-05, "loss": 0.9941, "step": 717 }, { "epoch": 0.45937300063979525, "grad_norm": 2.611713675419297, "learning_rate": 1.9986691889519878e-05, "loss": 0.9931, "step": 718 }, { "epoch": 0.4600127959053103, "grad_norm": 2.8371600035312268, "learning_rate": 1.9986584806376852e-05, "loss": 1.0716, "step": 719 }, { "epoch": 0.46065259117082535, "grad_norm": 2.3017675754130122, "learning_rate": 1.998647729442764e-05, "loss": 1.076, "step": 720 }, { "epoch": 0.46129238643634035, "grad_norm": 2.604204565770632, "learning_rate": 1.998636935367685e-05, "loss": 1.1097, "step": 721 }, { "epoch": 0.4619321817018554, "grad_norm": 2.7797399483587393, "learning_rate": 1.9986260984129132e-05, "loss": 1.0125, "step": 722 }, { "epoch": 0.46257197696737046, "grad_norm": 3.088835868969009, "learning_rate": 1.998615218578913e-05, "loss": 1.0901, "step": 723 }, { "epoch": 0.46321177223288545, "grad_norm": 2.4602001323919898, "learning_rate": 1.9986042958661515e-05, "loss": 1.0234, "step": 724 }, { "epoch": 0.4638515674984005, "grad_norm": 2.614688592517187, "learning_rate": 1.9985933302750983e-05, "loss": 1.0209, "step": 725 }, { "epoch": 0.46449136276391556, "grad_norm": 2.993273386209109, "learning_rate": 1.9985823218062236e-05, "loss": 1.0547, "step": 726 }, { "epoch": 0.46513115802943056, "grad_norm": 3.138113014499867, "learning_rate": 1.99857127046e-05, "loss": 1.0231, "step": 727 }, { "epoch": 0.4657709532949456, "grad_norm": 2.9131186420277175, "learning_rate": 1.998560176236903e-05, "loss": 1.0518, "step": 728 }, { "epoch": 0.46641074856046066, "grad_norm": 2.9833453938366543, "learning_rate": 1.9985490391374077e-05, "loss": 1.0495, "step": 729 }, { "epoch": 0.46705054382597566, "grad_norm": 2.7801329080372934, "learning_rate": 1.9985378591619933e-05, "loss": 1.0651, "step": 730 }, { "epoch": 0.4676903390914907, "grad_norm": 4.16944385477351, "learning_rate": 1.9985266363111397e-05, "loss": 1.0859, "step": 731 }, { "epoch": 0.46833013435700577, "grad_norm": 2.8281013221261713, "learning_rate": 1.998515370585328e-05, "loss": 1.045, "step": 732 }, { "epoch": 0.4689699296225208, "grad_norm": 2.668952486839899, "learning_rate": 1.9985040619850432e-05, "loss": 1.0741, "step": 733 }, { "epoch": 0.4696097248880358, "grad_norm": 3.074892571177553, "learning_rate": 1.9984927105107698e-05, "loss": 1.0119, "step": 734 }, { "epoch": 0.47024952015355087, "grad_norm": 3.5329252065495242, "learning_rate": 1.9984813161629957e-05, "loss": 1.096, "step": 735 }, { "epoch": 0.4708893154190659, "grad_norm": 3.1809838974015845, "learning_rate": 1.99846987894221e-05, "loss": 0.9944, "step": 736 }, { "epoch": 0.4715291106845809, "grad_norm": 3.1263095295137218, "learning_rate": 1.9984583988489037e-05, "loss": 1.0753, "step": 737 }, { "epoch": 0.472168905950096, "grad_norm": 2.978033869279466, "learning_rate": 1.99844687588357e-05, "loss": 1.1123, "step": 738 }, { "epoch": 0.472808701215611, "grad_norm": 3.0320032188048924, "learning_rate": 1.9984353100467035e-05, "loss": 1.0576, "step": 739 }, { "epoch": 0.473448496481126, "grad_norm": 2.618885765082849, "learning_rate": 1.998423701338801e-05, "loss": 1.0487, "step": 740 }, { "epoch": 0.4740882917466411, "grad_norm": 2.9653292150950734, "learning_rate": 1.998412049760361e-05, "loss": 1.0578, "step": 741 }, { "epoch": 0.47472808701215613, "grad_norm": 2.9750093006156293, "learning_rate": 1.998400355311883e-05, "loss": 1.0396, "step": 742 }, { "epoch": 0.4753678822776711, "grad_norm": 3.1795232689388793, "learning_rate": 1.9983886179938704e-05, "loss": 1.0202, "step": 743 }, { "epoch": 0.4760076775431862, "grad_norm": 2.7238567051120817, "learning_rate": 1.998376837806826e-05, "loss": 1.0656, "step": 744 }, { "epoch": 0.47664747280870123, "grad_norm": 3.0275273974733756, "learning_rate": 1.9983650147512566e-05, "loss": 1.011, "step": 745 }, { "epoch": 0.47728726807421623, "grad_norm": 3.152464106043066, "learning_rate": 1.998353148827669e-05, "loss": 1.0528, "step": 746 }, { "epoch": 0.4779270633397313, "grad_norm": 2.787950233455355, "learning_rate": 1.9983412400365734e-05, "loss": 1.0686, "step": 747 }, { "epoch": 0.47856685860524634, "grad_norm": 2.822501304605732, "learning_rate": 1.998329288378481e-05, "loss": 1.0534, "step": 748 }, { "epoch": 0.47920665387076133, "grad_norm": 2.796663741035814, "learning_rate": 1.9983172938539047e-05, "loss": 1.0992, "step": 749 }, { "epoch": 0.4798464491362764, "grad_norm": 2.8043126261841964, "learning_rate": 1.9983052564633593e-05, "loss": 1.1015, "step": 750 }, { "epoch": 0.48048624440179144, "grad_norm": 3.1101981160354497, "learning_rate": 1.9982931762073624e-05, "loss": 1.1122, "step": 751 }, { "epoch": 0.48112603966730644, "grad_norm": 2.93415247181202, "learning_rate": 1.9982810530864322e-05, "loss": 1.0397, "step": 752 }, { "epoch": 0.4817658349328215, "grad_norm": 2.896311789453156, "learning_rate": 1.9982688871010896e-05, "loss": 1.0743, "step": 753 }, { "epoch": 0.48240563019833654, "grad_norm": 2.4271041947356, "learning_rate": 1.9982566782518565e-05, "loss": 1.0646, "step": 754 }, { "epoch": 0.48304542546385154, "grad_norm": 2.4114087428810294, "learning_rate": 1.9982444265392577e-05, "loss": 1.0752, "step": 755 }, { "epoch": 0.4836852207293666, "grad_norm": 2.778840439663928, "learning_rate": 1.998232131963819e-05, "loss": 1.0961, "step": 756 }, { "epoch": 0.48432501599488165, "grad_norm": 3.3407545798181144, "learning_rate": 1.998219794526068e-05, "loss": 1.0617, "step": 757 }, { "epoch": 0.4849648112603967, "grad_norm": 2.719385199999821, "learning_rate": 1.998207414226535e-05, "loss": 1.0333, "step": 758 }, { "epoch": 0.4856046065259117, "grad_norm": 3.667796743609917, "learning_rate": 1.998194991065751e-05, "loss": 1.0378, "step": 759 }, { "epoch": 0.48624440179142675, "grad_norm": 3.978187117570256, "learning_rate": 1.9981825250442497e-05, "loss": 1.0523, "step": 760 }, { "epoch": 0.4868841970569418, "grad_norm": 2.2565124969637895, "learning_rate": 1.9981700161625666e-05, "loss": 0.9906, "step": 761 }, { "epoch": 0.4875239923224568, "grad_norm": 2.6630042076280076, "learning_rate": 1.9981574644212386e-05, "loss": 1.0156, "step": 762 }, { "epoch": 0.48816378758797185, "grad_norm": 3.155955507635851, "learning_rate": 1.9981448698208045e-05, "loss": 1.0297, "step": 763 }, { "epoch": 0.4888035828534869, "grad_norm": 3.839590942972022, "learning_rate": 1.9981322323618056e-05, "loss": 1.0406, "step": 764 }, { "epoch": 0.4894433781190019, "grad_norm": 2.9545969428485366, "learning_rate": 1.9981195520447838e-05, "loss": 1.0327, "step": 765 }, { "epoch": 0.49008317338451696, "grad_norm": 2.608078125108473, "learning_rate": 1.9981068288702842e-05, "loss": 1.0414, "step": 766 }, { "epoch": 0.490722968650032, "grad_norm": 2.3403389210239993, "learning_rate": 1.9980940628388526e-05, "loss": 1.048, "step": 767 }, { "epoch": 0.491362763915547, "grad_norm": 2.584450334600698, "learning_rate": 1.9980812539510378e-05, "loss": 1.0865, "step": 768 }, { "epoch": 0.49200255918106206, "grad_norm": 2.552759659068642, "learning_rate": 1.998068402207389e-05, "loss": 1.0328, "step": 769 }, { "epoch": 0.4926423544465771, "grad_norm": 2.88038160056437, "learning_rate": 1.9980555076084585e-05, "loss": 1.0298, "step": 770 }, { "epoch": 0.4932821497120921, "grad_norm": 2.9686432859755785, "learning_rate": 1.9980425701548002e-05, "loss": 1.0201, "step": 771 }, { "epoch": 0.49392194497760716, "grad_norm": 2.736316400060266, "learning_rate": 1.998029589846969e-05, "loss": 1.0538, "step": 772 }, { "epoch": 0.4945617402431222, "grad_norm": 2.9024251633762406, "learning_rate": 1.998016566685523e-05, "loss": 1.0923, "step": 773 }, { "epoch": 0.4952015355086372, "grad_norm": 2.8665044168941503, "learning_rate": 1.9980035006710206e-05, "loss": 1.0372, "step": 774 }, { "epoch": 0.49584133077415227, "grad_norm": 2.926390863154173, "learning_rate": 1.9979903918040234e-05, "loss": 1.0437, "step": 775 }, { "epoch": 0.4964811260396673, "grad_norm": 2.8556425235601486, "learning_rate": 1.9979772400850937e-05, "loss": 1.0565, "step": 776 }, { "epoch": 0.4971209213051823, "grad_norm": 2.922855774771315, "learning_rate": 1.997964045514797e-05, "loss": 1.0345, "step": 777 }, { "epoch": 0.49776071657069737, "grad_norm": 2.64950628918135, "learning_rate": 1.9979508080936992e-05, "loss": 1.0615, "step": 778 }, { "epoch": 0.4984005118362124, "grad_norm": 2.8476361122289653, "learning_rate": 1.9979375278223688e-05, "loss": 1.0831, "step": 779 }, { "epoch": 0.4990403071017274, "grad_norm": 3.033697682526449, "learning_rate": 1.9979242047013762e-05, "loss": 1.0051, "step": 780 }, { "epoch": 0.4996801023672425, "grad_norm": 3.5417782753875318, "learning_rate": 1.997910838731294e-05, "loss": 1.0753, "step": 781 }, { "epoch": 0.5003198976327575, "grad_norm": 2.6081601158734258, "learning_rate": 1.9978974299126946e-05, "loss": 0.9988, "step": 782 }, { "epoch": 0.5009596928982726, "grad_norm": 2.43826180222426, "learning_rate": 1.997883978246155e-05, "loss": 1.0521, "step": 783 }, { "epoch": 0.5015994881637876, "grad_norm": 2.579947861391404, "learning_rate": 1.9978704837322528e-05, "loss": 0.9711, "step": 784 }, { "epoch": 0.5022392834293026, "grad_norm": 2.5322891965314094, "learning_rate": 1.9978569463715672e-05, "loss": 1.0291, "step": 785 }, { "epoch": 0.5028790786948176, "grad_norm": 2.6569025735311342, "learning_rate": 1.997843366164679e-05, "loss": 1.0065, "step": 786 }, { "epoch": 0.5035188739603327, "grad_norm": 2.7691006364940525, "learning_rate": 1.9978297431121718e-05, "loss": 1.0742, "step": 787 }, { "epoch": 0.5041586692258477, "grad_norm": 2.757615027141603, "learning_rate": 1.9978160772146302e-05, "loss": 1.0084, "step": 788 }, { "epoch": 0.5047984644913628, "grad_norm": 3.1887115187402078, "learning_rate": 1.9978023684726415e-05, "loss": 1.0682, "step": 789 }, { "epoch": 0.5054382597568778, "grad_norm": 2.661857569144561, "learning_rate": 1.997788616886794e-05, "loss": 0.9973, "step": 790 }, { "epoch": 0.5060780550223928, "grad_norm": 3.1790245198680243, "learning_rate": 1.9977748224576784e-05, "loss": 0.9915, "step": 791 }, { "epoch": 0.5067178502879078, "grad_norm": 2.6821020941758227, "learning_rate": 1.9977609851858865e-05, "loss": 1.0168, "step": 792 }, { "epoch": 0.5073576455534229, "grad_norm": 2.892157443271336, "learning_rate": 1.9977471050720127e-05, "loss": 0.9752, "step": 793 }, { "epoch": 0.5079974408189379, "grad_norm": 2.5543644967904116, "learning_rate": 1.9977331821166534e-05, "loss": 1.067, "step": 794 }, { "epoch": 0.508637236084453, "grad_norm": 3.0101570813005405, "learning_rate": 1.997719216320406e-05, "loss": 1.0305, "step": 795 }, { "epoch": 0.509277031349968, "grad_norm": 2.641193761670583, "learning_rate": 1.99770520768387e-05, "loss": 1.0249, "step": 796 }, { "epoch": 0.5099168266154831, "grad_norm": 2.542627729391096, "learning_rate": 1.9976911562076476e-05, "loss": 1.0842, "step": 797 }, { "epoch": 0.510556621880998, "grad_norm": 2.747517507364122, "learning_rate": 1.997677061892341e-05, "loss": 1.0463, "step": 798 }, { "epoch": 0.5111964171465131, "grad_norm": 2.6825537457350626, "learning_rate": 1.9976629247385568e-05, "loss": 1.0416, "step": 799 }, { "epoch": 0.5118362124120281, "grad_norm": 2.806052959543928, "learning_rate": 1.9976487447469012e-05, "loss": 1.0336, "step": 800 }, { "epoch": 0.5124760076775432, "grad_norm": 2.9409086872858428, "learning_rate": 1.9976345219179828e-05, "loss": 1.0039, "step": 801 }, { "epoch": 0.5131158029430583, "grad_norm": 2.547159529984535, "learning_rate": 1.997620256252413e-05, "loss": 0.9914, "step": 802 }, { "epoch": 0.5137555982085733, "grad_norm": 2.684462539162385, "learning_rate": 1.9976059477508042e-05, "loss": 0.971, "step": 803 }, { "epoch": 0.5143953934740882, "grad_norm": 2.748021524427877, "learning_rate": 1.9975915964137702e-05, "loss": 1.0405, "step": 804 }, { "epoch": 0.5150351887396033, "grad_norm": 2.8683346163435637, "learning_rate": 1.997577202241928e-05, "loss": 1.0436, "step": 805 }, { "epoch": 0.5156749840051184, "grad_norm": 2.4971754468988214, "learning_rate": 1.997562765235895e-05, "loss": 1.0316, "step": 806 }, { "epoch": 0.5163147792706334, "grad_norm": 2.508323979113117, "learning_rate": 1.9975482853962913e-05, "loss": 1.0274, "step": 807 }, { "epoch": 0.5169545745361485, "grad_norm": 2.917548980225321, "learning_rate": 1.997533762723739e-05, "loss": 0.9931, "step": 808 }, { "epoch": 0.5175943698016635, "grad_norm": 2.5625367451077814, "learning_rate": 1.9975191972188615e-05, "loss": 1.0814, "step": 809 }, { "epoch": 0.5182341650671785, "grad_norm": 2.4632277775768214, "learning_rate": 1.9975045888822836e-05, "loss": 1.0765, "step": 810 }, { "epoch": 0.5188739603326935, "grad_norm": 2.430425841150665, "learning_rate": 1.9974899377146334e-05, "loss": 1.0496, "step": 811 }, { "epoch": 0.5195137555982086, "grad_norm": 2.8147791145613623, "learning_rate": 1.9974752437165396e-05, "loss": 1.042, "step": 812 }, { "epoch": 0.5201535508637236, "grad_norm": 2.736794585840475, "learning_rate": 1.9974605068886334e-05, "loss": 1.0357, "step": 813 }, { "epoch": 0.5207933461292387, "grad_norm": 2.698393303682229, "learning_rate": 1.9974457272315474e-05, "loss": 1.0534, "step": 814 }, { "epoch": 0.5214331413947537, "grad_norm": 2.9051573689745998, "learning_rate": 1.997430904745916e-05, "loss": 1.0465, "step": 815 }, { "epoch": 0.5220729366602687, "grad_norm": 2.598044886410086, "learning_rate": 1.9974160394323757e-05, "loss": 0.9996, "step": 816 }, { "epoch": 0.5227127319257837, "grad_norm": 2.6470670727907626, "learning_rate": 1.9974011312915654e-05, "loss": 1.0556, "step": 817 }, { "epoch": 0.5233525271912988, "grad_norm": 2.541772167527474, "learning_rate": 1.9973861803241244e-05, "loss": 1.0032, "step": 818 }, { "epoch": 0.5239923224568138, "grad_norm": 2.331805794754917, "learning_rate": 1.997371186530695e-05, "loss": 1.0515, "step": 819 }, { "epoch": 0.5246321177223289, "grad_norm": 2.492282237478171, "learning_rate": 1.997356149911921e-05, "loss": 1.0642, "step": 820 }, { "epoch": 0.5252719129878439, "grad_norm": 2.8088573429331456, "learning_rate": 1.9973410704684485e-05, "loss": 1.0406, "step": 821 }, { "epoch": 0.525911708253359, "grad_norm": 2.7680093755284982, "learning_rate": 1.997325948200924e-05, "loss": 1.0384, "step": 822 }, { "epoch": 0.5265515035188739, "grad_norm": 3.0685256983907503, "learning_rate": 1.997310783109998e-05, "loss": 1.0102, "step": 823 }, { "epoch": 0.527191298784389, "grad_norm": 2.9351494742063444, "learning_rate": 1.9972955751963206e-05, "loss": 1.0422, "step": 824 }, { "epoch": 0.527831094049904, "grad_norm": 2.679734316306714, "learning_rate": 1.9972803244605454e-05, "loss": 1.036, "step": 825 }, { "epoch": 0.5284708893154191, "grad_norm": 2.6856083474854526, "learning_rate": 1.9972650309033272e-05, "loss": 1.0132, "step": 826 }, { "epoch": 0.5291106845809341, "grad_norm": 3.5201872910749867, "learning_rate": 1.9972496945253222e-05, "loss": 1.0399, "step": 827 }, { "epoch": 0.5297504798464492, "grad_norm": 2.9180654550990326, "learning_rate": 1.9972343153271895e-05, "loss": 1.0075, "step": 828 }, { "epoch": 0.5303902751119641, "grad_norm": 2.913765810868923, "learning_rate": 1.9972188933095895e-05, "loss": 1.0068, "step": 829 }, { "epoch": 0.5310300703774792, "grad_norm": 2.6466033391356367, "learning_rate": 1.997203428473184e-05, "loss": 0.9904, "step": 830 }, { "epoch": 0.5316698656429942, "grad_norm": 2.9731350960688636, "learning_rate": 1.997187920818637e-05, "loss": 1.0505, "step": 831 }, { "epoch": 0.5323096609085093, "grad_norm": 3.051304618832314, "learning_rate": 1.997172370346615e-05, "loss": 0.998, "step": 832 }, { "epoch": 0.5329494561740243, "grad_norm": 2.947769356914641, "learning_rate": 1.9971567770577846e-05, "loss": 1.0432, "step": 833 }, { "epoch": 0.5335892514395394, "grad_norm": 2.490167715830177, "learning_rate": 1.997141140952816e-05, "loss": 0.9963, "step": 834 }, { "epoch": 0.5342290467050543, "grad_norm": 2.503222612197282, "learning_rate": 1.997125462032381e-05, "loss": 0.9902, "step": 835 }, { "epoch": 0.5348688419705694, "grad_norm": 2.762173915502348, "learning_rate": 1.997109740297152e-05, "loss": 0.9921, "step": 836 }, { "epoch": 0.5355086372360844, "grad_norm": 2.7049361358605863, "learning_rate": 1.997093975747805e-05, "loss": 0.9949, "step": 837 }, { "epoch": 0.5361484325015995, "grad_norm": 3.057776960554831, "learning_rate": 1.9970781683850163e-05, "loss": 1.0294, "step": 838 }, { "epoch": 0.5367882277671145, "grad_norm": 2.866280357522378, "learning_rate": 1.9970623182094643e-05, "loss": 1.0126, "step": 839 }, { "epoch": 0.5374280230326296, "grad_norm": 2.614833708115078, "learning_rate": 1.9970464252218303e-05, "loss": 1.0239, "step": 840 }, { "epoch": 0.5380678182981445, "grad_norm": 2.5208795052987325, "learning_rate": 1.9970304894227965e-05, "loss": 1.0587, "step": 841 }, { "epoch": 0.5387076135636596, "grad_norm": 3.0023252965024345, "learning_rate": 1.9970145108130468e-05, "loss": 1.042, "step": 842 }, { "epoch": 0.5393474088291746, "grad_norm": 2.586781558101705, "learning_rate": 1.9969984893932673e-05, "loss": 1.0808, "step": 843 }, { "epoch": 0.5399872040946897, "grad_norm": 2.875813607515651, "learning_rate": 1.9969824251641466e-05, "loss": 1.0267, "step": 844 }, { "epoch": 0.5406269993602048, "grad_norm": 3.078506677508897, "learning_rate": 1.9969663181263745e-05, "loss": 1.0322, "step": 845 }, { "epoch": 0.5412667946257198, "grad_norm": 2.827359684949135, "learning_rate": 1.9969501682806413e-05, "loss": 1.0278, "step": 846 }, { "epoch": 0.5419065898912349, "grad_norm": 2.854993770798177, "learning_rate": 1.996933975627642e-05, "loss": 1.073, "step": 847 }, { "epoch": 0.5425463851567498, "grad_norm": 2.6993887380152684, "learning_rate": 1.9969177401680705e-05, "loss": 0.9559, "step": 848 }, { "epoch": 0.5431861804222649, "grad_norm": 2.9153379289648877, "learning_rate": 1.996901461902625e-05, "loss": 1.0191, "step": 849 }, { "epoch": 0.5438259756877799, "grad_norm": 3.4307573195509833, "learning_rate": 1.9968851408320042e-05, "loss": 1.0659, "step": 850 }, { "epoch": 0.544465770953295, "grad_norm": 3.260022709592201, "learning_rate": 1.9968687769569082e-05, "loss": 1.0083, "step": 851 }, { "epoch": 0.54510556621881, "grad_norm": 3.2837350544059185, "learning_rate": 1.996852370278041e-05, "loss": 1.0585, "step": 852 }, { "epoch": 0.5457453614843251, "grad_norm": 2.61506348206185, "learning_rate": 1.9968359207961057e-05, "loss": 1.0567, "step": 853 }, { "epoch": 0.54638515674984, "grad_norm": 2.9061124518127754, "learning_rate": 1.9968194285118097e-05, "loss": 1.0239, "step": 854 }, { "epoch": 0.5470249520153551, "grad_norm": 2.705302879346602, "learning_rate": 1.9968028934258604e-05, "loss": 1.0562, "step": 855 }, { "epoch": 0.5476647472808701, "grad_norm": 2.8938249313897146, "learning_rate": 1.9967863155389676e-05, "loss": 1.0507, "step": 856 }, { "epoch": 0.5483045425463852, "grad_norm": 3.4693850887282975, "learning_rate": 1.9967696948518442e-05, "loss": 1.0498, "step": 857 }, { "epoch": 0.5489443378119002, "grad_norm": 2.6718384519825267, "learning_rate": 1.996753031365203e-05, "loss": 1.0209, "step": 858 }, { "epoch": 0.5495841330774153, "grad_norm": 2.6597511790152883, "learning_rate": 1.9967363250797596e-05, "loss": 1.0288, "step": 859 }, { "epoch": 0.5502239283429302, "grad_norm": 3.1105977565513214, "learning_rate": 1.9967195759962318e-05, "loss": 1.0312, "step": 860 }, { "epoch": 0.5508637236084453, "grad_norm": 2.735725604893267, "learning_rate": 1.996702784115338e-05, "loss": 1.001, "step": 861 }, { "epoch": 0.5515035188739603, "grad_norm": 2.7696072644843563, "learning_rate": 1.9966859494377998e-05, "loss": 1.0632, "step": 862 }, { "epoch": 0.5521433141394754, "grad_norm": 2.827964169325998, "learning_rate": 1.99666907196434e-05, "loss": 1.0281, "step": 863 }, { "epoch": 0.5527831094049904, "grad_norm": 2.9142224763435722, "learning_rate": 1.996652151695683e-05, "loss": 1.1067, "step": 864 }, { "epoch": 0.5534229046705055, "grad_norm": 2.83665281268364, "learning_rate": 1.9966351886325557e-05, "loss": 0.9894, "step": 865 }, { "epoch": 0.5540626999360204, "grad_norm": 2.636403620600593, "learning_rate": 1.996618182775686e-05, "loss": 1.0587, "step": 866 }, { "epoch": 0.5547024952015355, "grad_norm": 2.6918924470823424, "learning_rate": 1.9966011341258046e-05, "loss": 1.0185, "step": 867 }, { "epoch": 0.5553422904670505, "grad_norm": 3.1355802341363184, "learning_rate": 1.9965840426836433e-05, "loss": 1.0443, "step": 868 }, { "epoch": 0.5559820857325656, "grad_norm": 2.700867740463888, "learning_rate": 1.9965669084499363e-05, "loss": 1.0008, "step": 869 }, { "epoch": 0.5566218809980806, "grad_norm": 2.5263997199057924, "learning_rate": 1.9965497314254186e-05, "loss": 1.0092, "step": 870 }, { "epoch": 0.5572616762635957, "grad_norm": 2.531549054769641, "learning_rate": 1.9965325116108283e-05, "loss": 1.0432, "step": 871 }, { "epoch": 0.5579014715291107, "grad_norm": 2.4164723190251887, "learning_rate": 1.9965152490069043e-05, "loss": 1.0479, "step": 872 }, { "epoch": 0.5585412667946257, "grad_norm": 3.111694462496266, "learning_rate": 1.9964979436143887e-05, "loss": 1.0286, "step": 873 }, { "epoch": 0.5591810620601407, "grad_norm": 2.6161489431763103, "learning_rate": 1.9964805954340237e-05, "loss": 1.0222, "step": 874 }, { "epoch": 0.5598208573256558, "grad_norm": 2.435270949193825, "learning_rate": 1.9964632044665548e-05, "loss": 1.0208, "step": 875 }, { "epoch": 0.5604606525911708, "grad_norm": 2.6674569935360215, "learning_rate": 1.996445770712728e-05, "loss": 0.9871, "step": 876 }, { "epoch": 0.5611004478566859, "grad_norm": 3.5374563765923925, "learning_rate": 1.996428294173293e-05, "loss": 1.0747, "step": 877 }, { "epoch": 0.5617402431222009, "grad_norm": 2.7224776611988895, "learning_rate": 1.9964107748489988e-05, "loss": 0.9911, "step": 878 }, { "epoch": 0.5623800383877159, "grad_norm": 2.8532857049993905, "learning_rate": 1.9963932127405988e-05, "loss": 1.0594, "step": 879 }, { "epoch": 0.5630198336532309, "grad_norm": 2.498505971214164, "learning_rate": 1.996375607848847e-05, "loss": 1.0142, "step": 880 }, { "epoch": 0.563659628918746, "grad_norm": 2.4319037061611133, "learning_rate": 1.9963579601744985e-05, "loss": 1.0019, "step": 881 }, { "epoch": 0.564299424184261, "grad_norm": 2.9520572231563373, "learning_rate": 1.996340269718312e-05, "loss": 1.0131, "step": 882 }, { "epoch": 0.5649392194497761, "grad_norm": 2.690044186852168, "learning_rate": 1.9963225364810464e-05, "loss": 1.0416, "step": 883 }, { "epoch": 0.5655790147152912, "grad_norm": 2.575226390571022, "learning_rate": 1.9963047604634632e-05, "loss": 0.9877, "step": 884 }, { "epoch": 0.5662188099808061, "grad_norm": 3.075407300077085, "learning_rate": 1.9962869416663266e-05, "loss": 0.9855, "step": 885 }, { "epoch": 0.5668586052463211, "grad_norm": 2.4033980067827763, "learning_rate": 1.9962690800904e-05, "loss": 1.0286, "step": 886 }, { "epoch": 0.5674984005118362, "grad_norm": 2.7179096269692535, "learning_rate": 1.9962511757364522e-05, "loss": 0.9828, "step": 887 }, { "epoch": 0.5681381957773513, "grad_norm": 3.020877304053594, "learning_rate": 1.9962332286052507e-05, "loss": 1.0322, "step": 888 }, { "epoch": 0.5687779910428663, "grad_norm": 2.5848578626055145, "learning_rate": 1.9962152386975668e-05, "loss": 1.0027, "step": 889 }, { "epoch": 0.5694177863083814, "grad_norm": 2.8225980503188364, "learning_rate": 1.996197206014172e-05, "loss": 1.0099, "step": 890 }, { "epoch": 0.5700575815738963, "grad_norm": 2.7898226063251794, "learning_rate": 1.996179130555842e-05, "loss": 0.9926, "step": 891 }, { "epoch": 0.5706973768394114, "grad_norm": 2.576372105565397, "learning_rate": 1.996161012323352e-05, "loss": 1.0294, "step": 892 }, { "epoch": 0.5713371721049264, "grad_norm": 2.7739503900976668, "learning_rate": 1.9961428513174798e-05, "loss": 1.0242, "step": 893 }, { "epoch": 0.5719769673704415, "grad_norm": 3.3265400795441082, "learning_rate": 1.996124647539006e-05, "loss": 1.0096, "step": 894 }, { "epoch": 0.5726167626359565, "grad_norm": 2.7087777843022955, "learning_rate": 1.9961064009887117e-05, "loss": 1.0178, "step": 895 }, { "epoch": 0.5732565579014716, "grad_norm": 2.768386979754229, "learning_rate": 1.9960881116673803e-05, "loss": 0.9891, "step": 896 }, { "epoch": 0.5738963531669866, "grad_norm": 3.299599700800376, "learning_rate": 1.9960697795757976e-05, "loss": 0.9798, "step": 897 }, { "epoch": 0.5745361484325016, "grad_norm": 2.685335001770179, "learning_rate": 1.9960514047147502e-05, "loss": 1.0036, "step": 898 }, { "epoch": 0.5751759436980166, "grad_norm": 2.357806625720523, "learning_rate": 1.9960329870850272e-05, "loss": 1.0296, "step": 899 }, { "epoch": 0.5758157389635317, "grad_norm": 2.6581912566273815, "learning_rate": 1.99601452668742e-05, "loss": 1.0384, "step": 900 }, { "epoch": 0.5764555342290467, "grad_norm": 2.926337525152396, "learning_rate": 1.99599602352272e-05, "loss": 0.9846, "step": 901 }, { "epoch": 0.5770953294945618, "grad_norm": 3.125878852658488, "learning_rate": 1.9959774775917234e-05, "loss": 1.0005, "step": 902 }, { "epoch": 0.5777351247600768, "grad_norm": 2.754166249739187, "learning_rate": 1.995958888895225e-05, "loss": 0.9881, "step": 903 }, { "epoch": 0.5783749200255918, "grad_norm": 2.8690421403872612, "learning_rate": 1.995940257434024e-05, "loss": 1.0016, "step": 904 }, { "epoch": 0.5790147152911068, "grad_norm": 2.612942760726621, "learning_rate": 1.9959215832089197e-05, "loss": 1.0288, "step": 905 }, { "epoch": 0.5796545105566219, "grad_norm": 2.45186984741498, "learning_rate": 1.9959028662207146e-05, "loss": 0.9929, "step": 906 }, { "epoch": 0.5802943058221369, "grad_norm": 2.658758056150747, "learning_rate": 1.9958841064702114e-05, "loss": 1.0301, "step": 907 }, { "epoch": 0.580934101087652, "grad_norm": 2.9917334514164518, "learning_rate": 1.995865303958217e-05, "loss": 0.9805, "step": 908 }, { "epoch": 0.581573896353167, "grad_norm": 2.4037572331971346, "learning_rate": 1.995846458685537e-05, "loss": 0.9912, "step": 909 }, { "epoch": 0.582213691618682, "grad_norm": 2.783078829718741, "learning_rate": 1.9958275706529822e-05, "loss": 0.9882, "step": 910 }, { "epoch": 0.582853486884197, "grad_norm": 2.913145235472154, "learning_rate": 1.9958086398613627e-05, "loss": 1.0987, "step": 911 }, { "epoch": 0.5834932821497121, "grad_norm": 2.6084081138496606, "learning_rate": 1.9957896663114918e-05, "loss": 1.0268, "step": 912 }, { "epoch": 0.5841330774152271, "grad_norm": 2.7459597921150705, "learning_rate": 1.9957706500041838e-05, "loss": 1.0824, "step": 913 }, { "epoch": 0.5847728726807422, "grad_norm": 2.9821834410923715, "learning_rate": 1.9957515909402552e-05, "loss": 1.0569, "step": 914 }, { "epoch": 0.5854126679462572, "grad_norm": 2.5535658325055812, "learning_rate": 1.9957324891205253e-05, "loss": 1.0262, "step": 915 }, { "epoch": 0.5860524632117722, "grad_norm": 2.893792304919255, "learning_rate": 1.9957133445458133e-05, "loss": 1.0563, "step": 916 }, { "epoch": 0.5866922584772872, "grad_norm": 2.8840821106915002, "learning_rate": 1.995694157216941e-05, "loss": 0.9608, "step": 917 }, { "epoch": 0.5873320537428023, "grad_norm": 2.6917035998108934, "learning_rate": 1.9956749271347332e-05, "loss": 1.0034, "step": 918 }, { "epoch": 0.5879718490083173, "grad_norm": 3.087494189771474, "learning_rate": 1.9956556543000155e-05, "loss": 1.0741, "step": 919 }, { "epoch": 0.5886116442738324, "grad_norm": 2.7507351438454593, "learning_rate": 1.9956363387136147e-05, "loss": 1.0092, "step": 920 }, { "epoch": 0.5892514395393474, "grad_norm": 2.5234336841023732, "learning_rate": 1.9956169803763606e-05, "loss": 1.0208, "step": 921 }, { "epoch": 0.5898912348048625, "grad_norm": 2.830396605386334, "learning_rate": 1.9955975792890847e-05, "loss": 1.065, "step": 922 }, { "epoch": 0.5905310300703774, "grad_norm": 2.788270970394877, "learning_rate": 1.9955781354526197e-05, "loss": 0.9712, "step": 923 }, { "epoch": 0.5911708253358925, "grad_norm": 2.8112164101591324, "learning_rate": 1.9955586488678004e-05, "loss": 1.0471, "step": 924 }, { "epoch": 0.5918106206014075, "grad_norm": 2.735012899957593, "learning_rate": 1.9955391195354638e-05, "loss": 0.9522, "step": 925 }, { "epoch": 0.5924504158669226, "grad_norm": 3.0371989376094137, "learning_rate": 1.9955195474564488e-05, "loss": 1.0576, "step": 926 }, { "epoch": 0.5930902111324377, "grad_norm": 2.4865885310897116, "learning_rate": 1.9954999326315945e-05, "loss": 0.9906, "step": 927 }, { "epoch": 0.5937300063979527, "grad_norm": 2.3807148391682715, "learning_rate": 1.9954802750617443e-05, "loss": 0.9632, "step": 928 }, { "epoch": 0.5943698016634676, "grad_norm": 3.016259595637002, "learning_rate": 1.9954605747477422e-05, "loss": 1.0512, "step": 929 }, { "epoch": 0.5950095969289827, "grad_norm": 2.413163496158654, "learning_rate": 1.995440831690433e-05, "loss": 1.0165, "step": 930 }, { "epoch": 0.5956493921944978, "grad_norm": 2.46615927636866, "learning_rate": 1.995421045890666e-05, "loss": 1.0009, "step": 931 }, { "epoch": 0.5962891874600128, "grad_norm": 2.2064531041762767, "learning_rate": 1.9954012173492895e-05, "loss": 1.0788, "step": 932 }, { "epoch": 0.5969289827255279, "grad_norm": 3.4246419856918053, "learning_rate": 1.995381346067156e-05, "loss": 1.0368, "step": 933 }, { "epoch": 0.5975687779910429, "grad_norm": 2.890465958441493, "learning_rate": 1.9953614320451178e-05, "loss": 1.0388, "step": 934 }, { "epoch": 0.5982085732565579, "grad_norm": 2.481257292764561, "learning_rate": 1.9953414752840305e-05, "loss": 0.9623, "step": 935 }, { "epoch": 0.5988483685220729, "grad_norm": 2.7092750113257313, "learning_rate": 1.9953214757847505e-05, "loss": 0.9883, "step": 936 }, { "epoch": 0.599488163787588, "grad_norm": 2.275311799439164, "learning_rate": 1.995301433548137e-05, "loss": 0.9714, "step": 937 }, { "epoch": 0.600127959053103, "grad_norm": 2.6020387637128715, "learning_rate": 1.9952813485750507e-05, "loss": 1.0, "step": 938 }, { "epoch": 0.6007677543186181, "grad_norm": 2.8168209177860724, "learning_rate": 1.9952612208663533e-05, "loss": 0.9778, "step": 939 }, { "epoch": 0.6014075495841331, "grad_norm": 2.6664199793577588, "learning_rate": 1.9952410504229098e-05, "loss": 1.0063, "step": 940 }, { "epoch": 0.6020473448496481, "grad_norm": 2.547335630262168, "learning_rate": 1.9952208372455858e-05, "loss": 0.9372, "step": 941 }, { "epoch": 0.6026871401151631, "grad_norm": 2.2830698148778255, "learning_rate": 1.9952005813352497e-05, "loss": 1.0251, "step": 942 }, { "epoch": 0.6033269353806782, "grad_norm": 3.092336641694271, "learning_rate": 1.9951802826927705e-05, "loss": 1.0144, "step": 943 }, { "epoch": 0.6039667306461932, "grad_norm": 3.332942589878281, "learning_rate": 1.9951599413190208e-05, "loss": 1.0515, "step": 944 }, { "epoch": 0.6046065259117083, "grad_norm": 2.8120098136569034, "learning_rate": 1.9951395572148733e-05, "loss": 1.0095, "step": 945 }, { "epoch": 0.6052463211772233, "grad_norm": 2.537568346885153, "learning_rate": 1.995119130381203e-05, "loss": 1.0434, "step": 946 }, { "epoch": 0.6058861164427384, "grad_norm": 2.6446570149846127, "learning_rate": 1.995098660818888e-05, "loss": 1.0308, "step": 947 }, { "epoch": 0.6065259117082533, "grad_norm": 3.088540273428461, "learning_rate": 1.9950781485288066e-05, "loss": 1.0463, "step": 948 }, { "epoch": 0.6071657069737684, "grad_norm": 2.8123184953869487, "learning_rate": 1.9950575935118394e-05, "loss": 0.9926, "step": 949 }, { "epoch": 0.6078055022392834, "grad_norm": 3.1432436339901413, "learning_rate": 1.995036995768869e-05, "loss": 0.9981, "step": 950 }, { "epoch": 0.6084452975047985, "grad_norm": 2.350104529005463, "learning_rate": 1.9950163553007807e-05, "loss": 1.0595, "step": 951 }, { "epoch": 0.6090850927703135, "grad_norm": 2.9004270039777955, "learning_rate": 1.9949956721084596e-05, "loss": 1.0327, "step": 952 }, { "epoch": 0.6097248880358286, "grad_norm": 2.6060395907103606, "learning_rate": 1.9949749461927944e-05, "loss": 1.0325, "step": 953 }, { "epoch": 0.6103646833013435, "grad_norm": 2.9030056201048553, "learning_rate": 1.994954177554675e-05, "loss": 1.0412, "step": 954 }, { "epoch": 0.6110044785668586, "grad_norm": 2.831468988681581, "learning_rate": 1.994933366194993e-05, "loss": 1.0072, "step": 955 }, { "epoch": 0.6116442738323736, "grad_norm": 2.6273787354494638, "learning_rate": 1.994912512114642e-05, "loss": 1.0258, "step": 956 }, { "epoch": 0.6122840690978887, "grad_norm": 2.6829472759595028, "learning_rate": 1.9948916153145178e-05, "loss": 0.9589, "step": 957 }, { "epoch": 0.6129238643634037, "grad_norm": 2.532974309903781, "learning_rate": 1.9948706757955172e-05, "loss": 0.9844, "step": 958 }, { "epoch": 0.6135636596289188, "grad_norm": 3.2367929312357315, "learning_rate": 1.9948496935585396e-05, "loss": 1.0995, "step": 959 }, { "epoch": 0.6142034548944337, "grad_norm": 2.683738543956663, "learning_rate": 1.9948286686044857e-05, "loss": 0.9469, "step": 960 }, { "epoch": 0.6148432501599488, "grad_norm": 3.1145650590591263, "learning_rate": 1.9948076009342588e-05, "loss": 0.9806, "step": 961 }, { "epoch": 0.6154830454254638, "grad_norm": 2.6472464700589087, "learning_rate": 1.9947864905487627e-05, "loss": 1.0342, "step": 962 }, { "epoch": 0.6161228406909789, "grad_norm": 2.493150786498269, "learning_rate": 1.9947653374489048e-05, "loss": 0.9775, "step": 963 }, { "epoch": 0.6167626359564939, "grad_norm": 2.646872697254269, "learning_rate": 1.9947441416355925e-05, "loss": 1.0183, "step": 964 }, { "epoch": 0.617402431222009, "grad_norm": 2.825631103061218, "learning_rate": 1.994722903109736e-05, "loss": 1.0774, "step": 965 }, { "epoch": 0.6180422264875239, "grad_norm": 3.232077021430892, "learning_rate": 1.9947016218722482e-05, "loss": 0.9904, "step": 966 }, { "epoch": 0.618682021753039, "grad_norm": 2.8497058289253325, "learning_rate": 1.9946802979240417e-05, "loss": 1.0136, "step": 967 }, { "epoch": 0.619321817018554, "grad_norm": 2.5339593227403063, "learning_rate": 1.9946589312660328e-05, "loss": 1.1219, "step": 968 }, { "epoch": 0.6199616122840691, "grad_norm": 3.036726985743462, "learning_rate": 1.9946375218991388e-05, "loss": 1.0036, "step": 969 }, { "epoch": 0.6206014075495841, "grad_norm": 3.347878760194625, "learning_rate": 1.9946160698242787e-05, "loss": 1.0071, "step": 970 }, { "epoch": 0.6212412028150992, "grad_norm": 2.652553969513173, "learning_rate": 1.9945945750423742e-05, "loss": 0.9916, "step": 971 }, { "epoch": 0.6218809980806143, "grad_norm": 2.5236250302671537, "learning_rate": 1.9945730375543475e-05, "loss": 1.0044, "step": 972 }, { "epoch": 0.6225207933461292, "grad_norm": 2.955875305866405, "learning_rate": 1.994551457361124e-05, "loss": 0.9724, "step": 973 }, { "epoch": 0.6231605886116443, "grad_norm": 2.5746011023277084, "learning_rate": 1.9945298344636298e-05, "loss": 1.0662, "step": 974 }, { "epoch": 0.6238003838771593, "grad_norm": 2.621691993379091, "learning_rate": 1.994508168862794e-05, "loss": 1.0396, "step": 975 }, { "epoch": 0.6244401791426744, "grad_norm": 3.1814057065267387, "learning_rate": 1.9944864605595463e-05, "loss": 0.9978, "step": 976 }, { "epoch": 0.6250799744081894, "grad_norm": 2.6686059807371025, "learning_rate": 1.9944647095548194e-05, "loss": 1.0319, "step": 977 }, { "epoch": 0.6257197696737045, "grad_norm": 2.5836481091504795, "learning_rate": 1.9944429158495468e-05, "loss": 1.0138, "step": 978 }, { "epoch": 0.6263595649392194, "grad_norm": 3.1149836160896327, "learning_rate": 1.994421079444664e-05, "loss": 1.0061, "step": 979 }, { "epoch": 0.6269993602047345, "grad_norm": 3.120396998356657, "learning_rate": 1.9943992003411092e-05, "loss": 0.9896, "step": 980 }, { "epoch": 0.6276391554702495, "grad_norm": 2.3505671190297224, "learning_rate": 1.9943772785398216e-05, "loss": 0.9846, "step": 981 }, { "epoch": 0.6282789507357646, "grad_norm": 3.063857167292707, "learning_rate": 1.9943553140417427e-05, "loss": 1.021, "step": 982 }, { "epoch": 0.6289187460012796, "grad_norm": 2.33766905031754, "learning_rate": 1.994333306847815e-05, "loss": 1.0351, "step": 983 }, { "epoch": 0.6295585412667947, "grad_norm": 2.4009873985407, "learning_rate": 1.994311256958984e-05, "loss": 0.9971, "step": 984 }, { "epoch": 0.6301983365323096, "grad_norm": 2.4733700547024013, "learning_rate": 1.994289164376197e-05, "loss": 1.0427, "step": 985 }, { "epoch": 0.6308381317978247, "grad_norm": 3.557662963542747, "learning_rate": 1.9942670291004013e-05, "loss": 1.0132, "step": 986 }, { "epoch": 0.6314779270633397, "grad_norm": 2.856582960890221, "learning_rate": 1.9942448511325484e-05, "loss": 1.0342, "step": 987 }, { "epoch": 0.6321177223288548, "grad_norm": 2.6603517127066736, "learning_rate": 1.99422263047359e-05, "loss": 1.0465, "step": 988 }, { "epoch": 0.6327575175943698, "grad_norm": 2.5948227343091106, "learning_rate": 1.9942003671244806e-05, "loss": 1.0816, "step": 989 }, { "epoch": 0.6333973128598849, "grad_norm": 2.451801542421886, "learning_rate": 1.994178061086176e-05, "loss": 1.1114, "step": 990 }, { "epoch": 0.6340371081253998, "grad_norm": 2.7186655389581986, "learning_rate": 1.994155712359634e-05, "loss": 1.0472, "step": 991 }, { "epoch": 0.6346769033909149, "grad_norm": 2.7166752684075837, "learning_rate": 1.9941333209458138e-05, "loss": 0.9991, "step": 992 }, { "epoch": 0.6353166986564299, "grad_norm": 2.47224752495485, "learning_rate": 1.9941108868456775e-05, "loss": 1.0223, "step": 993 }, { "epoch": 0.635956493921945, "grad_norm": 2.7960186062056116, "learning_rate": 1.994088410060188e-05, "loss": 1.005, "step": 994 }, { "epoch": 0.63659628918746, "grad_norm": 2.7646738441685748, "learning_rate": 1.9940658905903105e-05, "loss": 1.0282, "step": 995 }, { "epoch": 0.6372360844529751, "grad_norm": 2.8824284658692054, "learning_rate": 1.994043328437012e-05, "loss": 1.0654, "step": 996 }, { "epoch": 0.6378758797184901, "grad_norm": 2.9271960757144595, "learning_rate": 1.994020723601261e-05, "loss": 1.0095, "step": 997 }, { "epoch": 0.6385156749840051, "grad_norm": 2.5182791032154217, "learning_rate": 1.9939980760840288e-05, "loss": 0.9839, "step": 998 }, { "epoch": 0.6391554702495201, "grad_norm": 2.7560022890827747, "learning_rate": 1.9939753858862874e-05, "loss": 0.9938, "step": 999 }, { "epoch": 0.6397952655150352, "grad_norm": 2.4608179739687714, "learning_rate": 1.993952653009011e-05, "loss": 1.0568, "step": 1000 }, { "epoch": 0.6404350607805502, "grad_norm": 2.2793282376973996, "learning_rate": 1.9939298774531757e-05, "loss": 1.0522, "step": 1001 }, { "epoch": 0.6410748560460653, "grad_norm": 2.561311553257534, "learning_rate": 1.9939070592197594e-05, "loss": 1.0581, "step": 1002 }, { "epoch": 0.6417146513115803, "grad_norm": 2.711682858087363, "learning_rate": 1.993884198309742e-05, "loss": 1.0639, "step": 1003 }, { "epoch": 0.6423544465770953, "grad_norm": 2.6394817632098255, "learning_rate": 1.9938612947241053e-05, "loss": 0.9767, "step": 1004 }, { "epoch": 0.6429942418426103, "grad_norm": 2.7171213877412015, "learning_rate": 1.9938383484638326e-05, "loss": 0.9834, "step": 1005 }, { "epoch": 0.6436340371081254, "grad_norm": 2.349026177367202, "learning_rate": 1.993815359529909e-05, "loss": 0.9834, "step": 1006 }, { "epoch": 0.6442738323736404, "grad_norm": 2.858730216249044, "learning_rate": 1.9937923279233214e-05, "loss": 1.0084, "step": 1007 }, { "epoch": 0.6449136276391555, "grad_norm": 3.004309664292986, "learning_rate": 1.9937692536450596e-05, "loss": 0.9832, "step": 1008 }, { "epoch": 0.6455534229046705, "grad_norm": 2.421217943143497, "learning_rate": 1.9937461366961138e-05, "loss": 1.003, "step": 1009 }, { "epoch": 0.6461932181701855, "grad_norm": 2.7694071387885137, "learning_rate": 1.9937229770774763e-05, "loss": 1.0139, "step": 1010 }, { "epoch": 0.6468330134357005, "grad_norm": 2.9394204510583877, "learning_rate": 1.993699774790142e-05, "loss": 1.0072, "step": 1011 }, { "epoch": 0.6474728087012156, "grad_norm": 2.6928242630410244, "learning_rate": 1.9936765298351072e-05, "loss": 1.0808, "step": 1012 }, { "epoch": 0.6481126039667306, "grad_norm": 2.8945449331656588, "learning_rate": 1.9936532422133695e-05, "loss": 1.0223, "step": 1013 }, { "epoch": 0.6487523992322457, "grad_norm": 2.322109871066615, "learning_rate": 1.9936299119259293e-05, "loss": 1.0011, "step": 1014 }, { "epoch": 0.6493921944977608, "grad_norm": 2.821632392583385, "learning_rate": 1.9936065389737882e-05, "loss": 0.9538, "step": 1015 }, { "epoch": 0.6500319897632757, "grad_norm": 2.978283427210038, "learning_rate": 1.9935831233579497e-05, "loss": 1.0033, "step": 1016 }, { "epoch": 0.6506717850287908, "grad_norm": 2.511234618119458, "learning_rate": 1.9935596650794196e-05, "loss": 1.0186, "step": 1017 }, { "epoch": 0.6513115802943058, "grad_norm": 2.3887059349656106, "learning_rate": 1.9935361641392046e-05, "loss": 1.0, "step": 1018 }, { "epoch": 0.6519513755598209, "grad_norm": 2.3686839278063148, "learning_rate": 1.993512620538314e-05, "loss": 1.0239, "step": 1019 }, { "epoch": 0.6525911708253359, "grad_norm": 3.2194960036189584, "learning_rate": 1.9934890342777592e-05, "loss": 1.0069, "step": 1020 }, { "epoch": 0.653230966090851, "grad_norm": 2.649752047419413, "learning_rate": 1.993465405358552e-05, "loss": 0.9938, "step": 1021 }, { "epoch": 0.653870761356366, "grad_norm": 2.523521961891083, "learning_rate": 1.993441733781708e-05, "loss": 1.0824, "step": 1022 }, { "epoch": 0.654510556621881, "grad_norm": 2.5398925010432203, "learning_rate": 1.9934180195482433e-05, "loss": 1.0147, "step": 1023 }, { "epoch": 0.655150351887396, "grad_norm": 2.2500870529966304, "learning_rate": 1.993394262659176e-05, "loss": 0.9956, "step": 1024 }, { "epoch": 0.6557901471529111, "grad_norm": 2.5302699970705524, "learning_rate": 1.9933704631155257e-05, "loss": 0.989, "step": 1025 }, { "epoch": 0.6564299424184261, "grad_norm": 3.1936710489637123, "learning_rate": 1.993346620918315e-05, "loss": 0.973, "step": 1026 }, { "epoch": 0.6570697376839412, "grad_norm": 2.521499591954754, "learning_rate": 1.993322736068567e-05, "loss": 0.9671, "step": 1027 }, { "epoch": 0.6577095329494562, "grad_norm": 2.5810718600890237, "learning_rate": 1.993298808567308e-05, "loss": 0.9441, "step": 1028 }, { "epoch": 0.6583493282149712, "grad_norm": 2.902866707882723, "learning_rate": 1.9932748384155657e-05, "loss": 0.9889, "step": 1029 }, { "epoch": 0.6589891234804862, "grad_norm": 2.2706605427404507, "learning_rate": 1.9932508256143683e-05, "loss": 1.0579, "step": 1030 }, { "epoch": 0.6596289187460013, "grad_norm": 2.6815216984682677, "learning_rate": 1.993226770164747e-05, "loss": 1.0088, "step": 1031 }, { "epoch": 0.6602687140115163, "grad_norm": 3.0337901780408263, "learning_rate": 1.9932026720677352e-05, "loss": 0.9188, "step": 1032 }, { "epoch": 0.6609085092770314, "grad_norm": 2.677462326623673, "learning_rate": 1.9931785313243677e-05, "loss": 1.039, "step": 1033 }, { "epoch": 0.6615483045425464, "grad_norm": 2.550462956359459, "learning_rate": 1.9931543479356804e-05, "loss": 1.0128, "step": 1034 }, { "epoch": 0.6621880998080614, "grad_norm": 2.1401969037847937, "learning_rate": 1.9931301219027124e-05, "loss": 0.9554, "step": 1035 }, { "epoch": 0.6628278950735764, "grad_norm": 3.46988053123975, "learning_rate": 1.9931058532265033e-05, "loss": 1.0277, "step": 1036 }, { "epoch": 0.6634676903390915, "grad_norm": 3.549334466777689, "learning_rate": 1.9930815419080955e-05, "loss": 0.9996, "step": 1037 }, { "epoch": 0.6641074856046065, "grad_norm": 3.416756914268974, "learning_rate": 1.9930571879485333e-05, "loss": 1.0137, "step": 1038 }, { "epoch": 0.6647472808701216, "grad_norm": 3.061313378219286, "learning_rate": 1.9930327913488613e-05, "loss": 0.9655, "step": 1039 }, { "epoch": 0.6653870761356366, "grad_norm": 2.485659212416012, "learning_rate": 1.993008352110128e-05, "loss": 0.9566, "step": 1040 }, { "epoch": 0.6660268714011516, "grad_norm": 2.61214004501287, "learning_rate": 1.9929838702333825e-05, "loss": 1.0365, "step": 1041 }, { "epoch": 0.6666666666666666, "grad_norm": 2.6858695840892906, "learning_rate": 1.992959345719676e-05, "loss": 0.9834, "step": 1042 }, { "epoch": 0.6673064619321817, "grad_norm": 2.809255066025928, "learning_rate": 1.9929347785700617e-05, "loss": 0.9885, "step": 1043 }, { "epoch": 0.6679462571976967, "grad_norm": 3.1835984486256454, "learning_rate": 1.9929101687855942e-05, "loss": 0.9958, "step": 1044 }, { "epoch": 0.6685860524632118, "grad_norm": 2.952907145223023, "learning_rate": 1.9928855163673303e-05, "loss": 1.0199, "step": 1045 }, { "epoch": 0.6692258477287268, "grad_norm": 2.6253271513804814, "learning_rate": 1.9928608213163284e-05, "loss": 0.9962, "step": 1046 }, { "epoch": 0.6698656429942419, "grad_norm": 2.1907618491503498, "learning_rate": 1.992836083633649e-05, "loss": 0.9903, "step": 1047 }, { "epoch": 0.6705054382597568, "grad_norm": 3.9766118398561123, "learning_rate": 1.9928113033203545e-05, "loss": 1.0397, "step": 1048 }, { "epoch": 0.6711452335252719, "grad_norm": 2.992836287603453, "learning_rate": 1.9927864803775084e-05, "loss": 0.9861, "step": 1049 }, { "epoch": 0.6717850287907869, "grad_norm": 2.5106201322427606, "learning_rate": 1.9927616148061772e-05, "loss": 0.9965, "step": 1050 }, { "epoch": 0.672424824056302, "grad_norm": 2.366703018746604, "learning_rate": 1.992736706607428e-05, "loss": 1.0168, "step": 1051 }, { "epoch": 0.673064619321817, "grad_norm": 2.30486210741006, "learning_rate": 1.9927117557823306e-05, "loss": 0.9945, "step": 1052 }, { "epoch": 0.6737044145873321, "grad_norm": 2.661900159573298, "learning_rate": 1.9926867623319562e-05, "loss": 1.0155, "step": 1053 }, { "epoch": 0.674344209852847, "grad_norm": 2.5667962689012427, "learning_rate": 1.9926617262573783e-05, "loss": 1.0174, "step": 1054 }, { "epoch": 0.6749840051183621, "grad_norm": 2.7949761283599943, "learning_rate": 1.9926366475596718e-05, "loss": 0.968, "step": 1055 }, { "epoch": 0.6756238003838771, "grad_norm": 2.826100672267484, "learning_rate": 1.992611526239913e-05, "loss": 1.029, "step": 1056 }, { "epoch": 0.6762635956493922, "grad_norm": 2.617877090321512, "learning_rate": 1.992586362299181e-05, "loss": 1.0112, "step": 1057 }, { "epoch": 0.6769033909149073, "grad_norm": 2.8135844221579083, "learning_rate": 1.9925611557385567e-05, "loss": 0.9742, "step": 1058 }, { "epoch": 0.6775431861804223, "grad_norm": 2.756186919911101, "learning_rate": 1.9925359065591218e-05, "loss": 0.9776, "step": 1059 }, { "epoch": 0.6781829814459372, "grad_norm": 2.489269518380488, "learning_rate": 1.9925106147619608e-05, "loss": 0.923, "step": 1060 }, { "epoch": 0.6788227767114523, "grad_norm": 2.187884808389131, "learning_rate": 1.9924852803481595e-05, "loss": 1.009, "step": 1061 }, { "epoch": 0.6794625719769674, "grad_norm": 2.6546102965817444, "learning_rate": 1.9924599033188055e-05, "loss": 0.9811, "step": 1062 }, { "epoch": 0.6801023672424824, "grad_norm": 2.726597813202643, "learning_rate": 1.9924344836749888e-05, "loss": 1.0185, "step": 1063 }, { "epoch": 0.6807421625079975, "grad_norm": 3.0751479910166375, "learning_rate": 1.992409021417801e-05, "loss": 1.0156, "step": 1064 }, { "epoch": 0.6813819577735125, "grad_norm": 2.4147456960779143, "learning_rate": 1.9923835165483346e-05, "loss": 1.0134, "step": 1065 }, { "epoch": 0.6820217530390275, "grad_norm": 2.357135958493448, "learning_rate": 1.9923579690676858e-05, "loss": 1.0353, "step": 1066 }, { "epoch": 0.6826615483045425, "grad_norm": 2.7107666624847107, "learning_rate": 1.992332378976951e-05, "loss": 1.0216, "step": 1067 }, { "epoch": 0.6833013435700576, "grad_norm": 2.89599274337727, "learning_rate": 1.9923067462772292e-05, "loss": 1.0056, "step": 1068 }, { "epoch": 0.6839411388355726, "grad_norm": 2.465587860797875, "learning_rate": 1.9922810709696207e-05, "loss": 1.0633, "step": 1069 }, { "epoch": 0.6845809341010877, "grad_norm": 2.6118462491104917, "learning_rate": 1.9922553530552283e-05, "loss": 1.0463, "step": 1070 }, { "epoch": 0.6852207293666027, "grad_norm": 2.381144545489971, "learning_rate": 1.992229592535156e-05, "loss": 1.0423, "step": 1071 }, { "epoch": 0.6858605246321178, "grad_norm": 2.496224208655369, "learning_rate": 1.99220378941051e-05, "loss": 0.9242, "step": 1072 }, { "epoch": 0.6865003198976327, "grad_norm": 2.5465563979186494, "learning_rate": 1.9921779436823983e-05, "loss": 1.0205, "step": 1073 }, { "epoch": 0.6871401151631478, "grad_norm": 2.533172458728813, "learning_rate": 1.9921520553519308e-05, "loss": 0.9679, "step": 1074 }, { "epoch": 0.6877799104286628, "grad_norm": 2.691520690156459, "learning_rate": 1.9921261244202186e-05, "loss": 0.9747, "step": 1075 }, { "epoch": 0.6884197056941779, "grad_norm": 2.189241570769951, "learning_rate": 1.9921001508883758e-05, "loss": 0.9778, "step": 1076 }, { "epoch": 0.6890595009596929, "grad_norm": 2.303683124122382, "learning_rate": 1.992074134757517e-05, "loss": 1.0261, "step": 1077 }, { "epoch": 0.689699296225208, "grad_norm": 2.3649015622265765, "learning_rate": 1.99204807602876e-05, "loss": 1.0044, "step": 1078 }, { "epoch": 0.6903390914907229, "grad_norm": 2.497441361767887, "learning_rate": 1.992021974703223e-05, "loss": 0.9977, "step": 1079 }, { "epoch": 0.690978886756238, "grad_norm": 2.4558918028221792, "learning_rate": 1.9919958307820275e-05, "loss": 0.9839, "step": 1080 }, { "epoch": 0.691618682021753, "grad_norm": 2.2859453196904274, "learning_rate": 1.9919696442662953e-05, "loss": 1.0169, "step": 1081 }, { "epoch": 0.6922584772872681, "grad_norm": 2.4542860050896635, "learning_rate": 1.9919434151571512e-05, "loss": 1.0244, "step": 1082 }, { "epoch": 0.6928982725527831, "grad_norm": 2.632400759582093, "learning_rate": 1.9919171434557217e-05, "loss": 0.9659, "step": 1083 }, { "epoch": 0.6935380678182982, "grad_norm": 2.6492968704791546, "learning_rate": 1.991890829163134e-05, "loss": 1.0283, "step": 1084 }, { "epoch": 0.6941778630838131, "grad_norm": 2.3922641799553213, "learning_rate": 1.991864472280519e-05, "loss": 0.9767, "step": 1085 }, { "epoch": 0.6948176583493282, "grad_norm": 2.412294320160818, "learning_rate": 1.9918380728090078e-05, "loss": 1.0608, "step": 1086 }, { "epoch": 0.6954574536148432, "grad_norm": 2.4907835473262545, "learning_rate": 1.9918116307497336e-05, "loss": 0.9884, "step": 1087 }, { "epoch": 0.6960972488803583, "grad_norm": 2.4824727537212348, "learning_rate": 1.991785146103833e-05, "loss": 0.9864, "step": 1088 }, { "epoch": 0.6967370441458733, "grad_norm": 2.5298861476091026, "learning_rate": 1.991758618872442e-05, "loss": 1.0406, "step": 1089 }, { "epoch": 0.6973768394113884, "grad_norm": 2.568121023051471, "learning_rate": 1.9917320490567003e-05, "loss": 0.9861, "step": 1090 }, { "epoch": 0.6980166346769033, "grad_norm": 2.610219108365668, "learning_rate": 1.991705436657749e-05, "loss": 1.0419, "step": 1091 }, { "epoch": 0.6986564299424184, "grad_norm": 2.3179833806349803, "learning_rate": 1.9916787816767294e-05, "loss": 1.0143, "step": 1092 }, { "epoch": 0.6992962252079334, "grad_norm": 2.707490125840079, "learning_rate": 1.991652084114788e-05, "loss": 1.0682, "step": 1093 }, { "epoch": 0.6999360204734485, "grad_norm": 2.163773181665865, "learning_rate": 1.9916253439730693e-05, "loss": 0.9694, "step": 1094 }, { "epoch": 0.7005758157389635, "grad_norm": 2.59132601097835, "learning_rate": 1.991598561252723e-05, "loss": 0.998, "step": 1095 }, { "epoch": 0.7012156110044786, "grad_norm": 2.368598041442015, "learning_rate": 1.9915717359548975e-05, "loss": 1.0065, "step": 1096 }, { "epoch": 0.7018554062699937, "grad_norm": 2.7906322133133887, "learning_rate": 1.991544868080746e-05, "loss": 1.0421, "step": 1097 }, { "epoch": 0.7024952015355086, "grad_norm": 2.7920488391779403, "learning_rate": 1.9915179576314222e-05, "loss": 1.0421, "step": 1098 }, { "epoch": 0.7031349968010236, "grad_norm": 2.3410050802634026, "learning_rate": 1.9914910046080803e-05, "loss": 1.0467, "step": 1099 }, { "epoch": 0.7037747920665387, "grad_norm": 2.782598019603046, "learning_rate": 1.991464009011879e-05, "loss": 1.0056, "step": 1100 }, { "epoch": 0.7044145873320538, "grad_norm": 2.6222324588309505, "learning_rate": 1.9914369708439762e-05, "loss": 1.0301, "step": 1101 }, { "epoch": 0.7050543825975688, "grad_norm": 2.189793027352479, "learning_rate": 1.9914098901055337e-05, "loss": 0.9939, "step": 1102 }, { "epoch": 0.7056941778630839, "grad_norm": 2.235511546571438, "learning_rate": 1.9913827667977144e-05, "loss": 1.0033, "step": 1103 }, { "epoch": 0.7063339731285988, "grad_norm": 2.980593614868651, "learning_rate": 1.9913556009216825e-05, "loss": 1.0162, "step": 1104 }, { "epoch": 0.7069737683941139, "grad_norm": 2.7896185716649984, "learning_rate": 1.9913283924786047e-05, "loss": 0.9442, "step": 1105 }, { "epoch": 0.7076135636596289, "grad_norm": 3.1374164348560107, "learning_rate": 1.991301141469649e-05, "loss": 0.9677, "step": 1106 }, { "epoch": 0.708253358925144, "grad_norm": 2.8795984738009506, "learning_rate": 1.9912738478959856e-05, "loss": 0.9639, "step": 1107 }, { "epoch": 0.708893154190659, "grad_norm": 2.477316710554268, "learning_rate": 1.9912465117587865e-05, "loss": 0.9717, "step": 1108 }, { "epoch": 0.7095329494561741, "grad_norm": 2.7297714064134464, "learning_rate": 1.9912191330592258e-05, "loss": 0.9735, "step": 1109 }, { "epoch": 0.710172744721689, "grad_norm": 2.567962496802519, "learning_rate": 1.9911917117984783e-05, "loss": 1.0064, "step": 1110 }, { "epoch": 0.7108125399872041, "grad_norm": 2.815902235794248, "learning_rate": 1.991164247977722e-05, "loss": 0.9912, "step": 1111 }, { "epoch": 0.7114523352527191, "grad_norm": 2.678999911418075, "learning_rate": 1.9911367415981364e-05, "loss": 1.0431, "step": 1112 }, { "epoch": 0.7120921305182342, "grad_norm": 2.861173520764034, "learning_rate": 1.9911091926609018e-05, "loss": 1.034, "step": 1113 }, { "epoch": 0.7127319257837492, "grad_norm": 2.6638195526347364, "learning_rate": 1.991081601167202e-05, "loss": 0.9723, "step": 1114 }, { "epoch": 0.7133717210492643, "grad_norm": 2.5521437976126133, "learning_rate": 1.9910539671182208e-05, "loss": 0.938, "step": 1115 }, { "epoch": 0.7140115163147792, "grad_norm": 2.4633842255767417, "learning_rate": 1.9910262905151453e-05, "loss": 1.056, "step": 1116 }, { "epoch": 0.7146513115802943, "grad_norm": 2.3577133432494612, "learning_rate": 1.9909985713591634e-05, "loss": 0.9897, "step": 1117 }, { "epoch": 0.7152911068458093, "grad_norm": 2.476528849034166, "learning_rate": 1.9909708096514663e-05, "loss": 1.0605, "step": 1118 }, { "epoch": 0.7159309021113244, "grad_norm": 2.394347471622061, "learning_rate": 1.9909430053932453e-05, "loss": 1.0354, "step": 1119 }, { "epoch": 0.7165706973768394, "grad_norm": 2.4596981382883096, "learning_rate": 1.9909151585856943e-05, "loss": 1.0547, "step": 1120 }, { "epoch": 0.7172104926423545, "grad_norm": 3.24610360895558, "learning_rate": 1.990887269230009e-05, "loss": 0.9886, "step": 1121 }, { "epoch": 0.7178502879078695, "grad_norm": 2.37063894378798, "learning_rate": 1.9908593373273873e-05, "loss": 0.9784, "step": 1122 }, { "epoch": 0.7184900831733845, "grad_norm": 2.919499971975954, "learning_rate": 1.990831362879028e-05, "loss": 0.941, "step": 1123 }, { "epoch": 0.7191298784388995, "grad_norm": 2.8336242671489957, "learning_rate": 1.9908033458861326e-05, "loss": 0.97, "step": 1124 }, { "epoch": 0.7197696737044146, "grad_norm": 2.6093331222429144, "learning_rate": 1.9907752863499042e-05, "loss": 0.9733, "step": 1125 }, { "epoch": 0.7204094689699296, "grad_norm": 2.7768015888325315, "learning_rate": 1.9907471842715474e-05, "loss": 0.9795, "step": 1126 }, { "epoch": 0.7210492642354447, "grad_norm": 2.2931705295961637, "learning_rate": 1.9907190396522686e-05, "loss": 0.9805, "step": 1127 }, { "epoch": 0.7216890595009597, "grad_norm": 2.6534314937111034, "learning_rate": 1.9906908524932768e-05, "loss": 1.0702, "step": 1128 }, { "epoch": 0.7223288547664747, "grad_norm": 2.4779641357156135, "learning_rate": 1.990662622795782e-05, "loss": 1.0086, "step": 1129 }, { "epoch": 0.7229686500319897, "grad_norm": 2.5211990464839733, "learning_rate": 1.990634350560997e-05, "loss": 0.9429, "step": 1130 }, { "epoch": 0.7236084452975048, "grad_norm": 2.5388404190783325, "learning_rate": 1.9906060357901344e-05, "loss": 0.9718, "step": 1131 }, { "epoch": 0.7242482405630198, "grad_norm": 3.0656464947270825, "learning_rate": 1.990577678484411e-05, "loss": 1.0709, "step": 1132 }, { "epoch": 0.7248880358285349, "grad_norm": 2.8147380374518396, "learning_rate": 1.9905492786450443e-05, "loss": 0.982, "step": 1133 }, { "epoch": 0.72552783109405, "grad_norm": 2.5669178043141314, "learning_rate": 1.9905208362732538e-05, "loss": 0.9434, "step": 1134 }, { "epoch": 0.7261676263595649, "grad_norm": 2.587002279757873, "learning_rate": 1.9904923513702602e-05, "loss": 1.0444, "step": 1135 }, { "epoch": 0.7268074216250799, "grad_norm": 2.636693117634626, "learning_rate": 1.9904638239372873e-05, "loss": 1.01, "step": 1136 }, { "epoch": 0.727447216890595, "grad_norm": 3.2707383258230327, "learning_rate": 1.9904352539755595e-05, "loss": 0.9975, "step": 1137 }, { "epoch": 0.72808701215611, "grad_norm": 2.3720412192810976, "learning_rate": 1.9904066414863038e-05, "loss": 1.0072, "step": 1138 }, { "epoch": 0.7287268074216251, "grad_norm": 2.6255755699462515, "learning_rate": 1.990377986470749e-05, "loss": 0.9833, "step": 1139 }, { "epoch": 0.7293666026871402, "grad_norm": 3.3066549279122817, "learning_rate": 1.9903492889301248e-05, "loss": 1.0186, "step": 1140 }, { "epoch": 0.7300063979526551, "grad_norm": 2.3900046490297653, "learning_rate": 1.9903205488656638e-05, "loss": 0.9855, "step": 1141 }, { "epoch": 0.7306461932181701, "grad_norm": 2.6137430055395954, "learning_rate": 1.9902917662786004e-05, "loss": 1.0032, "step": 1142 }, { "epoch": 0.7312859884836852, "grad_norm": 2.591653690828167, "learning_rate": 1.99026294117017e-05, "loss": 0.9766, "step": 1143 }, { "epoch": 0.7319257837492003, "grad_norm": 3.211213368471802, "learning_rate": 1.9902340735416102e-05, "loss": 1.0013, "step": 1144 }, { "epoch": 0.7325655790147153, "grad_norm": 2.553686795628611, "learning_rate": 1.9902051633941612e-05, "loss": 1.0256, "step": 1145 }, { "epoch": 0.7332053742802304, "grad_norm": 2.61652959247375, "learning_rate": 1.9901762107290637e-05, "loss": 1.0456, "step": 1146 }, { "epoch": 0.7338451695457454, "grad_norm": 2.6494700778665763, "learning_rate": 1.990147215547561e-05, "loss": 0.9967, "step": 1147 }, { "epoch": 0.7344849648112604, "grad_norm": 2.8110052405223724, "learning_rate": 1.9901181778508985e-05, "loss": 1.0226, "step": 1148 }, { "epoch": 0.7351247600767754, "grad_norm": 2.564588329553447, "learning_rate": 1.9900890976403224e-05, "loss": 0.9872, "step": 1149 }, { "epoch": 0.7357645553422905, "grad_norm": 2.505941659498274, "learning_rate": 1.990059974917082e-05, "loss": 0.9709, "step": 1150 }, { "epoch": 0.7364043506078055, "grad_norm": 2.9929727498651686, "learning_rate": 1.9900308096824274e-05, "loss": 0.9897, "step": 1151 }, { "epoch": 0.7370441458733206, "grad_norm": 2.8223370109270514, "learning_rate": 1.990001601937611e-05, "loss": 0.9686, "step": 1152 }, { "epoch": 0.7376839411388356, "grad_norm": 2.7202228883879815, "learning_rate": 1.9899723516838865e-05, "loss": 1.0223, "step": 1153 }, { "epoch": 0.7383237364043506, "grad_norm": 2.725982965945247, "learning_rate": 1.9899430589225104e-05, "loss": 1.0392, "step": 1154 }, { "epoch": 0.7389635316698656, "grad_norm": 2.849004850218111, "learning_rate": 1.9899137236547403e-05, "loss": 0.9903, "step": 1155 }, { "epoch": 0.7396033269353807, "grad_norm": 2.625348820977961, "learning_rate": 1.989884345881836e-05, "loss": 0.9758, "step": 1156 }, { "epoch": 0.7402431222008957, "grad_norm": 2.3885915389969297, "learning_rate": 1.989854925605059e-05, "loss": 0.9293, "step": 1157 }, { "epoch": 0.7408829174664108, "grad_norm": 2.4398813259051404, "learning_rate": 1.989825462825672e-05, "loss": 0.9636, "step": 1158 }, { "epoch": 0.7415227127319258, "grad_norm": 2.477673888996292, "learning_rate": 1.9897959575449403e-05, "loss": 1.0, "step": 1159 }, { "epoch": 0.7421625079974408, "grad_norm": 2.460795113536003, "learning_rate": 1.989766409764131e-05, "loss": 0.9835, "step": 1160 }, { "epoch": 0.7428023032629558, "grad_norm": 2.4830046693931176, "learning_rate": 1.9897368194845126e-05, "loss": 1.0477, "step": 1161 }, { "epoch": 0.7434420985284709, "grad_norm": 2.405840201592331, "learning_rate": 1.9897071867073557e-05, "loss": 1.051, "step": 1162 }, { "epoch": 0.7440818937939859, "grad_norm": 2.692121284207796, "learning_rate": 1.989677511433933e-05, "loss": 0.9685, "step": 1163 }, { "epoch": 0.744721689059501, "grad_norm": 2.8482893595313077, "learning_rate": 1.9896477936655185e-05, "loss": 0.9984, "step": 1164 }, { "epoch": 0.745361484325016, "grad_norm": 2.627138998696513, "learning_rate": 1.989618033403388e-05, "loss": 0.9774, "step": 1165 }, { "epoch": 0.746001279590531, "grad_norm": 2.4341425921115882, "learning_rate": 1.9895882306488197e-05, "loss": 0.9604, "step": 1166 }, { "epoch": 0.746641074856046, "grad_norm": 2.4921118731217766, "learning_rate": 1.989558385403093e-05, "loss": 0.9656, "step": 1167 }, { "epoch": 0.7472808701215611, "grad_norm": 3.2316277986566733, "learning_rate": 1.9895284976674896e-05, "loss": 0.9765, "step": 1168 }, { "epoch": 0.7479206653870761, "grad_norm": 2.236370604439613, "learning_rate": 1.9894985674432926e-05, "loss": 0.949, "step": 1169 }, { "epoch": 0.7485604606525912, "grad_norm": 2.294295052193467, "learning_rate": 1.9894685947317874e-05, "loss": 1.0074, "step": 1170 }, { "epoch": 0.7492002559181062, "grad_norm": 2.70390760539497, "learning_rate": 1.989438579534261e-05, "loss": 1.0001, "step": 1171 }, { "epoch": 0.7498400511836213, "grad_norm": 2.7276300340261095, "learning_rate": 1.9894085218520018e-05, "loss": 0.9598, "step": 1172 }, { "epoch": 0.7504798464491362, "grad_norm": 2.8107068637591666, "learning_rate": 1.9893784216863012e-05, "loss": 0.9682, "step": 1173 }, { "epoch": 0.7511196417146513, "grad_norm": 2.5006889335554257, "learning_rate": 1.9893482790384504e-05, "loss": 0.9927, "step": 1174 }, { "epoch": 0.7517594369801663, "grad_norm": 2.5994862381744372, "learning_rate": 1.989318093909745e-05, "loss": 1.0124, "step": 1175 }, { "epoch": 0.7523992322456814, "grad_norm": 2.609505334942117, "learning_rate": 1.98928786630148e-05, "loss": 1.0254, "step": 1176 }, { "epoch": 0.7530390275111964, "grad_norm": 2.5345460121410657, "learning_rate": 1.9892575962149544e-05, "loss": 0.9885, "step": 1177 }, { "epoch": 0.7536788227767115, "grad_norm": 2.5313510417892893, "learning_rate": 1.989227283651467e-05, "loss": 0.9629, "step": 1178 }, { "epoch": 0.7543186180422264, "grad_norm": 2.596159544145088, "learning_rate": 1.9891969286123197e-05, "loss": 1.0244, "step": 1179 }, { "epoch": 0.7549584133077415, "grad_norm": 2.0872934707587563, "learning_rate": 1.9891665310988162e-05, "loss": 1.026, "step": 1180 }, { "epoch": 0.7555982085732565, "grad_norm": 2.619779257369027, "learning_rate": 1.9891360911122612e-05, "loss": 0.9628, "step": 1181 }, { "epoch": 0.7562380038387716, "grad_norm": 2.4404414937282874, "learning_rate": 1.989105608653962e-05, "loss": 0.9831, "step": 1182 }, { "epoch": 0.7568777991042867, "grad_norm": 3.2513740481464586, "learning_rate": 1.9890750837252277e-05, "loss": 0.9934, "step": 1183 }, { "epoch": 0.7575175943698017, "grad_norm": 3.3390249958926073, "learning_rate": 1.9890445163273683e-05, "loss": 1.0016, "step": 1184 }, { "epoch": 0.7581573896353166, "grad_norm": 2.622464004007694, "learning_rate": 1.9890139064616973e-05, "loss": 1.0003, "step": 1185 }, { "epoch": 0.7587971849008317, "grad_norm": 2.3922379833262273, "learning_rate": 1.988983254129528e-05, "loss": 0.9348, "step": 1186 }, { "epoch": 0.7594369801663468, "grad_norm": 2.66609531969179, "learning_rate": 1.9889525593321773e-05, "loss": 0.9612, "step": 1187 }, { "epoch": 0.7600767754318618, "grad_norm": 2.6574823809921795, "learning_rate": 1.988921822070963e-05, "loss": 0.9646, "step": 1188 }, { "epoch": 0.7607165706973769, "grad_norm": 2.903405859583252, "learning_rate": 1.9888910423472044e-05, "loss": 1.0224, "step": 1189 }, { "epoch": 0.7613563659628919, "grad_norm": 2.8556812195712267, "learning_rate": 1.988860220162224e-05, "loss": 0.9648, "step": 1190 }, { "epoch": 0.761996161228407, "grad_norm": 2.5058818741117346, "learning_rate": 1.9888293555173445e-05, "loss": 0.9537, "step": 1191 }, { "epoch": 0.7626359564939219, "grad_norm": 2.0353248822214156, "learning_rate": 1.9887984484138916e-05, "loss": 0.9293, "step": 1192 }, { "epoch": 0.763275751759437, "grad_norm": 2.7821497351363242, "learning_rate": 1.9887674988531924e-05, "loss": 1.0261, "step": 1193 }, { "epoch": 0.763915547024952, "grad_norm": 2.3970800770185785, "learning_rate": 1.9887365068365753e-05, "loss": 1.0253, "step": 1194 }, { "epoch": 0.7645553422904671, "grad_norm": 2.2518581825895656, "learning_rate": 1.9887054723653717e-05, "loss": 1.0304, "step": 1195 }, { "epoch": 0.7651951375559821, "grad_norm": 2.4778324977256565, "learning_rate": 1.9886743954409137e-05, "loss": 1.0265, "step": 1196 }, { "epoch": 0.7658349328214972, "grad_norm": 2.62941168150571, "learning_rate": 1.988643276064536e-05, "loss": 1.0021, "step": 1197 }, { "epoch": 0.7664747280870121, "grad_norm": 2.5309135438290045, "learning_rate": 1.9886121142375745e-05, "loss": 0.9838, "step": 1198 }, { "epoch": 0.7671145233525272, "grad_norm": 2.4514734279170756, "learning_rate": 1.9885809099613677e-05, "loss": 0.9506, "step": 1199 }, { "epoch": 0.7677543186180422, "grad_norm": 2.4398033466110327, "learning_rate": 1.988549663237255e-05, "loss": 0.9385, "step": 1200 }, { "epoch": 0.7683941138835573, "grad_norm": 2.3531026323013933, "learning_rate": 1.9885183740665783e-05, "loss": 0.9884, "step": 1201 }, { "epoch": 0.7690339091490723, "grad_norm": 2.990179245580522, "learning_rate": 1.988487042450681e-05, "loss": 0.8931, "step": 1202 }, { "epoch": 0.7696737044145874, "grad_norm": 2.361973399619161, "learning_rate": 1.9884556683909088e-05, "loss": 0.993, "step": 1203 }, { "epoch": 0.7703134996801023, "grad_norm": 2.7475156291310583, "learning_rate": 1.988424251888608e-05, "loss": 0.9683, "step": 1204 }, { "epoch": 0.7709532949456174, "grad_norm": 2.5303419408070758, "learning_rate": 1.9883927929451284e-05, "loss": 0.9435, "step": 1205 }, { "epoch": 0.7715930902111324, "grad_norm": 2.4409453332943793, "learning_rate": 1.9883612915618204e-05, "loss": 1.0031, "step": 1206 }, { "epoch": 0.7722328854766475, "grad_norm": 2.478465918505081, "learning_rate": 1.988329747740037e-05, "loss": 0.9563, "step": 1207 }, { "epoch": 0.7728726807421625, "grad_norm": 2.6161021394742674, "learning_rate": 1.988298161481132e-05, "loss": 0.9672, "step": 1208 }, { "epoch": 0.7735124760076776, "grad_norm": 2.516885074187038, "learning_rate": 1.988266532786462e-05, "loss": 0.9503, "step": 1209 }, { "epoch": 0.7741522712731925, "grad_norm": 2.5601091535132268, "learning_rate": 1.988234861657385e-05, "loss": 0.9915, "step": 1210 }, { "epoch": 0.7747920665387076, "grad_norm": 2.006677031147774, "learning_rate": 1.9882031480952614e-05, "loss": 0.9147, "step": 1211 }, { "epoch": 0.7754318618042226, "grad_norm": 2.23694731624321, "learning_rate": 1.988171392101452e-05, "loss": 0.9571, "step": 1212 }, { "epoch": 0.7760716570697377, "grad_norm": 2.368142378158542, "learning_rate": 1.988139593677321e-05, "loss": 0.9508, "step": 1213 }, { "epoch": 0.7767114523352527, "grad_norm": 2.5977135312781274, "learning_rate": 1.9881077528242335e-05, "loss": 0.9985, "step": 1214 }, { "epoch": 0.7773512476007678, "grad_norm": 2.2273300719894706, "learning_rate": 1.988075869543557e-05, "loss": 0.9996, "step": 1215 }, { "epoch": 0.7779910428662828, "grad_norm": 2.254388205297791, "learning_rate": 1.9880439438366605e-05, "loss": 0.9587, "step": 1216 }, { "epoch": 0.7786308381317978, "grad_norm": 2.4585929654469263, "learning_rate": 1.9880119757049146e-05, "loss": 1.012, "step": 1217 }, { "epoch": 0.7792706333973128, "grad_norm": 2.407764622060685, "learning_rate": 1.9879799651496916e-05, "loss": 0.9942, "step": 1218 }, { "epoch": 0.7799104286628279, "grad_norm": 2.082594256932956, "learning_rate": 1.9879479121723667e-05, "loss": 0.9868, "step": 1219 }, { "epoch": 0.780550223928343, "grad_norm": 2.6054305202163612, "learning_rate": 1.987915816774316e-05, "loss": 0.9686, "step": 1220 }, { "epoch": 0.781190019193858, "grad_norm": 2.423058773050418, "learning_rate": 1.987883678956917e-05, "loss": 0.9975, "step": 1221 }, { "epoch": 0.781829814459373, "grad_norm": 2.2579186986913555, "learning_rate": 1.9878514987215505e-05, "loss": 0.9614, "step": 1222 }, { "epoch": 0.782469609724888, "grad_norm": 2.4679772256229064, "learning_rate": 1.9878192760695977e-05, "loss": 0.9983, "step": 1223 }, { "epoch": 0.783109404990403, "grad_norm": 2.436565759058268, "learning_rate": 1.9877870110024426e-05, "loss": 0.9811, "step": 1224 }, { "epoch": 0.7837492002559181, "grad_norm": 2.361469389369056, "learning_rate": 1.9877547035214703e-05, "loss": 0.9948, "step": 1225 }, { "epoch": 0.7843889955214332, "grad_norm": 2.258256926787194, "learning_rate": 1.9877223536280682e-05, "loss": 0.9478, "step": 1226 }, { "epoch": 0.7850287907869482, "grad_norm": 2.20648392453576, "learning_rate": 1.9876899613236248e-05, "loss": 0.9486, "step": 1227 }, { "epoch": 0.7856685860524633, "grad_norm": 2.204036990964189, "learning_rate": 1.987657526609532e-05, "loss": 0.9677, "step": 1228 }, { "epoch": 0.7863083813179782, "grad_norm": 2.607688071922758, "learning_rate": 1.9876250494871815e-05, "loss": 1.0367, "step": 1229 }, { "epoch": 0.7869481765834933, "grad_norm": 2.567493387497928, "learning_rate": 1.9875925299579686e-05, "loss": 0.9353, "step": 1230 }, { "epoch": 0.7875879718490083, "grad_norm": 2.6144293053569636, "learning_rate": 1.987559968023289e-05, "loss": 0.9678, "step": 1231 }, { "epoch": 0.7882277671145234, "grad_norm": 2.496560034727067, "learning_rate": 1.987527363684541e-05, "loss": 0.9868, "step": 1232 }, { "epoch": 0.7888675623800384, "grad_norm": 2.413333041033104, "learning_rate": 1.987494716943125e-05, "loss": 0.9804, "step": 1233 }, { "epoch": 0.7895073576455535, "grad_norm": 2.5393105270452563, "learning_rate": 1.9874620278004422e-05, "loss": 0.9973, "step": 1234 }, { "epoch": 0.7901471529110684, "grad_norm": 2.4341333322336034, "learning_rate": 1.9874292962578967e-05, "loss": 1.0125, "step": 1235 }, { "epoch": 0.7907869481765835, "grad_norm": 2.3948729196973364, "learning_rate": 1.9873965223168933e-05, "loss": 0.974, "step": 1236 }, { "epoch": 0.7914267434420985, "grad_norm": 3.055523256051832, "learning_rate": 1.9873637059788403e-05, "loss": 0.9976, "step": 1237 }, { "epoch": 0.7920665387076136, "grad_norm": 2.6737257360991067, "learning_rate": 1.9873308472451458e-05, "loss": 1.0337, "step": 1238 }, { "epoch": 0.7927063339731286, "grad_norm": 2.527062826312763, "learning_rate": 1.987297946117221e-05, "loss": 0.9759, "step": 1239 }, { "epoch": 0.7933461292386437, "grad_norm": 2.2628082870056634, "learning_rate": 1.9872650025964786e-05, "loss": 0.9856, "step": 1240 }, { "epoch": 0.7939859245041587, "grad_norm": 2.461842202777146, "learning_rate": 1.9872320166843336e-05, "loss": 1.0424, "step": 1241 }, { "epoch": 0.7946257197696737, "grad_norm": 2.8153410259426934, "learning_rate": 1.9871989883822016e-05, "loss": 1.0585, "step": 1242 }, { "epoch": 0.7952655150351887, "grad_norm": 2.5940706884995484, "learning_rate": 1.9871659176915014e-05, "loss": 1.0189, "step": 1243 }, { "epoch": 0.7959053103007038, "grad_norm": 2.4456987546103437, "learning_rate": 1.9871328046136523e-05, "loss": 0.9943, "step": 1244 }, { "epoch": 0.7965451055662188, "grad_norm": 2.338198774757478, "learning_rate": 1.9870996491500767e-05, "loss": 0.9803, "step": 1245 }, { "epoch": 0.7971849008317339, "grad_norm": 2.332765471177455, "learning_rate": 1.9870664513021985e-05, "loss": 0.9626, "step": 1246 }, { "epoch": 0.7978246960972489, "grad_norm": 2.701791954331873, "learning_rate": 1.9870332110714422e-05, "loss": 0.9971, "step": 1247 }, { "epoch": 0.7984644913627639, "grad_norm": 2.363907979234889, "learning_rate": 1.986999928459236e-05, "loss": 1.0011, "step": 1248 }, { "epoch": 0.7991042866282789, "grad_norm": 2.7650495812231846, "learning_rate": 1.9869666034670083e-05, "loss": 0.9713, "step": 1249 }, { "epoch": 0.799744081893794, "grad_norm": 2.373733827239882, "learning_rate": 1.9869332360961905e-05, "loss": 1.0345, "step": 1250 }, { "epoch": 0.800383877159309, "grad_norm": 2.4486650465585296, "learning_rate": 1.986899826348215e-05, "loss": 0.9941, "step": 1251 }, { "epoch": 0.8010236724248241, "grad_norm": 2.6281485496959367, "learning_rate": 1.9868663742245167e-05, "loss": 0.9594, "step": 1252 }, { "epoch": 0.8016634676903391, "grad_norm": 2.7644117709979494, "learning_rate": 1.9868328797265314e-05, "loss": 0.9962, "step": 1253 }, { "epoch": 0.8023032629558541, "grad_norm": 2.4965253582502522, "learning_rate": 1.986799342855698e-05, "loss": 1.0344, "step": 1254 }, { "epoch": 0.8029430582213691, "grad_norm": 2.1215837668755997, "learning_rate": 1.9867657636134563e-05, "loss": 0.954, "step": 1255 }, { "epoch": 0.8035828534868842, "grad_norm": 2.5038525978235744, "learning_rate": 1.986732142001248e-05, "loss": 1.006, "step": 1256 }, { "epoch": 0.8042226487523992, "grad_norm": 2.5144222406219834, "learning_rate": 1.9866984780205166e-05, "loss": 0.9891, "step": 1257 }, { "epoch": 0.8048624440179143, "grad_norm": 3.0125563689317914, "learning_rate": 1.9866647716727078e-05, "loss": 1.0104, "step": 1258 }, { "epoch": 0.8055022392834293, "grad_norm": 2.5984528811397376, "learning_rate": 1.9866310229592692e-05, "loss": 0.9178, "step": 1259 }, { "epoch": 0.8061420345489443, "grad_norm": 2.1369143680168765, "learning_rate": 1.9865972318816492e-05, "loss": 0.9922, "step": 1260 }, { "epoch": 0.8067818298144593, "grad_norm": 2.4581509549683043, "learning_rate": 1.986563398441299e-05, "loss": 0.9289, "step": 1261 }, { "epoch": 0.8074216250799744, "grad_norm": 2.3815303037093107, "learning_rate": 1.986529522639672e-05, "loss": 0.975, "step": 1262 }, { "epoch": 0.8080614203454894, "grad_norm": 3.084544014104132, "learning_rate": 1.9864956044782215e-05, "loss": 1.0181, "step": 1263 }, { "epoch": 0.8087012156110045, "grad_norm": 2.3545247920758983, "learning_rate": 1.9864616439584048e-05, "loss": 1.0167, "step": 1264 }, { "epoch": 0.8093410108765196, "grad_norm": 2.42878592869661, "learning_rate": 1.98642764108168e-05, "loss": 0.9943, "step": 1265 }, { "epoch": 0.8099808061420346, "grad_norm": 2.626800559671383, "learning_rate": 1.986393595849507e-05, "loss": 0.9528, "step": 1266 }, { "epoch": 0.8106206014075495, "grad_norm": 2.5484314579379275, "learning_rate": 1.9863595082633475e-05, "loss": 0.9751, "step": 1267 }, { "epoch": 0.8112603966730646, "grad_norm": 2.9286597596202975, "learning_rate": 1.9863253783246655e-05, "loss": 0.9384, "step": 1268 }, { "epoch": 0.8119001919385797, "grad_norm": 2.330794743972676, "learning_rate": 1.9862912060349263e-05, "loss": 1.0013, "step": 1269 }, { "epoch": 0.8125399872040947, "grad_norm": 2.49577181739862, "learning_rate": 1.986256991395597e-05, "loss": 1.0063, "step": 1270 }, { "epoch": 0.8131797824696098, "grad_norm": 2.8475928614562664, "learning_rate": 1.986222734408147e-05, "loss": 0.9792, "step": 1271 }, { "epoch": 0.8138195777351248, "grad_norm": 3.0095138200007283, "learning_rate": 1.986188435074047e-05, "loss": 0.929, "step": 1272 }, { "epoch": 0.8144593730006398, "grad_norm": 2.1038797407668075, "learning_rate": 1.98615409339477e-05, "loss": 0.9604, "step": 1273 }, { "epoch": 0.8150991682661548, "grad_norm": 2.2689836838312822, "learning_rate": 1.9861197093717908e-05, "loss": 0.9807, "step": 1274 }, { "epoch": 0.8157389635316699, "grad_norm": 2.907411395022464, "learning_rate": 1.986085283006585e-05, "loss": 1.0295, "step": 1275 }, { "epoch": 0.8163787587971849, "grad_norm": 2.6971174858856344, "learning_rate": 1.9860508143006317e-05, "loss": 0.9712, "step": 1276 }, { "epoch": 0.8170185540627, "grad_norm": 2.3065930961801615, "learning_rate": 1.98601630325541e-05, "loss": 0.8903, "step": 1277 }, { "epoch": 0.817658349328215, "grad_norm": 2.114853457854308, "learning_rate": 1.9859817498724023e-05, "loss": 1.0011, "step": 1278 }, { "epoch": 0.81829814459373, "grad_norm": 2.0435278025895873, "learning_rate": 1.9859471541530924e-05, "loss": 0.955, "step": 1279 }, { "epoch": 0.818937939859245, "grad_norm": 2.816778810995841, "learning_rate": 1.9859125160989653e-05, "loss": 1.0146, "step": 1280 }, { "epoch": 0.8195777351247601, "grad_norm": 2.9895977855400075, "learning_rate": 1.9858778357115088e-05, "loss": 1.0066, "step": 1281 }, { "epoch": 0.8202175303902751, "grad_norm": 2.32770688735366, "learning_rate": 1.985843112992212e-05, "loss": 0.9686, "step": 1282 }, { "epoch": 0.8208573256557902, "grad_norm": 2.162292530720862, "learning_rate": 1.985808347942565e-05, "loss": 0.9382, "step": 1283 }, { "epoch": 0.8214971209213052, "grad_norm": 3.521629352319627, "learning_rate": 1.9857735405640616e-05, "loss": 0.9625, "step": 1284 }, { "epoch": 0.8221369161868202, "grad_norm": 2.560909175683413, "learning_rate": 1.985738690858196e-05, "loss": 0.9488, "step": 1285 }, { "epoch": 0.8227767114523352, "grad_norm": 2.316962334685813, "learning_rate": 1.985703798826464e-05, "loss": 1.0077, "step": 1286 }, { "epoch": 0.8234165067178503, "grad_norm": 3.116579131581034, "learning_rate": 1.9856688644703647e-05, "loss": 0.9441, "step": 1287 }, { "epoch": 0.8240563019833653, "grad_norm": 2.826022998174468, "learning_rate": 1.9856338877913975e-05, "loss": 0.96, "step": 1288 }, { "epoch": 0.8246960972488804, "grad_norm": 2.473866634637742, "learning_rate": 1.9855988687910647e-05, "loss": 1.0195, "step": 1289 }, { "epoch": 0.8253358925143954, "grad_norm": 2.525146422618344, "learning_rate": 1.9855638074708693e-05, "loss": 1.0036, "step": 1290 }, { "epoch": 0.8259756877799105, "grad_norm": 3.065939623773754, "learning_rate": 1.985528703832318e-05, "loss": 0.9789, "step": 1291 }, { "epoch": 0.8266154830454254, "grad_norm": 3.5242681645316973, "learning_rate": 1.9854935578769164e-05, "loss": 0.9263, "step": 1292 }, { "epoch": 0.8272552783109405, "grad_norm": 2.057967111549053, "learning_rate": 1.985458369606175e-05, "loss": 0.9799, "step": 1293 }, { "epoch": 0.8278950735764555, "grad_norm": 2.5652470817602966, "learning_rate": 1.9854231390216043e-05, "loss": 0.941, "step": 1294 }, { "epoch": 0.8285348688419706, "grad_norm": 3.0133070567050773, "learning_rate": 1.9853878661247163e-05, "loss": 0.9986, "step": 1295 }, { "epoch": 0.8291746641074856, "grad_norm": 2.9769368536237373, "learning_rate": 1.985352550917027e-05, "loss": 0.9858, "step": 1296 }, { "epoch": 0.8298144593730007, "grad_norm": 2.8163007236934003, "learning_rate": 1.9853171934000516e-05, "loss": 1.0618, "step": 1297 }, { "epoch": 0.8304542546385156, "grad_norm": 2.0227623289190566, "learning_rate": 1.9852817935753083e-05, "loss": 0.9821, "step": 1298 }, { "epoch": 0.8310940499040307, "grad_norm": 2.524197801377639, "learning_rate": 1.985246351444318e-05, "loss": 0.9622, "step": 1299 }, { "epoch": 0.8317338451695457, "grad_norm": 2.5969070912174166, "learning_rate": 1.9852108670086017e-05, "loss": 0.9835, "step": 1300 }, { "epoch": 0.8323736404350608, "grad_norm": 2.2401840319795356, "learning_rate": 1.9851753402696834e-05, "loss": 1.0233, "step": 1301 }, { "epoch": 0.8330134357005758, "grad_norm": 2.726465695401673, "learning_rate": 1.9851397712290886e-05, "loss": 1.0008, "step": 1302 }, { "epoch": 0.8336532309660909, "grad_norm": 2.639398216558648, "learning_rate": 1.9851041598883443e-05, "loss": 0.9043, "step": 1303 }, { "epoch": 0.8342930262316058, "grad_norm": 2.52521424710568, "learning_rate": 1.98506850624898e-05, "loss": 1.0209, "step": 1304 }, { "epoch": 0.8349328214971209, "grad_norm": 2.3934380666617425, "learning_rate": 1.9850328103125257e-05, "loss": 0.9531, "step": 1305 }, { "epoch": 0.835572616762636, "grad_norm": 2.730220468331288, "learning_rate": 1.9849970720805152e-05, "loss": 0.9549, "step": 1306 }, { "epoch": 0.836212412028151, "grad_norm": 2.4226117882885023, "learning_rate": 1.9849612915544828e-05, "loss": 1.051, "step": 1307 }, { "epoch": 0.836852207293666, "grad_norm": 2.519066655939983, "learning_rate": 1.9849254687359645e-05, "loss": 0.9584, "step": 1308 }, { "epoch": 0.8374920025591811, "grad_norm": 3.045864829902733, "learning_rate": 1.9848896036264985e-05, "loss": 0.9888, "step": 1309 }, { "epoch": 0.838131797824696, "grad_norm": 2.388669848433297, "learning_rate": 1.984853696227625e-05, "loss": 0.9454, "step": 1310 }, { "epoch": 0.8387715930902111, "grad_norm": 2.1921886321451596, "learning_rate": 1.9848177465408854e-05, "loss": 0.9378, "step": 1311 }, { "epoch": 0.8394113883557262, "grad_norm": 2.365325798665154, "learning_rate": 1.9847817545678238e-05, "loss": 0.9665, "step": 1312 }, { "epoch": 0.8400511836212412, "grad_norm": 2.3555213407560043, "learning_rate": 1.9847457203099858e-05, "loss": 0.9773, "step": 1313 }, { "epoch": 0.8406909788867563, "grad_norm": 3.020890887140426, "learning_rate": 1.984709643768918e-05, "loss": 1.0164, "step": 1314 }, { "epoch": 0.8413307741522713, "grad_norm": 2.4289942218080776, "learning_rate": 1.9846735249461696e-05, "loss": 0.9583, "step": 1315 }, { "epoch": 0.8419705694177864, "grad_norm": 2.297842248672399, "learning_rate": 1.984637363843292e-05, "loss": 0.9604, "step": 1316 }, { "epoch": 0.8426103646833013, "grad_norm": 2.347295114781508, "learning_rate": 1.9846011604618375e-05, "loss": 0.8761, "step": 1317 }, { "epoch": 0.8432501599488164, "grad_norm": 2.290855031657481, "learning_rate": 1.9845649148033603e-05, "loss": 0.9777, "step": 1318 }, { "epoch": 0.8438899552143314, "grad_norm": 2.3219748208943467, "learning_rate": 1.984528626869417e-05, "loss": 0.9729, "step": 1319 }, { "epoch": 0.8445297504798465, "grad_norm": 2.5494818924856926, "learning_rate": 1.9844922966615663e-05, "loss": 0.9431, "step": 1320 }, { "epoch": 0.8451695457453615, "grad_norm": 2.889487555522606, "learning_rate": 1.9844559241813674e-05, "loss": 0.981, "step": 1321 }, { "epoch": 0.8458093410108766, "grad_norm": 3.024804288669281, "learning_rate": 1.9844195094303826e-05, "loss": 0.9477, "step": 1322 }, { "epoch": 0.8464491362763915, "grad_norm": 2.2547821790666407, "learning_rate": 1.9843830524101748e-05, "loss": 0.9101, "step": 1323 }, { "epoch": 0.8470889315419066, "grad_norm": 3.1140314514025214, "learning_rate": 1.9843465531223098e-05, "loss": 0.9602, "step": 1324 }, { "epoch": 0.8477287268074216, "grad_norm": 2.455590362429133, "learning_rate": 1.984310011568355e-05, "loss": 0.9886, "step": 1325 }, { "epoch": 0.8483685220729367, "grad_norm": 2.7908362010589753, "learning_rate": 1.9842734277498793e-05, "loss": 0.9279, "step": 1326 }, { "epoch": 0.8490083173384517, "grad_norm": 2.301512611535804, "learning_rate": 1.9842368016684534e-05, "loss": 0.9641, "step": 1327 }, { "epoch": 0.8496481126039668, "grad_norm": 2.6213096624126098, "learning_rate": 1.98420013332565e-05, "loss": 0.9451, "step": 1328 }, { "epoch": 0.8502879078694817, "grad_norm": 2.3366852136789245, "learning_rate": 1.984163422723044e-05, "loss": 0.9983, "step": 1329 }, { "epoch": 0.8509277031349968, "grad_norm": 2.94198088472575, "learning_rate": 1.984126669862211e-05, "loss": 0.9307, "step": 1330 }, { "epoch": 0.8515674984005118, "grad_norm": 2.6647932089349653, "learning_rate": 1.9840898747447293e-05, "loss": 0.9366, "step": 1331 }, { "epoch": 0.8522072936660269, "grad_norm": 2.345679845359141, "learning_rate": 1.984053037372179e-05, "loss": 0.9833, "step": 1332 }, { "epoch": 0.8528470889315419, "grad_norm": 2.151590953944923, "learning_rate": 1.9840161577461422e-05, "loss": 0.9748, "step": 1333 }, { "epoch": 0.853486884197057, "grad_norm": 2.556563337896248, "learning_rate": 1.9839792358682018e-05, "loss": 0.9198, "step": 1334 }, { "epoch": 0.8541266794625719, "grad_norm": 2.43071435388098, "learning_rate": 1.983942271739943e-05, "loss": 0.9275, "step": 1335 }, { "epoch": 0.854766474728087, "grad_norm": 2.3838188009420618, "learning_rate": 1.9839052653629535e-05, "loss": 0.9672, "step": 1336 }, { "epoch": 0.855406269993602, "grad_norm": 2.650822928971399, "learning_rate": 1.9838682167388223e-05, "loss": 0.9449, "step": 1337 }, { "epoch": 0.8560460652591171, "grad_norm": 2.540074978590469, "learning_rate": 1.98383112586914e-05, "loss": 0.9482, "step": 1338 }, { "epoch": 0.8566858605246321, "grad_norm": 2.551618586453057, "learning_rate": 1.9837939927554995e-05, "loss": 0.9276, "step": 1339 }, { "epoch": 0.8573256557901472, "grad_norm": 2.649094140034406, "learning_rate": 1.9837568173994947e-05, "loss": 1.0006, "step": 1340 }, { "epoch": 0.8579654510556622, "grad_norm": 2.609107374201131, "learning_rate": 1.9837195998027222e-05, "loss": 0.97, "step": 1341 }, { "epoch": 0.8586052463211772, "grad_norm": 2.443439546814091, "learning_rate": 1.98368233996678e-05, "loss": 0.9473, "step": 1342 }, { "epoch": 0.8592450415866922, "grad_norm": 2.32258754724658, "learning_rate": 1.983645037893268e-05, "loss": 0.9291, "step": 1343 }, { "epoch": 0.8598848368522073, "grad_norm": 2.3451940726151417, "learning_rate": 1.9836076935837876e-05, "loss": 0.9698, "step": 1344 }, { "epoch": 0.8605246321177223, "grad_norm": 2.475478829933603, "learning_rate": 1.9835703070399428e-05, "loss": 0.9631, "step": 1345 }, { "epoch": 0.8611644273832374, "grad_norm": 2.2954830321304196, "learning_rate": 1.9835328782633384e-05, "loss": 0.9806, "step": 1346 }, { "epoch": 0.8618042226487524, "grad_norm": 2.067070111789748, "learning_rate": 1.983495407255582e-05, "loss": 0.9244, "step": 1347 }, { "epoch": 0.8624440179142674, "grad_norm": 2.6736783037944956, "learning_rate": 1.9834578940182822e-05, "loss": 0.9943, "step": 1348 }, { "epoch": 0.8630838131797824, "grad_norm": 2.294197600222739, "learning_rate": 1.98342033855305e-05, "loss": 0.9648, "step": 1349 }, { "epoch": 0.8637236084452975, "grad_norm": 2.4454149975912123, "learning_rate": 1.9833827408614976e-05, "loss": 1.0292, "step": 1350 }, { "epoch": 0.8643634037108126, "grad_norm": 2.7083402051579526, "learning_rate": 1.9833451009452398e-05, "loss": 0.9896, "step": 1351 }, { "epoch": 0.8650031989763276, "grad_norm": 2.3207551641860733, "learning_rate": 1.9833074188058926e-05, "loss": 0.9941, "step": 1352 }, { "epoch": 0.8656429942418427, "grad_norm": 2.1810635305999546, "learning_rate": 1.9832696944450738e-05, "loss": 0.9606, "step": 1353 }, { "epoch": 0.8662827895073576, "grad_norm": 2.4861247083193283, "learning_rate": 1.9832319278644036e-05, "loss": 0.9334, "step": 1354 }, { "epoch": 0.8669225847728727, "grad_norm": 2.3674305854361815, "learning_rate": 1.9831941190655035e-05, "loss": 0.9147, "step": 1355 }, { "epoch": 0.8675623800383877, "grad_norm": 2.4550254400594755, "learning_rate": 1.9831562680499966e-05, "loss": 0.9586, "step": 1356 }, { "epoch": 0.8682021753039028, "grad_norm": 2.363310600809593, "learning_rate": 1.9831183748195086e-05, "loss": 0.9741, "step": 1357 }, { "epoch": 0.8688419705694178, "grad_norm": 2.529891459864506, "learning_rate": 1.9830804393756666e-05, "loss": 0.9307, "step": 1358 }, { "epoch": 0.8694817658349329, "grad_norm": 2.35882341118182, "learning_rate": 1.9830424617200994e-05, "loss": 0.9579, "step": 1359 }, { "epoch": 0.8701215611004478, "grad_norm": 2.1892830061599153, "learning_rate": 1.9830044418544372e-05, "loss": 0.976, "step": 1360 }, { "epoch": 0.8707613563659629, "grad_norm": 2.368792562834722, "learning_rate": 1.9829663797803128e-05, "loss": 0.9421, "step": 1361 }, { "epoch": 0.8714011516314779, "grad_norm": 2.5261988871018652, "learning_rate": 1.982928275499361e-05, "loss": 0.9294, "step": 1362 }, { "epoch": 0.872040946896993, "grad_norm": 2.4166335905290586, "learning_rate": 1.9828901290132175e-05, "loss": 0.9484, "step": 1363 }, { "epoch": 0.872680742162508, "grad_norm": 2.501229407731415, "learning_rate": 1.9828519403235202e-05, "loss": 0.9555, "step": 1364 }, { "epoch": 0.8733205374280231, "grad_norm": 2.396010429513849, "learning_rate": 1.9828137094319093e-05, "loss": 1.0023, "step": 1365 }, { "epoch": 0.8739603326935381, "grad_norm": 2.4878246153178063, "learning_rate": 1.9827754363400253e-05, "loss": 0.9298, "step": 1366 }, { "epoch": 0.8746001279590531, "grad_norm": 2.4852191243192854, "learning_rate": 1.9827371210495128e-05, "loss": 0.9274, "step": 1367 }, { "epoch": 0.8752399232245681, "grad_norm": 2.1238612611708994, "learning_rate": 1.9826987635620164e-05, "loss": 0.9636, "step": 1368 }, { "epoch": 0.8758797184900832, "grad_norm": 2.5645560912278316, "learning_rate": 1.982660363879183e-05, "loss": 0.9311, "step": 1369 }, { "epoch": 0.8765195137555982, "grad_norm": 1.9600131519740296, "learning_rate": 1.9826219220026615e-05, "loss": 0.996, "step": 1370 }, { "epoch": 0.8771593090211133, "grad_norm": 2.5936309204722163, "learning_rate": 1.9825834379341028e-05, "loss": 0.9451, "step": 1371 }, { "epoch": 0.8777991042866283, "grad_norm": 2.6426971779221073, "learning_rate": 1.9825449116751595e-05, "loss": 1.04, "step": 1372 }, { "epoch": 0.8784388995521433, "grad_norm": 2.4529572059845055, "learning_rate": 1.982506343227485e-05, "loss": 1.0168, "step": 1373 }, { "epoch": 0.8790786948176583, "grad_norm": 2.2099578880079007, "learning_rate": 1.9824677325927362e-05, "loss": 0.9459, "step": 1374 }, { "epoch": 0.8797184900831734, "grad_norm": 2.3568783780294127, "learning_rate": 1.98242907977257e-05, "loss": 0.9902, "step": 1375 }, { "epoch": 0.8803582853486884, "grad_norm": 2.400443139201545, "learning_rate": 1.9823903847686474e-05, "loss": 0.9463, "step": 1376 }, { "epoch": 0.8809980806142035, "grad_norm": 2.261496023541625, "learning_rate": 1.982351647582629e-05, "loss": 0.9197, "step": 1377 }, { "epoch": 0.8816378758797185, "grad_norm": 2.6433503578974333, "learning_rate": 1.9823128682161785e-05, "loss": 0.9518, "step": 1378 }, { "epoch": 0.8822776711452335, "grad_norm": 2.3586634317726523, "learning_rate": 1.9822740466709606e-05, "loss": 0.9821, "step": 1379 }, { "epoch": 0.8829174664107485, "grad_norm": 2.0817131779834472, "learning_rate": 1.9822351829486424e-05, "loss": 0.925, "step": 1380 }, { "epoch": 0.8835572616762636, "grad_norm": 2.3702448233386266, "learning_rate": 1.9821962770508928e-05, "loss": 0.9842, "step": 1381 }, { "epoch": 0.8841970569417786, "grad_norm": 2.076432504165223, "learning_rate": 1.9821573289793827e-05, "loss": 0.9012, "step": 1382 }, { "epoch": 0.8848368522072937, "grad_norm": 2.6409203139706876, "learning_rate": 1.9821183387357832e-05, "loss": 0.9836, "step": 1383 }, { "epoch": 0.8854766474728087, "grad_norm": 2.6868601712490308, "learning_rate": 1.9820793063217697e-05, "loss": 0.9649, "step": 1384 }, { "epoch": 0.8861164427383237, "grad_norm": 2.304793546470909, "learning_rate": 1.982040231739018e-05, "loss": 0.9732, "step": 1385 }, { "epoch": 0.8867562380038387, "grad_norm": 2.6324133305572586, "learning_rate": 1.9820011149892055e-05, "loss": 0.9757, "step": 1386 }, { "epoch": 0.8873960332693538, "grad_norm": 2.59334855979923, "learning_rate": 1.981961956074012e-05, "loss": 0.9887, "step": 1387 }, { "epoch": 0.8880358285348688, "grad_norm": 2.082969473425823, "learning_rate": 1.9819227549951186e-05, "loss": 0.9494, "step": 1388 }, { "epoch": 0.8886756238003839, "grad_norm": 2.345547651579458, "learning_rate": 1.9818835117542088e-05, "loss": 1.017, "step": 1389 }, { "epoch": 0.889315419065899, "grad_norm": 2.8548050327716945, "learning_rate": 1.9818442263529678e-05, "loss": 0.9902, "step": 1390 }, { "epoch": 0.889955214331414, "grad_norm": 2.4376601279186314, "learning_rate": 1.9818048987930824e-05, "loss": 0.9826, "step": 1391 }, { "epoch": 0.8905950095969289, "grad_norm": 2.2798972987562665, "learning_rate": 1.981765529076241e-05, "loss": 0.946, "step": 1392 }, { "epoch": 0.891234804862444, "grad_norm": 2.406272782446205, "learning_rate": 1.981726117204134e-05, "loss": 1.0061, "step": 1393 }, { "epoch": 0.891874600127959, "grad_norm": 2.2444083048377736, "learning_rate": 1.9816866631784545e-05, "loss": 0.9542, "step": 1394 }, { "epoch": 0.8925143953934741, "grad_norm": 2.566173110140673, "learning_rate": 1.9816471670008956e-05, "loss": 1.0033, "step": 1395 }, { "epoch": 0.8931541906589892, "grad_norm": 2.597255064913158, "learning_rate": 1.9816076286731535e-05, "loss": 0.9603, "step": 1396 }, { "epoch": 0.8937939859245042, "grad_norm": 2.789983691410866, "learning_rate": 1.9815680481969258e-05, "loss": 0.9612, "step": 1397 }, { "epoch": 0.8944337811900192, "grad_norm": 2.196464667574699, "learning_rate": 1.9815284255739128e-05, "loss": 1.0023, "step": 1398 }, { "epoch": 0.8950735764555342, "grad_norm": 2.2389501050318583, "learning_rate": 1.9814887608058146e-05, "loss": 0.9448, "step": 1399 }, { "epoch": 0.8957133717210493, "grad_norm": 2.7845560116506034, "learning_rate": 1.981449053894335e-05, "loss": 0.9918, "step": 1400 }, { "epoch": 0.8963531669865643, "grad_norm": 2.556347496080654, "learning_rate": 1.9814093048411793e-05, "loss": 0.9713, "step": 1401 }, { "epoch": 0.8969929622520794, "grad_norm": 2.460009621122525, "learning_rate": 1.9813695136480534e-05, "loss": 0.9771, "step": 1402 }, { "epoch": 0.8976327575175944, "grad_norm": 2.384847471192659, "learning_rate": 1.9813296803166667e-05, "loss": 0.9411, "step": 1403 }, { "epoch": 0.8982725527831094, "grad_norm": 2.3299989649786106, "learning_rate": 1.981289804848729e-05, "loss": 0.9638, "step": 1404 }, { "epoch": 0.8989123480486244, "grad_norm": 2.640654037856589, "learning_rate": 1.9812498872459526e-05, "loss": 0.9638, "step": 1405 }, { "epoch": 0.8995521433141395, "grad_norm": 2.2560098965298625, "learning_rate": 1.9812099275100516e-05, "loss": 0.9391, "step": 1406 }, { "epoch": 0.9001919385796545, "grad_norm": 2.1582715864368396, "learning_rate": 1.9811699256427416e-05, "loss": 0.926, "step": 1407 }, { "epoch": 0.9008317338451696, "grad_norm": 2.5966865370630416, "learning_rate": 1.9811298816457408e-05, "loss": 0.977, "step": 1408 }, { "epoch": 0.9014715291106846, "grad_norm": 2.3543918205080634, "learning_rate": 1.9810897955207676e-05, "loss": 0.9603, "step": 1409 }, { "epoch": 0.9021113243761996, "grad_norm": 2.1105817518576706, "learning_rate": 1.9810496672695442e-05, "loss": 0.96, "step": 1410 }, { "epoch": 0.9027511196417146, "grad_norm": 2.1900942707291415, "learning_rate": 1.981009496893793e-05, "loss": 0.9297, "step": 1411 }, { "epoch": 0.9033909149072297, "grad_norm": 2.6427547357869177, "learning_rate": 1.9809692843952396e-05, "loss": 0.9199, "step": 1412 }, { "epoch": 0.9040307101727447, "grad_norm": 2.741821584212075, "learning_rate": 1.9809290297756096e-05, "loss": 0.9615, "step": 1413 }, { "epoch": 0.9046705054382598, "grad_norm": 2.2157831605734213, "learning_rate": 1.9808887330366323e-05, "loss": 0.9676, "step": 1414 }, { "epoch": 0.9053103007037748, "grad_norm": 2.7134385280452613, "learning_rate": 1.980848394180037e-05, "loss": 0.9802, "step": 1415 }, { "epoch": 0.9059500959692899, "grad_norm": 2.470892006711248, "learning_rate": 1.980808013207557e-05, "loss": 0.9694, "step": 1416 }, { "epoch": 0.9065898912348048, "grad_norm": 2.364844759940977, "learning_rate": 1.9807675901209258e-05, "loss": 0.9564, "step": 1417 }, { "epoch": 0.9072296865003199, "grad_norm": 2.188065546213557, "learning_rate": 1.9807271249218787e-05, "loss": 0.8956, "step": 1418 }, { "epoch": 0.9078694817658349, "grad_norm": 2.3536813160899377, "learning_rate": 1.9806866176121537e-05, "loss": 0.9786, "step": 1419 }, { "epoch": 0.90850927703135, "grad_norm": 2.111812643052365, "learning_rate": 1.9806460681934894e-05, "loss": 0.912, "step": 1420 }, { "epoch": 0.909149072296865, "grad_norm": 2.401532640855597, "learning_rate": 1.9806054766676274e-05, "loss": 0.9745, "step": 1421 }, { "epoch": 0.9097888675623801, "grad_norm": 2.5331367248520165, "learning_rate": 1.980564843036311e-05, "loss": 0.9594, "step": 1422 }, { "epoch": 0.910428662827895, "grad_norm": 2.4021633996236975, "learning_rate": 1.9805241673012844e-05, "loss": 0.9785, "step": 1423 }, { "epoch": 0.9110684580934101, "grad_norm": 2.267574050962419, "learning_rate": 1.980483449464294e-05, "loss": 0.9807, "step": 1424 }, { "epoch": 0.9117082533589251, "grad_norm": 2.671307188750812, "learning_rate": 1.9804426895270885e-05, "loss": 0.9274, "step": 1425 }, { "epoch": 0.9123480486244402, "grad_norm": 2.597296464273354, "learning_rate": 1.9804018874914183e-05, "loss": 1.0182, "step": 1426 }, { "epoch": 0.9129878438899552, "grad_norm": 2.698830017630516, "learning_rate": 1.9803610433590347e-05, "loss": 0.9482, "step": 1427 }, { "epoch": 0.9136276391554703, "grad_norm": 2.2852963566410343, "learning_rate": 1.9803201571316923e-05, "loss": 0.9172, "step": 1428 }, { "epoch": 0.9142674344209852, "grad_norm": 2.3823006536581928, "learning_rate": 1.9802792288111458e-05, "loss": 0.9357, "step": 1429 }, { "epoch": 0.9149072296865003, "grad_norm": 2.220507103539851, "learning_rate": 1.9802382583991527e-05, "loss": 0.9792, "step": 1430 }, { "epoch": 0.9155470249520153, "grad_norm": 3.041191107669174, "learning_rate": 1.980197245897473e-05, "loss": 0.99, "step": 1431 }, { "epoch": 0.9161868202175304, "grad_norm": 2.47144714332395, "learning_rate": 1.980156191307867e-05, "loss": 0.9448, "step": 1432 }, { "epoch": 0.9168266154830454, "grad_norm": 2.9131031470319733, "learning_rate": 1.9801150946320976e-05, "loss": 0.9205, "step": 1433 }, { "epoch": 0.9174664107485605, "grad_norm": 2.7610421951949635, "learning_rate": 1.98007395587193e-05, "loss": 0.9532, "step": 1434 }, { "epoch": 0.9181062060140754, "grad_norm": 3.1939075643646864, "learning_rate": 1.9800327750291296e-05, "loss": 0.9368, "step": 1435 }, { "epoch": 0.9187460012795905, "grad_norm": 2.543279995606911, "learning_rate": 1.9799915521054656e-05, "loss": 0.9629, "step": 1436 }, { "epoch": 0.9193857965451055, "grad_norm": 2.6511005071393745, "learning_rate": 1.979950287102707e-05, "loss": 0.9969, "step": 1437 }, { "epoch": 0.9200255918106206, "grad_norm": 2.647584679198806, "learning_rate": 1.979908980022627e-05, "loss": 0.9323, "step": 1438 }, { "epoch": 0.9206653870761357, "grad_norm": 2.7906541614645395, "learning_rate": 1.9798676308669985e-05, "loss": 0.8685, "step": 1439 }, { "epoch": 0.9213051823416507, "grad_norm": 2.436762933133154, "learning_rate": 1.9798262396375965e-05, "loss": 0.9627, "step": 1440 }, { "epoch": 0.9219449776071658, "grad_norm": 3.183390681215855, "learning_rate": 1.9797848063361988e-05, "loss": 0.9181, "step": 1441 }, { "epoch": 0.9225847728726807, "grad_norm": 2.877098127069264, "learning_rate": 1.9797433309645848e-05, "loss": 1.0125, "step": 1442 }, { "epoch": 0.9232245681381958, "grad_norm": 2.3353769251559715, "learning_rate": 1.9797018135245344e-05, "loss": 0.9126, "step": 1443 }, { "epoch": 0.9238643634037108, "grad_norm": 2.8698663796205017, "learning_rate": 1.9796602540178313e-05, "loss": 0.99, "step": 1444 }, { "epoch": 0.9245041586692259, "grad_norm": 2.6274242663246152, "learning_rate": 1.9796186524462598e-05, "loss": 0.9186, "step": 1445 }, { "epoch": 0.9251439539347409, "grad_norm": 2.5910198021013304, "learning_rate": 1.9795770088116057e-05, "loss": 0.9734, "step": 1446 }, { "epoch": 0.925783749200256, "grad_norm": 2.6204614399853723, "learning_rate": 1.9795353231156574e-05, "loss": 0.9471, "step": 1447 }, { "epoch": 0.9264235444657709, "grad_norm": 2.448878300715096, "learning_rate": 1.9794935953602047e-05, "loss": 0.9795, "step": 1448 }, { "epoch": 0.927063339731286, "grad_norm": 2.2636119369014995, "learning_rate": 1.9794518255470393e-05, "loss": 0.9739, "step": 1449 }, { "epoch": 0.927703134996801, "grad_norm": 2.158658734689146, "learning_rate": 1.9794100136779552e-05, "loss": 0.9555, "step": 1450 }, { "epoch": 0.9283429302623161, "grad_norm": 2.413499291972092, "learning_rate": 1.9793681597547468e-05, "loss": 0.9745, "step": 1451 }, { "epoch": 0.9289827255278311, "grad_norm": 2.4570681290543486, "learning_rate": 1.9793262637792125e-05, "loss": 0.9849, "step": 1452 }, { "epoch": 0.9296225207933462, "grad_norm": 2.8626875422308657, "learning_rate": 1.9792843257531496e-05, "loss": 0.9441, "step": 1453 }, { "epoch": 0.9302623160588611, "grad_norm": 2.1503033510039167, "learning_rate": 1.9792423456783605e-05, "loss": 1.0375, "step": 1454 }, { "epoch": 0.9309021113243762, "grad_norm": 2.267555913893115, "learning_rate": 1.9792003235566467e-05, "loss": 0.9398, "step": 1455 }, { "epoch": 0.9315419065898912, "grad_norm": 2.143465232350783, "learning_rate": 1.979158259389813e-05, "loss": 0.9586, "step": 1456 }, { "epoch": 0.9321817018554063, "grad_norm": 2.1889290309031986, "learning_rate": 1.9791161531796654e-05, "loss": 0.897, "step": 1457 }, { "epoch": 0.9328214971209213, "grad_norm": 2.1810734755661842, "learning_rate": 1.9790740049280122e-05, "loss": 0.9242, "step": 1458 }, { "epoch": 0.9334612923864364, "grad_norm": 2.31655429696382, "learning_rate": 1.9790318146366624e-05, "loss": 0.9496, "step": 1459 }, { "epoch": 0.9341010876519513, "grad_norm": 2.4467931219879344, "learning_rate": 1.978989582307428e-05, "loss": 0.9387, "step": 1460 }, { "epoch": 0.9347408829174664, "grad_norm": 2.356004841022778, "learning_rate": 1.9789473079421228e-05, "loss": 0.9036, "step": 1461 }, { "epoch": 0.9353806781829814, "grad_norm": 2.790864761805249, "learning_rate": 1.9789049915425617e-05, "loss": 0.9745, "step": 1462 }, { "epoch": 0.9360204734484965, "grad_norm": 2.4480409970772765, "learning_rate": 1.978862633110561e-05, "loss": 0.9436, "step": 1463 }, { "epoch": 0.9366602687140115, "grad_norm": 2.5321234464676317, "learning_rate": 1.9788202326479404e-05, "loss": 0.9555, "step": 1464 }, { "epoch": 0.9373000639795266, "grad_norm": 2.3997151227799245, "learning_rate": 1.9787777901565204e-05, "loss": 1.0079, "step": 1465 }, { "epoch": 0.9379398592450416, "grad_norm": 2.263846526374863, "learning_rate": 1.9787353056381228e-05, "loss": 0.935, "step": 1466 }, { "epoch": 0.9385796545105566, "grad_norm": 2.311676949327423, "learning_rate": 1.9786927790945728e-05, "loss": 0.9617, "step": 1467 }, { "epoch": 0.9392194497760716, "grad_norm": 2.4124324190037516, "learning_rate": 1.9786502105276952e-05, "loss": 0.9344, "step": 1468 }, { "epoch": 0.9398592450415867, "grad_norm": 2.3259076601141233, "learning_rate": 1.9786075999393187e-05, "loss": 0.9576, "step": 1469 }, { "epoch": 0.9404990403071017, "grad_norm": 2.3935028255224022, "learning_rate": 1.9785649473312727e-05, "loss": 0.9359, "step": 1470 }, { "epoch": 0.9411388355726168, "grad_norm": 2.067176344643817, "learning_rate": 1.9785222527053887e-05, "loss": 0.907, "step": 1471 }, { "epoch": 0.9417786308381318, "grad_norm": 2.5360667601699936, "learning_rate": 1.9784795160634994e-05, "loss": 0.9586, "step": 1472 }, { "epoch": 0.9424184261036468, "grad_norm": 2.2381091485533875, "learning_rate": 1.9784367374074408e-05, "loss": 0.9234, "step": 1473 }, { "epoch": 0.9430582213691618, "grad_norm": 2.6596023603258288, "learning_rate": 1.978393916739049e-05, "loss": 0.9115, "step": 1474 }, { "epoch": 0.9436980166346769, "grad_norm": 2.214492942829724, "learning_rate": 1.9783510540601626e-05, "loss": 0.9341, "step": 1475 }, { "epoch": 0.944337811900192, "grad_norm": 2.054688349240105, "learning_rate": 1.9783081493726227e-05, "loss": 0.9687, "step": 1476 }, { "epoch": 0.944977607165707, "grad_norm": 2.317682566343241, "learning_rate": 1.9782652026782706e-05, "loss": 0.9039, "step": 1477 }, { "epoch": 0.945617402431222, "grad_norm": 2.486952676413047, "learning_rate": 1.9782222139789512e-05, "loss": 0.9083, "step": 1478 }, { "epoch": 0.946257197696737, "grad_norm": 2.4776184823840213, "learning_rate": 1.97817918327651e-05, "loss": 0.9725, "step": 1479 }, { "epoch": 0.946896992962252, "grad_norm": 2.703958049223321, "learning_rate": 1.978136110572795e-05, "loss": 0.9517, "step": 1480 }, { "epoch": 0.9475367882277671, "grad_norm": 2.610311959366463, "learning_rate": 1.978092995869655e-05, "loss": 0.9671, "step": 1481 }, { "epoch": 0.9481765834932822, "grad_norm": 2.204231771444204, "learning_rate": 1.978049839168942e-05, "loss": 0.9539, "step": 1482 }, { "epoch": 0.9488163787587972, "grad_norm": 2.4743940377287883, "learning_rate": 1.9780066404725086e-05, "loss": 0.9453, "step": 1483 }, { "epoch": 0.9494561740243123, "grad_norm": 2.804263202891838, "learning_rate": 1.9779633997822098e-05, "loss": 0.8828, "step": 1484 }, { "epoch": 0.9500959692898272, "grad_norm": 2.1446905173189172, "learning_rate": 1.977920117099902e-05, "loss": 0.9918, "step": 1485 }, { "epoch": 0.9507357645553423, "grad_norm": 2.5776872549205136, "learning_rate": 1.977876792427444e-05, "loss": 0.9485, "step": 1486 }, { "epoch": 0.9513755598208573, "grad_norm": 2.4400180313871713, "learning_rate": 1.977833425766696e-05, "loss": 0.9245, "step": 1487 }, { "epoch": 0.9520153550863724, "grad_norm": 2.251306459777591, "learning_rate": 1.9777900171195207e-05, "loss": 0.9421, "step": 1488 }, { "epoch": 0.9526551503518874, "grad_norm": 2.577730520072173, "learning_rate": 1.977746566487781e-05, "loss": 0.9058, "step": 1489 }, { "epoch": 0.9532949456174025, "grad_norm": 2.678223809059624, "learning_rate": 1.977703073873343e-05, "loss": 0.9968, "step": 1490 }, { "epoch": 0.9539347408829175, "grad_norm": 2.5594505117472557, "learning_rate": 1.9776595392780744e-05, "loss": 0.9315, "step": 1491 }, { "epoch": 0.9545745361484325, "grad_norm": 2.4220317730384355, "learning_rate": 1.977615962703844e-05, "loss": 0.9559, "step": 1492 }, { "epoch": 0.9552143314139475, "grad_norm": 2.2621246698616893, "learning_rate": 1.9775723441525235e-05, "loss": 0.9253, "step": 1493 }, { "epoch": 0.9558541266794626, "grad_norm": 2.4208314201235672, "learning_rate": 1.9775286836259856e-05, "loss": 0.9287, "step": 1494 }, { "epoch": 0.9564939219449776, "grad_norm": 2.562448366755688, "learning_rate": 1.9774849811261045e-05, "loss": 0.8921, "step": 1495 }, { "epoch": 0.9571337172104927, "grad_norm": 2.3830520740600925, "learning_rate": 1.9774412366547576e-05, "loss": 0.884, "step": 1496 }, { "epoch": 0.9577735124760077, "grad_norm": 2.4771475712433753, "learning_rate": 1.9773974502138224e-05, "loss": 0.9536, "step": 1497 }, { "epoch": 0.9584133077415227, "grad_norm": 2.653917682301746, "learning_rate": 1.9773536218051795e-05, "loss": 0.8996, "step": 1498 }, { "epoch": 0.9590531030070377, "grad_norm": 3.361543116316221, "learning_rate": 1.9773097514307103e-05, "loss": 0.9406, "step": 1499 }, { "epoch": 0.9596928982725528, "grad_norm": 3.494092257565959, "learning_rate": 1.977265839092299e-05, "loss": 0.9775, "step": 1500 }, { "epoch": 0.9603326935380678, "grad_norm": 2.4297009780168723, "learning_rate": 1.9772218847918313e-05, "loss": 0.9278, "step": 1501 }, { "epoch": 0.9609724888035829, "grad_norm": 2.9083395023609393, "learning_rate": 1.9771778885311937e-05, "loss": 0.9137, "step": 1502 }, { "epoch": 0.9616122840690979, "grad_norm": 3.3798329780284417, "learning_rate": 1.9771338503122767e-05, "loss": 0.9515, "step": 1503 }, { "epoch": 0.9622520793346129, "grad_norm": 2.5539845515892408, "learning_rate": 1.9770897701369697e-05, "loss": 0.9079, "step": 1504 }, { "epoch": 0.9628918746001279, "grad_norm": 2.573214047535533, "learning_rate": 1.9770456480071664e-05, "loss": 0.9742, "step": 1505 }, { "epoch": 0.963531669865643, "grad_norm": 2.3707065725200303, "learning_rate": 1.9770014839247604e-05, "loss": 0.982, "step": 1506 }, { "epoch": 0.964171465131158, "grad_norm": 2.76651137148857, "learning_rate": 1.976957277891649e-05, "loss": 0.9813, "step": 1507 }, { "epoch": 0.9648112603966731, "grad_norm": 2.801967737349366, "learning_rate": 1.97691302990973e-05, "loss": 0.9472, "step": 1508 }, { "epoch": 0.9654510556621881, "grad_norm": 2.7109753703841375, "learning_rate": 1.9768687399809035e-05, "loss": 0.9536, "step": 1509 }, { "epoch": 0.9660908509277031, "grad_norm": 2.2694457822089205, "learning_rate": 1.9768244081070708e-05, "loss": 0.9008, "step": 1510 }, { "epoch": 0.9667306461932181, "grad_norm": 2.667909568833561, "learning_rate": 1.9767800342901357e-05, "loss": 0.9529, "step": 1511 }, { "epoch": 0.9673704414587332, "grad_norm": 2.3320478411307453, "learning_rate": 1.9767356185320034e-05, "loss": 0.96, "step": 1512 }, { "epoch": 0.9680102367242482, "grad_norm": 2.4367392530470915, "learning_rate": 1.976691160834581e-05, "loss": 0.9168, "step": 1513 }, { "epoch": 0.9686500319897633, "grad_norm": 2.307307170294225, "learning_rate": 1.976646661199778e-05, "loss": 0.9573, "step": 1514 }, { "epoch": 0.9692898272552783, "grad_norm": 2.4710686967056414, "learning_rate": 1.976602119629504e-05, "loss": 0.9161, "step": 1515 }, { "epoch": 0.9699296225207934, "grad_norm": 2.1098811036723326, "learning_rate": 1.9765575361256727e-05, "loss": 0.9071, "step": 1516 }, { "epoch": 0.9705694177863083, "grad_norm": 2.729393027631854, "learning_rate": 1.976512910690198e-05, "loss": 0.9126, "step": 1517 }, { "epoch": 0.9712092130518234, "grad_norm": 2.617194283073077, "learning_rate": 1.9764682433249958e-05, "loss": 0.9247, "step": 1518 }, { "epoch": 0.9718490083173384, "grad_norm": 2.189815398246983, "learning_rate": 1.9764235340319842e-05, "loss": 0.9907, "step": 1519 }, { "epoch": 0.9724888035828535, "grad_norm": 2.375190695883395, "learning_rate": 1.976378782813083e-05, "loss": 0.9379, "step": 1520 }, { "epoch": 0.9731285988483686, "grad_norm": 2.6370170546621496, "learning_rate": 1.9763339896702137e-05, "loss": 0.8809, "step": 1521 }, { "epoch": 0.9737683941138836, "grad_norm": 2.412183979405243, "learning_rate": 1.9762891546053e-05, "loss": 0.9465, "step": 1522 }, { "epoch": 0.9744081893793985, "grad_norm": 2.2649328075019888, "learning_rate": 1.9762442776202664e-05, "loss": 0.9017, "step": 1523 }, { "epoch": 0.9750479846449136, "grad_norm": 2.439712506931133, "learning_rate": 1.97619935871704e-05, "loss": 0.9035, "step": 1524 }, { "epoch": 0.9756877799104287, "grad_norm": 2.129033777700421, "learning_rate": 1.9761543978975498e-05, "loss": 0.9888, "step": 1525 }, { "epoch": 0.9763275751759437, "grad_norm": 2.4797553837030195, "learning_rate": 1.9761093951637265e-05, "loss": 0.9364, "step": 1526 }, { "epoch": 0.9769673704414588, "grad_norm": 2.6258974322424535, "learning_rate": 1.9760643505175016e-05, "loss": 0.8783, "step": 1527 }, { "epoch": 0.9776071657069738, "grad_norm": 2.6873326116101754, "learning_rate": 1.9760192639608103e-05, "loss": 0.9381, "step": 1528 }, { "epoch": 0.9782469609724888, "grad_norm": 2.7941802656895587, "learning_rate": 1.9759741354955874e-05, "loss": 0.9718, "step": 1529 }, { "epoch": 0.9788867562380038, "grad_norm": 2.6572955767556707, "learning_rate": 1.9759289651237716e-05, "loss": 0.9411, "step": 1530 }, { "epoch": 0.9795265515035189, "grad_norm": 2.127908416186634, "learning_rate": 1.975883752847302e-05, "loss": 0.9166, "step": 1531 }, { "epoch": 0.9801663467690339, "grad_norm": 2.4407109904190776, "learning_rate": 1.9758384986681202e-05, "loss": 0.9165, "step": 1532 }, { "epoch": 0.980806142034549, "grad_norm": 2.51651504028775, "learning_rate": 1.9757932025881688e-05, "loss": 0.9592, "step": 1533 }, { "epoch": 0.981445937300064, "grad_norm": 2.339319925095269, "learning_rate": 1.9757478646093935e-05, "loss": 0.9305, "step": 1534 }, { "epoch": 0.982085732565579, "grad_norm": 2.3311198665106034, "learning_rate": 1.9757024847337404e-05, "loss": 0.9022, "step": 1535 }, { "epoch": 0.982725527831094, "grad_norm": 2.3639428014923793, "learning_rate": 1.975657062963158e-05, "loss": 0.9289, "step": 1536 }, { "epoch": 0.9833653230966091, "grad_norm": 2.623097039807324, "learning_rate": 1.9756115992995975e-05, "loss": 0.9178, "step": 1537 }, { "epoch": 0.9840051183621241, "grad_norm": 2.67458287690616, "learning_rate": 1.9755660937450098e-05, "loss": 0.9223, "step": 1538 }, { "epoch": 0.9846449136276392, "grad_norm": 2.3047758427920306, "learning_rate": 1.9755205463013495e-05, "loss": 0.8934, "step": 1539 }, { "epoch": 0.9852847088931542, "grad_norm": 2.2515124837729816, "learning_rate": 1.9754749569705727e-05, "loss": 0.9162, "step": 1540 }, { "epoch": 0.9859245041586693, "grad_norm": 2.3723316212155794, "learning_rate": 1.975429325754636e-05, "loss": 0.9422, "step": 1541 }, { "epoch": 0.9865642994241842, "grad_norm": 2.3515821079796138, "learning_rate": 1.9753836526554995e-05, "loss": 0.9331, "step": 1542 }, { "epoch": 0.9872040946896993, "grad_norm": 2.8040154102569024, "learning_rate": 1.975337937675124e-05, "loss": 0.9182, "step": 1543 }, { "epoch": 0.9878438899552143, "grad_norm": 2.509217282243699, "learning_rate": 1.9752921808154727e-05, "loss": 0.9732, "step": 1544 }, { "epoch": 0.9884836852207294, "grad_norm": 2.6590204554102717, "learning_rate": 1.9752463820785093e-05, "loss": 0.9506, "step": 1545 }, { "epoch": 0.9891234804862444, "grad_norm": 2.585374201785161, "learning_rate": 1.9752005414662015e-05, "loss": 0.9187, "step": 1546 }, { "epoch": 0.9897632757517595, "grad_norm": 2.6749957598653746, "learning_rate": 1.9751546589805173e-05, "loss": 0.9625, "step": 1547 }, { "epoch": 0.9904030710172744, "grad_norm": 2.8574962030052458, "learning_rate": 1.9751087346234267e-05, "loss": 0.9649, "step": 1548 }, { "epoch": 0.9910428662827895, "grad_norm": 2.2895234599301055, "learning_rate": 1.9750627683969016e-05, "loss": 0.9208, "step": 1549 }, { "epoch": 0.9916826615483045, "grad_norm": 2.5405900376371386, "learning_rate": 1.9750167603029155e-05, "loss": 0.9551, "step": 1550 }, { "epoch": 0.9923224568138196, "grad_norm": 2.731931011145661, "learning_rate": 1.9749707103434447e-05, "loss": 0.884, "step": 1551 }, { "epoch": 0.9929622520793346, "grad_norm": 2.292016149629074, "learning_rate": 1.9749246185204655e-05, "loss": 0.9429, "step": 1552 }, { "epoch": 0.9936020473448497, "grad_norm": 2.7525914453820337, "learning_rate": 1.9748784848359573e-05, "loss": 0.9722, "step": 1553 }, { "epoch": 0.9942418426103646, "grad_norm": 2.6223398853780298, "learning_rate": 1.9748323092919014e-05, "loss": 0.9206, "step": 1554 }, { "epoch": 0.9948816378758797, "grad_norm": 2.9941617790562014, "learning_rate": 1.97478609189028e-05, "loss": 0.9273, "step": 1555 }, { "epoch": 0.9955214331413947, "grad_norm": 2.4481835731167876, "learning_rate": 1.974739832633078e-05, "loss": 0.8722, "step": 1556 }, { "epoch": 0.9961612284069098, "grad_norm": 2.715648291493076, "learning_rate": 1.9746935315222818e-05, "loss": 0.9143, "step": 1557 }, { "epoch": 0.9968010236724248, "grad_norm": 2.2183803841315415, "learning_rate": 1.9746471885598786e-05, "loss": 0.9285, "step": 1558 }, { "epoch": 0.9974408189379399, "grad_norm": 2.3926480695116843, "learning_rate": 1.974600803747859e-05, "loss": 0.9164, "step": 1559 }, { "epoch": 0.9980806142034548, "grad_norm": 2.705966950785872, "learning_rate": 1.974554377088215e-05, "loss": 0.967, "step": 1560 }, { "epoch": 0.9987204094689699, "grad_norm": 2.2851326709303774, "learning_rate": 1.9745079085829395e-05, "loss": 0.8904, "step": 1561 }, { "epoch": 0.999360204734485, "grad_norm": 2.3364103879283045, "learning_rate": 1.974461398234028e-05, "loss": 0.9236, "step": 1562 }, { "epoch": 1.0, "grad_norm": 2.761119180280796, "learning_rate": 1.974414846043477e-05, "loss": 0.9183, "step": 1563 }, { "epoch": 1.000639795265515, "grad_norm": 2.9768332689274697, "learning_rate": 1.9743682520132863e-05, "loss": 0.8507, "step": 1564 }, { "epoch": 1.00127959053103, "grad_norm": 2.2417655801415837, "learning_rate": 1.974321616145456e-05, "loss": 0.8211, "step": 1565 }, { "epoch": 1.0019193857965452, "grad_norm": 2.253303434720466, "learning_rate": 1.9742749384419886e-05, "loss": 0.8592, "step": 1566 }, { "epoch": 1.0025591810620602, "grad_norm": 2.633784423759442, "learning_rate": 1.9742282189048887e-05, "loss": 0.8422, "step": 1567 }, { "epoch": 1.0031989763275753, "grad_norm": 2.5898850062459458, "learning_rate": 1.9741814575361617e-05, "loss": 0.8876, "step": 1568 }, { "epoch": 1.0038387715930903, "grad_norm": 2.7215922762348383, "learning_rate": 1.974134654337816e-05, "loss": 0.877, "step": 1569 }, { "epoch": 1.0044785668586051, "grad_norm": 2.282122461750821, "learning_rate": 1.9740878093118614e-05, "loss": 0.7849, "step": 1570 }, { "epoch": 1.0051183621241202, "grad_norm": 2.277004336496517, "learning_rate": 1.9740409224603088e-05, "loss": 0.8701, "step": 1571 }, { "epoch": 1.0057581573896353, "grad_norm": 2.969092210093472, "learning_rate": 1.973993993785172e-05, "loss": 0.8967, "step": 1572 }, { "epoch": 1.0063979526551503, "grad_norm": 2.61479667527491, "learning_rate": 1.973947023288465e-05, "loss": 0.8874, "step": 1573 }, { "epoch": 1.0070377479206654, "grad_norm": 2.494579596631536, "learning_rate": 1.9739000109722058e-05, "loss": 0.8616, "step": 1574 }, { "epoch": 1.0076775431861804, "grad_norm": 2.676990798039126, "learning_rate": 1.9738529568384122e-05, "loss": 0.8227, "step": 1575 }, { "epoch": 1.0083173384516955, "grad_norm": 2.679774064935096, "learning_rate": 1.9738058608891056e-05, "loss": 0.82, "step": 1576 }, { "epoch": 1.0089571337172105, "grad_norm": 2.336313382730229, "learning_rate": 1.973758723126307e-05, "loss": 0.8364, "step": 1577 }, { "epoch": 1.0095969289827256, "grad_norm": 2.3948281775951394, "learning_rate": 1.9737115435520417e-05, "loss": 0.857, "step": 1578 }, { "epoch": 1.0102367242482406, "grad_norm": 2.206499318714554, "learning_rate": 1.9736643221683343e-05, "loss": 0.7957, "step": 1579 }, { "epoch": 1.0108765195137557, "grad_norm": 2.71794071998503, "learning_rate": 1.973617058977213e-05, "loss": 0.8398, "step": 1580 }, { "epoch": 1.0115163147792707, "grad_norm": 2.850002310759826, "learning_rate": 1.973569753980707e-05, "loss": 0.8429, "step": 1581 }, { "epoch": 1.0121561100447856, "grad_norm": 2.3076906450381562, "learning_rate": 1.9735224071808478e-05, "loss": 0.8347, "step": 1582 }, { "epoch": 1.0127959053103006, "grad_norm": 2.39992449192474, "learning_rate": 1.973475018579668e-05, "loss": 0.7777, "step": 1583 }, { "epoch": 1.0134357005758157, "grad_norm": 2.620447968833844, "learning_rate": 1.973427588179203e-05, "loss": 0.8232, "step": 1584 }, { "epoch": 1.0140754958413307, "grad_norm": 2.35068812455199, "learning_rate": 1.9733801159814883e-05, "loss": 0.8005, "step": 1585 }, { "epoch": 1.0147152911068458, "grad_norm": 2.2797825536352376, "learning_rate": 1.9733326019885634e-05, "loss": 0.7624, "step": 1586 }, { "epoch": 1.0153550863723608, "grad_norm": 2.4206653891405563, "learning_rate": 1.973285046202468e-05, "loss": 0.8544, "step": 1587 }, { "epoch": 1.0159948816378759, "grad_norm": 2.5465419922205883, "learning_rate": 1.973237448625244e-05, "loss": 0.8152, "step": 1588 }, { "epoch": 1.016634676903391, "grad_norm": 2.631490511471477, "learning_rate": 1.9731898092589352e-05, "loss": 0.8759, "step": 1589 }, { "epoch": 1.017274472168906, "grad_norm": 2.6723889676532897, "learning_rate": 1.973142128105587e-05, "loss": 0.8763, "step": 1590 }, { "epoch": 1.017914267434421, "grad_norm": 2.3732524586367365, "learning_rate": 1.973094405167247e-05, "loss": 0.8437, "step": 1591 }, { "epoch": 1.018554062699936, "grad_norm": 2.3367735325209398, "learning_rate": 1.9730466404459646e-05, "loss": 0.9026, "step": 1592 }, { "epoch": 1.0191938579654511, "grad_norm": 2.2398776531925666, "learning_rate": 1.97299883394379e-05, "loss": 0.8524, "step": 1593 }, { "epoch": 1.0198336532309662, "grad_norm": 2.2525099030802065, "learning_rate": 1.9729509856627763e-05, "loss": 0.8663, "step": 1594 }, { "epoch": 1.020473448496481, "grad_norm": 2.4507629594317426, "learning_rate": 1.972903095604978e-05, "loss": 0.8583, "step": 1595 }, { "epoch": 1.021113243761996, "grad_norm": 2.086287088730467, "learning_rate": 1.9728551637724515e-05, "loss": 0.8253, "step": 1596 }, { "epoch": 1.0217530390275111, "grad_norm": 2.8886376626496246, "learning_rate": 1.9728071901672548e-05, "loss": 0.8571, "step": 1597 }, { "epoch": 1.0223928342930262, "grad_norm": 2.6191208578261587, "learning_rate": 1.972759174791448e-05, "loss": 0.8096, "step": 1598 }, { "epoch": 1.0230326295585412, "grad_norm": 2.649598263456148, "learning_rate": 1.9727111176470925e-05, "loss": 0.8193, "step": 1599 }, { "epoch": 1.0236724248240563, "grad_norm": 2.1193003929505827, "learning_rate": 1.972663018736252e-05, "loss": 0.8643, "step": 1600 }, { "epoch": 1.0243122200895713, "grad_norm": 2.547840100020664, "learning_rate": 1.9726148780609916e-05, "loss": 0.7875, "step": 1601 }, { "epoch": 1.0249520153550864, "grad_norm": 2.4213834131482406, "learning_rate": 1.9725666956233787e-05, "loss": 0.8768, "step": 1602 }, { "epoch": 1.0255918106206015, "grad_norm": 2.6504843863878564, "learning_rate": 1.9725184714254815e-05, "loss": 0.819, "step": 1603 }, { "epoch": 1.0262316058861165, "grad_norm": 2.193108604498024, "learning_rate": 1.9724702054693717e-05, "loss": 0.8593, "step": 1604 }, { "epoch": 1.0268714011516316, "grad_norm": 2.957469629927831, "learning_rate": 1.9724218977571207e-05, "loss": 0.8321, "step": 1605 }, { "epoch": 1.0275111964171466, "grad_norm": 2.378450832799987, "learning_rate": 1.9723735482908037e-05, "loss": 0.8052, "step": 1606 }, { "epoch": 1.0281509916826614, "grad_norm": 2.0925729340548145, "learning_rate": 1.9723251570724956e-05, "loss": 0.8146, "step": 1607 }, { "epoch": 1.0287907869481765, "grad_norm": 2.172501772238251, "learning_rate": 1.9722767241042754e-05, "loss": 0.8379, "step": 1608 }, { "epoch": 1.0294305822136915, "grad_norm": 3.1141323046763167, "learning_rate": 1.972228249388222e-05, "loss": 0.8489, "step": 1609 }, { "epoch": 1.0300703774792066, "grad_norm": 2.7122882801482153, "learning_rate": 1.9721797329264173e-05, "loss": 0.8308, "step": 1610 }, { "epoch": 1.0307101727447217, "grad_norm": 2.151666212303799, "learning_rate": 1.9721311747209437e-05, "loss": 0.8122, "step": 1611 }, { "epoch": 1.0313499680102367, "grad_norm": 2.050711765504288, "learning_rate": 1.972082574773887e-05, "loss": 0.7869, "step": 1612 }, { "epoch": 1.0319897632757518, "grad_norm": 2.614500873767797, "learning_rate": 1.9720339330873338e-05, "loss": 0.7775, "step": 1613 }, { "epoch": 1.0326295585412668, "grad_norm": 2.664334093968806, "learning_rate": 1.971985249663373e-05, "loss": 0.8388, "step": 1614 }, { "epoch": 1.0332693538067819, "grad_norm": 2.229553904047902, "learning_rate": 1.971936524504094e-05, "loss": 0.8216, "step": 1615 }, { "epoch": 1.033909149072297, "grad_norm": 2.642759271860383, "learning_rate": 1.97188775761159e-05, "loss": 0.8271, "step": 1616 }, { "epoch": 1.034548944337812, "grad_norm": 2.1694009491314397, "learning_rate": 1.9718389489879544e-05, "loss": 0.8537, "step": 1617 }, { "epoch": 1.035188739603327, "grad_norm": 2.457343445455418, "learning_rate": 1.971790098635283e-05, "loss": 0.8428, "step": 1618 }, { "epoch": 1.035828534868842, "grad_norm": 2.7697547077480653, "learning_rate": 1.9717412065556736e-05, "loss": 0.8467, "step": 1619 }, { "epoch": 1.036468330134357, "grad_norm": 2.503012679083421, "learning_rate": 1.9716922727512253e-05, "loss": 0.8378, "step": 1620 }, { "epoch": 1.037108125399872, "grad_norm": 2.4611257007787524, "learning_rate": 1.9716432972240393e-05, "loss": 0.8001, "step": 1621 }, { "epoch": 1.037747920665387, "grad_norm": 2.5552808146662156, "learning_rate": 1.9715942799762183e-05, "loss": 0.8231, "step": 1622 }, { "epoch": 1.038387715930902, "grad_norm": 2.927397920807121, "learning_rate": 1.971545221009868e-05, "loss": 0.806, "step": 1623 }, { "epoch": 1.0390275111964171, "grad_norm": 2.6567828602737245, "learning_rate": 1.9714961203270936e-05, "loss": 0.827, "step": 1624 }, { "epoch": 1.0396673064619322, "grad_norm": 2.8497629328271095, "learning_rate": 1.971446977930004e-05, "loss": 0.8578, "step": 1625 }, { "epoch": 1.0403071017274472, "grad_norm": 2.5380212921198204, "learning_rate": 1.971397793820709e-05, "loss": 0.8108, "step": 1626 }, { "epoch": 1.0409468969929623, "grad_norm": 2.3102994953088016, "learning_rate": 1.971348568001321e-05, "loss": 0.8672, "step": 1627 }, { "epoch": 1.0415866922584773, "grad_norm": 2.593635178120536, "learning_rate": 1.9712993004739532e-05, "loss": 0.8806, "step": 1628 }, { "epoch": 1.0422264875239924, "grad_norm": 2.3984918312606203, "learning_rate": 1.9712499912407213e-05, "loss": 0.8216, "step": 1629 }, { "epoch": 1.0428662827895074, "grad_norm": 2.1238313065090413, "learning_rate": 1.9712006403037426e-05, "loss": 0.8949, "step": 1630 }, { "epoch": 1.0435060780550225, "grad_norm": 2.338425226837913, "learning_rate": 1.971151247665136e-05, "loss": 0.8072, "step": 1631 }, { "epoch": 1.0441458733205373, "grad_norm": 2.2724798065127807, "learning_rate": 1.9711018133270227e-05, "loss": 0.8041, "step": 1632 }, { "epoch": 1.0447856685860524, "grad_norm": 2.731864978244014, "learning_rate": 1.9710523372915244e-05, "loss": 0.8627, "step": 1633 }, { "epoch": 1.0454254638515674, "grad_norm": 2.2290988834533727, "learning_rate": 1.9710028195607666e-05, "loss": 0.8312, "step": 1634 }, { "epoch": 1.0460652591170825, "grad_norm": 2.5099806731745358, "learning_rate": 1.970953260136875e-05, "loss": 0.8435, "step": 1635 }, { "epoch": 1.0467050543825975, "grad_norm": 2.603512838560982, "learning_rate": 1.9709036590219773e-05, "loss": 0.8902, "step": 1636 }, { "epoch": 1.0473448496481126, "grad_norm": 2.824242365416381, "learning_rate": 1.9708540162182034e-05, "loss": 0.866, "step": 1637 }, { "epoch": 1.0479846449136276, "grad_norm": 2.661333054824249, "learning_rate": 1.9708043317276854e-05, "loss": 0.8773, "step": 1638 }, { "epoch": 1.0486244401791427, "grad_norm": 2.213517619225951, "learning_rate": 1.9707546055525564e-05, "loss": 0.8237, "step": 1639 }, { "epoch": 1.0492642354446577, "grad_norm": 2.3781069843009086, "learning_rate": 1.9707048376949513e-05, "loss": 0.8214, "step": 1640 }, { "epoch": 1.0499040307101728, "grad_norm": 2.7838402730655, "learning_rate": 1.9706550281570074e-05, "loss": 0.8675, "step": 1641 }, { "epoch": 1.0505438259756879, "grad_norm": 2.7388066191283387, "learning_rate": 1.970605176940863e-05, "loss": 0.8298, "step": 1642 }, { "epoch": 1.051183621241203, "grad_norm": 2.4820054414138317, "learning_rate": 1.970555284048659e-05, "loss": 0.8135, "step": 1643 }, { "epoch": 1.051823416506718, "grad_norm": 2.419693752274172, "learning_rate": 1.9705053494825376e-05, "loss": 0.8045, "step": 1644 }, { "epoch": 1.0524632117722328, "grad_norm": 2.13921664141024, "learning_rate": 1.9704553732446428e-05, "loss": 0.8368, "step": 1645 }, { "epoch": 1.0531030070377478, "grad_norm": 2.4508053355823685, "learning_rate": 1.9704053553371207e-05, "loss": 0.8927, "step": 1646 }, { "epoch": 1.053742802303263, "grad_norm": 2.5757585631503783, "learning_rate": 1.9703552957621187e-05, "loss": 0.8456, "step": 1647 }, { "epoch": 1.054382597568778, "grad_norm": 2.1139490560571264, "learning_rate": 1.9703051945217866e-05, "loss": 0.8794, "step": 1648 }, { "epoch": 1.055022392834293, "grad_norm": 2.329336226086311, "learning_rate": 1.9702550516182756e-05, "loss": 0.85, "step": 1649 }, { "epoch": 1.055662188099808, "grad_norm": 2.5923515938189814, "learning_rate": 1.970204867053738e-05, "loss": 0.8733, "step": 1650 }, { "epoch": 1.056301983365323, "grad_norm": 2.0472959565618147, "learning_rate": 1.97015464083033e-05, "loss": 0.8239, "step": 1651 }, { "epoch": 1.0569417786308382, "grad_norm": 2.3511175301088674, "learning_rate": 1.970104372950207e-05, "loss": 0.8296, "step": 1652 }, { "epoch": 1.0575815738963532, "grad_norm": 2.4961068887783213, "learning_rate": 1.9700540634155283e-05, "loss": 0.8421, "step": 1653 }, { "epoch": 1.0582213691618683, "grad_norm": 2.7280900560391084, "learning_rate": 1.9700037122284533e-05, "loss": 0.8869, "step": 1654 }, { "epoch": 1.0588611644273833, "grad_norm": 2.474700003181345, "learning_rate": 1.969953319391145e-05, "loss": 0.8115, "step": 1655 }, { "epoch": 1.0595009596928984, "grad_norm": 2.600631277126323, "learning_rate": 1.969902884905766e-05, "loss": 0.8254, "step": 1656 }, { "epoch": 1.0601407549584132, "grad_norm": 2.2074321940365884, "learning_rate": 1.9698524087744828e-05, "loss": 0.8097, "step": 1657 }, { "epoch": 1.0607805502239283, "grad_norm": 2.2466247484142645, "learning_rate": 1.9698018909994623e-05, "loss": 0.8176, "step": 1658 }, { "epoch": 1.0614203454894433, "grad_norm": 2.746412778815921, "learning_rate": 1.9697513315828734e-05, "loss": 0.8624, "step": 1659 }, { "epoch": 1.0620601407549584, "grad_norm": 2.3507788787280615, "learning_rate": 1.969700730526888e-05, "loss": 0.812, "step": 1660 }, { "epoch": 1.0626999360204734, "grad_norm": 2.2029848865488924, "learning_rate": 1.9696500878336778e-05, "loss": 0.8692, "step": 1661 }, { "epoch": 1.0633397312859885, "grad_norm": 2.3458618229384522, "learning_rate": 1.9695994035054177e-05, "loss": 0.866, "step": 1662 }, { "epoch": 1.0639795265515035, "grad_norm": 2.4543691447289113, "learning_rate": 1.969548677544284e-05, "loss": 0.8294, "step": 1663 }, { "epoch": 1.0646193218170186, "grad_norm": 2.928974009007136, "learning_rate": 1.9694979099524548e-05, "loss": 0.8716, "step": 1664 }, { "epoch": 1.0652591170825336, "grad_norm": 2.396251678333461, "learning_rate": 1.9694471007321102e-05, "loss": 0.8446, "step": 1665 }, { "epoch": 1.0658989123480487, "grad_norm": 2.2580327998622622, "learning_rate": 1.9693962498854312e-05, "loss": 0.8425, "step": 1666 }, { "epoch": 1.0665387076135637, "grad_norm": 2.3082822159803023, "learning_rate": 1.9693453574146018e-05, "loss": 0.8082, "step": 1667 }, { "epoch": 1.0671785028790788, "grad_norm": 2.3700341735873534, "learning_rate": 1.9692944233218074e-05, "loss": 0.8387, "step": 1668 }, { "epoch": 1.0678182981445938, "grad_norm": 2.6080684989520955, "learning_rate": 1.9692434476092345e-05, "loss": 0.902, "step": 1669 }, { "epoch": 1.0684580934101087, "grad_norm": 2.2939094144505945, "learning_rate": 1.969192430279072e-05, "loss": 0.8561, "step": 1670 }, { "epoch": 1.0690978886756237, "grad_norm": 2.423041546996079, "learning_rate": 1.9691413713335107e-05, "loss": 0.861, "step": 1671 }, { "epoch": 1.0697376839411388, "grad_norm": 2.2883788811075134, "learning_rate": 1.969090270774743e-05, "loss": 0.8222, "step": 1672 }, { "epoch": 1.0703774792066538, "grad_norm": 2.182471783815088, "learning_rate": 1.9690391286049625e-05, "loss": 0.8186, "step": 1673 }, { "epoch": 1.0710172744721689, "grad_norm": 2.2966091953790064, "learning_rate": 1.968987944826366e-05, "loss": 0.8684, "step": 1674 }, { "epoch": 1.071657069737684, "grad_norm": 2.2228119850372883, "learning_rate": 1.9689367194411505e-05, "loss": 0.8207, "step": 1675 }, { "epoch": 1.072296865003199, "grad_norm": 2.316805344198216, "learning_rate": 1.9688854524515162e-05, "loss": 0.839, "step": 1676 }, { "epoch": 1.072936660268714, "grad_norm": 2.3802005643708717, "learning_rate": 1.968834143859664e-05, "loss": 0.8443, "step": 1677 }, { "epoch": 1.073576455534229, "grad_norm": 2.2746558666586867, "learning_rate": 1.968782793667797e-05, "loss": 0.8098, "step": 1678 }, { "epoch": 1.0742162507997441, "grad_norm": 2.548437487364569, "learning_rate": 1.9687314018781203e-05, "loss": 0.8692, "step": 1679 }, { "epoch": 1.0748560460652592, "grad_norm": 2.4246529806984523, "learning_rate": 1.9686799684928405e-05, "loss": 0.8605, "step": 1680 }, { "epoch": 1.0754958413307742, "grad_norm": 2.003959877888315, "learning_rate": 1.968628493514166e-05, "loss": 0.828, "step": 1681 }, { "epoch": 1.0761356365962893, "grad_norm": 2.4059169863082794, "learning_rate": 1.9685769769443068e-05, "loss": 0.8206, "step": 1682 }, { "epoch": 1.0767754318618041, "grad_norm": 2.108529626955218, "learning_rate": 1.9685254187854754e-05, "loss": 0.8649, "step": 1683 }, { "epoch": 1.0774152271273192, "grad_norm": 2.372899805253977, "learning_rate": 1.9684738190398853e-05, "loss": 0.8335, "step": 1684 }, { "epoch": 1.0780550223928342, "grad_norm": 2.185978672027218, "learning_rate": 1.9684221777097524e-05, "loss": 0.8569, "step": 1685 }, { "epoch": 1.0786948176583493, "grad_norm": 2.3820103394816843, "learning_rate": 1.9683704947972935e-05, "loss": 0.8434, "step": 1686 }, { "epoch": 1.0793346129238643, "grad_norm": 2.1937402868771776, "learning_rate": 1.9683187703047283e-05, "loss": 0.8402, "step": 1687 }, { "epoch": 1.0799744081893794, "grad_norm": 2.0867571328464836, "learning_rate": 1.9682670042342776e-05, "loss": 0.8498, "step": 1688 }, { "epoch": 1.0806142034548945, "grad_norm": 2.339272518655087, "learning_rate": 1.9682151965881643e-05, "loss": 0.825, "step": 1689 }, { "epoch": 1.0812539987204095, "grad_norm": 2.338483315919513, "learning_rate": 1.9681633473686125e-05, "loss": 0.8324, "step": 1690 }, { "epoch": 1.0818937939859246, "grad_norm": 2.0305690843440467, "learning_rate": 1.9681114565778488e-05, "loss": 0.836, "step": 1691 }, { "epoch": 1.0825335892514396, "grad_norm": 2.335752249549377, "learning_rate": 1.9680595242181016e-05, "loss": 0.8228, "step": 1692 }, { "epoch": 1.0831733845169547, "grad_norm": 2.068762109466192, "learning_rate": 1.9680075502916004e-05, "loss": 0.8755, "step": 1693 }, { "epoch": 1.0838131797824695, "grad_norm": 1.951598992749742, "learning_rate": 1.9679555348005767e-05, "loss": 0.8099, "step": 1694 }, { "epoch": 1.0844529750479845, "grad_norm": 2.194386774779888, "learning_rate": 1.9679034777472645e-05, "loss": 0.8493, "step": 1695 }, { "epoch": 1.0850927703134996, "grad_norm": 2.1382539909634324, "learning_rate": 1.9678513791338986e-05, "loss": 0.8675, "step": 1696 }, { "epoch": 1.0857325655790147, "grad_norm": 2.6581446475056634, "learning_rate": 1.967799238962716e-05, "loss": 0.8379, "step": 1697 }, { "epoch": 1.0863723608445297, "grad_norm": 2.453591304311365, "learning_rate": 1.967747057235956e-05, "loss": 0.8103, "step": 1698 }, { "epoch": 1.0870121561100448, "grad_norm": 2.242811456359985, "learning_rate": 1.9676948339558587e-05, "loss": 0.8559, "step": 1699 }, { "epoch": 1.0876519513755598, "grad_norm": 2.71699994430881, "learning_rate": 1.9676425691246666e-05, "loss": 0.8847, "step": 1700 }, { "epoch": 1.0882917466410749, "grad_norm": 2.3815932381803586, "learning_rate": 1.967590262744624e-05, "loss": 0.8452, "step": 1701 }, { "epoch": 1.08893154190659, "grad_norm": 2.434580582138944, "learning_rate": 1.967537914817976e-05, "loss": 0.8302, "step": 1702 }, { "epoch": 1.089571337172105, "grad_norm": 2.2988311268965496, "learning_rate": 1.967485525346972e-05, "loss": 0.7836, "step": 1703 }, { "epoch": 1.09021113243762, "grad_norm": 2.5840734329137027, "learning_rate": 1.9674330943338602e-05, "loss": 0.8702, "step": 1704 }, { "epoch": 1.090850927703135, "grad_norm": 2.108206269835554, "learning_rate": 1.9673806217808923e-05, "loss": 0.8388, "step": 1705 }, { "epoch": 1.0914907229686501, "grad_norm": 2.3452356216801635, "learning_rate": 1.9673281076903214e-05, "loss": 0.8276, "step": 1706 }, { "epoch": 1.092130518234165, "grad_norm": 2.1788121831817415, "learning_rate": 1.9672755520644023e-05, "loss": 0.8428, "step": 1707 }, { "epoch": 1.09277031349968, "grad_norm": 2.2805021529631833, "learning_rate": 1.9672229549053913e-05, "loss": 0.8375, "step": 1708 }, { "epoch": 1.093410108765195, "grad_norm": 2.3841500537218403, "learning_rate": 1.9671703162155474e-05, "loss": 0.7886, "step": 1709 }, { "epoch": 1.0940499040307101, "grad_norm": 2.700191191682592, "learning_rate": 1.9671176359971306e-05, "loss": 0.8744, "step": 1710 }, { "epoch": 1.0946896992962252, "grad_norm": 2.3254399649286257, "learning_rate": 1.967064914252403e-05, "loss": 0.8061, "step": 1711 }, { "epoch": 1.0953294945617402, "grad_norm": 2.336925996370437, "learning_rate": 1.9670121509836284e-05, "loss": 0.8197, "step": 1712 }, { "epoch": 1.0959692898272553, "grad_norm": 2.553052005033562, "learning_rate": 1.966959346193072e-05, "loss": 0.8656, "step": 1713 }, { "epoch": 1.0966090850927703, "grad_norm": 2.3108662758158194, "learning_rate": 1.9669064998830012e-05, "loss": 0.871, "step": 1714 }, { "epoch": 1.0972488803582854, "grad_norm": 2.734723928293006, "learning_rate": 1.9668536120556853e-05, "loss": 0.8308, "step": 1715 }, { "epoch": 1.0978886756238004, "grad_norm": 2.8324719294573466, "learning_rate": 1.9668006827133953e-05, "loss": 0.8259, "step": 1716 }, { "epoch": 1.0985284708893155, "grad_norm": 2.5670378355485517, "learning_rate": 1.9667477118584038e-05, "loss": 0.8467, "step": 1717 }, { "epoch": 1.0991682661548305, "grad_norm": 2.336326150572372, "learning_rate": 1.9666946994929856e-05, "loss": 0.8417, "step": 1718 }, { "epoch": 1.0998080614203456, "grad_norm": 2.242406090742273, "learning_rate": 1.966641645619416e-05, "loss": 0.8297, "step": 1719 }, { "epoch": 1.1004478566858604, "grad_norm": 2.5546608316125154, "learning_rate": 1.966588550239974e-05, "loss": 0.8246, "step": 1720 }, { "epoch": 1.1010876519513755, "grad_norm": 2.6228180527630762, "learning_rate": 1.966535413356939e-05, "loss": 0.8715, "step": 1721 }, { "epoch": 1.1017274472168905, "grad_norm": 2.6254539797651075, "learning_rate": 1.9664822349725927e-05, "loss": 0.9066, "step": 1722 }, { "epoch": 1.1023672424824056, "grad_norm": 2.5824875983564213, "learning_rate": 1.9664290150892187e-05, "loss": 0.7639, "step": 1723 }, { "epoch": 1.1030070377479206, "grad_norm": 2.52008329781677, "learning_rate": 1.9663757537091014e-05, "loss": 0.8412, "step": 1724 }, { "epoch": 1.1036468330134357, "grad_norm": 2.3230654385651546, "learning_rate": 1.966322450834529e-05, "loss": 0.8147, "step": 1725 }, { "epoch": 1.1042866282789507, "grad_norm": 2.4652948864429858, "learning_rate": 1.966269106467789e-05, "loss": 0.893, "step": 1726 }, { "epoch": 1.1049264235444658, "grad_norm": 2.304772185553636, "learning_rate": 1.9662157206111724e-05, "loss": 0.816, "step": 1727 }, { "epoch": 1.1055662188099808, "grad_norm": 2.2127995533879705, "learning_rate": 1.9661622932669718e-05, "loss": 0.8288, "step": 1728 }, { "epoch": 1.106206014075496, "grad_norm": 2.7097920877886796, "learning_rate": 1.966108824437481e-05, "loss": 0.8005, "step": 1729 }, { "epoch": 1.106845809341011, "grad_norm": 2.445405619562331, "learning_rate": 1.9660553141249955e-05, "loss": 0.8436, "step": 1730 }, { "epoch": 1.107485604606526, "grad_norm": 2.4640765667350024, "learning_rate": 1.9660017623318134e-05, "loss": 0.8335, "step": 1731 }, { "epoch": 1.108125399872041, "grad_norm": 2.251688151600021, "learning_rate": 1.9659481690602342e-05, "loss": 0.827, "step": 1732 }, { "epoch": 1.108765195137556, "grad_norm": 2.220874898177416, "learning_rate": 1.965894534312559e-05, "loss": 0.8227, "step": 1733 }, { "epoch": 1.109404990403071, "grad_norm": 2.3593903593032644, "learning_rate": 1.9658408580910903e-05, "loss": 0.8376, "step": 1734 }, { "epoch": 1.110044785668586, "grad_norm": 2.5818675000703455, "learning_rate": 1.9657871403981333e-05, "loss": 0.7579, "step": 1735 }, { "epoch": 1.110684580934101, "grad_norm": 2.3312664280126647, "learning_rate": 1.9657333812359948e-05, "loss": 0.8536, "step": 1736 }, { "epoch": 1.111324376199616, "grad_norm": 2.3387999443179273, "learning_rate": 1.965679580606983e-05, "loss": 0.8569, "step": 1737 }, { "epoch": 1.1119641714651312, "grad_norm": 2.3310254516848192, "learning_rate": 1.9656257385134073e-05, "loss": 0.8405, "step": 1738 }, { "epoch": 1.1126039667306462, "grad_norm": 2.315737081204945, "learning_rate": 1.96557185495758e-05, "loss": 0.8064, "step": 1739 }, { "epoch": 1.1132437619961613, "grad_norm": 2.569237708754139, "learning_rate": 1.9655179299418155e-05, "loss": 0.8371, "step": 1740 }, { "epoch": 1.1138835572616763, "grad_norm": 2.5670915452149496, "learning_rate": 1.965463963468428e-05, "loss": 0.8124, "step": 1741 }, { "epoch": 1.1145233525271914, "grad_norm": 2.1152088383934324, "learning_rate": 1.9654099555397358e-05, "loss": 0.8585, "step": 1742 }, { "epoch": 1.1151631477927064, "grad_norm": 2.337429617581099, "learning_rate": 1.9653559061580575e-05, "loss": 0.8382, "step": 1743 }, { "epoch": 1.1158029430582213, "grad_norm": 2.2205982157411954, "learning_rate": 1.9653018153257133e-05, "loss": 0.7933, "step": 1744 }, { "epoch": 1.1164427383237363, "grad_norm": 2.107516824104419, "learning_rate": 1.9652476830450268e-05, "loss": 0.8108, "step": 1745 }, { "epoch": 1.1170825335892514, "grad_norm": 2.269545760823515, "learning_rate": 1.9651935093183216e-05, "loss": 0.8609, "step": 1746 }, { "epoch": 1.1177223288547664, "grad_norm": 2.555512819476273, "learning_rate": 1.9651392941479236e-05, "loss": 0.7918, "step": 1747 }, { "epoch": 1.1183621241202815, "grad_norm": 2.048792446833198, "learning_rate": 1.965085037536162e-05, "loss": 0.8142, "step": 1748 }, { "epoch": 1.1190019193857965, "grad_norm": 2.49078687938013, "learning_rate": 1.965030739485365e-05, "loss": 0.8191, "step": 1749 }, { "epoch": 1.1196417146513116, "grad_norm": 2.3778216651039368, "learning_rate": 1.964976399997865e-05, "loss": 0.8349, "step": 1750 }, { "epoch": 1.1202815099168266, "grad_norm": 2.318769530322906, "learning_rate": 1.964922019075995e-05, "loss": 0.8755, "step": 1751 }, { "epoch": 1.1209213051823417, "grad_norm": 2.573150795813555, "learning_rate": 1.9648675967220897e-05, "loss": 0.8714, "step": 1752 }, { "epoch": 1.1215611004478567, "grad_norm": 2.472690151173234, "learning_rate": 1.9648131329384866e-05, "loss": 0.79, "step": 1753 }, { "epoch": 1.1222008957133718, "grad_norm": 2.621200252056306, "learning_rate": 1.9647586277275236e-05, "loss": 0.8064, "step": 1754 }, { "epoch": 1.1228406909788868, "grad_norm": 2.166192383201924, "learning_rate": 1.964704081091541e-05, "loss": 0.8483, "step": 1755 }, { "epoch": 1.1234804862444019, "grad_norm": 2.120456086921079, "learning_rate": 1.9646494930328815e-05, "loss": 0.805, "step": 1756 }, { "epoch": 1.1241202815099167, "grad_norm": 2.845883969964383, "learning_rate": 1.964594863553889e-05, "loss": 0.7611, "step": 1757 }, { "epoch": 1.1247600767754318, "grad_norm": 2.7288237657687593, "learning_rate": 1.9645401926569086e-05, "loss": 0.8798, "step": 1758 }, { "epoch": 1.1253998720409468, "grad_norm": 2.4509854796679553, "learning_rate": 1.964485480344288e-05, "loss": 0.8963, "step": 1759 }, { "epoch": 1.1260396673064619, "grad_norm": 2.371789623272321, "learning_rate": 1.964430726618377e-05, "loss": 0.7891, "step": 1760 }, { "epoch": 1.126679462571977, "grad_norm": 2.305596299727168, "learning_rate": 1.964375931481526e-05, "loss": 0.8171, "step": 1761 }, { "epoch": 1.127319257837492, "grad_norm": 2.4854517360503467, "learning_rate": 1.964321094936088e-05, "loss": 0.8109, "step": 1762 }, { "epoch": 1.127959053103007, "grad_norm": 2.0861808888704534, "learning_rate": 1.9642662169844174e-05, "loss": 0.751, "step": 1763 }, { "epoch": 1.128598848368522, "grad_norm": 2.473984138946467, "learning_rate": 1.9642112976288707e-05, "loss": 0.8416, "step": 1764 }, { "epoch": 1.1292386436340371, "grad_norm": 2.70736782439847, "learning_rate": 1.964156336871806e-05, "loss": 0.8457, "step": 1765 }, { "epoch": 1.1298784388995522, "grad_norm": 2.5248606959560873, "learning_rate": 1.9641013347155836e-05, "loss": 0.9016, "step": 1766 }, { "epoch": 1.1305182341650672, "grad_norm": 2.3805014052855475, "learning_rate": 1.964046291162565e-05, "loss": 0.8318, "step": 1767 }, { "epoch": 1.1311580294305823, "grad_norm": 2.6807997961298047, "learning_rate": 1.9639912062151135e-05, "loss": 0.8105, "step": 1768 }, { "epoch": 1.1317978246960974, "grad_norm": 2.510046982876012, "learning_rate": 1.9639360798755943e-05, "loss": 0.8221, "step": 1769 }, { "epoch": 1.1324376199616122, "grad_norm": 2.1835010426699286, "learning_rate": 1.9638809121463746e-05, "loss": 0.8634, "step": 1770 }, { "epoch": 1.1330774152271272, "grad_norm": 2.4824768671405173, "learning_rate": 1.963825703029823e-05, "loss": 0.8214, "step": 1771 }, { "epoch": 1.1337172104926423, "grad_norm": 2.803783214071671, "learning_rate": 1.9637704525283102e-05, "loss": 0.8782, "step": 1772 }, { "epoch": 1.1343570057581573, "grad_norm": 2.636190050388045, "learning_rate": 1.963715160644209e-05, "loss": 0.7555, "step": 1773 }, { "epoch": 1.1349968010236724, "grad_norm": 2.7334837619518964, "learning_rate": 1.9636598273798927e-05, "loss": 0.833, "step": 1774 }, { "epoch": 1.1356365962891874, "grad_norm": 2.165264571260731, "learning_rate": 1.9636044527377377e-05, "loss": 0.7967, "step": 1775 }, { "epoch": 1.1362763915547025, "grad_norm": 2.343903106201157, "learning_rate": 1.963549036720122e-05, "loss": 0.8187, "step": 1776 }, { "epoch": 1.1369161868202176, "grad_norm": 2.4974825811615493, "learning_rate": 1.963493579329424e-05, "loss": 0.7879, "step": 1777 }, { "epoch": 1.1375559820857326, "grad_norm": 2.348024972513641, "learning_rate": 1.9634380805680258e-05, "loss": 0.8186, "step": 1778 }, { "epoch": 1.1381957773512477, "grad_norm": 2.6757072688148127, "learning_rate": 1.9633825404383104e-05, "loss": 0.8291, "step": 1779 }, { "epoch": 1.1388355726167627, "grad_norm": 2.8081374079825228, "learning_rate": 1.9633269589426625e-05, "loss": 0.8383, "step": 1780 }, { "epoch": 1.1394753678822778, "grad_norm": 2.0999236379812634, "learning_rate": 1.9632713360834685e-05, "loss": 0.8011, "step": 1781 }, { "epoch": 1.1401151631477928, "grad_norm": 2.2837173382354146, "learning_rate": 1.963215671863117e-05, "loss": 0.8626, "step": 1782 }, { "epoch": 1.1407549584133077, "grad_norm": 2.579031804601103, "learning_rate": 1.963159966283998e-05, "loss": 0.8216, "step": 1783 }, { "epoch": 1.1413947536788227, "grad_norm": 2.5408476263569373, "learning_rate": 1.963104219348503e-05, "loss": 0.8418, "step": 1784 }, { "epoch": 1.1420345489443378, "grad_norm": 2.939315320554236, "learning_rate": 1.9630484310590265e-05, "loss": 0.8724, "step": 1785 }, { "epoch": 1.1426743442098528, "grad_norm": 2.766604484697838, "learning_rate": 1.9629926014179634e-05, "loss": 0.8121, "step": 1786 }, { "epoch": 1.1433141394753679, "grad_norm": 2.3067723810486327, "learning_rate": 1.962936730427711e-05, "loss": 0.7878, "step": 1787 }, { "epoch": 1.143953934740883, "grad_norm": 2.348295943509864, "learning_rate": 1.962880818090668e-05, "loss": 0.8249, "step": 1788 }, { "epoch": 1.144593730006398, "grad_norm": 2.4652001661280214, "learning_rate": 1.9628248644092358e-05, "loss": 0.8684, "step": 1789 }, { "epoch": 1.145233525271913, "grad_norm": 2.219545735628349, "learning_rate": 1.9627688693858168e-05, "loss": 0.7952, "step": 1790 }, { "epoch": 1.145873320537428, "grad_norm": 2.4379031788846897, "learning_rate": 1.962712833022815e-05, "loss": 0.8434, "step": 1791 }, { "epoch": 1.1465131158029431, "grad_norm": 2.5537335593464348, "learning_rate": 1.9626567553226365e-05, "loss": 0.8261, "step": 1792 }, { "epoch": 1.1471529110684582, "grad_norm": 2.422129925615811, "learning_rate": 1.9626006362876898e-05, "loss": 0.8644, "step": 1793 }, { "epoch": 1.147792706333973, "grad_norm": 2.479565061980055, "learning_rate": 1.9625444759203837e-05, "loss": 0.8557, "step": 1794 }, { "epoch": 1.1484325015994883, "grad_norm": 2.9102460186728636, "learning_rate": 1.9624882742231302e-05, "loss": 0.8394, "step": 1795 }, { "epoch": 1.1490722968650031, "grad_norm": 2.259256397542534, "learning_rate": 1.9624320311983423e-05, "loss": 0.8054, "step": 1796 }, { "epoch": 1.1497120921305182, "grad_norm": 2.370993251406777, "learning_rate": 1.962375746848435e-05, "loss": 0.808, "step": 1797 }, { "epoch": 1.1503518873960332, "grad_norm": 2.380169413178863, "learning_rate": 1.9623194211758253e-05, "loss": 0.861, "step": 1798 }, { "epoch": 1.1509916826615483, "grad_norm": 2.4420428986812914, "learning_rate": 1.9622630541829314e-05, "loss": 0.8857, "step": 1799 }, { "epoch": 1.1516314779270633, "grad_norm": 2.5230324761648846, "learning_rate": 1.9622066458721735e-05, "loss": 0.8507, "step": 1800 }, { "epoch": 1.1522712731925784, "grad_norm": 2.2424324399028372, "learning_rate": 1.962150196245974e-05, "loss": 0.7917, "step": 1801 }, { "epoch": 1.1529110684580934, "grad_norm": 2.5421249344670356, "learning_rate": 1.9620937053067565e-05, "loss": 0.857, "step": 1802 }, { "epoch": 1.1535508637236085, "grad_norm": 2.3901762139217775, "learning_rate": 1.9620371730569468e-05, "loss": 0.825, "step": 1803 }, { "epoch": 1.1541906589891235, "grad_norm": 2.3408083157180686, "learning_rate": 1.9619805994989723e-05, "loss": 0.8207, "step": 1804 }, { "epoch": 1.1548304542546386, "grad_norm": 2.6651899865358915, "learning_rate": 1.961923984635262e-05, "loss": 0.8469, "step": 1805 }, { "epoch": 1.1554702495201536, "grad_norm": 2.691200961104593, "learning_rate": 1.9618673284682466e-05, "loss": 0.8838, "step": 1806 }, { "epoch": 1.1561100447856685, "grad_norm": 2.369289260175582, "learning_rate": 1.9618106310003597e-05, "loss": 0.8633, "step": 1807 }, { "epoch": 1.1567498400511835, "grad_norm": 2.436570583635262, "learning_rate": 1.961753892234035e-05, "loss": 0.8571, "step": 1808 }, { "epoch": 1.1573896353166986, "grad_norm": 2.2366607932376272, "learning_rate": 1.961697112171709e-05, "loss": 0.7825, "step": 1809 }, { "epoch": 1.1580294305822136, "grad_norm": 2.593051162785129, "learning_rate": 1.9616402908158195e-05, "loss": 0.8099, "step": 1810 }, { "epoch": 1.1586692258477287, "grad_norm": 2.3176734959026972, "learning_rate": 1.961583428168807e-05, "loss": 0.8233, "step": 1811 }, { "epoch": 1.1593090211132437, "grad_norm": 2.163109573428445, "learning_rate": 1.9615265242331125e-05, "loss": 0.8573, "step": 1812 }, { "epoch": 1.1599488163787588, "grad_norm": 2.143840941446012, "learning_rate": 1.9614695790111795e-05, "loss": 0.8418, "step": 1813 }, { "epoch": 1.1605886116442738, "grad_norm": 2.1552641761343683, "learning_rate": 1.9614125925054525e-05, "loss": 0.8594, "step": 1814 }, { "epoch": 1.161228406909789, "grad_norm": 2.166159045077398, "learning_rate": 1.96135556471838e-05, "loss": 0.8329, "step": 1815 }, { "epoch": 1.161868202175304, "grad_norm": 2.051627276126034, "learning_rate": 1.9612984956524088e-05, "loss": 0.7806, "step": 1816 }, { "epoch": 1.162507997440819, "grad_norm": 2.2184726798328183, "learning_rate": 1.9612413853099906e-05, "loss": 0.7982, "step": 1817 }, { "epoch": 1.163147792706334, "grad_norm": 2.305541214754076, "learning_rate": 1.961184233693577e-05, "loss": 0.8995, "step": 1818 }, { "epoch": 1.1637875879718491, "grad_norm": 2.442793204250588, "learning_rate": 1.9611270408056226e-05, "loss": 0.8666, "step": 1819 }, { "epoch": 1.164427383237364, "grad_norm": 2.3319509782812387, "learning_rate": 1.9610698066485825e-05, "loss": 0.7762, "step": 1820 }, { "epoch": 1.165067178502879, "grad_norm": 2.2124626948609927, "learning_rate": 1.9610125312249147e-05, "loss": 0.8395, "step": 1821 }, { "epoch": 1.165706973768394, "grad_norm": 2.4811125416066315, "learning_rate": 1.960955214537078e-05, "loss": 0.7953, "step": 1822 }, { "epoch": 1.166346769033909, "grad_norm": 2.487056511323211, "learning_rate": 1.960897856587534e-05, "loss": 0.8068, "step": 1823 }, { "epoch": 1.1669865642994242, "grad_norm": 2.0927388483140614, "learning_rate": 1.9608404573787455e-05, "loss": 0.8019, "step": 1824 }, { "epoch": 1.1676263595649392, "grad_norm": 2.2308721464788124, "learning_rate": 1.9607830169131765e-05, "loss": 0.7479, "step": 1825 }, { "epoch": 1.1682661548304543, "grad_norm": 2.8748268658789393, "learning_rate": 1.9607255351932945e-05, "loss": 0.8158, "step": 1826 }, { "epoch": 1.1689059500959693, "grad_norm": 2.47993514890096, "learning_rate": 1.9606680122215665e-05, "loss": 0.8115, "step": 1827 }, { "epoch": 1.1695457453614844, "grad_norm": 2.962671013022815, "learning_rate": 1.960610448000463e-05, "loss": 0.7876, "step": 1828 }, { "epoch": 1.1701855406269994, "grad_norm": 2.4170410073794955, "learning_rate": 1.9605528425324557e-05, "loss": 0.8329, "step": 1829 }, { "epoch": 1.1708253358925145, "grad_norm": 2.3176318064629378, "learning_rate": 1.9604951958200186e-05, "loss": 0.8811, "step": 1830 }, { "epoch": 1.1714651311580295, "grad_norm": 2.928048880652178, "learning_rate": 1.9604375078656255e-05, "loss": 0.8596, "step": 1831 }, { "epoch": 1.1721049264235446, "grad_norm": 2.799406950266459, "learning_rate": 1.960379778671755e-05, "loss": 0.8611, "step": 1832 }, { "epoch": 1.1727447216890594, "grad_norm": 2.508635853647572, "learning_rate": 1.9603220082408845e-05, "loss": 0.8351, "step": 1833 }, { "epoch": 1.1733845169545745, "grad_norm": 2.2122574139884, "learning_rate": 1.960264196575496e-05, "loss": 0.8537, "step": 1834 }, { "epoch": 1.1740243122200895, "grad_norm": 2.3022288360543044, "learning_rate": 1.9602063436780703e-05, "loss": 0.8213, "step": 1835 }, { "epoch": 1.1746641074856046, "grad_norm": 2.426688670968072, "learning_rate": 1.9601484495510928e-05, "loss": 0.8162, "step": 1836 }, { "epoch": 1.1753039027511196, "grad_norm": 2.476032450641568, "learning_rate": 1.960090514197049e-05, "loss": 0.8209, "step": 1837 }, { "epoch": 1.1759436980166347, "grad_norm": 2.2376775791216996, "learning_rate": 1.960032537618426e-05, "loss": 0.7798, "step": 1838 }, { "epoch": 1.1765834932821497, "grad_norm": 2.3255833826076744, "learning_rate": 1.9599745198177135e-05, "loss": 0.8136, "step": 1839 }, { "epoch": 1.1772232885476648, "grad_norm": 2.798036180274826, "learning_rate": 1.9599164607974034e-05, "loss": 0.8183, "step": 1840 }, { "epoch": 1.1778630838131798, "grad_norm": 2.464273198291979, "learning_rate": 1.9598583605599876e-05, "loss": 0.8408, "step": 1841 }, { "epoch": 1.1785028790786949, "grad_norm": 2.320228893507996, "learning_rate": 1.959800219107961e-05, "loss": 0.866, "step": 1842 }, { "epoch": 1.17914267434421, "grad_norm": 2.0980509097809423, "learning_rate": 1.9597420364438207e-05, "loss": 0.8151, "step": 1843 }, { "epoch": 1.1797824696097248, "grad_norm": 2.5357728564130544, "learning_rate": 1.9596838125700644e-05, "loss": 0.7876, "step": 1844 }, { "epoch": 1.18042226487524, "grad_norm": 2.466891080020338, "learning_rate": 1.9596255474891926e-05, "loss": 0.8211, "step": 1845 }, { "epoch": 1.1810620601407549, "grad_norm": 2.275524895015393, "learning_rate": 1.9595672412037066e-05, "loss": 0.8779, "step": 1846 }, { "epoch": 1.18170185540627, "grad_norm": 2.586303103293257, "learning_rate": 1.95950889371611e-05, "loss": 0.8067, "step": 1847 }, { "epoch": 1.182341650671785, "grad_norm": 2.67062050303089, "learning_rate": 1.959450505028909e-05, "loss": 0.8748, "step": 1848 }, { "epoch": 1.1829814459373, "grad_norm": 2.142523193326157, "learning_rate": 1.9593920751446096e-05, "loss": 0.781, "step": 1849 }, { "epoch": 1.183621241202815, "grad_norm": 2.731614745132828, "learning_rate": 1.959333604065721e-05, "loss": 0.8331, "step": 1850 }, { "epoch": 1.1842610364683301, "grad_norm": 2.4296668613332053, "learning_rate": 1.9592750917947545e-05, "loss": 0.7741, "step": 1851 }, { "epoch": 1.1849008317338452, "grad_norm": 2.9613996332564048, "learning_rate": 1.9592165383342215e-05, "loss": 0.8467, "step": 1852 }, { "epoch": 1.1855406269993602, "grad_norm": 2.4477288186482737, "learning_rate": 1.9591579436866367e-05, "loss": 0.8799, "step": 1853 }, { "epoch": 1.1861804222648753, "grad_norm": 2.3078209307266704, "learning_rate": 1.9590993078545162e-05, "loss": 0.8362, "step": 1854 }, { "epoch": 1.1868202175303904, "grad_norm": 2.137294569117143, "learning_rate": 1.9590406308403772e-05, "loss": 0.7761, "step": 1855 }, { "epoch": 1.1874600127959054, "grad_norm": 2.3955575779017733, "learning_rate": 1.95898191264674e-05, "loss": 0.8656, "step": 1856 }, { "epoch": 1.1880998080614202, "grad_norm": 2.193076831877625, "learning_rate": 1.958923153276125e-05, "loss": 0.8292, "step": 1857 }, { "epoch": 1.1887396033269353, "grad_norm": 2.2389526057719893, "learning_rate": 1.9588643527310558e-05, "loss": 0.7939, "step": 1858 }, { "epoch": 1.1893793985924503, "grad_norm": 2.4622589931261767, "learning_rate": 1.9588055110140568e-05, "loss": 0.8155, "step": 1859 }, { "epoch": 1.1900191938579654, "grad_norm": 2.2542886138353637, "learning_rate": 1.9587466281276547e-05, "loss": 0.8666, "step": 1860 }, { "epoch": 1.1906589891234804, "grad_norm": 2.4712077430647903, "learning_rate": 1.958687704074378e-05, "loss": 0.842, "step": 1861 }, { "epoch": 1.1912987843889955, "grad_norm": 2.566889180402769, "learning_rate": 1.958628738856757e-05, "loss": 0.7861, "step": 1862 }, { "epoch": 1.1919385796545106, "grad_norm": 2.6328751946121804, "learning_rate": 1.9585697324773228e-05, "loss": 0.8644, "step": 1863 }, { "epoch": 1.1925783749200256, "grad_norm": 2.2557413061166747, "learning_rate": 1.9585106849386096e-05, "loss": 0.8498, "step": 1864 }, { "epoch": 1.1932181701855407, "grad_norm": 2.366164224817601, "learning_rate": 1.9584515962431523e-05, "loss": 0.8552, "step": 1865 }, { "epoch": 1.1938579654510557, "grad_norm": 2.2849990833231972, "learning_rate": 1.9583924663934886e-05, "loss": 0.8152, "step": 1866 }, { "epoch": 1.1944977607165708, "grad_norm": 2.5505160694690403, "learning_rate": 1.9583332953921573e-05, "loss": 0.8191, "step": 1867 }, { "epoch": 1.1951375559820858, "grad_norm": 2.632958319653156, "learning_rate": 1.9582740832416995e-05, "loss": 0.8692, "step": 1868 }, { "epoch": 1.1957773512476009, "grad_norm": 2.341420668556438, "learning_rate": 1.9582148299446565e-05, "loss": 0.7849, "step": 1869 }, { "epoch": 1.1964171465131157, "grad_norm": 2.6504073825676815, "learning_rate": 1.9581555355035733e-05, "loss": 0.8313, "step": 1870 }, { "epoch": 1.1970569417786308, "grad_norm": 2.202308934765863, "learning_rate": 1.9580961999209963e-05, "loss": 0.8091, "step": 1871 }, { "epoch": 1.1976967370441458, "grad_norm": 2.23253614301773, "learning_rate": 1.9580368231994724e-05, "loss": 0.7656, "step": 1872 }, { "epoch": 1.1983365323096609, "grad_norm": 2.242323580664685, "learning_rate": 1.9579774053415518e-05, "loss": 0.8101, "step": 1873 }, { "epoch": 1.198976327575176, "grad_norm": 2.3109738334337444, "learning_rate": 1.957917946349785e-05, "loss": 0.8049, "step": 1874 }, { "epoch": 1.199616122840691, "grad_norm": 2.21366803642531, "learning_rate": 1.9578584462267264e-05, "loss": 0.842, "step": 1875 }, { "epoch": 1.200255918106206, "grad_norm": 2.576925248287956, "learning_rate": 1.9577989049749292e-05, "loss": 0.8266, "step": 1876 }, { "epoch": 1.200895713371721, "grad_norm": 2.4617167787115832, "learning_rate": 1.9577393225969512e-05, "loss": 0.8412, "step": 1877 }, { "epoch": 1.2015355086372361, "grad_norm": 2.633057784023581, "learning_rate": 1.9576796990953502e-05, "loss": 0.785, "step": 1878 }, { "epoch": 1.2021753039027512, "grad_norm": 2.2561384250444014, "learning_rate": 1.957620034472687e-05, "loss": 0.8373, "step": 1879 }, { "epoch": 1.2028150991682662, "grad_norm": 2.2056876182333776, "learning_rate": 1.9575603287315227e-05, "loss": 0.8281, "step": 1880 }, { "epoch": 1.2034548944337813, "grad_norm": 3.107122984031301, "learning_rate": 1.957500581874421e-05, "loss": 0.8193, "step": 1881 }, { "epoch": 1.2040946896992963, "grad_norm": 2.1925945318914226, "learning_rate": 1.9574407939039478e-05, "loss": 0.8477, "step": 1882 }, { "epoch": 1.2047344849648112, "grad_norm": 2.5103823736013764, "learning_rate": 1.95738096482267e-05, "loss": 0.8302, "step": 1883 }, { "epoch": 1.2053742802303262, "grad_norm": 2.269354372580821, "learning_rate": 1.957321094633157e-05, "loss": 0.8554, "step": 1884 }, { "epoch": 1.2060140754958413, "grad_norm": 2.1763755515216583, "learning_rate": 1.9572611833379785e-05, "loss": 0.7712, "step": 1885 }, { "epoch": 1.2066538707613563, "grad_norm": 2.2187401126624504, "learning_rate": 1.9572012309397078e-05, "loss": 0.8153, "step": 1886 }, { "epoch": 1.2072936660268714, "grad_norm": 2.2387338990901626, "learning_rate": 1.9571412374409193e-05, "loss": 0.8286, "step": 1887 }, { "epoch": 1.2079334612923864, "grad_norm": 2.585385578063526, "learning_rate": 1.9570812028441884e-05, "loss": 0.8009, "step": 1888 }, { "epoch": 1.2085732565579015, "grad_norm": 2.1906547866429333, "learning_rate": 1.9570211271520932e-05, "loss": 0.837, "step": 1889 }, { "epoch": 1.2092130518234165, "grad_norm": 2.2625403713294925, "learning_rate": 1.9569610103672134e-05, "loss": 0.9221, "step": 1890 }, { "epoch": 1.2098528470889316, "grad_norm": 2.0791180770882365, "learning_rate": 1.95690085249213e-05, "loss": 0.8499, "step": 1891 }, { "epoch": 1.2104926423544466, "grad_norm": 3.0555223189652767, "learning_rate": 1.956840653529426e-05, "loss": 0.8298, "step": 1892 }, { "epoch": 1.2111324376199617, "grad_norm": 2.1532499193067576, "learning_rate": 1.9567804134816864e-05, "loss": 0.7876, "step": 1893 }, { "epoch": 1.2117722328854765, "grad_norm": 2.247747286465163, "learning_rate": 1.9567201323514985e-05, "loss": 0.8446, "step": 1894 }, { "epoch": 1.2124120281509918, "grad_norm": 2.3017431574837106, "learning_rate": 1.9566598101414494e-05, "loss": 0.839, "step": 1895 }, { "epoch": 1.2130518234165066, "grad_norm": 2.294331111128671, "learning_rate": 1.95659944685413e-05, "loss": 0.8438, "step": 1896 }, { "epoch": 1.2136916186820217, "grad_norm": 2.447763191778534, "learning_rate": 1.956539042492132e-05, "loss": 0.7978, "step": 1897 }, { "epoch": 1.2143314139475367, "grad_norm": 2.0864204117174596, "learning_rate": 1.9564785970580487e-05, "loss": 0.7986, "step": 1898 }, { "epoch": 1.2149712092130518, "grad_norm": 2.2104735525674104, "learning_rate": 1.956418110554476e-05, "loss": 0.8064, "step": 1899 }, { "epoch": 1.2156110044785668, "grad_norm": 2.133505015019704, "learning_rate": 1.9563575829840115e-05, "loss": 0.7882, "step": 1900 }, { "epoch": 1.216250799744082, "grad_norm": 2.24714298428157, "learning_rate": 1.956297014349253e-05, "loss": 0.8189, "step": 1901 }, { "epoch": 1.216890595009597, "grad_norm": 2.32422201640948, "learning_rate": 1.956236404652802e-05, "loss": 0.8019, "step": 1902 }, { "epoch": 1.217530390275112, "grad_norm": 2.217711844973494, "learning_rate": 1.9561757538972613e-05, "loss": 0.7817, "step": 1903 }, { "epoch": 1.218170185540627, "grad_norm": 2.1013533449853243, "learning_rate": 1.956115062085234e-05, "loss": 0.7773, "step": 1904 }, { "epoch": 1.2188099808061421, "grad_norm": 2.2987153204592157, "learning_rate": 1.956054329219327e-05, "loss": 0.8309, "step": 1905 }, { "epoch": 1.2194497760716572, "grad_norm": 2.1634207097237157, "learning_rate": 1.955993555302148e-05, "loss": 0.8091, "step": 1906 }, { "epoch": 1.220089571337172, "grad_norm": 2.4607916152559484, "learning_rate": 1.955932740336306e-05, "loss": 0.8306, "step": 1907 }, { "epoch": 1.220729366602687, "grad_norm": 2.769224162057785, "learning_rate": 1.9558718843244127e-05, "loss": 0.8071, "step": 1908 }, { "epoch": 1.221369161868202, "grad_norm": 2.3363580180969588, "learning_rate": 1.9558109872690808e-05, "loss": 0.8262, "step": 1909 }, { "epoch": 1.2220089571337172, "grad_norm": 2.2922447648140807, "learning_rate": 1.9557500491729257e-05, "loss": 0.8579, "step": 1910 }, { "epoch": 1.2226487523992322, "grad_norm": 2.3996285032817757, "learning_rate": 1.9556890700385637e-05, "loss": 0.7995, "step": 1911 }, { "epoch": 1.2232885476647473, "grad_norm": 2.6172868810586003, "learning_rate": 1.9556280498686128e-05, "loss": 0.8507, "step": 1912 }, { "epoch": 1.2239283429302623, "grad_norm": 2.284905970951021, "learning_rate": 1.955566988665693e-05, "loss": 0.8487, "step": 1913 }, { "epoch": 1.2245681381957774, "grad_norm": 2.179250922970401, "learning_rate": 1.955505886432427e-05, "loss": 0.8575, "step": 1914 }, { "epoch": 1.2252079334612924, "grad_norm": 2.0143068219158273, "learning_rate": 1.9554447431714383e-05, "loss": 0.7888, "step": 1915 }, { "epoch": 1.2258477287268075, "grad_norm": 2.2621732995465735, "learning_rate": 1.9553835588853515e-05, "loss": 0.8352, "step": 1916 }, { "epoch": 1.2264875239923225, "grad_norm": 2.232531646859888, "learning_rate": 1.955322333576794e-05, "loss": 0.8114, "step": 1917 }, { "epoch": 1.2271273192578376, "grad_norm": 2.4723369738169922, "learning_rate": 1.9552610672483954e-05, "loss": 0.885, "step": 1918 }, { "epoch": 1.2277671145233526, "grad_norm": 2.1812787502150646, "learning_rate": 1.9551997599027857e-05, "loss": 0.7438, "step": 1919 }, { "epoch": 1.2284069097888675, "grad_norm": 2.3699359930039603, "learning_rate": 1.9551384115425973e-05, "loss": 0.8615, "step": 1920 }, { "epoch": 1.2290467050543825, "grad_norm": 2.4398272380404107, "learning_rate": 1.9550770221704644e-05, "loss": 0.8142, "step": 1921 }, { "epoch": 1.2296865003198976, "grad_norm": 2.229129336669916, "learning_rate": 1.9550155917890233e-05, "loss": 0.8211, "step": 1922 }, { "epoch": 1.2303262955854126, "grad_norm": 2.4103382548961143, "learning_rate": 1.9549541204009117e-05, "loss": 0.8681, "step": 1923 }, { "epoch": 1.2309660908509277, "grad_norm": 2.488147078154664, "learning_rate": 1.954892608008769e-05, "loss": 0.7898, "step": 1924 }, { "epoch": 1.2316058861164427, "grad_norm": 2.187133924353774, "learning_rate": 1.954831054615236e-05, "loss": 0.8072, "step": 1925 }, { "epoch": 1.2322456813819578, "grad_norm": 2.172481517110744, "learning_rate": 1.9547694602229563e-05, "loss": 0.7996, "step": 1926 }, { "epoch": 1.2328854766474728, "grad_norm": 2.5098180002746444, "learning_rate": 1.9547078248345738e-05, "loss": 0.8007, "step": 1927 }, { "epoch": 1.2335252719129879, "grad_norm": 2.231516247754341, "learning_rate": 1.9546461484527362e-05, "loss": 0.7836, "step": 1928 }, { "epoch": 1.234165067178503, "grad_norm": 2.2683150279569175, "learning_rate": 1.954584431080091e-05, "loss": 0.8231, "step": 1929 }, { "epoch": 1.234804862444018, "grad_norm": 2.335060293975893, "learning_rate": 1.9545226727192884e-05, "loss": 0.8347, "step": 1930 }, { "epoch": 1.235444657709533, "grad_norm": 2.0705778667947454, "learning_rate": 1.95446087337298e-05, "loss": 0.7705, "step": 1931 }, { "epoch": 1.236084452975048, "grad_norm": 2.3943831439083727, "learning_rate": 1.95439903304382e-05, "loss": 0.801, "step": 1932 }, { "epoch": 1.236724248240563, "grad_norm": 2.0738438092122022, "learning_rate": 1.9543371517344628e-05, "loss": 0.7697, "step": 1933 }, { "epoch": 1.237364043506078, "grad_norm": 2.4608734046786256, "learning_rate": 1.954275229447566e-05, "loss": 0.8408, "step": 1934 }, { "epoch": 1.238003838771593, "grad_norm": 2.29704456037634, "learning_rate": 1.9542132661857885e-05, "loss": 0.8426, "step": 1935 }, { "epoch": 1.238643634037108, "grad_norm": 2.1107798474794204, "learning_rate": 1.954151261951791e-05, "loss": 0.7576, "step": 1936 }, { "epoch": 1.2392834293026231, "grad_norm": 2.165014471986409, "learning_rate": 1.9540892167482354e-05, "loss": 0.7837, "step": 1937 }, { "epoch": 1.2399232245681382, "grad_norm": 2.8060875025384324, "learning_rate": 1.954027130577786e-05, "loss": 0.8558, "step": 1938 }, { "epoch": 1.2405630198336532, "grad_norm": 2.490229015140693, "learning_rate": 1.9539650034431085e-05, "loss": 0.832, "step": 1939 }, { "epoch": 1.2412028150991683, "grad_norm": 2.1230371377707806, "learning_rate": 1.9539028353468713e-05, "loss": 0.7921, "step": 1940 }, { "epoch": 1.2418426103646834, "grad_norm": 2.0790527746133978, "learning_rate": 1.9538406262917427e-05, "loss": 0.7965, "step": 1941 }, { "epoch": 1.2424824056301984, "grad_norm": 2.1873711423992654, "learning_rate": 1.9537783762803945e-05, "loss": 0.8432, "step": 1942 }, { "epoch": 1.2431222008957135, "grad_norm": 2.1311150095523805, "learning_rate": 1.9537160853154994e-05, "loss": 0.8175, "step": 1943 }, { "epoch": 1.2437619961612283, "grad_norm": 2.632932928504426, "learning_rate": 1.9536537533997322e-05, "loss": 0.8102, "step": 1944 }, { "epoch": 1.2444017914267436, "grad_norm": 2.3327348418278238, "learning_rate": 1.9535913805357692e-05, "loss": 0.7929, "step": 1945 }, { "epoch": 1.2450415866922584, "grad_norm": 2.173961676453856, "learning_rate": 1.9535289667262885e-05, "loss": 0.7599, "step": 1946 }, { "epoch": 1.2456813819577734, "grad_norm": 2.456119511870974, "learning_rate": 1.9534665119739707e-05, "loss": 0.7886, "step": 1947 }, { "epoch": 1.2463211772232885, "grad_norm": 2.612446608241169, "learning_rate": 1.953404016281496e-05, "loss": 0.8347, "step": 1948 }, { "epoch": 1.2469609724888036, "grad_norm": 2.8071779241739603, "learning_rate": 1.95334147965155e-05, "loss": 0.8095, "step": 1949 }, { "epoch": 1.2476007677543186, "grad_norm": 2.0475782015212594, "learning_rate": 1.9532789020868158e-05, "loss": 0.7835, "step": 1950 }, { "epoch": 1.2482405630198337, "grad_norm": 2.9243186373043484, "learning_rate": 1.9532162835899816e-05, "loss": 0.891, "step": 1951 }, { "epoch": 1.2488803582853487, "grad_norm": 2.4653903351782236, "learning_rate": 1.9531536241637357e-05, "loss": 0.7869, "step": 1952 }, { "epoch": 1.2495201535508638, "grad_norm": 2.506826544244218, "learning_rate": 1.953090923810769e-05, "loss": 0.8434, "step": 1953 }, { "epoch": 1.2501599488163788, "grad_norm": 2.1841637161649565, "learning_rate": 1.9530281825337733e-05, "loss": 0.764, "step": 1954 }, { "epoch": 1.2507997440818939, "grad_norm": 2.639058722649408, "learning_rate": 1.952965400335443e-05, "loss": 0.8584, "step": 1955 }, { "epoch": 1.251439539347409, "grad_norm": 2.2951100834442824, "learning_rate": 1.9529025772184732e-05, "loss": 0.8325, "step": 1956 }, { "epoch": 1.2520793346129238, "grad_norm": 2.3365626128066257, "learning_rate": 1.952839713185562e-05, "loss": 0.7312, "step": 1957 }, { "epoch": 1.252719129878439, "grad_norm": 2.5082972789385756, "learning_rate": 1.9527768082394085e-05, "loss": 0.7806, "step": 1958 }, { "epoch": 1.2533589251439539, "grad_norm": 2.480147600962765, "learning_rate": 1.9527138623827137e-05, "loss": 0.839, "step": 1959 }, { "epoch": 1.253998720409469, "grad_norm": 2.3932349593019366, "learning_rate": 1.9526508756181805e-05, "loss": 0.8604, "step": 1960 }, { "epoch": 1.254638515674984, "grad_norm": 2.474445784360191, "learning_rate": 1.9525878479485133e-05, "loss": 0.8247, "step": 1961 }, { "epoch": 1.255278310940499, "grad_norm": 2.49127342328184, "learning_rate": 1.9525247793764183e-05, "loss": 0.8495, "step": 1962 }, { "epoch": 1.255918106206014, "grad_norm": 2.259011123258911, "learning_rate": 1.9524616699046035e-05, "loss": 0.8282, "step": 1963 }, { "epoch": 1.2565579014715291, "grad_norm": 2.4810008676913022, "learning_rate": 1.9523985195357792e-05, "loss": 0.8054, "step": 1964 }, { "epoch": 1.2571976967370442, "grad_norm": 2.2460526035984576, "learning_rate": 1.9523353282726565e-05, "loss": 0.8378, "step": 1965 }, { "epoch": 1.2578374920025592, "grad_norm": 2.3904700332140614, "learning_rate": 1.952272096117949e-05, "loss": 0.8234, "step": 1966 }, { "epoch": 1.2584772872680743, "grad_norm": 2.3569230583804335, "learning_rate": 1.9522088230743716e-05, "loss": 0.8411, "step": 1967 }, { "epoch": 1.2591170825335891, "grad_norm": 2.3275620962094816, "learning_rate": 1.9521455091446412e-05, "loss": 0.8479, "step": 1968 }, { "epoch": 1.2597568777991044, "grad_norm": 2.177882887046549, "learning_rate": 1.9520821543314764e-05, "loss": 0.8161, "step": 1969 }, { "epoch": 1.2603966730646192, "grad_norm": 2.5178349664354753, "learning_rate": 1.9520187586375974e-05, "loss": 0.7945, "step": 1970 }, { "epoch": 1.2610364683301343, "grad_norm": 2.227425050788682, "learning_rate": 1.9519553220657267e-05, "loss": 0.8156, "step": 1971 }, { "epoch": 1.2616762635956493, "grad_norm": 2.3636590286222137, "learning_rate": 1.9518918446185874e-05, "loss": 0.7649, "step": 1972 }, { "epoch": 1.2623160588611644, "grad_norm": 2.1455565014462987, "learning_rate": 1.951828326298906e-05, "loss": 0.8564, "step": 1973 }, { "epoch": 1.2629558541266794, "grad_norm": 2.4230616886455167, "learning_rate": 1.951764767109409e-05, "loss": 0.8121, "step": 1974 }, { "epoch": 1.2635956493921945, "grad_norm": 2.9344372037228066, "learning_rate": 1.9517011670528262e-05, "loss": 0.8099, "step": 1975 }, { "epoch": 1.2642354446577095, "grad_norm": 2.804886625137911, "learning_rate": 1.9516375261318883e-05, "loss": 0.7982, "step": 1976 }, { "epoch": 1.2648752399232246, "grad_norm": 2.5753910033448157, "learning_rate": 1.951573844349328e-05, "loss": 0.8851, "step": 1977 }, { "epoch": 1.2655150351887396, "grad_norm": 2.1181425201172552, "learning_rate": 1.9515101217078795e-05, "loss": 0.8268, "step": 1978 }, { "epoch": 1.2661548304542547, "grad_norm": 2.5290171686660123, "learning_rate": 1.9514463582102786e-05, "loss": 0.7872, "step": 1979 }, { "epoch": 1.2667946257197698, "grad_norm": 2.5720123909047614, "learning_rate": 1.9513825538592636e-05, "loss": 0.7635, "step": 1980 }, { "epoch": 1.2674344209852846, "grad_norm": 2.3360572434402425, "learning_rate": 1.9513187086575745e-05, "loss": 0.7761, "step": 1981 }, { "epoch": 1.2680742162507999, "grad_norm": 2.8627170814332703, "learning_rate": 1.9512548226079522e-05, "loss": 0.8058, "step": 1982 }, { "epoch": 1.2687140115163147, "grad_norm": 2.139872437833634, "learning_rate": 1.9511908957131398e-05, "loss": 0.8061, "step": 1983 }, { "epoch": 1.2693538067818297, "grad_norm": 2.3006900922543645, "learning_rate": 1.9511269279758824e-05, "loss": 0.8767, "step": 1984 }, { "epoch": 1.2699936020473448, "grad_norm": 2.3901088763329152, "learning_rate": 1.9510629193989266e-05, "loss": 0.8074, "step": 1985 }, { "epoch": 1.2706333973128598, "grad_norm": 2.565010602151121, "learning_rate": 1.9509988699850214e-05, "loss": 0.7919, "step": 1986 }, { "epoch": 1.271273192578375, "grad_norm": 2.6184571995022674, "learning_rate": 1.9509347797369157e-05, "loss": 0.8452, "step": 1987 }, { "epoch": 1.27191298784389, "grad_norm": 2.3093806468061966, "learning_rate": 1.9508706486573626e-05, "loss": 0.8662, "step": 1988 }, { "epoch": 1.272552783109405, "grad_norm": 2.4470651520856914, "learning_rate": 1.950806476749115e-05, "loss": 0.8132, "step": 1989 }, { "epoch": 1.27319257837492, "grad_norm": 2.5344609627342547, "learning_rate": 1.950742264014929e-05, "loss": 0.8008, "step": 1990 }, { "epoch": 1.2738323736404351, "grad_norm": 2.574555488797572, "learning_rate": 1.9506780104575612e-05, "loss": 0.8189, "step": 1991 }, { "epoch": 1.2744721689059502, "grad_norm": 2.4405806685638045, "learning_rate": 1.9506137160797706e-05, "loss": 0.8039, "step": 1992 }, { "epoch": 1.2751119641714652, "grad_norm": 2.106687494438838, "learning_rate": 1.9505493808843182e-05, "loss": 0.7976, "step": 1993 }, { "epoch": 1.27575175943698, "grad_norm": 2.867827970819382, "learning_rate": 1.950485004873966e-05, "loss": 0.797, "step": 1994 }, { "epoch": 1.2763915547024953, "grad_norm": 2.7868289830014175, "learning_rate": 1.950420588051479e-05, "loss": 0.8202, "step": 1995 }, { "epoch": 1.2770313499680102, "grad_norm": 2.2188240828138177, "learning_rate": 1.9503561304196227e-05, "loss": 0.7473, "step": 1996 }, { "epoch": 1.2776711452335252, "grad_norm": 2.4693858884496267, "learning_rate": 1.950291631981164e-05, "loss": 0.7922, "step": 1997 }, { "epoch": 1.2783109404990403, "grad_norm": 2.6543526690699735, "learning_rate": 1.9502270927388736e-05, "loss": 0.8283, "step": 1998 }, { "epoch": 1.2789507357645553, "grad_norm": 2.5734373406521804, "learning_rate": 1.950162512695522e-05, "loss": 0.8372, "step": 1999 }, { "epoch": 1.2795905310300704, "grad_norm": 2.61405679866298, "learning_rate": 1.9500978918538823e-05, "loss": 0.7932, "step": 2000 }, { "epoch": 1.2802303262955854, "grad_norm": 2.429252285165472, "learning_rate": 1.9500332302167292e-05, "loss": 0.7916, "step": 2001 }, { "epoch": 1.2808701215611005, "grad_norm": 2.30817852693649, "learning_rate": 1.9499685277868394e-05, "loss": 0.8022, "step": 2002 }, { "epoch": 1.2815099168266155, "grad_norm": 2.1524625384932654, "learning_rate": 1.9499037845669907e-05, "loss": 0.8367, "step": 2003 }, { "epoch": 1.2821497120921306, "grad_norm": 2.289416759212278, "learning_rate": 1.9498390005599632e-05, "loss": 0.8307, "step": 2004 }, { "epoch": 1.2827895073576456, "grad_norm": 2.5662934888909352, "learning_rate": 1.9497741757685388e-05, "loss": 0.7927, "step": 2005 }, { "epoch": 1.2834293026231607, "grad_norm": 2.8002545866838187, "learning_rate": 1.9497093101955005e-05, "loss": 0.7765, "step": 2006 }, { "epoch": 1.2840690978886755, "grad_norm": 2.524963017959201, "learning_rate": 1.9496444038436338e-05, "loss": 0.7934, "step": 2007 }, { "epoch": 1.2847088931541908, "grad_norm": 2.3407172972003636, "learning_rate": 1.9495794567157264e-05, "loss": 0.8068, "step": 2008 }, { "epoch": 1.2853486884197056, "grad_norm": 2.391231091516926, "learning_rate": 1.9495144688145656e-05, "loss": 0.754, "step": 2009 }, { "epoch": 1.2859884836852207, "grad_norm": 2.7829654868432825, "learning_rate": 1.9494494401429427e-05, "loss": 0.7971, "step": 2010 }, { "epoch": 1.2866282789507357, "grad_norm": 2.5015971498349043, "learning_rate": 1.94938437070365e-05, "loss": 0.7957, "step": 2011 }, { "epoch": 1.2872680742162508, "grad_norm": 2.2121586013455494, "learning_rate": 1.949319260499481e-05, "loss": 0.7618, "step": 2012 }, { "epoch": 1.2879078694817658, "grad_norm": 2.1506921850667386, "learning_rate": 1.9492541095332317e-05, "loss": 0.7809, "step": 2013 }, { "epoch": 1.2885476647472809, "grad_norm": 2.9729425635210647, "learning_rate": 1.9491889178076997e-05, "loss": 0.8715, "step": 2014 }, { "epoch": 1.289187460012796, "grad_norm": 2.4931522501479284, "learning_rate": 1.949123685325684e-05, "loss": 0.8059, "step": 2015 }, { "epoch": 1.289827255278311, "grad_norm": 2.53620322080421, "learning_rate": 1.9490584120899854e-05, "loss": 0.8315, "step": 2016 }, { "epoch": 1.290467050543826, "grad_norm": 2.1256106015480585, "learning_rate": 1.948993098103407e-05, "loss": 0.8085, "step": 2017 }, { "epoch": 1.2911068458093409, "grad_norm": 2.2902448331060756, "learning_rate": 1.948927743368753e-05, "loss": 0.7905, "step": 2018 }, { "epoch": 1.2917466410748562, "grad_norm": 2.496834186374089, "learning_rate": 1.94886234788883e-05, "loss": 0.797, "step": 2019 }, { "epoch": 1.292386436340371, "grad_norm": 2.2431217125416727, "learning_rate": 1.9487969116664452e-05, "loss": 0.8332, "step": 2020 }, { "epoch": 1.293026231605886, "grad_norm": 2.525194884938012, "learning_rate": 1.948731434704409e-05, "loss": 0.844, "step": 2021 }, { "epoch": 1.293666026871401, "grad_norm": 2.3939324534317623, "learning_rate": 1.9486659170055327e-05, "loss": 0.8248, "step": 2022 }, { "epoch": 1.2943058221369161, "grad_norm": 2.3917679846872737, "learning_rate": 1.9486003585726293e-05, "loss": 0.7742, "step": 2023 }, { "epoch": 1.2949456174024312, "grad_norm": 2.3733062085315857, "learning_rate": 1.9485347594085145e-05, "loss": 0.7616, "step": 2024 }, { "epoch": 1.2955854126679462, "grad_norm": 2.5138950747913724, "learning_rate": 1.9484691195160035e-05, "loss": 0.8489, "step": 2025 }, { "epoch": 1.2962252079334613, "grad_norm": 2.301088161856644, "learning_rate": 1.9484034388979167e-05, "loss": 0.8275, "step": 2026 }, { "epoch": 1.2968650031989764, "grad_norm": 2.412745059788646, "learning_rate": 1.9483377175570727e-05, "loss": 0.7978, "step": 2027 }, { "epoch": 1.2975047984644914, "grad_norm": 2.2591361578304165, "learning_rate": 1.9482719554962944e-05, "loss": 0.7946, "step": 2028 }, { "epoch": 1.2981445937300065, "grad_norm": 2.4426989096952467, "learning_rate": 1.9482061527184048e-05, "loss": 0.7967, "step": 2029 }, { "epoch": 1.2987843889955215, "grad_norm": 2.462749489340128, "learning_rate": 1.9481403092262303e-05, "loss": 0.77, "step": 2030 }, { "epoch": 1.2994241842610363, "grad_norm": 2.388217185739687, "learning_rate": 1.948074425022597e-05, "loss": 0.8182, "step": 2031 }, { "epoch": 1.3000639795265516, "grad_norm": 2.871517354247809, "learning_rate": 1.9480085001103344e-05, "loss": 0.776, "step": 2032 }, { "epoch": 1.3007037747920664, "grad_norm": 2.254518957658817, "learning_rate": 1.9479425344922733e-05, "loss": 0.8197, "step": 2033 }, { "epoch": 1.3013435700575815, "grad_norm": 2.3743737340496525, "learning_rate": 1.947876528171246e-05, "loss": 0.8011, "step": 2034 }, { "epoch": 1.3019833653230966, "grad_norm": 2.908197471092148, "learning_rate": 1.947810481150087e-05, "loss": 0.8623, "step": 2035 }, { "epoch": 1.3026231605886116, "grad_norm": 2.523419132668681, "learning_rate": 1.9477443934316316e-05, "loss": 0.833, "step": 2036 }, { "epoch": 1.3032629558541267, "grad_norm": 2.269421528149178, "learning_rate": 1.947678265018718e-05, "loss": 0.7729, "step": 2037 }, { "epoch": 1.3039027511196417, "grad_norm": 2.4446880910592634, "learning_rate": 1.9476120959141857e-05, "loss": 0.797, "step": 2038 }, { "epoch": 1.3045425463851568, "grad_norm": 2.1295634209311416, "learning_rate": 1.9475458861208754e-05, "loss": 0.7727, "step": 2039 }, { "epoch": 1.3051823416506718, "grad_norm": 2.307686356886326, "learning_rate": 1.94747963564163e-05, "loss": 0.8245, "step": 2040 }, { "epoch": 1.3058221369161869, "grad_norm": 2.118426054438129, "learning_rate": 1.947413344479295e-05, "loss": 0.8195, "step": 2041 }, { "epoch": 1.306461932181702, "grad_norm": 2.4466923189897467, "learning_rate": 1.9473470126367162e-05, "loss": 0.7983, "step": 2042 }, { "epoch": 1.307101727447217, "grad_norm": 2.4283622414459267, "learning_rate": 1.947280640116742e-05, "loss": 0.8529, "step": 2043 }, { "epoch": 1.3077415227127318, "grad_norm": 2.296963381090531, "learning_rate": 1.9472142269222217e-05, "loss": 0.795, "step": 2044 }, { "epoch": 1.308381317978247, "grad_norm": 2.2945139626599267, "learning_rate": 1.947147773056008e-05, "loss": 0.8492, "step": 2045 }, { "epoch": 1.309021113243762, "grad_norm": 2.5377724578007816, "learning_rate": 1.9470812785209533e-05, "loss": 0.8399, "step": 2046 }, { "epoch": 1.309660908509277, "grad_norm": 2.1520189850115, "learning_rate": 1.9470147433199135e-05, "loss": 0.8135, "step": 2047 }, { "epoch": 1.310300703774792, "grad_norm": 2.605348042219643, "learning_rate": 1.9469481674557454e-05, "loss": 0.8191, "step": 2048 }, { "epoch": 1.310940499040307, "grad_norm": 2.3177794294478904, "learning_rate": 1.946881550931307e-05, "loss": 0.7944, "step": 2049 }, { "epoch": 1.3115802943058221, "grad_norm": 2.098224169342518, "learning_rate": 1.9468148937494598e-05, "loss": 0.7921, "step": 2050 }, { "epoch": 1.3122200895713372, "grad_norm": 2.6431468416308332, "learning_rate": 1.946748195913065e-05, "loss": 0.7707, "step": 2051 }, { "epoch": 1.3128598848368522, "grad_norm": 2.3804805348594047, "learning_rate": 1.946681457424987e-05, "loss": 0.752, "step": 2052 }, { "epoch": 1.3134996801023673, "grad_norm": 2.161985714995594, "learning_rate": 1.9466146782880907e-05, "loss": 0.8087, "step": 2053 }, { "epoch": 1.3141394753678823, "grad_norm": 2.349294739035834, "learning_rate": 1.946547858505244e-05, "loss": 0.8108, "step": 2054 }, { "epoch": 1.3147792706333974, "grad_norm": 2.8849469949106896, "learning_rate": 1.9464809980793166e-05, "loss": 0.7629, "step": 2055 }, { "epoch": 1.3154190658989124, "grad_norm": 2.4789376918458244, "learning_rate": 1.9464140970131786e-05, "loss": 0.7627, "step": 2056 }, { "epoch": 1.3160588611644273, "grad_norm": 2.6883694625010017, "learning_rate": 1.946347155309703e-05, "loss": 0.7798, "step": 2057 }, { "epoch": 1.3166986564299425, "grad_norm": 2.654148429844913, "learning_rate": 1.9462801729717636e-05, "loss": 0.8059, "step": 2058 }, { "epoch": 1.3173384516954574, "grad_norm": 2.594569726209508, "learning_rate": 1.946213150002237e-05, "loss": 0.7881, "step": 2059 }, { "epoch": 1.3179782469609724, "grad_norm": 2.1829684285084934, "learning_rate": 1.946146086404001e-05, "loss": 0.7495, "step": 2060 }, { "epoch": 1.3186180422264875, "grad_norm": 2.24815218001933, "learning_rate": 1.946078982179935e-05, "loss": 0.7249, "step": 2061 }, { "epoch": 1.3192578374920025, "grad_norm": 2.2311781862075692, "learning_rate": 1.9460118373329205e-05, "loss": 0.8308, "step": 2062 }, { "epoch": 1.3198976327575176, "grad_norm": 2.2190878248737964, "learning_rate": 1.9459446518658404e-05, "loss": 0.7474, "step": 2063 }, { "epoch": 1.3205374280230326, "grad_norm": 2.1039491523678073, "learning_rate": 1.94587742578158e-05, "loss": 0.8044, "step": 2064 }, { "epoch": 1.3211772232885477, "grad_norm": 2.4995363692768398, "learning_rate": 1.9458101590830253e-05, "loss": 0.8287, "step": 2065 }, { "epoch": 1.3218170185540628, "grad_norm": 2.283240294000207, "learning_rate": 1.945742851773065e-05, "loss": 0.778, "step": 2066 }, { "epoch": 1.3224568138195778, "grad_norm": 2.132545851951541, "learning_rate": 1.9456755038545884e-05, "loss": 0.7902, "step": 2067 }, { "epoch": 1.3230966090850926, "grad_norm": 2.554680989398471, "learning_rate": 1.9456081153304886e-05, "loss": 0.8075, "step": 2068 }, { "epoch": 1.323736404350608, "grad_norm": 2.7314362273756996, "learning_rate": 1.945540686203658e-05, "loss": 0.7945, "step": 2069 }, { "epoch": 1.3243761996161227, "grad_norm": 2.334124617649052, "learning_rate": 1.9454732164769927e-05, "loss": 0.7362, "step": 2070 }, { "epoch": 1.3250159948816378, "grad_norm": 2.1721157901073673, "learning_rate": 1.9454057061533887e-05, "loss": 0.8022, "step": 2071 }, { "epoch": 1.3256557901471528, "grad_norm": 2.5055423801468577, "learning_rate": 1.9453381552357463e-05, "loss": 0.8017, "step": 2072 }, { "epoch": 1.326295585412668, "grad_norm": 2.4235650697057385, "learning_rate": 1.9452705637269647e-05, "loss": 0.7909, "step": 2073 }, { "epoch": 1.326935380678183, "grad_norm": 2.4471136684794903, "learning_rate": 1.9452029316299462e-05, "loss": 0.7773, "step": 2074 }, { "epoch": 1.327575175943698, "grad_norm": 2.4342283729736076, "learning_rate": 1.9451352589475958e-05, "loss": 0.7619, "step": 2075 }, { "epoch": 1.328214971209213, "grad_norm": 2.3518846058295124, "learning_rate": 1.9450675456828184e-05, "loss": 0.7856, "step": 2076 }, { "epoch": 1.328854766474728, "grad_norm": 2.1661666109055697, "learning_rate": 1.9449997918385217e-05, "loss": 0.8359, "step": 2077 }, { "epoch": 1.3294945617402432, "grad_norm": 2.6024615571107566, "learning_rate": 1.9449319974176152e-05, "loss": 0.8127, "step": 2078 }, { "epoch": 1.3301343570057582, "grad_norm": 2.3476594183928903, "learning_rate": 1.9448641624230093e-05, "loss": 0.8361, "step": 2079 }, { "epoch": 1.3307741522712733, "grad_norm": 2.361779717724429, "learning_rate": 1.9447962868576173e-05, "loss": 0.8314, "step": 2080 }, { "epoch": 1.331413947536788, "grad_norm": 2.178538547631646, "learning_rate": 1.9447283707243535e-05, "loss": 0.8213, "step": 2081 }, { "epoch": 1.3320537428023034, "grad_norm": 2.162940011968475, "learning_rate": 1.944660414026134e-05, "loss": 0.8327, "step": 2082 }, { "epoch": 1.3326935380678182, "grad_norm": 2.2651586365294594, "learning_rate": 1.9445924167658763e-05, "loss": 0.8091, "step": 2083 }, { "epoch": 1.3333333333333333, "grad_norm": 2.2702018483405637, "learning_rate": 1.944524378946501e-05, "loss": 0.8091, "step": 2084 }, { "epoch": 1.3339731285988483, "grad_norm": 2.2405641715510085, "learning_rate": 1.9444563005709286e-05, "loss": 0.7853, "step": 2085 }, { "epoch": 1.3346129238643634, "grad_norm": 2.498828428692908, "learning_rate": 1.9443881816420828e-05, "loss": 0.8259, "step": 2086 }, { "epoch": 1.3352527191298784, "grad_norm": 2.155888439266172, "learning_rate": 1.944320022162889e-05, "loss": 0.7503, "step": 2087 }, { "epoch": 1.3358925143953935, "grad_norm": 2.454296152055317, "learning_rate": 1.9442518221362725e-05, "loss": 0.7923, "step": 2088 }, { "epoch": 1.3365323096609085, "grad_norm": 2.518010721904445, "learning_rate": 1.944183581565163e-05, "loss": 0.7978, "step": 2089 }, { "epoch": 1.3371721049264236, "grad_norm": 2.051195870673785, "learning_rate": 1.94411530045249e-05, "loss": 0.7702, "step": 2090 }, { "epoch": 1.3378119001919386, "grad_norm": 2.08686697028604, "learning_rate": 1.944046978801185e-05, "loss": 0.7703, "step": 2091 }, { "epoch": 1.3384516954574537, "grad_norm": 2.1607977851255753, "learning_rate": 1.9439786166141824e-05, "loss": 0.817, "step": 2092 }, { "epoch": 1.3390914907229687, "grad_norm": 2.5318881847303234, "learning_rate": 1.943910213894417e-05, "loss": 0.791, "step": 2093 }, { "epoch": 1.3397312859884836, "grad_norm": 2.2804057156669533, "learning_rate": 1.9438417706448267e-05, "loss": 0.732, "step": 2094 }, { "epoch": 1.3403710812539988, "grad_norm": 2.3599113364551396, "learning_rate": 1.943773286868349e-05, "loss": 0.8159, "step": 2095 }, { "epoch": 1.3410108765195137, "grad_norm": 2.226485885524592, "learning_rate": 1.9437047625679256e-05, "loss": 0.8256, "step": 2096 }, { "epoch": 1.3416506717850287, "grad_norm": 2.7259592023105403, "learning_rate": 1.9436361977464985e-05, "loss": 0.8009, "step": 2097 }, { "epoch": 1.3422904670505438, "grad_norm": 2.4297963015747874, "learning_rate": 1.9435675924070115e-05, "loss": 0.7906, "step": 2098 }, { "epoch": 1.3429302623160588, "grad_norm": 2.2783334061961495, "learning_rate": 1.9434989465524103e-05, "loss": 0.796, "step": 2099 }, { "epoch": 1.3435700575815739, "grad_norm": 2.2818343534624685, "learning_rate": 1.943430260185643e-05, "loss": 0.7688, "step": 2100 }, { "epoch": 1.344209852847089, "grad_norm": 2.5370253197519945, "learning_rate": 1.9433615333096588e-05, "loss": 0.8067, "step": 2101 }, { "epoch": 1.344849648112604, "grad_norm": 2.3641472429412067, "learning_rate": 1.943292765927408e-05, "loss": 0.7917, "step": 2102 }, { "epoch": 1.345489443378119, "grad_norm": 2.5364578821630968, "learning_rate": 1.9432239580418438e-05, "loss": 0.7826, "step": 2103 }, { "epoch": 1.346129238643634, "grad_norm": 2.505073992570607, "learning_rate": 1.9431551096559213e-05, "loss": 0.7859, "step": 2104 }, { "epoch": 1.3467690339091491, "grad_norm": 2.3372808516085586, "learning_rate": 1.9430862207725957e-05, "loss": 0.7982, "step": 2105 }, { "epoch": 1.3474088291746642, "grad_norm": 2.4233559503787823, "learning_rate": 1.9430172913948254e-05, "loss": 0.8213, "step": 2106 }, { "epoch": 1.348048624440179, "grad_norm": 2.2693653662775377, "learning_rate": 1.94294832152557e-05, "loss": 0.7499, "step": 2107 }, { "epoch": 1.3486884197056943, "grad_norm": 3.0118227345905932, "learning_rate": 1.9428793111677913e-05, "loss": 0.8102, "step": 2108 }, { "epoch": 1.3493282149712091, "grad_norm": 2.7423577431571062, "learning_rate": 1.9428102603244522e-05, "loss": 0.7939, "step": 2109 }, { "epoch": 1.3499680102367242, "grad_norm": 2.3513484533116835, "learning_rate": 1.9427411689985175e-05, "loss": 0.7847, "step": 2110 }, { "epoch": 1.3506078055022392, "grad_norm": 2.4314885993901316, "learning_rate": 1.942672037192954e-05, "loss": 0.795, "step": 2111 }, { "epoch": 1.3512476007677543, "grad_norm": 2.5613672050084983, "learning_rate": 1.9426028649107304e-05, "loss": 0.774, "step": 2112 }, { "epoch": 1.3518873960332694, "grad_norm": 2.5276740841450582, "learning_rate": 1.942533652154816e-05, "loss": 0.7855, "step": 2113 }, { "epoch": 1.3525271912987844, "grad_norm": 2.524877135811689, "learning_rate": 1.942464398928184e-05, "loss": 0.8164, "step": 2114 }, { "epoch": 1.3531669865642995, "grad_norm": 2.362874033896682, "learning_rate": 1.9423951052338066e-05, "loss": 0.833, "step": 2115 }, { "epoch": 1.3538067818298145, "grad_norm": 2.241417107488204, "learning_rate": 1.9423257710746596e-05, "loss": 0.8523, "step": 2116 }, { "epoch": 1.3544465770953296, "grad_norm": 2.3864907419247894, "learning_rate": 1.9422563964537208e-05, "loss": 0.8215, "step": 2117 }, { "epoch": 1.3550863723608444, "grad_norm": 2.094897743262382, "learning_rate": 1.9421869813739683e-05, "loss": 0.7905, "step": 2118 }, { "epoch": 1.3557261676263597, "grad_norm": 2.389756007919711, "learning_rate": 1.9421175258383827e-05, "loss": 0.8465, "step": 2119 }, { "epoch": 1.3563659628918745, "grad_norm": 2.39015535762774, "learning_rate": 1.9420480298499466e-05, "loss": 0.8324, "step": 2120 }, { "epoch": 1.3570057581573896, "grad_norm": 2.0059451554826304, "learning_rate": 1.9419784934116438e-05, "loss": 0.7859, "step": 2121 }, { "epoch": 1.3576455534229046, "grad_norm": 1.940272654420114, "learning_rate": 1.9419089165264603e-05, "loss": 0.7666, "step": 2122 }, { "epoch": 1.3582853486884197, "grad_norm": 2.108518311253257, "learning_rate": 1.9418392991973832e-05, "loss": 0.7809, "step": 2123 }, { "epoch": 1.3589251439539347, "grad_norm": 2.250691533169469, "learning_rate": 1.9417696414274024e-05, "loss": 0.7775, "step": 2124 }, { "epoch": 1.3595649392194498, "grad_norm": 2.4081866973060615, "learning_rate": 1.941699943219508e-05, "loss": 0.7856, "step": 2125 }, { "epoch": 1.3602047344849648, "grad_norm": 2.445809878500096, "learning_rate": 1.9416302045766936e-05, "loss": 0.8077, "step": 2126 }, { "epoch": 1.3608445297504799, "grad_norm": 2.0511041139912844, "learning_rate": 1.9415604255019528e-05, "loss": 0.7524, "step": 2127 }, { "epoch": 1.361484325015995, "grad_norm": 2.262253895895992, "learning_rate": 1.9414906059982827e-05, "loss": 0.745, "step": 2128 }, { "epoch": 1.36212412028151, "grad_norm": 2.2637261258238857, "learning_rate": 1.9414207460686804e-05, "loss": 0.8583, "step": 2129 }, { "epoch": 1.362763915547025, "grad_norm": 2.5717322031333474, "learning_rate": 1.941350845716146e-05, "loss": 0.7902, "step": 2130 }, { "epoch": 1.3634037108125399, "grad_norm": 2.3152393297417238, "learning_rate": 1.941280904943681e-05, "loss": 0.7785, "step": 2131 }, { "epoch": 1.3640435060780551, "grad_norm": 2.1440715222925233, "learning_rate": 1.9412109237542883e-05, "loss": 0.7948, "step": 2132 }, { "epoch": 1.36468330134357, "grad_norm": 2.3720889563466216, "learning_rate": 1.9411409021509727e-05, "loss": 0.8112, "step": 2133 }, { "epoch": 1.365323096609085, "grad_norm": 2.2177893644296427, "learning_rate": 1.9410708401367408e-05, "loss": 0.7743, "step": 2134 }, { "epoch": 1.3659628918746, "grad_norm": 2.204501427864337, "learning_rate": 1.9410007377146012e-05, "loss": 0.7412, "step": 2135 }, { "epoch": 1.3666026871401151, "grad_norm": 2.4534921175985356, "learning_rate": 1.940930594887564e-05, "loss": 0.7778, "step": 2136 }, { "epoch": 1.3672424824056302, "grad_norm": 2.293901790715572, "learning_rate": 1.9408604116586404e-05, "loss": 0.7578, "step": 2137 }, { "epoch": 1.3678822776711452, "grad_norm": 2.3021630871927607, "learning_rate": 1.9407901880308447e-05, "loss": 0.8568, "step": 2138 }, { "epoch": 1.3685220729366603, "grad_norm": 2.2022494512621673, "learning_rate": 1.940719924007192e-05, "loss": 0.8048, "step": 2139 }, { "epoch": 1.3691618682021753, "grad_norm": 2.1465908073880535, "learning_rate": 1.940649619590699e-05, "loss": 0.7732, "step": 2140 }, { "epoch": 1.3698016634676904, "grad_norm": 2.0833057642318633, "learning_rate": 1.9405792747843845e-05, "loss": 0.7783, "step": 2141 }, { "epoch": 1.3704414587332054, "grad_norm": 2.145684638963198, "learning_rate": 1.9405088895912693e-05, "loss": 0.7865, "step": 2142 }, { "epoch": 1.3710812539987205, "grad_norm": 2.031026103842623, "learning_rate": 1.9404384640143754e-05, "loss": 0.8382, "step": 2143 }, { "epoch": 1.3717210492642353, "grad_norm": 2.5956022074334633, "learning_rate": 1.9403679980567266e-05, "loss": 0.8276, "step": 2144 }, { "epoch": 1.3723608445297506, "grad_norm": 2.5546313637167763, "learning_rate": 1.940297491721349e-05, "loss": 0.8268, "step": 2145 }, { "epoch": 1.3730006397952654, "grad_norm": 2.23601253257782, "learning_rate": 1.9402269450112696e-05, "loss": 0.7955, "step": 2146 }, { "epoch": 1.3736404350607805, "grad_norm": 2.8860553399252433, "learning_rate": 1.9401563579295177e-05, "loss": 0.7661, "step": 2147 }, { "epoch": 1.3742802303262955, "grad_norm": 2.471593806285718, "learning_rate": 1.940085730479124e-05, "loss": 0.8073, "step": 2148 }, { "epoch": 1.3749200255918106, "grad_norm": 2.6414807949812062, "learning_rate": 1.9400150626631218e-05, "loss": 0.8016, "step": 2149 }, { "epoch": 1.3755598208573256, "grad_norm": 2.6148448782171965, "learning_rate": 1.9399443544845446e-05, "loss": 0.807, "step": 2150 }, { "epoch": 1.3761996161228407, "grad_norm": 2.564999057241114, "learning_rate": 1.939873605946429e-05, "loss": 0.7647, "step": 2151 }, { "epoch": 1.3768394113883557, "grad_norm": 3.4722425432619692, "learning_rate": 1.939802817051813e-05, "loss": 0.7532, "step": 2152 }, { "epoch": 1.3774792066538708, "grad_norm": 2.2562579592632845, "learning_rate": 1.9397319878037352e-05, "loss": 0.7992, "step": 2153 }, { "epoch": 1.3781190019193859, "grad_norm": 2.315479171252177, "learning_rate": 1.9396611182052378e-05, "loss": 0.7795, "step": 2154 }, { "epoch": 1.378758797184901, "grad_norm": 2.464058902151907, "learning_rate": 1.9395902082593638e-05, "loss": 0.7856, "step": 2155 }, { "epoch": 1.379398592450416, "grad_norm": 2.4176000401172058, "learning_rate": 1.939519257969157e-05, "loss": 0.794, "step": 2156 }, { "epoch": 1.3800383877159308, "grad_norm": 2.343754324506537, "learning_rate": 1.9394482673376647e-05, "loss": 0.8107, "step": 2157 }, { "epoch": 1.380678182981446, "grad_norm": 2.3577379617426426, "learning_rate": 1.9393772363679353e-05, "loss": 0.8076, "step": 2158 }, { "epoch": 1.381317978246961, "grad_norm": 2.1205815841764513, "learning_rate": 1.9393061650630184e-05, "loss": 0.7796, "step": 2159 }, { "epoch": 1.381957773512476, "grad_norm": 2.321086036589019, "learning_rate": 1.9392350534259652e-05, "loss": 0.8002, "step": 2160 }, { "epoch": 1.382597568777991, "grad_norm": 2.413984948590053, "learning_rate": 1.9391639014598297e-05, "loss": 0.8269, "step": 2161 }, { "epoch": 1.383237364043506, "grad_norm": 2.3066436718265813, "learning_rate": 1.939092709167667e-05, "loss": 0.7851, "step": 2162 }, { "epoch": 1.383877159309021, "grad_norm": 2.3888626841353995, "learning_rate": 1.9390214765525343e-05, "loss": 0.7596, "step": 2163 }, { "epoch": 1.3845169545745362, "grad_norm": 2.103696852799195, "learning_rate": 1.9389502036174893e-05, "loss": 0.792, "step": 2164 }, { "epoch": 1.3851567498400512, "grad_norm": 2.0136719537791077, "learning_rate": 1.9388788903655928e-05, "loss": 0.8135, "step": 2165 }, { "epoch": 1.3857965451055663, "grad_norm": 2.204446482901249, "learning_rate": 1.938807536799907e-05, "loss": 0.746, "step": 2166 }, { "epoch": 1.3864363403710813, "grad_norm": 2.2998385738186657, "learning_rate": 1.938736142923496e-05, "loss": 0.802, "step": 2167 }, { "epoch": 1.3870761356365962, "grad_norm": 2.510438764455604, "learning_rate": 1.9386647087394244e-05, "loss": 0.8, "step": 2168 }, { "epoch": 1.3877159309021114, "grad_norm": 2.3047437266087027, "learning_rate": 1.9385932342507602e-05, "loss": 0.7961, "step": 2169 }, { "epoch": 1.3883557261676263, "grad_norm": 2.4300667883366147, "learning_rate": 1.938521719460572e-05, "loss": 0.8012, "step": 2170 }, { "epoch": 1.3889955214331415, "grad_norm": 2.288874181062469, "learning_rate": 1.9384501643719304e-05, "loss": 0.7838, "step": 2171 }, { "epoch": 1.3896353166986564, "grad_norm": 2.122649472086576, "learning_rate": 1.9383785689879087e-05, "loss": 0.7895, "step": 2172 }, { "epoch": 1.3902751119641714, "grad_norm": 2.4638499638910543, "learning_rate": 1.93830693331158e-05, "loss": 0.7526, "step": 2173 }, { "epoch": 1.3909149072296865, "grad_norm": 2.275171372265033, "learning_rate": 1.938235257346021e-05, "loss": 0.8149, "step": 2174 }, { "epoch": 1.3915547024952015, "grad_norm": 1.9431595042324612, "learning_rate": 1.938163541094309e-05, "loss": 0.7452, "step": 2175 }, { "epoch": 1.3921944977607166, "grad_norm": 2.1940402688155403, "learning_rate": 1.9380917845595233e-05, "loss": 0.7015, "step": 2176 }, { "epoch": 1.3928342930262316, "grad_norm": 2.396413806194683, "learning_rate": 1.9380199877447453e-05, "loss": 0.7618, "step": 2177 }, { "epoch": 1.3934740882917467, "grad_norm": 2.7886541008047008, "learning_rate": 1.9379481506530572e-05, "loss": 0.7812, "step": 2178 }, { "epoch": 1.3941138835572617, "grad_norm": 2.7116021912966715, "learning_rate": 1.9378762732875445e-05, "loss": 0.7447, "step": 2179 }, { "epoch": 1.3947536788227768, "grad_norm": 2.3892589451969517, "learning_rate": 1.937804355651293e-05, "loss": 0.7725, "step": 2180 }, { "epoch": 1.3953934740882916, "grad_norm": 2.471356244620341, "learning_rate": 1.93773239774739e-05, "loss": 0.7706, "step": 2181 }, { "epoch": 1.396033269353807, "grad_norm": 2.818472087705567, "learning_rate": 1.9376603995789268e-05, "loss": 0.7977, "step": 2182 }, { "epoch": 1.3966730646193217, "grad_norm": 2.958032202121251, "learning_rate": 1.9375883611489933e-05, "loss": 0.7893, "step": 2183 }, { "epoch": 1.3973128598848368, "grad_norm": 2.967392864347857, "learning_rate": 1.937516282460684e-05, "loss": 0.8165, "step": 2184 }, { "epoch": 1.3979526551503518, "grad_norm": 2.305264968647723, "learning_rate": 1.937444163517093e-05, "loss": 0.8127, "step": 2185 }, { "epoch": 1.3985924504158669, "grad_norm": 2.3781510834160993, "learning_rate": 1.937372004321317e-05, "loss": 0.7621, "step": 2186 }, { "epoch": 1.399232245681382, "grad_norm": 2.8421132659837145, "learning_rate": 1.9372998048764548e-05, "loss": 0.7589, "step": 2187 }, { "epoch": 1.399872040946897, "grad_norm": 2.752008036582886, "learning_rate": 1.937227565185606e-05, "loss": 0.7587, "step": 2188 }, { "epoch": 1.400511836212412, "grad_norm": 2.2804919316215777, "learning_rate": 1.937155285251873e-05, "loss": 0.854, "step": 2189 }, { "epoch": 1.401151631477927, "grad_norm": 2.235720698581284, "learning_rate": 1.9370829650783593e-05, "loss": 0.8065, "step": 2190 }, { "epoch": 1.4017914267434421, "grad_norm": 2.19325620166683, "learning_rate": 1.93701060466817e-05, "loss": 0.7434, "step": 2191 }, { "epoch": 1.4024312220089572, "grad_norm": 2.3370350432369515, "learning_rate": 1.936938204024412e-05, "loss": 0.7837, "step": 2192 }, { "epoch": 1.4030710172744723, "grad_norm": 2.2112837784482173, "learning_rate": 1.9368657631501943e-05, "loss": 0.8211, "step": 2193 }, { "epoch": 1.403710812539987, "grad_norm": 2.3398446661373518, "learning_rate": 1.936793282048627e-05, "loss": 0.7659, "step": 2194 }, { "epoch": 1.4043506078055024, "grad_norm": 2.20525276357062, "learning_rate": 1.936720760722823e-05, "loss": 0.7911, "step": 2195 }, { "epoch": 1.4049904030710172, "grad_norm": 2.071781814220588, "learning_rate": 1.9366481991758955e-05, "loss": 0.8155, "step": 2196 }, { "epoch": 1.4056301983365322, "grad_norm": 2.3163481134145387, "learning_rate": 1.9365755974109604e-05, "loss": 0.8238, "step": 2197 }, { "epoch": 1.4062699936020473, "grad_norm": 2.198572664239874, "learning_rate": 1.9365029554311355e-05, "loss": 0.7391, "step": 2198 }, { "epoch": 1.4069097888675623, "grad_norm": 2.1028843158427777, "learning_rate": 1.9364302732395395e-05, "loss": 0.7891, "step": 2199 }, { "epoch": 1.4075495841330774, "grad_norm": 2.337942795140262, "learning_rate": 1.9363575508392933e-05, "loss": 0.8065, "step": 2200 }, { "epoch": 1.4081893793985925, "grad_norm": 2.4167306351397877, "learning_rate": 1.9362847882335196e-05, "loss": 0.8116, "step": 2201 }, { "epoch": 1.4088291746641075, "grad_norm": 2.208706978860534, "learning_rate": 1.9362119854253426e-05, "loss": 0.7812, "step": 2202 }, { "epoch": 1.4094689699296226, "grad_norm": 2.5098662217795966, "learning_rate": 1.9361391424178883e-05, "loss": 0.7618, "step": 2203 }, { "epoch": 1.4101087651951376, "grad_norm": 2.2880535889332014, "learning_rate": 1.9360662592142846e-05, "loss": 0.8501, "step": 2204 }, { "epoch": 1.4107485604606527, "grad_norm": 2.68090310193555, "learning_rate": 1.9359933358176607e-05, "loss": 0.8081, "step": 2205 }, { "epoch": 1.4113883557261677, "grad_norm": 2.5310008865063454, "learning_rate": 1.935920372231148e-05, "loss": 0.7904, "step": 2206 }, { "epoch": 1.4120281509916826, "grad_norm": 2.3019599141656886, "learning_rate": 1.9358473684578795e-05, "loss": 0.7585, "step": 2207 }, { "epoch": 1.4126679462571978, "grad_norm": 2.2588740784191965, "learning_rate": 1.9357743245009898e-05, "loss": 0.8132, "step": 2208 }, { "epoch": 1.4133077415227127, "grad_norm": 2.5558205392607594, "learning_rate": 1.9357012403636147e-05, "loss": 0.8071, "step": 2209 }, { "epoch": 1.4139475367882277, "grad_norm": 2.518765469852632, "learning_rate": 1.9356281160488936e-05, "loss": 0.7475, "step": 2210 }, { "epoch": 1.4145873320537428, "grad_norm": 2.2308932626768714, "learning_rate": 1.935554951559965e-05, "loss": 0.7409, "step": 2211 }, { "epoch": 1.4152271273192578, "grad_norm": 2.36152059234552, "learning_rate": 1.935481746899971e-05, "loss": 0.7681, "step": 2212 }, { "epoch": 1.4158669225847729, "grad_norm": 2.276022923509964, "learning_rate": 1.935408502072055e-05, "loss": 0.7405, "step": 2213 }, { "epoch": 1.416506717850288, "grad_norm": 2.206855646061454, "learning_rate": 1.935335217079362e-05, "loss": 0.8269, "step": 2214 }, { "epoch": 1.417146513115803, "grad_norm": 2.2958219962812643, "learning_rate": 1.9352618919250384e-05, "loss": 0.7741, "step": 2215 }, { "epoch": 1.417786308381318, "grad_norm": 2.2284955867651104, "learning_rate": 1.9351885266122332e-05, "loss": 0.7734, "step": 2216 }, { "epoch": 1.418426103646833, "grad_norm": 2.107063297162008, "learning_rate": 1.935115121144096e-05, "loss": 0.8085, "step": 2217 }, { "epoch": 1.419065898912348, "grad_norm": 2.492686966382281, "learning_rate": 1.9350416755237792e-05, "loss": 0.8258, "step": 2218 }, { "epoch": 1.4197056941778632, "grad_norm": 2.4432006023525568, "learning_rate": 1.9349681897544358e-05, "loss": 0.7929, "step": 2219 }, { "epoch": 1.420345489443378, "grad_norm": 2.22854452088798, "learning_rate": 1.9348946638392218e-05, "loss": 0.7912, "step": 2220 }, { "epoch": 1.4209852847088933, "grad_norm": 1.999201833341955, "learning_rate": 1.9348210977812936e-05, "loss": 0.7584, "step": 2221 }, { "epoch": 1.4216250799744081, "grad_norm": 2.653470956478345, "learning_rate": 1.9347474915838107e-05, "loss": 0.8088, "step": 2222 }, { "epoch": 1.4222648752399232, "grad_norm": 2.5226569129584924, "learning_rate": 1.9346738452499332e-05, "loss": 0.7478, "step": 2223 }, { "epoch": 1.4229046705054382, "grad_norm": 2.4021847265886134, "learning_rate": 1.9346001587828235e-05, "loss": 0.7149, "step": 2224 }, { "epoch": 1.4235444657709533, "grad_norm": 2.367946210977854, "learning_rate": 1.9345264321856454e-05, "loss": 0.8308, "step": 2225 }, { "epoch": 1.4241842610364683, "grad_norm": 2.4642006973624033, "learning_rate": 1.9344526654615646e-05, "loss": 0.7844, "step": 2226 }, { "epoch": 1.4248240563019834, "grad_norm": 2.3886534319116923, "learning_rate": 1.934378858613749e-05, "loss": 0.7502, "step": 2227 }, { "epoch": 1.4254638515674984, "grad_norm": 2.622952614198573, "learning_rate": 1.934305011645367e-05, "loss": 0.8166, "step": 2228 }, { "epoch": 1.4261036468330135, "grad_norm": 2.3908061794635938, "learning_rate": 1.9342311245595898e-05, "loss": 0.7904, "step": 2229 }, { "epoch": 1.4267434420985285, "grad_norm": 2.0689540401298605, "learning_rate": 1.93415719735959e-05, "loss": 0.7799, "step": 2230 }, { "epoch": 1.4273832373640434, "grad_norm": 2.128612466659814, "learning_rate": 1.9340832300485423e-05, "loss": 0.8, "step": 2231 }, { "epoch": 1.4280230326295587, "grad_norm": 2.0909729078451913, "learning_rate": 1.934009222629622e-05, "loss": 0.8082, "step": 2232 }, { "epoch": 1.4286628278950735, "grad_norm": 2.371167242899515, "learning_rate": 1.9339351751060068e-05, "loss": 0.8, "step": 2233 }, { "epoch": 1.4293026231605885, "grad_norm": 2.3131331424770405, "learning_rate": 1.933861087480877e-05, "loss": 0.7739, "step": 2234 }, { "epoch": 1.4299424184261036, "grad_norm": 2.127861209884989, "learning_rate": 1.9337869597574133e-05, "loss": 0.7549, "step": 2235 }, { "epoch": 1.4305822136916186, "grad_norm": 2.0527549153831313, "learning_rate": 1.9337127919387984e-05, "loss": 0.7452, "step": 2236 }, { "epoch": 1.4312220089571337, "grad_norm": 2.196269520463173, "learning_rate": 1.933638584028217e-05, "loss": 0.7339, "step": 2237 }, { "epoch": 1.4318618042226487, "grad_norm": 2.5287535309392166, "learning_rate": 1.933564336028856e-05, "loss": 0.8171, "step": 2238 }, { "epoch": 1.4325015994881638, "grad_norm": 2.19955686570254, "learning_rate": 1.9334900479439032e-05, "loss": 0.7801, "step": 2239 }, { "epoch": 1.4331413947536789, "grad_norm": 2.2136204607718963, "learning_rate": 1.933415719776548e-05, "loss": 0.8053, "step": 2240 }, { "epoch": 1.433781190019194, "grad_norm": 2.832338393293692, "learning_rate": 1.933341351529982e-05, "loss": 0.7754, "step": 2241 }, { "epoch": 1.434420985284709, "grad_norm": 2.2993888562311526, "learning_rate": 1.933266943207399e-05, "loss": 0.7551, "step": 2242 }, { "epoch": 1.435060780550224, "grad_norm": 2.223317227899755, "learning_rate": 1.9331924948119935e-05, "loss": 0.758, "step": 2243 }, { "epoch": 1.4357005758157388, "grad_norm": 2.340016301359923, "learning_rate": 1.9331180063469623e-05, "loss": 0.7776, "step": 2244 }, { "epoch": 1.4363403710812541, "grad_norm": 2.4402901937807098, "learning_rate": 1.9330434778155037e-05, "loss": 0.7802, "step": 2245 }, { "epoch": 1.436980166346769, "grad_norm": 2.0376042891071444, "learning_rate": 1.932968909220818e-05, "loss": 0.8027, "step": 2246 }, { "epoch": 1.437619961612284, "grad_norm": 2.5170676481705616, "learning_rate": 1.9328943005661072e-05, "loss": 0.768, "step": 2247 }, { "epoch": 1.438259756877799, "grad_norm": 2.151728577765208, "learning_rate": 1.9328196518545742e-05, "loss": 0.7303, "step": 2248 }, { "epoch": 1.438899552143314, "grad_norm": 2.422663451061526, "learning_rate": 1.9327449630894254e-05, "loss": 0.7914, "step": 2249 }, { "epoch": 1.4395393474088292, "grad_norm": 2.2355831850641477, "learning_rate": 1.9326702342738663e-05, "loss": 0.7181, "step": 2250 }, { "epoch": 1.4401791426743442, "grad_norm": 2.1748138817322693, "learning_rate": 1.9325954654111068e-05, "loss": 0.7885, "step": 2251 }, { "epoch": 1.4408189379398593, "grad_norm": 2.157935291896458, "learning_rate": 1.932520656504357e-05, "loss": 0.7855, "step": 2252 }, { "epoch": 1.4414587332053743, "grad_norm": 2.371345142487744, "learning_rate": 1.9324458075568288e-05, "loss": 0.8066, "step": 2253 }, { "epoch": 1.4420985284708894, "grad_norm": 2.1964597541160606, "learning_rate": 1.9323709185717365e-05, "loss": 0.7392, "step": 2254 }, { "epoch": 1.4427383237364044, "grad_norm": 2.2594845184888444, "learning_rate": 1.9322959895522957e-05, "loss": 0.7931, "step": 2255 }, { "epoch": 1.4433781190019195, "grad_norm": 2.395969149043448, "learning_rate": 1.9322210205017234e-05, "loss": 0.8084, "step": 2256 }, { "epoch": 1.4440179142674343, "grad_norm": 2.2624945133433125, "learning_rate": 1.9321460114232383e-05, "loss": 0.772, "step": 2257 }, { "epoch": 1.4446577095329496, "grad_norm": 2.587166545861343, "learning_rate": 1.9320709623200623e-05, "loss": 0.835, "step": 2258 }, { "epoch": 1.4452975047984644, "grad_norm": 2.2639033145593985, "learning_rate": 1.931995873195417e-05, "loss": 0.7744, "step": 2259 }, { "epoch": 1.4459373000639795, "grad_norm": 2.2101445744370216, "learning_rate": 1.9319207440525267e-05, "loss": 0.7183, "step": 2260 }, { "epoch": 1.4465770953294945, "grad_norm": 2.0878658130820567, "learning_rate": 1.9318455748946175e-05, "loss": 0.683, "step": 2261 }, { "epoch": 1.4472168905950096, "grad_norm": 2.3107180685307895, "learning_rate": 1.931770365724917e-05, "loss": 0.8388, "step": 2262 }, { "epoch": 1.4478566858605246, "grad_norm": 2.6448203007000544, "learning_rate": 1.9316951165466542e-05, "loss": 0.801, "step": 2263 }, { "epoch": 1.4484964811260397, "grad_norm": 2.15731946884606, "learning_rate": 1.9316198273630603e-05, "loss": 0.7561, "step": 2264 }, { "epoch": 1.4491362763915547, "grad_norm": 2.447222725370756, "learning_rate": 1.931544498177369e-05, "loss": 0.7428, "step": 2265 }, { "epoch": 1.4497760716570698, "grad_norm": 2.041817782525194, "learning_rate": 1.9314691289928135e-05, "loss": 0.7226, "step": 2266 }, { "epoch": 1.4504158669225848, "grad_norm": 2.1255604380870534, "learning_rate": 1.9313937198126305e-05, "loss": 0.7743, "step": 2267 }, { "epoch": 1.4510556621880997, "grad_norm": 2.2847099808088016, "learning_rate": 1.9313182706400583e-05, "loss": 0.7607, "step": 2268 }, { "epoch": 1.451695457453615, "grad_norm": 2.331454001570407, "learning_rate": 1.9312427814783363e-05, "loss": 0.7316, "step": 2269 }, { "epoch": 1.4523352527191298, "grad_norm": 2.1105763980424816, "learning_rate": 1.931167252330705e-05, "loss": 0.7572, "step": 2270 }, { "epoch": 1.452975047984645, "grad_norm": 2.2764996391486405, "learning_rate": 1.931091683200409e-05, "loss": 0.7883, "step": 2271 }, { "epoch": 1.4536148432501599, "grad_norm": 2.0468718318772807, "learning_rate": 1.9310160740906926e-05, "loss": 0.786, "step": 2272 }, { "epoch": 1.454254638515675, "grad_norm": 2.1999449237706212, "learning_rate": 1.930940425004802e-05, "loss": 0.7616, "step": 2273 }, { "epoch": 1.45489443378119, "grad_norm": 2.1850948350367747, "learning_rate": 1.9308647359459853e-05, "loss": 0.7498, "step": 2274 }, { "epoch": 1.455534229046705, "grad_norm": 2.067274137889163, "learning_rate": 1.930789006917493e-05, "loss": 0.7801, "step": 2275 }, { "epoch": 1.45617402431222, "grad_norm": 2.560883306086443, "learning_rate": 1.930713237922576e-05, "loss": 0.7351, "step": 2276 }, { "epoch": 1.4568138195777351, "grad_norm": 2.0346123997551326, "learning_rate": 1.9306374289644886e-05, "loss": 0.7507, "step": 2277 }, { "epoch": 1.4574536148432502, "grad_norm": 2.3716470228006656, "learning_rate": 1.9305615800464853e-05, "loss": 0.7817, "step": 2278 }, { "epoch": 1.4580934101087653, "grad_norm": 2.483602164132493, "learning_rate": 1.930485691171823e-05, "loss": 0.8, "step": 2279 }, { "epoch": 1.4587332053742803, "grad_norm": 2.313122844883987, "learning_rate": 1.9304097623437607e-05, "loss": 0.7777, "step": 2280 }, { "epoch": 1.4593730006397951, "grad_norm": 2.3569994496563997, "learning_rate": 1.930333793565558e-05, "loss": 0.8276, "step": 2281 }, { "epoch": 1.4600127959053104, "grad_norm": 2.22937691327178, "learning_rate": 1.9302577848404772e-05, "loss": 0.7798, "step": 2282 }, { "epoch": 1.4606525911708252, "grad_norm": 2.271936169754283, "learning_rate": 1.930181736171782e-05, "loss": 0.7158, "step": 2283 }, { "epoch": 1.4612923864363403, "grad_norm": 2.103770829048776, "learning_rate": 1.9301056475627374e-05, "loss": 0.8122, "step": 2284 }, { "epoch": 1.4619321817018553, "grad_norm": 2.0851024967733336, "learning_rate": 1.9300295190166115e-05, "loss": 0.7792, "step": 2285 }, { "epoch": 1.4625719769673704, "grad_norm": 2.438381228962716, "learning_rate": 1.9299533505366718e-05, "loss": 0.82, "step": 2286 }, { "epoch": 1.4632117722328855, "grad_norm": 2.094591400621754, "learning_rate": 1.9298771421261895e-05, "loss": 0.7715, "step": 2287 }, { "epoch": 1.4638515674984005, "grad_norm": 2.1469136201125636, "learning_rate": 1.9298008937884372e-05, "loss": 0.7765, "step": 2288 }, { "epoch": 1.4644913627639156, "grad_norm": 2.0776779481355074, "learning_rate": 1.9297246055266882e-05, "loss": 0.712, "step": 2289 }, { "epoch": 1.4651311580294306, "grad_norm": 2.1140513447835367, "learning_rate": 1.9296482773442188e-05, "loss": 0.7556, "step": 2290 }, { "epoch": 1.4657709532949457, "grad_norm": 2.396236010223906, "learning_rate": 1.929571909244306e-05, "loss": 0.7443, "step": 2291 }, { "epoch": 1.4664107485604607, "grad_norm": 2.3241641557468316, "learning_rate": 1.9294955012302287e-05, "loss": 0.7439, "step": 2292 }, { "epoch": 1.4670505438259758, "grad_norm": 2.5504365628770476, "learning_rate": 1.9294190533052682e-05, "loss": 0.82, "step": 2293 }, { "epoch": 1.4676903390914906, "grad_norm": 2.7430123683328125, "learning_rate": 1.9293425654727072e-05, "loss": 0.7785, "step": 2294 }, { "epoch": 1.4683301343570059, "grad_norm": 2.2702433023558486, "learning_rate": 1.929266037735829e-05, "loss": 0.7759, "step": 2295 }, { "epoch": 1.4689699296225207, "grad_norm": 2.4199553902114808, "learning_rate": 1.9291894700979206e-05, "loss": 0.7754, "step": 2296 }, { "epoch": 1.4696097248880358, "grad_norm": 2.9753084295606613, "learning_rate": 1.929112862562269e-05, "loss": 0.7327, "step": 2297 }, { "epoch": 1.4702495201535508, "grad_norm": 2.6534574505113278, "learning_rate": 1.9290362151321643e-05, "loss": 0.8092, "step": 2298 }, { "epoch": 1.4708893154190659, "grad_norm": 2.1558736717817455, "learning_rate": 1.928959527810897e-05, "loss": 0.7683, "step": 2299 }, { "epoch": 1.471529110684581, "grad_norm": 2.2856110435193067, "learning_rate": 1.9288828006017594e-05, "loss": 0.7395, "step": 2300 }, { "epoch": 1.472168905950096, "grad_norm": 2.322897016724489, "learning_rate": 1.9288060335080474e-05, "loss": 0.8399, "step": 2301 }, { "epoch": 1.472808701215611, "grad_norm": 2.0961054454188197, "learning_rate": 1.9287292265330562e-05, "loss": 0.7507, "step": 2302 }, { "epoch": 1.473448496481126, "grad_norm": 2.184687633214704, "learning_rate": 1.9286523796800846e-05, "loss": 0.7888, "step": 2303 }, { "epoch": 1.4740882917466411, "grad_norm": 2.4381699987036924, "learning_rate": 1.928575492952431e-05, "loss": 0.7384, "step": 2304 }, { "epoch": 1.4747280870121562, "grad_norm": 2.150115024644405, "learning_rate": 1.928498566353398e-05, "loss": 0.7582, "step": 2305 }, { "epoch": 1.4753678822776712, "grad_norm": 2.3255669318525074, "learning_rate": 1.928421599886288e-05, "loss": 0.8353, "step": 2306 }, { "epoch": 1.476007677543186, "grad_norm": 2.3632396586625855, "learning_rate": 1.9283445935544064e-05, "loss": 0.7421, "step": 2307 }, { "epoch": 1.4766474728087013, "grad_norm": 2.3969879768938265, "learning_rate": 1.9282675473610594e-05, "loss": 0.7609, "step": 2308 }, { "epoch": 1.4772872680742162, "grad_norm": 2.1172344802635044, "learning_rate": 1.9281904613095546e-05, "loss": 0.7664, "step": 2309 }, { "epoch": 1.4779270633397312, "grad_norm": 2.0946996875160075, "learning_rate": 1.9281133354032028e-05, "loss": 0.7542, "step": 2310 }, { "epoch": 1.4785668586052463, "grad_norm": 2.402532262901314, "learning_rate": 1.9280361696453155e-05, "loss": 0.7557, "step": 2311 }, { "epoch": 1.4792066538707613, "grad_norm": 2.1578870735120943, "learning_rate": 1.9279589640392057e-05, "loss": 0.7569, "step": 2312 }, { "epoch": 1.4798464491362764, "grad_norm": 2.2507488917421066, "learning_rate": 1.9278817185881887e-05, "loss": 0.7224, "step": 2313 }, { "epoch": 1.4804862444017914, "grad_norm": 2.1826030572550175, "learning_rate": 1.9278044332955815e-05, "loss": 0.7731, "step": 2314 }, { "epoch": 1.4811260396673065, "grad_norm": 2.2805469296240783, "learning_rate": 1.9277271081647024e-05, "loss": 0.7876, "step": 2315 }, { "epoch": 1.4817658349328215, "grad_norm": 2.2082029641730694, "learning_rate": 1.9276497431988713e-05, "loss": 0.766, "step": 2316 }, { "epoch": 1.4824056301983366, "grad_norm": 2.1510974527412587, "learning_rate": 1.9275723384014108e-05, "loss": 0.7065, "step": 2317 }, { "epoch": 1.4830454254638514, "grad_norm": 2.0719197868098065, "learning_rate": 1.9274948937756436e-05, "loss": 0.7275, "step": 2318 }, { "epoch": 1.4836852207293667, "grad_norm": 2.238019851382662, "learning_rate": 1.9274174093248957e-05, "loss": 0.7542, "step": 2319 }, { "epoch": 1.4843250159948815, "grad_norm": 2.2515281241466902, "learning_rate": 1.927339885052494e-05, "loss": 0.7637, "step": 2320 }, { "epoch": 1.4849648112603968, "grad_norm": 2.414832297310626, "learning_rate": 1.9272623209617675e-05, "loss": 0.8567, "step": 2321 }, { "epoch": 1.4856046065259116, "grad_norm": 2.0286342172282446, "learning_rate": 1.927184717056046e-05, "loss": 0.7659, "step": 2322 }, { "epoch": 1.4862444017914267, "grad_norm": 2.0680631110633536, "learning_rate": 1.9271070733386625e-05, "loss": 0.7487, "step": 2323 }, { "epoch": 1.4868841970569417, "grad_norm": 2.1528181185016684, "learning_rate": 1.9270293898129503e-05, "loss": 0.7219, "step": 2324 }, { "epoch": 1.4875239923224568, "grad_norm": 2.0324859991748534, "learning_rate": 1.926951666482245e-05, "loss": 0.7584, "step": 2325 }, { "epoch": 1.4881637875879719, "grad_norm": 2.236855188232371, "learning_rate": 1.9268739033498846e-05, "loss": 0.7022, "step": 2326 }, { "epoch": 1.488803582853487, "grad_norm": 2.160981293527208, "learning_rate": 1.9267961004192067e-05, "loss": 0.6784, "step": 2327 }, { "epoch": 1.489443378119002, "grad_norm": 2.027270962501867, "learning_rate": 1.9267182576935535e-05, "loss": 0.707, "step": 2328 }, { "epoch": 1.490083173384517, "grad_norm": 2.2236275228025795, "learning_rate": 1.9266403751762666e-05, "loss": 0.7576, "step": 2329 }, { "epoch": 1.490722968650032, "grad_norm": 2.1852981714146575, "learning_rate": 1.9265624528706906e-05, "loss": 0.796, "step": 2330 }, { "epoch": 1.491362763915547, "grad_norm": 2.0419498390627626, "learning_rate": 1.926484490780171e-05, "loss": 0.7626, "step": 2331 }, { "epoch": 1.4920025591810622, "grad_norm": 2.0058099646055076, "learning_rate": 1.9264064889080553e-05, "loss": 0.7284, "step": 2332 }, { "epoch": 1.492642354446577, "grad_norm": 2.0668451447129863, "learning_rate": 1.9263284472576932e-05, "loss": 0.7765, "step": 2333 }, { "epoch": 1.493282149712092, "grad_norm": 2.632983825490252, "learning_rate": 1.9262503658324353e-05, "loss": 0.7891, "step": 2334 }, { "epoch": 1.493921944977607, "grad_norm": 2.230551615251418, "learning_rate": 1.926172244635634e-05, "loss": 0.7804, "step": 2335 }, { "epoch": 1.4945617402431222, "grad_norm": 2.1166885824159642, "learning_rate": 1.9260940836706443e-05, "loss": 0.7152, "step": 2336 }, { "epoch": 1.4952015355086372, "grad_norm": 2.316506299266081, "learning_rate": 1.926015882940822e-05, "loss": 0.7677, "step": 2337 }, { "epoch": 1.4958413307741523, "grad_norm": 2.174506249584753, "learning_rate": 1.9259376424495252e-05, "loss": 0.8176, "step": 2338 }, { "epoch": 1.4964811260396673, "grad_norm": 2.3789210424284715, "learning_rate": 1.9258593622001128e-05, "loss": 0.7462, "step": 2339 }, { "epoch": 1.4971209213051824, "grad_norm": 1.983321198644209, "learning_rate": 1.9257810421959466e-05, "loss": 0.7594, "step": 2340 }, { "epoch": 1.4977607165706974, "grad_norm": 2.245376206557549, "learning_rate": 1.925702682440389e-05, "loss": 0.7613, "step": 2341 }, { "epoch": 1.4984005118362125, "grad_norm": 1.997085365985703, "learning_rate": 1.9256242829368053e-05, "loss": 0.873, "step": 2342 }, { "epoch": 1.4990403071017275, "grad_norm": 2.333426628512666, "learning_rate": 1.925545843688561e-05, "loss": 0.7074, "step": 2343 }, { "epoch": 1.4996801023672424, "grad_norm": 2.2402431095328397, "learning_rate": 1.925467364699025e-05, "loss": 0.7321, "step": 2344 }, { "epoch": 1.5003198976327576, "grad_norm": 2.270687908471727, "learning_rate": 1.9253888459715663e-05, "loss": 0.7941, "step": 2345 }, { "epoch": 1.5009596928982725, "grad_norm": 2.2283001860999536, "learning_rate": 1.9253102875095568e-05, "loss": 0.8006, "step": 2346 }, { "epoch": 1.5015994881637877, "grad_norm": 2.088657441057797, "learning_rate": 1.92523168931637e-05, "loss": 0.7159, "step": 2347 }, { "epoch": 1.5022392834293026, "grad_norm": 2.0717228102210092, "learning_rate": 1.9251530513953794e-05, "loss": 0.7647, "step": 2348 }, { "epoch": 1.5028790786948176, "grad_norm": 2.061952740501845, "learning_rate": 1.9250743737499633e-05, "loss": 0.7006, "step": 2349 }, { "epoch": 1.5035188739603327, "grad_norm": 2.4834763304547094, "learning_rate": 1.924995656383499e-05, "loss": 0.8089, "step": 2350 }, { "epoch": 1.5041586692258477, "grad_norm": 2.2454567512259365, "learning_rate": 1.9249168992993666e-05, "loss": 0.7264, "step": 2351 }, { "epoch": 1.5047984644913628, "grad_norm": 2.5099460524814003, "learning_rate": 1.9248381025009476e-05, "loss": 0.7171, "step": 2352 }, { "epoch": 1.5054382597568778, "grad_norm": 2.144487924718672, "learning_rate": 1.924759265991626e-05, "loss": 0.792, "step": 2353 }, { "epoch": 1.506078055022393, "grad_norm": 2.4384510759621425, "learning_rate": 1.9246803897747864e-05, "loss": 0.8008, "step": 2354 }, { "epoch": 1.5067178502879077, "grad_norm": 2.239126240261688, "learning_rate": 1.9246014738538157e-05, "loss": 0.7434, "step": 2355 }, { "epoch": 1.507357645553423, "grad_norm": 2.500163304792401, "learning_rate": 1.9245225182321025e-05, "loss": 0.7294, "step": 2356 }, { "epoch": 1.5079974408189378, "grad_norm": 2.905807328395072, "learning_rate": 1.9244435229130368e-05, "loss": 0.7835, "step": 2357 }, { "epoch": 1.508637236084453, "grad_norm": 2.222281062468308, "learning_rate": 1.924364487900011e-05, "loss": 0.6733, "step": 2358 }, { "epoch": 1.509277031349968, "grad_norm": 8.398669353783658, "learning_rate": 1.9242854131964178e-05, "loss": 0.7672, "step": 2359 }, { "epoch": 1.5099168266154832, "grad_norm": 2.5738677371288783, "learning_rate": 1.924206298805654e-05, "loss": 0.7638, "step": 2360 }, { "epoch": 1.510556621880998, "grad_norm": 2.2714242059407828, "learning_rate": 1.9241271447311153e-05, "loss": 0.7684, "step": 2361 }, { "epoch": 1.511196417146513, "grad_norm": 2.5730899171571555, "learning_rate": 1.9240479509762008e-05, "loss": 0.793, "step": 2362 }, { "epoch": 1.5118362124120281, "grad_norm": 2.689368791193641, "learning_rate": 1.923968717544311e-05, "loss": 0.7565, "step": 2363 }, { "epoch": 1.5124760076775432, "grad_norm": 2.5339091674883, "learning_rate": 1.9238894444388484e-05, "loss": 0.8001, "step": 2364 }, { "epoch": 1.5131158029430583, "grad_norm": 2.3852271367428908, "learning_rate": 1.923810131663216e-05, "loss": 0.8629, "step": 2365 }, { "epoch": 1.5137555982085733, "grad_norm": 2.495352579630951, "learning_rate": 1.9237307792208203e-05, "loss": 0.7902, "step": 2366 }, { "epoch": 1.5143953934740884, "grad_norm": 2.493243278224722, "learning_rate": 1.9236513871150682e-05, "loss": 0.7968, "step": 2367 }, { "epoch": 1.5150351887396032, "grad_norm": 2.5849303651078666, "learning_rate": 1.9235719553493684e-05, "loss": 0.7263, "step": 2368 }, { "epoch": 1.5156749840051185, "grad_norm": 2.289646335716002, "learning_rate": 1.923492483927132e-05, "loss": 0.7742, "step": 2369 }, { "epoch": 1.5163147792706333, "grad_norm": 2.5526444252756417, "learning_rate": 1.923412972851771e-05, "loss": 0.7986, "step": 2370 }, { "epoch": 1.5169545745361486, "grad_norm": 2.091985286445872, "learning_rate": 1.9233334221266995e-05, "loss": 0.7684, "step": 2371 }, { "epoch": 1.5175943698016634, "grad_norm": 2.709156175240054, "learning_rate": 1.923253831755333e-05, "loss": 0.7194, "step": 2372 }, { "epoch": 1.5182341650671785, "grad_norm": 2.2371485639927244, "learning_rate": 1.9231742017410897e-05, "loss": 0.829, "step": 2373 }, { "epoch": 1.5188739603326935, "grad_norm": 2.1581124334211936, "learning_rate": 1.9230945320873884e-05, "loss": 0.7382, "step": 2374 }, { "epoch": 1.5195137555982086, "grad_norm": 2.002079211925528, "learning_rate": 1.92301482279765e-05, "loss": 0.7678, "step": 2375 }, { "epoch": 1.5201535508637236, "grad_norm": 2.1424543140866894, "learning_rate": 1.9229350738752967e-05, "loss": 0.8072, "step": 2376 }, { "epoch": 1.5207933461292387, "grad_norm": 2.1452263450382834, "learning_rate": 1.9228552853237534e-05, "loss": 0.7448, "step": 2377 }, { "epoch": 1.5214331413947537, "grad_norm": 1.9372261472802363, "learning_rate": 1.9227754571464456e-05, "loss": 0.7102, "step": 2378 }, { "epoch": 1.5220729366602685, "grad_norm": 2.269708014547212, "learning_rate": 1.9226955893468013e-05, "loss": 0.7927, "step": 2379 }, { "epoch": 1.5227127319257838, "grad_norm": 2.1047939562861058, "learning_rate": 1.9226156819282496e-05, "loss": 0.8022, "step": 2380 }, { "epoch": 1.5233525271912987, "grad_norm": 2.11515132461146, "learning_rate": 1.922535734894222e-05, "loss": 0.7844, "step": 2381 }, { "epoch": 1.523992322456814, "grad_norm": 2.1432342814558467, "learning_rate": 1.9224557482481505e-05, "loss": 0.7487, "step": 2382 }, { "epoch": 1.5246321177223288, "grad_norm": 2.074797393250597, "learning_rate": 1.9223757219934703e-05, "loss": 0.8007, "step": 2383 }, { "epoch": 1.525271912987844, "grad_norm": 2.199811379299639, "learning_rate": 1.9222956561336176e-05, "loss": 0.7971, "step": 2384 }, { "epoch": 1.5259117082533589, "grad_norm": 2.078320295807538, "learning_rate": 1.92221555067203e-05, "loss": 0.7375, "step": 2385 }, { "epoch": 1.526551503518874, "grad_norm": 2.0604591312311555, "learning_rate": 1.9221354056121472e-05, "loss": 0.7533, "step": 2386 }, { "epoch": 1.527191298784389, "grad_norm": 2.1658902940957376, "learning_rate": 1.92205522095741e-05, "loss": 0.7649, "step": 2387 }, { "epoch": 1.527831094049904, "grad_norm": 2.2282504493746127, "learning_rate": 1.9219749967112623e-05, "loss": 0.774, "step": 2388 }, { "epoch": 1.528470889315419, "grad_norm": 2.1982536734918696, "learning_rate": 1.921894732877148e-05, "loss": 0.7742, "step": 2389 }, { "epoch": 1.5291106845809341, "grad_norm": 2.498460458979844, "learning_rate": 1.9218144294585142e-05, "loss": 0.7976, "step": 2390 }, { "epoch": 1.5297504798464492, "grad_norm": 2.529459017744298, "learning_rate": 1.921734086458808e-05, "loss": 0.8143, "step": 2391 }, { "epoch": 1.530390275111964, "grad_norm": 2.077184894015473, "learning_rate": 1.9216537038814803e-05, "loss": 0.7472, "step": 2392 }, { "epoch": 1.5310300703774793, "grad_norm": 2.3588810664847437, "learning_rate": 1.9215732817299818e-05, "loss": 0.7276, "step": 2393 }, { "epoch": 1.5316698656429941, "grad_norm": 2.1763238227920083, "learning_rate": 1.9214928200077657e-05, "loss": 0.782, "step": 2394 }, { "epoch": 1.5323096609085094, "grad_norm": 2.079021197914135, "learning_rate": 1.9214123187182876e-05, "loss": 0.7608, "step": 2395 }, { "epoch": 1.5329494561740242, "grad_norm": 2.200511190627426, "learning_rate": 1.921331777865003e-05, "loss": 0.7884, "step": 2396 }, { "epoch": 1.5335892514395395, "grad_norm": 2.3027130531415496, "learning_rate": 1.9212511974513716e-05, "loss": 0.7668, "step": 2397 }, { "epoch": 1.5342290467050543, "grad_norm": 2.3071093433613847, "learning_rate": 1.921170577480852e-05, "loss": 0.7319, "step": 2398 }, { "epoch": 1.5348688419705694, "grad_norm": 2.235123070082575, "learning_rate": 1.9210899179569064e-05, "loss": 0.7783, "step": 2399 }, { "epoch": 1.5355086372360844, "grad_norm": 2.136336862485221, "learning_rate": 1.9210092188829986e-05, "loss": 0.782, "step": 2400 }, { "epoch": 1.5361484325015995, "grad_norm": 2.1559210557837716, "learning_rate": 1.920928480262593e-05, "loss": 0.7807, "step": 2401 }, { "epoch": 1.5367882277671145, "grad_norm": 2.678432886010397, "learning_rate": 1.9208477020991565e-05, "loss": 0.7014, "step": 2402 }, { "epoch": 1.5374280230326296, "grad_norm": 2.3945609687189364, "learning_rate": 1.9207668843961584e-05, "loss": 0.7756, "step": 2403 }, { "epoch": 1.5380678182981447, "grad_norm": 2.6336928695774984, "learning_rate": 1.9206860271570677e-05, "loss": 0.7974, "step": 2404 }, { "epoch": 1.5387076135636595, "grad_norm": 2.1479434166094125, "learning_rate": 1.9206051303853566e-05, "loss": 0.7548, "step": 2405 }, { "epoch": 1.5393474088291748, "grad_norm": 2.274518661718939, "learning_rate": 1.920524194084499e-05, "loss": 0.7528, "step": 2406 }, { "epoch": 1.5399872040946896, "grad_norm": 2.0980031077424663, "learning_rate": 1.9204432182579703e-05, "loss": 0.7573, "step": 2407 }, { "epoch": 1.5406269993602049, "grad_norm": 2.341622718689551, "learning_rate": 1.920362202909247e-05, "loss": 0.8007, "step": 2408 }, { "epoch": 1.5412667946257197, "grad_norm": 2.3348091283069876, "learning_rate": 1.9202811480418075e-05, "loss": 0.7544, "step": 2409 }, { "epoch": 1.541906589891235, "grad_norm": 2.1806448817668254, "learning_rate": 1.9202000536591327e-05, "loss": 0.7637, "step": 2410 }, { "epoch": 1.5425463851567498, "grad_norm": 2.284080604898056, "learning_rate": 1.920118919764705e-05, "loss": 0.8649, "step": 2411 }, { "epoch": 1.5431861804222649, "grad_norm": 2.4554551765245147, "learning_rate": 1.9200377463620072e-05, "loss": 0.7201, "step": 2412 }, { "epoch": 1.54382597568778, "grad_norm": 2.3327652464735644, "learning_rate": 1.919956533454525e-05, "loss": 0.7836, "step": 2413 }, { "epoch": 1.544465770953295, "grad_norm": 2.4953172255879728, "learning_rate": 1.919875281045746e-05, "loss": 0.7597, "step": 2414 }, { "epoch": 1.54510556621881, "grad_norm": 2.554478128974681, "learning_rate": 1.919793989139159e-05, "loss": 0.7678, "step": 2415 }, { "epoch": 1.545745361484325, "grad_norm": 1.9560984712415537, "learning_rate": 1.919712657738254e-05, "loss": 0.7262, "step": 2416 }, { "epoch": 1.5463851567498401, "grad_norm": 2.852118290111308, "learning_rate": 1.9196312868465236e-05, "loss": 0.7816, "step": 2417 }, { "epoch": 1.547024952015355, "grad_norm": 2.4106233092728, "learning_rate": 1.919549876467462e-05, "loss": 0.817, "step": 2418 }, { "epoch": 1.5476647472808702, "grad_norm": 2.291363134296066, "learning_rate": 1.919468426604564e-05, "loss": 0.817, "step": 2419 }, { "epoch": 1.548304542546385, "grad_norm": 2.174769310890795, "learning_rate": 1.919386937261327e-05, "loss": 0.7743, "step": 2420 }, { "epoch": 1.5489443378119003, "grad_norm": 2.5125947400461164, "learning_rate": 1.9193054084412513e-05, "loss": 0.7701, "step": 2421 }, { "epoch": 1.5495841330774152, "grad_norm": 2.5590339126627812, "learning_rate": 1.919223840147836e-05, "loss": 0.7454, "step": 2422 }, { "epoch": 1.5502239283429302, "grad_norm": 2.0477041169417878, "learning_rate": 1.9191422323845847e-05, "loss": 0.724, "step": 2423 }, { "epoch": 1.5508637236084453, "grad_norm": 2.6881671538053036, "learning_rate": 1.9190605851550007e-05, "loss": 0.7563, "step": 2424 }, { "epoch": 1.5515035188739603, "grad_norm": 2.2670411349932156, "learning_rate": 1.91897889846259e-05, "loss": 0.7405, "step": 2425 }, { "epoch": 1.5521433141394754, "grad_norm": 2.158267299000513, "learning_rate": 1.9188971723108598e-05, "loss": 0.7702, "step": 2426 }, { "epoch": 1.5527831094049904, "grad_norm": 2.187606054971913, "learning_rate": 1.91881540670332e-05, "loss": 0.7741, "step": 2427 }, { "epoch": 1.5534229046705055, "grad_norm": 2.133858736999164, "learning_rate": 1.918733601643481e-05, "loss": 0.822, "step": 2428 }, { "epoch": 1.5540626999360203, "grad_norm": 2.3197769696331108, "learning_rate": 1.9186517571348554e-05, "loss": 0.7356, "step": 2429 }, { "epoch": 1.5547024952015356, "grad_norm": 2.3261775271237384, "learning_rate": 1.9185698731809578e-05, "loss": 0.7664, "step": 2430 }, { "epoch": 1.5553422904670504, "grad_norm": 2.4305707337724627, "learning_rate": 1.9184879497853036e-05, "loss": 0.7308, "step": 2431 }, { "epoch": 1.5559820857325657, "grad_norm": 2.4309069270244628, "learning_rate": 1.918405986951411e-05, "loss": 0.6993, "step": 2432 }, { "epoch": 1.5566218809980805, "grad_norm": 2.2137191229845885, "learning_rate": 1.9183239846827985e-05, "loss": 0.7125, "step": 2433 }, { "epoch": 1.5572616762635958, "grad_norm": 2.4584727456837125, "learning_rate": 1.918241942982988e-05, "loss": 0.7706, "step": 2434 }, { "epoch": 1.5579014715291106, "grad_norm": 2.4605054877244363, "learning_rate": 1.918159861855502e-05, "loss": 0.6927, "step": 2435 }, { "epoch": 1.5585412667946257, "grad_norm": 2.545221553312472, "learning_rate": 1.9180777413038647e-05, "loss": 0.7703, "step": 2436 }, { "epoch": 1.5591810620601407, "grad_norm": 2.2588655253575483, "learning_rate": 1.917995581331602e-05, "loss": 0.7705, "step": 2437 }, { "epoch": 1.5598208573256558, "grad_norm": 2.5681326417616144, "learning_rate": 1.9179133819422425e-05, "loss": 0.7173, "step": 2438 }, { "epoch": 1.5604606525911708, "grad_norm": 2.7887587360543806, "learning_rate": 1.9178311431393154e-05, "loss": 0.7568, "step": 2439 }, { "epoch": 1.561100447856686, "grad_norm": 2.301214937217443, "learning_rate": 1.917748864926351e-05, "loss": 0.7021, "step": 2440 }, { "epoch": 1.561740243122201, "grad_norm": 2.330328722034925, "learning_rate": 1.9176665473068833e-05, "loss": 0.7543, "step": 2441 }, { "epoch": 1.5623800383877158, "grad_norm": 2.5359453600292317, "learning_rate": 1.9175841902844462e-05, "loss": 0.7672, "step": 2442 }, { "epoch": 1.563019833653231, "grad_norm": 2.154186766008533, "learning_rate": 1.9175017938625766e-05, "loss": 0.7339, "step": 2443 }, { "epoch": 1.5636596289187459, "grad_norm": 3.0433245621970144, "learning_rate": 1.917419358044812e-05, "loss": 0.7726, "step": 2444 }, { "epoch": 1.5642994241842612, "grad_norm": 2.7173994221966353, "learning_rate": 1.917336882834692e-05, "loss": 0.8045, "step": 2445 }, { "epoch": 1.564939219449776, "grad_norm": 2.0720814101810405, "learning_rate": 1.917254368235758e-05, "loss": 0.7074, "step": 2446 }, { "epoch": 1.5655790147152913, "grad_norm": 2.5256785758126625, "learning_rate": 1.917171814251553e-05, "loss": 0.7624, "step": 2447 }, { "epoch": 1.566218809980806, "grad_norm": 2.6743616484270425, "learning_rate": 1.9170892208856222e-05, "loss": 0.7104, "step": 2448 }, { "epoch": 1.5668586052463211, "grad_norm": 2.471542241174038, "learning_rate": 1.9170065881415115e-05, "loss": 0.8084, "step": 2449 }, { "epoch": 1.5674984005118362, "grad_norm": 2.497384127049962, "learning_rate": 1.916923916022769e-05, "loss": 0.7265, "step": 2450 }, { "epoch": 1.5681381957773513, "grad_norm": 2.2421861185158196, "learning_rate": 1.9168412045329446e-05, "loss": 0.7508, "step": 2451 }, { "epoch": 1.5687779910428663, "grad_norm": 2.6995424759783417, "learning_rate": 1.91675845367559e-05, "loss": 0.7596, "step": 2452 }, { "epoch": 1.5694177863083814, "grad_norm": 2.2688268382292254, "learning_rate": 1.9166756634542583e-05, "loss": 0.7008, "step": 2453 }, { "epoch": 1.5700575815738964, "grad_norm": 2.6744156207539915, "learning_rate": 1.916592833872504e-05, "loss": 0.7565, "step": 2454 }, { "epoch": 1.5706973768394112, "grad_norm": 2.297301702921786, "learning_rate": 1.916509964933884e-05, "loss": 0.7578, "step": 2455 }, { "epoch": 1.5713371721049265, "grad_norm": 2.095929161117997, "learning_rate": 1.916427056641956e-05, "loss": 0.7747, "step": 2456 }, { "epoch": 1.5719769673704413, "grad_norm": 2.2916537195555247, "learning_rate": 1.916344109000281e-05, "loss": 0.7594, "step": 2457 }, { "epoch": 1.5726167626359566, "grad_norm": 2.3409670064446066, "learning_rate": 1.91626112201242e-05, "loss": 0.7587, "step": 2458 }, { "epoch": 1.5732565579014715, "grad_norm": 2.2079962491026897, "learning_rate": 1.9161780956819363e-05, "loss": 0.7158, "step": 2459 }, { "epoch": 1.5738963531669867, "grad_norm": 3.147787601264533, "learning_rate": 1.9160950300123944e-05, "loss": 0.7579, "step": 2460 }, { "epoch": 1.5745361484325016, "grad_norm": 2.7019785973206605, "learning_rate": 1.916011925007362e-05, "loss": 0.7895, "step": 2461 }, { "epoch": 1.5751759436980166, "grad_norm": 1.994450644812252, "learning_rate": 1.9159287806704068e-05, "loss": 0.7174, "step": 2462 }, { "epoch": 1.5758157389635317, "grad_norm": 2.354355406829615, "learning_rate": 1.9158455970050995e-05, "loss": 0.7988, "step": 2463 }, { "epoch": 1.5764555342290467, "grad_norm": 2.317549832890297, "learning_rate": 1.9157623740150106e-05, "loss": 0.7317, "step": 2464 }, { "epoch": 1.5770953294945618, "grad_norm": 2.7029538405071087, "learning_rate": 1.915679111703715e-05, "loss": 0.7847, "step": 2465 }, { "epoch": 1.5777351247600768, "grad_norm": 2.291245941872039, "learning_rate": 1.915595810074787e-05, "loss": 0.7529, "step": 2466 }, { "epoch": 1.5783749200255919, "grad_norm": 2.3136923439962422, "learning_rate": 1.915512469131804e-05, "loss": 0.8154, "step": 2467 }, { "epoch": 1.5790147152911067, "grad_norm": 2.197339858793122, "learning_rate": 1.9154290888783435e-05, "loss": 0.7965, "step": 2468 }, { "epoch": 1.579654510556622, "grad_norm": 2.108788347754063, "learning_rate": 1.9153456693179867e-05, "loss": 0.8082, "step": 2469 }, { "epoch": 1.5802943058221368, "grad_norm": 2.0366792438673387, "learning_rate": 1.915262210454315e-05, "loss": 0.7288, "step": 2470 }, { "epoch": 1.580934101087652, "grad_norm": 2.2268512232091537, "learning_rate": 1.9151787122909122e-05, "loss": 0.7009, "step": 2471 }, { "epoch": 1.581573896353167, "grad_norm": 2.126632124537794, "learning_rate": 1.915095174831364e-05, "loss": 0.6577, "step": 2472 }, { "epoch": 1.582213691618682, "grad_norm": 2.368703708426203, "learning_rate": 1.9150115980792557e-05, "loss": 0.7873, "step": 2473 }, { "epoch": 1.582853486884197, "grad_norm": 2.1808710051468903, "learning_rate": 1.914927982038178e-05, "loss": 0.7768, "step": 2474 }, { "epoch": 1.583493282149712, "grad_norm": 2.485099895404823, "learning_rate": 1.91484432671172e-05, "loss": 0.7404, "step": 2475 }, { "epoch": 1.5841330774152271, "grad_norm": 2.2772113394464717, "learning_rate": 1.914760632103474e-05, "loss": 0.7449, "step": 2476 }, { "epoch": 1.5847728726807422, "grad_norm": 2.4271675264252153, "learning_rate": 1.9146768982170336e-05, "loss": 0.7061, "step": 2477 }, { "epoch": 1.5854126679462572, "grad_norm": 2.2904617590473157, "learning_rate": 1.9145931250559943e-05, "loss": 0.6984, "step": 2478 }, { "epoch": 1.586052463211772, "grad_norm": 2.268388954894367, "learning_rate": 1.914509312623953e-05, "loss": 0.8258, "step": 2479 }, { "epoch": 1.5866922584772873, "grad_norm": 2.284732927307559, "learning_rate": 1.9144254609245086e-05, "loss": 0.7739, "step": 2480 }, { "epoch": 1.5873320537428022, "grad_norm": 2.1544177503518327, "learning_rate": 1.9143415699612617e-05, "loss": 0.7424, "step": 2481 }, { "epoch": 1.5879718490083174, "grad_norm": 2.0863744453815687, "learning_rate": 1.9142576397378144e-05, "loss": 0.7301, "step": 2482 }, { "epoch": 1.5886116442738323, "grad_norm": 2.518206871377574, "learning_rate": 1.9141736702577704e-05, "loss": 0.7677, "step": 2483 }, { "epoch": 1.5892514395393476, "grad_norm": 2.303986477541, "learning_rate": 1.914089661524735e-05, "loss": 0.7475, "step": 2484 }, { "epoch": 1.5898912348048624, "grad_norm": 2.0045924977377787, "learning_rate": 1.9140056135423156e-05, "loss": 0.7237, "step": 2485 }, { "epoch": 1.5905310300703774, "grad_norm": 2.0998857942434324, "learning_rate": 1.913921526314121e-05, "loss": 0.7156, "step": 2486 }, { "epoch": 1.5911708253358925, "grad_norm": 2.1382410791770687, "learning_rate": 1.913837399843762e-05, "loss": 0.7691, "step": 2487 }, { "epoch": 1.5918106206014075, "grad_norm": 2.382528301691891, "learning_rate": 1.9137532341348503e-05, "loss": 0.7716, "step": 2488 }, { "epoch": 1.5924504158669226, "grad_norm": 2.6460972112915266, "learning_rate": 1.9136690291910005e-05, "loss": 0.774, "step": 2489 }, { "epoch": 1.5930902111324377, "grad_norm": 2.550626581715863, "learning_rate": 1.9135847850158278e-05, "loss": 0.7295, "step": 2490 }, { "epoch": 1.5937300063979527, "grad_norm": 1.9688621880436288, "learning_rate": 1.9135005016129495e-05, "loss": 0.7093, "step": 2491 }, { "epoch": 1.5943698016634675, "grad_norm": 2.2319827387126696, "learning_rate": 1.9134161789859847e-05, "loss": 0.7445, "step": 2492 }, { "epoch": 1.5950095969289828, "grad_norm": 2.7312339517114785, "learning_rate": 1.9133318171385537e-05, "loss": 0.8142, "step": 2493 }, { "epoch": 1.5956493921944976, "grad_norm": 2.3654559538245707, "learning_rate": 1.9132474160742797e-05, "loss": 0.7539, "step": 2494 }, { "epoch": 1.596289187460013, "grad_norm": 2.230255366241034, "learning_rate": 1.913162975796786e-05, "loss": 0.7708, "step": 2495 }, { "epoch": 1.5969289827255277, "grad_norm": 2.3196377628064924, "learning_rate": 1.9130784963096986e-05, "loss": 0.6941, "step": 2496 }, { "epoch": 1.597568777991043, "grad_norm": 2.012908134503499, "learning_rate": 1.9129939776166445e-05, "loss": 0.78, "step": 2497 }, { "epoch": 1.5982085732565579, "grad_norm": 2.1527229505148546, "learning_rate": 1.9129094197212533e-05, "loss": 0.7543, "step": 2498 }, { "epoch": 1.598848368522073, "grad_norm": 2.6451711586907916, "learning_rate": 1.9128248226271556e-05, "loss": 0.7901, "step": 2499 }, { "epoch": 1.599488163787588, "grad_norm": 2.3759651379566855, "learning_rate": 1.9127401863379835e-05, "loss": 0.6837, "step": 2500 }, { "epoch": 1.600127959053103, "grad_norm": 2.150099628677015, "learning_rate": 1.9126555108573718e-05, "loss": 0.748, "step": 2501 }, { "epoch": 1.600767754318618, "grad_norm": 2.3466666709054373, "learning_rate": 1.9125707961889556e-05, "loss": 0.7739, "step": 2502 }, { "epoch": 1.6014075495841331, "grad_norm": 2.279977383510648, "learning_rate": 1.912486042336373e-05, "loss": 0.7347, "step": 2503 }, { "epoch": 1.6020473448496482, "grad_norm": 2.618933801668599, "learning_rate": 1.912401249303263e-05, "loss": 0.7665, "step": 2504 }, { "epoch": 1.602687140115163, "grad_norm": 2.5559232762977806, "learning_rate": 1.9123164170932663e-05, "loss": 0.7758, "step": 2505 }, { "epoch": 1.6033269353806783, "grad_norm": 2.346354600925907, "learning_rate": 1.9122315457100255e-05, "loss": 0.7331, "step": 2506 }, { "epoch": 1.603966730646193, "grad_norm": 2.0408887066374306, "learning_rate": 1.9121466351571847e-05, "loss": 0.7103, "step": 2507 }, { "epoch": 1.6046065259117084, "grad_norm": 2.137443588626945, "learning_rate": 1.91206168543839e-05, "loss": 0.8056, "step": 2508 }, { "epoch": 1.6052463211772232, "grad_norm": 2.0553099041893215, "learning_rate": 1.9119766965572894e-05, "loss": 0.6921, "step": 2509 }, { "epoch": 1.6058861164427385, "grad_norm": 2.5313941462560954, "learning_rate": 1.9118916685175317e-05, "loss": 0.7421, "step": 2510 }, { "epoch": 1.6065259117082533, "grad_norm": 2.286086531764571, "learning_rate": 1.9118066013227673e-05, "loss": 0.7555, "step": 2511 }, { "epoch": 1.6071657069737684, "grad_norm": 2.333304896099671, "learning_rate": 1.91172149497665e-05, "loss": 0.7202, "step": 2512 }, { "epoch": 1.6078055022392834, "grad_norm": 2.4828053416943305, "learning_rate": 1.911636349482833e-05, "loss": 0.7658, "step": 2513 }, { "epoch": 1.6084452975047985, "grad_norm": 2.3039895993160635, "learning_rate": 1.9115511648449732e-05, "loss": 0.6821, "step": 2514 }, { "epoch": 1.6090850927703135, "grad_norm": 2.208879823806762, "learning_rate": 1.9114659410667275e-05, "loss": 0.7303, "step": 2515 }, { "epoch": 1.6097248880358286, "grad_norm": 2.0618428639155355, "learning_rate": 1.911380678151756e-05, "loss": 0.7758, "step": 2516 }, { "epoch": 1.6103646833013436, "grad_norm": 2.0535593307089757, "learning_rate": 1.9112953761037195e-05, "loss": 0.7155, "step": 2517 }, { "epoch": 1.6110044785668585, "grad_norm": 2.3288514157548716, "learning_rate": 1.9112100349262807e-05, "loss": 0.7382, "step": 2518 }, { "epoch": 1.6116442738323737, "grad_norm": 2.3277537851841448, "learning_rate": 1.9111246546231036e-05, "loss": 0.7538, "step": 2519 }, { "epoch": 1.6122840690978886, "grad_norm": 1.9927517717274297, "learning_rate": 1.9110392351978545e-05, "loss": 0.7532, "step": 2520 }, { "epoch": 1.6129238643634038, "grad_norm": 2.1981811573762835, "learning_rate": 1.9109537766542017e-05, "loss": 0.7336, "step": 2521 }, { "epoch": 1.6135636596289187, "grad_norm": 2.249542069824237, "learning_rate": 1.910868278995814e-05, "loss": 0.7596, "step": 2522 }, { "epoch": 1.6142034548944337, "grad_norm": 2.127498559813793, "learning_rate": 1.9107827422263625e-05, "loss": 0.7932, "step": 2523 }, { "epoch": 1.6148432501599488, "grad_norm": 1.9865359139988712, "learning_rate": 1.9106971663495205e-05, "loss": 0.7176, "step": 2524 }, { "epoch": 1.6154830454254638, "grad_norm": 2.190104697687148, "learning_rate": 1.910611551368962e-05, "loss": 0.6789, "step": 2525 }, { "epoch": 1.616122840690979, "grad_norm": 1.9697540602588965, "learning_rate": 1.9105258972883632e-05, "loss": 0.7248, "step": 2526 }, { "epoch": 1.616762635956494, "grad_norm": 2.1841075143148774, "learning_rate": 1.9104402041114022e-05, "loss": 0.6826, "step": 2527 }, { "epoch": 1.617402431222009, "grad_norm": 2.2069766394428183, "learning_rate": 1.9103544718417585e-05, "loss": 0.8047, "step": 2528 }, { "epoch": 1.6180422264875238, "grad_norm": 2.128306245648173, "learning_rate": 1.9102687004831134e-05, "loss": 0.7283, "step": 2529 }, { "epoch": 1.618682021753039, "grad_norm": 2.5956425586240415, "learning_rate": 1.910182890039149e-05, "loss": 0.7554, "step": 2530 }, { "epoch": 1.619321817018554, "grad_norm": 2.51907686741424, "learning_rate": 1.9100970405135506e-05, "loss": 0.7646, "step": 2531 }, { "epoch": 1.6199616122840692, "grad_norm": 2.0940707975008506, "learning_rate": 1.9100111519100042e-05, "loss": 0.7268, "step": 2532 }, { "epoch": 1.620601407549584, "grad_norm": 2.154288314164801, "learning_rate": 1.9099252242321978e-05, "loss": 0.7485, "step": 2533 }, { "epoch": 1.6212412028150993, "grad_norm": 2.2189860308683427, "learning_rate": 1.909839257483821e-05, "loss": 0.7673, "step": 2534 }, { "epoch": 1.6218809980806141, "grad_norm": 2.1661417299204095, "learning_rate": 1.9097532516685645e-05, "loss": 0.7526, "step": 2535 }, { "epoch": 1.6225207933461292, "grad_norm": 2.174590665819113, "learning_rate": 1.9096672067901224e-05, "loss": 0.7053, "step": 2536 }, { "epoch": 1.6231605886116443, "grad_norm": 2.329983180810017, "learning_rate": 1.9095811228521884e-05, "loss": 0.7452, "step": 2537 }, { "epoch": 1.6238003838771593, "grad_norm": 2.1181574984966334, "learning_rate": 1.9094949998584587e-05, "loss": 0.7173, "step": 2538 }, { "epoch": 1.6244401791426744, "grad_norm": 2.4444084873573604, "learning_rate": 1.9094088378126317e-05, "loss": 0.6976, "step": 2539 }, { "epoch": 1.6250799744081894, "grad_norm": 2.0754441031610065, "learning_rate": 1.909322636718407e-05, "loss": 0.7124, "step": 2540 }, { "epoch": 1.6257197696737045, "grad_norm": 2.112207638041281, "learning_rate": 1.9092363965794857e-05, "loss": 0.7253, "step": 2541 }, { "epoch": 1.6263595649392193, "grad_norm": 1.9461598969324931, "learning_rate": 1.909150117399571e-05, "loss": 0.6524, "step": 2542 }, { "epoch": 1.6269993602047346, "grad_norm": 2.049104588851503, "learning_rate": 1.9090637991823676e-05, "loss": 0.7337, "step": 2543 }, { "epoch": 1.6276391554702494, "grad_norm": 2.108041284917719, "learning_rate": 1.9089774419315817e-05, "loss": 0.7267, "step": 2544 }, { "epoch": 1.6282789507357647, "grad_norm": 2.1907668050998534, "learning_rate": 1.9088910456509213e-05, "loss": 0.715, "step": 2545 }, { "epoch": 1.6289187460012795, "grad_norm": 2.607643076934787, "learning_rate": 1.908804610344096e-05, "loss": 0.7294, "step": 2546 }, { "epoch": 1.6295585412667948, "grad_norm": 2.2646994725682927, "learning_rate": 1.908718136014818e-05, "loss": 0.7772, "step": 2547 }, { "epoch": 1.6301983365323096, "grad_norm": 2.010895844674732, "learning_rate": 1.9086316226667993e-05, "loss": 0.7509, "step": 2548 }, { "epoch": 1.6308381317978247, "grad_norm": 2.1129419422912963, "learning_rate": 1.908545070303755e-05, "loss": 0.7368, "step": 2549 }, { "epoch": 1.6314779270633397, "grad_norm": 2.355984947658953, "learning_rate": 1.9084584789294018e-05, "loss": 0.7442, "step": 2550 }, { "epoch": 1.6321177223288548, "grad_norm": 2.1220745331597577, "learning_rate": 1.9083718485474576e-05, "loss": 0.7401, "step": 2551 }, { "epoch": 1.6327575175943698, "grad_norm": 2.1975855893804987, "learning_rate": 1.908285179161642e-05, "loss": 0.7312, "step": 2552 }, { "epoch": 1.6333973128598849, "grad_norm": 2.107628078126282, "learning_rate": 1.908198470775676e-05, "loss": 0.7113, "step": 2553 }, { "epoch": 1.6340371081254, "grad_norm": 2.4032231380389244, "learning_rate": 1.908111723393284e-05, "loss": 0.6907, "step": 2554 }, { "epoch": 1.6346769033909148, "grad_norm": 2.402314732925841, "learning_rate": 1.9080249370181897e-05, "loss": 0.7752, "step": 2555 }, { "epoch": 1.63531669865643, "grad_norm": 2.4404083566787502, "learning_rate": 1.9079381116541197e-05, "loss": 0.6875, "step": 2556 }, { "epoch": 1.6359564939219449, "grad_norm": 3.0111068435523656, "learning_rate": 1.9078512473048023e-05, "loss": 0.7529, "step": 2557 }, { "epoch": 1.6365962891874601, "grad_norm": 2.2294951112490127, "learning_rate": 1.9077643439739675e-05, "loss": 0.8047, "step": 2558 }, { "epoch": 1.637236084452975, "grad_norm": 2.4989616517955113, "learning_rate": 1.9076774016653463e-05, "loss": 0.7406, "step": 2559 }, { "epoch": 1.6378758797184902, "grad_norm": 2.221509427885728, "learning_rate": 1.9075904203826724e-05, "loss": 0.7353, "step": 2560 }, { "epoch": 1.638515674984005, "grad_norm": 2.816847704049578, "learning_rate": 1.90750340012968e-05, "loss": 0.7187, "step": 2561 }, { "epoch": 1.6391554702495201, "grad_norm": 2.393607987672672, "learning_rate": 1.907416340910106e-05, "loss": 0.7521, "step": 2562 }, { "epoch": 1.6397952655150352, "grad_norm": 2.246069406812188, "learning_rate": 1.9073292427276887e-05, "loss": 0.6924, "step": 2563 }, { "epoch": 1.6404350607805502, "grad_norm": 2.2397872985628813, "learning_rate": 1.9072421055861675e-05, "loss": 0.7249, "step": 2564 }, { "epoch": 1.6410748560460653, "grad_norm": 2.1549569313902572, "learning_rate": 1.9071549294892843e-05, "loss": 0.7881, "step": 2565 }, { "epoch": 1.6417146513115803, "grad_norm": 2.2610368637523015, "learning_rate": 1.9070677144407822e-05, "loss": 0.708, "step": 2566 }, { "epoch": 1.6423544465770954, "grad_norm": 2.705433555078149, "learning_rate": 1.906980460444406e-05, "loss": 0.7851, "step": 2567 }, { "epoch": 1.6429942418426102, "grad_norm": 2.4697074534285592, "learning_rate": 1.906893167503902e-05, "loss": 0.7423, "step": 2568 }, { "epoch": 1.6436340371081255, "grad_norm": 2.504954906371923, "learning_rate": 1.906805835623019e-05, "loss": 0.7745, "step": 2569 }, { "epoch": 1.6442738323736403, "grad_norm": 2.088080281361612, "learning_rate": 1.906718464805506e-05, "loss": 0.7056, "step": 2570 }, { "epoch": 1.6449136276391556, "grad_norm": 2.243222126747617, "learning_rate": 1.9066310550551153e-05, "loss": 0.719, "step": 2571 }, { "epoch": 1.6455534229046704, "grad_norm": 2.399681188572811, "learning_rate": 1.9065436063755998e-05, "loss": 0.7201, "step": 2572 }, { "epoch": 1.6461932181701855, "grad_norm": 2.719729639094115, "learning_rate": 1.9064561187707143e-05, "loss": 0.7783, "step": 2573 }, { "epoch": 1.6468330134357005, "grad_norm": 2.4025850969961673, "learning_rate": 1.906368592244216e-05, "loss": 0.7476, "step": 2574 }, { "epoch": 1.6474728087012156, "grad_norm": 2.1417570610155625, "learning_rate": 1.9062810267998623e-05, "loss": 0.7611, "step": 2575 }, { "epoch": 1.6481126039667306, "grad_norm": 2.0047533229631034, "learning_rate": 1.9061934224414136e-05, "loss": 0.71, "step": 2576 }, { "epoch": 1.6487523992322457, "grad_norm": 1.9996692205794047, "learning_rate": 1.9061057791726312e-05, "loss": 0.6882, "step": 2577 }, { "epoch": 1.6493921944977608, "grad_norm": 2.3455534771655557, "learning_rate": 1.9060180969972783e-05, "loss": 0.7854, "step": 2578 }, { "epoch": 1.6500319897632756, "grad_norm": 2.257531308332574, "learning_rate": 1.9059303759191203e-05, "loss": 0.6926, "step": 2579 }, { "epoch": 1.6506717850287909, "grad_norm": 2.10731320436926, "learning_rate": 1.9058426159419236e-05, "loss": 0.7326, "step": 2580 }, { "epoch": 1.6513115802943057, "grad_norm": 2.5835340562892077, "learning_rate": 1.905754817069456e-05, "loss": 0.6983, "step": 2581 }, { "epoch": 1.651951375559821, "grad_norm": 2.051412854692541, "learning_rate": 1.905666979305488e-05, "loss": 0.6839, "step": 2582 }, { "epoch": 1.6525911708253358, "grad_norm": 2.289991708858189, "learning_rate": 1.9055791026537908e-05, "loss": 0.7116, "step": 2583 }, { "epoch": 1.653230966090851, "grad_norm": 2.3652220841356626, "learning_rate": 1.9054911871181382e-05, "loss": 0.7505, "step": 2584 }, { "epoch": 1.653870761356366, "grad_norm": 2.3618747139946925, "learning_rate": 1.9054032327023044e-05, "loss": 0.7438, "step": 2585 }, { "epoch": 1.654510556621881, "grad_norm": 2.166514077916101, "learning_rate": 1.9053152394100664e-05, "loss": 0.7813, "step": 2586 }, { "epoch": 1.655150351887396, "grad_norm": 2.1144034635163207, "learning_rate": 1.9052272072452024e-05, "loss": 0.7127, "step": 2587 }, { "epoch": 1.655790147152911, "grad_norm": 2.1868092669387047, "learning_rate": 1.9051391362114927e-05, "loss": 0.7138, "step": 2588 }, { "epoch": 1.6564299424184261, "grad_norm": 2.253600706155126, "learning_rate": 1.9050510263127184e-05, "loss": 0.7483, "step": 2589 }, { "epoch": 1.6570697376839412, "grad_norm": 2.0792961382490427, "learning_rate": 1.904962877552663e-05, "loss": 0.7284, "step": 2590 }, { "epoch": 1.6577095329494562, "grad_norm": 2.1041286454458312, "learning_rate": 1.9048746899351115e-05, "loss": 0.7226, "step": 2591 }, { "epoch": 1.658349328214971, "grad_norm": 2.0896094408410235, "learning_rate": 1.9047864634638503e-05, "loss": 0.7034, "step": 2592 }, { "epoch": 1.6589891234804863, "grad_norm": 2.1889631592389263, "learning_rate": 1.904698198142668e-05, "loss": 0.7222, "step": 2593 }, { "epoch": 1.6596289187460012, "grad_norm": 2.185869433312282, "learning_rate": 1.9046098939753546e-05, "loss": 0.6973, "step": 2594 }, { "epoch": 1.6602687140115164, "grad_norm": 2.163313040944795, "learning_rate": 1.904521550965701e-05, "loss": 0.7374, "step": 2595 }, { "epoch": 1.6609085092770313, "grad_norm": 2.479408872189801, "learning_rate": 1.9044331691175014e-05, "loss": 0.7699, "step": 2596 }, { "epoch": 1.6615483045425465, "grad_norm": 2.479049066725537, "learning_rate": 1.9043447484345504e-05, "loss": 0.7343, "step": 2597 }, { "epoch": 1.6621880998080614, "grad_norm": 2.276370524711705, "learning_rate": 1.9042562889206446e-05, "loss": 0.7379, "step": 2598 }, { "epoch": 1.6628278950735764, "grad_norm": 2.05170285738945, "learning_rate": 1.9041677905795822e-05, "loss": 0.6798, "step": 2599 }, { "epoch": 1.6634676903390915, "grad_norm": 2.327934756319044, "learning_rate": 1.9040792534151632e-05, "loss": 0.7307, "step": 2600 }, { "epoch": 1.6641074856046065, "grad_norm": 2.0712249516094503, "learning_rate": 1.9039906774311893e-05, "loss": 0.7176, "step": 2601 }, { "epoch": 1.6647472808701216, "grad_norm": 2.19145847399915, "learning_rate": 1.9039020626314637e-05, "loss": 0.7147, "step": 2602 }, { "epoch": 1.6653870761356366, "grad_norm": 2.089606698966587, "learning_rate": 1.9038134090197916e-05, "loss": 0.7158, "step": 2603 }, { "epoch": 1.6660268714011517, "grad_norm": 2.2034933388249054, "learning_rate": 1.9037247165999796e-05, "loss": 0.7311, "step": 2604 }, { "epoch": 1.6666666666666665, "grad_norm": 2.188451693744464, "learning_rate": 1.9036359853758356e-05, "loss": 0.7308, "step": 2605 }, { "epoch": 1.6673064619321818, "grad_norm": 2.0952012976897523, "learning_rate": 1.90354721535117e-05, "loss": 0.744, "step": 2606 }, { "epoch": 1.6679462571976966, "grad_norm": 2.3301687175534034, "learning_rate": 1.9034584065297945e-05, "loss": 0.7068, "step": 2607 }, { "epoch": 1.668586052463212, "grad_norm": 2.1400509803118237, "learning_rate": 1.9033695589155217e-05, "loss": 0.7856, "step": 2608 }, { "epoch": 1.6692258477287267, "grad_norm": 2.0604194103748705, "learning_rate": 1.9032806725121672e-05, "loss": 0.7115, "step": 2609 }, { "epoch": 1.669865642994242, "grad_norm": 1.98389416862119, "learning_rate": 1.9031917473235474e-05, "loss": 0.7753, "step": 2610 }, { "epoch": 1.6705054382597568, "grad_norm": 2.3599164372292813, "learning_rate": 1.9031027833534808e-05, "loss": 0.7536, "step": 2611 }, { "epoch": 1.671145233525272, "grad_norm": 2.1225389610701444, "learning_rate": 1.9030137806057868e-05, "loss": 0.7374, "step": 2612 }, { "epoch": 1.671785028790787, "grad_norm": 2.198134917360874, "learning_rate": 1.9029247390842878e-05, "loss": 0.7466, "step": 2613 }, { "epoch": 1.672424824056302, "grad_norm": 2.287951631577852, "learning_rate": 1.9028356587928064e-05, "loss": 0.7165, "step": 2614 }, { "epoch": 1.673064619321817, "grad_norm": 2.4663009305801182, "learning_rate": 1.902746539735168e-05, "loss": 0.7331, "step": 2615 }, { "epoch": 1.673704414587332, "grad_norm": 2.0988529519409167, "learning_rate": 1.902657381915199e-05, "loss": 0.7366, "step": 2616 }, { "epoch": 1.6743442098528472, "grad_norm": 2.3313782199594857, "learning_rate": 1.902568185336728e-05, "loss": 0.6805, "step": 2617 }, { "epoch": 1.674984005118362, "grad_norm": 2.1649120324214914, "learning_rate": 1.9024789500035842e-05, "loss": 0.66, "step": 2618 }, { "epoch": 1.6756238003838773, "grad_norm": 2.2227582088597604, "learning_rate": 1.9023896759195998e-05, "loss": 0.7376, "step": 2619 }, { "epoch": 1.676263595649392, "grad_norm": 2.1456177728222166, "learning_rate": 1.9023003630886082e-05, "loss": 0.7934, "step": 2620 }, { "epoch": 1.6769033909149074, "grad_norm": 2.2944095816572783, "learning_rate": 1.9022110115144435e-05, "loss": 0.7495, "step": 2621 }, { "epoch": 1.6775431861804222, "grad_norm": 2.5525899123153533, "learning_rate": 1.9021216212009433e-05, "loss": 0.713, "step": 2622 }, { "epoch": 1.6781829814459372, "grad_norm": 2.3531166423223056, "learning_rate": 1.9020321921519455e-05, "loss": 0.7782, "step": 2623 }, { "epoch": 1.6788227767114523, "grad_norm": 2.132000370243902, "learning_rate": 1.9019427243712897e-05, "loss": 0.7299, "step": 2624 }, { "epoch": 1.6794625719769674, "grad_norm": 2.214087470040354, "learning_rate": 1.901853217862818e-05, "loss": 0.7485, "step": 2625 }, { "epoch": 1.6801023672424824, "grad_norm": 2.0637307944270376, "learning_rate": 1.9017636726303733e-05, "loss": 0.7633, "step": 2626 }, { "epoch": 1.6807421625079975, "grad_norm": 2.1483477453525386, "learning_rate": 1.9016740886778006e-05, "loss": 0.6732, "step": 2627 }, { "epoch": 1.6813819577735125, "grad_norm": 2.194116677768287, "learning_rate": 1.9015844660089464e-05, "loss": 0.6902, "step": 2628 }, { "epoch": 1.6820217530390273, "grad_norm": 2.2330373386475726, "learning_rate": 1.9014948046276593e-05, "loss": 0.6782, "step": 2629 }, { "epoch": 1.6826615483045426, "grad_norm": 2.3632434443108323, "learning_rate": 1.9014051045377887e-05, "loss": 0.7612, "step": 2630 }, { "epoch": 1.6833013435700575, "grad_norm": 2.3478402912655407, "learning_rate": 1.9013153657431863e-05, "loss": 0.7603, "step": 2631 }, { "epoch": 1.6839411388355727, "grad_norm": 2.4221146106229523, "learning_rate": 1.9012255882477056e-05, "loss": 0.7275, "step": 2632 }, { "epoch": 1.6845809341010876, "grad_norm": 2.0811586334570595, "learning_rate": 1.9011357720552014e-05, "loss": 0.7617, "step": 2633 }, { "epoch": 1.6852207293666028, "grad_norm": 2.318708124992456, "learning_rate": 1.9010459171695298e-05, "loss": 0.7289, "step": 2634 }, { "epoch": 1.6858605246321177, "grad_norm": 2.128780990020201, "learning_rate": 1.9009560235945495e-05, "loss": 0.6765, "step": 2635 }, { "epoch": 1.6865003198976327, "grad_norm": 2.3616764334928653, "learning_rate": 1.90086609133412e-05, "loss": 0.6893, "step": 2636 }, { "epoch": 1.6871401151631478, "grad_norm": 2.1110704850284145, "learning_rate": 1.9007761203921035e-05, "loss": 0.7521, "step": 2637 }, { "epoch": 1.6877799104286628, "grad_norm": 2.3322773764257856, "learning_rate": 1.9006861107723623e-05, "loss": 0.7287, "step": 2638 }, { "epoch": 1.6884197056941779, "grad_norm": 2.5139716380913355, "learning_rate": 1.900596062478762e-05, "loss": 0.7167, "step": 2639 }, { "epoch": 1.689059500959693, "grad_norm": 2.4731606945563267, "learning_rate": 1.9005059755151692e-05, "loss": 0.6605, "step": 2640 }, { "epoch": 1.689699296225208, "grad_norm": 1.9838868039083148, "learning_rate": 1.900415849885451e-05, "loss": 0.6658, "step": 2641 }, { "epoch": 1.6903390914907228, "grad_norm": 2.3305043997774244, "learning_rate": 1.9003256855934784e-05, "loss": 0.7614, "step": 2642 }, { "epoch": 1.690978886756238, "grad_norm": 2.2148761844459353, "learning_rate": 1.900235482643122e-05, "loss": 0.6741, "step": 2643 }, { "epoch": 1.691618682021753, "grad_norm": 2.151403321122454, "learning_rate": 1.9001452410382553e-05, "loss": 0.6772, "step": 2644 }, { "epoch": 1.6922584772872682, "grad_norm": 2.069414949410158, "learning_rate": 1.9000549607827536e-05, "loss": 0.7142, "step": 2645 }, { "epoch": 1.692898272552783, "grad_norm": 2.22329587669618, "learning_rate": 1.899964641880493e-05, "loss": 0.7403, "step": 2646 }, { "epoch": 1.6935380678182983, "grad_norm": 1.9864240164973512, "learning_rate": 1.8998742843353514e-05, "loss": 0.7303, "step": 2647 }, { "epoch": 1.6941778630838131, "grad_norm": 2.166801995013646, "learning_rate": 1.8997838881512084e-05, "loss": 0.7057, "step": 2648 }, { "epoch": 1.6948176583493282, "grad_norm": 2.1006356899320493, "learning_rate": 1.8996934533319463e-05, "loss": 0.7261, "step": 2649 }, { "epoch": 1.6954574536148432, "grad_norm": 1.9716518126842166, "learning_rate": 1.8996029798814477e-05, "loss": 0.6767, "step": 2650 }, { "epoch": 1.6960972488803583, "grad_norm": 2.157627094789205, "learning_rate": 1.8995124678035976e-05, "loss": 0.7226, "step": 2651 }, { "epoch": 1.6967370441458733, "grad_norm": 2.198622485353027, "learning_rate": 1.8994219171022817e-05, "loss": 0.7661, "step": 2652 }, { "epoch": 1.6973768394113884, "grad_norm": 2.086563507155513, "learning_rate": 1.899331327781389e-05, "loss": 0.7353, "step": 2653 }, { "epoch": 1.6980166346769034, "grad_norm": 2.7829088827542976, "learning_rate": 1.8992406998448087e-05, "loss": 0.7173, "step": 2654 }, { "epoch": 1.6986564299424183, "grad_norm": 2.1434377589197795, "learning_rate": 1.8991500332964324e-05, "loss": 0.7253, "step": 2655 }, { "epoch": 1.6992962252079336, "grad_norm": 2.0060232378094334, "learning_rate": 1.899059328140153e-05, "loss": 0.6887, "step": 2656 }, { "epoch": 1.6999360204734484, "grad_norm": 2.0933727972869884, "learning_rate": 1.8989685843798658e-05, "loss": 0.6972, "step": 2657 }, { "epoch": 1.7005758157389637, "grad_norm": 2.3591050442013475, "learning_rate": 1.8988778020194668e-05, "loss": 0.7461, "step": 2658 }, { "epoch": 1.7012156110044785, "grad_norm": 2.2939798251155263, "learning_rate": 1.8987869810628535e-05, "loss": 0.7398, "step": 2659 }, { "epoch": 1.7018554062699938, "grad_norm": 2.2184106431768655, "learning_rate": 1.8986961215139263e-05, "loss": 0.6942, "step": 2660 }, { "epoch": 1.7024952015355086, "grad_norm": 2.085283465679251, "learning_rate": 1.898605223376586e-05, "loss": 0.6606, "step": 2661 }, { "epoch": 1.7031349968010236, "grad_norm": 2.256998368059722, "learning_rate": 1.898514286654736e-05, "loss": 0.6773, "step": 2662 }, { "epoch": 1.7037747920665387, "grad_norm": 2.708997348299704, "learning_rate": 1.8984233113522812e-05, "loss": 0.704, "step": 2663 }, { "epoch": 1.7044145873320538, "grad_norm": 2.4024114122822873, "learning_rate": 1.8983322974731274e-05, "loss": 0.6882, "step": 2664 }, { "epoch": 1.7050543825975688, "grad_norm": 2.1644111946898676, "learning_rate": 1.8982412450211827e-05, "loss": 0.6502, "step": 2665 }, { "epoch": 1.7056941778630839, "grad_norm": 2.288207059703266, "learning_rate": 1.898150154000357e-05, "loss": 0.6877, "step": 2666 }, { "epoch": 1.706333973128599, "grad_norm": 2.245374342312903, "learning_rate": 1.898059024414561e-05, "loss": 0.6622, "step": 2667 }, { "epoch": 1.7069737683941137, "grad_norm": 2.243058533293214, "learning_rate": 1.897967856267708e-05, "loss": 0.7362, "step": 2668 }, { "epoch": 1.707613563659629, "grad_norm": 2.1679677860507107, "learning_rate": 1.897876649563713e-05, "loss": 0.7142, "step": 2669 }, { "epoch": 1.7082533589251438, "grad_norm": 2.4149381231797484, "learning_rate": 1.8977854043064914e-05, "loss": 0.7107, "step": 2670 }, { "epoch": 1.7088931541906591, "grad_norm": 2.351314875473828, "learning_rate": 1.897694120499962e-05, "loss": 0.7177, "step": 2671 }, { "epoch": 1.709532949456174, "grad_norm": 2.1585842463488447, "learning_rate": 1.8976027981480437e-05, "loss": 0.6801, "step": 2672 }, { "epoch": 1.710172744721689, "grad_norm": 2.1394116965668117, "learning_rate": 1.897511437254658e-05, "loss": 0.7535, "step": 2673 }, { "epoch": 1.710812539987204, "grad_norm": 2.11843880334539, "learning_rate": 1.8974200378237276e-05, "loss": 0.7364, "step": 2674 }, { "epoch": 1.7114523352527191, "grad_norm": 1.9078829929612617, "learning_rate": 1.8973285998591776e-05, "loss": 0.712, "step": 2675 }, { "epoch": 1.7120921305182342, "grad_norm": 1.9788601331142264, "learning_rate": 1.8972371233649335e-05, "loss": 0.6765, "step": 2676 }, { "epoch": 1.7127319257837492, "grad_norm": 2.281408471481648, "learning_rate": 1.897145608344923e-05, "loss": 0.7372, "step": 2677 }, { "epoch": 1.7133717210492643, "grad_norm": 1.9655040523092815, "learning_rate": 1.8970540548030764e-05, "loss": 0.6979, "step": 2678 }, { "epoch": 1.714011516314779, "grad_norm": 2.2122957110859285, "learning_rate": 1.8969624627433246e-05, "loss": 0.6429, "step": 2679 }, { "epoch": 1.7146513115802944, "grad_norm": 1.9451843256529444, "learning_rate": 1.8968708321696e-05, "loss": 0.6583, "step": 2680 }, { "epoch": 1.7152911068458092, "grad_norm": 2.132315158438301, "learning_rate": 1.896779163085837e-05, "loss": 0.6839, "step": 2681 }, { "epoch": 1.7159309021113245, "grad_norm": 2.4921223887587316, "learning_rate": 1.8966874554959722e-05, "loss": 0.7057, "step": 2682 }, { "epoch": 1.7165706973768393, "grad_norm": 2.23731636976456, "learning_rate": 1.8965957094039434e-05, "loss": 0.7459, "step": 2683 }, { "epoch": 1.7172104926423546, "grad_norm": 2.3667515549995746, "learning_rate": 1.8965039248136895e-05, "loss": 0.6835, "step": 2684 }, { "epoch": 1.7178502879078694, "grad_norm": 2.332455809193711, "learning_rate": 1.8964121017291518e-05, "loss": 0.7218, "step": 2685 }, { "epoch": 1.7184900831733845, "grad_norm": 2.4511590956542646, "learning_rate": 1.8963202401542732e-05, "loss": 0.6756, "step": 2686 }, { "epoch": 1.7191298784388995, "grad_norm": 2.149005371051422, "learning_rate": 1.8962283400929977e-05, "loss": 0.6938, "step": 2687 }, { "epoch": 1.7197696737044146, "grad_norm": 2.3193256704059726, "learning_rate": 1.8961364015492716e-05, "loss": 0.7185, "step": 2688 }, { "epoch": 1.7204094689699296, "grad_norm": 2.2544381435031173, "learning_rate": 1.8960444245270427e-05, "loss": 0.7268, "step": 2689 }, { "epoch": 1.7210492642354447, "grad_norm": 2.1950254941449154, "learning_rate": 1.89595240903026e-05, "loss": 0.667, "step": 2690 }, { "epoch": 1.7216890595009597, "grad_norm": 2.107411204024598, "learning_rate": 1.8958603550628746e-05, "loss": 0.7529, "step": 2691 }, { "epoch": 1.7223288547664746, "grad_norm": 2.0486761899159847, "learning_rate": 1.8957682626288393e-05, "loss": 0.6829, "step": 2692 }, { "epoch": 1.7229686500319898, "grad_norm": 2.1685959863749082, "learning_rate": 1.8956761317321084e-05, "loss": 0.6209, "step": 2693 }, { "epoch": 1.7236084452975047, "grad_norm": 2.2993202334517306, "learning_rate": 1.8955839623766374e-05, "loss": 0.7168, "step": 2694 }, { "epoch": 1.72424824056302, "grad_norm": 2.349973886978011, "learning_rate": 1.895491754566384e-05, "loss": 0.7062, "step": 2695 }, { "epoch": 1.7248880358285348, "grad_norm": 2.4247271279313214, "learning_rate": 1.895399508305308e-05, "loss": 0.7503, "step": 2696 }, { "epoch": 1.72552783109405, "grad_norm": 2.109069210261524, "learning_rate": 1.89530722359737e-05, "loss": 0.6909, "step": 2697 }, { "epoch": 1.7261676263595649, "grad_norm": 2.0799929945505795, "learning_rate": 1.895214900446532e-05, "loss": 0.6982, "step": 2698 }, { "epoch": 1.72680742162508, "grad_norm": 2.4017425091025744, "learning_rate": 1.8951225388567587e-05, "loss": 0.747, "step": 2699 }, { "epoch": 1.727447216890595, "grad_norm": 2.090290140058114, "learning_rate": 1.8950301388320164e-05, "loss": 0.7396, "step": 2700 }, { "epoch": 1.72808701215611, "grad_norm": 2.288630131496329, "learning_rate": 1.8949377003762716e-05, "loss": 0.742, "step": 2701 }, { "epoch": 1.728726807421625, "grad_norm": 2.104591423703238, "learning_rate": 1.894845223493494e-05, "loss": 0.6464, "step": 2702 }, { "epoch": 1.7293666026871402, "grad_norm": 2.1278154834681393, "learning_rate": 1.8947527081876544e-05, "loss": 0.6612, "step": 2703 }, { "epoch": 1.7300063979526552, "grad_norm": 2.0840385793297607, "learning_rate": 1.894660154462725e-05, "loss": 0.7066, "step": 2704 }, { "epoch": 1.73064619321817, "grad_norm": 2.1561785473148873, "learning_rate": 1.89456756232268e-05, "loss": 0.6967, "step": 2705 }, { "epoch": 1.7312859884836853, "grad_norm": 2.1675478337759353, "learning_rate": 1.8944749317714956e-05, "loss": 0.7168, "step": 2706 }, { "epoch": 1.7319257837492001, "grad_norm": 2.1918954555497128, "learning_rate": 1.894382262813148e-05, "loss": 0.7124, "step": 2707 }, { "epoch": 1.7325655790147154, "grad_norm": 2.2116830883809313, "learning_rate": 1.8942895554516178e-05, "loss": 0.6765, "step": 2708 }, { "epoch": 1.7332053742802302, "grad_norm": 2.227207034691954, "learning_rate": 1.8941968096908848e-05, "loss": 0.6964, "step": 2709 }, { "epoch": 1.7338451695457455, "grad_norm": 2.2522598960015006, "learning_rate": 1.8941040255349313e-05, "loss": 0.7998, "step": 2710 }, { "epoch": 1.7344849648112604, "grad_norm": 2.049272566378981, "learning_rate": 1.8940112029877415e-05, "loss": 0.7048, "step": 2711 }, { "epoch": 1.7351247600767754, "grad_norm": 2.1642262437305226, "learning_rate": 1.8939183420533007e-05, "loss": 0.7271, "step": 2712 }, { "epoch": 1.7357645553422905, "grad_norm": 2.4726527786840924, "learning_rate": 1.8938254427355964e-05, "loss": 0.757, "step": 2713 }, { "epoch": 1.7364043506078055, "grad_norm": 2.246028946712368, "learning_rate": 1.8937325050386176e-05, "loss": 0.6807, "step": 2714 }, { "epoch": 1.7370441458733206, "grad_norm": 2.2147636438193024, "learning_rate": 1.893639528966355e-05, "loss": 0.6957, "step": 2715 }, { "epoch": 1.7376839411388356, "grad_norm": 2.245448005782251, "learning_rate": 1.893546514522801e-05, "loss": 0.6894, "step": 2716 }, { "epoch": 1.7383237364043507, "grad_norm": 2.209321795240257, "learning_rate": 1.893453461711949e-05, "loss": 0.7397, "step": 2717 }, { "epoch": 1.7389635316698655, "grad_norm": 1.938019292383386, "learning_rate": 1.8933603705377945e-05, "loss": 0.7119, "step": 2718 }, { "epoch": 1.7396033269353808, "grad_norm": 2.2841962487023917, "learning_rate": 1.893267241004335e-05, "loss": 0.7449, "step": 2719 }, { "epoch": 1.7402431222008956, "grad_norm": 2.224131320542056, "learning_rate": 1.893174073115569e-05, "loss": 0.7006, "step": 2720 }, { "epoch": 1.7408829174664109, "grad_norm": 2.4961277263572983, "learning_rate": 1.8930808668754972e-05, "loss": 0.6898, "step": 2721 }, { "epoch": 1.7415227127319257, "grad_norm": 2.042661754111418, "learning_rate": 1.8929876222881217e-05, "loss": 0.6886, "step": 2722 }, { "epoch": 1.7421625079974408, "grad_norm": 2.3695692949672384, "learning_rate": 1.8928943393574464e-05, "loss": 0.7158, "step": 2723 }, { "epoch": 1.7428023032629558, "grad_norm": 2.1306692460223617, "learning_rate": 1.892801018087476e-05, "loss": 0.7243, "step": 2724 }, { "epoch": 1.7434420985284709, "grad_norm": 2.324620550695687, "learning_rate": 1.8927076584822187e-05, "loss": 0.7555, "step": 2725 }, { "epoch": 1.744081893793986, "grad_norm": 2.2940575971618706, "learning_rate": 1.892614260545682e-05, "loss": 0.7394, "step": 2726 }, { "epoch": 1.744721689059501, "grad_norm": 2.164442702473583, "learning_rate": 1.892520824281877e-05, "loss": 0.7189, "step": 2727 }, { "epoch": 1.745361484325016, "grad_norm": 2.4686859177179774, "learning_rate": 1.8924273496948156e-05, "loss": 0.6768, "step": 2728 }, { "epoch": 1.7460012795905309, "grad_norm": 2.110039793242381, "learning_rate": 1.892333836788511e-05, "loss": 0.6914, "step": 2729 }, { "epoch": 1.7466410748560461, "grad_norm": 2.3822620052491463, "learning_rate": 1.892240285566979e-05, "loss": 0.7555, "step": 2730 }, { "epoch": 1.747280870121561, "grad_norm": 2.3360560502988084, "learning_rate": 1.8921466960342363e-05, "loss": 0.7056, "step": 2731 }, { "epoch": 1.7479206653870762, "grad_norm": 2.2165244967010316, "learning_rate": 1.8920530681943016e-05, "loss": 0.6924, "step": 2732 }, { "epoch": 1.748560460652591, "grad_norm": 2.169588063489331, "learning_rate": 1.8919594020511948e-05, "loss": 0.7628, "step": 2733 }, { "epoch": 1.7492002559181064, "grad_norm": 2.378444509564954, "learning_rate": 1.891865697608938e-05, "loss": 0.7052, "step": 2734 }, { "epoch": 1.7498400511836212, "grad_norm": 2.4162034421365086, "learning_rate": 1.8917719548715545e-05, "loss": 0.6854, "step": 2735 }, { "epoch": 1.7504798464491362, "grad_norm": 2.216022809468936, "learning_rate": 1.8916781738430695e-05, "loss": 0.6616, "step": 2736 }, { "epoch": 1.7511196417146513, "grad_norm": 2.1419328416805885, "learning_rate": 1.8915843545275102e-05, "loss": 0.7058, "step": 2737 }, { "epoch": 1.7517594369801663, "grad_norm": 2.0172232402526182, "learning_rate": 1.891490496928905e-05, "loss": 0.7129, "step": 2738 }, { "epoch": 1.7523992322456814, "grad_norm": 2.1432372087195684, "learning_rate": 1.8913966010512833e-05, "loss": 0.7118, "step": 2739 }, { "epoch": 1.7530390275111964, "grad_norm": 2.3218418643753695, "learning_rate": 1.8913026668986772e-05, "loss": 0.7049, "step": 2740 }, { "epoch": 1.7536788227767115, "grad_norm": 2.201872539480669, "learning_rate": 1.89120869447512e-05, "loss": 0.682, "step": 2741 }, { "epoch": 1.7543186180422263, "grad_norm": 2.1936461628994586, "learning_rate": 1.891114683784647e-05, "loss": 0.7007, "step": 2742 }, { "epoch": 1.7549584133077416, "grad_norm": 2.1373657837190345, "learning_rate": 1.8910206348312945e-05, "loss": 0.7423, "step": 2743 }, { "epoch": 1.7555982085732564, "grad_norm": 2.2584137326556393, "learning_rate": 1.8909265476191013e-05, "loss": 0.7155, "step": 2744 }, { "epoch": 1.7562380038387717, "grad_norm": 2.0857031953872007, "learning_rate": 1.8908324221521066e-05, "loss": 0.7266, "step": 2745 }, { "epoch": 1.7568777991042865, "grad_norm": 2.256050585161655, "learning_rate": 1.8907382584343526e-05, "loss": 0.7056, "step": 2746 }, { "epoch": 1.7575175943698018, "grad_norm": 2.0730537501835546, "learning_rate": 1.890644056469882e-05, "loss": 0.7254, "step": 2747 }, { "epoch": 1.7581573896353166, "grad_norm": 2.147147232424808, "learning_rate": 1.8905498162627404e-05, "loss": 0.6541, "step": 2748 }, { "epoch": 1.7587971849008317, "grad_norm": 2.0925632535274, "learning_rate": 1.8904555378169733e-05, "loss": 0.7703, "step": 2749 }, { "epoch": 1.7594369801663468, "grad_norm": 2.061510048679831, "learning_rate": 1.8903612211366297e-05, "loss": 0.7339, "step": 2750 }, { "epoch": 1.7600767754318618, "grad_norm": 2.4428330421153244, "learning_rate": 1.890266866225759e-05, "loss": 0.7039, "step": 2751 }, { "epoch": 1.7607165706973769, "grad_norm": 2.1400861400292412, "learning_rate": 1.8901724730884128e-05, "loss": 0.6914, "step": 2752 }, { "epoch": 1.761356365962892, "grad_norm": 2.0505328148098076, "learning_rate": 1.890078041728644e-05, "loss": 0.7524, "step": 2753 }, { "epoch": 1.761996161228407, "grad_norm": 1.9846622020959332, "learning_rate": 1.8899835721505074e-05, "loss": 0.7272, "step": 2754 }, { "epoch": 1.7626359564939218, "grad_norm": 2.0594137971400066, "learning_rate": 1.8898890643580596e-05, "loss": 0.6955, "step": 2755 }, { "epoch": 1.763275751759437, "grad_norm": 2.1239173487045844, "learning_rate": 1.8897945183553585e-05, "loss": 0.6977, "step": 2756 }, { "epoch": 1.763915547024952, "grad_norm": 2.017729017484144, "learning_rate": 1.8896999341464632e-05, "loss": 0.7043, "step": 2757 }, { "epoch": 1.7645553422904672, "grad_norm": 2.0618714931514166, "learning_rate": 1.8896053117354357e-05, "loss": 0.6769, "step": 2758 }, { "epoch": 1.765195137555982, "grad_norm": 2.2778025380441287, "learning_rate": 1.8895106511263383e-05, "loss": 0.6735, "step": 2759 }, { "epoch": 1.7658349328214973, "grad_norm": 2.000664395903638, "learning_rate": 1.889415952323236e-05, "loss": 0.6505, "step": 2760 }, { "epoch": 1.7664747280870121, "grad_norm": 2.1191153953803323, "learning_rate": 1.889321215330195e-05, "loss": 0.6859, "step": 2761 }, { "epoch": 1.7671145233525272, "grad_norm": 2.2436553865427404, "learning_rate": 1.8892264401512825e-05, "loss": 0.7492, "step": 2762 }, { "epoch": 1.7677543186180422, "grad_norm": 2.366918758374397, "learning_rate": 1.889131626790569e-05, "loss": 0.6909, "step": 2763 }, { "epoch": 1.7683941138835573, "grad_norm": 2.1277175764834166, "learning_rate": 1.889036775252125e-05, "loss": 0.6884, "step": 2764 }, { "epoch": 1.7690339091490723, "grad_norm": 2.1231825371384323, "learning_rate": 1.8889418855400232e-05, "loss": 0.6675, "step": 2765 }, { "epoch": 1.7696737044145874, "grad_norm": 2.1521380887766206, "learning_rate": 1.888846957658338e-05, "loss": 0.707, "step": 2766 }, { "epoch": 1.7703134996801024, "grad_norm": 2.224440719418938, "learning_rate": 1.888751991611146e-05, "loss": 0.7117, "step": 2767 }, { "epoch": 1.7709532949456173, "grad_norm": 2.176878191770557, "learning_rate": 1.888656987402524e-05, "loss": 0.6722, "step": 2768 }, { "epoch": 1.7715930902111325, "grad_norm": 2.250043121289396, "learning_rate": 1.888561945036552e-05, "loss": 0.646, "step": 2769 }, { "epoch": 1.7722328854766474, "grad_norm": 2.202503421952523, "learning_rate": 1.8884668645173105e-05, "loss": 0.6596, "step": 2770 }, { "epoch": 1.7728726807421626, "grad_norm": 2.1971065567125994, "learning_rate": 1.8883717458488828e-05, "loss": 0.7152, "step": 2771 }, { "epoch": 1.7735124760076775, "grad_norm": 2.306515272712011, "learning_rate": 1.888276589035352e-05, "loss": 0.7091, "step": 2772 }, { "epoch": 1.7741522712731925, "grad_norm": 2.2163214034480587, "learning_rate": 1.8881813940808048e-05, "loss": 0.6639, "step": 2773 }, { "epoch": 1.7747920665387076, "grad_norm": 2.0661693386784736, "learning_rate": 1.8880861609893288e-05, "loss": 0.6957, "step": 2774 }, { "epoch": 1.7754318618042226, "grad_norm": 2.2325014818994426, "learning_rate": 1.8879908897650126e-05, "loss": 0.7242, "step": 2775 }, { "epoch": 1.7760716570697377, "grad_norm": 2.163769004876767, "learning_rate": 1.887895580411947e-05, "loss": 0.6812, "step": 2776 }, { "epoch": 1.7767114523352527, "grad_norm": 2.002633469887497, "learning_rate": 1.887800232934225e-05, "loss": 0.6659, "step": 2777 }, { "epoch": 1.7773512476007678, "grad_norm": 2.3358597839860598, "learning_rate": 1.88770484733594e-05, "loss": 0.6581, "step": 2778 }, { "epoch": 1.7779910428662828, "grad_norm": 2.2546025484543, "learning_rate": 1.8876094236211882e-05, "loss": 0.7275, "step": 2779 }, { "epoch": 1.778630838131798, "grad_norm": 2.251586411194137, "learning_rate": 1.8875139617940664e-05, "loss": 0.6618, "step": 2780 }, { "epoch": 1.7792706333973127, "grad_norm": 2.257413838213511, "learning_rate": 1.887418461858674e-05, "loss": 0.7756, "step": 2781 }, { "epoch": 1.779910428662828, "grad_norm": 2.2301229648280887, "learning_rate": 1.8873229238191116e-05, "loss": 0.6933, "step": 2782 }, { "epoch": 1.7805502239283428, "grad_norm": 2.0645493754632076, "learning_rate": 1.887227347679481e-05, "loss": 0.6862, "step": 2783 }, { "epoch": 1.781190019193858, "grad_norm": 2.2989802374581867, "learning_rate": 1.8871317334438863e-05, "loss": 0.71, "step": 2784 }, { "epoch": 1.781829814459373, "grad_norm": 2.3490869162747603, "learning_rate": 1.8870360811164333e-05, "loss": 0.6713, "step": 2785 }, { "epoch": 1.782469609724888, "grad_norm": 2.192692579967415, "learning_rate": 1.886940390701229e-05, "loss": 0.7057, "step": 2786 }, { "epoch": 1.783109404990403, "grad_norm": 2.1198861296767495, "learning_rate": 1.8868446622023818e-05, "loss": 0.6881, "step": 2787 }, { "epoch": 1.783749200255918, "grad_norm": 2.152876566860982, "learning_rate": 1.8867488956240023e-05, "loss": 0.7557, "step": 2788 }, { "epoch": 1.7843889955214332, "grad_norm": 2.0728035063755867, "learning_rate": 1.8866530909702027e-05, "loss": 0.6895, "step": 2789 }, { "epoch": 1.7850287907869482, "grad_norm": 2.2960797073148727, "learning_rate": 1.8865572482450967e-05, "loss": 0.6255, "step": 2790 }, { "epoch": 1.7856685860524633, "grad_norm": 2.196450018266707, "learning_rate": 1.8864613674527997e-05, "loss": 0.6802, "step": 2791 }, { "epoch": 1.786308381317978, "grad_norm": 2.2910246285382767, "learning_rate": 1.8863654485974282e-05, "loss": 0.6569, "step": 2792 }, { "epoch": 1.7869481765834934, "grad_norm": 2.109641896553454, "learning_rate": 1.8862694916831012e-05, "loss": 0.6918, "step": 2793 }, { "epoch": 1.7875879718490082, "grad_norm": 2.0699793904505714, "learning_rate": 1.8861734967139386e-05, "loss": 0.6616, "step": 2794 }, { "epoch": 1.7882277671145235, "grad_norm": 2.4392740734837735, "learning_rate": 1.8860774636940627e-05, "loss": 0.69, "step": 2795 }, { "epoch": 1.7888675623800383, "grad_norm": 2.3883964956409196, "learning_rate": 1.8859813926275963e-05, "loss": 0.7261, "step": 2796 }, { "epoch": 1.7895073576455536, "grad_norm": 2.050922498232111, "learning_rate": 1.885885283518665e-05, "loss": 0.6886, "step": 2797 }, { "epoch": 1.7901471529110684, "grad_norm": 2.3750811589271605, "learning_rate": 1.8857891363713962e-05, "loss": 0.7375, "step": 2798 }, { "epoch": 1.7907869481765835, "grad_norm": 2.2328543953795568, "learning_rate": 1.885692951189917e-05, "loss": 0.7416, "step": 2799 }, { "epoch": 1.7914267434420985, "grad_norm": 2.1177088898631684, "learning_rate": 1.8855967279783578e-05, "loss": 0.6687, "step": 2800 }, { "epoch": 1.7920665387076136, "grad_norm": 2.1306535830940634, "learning_rate": 1.885500466740851e-05, "loss": 0.7003, "step": 2801 }, { "epoch": 1.7927063339731286, "grad_norm": 2.1530192502666465, "learning_rate": 1.8854041674815287e-05, "loss": 0.7007, "step": 2802 }, { "epoch": 1.7933461292386437, "grad_norm": 2.0857900118255643, "learning_rate": 1.885307830204527e-05, "loss": 0.7023, "step": 2803 }, { "epoch": 1.7939859245041587, "grad_norm": 2.13356725915473, "learning_rate": 1.885211454913982e-05, "loss": 0.6954, "step": 2804 }, { "epoch": 1.7946257197696736, "grad_norm": 1.97610585554252, "learning_rate": 1.8851150416140314e-05, "loss": 0.619, "step": 2805 }, { "epoch": 1.7952655150351888, "grad_norm": 2.3580190316938285, "learning_rate": 1.8850185903088155e-05, "loss": 0.6332, "step": 2806 }, { "epoch": 1.7959053103007037, "grad_norm": 2.1088400306903754, "learning_rate": 1.8849221010024757e-05, "loss": 0.6994, "step": 2807 }, { "epoch": 1.796545105566219, "grad_norm": 1.9557691752177155, "learning_rate": 1.884825573699155e-05, "loss": 0.6754, "step": 2808 }, { "epoch": 1.7971849008317338, "grad_norm": 2.1314736702312205, "learning_rate": 1.8847290084029982e-05, "loss": 0.6316, "step": 2809 }, { "epoch": 1.797824696097249, "grad_norm": 2.3007262323442923, "learning_rate": 1.8846324051181514e-05, "loss": 0.7311, "step": 2810 }, { "epoch": 1.7984644913627639, "grad_norm": 2.2479284471836762, "learning_rate": 1.884535763848763e-05, "loss": 0.6944, "step": 2811 }, { "epoch": 1.799104286628279, "grad_norm": 2.1394009833193897, "learning_rate": 1.8844390845989824e-05, "loss": 0.6625, "step": 2812 }, { "epoch": 1.799744081893794, "grad_norm": 2.1142021732674077, "learning_rate": 1.884342367372961e-05, "loss": 0.6503, "step": 2813 }, { "epoch": 1.800383877159309, "grad_norm": 2.490220562357213, "learning_rate": 1.884245612174851e-05, "loss": 0.7351, "step": 2814 }, { "epoch": 1.801023672424824, "grad_norm": 1.8895655603406811, "learning_rate": 1.8841488190088074e-05, "loss": 0.71, "step": 2815 }, { "epoch": 1.8016634676903391, "grad_norm": 1.986229274168058, "learning_rate": 1.8840519878789868e-05, "loss": 0.6039, "step": 2816 }, { "epoch": 1.8023032629558542, "grad_norm": 2.238940756628011, "learning_rate": 1.8839551187895456e-05, "loss": 0.716, "step": 2817 }, { "epoch": 1.802943058221369, "grad_norm": 2.2514394827675264, "learning_rate": 1.8838582117446448e-05, "loss": 0.6897, "step": 2818 }, { "epoch": 1.8035828534868843, "grad_norm": 2.2046981600553823, "learning_rate": 1.8837612667484443e-05, "loss": 0.738, "step": 2819 }, { "epoch": 1.8042226487523991, "grad_norm": 2.5013069353019546, "learning_rate": 1.883664283805107e-05, "loss": 0.6688, "step": 2820 }, { "epoch": 1.8048624440179144, "grad_norm": 1.9899786650328168, "learning_rate": 1.883567262918798e-05, "loss": 0.7008, "step": 2821 }, { "epoch": 1.8055022392834292, "grad_norm": 1.9731410195210644, "learning_rate": 1.8834702040936818e-05, "loss": 0.7518, "step": 2822 }, { "epoch": 1.8061420345489443, "grad_norm": 2.1332301540493064, "learning_rate": 1.8833731073339264e-05, "loss": 0.6971, "step": 2823 }, { "epoch": 1.8067818298144593, "grad_norm": 2.392818381652132, "learning_rate": 1.8832759726437016e-05, "loss": 0.6516, "step": 2824 }, { "epoch": 1.8074216250799744, "grad_norm": 2.0895344370653723, "learning_rate": 1.8831788000271774e-05, "loss": 0.6845, "step": 2825 }, { "epoch": 1.8080614203454894, "grad_norm": 2.43892437182762, "learning_rate": 1.8830815894885268e-05, "loss": 0.654, "step": 2826 }, { "epoch": 1.8087012156110045, "grad_norm": 2.1953087852651927, "learning_rate": 1.8829843410319238e-05, "loss": 0.7245, "step": 2827 }, { "epoch": 1.8093410108765196, "grad_norm": 2.1187627895414005, "learning_rate": 1.8828870546615436e-05, "loss": 0.6356, "step": 2828 }, { "epoch": 1.8099808061420346, "grad_norm": 2.2043208180264005, "learning_rate": 1.8827897303815643e-05, "loss": 0.6901, "step": 2829 }, { "epoch": 1.8106206014075497, "grad_norm": 2.3438843077164857, "learning_rate": 1.882692368196164e-05, "loss": 0.6504, "step": 2830 }, { "epoch": 1.8112603966730645, "grad_norm": 2.412341316517786, "learning_rate": 1.8825949681095235e-05, "loss": 0.6693, "step": 2831 }, { "epoch": 1.8119001919385798, "grad_norm": 2.3376281197010886, "learning_rate": 1.882497530125825e-05, "loss": 0.6835, "step": 2832 }, { "epoch": 1.8125399872040946, "grad_norm": 2.4483940380421063, "learning_rate": 1.8824000542492525e-05, "loss": 0.7502, "step": 2833 }, { "epoch": 1.8131797824696099, "grad_norm": 2.1141331891584043, "learning_rate": 1.8823025404839916e-05, "loss": 0.6815, "step": 2834 }, { "epoch": 1.8138195777351247, "grad_norm": 2.3401486721735165, "learning_rate": 1.882204988834229e-05, "loss": 0.6867, "step": 2835 }, { "epoch": 1.8144593730006398, "grad_norm": 2.0761527641162143, "learning_rate": 1.8821073993041534e-05, "loss": 0.6745, "step": 2836 }, { "epoch": 1.8150991682661548, "grad_norm": 2.495573795948765, "learning_rate": 1.882009771897955e-05, "loss": 0.6706, "step": 2837 }, { "epoch": 1.8157389635316699, "grad_norm": 2.0823265175084673, "learning_rate": 1.8819121066198262e-05, "loss": 0.6429, "step": 2838 }, { "epoch": 1.816378758797185, "grad_norm": 2.242686756642798, "learning_rate": 1.8818144034739605e-05, "loss": 0.6554, "step": 2839 }, { "epoch": 1.8170185540627, "grad_norm": 2.35521574084857, "learning_rate": 1.8817166624645528e-05, "loss": 0.7529, "step": 2840 }, { "epoch": 1.817658349328215, "grad_norm": 2.105173591841957, "learning_rate": 1.8816188835958003e-05, "loss": 0.7455, "step": 2841 }, { "epoch": 1.8182981445937298, "grad_norm": 2.1531553609011334, "learning_rate": 1.881521066871901e-05, "loss": 0.648, "step": 2842 }, { "epoch": 1.8189379398592451, "grad_norm": 2.0869245428352023, "learning_rate": 1.881423212297055e-05, "loss": 0.6713, "step": 2843 }, { "epoch": 1.81957773512476, "grad_norm": 2.266993760825072, "learning_rate": 1.8813253198754644e-05, "loss": 0.7192, "step": 2844 }, { "epoch": 1.8202175303902752, "grad_norm": 2.344519756308565, "learning_rate": 1.881227389611332e-05, "loss": 0.6333, "step": 2845 }, { "epoch": 1.82085732565579, "grad_norm": 2.265155235576091, "learning_rate": 1.8811294215088636e-05, "loss": 0.6783, "step": 2846 }, { "epoch": 1.8214971209213053, "grad_norm": 2.197462530087242, "learning_rate": 1.881031415572265e-05, "loss": 0.7218, "step": 2847 }, { "epoch": 1.8221369161868202, "grad_norm": 2.2531010423182516, "learning_rate": 1.8809333718057446e-05, "loss": 0.684, "step": 2848 }, { "epoch": 1.8227767114523352, "grad_norm": 2.4892459295050493, "learning_rate": 1.8808352902135124e-05, "loss": 0.6879, "step": 2849 }, { "epoch": 1.8234165067178503, "grad_norm": 2.346447926633639, "learning_rate": 1.8807371707997798e-05, "loss": 0.7009, "step": 2850 }, { "epoch": 1.8240563019833653, "grad_norm": 2.033897129340694, "learning_rate": 1.8806390135687596e-05, "loss": 0.6792, "step": 2851 }, { "epoch": 1.8246960972488804, "grad_norm": 2.702697024788221, "learning_rate": 1.8805408185246668e-05, "loss": 0.7222, "step": 2852 }, { "epoch": 1.8253358925143954, "grad_norm": 2.3307442322755176, "learning_rate": 1.8804425856717174e-05, "loss": 0.7355, "step": 2853 }, { "epoch": 1.8259756877799105, "grad_norm": 2.2193214734228293, "learning_rate": 1.8803443150141297e-05, "loss": 0.6453, "step": 2854 }, { "epoch": 1.8266154830454253, "grad_norm": 2.0374584001325733, "learning_rate": 1.8802460065561233e-05, "loss": 0.7077, "step": 2855 }, { "epoch": 1.8272552783109406, "grad_norm": 2.087117079369351, "learning_rate": 1.880147660301919e-05, "loss": 0.746, "step": 2856 }, { "epoch": 1.8278950735764554, "grad_norm": 2.0370979587379114, "learning_rate": 1.8800492762557398e-05, "loss": 0.683, "step": 2857 }, { "epoch": 1.8285348688419707, "grad_norm": 2.1871828603129737, "learning_rate": 1.8799508544218103e-05, "loss": 0.7518, "step": 2858 }, { "epoch": 1.8291746641074855, "grad_norm": 2.148986175073362, "learning_rate": 1.8798523948043563e-05, "loss": 0.6835, "step": 2859 }, { "epoch": 1.8298144593730008, "grad_norm": 2.041812038036196, "learning_rate": 1.8797538974076056e-05, "loss": 0.6354, "step": 2860 }, { "epoch": 1.8304542546385156, "grad_norm": 2.15062603041324, "learning_rate": 1.879655362235788e-05, "loss": 0.6562, "step": 2861 }, { "epoch": 1.8310940499040307, "grad_norm": 2.247473769836607, "learning_rate": 1.879556789293133e-05, "loss": 0.7045, "step": 2862 }, { "epoch": 1.8317338451695457, "grad_norm": 2.412556740674889, "learning_rate": 1.8794581785838746e-05, "loss": 0.7003, "step": 2863 }, { "epoch": 1.8323736404350608, "grad_norm": 2.1657198703134557, "learning_rate": 1.8793595301122468e-05, "loss": 0.685, "step": 2864 }, { "epoch": 1.8330134357005758, "grad_norm": 2.251468617971432, "learning_rate": 1.8792608438824845e-05, "loss": 0.6961, "step": 2865 }, { "epoch": 1.833653230966091, "grad_norm": 2.2088162008175387, "learning_rate": 1.879162119898826e-05, "loss": 0.7152, "step": 2866 }, { "epoch": 1.834293026231606, "grad_norm": 2.578494561933005, "learning_rate": 1.8790633581655095e-05, "loss": 0.7077, "step": 2867 }, { "epoch": 1.8349328214971208, "grad_norm": 2.490513636376645, "learning_rate": 1.8789645586867764e-05, "loss": 0.7075, "step": 2868 }, { "epoch": 1.835572616762636, "grad_norm": 2.28795449143814, "learning_rate": 1.8788657214668687e-05, "loss": 0.6713, "step": 2869 }, { "epoch": 1.8362124120281509, "grad_norm": 2.093061822977666, "learning_rate": 1.8787668465100305e-05, "loss": 0.685, "step": 2870 }, { "epoch": 1.8368522072936662, "grad_norm": 2.181127312639882, "learning_rate": 1.878667933820507e-05, "loss": 0.6086, "step": 2871 }, { "epoch": 1.837492002559181, "grad_norm": 2.3844266279329362, "learning_rate": 1.8785689834025454e-05, "loss": 0.6169, "step": 2872 }, { "epoch": 1.838131797824696, "grad_norm": 2.3293801136689862, "learning_rate": 1.8784699952603943e-05, "loss": 0.7111, "step": 2873 }, { "epoch": 1.838771593090211, "grad_norm": 2.316972410177252, "learning_rate": 1.8783709693983044e-05, "loss": 0.6859, "step": 2874 }, { "epoch": 1.8394113883557262, "grad_norm": 2.1591675984399674, "learning_rate": 1.878271905820528e-05, "loss": 0.6702, "step": 2875 }, { "epoch": 1.8400511836212412, "grad_norm": 2.1160479788223503, "learning_rate": 1.878172804531318e-05, "loss": 0.6514, "step": 2876 }, { "epoch": 1.8406909788867563, "grad_norm": 2.319677910626881, "learning_rate": 1.8780736655349296e-05, "loss": 0.6608, "step": 2877 }, { "epoch": 1.8413307741522713, "grad_norm": 2.2620931328169505, "learning_rate": 1.87797448883562e-05, "loss": 0.7272, "step": 2878 }, { "epoch": 1.8419705694177864, "grad_norm": 1.9709621800435568, "learning_rate": 1.8778752744376483e-05, "loss": 0.6156, "step": 2879 }, { "epoch": 1.8426103646833014, "grad_norm": 2.052230102119442, "learning_rate": 1.8777760223452735e-05, "loss": 0.6756, "step": 2880 }, { "epoch": 1.8432501599488162, "grad_norm": 2.1792949465268183, "learning_rate": 1.877676732562758e-05, "loss": 0.6807, "step": 2881 }, { "epoch": 1.8438899552143315, "grad_norm": 1.9934267798194083, "learning_rate": 1.8775774050943644e-05, "loss": 0.6895, "step": 2882 }, { "epoch": 1.8445297504798464, "grad_norm": 2.1240121351667285, "learning_rate": 1.877478039944358e-05, "loss": 0.6695, "step": 2883 }, { "epoch": 1.8451695457453616, "grad_norm": 2.1683366127326535, "learning_rate": 1.8773786371170063e-05, "loss": 0.7394, "step": 2884 }, { "epoch": 1.8458093410108765, "grad_norm": 2.1334987365921076, "learning_rate": 1.877279196616576e-05, "loss": 0.6915, "step": 2885 }, { "epoch": 1.8464491362763915, "grad_norm": 2.4119981212683257, "learning_rate": 1.8771797184473378e-05, "loss": 0.7155, "step": 2886 }, { "epoch": 1.8470889315419066, "grad_norm": 2.03741209990302, "learning_rate": 1.8770802026135627e-05, "loss": 0.6673, "step": 2887 }, { "epoch": 1.8477287268074216, "grad_norm": 2.007200161953685, "learning_rate": 1.876980649119524e-05, "loss": 0.6689, "step": 2888 }, { "epoch": 1.8483685220729367, "grad_norm": 2.192660029092288, "learning_rate": 1.8768810579694966e-05, "loss": 0.6507, "step": 2889 }, { "epoch": 1.8490083173384517, "grad_norm": 2.2037744692220538, "learning_rate": 1.876781429167756e-05, "loss": 0.6318, "step": 2890 }, { "epoch": 1.8496481126039668, "grad_norm": 2.1751294630888185, "learning_rate": 1.8766817627185804e-05, "loss": 0.6448, "step": 2891 }, { "epoch": 1.8502879078694816, "grad_norm": 2.5398059877421777, "learning_rate": 1.8765820586262498e-05, "loss": 0.6674, "step": 2892 }, { "epoch": 1.8509277031349969, "grad_norm": 2.1821796898325863, "learning_rate": 1.8764823168950445e-05, "loss": 0.6094, "step": 2893 }, { "epoch": 1.8515674984005117, "grad_norm": 2.1664158587542373, "learning_rate": 1.876382537529248e-05, "loss": 0.6674, "step": 2894 }, { "epoch": 1.852207293666027, "grad_norm": 2.2369568777246176, "learning_rate": 1.8762827205331442e-05, "loss": 0.73, "step": 2895 }, { "epoch": 1.8528470889315418, "grad_norm": 2.2823036605051064, "learning_rate": 1.876182865911019e-05, "loss": 0.6026, "step": 2896 }, { "epoch": 1.853486884197057, "grad_norm": 2.090586096178826, "learning_rate": 1.8760829736671603e-05, "loss": 0.6661, "step": 2897 }, { "epoch": 1.854126679462572, "grad_norm": 2.074196384404059, "learning_rate": 1.875983043805857e-05, "loss": 0.6446, "step": 2898 }, { "epoch": 1.854766474728087, "grad_norm": 2.0750588173562003, "learning_rate": 1.8758830763314006e-05, "loss": 0.727, "step": 2899 }, { "epoch": 1.855406269993602, "grad_norm": 2.1902334978714637, "learning_rate": 1.8757830712480826e-05, "loss": 0.6518, "step": 2900 }, { "epoch": 1.856046065259117, "grad_norm": 2.2445823318126923, "learning_rate": 1.875683028560197e-05, "loss": 0.712, "step": 2901 }, { "epoch": 1.8566858605246321, "grad_norm": 2.081578433140177, "learning_rate": 1.87558294827204e-05, "loss": 0.658, "step": 2902 }, { "epoch": 1.8573256557901472, "grad_norm": 2.364653889148393, "learning_rate": 1.8754828303879093e-05, "loss": 0.7193, "step": 2903 }, { "epoch": 1.8579654510556622, "grad_norm": 1.988361028058181, "learning_rate": 1.875382674912103e-05, "loss": 0.616, "step": 2904 }, { "epoch": 1.858605246321177, "grad_norm": 2.0786679769558125, "learning_rate": 1.8752824818489218e-05, "loss": 0.7185, "step": 2905 }, { "epoch": 1.8592450415866923, "grad_norm": 2.325161779812975, "learning_rate": 1.875182251202668e-05, "loss": 0.6858, "step": 2906 }, { "epoch": 1.8598848368522072, "grad_norm": 2.113181951024051, "learning_rate": 1.8750819829776445e-05, "loss": 0.6815, "step": 2907 }, { "epoch": 1.8605246321177225, "grad_norm": 2.149635239100949, "learning_rate": 1.874981677178158e-05, "loss": 0.6816, "step": 2908 }, { "epoch": 1.8611644273832373, "grad_norm": 2.2177542710643308, "learning_rate": 1.8748813338085142e-05, "loss": 0.6695, "step": 2909 }, { "epoch": 1.8618042226487526, "grad_norm": 1.917861149803569, "learning_rate": 1.874780952873023e-05, "loss": 0.6822, "step": 2910 }, { "epoch": 1.8624440179142674, "grad_norm": 2.0456702383067604, "learning_rate": 1.874680534375993e-05, "loss": 0.5799, "step": 2911 }, { "epoch": 1.8630838131797824, "grad_norm": 2.273677474373349, "learning_rate": 1.8745800783217373e-05, "loss": 0.6861, "step": 2912 }, { "epoch": 1.8637236084452975, "grad_norm": 2.2463766445844597, "learning_rate": 1.8744795847145684e-05, "loss": 0.6716, "step": 2913 }, { "epoch": 1.8643634037108126, "grad_norm": 2.197991154731108, "learning_rate": 1.8743790535588017e-05, "loss": 0.6803, "step": 2914 }, { "epoch": 1.8650031989763276, "grad_norm": 2.2835614278240364, "learning_rate": 1.8742784848587543e-05, "loss": 0.7266, "step": 2915 }, { "epoch": 1.8656429942418427, "grad_norm": 2.1502689372729207, "learning_rate": 1.8741778786187437e-05, "loss": 0.7073, "step": 2916 }, { "epoch": 1.8662827895073577, "grad_norm": 2.221673950569069, "learning_rate": 1.87407723484309e-05, "loss": 0.6144, "step": 2917 }, { "epoch": 1.8669225847728725, "grad_norm": 2.5386928629481056, "learning_rate": 1.8739765535361147e-05, "loss": 0.5982, "step": 2918 }, { "epoch": 1.8675623800383878, "grad_norm": 2.7379564580733597, "learning_rate": 1.8738758347021412e-05, "loss": 0.6503, "step": 2919 }, { "epoch": 1.8682021753039026, "grad_norm": 2.1722549039161327, "learning_rate": 1.8737750783454934e-05, "loss": 0.7517, "step": 2920 }, { "epoch": 1.868841970569418, "grad_norm": 2.2418227518090577, "learning_rate": 1.8736742844704984e-05, "loss": 0.5924, "step": 2921 }, { "epoch": 1.8694817658349328, "grad_norm": 2.2890483390404173, "learning_rate": 1.8735734530814837e-05, "loss": 0.6428, "step": 2922 }, { "epoch": 1.8701215611004478, "grad_norm": 2.750802781435516, "learning_rate": 1.8734725841827788e-05, "loss": 0.6007, "step": 2923 }, { "epoch": 1.8707613563659629, "grad_norm": 2.4995057630320265, "learning_rate": 1.873371677778715e-05, "loss": 0.6956, "step": 2924 }, { "epoch": 1.871401151631478, "grad_norm": 2.340695757911133, "learning_rate": 1.8732707338736248e-05, "loss": 0.6885, "step": 2925 }, { "epoch": 1.872040946896993, "grad_norm": 2.2857111123018448, "learning_rate": 1.873169752471843e-05, "loss": 0.69, "step": 2926 }, { "epoch": 1.872680742162508, "grad_norm": 3.1109949806382313, "learning_rate": 1.8730687335777048e-05, "loss": 0.6653, "step": 2927 }, { "epoch": 1.873320537428023, "grad_norm": 2.4106659701670154, "learning_rate": 1.8729676771955486e-05, "loss": 0.6682, "step": 2928 }, { "epoch": 1.8739603326935381, "grad_norm": 2.5249265165715875, "learning_rate": 1.8728665833297132e-05, "loss": 0.6357, "step": 2929 }, { "epoch": 1.8746001279590532, "grad_norm": 2.43192669443945, "learning_rate": 1.8727654519845395e-05, "loss": 0.6298, "step": 2930 }, { "epoch": 1.875239923224568, "grad_norm": 2.4462660553458337, "learning_rate": 1.8726642831643696e-05, "loss": 0.6823, "step": 2931 }, { "epoch": 1.8758797184900833, "grad_norm": 2.159134654129459, "learning_rate": 1.8725630768735483e-05, "loss": 0.6175, "step": 2932 }, { "epoch": 1.8765195137555981, "grad_norm": 2.4300472008143634, "learning_rate": 1.87246183311642e-05, "loss": 0.7224, "step": 2933 }, { "epoch": 1.8771593090211134, "grad_norm": 2.313914732221962, "learning_rate": 1.8723605518973328e-05, "loss": 0.6602, "step": 2934 }, { "epoch": 1.8777991042866282, "grad_norm": 2.0931136950928817, "learning_rate": 1.872259233220635e-05, "loss": 0.6473, "step": 2935 }, { "epoch": 1.8784388995521433, "grad_norm": 2.050178663824379, "learning_rate": 1.8721578770906776e-05, "loss": 0.6025, "step": 2936 }, { "epoch": 1.8790786948176583, "grad_norm": 2.060886999633449, "learning_rate": 1.8720564835118128e-05, "loss": 0.655, "step": 2937 }, { "epoch": 1.8797184900831734, "grad_norm": 2.4185926180663606, "learning_rate": 1.8719550524883935e-05, "loss": 0.7006, "step": 2938 }, { "epoch": 1.8803582853486884, "grad_norm": 2.1638263621592504, "learning_rate": 1.8718535840247752e-05, "loss": 0.5849, "step": 2939 }, { "epoch": 1.8809980806142035, "grad_norm": 2.1916776777931184, "learning_rate": 1.871752078125315e-05, "loss": 0.6955, "step": 2940 }, { "epoch": 1.8816378758797185, "grad_norm": 2.200804631611168, "learning_rate": 1.8716505347943713e-05, "loss": 0.6402, "step": 2941 }, { "epoch": 1.8822776711452334, "grad_norm": 2.2601442106813914, "learning_rate": 1.8715489540363043e-05, "loss": 0.6912, "step": 2942 }, { "epoch": 1.8829174664107486, "grad_norm": 2.0524323560267614, "learning_rate": 1.8714473358554753e-05, "loss": 0.6576, "step": 2943 }, { "epoch": 1.8835572616762635, "grad_norm": 2.079432224501133, "learning_rate": 1.8713456802562484e-05, "loss": 0.7262, "step": 2944 }, { "epoch": 1.8841970569417787, "grad_norm": 2.336443438647054, "learning_rate": 1.8712439872429876e-05, "loss": 0.7111, "step": 2945 }, { "epoch": 1.8848368522072936, "grad_norm": 2.0521090839982743, "learning_rate": 1.87114225682006e-05, "loss": 0.6636, "step": 2946 }, { "epoch": 1.8854766474728089, "grad_norm": 1.9765722609162115, "learning_rate": 1.8710404889918334e-05, "loss": 0.6126, "step": 2947 }, { "epoch": 1.8861164427383237, "grad_norm": 2.1473674659577333, "learning_rate": 1.8709386837626778e-05, "loss": 0.6155, "step": 2948 }, { "epoch": 1.8867562380038387, "grad_norm": 2.146302653571163, "learning_rate": 1.8708368411369643e-05, "loss": 0.6687, "step": 2949 }, { "epoch": 1.8873960332693538, "grad_norm": 2.190710568115309, "learning_rate": 1.870734961119066e-05, "loss": 0.6851, "step": 2950 }, { "epoch": 1.8880358285348688, "grad_norm": 2.032148282799541, "learning_rate": 1.870633043713357e-05, "loss": 0.6071, "step": 2951 }, { "epoch": 1.888675623800384, "grad_norm": 2.3792900664388412, "learning_rate": 1.8705310889242143e-05, "loss": 0.6912, "step": 2952 }, { "epoch": 1.889315419065899, "grad_norm": 2.454080391913765, "learning_rate": 1.870429096756015e-05, "loss": 0.6635, "step": 2953 }, { "epoch": 1.889955214331414, "grad_norm": 2.263220554419675, "learning_rate": 1.870327067213139e-05, "loss": 0.7141, "step": 2954 }, { "epoch": 1.8905950095969288, "grad_norm": 2.08738770871088, "learning_rate": 1.8702250002999664e-05, "loss": 0.6242, "step": 2955 }, { "epoch": 1.891234804862444, "grad_norm": 2.0300385534121688, "learning_rate": 1.8701228960208807e-05, "loss": 0.6146, "step": 2956 }, { "epoch": 1.891874600127959, "grad_norm": 2.090243423438904, "learning_rate": 1.8700207543802654e-05, "loss": 0.6725, "step": 2957 }, { "epoch": 1.8925143953934742, "grad_norm": 2.0526821292642543, "learning_rate": 1.8699185753825067e-05, "loss": 0.6215, "step": 2958 }, { "epoch": 1.893154190658989, "grad_norm": 2.333538279304294, "learning_rate": 1.8698163590319916e-05, "loss": 0.6914, "step": 2959 }, { "epoch": 1.8937939859245043, "grad_norm": 2.1486224704438333, "learning_rate": 1.8697141053331095e-05, "loss": 0.6611, "step": 2960 }, { "epoch": 1.8944337811900192, "grad_norm": 2.0332860131167845, "learning_rate": 1.869611814290251e-05, "loss": 0.5809, "step": 2961 }, { "epoch": 1.8950735764555342, "grad_norm": 2.0156463740165815, "learning_rate": 1.869509485907808e-05, "loss": 0.6456, "step": 2962 }, { "epoch": 1.8957133717210493, "grad_norm": 2.2712840425630674, "learning_rate": 1.8694071201901743e-05, "loss": 0.6906, "step": 2963 }, { "epoch": 1.8963531669865643, "grad_norm": 2.1495726292966624, "learning_rate": 1.8693047171417456e-05, "loss": 0.6969, "step": 2964 }, { "epoch": 1.8969929622520794, "grad_norm": 2.0314038842974336, "learning_rate": 1.8692022767669186e-05, "loss": 0.6081, "step": 2965 }, { "epoch": 1.8976327575175944, "grad_norm": 1.9731191895677889, "learning_rate": 1.869099799070092e-05, "loss": 0.6806, "step": 2966 }, { "epoch": 1.8982725527831095, "grad_norm": 2.210816088914468, "learning_rate": 1.868997284055666e-05, "loss": 0.6539, "step": 2967 }, { "epoch": 1.8989123480486243, "grad_norm": 2.0613328889927125, "learning_rate": 1.8688947317280428e-05, "loss": 0.6209, "step": 2968 }, { "epoch": 1.8995521433141396, "grad_norm": 2.170843838670733, "learning_rate": 1.868792142091625e-05, "loss": 0.6738, "step": 2969 }, { "epoch": 1.9001919385796544, "grad_norm": 2.3735751504962135, "learning_rate": 1.8686895151508186e-05, "loss": 0.6323, "step": 2970 }, { "epoch": 1.9008317338451697, "grad_norm": 2.051503214415855, "learning_rate": 1.868586850910029e-05, "loss": 0.6538, "step": 2971 }, { "epoch": 1.9014715291106845, "grad_norm": 2.215820148448397, "learning_rate": 1.8684841493736657e-05, "loss": 0.6727, "step": 2972 }, { "epoch": 1.9021113243761996, "grad_norm": 2.3098708726396064, "learning_rate": 1.8683814105461376e-05, "loss": 0.6908, "step": 2973 }, { "epoch": 1.9027511196417146, "grad_norm": 2.438306318637532, "learning_rate": 1.8682786344318566e-05, "loss": 0.6399, "step": 2974 }, { "epoch": 1.9033909149072297, "grad_norm": 2.090618557138416, "learning_rate": 1.8681758210352355e-05, "loss": 0.627, "step": 2975 }, { "epoch": 1.9040307101727447, "grad_norm": 2.1800515271904497, "learning_rate": 1.868072970360689e-05, "loss": 0.6772, "step": 2976 }, { "epoch": 1.9046705054382598, "grad_norm": 2.2467152748010952, "learning_rate": 1.8679700824126334e-05, "loss": 0.6368, "step": 2977 }, { "epoch": 1.9053103007037748, "grad_norm": 2.1193600476384207, "learning_rate": 1.8678671571954866e-05, "loss": 0.6829, "step": 2978 }, { "epoch": 1.9059500959692899, "grad_norm": 2.029505165512711, "learning_rate": 1.8677641947136678e-05, "loss": 0.6987, "step": 2979 }, { "epoch": 1.906589891234805, "grad_norm": 2.2674385169420708, "learning_rate": 1.867661194971598e-05, "loss": 0.6304, "step": 2980 }, { "epoch": 1.9072296865003198, "grad_norm": 2.011116466537542, "learning_rate": 1.8675581579737e-05, "loss": 0.6344, "step": 2981 }, { "epoch": 1.907869481765835, "grad_norm": 1.9407180592933113, "learning_rate": 1.867455083724398e-05, "loss": 0.6639, "step": 2982 }, { "epoch": 1.9085092770313499, "grad_norm": 2.2712319877353417, "learning_rate": 1.8673519722281178e-05, "loss": 0.6964, "step": 2983 }, { "epoch": 1.9091490722968651, "grad_norm": 1.9763372746690147, "learning_rate": 1.8672488234892867e-05, "loss": 0.6381, "step": 2984 }, { "epoch": 1.90978886756238, "grad_norm": 2.407272023409416, "learning_rate": 1.867145637512334e-05, "loss": 0.7, "step": 2985 }, { "epoch": 1.910428662827895, "grad_norm": 2.5340654444192046, "learning_rate": 1.86704241430169e-05, "loss": 0.671, "step": 2986 }, { "epoch": 1.91106845809341, "grad_norm": 2.1197975646963503, "learning_rate": 1.866939153861787e-05, "loss": 0.6852, "step": 2987 }, { "epoch": 1.9117082533589251, "grad_norm": 2.103301988792549, "learning_rate": 1.8668358561970593e-05, "loss": 0.6199, "step": 2988 }, { "epoch": 1.9123480486244402, "grad_norm": 2.367843847015376, "learning_rate": 1.8667325213119417e-05, "loss": 0.7141, "step": 2989 }, { "epoch": 1.9129878438899552, "grad_norm": 2.2616908068476347, "learning_rate": 1.8666291492108714e-05, "loss": 0.6734, "step": 2990 }, { "epoch": 1.9136276391554703, "grad_norm": 2.173973533095629, "learning_rate": 1.866525739898287e-05, "loss": 0.6517, "step": 2991 }, { "epoch": 1.9142674344209851, "grad_norm": 2.267698004925005, "learning_rate": 1.8664222933786285e-05, "loss": 0.6206, "step": 2992 }, { "epoch": 1.9149072296865004, "grad_norm": 2.0135288741317243, "learning_rate": 1.8663188096563384e-05, "loss": 0.6512, "step": 2993 }, { "epoch": 1.9155470249520152, "grad_norm": 1.9722602815201966, "learning_rate": 1.8662152887358594e-05, "loss": 0.6282, "step": 2994 }, { "epoch": 1.9161868202175305, "grad_norm": 1.8949395427984903, "learning_rate": 1.8661117306216373e-05, "loss": 0.5875, "step": 2995 }, { "epoch": 1.9168266154830453, "grad_norm": 2.824239939120822, "learning_rate": 1.8660081353181178e-05, "loss": 0.7041, "step": 2996 }, { "epoch": 1.9174664107485606, "grad_norm": 2.198881397320182, "learning_rate": 1.8659045028297492e-05, "loss": 0.6506, "step": 2997 }, { "epoch": 1.9181062060140754, "grad_norm": 2.184079771071433, "learning_rate": 1.865800833160982e-05, "loss": 0.6612, "step": 2998 }, { "epoch": 1.9187460012795905, "grad_norm": 2.1589945461947675, "learning_rate": 1.865697126316267e-05, "loss": 0.6447, "step": 2999 }, { "epoch": 1.9193857965451055, "grad_norm": 2.17102007575111, "learning_rate": 1.8655933823000572e-05, "loss": 0.6234, "step": 3000 }, { "epoch": 1.9200255918106206, "grad_norm": 2.334921824398694, "learning_rate": 1.8654896011168074e-05, "loss": 0.6364, "step": 3001 }, { "epoch": 1.9206653870761357, "grad_norm": 2.4836557747500767, "learning_rate": 1.865385782770974e-05, "loss": 0.6637, "step": 3002 }, { "epoch": 1.9213051823416507, "grad_norm": 2.141952005432971, "learning_rate": 1.8652819272670145e-05, "loss": 0.6586, "step": 3003 }, { "epoch": 1.9219449776071658, "grad_norm": 2.1244762329465643, "learning_rate": 1.8651780346093876e-05, "loss": 0.6354, "step": 3004 }, { "epoch": 1.9225847728726806, "grad_norm": 2.076368232066392, "learning_rate": 1.8650741048025555e-05, "loss": 0.6539, "step": 3005 }, { "epoch": 1.9232245681381959, "grad_norm": 2.429766869817612, "learning_rate": 1.8649701378509803e-05, "loss": 0.6762, "step": 3006 }, { "epoch": 1.9238643634037107, "grad_norm": 2.5262464087679546, "learning_rate": 1.864866133759126e-05, "loss": 0.6469, "step": 3007 }, { "epoch": 1.924504158669226, "grad_norm": 2.184953988846691, "learning_rate": 1.864762092531458e-05, "loss": 0.657, "step": 3008 }, { "epoch": 1.9251439539347408, "grad_norm": 2.1173539730769577, "learning_rate": 1.8646580141724442e-05, "loss": 0.6491, "step": 3009 }, { "epoch": 1.925783749200256, "grad_norm": 2.1199074842569465, "learning_rate": 1.8645538986865534e-05, "loss": 0.641, "step": 3010 }, { "epoch": 1.926423544465771, "grad_norm": 2.1076008499672994, "learning_rate": 1.864449746078256e-05, "loss": 0.6275, "step": 3011 }, { "epoch": 1.927063339731286, "grad_norm": 2.7197401831159875, "learning_rate": 1.8643455563520244e-05, "loss": 0.7247, "step": 3012 }, { "epoch": 1.927703134996801, "grad_norm": 2.4434261098681502, "learning_rate": 1.8642413295123326e-05, "loss": 0.6814, "step": 3013 }, { "epoch": 1.928342930262316, "grad_norm": 2.276560160350236, "learning_rate": 1.8641370655636548e-05, "loss": 0.6321, "step": 3014 }, { "epoch": 1.9289827255278311, "grad_norm": 2.1492919070885477, "learning_rate": 1.864032764510469e-05, "loss": 0.6393, "step": 3015 }, { "epoch": 1.9296225207933462, "grad_norm": 2.070004497250405, "learning_rate": 1.863928426357253e-05, "loss": 0.636, "step": 3016 }, { "epoch": 1.9302623160588612, "grad_norm": 2.0282984378694002, "learning_rate": 1.863824051108487e-05, "loss": 0.6927, "step": 3017 }, { "epoch": 1.930902111324376, "grad_norm": 2.290967954154839, "learning_rate": 1.8637196387686534e-05, "loss": 0.6246, "step": 3018 }, { "epoch": 1.9315419065898913, "grad_norm": 2.257509878427812, "learning_rate": 1.863615189342235e-05, "loss": 0.6203, "step": 3019 }, { "epoch": 1.9321817018554062, "grad_norm": 2.497288925812385, "learning_rate": 1.863510702833716e-05, "loss": 0.6442, "step": 3020 }, { "epoch": 1.9328214971209214, "grad_norm": 2.1505419104183727, "learning_rate": 1.8634061792475838e-05, "loss": 0.6236, "step": 3021 }, { "epoch": 1.9334612923864363, "grad_norm": 2.298040163256418, "learning_rate": 1.863301618588326e-05, "loss": 0.6989, "step": 3022 }, { "epoch": 1.9341010876519513, "grad_norm": 1.9974531591332707, "learning_rate": 1.8631970208604325e-05, "loss": 0.6677, "step": 3023 }, { "epoch": 1.9347408829174664, "grad_norm": 2.1071124867790627, "learning_rate": 1.8630923860683948e-05, "loss": 0.6504, "step": 3024 }, { "epoch": 1.9353806781829814, "grad_norm": 2.3656219066307638, "learning_rate": 1.862987714216705e-05, "loss": 0.6452, "step": 3025 }, { "epoch": 1.9360204734484965, "grad_norm": 1.9776958012901995, "learning_rate": 1.8628830053098577e-05, "loss": 0.6347, "step": 3026 }, { "epoch": 1.9366602687140115, "grad_norm": 2.2858346679307253, "learning_rate": 1.8627782593523493e-05, "loss": 0.6835, "step": 3027 }, { "epoch": 1.9373000639795266, "grad_norm": 2.0571351944420675, "learning_rate": 1.8626734763486773e-05, "loss": 0.6271, "step": 3028 }, { "epoch": 1.9379398592450416, "grad_norm": 2.121774294462634, "learning_rate": 1.8625686563033407e-05, "loss": 0.6948, "step": 3029 }, { "epoch": 1.9385796545105567, "grad_norm": 2.315685883962223, "learning_rate": 1.8624637992208407e-05, "loss": 0.6835, "step": 3030 }, { "epoch": 1.9392194497760715, "grad_norm": 2.1250716671891055, "learning_rate": 1.862358905105679e-05, "loss": 0.6276, "step": 3031 }, { "epoch": 1.9398592450415868, "grad_norm": 2.1445648305433878, "learning_rate": 1.8622539739623596e-05, "loss": 0.6746, "step": 3032 }, { "epoch": 1.9404990403071016, "grad_norm": 2.3387915125756003, "learning_rate": 1.8621490057953893e-05, "loss": 0.6577, "step": 3033 }, { "epoch": 1.941138835572617, "grad_norm": 2.3129297081591393, "learning_rate": 1.8620440006092737e-05, "loss": 0.6469, "step": 3034 }, { "epoch": 1.9417786308381317, "grad_norm": 2.384378711136424, "learning_rate": 1.8619389584085223e-05, "loss": 0.649, "step": 3035 }, { "epoch": 1.9424184261036468, "grad_norm": 2.1081705144817127, "learning_rate": 1.8618338791976454e-05, "loss": 0.6692, "step": 3036 }, { "epoch": 1.9430582213691618, "grad_norm": 2.1851589190994782, "learning_rate": 1.8617287629811548e-05, "loss": 0.6561, "step": 3037 }, { "epoch": 1.943698016634677, "grad_norm": 2.120440977040503, "learning_rate": 1.8616236097635636e-05, "loss": 0.6181, "step": 3038 }, { "epoch": 1.944337811900192, "grad_norm": 2.260054211994498, "learning_rate": 1.861518419549388e-05, "loss": 0.577, "step": 3039 }, { "epoch": 1.944977607165707, "grad_norm": 2.1275478415404887, "learning_rate": 1.8614131923431436e-05, "loss": 0.6886, "step": 3040 }, { "epoch": 1.945617402431222, "grad_norm": 2.219196230773216, "learning_rate": 1.861307928149349e-05, "loss": 0.6599, "step": 3041 }, { "epoch": 1.9462571976967369, "grad_norm": 2.244817214316665, "learning_rate": 1.8612026269725243e-05, "loss": 0.7189, "step": 3042 }, { "epoch": 1.9468969929622522, "grad_norm": 2.1585939070537883, "learning_rate": 1.8610972888171906e-05, "loss": 0.6487, "step": 3043 }, { "epoch": 1.947536788227767, "grad_norm": 2.2402796674635646, "learning_rate": 1.860991913687871e-05, "loss": 0.6721, "step": 3044 }, { "epoch": 1.9481765834932823, "grad_norm": 2.130525058512201, "learning_rate": 1.8608865015890906e-05, "loss": 0.6325, "step": 3045 }, { "epoch": 1.948816378758797, "grad_norm": 2.3194742939062505, "learning_rate": 1.8607810525253746e-05, "loss": 0.6227, "step": 3046 }, { "epoch": 1.9494561740243124, "grad_norm": 2.474408300734115, "learning_rate": 1.860675566501252e-05, "loss": 0.6921, "step": 3047 }, { "epoch": 1.9500959692898272, "grad_norm": 2.180287093202859, "learning_rate": 1.860570043521251e-05, "loss": 0.6877, "step": 3048 }, { "epoch": 1.9507357645553423, "grad_norm": 2.025923094661276, "learning_rate": 1.860464483589904e-05, "loss": 0.7042, "step": 3049 }, { "epoch": 1.9513755598208573, "grad_norm": 2.214963194785786, "learning_rate": 1.8603588867117423e-05, "loss": 0.6248, "step": 3050 }, { "epoch": 1.9520153550863724, "grad_norm": 2.0114123877678027, "learning_rate": 1.8602532528913005e-05, "loss": 0.6643, "step": 3051 }, { "epoch": 1.9526551503518874, "grad_norm": 1.954172977510535, "learning_rate": 1.860147582133114e-05, "loss": 0.6184, "step": 3052 }, { "epoch": 1.9532949456174025, "grad_norm": 2.136269950475071, "learning_rate": 1.8600418744417208e-05, "loss": 0.6461, "step": 3053 }, { "epoch": 1.9539347408829175, "grad_norm": 2.4750389574805944, "learning_rate": 1.859936129821659e-05, "loss": 0.6506, "step": 3054 }, { "epoch": 1.9545745361484324, "grad_norm": 2.480420893437806, "learning_rate": 1.85983034827747e-05, "loss": 0.6353, "step": 3055 }, { "epoch": 1.9552143314139476, "grad_norm": 2.105586533885776, "learning_rate": 1.859724529813695e-05, "loss": 0.6319, "step": 3056 }, { "epoch": 1.9558541266794625, "grad_norm": 2.085954862784325, "learning_rate": 1.8596186744348785e-05, "loss": 0.6081, "step": 3057 }, { "epoch": 1.9564939219449777, "grad_norm": 2.420940251870677, "learning_rate": 1.8595127821455644e-05, "loss": 0.6648, "step": 3058 }, { "epoch": 1.9571337172104926, "grad_norm": 2.3516366663352026, "learning_rate": 1.859406852950301e-05, "loss": 0.6817, "step": 3059 }, { "epoch": 1.9577735124760078, "grad_norm": 2.90129095286872, "learning_rate": 1.859300886853636e-05, "loss": 0.675, "step": 3060 }, { "epoch": 1.9584133077415227, "grad_norm": 2.3495481904611655, "learning_rate": 1.8591948838601194e-05, "loss": 0.6498, "step": 3061 }, { "epoch": 1.9590531030070377, "grad_norm": 1.8749586635431381, "learning_rate": 1.8590888439743032e-05, "loss": 0.6394, "step": 3062 }, { "epoch": 1.9596928982725528, "grad_norm": 2.1316226635566644, "learning_rate": 1.8589827672007402e-05, "loss": 0.6647, "step": 3063 }, { "epoch": 1.9603326935380678, "grad_norm": 2.1249137177326256, "learning_rate": 1.8588766535439847e-05, "loss": 0.6206, "step": 3064 }, { "epoch": 1.9609724888035829, "grad_norm": 2.236640679221623, "learning_rate": 1.8587705030085938e-05, "loss": 0.6757, "step": 3065 }, { "epoch": 1.961612284069098, "grad_norm": 2.400804870417536, "learning_rate": 1.858664315599125e-05, "loss": 0.7062, "step": 3066 }, { "epoch": 1.962252079334613, "grad_norm": 1.8800988735347506, "learning_rate": 1.858558091320138e-05, "loss": 0.6307, "step": 3067 }, { "epoch": 1.9628918746001278, "grad_norm": 1.9591567907362588, "learning_rate": 1.8584518301761942e-05, "loss": 0.6167, "step": 3068 }, { "epoch": 1.963531669865643, "grad_norm": 2.1343879385318156, "learning_rate": 1.8583455321718556e-05, "loss": 0.6788, "step": 3069 }, { "epoch": 1.964171465131158, "grad_norm": 1.9951267887187267, "learning_rate": 1.8582391973116867e-05, "loss": 0.6281, "step": 3070 }, { "epoch": 1.9648112603966732, "grad_norm": 2.2708860136456397, "learning_rate": 1.8581328256002536e-05, "loss": 0.6553, "step": 3071 }, { "epoch": 1.965451055662188, "grad_norm": 2.3032420115279733, "learning_rate": 1.858026417042123e-05, "loss": 0.687, "step": 3072 }, { "epoch": 1.966090850927703, "grad_norm": 2.319578977327806, "learning_rate": 1.8579199716418645e-05, "loss": 0.6793, "step": 3073 }, { "epoch": 1.9667306461932181, "grad_norm": 2.0452411278797817, "learning_rate": 1.8578134894040488e-05, "loss": 0.6481, "step": 3074 }, { "epoch": 1.9673704414587332, "grad_norm": 2.28168639651009, "learning_rate": 1.8577069703332477e-05, "loss": 0.6293, "step": 3075 }, { "epoch": 1.9680102367242482, "grad_norm": 2.1520308806463833, "learning_rate": 1.857600414434035e-05, "loss": 0.6856, "step": 3076 }, { "epoch": 1.9686500319897633, "grad_norm": 2.3055609255247944, "learning_rate": 1.857493821710986e-05, "loss": 0.5745, "step": 3077 }, { "epoch": 1.9692898272552783, "grad_norm": 2.198274056309223, "learning_rate": 1.8573871921686775e-05, "loss": 0.6148, "step": 3078 }, { "epoch": 1.9699296225207934, "grad_norm": 2.5909222957232307, "learning_rate": 1.8572805258116885e-05, "loss": 0.6413, "step": 3079 }, { "epoch": 1.9705694177863085, "grad_norm": 2.269848267227061, "learning_rate": 1.8571738226445982e-05, "loss": 0.6004, "step": 3080 }, { "epoch": 1.9712092130518233, "grad_norm": 1.9843019151788002, "learning_rate": 1.857067082671989e-05, "loss": 0.6243, "step": 3081 }, { "epoch": 1.9718490083173386, "grad_norm": 2.0945589277045054, "learning_rate": 1.856960305898444e-05, "loss": 0.6614, "step": 3082 }, { "epoch": 1.9724888035828534, "grad_norm": 2.3160341138862317, "learning_rate": 1.856853492328548e-05, "loss": 0.6306, "step": 3083 }, { "epoch": 1.9731285988483687, "grad_norm": 1.8418671217215097, "learning_rate": 1.8567466419668872e-05, "loss": 0.6305, "step": 3084 }, { "epoch": 1.9737683941138835, "grad_norm": 2.236092574926071, "learning_rate": 1.8566397548180493e-05, "loss": 0.6707, "step": 3085 }, { "epoch": 1.9744081893793985, "grad_norm": 2.191543273689647, "learning_rate": 1.8565328308866248e-05, "loss": 0.5885, "step": 3086 }, { "epoch": 1.9750479846449136, "grad_norm": 1.9933063922868695, "learning_rate": 1.8564258701772034e-05, "loss": 0.6078, "step": 3087 }, { "epoch": 1.9756877799104287, "grad_norm": 2.201531229217128, "learning_rate": 1.856318872694379e-05, "loss": 0.6483, "step": 3088 }, { "epoch": 1.9763275751759437, "grad_norm": 2.1091122531591626, "learning_rate": 1.8562118384427457e-05, "loss": 0.6281, "step": 3089 }, { "epoch": 1.9769673704414588, "grad_norm": 2.0340548181146456, "learning_rate": 1.856104767426899e-05, "loss": 0.6139, "step": 3090 }, { "epoch": 1.9776071657069738, "grad_norm": 2.17486437014115, "learning_rate": 1.855997659651436e-05, "loss": 0.6211, "step": 3091 }, { "epoch": 1.9782469609724886, "grad_norm": 2.457559424485255, "learning_rate": 1.8558905151209567e-05, "loss": 0.6974, "step": 3092 }, { "epoch": 1.978886756238004, "grad_norm": 2.137202113328863, "learning_rate": 1.855783333840061e-05, "loss": 0.5865, "step": 3093 }, { "epoch": 1.9795265515035187, "grad_norm": 2.2326611863892145, "learning_rate": 1.8556761158133514e-05, "loss": 0.6203, "step": 3094 }, { "epoch": 1.980166346769034, "grad_norm": 2.396305469388457, "learning_rate": 1.8555688610454314e-05, "loss": 0.6814, "step": 3095 }, { "epoch": 1.9808061420345489, "grad_norm": 2.7928156514793336, "learning_rate": 1.8554615695409067e-05, "loss": 0.6318, "step": 3096 }, { "epoch": 1.9814459373000641, "grad_norm": 2.179638260081235, "learning_rate": 1.8553542413043836e-05, "loss": 0.6708, "step": 3097 }, { "epoch": 1.982085732565579, "grad_norm": 2.0360192444882412, "learning_rate": 1.855246876340471e-05, "loss": 0.6743, "step": 3098 }, { "epoch": 1.982725527831094, "grad_norm": 2.1447277553012087, "learning_rate": 1.855139474653779e-05, "loss": 0.6612, "step": 3099 }, { "epoch": 1.983365323096609, "grad_norm": 2.329088192407545, "learning_rate": 1.8550320362489192e-05, "loss": 0.6469, "step": 3100 }, { "epoch": 1.9840051183621241, "grad_norm": 2.0440276108755246, "learning_rate": 1.8549245611305045e-05, "loss": 0.6001, "step": 3101 }, { "epoch": 1.9846449136276392, "grad_norm": 2.1212805215380586, "learning_rate": 1.8548170493031503e-05, "loss": 0.6438, "step": 3102 }, { "epoch": 1.9852847088931542, "grad_norm": 2.0550150285779707, "learning_rate": 1.8547095007714722e-05, "loss": 0.5619, "step": 3103 }, { "epoch": 1.9859245041586693, "grad_norm": 1.9778363488442199, "learning_rate": 1.854601915540089e-05, "loss": 0.6258, "step": 3104 }, { "epoch": 1.986564299424184, "grad_norm": 2.0717543610146243, "learning_rate": 1.8544942936136193e-05, "loss": 0.5955, "step": 3105 }, { "epoch": 1.9872040946896994, "grad_norm": 1.9943740281208968, "learning_rate": 1.8543866349966852e-05, "loss": 0.6365, "step": 3106 }, { "epoch": 1.9878438899552142, "grad_norm": 1.9816732791200284, "learning_rate": 1.854278939693908e-05, "loss": 0.6455, "step": 3107 }, { "epoch": 1.9884836852207295, "grad_norm": 2.1860033823239546, "learning_rate": 1.8541712077099133e-05, "loss": 0.6194, "step": 3108 }, { "epoch": 1.9891234804862443, "grad_norm": 2.023960309679814, "learning_rate": 1.8540634390493268e-05, "loss": 0.6346, "step": 3109 }, { "epoch": 1.9897632757517596, "grad_norm": 2.1300909771976477, "learning_rate": 1.8539556337167747e-05, "loss": 0.6503, "step": 3110 }, { "epoch": 1.9904030710172744, "grad_norm": 2.0720917761342603, "learning_rate": 1.8538477917168874e-05, "loss": 0.5946, "step": 3111 }, { "epoch": 1.9910428662827895, "grad_norm": 2.2025910173028906, "learning_rate": 1.8537399130542946e-05, "loss": 0.5508, "step": 3112 }, { "epoch": 1.9916826615483045, "grad_norm": 2.0581579311344766, "learning_rate": 1.8536319977336284e-05, "loss": 0.5896, "step": 3113 }, { "epoch": 1.9923224568138196, "grad_norm": 2.00349794533132, "learning_rate": 1.8535240457595226e-05, "loss": 0.6726, "step": 3114 }, { "epoch": 1.9929622520793346, "grad_norm": 2.4179462362503106, "learning_rate": 1.8534160571366132e-05, "loss": 0.6325, "step": 3115 }, { "epoch": 1.9936020473448497, "grad_norm": 2.447680951908359, "learning_rate": 1.853308031869536e-05, "loss": 0.6184, "step": 3116 }, { "epoch": 1.9942418426103647, "grad_norm": 2.3124778498475425, "learning_rate": 1.8531999699629298e-05, "loss": 0.6475, "step": 3117 }, { "epoch": 1.9948816378758796, "grad_norm": 2.2743584522127986, "learning_rate": 1.8530918714214347e-05, "loss": 0.6856, "step": 3118 }, { "epoch": 1.9955214331413949, "grad_norm": 2.130288863613967, "learning_rate": 1.852983736249692e-05, "loss": 0.6632, "step": 3119 }, { "epoch": 1.9961612284069097, "grad_norm": 2.3379616443549542, "learning_rate": 1.8528755644523454e-05, "loss": 0.5981, "step": 3120 }, { "epoch": 1.996801023672425, "grad_norm": 2.1840317418713076, "learning_rate": 1.852767356034039e-05, "loss": 0.6551, "step": 3121 }, { "epoch": 1.9974408189379398, "grad_norm": 2.494902486191954, "learning_rate": 1.8526591109994193e-05, "loss": 0.6461, "step": 3122 }, { "epoch": 1.9980806142034548, "grad_norm": 2.2212207473508467, "learning_rate": 1.852550829353134e-05, "loss": 0.6331, "step": 3123 }, { "epoch": 1.99872040946897, "grad_norm": 2.282734656329446, "learning_rate": 1.8524425110998328e-05, "loss": 0.5924, "step": 3124 }, { "epoch": 1.999360204734485, "grad_norm": 2.5739852704365322, "learning_rate": 1.8523341562441662e-05, "loss": 0.6479, "step": 3125 }, { "epoch": 2.0, "grad_norm": 2.363065169790372, "learning_rate": 1.8522257647907874e-05, "loss": 0.5932, "step": 3126 }, { "epoch": 2.000639795265515, "grad_norm": 1.9759613894602233, "learning_rate": 1.8521173367443507e-05, "loss": 0.4165, "step": 3127 }, { "epoch": 2.00127959053103, "grad_norm": 1.7465204117380886, "learning_rate": 1.8520088721095106e-05, "loss": 0.4374, "step": 3128 }, { "epoch": 2.001919385796545, "grad_norm": 2.040778607363496, "learning_rate": 1.8519003708909255e-05, "loss": 0.4426, "step": 3129 }, { "epoch": 2.00255918106206, "grad_norm": 1.884617551074153, "learning_rate": 1.851791833093254e-05, "loss": 0.4805, "step": 3130 }, { "epoch": 2.003198976327575, "grad_norm": 2.178133668079798, "learning_rate": 1.851683258721156e-05, "loss": 0.4378, "step": 3131 }, { "epoch": 2.0038387715930903, "grad_norm": 2.28318205722937, "learning_rate": 1.8515746477792944e-05, "loss": 0.4577, "step": 3132 }, { "epoch": 2.004478566858605, "grad_norm": 2.2464152566325133, "learning_rate": 1.851466000272332e-05, "loss": 0.4411, "step": 3133 }, { "epoch": 2.0051183621241204, "grad_norm": 2.002880950996141, "learning_rate": 1.8513573162049344e-05, "loss": 0.4315, "step": 3134 }, { "epoch": 2.0057581573896353, "grad_norm": 2.2526661725944623, "learning_rate": 1.8512485955817675e-05, "loss": 0.5119, "step": 3135 }, { "epoch": 2.0063979526551505, "grad_norm": 2.085285916690703, "learning_rate": 1.851139838407501e-05, "loss": 0.4476, "step": 3136 }, { "epoch": 2.0070377479206654, "grad_norm": 1.9086759350754818, "learning_rate": 1.8510310446868032e-05, "loss": 0.4463, "step": 3137 }, { "epoch": 2.0076775431861806, "grad_norm": 2.0251169736362953, "learning_rate": 1.8509222144243466e-05, "loss": 0.4455, "step": 3138 }, { "epoch": 2.0083173384516955, "grad_norm": 2.072565151855681, "learning_rate": 1.8508133476248036e-05, "loss": 0.468, "step": 3139 }, { "epoch": 2.0089571337172103, "grad_norm": 1.969157634626575, "learning_rate": 1.8507044442928488e-05, "loss": 0.4705, "step": 3140 }, { "epoch": 2.0095969289827256, "grad_norm": 2.063222214605023, "learning_rate": 1.8505955044331586e-05, "loss": 0.4471, "step": 3141 }, { "epoch": 2.0102367242482404, "grad_norm": 2.07270825267157, "learning_rate": 1.850486528050411e-05, "loss": 0.455, "step": 3142 }, { "epoch": 2.0108765195137557, "grad_norm": 2.2786212247141595, "learning_rate": 1.850377515149284e-05, "loss": 0.4416, "step": 3143 }, { "epoch": 2.0115163147792705, "grad_norm": 1.987700845126543, "learning_rate": 1.85026846573446e-05, "loss": 0.4503, "step": 3144 }, { "epoch": 2.012156110044786, "grad_norm": 1.9355206162014018, "learning_rate": 1.85015937981062e-05, "loss": 0.4451, "step": 3145 }, { "epoch": 2.0127959053103006, "grad_norm": 1.8589183014081454, "learning_rate": 1.8500502573824485e-05, "loss": 0.4042, "step": 3146 }, { "epoch": 2.013435700575816, "grad_norm": 1.8799262396080316, "learning_rate": 1.8499410984546312e-05, "loss": 0.4684, "step": 3147 }, { "epoch": 2.0140754958413307, "grad_norm": 1.9592218009267435, "learning_rate": 1.8498319030318552e-05, "loss": 0.4552, "step": 3148 }, { "epoch": 2.014715291106846, "grad_norm": 2.207339476419429, "learning_rate": 1.849722671118809e-05, "loss": 0.4614, "step": 3149 }, { "epoch": 2.015355086372361, "grad_norm": 1.9724993773058526, "learning_rate": 1.849613402720183e-05, "loss": 0.4645, "step": 3150 }, { "epoch": 2.0159948816378757, "grad_norm": 1.7957779490224934, "learning_rate": 1.849504097840668e-05, "loss": 0.4436, "step": 3151 }, { "epoch": 2.016634676903391, "grad_norm": 1.9421151023736123, "learning_rate": 1.849394756484959e-05, "loss": 0.4627, "step": 3152 }, { "epoch": 2.0172744721689058, "grad_norm": 1.9999284164587277, "learning_rate": 1.84928537865775e-05, "loss": 0.4342, "step": 3153 }, { "epoch": 2.017914267434421, "grad_norm": 2.0127607293949086, "learning_rate": 1.8491759643637378e-05, "loss": 0.4391, "step": 3154 }, { "epoch": 2.018554062699936, "grad_norm": 2.053968890776399, "learning_rate": 1.84906651360762e-05, "loss": 0.4507, "step": 3155 }, { "epoch": 2.019193857965451, "grad_norm": 1.8880263999799494, "learning_rate": 1.8489570263940963e-05, "loss": 0.463, "step": 3156 }, { "epoch": 2.019833653230966, "grad_norm": 1.836994413066695, "learning_rate": 1.8488475027278682e-05, "loss": 0.4439, "step": 3157 }, { "epoch": 2.0204734484964813, "grad_norm": 1.961569581968994, "learning_rate": 1.848737942613638e-05, "loss": 0.4409, "step": 3158 }, { "epoch": 2.021113243761996, "grad_norm": 1.954515608386902, "learning_rate": 1.848628346056111e-05, "loss": 0.484, "step": 3159 }, { "epoch": 2.0217530390275114, "grad_norm": 1.760978513856685, "learning_rate": 1.848518713059992e-05, "loss": 0.4339, "step": 3160 }, { "epoch": 2.022392834293026, "grad_norm": 1.835340160163322, "learning_rate": 1.848409043629989e-05, "loss": 0.478, "step": 3161 }, { "epoch": 2.0230326295585415, "grad_norm": 1.9750497010696006, "learning_rate": 1.848299337770811e-05, "loss": 0.4358, "step": 3162 }, { "epoch": 2.0236724248240563, "grad_norm": 1.9235241473119458, "learning_rate": 1.848189595487168e-05, "loss": 0.431, "step": 3163 }, { "epoch": 2.024312220089571, "grad_norm": 1.8473691962828407, "learning_rate": 1.848079816783773e-05, "loss": 0.4369, "step": 3164 }, { "epoch": 2.0249520153550864, "grad_norm": 1.9544611104642344, "learning_rate": 1.847970001665339e-05, "loss": 0.4506, "step": 3165 }, { "epoch": 2.0255918106206012, "grad_norm": 1.9380757072643309, "learning_rate": 1.8478601501365817e-05, "loss": 0.4365, "step": 3166 }, { "epoch": 2.0262316058861165, "grad_norm": 1.926705614922382, "learning_rate": 1.8477502622022178e-05, "loss": 0.4956, "step": 3167 }, { "epoch": 2.0268714011516313, "grad_norm": 2.226535498864768, "learning_rate": 1.8476403378669653e-05, "loss": 0.4486, "step": 3168 }, { "epoch": 2.0275111964171466, "grad_norm": 1.951032186620974, "learning_rate": 1.847530377135545e-05, "loss": 0.4243, "step": 3169 }, { "epoch": 2.0281509916826614, "grad_norm": 2.0294875784565365, "learning_rate": 1.8474203800126778e-05, "loss": 0.4351, "step": 3170 }, { "epoch": 2.0287907869481767, "grad_norm": 1.931266600853799, "learning_rate": 1.847310346503087e-05, "loss": 0.4691, "step": 3171 }, { "epoch": 2.0294305822136915, "grad_norm": 1.7626012563118008, "learning_rate": 1.847200276611497e-05, "loss": 0.446, "step": 3172 }, { "epoch": 2.030070377479207, "grad_norm": 1.7846184326340595, "learning_rate": 1.8470901703426343e-05, "loss": 0.4283, "step": 3173 }, { "epoch": 2.0307101727447217, "grad_norm": 1.787908355155538, "learning_rate": 1.8469800277012263e-05, "loss": 0.4235, "step": 3174 }, { "epoch": 2.031349968010237, "grad_norm": 1.9406733647618069, "learning_rate": 1.846869848692003e-05, "loss": 0.4575, "step": 3175 }, { "epoch": 2.0319897632757518, "grad_norm": 1.9635703424945432, "learning_rate": 1.8467596333196945e-05, "loss": 0.442, "step": 3176 }, { "epoch": 2.0326295585412666, "grad_norm": 1.8905345073592497, "learning_rate": 1.846649381589034e-05, "loss": 0.4725, "step": 3177 }, { "epoch": 2.033269353806782, "grad_norm": 2.0166248299801746, "learning_rate": 1.8465390935047548e-05, "loss": 0.489, "step": 3178 }, { "epoch": 2.0339091490722967, "grad_norm": 2.078824684396249, "learning_rate": 1.8464287690715927e-05, "loss": 0.5089, "step": 3179 }, { "epoch": 2.034548944337812, "grad_norm": 1.9263779332703237, "learning_rate": 1.8463184082942855e-05, "loss": 0.4504, "step": 3180 }, { "epoch": 2.035188739603327, "grad_norm": 1.9109030614151163, "learning_rate": 1.8462080111775707e-05, "loss": 0.4584, "step": 3181 }, { "epoch": 2.035828534868842, "grad_norm": 1.9456681303351266, "learning_rate": 1.8460975777261895e-05, "loss": 0.4675, "step": 3182 }, { "epoch": 2.036468330134357, "grad_norm": 1.9406572120952825, "learning_rate": 1.845987107944883e-05, "loss": 0.4627, "step": 3183 }, { "epoch": 2.037108125399872, "grad_norm": 1.87336340619355, "learning_rate": 1.8458766018383956e-05, "loss": 0.4178, "step": 3184 }, { "epoch": 2.037747920665387, "grad_norm": 1.993234805704716, "learning_rate": 1.845766059411471e-05, "loss": 0.4523, "step": 3185 }, { "epoch": 2.0383877159309023, "grad_norm": 1.912508978150446, "learning_rate": 1.8456554806688568e-05, "loss": 0.4334, "step": 3186 }, { "epoch": 2.039027511196417, "grad_norm": 1.9363324851648924, "learning_rate": 1.8455448656153003e-05, "loss": 0.4673, "step": 3187 }, { "epoch": 2.0396673064619324, "grad_norm": 2.3296515272708964, "learning_rate": 1.845434214255552e-05, "loss": 0.4419, "step": 3188 }, { "epoch": 2.0403071017274472, "grad_norm": 1.9225100817660765, "learning_rate": 1.8453235265943616e-05, "loss": 0.4352, "step": 3189 }, { "epoch": 2.040946896992962, "grad_norm": 2.023442422035087, "learning_rate": 1.845212802636483e-05, "loss": 0.4691, "step": 3190 }, { "epoch": 2.0415866922584773, "grad_norm": 1.9895135951390501, "learning_rate": 1.8451020423866703e-05, "loss": 0.4174, "step": 3191 }, { "epoch": 2.042226487523992, "grad_norm": 1.9784883481151179, "learning_rate": 1.844991245849679e-05, "loss": 0.4174, "step": 3192 }, { "epoch": 2.0428662827895074, "grad_norm": 1.7408602742639034, "learning_rate": 1.844880413030267e-05, "loss": 0.4198, "step": 3193 }, { "epoch": 2.0435060780550223, "grad_norm": 2.037659769191533, "learning_rate": 1.8447695439331926e-05, "loss": 0.4623, "step": 3194 }, { "epoch": 2.0441458733205375, "grad_norm": 1.9445130287392516, "learning_rate": 1.8446586385632172e-05, "loss": 0.4485, "step": 3195 }, { "epoch": 2.0447856685860524, "grad_norm": 2.081794123426113, "learning_rate": 1.844547696925102e-05, "loss": 0.4662, "step": 3196 }, { "epoch": 2.0454254638515676, "grad_norm": 1.9897185200589487, "learning_rate": 1.844436719023611e-05, "loss": 0.501, "step": 3197 }, { "epoch": 2.0460652591170825, "grad_norm": 1.8230454206342834, "learning_rate": 1.8443257048635096e-05, "loss": 0.4092, "step": 3198 }, { "epoch": 2.0467050543825978, "grad_norm": 1.9832835067111, "learning_rate": 1.8442146544495645e-05, "loss": 0.4529, "step": 3199 }, { "epoch": 2.0473448496481126, "grad_norm": 1.9760814928058703, "learning_rate": 1.844103567786544e-05, "loss": 0.4632, "step": 3200 }, { "epoch": 2.047984644913628, "grad_norm": 2.0272514864004187, "learning_rate": 1.8439924448792172e-05, "loss": 0.4504, "step": 3201 }, { "epoch": 2.0486244401791427, "grad_norm": 2.2243027624189358, "learning_rate": 1.8438812857323564e-05, "loss": 0.4425, "step": 3202 }, { "epoch": 2.0492642354446575, "grad_norm": 2.0264253183701006, "learning_rate": 1.8437700903507346e-05, "loss": 0.4417, "step": 3203 }, { "epoch": 2.049904030710173, "grad_norm": 1.727966305782272, "learning_rate": 1.843658858739126e-05, "loss": 0.428, "step": 3204 }, { "epoch": 2.0505438259756876, "grad_norm": 2.148262383650573, "learning_rate": 1.843547590902307e-05, "loss": 0.4421, "step": 3205 }, { "epoch": 2.051183621241203, "grad_norm": 1.9353989899212036, "learning_rate": 1.8434362868450545e-05, "loss": 0.524, "step": 3206 }, { "epoch": 2.0518234165067177, "grad_norm": 1.935121731961289, "learning_rate": 1.843324946572148e-05, "loss": 0.4281, "step": 3207 }, { "epoch": 2.052463211772233, "grad_norm": 2.178271345987937, "learning_rate": 1.8432135700883692e-05, "loss": 0.4758, "step": 3208 }, { "epoch": 2.053103007037748, "grad_norm": 1.941308499353263, "learning_rate": 1.8431021573984993e-05, "loss": 0.4381, "step": 3209 }, { "epoch": 2.053742802303263, "grad_norm": 2.1393432143977744, "learning_rate": 1.8429907085073225e-05, "loss": 0.4489, "step": 3210 }, { "epoch": 2.054382597568778, "grad_norm": 1.9622202292644724, "learning_rate": 1.842879223419624e-05, "loss": 0.4432, "step": 3211 }, { "epoch": 2.055022392834293, "grad_norm": 1.9412152095039898, "learning_rate": 1.8427677021401914e-05, "loss": 0.5006, "step": 3212 }, { "epoch": 2.055662188099808, "grad_norm": 2.0728114639680904, "learning_rate": 1.842656144673813e-05, "loss": 0.4526, "step": 3213 }, { "epoch": 2.056301983365323, "grad_norm": 1.9078691211396173, "learning_rate": 1.842544551025278e-05, "loss": 0.4375, "step": 3214 }, { "epoch": 2.056941778630838, "grad_norm": 2.1808143873525814, "learning_rate": 1.8424329211993795e-05, "loss": 0.4348, "step": 3215 }, { "epoch": 2.057581573896353, "grad_norm": 2.196652482944125, "learning_rate": 1.8423212552009094e-05, "loss": 0.494, "step": 3216 }, { "epoch": 2.0582213691618683, "grad_norm": 1.968018658435172, "learning_rate": 1.8422095530346635e-05, "loss": 0.4613, "step": 3217 }, { "epoch": 2.058861164427383, "grad_norm": 1.8842461412878446, "learning_rate": 1.842097814705437e-05, "loss": 0.4185, "step": 3218 }, { "epoch": 2.0595009596928984, "grad_norm": 1.9868061189109576, "learning_rate": 1.8419860402180286e-05, "loss": 0.4188, "step": 3219 }, { "epoch": 2.060140754958413, "grad_norm": 2.021837939349601, "learning_rate": 1.8418742295772377e-05, "loss": 0.463, "step": 3220 }, { "epoch": 2.0607805502239285, "grad_norm": 1.899854189602249, "learning_rate": 1.8417623827878647e-05, "loss": 0.4127, "step": 3221 }, { "epoch": 2.0614203454894433, "grad_norm": 1.9295793258931333, "learning_rate": 1.841650499854712e-05, "loss": 0.4389, "step": 3222 }, { "epoch": 2.0620601407549586, "grad_norm": 2.108835889119033, "learning_rate": 1.8415385807825844e-05, "loss": 0.4771, "step": 3223 }, { "epoch": 2.0626999360204734, "grad_norm": 2.0008877544562838, "learning_rate": 1.841426625576287e-05, "loss": 0.4066, "step": 3224 }, { "epoch": 2.0633397312859887, "grad_norm": 1.847932178130914, "learning_rate": 1.8413146342406273e-05, "loss": 0.4274, "step": 3225 }, { "epoch": 2.0639795265515035, "grad_norm": 1.9269520561807996, "learning_rate": 1.8412026067804137e-05, "loss": 0.4737, "step": 3226 }, { "epoch": 2.0646193218170183, "grad_norm": 2.009119511586518, "learning_rate": 1.8410905432004564e-05, "loss": 0.4476, "step": 3227 }, { "epoch": 2.0652591170825336, "grad_norm": 2.2124307045177827, "learning_rate": 1.8409784435055672e-05, "loss": 0.456, "step": 3228 }, { "epoch": 2.0658989123480485, "grad_norm": 2.084328704798374, "learning_rate": 1.84086630770056e-05, "loss": 0.4957, "step": 3229 }, { "epoch": 2.0665387076135637, "grad_norm": 1.9071721851783938, "learning_rate": 1.840754135790249e-05, "loss": 0.502, "step": 3230 }, { "epoch": 2.0671785028790786, "grad_norm": 1.9115004767926014, "learning_rate": 1.8406419277794514e-05, "loss": 0.4757, "step": 3231 }, { "epoch": 2.067818298144594, "grad_norm": 1.818481921037835, "learning_rate": 1.8405296836729846e-05, "loss": 0.4097, "step": 3232 }, { "epoch": 2.0684580934101087, "grad_norm": 2.119634233299938, "learning_rate": 1.8404174034756684e-05, "loss": 0.4354, "step": 3233 }, { "epoch": 2.069097888675624, "grad_norm": 1.960906460982008, "learning_rate": 1.8403050871923237e-05, "loss": 0.4398, "step": 3234 }, { "epoch": 2.0697376839411388, "grad_norm": 1.8402217427051903, "learning_rate": 1.8401927348277737e-05, "loss": 0.4466, "step": 3235 }, { "epoch": 2.070377479206654, "grad_norm": 1.8597344100811093, "learning_rate": 1.840080346386842e-05, "loss": 0.4613, "step": 3236 }, { "epoch": 2.071017274472169, "grad_norm": 2.0172357744636855, "learning_rate": 1.8399679218743543e-05, "loss": 0.4404, "step": 3237 }, { "epoch": 2.071657069737684, "grad_norm": 2.2291822761116666, "learning_rate": 1.839855461295139e-05, "loss": 0.4807, "step": 3238 }, { "epoch": 2.072296865003199, "grad_norm": 1.9682450364286697, "learning_rate": 1.8397429646540235e-05, "loss": 0.4709, "step": 3239 }, { "epoch": 2.072936660268714, "grad_norm": 1.9640889205727066, "learning_rate": 1.8396304319558395e-05, "loss": 0.4338, "step": 3240 }, { "epoch": 2.073576455534229, "grad_norm": 1.9571245528471313, "learning_rate": 1.839517863205418e-05, "loss": 0.4187, "step": 3241 }, { "epoch": 2.074216250799744, "grad_norm": 1.9844163714942082, "learning_rate": 1.8394052584075925e-05, "loss": 0.4785, "step": 3242 }, { "epoch": 2.074856046065259, "grad_norm": 1.9615867231443689, "learning_rate": 1.8392926175671986e-05, "loss": 0.4795, "step": 3243 }, { "epoch": 2.075495841330774, "grad_norm": 1.8933627557487245, "learning_rate": 1.8391799406890728e-05, "loss": 0.3943, "step": 3244 }, { "epoch": 2.0761356365962893, "grad_norm": 1.9352549570714088, "learning_rate": 1.839067227778053e-05, "loss": 0.4227, "step": 3245 }, { "epoch": 2.076775431861804, "grad_norm": 1.8091297764605958, "learning_rate": 1.8389544788389795e-05, "loss": 0.4159, "step": 3246 }, { "epoch": 2.0774152271273194, "grad_norm": 1.8957269473499194, "learning_rate": 1.8388416938766926e-05, "loss": 0.4578, "step": 3247 }, { "epoch": 2.0780550223928342, "grad_norm": 1.9686410160296086, "learning_rate": 1.838728872896036e-05, "loss": 0.4677, "step": 3248 }, { "epoch": 2.0786948176583495, "grad_norm": 2.0099993581597237, "learning_rate": 1.8386160159018527e-05, "loss": 0.4553, "step": 3249 }, { "epoch": 2.0793346129238643, "grad_norm": 1.973503311554468, "learning_rate": 1.83850312289899e-05, "loss": 0.4678, "step": 3250 }, { "epoch": 2.079974408189379, "grad_norm": 1.9974920180685714, "learning_rate": 1.838390193892295e-05, "loss": 0.461, "step": 3251 }, { "epoch": 2.0806142034548945, "grad_norm": 2.050651879951216, "learning_rate": 1.8382772288866163e-05, "loss": 0.4759, "step": 3252 }, { "epoch": 2.0812539987204093, "grad_norm": 1.9347236295014993, "learning_rate": 1.8381642278868042e-05, "loss": 0.4492, "step": 3253 }, { "epoch": 2.0818937939859246, "grad_norm": 1.9547419336242586, "learning_rate": 1.8380511908977115e-05, "loss": 0.4658, "step": 3254 }, { "epoch": 2.0825335892514394, "grad_norm": 2.064977611498674, "learning_rate": 1.8379381179241916e-05, "loss": 0.4547, "step": 3255 }, { "epoch": 2.0831733845169547, "grad_norm": 1.9516207824602865, "learning_rate": 1.837825008971099e-05, "loss": 0.41, "step": 3256 }, { "epoch": 2.0838131797824695, "grad_norm": 1.9841762760134924, "learning_rate": 1.8377118640432914e-05, "loss": 0.4457, "step": 3257 }, { "epoch": 2.0844529750479848, "grad_norm": 1.9180549618776255, "learning_rate": 1.8375986831456262e-05, "loss": 0.4482, "step": 3258 }, { "epoch": 2.0850927703134996, "grad_norm": 1.9927499563652362, "learning_rate": 1.8374854662829634e-05, "loss": 0.4173, "step": 3259 }, { "epoch": 2.085732565579015, "grad_norm": 2.2028745580936104, "learning_rate": 1.837372213460165e-05, "loss": 0.4446, "step": 3260 }, { "epoch": 2.0863723608445297, "grad_norm": 2.0710333627765096, "learning_rate": 1.837258924682093e-05, "loss": 0.4572, "step": 3261 }, { "epoch": 2.087012156110045, "grad_norm": 1.8946803343472511, "learning_rate": 1.8371455999536122e-05, "loss": 0.4587, "step": 3262 }, { "epoch": 2.08765195137556, "grad_norm": 1.756573705226281, "learning_rate": 1.8370322392795882e-05, "loss": 0.4236, "step": 3263 }, { "epoch": 2.0882917466410746, "grad_norm": 1.835481391954523, "learning_rate": 1.836918842664889e-05, "loss": 0.4366, "step": 3264 }, { "epoch": 2.08893154190659, "grad_norm": 2.2834531128276887, "learning_rate": 1.8368054101143834e-05, "loss": 0.4061, "step": 3265 }, { "epoch": 2.0895713371721047, "grad_norm": 2.0561108632336227, "learning_rate": 1.8366919416329427e-05, "loss": 0.4436, "step": 3266 }, { "epoch": 2.09021113243762, "grad_norm": 2.043316793712683, "learning_rate": 1.8365784372254378e-05, "loss": 0.4883, "step": 3267 }, { "epoch": 2.090850927703135, "grad_norm": 1.8888089394304657, "learning_rate": 1.8364648968967428e-05, "loss": 0.4107, "step": 3268 }, { "epoch": 2.09149072296865, "grad_norm": 1.9889713365555213, "learning_rate": 1.8363513206517332e-05, "loss": 0.4604, "step": 3269 }, { "epoch": 2.092130518234165, "grad_norm": 1.998186188957879, "learning_rate": 1.8362377084952857e-05, "loss": 0.454, "step": 3270 }, { "epoch": 2.0927703134996802, "grad_norm": 1.8557679123598358, "learning_rate": 1.8361240604322784e-05, "loss": 0.4208, "step": 3271 }, { "epoch": 2.093410108765195, "grad_norm": 1.8455937658651842, "learning_rate": 1.8360103764675916e-05, "loss": 0.4726, "step": 3272 }, { "epoch": 2.0940499040307103, "grad_norm": 1.7520898069836854, "learning_rate": 1.8358966566061062e-05, "loss": 0.4202, "step": 3273 }, { "epoch": 2.094689699296225, "grad_norm": 2.024416673030983, "learning_rate": 1.8357829008527055e-05, "loss": 0.4773, "step": 3274 }, { "epoch": 2.0953294945617404, "grad_norm": 1.8754152412477136, "learning_rate": 1.8356691092122736e-05, "loss": 0.4793, "step": 3275 }, { "epoch": 2.0959692898272553, "grad_norm": 1.859705544986386, "learning_rate": 1.8355552816896968e-05, "loss": 0.3981, "step": 3276 }, { "epoch": 2.09660908509277, "grad_norm": 2.1216477445342172, "learning_rate": 1.8354414182898625e-05, "loss": 0.4127, "step": 3277 }, { "epoch": 2.0972488803582854, "grad_norm": 2.048362153134686, "learning_rate": 1.8353275190176597e-05, "loss": 0.4609, "step": 3278 }, { "epoch": 2.0978886756238, "grad_norm": 1.9998342830277056, "learning_rate": 1.8352135838779793e-05, "loss": 0.4216, "step": 3279 }, { "epoch": 2.0985284708893155, "grad_norm": 1.9327016413886342, "learning_rate": 1.8350996128757136e-05, "loss": 0.4966, "step": 3280 }, { "epoch": 2.0991682661548303, "grad_norm": 1.9031786012418328, "learning_rate": 1.8349856060157555e-05, "loss": 0.4724, "step": 3281 }, { "epoch": 2.0998080614203456, "grad_norm": 1.7995615235063331, "learning_rate": 1.834871563303001e-05, "loss": 0.3809, "step": 3282 }, { "epoch": 2.1004478566858604, "grad_norm": 1.9444598435206657, "learning_rate": 1.834757484742347e-05, "loss": 0.4533, "step": 3283 }, { "epoch": 2.1010876519513757, "grad_norm": 2.0650691604808524, "learning_rate": 1.834643370338691e-05, "loss": 0.4401, "step": 3284 }, { "epoch": 2.1017274472168905, "grad_norm": 1.8281952856474468, "learning_rate": 1.8345292200969334e-05, "loss": 0.4146, "step": 3285 }, { "epoch": 2.102367242482406, "grad_norm": 1.9380572192061218, "learning_rate": 1.834415034021976e-05, "loss": 0.4952, "step": 3286 }, { "epoch": 2.1030070377479206, "grad_norm": 2.0027103615997235, "learning_rate": 1.8343008121187215e-05, "loss": 0.4461, "step": 3287 }, { "epoch": 2.103646833013436, "grad_norm": 2.160517801282483, "learning_rate": 1.834186554392074e-05, "loss": 0.4876, "step": 3288 }, { "epoch": 2.1042866282789507, "grad_norm": 1.9955789321168953, "learning_rate": 1.8340722608469396e-05, "loss": 0.428, "step": 3289 }, { "epoch": 2.1049264235444656, "grad_norm": 1.9853502891178882, "learning_rate": 1.833957931488226e-05, "loss": 0.4419, "step": 3290 }, { "epoch": 2.105566218809981, "grad_norm": 2.051673136756737, "learning_rate": 1.8338435663208427e-05, "loss": 0.4004, "step": 3291 }, { "epoch": 2.1062060140754957, "grad_norm": 2.102836911864055, "learning_rate": 1.8337291653496995e-05, "loss": 0.5, "step": 3292 }, { "epoch": 2.106845809341011, "grad_norm": 2.110146626428777, "learning_rate": 1.8336147285797096e-05, "loss": 0.4427, "step": 3293 }, { "epoch": 2.107485604606526, "grad_norm": 1.790608534627079, "learning_rate": 1.8335002560157855e-05, "loss": 0.4626, "step": 3294 }, { "epoch": 2.108125399872041, "grad_norm": 1.83666388675089, "learning_rate": 1.8333857476628434e-05, "loss": 0.4692, "step": 3295 }, { "epoch": 2.108765195137556, "grad_norm": 2.133751688608335, "learning_rate": 1.8332712035257998e-05, "loss": 0.4593, "step": 3296 }, { "epoch": 2.109404990403071, "grad_norm": 1.8340118077825607, "learning_rate": 1.833156623609573e-05, "loss": 0.4918, "step": 3297 }, { "epoch": 2.110044785668586, "grad_norm": 1.8360854168782856, "learning_rate": 1.8330420079190826e-05, "loss": 0.4268, "step": 3298 }, { "epoch": 2.1106845809341013, "grad_norm": 1.9300952151396153, "learning_rate": 1.8329273564592505e-05, "loss": 0.4591, "step": 3299 }, { "epoch": 2.111324376199616, "grad_norm": 2.1620679599097348, "learning_rate": 1.8328126692349992e-05, "loss": 0.505, "step": 3300 }, { "epoch": 2.1119641714651314, "grad_norm": 1.807855727218728, "learning_rate": 1.8326979462512534e-05, "loss": 0.425, "step": 3301 }, { "epoch": 2.112603966730646, "grad_norm": 1.824558674453988, "learning_rate": 1.832583187512939e-05, "loss": 0.4687, "step": 3302 }, { "epoch": 2.113243761996161, "grad_norm": 1.9004054124491696, "learning_rate": 1.832468393024984e-05, "loss": 0.4734, "step": 3303 }, { "epoch": 2.1138835572616763, "grad_norm": 1.9564091997286204, "learning_rate": 1.8323535627923167e-05, "loss": 0.4532, "step": 3304 }, { "epoch": 2.114523352527191, "grad_norm": 1.9543769419909567, "learning_rate": 1.832238696819868e-05, "loss": 0.4376, "step": 3305 }, { "epoch": 2.1151631477927064, "grad_norm": 1.9515988154771804, "learning_rate": 1.83212379511257e-05, "loss": 0.4406, "step": 3306 }, { "epoch": 2.1158029430582213, "grad_norm": 1.7855998321154916, "learning_rate": 1.8320088576753568e-05, "loss": 0.4052, "step": 3307 }, { "epoch": 2.1164427383237365, "grad_norm": 1.8225608711337487, "learning_rate": 1.8318938845131632e-05, "loss": 0.431, "step": 3308 }, { "epoch": 2.1170825335892514, "grad_norm": 1.9885392491497278, "learning_rate": 1.8317788756309257e-05, "loss": 0.4666, "step": 3309 }, { "epoch": 2.1177223288547666, "grad_norm": 2.046458119670318, "learning_rate": 1.8316638310335834e-05, "loss": 0.4646, "step": 3310 }, { "epoch": 2.1183621241202815, "grad_norm": 2.0986575074743103, "learning_rate": 1.8315487507260753e-05, "loss": 0.4088, "step": 3311 }, { "epoch": 2.1190019193857967, "grad_norm": 1.9357141823921213, "learning_rate": 1.831433634713343e-05, "loss": 0.474, "step": 3312 }, { "epoch": 2.1196417146513116, "grad_norm": 2.055220981627828, "learning_rate": 1.8313184830003298e-05, "loss": 0.4671, "step": 3313 }, { "epoch": 2.1202815099168264, "grad_norm": 2.0829750062228283, "learning_rate": 1.8312032955919793e-05, "loss": 0.4704, "step": 3314 }, { "epoch": 2.1209213051823417, "grad_norm": 2.060602773989726, "learning_rate": 1.8310880724932382e-05, "loss": 0.4337, "step": 3315 }, { "epoch": 2.1215611004478565, "grad_norm": 1.9745533404186384, "learning_rate": 1.8309728137090536e-05, "loss": 0.461, "step": 3316 }, { "epoch": 2.122200895713372, "grad_norm": 2.1801737593194197, "learning_rate": 1.8308575192443742e-05, "loss": 0.4613, "step": 3317 }, { "epoch": 2.1228406909788866, "grad_norm": 2.2021723422575747, "learning_rate": 1.8307421891041514e-05, "loss": 0.4878, "step": 3318 }, { "epoch": 2.123480486244402, "grad_norm": 1.9252271084073704, "learning_rate": 1.8306268232933363e-05, "loss": 0.4403, "step": 3319 }, { "epoch": 2.1241202815099167, "grad_norm": 2.0397960676829303, "learning_rate": 1.8305114218168833e-05, "loss": 0.4664, "step": 3320 }, { "epoch": 2.124760076775432, "grad_norm": 1.9887380269904626, "learning_rate": 1.8303959846797473e-05, "loss": 0.4721, "step": 3321 }, { "epoch": 2.125399872040947, "grad_norm": 1.8888958036482548, "learning_rate": 1.8302805118868847e-05, "loss": 0.4688, "step": 3322 }, { "epoch": 2.126039667306462, "grad_norm": 1.9359891899471846, "learning_rate": 1.8301650034432538e-05, "loss": 0.4347, "step": 3323 }, { "epoch": 2.126679462571977, "grad_norm": 1.9666089022770885, "learning_rate": 1.8300494593538147e-05, "loss": 0.4352, "step": 3324 }, { "epoch": 2.127319257837492, "grad_norm": 1.9253348877407135, "learning_rate": 1.829933879623528e-05, "loss": 0.4953, "step": 3325 }, { "epoch": 2.127959053103007, "grad_norm": 1.864381415251645, "learning_rate": 1.829818264257357e-05, "loss": 0.3839, "step": 3326 }, { "epoch": 2.128598848368522, "grad_norm": 1.9797689433000312, "learning_rate": 1.829702613260266e-05, "loss": 0.4414, "step": 3327 }, { "epoch": 2.129238643634037, "grad_norm": 2.0166437016435967, "learning_rate": 1.8295869266372208e-05, "loss": 0.4618, "step": 3328 }, { "epoch": 2.129878438899552, "grad_norm": 1.9499771743223675, "learning_rate": 1.8294712043931885e-05, "loss": 0.4547, "step": 3329 }, { "epoch": 2.1305182341650672, "grad_norm": 2.0709236918559406, "learning_rate": 1.829355446533138e-05, "loss": 0.4521, "step": 3330 }, { "epoch": 2.131158029430582, "grad_norm": 1.9493701196336248, "learning_rate": 1.8292396530620402e-05, "loss": 0.473, "step": 3331 }, { "epoch": 2.1317978246960974, "grad_norm": 1.9684735433706473, "learning_rate": 1.8291238239848668e-05, "loss": 0.4199, "step": 3332 }, { "epoch": 2.132437619961612, "grad_norm": 1.9648616466708617, "learning_rate": 1.829007959306591e-05, "loss": 0.4167, "step": 3333 }, { "epoch": 2.1330774152271275, "grad_norm": 1.835924413280501, "learning_rate": 1.8288920590321884e-05, "loss": 0.4256, "step": 3334 }, { "epoch": 2.1337172104926423, "grad_norm": 2.0286001707476538, "learning_rate": 1.828776123166635e-05, "loss": 0.4541, "step": 3335 }, { "epoch": 2.1343570057581576, "grad_norm": 1.9906934060533883, "learning_rate": 1.8286601517149096e-05, "loss": 0.4262, "step": 3336 }, { "epoch": 2.1349968010236724, "grad_norm": 1.966596984452537, "learning_rate": 1.828544144681991e-05, "loss": 0.431, "step": 3337 }, { "epoch": 2.1356365962891877, "grad_norm": 1.9651577533070088, "learning_rate": 1.828428102072861e-05, "loss": 0.4321, "step": 3338 }, { "epoch": 2.1362763915547025, "grad_norm": 2.0730483422800443, "learning_rate": 1.828312023892502e-05, "loss": 0.5072, "step": 3339 }, { "epoch": 2.1369161868202173, "grad_norm": 2.0555112236861666, "learning_rate": 1.8281959101458975e-05, "loss": 0.4341, "step": 3340 }, { "epoch": 2.1375559820857326, "grad_norm": 1.900292622976968, "learning_rate": 1.828079760838034e-05, "loss": 0.4655, "step": 3341 }, { "epoch": 2.1381957773512474, "grad_norm": 1.9317033993685546, "learning_rate": 1.8279635759738995e-05, "loss": 0.483, "step": 3342 }, { "epoch": 2.1388355726167627, "grad_norm": 1.9430670231531377, "learning_rate": 1.827847355558481e-05, "loss": 0.4496, "step": 3343 }, { "epoch": 2.1394753678822775, "grad_norm": 1.8155195993966893, "learning_rate": 1.8277310995967702e-05, "loss": 0.4296, "step": 3344 }, { "epoch": 2.140115163147793, "grad_norm": 2.0712263097187873, "learning_rate": 1.827614808093758e-05, "loss": 0.4746, "step": 3345 }, { "epoch": 2.1407549584133077, "grad_norm": 1.9756912277347474, "learning_rate": 1.827498481054438e-05, "loss": 0.4849, "step": 3346 }, { "epoch": 2.141394753678823, "grad_norm": 1.8815272659353846, "learning_rate": 1.8273821184838058e-05, "loss": 0.4136, "step": 3347 }, { "epoch": 2.1420345489443378, "grad_norm": 1.9692563949267046, "learning_rate": 1.827265720386857e-05, "loss": 0.4397, "step": 3348 }, { "epoch": 2.142674344209853, "grad_norm": 2.103684902562109, "learning_rate": 1.82714928676859e-05, "loss": 0.4763, "step": 3349 }, { "epoch": 2.143314139475368, "grad_norm": 2.081254062616983, "learning_rate": 1.8270328176340032e-05, "loss": 0.4471, "step": 3350 }, { "epoch": 2.1439539347408827, "grad_norm": 2.0018422768936532, "learning_rate": 1.8269163129880993e-05, "loss": 0.4592, "step": 3351 }, { "epoch": 2.144593730006398, "grad_norm": 1.9448714247742964, "learning_rate": 1.8267997728358793e-05, "loss": 0.4532, "step": 3352 }, { "epoch": 2.145233525271913, "grad_norm": 1.9285881366436353, "learning_rate": 1.826683197182348e-05, "loss": 0.44, "step": 3353 }, { "epoch": 2.145873320537428, "grad_norm": 2.04892634539826, "learning_rate": 1.8265665860325108e-05, "loss": 0.4667, "step": 3354 }, { "epoch": 2.146513115802943, "grad_norm": 1.8621543892060413, "learning_rate": 1.8264499393913748e-05, "loss": 0.48, "step": 3355 }, { "epoch": 2.147152911068458, "grad_norm": 1.9746898366354038, "learning_rate": 1.8263332572639486e-05, "loss": 0.4498, "step": 3356 }, { "epoch": 2.147792706333973, "grad_norm": 1.8616570780132435, "learning_rate": 1.826216539655242e-05, "loss": 0.465, "step": 3357 }, { "epoch": 2.1484325015994883, "grad_norm": 1.9685286155110369, "learning_rate": 1.826099786570267e-05, "loss": 0.4669, "step": 3358 }, { "epoch": 2.149072296865003, "grad_norm": 1.8327917359894588, "learning_rate": 1.8259829980140367e-05, "loss": 0.4378, "step": 3359 }, { "epoch": 2.1497120921305184, "grad_norm": 1.9048584660842924, "learning_rate": 1.8258661739915658e-05, "loss": 0.4915, "step": 3360 }, { "epoch": 2.1503518873960332, "grad_norm": 2.057241434693828, "learning_rate": 1.8257493145078706e-05, "loss": 0.4178, "step": 3361 }, { "epoch": 2.1509916826615485, "grad_norm": 2.019133054088613, "learning_rate": 1.8256324195679687e-05, "loss": 0.4135, "step": 3362 }, { "epoch": 2.1516314779270633, "grad_norm": 1.925679186885445, "learning_rate": 1.8255154891768794e-05, "loss": 0.4622, "step": 3363 }, { "epoch": 2.1522712731925786, "grad_norm": 1.9089521726256433, "learning_rate": 1.8253985233396233e-05, "loss": 0.474, "step": 3364 }, { "epoch": 2.1529110684580934, "grad_norm": 2.0044875008917944, "learning_rate": 1.825281522061223e-05, "loss": 0.3977, "step": 3365 }, { "epoch": 2.1535508637236083, "grad_norm": 1.8375140189445642, "learning_rate": 1.825164485346702e-05, "loss": 0.4072, "step": 3366 }, { "epoch": 2.1541906589891235, "grad_norm": 2.0013198833998143, "learning_rate": 1.825047413201086e-05, "loss": 0.4479, "step": 3367 }, { "epoch": 2.1548304542546384, "grad_norm": 1.8295172457926931, "learning_rate": 1.824930305629402e-05, "loss": 0.4057, "step": 3368 }, { "epoch": 2.1554702495201536, "grad_norm": 2.019682956202068, "learning_rate": 1.8248131626366777e-05, "loss": 0.4771, "step": 3369 }, { "epoch": 2.1561100447856685, "grad_norm": 1.8929230095238527, "learning_rate": 1.8246959842279437e-05, "loss": 0.4613, "step": 3370 }, { "epoch": 2.1567498400511838, "grad_norm": 1.8104313918241997, "learning_rate": 1.824578770408231e-05, "loss": 0.4401, "step": 3371 }, { "epoch": 2.1573896353166986, "grad_norm": 1.7695515171238634, "learning_rate": 1.8244615211825728e-05, "loss": 0.4502, "step": 3372 }, { "epoch": 2.158029430582214, "grad_norm": 1.8001028703050026, "learning_rate": 1.8243442365560034e-05, "loss": 0.4582, "step": 3373 }, { "epoch": 2.1586692258477287, "grad_norm": 1.8715073773647983, "learning_rate": 1.8242269165335588e-05, "loss": 0.4364, "step": 3374 }, { "epoch": 2.159309021113244, "grad_norm": 1.8396897944281776, "learning_rate": 1.824109561120277e-05, "loss": 0.4486, "step": 3375 }, { "epoch": 2.159948816378759, "grad_norm": 2.1233954439959843, "learning_rate": 1.823992170321196e-05, "loss": 0.4844, "step": 3376 }, { "epoch": 2.1605886116442736, "grad_norm": 1.899458875112805, "learning_rate": 1.8238747441413576e-05, "loss": 0.4329, "step": 3377 }, { "epoch": 2.161228406909789, "grad_norm": 1.9216711173033794, "learning_rate": 1.823757282585803e-05, "loss": 0.4301, "step": 3378 }, { "epoch": 2.1618682021753037, "grad_norm": 1.963087027338093, "learning_rate": 1.823639785659576e-05, "loss": 0.4289, "step": 3379 }, { "epoch": 2.162507997440819, "grad_norm": 1.9640170842635838, "learning_rate": 1.8235222533677216e-05, "loss": 0.4313, "step": 3380 }, { "epoch": 2.163147792706334, "grad_norm": 1.9333696564127896, "learning_rate": 1.8234046857152868e-05, "loss": 0.4117, "step": 3381 }, { "epoch": 2.163787587971849, "grad_norm": 2.070207453959844, "learning_rate": 1.8232870827073196e-05, "loss": 0.4264, "step": 3382 }, { "epoch": 2.164427383237364, "grad_norm": 2.053053969521543, "learning_rate": 1.823169444348869e-05, "loss": 0.4194, "step": 3383 }, { "epoch": 2.165067178502879, "grad_norm": 2.0349350756308704, "learning_rate": 1.8230517706449873e-05, "loss": 0.438, "step": 3384 }, { "epoch": 2.165706973768394, "grad_norm": 1.9813793968307303, "learning_rate": 1.8229340616007263e-05, "loss": 0.4554, "step": 3385 }, { "epoch": 2.1663467690339093, "grad_norm": 2.0443583233956613, "learning_rate": 1.8228163172211412e-05, "loss": 0.5014, "step": 3386 }, { "epoch": 2.166986564299424, "grad_norm": 1.9572962770369169, "learning_rate": 1.8226985375112866e-05, "loss": 0.4474, "step": 3387 }, { "epoch": 2.167626359564939, "grad_norm": 1.8476599184881441, "learning_rate": 1.8225807224762205e-05, "loss": 0.4434, "step": 3388 }, { "epoch": 2.1682661548304543, "grad_norm": 1.9085672313622477, "learning_rate": 1.8224628721210013e-05, "loss": 0.4037, "step": 3389 }, { "epoch": 2.168905950095969, "grad_norm": 1.9643224250808666, "learning_rate": 1.8223449864506895e-05, "loss": 0.4331, "step": 3390 }, { "epoch": 2.1695457453614844, "grad_norm": 2.1275972293031264, "learning_rate": 1.8222270654703466e-05, "loss": 0.4444, "step": 3391 }, { "epoch": 2.170185540626999, "grad_norm": 2.1312663410346726, "learning_rate": 1.8221091091850365e-05, "loss": 0.4669, "step": 3392 }, { "epoch": 2.1708253358925145, "grad_norm": 1.972142196178118, "learning_rate": 1.8219911175998235e-05, "loss": 0.4144, "step": 3393 }, { "epoch": 2.1714651311580293, "grad_norm": 2.163362695822771, "learning_rate": 1.8218730907197743e-05, "loss": 0.5334, "step": 3394 }, { "epoch": 2.1721049264235446, "grad_norm": 2.0401763480497204, "learning_rate": 1.821755028549956e-05, "loss": 0.3948, "step": 3395 }, { "epoch": 2.1727447216890594, "grad_norm": 2.1729578103820444, "learning_rate": 1.821636931095439e-05, "loss": 0.4455, "step": 3396 }, { "epoch": 2.1733845169545747, "grad_norm": 2.0091909993062957, "learning_rate": 1.821518798361294e-05, "loss": 0.454, "step": 3397 }, { "epoch": 2.1740243122200895, "grad_norm": 2.1914537030701675, "learning_rate": 1.8214006303525927e-05, "loss": 0.4981, "step": 3398 }, { "epoch": 2.174664107485605, "grad_norm": 2.0179658854823552, "learning_rate": 1.8212824270744095e-05, "loss": 0.4452, "step": 3399 }, { "epoch": 2.1753039027511196, "grad_norm": 2.0742124187258812, "learning_rate": 1.8211641885318202e-05, "loss": 0.4772, "step": 3400 }, { "epoch": 2.175943698016635, "grad_norm": 2.0133205586599257, "learning_rate": 1.821045914729901e-05, "loss": 0.4038, "step": 3401 }, { "epoch": 2.1765834932821497, "grad_norm": 2.1089990905688123, "learning_rate": 1.820927605673731e-05, "loss": 0.4624, "step": 3402 }, { "epoch": 2.1772232885476646, "grad_norm": 1.8485154202019505, "learning_rate": 1.82080926136839e-05, "loss": 0.4105, "step": 3403 }, { "epoch": 2.17786308381318, "grad_norm": 2.042305957001792, "learning_rate": 1.8206908818189594e-05, "loss": 0.4427, "step": 3404 }, { "epoch": 2.1785028790786947, "grad_norm": 2.1112713934617267, "learning_rate": 1.8205724670305218e-05, "loss": 0.4622, "step": 3405 }, { "epoch": 2.17914267434421, "grad_norm": 1.867201025998322, "learning_rate": 1.8204540170081625e-05, "loss": 0.4086, "step": 3406 }, { "epoch": 2.1797824696097248, "grad_norm": 2.128876181445566, "learning_rate": 1.820335531756967e-05, "loss": 0.5022, "step": 3407 }, { "epoch": 2.18042226487524, "grad_norm": 2.005760023258529, "learning_rate": 1.8202170112820233e-05, "loss": 0.4593, "step": 3408 }, { "epoch": 2.181062060140755, "grad_norm": 1.950935223081763, "learning_rate": 1.82009845558842e-05, "loss": 0.4577, "step": 3409 }, { "epoch": 2.18170185540627, "grad_norm": 1.9535648993161467, "learning_rate": 1.819979864681248e-05, "loss": 0.4797, "step": 3410 }, { "epoch": 2.182341650671785, "grad_norm": 2.0461925513057286, "learning_rate": 1.8198612385655993e-05, "loss": 0.5177, "step": 3411 }, { "epoch": 2.1829814459373003, "grad_norm": 2.0399073576581443, "learning_rate": 1.819742577246567e-05, "loss": 0.4374, "step": 3412 }, { "epoch": 2.183621241202815, "grad_norm": 2.038423686422463, "learning_rate": 1.819623880729247e-05, "loss": 0.4303, "step": 3413 }, { "epoch": 2.18426103646833, "grad_norm": 1.9042163272371502, "learning_rate": 1.8195051490187356e-05, "loss": 0.4469, "step": 3414 }, { "epoch": 2.184900831733845, "grad_norm": 1.9666816405040768, "learning_rate": 1.8193863821201306e-05, "loss": 0.4743, "step": 3415 }, { "epoch": 2.18554062699936, "grad_norm": 1.982274650479486, "learning_rate": 1.8192675800385323e-05, "loss": 0.425, "step": 3416 }, { "epoch": 2.1861804222648753, "grad_norm": 2.183634584132621, "learning_rate": 1.8191487427790413e-05, "loss": 0.4498, "step": 3417 }, { "epoch": 2.18682021753039, "grad_norm": 2.0531667632646187, "learning_rate": 1.8190298703467604e-05, "loss": 0.4535, "step": 3418 }, { "epoch": 2.1874600127959054, "grad_norm": 1.912669765308496, "learning_rate": 1.818910962746794e-05, "loss": 0.465, "step": 3419 }, { "epoch": 2.1880998080614202, "grad_norm": 2.0051800593780755, "learning_rate": 1.8187920199842473e-05, "loss": 0.4404, "step": 3420 }, { "epoch": 2.1887396033269355, "grad_norm": 2.074873285838174, "learning_rate": 1.818673042064228e-05, "loss": 0.4366, "step": 3421 }, { "epoch": 2.1893793985924503, "grad_norm": 1.9520764106368198, "learning_rate": 1.8185540289918443e-05, "loss": 0.4837, "step": 3422 }, { "epoch": 2.1900191938579656, "grad_norm": 2.0225588705581385, "learning_rate": 1.818434980772207e-05, "loss": 0.4449, "step": 3423 }, { "epoch": 2.1906589891234804, "grad_norm": 1.9342771156805671, "learning_rate": 1.8183158974104275e-05, "loss": 0.4708, "step": 3424 }, { "epoch": 2.1912987843889957, "grad_norm": 1.9923396082474771, "learning_rate": 1.8181967789116188e-05, "loss": 0.4364, "step": 3425 }, { "epoch": 2.1919385796545106, "grad_norm": 1.9912337577026535, "learning_rate": 1.8180776252808958e-05, "loss": 0.4375, "step": 3426 }, { "epoch": 2.1925783749200254, "grad_norm": 2.103416545739382, "learning_rate": 1.817958436523375e-05, "loss": 0.4114, "step": 3427 }, { "epoch": 2.1932181701855407, "grad_norm": 1.8640226554895665, "learning_rate": 1.817839212644174e-05, "loss": 0.4313, "step": 3428 }, { "epoch": 2.1938579654510555, "grad_norm": 1.7954666534660437, "learning_rate": 1.8177199536484123e-05, "loss": 0.4401, "step": 3429 }, { "epoch": 2.1944977607165708, "grad_norm": 2.2619687924778282, "learning_rate": 1.81760065954121e-05, "loss": 0.4609, "step": 3430 }, { "epoch": 2.1951375559820856, "grad_norm": 1.9415540717858528, "learning_rate": 1.81748133032769e-05, "loss": 0.444, "step": 3431 }, { "epoch": 2.195777351247601, "grad_norm": 1.8454429503168843, "learning_rate": 1.8173619660129754e-05, "loss": 0.4564, "step": 3432 }, { "epoch": 2.1964171465131157, "grad_norm": 1.908199559596295, "learning_rate": 1.8172425666021926e-05, "loss": 0.4597, "step": 3433 }, { "epoch": 2.197056941778631, "grad_norm": 2.024688965387357, "learning_rate": 1.8171231321004675e-05, "loss": 0.4696, "step": 3434 }, { "epoch": 2.197696737044146, "grad_norm": 2.1527210149463496, "learning_rate": 1.8170036625129284e-05, "loss": 0.458, "step": 3435 }, { "epoch": 2.198336532309661, "grad_norm": 1.830771717940031, "learning_rate": 1.816884157844706e-05, "loss": 0.3954, "step": 3436 }, { "epoch": 2.198976327575176, "grad_norm": 1.9705449872807608, "learning_rate": 1.8167646181009302e-05, "loss": 0.4623, "step": 3437 }, { "epoch": 2.199616122840691, "grad_norm": 2.1420068912812384, "learning_rate": 1.816645043286735e-05, "loss": 0.4654, "step": 3438 }, { "epoch": 2.200255918106206, "grad_norm": 2.051723300727396, "learning_rate": 1.8165254334072543e-05, "loss": 0.4616, "step": 3439 }, { "epoch": 2.200895713371721, "grad_norm": 2.041954928582068, "learning_rate": 1.816405788467624e-05, "loss": 0.4479, "step": 3440 }, { "epoch": 2.201535508637236, "grad_norm": 1.849690912119803, "learning_rate": 1.8162861084729816e-05, "loss": 0.4122, "step": 3441 }, { "epoch": 2.202175303902751, "grad_norm": 1.984635486582728, "learning_rate": 1.8161663934284654e-05, "loss": 0.4512, "step": 3442 }, { "epoch": 2.2028150991682662, "grad_norm": 1.8422445938007028, "learning_rate": 1.8160466433392163e-05, "loss": 0.4382, "step": 3443 }, { "epoch": 2.203454894433781, "grad_norm": 1.8938556761546652, "learning_rate": 1.815926858210376e-05, "loss": 0.4159, "step": 3444 }, { "epoch": 2.2040946896992963, "grad_norm": 1.9066524694752303, "learning_rate": 1.815807038047088e-05, "loss": 0.4185, "step": 3445 }, { "epoch": 2.204734484964811, "grad_norm": 1.9738628519752317, "learning_rate": 1.8156871828544965e-05, "loss": 0.4321, "step": 3446 }, { "epoch": 2.2053742802303264, "grad_norm": 1.8960407914336028, "learning_rate": 1.8155672926377488e-05, "loss": 0.4226, "step": 3447 }, { "epoch": 2.2060140754958413, "grad_norm": 1.9884292078965202, "learning_rate": 1.8154473674019922e-05, "loss": 0.4858, "step": 3448 }, { "epoch": 2.2066538707613566, "grad_norm": 1.940010206215438, "learning_rate": 1.815327407152376e-05, "loss": 0.4294, "step": 3449 }, { "epoch": 2.2072936660268714, "grad_norm": 2.240988495368133, "learning_rate": 1.8152074118940516e-05, "loss": 0.461, "step": 3450 }, { "epoch": 2.207933461292386, "grad_norm": 1.94866659235838, "learning_rate": 1.815087381632171e-05, "loss": 0.4874, "step": 3451 }, { "epoch": 2.2085732565579015, "grad_norm": 1.9309691410341763, "learning_rate": 1.814967316371888e-05, "loss": 0.4563, "step": 3452 }, { "epoch": 2.2092130518234163, "grad_norm": 1.8488164120949218, "learning_rate": 1.8148472161183584e-05, "loss": 0.4273, "step": 3453 }, { "epoch": 2.2098528470889316, "grad_norm": 1.9800472423518654, "learning_rate": 1.8147270808767386e-05, "loss": 0.4504, "step": 3454 }, { "epoch": 2.2104926423544464, "grad_norm": 2.2267816377466554, "learning_rate": 1.8146069106521875e-05, "loss": 0.4724, "step": 3455 }, { "epoch": 2.2111324376199617, "grad_norm": 1.9057253196398645, "learning_rate": 1.8144867054498645e-05, "loss": 0.3957, "step": 3456 }, { "epoch": 2.2117722328854765, "grad_norm": 1.8484837825019211, "learning_rate": 1.8143664652749314e-05, "loss": 0.3858, "step": 3457 }, { "epoch": 2.212412028150992, "grad_norm": 1.8780532546409825, "learning_rate": 1.8142461901325507e-05, "loss": 0.4705, "step": 3458 }, { "epoch": 2.2130518234165066, "grad_norm": 2.0178326767035752, "learning_rate": 1.814125880027887e-05, "loss": 0.4537, "step": 3459 }, { "epoch": 2.213691618682022, "grad_norm": 1.9672185046493826, "learning_rate": 1.8140055349661065e-05, "loss": 0.4813, "step": 3460 }, { "epoch": 2.2143314139475367, "grad_norm": 1.8662230486917915, "learning_rate": 1.8138851549523764e-05, "loss": 0.4302, "step": 3461 }, { "epoch": 2.214971209213052, "grad_norm": 1.9638212846238186, "learning_rate": 1.8137647399918653e-05, "loss": 0.4434, "step": 3462 }, { "epoch": 2.215611004478567, "grad_norm": 1.8626586234730118, "learning_rate": 1.813644290089744e-05, "loss": 0.4435, "step": 3463 }, { "epoch": 2.216250799744082, "grad_norm": 1.8544933789889975, "learning_rate": 1.813523805251184e-05, "loss": 0.4662, "step": 3464 }, { "epoch": 2.216890595009597, "grad_norm": 1.8121773103261345, "learning_rate": 1.8134032854813592e-05, "loss": 0.4235, "step": 3465 }, { "epoch": 2.217530390275112, "grad_norm": 1.9318171192977875, "learning_rate": 1.813282730785444e-05, "loss": 0.4472, "step": 3466 }, { "epoch": 2.218170185540627, "grad_norm": 1.8235654315897771, "learning_rate": 1.8131621411686153e-05, "loss": 0.4189, "step": 3467 }, { "epoch": 2.218809980806142, "grad_norm": 1.8926388588651668, "learning_rate": 1.8130415166360506e-05, "loss": 0.3918, "step": 3468 }, { "epoch": 2.219449776071657, "grad_norm": 1.9844183157641646, "learning_rate": 1.8129208571929295e-05, "loss": 0.4089, "step": 3469 }, { "epoch": 2.220089571337172, "grad_norm": 2.2196616883547677, "learning_rate": 1.812800162844433e-05, "loss": 0.4421, "step": 3470 }, { "epoch": 2.2207293666026873, "grad_norm": 2.0549711849208063, "learning_rate": 1.8126794335957432e-05, "loss": 0.4204, "step": 3471 }, { "epoch": 2.221369161868202, "grad_norm": 1.8723777478707995, "learning_rate": 1.8125586694520443e-05, "loss": 0.3777, "step": 3472 }, { "epoch": 2.2220089571337174, "grad_norm": 1.9970715779179198, "learning_rate": 1.8124378704185213e-05, "loss": 0.4721, "step": 3473 }, { "epoch": 2.222648752399232, "grad_norm": 1.7808793560826557, "learning_rate": 1.8123170365003618e-05, "loss": 0.3938, "step": 3474 }, { "epoch": 2.2232885476647475, "grad_norm": 2.0437121377944476, "learning_rate": 1.8121961677027533e-05, "loss": 0.4554, "step": 3475 }, { "epoch": 2.2239283429302623, "grad_norm": 2.124878677864941, "learning_rate": 1.8120752640308864e-05, "loss": 0.4477, "step": 3476 }, { "epoch": 2.224568138195777, "grad_norm": 2.1819840836697706, "learning_rate": 1.811954325489952e-05, "loss": 0.4183, "step": 3477 }, { "epoch": 2.2252079334612924, "grad_norm": 2.008931905501649, "learning_rate": 1.8118333520851438e-05, "loss": 0.4389, "step": 3478 }, { "epoch": 2.2258477287268073, "grad_norm": 1.9048395347592293, "learning_rate": 1.8117123438216547e-05, "loss": 0.4475, "step": 3479 }, { "epoch": 2.2264875239923225, "grad_norm": 2.0684924972442524, "learning_rate": 1.8115913007046822e-05, "loss": 0.4135, "step": 3480 }, { "epoch": 2.2271273192578374, "grad_norm": 2.273222092796233, "learning_rate": 1.811470222739423e-05, "loss": 0.5096, "step": 3481 }, { "epoch": 2.2277671145233526, "grad_norm": 2.158205556360603, "learning_rate": 1.8113491099310755e-05, "loss": 0.4528, "step": 3482 }, { "epoch": 2.2284069097888675, "grad_norm": 1.8031656011392676, "learning_rate": 1.8112279622848407e-05, "loss": 0.4256, "step": 3483 }, { "epoch": 2.2290467050543827, "grad_norm": 1.7838922683643454, "learning_rate": 1.81110677980592e-05, "loss": 0.4473, "step": 3484 }, { "epoch": 2.2296865003198976, "grad_norm": 2.071887512492206, "learning_rate": 1.8109855624995174e-05, "loss": 0.5088, "step": 3485 }, { "epoch": 2.230326295585413, "grad_norm": 2.0032473142292004, "learning_rate": 1.8108643103708372e-05, "loss": 0.4313, "step": 3486 }, { "epoch": 2.2309660908509277, "grad_norm": 2.300603423747754, "learning_rate": 1.810743023425086e-05, "loss": 0.4665, "step": 3487 }, { "epoch": 2.2316058861164425, "grad_norm": 2.1127520149260914, "learning_rate": 1.8106217016674717e-05, "loss": 0.4743, "step": 3488 }, { "epoch": 2.232245681381958, "grad_norm": 1.8296734218421007, "learning_rate": 1.810500345103203e-05, "loss": 0.4096, "step": 3489 }, { "epoch": 2.2328854766474726, "grad_norm": 1.9297134146031856, "learning_rate": 1.8103789537374918e-05, "loss": 0.4486, "step": 3490 }, { "epoch": 2.233525271912988, "grad_norm": 2.073700697711357, "learning_rate": 1.8102575275755496e-05, "loss": 0.4191, "step": 3491 }, { "epoch": 2.2341650671785027, "grad_norm": 2.157194059380139, "learning_rate": 1.8101360666225907e-05, "loss": 0.4876, "step": 3492 }, { "epoch": 2.234804862444018, "grad_norm": 2.0722520629492833, "learning_rate": 1.8100145708838298e-05, "loss": 0.4619, "step": 3493 }, { "epoch": 2.235444657709533, "grad_norm": 1.9325376535762877, "learning_rate": 1.8098930403644846e-05, "loss": 0.4248, "step": 3494 }, { "epoch": 2.236084452975048, "grad_norm": 1.9406096358443459, "learning_rate": 1.8097714750697725e-05, "loss": 0.4115, "step": 3495 }, { "epoch": 2.236724248240563, "grad_norm": 2.035603701534115, "learning_rate": 1.809649875004914e-05, "loss": 0.4548, "step": 3496 }, { "epoch": 2.237364043506078, "grad_norm": 1.8799339875551104, "learning_rate": 1.8095282401751305e-05, "loss": 0.4706, "step": 3497 }, { "epoch": 2.238003838771593, "grad_norm": 1.846489132928023, "learning_rate": 1.809406570585644e-05, "loss": 0.4199, "step": 3498 }, { "epoch": 2.2386436340371083, "grad_norm": 1.9617697659167637, "learning_rate": 1.809284866241679e-05, "loss": 0.4966, "step": 3499 }, { "epoch": 2.239283429302623, "grad_norm": 2.2073473934433108, "learning_rate": 1.8091631271484617e-05, "loss": 0.4582, "step": 3500 }, { "epoch": 2.2399232245681384, "grad_norm": 1.9750071298474003, "learning_rate": 1.809041353311219e-05, "loss": 0.4528, "step": 3501 }, { "epoch": 2.2405630198336532, "grad_norm": 1.8862776851802345, "learning_rate": 1.80891954473518e-05, "loss": 0.414, "step": 3502 }, { "epoch": 2.241202815099168, "grad_norm": 2.0091069993304465, "learning_rate": 1.8087977014255745e-05, "loss": 0.4761, "step": 3503 }, { "epoch": 2.2418426103646834, "grad_norm": 1.778700610477075, "learning_rate": 1.8086758233876343e-05, "loss": 0.4368, "step": 3504 }, { "epoch": 2.242482405630198, "grad_norm": 1.948020145114326, "learning_rate": 1.808553910626593e-05, "loss": 0.4575, "step": 3505 }, { "epoch": 2.2431222008957135, "grad_norm": 1.9935055986158006, "learning_rate": 1.808431963147685e-05, "loss": 0.4667, "step": 3506 }, { "epoch": 2.2437619961612283, "grad_norm": 2.1504722225412896, "learning_rate": 1.8083099809561467e-05, "loss": 0.4845, "step": 3507 }, { "epoch": 2.2444017914267436, "grad_norm": 1.937969331620018, "learning_rate": 1.8081879640572156e-05, "loss": 0.4341, "step": 3508 }, { "epoch": 2.2450415866922584, "grad_norm": 1.936080258173945, "learning_rate": 1.808065912456131e-05, "loss": 0.427, "step": 3509 }, { "epoch": 2.2456813819577737, "grad_norm": 1.9389915937532098, "learning_rate": 1.8079438261581334e-05, "loss": 0.4268, "step": 3510 }, { "epoch": 2.2463211772232885, "grad_norm": 2.0057922993916746, "learning_rate": 1.8078217051684652e-05, "loss": 0.4689, "step": 3511 }, { "epoch": 2.2469609724888038, "grad_norm": 2.0971930244783583, "learning_rate": 1.8076995494923702e-05, "loss": 0.4471, "step": 3512 }, { "epoch": 2.2476007677543186, "grad_norm": 1.9756521736618593, "learning_rate": 1.8075773591350933e-05, "loss": 0.4771, "step": 3513 }, { "epoch": 2.2482405630198334, "grad_norm": 1.9877860723697292, "learning_rate": 1.807455134101881e-05, "loss": 0.4302, "step": 3514 }, { "epoch": 2.2488803582853487, "grad_norm": 2.02919244060732, "learning_rate": 1.8073328743979814e-05, "loss": 0.4283, "step": 3515 }, { "epoch": 2.2495201535508635, "grad_norm": 1.9190266127751654, "learning_rate": 1.8072105800286445e-05, "loss": 0.4259, "step": 3516 }, { "epoch": 2.250159948816379, "grad_norm": 1.935821864581953, "learning_rate": 1.8070882509991213e-05, "loss": 0.432, "step": 3517 }, { "epoch": 2.2507997440818936, "grad_norm": 1.909436051392451, "learning_rate": 1.8069658873146644e-05, "loss": 0.4565, "step": 3518 }, { "epoch": 2.251439539347409, "grad_norm": 2.092490661382176, "learning_rate": 1.8068434889805276e-05, "loss": 0.451, "step": 3519 }, { "epoch": 2.2520793346129238, "grad_norm": 2.112497146472536, "learning_rate": 1.8067210560019664e-05, "loss": 0.4527, "step": 3520 }, { "epoch": 2.252719129878439, "grad_norm": 1.874184326665986, "learning_rate": 1.8065985883842385e-05, "loss": 0.4306, "step": 3521 }, { "epoch": 2.253358925143954, "grad_norm": 1.9512715547561637, "learning_rate": 1.806476086132602e-05, "loss": 0.4176, "step": 3522 }, { "epoch": 2.253998720409469, "grad_norm": 2.120198677043199, "learning_rate": 1.8063535492523167e-05, "loss": 0.5238, "step": 3523 }, { "epoch": 2.254638515674984, "grad_norm": 1.8661490938230911, "learning_rate": 1.8062309777486446e-05, "loss": 0.3934, "step": 3524 }, { "epoch": 2.255278310940499, "grad_norm": 2.0074539738960855, "learning_rate": 1.8061083716268484e-05, "loss": 0.4529, "step": 3525 }, { "epoch": 2.255918106206014, "grad_norm": 1.8417800889283278, "learning_rate": 1.805985730892193e-05, "loss": 0.4455, "step": 3526 }, { "epoch": 2.2565579014715293, "grad_norm": 2.0279311183207045, "learning_rate": 1.8058630555499435e-05, "loss": 0.4127, "step": 3527 }, { "epoch": 2.257197696737044, "grad_norm": 2.112863771459073, "learning_rate": 1.8057403456053685e-05, "loss": 0.475, "step": 3528 }, { "epoch": 2.257837492002559, "grad_norm": 2.1376454370560443, "learning_rate": 1.805617601063736e-05, "loss": 0.497, "step": 3529 }, { "epoch": 2.2584772872680743, "grad_norm": 1.9088698228994447, "learning_rate": 1.805494821930317e-05, "loss": 0.4034, "step": 3530 }, { "epoch": 2.259117082533589, "grad_norm": 1.8587925104515652, "learning_rate": 1.805372008210383e-05, "loss": 0.4237, "step": 3531 }, { "epoch": 2.2597568777991044, "grad_norm": 1.9534211562492645, "learning_rate": 1.8052491599092078e-05, "loss": 0.4824, "step": 3532 }, { "epoch": 2.260396673064619, "grad_norm": 1.952686442758173, "learning_rate": 1.805126277032066e-05, "loss": 0.4616, "step": 3533 }, { "epoch": 2.2610364683301345, "grad_norm": 1.9098834540455512, "learning_rate": 1.8050033595842337e-05, "loss": 0.4601, "step": 3534 }, { "epoch": 2.2616762635956493, "grad_norm": 1.963988762444253, "learning_rate": 1.8048804075709895e-05, "loss": 0.4457, "step": 3535 }, { "epoch": 2.2623160588611646, "grad_norm": 1.989128799776346, "learning_rate": 1.8047574209976126e-05, "loss": 0.4345, "step": 3536 }, { "epoch": 2.2629558541266794, "grad_norm": 1.960028417975544, "learning_rate": 1.8046343998693832e-05, "loss": 0.4183, "step": 3537 }, { "epoch": 2.2635956493921947, "grad_norm": 2.107038232774411, "learning_rate": 1.8045113441915842e-05, "loss": 0.4264, "step": 3538 }, { "epoch": 2.2642354446577095, "grad_norm": 2.066383848551461, "learning_rate": 1.8043882539694994e-05, "loss": 0.4428, "step": 3539 }, { "epoch": 2.2648752399232244, "grad_norm": 1.8174865470464248, "learning_rate": 1.8042651292084135e-05, "loss": 0.4365, "step": 3540 }, { "epoch": 2.2655150351887396, "grad_norm": 2.040918624889689, "learning_rate": 1.8041419699136134e-05, "loss": 0.4816, "step": 3541 }, { "epoch": 2.2661548304542545, "grad_norm": 2.0091174342331826, "learning_rate": 1.804018776090388e-05, "loss": 0.4191, "step": 3542 }, { "epoch": 2.2667946257197698, "grad_norm": 1.9670760294460672, "learning_rate": 1.8038955477440266e-05, "loss": 0.423, "step": 3543 }, { "epoch": 2.2674344209852846, "grad_norm": 1.9333249736722233, "learning_rate": 1.8037722848798203e-05, "loss": 0.3944, "step": 3544 }, { "epoch": 2.2680742162508, "grad_norm": 2.0022801505244785, "learning_rate": 1.803648987503062e-05, "loss": 0.4366, "step": 3545 }, { "epoch": 2.2687140115163147, "grad_norm": 2.3882545594249414, "learning_rate": 1.8035256556190456e-05, "loss": 0.4272, "step": 3546 }, { "epoch": 2.26935380678183, "grad_norm": 1.9630999245067025, "learning_rate": 1.803402289233067e-05, "loss": 0.4264, "step": 3547 }, { "epoch": 2.269993602047345, "grad_norm": 1.8639388674387867, "learning_rate": 1.8032788883504233e-05, "loss": 0.4441, "step": 3548 }, { "epoch": 2.27063339731286, "grad_norm": 2.1071331842478913, "learning_rate": 1.803155452976413e-05, "loss": 0.4524, "step": 3549 }, { "epoch": 2.271273192578375, "grad_norm": 1.7820259536770773, "learning_rate": 1.8030319831163363e-05, "loss": 0.4473, "step": 3550 }, { "epoch": 2.2719129878438897, "grad_norm": 1.9165853294304234, "learning_rate": 1.8029084787754945e-05, "loss": 0.4106, "step": 3551 }, { "epoch": 2.272552783109405, "grad_norm": 1.9752746071457754, "learning_rate": 1.802784939959191e-05, "loss": 0.4173, "step": 3552 }, { "epoch": 2.27319257837492, "grad_norm": 2.015210133688076, "learning_rate": 1.80266136667273e-05, "loss": 0.4651, "step": 3553 }, { "epoch": 2.273832373640435, "grad_norm": 1.8364701279390416, "learning_rate": 1.8025377589214183e-05, "loss": 0.3773, "step": 3554 }, { "epoch": 2.27447216890595, "grad_norm": 1.978072026155289, "learning_rate": 1.8024141167105626e-05, "loss": 0.4287, "step": 3555 }, { "epoch": 2.275111964171465, "grad_norm": 1.8251816839586708, "learning_rate": 1.802290440045472e-05, "loss": 0.4323, "step": 3556 }, { "epoch": 2.27575175943698, "grad_norm": 1.8560309546437055, "learning_rate": 1.802166728931457e-05, "loss": 0.4134, "step": 3557 }, { "epoch": 2.2763915547024953, "grad_norm": 2.1987816359555126, "learning_rate": 1.8020429833738298e-05, "loss": 0.4153, "step": 3558 }, { "epoch": 2.27703134996801, "grad_norm": 2.1123343081238337, "learning_rate": 1.8019192033779034e-05, "loss": 0.4723, "step": 3559 }, { "epoch": 2.2776711452335254, "grad_norm": 2.086124181712707, "learning_rate": 1.801795388948993e-05, "loss": 0.4415, "step": 3560 }, { "epoch": 2.2783109404990403, "grad_norm": 1.9458553727380692, "learning_rate": 1.801671540092415e-05, "loss": 0.4271, "step": 3561 }, { "epoch": 2.2789507357645555, "grad_norm": 1.7775540472964935, "learning_rate": 1.8015476568134872e-05, "loss": 0.4153, "step": 3562 }, { "epoch": 2.2795905310300704, "grad_norm": 2.206056677133341, "learning_rate": 1.8014237391175284e-05, "loss": 0.4851, "step": 3563 }, { "epoch": 2.2802303262955856, "grad_norm": 1.9399900327052457, "learning_rate": 1.8012997870098603e-05, "loss": 0.4211, "step": 3564 }, { "epoch": 2.2808701215611005, "grad_norm": 1.9292125163080522, "learning_rate": 1.8011758004958047e-05, "loss": 0.4546, "step": 3565 }, { "epoch": 2.2815099168266153, "grad_norm": 2.0393307046343447, "learning_rate": 1.801051779580685e-05, "loss": 0.4616, "step": 3566 }, { "epoch": 2.2821497120921306, "grad_norm": 2.040444286876987, "learning_rate": 1.8009277242698274e-05, "loss": 0.3896, "step": 3567 }, { "epoch": 2.2827895073576454, "grad_norm": 1.9341574298882491, "learning_rate": 1.800803634568558e-05, "loss": 0.4155, "step": 3568 }, { "epoch": 2.2834293026231607, "grad_norm": 1.8414420430597211, "learning_rate": 1.8006795104822046e-05, "loss": 0.4176, "step": 3569 }, { "epoch": 2.2840690978886755, "grad_norm": 1.8051895047056363, "learning_rate": 1.800555352016098e-05, "loss": 0.4409, "step": 3570 }, { "epoch": 2.284708893154191, "grad_norm": 1.8128678455126186, "learning_rate": 1.8004311591755682e-05, "loss": 0.44, "step": 3571 }, { "epoch": 2.2853486884197056, "grad_norm": 2.093233257661318, "learning_rate": 1.800306931965948e-05, "loss": 0.4531, "step": 3572 }, { "epoch": 2.285988483685221, "grad_norm": 1.853990675069797, "learning_rate": 1.8001826703925723e-05, "loss": 0.4129, "step": 3573 }, { "epoch": 2.2866282789507357, "grad_norm": 1.886499211022605, "learning_rate": 1.800058374460776e-05, "loss": 0.4181, "step": 3574 }, { "epoch": 2.287268074216251, "grad_norm": 1.8356199137617282, "learning_rate": 1.7999340441758963e-05, "loss": 0.388, "step": 3575 }, { "epoch": 2.287907869481766, "grad_norm": 1.9190775025003026, "learning_rate": 1.799809679543272e-05, "loss": 0.4597, "step": 3576 }, { "epoch": 2.2885476647472807, "grad_norm": 2.110509568938325, "learning_rate": 1.7996852805682424e-05, "loss": 0.4722, "step": 3577 }, { "epoch": 2.289187460012796, "grad_norm": 1.7390544590369885, "learning_rate": 1.7995608472561493e-05, "loss": 0.3936, "step": 3578 }, { "epoch": 2.2898272552783108, "grad_norm": 1.9484131038886623, "learning_rate": 1.799436379612336e-05, "loss": 0.4358, "step": 3579 }, { "epoch": 2.290467050543826, "grad_norm": 1.7965781892561903, "learning_rate": 1.7993118776421467e-05, "loss": 0.4273, "step": 3580 }, { "epoch": 2.291106845809341, "grad_norm": 2.0027752934017293, "learning_rate": 1.7991873413509266e-05, "loss": 0.5007, "step": 3581 }, { "epoch": 2.291746641074856, "grad_norm": 1.7590213866737126, "learning_rate": 1.7990627707440245e-05, "loss": 0.4057, "step": 3582 }, { "epoch": 2.292386436340371, "grad_norm": 1.766914759746092, "learning_rate": 1.7989381658267883e-05, "loss": 0.4006, "step": 3583 }, { "epoch": 2.2930262316058863, "grad_norm": 1.9997812378397921, "learning_rate": 1.798813526604568e-05, "loss": 0.4373, "step": 3584 }, { "epoch": 2.293666026871401, "grad_norm": 1.926845845485588, "learning_rate": 1.7986888530827162e-05, "loss": 0.4955, "step": 3585 }, { "epoch": 2.2943058221369164, "grad_norm": 1.950014745688012, "learning_rate": 1.7985641452665857e-05, "loss": 0.4854, "step": 3586 }, { "epoch": 2.294945617402431, "grad_norm": 1.914367471160276, "learning_rate": 1.7984394031615315e-05, "loss": 0.4444, "step": 3587 }, { "epoch": 2.295585412667946, "grad_norm": 1.9759472953941617, "learning_rate": 1.7983146267729093e-05, "loss": 0.4436, "step": 3588 }, { "epoch": 2.2962252079334613, "grad_norm": 1.8441244443915914, "learning_rate": 1.7981898161060776e-05, "loss": 0.4475, "step": 3589 }, { "epoch": 2.2968650031989766, "grad_norm": 2.0480148390337094, "learning_rate": 1.7980649711663946e-05, "loss": 0.4504, "step": 3590 }, { "epoch": 2.2975047984644914, "grad_norm": 1.999012322270647, "learning_rate": 1.7979400919592218e-05, "loss": 0.4774, "step": 3591 }, { "epoch": 2.2981445937300062, "grad_norm": 1.8299893671639906, "learning_rate": 1.7978151784899207e-05, "loss": 0.4158, "step": 3592 }, { "epoch": 2.2987843889955215, "grad_norm": 1.9249931813794148, "learning_rate": 1.797690230763855e-05, "loss": 0.4867, "step": 3593 }, { "epoch": 2.2994241842610363, "grad_norm": 1.8041385139335986, "learning_rate": 1.7975652487863896e-05, "loss": 0.4281, "step": 3594 }, { "epoch": 2.3000639795265516, "grad_norm": 1.711034047186077, "learning_rate": 1.7974402325628915e-05, "loss": 0.4201, "step": 3595 }, { "epoch": 2.3007037747920664, "grad_norm": 1.9388347057791175, "learning_rate": 1.7973151820987282e-05, "loss": 0.4348, "step": 3596 }, { "epoch": 2.3013435700575817, "grad_norm": 1.9682031120068524, "learning_rate": 1.7971900973992693e-05, "loss": 0.418, "step": 3597 }, { "epoch": 2.3019833653230966, "grad_norm": 1.9087667266219188, "learning_rate": 1.797064978469886e-05, "loss": 0.3892, "step": 3598 }, { "epoch": 2.302623160588612, "grad_norm": 2.0470497507370653, "learning_rate": 1.79693982531595e-05, "loss": 0.4272, "step": 3599 }, { "epoch": 2.3032629558541267, "grad_norm": 1.9073046646521865, "learning_rate": 1.7968146379428355e-05, "loss": 0.3995, "step": 3600 }, { "epoch": 2.303902751119642, "grad_norm": 1.973222588265547, "learning_rate": 1.796689416355918e-05, "loss": 0.4412, "step": 3601 }, { "epoch": 2.3045425463851568, "grad_norm": 2.0350564592467806, "learning_rate": 1.7965641605605743e-05, "loss": 0.4202, "step": 3602 }, { "epoch": 2.3051823416506716, "grad_norm": 2.017200399912146, "learning_rate": 1.7964388705621824e-05, "loss": 0.4468, "step": 3603 }, { "epoch": 2.305822136916187, "grad_norm": 2.0397152166185197, "learning_rate": 1.7963135463661224e-05, "loss": 0.4291, "step": 3604 }, { "epoch": 2.3064619321817017, "grad_norm": 1.911880766482005, "learning_rate": 1.7961881879777746e-05, "loss": 0.4359, "step": 3605 }, { "epoch": 2.307101727447217, "grad_norm": 1.9104331288076728, "learning_rate": 1.796062795402523e-05, "loss": 0.4573, "step": 3606 }, { "epoch": 2.307741522712732, "grad_norm": 2.1482206508024504, "learning_rate": 1.795937368645751e-05, "loss": 0.4917, "step": 3607 }, { "epoch": 2.308381317978247, "grad_norm": 1.899411113509105, "learning_rate": 1.7958119077128437e-05, "loss": 0.4323, "step": 3608 }, { "epoch": 2.309021113243762, "grad_norm": 1.8049848757745435, "learning_rate": 1.7956864126091892e-05, "loss": 0.4128, "step": 3609 }, { "epoch": 2.309660908509277, "grad_norm": 1.8473135069040725, "learning_rate": 1.7955608833401754e-05, "loss": 0.47, "step": 3610 }, { "epoch": 2.310300703774792, "grad_norm": 1.9425809255003936, "learning_rate": 1.7954353199111923e-05, "loss": 0.4323, "step": 3611 }, { "epoch": 2.3109404990403073, "grad_norm": 1.8407258053154685, "learning_rate": 1.7953097223276317e-05, "loss": 0.4491, "step": 3612 }, { "epoch": 2.311580294305822, "grad_norm": 1.9049148778622778, "learning_rate": 1.7951840905948865e-05, "loss": 0.4341, "step": 3613 }, { "epoch": 2.312220089571337, "grad_norm": 1.78339868803954, "learning_rate": 1.795058424718351e-05, "loss": 0.3828, "step": 3614 }, { "epoch": 2.3128598848368522, "grad_norm": 1.9605142950444236, "learning_rate": 1.7949327247034206e-05, "loss": 0.4586, "step": 3615 }, { "epoch": 2.313499680102367, "grad_norm": 1.9459305897453703, "learning_rate": 1.794806990555494e-05, "loss": 0.4316, "step": 3616 }, { "epoch": 2.3141394753678823, "grad_norm": 1.9306522694059898, "learning_rate": 1.7946812222799682e-05, "loss": 0.4139, "step": 3617 }, { "epoch": 2.314779270633397, "grad_norm": 2.3918415215320157, "learning_rate": 1.7945554198822445e-05, "loss": 0.4586, "step": 3618 }, { "epoch": 2.3154190658989124, "grad_norm": 1.8846245062505707, "learning_rate": 1.7944295833677247e-05, "loss": 0.4133, "step": 3619 }, { "epoch": 2.3160588611644273, "grad_norm": 1.8968277337547348, "learning_rate": 1.7943037127418115e-05, "loss": 0.4385, "step": 3620 }, { "epoch": 2.3166986564299425, "grad_norm": 1.8149919062979902, "learning_rate": 1.79417780800991e-05, "loss": 0.4526, "step": 3621 }, { "epoch": 2.3173384516954574, "grad_norm": 1.8182348431825992, "learning_rate": 1.794051869177426e-05, "loss": 0.4374, "step": 3622 }, { "epoch": 2.3179782469609727, "grad_norm": 1.9201256389689585, "learning_rate": 1.7939258962497675e-05, "loss": 0.4363, "step": 3623 }, { "epoch": 2.3186180422264875, "grad_norm": 1.9271153944766564, "learning_rate": 1.793799889232343e-05, "loss": 0.3864, "step": 3624 }, { "epoch": 2.3192578374920023, "grad_norm": 2.1157185892690236, "learning_rate": 1.7936738481305636e-05, "loss": 0.4515, "step": 3625 }, { "epoch": 2.3198976327575176, "grad_norm": 1.8589451642010741, "learning_rate": 1.793547772949841e-05, "loss": 0.4019, "step": 3626 }, { "epoch": 2.320537428023033, "grad_norm": 1.9581816327374237, "learning_rate": 1.7934216636955883e-05, "loss": 0.4103, "step": 3627 }, { "epoch": 2.3211772232885477, "grad_norm": 1.8998119041311878, "learning_rate": 1.7932955203732213e-05, "loss": 0.4217, "step": 3628 }, { "epoch": 2.3218170185540625, "grad_norm": 2.1565805641994116, "learning_rate": 1.793169342988155e-05, "loss": 0.4638, "step": 3629 }, { "epoch": 2.322456813819578, "grad_norm": 1.94405255106242, "learning_rate": 1.793043131545809e-05, "loss": 0.4592, "step": 3630 }, { "epoch": 2.3230966090850926, "grad_norm": 1.7453410264581113, "learning_rate": 1.792916886051601e-05, "loss": 0.4519, "step": 3631 }, { "epoch": 2.323736404350608, "grad_norm": 1.9109084651667758, "learning_rate": 1.7927906065109527e-05, "loss": 0.39, "step": 3632 }, { "epoch": 2.3243761996161227, "grad_norm": 1.826366029440867, "learning_rate": 1.792664292929286e-05, "loss": 0.4224, "step": 3633 }, { "epoch": 2.325015994881638, "grad_norm": 1.8067313810386363, "learning_rate": 1.7925379453120243e-05, "loss": 0.4668, "step": 3634 }, { "epoch": 2.325655790147153, "grad_norm": 1.9583166196172845, "learning_rate": 1.7924115636645933e-05, "loss": 0.4255, "step": 3635 }, { "epoch": 2.326295585412668, "grad_norm": 2.1424722604613784, "learning_rate": 1.7922851479924196e-05, "loss": 0.4481, "step": 3636 }, { "epoch": 2.326935380678183, "grad_norm": 1.9518843166762652, "learning_rate": 1.7921586983009304e-05, "loss": 0.4497, "step": 3637 }, { "epoch": 2.3275751759436982, "grad_norm": 1.92994060271758, "learning_rate": 1.7920322145955563e-05, "loss": 0.4175, "step": 3638 }, { "epoch": 2.328214971209213, "grad_norm": 2.015129843948813, "learning_rate": 1.7919056968817277e-05, "loss": 0.4327, "step": 3639 }, { "epoch": 2.328854766474728, "grad_norm": 1.8759464476416716, "learning_rate": 1.791779145164877e-05, "loss": 0.3931, "step": 3640 }, { "epoch": 2.329494561740243, "grad_norm": 1.8997854226415007, "learning_rate": 1.7916525594504388e-05, "loss": 0.4432, "step": 3641 }, { "epoch": 2.330134357005758, "grad_norm": 1.8348447238798875, "learning_rate": 1.7915259397438473e-05, "loss": 0.4536, "step": 3642 }, { "epoch": 2.3307741522712733, "grad_norm": 2.0307235758983166, "learning_rate": 1.7913992860505402e-05, "loss": 0.4575, "step": 3643 }, { "epoch": 2.331413947536788, "grad_norm": 1.9294798309406667, "learning_rate": 1.7912725983759556e-05, "loss": 0.4864, "step": 3644 }, { "epoch": 2.3320537428023034, "grad_norm": 1.9812447274313991, "learning_rate": 1.7911458767255333e-05, "loss": 0.4598, "step": 3645 }, { "epoch": 2.332693538067818, "grad_norm": 1.7531573319638818, "learning_rate": 1.791019121104714e-05, "loss": 0.4389, "step": 3646 }, { "epoch": 2.3333333333333335, "grad_norm": 1.8876780288993342, "learning_rate": 1.7908923315189413e-05, "loss": 0.4174, "step": 3647 }, { "epoch": 2.3339731285988483, "grad_norm": 2.10761540932619, "learning_rate": 1.7907655079736583e-05, "loss": 0.4566, "step": 3648 }, { "epoch": 2.3346129238643636, "grad_norm": 1.9923236557801765, "learning_rate": 1.7906386504743115e-05, "loss": 0.4538, "step": 3649 }, { "epoch": 2.3352527191298784, "grad_norm": 1.9453990501780722, "learning_rate": 1.790511759026347e-05, "loss": 0.4348, "step": 3650 }, { "epoch": 2.3358925143953932, "grad_norm": 2.0042359280699116, "learning_rate": 1.790384833635214e-05, "loss": 0.4757, "step": 3651 }, { "epoch": 2.3365323096609085, "grad_norm": 2.061282643709496, "learning_rate": 1.790257874306362e-05, "loss": 0.4488, "step": 3652 }, { "epoch": 2.3371721049264234, "grad_norm": 2.2362133992649555, "learning_rate": 1.7901308810452433e-05, "loss": 0.4031, "step": 3653 }, { "epoch": 2.3378119001919386, "grad_norm": 1.9821074024735736, "learning_rate": 1.7900038538573094e-05, "loss": 0.4304, "step": 3654 }, { "epoch": 2.3384516954574535, "grad_norm": 2.0638400537594963, "learning_rate": 1.7898767927480157e-05, "loss": 0.4379, "step": 3655 }, { "epoch": 2.3390914907229687, "grad_norm": 2.0214077455785824, "learning_rate": 1.7897496977228174e-05, "loss": 0.4691, "step": 3656 }, { "epoch": 2.3397312859884836, "grad_norm": 2.078694818441723, "learning_rate": 1.789622568787172e-05, "loss": 0.3975, "step": 3657 }, { "epoch": 2.340371081253999, "grad_norm": 2.051243971251175, "learning_rate": 1.789495405946538e-05, "loss": 0.4685, "step": 3658 }, { "epoch": 2.3410108765195137, "grad_norm": 2.184618169665133, "learning_rate": 1.789368209206376e-05, "loss": 0.5008, "step": 3659 }, { "epoch": 2.341650671785029, "grad_norm": 1.9116528456585864, "learning_rate": 1.789240978572147e-05, "loss": 0.4668, "step": 3660 }, { "epoch": 2.342290467050544, "grad_norm": 1.7818383378592608, "learning_rate": 1.7891137140493144e-05, "loss": 0.3624, "step": 3661 }, { "epoch": 2.342930262316059, "grad_norm": 1.7661319962777513, "learning_rate": 1.7889864156433427e-05, "loss": 0.4289, "step": 3662 }, { "epoch": 2.343570057581574, "grad_norm": 1.8438212842798745, "learning_rate": 1.7888590833596978e-05, "loss": 0.4533, "step": 3663 }, { "epoch": 2.344209852847089, "grad_norm": 1.8915018205876306, "learning_rate": 1.788731717203847e-05, "loss": 0.4058, "step": 3664 }, { "epoch": 2.344849648112604, "grad_norm": 1.970602481465609, "learning_rate": 1.7886043171812595e-05, "loss": 0.4197, "step": 3665 }, { "epoch": 2.345489443378119, "grad_norm": 1.9254473591087122, "learning_rate": 1.7884768832974055e-05, "loss": 0.3826, "step": 3666 }, { "epoch": 2.346129238643634, "grad_norm": 1.8728308330762071, "learning_rate": 1.7883494155577566e-05, "loss": 0.3916, "step": 3667 }, { "epoch": 2.346769033909149, "grad_norm": 1.8027499116962766, "learning_rate": 1.7882219139677862e-05, "loss": 0.4436, "step": 3668 }, { "epoch": 2.347408829174664, "grad_norm": 1.9716039935009901, "learning_rate": 1.7880943785329688e-05, "loss": 0.4657, "step": 3669 }, { "epoch": 2.348048624440179, "grad_norm": 1.940150366003284, "learning_rate": 1.7879668092587806e-05, "loss": 0.4613, "step": 3670 }, { "epoch": 2.3486884197056943, "grad_norm": 1.942889940738484, "learning_rate": 1.7878392061507e-05, "loss": 0.4566, "step": 3671 }, { "epoch": 2.349328214971209, "grad_norm": 1.9025954503581495, "learning_rate": 1.7877115692142046e-05, "loss": 0.426, "step": 3672 }, { "epoch": 2.3499680102367244, "grad_norm": 1.7912696589129484, "learning_rate": 1.7875838984547757e-05, "loss": 0.4317, "step": 3673 }, { "epoch": 2.3506078055022392, "grad_norm": 1.838942378466807, "learning_rate": 1.7874561938778955e-05, "loss": 0.4623, "step": 3674 }, { "epoch": 2.3512476007677545, "grad_norm": 1.9435326370098522, "learning_rate": 1.787328455489047e-05, "loss": 0.4588, "step": 3675 }, { "epoch": 2.3518873960332694, "grad_norm": 1.949564419375124, "learning_rate": 1.787200683293715e-05, "loss": 0.4583, "step": 3676 }, { "epoch": 2.352527191298784, "grad_norm": 1.9515127326329234, "learning_rate": 1.7870728772973863e-05, "loss": 0.4722, "step": 3677 }, { "epoch": 2.3531669865642995, "grad_norm": 1.9127884901236167, "learning_rate": 1.786945037505548e-05, "loss": 0.427, "step": 3678 }, { "epoch": 2.3538067818298143, "grad_norm": 2.1124350570407597, "learning_rate": 1.78681716392369e-05, "loss": 0.4277, "step": 3679 }, { "epoch": 2.3544465770953296, "grad_norm": 1.817240713732258, "learning_rate": 1.7866892565573025e-05, "loss": 0.4114, "step": 3680 }, { "epoch": 2.3550863723608444, "grad_norm": 2.0617159712269078, "learning_rate": 1.7865613154118776e-05, "loss": 0.4588, "step": 3681 }, { "epoch": 2.3557261676263597, "grad_norm": 1.9581295157326375, "learning_rate": 1.7864333404929092e-05, "loss": 0.4476, "step": 3682 }, { "epoch": 2.3563659628918745, "grad_norm": 1.8700151022556322, "learning_rate": 1.786305331805892e-05, "loss": 0.4424, "step": 3683 }, { "epoch": 2.3570057581573898, "grad_norm": 1.9191829382824719, "learning_rate": 1.7861772893563224e-05, "loss": 0.4892, "step": 3684 }, { "epoch": 2.3576455534229046, "grad_norm": 2.0772409258196887, "learning_rate": 1.7860492131496987e-05, "loss": 0.4528, "step": 3685 }, { "epoch": 2.35828534868842, "grad_norm": 1.9952283780760702, "learning_rate": 1.7859211031915202e-05, "loss": 0.4278, "step": 3686 }, { "epoch": 2.3589251439539347, "grad_norm": 1.8352956016183715, "learning_rate": 1.7857929594872875e-05, "loss": 0.4162, "step": 3687 }, { "epoch": 2.3595649392194495, "grad_norm": 1.9745602118288426, "learning_rate": 1.785664782042503e-05, "loss": 0.4551, "step": 3688 }, { "epoch": 2.360204734484965, "grad_norm": 1.9807968916974712, "learning_rate": 1.78553657086267e-05, "loss": 0.4896, "step": 3689 }, { "epoch": 2.36084452975048, "grad_norm": 1.9138058214539637, "learning_rate": 1.785408325953294e-05, "loss": 0.4021, "step": 3690 }, { "epoch": 2.361484325015995, "grad_norm": 1.9695735006695907, "learning_rate": 1.785280047319882e-05, "loss": 0.4438, "step": 3691 }, { "epoch": 2.3621241202815098, "grad_norm": 1.9212499253908224, "learning_rate": 1.7851517349679416e-05, "loss": 0.4256, "step": 3692 }, { "epoch": 2.362763915547025, "grad_norm": 1.9953945721813307, "learning_rate": 1.785023388902982e-05, "loss": 0.4336, "step": 3693 }, { "epoch": 2.36340371081254, "grad_norm": 1.8460316277575348, "learning_rate": 1.7848950091305152e-05, "loss": 0.3891, "step": 3694 }, { "epoch": 2.364043506078055, "grad_norm": 1.9172734544797079, "learning_rate": 1.7847665956560524e-05, "loss": 0.3984, "step": 3695 }, { "epoch": 2.36468330134357, "grad_norm": 2.0314134790911083, "learning_rate": 1.7846381484851082e-05, "loss": 0.4315, "step": 3696 }, { "epoch": 2.3653230966090852, "grad_norm": 1.9635206481888068, "learning_rate": 1.7845096676231974e-05, "loss": 0.4117, "step": 3697 }, { "epoch": 2.3659628918746, "grad_norm": 1.8043101079041088, "learning_rate": 1.7843811530758373e-05, "loss": 0.4062, "step": 3698 }, { "epoch": 2.3666026871401153, "grad_norm": 2.11662569237366, "learning_rate": 1.7842526048485455e-05, "loss": 0.4634, "step": 3699 }, { "epoch": 2.36724248240563, "grad_norm": 2.102667803046794, "learning_rate": 1.7841240229468422e-05, "loss": 0.4535, "step": 3700 }, { "epoch": 2.3678822776711455, "grad_norm": 1.8728146541349302, "learning_rate": 1.783995407376248e-05, "loss": 0.4471, "step": 3701 }, { "epoch": 2.3685220729366603, "grad_norm": 2.0626696609968778, "learning_rate": 1.783866758142286e-05, "loss": 0.3995, "step": 3702 }, { "epoch": 2.369161868202175, "grad_norm": 1.9378660670425274, "learning_rate": 1.7837380752504794e-05, "loss": 0.4287, "step": 3703 }, { "epoch": 2.3698016634676904, "grad_norm": 1.798600880665813, "learning_rate": 1.7836093587063544e-05, "loss": 0.4302, "step": 3704 }, { "epoch": 2.370441458733205, "grad_norm": 2.016757415499821, "learning_rate": 1.783480608515437e-05, "loss": 0.4151, "step": 3705 }, { "epoch": 2.3710812539987205, "grad_norm": 1.9981581406669506, "learning_rate": 1.783351824683256e-05, "loss": 0.4356, "step": 3706 }, { "epoch": 2.3717210492642353, "grad_norm": 1.8651662567908773, "learning_rate": 1.7832230072153415e-05, "loss": 0.4483, "step": 3707 }, { "epoch": 2.3723608445297506, "grad_norm": 2.185701153398477, "learning_rate": 1.783094156117224e-05, "loss": 0.4377, "step": 3708 }, { "epoch": 2.3730006397952654, "grad_norm": 1.929844872819424, "learning_rate": 1.7829652713944368e-05, "loss": 0.4512, "step": 3709 }, { "epoch": 2.3736404350607807, "grad_norm": 1.8262582224324884, "learning_rate": 1.7828363530525133e-05, "loss": 0.3985, "step": 3710 }, { "epoch": 2.3742802303262955, "grad_norm": 1.9468489428123117, "learning_rate": 1.7827074010969893e-05, "loss": 0.476, "step": 3711 }, { "epoch": 2.374920025591811, "grad_norm": 1.8293617326140739, "learning_rate": 1.782578415533402e-05, "loss": 0.4187, "step": 3712 }, { "epoch": 2.3755598208573256, "grad_norm": 1.9581918784987187, "learning_rate": 1.78244939636729e-05, "loss": 0.4334, "step": 3713 }, { "epoch": 2.3761996161228405, "grad_norm": 1.8836800911646183, "learning_rate": 1.7823203436041923e-05, "loss": 0.4166, "step": 3714 }, { "epoch": 2.3768394113883557, "grad_norm": 1.9329338413138202, "learning_rate": 1.7821912572496507e-05, "loss": 0.4632, "step": 3715 }, { "epoch": 2.3774792066538706, "grad_norm": 1.9977885747293103, "learning_rate": 1.782062137309208e-05, "loss": 0.4497, "step": 3716 }, { "epoch": 2.378119001919386, "grad_norm": 1.8905068336460067, "learning_rate": 1.7819329837884087e-05, "loss": 0.4285, "step": 3717 }, { "epoch": 2.3787587971849007, "grad_norm": 2.0368093767439026, "learning_rate": 1.7818037966927973e-05, "loss": 0.4688, "step": 3718 }, { "epoch": 2.379398592450416, "grad_norm": 1.8882551620497399, "learning_rate": 1.781674576027922e-05, "loss": 0.4815, "step": 3719 }, { "epoch": 2.380038387715931, "grad_norm": 2.0073568249514113, "learning_rate": 1.7815453217993308e-05, "loss": 0.4832, "step": 3720 }, { "epoch": 2.380678182981446, "grad_norm": 1.9149595771964025, "learning_rate": 1.7814160340125738e-05, "loss": 0.4662, "step": 3721 }, { "epoch": 2.381317978246961, "grad_norm": 1.8516790649273445, "learning_rate": 1.7812867126732026e-05, "loss": 0.4539, "step": 3722 }, { "epoch": 2.381957773512476, "grad_norm": 1.7939906423255854, "learning_rate": 1.7811573577867693e-05, "loss": 0.4301, "step": 3723 }, { "epoch": 2.382597568777991, "grad_norm": 1.904754755249588, "learning_rate": 1.7810279693588288e-05, "loss": 0.4226, "step": 3724 }, { "epoch": 2.383237364043506, "grad_norm": 1.9840825941455167, "learning_rate": 1.7808985473949368e-05, "loss": 0.426, "step": 3725 }, { "epoch": 2.383877159309021, "grad_norm": 1.9107867736252875, "learning_rate": 1.78076909190065e-05, "loss": 0.4238, "step": 3726 }, { "epoch": 2.3845169545745364, "grad_norm": 1.8741415216706703, "learning_rate": 1.7806396028815273e-05, "loss": 0.4288, "step": 3727 }, { "epoch": 2.385156749840051, "grad_norm": 2.010029376927254, "learning_rate": 1.780510080343129e-05, "loss": 0.4277, "step": 3728 }, { "epoch": 2.385796545105566, "grad_norm": 1.8494713146585406, "learning_rate": 1.780380524291016e-05, "loss": 0.4039, "step": 3729 }, { "epoch": 2.3864363403710813, "grad_norm": 1.960543681991811, "learning_rate": 1.7802509347307516e-05, "loss": 0.4297, "step": 3730 }, { "epoch": 2.387076135636596, "grad_norm": 2.0149824889662855, "learning_rate": 1.7801213116679e-05, "loss": 0.4413, "step": 3731 }, { "epoch": 2.3877159309021114, "grad_norm": 1.9921819613329415, "learning_rate": 1.7799916551080268e-05, "loss": 0.4324, "step": 3732 }, { "epoch": 2.3883557261676263, "grad_norm": 1.8186136508992843, "learning_rate": 1.7798619650567e-05, "loss": 0.4188, "step": 3733 }, { "epoch": 2.3889955214331415, "grad_norm": 1.7968958517317906, "learning_rate": 1.779732241519487e-05, "loss": 0.4214, "step": 3734 }, { "epoch": 2.3896353166986564, "grad_norm": 1.8039834016528367, "learning_rate": 1.779602484501959e-05, "loss": 0.3981, "step": 3735 }, { "epoch": 2.3902751119641716, "grad_norm": 1.7385719413238125, "learning_rate": 1.779472694009687e-05, "loss": 0.3913, "step": 3736 }, { "epoch": 2.3909149072296865, "grad_norm": 1.77672841462962, "learning_rate": 1.779342870048244e-05, "loss": 0.3647, "step": 3737 }, { "epoch": 2.3915547024952017, "grad_norm": 1.926852009762949, "learning_rate": 1.7792130126232047e-05, "loss": 0.3942, "step": 3738 }, { "epoch": 2.3921944977607166, "grad_norm": 1.9133642511487678, "learning_rate": 1.779083121740145e-05, "loss": 0.387, "step": 3739 }, { "epoch": 2.3928342930262314, "grad_norm": 2.0951504605188775, "learning_rate": 1.7789531974046413e-05, "loss": 0.4774, "step": 3740 }, { "epoch": 2.3934740882917467, "grad_norm": 1.7247100884011235, "learning_rate": 1.7788232396222732e-05, "loss": 0.4069, "step": 3741 }, { "epoch": 2.3941138835572615, "grad_norm": 1.943098761554885, "learning_rate": 1.7786932483986205e-05, "loss": 0.4341, "step": 3742 }, { "epoch": 2.394753678822777, "grad_norm": 1.9134905120303314, "learning_rate": 1.778563223739265e-05, "loss": 0.4418, "step": 3743 }, { "epoch": 2.3953934740882916, "grad_norm": 1.9642386874426416, "learning_rate": 1.7784331656497897e-05, "loss": 0.4211, "step": 3744 }, { "epoch": 2.396033269353807, "grad_norm": 1.933613992651898, "learning_rate": 1.7783030741357792e-05, "loss": 0.4268, "step": 3745 }, { "epoch": 2.3966730646193217, "grad_norm": 1.9140737095495621, "learning_rate": 1.7781729492028187e-05, "loss": 0.4542, "step": 3746 }, { "epoch": 2.397312859884837, "grad_norm": 1.926399345589959, "learning_rate": 1.7780427908564965e-05, "loss": 0.4746, "step": 3747 }, { "epoch": 2.397952655150352, "grad_norm": 1.827370471081967, "learning_rate": 1.7779125991024007e-05, "loss": 0.3914, "step": 3748 }, { "epoch": 2.398592450415867, "grad_norm": 1.990193274844511, "learning_rate": 1.7777823739461214e-05, "loss": 0.4187, "step": 3749 }, { "epoch": 2.399232245681382, "grad_norm": 1.9423357363476503, "learning_rate": 1.777652115393251e-05, "loss": 0.4343, "step": 3750 }, { "epoch": 2.3998720409468968, "grad_norm": 1.8983314343910527, "learning_rate": 1.777521823449382e-05, "loss": 0.443, "step": 3751 }, { "epoch": 2.400511836212412, "grad_norm": 2.1506065698753303, "learning_rate": 1.7773914981201087e-05, "loss": 0.4304, "step": 3752 }, { "epoch": 2.401151631477927, "grad_norm": 1.9426999145534323, "learning_rate": 1.7772611394110275e-05, "loss": 0.4231, "step": 3753 }, { "epoch": 2.401791426743442, "grad_norm": 2.037880462068111, "learning_rate": 1.7771307473277358e-05, "loss": 0.4469, "step": 3754 }, { "epoch": 2.402431222008957, "grad_norm": 1.8153484296193139, "learning_rate": 1.777000321875832e-05, "loss": 0.4154, "step": 3755 }, { "epoch": 2.4030710172744723, "grad_norm": 2.0608421298947595, "learning_rate": 1.7768698630609168e-05, "loss": 0.3963, "step": 3756 }, { "epoch": 2.403710812539987, "grad_norm": 2.054841250637761, "learning_rate": 1.7767393708885915e-05, "loss": 0.435, "step": 3757 }, { "epoch": 2.4043506078055024, "grad_norm": 1.9594657601975378, "learning_rate": 1.776608845364459e-05, "loss": 0.4506, "step": 3758 }, { "epoch": 2.404990403071017, "grad_norm": 1.862275232400059, "learning_rate": 1.7764782864941247e-05, "loss": 0.3769, "step": 3759 }, { "epoch": 2.4056301983365325, "grad_norm": 1.993333513739219, "learning_rate": 1.7763476942831938e-05, "loss": 0.4518, "step": 3760 }, { "epoch": 2.4062699936020473, "grad_norm": 1.7808287291856275, "learning_rate": 1.7762170687372738e-05, "loss": 0.4146, "step": 3761 }, { "epoch": 2.4069097888675626, "grad_norm": 2.037831433934209, "learning_rate": 1.7760864098619737e-05, "loss": 0.476, "step": 3762 }, { "epoch": 2.4075495841330774, "grad_norm": 1.7761575972580759, "learning_rate": 1.7759557176629037e-05, "loss": 0.3912, "step": 3763 }, { "epoch": 2.4081893793985927, "grad_norm": 1.8618491160331978, "learning_rate": 1.7758249921456755e-05, "loss": 0.3904, "step": 3764 }, { "epoch": 2.4088291746641075, "grad_norm": 1.8889690641133923, "learning_rate": 1.7756942333159025e-05, "loss": 0.4181, "step": 3765 }, { "epoch": 2.4094689699296223, "grad_norm": 2.088783207774328, "learning_rate": 1.7755634411791985e-05, "loss": 0.4177, "step": 3766 }, { "epoch": 2.4101087651951376, "grad_norm": 1.8527646537297464, "learning_rate": 1.77543261574118e-05, "loss": 0.4131, "step": 3767 }, { "epoch": 2.4107485604606524, "grad_norm": 1.8045811954007587, "learning_rate": 1.775301757007465e-05, "loss": 0.4368, "step": 3768 }, { "epoch": 2.4113883557261677, "grad_norm": 2.0225255177077703, "learning_rate": 1.7751708649836715e-05, "loss": 0.4505, "step": 3769 }, { "epoch": 2.4120281509916826, "grad_norm": 2.0227301130569266, "learning_rate": 1.7750399396754196e-05, "loss": 0.4304, "step": 3770 }, { "epoch": 2.412667946257198, "grad_norm": 2.1212038427107864, "learning_rate": 1.7749089810883315e-05, "loss": 0.4539, "step": 3771 }, { "epoch": 2.4133077415227127, "grad_norm": 1.87112530833932, "learning_rate": 1.77477798922803e-05, "loss": 0.4214, "step": 3772 }, { "epoch": 2.413947536788228, "grad_norm": 1.7948796704336587, "learning_rate": 1.7746469641001404e-05, "loss": 0.437, "step": 3773 }, { "epoch": 2.4145873320537428, "grad_norm": 1.7860702110444069, "learning_rate": 1.7745159057102877e-05, "loss": 0.4063, "step": 3774 }, { "epoch": 2.415227127319258, "grad_norm": 1.6963534553779138, "learning_rate": 1.7743848140641004e-05, "loss": 0.3922, "step": 3775 }, { "epoch": 2.415866922584773, "grad_norm": 1.9853376323050036, "learning_rate": 1.774253689167206e-05, "loss": 0.428, "step": 3776 }, { "epoch": 2.4165067178502877, "grad_norm": 2.070434586925149, "learning_rate": 1.7741225310252357e-05, "loss": 0.4214, "step": 3777 }, { "epoch": 2.417146513115803, "grad_norm": 1.8082238405241153, "learning_rate": 1.773991339643821e-05, "loss": 0.4247, "step": 3778 }, { "epoch": 2.417786308381318, "grad_norm": 2.0081494744156085, "learning_rate": 1.773860115028595e-05, "loss": 0.4269, "step": 3779 }, { "epoch": 2.418426103646833, "grad_norm": 1.904840386860165, "learning_rate": 1.7737288571851923e-05, "loss": 0.4086, "step": 3780 }, { "epoch": 2.419065898912348, "grad_norm": 2.2595789502352326, "learning_rate": 1.7735975661192485e-05, "loss": 0.463, "step": 3781 }, { "epoch": 2.419705694177863, "grad_norm": 1.7853015619796522, "learning_rate": 1.773466241836402e-05, "loss": 0.399, "step": 3782 }, { "epoch": 2.420345489443378, "grad_norm": 1.8900592194752965, "learning_rate": 1.7733348843422903e-05, "loss": 0.3978, "step": 3783 }, { "epoch": 2.4209852847088933, "grad_norm": 1.8434951095457748, "learning_rate": 1.7732034936425547e-05, "loss": 0.4258, "step": 3784 }, { "epoch": 2.421625079974408, "grad_norm": 1.8420321755655025, "learning_rate": 1.7730720697428366e-05, "loss": 0.4197, "step": 3785 }, { "epoch": 2.4222648752399234, "grad_norm": 1.8616655617085667, "learning_rate": 1.7729406126487786e-05, "loss": 0.438, "step": 3786 }, { "epoch": 2.4229046705054382, "grad_norm": 1.9054336392082876, "learning_rate": 1.772809122366026e-05, "loss": 0.4066, "step": 3787 }, { "epoch": 2.423544465770953, "grad_norm": 1.855900289076722, "learning_rate": 1.772677598900224e-05, "loss": 0.4252, "step": 3788 }, { "epoch": 2.4241842610364683, "grad_norm": 1.932660900915697, "learning_rate": 1.7725460422570208e-05, "loss": 0.4079, "step": 3789 }, { "epoch": 2.4248240563019836, "grad_norm": 2.028079093516144, "learning_rate": 1.7724144524420648e-05, "loss": 0.4345, "step": 3790 }, { "epoch": 2.4254638515674984, "grad_norm": 2.0384321068481785, "learning_rate": 1.772282829461006e-05, "loss": 0.4643, "step": 3791 }, { "epoch": 2.4261036468330133, "grad_norm": 1.9317475905165449, "learning_rate": 1.7721511733194965e-05, "loss": 0.4289, "step": 3792 }, { "epoch": 2.4267434420985285, "grad_norm": 1.8515662491713767, "learning_rate": 1.772019484023189e-05, "loss": 0.4312, "step": 3793 }, { "epoch": 2.4273832373640434, "grad_norm": 1.8239146393473558, "learning_rate": 1.7718877615777382e-05, "loss": 0.422, "step": 3794 }, { "epoch": 2.4280230326295587, "grad_norm": 1.7811476021491603, "learning_rate": 1.7717560059888003e-05, "loss": 0.4287, "step": 3795 }, { "epoch": 2.4286628278950735, "grad_norm": 1.9097396969053477, "learning_rate": 1.771624217262032e-05, "loss": 0.4164, "step": 3796 }, { "epoch": 2.4293026231605888, "grad_norm": 1.9540473980854167, "learning_rate": 1.7714923954030923e-05, "loss": 0.4355, "step": 3797 }, { "epoch": 2.4299424184261036, "grad_norm": 1.90156908851339, "learning_rate": 1.7713605404176417e-05, "loss": 0.4427, "step": 3798 }, { "epoch": 2.430582213691619, "grad_norm": 2.2266871368474908, "learning_rate": 1.7712286523113418e-05, "loss": 0.411, "step": 3799 }, { "epoch": 2.4312220089571337, "grad_norm": 1.794621403530298, "learning_rate": 1.771096731089855e-05, "loss": 0.4007, "step": 3800 }, { "epoch": 2.431861804222649, "grad_norm": 1.8465905111105196, "learning_rate": 1.770964776758847e-05, "loss": 0.4331, "step": 3801 }, { "epoch": 2.432501599488164, "grad_norm": 1.7589650741469829, "learning_rate": 1.7708327893239823e-05, "loss": 0.4098, "step": 3802 }, { "epoch": 2.4331413947536786, "grad_norm": 1.920702435467733, "learning_rate": 1.770700768790929e-05, "loss": 0.4166, "step": 3803 }, { "epoch": 2.433781190019194, "grad_norm": 2.3396765244748154, "learning_rate": 1.7705687151653558e-05, "loss": 0.4595, "step": 3804 }, { "epoch": 2.4344209852847087, "grad_norm": 1.7637621482887045, "learning_rate": 1.7704366284529327e-05, "loss": 0.4553, "step": 3805 }, { "epoch": 2.435060780550224, "grad_norm": 1.8705175399062615, "learning_rate": 1.7703045086593313e-05, "loss": 0.4595, "step": 3806 }, { "epoch": 2.435700575815739, "grad_norm": 1.9750993716906207, "learning_rate": 1.7701723557902242e-05, "loss": 0.4697, "step": 3807 }, { "epoch": 2.436340371081254, "grad_norm": 1.6390793549455493, "learning_rate": 1.7700401698512862e-05, "loss": 0.3894, "step": 3808 }, { "epoch": 2.436980166346769, "grad_norm": 1.8231291399746974, "learning_rate": 1.7699079508481936e-05, "loss": 0.3735, "step": 3809 }, { "epoch": 2.4376199616122842, "grad_norm": 2.0697622793287005, "learning_rate": 1.7697756987866227e-05, "loss": 0.4451, "step": 3810 }, { "epoch": 2.438259756877799, "grad_norm": 1.8507530083041943, "learning_rate": 1.7696434136722527e-05, "loss": 0.4434, "step": 3811 }, { "epoch": 2.4388995521433143, "grad_norm": 1.8264789230413812, "learning_rate": 1.769511095510764e-05, "loss": 0.4018, "step": 3812 }, { "epoch": 2.439539347408829, "grad_norm": 1.8564976692732966, "learning_rate": 1.7693787443078372e-05, "loss": 0.4262, "step": 3813 }, { "epoch": 2.440179142674344, "grad_norm": 1.8059256899388447, "learning_rate": 1.7692463600691557e-05, "loss": 0.4181, "step": 3814 }, { "epoch": 2.4408189379398593, "grad_norm": 1.9321493521525999, "learning_rate": 1.7691139428004045e-05, "loss": 0.4288, "step": 3815 }, { "epoch": 2.441458733205374, "grad_norm": 2.2337222544436997, "learning_rate": 1.768981492507268e-05, "loss": 0.4549, "step": 3816 }, { "epoch": 2.4420985284708894, "grad_norm": 2.2463534919946335, "learning_rate": 1.7688490091954345e-05, "loss": 0.4431, "step": 3817 }, { "epoch": 2.442738323736404, "grad_norm": 1.9899823253517122, "learning_rate": 1.7687164928705925e-05, "loss": 0.4326, "step": 3818 }, { "epoch": 2.4433781190019195, "grad_norm": 2.059873686348673, "learning_rate": 1.7685839435384316e-05, "loss": 0.4113, "step": 3819 }, { "epoch": 2.4440179142674343, "grad_norm": 2.1286225234706273, "learning_rate": 1.768451361204643e-05, "loss": 0.433, "step": 3820 }, { "epoch": 2.4446577095329496, "grad_norm": 2.1800483862570714, "learning_rate": 1.76831874587492e-05, "loss": 0.3973, "step": 3821 }, { "epoch": 2.4452975047984644, "grad_norm": 1.8250453821290442, "learning_rate": 1.768186097554957e-05, "loss": 0.4416, "step": 3822 }, { "epoch": 2.4459373000639797, "grad_norm": 1.7708121480618193, "learning_rate": 1.7680534162504495e-05, "loss": 0.4569, "step": 3823 }, { "epoch": 2.4465770953294945, "grad_norm": 1.8533477712697155, "learning_rate": 1.7679207019670945e-05, "loss": 0.4147, "step": 3824 }, { "epoch": 2.4472168905950094, "grad_norm": 1.873278496244333, "learning_rate": 1.7677879547105904e-05, "loss": 0.3943, "step": 3825 }, { "epoch": 2.4478566858605246, "grad_norm": 1.8749310635332779, "learning_rate": 1.7676551744866376e-05, "loss": 0.435, "step": 3826 }, { "epoch": 2.44849648112604, "grad_norm": 2.230627955261281, "learning_rate": 1.7675223613009368e-05, "loss": 0.4668, "step": 3827 }, { "epoch": 2.4491362763915547, "grad_norm": 1.8743066198847655, "learning_rate": 1.7673895151591914e-05, "loss": 0.4176, "step": 3828 }, { "epoch": 2.4497760716570696, "grad_norm": 1.9573297758531722, "learning_rate": 1.767256636067105e-05, "loss": 0.4008, "step": 3829 }, { "epoch": 2.450415866922585, "grad_norm": 1.9548586287766212, "learning_rate": 1.767123724030384e-05, "loss": 0.4066, "step": 3830 }, { "epoch": 2.4510556621880997, "grad_norm": 1.9896235776405486, "learning_rate": 1.7669907790547346e-05, "loss": 0.4809, "step": 3831 }, { "epoch": 2.451695457453615, "grad_norm": 1.8856044623820032, "learning_rate": 1.7668578011458657e-05, "loss": 0.3962, "step": 3832 }, { "epoch": 2.4523352527191298, "grad_norm": 2.071293069696439, "learning_rate": 1.7667247903094868e-05, "loss": 0.4812, "step": 3833 }, { "epoch": 2.452975047984645, "grad_norm": 1.7840745468624073, "learning_rate": 1.7665917465513092e-05, "loss": 0.4002, "step": 3834 }, { "epoch": 2.45361484325016, "grad_norm": 1.854787735363968, "learning_rate": 1.7664586698770457e-05, "loss": 0.3945, "step": 3835 }, { "epoch": 2.454254638515675, "grad_norm": 2.0344180627671573, "learning_rate": 1.7663255602924104e-05, "loss": 0.3937, "step": 3836 }, { "epoch": 2.45489443378119, "grad_norm": 1.9668045243399956, "learning_rate": 1.7661924178031187e-05, "loss": 0.443, "step": 3837 }, { "epoch": 2.4555342290467053, "grad_norm": 2.2283134568202247, "learning_rate": 1.766059242414888e-05, "loss": 0.4502, "step": 3838 }, { "epoch": 2.45617402431222, "grad_norm": 1.8823232722971357, "learning_rate": 1.765926034133435e-05, "loss": 0.406, "step": 3839 }, { "epoch": 2.456813819577735, "grad_norm": 1.8613102496765839, "learning_rate": 1.7657927929644816e-05, "loss": 0.3829, "step": 3840 }, { "epoch": 2.45745361484325, "grad_norm": 1.9207592892988385, "learning_rate": 1.7656595189137476e-05, "loss": 0.3939, "step": 3841 }, { "epoch": 2.458093410108765, "grad_norm": 1.946979448016993, "learning_rate": 1.7655262119869558e-05, "loss": 0.4457, "step": 3842 }, { "epoch": 2.4587332053742803, "grad_norm": 1.8838654422723742, "learning_rate": 1.76539287218983e-05, "loss": 0.3957, "step": 3843 }, { "epoch": 2.459373000639795, "grad_norm": 1.8896886431851878, "learning_rate": 1.7652594995280963e-05, "loss": 0.4497, "step": 3844 }, { "epoch": 2.4600127959053104, "grad_norm": 1.7720554252803609, "learning_rate": 1.7651260940074806e-05, "loss": 0.4044, "step": 3845 }, { "epoch": 2.4606525911708252, "grad_norm": 2.094675152245673, "learning_rate": 1.7649926556337118e-05, "loss": 0.4582, "step": 3846 }, { "epoch": 2.4612923864363405, "grad_norm": 2.0568318074583, "learning_rate": 1.764859184412519e-05, "loss": 0.4398, "step": 3847 }, { "epoch": 2.4619321817018553, "grad_norm": 2.005544596639724, "learning_rate": 1.7647256803496334e-05, "loss": 0.3854, "step": 3848 }, { "epoch": 2.4625719769673706, "grad_norm": 1.8393992112815576, "learning_rate": 1.7645921434507872e-05, "loss": 0.4273, "step": 3849 }, { "epoch": 2.4632117722328855, "grad_norm": 1.8423222280221256, "learning_rate": 1.7644585737217148e-05, "loss": 0.4549, "step": 3850 }, { "epoch": 2.4638515674984003, "grad_norm": 1.935221037054557, "learning_rate": 1.7643249711681512e-05, "loss": 0.3848, "step": 3851 }, { "epoch": 2.4644913627639156, "grad_norm": 1.9726348584302384, "learning_rate": 1.7641913357958326e-05, "loss": 0.3624, "step": 3852 }, { "epoch": 2.4651311580294304, "grad_norm": 2.060321731276766, "learning_rate": 1.7640576676104974e-05, "loss": 0.458, "step": 3853 }, { "epoch": 2.4657709532949457, "grad_norm": 1.9918280487980347, "learning_rate": 1.7639239666178855e-05, "loss": 0.4017, "step": 3854 }, { "epoch": 2.4664107485604605, "grad_norm": 1.8466843278209029, "learning_rate": 1.7637902328237374e-05, "loss": 0.4206, "step": 3855 }, { "epoch": 2.4670505438259758, "grad_norm": 1.8173784789509293, "learning_rate": 1.763656466233795e-05, "loss": 0.4249, "step": 3856 }, { "epoch": 2.4676903390914906, "grad_norm": 1.9446945690064072, "learning_rate": 1.7635226668538024e-05, "loss": 0.4235, "step": 3857 }, { "epoch": 2.468330134357006, "grad_norm": 1.935740952936062, "learning_rate": 1.763388834689505e-05, "loss": 0.4258, "step": 3858 }, { "epoch": 2.4689699296225207, "grad_norm": 1.9838567892283012, "learning_rate": 1.7632549697466487e-05, "loss": 0.4447, "step": 3859 }, { "epoch": 2.469609724888036, "grad_norm": 2.0619017487952673, "learning_rate": 1.763121072030982e-05, "loss": 0.4137, "step": 3860 }, { "epoch": 2.470249520153551, "grad_norm": 1.7495293110900902, "learning_rate": 1.762987141548254e-05, "loss": 0.4282, "step": 3861 }, { "epoch": 2.470889315419066, "grad_norm": 1.8950704854004696, "learning_rate": 1.762853178304215e-05, "loss": 0.4138, "step": 3862 }, { "epoch": 2.471529110684581, "grad_norm": 2.0019047938080563, "learning_rate": 1.762719182304618e-05, "loss": 0.3973, "step": 3863 }, { "epoch": 2.472168905950096, "grad_norm": 1.9098768489953206, "learning_rate": 1.7625851535552156e-05, "loss": 0.4256, "step": 3864 }, { "epoch": 2.472808701215611, "grad_norm": 1.854334299332383, "learning_rate": 1.7624510920617638e-05, "loss": 0.3521, "step": 3865 }, { "epoch": 2.473448496481126, "grad_norm": 1.7555245566289774, "learning_rate": 1.7623169978300177e-05, "loss": 0.3895, "step": 3866 }, { "epoch": 2.474088291746641, "grad_norm": 1.9373841117219581, "learning_rate": 1.7621828708657364e-05, "loss": 0.4187, "step": 3867 }, { "epoch": 2.474728087012156, "grad_norm": 2.066786257276827, "learning_rate": 1.762048711174678e-05, "loss": 0.4718, "step": 3868 }, { "epoch": 2.4753678822776712, "grad_norm": 2.017756276022328, "learning_rate": 1.7619145187626036e-05, "loss": 0.4619, "step": 3869 }, { "epoch": 2.476007677543186, "grad_norm": 1.789208797127696, "learning_rate": 1.7617802936352753e-05, "loss": 0.3852, "step": 3870 }, { "epoch": 2.4766474728087013, "grad_norm": 1.7387682377797389, "learning_rate": 1.761646035798456e-05, "loss": 0.407, "step": 3871 }, { "epoch": 2.477287268074216, "grad_norm": 1.8726088729746397, "learning_rate": 1.7615117452579113e-05, "loss": 0.45, "step": 3872 }, { "epoch": 2.4779270633397315, "grad_norm": 1.7560574517137029, "learning_rate": 1.7613774220194068e-05, "loss": 0.4015, "step": 3873 }, { "epoch": 2.4785668586052463, "grad_norm": 1.7548747096587116, "learning_rate": 1.7612430660887096e-05, "loss": 0.3912, "step": 3874 }, { "epoch": 2.4792066538707616, "grad_norm": 1.8845991432420006, "learning_rate": 1.7611086774715896e-05, "loss": 0.4456, "step": 3875 }, { "epoch": 2.4798464491362764, "grad_norm": 1.7707278157905666, "learning_rate": 1.760974256173817e-05, "loss": 0.3974, "step": 3876 }, { "epoch": 2.480486244401791, "grad_norm": 1.891649179389692, "learning_rate": 1.7608398022011635e-05, "loss": 0.4061, "step": 3877 }, { "epoch": 2.4811260396673065, "grad_norm": 1.8096009759490144, "learning_rate": 1.7607053155594024e-05, "loss": 0.3851, "step": 3878 }, { "epoch": 2.4817658349328213, "grad_norm": 1.7523679911512382, "learning_rate": 1.760570796254308e-05, "loss": 0.4656, "step": 3879 }, { "epoch": 2.4824056301983366, "grad_norm": 1.682511367668392, "learning_rate": 1.7604362442916567e-05, "loss": 0.4008, "step": 3880 }, { "epoch": 2.4830454254638514, "grad_norm": 1.7913063968695033, "learning_rate": 1.7603016596772256e-05, "loss": 0.4042, "step": 3881 }, { "epoch": 2.4836852207293667, "grad_norm": 1.7230509357402966, "learning_rate": 1.7601670424167943e-05, "loss": 0.3842, "step": 3882 }, { "epoch": 2.4843250159948815, "grad_norm": 1.8094919634421134, "learning_rate": 1.7600323925161418e-05, "loss": 0.3703, "step": 3883 }, { "epoch": 2.484964811260397, "grad_norm": 1.9715163146017707, "learning_rate": 1.7598977099810506e-05, "loss": 0.4421, "step": 3884 }, { "epoch": 2.4856046065259116, "grad_norm": 1.8280527091267196, "learning_rate": 1.7597629948173033e-05, "loss": 0.3644, "step": 3885 }, { "epoch": 2.486244401791427, "grad_norm": 1.8192910712027277, "learning_rate": 1.759628247030685e-05, "loss": 0.4185, "step": 3886 }, { "epoch": 2.4868841970569417, "grad_norm": 1.7843249951452664, "learning_rate": 1.7594934666269804e-05, "loss": 0.4562, "step": 3887 }, { "epoch": 2.4875239923224566, "grad_norm": 1.9654341687486307, "learning_rate": 1.759358653611978e-05, "loss": 0.4464, "step": 3888 }, { "epoch": 2.488163787587972, "grad_norm": 1.8276835883164217, "learning_rate": 1.7592238079914653e-05, "loss": 0.4117, "step": 3889 }, { "epoch": 2.488803582853487, "grad_norm": 1.7793709134837306, "learning_rate": 1.759088929771233e-05, "loss": 0.401, "step": 3890 }, { "epoch": 2.489443378119002, "grad_norm": 2.1218367768993116, "learning_rate": 1.758954018957072e-05, "loss": 0.4648, "step": 3891 }, { "epoch": 2.490083173384517, "grad_norm": 1.8704957119356571, "learning_rate": 1.7588190755547763e-05, "loss": 0.3723, "step": 3892 }, { "epoch": 2.490722968650032, "grad_norm": 1.9429240551349989, "learning_rate": 1.7586840995701387e-05, "loss": 0.433, "step": 3893 }, { "epoch": 2.491362763915547, "grad_norm": 2.2521880466463426, "learning_rate": 1.758549091008956e-05, "loss": 0.4437, "step": 3894 }, { "epoch": 2.492002559181062, "grad_norm": 2.0111644536842217, "learning_rate": 1.758414049877024e-05, "loss": 0.4413, "step": 3895 }, { "epoch": 2.492642354446577, "grad_norm": 1.8439649158344917, "learning_rate": 1.7582789761801423e-05, "loss": 0.4234, "step": 3896 }, { "epoch": 2.4932821497120923, "grad_norm": 1.7303153978395094, "learning_rate": 1.7581438699241102e-05, "loss": 0.4355, "step": 3897 }, { "epoch": 2.493921944977607, "grad_norm": 1.8645578231005473, "learning_rate": 1.758008731114729e-05, "loss": 0.4718, "step": 3898 }, { "epoch": 2.4945617402431224, "grad_norm": 1.8395614666724025, "learning_rate": 1.7578735597578008e-05, "loss": 0.4249, "step": 3899 }, { "epoch": 2.495201535508637, "grad_norm": 1.8280076061429649, "learning_rate": 1.7577383558591306e-05, "loss": 0.4191, "step": 3900 }, { "epoch": 2.4958413307741525, "grad_norm": 1.7072162205004517, "learning_rate": 1.7576031194245235e-05, "loss": 0.4134, "step": 3901 }, { "epoch": 2.4964811260396673, "grad_norm": 1.7524423681399155, "learning_rate": 1.7574678504597855e-05, "loss": 0.3799, "step": 3902 }, { "epoch": 2.497120921305182, "grad_norm": 1.7661561501267036, "learning_rate": 1.757332548970726e-05, "loss": 0.3475, "step": 3903 }, { "epoch": 2.4977607165706974, "grad_norm": 1.781254325329872, "learning_rate": 1.7571972149631532e-05, "loss": 0.3944, "step": 3904 }, { "epoch": 2.4984005118362123, "grad_norm": 1.8892519129632626, "learning_rate": 1.7570618484428798e-05, "loss": 0.4159, "step": 3905 }, { "epoch": 2.4990403071017275, "grad_norm": 1.7814578712374245, "learning_rate": 1.756926449415717e-05, "loss": 0.3802, "step": 3906 }, { "epoch": 2.4996801023672424, "grad_norm": 2.066135992088317, "learning_rate": 1.756791017887479e-05, "loss": 0.4142, "step": 3907 }, { "epoch": 2.5003198976327576, "grad_norm": 1.7376327641664109, "learning_rate": 1.756655553863981e-05, "loss": 0.3931, "step": 3908 }, { "epoch": 2.5009596928982725, "grad_norm": 2.1923553404107423, "learning_rate": 1.7565200573510392e-05, "loss": 0.4173, "step": 3909 }, { "epoch": 2.5015994881637877, "grad_norm": 1.9263608787227706, "learning_rate": 1.756384528354472e-05, "loss": 0.431, "step": 3910 }, { "epoch": 2.5022392834293026, "grad_norm": 1.7690748584753098, "learning_rate": 1.756248966880099e-05, "loss": 0.3801, "step": 3911 }, { "epoch": 2.502879078694818, "grad_norm": 1.7930201682858418, "learning_rate": 1.75611337293374e-05, "loss": 0.414, "step": 3912 }, { "epoch": 2.5035188739603327, "grad_norm": 1.8627360552784333, "learning_rate": 1.7559777465212183e-05, "loss": 0.3734, "step": 3913 }, { "epoch": 2.5041586692258475, "grad_norm": 1.875099628428326, "learning_rate": 1.7558420876483567e-05, "loss": 0.4138, "step": 3914 }, { "epoch": 2.504798464491363, "grad_norm": 1.7460052146174998, "learning_rate": 1.7557063963209805e-05, "loss": 0.3788, "step": 3915 }, { "epoch": 2.505438259756878, "grad_norm": 1.7783006871215061, "learning_rate": 1.755570672544916e-05, "loss": 0.4538, "step": 3916 }, { "epoch": 2.506078055022393, "grad_norm": 1.8775910173826404, "learning_rate": 1.7554349163259908e-05, "loss": 0.4398, "step": 3917 }, { "epoch": 2.5067178502879077, "grad_norm": 1.8045038529083772, "learning_rate": 1.7552991276700343e-05, "loss": 0.4081, "step": 3918 }, { "epoch": 2.507357645553423, "grad_norm": 1.9802572211389804, "learning_rate": 1.7551633065828765e-05, "loss": 0.4134, "step": 3919 }, { "epoch": 2.507997440818938, "grad_norm": 1.7800617671415517, "learning_rate": 1.7550274530703495e-05, "loss": 0.4346, "step": 3920 }, { "epoch": 2.508637236084453, "grad_norm": 1.8699097807457732, "learning_rate": 1.7548915671382873e-05, "loss": 0.3954, "step": 3921 }, { "epoch": 2.509277031349968, "grad_norm": 1.7963626198504072, "learning_rate": 1.7547556487925236e-05, "loss": 0.3955, "step": 3922 }, { "epoch": 2.509916826615483, "grad_norm": 1.963226312557015, "learning_rate": 1.754619698038895e-05, "loss": 0.4452, "step": 3923 }, { "epoch": 2.510556621880998, "grad_norm": 1.8652002589215522, "learning_rate": 1.7544837148832393e-05, "loss": 0.4641, "step": 3924 }, { "epoch": 2.511196417146513, "grad_norm": 1.9949438584385326, "learning_rate": 1.7543476993313948e-05, "loss": 0.4208, "step": 3925 }, { "epoch": 2.511836212412028, "grad_norm": 1.775853251178267, "learning_rate": 1.754211651389202e-05, "loss": 0.4214, "step": 3926 }, { "epoch": 2.5124760076775434, "grad_norm": 1.7688603304700443, "learning_rate": 1.7540755710625023e-05, "loss": 0.39, "step": 3927 }, { "epoch": 2.5131158029430583, "grad_norm": 1.8727519913992505, "learning_rate": 1.7539394583571393e-05, "loss": 0.4409, "step": 3928 }, { "epoch": 2.513755598208573, "grad_norm": 1.8148675375474272, "learning_rate": 1.753803313278957e-05, "loss": 0.429, "step": 3929 }, { "epoch": 2.5143953934740884, "grad_norm": 1.8200798741886124, "learning_rate": 1.7536671358338012e-05, "loss": 0.4285, "step": 3930 }, { "epoch": 2.515035188739603, "grad_norm": 1.848069202344665, "learning_rate": 1.7535309260275192e-05, "loss": 0.3937, "step": 3931 }, { "epoch": 2.5156749840051185, "grad_norm": 1.8472713214352017, "learning_rate": 1.75339468386596e-05, "loss": 0.434, "step": 3932 }, { "epoch": 2.5163147792706333, "grad_norm": 2.0904762615029884, "learning_rate": 1.753258409354973e-05, "loss": 0.4205, "step": 3933 }, { "epoch": 2.5169545745361486, "grad_norm": 1.8474064961956007, "learning_rate": 1.75312210250041e-05, "loss": 0.4005, "step": 3934 }, { "epoch": 2.5175943698016634, "grad_norm": 1.627951479126178, "learning_rate": 1.7529857633081233e-05, "loss": 0.4029, "step": 3935 }, { "epoch": 2.5182341650671782, "grad_norm": 1.7715623468934043, "learning_rate": 1.7528493917839678e-05, "loss": 0.388, "step": 3936 }, { "epoch": 2.5188739603326935, "grad_norm": 1.763480913339221, "learning_rate": 1.7527129879337984e-05, "loss": 0.4055, "step": 3937 }, { "epoch": 2.519513755598209, "grad_norm": 2.0624906128939475, "learning_rate": 1.7525765517634724e-05, "loss": 0.4539, "step": 3938 }, { "epoch": 2.5201535508637236, "grad_norm": 1.9829113052086234, "learning_rate": 1.752440083278848e-05, "loss": 0.4062, "step": 3939 }, { "epoch": 2.5207933461292384, "grad_norm": 1.7669614891107772, "learning_rate": 1.7523035824857845e-05, "loss": 0.3892, "step": 3940 }, { "epoch": 2.5214331413947537, "grad_norm": 1.7994098925220343, "learning_rate": 1.752167049390144e-05, "loss": 0.4298, "step": 3941 }, { "epoch": 2.5220729366602685, "grad_norm": 1.8033545328325091, "learning_rate": 1.752030483997788e-05, "loss": 0.4119, "step": 3942 }, { "epoch": 2.522712731925784, "grad_norm": 1.8874970560065438, "learning_rate": 1.7518938863145808e-05, "loss": 0.4269, "step": 3943 }, { "epoch": 2.5233525271912987, "grad_norm": 1.7482206818112818, "learning_rate": 1.7517572563463877e-05, "loss": 0.3725, "step": 3944 }, { "epoch": 2.523992322456814, "grad_norm": 2.0919876447802155, "learning_rate": 1.7516205940990755e-05, "loss": 0.4504, "step": 3945 }, { "epoch": 2.5246321177223288, "grad_norm": 1.7799226159780832, "learning_rate": 1.751483899578512e-05, "loss": 0.414, "step": 3946 }, { "epoch": 2.525271912987844, "grad_norm": 1.942937465227561, "learning_rate": 1.7513471727905666e-05, "loss": 0.3887, "step": 3947 }, { "epoch": 2.525911708253359, "grad_norm": 1.820663451388551, "learning_rate": 1.7512104137411098e-05, "loss": 0.3901, "step": 3948 }, { "epoch": 2.526551503518874, "grad_norm": 1.8441446109372477, "learning_rate": 1.7510736224360146e-05, "loss": 0.4127, "step": 3949 }, { "epoch": 2.527191298784389, "grad_norm": 1.770853100821273, "learning_rate": 1.750936798881154e-05, "loss": 0.4032, "step": 3950 }, { "epoch": 2.527831094049904, "grad_norm": 1.8094570899188045, "learning_rate": 1.7507999430824026e-05, "loss": 0.439, "step": 3951 }, { "epoch": 2.528470889315419, "grad_norm": 1.7965544356235763, "learning_rate": 1.750663055045638e-05, "loss": 0.4151, "step": 3952 }, { "epoch": 2.5291106845809344, "grad_norm": 1.9366655534485697, "learning_rate": 1.7505261347767364e-05, "loss": 0.469, "step": 3953 }, { "epoch": 2.529750479846449, "grad_norm": 1.8084160063851327, "learning_rate": 1.7503891822815783e-05, "loss": 0.3826, "step": 3954 }, { "epoch": 2.530390275111964, "grad_norm": 1.7784126111995564, "learning_rate": 1.7502521975660428e-05, "loss": 0.3977, "step": 3955 }, { "epoch": 2.5310300703774793, "grad_norm": 1.7636774789964555, "learning_rate": 1.750115180636013e-05, "loss": 0.4154, "step": 3956 }, { "epoch": 2.531669865642994, "grad_norm": 1.8259726740560371, "learning_rate": 1.7499781314973717e-05, "loss": 0.4116, "step": 3957 }, { "epoch": 2.5323096609085094, "grad_norm": 1.8054698012423194, "learning_rate": 1.7498410501560035e-05, "loss": 0.4236, "step": 3958 }, { "epoch": 2.5329494561740242, "grad_norm": 1.8139902325321613, "learning_rate": 1.7497039366177942e-05, "loss": 0.4369, "step": 3959 }, { "epoch": 2.5335892514395395, "grad_norm": 1.8238311138165937, "learning_rate": 1.7495667908886318e-05, "loss": 0.3894, "step": 3960 }, { "epoch": 2.5342290467050543, "grad_norm": 1.8382412255043037, "learning_rate": 1.749429612974405e-05, "loss": 0.3887, "step": 3961 }, { "epoch": 2.534868841970569, "grad_norm": 1.5629313302447387, "learning_rate": 1.7492924028810036e-05, "loss": 0.3663, "step": 3962 }, { "epoch": 2.5355086372360844, "grad_norm": 1.7688739802445124, "learning_rate": 1.749155160614319e-05, "loss": 0.4194, "step": 3963 }, { "epoch": 2.5361484325015997, "grad_norm": 1.872920650809381, "learning_rate": 1.7490178861802444e-05, "loss": 0.4089, "step": 3964 }, { "epoch": 2.5367882277671145, "grad_norm": 1.8308976544263238, "learning_rate": 1.7488805795846744e-05, "loss": 0.3899, "step": 3965 }, { "epoch": 2.5374280230326294, "grad_norm": 1.885073088943752, "learning_rate": 1.7487432408335047e-05, "loss": 0.4165, "step": 3966 }, { "epoch": 2.5380678182981447, "grad_norm": 1.8304598504311091, "learning_rate": 1.748605869932632e-05, "loss": 0.419, "step": 3967 }, { "epoch": 2.5387076135636595, "grad_norm": 1.6917623487615554, "learning_rate": 1.7484684668879545e-05, "loss": 0.4421, "step": 3968 }, { "epoch": 2.5393474088291748, "grad_norm": 1.826790524592967, "learning_rate": 1.748331031705373e-05, "loss": 0.3928, "step": 3969 }, { "epoch": 2.5399872040946896, "grad_norm": 1.8863999457025382, "learning_rate": 1.7481935643907875e-05, "loss": 0.4515, "step": 3970 }, { "epoch": 2.540626999360205, "grad_norm": 1.9977900261916914, "learning_rate": 1.748056064950102e-05, "loss": 0.4466, "step": 3971 }, { "epoch": 2.5412667946257197, "grad_norm": 1.9191587368587126, "learning_rate": 1.7479185333892195e-05, "loss": 0.4539, "step": 3972 }, { "epoch": 2.541906589891235, "grad_norm": 1.922892884334914, "learning_rate": 1.7477809697140454e-05, "loss": 0.4267, "step": 3973 }, { "epoch": 2.54254638515675, "grad_norm": 1.7810504650030994, "learning_rate": 1.747643373930487e-05, "loss": 0.3976, "step": 3974 }, { "epoch": 2.543186180422265, "grad_norm": 1.9958402794907488, "learning_rate": 1.747505746044452e-05, "loss": 0.4258, "step": 3975 }, { "epoch": 2.54382597568778, "grad_norm": 1.7951277840478994, "learning_rate": 1.74736808606185e-05, "loss": 0.4413, "step": 3976 }, { "epoch": 2.5444657709532947, "grad_norm": 1.8149414522053529, "learning_rate": 1.747230393988592e-05, "loss": 0.385, "step": 3977 }, { "epoch": 2.54510556621881, "grad_norm": 1.9133902857631842, "learning_rate": 1.7470926698305895e-05, "loss": 0.4464, "step": 3978 }, { "epoch": 2.5457453614843253, "grad_norm": 1.7582138414452544, "learning_rate": 1.746954913593757e-05, "loss": 0.3902, "step": 3979 }, { "epoch": 2.54638515674984, "grad_norm": 1.7148409041913197, "learning_rate": 1.7468171252840094e-05, "loss": 0.3889, "step": 3980 }, { "epoch": 2.547024952015355, "grad_norm": 1.8139544010649002, "learning_rate": 1.746679304907263e-05, "loss": 0.4495, "step": 3981 }, { "epoch": 2.5476647472808702, "grad_norm": 1.8362239422604834, "learning_rate": 1.746541452469435e-05, "loss": 0.4173, "step": 3982 }, { "epoch": 2.548304542546385, "grad_norm": 1.8121461729410238, "learning_rate": 1.7464035679764455e-05, "loss": 0.3665, "step": 3983 }, { "epoch": 2.5489443378119003, "grad_norm": 1.804444107387752, "learning_rate": 1.7462656514342144e-05, "loss": 0.4367, "step": 3984 }, { "epoch": 2.549584133077415, "grad_norm": 1.8364094442797307, "learning_rate": 1.746127702848664e-05, "loss": 0.3872, "step": 3985 }, { "epoch": 2.5502239283429304, "grad_norm": 1.7301849488469543, "learning_rate": 1.745989722225717e-05, "loss": 0.3753, "step": 3986 }, { "epoch": 2.5508637236084453, "grad_norm": 1.8242353962633726, "learning_rate": 1.7458517095712984e-05, "loss": 0.4, "step": 3987 }, { "epoch": 2.55150351887396, "grad_norm": 1.7415913581685172, "learning_rate": 1.745713664891334e-05, "loss": 0.3402, "step": 3988 }, { "epoch": 2.5521433141394754, "grad_norm": 1.8143767688575099, "learning_rate": 1.7455755881917515e-05, "loss": 0.4169, "step": 3989 }, { "epoch": 2.5527831094049906, "grad_norm": 1.8888462422670687, "learning_rate": 1.7454374794784796e-05, "loss": 0.4029, "step": 3990 }, { "epoch": 2.5534229046705055, "grad_norm": 1.932784874748598, "learning_rate": 1.7452993387574482e-05, "loss": 0.4108, "step": 3991 }, { "epoch": 2.5540626999360203, "grad_norm": 1.7883097652252595, "learning_rate": 1.745161166034589e-05, "loss": 0.3745, "step": 3992 }, { "epoch": 2.5547024952015356, "grad_norm": 1.809691473725621, "learning_rate": 1.745022961315835e-05, "loss": 0.4258, "step": 3993 }, { "epoch": 2.5553422904670504, "grad_norm": 1.885123398291673, "learning_rate": 1.7448847246071203e-05, "loss": 0.4101, "step": 3994 }, { "epoch": 2.5559820857325657, "grad_norm": 1.817640343509578, "learning_rate": 1.7447464559143806e-05, "loss": 0.4129, "step": 3995 }, { "epoch": 2.5566218809980805, "grad_norm": 1.858771870874405, "learning_rate": 1.744608155243553e-05, "loss": 0.413, "step": 3996 }, { "epoch": 2.557261676263596, "grad_norm": 1.7376382439130624, "learning_rate": 1.744469822600575e-05, "loss": 0.4172, "step": 3997 }, { "epoch": 2.5579014715291106, "grad_norm": 1.7980369244878538, "learning_rate": 1.7443314579913878e-05, "loss": 0.405, "step": 3998 }, { "epoch": 2.5585412667946255, "grad_norm": 1.9291475502653552, "learning_rate": 1.744193061421932e-05, "loss": 0.4277, "step": 3999 }, { "epoch": 2.5591810620601407, "grad_norm": 1.9377925505519422, "learning_rate": 1.7440546328981497e-05, "loss": 0.4751, "step": 4000 }, { "epoch": 2.559820857325656, "grad_norm": 1.7969249973916683, "learning_rate": 1.743916172425985e-05, "loss": 0.3859, "step": 4001 }, { "epoch": 2.560460652591171, "grad_norm": 1.7459478432868611, "learning_rate": 1.7437776800113833e-05, "loss": 0.426, "step": 4002 }, { "epoch": 2.5611004478566857, "grad_norm": 1.8171536069087117, "learning_rate": 1.743639155660291e-05, "loss": 0.401, "step": 4003 }, { "epoch": 2.561740243122201, "grad_norm": 1.7412498430835714, "learning_rate": 1.743500599378656e-05, "loss": 0.3855, "step": 4004 }, { "epoch": 2.5623800383877158, "grad_norm": 1.737976368011274, "learning_rate": 1.7433620111724283e-05, "loss": 0.3819, "step": 4005 }, { "epoch": 2.563019833653231, "grad_norm": 1.875361686116006, "learning_rate": 1.743223391047558e-05, "loss": 0.4093, "step": 4006 }, { "epoch": 2.563659628918746, "grad_norm": 1.9179405516234878, "learning_rate": 1.743084739009997e-05, "loss": 0.4345, "step": 4007 }, { "epoch": 2.564299424184261, "grad_norm": 1.7785808796218179, "learning_rate": 1.7429460550656997e-05, "loss": 0.3537, "step": 4008 }, { "epoch": 2.564939219449776, "grad_norm": 1.8655795521499219, "learning_rate": 1.7428073392206202e-05, "loss": 0.4111, "step": 4009 }, { "epoch": 2.5655790147152913, "grad_norm": 1.8212347825823216, "learning_rate": 1.742668591480715e-05, "loss": 0.3749, "step": 4010 }, { "epoch": 2.566218809980806, "grad_norm": 1.6799617958224078, "learning_rate": 1.7425298118519413e-05, "loss": 0.4287, "step": 4011 }, { "epoch": 2.5668586052463214, "grad_norm": 1.7917970830464973, "learning_rate": 1.7423910003402587e-05, "loss": 0.4014, "step": 4012 }, { "epoch": 2.567498400511836, "grad_norm": 1.7308578792929115, "learning_rate": 1.742252156951627e-05, "loss": 0.3936, "step": 4013 }, { "epoch": 2.568138195777351, "grad_norm": 1.8575238949305537, "learning_rate": 1.7421132816920078e-05, "loss": 0.387, "step": 4014 }, { "epoch": 2.5687779910428663, "grad_norm": 1.7108020548523493, "learning_rate": 1.7419743745673647e-05, "loss": 0.388, "step": 4015 }, { "epoch": 2.5694177863083816, "grad_norm": 1.8027040505818654, "learning_rate": 1.7418354355836617e-05, "loss": 0.4179, "step": 4016 }, { "epoch": 2.5700575815738964, "grad_norm": 1.8667910789865418, "learning_rate": 1.741696464746865e-05, "loss": 0.4435, "step": 4017 }, { "epoch": 2.5706973768394112, "grad_norm": 2.0838174996720737, "learning_rate": 1.741557462062941e-05, "loss": 0.4354, "step": 4018 }, { "epoch": 2.5713371721049265, "grad_norm": 1.822611396868532, "learning_rate": 1.741418427537859e-05, "loss": 0.4316, "step": 4019 }, { "epoch": 2.5719769673704413, "grad_norm": 1.772873816264757, "learning_rate": 1.7412793611775883e-05, "loss": 0.3656, "step": 4020 }, { "epoch": 2.5726167626359566, "grad_norm": 1.7321275433427648, "learning_rate": 1.7411402629881007e-05, "loss": 0.4244, "step": 4021 }, { "epoch": 2.5732565579014715, "grad_norm": 1.693345002584498, "learning_rate": 1.741001132975369e-05, "loss": 0.3983, "step": 4022 }, { "epoch": 2.5738963531669867, "grad_norm": 1.7390515086434732, "learning_rate": 1.7408619711453662e-05, "loss": 0.3924, "step": 4023 }, { "epoch": 2.5745361484325016, "grad_norm": 2.0844692891166314, "learning_rate": 1.7407227775040682e-05, "loss": 0.4619, "step": 4024 }, { "epoch": 2.5751759436980164, "grad_norm": 2.0364728317470324, "learning_rate": 1.7405835520574524e-05, "loss": 0.4226, "step": 4025 }, { "epoch": 2.5758157389635317, "grad_norm": 2.04472787359922, "learning_rate": 1.7404442948114956e-05, "loss": 0.3935, "step": 4026 }, { "epoch": 2.576455534229047, "grad_norm": 1.9087060847045483, "learning_rate": 1.740305005772178e-05, "loss": 0.4542, "step": 4027 }, { "epoch": 2.5770953294945618, "grad_norm": 1.7517561533962844, "learning_rate": 1.740165684945481e-05, "loss": 0.424, "step": 4028 }, { "epoch": 2.5777351247600766, "grad_norm": 1.7751470932618447, "learning_rate": 1.7400263323373858e-05, "loss": 0.4261, "step": 4029 }, { "epoch": 2.578374920025592, "grad_norm": 1.8580926631295724, "learning_rate": 1.739886947953876e-05, "loss": 0.3938, "step": 4030 }, { "epoch": 2.5790147152911067, "grad_norm": 1.969657943198772, "learning_rate": 1.739747531800937e-05, "loss": 0.388, "step": 4031 }, { "epoch": 2.579654510556622, "grad_norm": 1.7785676253190015, "learning_rate": 1.739608083884555e-05, "loss": 0.4149, "step": 4032 }, { "epoch": 2.580294305822137, "grad_norm": 1.8079951990917222, "learning_rate": 1.7394686042107173e-05, "loss": 0.4106, "step": 4033 }, { "epoch": 2.580934101087652, "grad_norm": 1.7688387647763544, "learning_rate": 1.7393290927854135e-05, "loss": 0.4153, "step": 4034 }, { "epoch": 2.581573896353167, "grad_norm": 1.832932183748095, "learning_rate": 1.7391895496146332e-05, "loss": 0.3866, "step": 4035 }, { "epoch": 2.5822136916186818, "grad_norm": 1.96650267117124, "learning_rate": 1.739049974704369e-05, "loss": 0.4186, "step": 4036 }, { "epoch": 2.582853486884197, "grad_norm": 1.81639460288407, "learning_rate": 1.7389103680606132e-05, "loss": 0.4041, "step": 4037 }, { "epoch": 2.5834932821497123, "grad_norm": 1.8658093943266596, "learning_rate": 1.738770729689361e-05, "loss": 0.4307, "step": 4038 }, { "epoch": 2.584133077415227, "grad_norm": 1.8055089712687906, "learning_rate": 1.738631059596607e-05, "loss": 0.4064, "step": 4039 }, { "epoch": 2.584772872680742, "grad_norm": 1.5871145467031176, "learning_rate": 1.73849135778835e-05, "loss": 0.4057, "step": 4040 }, { "epoch": 2.5854126679462572, "grad_norm": 1.8878844398062677, "learning_rate": 1.7383516242705874e-05, "loss": 0.4039, "step": 4041 }, { "epoch": 2.586052463211772, "grad_norm": 1.7866502507977529, "learning_rate": 1.73821185904932e-05, "loss": 0.442, "step": 4042 }, { "epoch": 2.5866922584772873, "grad_norm": 1.7097593540190494, "learning_rate": 1.7380720621305483e-05, "loss": 0.3518, "step": 4043 }, { "epoch": 2.587332053742802, "grad_norm": 1.8578990403812115, "learning_rate": 1.737932233520275e-05, "loss": 0.4162, "step": 4044 }, { "epoch": 2.5879718490083174, "grad_norm": 2.154827354949683, "learning_rate": 1.7377923732245043e-05, "loss": 0.4274, "step": 4045 }, { "epoch": 2.5886116442738323, "grad_norm": 1.8936670762678551, "learning_rate": 1.7376524812492415e-05, "loss": 0.3926, "step": 4046 }, { "epoch": 2.5892514395393476, "grad_norm": 1.8967363571053144, "learning_rate": 1.7375125576004933e-05, "loss": 0.4277, "step": 4047 }, { "epoch": 2.5898912348048624, "grad_norm": 1.940577915878539, "learning_rate": 1.737372602284268e-05, "loss": 0.4216, "step": 4048 }, { "epoch": 2.5905310300703777, "grad_norm": 2.0382785247377364, "learning_rate": 1.737232615306575e-05, "loss": 0.4714, "step": 4049 }, { "epoch": 2.5911708253358925, "grad_norm": 1.8168996209437458, "learning_rate": 1.7370925966734246e-05, "loss": 0.4104, "step": 4050 }, { "epoch": 2.5918106206014073, "grad_norm": 1.7563110641377317, "learning_rate": 1.7369525463908292e-05, "loss": 0.4077, "step": 4051 }, { "epoch": 2.5924504158669226, "grad_norm": 1.7733896513757421, "learning_rate": 1.7368124644648027e-05, "loss": 0.3644, "step": 4052 }, { "epoch": 2.593090211132438, "grad_norm": 1.7327809669878647, "learning_rate": 1.7366723509013598e-05, "loss": 0.4018, "step": 4053 }, { "epoch": 2.5937300063979527, "grad_norm": 1.8141101318140633, "learning_rate": 1.736532205706516e-05, "loss": 0.4527, "step": 4054 }, { "epoch": 2.5943698016634675, "grad_norm": 1.9606603435463872, "learning_rate": 1.73639202888629e-05, "loss": 0.4358, "step": 4055 }, { "epoch": 2.595009596928983, "grad_norm": 1.7957048215483649, "learning_rate": 1.7362518204467e-05, "loss": 0.4282, "step": 4056 }, { "epoch": 2.5956493921944976, "grad_norm": 1.7272296978816322, "learning_rate": 1.7361115803937666e-05, "loss": 0.3831, "step": 4057 }, { "epoch": 2.596289187460013, "grad_norm": 1.8401226095324341, "learning_rate": 1.7359713087335112e-05, "loss": 0.3601, "step": 4058 }, { "epoch": 2.5969289827255277, "grad_norm": 1.9116205432731976, "learning_rate": 1.7358310054719572e-05, "loss": 0.46, "step": 4059 }, { "epoch": 2.597568777991043, "grad_norm": 2.1303145226009357, "learning_rate": 1.7356906706151286e-05, "loss": 0.4263, "step": 4060 }, { "epoch": 2.598208573256558, "grad_norm": 1.8860848451237868, "learning_rate": 1.7355503041690513e-05, "loss": 0.4473, "step": 4061 }, { "epoch": 2.5988483685220727, "grad_norm": 1.7436687884753599, "learning_rate": 1.7354099061397523e-05, "loss": 0.3644, "step": 4062 }, { "epoch": 2.599488163787588, "grad_norm": 1.876260559211986, "learning_rate": 1.7352694765332606e-05, "loss": 0.4068, "step": 4063 }, { "epoch": 2.6001279590531032, "grad_norm": 2.0306781112762953, "learning_rate": 1.7351290153556052e-05, "loss": 0.4096, "step": 4064 }, { "epoch": 2.600767754318618, "grad_norm": 1.908694870354115, "learning_rate": 1.7349885226128175e-05, "loss": 0.4307, "step": 4065 }, { "epoch": 2.601407549584133, "grad_norm": 1.8421227479067457, "learning_rate": 1.7348479983109303e-05, "loss": 0.4237, "step": 4066 }, { "epoch": 2.602047344849648, "grad_norm": 1.795153803958622, "learning_rate": 1.7347074424559772e-05, "loss": 0.4473, "step": 4067 }, { "epoch": 2.602687140115163, "grad_norm": 1.8302223782766678, "learning_rate": 1.734566855053993e-05, "loss": 0.4238, "step": 4068 }, { "epoch": 2.6033269353806783, "grad_norm": 2.109034386580635, "learning_rate": 1.7344262361110154e-05, "loss": 0.4513, "step": 4069 }, { "epoch": 2.603966730646193, "grad_norm": 1.944264166512421, "learning_rate": 1.7342855856330814e-05, "loss": 0.398, "step": 4070 }, { "epoch": 2.6046065259117084, "grad_norm": 1.9870215067676915, "learning_rate": 1.7341449036262304e-05, "loss": 0.4424, "step": 4071 }, { "epoch": 2.605246321177223, "grad_norm": 1.6619851402847017, "learning_rate": 1.7340041900965035e-05, "loss": 0.4048, "step": 4072 }, { "epoch": 2.6058861164427385, "grad_norm": 1.9164222804355566, "learning_rate": 1.733863445049942e-05, "loss": 0.3896, "step": 4073 }, { "epoch": 2.6065259117082533, "grad_norm": 2.030244649422831, "learning_rate": 1.73372266849259e-05, "loss": 0.4177, "step": 4074 }, { "epoch": 2.6071657069737686, "grad_norm": 1.9359525956100758, "learning_rate": 1.7335818604304916e-05, "loss": 0.4368, "step": 4075 }, { "epoch": 2.6078055022392834, "grad_norm": 1.7350711228616973, "learning_rate": 1.733441020869693e-05, "loss": 0.3812, "step": 4076 }, { "epoch": 2.6084452975047983, "grad_norm": 1.808663702070607, "learning_rate": 1.7333001498162415e-05, "loss": 0.4326, "step": 4077 }, { "epoch": 2.6090850927703135, "grad_norm": 1.828216719930257, "learning_rate": 1.7331592472761865e-05, "loss": 0.4227, "step": 4078 }, { "epoch": 2.609724888035829, "grad_norm": 1.9097623701990938, "learning_rate": 1.733018313255577e-05, "loss": 0.4684, "step": 4079 }, { "epoch": 2.6103646833013436, "grad_norm": 1.7777509089699606, "learning_rate": 1.7328773477604656e-05, "loss": 0.3863, "step": 4080 }, { "epoch": 2.6110044785668585, "grad_norm": 1.7238365254646186, "learning_rate": 1.732736350796904e-05, "loss": 0.3902, "step": 4081 }, { "epoch": 2.6116442738323737, "grad_norm": 1.765665295338223, "learning_rate": 1.7325953223709477e-05, "loss": 0.4477, "step": 4082 }, { "epoch": 2.6122840690978886, "grad_norm": 2.018488610366886, "learning_rate": 1.732454262488651e-05, "loss": 0.491, "step": 4083 }, { "epoch": 2.612923864363404, "grad_norm": 1.8988517528923377, "learning_rate": 1.7323131711560712e-05, "loss": 0.4355, "step": 4084 }, { "epoch": 2.6135636596289187, "grad_norm": 1.7666921407788163, "learning_rate": 1.7321720483792665e-05, "loss": 0.3745, "step": 4085 }, { "epoch": 2.614203454894434, "grad_norm": 1.82618603005837, "learning_rate": 1.7320308941642966e-05, "loss": 0.4574, "step": 4086 }, { "epoch": 2.614843250159949, "grad_norm": 1.8982794863253365, "learning_rate": 1.7318897085172222e-05, "loss": 0.4424, "step": 4087 }, { "epoch": 2.6154830454254636, "grad_norm": 1.8417595217770137, "learning_rate": 1.7317484914441057e-05, "loss": 0.4236, "step": 4088 }, { "epoch": 2.616122840690979, "grad_norm": 1.8387936407986127, "learning_rate": 1.7316072429510104e-05, "loss": 0.4082, "step": 4089 }, { "epoch": 2.616762635956494, "grad_norm": 1.7396640690400413, "learning_rate": 1.731465963044002e-05, "loss": 0.3834, "step": 4090 }, { "epoch": 2.617402431222009, "grad_norm": 1.7645235658197391, "learning_rate": 1.731324651729146e-05, "loss": 0.4221, "step": 4091 }, { "epoch": 2.618042226487524, "grad_norm": 1.7776558283355506, "learning_rate": 1.7311833090125105e-05, "loss": 0.4633, "step": 4092 }, { "epoch": 2.618682021753039, "grad_norm": 1.8858661935655654, "learning_rate": 1.7310419349001645e-05, "loss": 0.3847, "step": 4093 }, { "epoch": 2.619321817018554, "grad_norm": 2.004115114624288, "learning_rate": 1.730900529398178e-05, "loss": 0.4061, "step": 4094 }, { "epoch": 2.619961612284069, "grad_norm": 1.8530484792512543, "learning_rate": 1.7307590925126227e-05, "loss": 0.3648, "step": 4095 }, { "epoch": 2.620601407549584, "grad_norm": 1.6734724083615247, "learning_rate": 1.7306176242495722e-05, "loss": 0.3862, "step": 4096 }, { "epoch": 2.6212412028150993, "grad_norm": 1.8383941308590568, "learning_rate": 1.7304761246151007e-05, "loss": 0.3812, "step": 4097 }, { "epoch": 2.621880998080614, "grad_norm": 1.9011667320358427, "learning_rate": 1.730334593615284e-05, "loss": 0.4361, "step": 4098 }, { "epoch": 2.622520793346129, "grad_norm": 1.9811137967863437, "learning_rate": 1.7301930312561984e-05, "loss": 0.4478, "step": 4099 }, { "epoch": 2.6231605886116443, "grad_norm": 1.9019211751766871, "learning_rate": 1.7300514375439232e-05, "loss": 0.4495, "step": 4100 }, { "epoch": 2.6238003838771595, "grad_norm": 1.8989100888438313, "learning_rate": 1.7299098124845382e-05, "loss": 0.4474, "step": 4101 }, { "epoch": 2.6244401791426744, "grad_norm": 1.7862065413194785, "learning_rate": 1.729768156084124e-05, "loss": 0.4225, "step": 4102 }, { "epoch": 2.625079974408189, "grad_norm": 1.9365146415704186, "learning_rate": 1.7296264683487633e-05, "loss": 0.3933, "step": 4103 }, { "epoch": 2.6257197696737045, "grad_norm": 1.7059310395229639, "learning_rate": 1.72948474928454e-05, "loss": 0.3609, "step": 4104 }, { "epoch": 2.6263595649392193, "grad_norm": 1.9196306653322495, "learning_rate": 1.7293429988975393e-05, "loss": 0.4156, "step": 4105 }, { "epoch": 2.6269993602047346, "grad_norm": 2.0264245627797624, "learning_rate": 1.729201217193848e-05, "loss": 0.4302, "step": 4106 }, { "epoch": 2.6276391554702494, "grad_norm": 1.608750212246827, "learning_rate": 1.7290594041795527e-05, "loss": 0.3768, "step": 4107 }, { "epoch": 2.6282789507357647, "grad_norm": 1.8882129972982982, "learning_rate": 1.7289175598607442e-05, "loss": 0.4106, "step": 4108 }, { "epoch": 2.6289187460012795, "grad_norm": 1.8313992294009216, "learning_rate": 1.728775684243512e-05, "loss": 0.3741, "step": 4109 }, { "epoch": 2.629558541266795, "grad_norm": 1.7243710542554376, "learning_rate": 1.7286337773339484e-05, "loss": 0.3893, "step": 4110 }, { "epoch": 2.6301983365323096, "grad_norm": 1.8433979886600038, "learning_rate": 1.7284918391381464e-05, "loss": 0.4053, "step": 4111 }, { "epoch": 2.630838131797825, "grad_norm": 1.7546577707617421, "learning_rate": 1.728349869662201e-05, "loss": 0.3777, "step": 4112 }, { "epoch": 2.6314779270633397, "grad_norm": 1.8533490774810317, "learning_rate": 1.7282078689122074e-05, "loss": 0.4361, "step": 4113 }, { "epoch": 2.6321177223288545, "grad_norm": 1.8780483777122199, "learning_rate": 1.7280658368942636e-05, "loss": 0.4192, "step": 4114 }, { "epoch": 2.63275751759437, "grad_norm": 1.7562066720480658, "learning_rate": 1.727923773614468e-05, "loss": 0.3322, "step": 4115 }, { "epoch": 2.633397312859885, "grad_norm": 2.0186606035169894, "learning_rate": 1.72778167907892e-05, "loss": 0.3618, "step": 4116 }, { "epoch": 2.6340371081254, "grad_norm": 1.7008733690276965, "learning_rate": 1.727639553293722e-05, "loss": 0.4159, "step": 4117 }, { "epoch": 2.6346769033909148, "grad_norm": 1.9873739450754178, "learning_rate": 1.7274973962649756e-05, "loss": 0.4324, "step": 4118 }, { "epoch": 2.63531669865643, "grad_norm": 1.8432418636371788, "learning_rate": 1.727355207998785e-05, "loss": 0.377, "step": 4119 }, { "epoch": 2.635956493921945, "grad_norm": 1.883281120329507, "learning_rate": 1.727212988501256e-05, "loss": 0.43, "step": 4120 }, { "epoch": 2.63659628918746, "grad_norm": 1.8618049955690943, "learning_rate": 1.7270707377784944e-05, "loss": 0.4414, "step": 4121 }, { "epoch": 2.637236084452975, "grad_norm": 1.8398743754474522, "learning_rate": 1.7269284558366088e-05, "loss": 0.3813, "step": 4122 }, { "epoch": 2.6378758797184902, "grad_norm": 1.8455871302174778, "learning_rate": 1.7267861426817084e-05, "loss": 0.4517, "step": 4123 }, { "epoch": 2.638515674984005, "grad_norm": 1.7377416600589857, "learning_rate": 1.7266437983199042e-05, "loss": 0.3854, "step": 4124 }, { "epoch": 2.63915547024952, "grad_norm": 1.7765142702317518, "learning_rate": 1.7265014227573074e-05, "loss": 0.406, "step": 4125 }, { "epoch": 2.639795265515035, "grad_norm": 1.8683064023325113, "learning_rate": 1.7263590160000324e-05, "loss": 0.3754, "step": 4126 }, { "epoch": 2.6404350607805505, "grad_norm": 1.8261926017236694, "learning_rate": 1.726216578054193e-05, "loss": 0.4312, "step": 4127 }, { "epoch": 2.6410748560460653, "grad_norm": 1.8595352637014289, "learning_rate": 1.7260741089259056e-05, "loss": 0.4036, "step": 4128 }, { "epoch": 2.64171465131158, "grad_norm": 1.8164272607732508, "learning_rate": 1.725931608621288e-05, "loss": 0.4116, "step": 4129 }, { "epoch": 2.6423544465770954, "grad_norm": 1.7543151118665694, "learning_rate": 1.7257890771464576e-05, "loss": 0.4076, "step": 4130 }, { "epoch": 2.6429942418426102, "grad_norm": 1.8186930280368963, "learning_rate": 1.7256465145075362e-05, "loss": 0.4643, "step": 4131 }, { "epoch": 2.6436340371081255, "grad_norm": 1.8925636601656706, "learning_rate": 1.7255039207106437e-05, "loss": 0.3876, "step": 4132 }, { "epoch": 2.6442738323736403, "grad_norm": 1.8300593660520441, "learning_rate": 1.7253612957619034e-05, "loss": 0.3927, "step": 4133 }, { "epoch": 2.6449136276391556, "grad_norm": 1.6878257546281592, "learning_rate": 1.7252186396674395e-05, "loss": 0.419, "step": 4134 }, { "epoch": 2.6455534229046704, "grad_norm": 1.818142551058083, "learning_rate": 1.7250759524333776e-05, "loss": 0.3958, "step": 4135 }, { "epoch": 2.6461932181701853, "grad_norm": 1.851554432008729, "learning_rate": 1.724933234065844e-05, "loss": 0.3712, "step": 4136 }, { "epoch": 2.6468330134357005, "grad_norm": 1.883443839508552, "learning_rate": 1.7247904845709664e-05, "loss": 0.4494, "step": 4137 }, { "epoch": 2.647472808701216, "grad_norm": 1.8606710489286775, "learning_rate": 1.724647703954875e-05, "loss": 0.3736, "step": 4138 }, { "epoch": 2.6481126039667306, "grad_norm": 1.849374552208842, "learning_rate": 1.7245048922237004e-05, "loss": 0.42, "step": 4139 }, { "epoch": 2.6487523992322455, "grad_norm": 1.8523445486421173, "learning_rate": 1.7243620493835743e-05, "loss": 0.3693, "step": 4140 }, { "epoch": 2.6493921944977608, "grad_norm": 1.7029006192457992, "learning_rate": 1.7242191754406304e-05, "loss": 0.3629, "step": 4141 }, { "epoch": 2.6500319897632756, "grad_norm": 1.7442035654399093, "learning_rate": 1.7240762704010034e-05, "loss": 0.3862, "step": 4142 }, { "epoch": 2.650671785028791, "grad_norm": 1.7803504840292985, "learning_rate": 1.7239333342708297e-05, "loss": 0.4257, "step": 4143 }, { "epoch": 2.6513115802943057, "grad_norm": 1.8920474158746556, "learning_rate": 1.723790367056246e-05, "loss": 0.4375, "step": 4144 }, { "epoch": 2.651951375559821, "grad_norm": 1.6528949896084089, "learning_rate": 1.7236473687633916e-05, "loss": 0.3971, "step": 4145 }, { "epoch": 2.652591170825336, "grad_norm": 1.8529571924323684, "learning_rate": 1.723504339398406e-05, "loss": 0.3963, "step": 4146 }, { "epoch": 2.653230966090851, "grad_norm": 1.6989348793569545, "learning_rate": 1.7233612789674315e-05, "loss": 0.3883, "step": 4147 }, { "epoch": 2.653870761356366, "grad_norm": 1.6715207405731975, "learning_rate": 1.7232181874766106e-05, "loss": 0.3741, "step": 4148 }, { "epoch": 2.654510556621881, "grad_norm": 2.777696436624631, "learning_rate": 1.7230750649320872e-05, "loss": 0.3965, "step": 4149 }, { "epoch": 2.655150351887396, "grad_norm": 1.7408497751967884, "learning_rate": 1.7229319113400066e-05, "loss": 0.3905, "step": 4150 }, { "epoch": 2.655790147152911, "grad_norm": 1.8161114900583832, "learning_rate": 1.7227887267065158e-05, "loss": 0.3859, "step": 4151 }, { "epoch": 2.656429942418426, "grad_norm": 1.7901545030640804, "learning_rate": 1.7226455110377628e-05, "loss": 0.3541, "step": 4152 }, { "epoch": 2.6570697376839414, "grad_norm": 1.7686333303569648, "learning_rate": 1.722502264339897e-05, "loss": 0.4263, "step": 4153 }, { "epoch": 2.657709532949456, "grad_norm": 1.957230313575568, "learning_rate": 1.7223589866190693e-05, "loss": 0.4088, "step": 4154 }, { "epoch": 2.658349328214971, "grad_norm": 1.9891160919623712, "learning_rate": 1.7222156778814315e-05, "loss": 0.4516, "step": 4155 }, { "epoch": 2.6589891234804863, "grad_norm": 2.0106203374054323, "learning_rate": 1.7220723381331376e-05, "loss": 0.4741, "step": 4156 }, { "epoch": 2.659628918746001, "grad_norm": 2.0840077220884545, "learning_rate": 1.7219289673803413e-05, "loss": 0.4348, "step": 4157 }, { "epoch": 2.6602687140115164, "grad_norm": 1.6832313073780751, "learning_rate": 1.7217855656292e-05, "loss": 0.365, "step": 4158 }, { "epoch": 2.6609085092770313, "grad_norm": 1.8440020260635321, "learning_rate": 1.7216421328858703e-05, "loss": 0.3871, "step": 4159 }, { "epoch": 2.6615483045425465, "grad_norm": 1.7436862160597697, "learning_rate": 1.721498669156511e-05, "loss": 0.3397, "step": 4160 }, { "epoch": 2.6621880998080614, "grad_norm": 1.745869122434296, "learning_rate": 1.7213551744472822e-05, "loss": 0.3696, "step": 4161 }, { "epoch": 2.662827895073576, "grad_norm": 1.904903854357807, "learning_rate": 1.7212116487643455e-05, "loss": 0.4645, "step": 4162 }, { "epoch": 2.6634676903390915, "grad_norm": 1.7561956851240472, "learning_rate": 1.7210680921138636e-05, "loss": 0.3914, "step": 4163 }, { "epoch": 2.6641074856046068, "grad_norm": 1.7482544930471648, "learning_rate": 1.7209245045020005e-05, "loss": 0.3727, "step": 4164 }, { "epoch": 2.6647472808701216, "grad_norm": 1.6895250527227774, "learning_rate": 1.7207808859349216e-05, "loss": 0.3255, "step": 4165 }, { "epoch": 2.6653870761356364, "grad_norm": 1.7530051776820412, "learning_rate": 1.7206372364187937e-05, "loss": 0.3633, "step": 4166 }, { "epoch": 2.6660268714011517, "grad_norm": 1.7991956104434346, "learning_rate": 1.7204935559597847e-05, "loss": 0.4327, "step": 4167 }, { "epoch": 2.6666666666666665, "grad_norm": 1.723411467415163, "learning_rate": 1.720349844564064e-05, "loss": 0.3733, "step": 4168 }, { "epoch": 2.667306461932182, "grad_norm": 1.7258162937418673, "learning_rate": 1.7202061022378023e-05, "loss": 0.3856, "step": 4169 }, { "epoch": 2.6679462571976966, "grad_norm": 1.9359064826825856, "learning_rate": 1.720062328987172e-05, "loss": 0.3752, "step": 4170 }, { "epoch": 2.668586052463212, "grad_norm": 1.7831371450959503, "learning_rate": 1.719918524818346e-05, "loss": 0.3685, "step": 4171 }, { "epoch": 2.6692258477287267, "grad_norm": 1.8657605778284896, "learning_rate": 1.719774689737499e-05, "loss": 0.4131, "step": 4172 }, { "epoch": 2.669865642994242, "grad_norm": 1.9251687067074355, "learning_rate": 1.7196308237508078e-05, "loss": 0.4122, "step": 4173 }, { "epoch": 2.670505438259757, "grad_norm": 1.7844987970495454, "learning_rate": 1.7194869268644484e-05, "loss": 0.3813, "step": 4174 }, { "epoch": 2.671145233525272, "grad_norm": 1.8608326961901962, "learning_rate": 1.7193429990846006e-05, "loss": 0.419, "step": 4175 }, { "epoch": 2.671785028790787, "grad_norm": 1.7763188018751945, "learning_rate": 1.7191990404174442e-05, "loss": 0.4152, "step": 4176 }, { "epoch": 2.6724248240563018, "grad_norm": 1.7282525321952371, "learning_rate": 1.71905505086916e-05, "loss": 0.4237, "step": 4177 }, { "epoch": 2.673064619321817, "grad_norm": 1.8746717619318025, "learning_rate": 1.7189110304459313e-05, "loss": 0.486, "step": 4178 }, { "epoch": 2.6737044145873323, "grad_norm": 1.8328935056768518, "learning_rate": 1.7187669791539418e-05, "loss": 0.4645, "step": 4179 }, { "epoch": 2.674344209852847, "grad_norm": 1.9967541238988635, "learning_rate": 1.7186228969993767e-05, "loss": 0.397, "step": 4180 }, { "epoch": 2.674984005118362, "grad_norm": 1.862419182700251, "learning_rate": 1.7184787839884226e-05, "loss": 0.433, "step": 4181 }, { "epoch": 2.6756238003838773, "grad_norm": 1.8546790717043748, "learning_rate": 1.7183346401272678e-05, "loss": 0.4218, "step": 4182 }, { "epoch": 2.676263595649392, "grad_norm": 1.7659414979195873, "learning_rate": 1.7181904654221015e-05, "loss": 0.3631, "step": 4183 }, { "epoch": 2.6769033909149074, "grad_norm": 1.7849934362051458, "learning_rate": 1.7180462598791137e-05, "loss": 0.3901, "step": 4184 }, { "epoch": 2.677543186180422, "grad_norm": 1.6581463959784375, "learning_rate": 1.7179020235044972e-05, "loss": 0.3877, "step": 4185 }, { "epoch": 2.6781829814459375, "grad_norm": 1.5787286996625278, "learning_rate": 1.7177577563044447e-05, "loss": 0.3295, "step": 4186 }, { "epoch": 2.6788227767114523, "grad_norm": 1.8218056830508813, "learning_rate": 1.717613458285151e-05, "loss": 0.3998, "step": 4187 }, { "epoch": 2.679462571976967, "grad_norm": 1.933378015070458, "learning_rate": 1.717469129452812e-05, "loss": 0.4179, "step": 4188 }, { "epoch": 2.6801023672424824, "grad_norm": 1.8496910409483853, "learning_rate": 1.7173247698136246e-05, "loss": 0.4158, "step": 4189 }, { "epoch": 2.6807421625079977, "grad_norm": 1.9155698612669942, "learning_rate": 1.717180379373788e-05, "loss": 0.4197, "step": 4190 }, { "epoch": 2.6813819577735125, "grad_norm": 1.9673908623555179, "learning_rate": 1.7170359581395014e-05, "loss": 0.3686, "step": 4191 }, { "epoch": 2.6820217530390273, "grad_norm": 1.8421160903560259, "learning_rate": 1.7168915061169662e-05, "loss": 0.4133, "step": 4192 }, { "epoch": 2.6826615483045426, "grad_norm": 1.876127214872949, "learning_rate": 1.716747023312385e-05, "loss": 0.4382, "step": 4193 }, { "epoch": 2.6833013435700575, "grad_norm": 1.6530580168623417, "learning_rate": 1.7166025097319616e-05, "loss": 0.3839, "step": 4194 }, { "epoch": 2.6839411388355727, "grad_norm": 2.0069123855373343, "learning_rate": 1.7164579653819014e-05, "loss": 0.4229, "step": 4195 }, { "epoch": 2.6845809341010876, "grad_norm": 1.8711043522162647, "learning_rate": 1.7163133902684105e-05, "loss": 0.3775, "step": 4196 }, { "epoch": 2.685220729366603, "grad_norm": 1.8963767819782353, "learning_rate": 1.7161687843976967e-05, "loss": 0.393, "step": 4197 }, { "epoch": 2.6858605246321177, "grad_norm": 2.022737581673144, "learning_rate": 1.71602414777597e-05, "loss": 0.3945, "step": 4198 }, { "epoch": 2.6865003198976325, "grad_norm": 1.736161755546126, "learning_rate": 1.7158794804094393e-05, "loss": 0.38, "step": 4199 }, { "epoch": 2.6871401151631478, "grad_norm": 1.7955235494130406, "learning_rate": 1.7157347823043173e-05, "loss": 0.4444, "step": 4200 }, { "epoch": 2.687779910428663, "grad_norm": 1.8178817187975251, "learning_rate": 1.7155900534668172e-05, "loss": 0.3969, "step": 4201 }, { "epoch": 2.688419705694178, "grad_norm": 1.8753845525434916, "learning_rate": 1.715445293903153e-05, "loss": 0.3751, "step": 4202 }, { "epoch": 2.6890595009596927, "grad_norm": 2.0668339022601514, "learning_rate": 1.7153005036195402e-05, "loss": 0.4453, "step": 4203 }, { "epoch": 2.689699296225208, "grad_norm": 1.7671967228180419, "learning_rate": 1.7151556826221968e-05, "loss": 0.4153, "step": 4204 }, { "epoch": 2.690339091490723, "grad_norm": 1.7388009522249024, "learning_rate": 1.7150108309173404e-05, "loss": 0.3981, "step": 4205 }, { "epoch": 2.690978886756238, "grad_norm": 1.7096346756554248, "learning_rate": 1.7148659485111905e-05, "loss": 0.3828, "step": 4206 }, { "epoch": 2.691618682021753, "grad_norm": 1.674689596557126, "learning_rate": 1.7147210354099682e-05, "loss": 0.3552, "step": 4207 }, { "epoch": 2.692258477287268, "grad_norm": 1.7108317714567953, "learning_rate": 1.714576091619897e-05, "loss": 0.3727, "step": 4208 }, { "epoch": 2.692898272552783, "grad_norm": 1.7720785432426587, "learning_rate": 1.714431117147199e-05, "loss": 0.3693, "step": 4209 }, { "epoch": 2.6935380678182983, "grad_norm": 1.6288418623997707, "learning_rate": 1.7142861119980995e-05, "loss": 0.424, "step": 4210 }, { "epoch": 2.694177863083813, "grad_norm": 1.6986639565523378, "learning_rate": 1.714141076178825e-05, "loss": 0.362, "step": 4211 }, { "epoch": 2.6948176583493284, "grad_norm": 1.7484252682824954, "learning_rate": 1.7139960096956035e-05, "loss": 0.3995, "step": 4212 }, { "epoch": 2.6954574536148432, "grad_norm": 1.7779503422628526, "learning_rate": 1.713850912554663e-05, "loss": 0.4428, "step": 4213 }, { "epoch": 2.696097248880358, "grad_norm": 1.7079887391319475, "learning_rate": 1.7137057847622345e-05, "loss": 0.3325, "step": 4214 }, { "epoch": 2.6967370441458733, "grad_norm": 1.7478342616458407, "learning_rate": 1.713560626324549e-05, "loss": 0.4097, "step": 4215 }, { "epoch": 2.6973768394113886, "grad_norm": 1.6816550435621014, "learning_rate": 1.7134154372478395e-05, "loss": 0.4481, "step": 4216 }, { "epoch": 2.6980166346769034, "grad_norm": 1.6816357625940856, "learning_rate": 1.7132702175383405e-05, "loss": 0.3935, "step": 4217 }, { "epoch": 2.6986564299424183, "grad_norm": 1.743303228600705, "learning_rate": 1.7131249672022866e-05, "loss": 0.4157, "step": 4218 }, { "epoch": 2.6992962252079336, "grad_norm": 1.7932236618245483, "learning_rate": 1.7129796862459158e-05, "loss": 0.3754, "step": 4219 }, { "epoch": 2.6999360204734484, "grad_norm": 1.9165859240641152, "learning_rate": 1.712834374675465e-05, "loss": 0.3938, "step": 4220 }, { "epoch": 2.7005758157389637, "grad_norm": 1.8139285017078621, "learning_rate": 1.7126890324971745e-05, "loss": 0.3816, "step": 4221 }, { "epoch": 2.7012156110044785, "grad_norm": 1.8004196342422976, "learning_rate": 1.7125436597172846e-05, "loss": 0.3834, "step": 4222 }, { "epoch": 2.7018554062699938, "grad_norm": 1.9013938945919053, "learning_rate": 1.7123982563420375e-05, "loss": 0.4046, "step": 4223 }, { "epoch": 2.7024952015355086, "grad_norm": 1.7862998297804389, "learning_rate": 1.7122528223776768e-05, "loss": 0.39, "step": 4224 }, { "epoch": 2.7031349968010234, "grad_norm": 1.7942807908494787, "learning_rate": 1.7121073578304464e-05, "loss": 0.3937, "step": 4225 }, { "epoch": 2.7037747920665387, "grad_norm": 1.6794243777676867, "learning_rate": 1.711961862706593e-05, "loss": 0.4139, "step": 4226 }, { "epoch": 2.704414587332054, "grad_norm": 1.718713864135743, "learning_rate": 1.7118163370123637e-05, "loss": 0.3852, "step": 4227 }, { "epoch": 2.705054382597569, "grad_norm": 1.9411307031740404, "learning_rate": 1.711670780754007e-05, "loss": 0.4204, "step": 4228 }, { "epoch": 2.7056941778630836, "grad_norm": 1.903556157618042, "learning_rate": 1.7115251939377734e-05, "loss": 0.4451, "step": 4229 }, { "epoch": 2.706333973128599, "grad_norm": 1.8706218592427897, "learning_rate": 1.7113795765699134e-05, "loss": 0.4289, "step": 4230 }, { "epoch": 2.7069737683941137, "grad_norm": 1.7681957458448254, "learning_rate": 1.7112339286566796e-05, "loss": 0.3776, "step": 4231 }, { "epoch": 2.707613563659629, "grad_norm": 1.8547973270112774, "learning_rate": 1.7110882502043265e-05, "loss": 0.3765, "step": 4232 }, { "epoch": 2.708253358925144, "grad_norm": 1.7442600562666055, "learning_rate": 1.7109425412191085e-05, "loss": 0.3855, "step": 4233 }, { "epoch": 2.708893154190659, "grad_norm": 1.7084336824440378, "learning_rate": 1.7107968017072825e-05, "loss": 0.4153, "step": 4234 }, { "epoch": 2.709532949456174, "grad_norm": 1.7068850363521466, "learning_rate": 1.7106510316751068e-05, "loss": 0.4559, "step": 4235 }, { "epoch": 2.710172744721689, "grad_norm": 1.6419592328770287, "learning_rate": 1.710505231128839e-05, "loss": 0.3791, "step": 4236 }, { "epoch": 2.710812539987204, "grad_norm": 1.6693914747994154, "learning_rate": 1.710359400074741e-05, "loss": 0.3569, "step": 4237 }, { "epoch": 2.7114523352527193, "grad_norm": 1.9667913206704097, "learning_rate": 1.710213538519074e-05, "loss": 0.3865, "step": 4238 }, { "epoch": 2.712092130518234, "grad_norm": 1.9069892312109455, "learning_rate": 1.710067646468101e-05, "loss": 0.4234, "step": 4239 }, { "epoch": 2.712731925783749, "grad_norm": 1.6608141969358967, "learning_rate": 1.7099217239280865e-05, "loss": 0.3831, "step": 4240 }, { "epoch": 2.7133717210492643, "grad_norm": 1.9223023178892698, "learning_rate": 1.7097757709052957e-05, "loss": 0.3693, "step": 4241 }, { "epoch": 2.714011516314779, "grad_norm": 1.9407439463193348, "learning_rate": 1.7096297874059962e-05, "loss": 0.4154, "step": 4242 }, { "epoch": 2.7146513115802944, "grad_norm": 1.7053297128252987, "learning_rate": 1.7094837734364557e-05, "loss": 0.3643, "step": 4243 }, { "epoch": 2.715291106845809, "grad_norm": 1.794655358333748, "learning_rate": 1.7093377290029443e-05, "loss": 0.4119, "step": 4244 }, { "epoch": 2.7159309021113245, "grad_norm": 1.7889814283185652, "learning_rate": 1.7091916541117324e-05, "loss": 0.4317, "step": 4245 }, { "epoch": 2.7165706973768393, "grad_norm": 1.8728696332716204, "learning_rate": 1.7090455487690927e-05, "loss": 0.3711, "step": 4246 }, { "epoch": 2.7172104926423546, "grad_norm": 2.107131399842562, "learning_rate": 1.7088994129812985e-05, "loss": 0.4664, "step": 4247 }, { "epoch": 2.7178502879078694, "grad_norm": 1.723091175178344, "learning_rate": 1.708753246754624e-05, "loss": 0.4026, "step": 4248 }, { "epoch": 2.7184900831733847, "grad_norm": 1.7422624007724588, "learning_rate": 1.708607050095346e-05, "loss": 0.3859, "step": 4249 }, { "epoch": 2.7191298784388995, "grad_norm": 1.920168426144395, "learning_rate": 1.7084608230097423e-05, "loss": 0.4257, "step": 4250 }, { "epoch": 2.7197696737044144, "grad_norm": 1.7843638426155934, "learning_rate": 1.7083145655040904e-05, "loss": 0.3973, "step": 4251 }, { "epoch": 2.7204094689699296, "grad_norm": 1.8995110280286853, "learning_rate": 1.7081682775846716e-05, "loss": 0.4361, "step": 4252 }, { "epoch": 2.721049264235445, "grad_norm": 1.7946139335186349, "learning_rate": 1.708021959257766e-05, "loss": 0.387, "step": 4253 }, { "epoch": 2.7216890595009597, "grad_norm": 1.7403194749334794, "learning_rate": 1.707875610529658e-05, "loss": 0.3924, "step": 4254 }, { "epoch": 2.7223288547664746, "grad_norm": 1.8612195646518757, "learning_rate": 1.70772923140663e-05, "loss": 0.3821, "step": 4255 }, { "epoch": 2.72296865003199, "grad_norm": 1.8596850350815424, "learning_rate": 1.7075828218949675e-05, "loss": 0.4211, "step": 4256 }, { "epoch": 2.7236084452975047, "grad_norm": 1.8150942574304416, "learning_rate": 1.7074363820009577e-05, "loss": 0.4492, "step": 4257 }, { "epoch": 2.72424824056302, "grad_norm": 1.8477247934303194, "learning_rate": 1.7072899117308877e-05, "loss": 0.4013, "step": 4258 }, { "epoch": 2.724888035828535, "grad_norm": 1.9077282621630034, "learning_rate": 1.7071434110910474e-05, "loss": 0.4409, "step": 4259 }, { "epoch": 2.72552783109405, "grad_norm": 1.8428243961586588, "learning_rate": 1.7069968800877265e-05, "loss": 0.3987, "step": 4260 }, { "epoch": 2.726167626359565, "grad_norm": 1.8434376021475833, "learning_rate": 1.7068503187272177e-05, "loss": 0.4176, "step": 4261 }, { "epoch": 2.7268074216250797, "grad_norm": 1.757380577774949, "learning_rate": 1.7067037270158135e-05, "loss": 0.4037, "step": 4262 }, { "epoch": 2.727447216890595, "grad_norm": 1.8810210293041516, "learning_rate": 1.706557104959808e-05, "loss": 0.5008, "step": 4263 }, { "epoch": 2.7280870121561103, "grad_norm": 1.7016491848239161, "learning_rate": 1.7064104525654982e-05, "loss": 0.3633, "step": 4264 }, { "epoch": 2.728726807421625, "grad_norm": 1.8147570797885275, "learning_rate": 1.7062637698391793e-05, "loss": 0.4177, "step": 4265 }, { "epoch": 2.72936660268714, "grad_norm": 1.9012859954867358, "learning_rate": 1.7061170567871504e-05, "loss": 0.3663, "step": 4266 }, { "epoch": 2.730006397952655, "grad_norm": 1.7288731466354632, "learning_rate": 1.705970313415712e-05, "loss": 0.3454, "step": 4267 }, { "epoch": 2.73064619321817, "grad_norm": 2.088433590161815, "learning_rate": 1.7058235397311637e-05, "loss": 0.4374, "step": 4268 }, { "epoch": 2.7312859884836853, "grad_norm": 1.7786290120387105, "learning_rate": 1.705676735739808e-05, "loss": 0.4052, "step": 4269 }, { "epoch": 2.7319257837492, "grad_norm": 1.926176160315549, "learning_rate": 1.7055299014479486e-05, "loss": 0.3813, "step": 4270 }, { "epoch": 2.7325655790147154, "grad_norm": 2.0372277003227146, "learning_rate": 1.7053830368618904e-05, "loss": 0.4311, "step": 4271 }, { "epoch": 2.7332053742802302, "grad_norm": 1.8493200997205212, "learning_rate": 1.7052361419879395e-05, "loss": 0.3975, "step": 4272 }, { "epoch": 2.7338451695457455, "grad_norm": 1.8394904320120802, "learning_rate": 1.7050892168324028e-05, "loss": 0.4533, "step": 4273 }, { "epoch": 2.7344849648112604, "grad_norm": 1.9292433370945954, "learning_rate": 1.7049422614015898e-05, "loss": 0.403, "step": 4274 }, { "epoch": 2.7351247600767756, "grad_norm": 1.824560387081592, "learning_rate": 1.7047952757018097e-05, "loss": 0.4259, "step": 4275 }, { "epoch": 2.7357645553422905, "grad_norm": 2.002531705664439, "learning_rate": 1.7046482597393745e-05, "loss": 0.3979, "step": 4276 }, { "epoch": 2.7364043506078053, "grad_norm": 1.8231036829677247, "learning_rate": 1.7045012135205964e-05, "loss": 0.3586, "step": 4277 }, { "epoch": 2.7370441458733206, "grad_norm": 1.6704424423244741, "learning_rate": 1.7043541370517896e-05, "loss": 0.4735, "step": 4278 }, { "epoch": 2.737683941138836, "grad_norm": 1.9798563792543131, "learning_rate": 1.704207030339269e-05, "loss": 0.4202, "step": 4279 }, { "epoch": 2.7383237364043507, "grad_norm": 1.721854498657253, "learning_rate": 1.704059893389351e-05, "loss": 0.3767, "step": 4280 }, { "epoch": 2.7389635316698655, "grad_norm": 1.6736898192930851, "learning_rate": 1.703912726208354e-05, "loss": 0.4177, "step": 4281 }, { "epoch": 2.739603326935381, "grad_norm": 1.7754217416674347, "learning_rate": 1.7037655288025962e-05, "loss": 0.4122, "step": 4282 }, { "epoch": 2.7402431222008956, "grad_norm": 1.6809650385875206, "learning_rate": 1.7036183011783985e-05, "loss": 0.3728, "step": 4283 }, { "epoch": 2.740882917466411, "grad_norm": 1.7581656728783177, "learning_rate": 1.703471043342083e-05, "loss": 0.3637, "step": 4284 }, { "epoch": 2.7415227127319257, "grad_norm": 1.7311465895433105, "learning_rate": 1.703323755299972e-05, "loss": 0.3544, "step": 4285 }, { "epoch": 2.742162507997441, "grad_norm": 1.8125252647384398, "learning_rate": 1.70317643705839e-05, "loss": 0.4345, "step": 4286 }, { "epoch": 2.742802303262956, "grad_norm": 1.6956595061194473, "learning_rate": 1.703029088623663e-05, "loss": 0.399, "step": 4287 }, { "epoch": 2.7434420985284707, "grad_norm": 1.714150278401996, "learning_rate": 1.7028817100021173e-05, "loss": 0.3621, "step": 4288 }, { "epoch": 2.744081893793986, "grad_norm": 1.8680762442049277, "learning_rate": 1.7027343012000813e-05, "loss": 0.3795, "step": 4289 }, { "epoch": 2.744721689059501, "grad_norm": 1.8118579266908994, "learning_rate": 1.7025868622238844e-05, "loss": 0.3947, "step": 4290 }, { "epoch": 2.745361484325016, "grad_norm": 2.003597479309518, "learning_rate": 1.7024393930798577e-05, "loss": 0.4335, "step": 4291 }, { "epoch": 2.746001279590531, "grad_norm": 1.837734054641697, "learning_rate": 1.7022918937743327e-05, "loss": 0.4307, "step": 4292 }, { "epoch": 2.746641074856046, "grad_norm": 1.8133058610253194, "learning_rate": 1.7021443643136434e-05, "loss": 0.3737, "step": 4293 }, { "epoch": 2.747280870121561, "grad_norm": 1.7553497617000766, "learning_rate": 1.7019968047041237e-05, "loss": 0.4103, "step": 4294 }, { "epoch": 2.7479206653870762, "grad_norm": 1.731408025722716, "learning_rate": 1.7018492149521103e-05, "loss": 0.4217, "step": 4295 }, { "epoch": 2.748560460652591, "grad_norm": 1.6658392791513341, "learning_rate": 1.70170159506394e-05, "loss": 0.4294, "step": 4296 }, { "epoch": 2.7492002559181064, "grad_norm": 1.731868077054078, "learning_rate": 1.7015539450459514e-05, "loss": 0.3864, "step": 4297 }, { "epoch": 2.749840051183621, "grad_norm": 1.8721048547659467, "learning_rate": 1.7014062649044843e-05, "loss": 0.4238, "step": 4298 }, { "epoch": 2.750479846449136, "grad_norm": 1.8483655015870095, "learning_rate": 1.70125855464588e-05, "loss": 0.409, "step": 4299 }, { "epoch": 2.7511196417146513, "grad_norm": 1.8627425820373098, "learning_rate": 1.701110814276481e-05, "loss": 0.4524, "step": 4300 }, { "epoch": 2.7517594369801666, "grad_norm": 1.7486540627426375, "learning_rate": 1.7009630438026303e-05, "loss": 0.3567, "step": 4301 }, { "epoch": 2.7523992322456814, "grad_norm": 1.8655972067670208, "learning_rate": 1.700815243230674e-05, "loss": 0.4342, "step": 4302 }, { "epoch": 2.7530390275111962, "grad_norm": 1.8290582391275507, "learning_rate": 1.7006674125669572e-05, "loss": 0.4073, "step": 4303 }, { "epoch": 2.7536788227767115, "grad_norm": 1.8176133529685956, "learning_rate": 1.7005195518178285e-05, "loss": 0.3902, "step": 4304 }, { "epoch": 2.7543186180422263, "grad_norm": 1.7700621887546675, "learning_rate": 1.700371660989636e-05, "loss": 0.3987, "step": 4305 }, { "epoch": 2.7549584133077416, "grad_norm": 1.7190743088567884, "learning_rate": 1.700223740088731e-05, "loss": 0.4148, "step": 4306 }, { "epoch": 2.7555982085732564, "grad_norm": 1.9396487161305815, "learning_rate": 1.7000757891214637e-05, "loss": 0.4384, "step": 4307 }, { "epoch": 2.7562380038387717, "grad_norm": 1.7219899737923219, "learning_rate": 1.6999278080941874e-05, "loss": 0.4, "step": 4308 }, { "epoch": 2.7568777991042865, "grad_norm": 1.6749528390664268, "learning_rate": 1.6997797970132557e-05, "loss": 0.3833, "step": 4309 }, { "epoch": 2.757517594369802, "grad_norm": 1.639450929339447, "learning_rate": 1.699631755885025e-05, "loss": 0.4246, "step": 4310 }, { "epoch": 2.7581573896353166, "grad_norm": 1.8625607113076248, "learning_rate": 1.6994836847158512e-05, "loss": 0.445, "step": 4311 }, { "epoch": 2.758797184900832, "grad_norm": 1.8216427223695224, "learning_rate": 1.699335583512092e-05, "loss": 0.3974, "step": 4312 }, { "epoch": 2.7594369801663468, "grad_norm": 1.8231707861675717, "learning_rate": 1.6991874522801066e-05, "loss": 0.4, "step": 4313 }, { "epoch": 2.7600767754318616, "grad_norm": 1.7849249830925975, "learning_rate": 1.699039291026256e-05, "loss": 0.4033, "step": 4314 }, { "epoch": 2.760716570697377, "grad_norm": 1.7027602941418898, "learning_rate": 1.6988910997569018e-05, "loss": 0.4028, "step": 4315 }, { "epoch": 2.761356365962892, "grad_norm": 1.7878334374924194, "learning_rate": 1.6987428784784072e-05, "loss": 0.3982, "step": 4316 }, { "epoch": 2.761996161228407, "grad_norm": 1.6690047497832914, "learning_rate": 1.698594627197136e-05, "loss": 0.361, "step": 4317 }, { "epoch": 2.762635956493922, "grad_norm": 1.7082920642796389, "learning_rate": 1.6984463459194543e-05, "loss": 0.375, "step": 4318 }, { "epoch": 2.763275751759437, "grad_norm": 1.5984955824480274, "learning_rate": 1.698298034651729e-05, "loss": 0.3164, "step": 4319 }, { "epoch": 2.763915547024952, "grad_norm": 1.7839189887575246, "learning_rate": 1.6981496934003284e-05, "loss": 0.4063, "step": 4320 }, { "epoch": 2.764555342290467, "grad_norm": 1.829316835667242, "learning_rate": 1.6980013221716216e-05, "loss": 0.4822, "step": 4321 }, { "epoch": 2.765195137555982, "grad_norm": 1.907144750844032, "learning_rate": 1.6978529209719797e-05, "loss": 0.3749, "step": 4322 }, { "epoch": 2.7658349328214973, "grad_norm": 1.6623371631358976, "learning_rate": 1.6977044898077752e-05, "loss": 0.3332, "step": 4323 }, { "epoch": 2.766474728087012, "grad_norm": 1.8012122614186241, "learning_rate": 1.6975560286853807e-05, "loss": 0.3742, "step": 4324 }, { "epoch": 2.767114523352527, "grad_norm": 1.7377092460492116, "learning_rate": 1.697407537611171e-05, "loss": 0.4263, "step": 4325 }, { "epoch": 2.767754318618042, "grad_norm": 1.546311245284248, "learning_rate": 1.6972590165915228e-05, "loss": 0.3364, "step": 4326 }, { "epoch": 2.7683941138835575, "grad_norm": 1.9699947356755665, "learning_rate": 1.6971104656328123e-05, "loss": 0.379, "step": 4327 }, { "epoch": 2.7690339091490723, "grad_norm": 1.8705555455128275, "learning_rate": 1.6969618847414183e-05, "loss": 0.3877, "step": 4328 }, { "epoch": 2.769673704414587, "grad_norm": 1.740497596259295, "learning_rate": 1.6968132739237212e-05, "loss": 0.4431, "step": 4329 }, { "epoch": 2.7703134996801024, "grad_norm": 1.737915287208973, "learning_rate": 1.6966646331861014e-05, "loss": 0.356, "step": 4330 }, { "epoch": 2.7709532949456173, "grad_norm": 1.8283816793535788, "learning_rate": 1.6965159625349418e-05, "loss": 0.4151, "step": 4331 }, { "epoch": 2.7715930902111325, "grad_norm": 1.6404964734054575, "learning_rate": 1.6963672619766254e-05, "loss": 0.3738, "step": 4332 }, { "epoch": 2.7722328854766474, "grad_norm": 1.777414339804438, "learning_rate": 1.6962185315175377e-05, "loss": 0.3848, "step": 4333 }, { "epoch": 2.7728726807421626, "grad_norm": 1.7920629765602736, "learning_rate": 1.6960697711640647e-05, "loss": 0.3648, "step": 4334 }, { "epoch": 2.7735124760076775, "grad_norm": 1.8689722345219666, "learning_rate": 1.695920980922594e-05, "loss": 0.3733, "step": 4335 }, { "epoch": 2.7741522712731923, "grad_norm": 1.7045089800611222, "learning_rate": 1.695772160799514e-05, "loss": 0.3942, "step": 4336 }, { "epoch": 2.7747920665387076, "grad_norm": 1.7385760504168601, "learning_rate": 1.6956233108012156e-05, "loss": 0.4012, "step": 4337 }, { "epoch": 2.775431861804223, "grad_norm": 1.5708677546026573, "learning_rate": 1.6954744309340894e-05, "loss": 0.3332, "step": 4338 }, { "epoch": 2.7760716570697377, "grad_norm": 1.7527346854343866, "learning_rate": 1.6953255212045282e-05, "loss": 0.3795, "step": 4339 }, { "epoch": 2.7767114523352525, "grad_norm": 1.761130981482091, "learning_rate": 1.6951765816189262e-05, "loss": 0.3686, "step": 4340 }, { "epoch": 2.777351247600768, "grad_norm": 1.8742611947690218, "learning_rate": 1.6950276121836786e-05, "loss": 0.4209, "step": 4341 }, { "epoch": 2.777991042866283, "grad_norm": 1.8689223850234387, "learning_rate": 1.694878612905181e-05, "loss": 0.4204, "step": 4342 }, { "epoch": 2.778630838131798, "grad_norm": 1.8345072863862701, "learning_rate": 1.6947295837898323e-05, "loss": 0.4134, "step": 4343 }, { "epoch": 2.7792706333973127, "grad_norm": 1.8783268367332366, "learning_rate": 1.6945805248440313e-05, "loss": 0.4609, "step": 4344 }, { "epoch": 2.779910428662828, "grad_norm": 1.7830026614643002, "learning_rate": 1.694431436074178e-05, "loss": 0.4111, "step": 4345 }, { "epoch": 2.780550223928343, "grad_norm": 1.873314559695879, "learning_rate": 1.694282317486674e-05, "loss": 0.3773, "step": 4346 }, { "epoch": 2.781190019193858, "grad_norm": 1.720194786460445, "learning_rate": 1.6941331690879223e-05, "loss": 0.3866, "step": 4347 }, { "epoch": 2.781829814459373, "grad_norm": 1.7879672580486596, "learning_rate": 1.693983990884327e-05, "loss": 0.4473, "step": 4348 }, { "epoch": 2.782469609724888, "grad_norm": 1.785748498688766, "learning_rate": 1.693834782882294e-05, "loss": 0.4091, "step": 4349 }, { "epoch": 2.783109404990403, "grad_norm": 1.775724043086051, "learning_rate": 1.6936855450882293e-05, "loss": 0.3849, "step": 4350 }, { "epoch": 2.783749200255918, "grad_norm": 1.7424182574236764, "learning_rate": 1.6935362775085416e-05, "loss": 0.3793, "step": 4351 }, { "epoch": 2.784388995521433, "grad_norm": 1.9267106755857513, "learning_rate": 1.6933869801496398e-05, "loss": 0.3852, "step": 4352 }, { "epoch": 2.7850287907869484, "grad_norm": 1.58678504782333, "learning_rate": 1.693237653017934e-05, "loss": 0.3764, "step": 4353 }, { "epoch": 2.7856685860524633, "grad_norm": 1.7794289917703605, "learning_rate": 1.693088296119837e-05, "loss": 0.3486, "step": 4354 }, { "epoch": 2.786308381317978, "grad_norm": 1.6671514912441203, "learning_rate": 1.6929389094617613e-05, "loss": 0.3688, "step": 4355 }, { "epoch": 2.7869481765834934, "grad_norm": 1.8625322736500023, "learning_rate": 1.6927894930501215e-05, "loss": 0.4687, "step": 4356 }, { "epoch": 2.787587971849008, "grad_norm": 1.6902994134328209, "learning_rate": 1.6926400468913334e-05, "loss": 0.3916, "step": 4357 }, { "epoch": 2.7882277671145235, "grad_norm": 1.8430762468395916, "learning_rate": 1.692490570991814e-05, "loss": 0.3666, "step": 4358 }, { "epoch": 2.7888675623800383, "grad_norm": 1.9074007971715623, "learning_rate": 1.692341065357981e-05, "loss": 0.4126, "step": 4359 }, { "epoch": 2.7895073576455536, "grad_norm": 1.71266116978192, "learning_rate": 1.6921915299962538e-05, "loss": 0.3831, "step": 4360 }, { "epoch": 2.7901471529110684, "grad_norm": 1.9208915382511809, "learning_rate": 1.6920419649130543e-05, "loss": 0.4223, "step": 4361 }, { "epoch": 2.7907869481765832, "grad_norm": 1.8338823461012805, "learning_rate": 1.6918923701148037e-05, "loss": 0.4036, "step": 4362 }, { "epoch": 2.7914267434420985, "grad_norm": 1.8171704135161217, "learning_rate": 1.6917427456079255e-05, "loss": 0.3819, "step": 4363 }, { "epoch": 2.792066538707614, "grad_norm": 1.947995036696716, "learning_rate": 1.6915930913988442e-05, "loss": 0.4812, "step": 4364 }, { "epoch": 2.7927063339731286, "grad_norm": 1.8249764250886475, "learning_rate": 1.691443407493986e-05, "loss": 0.3975, "step": 4365 }, { "epoch": 2.7933461292386434, "grad_norm": 1.7696169791216863, "learning_rate": 1.6912936938997778e-05, "loss": 0.4031, "step": 4366 }, { "epoch": 2.7939859245041587, "grad_norm": 1.704195473855155, "learning_rate": 1.691143950622648e-05, "loss": 0.3744, "step": 4367 }, { "epoch": 2.7946257197696736, "grad_norm": 1.7346896792325728, "learning_rate": 1.6909941776690264e-05, "loss": 0.4443, "step": 4368 }, { "epoch": 2.795265515035189, "grad_norm": 1.8103822445538895, "learning_rate": 1.690844375045344e-05, "loss": 0.4052, "step": 4369 }, { "epoch": 2.7959053103007037, "grad_norm": 1.8347140205982713, "learning_rate": 1.690694542758033e-05, "loss": 0.3901, "step": 4370 }, { "epoch": 2.796545105566219, "grad_norm": 1.9191759487266349, "learning_rate": 1.690544680813527e-05, "loss": 0.4002, "step": 4371 }, { "epoch": 2.7971849008317338, "grad_norm": 1.6775225034996557, "learning_rate": 1.690394789218261e-05, "loss": 0.3563, "step": 4372 }, { "epoch": 2.797824696097249, "grad_norm": 1.7760874314306119, "learning_rate": 1.690244867978671e-05, "loss": 0.3638, "step": 4373 }, { "epoch": 2.798464491362764, "grad_norm": 1.7701442668670433, "learning_rate": 1.6900949171011937e-05, "loss": 0.4002, "step": 4374 }, { "epoch": 2.799104286628279, "grad_norm": 1.6755618546293043, "learning_rate": 1.6899449365922685e-05, "loss": 0.3416, "step": 4375 }, { "epoch": 2.799744081893794, "grad_norm": 1.7306169212357347, "learning_rate": 1.6897949264583352e-05, "loss": 0.3996, "step": 4376 }, { "epoch": 2.800383877159309, "grad_norm": 1.6475217290871673, "learning_rate": 1.6896448867058345e-05, "loss": 0.4195, "step": 4377 }, { "epoch": 2.801023672424824, "grad_norm": 1.721066512072694, "learning_rate": 1.6894948173412094e-05, "loss": 0.3635, "step": 4378 }, { "epoch": 2.8016634676903394, "grad_norm": 1.6719805817049183, "learning_rate": 1.6893447183709034e-05, "loss": 0.3723, "step": 4379 }, { "epoch": 2.802303262955854, "grad_norm": 1.7235663973040567, "learning_rate": 1.689194589801361e-05, "loss": 0.3965, "step": 4380 }, { "epoch": 2.802943058221369, "grad_norm": 1.76175313435798, "learning_rate": 1.6890444316390296e-05, "loss": 0.3973, "step": 4381 }, { "epoch": 2.8035828534868843, "grad_norm": 1.7442969553564676, "learning_rate": 1.6888942438903554e-05, "loss": 0.3817, "step": 4382 }, { "epoch": 2.804222648752399, "grad_norm": 1.6758333353453643, "learning_rate": 1.6887440265617884e-05, "loss": 0.3756, "step": 4383 }, { "epoch": 2.8048624440179144, "grad_norm": 1.7703871427851818, "learning_rate": 1.6885937796597775e-05, "loss": 0.3698, "step": 4384 }, { "epoch": 2.8055022392834292, "grad_norm": 1.7380387563884818, "learning_rate": 1.6884435031907748e-05, "loss": 0.4246, "step": 4385 }, { "epoch": 2.8061420345489445, "grad_norm": 1.7436481252470828, "learning_rate": 1.6882931971612326e-05, "loss": 0.3923, "step": 4386 }, { "epoch": 2.8067818298144593, "grad_norm": 1.6524408733771339, "learning_rate": 1.688142861577605e-05, "loss": 0.4027, "step": 4387 }, { "epoch": 2.807421625079974, "grad_norm": 1.8010735299903478, "learning_rate": 1.6879924964463474e-05, "loss": 0.3712, "step": 4388 }, { "epoch": 2.8080614203454894, "grad_norm": 1.8517273535497598, "learning_rate": 1.6878421017739152e-05, "loss": 0.4133, "step": 4389 }, { "epoch": 2.8087012156110047, "grad_norm": 1.9026877545182612, "learning_rate": 1.6876916775667672e-05, "loss": 0.4129, "step": 4390 }, { "epoch": 2.8093410108765196, "grad_norm": 1.5965450527739216, "learning_rate": 1.6875412238313615e-05, "loss": 0.3941, "step": 4391 }, { "epoch": 2.8099808061420344, "grad_norm": 1.7225167508300154, "learning_rate": 1.6873907405741593e-05, "loss": 0.4073, "step": 4392 }, { "epoch": 2.8106206014075497, "grad_norm": 1.8528391668851687, "learning_rate": 1.687240227801621e-05, "loss": 0.4269, "step": 4393 }, { "epoch": 2.8112603966730645, "grad_norm": 1.712798563023517, "learning_rate": 1.68708968552021e-05, "loss": 0.3876, "step": 4394 }, { "epoch": 2.8119001919385798, "grad_norm": 1.70301320748217, "learning_rate": 1.68693911373639e-05, "loss": 0.4002, "step": 4395 }, { "epoch": 2.8125399872040946, "grad_norm": 1.714912305079844, "learning_rate": 1.6867885124566265e-05, "loss": 0.3523, "step": 4396 }, { "epoch": 2.81317978246961, "grad_norm": 1.7747031128170563, "learning_rate": 1.6866378816873856e-05, "loss": 0.3948, "step": 4397 }, { "epoch": 2.8138195777351247, "grad_norm": 1.6425096784721185, "learning_rate": 1.6864872214351362e-05, "loss": 0.4007, "step": 4398 }, { "epoch": 2.8144593730006395, "grad_norm": 1.6303864745704413, "learning_rate": 1.686336531706346e-05, "loss": 0.387, "step": 4399 }, { "epoch": 2.815099168266155, "grad_norm": 1.8801222063976752, "learning_rate": 1.6861858125074864e-05, "loss": 0.405, "step": 4400 }, { "epoch": 2.81573896353167, "grad_norm": 1.797338383246518, "learning_rate": 1.686035063845029e-05, "loss": 0.3979, "step": 4401 }, { "epoch": 2.816378758797185, "grad_norm": 1.7234189993570705, "learning_rate": 1.6858842857254457e-05, "loss": 0.3923, "step": 4402 }, { "epoch": 2.8170185540626997, "grad_norm": 1.73365295010304, "learning_rate": 1.6857334781552115e-05, "loss": 0.3783, "step": 4403 }, { "epoch": 2.817658349328215, "grad_norm": 1.7749884352674887, "learning_rate": 1.6855826411408016e-05, "loss": 0.3833, "step": 4404 }, { "epoch": 2.81829814459373, "grad_norm": 1.8017304974246284, "learning_rate": 1.6854317746886923e-05, "loss": 0.3382, "step": 4405 }, { "epoch": 2.818937939859245, "grad_norm": 1.7442595590900671, "learning_rate": 1.6852808788053624e-05, "loss": 0.4059, "step": 4406 }, { "epoch": 2.81957773512476, "grad_norm": 1.755576190689157, "learning_rate": 1.6851299534972904e-05, "loss": 0.3901, "step": 4407 }, { "epoch": 2.8202175303902752, "grad_norm": 1.8004884622706843, "learning_rate": 1.6849789987709567e-05, "loss": 0.3988, "step": 4408 }, { "epoch": 2.82085732565579, "grad_norm": 1.6906398431385596, "learning_rate": 1.6848280146328436e-05, "loss": 0.3485, "step": 4409 }, { "epoch": 2.8214971209213053, "grad_norm": 1.751614627022824, "learning_rate": 1.6846770010894335e-05, "loss": 0.3735, "step": 4410 }, { "epoch": 2.82213691618682, "grad_norm": 1.8164771613849202, "learning_rate": 1.6845259581472113e-05, "loss": 0.4564, "step": 4411 }, { "epoch": 2.8227767114523354, "grad_norm": 1.7777450065426843, "learning_rate": 1.684374885812662e-05, "loss": 0.35, "step": 4412 }, { "epoch": 2.8234165067178503, "grad_norm": 1.589732619373658, "learning_rate": 1.6842237840922723e-05, "loss": 0.3452, "step": 4413 }, { "epoch": 2.824056301983365, "grad_norm": 1.7422857880884335, "learning_rate": 1.6840726529925303e-05, "loss": 0.3357, "step": 4414 }, { "epoch": 2.8246960972488804, "grad_norm": 1.7645610118926827, "learning_rate": 1.683921492519926e-05, "loss": 0.3408, "step": 4415 }, { "epoch": 2.8253358925143957, "grad_norm": 1.7893834489056795, "learning_rate": 1.6837703026809492e-05, "loss": 0.4008, "step": 4416 }, { "epoch": 2.8259756877799105, "grad_norm": 1.7827218719203684, "learning_rate": 1.6836190834820917e-05, "loss": 0.3726, "step": 4417 }, { "epoch": 2.8266154830454253, "grad_norm": 1.7517027441855266, "learning_rate": 1.683467834929847e-05, "loss": 0.3919, "step": 4418 }, { "epoch": 2.8272552783109406, "grad_norm": 1.8529872079535976, "learning_rate": 1.683316557030709e-05, "loss": 0.3881, "step": 4419 }, { "epoch": 2.8278950735764554, "grad_norm": 1.725477586996637, "learning_rate": 1.6831652497911742e-05, "loss": 0.4162, "step": 4420 }, { "epoch": 2.8285348688419707, "grad_norm": 1.8395672252160458, "learning_rate": 1.6830139132177384e-05, "loss": 0.3823, "step": 4421 }, { "epoch": 2.8291746641074855, "grad_norm": 1.7186619940825802, "learning_rate": 1.6828625473169e-05, "loss": 0.3756, "step": 4422 }, { "epoch": 2.829814459373001, "grad_norm": 1.667924767526383, "learning_rate": 1.6827111520951592e-05, "loss": 0.342, "step": 4423 }, { "epoch": 2.8304542546385156, "grad_norm": 1.6286425368913449, "learning_rate": 1.6825597275590154e-05, "loss": 0.4078, "step": 4424 }, { "epoch": 2.8310940499040305, "grad_norm": 1.7305893333627929, "learning_rate": 1.682408273714971e-05, "loss": 0.3418, "step": 4425 }, { "epoch": 2.8317338451695457, "grad_norm": 1.6209879862578298, "learning_rate": 1.6822567905695295e-05, "loss": 0.3567, "step": 4426 }, { "epoch": 2.832373640435061, "grad_norm": 1.5364174482126236, "learning_rate": 1.6821052781291948e-05, "loss": 0.3697, "step": 4427 }, { "epoch": 2.833013435700576, "grad_norm": 2.071824332389436, "learning_rate": 1.6819537364004733e-05, "loss": 0.4525, "step": 4428 }, { "epoch": 2.8336532309660907, "grad_norm": 1.8555197075280023, "learning_rate": 1.681802165389871e-05, "loss": 0.4498, "step": 4429 }, { "epoch": 2.834293026231606, "grad_norm": 1.6019328437486111, "learning_rate": 1.6816505651038967e-05, "loss": 0.3695, "step": 4430 }, { "epoch": 2.834932821497121, "grad_norm": 1.8125597251498011, "learning_rate": 1.68149893554906e-05, "loss": 0.4031, "step": 4431 }, { "epoch": 2.835572616762636, "grad_norm": 1.6787036441684569, "learning_rate": 1.6813472767318708e-05, "loss": 0.4259, "step": 4432 }, { "epoch": 2.836212412028151, "grad_norm": 1.6387742985996896, "learning_rate": 1.681195588658842e-05, "loss": 0.3839, "step": 4433 }, { "epoch": 2.836852207293666, "grad_norm": 1.8570478123974758, "learning_rate": 1.681043871336486e-05, "loss": 0.4096, "step": 4434 }, { "epoch": 2.837492002559181, "grad_norm": 1.8756948321683946, "learning_rate": 1.6808921247713178e-05, "loss": 0.3606, "step": 4435 }, { "epoch": 2.838131797824696, "grad_norm": 1.7299777162334418, "learning_rate": 1.680740348969853e-05, "loss": 0.3677, "step": 4436 }, { "epoch": 2.838771593090211, "grad_norm": 1.6624673266988166, "learning_rate": 1.6805885439386083e-05, "loss": 0.3803, "step": 4437 }, { "epoch": 2.8394113883557264, "grad_norm": 1.7829678883356115, "learning_rate": 1.6804367096841027e-05, "loss": 0.3478, "step": 4438 }, { "epoch": 2.840051183621241, "grad_norm": 1.821585911421057, "learning_rate": 1.680284846212855e-05, "loss": 0.3475, "step": 4439 }, { "epoch": 2.840690978886756, "grad_norm": 1.7657240285069562, "learning_rate": 1.680132953531386e-05, "loss": 0.3632, "step": 4440 }, { "epoch": 2.8413307741522713, "grad_norm": 1.6970465569664217, "learning_rate": 1.679981031646218e-05, "loss": 0.3451, "step": 4441 }, { "epoch": 2.8419705694177866, "grad_norm": 1.878344015342109, "learning_rate": 1.679829080563874e-05, "loss": 0.3962, "step": 4442 }, { "epoch": 2.8426103646833014, "grad_norm": 1.6617934515973716, "learning_rate": 1.6796771002908788e-05, "loss": 0.3291, "step": 4443 }, { "epoch": 2.8432501599488162, "grad_norm": 1.665992620710727, "learning_rate": 1.6795250908337575e-05, "loss": 0.3748, "step": 4444 }, { "epoch": 2.8438899552143315, "grad_norm": 1.7053275096880827, "learning_rate": 1.679373052199038e-05, "loss": 0.3968, "step": 4445 }, { "epoch": 2.8445297504798464, "grad_norm": 1.9638093342198122, "learning_rate": 1.6792209843932478e-05, "loss": 0.3981, "step": 4446 }, { "epoch": 2.8451695457453616, "grad_norm": 1.8324124914934454, "learning_rate": 1.679068887422917e-05, "loss": 0.3952, "step": 4447 }, { "epoch": 2.8458093410108765, "grad_norm": 1.586584384268535, "learning_rate": 1.678916761294576e-05, "loss": 0.3721, "step": 4448 }, { "epoch": 2.8464491362763917, "grad_norm": 1.8322139712932384, "learning_rate": 1.6787646060147573e-05, "loss": 0.3925, "step": 4449 }, { "epoch": 2.8470889315419066, "grad_norm": 1.77202881371447, "learning_rate": 1.6786124215899934e-05, "loss": 0.3845, "step": 4450 }, { "epoch": 2.8477287268074214, "grad_norm": 1.6999399625536404, "learning_rate": 1.6784602080268197e-05, "loss": 0.3417, "step": 4451 }, { "epoch": 2.8483685220729367, "grad_norm": 1.6383293739663398, "learning_rate": 1.6783079653317713e-05, "loss": 0.3751, "step": 4452 }, { "epoch": 2.849008317338452, "grad_norm": 1.6968293840127138, "learning_rate": 1.6781556935113855e-05, "loss": 0.3623, "step": 4453 }, { "epoch": 2.8496481126039668, "grad_norm": 1.7461670358672978, "learning_rate": 1.6780033925722005e-05, "loss": 0.3478, "step": 4454 }, { "epoch": 2.8502879078694816, "grad_norm": 2.059198153569621, "learning_rate": 1.6778510625207564e-05, "loss": 0.4276, "step": 4455 }, { "epoch": 2.850927703134997, "grad_norm": 1.7510337940727823, "learning_rate": 1.677698703363593e-05, "loss": 0.3988, "step": 4456 }, { "epoch": 2.8515674984005117, "grad_norm": 1.6988801668838636, "learning_rate": 1.6775463151072525e-05, "loss": 0.3775, "step": 4457 }, { "epoch": 2.852207293666027, "grad_norm": 1.7231388852854204, "learning_rate": 1.677393897758279e-05, "loss": 0.3866, "step": 4458 }, { "epoch": 2.852847088931542, "grad_norm": 1.7713452226107507, "learning_rate": 1.6772414513232164e-05, "loss": 0.3521, "step": 4459 }, { "epoch": 2.853486884197057, "grad_norm": 1.7062490058616406, "learning_rate": 1.6770889758086108e-05, "loss": 0.3327, "step": 4460 }, { "epoch": 2.854126679462572, "grad_norm": 1.688891717975907, "learning_rate": 1.6769364712210087e-05, "loss": 0.356, "step": 4461 }, { "epoch": 2.8547664747280868, "grad_norm": 1.9000526225413397, "learning_rate": 1.6767839375669587e-05, "loss": 0.445, "step": 4462 }, { "epoch": 2.855406269993602, "grad_norm": 1.734173752359921, "learning_rate": 1.6766313748530105e-05, "loss": 0.4053, "step": 4463 }, { "epoch": 2.8560460652591173, "grad_norm": 1.6846504146789794, "learning_rate": 1.676478783085715e-05, "loss": 0.3458, "step": 4464 }, { "epoch": 2.856685860524632, "grad_norm": 1.7961978825436273, "learning_rate": 1.6763261622716233e-05, "loss": 0.3992, "step": 4465 }, { "epoch": 2.857325655790147, "grad_norm": 1.8537345357912298, "learning_rate": 1.6761735124172894e-05, "loss": 0.3817, "step": 4466 }, { "epoch": 2.8579654510556622, "grad_norm": 1.7461834245412653, "learning_rate": 1.676020833529268e-05, "loss": 0.3811, "step": 4467 }, { "epoch": 2.858605246321177, "grad_norm": 1.7536925131929273, "learning_rate": 1.6758681256141144e-05, "loss": 0.3678, "step": 4468 }, { "epoch": 2.8592450415866923, "grad_norm": 1.7116869605539642, "learning_rate": 1.675715388678386e-05, "loss": 0.3335, "step": 4469 }, { "epoch": 2.859884836852207, "grad_norm": 1.6904519437450367, "learning_rate": 1.6755626227286404e-05, "loss": 0.3932, "step": 4470 }, { "epoch": 2.8605246321177225, "grad_norm": 1.699676573538381, "learning_rate": 1.6754098277714376e-05, "loss": 0.3652, "step": 4471 }, { "epoch": 2.8611644273832373, "grad_norm": 1.6386909440153832, "learning_rate": 1.6752570038133383e-05, "loss": 0.3785, "step": 4472 }, { "epoch": 2.8618042226487526, "grad_norm": 1.8807546796287824, "learning_rate": 1.6751041508609046e-05, "loss": 0.3908, "step": 4473 }, { "epoch": 2.8624440179142674, "grad_norm": 1.7966076622724463, "learning_rate": 1.6749512689206995e-05, "loss": 0.3904, "step": 4474 }, { "epoch": 2.8630838131797827, "grad_norm": 1.6684471964846055, "learning_rate": 1.6747983579992878e-05, "loss": 0.3792, "step": 4475 }, { "epoch": 2.8637236084452975, "grad_norm": 1.732436442977426, "learning_rate": 1.6746454181032344e-05, "loss": 0.4113, "step": 4476 }, { "epoch": 2.8643634037108123, "grad_norm": 1.730828267256522, "learning_rate": 1.6744924492391072e-05, "loss": 0.3985, "step": 4477 }, { "epoch": 2.8650031989763276, "grad_norm": 1.700064234655729, "learning_rate": 1.674339451413474e-05, "loss": 0.4025, "step": 4478 }, { "epoch": 2.865642994241843, "grad_norm": 1.7623160974523429, "learning_rate": 1.6741864246329044e-05, "loss": 0.393, "step": 4479 }, { "epoch": 2.8662827895073577, "grad_norm": 1.7770765084284705, "learning_rate": 1.674033368903969e-05, "loss": 0.3773, "step": 4480 }, { "epoch": 2.8669225847728725, "grad_norm": 1.7669259610441135, "learning_rate": 1.6738802842332397e-05, "loss": 0.3753, "step": 4481 }, { "epoch": 2.867562380038388, "grad_norm": 1.6522619847510738, "learning_rate": 1.6737271706272897e-05, "loss": 0.4185, "step": 4482 }, { "epoch": 2.8682021753039026, "grad_norm": 1.6776659765235484, "learning_rate": 1.6735740280926936e-05, "loss": 0.3271, "step": 4483 }, { "epoch": 2.868841970569418, "grad_norm": 1.8541287283468724, "learning_rate": 1.6734208566360267e-05, "loss": 0.3976, "step": 4484 }, { "epoch": 2.8694817658349328, "grad_norm": 1.6883656067577488, "learning_rate": 1.6732676562638665e-05, "loss": 0.347, "step": 4485 }, { "epoch": 2.870121561100448, "grad_norm": 1.6752038470008548, "learning_rate": 1.6731144269827902e-05, "loss": 0.3526, "step": 4486 }, { "epoch": 2.870761356365963, "grad_norm": 1.749658277136559, "learning_rate": 1.672961168799378e-05, "loss": 0.4208, "step": 4487 }, { "epoch": 2.8714011516314777, "grad_norm": 1.7176566648804337, "learning_rate": 1.672807881720211e-05, "loss": 0.3341, "step": 4488 }, { "epoch": 2.872040946896993, "grad_norm": 1.9001037989608114, "learning_rate": 1.6726545657518696e-05, "loss": 0.4424, "step": 4489 }, { "epoch": 2.8726807421625082, "grad_norm": 1.6608691433350151, "learning_rate": 1.672501220900938e-05, "loss": 0.3463, "step": 4490 }, { "epoch": 2.873320537428023, "grad_norm": 1.7273054567181, "learning_rate": 1.672347847174e-05, "loss": 0.3801, "step": 4491 }, { "epoch": 2.873960332693538, "grad_norm": 1.7417857186933348, "learning_rate": 1.672194444577642e-05, "loss": 0.3873, "step": 4492 }, { "epoch": 2.874600127959053, "grad_norm": 1.6556069100606334, "learning_rate": 1.6720410131184496e-05, "loss": 0.3943, "step": 4493 }, { "epoch": 2.875239923224568, "grad_norm": 1.6821470206859805, "learning_rate": 1.671887552803012e-05, "loss": 0.3662, "step": 4494 }, { "epoch": 2.8758797184900833, "grad_norm": 1.6979235748949557, "learning_rate": 1.671734063637918e-05, "loss": 0.3465, "step": 4495 }, { "epoch": 2.876519513755598, "grad_norm": 1.7998205780253214, "learning_rate": 1.6715805456297584e-05, "loss": 0.4096, "step": 4496 }, { "epoch": 2.8771593090211134, "grad_norm": 1.5841732893035785, "learning_rate": 1.6714269987851245e-05, "loss": 0.3977, "step": 4497 }, { "epoch": 2.877799104286628, "grad_norm": 1.5295302150930077, "learning_rate": 1.67127342311061e-05, "loss": 0.3239, "step": 4498 }, { "epoch": 2.878438899552143, "grad_norm": 1.7205034590321475, "learning_rate": 1.6711198186128085e-05, "loss": 0.4267, "step": 4499 }, { "epoch": 2.8790786948176583, "grad_norm": 1.7624175033199816, "learning_rate": 1.6709661852983157e-05, "loss": 0.4094, "step": 4500 }, { "epoch": 2.8797184900831736, "grad_norm": 1.8471859566303304, "learning_rate": 1.670812523173729e-05, "loss": 0.3884, "step": 4501 }, { "epoch": 2.8803582853486884, "grad_norm": 1.7878404967892885, "learning_rate": 1.6706588322456455e-05, "loss": 0.3772, "step": 4502 }, { "epoch": 2.8809980806142033, "grad_norm": 2.062337184009804, "learning_rate": 1.6705051125206648e-05, "loss": 0.4093, "step": 4503 }, { "epoch": 2.8816378758797185, "grad_norm": 1.8396818121069254, "learning_rate": 1.6703513640053873e-05, "loss": 0.373, "step": 4504 }, { "epoch": 2.8822776711452334, "grad_norm": 1.7153963131663228, "learning_rate": 1.6701975867064147e-05, "loss": 0.3722, "step": 4505 }, { "epoch": 2.8829174664107486, "grad_norm": 1.8707792457961196, "learning_rate": 1.6700437806303498e-05, "loss": 0.3499, "step": 4506 }, { "epoch": 2.8835572616762635, "grad_norm": 1.8820278583188295, "learning_rate": 1.669889945783797e-05, "loss": 0.4541, "step": 4507 }, { "epoch": 2.8841970569417787, "grad_norm": 1.745526528151345, "learning_rate": 1.6697360821733616e-05, "loss": 0.3531, "step": 4508 }, { "epoch": 2.8848368522072936, "grad_norm": 1.783144837533218, "learning_rate": 1.6695821898056502e-05, "loss": 0.3401, "step": 4509 }, { "epoch": 2.885476647472809, "grad_norm": 1.742912718451244, "learning_rate": 1.6694282686872706e-05, "loss": 0.3544, "step": 4510 }, { "epoch": 2.8861164427383237, "grad_norm": 1.9315582419531145, "learning_rate": 1.6692743188248317e-05, "loss": 0.4527, "step": 4511 }, { "epoch": 2.886756238003839, "grad_norm": 1.5161085052562537, "learning_rate": 1.6691203402249443e-05, "loss": 0.3391, "step": 4512 }, { "epoch": 2.887396033269354, "grad_norm": 1.8143164476832407, "learning_rate": 1.6689663328942202e-05, "loss": 0.4017, "step": 4513 }, { "epoch": 2.8880358285348686, "grad_norm": 1.7285022626122215, "learning_rate": 1.668812296839271e-05, "loss": 0.3755, "step": 4514 }, { "epoch": 2.888675623800384, "grad_norm": 1.8027857242746927, "learning_rate": 1.6686582320667122e-05, "loss": 0.4102, "step": 4515 }, { "epoch": 2.889315419065899, "grad_norm": 1.6960436128202816, "learning_rate": 1.668504138583158e-05, "loss": 0.3783, "step": 4516 }, { "epoch": 2.889955214331414, "grad_norm": 1.7783278610374431, "learning_rate": 1.6683500163952253e-05, "loss": 0.4168, "step": 4517 }, { "epoch": 2.890595009596929, "grad_norm": 1.809512923650157, "learning_rate": 1.6681958655095318e-05, "loss": 0.3716, "step": 4518 }, { "epoch": 2.891234804862444, "grad_norm": 1.6382224856824914, "learning_rate": 1.6680416859326964e-05, "loss": 0.324, "step": 4519 }, { "epoch": 2.891874600127959, "grad_norm": 1.7566869295865728, "learning_rate": 1.6678874776713393e-05, "loss": 0.366, "step": 4520 }, { "epoch": 2.892514395393474, "grad_norm": 1.6883540825219647, "learning_rate": 1.6677332407320822e-05, "loss": 0.4255, "step": 4521 }, { "epoch": 2.893154190658989, "grad_norm": 1.7327204536296048, "learning_rate": 1.667578975121548e-05, "loss": 0.3395, "step": 4522 }, { "epoch": 2.8937939859245043, "grad_norm": 1.820408096376717, "learning_rate": 1.6674246808463593e-05, "loss": 0.4138, "step": 4523 }, { "epoch": 2.894433781190019, "grad_norm": 1.7948515790159973, "learning_rate": 1.667270357913142e-05, "loss": 0.3521, "step": 4524 }, { "epoch": 2.895073576455534, "grad_norm": 1.8500884078768196, "learning_rate": 1.667116006328523e-05, "loss": 0.3738, "step": 4525 }, { "epoch": 2.8957133717210493, "grad_norm": 1.70537254175925, "learning_rate": 1.6669616260991296e-05, "loss": 0.3305, "step": 4526 }, { "epoch": 2.8963531669865645, "grad_norm": 1.736755991797405, "learning_rate": 1.66680721723159e-05, "loss": 0.3709, "step": 4527 }, { "epoch": 2.8969929622520794, "grad_norm": 1.7146081079189193, "learning_rate": 1.6666527797325348e-05, "loss": 0.3489, "step": 4528 }, { "epoch": 2.897632757517594, "grad_norm": 1.7401145610558097, "learning_rate": 1.666498313608595e-05, "loss": 0.3892, "step": 4529 }, { "epoch": 2.8982725527831095, "grad_norm": 1.6765468337077918, "learning_rate": 1.666343818866403e-05, "loss": 0.3989, "step": 4530 }, { "epoch": 2.8989123480486243, "grad_norm": 1.782993624841976, "learning_rate": 1.6661892955125934e-05, "loss": 0.3408, "step": 4531 }, { "epoch": 2.8995521433141396, "grad_norm": 1.7162466171168265, "learning_rate": 1.6660347435538e-05, "loss": 0.4075, "step": 4532 }, { "epoch": 2.9001919385796544, "grad_norm": 1.7373899080276356, "learning_rate": 1.66588016299666e-05, "loss": 0.3854, "step": 4533 }, { "epoch": 2.9008317338451697, "grad_norm": 1.6446005212654227, "learning_rate": 1.6657255538478097e-05, "loss": 0.3544, "step": 4534 }, { "epoch": 2.9014715291106845, "grad_norm": 1.7050443582783208, "learning_rate": 1.6655709161138886e-05, "loss": 0.4323, "step": 4535 }, { "epoch": 2.9021113243761993, "grad_norm": 1.668917028325132, "learning_rate": 1.6654162498015366e-05, "loss": 0.3766, "step": 4536 }, { "epoch": 2.9027511196417146, "grad_norm": 1.6777089311617763, "learning_rate": 1.665261554917394e-05, "loss": 0.3451, "step": 4537 }, { "epoch": 2.90339091490723, "grad_norm": 1.8065091920868257, "learning_rate": 1.6651068314681037e-05, "loss": 0.3509, "step": 4538 }, { "epoch": 2.9040307101727447, "grad_norm": 1.7404786022699474, "learning_rate": 1.6649520794603097e-05, "loss": 0.3641, "step": 4539 }, { "epoch": 2.9046705054382596, "grad_norm": 1.8415432920196468, "learning_rate": 1.664797298900656e-05, "loss": 0.3627, "step": 4540 }, { "epoch": 2.905310300703775, "grad_norm": 1.8041032226005187, "learning_rate": 1.6646424897957885e-05, "loss": 0.3436, "step": 4541 }, { "epoch": 2.90595009596929, "grad_norm": 1.7702687320712247, "learning_rate": 1.664487652152355e-05, "loss": 0.3623, "step": 4542 }, { "epoch": 2.906589891234805, "grad_norm": 1.8335001332005643, "learning_rate": 1.664332785977004e-05, "loss": 0.3698, "step": 4543 }, { "epoch": 2.9072296865003198, "grad_norm": 1.7416970083690695, "learning_rate": 1.664177891276385e-05, "loss": 0.3832, "step": 4544 }, { "epoch": 2.907869481765835, "grad_norm": 1.6528226334896516, "learning_rate": 1.6640229680571484e-05, "loss": 0.3838, "step": 4545 }, { "epoch": 2.90850927703135, "grad_norm": 1.610323662887497, "learning_rate": 1.6638680163259467e-05, "loss": 0.3249, "step": 4546 }, { "epoch": 2.909149072296865, "grad_norm": 1.6602509466260338, "learning_rate": 1.6637130360894338e-05, "loss": 0.3448, "step": 4547 }, { "epoch": 2.90978886756238, "grad_norm": 1.5562956242967536, "learning_rate": 1.6635580273542635e-05, "loss": 0.3988, "step": 4548 }, { "epoch": 2.9104286628278953, "grad_norm": 1.7614448952264041, "learning_rate": 1.663402990127092e-05, "loss": 0.3906, "step": 4549 }, { "epoch": 2.91106845809341, "grad_norm": 1.6124032464782105, "learning_rate": 1.6632479244145764e-05, "loss": 0.3907, "step": 4550 }, { "epoch": 2.911708253358925, "grad_norm": 1.5823138821135398, "learning_rate": 1.6630928302233743e-05, "loss": 0.332, "step": 4551 }, { "epoch": 2.91234804862444, "grad_norm": 1.659498420187062, "learning_rate": 1.662937707560146e-05, "loss": 0.3208, "step": 4552 }, { "epoch": 2.9129878438899555, "grad_norm": 1.7181863685301138, "learning_rate": 1.6627825564315517e-05, "loss": 0.325, "step": 4553 }, { "epoch": 2.9136276391554703, "grad_norm": 2.0721294631779865, "learning_rate": 1.662627376844253e-05, "loss": 0.4193, "step": 4554 }, { "epoch": 2.914267434420985, "grad_norm": 1.736886457525362, "learning_rate": 1.662472168804914e-05, "loss": 0.339, "step": 4555 }, { "epoch": 2.9149072296865004, "grad_norm": 1.6819076013273901, "learning_rate": 1.662316932320198e-05, "loss": 0.3522, "step": 4556 }, { "epoch": 2.9155470249520152, "grad_norm": 1.7209205757106982, "learning_rate": 1.6621616673967718e-05, "loss": 0.3727, "step": 4557 }, { "epoch": 2.9161868202175305, "grad_norm": 1.6052108791772324, "learning_rate": 1.662006374041301e-05, "loss": 0.3462, "step": 4558 }, { "epoch": 2.9168266154830453, "grad_norm": 1.7198452913516986, "learning_rate": 1.6618510522604544e-05, "loss": 0.4274, "step": 4559 }, { "epoch": 2.9174664107485606, "grad_norm": 1.683365240438474, "learning_rate": 1.661695702060901e-05, "loss": 0.3299, "step": 4560 }, { "epoch": 2.9181062060140754, "grad_norm": 1.7165187270032483, "learning_rate": 1.6615403234493104e-05, "loss": 0.3819, "step": 4561 }, { "epoch": 2.9187460012795903, "grad_norm": 1.7623012091340797, "learning_rate": 1.6613849164323562e-05, "loss": 0.3931, "step": 4562 }, { "epoch": 2.9193857965451055, "grad_norm": 1.5622999689774406, "learning_rate": 1.66122948101671e-05, "loss": 0.346, "step": 4563 }, { "epoch": 2.920025591810621, "grad_norm": 1.749504415767081, "learning_rate": 1.6610740172090453e-05, "loss": 0.3999, "step": 4564 }, { "epoch": 2.9206653870761357, "grad_norm": 1.6750244507377081, "learning_rate": 1.660918525016039e-05, "loss": 0.3485, "step": 4565 }, { "epoch": 2.9213051823416505, "grad_norm": 1.6162881622515795, "learning_rate": 1.6607630044443665e-05, "loss": 0.3169, "step": 4566 }, { "epoch": 2.9219449776071658, "grad_norm": 1.6581142634366552, "learning_rate": 1.6606074555007065e-05, "loss": 0.3601, "step": 4567 }, { "epoch": 2.9225847728726806, "grad_norm": 1.813521928300432, "learning_rate": 1.660451878191737e-05, "loss": 0.3945, "step": 4568 }, { "epoch": 2.923224568138196, "grad_norm": 1.8474822815147154, "learning_rate": 1.6602962725241393e-05, "loss": 0.3625, "step": 4569 }, { "epoch": 2.9238643634037107, "grad_norm": 1.6605886340879856, "learning_rate": 1.6601406385045938e-05, "loss": 0.3612, "step": 4570 }, { "epoch": 2.924504158669226, "grad_norm": 1.682816700355653, "learning_rate": 1.6599849761397836e-05, "loss": 0.3305, "step": 4571 }, { "epoch": 2.925143953934741, "grad_norm": 1.8997525388666934, "learning_rate": 1.6598292854363924e-05, "loss": 0.4098, "step": 4572 }, { "epoch": 2.925783749200256, "grad_norm": 1.561916872553066, "learning_rate": 1.6596735664011054e-05, "loss": 0.3404, "step": 4573 }, { "epoch": 2.926423544465771, "grad_norm": 1.635540679666522, "learning_rate": 1.6595178190406095e-05, "loss": 0.3649, "step": 4574 }, { "epoch": 2.927063339731286, "grad_norm": 1.837592993663815, "learning_rate": 1.659362043361591e-05, "loss": 0.3792, "step": 4575 }, { "epoch": 2.927703134996801, "grad_norm": 1.8439070918466238, "learning_rate": 1.6592062393707395e-05, "loss": 0.4032, "step": 4576 }, { "epoch": 2.928342930262316, "grad_norm": 1.6388854730359426, "learning_rate": 1.659050407074745e-05, "loss": 0.3892, "step": 4577 }, { "epoch": 2.928982725527831, "grad_norm": 1.8004615196065972, "learning_rate": 1.658894546480298e-05, "loss": 0.3621, "step": 4578 }, { "epoch": 2.9296225207933464, "grad_norm": 1.7348898810753381, "learning_rate": 1.6587386575940915e-05, "loss": 0.4072, "step": 4579 }, { "epoch": 2.9302623160588612, "grad_norm": 1.6789653439514167, "learning_rate": 1.658582740422819e-05, "loss": 0.3424, "step": 4580 }, { "epoch": 2.930902111324376, "grad_norm": 1.7908151713064433, "learning_rate": 1.6584267949731744e-05, "loss": 0.3431, "step": 4581 }, { "epoch": 2.9315419065898913, "grad_norm": 1.7508081104891864, "learning_rate": 1.6582708212518555e-05, "loss": 0.3314, "step": 4582 }, { "epoch": 2.932181701855406, "grad_norm": 1.6198765482874558, "learning_rate": 1.6581148192655575e-05, "loss": 0.3697, "step": 4583 }, { "epoch": 2.9328214971209214, "grad_norm": 1.5980282838189672, "learning_rate": 1.6579587890209805e-05, "loss": 0.4239, "step": 4584 }, { "epoch": 2.9334612923864363, "grad_norm": 1.674396809331926, "learning_rate": 1.6578027305248235e-05, "loss": 0.4062, "step": 4585 }, { "epoch": 2.9341010876519515, "grad_norm": 1.683775164513597, "learning_rate": 1.657646643783787e-05, "loss": 0.3818, "step": 4586 }, { "epoch": 2.9347408829174664, "grad_norm": 1.6115901983982273, "learning_rate": 1.657490528804574e-05, "loss": 0.3747, "step": 4587 }, { "epoch": 2.935380678182981, "grad_norm": 1.7181600063268676, "learning_rate": 1.657334385593887e-05, "loss": 0.4103, "step": 4588 }, { "epoch": 2.9360204734484965, "grad_norm": 1.6153706184287706, "learning_rate": 1.657178214158431e-05, "loss": 0.3578, "step": 4589 }, { "epoch": 2.9366602687140118, "grad_norm": 1.6777218099109097, "learning_rate": 1.657022014504911e-05, "loss": 0.3717, "step": 4590 }, { "epoch": 2.9373000639795266, "grad_norm": 1.805261579246797, "learning_rate": 1.6568657866400347e-05, "loss": 0.3798, "step": 4591 }, { "epoch": 2.9379398592450414, "grad_norm": 1.7502296444188754, "learning_rate": 1.65670953057051e-05, "loss": 0.3803, "step": 4592 }, { "epoch": 2.9385796545105567, "grad_norm": 1.7828019624797087, "learning_rate": 1.6565532463030466e-05, "loss": 0.352, "step": 4593 }, { "epoch": 2.9392194497760715, "grad_norm": 1.5990112956089386, "learning_rate": 1.6563969338443545e-05, "loss": 0.3855, "step": 4594 }, { "epoch": 2.939859245041587, "grad_norm": 1.7509986535812885, "learning_rate": 1.6562405932011453e-05, "loss": 0.3417, "step": 4595 }, { "epoch": 2.9404990403071016, "grad_norm": 1.7321891743901976, "learning_rate": 1.6560842243801328e-05, "loss": 0.3658, "step": 4596 }, { "epoch": 2.941138835572617, "grad_norm": 1.6828595451042483, "learning_rate": 1.655927827388031e-05, "loss": 0.3951, "step": 4597 }, { "epoch": 2.9417786308381317, "grad_norm": 1.7620160559555504, "learning_rate": 1.6557714022315544e-05, "loss": 0.3588, "step": 4598 }, { "epoch": 2.9424184261036466, "grad_norm": 1.6880136831258055, "learning_rate": 1.655614948917421e-05, "loss": 0.3481, "step": 4599 }, { "epoch": 2.943058221369162, "grad_norm": 1.6420887080801092, "learning_rate": 1.6554584674523475e-05, "loss": 0.3319, "step": 4600 }, { "epoch": 2.943698016634677, "grad_norm": 1.6728965149756745, "learning_rate": 1.6553019578430537e-05, "loss": 0.3975, "step": 4601 }, { "epoch": 2.944337811900192, "grad_norm": 1.6639342287213132, "learning_rate": 1.655145420096259e-05, "loss": 0.373, "step": 4602 }, { "epoch": 2.944977607165707, "grad_norm": 1.636342173619136, "learning_rate": 1.654988854218686e-05, "loss": 0.3533, "step": 4603 }, { "epoch": 2.945617402431222, "grad_norm": 1.7029995810007077, "learning_rate": 1.6548322602170565e-05, "loss": 0.4266, "step": 4604 }, { "epoch": 2.946257197696737, "grad_norm": 1.7306037922534747, "learning_rate": 1.6546756380980947e-05, "loss": 0.3631, "step": 4605 }, { "epoch": 2.946896992962252, "grad_norm": 1.5498303860484444, "learning_rate": 1.6545189878685253e-05, "loss": 0.4045, "step": 4606 }, { "epoch": 2.947536788227767, "grad_norm": 1.7199405618213865, "learning_rate": 1.6543623095350748e-05, "loss": 0.3378, "step": 4607 }, { "epoch": 2.9481765834932823, "grad_norm": 1.6107987784647857, "learning_rate": 1.654205603104471e-05, "loss": 0.3616, "step": 4608 }, { "epoch": 2.948816378758797, "grad_norm": 1.772400563391084, "learning_rate": 1.6540488685834425e-05, "loss": 0.3796, "step": 4609 }, { "epoch": 2.9494561740243124, "grad_norm": 1.7129444856681497, "learning_rate": 1.6538921059787186e-05, "loss": 0.4518, "step": 4610 }, { "epoch": 2.950095969289827, "grad_norm": 1.5483471751572986, "learning_rate": 1.653735315297031e-05, "loss": 0.3381, "step": 4611 }, { "epoch": 2.9507357645553425, "grad_norm": 1.6551250875425774, "learning_rate": 1.6535784965451123e-05, "loss": 0.3455, "step": 4612 }, { "epoch": 2.9513755598208573, "grad_norm": 1.7485009828948086, "learning_rate": 1.6534216497296952e-05, "loss": 0.4195, "step": 4613 }, { "epoch": 2.952015355086372, "grad_norm": 1.645215033628823, "learning_rate": 1.653264774857515e-05, "loss": 0.3443, "step": 4614 }, { "epoch": 2.9526551503518874, "grad_norm": 1.842747904873728, "learning_rate": 1.653107871935307e-05, "loss": 0.3983, "step": 4615 }, { "epoch": 2.9532949456174027, "grad_norm": 1.700911831455477, "learning_rate": 1.6529509409698093e-05, "loss": 0.3535, "step": 4616 }, { "epoch": 2.9539347408829175, "grad_norm": 1.8292969379111026, "learning_rate": 1.6527939819677593e-05, "loss": 0.3861, "step": 4617 }, { "epoch": 2.9545745361484324, "grad_norm": 1.5830109357187057, "learning_rate": 1.6526369949358973e-05, "loss": 0.3812, "step": 4618 }, { "epoch": 2.9552143314139476, "grad_norm": 1.5467028692735538, "learning_rate": 1.6524799798809636e-05, "loss": 0.3074, "step": 4619 }, { "epoch": 2.9558541266794625, "grad_norm": 1.6695842346884366, "learning_rate": 1.6523229368097e-05, "loss": 0.3807, "step": 4620 }, { "epoch": 2.9564939219449777, "grad_norm": 1.6888975109398492, "learning_rate": 1.65216586572885e-05, "loss": 0.3714, "step": 4621 }, { "epoch": 2.9571337172104926, "grad_norm": 1.7105416640273254, "learning_rate": 1.652008766645158e-05, "loss": 0.3737, "step": 4622 }, { "epoch": 2.957773512476008, "grad_norm": 1.7446469605113841, "learning_rate": 1.651851639565369e-05, "loss": 0.3579, "step": 4623 }, { "epoch": 2.9584133077415227, "grad_norm": 1.8531101631979978, "learning_rate": 1.6516944844962305e-05, "loss": 0.3943, "step": 4624 }, { "epoch": 2.9590531030070375, "grad_norm": 1.8770530445410394, "learning_rate": 1.65153730144449e-05, "loss": 0.4359, "step": 4625 }, { "epoch": 2.9596928982725528, "grad_norm": 1.6975152417503023, "learning_rate": 1.651380090416897e-05, "loss": 0.3402, "step": 4626 }, { "epoch": 2.960332693538068, "grad_norm": 1.5247484582694093, "learning_rate": 1.6512228514202014e-05, "loss": 0.348, "step": 4627 }, { "epoch": 2.960972488803583, "grad_norm": 1.6257766278354713, "learning_rate": 1.6510655844611546e-05, "loss": 0.3504, "step": 4628 }, { "epoch": 2.9616122840690977, "grad_norm": 1.5502873119461291, "learning_rate": 1.6509082895465104e-05, "loss": 0.3462, "step": 4629 }, { "epoch": 2.962252079334613, "grad_norm": 1.661054803547218, "learning_rate": 1.650750966683022e-05, "loss": 0.3649, "step": 4630 }, { "epoch": 2.962891874600128, "grad_norm": 1.5407881330987947, "learning_rate": 1.6505936158774445e-05, "loss": 0.308, "step": 4631 }, { "epoch": 2.963531669865643, "grad_norm": 1.6891395324962355, "learning_rate": 1.6504362371365346e-05, "loss": 0.3947, "step": 4632 }, { "epoch": 2.964171465131158, "grad_norm": 1.649856797398211, "learning_rate": 1.6502788304670495e-05, "loss": 0.3849, "step": 4633 }, { "epoch": 2.964811260396673, "grad_norm": 1.7843935008377558, "learning_rate": 1.6501213958757483e-05, "loss": 0.3799, "step": 4634 }, { "epoch": 2.965451055662188, "grad_norm": 1.6293315613601878, "learning_rate": 1.649963933369391e-05, "loss": 0.3662, "step": 4635 }, { "epoch": 2.966090850927703, "grad_norm": 1.7690152782924584, "learning_rate": 1.649806442954738e-05, "loss": 0.3364, "step": 4636 }, { "epoch": 2.966730646193218, "grad_norm": 1.6670389330607844, "learning_rate": 1.6496489246385534e-05, "loss": 0.353, "step": 4637 }, { "epoch": 2.9673704414587334, "grad_norm": 1.7106664845510668, "learning_rate": 1.6494913784275984e-05, "loss": 0.3403, "step": 4638 }, { "epoch": 2.9680102367242482, "grad_norm": 1.8111283506107791, "learning_rate": 1.6493338043286396e-05, "loss": 0.3656, "step": 4639 }, { "epoch": 2.968650031989763, "grad_norm": 1.7518286708039577, "learning_rate": 1.6491762023484422e-05, "loss": 0.3975, "step": 4640 }, { "epoch": 2.9692898272552783, "grad_norm": 1.6703777188206408, "learning_rate": 1.6490185724937738e-05, "loss": 0.3524, "step": 4641 }, { "epoch": 2.9699296225207936, "grad_norm": 1.7514891034727258, "learning_rate": 1.648860914771402e-05, "loss": 0.4068, "step": 4642 }, { "epoch": 2.9705694177863085, "grad_norm": 1.6232351424023583, "learning_rate": 1.6487032291880968e-05, "loss": 0.323, "step": 4643 }, { "epoch": 2.9712092130518233, "grad_norm": 1.7042279845969794, "learning_rate": 1.648545515750629e-05, "loss": 0.3266, "step": 4644 }, { "epoch": 2.9718490083173386, "grad_norm": 1.550451516558887, "learning_rate": 1.6483877744657703e-05, "loss": 0.3767, "step": 4645 }, { "epoch": 2.9724888035828534, "grad_norm": 1.6862266381590527, "learning_rate": 1.648230005340294e-05, "loss": 0.3625, "step": 4646 }, { "epoch": 2.9731285988483687, "grad_norm": 1.5301604639997093, "learning_rate": 1.6480722083809743e-05, "loss": 0.3522, "step": 4647 }, { "epoch": 2.9737683941138835, "grad_norm": 1.6800947038874832, "learning_rate": 1.647914383594587e-05, "loss": 0.3635, "step": 4648 }, { "epoch": 2.9744081893793988, "grad_norm": 1.7546667449518958, "learning_rate": 1.6477565309879086e-05, "loss": 0.3578, "step": 4649 }, { "epoch": 2.9750479846449136, "grad_norm": 1.5886169688916851, "learning_rate": 1.647598650567717e-05, "loss": 0.3402, "step": 4650 }, { "epoch": 2.9756877799104284, "grad_norm": 1.5699169525124894, "learning_rate": 1.6474407423407913e-05, "loss": 0.3566, "step": 4651 }, { "epoch": 2.9763275751759437, "grad_norm": 1.7731747793158992, "learning_rate": 1.6472828063139116e-05, "loss": 0.4022, "step": 4652 }, { "epoch": 2.976967370441459, "grad_norm": 1.7491724309304677, "learning_rate": 1.64712484249386e-05, "loss": 0.3987, "step": 4653 }, { "epoch": 2.977607165706974, "grad_norm": 1.799010382799031, "learning_rate": 1.6469668508874188e-05, "loss": 0.398, "step": 4654 }, { "epoch": 2.9782469609724886, "grad_norm": 1.8584322446534143, "learning_rate": 1.6468088315013714e-05, "loss": 0.3814, "step": 4655 }, { "epoch": 2.978886756238004, "grad_norm": 1.7147006023102975, "learning_rate": 1.6466507843425036e-05, "loss": 0.366, "step": 4656 }, { "epoch": 2.9795265515035187, "grad_norm": 1.6565860989876868, "learning_rate": 1.6464927094176016e-05, "loss": 0.4089, "step": 4657 }, { "epoch": 2.980166346769034, "grad_norm": 1.709652913311459, "learning_rate": 1.6463346067334526e-05, "loss": 0.3582, "step": 4658 }, { "epoch": 2.980806142034549, "grad_norm": 1.677690095717981, "learning_rate": 1.6461764762968455e-05, "loss": 0.342, "step": 4659 }, { "epoch": 2.981445937300064, "grad_norm": 1.6749488274846966, "learning_rate": 1.6460183181145696e-05, "loss": 0.3445, "step": 4660 }, { "epoch": 2.982085732565579, "grad_norm": 1.5706936632929955, "learning_rate": 1.645860132193417e-05, "loss": 0.3645, "step": 4661 }, { "epoch": 2.982725527831094, "grad_norm": 1.764169460651485, "learning_rate": 1.6457019185401784e-05, "loss": 0.3359, "step": 4662 }, { "epoch": 2.983365323096609, "grad_norm": 1.8571630123340988, "learning_rate": 1.645543677161648e-05, "loss": 0.4098, "step": 4663 }, { "epoch": 2.9840051183621243, "grad_norm": 1.8093606739128274, "learning_rate": 1.6453854080646213e-05, "loss": 0.3626, "step": 4664 }, { "epoch": 2.984644913627639, "grad_norm": 1.7430470537981468, "learning_rate": 1.6452271112558924e-05, "loss": 0.3637, "step": 4665 }, { "epoch": 2.985284708893154, "grad_norm": 1.6858971662388649, "learning_rate": 1.6450687867422595e-05, "loss": 0.3692, "step": 4666 }, { "epoch": 2.9859245041586693, "grad_norm": 1.6786330436766794, "learning_rate": 1.6449104345305204e-05, "loss": 0.3382, "step": 4667 }, { "epoch": 2.986564299424184, "grad_norm": 1.5878894696998604, "learning_rate": 1.6447520546274742e-05, "loss": 0.3466, "step": 4668 }, { "epoch": 2.9872040946896994, "grad_norm": 1.7047976852288993, "learning_rate": 1.644593647039922e-05, "loss": 0.3744, "step": 4669 }, { "epoch": 2.987843889955214, "grad_norm": 1.647775585612246, "learning_rate": 1.644435211774665e-05, "loss": 0.3387, "step": 4670 }, { "epoch": 2.9884836852207295, "grad_norm": 1.7342139724642216, "learning_rate": 1.6442767488385066e-05, "loss": 0.3571, "step": 4671 }, { "epoch": 2.9891234804862443, "grad_norm": 1.7768363456436744, "learning_rate": 1.6441182582382506e-05, "loss": 0.3572, "step": 4672 }, { "epoch": 2.9897632757517596, "grad_norm": 1.600374324748444, "learning_rate": 1.643959739980702e-05, "loss": 0.3332, "step": 4673 }, { "epoch": 2.9904030710172744, "grad_norm": 1.7439031988415217, "learning_rate": 1.643801194072668e-05, "loss": 0.342, "step": 4674 }, { "epoch": 2.9910428662827897, "grad_norm": 1.805980815774166, "learning_rate": 1.6436426205209563e-05, "loss": 0.3642, "step": 4675 }, { "epoch": 2.9916826615483045, "grad_norm": 1.8630430501974862, "learning_rate": 1.643484019332375e-05, "loss": 0.397, "step": 4676 }, { "epoch": 2.9923224568138194, "grad_norm": 1.9218581289178513, "learning_rate": 1.6433253905137346e-05, "loss": 0.4086, "step": 4677 }, { "epoch": 2.9929622520793346, "grad_norm": 1.6799329687464766, "learning_rate": 1.6431667340718465e-05, "loss": 0.3854, "step": 4678 }, { "epoch": 2.99360204734485, "grad_norm": 1.6846823489512024, "learning_rate": 1.6430080500135228e-05, "loss": 0.3813, "step": 4679 }, { "epoch": 2.9942418426103647, "grad_norm": 1.7324135564980463, "learning_rate": 1.6428493383455774e-05, "loss": 0.392, "step": 4680 }, { "epoch": 2.9948816378758796, "grad_norm": 1.7219771141049516, "learning_rate": 1.642690599074825e-05, "loss": 0.3187, "step": 4681 }, { "epoch": 2.995521433141395, "grad_norm": 1.8465707444908295, "learning_rate": 1.642531832208081e-05, "loss": 0.3608, "step": 4682 }, { "epoch": 2.9961612284069097, "grad_norm": 1.657091601630111, "learning_rate": 1.6423730377521635e-05, "loss": 0.3778, "step": 4683 }, { "epoch": 2.996801023672425, "grad_norm": 1.5392373131610195, "learning_rate": 1.6422142157138904e-05, "loss": 0.3676, "step": 4684 }, { "epoch": 2.99744081893794, "grad_norm": 1.6944462975169425, "learning_rate": 1.6420553661000813e-05, "loss": 0.372, "step": 4685 }, { "epoch": 2.998080614203455, "grad_norm": 1.7357512170295133, "learning_rate": 1.6418964889175572e-05, "loss": 0.3246, "step": 4686 }, { "epoch": 2.99872040946897, "grad_norm": 1.60993478425556, "learning_rate": 1.6417375841731393e-05, "loss": 0.3317, "step": 4687 }, { "epoch": 2.9993602047344847, "grad_norm": 1.7582331440303662, "learning_rate": 1.641578651873651e-05, "loss": 0.3457, "step": 4688 }, { "epoch": 3.0, "grad_norm": 1.6550944744000413, "learning_rate": 1.641419692025917e-05, "loss": 0.3344, "step": 4689 }, { "epoch": 3.000639795265515, "grad_norm": 1.2845832553095478, "learning_rate": 1.6412607046367622e-05, "loss": 0.2229, "step": 4690 }, { "epoch": 3.00127959053103, "grad_norm": 1.5219744170889495, "learning_rate": 1.6411016897130138e-05, "loss": 0.2553, "step": 4691 }, { "epoch": 3.001919385796545, "grad_norm": 1.367338673807188, "learning_rate": 1.640942647261499e-05, "loss": 0.2472, "step": 4692 }, { "epoch": 3.00255918106206, "grad_norm": 1.5052422843007645, "learning_rate": 1.640783577289047e-05, "loss": 0.2578, "step": 4693 }, { "epoch": 3.003198976327575, "grad_norm": 1.5858479019034701, "learning_rate": 1.6406244798024884e-05, "loss": 0.2676, "step": 4694 }, { "epoch": 3.0038387715930903, "grad_norm": 1.631889831678651, "learning_rate": 1.640465354808654e-05, "loss": 0.267, "step": 4695 }, { "epoch": 3.004478566858605, "grad_norm": 1.6956572472919964, "learning_rate": 1.6403062023143767e-05, "loss": 0.2728, "step": 4696 }, { "epoch": 3.0051183621241204, "grad_norm": 1.7132803945909023, "learning_rate": 1.6401470223264898e-05, "loss": 0.2699, "step": 4697 }, { "epoch": 3.0057581573896353, "grad_norm": 1.4846805850550213, "learning_rate": 1.6399878148518288e-05, "loss": 0.2363, "step": 4698 }, { "epoch": 3.0063979526551505, "grad_norm": 1.544174205408368, "learning_rate": 1.6398285798972292e-05, "loss": 0.2435, "step": 4699 }, { "epoch": 3.0070377479206654, "grad_norm": 1.478257492436043, "learning_rate": 1.6396693174695286e-05, "loss": 0.2548, "step": 4700 }, { "epoch": 3.0076775431861806, "grad_norm": 1.540019205030196, "learning_rate": 1.6395100275755657e-05, "loss": 0.2414, "step": 4701 }, { "epoch": 3.0083173384516955, "grad_norm": 1.4146354453399679, "learning_rate": 1.6393507102221793e-05, "loss": 0.2433, "step": 4702 }, { "epoch": 3.0089571337172103, "grad_norm": 1.5230724503863546, "learning_rate": 1.639191365416211e-05, "loss": 0.2488, "step": 4703 }, { "epoch": 3.0095969289827256, "grad_norm": 1.4286125814572208, "learning_rate": 1.639031993164503e-05, "loss": 0.2412, "step": 4704 }, { "epoch": 3.0102367242482404, "grad_norm": 1.3996716892310364, "learning_rate": 1.6388725934738973e-05, "loss": 0.2441, "step": 4705 }, { "epoch": 3.0108765195137557, "grad_norm": 1.5402322188686353, "learning_rate": 1.6387131663512393e-05, "loss": 0.2637, "step": 4706 }, { "epoch": 3.0115163147792705, "grad_norm": 1.5358562725981693, "learning_rate": 1.6385537118033735e-05, "loss": 0.2434, "step": 4707 }, { "epoch": 3.012156110044786, "grad_norm": 1.677254459091735, "learning_rate": 1.6383942298371475e-05, "loss": 0.2761, "step": 4708 }, { "epoch": 3.0127959053103006, "grad_norm": 1.5431780493254155, "learning_rate": 1.6382347204594092e-05, "loss": 0.2362, "step": 4709 }, { "epoch": 3.013435700575816, "grad_norm": 1.6310004785296672, "learning_rate": 1.6380751836770073e-05, "loss": 0.2479, "step": 4710 }, { "epoch": 3.0140754958413307, "grad_norm": 1.6061403977163313, "learning_rate": 1.6379156194967916e-05, "loss": 0.278, "step": 4711 }, { "epoch": 3.014715291106846, "grad_norm": 1.4972209717174145, "learning_rate": 1.6377560279256145e-05, "loss": 0.2356, "step": 4712 }, { "epoch": 3.015355086372361, "grad_norm": 1.4586264225367764, "learning_rate": 1.6375964089703278e-05, "loss": 0.2375, "step": 4713 }, { "epoch": 3.0159948816378757, "grad_norm": 1.6221619079970713, "learning_rate": 1.6374367626377852e-05, "loss": 0.2448, "step": 4714 }, { "epoch": 3.016634676903391, "grad_norm": 1.3803206563330703, "learning_rate": 1.6372770889348424e-05, "loss": 0.2468, "step": 4715 }, { "epoch": 3.0172744721689058, "grad_norm": 1.4337775876546381, "learning_rate": 1.637117387868355e-05, "loss": 0.2654, "step": 4716 }, { "epoch": 3.017914267434421, "grad_norm": 1.3958116592681544, "learning_rate": 1.6369576594451803e-05, "loss": 0.2161, "step": 4717 }, { "epoch": 3.018554062699936, "grad_norm": 1.52633967252805, "learning_rate": 1.6367979036721764e-05, "loss": 0.2449, "step": 4718 }, { "epoch": 3.019193857965451, "grad_norm": 1.5280814200618842, "learning_rate": 1.6366381205562033e-05, "loss": 0.2511, "step": 4719 }, { "epoch": 3.019833653230966, "grad_norm": 1.6409653046923043, "learning_rate": 1.6364783101041222e-05, "loss": 0.267, "step": 4720 }, { "epoch": 3.0204734484964813, "grad_norm": 1.6071480926570956, "learning_rate": 1.6363184723227944e-05, "loss": 0.2717, "step": 4721 }, { "epoch": 3.021113243761996, "grad_norm": 1.3546492739126603, "learning_rate": 1.6361586072190834e-05, "loss": 0.2129, "step": 4722 }, { "epoch": 3.0217530390275114, "grad_norm": 1.5613382155979512, "learning_rate": 1.6359987147998533e-05, "loss": 0.2899, "step": 4723 }, { "epoch": 3.022392834293026, "grad_norm": 1.404634766036156, "learning_rate": 1.63583879507197e-05, "loss": 0.2414, "step": 4724 }, { "epoch": 3.0230326295585415, "grad_norm": 1.399200393530498, "learning_rate": 1.6356788480423e-05, "loss": 0.2361, "step": 4725 }, { "epoch": 3.0236724248240563, "grad_norm": 1.3820470800303253, "learning_rate": 1.6355188737177107e-05, "loss": 0.2119, "step": 4726 }, { "epoch": 3.024312220089571, "grad_norm": 1.4281706182320337, "learning_rate": 1.6353588721050714e-05, "loss": 0.228, "step": 4727 }, { "epoch": 3.0249520153550864, "grad_norm": 1.529081086069791, "learning_rate": 1.6351988432112526e-05, "loss": 0.2409, "step": 4728 }, { "epoch": 3.0255918106206012, "grad_norm": 1.4867236931348575, "learning_rate": 1.6350387870431256e-05, "loss": 0.2412, "step": 4729 }, { "epoch": 3.0262316058861165, "grad_norm": 1.4900968712739873, "learning_rate": 1.634878703607562e-05, "loss": 0.2464, "step": 4730 }, { "epoch": 3.0268714011516313, "grad_norm": 1.484829455025725, "learning_rate": 1.634718592911437e-05, "loss": 0.2251, "step": 4731 }, { "epoch": 3.0275111964171466, "grad_norm": 1.5533467948240758, "learning_rate": 1.6345584549616248e-05, "loss": 0.2627, "step": 4732 }, { "epoch": 3.0281509916826614, "grad_norm": 1.5760800790961946, "learning_rate": 1.6343982897650005e-05, "loss": 0.2609, "step": 4733 }, { "epoch": 3.0287907869481767, "grad_norm": 1.6016415055908693, "learning_rate": 1.6342380973284428e-05, "loss": 0.2475, "step": 4734 }, { "epoch": 3.0294305822136915, "grad_norm": 1.44876377664777, "learning_rate": 1.6340778776588294e-05, "loss": 0.278, "step": 4735 }, { "epoch": 3.030070377479207, "grad_norm": 1.3989820809491877, "learning_rate": 1.6339176307630395e-05, "loss": 0.2173, "step": 4736 }, { "epoch": 3.0307101727447217, "grad_norm": 1.3515312979369745, "learning_rate": 1.6337573566479542e-05, "loss": 0.2219, "step": 4737 }, { "epoch": 3.031349968010237, "grad_norm": 1.5042926779291643, "learning_rate": 1.6335970553204555e-05, "loss": 0.2449, "step": 4738 }, { "epoch": 3.0319897632757518, "grad_norm": 1.4678079100330381, "learning_rate": 1.6334367267874265e-05, "loss": 0.2341, "step": 4739 }, { "epoch": 3.0326295585412666, "grad_norm": 1.4955872852864318, "learning_rate": 1.633276371055751e-05, "loss": 0.2537, "step": 4740 }, { "epoch": 3.033269353806782, "grad_norm": 1.4625976751091294, "learning_rate": 1.633115988132315e-05, "loss": 0.2354, "step": 4741 }, { "epoch": 3.0339091490722967, "grad_norm": 1.4295458398556886, "learning_rate": 1.6329555780240043e-05, "loss": 0.2487, "step": 4742 }, { "epoch": 3.034548944337812, "grad_norm": 1.3942063792847235, "learning_rate": 1.632795140737707e-05, "loss": 0.2014, "step": 4743 }, { "epoch": 3.035188739603327, "grad_norm": 1.626280100927758, "learning_rate": 1.6326346762803126e-05, "loss": 0.2663, "step": 4744 }, { "epoch": 3.035828534868842, "grad_norm": 1.43791478632137, "learning_rate": 1.63247418465871e-05, "loss": 0.2349, "step": 4745 }, { "epoch": 3.036468330134357, "grad_norm": 1.4695809110560527, "learning_rate": 1.632313665879791e-05, "loss": 0.2609, "step": 4746 }, { "epoch": 3.037108125399872, "grad_norm": 1.4837148699301967, "learning_rate": 1.6321531199504483e-05, "loss": 0.2568, "step": 4747 }, { "epoch": 3.037747920665387, "grad_norm": 1.3873021754540233, "learning_rate": 1.631992546877575e-05, "loss": 0.2365, "step": 4748 }, { "epoch": 3.0383877159309023, "grad_norm": 1.3862779974277342, "learning_rate": 1.6318319466680657e-05, "loss": 0.2438, "step": 4749 }, { "epoch": 3.039027511196417, "grad_norm": 1.3862704086251485, "learning_rate": 1.631671319328817e-05, "loss": 0.2256, "step": 4750 }, { "epoch": 3.0396673064619324, "grad_norm": 1.5841377917567443, "learning_rate": 1.6315106648667252e-05, "loss": 0.258, "step": 4751 }, { "epoch": 3.0403071017274472, "grad_norm": 1.618102058417711, "learning_rate": 1.631349983288689e-05, "loss": 0.2818, "step": 4752 }, { "epoch": 3.040946896992962, "grad_norm": 1.54023008604072, "learning_rate": 1.6311892746016076e-05, "loss": 0.2562, "step": 4753 }, { "epoch": 3.0415866922584773, "grad_norm": 1.4588922436407163, "learning_rate": 1.6310285388123813e-05, "loss": 0.24, "step": 4754 }, { "epoch": 3.042226487523992, "grad_norm": 1.530361169416888, "learning_rate": 1.6308677759279123e-05, "loss": 0.2581, "step": 4755 }, { "epoch": 3.0428662827895074, "grad_norm": 1.556900708040586, "learning_rate": 1.6307069859551034e-05, "loss": 0.2717, "step": 4756 }, { "epoch": 3.0435060780550223, "grad_norm": 1.3589901559180588, "learning_rate": 1.6305461689008578e-05, "loss": 0.2219, "step": 4757 }, { "epoch": 3.0441458733205375, "grad_norm": 1.4756288882666588, "learning_rate": 1.630385324772082e-05, "loss": 0.2381, "step": 4758 }, { "epoch": 3.0447856685860524, "grad_norm": 1.524087285620321, "learning_rate": 1.6302244535756814e-05, "loss": 0.2473, "step": 4759 }, { "epoch": 3.0454254638515676, "grad_norm": 1.4501026372274357, "learning_rate": 1.630063555318564e-05, "loss": 0.2388, "step": 4760 }, { "epoch": 3.0460652591170825, "grad_norm": 1.4674103307007027, "learning_rate": 1.6299026300076382e-05, "loss": 0.2358, "step": 4761 }, { "epoch": 3.0467050543825978, "grad_norm": 1.4983274272414588, "learning_rate": 1.6297416776498143e-05, "loss": 0.2433, "step": 4762 }, { "epoch": 3.0473448496481126, "grad_norm": 1.680144587624245, "learning_rate": 1.6295806982520028e-05, "loss": 0.2768, "step": 4763 }, { "epoch": 3.047984644913628, "grad_norm": 1.552844178792695, "learning_rate": 1.629419691821116e-05, "loss": 0.2425, "step": 4764 }, { "epoch": 3.0486244401791427, "grad_norm": 1.4655542377541901, "learning_rate": 1.6292586583640672e-05, "loss": 0.2162, "step": 4765 }, { "epoch": 3.0492642354446575, "grad_norm": 1.519572099828501, "learning_rate": 1.629097597887771e-05, "loss": 0.2456, "step": 4766 }, { "epoch": 3.049904030710173, "grad_norm": 1.5358785053529114, "learning_rate": 1.628936510399143e-05, "loss": 0.2712, "step": 4767 }, { "epoch": 3.0505438259756876, "grad_norm": 1.5760444220281014, "learning_rate": 1.6287753959051003e-05, "loss": 0.2542, "step": 4768 }, { "epoch": 3.051183621241203, "grad_norm": 1.4952514327518724, "learning_rate": 1.62861425441256e-05, "loss": 0.2397, "step": 4769 }, { "epoch": 3.0518234165067177, "grad_norm": 1.57660760135019, "learning_rate": 1.6284530859284427e-05, "loss": 0.2572, "step": 4770 }, { "epoch": 3.052463211772233, "grad_norm": 1.3951725895286213, "learning_rate": 1.6282918904596675e-05, "loss": 0.225, "step": 4771 }, { "epoch": 3.053103007037748, "grad_norm": 1.4229005633388299, "learning_rate": 1.6281306680131555e-05, "loss": 0.2658, "step": 4772 }, { "epoch": 3.053742802303263, "grad_norm": 1.4547551576598556, "learning_rate": 1.6279694185958307e-05, "loss": 0.2544, "step": 4773 }, { "epoch": 3.054382597568778, "grad_norm": 1.3848856262693745, "learning_rate": 1.627808142214616e-05, "loss": 0.224, "step": 4774 }, { "epoch": 3.055022392834293, "grad_norm": 1.4788208978611495, "learning_rate": 1.6276468388764363e-05, "loss": 0.2544, "step": 4775 }, { "epoch": 3.055662188099808, "grad_norm": 1.460637086681454, "learning_rate": 1.627485508588218e-05, "loss": 0.2339, "step": 4776 }, { "epoch": 3.056301983365323, "grad_norm": 1.464873293448338, "learning_rate": 1.6273241513568884e-05, "loss": 0.2334, "step": 4777 }, { "epoch": 3.056941778630838, "grad_norm": 1.5557877419523112, "learning_rate": 1.6271627671893753e-05, "loss": 0.2474, "step": 4778 }, { "epoch": 3.057581573896353, "grad_norm": 1.5339872308693816, "learning_rate": 1.6270013560926086e-05, "loss": 0.2474, "step": 4779 }, { "epoch": 3.0582213691618683, "grad_norm": 1.6698188828360048, "learning_rate": 1.626839918073519e-05, "loss": 0.2613, "step": 4780 }, { "epoch": 3.058861164427383, "grad_norm": 1.6861632620373093, "learning_rate": 1.6266784531390386e-05, "loss": 0.2864, "step": 4781 }, { "epoch": 3.0595009596928984, "grad_norm": 1.6823708615416266, "learning_rate": 1.6265169612961e-05, "loss": 0.2892, "step": 4782 }, { "epoch": 3.060140754958413, "grad_norm": 1.4249305183755978, "learning_rate": 1.6263554425516378e-05, "loss": 0.2211, "step": 4783 }, { "epoch": 3.0607805502239285, "grad_norm": 1.4030769885120398, "learning_rate": 1.6261938969125867e-05, "loss": 0.2364, "step": 4784 }, { "epoch": 3.0614203454894433, "grad_norm": 1.508037637359766, "learning_rate": 1.626032324385884e-05, "loss": 0.2273, "step": 4785 }, { "epoch": 3.0620601407549586, "grad_norm": 1.4297279401713971, "learning_rate": 1.6258707249784666e-05, "loss": 0.2389, "step": 4786 }, { "epoch": 3.0626999360204734, "grad_norm": 1.3997750071348949, "learning_rate": 1.6257090986972738e-05, "loss": 0.2199, "step": 4787 }, { "epoch": 3.0633397312859887, "grad_norm": 1.4441686469535275, "learning_rate": 1.6255474455492455e-05, "loss": 0.2197, "step": 4788 }, { "epoch": 3.0639795265515035, "grad_norm": 1.5644980844462315, "learning_rate": 1.6253857655413226e-05, "loss": 0.2579, "step": 4789 }, { "epoch": 3.0646193218170183, "grad_norm": 1.7191512011151326, "learning_rate": 1.625224058680447e-05, "loss": 0.2534, "step": 4790 }, { "epoch": 3.0652591170825336, "grad_norm": 1.6360669692018845, "learning_rate": 1.625062324973563e-05, "loss": 0.2573, "step": 4791 }, { "epoch": 3.0658989123480485, "grad_norm": 1.436907319457157, "learning_rate": 1.6249005644276143e-05, "loss": 0.2546, "step": 4792 }, { "epoch": 3.0665387076135637, "grad_norm": 1.5516709240927027, "learning_rate": 1.6247387770495474e-05, "loss": 0.2314, "step": 4793 }, { "epoch": 3.0671785028790786, "grad_norm": 1.440667468547122, "learning_rate": 1.624576962846308e-05, "loss": 0.2586, "step": 4794 }, { "epoch": 3.067818298144594, "grad_norm": 1.5430883853368313, "learning_rate": 1.6244151218248453e-05, "loss": 0.2427, "step": 4795 }, { "epoch": 3.0684580934101087, "grad_norm": 1.5509957740333815, "learning_rate": 1.6242532539921084e-05, "loss": 0.2628, "step": 4796 }, { "epoch": 3.069097888675624, "grad_norm": 1.4166703174723057, "learning_rate": 1.624091359355047e-05, "loss": 0.2361, "step": 4797 }, { "epoch": 3.0697376839411388, "grad_norm": 1.5794914231784614, "learning_rate": 1.6239294379206123e-05, "loss": 0.2293, "step": 4798 }, { "epoch": 3.070377479206654, "grad_norm": 1.5550128762454116, "learning_rate": 1.623767489695758e-05, "loss": 0.2525, "step": 4799 }, { "epoch": 3.071017274472169, "grad_norm": 1.4777995400340378, "learning_rate": 1.6236055146874367e-05, "loss": 0.254, "step": 4800 }, { "epoch": 3.071657069737684, "grad_norm": 1.4971006598324388, "learning_rate": 1.6234435129026046e-05, "loss": 0.2412, "step": 4801 }, { "epoch": 3.072296865003199, "grad_norm": 1.633932047909267, "learning_rate": 1.6232814843482165e-05, "loss": 0.2705, "step": 4802 }, { "epoch": 3.072936660268714, "grad_norm": 1.559247595089351, "learning_rate": 1.6231194290312304e-05, "loss": 0.2598, "step": 4803 }, { "epoch": 3.073576455534229, "grad_norm": 1.52640118051112, "learning_rate": 1.6229573469586047e-05, "loss": 0.2712, "step": 4804 }, { "epoch": 3.074216250799744, "grad_norm": 1.5725202163191812, "learning_rate": 1.6227952381372983e-05, "loss": 0.2699, "step": 4805 }, { "epoch": 3.074856046065259, "grad_norm": 1.5337424822561934, "learning_rate": 1.6226331025742723e-05, "loss": 0.2385, "step": 4806 }, { "epoch": 3.075495841330774, "grad_norm": 1.4891291032954745, "learning_rate": 1.6224709402764887e-05, "loss": 0.234, "step": 4807 }, { "epoch": 3.0761356365962893, "grad_norm": 1.4924827625534434, "learning_rate": 1.62230875125091e-05, "loss": 0.2557, "step": 4808 }, { "epoch": 3.076775431861804, "grad_norm": 1.4883729288839274, "learning_rate": 1.6221465355045008e-05, "loss": 0.246, "step": 4809 }, { "epoch": 3.0774152271273194, "grad_norm": 1.570029912776763, "learning_rate": 1.621984293044226e-05, "loss": 0.2771, "step": 4810 }, { "epoch": 3.0780550223928342, "grad_norm": 1.513867851112407, "learning_rate": 1.621822023877052e-05, "loss": 0.2368, "step": 4811 }, { "epoch": 3.0786948176583495, "grad_norm": 1.3724375940703528, "learning_rate": 1.621659728009946e-05, "loss": 0.2396, "step": 4812 }, { "epoch": 3.0793346129238643, "grad_norm": 1.4936253568946023, "learning_rate": 1.6214974054498776e-05, "loss": 0.2522, "step": 4813 }, { "epoch": 3.079974408189379, "grad_norm": 1.5901003034133279, "learning_rate": 1.621335056203816e-05, "loss": 0.2678, "step": 4814 }, { "epoch": 3.0806142034548945, "grad_norm": 1.544082015291486, "learning_rate": 1.6211726802787328e-05, "loss": 0.2581, "step": 4815 }, { "epoch": 3.0812539987204093, "grad_norm": 1.600000124747705, "learning_rate": 1.621010277681599e-05, "loss": 0.2384, "step": 4816 }, { "epoch": 3.0818937939859246, "grad_norm": 1.507427652949829, "learning_rate": 1.620847848419389e-05, "loss": 0.2652, "step": 4817 }, { "epoch": 3.0825335892514394, "grad_norm": 1.4694364697185942, "learning_rate": 1.620685392499077e-05, "loss": 0.2565, "step": 4818 }, { "epoch": 3.0831733845169547, "grad_norm": 1.5201262018331287, "learning_rate": 1.620522909927638e-05, "loss": 0.2456, "step": 4819 }, { "epoch": 3.0838131797824695, "grad_norm": 1.4787569165185868, "learning_rate": 1.6203604007120487e-05, "loss": 0.2317, "step": 4820 }, { "epoch": 3.0844529750479848, "grad_norm": 1.472044741257508, "learning_rate": 1.620197864859288e-05, "loss": 0.2419, "step": 4821 }, { "epoch": 3.0850927703134996, "grad_norm": 1.4614271469234836, "learning_rate": 1.620035302376334e-05, "loss": 0.2453, "step": 4822 }, { "epoch": 3.085732565579015, "grad_norm": 1.5440405368144505, "learning_rate": 1.619872713270167e-05, "loss": 0.2427, "step": 4823 }, { "epoch": 3.0863723608445297, "grad_norm": 1.494163686325222, "learning_rate": 1.6197100975477685e-05, "loss": 0.2575, "step": 4824 }, { "epoch": 3.087012156110045, "grad_norm": 1.4722743675367032, "learning_rate": 1.619547455216121e-05, "loss": 0.2104, "step": 4825 }, { "epoch": 3.08765195137556, "grad_norm": 1.5430968868136115, "learning_rate": 1.619384786282207e-05, "loss": 0.2813, "step": 4826 }, { "epoch": 3.0882917466410746, "grad_norm": 1.508176416668283, "learning_rate": 1.619222090753013e-05, "loss": 0.2314, "step": 4827 }, { "epoch": 3.08893154190659, "grad_norm": 1.5403448666623312, "learning_rate": 1.6190593686355237e-05, "loss": 0.2411, "step": 4828 }, { "epoch": 3.0895713371721047, "grad_norm": 1.3799465397548119, "learning_rate": 1.618896619936726e-05, "loss": 0.2174, "step": 4829 }, { "epoch": 3.09021113243762, "grad_norm": 1.4320418141142122, "learning_rate": 1.6187338446636086e-05, "loss": 0.239, "step": 4830 }, { "epoch": 3.090850927703135, "grad_norm": 1.5622290269496555, "learning_rate": 1.6185710428231605e-05, "loss": 0.2502, "step": 4831 }, { "epoch": 3.09149072296865, "grad_norm": 1.510797191090967, "learning_rate": 1.618408214422372e-05, "loss": 0.2526, "step": 4832 }, { "epoch": 3.092130518234165, "grad_norm": 1.380209094932142, "learning_rate": 1.618245359468235e-05, "loss": 0.2301, "step": 4833 }, { "epoch": 3.0927703134996802, "grad_norm": 1.6327311943231555, "learning_rate": 1.618082477967742e-05, "loss": 0.2824, "step": 4834 }, { "epoch": 3.093410108765195, "grad_norm": 1.6576800127368987, "learning_rate": 1.6179195699278874e-05, "loss": 0.269, "step": 4835 }, { "epoch": 3.0940499040307103, "grad_norm": 1.4942901095796266, "learning_rate": 1.617756635355665e-05, "loss": 0.2311, "step": 4836 }, { "epoch": 3.094689699296225, "grad_norm": 1.446119369788071, "learning_rate": 1.617593674258072e-05, "loss": 0.2458, "step": 4837 }, { "epoch": 3.0953294945617404, "grad_norm": 1.5534766503814983, "learning_rate": 1.617430686642105e-05, "loss": 0.2711, "step": 4838 }, { "epoch": 3.0959692898272553, "grad_norm": 1.4474475246441467, "learning_rate": 1.6172676725147626e-05, "loss": 0.2375, "step": 4839 }, { "epoch": 3.09660908509277, "grad_norm": 1.451023144737553, "learning_rate": 1.6171046318830446e-05, "loss": 0.2351, "step": 4840 }, { "epoch": 3.0972488803582854, "grad_norm": 1.464876630708972, "learning_rate": 1.6169415647539516e-05, "loss": 0.2191, "step": 4841 }, { "epoch": 3.0978886756238, "grad_norm": 1.5036030546525587, "learning_rate": 1.6167784711344846e-05, "loss": 0.2706, "step": 4842 }, { "epoch": 3.0985284708893155, "grad_norm": 1.4675526036189168, "learning_rate": 1.616615351031648e-05, "loss": 0.2372, "step": 4843 }, { "epoch": 3.0991682661548303, "grad_norm": 1.669881690037458, "learning_rate": 1.6164522044524444e-05, "loss": 0.2604, "step": 4844 }, { "epoch": 3.0998080614203456, "grad_norm": 1.6648791058518375, "learning_rate": 1.61628903140388e-05, "loss": 0.2907, "step": 4845 }, { "epoch": 3.1004478566858604, "grad_norm": 1.5205644622763677, "learning_rate": 1.6161258318929608e-05, "loss": 0.2394, "step": 4846 }, { "epoch": 3.1010876519513757, "grad_norm": 1.495569235012655, "learning_rate": 1.6159626059266944e-05, "loss": 0.2661, "step": 4847 }, { "epoch": 3.1017274472168905, "grad_norm": 1.5302491994604235, "learning_rate": 1.6157993535120895e-05, "loss": 0.2479, "step": 4848 }, { "epoch": 3.102367242482406, "grad_norm": 1.394287737989242, "learning_rate": 1.615636074656156e-05, "loss": 0.2329, "step": 4849 }, { "epoch": 3.1030070377479206, "grad_norm": 1.4757475035532241, "learning_rate": 1.6154727693659045e-05, "loss": 0.2517, "step": 4850 }, { "epoch": 3.103646833013436, "grad_norm": 1.522301734203222, "learning_rate": 1.6153094376483468e-05, "loss": 0.255, "step": 4851 }, { "epoch": 3.1042866282789507, "grad_norm": 1.446018352771596, "learning_rate": 1.6151460795104966e-05, "loss": 0.2195, "step": 4852 }, { "epoch": 3.1049264235444656, "grad_norm": 1.706410108970667, "learning_rate": 1.614982694959368e-05, "loss": 0.274, "step": 4853 }, { "epoch": 3.105566218809981, "grad_norm": 1.6658951591709847, "learning_rate": 1.6148192840019766e-05, "loss": 0.2667, "step": 4854 }, { "epoch": 3.1062060140754957, "grad_norm": 1.5435137012332778, "learning_rate": 1.6146558466453384e-05, "loss": 0.2695, "step": 4855 }, { "epoch": 3.106845809341011, "grad_norm": 1.6081783624530623, "learning_rate": 1.614492382896472e-05, "loss": 0.2691, "step": 4856 }, { "epoch": 3.107485604606526, "grad_norm": 1.457168383382789, "learning_rate": 1.6143288927623954e-05, "loss": 0.2524, "step": 4857 }, { "epoch": 3.108125399872041, "grad_norm": 1.478658742852597, "learning_rate": 1.614165376250129e-05, "loss": 0.2325, "step": 4858 }, { "epoch": 3.108765195137556, "grad_norm": 1.5750230774584515, "learning_rate": 1.614001833366694e-05, "loss": 0.2309, "step": 4859 }, { "epoch": 3.109404990403071, "grad_norm": 1.40064364096898, "learning_rate": 1.613838264119113e-05, "loss": 0.2273, "step": 4860 }, { "epoch": 3.110044785668586, "grad_norm": 1.534642814025407, "learning_rate": 1.613674668514408e-05, "loss": 0.2697, "step": 4861 }, { "epoch": 3.1106845809341013, "grad_norm": 1.5264441848541657, "learning_rate": 1.6135110465596045e-05, "loss": 0.2495, "step": 4862 }, { "epoch": 3.111324376199616, "grad_norm": 1.5379790945907832, "learning_rate": 1.6133473982617284e-05, "loss": 0.2494, "step": 4863 }, { "epoch": 3.1119641714651314, "grad_norm": 1.4876656359912137, "learning_rate": 1.613183723627806e-05, "loss": 0.2482, "step": 4864 }, { "epoch": 3.112603966730646, "grad_norm": 1.494868325724146, "learning_rate": 1.6130200226648653e-05, "loss": 0.2737, "step": 4865 }, { "epoch": 3.113243761996161, "grad_norm": 1.6268224323799703, "learning_rate": 1.6128562953799352e-05, "loss": 0.2445, "step": 4866 }, { "epoch": 3.1138835572616763, "grad_norm": 1.560106004111509, "learning_rate": 1.612692541780046e-05, "loss": 0.2331, "step": 4867 }, { "epoch": 3.114523352527191, "grad_norm": 1.4834024328410955, "learning_rate": 1.6125287618722288e-05, "loss": 0.2309, "step": 4868 }, { "epoch": 3.1151631477927064, "grad_norm": 1.5709394190255501, "learning_rate": 1.6123649556635165e-05, "loss": 0.2553, "step": 4869 }, { "epoch": 3.1158029430582213, "grad_norm": 1.4304024012503687, "learning_rate": 1.612201123160942e-05, "loss": 0.239, "step": 4870 }, { "epoch": 3.1164427383237365, "grad_norm": 1.4774176725689026, "learning_rate": 1.6120372643715403e-05, "loss": 0.2407, "step": 4871 }, { "epoch": 3.1170825335892514, "grad_norm": 1.3917886061024265, "learning_rate": 1.6118733793023473e-05, "loss": 0.2194, "step": 4872 }, { "epoch": 3.1177223288547666, "grad_norm": 1.5191823403057716, "learning_rate": 1.6117094679604e-05, "loss": 0.2531, "step": 4873 }, { "epoch": 3.1183621241202815, "grad_norm": 1.4419876685262514, "learning_rate": 1.611545530352736e-05, "loss": 0.2501, "step": 4874 }, { "epoch": 3.1190019193857967, "grad_norm": 1.542453167614456, "learning_rate": 1.611381566486395e-05, "loss": 0.2798, "step": 4875 }, { "epoch": 3.1196417146513116, "grad_norm": 1.5694848126251164, "learning_rate": 1.611217576368417e-05, "loss": 0.2618, "step": 4876 }, { "epoch": 3.1202815099168264, "grad_norm": 1.616851079759159, "learning_rate": 1.611053560005843e-05, "loss": 0.2532, "step": 4877 }, { "epoch": 3.1209213051823417, "grad_norm": 1.691459184992037, "learning_rate": 1.610889517405717e-05, "loss": 0.2858, "step": 4878 }, { "epoch": 3.1215611004478565, "grad_norm": 1.6076855141808668, "learning_rate": 1.6107254485750815e-05, "loss": 0.2551, "step": 4879 }, { "epoch": 3.122200895713372, "grad_norm": 1.6142067799757498, "learning_rate": 1.6105613535209815e-05, "loss": 0.2749, "step": 4880 }, { "epoch": 3.1228406909788866, "grad_norm": 1.3620481685897394, "learning_rate": 1.6103972322504634e-05, "loss": 0.2191, "step": 4881 }, { "epoch": 3.123480486244402, "grad_norm": 1.5608105331601003, "learning_rate": 1.6102330847705737e-05, "loss": 0.2642, "step": 4882 }, { "epoch": 3.1241202815099167, "grad_norm": 1.5059313885559693, "learning_rate": 1.610068911088361e-05, "loss": 0.2619, "step": 4883 }, { "epoch": 3.124760076775432, "grad_norm": 1.4856295716039591, "learning_rate": 1.6099047112108742e-05, "loss": 0.252, "step": 4884 }, { "epoch": 3.125399872040947, "grad_norm": 1.5104818016644588, "learning_rate": 1.6097404851451644e-05, "loss": 0.2444, "step": 4885 }, { "epoch": 3.126039667306462, "grad_norm": 1.5044298030111785, "learning_rate": 1.6095762328982828e-05, "loss": 0.2322, "step": 4886 }, { "epoch": 3.126679462571977, "grad_norm": 1.5151834981887466, "learning_rate": 1.6094119544772815e-05, "loss": 0.2679, "step": 4887 }, { "epoch": 3.127319257837492, "grad_norm": 1.4685059616870784, "learning_rate": 1.6092476498892157e-05, "loss": 0.2463, "step": 4888 }, { "epoch": 3.127959053103007, "grad_norm": 1.5601365570084629, "learning_rate": 1.609083319141139e-05, "loss": 0.2625, "step": 4889 }, { "epoch": 3.128598848368522, "grad_norm": 1.5291791819513942, "learning_rate": 1.6089189622401085e-05, "loss": 0.2487, "step": 4890 }, { "epoch": 3.129238643634037, "grad_norm": 1.4631616013104995, "learning_rate": 1.6087545791931805e-05, "loss": 0.2353, "step": 4891 }, { "epoch": 3.129878438899552, "grad_norm": 1.4951412916239715, "learning_rate": 1.608590170007414e-05, "loss": 0.2611, "step": 4892 }, { "epoch": 3.1305182341650672, "grad_norm": 1.5375153473903134, "learning_rate": 1.6084257346898685e-05, "loss": 0.2383, "step": 4893 }, { "epoch": 3.131158029430582, "grad_norm": 1.3823261134731515, "learning_rate": 1.608261273247604e-05, "loss": 0.2181, "step": 4894 }, { "epoch": 3.1317978246960974, "grad_norm": 1.5036400256014348, "learning_rate": 1.6080967856876822e-05, "loss": 0.2578, "step": 4895 }, { "epoch": 3.132437619961612, "grad_norm": 1.4529443028117754, "learning_rate": 1.6079322720171665e-05, "loss": 0.2321, "step": 4896 }, { "epoch": 3.1330774152271275, "grad_norm": 1.6044323026750795, "learning_rate": 1.6077677322431203e-05, "loss": 0.2637, "step": 4897 }, { "epoch": 3.1337172104926423, "grad_norm": 1.514883587309794, "learning_rate": 1.6076031663726088e-05, "loss": 0.2644, "step": 4898 }, { "epoch": 3.1343570057581576, "grad_norm": 1.4169349111840939, "learning_rate": 1.607438574412698e-05, "loss": 0.2151, "step": 4899 }, { "epoch": 3.1349968010236724, "grad_norm": 1.4563263681508392, "learning_rate": 1.607273956370456e-05, "loss": 0.228, "step": 4900 }, { "epoch": 3.1356365962891877, "grad_norm": 1.5961236564162535, "learning_rate": 1.6071093122529503e-05, "loss": 0.2827, "step": 4901 }, { "epoch": 3.1362763915547025, "grad_norm": 1.5086007211981314, "learning_rate": 1.60694464206725e-05, "loss": 0.2603, "step": 4902 }, { "epoch": 3.1369161868202173, "grad_norm": 1.6333831191223764, "learning_rate": 1.6067799458204275e-05, "loss": 0.26, "step": 4903 }, { "epoch": 3.1375559820857326, "grad_norm": 1.4205029949166599, "learning_rate": 1.6066152235195527e-05, "loss": 0.2094, "step": 4904 }, { "epoch": 3.1381957773512474, "grad_norm": 1.538400382459196, "learning_rate": 1.6064504751716997e-05, "loss": 0.2308, "step": 4905 }, { "epoch": 3.1388355726167627, "grad_norm": 1.4495458389315927, "learning_rate": 1.606285700783942e-05, "loss": 0.2345, "step": 4906 }, { "epoch": 3.1394753678822775, "grad_norm": 1.5493395392412401, "learning_rate": 1.6061209003633544e-05, "loss": 0.2498, "step": 4907 }, { "epoch": 3.140115163147793, "grad_norm": 1.5011008259883116, "learning_rate": 1.605956073917014e-05, "loss": 0.254, "step": 4908 }, { "epoch": 3.1407549584133077, "grad_norm": 1.670887570007231, "learning_rate": 1.6057912214519978e-05, "loss": 0.2477, "step": 4909 }, { "epoch": 3.141394753678823, "grad_norm": 1.5299863210407076, "learning_rate": 1.605626342975384e-05, "loss": 0.2679, "step": 4910 }, { "epoch": 3.1420345489443378, "grad_norm": 1.391690897306588, "learning_rate": 1.605461438494252e-05, "loss": 0.2407, "step": 4911 }, { "epoch": 3.142674344209853, "grad_norm": 1.5164988095995222, "learning_rate": 1.6052965080156836e-05, "loss": 0.2501, "step": 4912 }, { "epoch": 3.143314139475368, "grad_norm": 1.492307253913446, "learning_rate": 1.6051315515467597e-05, "loss": 0.2392, "step": 4913 }, { "epoch": 3.1439539347408827, "grad_norm": 1.4277083947067029, "learning_rate": 1.604966569094563e-05, "loss": 0.2257, "step": 4914 }, { "epoch": 3.144593730006398, "grad_norm": 1.4902263426911966, "learning_rate": 1.6048015606661782e-05, "loss": 0.2553, "step": 4915 }, { "epoch": 3.145233525271913, "grad_norm": 1.4411972364399126, "learning_rate": 1.6046365262686903e-05, "loss": 0.2301, "step": 4916 }, { "epoch": 3.145873320537428, "grad_norm": 1.5272809852735378, "learning_rate": 1.6044714659091855e-05, "loss": 0.251, "step": 4917 }, { "epoch": 3.146513115802943, "grad_norm": 1.5013139001342635, "learning_rate": 1.604306379594751e-05, "loss": 0.2533, "step": 4918 }, { "epoch": 3.147152911068458, "grad_norm": 1.421215437256302, "learning_rate": 1.6041412673324757e-05, "loss": 0.2476, "step": 4919 }, { "epoch": 3.147792706333973, "grad_norm": 1.507282248981991, "learning_rate": 1.6039761291294493e-05, "loss": 0.2569, "step": 4920 }, { "epoch": 3.1484325015994883, "grad_norm": 1.4873677589087264, "learning_rate": 1.603810964992762e-05, "loss": 0.2425, "step": 4921 }, { "epoch": 3.149072296865003, "grad_norm": 1.5176637969912394, "learning_rate": 1.6036457749295063e-05, "loss": 0.2531, "step": 4922 }, { "epoch": 3.1497120921305184, "grad_norm": 1.4724211721147948, "learning_rate": 1.6034805589467743e-05, "loss": 0.2418, "step": 4923 }, { "epoch": 3.1503518873960332, "grad_norm": 1.526113730754682, "learning_rate": 1.6033153170516608e-05, "loss": 0.2684, "step": 4924 }, { "epoch": 3.1509916826615485, "grad_norm": 1.606952192344439, "learning_rate": 1.603150049251261e-05, "loss": 0.2569, "step": 4925 }, { "epoch": 3.1516314779270633, "grad_norm": 1.4706243528526484, "learning_rate": 1.602984755552671e-05, "loss": 0.2531, "step": 4926 }, { "epoch": 3.1522712731925786, "grad_norm": 1.589488093599225, "learning_rate": 1.602819435962988e-05, "loss": 0.2623, "step": 4927 }, { "epoch": 3.1529110684580934, "grad_norm": 1.5991592922369915, "learning_rate": 1.602654090489311e-05, "loss": 0.2669, "step": 4928 }, { "epoch": 3.1535508637236083, "grad_norm": 1.5192113558324527, "learning_rate": 1.602488719138739e-05, "loss": 0.2408, "step": 4929 }, { "epoch": 3.1541906589891235, "grad_norm": 1.4061757028590824, "learning_rate": 1.6023233219183737e-05, "loss": 0.2585, "step": 4930 }, { "epoch": 3.1548304542546384, "grad_norm": 1.5093435684502867, "learning_rate": 1.602157898835316e-05, "loss": 0.2343, "step": 4931 }, { "epoch": 3.1554702495201536, "grad_norm": 1.575661878081255, "learning_rate": 1.6019924498966694e-05, "loss": 0.2695, "step": 4932 }, { "epoch": 3.1561100447856685, "grad_norm": 1.4787735859326425, "learning_rate": 1.6018269751095377e-05, "loss": 0.2249, "step": 4933 }, { "epoch": 3.1567498400511838, "grad_norm": 1.5907160979391, "learning_rate": 1.6016614744810267e-05, "loss": 0.2572, "step": 4934 }, { "epoch": 3.1573896353166986, "grad_norm": 1.4209498935905525, "learning_rate": 1.6014959480182418e-05, "loss": 0.2207, "step": 4935 }, { "epoch": 3.158029430582214, "grad_norm": 1.5716877228786437, "learning_rate": 1.601330395728291e-05, "loss": 0.2636, "step": 4936 }, { "epoch": 3.1586692258477287, "grad_norm": 1.548730885601437, "learning_rate": 1.6011648176182828e-05, "loss": 0.2423, "step": 4937 }, { "epoch": 3.159309021113244, "grad_norm": 1.472059484632849, "learning_rate": 1.6009992136953266e-05, "loss": 0.248, "step": 4938 }, { "epoch": 3.159948816378759, "grad_norm": 1.5073035749208448, "learning_rate": 1.6008335839665332e-05, "loss": 0.223, "step": 4939 }, { "epoch": 3.1605886116442736, "grad_norm": 1.6347232423373281, "learning_rate": 1.6006679284390146e-05, "loss": 0.2536, "step": 4940 }, { "epoch": 3.161228406909789, "grad_norm": 1.4472143687286954, "learning_rate": 1.600502247119884e-05, "loss": 0.2418, "step": 4941 }, { "epoch": 3.1618682021753037, "grad_norm": 1.4418047078889544, "learning_rate": 1.6003365400162546e-05, "loss": 0.2561, "step": 4942 }, { "epoch": 3.162507997440819, "grad_norm": 1.4286856211878127, "learning_rate": 1.600170807135242e-05, "loss": 0.2186, "step": 4943 }, { "epoch": 3.163147792706334, "grad_norm": 1.5802853666826602, "learning_rate": 1.6000050484839632e-05, "loss": 0.2835, "step": 4944 }, { "epoch": 3.163787587971849, "grad_norm": 1.5355182816510968, "learning_rate": 1.5998392640695346e-05, "loss": 0.2689, "step": 4945 }, { "epoch": 3.164427383237364, "grad_norm": 1.5384710603965326, "learning_rate": 1.599673453899075e-05, "loss": 0.2245, "step": 4946 }, { "epoch": 3.165067178502879, "grad_norm": 1.47027789237415, "learning_rate": 1.599507617979704e-05, "loss": 0.2416, "step": 4947 }, { "epoch": 3.165706973768394, "grad_norm": 1.4962248694632587, "learning_rate": 1.5993417563185425e-05, "loss": 0.2481, "step": 4948 }, { "epoch": 3.1663467690339093, "grad_norm": 1.5415195387688636, "learning_rate": 1.5991758689227118e-05, "loss": 0.2641, "step": 4949 }, { "epoch": 3.166986564299424, "grad_norm": 1.5937126580701935, "learning_rate": 1.5990099557993354e-05, "loss": 0.2493, "step": 4950 }, { "epoch": 3.167626359564939, "grad_norm": 1.5285139138388006, "learning_rate": 1.598844016955537e-05, "loss": 0.2656, "step": 4951 }, { "epoch": 3.1682661548304543, "grad_norm": 1.5591992314976277, "learning_rate": 1.5986780523984416e-05, "loss": 0.2516, "step": 4952 }, { "epoch": 3.168905950095969, "grad_norm": 1.517925187790386, "learning_rate": 1.598512062135176e-05, "loss": 0.2508, "step": 4953 }, { "epoch": 3.1695457453614844, "grad_norm": 1.519719756234013, "learning_rate": 1.598346046172867e-05, "loss": 0.2562, "step": 4954 }, { "epoch": 3.170185540626999, "grad_norm": 1.3891677440245016, "learning_rate": 1.598180004518643e-05, "loss": 0.2277, "step": 4955 }, { "epoch": 3.1708253358925145, "grad_norm": 1.651103408296251, "learning_rate": 1.5980139371796333e-05, "loss": 0.2574, "step": 4956 }, { "epoch": 3.1714651311580293, "grad_norm": 1.7834582139502415, "learning_rate": 1.5978478441629697e-05, "loss": 0.2793, "step": 4957 }, { "epoch": 3.1721049264235446, "grad_norm": 1.4631310317237023, "learning_rate": 1.597681725475783e-05, "loss": 0.2305, "step": 4958 }, { "epoch": 3.1727447216890594, "grad_norm": 1.650164152652009, "learning_rate": 1.5975155811252056e-05, "loss": 0.2704, "step": 4959 }, { "epoch": 3.1733845169545747, "grad_norm": 1.7750142866019405, "learning_rate": 1.5973494111183726e-05, "loss": 0.2981, "step": 4960 }, { "epoch": 3.1740243122200895, "grad_norm": 1.3174629816759738, "learning_rate": 1.5971832154624187e-05, "loss": 0.2164, "step": 4961 }, { "epoch": 3.174664107485605, "grad_norm": 1.6137997649919624, "learning_rate": 1.5970169941644793e-05, "loss": 0.2645, "step": 4962 }, { "epoch": 3.1753039027511196, "grad_norm": 1.515454823797831, "learning_rate": 1.5968507472316922e-05, "loss": 0.2356, "step": 4963 }, { "epoch": 3.175943698016635, "grad_norm": 1.4830209237108296, "learning_rate": 1.596684474671196e-05, "loss": 0.2449, "step": 4964 }, { "epoch": 3.1765834932821497, "grad_norm": 1.4499859694990314, "learning_rate": 1.5965181764901298e-05, "loss": 0.2287, "step": 4965 }, { "epoch": 3.1772232885476646, "grad_norm": 1.5357408050416321, "learning_rate": 1.596351852695634e-05, "loss": 0.2481, "step": 4966 }, { "epoch": 3.17786308381318, "grad_norm": 1.5498570450012743, "learning_rate": 1.5961855032948508e-05, "loss": 0.2535, "step": 4967 }, { "epoch": 3.1785028790786947, "grad_norm": 1.576088557045589, "learning_rate": 1.5960191282949224e-05, "loss": 0.2842, "step": 4968 }, { "epoch": 3.17914267434421, "grad_norm": 1.5261088923075627, "learning_rate": 1.595852727702993e-05, "loss": 0.2382, "step": 4969 }, { "epoch": 3.1797824696097248, "grad_norm": 1.3960879658605352, "learning_rate": 1.5956863015262075e-05, "loss": 0.2357, "step": 4970 }, { "epoch": 3.18042226487524, "grad_norm": 1.4921281472661836, "learning_rate": 1.5955198497717116e-05, "loss": 0.2574, "step": 4971 }, { "epoch": 3.181062060140755, "grad_norm": 1.5666448539000244, "learning_rate": 1.5953533724466527e-05, "loss": 0.2646, "step": 4972 }, { "epoch": 3.18170185540627, "grad_norm": 1.636764834910487, "learning_rate": 1.5951868695581794e-05, "loss": 0.2757, "step": 4973 }, { "epoch": 3.182341650671785, "grad_norm": 1.5167847757858344, "learning_rate": 1.59502034111344e-05, "loss": 0.2507, "step": 4974 }, { "epoch": 3.1829814459373003, "grad_norm": 1.4610224968059384, "learning_rate": 1.5948537871195863e-05, "loss": 0.2378, "step": 4975 }, { "epoch": 3.183621241202815, "grad_norm": 1.617629041458487, "learning_rate": 1.594687207583769e-05, "loss": 0.2574, "step": 4976 }, { "epoch": 3.18426103646833, "grad_norm": 1.5602956252795013, "learning_rate": 1.5945206025131404e-05, "loss": 0.245, "step": 4977 }, { "epoch": 3.184900831733845, "grad_norm": 1.4960738871354526, "learning_rate": 1.5943539719148554e-05, "loss": 0.2448, "step": 4978 }, { "epoch": 3.18554062699936, "grad_norm": 1.522623462221726, "learning_rate": 1.5941873157960677e-05, "loss": 0.2736, "step": 4979 }, { "epoch": 3.1861804222648753, "grad_norm": 1.6543253595178387, "learning_rate": 1.5940206341639338e-05, "loss": 0.2779, "step": 4980 }, { "epoch": 3.18682021753039, "grad_norm": 1.4339880599122812, "learning_rate": 1.5938539270256104e-05, "loss": 0.2268, "step": 4981 }, { "epoch": 3.1874600127959054, "grad_norm": 1.4934064326353698, "learning_rate": 1.593687194388256e-05, "loss": 0.2508, "step": 4982 }, { "epoch": 3.1880998080614202, "grad_norm": 1.5451369301198536, "learning_rate": 1.5935204362590292e-05, "loss": 0.2537, "step": 4983 }, { "epoch": 3.1887396033269355, "grad_norm": 1.5007059111647902, "learning_rate": 1.5933536526450907e-05, "loss": 0.2344, "step": 4984 }, { "epoch": 3.1893793985924503, "grad_norm": 1.5333225940014874, "learning_rate": 1.593186843553602e-05, "loss": 0.2465, "step": 4985 }, { "epoch": 3.1900191938579656, "grad_norm": 1.6068945498803269, "learning_rate": 1.5930200089917252e-05, "loss": 0.2884, "step": 4986 }, { "epoch": 3.1906589891234804, "grad_norm": 1.4084612044870581, "learning_rate": 1.5928531489666247e-05, "loss": 0.2278, "step": 4987 }, { "epoch": 3.1912987843889957, "grad_norm": 1.6918603076386876, "learning_rate": 1.5926862634854644e-05, "loss": 0.2769, "step": 4988 }, { "epoch": 3.1919385796545106, "grad_norm": 1.549234235322282, "learning_rate": 1.59251935255541e-05, "loss": 0.2576, "step": 4989 }, { "epoch": 3.1925783749200254, "grad_norm": 1.4181412077886508, "learning_rate": 1.5923524161836287e-05, "loss": 0.2224, "step": 4990 }, { "epoch": 3.1932181701855407, "grad_norm": 1.520937854349658, "learning_rate": 1.592185454377288e-05, "loss": 0.2467, "step": 4991 }, { "epoch": 3.1938579654510555, "grad_norm": 1.5913927174498865, "learning_rate": 1.592018467143558e-05, "loss": 0.2395, "step": 4992 }, { "epoch": 3.1944977607165708, "grad_norm": 1.479264511081746, "learning_rate": 1.5918514544896077e-05, "loss": 0.2239, "step": 4993 }, { "epoch": 3.1951375559820856, "grad_norm": 1.4258171019693535, "learning_rate": 1.591684416422609e-05, "loss": 0.2294, "step": 4994 }, { "epoch": 3.195777351247601, "grad_norm": 1.688344304870728, "learning_rate": 1.5915173529497337e-05, "loss": 0.2592, "step": 4995 }, { "epoch": 3.1964171465131157, "grad_norm": 1.5251958162829937, "learning_rate": 1.5913502640781554e-05, "loss": 0.2457, "step": 4996 }, { "epoch": 3.197056941778631, "grad_norm": 1.5196549855463326, "learning_rate": 1.5911831498150492e-05, "loss": 0.2407, "step": 4997 }, { "epoch": 3.197696737044146, "grad_norm": 1.5996464430234074, "learning_rate": 1.5910160101675896e-05, "loss": 0.264, "step": 4998 }, { "epoch": 3.198336532309661, "grad_norm": 1.553433705259062, "learning_rate": 1.5908488451429546e-05, "loss": 0.2561, "step": 4999 }, { "epoch": 3.198976327575176, "grad_norm": 1.384369246269086, "learning_rate": 1.5906816547483207e-05, "loss": 0.2607, "step": 5000 }, { "epoch": 3.199616122840691, "grad_norm": 1.561316689788971, "learning_rate": 1.590514438990867e-05, "loss": 0.2543, "step": 5001 }, { "epoch": 3.200255918106206, "grad_norm": 1.515767379504214, "learning_rate": 1.5903471978777743e-05, "loss": 0.2452, "step": 5002 }, { "epoch": 3.200895713371721, "grad_norm": 1.4509714889211538, "learning_rate": 1.590179931416223e-05, "loss": 0.2463, "step": 5003 }, { "epoch": 3.201535508637236, "grad_norm": 1.46854435240591, "learning_rate": 1.5900126396133953e-05, "loss": 0.2351, "step": 5004 }, { "epoch": 3.202175303902751, "grad_norm": 1.5412612190794301, "learning_rate": 1.589845322476474e-05, "loss": 0.2513, "step": 5005 }, { "epoch": 3.2028150991682662, "grad_norm": 1.4957979944177626, "learning_rate": 1.5896779800126444e-05, "loss": 0.2344, "step": 5006 }, { "epoch": 3.203454894433781, "grad_norm": 1.4766718254412585, "learning_rate": 1.589510612229091e-05, "loss": 0.2558, "step": 5007 }, { "epoch": 3.2040946896992963, "grad_norm": 1.506035644476434, "learning_rate": 1.589343219133e-05, "loss": 0.2419, "step": 5008 }, { "epoch": 3.204734484964811, "grad_norm": 1.5354552114057272, "learning_rate": 1.5891758007315604e-05, "loss": 0.2286, "step": 5009 }, { "epoch": 3.2053742802303264, "grad_norm": 1.6934166294125825, "learning_rate": 1.5890083570319598e-05, "loss": 0.302, "step": 5010 }, { "epoch": 3.2060140754958413, "grad_norm": 1.5377962014410613, "learning_rate": 1.588840888041388e-05, "loss": 0.2602, "step": 5011 }, { "epoch": 3.2066538707613566, "grad_norm": 1.427500175163889, "learning_rate": 1.588673393767036e-05, "loss": 0.2492, "step": 5012 }, { "epoch": 3.2072936660268714, "grad_norm": 1.5627699424870432, "learning_rate": 1.5885058742160954e-05, "loss": 0.2524, "step": 5013 }, { "epoch": 3.207933461292386, "grad_norm": 1.569079759494596, "learning_rate": 1.58833832939576e-05, "loss": 0.2738, "step": 5014 }, { "epoch": 3.2085732565579015, "grad_norm": 1.3449910109279775, "learning_rate": 1.5881707593132226e-05, "loss": 0.2291, "step": 5015 }, { "epoch": 3.2092130518234163, "grad_norm": 1.4804806044871233, "learning_rate": 1.588003163975679e-05, "loss": 0.2379, "step": 5016 }, { "epoch": 3.2098528470889316, "grad_norm": 1.6285226502688648, "learning_rate": 1.5878355433903263e-05, "loss": 0.271, "step": 5017 }, { "epoch": 3.2104926423544464, "grad_norm": 1.4900079843611953, "learning_rate": 1.5876678975643604e-05, "loss": 0.2344, "step": 5018 }, { "epoch": 3.2111324376199617, "grad_norm": 1.5372720391209191, "learning_rate": 1.5875002265049808e-05, "loss": 0.2659, "step": 5019 }, { "epoch": 3.2117722328854765, "grad_norm": 1.5338861642905444, "learning_rate": 1.5873325302193862e-05, "loss": 0.2318, "step": 5020 }, { "epoch": 3.212412028150992, "grad_norm": 1.4371844154824218, "learning_rate": 1.5871648087147774e-05, "loss": 0.2427, "step": 5021 }, { "epoch": 3.2130518234165066, "grad_norm": 1.4767819755553582, "learning_rate": 1.5869970619983567e-05, "loss": 0.2363, "step": 5022 }, { "epoch": 3.213691618682022, "grad_norm": 1.4961409548549114, "learning_rate": 1.586829290077326e-05, "loss": 0.2791, "step": 5023 }, { "epoch": 3.2143314139475367, "grad_norm": 1.5027675913619645, "learning_rate": 1.5866614929588898e-05, "loss": 0.2469, "step": 5024 }, { "epoch": 3.214971209213052, "grad_norm": 1.5275474164846263, "learning_rate": 1.5864936706502524e-05, "loss": 0.2604, "step": 5025 }, { "epoch": 3.215611004478567, "grad_norm": 1.7416872596121125, "learning_rate": 1.58632582315862e-05, "loss": 0.2606, "step": 5026 }, { "epoch": 3.216250799744082, "grad_norm": 1.4982808440903956, "learning_rate": 1.5861579504912003e-05, "loss": 0.2389, "step": 5027 }, { "epoch": 3.216890595009597, "grad_norm": 1.6928027973826896, "learning_rate": 1.5859900526552005e-05, "loss": 0.299, "step": 5028 }, { "epoch": 3.217530390275112, "grad_norm": 1.5983788197299447, "learning_rate": 1.58582212965783e-05, "loss": 0.2769, "step": 5029 }, { "epoch": 3.218170185540627, "grad_norm": 1.6994645149336063, "learning_rate": 1.5856541815062995e-05, "loss": 0.2488, "step": 5030 }, { "epoch": 3.218809980806142, "grad_norm": 1.6072664351820367, "learning_rate": 1.5854862082078203e-05, "loss": 0.2615, "step": 5031 }, { "epoch": 3.219449776071657, "grad_norm": 1.5413090687089919, "learning_rate": 1.585318209769605e-05, "loss": 0.2735, "step": 5032 }, { "epoch": 3.220089571337172, "grad_norm": 1.5579554917625231, "learning_rate": 1.5851501861988668e-05, "loss": 0.2407, "step": 5033 }, { "epoch": 3.2207293666026873, "grad_norm": 1.3472948078264868, "learning_rate": 1.5849821375028204e-05, "loss": 0.2144, "step": 5034 }, { "epoch": 3.221369161868202, "grad_norm": 1.582570720909756, "learning_rate": 1.584814063688682e-05, "loss": 0.2871, "step": 5035 }, { "epoch": 3.2220089571337174, "grad_norm": 1.5997761336726595, "learning_rate": 1.5846459647636675e-05, "loss": 0.2561, "step": 5036 }, { "epoch": 3.222648752399232, "grad_norm": 1.4585837805560133, "learning_rate": 1.5844778407349956e-05, "loss": 0.2524, "step": 5037 }, { "epoch": 3.2232885476647475, "grad_norm": 1.616690876386779, "learning_rate": 1.5843096916098848e-05, "loss": 0.2942, "step": 5038 }, { "epoch": 3.2239283429302623, "grad_norm": 1.5364179021993285, "learning_rate": 1.5841415173955554e-05, "loss": 0.2423, "step": 5039 }, { "epoch": 3.224568138195777, "grad_norm": 1.5883778129697173, "learning_rate": 1.5839733180992283e-05, "loss": 0.2626, "step": 5040 }, { "epoch": 3.2252079334612924, "grad_norm": 1.5140771941868278, "learning_rate": 1.5838050937281255e-05, "loss": 0.2394, "step": 5041 }, { "epoch": 3.2258477287268073, "grad_norm": 1.3643116582260755, "learning_rate": 1.5836368442894706e-05, "loss": 0.2402, "step": 5042 }, { "epoch": 3.2264875239923225, "grad_norm": 1.4736215990322357, "learning_rate": 1.583468569790488e-05, "loss": 0.236, "step": 5043 }, { "epoch": 3.2271273192578374, "grad_norm": 1.567612525370598, "learning_rate": 1.5833002702384028e-05, "loss": 0.2486, "step": 5044 }, { "epoch": 3.2277671145233526, "grad_norm": 1.6509906583351521, "learning_rate": 1.5831319456404416e-05, "loss": 0.2934, "step": 5045 }, { "epoch": 3.2284069097888675, "grad_norm": 1.5043774096657498, "learning_rate": 1.5829635960038317e-05, "loss": 0.2644, "step": 5046 }, { "epoch": 3.2290467050543827, "grad_norm": 1.5568105185820733, "learning_rate": 1.5827952213358028e-05, "loss": 0.2679, "step": 5047 }, { "epoch": 3.2296865003198976, "grad_norm": 1.614723389112497, "learning_rate": 1.582626821643583e-05, "loss": 0.2407, "step": 5048 }, { "epoch": 3.230326295585413, "grad_norm": 1.591869125897646, "learning_rate": 1.5824583969344042e-05, "loss": 0.2585, "step": 5049 }, { "epoch": 3.2309660908509277, "grad_norm": 1.5455522536899429, "learning_rate": 1.582289947215498e-05, "loss": 0.2425, "step": 5050 }, { "epoch": 3.2316058861164425, "grad_norm": 1.5692364898647293, "learning_rate": 1.5821214724940973e-05, "loss": 0.2547, "step": 5051 }, { "epoch": 3.232245681381958, "grad_norm": 1.471761042282956, "learning_rate": 1.581952972777436e-05, "loss": 0.2425, "step": 5052 }, { "epoch": 3.2328854766474726, "grad_norm": 1.5989699419962937, "learning_rate": 1.581784448072749e-05, "loss": 0.2772, "step": 5053 }, { "epoch": 3.233525271912988, "grad_norm": 1.3780529909682735, "learning_rate": 1.581615898387273e-05, "loss": 0.2306, "step": 5054 }, { "epoch": 3.2341650671785027, "grad_norm": 1.4176002755014823, "learning_rate": 1.5814473237282452e-05, "loss": 0.2453, "step": 5055 }, { "epoch": 3.234804862444018, "grad_norm": 1.6196952769936777, "learning_rate": 1.581278724102903e-05, "loss": 0.2807, "step": 5056 }, { "epoch": 3.235444657709533, "grad_norm": 1.576090429824896, "learning_rate": 1.5811100995184866e-05, "loss": 0.2644, "step": 5057 }, { "epoch": 3.236084452975048, "grad_norm": 1.4506746676316398, "learning_rate": 1.5809414499822365e-05, "loss": 0.2417, "step": 5058 }, { "epoch": 3.236724248240563, "grad_norm": 1.4120002243854055, "learning_rate": 1.580772775501394e-05, "loss": 0.2297, "step": 5059 }, { "epoch": 3.237364043506078, "grad_norm": 1.4238399106900472, "learning_rate": 1.5806040760832014e-05, "loss": 0.2323, "step": 5060 }, { "epoch": 3.238003838771593, "grad_norm": 2.159735420262152, "learning_rate": 1.5804353517349024e-05, "loss": 0.2721, "step": 5061 }, { "epoch": 3.2386436340371083, "grad_norm": 1.6881526348819698, "learning_rate": 1.5802666024637424e-05, "loss": 0.2586, "step": 5062 }, { "epoch": 3.239283429302623, "grad_norm": 1.484569371227152, "learning_rate": 1.5800978282769667e-05, "loss": 0.2322, "step": 5063 }, { "epoch": 3.2399232245681384, "grad_norm": 1.4819787973681564, "learning_rate": 1.5799290291818218e-05, "loss": 0.255, "step": 5064 }, { "epoch": 3.2405630198336532, "grad_norm": 1.6006401656676053, "learning_rate": 1.5797602051855562e-05, "loss": 0.2392, "step": 5065 }, { "epoch": 3.241202815099168, "grad_norm": 1.6493717296615777, "learning_rate": 1.5795913562954185e-05, "loss": 0.2764, "step": 5066 }, { "epoch": 3.2418426103646834, "grad_norm": 1.466642814553788, "learning_rate": 1.579422482518659e-05, "loss": 0.2533, "step": 5067 }, { "epoch": 3.242482405630198, "grad_norm": 1.4799843695601076, "learning_rate": 1.579253583862529e-05, "loss": 0.2517, "step": 5068 }, { "epoch": 3.2431222008957135, "grad_norm": 1.4957822581098092, "learning_rate": 1.5790846603342805e-05, "loss": 0.271, "step": 5069 }, { "epoch": 3.2437619961612283, "grad_norm": 1.6100196819607626, "learning_rate": 1.5789157119411666e-05, "loss": 0.2667, "step": 5070 }, { "epoch": 3.2444017914267436, "grad_norm": 1.462501572952313, "learning_rate": 1.578746738690442e-05, "loss": 0.23, "step": 5071 }, { "epoch": 3.2450415866922584, "grad_norm": 1.5924741074195357, "learning_rate": 1.5785777405893622e-05, "loss": 0.2693, "step": 5072 }, { "epoch": 3.2456813819577737, "grad_norm": 1.67734376262359, "learning_rate": 1.578408717645183e-05, "loss": 0.2772, "step": 5073 }, { "epoch": 3.2463211772232885, "grad_norm": 1.5895845853278963, "learning_rate": 1.5782396698651625e-05, "loss": 0.281, "step": 5074 }, { "epoch": 3.2469609724888038, "grad_norm": 1.549400603988712, "learning_rate": 1.578070597256559e-05, "loss": 0.2399, "step": 5075 }, { "epoch": 3.2476007677543186, "grad_norm": 1.532388340866263, "learning_rate": 1.5779014998266327e-05, "loss": 0.2504, "step": 5076 }, { "epoch": 3.2482405630198334, "grad_norm": 1.5147549834645684, "learning_rate": 1.577732377582644e-05, "loss": 0.2716, "step": 5077 }, { "epoch": 3.2488803582853487, "grad_norm": 1.6070274098423445, "learning_rate": 1.5775632305318548e-05, "loss": 0.2731, "step": 5078 }, { "epoch": 3.2495201535508635, "grad_norm": 1.5957839916430934, "learning_rate": 1.5773940586815275e-05, "loss": 0.2639, "step": 5079 }, { "epoch": 3.250159948816379, "grad_norm": 1.5325547929108425, "learning_rate": 1.5772248620389264e-05, "loss": 0.2369, "step": 5080 }, { "epoch": 3.2507997440818936, "grad_norm": 1.6239648563644373, "learning_rate": 1.577055640611317e-05, "loss": 0.2778, "step": 5081 }, { "epoch": 3.251439539347409, "grad_norm": 1.5525366681850155, "learning_rate": 1.5768863944059646e-05, "loss": 0.2489, "step": 5082 }, { "epoch": 3.2520793346129238, "grad_norm": 1.5162530041550613, "learning_rate": 1.5767171234301363e-05, "loss": 0.2603, "step": 5083 }, { "epoch": 3.252719129878439, "grad_norm": 1.3984640212176849, "learning_rate": 1.576547827691101e-05, "loss": 0.2443, "step": 5084 }, { "epoch": 3.253358925143954, "grad_norm": 1.6254789044146989, "learning_rate": 1.576378507196128e-05, "loss": 0.2937, "step": 5085 }, { "epoch": 3.253998720409469, "grad_norm": 1.6064090725374023, "learning_rate": 1.5762091619524866e-05, "loss": 0.2594, "step": 5086 }, { "epoch": 3.254638515674984, "grad_norm": 1.5959113426254232, "learning_rate": 1.5760397919674487e-05, "loss": 0.2535, "step": 5087 }, { "epoch": 3.255278310940499, "grad_norm": 1.5626095718438553, "learning_rate": 1.5758703972482874e-05, "loss": 0.2266, "step": 5088 }, { "epoch": 3.255918106206014, "grad_norm": 1.5421705466582496, "learning_rate": 1.575700977802275e-05, "loss": 0.2313, "step": 5089 }, { "epoch": 3.2565579014715293, "grad_norm": 1.5965655916944597, "learning_rate": 1.5755315336366872e-05, "loss": 0.2681, "step": 5090 }, { "epoch": 3.257197696737044, "grad_norm": 1.6720673875775693, "learning_rate": 1.575362064758799e-05, "loss": 0.2732, "step": 5091 }, { "epoch": 3.257837492002559, "grad_norm": 1.6765037770310531, "learning_rate": 1.575192571175887e-05, "loss": 0.2815, "step": 5092 }, { "epoch": 3.2584772872680743, "grad_norm": 1.5548902503319084, "learning_rate": 1.5750230528952295e-05, "loss": 0.2578, "step": 5093 }, { "epoch": 3.259117082533589, "grad_norm": 1.5012852157519496, "learning_rate": 1.574853509924105e-05, "loss": 0.2499, "step": 5094 }, { "epoch": 3.2597568777991044, "grad_norm": 1.6157937199151284, "learning_rate": 1.574683942269793e-05, "loss": 0.2595, "step": 5095 }, { "epoch": 3.260396673064619, "grad_norm": 1.5914313250566914, "learning_rate": 1.574514349939575e-05, "loss": 0.2687, "step": 5096 }, { "epoch": 3.2610364683301345, "grad_norm": 1.584366444632868, "learning_rate": 1.574344732940733e-05, "loss": 0.2715, "step": 5097 }, { "epoch": 3.2616762635956493, "grad_norm": 1.5703933356099193, "learning_rate": 1.5741750912805495e-05, "loss": 0.2907, "step": 5098 }, { "epoch": 3.2623160588611646, "grad_norm": 1.487038304476786, "learning_rate": 1.574005424966309e-05, "loss": 0.25, "step": 5099 }, { "epoch": 3.2629558541266794, "grad_norm": 1.5173055382012983, "learning_rate": 1.573835734005297e-05, "loss": 0.2837, "step": 5100 }, { "epoch": 3.2635956493921947, "grad_norm": 1.5248232360590226, "learning_rate": 1.5736660184047993e-05, "loss": 0.2698, "step": 5101 }, { "epoch": 3.2642354446577095, "grad_norm": 1.591976592558838, "learning_rate": 1.573496278172103e-05, "loss": 0.2944, "step": 5102 }, { "epoch": 3.2648752399232244, "grad_norm": 1.5791559651373763, "learning_rate": 1.5733265133144965e-05, "loss": 0.2644, "step": 5103 }, { "epoch": 3.2655150351887396, "grad_norm": 1.574761036205279, "learning_rate": 1.5731567238392697e-05, "loss": 0.266, "step": 5104 }, { "epoch": 3.2661548304542545, "grad_norm": 1.4425029101851718, "learning_rate": 1.5729869097537126e-05, "loss": 0.2479, "step": 5105 }, { "epoch": 3.2667946257197698, "grad_norm": 1.4546527694927942, "learning_rate": 1.572817071065117e-05, "loss": 0.244, "step": 5106 }, { "epoch": 3.2674344209852846, "grad_norm": 1.5869131101738734, "learning_rate": 1.572647207780775e-05, "loss": 0.275, "step": 5107 }, { "epoch": 3.2680742162508, "grad_norm": 1.5703839140219962, "learning_rate": 1.572477319907981e-05, "loss": 0.2682, "step": 5108 }, { "epoch": 3.2687140115163147, "grad_norm": 1.412863362598668, "learning_rate": 1.5723074074540293e-05, "loss": 0.2123, "step": 5109 }, { "epoch": 3.26935380678183, "grad_norm": 1.5046127753438676, "learning_rate": 1.572137470426215e-05, "loss": 0.2462, "step": 5110 }, { "epoch": 3.269993602047345, "grad_norm": 1.5412229956573666, "learning_rate": 1.5719675088318356e-05, "loss": 0.2449, "step": 5111 }, { "epoch": 3.27063339731286, "grad_norm": 1.4823916503499048, "learning_rate": 1.5717975226781893e-05, "loss": 0.2522, "step": 5112 }, { "epoch": 3.271273192578375, "grad_norm": 1.5563332345194882, "learning_rate": 1.571627511972574e-05, "loss": 0.2814, "step": 5113 }, { "epoch": 3.2719129878438897, "grad_norm": 1.5367560277086056, "learning_rate": 1.5714574767222907e-05, "loss": 0.2598, "step": 5114 }, { "epoch": 3.272552783109405, "grad_norm": 1.5734798428839463, "learning_rate": 1.5712874169346398e-05, "loss": 0.2544, "step": 5115 }, { "epoch": 3.27319257837492, "grad_norm": 1.4929792472060504, "learning_rate": 1.5711173326169233e-05, "loss": 0.2397, "step": 5116 }, { "epoch": 3.273832373640435, "grad_norm": 1.4853957174399794, "learning_rate": 1.5709472237764445e-05, "loss": 0.2456, "step": 5117 }, { "epoch": 3.27447216890595, "grad_norm": 1.4476672616537818, "learning_rate": 1.5707770904205073e-05, "loss": 0.2365, "step": 5118 }, { "epoch": 3.275111964171465, "grad_norm": 1.4629062330610416, "learning_rate": 1.5706069325564175e-05, "loss": 0.2447, "step": 5119 }, { "epoch": 3.27575175943698, "grad_norm": 1.4709167457977426, "learning_rate": 1.570436750191481e-05, "loss": 0.2418, "step": 5120 }, { "epoch": 3.2763915547024953, "grad_norm": 1.5259009713157161, "learning_rate": 1.570266543333005e-05, "loss": 0.2338, "step": 5121 }, { "epoch": 3.27703134996801, "grad_norm": 1.4258671400408076, "learning_rate": 1.5700963119882982e-05, "loss": 0.2227, "step": 5122 }, { "epoch": 3.2776711452335254, "grad_norm": 1.6585301745623136, "learning_rate": 1.5699260561646698e-05, "loss": 0.2357, "step": 5123 }, { "epoch": 3.2783109404990403, "grad_norm": 1.4922838397733005, "learning_rate": 1.5697557758694305e-05, "loss": 0.2226, "step": 5124 }, { "epoch": 3.2789507357645555, "grad_norm": 1.4298711202904766, "learning_rate": 1.5695854711098915e-05, "loss": 0.2241, "step": 5125 }, { "epoch": 3.2795905310300704, "grad_norm": 1.5383387312743408, "learning_rate": 1.569415141893366e-05, "loss": 0.2548, "step": 5126 }, { "epoch": 3.2802303262955856, "grad_norm": 1.6405462379540372, "learning_rate": 1.5692447882271667e-05, "loss": 0.2749, "step": 5127 }, { "epoch": 3.2808701215611005, "grad_norm": 1.635508218464127, "learning_rate": 1.569074410118609e-05, "loss": 0.2423, "step": 5128 }, { "epoch": 3.2815099168266153, "grad_norm": 1.6889713919377551, "learning_rate": 1.5689040075750086e-05, "loss": 0.2582, "step": 5129 }, { "epoch": 3.2821497120921306, "grad_norm": 1.7098354124978483, "learning_rate": 1.5687335806036817e-05, "loss": 0.2787, "step": 5130 }, { "epoch": 3.2827895073576454, "grad_norm": 1.5183619085789244, "learning_rate": 1.568563129211947e-05, "loss": 0.2447, "step": 5131 }, { "epoch": 3.2834293026231607, "grad_norm": 1.597800463704485, "learning_rate": 1.5683926534071226e-05, "loss": 0.2582, "step": 5132 }, { "epoch": 3.2840690978886755, "grad_norm": 1.4801401600176798, "learning_rate": 1.5682221531965288e-05, "loss": 0.2385, "step": 5133 }, { "epoch": 3.284708893154191, "grad_norm": 1.753156131994834, "learning_rate": 1.568051628587487e-05, "loss": 0.2909, "step": 5134 }, { "epoch": 3.2853486884197056, "grad_norm": 1.5559410500863577, "learning_rate": 1.567881079587318e-05, "loss": 0.2703, "step": 5135 }, { "epoch": 3.285988483685221, "grad_norm": 1.6435811218859422, "learning_rate": 1.5677105062033464e-05, "loss": 0.2341, "step": 5136 }, { "epoch": 3.2866282789507357, "grad_norm": 1.4560250929223193, "learning_rate": 1.5675399084428947e-05, "loss": 0.2361, "step": 5137 }, { "epoch": 3.287268074216251, "grad_norm": 1.6415231360497495, "learning_rate": 1.5673692863132896e-05, "loss": 0.2556, "step": 5138 }, { "epoch": 3.287907869481766, "grad_norm": 1.5961678079788992, "learning_rate": 1.5671986398218564e-05, "loss": 0.256, "step": 5139 }, { "epoch": 3.2885476647472807, "grad_norm": 1.6203601767125577, "learning_rate": 1.5670279689759228e-05, "loss": 0.2917, "step": 5140 }, { "epoch": 3.289187460012796, "grad_norm": 1.520317775514317, "learning_rate": 1.5668572737828165e-05, "loss": 0.2396, "step": 5141 }, { "epoch": 3.2898272552783108, "grad_norm": 1.6063581043750534, "learning_rate": 1.566686554249867e-05, "loss": 0.2668, "step": 5142 }, { "epoch": 3.290467050543826, "grad_norm": 1.5072482013035793, "learning_rate": 1.566515810384406e-05, "loss": 0.2602, "step": 5143 }, { "epoch": 3.291106845809341, "grad_norm": 1.5496833462944615, "learning_rate": 1.5663450421937628e-05, "loss": 0.2351, "step": 5144 }, { "epoch": 3.291746641074856, "grad_norm": 1.6003408736888636, "learning_rate": 1.566174249685271e-05, "loss": 0.2445, "step": 5145 }, { "epoch": 3.292386436340371, "grad_norm": 1.56052197008398, "learning_rate": 1.5660034328662647e-05, "loss": 0.2532, "step": 5146 }, { "epoch": 3.2930262316058863, "grad_norm": 1.5808677106109759, "learning_rate": 1.5658325917440773e-05, "loss": 0.2618, "step": 5147 }, { "epoch": 3.293666026871401, "grad_norm": 1.530557396024165, "learning_rate": 1.565661726326045e-05, "loss": 0.2509, "step": 5148 }, { "epoch": 3.2943058221369164, "grad_norm": 1.5673466269039644, "learning_rate": 1.5654908366195044e-05, "loss": 0.2671, "step": 5149 }, { "epoch": 3.294945617402431, "grad_norm": 1.520408842806667, "learning_rate": 1.5653199226317935e-05, "loss": 0.2361, "step": 5150 }, { "epoch": 3.295585412667946, "grad_norm": 1.4932094361456376, "learning_rate": 1.5651489843702506e-05, "loss": 0.2485, "step": 5151 }, { "epoch": 3.2962252079334613, "grad_norm": 1.4798728930033507, "learning_rate": 1.5649780218422156e-05, "loss": 0.2646, "step": 5152 }, { "epoch": 3.2968650031989766, "grad_norm": 1.5187006579240279, "learning_rate": 1.5648070350550292e-05, "loss": 0.2517, "step": 5153 }, { "epoch": 3.2975047984644914, "grad_norm": 1.486884342171351, "learning_rate": 1.564636024016034e-05, "loss": 0.2704, "step": 5154 }, { "epoch": 3.2981445937300062, "grad_norm": 1.502350238412886, "learning_rate": 1.5644649887325715e-05, "loss": 0.2218, "step": 5155 }, { "epoch": 3.2987843889955215, "grad_norm": 1.5256124043617336, "learning_rate": 1.5642939292119873e-05, "loss": 0.2342, "step": 5156 }, { "epoch": 3.2994241842610363, "grad_norm": 1.6247054097884952, "learning_rate": 1.5641228454616252e-05, "loss": 0.247, "step": 5157 }, { "epoch": 3.3000639795265516, "grad_norm": 1.6617581738168625, "learning_rate": 1.5639517374888317e-05, "loss": 0.2757, "step": 5158 }, { "epoch": 3.3007037747920664, "grad_norm": 1.5774389011649272, "learning_rate": 1.5637806053009535e-05, "loss": 0.2626, "step": 5159 }, { "epoch": 3.3013435700575817, "grad_norm": 1.6012939715715224, "learning_rate": 1.563609448905339e-05, "loss": 0.2596, "step": 5160 }, { "epoch": 3.3019833653230966, "grad_norm": 1.5451752810203583, "learning_rate": 1.5634382683093377e-05, "loss": 0.2298, "step": 5161 }, { "epoch": 3.302623160588612, "grad_norm": 1.5767116676516653, "learning_rate": 1.5632670635202992e-05, "loss": 0.2271, "step": 5162 }, { "epoch": 3.3032629558541267, "grad_norm": 1.6587049316576923, "learning_rate": 1.5630958345455752e-05, "loss": 0.3156, "step": 5163 }, { "epoch": 3.303902751119642, "grad_norm": 1.5627751663673983, "learning_rate": 1.5629245813925176e-05, "loss": 0.2726, "step": 5164 }, { "epoch": 3.3045425463851568, "grad_norm": 1.336085851878503, "learning_rate": 1.5627533040684798e-05, "loss": 0.2339, "step": 5165 }, { "epoch": 3.3051823416506716, "grad_norm": 1.477767248113972, "learning_rate": 1.562582002580816e-05, "loss": 0.274, "step": 5166 }, { "epoch": 3.305822136916187, "grad_norm": 1.664661242717586, "learning_rate": 1.562410676936882e-05, "loss": 0.2586, "step": 5167 }, { "epoch": 3.3064619321817017, "grad_norm": 1.5350957832219703, "learning_rate": 1.562239327144034e-05, "loss": 0.2362, "step": 5168 }, { "epoch": 3.307101727447217, "grad_norm": 1.5418326065387786, "learning_rate": 1.5620679532096296e-05, "loss": 0.2679, "step": 5169 }, { "epoch": 3.307741522712732, "grad_norm": 1.6830959381638706, "learning_rate": 1.5618965551410265e-05, "loss": 0.2851, "step": 5170 }, { "epoch": 3.308381317978247, "grad_norm": 1.5393862686621806, "learning_rate": 1.5617251329455857e-05, "loss": 0.2669, "step": 5171 }, { "epoch": 3.309021113243762, "grad_norm": 1.5372144800216347, "learning_rate": 1.5615536866306662e-05, "loss": 0.2498, "step": 5172 }, { "epoch": 3.309660908509277, "grad_norm": 1.5879269211668026, "learning_rate": 1.5613822162036308e-05, "loss": 0.2657, "step": 5173 }, { "epoch": 3.310300703774792, "grad_norm": 1.602656654157091, "learning_rate": 1.5612107216718416e-05, "loss": 0.2635, "step": 5174 }, { "epoch": 3.3109404990403073, "grad_norm": 1.6508568871460225, "learning_rate": 1.561039203042662e-05, "loss": 0.3015, "step": 5175 }, { "epoch": 3.311580294305822, "grad_norm": 1.4828194461826356, "learning_rate": 1.5608676603234574e-05, "loss": 0.2334, "step": 5176 }, { "epoch": 3.312220089571337, "grad_norm": 1.5321055231039173, "learning_rate": 1.560696093521593e-05, "loss": 0.257, "step": 5177 }, { "epoch": 3.3128598848368522, "grad_norm": 1.456856554013956, "learning_rate": 1.560524502644436e-05, "loss": 0.2492, "step": 5178 }, { "epoch": 3.313499680102367, "grad_norm": 1.7294580204006142, "learning_rate": 1.5603528876993533e-05, "loss": 0.287, "step": 5179 }, { "epoch": 3.3141394753678823, "grad_norm": 1.4471605544905222, "learning_rate": 1.560181248693715e-05, "loss": 0.2477, "step": 5180 }, { "epoch": 3.314779270633397, "grad_norm": 1.4328531887951892, "learning_rate": 1.56000958563489e-05, "loss": 0.2492, "step": 5181 }, { "epoch": 3.3154190658989124, "grad_norm": 1.538133402169623, "learning_rate": 1.5598378985302497e-05, "loss": 0.2443, "step": 5182 }, { "epoch": 3.3160588611644273, "grad_norm": 1.408360421735929, "learning_rate": 1.559666187387166e-05, "loss": 0.2283, "step": 5183 }, { "epoch": 3.3166986564299425, "grad_norm": 1.573083310979267, "learning_rate": 1.559494452213012e-05, "loss": 0.246, "step": 5184 }, { "epoch": 3.3173384516954574, "grad_norm": 1.5316698925839742, "learning_rate": 1.559322693015161e-05, "loss": 0.266, "step": 5185 }, { "epoch": 3.3179782469609727, "grad_norm": 1.5728824315426198, "learning_rate": 1.5591509098009887e-05, "loss": 0.2745, "step": 5186 }, { "epoch": 3.3186180422264875, "grad_norm": 1.6365699731287557, "learning_rate": 1.5589791025778712e-05, "loss": 0.2575, "step": 5187 }, { "epoch": 3.3192578374920023, "grad_norm": 1.4106118840741408, "learning_rate": 1.558807271353185e-05, "loss": 0.2312, "step": 5188 }, { "epoch": 3.3198976327575176, "grad_norm": 1.5194824332152146, "learning_rate": 1.558635416134309e-05, "loss": 0.2414, "step": 5189 }, { "epoch": 3.320537428023033, "grad_norm": 1.593830359096206, "learning_rate": 1.5584635369286218e-05, "loss": 0.2517, "step": 5190 }, { "epoch": 3.3211772232885477, "grad_norm": 1.599235987044249, "learning_rate": 1.5582916337435036e-05, "loss": 0.2669, "step": 5191 }, { "epoch": 3.3218170185540625, "grad_norm": 1.5775124820388347, "learning_rate": 1.5581197065863362e-05, "loss": 0.2546, "step": 5192 }, { "epoch": 3.322456813819578, "grad_norm": 1.5520418088255001, "learning_rate": 1.557947755464501e-05, "loss": 0.2636, "step": 5193 }, { "epoch": 3.3230966090850926, "grad_norm": 1.5751477923491932, "learning_rate": 1.5577757803853817e-05, "loss": 0.2671, "step": 5194 }, { "epoch": 3.323736404350608, "grad_norm": 1.5570168594526508, "learning_rate": 1.5576037813563625e-05, "loss": 0.2624, "step": 5195 }, { "epoch": 3.3243761996161227, "grad_norm": 1.4369991695084103, "learning_rate": 1.5574317583848295e-05, "loss": 0.2414, "step": 5196 }, { "epoch": 3.325015994881638, "grad_norm": 1.4296852805889828, "learning_rate": 1.5572597114781678e-05, "loss": 0.2344, "step": 5197 }, { "epoch": 3.325655790147153, "grad_norm": 1.6379689767688879, "learning_rate": 1.5570876406437655e-05, "loss": 0.2593, "step": 5198 }, { "epoch": 3.326295585412668, "grad_norm": 1.4155751793125917, "learning_rate": 1.556915545889011e-05, "loss": 0.221, "step": 5199 }, { "epoch": 3.326935380678183, "grad_norm": 1.5599420798852468, "learning_rate": 1.5567434272212932e-05, "loss": 0.2507, "step": 5200 }, { "epoch": 3.3275751759436982, "grad_norm": 1.5992373449942163, "learning_rate": 1.5565712846480037e-05, "loss": 0.249, "step": 5201 }, { "epoch": 3.328214971209213, "grad_norm": 1.5390989402682256, "learning_rate": 1.556399118176533e-05, "loss": 0.2623, "step": 5202 }, { "epoch": 3.328854766474728, "grad_norm": 1.5933013405026535, "learning_rate": 1.556226927814274e-05, "loss": 0.2624, "step": 5203 }, { "epoch": 3.329494561740243, "grad_norm": 1.6224903751864932, "learning_rate": 1.55605471356862e-05, "loss": 0.2517, "step": 5204 }, { "epoch": 3.330134357005758, "grad_norm": 1.6154056465999855, "learning_rate": 1.5558824754469654e-05, "loss": 0.2712, "step": 5205 }, { "epoch": 3.3307741522712733, "grad_norm": 1.4326613285472813, "learning_rate": 1.5557102134567067e-05, "loss": 0.2234, "step": 5206 }, { "epoch": 3.331413947536788, "grad_norm": 1.5523007200910246, "learning_rate": 1.5555379276052397e-05, "loss": 0.2912, "step": 5207 }, { "epoch": 3.3320537428023034, "grad_norm": 1.5150392272698716, "learning_rate": 1.5553656178999627e-05, "loss": 0.2486, "step": 5208 }, { "epoch": 3.332693538067818, "grad_norm": 1.4341814172054357, "learning_rate": 1.5551932843482734e-05, "loss": 0.2606, "step": 5209 }, { "epoch": 3.3333333333333335, "grad_norm": 1.5635433520502489, "learning_rate": 1.5550209269575723e-05, "loss": 0.2476, "step": 5210 }, { "epoch": 3.3339731285988483, "grad_norm": 1.4361032932272184, "learning_rate": 1.55484854573526e-05, "loss": 0.2428, "step": 5211 }, { "epoch": 3.3346129238643636, "grad_norm": 1.5978271843706888, "learning_rate": 1.5546761406887382e-05, "loss": 0.2999, "step": 5212 }, { "epoch": 3.3352527191298784, "grad_norm": 1.4256828504209038, "learning_rate": 1.55450371182541e-05, "loss": 0.2606, "step": 5213 }, { "epoch": 3.3358925143953932, "grad_norm": 1.5133299680340662, "learning_rate": 1.554331259152678e-05, "loss": 0.2681, "step": 5214 }, { "epoch": 3.3365323096609085, "grad_norm": 1.4871197266229763, "learning_rate": 1.554158782677948e-05, "loss": 0.2467, "step": 5215 }, { "epoch": 3.3371721049264234, "grad_norm": 1.624624127068688, "learning_rate": 1.5539862824086262e-05, "loss": 0.2717, "step": 5216 }, { "epoch": 3.3378119001919386, "grad_norm": 1.4738095947227599, "learning_rate": 1.553813758352118e-05, "loss": 0.2413, "step": 5217 }, { "epoch": 3.3384516954574535, "grad_norm": 1.55422536124204, "learning_rate": 1.553641210515833e-05, "loss": 0.2621, "step": 5218 }, { "epoch": 3.3390914907229687, "grad_norm": 1.5260945050553822, "learning_rate": 1.5534686389071788e-05, "loss": 0.2192, "step": 5219 }, { "epoch": 3.3397312859884836, "grad_norm": 1.5557185274763605, "learning_rate": 1.553296043533566e-05, "loss": 0.2408, "step": 5220 }, { "epoch": 3.340371081253999, "grad_norm": 1.4130460192380503, "learning_rate": 1.553123424402405e-05, "loss": 0.2161, "step": 5221 }, { "epoch": 3.3410108765195137, "grad_norm": 1.561582119978973, "learning_rate": 1.5529507815211085e-05, "loss": 0.2382, "step": 5222 }, { "epoch": 3.341650671785029, "grad_norm": 1.5017600336073933, "learning_rate": 1.5527781148970888e-05, "loss": 0.2536, "step": 5223 }, { "epoch": 3.342290467050544, "grad_norm": 1.4924280123052087, "learning_rate": 1.5526054245377606e-05, "loss": 0.2617, "step": 5224 }, { "epoch": 3.342930262316059, "grad_norm": 1.4730732196033929, "learning_rate": 1.552432710450538e-05, "loss": 0.2482, "step": 5225 }, { "epoch": 3.343570057581574, "grad_norm": 1.4338342904797998, "learning_rate": 1.5522599726428375e-05, "loss": 0.2506, "step": 5226 }, { "epoch": 3.344209852847089, "grad_norm": 1.5446903017137057, "learning_rate": 1.5520872111220765e-05, "loss": 0.2605, "step": 5227 }, { "epoch": 3.344849648112604, "grad_norm": 1.5425704941992422, "learning_rate": 1.5519144258956728e-05, "loss": 0.288, "step": 5228 }, { "epoch": 3.345489443378119, "grad_norm": 1.5287579172601091, "learning_rate": 1.551741616971045e-05, "loss": 0.2448, "step": 5229 }, { "epoch": 3.346129238643634, "grad_norm": 1.59819823999383, "learning_rate": 1.5515687843556137e-05, "loss": 0.2485, "step": 5230 }, { "epoch": 3.346769033909149, "grad_norm": 1.5233517582685656, "learning_rate": 1.5513959280568004e-05, "loss": 0.2474, "step": 5231 }, { "epoch": 3.347408829174664, "grad_norm": 1.5511585001992734, "learning_rate": 1.5512230480820264e-05, "loss": 0.2414, "step": 5232 }, { "epoch": 3.348048624440179, "grad_norm": 1.5699378348845894, "learning_rate": 1.5510501444387155e-05, "loss": 0.2721, "step": 5233 }, { "epoch": 3.3486884197056943, "grad_norm": 1.4929454433108962, "learning_rate": 1.5508772171342913e-05, "loss": 0.2295, "step": 5234 }, { "epoch": 3.349328214971209, "grad_norm": 1.6105238026453095, "learning_rate": 1.5507042661761794e-05, "loss": 0.2663, "step": 5235 }, { "epoch": 3.3499680102367244, "grad_norm": 1.5482788651562043, "learning_rate": 1.550531291571806e-05, "loss": 0.2346, "step": 5236 }, { "epoch": 3.3506078055022392, "grad_norm": 1.550411441632083, "learning_rate": 1.5503582933285986e-05, "loss": 0.2456, "step": 5237 }, { "epoch": 3.3512476007677545, "grad_norm": 1.5532102437828164, "learning_rate": 1.5501852714539847e-05, "loss": 0.2549, "step": 5238 }, { "epoch": 3.3518873960332694, "grad_norm": 1.362692770518543, "learning_rate": 1.5500122259553942e-05, "loss": 0.2497, "step": 5239 }, { "epoch": 3.352527191298784, "grad_norm": 1.2536037469563712, "learning_rate": 1.549839156840257e-05, "loss": 0.1914, "step": 5240 }, { "epoch": 3.3531669865642995, "grad_norm": 1.5812419169745016, "learning_rate": 1.5496660641160045e-05, "loss": 0.2756, "step": 5241 }, { "epoch": 3.3538067818298143, "grad_norm": 1.564098334307195, "learning_rate": 1.549492947790069e-05, "loss": 0.2576, "step": 5242 }, { "epoch": 3.3544465770953296, "grad_norm": 1.5949350360534262, "learning_rate": 1.5493198078698838e-05, "loss": 0.2746, "step": 5243 }, { "epoch": 3.3550863723608444, "grad_norm": 1.4895774664369514, "learning_rate": 1.549146644362883e-05, "loss": 0.2386, "step": 5244 }, { "epoch": 3.3557261676263597, "grad_norm": 1.5914947439074874, "learning_rate": 1.5489734572765025e-05, "loss": 0.2618, "step": 5245 }, { "epoch": 3.3563659628918745, "grad_norm": 1.5598825918219308, "learning_rate": 1.548800246618178e-05, "loss": 0.2653, "step": 5246 }, { "epoch": 3.3570057581573898, "grad_norm": 1.5989284722120534, "learning_rate": 1.5486270123953472e-05, "loss": 0.2597, "step": 5247 }, { "epoch": 3.3576455534229046, "grad_norm": 1.6041309029462325, "learning_rate": 1.5484537546154486e-05, "loss": 0.2536, "step": 5248 }, { "epoch": 3.35828534868842, "grad_norm": 1.5062818312965724, "learning_rate": 1.5482804732859212e-05, "loss": 0.2417, "step": 5249 }, { "epoch": 3.3589251439539347, "grad_norm": 1.5782804772219727, "learning_rate": 1.5481071684142058e-05, "loss": 0.255, "step": 5250 }, { "epoch": 3.3595649392194495, "grad_norm": 1.4833866168393506, "learning_rate": 1.5479338400077435e-05, "loss": 0.2455, "step": 5251 }, { "epoch": 3.360204734484965, "grad_norm": 1.4340508534173695, "learning_rate": 1.5477604880739766e-05, "loss": 0.2547, "step": 5252 }, { "epoch": 3.36084452975048, "grad_norm": 1.6630525939867726, "learning_rate": 1.5475871126203494e-05, "loss": 0.2975, "step": 5253 }, { "epoch": 3.361484325015995, "grad_norm": 1.5725215532129508, "learning_rate": 1.547413713654305e-05, "loss": 0.2658, "step": 5254 }, { "epoch": 3.3621241202815098, "grad_norm": 1.5552917091493674, "learning_rate": 1.54724029118329e-05, "loss": 0.2639, "step": 5255 }, { "epoch": 3.362763915547025, "grad_norm": 1.499822090894676, "learning_rate": 1.54706684521475e-05, "loss": 0.2593, "step": 5256 }, { "epoch": 3.36340371081254, "grad_norm": 1.4875932800947573, "learning_rate": 1.5468933757561325e-05, "loss": 0.2585, "step": 5257 }, { "epoch": 3.364043506078055, "grad_norm": 1.484569761179585, "learning_rate": 1.5467198828148867e-05, "loss": 0.2254, "step": 5258 }, { "epoch": 3.36468330134357, "grad_norm": 1.6115028552857782, "learning_rate": 1.5465463663984616e-05, "loss": 0.2677, "step": 5259 }, { "epoch": 3.3653230966090852, "grad_norm": 1.4251929275627908, "learning_rate": 1.546372826514308e-05, "loss": 0.23, "step": 5260 }, { "epoch": 3.3659628918746, "grad_norm": 1.5566661118560414, "learning_rate": 1.5461992631698768e-05, "loss": 0.2625, "step": 5261 }, { "epoch": 3.3666026871401153, "grad_norm": 1.4678029311060594, "learning_rate": 1.546025676372621e-05, "loss": 0.2641, "step": 5262 }, { "epoch": 3.36724248240563, "grad_norm": 1.5859334374303673, "learning_rate": 1.545852066129994e-05, "loss": 0.2715, "step": 5263 }, { "epoch": 3.3678822776711455, "grad_norm": 1.4894921763645348, "learning_rate": 1.5456784324494504e-05, "loss": 0.259, "step": 5264 }, { "epoch": 3.3685220729366603, "grad_norm": 1.715120425330633, "learning_rate": 1.5455047753384455e-05, "loss": 0.305, "step": 5265 }, { "epoch": 3.369161868202175, "grad_norm": 1.4482846001946705, "learning_rate": 1.5453310948044358e-05, "loss": 0.2643, "step": 5266 }, { "epoch": 3.3698016634676904, "grad_norm": 1.408573183843712, "learning_rate": 1.545157390854879e-05, "loss": 0.2288, "step": 5267 }, { "epoch": 3.370441458733205, "grad_norm": 1.5329077969853573, "learning_rate": 1.5449836634972334e-05, "loss": 0.2709, "step": 5268 }, { "epoch": 3.3710812539987205, "grad_norm": 1.628239448367897, "learning_rate": 1.544809912738959e-05, "loss": 0.2702, "step": 5269 }, { "epoch": 3.3717210492642353, "grad_norm": 1.4818953626480749, "learning_rate": 1.5446361385875165e-05, "loss": 0.2635, "step": 5270 }, { "epoch": 3.3723608445297506, "grad_norm": 1.593753822298113, "learning_rate": 1.5444623410503663e-05, "loss": 0.2749, "step": 5271 }, { "epoch": 3.3730006397952654, "grad_norm": 1.4933765179537795, "learning_rate": 1.544288520134972e-05, "loss": 0.2322, "step": 5272 }, { "epoch": 3.3736404350607807, "grad_norm": 1.5476154436001663, "learning_rate": 1.544114675848797e-05, "loss": 0.2507, "step": 5273 }, { "epoch": 3.3742802303262955, "grad_norm": 1.5862058764233415, "learning_rate": 1.5439408081993055e-05, "loss": 0.2383, "step": 5274 }, { "epoch": 3.374920025591811, "grad_norm": 1.548577587963204, "learning_rate": 1.5437669171939633e-05, "loss": 0.2522, "step": 5275 }, { "epoch": 3.3755598208573256, "grad_norm": 1.6682147044162408, "learning_rate": 1.543593002840237e-05, "loss": 0.2663, "step": 5276 }, { "epoch": 3.3761996161228405, "grad_norm": 1.5364173060801514, "learning_rate": 1.543419065145594e-05, "loss": 0.2485, "step": 5277 }, { "epoch": 3.3768394113883557, "grad_norm": 1.4146550325222575, "learning_rate": 1.5432451041175035e-05, "loss": 0.2189, "step": 5278 }, { "epoch": 3.3774792066538706, "grad_norm": 1.7190084350514125, "learning_rate": 1.5430711197634343e-05, "loss": 0.3052, "step": 5279 }, { "epoch": 3.378119001919386, "grad_norm": 1.496284891387144, "learning_rate": 1.542897112090857e-05, "loss": 0.2474, "step": 5280 }, { "epoch": 3.3787587971849007, "grad_norm": 1.6107990085497428, "learning_rate": 1.5427230811072437e-05, "loss": 0.263, "step": 5281 }, { "epoch": 3.379398592450416, "grad_norm": 1.5839815915712032, "learning_rate": 1.5425490268200666e-05, "loss": 0.2743, "step": 5282 }, { "epoch": 3.380038387715931, "grad_norm": 1.653434666974168, "learning_rate": 1.5423749492367997e-05, "loss": 0.2656, "step": 5283 }, { "epoch": 3.380678182981446, "grad_norm": 1.4733315587209435, "learning_rate": 1.542200848364917e-05, "loss": 0.259, "step": 5284 }, { "epoch": 3.381317978246961, "grad_norm": 1.601887842448808, "learning_rate": 1.5420267242118945e-05, "loss": 0.2713, "step": 5285 }, { "epoch": 3.381957773512476, "grad_norm": 1.5866957039578966, "learning_rate": 1.5418525767852087e-05, "loss": 0.2777, "step": 5286 }, { "epoch": 3.382597568777991, "grad_norm": 1.5209316420584185, "learning_rate": 1.541678406092337e-05, "loss": 0.2621, "step": 5287 }, { "epoch": 3.383237364043506, "grad_norm": 1.5351369706131273, "learning_rate": 1.5415042121407583e-05, "loss": 0.2339, "step": 5288 }, { "epoch": 3.383877159309021, "grad_norm": 1.4304698276672299, "learning_rate": 1.541329994937952e-05, "loss": 0.241, "step": 5289 }, { "epoch": 3.3845169545745364, "grad_norm": 1.4904504679346233, "learning_rate": 1.5411557544913983e-05, "loss": 0.223, "step": 5290 }, { "epoch": 3.385156749840051, "grad_norm": 1.559514929177012, "learning_rate": 1.5409814908085792e-05, "loss": 0.2671, "step": 5291 }, { "epoch": 3.385796545105566, "grad_norm": 1.4329381672679138, "learning_rate": 1.5408072038969776e-05, "loss": 0.2535, "step": 5292 }, { "epoch": 3.3864363403710813, "grad_norm": 1.486232448190133, "learning_rate": 1.5406328937640768e-05, "loss": 0.2522, "step": 5293 }, { "epoch": 3.387076135636596, "grad_norm": 1.4934735141983613, "learning_rate": 1.5404585604173606e-05, "loss": 0.2582, "step": 5294 }, { "epoch": 3.3877159309021114, "grad_norm": 1.526076064259839, "learning_rate": 1.5402842038643157e-05, "loss": 0.2528, "step": 5295 }, { "epoch": 3.3883557261676263, "grad_norm": 1.5613020199558885, "learning_rate": 1.5401098241124278e-05, "loss": 0.2916, "step": 5296 }, { "epoch": 3.3889955214331415, "grad_norm": 1.5205605102665645, "learning_rate": 1.5399354211691854e-05, "loss": 0.2177, "step": 5297 }, { "epoch": 3.3896353166986564, "grad_norm": 1.5176741627945358, "learning_rate": 1.539760995042076e-05, "loss": 0.2781, "step": 5298 }, { "epoch": 3.3902751119641716, "grad_norm": 1.5230852439725224, "learning_rate": 1.5395865457385896e-05, "loss": 0.2703, "step": 5299 }, { "epoch": 3.3909149072296865, "grad_norm": 1.529559372674395, "learning_rate": 1.539412073266217e-05, "loss": 0.2525, "step": 5300 }, { "epoch": 3.3915547024952017, "grad_norm": 1.561205937429571, "learning_rate": 1.5392375776324494e-05, "loss": 0.2524, "step": 5301 }, { "epoch": 3.3921944977607166, "grad_norm": 1.6075872852459698, "learning_rate": 1.53906305884478e-05, "loss": 0.2765, "step": 5302 }, { "epoch": 3.3928342930262314, "grad_norm": 1.5854118402292567, "learning_rate": 1.538888516910701e-05, "loss": 0.2497, "step": 5303 }, { "epoch": 3.3934740882917467, "grad_norm": 1.55736942223867, "learning_rate": 1.538713951837708e-05, "loss": 0.2443, "step": 5304 }, { "epoch": 3.3941138835572615, "grad_norm": 1.4441779065476124, "learning_rate": 1.538539363633296e-05, "loss": 0.2561, "step": 5305 }, { "epoch": 3.394753678822777, "grad_norm": 1.627676416382759, "learning_rate": 1.538364752304962e-05, "loss": 0.2548, "step": 5306 }, { "epoch": 3.3953934740882916, "grad_norm": 1.4727726423426593, "learning_rate": 1.5381901178602032e-05, "loss": 0.2631, "step": 5307 }, { "epoch": 3.396033269353807, "grad_norm": 1.591165265756008, "learning_rate": 1.5380154603065178e-05, "loss": 0.2469, "step": 5308 }, { "epoch": 3.3966730646193217, "grad_norm": 1.579373944849724, "learning_rate": 1.537840779651406e-05, "loss": 0.2349, "step": 5309 }, { "epoch": 3.397312859884837, "grad_norm": 1.4170245644460209, "learning_rate": 1.5376660759023678e-05, "loss": 0.2541, "step": 5310 }, { "epoch": 3.397952655150352, "grad_norm": 1.469072298894303, "learning_rate": 1.5374913490669045e-05, "loss": 0.2329, "step": 5311 }, { "epoch": 3.398592450415867, "grad_norm": 1.536009942070805, "learning_rate": 1.537316599152519e-05, "loss": 0.2339, "step": 5312 }, { "epoch": 3.399232245681382, "grad_norm": 1.474875321953713, "learning_rate": 1.5371418261667145e-05, "loss": 0.2366, "step": 5313 }, { "epoch": 3.3998720409468968, "grad_norm": 1.4311551593413994, "learning_rate": 1.536967030116995e-05, "loss": 0.2553, "step": 5314 }, { "epoch": 3.400511836212412, "grad_norm": 1.524268517999237, "learning_rate": 1.536792211010867e-05, "loss": 0.2497, "step": 5315 }, { "epoch": 3.401151631477927, "grad_norm": 1.5590492340558393, "learning_rate": 1.5366173688558363e-05, "loss": 0.2689, "step": 5316 }, { "epoch": 3.401791426743442, "grad_norm": 1.5837313822529004, "learning_rate": 1.5364425036594103e-05, "loss": 0.2799, "step": 5317 }, { "epoch": 3.402431222008957, "grad_norm": 1.373367094111966, "learning_rate": 1.5362676154290974e-05, "loss": 0.2253, "step": 5318 }, { "epoch": 3.4030710172744723, "grad_norm": 1.5783395633208284, "learning_rate": 1.5360927041724068e-05, "loss": 0.2644, "step": 5319 }, { "epoch": 3.403710812539987, "grad_norm": 1.4164023967450494, "learning_rate": 1.5359177698968495e-05, "loss": 0.2288, "step": 5320 }, { "epoch": 3.4043506078055024, "grad_norm": 1.5446049805504922, "learning_rate": 1.5357428126099363e-05, "loss": 0.2495, "step": 5321 }, { "epoch": 3.404990403071017, "grad_norm": 1.400718484020213, "learning_rate": 1.5355678323191797e-05, "loss": 0.2371, "step": 5322 }, { "epoch": 3.4056301983365325, "grad_norm": 1.6368173410614812, "learning_rate": 1.535392829032093e-05, "loss": 0.2585, "step": 5323 }, { "epoch": 3.4062699936020473, "grad_norm": 1.5483650952207155, "learning_rate": 1.5352178027561908e-05, "loss": 0.2412, "step": 5324 }, { "epoch": 3.4069097888675626, "grad_norm": 1.622134354524735, "learning_rate": 1.5350427534989884e-05, "loss": 0.2501, "step": 5325 }, { "epoch": 3.4075495841330774, "grad_norm": 1.5884372731251772, "learning_rate": 1.5348676812680017e-05, "loss": 0.2595, "step": 5326 }, { "epoch": 3.4081893793985927, "grad_norm": 1.4606838030656972, "learning_rate": 1.5346925860707483e-05, "loss": 0.2543, "step": 5327 }, { "epoch": 3.4088291746641075, "grad_norm": 1.5311146596100251, "learning_rate": 1.5345174679147462e-05, "loss": 0.2628, "step": 5328 }, { "epoch": 3.4094689699296223, "grad_norm": 1.7330066500994075, "learning_rate": 1.534342326807515e-05, "loss": 0.2632, "step": 5329 }, { "epoch": 3.4101087651951376, "grad_norm": 1.5236532771343205, "learning_rate": 1.5341671627565748e-05, "loss": 0.2528, "step": 5330 }, { "epoch": 3.4107485604606524, "grad_norm": 1.5684185381109623, "learning_rate": 1.533991975769447e-05, "loss": 0.2689, "step": 5331 }, { "epoch": 3.4113883557261677, "grad_norm": 1.4344928473791678, "learning_rate": 1.5338167658536537e-05, "loss": 0.2498, "step": 5332 }, { "epoch": 3.4120281509916826, "grad_norm": 1.4487042290786487, "learning_rate": 1.5336415330167177e-05, "loss": 0.2335, "step": 5333 }, { "epoch": 3.412667946257198, "grad_norm": 1.3993379099403367, "learning_rate": 1.5334662772661635e-05, "loss": 0.2517, "step": 5334 }, { "epoch": 3.4133077415227127, "grad_norm": 1.584439010571244, "learning_rate": 1.5332909986095166e-05, "loss": 0.2654, "step": 5335 }, { "epoch": 3.413947536788228, "grad_norm": 1.6000132091739072, "learning_rate": 1.533115697054303e-05, "loss": 0.2741, "step": 5336 }, { "epoch": 3.4145873320537428, "grad_norm": 1.4195367894224669, "learning_rate": 1.5329403726080496e-05, "loss": 0.2426, "step": 5337 }, { "epoch": 3.415227127319258, "grad_norm": 1.3701335129953207, "learning_rate": 1.5327650252782843e-05, "loss": 0.2416, "step": 5338 }, { "epoch": 3.415866922584773, "grad_norm": 1.6195638535998238, "learning_rate": 1.5325896550725366e-05, "loss": 0.2297, "step": 5339 }, { "epoch": 3.4165067178502877, "grad_norm": 1.5353329243917253, "learning_rate": 1.532414261998337e-05, "loss": 0.278, "step": 5340 }, { "epoch": 3.417146513115803, "grad_norm": 1.4739189635440992, "learning_rate": 1.5322388460632154e-05, "loss": 0.2568, "step": 5341 }, { "epoch": 3.417786308381318, "grad_norm": 1.7044412714106516, "learning_rate": 1.5320634072747052e-05, "loss": 0.2868, "step": 5342 }, { "epoch": 3.418426103646833, "grad_norm": 1.5678307266668872, "learning_rate": 1.5318879456403384e-05, "loss": 0.2633, "step": 5343 }, { "epoch": 3.419065898912348, "grad_norm": 1.5477990232724803, "learning_rate": 1.5317124611676493e-05, "loss": 0.3022, "step": 5344 }, { "epoch": 3.419705694177863, "grad_norm": 1.45272567371546, "learning_rate": 1.531536953864173e-05, "loss": 0.2446, "step": 5345 }, { "epoch": 3.420345489443378, "grad_norm": 1.4454935350690945, "learning_rate": 1.5313614237374458e-05, "loss": 0.2135, "step": 5346 }, { "epoch": 3.4209852847088933, "grad_norm": 1.5338039221397504, "learning_rate": 1.531185870795004e-05, "loss": 0.2631, "step": 5347 }, { "epoch": 3.421625079974408, "grad_norm": 1.546412464256527, "learning_rate": 1.5310102950443856e-05, "loss": 0.2594, "step": 5348 }, { "epoch": 3.4222648752399234, "grad_norm": 1.6724150575182068, "learning_rate": 1.53083469649313e-05, "loss": 0.3029, "step": 5349 }, { "epoch": 3.4229046705054382, "grad_norm": 1.5059726428607147, "learning_rate": 1.5306590751487766e-05, "loss": 0.2424, "step": 5350 }, { "epoch": 3.423544465770953, "grad_norm": 1.6222356693194713, "learning_rate": 1.530483431018867e-05, "loss": 0.2426, "step": 5351 }, { "epoch": 3.4241842610364683, "grad_norm": 1.3896651960766015, "learning_rate": 1.530307764110942e-05, "loss": 0.2444, "step": 5352 }, { "epoch": 3.4248240563019836, "grad_norm": 1.6553907297445154, "learning_rate": 1.530132074432545e-05, "loss": 0.279, "step": 5353 }, { "epoch": 3.4254638515674984, "grad_norm": 1.4160416971286214, "learning_rate": 1.52995636199122e-05, "loss": 0.2287, "step": 5354 }, { "epoch": 3.4261036468330133, "grad_norm": 1.3681529690049614, "learning_rate": 1.5297806267945115e-05, "loss": 0.2441, "step": 5355 }, { "epoch": 3.4267434420985285, "grad_norm": 1.361418048473922, "learning_rate": 1.529604868849965e-05, "loss": 0.2148, "step": 5356 }, { "epoch": 3.4273832373640434, "grad_norm": 1.5960601165239658, "learning_rate": 1.5294290881651278e-05, "loss": 0.2548, "step": 5357 }, { "epoch": 3.4280230326295587, "grad_norm": 1.676904632449928, "learning_rate": 1.5292532847475472e-05, "loss": 0.2727, "step": 5358 }, { "epoch": 3.4286628278950735, "grad_norm": 1.6194976234562115, "learning_rate": 1.5290774586047722e-05, "loss": 0.2823, "step": 5359 }, { "epoch": 3.4293026231605888, "grad_norm": 1.5188233531362596, "learning_rate": 1.528901609744352e-05, "loss": 0.2412, "step": 5360 }, { "epoch": 3.4299424184261036, "grad_norm": 1.581282881040312, "learning_rate": 1.5287257381738377e-05, "loss": 0.2582, "step": 5361 }, { "epoch": 3.430582213691619, "grad_norm": 1.415234629472986, "learning_rate": 1.5285498439007807e-05, "loss": 0.2304, "step": 5362 }, { "epoch": 3.4312220089571337, "grad_norm": 1.5564603018972407, "learning_rate": 1.5283739269327337e-05, "loss": 0.2575, "step": 5363 }, { "epoch": 3.431861804222649, "grad_norm": 1.532161172372355, "learning_rate": 1.5281979872772503e-05, "loss": 0.2532, "step": 5364 }, { "epoch": 3.432501599488164, "grad_norm": 1.6599096882843205, "learning_rate": 1.5280220249418848e-05, "loss": 0.2963, "step": 5365 }, { "epoch": 3.4331413947536786, "grad_norm": 1.5471912032369803, "learning_rate": 1.5278460399341923e-05, "loss": 0.2635, "step": 5366 }, { "epoch": 3.433781190019194, "grad_norm": 1.5207453322817048, "learning_rate": 1.5276700322617305e-05, "loss": 0.2177, "step": 5367 }, { "epoch": 3.4344209852847087, "grad_norm": 1.5239452138119334, "learning_rate": 1.527494001932056e-05, "loss": 0.2397, "step": 5368 }, { "epoch": 3.435060780550224, "grad_norm": 1.509944225172137, "learning_rate": 1.5273179489527273e-05, "loss": 0.2494, "step": 5369 }, { "epoch": 3.435700575815739, "grad_norm": 1.4302624859088855, "learning_rate": 1.527141873331304e-05, "loss": 0.2602, "step": 5370 }, { "epoch": 3.436340371081254, "grad_norm": 1.5897792938492015, "learning_rate": 1.5269657750753463e-05, "loss": 0.2503, "step": 5371 }, { "epoch": 3.436980166346769, "grad_norm": 1.4888022027118861, "learning_rate": 1.5267896541924157e-05, "loss": 0.2543, "step": 5372 }, { "epoch": 3.4376199616122842, "grad_norm": 1.4844918205837174, "learning_rate": 1.526613510690074e-05, "loss": 0.2512, "step": 5373 }, { "epoch": 3.438259756877799, "grad_norm": 1.4563266691783605, "learning_rate": 1.5264373445758853e-05, "loss": 0.2366, "step": 5374 }, { "epoch": 3.4388995521433143, "grad_norm": 1.3890706853834296, "learning_rate": 1.5262611558574138e-05, "loss": 0.2448, "step": 5375 }, { "epoch": 3.439539347408829, "grad_norm": 1.5566074695716912, "learning_rate": 1.5260849445422242e-05, "loss": 0.2568, "step": 5376 }, { "epoch": 3.440179142674344, "grad_norm": 1.562283235333789, "learning_rate": 1.5259087106378826e-05, "loss": 0.2792, "step": 5377 }, { "epoch": 3.4408189379398593, "grad_norm": 1.5478311610401236, "learning_rate": 1.5257324541519567e-05, "loss": 0.2554, "step": 5378 }, { "epoch": 3.441458733205374, "grad_norm": 1.5946179345349816, "learning_rate": 1.5255561750920144e-05, "loss": 0.2832, "step": 5379 }, { "epoch": 3.4420985284708894, "grad_norm": 1.569828123400452, "learning_rate": 1.5253798734656248e-05, "loss": 0.2494, "step": 5380 }, { "epoch": 3.442738323736404, "grad_norm": 1.7148820652599337, "learning_rate": 1.525203549280358e-05, "loss": 0.2539, "step": 5381 }, { "epoch": 3.4433781190019195, "grad_norm": 1.6202415667160661, "learning_rate": 1.5250272025437852e-05, "loss": 0.2215, "step": 5382 }, { "epoch": 3.4440179142674343, "grad_norm": 1.4856221200609907, "learning_rate": 1.5248508332634782e-05, "loss": 0.2364, "step": 5383 }, { "epoch": 3.4446577095329496, "grad_norm": 1.458085349259238, "learning_rate": 1.5246744414470099e-05, "loss": 0.2307, "step": 5384 }, { "epoch": 3.4452975047984644, "grad_norm": 1.4848113026022753, "learning_rate": 1.5244980271019548e-05, "loss": 0.2215, "step": 5385 }, { "epoch": 3.4459373000639797, "grad_norm": 1.3792469494806119, "learning_rate": 1.524321590235887e-05, "loss": 0.2384, "step": 5386 }, { "epoch": 3.4465770953294945, "grad_norm": 1.4726471007751198, "learning_rate": 1.524145130856383e-05, "loss": 0.2476, "step": 5387 }, { "epoch": 3.4472168905950094, "grad_norm": 1.4970536073961553, "learning_rate": 1.5239686489710192e-05, "loss": 0.2579, "step": 5388 }, { "epoch": 3.4478566858605246, "grad_norm": 1.5811258389415896, "learning_rate": 1.5237921445873742e-05, "loss": 0.2726, "step": 5389 }, { "epoch": 3.44849648112604, "grad_norm": 1.5319599942690685, "learning_rate": 1.5236156177130259e-05, "loss": 0.2437, "step": 5390 }, { "epoch": 3.4491362763915547, "grad_norm": 1.4600428914952928, "learning_rate": 1.5234390683555546e-05, "loss": 0.2792, "step": 5391 }, { "epoch": 3.4497760716570696, "grad_norm": 1.5605280371092383, "learning_rate": 1.5232624965225408e-05, "loss": 0.2653, "step": 5392 }, { "epoch": 3.450415866922585, "grad_norm": 1.5728834144141146, "learning_rate": 1.5230859022215661e-05, "loss": 0.2738, "step": 5393 }, { "epoch": 3.4510556621880997, "grad_norm": 1.4565006591368181, "learning_rate": 1.5229092854602135e-05, "loss": 0.2332, "step": 5394 }, { "epoch": 3.451695457453615, "grad_norm": 1.49493521425874, "learning_rate": 1.522732646246066e-05, "loss": 0.2532, "step": 5395 }, { "epoch": 3.4523352527191298, "grad_norm": 1.445683545791891, "learning_rate": 1.5225559845867086e-05, "loss": 0.2303, "step": 5396 }, { "epoch": 3.452975047984645, "grad_norm": 1.6762558741271771, "learning_rate": 1.5223793004897269e-05, "loss": 0.2895, "step": 5397 }, { "epoch": 3.45361484325016, "grad_norm": 1.63156481449038, "learning_rate": 1.5222025939627071e-05, "loss": 0.2722, "step": 5398 }, { "epoch": 3.454254638515675, "grad_norm": 1.4731330056016474, "learning_rate": 1.5220258650132374e-05, "loss": 0.2458, "step": 5399 }, { "epoch": 3.45489443378119, "grad_norm": 1.5524018161667648, "learning_rate": 1.5218491136489052e-05, "loss": 0.2455, "step": 5400 }, { "epoch": 3.4555342290467053, "grad_norm": 1.6429228730277439, "learning_rate": 1.5216723398773004e-05, "loss": 0.2803, "step": 5401 }, { "epoch": 3.45617402431222, "grad_norm": 1.4148910447635077, "learning_rate": 1.5214955437060135e-05, "loss": 0.2214, "step": 5402 }, { "epoch": 3.456813819577735, "grad_norm": 1.5252098824091849, "learning_rate": 1.5213187251426353e-05, "loss": 0.2414, "step": 5403 }, { "epoch": 3.45745361484325, "grad_norm": 1.502308908675332, "learning_rate": 1.5211418841947586e-05, "loss": 0.2763, "step": 5404 }, { "epoch": 3.458093410108765, "grad_norm": 1.49175712248759, "learning_rate": 1.5209650208699763e-05, "loss": 0.2682, "step": 5405 }, { "epoch": 3.4587332053742803, "grad_norm": 1.6065581230595394, "learning_rate": 1.520788135175883e-05, "loss": 0.2825, "step": 5406 }, { "epoch": 3.459373000639795, "grad_norm": 1.5256922685644614, "learning_rate": 1.5206112271200732e-05, "loss": 0.245, "step": 5407 }, { "epoch": 3.4600127959053104, "grad_norm": 1.4507039210680377, "learning_rate": 1.5204342967101436e-05, "loss": 0.2321, "step": 5408 }, { "epoch": 3.4606525911708252, "grad_norm": 1.4668605417409921, "learning_rate": 1.5202573439536911e-05, "loss": 0.2403, "step": 5409 }, { "epoch": 3.4612923864363405, "grad_norm": 1.7152252809115844, "learning_rate": 1.5200803688583139e-05, "loss": 0.279, "step": 5410 }, { "epoch": 3.4619321817018553, "grad_norm": 1.5668799731971137, "learning_rate": 1.5199033714316103e-05, "loss": 0.2605, "step": 5411 }, { "epoch": 3.4625719769673706, "grad_norm": 1.6447935609204039, "learning_rate": 1.5197263516811811e-05, "loss": 0.261, "step": 5412 }, { "epoch": 3.4632117722328855, "grad_norm": 1.507549977949219, "learning_rate": 1.519549309614627e-05, "loss": 0.2395, "step": 5413 }, { "epoch": 3.4638515674984003, "grad_norm": 1.510927657653467, "learning_rate": 1.5193722452395496e-05, "loss": 0.2476, "step": 5414 }, { "epoch": 3.4644913627639156, "grad_norm": 1.609713451344526, "learning_rate": 1.519195158563552e-05, "loss": 0.288, "step": 5415 }, { "epoch": 3.4651311580294304, "grad_norm": 1.6139398822533095, "learning_rate": 1.5190180495942377e-05, "loss": 0.2785, "step": 5416 }, { "epoch": 3.4657709532949457, "grad_norm": 1.5636386574516656, "learning_rate": 1.518840918339212e-05, "loss": 0.2489, "step": 5417 }, { "epoch": 3.4664107485604605, "grad_norm": 1.5256580789195704, "learning_rate": 1.5186637648060796e-05, "loss": 0.223, "step": 5418 }, { "epoch": 3.4670505438259758, "grad_norm": 1.6484798151280236, "learning_rate": 1.518486589002448e-05, "loss": 0.2747, "step": 5419 }, { "epoch": 3.4676903390914906, "grad_norm": 1.5799960987683004, "learning_rate": 1.5183093909359246e-05, "loss": 0.2602, "step": 5420 }, { "epoch": 3.468330134357006, "grad_norm": 1.4839582627243955, "learning_rate": 1.518132170614118e-05, "loss": 0.2368, "step": 5421 }, { "epoch": 3.4689699296225207, "grad_norm": 1.4808047947785554, "learning_rate": 1.5179549280446378e-05, "loss": 0.2548, "step": 5422 }, { "epoch": 3.469609724888036, "grad_norm": 1.708476684077324, "learning_rate": 1.5177776632350942e-05, "loss": 0.2877, "step": 5423 }, { "epoch": 3.470249520153551, "grad_norm": 1.5599445341391687, "learning_rate": 1.5176003761930988e-05, "loss": 0.2729, "step": 5424 }, { "epoch": 3.470889315419066, "grad_norm": 1.5347696448527206, "learning_rate": 1.5174230669262643e-05, "loss": 0.2556, "step": 5425 }, { "epoch": 3.471529110684581, "grad_norm": 1.4307457427565389, "learning_rate": 1.5172457354422033e-05, "loss": 0.2587, "step": 5426 }, { "epoch": 3.472168905950096, "grad_norm": 1.4529909712596094, "learning_rate": 1.5170683817485306e-05, "loss": 0.2419, "step": 5427 }, { "epoch": 3.472808701215611, "grad_norm": 1.4453176873470088, "learning_rate": 1.5168910058528612e-05, "loss": 0.2365, "step": 5428 }, { "epoch": 3.473448496481126, "grad_norm": 1.5912259491072174, "learning_rate": 1.516713607762812e-05, "loss": 0.2324, "step": 5429 }, { "epoch": 3.474088291746641, "grad_norm": 1.6107387082974325, "learning_rate": 1.5165361874859996e-05, "loss": 0.247, "step": 5430 }, { "epoch": 3.474728087012156, "grad_norm": 1.5194685047468646, "learning_rate": 1.5163587450300417e-05, "loss": 0.2563, "step": 5431 }, { "epoch": 3.4753678822776712, "grad_norm": 1.463103020473843, "learning_rate": 1.5161812804025582e-05, "loss": 0.2654, "step": 5432 }, { "epoch": 3.476007677543186, "grad_norm": 1.472174815706956, "learning_rate": 1.5160037936111684e-05, "loss": 0.2373, "step": 5433 }, { "epoch": 3.4766474728087013, "grad_norm": 1.5210397297259814, "learning_rate": 1.515826284663494e-05, "loss": 0.254, "step": 5434 }, { "epoch": 3.477287268074216, "grad_norm": 1.553856710467131, "learning_rate": 1.5156487535671563e-05, "loss": 0.2573, "step": 5435 }, { "epoch": 3.4779270633397315, "grad_norm": 1.536158196372737, "learning_rate": 1.5154712003297784e-05, "loss": 0.2596, "step": 5436 }, { "epoch": 3.4785668586052463, "grad_norm": 1.524146573217488, "learning_rate": 1.5152936249589842e-05, "loss": 0.2481, "step": 5437 }, { "epoch": 3.4792066538707616, "grad_norm": 1.5045181324808825, "learning_rate": 1.5151160274623982e-05, "loss": 0.2461, "step": 5438 }, { "epoch": 3.4798464491362764, "grad_norm": 1.479529698937503, "learning_rate": 1.5149384078476463e-05, "loss": 0.2422, "step": 5439 }, { "epoch": 3.480486244401791, "grad_norm": 1.6257459747430814, "learning_rate": 1.5147607661223554e-05, "loss": 0.2486, "step": 5440 }, { "epoch": 3.4811260396673065, "grad_norm": 1.5526669667592727, "learning_rate": 1.5145831022941526e-05, "loss": 0.273, "step": 5441 }, { "epoch": 3.4817658349328213, "grad_norm": 1.6190453794784623, "learning_rate": 1.514405416370667e-05, "loss": 0.2546, "step": 5442 }, { "epoch": 3.4824056301983366, "grad_norm": 1.6631540002942653, "learning_rate": 1.5142277083595278e-05, "loss": 0.2701, "step": 5443 }, { "epoch": 3.4830454254638514, "grad_norm": 1.6175645597275061, "learning_rate": 1.5140499782683655e-05, "loss": 0.26, "step": 5444 }, { "epoch": 3.4836852207293667, "grad_norm": 1.5610074469535486, "learning_rate": 1.5138722261048117e-05, "loss": 0.2581, "step": 5445 }, { "epoch": 3.4843250159948815, "grad_norm": 1.5595827848997423, "learning_rate": 1.5136944518764987e-05, "loss": 0.2757, "step": 5446 }, { "epoch": 3.484964811260397, "grad_norm": 1.6268804225394125, "learning_rate": 1.5135166555910594e-05, "loss": 0.2476, "step": 5447 }, { "epoch": 3.4856046065259116, "grad_norm": 1.5618336030348057, "learning_rate": 1.5133388372561286e-05, "loss": 0.2307, "step": 5448 }, { "epoch": 3.486244401791427, "grad_norm": 1.583719194391384, "learning_rate": 1.5131609968793412e-05, "loss": 0.2539, "step": 5449 }, { "epoch": 3.4868841970569417, "grad_norm": 1.5579225609869638, "learning_rate": 1.5129831344683333e-05, "loss": 0.2555, "step": 5450 }, { "epoch": 3.4875239923224566, "grad_norm": 1.751663265847215, "learning_rate": 1.5128052500307426e-05, "loss": 0.2832, "step": 5451 }, { "epoch": 3.488163787587972, "grad_norm": 1.7525362643889828, "learning_rate": 1.5126273435742063e-05, "loss": 0.2929, "step": 5452 }, { "epoch": 3.488803582853487, "grad_norm": 1.5081672062401634, "learning_rate": 1.5124494151063638e-05, "loss": 0.2641, "step": 5453 }, { "epoch": 3.489443378119002, "grad_norm": 1.5169852037342766, "learning_rate": 1.5122714646348552e-05, "loss": 0.2616, "step": 5454 }, { "epoch": 3.490083173384517, "grad_norm": 1.700033733836432, "learning_rate": 1.5120934921673207e-05, "loss": 0.2407, "step": 5455 }, { "epoch": 3.490722968650032, "grad_norm": 1.4953658627482824, "learning_rate": 1.5119154977114028e-05, "loss": 0.2585, "step": 5456 }, { "epoch": 3.491362763915547, "grad_norm": 1.4242629002269747, "learning_rate": 1.5117374812747443e-05, "loss": 0.2402, "step": 5457 }, { "epoch": 3.492002559181062, "grad_norm": 1.413257180292085, "learning_rate": 1.5115594428649885e-05, "loss": 0.2475, "step": 5458 }, { "epoch": 3.492642354446577, "grad_norm": 1.3942725294682379, "learning_rate": 1.5113813824897801e-05, "loss": 0.2428, "step": 5459 }, { "epoch": 3.4932821497120923, "grad_norm": 1.5263490139982134, "learning_rate": 1.511203300156765e-05, "loss": 0.2277, "step": 5460 }, { "epoch": 3.493921944977607, "grad_norm": 1.6483145900225633, "learning_rate": 1.5110251958735896e-05, "loss": 0.2733, "step": 5461 }, { "epoch": 3.4945617402431224, "grad_norm": 1.4976800746268089, "learning_rate": 1.5108470696479011e-05, "loss": 0.258, "step": 5462 }, { "epoch": 3.495201535508637, "grad_norm": 1.5400959501696017, "learning_rate": 1.5106689214873482e-05, "loss": 0.2697, "step": 5463 }, { "epoch": 3.4958413307741525, "grad_norm": 1.6890132507072049, "learning_rate": 1.5104907513995804e-05, "loss": 0.2835, "step": 5464 }, { "epoch": 3.4964811260396673, "grad_norm": 1.5869429425213812, "learning_rate": 1.5103125593922477e-05, "loss": 0.2686, "step": 5465 }, { "epoch": 3.497120921305182, "grad_norm": 1.456170390150537, "learning_rate": 1.5101343454730015e-05, "loss": 0.2433, "step": 5466 }, { "epoch": 3.4977607165706974, "grad_norm": 1.5405309641330345, "learning_rate": 1.5099561096494939e-05, "loss": 0.2259, "step": 5467 }, { "epoch": 3.4984005118362123, "grad_norm": 1.6668572023294348, "learning_rate": 1.5097778519293782e-05, "loss": 0.2344, "step": 5468 }, { "epoch": 3.4990403071017275, "grad_norm": 1.6255275658415613, "learning_rate": 1.5095995723203084e-05, "loss": 0.2679, "step": 5469 }, { "epoch": 3.4996801023672424, "grad_norm": 1.5819189686648867, "learning_rate": 1.5094212708299392e-05, "loss": 0.2993, "step": 5470 }, { "epoch": 3.5003198976327576, "grad_norm": 1.5610957951773161, "learning_rate": 1.5092429474659272e-05, "loss": 0.2458, "step": 5471 }, { "epoch": 3.5009596928982725, "grad_norm": 1.5414113251970467, "learning_rate": 1.5090646022359286e-05, "loss": 0.2818, "step": 5472 }, { "epoch": 3.5015994881637877, "grad_norm": 1.5303358592428182, "learning_rate": 1.5088862351476015e-05, "loss": 0.2722, "step": 5473 }, { "epoch": 3.5022392834293026, "grad_norm": 1.4053447126513148, "learning_rate": 1.508707846208605e-05, "loss": 0.2484, "step": 5474 }, { "epoch": 3.502879078694818, "grad_norm": 1.523076983146441, "learning_rate": 1.5085294354265982e-05, "loss": 0.2615, "step": 5475 }, { "epoch": 3.5035188739603327, "grad_norm": 1.5644668338172818, "learning_rate": 1.5083510028092425e-05, "loss": 0.281, "step": 5476 }, { "epoch": 3.5041586692258475, "grad_norm": 1.5012992047961629, "learning_rate": 1.5081725483641988e-05, "loss": 0.2823, "step": 5477 }, { "epoch": 3.504798464491363, "grad_norm": 1.5749404525073232, "learning_rate": 1.5079940720991296e-05, "loss": 0.2524, "step": 5478 }, { "epoch": 3.505438259756878, "grad_norm": 1.457275367461925, "learning_rate": 1.5078155740216986e-05, "loss": 0.2461, "step": 5479 }, { "epoch": 3.506078055022393, "grad_norm": 1.4825045516854973, "learning_rate": 1.5076370541395704e-05, "loss": 0.2589, "step": 5480 }, { "epoch": 3.5067178502879077, "grad_norm": 1.6196713396316054, "learning_rate": 1.5074585124604102e-05, "loss": 0.275, "step": 5481 }, { "epoch": 3.507357645553423, "grad_norm": 1.545733815957944, "learning_rate": 1.5072799489918839e-05, "loss": 0.2194, "step": 5482 }, { "epoch": 3.507997440818938, "grad_norm": 1.4620825415795, "learning_rate": 1.5071013637416592e-05, "loss": 0.247, "step": 5483 }, { "epoch": 3.508637236084453, "grad_norm": 1.6190412472759972, "learning_rate": 1.506922756717404e-05, "loss": 0.2619, "step": 5484 }, { "epoch": 3.509277031349968, "grad_norm": 1.600972511573452, "learning_rate": 1.5067441279267872e-05, "loss": 0.2663, "step": 5485 }, { "epoch": 3.509916826615483, "grad_norm": 1.5269123898706944, "learning_rate": 1.5065654773774792e-05, "loss": 0.2876, "step": 5486 }, { "epoch": 3.510556621880998, "grad_norm": 1.5878028622997504, "learning_rate": 1.5063868050771505e-05, "loss": 0.2682, "step": 5487 }, { "epoch": 3.511196417146513, "grad_norm": 1.4072651382828583, "learning_rate": 1.506208111033473e-05, "loss": 0.2247, "step": 5488 }, { "epoch": 3.511836212412028, "grad_norm": 1.4861676733365627, "learning_rate": 1.50602939525412e-05, "loss": 0.2496, "step": 5489 }, { "epoch": 3.5124760076775434, "grad_norm": 1.4547922970117197, "learning_rate": 1.5058506577467649e-05, "loss": 0.2592, "step": 5490 }, { "epoch": 3.5131158029430583, "grad_norm": 1.5370011766098053, "learning_rate": 1.5056718985190826e-05, "loss": 0.2701, "step": 5491 }, { "epoch": 3.513755598208573, "grad_norm": 1.4563272084409407, "learning_rate": 1.5054931175787482e-05, "loss": 0.2434, "step": 5492 }, { "epoch": 3.5143953934740884, "grad_norm": 1.4252489673298383, "learning_rate": 1.5053143149334385e-05, "loss": 0.228, "step": 5493 }, { "epoch": 3.515035188739603, "grad_norm": 1.5909059823710974, "learning_rate": 1.505135490590831e-05, "loss": 0.2759, "step": 5494 }, { "epoch": 3.5156749840051185, "grad_norm": 1.5403724012391002, "learning_rate": 1.5049566445586042e-05, "loss": 0.2447, "step": 5495 }, { "epoch": 3.5163147792706333, "grad_norm": 1.5233222029291817, "learning_rate": 1.5047777768444374e-05, "loss": 0.2301, "step": 5496 }, { "epoch": 3.5169545745361486, "grad_norm": 1.5179010297939084, "learning_rate": 1.5045988874560107e-05, "loss": 0.2413, "step": 5497 }, { "epoch": 3.5175943698016634, "grad_norm": 1.6070217210097129, "learning_rate": 1.5044199764010054e-05, "loss": 0.2762, "step": 5498 }, { "epoch": 3.5182341650671782, "grad_norm": 1.5850648508935121, "learning_rate": 1.5042410436871038e-05, "loss": 0.25, "step": 5499 }, { "epoch": 3.5188739603326935, "grad_norm": 1.5753466149819446, "learning_rate": 1.5040620893219885e-05, "loss": 0.2374, "step": 5500 }, { "epoch": 3.519513755598209, "grad_norm": 1.5154167532213865, "learning_rate": 1.5038831133133437e-05, "loss": 0.2466, "step": 5501 }, { "epoch": 3.5201535508637236, "grad_norm": 1.4494179753981131, "learning_rate": 1.5037041156688543e-05, "loss": 0.2271, "step": 5502 }, { "epoch": 3.5207933461292384, "grad_norm": 1.5469628728973481, "learning_rate": 1.5035250963962064e-05, "loss": 0.2393, "step": 5503 }, { "epoch": 3.5214331413947537, "grad_norm": 1.4327132569503016, "learning_rate": 1.5033460555030866e-05, "loss": 0.2504, "step": 5504 }, { "epoch": 3.5220729366602685, "grad_norm": 1.4917045966761258, "learning_rate": 1.5031669929971826e-05, "loss": 0.2574, "step": 5505 }, { "epoch": 3.522712731925784, "grad_norm": 1.5560298931583472, "learning_rate": 1.502987908886183e-05, "loss": 0.2425, "step": 5506 }, { "epoch": 3.5233525271912987, "grad_norm": 1.5077714143986822, "learning_rate": 1.5028088031777767e-05, "loss": 0.2484, "step": 5507 }, { "epoch": 3.523992322456814, "grad_norm": 1.640316207980459, "learning_rate": 1.5026296758796554e-05, "loss": 0.2851, "step": 5508 }, { "epoch": 3.5246321177223288, "grad_norm": 1.6006441366195443, "learning_rate": 1.50245052699951e-05, "loss": 0.2516, "step": 5509 }, { "epoch": 3.525271912987844, "grad_norm": 1.5464682043197238, "learning_rate": 1.5022713565450324e-05, "loss": 0.2261, "step": 5510 }, { "epoch": 3.525911708253359, "grad_norm": 1.5504443456085832, "learning_rate": 1.5020921645239167e-05, "loss": 0.2753, "step": 5511 }, { "epoch": 3.526551503518874, "grad_norm": 1.5318334975775783, "learning_rate": 1.5019129509438562e-05, "loss": 0.2339, "step": 5512 }, { "epoch": 3.527191298784389, "grad_norm": 1.5170215969198162, "learning_rate": 1.5017337158125465e-05, "loss": 0.2572, "step": 5513 }, { "epoch": 3.527831094049904, "grad_norm": 1.482166235384483, "learning_rate": 1.5015544591376836e-05, "loss": 0.2215, "step": 5514 }, { "epoch": 3.528470889315419, "grad_norm": 1.4845933594676923, "learning_rate": 1.5013751809269646e-05, "loss": 0.2497, "step": 5515 }, { "epoch": 3.5291106845809344, "grad_norm": 1.5590619618521715, "learning_rate": 1.5011958811880871e-05, "loss": 0.2675, "step": 5516 }, { "epoch": 3.529750479846449, "grad_norm": 1.5032102102740068, "learning_rate": 1.5010165599287496e-05, "loss": 0.2537, "step": 5517 }, { "epoch": 3.530390275111964, "grad_norm": 1.5421761042809452, "learning_rate": 1.5008372171566525e-05, "loss": 0.2577, "step": 5518 }, { "epoch": 3.5310300703774793, "grad_norm": 1.366897588631478, "learning_rate": 1.5006578528794965e-05, "loss": 0.2425, "step": 5519 }, { "epoch": 3.531669865642994, "grad_norm": 1.531144778677596, "learning_rate": 1.500478467104983e-05, "loss": 0.2858, "step": 5520 }, { "epoch": 3.5323096609085094, "grad_norm": 1.5306342047523183, "learning_rate": 1.5002990598408143e-05, "loss": 0.2523, "step": 5521 }, { "epoch": 3.5329494561740242, "grad_norm": 1.6651790469824834, "learning_rate": 1.5001196310946935e-05, "loss": 0.2637, "step": 5522 }, { "epoch": 3.5335892514395395, "grad_norm": 1.3861279950027507, "learning_rate": 1.4999401808743255e-05, "loss": 0.2333, "step": 5523 }, { "epoch": 3.5342290467050543, "grad_norm": 1.4153765134143679, "learning_rate": 1.4997607091874158e-05, "loss": 0.2448, "step": 5524 }, { "epoch": 3.534868841970569, "grad_norm": 1.5888975888689398, "learning_rate": 1.4995812160416701e-05, "loss": 0.2651, "step": 5525 }, { "epoch": 3.5355086372360844, "grad_norm": 1.5862260551339347, "learning_rate": 1.4994017014447955e-05, "loss": 0.2604, "step": 5526 }, { "epoch": 3.5361484325015997, "grad_norm": 1.52804765033554, "learning_rate": 1.4992221654045006e-05, "loss": 0.2336, "step": 5527 }, { "epoch": 3.5367882277671145, "grad_norm": 1.5249890458747943, "learning_rate": 1.4990426079284934e-05, "loss": 0.2657, "step": 5528 }, { "epoch": 3.5374280230326294, "grad_norm": 1.5956878818991498, "learning_rate": 1.4988630290244848e-05, "loss": 0.2412, "step": 5529 }, { "epoch": 3.5380678182981447, "grad_norm": 1.3565267842714148, "learning_rate": 1.498683428700185e-05, "loss": 0.228, "step": 5530 }, { "epoch": 3.5387076135636595, "grad_norm": 1.5683371313145598, "learning_rate": 1.498503806963306e-05, "loss": 0.2404, "step": 5531 }, { "epoch": 3.5393474088291748, "grad_norm": 1.4551411241996801, "learning_rate": 1.49832416382156e-05, "loss": 0.2391, "step": 5532 }, { "epoch": 3.5399872040946896, "grad_norm": 1.451119795535858, "learning_rate": 1.4981444992826608e-05, "loss": 0.2345, "step": 5533 }, { "epoch": 3.540626999360205, "grad_norm": 1.6022735921487312, "learning_rate": 1.4979648133543232e-05, "loss": 0.2597, "step": 5534 }, { "epoch": 3.5412667946257197, "grad_norm": 1.425788012661562, "learning_rate": 1.4977851060442623e-05, "loss": 0.2525, "step": 5535 }, { "epoch": 3.541906589891235, "grad_norm": 1.5027337846444488, "learning_rate": 1.4976053773601947e-05, "loss": 0.2581, "step": 5536 }, { "epoch": 3.54254638515675, "grad_norm": 1.4700692475850206, "learning_rate": 1.4974256273098369e-05, "loss": 0.24, "step": 5537 }, { "epoch": 3.543186180422265, "grad_norm": 1.4540282973815686, "learning_rate": 1.4972458559009078e-05, "loss": 0.2708, "step": 5538 }, { "epoch": 3.54382597568778, "grad_norm": 1.4786163645802233, "learning_rate": 1.497066063141126e-05, "loss": 0.2569, "step": 5539 }, { "epoch": 3.5444657709532947, "grad_norm": 1.4328761466452893, "learning_rate": 1.4968862490382114e-05, "loss": 0.23, "step": 5540 }, { "epoch": 3.54510556621881, "grad_norm": 1.5488923882964456, "learning_rate": 1.4967064135998857e-05, "loss": 0.2516, "step": 5541 }, { "epoch": 3.5457453614843253, "grad_norm": 1.635146450761899, "learning_rate": 1.49652655683387e-05, "loss": 0.2768, "step": 5542 }, { "epoch": 3.54638515674984, "grad_norm": 1.4119412457645122, "learning_rate": 1.4963466787478872e-05, "loss": 0.248, "step": 5543 }, { "epoch": 3.547024952015355, "grad_norm": 1.5290688378691835, "learning_rate": 1.4961667793496608e-05, "loss": 0.2625, "step": 5544 }, { "epoch": 3.5476647472808702, "grad_norm": 1.4414931492899867, "learning_rate": 1.4959868586469154e-05, "loss": 0.2503, "step": 5545 }, { "epoch": 3.548304542546385, "grad_norm": 1.67048828080934, "learning_rate": 1.495806916647377e-05, "loss": 0.2802, "step": 5546 }, { "epoch": 3.5489443378119003, "grad_norm": 1.418198967093182, "learning_rate": 1.4956269533587714e-05, "loss": 0.2383, "step": 5547 }, { "epoch": 3.549584133077415, "grad_norm": 1.5133370627112914, "learning_rate": 1.495446968788826e-05, "loss": 0.2411, "step": 5548 }, { "epoch": 3.5502239283429304, "grad_norm": 1.5978893054271486, "learning_rate": 1.4952669629452689e-05, "loss": 0.2922, "step": 5549 }, { "epoch": 3.5508637236084453, "grad_norm": 1.4428231573769963, "learning_rate": 1.49508693583583e-05, "loss": 0.2455, "step": 5550 }, { "epoch": 3.55150351887396, "grad_norm": 1.5800354782190689, "learning_rate": 1.4949068874682384e-05, "loss": 0.2571, "step": 5551 }, { "epoch": 3.5521433141394754, "grad_norm": 1.5347830535503744, "learning_rate": 1.4947268178502254e-05, "loss": 0.2381, "step": 5552 }, { "epoch": 3.5527831094049906, "grad_norm": 1.5403165522260542, "learning_rate": 1.494546726989523e-05, "loss": 0.2805, "step": 5553 }, { "epoch": 3.5534229046705055, "grad_norm": 1.4493293065591053, "learning_rate": 1.4943666148938636e-05, "loss": 0.2496, "step": 5554 }, { "epoch": 3.5540626999360203, "grad_norm": 1.6024127950300664, "learning_rate": 1.4941864815709814e-05, "loss": 0.2356, "step": 5555 }, { "epoch": 3.5547024952015356, "grad_norm": 1.5280704351752166, "learning_rate": 1.4940063270286108e-05, "loss": 0.2429, "step": 5556 }, { "epoch": 3.5553422904670504, "grad_norm": 1.4992898989059786, "learning_rate": 1.4938261512744871e-05, "loss": 0.249, "step": 5557 }, { "epoch": 3.5559820857325657, "grad_norm": 1.639589530346848, "learning_rate": 1.4936459543163473e-05, "loss": 0.2756, "step": 5558 }, { "epoch": 3.5566218809980805, "grad_norm": 1.612672230664748, "learning_rate": 1.493465736161928e-05, "loss": 0.2822, "step": 5559 }, { "epoch": 3.557261676263596, "grad_norm": 1.5216726043740771, "learning_rate": 1.4932854968189679e-05, "loss": 0.2492, "step": 5560 }, { "epoch": 3.5579014715291106, "grad_norm": 1.6415141234983006, "learning_rate": 1.4931052362952061e-05, "loss": 0.2709, "step": 5561 }, { "epoch": 3.5585412667946255, "grad_norm": 1.746780328787221, "learning_rate": 1.4929249545983824e-05, "loss": 0.2975, "step": 5562 }, { "epoch": 3.5591810620601407, "grad_norm": 1.4457880821739844, "learning_rate": 1.4927446517362382e-05, "loss": 0.2542, "step": 5563 }, { "epoch": 3.559820857325656, "grad_norm": 1.436235872754688, "learning_rate": 1.4925643277165151e-05, "loss": 0.2341, "step": 5564 }, { "epoch": 3.560460652591171, "grad_norm": 1.5982299137290215, "learning_rate": 1.4923839825469561e-05, "loss": 0.2941, "step": 5565 }, { "epoch": 3.5611004478566857, "grad_norm": 1.5465697459719652, "learning_rate": 1.4922036162353047e-05, "loss": 0.248, "step": 5566 }, { "epoch": 3.561740243122201, "grad_norm": 1.460617205163161, "learning_rate": 1.4920232287893055e-05, "loss": 0.2323, "step": 5567 }, { "epoch": 3.5623800383877158, "grad_norm": 1.5291793701193508, "learning_rate": 1.4918428202167041e-05, "loss": 0.2438, "step": 5568 }, { "epoch": 3.563019833653231, "grad_norm": 1.5956042577524157, "learning_rate": 1.491662390525247e-05, "loss": 0.2998, "step": 5569 }, { "epoch": 3.563659628918746, "grad_norm": 1.477113624329683, "learning_rate": 1.4914819397226814e-05, "loss": 0.2545, "step": 5570 }, { "epoch": 3.564299424184261, "grad_norm": 1.5177244963835301, "learning_rate": 1.4913014678167557e-05, "loss": 0.2754, "step": 5571 }, { "epoch": 3.564939219449776, "grad_norm": 1.554092225396191, "learning_rate": 1.491120974815219e-05, "loss": 0.2589, "step": 5572 }, { "epoch": 3.5655790147152913, "grad_norm": 1.6196041765768572, "learning_rate": 1.490940460725821e-05, "loss": 0.3052, "step": 5573 }, { "epoch": 3.566218809980806, "grad_norm": 1.598385408945133, "learning_rate": 1.490759925556313e-05, "loss": 0.2586, "step": 5574 }, { "epoch": 3.5668586052463214, "grad_norm": 1.387696937664546, "learning_rate": 1.4905793693144469e-05, "loss": 0.2318, "step": 5575 }, { "epoch": 3.567498400511836, "grad_norm": 1.6592877171868388, "learning_rate": 1.4903987920079752e-05, "loss": 0.2483, "step": 5576 }, { "epoch": 3.568138195777351, "grad_norm": 1.5153904110219427, "learning_rate": 1.4902181936446518e-05, "loss": 0.2621, "step": 5577 }, { "epoch": 3.5687779910428663, "grad_norm": 1.349477706627683, "learning_rate": 1.4900375742322313e-05, "loss": 0.2368, "step": 5578 }, { "epoch": 3.5694177863083816, "grad_norm": 1.4621108310521778, "learning_rate": 1.489856933778469e-05, "loss": 0.2331, "step": 5579 }, { "epoch": 3.5700575815738964, "grad_norm": 1.4946282462975504, "learning_rate": 1.4896762722911213e-05, "loss": 0.2548, "step": 5580 }, { "epoch": 3.5706973768394112, "grad_norm": 1.6230913461381653, "learning_rate": 1.4894955897779458e-05, "loss": 0.2517, "step": 5581 }, { "epoch": 3.5713371721049265, "grad_norm": 1.4672053881116094, "learning_rate": 1.4893148862467e-05, "loss": 0.2547, "step": 5582 }, { "epoch": 3.5719769673704413, "grad_norm": 1.4756678651936246, "learning_rate": 1.4891341617051435e-05, "loss": 0.2626, "step": 5583 }, { "epoch": 3.5726167626359566, "grad_norm": 1.5253050532451295, "learning_rate": 1.4889534161610364e-05, "loss": 0.2455, "step": 5584 }, { "epoch": 3.5732565579014715, "grad_norm": 1.52594790961158, "learning_rate": 1.4887726496221393e-05, "loss": 0.2339, "step": 5585 }, { "epoch": 3.5738963531669867, "grad_norm": 1.6128365072091002, "learning_rate": 1.488591862096214e-05, "loss": 0.2796, "step": 5586 }, { "epoch": 3.5745361484325016, "grad_norm": 1.583420618756359, "learning_rate": 1.4884110535910235e-05, "loss": 0.2567, "step": 5587 }, { "epoch": 3.5751759436980164, "grad_norm": 1.7595643988879093, "learning_rate": 1.4882302241143311e-05, "loss": 0.2567, "step": 5588 }, { "epoch": 3.5758157389635317, "grad_norm": 1.4893980025102977, "learning_rate": 1.4880493736739012e-05, "loss": 0.2474, "step": 5589 }, { "epoch": 3.576455534229047, "grad_norm": 1.4910058795600771, "learning_rate": 1.4878685022774994e-05, "loss": 0.2399, "step": 5590 }, { "epoch": 3.5770953294945618, "grad_norm": 1.29850981641767, "learning_rate": 1.487687609932892e-05, "loss": 0.215, "step": 5591 }, { "epoch": 3.5777351247600766, "grad_norm": 1.3829108740038443, "learning_rate": 1.4875066966478462e-05, "loss": 0.2297, "step": 5592 }, { "epoch": 3.578374920025592, "grad_norm": 1.4180782264549157, "learning_rate": 1.48732576243013e-05, "loss": 0.2618, "step": 5593 }, { "epoch": 3.5790147152911067, "grad_norm": 1.469634313342905, "learning_rate": 1.4871448072875126e-05, "loss": 0.2674, "step": 5594 }, { "epoch": 3.579654510556622, "grad_norm": 1.3612375620862496, "learning_rate": 1.4869638312277637e-05, "loss": 0.2347, "step": 5595 }, { "epoch": 3.580294305822137, "grad_norm": 1.4653670885721581, "learning_rate": 1.4867828342586541e-05, "loss": 0.2621, "step": 5596 }, { "epoch": 3.580934101087652, "grad_norm": 1.6261228422717318, "learning_rate": 1.4866018163879554e-05, "loss": 0.2727, "step": 5597 }, { "epoch": 3.581573896353167, "grad_norm": 1.6753459045713157, "learning_rate": 1.4864207776234405e-05, "loss": 0.2772, "step": 5598 }, { "epoch": 3.5822136916186818, "grad_norm": 1.5776771970981747, "learning_rate": 1.4862397179728825e-05, "loss": 0.2543, "step": 5599 }, { "epoch": 3.582853486884197, "grad_norm": 1.5748399146807133, "learning_rate": 1.4860586374440558e-05, "loss": 0.2724, "step": 5600 }, { "epoch": 3.5834932821497123, "grad_norm": 1.465501347642079, "learning_rate": 1.4858775360447362e-05, "loss": 0.2485, "step": 5601 }, { "epoch": 3.584133077415227, "grad_norm": 1.5961212992525422, "learning_rate": 1.4856964137826994e-05, "loss": 0.2494, "step": 5602 }, { "epoch": 3.584772872680742, "grad_norm": 1.4131929602626883, "learning_rate": 1.4855152706657228e-05, "loss": 0.216, "step": 5603 }, { "epoch": 3.5854126679462572, "grad_norm": 1.4470160873879596, "learning_rate": 1.4853341067015837e-05, "loss": 0.2294, "step": 5604 }, { "epoch": 3.586052463211772, "grad_norm": 1.6312954509053899, "learning_rate": 1.4851529218980613e-05, "loss": 0.2687, "step": 5605 }, { "epoch": 3.5866922584772873, "grad_norm": 1.5609677705756224, "learning_rate": 1.484971716262936e-05, "loss": 0.2656, "step": 5606 }, { "epoch": 3.587332053742802, "grad_norm": 1.5136583243215602, "learning_rate": 1.4847904898039875e-05, "loss": 0.2708, "step": 5607 }, { "epoch": 3.5879718490083174, "grad_norm": 1.425699969933614, "learning_rate": 1.4846092425289983e-05, "loss": 0.2586, "step": 5608 }, { "epoch": 3.5886116442738323, "grad_norm": 1.4657010898381209, "learning_rate": 1.4844279744457495e-05, "loss": 0.2576, "step": 5609 }, { "epoch": 3.5892514395393476, "grad_norm": 1.6306368946676097, "learning_rate": 1.4842466855620256e-05, "loss": 0.2566, "step": 5610 }, { "epoch": 3.5898912348048624, "grad_norm": 1.460096692472347, "learning_rate": 1.4840653758856106e-05, "loss": 0.2388, "step": 5611 }, { "epoch": 3.5905310300703777, "grad_norm": 1.5939254302367611, "learning_rate": 1.483884045424289e-05, "loss": 0.2495, "step": 5612 }, { "epoch": 3.5911708253358925, "grad_norm": 1.657220843634073, "learning_rate": 1.4837026941858477e-05, "loss": 0.2661, "step": 5613 }, { "epoch": 3.5918106206014073, "grad_norm": 1.5012437968211663, "learning_rate": 1.483521322178073e-05, "loss": 0.2578, "step": 5614 }, { "epoch": 3.5924504158669226, "grad_norm": 1.5018129362650359, "learning_rate": 1.4833399294087527e-05, "loss": 0.2387, "step": 5615 }, { "epoch": 3.593090211132438, "grad_norm": 1.3949744924615446, "learning_rate": 1.4831585158856756e-05, "loss": 0.2414, "step": 5616 }, { "epoch": 3.5937300063979527, "grad_norm": 1.637234914941752, "learning_rate": 1.4829770816166317e-05, "loss": 0.2966, "step": 5617 }, { "epoch": 3.5943698016634675, "grad_norm": 1.5425760449432946, "learning_rate": 1.4827956266094108e-05, "loss": 0.2919, "step": 5618 }, { "epoch": 3.595009596928983, "grad_norm": 1.4568637372695992, "learning_rate": 1.4826141508718043e-05, "loss": 0.256, "step": 5619 }, { "epoch": 3.5956493921944976, "grad_norm": 1.6447274555407656, "learning_rate": 1.4824326544116051e-05, "loss": 0.2574, "step": 5620 }, { "epoch": 3.596289187460013, "grad_norm": 1.5351202142362375, "learning_rate": 1.4822511372366056e-05, "loss": 0.2544, "step": 5621 }, { "epoch": 3.5969289827255277, "grad_norm": 1.3452362038798413, "learning_rate": 1.4820695993546001e-05, "loss": 0.2177, "step": 5622 }, { "epoch": 3.597568777991043, "grad_norm": 1.4815570693412736, "learning_rate": 1.4818880407733837e-05, "loss": 0.2518, "step": 5623 }, { "epoch": 3.598208573256558, "grad_norm": 1.3261268700575732, "learning_rate": 1.4817064615007522e-05, "loss": 0.219, "step": 5624 }, { "epoch": 3.5988483685220727, "grad_norm": 1.604906268769164, "learning_rate": 1.481524861544502e-05, "loss": 0.2664, "step": 5625 }, { "epoch": 3.599488163787588, "grad_norm": 1.5536152202989457, "learning_rate": 1.4813432409124307e-05, "loss": 0.2417, "step": 5626 }, { "epoch": 3.6001279590531032, "grad_norm": 1.5629538768677336, "learning_rate": 1.4811615996123367e-05, "loss": 0.2485, "step": 5627 }, { "epoch": 3.600767754318618, "grad_norm": 1.4587295144553267, "learning_rate": 1.4809799376520198e-05, "loss": 0.2496, "step": 5628 }, { "epoch": 3.601407549584133, "grad_norm": 1.591029017666626, "learning_rate": 1.4807982550392799e-05, "loss": 0.2637, "step": 5629 }, { "epoch": 3.602047344849648, "grad_norm": 1.4635188389325688, "learning_rate": 1.4806165517819182e-05, "loss": 0.2598, "step": 5630 }, { "epoch": 3.602687140115163, "grad_norm": 1.4582140676720772, "learning_rate": 1.4804348278877368e-05, "loss": 0.2098, "step": 5631 }, { "epoch": 3.6033269353806783, "grad_norm": 1.560724283295828, "learning_rate": 1.4802530833645383e-05, "loss": 0.2534, "step": 5632 }, { "epoch": 3.603966730646193, "grad_norm": 1.60761525369124, "learning_rate": 1.4800713182201269e-05, "loss": 0.2601, "step": 5633 }, { "epoch": 3.6046065259117084, "grad_norm": 1.563979409690963, "learning_rate": 1.4798895324623071e-05, "loss": 0.2719, "step": 5634 }, { "epoch": 3.605246321177223, "grad_norm": 1.528293681200819, "learning_rate": 1.4797077260988845e-05, "loss": 0.2519, "step": 5635 }, { "epoch": 3.6058861164427385, "grad_norm": 1.493632642956175, "learning_rate": 1.479525899137665e-05, "loss": 0.2505, "step": 5636 }, { "epoch": 3.6065259117082533, "grad_norm": 1.543386777109153, "learning_rate": 1.4793440515864568e-05, "loss": 0.2739, "step": 5637 }, { "epoch": 3.6071657069737686, "grad_norm": 1.3505082854406665, "learning_rate": 1.4791621834530678e-05, "loss": 0.2558, "step": 5638 }, { "epoch": 3.6078055022392834, "grad_norm": 1.592732261632351, "learning_rate": 1.478980294745307e-05, "loss": 0.289, "step": 5639 }, { "epoch": 3.6084452975047983, "grad_norm": 1.5067253850198596, "learning_rate": 1.4787983854709842e-05, "loss": 0.2693, "step": 5640 }, { "epoch": 3.6090850927703135, "grad_norm": 1.512031984910665, "learning_rate": 1.4786164556379103e-05, "loss": 0.2442, "step": 5641 }, { "epoch": 3.609724888035829, "grad_norm": 1.5429515754979213, "learning_rate": 1.4784345052538975e-05, "loss": 0.2512, "step": 5642 }, { "epoch": 3.6103646833013436, "grad_norm": 1.5140109745549237, "learning_rate": 1.4782525343267579e-05, "loss": 0.2505, "step": 5643 }, { "epoch": 3.6110044785668585, "grad_norm": 1.5083149543792416, "learning_rate": 1.4780705428643053e-05, "loss": 0.2436, "step": 5644 }, { "epoch": 3.6116442738323737, "grad_norm": 1.546176657171157, "learning_rate": 1.477888530874354e-05, "loss": 0.2637, "step": 5645 }, { "epoch": 3.6122840690978886, "grad_norm": 1.503872636774367, "learning_rate": 1.4777064983647194e-05, "loss": 0.2517, "step": 5646 }, { "epoch": 3.612923864363404, "grad_norm": 1.5938311791046607, "learning_rate": 1.4775244453432174e-05, "loss": 0.2682, "step": 5647 }, { "epoch": 3.6135636596289187, "grad_norm": 1.624526308783857, "learning_rate": 1.4773423718176654e-05, "loss": 0.2733, "step": 5648 }, { "epoch": 3.614203454894434, "grad_norm": 1.5308357283079703, "learning_rate": 1.4771602777958806e-05, "loss": 0.2512, "step": 5649 }, { "epoch": 3.614843250159949, "grad_norm": 1.420106147788577, "learning_rate": 1.4769781632856828e-05, "loss": 0.2427, "step": 5650 }, { "epoch": 3.6154830454254636, "grad_norm": 1.501589346410283, "learning_rate": 1.4767960282948907e-05, "loss": 0.26, "step": 5651 }, { "epoch": 3.616122840690979, "grad_norm": 1.556188311936606, "learning_rate": 1.4766138728313255e-05, "loss": 0.2685, "step": 5652 }, { "epoch": 3.616762635956494, "grad_norm": 1.3800823769821922, "learning_rate": 1.4764316969028085e-05, "loss": 0.2266, "step": 5653 }, { "epoch": 3.617402431222009, "grad_norm": 1.3740307869501356, "learning_rate": 1.476249500517162e-05, "loss": 0.2268, "step": 5654 }, { "epoch": 3.618042226487524, "grad_norm": 1.5116776225161355, "learning_rate": 1.4760672836822091e-05, "loss": 0.2423, "step": 5655 }, { "epoch": 3.618682021753039, "grad_norm": 1.4892549008162916, "learning_rate": 1.4758850464057737e-05, "loss": 0.2718, "step": 5656 }, { "epoch": 3.619321817018554, "grad_norm": 1.4767506754158046, "learning_rate": 1.475702788695681e-05, "loss": 0.2436, "step": 5657 }, { "epoch": 3.619961612284069, "grad_norm": 1.5202208492122549, "learning_rate": 1.4755205105597567e-05, "loss": 0.2344, "step": 5658 }, { "epoch": 3.620601407549584, "grad_norm": 1.4949479327094568, "learning_rate": 1.4753382120058278e-05, "loss": 0.2387, "step": 5659 }, { "epoch": 3.6212412028150993, "grad_norm": 1.558625773674396, "learning_rate": 1.4751558930417216e-05, "loss": 0.2871, "step": 5660 }, { "epoch": 3.621880998080614, "grad_norm": 1.5897946060674177, "learning_rate": 1.4749735536752662e-05, "loss": 0.2566, "step": 5661 }, { "epoch": 3.622520793346129, "grad_norm": 1.416120853154634, "learning_rate": 1.4747911939142918e-05, "loss": 0.225, "step": 5662 }, { "epoch": 3.6231605886116443, "grad_norm": 1.4754671254585294, "learning_rate": 1.4746088137666281e-05, "loss": 0.2357, "step": 5663 }, { "epoch": 3.6238003838771595, "grad_norm": 1.5078253594878643, "learning_rate": 1.474426413240106e-05, "loss": 0.2624, "step": 5664 }, { "epoch": 3.6244401791426744, "grad_norm": 1.395440801291076, "learning_rate": 1.4742439923425578e-05, "loss": 0.2372, "step": 5665 }, { "epoch": 3.625079974408189, "grad_norm": 1.4199857702536702, "learning_rate": 1.4740615510818159e-05, "loss": 0.2518, "step": 5666 }, { "epoch": 3.6257197696737045, "grad_norm": 1.5028633797599857, "learning_rate": 1.4738790894657142e-05, "loss": 0.2479, "step": 5667 }, { "epoch": 3.6263595649392193, "grad_norm": 1.3749100249813395, "learning_rate": 1.4736966075020878e-05, "loss": 0.2193, "step": 5668 }, { "epoch": 3.6269993602047346, "grad_norm": 1.6959474176926035, "learning_rate": 1.4735141051987719e-05, "loss": 0.2863, "step": 5669 }, { "epoch": 3.6276391554702494, "grad_norm": 1.465660437475591, "learning_rate": 1.473331582563602e-05, "loss": 0.283, "step": 5670 }, { "epoch": 3.6282789507357647, "grad_norm": 1.4951783870653879, "learning_rate": 1.4731490396044162e-05, "loss": 0.2354, "step": 5671 }, { "epoch": 3.6289187460012795, "grad_norm": 1.530778916748257, "learning_rate": 1.4729664763290524e-05, "loss": 0.2473, "step": 5672 }, { "epoch": 3.629558541266795, "grad_norm": 1.5246864392587018, "learning_rate": 1.4727838927453493e-05, "loss": 0.244, "step": 5673 }, { "epoch": 3.6301983365323096, "grad_norm": 1.4138430064076366, "learning_rate": 1.4726012888611469e-05, "loss": 0.2234, "step": 5674 }, { "epoch": 3.630838131797825, "grad_norm": 1.3951947196495682, "learning_rate": 1.472418664684286e-05, "loss": 0.2325, "step": 5675 }, { "epoch": 3.6314779270633397, "grad_norm": 1.6003425143044199, "learning_rate": 1.4722360202226077e-05, "loss": 0.2613, "step": 5676 }, { "epoch": 3.6321177223288545, "grad_norm": 1.514442467688298, "learning_rate": 1.4720533554839551e-05, "loss": 0.2437, "step": 5677 }, { "epoch": 3.63275751759437, "grad_norm": 1.5246689450066375, "learning_rate": 1.4718706704761709e-05, "loss": 0.2622, "step": 5678 }, { "epoch": 3.633397312859885, "grad_norm": 1.5136833715317455, "learning_rate": 1.4716879652070995e-05, "loss": 0.2521, "step": 5679 }, { "epoch": 3.6340371081254, "grad_norm": 1.5001673221557943, "learning_rate": 1.4715052396845862e-05, "loss": 0.24, "step": 5680 }, { "epoch": 3.6346769033909148, "grad_norm": 1.5019919379651723, "learning_rate": 1.4713224939164762e-05, "loss": 0.2488, "step": 5681 }, { "epoch": 3.63531669865643, "grad_norm": 1.5259675013748744, "learning_rate": 1.471139727910617e-05, "loss": 0.2516, "step": 5682 }, { "epoch": 3.635956493921945, "grad_norm": 1.4292118752001999, "learning_rate": 1.4709569416748561e-05, "loss": 0.2474, "step": 5683 }, { "epoch": 3.63659628918746, "grad_norm": 1.5693356250046846, "learning_rate": 1.4707741352170418e-05, "loss": 0.2609, "step": 5684 }, { "epoch": 3.637236084452975, "grad_norm": 1.6894691407748397, "learning_rate": 1.4705913085450237e-05, "loss": 0.2858, "step": 5685 }, { "epoch": 3.6378758797184902, "grad_norm": 1.4939849030313512, "learning_rate": 1.4704084616666516e-05, "loss": 0.2284, "step": 5686 }, { "epoch": 3.638515674984005, "grad_norm": 1.4309866762381094, "learning_rate": 1.470225594589777e-05, "loss": 0.2387, "step": 5687 }, { "epoch": 3.63915547024952, "grad_norm": 1.452443627273521, "learning_rate": 1.4700427073222518e-05, "loss": 0.237, "step": 5688 }, { "epoch": 3.639795265515035, "grad_norm": 1.5617503615994621, "learning_rate": 1.4698597998719291e-05, "loss": 0.2406, "step": 5689 }, { "epoch": 3.6404350607805505, "grad_norm": 1.5473906748319666, "learning_rate": 1.4696768722466624e-05, "loss": 0.2511, "step": 5690 }, { "epoch": 3.6410748560460653, "grad_norm": 1.3985641421575012, "learning_rate": 1.4694939244543061e-05, "loss": 0.2105, "step": 5691 }, { "epoch": 3.64171465131158, "grad_norm": 1.4800962318289401, "learning_rate": 1.469310956502716e-05, "loss": 0.2591, "step": 5692 }, { "epoch": 3.6423544465770954, "grad_norm": 1.5622575628369066, "learning_rate": 1.4691279683997484e-05, "loss": 0.2364, "step": 5693 }, { "epoch": 3.6429942418426102, "grad_norm": 1.4799933529223888, "learning_rate": 1.4689449601532602e-05, "loss": 0.2405, "step": 5694 }, { "epoch": 3.6436340371081255, "grad_norm": 1.4963682476349245, "learning_rate": 1.4687619317711096e-05, "loss": 0.2463, "step": 5695 }, { "epoch": 3.6442738323736403, "grad_norm": 1.487627106979004, "learning_rate": 1.4685788832611554e-05, "loss": 0.2515, "step": 5696 }, { "epoch": 3.6449136276391556, "grad_norm": 1.5685507263456302, "learning_rate": 1.4683958146312579e-05, "loss": 0.2429, "step": 5697 }, { "epoch": 3.6455534229046704, "grad_norm": 1.5890799587576525, "learning_rate": 1.4682127258892766e-05, "loss": 0.2537, "step": 5698 }, { "epoch": 3.6461932181701853, "grad_norm": 1.4145987647868428, "learning_rate": 1.4680296170430744e-05, "loss": 0.226, "step": 5699 }, { "epoch": 3.6468330134357005, "grad_norm": 1.499510008497521, "learning_rate": 1.4678464881005128e-05, "loss": 0.2755, "step": 5700 }, { "epoch": 3.647472808701216, "grad_norm": 1.515418200124116, "learning_rate": 1.4676633390694553e-05, "loss": 0.2579, "step": 5701 }, { "epoch": 3.6481126039667306, "grad_norm": 1.598763582292168, "learning_rate": 1.4674801699577659e-05, "loss": 0.2435, "step": 5702 }, { "epoch": 3.6487523992322455, "grad_norm": 1.50384122623654, "learning_rate": 1.4672969807733094e-05, "loss": 0.2445, "step": 5703 }, { "epoch": 3.6493921944977608, "grad_norm": 1.488187451822872, "learning_rate": 1.4671137715239519e-05, "loss": 0.249, "step": 5704 }, { "epoch": 3.6500319897632756, "grad_norm": 1.6180051945922767, "learning_rate": 1.4669305422175599e-05, "loss": 0.2679, "step": 5705 }, { "epoch": 3.650671785028791, "grad_norm": 1.524084122694001, "learning_rate": 1.4667472928620011e-05, "loss": 0.2454, "step": 5706 }, { "epoch": 3.6513115802943057, "grad_norm": 1.525541559714884, "learning_rate": 1.4665640234651439e-05, "loss": 0.2594, "step": 5707 }, { "epoch": 3.651951375559821, "grad_norm": 1.5402595740929075, "learning_rate": 1.4663807340348574e-05, "loss": 0.286, "step": 5708 }, { "epoch": 3.652591170825336, "grad_norm": 1.3628908198805005, "learning_rate": 1.4661974245790118e-05, "loss": 0.245, "step": 5709 }, { "epoch": 3.653230966090851, "grad_norm": 1.3555546994740353, "learning_rate": 1.4660140951054781e-05, "loss": 0.2099, "step": 5710 }, { "epoch": 3.653870761356366, "grad_norm": 1.4137766387145532, "learning_rate": 1.465830745622128e-05, "loss": 0.2559, "step": 5711 }, { "epoch": 3.654510556621881, "grad_norm": 1.4120911473655173, "learning_rate": 1.4656473761368349e-05, "loss": 0.2322, "step": 5712 }, { "epoch": 3.655150351887396, "grad_norm": 1.52127865971382, "learning_rate": 1.465463986657471e-05, "loss": 0.2295, "step": 5713 }, { "epoch": 3.655790147152911, "grad_norm": 1.537868532889274, "learning_rate": 1.4652805771919121e-05, "loss": 0.2465, "step": 5714 }, { "epoch": 3.656429942418426, "grad_norm": 1.504125332210471, "learning_rate": 1.465097147748033e-05, "loss": 0.2653, "step": 5715 }, { "epoch": 3.6570697376839414, "grad_norm": 1.5421318916914795, "learning_rate": 1.4649136983337091e-05, "loss": 0.2461, "step": 5716 }, { "epoch": 3.657709532949456, "grad_norm": 1.54037522740864, "learning_rate": 1.4647302289568184e-05, "loss": 0.263, "step": 5717 }, { "epoch": 3.658349328214971, "grad_norm": 1.600172319937132, "learning_rate": 1.4645467396252382e-05, "loss": 0.2691, "step": 5718 }, { "epoch": 3.6589891234804863, "grad_norm": 1.4483138195248992, "learning_rate": 1.4643632303468473e-05, "loss": 0.2465, "step": 5719 }, { "epoch": 3.659628918746001, "grad_norm": 1.577543688534061, "learning_rate": 1.4641797011295255e-05, "loss": 0.2681, "step": 5720 }, { "epoch": 3.6602687140115164, "grad_norm": 1.3418604238488594, "learning_rate": 1.463996151981153e-05, "loss": 0.2172, "step": 5721 }, { "epoch": 3.6609085092770313, "grad_norm": 1.594310736944652, "learning_rate": 1.4638125829096113e-05, "loss": 0.2711, "step": 5722 }, { "epoch": 3.6615483045425465, "grad_norm": 1.4341207264320868, "learning_rate": 1.463628993922782e-05, "loss": 0.2476, "step": 5723 }, { "epoch": 3.6621880998080614, "grad_norm": 1.4368346712460145, "learning_rate": 1.4634453850285488e-05, "loss": 0.2502, "step": 5724 }, { "epoch": 3.662827895073576, "grad_norm": 1.3717783729492552, "learning_rate": 1.4632617562347951e-05, "loss": 0.2458, "step": 5725 }, { "epoch": 3.6634676903390915, "grad_norm": 1.3675064144318942, "learning_rate": 1.4630781075494053e-05, "loss": 0.2292, "step": 5726 }, { "epoch": 3.6641074856046068, "grad_norm": 1.626171834371159, "learning_rate": 1.4628944389802657e-05, "loss": 0.2607, "step": 5727 }, { "epoch": 3.6647472808701216, "grad_norm": 1.6352864214696101, "learning_rate": 1.462710750535262e-05, "loss": 0.252, "step": 5728 }, { "epoch": 3.6653870761356364, "grad_norm": 1.4900343807115943, "learning_rate": 1.4625270422222822e-05, "loss": 0.2549, "step": 5729 }, { "epoch": 3.6660268714011517, "grad_norm": 1.5540113410751257, "learning_rate": 1.4623433140492135e-05, "loss": 0.2749, "step": 5730 }, { "epoch": 3.6666666666666665, "grad_norm": 1.6327292773201139, "learning_rate": 1.4621595660239455e-05, "loss": 0.2628, "step": 5731 }, { "epoch": 3.667306461932182, "grad_norm": 1.3962880028749785, "learning_rate": 1.4619757981543681e-05, "loss": 0.2229, "step": 5732 }, { "epoch": 3.6679462571976966, "grad_norm": 1.6776766455625245, "learning_rate": 1.4617920104483715e-05, "loss": 0.282, "step": 5733 }, { "epoch": 3.668586052463212, "grad_norm": 1.4147946099347835, "learning_rate": 1.4616082029138473e-05, "loss": 0.2217, "step": 5734 }, { "epoch": 3.6692258477287267, "grad_norm": 1.5292371131205242, "learning_rate": 1.4614243755586883e-05, "loss": 0.2795, "step": 5735 }, { "epoch": 3.669865642994242, "grad_norm": 1.5994710025261338, "learning_rate": 1.4612405283907873e-05, "loss": 0.2727, "step": 5736 }, { "epoch": 3.670505438259757, "grad_norm": 1.593078646560135, "learning_rate": 1.4610566614180384e-05, "loss": 0.2718, "step": 5737 }, { "epoch": 3.671145233525272, "grad_norm": 1.6122493713055592, "learning_rate": 1.4608727746483369e-05, "loss": 0.2719, "step": 5738 }, { "epoch": 3.671785028790787, "grad_norm": 1.4183062056925797, "learning_rate": 1.4606888680895779e-05, "loss": 0.2269, "step": 5739 }, { "epoch": 3.6724248240563018, "grad_norm": 1.4474227782715485, "learning_rate": 1.4605049417496589e-05, "loss": 0.248, "step": 5740 }, { "epoch": 3.673064619321817, "grad_norm": 1.535693767099723, "learning_rate": 1.4603209956364764e-05, "loss": 0.2597, "step": 5741 }, { "epoch": 3.6737044145873323, "grad_norm": 1.4914375525128503, "learning_rate": 1.4601370297579293e-05, "loss": 0.2791, "step": 5742 }, { "epoch": 3.674344209852847, "grad_norm": 1.5129414793805807, "learning_rate": 1.4599530441219168e-05, "loss": 0.2436, "step": 5743 }, { "epoch": 3.674984005118362, "grad_norm": 1.5937066417982575, "learning_rate": 1.4597690387363388e-05, "loss": 0.2624, "step": 5744 }, { "epoch": 3.6756238003838773, "grad_norm": 1.5500684877010458, "learning_rate": 1.4595850136090961e-05, "loss": 0.2576, "step": 5745 }, { "epoch": 3.676263595649392, "grad_norm": 1.6874268802153252, "learning_rate": 1.4594009687480903e-05, "loss": 0.2656, "step": 5746 }, { "epoch": 3.6769033909149074, "grad_norm": 1.6376924721745092, "learning_rate": 1.4592169041612243e-05, "loss": 0.2373, "step": 5747 }, { "epoch": 3.677543186180422, "grad_norm": 1.6391713638852123, "learning_rate": 1.4590328198564013e-05, "loss": 0.2724, "step": 5748 }, { "epoch": 3.6781829814459375, "grad_norm": 1.4297558830780916, "learning_rate": 1.4588487158415252e-05, "loss": 0.2605, "step": 5749 }, { "epoch": 3.6788227767114523, "grad_norm": 1.6487611952987955, "learning_rate": 1.4586645921245019e-05, "loss": 0.2695, "step": 5750 }, { "epoch": 3.679462571976967, "grad_norm": 1.486924606635526, "learning_rate": 1.4584804487132367e-05, "loss": 0.2642, "step": 5751 }, { "epoch": 3.6801023672424824, "grad_norm": 1.5187513960481762, "learning_rate": 1.4582962856156368e-05, "loss": 0.2264, "step": 5752 }, { "epoch": 3.6807421625079977, "grad_norm": 1.6769121008813233, "learning_rate": 1.4581121028396098e-05, "loss": 0.2926, "step": 5753 }, { "epoch": 3.6813819577735125, "grad_norm": 1.4102542875801332, "learning_rate": 1.4579279003930635e-05, "loss": 0.2282, "step": 5754 }, { "epoch": 3.6820217530390273, "grad_norm": 1.4444486255039841, "learning_rate": 1.457743678283908e-05, "loss": 0.2279, "step": 5755 }, { "epoch": 3.6826615483045426, "grad_norm": 1.4068733540535956, "learning_rate": 1.4575594365200532e-05, "loss": 0.2268, "step": 5756 }, { "epoch": 3.6833013435700575, "grad_norm": 1.3623132090173748, "learning_rate": 1.4573751751094103e-05, "loss": 0.2517, "step": 5757 }, { "epoch": 3.6839411388355727, "grad_norm": 1.4538620738329755, "learning_rate": 1.4571908940598907e-05, "loss": 0.2594, "step": 5758 }, { "epoch": 3.6845809341010876, "grad_norm": 1.5579507904887764, "learning_rate": 1.4570065933794077e-05, "loss": 0.2776, "step": 5759 }, { "epoch": 3.685220729366603, "grad_norm": 1.5122067330313593, "learning_rate": 1.4568222730758743e-05, "loss": 0.2641, "step": 5760 }, { "epoch": 3.6858605246321177, "grad_norm": 1.5085809680826254, "learning_rate": 1.4566379331572051e-05, "loss": 0.2447, "step": 5761 }, { "epoch": 3.6865003198976325, "grad_norm": 1.5502809143537803, "learning_rate": 1.4564535736313157e-05, "loss": 0.2591, "step": 5762 }, { "epoch": 3.6871401151631478, "grad_norm": 1.5971427795981972, "learning_rate": 1.4562691945061214e-05, "loss": 0.2286, "step": 5763 }, { "epoch": 3.687779910428663, "grad_norm": 1.5071450535412958, "learning_rate": 1.4560847957895394e-05, "loss": 0.2348, "step": 5764 }, { "epoch": 3.688419705694178, "grad_norm": 1.491731393794149, "learning_rate": 1.455900377489488e-05, "loss": 0.2249, "step": 5765 }, { "epoch": 3.6890595009596927, "grad_norm": 1.5470359497878208, "learning_rate": 1.455715939613885e-05, "loss": 0.2241, "step": 5766 }, { "epoch": 3.689699296225208, "grad_norm": 1.5373781769842758, "learning_rate": 1.4555314821706507e-05, "loss": 0.2432, "step": 5767 }, { "epoch": 3.690339091490723, "grad_norm": 1.4508680910434026, "learning_rate": 1.4553470051677048e-05, "loss": 0.2242, "step": 5768 }, { "epoch": 3.690978886756238, "grad_norm": 1.5756077264002641, "learning_rate": 1.4551625086129681e-05, "loss": 0.2831, "step": 5769 }, { "epoch": 3.691618682021753, "grad_norm": 1.5296521694810705, "learning_rate": 1.4549779925143634e-05, "loss": 0.2585, "step": 5770 }, { "epoch": 3.692258477287268, "grad_norm": 1.5807055326422985, "learning_rate": 1.4547934568798128e-05, "loss": 0.2675, "step": 5771 }, { "epoch": 3.692898272552783, "grad_norm": 1.4657068408799843, "learning_rate": 1.4546089017172402e-05, "loss": 0.2541, "step": 5772 }, { "epoch": 3.6935380678182983, "grad_norm": 1.6015318648046277, "learning_rate": 1.45442432703457e-05, "loss": 0.2694, "step": 5773 }, { "epoch": 3.694177863083813, "grad_norm": 1.5048210172566368, "learning_rate": 1.4542397328397276e-05, "loss": 0.2358, "step": 5774 }, { "epoch": 3.6948176583493284, "grad_norm": 1.4441693225426087, "learning_rate": 1.4540551191406392e-05, "loss": 0.2551, "step": 5775 }, { "epoch": 3.6954574536148432, "grad_norm": 1.3985018737574875, "learning_rate": 1.4538704859452315e-05, "loss": 0.2521, "step": 5776 }, { "epoch": 3.696097248880358, "grad_norm": 1.4842591772187228, "learning_rate": 1.4536858332614327e-05, "loss": 0.2409, "step": 5777 }, { "epoch": 3.6967370441458733, "grad_norm": 1.3284395271950982, "learning_rate": 1.453501161097171e-05, "loss": 0.2232, "step": 5778 }, { "epoch": 3.6973768394113886, "grad_norm": 1.4497252190858674, "learning_rate": 1.4533164694603766e-05, "loss": 0.2201, "step": 5779 }, { "epoch": 3.6980166346769034, "grad_norm": 1.3950506100745483, "learning_rate": 1.4531317583589787e-05, "loss": 0.2284, "step": 5780 }, { "epoch": 3.6986564299424183, "grad_norm": 1.4315158500164233, "learning_rate": 1.4529470278009097e-05, "loss": 0.2396, "step": 5781 }, { "epoch": 3.6992962252079336, "grad_norm": 1.590517020358406, "learning_rate": 1.4527622777941008e-05, "loss": 0.2589, "step": 5782 }, { "epoch": 3.6999360204734484, "grad_norm": 1.4982153991393867, "learning_rate": 1.4525775083464852e-05, "loss": 0.2581, "step": 5783 }, { "epoch": 3.7005758157389637, "grad_norm": 1.426199055601238, "learning_rate": 1.4523927194659968e-05, "loss": 0.2396, "step": 5784 }, { "epoch": 3.7012156110044785, "grad_norm": 1.3303993961144405, "learning_rate": 1.4522079111605693e-05, "loss": 0.2258, "step": 5785 }, { "epoch": 3.7018554062699938, "grad_norm": 1.438156971456895, "learning_rate": 1.4520230834381386e-05, "loss": 0.2462, "step": 5786 }, { "epoch": 3.7024952015355086, "grad_norm": 1.5596996654328223, "learning_rate": 1.451838236306641e-05, "loss": 0.2776, "step": 5787 }, { "epoch": 3.7031349968010234, "grad_norm": 1.447757327474461, "learning_rate": 1.451653369774013e-05, "loss": 0.2405, "step": 5788 }, { "epoch": 3.7037747920665387, "grad_norm": 1.4798986733431316, "learning_rate": 1.4514684838481929e-05, "loss": 0.2816, "step": 5789 }, { "epoch": 3.704414587332054, "grad_norm": 1.4100014428243026, "learning_rate": 1.4512835785371193e-05, "loss": 0.2073, "step": 5790 }, { "epoch": 3.705054382597569, "grad_norm": 1.474281262131324, "learning_rate": 1.4510986538487312e-05, "loss": 0.2424, "step": 5791 }, { "epoch": 3.7056941778630836, "grad_norm": 1.5526867490334526, "learning_rate": 1.4509137097909698e-05, "loss": 0.2781, "step": 5792 }, { "epoch": 3.706333973128599, "grad_norm": 1.5500445232765458, "learning_rate": 1.4507287463717754e-05, "loss": 0.2395, "step": 5793 }, { "epoch": 3.7069737683941137, "grad_norm": 1.5456181208890876, "learning_rate": 1.4505437635990909e-05, "loss": 0.2336, "step": 5794 }, { "epoch": 3.707613563659629, "grad_norm": 1.5885055472794514, "learning_rate": 1.4503587614808582e-05, "loss": 0.2859, "step": 5795 }, { "epoch": 3.708253358925144, "grad_norm": 1.635808253391204, "learning_rate": 1.4501737400250215e-05, "loss": 0.3278, "step": 5796 }, { "epoch": 3.708893154190659, "grad_norm": 1.540312176416634, "learning_rate": 1.4499886992395253e-05, "loss": 0.2407, "step": 5797 }, { "epoch": 3.709532949456174, "grad_norm": 1.4037743371635296, "learning_rate": 1.449803639132315e-05, "loss": 0.2288, "step": 5798 }, { "epoch": 3.710172744721689, "grad_norm": 1.461170313830175, "learning_rate": 1.4496185597113366e-05, "loss": 0.2863, "step": 5799 }, { "epoch": 3.710812539987204, "grad_norm": 1.5562501039236154, "learning_rate": 1.4494334609845366e-05, "loss": 0.2786, "step": 5800 }, { "epoch": 3.7114523352527193, "grad_norm": 1.429667539604825, "learning_rate": 1.4492483429598637e-05, "loss": 0.241, "step": 5801 }, { "epoch": 3.712092130518234, "grad_norm": 1.2983166894303872, "learning_rate": 1.4490632056452658e-05, "loss": 0.2186, "step": 5802 }, { "epoch": 3.712731925783749, "grad_norm": 1.4889194517879423, "learning_rate": 1.4488780490486926e-05, "loss": 0.2291, "step": 5803 }, { "epoch": 3.7133717210492643, "grad_norm": 1.471069468521929, "learning_rate": 1.4486928731780947e-05, "loss": 0.2563, "step": 5804 }, { "epoch": 3.714011516314779, "grad_norm": 1.4804116510019554, "learning_rate": 1.4485076780414232e-05, "loss": 0.2317, "step": 5805 }, { "epoch": 3.7146513115802944, "grad_norm": 1.6143154236914627, "learning_rate": 1.4483224636466296e-05, "loss": 0.2943, "step": 5806 }, { "epoch": 3.715291106845809, "grad_norm": 1.566224153787276, "learning_rate": 1.4481372300016668e-05, "loss": 0.2542, "step": 5807 }, { "epoch": 3.7159309021113245, "grad_norm": 1.638175207243326, "learning_rate": 1.4479519771144885e-05, "loss": 0.2749, "step": 5808 }, { "epoch": 3.7165706973768393, "grad_norm": 1.416161909721996, "learning_rate": 1.4477667049930495e-05, "loss": 0.2498, "step": 5809 }, { "epoch": 3.7172104926423546, "grad_norm": 1.3910471706352545, "learning_rate": 1.4475814136453043e-05, "loss": 0.2285, "step": 5810 }, { "epoch": 3.7178502879078694, "grad_norm": 1.6099582507671817, "learning_rate": 1.4473961030792095e-05, "loss": 0.2573, "step": 5811 }, { "epoch": 3.7184900831733847, "grad_norm": 1.4218311254555014, "learning_rate": 1.4472107733027221e-05, "loss": 0.2399, "step": 5812 }, { "epoch": 3.7191298784388995, "grad_norm": 1.3642269553842694, "learning_rate": 1.4470254243237993e-05, "loss": 0.2418, "step": 5813 }, { "epoch": 3.7197696737044144, "grad_norm": 1.5194324760264197, "learning_rate": 1.4468400561504004e-05, "loss": 0.2652, "step": 5814 }, { "epoch": 3.7204094689699296, "grad_norm": 1.5683825465642363, "learning_rate": 1.4466546687904839e-05, "loss": 0.2508, "step": 5815 }, { "epoch": 3.721049264235445, "grad_norm": 1.3701638281979092, "learning_rate": 1.4464692622520107e-05, "loss": 0.2259, "step": 5816 }, { "epoch": 3.7216890595009597, "grad_norm": 1.6116858358117014, "learning_rate": 1.4462838365429414e-05, "loss": 0.2593, "step": 5817 }, { "epoch": 3.7223288547664746, "grad_norm": 1.5716181463784453, "learning_rate": 1.4460983916712384e-05, "loss": 0.2619, "step": 5818 }, { "epoch": 3.72296865003199, "grad_norm": 1.7056995136736675, "learning_rate": 1.4459129276448636e-05, "loss": 0.3047, "step": 5819 }, { "epoch": 3.7236084452975047, "grad_norm": 1.5431063862023602, "learning_rate": 1.4457274444717812e-05, "loss": 0.2723, "step": 5820 }, { "epoch": 3.72424824056302, "grad_norm": 1.6263441837401547, "learning_rate": 1.445541942159955e-05, "loss": 0.2609, "step": 5821 }, { "epoch": 3.724888035828535, "grad_norm": 1.5294068518141322, "learning_rate": 1.4453564207173507e-05, "loss": 0.2564, "step": 5822 }, { "epoch": 3.72552783109405, "grad_norm": 1.5014368174997126, "learning_rate": 1.4451708801519335e-05, "loss": 0.245, "step": 5823 }, { "epoch": 3.726167626359565, "grad_norm": 1.3540750128795738, "learning_rate": 1.4449853204716711e-05, "loss": 0.2364, "step": 5824 }, { "epoch": 3.7268074216250797, "grad_norm": 1.5635664539633047, "learning_rate": 1.4447997416845301e-05, "loss": 0.2646, "step": 5825 }, { "epoch": 3.727447216890595, "grad_norm": 1.4395147378525317, "learning_rate": 1.4446141437984795e-05, "loss": 0.2322, "step": 5826 }, { "epoch": 3.7280870121561103, "grad_norm": 1.4769558394933815, "learning_rate": 1.4444285268214886e-05, "loss": 0.2552, "step": 5827 }, { "epoch": 3.728726807421625, "grad_norm": 1.4952926494148693, "learning_rate": 1.4442428907615275e-05, "loss": 0.2308, "step": 5828 }, { "epoch": 3.72936660268714, "grad_norm": 1.4600172566439054, "learning_rate": 1.4440572356265668e-05, "loss": 0.2301, "step": 5829 }, { "epoch": 3.730006397952655, "grad_norm": 1.6281311121312358, "learning_rate": 1.4438715614245781e-05, "loss": 0.2347, "step": 5830 }, { "epoch": 3.73064619321817, "grad_norm": 1.467792326487647, "learning_rate": 1.4436858681635343e-05, "loss": 0.2438, "step": 5831 }, { "epoch": 3.7312859884836853, "grad_norm": 1.5443297011156103, "learning_rate": 1.4435001558514089e-05, "loss": 0.2654, "step": 5832 }, { "epoch": 3.7319257837492, "grad_norm": 1.5163688220921099, "learning_rate": 1.4433144244961752e-05, "loss": 0.2486, "step": 5833 }, { "epoch": 3.7325655790147154, "grad_norm": 1.4338272239033205, "learning_rate": 1.443128674105809e-05, "loss": 0.2369, "step": 5834 }, { "epoch": 3.7332053742802302, "grad_norm": 1.679213584413335, "learning_rate": 1.4429429046882857e-05, "loss": 0.2738, "step": 5835 }, { "epoch": 3.7338451695457455, "grad_norm": 1.6062097402418003, "learning_rate": 1.4427571162515822e-05, "loss": 0.2692, "step": 5836 }, { "epoch": 3.7344849648112604, "grad_norm": 1.5634281394487088, "learning_rate": 1.4425713088036754e-05, "loss": 0.2568, "step": 5837 }, { "epoch": 3.7351247600767756, "grad_norm": 1.5310258907419523, "learning_rate": 1.442385482352544e-05, "loss": 0.2835, "step": 5838 }, { "epoch": 3.7357645553422905, "grad_norm": 1.4139568929961328, "learning_rate": 1.442199636906167e-05, "loss": 0.2436, "step": 5839 }, { "epoch": 3.7364043506078053, "grad_norm": 1.4640397470666882, "learning_rate": 1.4420137724725243e-05, "loss": 0.2589, "step": 5840 }, { "epoch": 3.7370441458733206, "grad_norm": 1.4516411916459266, "learning_rate": 1.4418278890595962e-05, "loss": 0.2479, "step": 5841 }, { "epoch": 3.737683941138836, "grad_norm": 1.587130044523004, "learning_rate": 1.4416419866753648e-05, "loss": 0.2679, "step": 5842 }, { "epoch": 3.7383237364043507, "grad_norm": 1.5133403580451483, "learning_rate": 1.441456065327812e-05, "loss": 0.2793, "step": 5843 }, { "epoch": 3.7389635316698655, "grad_norm": 1.6397712747811592, "learning_rate": 1.441270125024921e-05, "loss": 0.2923, "step": 5844 }, { "epoch": 3.739603326935381, "grad_norm": 1.5165777106467728, "learning_rate": 1.4410841657746759e-05, "loss": 0.2501, "step": 5845 }, { "epoch": 3.7402431222008956, "grad_norm": 1.6924071783109425, "learning_rate": 1.4408981875850613e-05, "loss": 0.2871, "step": 5846 }, { "epoch": 3.740882917466411, "grad_norm": 1.2568400358376157, "learning_rate": 1.4407121904640626e-05, "loss": 0.1993, "step": 5847 }, { "epoch": 3.7415227127319257, "grad_norm": 1.5534574566224975, "learning_rate": 1.4405261744196665e-05, "loss": 0.2571, "step": 5848 }, { "epoch": 3.742162507997441, "grad_norm": 1.4381856530268489, "learning_rate": 1.4403401394598604e-05, "loss": 0.2332, "step": 5849 }, { "epoch": 3.742802303262956, "grad_norm": 1.486247763819812, "learning_rate": 1.440154085592632e-05, "loss": 0.2419, "step": 5850 }, { "epoch": 3.7434420985284707, "grad_norm": 1.4401887542248162, "learning_rate": 1.4399680128259698e-05, "loss": 0.2443, "step": 5851 }, { "epoch": 3.744081893793986, "grad_norm": 1.4101707599188507, "learning_rate": 1.439781921167864e-05, "loss": 0.2338, "step": 5852 }, { "epoch": 3.744721689059501, "grad_norm": 1.4604193466419495, "learning_rate": 1.4395958106263045e-05, "loss": 0.2295, "step": 5853 }, { "epoch": 3.745361484325016, "grad_norm": 1.3624741963904374, "learning_rate": 1.4394096812092832e-05, "loss": 0.212, "step": 5854 }, { "epoch": 3.746001279590531, "grad_norm": 1.5056109287868926, "learning_rate": 1.4392235329247917e-05, "loss": 0.2879, "step": 5855 }, { "epoch": 3.746641074856046, "grad_norm": 1.5999194456187456, "learning_rate": 1.4390373657808233e-05, "loss": 0.2546, "step": 5856 }, { "epoch": 3.747280870121561, "grad_norm": 1.537501434412052, "learning_rate": 1.4388511797853707e-05, "loss": 0.2482, "step": 5857 }, { "epoch": 3.7479206653870762, "grad_norm": 1.592003261958615, "learning_rate": 1.4386649749464299e-05, "loss": 0.2583, "step": 5858 }, { "epoch": 3.748560460652591, "grad_norm": 1.488067276546459, "learning_rate": 1.438478751271995e-05, "loss": 0.2762, "step": 5859 }, { "epoch": 3.7492002559181064, "grad_norm": 1.3566073685542297, "learning_rate": 1.4382925087700625e-05, "loss": 0.2189, "step": 5860 }, { "epoch": 3.749840051183621, "grad_norm": 1.3336662829007633, "learning_rate": 1.4381062474486296e-05, "loss": 0.2019, "step": 5861 }, { "epoch": 3.750479846449136, "grad_norm": 1.533713378649786, "learning_rate": 1.4379199673156934e-05, "loss": 0.2336, "step": 5862 }, { "epoch": 3.7511196417146513, "grad_norm": 1.4622721479164222, "learning_rate": 1.437733668379253e-05, "loss": 0.2588, "step": 5863 }, { "epoch": 3.7517594369801666, "grad_norm": 1.5331104727848333, "learning_rate": 1.4375473506473077e-05, "loss": 0.2432, "step": 5864 }, { "epoch": 3.7523992322456814, "grad_norm": 1.508509462804399, "learning_rate": 1.4373610141278573e-05, "loss": 0.2599, "step": 5865 }, { "epoch": 3.7530390275111962, "grad_norm": 1.495722478455193, "learning_rate": 1.4371746588289031e-05, "loss": 0.2304, "step": 5866 }, { "epoch": 3.7536788227767115, "grad_norm": 1.464229599250598, "learning_rate": 1.4369882847584464e-05, "loss": 0.258, "step": 5867 }, { "epoch": 3.7543186180422263, "grad_norm": 1.4544902514946076, "learning_rate": 1.4368018919244904e-05, "loss": 0.2537, "step": 5868 }, { "epoch": 3.7549584133077416, "grad_norm": 1.495093868295168, "learning_rate": 1.4366154803350381e-05, "loss": 0.2216, "step": 5869 }, { "epoch": 3.7555982085732564, "grad_norm": 1.4101170632747777, "learning_rate": 1.4364290499980935e-05, "loss": 0.2423, "step": 5870 }, { "epoch": 3.7562380038387717, "grad_norm": 1.4922929775461928, "learning_rate": 1.436242600921662e-05, "loss": 0.2261, "step": 5871 }, { "epoch": 3.7568777991042865, "grad_norm": 1.5527192365640106, "learning_rate": 1.4360561331137493e-05, "loss": 0.2792, "step": 5872 }, { "epoch": 3.757517594369802, "grad_norm": 1.5501996630406447, "learning_rate": 1.435869646582362e-05, "loss": 0.2614, "step": 5873 }, { "epoch": 3.7581573896353166, "grad_norm": 1.5100876399112677, "learning_rate": 1.4356831413355074e-05, "loss": 0.2704, "step": 5874 }, { "epoch": 3.758797184900832, "grad_norm": 1.528708512148475, "learning_rate": 1.4354966173811938e-05, "loss": 0.2479, "step": 5875 }, { "epoch": 3.7594369801663468, "grad_norm": 1.3890982245970722, "learning_rate": 1.43531007472743e-05, "loss": 0.2178, "step": 5876 }, { "epoch": 3.7600767754318616, "grad_norm": 1.46960691356299, "learning_rate": 1.4351235133822259e-05, "loss": 0.2456, "step": 5877 }, { "epoch": 3.760716570697377, "grad_norm": 1.5088283530809692, "learning_rate": 1.434936933353592e-05, "loss": 0.2474, "step": 5878 }, { "epoch": 3.761356365962892, "grad_norm": 1.6310787212446771, "learning_rate": 1.4347503346495405e-05, "loss": 0.2746, "step": 5879 }, { "epoch": 3.761996161228407, "grad_norm": 1.4531642385543997, "learning_rate": 1.4345637172780826e-05, "loss": 0.242, "step": 5880 }, { "epoch": 3.762635956493922, "grad_norm": 1.649288272317787, "learning_rate": 1.4343770812472317e-05, "loss": 0.2803, "step": 5881 }, { "epoch": 3.763275751759437, "grad_norm": 1.3839116634804252, "learning_rate": 1.4341904265650015e-05, "loss": 0.2342, "step": 5882 }, { "epoch": 3.763915547024952, "grad_norm": 1.4611528357997516, "learning_rate": 1.434003753239407e-05, "loss": 0.2205, "step": 5883 }, { "epoch": 3.764555342290467, "grad_norm": 1.4200042354518423, "learning_rate": 1.4338170612784633e-05, "loss": 0.2063, "step": 5884 }, { "epoch": 3.765195137555982, "grad_norm": 1.4396270591878009, "learning_rate": 1.4336303506901863e-05, "loss": 0.2395, "step": 5885 }, { "epoch": 3.7658349328214973, "grad_norm": 1.3864604934430178, "learning_rate": 1.4334436214825939e-05, "loss": 0.2151, "step": 5886 }, { "epoch": 3.766474728087012, "grad_norm": 1.5074339393152887, "learning_rate": 1.4332568736637032e-05, "loss": 0.2399, "step": 5887 }, { "epoch": 3.767114523352527, "grad_norm": 1.4907814020894044, "learning_rate": 1.4330701072415331e-05, "loss": 0.2414, "step": 5888 }, { "epoch": 3.767754318618042, "grad_norm": 1.5022309183146358, "learning_rate": 1.4328833222241031e-05, "loss": 0.2443, "step": 5889 }, { "epoch": 3.7683941138835575, "grad_norm": 1.61571969942348, "learning_rate": 1.4326965186194333e-05, "loss": 0.2902, "step": 5890 }, { "epoch": 3.7690339091490723, "grad_norm": 1.4738458211915344, "learning_rate": 1.4325096964355445e-05, "loss": 0.2608, "step": 5891 }, { "epoch": 3.769673704414587, "grad_norm": 1.4380368217838007, "learning_rate": 1.4323228556804587e-05, "loss": 0.2417, "step": 5892 }, { "epoch": 3.7703134996801024, "grad_norm": 1.4691853679417812, "learning_rate": 1.4321359963621985e-05, "loss": 0.2568, "step": 5893 }, { "epoch": 3.7709532949456173, "grad_norm": 1.4155258573096674, "learning_rate": 1.4319491184887874e-05, "loss": 0.2208, "step": 5894 }, { "epoch": 3.7715930902111325, "grad_norm": 1.580304993289639, "learning_rate": 1.4317622220682498e-05, "loss": 0.2924, "step": 5895 }, { "epoch": 3.7722328854766474, "grad_norm": 1.3995596975725226, "learning_rate": 1.4315753071086103e-05, "loss": 0.2381, "step": 5896 }, { "epoch": 3.7728726807421626, "grad_norm": 1.4764413611312117, "learning_rate": 1.4313883736178946e-05, "loss": 0.2731, "step": 5897 }, { "epoch": 3.7735124760076775, "grad_norm": 1.549623589295013, "learning_rate": 1.4312014216041297e-05, "loss": 0.2721, "step": 5898 }, { "epoch": 3.7741522712731923, "grad_norm": 1.5866363080788544, "learning_rate": 1.4310144510753428e-05, "loss": 0.2774, "step": 5899 }, { "epoch": 3.7747920665387076, "grad_norm": 1.377983307945758, "learning_rate": 1.4308274620395617e-05, "loss": 0.2131, "step": 5900 }, { "epoch": 3.775431861804223, "grad_norm": 1.4400299611866862, "learning_rate": 1.4306404545048163e-05, "loss": 0.2309, "step": 5901 }, { "epoch": 3.7760716570697377, "grad_norm": 1.7420569529779766, "learning_rate": 1.4304534284791354e-05, "loss": 0.3205, "step": 5902 }, { "epoch": 3.7767114523352525, "grad_norm": 1.493513035884785, "learning_rate": 1.4302663839705504e-05, "loss": 0.2564, "step": 5903 }, { "epoch": 3.777351247600768, "grad_norm": 1.5722520260230195, "learning_rate": 1.4300793209870922e-05, "loss": 0.2714, "step": 5904 }, { "epoch": 3.777991042866283, "grad_norm": 1.5351030372148813, "learning_rate": 1.4298922395367927e-05, "loss": 0.2519, "step": 5905 }, { "epoch": 3.778630838131798, "grad_norm": 1.7227365100918595, "learning_rate": 1.4297051396276855e-05, "loss": 0.2837, "step": 5906 }, { "epoch": 3.7792706333973127, "grad_norm": 1.4714904156539863, "learning_rate": 1.4295180212678037e-05, "loss": 0.2582, "step": 5907 }, { "epoch": 3.779910428662828, "grad_norm": 1.5433982054234239, "learning_rate": 1.429330884465182e-05, "loss": 0.2697, "step": 5908 }, { "epoch": 3.780550223928343, "grad_norm": 1.4786936777845252, "learning_rate": 1.4291437292278562e-05, "loss": 0.2499, "step": 5909 }, { "epoch": 3.781190019193858, "grad_norm": 1.6668642278636707, "learning_rate": 1.428956555563862e-05, "loss": 0.2831, "step": 5910 }, { "epoch": 3.781829814459373, "grad_norm": 1.4965591369643587, "learning_rate": 1.4287693634812366e-05, "loss": 0.2728, "step": 5911 }, { "epoch": 3.782469609724888, "grad_norm": 1.4564371369327413, "learning_rate": 1.428582152988017e-05, "loss": 0.2358, "step": 5912 }, { "epoch": 3.783109404990403, "grad_norm": 1.5145568283083608, "learning_rate": 1.4283949240922423e-05, "loss": 0.2819, "step": 5913 }, { "epoch": 3.783749200255918, "grad_norm": 1.4479202929310468, "learning_rate": 1.4282076768019514e-05, "loss": 0.2248, "step": 5914 }, { "epoch": 3.784388995521433, "grad_norm": 1.4119209900901128, "learning_rate": 1.428020411125185e-05, "loss": 0.2461, "step": 5915 }, { "epoch": 3.7850287907869484, "grad_norm": 1.2842193572141898, "learning_rate": 1.4278331270699832e-05, "loss": 0.1987, "step": 5916 }, { "epoch": 3.7856685860524633, "grad_norm": 1.5632263016171628, "learning_rate": 1.427645824644388e-05, "loss": 0.2563, "step": 5917 }, { "epoch": 3.786308381317978, "grad_norm": 1.525470885377374, "learning_rate": 1.4274585038564422e-05, "loss": 0.2457, "step": 5918 }, { "epoch": 3.7869481765834934, "grad_norm": 1.4398111205133344, "learning_rate": 1.4272711647141884e-05, "loss": 0.2468, "step": 5919 }, { "epoch": 3.787587971849008, "grad_norm": 1.635480138415271, "learning_rate": 1.4270838072256708e-05, "loss": 0.2517, "step": 5920 }, { "epoch": 3.7882277671145235, "grad_norm": 1.534139400680292, "learning_rate": 1.4268964313989342e-05, "loss": 0.2668, "step": 5921 }, { "epoch": 3.7888675623800383, "grad_norm": 1.5729721467680482, "learning_rate": 1.426709037242024e-05, "loss": 0.2474, "step": 5922 }, { "epoch": 3.7895073576455536, "grad_norm": 1.4960656782613069, "learning_rate": 1.4265216247629871e-05, "loss": 0.2481, "step": 5923 }, { "epoch": 3.7901471529110684, "grad_norm": 1.6545437860621508, "learning_rate": 1.4263341939698706e-05, "loss": 0.22, "step": 5924 }, { "epoch": 3.7907869481765832, "grad_norm": 1.6664427621823226, "learning_rate": 1.4261467448707219e-05, "loss": 0.2481, "step": 5925 }, { "epoch": 3.7914267434420985, "grad_norm": 1.6393638555434804, "learning_rate": 1.4259592774735901e-05, "loss": 0.279, "step": 5926 }, { "epoch": 3.792066538707614, "grad_norm": 1.4699139216902701, "learning_rate": 1.4257717917865246e-05, "loss": 0.2556, "step": 5927 }, { "epoch": 3.7927063339731286, "grad_norm": 1.5592977380616158, "learning_rate": 1.4255842878175753e-05, "loss": 0.2694, "step": 5928 }, { "epoch": 3.7933461292386434, "grad_norm": 1.4366753113770672, "learning_rate": 1.4253967655747945e-05, "loss": 0.2221, "step": 5929 }, { "epoch": 3.7939859245041587, "grad_norm": 1.5168080147651137, "learning_rate": 1.4252092250662328e-05, "loss": 0.2445, "step": 5930 }, { "epoch": 3.7946257197696736, "grad_norm": 1.6178201162072805, "learning_rate": 1.4250216662999437e-05, "loss": 0.2847, "step": 5931 }, { "epoch": 3.795265515035189, "grad_norm": 1.4681432321781507, "learning_rate": 1.4248340892839801e-05, "loss": 0.2348, "step": 5932 }, { "epoch": 3.7959053103007037, "grad_norm": 1.4970270969323378, "learning_rate": 1.4246464940263964e-05, "loss": 0.2589, "step": 5933 }, { "epoch": 3.796545105566219, "grad_norm": 1.3740454444434502, "learning_rate": 1.424458880535248e-05, "loss": 0.2505, "step": 5934 }, { "epoch": 3.7971849008317338, "grad_norm": 1.4501533445895867, "learning_rate": 1.4242712488185899e-05, "loss": 0.2572, "step": 5935 }, { "epoch": 3.797824696097249, "grad_norm": 1.5241905229398753, "learning_rate": 1.4240835988844794e-05, "loss": 0.24, "step": 5936 }, { "epoch": 3.798464491362764, "grad_norm": 1.6104256981351717, "learning_rate": 1.4238959307409732e-05, "loss": 0.274, "step": 5937 }, { "epoch": 3.799104286628279, "grad_norm": 1.4768148690174743, "learning_rate": 1.4237082443961301e-05, "loss": 0.2525, "step": 5938 }, { "epoch": 3.799744081893794, "grad_norm": 1.600782974419365, "learning_rate": 1.4235205398580085e-05, "loss": 0.2705, "step": 5939 }, { "epoch": 3.800383877159309, "grad_norm": 1.49286055656257, "learning_rate": 1.4233328171346686e-05, "loss": 0.2521, "step": 5940 }, { "epoch": 3.801023672424824, "grad_norm": 1.6060920237875893, "learning_rate": 1.4231450762341703e-05, "loss": 0.2515, "step": 5941 }, { "epoch": 3.8016634676903394, "grad_norm": 1.4157536000283635, "learning_rate": 1.4229573171645751e-05, "loss": 0.2328, "step": 5942 }, { "epoch": 3.802303262955854, "grad_norm": 1.5412883048683963, "learning_rate": 1.4227695399339454e-05, "loss": 0.2642, "step": 5943 }, { "epoch": 3.802943058221369, "grad_norm": 1.3971783451446904, "learning_rate": 1.4225817445503433e-05, "loss": 0.2435, "step": 5944 }, { "epoch": 3.8035828534868843, "grad_norm": 1.4528620899903741, "learning_rate": 1.4223939310218329e-05, "loss": 0.2561, "step": 5945 }, { "epoch": 3.804222648752399, "grad_norm": 1.3290591247029917, "learning_rate": 1.4222060993564788e-05, "loss": 0.2244, "step": 5946 }, { "epoch": 3.8048624440179144, "grad_norm": 1.4479347628402066, "learning_rate": 1.4220182495623454e-05, "loss": 0.2408, "step": 5947 }, { "epoch": 3.8055022392834292, "grad_norm": 1.4940813699354283, "learning_rate": 1.4218303816474994e-05, "loss": 0.2462, "step": 5948 }, { "epoch": 3.8061420345489445, "grad_norm": 1.475478268783757, "learning_rate": 1.4216424956200072e-05, "loss": 0.2454, "step": 5949 }, { "epoch": 3.8067818298144593, "grad_norm": 1.5324530568382577, "learning_rate": 1.421454591487936e-05, "loss": 0.2533, "step": 5950 }, { "epoch": 3.807421625079974, "grad_norm": 1.5698331276220465, "learning_rate": 1.4212666692593546e-05, "loss": 0.2482, "step": 5951 }, { "epoch": 3.8080614203454894, "grad_norm": 1.4807074031399101, "learning_rate": 1.4210787289423315e-05, "loss": 0.2581, "step": 5952 }, { "epoch": 3.8087012156110047, "grad_norm": 1.480291015399597, "learning_rate": 1.4208907705449374e-05, "loss": 0.2266, "step": 5953 }, { "epoch": 3.8093410108765196, "grad_norm": 1.4061995344579863, "learning_rate": 1.4207027940752418e-05, "loss": 0.2244, "step": 5954 }, { "epoch": 3.8099808061420344, "grad_norm": 1.39422202489375, "learning_rate": 1.420514799541317e-05, "loss": 0.2425, "step": 5955 }, { "epoch": 3.8106206014075497, "grad_norm": 1.4353391185316413, "learning_rate": 1.4203267869512343e-05, "loss": 0.2399, "step": 5956 }, { "epoch": 3.8112603966730645, "grad_norm": 1.3685383097746542, "learning_rate": 1.4201387563130673e-05, "loss": 0.217, "step": 5957 }, { "epoch": 3.8119001919385798, "grad_norm": 1.4757994099826826, "learning_rate": 1.4199507076348895e-05, "loss": 0.2495, "step": 5958 }, { "epoch": 3.8125399872040946, "grad_norm": 1.5072211630901755, "learning_rate": 1.4197626409247752e-05, "loss": 0.2552, "step": 5959 }, { "epoch": 3.81317978246961, "grad_norm": 1.589800280665043, "learning_rate": 1.4195745561907996e-05, "loss": 0.3056, "step": 5960 }, { "epoch": 3.8138195777351247, "grad_norm": 1.5459967462203683, "learning_rate": 1.4193864534410396e-05, "loss": 0.2644, "step": 5961 }, { "epoch": 3.8144593730006395, "grad_norm": 1.5719248038265128, "learning_rate": 1.4191983326835707e-05, "loss": 0.2499, "step": 5962 }, { "epoch": 3.815099168266155, "grad_norm": 1.450866638304806, "learning_rate": 1.4190101939264716e-05, "loss": 0.2299, "step": 5963 }, { "epoch": 3.81573896353167, "grad_norm": 1.4457374291012128, "learning_rate": 1.4188220371778199e-05, "loss": 0.2611, "step": 5964 }, { "epoch": 3.816378758797185, "grad_norm": 1.6113919308420697, "learning_rate": 1.4186338624456948e-05, "loss": 0.2794, "step": 5965 }, { "epoch": 3.8170185540626997, "grad_norm": 1.5845123296807169, "learning_rate": 1.4184456697381764e-05, "loss": 0.2552, "step": 5966 }, { "epoch": 3.817658349328215, "grad_norm": 1.5122638707131952, "learning_rate": 1.4182574590633453e-05, "loss": 0.2278, "step": 5967 }, { "epoch": 3.81829814459373, "grad_norm": 1.4635526878984526, "learning_rate": 1.4180692304292828e-05, "loss": 0.2568, "step": 5968 }, { "epoch": 3.818937939859245, "grad_norm": 1.4004890004025798, "learning_rate": 1.4178809838440715e-05, "loss": 0.2379, "step": 5969 }, { "epoch": 3.81957773512476, "grad_norm": 1.2641329840219429, "learning_rate": 1.417692719315794e-05, "loss": 0.2228, "step": 5970 }, { "epoch": 3.8202175303902752, "grad_norm": 1.3995683340011273, "learning_rate": 1.4175044368525344e-05, "loss": 0.2311, "step": 5971 }, { "epoch": 3.82085732565579, "grad_norm": 1.442437990213433, "learning_rate": 1.4173161364623767e-05, "loss": 0.2466, "step": 5972 }, { "epoch": 3.8214971209213053, "grad_norm": 1.4609715678579775, "learning_rate": 1.4171278181534066e-05, "loss": 0.2516, "step": 5973 }, { "epoch": 3.82213691618682, "grad_norm": 1.4858848980735002, "learning_rate": 1.4169394819337096e-05, "loss": 0.258, "step": 5974 }, { "epoch": 3.8227767114523354, "grad_norm": 1.407194291616144, "learning_rate": 1.4167511278113732e-05, "loss": 0.2332, "step": 5975 }, { "epoch": 3.8234165067178503, "grad_norm": 1.4660424919495683, "learning_rate": 1.4165627557944847e-05, "loss": 0.2566, "step": 5976 }, { "epoch": 3.824056301983365, "grad_norm": 1.4731568613932264, "learning_rate": 1.4163743658911325e-05, "loss": 0.2461, "step": 5977 }, { "epoch": 3.8246960972488804, "grad_norm": 1.5076120960250414, "learning_rate": 1.4161859581094057e-05, "loss": 0.2377, "step": 5978 }, { "epoch": 3.8253358925143957, "grad_norm": 1.4346686645885167, "learning_rate": 1.4159975324573942e-05, "loss": 0.2396, "step": 5979 }, { "epoch": 3.8259756877799105, "grad_norm": 1.4325749538453307, "learning_rate": 1.4158090889431888e-05, "loss": 0.2487, "step": 5980 }, { "epoch": 3.8266154830454253, "grad_norm": 1.4230430900780013, "learning_rate": 1.4156206275748808e-05, "loss": 0.2354, "step": 5981 }, { "epoch": 3.8272552783109406, "grad_norm": 1.2625238886388759, "learning_rate": 1.415432148360562e-05, "loss": 0.2104, "step": 5982 }, { "epoch": 3.8278950735764554, "grad_norm": 1.5184076043257633, "learning_rate": 1.4152436513083259e-05, "loss": 0.2468, "step": 5983 }, { "epoch": 3.8285348688419707, "grad_norm": 1.410860099346919, "learning_rate": 1.415055136426266e-05, "loss": 0.2371, "step": 5984 }, { "epoch": 3.8291746641074855, "grad_norm": 1.3818525692770447, "learning_rate": 1.4148666037224771e-05, "loss": 0.2421, "step": 5985 }, { "epoch": 3.829814459373001, "grad_norm": 1.6231948473382702, "learning_rate": 1.4146780532050544e-05, "loss": 0.2689, "step": 5986 }, { "epoch": 3.8304542546385156, "grad_norm": 1.5092365455935768, "learning_rate": 1.4144894848820933e-05, "loss": 0.2669, "step": 5987 }, { "epoch": 3.8310940499040305, "grad_norm": 1.4054578427035482, "learning_rate": 1.4143008987616913e-05, "loss": 0.2388, "step": 5988 }, { "epoch": 3.8317338451695457, "grad_norm": 1.5298976900732963, "learning_rate": 1.4141122948519454e-05, "loss": 0.2442, "step": 5989 }, { "epoch": 3.832373640435061, "grad_norm": 1.4914997084960393, "learning_rate": 1.4139236731609545e-05, "loss": 0.2522, "step": 5990 }, { "epoch": 3.833013435700576, "grad_norm": 1.4637143520495377, "learning_rate": 1.4137350336968175e-05, "loss": 0.2726, "step": 5991 }, { "epoch": 3.8336532309660907, "grad_norm": 1.3663235677669634, "learning_rate": 1.4135463764676338e-05, "loss": 0.243, "step": 5992 }, { "epoch": 3.834293026231606, "grad_norm": 1.4659615805561375, "learning_rate": 1.4133577014815045e-05, "loss": 0.259, "step": 5993 }, { "epoch": 3.834932821497121, "grad_norm": 1.345495981440185, "learning_rate": 1.4131690087465308e-05, "loss": 0.2369, "step": 5994 }, { "epoch": 3.835572616762636, "grad_norm": 1.3427029841777491, "learning_rate": 1.4129802982708147e-05, "loss": 0.2296, "step": 5995 }, { "epoch": 3.836212412028151, "grad_norm": 1.6037541395509738, "learning_rate": 1.4127915700624595e-05, "loss": 0.2759, "step": 5996 }, { "epoch": 3.836852207293666, "grad_norm": 1.5045955808655767, "learning_rate": 1.4126028241295683e-05, "loss": 0.2573, "step": 5997 }, { "epoch": 3.837492002559181, "grad_norm": 1.2995804328170713, "learning_rate": 1.4124140604802458e-05, "loss": 0.2028, "step": 5998 }, { "epoch": 3.838131797824696, "grad_norm": 1.5600753251718444, "learning_rate": 1.4122252791225974e-05, "loss": 0.2593, "step": 5999 }, { "epoch": 3.838771593090211, "grad_norm": 1.6264039627046754, "learning_rate": 1.4120364800647288e-05, "loss": 0.292, "step": 6000 }, { "epoch": 3.8394113883557264, "grad_norm": 1.5107925594366842, "learning_rate": 1.4118476633147467e-05, "loss": 0.2527, "step": 6001 }, { "epoch": 3.840051183621241, "grad_norm": 1.409765988257494, "learning_rate": 1.4116588288807584e-05, "loss": 0.2457, "step": 6002 }, { "epoch": 3.840690978886756, "grad_norm": 1.4831870694155496, "learning_rate": 1.4114699767708725e-05, "loss": 0.2313, "step": 6003 }, { "epoch": 3.8413307741522713, "grad_norm": 1.6421076053434644, "learning_rate": 1.4112811069931973e-05, "loss": 0.2697, "step": 6004 }, { "epoch": 3.8419705694177866, "grad_norm": 1.590847412685493, "learning_rate": 1.4110922195558433e-05, "loss": 0.2515, "step": 6005 }, { "epoch": 3.8426103646833014, "grad_norm": 1.4631262378571186, "learning_rate": 1.4109033144669206e-05, "loss": 0.2436, "step": 6006 }, { "epoch": 3.8432501599488162, "grad_norm": 1.396278093396846, "learning_rate": 1.4107143917345405e-05, "loss": 0.2436, "step": 6007 }, { "epoch": 3.8438899552143315, "grad_norm": 1.4881595135374266, "learning_rate": 1.410525451366815e-05, "loss": 0.2405, "step": 6008 }, { "epoch": 3.8445297504798464, "grad_norm": 1.4017854422928948, "learning_rate": 1.4103364933718571e-05, "loss": 0.2083, "step": 6009 }, { "epoch": 3.8451695457453616, "grad_norm": 1.5476296782661432, "learning_rate": 1.4101475177577801e-05, "loss": 0.241, "step": 6010 }, { "epoch": 3.8458093410108765, "grad_norm": 1.556992210734682, "learning_rate": 1.409958524532698e-05, "loss": 0.2697, "step": 6011 }, { "epoch": 3.8464491362763917, "grad_norm": 1.3823158827400572, "learning_rate": 1.4097695137047263e-05, "loss": 0.2334, "step": 6012 }, { "epoch": 3.8470889315419066, "grad_norm": 1.4404494285611738, "learning_rate": 1.4095804852819806e-05, "loss": 0.2324, "step": 6013 }, { "epoch": 3.8477287268074214, "grad_norm": 1.5710535682643072, "learning_rate": 1.4093914392725771e-05, "loss": 0.2471, "step": 6014 }, { "epoch": 3.8483685220729367, "grad_norm": 1.5646469356070516, "learning_rate": 1.4092023756846339e-05, "loss": 0.2699, "step": 6015 }, { "epoch": 3.849008317338452, "grad_norm": 1.366473397257186, "learning_rate": 1.4090132945262687e-05, "loss": 0.2403, "step": 6016 }, { "epoch": 3.8496481126039668, "grad_norm": 1.5138572641025314, "learning_rate": 1.4088241958055997e-05, "loss": 0.2539, "step": 6017 }, { "epoch": 3.8502879078694816, "grad_norm": 1.459186596092479, "learning_rate": 1.4086350795307474e-05, "loss": 0.2344, "step": 6018 }, { "epoch": 3.850927703134997, "grad_norm": 1.5103234345761394, "learning_rate": 1.4084459457098314e-05, "loss": 0.2565, "step": 6019 }, { "epoch": 3.8515674984005117, "grad_norm": 1.4876007908392526, "learning_rate": 1.4082567943509733e-05, "loss": 0.2348, "step": 6020 }, { "epoch": 3.852207293666027, "grad_norm": 1.5261662373471017, "learning_rate": 1.4080676254622945e-05, "loss": 0.2503, "step": 6021 }, { "epoch": 3.852847088931542, "grad_norm": 1.5513214298137048, "learning_rate": 1.4078784390519178e-05, "loss": 0.2701, "step": 6022 }, { "epoch": 3.853486884197057, "grad_norm": 1.5997352207822828, "learning_rate": 1.4076892351279665e-05, "loss": 0.2511, "step": 6023 }, { "epoch": 3.854126679462572, "grad_norm": 1.5251633568510825, "learning_rate": 1.4075000136985648e-05, "loss": 0.2581, "step": 6024 }, { "epoch": 3.8547664747280868, "grad_norm": 1.4874675253272223, "learning_rate": 1.4073107747718373e-05, "loss": 0.2394, "step": 6025 }, { "epoch": 3.855406269993602, "grad_norm": 1.5107925909169857, "learning_rate": 1.4071215183559095e-05, "loss": 0.2733, "step": 6026 }, { "epoch": 3.8560460652591173, "grad_norm": 1.5187765973432639, "learning_rate": 1.4069322444589082e-05, "loss": 0.2636, "step": 6027 }, { "epoch": 3.856685860524632, "grad_norm": 1.4926810796127965, "learning_rate": 1.40674295308896e-05, "loss": 0.2569, "step": 6028 }, { "epoch": 3.857325655790147, "grad_norm": 1.4483004937861286, "learning_rate": 1.406553644254193e-05, "loss": 0.2539, "step": 6029 }, { "epoch": 3.8579654510556622, "grad_norm": 1.4369265720293125, "learning_rate": 1.4063643179627358e-05, "loss": 0.2447, "step": 6030 }, { "epoch": 3.858605246321177, "grad_norm": 1.332768816118769, "learning_rate": 1.4061749742227177e-05, "loss": 0.2214, "step": 6031 }, { "epoch": 3.8592450415866923, "grad_norm": 1.6157399872518077, "learning_rate": 1.4059856130422684e-05, "loss": 0.2862, "step": 6032 }, { "epoch": 3.859884836852207, "grad_norm": 1.4333131894641198, "learning_rate": 1.4057962344295196e-05, "loss": 0.2423, "step": 6033 }, { "epoch": 3.8605246321177225, "grad_norm": 1.4417280127376197, "learning_rate": 1.4056068383926018e-05, "loss": 0.2362, "step": 6034 }, { "epoch": 3.8611644273832373, "grad_norm": 1.499352163530904, "learning_rate": 1.4054174249396482e-05, "loss": 0.2127, "step": 6035 }, { "epoch": 3.8618042226487526, "grad_norm": 1.6145002266808315, "learning_rate": 1.4052279940787914e-05, "loss": 0.2603, "step": 6036 }, { "epoch": 3.8624440179142674, "grad_norm": 1.4027235485917124, "learning_rate": 1.4050385458181652e-05, "loss": 0.2378, "step": 6037 }, { "epoch": 3.8630838131797827, "grad_norm": 1.4642065539249283, "learning_rate": 1.4048490801659048e-05, "loss": 0.2439, "step": 6038 }, { "epoch": 3.8637236084452975, "grad_norm": 1.55674948447412, "learning_rate": 1.404659597130145e-05, "loss": 0.2569, "step": 6039 }, { "epoch": 3.8643634037108123, "grad_norm": 1.4409964550961938, "learning_rate": 1.4044700967190218e-05, "loss": 0.2626, "step": 6040 }, { "epoch": 3.8650031989763276, "grad_norm": 1.4535662006712442, "learning_rate": 1.4042805789406722e-05, "loss": 0.2404, "step": 6041 }, { "epoch": 3.865642994241843, "grad_norm": 1.3235415275232028, "learning_rate": 1.4040910438032333e-05, "loss": 0.2208, "step": 6042 }, { "epoch": 3.8662827895073577, "grad_norm": 1.4858985947075993, "learning_rate": 1.4039014913148443e-05, "loss": 0.2121, "step": 6043 }, { "epoch": 3.8669225847728725, "grad_norm": 1.4288078521628238, "learning_rate": 1.4037119214836436e-05, "loss": 0.2324, "step": 6044 }, { "epoch": 3.867562380038388, "grad_norm": 1.5001673386038763, "learning_rate": 1.4035223343177713e-05, "loss": 0.2468, "step": 6045 }, { "epoch": 3.8682021753039026, "grad_norm": 1.4615811764659614, "learning_rate": 1.4033327298253676e-05, "loss": 0.2507, "step": 6046 }, { "epoch": 3.868841970569418, "grad_norm": 1.493503788931885, "learning_rate": 1.4031431080145738e-05, "loss": 0.2634, "step": 6047 }, { "epoch": 3.8694817658349328, "grad_norm": 1.4524207095668102, "learning_rate": 1.4029534688935325e-05, "loss": 0.2627, "step": 6048 }, { "epoch": 3.870121561100448, "grad_norm": 1.3939293989733554, "learning_rate": 1.4027638124703855e-05, "loss": 0.2085, "step": 6049 }, { "epoch": 3.870761356365963, "grad_norm": 1.5026502604940573, "learning_rate": 1.4025741387532775e-05, "loss": 0.2534, "step": 6050 }, { "epoch": 3.8714011516314777, "grad_norm": 1.592684012239907, "learning_rate": 1.4023844477503514e-05, "loss": 0.2642, "step": 6051 }, { "epoch": 3.872040946896993, "grad_norm": 1.4836248384008501, "learning_rate": 1.4021947394697533e-05, "loss": 0.2357, "step": 6052 }, { "epoch": 3.8726807421625082, "grad_norm": 1.4877522807212693, "learning_rate": 1.4020050139196287e-05, "loss": 0.2465, "step": 6053 }, { "epoch": 3.873320537428023, "grad_norm": 1.6434185968306103, "learning_rate": 1.4018152711081236e-05, "loss": 0.2847, "step": 6054 }, { "epoch": 3.873960332693538, "grad_norm": 1.48372793098881, "learning_rate": 1.4016255110433857e-05, "loss": 0.2559, "step": 6055 }, { "epoch": 3.874600127959053, "grad_norm": 1.4520109675137816, "learning_rate": 1.4014357337335626e-05, "loss": 0.2441, "step": 6056 }, { "epoch": 3.875239923224568, "grad_norm": 1.5028747670468836, "learning_rate": 1.4012459391868032e-05, "loss": 0.2504, "step": 6057 }, { "epoch": 3.8758797184900833, "grad_norm": 1.426179771726455, "learning_rate": 1.4010561274112573e-05, "loss": 0.2403, "step": 6058 }, { "epoch": 3.876519513755598, "grad_norm": 1.3978017540643828, "learning_rate": 1.4008662984150742e-05, "loss": 0.2476, "step": 6059 }, { "epoch": 3.8771593090211134, "grad_norm": 1.3907245811571238, "learning_rate": 1.4006764522064059e-05, "loss": 0.2111, "step": 6060 }, { "epoch": 3.877799104286628, "grad_norm": 1.338123852875934, "learning_rate": 1.4004865887934032e-05, "loss": 0.2398, "step": 6061 }, { "epoch": 3.878438899552143, "grad_norm": 1.4076493533212318, "learning_rate": 1.4002967081842184e-05, "loss": 0.2309, "step": 6062 }, { "epoch": 3.8790786948176583, "grad_norm": 1.3993099648530312, "learning_rate": 1.4001068103870058e-05, "loss": 0.2557, "step": 6063 }, { "epoch": 3.8797184900831736, "grad_norm": 1.3525728590536241, "learning_rate": 1.3999168954099178e-05, "loss": 0.2221, "step": 6064 }, { "epoch": 3.8803582853486884, "grad_norm": 1.4590181741703716, "learning_rate": 1.3997269632611101e-05, "loss": 0.2561, "step": 6065 }, { "epoch": 3.8809980806142033, "grad_norm": 1.4024068760670712, "learning_rate": 1.3995370139487374e-05, "loss": 0.2268, "step": 6066 }, { "epoch": 3.8816378758797185, "grad_norm": 1.5286178883019308, "learning_rate": 1.3993470474809562e-05, "loss": 0.2606, "step": 6067 }, { "epoch": 3.8822776711452334, "grad_norm": 1.5175974918788873, "learning_rate": 1.3991570638659232e-05, "loss": 0.272, "step": 6068 }, { "epoch": 3.8829174664107486, "grad_norm": 1.5374673980868552, "learning_rate": 1.398967063111796e-05, "loss": 0.2535, "step": 6069 }, { "epoch": 3.8835572616762635, "grad_norm": 1.4542870453376142, "learning_rate": 1.3987770452267328e-05, "loss": 0.2492, "step": 6070 }, { "epoch": 3.8841970569417787, "grad_norm": 1.3569706207693646, "learning_rate": 1.3985870102188925e-05, "loss": 0.2082, "step": 6071 }, { "epoch": 3.8848368522072936, "grad_norm": 1.394786318973871, "learning_rate": 1.398396958096435e-05, "loss": 0.2381, "step": 6072 }, { "epoch": 3.885476647472809, "grad_norm": 1.5522883363336912, "learning_rate": 1.3982068888675208e-05, "loss": 0.2574, "step": 6073 }, { "epoch": 3.8861164427383237, "grad_norm": 1.6381026807692434, "learning_rate": 1.398016802540311e-05, "loss": 0.2742, "step": 6074 }, { "epoch": 3.886756238003839, "grad_norm": 1.6023479702072472, "learning_rate": 1.3978266991229683e-05, "loss": 0.2777, "step": 6075 }, { "epoch": 3.887396033269354, "grad_norm": 1.440430150209178, "learning_rate": 1.3976365786236542e-05, "loss": 0.2375, "step": 6076 }, { "epoch": 3.8880358285348686, "grad_norm": 1.4402770333043422, "learning_rate": 1.397446441050533e-05, "loss": 0.2354, "step": 6077 }, { "epoch": 3.888675623800384, "grad_norm": 1.400832936278878, "learning_rate": 1.397256286411769e-05, "loss": 0.2454, "step": 6078 }, { "epoch": 3.889315419065899, "grad_norm": 1.4714802458008152, "learning_rate": 1.3970661147155263e-05, "loss": 0.2441, "step": 6079 }, { "epoch": 3.889955214331414, "grad_norm": 1.6140131277133964, "learning_rate": 1.3968759259699709e-05, "loss": 0.2352, "step": 6080 }, { "epoch": 3.890595009596929, "grad_norm": 1.4926256664961572, "learning_rate": 1.3966857201832694e-05, "loss": 0.22, "step": 6081 }, { "epoch": 3.891234804862444, "grad_norm": 1.431888934958207, "learning_rate": 1.3964954973635884e-05, "loss": 0.2399, "step": 6082 }, { "epoch": 3.891874600127959, "grad_norm": 1.7038586637152957, "learning_rate": 1.3963052575190963e-05, "loss": 0.3007, "step": 6083 }, { "epoch": 3.892514395393474, "grad_norm": 1.4095888037833468, "learning_rate": 1.3961150006579617e-05, "loss": 0.2111, "step": 6084 }, { "epoch": 3.893154190658989, "grad_norm": 1.5854758797752933, "learning_rate": 1.3959247267883533e-05, "loss": 0.2731, "step": 6085 }, { "epoch": 3.8937939859245043, "grad_norm": 1.493128273913369, "learning_rate": 1.3957344359184414e-05, "loss": 0.2635, "step": 6086 }, { "epoch": 3.894433781190019, "grad_norm": 1.442347458858689, "learning_rate": 1.395544128056397e-05, "loss": 0.2614, "step": 6087 }, { "epoch": 3.895073576455534, "grad_norm": 1.436781563930465, "learning_rate": 1.395353803210391e-05, "loss": 0.2358, "step": 6088 }, { "epoch": 3.8957133717210493, "grad_norm": 1.493042780716637, "learning_rate": 1.395163461388596e-05, "loss": 0.2192, "step": 6089 }, { "epoch": 3.8963531669865645, "grad_norm": 1.4599760972328992, "learning_rate": 1.3949731025991853e-05, "loss": 0.2597, "step": 6090 }, { "epoch": 3.8969929622520794, "grad_norm": 1.3665184248202067, "learning_rate": 1.3947827268503323e-05, "loss": 0.2176, "step": 6091 }, { "epoch": 3.897632757517594, "grad_norm": 1.360121362708436, "learning_rate": 1.3945923341502108e-05, "loss": 0.2069, "step": 6092 }, { "epoch": 3.8982725527831095, "grad_norm": 1.498904242301008, "learning_rate": 1.3944019245069963e-05, "loss": 0.2296, "step": 6093 }, { "epoch": 3.8989123480486243, "grad_norm": 1.4171426404683183, "learning_rate": 1.394211497928865e-05, "loss": 0.2433, "step": 6094 }, { "epoch": 3.8995521433141396, "grad_norm": 1.4890815910849664, "learning_rate": 1.3940210544239934e-05, "loss": 0.2452, "step": 6095 }, { "epoch": 3.9001919385796544, "grad_norm": 1.5713996885065553, "learning_rate": 1.3938305940005583e-05, "loss": 0.2639, "step": 6096 }, { "epoch": 3.9008317338451697, "grad_norm": 1.5918509103556369, "learning_rate": 1.3936401166667382e-05, "loss": 0.2453, "step": 6097 }, { "epoch": 3.9014715291106845, "grad_norm": 1.629562678653458, "learning_rate": 1.3934496224307117e-05, "loss": 0.2877, "step": 6098 }, { "epoch": 3.9021113243761993, "grad_norm": 1.4447583500118737, "learning_rate": 1.3932591113006584e-05, "loss": 0.2164, "step": 6099 }, { "epoch": 3.9027511196417146, "grad_norm": 1.3746528470840205, "learning_rate": 1.3930685832847584e-05, "loss": 0.2241, "step": 6100 }, { "epoch": 3.90339091490723, "grad_norm": 1.770756350572624, "learning_rate": 1.3928780383911927e-05, "loss": 0.3013, "step": 6101 }, { "epoch": 3.9040307101727447, "grad_norm": 1.3882704465653875, "learning_rate": 1.3926874766281428e-05, "loss": 0.232, "step": 6102 }, { "epoch": 3.9046705054382596, "grad_norm": 1.3131955328614007, "learning_rate": 1.392496898003791e-05, "loss": 0.2057, "step": 6103 }, { "epoch": 3.905310300703775, "grad_norm": 1.3523983334548841, "learning_rate": 1.3923063025263206e-05, "loss": 0.2163, "step": 6104 }, { "epoch": 3.90595009596929, "grad_norm": 1.3854515815942727, "learning_rate": 1.392115690203916e-05, "loss": 0.2284, "step": 6105 }, { "epoch": 3.906589891234805, "grad_norm": 1.3750171374687066, "learning_rate": 1.3919250610447608e-05, "loss": 0.2254, "step": 6106 }, { "epoch": 3.9072296865003198, "grad_norm": 1.4942140656346448, "learning_rate": 1.3917344150570406e-05, "loss": 0.2289, "step": 6107 }, { "epoch": 3.907869481765835, "grad_norm": 1.5277767208279507, "learning_rate": 1.3915437522489414e-05, "loss": 0.2399, "step": 6108 }, { "epoch": 3.90850927703135, "grad_norm": 1.4657980941984765, "learning_rate": 1.3913530726286502e-05, "loss": 0.2186, "step": 6109 }, { "epoch": 3.909149072296865, "grad_norm": 1.5193318296124911, "learning_rate": 1.391162376204354e-05, "loss": 0.2269, "step": 6110 }, { "epoch": 3.90978886756238, "grad_norm": 1.4091680170872523, "learning_rate": 1.3909716629842414e-05, "loss": 0.2253, "step": 6111 }, { "epoch": 3.9104286628278953, "grad_norm": 1.4085235460556278, "learning_rate": 1.390780932976501e-05, "loss": 0.2132, "step": 6112 }, { "epoch": 3.91106845809341, "grad_norm": 1.4639056243434354, "learning_rate": 1.3905901861893223e-05, "loss": 0.2356, "step": 6113 }, { "epoch": 3.911708253358925, "grad_norm": 1.4635775914091214, "learning_rate": 1.3903994226308961e-05, "loss": 0.2571, "step": 6114 }, { "epoch": 3.91234804862444, "grad_norm": 1.4346152285115146, "learning_rate": 1.3902086423094131e-05, "loss": 0.2473, "step": 6115 }, { "epoch": 3.9129878438899555, "grad_norm": 1.392449192761787, "learning_rate": 1.390017845233065e-05, "loss": 0.241, "step": 6116 }, { "epoch": 3.9136276391554703, "grad_norm": 1.5217557759237312, "learning_rate": 1.3898270314100445e-05, "loss": 0.26, "step": 6117 }, { "epoch": 3.914267434420985, "grad_norm": 1.4443562683168478, "learning_rate": 1.3896362008485446e-05, "loss": 0.2397, "step": 6118 }, { "epoch": 3.9149072296865004, "grad_norm": 1.4928444596340564, "learning_rate": 1.3894453535567593e-05, "loss": 0.2505, "step": 6119 }, { "epoch": 3.9155470249520152, "grad_norm": 1.4048221429814378, "learning_rate": 1.3892544895428837e-05, "loss": 0.2177, "step": 6120 }, { "epoch": 3.9161868202175305, "grad_norm": 1.5065107442377628, "learning_rate": 1.3890636088151126e-05, "loss": 0.2484, "step": 6121 }, { "epoch": 3.9168266154830453, "grad_norm": 1.5246479828412622, "learning_rate": 1.3888727113816422e-05, "loss": 0.2478, "step": 6122 }, { "epoch": 3.9174664107485606, "grad_norm": 1.612510685747209, "learning_rate": 1.3886817972506693e-05, "loss": 0.2588, "step": 6123 }, { "epoch": 3.9181062060140754, "grad_norm": 1.405013311969959, "learning_rate": 1.388490866430391e-05, "loss": 0.2325, "step": 6124 }, { "epoch": 3.9187460012795903, "grad_norm": 1.4793845608479115, "learning_rate": 1.3882999189290066e-05, "loss": 0.2398, "step": 6125 }, { "epoch": 3.9193857965451055, "grad_norm": 1.4698789088939124, "learning_rate": 1.388108954754714e-05, "loss": 0.2499, "step": 6126 }, { "epoch": 3.920025591810621, "grad_norm": 1.5155434385634732, "learning_rate": 1.3879179739157133e-05, "loss": 0.24, "step": 6127 }, { "epoch": 3.9206653870761357, "grad_norm": 1.431516483385826, "learning_rate": 1.387726976420205e-05, "loss": 0.2266, "step": 6128 }, { "epoch": 3.9213051823416505, "grad_norm": 1.490463705726812, "learning_rate": 1.3875359622763902e-05, "loss": 0.266, "step": 6129 }, { "epoch": 3.9219449776071658, "grad_norm": 1.4338616078822612, "learning_rate": 1.3873449314924706e-05, "loss": 0.242, "step": 6130 }, { "epoch": 3.9225847728726806, "grad_norm": 1.3964535658990787, "learning_rate": 1.387153884076648e-05, "loss": 0.232, "step": 6131 }, { "epoch": 3.923224568138196, "grad_norm": 1.4222796364778467, "learning_rate": 1.386962820037127e-05, "loss": 0.224, "step": 6132 }, { "epoch": 3.9238643634037107, "grad_norm": 1.458419681343468, "learning_rate": 1.3867717393821105e-05, "loss": 0.2515, "step": 6133 }, { "epoch": 3.924504158669226, "grad_norm": 1.530062540479002, "learning_rate": 1.3865806421198036e-05, "loss": 0.2563, "step": 6134 }, { "epoch": 3.925143953934741, "grad_norm": 1.377141778610655, "learning_rate": 1.3863895282584117e-05, "loss": 0.2229, "step": 6135 }, { "epoch": 3.925783749200256, "grad_norm": 1.6283769365961331, "learning_rate": 1.3861983978061408e-05, "loss": 0.2515, "step": 6136 }, { "epoch": 3.926423544465771, "grad_norm": 1.5916572058783047, "learning_rate": 1.3860072507711979e-05, "loss": 0.2695, "step": 6137 }, { "epoch": 3.927063339731286, "grad_norm": 1.5972972305165682, "learning_rate": 1.3858160871617898e-05, "loss": 0.2812, "step": 6138 }, { "epoch": 3.927703134996801, "grad_norm": 1.652336749733201, "learning_rate": 1.3856249069861253e-05, "loss": 0.248, "step": 6139 }, { "epoch": 3.928342930262316, "grad_norm": 1.5268500223412644, "learning_rate": 1.3854337102524138e-05, "loss": 0.2463, "step": 6140 }, { "epoch": 3.928982725527831, "grad_norm": 1.5392915673513656, "learning_rate": 1.3852424969688638e-05, "loss": 0.2384, "step": 6141 }, { "epoch": 3.9296225207933464, "grad_norm": 1.4874766545543576, "learning_rate": 1.3850512671436866e-05, "loss": 0.2608, "step": 6142 }, { "epoch": 3.9302623160588612, "grad_norm": 1.3917849629988963, "learning_rate": 1.3848600207850928e-05, "loss": 0.2177, "step": 6143 }, { "epoch": 3.930902111324376, "grad_norm": 1.4787038989736832, "learning_rate": 1.3846687579012946e-05, "loss": 0.2424, "step": 6144 }, { "epoch": 3.9315419065898913, "grad_norm": 1.6374294379542385, "learning_rate": 1.384477478500504e-05, "loss": 0.2713, "step": 6145 }, { "epoch": 3.932181701855406, "grad_norm": 1.4559718399953394, "learning_rate": 1.3842861825909344e-05, "loss": 0.2385, "step": 6146 }, { "epoch": 3.9328214971209214, "grad_norm": 1.4700790240370476, "learning_rate": 1.3840948701807998e-05, "loss": 0.2357, "step": 6147 }, { "epoch": 3.9334612923864363, "grad_norm": 1.5603336207961829, "learning_rate": 1.3839035412783143e-05, "loss": 0.2651, "step": 6148 }, { "epoch": 3.9341010876519515, "grad_norm": 1.514554074688383, "learning_rate": 1.3837121958916939e-05, "loss": 0.2492, "step": 6149 }, { "epoch": 3.9347408829174664, "grad_norm": 1.343824659808503, "learning_rate": 1.3835208340291546e-05, "loss": 0.2177, "step": 6150 }, { "epoch": 3.935380678182981, "grad_norm": 1.5982219912083222, "learning_rate": 1.3833294556989129e-05, "loss": 0.2602, "step": 6151 }, { "epoch": 3.9360204734484965, "grad_norm": 1.4537175484279505, "learning_rate": 1.3831380609091863e-05, "loss": 0.2598, "step": 6152 }, { "epoch": 3.9366602687140118, "grad_norm": 1.3466116401412618, "learning_rate": 1.3829466496681927e-05, "loss": 0.2115, "step": 6153 }, { "epoch": 3.9373000639795266, "grad_norm": 1.4309771337480468, "learning_rate": 1.3827552219841511e-05, "loss": 0.2525, "step": 6154 }, { "epoch": 3.9379398592450414, "grad_norm": 1.6673948648816916, "learning_rate": 1.3825637778652814e-05, "loss": 0.2803, "step": 6155 }, { "epoch": 3.9385796545105567, "grad_norm": 1.4548776323877293, "learning_rate": 1.3823723173198033e-05, "loss": 0.232, "step": 6156 }, { "epoch": 3.9392194497760715, "grad_norm": 1.5325274415114891, "learning_rate": 1.3821808403559382e-05, "loss": 0.2498, "step": 6157 }, { "epoch": 3.939859245041587, "grad_norm": 1.5529915793833393, "learning_rate": 1.3819893469819081e-05, "loss": 0.2828, "step": 6158 }, { "epoch": 3.9404990403071016, "grad_norm": 1.41528307025705, "learning_rate": 1.3817978372059344e-05, "loss": 0.2339, "step": 6159 }, { "epoch": 3.941138835572617, "grad_norm": 1.4842990157933733, "learning_rate": 1.381606311036241e-05, "loss": 0.2693, "step": 6160 }, { "epoch": 3.9417786308381317, "grad_norm": 1.5340515656134321, "learning_rate": 1.3814147684810512e-05, "loss": 0.2536, "step": 6161 }, { "epoch": 3.9424184261036466, "grad_norm": 1.4979765227540562, "learning_rate": 1.38122320954859e-05, "loss": 0.2701, "step": 6162 }, { "epoch": 3.943058221369162, "grad_norm": 1.2928240389513566, "learning_rate": 1.3810316342470819e-05, "loss": 0.2107, "step": 6163 }, { "epoch": 3.943698016634677, "grad_norm": 1.4604979274619032, "learning_rate": 1.3808400425847535e-05, "loss": 0.2309, "step": 6164 }, { "epoch": 3.944337811900192, "grad_norm": 1.3518531521051484, "learning_rate": 1.3806484345698312e-05, "loss": 0.2423, "step": 6165 }, { "epoch": 3.944977607165707, "grad_norm": 1.5124932065519259, "learning_rate": 1.3804568102105425e-05, "loss": 0.2437, "step": 6166 }, { "epoch": 3.945617402431222, "grad_norm": 1.51242552697997, "learning_rate": 1.3802651695151148e-05, "loss": 0.2552, "step": 6167 }, { "epoch": 3.946257197696737, "grad_norm": 1.4104330850840583, "learning_rate": 1.380073512491777e-05, "loss": 0.248, "step": 6168 }, { "epoch": 3.946896992962252, "grad_norm": 1.4359324666531585, "learning_rate": 1.3798818391487591e-05, "loss": 0.2476, "step": 6169 }, { "epoch": 3.947536788227767, "grad_norm": 1.369900961428668, "learning_rate": 1.3796901494942904e-05, "loss": 0.2504, "step": 6170 }, { "epoch": 3.9481765834932823, "grad_norm": 1.451842321714788, "learning_rate": 1.3794984435366021e-05, "loss": 0.241, "step": 6171 }, { "epoch": 3.948816378758797, "grad_norm": 1.4612248494595974, "learning_rate": 1.3793067212839259e-05, "loss": 0.2511, "step": 6172 }, { "epoch": 3.9494561740243124, "grad_norm": 1.467544711882886, "learning_rate": 1.379114982744494e-05, "loss": 0.2281, "step": 6173 }, { "epoch": 3.950095969289827, "grad_norm": 1.3937402041259814, "learning_rate": 1.378923227926539e-05, "loss": 0.2408, "step": 6174 }, { "epoch": 3.9507357645553425, "grad_norm": 1.4743759256568412, "learning_rate": 1.3787314568382946e-05, "loss": 0.2344, "step": 6175 }, { "epoch": 3.9513755598208573, "grad_norm": 1.5121579592192684, "learning_rate": 1.3785396694879949e-05, "loss": 0.2314, "step": 6176 }, { "epoch": 3.952015355086372, "grad_norm": 1.3856392578389187, "learning_rate": 1.3783478658838753e-05, "loss": 0.2126, "step": 6177 }, { "epoch": 3.9526551503518874, "grad_norm": 1.5543475037493417, "learning_rate": 1.3781560460341715e-05, "loss": 0.2581, "step": 6178 }, { "epoch": 3.9532949456174027, "grad_norm": 1.6185455633031607, "learning_rate": 1.3779642099471198e-05, "loss": 0.2627, "step": 6179 }, { "epoch": 3.9539347408829175, "grad_norm": 1.625919654891878, "learning_rate": 1.3777723576309572e-05, "loss": 0.2781, "step": 6180 }, { "epoch": 3.9545745361484324, "grad_norm": 1.3796367516390324, "learning_rate": 1.3775804890939216e-05, "loss": 0.2209, "step": 6181 }, { "epoch": 3.9552143314139476, "grad_norm": 1.3709701878669946, "learning_rate": 1.3773886043442514e-05, "loss": 0.2262, "step": 6182 }, { "epoch": 3.9558541266794625, "grad_norm": 1.3813119708228516, "learning_rate": 1.3771967033901858e-05, "loss": 0.2497, "step": 6183 }, { "epoch": 3.9564939219449777, "grad_norm": 1.355743614335941, "learning_rate": 1.3770047862399648e-05, "loss": 0.2393, "step": 6184 }, { "epoch": 3.9571337172104926, "grad_norm": 1.4029321000326074, "learning_rate": 1.3768128529018289e-05, "loss": 0.2453, "step": 6185 }, { "epoch": 3.957773512476008, "grad_norm": 1.3879244437202027, "learning_rate": 1.3766209033840193e-05, "loss": 0.2224, "step": 6186 }, { "epoch": 3.9584133077415227, "grad_norm": 1.5077117942070337, "learning_rate": 1.3764289376947784e-05, "loss": 0.2675, "step": 6187 }, { "epoch": 3.9590531030070375, "grad_norm": 1.3262323035348242, "learning_rate": 1.3762369558423483e-05, "loss": 0.2159, "step": 6188 }, { "epoch": 3.9596928982725528, "grad_norm": 1.3981657394192482, "learning_rate": 1.3760449578349728e-05, "loss": 0.2098, "step": 6189 }, { "epoch": 3.960332693538068, "grad_norm": 1.4768364216105592, "learning_rate": 1.3758529436808955e-05, "loss": 0.2556, "step": 6190 }, { "epoch": 3.960972488803583, "grad_norm": 1.3921991443961668, "learning_rate": 1.3756609133883611e-05, "loss": 0.229, "step": 6191 }, { "epoch": 3.9616122840690977, "grad_norm": 1.5738359713411578, "learning_rate": 1.375468866965616e-05, "loss": 0.2421, "step": 6192 }, { "epoch": 3.962252079334613, "grad_norm": 1.4850823851959087, "learning_rate": 1.375276804420905e-05, "loss": 0.2515, "step": 6193 }, { "epoch": 3.962891874600128, "grad_norm": 1.56044482776237, "learning_rate": 1.3750847257624762e-05, "loss": 0.2741, "step": 6194 }, { "epoch": 3.963531669865643, "grad_norm": 1.5084515762041972, "learning_rate": 1.374892630998576e-05, "loss": 0.2426, "step": 6195 }, { "epoch": 3.964171465131158, "grad_norm": 1.4174029044967857, "learning_rate": 1.3747005201374535e-05, "loss": 0.2692, "step": 6196 }, { "epoch": 3.964811260396673, "grad_norm": 1.4300980561028571, "learning_rate": 1.3745083931873572e-05, "loss": 0.2325, "step": 6197 }, { "epoch": 3.965451055662188, "grad_norm": 1.4198979435022594, "learning_rate": 1.3743162501565364e-05, "loss": 0.2539, "step": 6198 }, { "epoch": 3.966090850927703, "grad_norm": 1.3876821787302174, "learning_rate": 1.3741240910532416e-05, "loss": 0.2334, "step": 6199 }, { "epoch": 3.966730646193218, "grad_norm": 1.4272726542342438, "learning_rate": 1.373931915885724e-05, "loss": 0.2548, "step": 6200 }, { "epoch": 3.9673704414587334, "grad_norm": 1.4295717328439113, "learning_rate": 1.373739724662235e-05, "loss": 0.2371, "step": 6201 }, { "epoch": 3.9680102367242482, "grad_norm": 1.304326680749786, "learning_rate": 1.3735475173910273e-05, "loss": 0.2214, "step": 6202 }, { "epoch": 3.968650031989763, "grad_norm": 1.4050965183988167, "learning_rate": 1.3733552940803539e-05, "loss": 0.2395, "step": 6203 }, { "epoch": 3.9692898272552783, "grad_norm": 1.5111968463112055, "learning_rate": 1.373163054738468e-05, "loss": 0.2687, "step": 6204 }, { "epoch": 3.9699296225207936, "grad_norm": 1.3792350678109735, "learning_rate": 1.3729707993736242e-05, "loss": 0.2277, "step": 6205 }, { "epoch": 3.9705694177863085, "grad_norm": 1.5147984574922153, "learning_rate": 1.3727785279940777e-05, "loss": 0.2457, "step": 6206 }, { "epoch": 3.9712092130518233, "grad_norm": 1.452794712093282, "learning_rate": 1.3725862406080847e-05, "loss": 0.2514, "step": 6207 }, { "epoch": 3.9718490083173386, "grad_norm": 1.6785731572288862, "learning_rate": 1.372393937223901e-05, "loss": 0.2806, "step": 6208 }, { "epoch": 3.9724888035828534, "grad_norm": 1.5471089152986413, "learning_rate": 1.3722016178497844e-05, "loss": 0.2203, "step": 6209 }, { "epoch": 3.9731285988483687, "grad_norm": 1.539383193669903, "learning_rate": 1.3720092824939921e-05, "loss": 0.2426, "step": 6210 }, { "epoch": 3.9737683941138835, "grad_norm": 1.661025499607737, "learning_rate": 1.3718169311647832e-05, "loss": 0.2667, "step": 6211 }, { "epoch": 3.9744081893793988, "grad_norm": 1.5496759007832521, "learning_rate": 1.3716245638704166e-05, "loss": 0.2555, "step": 6212 }, { "epoch": 3.9750479846449136, "grad_norm": 1.5189989522772838, "learning_rate": 1.3714321806191522e-05, "loss": 0.2542, "step": 6213 }, { "epoch": 3.9756877799104284, "grad_norm": 1.3978552835832943, "learning_rate": 1.3712397814192509e-05, "loss": 0.2337, "step": 6214 }, { "epoch": 3.9763275751759437, "grad_norm": 1.510643262377595, "learning_rate": 1.3710473662789734e-05, "loss": 0.2804, "step": 6215 }, { "epoch": 3.976967370441459, "grad_norm": 1.3795242022672871, "learning_rate": 1.370854935206582e-05, "loss": 0.2331, "step": 6216 }, { "epoch": 3.977607165706974, "grad_norm": 1.5176913718610816, "learning_rate": 1.3706624882103399e-05, "loss": 0.2359, "step": 6217 }, { "epoch": 3.9782469609724886, "grad_norm": 1.3729830608255658, "learning_rate": 1.3704700252985097e-05, "loss": 0.2289, "step": 6218 }, { "epoch": 3.978886756238004, "grad_norm": 1.505166832761831, "learning_rate": 1.3702775464793558e-05, "loss": 0.2477, "step": 6219 }, { "epoch": 3.9795265515035187, "grad_norm": 1.6003475688265105, "learning_rate": 1.3700850517611423e-05, "loss": 0.2605, "step": 6220 }, { "epoch": 3.980166346769034, "grad_norm": 1.5768637984424587, "learning_rate": 1.369892541152135e-05, "loss": 0.2622, "step": 6221 }, { "epoch": 3.980806142034549, "grad_norm": 1.4807925986347947, "learning_rate": 1.3697000146606004e-05, "loss": 0.26, "step": 6222 }, { "epoch": 3.981445937300064, "grad_norm": 1.4249780055386951, "learning_rate": 1.3695074722948046e-05, "loss": 0.2333, "step": 6223 }, { "epoch": 3.982085732565579, "grad_norm": 1.4604002503748914, "learning_rate": 1.3693149140630153e-05, "loss": 0.2623, "step": 6224 }, { "epoch": 3.982725527831094, "grad_norm": 1.4255056678008438, "learning_rate": 1.3691223399735004e-05, "loss": 0.2406, "step": 6225 }, { "epoch": 3.983365323096609, "grad_norm": 1.4480678427818123, "learning_rate": 1.3689297500345289e-05, "loss": 0.2503, "step": 6226 }, { "epoch": 3.9840051183621243, "grad_norm": 1.4732013658946088, "learning_rate": 1.3687371442543704e-05, "loss": 0.2247, "step": 6227 }, { "epoch": 3.984644913627639, "grad_norm": 1.2777237689988628, "learning_rate": 1.3685445226412946e-05, "loss": 0.226, "step": 6228 }, { "epoch": 3.985284708893154, "grad_norm": 1.408662796226623, "learning_rate": 1.3683518852035728e-05, "loss": 0.2167, "step": 6229 }, { "epoch": 3.9859245041586693, "grad_norm": 1.4709177644949956, "learning_rate": 1.368159231949476e-05, "loss": 0.2217, "step": 6230 }, { "epoch": 3.986564299424184, "grad_norm": 1.6545223516650838, "learning_rate": 1.3679665628872769e-05, "loss": 0.255, "step": 6231 }, { "epoch": 3.9872040946896994, "grad_norm": 1.4832045643021077, "learning_rate": 1.3677738780252484e-05, "loss": 0.2692, "step": 6232 }, { "epoch": 3.987843889955214, "grad_norm": 1.4693727947156223, "learning_rate": 1.3675811773716634e-05, "loss": 0.2503, "step": 6233 }, { "epoch": 3.9884836852207295, "grad_norm": 1.4212355770980065, "learning_rate": 1.3673884609347968e-05, "loss": 0.2438, "step": 6234 }, { "epoch": 3.9891234804862443, "grad_norm": 1.4532860198698292, "learning_rate": 1.367195728722923e-05, "loss": 0.2422, "step": 6235 }, { "epoch": 3.9897632757517596, "grad_norm": 1.4281561931097189, "learning_rate": 1.3670029807443177e-05, "loss": 0.2564, "step": 6236 }, { "epoch": 3.9904030710172744, "grad_norm": 1.5229988712326783, "learning_rate": 1.3668102170072575e-05, "loss": 0.2709, "step": 6237 }, { "epoch": 3.9910428662827897, "grad_norm": 1.3991162072003311, "learning_rate": 1.3666174375200188e-05, "loss": 0.2257, "step": 6238 }, { "epoch": 3.9916826615483045, "grad_norm": 1.3759521984745122, "learning_rate": 1.3664246422908798e-05, "loss": 0.2392, "step": 6239 }, { "epoch": 3.9923224568138194, "grad_norm": 1.5554571623383657, "learning_rate": 1.3662318313281184e-05, "loss": 0.2387, "step": 6240 }, { "epoch": 3.9929622520793346, "grad_norm": 1.430309987195469, "learning_rate": 1.3660390046400136e-05, "loss": 0.2313, "step": 6241 }, { "epoch": 3.99360204734485, "grad_norm": 1.5550696844557126, "learning_rate": 1.365846162234845e-05, "loss": 0.2635, "step": 6242 }, { "epoch": 3.9942418426103647, "grad_norm": 1.4970432286620046, "learning_rate": 1.3656533041208933e-05, "loss": 0.2346, "step": 6243 }, { "epoch": 3.9948816378758796, "grad_norm": 1.493407495175054, "learning_rate": 1.365460430306439e-05, "loss": 0.2296, "step": 6244 }, { "epoch": 3.995521433141395, "grad_norm": 1.3444104011326763, "learning_rate": 1.3652675407997635e-05, "loss": 0.2313, "step": 6245 }, { "epoch": 3.9961612284069097, "grad_norm": 1.5271202714996803, "learning_rate": 1.3650746356091503e-05, "loss": 0.2533, "step": 6246 }, { "epoch": 3.996801023672425, "grad_norm": 1.4002018094027366, "learning_rate": 1.364881714742881e-05, "loss": 0.2508, "step": 6247 }, { "epoch": 3.99744081893794, "grad_norm": 1.5643598995995358, "learning_rate": 1.3646887782092404e-05, "loss": 0.2519, "step": 6248 }, { "epoch": 3.998080614203455, "grad_norm": 1.4000691311291358, "learning_rate": 1.3644958260165124e-05, "loss": 0.2357, "step": 6249 }, { "epoch": 3.99872040946897, "grad_norm": 1.4695475551317425, "learning_rate": 1.3643028581729815e-05, "loss": 0.2699, "step": 6250 }, { "epoch": 3.9993602047344847, "grad_norm": 1.3058062854766184, "learning_rate": 1.3641098746869346e-05, "loss": 0.2264, "step": 6251 }, { "epoch": 4.0, "grad_norm": 1.3252750581543637, "learning_rate": 1.3639168755666567e-05, "loss": 0.1968, "step": 6252 }, { "epoch": 4.000639795265515, "grad_norm": 1.1324976032704726, "learning_rate": 1.363723860820436e-05, "loss": 0.1708, "step": 6253 }, { "epoch": 4.00127959053103, "grad_norm": 1.188705072400341, "learning_rate": 1.3635308304565595e-05, "loss": 0.1665, "step": 6254 }, { "epoch": 4.001919385796545, "grad_norm": 1.1295240422096857, "learning_rate": 1.3633377844833158e-05, "loss": 0.1746, "step": 6255 }, { "epoch": 4.00255918106206, "grad_norm": 1.1624292247635768, "learning_rate": 1.3631447229089942e-05, "loss": 0.1698, "step": 6256 }, { "epoch": 4.0031989763275755, "grad_norm": 1.1476897844224692, "learning_rate": 1.3629516457418843e-05, "loss": 0.1643, "step": 6257 }, { "epoch": 4.00383877159309, "grad_norm": 1.2547195350915727, "learning_rate": 1.3627585529902759e-05, "loss": 0.1575, "step": 6258 }, { "epoch": 4.004478566858605, "grad_norm": 1.286316056855629, "learning_rate": 1.3625654446624608e-05, "loss": 0.1573, "step": 6259 }, { "epoch": 4.00511836212412, "grad_norm": 1.406227227355755, "learning_rate": 1.3623723207667302e-05, "loss": 0.1636, "step": 6260 }, { "epoch": 4.005758157389636, "grad_norm": 1.4086802920732082, "learning_rate": 1.3621791813113771e-05, "loss": 0.1573, "step": 6261 }, { "epoch": 4.00639795265515, "grad_norm": 1.3238339809896849, "learning_rate": 1.361986026304694e-05, "loss": 0.1693, "step": 6262 }, { "epoch": 4.007037747920665, "grad_norm": 1.254674904903472, "learning_rate": 1.3617928557549751e-05, "loss": 0.1521, "step": 6263 }, { "epoch": 4.007677543186181, "grad_norm": 1.327141869490503, "learning_rate": 1.3615996696705145e-05, "loss": 0.1637, "step": 6264 }, { "epoch": 4.008317338451695, "grad_norm": 1.3176603874186723, "learning_rate": 1.3614064680596073e-05, "loss": 0.1544, "step": 6265 }, { "epoch": 4.00895713371721, "grad_norm": 1.2334914344742969, "learning_rate": 1.361213250930549e-05, "loss": 0.1653, "step": 6266 }, { "epoch": 4.009596928982726, "grad_norm": 1.2299366045550633, "learning_rate": 1.3610200182916364e-05, "loss": 0.1519, "step": 6267 }, { "epoch": 4.010236724248241, "grad_norm": 1.2056515883766221, "learning_rate": 1.3608267701511663e-05, "loss": 0.1516, "step": 6268 }, { "epoch": 4.010876519513755, "grad_norm": 1.1818569090889053, "learning_rate": 1.360633506517437e-05, "loss": 0.167, "step": 6269 }, { "epoch": 4.0115163147792705, "grad_norm": 1.2105124386591095, "learning_rate": 1.3604402273987459e-05, "loss": 0.1608, "step": 6270 }, { "epoch": 4.012156110044786, "grad_norm": 1.2974548444854428, "learning_rate": 1.360246932803393e-05, "loss": 0.1646, "step": 6271 }, { "epoch": 4.012795905310301, "grad_norm": 1.263383680848024, "learning_rate": 1.3600536227396774e-05, "loss": 0.1639, "step": 6272 }, { "epoch": 4.0134357005758154, "grad_norm": 1.2727892404099221, "learning_rate": 1.3598602972158998e-05, "loss": 0.1641, "step": 6273 }, { "epoch": 4.014075495841331, "grad_norm": 1.2177549211889689, "learning_rate": 1.3596669562403616e-05, "loss": 0.167, "step": 6274 }, { "epoch": 4.014715291106846, "grad_norm": 1.1759945906502969, "learning_rate": 1.3594735998213634e-05, "loss": 0.1537, "step": 6275 }, { "epoch": 4.015355086372361, "grad_norm": 1.223979212747409, "learning_rate": 1.3592802279672087e-05, "loss": 0.1547, "step": 6276 }, { "epoch": 4.015994881637876, "grad_norm": 1.1959061766429735, "learning_rate": 1.3590868406862e-05, "loss": 0.1621, "step": 6277 }, { "epoch": 4.016634676903391, "grad_norm": 1.209069344888269, "learning_rate": 1.358893437986641e-05, "loss": 0.1553, "step": 6278 }, { "epoch": 4.017274472168906, "grad_norm": 1.1729381124718512, "learning_rate": 1.3587000198768365e-05, "loss": 0.154, "step": 6279 }, { "epoch": 4.017914267434421, "grad_norm": 1.3275147334030686, "learning_rate": 1.3585065863650908e-05, "loss": 0.1699, "step": 6280 }, { "epoch": 4.018554062699936, "grad_norm": 1.2954957003854393, "learning_rate": 1.3583131374597102e-05, "loss": 0.1659, "step": 6281 }, { "epoch": 4.019193857965451, "grad_norm": 1.240902060409001, "learning_rate": 1.358119673169001e-05, "loss": 0.1651, "step": 6282 }, { "epoch": 4.019833653230966, "grad_norm": 1.3007293530708048, "learning_rate": 1.3579261935012698e-05, "loss": 0.1744, "step": 6283 }, { "epoch": 4.020473448496481, "grad_norm": 1.2940581497889823, "learning_rate": 1.357732698464825e-05, "loss": 0.162, "step": 6284 }, { "epoch": 4.021113243761996, "grad_norm": 1.2643054416911461, "learning_rate": 1.357539188067974e-05, "loss": 0.1565, "step": 6285 }, { "epoch": 4.021753039027511, "grad_norm": 1.1736883753834582, "learning_rate": 1.3573456623190267e-05, "loss": 0.1544, "step": 6286 }, { "epoch": 4.022392834293027, "grad_norm": 1.2312825505191176, "learning_rate": 1.357152121226292e-05, "loss": 0.1609, "step": 6287 }, { "epoch": 4.023032629558541, "grad_norm": 1.1745947448762928, "learning_rate": 1.3569585647980808e-05, "loss": 0.1536, "step": 6288 }, { "epoch": 4.023672424824056, "grad_norm": 1.2669074057558525, "learning_rate": 1.3567649930427038e-05, "loss": 0.1626, "step": 6289 }, { "epoch": 4.024312220089572, "grad_norm": 1.3142612595128889, "learning_rate": 1.3565714059684724e-05, "loss": 0.162, "step": 6290 }, { "epoch": 4.024952015355086, "grad_norm": 1.2254386496185863, "learning_rate": 1.3563778035836994e-05, "loss": 0.164, "step": 6291 }, { "epoch": 4.025591810620601, "grad_norm": 1.259350771684172, "learning_rate": 1.356184185896697e-05, "loss": 0.1571, "step": 6292 }, { "epoch": 4.0262316058861165, "grad_norm": 1.1958587362566448, "learning_rate": 1.3559905529157797e-05, "loss": 0.1557, "step": 6293 }, { "epoch": 4.026871401151632, "grad_norm": 1.3453051897092965, "learning_rate": 1.3557969046492613e-05, "loss": 0.1596, "step": 6294 }, { "epoch": 4.027511196417146, "grad_norm": 1.2622362475410966, "learning_rate": 1.3556032411054565e-05, "loss": 0.1516, "step": 6295 }, { "epoch": 4.028150991682661, "grad_norm": 1.165832683258824, "learning_rate": 1.3554095622926812e-05, "loss": 0.1472, "step": 6296 }, { "epoch": 4.028790786948177, "grad_norm": 1.2899880833437185, "learning_rate": 1.3552158682192515e-05, "loss": 0.1622, "step": 6297 }, { "epoch": 4.029430582213692, "grad_norm": 1.2206061142091378, "learning_rate": 1.3550221588934843e-05, "loss": 0.1678, "step": 6298 }, { "epoch": 4.030070377479206, "grad_norm": 1.182281443202498, "learning_rate": 1.3548284343236973e-05, "loss": 0.1507, "step": 6299 }, { "epoch": 4.030710172744722, "grad_norm": 1.29510538786706, "learning_rate": 1.3546346945182082e-05, "loss": 0.1616, "step": 6300 }, { "epoch": 4.031349968010237, "grad_norm": 1.2717343062503605, "learning_rate": 1.3544409394853363e-05, "loss": 0.1699, "step": 6301 }, { "epoch": 4.031989763275751, "grad_norm": 1.2706333522631674, "learning_rate": 1.3542471692334011e-05, "loss": 0.1706, "step": 6302 }, { "epoch": 4.032629558541267, "grad_norm": 1.1456803420168118, "learning_rate": 1.3540533837707223e-05, "loss": 0.1603, "step": 6303 }, { "epoch": 4.033269353806782, "grad_norm": 1.2689527777067537, "learning_rate": 1.3538595831056216e-05, "loss": 0.1595, "step": 6304 }, { "epoch": 4.033909149072297, "grad_norm": 1.2992682794284294, "learning_rate": 1.3536657672464193e-05, "loss": 0.1558, "step": 6305 }, { "epoch": 4.0345489443378115, "grad_norm": 1.1336995430060068, "learning_rate": 1.3534719362014383e-05, "loss": 0.1549, "step": 6306 }, { "epoch": 4.035188739603327, "grad_norm": 1.2384658680543996, "learning_rate": 1.353278089979001e-05, "loss": 0.1564, "step": 6307 }, { "epoch": 4.035828534868842, "grad_norm": 1.3410639590153264, "learning_rate": 1.3530842285874313e-05, "loss": 0.1732, "step": 6308 }, { "epoch": 4.036468330134357, "grad_norm": 1.17961100168347, "learning_rate": 1.3528903520350527e-05, "loss": 0.1608, "step": 6309 }, { "epoch": 4.037108125399872, "grad_norm": 1.2356904599472613, "learning_rate": 1.35269646033019e-05, "loss": 0.1708, "step": 6310 }, { "epoch": 4.037747920665387, "grad_norm": 1.2839176015956095, "learning_rate": 1.3525025534811689e-05, "loss": 0.1689, "step": 6311 }, { "epoch": 4.038387715930902, "grad_norm": 1.2914882233595226, "learning_rate": 1.3523086314963148e-05, "loss": 0.1586, "step": 6312 }, { "epoch": 4.039027511196418, "grad_norm": 1.318885821001228, "learning_rate": 1.352114694383955e-05, "loss": 0.1737, "step": 6313 }, { "epoch": 4.039667306461932, "grad_norm": 1.1590246750998496, "learning_rate": 1.3519207421524168e-05, "loss": 0.1555, "step": 6314 }, { "epoch": 4.040307101727447, "grad_norm": 1.2025472826158514, "learning_rate": 1.3517267748100276e-05, "loss": 0.1662, "step": 6315 }, { "epoch": 4.0409468969929625, "grad_norm": 1.2658004936226885, "learning_rate": 1.3515327923651167e-05, "loss": 0.1678, "step": 6316 }, { "epoch": 4.041586692258477, "grad_norm": 1.2353141785206736, "learning_rate": 1.351338794826013e-05, "loss": 0.1631, "step": 6317 }, { "epoch": 4.042226487523992, "grad_norm": 1.3649929847003208, "learning_rate": 1.3511447822010462e-05, "loss": 0.1628, "step": 6318 }, { "epoch": 4.042866282789507, "grad_norm": 1.2412725405363192, "learning_rate": 1.3509507544985475e-05, "loss": 0.1576, "step": 6319 }, { "epoch": 4.043506078055023, "grad_norm": 1.2932376576363258, "learning_rate": 1.3507567117268471e-05, "loss": 0.1685, "step": 6320 }, { "epoch": 4.044145873320537, "grad_norm": 1.2733647617652935, "learning_rate": 1.350562653894278e-05, "loss": 0.1578, "step": 6321 }, { "epoch": 4.044785668586052, "grad_norm": 1.279220257813644, "learning_rate": 1.3503685810091716e-05, "loss": 0.1638, "step": 6322 }, { "epoch": 4.045425463851568, "grad_norm": 1.233923771615668, "learning_rate": 1.3501744930798619e-05, "loss": 0.1561, "step": 6323 }, { "epoch": 4.046065259117083, "grad_norm": 1.1757360834091952, "learning_rate": 1.3499803901146827e-05, "loss": 0.1533, "step": 6324 }, { "epoch": 4.046705054382597, "grad_norm": 1.2282767861237551, "learning_rate": 1.3497862721219677e-05, "loss": 0.1544, "step": 6325 }, { "epoch": 4.047344849648113, "grad_norm": 1.2382361835221902, "learning_rate": 1.3495921391100524e-05, "loss": 0.1553, "step": 6326 }, { "epoch": 4.047984644913628, "grad_norm": 1.362318882495105, "learning_rate": 1.3493979910872725e-05, "loss": 0.1654, "step": 6327 }, { "epoch": 4.048624440179142, "grad_norm": 1.2135422335486035, "learning_rate": 1.3492038280619647e-05, "loss": 0.1581, "step": 6328 }, { "epoch": 4.0492642354446575, "grad_norm": 1.1131808468977122, "learning_rate": 1.3490096500424653e-05, "loss": 0.1595, "step": 6329 }, { "epoch": 4.049904030710173, "grad_norm": 1.2511166936874079, "learning_rate": 1.3488154570371124e-05, "loss": 0.1691, "step": 6330 }, { "epoch": 4.050543825975688, "grad_norm": 1.209006123497709, "learning_rate": 1.3486212490542446e-05, "loss": 0.1627, "step": 6331 }, { "epoch": 4.0511836212412025, "grad_norm": 1.2349602627170515, "learning_rate": 1.3484270261022005e-05, "loss": 0.1671, "step": 6332 }, { "epoch": 4.051823416506718, "grad_norm": 1.1472878134361901, "learning_rate": 1.3482327881893197e-05, "loss": 0.1569, "step": 6333 }, { "epoch": 4.052463211772233, "grad_norm": 1.1997330593717144, "learning_rate": 1.348038535323942e-05, "loss": 0.1542, "step": 6334 }, { "epoch": 4.053103007037748, "grad_norm": 1.181161069297177, "learning_rate": 1.347844267514409e-05, "loss": 0.1551, "step": 6335 }, { "epoch": 4.053742802303263, "grad_norm": 1.3157907029148372, "learning_rate": 1.347649984769062e-05, "loss": 0.1741, "step": 6336 }, { "epoch": 4.054382597568778, "grad_norm": 1.2551548495580316, "learning_rate": 1.3474556870962427e-05, "loss": 0.1679, "step": 6337 }, { "epoch": 4.055022392834293, "grad_norm": 1.1266048534206206, "learning_rate": 1.3472613745042948e-05, "loss": 0.1522, "step": 6338 }, { "epoch": 4.0556621880998085, "grad_norm": 1.1662993108264958, "learning_rate": 1.347067047001561e-05, "loss": 0.1538, "step": 6339 }, { "epoch": 4.056301983365323, "grad_norm": 1.157830905699519, "learning_rate": 1.3468727045963856e-05, "loss": 0.1498, "step": 6340 }, { "epoch": 4.056941778630838, "grad_norm": 1.2155655096608564, "learning_rate": 1.3466783472971132e-05, "loss": 0.161, "step": 6341 }, { "epoch": 4.057581573896353, "grad_norm": 1.260347255838817, "learning_rate": 1.3464839751120893e-05, "loss": 0.1579, "step": 6342 }, { "epoch": 4.058221369161868, "grad_norm": 1.3005823295614813, "learning_rate": 1.34628958804966e-05, "loss": 0.1856, "step": 6343 }, { "epoch": 4.058861164427383, "grad_norm": 1.2323217232343608, "learning_rate": 1.3460951861181714e-05, "loss": 0.1633, "step": 6344 }, { "epoch": 4.059500959692898, "grad_norm": 1.1927654682694782, "learning_rate": 1.3459007693259713e-05, "loss": 0.1625, "step": 6345 }, { "epoch": 4.060140754958414, "grad_norm": 1.264994195349691, "learning_rate": 1.345706337681408e-05, "loss": 0.1531, "step": 6346 }, { "epoch": 4.060780550223928, "grad_norm": 1.2063383154394272, "learning_rate": 1.345511891192829e-05, "loss": 0.1697, "step": 6347 }, { "epoch": 4.061420345489443, "grad_norm": 1.372569910669759, "learning_rate": 1.3453174298685843e-05, "loss": 0.1783, "step": 6348 }, { "epoch": 4.062060140754959, "grad_norm": 1.3709061842128623, "learning_rate": 1.3451229537170232e-05, "loss": 0.1699, "step": 6349 }, { "epoch": 4.062699936020474, "grad_norm": 1.1927204445927408, "learning_rate": 1.3449284627464963e-05, "loss": 0.1608, "step": 6350 }, { "epoch": 4.063339731285988, "grad_norm": 1.0728814897713075, "learning_rate": 1.3447339569653551e-05, "loss": 0.1447, "step": 6351 }, { "epoch": 4.0639795265515035, "grad_norm": 1.0430585856349066, "learning_rate": 1.3445394363819508e-05, "loss": 0.1386, "step": 6352 }, { "epoch": 4.064619321817019, "grad_norm": 1.2202534552256288, "learning_rate": 1.3443449010046361e-05, "loss": 0.1617, "step": 6353 }, { "epoch": 4.065259117082533, "grad_norm": 1.2318887749256024, "learning_rate": 1.3441503508417638e-05, "loss": 0.1631, "step": 6354 }, { "epoch": 4.0658989123480485, "grad_norm": 1.2019977293959214, "learning_rate": 1.3439557859016876e-05, "loss": 0.1549, "step": 6355 }, { "epoch": 4.066538707613564, "grad_norm": 1.2047864664588577, "learning_rate": 1.343761206192762e-05, "loss": 0.1609, "step": 6356 }, { "epoch": 4.067178502879079, "grad_norm": 1.2843548515975454, "learning_rate": 1.3435666117233417e-05, "loss": 0.1522, "step": 6357 }, { "epoch": 4.067818298144593, "grad_norm": 1.2921621468147924, "learning_rate": 1.3433720025017822e-05, "loss": 0.164, "step": 6358 }, { "epoch": 4.068458093410109, "grad_norm": 1.2924905421632593, "learning_rate": 1.3431773785364395e-05, "loss": 0.1751, "step": 6359 }, { "epoch": 4.069097888675624, "grad_norm": 1.1465800754716313, "learning_rate": 1.3429827398356707e-05, "loss": 0.1463, "step": 6360 }, { "epoch": 4.069737683941139, "grad_norm": 1.2897783509208478, "learning_rate": 1.3427880864078336e-05, "loss": 0.1502, "step": 6361 }, { "epoch": 4.070377479206654, "grad_norm": 1.3064223879313015, "learning_rate": 1.3425934182612854e-05, "loss": 0.1575, "step": 6362 }, { "epoch": 4.071017274472169, "grad_norm": 1.3649182781261504, "learning_rate": 1.3423987354043854e-05, "loss": 0.159, "step": 6363 }, { "epoch": 4.071657069737684, "grad_norm": 1.227230408111429, "learning_rate": 1.3422040378454928e-05, "loss": 0.1605, "step": 6364 }, { "epoch": 4.0722968650031985, "grad_norm": 1.2452106298152632, "learning_rate": 1.3420093255929673e-05, "loss": 0.1526, "step": 6365 }, { "epoch": 4.072936660268714, "grad_norm": 1.2252905742822437, "learning_rate": 1.34181459865517e-05, "loss": 0.1624, "step": 6366 }, { "epoch": 4.073576455534229, "grad_norm": 1.2108195746821717, "learning_rate": 1.3416198570404618e-05, "loss": 0.1621, "step": 6367 }, { "epoch": 4.074216250799744, "grad_norm": 1.2300434697162315, "learning_rate": 1.3414251007572048e-05, "loss": 0.1627, "step": 6368 }, { "epoch": 4.074856046065259, "grad_norm": 1.2227501640747962, "learning_rate": 1.3412303298137612e-05, "loss": 0.1582, "step": 6369 }, { "epoch": 4.075495841330774, "grad_norm": 1.2025220024840413, "learning_rate": 1.341035544218494e-05, "loss": 0.1509, "step": 6370 }, { "epoch": 4.076135636596289, "grad_norm": 1.2291214592977007, "learning_rate": 1.3408407439797678e-05, "loss": 0.171, "step": 6371 }, { "epoch": 4.076775431861805, "grad_norm": 1.2385437668567827, "learning_rate": 1.3406459291059458e-05, "loss": 0.1615, "step": 6372 }, { "epoch": 4.077415227127319, "grad_norm": 1.1987609221420792, "learning_rate": 1.3404510996053936e-05, "loss": 0.1504, "step": 6373 }, { "epoch": 4.078055022392834, "grad_norm": 1.3357427249248641, "learning_rate": 1.3402562554864769e-05, "loss": 0.1745, "step": 6374 }, { "epoch": 4.0786948176583495, "grad_norm": 1.2497828731022023, "learning_rate": 1.3400613967575618e-05, "loss": 0.1593, "step": 6375 }, { "epoch": 4.079334612923865, "grad_norm": 1.2838561617574982, "learning_rate": 1.3398665234270154e-05, "loss": 0.174, "step": 6376 }, { "epoch": 4.079974408189379, "grad_norm": 1.090842748228203, "learning_rate": 1.3396716355032051e-05, "loss": 0.145, "step": 6377 }, { "epoch": 4.0806142034548945, "grad_norm": 1.3223261465556169, "learning_rate": 1.3394767329944989e-05, "loss": 0.1732, "step": 6378 }, { "epoch": 4.08125399872041, "grad_norm": 1.209775462859589, "learning_rate": 1.3392818159092658e-05, "loss": 0.1548, "step": 6379 }, { "epoch": 4.081893793985924, "grad_norm": 1.2159130678931958, "learning_rate": 1.3390868842558746e-05, "loss": 0.1473, "step": 6380 }, { "epoch": 4.082533589251439, "grad_norm": 1.2100944587559683, "learning_rate": 1.3388919380426964e-05, "loss": 0.1597, "step": 6381 }, { "epoch": 4.083173384516955, "grad_norm": 1.14860598341372, "learning_rate": 1.3386969772781007e-05, "loss": 0.1589, "step": 6382 }, { "epoch": 4.08381317978247, "grad_norm": 1.2930806632890413, "learning_rate": 1.3385020019704595e-05, "loss": 0.173, "step": 6383 }, { "epoch": 4.084452975047984, "grad_norm": 1.1809688961232963, "learning_rate": 1.3383070121281447e-05, "loss": 0.1446, "step": 6384 }, { "epoch": 4.0850927703135, "grad_norm": 1.2008100705154174, "learning_rate": 1.3381120077595282e-05, "loss": 0.1521, "step": 6385 }, { "epoch": 4.085732565579015, "grad_norm": 1.2522026361114837, "learning_rate": 1.3379169888729839e-05, "loss": 0.1704, "step": 6386 }, { "epoch": 4.08637236084453, "grad_norm": 1.0572191833312394, "learning_rate": 1.337721955476885e-05, "loss": 0.1438, "step": 6387 }, { "epoch": 4.0870121561100445, "grad_norm": 1.3126615849260859, "learning_rate": 1.3375269075796063e-05, "loss": 0.1596, "step": 6388 }, { "epoch": 4.08765195137556, "grad_norm": 1.1080333612361588, "learning_rate": 1.337331845189522e-05, "loss": 0.1506, "step": 6389 }, { "epoch": 4.088291746641075, "grad_norm": 1.1813430110978573, "learning_rate": 1.3371367683150087e-05, "loss": 0.1666, "step": 6390 }, { "epoch": 4.0889315419065895, "grad_norm": 1.4040528566122314, "learning_rate": 1.3369416769644426e-05, "loss": 0.1708, "step": 6391 }, { "epoch": 4.089571337172105, "grad_norm": 1.1311545136653833, "learning_rate": 1.3367465711462e-05, "loss": 0.1574, "step": 6392 }, { "epoch": 4.09021113243762, "grad_norm": 1.1828457578482046, "learning_rate": 1.336551450868659e-05, "loss": 0.1609, "step": 6393 }, { "epoch": 4.090850927703135, "grad_norm": 1.1488428204236492, "learning_rate": 1.336356316140197e-05, "loss": 0.1642, "step": 6394 }, { "epoch": 4.09149072296865, "grad_norm": 1.1654355360133952, "learning_rate": 1.336161166969193e-05, "loss": 0.165, "step": 6395 }, { "epoch": 4.092130518234165, "grad_norm": 1.1381252892903664, "learning_rate": 1.3359660033640267e-05, "loss": 0.1406, "step": 6396 }, { "epoch": 4.09277031349968, "grad_norm": 1.081238571620948, "learning_rate": 1.3357708253330776e-05, "loss": 0.1471, "step": 6397 }, { "epoch": 4.0934101087651955, "grad_norm": 1.2551936461486568, "learning_rate": 1.3355756328847268e-05, "loss": 0.1665, "step": 6398 }, { "epoch": 4.09404990403071, "grad_norm": 1.2641594594967278, "learning_rate": 1.3353804260273555e-05, "loss": 0.1555, "step": 6399 }, { "epoch": 4.094689699296225, "grad_norm": 1.2122656850216127, "learning_rate": 1.335185204769345e-05, "loss": 0.1607, "step": 6400 }, { "epoch": 4.0953294945617404, "grad_norm": 1.1889104019196752, "learning_rate": 1.334989969119078e-05, "loss": 0.1635, "step": 6401 }, { "epoch": 4.095969289827256, "grad_norm": 1.2074327641055604, "learning_rate": 1.3347947190849376e-05, "loss": 0.1574, "step": 6402 }, { "epoch": 4.09660908509277, "grad_norm": 1.1375276070312694, "learning_rate": 1.3345994546753078e-05, "loss": 0.1427, "step": 6403 }, { "epoch": 4.097248880358285, "grad_norm": 1.1306104706170756, "learning_rate": 1.3344041758985724e-05, "loss": 0.1489, "step": 6404 }, { "epoch": 4.097888675623801, "grad_norm": 1.1925188137241056, "learning_rate": 1.3342088827631166e-05, "loss": 0.1473, "step": 6405 }, { "epoch": 4.098528470889315, "grad_norm": 1.2849536400580095, "learning_rate": 1.3340135752773258e-05, "loss": 0.1748, "step": 6406 }, { "epoch": 4.09916826615483, "grad_norm": 1.2306366222798695, "learning_rate": 1.3338182534495863e-05, "loss": 0.1644, "step": 6407 }, { "epoch": 4.099808061420346, "grad_norm": 1.2009417761259844, "learning_rate": 1.333622917288285e-05, "loss": 0.1698, "step": 6408 }, { "epoch": 4.100447856685861, "grad_norm": 1.1522022049966714, "learning_rate": 1.3334275668018089e-05, "loss": 0.1527, "step": 6409 }, { "epoch": 4.101087651951375, "grad_norm": 1.220473831931454, "learning_rate": 1.3332322019985462e-05, "loss": 0.1548, "step": 6410 }, { "epoch": 4.1017274472168905, "grad_norm": 1.1569095136668397, "learning_rate": 1.3330368228868856e-05, "loss": 0.1469, "step": 6411 }, { "epoch": 4.102367242482406, "grad_norm": 1.215040743157351, "learning_rate": 1.332841429475216e-05, "loss": 0.1538, "step": 6412 }, { "epoch": 4.103007037747921, "grad_norm": 1.2405363227363244, "learning_rate": 1.332646021771928e-05, "loss": 0.157, "step": 6413 }, { "epoch": 4.1036468330134355, "grad_norm": 1.2776248295181327, "learning_rate": 1.3324505997854112e-05, "loss": 0.1536, "step": 6414 }, { "epoch": 4.104286628278951, "grad_norm": 1.228302220341105, "learning_rate": 1.3322551635240571e-05, "loss": 0.16, "step": 6415 }, { "epoch": 4.104926423544466, "grad_norm": 1.2590078227417687, "learning_rate": 1.3320597129962573e-05, "loss": 0.1596, "step": 6416 }, { "epoch": 4.10556621880998, "grad_norm": 1.2534475967036205, "learning_rate": 1.3318642482104038e-05, "loss": 0.1628, "step": 6417 }, { "epoch": 4.106206014075496, "grad_norm": 1.3361538472324463, "learning_rate": 1.3316687691748904e-05, "loss": 0.16, "step": 6418 }, { "epoch": 4.106845809341011, "grad_norm": 1.187441764789322, "learning_rate": 1.3314732758981096e-05, "loss": 0.1482, "step": 6419 }, { "epoch": 4.107485604606526, "grad_norm": 1.173821685138082, "learning_rate": 1.3312777683884564e-05, "loss": 0.1484, "step": 6420 }, { "epoch": 4.108125399872041, "grad_norm": 1.2857291512858975, "learning_rate": 1.3310822466543247e-05, "loss": 0.1539, "step": 6421 }, { "epoch": 4.108765195137556, "grad_norm": 1.2540797910523693, "learning_rate": 1.3308867107041106e-05, "loss": 0.1604, "step": 6422 }, { "epoch": 4.109404990403071, "grad_norm": 1.1989182360545096, "learning_rate": 1.3306911605462096e-05, "loss": 0.1685, "step": 6423 }, { "epoch": 4.110044785668586, "grad_norm": 1.257935719578642, "learning_rate": 1.3304955961890185e-05, "loss": 0.158, "step": 6424 }, { "epoch": 4.110684580934101, "grad_norm": 1.2822707249395109, "learning_rate": 1.3303000176409344e-05, "loss": 0.162, "step": 6425 }, { "epoch": 4.111324376199616, "grad_norm": 1.2111911628028529, "learning_rate": 1.3301044249103549e-05, "loss": 0.1487, "step": 6426 }, { "epoch": 4.111964171465131, "grad_norm": 1.3010462269930108, "learning_rate": 1.3299088180056785e-05, "loss": 0.1609, "step": 6427 }, { "epoch": 4.112603966730646, "grad_norm": 1.2483088635888746, "learning_rate": 1.3297131969353048e-05, "loss": 0.1682, "step": 6428 }, { "epoch": 4.113243761996161, "grad_norm": 1.118814572110439, "learning_rate": 1.3295175617076327e-05, "loss": 0.157, "step": 6429 }, { "epoch": 4.113883557261676, "grad_norm": 1.2416072094094397, "learning_rate": 1.3293219123310628e-05, "loss": 0.1602, "step": 6430 }, { "epoch": 4.114523352527192, "grad_norm": 1.3313639373268187, "learning_rate": 1.3291262488139955e-05, "loss": 0.1623, "step": 6431 }, { "epoch": 4.115163147792706, "grad_norm": 1.1297753516979332, "learning_rate": 1.3289305711648326e-05, "loss": 0.1517, "step": 6432 }, { "epoch": 4.115802943058221, "grad_norm": 1.1172813306252798, "learning_rate": 1.3287348793919764e-05, "loss": 0.1457, "step": 6433 }, { "epoch": 4.1164427383237365, "grad_norm": 1.2800465427297465, "learning_rate": 1.3285391735038289e-05, "loss": 0.1791, "step": 6434 }, { "epoch": 4.117082533589252, "grad_norm": 1.2515318242250917, "learning_rate": 1.328343453508794e-05, "loss": 0.1667, "step": 6435 }, { "epoch": 4.117722328854766, "grad_norm": 1.3482444432419158, "learning_rate": 1.328147719415275e-05, "loss": 0.1725, "step": 6436 }, { "epoch": 4.1183621241202815, "grad_norm": 1.2493966876607152, "learning_rate": 1.327951971231677e-05, "loss": 0.1635, "step": 6437 }, { "epoch": 4.119001919385797, "grad_norm": 1.208239163125236, "learning_rate": 1.3277562089664047e-05, "loss": 0.161, "step": 6438 }, { "epoch": 4.119641714651312, "grad_norm": 1.1363888686586807, "learning_rate": 1.3275604326278634e-05, "loss": 0.1472, "step": 6439 }, { "epoch": 4.120281509916826, "grad_norm": 1.2869987990659226, "learning_rate": 1.3273646422244601e-05, "loss": 0.1606, "step": 6440 }, { "epoch": 4.120921305182342, "grad_norm": 1.1449510926675937, "learning_rate": 1.3271688377646013e-05, "loss": 0.15, "step": 6441 }, { "epoch": 4.121561100447857, "grad_norm": 1.2329769480373574, "learning_rate": 1.3269730192566947e-05, "loss": 0.1578, "step": 6442 }, { "epoch": 4.122200895713371, "grad_norm": 1.2186926362788042, "learning_rate": 1.3267771867091483e-05, "loss": 0.1539, "step": 6443 }, { "epoch": 4.122840690978887, "grad_norm": 1.2179811527461402, "learning_rate": 1.3265813401303709e-05, "loss": 0.1552, "step": 6444 }, { "epoch": 4.123480486244402, "grad_norm": 1.2595387015697688, "learning_rate": 1.3263854795287716e-05, "loss": 0.1657, "step": 6445 }, { "epoch": 4.124120281509917, "grad_norm": 1.2186389736014065, "learning_rate": 1.3261896049127603e-05, "loss": 0.1587, "step": 6446 }, { "epoch": 4.1247600767754315, "grad_norm": 1.2639316548293047, "learning_rate": 1.3259937162907476e-05, "loss": 0.1613, "step": 6447 }, { "epoch": 4.125399872040947, "grad_norm": 1.184978165993765, "learning_rate": 1.3257978136711447e-05, "loss": 0.1509, "step": 6448 }, { "epoch": 4.126039667306462, "grad_norm": 1.2613366004351598, "learning_rate": 1.3256018970623632e-05, "loss": 0.1667, "step": 6449 }, { "epoch": 4.126679462571977, "grad_norm": 1.3322064825778142, "learning_rate": 1.3254059664728155e-05, "loss": 0.1662, "step": 6450 }, { "epoch": 4.127319257837492, "grad_norm": 1.2784765066219885, "learning_rate": 1.3252100219109143e-05, "loss": 0.1614, "step": 6451 }, { "epoch": 4.127959053103007, "grad_norm": 1.3430558681219846, "learning_rate": 1.3250140633850736e-05, "loss": 0.1779, "step": 6452 }, { "epoch": 4.128598848368522, "grad_norm": 1.3448056120063687, "learning_rate": 1.324818090903707e-05, "loss": 0.1742, "step": 6453 }, { "epoch": 4.129238643634037, "grad_norm": 1.1847875309515636, "learning_rate": 1.3246221044752292e-05, "loss": 0.1678, "step": 6454 }, { "epoch": 4.129878438899552, "grad_norm": 1.2650422154808645, "learning_rate": 1.3244261041080559e-05, "loss": 0.1689, "step": 6455 }, { "epoch": 4.130518234165067, "grad_norm": 1.314764644739118, "learning_rate": 1.3242300898106026e-05, "loss": 0.1698, "step": 6456 }, { "epoch": 4.1311580294305825, "grad_norm": 1.331276399605815, "learning_rate": 1.324034061591286e-05, "loss": 0.1727, "step": 6457 }, { "epoch": 4.131797824696097, "grad_norm": 1.3557657837773047, "learning_rate": 1.3238380194585237e-05, "loss": 0.1645, "step": 6458 }, { "epoch": 4.132437619961612, "grad_norm": 1.2038476003561602, "learning_rate": 1.3236419634207325e-05, "loss": 0.1524, "step": 6459 }, { "epoch": 4.1330774152271275, "grad_norm": 1.3665826999938075, "learning_rate": 1.3234458934863312e-05, "loss": 0.1595, "step": 6460 }, { "epoch": 4.133717210492643, "grad_norm": 1.4545589635984493, "learning_rate": 1.3232498096637384e-05, "loss": 0.186, "step": 6461 }, { "epoch": 4.134357005758157, "grad_norm": 1.1740388135713302, "learning_rate": 1.3230537119613738e-05, "loss": 0.151, "step": 6462 }, { "epoch": 4.134996801023672, "grad_norm": 1.2972320458140418, "learning_rate": 1.3228576003876578e-05, "loss": 0.1723, "step": 6463 }, { "epoch": 4.135636596289188, "grad_norm": 1.2615757324555232, "learning_rate": 1.3226614749510105e-05, "loss": 0.1619, "step": 6464 }, { "epoch": 4.136276391554702, "grad_norm": 1.2571120849872157, "learning_rate": 1.3224653356598535e-05, "loss": 0.1575, "step": 6465 }, { "epoch": 4.136916186820217, "grad_norm": 1.3198345556377225, "learning_rate": 1.3222691825226086e-05, "loss": 0.1639, "step": 6466 }, { "epoch": 4.137555982085733, "grad_norm": 1.1853498170663879, "learning_rate": 1.3220730155476984e-05, "loss": 0.1479, "step": 6467 }, { "epoch": 4.138195777351248, "grad_norm": 1.2410576361169041, "learning_rate": 1.3218768347435458e-05, "loss": 0.1717, "step": 6468 }, { "epoch": 4.138835572616762, "grad_norm": 1.2878664728269094, "learning_rate": 1.3216806401185744e-05, "loss": 0.1624, "step": 6469 }, { "epoch": 4.1394753678822775, "grad_norm": 1.2241975931769011, "learning_rate": 1.3214844316812086e-05, "loss": 0.1727, "step": 6470 }, { "epoch": 4.140115163147793, "grad_norm": 1.188990776069567, "learning_rate": 1.3212882094398731e-05, "loss": 0.1621, "step": 6471 }, { "epoch": 4.140754958413308, "grad_norm": 1.1059751562575486, "learning_rate": 1.3210919734029932e-05, "loss": 0.1579, "step": 6472 }, { "epoch": 4.1413947536788225, "grad_norm": 1.313950479792437, "learning_rate": 1.3208957235789957e-05, "loss": 0.1615, "step": 6473 }, { "epoch": 4.142034548944338, "grad_norm": 1.300444118498972, "learning_rate": 1.3206994599763064e-05, "loss": 0.1576, "step": 6474 }, { "epoch": 4.142674344209853, "grad_norm": 1.2352372621845216, "learning_rate": 1.3205031826033528e-05, "loss": 0.1652, "step": 6475 }, { "epoch": 4.143314139475368, "grad_norm": 1.1641331110120077, "learning_rate": 1.3203068914685624e-05, "loss": 0.1644, "step": 6476 }, { "epoch": 4.143953934740883, "grad_norm": 1.2900472095241577, "learning_rate": 1.3201105865803641e-05, "loss": 0.1718, "step": 6477 }, { "epoch": 4.144593730006398, "grad_norm": 1.3061157329443842, "learning_rate": 1.3199142679471865e-05, "loss": 0.174, "step": 6478 }, { "epoch": 4.145233525271913, "grad_norm": 1.1697989314001882, "learning_rate": 1.3197179355774594e-05, "loss": 0.1511, "step": 6479 }, { "epoch": 4.145873320537428, "grad_norm": 1.1892675446534318, "learning_rate": 1.319521589479613e-05, "loss": 0.1583, "step": 6480 }, { "epoch": 4.146513115802943, "grad_norm": 1.2986204253410705, "learning_rate": 1.3193252296620775e-05, "loss": 0.1619, "step": 6481 }, { "epoch": 4.147152911068458, "grad_norm": 1.430137619133267, "learning_rate": 1.3191288561332853e-05, "loss": 0.1866, "step": 6482 }, { "epoch": 4.1477927063339735, "grad_norm": 1.4005002017710666, "learning_rate": 1.3189324689016676e-05, "loss": 0.1674, "step": 6483 }, { "epoch": 4.148432501599488, "grad_norm": 1.3067243652192853, "learning_rate": 1.3187360679756567e-05, "loss": 0.1722, "step": 6484 }, { "epoch": 4.149072296865003, "grad_norm": 1.2737637523285072, "learning_rate": 1.3185396533636862e-05, "loss": 0.1623, "step": 6485 }, { "epoch": 4.149712092130518, "grad_norm": 1.245791451276545, "learning_rate": 1.3183432250741894e-05, "loss": 0.1615, "step": 6486 }, { "epoch": 4.150351887396034, "grad_norm": 1.3134273895200415, "learning_rate": 1.3181467831156011e-05, "loss": 0.1636, "step": 6487 }, { "epoch": 4.150991682661548, "grad_norm": 1.3661451372527713, "learning_rate": 1.3179503274963557e-05, "loss": 0.1741, "step": 6488 }, { "epoch": 4.151631477927063, "grad_norm": 1.3135804936218185, "learning_rate": 1.317753858224889e-05, "loss": 0.1697, "step": 6489 }, { "epoch": 4.152271273192579, "grad_norm": 1.2812571997649815, "learning_rate": 1.3175573753096367e-05, "loss": 0.1496, "step": 6490 }, { "epoch": 4.152911068458093, "grad_norm": 1.1952656294436317, "learning_rate": 1.3173608787590358e-05, "loss": 0.1632, "step": 6491 }, { "epoch": 4.153550863723608, "grad_norm": 1.1822966605732426, "learning_rate": 1.3171643685815232e-05, "loss": 0.1549, "step": 6492 }, { "epoch": 4.1541906589891235, "grad_norm": 1.1669611783711529, "learning_rate": 1.3169678447855366e-05, "loss": 0.1461, "step": 6493 }, { "epoch": 4.154830454254639, "grad_norm": 1.1440029032541899, "learning_rate": 1.3167713073795145e-05, "loss": 0.1484, "step": 6494 }, { "epoch": 4.155470249520153, "grad_norm": 1.2809709259976196, "learning_rate": 1.3165747563718965e-05, "loss": 0.1622, "step": 6495 }, { "epoch": 4.1561100447856685, "grad_norm": 1.343207155312542, "learning_rate": 1.3163781917711212e-05, "loss": 0.1697, "step": 6496 }, { "epoch": 4.156749840051184, "grad_norm": 1.2598723414737165, "learning_rate": 1.3161816135856294e-05, "loss": 0.1707, "step": 6497 }, { "epoch": 4.157389635316699, "grad_norm": 1.2192973032423045, "learning_rate": 1.3159850218238614e-05, "loss": 0.1591, "step": 6498 }, { "epoch": 4.158029430582213, "grad_norm": 1.262446084538289, "learning_rate": 1.3157884164942585e-05, "loss": 0.162, "step": 6499 }, { "epoch": 4.158669225847729, "grad_norm": 1.2796487629424502, "learning_rate": 1.315591797605263e-05, "loss": 0.1684, "step": 6500 }, { "epoch": 4.159309021113244, "grad_norm": 1.3826110611640392, "learning_rate": 1.3153951651653167e-05, "loss": 0.1684, "step": 6501 }, { "epoch": 4.159948816378758, "grad_norm": 1.190316622355004, "learning_rate": 1.3151985191828635e-05, "loss": 0.1612, "step": 6502 }, { "epoch": 4.160588611644274, "grad_norm": 1.2592301401266401, "learning_rate": 1.3150018596663463e-05, "loss": 0.1643, "step": 6503 }, { "epoch": 4.161228406909789, "grad_norm": 1.287244481155769, "learning_rate": 1.3148051866242099e-05, "loss": 0.1685, "step": 6504 }, { "epoch": 4.161868202175304, "grad_norm": 1.3015950681651256, "learning_rate": 1.3146085000648987e-05, "loss": 0.1763, "step": 6505 }, { "epoch": 4.162507997440819, "grad_norm": 1.3200737065480743, "learning_rate": 1.314411799996858e-05, "loss": 0.1759, "step": 6506 }, { "epoch": 4.163147792706334, "grad_norm": 1.1958969142158615, "learning_rate": 1.314215086428534e-05, "loss": 0.1627, "step": 6507 }, { "epoch": 4.163787587971849, "grad_norm": 1.2833914586401536, "learning_rate": 1.3140183593683731e-05, "loss": 0.1751, "step": 6508 }, { "epoch": 4.164427383237364, "grad_norm": 1.3119956402493378, "learning_rate": 1.3138216188248224e-05, "loss": 0.1752, "step": 6509 }, { "epoch": 4.165067178502879, "grad_norm": 1.2768191537193496, "learning_rate": 1.3136248648063296e-05, "loss": 0.1729, "step": 6510 }, { "epoch": 4.165706973768394, "grad_norm": 1.1830057948060198, "learning_rate": 1.3134280973213432e-05, "loss": 0.1603, "step": 6511 }, { "epoch": 4.166346769033909, "grad_norm": 1.2466676822283342, "learning_rate": 1.3132313163783118e-05, "loss": 0.1489, "step": 6512 }, { "epoch": 4.166986564299425, "grad_norm": 1.2613609602849065, "learning_rate": 1.3130345219856846e-05, "loss": 0.1616, "step": 6513 }, { "epoch": 4.167626359564939, "grad_norm": 1.1442271765205871, "learning_rate": 1.3128377141519119e-05, "loss": 0.1569, "step": 6514 }, { "epoch": 4.168266154830454, "grad_norm": 1.2028776445432094, "learning_rate": 1.3126408928854447e-05, "loss": 0.16, "step": 6515 }, { "epoch": 4.1689059500959695, "grad_norm": 1.1304716703088278, "learning_rate": 1.3124440581947332e-05, "loss": 0.1472, "step": 6516 }, { "epoch": 4.169545745361484, "grad_norm": 1.3515871523799003, "learning_rate": 1.31224721008823e-05, "loss": 0.1637, "step": 6517 }, { "epoch": 4.170185540626999, "grad_norm": 1.1519753853195331, "learning_rate": 1.3120503485743867e-05, "loss": 0.1422, "step": 6518 }, { "epoch": 4.1708253358925145, "grad_norm": 1.1586244150488947, "learning_rate": 1.3118534736616568e-05, "loss": 0.1467, "step": 6519 }, { "epoch": 4.17146513115803, "grad_norm": 1.1221902855953965, "learning_rate": 1.3116565853584937e-05, "loss": 0.1404, "step": 6520 }, { "epoch": 4.172104926423544, "grad_norm": 1.3930889692317465, "learning_rate": 1.3114596836733507e-05, "loss": 0.1818, "step": 6521 }, { "epoch": 4.172744721689059, "grad_norm": 1.1038238380037682, "learning_rate": 1.3112627686146833e-05, "loss": 0.1506, "step": 6522 }, { "epoch": 4.173384516954575, "grad_norm": 1.3025905478154438, "learning_rate": 1.3110658401909459e-05, "loss": 0.1622, "step": 6523 }, { "epoch": 4.17402431222009, "grad_norm": 1.214893290322716, "learning_rate": 1.310868898410595e-05, "loss": 0.1678, "step": 6524 }, { "epoch": 4.174664107485604, "grad_norm": 1.22848066885572, "learning_rate": 1.3106719432820864e-05, "loss": 0.1711, "step": 6525 }, { "epoch": 4.17530390275112, "grad_norm": 1.294473123642591, "learning_rate": 1.3104749748138775e-05, "loss": 0.1623, "step": 6526 }, { "epoch": 4.175943698016635, "grad_norm": 1.1993140611659945, "learning_rate": 1.3102779930144252e-05, "loss": 0.1579, "step": 6527 }, { "epoch": 4.176583493282149, "grad_norm": 1.2229754215247057, "learning_rate": 1.3100809978921882e-05, "loss": 0.1615, "step": 6528 }, { "epoch": 4.177223288547665, "grad_norm": 1.2460744983083358, "learning_rate": 1.3098839894556243e-05, "loss": 0.1672, "step": 6529 }, { "epoch": 4.17786308381318, "grad_norm": 1.228756845368751, "learning_rate": 1.3096869677131933e-05, "loss": 0.1607, "step": 6530 }, { "epoch": 4.178502879078695, "grad_norm": 1.189668939850081, "learning_rate": 1.3094899326733548e-05, "loss": 0.1531, "step": 6531 }, { "epoch": 4.1791426743442095, "grad_norm": 1.278402983834548, "learning_rate": 1.3092928843445691e-05, "loss": 0.1627, "step": 6532 }, { "epoch": 4.179782469609725, "grad_norm": 1.2099283112610846, "learning_rate": 1.309095822735297e-05, "loss": 0.1527, "step": 6533 }, { "epoch": 4.18042226487524, "grad_norm": 1.3079400301964526, "learning_rate": 1.3088987478540006e-05, "loss": 0.1673, "step": 6534 }, { "epoch": 4.181062060140755, "grad_norm": 1.1717982460451635, "learning_rate": 1.3087016597091411e-05, "loss": 0.1584, "step": 6535 }, { "epoch": 4.18170185540627, "grad_norm": 1.1412898353262986, "learning_rate": 1.3085045583091813e-05, "loss": 0.1494, "step": 6536 }, { "epoch": 4.182341650671785, "grad_norm": 1.2241680981937595, "learning_rate": 1.3083074436625849e-05, "loss": 0.1641, "step": 6537 }, { "epoch": 4.1829814459373, "grad_norm": 1.066239394878192, "learning_rate": 1.308110315777815e-05, "loss": 0.1518, "step": 6538 }, { "epoch": 4.183621241202815, "grad_norm": 1.1419501340003053, "learning_rate": 1.3079131746633363e-05, "loss": 0.1533, "step": 6539 }, { "epoch": 4.18426103646833, "grad_norm": 1.2723087628639158, "learning_rate": 1.3077160203276138e-05, "loss": 0.1688, "step": 6540 }, { "epoch": 4.184900831733845, "grad_norm": 1.1805760173029813, "learning_rate": 1.3075188527791126e-05, "loss": 0.1609, "step": 6541 }, { "epoch": 4.1855406269993605, "grad_norm": 1.2100907470025848, "learning_rate": 1.3073216720262989e-05, "loss": 0.1564, "step": 6542 }, { "epoch": 4.186180422264875, "grad_norm": 1.2778553876907905, "learning_rate": 1.3071244780776395e-05, "loss": 0.1715, "step": 6543 }, { "epoch": 4.18682021753039, "grad_norm": 1.2288121158186043, "learning_rate": 1.3069272709416011e-05, "loss": 0.1547, "step": 6544 }, { "epoch": 4.187460012795905, "grad_norm": 1.2413440883010785, "learning_rate": 1.3067300506266518e-05, "loss": 0.1595, "step": 6545 }, { "epoch": 4.188099808061421, "grad_norm": 1.2972534248730998, "learning_rate": 1.3065328171412597e-05, "loss": 0.1657, "step": 6546 }, { "epoch": 4.188739603326935, "grad_norm": 1.2865963975403507, "learning_rate": 1.3063355704938938e-05, "loss": 0.1591, "step": 6547 }, { "epoch": 4.18937939859245, "grad_norm": 1.2923492039069453, "learning_rate": 1.3061383106930232e-05, "loss": 0.158, "step": 6548 }, { "epoch": 4.190019193857966, "grad_norm": 1.2964228493331393, "learning_rate": 1.3059410377471185e-05, "loss": 0.1681, "step": 6549 }, { "epoch": 4.190658989123481, "grad_norm": 1.237187540977836, "learning_rate": 1.3057437516646497e-05, "loss": 0.1617, "step": 6550 }, { "epoch": 4.191298784388995, "grad_norm": 1.286008573575109, "learning_rate": 1.3055464524540877e-05, "loss": 0.1622, "step": 6551 }, { "epoch": 4.1919385796545106, "grad_norm": 1.2281101174226954, "learning_rate": 1.3053491401239051e-05, "loss": 0.168, "step": 6552 }, { "epoch": 4.192578374920026, "grad_norm": 1.2236741132552866, "learning_rate": 1.3051518146825732e-05, "loss": 0.1672, "step": 6553 }, { "epoch": 4.19321817018554, "grad_norm": 1.2096117243674263, "learning_rate": 1.3049544761385653e-05, "loss": 0.1633, "step": 6554 }, { "epoch": 4.1938579654510555, "grad_norm": 1.4109704233097147, "learning_rate": 1.3047571245003549e-05, "loss": 0.1693, "step": 6555 }, { "epoch": 4.194497760716571, "grad_norm": 1.1934879215551035, "learning_rate": 1.3045597597764154e-05, "loss": 0.1569, "step": 6556 }, { "epoch": 4.195137555982086, "grad_norm": 1.2017759685875145, "learning_rate": 1.3043623819752218e-05, "loss": 0.1653, "step": 6557 }, { "epoch": 4.1957773512476, "grad_norm": 1.251888191435758, "learning_rate": 1.304164991105249e-05, "loss": 0.1672, "step": 6558 }, { "epoch": 4.196417146513116, "grad_norm": 1.290973406574881, "learning_rate": 1.3039675871749724e-05, "loss": 0.1715, "step": 6559 }, { "epoch": 4.197056941778631, "grad_norm": 1.1894590719547686, "learning_rate": 1.3037701701928682e-05, "loss": 0.1541, "step": 6560 }, { "epoch": 4.197696737044146, "grad_norm": 1.301758590307918, "learning_rate": 1.3035727401674132e-05, "loss": 0.1715, "step": 6561 }, { "epoch": 4.198336532309661, "grad_norm": 1.1761253138868277, "learning_rate": 1.3033752971070852e-05, "loss": 0.1577, "step": 6562 }, { "epoch": 4.198976327575176, "grad_norm": 1.3216050498649095, "learning_rate": 1.3031778410203612e-05, "loss": 0.1728, "step": 6563 }, { "epoch": 4.199616122840691, "grad_norm": 1.273807335707573, "learning_rate": 1.3029803719157201e-05, "loss": 0.168, "step": 6564 }, { "epoch": 4.2002559181062065, "grad_norm": 1.284867862965576, "learning_rate": 1.302782889801641e-05, "loss": 0.1647, "step": 6565 }, { "epoch": 4.200895713371721, "grad_norm": 1.3119565715953088, "learning_rate": 1.3025853946866029e-05, "loss": 0.1731, "step": 6566 }, { "epoch": 4.201535508637236, "grad_norm": 1.2114903999853894, "learning_rate": 1.3023878865790867e-05, "loss": 0.1637, "step": 6567 }, { "epoch": 4.202175303902751, "grad_norm": 1.2686363710388848, "learning_rate": 1.302190365487572e-05, "loss": 0.1578, "step": 6568 }, { "epoch": 4.202815099168266, "grad_norm": 1.295960929028581, "learning_rate": 1.3019928314205409e-05, "loss": 0.173, "step": 6569 }, { "epoch": 4.203454894433781, "grad_norm": 1.3372593588625779, "learning_rate": 1.3017952843864748e-05, "loss": 0.1649, "step": 6570 }, { "epoch": 4.204094689699296, "grad_norm": 1.333677788758845, "learning_rate": 1.3015977243938558e-05, "loss": 0.1687, "step": 6571 }, { "epoch": 4.204734484964812, "grad_norm": 1.21489440916873, "learning_rate": 1.3014001514511672e-05, "loss": 0.1504, "step": 6572 }, { "epoch": 4.205374280230326, "grad_norm": 1.1588689160905081, "learning_rate": 1.3012025655668923e-05, "loss": 0.1549, "step": 6573 }, { "epoch": 4.206014075495841, "grad_norm": 1.198572291812961, "learning_rate": 1.301004966749515e-05, "loss": 0.1553, "step": 6574 }, { "epoch": 4.2066538707613566, "grad_norm": 1.2888791930787205, "learning_rate": 1.3008073550075198e-05, "loss": 0.169, "step": 6575 }, { "epoch": 4.207293666026872, "grad_norm": 1.205414522779977, "learning_rate": 1.3006097303493917e-05, "loss": 0.1627, "step": 6576 }, { "epoch": 4.207933461292386, "grad_norm": 1.1853220074624051, "learning_rate": 1.3004120927836168e-05, "loss": 0.1474, "step": 6577 }, { "epoch": 4.2085732565579015, "grad_norm": 1.3154118610331282, "learning_rate": 1.300214442318681e-05, "loss": 0.1605, "step": 6578 }, { "epoch": 4.209213051823417, "grad_norm": 1.2232153868368876, "learning_rate": 1.300016778963071e-05, "loss": 0.1602, "step": 6579 }, { "epoch": 4.209852847088931, "grad_norm": 1.226578734660504, "learning_rate": 1.2998191027252742e-05, "loss": 0.1607, "step": 6580 }, { "epoch": 4.210492642354446, "grad_norm": 1.352366952727773, "learning_rate": 1.2996214136137786e-05, "loss": 0.1608, "step": 6581 }, { "epoch": 4.211132437619962, "grad_norm": 1.2072748905095745, "learning_rate": 1.2994237116370726e-05, "loss": 0.1538, "step": 6582 }, { "epoch": 4.211772232885477, "grad_norm": 1.2965011301287588, "learning_rate": 1.2992259968036444e-05, "loss": 0.1771, "step": 6583 }, { "epoch": 4.212412028150991, "grad_norm": 1.2394120229302712, "learning_rate": 1.2990282691219848e-05, "loss": 0.1562, "step": 6584 }, { "epoch": 4.213051823416507, "grad_norm": 1.3173501796307137, "learning_rate": 1.298830528600583e-05, "loss": 0.166, "step": 6585 }, { "epoch": 4.213691618682022, "grad_norm": 1.3708137537045053, "learning_rate": 1.2986327752479298e-05, "loss": 0.1787, "step": 6586 }, { "epoch": 4.214331413947537, "grad_norm": 1.2110732251403848, "learning_rate": 1.2984350090725166e-05, "loss": 0.1654, "step": 6587 }, { "epoch": 4.214971209213052, "grad_norm": 1.2668997078857773, "learning_rate": 1.2982372300828352e-05, "loss": 0.163, "step": 6588 }, { "epoch": 4.215611004478567, "grad_norm": 1.2492383305649086, "learning_rate": 1.2980394382873775e-05, "loss": 0.1661, "step": 6589 }, { "epoch": 4.216250799744082, "grad_norm": 1.2290700655752476, "learning_rate": 1.2978416336946364e-05, "loss": 0.1724, "step": 6590 }, { "epoch": 4.2168905950095965, "grad_norm": 1.217828708108643, "learning_rate": 1.2976438163131053e-05, "loss": 0.1629, "step": 6591 }, { "epoch": 4.217530390275112, "grad_norm": 1.2129998970539395, "learning_rate": 1.2974459861512784e-05, "loss": 0.1613, "step": 6592 }, { "epoch": 4.218170185540627, "grad_norm": 1.33742270705121, "learning_rate": 1.2972481432176497e-05, "loss": 0.1646, "step": 6593 }, { "epoch": 4.218809980806142, "grad_norm": 1.2149842126080643, "learning_rate": 1.297050287520715e-05, "loss": 0.162, "step": 6594 }, { "epoch": 4.219449776071657, "grad_norm": 1.254029267156046, "learning_rate": 1.296852419068969e-05, "loss": 0.1599, "step": 6595 }, { "epoch": 4.220089571337172, "grad_norm": 1.2404595827779374, "learning_rate": 1.2966545378709081e-05, "loss": 0.1626, "step": 6596 }, { "epoch": 4.220729366602687, "grad_norm": 1.3133531678338155, "learning_rate": 1.2964566439350292e-05, "loss": 0.1676, "step": 6597 }, { "epoch": 4.2213691618682025, "grad_norm": 1.1960141384268346, "learning_rate": 1.2962587372698293e-05, "loss": 0.1551, "step": 6598 }, { "epoch": 4.222008957133717, "grad_norm": 1.2368351732352458, "learning_rate": 1.2960608178838064e-05, "loss": 0.1561, "step": 6599 }, { "epoch": 4.222648752399232, "grad_norm": 1.24384910501819, "learning_rate": 1.2958628857854584e-05, "loss": 0.1665, "step": 6600 }, { "epoch": 4.2232885476647475, "grad_norm": 1.248107897703978, "learning_rate": 1.2956649409832842e-05, "loss": 0.1608, "step": 6601 }, { "epoch": 4.223928342930263, "grad_norm": 1.2624941915579844, "learning_rate": 1.2954669834857838e-05, "loss": 0.1565, "step": 6602 }, { "epoch": 4.224568138195777, "grad_norm": 1.3163056390605525, "learning_rate": 1.2952690133014567e-05, "loss": 0.169, "step": 6603 }, { "epoch": 4.225207933461292, "grad_norm": 1.2700779536118039, "learning_rate": 1.2950710304388034e-05, "loss": 0.1631, "step": 6604 }, { "epoch": 4.225847728726808, "grad_norm": 1.3661623040644884, "learning_rate": 1.2948730349063247e-05, "loss": 0.1528, "step": 6605 }, { "epoch": 4.226487523992322, "grad_norm": 1.2681049745808342, "learning_rate": 1.2946750267125226e-05, "loss": 0.1743, "step": 6606 }, { "epoch": 4.227127319257837, "grad_norm": 1.2795181817290646, "learning_rate": 1.2944770058658991e-05, "loss": 0.1713, "step": 6607 }, { "epoch": 4.227767114523353, "grad_norm": 1.278370823245842, "learning_rate": 1.2942789723749564e-05, "loss": 0.1678, "step": 6608 }, { "epoch": 4.228406909788868, "grad_norm": 1.35179464181157, "learning_rate": 1.2940809262481988e-05, "loss": 0.1689, "step": 6609 }, { "epoch": 4.229046705054382, "grad_norm": 1.2629401024075153, "learning_rate": 1.2938828674941291e-05, "loss": 0.1657, "step": 6610 }, { "epoch": 4.229686500319898, "grad_norm": 1.2761854202216036, "learning_rate": 1.2936847961212516e-05, "loss": 0.1688, "step": 6611 }, { "epoch": 4.230326295585413, "grad_norm": 1.306333668285541, "learning_rate": 1.2934867121380714e-05, "loss": 0.1612, "step": 6612 }, { "epoch": 4.230966090850928, "grad_norm": 1.3340786028428142, "learning_rate": 1.2932886155530939e-05, "loss": 0.1835, "step": 6613 }, { "epoch": 4.2316058861164425, "grad_norm": 1.2232312857433865, "learning_rate": 1.293090506374825e-05, "loss": 0.1659, "step": 6614 }, { "epoch": 4.232245681381958, "grad_norm": 1.1954357407745162, "learning_rate": 1.2928923846117712e-05, "loss": 0.1584, "step": 6615 }, { "epoch": 4.232885476647473, "grad_norm": 1.1724762271962457, "learning_rate": 1.2926942502724391e-05, "loss": 0.1587, "step": 6616 }, { "epoch": 4.233525271912987, "grad_norm": 1.1878232898501702, "learning_rate": 1.2924961033653369e-05, "loss": 0.1567, "step": 6617 }, { "epoch": 4.234165067178503, "grad_norm": 1.1797628984738846, "learning_rate": 1.2922979438989725e-05, "loss": 0.1518, "step": 6618 }, { "epoch": 4.234804862444018, "grad_norm": 1.305968632814567, "learning_rate": 1.292099771881854e-05, "loss": 0.1638, "step": 6619 }, { "epoch": 4.235444657709533, "grad_norm": 1.2262939494725595, "learning_rate": 1.2919015873224909e-05, "loss": 0.1642, "step": 6620 }, { "epoch": 4.236084452975048, "grad_norm": 1.2069896310954833, "learning_rate": 1.291703390229393e-05, "loss": 0.1586, "step": 6621 }, { "epoch": 4.236724248240563, "grad_norm": 1.2677267434666848, "learning_rate": 1.2915051806110706e-05, "loss": 0.1599, "step": 6622 }, { "epoch": 4.237364043506078, "grad_norm": 1.1799402033693345, "learning_rate": 1.2913069584760339e-05, "loss": 0.1489, "step": 6623 }, { "epoch": 4.2380038387715935, "grad_norm": 1.2701503124589149, "learning_rate": 1.2911087238327948e-05, "loss": 0.1582, "step": 6624 }, { "epoch": 4.238643634037108, "grad_norm": 1.3404085664564438, "learning_rate": 1.290910476689865e-05, "loss": 0.1741, "step": 6625 }, { "epoch": 4.239283429302623, "grad_norm": 1.360394620908922, "learning_rate": 1.2907122170557567e-05, "loss": 0.1737, "step": 6626 }, { "epoch": 4.239923224568138, "grad_norm": 1.3620176171620388, "learning_rate": 1.2905139449389831e-05, "loss": 0.1785, "step": 6627 }, { "epoch": 4.240563019833653, "grad_norm": 1.2644780373976567, "learning_rate": 1.2903156603480573e-05, "loss": 0.1602, "step": 6628 }, { "epoch": 4.241202815099168, "grad_norm": 1.2082292567834136, "learning_rate": 1.2901173632914937e-05, "loss": 0.1629, "step": 6629 }, { "epoch": 4.241842610364683, "grad_norm": 1.2897983650617633, "learning_rate": 1.2899190537778063e-05, "loss": 0.1647, "step": 6630 }, { "epoch": 4.242482405630199, "grad_norm": 1.1767733768955686, "learning_rate": 1.2897207318155106e-05, "loss": 0.1518, "step": 6631 }, { "epoch": 4.243122200895713, "grad_norm": 1.3035295998098457, "learning_rate": 1.2895223974131221e-05, "loss": 0.174, "step": 6632 }, { "epoch": 4.243761996161228, "grad_norm": 1.1862915263220193, "learning_rate": 1.2893240505791569e-05, "loss": 0.1447, "step": 6633 }, { "epoch": 4.244401791426744, "grad_norm": 1.2693831442081922, "learning_rate": 1.2891256913221317e-05, "loss": 0.1703, "step": 6634 }, { "epoch": 4.245041586692259, "grad_norm": 1.2487101343376155, "learning_rate": 1.2889273196505633e-05, "loss": 0.1715, "step": 6635 }, { "epoch": 4.245681381957773, "grad_norm": 1.310000423233605, "learning_rate": 1.2887289355729702e-05, "loss": 0.1715, "step": 6636 }, { "epoch": 4.2463211772232885, "grad_norm": 1.360330831715564, "learning_rate": 1.2885305390978698e-05, "loss": 0.1714, "step": 6637 }, { "epoch": 4.246960972488804, "grad_norm": 1.2246119857219593, "learning_rate": 1.2883321302337814e-05, "loss": 0.1622, "step": 6638 }, { "epoch": 4.247600767754319, "grad_norm": 1.280395790167544, "learning_rate": 1.2881337089892245e-05, "loss": 0.1699, "step": 6639 }, { "epoch": 4.248240563019833, "grad_norm": 1.1345166025608535, "learning_rate": 1.2879352753727185e-05, "loss": 0.1544, "step": 6640 }, { "epoch": 4.248880358285349, "grad_norm": 1.2827078287255507, "learning_rate": 1.2877368293927841e-05, "loss": 0.1605, "step": 6641 }, { "epoch": 4.249520153550864, "grad_norm": 1.2714355121566379, "learning_rate": 1.287538371057942e-05, "loss": 0.1803, "step": 6642 }, { "epoch": 4.250159948816378, "grad_norm": 1.1781356934675375, "learning_rate": 1.2873399003767135e-05, "loss": 0.1506, "step": 6643 }, { "epoch": 4.250799744081894, "grad_norm": 1.2132453352798984, "learning_rate": 1.2871414173576211e-05, "loss": 0.1649, "step": 6644 }, { "epoch": 4.251439539347409, "grad_norm": 1.2324793223212742, "learning_rate": 1.2869429220091867e-05, "loss": 0.1578, "step": 6645 }, { "epoch": 4.252079334612924, "grad_norm": 1.1262986142186482, "learning_rate": 1.2867444143399341e-05, "loss": 0.1501, "step": 6646 }, { "epoch": 4.252719129878439, "grad_norm": 1.1266631880024427, "learning_rate": 1.2865458943583862e-05, "loss": 0.1478, "step": 6647 }, { "epoch": 4.253358925143954, "grad_norm": 1.20001655430298, "learning_rate": 1.2863473620730673e-05, "loss": 0.1586, "step": 6648 }, { "epoch": 4.253998720409469, "grad_norm": 1.2630132182642908, "learning_rate": 1.2861488174925024e-05, "loss": 0.1701, "step": 6649 }, { "epoch": 4.254638515674984, "grad_norm": 1.3424111833838563, "learning_rate": 1.2859502606252158e-05, "loss": 0.1703, "step": 6650 }, { "epoch": 4.255278310940499, "grad_norm": 1.333419713085515, "learning_rate": 1.2857516914797339e-05, "loss": 0.1656, "step": 6651 }, { "epoch": 4.255918106206014, "grad_norm": 1.177302340057116, "learning_rate": 1.2855531100645826e-05, "loss": 0.1528, "step": 6652 }, { "epoch": 4.256557901471529, "grad_norm": 1.4064879807077546, "learning_rate": 1.2853545163882886e-05, "loss": 0.1701, "step": 6653 }, { "epoch": 4.257197696737044, "grad_norm": 1.4297284039647746, "learning_rate": 1.2851559104593795e-05, "loss": 0.1782, "step": 6654 }, { "epoch": 4.257837492002559, "grad_norm": 1.3801348446690065, "learning_rate": 1.284957292286383e-05, "loss": 0.181, "step": 6655 }, { "epoch": 4.258477287268074, "grad_norm": 1.321836117518134, "learning_rate": 1.2847586618778272e-05, "loss": 0.1652, "step": 6656 }, { "epoch": 4.25911708253359, "grad_norm": 1.18428459979557, "learning_rate": 1.2845600192422407e-05, "loss": 0.1599, "step": 6657 }, { "epoch": 4.259756877799104, "grad_norm": 1.3247402719966686, "learning_rate": 1.2843613643881532e-05, "loss": 0.1794, "step": 6658 }, { "epoch": 4.260396673064619, "grad_norm": 1.2607171050298325, "learning_rate": 1.2841626973240948e-05, "loss": 0.1672, "step": 6659 }, { "epoch": 4.2610364683301345, "grad_norm": 1.265911313411634, "learning_rate": 1.2839640180585955e-05, "loss": 0.1651, "step": 6660 }, { "epoch": 4.26167626359565, "grad_norm": 1.359001942790187, "learning_rate": 1.2837653266001863e-05, "loss": 0.1708, "step": 6661 }, { "epoch": 4.262316058861164, "grad_norm": 1.1633624425564917, "learning_rate": 1.2835666229573991e-05, "loss": 0.1531, "step": 6662 }, { "epoch": 4.262955854126679, "grad_norm": 1.272066387015008, "learning_rate": 1.2833679071387652e-05, "loss": 0.1611, "step": 6663 }, { "epoch": 4.263595649392195, "grad_norm": 1.280888638974207, "learning_rate": 1.2831691791528177e-05, "loss": 0.1729, "step": 6664 }, { "epoch": 4.264235444657709, "grad_norm": 1.2625555744504122, "learning_rate": 1.2829704390080893e-05, "loss": 0.1618, "step": 6665 }, { "epoch": 4.264875239923224, "grad_norm": 1.2138647649695478, "learning_rate": 1.2827716867131135e-05, "loss": 0.1645, "step": 6666 }, { "epoch": 4.26551503518874, "grad_norm": 1.1800576349058847, "learning_rate": 1.2825729222764245e-05, "loss": 0.156, "step": 6667 }, { "epoch": 4.266154830454255, "grad_norm": 1.2361592006862478, "learning_rate": 1.2823741457065568e-05, "loss": 0.1471, "step": 6668 }, { "epoch": 4.266794625719769, "grad_norm": 1.3015022264150489, "learning_rate": 1.2821753570120458e-05, "loss": 0.1725, "step": 6669 }, { "epoch": 4.267434420985285, "grad_norm": 1.1502158851115463, "learning_rate": 1.281976556201427e-05, "loss": 0.1527, "step": 6670 }, { "epoch": 4.2680742162508, "grad_norm": 1.3702381105041603, "learning_rate": 1.2817777432832364e-05, "loss": 0.177, "step": 6671 }, { "epoch": 4.268714011516315, "grad_norm": 1.2622224053436606, "learning_rate": 1.2815789182660106e-05, "loss": 0.1732, "step": 6672 }, { "epoch": 4.2693538067818295, "grad_norm": 1.164898963548393, "learning_rate": 1.2813800811582868e-05, "loss": 0.1617, "step": 6673 }, { "epoch": 4.269993602047345, "grad_norm": 1.3977551327889686, "learning_rate": 1.2811812319686033e-05, "loss": 0.1719, "step": 6674 }, { "epoch": 4.27063339731286, "grad_norm": 1.2252119763847729, "learning_rate": 1.2809823707054974e-05, "loss": 0.1605, "step": 6675 }, { "epoch": 4.271273192578375, "grad_norm": 1.147156121782692, "learning_rate": 1.2807834973775086e-05, "loss": 0.1523, "step": 6676 }, { "epoch": 4.27191298784389, "grad_norm": 1.2185240344161472, "learning_rate": 1.2805846119931757e-05, "loss": 0.1589, "step": 6677 }, { "epoch": 4.272552783109405, "grad_norm": 1.2836383236199767, "learning_rate": 1.280385714561039e-05, "loss": 0.1731, "step": 6678 }, { "epoch": 4.27319257837492, "grad_norm": 1.1930195351550787, "learning_rate": 1.2801868050896383e-05, "loss": 0.1503, "step": 6679 }, { "epoch": 4.273832373640435, "grad_norm": 1.2880877691229196, "learning_rate": 1.2799878835875146e-05, "loss": 0.1582, "step": 6680 }, { "epoch": 4.27447216890595, "grad_norm": 1.259516720676185, "learning_rate": 1.2797889500632093e-05, "loss": 0.1752, "step": 6681 }, { "epoch": 4.275111964171465, "grad_norm": 1.2676987693406034, "learning_rate": 1.2795900045252641e-05, "loss": 0.174, "step": 6682 }, { "epoch": 4.2757517594369805, "grad_norm": 1.2812897890198887, "learning_rate": 1.2793910469822213e-05, "loss": 0.1671, "step": 6683 }, { "epoch": 4.276391554702495, "grad_norm": 1.338430715152242, "learning_rate": 1.2791920774426244e-05, "loss": 0.1759, "step": 6684 }, { "epoch": 4.27703134996801, "grad_norm": 1.2700406210038757, "learning_rate": 1.2789930959150161e-05, "loss": 0.158, "step": 6685 }, { "epoch": 4.277671145233525, "grad_norm": 1.5007353437701214, "learning_rate": 1.2787941024079407e-05, "loss": 0.1899, "step": 6686 }, { "epoch": 4.278310940499041, "grad_norm": 1.1451302220678785, "learning_rate": 1.2785950969299421e-05, "loss": 0.1559, "step": 6687 }, { "epoch": 4.278950735764555, "grad_norm": 1.281124348183149, "learning_rate": 1.2783960794895658e-05, "loss": 0.1747, "step": 6688 }, { "epoch": 4.27959053103007, "grad_norm": 1.216106650781901, "learning_rate": 1.2781970500953571e-05, "loss": 0.151, "step": 6689 }, { "epoch": 4.280230326295586, "grad_norm": 1.173562180857677, "learning_rate": 1.2779980087558619e-05, "loss": 0.159, "step": 6690 }, { "epoch": 4.280870121561101, "grad_norm": 1.2056475792969252, "learning_rate": 1.2777989554796269e-05, "loss": 0.1533, "step": 6691 }, { "epoch": 4.281509916826615, "grad_norm": 1.3249338353495101, "learning_rate": 1.2775998902751986e-05, "loss": 0.1729, "step": 6692 }, { "epoch": 4.282149712092131, "grad_norm": 1.384571718880031, "learning_rate": 1.2774008131511254e-05, "loss": 0.1818, "step": 6693 }, { "epoch": 4.282789507357646, "grad_norm": 1.4010080449809437, "learning_rate": 1.2772017241159542e-05, "loss": 0.1782, "step": 6694 }, { "epoch": 4.28342930262316, "grad_norm": 1.333961469421224, "learning_rate": 1.2770026231782342e-05, "loss": 0.1692, "step": 6695 }, { "epoch": 4.2840690978886755, "grad_norm": 1.2032358262261404, "learning_rate": 1.2768035103465143e-05, "loss": 0.1556, "step": 6696 }, { "epoch": 4.284708893154191, "grad_norm": 1.1639201260627878, "learning_rate": 1.276604385629344e-05, "loss": 0.1569, "step": 6697 }, { "epoch": 4.285348688419706, "grad_norm": 1.2537829306747343, "learning_rate": 1.2764052490352734e-05, "loss": 0.1694, "step": 6698 }, { "epoch": 4.2859884836852205, "grad_norm": 1.2104384446807261, "learning_rate": 1.2762061005728534e-05, "loss": 0.1581, "step": 6699 }, { "epoch": 4.286628278950736, "grad_norm": 1.395552970188915, "learning_rate": 1.2760069402506348e-05, "loss": 0.1789, "step": 6700 }, { "epoch": 4.287268074216251, "grad_norm": 1.255606980125741, "learning_rate": 1.275807768077169e-05, "loss": 0.1531, "step": 6701 }, { "epoch": 4.287907869481765, "grad_norm": 1.3897868101965845, "learning_rate": 1.2756085840610081e-05, "loss": 0.1696, "step": 6702 }, { "epoch": 4.288547664747281, "grad_norm": 1.2683180366272455, "learning_rate": 1.2754093882107048e-05, "loss": 0.1694, "step": 6703 }, { "epoch": 4.289187460012796, "grad_norm": 1.2392538415312055, "learning_rate": 1.2752101805348127e-05, "loss": 0.153, "step": 6704 }, { "epoch": 4.289827255278311, "grad_norm": 1.2137976117473956, "learning_rate": 1.2750109610418847e-05, "loss": 0.1614, "step": 6705 }, { "epoch": 4.290467050543826, "grad_norm": 1.1799878224146876, "learning_rate": 1.2748117297404754e-05, "loss": 0.157, "step": 6706 }, { "epoch": 4.291106845809341, "grad_norm": 1.2918243781928305, "learning_rate": 1.2746124866391391e-05, "loss": 0.1669, "step": 6707 }, { "epoch": 4.291746641074856, "grad_norm": 1.3842858778789764, "learning_rate": 1.2744132317464313e-05, "loss": 0.1771, "step": 6708 }, { "epoch": 4.292386436340371, "grad_norm": 1.3016547074275702, "learning_rate": 1.2742139650709073e-05, "loss": 0.1642, "step": 6709 }, { "epoch": 4.293026231605886, "grad_norm": 1.3214874351786605, "learning_rate": 1.2740146866211236e-05, "loss": 0.1649, "step": 6710 }, { "epoch": 4.293666026871401, "grad_norm": 1.257286897916871, "learning_rate": 1.2738153964056367e-05, "loss": 0.169, "step": 6711 }, { "epoch": 4.294305822136916, "grad_norm": 1.3649488240479075, "learning_rate": 1.2736160944330034e-05, "loss": 0.1739, "step": 6712 }, { "epoch": 4.294945617402432, "grad_norm": 1.219269430638931, "learning_rate": 1.2734167807117821e-05, "loss": 0.1737, "step": 6713 }, { "epoch": 4.295585412667946, "grad_norm": 1.1217237349549205, "learning_rate": 1.2732174552505306e-05, "loss": 0.154, "step": 6714 }, { "epoch": 4.296225207933461, "grad_norm": 1.1818207652419208, "learning_rate": 1.2730181180578077e-05, "loss": 0.1594, "step": 6715 }, { "epoch": 4.296865003198977, "grad_norm": 1.2843647467281918, "learning_rate": 1.2728187691421723e-05, "loss": 0.1544, "step": 6716 }, { "epoch": 4.297504798464491, "grad_norm": 1.29976183242576, "learning_rate": 1.272619408512184e-05, "loss": 0.1595, "step": 6717 }, { "epoch": 4.298144593730006, "grad_norm": 1.3230459865844069, "learning_rate": 1.2724200361764036e-05, "loss": 0.1785, "step": 6718 }, { "epoch": 4.2987843889955215, "grad_norm": 1.4021525861914108, "learning_rate": 1.2722206521433912e-05, "loss": 0.1616, "step": 6719 }, { "epoch": 4.299424184261037, "grad_norm": 1.3197107529568433, "learning_rate": 1.2720212564217083e-05, "loss": 0.175, "step": 6720 }, { "epoch": 4.300063979526551, "grad_norm": 1.1399988823445715, "learning_rate": 1.2718218490199165e-05, "loss": 0.1547, "step": 6721 }, { "epoch": 4.3007037747920664, "grad_norm": 1.378807449581057, "learning_rate": 1.2716224299465782e-05, "loss": 0.1653, "step": 6722 }, { "epoch": 4.301343570057582, "grad_norm": 1.0684905806632206, "learning_rate": 1.2714229992102563e-05, "loss": 0.1509, "step": 6723 }, { "epoch": 4.301983365323097, "grad_norm": 1.3002568010517528, "learning_rate": 1.2712235568195128e-05, "loss": 0.1662, "step": 6724 }, { "epoch": 4.302623160588611, "grad_norm": 1.377158631982672, "learning_rate": 1.2710241027829126e-05, "loss": 0.1719, "step": 6725 }, { "epoch": 4.303262955854127, "grad_norm": 1.2140577656821965, "learning_rate": 1.2708246371090198e-05, "loss": 0.164, "step": 6726 }, { "epoch": 4.303902751119642, "grad_norm": 1.2793020919172404, "learning_rate": 1.2706251598063985e-05, "loss": 0.1649, "step": 6727 }, { "epoch": 4.304542546385157, "grad_norm": 1.274300701205157, "learning_rate": 1.2704256708836145e-05, "loss": 0.1644, "step": 6728 }, { "epoch": 4.305182341650672, "grad_norm": 1.2878155612150646, "learning_rate": 1.2702261703492329e-05, "loss": 0.1607, "step": 6729 }, { "epoch": 4.305822136916187, "grad_norm": 1.3327665770078498, "learning_rate": 1.2700266582118205e-05, "loss": 0.1719, "step": 6730 }, { "epoch": 4.306461932181702, "grad_norm": 1.3011750925481949, "learning_rate": 1.269827134479944e-05, "loss": 0.1651, "step": 6731 }, { "epoch": 4.3071017274472165, "grad_norm": 1.2533962864230328, "learning_rate": 1.2696275991621699e-05, "loss": 0.1673, "step": 6732 }, { "epoch": 4.307741522712732, "grad_norm": 1.2872709828532276, "learning_rate": 1.2694280522670665e-05, "loss": 0.1641, "step": 6733 }, { "epoch": 4.308381317978247, "grad_norm": 1.289578816613717, "learning_rate": 1.2692284938032016e-05, "loss": 0.1616, "step": 6734 }, { "epoch": 4.309021113243762, "grad_norm": 1.2653495102874346, "learning_rate": 1.2690289237791441e-05, "loss": 0.1525, "step": 6735 }, { "epoch": 4.309660908509277, "grad_norm": 1.1943169778364124, "learning_rate": 1.2688293422034635e-05, "loss": 0.156, "step": 6736 }, { "epoch": 4.310300703774792, "grad_norm": 1.402411353583125, "learning_rate": 1.268629749084729e-05, "loss": 0.1679, "step": 6737 }, { "epoch": 4.310940499040307, "grad_norm": 1.4056854925986992, "learning_rate": 1.2684301444315113e-05, "loss": 0.1819, "step": 6738 }, { "epoch": 4.311580294305822, "grad_norm": 1.3228862062563418, "learning_rate": 1.2682305282523802e-05, "loss": 0.1791, "step": 6739 }, { "epoch": 4.312220089571337, "grad_norm": 1.1719988670119217, "learning_rate": 1.2680309005559073e-05, "loss": 0.1613, "step": 6740 }, { "epoch": 4.312859884836852, "grad_norm": 1.3548362462498804, "learning_rate": 1.2678312613506644e-05, "loss": 0.1761, "step": 6741 }, { "epoch": 4.3134996801023675, "grad_norm": 1.2563535052147126, "learning_rate": 1.2676316106452236e-05, "loss": 0.1558, "step": 6742 }, { "epoch": 4.314139475367882, "grad_norm": 1.3225744232327115, "learning_rate": 1.2674319484481577e-05, "loss": 0.1762, "step": 6743 }, { "epoch": 4.314779270633397, "grad_norm": 1.2052571401073517, "learning_rate": 1.2672322747680395e-05, "loss": 0.16, "step": 6744 }, { "epoch": 4.315419065898912, "grad_norm": 1.267501306832659, "learning_rate": 1.267032589613443e-05, "loss": 0.1658, "step": 6745 }, { "epoch": 4.316058861164428, "grad_norm": 1.1623450545497138, "learning_rate": 1.266832892992942e-05, "loss": 0.1696, "step": 6746 }, { "epoch": 4.316698656429942, "grad_norm": 1.3264769685728988, "learning_rate": 1.266633184915111e-05, "loss": 0.1787, "step": 6747 }, { "epoch": 4.317338451695457, "grad_norm": 1.27660704298608, "learning_rate": 1.2664334653885257e-05, "loss": 0.1812, "step": 6748 }, { "epoch": 4.317978246960973, "grad_norm": 1.3611807976538768, "learning_rate": 1.266233734421761e-05, "loss": 0.1656, "step": 6749 }, { "epoch": 4.318618042226488, "grad_norm": 1.1922111945471732, "learning_rate": 1.2660339920233936e-05, "loss": 0.1658, "step": 6750 }, { "epoch": 4.319257837492002, "grad_norm": 1.3419667416577348, "learning_rate": 1.2658342382019996e-05, "loss": 0.1683, "step": 6751 }, { "epoch": 4.319897632757518, "grad_norm": 1.2757595141113407, "learning_rate": 1.2656344729661567e-05, "loss": 0.1634, "step": 6752 }, { "epoch": 4.320537428023033, "grad_norm": 1.3325956804642818, "learning_rate": 1.2654346963244417e-05, "loss": 0.1685, "step": 6753 }, { "epoch": 4.321177223288547, "grad_norm": 1.1907024409547522, "learning_rate": 1.2652349082854332e-05, "loss": 0.1577, "step": 6754 }, { "epoch": 4.3218170185540625, "grad_norm": 1.4534319341345157, "learning_rate": 1.2650351088577093e-05, "loss": 0.173, "step": 6755 }, { "epoch": 4.322456813819578, "grad_norm": 1.2134672403084885, "learning_rate": 1.2648352980498495e-05, "loss": 0.1548, "step": 6756 }, { "epoch": 4.323096609085093, "grad_norm": 1.1600764439718454, "learning_rate": 1.2646354758704328e-05, "loss": 0.1478, "step": 6757 }, { "epoch": 4.3237364043506075, "grad_norm": 1.2791088725342565, "learning_rate": 1.2644356423280399e-05, "loss": 0.1582, "step": 6758 }, { "epoch": 4.324376199616123, "grad_norm": 1.2257649735792828, "learning_rate": 1.2642357974312507e-05, "loss": 0.1584, "step": 6759 }, { "epoch": 4.325015994881638, "grad_norm": 1.2216308102242182, "learning_rate": 1.2640359411886467e-05, "loss": 0.1557, "step": 6760 }, { "epoch": 4.325655790147153, "grad_norm": 1.2179116209865342, "learning_rate": 1.2638360736088089e-05, "loss": 0.1561, "step": 6761 }, { "epoch": 4.326295585412668, "grad_norm": 1.4036698702667452, "learning_rate": 1.2636361947003192e-05, "loss": 0.157, "step": 6762 }, { "epoch": 4.326935380678183, "grad_norm": 1.192436345453233, "learning_rate": 1.2634363044717606e-05, "loss": 0.1598, "step": 6763 }, { "epoch": 4.327575175943698, "grad_norm": 1.2839846030543833, "learning_rate": 1.2632364029317155e-05, "loss": 0.1674, "step": 6764 }, { "epoch": 4.3282149712092135, "grad_norm": 1.229677633698315, "learning_rate": 1.2630364900887673e-05, "loss": 0.1632, "step": 6765 }, { "epoch": 4.328854766474728, "grad_norm": 1.2059652367619398, "learning_rate": 1.2628365659515005e-05, "loss": 0.1556, "step": 6766 }, { "epoch": 4.329494561740243, "grad_norm": 1.2303930681275173, "learning_rate": 1.2626366305284991e-05, "loss": 0.1713, "step": 6767 }, { "epoch": 4.330134357005758, "grad_norm": 1.2055900596343867, "learning_rate": 1.2624366838283479e-05, "loss": 0.1606, "step": 6768 }, { "epoch": 4.330774152271273, "grad_norm": 1.3000162774481507, "learning_rate": 1.2622367258596322e-05, "loss": 0.1655, "step": 6769 }, { "epoch": 4.331413947536788, "grad_norm": 1.2581973556605, "learning_rate": 1.2620367566309381e-05, "loss": 0.154, "step": 6770 }, { "epoch": 4.332053742802303, "grad_norm": 1.1604598898623428, "learning_rate": 1.261836776150852e-05, "loss": 0.1502, "step": 6771 }, { "epoch": 4.332693538067819, "grad_norm": 1.2097545225779605, "learning_rate": 1.26163678442796e-05, "loss": 0.1491, "step": 6772 }, { "epoch": 4.333333333333333, "grad_norm": 1.3313326161804793, "learning_rate": 1.2614367814708503e-05, "loss": 0.1586, "step": 6773 }, { "epoch": 4.333973128598848, "grad_norm": 1.3511883126724542, "learning_rate": 1.26123676728811e-05, "loss": 0.1698, "step": 6774 }, { "epoch": 4.334612923864364, "grad_norm": 1.1492179786910595, "learning_rate": 1.2610367418883278e-05, "loss": 0.1457, "step": 6775 }, { "epoch": 4.335252719129878, "grad_norm": 1.2311967548359801, "learning_rate": 1.2608367052800924e-05, "loss": 0.1587, "step": 6776 }, { "epoch": 4.335892514395393, "grad_norm": 1.3702823487577809, "learning_rate": 1.2606366574719926e-05, "loss": 0.1685, "step": 6777 }, { "epoch": 4.3365323096609085, "grad_norm": 1.3198670176959169, "learning_rate": 1.2604365984726186e-05, "loss": 0.1676, "step": 6778 }, { "epoch": 4.337172104926424, "grad_norm": 1.3095025591323448, "learning_rate": 1.26023652829056e-05, "loss": 0.1677, "step": 6779 }, { "epoch": 4.337811900191938, "grad_norm": 1.3957379728894315, "learning_rate": 1.2600364469344077e-05, "loss": 0.1723, "step": 6780 }, { "epoch": 4.3384516954574535, "grad_norm": 1.3316050417965963, "learning_rate": 1.2598363544127535e-05, "loss": 0.1719, "step": 6781 }, { "epoch": 4.339091490722969, "grad_norm": 1.2742943556936708, "learning_rate": 1.2596362507341881e-05, "loss": 0.164, "step": 6782 }, { "epoch": 4.339731285988484, "grad_norm": 1.3469104383396566, "learning_rate": 1.2594361359073042e-05, "loss": 0.1667, "step": 6783 }, { "epoch": 4.340371081253998, "grad_norm": 1.2572027485359998, "learning_rate": 1.259236009940694e-05, "loss": 0.1742, "step": 6784 }, { "epoch": 4.341010876519514, "grad_norm": 1.2003535351921963, "learning_rate": 1.2590358728429504e-05, "loss": 0.157, "step": 6785 }, { "epoch": 4.341650671785029, "grad_norm": 1.1567048801900304, "learning_rate": 1.2588357246226676e-05, "loss": 0.1603, "step": 6786 }, { "epoch": 4.342290467050544, "grad_norm": 1.213511988223473, "learning_rate": 1.2586355652884387e-05, "loss": 0.1658, "step": 6787 }, { "epoch": 4.342930262316059, "grad_norm": 1.2506102477036758, "learning_rate": 1.2584353948488594e-05, "loss": 0.1575, "step": 6788 }, { "epoch": 4.343570057581574, "grad_norm": 1.2118978084307177, "learning_rate": 1.2582352133125235e-05, "loss": 0.1587, "step": 6789 }, { "epoch": 4.344209852847089, "grad_norm": 1.2439377473361741, "learning_rate": 1.258035020688027e-05, "loss": 0.1584, "step": 6790 }, { "epoch": 4.3448496481126035, "grad_norm": 1.3843735465778155, "learning_rate": 1.2578348169839658e-05, "loss": 0.1723, "step": 6791 }, { "epoch": 4.345489443378119, "grad_norm": 1.3031102590592691, "learning_rate": 1.257634602208936e-05, "loss": 0.1655, "step": 6792 }, { "epoch": 4.346129238643634, "grad_norm": 1.446696593011016, "learning_rate": 1.2574343763715351e-05, "loss": 0.169, "step": 6793 }, { "epoch": 4.346769033909149, "grad_norm": 1.216228955874002, "learning_rate": 1.2572341394803597e-05, "loss": 0.1654, "step": 6794 }, { "epoch": 4.347408829174664, "grad_norm": 1.2536124574692418, "learning_rate": 1.2570338915440081e-05, "loss": 0.1642, "step": 6795 }, { "epoch": 4.348048624440179, "grad_norm": 1.233255466210166, "learning_rate": 1.2568336325710782e-05, "loss": 0.1658, "step": 6796 }, { "epoch": 4.348688419705694, "grad_norm": 1.3545610637840757, "learning_rate": 1.2566333625701692e-05, "loss": 0.182, "step": 6797 }, { "epoch": 4.34932821497121, "grad_norm": 1.342475705240299, "learning_rate": 1.2564330815498801e-05, "loss": 0.1748, "step": 6798 }, { "epoch": 4.349968010236724, "grad_norm": 1.3712252942810097, "learning_rate": 1.2562327895188104e-05, "loss": 0.1648, "step": 6799 }, { "epoch": 4.350607805502239, "grad_norm": 1.1798413960743266, "learning_rate": 1.256032486485561e-05, "loss": 0.1625, "step": 6800 }, { "epoch": 4.3512476007677545, "grad_norm": 1.2423601819160475, "learning_rate": 1.2558321724587315e-05, "loss": 0.1576, "step": 6801 }, { "epoch": 4.35188739603327, "grad_norm": 1.3079574052021992, "learning_rate": 1.2556318474469239e-05, "loss": 0.1774, "step": 6802 }, { "epoch": 4.352527191298784, "grad_norm": 1.134626854257699, "learning_rate": 1.2554315114587395e-05, "loss": 0.1496, "step": 6803 }, { "epoch": 4.3531669865642995, "grad_norm": 1.2630804286470871, "learning_rate": 1.2552311645027802e-05, "loss": 0.1724, "step": 6804 }, { "epoch": 4.353806781829815, "grad_norm": 1.229332621810343, "learning_rate": 1.255030806587649e-05, "loss": 0.171, "step": 6805 }, { "epoch": 4.354446577095329, "grad_norm": 1.2321697660335547, "learning_rate": 1.2548304377219486e-05, "loss": 0.1602, "step": 6806 }, { "epoch": 4.355086372360844, "grad_norm": 1.3290483157433182, "learning_rate": 1.2546300579142823e-05, "loss": 0.1794, "step": 6807 }, { "epoch": 4.35572616762636, "grad_norm": 1.1991438785814772, "learning_rate": 1.2544296671732545e-05, "loss": 0.1641, "step": 6808 }, { "epoch": 4.356365962891875, "grad_norm": 1.3857957047757188, "learning_rate": 1.254229265507469e-05, "loss": 0.1694, "step": 6809 }, { "epoch": 4.357005758157389, "grad_norm": 1.2999365947377222, "learning_rate": 1.2540288529255316e-05, "loss": 0.1615, "step": 6810 }, { "epoch": 4.357645553422905, "grad_norm": 1.3459762934793298, "learning_rate": 1.2538284294360467e-05, "loss": 0.1719, "step": 6811 }, { "epoch": 4.35828534868842, "grad_norm": 1.2753296681911106, "learning_rate": 1.2536279950476209e-05, "loss": 0.1625, "step": 6812 }, { "epoch": 4.358925143953935, "grad_norm": 1.1651102115903766, "learning_rate": 1.2534275497688602e-05, "loss": 0.1402, "step": 6813 }, { "epoch": 4.3595649392194495, "grad_norm": 1.260636170624608, "learning_rate": 1.2532270936083708e-05, "loss": 0.1562, "step": 6814 }, { "epoch": 4.360204734484965, "grad_norm": 1.083999631226031, "learning_rate": 1.253026626574761e-05, "loss": 0.1514, "step": 6815 }, { "epoch": 4.36084452975048, "grad_norm": 1.3353250236025245, "learning_rate": 1.2528261486766375e-05, "loss": 0.1629, "step": 6816 }, { "epoch": 4.3614843250159945, "grad_norm": 1.2691405500349358, "learning_rate": 1.2526256599226091e-05, "loss": 0.1696, "step": 6817 }, { "epoch": 4.36212412028151, "grad_norm": 1.1976366368736118, "learning_rate": 1.2524251603212845e-05, "loss": 0.1506, "step": 6818 }, { "epoch": 4.362763915547025, "grad_norm": 1.2209911780943887, "learning_rate": 1.2522246498812723e-05, "loss": 0.1683, "step": 6819 }, { "epoch": 4.36340371081254, "grad_norm": 1.2734814243172885, "learning_rate": 1.2520241286111822e-05, "loss": 0.153, "step": 6820 }, { "epoch": 4.364043506078055, "grad_norm": 1.1403632893227273, "learning_rate": 1.2518235965196248e-05, "loss": 0.1519, "step": 6821 }, { "epoch": 4.36468330134357, "grad_norm": 1.301801580257763, "learning_rate": 1.2516230536152097e-05, "loss": 0.1697, "step": 6822 }, { "epoch": 4.365323096609085, "grad_norm": 1.4497540411904197, "learning_rate": 1.2514224999065485e-05, "loss": 0.1857, "step": 6823 }, { "epoch": 4.3659628918746005, "grad_norm": 1.2600493731533426, "learning_rate": 1.2512219354022522e-05, "loss": 0.1655, "step": 6824 }, { "epoch": 4.366602687140115, "grad_norm": 1.4111733447574553, "learning_rate": 1.2510213601109331e-05, "loss": 0.1819, "step": 6825 }, { "epoch": 4.36724248240563, "grad_norm": 1.2492627347935712, "learning_rate": 1.250820774041203e-05, "loss": 0.1568, "step": 6826 }, { "epoch": 4.3678822776711455, "grad_norm": 1.3040907126383787, "learning_rate": 1.2506201772016755e-05, "loss": 0.1735, "step": 6827 }, { "epoch": 4.36852207293666, "grad_norm": 1.355706172033495, "learning_rate": 1.2504195696009633e-05, "loss": 0.1759, "step": 6828 }, { "epoch": 4.369161868202175, "grad_norm": 1.2618148698615779, "learning_rate": 1.2502189512476799e-05, "loss": 0.1773, "step": 6829 }, { "epoch": 4.36980166346769, "grad_norm": 1.1174202144283534, "learning_rate": 1.25001832215044e-05, "loss": 0.1529, "step": 6830 }, { "epoch": 4.370441458733206, "grad_norm": 1.2883117065224035, "learning_rate": 1.2498176823178581e-05, "loss": 0.1701, "step": 6831 }, { "epoch": 4.37108125399872, "grad_norm": 1.193858150870418, "learning_rate": 1.2496170317585493e-05, "loss": 0.1462, "step": 6832 }, { "epoch": 4.371721049264235, "grad_norm": 1.263332340733078, "learning_rate": 1.2494163704811293e-05, "loss": 0.151, "step": 6833 }, { "epoch": 4.372360844529751, "grad_norm": 1.300172175302495, "learning_rate": 1.2492156984942139e-05, "loss": 0.1661, "step": 6834 }, { "epoch": 4.373000639795266, "grad_norm": 1.2042462548380544, "learning_rate": 1.2490150158064198e-05, "loss": 0.1608, "step": 6835 }, { "epoch": 4.37364043506078, "grad_norm": 1.4138279887612801, "learning_rate": 1.248814322426364e-05, "loss": 0.1972, "step": 6836 }, { "epoch": 4.3742802303262955, "grad_norm": 1.322145710150349, "learning_rate": 1.2486136183626635e-05, "loss": 0.1724, "step": 6837 }, { "epoch": 4.374920025591811, "grad_norm": 1.2367037539897001, "learning_rate": 1.248412903623937e-05, "loss": 0.1652, "step": 6838 }, { "epoch": 4.375559820857326, "grad_norm": 1.2632985408857742, "learning_rate": 1.2482121782188018e-05, "loss": 0.1698, "step": 6839 }, { "epoch": 4.3761996161228405, "grad_norm": 1.2342290631409403, "learning_rate": 1.2480114421558775e-05, "loss": 0.1585, "step": 6840 }, { "epoch": 4.376839411388356, "grad_norm": 1.260042580100536, "learning_rate": 1.2478106954437828e-05, "loss": 0.1699, "step": 6841 }, { "epoch": 4.377479206653871, "grad_norm": 1.3267130141600574, "learning_rate": 1.247609938091138e-05, "loss": 0.1751, "step": 6842 }, { "epoch": 4.378119001919385, "grad_norm": 1.2383227508560486, "learning_rate": 1.2474091701065626e-05, "loss": 0.1646, "step": 6843 }, { "epoch": 4.378758797184901, "grad_norm": 1.210073053855163, "learning_rate": 1.2472083914986776e-05, "loss": 0.1589, "step": 6844 }, { "epoch": 4.379398592450416, "grad_norm": 1.346183321251074, "learning_rate": 1.2470076022761041e-05, "loss": 0.1704, "step": 6845 }, { "epoch": 4.380038387715931, "grad_norm": 1.4432484177268847, "learning_rate": 1.2468068024474635e-05, "loss": 0.1842, "step": 6846 }, { "epoch": 4.380678182981446, "grad_norm": 1.3667574904810322, "learning_rate": 1.2466059920213777e-05, "loss": 0.1858, "step": 6847 }, { "epoch": 4.381317978246961, "grad_norm": 1.316661333915213, "learning_rate": 1.2464051710064695e-05, "loss": 0.1687, "step": 6848 }, { "epoch": 4.381957773512476, "grad_norm": 1.2597558611465935, "learning_rate": 1.2462043394113615e-05, "loss": 0.1663, "step": 6849 }, { "epoch": 4.3825975687779914, "grad_norm": 1.2431338462135408, "learning_rate": 1.2460034972446772e-05, "loss": 0.1508, "step": 6850 }, { "epoch": 4.383237364043506, "grad_norm": 1.3093041016947258, "learning_rate": 1.2458026445150404e-05, "loss": 0.1706, "step": 6851 }, { "epoch": 4.383877159309021, "grad_norm": 1.315074383739553, "learning_rate": 1.245601781231075e-05, "loss": 0.1727, "step": 6852 }, { "epoch": 4.384516954574536, "grad_norm": 1.2453369263981846, "learning_rate": 1.2454009074014063e-05, "loss": 0.1534, "step": 6853 }, { "epoch": 4.385156749840051, "grad_norm": 1.2490206226390033, "learning_rate": 1.245200023034659e-05, "loss": 0.1585, "step": 6854 }, { "epoch": 4.385796545105566, "grad_norm": 1.2093577784281266, "learning_rate": 1.244999128139459e-05, "loss": 0.1507, "step": 6855 }, { "epoch": 4.386436340371081, "grad_norm": 1.3544833346600558, "learning_rate": 1.2447982227244323e-05, "loss": 0.1729, "step": 6856 }, { "epoch": 4.387076135636597, "grad_norm": 1.31095618524397, "learning_rate": 1.2445973067982051e-05, "loss": 0.1729, "step": 6857 }, { "epoch": 4.387715930902111, "grad_norm": 1.210312362442315, "learning_rate": 1.2443963803694051e-05, "loss": 0.1552, "step": 6858 }, { "epoch": 4.388355726167626, "grad_norm": 1.2488070097292823, "learning_rate": 1.2441954434466588e-05, "loss": 0.1558, "step": 6859 }, { "epoch": 4.3889955214331415, "grad_norm": 1.1483057513355728, "learning_rate": 1.2439944960385949e-05, "loss": 0.1604, "step": 6860 }, { "epoch": 4.389635316698657, "grad_norm": 1.4516511101462912, "learning_rate": 1.2437935381538411e-05, "loss": 0.1736, "step": 6861 }, { "epoch": 4.390275111964171, "grad_norm": 1.2792909733126636, "learning_rate": 1.2435925698010266e-05, "loss": 0.1717, "step": 6862 }, { "epoch": 4.3909149072296865, "grad_norm": 1.3508078560068852, "learning_rate": 1.2433915909887804e-05, "loss": 0.1751, "step": 6863 }, { "epoch": 4.391554702495202, "grad_norm": 1.1825417708842803, "learning_rate": 1.243190601725732e-05, "loss": 0.1524, "step": 6864 }, { "epoch": 4.392194497760716, "grad_norm": 1.3929173196953355, "learning_rate": 1.2429896020205124e-05, "loss": 0.1693, "step": 6865 }, { "epoch": 4.392834293026231, "grad_norm": 1.2546229893685923, "learning_rate": 1.2427885918817511e-05, "loss": 0.1667, "step": 6866 }, { "epoch": 4.393474088291747, "grad_norm": 1.2023859328633049, "learning_rate": 1.2425875713180795e-05, "loss": 0.1563, "step": 6867 }, { "epoch": 4.394113883557262, "grad_norm": 1.285065602405152, "learning_rate": 1.2423865403381295e-05, "loss": 0.1733, "step": 6868 }, { "epoch": 4.394753678822776, "grad_norm": 1.2422042623915444, "learning_rate": 1.2421854989505322e-05, "loss": 0.1611, "step": 6869 }, { "epoch": 4.395393474088292, "grad_norm": 1.2155693525985936, "learning_rate": 1.2419844471639207e-05, "loss": 0.1649, "step": 6870 }, { "epoch": 4.396033269353807, "grad_norm": 1.2137633226088003, "learning_rate": 1.2417833849869271e-05, "loss": 0.1652, "step": 6871 }, { "epoch": 4.396673064619322, "grad_norm": 1.2781845562101384, "learning_rate": 1.2415823124281855e-05, "loss": 0.1632, "step": 6872 }, { "epoch": 4.3973128598848366, "grad_norm": 1.2589378357478556, "learning_rate": 1.241381229496329e-05, "loss": 0.1629, "step": 6873 }, { "epoch": 4.397952655150352, "grad_norm": 1.2976223997808503, "learning_rate": 1.2411801361999918e-05, "loss": 0.1643, "step": 6874 }, { "epoch": 4.398592450415867, "grad_norm": 1.3278294371776715, "learning_rate": 1.2409790325478085e-05, "loss": 0.1636, "step": 6875 }, { "epoch": 4.399232245681382, "grad_norm": 1.1403989697771275, "learning_rate": 1.2407779185484143e-05, "loss": 0.158, "step": 6876 }, { "epoch": 4.399872040946897, "grad_norm": 1.3349320756092018, "learning_rate": 1.2405767942104447e-05, "loss": 0.1704, "step": 6877 }, { "epoch": 4.400511836212412, "grad_norm": 1.2955257070792685, "learning_rate": 1.2403756595425354e-05, "loss": 0.1674, "step": 6878 }, { "epoch": 4.401151631477927, "grad_norm": 1.2511435470237648, "learning_rate": 1.2401745145533228e-05, "loss": 0.1538, "step": 6879 }, { "epoch": 4.401791426743442, "grad_norm": 1.1449451903299972, "learning_rate": 1.239973359251444e-05, "loss": 0.1451, "step": 6880 }, { "epoch": 4.402431222008957, "grad_norm": 1.2686447781400916, "learning_rate": 1.239772193645536e-05, "loss": 0.1661, "step": 6881 }, { "epoch": 4.403071017274472, "grad_norm": 1.3138863376654937, "learning_rate": 1.2395710177442365e-05, "loss": 0.16, "step": 6882 }, { "epoch": 4.4037108125399875, "grad_norm": 1.3576670605004424, "learning_rate": 1.2393698315561836e-05, "loss": 0.1664, "step": 6883 }, { "epoch": 4.404350607805502, "grad_norm": 1.4047109778260731, "learning_rate": 1.2391686350900158e-05, "loss": 0.1646, "step": 6884 }, { "epoch": 4.404990403071017, "grad_norm": 1.2309399423044156, "learning_rate": 1.2389674283543725e-05, "loss": 0.1531, "step": 6885 }, { "epoch": 4.4056301983365325, "grad_norm": 1.2200553697314704, "learning_rate": 1.2387662113578927e-05, "loss": 0.1581, "step": 6886 }, { "epoch": 4.406269993602048, "grad_norm": 1.4826056627639794, "learning_rate": 1.238564984109217e-05, "loss": 0.1816, "step": 6887 }, { "epoch": 4.406909788867562, "grad_norm": 1.2890781875187056, "learning_rate": 1.2383637466169848e-05, "loss": 0.1542, "step": 6888 }, { "epoch": 4.407549584133077, "grad_norm": 1.295186466674108, "learning_rate": 1.2381624988898373e-05, "loss": 0.1673, "step": 6889 }, { "epoch": 4.408189379398593, "grad_norm": 1.2819757041257933, "learning_rate": 1.2379612409364161e-05, "loss": 0.1599, "step": 6890 }, { "epoch": 4.408829174664108, "grad_norm": 1.3357530824417445, "learning_rate": 1.2377599727653622e-05, "loss": 0.1661, "step": 6891 }, { "epoch": 4.409468969929622, "grad_norm": 1.3163816678765172, "learning_rate": 1.2375586943853181e-05, "loss": 0.1641, "step": 6892 }, { "epoch": 4.410108765195138, "grad_norm": 1.318521629233444, "learning_rate": 1.237357405804926e-05, "loss": 0.1779, "step": 6893 }, { "epoch": 4.410748560460653, "grad_norm": 1.3629784691070546, "learning_rate": 1.2371561070328295e-05, "loss": 0.1804, "step": 6894 }, { "epoch": 4.411388355726167, "grad_norm": 1.3323363604436473, "learning_rate": 1.2369547980776715e-05, "loss": 0.1738, "step": 6895 }, { "epoch": 4.4120281509916826, "grad_norm": 1.4251549294173402, "learning_rate": 1.2367534789480961e-05, "loss": 0.1768, "step": 6896 }, { "epoch": 4.412667946257198, "grad_norm": 1.3002869009738873, "learning_rate": 1.2365521496527476e-05, "loss": 0.1701, "step": 6897 }, { "epoch": 4.413307741522713, "grad_norm": 1.434318827788126, "learning_rate": 1.2363508102002702e-05, "loss": 0.1804, "step": 6898 }, { "epoch": 4.4139475367882275, "grad_norm": 1.2387077783402254, "learning_rate": 1.2361494605993094e-05, "loss": 0.165, "step": 6899 }, { "epoch": 4.414587332053743, "grad_norm": 1.2928386761212465, "learning_rate": 1.235948100858511e-05, "loss": 0.1684, "step": 6900 }, { "epoch": 4.415227127319258, "grad_norm": 1.3049179193140608, "learning_rate": 1.235746730986521e-05, "loss": 0.1665, "step": 6901 }, { "epoch": 4.415866922584772, "grad_norm": 1.3778586324929707, "learning_rate": 1.2355453509919857e-05, "loss": 0.1798, "step": 6902 }, { "epoch": 4.416506717850288, "grad_norm": 1.241695765421609, "learning_rate": 1.235343960883552e-05, "loss": 0.1733, "step": 6903 }, { "epoch": 4.417146513115803, "grad_norm": 1.1821379616093832, "learning_rate": 1.2351425606698673e-05, "loss": 0.1567, "step": 6904 }, { "epoch": 4.417786308381318, "grad_norm": 1.228122347004859, "learning_rate": 1.2349411503595793e-05, "loss": 0.1668, "step": 6905 }, { "epoch": 4.418426103646833, "grad_norm": 1.2094385270230734, "learning_rate": 1.234739729961336e-05, "loss": 0.1604, "step": 6906 }, { "epoch": 4.419065898912348, "grad_norm": 1.1925618089536631, "learning_rate": 1.2345382994837865e-05, "loss": 0.1503, "step": 6907 }, { "epoch": 4.419705694177863, "grad_norm": 1.3064933433637063, "learning_rate": 1.2343368589355796e-05, "loss": 0.1708, "step": 6908 }, { "epoch": 4.4203454894433785, "grad_norm": 1.3002409724706367, "learning_rate": 1.2341354083253648e-05, "loss": 0.175, "step": 6909 }, { "epoch": 4.420985284708893, "grad_norm": 1.2075563933451168, "learning_rate": 1.233933947661792e-05, "loss": 0.1594, "step": 6910 }, { "epoch": 4.421625079974408, "grad_norm": 1.2805700977267842, "learning_rate": 1.2337324769535116e-05, "loss": 0.1655, "step": 6911 }, { "epoch": 4.422264875239923, "grad_norm": 1.2521274198681895, "learning_rate": 1.2335309962091748e-05, "loss": 0.1592, "step": 6912 }, { "epoch": 4.422904670505439, "grad_norm": 1.2301984376929873, "learning_rate": 1.2333295054374322e-05, "loss": 0.1536, "step": 6913 }, { "epoch": 4.423544465770953, "grad_norm": 1.312449793112651, "learning_rate": 1.2331280046469354e-05, "loss": 0.167, "step": 6914 }, { "epoch": 4.424184261036468, "grad_norm": 1.3739145242899378, "learning_rate": 1.2329264938463371e-05, "loss": 0.1806, "step": 6915 }, { "epoch": 4.424824056301984, "grad_norm": 1.331388851466966, "learning_rate": 1.2327249730442894e-05, "loss": 0.1683, "step": 6916 }, { "epoch": 4.425463851567498, "grad_norm": 1.226475649474268, "learning_rate": 1.2325234422494454e-05, "loss": 0.1627, "step": 6917 }, { "epoch": 4.426103646833013, "grad_norm": 1.208215345440213, "learning_rate": 1.2323219014704586e-05, "loss": 0.1634, "step": 6918 }, { "epoch": 4.4267434420985285, "grad_norm": 1.2432917277673494, "learning_rate": 1.2321203507159822e-05, "loss": 0.1609, "step": 6919 }, { "epoch": 4.427383237364044, "grad_norm": 1.3711762030954913, "learning_rate": 1.2319187899946713e-05, "loss": 0.1657, "step": 6920 }, { "epoch": 4.428023032629558, "grad_norm": 1.3169561908048855, "learning_rate": 1.2317172193151796e-05, "loss": 0.1686, "step": 6921 }, { "epoch": 4.4286628278950735, "grad_norm": 1.2238311218014706, "learning_rate": 1.2315156386861633e-05, "loss": 0.1564, "step": 6922 }, { "epoch": 4.429302623160589, "grad_norm": 1.387131012027359, "learning_rate": 1.2313140481162768e-05, "loss": 0.1673, "step": 6923 }, { "epoch": 4.429942418426104, "grad_norm": 1.38640441105928, "learning_rate": 1.2311124476141767e-05, "loss": 0.1788, "step": 6924 }, { "epoch": 4.430582213691618, "grad_norm": 1.3252546258421662, "learning_rate": 1.2309108371885193e-05, "loss": 0.1683, "step": 6925 }, { "epoch": 4.431222008957134, "grad_norm": 1.1608716486667061, "learning_rate": 1.2307092168479613e-05, "loss": 0.1526, "step": 6926 }, { "epoch": 4.431861804222649, "grad_norm": 1.4411749157218883, "learning_rate": 1.23050758660116e-05, "loss": 0.1721, "step": 6927 }, { "epoch": 4.432501599488164, "grad_norm": 1.152307468246122, "learning_rate": 1.2303059464567728e-05, "loss": 0.1491, "step": 6928 }, { "epoch": 4.433141394753679, "grad_norm": 1.1980127339624764, "learning_rate": 1.2301042964234578e-05, "loss": 0.153, "step": 6929 }, { "epoch": 4.433781190019194, "grad_norm": 1.3341528722611944, "learning_rate": 1.229902636509874e-05, "loss": 0.1699, "step": 6930 }, { "epoch": 4.434420985284709, "grad_norm": 1.343284312530153, "learning_rate": 1.2297009667246798e-05, "loss": 0.1693, "step": 6931 }, { "epoch": 4.435060780550224, "grad_norm": 1.3180106425312297, "learning_rate": 1.2294992870765348e-05, "loss": 0.1754, "step": 6932 }, { "epoch": 4.435700575815739, "grad_norm": 1.3227858823172762, "learning_rate": 1.2292975975740987e-05, "loss": 0.1689, "step": 6933 }, { "epoch": 4.436340371081254, "grad_norm": 1.353554610356207, "learning_rate": 1.2290958982260312e-05, "loss": 0.1667, "step": 6934 }, { "epoch": 4.436980166346769, "grad_norm": 1.279563515488318, "learning_rate": 1.2288941890409937e-05, "loss": 0.1627, "step": 6935 }, { "epoch": 4.437619961612284, "grad_norm": 1.346267911624519, "learning_rate": 1.2286924700276468e-05, "loss": 0.1766, "step": 6936 }, { "epoch": 4.438259756877799, "grad_norm": 1.2695528043010018, "learning_rate": 1.2284907411946523e-05, "loss": 0.1683, "step": 6937 }, { "epoch": 4.438899552143314, "grad_norm": 1.2139962731237914, "learning_rate": 1.2282890025506715e-05, "loss": 0.1666, "step": 6938 }, { "epoch": 4.439539347408829, "grad_norm": 1.31445692693682, "learning_rate": 1.2280872541043669e-05, "loss": 0.1683, "step": 6939 }, { "epoch": 4.440179142674344, "grad_norm": 1.3115038394492107, "learning_rate": 1.2278854958644018e-05, "loss": 0.1704, "step": 6940 }, { "epoch": 4.440818937939859, "grad_norm": 1.3162170775601298, "learning_rate": 1.2276837278394387e-05, "loss": 0.1666, "step": 6941 }, { "epoch": 4.4414587332053745, "grad_norm": 1.225150197470924, "learning_rate": 1.2274819500381414e-05, "loss": 0.1629, "step": 6942 }, { "epoch": 4.442098528470889, "grad_norm": 1.368226025638296, "learning_rate": 1.2272801624691737e-05, "loss": 0.1706, "step": 6943 }, { "epoch": 4.442738323736404, "grad_norm": 1.3831973170498955, "learning_rate": 1.2270783651412002e-05, "loss": 0.1676, "step": 6944 }, { "epoch": 4.4433781190019195, "grad_norm": 1.3840081458061113, "learning_rate": 1.2268765580628857e-05, "loss": 0.1667, "step": 6945 }, { "epoch": 4.444017914267435, "grad_norm": 1.3311781978795443, "learning_rate": 1.226674741242895e-05, "loss": 0.1722, "step": 6946 }, { "epoch": 4.444657709532949, "grad_norm": 1.2008569047610356, "learning_rate": 1.2264729146898946e-05, "loss": 0.1533, "step": 6947 }, { "epoch": 4.445297504798464, "grad_norm": 1.283902303746145, "learning_rate": 1.22627107841255e-05, "loss": 0.1621, "step": 6948 }, { "epoch": 4.44593730006398, "grad_norm": 1.2716530886594002, "learning_rate": 1.2260692324195277e-05, "loss": 0.174, "step": 6949 }, { "epoch": 4.446577095329495, "grad_norm": 1.3324986107523882, "learning_rate": 1.2258673767194948e-05, "loss": 0.1682, "step": 6950 }, { "epoch": 4.447216890595009, "grad_norm": 1.2702616774965694, "learning_rate": 1.2256655113211182e-05, "loss": 0.175, "step": 6951 }, { "epoch": 4.447856685860525, "grad_norm": 1.1491916185097646, "learning_rate": 1.2254636362330665e-05, "loss": 0.1571, "step": 6952 }, { "epoch": 4.44849648112604, "grad_norm": 1.245028460537723, "learning_rate": 1.2252617514640068e-05, "loss": 0.151, "step": 6953 }, { "epoch": 4.449136276391554, "grad_norm": 1.3328989784299936, "learning_rate": 1.2250598570226087e-05, "loss": 0.1732, "step": 6954 }, { "epoch": 4.44977607165707, "grad_norm": 1.3622510909032366, "learning_rate": 1.2248579529175403e-05, "loss": 0.1823, "step": 6955 }, { "epoch": 4.450415866922585, "grad_norm": 1.2771495492930363, "learning_rate": 1.2246560391574718e-05, "loss": 0.1818, "step": 6956 }, { "epoch": 4.4510556621881, "grad_norm": 1.2716771540229548, "learning_rate": 1.2244541157510723e-05, "loss": 0.159, "step": 6957 }, { "epoch": 4.4516954574536145, "grad_norm": 1.4367594752164792, "learning_rate": 1.2242521827070125e-05, "loss": 0.1767, "step": 6958 }, { "epoch": 4.45233525271913, "grad_norm": 1.213704380314264, "learning_rate": 1.2240502400339627e-05, "loss": 0.1593, "step": 6959 }, { "epoch": 4.452975047984645, "grad_norm": 1.2612117035869859, "learning_rate": 1.2238482877405942e-05, "loss": 0.1571, "step": 6960 }, { "epoch": 4.45361484325016, "grad_norm": 1.279118043328474, "learning_rate": 1.2236463258355783e-05, "loss": 0.1656, "step": 6961 }, { "epoch": 4.454254638515675, "grad_norm": 1.1904200970201662, "learning_rate": 1.2234443543275874e-05, "loss": 0.1451, "step": 6962 }, { "epoch": 4.45489443378119, "grad_norm": 1.186827645164808, "learning_rate": 1.2232423732252932e-05, "loss": 0.1569, "step": 6963 }, { "epoch": 4.455534229046705, "grad_norm": 1.4614692367055644, "learning_rate": 1.2230403825373687e-05, "loss": 0.1731, "step": 6964 }, { "epoch": 4.4561740243122205, "grad_norm": 1.4188119655521536, "learning_rate": 1.2228383822724867e-05, "loss": 0.1714, "step": 6965 }, { "epoch": 4.456813819577735, "grad_norm": 1.3539818650289113, "learning_rate": 1.222636372439321e-05, "loss": 0.1671, "step": 6966 }, { "epoch": 4.45745361484325, "grad_norm": 1.254735765780995, "learning_rate": 1.2224343530465457e-05, "loss": 0.1675, "step": 6967 }, { "epoch": 4.4580934101087655, "grad_norm": 1.393482452755131, "learning_rate": 1.2222323241028346e-05, "loss": 0.1646, "step": 6968 }, { "epoch": 4.45873320537428, "grad_norm": 1.2902414268054885, "learning_rate": 1.2220302856168634e-05, "loss": 0.1731, "step": 6969 }, { "epoch": 4.459373000639795, "grad_norm": 1.1073930968047463, "learning_rate": 1.221828237597306e-05, "loss": 0.1632, "step": 6970 }, { "epoch": 4.46001279590531, "grad_norm": 1.4029395505299997, "learning_rate": 1.2216261800528392e-05, "loss": 0.1744, "step": 6971 }, { "epoch": 4.460652591170826, "grad_norm": 1.2854794439988921, "learning_rate": 1.2214241129921384e-05, "loss": 0.1591, "step": 6972 }, { "epoch": 4.46129238643634, "grad_norm": 1.290461218458804, "learning_rate": 1.2212220364238798e-05, "loss": 0.1633, "step": 6973 }, { "epoch": 4.461932181701855, "grad_norm": 1.2718726421527726, "learning_rate": 1.2210199503567409e-05, "loss": 0.1738, "step": 6974 }, { "epoch": 4.462571976967371, "grad_norm": 1.2692998965115294, "learning_rate": 1.220817854799398e-05, "loss": 0.1672, "step": 6975 }, { "epoch": 4.463211772232885, "grad_norm": 1.3203589926489452, "learning_rate": 1.2206157497605296e-05, "loss": 0.1685, "step": 6976 }, { "epoch": 4.4638515674984, "grad_norm": 1.1981872294672953, "learning_rate": 1.2204136352488132e-05, "loss": 0.1562, "step": 6977 }, { "epoch": 4.464491362763916, "grad_norm": 1.3721170376022456, "learning_rate": 1.2202115112729276e-05, "loss": 0.1713, "step": 6978 }, { "epoch": 4.465131158029431, "grad_norm": 1.2143600501767184, "learning_rate": 1.2200093778415512e-05, "loss": 0.162, "step": 6979 }, { "epoch": 4.465770953294945, "grad_norm": 1.237674523533755, "learning_rate": 1.2198072349633635e-05, "loss": 0.16, "step": 6980 }, { "epoch": 4.4664107485604605, "grad_norm": 1.1718184135974417, "learning_rate": 1.219605082647044e-05, "loss": 0.1519, "step": 6981 }, { "epoch": 4.467050543825976, "grad_norm": 1.340007692700591, "learning_rate": 1.2194029209012732e-05, "loss": 0.1648, "step": 6982 }, { "epoch": 4.467690339091491, "grad_norm": 1.2335464411641919, "learning_rate": 1.2192007497347311e-05, "loss": 0.1558, "step": 6983 }, { "epoch": 4.468330134357005, "grad_norm": 1.2685423361742327, "learning_rate": 1.2189985691560987e-05, "loss": 0.174, "step": 6984 }, { "epoch": 4.468969929622521, "grad_norm": 1.27009016503837, "learning_rate": 1.2187963791740574e-05, "loss": 0.1665, "step": 6985 }, { "epoch": 4.469609724888036, "grad_norm": 1.310841321534035, "learning_rate": 1.2185941797972889e-05, "loss": 0.1705, "step": 6986 }, { "epoch": 4.470249520153551, "grad_norm": 1.3486855488678053, "learning_rate": 1.2183919710344751e-05, "loss": 0.1769, "step": 6987 }, { "epoch": 4.470889315419066, "grad_norm": 1.313049896381712, "learning_rate": 1.2181897528942983e-05, "loss": 0.1674, "step": 6988 }, { "epoch": 4.471529110684581, "grad_norm": 1.2052173717378334, "learning_rate": 1.2179875253854419e-05, "loss": 0.1538, "step": 6989 }, { "epoch": 4.472168905950096, "grad_norm": 1.3246457237540696, "learning_rate": 1.2177852885165889e-05, "loss": 0.1721, "step": 6990 }, { "epoch": 4.472808701215611, "grad_norm": 1.2642741602754692, "learning_rate": 1.2175830422964228e-05, "loss": 0.1607, "step": 6991 }, { "epoch": 4.473448496481126, "grad_norm": 1.18308534264926, "learning_rate": 1.217380786733628e-05, "loss": 0.1556, "step": 6992 }, { "epoch": 4.474088291746641, "grad_norm": 1.2932463197282955, "learning_rate": 1.2171785218368892e-05, "loss": 0.169, "step": 6993 }, { "epoch": 4.474728087012156, "grad_norm": 1.2024831014984885, "learning_rate": 1.216976247614891e-05, "loss": 0.1503, "step": 6994 }, { "epoch": 4.475367882277671, "grad_norm": 1.2489699074537184, "learning_rate": 1.2167739640763183e-05, "loss": 0.1628, "step": 6995 }, { "epoch": 4.476007677543186, "grad_norm": 1.3832429094432834, "learning_rate": 1.2165716712298572e-05, "loss": 0.1841, "step": 6996 }, { "epoch": 4.476647472808701, "grad_norm": 1.2176658528886142, "learning_rate": 1.2163693690841943e-05, "loss": 0.1569, "step": 6997 }, { "epoch": 4.477287268074217, "grad_norm": 1.3303591909237424, "learning_rate": 1.2161670576480148e-05, "loss": 0.1724, "step": 6998 }, { "epoch": 4.477927063339731, "grad_norm": 1.252103515659782, "learning_rate": 1.2159647369300071e-05, "loss": 0.1615, "step": 6999 }, { "epoch": 4.478566858605246, "grad_norm": 1.2224758800636815, "learning_rate": 1.2157624069388574e-05, "loss": 0.1642, "step": 7000 }, { "epoch": 4.479206653870762, "grad_norm": 1.2316940438736754, "learning_rate": 1.215560067683254e-05, "loss": 0.1671, "step": 7001 }, { "epoch": 4.479846449136277, "grad_norm": 1.3363418167396328, "learning_rate": 1.2153577191718849e-05, "loss": 0.1817, "step": 7002 }, { "epoch": 4.480486244401791, "grad_norm": 1.1967979803809143, "learning_rate": 1.215155361413438e-05, "loss": 0.1532, "step": 7003 }, { "epoch": 4.4811260396673065, "grad_norm": 1.259978463461553, "learning_rate": 1.2149529944166027e-05, "loss": 0.1647, "step": 7004 }, { "epoch": 4.481765834932822, "grad_norm": 1.321982266043522, "learning_rate": 1.2147506181900681e-05, "loss": 0.1732, "step": 7005 }, { "epoch": 4.482405630198336, "grad_norm": 1.1955536733930663, "learning_rate": 1.214548232742524e-05, "loss": 0.1671, "step": 7006 }, { "epoch": 4.483045425463851, "grad_norm": 1.1710798007493888, "learning_rate": 1.2143458380826608e-05, "loss": 0.1608, "step": 7007 }, { "epoch": 4.483685220729367, "grad_norm": 1.2298403027119804, "learning_rate": 1.2141434342191683e-05, "loss": 0.1612, "step": 7008 }, { "epoch": 4.484325015994882, "grad_norm": 1.244879821359995, "learning_rate": 1.2139410211607379e-05, "loss": 0.1581, "step": 7009 }, { "epoch": 4.484964811260396, "grad_norm": 1.2134052931286352, "learning_rate": 1.2137385989160602e-05, "loss": 0.163, "step": 7010 }, { "epoch": 4.485604606525912, "grad_norm": 1.371736531209274, "learning_rate": 1.2135361674938276e-05, "loss": 0.1768, "step": 7011 }, { "epoch": 4.486244401791427, "grad_norm": 1.1906288800366063, "learning_rate": 1.2133337269027316e-05, "loss": 0.1508, "step": 7012 }, { "epoch": 4.486884197056942, "grad_norm": 1.334287081953613, "learning_rate": 1.2131312771514649e-05, "loss": 0.1665, "step": 7013 }, { "epoch": 4.487523992322457, "grad_norm": 1.2794973317098222, "learning_rate": 1.2129288182487207e-05, "loss": 0.156, "step": 7014 }, { "epoch": 4.488163787587972, "grad_norm": 1.319774819839097, "learning_rate": 1.2127263502031912e-05, "loss": 0.1614, "step": 7015 }, { "epoch": 4.488803582853487, "grad_norm": 1.3147895217345291, "learning_rate": 1.212523873023571e-05, "loss": 0.1709, "step": 7016 }, { "epoch": 4.4894433781190015, "grad_norm": 1.241529514155868, "learning_rate": 1.212321386718554e-05, "loss": 0.1626, "step": 7017 }, { "epoch": 4.490083173384517, "grad_norm": 1.3616090053077539, "learning_rate": 1.2121188912968339e-05, "loss": 0.1733, "step": 7018 }, { "epoch": 4.490722968650032, "grad_norm": 1.1349648857100403, "learning_rate": 1.211916386767106e-05, "loss": 0.1559, "step": 7019 }, { "epoch": 4.491362763915547, "grad_norm": 1.3509456931427375, "learning_rate": 1.2117138731380653e-05, "loss": 0.1758, "step": 7020 }, { "epoch": 4.492002559181062, "grad_norm": 1.3141321149751037, "learning_rate": 1.2115113504184077e-05, "loss": 0.1768, "step": 7021 }, { "epoch": 4.492642354446577, "grad_norm": 1.257773544029613, "learning_rate": 1.2113088186168289e-05, "loss": 0.164, "step": 7022 }, { "epoch": 4.493282149712092, "grad_norm": 1.2352543174163007, "learning_rate": 1.2111062777420255e-05, "loss": 0.1745, "step": 7023 }, { "epoch": 4.4939219449776076, "grad_norm": 1.2737469727241468, "learning_rate": 1.210903727802694e-05, "loss": 0.1612, "step": 7024 }, { "epoch": 4.494561740243122, "grad_norm": 1.183053303442891, "learning_rate": 1.2107011688075313e-05, "loss": 0.1575, "step": 7025 }, { "epoch": 4.495201535508637, "grad_norm": 1.2700919584820212, "learning_rate": 1.2104986007652355e-05, "loss": 0.166, "step": 7026 }, { "epoch": 4.4958413307741525, "grad_norm": 1.331920099710167, "learning_rate": 1.2102960236845043e-05, "loss": 0.1592, "step": 7027 }, { "epoch": 4.496481126039667, "grad_norm": 1.1640676085587551, "learning_rate": 1.2100934375740357e-05, "loss": 0.1551, "step": 7028 }, { "epoch": 4.497120921305182, "grad_norm": 1.2273199253353209, "learning_rate": 1.2098908424425288e-05, "loss": 0.1519, "step": 7029 }, { "epoch": 4.497760716570697, "grad_norm": 1.2826247116264757, "learning_rate": 1.2096882382986823e-05, "loss": 0.1572, "step": 7030 }, { "epoch": 4.498400511836213, "grad_norm": 1.3418705520180976, "learning_rate": 1.209485625151196e-05, "loss": 0.1709, "step": 7031 }, { "epoch": 4.499040307101727, "grad_norm": 1.2236313873649831, "learning_rate": 1.2092830030087697e-05, "loss": 0.1569, "step": 7032 }, { "epoch": 4.499680102367242, "grad_norm": 1.2920634693199466, "learning_rate": 1.2090803718801032e-05, "loss": 0.1626, "step": 7033 }, { "epoch": 4.500319897632758, "grad_norm": 1.290635005192406, "learning_rate": 1.2088777317738978e-05, "loss": 0.1662, "step": 7034 }, { "epoch": 4.500959692898273, "grad_norm": 1.211328006886769, "learning_rate": 1.208675082698854e-05, "loss": 0.1692, "step": 7035 }, { "epoch": 4.501599488163787, "grad_norm": 1.3097494708839612, "learning_rate": 1.2084724246636735e-05, "loss": 0.1642, "step": 7036 }, { "epoch": 4.502239283429303, "grad_norm": 1.3060921087642374, "learning_rate": 1.2082697576770577e-05, "loss": 0.1622, "step": 7037 }, { "epoch": 4.502879078694818, "grad_norm": 1.2204266754313413, "learning_rate": 1.2080670817477092e-05, "loss": 0.1588, "step": 7038 }, { "epoch": 4.503518873960333, "grad_norm": 1.3848891718918468, "learning_rate": 1.2078643968843303e-05, "loss": 0.1858, "step": 7039 }, { "epoch": 4.5041586692258475, "grad_norm": 1.3456924705278237, "learning_rate": 1.2076617030956237e-05, "loss": 0.167, "step": 7040 }, { "epoch": 4.504798464491363, "grad_norm": 1.261480298899435, "learning_rate": 1.2074590003902934e-05, "loss": 0.163, "step": 7041 }, { "epoch": 4.505438259756878, "grad_norm": 1.3236389923782066, "learning_rate": 1.207256288777042e-05, "loss": 0.1768, "step": 7042 }, { "epoch": 4.5060780550223924, "grad_norm": 1.2771303913795937, "learning_rate": 1.2070535682645748e-05, "loss": 0.1622, "step": 7043 }, { "epoch": 4.506717850287908, "grad_norm": 1.2666527393305664, "learning_rate": 1.2068508388615957e-05, "loss": 0.1646, "step": 7044 }, { "epoch": 4.507357645553423, "grad_norm": 1.2117543478299262, "learning_rate": 1.2066481005768091e-05, "loss": 0.1602, "step": 7045 }, { "epoch": 4.507997440818938, "grad_norm": 1.1672302062821638, "learning_rate": 1.2064453534189212e-05, "loss": 0.1572, "step": 7046 }, { "epoch": 4.508637236084453, "grad_norm": 1.6924842915731908, "learning_rate": 1.2062425973966367e-05, "loss": 0.1676, "step": 7047 }, { "epoch": 4.509277031349968, "grad_norm": 1.216341783157426, "learning_rate": 1.206039832518662e-05, "loss": 0.161, "step": 7048 }, { "epoch": 4.509916826615483, "grad_norm": 1.314421859964503, "learning_rate": 1.2058370587937037e-05, "loss": 0.1718, "step": 7049 }, { "epoch": 4.510556621880998, "grad_norm": 1.2684333583858578, "learning_rate": 1.2056342762304679e-05, "loss": 0.1589, "step": 7050 }, { "epoch": 4.511196417146513, "grad_norm": 1.350560351679826, "learning_rate": 1.2054314848376622e-05, "loss": 0.1731, "step": 7051 }, { "epoch": 4.511836212412028, "grad_norm": 1.2591833109914146, "learning_rate": 1.2052286846239938e-05, "loss": 0.165, "step": 7052 }, { "epoch": 4.512476007677543, "grad_norm": 1.2783946804772186, "learning_rate": 1.205025875598171e-05, "loss": 0.1651, "step": 7053 }, { "epoch": 4.513115802943059, "grad_norm": 1.3177430469715117, "learning_rate": 1.204823057768902e-05, "loss": 0.1563, "step": 7054 }, { "epoch": 4.513755598208573, "grad_norm": 1.3973490133099966, "learning_rate": 1.2046202311448947e-05, "loss": 0.166, "step": 7055 }, { "epoch": 4.514395393474088, "grad_norm": 1.2186043170823904, "learning_rate": 1.204417395734859e-05, "loss": 0.1678, "step": 7056 }, { "epoch": 4.515035188739604, "grad_norm": 1.2766120549468358, "learning_rate": 1.2042145515475038e-05, "loss": 0.1577, "step": 7057 }, { "epoch": 4.515674984005118, "grad_norm": 1.2597521312485045, "learning_rate": 1.204011698591539e-05, "loss": 0.1591, "step": 7058 }, { "epoch": 4.516314779270633, "grad_norm": 1.3502187092983478, "learning_rate": 1.203808836875675e-05, "loss": 0.176, "step": 7059 }, { "epoch": 4.516954574536149, "grad_norm": 1.2078780458196372, "learning_rate": 1.2036059664086222e-05, "loss": 0.1513, "step": 7060 }, { "epoch": 4.517594369801664, "grad_norm": 1.2897890171053035, "learning_rate": 1.2034030871990911e-05, "loss": 0.1734, "step": 7061 }, { "epoch": 4.518234165067178, "grad_norm": 1.2353544788301984, "learning_rate": 1.2032001992557936e-05, "loss": 0.1478, "step": 7062 }, { "epoch": 4.5188739603326935, "grad_norm": 1.2584042900534202, "learning_rate": 1.2029973025874406e-05, "loss": 0.1733, "step": 7063 }, { "epoch": 4.519513755598209, "grad_norm": 1.307909351852944, "learning_rate": 1.2027943972027448e-05, "loss": 0.1781, "step": 7064 }, { "epoch": 4.520153550863723, "grad_norm": 1.3566392064599981, "learning_rate": 1.2025914831104179e-05, "loss": 0.1737, "step": 7065 }, { "epoch": 4.520793346129238, "grad_norm": 1.2419716995306587, "learning_rate": 1.2023885603191736e-05, "loss": 0.1641, "step": 7066 }, { "epoch": 4.521433141394754, "grad_norm": 1.4505087300593222, "learning_rate": 1.2021856288377243e-05, "loss": 0.1749, "step": 7067 }, { "epoch": 4.522072936660269, "grad_norm": 1.2969832575464308, "learning_rate": 1.2019826886747841e-05, "loss": 0.1671, "step": 7068 }, { "epoch": 4.522712731925783, "grad_norm": 1.3531917179581106, "learning_rate": 1.2017797398390663e-05, "loss": 0.1532, "step": 7069 }, { "epoch": 4.523352527191299, "grad_norm": 1.1390949783421882, "learning_rate": 1.2015767823392851e-05, "loss": 0.149, "step": 7070 }, { "epoch": 4.523992322456814, "grad_norm": 1.2283386492435449, "learning_rate": 1.2013738161841559e-05, "loss": 0.1571, "step": 7071 }, { "epoch": 4.524632117722329, "grad_norm": 1.2028217546205673, "learning_rate": 1.201170841382393e-05, "loss": 0.1522, "step": 7072 }, { "epoch": 4.525271912987844, "grad_norm": 1.3403832962031108, "learning_rate": 1.200967857942712e-05, "loss": 0.166, "step": 7073 }, { "epoch": 4.525911708253359, "grad_norm": 1.2682191538349525, "learning_rate": 1.2007648658738288e-05, "loss": 0.1623, "step": 7074 }, { "epoch": 4.526551503518874, "grad_norm": 1.306160145702815, "learning_rate": 1.2005618651844593e-05, "loss": 0.1678, "step": 7075 }, { "epoch": 4.527191298784389, "grad_norm": 1.2997170375767155, "learning_rate": 1.20035885588332e-05, "loss": 0.1739, "step": 7076 }, { "epoch": 4.527831094049904, "grad_norm": 1.1826633067467602, "learning_rate": 1.2001558379791283e-05, "loss": 0.1579, "step": 7077 }, { "epoch": 4.528470889315419, "grad_norm": 1.3576511859846296, "learning_rate": 1.1999528114806005e-05, "loss": 0.1869, "step": 7078 }, { "epoch": 4.529110684580934, "grad_norm": 1.2177746104034013, "learning_rate": 1.199749776396455e-05, "loss": 0.1658, "step": 7079 }, { "epoch": 4.529750479846449, "grad_norm": 1.2328812831433085, "learning_rate": 1.199546732735409e-05, "loss": 0.1533, "step": 7080 }, { "epoch": 4.530390275111964, "grad_norm": 1.4122045144749291, "learning_rate": 1.1993436805061815e-05, "loss": 0.1811, "step": 7081 }, { "epoch": 4.531030070377479, "grad_norm": 1.1775485001241448, "learning_rate": 1.1991406197174911e-05, "loss": 0.1574, "step": 7082 }, { "epoch": 4.531669865642995, "grad_norm": 1.3731200328965636, "learning_rate": 1.1989375503780568e-05, "loss": 0.1632, "step": 7083 }, { "epoch": 4.532309660908509, "grad_norm": 1.2866756730393598, "learning_rate": 1.1987344724965977e-05, "loss": 0.1577, "step": 7084 }, { "epoch": 4.532949456174024, "grad_norm": 1.3421850912868962, "learning_rate": 1.1985313860818336e-05, "loss": 0.167, "step": 7085 }, { "epoch": 4.5335892514395395, "grad_norm": 1.2455792431846349, "learning_rate": 1.1983282911424856e-05, "loss": 0.161, "step": 7086 }, { "epoch": 4.534229046705054, "grad_norm": 1.3270472445680572, "learning_rate": 1.198125187687273e-05, "loss": 0.1717, "step": 7087 }, { "epoch": 4.534868841970569, "grad_norm": 1.2164016169236158, "learning_rate": 1.1979220757249174e-05, "loss": 0.1713, "step": 7088 }, { "epoch": 4.535508637236084, "grad_norm": 1.2418853659500824, "learning_rate": 1.1977189552641404e-05, "loss": 0.1687, "step": 7089 }, { "epoch": 4.5361484325016, "grad_norm": 1.3074479288731895, "learning_rate": 1.1975158263136626e-05, "loss": 0.1679, "step": 7090 }, { "epoch": 4.536788227767115, "grad_norm": 1.2849730800125712, "learning_rate": 1.1973126888822072e-05, "loss": 0.1647, "step": 7091 }, { "epoch": 4.537428023032629, "grad_norm": 1.314916057342568, "learning_rate": 1.1971095429784954e-05, "loss": 0.1748, "step": 7092 }, { "epoch": 4.538067818298145, "grad_norm": 1.2120017828161542, "learning_rate": 1.1969063886112507e-05, "loss": 0.1648, "step": 7093 }, { "epoch": 4.53870761356366, "grad_norm": 1.2236283641047692, "learning_rate": 1.1967032257891957e-05, "loss": 0.1586, "step": 7094 }, { "epoch": 4.539347408829174, "grad_norm": 1.4151160582345172, "learning_rate": 1.1965000545210544e-05, "loss": 0.1925, "step": 7095 }, { "epoch": 4.53998720409469, "grad_norm": 1.2113011590087195, "learning_rate": 1.1962968748155501e-05, "loss": 0.1589, "step": 7096 }, { "epoch": 4.540626999360205, "grad_norm": 1.183998772335485, "learning_rate": 1.1960936866814073e-05, "loss": 0.1603, "step": 7097 }, { "epoch": 4.54126679462572, "grad_norm": 1.3655817281622957, "learning_rate": 1.1958904901273503e-05, "loss": 0.1613, "step": 7098 }, { "epoch": 4.5419065898912345, "grad_norm": 1.2696011443794921, "learning_rate": 1.1956872851621043e-05, "loss": 0.1635, "step": 7099 }, { "epoch": 4.54254638515675, "grad_norm": 1.2354838953927143, "learning_rate": 1.195484071794394e-05, "loss": 0.1596, "step": 7100 }, { "epoch": 4.543186180422265, "grad_norm": 1.2906120217124126, "learning_rate": 1.1952808500329459e-05, "loss": 0.1731, "step": 7101 }, { "epoch": 4.5438259756877795, "grad_norm": 1.244169971152135, "learning_rate": 1.195077619886485e-05, "loss": 0.1552, "step": 7102 }, { "epoch": 4.544465770953295, "grad_norm": 1.323909259521261, "learning_rate": 1.1948743813637383e-05, "loss": 0.1754, "step": 7103 }, { "epoch": 4.54510556621881, "grad_norm": 1.2380934819331988, "learning_rate": 1.1946711344734326e-05, "loss": 0.1706, "step": 7104 }, { "epoch": 4.545745361484325, "grad_norm": 1.365772539603914, "learning_rate": 1.194467879224294e-05, "loss": 0.173, "step": 7105 }, { "epoch": 4.54638515674984, "grad_norm": 1.2274910990885781, "learning_rate": 1.1942646156250513e-05, "loss": 0.165, "step": 7106 }, { "epoch": 4.547024952015355, "grad_norm": 1.3315097251793957, "learning_rate": 1.1940613436844315e-05, "loss": 0.1713, "step": 7107 }, { "epoch": 4.54766474728087, "grad_norm": 1.2845749970580915, "learning_rate": 1.1938580634111627e-05, "loss": 0.1745, "step": 7108 }, { "epoch": 4.5483045425463855, "grad_norm": 1.4223008981503382, "learning_rate": 1.1936547748139733e-05, "loss": 0.1683, "step": 7109 }, { "epoch": 4.5489443378119, "grad_norm": 1.3110828141061073, "learning_rate": 1.1934514779015923e-05, "loss": 0.1635, "step": 7110 }, { "epoch": 4.549584133077415, "grad_norm": 1.2447024158599536, "learning_rate": 1.1932481726827492e-05, "loss": 0.1697, "step": 7111 }, { "epoch": 4.55022392834293, "grad_norm": 1.4485559166905138, "learning_rate": 1.1930448591661733e-05, "loss": 0.1913, "step": 7112 }, { "epoch": 4.550863723608446, "grad_norm": 1.2641414056236657, "learning_rate": 1.1928415373605943e-05, "loss": 0.167, "step": 7113 }, { "epoch": 4.55150351887396, "grad_norm": 1.201751051822829, "learning_rate": 1.1926382072747434e-05, "loss": 0.16, "step": 7114 }, { "epoch": 4.552143314139475, "grad_norm": 1.3420243201850932, "learning_rate": 1.1924348689173497e-05, "loss": 0.1844, "step": 7115 }, { "epoch": 4.552783109404991, "grad_norm": 1.1740856218086704, "learning_rate": 1.1922315222971458e-05, "loss": 0.1602, "step": 7116 }, { "epoch": 4.553422904670505, "grad_norm": 1.394395980225942, "learning_rate": 1.1920281674228617e-05, "loss": 0.1841, "step": 7117 }, { "epoch": 4.55406269993602, "grad_norm": 1.2614193317764488, "learning_rate": 1.19182480430323e-05, "loss": 0.1731, "step": 7118 }, { "epoch": 4.554702495201536, "grad_norm": 1.3049758182843998, "learning_rate": 1.1916214329469822e-05, "loss": 0.1724, "step": 7119 }, { "epoch": 4.555342290467051, "grad_norm": 1.2271457670370243, "learning_rate": 1.191418053362851e-05, "loss": 0.1565, "step": 7120 }, { "epoch": 4.555982085732565, "grad_norm": 1.2793000312068172, "learning_rate": 1.1912146655595693e-05, "loss": 0.1569, "step": 7121 }, { "epoch": 4.5566218809980805, "grad_norm": 1.2896041244123233, "learning_rate": 1.19101126954587e-05, "loss": 0.1683, "step": 7122 }, { "epoch": 4.557261676263596, "grad_norm": 1.3358121388994126, "learning_rate": 1.1908078653304864e-05, "loss": 0.1727, "step": 7123 }, { "epoch": 4.557901471529111, "grad_norm": 1.4051077319052898, "learning_rate": 1.1906044529221525e-05, "loss": 0.1756, "step": 7124 }, { "epoch": 4.5585412667946255, "grad_norm": 1.1687928744880292, "learning_rate": 1.1904010323296022e-05, "loss": 0.1639, "step": 7125 }, { "epoch": 4.559181062060141, "grad_norm": 1.1882651682616778, "learning_rate": 1.1901976035615708e-05, "loss": 0.1649, "step": 7126 }, { "epoch": 4.559820857325656, "grad_norm": 1.2200669260943564, "learning_rate": 1.1899941666267922e-05, "loss": 0.1595, "step": 7127 }, { "epoch": 4.560460652591171, "grad_norm": 1.2568685637632568, "learning_rate": 1.1897907215340023e-05, "loss": 0.1628, "step": 7128 }, { "epoch": 4.561100447856686, "grad_norm": 1.2596815601775881, "learning_rate": 1.1895872682919364e-05, "loss": 0.1514, "step": 7129 }, { "epoch": 4.561740243122201, "grad_norm": 1.2916005191911966, "learning_rate": 1.1893838069093304e-05, "loss": 0.1637, "step": 7130 }, { "epoch": 4.562380038387716, "grad_norm": 1.4284195508771897, "learning_rate": 1.1891803373949208e-05, "loss": 0.1744, "step": 7131 }, { "epoch": 4.563019833653231, "grad_norm": 1.2969525275676392, "learning_rate": 1.1889768597574438e-05, "loss": 0.1752, "step": 7132 }, { "epoch": 4.563659628918746, "grad_norm": 1.1401308631091334, "learning_rate": 1.1887733740056369e-05, "loss": 0.1564, "step": 7133 }, { "epoch": 4.564299424184261, "grad_norm": 1.2479513610682287, "learning_rate": 1.1885698801482366e-05, "loss": 0.1692, "step": 7134 }, { "epoch": 4.564939219449776, "grad_norm": 1.2502738095387291, "learning_rate": 1.1883663781939814e-05, "loss": 0.1584, "step": 7135 }, { "epoch": 4.565579014715291, "grad_norm": 1.4784182266994832, "learning_rate": 1.188162868151609e-05, "loss": 0.1821, "step": 7136 }, { "epoch": 4.566218809980806, "grad_norm": 1.3057947877044145, "learning_rate": 1.1879593500298579e-05, "loss": 0.1669, "step": 7137 }, { "epoch": 4.566858605246321, "grad_norm": 1.2567204988372391, "learning_rate": 1.1877558238374666e-05, "loss": 0.1598, "step": 7138 }, { "epoch": 4.567498400511836, "grad_norm": 1.109965025290623, "learning_rate": 1.1875522895831742e-05, "loss": 0.153, "step": 7139 }, { "epoch": 4.568138195777351, "grad_norm": 1.4113074272737853, "learning_rate": 1.1873487472757198e-05, "loss": 0.178, "step": 7140 }, { "epoch": 4.568777991042866, "grad_norm": 1.4084060152705284, "learning_rate": 1.1871451969238437e-05, "loss": 0.1822, "step": 7141 }, { "epoch": 4.569417786308382, "grad_norm": 1.2564638692913486, "learning_rate": 1.1869416385362858e-05, "loss": 0.1519, "step": 7142 }, { "epoch": 4.570057581573896, "grad_norm": 1.2524657118448717, "learning_rate": 1.1867380721217864e-05, "loss": 0.1669, "step": 7143 }, { "epoch": 4.570697376839411, "grad_norm": 1.4153015703881633, "learning_rate": 1.1865344976890865e-05, "loss": 0.1706, "step": 7144 }, { "epoch": 4.5713371721049265, "grad_norm": 1.2516808508495867, "learning_rate": 1.186330915246927e-05, "loss": 0.1628, "step": 7145 }, { "epoch": 4.571976967370442, "grad_norm": 1.271503270254983, "learning_rate": 1.1861273248040494e-05, "loss": 0.1553, "step": 7146 }, { "epoch": 4.572616762635956, "grad_norm": 1.2989848169408966, "learning_rate": 1.1859237263691956e-05, "loss": 0.1627, "step": 7147 }, { "epoch": 4.5732565579014715, "grad_norm": 1.3735544784662914, "learning_rate": 1.1857201199511077e-05, "loss": 0.1619, "step": 7148 }, { "epoch": 4.573896353166987, "grad_norm": 1.2406439009260357, "learning_rate": 1.1855165055585282e-05, "loss": 0.1596, "step": 7149 }, { "epoch": 4.574536148432502, "grad_norm": 1.2942290518729005, "learning_rate": 1.1853128832001999e-05, "loss": 0.1768, "step": 7150 }, { "epoch": 4.575175943698016, "grad_norm": 1.3444806486587721, "learning_rate": 1.1851092528848662e-05, "loss": 0.1802, "step": 7151 }, { "epoch": 4.575815738963532, "grad_norm": 1.3334895179831163, "learning_rate": 1.1849056146212707e-05, "loss": 0.1828, "step": 7152 }, { "epoch": 4.576455534229047, "grad_norm": 1.302033265048259, "learning_rate": 1.1847019684181568e-05, "loss": 0.1636, "step": 7153 }, { "epoch": 4.577095329494561, "grad_norm": 1.5517292994687466, "learning_rate": 1.1844983142842687e-05, "loss": 0.1933, "step": 7154 }, { "epoch": 4.577735124760077, "grad_norm": 1.1844268247669032, "learning_rate": 1.1842946522283514e-05, "loss": 0.1577, "step": 7155 }, { "epoch": 4.578374920025592, "grad_norm": 1.1904686586840938, "learning_rate": 1.1840909822591495e-05, "loss": 0.1441, "step": 7156 }, { "epoch": 4.579014715291107, "grad_norm": 1.4433778584522214, "learning_rate": 1.1838873043854083e-05, "loss": 0.1752, "step": 7157 }, { "epoch": 4.5796545105566215, "grad_norm": 1.2747431326620373, "learning_rate": 1.1836836186158736e-05, "loss": 0.1564, "step": 7158 }, { "epoch": 4.580294305822137, "grad_norm": 1.3298085149237586, "learning_rate": 1.1834799249592908e-05, "loss": 0.174, "step": 7159 }, { "epoch": 4.580934101087652, "grad_norm": 1.2860594096022397, "learning_rate": 1.1832762234244064e-05, "loss": 0.1576, "step": 7160 }, { "epoch": 4.581573896353167, "grad_norm": 1.1701012400489965, "learning_rate": 1.1830725140199672e-05, "loss": 0.1486, "step": 7161 }, { "epoch": 4.582213691618682, "grad_norm": 1.2644318447698237, "learning_rate": 1.1828687967547196e-05, "loss": 0.1607, "step": 7162 }, { "epoch": 4.582853486884197, "grad_norm": 1.3073386261407962, "learning_rate": 1.1826650716374115e-05, "loss": 0.1729, "step": 7163 }, { "epoch": 4.583493282149712, "grad_norm": 1.3690766852683118, "learning_rate": 1.1824613386767898e-05, "loss": 0.1647, "step": 7164 }, { "epoch": 4.584133077415228, "grad_norm": 1.1991133836718704, "learning_rate": 1.1822575978816028e-05, "loss": 0.1634, "step": 7165 }, { "epoch": 4.584772872680742, "grad_norm": 1.2387035363769752, "learning_rate": 1.182053849260599e-05, "loss": 0.1596, "step": 7166 }, { "epoch": 4.585412667946257, "grad_norm": 1.3813091561494335, "learning_rate": 1.1818500928225269e-05, "loss": 0.168, "step": 7167 }, { "epoch": 4.5860524632117725, "grad_norm": 1.177131606849001, "learning_rate": 1.1816463285761352e-05, "loss": 0.1542, "step": 7168 }, { "epoch": 4.586692258477287, "grad_norm": 1.3626231690574822, "learning_rate": 1.181442556530173e-05, "loss": 0.1752, "step": 7169 }, { "epoch": 4.587332053742802, "grad_norm": 1.258126740595179, "learning_rate": 1.1812387766933903e-05, "loss": 0.1586, "step": 7170 }, { "epoch": 4.5879718490083174, "grad_norm": 1.2695049590618157, "learning_rate": 1.1810349890745372e-05, "loss": 0.1564, "step": 7171 }, { "epoch": 4.588611644273833, "grad_norm": 1.2897245181567127, "learning_rate": 1.1808311936823634e-05, "loss": 0.1686, "step": 7172 }, { "epoch": 4.589251439539347, "grad_norm": 1.289540126892877, "learning_rate": 1.1806273905256198e-05, "loss": 0.1601, "step": 7173 }, { "epoch": 4.589891234804862, "grad_norm": 1.2171714184661848, "learning_rate": 1.1804235796130577e-05, "loss": 0.1619, "step": 7174 }, { "epoch": 4.590531030070378, "grad_norm": 1.1681207116671464, "learning_rate": 1.1802197609534276e-05, "loss": 0.1523, "step": 7175 }, { "epoch": 4.591170825335892, "grad_norm": 1.3132126340281527, "learning_rate": 1.1800159345554818e-05, "loss": 0.1717, "step": 7176 }, { "epoch": 4.591810620601407, "grad_norm": 1.280091571626261, "learning_rate": 1.179812100427972e-05, "loss": 0.1628, "step": 7177 }, { "epoch": 4.592450415866923, "grad_norm": 1.2256214984385028, "learning_rate": 1.1796082585796505e-05, "loss": 0.151, "step": 7178 }, { "epoch": 4.593090211132438, "grad_norm": 1.2389895270112334, "learning_rate": 1.1794044090192696e-05, "loss": 0.1569, "step": 7179 }, { "epoch": 4.593730006397953, "grad_norm": 1.2536061430638636, "learning_rate": 1.1792005517555827e-05, "loss": 0.1621, "step": 7180 }, { "epoch": 4.5943698016634675, "grad_norm": 1.3987495729505588, "learning_rate": 1.1789966867973429e-05, "loss": 0.1755, "step": 7181 }, { "epoch": 4.595009596928983, "grad_norm": 1.1697022964557855, "learning_rate": 1.1787928141533038e-05, "loss": 0.161, "step": 7182 }, { "epoch": 4.595649392194498, "grad_norm": 1.3322643860603778, "learning_rate": 1.1785889338322195e-05, "loss": 0.1649, "step": 7183 }, { "epoch": 4.5962891874600125, "grad_norm": 1.3389951950084402, "learning_rate": 1.1783850458428436e-05, "loss": 0.1576, "step": 7184 }, { "epoch": 4.596928982725528, "grad_norm": 1.2940769921205735, "learning_rate": 1.1781811501939313e-05, "loss": 0.1621, "step": 7185 }, { "epoch": 4.597568777991043, "grad_norm": 1.3097834333163132, "learning_rate": 1.1779772468942377e-05, "loss": 0.1748, "step": 7186 }, { "epoch": 4.598208573256558, "grad_norm": 1.416189675450451, "learning_rate": 1.1777733359525172e-05, "loss": 0.1837, "step": 7187 }, { "epoch": 4.598848368522073, "grad_norm": 1.2112834742465144, "learning_rate": 1.1775694173775265e-05, "loss": 0.1581, "step": 7188 }, { "epoch": 4.599488163787588, "grad_norm": 1.4852180717019712, "learning_rate": 1.1773654911780206e-05, "loss": 0.1844, "step": 7189 }, { "epoch": 4.600127959053103, "grad_norm": 1.3516605906059347, "learning_rate": 1.177161557362756e-05, "loss": 0.172, "step": 7190 }, { "epoch": 4.600767754318618, "grad_norm": 1.2420332290384988, "learning_rate": 1.1769576159404891e-05, "loss": 0.1662, "step": 7191 }, { "epoch": 4.601407549584133, "grad_norm": 1.2707585190267157, "learning_rate": 1.1767536669199771e-05, "loss": 0.1726, "step": 7192 }, { "epoch": 4.602047344849648, "grad_norm": 1.1596756322246433, "learning_rate": 1.1765497103099772e-05, "loss": 0.1532, "step": 7193 }, { "epoch": 4.6026871401151634, "grad_norm": 1.2653363040227905, "learning_rate": 1.1763457461192464e-05, "loss": 0.1641, "step": 7194 }, { "epoch": 4.603326935380678, "grad_norm": 1.2007301643496997, "learning_rate": 1.1761417743565434e-05, "loss": 0.1625, "step": 7195 }, { "epoch": 4.603966730646193, "grad_norm": 1.1513524473570476, "learning_rate": 1.1759377950306257e-05, "loss": 0.1465, "step": 7196 }, { "epoch": 4.604606525911708, "grad_norm": 1.245216538048645, "learning_rate": 1.1757338081502522e-05, "loss": 0.1623, "step": 7197 }, { "epoch": 4.605246321177224, "grad_norm": 1.2116507684319902, "learning_rate": 1.1755298137241815e-05, "loss": 0.1665, "step": 7198 }, { "epoch": 4.605886116442738, "grad_norm": 1.2068349422248612, "learning_rate": 1.1753258117611727e-05, "loss": 0.1615, "step": 7199 }, { "epoch": 4.606525911708253, "grad_norm": 1.3311390872382574, "learning_rate": 1.1751218022699857e-05, "loss": 0.1631, "step": 7200 }, { "epoch": 4.607165706973769, "grad_norm": 1.1866877418522823, "learning_rate": 1.1749177852593798e-05, "loss": 0.1609, "step": 7201 }, { "epoch": 4.607805502239284, "grad_norm": 1.2352544093023838, "learning_rate": 1.1747137607381154e-05, "loss": 0.1662, "step": 7202 }, { "epoch": 4.608445297504798, "grad_norm": 1.1834892016982863, "learning_rate": 1.174509728714953e-05, "loss": 0.1599, "step": 7203 }, { "epoch": 4.6090850927703135, "grad_norm": 1.282258035922965, "learning_rate": 1.1743056891986533e-05, "loss": 0.159, "step": 7204 }, { "epoch": 4.609724888035829, "grad_norm": 1.3081407753149092, "learning_rate": 1.1741016421979773e-05, "loss": 0.1786, "step": 7205 }, { "epoch": 4.610364683301343, "grad_norm": 1.2428006930172997, "learning_rate": 1.1738975877216865e-05, "loss": 0.1538, "step": 7206 }, { "epoch": 4.6110044785668585, "grad_norm": 1.2836057506956298, "learning_rate": 1.1736935257785425e-05, "loss": 0.1633, "step": 7207 }, { "epoch": 4.611644273832374, "grad_norm": 1.2925877979604647, "learning_rate": 1.1734894563773077e-05, "loss": 0.1715, "step": 7208 }, { "epoch": 4.612284069097889, "grad_norm": 1.1805116164751497, "learning_rate": 1.173285379526744e-05, "loss": 0.1538, "step": 7209 }, { "epoch": 4.612923864363403, "grad_norm": 1.2850096712177992, "learning_rate": 1.1730812952356145e-05, "loss": 0.1633, "step": 7210 }, { "epoch": 4.613563659628919, "grad_norm": 1.2701571656069697, "learning_rate": 1.1728772035126821e-05, "loss": 0.1745, "step": 7211 }, { "epoch": 4.614203454894434, "grad_norm": 1.348116600593694, "learning_rate": 1.1726731043667102e-05, "loss": 0.1633, "step": 7212 }, { "epoch": 4.614843250159948, "grad_norm": 1.2569465391831642, "learning_rate": 1.1724689978064622e-05, "loss": 0.1588, "step": 7213 }, { "epoch": 4.615483045425464, "grad_norm": 1.2298333317285672, "learning_rate": 1.1722648838407023e-05, "loss": 0.1632, "step": 7214 }, { "epoch": 4.616122840690979, "grad_norm": 1.3075101523662904, "learning_rate": 1.1720607624781948e-05, "loss": 0.1592, "step": 7215 }, { "epoch": 4.616762635956494, "grad_norm": 1.2394552061513433, "learning_rate": 1.171856633727704e-05, "loss": 0.1498, "step": 7216 }, { "epoch": 4.617402431222009, "grad_norm": 1.3700596250107733, "learning_rate": 1.171652497597995e-05, "loss": 0.1683, "step": 7217 }, { "epoch": 4.618042226487524, "grad_norm": 1.301066788838369, "learning_rate": 1.1714483540978333e-05, "loss": 0.1584, "step": 7218 }, { "epoch": 4.618682021753039, "grad_norm": 1.3358026922102912, "learning_rate": 1.1712442032359845e-05, "loss": 0.1644, "step": 7219 }, { "epoch": 4.619321817018554, "grad_norm": 1.3199066429861357, "learning_rate": 1.171040045021214e-05, "loss": 0.1649, "step": 7220 }, { "epoch": 4.619961612284069, "grad_norm": 1.252883924449363, "learning_rate": 1.1708358794622883e-05, "loss": 0.1669, "step": 7221 }, { "epoch": 4.620601407549584, "grad_norm": 1.1859300055263833, "learning_rate": 1.1706317065679736e-05, "loss": 0.1529, "step": 7222 }, { "epoch": 4.621241202815099, "grad_norm": 1.3546743416053866, "learning_rate": 1.170427526347037e-05, "loss": 0.1673, "step": 7223 }, { "epoch": 4.621880998080615, "grad_norm": 1.2244840468137197, "learning_rate": 1.1702233388082456e-05, "loss": 0.1694, "step": 7224 }, { "epoch": 4.622520793346129, "grad_norm": 1.3303912164851202, "learning_rate": 1.170019143960367e-05, "loss": 0.1615, "step": 7225 }, { "epoch": 4.623160588611644, "grad_norm": 1.412260187380645, "learning_rate": 1.1698149418121685e-05, "loss": 0.1825, "step": 7226 }, { "epoch": 4.6238003838771595, "grad_norm": 1.2860052925314698, "learning_rate": 1.1696107323724189e-05, "loss": 0.1656, "step": 7227 }, { "epoch": 4.624440179142674, "grad_norm": 1.2971280074745815, "learning_rate": 1.1694065156498859e-05, "loss": 0.1763, "step": 7228 }, { "epoch": 4.625079974408189, "grad_norm": 1.3766376127754627, "learning_rate": 1.1692022916533386e-05, "loss": 0.1697, "step": 7229 }, { "epoch": 4.6257197696737045, "grad_norm": 1.2851626108931367, "learning_rate": 1.1689980603915459e-05, "loss": 0.166, "step": 7230 }, { "epoch": 4.62635956493922, "grad_norm": 1.3165213295118217, "learning_rate": 1.1687938218732768e-05, "loss": 0.1616, "step": 7231 }, { "epoch": 4.626999360204734, "grad_norm": 1.4168014234191697, "learning_rate": 1.1685895761073014e-05, "loss": 0.1746, "step": 7232 }, { "epoch": 4.627639155470249, "grad_norm": 1.167209823556835, "learning_rate": 1.1683853231023896e-05, "loss": 0.1582, "step": 7233 }, { "epoch": 4.628278950735765, "grad_norm": 1.2229070406623075, "learning_rate": 1.1681810628673116e-05, "loss": 0.1578, "step": 7234 }, { "epoch": 4.62891874600128, "grad_norm": 1.2921951908557485, "learning_rate": 1.1679767954108379e-05, "loss": 0.1612, "step": 7235 }, { "epoch": 4.629558541266794, "grad_norm": 1.2814364381304486, "learning_rate": 1.1677725207417391e-05, "loss": 0.1684, "step": 7236 }, { "epoch": 4.63019833653231, "grad_norm": 1.333633466476441, "learning_rate": 1.167568238868787e-05, "loss": 0.1661, "step": 7237 }, { "epoch": 4.630838131797825, "grad_norm": 1.2417348496368632, "learning_rate": 1.1673639498007527e-05, "loss": 0.1637, "step": 7238 }, { "epoch": 4.63147792706334, "grad_norm": 1.2440873208402856, "learning_rate": 1.1671596535464083e-05, "loss": 0.1622, "step": 7239 }, { "epoch": 4.6321177223288545, "grad_norm": 1.4064327337063862, "learning_rate": 1.1669553501145257e-05, "loss": 0.1857, "step": 7240 }, { "epoch": 4.63275751759437, "grad_norm": 1.3203610653840547, "learning_rate": 1.1667510395138773e-05, "loss": 0.1705, "step": 7241 }, { "epoch": 4.633397312859885, "grad_norm": 1.352754758944774, "learning_rate": 1.1665467217532357e-05, "loss": 0.1704, "step": 7242 }, { "epoch": 4.6340371081253995, "grad_norm": 1.250150037159819, "learning_rate": 1.1663423968413743e-05, "loss": 0.1686, "step": 7243 }, { "epoch": 4.634676903390915, "grad_norm": 1.2173348396568953, "learning_rate": 1.1661380647870661e-05, "loss": 0.1612, "step": 7244 }, { "epoch": 4.63531669865643, "grad_norm": 1.2471476787523432, "learning_rate": 1.1659337255990853e-05, "loss": 0.1611, "step": 7245 }, { "epoch": 4.635956493921945, "grad_norm": 1.1436864241717881, "learning_rate": 1.1657293792862051e-05, "loss": 0.1576, "step": 7246 }, { "epoch": 4.63659628918746, "grad_norm": 1.2176504336284248, "learning_rate": 1.1655250258572005e-05, "loss": 0.1635, "step": 7247 }, { "epoch": 4.637236084452975, "grad_norm": 1.294513823560547, "learning_rate": 1.1653206653208455e-05, "loss": 0.1774, "step": 7248 }, { "epoch": 4.63787587971849, "grad_norm": 1.426298230551526, "learning_rate": 1.1651162976859155e-05, "loss": 0.1811, "step": 7249 }, { "epoch": 4.638515674984005, "grad_norm": 1.3079148605767539, "learning_rate": 1.1649119229611854e-05, "loss": 0.1634, "step": 7250 }, { "epoch": 4.63915547024952, "grad_norm": 1.4351207326014805, "learning_rate": 1.1647075411554302e-05, "loss": 0.1845, "step": 7251 }, { "epoch": 4.639795265515035, "grad_norm": 1.299101964878041, "learning_rate": 1.1645031522774264e-05, "loss": 0.1716, "step": 7252 }, { "epoch": 4.6404350607805505, "grad_norm": 1.3075541903024506, "learning_rate": 1.16429875633595e-05, "loss": 0.1716, "step": 7253 }, { "epoch": 4.641074856046066, "grad_norm": 1.2213351171021265, "learning_rate": 1.164094353339777e-05, "loss": 0.1637, "step": 7254 }, { "epoch": 4.64171465131158, "grad_norm": 1.4180337703922976, "learning_rate": 1.1638899432976846e-05, "loss": 0.1732, "step": 7255 }, { "epoch": 4.642354446577095, "grad_norm": 1.263300566903925, "learning_rate": 1.1636855262184499e-05, "loss": 0.1549, "step": 7256 }, { "epoch": 4.642994241842611, "grad_norm": 1.2199476454585156, "learning_rate": 1.1634811021108494e-05, "loss": 0.1538, "step": 7257 }, { "epoch": 4.643634037108125, "grad_norm": 1.294616983877901, "learning_rate": 1.1632766709836614e-05, "loss": 0.17, "step": 7258 }, { "epoch": 4.64427383237364, "grad_norm": 1.1483304146845152, "learning_rate": 1.1630722328456635e-05, "loss": 0.1564, "step": 7259 }, { "epoch": 4.644913627639156, "grad_norm": 1.2645258067029808, "learning_rate": 1.162867787705634e-05, "loss": 0.1624, "step": 7260 }, { "epoch": 4.645553422904671, "grad_norm": 1.4292004587118368, "learning_rate": 1.1626633355723514e-05, "loss": 0.1833, "step": 7261 }, { "epoch": 4.646193218170185, "grad_norm": 1.3638450754204927, "learning_rate": 1.1624588764545944e-05, "loss": 0.1847, "step": 7262 }, { "epoch": 4.6468330134357005, "grad_norm": 1.4444181284227973, "learning_rate": 1.1622544103611424e-05, "loss": 0.1783, "step": 7263 }, { "epoch": 4.647472808701216, "grad_norm": 1.2699424315007821, "learning_rate": 1.1620499373007747e-05, "loss": 0.1685, "step": 7264 }, { "epoch": 4.64811260396673, "grad_norm": 1.201372416021439, "learning_rate": 1.161845457282271e-05, "loss": 0.1542, "step": 7265 }, { "epoch": 4.6487523992322455, "grad_norm": 1.2657908226499146, "learning_rate": 1.1616409703144109e-05, "loss": 0.161, "step": 7266 }, { "epoch": 4.649392194497761, "grad_norm": 1.3105792406333179, "learning_rate": 1.1614364764059756e-05, "loss": 0.1651, "step": 7267 }, { "epoch": 4.650031989763276, "grad_norm": 1.2913219529604196, "learning_rate": 1.1612319755657446e-05, "loss": 0.1709, "step": 7268 }, { "epoch": 4.65067178502879, "grad_norm": 1.3421916884031204, "learning_rate": 1.1610274678024997e-05, "loss": 0.1669, "step": 7269 }, { "epoch": 4.651311580294306, "grad_norm": 1.2702142058877812, "learning_rate": 1.160822953125022e-05, "loss": 0.1588, "step": 7270 }, { "epoch": 4.651951375559821, "grad_norm": 1.2405483881226829, "learning_rate": 1.1606184315420924e-05, "loss": 0.1495, "step": 7271 }, { "epoch": 4.652591170825336, "grad_norm": 1.1659102684903369, "learning_rate": 1.1604139030624935e-05, "loss": 0.1535, "step": 7272 }, { "epoch": 4.653230966090851, "grad_norm": 1.3689009415631217, "learning_rate": 1.1602093676950064e-05, "loss": 0.1854, "step": 7273 }, { "epoch": 4.653870761356366, "grad_norm": 1.3853765036090715, "learning_rate": 1.1600048254484143e-05, "loss": 0.1762, "step": 7274 }, { "epoch": 4.654510556621881, "grad_norm": 1.2850122072687842, "learning_rate": 1.1598002763314998e-05, "loss": 0.1654, "step": 7275 }, { "epoch": 4.6551503518873965, "grad_norm": 1.1259030103651784, "learning_rate": 1.1595957203530453e-05, "loss": 0.1471, "step": 7276 }, { "epoch": 4.655790147152911, "grad_norm": 1.297496772292951, "learning_rate": 1.1593911575218346e-05, "loss": 0.1596, "step": 7277 }, { "epoch": 4.656429942418426, "grad_norm": 1.5165018697576667, "learning_rate": 1.1591865878466514e-05, "loss": 0.1919, "step": 7278 }, { "epoch": 4.657069737683941, "grad_norm": 1.2142869212768614, "learning_rate": 1.158982011336279e-05, "loss": 0.162, "step": 7279 }, { "epoch": 4.657709532949456, "grad_norm": 1.2910164070411252, "learning_rate": 1.1587774279995021e-05, "loss": 0.1499, "step": 7280 }, { "epoch": 4.658349328214971, "grad_norm": 1.3292300344864316, "learning_rate": 1.1585728378451046e-05, "loss": 0.1561, "step": 7281 }, { "epoch": 4.658989123480486, "grad_norm": 1.2858521711875481, "learning_rate": 1.1583682408818717e-05, "loss": 0.1514, "step": 7282 }, { "epoch": 4.659628918746002, "grad_norm": 1.3285643750134584, "learning_rate": 1.158163637118588e-05, "loss": 0.1772, "step": 7283 }, { "epoch": 4.660268714011516, "grad_norm": 1.4026531094542212, "learning_rate": 1.1579590265640392e-05, "loss": 0.1686, "step": 7284 }, { "epoch": 4.660908509277031, "grad_norm": 1.3438485238708249, "learning_rate": 1.157754409227011e-05, "loss": 0.1764, "step": 7285 }, { "epoch": 4.6615483045425465, "grad_norm": 1.3279977091111999, "learning_rate": 1.1575497851162892e-05, "loss": 0.1785, "step": 7286 }, { "epoch": 4.662188099808061, "grad_norm": 1.199064628414574, "learning_rate": 1.1573451542406597e-05, "loss": 0.1626, "step": 7287 }, { "epoch": 4.662827895073576, "grad_norm": 1.1637673307568062, "learning_rate": 1.157140516608909e-05, "loss": 0.1537, "step": 7288 }, { "epoch": 4.6634676903390915, "grad_norm": 1.2181524087295765, "learning_rate": 1.156935872229824e-05, "loss": 0.1735, "step": 7289 }, { "epoch": 4.664107485604607, "grad_norm": 1.15882624166534, "learning_rate": 1.156731221112192e-05, "loss": 0.1469, "step": 7290 }, { "epoch": 4.664747280870122, "grad_norm": 1.2434357758051153, "learning_rate": 1.1565265632648002e-05, "loss": 0.1582, "step": 7291 }, { "epoch": 4.665387076135636, "grad_norm": 1.3284939583944169, "learning_rate": 1.1563218986964361e-05, "loss": 0.1719, "step": 7292 }, { "epoch": 4.666026871401152, "grad_norm": 1.3416491398012969, "learning_rate": 1.1561172274158877e-05, "loss": 0.172, "step": 7293 }, { "epoch": 4.666666666666667, "grad_norm": 1.2597754202868057, "learning_rate": 1.1559125494319433e-05, "loss": 0.1655, "step": 7294 }, { "epoch": 4.667306461932181, "grad_norm": 1.2896265072384459, "learning_rate": 1.1557078647533916e-05, "loss": 0.1709, "step": 7295 }, { "epoch": 4.667946257197697, "grad_norm": 1.22671582985237, "learning_rate": 1.155503173389021e-05, "loss": 0.1564, "step": 7296 }, { "epoch": 4.668586052463212, "grad_norm": 1.329718281071188, "learning_rate": 1.1552984753476209e-05, "loss": 0.1693, "step": 7297 }, { "epoch": 4.669225847728727, "grad_norm": 1.2964509158304565, "learning_rate": 1.1550937706379804e-05, "loss": 0.1589, "step": 7298 }, { "epoch": 4.669865642994242, "grad_norm": 1.3721849754952358, "learning_rate": 1.154889059268889e-05, "loss": 0.1838, "step": 7299 }, { "epoch": 4.670505438259757, "grad_norm": 1.1552116467566096, "learning_rate": 1.1546843412491373e-05, "loss": 0.1543, "step": 7300 }, { "epoch": 4.671145233525272, "grad_norm": 1.3658630902140905, "learning_rate": 1.1544796165875152e-05, "loss": 0.1802, "step": 7301 }, { "epoch": 4.6717850287907865, "grad_norm": 1.3392344182258062, "learning_rate": 1.1542748852928129e-05, "loss": 0.163, "step": 7302 }, { "epoch": 4.672424824056302, "grad_norm": 1.187127051564948, "learning_rate": 1.1540701473738216e-05, "loss": 0.1679, "step": 7303 }, { "epoch": 4.673064619321817, "grad_norm": 1.1876840538193711, "learning_rate": 1.1538654028393321e-05, "loss": 0.1589, "step": 7304 }, { "epoch": 4.673704414587332, "grad_norm": 1.3327094694926682, "learning_rate": 1.153660651698136e-05, "loss": 0.1701, "step": 7305 }, { "epoch": 4.674344209852847, "grad_norm": 1.3836594908282722, "learning_rate": 1.1534558939590248e-05, "loss": 0.1768, "step": 7306 }, { "epoch": 4.674984005118362, "grad_norm": 1.4352634243206837, "learning_rate": 1.1532511296307907e-05, "loss": 0.19, "step": 7307 }, { "epoch": 4.675623800383877, "grad_norm": 1.3441156312738665, "learning_rate": 1.1530463587222253e-05, "loss": 0.1745, "step": 7308 }, { "epoch": 4.6762635956493925, "grad_norm": 1.2848088771041775, "learning_rate": 1.1528415812421222e-05, "loss": 0.1738, "step": 7309 }, { "epoch": 4.676903390914907, "grad_norm": 1.343835252813624, "learning_rate": 1.152636797199273e-05, "loss": 0.1843, "step": 7310 }, { "epoch": 4.677543186180422, "grad_norm": 1.1813708246059047, "learning_rate": 1.1524320066024713e-05, "loss": 0.1499, "step": 7311 }, { "epoch": 4.6781829814459375, "grad_norm": 1.3428227347510038, "learning_rate": 1.1522272094605103e-05, "loss": 0.1761, "step": 7312 }, { "epoch": 4.678822776711453, "grad_norm": 1.2471184677544154, "learning_rate": 1.1520224057821836e-05, "loss": 0.1567, "step": 7313 }, { "epoch": 4.679462571976967, "grad_norm": 1.234829098045305, "learning_rate": 1.1518175955762854e-05, "loss": 0.1534, "step": 7314 }, { "epoch": 4.680102367242482, "grad_norm": 1.2003844578457485, "learning_rate": 1.1516127788516098e-05, "loss": 0.1551, "step": 7315 }, { "epoch": 4.680742162507998, "grad_norm": 1.1194340986838032, "learning_rate": 1.1514079556169514e-05, "loss": 0.1391, "step": 7316 }, { "epoch": 4.681381957773512, "grad_norm": 1.2923917833061007, "learning_rate": 1.1512031258811045e-05, "loss": 0.1552, "step": 7317 }, { "epoch": 4.682021753039027, "grad_norm": 1.3412338486365087, "learning_rate": 1.150998289652864e-05, "loss": 0.1646, "step": 7318 }, { "epoch": 4.682661548304543, "grad_norm": 1.3105656467197582, "learning_rate": 1.1507934469410259e-05, "loss": 0.1766, "step": 7319 }, { "epoch": 4.683301343570058, "grad_norm": 1.234042553167532, "learning_rate": 1.1505885977543853e-05, "loss": 0.1547, "step": 7320 }, { "epoch": 4.683941138835572, "grad_norm": 1.19816957133605, "learning_rate": 1.150383742101738e-05, "loss": 0.1566, "step": 7321 }, { "epoch": 4.684580934101088, "grad_norm": 1.4313815423221916, "learning_rate": 1.1501788799918807e-05, "loss": 0.1747, "step": 7322 }, { "epoch": 4.685220729366603, "grad_norm": 1.2051967633743839, "learning_rate": 1.1499740114336092e-05, "loss": 0.1628, "step": 7323 }, { "epoch": 4.685860524632118, "grad_norm": 1.3602875947312016, "learning_rate": 1.1497691364357206e-05, "loss": 0.1693, "step": 7324 }, { "epoch": 4.6865003198976325, "grad_norm": 1.481417756209134, "learning_rate": 1.1495642550070118e-05, "loss": 0.1893, "step": 7325 }, { "epoch": 4.687140115163148, "grad_norm": 1.3995173215299648, "learning_rate": 1.1493593671562797e-05, "loss": 0.17, "step": 7326 }, { "epoch": 4.687779910428663, "grad_norm": 1.305974163575679, "learning_rate": 1.1491544728923222e-05, "loss": 0.1754, "step": 7327 }, { "epoch": 4.688419705694178, "grad_norm": 1.2999801443992731, "learning_rate": 1.148949572223937e-05, "loss": 0.1712, "step": 7328 }, { "epoch": 4.689059500959693, "grad_norm": 1.1937654013180505, "learning_rate": 1.148744665159922e-05, "loss": 0.1564, "step": 7329 }, { "epoch": 4.689699296225208, "grad_norm": 1.2138025814438147, "learning_rate": 1.1485397517090757e-05, "loss": 0.1602, "step": 7330 }, { "epoch": 4.690339091490723, "grad_norm": 1.129154403026307, "learning_rate": 1.1483348318801968e-05, "loss": 0.1455, "step": 7331 }, { "epoch": 4.690978886756238, "grad_norm": 1.1925174593782013, "learning_rate": 1.148129905682084e-05, "loss": 0.1613, "step": 7332 }, { "epoch": 4.691618682021753, "grad_norm": 1.3325727720197358, "learning_rate": 1.1479249731235365e-05, "loss": 0.1731, "step": 7333 }, { "epoch": 4.692258477287268, "grad_norm": 1.141781368205343, "learning_rate": 1.1477200342133537e-05, "loss": 0.1552, "step": 7334 }, { "epoch": 4.6928982725527835, "grad_norm": 1.3066939912255602, "learning_rate": 1.1475150889603358e-05, "loss": 0.1646, "step": 7335 }, { "epoch": 4.693538067818298, "grad_norm": 1.2748028952394261, "learning_rate": 1.1473101373732818e-05, "loss": 0.1688, "step": 7336 }, { "epoch": 4.694177863083813, "grad_norm": 1.3074674988153245, "learning_rate": 1.1471051794609931e-05, "loss": 0.1631, "step": 7337 }, { "epoch": 4.694817658349328, "grad_norm": 1.310772452403735, "learning_rate": 1.1469002152322692e-05, "loss": 0.1697, "step": 7338 }, { "epoch": 4.695457453614843, "grad_norm": 1.2556915638570836, "learning_rate": 1.146695244695912e-05, "loss": 0.1531, "step": 7339 }, { "epoch": 4.696097248880358, "grad_norm": 1.3212350176280896, "learning_rate": 1.1464902678607215e-05, "loss": 0.17, "step": 7340 }, { "epoch": 4.696737044145873, "grad_norm": 1.2616901949986448, "learning_rate": 1.1462852847354995e-05, "loss": 0.1573, "step": 7341 }, { "epoch": 4.697376839411389, "grad_norm": 1.2580488897431883, "learning_rate": 1.1460802953290475e-05, "loss": 0.1502, "step": 7342 }, { "epoch": 4.698016634676903, "grad_norm": 1.1398626108906396, "learning_rate": 1.1458752996501677e-05, "loss": 0.147, "step": 7343 }, { "epoch": 4.698656429942418, "grad_norm": 1.2602016038127877, "learning_rate": 1.1456702977076618e-05, "loss": 0.1703, "step": 7344 }, { "epoch": 4.6992962252079336, "grad_norm": 1.3024819861239811, "learning_rate": 1.1454652895103327e-05, "loss": 0.1689, "step": 7345 }, { "epoch": 4.699936020473449, "grad_norm": 1.3444070897790952, "learning_rate": 1.1452602750669827e-05, "loss": 0.1727, "step": 7346 }, { "epoch": 4.700575815738963, "grad_norm": 1.2524656411807351, "learning_rate": 1.145055254386415e-05, "loss": 0.1632, "step": 7347 }, { "epoch": 4.7012156110044785, "grad_norm": 1.3220524125039228, "learning_rate": 1.1448502274774325e-05, "loss": 0.17, "step": 7348 }, { "epoch": 4.701855406269994, "grad_norm": 1.3063510823178346, "learning_rate": 1.1446451943488393e-05, "loss": 0.1687, "step": 7349 }, { "epoch": 4.702495201535509, "grad_norm": 1.1083537328386688, "learning_rate": 1.1444401550094381e-05, "loss": 0.1575, "step": 7350 }, { "epoch": 4.703134996801023, "grad_norm": 1.3358253631398969, "learning_rate": 1.144235109468034e-05, "loss": 0.1706, "step": 7351 }, { "epoch": 4.703774792066539, "grad_norm": 1.2627973568416135, "learning_rate": 1.1440300577334309e-05, "loss": 0.1692, "step": 7352 }, { "epoch": 4.704414587332054, "grad_norm": 1.3352024001254263, "learning_rate": 1.143824999814433e-05, "loss": 0.1669, "step": 7353 }, { "epoch": 4.705054382597568, "grad_norm": 1.1654686100205716, "learning_rate": 1.143619935719846e-05, "loss": 0.1631, "step": 7354 }, { "epoch": 4.705694177863084, "grad_norm": 1.2263779438286027, "learning_rate": 1.1434148654584742e-05, "loss": 0.1491, "step": 7355 }, { "epoch": 4.706333973128599, "grad_norm": 1.337794274991258, "learning_rate": 1.1432097890391231e-05, "loss": 0.1707, "step": 7356 }, { "epoch": 4.706973768394114, "grad_norm": 1.2643392903116684, "learning_rate": 1.1430047064705984e-05, "loss": 0.1585, "step": 7357 }, { "epoch": 4.707613563659629, "grad_norm": 1.2949403321310984, "learning_rate": 1.142799617761706e-05, "loss": 0.167, "step": 7358 }, { "epoch": 4.708253358925144, "grad_norm": 1.0739588293976396, "learning_rate": 1.1425945229212523e-05, "loss": 0.1446, "step": 7359 }, { "epoch": 4.708893154190659, "grad_norm": 1.2338276437576803, "learning_rate": 1.1423894219580431e-05, "loss": 0.1625, "step": 7360 }, { "epoch": 4.709532949456174, "grad_norm": 1.3630218674030548, "learning_rate": 1.1421843148808856e-05, "loss": 0.1777, "step": 7361 }, { "epoch": 4.710172744721689, "grad_norm": 1.3602860861489807, "learning_rate": 1.1419792016985867e-05, "loss": 0.1797, "step": 7362 }, { "epoch": 4.710812539987204, "grad_norm": 1.2687861169740844, "learning_rate": 1.141774082419953e-05, "loss": 0.1573, "step": 7363 }, { "epoch": 4.711452335252719, "grad_norm": 1.267882017032737, "learning_rate": 1.1415689570537927e-05, "loss": 0.1539, "step": 7364 }, { "epoch": 4.712092130518235, "grad_norm": 1.260522201487805, "learning_rate": 1.141363825608913e-05, "loss": 0.1685, "step": 7365 }, { "epoch": 4.712731925783749, "grad_norm": 1.3095175589382808, "learning_rate": 1.141158688094122e-05, "loss": 0.1729, "step": 7366 }, { "epoch": 4.713371721049264, "grad_norm": 1.4789506036611082, "learning_rate": 1.1409535445182284e-05, "loss": 0.1812, "step": 7367 }, { "epoch": 4.7140115163147795, "grad_norm": 1.3580166720076492, "learning_rate": 1.1407483948900401e-05, "loss": 0.1692, "step": 7368 }, { "epoch": 4.714651311580294, "grad_norm": 1.2160293107665845, "learning_rate": 1.1405432392183665e-05, "loss": 0.1659, "step": 7369 }, { "epoch": 4.715291106845809, "grad_norm": 1.116767675265293, "learning_rate": 1.1403380775120159e-05, "loss": 0.1611, "step": 7370 }, { "epoch": 4.7159309021113245, "grad_norm": 1.181797689239338, "learning_rate": 1.1401329097797977e-05, "loss": 0.1505, "step": 7371 }, { "epoch": 4.71657069737684, "grad_norm": 1.3946707625079973, "learning_rate": 1.139927736030522e-05, "loss": 0.1774, "step": 7372 }, { "epoch": 4.717210492642354, "grad_norm": 1.2561698516724706, "learning_rate": 1.1397225562729979e-05, "loss": 0.1608, "step": 7373 }, { "epoch": 4.717850287907869, "grad_norm": 1.3022260088437838, "learning_rate": 1.139517370516036e-05, "loss": 0.1701, "step": 7374 }, { "epoch": 4.718490083173385, "grad_norm": 1.2993241479344222, "learning_rate": 1.1393121787684463e-05, "loss": 0.1702, "step": 7375 }, { "epoch": 4.719129878438899, "grad_norm": 1.1553036056515014, "learning_rate": 1.1391069810390395e-05, "loss": 0.1466, "step": 7376 }, { "epoch": 4.719769673704414, "grad_norm": 1.367778021625046, "learning_rate": 1.1389017773366267e-05, "loss": 0.1824, "step": 7377 }, { "epoch": 4.72040946896993, "grad_norm": 1.273774116540679, "learning_rate": 1.1386965676700182e-05, "loss": 0.1634, "step": 7378 }, { "epoch": 4.721049264235445, "grad_norm": 1.3752701112572472, "learning_rate": 1.1384913520480261e-05, "loss": 0.1735, "step": 7379 }, { "epoch": 4.72168905950096, "grad_norm": 1.1747664319463136, "learning_rate": 1.1382861304794615e-05, "loss": 0.1604, "step": 7380 }, { "epoch": 4.722328854766475, "grad_norm": 1.2587459575803173, "learning_rate": 1.1380809029731364e-05, "loss": 0.1619, "step": 7381 }, { "epoch": 4.72296865003199, "grad_norm": 1.3253333827408722, "learning_rate": 1.137875669537863e-05, "loss": 0.1739, "step": 7382 }, { "epoch": 4.723608445297505, "grad_norm": 1.363125020696766, "learning_rate": 1.1376704301824534e-05, "loss": 0.1564, "step": 7383 }, { "epoch": 4.7242482405630195, "grad_norm": 1.3452696358307117, "learning_rate": 1.137465184915721e-05, "loss": 0.1715, "step": 7384 }, { "epoch": 4.724888035828535, "grad_norm": 1.3220423225654452, "learning_rate": 1.1372599337464778e-05, "loss": 0.1683, "step": 7385 }, { "epoch": 4.72552783109405, "grad_norm": 1.1954856784590546, "learning_rate": 1.1370546766835371e-05, "loss": 0.1605, "step": 7386 }, { "epoch": 4.726167626359565, "grad_norm": 1.2807436942819106, "learning_rate": 1.1368494137357126e-05, "loss": 0.1622, "step": 7387 }, { "epoch": 4.72680742162508, "grad_norm": 1.2763131152055474, "learning_rate": 1.1366441449118176e-05, "loss": 0.1613, "step": 7388 }, { "epoch": 4.727447216890595, "grad_norm": 1.2465034255695953, "learning_rate": 1.136438870220666e-05, "loss": 0.1619, "step": 7389 }, { "epoch": 4.72808701215611, "grad_norm": 1.3114085834847362, "learning_rate": 1.1362335896710722e-05, "loss": 0.161, "step": 7390 }, { "epoch": 4.728726807421625, "grad_norm": 1.4130210080440762, "learning_rate": 1.1360283032718501e-05, "loss": 0.1632, "step": 7391 }, { "epoch": 4.72936660268714, "grad_norm": 1.2786510106031261, "learning_rate": 1.1358230110318152e-05, "loss": 0.1604, "step": 7392 }, { "epoch": 4.730006397952655, "grad_norm": 1.2123518852763926, "learning_rate": 1.1356177129597811e-05, "loss": 0.1548, "step": 7393 }, { "epoch": 4.7306461932181705, "grad_norm": 1.2722962019827901, "learning_rate": 1.1354124090645642e-05, "loss": 0.1792, "step": 7394 }, { "epoch": 4.731285988483685, "grad_norm": 1.3061069966174956, "learning_rate": 1.1352070993549792e-05, "loss": 0.1676, "step": 7395 }, { "epoch": 4.7319257837492, "grad_norm": 1.201813940369873, "learning_rate": 1.1350017838398415e-05, "loss": 0.15, "step": 7396 }, { "epoch": 4.732565579014715, "grad_norm": 1.2857333382434621, "learning_rate": 1.1347964625279678e-05, "loss": 0.1737, "step": 7397 }, { "epoch": 4.733205374280231, "grad_norm": 1.2715169439975016, "learning_rate": 1.1345911354281736e-05, "loss": 0.1645, "step": 7398 }, { "epoch": 4.733845169545745, "grad_norm": 1.1712118275066312, "learning_rate": 1.1343858025492756e-05, "loss": 0.1595, "step": 7399 }, { "epoch": 4.73448496481126, "grad_norm": 1.2413945447822754, "learning_rate": 1.1341804639000901e-05, "loss": 0.1579, "step": 7400 }, { "epoch": 4.735124760076776, "grad_norm": 1.2022222015846675, "learning_rate": 1.1339751194894341e-05, "loss": 0.157, "step": 7401 }, { "epoch": 4.735764555342291, "grad_norm": 1.3698615598660318, "learning_rate": 1.1337697693261251e-05, "loss": 0.1589, "step": 7402 }, { "epoch": 4.736404350607805, "grad_norm": 1.4185155304991877, "learning_rate": 1.1335644134189801e-05, "loss": 0.1737, "step": 7403 }, { "epoch": 4.737044145873321, "grad_norm": 1.2738355757516433, "learning_rate": 1.1333590517768168e-05, "loss": 0.1555, "step": 7404 }, { "epoch": 4.737683941138836, "grad_norm": 1.1265686589262194, "learning_rate": 1.1331536844084527e-05, "loss": 0.1501, "step": 7405 }, { "epoch": 4.73832373640435, "grad_norm": 1.0993423664518476, "learning_rate": 1.1329483113227067e-05, "loss": 0.1547, "step": 7406 }, { "epoch": 4.7389635316698655, "grad_norm": 1.2267111731454061, "learning_rate": 1.1327429325283966e-05, "loss": 0.1605, "step": 7407 }, { "epoch": 4.739603326935381, "grad_norm": 1.170417777973171, "learning_rate": 1.1325375480343407e-05, "loss": 0.1545, "step": 7408 }, { "epoch": 4.740243122200896, "grad_norm": 1.3499581746399065, "learning_rate": 1.1323321578493589e-05, "loss": 0.1723, "step": 7409 }, { "epoch": 4.74088291746641, "grad_norm": 1.3388270924966612, "learning_rate": 1.132126761982269e-05, "loss": 0.1823, "step": 7410 }, { "epoch": 4.741522712731926, "grad_norm": 1.2759802924046064, "learning_rate": 1.1319213604418913e-05, "loss": 0.1585, "step": 7411 }, { "epoch": 4.742162507997441, "grad_norm": 1.3776160015228662, "learning_rate": 1.1317159532370452e-05, "loss": 0.1766, "step": 7412 }, { "epoch": 4.742802303262955, "grad_norm": 1.3545127212061967, "learning_rate": 1.13151054037655e-05, "loss": 0.1745, "step": 7413 }, { "epoch": 4.743442098528471, "grad_norm": 1.250001919806068, "learning_rate": 1.1313051218692266e-05, "loss": 0.1644, "step": 7414 }, { "epoch": 4.744081893793986, "grad_norm": 1.2846508857280297, "learning_rate": 1.1310996977238946e-05, "loss": 0.1721, "step": 7415 }, { "epoch": 4.744721689059501, "grad_norm": 1.2447121216052086, "learning_rate": 1.1308942679493746e-05, "loss": 0.1619, "step": 7416 }, { "epoch": 4.7453614843250165, "grad_norm": 1.299141616845474, "learning_rate": 1.1306888325544878e-05, "loss": 0.164, "step": 7417 }, { "epoch": 4.746001279590531, "grad_norm": 1.30664200405118, "learning_rate": 1.1304833915480549e-05, "loss": 0.1684, "step": 7418 }, { "epoch": 4.746641074856046, "grad_norm": 1.4095313228603106, "learning_rate": 1.1302779449388974e-05, "loss": 0.1825, "step": 7419 }, { "epoch": 4.747280870121561, "grad_norm": 1.3081039145332276, "learning_rate": 1.1300724927358366e-05, "loss": 0.1698, "step": 7420 }, { "epoch": 4.747920665387076, "grad_norm": 1.265690144174974, "learning_rate": 1.1298670349476943e-05, "loss": 0.1685, "step": 7421 }, { "epoch": 4.748560460652591, "grad_norm": 1.3679053920233897, "learning_rate": 1.1296615715832928e-05, "loss": 0.1729, "step": 7422 }, { "epoch": 4.749200255918106, "grad_norm": 1.2549304002314583, "learning_rate": 1.1294561026514539e-05, "loss": 0.1588, "step": 7423 }, { "epoch": 4.749840051183622, "grad_norm": 1.2090029179871624, "learning_rate": 1.1292506281610004e-05, "loss": 0.1674, "step": 7424 }, { "epoch": 4.750479846449136, "grad_norm": 1.2495283667199437, "learning_rate": 1.1290451481207543e-05, "loss": 0.1649, "step": 7425 }, { "epoch": 4.751119641714651, "grad_norm": 1.361088735706102, "learning_rate": 1.1288396625395397e-05, "loss": 0.1739, "step": 7426 }, { "epoch": 4.751759436980167, "grad_norm": 1.203242489510536, "learning_rate": 1.1286341714261786e-05, "loss": 0.1454, "step": 7427 }, { "epoch": 4.752399232245681, "grad_norm": 1.277749928778883, "learning_rate": 1.1284286747894952e-05, "loss": 0.1689, "step": 7428 }, { "epoch": 4.753039027511196, "grad_norm": 1.20597726925564, "learning_rate": 1.1282231726383133e-05, "loss": 0.1488, "step": 7429 }, { "epoch": 4.7536788227767115, "grad_norm": 1.2542865667353627, "learning_rate": 1.1280176649814565e-05, "loss": 0.1583, "step": 7430 }, { "epoch": 4.754318618042227, "grad_norm": 1.2639922781621549, "learning_rate": 1.1278121518277487e-05, "loss": 0.1651, "step": 7431 }, { "epoch": 4.754958413307741, "grad_norm": 1.3258893000425784, "learning_rate": 1.1276066331860142e-05, "loss": 0.1852, "step": 7432 }, { "epoch": 4.755598208573256, "grad_norm": 1.3006459725776232, "learning_rate": 1.1274011090650778e-05, "loss": 0.1736, "step": 7433 }, { "epoch": 4.756238003838772, "grad_norm": 1.2050144990267029, "learning_rate": 1.127195579473765e-05, "loss": 0.1687, "step": 7434 }, { "epoch": 4.756877799104287, "grad_norm": 1.143816971946004, "learning_rate": 1.1269900444208996e-05, "loss": 0.155, "step": 7435 }, { "epoch": 4.757517594369801, "grad_norm": 1.1940113775220602, "learning_rate": 1.1267845039153082e-05, "loss": 0.1618, "step": 7436 }, { "epoch": 4.758157389635317, "grad_norm": 1.2109974459877755, "learning_rate": 1.1265789579658156e-05, "loss": 0.1622, "step": 7437 }, { "epoch": 4.758797184900832, "grad_norm": 1.1767121849056286, "learning_rate": 1.1263734065812474e-05, "loss": 0.151, "step": 7438 }, { "epoch": 4.759436980166347, "grad_norm": 1.1190632319554432, "learning_rate": 1.1261678497704302e-05, "loss": 0.157, "step": 7439 }, { "epoch": 4.760076775431862, "grad_norm": 1.1780348448312175, "learning_rate": 1.1259622875421898e-05, "loss": 0.1529, "step": 7440 }, { "epoch": 4.760716570697377, "grad_norm": 1.2454649814759655, "learning_rate": 1.1257567199053529e-05, "loss": 0.1638, "step": 7441 }, { "epoch": 4.761356365962892, "grad_norm": 1.3273940448172896, "learning_rate": 1.1255511468687461e-05, "loss": 0.1688, "step": 7442 }, { "epoch": 4.7619961612284065, "grad_norm": 1.4017890818287229, "learning_rate": 1.1253455684411964e-05, "loss": 0.1739, "step": 7443 }, { "epoch": 4.762635956493922, "grad_norm": 1.2633878816905095, "learning_rate": 1.1251399846315311e-05, "loss": 0.1648, "step": 7444 }, { "epoch": 4.763275751759437, "grad_norm": 1.475299451648963, "learning_rate": 1.1249343954485777e-05, "loss": 0.1894, "step": 7445 }, { "epoch": 4.763915547024952, "grad_norm": 1.319347577502736, "learning_rate": 1.1247288009011631e-05, "loss": 0.1722, "step": 7446 }, { "epoch": 4.764555342290467, "grad_norm": 1.3057893875204178, "learning_rate": 1.124523200998116e-05, "loss": 0.1623, "step": 7447 }, { "epoch": 4.765195137555982, "grad_norm": 1.3543961946873595, "learning_rate": 1.1243175957482638e-05, "loss": 0.1728, "step": 7448 }, { "epoch": 4.765834932821497, "grad_norm": 1.2260835040322324, "learning_rate": 1.1241119851604355e-05, "loss": 0.1498, "step": 7449 }, { "epoch": 4.766474728087012, "grad_norm": 1.3167064734282619, "learning_rate": 1.123906369243459e-05, "loss": 0.1615, "step": 7450 }, { "epoch": 4.767114523352527, "grad_norm": 1.315723080488913, "learning_rate": 1.1237007480061639e-05, "loss": 0.1718, "step": 7451 }, { "epoch": 4.767754318618042, "grad_norm": 1.3086330538065742, "learning_rate": 1.1234951214573785e-05, "loss": 0.1641, "step": 7452 }, { "epoch": 4.7683941138835575, "grad_norm": 1.3601058356325821, "learning_rate": 1.1232894896059322e-05, "loss": 0.1634, "step": 7453 }, { "epoch": 4.769033909149073, "grad_norm": 1.4309368415705415, "learning_rate": 1.1230838524606547e-05, "loss": 0.1841, "step": 7454 }, { "epoch": 4.769673704414587, "grad_norm": 1.283590309147873, "learning_rate": 1.1228782100303752e-05, "loss": 0.1577, "step": 7455 }, { "epoch": 4.770313499680102, "grad_norm": 1.2305731228779155, "learning_rate": 1.1226725623239244e-05, "loss": 0.1568, "step": 7456 }, { "epoch": 4.770953294945618, "grad_norm": 1.4439882556782684, "learning_rate": 1.1224669093501315e-05, "loss": 0.1889, "step": 7457 }, { "epoch": 4.771593090211132, "grad_norm": 1.3483393547901075, "learning_rate": 1.1222612511178274e-05, "loss": 0.1724, "step": 7458 }, { "epoch": 4.772232885476647, "grad_norm": 1.2733189824799682, "learning_rate": 1.122055587635843e-05, "loss": 0.1576, "step": 7459 }, { "epoch": 4.772872680742163, "grad_norm": 1.2063177101355245, "learning_rate": 1.1218499189130087e-05, "loss": 0.1596, "step": 7460 }, { "epoch": 4.773512476007678, "grad_norm": 1.3389874380220383, "learning_rate": 1.1216442449581558e-05, "loss": 0.1629, "step": 7461 }, { "epoch": 4.774152271273192, "grad_norm": 1.222455493757422, "learning_rate": 1.121438565780115e-05, "loss": 0.1659, "step": 7462 }, { "epoch": 4.774792066538708, "grad_norm": 1.3013710180550184, "learning_rate": 1.1212328813877185e-05, "loss": 0.1659, "step": 7463 }, { "epoch": 4.775431861804223, "grad_norm": 1.3393350498004617, "learning_rate": 1.1210271917897977e-05, "loss": 0.1731, "step": 7464 }, { "epoch": 4.776071657069737, "grad_norm": 1.2826427140812258, "learning_rate": 1.1208214969951843e-05, "loss": 0.1626, "step": 7465 }, { "epoch": 4.7767114523352525, "grad_norm": 1.2465640346775808, "learning_rate": 1.120615797012711e-05, "loss": 0.1597, "step": 7466 }, { "epoch": 4.777351247600768, "grad_norm": 1.174435475414997, "learning_rate": 1.12041009185121e-05, "loss": 0.1546, "step": 7467 }, { "epoch": 4.777991042866283, "grad_norm": 1.3347696915757004, "learning_rate": 1.1202043815195137e-05, "loss": 0.1638, "step": 7468 }, { "epoch": 4.7786308381317975, "grad_norm": 1.304360857899911, "learning_rate": 1.1199986660264552e-05, "loss": 0.176, "step": 7469 }, { "epoch": 4.779270633397313, "grad_norm": 1.2231613603782538, "learning_rate": 1.1197929453808672e-05, "loss": 0.1602, "step": 7470 }, { "epoch": 4.779910428662828, "grad_norm": 1.2551349525123443, "learning_rate": 1.1195872195915836e-05, "loss": 0.1596, "step": 7471 }, { "epoch": 4.780550223928343, "grad_norm": 1.2495346968614907, "learning_rate": 1.119381488667437e-05, "loss": 0.1578, "step": 7472 }, { "epoch": 4.781190019193858, "grad_norm": 1.0812049519956277, "learning_rate": 1.1191757526172619e-05, "loss": 0.1412, "step": 7473 }, { "epoch": 4.781829814459373, "grad_norm": 1.2489917337470897, "learning_rate": 1.1189700114498923e-05, "loss": 0.1601, "step": 7474 }, { "epoch": 4.782469609724888, "grad_norm": 1.2774243116866306, "learning_rate": 1.1187642651741616e-05, "loss": 0.1535, "step": 7475 }, { "epoch": 4.7831094049904035, "grad_norm": 1.2769470503257305, "learning_rate": 1.118558513798905e-05, "loss": 0.167, "step": 7476 }, { "epoch": 4.783749200255918, "grad_norm": 1.237764065628623, "learning_rate": 1.1183527573329564e-05, "loss": 0.1581, "step": 7477 }, { "epoch": 4.784388995521433, "grad_norm": 1.334547476295624, "learning_rate": 1.118146995785151e-05, "loss": 0.1627, "step": 7478 }, { "epoch": 4.785028790786948, "grad_norm": 1.2372917158590484, "learning_rate": 1.1179412291643237e-05, "loss": 0.15, "step": 7479 }, { "epoch": 4.785668586052463, "grad_norm": 1.3954682700503163, "learning_rate": 1.11773545747931e-05, "loss": 0.1725, "step": 7480 }, { "epoch": 4.786308381317978, "grad_norm": 1.3616056921117745, "learning_rate": 1.1175296807389451e-05, "loss": 0.1619, "step": 7481 }, { "epoch": 4.786948176583493, "grad_norm": 1.232869536075111, "learning_rate": 1.1173238989520649e-05, "loss": 0.1535, "step": 7482 }, { "epoch": 4.787587971849009, "grad_norm": 1.417521131040746, "learning_rate": 1.117118112127505e-05, "loss": 0.1832, "step": 7483 }, { "epoch": 4.788227767114523, "grad_norm": 1.1714017947981707, "learning_rate": 1.116912320274102e-05, "loss": 0.1501, "step": 7484 }, { "epoch": 4.788867562380038, "grad_norm": 1.4064531680769377, "learning_rate": 1.1167065234006915e-05, "loss": 0.1701, "step": 7485 }, { "epoch": 4.789507357645554, "grad_norm": 1.2255450094583464, "learning_rate": 1.1165007215161108e-05, "loss": 0.1669, "step": 7486 }, { "epoch": 4.790147152911068, "grad_norm": 1.4407670152742356, "learning_rate": 1.1162949146291962e-05, "loss": 0.1815, "step": 7487 }, { "epoch": 4.790786948176583, "grad_norm": 1.4211931439964938, "learning_rate": 1.1160891027487846e-05, "loss": 0.188, "step": 7488 }, { "epoch": 4.7914267434420985, "grad_norm": 1.507880684877614, "learning_rate": 1.1158832858837139e-05, "loss": 0.1894, "step": 7489 }, { "epoch": 4.792066538707614, "grad_norm": 1.2270344956150707, "learning_rate": 1.115677464042821e-05, "loss": 0.1557, "step": 7490 }, { "epoch": 4.792706333973129, "grad_norm": 1.4098754991623577, "learning_rate": 1.1154716372349435e-05, "loss": 0.1822, "step": 7491 }, { "epoch": 4.7933461292386434, "grad_norm": 1.1881464295912152, "learning_rate": 1.1152658054689191e-05, "loss": 0.1482, "step": 7492 }, { "epoch": 4.793985924504159, "grad_norm": 1.2324176836449379, "learning_rate": 1.115059968753586e-05, "loss": 0.1574, "step": 7493 }, { "epoch": 4.794625719769674, "grad_norm": 1.2850684784412365, "learning_rate": 1.114854127097783e-05, "loss": 0.1711, "step": 7494 }, { "epoch": 4.795265515035188, "grad_norm": 1.2496669291680593, "learning_rate": 1.1146482805103476e-05, "loss": 0.1623, "step": 7495 }, { "epoch": 4.795905310300704, "grad_norm": 1.1837939106563773, "learning_rate": 1.1144424290001195e-05, "loss": 0.1553, "step": 7496 }, { "epoch": 4.796545105566219, "grad_norm": 1.0935717565262937, "learning_rate": 1.1142365725759367e-05, "loss": 0.1492, "step": 7497 }, { "epoch": 4.797184900831734, "grad_norm": 1.4314334233341837, "learning_rate": 1.1140307112466387e-05, "loss": 0.1738, "step": 7498 }, { "epoch": 4.797824696097249, "grad_norm": 1.2610913314118128, "learning_rate": 1.1138248450210648e-05, "loss": 0.1626, "step": 7499 }, { "epoch": 4.798464491362764, "grad_norm": 1.2270746677670012, "learning_rate": 1.1136189739080545e-05, "loss": 0.161, "step": 7500 }, { "epoch": 4.799104286628279, "grad_norm": 1.2281910935369056, "learning_rate": 1.1134130979164479e-05, "loss": 0.1441, "step": 7501 }, { "epoch": 4.7997440818937935, "grad_norm": 1.2999868412450386, "learning_rate": 1.1132072170550843e-05, "loss": 0.1541, "step": 7502 }, { "epoch": 4.800383877159309, "grad_norm": 1.2375194741115616, "learning_rate": 1.1130013313328044e-05, "loss": 0.1603, "step": 7503 }, { "epoch": 4.801023672424824, "grad_norm": 1.23308819002351, "learning_rate": 1.112795440758448e-05, "loss": 0.1635, "step": 7504 }, { "epoch": 4.801663467690339, "grad_norm": 1.2323202104067412, "learning_rate": 1.1125895453408564e-05, "loss": 0.1565, "step": 7505 }, { "epoch": 4.802303262955854, "grad_norm": 1.273713508531223, "learning_rate": 1.11238364508887e-05, "loss": 0.152, "step": 7506 }, { "epoch": 4.802943058221369, "grad_norm": 1.2815969257270956, "learning_rate": 1.1121777400113293e-05, "loss": 0.1653, "step": 7507 }, { "epoch": 4.803582853486884, "grad_norm": 1.2981129939497984, "learning_rate": 1.1119718301170766e-05, "loss": 0.1685, "step": 7508 }, { "epoch": 4.8042226487524, "grad_norm": 1.325615247037929, "learning_rate": 1.1117659154149523e-05, "loss": 0.1618, "step": 7509 }, { "epoch": 4.804862444017914, "grad_norm": 1.1423627491023662, "learning_rate": 1.1115599959137985e-05, "loss": 0.1417, "step": 7510 }, { "epoch": 4.805502239283429, "grad_norm": 1.3724758722491774, "learning_rate": 1.1113540716224569e-05, "loss": 0.1683, "step": 7511 }, { "epoch": 4.8061420345489445, "grad_norm": 1.2261220774628914, "learning_rate": 1.1111481425497697e-05, "loss": 0.1575, "step": 7512 }, { "epoch": 4.80678182981446, "grad_norm": 1.4062147600820865, "learning_rate": 1.1109422087045789e-05, "loss": 0.1779, "step": 7513 }, { "epoch": 4.807421625079974, "grad_norm": 1.2349524356051897, "learning_rate": 1.110736270095727e-05, "loss": 0.1626, "step": 7514 }, { "epoch": 4.8080614203454894, "grad_norm": 1.4271828698460713, "learning_rate": 1.1105303267320564e-05, "loss": 0.1832, "step": 7515 }, { "epoch": 4.808701215611005, "grad_norm": 1.1669141668096616, "learning_rate": 1.1103243786224105e-05, "loss": 0.1517, "step": 7516 }, { "epoch": 4.809341010876519, "grad_norm": 1.2764207176841156, "learning_rate": 1.110118425775632e-05, "loss": 0.1633, "step": 7517 }, { "epoch": 4.809980806142034, "grad_norm": 1.2166862647632437, "learning_rate": 1.1099124682005643e-05, "loss": 0.154, "step": 7518 }, { "epoch": 4.81062060140755, "grad_norm": 1.21590086274166, "learning_rate": 1.1097065059060504e-05, "loss": 0.157, "step": 7519 }, { "epoch": 4.811260396673065, "grad_norm": 1.3224145118466881, "learning_rate": 1.1095005389009346e-05, "loss": 0.1782, "step": 7520 }, { "epoch": 4.811900191938579, "grad_norm": 1.3817975998206569, "learning_rate": 1.1092945671940603e-05, "loss": 0.1775, "step": 7521 }, { "epoch": 4.812539987204095, "grad_norm": 1.4309962530987757, "learning_rate": 1.1090885907942717e-05, "loss": 0.1661, "step": 7522 }, { "epoch": 4.81317978246961, "grad_norm": 1.172617605929558, "learning_rate": 1.108882609710413e-05, "loss": 0.1598, "step": 7523 }, { "epoch": 4.813819577735125, "grad_norm": 1.3585621049566132, "learning_rate": 1.1086766239513286e-05, "loss": 0.161, "step": 7524 }, { "epoch": 4.8144593730006395, "grad_norm": 1.3094141694348989, "learning_rate": 1.1084706335258634e-05, "loss": 0.1666, "step": 7525 }, { "epoch": 4.815099168266155, "grad_norm": 1.2116429779051996, "learning_rate": 1.1082646384428626e-05, "loss": 0.1564, "step": 7526 }, { "epoch": 4.81573896353167, "grad_norm": 1.2481052277693239, "learning_rate": 1.1080586387111704e-05, "loss": 0.1716, "step": 7527 }, { "epoch": 4.816378758797185, "grad_norm": 1.233163108512636, "learning_rate": 1.1078526343396327e-05, "loss": 0.1605, "step": 7528 }, { "epoch": 4.8170185540627, "grad_norm": 1.3094228232134109, "learning_rate": 1.1076466253370946e-05, "loss": 0.1736, "step": 7529 }, { "epoch": 4.817658349328215, "grad_norm": 1.290403316553438, "learning_rate": 1.1074406117124018e-05, "loss": 0.1706, "step": 7530 }, { "epoch": 4.81829814459373, "grad_norm": 1.3378142907796235, "learning_rate": 1.1072345934744002e-05, "loss": 0.1565, "step": 7531 }, { "epoch": 4.818937939859245, "grad_norm": 1.2847447868596207, "learning_rate": 1.1070285706319359e-05, "loss": 0.1669, "step": 7532 }, { "epoch": 4.81957773512476, "grad_norm": 1.3596972052497143, "learning_rate": 1.1068225431938555e-05, "loss": 0.1657, "step": 7533 }, { "epoch": 4.820217530390275, "grad_norm": 1.4762438454837779, "learning_rate": 1.1066165111690047e-05, "loss": 0.1893, "step": 7534 }, { "epoch": 4.8208573256557905, "grad_norm": 1.180176179689372, "learning_rate": 1.106410474566231e-05, "loss": 0.1603, "step": 7535 }, { "epoch": 4.821497120921305, "grad_norm": 1.2410153932124324, "learning_rate": 1.1062044333943807e-05, "loss": 0.1535, "step": 7536 }, { "epoch": 4.82213691618682, "grad_norm": 1.4158508273687278, "learning_rate": 1.1059983876623006e-05, "loss": 0.1787, "step": 7537 }, { "epoch": 4.822776711452335, "grad_norm": 1.4350869743913162, "learning_rate": 1.1057923373788387e-05, "loss": 0.1841, "step": 7538 }, { "epoch": 4.82341650671785, "grad_norm": 1.3721419004408173, "learning_rate": 1.1055862825528419e-05, "loss": 0.1848, "step": 7539 }, { "epoch": 4.824056301983365, "grad_norm": 1.2633427136239814, "learning_rate": 1.1053802231931576e-05, "loss": 0.1652, "step": 7540 }, { "epoch": 4.82469609724888, "grad_norm": 1.305668667443098, "learning_rate": 1.1051741593086344e-05, "loss": 0.1683, "step": 7541 }, { "epoch": 4.825335892514396, "grad_norm": 1.170999651348444, "learning_rate": 1.10496809090812e-05, "loss": 0.1576, "step": 7542 }, { "epoch": 4.825975687779911, "grad_norm": 1.2264596998730715, "learning_rate": 1.1047620180004624e-05, "loss": 0.1476, "step": 7543 }, { "epoch": 4.826615483045425, "grad_norm": 1.0881613756184991, "learning_rate": 1.1045559405945096e-05, "loss": 0.1548, "step": 7544 }, { "epoch": 4.827255278310941, "grad_norm": 1.2238856434681162, "learning_rate": 1.104349858699111e-05, "loss": 0.1525, "step": 7545 }, { "epoch": 4.827895073576456, "grad_norm": 1.28069031861836, "learning_rate": 1.1041437723231154e-05, "loss": 0.1625, "step": 7546 }, { "epoch": 4.82853486884197, "grad_norm": 1.22900454045293, "learning_rate": 1.103937681475371e-05, "loss": 0.1664, "step": 7547 }, { "epoch": 4.8291746641074855, "grad_norm": 1.255720273541348, "learning_rate": 1.103731586164728e-05, "loss": 0.165, "step": 7548 }, { "epoch": 4.829814459373001, "grad_norm": 1.3553629997975503, "learning_rate": 1.1035254864000347e-05, "loss": 0.1771, "step": 7549 }, { "epoch": 4.830454254638516, "grad_norm": 1.3503076091001094, "learning_rate": 1.1033193821901414e-05, "loss": 0.1609, "step": 7550 }, { "epoch": 4.8310940499040305, "grad_norm": 1.28560722629189, "learning_rate": 1.1031132735438978e-05, "loss": 0.1729, "step": 7551 }, { "epoch": 4.831733845169546, "grad_norm": 1.3379934793261614, "learning_rate": 1.1029071604701534e-05, "loss": 0.1855, "step": 7552 }, { "epoch": 4.832373640435061, "grad_norm": 1.2227009100216735, "learning_rate": 1.1027010429777587e-05, "loss": 0.1674, "step": 7553 }, { "epoch": 4.833013435700575, "grad_norm": 1.2637436281834697, "learning_rate": 1.1024949210755637e-05, "loss": 0.1608, "step": 7554 }, { "epoch": 4.833653230966091, "grad_norm": 1.200970954845747, "learning_rate": 1.1022887947724192e-05, "loss": 0.152, "step": 7555 }, { "epoch": 4.834293026231606, "grad_norm": 1.166228484795231, "learning_rate": 1.1020826640771758e-05, "loss": 0.1568, "step": 7556 }, { "epoch": 4.834932821497121, "grad_norm": 1.221679603358316, "learning_rate": 1.1018765289986845e-05, "loss": 0.1694, "step": 7557 }, { "epoch": 4.835572616762636, "grad_norm": 1.2382588837267694, "learning_rate": 1.1016703895457965e-05, "loss": 0.1562, "step": 7558 }, { "epoch": 4.836212412028151, "grad_norm": 1.4042231754581, "learning_rate": 1.1014642457273626e-05, "loss": 0.1653, "step": 7559 }, { "epoch": 4.836852207293666, "grad_norm": 1.3664084312748441, "learning_rate": 1.1012580975522342e-05, "loss": 0.1799, "step": 7560 }, { "epoch": 4.837492002559181, "grad_norm": 1.1925296564832473, "learning_rate": 1.1010519450292636e-05, "loss": 0.1527, "step": 7561 }, { "epoch": 4.838131797824696, "grad_norm": 1.1887214521982623, "learning_rate": 1.100845788167302e-05, "loss": 0.1539, "step": 7562 }, { "epoch": 4.838771593090211, "grad_norm": 1.3420471025919214, "learning_rate": 1.1006396269752017e-05, "loss": 0.1799, "step": 7563 }, { "epoch": 4.839411388355726, "grad_norm": 1.3567474852201518, "learning_rate": 1.1004334614618149e-05, "loss": 0.1696, "step": 7564 }, { "epoch": 4.840051183621242, "grad_norm": 1.3289770430673182, "learning_rate": 1.1002272916359939e-05, "loss": 0.1655, "step": 7565 }, { "epoch": 4.840690978886756, "grad_norm": 1.2627283190176848, "learning_rate": 1.1000211175065915e-05, "loss": 0.1725, "step": 7566 }, { "epoch": 4.841330774152271, "grad_norm": 1.2675678911878347, "learning_rate": 1.09981493908246e-05, "loss": 0.1631, "step": 7567 }, { "epoch": 4.841970569417787, "grad_norm": 1.3017248017913927, "learning_rate": 1.0996087563724527e-05, "loss": 0.1566, "step": 7568 }, { "epoch": 4.842610364683301, "grad_norm": 1.286478001169205, "learning_rate": 1.0994025693854222e-05, "loss": 0.1625, "step": 7569 }, { "epoch": 4.843250159948816, "grad_norm": 1.315736923372412, "learning_rate": 1.0991963781302225e-05, "loss": 0.1655, "step": 7570 }, { "epoch": 4.8438899552143315, "grad_norm": 1.1420763409628596, "learning_rate": 1.0989901826157067e-05, "loss": 0.1641, "step": 7571 }, { "epoch": 4.844529750479847, "grad_norm": 1.2692420406389922, "learning_rate": 1.098783982850729e-05, "loss": 0.1648, "step": 7572 }, { "epoch": 4.845169545745361, "grad_norm": 1.2149867069960987, "learning_rate": 1.0985777788441423e-05, "loss": 0.1577, "step": 7573 }, { "epoch": 4.8458093410108765, "grad_norm": 1.288039680413401, "learning_rate": 1.0983715706048012e-05, "loss": 0.1601, "step": 7574 }, { "epoch": 4.846449136276392, "grad_norm": 1.2385131910227847, "learning_rate": 1.0981653581415599e-05, "loss": 0.1615, "step": 7575 }, { "epoch": 4.847088931541906, "grad_norm": 1.2312063523424581, "learning_rate": 1.0979591414632729e-05, "loss": 0.1728, "step": 7576 }, { "epoch": 4.847728726807421, "grad_norm": 1.1378541598212688, "learning_rate": 1.0977529205787943e-05, "loss": 0.1482, "step": 7577 }, { "epoch": 4.848368522072937, "grad_norm": 1.2598094959431185, "learning_rate": 1.0975466954969794e-05, "loss": 0.1624, "step": 7578 }, { "epoch": 4.849008317338452, "grad_norm": 1.3271769995688147, "learning_rate": 1.0973404662266826e-05, "loss": 0.163, "step": 7579 }, { "epoch": 4.849648112603967, "grad_norm": 1.2826748968635868, "learning_rate": 1.0971342327767598e-05, "loss": 0.1678, "step": 7580 }, { "epoch": 4.850287907869482, "grad_norm": 1.127679754468526, "learning_rate": 1.0969279951560658e-05, "loss": 0.1569, "step": 7581 }, { "epoch": 4.850927703134997, "grad_norm": 1.2884145368405566, "learning_rate": 1.0967217533734558e-05, "loss": 0.168, "step": 7582 }, { "epoch": 4.851567498400512, "grad_norm": 1.2733255805925898, "learning_rate": 1.096515507437786e-05, "loss": 0.1637, "step": 7583 }, { "epoch": 4.8522072936660265, "grad_norm": 1.163455744979934, "learning_rate": 1.0963092573579117e-05, "loss": 0.1581, "step": 7584 }, { "epoch": 4.852847088931542, "grad_norm": 1.3519348959638284, "learning_rate": 1.0961030031426895e-05, "loss": 0.1687, "step": 7585 }, { "epoch": 4.853486884197057, "grad_norm": 1.278232330619599, "learning_rate": 1.0958967448009752e-05, "loss": 0.161, "step": 7586 }, { "epoch": 4.854126679462572, "grad_norm": 1.1990648952402103, "learning_rate": 1.0956904823416255e-05, "loss": 0.1446, "step": 7587 }, { "epoch": 4.854766474728087, "grad_norm": 1.3506547469125367, "learning_rate": 1.0954842157734964e-05, "loss": 0.1734, "step": 7588 }, { "epoch": 4.855406269993602, "grad_norm": 1.2847294982878135, "learning_rate": 1.0952779451054449e-05, "loss": 0.1614, "step": 7589 }, { "epoch": 4.856046065259117, "grad_norm": 1.366513807324132, "learning_rate": 1.0950716703463282e-05, "loss": 0.1752, "step": 7590 }, { "epoch": 4.856685860524632, "grad_norm": 1.3305576365586702, "learning_rate": 1.0948653915050029e-05, "loss": 0.1801, "step": 7591 }, { "epoch": 4.857325655790147, "grad_norm": 1.1544138362398484, "learning_rate": 1.0946591085903262e-05, "loss": 0.1575, "step": 7592 }, { "epoch": 4.857965451055662, "grad_norm": 1.2487143370384557, "learning_rate": 1.0944528216111563e-05, "loss": 0.1708, "step": 7593 }, { "epoch": 4.8586052463211775, "grad_norm": 1.2089942520335455, "learning_rate": 1.0942465305763498e-05, "loss": 0.162, "step": 7594 }, { "epoch": 4.859245041586692, "grad_norm": 1.2383168421328343, "learning_rate": 1.0940402354947652e-05, "loss": 0.1603, "step": 7595 }, { "epoch": 4.859884836852207, "grad_norm": 1.2260743414544277, "learning_rate": 1.0938339363752604e-05, "loss": 0.1652, "step": 7596 }, { "epoch": 4.8605246321177225, "grad_norm": 1.1849868149412297, "learning_rate": 1.0936276332266927e-05, "loss": 0.1543, "step": 7597 }, { "epoch": 4.861164427383238, "grad_norm": 1.157321066470432, "learning_rate": 1.0934213260579214e-05, "loss": 0.1666, "step": 7598 }, { "epoch": 4.861804222648752, "grad_norm": 1.2477174861900748, "learning_rate": 1.0932150148778042e-05, "loss": 0.1615, "step": 7599 }, { "epoch": 4.862444017914267, "grad_norm": 1.247448048930915, "learning_rate": 1.0930086996952004e-05, "loss": 0.1619, "step": 7600 }, { "epoch": 4.863083813179783, "grad_norm": 1.2359404159997138, "learning_rate": 1.0928023805189683e-05, "loss": 0.1625, "step": 7601 }, { "epoch": 4.863723608445298, "grad_norm": 1.263940334822591, "learning_rate": 1.0925960573579672e-05, "loss": 0.1546, "step": 7602 }, { "epoch": 4.864363403710812, "grad_norm": 1.1667580671758695, "learning_rate": 1.0923897302210562e-05, "loss": 0.1491, "step": 7603 }, { "epoch": 4.865003198976328, "grad_norm": 1.246740557675499, "learning_rate": 1.0921833991170943e-05, "loss": 0.1557, "step": 7604 }, { "epoch": 4.865642994241843, "grad_norm": 1.3532550172260789, "learning_rate": 1.0919770640549415e-05, "loss": 0.1673, "step": 7605 }, { "epoch": 4.866282789507357, "grad_norm": 1.1994717172378897, "learning_rate": 1.0917707250434569e-05, "loss": 0.1677, "step": 7606 }, { "epoch": 4.8669225847728725, "grad_norm": 1.3163499893434019, "learning_rate": 1.0915643820915005e-05, "loss": 0.1658, "step": 7607 }, { "epoch": 4.867562380038388, "grad_norm": 1.2307133374982522, "learning_rate": 1.0913580352079328e-05, "loss": 0.1587, "step": 7608 }, { "epoch": 4.868202175303903, "grad_norm": 1.1862739113863274, "learning_rate": 1.0911516844016132e-05, "loss": 0.1659, "step": 7609 }, { "epoch": 4.8688419705694175, "grad_norm": 1.2972669292332393, "learning_rate": 1.0909453296814027e-05, "loss": 0.1702, "step": 7610 }, { "epoch": 4.869481765834933, "grad_norm": 1.2504173051327727, "learning_rate": 1.0907389710561616e-05, "loss": 0.1607, "step": 7611 }, { "epoch": 4.870121561100448, "grad_norm": 1.309026005780416, "learning_rate": 1.09053260853475e-05, "loss": 0.1616, "step": 7612 }, { "epoch": 4.870761356365962, "grad_norm": 1.2792882885380557, "learning_rate": 1.0903262421260299e-05, "loss": 0.1643, "step": 7613 }, { "epoch": 4.871401151631478, "grad_norm": 1.2270438129424155, "learning_rate": 1.090119871838861e-05, "loss": 0.1671, "step": 7614 }, { "epoch": 4.872040946896993, "grad_norm": 1.2151197564229141, "learning_rate": 1.0899134976821058e-05, "loss": 0.1504, "step": 7615 }, { "epoch": 4.872680742162508, "grad_norm": 1.233244337380909, "learning_rate": 1.0897071196646242e-05, "loss": 0.1534, "step": 7616 }, { "epoch": 4.8733205374280235, "grad_norm": 1.3682511533839818, "learning_rate": 1.0895007377952792e-05, "loss": 0.1543, "step": 7617 }, { "epoch": 4.873960332693538, "grad_norm": 1.2046966206253893, "learning_rate": 1.0892943520829314e-05, "loss": 0.162, "step": 7618 }, { "epoch": 4.874600127959053, "grad_norm": 1.2137466616874024, "learning_rate": 1.0890879625364429e-05, "loss": 0.1502, "step": 7619 }, { "epoch": 4.8752399232245685, "grad_norm": 1.1984143229671849, "learning_rate": 1.0888815691646756e-05, "loss": 0.1552, "step": 7620 }, { "epoch": 4.875879718490083, "grad_norm": 1.2582370877680518, "learning_rate": 1.088675171976492e-05, "loss": 0.1635, "step": 7621 }, { "epoch": 4.876519513755598, "grad_norm": 1.2658465913907206, "learning_rate": 1.0884687709807542e-05, "loss": 0.1631, "step": 7622 }, { "epoch": 4.877159309021113, "grad_norm": 1.05332341339562, "learning_rate": 1.088262366186325e-05, "loss": 0.1533, "step": 7623 }, { "epoch": 4.877799104286629, "grad_norm": 1.16406712479408, "learning_rate": 1.0880559576020665e-05, "loss": 0.1507, "step": 7624 }, { "epoch": 4.878438899552143, "grad_norm": 1.1735365437418457, "learning_rate": 1.087849545236842e-05, "loss": 0.1446, "step": 7625 }, { "epoch": 4.879078694817658, "grad_norm": 1.2093056089473648, "learning_rate": 1.0876431290995145e-05, "loss": 0.1637, "step": 7626 }, { "epoch": 4.879718490083174, "grad_norm": 1.2059595828387564, "learning_rate": 1.0874367091989464e-05, "loss": 0.1598, "step": 7627 }, { "epoch": 4.880358285348688, "grad_norm": 1.2307262259387641, "learning_rate": 1.087230285544002e-05, "loss": 0.1603, "step": 7628 }, { "epoch": 4.880998080614203, "grad_norm": 1.3946599311578456, "learning_rate": 1.087023858143544e-05, "loss": 0.1772, "step": 7629 }, { "epoch": 4.8816378758797185, "grad_norm": 1.2385520446476452, "learning_rate": 1.0868174270064364e-05, "loss": 0.1544, "step": 7630 }, { "epoch": 4.882277671145234, "grad_norm": 1.192268754999436, "learning_rate": 1.086610992141543e-05, "loss": 0.1561, "step": 7631 }, { "epoch": 4.882917466410748, "grad_norm": 1.2170317891853686, "learning_rate": 1.0864045535577277e-05, "loss": 0.1623, "step": 7632 }, { "epoch": 4.8835572616762635, "grad_norm": 1.1889279490143847, "learning_rate": 1.0861981112638546e-05, "loss": 0.153, "step": 7633 }, { "epoch": 4.884197056941779, "grad_norm": 1.3214562108252432, "learning_rate": 1.0859916652687878e-05, "loss": 0.17, "step": 7634 }, { "epoch": 4.884836852207294, "grad_norm": 1.2168146696266482, "learning_rate": 1.0857852155813922e-05, "loss": 0.1632, "step": 7635 }, { "epoch": 4.885476647472808, "grad_norm": 1.2988416857640377, "learning_rate": 1.0855787622105316e-05, "loss": 0.1659, "step": 7636 }, { "epoch": 4.886116442738324, "grad_norm": 1.2385260470920842, "learning_rate": 1.0853723051650713e-05, "loss": 0.1562, "step": 7637 }, { "epoch": 4.886756238003839, "grad_norm": 1.2859601176440565, "learning_rate": 1.0851658444538762e-05, "loss": 0.1723, "step": 7638 }, { "epoch": 4.887396033269354, "grad_norm": 1.2520073727618097, "learning_rate": 1.084959380085811e-05, "loss": 0.1517, "step": 7639 }, { "epoch": 4.888035828534869, "grad_norm": 1.262862614668403, "learning_rate": 1.0847529120697416e-05, "loss": 0.1708, "step": 7640 }, { "epoch": 4.888675623800384, "grad_norm": 1.206136491704165, "learning_rate": 1.0845464404145324e-05, "loss": 0.1562, "step": 7641 }, { "epoch": 4.889315419065899, "grad_norm": 1.3334255832503086, "learning_rate": 1.0843399651290498e-05, "loss": 0.1783, "step": 7642 }, { "epoch": 4.889955214331414, "grad_norm": 1.1185487652010968, "learning_rate": 1.084133486222159e-05, "loss": 0.1505, "step": 7643 }, { "epoch": 4.890595009596929, "grad_norm": 1.2783078483270327, "learning_rate": 1.0839270037027259e-05, "loss": 0.1654, "step": 7644 }, { "epoch": 4.891234804862444, "grad_norm": 1.3179327057000778, "learning_rate": 1.0837205175796168e-05, "loss": 0.1719, "step": 7645 }, { "epoch": 4.891874600127959, "grad_norm": 1.2751672574018256, "learning_rate": 1.0835140278616973e-05, "loss": 0.1662, "step": 7646 }, { "epoch": 4.892514395393474, "grad_norm": 1.1610907264010093, "learning_rate": 1.0833075345578343e-05, "loss": 0.1485, "step": 7647 }, { "epoch": 4.893154190658989, "grad_norm": 1.0924537575313773, "learning_rate": 1.0831010376768938e-05, "loss": 0.1462, "step": 7648 }, { "epoch": 4.893793985924504, "grad_norm": 1.2286362882527955, "learning_rate": 1.0828945372277427e-05, "loss": 0.1705, "step": 7649 }, { "epoch": 4.894433781190019, "grad_norm": 1.1503051303744882, "learning_rate": 1.0826880332192475e-05, "loss": 0.1565, "step": 7650 }, { "epoch": 4.895073576455534, "grad_norm": 1.2427803669520274, "learning_rate": 1.0824815256602749e-05, "loss": 0.1714, "step": 7651 }, { "epoch": 4.895713371721049, "grad_norm": 1.285897367030437, "learning_rate": 1.0822750145596928e-05, "loss": 0.1719, "step": 7652 }, { "epoch": 4.8963531669865645, "grad_norm": 1.1921725277735336, "learning_rate": 1.0820684999263678e-05, "loss": 0.155, "step": 7653 }, { "epoch": 4.89699296225208, "grad_norm": 1.307576996103542, "learning_rate": 1.0818619817691674e-05, "loss": 0.1701, "step": 7654 }, { "epoch": 4.897632757517594, "grad_norm": 1.2395593048453566, "learning_rate": 1.0816554600969591e-05, "loss": 0.158, "step": 7655 }, { "epoch": 4.8982725527831095, "grad_norm": 1.3061276514018723, "learning_rate": 1.0814489349186106e-05, "loss": 0.1555, "step": 7656 }, { "epoch": 4.898912348048625, "grad_norm": 1.3280274259554825, "learning_rate": 1.08124240624299e-05, "loss": 0.1704, "step": 7657 }, { "epoch": 4.899552143314139, "grad_norm": 1.2489935103445635, "learning_rate": 1.0810358740789646e-05, "loss": 0.1678, "step": 7658 }, { "epoch": 4.900191938579654, "grad_norm": 1.3148289508387558, "learning_rate": 1.080829338435403e-05, "loss": 0.1679, "step": 7659 }, { "epoch": 4.90083173384517, "grad_norm": 1.3541035876685181, "learning_rate": 1.0806227993211734e-05, "loss": 0.1758, "step": 7660 }, { "epoch": 4.901471529110685, "grad_norm": 1.1986724481176079, "learning_rate": 1.0804162567451442e-05, "loss": 0.1644, "step": 7661 }, { "epoch": 4.902111324376199, "grad_norm": 1.097099702895601, "learning_rate": 1.0802097107161843e-05, "loss": 0.1476, "step": 7662 }, { "epoch": 4.902751119641715, "grad_norm": 1.1613662666007536, "learning_rate": 1.0800031612431616e-05, "loss": 0.1518, "step": 7663 }, { "epoch": 4.90339091490723, "grad_norm": 1.1804783986812757, "learning_rate": 1.0797966083349456e-05, "loss": 0.153, "step": 7664 }, { "epoch": 4.904030710172744, "grad_norm": 1.2618875071317794, "learning_rate": 1.0795900520004054e-05, "loss": 0.1535, "step": 7665 }, { "epoch": 4.9046705054382596, "grad_norm": 1.2364833053545965, "learning_rate": 1.0793834922484099e-05, "loss": 0.16, "step": 7666 }, { "epoch": 4.905310300703775, "grad_norm": 1.3816299953546802, "learning_rate": 1.0791769290878284e-05, "loss": 0.1563, "step": 7667 }, { "epoch": 4.90595009596929, "grad_norm": 1.097844147908669, "learning_rate": 1.0789703625275303e-05, "loss": 0.1421, "step": 7668 }, { "epoch": 4.9065898912348045, "grad_norm": 1.193131302397564, "learning_rate": 1.0787637925763854e-05, "loss": 0.1593, "step": 7669 }, { "epoch": 4.90722968650032, "grad_norm": 1.259432886045788, "learning_rate": 1.0785572192432634e-05, "loss": 0.1583, "step": 7670 }, { "epoch": 4.907869481765835, "grad_norm": 1.1905763410363441, "learning_rate": 1.0783506425370343e-05, "loss": 0.1522, "step": 7671 }, { "epoch": 4.90850927703135, "grad_norm": 1.147014551167989, "learning_rate": 1.0781440624665678e-05, "loss": 0.1473, "step": 7672 }, { "epoch": 4.909149072296865, "grad_norm": 1.313412021393919, "learning_rate": 1.0779374790407342e-05, "loss": 0.1675, "step": 7673 }, { "epoch": 4.90978886756238, "grad_norm": 1.2286847166799981, "learning_rate": 1.0777308922684038e-05, "loss": 0.1558, "step": 7674 }, { "epoch": 4.910428662827895, "grad_norm": 1.2272982859247294, "learning_rate": 1.0775243021584474e-05, "loss": 0.1538, "step": 7675 }, { "epoch": 4.9110684580934105, "grad_norm": 1.2487488529139443, "learning_rate": 1.0773177087197354e-05, "loss": 0.1611, "step": 7676 }, { "epoch": 4.911708253358925, "grad_norm": 1.4490121704484897, "learning_rate": 1.0771111119611384e-05, "loss": 0.1725, "step": 7677 }, { "epoch": 4.91234804862444, "grad_norm": 1.3413639559128745, "learning_rate": 1.0769045118915275e-05, "loss": 0.1707, "step": 7678 }, { "epoch": 4.9129878438899555, "grad_norm": 2.705610455812009, "learning_rate": 1.0766979085197735e-05, "loss": 0.1679, "step": 7679 }, { "epoch": 4.91362763915547, "grad_norm": 1.2812077595517413, "learning_rate": 1.0764913018547479e-05, "loss": 0.1547, "step": 7680 }, { "epoch": 4.914267434420985, "grad_norm": 1.1836063791521287, "learning_rate": 1.0762846919053218e-05, "loss": 0.1659, "step": 7681 }, { "epoch": 4.9149072296865, "grad_norm": 1.2129603317280941, "learning_rate": 1.0760780786803669e-05, "loss": 0.1522, "step": 7682 }, { "epoch": 4.915547024952016, "grad_norm": 1.3080985976186286, "learning_rate": 1.0758714621887542e-05, "loss": 0.169, "step": 7683 }, { "epoch": 4.91618682021753, "grad_norm": 1.2602304401445588, "learning_rate": 1.075664842439356e-05, "loss": 0.163, "step": 7684 }, { "epoch": 4.916826615483045, "grad_norm": 1.2145959695131623, "learning_rate": 1.0754582194410444e-05, "loss": 0.1662, "step": 7685 }, { "epoch": 4.917466410748561, "grad_norm": 1.2608205871234084, "learning_rate": 1.075251593202691e-05, "loss": 0.1573, "step": 7686 }, { "epoch": 4.918106206014075, "grad_norm": 1.3370398959522602, "learning_rate": 1.075044963733168e-05, "loss": 0.1695, "step": 7687 }, { "epoch": 4.91874600127959, "grad_norm": 1.1772064969030915, "learning_rate": 1.0748383310413474e-05, "loss": 0.1504, "step": 7688 }, { "epoch": 4.9193857965451055, "grad_norm": 1.3088064461141709, "learning_rate": 1.0746316951361024e-05, "loss": 0.1593, "step": 7689 }, { "epoch": 4.920025591810621, "grad_norm": 1.2243477665004163, "learning_rate": 1.0744250560263053e-05, "loss": 0.1561, "step": 7690 }, { "epoch": 4.920665387076136, "grad_norm": 1.2750233274390144, "learning_rate": 1.0742184137208282e-05, "loss": 0.1637, "step": 7691 }, { "epoch": 4.9213051823416505, "grad_norm": 1.3255873312517714, "learning_rate": 1.0740117682285446e-05, "loss": 0.1762, "step": 7692 }, { "epoch": 4.921944977607166, "grad_norm": 1.2533069645932298, "learning_rate": 1.0738051195583276e-05, "loss": 0.1656, "step": 7693 }, { "epoch": 4.922584772872681, "grad_norm": 1.16544162327522, "learning_rate": 1.0735984677190498e-05, "loss": 0.1484, "step": 7694 }, { "epoch": 4.923224568138195, "grad_norm": 1.1557899661278697, "learning_rate": 1.0733918127195851e-05, "loss": 0.1561, "step": 7695 }, { "epoch": 4.923864363403711, "grad_norm": 1.1650790535864948, "learning_rate": 1.073185154568806e-05, "loss": 0.1432, "step": 7696 }, { "epoch": 4.924504158669226, "grad_norm": 1.2925666670638372, "learning_rate": 1.072978493275587e-05, "loss": 0.1619, "step": 7697 }, { "epoch": 4.925143953934741, "grad_norm": 1.1712597963496612, "learning_rate": 1.0727718288488007e-05, "loss": 0.1535, "step": 7698 }, { "epoch": 4.925783749200256, "grad_norm": 1.3962640201894043, "learning_rate": 1.0725651612973218e-05, "loss": 0.1866, "step": 7699 }, { "epoch": 4.926423544465771, "grad_norm": 1.4007655559755534, "learning_rate": 1.0723584906300242e-05, "loss": 0.1631, "step": 7700 }, { "epoch": 4.927063339731286, "grad_norm": 1.2952954601469158, "learning_rate": 1.0721518168557817e-05, "loss": 0.165, "step": 7701 }, { "epoch": 4.927703134996801, "grad_norm": 1.224273660945667, "learning_rate": 1.0719451399834683e-05, "loss": 0.1679, "step": 7702 }, { "epoch": 4.928342930262316, "grad_norm": 1.2699829239658689, "learning_rate": 1.0717384600219587e-05, "loss": 0.1694, "step": 7703 }, { "epoch": 4.928982725527831, "grad_norm": 1.2689156974440814, "learning_rate": 1.0715317769801273e-05, "loss": 0.1622, "step": 7704 }, { "epoch": 4.929622520793346, "grad_norm": 1.3302346225907324, "learning_rate": 1.0713250908668485e-05, "loss": 0.1787, "step": 7705 }, { "epoch": 4.930262316058861, "grad_norm": 1.3204469516109112, "learning_rate": 1.0711184016909972e-05, "loss": 0.1803, "step": 7706 }, { "epoch": 4.930902111324376, "grad_norm": 1.343394559205648, "learning_rate": 1.0709117094614483e-05, "loss": 0.1736, "step": 7707 }, { "epoch": 4.931541906589891, "grad_norm": 1.2066545700420153, "learning_rate": 1.070705014187077e-05, "loss": 0.162, "step": 7708 }, { "epoch": 4.932181701855407, "grad_norm": 1.2221321219437722, "learning_rate": 1.0704983158767576e-05, "loss": 0.1612, "step": 7709 }, { "epoch": 4.932821497120921, "grad_norm": 1.2070640444653826, "learning_rate": 1.0702916145393662e-05, "loss": 0.1506, "step": 7710 }, { "epoch": 4.933461292386436, "grad_norm": 1.204992863939726, "learning_rate": 1.070084910183778e-05, "loss": 0.1662, "step": 7711 }, { "epoch": 4.9341010876519515, "grad_norm": 1.17271774075882, "learning_rate": 1.0698782028188683e-05, "loss": 0.1638, "step": 7712 }, { "epoch": 4.934740882917467, "grad_norm": 1.2510418467479116, "learning_rate": 1.0696714924535129e-05, "loss": 0.1544, "step": 7713 }, { "epoch": 4.935380678182981, "grad_norm": 1.1157446740177583, "learning_rate": 1.0694647790965876e-05, "loss": 0.1509, "step": 7714 }, { "epoch": 4.9360204734484965, "grad_norm": 1.3679437436047002, "learning_rate": 1.0692580627569682e-05, "loss": 0.1801, "step": 7715 }, { "epoch": 4.936660268714012, "grad_norm": 1.203954927406623, "learning_rate": 1.0690513434435309e-05, "loss": 0.1614, "step": 7716 }, { "epoch": 4.937300063979526, "grad_norm": 1.1713478975189846, "learning_rate": 1.0688446211651517e-05, "loss": 0.1622, "step": 7717 }, { "epoch": 4.937939859245041, "grad_norm": 1.1336859397198031, "learning_rate": 1.068637895930707e-05, "loss": 0.1558, "step": 7718 }, { "epoch": 4.938579654510557, "grad_norm": 1.1961966961142851, "learning_rate": 1.0684311677490726e-05, "loss": 0.1609, "step": 7719 }, { "epoch": 4.939219449776072, "grad_norm": 1.2371770842642156, "learning_rate": 1.0682244366291263e-05, "loss": 0.1608, "step": 7720 }, { "epoch": 4.939859245041586, "grad_norm": 1.2671375689338151, "learning_rate": 1.0680177025797436e-05, "loss": 0.1692, "step": 7721 }, { "epoch": 4.940499040307102, "grad_norm": 1.1258679565477987, "learning_rate": 1.0678109656098021e-05, "loss": 0.139, "step": 7722 }, { "epoch": 4.941138835572617, "grad_norm": 1.2143732080536564, "learning_rate": 1.0676042257281783e-05, "loss": 0.1554, "step": 7723 }, { "epoch": 4.941778630838132, "grad_norm": 1.1413495260595774, "learning_rate": 1.067397482943749e-05, "loss": 0.153, "step": 7724 }, { "epoch": 4.942418426103647, "grad_norm": 3.0219852207487543, "learning_rate": 1.067190737265392e-05, "loss": 0.1565, "step": 7725 }, { "epoch": 4.943058221369162, "grad_norm": 1.2530952803718376, "learning_rate": 1.0669839887019839e-05, "loss": 0.166, "step": 7726 }, { "epoch": 4.943698016634677, "grad_norm": 1.1893074280491802, "learning_rate": 1.0667772372624029e-05, "loss": 0.1504, "step": 7727 }, { "epoch": 4.944337811900192, "grad_norm": 1.3106402289525556, "learning_rate": 1.0665704829555257e-05, "loss": 0.1717, "step": 7728 }, { "epoch": 4.944977607165707, "grad_norm": 1.3313411855536792, "learning_rate": 1.0663637257902307e-05, "loss": 0.1659, "step": 7729 }, { "epoch": 4.945617402431222, "grad_norm": 1.3435156533638741, "learning_rate": 1.0661569657753952e-05, "loss": 0.1694, "step": 7730 }, { "epoch": 4.946257197696737, "grad_norm": 1.2257263590957315, "learning_rate": 1.0659502029198973e-05, "loss": 0.1573, "step": 7731 }, { "epoch": 4.946896992962252, "grad_norm": 1.2616664594086524, "learning_rate": 1.0657434372326155e-05, "loss": 0.1663, "step": 7732 }, { "epoch": 4.947536788227767, "grad_norm": 1.2228263684907548, "learning_rate": 1.0655366687224266e-05, "loss": 0.167, "step": 7733 }, { "epoch": 4.948176583493282, "grad_norm": 1.315996384874027, "learning_rate": 1.06532989739821e-05, "loss": 0.1555, "step": 7734 }, { "epoch": 4.9488163787587975, "grad_norm": 1.2706259520944096, "learning_rate": 1.0651231232688442e-05, "loss": 0.1597, "step": 7735 }, { "epoch": 4.949456174024312, "grad_norm": 1.3842726173162094, "learning_rate": 1.0649163463432068e-05, "loss": 0.1708, "step": 7736 }, { "epoch": 4.950095969289827, "grad_norm": 1.2228489048960536, "learning_rate": 1.0647095666301774e-05, "loss": 0.1504, "step": 7737 }, { "epoch": 4.9507357645553425, "grad_norm": 1.3683206018164855, "learning_rate": 1.0645027841386344e-05, "loss": 0.1722, "step": 7738 }, { "epoch": 4.951375559820857, "grad_norm": 1.2455417432340712, "learning_rate": 1.0642959988774564e-05, "loss": 0.1671, "step": 7739 }, { "epoch": 4.952015355086372, "grad_norm": 1.2830070017102757, "learning_rate": 1.0640892108555226e-05, "loss": 0.1716, "step": 7740 }, { "epoch": 4.952655150351887, "grad_norm": 1.1684003288868297, "learning_rate": 1.0638824200817117e-05, "loss": 0.1491, "step": 7741 }, { "epoch": 4.953294945617403, "grad_norm": 1.5434695654598702, "learning_rate": 1.0636756265649037e-05, "loss": 0.16, "step": 7742 }, { "epoch": 4.953934740882918, "grad_norm": 1.1747671230888121, "learning_rate": 1.0634688303139775e-05, "loss": 0.1508, "step": 7743 }, { "epoch": 4.954574536148432, "grad_norm": 1.2497858373817157, "learning_rate": 1.0632620313378128e-05, "loss": 0.1567, "step": 7744 }, { "epoch": 4.955214331413948, "grad_norm": 1.2849737134783166, "learning_rate": 1.0630552296452887e-05, "loss": 0.1578, "step": 7745 }, { "epoch": 4.955854126679463, "grad_norm": 1.253982644518333, "learning_rate": 1.0628484252452855e-05, "loss": 0.1485, "step": 7746 }, { "epoch": 4.956493921944977, "grad_norm": 1.182619458175665, "learning_rate": 1.0626416181466825e-05, "loss": 0.1627, "step": 7747 }, { "epoch": 4.957133717210493, "grad_norm": 1.2319291637621141, "learning_rate": 1.06243480835836e-05, "loss": 0.1518, "step": 7748 }, { "epoch": 4.957773512476008, "grad_norm": 1.2846855625786702, "learning_rate": 1.0622279958891978e-05, "loss": 0.1576, "step": 7749 }, { "epoch": 4.958413307741523, "grad_norm": 1.3124895736912119, "learning_rate": 1.062021180748076e-05, "loss": 0.163, "step": 7750 }, { "epoch": 4.9590531030070375, "grad_norm": 1.3023971956533735, "learning_rate": 1.061814362943875e-05, "loss": 0.1615, "step": 7751 }, { "epoch": 4.959692898272553, "grad_norm": 1.4075703470699064, "learning_rate": 1.0616075424854757e-05, "loss": 0.1799, "step": 7752 }, { "epoch": 4.960332693538068, "grad_norm": 1.2746072426898516, "learning_rate": 1.0614007193817579e-05, "loss": 0.1606, "step": 7753 }, { "epoch": 4.960972488803582, "grad_norm": 1.2652781180840704, "learning_rate": 1.0611938936416024e-05, "loss": 0.167, "step": 7754 }, { "epoch": 4.961612284069098, "grad_norm": 1.4358075547750553, "learning_rate": 1.0609870652738898e-05, "loss": 0.1648, "step": 7755 }, { "epoch": 4.962252079334613, "grad_norm": 1.3153173435110337, "learning_rate": 1.0607802342875011e-05, "loss": 0.1722, "step": 7756 }, { "epoch": 4.962891874600128, "grad_norm": 1.443113514787158, "learning_rate": 1.0605734006913175e-05, "loss": 0.188, "step": 7757 }, { "epoch": 4.963531669865643, "grad_norm": 1.263319901713287, "learning_rate": 1.0603665644942196e-05, "loss": 0.1759, "step": 7758 }, { "epoch": 4.964171465131158, "grad_norm": 1.297914800312607, "learning_rate": 1.0601597257050891e-05, "loss": 0.1581, "step": 7759 }, { "epoch": 4.964811260396673, "grad_norm": 1.290297163755114, "learning_rate": 1.0599528843328066e-05, "loss": 0.1598, "step": 7760 }, { "epoch": 4.9654510556621885, "grad_norm": 1.2558577674401923, "learning_rate": 1.0597460403862543e-05, "loss": 0.1688, "step": 7761 }, { "epoch": 4.966090850927703, "grad_norm": 1.238369673156077, "learning_rate": 1.0595391938743131e-05, "loss": 0.1686, "step": 7762 }, { "epoch": 4.966730646193218, "grad_norm": 1.3702617505218582, "learning_rate": 1.0593323448058647e-05, "loss": 0.169, "step": 7763 }, { "epoch": 4.967370441458733, "grad_norm": 1.270856408231226, "learning_rate": 1.0591254931897912e-05, "loss": 0.1612, "step": 7764 }, { "epoch": 4.968010236724249, "grad_norm": 1.323517765928652, "learning_rate": 1.058918639034974e-05, "loss": 0.1634, "step": 7765 }, { "epoch": 4.968650031989763, "grad_norm": 1.3188418837767657, "learning_rate": 1.0587117823502952e-05, "loss": 0.168, "step": 7766 }, { "epoch": 4.969289827255278, "grad_norm": 1.3788423560008978, "learning_rate": 1.0585049231446371e-05, "loss": 0.1759, "step": 7767 }, { "epoch": 4.969929622520794, "grad_norm": 1.31002821361093, "learning_rate": 1.0582980614268819e-05, "loss": 0.165, "step": 7768 }, { "epoch": 4.970569417786308, "grad_norm": 1.3040245099259116, "learning_rate": 1.0580911972059114e-05, "loss": 0.1542, "step": 7769 }, { "epoch": 4.971209213051823, "grad_norm": 1.2813272058481726, "learning_rate": 1.0578843304906081e-05, "loss": 0.1733, "step": 7770 }, { "epoch": 4.971849008317339, "grad_norm": 1.084328471232194, "learning_rate": 1.0576774612898545e-05, "loss": 0.1417, "step": 7771 }, { "epoch": 4.972488803582854, "grad_norm": 1.4484989058944586, "learning_rate": 1.0574705896125336e-05, "loss": 0.1686, "step": 7772 }, { "epoch": 4.973128598848368, "grad_norm": 1.247708402523089, "learning_rate": 1.0572637154675276e-05, "loss": 0.1488, "step": 7773 }, { "epoch": 4.9737683941138835, "grad_norm": 1.2754813157429887, "learning_rate": 1.05705683886372e-05, "loss": 0.1548, "step": 7774 }, { "epoch": 4.974408189379399, "grad_norm": 1.2515623400444769, "learning_rate": 1.0568499598099926e-05, "loss": 0.1576, "step": 7775 }, { "epoch": 4.975047984644913, "grad_norm": 1.110437495561153, "learning_rate": 1.0566430783152295e-05, "loss": 0.1505, "step": 7776 }, { "epoch": 4.975687779910428, "grad_norm": 1.3654610485316754, "learning_rate": 1.0564361943883133e-05, "loss": 0.1755, "step": 7777 }, { "epoch": 4.976327575175944, "grad_norm": 1.1591689773150284, "learning_rate": 1.0562293080381272e-05, "loss": 0.1584, "step": 7778 }, { "epoch": 4.976967370441459, "grad_norm": 1.2107082927134853, "learning_rate": 1.0560224192735547e-05, "loss": 0.1587, "step": 7779 }, { "epoch": 4.977607165706974, "grad_norm": 1.193208587257503, "learning_rate": 1.055815528103479e-05, "loss": 0.1664, "step": 7780 }, { "epoch": 4.978246960972489, "grad_norm": 1.139281280822756, "learning_rate": 1.0556086345367837e-05, "loss": 0.1473, "step": 7781 }, { "epoch": 4.978886756238004, "grad_norm": 1.2161826430414127, "learning_rate": 1.055401738582353e-05, "loss": 0.1532, "step": 7782 }, { "epoch": 4.979526551503519, "grad_norm": 1.2498698699035018, "learning_rate": 1.0551948402490701e-05, "loss": 0.1672, "step": 7783 }, { "epoch": 4.980166346769034, "grad_norm": 1.1695768238868156, "learning_rate": 1.0549879395458188e-05, "loss": 0.1566, "step": 7784 }, { "epoch": 4.980806142034549, "grad_norm": 1.186394689802295, "learning_rate": 1.0547810364814832e-05, "loss": 0.1513, "step": 7785 }, { "epoch": 4.981445937300064, "grad_norm": 1.2567557162448257, "learning_rate": 1.0545741310649472e-05, "loss": 0.1587, "step": 7786 }, { "epoch": 4.982085732565579, "grad_norm": 1.3093128904900555, "learning_rate": 1.0543672233050955e-05, "loss": 0.1661, "step": 7787 }, { "epoch": 4.982725527831094, "grad_norm": 1.3324341405904987, "learning_rate": 1.0541603132108116e-05, "loss": 0.1596, "step": 7788 }, { "epoch": 4.983365323096609, "grad_norm": 1.4200769190483327, "learning_rate": 1.0539534007909805e-05, "loss": 0.1709, "step": 7789 }, { "epoch": 4.984005118362124, "grad_norm": 1.1366929982027776, "learning_rate": 1.0537464860544861e-05, "loss": 0.1507, "step": 7790 }, { "epoch": 4.984644913627639, "grad_norm": 1.2928892502274805, "learning_rate": 1.0535395690102132e-05, "loss": 0.1635, "step": 7791 }, { "epoch": 4.985284708893154, "grad_norm": 1.203389007628211, "learning_rate": 1.0533326496670466e-05, "loss": 0.1614, "step": 7792 }, { "epoch": 4.985924504158669, "grad_norm": 1.2162946497091807, "learning_rate": 1.0531257280338708e-05, "loss": 0.1554, "step": 7793 }, { "epoch": 4.9865642994241846, "grad_norm": 1.050079910550055, "learning_rate": 1.0529188041195708e-05, "loss": 0.1443, "step": 7794 }, { "epoch": 4.987204094689699, "grad_norm": 1.228363930819911, "learning_rate": 1.0527118779330312e-05, "loss": 0.1666, "step": 7795 }, { "epoch": 4.987843889955214, "grad_norm": 1.2665013210499199, "learning_rate": 1.0525049494831375e-05, "loss": 0.1641, "step": 7796 }, { "epoch": 4.9884836852207295, "grad_norm": 1.103644898936429, "learning_rate": 1.0522980187787748e-05, "loss": 0.1482, "step": 7797 }, { "epoch": 4.989123480486245, "grad_norm": 1.204061775779016, "learning_rate": 1.0520910858288282e-05, "loss": 0.1603, "step": 7798 }, { "epoch": 4.989763275751759, "grad_norm": 1.2707137306050713, "learning_rate": 1.0518841506421831e-05, "loss": 0.1616, "step": 7799 }, { "epoch": 4.990403071017274, "grad_norm": 1.253875063234316, "learning_rate": 1.0516772132277247e-05, "loss": 0.1605, "step": 7800 }, { "epoch": 4.99104286628279, "grad_norm": 1.3077065919204416, "learning_rate": 1.0514702735943385e-05, "loss": 0.172, "step": 7801 }, { "epoch": 4.991682661548305, "grad_norm": 1.1720500397164568, "learning_rate": 1.0512633317509106e-05, "loss": 0.1401, "step": 7802 }, { "epoch": 4.992322456813819, "grad_norm": 1.2654101346480955, "learning_rate": 1.0510563877063263e-05, "loss": 0.1657, "step": 7803 }, { "epoch": 4.992962252079335, "grad_norm": 1.2459792719419047, "learning_rate": 1.0508494414694717e-05, "loss": 0.167, "step": 7804 }, { "epoch": 4.99360204734485, "grad_norm": 1.2384640926783461, "learning_rate": 1.0506424930492326e-05, "loss": 0.1648, "step": 7805 }, { "epoch": 4.994241842610364, "grad_norm": 1.2291524681266321, "learning_rate": 1.0504355424544945e-05, "loss": 0.1537, "step": 7806 }, { "epoch": 4.99488163787588, "grad_norm": 1.1183653798009312, "learning_rate": 1.0502285896941441e-05, "loss": 0.1384, "step": 7807 }, { "epoch": 4.995521433141395, "grad_norm": 1.271461952614941, "learning_rate": 1.0500216347770675e-05, "loss": 0.1762, "step": 7808 }, { "epoch": 4.99616122840691, "grad_norm": 1.1553305119072017, "learning_rate": 1.049814677712151e-05, "loss": 0.154, "step": 7809 }, { "epoch": 4.9968010236724245, "grad_norm": 1.293757626946663, "learning_rate": 1.0496077185082808e-05, "loss": 0.1642, "step": 7810 }, { "epoch": 4.99744081893794, "grad_norm": 1.315237287616035, "learning_rate": 1.049400757174343e-05, "loss": 0.1649, "step": 7811 }, { "epoch": 4.998080614203455, "grad_norm": 1.1041034645156385, "learning_rate": 1.0491937937192252e-05, "loss": 0.1477, "step": 7812 }, { "epoch": 4.9987204094689695, "grad_norm": 1.1850254518582408, "learning_rate": 1.0489868281518129e-05, "loss": 0.1616, "step": 7813 }, { "epoch": 4.999360204734485, "grad_norm": 1.134263028917835, "learning_rate": 1.0487798604809938e-05, "loss": 0.1483, "step": 7814 }, { "epoch": 5.0, "grad_norm": 1.1549245927814091, "learning_rate": 1.0485728907156539e-05, "loss": 0.1552, "step": 7815 }, { "epoch": 5.000639795265515, "grad_norm": 0.940814419556263, "learning_rate": 1.0483659188646806e-05, "loss": 0.1215, "step": 7816 }, { "epoch": 5.00127959053103, "grad_norm": 0.8905067443791923, "learning_rate": 1.0481589449369606e-05, "loss": 0.1244, "step": 7817 }, { "epoch": 5.001919385796545, "grad_norm": 1.0167757957469885, "learning_rate": 1.0479519689413813e-05, "loss": 0.1247, "step": 7818 }, { "epoch": 5.00255918106206, "grad_norm": 0.9443874269768971, "learning_rate": 1.04774499088683e-05, "loss": 0.1229, "step": 7819 }, { "epoch": 5.0031989763275755, "grad_norm": 0.8430588552057678, "learning_rate": 1.0475380107821936e-05, "loss": 0.1151, "step": 7820 }, { "epoch": 5.00383877159309, "grad_norm": 0.9378234315225789, "learning_rate": 1.0473310286363595e-05, "loss": 0.1207, "step": 7821 }, { "epoch": 5.004478566858605, "grad_norm": 1.0343868198950101, "learning_rate": 1.0471240444582151e-05, "loss": 0.118, "step": 7822 }, { "epoch": 5.00511836212412, "grad_norm": 0.9939167415025975, "learning_rate": 1.0469170582566482e-05, "loss": 0.1165, "step": 7823 }, { "epoch": 5.005758157389636, "grad_norm": 0.9384711482329792, "learning_rate": 1.0467100700405464e-05, "loss": 0.1117, "step": 7824 }, { "epoch": 5.00639795265515, "grad_norm": 0.9505656866963262, "learning_rate": 1.046503079818797e-05, "loss": 0.1092, "step": 7825 }, { "epoch": 5.007037747920665, "grad_norm": 1.064553198392962, "learning_rate": 1.0462960876002884e-05, "loss": 0.1186, "step": 7826 }, { "epoch": 5.007677543186181, "grad_norm": 1.1697018703774575, "learning_rate": 1.046089093393908e-05, "loss": 0.1194, "step": 7827 }, { "epoch": 5.008317338451695, "grad_norm": 1.0276672850769524, "learning_rate": 1.0458820972085438e-05, "loss": 0.119, "step": 7828 }, { "epoch": 5.00895713371721, "grad_norm": 1.0731153240494569, "learning_rate": 1.0456750990530841e-05, "loss": 0.1139, "step": 7829 }, { "epoch": 5.009596928982726, "grad_norm": 1.0869291767482419, "learning_rate": 1.0454680989364168e-05, "loss": 0.1227, "step": 7830 }, { "epoch": 5.010236724248241, "grad_norm": 1.0114938898431873, "learning_rate": 1.0452610968674303e-05, "loss": 0.1175, "step": 7831 }, { "epoch": 5.010876519513755, "grad_norm": 1.0152457309936913, "learning_rate": 1.0450540928550126e-05, "loss": 0.1177, "step": 7832 }, { "epoch": 5.0115163147792705, "grad_norm": 1.149274728039979, "learning_rate": 1.0448470869080525e-05, "loss": 0.1222, "step": 7833 }, { "epoch": 5.012156110044786, "grad_norm": 1.0139191701914982, "learning_rate": 1.0446400790354384e-05, "loss": 0.1179, "step": 7834 }, { "epoch": 5.012795905310301, "grad_norm": 1.023966439180677, "learning_rate": 1.0444330692460584e-05, "loss": 0.1078, "step": 7835 }, { "epoch": 5.0134357005758154, "grad_norm": 1.0390909805491522, "learning_rate": 1.0442260575488019e-05, "loss": 0.1122, "step": 7836 }, { "epoch": 5.014075495841331, "grad_norm": 1.05129692624156, "learning_rate": 1.0440190439525565e-05, "loss": 0.1231, "step": 7837 }, { "epoch": 5.014715291106846, "grad_norm": 0.9973991179676686, "learning_rate": 1.0438120284662121e-05, "loss": 0.1084, "step": 7838 }, { "epoch": 5.015355086372361, "grad_norm": 0.9085993585290661, "learning_rate": 1.0436050110986571e-05, "loss": 0.1115, "step": 7839 }, { "epoch": 5.015994881637876, "grad_norm": 0.9805272702668791, "learning_rate": 1.0433979918587802e-05, "loss": 0.1161, "step": 7840 }, { "epoch": 5.016634676903391, "grad_norm": 1.1056742100678254, "learning_rate": 1.0431909707554712e-05, "loss": 0.1159, "step": 7841 }, { "epoch": 5.017274472168906, "grad_norm": 0.9374943457863827, "learning_rate": 1.0429839477976184e-05, "loss": 0.1132, "step": 7842 }, { "epoch": 5.017914267434421, "grad_norm": 1.0585923509959632, "learning_rate": 1.0427769229941116e-05, "loss": 0.1173, "step": 7843 }, { "epoch": 5.018554062699936, "grad_norm": 0.9959774503908576, "learning_rate": 1.0425698963538398e-05, "loss": 0.1152, "step": 7844 }, { "epoch": 5.019193857965451, "grad_norm": 1.0524242310050718, "learning_rate": 1.042362867885692e-05, "loss": 0.1142, "step": 7845 }, { "epoch": 5.019833653230966, "grad_norm": 1.0559478315803144, "learning_rate": 1.0421558375985586e-05, "loss": 0.122, "step": 7846 }, { "epoch": 5.020473448496481, "grad_norm": 1.0659298457787207, "learning_rate": 1.0419488055013278e-05, "loss": 0.111, "step": 7847 }, { "epoch": 5.021113243761996, "grad_norm": 1.0091479355663235, "learning_rate": 1.0417417716028903e-05, "loss": 0.1122, "step": 7848 }, { "epoch": 5.021753039027511, "grad_norm": 0.9709078428287903, "learning_rate": 1.0415347359121355e-05, "loss": 0.1146, "step": 7849 }, { "epoch": 5.022392834293027, "grad_norm": 1.0717919497133952, "learning_rate": 1.041327698437953e-05, "loss": 0.1156, "step": 7850 }, { "epoch": 5.023032629558541, "grad_norm": 0.9711757569972258, "learning_rate": 1.0411206591892327e-05, "loss": 0.1157, "step": 7851 }, { "epoch": 5.023672424824056, "grad_norm": 1.0736493415014945, "learning_rate": 1.040913618174864e-05, "loss": 0.1124, "step": 7852 }, { "epoch": 5.024312220089572, "grad_norm": 0.9484891114264588, "learning_rate": 1.0407065754037376e-05, "loss": 0.1125, "step": 7853 }, { "epoch": 5.024952015355086, "grad_norm": 0.8713999126242264, "learning_rate": 1.0404995308847434e-05, "loss": 0.1038, "step": 7854 }, { "epoch": 5.025591810620601, "grad_norm": 1.0160356980232264, "learning_rate": 1.0402924846267713e-05, "loss": 0.1163, "step": 7855 }, { "epoch": 5.0262316058861165, "grad_norm": 1.0003132614876549, "learning_rate": 1.0400854366387115e-05, "loss": 0.116, "step": 7856 }, { "epoch": 5.026871401151632, "grad_norm": 0.9877266710213994, "learning_rate": 1.0398783869294541e-05, "loss": 0.1108, "step": 7857 }, { "epoch": 5.027511196417146, "grad_norm": 1.0833954035985758, "learning_rate": 1.0396713355078902e-05, "loss": 0.1126, "step": 7858 }, { "epoch": 5.028150991682661, "grad_norm": 1.0142278013514496, "learning_rate": 1.0394642823829098e-05, "loss": 0.1166, "step": 7859 }, { "epoch": 5.028790786948177, "grad_norm": 1.0085139881997205, "learning_rate": 1.039257227563403e-05, "loss": 0.1139, "step": 7860 }, { "epoch": 5.029430582213692, "grad_norm": 0.8649856328471697, "learning_rate": 1.0390501710582608e-05, "loss": 0.109, "step": 7861 }, { "epoch": 5.030070377479206, "grad_norm": 1.0534819606907146, "learning_rate": 1.0388431128763735e-05, "loss": 0.12, "step": 7862 }, { "epoch": 5.030710172744722, "grad_norm": 0.9333645663888485, "learning_rate": 1.038636053026632e-05, "loss": 0.1129, "step": 7863 }, { "epoch": 5.031349968010237, "grad_norm": 0.9597643844298892, "learning_rate": 1.0384289915179273e-05, "loss": 0.1158, "step": 7864 }, { "epoch": 5.031989763275751, "grad_norm": 1.0206535357277275, "learning_rate": 1.0382219283591502e-05, "loss": 0.115, "step": 7865 }, { "epoch": 5.032629558541267, "grad_norm": 1.067131918299212, "learning_rate": 1.0380148635591916e-05, "loss": 0.1173, "step": 7866 }, { "epoch": 5.033269353806782, "grad_norm": 0.9653001879933599, "learning_rate": 1.0378077971269418e-05, "loss": 0.1208, "step": 7867 }, { "epoch": 5.033909149072297, "grad_norm": 1.0027578269767594, "learning_rate": 1.0376007290712928e-05, "loss": 0.1082, "step": 7868 }, { "epoch": 5.0345489443378115, "grad_norm": 0.9938356291409577, "learning_rate": 1.0373936594011354e-05, "loss": 0.1149, "step": 7869 }, { "epoch": 5.035188739603327, "grad_norm": 1.0415511791872878, "learning_rate": 1.0371865881253606e-05, "loss": 0.1193, "step": 7870 }, { "epoch": 5.035828534868842, "grad_norm": 1.0311008175126126, "learning_rate": 1.03697951525286e-05, "loss": 0.1194, "step": 7871 }, { "epoch": 5.036468330134357, "grad_norm": 0.944182352977368, "learning_rate": 1.0367724407925248e-05, "loss": 0.11, "step": 7872 }, { "epoch": 5.037108125399872, "grad_norm": 1.008566567619989, "learning_rate": 1.0365653647532466e-05, "loss": 0.1147, "step": 7873 }, { "epoch": 5.037747920665387, "grad_norm": 1.0129772654794829, "learning_rate": 1.0363582871439164e-05, "loss": 0.1138, "step": 7874 }, { "epoch": 5.038387715930902, "grad_norm": 1.1442279724901765, "learning_rate": 1.0361512079734259e-05, "loss": 0.1241, "step": 7875 }, { "epoch": 5.039027511196418, "grad_norm": 0.9892634941816708, "learning_rate": 1.0359441272506672e-05, "loss": 0.1082, "step": 7876 }, { "epoch": 5.039667306461932, "grad_norm": 0.985707652746079, "learning_rate": 1.0357370449845313e-05, "loss": 0.1105, "step": 7877 }, { "epoch": 5.040307101727447, "grad_norm": 0.953803343027863, "learning_rate": 1.0355299611839104e-05, "loss": 0.1119, "step": 7878 }, { "epoch": 5.0409468969929625, "grad_norm": 1.0087279659085804, "learning_rate": 1.0353228758576964e-05, "loss": 0.1184, "step": 7879 }, { "epoch": 5.041586692258477, "grad_norm": 1.0910458642353227, "learning_rate": 1.0351157890147808e-05, "loss": 0.1201, "step": 7880 }, { "epoch": 5.042226487523992, "grad_norm": 1.0662000171749257, "learning_rate": 1.0349087006640557e-05, "loss": 0.1164, "step": 7881 }, { "epoch": 5.042866282789507, "grad_norm": 1.0153325944783176, "learning_rate": 1.0347016108144128e-05, "loss": 0.113, "step": 7882 }, { "epoch": 5.043506078055023, "grad_norm": 0.9393291358543562, "learning_rate": 1.0344945194747445e-05, "loss": 0.1153, "step": 7883 }, { "epoch": 5.044145873320537, "grad_norm": 1.0647339853083302, "learning_rate": 1.0342874266539431e-05, "loss": 0.1183, "step": 7884 }, { "epoch": 5.044785668586052, "grad_norm": 0.9586133817051858, "learning_rate": 1.0340803323609006e-05, "loss": 0.1146, "step": 7885 }, { "epoch": 5.045425463851568, "grad_norm": 1.097091133137834, "learning_rate": 1.0338732366045092e-05, "loss": 0.1187, "step": 7886 }, { "epoch": 5.046065259117083, "grad_norm": 0.9996779244811895, "learning_rate": 1.0336661393936611e-05, "loss": 0.1117, "step": 7887 }, { "epoch": 5.046705054382597, "grad_norm": 0.9839369663538562, "learning_rate": 1.0334590407372493e-05, "loss": 0.1104, "step": 7888 }, { "epoch": 5.047344849648113, "grad_norm": 0.961315335212631, "learning_rate": 1.0332519406441657e-05, "loss": 0.1204, "step": 7889 }, { "epoch": 5.047984644913628, "grad_norm": 1.0815846057701413, "learning_rate": 1.0330448391233026e-05, "loss": 0.1155, "step": 7890 }, { "epoch": 5.048624440179142, "grad_norm": 1.067486076986388, "learning_rate": 1.032837736183553e-05, "loss": 0.1171, "step": 7891 }, { "epoch": 5.0492642354446575, "grad_norm": 1.0394601734629172, "learning_rate": 1.0326306318338095e-05, "loss": 0.1187, "step": 7892 }, { "epoch": 5.049904030710173, "grad_norm": 1.0163713986168985, "learning_rate": 1.0324235260829647e-05, "loss": 0.1126, "step": 7893 }, { "epoch": 5.050543825975688, "grad_norm": 0.9834668289897829, "learning_rate": 1.0322164189399114e-05, "loss": 0.1101, "step": 7894 }, { "epoch": 5.0511836212412025, "grad_norm": 0.9613495753272637, "learning_rate": 1.0320093104135425e-05, "loss": 0.105, "step": 7895 }, { "epoch": 5.051823416506718, "grad_norm": 1.092444848223603, "learning_rate": 1.0318022005127506e-05, "loss": 0.1089, "step": 7896 }, { "epoch": 5.052463211772233, "grad_norm": 1.0294521997540618, "learning_rate": 1.0315950892464285e-05, "loss": 0.1165, "step": 7897 }, { "epoch": 5.053103007037748, "grad_norm": 1.0703187511924197, "learning_rate": 1.03138797662347e-05, "loss": 0.1191, "step": 7898 }, { "epoch": 5.053742802303263, "grad_norm": 1.0419336979421052, "learning_rate": 1.0311808626527673e-05, "loss": 0.1235, "step": 7899 }, { "epoch": 5.054382597568778, "grad_norm": 1.1209832456453013, "learning_rate": 1.0309737473432137e-05, "loss": 0.1274, "step": 7900 }, { "epoch": 5.055022392834293, "grad_norm": 1.0277682337431369, "learning_rate": 1.0307666307037027e-05, "loss": 0.1142, "step": 7901 }, { "epoch": 5.0556621880998085, "grad_norm": 1.1276119714351853, "learning_rate": 1.0305595127431272e-05, "loss": 0.1216, "step": 7902 }, { "epoch": 5.056301983365323, "grad_norm": 1.0178195959920233, "learning_rate": 1.0303523934703807e-05, "loss": 0.1167, "step": 7903 }, { "epoch": 5.056941778630838, "grad_norm": 1.110912134672578, "learning_rate": 1.0301452728943566e-05, "loss": 0.1214, "step": 7904 }, { "epoch": 5.057581573896353, "grad_norm": 1.02703506395037, "learning_rate": 1.029938151023948e-05, "loss": 0.1101, "step": 7905 }, { "epoch": 5.058221369161868, "grad_norm": 1.056035242118623, "learning_rate": 1.0297310278680482e-05, "loss": 0.1149, "step": 7906 }, { "epoch": 5.058861164427383, "grad_norm": 1.080755583716169, "learning_rate": 1.0295239034355512e-05, "loss": 0.1183, "step": 7907 }, { "epoch": 5.059500959692898, "grad_norm": 0.9852582579514056, "learning_rate": 1.0293167777353501e-05, "loss": 0.118, "step": 7908 }, { "epoch": 5.060140754958414, "grad_norm": 1.0389198104154749, "learning_rate": 1.0291096507763386e-05, "loss": 0.1195, "step": 7909 }, { "epoch": 5.060780550223928, "grad_norm": 1.0356682278981524, "learning_rate": 1.028902522567411e-05, "loss": 0.1184, "step": 7910 }, { "epoch": 5.061420345489443, "grad_norm": 0.959067438342726, "learning_rate": 1.0286953931174601e-05, "loss": 0.1121, "step": 7911 }, { "epoch": 5.062060140754959, "grad_norm": 0.9469143889710667, "learning_rate": 1.02848826243538e-05, "loss": 0.1126, "step": 7912 }, { "epoch": 5.062699936020474, "grad_norm": 0.9646403224176562, "learning_rate": 1.028281130530065e-05, "loss": 0.1187, "step": 7913 }, { "epoch": 5.063339731285988, "grad_norm": 0.9845907834751397, "learning_rate": 1.0280739974104081e-05, "loss": 0.1088, "step": 7914 }, { "epoch": 5.0639795265515035, "grad_norm": 0.9515065968600305, "learning_rate": 1.0278668630853039e-05, "loss": 0.1141, "step": 7915 }, { "epoch": 5.064619321817019, "grad_norm": 1.08851681659945, "learning_rate": 1.0276597275636462e-05, "loss": 0.115, "step": 7916 }, { "epoch": 5.065259117082533, "grad_norm": 1.106836485832815, "learning_rate": 1.0274525908543289e-05, "loss": 0.1202, "step": 7917 }, { "epoch": 5.0658989123480485, "grad_norm": 0.9991068177207882, "learning_rate": 1.0272454529662463e-05, "loss": 0.1058, "step": 7918 }, { "epoch": 5.066538707613564, "grad_norm": 0.9432910528253435, "learning_rate": 1.0270383139082924e-05, "loss": 0.112, "step": 7919 }, { "epoch": 5.067178502879079, "grad_norm": 1.0770375568765247, "learning_rate": 1.0268311736893612e-05, "loss": 0.118, "step": 7920 }, { "epoch": 5.067818298144593, "grad_norm": 1.0078361674162666, "learning_rate": 1.0266240323183474e-05, "loss": 0.1153, "step": 7921 }, { "epoch": 5.068458093410109, "grad_norm": 1.0008915239996985, "learning_rate": 1.0264168898041449e-05, "loss": 0.1123, "step": 7922 }, { "epoch": 5.069097888675624, "grad_norm": 1.031115505966658, "learning_rate": 1.0262097461556482e-05, "loss": 0.1129, "step": 7923 }, { "epoch": 5.069737683941139, "grad_norm": 0.9306271355302201, "learning_rate": 1.0260026013817514e-05, "loss": 0.1127, "step": 7924 }, { "epoch": 5.070377479206654, "grad_norm": 1.0591024446236994, "learning_rate": 1.0257954554913496e-05, "loss": 0.1187, "step": 7925 }, { "epoch": 5.071017274472169, "grad_norm": 1.1387130300724846, "learning_rate": 1.0255883084933364e-05, "loss": 0.1206, "step": 7926 }, { "epoch": 5.071657069737684, "grad_norm": 0.9387760061526911, "learning_rate": 1.0253811603966069e-05, "loss": 0.1134, "step": 7927 }, { "epoch": 5.0722968650031985, "grad_norm": 1.0780530779597917, "learning_rate": 1.0251740112100557e-05, "loss": 0.1226, "step": 7928 }, { "epoch": 5.072936660268714, "grad_norm": 1.053069399026103, "learning_rate": 1.0249668609425769e-05, "loss": 0.1216, "step": 7929 }, { "epoch": 5.073576455534229, "grad_norm": 1.0094118929306726, "learning_rate": 1.0247597096030655e-05, "loss": 0.1137, "step": 7930 }, { "epoch": 5.074216250799744, "grad_norm": 1.134365537431779, "learning_rate": 1.0245525572004165e-05, "loss": 0.1237, "step": 7931 }, { "epoch": 5.074856046065259, "grad_norm": 1.0900235048116629, "learning_rate": 1.0243454037435241e-05, "loss": 0.1216, "step": 7932 }, { "epoch": 5.075495841330774, "grad_norm": 0.9985475572969188, "learning_rate": 1.0241382492412837e-05, "loss": 0.1113, "step": 7933 }, { "epoch": 5.076135636596289, "grad_norm": 0.9632600708516411, "learning_rate": 1.0239310937025898e-05, "loss": 0.1115, "step": 7934 }, { "epoch": 5.076775431861805, "grad_norm": 1.0636560583936656, "learning_rate": 1.023723937136337e-05, "loss": 0.121, "step": 7935 }, { "epoch": 5.077415227127319, "grad_norm": 1.0115790837502803, "learning_rate": 1.0235167795514207e-05, "loss": 0.1156, "step": 7936 }, { "epoch": 5.078055022392834, "grad_norm": 0.973038910329687, "learning_rate": 1.0233096209567357e-05, "loss": 0.1133, "step": 7937 }, { "epoch": 5.0786948176583495, "grad_norm": 1.100701870845409, "learning_rate": 1.0231024613611771e-05, "loss": 0.1179, "step": 7938 }, { "epoch": 5.079334612923865, "grad_norm": 0.8697042425314022, "learning_rate": 1.0228953007736397e-05, "loss": 0.1059, "step": 7939 }, { "epoch": 5.079974408189379, "grad_norm": 0.9595341394865535, "learning_rate": 1.022688139203019e-05, "loss": 0.1083, "step": 7940 }, { "epoch": 5.0806142034548945, "grad_norm": 1.0202349506386355, "learning_rate": 1.0224809766582102e-05, "loss": 0.1122, "step": 7941 }, { "epoch": 5.08125399872041, "grad_norm": 0.9617730693248161, "learning_rate": 1.0222738131481077e-05, "loss": 0.1092, "step": 7942 }, { "epoch": 5.081893793985924, "grad_norm": 0.9423706700129928, "learning_rate": 1.0220666486816078e-05, "loss": 0.1064, "step": 7943 }, { "epoch": 5.082533589251439, "grad_norm": 1.0319802223541086, "learning_rate": 1.021859483267605e-05, "loss": 0.1161, "step": 7944 }, { "epoch": 5.083173384516955, "grad_norm": 1.0159366128518665, "learning_rate": 1.0216523169149946e-05, "loss": 0.1118, "step": 7945 }, { "epoch": 5.08381317978247, "grad_norm": 1.0121456239118904, "learning_rate": 1.0214451496326727e-05, "loss": 0.1065, "step": 7946 }, { "epoch": 5.084452975047984, "grad_norm": 1.0316947250795407, "learning_rate": 1.0212379814295339e-05, "loss": 0.1179, "step": 7947 }, { "epoch": 5.0850927703135, "grad_norm": 0.9526418960097691, "learning_rate": 1.0210308123144743e-05, "loss": 0.1128, "step": 7948 }, { "epoch": 5.085732565579015, "grad_norm": 0.9358471270583504, "learning_rate": 1.0208236422963888e-05, "loss": 0.1109, "step": 7949 }, { "epoch": 5.08637236084453, "grad_norm": 0.9451076245781427, "learning_rate": 1.0206164713841731e-05, "loss": 0.1165, "step": 7950 }, { "epoch": 5.0870121561100445, "grad_norm": 0.936961786279276, "learning_rate": 1.020409299586723e-05, "loss": 0.1066, "step": 7951 }, { "epoch": 5.08765195137556, "grad_norm": 0.9440459224325375, "learning_rate": 1.0202021269129336e-05, "loss": 0.1156, "step": 7952 }, { "epoch": 5.088291746641075, "grad_norm": 1.0651794146647031, "learning_rate": 1.0199949533717011e-05, "loss": 0.1127, "step": 7953 }, { "epoch": 5.0889315419065895, "grad_norm": 1.0503575980024022, "learning_rate": 1.0197877789719206e-05, "loss": 0.1219, "step": 7954 }, { "epoch": 5.089571337172105, "grad_norm": 1.0213480999387927, "learning_rate": 1.0195806037224882e-05, "loss": 0.1158, "step": 7955 }, { "epoch": 5.09021113243762, "grad_norm": 1.053831183502866, "learning_rate": 1.0193734276322998e-05, "loss": 0.1192, "step": 7956 }, { "epoch": 5.090850927703135, "grad_norm": 1.0056794243163074, "learning_rate": 1.0191662507102505e-05, "loss": 0.113, "step": 7957 }, { "epoch": 5.09149072296865, "grad_norm": 0.9877025916247877, "learning_rate": 1.0189590729652366e-05, "loss": 0.1159, "step": 7958 }, { "epoch": 5.092130518234165, "grad_norm": 0.9892618574013575, "learning_rate": 1.0187518944061536e-05, "loss": 0.1179, "step": 7959 }, { "epoch": 5.09277031349968, "grad_norm": 1.0699211104693913, "learning_rate": 1.0185447150418976e-05, "loss": 0.1173, "step": 7960 }, { "epoch": 5.0934101087651955, "grad_norm": 0.8671292885129267, "learning_rate": 1.018337534881365e-05, "loss": 0.1172, "step": 7961 }, { "epoch": 5.09404990403071, "grad_norm": 0.9851814579252675, "learning_rate": 1.0181303539334509e-05, "loss": 0.1053, "step": 7962 }, { "epoch": 5.094689699296225, "grad_norm": 1.0463769361979218, "learning_rate": 1.0179231722070518e-05, "loss": 0.1183, "step": 7963 }, { "epoch": 5.0953294945617404, "grad_norm": 0.9737100956352368, "learning_rate": 1.0177159897110637e-05, "loss": 0.1172, "step": 7964 }, { "epoch": 5.095969289827256, "grad_norm": 0.990080845321647, "learning_rate": 1.0175088064543822e-05, "loss": 0.1178, "step": 7965 }, { "epoch": 5.09660908509277, "grad_norm": 0.9690760078431959, "learning_rate": 1.017301622445904e-05, "loss": 0.1239, "step": 7966 }, { "epoch": 5.097248880358285, "grad_norm": 1.0393238561056102, "learning_rate": 1.0170944376945247e-05, "loss": 0.1279, "step": 7967 }, { "epoch": 5.097888675623801, "grad_norm": 1.050918847807036, "learning_rate": 1.0168872522091407e-05, "loss": 0.1184, "step": 7968 }, { "epoch": 5.098528470889315, "grad_norm": 0.9550207959082851, "learning_rate": 1.016680065998648e-05, "loss": 0.1098, "step": 7969 }, { "epoch": 5.09916826615483, "grad_norm": 0.9814709879350093, "learning_rate": 1.0164728790719432e-05, "loss": 0.1129, "step": 7970 }, { "epoch": 5.099808061420346, "grad_norm": 0.9608589102500762, "learning_rate": 1.0162656914379224e-05, "loss": 0.1117, "step": 7971 }, { "epoch": 5.100447856685861, "grad_norm": 1.121894512216251, "learning_rate": 1.0160585031054813e-05, "loss": 0.1162, "step": 7972 }, { "epoch": 5.101087651951375, "grad_norm": 1.0459202451871914, "learning_rate": 1.0158513140835169e-05, "loss": 0.1181, "step": 7973 }, { "epoch": 5.1017274472168905, "grad_norm": 1.0301937864288244, "learning_rate": 1.0156441243809252e-05, "loss": 0.1039, "step": 7974 }, { "epoch": 5.102367242482406, "grad_norm": 1.1066227045677925, "learning_rate": 1.0154369340066026e-05, "loss": 0.1093, "step": 7975 }, { "epoch": 5.103007037747921, "grad_norm": 1.063588075269982, "learning_rate": 1.0152297429694455e-05, "loss": 0.1195, "step": 7976 }, { "epoch": 5.1036468330134355, "grad_norm": 1.1686312932334115, "learning_rate": 1.0150225512783502e-05, "loss": 0.1175, "step": 7977 }, { "epoch": 5.104286628278951, "grad_norm": 1.0260720585800136, "learning_rate": 1.0148153589422136e-05, "loss": 0.1196, "step": 7978 }, { "epoch": 5.104926423544466, "grad_norm": 1.113325041937606, "learning_rate": 1.0146081659699318e-05, "loss": 0.1205, "step": 7979 }, { "epoch": 5.10556621880998, "grad_norm": 0.9677140492569327, "learning_rate": 1.0144009723704014e-05, "loss": 0.1059, "step": 7980 }, { "epoch": 5.106206014075496, "grad_norm": 0.9760142312588932, "learning_rate": 1.0141937781525186e-05, "loss": 0.1167, "step": 7981 }, { "epoch": 5.106845809341011, "grad_norm": 1.0626097670520716, "learning_rate": 1.0139865833251801e-05, "loss": 0.1141, "step": 7982 }, { "epoch": 5.107485604606526, "grad_norm": 1.0069349684443543, "learning_rate": 1.013779387897283e-05, "loss": 0.115, "step": 7983 }, { "epoch": 5.108125399872041, "grad_norm": 1.0616479195344182, "learning_rate": 1.013572191877723e-05, "loss": 0.1186, "step": 7984 }, { "epoch": 5.108765195137556, "grad_norm": 1.1173983502794498, "learning_rate": 1.0133649952753977e-05, "loss": 0.1169, "step": 7985 }, { "epoch": 5.109404990403071, "grad_norm": 1.0168904484187258, "learning_rate": 1.0131577980992033e-05, "loss": 0.1175, "step": 7986 }, { "epoch": 5.110044785668586, "grad_norm": 1.0673317448657342, "learning_rate": 1.012950600358036e-05, "loss": 0.1214, "step": 7987 }, { "epoch": 5.110684580934101, "grad_norm": 1.0148844530165577, "learning_rate": 1.0127434020607934e-05, "loss": 0.1147, "step": 7988 }, { "epoch": 5.111324376199616, "grad_norm": 1.0401538881509222, "learning_rate": 1.0125362032163715e-05, "loss": 0.1115, "step": 7989 }, { "epoch": 5.111964171465131, "grad_norm": 1.0279064704646828, "learning_rate": 1.0123290038336675e-05, "loss": 0.1176, "step": 7990 }, { "epoch": 5.112603966730646, "grad_norm": 1.0637144273915413, "learning_rate": 1.0121218039215779e-05, "loss": 0.1211, "step": 7991 }, { "epoch": 5.113243761996161, "grad_norm": 1.0345064715872745, "learning_rate": 1.0119146034889995e-05, "loss": 0.1139, "step": 7992 }, { "epoch": 5.113883557261676, "grad_norm": 0.8800198897927802, "learning_rate": 1.0117074025448294e-05, "loss": 0.1066, "step": 7993 }, { "epoch": 5.114523352527192, "grad_norm": 0.9693054649058072, "learning_rate": 1.0115002010979644e-05, "loss": 0.1228, "step": 7994 }, { "epoch": 5.115163147792706, "grad_norm": 1.1064169452438273, "learning_rate": 1.0112929991573011e-05, "loss": 0.1173, "step": 7995 }, { "epoch": 5.115802943058221, "grad_norm": 0.9694924262332707, "learning_rate": 1.0110857967317364e-05, "loss": 0.113, "step": 7996 }, { "epoch": 5.1164427383237365, "grad_norm": 1.0006293801030164, "learning_rate": 1.0108785938301673e-05, "loss": 0.1106, "step": 7997 }, { "epoch": 5.117082533589252, "grad_norm": 1.1256164772724788, "learning_rate": 1.010671390461491e-05, "loss": 0.1267, "step": 7998 }, { "epoch": 5.117722328854766, "grad_norm": 1.0322213773105116, "learning_rate": 1.0104641866346037e-05, "loss": 0.1213, "step": 7999 }, { "epoch": 5.1183621241202815, "grad_norm": 0.9772991966664234, "learning_rate": 1.0102569823584032e-05, "loss": 0.1177, "step": 8000 }, { "epoch": 5.119001919385797, "grad_norm": 1.006322242003033, "learning_rate": 1.0100497776417861e-05, "loss": 0.1162, "step": 8001 }, { "epoch": 5.119641714651312, "grad_norm": 1.058477822630113, "learning_rate": 1.0098425724936492e-05, "loss": 0.1093, "step": 8002 }, { "epoch": 5.120281509916826, "grad_norm": 0.8641225164424916, "learning_rate": 1.00963536692289e-05, "loss": 0.1038, "step": 8003 }, { "epoch": 5.120921305182342, "grad_norm": 0.9359989497957384, "learning_rate": 1.0094281609384051e-05, "loss": 0.1084, "step": 8004 }, { "epoch": 5.121561100447857, "grad_norm": 0.8931073434131437, "learning_rate": 1.0092209545490918e-05, "loss": 0.109, "step": 8005 }, { "epoch": 5.122200895713371, "grad_norm": 0.9323483232819987, "learning_rate": 1.009013747763847e-05, "loss": 0.1041, "step": 8006 }, { "epoch": 5.122840690978887, "grad_norm": 1.0207893387612672, "learning_rate": 1.0088065405915681e-05, "loss": 0.1182, "step": 8007 }, { "epoch": 5.123480486244402, "grad_norm": 1.0979686800077466, "learning_rate": 1.008599333041152e-05, "loss": 0.1213, "step": 8008 }, { "epoch": 5.124120281509917, "grad_norm": 1.0380651234668152, "learning_rate": 1.0083921251214959e-05, "loss": 0.1097, "step": 8009 }, { "epoch": 5.1247600767754315, "grad_norm": 1.016626000328172, "learning_rate": 1.008184916841497e-05, "loss": 0.1179, "step": 8010 }, { "epoch": 5.125399872040947, "grad_norm": 1.0442832178441748, "learning_rate": 1.0079777082100518e-05, "loss": 0.1173, "step": 8011 }, { "epoch": 5.126039667306462, "grad_norm": 1.0130278423472223, "learning_rate": 1.0077704992360585e-05, "loss": 0.1145, "step": 8012 }, { "epoch": 5.126679462571977, "grad_norm": 1.0866356204967191, "learning_rate": 1.0075632899284136e-05, "loss": 0.1169, "step": 8013 }, { "epoch": 5.127319257837492, "grad_norm": 1.1472431020784972, "learning_rate": 1.0073560802960143e-05, "loss": 0.1187, "step": 8014 }, { "epoch": 5.127959053103007, "grad_norm": 1.014807130744692, "learning_rate": 1.0071488703477584e-05, "loss": 0.1125, "step": 8015 }, { "epoch": 5.128598848368522, "grad_norm": 0.973783596796498, "learning_rate": 1.0069416600925423e-05, "loss": 0.1113, "step": 8016 }, { "epoch": 5.129238643634037, "grad_norm": 1.0281914162874934, "learning_rate": 1.006734449539264e-05, "loss": 0.116, "step": 8017 }, { "epoch": 5.129878438899552, "grad_norm": 1.0763562845399526, "learning_rate": 1.0065272386968203e-05, "loss": 0.1183, "step": 8018 }, { "epoch": 5.130518234165067, "grad_norm": 1.061517910119846, "learning_rate": 1.0063200275741083e-05, "loss": 0.1153, "step": 8019 }, { "epoch": 5.1311580294305825, "grad_norm": 1.0406574041585566, "learning_rate": 1.006112816180026e-05, "loss": 0.1168, "step": 8020 }, { "epoch": 5.131797824696097, "grad_norm": 0.954554841122485, "learning_rate": 1.0059056045234698e-05, "loss": 0.1074, "step": 8021 }, { "epoch": 5.132437619961612, "grad_norm": 1.1682167894516429, "learning_rate": 1.0056983926133377e-05, "loss": 0.1267, "step": 8022 }, { "epoch": 5.1330774152271275, "grad_norm": 1.0149106639274947, "learning_rate": 1.005491180458527e-05, "loss": 0.1165, "step": 8023 }, { "epoch": 5.133717210492643, "grad_norm": 1.1677980776265717, "learning_rate": 1.0052839680679345e-05, "loss": 0.1255, "step": 8024 }, { "epoch": 5.134357005758157, "grad_norm": 0.987687080653227, "learning_rate": 1.0050767554504581e-05, "loss": 0.1169, "step": 8025 }, { "epoch": 5.134996801023672, "grad_norm": 1.040952795296312, "learning_rate": 1.0048695426149944e-05, "loss": 0.1158, "step": 8026 }, { "epoch": 5.135636596289188, "grad_norm": 1.095468754379721, "learning_rate": 1.0046623295704414e-05, "loss": 0.1205, "step": 8027 }, { "epoch": 5.136276391554702, "grad_norm": 1.026554808170049, "learning_rate": 1.0044551163256964e-05, "loss": 0.1219, "step": 8028 }, { "epoch": 5.136916186820217, "grad_norm": 1.0366512173130795, "learning_rate": 1.0042479028896563e-05, "loss": 0.1158, "step": 8029 }, { "epoch": 5.137555982085733, "grad_norm": 1.0528881684353768, "learning_rate": 1.0040406892712192e-05, "loss": 0.1167, "step": 8030 }, { "epoch": 5.138195777351248, "grad_norm": 1.0543986910415175, "learning_rate": 1.0038334754792821e-05, "loss": 0.1195, "step": 8031 }, { "epoch": 5.138835572616762, "grad_norm": 0.9682173312370875, "learning_rate": 1.003626261522742e-05, "loss": 0.1128, "step": 8032 }, { "epoch": 5.1394753678822775, "grad_norm": 0.9842685525122242, "learning_rate": 1.0034190474104968e-05, "loss": 0.1168, "step": 8033 }, { "epoch": 5.140115163147793, "grad_norm": 1.0791425974300484, "learning_rate": 1.0032118331514438e-05, "loss": 0.1246, "step": 8034 }, { "epoch": 5.140754958413308, "grad_norm": 0.9897324043488364, "learning_rate": 1.0030046187544804e-05, "loss": 0.1086, "step": 8035 }, { "epoch": 5.1413947536788225, "grad_norm": 0.8965820487431706, "learning_rate": 1.002797404228504e-05, "loss": 0.1032, "step": 8036 }, { "epoch": 5.142034548944338, "grad_norm": 1.0789727076928957, "learning_rate": 1.002590189582412e-05, "loss": 0.1161, "step": 8037 }, { "epoch": 5.142674344209853, "grad_norm": 0.911241229986568, "learning_rate": 1.0023829748251021e-05, "loss": 0.119, "step": 8038 }, { "epoch": 5.143314139475368, "grad_norm": 1.113124407631575, "learning_rate": 1.0021757599654715e-05, "loss": 0.1221, "step": 8039 }, { "epoch": 5.143953934740883, "grad_norm": 1.0051169472855424, "learning_rate": 1.0019685450124176e-05, "loss": 0.1196, "step": 8040 }, { "epoch": 5.144593730006398, "grad_norm": 0.8501500435378891, "learning_rate": 1.0017613299748376e-05, "loss": 0.1092, "step": 8041 }, { "epoch": 5.145233525271913, "grad_norm": 1.0223823076281813, "learning_rate": 1.0015541148616294e-05, "loss": 0.1208, "step": 8042 }, { "epoch": 5.145873320537428, "grad_norm": 0.9865106584276947, "learning_rate": 1.0013468996816905e-05, "loss": 0.1148, "step": 8043 }, { "epoch": 5.146513115802943, "grad_norm": 1.002825028304972, "learning_rate": 1.0011396844439176e-05, "loss": 0.1175, "step": 8044 }, { "epoch": 5.147152911068458, "grad_norm": 0.905382896879988, "learning_rate": 1.0009324691572091e-05, "loss": 0.1103, "step": 8045 }, { "epoch": 5.1477927063339735, "grad_norm": 1.0642820485306628, "learning_rate": 1.0007252538304619e-05, "loss": 0.1182, "step": 8046 }, { "epoch": 5.148432501599488, "grad_norm": 1.0176266331857675, "learning_rate": 1.0005180384725736e-05, "loss": 0.115, "step": 8047 }, { "epoch": 5.149072296865003, "grad_norm": 1.070044197638247, "learning_rate": 1.0003108230924418e-05, "loss": 0.1265, "step": 8048 }, { "epoch": 5.149712092130518, "grad_norm": 1.2532693713872967, "learning_rate": 1.0001036076989635e-05, "loss": 0.1221, "step": 8049 }, { "epoch": 5.150351887396034, "grad_norm": 1.0769263984043023, "learning_rate": 9.998963923010365e-06, "loss": 0.1139, "step": 8050 }, { "epoch": 5.150991682661548, "grad_norm": 1.0824842864853645, "learning_rate": 9.996891769075585e-06, "loss": 0.1228, "step": 8051 }, { "epoch": 5.151631477927063, "grad_norm": 0.9337982925674505, "learning_rate": 9.994819615274267e-06, "loss": 0.1151, "step": 8052 }, { "epoch": 5.152271273192579, "grad_norm": 1.0901428118365828, "learning_rate": 9.992747461695383e-06, "loss": 0.1219, "step": 8053 }, { "epoch": 5.152911068458093, "grad_norm": 1.0392960341400879, "learning_rate": 9.990675308427912e-06, "loss": 0.1157, "step": 8054 }, { "epoch": 5.153550863723608, "grad_norm": 1.0807355188514824, "learning_rate": 9.988603155560826e-06, "loss": 0.1202, "step": 8055 }, { "epoch": 5.1541906589891235, "grad_norm": 1.0240971437560085, "learning_rate": 9.986531003183102e-06, "loss": 0.1167, "step": 8056 }, { "epoch": 5.154830454254639, "grad_norm": 1.02781841408588, "learning_rate": 9.984458851383707e-06, "loss": 0.1107, "step": 8057 }, { "epoch": 5.155470249520153, "grad_norm": 1.024508790314054, "learning_rate": 9.982386700251625e-06, "loss": 0.1228, "step": 8058 }, { "epoch": 5.1561100447856685, "grad_norm": 1.040013436869932, "learning_rate": 9.980314549875829e-06, "loss": 0.1119, "step": 8059 }, { "epoch": 5.156749840051184, "grad_norm": 1.060540252088371, "learning_rate": 9.978242400345288e-06, "loss": 0.1252, "step": 8060 }, { "epoch": 5.157389635316699, "grad_norm": 1.0182567767702526, "learning_rate": 9.976170251748982e-06, "loss": 0.1112, "step": 8061 }, { "epoch": 5.158029430582213, "grad_norm": 1.0665305173917328, "learning_rate": 9.974098104175882e-06, "loss": 0.1197, "step": 8062 }, { "epoch": 5.158669225847729, "grad_norm": 1.0816315741871356, "learning_rate": 9.97202595771496e-06, "loss": 0.1197, "step": 8063 }, { "epoch": 5.159309021113244, "grad_norm": 1.0294385336453584, "learning_rate": 9.969953812455198e-06, "loss": 0.1146, "step": 8064 }, { "epoch": 5.159948816378758, "grad_norm": 0.9520544120590931, "learning_rate": 9.967881668485566e-06, "loss": 0.1171, "step": 8065 }, { "epoch": 5.160588611644274, "grad_norm": 1.0346208846753842, "learning_rate": 9.965809525895037e-06, "loss": 0.1204, "step": 8066 }, { "epoch": 5.161228406909789, "grad_norm": 1.0674130019713017, "learning_rate": 9.963737384772582e-06, "loss": 0.1217, "step": 8067 }, { "epoch": 5.161868202175304, "grad_norm": 0.9333750908546272, "learning_rate": 9.961665245207184e-06, "loss": 0.1132, "step": 8068 }, { "epoch": 5.162507997440819, "grad_norm": 0.9604075705481595, "learning_rate": 9.959593107287812e-06, "loss": 0.1144, "step": 8069 }, { "epoch": 5.163147792706334, "grad_norm": 1.1329433791916501, "learning_rate": 9.957520971103439e-06, "loss": 0.116, "step": 8070 }, { "epoch": 5.163787587971849, "grad_norm": 0.9814222894942088, "learning_rate": 9.95544883674304e-06, "loss": 0.1055, "step": 8071 }, { "epoch": 5.164427383237364, "grad_norm": 1.007491312498022, "learning_rate": 9.95337670429559e-06, "loss": 0.1124, "step": 8072 }, { "epoch": 5.165067178502879, "grad_norm": 1.0696004470387117, "learning_rate": 9.951304573850056e-06, "loss": 0.1212, "step": 8073 }, { "epoch": 5.165706973768394, "grad_norm": 0.9162110416710741, "learning_rate": 9.949232445495422e-06, "loss": 0.1185, "step": 8074 }, { "epoch": 5.166346769033909, "grad_norm": 1.0070572665518216, "learning_rate": 9.947160319320656e-06, "loss": 0.1111, "step": 8075 }, { "epoch": 5.166986564299425, "grad_norm": 0.9480196159422993, "learning_rate": 9.945088195414734e-06, "loss": 0.109, "step": 8076 }, { "epoch": 5.167626359564939, "grad_norm": 1.0223849981193032, "learning_rate": 9.943016073866624e-06, "loss": 0.1162, "step": 8077 }, { "epoch": 5.168266154830454, "grad_norm": 1.1320171085583361, "learning_rate": 9.940943954765304e-06, "loss": 0.1207, "step": 8078 }, { "epoch": 5.1689059500959695, "grad_norm": 0.9215819494307856, "learning_rate": 9.938871838199745e-06, "loss": 0.1051, "step": 8079 }, { "epoch": 5.169545745361484, "grad_norm": 1.0126319201858074, "learning_rate": 9.936799724258917e-06, "loss": 0.1204, "step": 8080 }, { "epoch": 5.170185540626999, "grad_norm": 0.9794533288584173, "learning_rate": 9.9347276130318e-06, "loss": 0.1105, "step": 8081 }, { "epoch": 5.1708253358925145, "grad_norm": 1.2413540294186933, "learning_rate": 9.932655504607365e-06, "loss": 0.1248, "step": 8082 }, { "epoch": 5.17146513115803, "grad_norm": 0.991848885907737, "learning_rate": 9.930583399074579e-06, "loss": 0.1114, "step": 8083 }, { "epoch": 5.172104926423544, "grad_norm": 1.004344650193958, "learning_rate": 9.928511296522421e-06, "loss": 0.1149, "step": 8084 }, { "epoch": 5.172744721689059, "grad_norm": 1.0496930040660368, "learning_rate": 9.92643919703986e-06, "loss": 0.1182, "step": 8085 }, { "epoch": 5.173384516954575, "grad_norm": 1.0044747384946413, "learning_rate": 9.92436710071587e-06, "loss": 0.1119, "step": 8086 }, { "epoch": 5.17402431222009, "grad_norm": 1.0457829385518889, "learning_rate": 9.922295007639417e-06, "loss": 0.1245, "step": 8087 }, { "epoch": 5.174664107485604, "grad_norm": 0.9989551089180768, "learning_rate": 9.920222917899483e-06, "loss": 0.1205, "step": 8088 }, { "epoch": 5.17530390275112, "grad_norm": 1.1419477914404828, "learning_rate": 9.918150831585036e-06, "loss": 0.1135, "step": 8089 }, { "epoch": 5.175943698016635, "grad_norm": 0.979158218661664, "learning_rate": 9.916078748785043e-06, "loss": 0.1125, "step": 8090 }, { "epoch": 5.176583493282149, "grad_norm": 1.1452751392460816, "learning_rate": 9.914006669588481e-06, "loss": 0.1227, "step": 8091 }, { "epoch": 5.177223288547665, "grad_norm": 0.972493251414391, "learning_rate": 9.911934594084322e-06, "loss": 0.1199, "step": 8092 }, { "epoch": 5.17786308381318, "grad_norm": 1.0643088325196404, "learning_rate": 9.909862522361533e-06, "loss": 0.123, "step": 8093 }, { "epoch": 5.178502879078695, "grad_norm": 0.9522364953279797, "learning_rate": 9.907790454509083e-06, "loss": 0.1097, "step": 8094 }, { "epoch": 5.1791426743442095, "grad_norm": 1.118743791846553, "learning_rate": 9.905718390615952e-06, "loss": 0.1196, "step": 8095 }, { "epoch": 5.179782469609725, "grad_norm": 1.101445184302432, "learning_rate": 9.903646330771105e-06, "loss": 0.1228, "step": 8096 }, { "epoch": 5.18042226487524, "grad_norm": 1.0496265139219267, "learning_rate": 9.901574275063508e-06, "loss": 0.1179, "step": 8097 }, { "epoch": 5.181062060140755, "grad_norm": 1.0753489334092166, "learning_rate": 9.899502223582142e-06, "loss": 0.1129, "step": 8098 }, { "epoch": 5.18170185540627, "grad_norm": 1.0116956548494815, "learning_rate": 9.897430176415972e-06, "loss": 0.1189, "step": 8099 }, { "epoch": 5.182341650671785, "grad_norm": 1.0415501114255648, "learning_rate": 9.895358133653964e-06, "loss": 0.116, "step": 8100 }, { "epoch": 5.1829814459373, "grad_norm": 1.0442578699077611, "learning_rate": 9.893286095385096e-06, "loss": 0.1167, "step": 8101 }, { "epoch": 5.183621241202815, "grad_norm": 1.0696352551788793, "learning_rate": 9.891214061698332e-06, "loss": 0.1176, "step": 8102 }, { "epoch": 5.18426103646833, "grad_norm": 0.9590018068165038, "learning_rate": 9.88914203268264e-06, "loss": 0.1163, "step": 8103 }, { "epoch": 5.184900831733845, "grad_norm": 1.0370350028107855, "learning_rate": 9.887070008426992e-06, "loss": 0.1197, "step": 8104 }, { "epoch": 5.1855406269993605, "grad_norm": 0.9520062286434257, "learning_rate": 9.88499798902036e-06, "loss": 0.1077, "step": 8105 }, { "epoch": 5.186180422264875, "grad_norm": 0.8729679558677084, "learning_rate": 9.882925974551708e-06, "loss": 0.1087, "step": 8106 }, { "epoch": 5.18682021753039, "grad_norm": 1.0451114309767375, "learning_rate": 9.880853965110006e-06, "loss": 0.1238, "step": 8107 }, { "epoch": 5.187460012795905, "grad_norm": 1.0689020560469888, "learning_rate": 9.878781960784225e-06, "loss": 0.1224, "step": 8108 }, { "epoch": 5.188099808061421, "grad_norm": 1.022379806553774, "learning_rate": 9.87670996166333e-06, "loss": 0.1159, "step": 8109 }, { "epoch": 5.188739603326935, "grad_norm": 0.786177548240257, "learning_rate": 9.874637967836286e-06, "loss": 0.104, "step": 8110 }, { "epoch": 5.18937939859245, "grad_norm": 1.086578548136375, "learning_rate": 9.87256597939207e-06, "loss": 0.1201, "step": 8111 }, { "epoch": 5.190019193857966, "grad_norm": 0.9393026996282715, "learning_rate": 9.870493996419643e-06, "loss": 0.1089, "step": 8112 }, { "epoch": 5.190658989123481, "grad_norm": 0.9980091175371484, "learning_rate": 9.868422019007972e-06, "loss": 0.1143, "step": 8113 }, { "epoch": 5.191298784388995, "grad_norm": 1.0795742607036278, "learning_rate": 9.866350047246025e-06, "loss": 0.1164, "step": 8114 }, { "epoch": 5.1919385796545106, "grad_norm": 1.0473292394805371, "learning_rate": 9.864278081222772e-06, "loss": 0.1193, "step": 8115 }, { "epoch": 5.192578374920026, "grad_norm": 0.9878645059043081, "learning_rate": 9.862206121027176e-06, "loss": 0.1042, "step": 8116 }, { "epoch": 5.19321817018554, "grad_norm": 1.093216591482185, "learning_rate": 9.8601341667482e-06, "loss": 0.1225, "step": 8117 }, { "epoch": 5.1938579654510555, "grad_norm": 0.9571106094975135, "learning_rate": 9.858062218474818e-06, "loss": 0.1173, "step": 8118 }, { "epoch": 5.194497760716571, "grad_norm": 0.897300901090195, "learning_rate": 9.85599027629599e-06, "loss": 0.1071, "step": 8119 }, { "epoch": 5.195137555982086, "grad_norm": 1.0756176945396965, "learning_rate": 9.853918340300684e-06, "loss": 0.1145, "step": 8120 }, { "epoch": 5.1957773512476, "grad_norm": 1.0450953387885051, "learning_rate": 9.851846410577866e-06, "loss": 0.1106, "step": 8121 }, { "epoch": 5.196417146513116, "grad_norm": 1.1328947257115156, "learning_rate": 9.8497744872165e-06, "loss": 0.1189, "step": 8122 }, { "epoch": 5.197056941778631, "grad_norm": 0.9461510871414918, "learning_rate": 9.847702570305551e-06, "loss": 0.1073, "step": 8123 }, { "epoch": 5.197696737044146, "grad_norm": 0.9935700261003094, "learning_rate": 9.845630659933977e-06, "loss": 0.1146, "step": 8124 }, { "epoch": 5.198336532309661, "grad_norm": 1.011091047939895, "learning_rate": 9.843558756190753e-06, "loss": 0.1117, "step": 8125 }, { "epoch": 5.198976327575176, "grad_norm": 1.0119703861318385, "learning_rate": 9.841486859164836e-06, "loss": 0.1078, "step": 8126 }, { "epoch": 5.199616122840691, "grad_norm": 0.9641903764821065, "learning_rate": 9.839414968945187e-06, "loss": 0.1171, "step": 8127 }, { "epoch": 5.2002559181062065, "grad_norm": 1.00455668572558, "learning_rate": 9.83734308562078e-06, "loss": 0.1113, "step": 8128 }, { "epoch": 5.200895713371721, "grad_norm": 1.0063180853339178, "learning_rate": 9.83527120928057e-06, "loss": 0.1118, "step": 8129 }, { "epoch": 5.201535508637236, "grad_norm": 0.9620542711153098, "learning_rate": 9.833199340013521e-06, "loss": 0.1017, "step": 8130 }, { "epoch": 5.202175303902751, "grad_norm": 1.0395994926637484, "learning_rate": 9.831127477908595e-06, "loss": 0.1126, "step": 8131 }, { "epoch": 5.202815099168266, "grad_norm": 0.9349369645791599, "learning_rate": 9.829055623054757e-06, "loss": 0.1084, "step": 8132 }, { "epoch": 5.203454894433781, "grad_norm": 0.992130264261145, "learning_rate": 9.826983775540965e-06, "loss": 0.1219, "step": 8133 }, { "epoch": 5.204094689699296, "grad_norm": 2.106118176070999, "learning_rate": 9.824911935456178e-06, "loss": 0.1244, "step": 8134 }, { "epoch": 5.204734484964812, "grad_norm": 1.0664537935594751, "learning_rate": 9.822840102889366e-06, "loss": 0.1129, "step": 8135 }, { "epoch": 5.205374280230326, "grad_norm": 1.0078682586113503, "learning_rate": 9.820768277929484e-06, "loss": 0.1076, "step": 8136 }, { "epoch": 5.206014075495841, "grad_norm": 0.994678249273694, "learning_rate": 9.818696460665493e-06, "loss": 0.1118, "step": 8137 }, { "epoch": 5.2066538707613566, "grad_norm": 0.9930220395845393, "learning_rate": 9.816624651186352e-06, "loss": 0.119, "step": 8138 }, { "epoch": 5.207293666026872, "grad_norm": 0.986265618755429, "learning_rate": 9.814552849581027e-06, "loss": 0.1171, "step": 8139 }, { "epoch": 5.207933461292386, "grad_norm": 0.9457679150502643, "learning_rate": 9.812481055938463e-06, "loss": 0.1137, "step": 8140 }, { "epoch": 5.2085732565579015, "grad_norm": 0.9983116308371652, "learning_rate": 9.810409270347638e-06, "loss": 0.1266, "step": 8141 }, { "epoch": 5.209213051823417, "grad_norm": 0.8987445371699879, "learning_rate": 9.8083374928975e-06, "loss": 0.1084, "step": 8142 }, { "epoch": 5.209852847088931, "grad_norm": 1.1179617491543246, "learning_rate": 9.806265723677009e-06, "loss": 0.1206, "step": 8143 }, { "epoch": 5.210492642354446, "grad_norm": 1.0840821347540845, "learning_rate": 9.80419396277512e-06, "loss": 0.124, "step": 8144 }, { "epoch": 5.211132437619962, "grad_norm": 1.091591610415982, "learning_rate": 9.802122210280797e-06, "loss": 0.1199, "step": 8145 }, { "epoch": 5.211772232885477, "grad_norm": 1.0325415925163746, "learning_rate": 9.800050466282994e-06, "loss": 0.116, "step": 8146 }, { "epoch": 5.212412028150991, "grad_norm": 1.0476634818443984, "learning_rate": 9.797978730870664e-06, "loss": 0.1142, "step": 8147 }, { "epoch": 5.213051823416507, "grad_norm": 1.0479833053639265, "learning_rate": 9.795907004132773e-06, "loss": 0.1205, "step": 8148 }, { "epoch": 5.213691618682022, "grad_norm": 1.022599795694168, "learning_rate": 9.793835286158272e-06, "loss": 0.1203, "step": 8149 }, { "epoch": 5.214331413947537, "grad_norm": 1.0086948308205077, "learning_rate": 9.791763577036114e-06, "loss": 0.1144, "step": 8150 }, { "epoch": 5.214971209213052, "grad_norm": 0.9949986527487351, "learning_rate": 9.78969187685526e-06, "loss": 0.1134, "step": 8151 }, { "epoch": 5.215611004478567, "grad_norm": 1.0695360943768977, "learning_rate": 9.787620185704664e-06, "loss": 0.1172, "step": 8152 }, { "epoch": 5.216250799744082, "grad_norm": 1.0011035372240558, "learning_rate": 9.785548503673278e-06, "loss": 0.1151, "step": 8153 }, { "epoch": 5.2168905950095965, "grad_norm": 0.9514770008039474, "learning_rate": 9.783476830850054e-06, "loss": 0.113, "step": 8154 }, { "epoch": 5.217530390275112, "grad_norm": 1.0494243044568277, "learning_rate": 9.781405167323954e-06, "loss": 0.1203, "step": 8155 }, { "epoch": 5.218170185540627, "grad_norm": 1.1529399357154293, "learning_rate": 9.779333513183927e-06, "loss": 0.1198, "step": 8156 }, { "epoch": 5.218809980806142, "grad_norm": 0.9625076588905449, "learning_rate": 9.777261868518923e-06, "loss": 0.1186, "step": 8157 }, { "epoch": 5.219449776071657, "grad_norm": 1.084212444731014, "learning_rate": 9.775190233417901e-06, "loss": 0.1278, "step": 8158 }, { "epoch": 5.220089571337172, "grad_norm": 1.0438008130201382, "learning_rate": 9.773118607969811e-06, "loss": 0.1144, "step": 8159 }, { "epoch": 5.220729366602687, "grad_norm": 1.122194811271703, "learning_rate": 9.771046992263603e-06, "loss": 0.1188, "step": 8160 }, { "epoch": 5.2213691618682025, "grad_norm": 1.0079975178792815, "learning_rate": 9.768975386388232e-06, "loss": 0.1212, "step": 8161 }, { "epoch": 5.222008957133717, "grad_norm": 0.9732926674129793, "learning_rate": 9.766903790432645e-06, "loss": 0.1122, "step": 8162 }, { "epoch": 5.222648752399232, "grad_norm": 1.0756923547034551, "learning_rate": 9.764832204485798e-06, "loss": 0.1132, "step": 8163 }, { "epoch": 5.2232885476647475, "grad_norm": 0.9695140635757061, "learning_rate": 9.76276062863663e-06, "loss": 0.1168, "step": 8164 }, { "epoch": 5.223928342930263, "grad_norm": 0.9164122669287381, "learning_rate": 9.760689062974104e-06, "loss": 0.1118, "step": 8165 }, { "epoch": 5.224568138195777, "grad_norm": 1.0490023218784719, "learning_rate": 9.758617507587165e-06, "loss": 0.1211, "step": 8166 }, { "epoch": 5.225207933461292, "grad_norm": 1.0849211054900563, "learning_rate": 9.756545962564759e-06, "loss": 0.1131, "step": 8167 }, { "epoch": 5.225847728726808, "grad_norm": 0.9969164176212764, "learning_rate": 9.754474427995836e-06, "loss": 0.1121, "step": 8168 }, { "epoch": 5.226487523992322, "grad_norm": 0.9695265573364434, "learning_rate": 9.752402903969347e-06, "loss": 0.1142, "step": 8169 }, { "epoch": 5.227127319257837, "grad_norm": 0.9810597255624323, "learning_rate": 9.750331390574236e-06, "loss": 0.1102, "step": 8170 }, { "epoch": 5.227767114523353, "grad_norm": 1.0573596261233145, "learning_rate": 9.748259887899447e-06, "loss": 0.1177, "step": 8171 }, { "epoch": 5.228406909788868, "grad_norm": 0.981153485281957, "learning_rate": 9.746188396033933e-06, "loss": 0.1156, "step": 8172 }, { "epoch": 5.229046705054382, "grad_norm": 1.0066426610433863, "learning_rate": 9.74411691506664e-06, "loss": 0.1146, "step": 8173 }, { "epoch": 5.229686500319898, "grad_norm": 0.9226073160523793, "learning_rate": 9.742045445086507e-06, "loss": 0.1102, "step": 8174 }, { "epoch": 5.230326295585413, "grad_norm": 1.0310607728296242, "learning_rate": 9.739973986182488e-06, "loss": 0.1163, "step": 8175 }, { "epoch": 5.230966090850928, "grad_norm": 1.0081990744089993, "learning_rate": 9.737902538443523e-06, "loss": 0.1165, "step": 8176 }, { "epoch": 5.2316058861164425, "grad_norm": 1.0026430997509101, "learning_rate": 9.735831101958551e-06, "loss": 0.1225, "step": 8177 }, { "epoch": 5.232245681381958, "grad_norm": 1.0612648837921355, "learning_rate": 9.733759676816528e-06, "loss": 0.1197, "step": 8178 }, { "epoch": 5.232885476647473, "grad_norm": 1.0252028871237837, "learning_rate": 9.73168826310639e-06, "loss": 0.1142, "step": 8179 }, { "epoch": 5.233525271912987, "grad_norm": 0.9488406610963254, "learning_rate": 9.72961686091708e-06, "loss": 0.1194, "step": 8180 }, { "epoch": 5.234165067178503, "grad_norm": 1.1141187980126082, "learning_rate": 9.72754547033754e-06, "loss": 0.121, "step": 8181 }, { "epoch": 5.234804862444018, "grad_norm": 1.0275230111059246, "learning_rate": 9.725474091456715e-06, "loss": 0.1184, "step": 8182 }, { "epoch": 5.235444657709533, "grad_norm": 0.9472713030882522, "learning_rate": 9.723402724363543e-06, "loss": 0.1102, "step": 8183 }, { "epoch": 5.236084452975048, "grad_norm": 0.9338289845352411, "learning_rate": 9.721331369146963e-06, "loss": 0.1093, "step": 8184 }, { "epoch": 5.236724248240563, "grad_norm": 1.0757653289543145, "learning_rate": 9.71926002589592e-06, "loss": 0.1205, "step": 8185 }, { "epoch": 5.237364043506078, "grad_norm": 0.9731445727181965, "learning_rate": 9.717188694699355e-06, "loss": 0.1126, "step": 8186 }, { "epoch": 5.2380038387715935, "grad_norm": 0.9832564770604356, "learning_rate": 9.7151173756462e-06, "loss": 0.1216, "step": 8187 }, { "epoch": 5.238643634037108, "grad_norm": 1.0329614377785243, "learning_rate": 9.713046068825402e-06, "loss": 0.1172, "step": 8188 }, { "epoch": 5.239283429302623, "grad_norm": 1.0130088445484668, "learning_rate": 9.710974774325893e-06, "loss": 0.1169, "step": 8189 }, { "epoch": 5.239923224568138, "grad_norm": 1.165562019364293, "learning_rate": 9.708903492236615e-06, "loss": 0.1227, "step": 8190 }, { "epoch": 5.240563019833653, "grad_norm": 1.0913600110574535, "learning_rate": 9.706832222646502e-06, "loss": 0.1248, "step": 8191 }, { "epoch": 5.241202815099168, "grad_norm": 1.034104022037105, "learning_rate": 9.704760965644493e-06, "loss": 0.1168, "step": 8192 }, { "epoch": 5.241842610364683, "grad_norm": 0.9403061980621349, "learning_rate": 9.702689721319522e-06, "loss": 0.1135, "step": 8193 }, { "epoch": 5.242482405630199, "grad_norm": 0.9129943067938973, "learning_rate": 9.700618489760522e-06, "loss": 0.111, "step": 8194 }, { "epoch": 5.243122200895713, "grad_norm": 1.028599427788006, "learning_rate": 9.698547271056436e-06, "loss": 0.1122, "step": 8195 }, { "epoch": 5.243761996161228, "grad_norm": 1.0512768446768308, "learning_rate": 9.696476065296195e-06, "loss": 0.1166, "step": 8196 }, { "epoch": 5.244401791426744, "grad_norm": 1.0580398734572354, "learning_rate": 9.694404872568728e-06, "loss": 0.1168, "step": 8197 }, { "epoch": 5.245041586692259, "grad_norm": 1.1469852761456452, "learning_rate": 9.692333692962976e-06, "loss": 0.1148, "step": 8198 }, { "epoch": 5.245681381957773, "grad_norm": 0.9654228627650513, "learning_rate": 9.690262526567866e-06, "loss": 0.113, "step": 8199 }, { "epoch": 5.2463211772232885, "grad_norm": 1.1432419237040294, "learning_rate": 9.688191373472332e-06, "loss": 0.1222, "step": 8200 }, { "epoch": 5.246960972488804, "grad_norm": 1.0931148172531369, "learning_rate": 9.686120233765303e-06, "loss": 0.1171, "step": 8201 }, { "epoch": 5.247600767754319, "grad_norm": 0.9715411709898931, "learning_rate": 9.684049107535716e-06, "loss": 0.1228, "step": 8202 }, { "epoch": 5.248240563019833, "grad_norm": 1.1450371967269755, "learning_rate": 9.6819779948725e-06, "loss": 0.1256, "step": 8203 }, { "epoch": 5.248880358285349, "grad_norm": 1.0502672998228737, "learning_rate": 9.67990689586458e-06, "loss": 0.1126, "step": 8204 }, { "epoch": 5.249520153550864, "grad_norm": 1.0280610358058566, "learning_rate": 9.67783581060089e-06, "loss": 0.12, "step": 8205 }, { "epoch": 5.250159948816378, "grad_norm": 1.11933350779016, "learning_rate": 9.675764739170357e-06, "loss": 0.1151, "step": 8206 }, { "epoch": 5.250799744081894, "grad_norm": 1.164336765327169, "learning_rate": 9.673693681661905e-06, "loss": 0.1309, "step": 8207 }, { "epoch": 5.251439539347409, "grad_norm": 1.0242121292398678, "learning_rate": 9.671622638164471e-06, "loss": 0.1165, "step": 8208 }, { "epoch": 5.252079334612924, "grad_norm": 1.050726896813441, "learning_rate": 9.669551608766978e-06, "loss": 0.1243, "step": 8209 }, { "epoch": 5.252719129878439, "grad_norm": 1.0087359344588944, "learning_rate": 9.66748059355835e-06, "loss": 0.1041, "step": 8210 }, { "epoch": 5.253358925143954, "grad_norm": 1.084577124431386, "learning_rate": 9.66540959262751e-06, "loss": 0.1165, "step": 8211 }, { "epoch": 5.253998720409469, "grad_norm": 1.0646036001400239, "learning_rate": 9.66333860606339e-06, "loss": 0.1197, "step": 8212 }, { "epoch": 5.254638515674984, "grad_norm": 1.1005464413778927, "learning_rate": 9.661267633954913e-06, "loss": 0.1141, "step": 8213 }, { "epoch": 5.255278310940499, "grad_norm": 0.9198169378119402, "learning_rate": 9.659196676390995e-06, "loss": 0.1113, "step": 8214 }, { "epoch": 5.255918106206014, "grad_norm": 1.0446606531195375, "learning_rate": 9.65712573346057e-06, "loss": 0.1174, "step": 8215 }, { "epoch": 5.256557901471529, "grad_norm": 1.032317148308679, "learning_rate": 9.655054805252557e-06, "loss": 0.1122, "step": 8216 }, { "epoch": 5.257197696737044, "grad_norm": 0.9805224810463986, "learning_rate": 9.652983891855873e-06, "loss": 0.1164, "step": 8217 }, { "epoch": 5.257837492002559, "grad_norm": 0.9230670179885486, "learning_rate": 9.650912993359446e-06, "loss": 0.1087, "step": 8218 }, { "epoch": 5.258477287268074, "grad_norm": 0.9889862543809168, "learning_rate": 9.648842109852196e-06, "loss": 0.115, "step": 8219 }, { "epoch": 5.25911708253359, "grad_norm": 1.1168828505911927, "learning_rate": 9.646771241423041e-06, "loss": 0.1277, "step": 8220 }, { "epoch": 5.259756877799104, "grad_norm": 0.9956347746755486, "learning_rate": 9.644700388160898e-06, "loss": 0.1091, "step": 8221 }, { "epoch": 5.260396673064619, "grad_norm": 1.0603497980587664, "learning_rate": 9.64262955015469e-06, "loss": 0.1295, "step": 8222 }, { "epoch": 5.2610364683301345, "grad_norm": 1.1048199598432031, "learning_rate": 9.640558727493333e-06, "loss": 0.1183, "step": 8223 }, { "epoch": 5.26167626359565, "grad_norm": 0.9663982115474623, "learning_rate": 9.638487920265741e-06, "loss": 0.1166, "step": 8224 }, { "epoch": 5.262316058861164, "grad_norm": 0.9756937929981426, "learning_rate": 9.63641712856084e-06, "loss": 0.1117, "step": 8225 }, { "epoch": 5.262955854126679, "grad_norm": 0.987857807451542, "learning_rate": 9.634346352467539e-06, "loss": 0.1202, "step": 8226 }, { "epoch": 5.263595649392195, "grad_norm": 0.9279552957722158, "learning_rate": 9.632275592074754e-06, "loss": 0.1157, "step": 8227 }, { "epoch": 5.264235444657709, "grad_norm": 1.12961872064735, "learning_rate": 9.630204847471401e-06, "loss": 0.117, "step": 8228 }, { "epoch": 5.264875239923224, "grad_norm": 1.0242159341415766, "learning_rate": 9.628134118746397e-06, "loss": 0.114, "step": 8229 }, { "epoch": 5.26551503518874, "grad_norm": 0.9996173230983143, "learning_rate": 9.626063405988651e-06, "loss": 0.1156, "step": 8230 }, { "epoch": 5.266154830454255, "grad_norm": 0.9573718104672873, "learning_rate": 9.623992709287073e-06, "loss": 0.1188, "step": 8231 }, { "epoch": 5.266794625719769, "grad_norm": 1.0067215594620889, "learning_rate": 9.621922028730584e-06, "loss": 0.1164, "step": 8232 }, { "epoch": 5.267434420985285, "grad_norm": 0.9593064207048821, "learning_rate": 9.61985136440809e-06, "loss": 0.1106, "step": 8233 }, { "epoch": 5.2680742162508, "grad_norm": 1.0969421998123974, "learning_rate": 9.6177807164085e-06, "loss": 0.1247, "step": 8234 }, { "epoch": 5.268714011516315, "grad_norm": 1.2374375022008386, "learning_rate": 9.615710084820728e-06, "loss": 0.1187, "step": 8235 }, { "epoch": 5.2693538067818295, "grad_norm": 1.0848916755837497, "learning_rate": 9.613639469733684e-06, "loss": 0.1191, "step": 8236 }, { "epoch": 5.269993602047345, "grad_norm": 1.05154599597217, "learning_rate": 9.611568871236267e-06, "loss": 0.1215, "step": 8237 }, { "epoch": 5.27063339731286, "grad_norm": 1.0605954721660495, "learning_rate": 9.609498289417397e-06, "loss": 0.1131, "step": 8238 }, { "epoch": 5.271273192578375, "grad_norm": 0.962135570518123, "learning_rate": 9.607427724365974e-06, "loss": 0.1125, "step": 8239 }, { "epoch": 5.27191298784389, "grad_norm": 1.0635481069734494, "learning_rate": 9.605357176170907e-06, "loss": 0.1251, "step": 8240 }, { "epoch": 5.272552783109405, "grad_norm": 0.9598021308606405, "learning_rate": 9.6032866449211e-06, "loss": 0.1089, "step": 8241 }, { "epoch": 5.27319257837492, "grad_norm": 1.0049437199547244, "learning_rate": 9.60121613070546e-06, "loss": 0.1162, "step": 8242 }, { "epoch": 5.273832373640435, "grad_norm": 0.9810939813776327, "learning_rate": 9.59914563361289e-06, "loss": 0.1113, "step": 8243 }, { "epoch": 5.27447216890595, "grad_norm": 0.9452640472841429, "learning_rate": 9.597075153732289e-06, "loss": 0.1045, "step": 8244 }, { "epoch": 5.275111964171465, "grad_norm": 1.1511611930866026, "learning_rate": 9.595004691152569e-06, "loss": 0.1261, "step": 8245 }, { "epoch": 5.2757517594369805, "grad_norm": 1.1551512618912434, "learning_rate": 9.592934245962626e-06, "loss": 0.1297, "step": 8246 }, { "epoch": 5.276391554702495, "grad_norm": 1.0760051954285748, "learning_rate": 9.590863818251364e-06, "loss": 0.1142, "step": 8247 }, { "epoch": 5.27703134996801, "grad_norm": 1.0219156807165206, "learning_rate": 9.588793408107677e-06, "loss": 0.113, "step": 8248 }, { "epoch": 5.277671145233525, "grad_norm": 1.1072814941045086, "learning_rate": 9.586723015620472e-06, "loss": 0.1323, "step": 8249 }, { "epoch": 5.278310940499041, "grad_norm": 0.8966590220571578, "learning_rate": 9.584652640878649e-06, "loss": 0.1134, "step": 8250 }, { "epoch": 5.278950735764555, "grad_norm": 0.983219963260631, "learning_rate": 9.582582283971098e-06, "loss": 0.1248, "step": 8251 }, { "epoch": 5.27959053103007, "grad_norm": 1.029068274489297, "learning_rate": 9.580511944986723e-06, "loss": 0.1192, "step": 8252 }, { "epoch": 5.280230326295586, "grad_norm": 1.0415522775737334, "learning_rate": 9.57844162401442e-06, "loss": 0.1325, "step": 8253 }, { "epoch": 5.280870121561101, "grad_norm": 0.9460573664477774, "learning_rate": 9.57637132114308e-06, "loss": 0.1117, "step": 8254 }, { "epoch": 5.281509916826615, "grad_norm": 1.0074131432687052, "learning_rate": 9.574301036461606e-06, "loss": 0.1306, "step": 8255 }, { "epoch": 5.282149712092131, "grad_norm": 0.8684854514519084, "learning_rate": 9.572230770058887e-06, "loss": 0.1093, "step": 8256 }, { "epoch": 5.282789507357646, "grad_norm": 0.9125214774029012, "learning_rate": 9.57016052202382e-06, "loss": 0.1109, "step": 8257 }, { "epoch": 5.28342930262316, "grad_norm": 0.9991794638256625, "learning_rate": 9.568090292445291e-06, "loss": 0.1163, "step": 8258 }, { "epoch": 5.2840690978886755, "grad_norm": 1.107944350076276, "learning_rate": 9.5660200814122e-06, "loss": 0.123, "step": 8259 }, { "epoch": 5.284708893154191, "grad_norm": 0.9622507429078584, "learning_rate": 9.563949889013434e-06, "loss": 0.1115, "step": 8260 }, { "epoch": 5.285348688419706, "grad_norm": 1.1025387962216362, "learning_rate": 9.56187971533788e-06, "loss": 0.1061, "step": 8261 }, { "epoch": 5.2859884836852205, "grad_norm": 0.9713674051470851, "learning_rate": 9.559809560474436e-06, "loss": 0.1117, "step": 8262 }, { "epoch": 5.286628278950736, "grad_norm": 0.9593453213585327, "learning_rate": 9.557739424511986e-06, "loss": 0.11, "step": 8263 }, { "epoch": 5.287268074216251, "grad_norm": 1.0121594694905567, "learning_rate": 9.555669307539417e-06, "loss": 0.1172, "step": 8264 }, { "epoch": 5.287907869481765, "grad_norm": 1.0677938976999317, "learning_rate": 9.553599209645622e-06, "loss": 0.1165, "step": 8265 }, { "epoch": 5.288547664747281, "grad_norm": 0.9607138130625008, "learning_rate": 9.551529130919478e-06, "loss": 0.1074, "step": 8266 }, { "epoch": 5.289187460012796, "grad_norm": 1.0482181459050013, "learning_rate": 9.549459071449879e-06, "loss": 0.1208, "step": 8267 }, { "epoch": 5.289827255278311, "grad_norm": 1.0666041611735635, "learning_rate": 9.5473890313257e-06, "loss": 0.1241, "step": 8268 }, { "epoch": 5.290467050543826, "grad_norm": 0.8777751119919892, "learning_rate": 9.545319010635835e-06, "loss": 0.1059, "step": 8269 }, { "epoch": 5.291106845809341, "grad_norm": 1.0507943551204548, "learning_rate": 9.543249009469164e-06, "loss": 0.1175, "step": 8270 }, { "epoch": 5.291746641074856, "grad_norm": 1.0772109815945639, "learning_rate": 9.541179027914566e-06, "loss": 0.12, "step": 8271 }, { "epoch": 5.292386436340371, "grad_norm": 1.0603696418246629, "learning_rate": 9.539109066060926e-06, "loss": 0.1193, "step": 8272 }, { "epoch": 5.293026231605886, "grad_norm": 1.1023570706967702, "learning_rate": 9.537039123997121e-06, "loss": 0.1248, "step": 8273 }, { "epoch": 5.293666026871401, "grad_norm": 1.0137949042188894, "learning_rate": 9.534969201812032e-06, "loss": 0.1175, "step": 8274 }, { "epoch": 5.294305822136916, "grad_norm": 0.913193319170743, "learning_rate": 9.532899299594539e-06, "loss": 0.1121, "step": 8275 }, { "epoch": 5.294945617402432, "grad_norm": 0.9647100449193401, "learning_rate": 9.530829417433521e-06, "loss": 0.1111, "step": 8276 }, { "epoch": 5.295585412667946, "grad_norm": 1.1384842654467322, "learning_rate": 9.528759555417852e-06, "loss": 0.1336, "step": 8277 }, { "epoch": 5.296225207933461, "grad_norm": 0.9791676096348406, "learning_rate": 9.526689713636406e-06, "loss": 0.1142, "step": 8278 }, { "epoch": 5.296865003198977, "grad_norm": 1.0013588538551323, "learning_rate": 9.524619892178066e-06, "loss": 0.1179, "step": 8279 }, { "epoch": 5.297504798464491, "grad_norm": 0.9596201575033927, "learning_rate": 9.522550091131703e-06, "loss": 0.1135, "step": 8280 }, { "epoch": 5.298144593730006, "grad_norm": 0.9836237265830995, "learning_rate": 9.520480310586187e-06, "loss": 0.1169, "step": 8281 }, { "epoch": 5.2987843889955215, "grad_norm": 0.9980419504990328, "learning_rate": 9.518410550630396e-06, "loss": 0.112, "step": 8282 }, { "epoch": 5.299424184261037, "grad_norm": 0.9526460565810596, "learning_rate": 9.516340811353197e-06, "loss": 0.1123, "step": 8283 }, { "epoch": 5.300063979526551, "grad_norm": 0.9175136015029458, "learning_rate": 9.514271092843461e-06, "loss": 0.1149, "step": 8284 }, { "epoch": 5.3007037747920664, "grad_norm": 0.9100094657682735, "learning_rate": 9.512201395190065e-06, "loss": 0.113, "step": 8285 }, { "epoch": 5.301343570057582, "grad_norm": 0.9768313444216807, "learning_rate": 9.510131718481873e-06, "loss": 0.1162, "step": 8286 }, { "epoch": 5.301983365323097, "grad_norm": 1.1591678893141784, "learning_rate": 9.508062062807753e-06, "loss": 0.1294, "step": 8287 }, { "epoch": 5.302623160588611, "grad_norm": 0.9519147324898897, "learning_rate": 9.50599242825657e-06, "loss": 0.1148, "step": 8288 }, { "epoch": 5.303262955854127, "grad_norm": 1.0836825228478213, "learning_rate": 9.503922814917197e-06, "loss": 0.1153, "step": 8289 }, { "epoch": 5.303902751119642, "grad_norm": 1.0609868859972758, "learning_rate": 9.501853222878495e-06, "loss": 0.1192, "step": 8290 }, { "epoch": 5.304542546385157, "grad_norm": 1.1274194866710268, "learning_rate": 9.499783652229325e-06, "loss": 0.1198, "step": 8291 }, { "epoch": 5.305182341650672, "grad_norm": 1.0032435578058312, "learning_rate": 9.49771410305856e-06, "loss": 0.124, "step": 8292 }, { "epoch": 5.305822136916187, "grad_norm": 1.0267504014343831, "learning_rate": 9.495644575455058e-06, "loss": 0.1238, "step": 8293 }, { "epoch": 5.306461932181702, "grad_norm": 0.9780171055766947, "learning_rate": 9.493575069507678e-06, "loss": 0.1163, "step": 8294 }, { "epoch": 5.3071017274472165, "grad_norm": 1.0063362970734393, "learning_rate": 9.491505585305287e-06, "loss": 0.1172, "step": 8295 }, { "epoch": 5.307741522712732, "grad_norm": 0.9633475129574276, "learning_rate": 9.48943612293674e-06, "loss": 0.1125, "step": 8296 }, { "epoch": 5.308381317978247, "grad_norm": 1.0125236459657634, "learning_rate": 9.487366682490899e-06, "loss": 0.1138, "step": 8297 }, { "epoch": 5.309021113243762, "grad_norm": 1.0309733092900435, "learning_rate": 9.485297264056615e-06, "loss": 0.1115, "step": 8298 }, { "epoch": 5.309660908509277, "grad_norm": 0.9048437511835599, "learning_rate": 9.483227867722757e-06, "loss": 0.1091, "step": 8299 }, { "epoch": 5.310300703774792, "grad_norm": 0.9753359490822808, "learning_rate": 9.481158493578174e-06, "loss": 0.1108, "step": 8300 }, { "epoch": 5.310940499040307, "grad_norm": 1.0515361404451755, "learning_rate": 9.479089141711721e-06, "loss": 0.1153, "step": 8301 }, { "epoch": 5.311580294305822, "grad_norm": 0.822508098758857, "learning_rate": 9.477019812212254e-06, "loss": 0.1039, "step": 8302 }, { "epoch": 5.312220089571337, "grad_norm": 1.1133154260125873, "learning_rate": 9.474950505168628e-06, "loss": 0.1321, "step": 8303 }, { "epoch": 5.312859884836852, "grad_norm": 1.0863195220721402, "learning_rate": 9.47288122066969e-06, "loss": 0.1143, "step": 8304 }, { "epoch": 5.3134996801023675, "grad_norm": 0.9784672610286893, "learning_rate": 9.470811958804295e-06, "loss": 0.1093, "step": 8305 }, { "epoch": 5.314139475367882, "grad_norm": 1.1301697569672668, "learning_rate": 9.468742719661296e-06, "loss": 0.121, "step": 8306 }, { "epoch": 5.314779270633397, "grad_norm": 0.9892824355174176, "learning_rate": 9.466673503329539e-06, "loss": 0.1088, "step": 8307 }, { "epoch": 5.315419065898912, "grad_norm": 1.039988969038142, "learning_rate": 9.464604309897868e-06, "loss": 0.1145, "step": 8308 }, { "epoch": 5.316058861164428, "grad_norm": 1.0329124062289161, "learning_rate": 9.46253513945514e-06, "loss": 0.1149, "step": 8309 }, { "epoch": 5.316698656429942, "grad_norm": 1.0935889706357582, "learning_rate": 9.4604659920902e-06, "loss": 0.1249, "step": 8310 }, { "epoch": 5.317338451695457, "grad_norm": 1.209192962188853, "learning_rate": 9.458396867891884e-06, "loss": 0.1295, "step": 8311 }, { "epoch": 5.317978246960973, "grad_norm": 1.0760467536373681, "learning_rate": 9.456327766949047e-06, "loss": 0.1151, "step": 8312 }, { "epoch": 5.318618042226488, "grad_norm": 0.9538025053256874, "learning_rate": 9.454258689350529e-06, "loss": 0.1067, "step": 8313 }, { "epoch": 5.319257837492002, "grad_norm": 1.1560044728034884, "learning_rate": 9.452189635185168e-06, "loss": 0.1255, "step": 8314 }, { "epoch": 5.319897632757518, "grad_norm": 1.0362792704725492, "learning_rate": 9.450120604541813e-06, "loss": 0.1186, "step": 8315 }, { "epoch": 5.320537428023033, "grad_norm": 1.0143733132842068, "learning_rate": 9.4480515975093e-06, "loss": 0.1175, "step": 8316 }, { "epoch": 5.321177223288547, "grad_norm": 1.0160623110802098, "learning_rate": 9.445982614176471e-06, "loss": 0.1158, "step": 8317 }, { "epoch": 5.3218170185540625, "grad_norm": 0.8809920022084261, "learning_rate": 9.443913654632161e-06, "loss": 0.1106, "step": 8318 }, { "epoch": 5.322456813819578, "grad_norm": 1.1820785011068322, "learning_rate": 9.441844718965213e-06, "loss": 0.1297, "step": 8319 }, { "epoch": 5.323096609085093, "grad_norm": 0.9753495287230857, "learning_rate": 9.439775807264458e-06, "loss": 0.1204, "step": 8320 }, { "epoch": 5.3237364043506075, "grad_norm": 1.0584050073395002, "learning_rate": 9.43770691961873e-06, "loss": 0.123, "step": 8321 }, { "epoch": 5.324376199616123, "grad_norm": 1.1078075154135074, "learning_rate": 9.43563805611687e-06, "loss": 0.1144, "step": 8322 }, { "epoch": 5.325015994881638, "grad_norm": 1.0326602332049681, "learning_rate": 9.433569216847708e-06, "loss": 0.1147, "step": 8323 }, { "epoch": 5.325655790147153, "grad_norm": 0.9343577040081782, "learning_rate": 9.431500401900074e-06, "loss": 0.1108, "step": 8324 }, { "epoch": 5.326295585412668, "grad_norm": 0.9828170139204182, "learning_rate": 9.429431611362804e-06, "loss": 0.1128, "step": 8325 }, { "epoch": 5.326935380678183, "grad_norm": 0.9355296940176066, "learning_rate": 9.427362845324726e-06, "loss": 0.1098, "step": 8326 }, { "epoch": 5.327575175943698, "grad_norm": 1.003704823382344, "learning_rate": 9.425294103874669e-06, "loss": 0.1189, "step": 8327 }, { "epoch": 5.3282149712092135, "grad_norm": 0.8704470814279027, "learning_rate": 9.423225387101456e-06, "loss": 0.1142, "step": 8328 }, { "epoch": 5.328854766474728, "grad_norm": 1.114053415493166, "learning_rate": 9.421156695093922e-06, "loss": 0.125, "step": 8329 }, { "epoch": 5.329494561740243, "grad_norm": 0.9573034136625874, "learning_rate": 9.419088027940891e-06, "loss": 0.1013, "step": 8330 }, { "epoch": 5.330134357005758, "grad_norm": 1.043477671025481, "learning_rate": 9.417019385731186e-06, "loss": 0.1175, "step": 8331 }, { "epoch": 5.330774152271273, "grad_norm": 1.0270093999065697, "learning_rate": 9.41495076855363e-06, "loss": 0.1135, "step": 8332 }, { "epoch": 5.331413947536788, "grad_norm": 1.0219759221999254, "learning_rate": 9.41288217649705e-06, "loss": 0.1105, "step": 8333 }, { "epoch": 5.332053742802303, "grad_norm": 0.9961159455933339, "learning_rate": 9.410813609650265e-06, "loss": 0.1174, "step": 8334 }, { "epoch": 5.332693538067819, "grad_norm": 0.9896531136677029, "learning_rate": 9.408745068102091e-06, "loss": 0.116, "step": 8335 }, { "epoch": 5.333333333333333, "grad_norm": 1.0932283674712813, "learning_rate": 9.406676551941355e-06, "loss": 0.116, "step": 8336 }, { "epoch": 5.333973128598848, "grad_norm": 1.0947837124827897, "learning_rate": 9.404608061256874e-06, "loss": 0.1235, "step": 8337 }, { "epoch": 5.334612923864364, "grad_norm": 1.0138125595774032, "learning_rate": 9.402539596137458e-06, "loss": 0.1124, "step": 8338 }, { "epoch": 5.335252719129878, "grad_norm": 0.977830348970504, "learning_rate": 9.400471156671937e-06, "loss": 0.1145, "step": 8339 }, { "epoch": 5.335892514395393, "grad_norm": 0.9371858537602707, "learning_rate": 9.398402742949115e-06, "loss": 0.1062, "step": 8340 }, { "epoch": 5.3365323096609085, "grad_norm": 1.092521094176087, "learning_rate": 9.396334355057805e-06, "loss": 0.1195, "step": 8341 }, { "epoch": 5.337172104926424, "grad_norm": 0.8951639746070967, "learning_rate": 9.394265993086827e-06, "loss": 0.1156, "step": 8342 }, { "epoch": 5.337811900191938, "grad_norm": 0.9844905622821913, "learning_rate": 9.392197657124992e-06, "loss": 0.1149, "step": 8343 }, { "epoch": 5.3384516954574535, "grad_norm": 1.0111838800631419, "learning_rate": 9.390129347261107e-06, "loss": 0.1171, "step": 8344 }, { "epoch": 5.339091490722969, "grad_norm": 1.1972985545928463, "learning_rate": 9.388061063583979e-06, "loss": 0.1235, "step": 8345 }, { "epoch": 5.339731285988484, "grad_norm": 0.9687115449475836, "learning_rate": 9.385992806182425e-06, "loss": 0.1197, "step": 8346 }, { "epoch": 5.340371081253998, "grad_norm": 1.0024973308751879, "learning_rate": 9.383924575145247e-06, "loss": 0.1133, "step": 8347 }, { "epoch": 5.341010876519514, "grad_norm": 1.1589912630363362, "learning_rate": 9.38185637056125e-06, "loss": 0.1178, "step": 8348 }, { "epoch": 5.341650671785029, "grad_norm": 0.9418757084695811, "learning_rate": 9.379788192519242e-06, "loss": 0.1094, "step": 8349 }, { "epoch": 5.342290467050544, "grad_norm": 1.0104731255944752, "learning_rate": 9.377720041108027e-06, "loss": 0.121, "step": 8350 }, { "epoch": 5.342930262316059, "grad_norm": 1.081184732590898, "learning_rate": 9.375651916416401e-06, "loss": 0.1216, "step": 8351 }, { "epoch": 5.343570057581574, "grad_norm": 0.9951463617851168, "learning_rate": 9.373583818533177e-06, "loss": 0.1148, "step": 8352 }, { "epoch": 5.344209852847089, "grad_norm": 1.2410574173659008, "learning_rate": 9.371515747547148e-06, "loss": 0.1212, "step": 8353 }, { "epoch": 5.3448496481126035, "grad_norm": 0.9558500396599073, "learning_rate": 9.369447703547117e-06, "loss": 0.1177, "step": 8354 }, { "epoch": 5.345489443378119, "grad_norm": 1.1011905228134236, "learning_rate": 9.367379686621875e-06, "loss": 0.1126, "step": 8355 }, { "epoch": 5.346129238643634, "grad_norm": 0.9842460015761488, "learning_rate": 9.365311696860228e-06, "loss": 0.1136, "step": 8356 }, { "epoch": 5.346769033909149, "grad_norm": 1.1131031540403995, "learning_rate": 9.363243734350968e-06, "loss": 0.1103, "step": 8357 }, { "epoch": 5.347408829174664, "grad_norm": 0.9788210206054153, "learning_rate": 9.361175799182883e-06, "loss": 0.1203, "step": 8358 }, { "epoch": 5.348048624440179, "grad_norm": 0.94977588170817, "learning_rate": 9.35910789144478e-06, "loss": 0.1111, "step": 8359 }, { "epoch": 5.348688419705694, "grad_norm": 0.9942154587712301, "learning_rate": 9.357040011225441e-06, "loss": 0.1112, "step": 8360 }, { "epoch": 5.34932821497121, "grad_norm": 1.0550110386425673, "learning_rate": 9.35497215861366e-06, "loss": 0.1227, "step": 8361 }, { "epoch": 5.349968010236724, "grad_norm": 1.1253397074742422, "learning_rate": 9.352904333698229e-06, "loss": 0.1214, "step": 8362 }, { "epoch": 5.350607805502239, "grad_norm": 1.0160981567481757, "learning_rate": 9.350836536567934e-06, "loss": 0.1207, "step": 8363 }, { "epoch": 5.3512476007677545, "grad_norm": 0.8543341330813065, "learning_rate": 9.348768767311563e-06, "loss": 0.1047, "step": 8364 }, { "epoch": 5.35188739603327, "grad_norm": 0.9320748376268589, "learning_rate": 9.346701026017899e-06, "loss": 0.1074, "step": 8365 }, { "epoch": 5.352527191298784, "grad_norm": 1.0921973696990661, "learning_rate": 9.344633312775736e-06, "loss": 0.1239, "step": 8366 }, { "epoch": 5.3531669865642995, "grad_norm": 1.0422501699876945, "learning_rate": 9.342565627673852e-06, "loss": 0.1206, "step": 8367 }, { "epoch": 5.353806781829815, "grad_norm": 0.9755299898047562, "learning_rate": 9.340497970801029e-06, "loss": 0.1163, "step": 8368 }, { "epoch": 5.354446577095329, "grad_norm": 1.0824833380963186, "learning_rate": 9.338430342246052e-06, "loss": 0.1216, "step": 8369 }, { "epoch": 5.355086372360844, "grad_norm": 0.9679371817370218, "learning_rate": 9.336362742097698e-06, "loss": 0.1057, "step": 8370 }, { "epoch": 5.35572616762636, "grad_norm": 1.0339933425158923, "learning_rate": 9.334295170444743e-06, "loss": 0.1151, "step": 8371 }, { "epoch": 5.356365962891875, "grad_norm": 0.9210244101335768, "learning_rate": 9.332227627375973e-06, "loss": 0.1061, "step": 8372 }, { "epoch": 5.357005758157389, "grad_norm": 1.0067777969832097, "learning_rate": 9.330160112980163e-06, "loss": 0.1159, "step": 8373 }, { "epoch": 5.357645553422905, "grad_norm": 1.0419431129415957, "learning_rate": 9.328092627346085e-06, "loss": 0.1246, "step": 8374 }, { "epoch": 5.35828534868842, "grad_norm": 0.978816249371994, "learning_rate": 9.326025170562512e-06, "loss": 0.1141, "step": 8375 }, { "epoch": 5.358925143953935, "grad_norm": 1.0112622575058619, "learning_rate": 9.32395774271822e-06, "loss": 0.1119, "step": 8376 }, { "epoch": 5.3595649392194495, "grad_norm": 0.9103303430520029, "learning_rate": 9.321890343901982e-06, "loss": 0.1082, "step": 8377 }, { "epoch": 5.360204734484965, "grad_norm": 1.068983236373382, "learning_rate": 9.319822974202564e-06, "loss": 0.1148, "step": 8378 }, { "epoch": 5.36084452975048, "grad_norm": 1.056947432655907, "learning_rate": 9.31775563370874e-06, "loss": 0.1225, "step": 8379 }, { "epoch": 5.3614843250159945, "grad_norm": 1.0663290482784702, "learning_rate": 9.315688322509275e-06, "loss": 0.1149, "step": 8380 }, { "epoch": 5.36212412028151, "grad_norm": 1.0790591207072537, "learning_rate": 9.313621040692932e-06, "loss": 0.1254, "step": 8381 }, { "epoch": 5.362763915547025, "grad_norm": 1.071665565398459, "learning_rate": 9.311553788348486e-06, "loss": 0.119, "step": 8382 }, { "epoch": 5.36340371081254, "grad_norm": 0.8855585703347182, "learning_rate": 9.309486565564693e-06, "loss": 0.1112, "step": 8383 }, { "epoch": 5.364043506078055, "grad_norm": 1.088098948326595, "learning_rate": 9.307419372430321e-06, "loss": 0.1148, "step": 8384 }, { "epoch": 5.36468330134357, "grad_norm": 0.9658668163417393, "learning_rate": 9.305352209034126e-06, "loss": 0.1101, "step": 8385 }, { "epoch": 5.365323096609085, "grad_norm": 0.98964092188669, "learning_rate": 9.303285075464874e-06, "loss": 0.1144, "step": 8386 }, { "epoch": 5.3659628918746005, "grad_norm": 1.069829083421111, "learning_rate": 9.301217971811322e-06, "loss": 0.1127, "step": 8387 }, { "epoch": 5.366602687140115, "grad_norm": 0.9439528035445234, "learning_rate": 9.299150898162221e-06, "loss": 0.1146, "step": 8388 }, { "epoch": 5.36724248240563, "grad_norm": 0.9354807315049204, "learning_rate": 9.29708385460634e-06, "loss": 0.1084, "step": 8389 }, { "epoch": 5.3678822776711455, "grad_norm": 1.0134003903143065, "learning_rate": 9.295016841232427e-06, "loss": 0.1111, "step": 8390 }, { "epoch": 5.36852207293666, "grad_norm": 0.9278183668483752, "learning_rate": 9.292949858129234e-06, "loss": 0.1135, "step": 8391 }, { "epoch": 5.369161868202175, "grad_norm": 0.8440156944573557, "learning_rate": 9.29088290538552e-06, "loss": 0.1082, "step": 8392 }, { "epoch": 5.36980166346769, "grad_norm": 1.0373019287493317, "learning_rate": 9.288815983090031e-06, "loss": 0.1211, "step": 8393 }, { "epoch": 5.370441458733206, "grad_norm": 1.0423179507102822, "learning_rate": 9.28674909133152e-06, "loss": 0.1193, "step": 8394 }, { "epoch": 5.37108125399872, "grad_norm": 1.099525562685934, "learning_rate": 9.284682230198729e-06, "loss": 0.1279, "step": 8395 }, { "epoch": 5.371721049264235, "grad_norm": 0.9424771940362271, "learning_rate": 9.282615399780414e-06, "loss": 0.1072, "step": 8396 }, { "epoch": 5.372360844529751, "grad_norm": 1.037020783386695, "learning_rate": 9.28054860016532e-06, "loss": 0.1167, "step": 8397 }, { "epoch": 5.373000639795266, "grad_norm": 1.0192800808725777, "learning_rate": 9.278481831442187e-06, "loss": 0.1082, "step": 8398 }, { "epoch": 5.37364043506078, "grad_norm": 0.9291479304475059, "learning_rate": 9.276415093699761e-06, "loss": 0.1081, "step": 8399 }, { "epoch": 5.3742802303262955, "grad_norm": 0.9958768497633842, "learning_rate": 9.274348387026785e-06, "loss": 0.1161, "step": 8400 }, { "epoch": 5.374920025591811, "grad_norm": 1.004010260133602, "learning_rate": 9.272281711511993e-06, "loss": 0.1183, "step": 8401 }, { "epoch": 5.375559820857326, "grad_norm": 1.1075869250601396, "learning_rate": 9.270215067244136e-06, "loss": 0.1213, "step": 8402 }, { "epoch": 5.3761996161228405, "grad_norm": 0.9295451499052019, "learning_rate": 9.268148454311944e-06, "loss": 0.1159, "step": 8403 }, { "epoch": 5.376839411388356, "grad_norm": 0.9902166121144572, "learning_rate": 9.266081872804156e-06, "loss": 0.1136, "step": 8404 }, { "epoch": 5.377479206653871, "grad_norm": 1.0548985521049856, "learning_rate": 9.264015322809502e-06, "loss": 0.1243, "step": 8405 }, { "epoch": 5.378119001919385, "grad_norm": 0.988027675345641, "learning_rate": 9.261948804416726e-06, "loss": 0.1125, "step": 8406 }, { "epoch": 5.378758797184901, "grad_norm": 1.070283827173667, "learning_rate": 9.259882317714555e-06, "loss": 0.1234, "step": 8407 }, { "epoch": 5.379398592450416, "grad_norm": 1.0239493339836545, "learning_rate": 9.25781586279172e-06, "loss": 0.1109, "step": 8408 }, { "epoch": 5.380038387715931, "grad_norm": 1.0223420423291825, "learning_rate": 9.255749439736952e-06, "loss": 0.116, "step": 8409 }, { "epoch": 5.380678182981446, "grad_norm": 0.9761044569189695, "learning_rate": 9.253683048638979e-06, "loss": 0.1135, "step": 8410 }, { "epoch": 5.381317978246961, "grad_norm": 1.0710520441011637, "learning_rate": 9.25161668958653e-06, "loss": 0.1232, "step": 8411 }, { "epoch": 5.381957773512476, "grad_norm": 0.9560906341134412, "learning_rate": 9.249550362668322e-06, "loss": 0.116, "step": 8412 }, { "epoch": 5.3825975687779914, "grad_norm": 1.1425611858244331, "learning_rate": 9.247484067973093e-06, "loss": 0.122, "step": 8413 }, { "epoch": 5.383237364043506, "grad_norm": 1.0460660883690165, "learning_rate": 9.24541780558956e-06, "loss": 0.1199, "step": 8414 }, { "epoch": 5.383877159309021, "grad_norm": 1.051514877112949, "learning_rate": 9.243351575606441e-06, "loss": 0.1191, "step": 8415 }, { "epoch": 5.384516954574536, "grad_norm": 0.8963135693732424, "learning_rate": 9.241285378112461e-06, "loss": 0.1071, "step": 8416 }, { "epoch": 5.385156749840051, "grad_norm": 0.9898273720785984, "learning_rate": 9.239219213196338e-06, "loss": 0.1097, "step": 8417 }, { "epoch": 5.385796545105566, "grad_norm": 1.1853292927620052, "learning_rate": 9.237153080946783e-06, "loss": 0.1147, "step": 8418 }, { "epoch": 5.386436340371081, "grad_norm": 1.0249917039660765, "learning_rate": 9.235086981452523e-06, "loss": 0.1228, "step": 8419 }, { "epoch": 5.387076135636597, "grad_norm": 1.0727460786244765, "learning_rate": 9.233020914802269e-06, "loss": 0.123, "step": 8420 }, { "epoch": 5.387715930902111, "grad_norm": 1.1036916030767456, "learning_rate": 9.23095488108473e-06, "loss": 0.1127, "step": 8421 }, { "epoch": 5.388355726167626, "grad_norm": 1.1014551978387659, "learning_rate": 9.22888888038862e-06, "loss": 0.1157, "step": 8422 }, { "epoch": 5.3889955214331415, "grad_norm": 0.9718242597278681, "learning_rate": 9.226822912802651e-06, "loss": 0.1094, "step": 8423 }, { "epoch": 5.389635316698657, "grad_norm": 1.0899994929754055, "learning_rate": 9.22475697841553e-06, "loss": 0.1209, "step": 8424 }, { "epoch": 5.390275111964171, "grad_norm": 1.0037141835422703, "learning_rate": 9.222691077315962e-06, "loss": 0.1153, "step": 8425 }, { "epoch": 5.3909149072296865, "grad_norm": 1.0802501892370044, "learning_rate": 9.220625209592661e-06, "loss": 0.1185, "step": 8426 }, { "epoch": 5.391554702495202, "grad_norm": 1.0265513096408543, "learning_rate": 9.218559375334327e-06, "loss": 0.1141, "step": 8427 }, { "epoch": 5.392194497760716, "grad_norm": 0.9669289174517871, "learning_rate": 9.21649357462966e-06, "loss": 0.0996, "step": 8428 }, { "epoch": 5.392834293026231, "grad_norm": 1.0633700742611727, "learning_rate": 9.214427807567368e-06, "loss": 0.1125, "step": 8429 }, { "epoch": 5.393474088291747, "grad_norm": 1.0694163627653126, "learning_rate": 9.212362074236149e-06, "loss": 0.1136, "step": 8430 }, { "epoch": 5.394113883557262, "grad_norm": 1.0298208065957264, "learning_rate": 9.210296374724702e-06, "loss": 0.1194, "step": 8431 }, { "epoch": 5.394753678822776, "grad_norm": 1.0173377668780434, "learning_rate": 9.208230709121717e-06, "loss": 0.1178, "step": 8432 }, { "epoch": 5.395393474088292, "grad_norm": 0.9477367426532511, "learning_rate": 9.206165077515904e-06, "loss": 0.1196, "step": 8433 }, { "epoch": 5.396033269353807, "grad_norm": 1.1098541038084344, "learning_rate": 9.20409947999595e-06, "loss": 0.1267, "step": 8434 }, { "epoch": 5.396673064619322, "grad_norm": 0.9900553500699233, "learning_rate": 9.202033916650544e-06, "loss": 0.1124, "step": 8435 }, { "epoch": 5.3973128598848366, "grad_norm": 0.9184795011726561, "learning_rate": 9.199968387568386e-06, "loss": 0.1085, "step": 8436 }, { "epoch": 5.397952655150352, "grad_norm": 0.9824165101740195, "learning_rate": 9.197902892838162e-06, "loss": 0.1124, "step": 8437 }, { "epoch": 5.398592450415867, "grad_norm": 1.1878883996688605, "learning_rate": 9.19583743254856e-06, "loss": 0.1254, "step": 8438 }, { "epoch": 5.399232245681382, "grad_norm": 1.0459607525679158, "learning_rate": 9.193772006788268e-06, "loss": 0.1193, "step": 8439 }, { "epoch": 5.399872040946897, "grad_norm": 0.9846027880264459, "learning_rate": 9.191706615645974e-06, "loss": 0.1101, "step": 8440 }, { "epoch": 5.400511836212412, "grad_norm": 1.0939279166559932, "learning_rate": 9.18964125921036e-06, "loss": 0.1227, "step": 8441 }, { "epoch": 5.401151631477927, "grad_norm": 0.8940144527560175, "learning_rate": 9.187575937570103e-06, "loss": 0.1048, "step": 8442 }, { "epoch": 5.401791426743442, "grad_norm": 0.9365699013863918, "learning_rate": 9.185510650813895e-06, "loss": 0.1127, "step": 8443 }, { "epoch": 5.402431222008957, "grad_norm": 1.0609675811396366, "learning_rate": 9.183445399030412e-06, "loss": 0.1173, "step": 8444 }, { "epoch": 5.403071017274472, "grad_norm": 1.083252311482882, "learning_rate": 9.181380182308328e-06, "loss": 0.1193, "step": 8445 }, { "epoch": 5.4037108125399875, "grad_norm": 1.0057358204774645, "learning_rate": 9.179315000736325e-06, "loss": 0.1145, "step": 8446 }, { "epoch": 5.404350607805502, "grad_norm": 1.072200832906178, "learning_rate": 9.177249854403076e-06, "loss": 0.1197, "step": 8447 }, { "epoch": 5.404990403071017, "grad_norm": 0.9588694606995636, "learning_rate": 9.17518474339725e-06, "loss": 0.125, "step": 8448 }, { "epoch": 5.4056301983365325, "grad_norm": 0.9491233380547487, "learning_rate": 9.173119667807529e-06, "loss": 0.1087, "step": 8449 }, { "epoch": 5.406269993602048, "grad_norm": 1.0887655986185358, "learning_rate": 9.171054627722578e-06, "loss": 0.115, "step": 8450 }, { "epoch": 5.406909788867562, "grad_norm": 1.048746381705736, "learning_rate": 9.168989623231065e-06, "loss": 0.1211, "step": 8451 }, { "epoch": 5.407549584133077, "grad_norm": 0.9612093934921316, "learning_rate": 9.16692465442166e-06, "loss": 0.1062, "step": 8452 }, { "epoch": 5.408189379398593, "grad_norm": 0.90276632864731, "learning_rate": 9.16485972138303e-06, "loss": 0.1106, "step": 8453 }, { "epoch": 5.408829174664108, "grad_norm": 1.0328910652262757, "learning_rate": 9.162794824203837e-06, "loss": 0.1172, "step": 8454 }, { "epoch": 5.409468969929622, "grad_norm": 0.9725902301601868, "learning_rate": 9.160729962972741e-06, "loss": 0.1154, "step": 8455 }, { "epoch": 5.410108765195138, "grad_norm": 0.9365463045320729, "learning_rate": 9.158665137778412e-06, "loss": 0.1128, "step": 8456 }, { "epoch": 5.410748560460653, "grad_norm": 1.0764735592020231, "learning_rate": 9.156600348709505e-06, "loss": 0.1274, "step": 8457 }, { "epoch": 5.411388355726167, "grad_norm": 0.935730662666566, "learning_rate": 9.154535595854677e-06, "loss": 0.108, "step": 8458 }, { "epoch": 5.4120281509916826, "grad_norm": 1.1090552160653881, "learning_rate": 9.152470879302589e-06, "loss": 0.1193, "step": 8459 }, { "epoch": 5.412667946257198, "grad_norm": 0.9640181241230782, "learning_rate": 9.150406199141892e-06, "loss": 0.1122, "step": 8460 }, { "epoch": 5.413307741522713, "grad_norm": 1.1060311779158856, "learning_rate": 9.148341555461243e-06, "loss": 0.1208, "step": 8461 }, { "epoch": 5.4139475367882275, "grad_norm": 1.0017871159614178, "learning_rate": 9.146276948349289e-06, "loss": 0.1099, "step": 8462 }, { "epoch": 5.414587332053743, "grad_norm": 1.075225615236638, "learning_rate": 9.144212377894689e-06, "loss": 0.1204, "step": 8463 }, { "epoch": 5.415227127319258, "grad_norm": 0.9780742815353353, "learning_rate": 9.142147844186085e-06, "loss": 0.111, "step": 8464 }, { "epoch": 5.415866922584772, "grad_norm": 1.1140484830094979, "learning_rate": 9.140083347312122e-06, "loss": 0.1214, "step": 8465 }, { "epoch": 5.416506717850288, "grad_norm": 0.9751125147178078, "learning_rate": 9.138018887361457e-06, "loss": 0.1198, "step": 8466 }, { "epoch": 5.417146513115803, "grad_norm": 1.0538212454558968, "learning_rate": 9.135954464422725e-06, "loss": 0.1155, "step": 8467 }, { "epoch": 5.417786308381318, "grad_norm": 0.9605060714474122, "learning_rate": 9.133890078584571e-06, "loss": 0.1107, "step": 8468 }, { "epoch": 5.418426103646833, "grad_norm": 1.0062480290798217, "learning_rate": 9.131825729935638e-06, "loss": 0.1139, "step": 8469 }, { "epoch": 5.419065898912348, "grad_norm": 0.955313004391371, "learning_rate": 9.129761418564563e-06, "loss": 0.1148, "step": 8470 }, { "epoch": 5.419705694177863, "grad_norm": 0.9731900442025234, "learning_rate": 9.127697144559987e-06, "loss": 0.1199, "step": 8471 }, { "epoch": 5.4203454894433785, "grad_norm": 1.0067418289670864, "learning_rate": 9.125632908010536e-06, "loss": 0.1154, "step": 8472 }, { "epoch": 5.420985284708893, "grad_norm": 0.9549800131538363, "learning_rate": 9.12356870900486e-06, "loss": 0.1231, "step": 8473 }, { "epoch": 5.421625079974408, "grad_norm": 1.0142583910151795, "learning_rate": 9.121504547631583e-06, "loss": 0.1227, "step": 8474 }, { "epoch": 5.422264875239923, "grad_norm": 0.9843436702416423, "learning_rate": 9.119440423979337e-06, "loss": 0.1185, "step": 8475 }, { "epoch": 5.422904670505439, "grad_norm": 1.0084931173888254, "learning_rate": 9.117376338136755e-06, "loss": 0.1195, "step": 8476 }, { "epoch": 5.423544465770953, "grad_norm": 0.9489217886783816, "learning_rate": 9.115312290192461e-06, "loss": 0.1109, "step": 8477 }, { "epoch": 5.424184261036468, "grad_norm": 1.0449763914239294, "learning_rate": 9.113248280235081e-06, "loss": 0.1213, "step": 8478 }, { "epoch": 5.424824056301984, "grad_norm": 1.0233997778820607, "learning_rate": 9.111184308353245e-06, "loss": 0.1177, "step": 8479 }, { "epoch": 5.425463851567498, "grad_norm": 1.0781312830789453, "learning_rate": 9.109120374635576e-06, "loss": 0.1228, "step": 8480 }, { "epoch": 5.426103646833013, "grad_norm": 1.0072882022140812, "learning_rate": 9.107056479170691e-06, "loss": 0.1177, "step": 8481 }, { "epoch": 5.4267434420985285, "grad_norm": 0.9403323027856426, "learning_rate": 9.104992622047213e-06, "loss": 0.1132, "step": 8482 }, { "epoch": 5.427383237364044, "grad_norm": 0.9217744182718679, "learning_rate": 9.10292880335376e-06, "loss": 0.1063, "step": 8483 }, { "epoch": 5.428023032629558, "grad_norm": 0.882449481488288, "learning_rate": 9.100865023178949e-06, "loss": 0.1038, "step": 8484 }, { "epoch": 5.4286628278950735, "grad_norm": 0.9876311090714355, "learning_rate": 9.098801281611388e-06, "loss": 0.1145, "step": 8485 }, { "epoch": 5.429302623160589, "grad_norm": 0.9457656203545891, "learning_rate": 9.096737578739705e-06, "loss": 0.1076, "step": 8486 }, { "epoch": 5.429942418426104, "grad_norm": 1.0453170404092589, "learning_rate": 9.094673914652501e-06, "loss": 0.122, "step": 8487 }, { "epoch": 5.430582213691618, "grad_norm": 0.9882232880144729, "learning_rate": 9.092610289438389e-06, "loss": 0.1117, "step": 8488 }, { "epoch": 5.431222008957134, "grad_norm": 1.0530271215740095, "learning_rate": 9.090546703185976e-06, "loss": 0.1232, "step": 8489 }, { "epoch": 5.431861804222649, "grad_norm": 1.119351350421275, "learning_rate": 9.088483155983871e-06, "loss": 0.1169, "step": 8490 }, { "epoch": 5.432501599488164, "grad_norm": 0.941883110737294, "learning_rate": 9.086419647920677e-06, "loss": 0.1134, "step": 8491 }, { "epoch": 5.433141394753679, "grad_norm": 1.0278022268727542, "learning_rate": 9.084356179084995e-06, "loss": 0.1108, "step": 8492 }, { "epoch": 5.433781190019194, "grad_norm": 0.9674154764847943, "learning_rate": 9.082292749565436e-06, "loss": 0.1154, "step": 8493 }, { "epoch": 5.434420985284709, "grad_norm": 1.0056306618518076, "learning_rate": 9.08022935945059e-06, "loss": 0.1096, "step": 8494 }, { "epoch": 5.435060780550224, "grad_norm": 0.9638108488605301, "learning_rate": 9.078166008829057e-06, "loss": 0.1187, "step": 8495 }, { "epoch": 5.435700575815739, "grad_norm": 1.0450221059885783, "learning_rate": 9.07610269778944e-06, "loss": 0.1177, "step": 8496 }, { "epoch": 5.436340371081254, "grad_norm": 1.022232104994417, "learning_rate": 9.07403942642033e-06, "loss": 0.1091, "step": 8497 }, { "epoch": 5.436980166346769, "grad_norm": 1.0434440264177391, "learning_rate": 9.071976194810319e-06, "loss": 0.1207, "step": 8498 }, { "epoch": 5.437619961612284, "grad_norm": 1.030905977084545, "learning_rate": 9.069913003047997e-06, "loss": 0.11, "step": 8499 }, { "epoch": 5.438259756877799, "grad_norm": 0.9521897206066946, "learning_rate": 9.06784985122196e-06, "loss": 0.1226, "step": 8500 }, { "epoch": 5.438899552143314, "grad_norm": 0.9532096732628638, "learning_rate": 9.065786739420792e-06, "loss": 0.11, "step": 8501 }, { "epoch": 5.439539347408829, "grad_norm": 1.0535697770722943, "learning_rate": 9.063723667733073e-06, "loss": 0.1224, "step": 8502 }, { "epoch": 5.440179142674344, "grad_norm": 0.9921014758499255, "learning_rate": 9.0616606362474e-06, "loss": 0.1273, "step": 8503 }, { "epoch": 5.440818937939859, "grad_norm": 0.9507444909200371, "learning_rate": 9.059597645052351e-06, "loss": 0.1122, "step": 8504 }, { "epoch": 5.4414587332053745, "grad_norm": 1.1193135318033742, "learning_rate": 9.057534694236503e-06, "loss": 0.124, "step": 8505 }, { "epoch": 5.442098528470889, "grad_norm": 0.9536758641296466, "learning_rate": 9.05547178388844e-06, "loss": 0.1219, "step": 8506 }, { "epoch": 5.442738323736404, "grad_norm": 1.0492321198061163, "learning_rate": 9.053408914096739e-06, "loss": 0.1182, "step": 8507 }, { "epoch": 5.4433781190019195, "grad_norm": 1.0674677844686085, "learning_rate": 9.051346084949976e-06, "loss": 0.1199, "step": 8508 }, { "epoch": 5.444017914267435, "grad_norm": 1.0273220733328632, "learning_rate": 9.049283296536721e-06, "loss": 0.1192, "step": 8509 }, { "epoch": 5.444657709532949, "grad_norm": 1.0877970799505454, "learning_rate": 9.047220548945553e-06, "loss": 0.1232, "step": 8510 }, { "epoch": 5.445297504798464, "grad_norm": 0.9973031482519544, "learning_rate": 9.04515784226504e-06, "loss": 0.1169, "step": 8511 }, { "epoch": 5.44593730006398, "grad_norm": 1.0238144373658462, "learning_rate": 9.04309517658375e-06, "loss": 0.1159, "step": 8512 }, { "epoch": 5.446577095329495, "grad_norm": 1.1555621231303532, "learning_rate": 9.041032551990251e-06, "loss": 0.1175, "step": 8513 }, { "epoch": 5.447216890595009, "grad_norm": 0.9822984632798758, "learning_rate": 9.03896996857311e-06, "loss": 0.1139, "step": 8514 }, { "epoch": 5.447856685860525, "grad_norm": 0.981540347290958, "learning_rate": 9.036907426420882e-06, "loss": 0.1188, "step": 8515 }, { "epoch": 5.44849648112604, "grad_norm": 1.161778319099778, "learning_rate": 9.034844925622142e-06, "loss": 0.1265, "step": 8516 }, { "epoch": 5.449136276391554, "grad_norm": 0.9116931708028941, "learning_rate": 9.032782466265446e-06, "loss": 0.1189, "step": 8517 }, { "epoch": 5.44977607165707, "grad_norm": 1.0063077439276213, "learning_rate": 9.030720048439347e-06, "loss": 0.1141, "step": 8518 }, { "epoch": 5.450415866922585, "grad_norm": 0.9560156487777571, "learning_rate": 9.028657672232404e-06, "loss": 0.1154, "step": 8519 }, { "epoch": 5.4510556621881, "grad_norm": 0.9677079931520693, "learning_rate": 9.026595337733175e-06, "loss": 0.1176, "step": 8520 }, { "epoch": 5.4516954574536145, "grad_norm": 1.0491787292246832, "learning_rate": 9.024533045030211e-06, "loss": 0.1201, "step": 8521 }, { "epoch": 5.45233525271913, "grad_norm": 1.0281684562180289, "learning_rate": 9.022470794212057e-06, "loss": 0.1198, "step": 8522 }, { "epoch": 5.452975047984645, "grad_norm": 1.0023118989928985, "learning_rate": 9.020408585367275e-06, "loss": 0.1137, "step": 8523 }, { "epoch": 5.45361484325016, "grad_norm": 1.061738718847853, "learning_rate": 9.018346418584405e-06, "loss": 0.1188, "step": 8524 }, { "epoch": 5.454254638515675, "grad_norm": 1.1960367908290122, "learning_rate": 9.016284293951988e-06, "loss": 0.1231, "step": 8525 }, { "epoch": 5.45489443378119, "grad_norm": 1.0177811125003418, "learning_rate": 9.014222211558579e-06, "loss": 0.1121, "step": 8526 }, { "epoch": 5.455534229046705, "grad_norm": 1.0224165873397666, "learning_rate": 9.012160171492713e-06, "loss": 0.1143, "step": 8527 }, { "epoch": 5.4561740243122205, "grad_norm": 1.0203216616468258, "learning_rate": 9.010098173842935e-06, "loss": 0.1123, "step": 8528 }, { "epoch": 5.456813819577735, "grad_norm": 1.047336822655913, "learning_rate": 9.008036218697775e-06, "loss": 0.124, "step": 8529 }, { "epoch": 5.45745361484325, "grad_norm": 0.9474295630542742, "learning_rate": 9.00597430614578e-06, "loss": 0.112, "step": 8530 }, { "epoch": 5.4580934101087655, "grad_norm": 1.0761744409511833, "learning_rate": 9.00391243627548e-06, "loss": 0.1171, "step": 8531 }, { "epoch": 5.45873320537428, "grad_norm": 0.9850996282678823, "learning_rate": 9.001850609175402e-06, "loss": 0.1186, "step": 8532 }, { "epoch": 5.459373000639795, "grad_norm": 1.1101356647585645, "learning_rate": 8.999788824934088e-06, "loss": 0.1187, "step": 8533 }, { "epoch": 5.46001279590531, "grad_norm": 0.9143052226309941, "learning_rate": 8.997727083640063e-06, "loss": 0.1047, "step": 8534 }, { "epoch": 5.460652591170826, "grad_norm": 0.983695456380568, "learning_rate": 8.995665385381853e-06, "loss": 0.115, "step": 8535 }, { "epoch": 5.46129238643634, "grad_norm": 1.079552502889986, "learning_rate": 8.993603730247985e-06, "loss": 0.122, "step": 8536 }, { "epoch": 5.461932181701855, "grad_norm": 1.0760294381051143, "learning_rate": 8.991542118326984e-06, "loss": 0.12, "step": 8537 }, { "epoch": 5.462571976967371, "grad_norm": 1.0953203836499064, "learning_rate": 8.98948054970737e-06, "loss": 0.1158, "step": 8538 }, { "epoch": 5.463211772232885, "grad_norm": 1.006388516061164, "learning_rate": 8.98741902447766e-06, "loss": 0.1188, "step": 8539 }, { "epoch": 5.4638515674984, "grad_norm": 1.034049755375269, "learning_rate": 8.985357542726379e-06, "loss": 0.121, "step": 8540 }, { "epoch": 5.464491362763916, "grad_norm": 1.1036010780521337, "learning_rate": 8.98329610454204e-06, "loss": 0.1218, "step": 8541 }, { "epoch": 5.465131158029431, "grad_norm": 1.0231480907482928, "learning_rate": 8.981234710013157e-06, "loss": 0.1066, "step": 8542 }, { "epoch": 5.465770953294945, "grad_norm": 0.9860129081218946, "learning_rate": 8.979173359228243e-06, "loss": 0.1172, "step": 8543 }, { "epoch": 5.4664107485604605, "grad_norm": 1.06012321448041, "learning_rate": 8.977112052275811e-06, "loss": 0.1162, "step": 8544 }, { "epoch": 5.467050543825976, "grad_norm": 0.9380252573353497, "learning_rate": 8.975050789244365e-06, "loss": 0.1085, "step": 8545 }, { "epoch": 5.467690339091491, "grad_norm": 1.057570906934088, "learning_rate": 8.972989570222416e-06, "loss": 0.1209, "step": 8546 }, { "epoch": 5.468330134357005, "grad_norm": 1.0462636607958409, "learning_rate": 8.970928395298471e-06, "loss": 0.1164, "step": 8547 }, { "epoch": 5.468969929622521, "grad_norm": 1.0954448933043723, "learning_rate": 8.968867264561027e-06, "loss": 0.123, "step": 8548 }, { "epoch": 5.469609724888036, "grad_norm": 0.9528983696915224, "learning_rate": 8.966806178098587e-06, "loss": 0.1158, "step": 8549 }, { "epoch": 5.470249520153551, "grad_norm": 0.9995792775453612, "learning_rate": 8.964745135999656e-06, "loss": 0.1107, "step": 8550 }, { "epoch": 5.470889315419066, "grad_norm": 1.0693825847131697, "learning_rate": 8.962684138352726e-06, "loss": 0.1226, "step": 8551 }, { "epoch": 5.471529110684581, "grad_norm": 1.0549874115741078, "learning_rate": 8.96062318524629e-06, "loss": 0.1168, "step": 8552 }, { "epoch": 5.472168905950096, "grad_norm": 1.172144492458643, "learning_rate": 8.95856227676885e-06, "loss": 0.1257, "step": 8553 }, { "epoch": 5.472808701215611, "grad_norm": 1.0084530057590486, "learning_rate": 8.956501413008892e-06, "loss": 0.1144, "step": 8554 }, { "epoch": 5.473448496481126, "grad_norm": 0.9679995854999537, "learning_rate": 8.954440594054904e-06, "loss": 0.112, "step": 8555 }, { "epoch": 5.474088291746641, "grad_norm": 1.0590637904128175, "learning_rate": 8.952379819995381e-06, "loss": 0.1204, "step": 8556 }, { "epoch": 5.474728087012156, "grad_norm": 1.0542628934679714, "learning_rate": 8.950319090918804e-06, "loss": 0.1116, "step": 8557 }, { "epoch": 5.475367882277671, "grad_norm": 0.9896715057445892, "learning_rate": 8.94825840691366e-06, "loss": 0.1218, "step": 8558 }, { "epoch": 5.476007677543186, "grad_norm": 0.9166793940565697, "learning_rate": 8.946197768068424e-06, "loss": 0.114, "step": 8559 }, { "epoch": 5.476647472808701, "grad_norm": 0.9883322896318861, "learning_rate": 8.944137174471585e-06, "loss": 0.1173, "step": 8560 }, { "epoch": 5.477287268074217, "grad_norm": 1.0237174942301064, "learning_rate": 8.942076626211617e-06, "loss": 0.1173, "step": 8561 }, { "epoch": 5.477927063339731, "grad_norm": 1.0763909350799643, "learning_rate": 8.940016123376994e-06, "loss": 0.1203, "step": 8562 }, { "epoch": 5.478566858605246, "grad_norm": 0.9015788530999974, "learning_rate": 8.937955666056198e-06, "loss": 0.116, "step": 8563 }, { "epoch": 5.479206653870762, "grad_norm": 0.9997756957180803, "learning_rate": 8.935895254337694e-06, "loss": 0.1177, "step": 8564 }, { "epoch": 5.479846449136277, "grad_norm": 1.0305512571500712, "learning_rate": 8.933834888309956e-06, "loss": 0.1211, "step": 8565 }, { "epoch": 5.480486244401791, "grad_norm": 1.065003847703054, "learning_rate": 8.931774568061449e-06, "loss": 0.1283, "step": 8566 }, { "epoch": 5.4811260396673065, "grad_norm": 1.0487591489599852, "learning_rate": 8.929714293680643e-06, "loss": 0.1189, "step": 8567 }, { "epoch": 5.481765834932822, "grad_norm": 1.063890270805203, "learning_rate": 8.927654065256003e-06, "loss": 0.1111, "step": 8568 }, { "epoch": 5.482405630198336, "grad_norm": 1.0026131472505484, "learning_rate": 8.925593882875984e-06, "loss": 0.1203, "step": 8569 }, { "epoch": 5.483045425463851, "grad_norm": 0.9992788611829413, "learning_rate": 8.923533746629059e-06, "loss": 0.1197, "step": 8570 }, { "epoch": 5.483685220729367, "grad_norm": 0.939947401991882, "learning_rate": 8.921473656603678e-06, "loss": 0.1143, "step": 8571 }, { "epoch": 5.484325015994882, "grad_norm": 1.007550101581741, "learning_rate": 8.919413612888297e-06, "loss": 0.1186, "step": 8572 }, { "epoch": 5.484964811260396, "grad_norm": 0.9717509144104854, "learning_rate": 8.917353615571377e-06, "loss": 0.1125, "step": 8573 }, { "epoch": 5.485604606525912, "grad_norm": 1.0361877059776237, "learning_rate": 8.915293664741367e-06, "loss": 0.1137, "step": 8574 }, { "epoch": 5.486244401791427, "grad_norm": 1.0447383047448509, "learning_rate": 8.913233760486717e-06, "loss": 0.1126, "step": 8575 }, { "epoch": 5.486884197056942, "grad_norm": 0.9794439253506724, "learning_rate": 8.911173902895873e-06, "loss": 0.1172, "step": 8576 }, { "epoch": 5.487523992322457, "grad_norm": 1.0288397397063234, "learning_rate": 8.909114092057286e-06, "loss": 0.1125, "step": 8577 }, { "epoch": 5.488163787587972, "grad_norm": 1.007332478278753, "learning_rate": 8.907054328059402e-06, "loss": 0.1207, "step": 8578 }, { "epoch": 5.488803582853487, "grad_norm": 0.9894907451566008, "learning_rate": 8.904994610990658e-06, "loss": 0.1075, "step": 8579 }, { "epoch": 5.4894433781190015, "grad_norm": 1.0442110773992175, "learning_rate": 8.9029349409395e-06, "loss": 0.115, "step": 8580 }, { "epoch": 5.490083173384517, "grad_norm": 1.0424145329496732, "learning_rate": 8.900875317994362e-06, "loss": 0.1161, "step": 8581 }, { "epoch": 5.490722968650032, "grad_norm": 0.9805082671219221, "learning_rate": 8.898815742243682e-06, "loss": 0.1105, "step": 8582 }, { "epoch": 5.491362763915547, "grad_norm": 1.0423047910297127, "learning_rate": 8.896756213775897e-06, "loss": 0.1115, "step": 8583 }, { "epoch": 5.492002559181062, "grad_norm": 1.039305367120738, "learning_rate": 8.89469673267944e-06, "loss": 0.1153, "step": 8584 }, { "epoch": 5.492642354446577, "grad_norm": 1.053298666114086, "learning_rate": 8.892637299042735e-06, "loss": 0.1193, "step": 8585 }, { "epoch": 5.493282149712092, "grad_norm": 1.0097831832139987, "learning_rate": 8.890577912954214e-06, "loss": 0.117, "step": 8586 }, { "epoch": 5.4939219449776076, "grad_norm": 0.9807720089565246, "learning_rate": 8.888518574502306e-06, "loss": 0.1178, "step": 8587 }, { "epoch": 5.494561740243122, "grad_norm": 1.0859841050268229, "learning_rate": 8.886459283775434e-06, "loss": 0.1159, "step": 8588 }, { "epoch": 5.495201535508637, "grad_norm": 0.9863004780467306, "learning_rate": 8.884400040862017e-06, "loss": 0.1186, "step": 8589 }, { "epoch": 5.4958413307741525, "grad_norm": 1.102710605481491, "learning_rate": 8.882340845850482e-06, "loss": 0.1179, "step": 8590 }, { "epoch": 5.496481126039667, "grad_norm": 1.0163400328630148, "learning_rate": 8.880281698829239e-06, "loss": 0.1135, "step": 8591 }, { "epoch": 5.497120921305182, "grad_norm": 1.0239186904235944, "learning_rate": 8.878222599886707e-06, "loss": 0.1186, "step": 8592 }, { "epoch": 5.497760716570697, "grad_norm": 1.0671472729887563, "learning_rate": 8.876163549111304e-06, "loss": 0.1268, "step": 8593 }, { "epoch": 5.498400511836213, "grad_norm": 1.065741107181935, "learning_rate": 8.87410454659144e-06, "loss": 0.1277, "step": 8594 }, { "epoch": 5.499040307101727, "grad_norm": 1.0345504350747765, "learning_rate": 8.872045592415523e-06, "loss": 0.1154, "step": 8595 }, { "epoch": 5.499680102367242, "grad_norm": 0.9546663112441295, "learning_rate": 8.86998668667196e-06, "loss": 0.1135, "step": 8596 }, { "epoch": 5.500319897632758, "grad_norm": 0.9846856619087199, "learning_rate": 8.86792782944916e-06, "loss": 0.1165, "step": 8597 }, { "epoch": 5.500959692898273, "grad_norm": 1.043736967535953, "learning_rate": 8.865869020835528e-06, "loss": 0.1196, "step": 8598 }, { "epoch": 5.501599488163787, "grad_norm": 1.0055605852896712, "learning_rate": 8.863810260919454e-06, "loss": 0.1131, "step": 8599 }, { "epoch": 5.502239283429303, "grad_norm": 1.0581664725630022, "learning_rate": 8.861751549789353e-06, "loss": 0.1141, "step": 8600 }, { "epoch": 5.502879078694818, "grad_norm": 1.0691988094469522, "learning_rate": 8.859692887533617e-06, "loss": 0.1125, "step": 8601 }, { "epoch": 5.503518873960333, "grad_norm": 1.031571942942148, "learning_rate": 8.857634274240635e-06, "loss": 0.1187, "step": 8602 }, { "epoch": 5.5041586692258475, "grad_norm": 1.0847461412641144, "learning_rate": 8.855575709998809e-06, "loss": 0.1164, "step": 8603 }, { "epoch": 5.504798464491363, "grad_norm": 1.1002760197533317, "learning_rate": 8.853517194896525e-06, "loss": 0.1161, "step": 8604 }, { "epoch": 5.505438259756878, "grad_norm": 1.1598268636769515, "learning_rate": 8.851458729022176e-06, "loss": 0.1218, "step": 8605 }, { "epoch": 5.5060780550223924, "grad_norm": 1.015580106430825, "learning_rate": 8.849400312464139e-06, "loss": 0.1107, "step": 8606 }, { "epoch": 5.506717850287908, "grad_norm": 1.0990267838456649, "learning_rate": 8.847341945310812e-06, "loss": 0.117, "step": 8607 }, { "epoch": 5.507357645553423, "grad_norm": 0.9939242112901885, "learning_rate": 8.845283627650569e-06, "loss": 0.108, "step": 8608 }, { "epoch": 5.507997440818938, "grad_norm": 1.0375898674016113, "learning_rate": 8.843225359571793e-06, "loss": 0.116, "step": 8609 }, { "epoch": 5.508637236084453, "grad_norm": 1.0486839028383161, "learning_rate": 8.841167141162864e-06, "loss": 0.1196, "step": 8610 }, { "epoch": 5.509277031349968, "grad_norm": 1.074678284641484, "learning_rate": 8.839108972512156e-06, "loss": 0.1231, "step": 8611 }, { "epoch": 5.509916826615483, "grad_norm": 1.0951553329925587, "learning_rate": 8.83705085370804e-06, "loss": 0.1113, "step": 8612 }, { "epoch": 5.510556621880998, "grad_norm": 0.9628459693984505, "learning_rate": 8.834992784838895e-06, "loss": 0.1164, "step": 8613 }, { "epoch": 5.511196417146513, "grad_norm": 0.9379453551716275, "learning_rate": 8.832934765993088e-06, "loss": 0.1113, "step": 8614 }, { "epoch": 5.511836212412028, "grad_norm": 1.02327178856857, "learning_rate": 8.830876797258987e-06, "loss": 0.106, "step": 8615 }, { "epoch": 5.512476007677543, "grad_norm": 0.959278679562849, "learning_rate": 8.828818878724951e-06, "loss": 0.1106, "step": 8616 }, { "epoch": 5.513115802943059, "grad_norm": 1.1008746958373983, "learning_rate": 8.826761010479353e-06, "loss": 0.124, "step": 8617 }, { "epoch": 5.513755598208573, "grad_norm": 0.9065079402871525, "learning_rate": 8.824703192610552e-06, "loss": 0.1102, "step": 8618 }, { "epoch": 5.514395393474088, "grad_norm": 1.1288046045998479, "learning_rate": 8.822645425206901e-06, "loss": 0.1248, "step": 8619 }, { "epoch": 5.515035188739604, "grad_norm": 1.2092085528779648, "learning_rate": 8.820587708356765e-06, "loss": 0.1236, "step": 8620 }, { "epoch": 5.515674984005118, "grad_norm": 1.1683616023358656, "learning_rate": 8.818530042148494e-06, "loss": 0.125, "step": 8621 }, { "epoch": 5.516314779270633, "grad_norm": 1.039643030208376, "learning_rate": 8.816472426670438e-06, "loss": 0.118, "step": 8622 }, { "epoch": 5.516954574536149, "grad_norm": 0.9711840467498344, "learning_rate": 8.814414862010953e-06, "loss": 0.1206, "step": 8623 }, { "epoch": 5.517594369801664, "grad_norm": 1.1517526308206452, "learning_rate": 8.812357348258385e-06, "loss": 0.1219, "step": 8624 }, { "epoch": 5.518234165067178, "grad_norm": 0.9665013182476041, "learning_rate": 8.81029988550108e-06, "loss": 0.1132, "step": 8625 }, { "epoch": 5.5188739603326935, "grad_norm": 0.9831371791561705, "learning_rate": 8.808242473827383e-06, "loss": 0.1109, "step": 8626 }, { "epoch": 5.519513755598209, "grad_norm": 0.9905830166638245, "learning_rate": 8.806185113325632e-06, "loss": 0.1198, "step": 8627 }, { "epoch": 5.520153550863723, "grad_norm": 1.089183369245978, "learning_rate": 8.80412780408417e-06, "loss": 0.1182, "step": 8628 }, { "epoch": 5.520793346129238, "grad_norm": 0.9201831823667229, "learning_rate": 8.802070546191328e-06, "loss": 0.1092, "step": 8629 }, { "epoch": 5.521433141394754, "grad_norm": 0.9202077595678609, "learning_rate": 8.800013339735451e-06, "loss": 0.1062, "step": 8630 }, { "epoch": 5.522072936660269, "grad_norm": 0.9941475638736732, "learning_rate": 8.797956184804867e-06, "loss": 0.1117, "step": 8631 }, { "epoch": 5.522712731925783, "grad_norm": 1.0669189395832317, "learning_rate": 8.795899081487902e-06, "loss": 0.1101, "step": 8632 }, { "epoch": 5.523352527191299, "grad_norm": 1.1398538932976976, "learning_rate": 8.793842029872892e-06, "loss": 0.1215, "step": 8633 }, { "epoch": 5.523992322456814, "grad_norm": 1.1227103161844023, "learning_rate": 8.79178503004816e-06, "loss": 0.1226, "step": 8634 }, { "epoch": 5.524632117722329, "grad_norm": 0.9927846968094074, "learning_rate": 8.78972808210203e-06, "loss": 0.1144, "step": 8635 }, { "epoch": 5.525271912987844, "grad_norm": 1.033133497010597, "learning_rate": 8.787671186122816e-06, "loss": 0.1217, "step": 8636 }, { "epoch": 5.525911708253359, "grad_norm": 0.9537064127013403, "learning_rate": 8.785614342198852e-06, "loss": 0.1103, "step": 8637 }, { "epoch": 5.526551503518874, "grad_norm": 0.9305393036342757, "learning_rate": 8.783557550418447e-06, "loss": 0.1141, "step": 8638 }, { "epoch": 5.527191298784389, "grad_norm": 1.034200580577279, "learning_rate": 8.781500810869915e-06, "loss": 0.1177, "step": 8639 }, { "epoch": 5.527831094049904, "grad_norm": 0.9540682223245348, "learning_rate": 8.779444123641572e-06, "loss": 0.1125, "step": 8640 }, { "epoch": 5.528470889315419, "grad_norm": 1.0974883638909332, "learning_rate": 8.777387488821727e-06, "loss": 0.118, "step": 8641 }, { "epoch": 5.529110684580934, "grad_norm": 1.0498890105412653, "learning_rate": 8.77533090649869e-06, "loss": 0.1151, "step": 8642 }, { "epoch": 5.529750479846449, "grad_norm": 1.023327917020468, "learning_rate": 8.77327437676076e-06, "loss": 0.117, "step": 8643 }, { "epoch": 5.530390275111964, "grad_norm": 0.9942399898128773, "learning_rate": 8.77121789969625e-06, "loss": 0.1168, "step": 8644 }, { "epoch": 5.531030070377479, "grad_norm": 0.8610302643364225, "learning_rate": 8.769161475393458e-06, "loss": 0.1078, "step": 8645 }, { "epoch": 5.531669865642995, "grad_norm": 1.0952703394922565, "learning_rate": 8.767105103940678e-06, "loss": 0.1183, "step": 8646 }, { "epoch": 5.532309660908509, "grad_norm": 1.0149913594927817, "learning_rate": 8.765048785426216e-06, "loss": 0.1154, "step": 8647 }, { "epoch": 5.532949456174024, "grad_norm": 1.168587978110621, "learning_rate": 8.762992519938363e-06, "loss": 0.1243, "step": 8648 }, { "epoch": 5.5335892514395395, "grad_norm": 1.0936586793323813, "learning_rate": 8.76093630756541e-06, "loss": 0.1192, "step": 8649 }, { "epoch": 5.534229046705054, "grad_norm": 1.0624998363345612, "learning_rate": 8.758880148395648e-06, "loss": 0.1081, "step": 8650 }, { "epoch": 5.534868841970569, "grad_norm": 1.0664297608170386, "learning_rate": 8.756824042517364e-06, "loss": 0.1205, "step": 8651 }, { "epoch": 5.535508637236084, "grad_norm": 1.0423680132698665, "learning_rate": 8.754767990018847e-06, "loss": 0.12, "step": 8652 }, { "epoch": 5.5361484325016, "grad_norm": 1.1898462330696997, "learning_rate": 8.75271199098837e-06, "loss": 0.1203, "step": 8653 }, { "epoch": 5.536788227767115, "grad_norm": 1.1497467101432508, "learning_rate": 8.750656045514228e-06, "loss": 0.1205, "step": 8654 }, { "epoch": 5.537428023032629, "grad_norm": 1.0366377108475526, "learning_rate": 8.74860015368469e-06, "loss": 0.1147, "step": 8655 }, { "epoch": 5.538067818298145, "grad_norm": 0.9487570422969798, "learning_rate": 8.746544315588036e-06, "loss": 0.1096, "step": 8656 }, { "epoch": 5.53870761356366, "grad_norm": 1.0626437888685494, "learning_rate": 8.74448853131254e-06, "loss": 0.1262, "step": 8657 }, { "epoch": 5.539347408829174, "grad_norm": 0.9595207132855382, "learning_rate": 8.742432800946475e-06, "loss": 0.1128, "step": 8658 }, { "epoch": 5.53998720409469, "grad_norm": 0.9761639811545003, "learning_rate": 8.740377124578101e-06, "loss": 0.1157, "step": 8659 }, { "epoch": 5.540626999360205, "grad_norm": 1.0752110901910255, "learning_rate": 8.738321502295701e-06, "loss": 0.1227, "step": 8660 }, { "epoch": 5.54126679462572, "grad_norm": 0.955125137384954, "learning_rate": 8.736265934187529e-06, "loss": 0.1134, "step": 8661 }, { "epoch": 5.5419065898912345, "grad_norm": 1.034687156071993, "learning_rate": 8.73421042034185e-06, "loss": 0.1225, "step": 8662 }, { "epoch": 5.54254638515675, "grad_norm": 1.0046084314522326, "learning_rate": 8.732154960846921e-06, "loss": 0.1196, "step": 8663 }, { "epoch": 5.543186180422265, "grad_norm": 0.9557281875913477, "learning_rate": 8.730099555791006e-06, "loss": 0.1188, "step": 8664 }, { "epoch": 5.5438259756877795, "grad_norm": 0.9100352856707241, "learning_rate": 8.728044205262355e-06, "loss": 0.1209, "step": 8665 }, { "epoch": 5.544465770953295, "grad_norm": 0.9155458942673975, "learning_rate": 8.725988909349222e-06, "loss": 0.1186, "step": 8666 }, { "epoch": 5.54510556621881, "grad_norm": 1.1266283940263866, "learning_rate": 8.723933668139861e-06, "loss": 0.1211, "step": 8667 }, { "epoch": 5.545745361484325, "grad_norm": 1.0637609450095047, "learning_rate": 8.72187848172252e-06, "loss": 0.1118, "step": 8668 }, { "epoch": 5.54638515674984, "grad_norm": 1.0711652002413476, "learning_rate": 8.719823350185439e-06, "loss": 0.1148, "step": 8669 }, { "epoch": 5.547024952015355, "grad_norm": 0.9796605562379623, "learning_rate": 8.717768273616869e-06, "loss": 0.1146, "step": 8670 }, { "epoch": 5.54766474728087, "grad_norm": 1.1035199347217857, "learning_rate": 8.71571325210505e-06, "loss": 0.1225, "step": 8671 }, { "epoch": 5.5483045425463855, "grad_norm": 0.8873561606481924, "learning_rate": 8.713658285738217e-06, "loss": 0.1173, "step": 8672 }, { "epoch": 5.5489443378119, "grad_norm": 1.0929128818260765, "learning_rate": 8.711603374604607e-06, "loss": 0.1229, "step": 8673 }, { "epoch": 5.549584133077415, "grad_norm": 1.0297746791337228, "learning_rate": 8.709548518792459e-06, "loss": 0.12, "step": 8674 }, { "epoch": 5.55022392834293, "grad_norm": 0.9128061080286747, "learning_rate": 8.707493718390003e-06, "loss": 0.1089, "step": 8675 }, { "epoch": 5.550863723608446, "grad_norm": 1.0156267635478566, "learning_rate": 8.705438973485463e-06, "loss": 0.1155, "step": 8676 }, { "epoch": 5.55150351887396, "grad_norm": 1.0122316424538176, "learning_rate": 8.703384284167075e-06, "loss": 0.1202, "step": 8677 }, { "epoch": 5.552143314139475, "grad_norm": 1.0009299124023083, "learning_rate": 8.701329650523059e-06, "loss": 0.1238, "step": 8678 }, { "epoch": 5.552783109404991, "grad_norm": 1.0409268795863291, "learning_rate": 8.699275072641635e-06, "loss": 0.1184, "step": 8679 }, { "epoch": 5.553422904670505, "grad_norm": 0.8669860103417769, "learning_rate": 8.697220550611028e-06, "loss": 0.1094, "step": 8680 }, { "epoch": 5.55406269993602, "grad_norm": 0.9683911018184929, "learning_rate": 8.695166084519454e-06, "loss": 0.1139, "step": 8681 }, { "epoch": 5.554702495201536, "grad_norm": 0.955836590441477, "learning_rate": 8.693111674455126e-06, "loss": 0.1114, "step": 8682 }, { "epoch": 5.555342290467051, "grad_norm": 1.0527850073112308, "learning_rate": 8.691057320506254e-06, "loss": 0.1155, "step": 8683 }, { "epoch": 5.555982085732565, "grad_norm": 0.9840073354373463, "learning_rate": 8.689003022761059e-06, "loss": 0.111, "step": 8684 }, { "epoch": 5.5566218809980805, "grad_norm": 1.0127126882376185, "learning_rate": 8.686948781307739e-06, "loss": 0.1103, "step": 8685 }, { "epoch": 5.557261676263596, "grad_norm": 1.0359008524679967, "learning_rate": 8.684894596234502e-06, "loss": 0.1212, "step": 8686 }, { "epoch": 5.557901471529111, "grad_norm": 1.0632554031878096, "learning_rate": 8.682840467629552e-06, "loss": 0.1192, "step": 8687 }, { "epoch": 5.5585412667946255, "grad_norm": 0.9927025576349434, "learning_rate": 8.68078639558109e-06, "loss": 0.1165, "step": 8688 }, { "epoch": 5.559181062060141, "grad_norm": 0.8675420846413581, "learning_rate": 8.678732380177309e-06, "loss": 0.1133, "step": 8689 }, { "epoch": 5.559820857325656, "grad_norm": 1.01081443664948, "learning_rate": 8.676678421506415e-06, "loss": 0.1173, "step": 8690 }, { "epoch": 5.560460652591171, "grad_norm": 0.9869712681331362, "learning_rate": 8.674624519656594e-06, "loss": 0.1091, "step": 8691 }, { "epoch": 5.561100447856686, "grad_norm": 0.9813393566519312, "learning_rate": 8.67257067471604e-06, "loss": 0.1162, "step": 8692 }, { "epoch": 5.561740243122201, "grad_norm": 0.9339306189752115, "learning_rate": 8.670516886772936e-06, "loss": 0.107, "step": 8693 }, { "epoch": 5.562380038387716, "grad_norm": 1.0274696352155783, "learning_rate": 8.668463155915474e-06, "loss": 0.117, "step": 8694 }, { "epoch": 5.563019833653231, "grad_norm": 0.9986505043296378, "learning_rate": 8.666409482231837e-06, "loss": 0.1161, "step": 8695 }, { "epoch": 5.563659628918746, "grad_norm": 0.9139121006352006, "learning_rate": 8.6643558658102e-06, "loss": 0.113, "step": 8696 }, { "epoch": 5.564299424184261, "grad_norm": 0.9990325806632742, "learning_rate": 8.66230230673875e-06, "loss": 0.113, "step": 8697 }, { "epoch": 5.564939219449776, "grad_norm": 0.9995231149267002, "learning_rate": 8.66024880510566e-06, "loss": 0.1227, "step": 8698 }, { "epoch": 5.565579014715291, "grad_norm": 0.9649785746564867, "learning_rate": 8.6581953609991e-06, "loss": 0.1064, "step": 8699 }, { "epoch": 5.566218809980806, "grad_norm": 1.009280041610791, "learning_rate": 8.656141974507249e-06, "loss": 0.1147, "step": 8700 }, { "epoch": 5.566858605246321, "grad_norm": 1.0211621246217752, "learning_rate": 8.654088645718269e-06, "loss": 0.1137, "step": 8701 }, { "epoch": 5.567498400511836, "grad_norm": 1.1305964042046612, "learning_rate": 8.652035374720327e-06, "loss": 0.1291, "step": 8702 }, { "epoch": 5.568138195777351, "grad_norm": 1.0348353623015922, "learning_rate": 8.649982161601585e-06, "loss": 0.1144, "step": 8703 }, { "epoch": 5.568777991042866, "grad_norm": 0.9118349387455105, "learning_rate": 8.647929006450213e-06, "loss": 0.1193, "step": 8704 }, { "epoch": 5.569417786308382, "grad_norm": 1.047074038591954, "learning_rate": 8.645875909354363e-06, "loss": 0.1194, "step": 8705 }, { "epoch": 5.570057581573896, "grad_norm": 1.0573270158169206, "learning_rate": 8.643822870402188e-06, "loss": 0.1131, "step": 8706 }, { "epoch": 5.570697376839411, "grad_norm": 1.0692566580816212, "learning_rate": 8.641769889681852e-06, "loss": 0.1208, "step": 8707 }, { "epoch": 5.5713371721049265, "grad_norm": 0.9422387493771881, "learning_rate": 8.6397169672815e-06, "loss": 0.1145, "step": 8708 }, { "epoch": 5.571976967370442, "grad_norm": 0.9970539170082168, "learning_rate": 8.63766410328928e-06, "loss": 0.1185, "step": 8709 }, { "epoch": 5.572616762635956, "grad_norm": 0.9527023639755118, "learning_rate": 8.635611297793343e-06, "loss": 0.1168, "step": 8710 }, { "epoch": 5.5732565579014715, "grad_norm": 0.9777654654333594, "learning_rate": 8.63355855088183e-06, "loss": 0.1139, "step": 8711 }, { "epoch": 5.573896353166987, "grad_norm": 0.9449465771875059, "learning_rate": 8.631505862642879e-06, "loss": 0.1155, "step": 8712 }, { "epoch": 5.574536148432502, "grad_norm": 1.0009073261738408, "learning_rate": 8.62945323316463e-06, "loss": 0.1111, "step": 8713 }, { "epoch": 5.575175943698016, "grad_norm": 0.9579724211423387, "learning_rate": 8.627400662535224e-06, "loss": 0.1124, "step": 8714 }, { "epoch": 5.575815738963532, "grad_norm": 0.9976703508793979, "learning_rate": 8.625348150842793e-06, "loss": 0.1167, "step": 8715 }, { "epoch": 5.576455534229047, "grad_norm": 1.0566977646765219, "learning_rate": 8.623295698175466e-06, "loss": 0.1172, "step": 8716 }, { "epoch": 5.577095329494561, "grad_norm": 0.9443106763392664, "learning_rate": 8.621243304621373e-06, "loss": 0.1135, "step": 8717 }, { "epoch": 5.577735124760077, "grad_norm": 1.0236343147117297, "learning_rate": 8.619190970268641e-06, "loss": 0.1133, "step": 8718 }, { "epoch": 5.578374920025592, "grad_norm": 1.1101954636406075, "learning_rate": 8.617138695205391e-06, "loss": 0.1165, "step": 8719 }, { "epoch": 5.579014715291107, "grad_norm": 1.0274544973647048, "learning_rate": 8.615086479519742e-06, "loss": 0.1141, "step": 8720 }, { "epoch": 5.5796545105566215, "grad_norm": 1.0531778547034174, "learning_rate": 8.613034323299821e-06, "loss": 0.1159, "step": 8721 }, { "epoch": 5.580294305822137, "grad_norm": 0.9486591174343667, "learning_rate": 8.61098222663374e-06, "loss": 0.1139, "step": 8722 }, { "epoch": 5.580934101087652, "grad_norm": 0.9741967911328342, "learning_rate": 8.608930189609606e-06, "loss": 0.1182, "step": 8723 }, { "epoch": 5.581573896353167, "grad_norm": 1.062532011259621, "learning_rate": 8.606878212315539e-06, "loss": 0.1105, "step": 8724 }, { "epoch": 5.582213691618682, "grad_norm": 0.9025183905567047, "learning_rate": 8.604826294839645e-06, "loss": 0.1086, "step": 8725 }, { "epoch": 5.582853486884197, "grad_norm": 1.0470392467443452, "learning_rate": 8.602774437270021e-06, "loss": 0.1169, "step": 8726 }, { "epoch": 5.583493282149712, "grad_norm": 1.092584380442984, "learning_rate": 8.600722639694783e-06, "loss": 0.1143, "step": 8727 }, { "epoch": 5.584133077415228, "grad_norm": 0.9751680357172489, "learning_rate": 8.598670902202025e-06, "loss": 0.1125, "step": 8728 }, { "epoch": 5.584772872680742, "grad_norm": 0.9636133535959601, "learning_rate": 8.596619224879846e-06, "loss": 0.1161, "step": 8729 }, { "epoch": 5.585412667946257, "grad_norm": 0.8880736986004462, "learning_rate": 8.594567607816339e-06, "loss": 0.1115, "step": 8730 }, { "epoch": 5.5860524632117725, "grad_norm": 0.9679654212028809, "learning_rate": 8.5925160510996e-06, "loss": 0.1178, "step": 8731 }, { "epoch": 5.586692258477287, "grad_norm": 0.9575788338054998, "learning_rate": 8.59046455481772e-06, "loss": 0.1163, "step": 8732 }, { "epoch": 5.587332053742802, "grad_norm": 1.1038843963968081, "learning_rate": 8.58841311905878e-06, "loss": 0.125, "step": 8733 }, { "epoch": 5.5879718490083174, "grad_norm": 1.1451898202756003, "learning_rate": 8.586361743910873e-06, "loss": 0.1247, "step": 8734 }, { "epoch": 5.588611644273833, "grad_norm": 1.0218954327252965, "learning_rate": 8.584310429462078e-06, "loss": 0.1142, "step": 8735 }, { "epoch": 5.589251439539347, "grad_norm": 1.1293031560389821, "learning_rate": 8.582259175800472e-06, "loss": 0.119, "step": 8736 }, { "epoch": 5.589891234804862, "grad_norm": 1.0738982419236005, "learning_rate": 8.580207983014136e-06, "loss": 0.1181, "step": 8737 }, { "epoch": 5.590531030070378, "grad_norm": 1.021993700924603, "learning_rate": 8.578156851191145e-06, "loss": 0.1164, "step": 8738 }, { "epoch": 5.591170825335892, "grad_norm": 1.1032010499011449, "learning_rate": 8.576105780419572e-06, "loss": 0.1139, "step": 8739 }, { "epoch": 5.591810620601407, "grad_norm": 1.0183588508477983, "learning_rate": 8.57405477078748e-06, "loss": 0.1201, "step": 8740 }, { "epoch": 5.592450415866923, "grad_norm": 1.0642142289411949, "learning_rate": 8.572003822382942e-06, "loss": 0.1222, "step": 8741 }, { "epoch": 5.593090211132438, "grad_norm": 1.1805788614263009, "learning_rate": 8.569952935294019e-06, "loss": 0.1211, "step": 8742 }, { "epoch": 5.593730006397953, "grad_norm": 0.849929540223136, "learning_rate": 8.56790210960877e-06, "loss": 0.1086, "step": 8743 }, { "epoch": 5.5943698016634675, "grad_norm": 1.063158319995786, "learning_rate": 8.56585134541526e-06, "loss": 0.1235, "step": 8744 }, { "epoch": 5.595009596928983, "grad_norm": 0.9515660849403506, "learning_rate": 8.563800642801543e-06, "loss": 0.1146, "step": 8745 }, { "epoch": 5.595649392194498, "grad_norm": 1.036111335240055, "learning_rate": 8.56175000185567e-06, "loss": 0.1243, "step": 8746 }, { "epoch": 5.5962891874600125, "grad_norm": 1.11406595193279, "learning_rate": 8.559699422665694e-06, "loss": 0.1182, "step": 8747 }, { "epoch": 5.596928982725528, "grad_norm": 1.0313707651247412, "learning_rate": 8.557648905319663e-06, "loss": 0.1123, "step": 8748 }, { "epoch": 5.597568777991043, "grad_norm": 1.0492652374724536, "learning_rate": 8.555598449905622e-06, "loss": 0.1179, "step": 8749 }, { "epoch": 5.598208573256558, "grad_norm": 0.9370463728254674, "learning_rate": 8.55354805651161e-06, "loss": 0.115, "step": 8750 }, { "epoch": 5.598848368522073, "grad_norm": 1.0302523050727723, "learning_rate": 8.551497725225677e-06, "loss": 0.1192, "step": 8751 }, { "epoch": 5.599488163787588, "grad_norm": 1.0658477651629772, "learning_rate": 8.549447456135855e-06, "loss": 0.1109, "step": 8752 }, { "epoch": 5.600127959053103, "grad_norm": 1.0491394201041098, "learning_rate": 8.547397249330174e-06, "loss": 0.1114, "step": 8753 }, { "epoch": 5.600767754318618, "grad_norm": 1.0204955450982602, "learning_rate": 8.545347104896677e-06, "loss": 0.1165, "step": 8754 }, { "epoch": 5.601407549584133, "grad_norm": 1.0031132502934852, "learning_rate": 8.543297022923385e-06, "loss": 0.1108, "step": 8755 }, { "epoch": 5.602047344849648, "grad_norm": 0.9405851260294399, "learning_rate": 8.541247003498325e-06, "loss": 0.1048, "step": 8756 }, { "epoch": 5.6026871401151634, "grad_norm": 1.0037748992257287, "learning_rate": 8.539197046709526e-06, "loss": 0.122, "step": 8757 }, { "epoch": 5.603326935380678, "grad_norm": 1.0983218377267296, "learning_rate": 8.537147152645009e-06, "loss": 0.125, "step": 8758 }, { "epoch": 5.603966730646193, "grad_norm": 0.9574537973982823, "learning_rate": 8.535097321392792e-06, "loss": 0.107, "step": 8759 }, { "epoch": 5.604606525911708, "grad_norm": 0.9818535543283722, "learning_rate": 8.533047553040886e-06, "loss": 0.115, "step": 8760 }, { "epoch": 5.605246321177224, "grad_norm": 0.9080904784115522, "learning_rate": 8.530997847677311e-06, "loss": 0.1042, "step": 8761 }, { "epoch": 5.605886116442738, "grad_norm": 1.0745564190440564, "learning_rate": 8.528948205390076e-06, "loss": 0.12, "step": 8762 }, { "epoch": 5.606525911708253, "grad_norm": 1.1161329722200375, "learning_rate": 8.526898626267182e-06, "loss": 0.1252, "step": 8763 }, { "epoch": 5.607165706973769, "grad_norm": 1.0035446562615968, "learning_rate": 8.524849110396647e-06, "loss": 0.1137, "step": 8764 }, { "epoch": 5.607805502239284, "grad_norm": 1.0757209658752391, "learning_rate": 8.522799657866466e-06, "loss": 0.1224, "step": 8765 }, { "epoch": 5.608445297504798, "grad_norm": 0.9619295229803817, "learning_rate": 8.520750268764636e-06, "loss": 0.1183, "step": 8766 }, { "epoch": 5.6090850927703135, "grad_norm": 1.017586667864231, "learning_rate": 8.518700943179163e-06, "loss": 0.1189, "step": 8767 }, { "epoch": 5.609724888035829, "grad_norm": 1.010665219233189, "learning_rate": 8.516651681198037e-06, "loss": 0.1117, "step": 8768 }, { "epoch": 5.610364683301343, "grad_norm": 0.9440731299399948, "learning_rate": 8.514602482909247e-06, "loss": 0.1185, "step": 8769 }, { "epoch": 5.6110044785668585, "grad_norm": 1.0874697824777604, "learning_rate": 8.512553348400782e-06, "loss": 0.1178, "step": 8770 }, { "epoch": 5.611644273832374, "grad_norm": 0.9711628209891962, "learning_rate": 8.510504277760633e-06, "loss": 0.1144, "step": 8771 }, { "epoch": 5.612284069097889, "grad_norm": 0.8713550170162432, "learning_rate": 8.508455271076783e-06, "loss": 0.1109, "step": 8772 }, { "epoch": 5.612923864363403, "grad_norm": 1.0535235345700478, "learning_rate": 8.506406328437203e-06, "loss": 0.1192, "step": 8773 }, { "epoch": 5.613563659628919, "grad_norm": 0.9746438993534831, "learning_rate": 8.504357449929885e-06, "loss": 0.1125, "step": 8774 }, { "epoch": 5.614203454894434, "grad_norm": 0.9207649558905232, "learning_rate": 8.502308635642795e-06, "loss": 0.1153, "step": 8775 }, { "epoch": 5.614843250159948, "grad_norm": 0.9712596891903995, "learning_rate": 8.500259885663908e-06, "loss": 0.1188, "step": 8776 }, { "epoch": 5.615483045425464, "grad_norm": 1.1126798386002958, "learning_rate": 8.498211200081194e-06, "loss": 0.1095, "step": 8777 }, { "epoch": 5.616122840690979, "grad_norm": 1.0182436398176073, "learning_rate": 8.496162578982621e-06, "loss": 0.1132, "step": 8778 }, { "epoch": 5.616762635956494, "grad_norm": 1.0193337258184456, "learning_rate": 8.494114022456152e-06, "loss": 0.1196, "step": 8779 }, { "epoch": 5.617402431222009, "grad_norm": 1.1173529820806385, "learning_rate": 8.492065530589743e-06, "loss": 0.1176, "step": 8780 }, { "epoch": 5.618042226487524, "grad_norm": 1.0686071662686978, "learning_rate": 8.490017103471361e-06, "loss": 0.1199, "step": 8781 }, { "epoch": 5.618682021753039, "grad_norm": 0.97043608504125, "learning_rate": 8.487968741188962e-06, "loss": 0.1121, "step": 8782 }, { "epoch": 5.619321817018554, "grad_norm": 0.9917561079382213, "learning_rate": 8.48592044383049e-06, "loss": 0.1063, "step": 8783 }, { "epoch": 5.619961612284069, "grad_norm": 0.98068541654141, "learning_rate": 8.483872211483905e-06, "loss": 0.1155, "step": 8784 }, { "epoch": 5.620601407549584, "grad_norm": 0.989528460949211, "learning_rate": 8.481824044237147e-06, "loss": 0.1151, "step": 8785 }, { "epoch": 5.621241202815099, "grad_norm": 0.9714531731663465, "learning_rate": 8.479775942178164e-06, "loss": 0.1173, "step": 8786 }, { "epoch": 5.621880998080615, "grad_norm": 1.0708602695021376, "learning_rate": 8.477727905394899e-06, "loss": 0.1205, "step": 8787 }, { "epoch": 5.622520793346129, "grad_norm": 0.9900291018601197, "learning_rate": 8.475679933975292e-06, "loss": 0.1205, "step": 8788 }, { "epoch": 5.623160588611644, "grad_norm": 1.1124325207613135, "learning_rate": 8.473632028007275e-06, "loss": 0.1208, "step": 8789 }, { "epoch": 5.6238003838771595, "grad_norm": 1.0318974530542881, "learning_rate": 8.471584187578783e-06, "loss": 0.1212, "step": 8790 }, { "epoch": 5.624440179142674, "grad_norm": 1.1138465852489368, "learning_rate": 8.469536412777749e-06, "loss": 0.1259, "step": 8791 }, { "epoch": 5.625079974408189, "grad_norm": 1.0465532829286108, "learning_rate": 8.467488703692098e-06, "loss": 0.1128, "step": 8792 }, { "epoch": 5.6257197696737045, "grad_norm": 1.142462163524108, "learning_rate": 8.465441060409752e-06, "loss": 0.1177, "step": 8793 }, { "epoch": 5.62635956493922, "grad_norm": 1.0673934183370277, "learning_rate": 8.463393483018643e-06, "loss": 0.1159, "step": 8794 }, { "epoch": 5.626999360204734, "grad_norm": 1.1264115078506713, "learning_rate": 8.461345971606682e-06, "loss": 0.1128, "step": 8795 }, { "epoch": 5.627639155470249, "grad_norm": 1.0592973508912218, "learning_rate": 8.459298526261789e-06, "loss": 0.1159, "step": 8796 }, { "epoch": 5.628278950735765, "grad_norm": 1.0941190333521842, "learning_rate": 8.457251147071873e-06, "loss": 0.1158, "step": 8797 }, { "epoch": 5.62891874600128, "grad_norm": 0.9971374165187992, "learning_rate": 8.455203834124851e-06, "loss": 0.1189, "step": 8798 }, { "epoch": 5.629558541266794, "grad_norm": 0.9437652362058666, "learning_rate": 8.453156587508629e-06, "loss": 0.1198, "step": 8799 }, { "epoch": 5.63019833653231, "grad_norm": 0.9487541690834334, "learning_rate": 8.451109407311111e-06, "loss": 0.1127, "step": 8800 }, { "epoch": 5.630838131797825, "grad_norm": 1.2131889596149033, "learning_rate": 8.4490622936202e-06, "loss": 0.1381, "step": 8801 }, { "epoch": 5.63147792706334, "grad_norm": 1.0689810055345346, "learning_rate": 8.447015246523796e-06, "loss": 0.1196, "step": 8802 }, { "epoch": 5.6321177223288545, "grad_norm": 0.9684732009049053, "learning_rate": 8.44496826610979e-06, "loss": 0.1211, "step": 8803 }, { "epoch": 5.63275751759437, "grad_norm": 0.9330183015997993, "learning_rate": 8.442921352466085e-06, "loss": 0.1143, "step": 8804 }, { "epoch": 5.633397312859885, "grad_norm": 1.0370707615837769, "learning_rate": 8.440874505680568e-06, "loss": 0.1216, "step": 8805 }, { "epoch": 5.6340371081253995, "grad_norm": 0.9299336094014367, "learning_rate": 8.438827725841125e-06, "loss": 0.1066, "step": 8806 }, { "epoch": 5.634676903390915, "grad_norm": 0.9898166989146513, "learning_rate": 8.436781013035642e-06, "loss": 0.108, "step": 8807 }, { "epoch": 5.63531669865643, "grad_norm": 1.1975082811582047, "learning_rate": 8.434734367352003e-06, "loss": 0.1234, "step": 8808 }, { "epoch": 5.635956493921945, "grad_norm": 0.9451965981372658, "learning_rate": 8.432687788878085e-06, "loss": 0.1143, "step": 8809 }, { "epoch": 5.63659628918746, "grad_norm": 1.0048436495942294, "learning_rate": 8.430641277701762e-06, "loss": 0.1115, "step": 8810 }, { "epoch": 5.637236084452975, "grad_norm": 1.0839219976733425, "learning_rate": 8.428594833910914e-06, "loss": 0.1178, "step": 8811 }, { "epoch": 5.63787587971849, "grad_norm": 1.0180351327036923, "learning_rate": 8.426548457593408e-06, "loss": 0.1075, "step": 8812 }, { "epoch": 5.638515674984005, "grad_norm": 1.0671386254327566, "learning_rate": 8.424502148837112e-06, "loss": 0.1152, "step": 8813 }, { "epoch": 5.63915547024952, "grad_norm": 1.1603149726840112, "learning_rate": 8.422455907729891e-06, "loss": 0.1249, "step": 8814 }, { "epoch": 5.639795265515035, "grad_norm": 0.9923316140790519, "learning_rate": 8.420409734359611e-06, "loss": 0.1141, "step": 8815 }, { "epoch": 5.6404350607805505, "grad_norm": 1.0589142371965747, "learning_rate": 8.418363628814124e-06, "loss": 0.1143, "step": 8816 }, { "epoch": 5.641074856046066, "grad_norm": 1.107714863031925, "learning_rate": 8.416317591181285e-06, "loss": 0.1277, "step": 8817 }, { "epoch": 5.64171465131158, "grad_norm": 1.0435344214373865, "learning_rate": 8.414271621548957e-06, "loss": 0.1213, "step": 8818 }, { "epoch": 5.642354446577095, "grad_norm": 1.0159504508432708, "learning_rate": 8.412225720004984e-06, "loss": 0.1257, "step": 8819 }, { "epoch": 5.642994241842611, "grad_norm": 0.9438343802510534, "learning_rate": 8.410179886637213e-06, "loss": 0.1145, "step": 8820 }, { "epoch": 5.643634037108125, "grad_norm": 1.0254056188097436, "learning_rate": 8.408134121533491e-06, "loss": 0.1252, "step": 8821 }, { "epoch": 5.64427383237364, "grad_norm": 1.0403140065591687, "learning_rate": 8.406088424781657e-06, "loss": 0.1167, "step": 8822 }, { "epoch": 5.644913627639156, "grad_norm": 1.0008284376603862, "learning_rate": 8.404042796469549e-06, "loss": 0.1108, "step": 8823 }, { "epoch": 5.645553422904671, "grad_norm": 1.0348550326158774, "learning_rate": 8.401997236685007e-06, "loss": 0.1191, "step": 8824 }, { "epoch": 5.646193218170185, "grad_norm": 1.0607046355797607, "learning_rate": 8.39995174551586e-06, "loss": 0.1147, "step": 8825 }, { "epoch": 5.6468330134357005, "grad_norm": 1.012482784443604, "learning_rate": 8.397906323049941e-06, "loss": 0.1109, "step": 8826 }, { "epoch": 5.647472808701216, "grad_norm": 1.0479976538631453, "learning_rate": 8.395860969375068e-06, "loss": 0.1187, "step": 8827 }, { "epoch": 5.64811260396673, "grad_norm": 0.9735328361106971, "learning_rate": 8.393815684579079e-06, "loss": 0.1087, "step": 8828 }, { "epoch": 5.6487523992322455, "grad_norm": 1.000516656097568, "learning_rate": 8.391770468749784e-06, "loss": 0.1181, "step": 8829 }, { "epoch": 5.649392194497761, "grad_norm": 1.044303253714178, "learning_rate": 8.389725321975005e-06, "loss": 0.1208, "step": 8830 }, { "epoch": 5.650031989763276, "grad_norm": 1.0412887714799155, "learning_rate": 8.387680244342555e-06, "loss": 0.1183, "step": 8831 }, { "epoch": 5.65067178502879, "grad_norm": 1.0320910719767806, "learning_rate": 8.385635235940249e-06, "loss": 0.1144, "step": 8832 }, { "epoch": 5.651311580294306, "grad_norm": 1.1027884925069575, "learning_rate": 8.38359029685589e-06, "loss": 0.1218, "step": 8833 }, { "epoch": 5.651951375559821, "grad_norm": 0.9900079745561176, "learning_rate": 8.381545427177293e-06, "loss": 0.1124, "step": 8834 }, { "epoch": 5.652591170825336, "grad_norm": 1.0500179005690118, "learning_rate": 8.379500626992256e-06, "loss": 0.1149, "step": 8835 }, { "epoch": 5.653230966090851, "grad_norm": 0.9459400241443624, "learning_rate": 8.377455896388579e-06, "loss": 0.1149, "step": 8836 }, { "epoch": 5.653870761356366, "grad_norm": 0.9615636168577045, "learning_rate": 8.375411235454058e-06, "loss": 0.113, "step": 8837 }, { "epoch": 5.654510556621881, "grad_norm": 1.0642794234739408, "learning_rate": 8.373366644276491e-06, "loss": 0.1165, "step": 8838 }, { "epoch": 5.6551503518873965, "grad_norm": 0.9911391491335425, "learning_rate": 8.371322122943665e-06, "loss": 0.1174, "step": 8839 }, { "epoch": 5.655790147152911, "grad_norm": 0.9340265333576258, "learning_rate": 8.369277671543367e-06, "loss": 0.11, "step": 8840 }, { "epoch": 5.656429942418426, "grad_norm": 1.0812760753625135, "learning_rate": 8.36723329016339e-06, "loss": 0.1253, "step": 8841 }, { "epoch": 5.657069737683941, "grad_norm": 0.9785607542895911, "learning_rate": 8.36518897889151e-06, "loss": 0.111, "step": 8842 }, { "epoch": 5.657709532949456, "grad_norm": 1.033194600575191, "learning_rate": 8.363144737815504e-06, "loss": 0.1109, "step": 8843 }, { "epoch": 5.658349328214971, "grad_norm": 1.0254330018038356, "learning_rate": 8.361100567023156e-06, "loss": 0.1178, "step": 8844 }, { "epoch": 5.658989123480486, "grad_norm": 1.0759626587363083, "learning_rate": 8.359056466602232e-06, "loss": 0.1138, "step": 8845 }, { "epoch": 5.659628918746002, "grad_norm": 1.1248646987735988, "learning_rate": 8.357012436640505e-06, "loss": 0.1201, "step": 8846 }, { "epoch": 5.660268714011516, "grad_norm": 1.0968461636192166, "learning_rate": 8.354968477225736e-06, "loss": 0.1237, "step": 8847 }, { "epoch": 5.660908509277031, "grad_norm": 0.9509402837405708, "learning_rate": 8.3529245884457e-06, "loss": 0.11, "step": 8848 }, { "epoch": 5.6615483045425465, "grad_norm": 1.0000111602078563, "learning_rate": 8.350880770388151e-06, "loss": 0.1127, "step": 8849 }, { "epoch": 5.662188099808061, "grad_norm": 1.0760779523474906, "learning_rate": 8.348837023140847e-06, "loss": 0.1149, "step": 8850 }, { "epoch": 5.662827895073576, "grad_norm": 1.0271329954749844, "learning_rate": 8.346793346791547e-06, "loss": 0.1143, "step": 8851 }, { "epoch": 5.6634676903390915, "grad_norm": 0.9832358374277858, "learning_rate": 8.344749741427999e-06, "loss": 0.1136, "step": 8852 }, { "epoch": 5.664107485604607, "grad_norm": 1.0493433961653824, "learning_rate": 8.342706207137949e-06, "loss": 0.1186, "step": 8853 }, { "epoch": 5.664747280870122, "grad_norm": 0.8857247633511792, "learning_rate": 8.340662744009149e-06, "loss": 0.1129, "step": 8854 }, { "epoch": 5.665387076135636, "grad_norm": 1.0082501449580994, "learning_rate": 8.33861935212934e-06, "loss": 0.1102, "step": 8855 }, { "epoch": 5.666026871401152, "grad_norm": 1.000983973474625, "learning_rate": 8.33657603158626e-06, "loss": 0.1202, "step": 8856 }, { "epoch": 5.666666666666667, "grad_norm": 0.9425279172539344, "learning_rate": 8.334532782467644e-06, "loss": 0.1088, "step": 8857 }, { "epoch": 5.667306461932181, "grad_norm": 1.0597294177575132, "learning_rate": 8.33248960486123e-06, "loss": 0.1167, "step": 8858 }, { "epoch": 5.667946257197697, "grad_norm": 1.00183140906409, "learning_rate": 8.33044649885475e-06, "loss": 0.1176, "step": 8859 }, { "epoch": 5.668586052463212, "grad_norm": 1.0581907878600452, "learning_rate": 8.328403464535919e-06, "loss": 0.1241, "step": 8860 }, { "epoch": 5.669225847728727, "grad_norm": 1.0155270793739533, "learning_rate": 8.326360501992475e-06, "loss": 0.1169, "step": 8861 }, { "epoch": 5.669865642994242, "grad_norm": 1.0072789176851893, "learning_rate": 8.324317611312132e-06, "loss": 0.1182, "step": 8862 }, { "epoch": 5.670505438259757, "grad_norm": 0.8309715773601696, "learning_rate": 8.322274792582609e-06, "loss": 0.0967, "step": 8863 }, { "epoch": 5.671145233525272, "grad_norm": 1.0507233303577517, "learning_rate": 8.320232045891625e-06, "loss": 0.1162, "step": 8864 }, { "epoch": 5.6717850287907865, "grad_norm": 0.9191024533148509, "learning_rate": 8.318189371326889e-06, "loss": 0.1057, "step": 8865 }, { "epoch": 5.672424824056302, "grad_norm": 0.9813147719783655, "learning_rate": 8.316146768976107e-06, "loss": 0.1155, "step": 8866 }, { "epoch": 5.673064619321817, "grad_norm": 0.9932662439421193, "learning_rate": 8.314104238926988e-06, "loss": 0.1178, "step": 8867 }, { "epoch": 5.673704414587332, "grad_norm": 1.0296210665061145, "learning_rate": 8.312061781267235e-06, "loss": 0.1178, "step": 8868 }, { "epoch": 5.674344209852847, "grad_norm": 1.0165003301111244, "learning_rate": 8.310019396084548e-06, "loss": 0.1124, "step": 8869 }, { "epoch": 5.674984005118362, "grad_norm": 1.0140375465540081, "learning_rate": 8.307977083466616e-06, "loss": 0.1145, "step": 8870 }, { "epoch": 5.675623800383877, "grad_norm": 1.0154037000563583, "learning_rate": 8.305934843501143e-06, "loss": 0.1177, "step": 8871 }, { "epoch": 5.6762635956493925, "grad_norm": 1.0439878111933956, "learning_rate": 8.303892676275815e-06, "loss": 0.1221, "step": 8872 }, { "epoch": 5.676903390914907, "grad_norm": 1.0322368584877644, "learning_rate": 8.301850581878315e-06, "loss": 0.1201, "step": 8873 }, { "epoch": 5.677543186180422, "grad_norm": 1.1741488600865828, "learning_rate": 8.299808560396334e-06, "loss": 0.1245, "step": 8874 }, { "epoch": 5.6781829814459375, "grad_norm": 1.0285835016901808, "learning_rate": 8.297766611917547e-06, "loss": 0.1191, "step": 8875 }, { "epoch": 5.678822776711453, "grad_norm": 1.0354595043579744, "learning_rate": 8.295724736529635e-06, "loss": 0.1097, "step": 8876 }, { "epoch": 5.679462571976967, "grad_norm": 1.0808101522444007, "learning_rate": 8.293682934320266e-06, "loss": 0.118, "step": 8877 }, { "epoch": 5.680102367242482, "grad_norm": 1.0390048045147655, "learning_rate": 8.291641205377122e-06, "loss": 0.1219, "step": 8878 }, { "epoch": 5.680742162507998, "grad_norm": 0.9344733396553653, "learning_rate": 8.289599549787864e-06, "loss": 0.1096, "step": 8879 }, { "epoch": 5.681381957773512, "grad_norm": 0.9743838770641966, "learning_rate": 8.287557967640158e-06, "loss": 0.1104, "step": 8880 }, { "epoch": 5.682021753039027, "grad_norm": 1.0171900488382553, "learning_rate": 8.285516459021668e-06, "loss": 0.115, "step": 8881 }, { "epoch": 5.682661548304543, "grad_norm": 1.0478737829522076, "learning_rate": 8.283475024020053e-06, "loss": 0.1215, "step": 8882 }, { "epoch": 5.683301343570058, "grad_norm": 0.9620830051826482, "learning_rate": 8.281433662722966e-06, "loss": 0.1091, "step": 8883 }, { "epoch": 5.683941138835572, "grad_norm": 0.9779590934700113, "learning_rate": 8.279392375218056e-06, "loss": 0.1115, "step": 8884 }, { "epoch": 5.684580934101088, "grad_norm": 1.0407214024324625, "learning_rate": 8.27735116159298e-06, "loss": 0.1254, "step": 8885 }, { "epoch": 5.685220729366603, "grad_norm": 1.0152136788214319, "learning_rate": 8.275310021935381e-06, "loss": 0.1255, "step": 8886 }, { "epoch": 5.685860524632118, "grad_norm": 0.9478063615374123, "learning_rate": 8.2732689563329e-06, "loss": 0.1182, "step": 8887 }, { "epoch": 5.6865003198976325, "grad_norm": 0.9832306142522499, "learning_rate": 8.271227964873182e-06, "loss": 0.1114, "step": 8888 }, { "epoch": 5.687140115163148, "grad_norm": 0.9636482474167609, "learning_rate": 8.269187047643858e-06, "loss": 0.1134, "step": 8889 }, { "epoch": 5.687779910428663, "grad_norm": 0.9294973105907638, "learning_rate": 8.26714620473256e-06, "loss": 0.1128, "step": 8890 }, { "epoch": 5.688419705694178, "grad_norm": 0.9194549792448312, "learning_rate": 8.265105436226925e-06, "loss": 0.1083, "step": 8891 }, { "epoch": 5.689059500959693, "grad_norm": 1.1961741326051043, "learning_rate": 8.263064742214577e-06, "loss": 0.1246, "step": 8892 }, { "epoch": 5.689699296225208, "grad_norm": 1.0366611321567103, "learning_rate": 8.26102412278314e-06, "loss": 0.1107, "step": 8893 }, { "epoch": 5.690339091490723, "grad_norm": 0.9842614584345459, "learning_rate": 8.258983578020228e-06, "loss": 0.1144, "step": 8894 }, { "epoch": 5.690978886756238, "grad_norm": 1.0748488725148566, "learning_rate": 8.256943108013468e-06, "loss": 0.1139, "step": 8895 }, { "epoch": 5.691618682021753, "grad_norm": 0.8978799758959636, "learning_rate": 8.254902712850472e-06, "loss": 0.107, "step": 8896 }, { "epoch": 5.692258477287268, "grad_norm": 1.0956334129179544, "learning_rate": 8.252862392618848e-06, "loss": 0.1127, "step": 8897 }, { "epoch": 5.6928982725527835, "grad_norm": 1.0534241840206768, "learning_rate": 8.250822147406204e-06, "loss": 0.1159, "step": 8898 }, { "epoch": 5.693538067818298, "grad_norm": 0.9443382932579801, "learning_rate": 8.248781977300148e-06, "loss": 0.1085, "step": 8899 }, { "epoch": 5.694177863083813, "grad_norm": 1.0197064885571492, "learning_rate": 8.246741882388273e-06, "loss": 0.118, "step": 8900 }, { "epoch": 5.694817658349328, "grad_norm": 0.9946016870436485, "learning_rate": 8.244701862758187e-06, "loss": 0.1125, "step": 8901 }, { "epoch": 5.695457453614843, "grad_norm": 0.9831262908157846, "learning_rate": 8.242661918497481e-06, "loss": 0.1116, "step": 8902 }, { "epoch": 5.696097248880358, "grad_norm": 0.9370521050592121, "learning_rate": 8.240622049693746e-06, "loss": 0.1107, "step": 8903 }, { "epoch": 5.696737044145873, "grad_norm": 0.9935105020370086, "learning_rate": 8.238582256434568e-06, "loss": 0.1118, "step": 8904 }, { "epoch": 5.697376839411389, "grad_norm": 1.0494080822103122, "learning_rate": 8.236542538807539e-06, "loss": 0.1235, "step": 8905 }, { "epoch": 5.698016634676903, "grad_norm": 1.119503495149436, "learning_rate": 8.234502896900234e-06, "loss": 0.1211, "step": 8906 }, { "epoch": 5.698656429942418, "grad_norm": 1.1109923014229353, "learning_rate": 8.23246333080023e-06, "loss": 0.1196, "step": 8907 }, { "epoch": 5.6992962252079336, "grad_norm": 1.096124483468988, "learning_rate": 8.23042384059511e-06, "loss": 0.124, "step": 8908 }, { "epoch": 5.699936020473449, "grad_norm": 1.074762106280918, "learning_rate": 8.228384426372444e-06, "loss": 0.1266, "step": 8909 }, { "epoch": 5.700575815738963, "grad_norm": 1.0276943884479042, "learning_rate": 8.226345088219797e-06, "loss": 0.108, "step": 8910 }, { "epoch": 5.7012156110044785, "grad_norm": 1.0842648589247632, "learning_rate": 8.224305826224738e-06, "loss": 0.1203, "step": 8911 }, { "epoch": 5.701855406269994, "grad_norm": 1.0831064430313715, "learning_rate": 8.22226664047483e-06, "loss": 0.1235, "step": 8912 }, { "epoch": 5.702495201535509, "grad_norm": 0.9595437105113425, "learning_rate": 8.220227531057628e-06, "loss": 0.1129, "step": 8913 }, { "epoch": 5.703134996801023, "grad_norm": 0.9026632362737, "learning_rate": 8.218188498060686e-06, "loss": 0.112, "step": 8914 }, { "epoch": 5.703774792066539, "grad_norm": 0.9539291504561732, "learning_rate": 8.216149541571565e-06, "loss": 0.1099, "step": 8915 }, { "epoch": 5.704414587332054, "grad_norm": 0.9080656659852192, "learning_rate": 8.21411066167781e-06, "loss": 0.1099, "step": 8916 }, { "epoch": 5.705054382597568, "grad_norm": 0.9912024569828756, "learning_rate": 8.212071858466963e-06, "loss": 0.1194, "step": 8917 }, { "epoch": 5.705694177863084, "grad_norm": 1.147314639525653, "learning_rate": 8.210033132026573e-06, "loss": 0.1245, "step": 8918 }, { "epoch": 5.706333973128599, "grad_norm": 0.9176677329867067, "learning_rate": 8.207994482444177e-06, "loss": 0.1114, "step": 8919 }, { "epoch": 5.706973768394114, "grad_norm": 0.9701457961718705, "learning_rate": 8.205955909807304e-06, "loss": 0.114, "step": 8920 }, { "epoch": 5.707613563659629, "grad_norm": 1.0682874730006058, "learning_rate": 8.203917414203498e-06, "loss": 0.1132, "step": 8921 }, { "epoch": 5.708253358925144, "grad_norm": 0.8434598156629344, "learning_rate": 8.201878995720284e-06, "loss": 0.106, "step": 8922 }, { "epoch": 5.708893154190659, "grad_norm": 1.093311974900632, "learning_rate": 8.199840654445185e-06, "loss": 0.126, "step": 8923 }, { "epoch": 5.709532949456174, "grad_norm": 0.9714605355147652, "learning_rate": 8.197802390465724e-06, "loss": 0.1191, "step": 8924 }, { "epoch": 5.710172744721689, "grad_norm": 1.1701941486338965, "learning_rate": 8.195764203869426e-06, "loss": 0.1115, "step": 8925 }, { "epoch": 5.710812539987204, "grad_norm": 0.9797563761921495, "learning_rate": 8.193726094743803e-06, "loss": 0.1157, "step": 8926 }, { "epoch": 5.711452335252719, "grad_norm": 0.9534607707176825, "learning_rate": 8.191688063176368e-06, "loss": 0.1167, "step": 8927 }, { "epoch": 5.712092130518235, "grad_norm": 0.9580069366555723, "learning_rate": 8.189650109254632e-06, "loss": 0.1103, "step": 8928 }, { "epoch": 5.712731925783749, "grad_norm": 1.0039439034412492, "learning_rate": 8.1876122330661e-06, "loss": 0.1067, "step": 8929 }, { "epoch": 5.713371721049264, "grad_norm": 1.1251094223440241, "learning_rate": 8.18557443469827e-06, "loss": 0.1286, "step": 8930 }, { "epoch": 5.7140115163147795, "grad_norm": 1.087436293524156, "learning_rate": 8.183536714238651e-06, "loss": 0.1148, "step": 8931 }, { "epoch": 5.714651311580294, "grad_norm": 1.0120922050490744, "learning_rate": 8.181499071774735e-06, "loss": 0.1138, "step": 8932 }, { "epoch": 5.715291106845809, "grad_norm": 0.9670401599550079, "learning_rate": 8.179461507394011e-06, "loss": 0.1088, "step": 8933 }, { "epoch": 5.7159309021113245, "grad_norm": 0.9882311039095926, "learning_rate": 8.177424021183972e-06, "loss": 0.1103, "step": 8934 }, { "epoch": 5.71657069737684, "grad_norm": 1.057900876462242, "learning_rate": 8.175386613232104e-06, "loss": 0.119, "step": 8935 }, { "epoch": 5.717210492642354, "grad_norm": 1.0256987476908235, "learning_rate": 8.173349283625892e-06, "loss": 0.1174, "step": 8936 }, { "epoch": 5.717850287907869, "grad_norm": 1.081781362719259, "learning_rate": 8.171312032452806e-06, "loss": 0.1193, "step": 8937 }, { "epoch": 5.718490083173385, "grad_norm": 0.9042253932666323, "learning_rate": 8.169274859800331e-06, "loss": 0.1119, "step": 8938 }, { "epoch": 5.719129878438899, "grad_norm": 1.0110085570863516, "learning_rate": 8.167237765755938e-06, "loss": 0.1167, "step": 8939 }, { "epoch": 5.719769673704414, "grad_norm": 0.9930259341988845, "learning_rate": 8.165200750407094e-06, "loss": 0.117, "step": 8940 }, { "epoch": 5.72040946896993, "grad_norm": 1.1266791980030877, "learning_rate": 8.163163813841269e-06, "loss": 0.1212, "step": 8941 }, { "epoch": 5.721049264235445, "grad_norm": 1.0512327854527017, "learning_rate": 8.161126956145919e-06, "loss": 0.1229, "step": 8942 }, { "epoch": 5.72168905950096, "grad_norm": 1.2569380806358403, "learning_rate": 8.15909017740851e-06, "loss": 0.1404, "step": 8943 }, { "epoch": 5.722328854766475, "grad_norm": 1.0510240297373434, "learning_rate": 8.157053477716487e-06, "loss": 0.1203, "step": 8944 }, { "epoch": 5.72296865003199, "grad_norm": 1.002244793545377, "learning_rate": 8.155016857157315e-06, "loss": 0.1153, "step": 8945 }, { "epoch": 5.723608445297505, "grad_norm": 1.0412467073470189, "learning_rate": 8.152980315818437e-06, "loss": 0.1172, "step": 8946 }, { "epoch": 5.7242482405630195, "grad_norm": 1.1042812745016846, "learning_rate": 8.150943853787296e-06, "loss": 0.1118, "step": 8947 }, { "epoch": 5.724888035828535, "grad_norm": 1.0431737739813631, "learning_rate": 8.14890747115134e-06, "loss": 0.1148, "step": 8948 }, { "epoch": 5.72552783109405, "grad_norm": 1.1179966793880791, "learning_rate": 8.146871167998003e-06, "loss": 0.1212, "step": 8949 }, { "epoch": 5.726167626359565, "grad_norm": 1.0374063556514916, "learning_rate": 8.144834944414718e-06, "loss": 0.1214, "step": 8950 }, { "epoch": 5.72680742162508, "grad_norm": 0.9334493119960315, "learning_rate": 8.142798800488924e-06, "loss": 0.1158, "step": 8951 }, { "epoch": 5.727447216890595, "grad_norm": 1.0562516445391634, "learning_rate": 8.140762736308048e-06, "loss": 0.1135, "step": 8952 }, { "epoch": 5.72808701215611, "grad_norm": 1.1036003103924052, "learning_rate": 8.138726751959511e-06, "loss": 0.1226, "step": 8953 }, { "epoch": 5.728726807421625, "grad_norm": 0.9090061385788163, "learning_rate": 8.136690847530732e-06, "loss": 0.1101, "step": 8954 }, { "epoch": 5.72936660268714, "grad_norm": 1.128416906910522, "learning_rate": 8.134655023109136e-06, "loss": 0.1191, "step": 8955 }, { "epoch": 5.730006397952655, "grad_norm": 1.0074459232237414, "learning_rate": 8.132619278782139e-06, "loss": 0.1109, "step": 8956 }, { "epoch": 5.7306461932181705, "grad_norm": 1.02545907623133, "learning_rate": 8.130583614637144e-06, "loss": 0.1141, "step": 8957 }, { "epoch": 5.731285988483685, "grad_norm": 1.035002136610598, "learning_rate": 8.128548030761565e-06, "loss": 0.1151, "step": 8958 }, { "epoch": 5.7319257837492, "grad_norm": 0.9783357220681893, "learning_rate": 8.126512527242804e-06, "loss": 0.1124, "step": 8959 }, { "epoch": 5.732565579014715, "grad_norm": 1.0128683479684897, "learning_rate": 8.124477104168265e-06, "loss": 0.1107, "step": 8960 }, { "epoch": 5.733205374280231, "grad_norm": 1.0546096027135201, "learning_rate": 8.122441761625337e-06, "loss": 0.114, "step": 8961 }, { "epoch": 5.733845169545745, "grad_norm": 0.987801630254737, "learning_rate": 8.120406499701423e-06, "loss": 0.1184, "step": 8962 }, { "epoch": 5.73448496481126, "grad_norm": 1.1531234071905199, "learning_rate": 8.118371318483911e-06, "loss": 0.1238, "step": 8963 }, { "epoch": 5.735124760076776, "grad_norm": 0.9292487440484222, "learning_rate": 8.116336218060186e-06, "loss": 0.106, "step": 8964 }, { "epoch": 5.735764555342291, "grad_norm": 1.0231108664539503, "learning_rate": 8.114301198517635e-06, "loss": 0.1104, "step": 8965 }, { "epoch": 5.736404350607805, "grad_norm": 0.9282995707190467, "learning_rate": 8.112266259943638e-06, "loss": 0.111, "step": 8966 }, { "epoch": 5.737044145873321, "grad_norm": 1.0146656670752505, "learning_rate": 8.110231402425562e-06, "loss": 0.116, "step": 8967 }, { "epoch": 5.737683941138836, "grad_norm": 0.9858242899163361, "learning_rate": 8.108196626050795e-06, "loss": 0.1207, "step": 8968 }, { "epoch": 5.73832373640435, "grad_norm": 0.9995532360494911, "learning_rate": 8.106161930906697e-06, "loss": 0.1102, "step": 8969 }, { "epoch": 5.7389635316698655, "grad_norm": 1.1017302090447547, "learning_rate": 8.10412731708064e-06, "loss": 0.1218, "step": 8970 }, { "epoch": 5.739603326935381, "grad_norm": 1.1473553557176912, "learning_rate": 8.102092784659978e-06, "loss": 0.1235, "step": 8971 }, { "epoch": 5.740243122200896, "grad_norm": 1.1588227236435666, "learning_rate": 8.10005833373208e-06, "loss": 0.1226, "step": 8972 }, { "epoch": 5.74088291746641, "grad_norm": 1.0523232110244964, "learning_rate": 8.098023964384297e-06, "loss": 0.1214, "step": 8973 }, { "epoch": 5.741522712731926, "grad_norm": 1.0188872203223334, "learning_rate": 8.095989676703976e-06, "loss": 0.1128, "step": 8974 }, { "epoch": 5.742162507997441, "grad_norm": 1.052249735188964, "learning_rate": 8.093955470778479e-06, "loss": 0.1154, "step": 8975 }, { "epoch": 5.742802303262955, "grad_norm": 1.0688758648534127, "learning_rate": 8.091921346695141e-06, "loss": 0.1191, "step": 8976 }, { "epoch": 5.743442098528471, "grad_norm": 1.1700900721097933, "learning_rate": 8.089887304541304e-06, "loss": 0.1357, "step": 8977 }, { "epoch": 5.744081893793986, "grad_norm": 0.9631595964008192, "learning_rate": 8.08785334440431e-06, "loss": 0.1125, "step": 8978 }, { "epoch": 5.744721689059501, "grad_norm": 0.886074113242492, "learning_rate": 8.085819466371492e-06, "loss": 0.1138, "step": 8979 }, { "epoch": 5.7453614843250165, "grad_norm": 0.9780031216322431, "learning_rate": 8.083785670530183e-06, "loss": 0.1184, "step": 8980 }, { "epoch": 5.746001279590531, "grad_norm": 0.9673148776692944, "learning_rate": 8.081751956967702e-06, "loss": 0.1049, "step": 8981 }, { "epoch": 5.746641074856046, "grad_norm": 1.0320039890756698, "learning_rate": 8.079718325771386e-06, "loss": 0.1146, "step": 8982 }, { "epoch": 5.747280870121561, "grad_norm": 0.9716984241023073, "learning_rate": 8.077684777028547e-06, "loss": 0.1178, "step": 8983 }, { "epoch": 5.747920665387076, "grad_norm": 1.1490191648737627, "learning_rate": 8.075651310826503e-06, "loss": 0.126, "step": 8984 }, { "epoch": 5.748560460652591, "grad_norm": 0.9938326952552022, "learning_rate": 8.073617927252571e-06, "loss": 0.1193, "step": 8985 }, { "epoch": 5.749200255918106, "grad_norm": 0.9539521875801484, "learning_rate": 8.071584626394058e-06, "loss": 0.1036, "step": 8986 }, { "epoch": 5.749840051183622, "grad_norm": 1.025539124807392, "learning_rate": 8.069551408338269e-06, "loss": 0.1216, "step": 8987 }, { "epoch": 5.750479846449136, "grad_norm": 0.9847451950157746, "learning_rate": 8.06751827317251e-06, "loss": 0.1164, "step": 8988 }, { "epoch": 5.751119641714651, "grad_norm": 1.056970656364753, "learning_rate": 8.065485220984079e-06, "loss": 0.1159, "step": 8989 }, { "epoch": 5.751759436980167, "grad_norm": 1.0792158555606783, "learning_rate": 8.063452251860273e-06, "loss": 0.1178, "step": 8990 }, { "epoch": 5.752399232245681, "grad_norm": 1.1392121076681043, "learning_rate": 8.061419365888376e-06, "loss": 0.1334, "step": 8991 }, { "epoch": 5.753039027511196, "grad_norm": 0.972419058458258, "learning_rate": 8.059386563155689e-06, "loss": 0.1121, "step": 8992 }, { "epoch": 5.7536788227767115, "grad_norm": 1.0152538022612465, "learning_rate": 8.05735384374949e-06, "loss": 0.1172, "step": 8993 }, { "epoch": 5.754318618042227, "grad_norm": 0.919184436769533, "learning_rate": 8.055321207757058e-06, "loss": 0.1146, "step": 8994 }, { "epoch": 5.754958413307741, "grad_norm": 1.0271122507411423, "learning_rate": 8.05328865526568e-06, "loss": 0.1189, "step": 8995 }, { "epoch": 5.755598208573256, "grad_norm": 1.074297637776255, "learning_rate": 8.05125618636262e-06, "loss": 0.129, "step": 8996 }, { "epoch": 5.756238003838772, "grad_norm": 1.1480780590873547, "learning_rate": 8.04922380113515e-06, "loss": 0.1213, "step": 8997 }, { "epoch": 5.756877799104287, "grad_norm": 0.9478222712757396, "learning_rate": 8.047191499670545e-06, "loss": 0.113, "step": 8998 }, { "epoch": 5.757517594369801, "grad_norm": 0.9964714217741597, "learning_rate": 8.045159282056063e-06, "loss": 0.1143, "step": 8999 }, { "epoch": 5.758157389635317, "grad_norm": 0.9196530835161123, "learning_rate": 8.043127148378962e-06, "loss": 0.1191, "step": 9000 }, { "epoch": 5.758797184900832, "grad_norm": 0.9930821236737386, "learning_rate": 8.0410950987265e-06, "loss": 0.1177, "step": 9001 }, { "epoch": 5.759436980166347, "grad_norm": 0.9996795845017886, "learning_rate": 8.039063133185932e-06, "loss": 0.1214, "step": 9002 }, { "epoch": 5.760076775431862, "grad_norm": 1.0722964185726351, "learning_rate": 8.037031251844504e-06, "loss": 0.1205, "step": 9003 }, { "epoch": 5.760716570697377, "grad_norm": 1.0987932128791242, "learning_rate": 8.034999454789458e-06, "loss": 0.1137, "step": 9004 }, { "epoch": 5.761356365962892, "grad_norm": 1.0714887325991092, "learning_rate": 8.032967742108044e-06, "loss": 0.1234, "step": 9005 }, { "epoch": 5.7619961612284065, "grad_norm": 1.084817018528206, "learning_rate": 8.030936113887498e-06, "loss": 0.1293, "step": 9006 }, { "epoch": 5.762635956493922, "grad_norm": 1.0071516057738341, "learning_rate": 8.028904570215048e-06, "loss": 0.1111, "step": 9007 }, { "epoch": 5.763275751759437, "grad_norm": 0.9652445667458537, "learning_rate": 8.026873111177933e-06, "loss": 0.1124, "step": 9008 }, { "epoch": 5.763915547024952, "grad_norm": 0.9691938206870123, "learning_rate": 8.024841736863375e-06, "loss": 0.1021, "step": 9009 }, { "epoch": 5.764555342290467, "grad_norm": 0.9720372011764495, "learning_rate": 8.022810447358602e-06, "loss": 0.1163, "step": 9010 }, { "epoch": 5.765195137555982, "grad_norm": 0.9851658949843173, "learning_rate": 8.020779242750825e-06, "loss": 0.1126, "step": 9011 }, { "epoch": 5.765834932821497, "grad_norm": 0.9893072562911477, "learning_rate": 8.018748123127272e-06, "loss": 0.1194, "step": 9012 }, { "epoch": 5.766474728087012, "grad_norm": 0.9982470515120623, "learning_rate": 8.016717088575149e-06, "loss": 0.1072, "step": 9013 }, { "epoch": 5.767114523352527, "grad_norm": 0.9153898796877853, "learning_rate": 8.014686139181662e-06, "loss": 0.1123, "step": 9014 }, { "epoch": 5.767754318618042, "grad_norm": 1.097988185548743, "learning_rate": 8.012655275034026e-06, "loss": 0.1133, "step": 9015 }, { "epoch": 5.7683941138835575, "grad_norm": 0.9827776926000312, "learning_rate": 8.010624496219437e-06, "loss": 0.1182, "step": 9016 }, { "epoch": 5.769033909149073, "grad_norm": 1.0736658110067188, "learning_rate": 8.00859380282509e-06, "loss": 0.1179, "step": 9017 }, { "epoch": 5.769673704414587, "grad_norm": 1.004885186832513, "learning_rate": 8.006563194938186e-06, "loss": 0.1043, "step": 9018 }, { "epoch": 5.770313499680102, "grad_norm": 1.1148638607969348, "learning_rate": 8.004532672645913e-06, "loss": 0.1169, "step": 9019 }, { "epoch": 5.770953294945618, "grad_norm": 0.9824892673277372, "learning_rate": 8.002502236035455e-06, "loss": 0.1139, "step": 9020 }, { "epoch": 5.771593090211132, "grad_norm": 0.9366517554005326, "learning_rate": 8.000471885193995e-06, "loss": 0.1249, "step": 9021 }, { "epoch": 5.772232885476647, "grad_norm": 1.0090901709585762, "learning_rate": 7.99844162020872e-06, "loss": 0.1161, "step": 9022 }, { "epoch": 5.772872680742163, "grad_norm": 0.9974284924395067, "learning_rate": 7.9964114411668e-06, "loss": 0.1156, "step": 9023 }, { "epoch": 5.773512476007678, "grad_norm": 0.9299079530731558, "learning_rate": 7.994381348155407e-06, "loss": 0.1166, "step": 9024 }, { "epoch": 5.774152271273192, "grad_norm": 1.064652174447242, "learning_rate": 7.992351341261715e-06, "loss": 0.1191, "step": 9025 }, { "epoch": 5.774792066538708, "grad_norm": 0.9471974383989076, "learning_rate": 7.990321420572885e-06, "loss": 0.1195, "step": 9026 }, { "epoch": 5.775431861804223, "grad_norm": 0.9571436129557181, "learning_rate": 7.988291586176071e-06, "loss": 0.1125, "step": 9027 }, { "epoch": 5.776071657069737, "grad_norm": 1.0031610439529273, "learning_rate": 7.986261838158443e-06, "loss": 0.1181, "step": 9028 }, { "epoch": 5.7767114523352525, "grad_norm": 0.9166167692350073, "learning_rate": 7.98423217660715e-06, "loss": 0.1135, "step": 9029 }, { "epoch": 5.777351247600768, "grad_norm": 0.9752561834268949, "learning_rate": 7.982202601609342e-06, "loss": 0.116, "step": 9030 }, { "epoch": 5.777991042866283, "grad_norm": 1.085051931581387, "learning_rate": 7.980173113252164e-06, "loss": 0.1272, "step": 9031 }, { "epoch": 5.7786308381317975, "grad_norm": 0.9385946303860556, "learning_rate": 7.978143711622759e-06, "loss": 0.1114, "step": 9032 }, { "epoch": 5.779270633397313, "grad_norm": 1.0036768841510135, "learning_rate": 7.976114396808268e-06, "loss": 0.111, "step": 9033 }, { "epoch": 5.779910428662828, "grad_norm": 1.1376804380792542, "learning_rate": 7.97408516889582e-06, "loss": 0.127, "step": 9034 }, { "epoch": 5.780550223928343, "grad_norm": 1.0893726994941855, "learning_rate": 7.972056027972554e-06, "loss": 0.1216, "step": 9035 }, { "epoch": 5.781190019193858, "grad_norm": 0.9239666335025397, "learning_rate": 7.970026974125597e-06, "loss": 0.1112, "step": 9036 }, { "epoch": 5.781829814459373, "grad_norm": 1.1194062115880148, "learning_rate": 7.96799800744207e-06, "loss": 0.1232, "step": 9037 }, { "epoch": 5.782469609724888, "grad_norm": 0.9849140760663979, "learning_rate": 7.96596912800909e-06, "loss": 0.1102, "step": 9038 }, { "epoch": 5.7831094049904035, "grad_norm": 1.0201451869988867, "learning_rate": 7.963940335913783e-06, "loss": 0.1128, "step": 9039 }, { "epoch": 5.783749200255918, "grad_norm": 1.054889342756365, "learning_rate": 7.961911631243253e-06, "loss": 0.115, "step": 9040 }, { "epoch": 5.784388995521433, "grad_norm": 0.9074482677007938, "learning_rate": 7.959883014084608e-06, "loss": 0.1065, "step": 9041 }, { "epoch": 5.785028790786948, "grad_norm": 1.032990488110114, "learning_rate": 7.957854484524963e-06, "loss": 0.1222, "step": 9042 }, { "epoch": 5.785668586052463, "grad_norm": 1.0530962584021426, "learning_rate": 7.955826042651413e-06, "loss": 0.1212, "step": 9043 }, { "epoch": 5.786308381317978, "grad_norm": 1.0012484445383114, "learning_rate": 7.953797688551053e-06, "loss": 0.1189, "step": 9044 }, { "epoch": 5.786948176583493, "grad_norm": 1.0073441451517136, "learning_rate": 7.951769422310984e-06, "loss": 0.1154, "step": 9045 }, { "epoch": 5.787587971849009, "grad_norm": 1.0270496425685058, "learning_rate": 7.949741244018291e-06, "loss": 0.1183, "step": 9046 }, { "epoch": 5.788227767114523, "grad_norm": 0.9864729997966222, "learning_rate": 7.947713153760064e-06, "loss": 0.1025, "step": 9047 }, { "epoch": 5.788867562380038, "grad_norm": 1.0416702542552496, "learning_rate": 7.945685151623382e-06, "loss": 0.1208, "step": 9048 }, { "epoch": 5.789507357645554, "grad_norm": 1.0291899785827914, "learning_rate": 7.943657237695325e-06, "loss": 0.1178, "step": 9049 }, { "epoch": 5.790147152911068, "grad_norm": 0.9885160775473333, "learning_rate": 7.941629412062968e-06, "loss": 0.1114, "step": 9050 }, { "epoch": 5.790786948176583, "grad_norm": 0.9884301105870192, "learning_rate": 7.93960167481338e-06, "loss": 0.1273, "step": 9051 }, { "epoch": 5.7914267434420985, "grad_norm": 0.9469985268996232, "learning_rate": 7.937574026033636e-06, "loss": 0.1035, "step": 9052 }, { "epoch": 5.792066538707614, "grad_norm": 1.1301036526772206, "learning_rate": 7.935546465810793e-06, "loss": 0.1273, "step": 9053 }, { "epoch": 5.792706333973129, "grad_norm": 0.991535363650778, "learning_rate": 7.93351899423191e-06, "loss": 0.1198, "step": 9054 }, { "epoch": 5.7933461292386434, "grad_norm": 0.9904095975374655, "learning_rate": 7.931491611384049e-06, "loss": 0.1084, "step": 9055 }, { "epoch": 5.793985924504159, "grad_norm": 0.9100976469218646, "learning_rate": 7.929464317354256e-06, "loss": 0.1128, "step": 9056 }, { "epoch": 5.794625719769674, "grad_norm": 0.910018351085235, "learning_rate": 7.927437112229583e-06, "loss": 0.1106, "step": 9057 }, { "epoch": 5.795265515035188, "grad_norm": 1.1759394449572804, "learning_rate": 7.925409996097071e-06, "loss": 0.1248, "step": 9058 }, { "epoch": 5.795905310300704, "grad_norm": 1.0217468390842759, "learning_rate": 7.923382969043766e-06, "loss": 0.1166, "step": 9059 }, { "epoch": 5.796545105566219, "grad_norm": 1.0129861586212956, "learning_rate": 7.921356031156702e-06, "loss": 0.1135, "step": 9060 }, { "epoch": 5.797184900831734, "grad_norm": 0.9916420026107055, "learning_rate": 7.919329182522913e-06, "loss": 0.114, "step": 9061 }, { "epoch": 5.797824696097249, "grad_norm": 0.9638747742129178, "learning_rate": 7.917302423229428e-06, "loss": 0.1177, "step": 9062 }, { "epoch": 5.798464491362764, "grad_norm": 1.0026497231193068, "learning_rate": 7.91527575336327e-06, "loss": 0.1167, "step": 9063 }, { "epoch": 5.799104286628279, "grad_norm": 1.0197699266471052, "learning_rate": 7.913249173011461e-06, "loss": 0.1143, "step": 9064 }, { "epoch": 5.7997440818937935, "grad_norm": 1.0527542037008744, "learning_rate": 7.911222682261024e-06, "loss": 0.1153, "step": 9065 }, { "epoch": 5.800383877159309, "grad_norm": 0.9211023676276867, "learning_rate": 7.90919628119897e-06, "loss": 0.1101, "step": 9066 }, { "epoch": 5.801023672424824, "grad_norm": 1.0799042631616385, "learning_rate": 7.907169969912308e-06, "loss": 0.1304, "step": 9067 }, { "epoch": 5.801663467690339, "grad_norm": 0.9894567240328371, "learning_rate": 7.905143748488042e-06, "loss": 0.1147, "step": 9068 }, { "epoch": 5.802303262955854, "grad_norm": 1.034275177832266, "learning_rate": 7.90311761701318e-06, "loss": 0.1244, "step": 9069 }, { "epoch": 5.802943058221369, "grad_norm": 0.9754168827253926, "learning_rate": 7.901091575574717e-06, "loss": 0.1185, "step": 9070 }, { "epoch": 5.803582853486884, "grad_norm": 1.0178443326534328, "learning_rate": 7.899065624259645e-06, "loss": 0.1218, "step": 9071 }, { "epoch": 5.8042226487524, "grad_norm": 1.0918740538059752, "learning_rate": 7.89703976315496e-06, "loss": 0.1142, "step": 9072 }, { "epoch": 5.804862444017914, "grad_norm": 1.058901132997107, "learning_rate": 7.895013992347648e-06, "loss": 0.1179, "step": 9073 }, { "epoch": 5.805502239283429, "grad_norm": 1.0344498728161184, "learning_rate": 7.892988311924685e-06, "loss": 0.1117, "step": 9074 }, { "epoch": 5.8061420345489445, "grad_norm": 1.0126319949102784, "learning_rate": 7.890962721973061e-06, "loss": 0.1173, "step": 9075 }, { "epoch": 5.80678182981446, "grad_norm": 0.9509045652184694, "learning_rate": 7.888937222579746e-06, "loss": 0.1124, "step": 9076 }, { "epoch": 5.807421625079974, "grad_norm": 1.1138441617435606, "learning_rate": 7.886911813831713e-06, "loss": 0.1159, "step": 9077 }, { "epoch": 5.8080614203454894, "grad_norm": 0.937429966092186, "learning_rate": 7.884886495815924e-06, "loss": 0.1171, "step": 9078 }, { "epoch": 5.808701215611005, "grad_norm": 1.0106075333377709, "learning_rate": 7.882861268619349e-06, "loss": 0.1082, "step": 9079 }, { "epoch": 5.809341010876519, "grad_norm": 0.9512921957404648, "learning_rate": 7.880836132328945e-06, "loss": 0.1145, "step": 9080 }, { "epoch": 5.809980806142034, "grad_norm": 1.0286688227513243, "learning_rate": 7.878811087031663e-06, "loss": 0.1155, "step": 9081 }, { "epoch": 5.81062060140755, "grad_norm": 0.9736469630142376, "learning_rate": 7.876786132814465e-06, "loss": 0.1088, "step": 9082 }, { "epoch": 5.811260396673065, "grad_norm": 1.0129202298605395, "learning_rate": 7.874761269764291e-06, "loss": 0.1154, "step": 9083 }, { "epoch": 5.811900191938579, "grad_norm": 0.9708123939762724, "learning_rate": 7.872736497968088e-06, "loss": 0.118, "step": 9084 }, { "epoch": 5.812539987204095, "grad_norm": 0.9465537074401815, "learning_rate": 7.870711817512798e-06, "loss": 0.1152, "step": 9085 }, { "epoch": 5.81317978246961, "grad_norm": 1.065350690768018, "learning_rate": 7.868687228485353e-06, "loss": 0.1163, "step": 9086 }, { "epoch": 5.813819577735125, "grad_norm": 1.0022726216828655, "learning_rate": 7.866662730972687e-06, "loss": 0.1171, "step": 9087 }, { "epoch": 5.8144593730006395, "grad_norm": 0.9104955960497334, "learning_rate": 7.864638325061726e-06, "loss": 0.1072, "step": 9088 }, { "epoch": 5.815099168266155, "grad_norm": 0.911640290606244, "learning_rate": 7.8626140108394e-06, "loss": 0.1148, "step": 9089 }, { "epoch": 5.81573896353167, "grad_norm": 0.9075403117160641, "learning_rate": 7.860589788392626e-06, "loss": 0.1101, "step": 9090 }, { "epoch": 5.816378758797185, "grad_norm": 0.9705509658522221, "learning_rate": 7.858565657808318e-06, "loss": 0.1207, "step": 9091 }, { "epoch": 5.8170185540627, "grad_norm": 0.9577327971465832, "learning_rate": 7.856541619173396e-06, "loss": 0.1078, "step": 9092 }, { "epoch": 5.817658349328215, "grad_norm": 0.8709439948796396, "learning_rate": 7.854517672574762e-06, "loss": 0.1143, "step": 9093 }, { "epoch": 5.81829814459373, "grad_norm": 1.062993153505256, "learning_rate": 7.852493818099319e-06, "loss": 0.1193, "step": 9094 }, { "epoch": 5.818937939859245, "grad_norm": 0.9653944104273453, "learning_rate": 7.850470055833976e-06, "loss": 0.1136, "step": 9095 }, { "epoch": 5.81957773512476, "grad_norm": 1.030252690262487, "learning_rate": 7.848446385865624e-06, "loss": 0.1158, "step": 9096 }, { "epoch": 5.820217530390275, "grad_norm": 1.0111880489783938, "learning_rate": 7.846422808281158e-06, "loss": 0.1118, "step": 9097 }, { "epoch": 5.8208573256557905, "grad_norm": 0.9633461887992782, "learning_rate": 7.844399323167462e-06, "loss": 0.1113, "step": 9098 }, { "epoch": 5.821497120921305, "grad_norm": 0.9886208015518104, "learning_rate": 7.842375930611428e-06, "loss": 0.1166, "step": 9099 }, { "epoch": 5.82213691618682, "grad_norm": 1.0273991162133476, "learning_rate": 7.840352630699934e-06, "loss": 0.117, "step": 9100 }, { "epoch": 5.822776711452335, "grad_norm": 0.8887454243368081, "learning_rate": 7.83832942351985e-06, "loss": 0.1101, "step": 9101 }, { "epoch": 5.82341650671785, "grad_norm": 1.0303557186455454, "learning_rate": 7.836306309158062e-06, "loss": 0.1135, "step": 9102 }, { "epoch": 5.824056301983365, "grad_norm": 1.0522606288875307, "learning_rate": 7.83428328770143e-06, "loss": 0.1179, "step": 9103 }, { "epoch": 5.82469609724888, "grad_norm": 1.041191899134558, "learning_rate": 7.832260359236817e-06, "loss": 0.1132, "step": 9104 }, { "epoch": 5.825335892514396, "grad_norm": 1.0216545481803103, "learning_rate": 7.830237523851093e-06, "loss": 0.119, "step": 9105 }, { "epoch": 5.825975687779911, "grad_norm": 0.9560798657996878, "learning_rate": 7.82821478163111e-06, "loss": 0.1141, "step": 9106 }, { "epoch": 5.826615483045425, "grad_norm": 0.9621284376542798, "learning_rate": 7.826192132663721e-06, "loss": 0.1216, "step": 9107 }, { "epoch": 5.827255278310941, "grad_norm": 1.0664360633717496, "learning_rate": 7.824169577035773e-06, "loss": 0.1177, "step": 9108 }, { "epoch": 5.827895073576456, "grad_norm": 0.9634873758374867, "learning_rate": 7.822147114834116e-06, "loss": 0.1156, "step": 9109 }, { "epoch": 5.82853486884197, "grad_norm": 1.007221789610578, "learning_rate": 7.820124746145586e-06, "loss": 0.1142, "step": 9110 }, { "epoch": 5.8291746641074855, "grad_norm": 1.0452911839606394, "learning_rate": 7.818102471057017e-06, "loss": 0.1137, "step": 9111 }, { "epoch": 5.829814459373001, "grad_norm": 0.9412505587608914, "learning_rate": 7.816080289655252e-06, "loss": 0.1192, "step": 9112 }, { "epoch": 5.830454254638516, "grad_norm": 1.0659731950103846, "learning_rate": 7.814058202027114e-06, "loss": 0.1219, "step": 9113 }, { "epoch": 5.8310940499040305, "grad_norm": 0.9476668608929077, "learning_rate": 7.81203620825943e-06, "loss": 0.1131, "step": 9114 }, { "epoch": 5.831733845169546, "grad_norm": 0.913530206820743, "learning_rate": 7.810014308439014e-06, "loss": 0.1104, "step": 9115 }, { "epoch": 5.832373640435061, "grad_norm": 0.90794226530418, "learning_rate": 7.807992502652692e-06, "loss": 0.1145, "step": 9116 }, { "epoch": 5.833013435700575, "grad_norm": 1.1530064015117054, "learning_rate": 7.805970790987273e-06, "loss": 0.1228, "step": 9117 }, { "epoch": 5.833653230966091, "grad_norm": 0.992569999650312, "learning_rate": 7.80394917352956e-06, "loss": 0.1187, "step": 9118 }, { "epoch": 5.834293026231606, "grad_norm": 0.8641032157502649, "learning_rate": 7.801927650366367e-06, "loss": 0.1001, "step": 9119 }, { "epoch": 5.834932821497121, "grad_norm": 1.019410297009878, "learning_rate": 7.799906221584493e-06, "loss": 0.1175, "step": 9120 }, { "epoch": 5.835572616762636, "grad_norm": 0.9651474594928008, "learning_rate": 7.797884887270727e-06, "loss": 0.1107, "step": 9121 }, { "epoch": 5.836212412028151, "grad_norm": 0.9235010182727807, "learning_rate": 7.795863647511871e-06, "loss": 0.1164, "step": 9122 }, { "epoch": 5.836852207293666, "grad_norm": 0.9322071717346037, "learning_rate": 7.793842502394709e-06, "loss": 0.1095, "step": 9123 }, { "epoch": 5.837492002559181, "grad_norm": 0.8833690273436702, "learning_rate": 7.791821452006025e-06, "loss": 0.1067, "step": 9124 }, { "epoch": 5.838131797824696, "grad_norm": 0.9861467724062547, "learning_rate": 7.789800496432595e-06, "loss": 0.1141, "step": 9125 }, { "epoch": 5.838771593090211, "grad_norm": 0.9714854744240711, "learning_rate": 7.787779635761203e-06, "loss": 0.1133, "step": 9126 }, { "epoch": 5.839411388355726, "grad_norm": 0.9344914877193888, "learning_rate": 7.785758870078621e-06, "loss": 0.1049, "step": 9127 }, { "epoch": 5.840051183621242, "grad_norm": 0.9826227249253915, "learning_rate": 7.783738199471611e-06, "loss": 0.1157, "step": 9128 }, { "epoch": 5.840690978886756, "grad_norm": 0.914756559311995, "learning_rate": 7.781717624026942e-06, "loss": 0.1065, "step": 9129 }, { "epoch": 5.841330774152271, "grad_norm": 1.0229577098843805, "learning_rate": 7.779697143831373e-06, "loss": 0.1197, "step": 9130 }, { "epoch": 5.841970569417787, "grad_norm": 0.9229406971583288, "learning_rate": 7.777676758971654e-06, "loss": 0.1138, "step": 9131 }, { "epoch": 5.842610364683301, "grad_norm": 1.031343320877659, "learning_rate": 7.775656469534548e-06, "loss": 0.1261, "step": 9132 }, { "epoch": 5.843250159948816, "grad_norm": 1.037767795021841, "learning_rate": 7.773636275606794e-06, "loss": 0.119, "step": 9133 }, { "epoch": 5.8438899552143315, "grad_norm": 1.081609454394017, "learning_rate": 7.771616177275138e-06, "loss": 0.1203, "step": 9134 }, { "epoch": 5.844529750479847, "grad_norm": 1.0739374342683419, "learning_rate": 7.769596174626316e-06, "loss": 0.121, "step": 9135 }, { "epoch": 5.845169545745361, "grad_norm": 0.9988237897402209, "learning_rate": 7.767576267747072e-06, "loss": 0.1091, "step": 9136 }, { "epoch": 5.8458093410108765, "grad_norm": 0.9789856776189461, "learning_rate": 7.76555645672413e-06, "loss": 0.115, "step": 9137 }, { "epoch": 5.846449136276392, "grad_norm": 1.0236133213800847, "learning_rate": 7.763536741644218e-06, "loss": 0.1122, "step": 9138 }, { "epoch": 5.847088931541906, "grad_norm": 0.9701975077332627, "learning_rate": 7.761517122594062e-06, "loss": 0.1191, "step": 9139 }, { "epoch": 5.847728726807421, "grad_norm": 1.0689085317768907, "learning_rate": 7.759497599660378e-06, "loss": 0.1175, "step": 9140 }, { "epoch": 5.848368522072937, "grad_norm": 1.186044167082316, "learning_rate": 7.757478172929877e-06, "loss": 0.1218, "step": 9141 }, { "epoch": 5.849008317338452, "grad_norm": 0.9669045536335954, "learning_rate": 7.75545884248928e-06, "loss": 0.1115, "step": 9142 }, { "epoch": 5.849648112603967, "grad_norm": 0.9941045941594588, "learning_rate": 7.753439608425287e-06, "loss": 0.1238, "step": 9143 }, { "epoch": 5.850287907869482, "grad_norm": 1.000312875448517, "learning_rate": 7.7514204708246e-06, "loss": 0.1166, "step": 9144 }, { "epoch": 5.850927703134997, "grad_norm": 1.0131228123245315, "learning_rate": 7.749401429773916e-06, "loss": 0.1146, "step": 9145 }, { "epoch": 5.851567498400512, "grad_norm": 0.9783145091607844, "learning_rate": 7.747382485359934e-06, "loss": 0.1231, "step": 9146 }, { "epoch": 5.8522072936660265, "grad_norm": 1.042467578740683, "learning_rate": 7.745363637669341e-06, "loss": 0.1123, "step": 9147 }, { "epoch": 5.852847088931542, "grad_norm": 1.002652560918847, "learning_rate": 7.743344886788818e-06, "loss": 0.1132, "step": 9148 }, { "epoch": 5.853486884197057, "grad_norm": 0.9708731198780243, "learning_rate": 7.741326232805055e-06, "loss": 0.1061, "step": 9149 }, { "epoch": 5.854126679462572, "grad_norm": 1.0248849366364863, "learning_rate": 7.739307675804727e-06, "loss": 0.126, "step": 9150 }, { "epoch": 5.854766474728087, "grad_norm": 0.923271754581552, "learning_rate": 7.737289215874503e-06, "loss": 0.1105, "step": 9151 }, { "epoch": 5.855406269993602, "grad_norm": 1.0374474037307737, "learning_rate": 7.735270853101057e-06, "loss": 0.1159, "step": 9152 }, { "epoch": 5.856046065259117, "grad_norm": 0.9881948708460825, "learning_rate": 7.733252587571053e-06, "loss": 0.1152, "step": 9153 }, { "epoch": 5.856685860524632, "grad_norm": 0.9759958911505414, "learning_rate": 7.73123441937115e-06, "loss": 0.1136, "step": 9154 }, { "epoch": 5.857325655790147, "grad_norm": 1.1494697107025709, "learning_rate": 7.729216348588e-06, "loss": 0.1242, "step": 9155 }, { "epoch": 5.857965451055662, "grad_norm": 0.8948731790144776, "learning_rate": 7.727198375308265e-06, "loss": 0.1116, "step": 9156 }, { "epoch": 5.8586052463211775, "grad_norm": 1.0921893695078022, "learning_rate": 7.725180499618591e-06, "loss": 0.1188, "step": 9157 }, { "epoch": 5.859245041586692, "grad_norm": 1.1194190416848826, "learning_rate": 7.723162721605616e-06, "loss": 0.1213, "step": 9158 }, { "epoch": 5.859884836852207, "grad_norm": 0.9762039779597126, "learning_rate": 7.721145041355985e-06, "loss": 0.1121, "step": 9159 }, { "epoch": 5.8605246321177225, "grad_norm": 1.1191817876376204, "learning_rate": 7.719127458956333e-06, "loss": 0.1188, "step": 9160 }, { "epoch": 5.861164427383238, "grad_norm": 0.8816110094904037, "learning_rate": 7.717109974493287e-06, "loss": 0.1092, "step": 9161 }, { "epoch": 5.861804222648752, "grad_norm": 0.9486487390252865, "learning_rate": 7.71509258805348e-06, "loss": 0.1144, "step": 9162 }, { "epoch": 5.862444017914267, "grad_norm": 0.9420326148781463, "learning_rate": 7.713075299723534e-06, "loss": 0.1128, "step": 9163 }, { "epoch": 5.863083813179783, "grad_norm": 1.0058360941895206, "learning_rate": 7.711058109590068e-06, "loss": 0.1109, "step": 9164 }, { "epoch": 5.863723608445298, "grad_norm": 0.9052148127420432, "learning_rate": 7.70904101773969e-06, "loss": 0.1155, "step": 9165 }, { "epoch": 5.864363403710812, "grad_norm": 0.9767928629115125, "learning_rate": 7.707024024259018e-06, "loss": 0.1122, "step": 9166 }, { "epoch": 5.865003198976328, "grad_norm": 0.954940328138921, "learning_rate": 7.705007129234657e-06, "loss": 0.1085, "step": 9167 }, { "epoch": 5.865642994241843, "grad_norm": 0.9463849012318979, "learning_rate": 7.702990332753204e-06, "loss": 0.1232, "step": 9168 }, { "epoch": 5.866282789507357, "grad_norm": 1.160530548560919, "learning_rate": 7.700973634901263e-06, "loss": 0.1217, "step": 9169 }, { "epoch": 5.8669225847728725, "grad_norm": 0.9290028484739467, "learning_rate": 7.698957035765423e-06, "loss": 0.1201, "step": 9170 }, { "epoch": 5.867562380038388, "grad_norm": 1.0351989433896875, "learning_rate": 7.696940535432274e-06, "loss": 0.112, "step": 9171 }, { "epoch": 5.868202175303903, "grad_norm": 0.9232825325867651, "learning_rate": 7.694924133988404e-06, "loss": 0.1159, "step": 9172 }, { "epoch": 5.8688419705694175, "grad_norm": 1.0242220418916603, "learning_rate": 7.69290783152039e-06, "loss": 0.1183, "step": 9173 }, { "epoch": 5.869481765834933, "grad_norm": 0.9208052271268177, "learning_rate": 7.69089162811481e-06, "loss": 0.1046, "step": 9174 }, { "epoch": 5.870121561100448, "grad_norm": 0.9782873420771451, "learning_rate": 7.688875523858235e-06, "loss": 0.1175, "step": 9175 }, { "epoch": 5.870761356365962, "grad_norm": 0.8924539419286697, "learning_rate": 7.686859518837235e-06, "loss": 0.1096, "step": 9176 }, { "epoch": 5.871401151631478, "grad_norm": 0.9784109623805368, "learning_rate": 7.684843613138372e-06, "loss": 0.1175, "step": 9177 }, { "epoch": 5.872040946896993, "grad_norm": 0.9570599228011406, "learning_rate": 7.682827806848203e-06, "loss": 0.1085, "step": 9178 }, { "epoch": 5.872680742162508, "grad_norm": 1.0515531587272593, "learning_rate": 7.680812100053289e-06, "loss": 0.1138, "step": 9179 }, { "epoch": 5.8733205374280235, "grad_norm": 0.9449645961568001, "learning_rate": 7.67879649284018e-06, "loss": 0.1079, "step": 9180 }, { "epoch": 5.873960332693538, "grad_norm": 1.06764315295465, "learning_rate": 7.676780985295417e-06, "loss": 0.1129, "step": 9181 }, { "epoch": 5.874600127959053, "grad_norm": 0.939966774369353, "learning_rate": 7.674765577505548e-06, "loss": 0.1157, "step": 9182 }, { "epoch": 5.8752399232245685, "grad_norm": 1.0408738930319035, "learning_rate": 7.67275026955711e-06, "loss": 0.1177, "step": 9183 }, { "epoch": 5.875879718490083, "grad_norm": 0.9715182883784048, "learning_rate": 7.670735061536634e-06, "loss": 0.1125, "step": 9184 }, { "epoch": 5.876519513755598, "grad_norm": 1.0056145241577894, "learning_rate": 7.668719953530647e-06, "loss": 0.1157, "step": 9185 }, { "epoch": 5.877159309021113, "grad_norm": 1.0073854918945726, "learning_rate": 7.666704945625683e-06, "loss": 0.1269, "step": 9186 }, { "epoch": 5.877799104286629, "grad_norm": 1.08704352774192, "learning_rate": 7.664690037908257e-06, "loss": 0.1157, "step": 9187 }, { "epoch": 5.878438899552143, "grad_norm": 0.9457693099701157, "learning_rate": 7.662675230464885e-06, "loss": 0.1125, "step": 9188 }, { "epoch": 5.879078694817658, "grad_norm": 1.1565804306072875, "learning_rate": 7.660660523382083e-06, "loss": 0.1239, "step": 9189 }, { "epoch": 5.879718490083174, "grad_norm": 0.9649671040347777, "learning_rate": 7.658645916746357e-06, "loss": 0.1069, "step": 9190 }, { "epoch": 5.880358285348688, "grad_norm": 1.0394892815718657, "learning_rate": 7.656631410644209e-06, "loss": 0.1186, "step": 9191 }, { "epoch": 5.880998080614203, "grad_norm": 1.0213168910227155, "learning_rate": 7.654617005162137e-06, "loss": 0.1197, "step": 9192 }, { "epoch": 5.8816378758797185, "grad_norm": 1.014674029456163, "learning_rate": 7.652602700386644e-06, "loss": 0.1101, "step": 9193 }, { "epoch": 5.882277671145234, "grad_norm": 0.9914653076149416, "learning_rate": 7.650588496404212e-06, "loss": 0.1134, "step": 9194 }, { "epoch": 5.882917466410748, "grad_norm": 1.0143536286213484, "learning_rate": 7.648574393301328e-06, "loss": 0.111, "step": 9195 }, { "epoch": 5.8835572616762635, "grad_norm": 0.882280529295511, "learning_rate": 7.646560391164483e-06, "loss": 0.1091, "step": 9196 }, { "epoch": 5.884197056941779, "grad_norm": 0.9727075354356709, "learning_rate": 7.644546490080146e-06, "loss": 0.1137, "step": 9197 }, { "epoch": 5.884836852207294, "grad_norm": 1.1015176486621594, "learning_rate": 7.642532690134792e-06, "loss": 0.1235, "step": 9198 }, { "epoch": 5.885476647472808, "grad_norm": 1.0471162598425587, "learning_rate": 7.640518991414892e-06, "loss": 0.1149, "step": 9199 }, { "epoch": 5.886116442738324, "grad_norm": 0.9929807100500128, "learning_rate": 7.638505394006908e-06, "loss": 0.114, "step": 9200 }, { "epoch": 5.886756238003839, "grad_norm": 1.0224971486999233, "learning_rate": 7.636491897997303e-06, "loss": 0.1137, "step": 9201 }, { "epoch": 5.887396033269354, "grad_norm": 1.1125813839441017, "learning_rate": 7.634478503472527e-06, "loss": 0.1188, "step": 9202 }, { "epoch": 5.888035828534869, "grad_norm": 0.9394625984191428, "learning_rate": 7.63246521051904e-06, "loss": 0.1096, "step": 9203 }, { "epoch": 5.888675623800384, "grad_norm": 0.9635380554899555, "learning_rate": 7.630452019223289e-06, "loss": 0.1113, "step": 9204 }, { "epoch": 5.889315419065899, "grad_norm": 1.0663629017894116, "learning_rate": 7.628438929671706e-06, "loss": 0.1208, "step": 9205 }, { "epoch": 5.889955214331414, "grad_norm": 0.8215895082852951, "learning_rate": 7.6264259419507415e-06, "loss": 0.1081, "step": 9206 }, { "epoch": 5.890595009596929, "grad_norm": 1.0961981898048363, "learning_rate": 7.624413056146824e-06, "loss": 0.1236, "step": 9207 }, { "epoch": 5.891234804862444, "grad_norm": 0.8978802191423342, "learning_rate": 7.62240027234638e-06, "loss": 0.101, "step": 9208 }, { "epoch": 5.891874600127959, "grad_norm": 1.031151603521835, "learning_rate": 7.620387590635843e-06, "loss": 0.1143, "step": 9209 }, { "epoch": 5.892514395393474, "grad_norm": 0.8242682575707692, "learning_rate": 7.618375011101629e-06, "loss": 0.1083, "step": 9210 }, { "epoch": 5.893154190658989, "grad_norm": 0.9391185494523253, "learning_rate": 7.6163625338301575e-06, "loss": 0.1134, "step": 9211 }, { "epoch": 5.893793985924504, "grad_norm": 0.9722592878021433, "learning_rate": 7.614350158907835e-06, "loss": 0.1115, "step": 9212 }, { "epoch": 5.894433781190019, "grad_norm": 0.9954882642955254, "learning_rate": 7.612337886421075e-06, "loss": 0.1027, "step": 9213 }, { "epoch": 5.895073576455534, "grad_norm": 1.0299973541534695, "learning_rate": 7.610325716456279e-06, "loss": 0.1153, "step": 9214 }, { "epoch": 5.895713371721049, "grad_norm": 0.9987681205345356, "learning_rate": 7.608313649099842e-06, "loss": 0.1144, "step": 9215 }, { "epoch": 5.8963531669865645, "grad_norm": 0.9474346098976615, "learning_rate": 7.606301684438167e-06, "loss": 0.1185, "step": 9216 }, { "epoch": 5.89699296225208, "grad_norm": 1.0064134022457702, "learning_rate": 7.604289822557639e-06, "loss": 0.1134, "step": 9217 }, { "epoch": 5.897632757517594, "grad_norm": 1.1162789848928916, "learning_rate": 7.602278063544643e-06, "loss": 0.1118, "step": 9218 }, { "epoch": 5.8982725527831095, "grad_norm": 0.9639597207455529, "learning_rate": 7.600266407485562e-06, "loss": 0.1129, "step": 9219 }, { "epoch": 5.898912348048625, "grad_norm": 0.9453622692807283, "learning_rate": 7.598254854466774e-06, "loss": 0.1105, "step": 9220 }, { "epoch": 5.899552143314139, "grad_norm": 1.079931889520518, "learning_rate": 7.5962434045746506e-06, "loss": 0.1111, "step": 9221 }, { "epoch": 5.900191938579654, "grad_norm": 0.9677952519307897, "learning_rate": 7.594232057895554e-06, "loss": 0.1112, "step": 9222 }, { "epoch": 5.90083173384517, "grad_norm": 1.0386193027791426, "learning_rate": 7.5922208145158586e-06, "loss": 0.1226, "step": 9223 }, { "epoch": 5.901471529110685, "grad_norm": 1.306832765875905, "learning_rate": 7.590209674521918e-06, "loss": 0.1128, "step": 9224 }, { "epoch": 5.902111324376199, "grad_norm": 0.9801901937648578, "learning_rate": 7.588198638000084e-06, "loss": 0.1099, "step": 9225 }, { "epoch": 5.902751119641715, "grad_norm": 1.0453035282061125, "learning_rate": 7.586187705036713e-06, "loss": 0.1137, "step": 9226 }, { "epoch": 5.90339091490723, "grad_norm": 1.0754174460449202, "learning_rate": 7.584176875718148e-06, "loss": 0.122, "step": 9227 }, { "epoch": 5.904030710172744, "grad_norm": 0.964099818382103, "learning_rate": 7.58216615013073e-06, "loss": 0.1162, "step": 9228 }, { "epoch": 5.9046705054382596, "grad_norm": 0.9826427673775057, "learning_rate": 7.580155528360797e-06, "loss": 0.1166, "step": 9229 }, { "epoch": 5.905310300703775, "grad_norm": 1.0572247839166726, "learning_rate": 7.5781450104946816e-06, "loss": 0.1243, "step": 9230 }, { "epoch": 5.90595009596929, "grad_norm": 1.081335549632693, "learning_rate": 7.576134596618712e-06, "loss": 0.1158, "step": 9231 }, { "epoch": 5.9065898912348045, "grad_norm": 1.0355584068256234, "learning_rate": 7.574124286819206e-06, "loss": 0.1242, "step": 9232 }, { "epoch": 5.90722968650032, "grad_norm": 0.9720222487341398, "learning_rate": 7.572114081182491e-06, "loss": 0.1141, "step": 9233 }, { "epoch": 5.907869481765835, "grad_norm": 0.9768100530113344, "learning_rate": 7.57010397979488e-06, "loss": 0.1118, "step": 9234 }, { "epoch": 5.90850927703135, "grad_norm": 0.9420247866507951, "learning_rate": 7.56809398274268e-06, "loss": 0.1149, "step": 9235 }, { "epoch": 5.909149072296865, "grad_norm": 1.1268934131051382, "learning_rate": 7.5660840901122005e-06, "loss": 0.1112, "step": 9236 }, { "epoch": 5.90978886756238, "grad_norm": 0.9682729127112665, "learning_rate": 7.5640743019897385e-06, "loss": 0.1102, "step": 9237 }, { "epoch": 5.910428662827895, "grad_norm": 0.9713049817450413, "learning_rate": 7.562064618461591e-06, "loss": 0.1094, "step": 9238 }, { "epoch": 5.9110684580934105, "grad_norm": 0.8744601696041482, "learning_rate": 7.560055039614055e-06, "loss": 0.1095, "step": 9239 }, { "epoch": 5.911708253358925, "grad_norm": 0.9620751915414846, "learning_rate": 7.558045565533415e-06, "loss": 0.113, "step": 9240 }, { "epoch": 5.91234804862444, "grad_norm": 0.9938866183859772, "learning_rate": 7.556036196305955e-06, "loss": 0.1162, "step": 9241 }, { "epoch": 5.9129878438899555, "grad_norm": 0.9868806537099373, "learning_rate": 7.554026932017952e-06, "loss": 0.1125, "step": 9242 }, { "epoch": 5.91362763915547, "grad_norm": 0.9816404544559468, "learning_rate": 7.552017772755682e-06, "loss": 0.1161, "step": 9243 }, { "epoch": 5.914267434420985, "grad_norm": 1.0307597595771807, "learning_rate": 7.550008718605415e-06, "loss": 0.1156, "step": 9244 }, { "epoch": 5.9149072296865, "grad_norm": 1.135097164078826, "learning_rate": 7.547999769653411e-06, "loss": 0.1233, "step": 9245 }, { "epoch": 5.915547024952016, "grad_norm": 0.9405285302060286, "learning_rate": 7.54599092598594e-06, "loss": 0.116, "step": 9246 }, { "epoch": 5.91618682021753, "grad_norm": 0.9005152177033968, "learning_rate": 7.543982187689253e-06, "loss": 0.1108, "step": 9247 }, { "epoch": 5.916826615483045, "grad_norm": 1.0181024374267886, "learning_rate": 7.541973554849599e-06, "loss": 0.1201, "step": 9248 }, { "epoch": 5.917466410748561, "grad_norm": 1.0255889535909846, "learning_rate": 7.539965027553231e-06, "loss": 0.1164, "step": 9249 }, { "epoch": 5.918106206014075, "grad_norm": 0.9685122934798608, "learning_rate": 7.537956605886388e-06, "loss": 0.1132, "step": 9250 }, { "epoch": 5.91874600127959, "grad_norm": 0.9981694220372358, "learning_rate": 7.5359482899353085e-06, "loss": 0.1234, "step": 9251 }, { "epoch": 5.9193857965451055, "grad_norm": 0.8500560815633555, "learning_rate": 7.533940079786224e-06, "loss": 0.1079, "step": 9252 }, { "epoch": 5.920025591810621, "grad_norm": 0.9402690747038216, "learning_rate": 7.531931975525368e-06, "loss": 0.1099, "step": 9253 }, { "epoch": 5.920665387076136, "grad_norm": 0.9376222195008753, "learning_rate": 7.529923977238963e-06, "loss": 0.1106, "step": 9254 }, { "epoch": 5.9213051823416505, "grad_norm": 0.9853503323335147, "learning_rate": 7.527916085013225e-06, "loss": 0.1157, "step": 9255 }, { "epoch": 5.921944977607166, "grad_norm": 0.9923525174945433, "learning_rate": 7.525908298934377e-06, "loss": 0.1191, "step": 9256 }, { "epoch": 5.922584772872681, "grad_norm": 0.9641898123237932, "learning_rate": 7.523900619088624e-06, "loss": 0.1125, "step": 9257 }, { "epoch": 5.923224568138195, "grad_norm": 1.132156773645548, "learning_rate": 7.521893045562173e-06, "loss": 0.1127, "step": 9258 }, { "epoch": 5.923864363403711, "grad_norm": 0.9570525081637358, "learning_rate": 7.5198855784412285e-06, "loss": 0.1168, "step": 9259 }, { "epoch": 5.924504158669226, "grad_norm": 1.0085752595798303, "learning_rate": 7.517878217811985e-06, "loss": 0.1129, "step": 9260 }, { "epoch": 5.925143953934741, "grad_norm": 0.9384671192246481, "learning_rate": 7.5158709637606366e-06, "loss": 0.1103, "step": 9261 }, { "epoch": 5.925783749200256, "grad_norm": 0.9826624233407719, "learning_rate": 7.513863816373364e-06, "loss": 0.1221, "step": 9262 }, { "epoch": 5.926423544465771, "grad_norm": 1.0238923704846328, "learning_rate": 7.511856775736362e-06, "loss": 0.1207, "step": 9263 }, { "epoch": 5.927063339731286, "grad_norm": 0.9751438026964013, "learning_rate": 7.509849841935803e-06, "loss": 0.1118, "step": 9264 }, { "epoch": 5.927703134996801, "grad_norm": 0.9887918967748551, "learning_rate": 7.507843015057862e-06, "loss": 0.1133, "step": 9265 }, { "epoch": 5.928342930262316, "grad_norm": 1.0472262881430459, "learning_rate": 7.505836295188709e-06, "loss": 0.1093, "step": 9266 }, { "epoch": 5.928982725527831, "grad_norm": 1.06239342401774, "learning_rate": 7.50382968241451e-06, "loss": 0.1176, "step": 9267 }, { "epoch": 5.929622520793346, "grad_norm": 1.055819957965036, "learning_rate": 7.501823176821423e-06, "loss": 0.1161, "step": 9268 }, { "epoch": 5.930262316058861, "grad_norm": 0.9706501879023535, "learning_rate": 7.499816778495601e-06, "loss": 0.1097, "step": 9269 }, { "epoch": 5.930902111324376, "grad_norm": 1.0400032216605306, "learning_rate": 7.497810487523204e-06, "loss": 0.1195, "step": 9270 }, { "epoch": 5.931541906589891, "grad_norm": 0.8971739789988071, "learning_rate": 7.4958043039903724e-06, "loss": 0.105, "step": 9271 }, { "epoch": 5.932181701855407, "grad_norm": 1.025695098035966, "learning_rate": 7.493798227983249e-06, "loss": 0.1172, "step": 9272 }, { "epoch": 5.932821497120921, "grad_norm": 0.9501466753774462, "learning_rate": 7.491792259587972e-06, "loss": 0.1073, "step": 9273 }, { "epoch": 5.933461292386436, "grad_norm": 1.0261287847926368, "learning_rate": 7.489786398890675e-06, "loss": 0.1228, "step": 9274 }, { "epoch": 5.9341010876519515, "grad_norm": 0.8970653993836594, "learning_rate": 7.487780645977478e-06, "loss": 0.1128, "step": 9275 }, { "epoch": 5.934740882917467, "grad_norm": 0.9563389019146558, "learning_rate": 7.485775000934517e-06, "loss": 0.1113, "step": 9276 }, { "epoch": 5.935380678182981, "grad_norm": 0.9557858402730308, "learning_rate": 7.4837694638479055e-06, "loss": 0.1071, "step": 9277 }, { "epoch": 5.9360204734484965, "grad_norm": 0.9712458420167949, "learning_rate": 7.481764034803757e-06, "loss": 0.1148, "step": 9278 }, { "epoch": 5.936660268714012, "grad_norm": 1.1356043703880454, "learning_rate": 7.479758713888179e-06, "loss": 0.1213, "step": 9279 }, { "epoch": 5.937300063979526, "grad_norm": 1.018564859725171, "learning_rate": 7.47775350118728e-06, "loss": 0.1139, "step": 9280 }, { "epoch": 5.937939859245041, "grad_norm": 0.9923419834894192, "learning_rate": 7.4757483967871605e-06, "loss": 0.1086, "step": 9281 }, { "epoch": 5.938579654510557, "grad_norm": 0.8961142484746597, "learning_rate": 7.4737434007739094e-06, "loss": 0.1072, "step": 9282 }, { "epoch": 5.939219449776072, "grad_norm": 1.0878736471522419, "learning_rate": 7.471738513233626e-06, "loss": 0.1258, "step": 9283 }, { "epoch": 5.939859245041586, "grad_norm": 0.8656682627747387, "learning_rate": 7.469733734252396e-06, "loss": 0.1113, "step": 9284 }, { "epoch": 5.940499040307102, "grad_norm": 1.0855709606607395, "learning_rate": 7.467729063916291e-06, "loss": 0.1056, "step": 9285 }, { "epoch": 5.941138835572617, "grad_norm": 0.966874015330187, "learning_rate": 7.465724502311402e-06, "loss": 0.1107, "step": 9286 }, { "epoch": 5.941778630838132, "grad_norm": 0.9914351224385549, "learning_rate": 7.463720049523795e-06, "loss": 0.1092, "step": 9287 }, { "epoch": 5.942418426103647, "grad_norm": 0.9052818520626746, "learning_rate": 7.461715705639536e-06, "loss": 0.1131, "step": 9288 }, { "epoch": 5.943058221369162, "grad_norm": 0.9487118398385873, "learning_rate": 7.459711470744687e-06, "loss": 0.1116, "step": 9289 }, { "epoch": 5.943698016634677, "grad_norm": 1.0115719406744523, "learning_rate": 7.457707344925312e-06, "loss": 0.1077, "step": 9290 }, { "epoch": 5.944337811900192, "grad_norm": 1.111608132046958, "learning_rate": 7.455703328267461e-06, "loss": 0.1231, "step": 9291 }, { "epoch": 5.944977607165707, "grad_norm": 0.9802826127575154, "learning_rate": 7.453699420857178e-06, "loss": 0.1098, "step": 9292 }, { "epoch": 5.945617402431222, "grad_norm": 0.9507901605376352, "learning_rate": 7.451695622780517e-06, "loss": 0.1195, "step": 9293 }, { "epoch": 5.946257197696737, "grad_norm": 1.0429664332765856, "learning_rate": 7.449691934123513e-06, "loss": 0.1231, "step": 9294 }, { "epoch": 5.946896992962252, "grad_norm": 1.0738818336904945, "learning_rate": 7.447688354972198e-06, "loss": 0.1199, "step": 9295 }, { "epoch": 5.947536788227767, "grad_norm": 1.0550371672353431, "learning_rate": 7.445684885412607e-06, "loss": 0.1175, "step": 9296 }, { "epoch": 5.948176583493282, "grad_norm": 1.0105734355226024, "learning_rate": 7.443681525530765e-06, "loss": 0.1133, "step": 9297 }, { "epoch": 5.9488163787587975, "grad_norm": 0.9279025541890814, "learning_rate": 7.441678275412689e-06, "loss": 0.1106, "step": 9298 }, { "epoch": 5.949456174024312, "grad_norm": 0.994886358235835, "learning_rate": 7.439675135144394e-06, "loss": 0.1081, "step": 9299 }, { "epoch": 5.950095969289827, "grad_norm": 0.9838447981919873, "learning_rate": 7.4376721048118984e-06, "loss": 0.1083, "step": 9300 }, { "epoch": 5.9507357645553425, "grad_norm": 0.9669599815042826, "learning_rate": 7.4356691845012045e-06, "loss": 0.1132, "step": 9301 }, { "epoch": 5.951375559820857, "grad_norm": 0.9254628954202839, "learning_rate": 7.433666374298311e-06, "loss": 0.1068, "step": 9302 }, { "epoch": 5.952015355086372, "grad_norm": 0.9927075233233567, "learning_rate": 7.431663674289222e-06, "loss": 0.1157, "step": 9303 }, { "epoch": 5.952655150351887, "grad_norm": 0.9801162389887931, "learning_rate": 7.429661084559925e-06, "loss": 0.1089, "step": 9304 }, { "epoch": 5.953294945617403, "grad_norm": 1.0478530086692062, "learning_rate": 7.427658605196404e-06, "loss": 0.1178, "step": 9305 }, { "epoch": 5.953934740882918, "grad_norm": 1.0286590995552807, "learning_rate": 7.4256562362846505e-06, "loss": 0.1137, "step": 9306 }, { "epoch": 5.954574536148432, "grad_norm": 1.080800545077076, "learning_rate": 7.423653977910641e-06, "loss": 0.1228, "step": 9307 }, { "epoch": 5.955214331413948, "grad_norm": 1.0469382866915153, "learning_rate": 7.4216518301603454e-06, "loss": 0.1184, "step": 9308 }, { "epoch": 5.955854126679463, "grad_norm": 0.9933768770288096, "learning_rate": 7.419649793119732e-06, "loss": 0.1098, "step": 9309 }, { "epoch": 5.956493921944977, "grad_norm": 0.956350903400025, "learning_rate": 7.417647866874769e-06, "loss": 0.1115, "step": 9310 }, { "epoch": 5.957133717210493, "grad_norm": 0.9326117734896082, "learning_rate": 7.4156460515114115e-06, "loss": 0.1119, "step": 9311 }, { "epoch": 5.957773512476008, "grad_norm": 1.0229684121350375, "learning_rate": 7.413644347115612e-06, "loss": 0.1201, "step": 9312 }, { "epoch": 5.958413307741523, "grad_norm": 0.9343242338917402, "learning_rate": 7.411642753773327e-06, "loss": 0.114, "step": 9313 }, { "epoch": 5.9590531030070375, "grad_norm": 1.0132373063219398, "learning_rate": 7.409641271570498e-06, "loss": 0.1125, "step": 9314 }, { "epoch": 5.959692898272553, "grad_norm": 0.9888751455685838, "learning_rate": 7.407639900593062e-06, "loss": 0.121, "step": 9315 }, { "epoch": 5.960332693538068, "grad_norm": 1.0169892521373294, "learning_rate": 7.40563864092696e-06, "loss": 0.1221, "step": 9316 }, { "epoch": 5.960972488803582, "grad_norm": 0.9821192287594804, "learning_rate": 7.40363749265812e-06, "loss": 0.1202, "step": 9317 }, { "epoch": 5.961612284069098, "grad_norm": 1.0549715504398625, "learning_rate": 7.401636455872469e-06, "loss": 0.1203, "step": 9318 }, { "epoch": 5.962252079334613, "grad_norm": 0.9217529043390548, "learning_rate": 7.399635530655922e-06, "loss": 0.1162, "step": 9319 }, { "epoch": 5.962891874600128, "grad_norm": 0.8615399307836675, "learning_rate": 7.397634717094403e-06, "loss": 0.1075, "step": 9320 }, { "epoch": 5.963531669865643, "grad_norm": 1.0173611047610218, "learning_rate": 7.3956340152738206e-06, "loss": 0.1137, "step": 9321 }, { "epoch": 5.964171465131158, "grad_norm": 0.9499714494123879, "learning_rate": 7.393633425280076e-06, "loss": 0.109, "step": 9322 }, { "epoch": 5.964811260396673, "grad_norm": 0.9899145694482514, "learning_rate": 7.391632947199079e-06, "loss": 0.1203, "step": 9323 }, { "epoch": 5.9654510556621885, "grad_norm": 1.1297773525566035, "learning_rate": 7.3896325811167235e-06, "loss": 0.1124, "step": 9324 }, { "epoch": 5.966090850927703, "grad_norm": 1.056028048879979, "learning_rate": 7.387632327118901e-06, "loss": 0.1149, "step": 9325 }, { "epoch": 5.966730646193218, "grad_norm": 1.0477877595644518, "learning_rate": 7.3856321852915e-06, "loss": 0.1142, "step": 9326 }, { "epoch": 5.967370441458733, "grad_norm": 0.9215550998043548, "learning_rate": 7.383632155720403e-06, "loss": 0.112, "step": 9327 }, { "epoch": 5.968010236724249, "grad_norm": 1.150321861564029, "learning_rate": 7.381632238491487e-06, "loss": 0.1139, "step": 9328 }, { "epoch": 5.968650031989763, "grad_norm": 1.0575056739046689, "learning_rate": 7.37963243369062e-06, "loss": 0.1116, "step": 9329 }, { "epoch": 5.969289827255278, "grad_norm": 0.9084703124397002, "learning_rate": 7.377632741403679e-06, "loss": 0.1113, "step": 9330 }, { "epoch": 5.969929622520794, "grad_norm": 0.9822113501175372, "learning_rate": 7.375633161716524e-06, "loss": 0.1104, "step": 9331 }, { "epoch": 5.970569417786308, "grad_norm": 0.9743608596530964, "learning_rate": 7.373633694715012e-06, "loss": 0.1104, "step": 9332 }, { "epoch": 5.971209213051823, "grad_norm": 0.9929439737602931, "learning_rate": 7.371634340484998e-06, "loss": 0.1166, "step": 9333 }, { "epoch": 5.971849008317339, "grad_norm": 1.0223002851873002, "learning_rate": 7.3696350991123285e-06, "loss": 0.1093, "step": 9334 }, { "epoch": 5.972488803582854, "grad_norm": 0.9584661837813339, "learning_rate": 7.367635970682847e-06, "loss": 0.1141, "step": 9335 }, { "epoch": 5.973128598848368, "grad_norm": 1.0921127029884827, "learning_rate": 7.365636955282397e-06, "loss": 0.1187, "step": 9336 }, { "epoch": 5.9737683941138835, "grad_norm": 0.9317715406821995, "learning_rate": 7.363638052996811e-06, "loss": 0.1065, "step": 9337 }, { "epoch": 5.974408189379399, "grad_norm": 1.0143883922099561, "learning_rate": 7.361639263911916e-06, "loss": 0.1163, "step": 9338 }, { "epoch": 5.975047984644913, "grad_norm": 0.9512884640040878, "learning_rate": 7.359640588113537e-06, "loss": 0.1137, "step": 9339 }, { "epoch": 5.975687779910428, "grad_norm": 0.93057815437218, "learning_rate": 7.357642025687495e-06, "loss": 0.1085, "step": 9340 }, { "epoch": 5.976327575175944, "grad_norm": 0.8997545615767869, "learning_rate": 7.3556435767196045e-06, "loss": 0.1112, "step": 9341 }, { "epoch": 5.976967370441459, "grad_norm": 0.9281818874371436, "learning_rate": 7.3536452412956716e-06, "loss": 0.1083, "step": 9342 }, { "epoch": 5.977607165706974, "grad_norm": 0.9702972227885065, "learning_rate": 7.3516470195015075e-06, "loss": 0.1094, "step": 9343 }, { "epoch": 5.978246960972489, "grad_norm": 1.1098512343668125, "learning_rate": 7.349648911422911e-06, "loss": 0.1145, "step": 9344 }, { "epoch": 5.978886756238004, "grad_norm": 0.9994787421968134, "learning_rate": 7.34765091714567e-06, "loss": 0.1111, "step": 9345 }, { "epoch": 5.979526551503519, "grad_norm": 1.0761747959601935, "learning_rate": 7.345653036755585e-06, "loss": 0.1149, "step": 9346 }, { "epoch": 5.980166346769034, "grad_norm": 0.9975575270762479, "learning_rate": 7.3436552703384365e-06, "loss": 0.1103, "step": 9347 }, { "epoch": 5.980806142034549, "grad_norm": 0.9810130600585305, "learning_rate": 7.3416576179800055e-06, "loss": 0.1158, "step": 9348 }, { "epoch": 5.981445937300064, "grad_norm": 1.039511669821398, "learning_rate": 7.339660079766067e-06, "loss": 0.1183, "step": 9349 }, { "epoch": 5.982085732565579, "grad_norm": 0.9489098156137096, "learning_rate": 7.3376626557823914e-06, "loss": 0.1146, "step": 9350 }, { "epoch": 5.982725527831094, "grad_norm": 0.9063163406144235, "learning_rate": 7.335665346114749e-06, "loss": 0.1097, "step": 9351 }, { "epoch": 5.983365323096609, "grad_norm": 1.0532060630430682, "learning_rate": 7.333668150848889e-06, "loss": 0.1192, "step": 9352 }, { "epoch": 5.984005118362124, "grad_norm": 1.0981685990805785, "learning_rate": 7.331671070070583e-06, "loss": 0.1142, "step": 9353 }, { "epoch": 5.984644913627639, "grad_norm": 1.048743405877712, "learning_rate": 7.329674103865574e-06, "loss": 0.1241, "step": 9354 }, { "epoch": 5.985284708893154, "grad_norm": 1.0202348393693894, "learning_rate": 7.327677252319607e-06, "loss": 0.1199, "step": 9355 }, { "epoch": 5.985924504158669, "grad_norm": 0.915201954392356, "learning_rate": 7.325680515518425e-06, "loss": 0.1097, "step": 9356 }, { "epoch": 5.9865642994241846, "grad_norm": 1.1255441996280902, "learning_rate": 7.323683893547766e-06, "loss": 0.1248, "step": 9357 }, { "epoch": 5.987204094689699, "grad_norm": 1.0485239849589918, "learning_rate": 7.321687386493361e-06, "loss": 0.1188, "step": 9358 }, { "epoch": 5.987843889955214, "grad_norm": 0.8743669655715212, "learning_rate": 7.319690994440929e-06, "loss": 0.1094, "step": 9359 }, { "epoch": 5.9884836852207295, "grad_norm": 1.037031375498054, "learning_rate": 7.317694717476202e-06, "loss": 0.1141, "step": 9360 }, { "epoch": 5.989123480486245, "grad_norm": 0.9993660646242472, "learning_rate": 7.315698555684894e-06, "loss": 0.1195, "step": 9361 }, { "epoch": 5.989763275751759, "grad_norm": 1.1383436590709464, "learning_rate": 7.313702509152711e-06, "loss": 0.1275, "step": 9362 }, { "epoch": 5.990403071017274, "grad_norm": 1.1015032340067887, "learning_rate": 7.311706577965367e-06, "loss": 0.1248, "step": 9363 }, { "epoch": 5.99104286628279, "grad_norm": 0.8969062453961432, "learning_rate": 7.309710762208562e-06, "loss": 0.1118, "step": 9364 }, { "epoch": 5.991682661548305, "grad_norm": 0.9766853483341007, "learning_rate": 7.307715061967988e-06, "loss": 0.1089, "step": 9365 }, { "epoch": 5.992322456813819, "grad_norm": 0.8904214169488947, "learning_rate": 7.305719477329338e-06, "loss": 0.109, "step": 9366 }, { "epoch": 5.992962252079335, "grad_norm": 1.0134784411601692, "learning_rate": 7.303724008378304e-06, "loss": 0.1154, "step": 9367 }, { "epoch": 5.99360204734485, "grad_norm": 1.1722219832637124, "learning_rate": 7.301728655200567e-06, "loss": 0.1178, "step": 9368 }, { "epoch": 5.994241842610364, "grad_norm": 1.0598975635103285, "learning_rate": 7.299733417881796e-06, "loss": 0.1151, "step": 9369 }, { "epoch": 5.99488163787588, "grad_norm": 0.9892577717225994, "learning_rate": 7.297738296507673e-06, "loss": 0.1144, "step": 9370 }, { "epoch": 5.995521433141395, "grad_norm": 0.9468610249185871, "learning_rate": 7.29574329116386e-06, "loss": 0.1129, "step": 9371 }, { "epoch": 5.99616122840691, "grad_norm": 1.0667765330203534, "learning_rate": 7.293748401936016e-06, "loss": 0.1188, "step": 9372 }, { "epoch": 5.9968010236724245, "grad_norm": 0.9689323993166701, "learning_rate": 7.291753628909806e-06, "loss": 0.1063, "step": 9373 }, { "epoch": 5.99744081893794, "grad_norm": 0.9509119528982233, "learning_rate": 7.289758972170876e-06, "loss": 0.1172, "step": 9374 }, { "epoch": 5.998080614203455, "grad_norm": 0.957957730493538, "learning_rate": 7.2877644318048754e-06, "loss": 0.1162, "step": 9375 }, { "epoch": 5.9987204094689695, "grad_norm": 1.0176547273073373, "learning_rate": 7.2857700078974415e-06, "loss": 0.1154, "step": 9376 }, { "epoch": 5.999360204734485, "grad_norm": 0.9537177278012301, "learning_rate": 7.2837757005342194e-06, "loss": 0.1108, "step": 9377 }, { "epoch": 6.0, "grad_norm": 0.8504941397651667, "learning_rate": 7.281781509800838e-06, "loss": 0.111, "step": 9378 }, { "epoch": 6.000639795265515, "grad_norm": 0.8370091384910474, "learning_rate": 7.279787435782919e-06, "loss": 0.0916, "step": 9379 }, { "epoch": 6.00127959053103, "grad_norm": 0.7174997301528919, "learning_rate": 7.277793478566091e-06, "loss": 0.0895, "step": 9380 }, { "epoch": 6.001919385796545, "grad_norm": 0.6928847344837098, "learning_rate": 7.2757996382359694e-06, "loss": 0.0907, "step": 9381 }, { "epoch": 6.00255918106206, "grad_norm": 0.7312872997351876, "learning_rate": 7.27380591487816e-06, "loss": 0.0885, "step": 9382 }, { "epoch": 6.0031989763275755, "grad_norm": 0.8548932255441303, "learning_rate": 7.271812308578281e-06, "loss": 0.088, "step": 9383 }, { "epoch": 6.00383877159309, "grad_norm": 0.7436295168497848, "learning_rate": 7.2698188194219275e-06, "loss": 0.0849, "step": 9384 }, { "epoch": 6.004478566858605, "grad_norm": 0.8481150601986784, "learning_rate": 7.267825447494698e-06, "loss": 0.0883, "step": 9385 }, { "epoch": 6.00511836212412, "grad_norm": 0.7392915605095617, "learning_rate": 7.2658321928821804e-06, "loss": 0.0846, "step": 9386 }, { "epoch": 6.005758157389636, "grad_norm": 0.7655675310316099, "learning_rate": 7.263839055669968e-06, "loss": 0.0904, "step": 9387 }, { "epoch": 6.00639795265515, "grad_norm": 0.8298443687372424, "learning_rate": 7.261846035943639e-06, "loss": 0.0853, "step": 9388 }, { "epoch": 6.007037747920665, "grad_norm": 0.9246963673330496, "learning_rate": 7.259853133788765e-06, "loss": 0.082, "step": 9389 }, { "epoch": 6.007677543186181, "grad_norm": 0.8497372188163419, "learning_rate": 7.257860349290928e-06, "loss": 0.0831, "step": 9390 }, { "epoch": 6.008317338451695, "grad_norm": 0.8410433384304873, "learning_rate": 7.255867682535691e-06, "loss": 0.0842, "step": 9391 }, { "epoch": 6.00895713371721, "grad_norm": 0.8285503626479975, "learning_rate": 7.25387513360861e-06, "loss": 0.0848, "step": 9392 }, { "epoch": 6.009596928982726, "grad_norm": 0.7965076895737881, "learning_rate": 7.25188270259525e-06, "loss": 0.0844, "step": 9393 }, { "epoch": 6.010236724248241, "grad_norm": 0.8480632903718872, "learning_rate": 7.2498903895811565e-06, "loss": 0.0898, "step": 9394 }, { "epoch": 6.010876519513755, "grad_norm": 0.8678785675256477, "learning_rate": 7.247898194651879e-06, "loss": 0.0812, "step": 9395 }, { "epoch": 6.0115163147792705, "grad_norm": 0.9298610549975936, "learning_rate": 7.245906117892951e-06, "loss": 0.0835, "step": 9396 }, { "epoch": 6.012156110044786, "grad_norm": 0.9366115208770793, "learning_rate": 7.243914159389922e-06, "loss": 0.0856, "step": 9397 }, { "epoch": 6.012795905310301, "grad_norm": 0.9123949995704426, "learning_rate": 7.2419223192283154e-06, "loss": 0.087, "step": 9398 }, { "epoch": 6.0134357005758154, "grad_norm": 0.9054386314980222, "learning_rate": 7.2399305974936566e-06, "loss": 0.0856, "step": 9399 }, { "epoch": 6.014075495841331, "grad_norm": 0.8978943600016004, "learning_rate": 7.2379389942714685e-06, "loss": 0.0882, "step": 9400 }, { "epoch": 6.014715291106846, "grad_norm": 1.0106333601455513, "learning_rate": 7.235947509647268e-06, "loss": 0.0877, "step": 9401 }, { "epoch": 6.015355086372361, "grad_norm": 0.9202975981627161, "learning_rate": 7.23395614370656e-06, "loss": 0.0939, "step": 9402 }, { "epoch": 6.015994881637876, "grad_norm": 0.8529078307343921, "learning_rate": 7.231964896534859e-06, "loss": 0.085, "step": 9403 }, { "epoch": 6.016634676903391, "grad_norm": 0.8077773995848344, "learning_rate": 7.229973768217662e-06, "loss": 0.0863, "step": 9404 }, { "epoch": 6.017274472168906, "grad_norm": 0.8965102893467989, "learning_rate": 7.227982758840461e-06, "loss": 0.0926, "step": 9405 }, { "epoch": 6.017914267434421, "grad_norm": 0.9565317958329721, "learning_rate": 7.225991868488749e-06, "loss": 0.0928, "step": 9406 }, { "epoch": 6.018554062699936, "grad_norm": 0.781080789882799, "learning_rate": 7.224001097248017e-06, "loss": 0.0783, "step": 9407 }, { "epoch": 6.019193857965451, "grad_norm": 0.7961724538727678, "learning_rate": 7.222010445203735e-06, "loss": 0.0844, "step": 9408 }, { "epoch": 6.019833653230966, "grad_norm": 0.8270072845749146, "learning_rate": 7.220019912441382e-06, "loss": 0.0802, "step": 9409 }, { "epoch": 6.020473448496481, "grad_norm": 0.895126998072036, "learning_rate": 7.218029499046431e-06, "loss": 0.0989, "step": 9410 }, { "epoch": 6.021113243761996, "grad_norm": 0.7606121981222776, "learning_rate": 7.216039205104346e-06, "loss": 0.0877, "step": 9411 }, { "epoch": 6.021753039027511, "grad_norm": 0.8973352427964714, "learning_rate": 7.214049030700579e-06, "loss": 0.0905, "step": 9412 }, { "epoch": 6.022392834293027, "grad_norm": 0.850370886405185, "learning_rate": 7.212058975920598e-06, "loss": 0.0814, "step": 9413 }, { "epoch": 6.023032629558541, "grad_norm": 0.8906447780109468, "learning_rate": 7.210069040849842e-06, "loss": 0.088, "step": 9414 }, { "epoch": 6.023672424824056, "grad_norm": 0.9379654003100156, "learning_rate": 7.20807922557376e-06, "loss": 0.0878, "step": 9415 }, { "epoch": 6.024312220089572, "grad_norm": 0.7672708193091696, "learning_rate": 7.2060895301777866e-06, "loss": 0.082, "step": 9416 }, { "epoch": 6.024952015355086, "grad_norm": 0.8428544931744152, "learning_rate": 7.204099954747362e-06, "loss": 0.0833, "step": 9417 }, { "epoch": 6.025591810620601, "grad_norm": 0.8344059749181805, "learning_rate": 7.202110499367911e-06, "loss": 0.0771, "step": 9418 }, { "epoch": 6.0262316058861165, "grad_norm": 0.8052627590813389, "learning_rate": 7.200121164124856e-06, "loss": 0.0827, "step": 9419 }, { "epoch": 6.026871401151632, "grad_norm": 0.8414293699601202, "learning_rate": 7.1981319491036196e-06, "loss": 0.0802, "step": 9420 }, { "epoch": 6.027511196417146, "grad_norm": 0.9436565900560994, "learning_rate": 7.196142854389614e-06, "loss": 0.0915, "step": 9421 }, { "epoch": 6.028150991682661, "grad_norm": 0.8964634672488321, "learning_rate": 7.194153880068244e-06, "loss": 0.0895, "step": 9422 }, { "epoch": 6.028790786948177, "grad_norm": 0.8592064368399637, "learning_rate": 7.192165026224917e-06, "loss": 0.0896, "step": 9423 }, { "epoch": 6.029430582213692, "grad_norm": 0.8522793050585616, "learning_rate": 7.190176292945031e-06, "loss": 0.082, "step": 9424 }, { "epoch": 6.030070377479206, "grad_norm": 0.851456016264091, "learning_rate": 7.188187680313974e-06, "loss": 0.0814, "step": 9425 }, { "epoch": 6.030710172744722, "grad_norm": 0.829070004500614, "learning_rate": 7.186199188417133e-06, "loss": 0.0859, "step": 9426 }, { "epoch": 6.031349968010237, "grad_norm": 0.8974963997903407, "learning_rate": 7.184210817339899e-06, "loss": 0.0785, "step": 9427 }, { "epoch": 6.031989763275751, "grad_norm": 0.9164392229028261, "learning_rate": 7.182222567167641e-06, "loss": 0.0901, "step": 9428 }, { "epoch": 6.032629558541267, "grad_norm": 0.9163425615035592, "learning_rate": 7.1802344379857335e-06, "loss": 0.0877, "step": 9429 }, { "epoch": 6.033269353806782, "grad_norm": 0.7974644926942793, "learning_rate": 7.178246429879545e-06, "loss": 0.0728, "step": 9430 }, { "epoch": 6.033909149072297, "grad_norm": 0.7939375331015079, "learning_rate": 7.176258542934436e-06, "loss": 0.0728, "step": 9431 }, { "epoch": 6.0345489443378115, "grad_norm": 0.833710551429354, "learning_rate": 7.17427077723576e-06, "loss": 0.0842, "step": 9432 }, { "epoch": 6.035188739603327, "grad_norm": 0.7841747639075819, "learning_rate": 7.172283132868867e-06, "loss": 0.0798, "step": 9433 }, { "epoch": 6.035828534868842, "grad_norm": 0.82255949088676, "learning_rate": 7.1702956099191115e-06, "loss": 0.0808, "step": 9434 }, { "epoch": 6.036468330134357, "grad_norm": 0.8416754944704189, "learning_rate": 7.168308208471827e-06, "loss": 0.0761, "step": 9435 }, { "epoch": 6.037108125399872, "grad_norm": 0.818531057967006, "learning_rate": 7.166320928612349e-06, "loss": 0.0801, "step": 9436 }, { "epoch": 6.037747920665387, "grad_norm": 0.9013298785233699, "learning_rate": 7.164333770426013e-06, "loss": 0.0866, "step": 9437 }, { "epoch": 6.038387715930902, "grad_norm": 0.8883839408731188, "learning_rate": 7.16234673399814e-06, "loss": 0.0824, "step": 9438 }, { "epoch": 6.039027511196418, "grad_norm": 0.7931101430942682, "learning_rate": 7.160359819414047e-06, "loss": 0.085, "step": 9439 }, { "epoch": 6.039667306461932, "grad_norm": 0.841220105846425, "learning_rate": 7.158373026759056e-06, "loss": 0.0836, "step": 9440 }, { "epoch": 6.040307101727447, "grad_norm": 0.7696092549739952, "learning_rate": 7.156386356118471e-06, "loss": 0.0797, "step": 9441 }, { "epoch": 6.0409468969929625, "grad_norm": 0.8143124516765191, "learning_rate": 7.154399807577599e-06, "loss": 0.083, "step": 9442 }, { "epoch": 6.041586692258477, "grad_norm": 1.0106447512913714, "learning_rate": 7.1524133812217325e-06, "loss": 0.0956, "step": 9443 }, { "epoch": 6.042226487523992, "grad_norm": 1.0082999522567362, "learning_rate": 7.1504270771361724e-06, "loss": 0.0912, "step": 9444 }, { "epoch": 6.042866282789507, "grad_norm": 0.8022134774981328, "learning_rate": 7.148440895406208e-06, "loss": 0.0764, "step": 9445 }, { "epoch": 6.043506078055023, "grad_norm": 0.8872948438644006, "learning_rate": 7.1464548361171135e-06, "loss": 0.0875, "step": 9446 }, { "epoch": 6.044145873320537, "grad_norm": 0.8329330640768706, "learning_rate": 7.144468899354177e-06, "loss": 0.0859, "step": 9447 }, { "epoch": 6.044785668586052, "grad_norm": 0.8841966474728453, "learning_rate": 7.142483085202665e-06, "loss": 0.0868, "step": 9448 }, { "epoch": 6.045425463851568, "grad_norm": 0.8796480476819563, "learning_rate": 7.140497393747843e-06, "loss": 0.081, "step": 9449 }, { "epoch": 6.046065259117083, "grad_norm": 0.9632668831874441, "learning_rate": 7.13851182507498e-06, "loss": 0.0819, "step": 9450 }, { "epoch": 6.046705054382597, "grad_norm": 0.8736009376150408, "learning_rate": 7.136526379269329e-06, "loss": 0.0761, "step": 9451 }, { "epoch": 6.047344849648113, "grad_norm": 0.9111536842518185, "learning_rate": 7.134541056416142e-06, "loss": 0.0876, "step": 9452 }, { "epoch": 6.047984644913628, "grad_norm": 0.8141196552503226, "learning_rate": 7.132555856600662e-06, "loss": 0.0834, "step": 9453 }, { "epoch": 6.048624440179142, "grad_norm": 0.8890377431045339, "learning_rate": 7.1305707799081345e-06, "loss": 0.0848, "step": 9454 }, { "epoch": 6.0492642354446575, "grad_norm": 0.8365905870983572, "learning_rate": 7.128585826423794e-06, "loss": 0.0779, "step": 9455 }, { "epoch": 6.049904030710173, "grad_norm": 0.9362779839498915, "learning_rate": 7.126600996232867e-06, "loss": 0.0824, "step": 9456 }, { "epoch": 6.050543825975688, "grad_norm": 0.8979293861553113, "learning_rate": 7.124616289420584e-06, "loss": 0.0932, "step": 9457 }, { "epoch": 6.0511836212412025, "grad_norm": 0.8933926100440981, "learning_rate": 7.122631706072164e-06, "loss": 0.0881, "step": 9458 }, { "epoch": 6.051823416506718, "grad_norm": 0.9111004459191752, "learning_rate": 7.120647246272817e-06, "loss": 0.0908, "step": 9459 }, { "epoch": 6.052463211772233, "grad_norm": 0.815915853468158, "learning_rate": 7.118662910107758e-06, "loss": 0.0789, "step": 9460 }, { "epoch": 6.053103007037748, "grad_norm": 0.9749640832160602, "learning_rate": 7.116678697662189e-06, "loss": 0.0823, "step": 9461 }, { "epoch": 6.053742802303263, "grad_norm": 0.9073712584379819, "learning_rate": 7.114694609021306e-06, "loss": 0.0827, "step": 9462 }, { "epoch": 6.054382597568778, "grad_norm": 0.835698171870257, "learning_rate": 7.112710644270301e-06, "loss": 0.0791, "step": 9463 }, { "epoch": 6.055022392834293, "grad_norm": 0.8325374004636257, "learning_rate": 7.1107268034943685e-06, "loss": 0.0797, "step": 9464 }, { "epoch": 6.0556621880998085, "grad_norm": 0.8235852936879271, "learning_rate": 7.108743086778688e-06, "loss": 0.0853, "step": 9465 }, { "epoch": 6.056301983365323, "grad_norm": 0.8021540259053427, "learning_rate": 7.106759494208433e-06, "loss": 0.081, "step": 9466 }, { "epoch": 6.056941778630838, "grad_norm": 0.7168332140324217, "learning_rate": 7.104776025868781e-06, "loss": 0.0886, "step": 9467 }, { "epoch": 6.057581573896353, "grad_norm": 0.8946836633961857, "learning_rate": 7.102792681844897e-06, "loss": 0.0817, "step": 9468 }, { "epoch": 6.058221369161868, "grad_norm": 0.8242286652306807, "learning_rate": 7.100809462221938e-06, "loss": 0.0827, "step": 9469 }, { "epoch": 6.058861164427383, "grad_norm": 0.8725641011377224, "learning_rate": 7.098826367085065e-06, "loss": 0.0855, "step": 9470 }, { "epoch": 6.059500959692898, "grad_norm": 0.7864560213309849, "learning_rate": 7.096843396519429e-06, "loss": 0.0815, "step": 9471 }, { "epoch": 6.060140754958414, "grad_norm": 0.8203274754662465, "learning_rate": 7.0948605506101745e-06, "loss": 0.0802, "step": 9472 }, { "epoch": 6.060780550223928, "grad_norm": 0.8855204939754011, "learning_rate": 7.092877829442433e-06, "loss": 0.0884, "step": 9473 }, { "epoch": 6.061420345489443, "grad_norm": 0.7972407190626234, "learning_rate": 7.090895233101351e-06, "loss": 0.0783, "step": 9474 }, { "epoch": 6.062060140754959, "grad_norm": 0.8259369441950009, "learning_rate": 7.088912761672055e-06, "loss": 0.0831, "step": 9475 }, { "epoch": 6.062699936020474, "grad_norm": 0.9146943098780972, "learning_rate": 7.086930415239663e-06, "loss": 0.087, "step": 9476 }, { "epoch": 6.063339731285988, "grad_norm": 0.8357498212275113, "learning_rate": 7.0849481938892986e-06, "loss": 0.0876, "step": 9477 }, { "epoch": 6.0639795265515035, "grad_norm": 0.8231195204718154, "learning_rate": 7.082966097706073e-06, "loss": 0.0834, "step": 9478 }, { "epoch": 6.064619321817019, "grad_norm": 0.9139314628300363, "learning_rate": 7.08098412677509e-06, "loss": 0.0877, "step": 9479 }, { "epoch": 6.065259117082533, "grad_norm": 0.8446533445944744, "learning_rate": 7.079002281181462e-06, "loss": 0.0812, "step": 9480 }, { "epoch": 6.0658989123480485, "grad_norm": 0.8942706757018637, "learning_rate": 7.077020561010279e-06, "loss": 0.0791, "step": 9481 }, { "epoch": 6.066538707613564, "grad_norm": 0.8393102387471093, "learning_rate": 7.075038966346632e-06, "loss": 0.0865, "step": 9482 }, { "epoch": 6.067178502879079, "grad_norm": 0.8875638354699968, "learning_rate": 7.073057497275608e-06, "loss": 0.0862, "step": 9483 }, { "epoch": 6.067818298144593, "grad_norm": 0.8955266276195396, "learning_rate": 7.071076153882292e-06, "loss": 0.0885, "step": 9484 }, { "epoch": 6.068458093410109, "grad_norm": 1.0159930340348398, "learning_rate": 7.069094936251754e-06, "loss": 0.0862, "step": 9485 }, { "epoch": 6.069097888675624, "grad_norm": 0.9145874923502736, "learning_rate": 7.067113844469063e-06, "loss": 0.0878, "step": 9486 }, { "epoch": 6.069737683941139, "grad_norm": 0.8431307375799316, "learning_rate": 7.065132878619288e-06, "loss": 0.0831, "step": 9487 }, { "epoch": 6.070377479206654, "grad_norm": 0.8061320590025457, "learning_rate": 7.0631520387874885e-06, "loss": 0.082, "step": 9488 }, { "epoch": 6.071017274472169, "grad_norm": 0.7881673678973447, "learning_rate": 7.061171325058713e-06, "loss": 0.0863, "step": 9489 }, { "epoch": 6.071657069737684, "grad_norm": 0.8590000930709987, "learning_rate": 7.0591907375180156e-06, "loss": 0.0805, "step": 9490 }, { "epoch": 6.0722968650031985, "grad_norm": 0.776278500753944, "learning_rate": 7.057210276250437e-06, "loss": 0.0771, "step": 9491 }, { "epoch": 6.072936660268714, "grad_norm": 0.7612476407169757, "learning_rate": 7.0552299413410145e-06, "loss": 0.0862, "step": 9492 }, { "epoch": 6.073576455534229, "grad_norm": 0.8137251381678527, "learning_rate": 7.053249732874775e-06, "loss": 0.0822, "step": 9493 }, { "epoch": 6.074216250799744, "grad_norm": 0.8244963493921158, "learning_rate": 7.051269650936754e-06, "loss": 0.0817, "step": 9494 }, { "epoch": 6.074856046065259, "grad_norm": 0.9226550059097127, "learning_rate": 7.0492896956119694e-06, "loss": 0.0873, "step": 9495 }, { "epoch": 6.075495841330774, "grad_norm": 0.8302593419169483, "learning_rate": 7.047309866985434e-06, "loss": 0.0842, "step": 9496 }, { "epoch": 6.076135636596289, "grad_norm": 0.8911931283651148, "learning_rate": 7.045330165142164e-06, "loss": 0.0876, "step": 9497 }, { "epoch": 6.076775431861805, "grad_norm": 0.8053593619270529, "learning_rate": 7.043350590167159e-06, "loss": 0.0778, "step": 9498 }, { "epoch": 6.077415227127319, "grad_norm": 0.7691186988039613, "learning_rate": 7.041371142145417e-06, "loss": 0.0733, "step": 9499 }, { "epoch": 6.078055022392834, "grad_norm": 0.9125727268494518, "learning_rate": 7.039391821161938e-06, "loss": 0.0812, "step": 9500 }, { "epoch": 6.0786948176583495, "grad_norm": 0.8018323859388663, "learning_rate": 7.03741262730171e-06, "loss": 0.0791, "step": 9501 }, { "epoch": 6.079334612923865, "grad_norm": 0.9584051218542483, "learning_rate": 7.035433560649712e-06, "loss": 0.089, "step": 9502 }, { "epoch": 6.079974408189379, "grad_norm": 0.8334280753366299, "learning_rate": 7.0334546212909195e-06, "loss": 0.0891, "step": 9503 }, { "epoch": 6.0806142034548945, "grad_norm": 0.9311258088211752, "learning_rate": 7.031475809310313e-06, "loss": 0.085, "step": 9504 }, { "epoch": 6.08125399872041, "grad_norm": 0.8406141626829124, "learning_rate": 7.0294971247928545e-06, "loss": 0.0739, "step": 9505 }, { "epoch": 6.081893793985924, "grad_norm": 0.9086078511904185, "learning_rate": 7.0275185678235025e-06, "loss": 0.088, "step": 9506 }, { "epoch": 6.082533589251439, "grad_norm": 0.7650655891950888, "learning_rate": 7.025540138487219e-06, "loss": 0.0808, "step": 9507 }, { "epoch": 6.083173384516955, "grad_norm": 0.7690482777533708, "learning_rate": 7.02356183686895e-06, "loss": 0.078, "step": 9508 }, { "epoch": 6.08381317978247, "grad_norm": 0.9048392239316628, "learning_rate": 7.021583663053641e-06, "loss": 0.0858, "step": 9509 }, { "epoch": 6.084452975047984, "grad_norm": 0.7311036754635386, "learning_rate": 7.0196056171262274e-06, "loss": 0.0776, "step": 9510 }, { "epoch": 6.0850927703135, "grad_norm": 0.8358489300870459, "learning_rate": 7.0176276991716505e-06, "loss": 0.081, "step": 9511 }, { "epoch": 6.085732565579015, "grad_norm": 1.0397505959629778, "learning_rate": 7.015649909274835e-06, "loss": 0.0932, "step": 9512 }, { "epoch": 6.08637236084453, "grad_norm": 0.875442461431722, "learning_rate": 7.013672247520702e-06, "loss": 0.0815, "step": 9513 }, { "epoch": 6.0870121561100445, "grad_norm": 0.820802448279815, "learning_rate": 7.011694713994173e-06, "loss": 0.076, "step": 9514 }, { "epoch": 6.08765195137556, "grad_norm": 0.7815428669930563, "learning_rate": 7.009717308780158e-06, "loss": 0.089, "step": 9515 }, { "epoch": 6.088291746641075, "grad_norm": 0.8982275503963678, "learning_rate": 7.007740031963556e-06, "loss": 0.0864, "step": 9516 }, { "epoch": 6.0889315419065895, "grad_norm": 0.9616543377990506, "learning_rate": 7.005762883629279e-06, "loss": 0.0857, "step": 9517 }, { "epoch": 6.089571337172105, "grad_norm": 0.9004837040134775, "learning_rate": 7.003785863862218e-06, "loss": 0.0802, "step": 9518 }, { "epoch": 6.09021113243762, "grad_norm": 1.034747216659692, "learning_rate": 7.001808972747261e-06, "loss": 0.0888, "step": 9519 }, { "epoch": 6.090850927703135, "grad_norm": 0.8335268106741626, "learning_rate": 6.999832210369293e-06, "loss": 0.0798, "step": 9520 }, { "epoch": 6.09149072296865, "grad_norm": 0.8899295714449789, "learning_rate": 6.997855576813195e-06, "loss": 0.0743, "step": 9521 }, { "epoch": 6.092130518234165, "grad_norm": 0.9127137633234158, "learning_rate": 6.995879072163837e-06, "loss": 0.0889, "step": 9522 }, { "epoch": 6.09277031349968, "grad_norm": 0.8876672893422167, "learning_rate": 6.993902696506083e-06, "loss": 0.0894, "step": 9523 }, { "epoch": 6.0934101087651955, "grad_norm": 0.8877772330825997, "learning_rate": 6.991926449924806e-06, "loss": 0.0853, "step": 9524 }, { "epoch": 6.09404990403071, "grad_norm": 0.8492467273704203, "learning_rate": 6.989950332504856e-06, "loss": 0.081, "step": 9525 }, { "epoch": 6.094689699296225, "grad_norm": 0.9295362504519946, "learning_rate": 6.987974344331081e-06, "loss": 0.084, "step": 9526 }, { "epoch": 6.0953294945617404, "grad_norm": 0.9290946629022631, "learning_rate": 6.985998485488332e-06, "loss": 0.088, "step": 9527 }, { "epoch": 6.095969289827256, "grad_norm": 0.8082811036596483, "learning_rate": 6.984022756061446e-06, "loss": 0.0844, "step": 9528 }, { "epoch": 6.09660908509277, "grad_norm": 0.8366270257046255, "learning_rate": 6.982047156135259e-06, "loss": 0.0829, "step": 9529 }, { "epoch": 6.097248880358285, "grad_norm": 0.9957419540144841, "learning_rate": 6.980071685794594e-06, "loss": 0.0898, "step": 9530 }, { "epoch": 6.097888675623801, "grad_norm": 0.8642061383973823, "learning_rate": 6.978096345124283e-06, "loss": 0.083, "step": 9531 }, { "epoch": 6.098528470889315, "grad_norm": 0.9195793684349226, "learning_rate": 6.97612113420914e-06, "loss": 0.0832, "step": 9532 }, { "epoch": 6.09916826615483, "grad_norm": 0.8426706357692374, "learning_rate": 6.97414605313397e-06, "loss": 0.0848, "step": 9533 }, { "epoch": 6.099808061420346, "grad_norm": 0.7515881840972694, "learning_rate": 6.9721711019835925e-06, "loss": 0.0797, "step": 9534 }, { "epoch": 6.100447856685861, "grad_norm": 0.7894647248712752, "learning_rate": 6.9701962808428e-06, "loss": 0.0813, "step": 9535 }, { "epoch": 6.101087651951375, "grad_norm": 0.8959011033006521, "learning_rate": 6.968221589796389e-06, "loss": 0.0896, "step": 9536 }, { "epoch": 6.1017274472168905, "grad_norm": 0.848262597152672, "learning_rate": 6.966247028929151e-06, "loss": 0.084, "step": 9537 }, { "epoch": 6.102367242482406, "grad_norm": 0.8133214203633731, "learning_rate": 6.96427259832587e-06, "loss": 0.0743, "step": 9538 }, { "epoch": 6.103007037747921, "grad_norm": 0.9656039911918415, "learning_rate": 6.9622982980713225e-06, "loss": 0.0949, "step": 9539 }, { "epoch": 6.1036468330134355, "grad_norm": 0.9089627819134828, "learning_rate": 6.960324128250279e-06, "loss": 0.0854, "step": 9540 }, { "epoch": 6.104286628278951, "grad_norm": 0.777276107162762, "learning_rate": 6.958350088947514e-06, "loss": 0.0781, "step": 9541 }, { "epoch": 6.104926423544466, "grad_norm": 1.0233168100709056, "learning_rate": 6.9563761802477855e-06, "loss": 0.0839, "step": 9542 }, { "epoch": 6.10556621880998, "grad_norm": 0.8673636699701864, "learning_rate": 6.954402402235847e-06, "loss": 0.082, "step": 9543 }, { "epoch": 6.106206014075496, "grad_norm": 0.7623746126162481, "learning_rate": 6.952428754996454e-06, "loss": 0.0842, "step": 9544 }, { "epoch": 6.106845809341011, "grad_norm": 0.8811509116751363, "learning_rate": 6.95045523861435e-06, "loss": 0.0824, "step": 9545 }, { "epoch": 6.107485604606526, "grad_norm": 0.7454388431296274, "learning_rate": 6.948481853174268e-06, "loss": 0.0723, "step": 9546 }, { "epoch": 6.108125399872041, "grad_norm": 0.8901927742769565, "learning_rate": 6.946508598760953e-06, "loss": 0.0918, "step": 9547 }, { "epoch": 6.108765195137556, "grad_norm": 0.9098237603293674, "learning_rate": 6.944535475459124e-06, "loss": 0.0936, "step": 9548 }, { "epoch": 6.109404990403071, "grad_norm": 0.8189571589408132, "learning_rate": 6.942562483353509e-06, "loss": 0.0837, "step": 9549 }, { "epoch": 6.110044785668586, "grad_norm": 0.8981666316033782, "learning_rate": 6.9405896225288195e-06, "loss": 0.0808, "step": 9550 }, { "epoch": 6.110684580934101, "grad_norm": 0.9652750152913541, "learning_rate": 6.93861689306977e-06, "loss": 0.088, "step": 9551 }, { "epoch": 6.111324376199616, "grad_norm": 0.8840730255272724, "learning_rate": 6.936644295061068e-06, "loss": 0.081, "step": 9552 }, { "epoch": 6.111964171465131, "grad_norm": 0.8312278714238133, "learning_rate": 6.934671828587405e-06, "loss": 0.0867, "step": 9553 }, { "epoch": 6.112603966730646, "grad_norm": 0.9669713388526704, "learning_rate": 6.932699493733484e-06, "loss": 0.0964, "step": 9554 }, { "epoch": 6.113243761996161, "grad_norm": 0.7525086565243867, "learning_rate": 6.930727290583992e-06, "loss": 0.0797, "step": 9555 }, { "epoch": 6.113883557261676, "grad_norm": 0.7556878581584018, "learning_rate": 6.928755219223607e-06, "loss": 0.0788, "step": 9556 }, { "epoch": 6.114523352527192, "grad_norm": 0.7858972601269597, "learning_rate": 6.926783279737013e-06, "loss": 0.0799, "step": 9557 }, { "epoch": 6.115163147792706, "grad_norm": 0.8690412107140296, "learning_rate": 6.924811472208878e-06, "loss": 0.0812, "step": 9558 }, { "epoch": 6.115802943058221, "grad_norm": 1.049728737784977, "learning_rate": 6.922839796723867e-06, "loss": 0.096, "step": 9559 }, { "epoch": 6.1164427383237365, "grad_norm": 0.8934785320953011, "learning_rate": 6.920868253366637e-06, "loss": 0.0848, "step": 9560 }, { "epoch": 6.117082533589252, "grad_norm": 0.8697511374966789, "learning_rate": 6.9188968422218525e-06, "loss": 0.0798, "step": 9561 }, { "epoch": 6.117722328854766, "grad_norm": 0.8469254127406538, "learning_rate": 6.916925563374156e-06, "loss": 0.0869, "step": 9562 }, { "epoch": 6.1183621241202815, "grad_norm": 0.920989526980982, "learning_rate": 6.9149544169081876e-06, "loss": 0.0875, "step": 9563 }, { "epoch": 6.119001919385797, "grad_norm": 0.7708942773016778, "learning_rate": 6.912983402908593e-06, "loss": 0.0813, "step": 9564 }, { "epoch": 6.119641714651312, "grad_norm": 0.8478495319857507, "learning_rate": 6.911012521459999e-06, "loss": 0.0833, "step": 9565 }, { "epoch": 6.120281509916826, "grad_norm": 0.8860736862772258, "learning_rate": 6.90904177264703e-06, "loss": 0.0869, "step": 9566 }, { "epoch": 6.120921305182342, "grad_norm": 0.8432343328678614, "learning_rate": 6.907071156554311e-06, "loss": 0.0852, "step": 9567 }, { "epoch": 6.121561100447857, "grad_norm": 0.8825416819401068, "learning_rate": 6.905100673266456e-06, "loss": 0.0818, "step": 9568 }, { "epoch": 6.122200895713371, "grad_norm": 0.9343416038961442, "learning_rate": 6.903130322868071e-06, "loss": 0.0907, "step": 9569 }, { "epoch": 6.122840690978887, "grad_norm": 0.7918042562164334, "learning_rate": 6.9011601054437586e-06, "loss": 0.0863, "step": 9570 }, { "epoch": 6.123480486244402, "grad_norm": 0.8969633491124674, "learning_rate": 6.899190021078122e-06, "loss": 0.0895, "step": 9571 }, { "epoch": 6.124120281509917, "grad_norm": 0.9379774361681817, "learning_rate": 6.89722006985575e-06, "loss": 0.0887, "step": 9572 }, { "epoch": 6.1247600767754315, "grad_norm": 0.8548398447673408, "learning_rate": 6.8952502518612274e-06, "loss": 0.084, "step": 9573 }, { "epoch": 6.125399872040947, "grad_norm": 0.8919949202030139, "learning_rate": 6.893280567179137e-06, "loss": 0.0914, "step": 9574 }, { "epoch": 6.126039667306462, "grad_norm": 0.8667946740032232, "learning_rate": 6.891311015894054e-06, "loss": 0.0867, "step": 9575 }, { "epoch": 6.126679462571977, "grad_norm": 0.9310080719172501, "learning_rate": 6.88934159809054e-06, "loss": 0.0794, "step": 9576 }, { "epoch": 6.127319257837492, "grad_norm": 0.9463254552531095, "learning_rate": 6.887372313853169e-06, "loss": 0.0893, "step": 9577 }, { "epoch": 6.127959053103007, "grad_norm": 0.8732994883136508, "learning_rate": 6.885403163266495e-06, "loss": 0.0873, "step": 9578 }, { "epoch": 6.128598848368522, "grad_norm": 0.8734857490278163, "learning_rate": 6.883434146415069e-06, "loss": 0.0812, "step": 9579 }, { "epoch": 6.129238643634037, "grad_norm": 0.9696553307590452, "learning_rate": 6.881465263383433e-06, "loss": 0.0932, "step": 9580 }, { "epoch": 6.129878438899552, "grad_norm": 0.9430817836265776, "learning_rate": 6.879496514256134e-06, "loss": 0.086, "step": 9581 }, { "epoch": 6.130518234165067, "grad_norm": 0.8936107347987372, "learning_rate": 6.877527899117704e-06, "loss": 0.0875, "step": 9582 }, { "epoch": 6.1311580294305825, "grad_norm": 0.8621974650036668, "learning_rate": 6.875559418052668e-06, "loss": 0.0794, "step": 9583 }, { "epoch": 6.131797824696097, "grad_norm": 0.9423945901405174, "learning_rate": 6.873591071145556e-06, "loss": 0.0824, "step": 9584 }, { "epoch": 6.132437619961612, "grad_norm": 0.8808389339833419, "learning_rate": 6.871622858480882e-06, "loss": 0.0842, "step": 9585 }, { "epoch": 6.1330774152271275, "grad_norm": 0.8399078127511158, "learning_rate": 6.869654780143158e-06, "loss": 0.0782, "step": 9586 }, { "epoch": 6.133717210492643, "grad_norm": 0.8637475742075394, "learning_rate": 6.867686836216887e-06, "loss": 0.0836, "step": 9587 }, { "epoch": 6.134357005758157, "grad_norm": 1.0053619375785534, "learning_rate": 6.865719026786573e-06, "loss": 0.0842, "step": 9588 }, { "epoch": 6.134996801023672, "grad_norm": 0.8945524808748592, "learning_rate": 6.863751351936708e-06, "loss": 0.0872, "step": 9589 }, { "epoch": 6.135636596289188, "grad_norm": 0.8728814172518682, "learning_rate": 6.861783811751779e-06, "loss": 0.0852, "step": 9590 }, { "epoch": 6.136276391554702, "grad_norm": 0.9480892418357846, "learning_rate": 6.859816406316273e-06, "loss": 0.089, "step": 9591 }, { "epoch": 6.136916186820217, "grad_norm": 0.9760844935736823, "learning_rate": 6.857849135714664e-06, "loss": 0.0829, "step": 9592 }, { "epoch": 6.137555982085733, "grad_norm": 0.8872867722581427, "learning_rate": 6.8558820000314205e-06, "loss": 0.0846, "step": 9593 }, { "epoch": 6.138195777351248, "grad_norm": 0.7391677538498208, "learning_rate": 6.853914999351014e-06, "loss": 0.0734, "step": 9594 }, { "epoch": 6.138835572616762, "grad_norm": 0.7924133849974729, "learning_rate": 6.851948133757904e-06, "loss": 0.0866, "step": 9595 }, { "epoch": 6.1394753678822775, "grad_norm": 0.7837226791042938, "learning_rate": 6.849981403336539e-06, "loss": 0.0761, "step": 9596 }, { "epoch": 6.140115163147793, "grad_norm": 0.9152477022947288, "learning_rate": 6.848014808171366e-06, "loss": 0.0872, "step": 9597 }, { "epoch": 6.140754958413308, "grad_norm": 0.7756263016226456, "learning_rate": 6.846048348346834e-06, "loss": 0.0777, "step": 9598 }, { "epoch": 6.1413947536788225, "grad_norm": 0.868626019492698, "learning_rate": 6.844082023947376e-06, "loss": 0.0885, "step": 9599 }, { "epoch": 6.142034548944338, "grad_norm": 0.7998409133790553, "learning_rate": 6.842115835057417e-06, "loss": 0.0801, "step": 9600 }, { "epoch": 6.142674344209853, "grad_norm": 0.8864189477555738, "learning_rate": 6.8401497817613895e-06, "loss": 0.0921, "step": 9601 }, { "epoch": 6.143314139475368, "grad_norm": 0.9452601035239194, "learning_rate": 6.83818386414371e-06, "loss": 0.0856, "step": 9602 }, { "epoch": 6.143953934740883, "grad_norm": 0.8354452968130265, "learning_rate": 6.83621808228879e-06, "loss": 0.0855, "step": 9603 }, { "epoch": 6.144593730006398, "grad_norm": 0.981153264328192, "learning_rate": 6.834252436281039e-06, "loss": 0.09, "step": 9604 }, { "epoch": 6.145233525271913, "grad_norm": 0.816118005169293, "learning_rate": 6.832286926204857e-06, "loss": 0.0829, "step": 9605 }, { "epoch": 6.145873320537428, "grad_norm": 0.7935976423770846, "learning_rate": 6.830321552144638e-06, "loss": 0.0833, "step": 9606 }, { "epoch": 6.146513115802943, "grad_norm": 0.7852458595338918, "learning_rate": 6.828356314184772e-06, "loss": 0.0763, "step": 9607 }, { "epoch": 6.147152911068458, "grad_norm": 0.8702797276530191, "learning_rate": 6.826391212409647e-06, "loss": 0.087, "step": 9608 }, { "epoch": 6.1477927063339735, "grad_norm": 0.8479870353991399, "learning_rate": 6.824426246903637e-06, "loss": 0.085, "step": 9609 }, { "epoch": 6.148432501599488, "grad_norm": 0.9144760698215357, "learning_rate": 6.822461417751114e-06, "loss": 0.084, "step": 9610 }, { "epoch": 6.149072296865003, "grad_norm": 0.8428422473037608, "learning_rate": 6.820496725036446e-06, "loss": 0.0872, "step": 9611 }, { "epoch": 6.149712092130518, "grad_norm": 0.8765075207690189, "learning_rate": 6.8185321688439945e-06, "loss": 0.0844, "step": 9612 }, { "epoch": 6.150351887396034, "grad_norm": 0.943179573702618, "learning_rate": 6.816567749258106e-06, "loss": 0.0895, "step": 9613 }, { "epoch": 6.150991682661548, "grad_norm": 0.8241503879138349, "learning_rate": 6.814603466363141e-06, "loss": 0.0866, "step": 9614 }, { "epoch": 6.151631477927063, "grad_norm": 0.7809365545465127, "learning_rate": 6.812639320243437e-06, "loss": 0.0846, "step": 9615 }, { "epoch": 6.152271273192579, "grad_norm": 0.97866007426194, "learning_rate": 6.81067531098333e-06, "loss": 0.0896, "step": 9616 }, { "epoch": 6.152911068458093, "grad_norm": 0.892149346593201, "learning_rate": 6.80871143866715e-06, "loss": 0.0846, "step": 9617 }, { "epoch": 6.153550863723608, "grad_norm": 0.9233128246059387, "learning_rate": 6.806747703379226e-06, "loss": 0.0857, "step": 9618 }, { "epoch": 6.1541906589891235, "grad_norm": 0.8620792098265639, "learning_rate": 6.804784105203876e-06, "loss": 0.0841, "step": 9619 }, { "epoch": 6.154830454254639, "grad_norm": 0.8050488930206193, "learning_rate": 6.8028206442254076e-06, "loss": 0.0852, "step": 9620 }, { "epoch": 6.155470249520153, "grad_norm": 0.8043179283056935, "learning_rate": 6.800857320528137e-06, "loss": 0.0801, "step": 9621 }, { "epoch": 6.1561100447856685, "grad_norm": 0.9089686674980894, "learning_rate": 6.798894134196363e-06, "loss": 0.0826, "step": 9622 }, { "epoch": 6.156749840051184, "grad_norm": 0.8164295225479178, "learning_rate": 6.796931085314376e-06, "loss": 0.0871, "step": 9623 }, { "epoch": 6.157389635316699, "grad_norm": 0.8921374722249561, "learning_rate": 6.794968173966475e-06, "loss": 0.0789, "step": 9624 }, { "epoch": 6.158029430582213, "grad_norm": 0.9825512452864467, "learning_rate": 6.79300540023694e-06, "loss": 0.0902, "step": 9625 }, { "epoch": 6.158669225847729, "grad_norm": 0.9628805687287292, "learning_rate": 6.791042764210048e-06, "loss": 0.0839, "step": 9626 }, { "epoch": 6.159309021113244, "grad_norm": 0.8537732636375565, "learning_rate": 6.7890802659700685e-06, "loss": 0.0897, "step": 9627 }, { "epoch": 6.159948816378758, "grad_norm": 0.8614225676522652, "learning_rate": 6.787117905601273e-06, "loss": 0.0845, "step": 9628 }, { "epoch": 6.160588611644274, "grad_norm": 0.8952822922223808, "learning_rate": 6.785155683187919e-06, "loss": 0.0825, "step": 9629 }, { "epoch": 6.161228406909789, "grad_norm": 0.7710404073108916, "learning_rate": 6.783193598814257e-06, "loss": 0.0802, "step": 9630 }, { "epoch": 6.161868202175304, "grad_norm": 0.9134561036645193, "learning_rate": 6.781231652564544e-06, "loss": 0.0844, "step": 9631 }, { "epoch": 6.162507997440819, "grad_norm": 0.8307155195671725, "learning_rate": 6.779269844523019e-06, "loss": 0.0815, "step": 9632 }, { "epoch": 6.163147792706334, "grad_norm": 0.8108286882209967, "learning_rate": 6.777308174773916e-06, "loss": 0.0854, "step": 9633 }, { "epoch": 6.163787587971849, "grad_norm": 0.7530888494163838, "learning_rate": 6.7753466434014666e-06, "loss": 0.0792, "step": 9634 }, { "epoch": 6.164427383237364, "grad_norm": 0.8824182772527877, "learning_rate": 6.773385250489899e-06, "loss": 0.0823, "step": 9635 }, { "epoch": 6.165067178502879, "grad_norm": 0.8007081346159951, "learning_rate": 6.771423996123427e-06, "loss": 0.0849, "step": 9636 }, { "epoch": 6.165706973768394, "grad_norm": 0.7839092373504415, "learning_rate": 6.769462880386262e-06, "loss": 0.0806, "step": 9637 }, { "epoch": 6.166346769033909, "grad_norm": 0.7702454014815655, "learning_rate": 6.767501903362619e-06, "loss": 0.0745, "step": 9638 }, { "epoch": 6.166986564299425, "grad_norm": 0.9190035701779352, "learning_rate": 6.7655410651366925e-06, "loss": 0.0864, "step": 9639 }, { "epoch": 6.167626359564939, "grad_norm": 0.760662123851886, "learning_rate": 6.763580365792678e-06, "loss": 0.0787, "step": 9640 }, { "epoch": 6.168266154830454, "grad_norm": 0.8634958344319978, "learning_rate": 6.761619805414768e-06, "loss": 0.0876, "step": 9641 }, { "epoch": 6.1689059500959695, "grad_norm": 0.813885895334418, "learning_rate": 6.759659384087143e-06, "loss": 0.0858, "step": 9642 }, { "epoch": 6.169545745361484, "grad_norm": 0.8710465403265689, "learning_rate": 6.757699101893974e-06, "loss": 0.0864, "step": 9643 }, { "epoch": 6.170185540626999, "grad_norm": 0.8782190251149358, "learning_rate": 6.755738958919444e-06, "loss": 0.0846, "step": 9644 }, { "epoch": 6.1708253358925145, "grad_norm": 0.9134765817854124, "learning_rate": 6.753778955247711e-06, "loss": 0.0906, "step": 9645 }, { "epoch": 6.17146513115803, "grad_norm": 0.8791423286332553, "learning_rate": 6.751819090962934e-06, "loss": 0.0866, "step": 9646 }, { "epoch": 6.172104926423544, "grad_norm": 0.8866817152414755, "learning_rate": 6.749859366149267e-06, "loss": 0.0817, "step": 9647 }, { "epoch": 6.172744721689059, "grad_norm": 0.9287856575097657, "learning_rate": 6.747899780890858e-06, "loss": 0.089, "step": 9648 }, { "epoch": 6.173384516954575, "grad_norm": 0.9343116741573875, "learning_rate": 6.74594033527185e-06, "loss": 0.0948, "step": 9649 }, { "epoch": 6.17402431222009, "grad_norm": 0.8782577162397484, "learning_rate": 6.74398102937637e-06, "loss": 0.0844, "step": 9650 }, { "epoch": 6.174664107485604, "grad_norm": 0.8747950216589918, "learning_rate": 6.742021863288555e-06, "loss": 0.0842, "step": 9651 }, { "epoch": 6.17530390275112, "grad_norm": 0.9553392777379311, "learning_rate": 6.740062837092527e-06, "loss": 0.0888, "step": 9652 }, { "epoch": 6.175943698016635, "grad_norm": 0.8866121369768474, "learning_rate": 6.738103950872397e-06, "loss": 0.0832, "step": 9653 }, { "epoch": 6.176583493282149, "grad_norm": 0.815351332043313, "learning_rate": 6.736145204712287e-06, "loss": 0.0828, "step": 9654 }, { "epoch": 6.177223288547665, "grad_norm": 0.8998046965718641, "learning_rate": 6.734186598696294e-06, "loss": 0.0931, "step": 9655 }, { "epoch": 6.17786308381318, "grad_norm": 0.8963590418449904, "learning_rate": 6.7322281329085204e-06, "loss": 0.0877, "step": 9656 }, { "epoch": 6.178502879078695, "grad_norm": 0.724951673223811, "learning_rate": 6.730269807433054e-06, "loss": 0.0822, "step": 9657 }, { "epoch": 6.1791426743442095, "grad_norm": 0.8843954019423997, "learning_rate": 6.728311622353989e-06, "loss": 0.0872, "step": 9658 }, { "epoch": 6.179782469609725, "grad_norm": 0.8328034593808903, "learning_rate": 6.726353577755402e-06, "loss": 0.0823, "step": 9659 }, { "epoch": 6.18042226487524, "grad_norm": 0.8064611508774644, "learning_rate": 6.7243956737213665e-06, "loss": 0.0773, "step": 9660 }, { "epoch": 6.181062060140755, "grad_norm": 0.9568772448053634, "learning_rate": 6.722437910335957e-06, "loss": 0.0855, "step": 9661 }, { "epoch": 6.18170185540627, "grad_norm": 0.7972108427093275, "learning_rate": 6.720480287683233e-06, "loss": 0.0869, "step": 9662 }, { "epoch": 6.182341650671785, "grad_norm": 0.8464380181546409, "learning_rate": 6.718522805847253e-06, "loss": 0.0802, "step": 9663 }, { "epoch": 6.1829814459373, "grad_norm": 0.8830691315293588, "learning_rate": 6.716565464912063e-06, "loss": 0.0895, "step": 9664 }, { "epoch": 6.183621241202815, "grad_norm": 0.8871958900537995, "learning_rate": 6.714608264961714e-06, "loss": 0.0829, "step": 9665 }, { "epoch": 6.18426103646833, "grad_norm": 0.9419606110527863, "learning_rate": 6.712651206080241e-06, "loss": 0.0871, "step": 9666 }, { "epoch": 6.184900831733845, "grad_norm": 0.9537749322170315, "learning_rate": 6.710694288351675e-06, "loss": 0.0815, "step": 9667 }, { "epoch": 6.1855406269993605, "grad_norm": 0.9005864846722037, "learning_rate": 6.708737511860048e-06, "loss": 0.0878, "step": 9668 }, { "epoch": 6.186180422264875, "grad_norm": 0.9791136721112372, "learning_rate": 6.706780876689377e-06, "loss": 0.0846, "step": 9669 }, { "epoch": 6.18682021753039, "grad_norm": 0.795215628523619, "learning_rate": 6.704824382923676e-06, "loss": 0.0774, "step": 9670 }, { "epoch": 6.187460012795905, "grad_norm": 1.0227852491343805, "learning_rate": 6.702868030646956e-06, "loss": 0.0917, "step": 9671 }, { "epoch": 6.188099808061421, "grad_norm": 0.941280595931747, "learning_rate": 6.700911819943216e-06, "loss": 0.0842, "step": 9672 }, { "epoch": 6.188739603326935, "grad_norm": 0.8806826709371459, "learning_rate": 6.698955750896457e-06, "loss": 0.087, "step": 9673 }, { "epoch": 6.18937939859245, "grad_norm": 0.8929989311879414, "learning_rate": 6.69699982359066e-06, "loss": 0.0901, "step": 9674 }, { "epoch": 6.190019193857966, "grad_norm": 0.8441489289175781, "learning_rate": 6.695044038109819e-06, "loss": 0.0795, "step": 9675 }, { "epoch": 6.190658989123481, "grad_norm": 0.9107188220982921, "learning_rate": 6.693088394537908e-06, "loss": 0.0845, "step": 9676 }, { "epoch": 6.191298784388995, "grad_norm": 0.8681370926005922, "learning_rate": 6.691132892958897e-06, "loss": 0.0867, "step": 9677 }, { "epoch": 6.1919385796545106, "grad_norm": 0.8197221850940757, "learning_rate": 6.6891775334567566e-06, "loss": 0.0849, "step": 9678 }, { "epoch": 6.192578374920026, "grad_norm": 0.8810335511579945, "learning_rate": 6.6872223161154414e-06, "loss": 0.0836, "step": 9679 }, { "epoch": 6.19321817018554, "grad_norm": 0.8083184745903906, "learning_rate": 6.685267241018904e-06, "loss": 0.0838, "step": 9680 }, { "epoch": 6.1938579654510555, "grad_norm": 0.8921260836953493, "learning_rate": 6.683312308251099e-06, "loss": 0.0873, "step": 9681 }, { "epoch": 6.194497760716571, "grad_norm": 0.9246839811900919, "learning_rate": 6.6813575178959635e-06, "loss": 0.0862, "step": 9682 }, { "epoch": 6.195137555982086, "grad_norm": 0.8893416956873573, "learning_rate": 6.6794028700374325e-06, "loss": 0.0723, "step": 9683 }, { "epoch": 6.1957773512476, "grad_norm": 0.8884810824846003, "learning_rate": 6.6774483647594314e-06, "loss": 0.0796, "step": 9684 }, { "epoch": 6.196417146513116, "grad_norm": 0.8861281370230694, "learning_rate": 6.675494002145891e-06, "loss": 0.083, "step": 9685 }, { "epoch": 6.197056941778631, "grad_norm": 0.895370961657227, "learning_rate": 6.673539782280725e-06, "loss": 0.088, "step": 9686 }, { "epoch": 6.197696737044146, "grad_norm": 0.9772746930487214, "learning_rate": 6.671585705247841e-06, "loss": 0.0894, "step": 9687 }, { "epoch": 6.198336532309661, "grad_norm": 0.8611980568630413, "learning_rate": 6.6696317711311484e-06, "loss": 0.0891, "step": 9688 }, { "epoch": 6.198976327575176, "grad_norm": 0.9458692791620696, "learning_rate": 6.667677980014542e-06, "loss": 0.0872, "step": 9689 }, { "epoch": 6.199616122840691, "grad_norm": 0.8061006875846533, "learning_rate": 6.6657243319819135e-06, "loss": 0.0849, "step": 9690 }, { "epoch": 6.2002559181062065, "grad_norm": 0.867370814727174, "learning_rate": 6.663770827117153e-06, "loss": 0.0842, "step": 9691 }, { "epoch": 6.200895713371721, "grad_norm": 0.859433966247611, "learning_rate": 6.661817465504139e-06, "loss": 0.0803, "step": 9692 }, { "epoch": 6.201535508637236, "grad_norm": 0.8664736569810273, "learning_rate": 6.6598642472267465e-06, "loss": 0.0925, "step": 9693 }, { "epoch": 6.202175303902751, "grad_norm": 0.8686024496975748, "learning_rate": 6.657911172368837e-06, "loss": 0.0801, "step": 9694 }, { "epoch": 6.202815099168266, "grad_norm": 0.7931973864525954, "learning_rate": 6.65595824101428e-06, "loss": 0.088, "step": 9695 }, { "epoch": 6.203454894433781, "grad_norm": 0.8854337820515022, "learning_rate": 6.654005453246927e-06, "loss": 0.0847, "step": 9696 }, { "epoch": 6.204094689699296, "grad_norm": 0.7991103082613568, "learning_rate": 6.652052809150624e-06, "loss": 0.081, "step": 9697 }, { "epoch": 6.204734484964812, "grad_norm": 0.7644560185074268, "learning_rate": 6.650100308809222e-06, "loss": 0.0799, "step": 9698 }, { "epoch": 6.205374280230326, "grad_norm": 0.8142447698206025, "learning_rate": 6.648147952306554e-06, "loss": 0.0827, "step": 9699 }, { "epoch": 6.206014075495841, "grad_norm": 0.8773660978223368, "learning_rate": 6.646195739726448e-06, "loss": 0.0922, "step": 9700 }, { "epoch": 6.2066538707613566, "grad_norm": 0.784000798169538, "learning_rate": 6.644243671152734e-06, "loss": 0.0782, "step": 9701 }, { "epoch": 6.207293666026872, "grad_norm": 0.8027719920171482, "learning_rate": 6.642291746669226e-06, "loss": 0.0798, "step": 9702 }, { "epoch": 6.207933461292386, "grad_norm": 0.8886241969806107, "learning_rate": 6.640339966359739e-06, "loss": 0.0805, "step": 9703 }, { "epoch": 6.2085732565579015, "grad_norm": 0.9081067861615885, "learning_rate": 6.638388330308072e-06, "loss": 0.0852, "step": 9704 }, { "epoch": 6.209213051823417, "grad_norm": 0.9017166357954288, "learning_rate": 6.636436838598035e-06, "loss": 0.0819, "step": 9705 }, { "epoch": 6.209852847088931, "grad_norm": 0.8159545604460159, "learning_rate": 6.634485491313416e-06, "loss": 0.0824, "step": 9706 }, { "epoch": 6.210492642354446, "grad_norm": 0.8198355773467397, "learning_rate": 6.632534288538002e-06, "loss": 0.0803, "step": 9707 }, { "epoch": 6.211132437619962, "grad_norm": 0.9203618485943043, "learning_rate": 6.630583230355576e-06, "loss": 0.0773, "step": 9708 }, { "epoch": 6.211772232885477, "grad_norm": 0.8558102119782146, "learning_rate": 6.628632316849915e-06, "loss": 0.0754, "step": 9709 }, { "epoch": 6.212412028150991, "grad_norm": 0.8358698952155126, "learning_rate": 6.626681548104779e-06, "loss": 0.0813, "step": 9710 }, { "epoch": 6.213051823416507, "grad_norm": 0.8998188088607502, "learning_rate": 6.624730924203941e-06, "loss": 0.0854, "step": 9711 }, { "epoch": 6.213691618682022, "grad_norm": 0.9391791318457983, "learning_rate": 6.622780445231154e-06, "loss": 0.0915, "step": 9712 }, { "epoch": 6.214331413947537, "grad_norm": 0.7648625105207951, "learning_rate": 6.620830111270167e-06, "loss": 0.0822, "step": 9713 }, { "epoch": 6.214971209213052, "grad_norm": 0.8475174501372241, "learning_rate": 6.618879922404718e-06, "loss": 0.0826, "step": 9714 }, { "epoch": 6.215611004478567, "grad_norm": 0.9348389010483542, "learning_rate": 6.616929878718556e-06, "loss": 0.0862, "step": 9715 }, { "epoch": 6.216250799744082, "grad_norm": 0.7959934352758289, "learning_rate": 6.614979980295407e-06, "loss": 0.0884, "step": 9716 }, { "epoch": 6.2168905950095965, "grad_norm": 0.901251146783366, "learning_rate": 6.613030227218994e-06, "loss": 0.0861, "step": 9717 }, { "epoch": 6.217530390275112, "grad_norm": 0.9700267417829812, "learning_rate": 6.61108061957304e-06, "loss": 0.0931, "step": 9718 }, { "epoch": 6.218170185540627, "grad_norm": 0.8126073447973291, "learning_rate": 6.609131157441256e-06, "loss": 0.0824, "step": 9719 }, { "epoch": 6.218809980806142, "grad_norm": 0.7827971215036885, "learning_rate": 6.6071818409073445e-06, "loss": 0.0843, "step": 9720 }, { "epoch": 6.219449776071657, "grad_norm": 0.9296751819071752, "learning_rate": 6.605232670055013e-06, "loss": 0.0929, "step": 9721 }, { "epoch": 6.220089571337172, "grad_norm": 0.7769363737138508, "learning_rate": 6.6032836449679515e-06, "loss": 0.0815, "step": 9722 }, { "epoch": 6.220729366602687, "grad_norm": 0.8960218406127897, "learning_rate": 6.601334765729849e-06, "loss": 0.0849, "step": 9723 }, { "epoch": 6.2213691618682025, "grad_norm": 0.8782514821088915, "learning_rate": 6.599386032424383e-06, "loss": 0.0826, "step": 9724 }, { "epoch": 6.222008957133717, "grad_norm": 0.9238369657742538, "learning_rate": 6.597437445135233e-06, "loss": 0.0883, "step": 9725 }, { "epoch": 6.222648752399232, "grad_norm": 0.8397541446430015, "learning_rate": 6.595489003946067e-06, "loss": 0.088, "step": 9726 }, { "epoch": 6.2232885476647475, "grad_norm": 0.8532106832664447, "learning_rate": 6.593540708940545e-06, "loss": 0.0825, "step": 9727 }, { "epoch": 6.223928342930263, "grad_norm": 0.8984286711805318, "learning_rate": 6.591592560202326e-06, "loss": 0.0845, "step": 9728 }, { "epoch": 6.224568138195777, "grad_norm": 0.9522974414587646, "learning_rate": 6.589644557815061e-06, "loss": 0.086, "step": 9729 }, { "epoch": 6.225207933461292, "grad_norm": 0.9500561625685543, "learning_rate": 6.587696701862391e-06, "loss": 0.0871, "step": 9730 }, { "epoch": 6.225847728726808, "grad_norm": 0.8435656297384425, "learning_rate": 6.585748992427956e-06, "loss": 0.08, "step": 9731 }, { "epoch": 6.226487523992322, "grad_norm": 0.9187904539379002, "learning_rate": 6.583801429595384e-06, "loss": 0.0799, "step": 9732 }, { "epoch": 6.227127319257837, "grad_norm": 1.0090329879643503, "learning_rate": 6.581854013448304e-06, "loss": 0.0916, "step": 9733 }, { "epoch": 6.227767114523353, "grad_norm": 0.8331647957013547, "learning_rate": 6.579906744070328e-06, "loss": 0.0818, "step": 9734 }, { "epoch": 6.228406909788868, "grad_norm": 0.8973490642896942, "learning_rate": 6.577959621545076e-06, "loss": 0.0843, "step": 9735 }, { "epoch": 6.229046705054382, "grad_norm": 0.9490032523145153, "learning_rate": 6.576012645956149e-06, "loss": 0.0854, "step": 9736 }, { "epoch": 6.229686500319898, "grad_norm": 0.9298007252733005, "learning_rate": 6.5740658173871485e-06, "loss": 0.0875, "step": 9737 }, { "epoch": 6.230326295585413, "grad_norm": 0.8792761961947304, "learning_rate": 6.572119135921668e-06, "loss": 0.0853, "step": 9738 }, { "epoch": 6.230966090850928, "grad_norm": 0.8350796042307911, "learning_rate": 6.5701726016432955e-06, "loss": 0.0877, "step": 9739 }, { "epoch": 6.2316058861164425, "grad_norm": 0.909771097160091, "learning_rate": 6.568226214635609e-06, "loss": 0.0839, "step": 9740 }, { "epoch": 6.232245681381958, "grad_norm": 0.7486299871355383, "learning_rate": 6.566279974982182e-06, "loss": 0.0765, "step": 9741 }, { "epoch": 6.232885476647473, "grad_norm": 0.7843959449238176, "learning_rate": 6.564333882766587e-06, "loss": 0.0859, "step": 9742 }, { "epoch": 6.233525271912987, "grad_norm": 0.9412896740800445, "learning_rate": 6.562387938072383e-06, "loss": 0.0857, "step": 9743 }, { "epoch": 6.234165067178503, "grad_norm": 0.8171069233254153, "learning_rate": 6.560442140983123e-06, "loss": 0.0856, "step": 9744 }, { "epoch": 6.234804862444018, "grad_norm": 0.9686418641866718, "learning_rate": 6.558496491582362e-06, "loss": 0.0912, "step": 9745 }, { "epoch": 6.235444657709533, "grad_norm": 0.9003756180931711, "learning_rate": 6.556550989953641e-06, "loss": 0.0831, "step": 9746 }, { "epoch": 6.236084452975048, "grad_norm": 0.8751584535611192, "learning_rate": 6.554605636180493e-06, "loss": 0.0851, "step": 9747 }, { "epoch": 6.236724248240563, "grad_norm": 0.9390336685121685, "learning_rate": 6.5526604303464515e-06, "loss": 0.0919, "step": 9748 }, { "epoch": 6.237364043506078, "grad_norm": 0.8115883056568166, "learning_rate": 6.5507153725350395e-06, "loss": 0.0882, "step": 9749 }, { "epoch": 6.2380038387715935, "grad_norm": 0.9329149680771004, "learning_rate": 6.5487704628297735e-06, "loss": 0.0915, "step": 9750 }, { "epoch": 6.238643634037108, "grad_norm": 0.9657228424908184, "learning_rate": 6.546825701314159e-06, "loss": 0.0868, "step": 9751 }, { "epoch": 6.239283429302623, "grad_norm": 0.9730001690357191, "learning_rate": 6.544881088071712e-06, "loss": 0.088, "step": 9752 }, { "epoch": 6.239923224568138, "grad_norm": 0.9238617921937898, "learning_rate": 6.542936623185925e-06, "loss": 0.0883, "step": 9753 }, { "epoch": 6.240563019833653, "grad_norm": 0.9551241750962817, "learning_rate": 6.540992306740286e-06, "loss": 0.0851, "step": 9754 }, { "epoch": 6.241202815099168, "grad_norm": 0.9474557272354056, "learning_rate": 6.539048138818287e-06, "loss": 0.0851, "step": 9755 }, { "epoch": 6.241842610364683, "grad_norm": 0.8273613951786747, "learning_rate": 6.537104119503406e-06, "loss": 0.0826, "step": 9756 }, { "epoch": 6.242482405630199, "grad_norm": 1.0002363618774568, "learning_rate": 6.535160248879108e-06, "loss": 0.086, "step": 9757 }, { "epoch": 6.243122200895713, "grad_norm": 0.9918817697351826, "learning_rate": 6.533216527028869e-06, "loss": 0.0925, "step": 9758 }, { "epoch": 6.243761996161228, "grad_norm": 1.0318209563970406, "learning_rate": 6.5312729540361475e-06, "loss": 0.0885, "step": 9759 }, { "epoch": 6.244401791426744, "grad_norm": 0.8666362475153809, "learning_rate": 6.529329529984394e-06, "loss": 0.0812, "step": 9760 }, { "epoch": 6.245041586692259, "grad_norm": 0.8262304124817808, "learning_rate": 6.527386254957055e-06, "loss": 0.0807, "step": 9761 }, { "epoch": 6.245681381957773, "grad_norm": 0.8827023174282166, "learning_rate": 6.525443129037574e-06, "loss": 0.078, "step": 9762 }, { "epoch": 6.2463211772232885, "grad_norm": 0.8918447111168563, "learning_rate": 6.5235001523093856e-06, "loss": 0.0881, "step": 9763 }, { "epoch": 6.246960972488804, "grad_norm": 0.9060778300494655, "learning_rate": 6.521557324855912e-06, "loss": 0.0896, "step": 9764 }, { "epoch": 6.247600767754319, "grad_norm": 0.8838343990383235, "learning_rate": 6.519614646760582e-06, "loss": 0.082, "step": 9765 }, { "epoch": 6.248240563019833, "grad_norm": 0.9487697164346245, "learning_rate": 6.5176721181068085e-06, "loss": 0.0877, "step": 9766 }, { "epoch": 6.248880358285349, "grad_norm": 0.8487209389992888, "learning_rate": 6.515729738977998e-06, "loss": 0.0817, "step": 9767 }, { "epoch": 6.249520153550864, "grad_norm": 0.8662375631879491, "learning_rate": 6.513787509457557e-06, "loss": 0.0916, "step": 9768 }, { "epoch": 6.250159948816378, "grad_norm": 0.8232742861298733, "learning_rate": 6.511845429628878e-06, "loss": 0.082, "step": 9769 }, { "epoch": 6.250799744081894, "grad_norm": 0.8372960567675041, "learning_rate": 6.509903499575352e-06, "loss": 0.0776, "step": 9770 }, { "epoch": 6.251439539347409, "grad_norm": 0.8223756180573414, "learning_rate": 6.507961719380356e-06, "loss": 0.089, "step": 9771 }, { "epoch": 6.252079334612924, "grad_norm": 0.9014406775492028, "learning_rate": 6.506020089127277e-06, "loss": 0.0927, "step": 9772 }, { "epoch": 6.252719129878439, "grad_norm": 0.7785881224390108, "learning_rate": 6.50407860889948e-06, "loss": 0.0848, "step": 9773 }, { "epoch": 6.253358925143954, "grad_norm": 0.9077097416893629, "learning_rate": 6.502137278780325e-06, "loss": 0.0888, "step": 9774 }, { "epoch": 6.253998720409469, "grad_norm": 0.8417822515423228, "learning_rate": 6.5001960988531765e-06, "loss": 0.0842, "step": 9775 }, { "epoch": 6.254638515674984, "grad_norm": 0.943839987269391, "learning_rate": 6.4982550692013826e-06, "loss": 0.0867, "step": 9776 }, { "epoch": 6.255278310940499, "grad_norm": 1.0204586525972235, "learning_rate": 6.496314189908284e-06, "loss": 0.0871, "step": 9777 }, { "epoch": 6.255918106206014, "grad_norm": 0.910518038515042, "learning_rate": 6.494373461057224e-06, "loss": 0.0916, "step": 9778 }, { "epoch": 6.256557901471529, "grad_norm": 0.9438882246187337, "learning_rate": 6.492432882731532e-06, "loss": 0.0905, "step": 9779 }, { "epoch": 6.257197696737044, "grad_norm": 0.7787875133474563, "learning_rate": 6.490492455014532e-06, "loss": 0.0799, "step": 9780 }, { "epoch": 6.257837492002559, "grad_norm": 0.7904550819653138, "learning_rate": 6.488552177989537e-06, "loss": 0.0824, "step": 9781 }, { "epoch": 6.258477287268074, "grad_norm": 0.8231202458645169, "learning_rate": 6.486612051739872e-06, "loss": 0.0855, "step": 9782 }, { "epoch": 6.25911708253359, "grad_norm": 0.7511406775827066, "learning_rate": 6.484672076348834e-06, "loss": 0.0752, "step": 9783 }, { "epoch": 6.259756877799104, "grad_norm": 0.8504484266042512, "learning_rate": 6.4827322518997235e-06, "loss": 0.0819, "step": 9784 }, { "epoch": 6.260396673064619, "grad_norm": 0.8399835085375223, "learning_rate": 6.480792578475834e-06, "loss": 0.083, "step": 9785 }, { "epoch": 6.2610364683301345, "grad_norm": 0.9188741527814474, "learning_rate": 6.4788530561604525e-06, "loss": 0.0867, "step": 9786 }, { "epoch": 6.26167626359565, "grad_norm": 0.8728594990973828, "learning_rate": 6.476913685036852e-06, "loss": 0.0833, "step": 9787 }, { "epoch": 6.262316058861164, "grad_norm": 0.8481914871214494, "learning_rate": 6.474974465188315e-06, "loss": 0.0809, "step": 9788 }, { "epoch": 6.262955854126679, "grad_norm": 0.9240651517938584, "learning_rate": 6.473035396698104e-06, "loss": 0.083, "step": 9789 }, { "epoch": 6.263595649392195, "grad_norm": 0.9087123781039135, "learning_rate": 6.471096479649479e-06, "loss": 0.0897, "step": 9790 }, { "epoch": 6.264235444657709, "grad_norm": 0.8602223633891884, "learning_rate": 6.469157714125691e-06, "loss": 0.0771, "step": 9791 }, { "epoch": 6.264875239923224, "grad_norm": 0.9337882343679424, "learning_rate": 6.467219100209993e-06, "loss": 0.0895, "step": 9792 }, { "epoch": 6.26551503518874, "grad_norm": 0.8941837796980385, "learning_rate": 6.465280637985622e-06, "loss": 0.0966, "step": 9793 }, { "epoch": 6.266154830454255, "grad_norm": 0.873967484238959, "learning_rate": 6.463342327535809e-06, "loss": 0.0856, "step": 9794 }, { "epoch": 6.266794625719769, "grad_norm": 0.9475058961494726, "learning_rate": 6.461404168943788e-06, "loss": 0.0828, "step": 9795 }, { "epoch": 6.267434420985285, "grad_norm": 0.8029678669463491, "learning_rate": 6.459466162292778e-06, "loss": 0.0852, "step": 9796 }, { "epoch": 6.2680742162508, "grad_norm": 0.8815543162628059, "learning_rate": 6.457528307665991e-06, "loss": 0.086, "step": 9797 }, { "epoch": 6.268714011516315, "grad_norm": 0.930428386483308, "learning_rate": 6.4555906051466385e-06, "loss": 0.0896, "step": 9798 }, { "epoch": 6.2693538067818295, "grad_norm": 0.8112605292621796, "learning_rate": 6.453653054817921e-06, "loss": 0.0782, "step": 9799 }, { "epoch": 6.269993602047345, "grad_norm": 0.922049782370903, "learning_rate": 6.451715656763033e-06, "loss": 0.0876, "step": 9800 }, { "epoch": 6.27063339731286, "grad_norm": 0.9517310862180589, "learning_rate": 6.449778411065158e-06, "loss": 0.0896, "step": 9801 }, { "epoch": 6.271273192578375, "grad_norm": 0.9607874110503014, "learning_rate": 6.447841317807487e-06, "loss": 0.0853, "step": 9802 }, { "epoch": 6.27191298784389, "grad_norm": 0.932057707833958, "learning_rate": 6.445904377073191e-06, "loss": 0.0885, "step": 9803 }, { "epoch": 6.272552783109405, "grad_norm": 0.9640182134817893, "learning_rate": 6.4439675889454355e-06, "loss": 0.0884, "step": 9804 }, { "epoch": 6.27319257837492, "grad_norm": 0.8790094586567517, "learning_rate": 6.44203095350739e-06, "loss": 0.0813, "step": 9805 }, { "epoch": 6.273832373640435, "grad_norm": 0.8993935243673669, "learning_rate": 6.440094470842205e-06, "loss": 0.0919, "step": 9806 }, { "epoch": 6.27447216890595, "grad_norm": 0.8994702093421941, "learning_rate": 6.438158141033029e-06, "loss": 0.084, "step": 9807 }, { "epoch": 6.275111964171465, "grad_norm": 0.8982905440264999, "learning_rate": 6.43622196416301e-06, "loss": 0.0898, "step": 9808 }, { "epoch": 6.2757517594369805, "grad_norm": 0.8945222034834275, "learning_rate": 6.434285940315279e-06, "loss": 0.0916, "step": 9809 }, { "epoch": 6.276391554702495, "grad_norm": 0.8283982649089904, "learning_rate": 6.432350069572968e-06, "loss": 0.0853, "step": 9810 }, { "epoch": 6.27703134996801, "grad_norm": 0.8496484181267958, "learning_rate": 6.430414352019193e-06, "loss": 0.0875, "step": 9811 }, { "epoch": 6.277671145233525, "grad_norm": 0.9281254530143914, "learning_rate": 6.42847878773708e-06, "loss": 0.0902, "step": 9812 }, { "epoch": 6.278310940499041, "grad_norm": 0.9186126985137995, "learning_rate": 6.426543376809736e-06, "loss": 0.0867, "step": 9813 }, { "epoch": 6.278950735764555, "grad_norm": 0.8662397789101536, "learning_rate": 6.42460811932026e-06, "loss": 0.0847, "step": 9814 }, { "epoch": 6.27959053103007, "grad_norm": 0.7807375626956408, "learning_rate": 6.422673015351754e-06, "loss": 0.0835, "step": 9815 }, { "epoch": 6.280230326295586, "grad_norm": 0.7689382073731379, "learning_rate": 6.420738064987304e-06, "loss": 0.0852, "step": 9816 }, { "epoch": 6.280870121561101, "grad_norm": 0.8971967242235063, "learning_rate": 6.4188032683099944e-06, "loss": 0.0962, "step": 9817 }, { "epoch": 6.281509916826615, "grad_norm": 0.8784392982205402, "learning_rate": 6.416868625402899e-06, "loss": 0.0867, "step": 9818 }, { "epoch": 6.282149712092131, "grad_norm": 0.8470854797358204, "learning_rate": 6.414934136349095e-06, "loss": 0.0887, "step": 9819 }, { "epoch": 6.282789507357646, "grad_norm": 0.8872003630424458, "learning_rate": 6.4129998012316405e-06, "loss": 0.0807, "step": 9820 }, { "epoch": 6.28342930262316, "grad_norm": 0.8641150809598735, "learning_rate": 6.411065620133593e-06, "loss": 0.0889, "step": 9821 }, { "epoch": 6.2840690978886755, "grad_norm": 0.853724104655038, "learning_rate": 6.409131593138005e-06, "loss": 0.0889, "step": 9822 }, { "epoch": 6.284708893154191, "grad_norm": 0.9123573383362722, "learning_rate": 6.407197720327919e-06, "loss": 0.0887, "step": 9823 }, { "epoch": 6.285348688419706, "grad_norm": 0.915130328251902, "learning_rate": 6.405264001786366e-06, "loss": 0.0947, "step": 9824 }, { "epoch": 6.2859884836852205, "grad_norm": 0.9701648544664655, "learning_rate": 6.403330437596389e-06, "loss": 0.0893, "step": 9825 }, { "epoch": 6.286628278950736, "grad_norm": 0.9147341393102286, "learning_rate": 6.401397027841004e-06, "loss": 0.0948, "step": 9826 }, { "epoch": 6.287268074216251, "grad_norm": 0.7940801364705605, "learning_rate": 6.3994637726032295e-06, "loss": 0.0875, "step": 9827 }, { "epoch": 6.287907869481765, "grad_norm": 0.8846199492189517, "learning_rate": 6.3975306719660726e-06, "loss": 0.086, "step": 9828 }, { "epoch": 6.288547664747281, "grad_norm": 0.8052106229431396, "learning_rate": 6.3955977260125435e-06, "loss": 0.0869, "step": 9829 }, { "epoch": 6.289187460012796, "grad_norm": 0.8537912659456248, "learning_rate": 6.393664934825636e-06, "loss": 0.0868, "step": 9830 }, { "epoch": 6.289827255278311, "grad_norm": 0.8507454746321723, "learning_rate": 6.391732298488337e-06, "loss": 0.0833, "step": 9831 }, { "epoch": 6.290467050543826, "grad_norm": 0.8797160211370831, "learning_rate": 6.3897998170836384e-06, "loss": 0.0802, "step": 9832 }, { "epoch": 6.291106845809341, "grad_norm": 0.9777496324509816, "learning_rate": 6.387867490694513e-06, "loss": 0.0834, "step": 9833 }, { "epoch": 6.291746641074856, "grad_norm": 0.8619194777515012, "learning_rate": 6.385935319403929e-06, "loss": 0.0881, "step": 9834 }, { "epoch": 6.292386436340371, "grad_norm": 0.8854163935475833, "learning_rate": 6.384003303294856e-06, "loss": 0.0886, "step": 9835 }, { "epoch": 6.293026231605886, "grad_norm": 0.8531320539171525, "learning_rate": 6.38207144245025e-06, "loss": 0.0788, "step": 9836 }, { "epoch": 6.293666026871401, "grad_norm": 0.7532135732356509, "learning_rate": 6.380139736953061e-06, "loss": 0.0773, "step": 9837 }, { "epoch": 6.294305822136916, "grad_norm": 0.8998823493010054, "learning_rate": 6.37820818688623e-06, "loss": 0.0891, "step": 9838 }, { "epoch": 6.294945617402432, "grad_norm": 1.054632837160228, "learning_rate": 6.376276792332698e-06, "loss": 0.0853, "step": 9839 }, { "epoch": 6.295585412667946, "grad_norm": 0.8711886176874212, "learning_rate": 6.374345553375396e-06, "loss": 0.0768, "step": 9840 }, { "epoch": 6.296225207933461, "grad_norm": 0.8640011245263508, "learning_rate": 6.3724144700972414e-06, "loss": 0.0821, "step": 9841 }, { "epoch": 6.296865003198977, "grad_norm": 0.886710925429581, "learning_rate": 6.37048354258116e-06, "loss": 0.0803, "step": 9842 }, { "epoch": 6.297504798464491, "grad_norm": 0.9726527042801292, "learning_rate": 6.3685527709100605e-06, "loss": 0.0912, "step": 9843 }, { "epoch": 6.298144593730006, "grad_norm": 0.9484550419614961, "learning_rate": 6.366622155166842e-06, "loss": 0.0852, "step": 9844 }, { "epoch": 6.2987843889955215, "grad_norm": 0.9382420350896017, "learning_rate": 6.364691695434406e-06, "loss": 0.0902, "step": 9845 }, { "epoch": 6.299424184261037, "grad_norm": 0.8194963306302878, "learning_rate": 6.362761391795644e-06, "loss": 0.0863, "step": 9846 }, { "epoch": 6.300063979526551, "grad_norm": 1.002749674709401, "learning_rate": 6.360831244333437e-06, "loss": 0.0854, "step": 9847 }, { "epoch": 6.3007037747920664, "grad_norm": 0.8959032932986146, "learning_rate": 6.358901253130658e-06, "loss": 0.0869, "step": 9848 }, { "epoch": 6.301343570057582, "grad_norm": 0.7657969553434879, "learning_rate": 6.3569714182701855e-06, "loss": 0.0828, "step": 9849 }, { "epoch": 6.301983365323097, "grad_norm": 0.8777073028244214, "learning_rate": 6.3550417398348814e-06, "loss": 0.0789, "step": 9850 }, { "epoch": 6.302623160588611, "grad_norm": 0.8235318517420747, "learning_rate": 6.353112217907599e-06, "loss": 0.0884, "step": 9851 }, { "epoch": 6.303262955854127, "grad_norm": 0.887877881046022, "learning_rate": 6.351182852571192e-06, "loss": 0.0824, "step": 9852 }, { "epoch": 6.303902751119642, "grad_norm": 0.8410553780138086, "learning_rate": 6.349253643908503e-06, "loss": 0.0817, "step": 9853 }, { "epoch": 6.304542546385157, "grad_norm": 0.8201403486050346, "learning_rate": 6.347324592002364e-06, "loss": 0.0826, "step": 9854 }, { "epoch": 6.305182341650672, "grad_norm": 0.8212582417416845, "learning_rate": 6.345395696935613e-06, "loss": 0.0818, "step": 9855 }, { "epoch": 6.305822136916187, "grad_norm": 0.8747667150448311, "learning_rate": 6.343466958791071e-06, "loss": 0.0892, "step": 9856 }, { "epoch": 6.306461932181702, "grad_norm": 0.8433499648005915, "learning_rate": 6.341538377651553e-06, "loss": 0.0808, "step": 9857 }, { "epoch": 6.3071017274472165, "grad_norm": 0.7636183892859909, "learning_rate": 6.3396099535998664e-06, "loss": 0.0845, "step": 9858 }, { "epoch": 6.307741522712732, "grad_norm": 0.8009475400376268, "learning_rate": 6.3376816867188194e-06, "loss": 0.0816, "step": 9859 }, { "epoch": 6.308381317978247, "grad_norm": 0.8544394344027115, "learning_rate": 6.335753577091206e-06, "loss": 0.0916, "step": 9860 }, { "epoch": 6.309021113243762, "grad_norm": 0.9368100612385246, "learning_rate": 6.333825624799812e-06, "loss": 0.0934, "step": 9861 }, { "epoch": 6.309660908509277, "grad_norm": 0.8623730613013757, "learning_rate": 6.331897829927428e-06, "loss": 0.086, "step": 9862 }, { "epoch": 6.310300703774792, "grad_norm": 0.8094383780643021, "learning_rate": 6.329970192556827e-06, "loss": 0.0803, "step": 9863 }, { "epoch": 6.310940499040307, "grad_norm": 0.9013762545274615, "learning_rate": 6.328042712770772e-06, "loss": 0.0882, "step": 9864 }, { "epoch": 6.311580294305822, "grad_norm": 0.9179635644521111, "learning_rate": 6.3261153906520345e-06, "loss": 0.0899, "step": 9865 }, { "epoch": 6.312220089571337, "grad_norm": 1.0151963118674174, "learning_rate": 6.324188226283369e-06, "loss": 0.0889, "step": 9866 }, { "epoch": 6.312859884836852, "grad_norm": 0.8834149359997898, "learning_rate": 6.322261219747521e-06, "loss": 0.0861, "step": 9867 }, { "epoch": 6.3134996801023675, "grad_norm": 0.9781358352753561, "learning_rate": 6.3203343711272325e-06, "loss": 0.0853, "step": 9868 }, { "epoch": 6.314139475367882, "grad_norm": 0.9459928930719596, "learning_rate": 6.318407680505242e-06, "loss": 0.0917, "step": 9869 }, { "epoch": 6.314779270633397, "grad_norm": 0.8022672207957057, "learning_rate": 6.3164811479642775e-06, "loss": 0.0802, "step": 9870 }, { "epoch": 6.315419065898912, "grad_norm": 0.8402117088085147, "learning_rate": 6.314554773587054e-06, "loss": 0.0897, "step": 9871 }, { "epoch": 6.316058861164428, "grad_norm": 0.8771975417734913, "learning_rate": 6.312628557456299e-06, "loss": 0.0839, "step": 9872 }, { "epoch": 6.316698656429942, "grad_norm": 0.8524087662799616, "learning_rate": 6.310702499654713e-06, "loss": 0.0826, "step": 9873 }, { "epoch": 6.317338451695457, "grad_norm": 0.8816220949286584, "learning_rate": 6.308776600264996e-06, "loss": 0.0838, "step": 9874 }, { "epoch": 6.317978246960973, "grad_norm": 0.8818416523500756, "learning_rate": 6.3068508593698504e-06, "loss": 0.0886, "step": 9875 }, { "epoch": 6.318618042226488, "grad_norm": 1.0190787947416209, "learning_rate": 6.3049252770519575e-06, "loss": 0.0929, "step": 9876 }, { "epoch": 6.319257837492002, "grad_norm": 0.8448741930182876, "learning_rate": 6.302999853394001e-06, "loss": 0.084, "step": 9877 }, { "epoch": 6.319897632757518, "grad_norm": 1.014101264058774, "learning_rate": 6.301074588478649e-06, "loss": 0.0922, "step": 9878 }, { "epoch": 6.320537428023033, "grad_norm": 0.9632302317781448, "learning_rate": 6.29914948238858e-06, "loss": 0.0909, "step": 9879 }, { "epoch": 6.321177223288547, "grad_norm": 0.8406101518331953, "learning_rate": 6.2972245352064475e-06, "loss": 0.0851, "step": 9880 }, { "epoch": 6.3218170185540625, "grad_norm": 0.7466292947320996, "learning_rate": 6.295299747014905e-06, "loss": 0.0784, "step": 9881 }, { "epoch": 6.322456813819578, "grad_norm": 0.8825969354241954, "learning_rate": 6.293375117896604e-06, "loss": 0.0834, "step": 9882 }, { "epoch": 6.323096609085093, "grad_norm": 0.9615405491504229, "learning_rate": 6.291450647934181e-06, "loss": 0.0956, "step": 9883 }, { "epoch": 6.3237364043506075, "grad_norm": 0.8670340575459532, "learning_rate": 6.289526337210266e-06, "loss": 0.0826, "step": 9884 }, { "epoch": 6.324376199616123, "grad_norm": 0.7985103317874723, "learning_rate": 6.2876021858074955e-06, "loss": 0.084, "step": 9885 }, { "epoch": 6.325015994881638, "grad_norm": 0.8963136575025168, "learning_rate": 6.285678193808482e-06, "loss": 0.089, "step": 9886 }, { "epoch": 6.325655790147153, "grad_norm": 0.9226654864920942, "learning_rate": 6.28375436129584e-06, "loss": 0.0844, "step": 9887 }, { "epoch": 6.326295585412668, "grad_norm": 0.8501250224146972, "learning_rate": 6.281830688352172e-06, "loss": 0.0898, "step": 9888 }, { "epoch": 6.326935380678183, "grad_norm": 0.7926267216893741, "learning_rate": 6.279907175060083e-06, "loss": 0.0824, "step": 9889 }, { "epoch": 6.327575175943698, "grad_norm": 0.80434847941905, "learning_rate": 6.277983821502161e-06, "loss": 0.0849, "step": 9890 }, { "epoch": 6.3282149712092135, "grad_norm": 0.975226399559877, "learning_rate": 6.27606062776099e-06, "loss": 0.0879, "step": 9891 }, { "epoch": 6.328854766474728, "grad_norm": 0.8926386827429967, "learning_rate": 6.274137593919156e-06, "loss": 0.081, "step": 9892 }, { "epoch": 6.329494561740243, "grad_norm": 0.8922513182081244, "learning_rate": 6.2722147200592244e-06, "loss": 0.088, "step": 9893 }, { "epoch": 6.330134357005758, "grad_norm": 0.8864731136992111, "learning_rate": 6.270292006263758e-06, "loss": 0.0839, "step": 9894 }, { "epoch": 6.330774152271273, "grad_norm": 0.8530239671764798, "learning_rate": 6.2683694526153226e-06, "loss": 0.0888, "step": 9895 }, { "epoch": 6.331413947536788, "grad_norm": 0.8327915466282798, "learning_rate": 6.2664470591964655e-06, "loss": 0.0832, "step": 9896 }, { "epoch": 6.332053742802303, "grad_norm": 0.9964428625677308, "learning_rate": 6.264524826089729e-06, "loss": 0.0852, "step": 9897 }, { "epoch": 6.332693538067819, "grad_norm": 0.9630203676218839, "learning_rate": 6.262602753377649e-06, "loss": 0.0891, "step": 9898 }, { "epoch": 6.333333333333333, "grad_norm": 0.9287559537836523, "learning_rate": 6.260680841142762e-06, "loss": 0.078, "step": 9899 }, { "epoch": 6.333973128598848, "grad_norm": 0.8774766158186575, "learning_rate": 6.258759089467588e-06, "loss": 0.087, "step": 9900 }, { "epoch": 6.334612923864364, "grad_norm": 0.9055887518198332, "learning_rate": 6.256837498434638e-06, "loss": 0.0819, "step": 9901 }, { "epoch": 6.335252719129878, "grad_norm": 0.9173323429769829, "learning_rate": 6.2549160681264335e-06, "loss": 0.0818, "step": 9902 }, { "epoch": 6.335892514395393, "grad_norm": 0.8281232246844025, "learning_rate": 6.252994798625469e-06, "loss": 0.0802, "step": 9903 }, { "epoch": 6.3365323096609085, "grad_norm": 0.8300050753802651, "learning_rate": 6.251073690014242e-06, "loss": 0.0906, "step": 9904 }, { "epoch": 6.337172104926424, "grad_norm": 0.8357922110783597, "learning_rate": 6.249152742375242e-06, "loss": 0.0885, "step": 9905 }, { "epoch": 6.337811900191938, "grad_norm": 0.9967524127965585, "learning_rate": 6.247231955790951e-06, "loss": 0.0954, "step": 9906 }, { "epoch": 6.3384516954574535, "grad_norm": 0.9545306897907291, "learning_rate": 6.245311330343846e-06, "loss": 0.0831, "step": 9907 }, { "epoch": 6.339091490722969, "grad_norm": 0.8972159808591181, "learning_rate": 6.243390866116388e-06, "loss": 0.0783, "step": 9908 }, { "epoch": 6.339731285988484, "grad_norm": 0.9816413098506679, "learning_rate": 6.241470563191049e-06, "loss": 0.0896, "step": 9909 }, { "epoch": 6.340371081253998, "grad_norm": 0.8524835720259595, "learning_rate": 6.239550421650278e-06, "loss": 0.0812, "step": 9910 }, { "epoch": 6.341010876519514, "grad_norm": 0.8285814569699993, "learning_rate": 6.2376304415765186e-06, "loss": 0.0796, "step": 9911 }, { "epoch": 6.341650671785029, "grad_norm": 0.7955449021185061, "learning_rate": 6.23571062305222e-06, "loss": 0.0876, "step": 9912 }, { "epoch": 6.342290467050544, "grad_norm": 0.8711231577099907, "learning_rate": 6.23379096615981e-06, "loss": 0.0885, "step": 9913 }, { "epoch": 6.342930262316059, "grad_norm": 0.852486275773163, "learning_rate": 6.231871470981716e-06, "loss": 0.0846, "step": 9914 }, { "epoch": 6.343570057581574, "grad_norm": 0.9158967127813551, "learning_rate": 6.229952137600353e-06, "loss": 0.0822, "step": 9915 }, { "epoch": 6.344209852847089, "grad_norm": 0.9881238649876923, "learning_rate": 6.2280329660981456e-06, "loss": 0.0858, "step": 9916 }, { "epoch": 6.3448496481126035, "grad_norm": 0.9263044724505408, "learning_rate": 6.22611395655749e-06, "loss": 0.0847, "step": 9917 }, { "epoch": 6.345489443378119, "grad_norm": 0.875220501302863, "learning_rate": 6.224195109060788e-06, "loss": 0.0889, "step": 9918 }, { "epoch": 6.346129238643634, "grad_norm": 0.9167620851883559, "learning_rate": 6.222276423690432e-06, "loss": 0.083, "step": 9919 }, { "epoch": 6.346769033909149, "grad_norm": 0.9320452647944212, "learning_rate": 6.220357900528806e-06, "loss": 0.0955, "step": 9920 }, { "epoch": 6.347408829174664, "grad_norm": 0.7985517849130952, "learning_rate": 6.218439539658285e-06, "loss": 0.0754, "step": 9921 }, { "epoch": 6.348048624440179, "grad_norm": 0.7980677063236166, "learning_rate": 6.216521341161248e-06, "loss": 0.0773, "step": 9922 }, { "epoch": 6.348688419705694, "grad_norm": 0.8314195291275595, "learning_rate": 6.214603305120055e-06, "loss": 0.082, "step": 9923 }, { "epoch": 6.34932821497121, "grad_norm": 0.8537825971373829, "learning_rate": 6.21268543161706e-06, "loss": 0.0816, "step": 9924 }, { "epoch": 6.349968010236724, "grad_norm": 0.8561106820109053, "learning_rate": 6.210767720734613e-06, "loss": 0.0776, "step": 9925 }, { "epoch": 6.350607805502239, "grad_norm": 0.8326941919806236, "learning_rate": 6.208850172555063e-06, "loss": 0.0846, "step": 9926 }, { "epoch": 6.3512476007677545, "grad_norm": 0.8938486515027408, "learning_rate": 6.206932787160744e-06, "loss": 0.09, "step": 9927 }, { "epoch": 6.35188739603327, "grad_norm": 0.8444185981493995, "learning_rate": 6.20501556463398e-06, "loss": 0.0822, "step": 9928 }, { "epoch": 6.352527191298784, "grad_norm": 0.9166447062481284, "learning_rate": 6.203098505057099e-06, "loss": 0.0855, "step": 9929 }, { "epoch": 6.3531669865642995, "grad_norm": 0.7950369851006254, "learning_rate": 6.2011816085124145e-06, "loss": 0.0802, "step": 9930 }, { "epoch": 6.353806781829815, "grad_norm": 0.9568099320035552, "learning_rate": 6.199264875082229e-06, "loss": 0.0897, "step": 9931 }, { "epoch": 6.354446577095329, "grad_norm": 0.9519024488091856, "learning_rate": 6.197348304848855e-06, "loss": 0.0943, "step": 9932 }, { "epoch": 6.355086372360844, "grad_norm": 0.8180623886329137, "learning_rate": 6.1954318978945794e-06, "loss": 0.0763, "step": 9933 }, { "epoch": 6.35572616762636, "grad_norm": 0.8412218624349815, "learning_rate": 6.19351565430169e-06, "loss": 0.0825, "step": 9934 }, { "epoch": 6.356365962891875, "grad_norm": 1.0006620102257613, "learning_rate": 6.191599574152465e-06, "loss": 0.0889, "step": 9935 }, { "epoch": 6.357005758157389, "grad_norm": 0.8192435841390012, "learning_rate": 6.189683657529182e-06, "loss": 0.0812, "step": 9936 }, { "epoch": 6.357645553422905, "grad_norm": 0.9144875675868579, "learning_rate": 6.187767904514106e-06, "loss": 0.0889, "step": 9937 }, { "epoch": 6.35828534868842, "grad_norm": 0.9018655934995884, "learning_rate": 6.185852315189489e-06, "loss": 0.0929, "step": 9938 }, { "epoch": 6.358925143953935, "grad_norm": 0.9472782185252779, "learning_rate": 6.183936889637593e-06, "loss": 0.0874, "step": 9939 }, { "epoch": 6.3595649392194495, "grad_norm": 0.9934732360983423, "learning_rate": 6.18202162794066e-06, "loss": 0.0861, "step": 9940 }, { "epoch": 6.360204734484965, "grad_norm": 0.9966112357523148, "learning_rate": 6.180106530180923e-06, "loss": 0.0822, "step": 9941 }, { "epoch": 6.36084452975048, "grad_norm": 0.8766640779905034, "learning_rate": 6.178191596440619e-06, "loss": 0.0847, "step": 9942 }, { "epoch": 6.3614843250159945, "grad_norm": 0.8866253368952017, "learning_rate": 6.17627682680197e-06, "loss": 0.0872, "step": 9943 }, { "epoch": 6.36212412028151, "grad_norm": 0.7458421073206418, "learning_rate": 6.174362221347192e-06, "loss": 0.0815, "step": 9944 }, { "epoch": 6.362763915547025, "grad_norm": 0.9464153718686659, "learning_rate": 6.1724477801584905e-06, "loss": 0.0857, "step": 9945 }, { "epoch": 6.36340371081254, "grad_norm": 0.80786604577411, "learning_rate": 6.170533503318077e-06, "loss": 0.0846, "step": 9946 }, { "epoch": 6.364043506078055, "grad_norm": 0.811134412226495, "learning_rate": 6.168619390908141e-06, "loss": 0.0867, "step": 9947 }, { "epoch": 6.36468330134357, "grad_norm": 0.7277004647320798, "learning_rate": 6.166705443010873e-06, "loss": 0.0787, "step": 9948 }, { "epoch": 6.365323096609085, "grad_norm": 0.9285734002305459, "learning_rate": 6.164791659708456e-06, "loss": 0.0825, "step": 9949 }, { "epoch": 6.3659628918746005, "grad_norm": 0.8134738019635843, "learning_rate": 6.162878041083063e-06, "loss": 0.0816, "step": 9950 }, { "epoch": 6.366602687140115, "grad_norm": 0.9724898577905596, "learning_rate": 6.160964587216857e-06, "loss": 0.0901, "step": 9951 }, { "epoch": 6.36724248240563, "grad_norm": 0.925690771969187, "learning_rate": 6.159051298192006e-06, "loss": 0.0957, "step": 9952 }, { "epoch": 6.3678822776711455, "grad_norm": 0.8077712629329677, "learning_rate": 6.15713817409066e-06, "loss": 0.0773, "step": 9953 }, { "epoch": 6.36852207293666, "grad_norm": 0.8160386805739878, "learning_rate": 6.155225214994964e-06, "loss": 0.0798, "step": 9954 }, { "epoch": 6.369161868202175, "grad_norm": 0.8833396288248394, "learning_rate": 6.1533124209870566e-06, "loss": 0.0846, "step": 9955 }, { "epoch": 6.36980166346769, "grad_norm": 0.8862864203255529, "learning_rate": 6.1513997921490754e-06, "loss": 0.0842, "step": 9956 }, { "epoch": 6.370441458733206, "grad_norm": 0.957100398852324, "learning_rate": 6.149487328563139e-06, "loss": 0.0918, "step": 9957 }, { "epoch": 6.37108125399872, "grad_norm": 0.8053945618980128, "learning_rate": 6.147575030311362e-06, "loss": 0.0855, "step": 9958 }, { "epoch": 6.371721049264235, "grad_norm": 0.8449018785737483, "learning_rate": 6.1456628974758655e-06, "loss": 0.0833, "step": 9959 }, { "epoch": 6.372360844529751, "grad_norm": 0.9892814062007385, "learning_rate": 6.143750930138747e-06, "loss": 0.0827, "step": 9960 }, { "epoch": 6.373000639795266, "grad_norm": 0.7881482284145973, "learning_rate": 6.141839128382102e-06, "loss": 0.0745, "step": 9961 }, { "epoch": 6.37364043506078, "grad_norm": 0.895159422124989, "learning_rate": 6.139927492288025e-06, "loss": 0.0869, "step": 9962 }, { "epoch": 6.3742802303262955, "grad_norm": 0.7917205916009562, "learning_rate": 6.138016021938593e-06, "loss": 0.0813, "step": 9963 }, { "epoch": 6.374920025591811, "grad_norm": 0.9871895368872158, "learning_rate": 6.136104717415885e-06, "loss": 0.088, "step": 9964 }, { "epoch": 6.375559820857326, "grad_norm": 0.9559540608528179, "learning_rate": 6.134193578801965e-06, "loss": 0.0875, "step": 9965 }, { "epoch": 6.3761996161228405, "grad_norm": 0.7975678707803999, "learning_rate": 6.132282606178896e-06, "loss": 0.0835, "step": 9966 }, { "epoch": 6.376839411388356, "grad_norm": 0.8779385162829818, "learning_rate": 6.130371799628734e-06, "loss": 0.0841, "step": 9967 }, { "epoch": 6.377479206653871, "grad_norm": 1.0290006554292488, "learning_rate": 6.128461159233519e-06, "loss": 0.0891, "step": 9968 }, { "epoch": 6.378119001919385, "grad_norm": 0.8379872167244744, "learning_rate": 6.126550685075298e-06, "loss": 0.0874, "step": 9969 }, { "epoch": 6.378758797184901, "grad_norm": 0.8440060884309935, "learning_rate": 6.124640377236102e-06, "loss": 0.0847, "step": 9970 }, { "epoch": 6.379398592450416, "grad_norm": 0.9504875021417133, "learning_rate": 6.12273023579795e-06, "loss": 0.0872, "step": 9971 }, { "epoch": 6.380038387715931, "grad_norm": 0.943210011744938, "learning_rate": 6.120820260842868e-06, "loss": 0.0915, "step": 9972 }, { "epoch": 6.380678182981446, "grad_norm": 1.121298855529001, "learning_rate": 6.118910452452863e-06, "loss": 0.0901, "step": 9973 }, { "epoch": 6.381317978246961, "grad_norm": 0.8152720377555327, "learning_rate": 6.11700081070994e-06, "loss": 0.083, "step": 9974 }, { "epoch": 6.381957773512476, "grad_norm": 0.9044510735356233, "learning_rate": 6.1150913356960905e-06, "loss": 0.0796, "step": 9975 }, { "epoch": 6.3825975687779914, "grad_norm": 1.0229503679018774, "learning_rate": 6.1131820274933115e-06, "loss": 0.0994, "step": 9976 }, { "epoch": 6.383237364043506, "grad_norm": 0.9435972422859421, "learning_rate": 6.111272886183584e-06, "loss": 0.0912, "step": 9977 }, { "epoch": 6.383877159309021, "grad_norm": 0.8403716572718822, "learning_rate": 6.109363911848878e-06, "loss": 0.0862, "step": 9978 }, { "epoch": 6.384516954574536, "grad_norm": 0.8346822850045473, "learning_rate": 6.107455104571166e-06, "loss": 0.0891, "step": 9979 }, { "epoch": 6.385156749840051, "grad_norm": 0.8684683349362836, "learning_rate": 6.105546464432408e-06, "loss": 0.0891, "step": 9980 }, { "epoch": 6.385796545105566, "grad_norm": 0.8363381963679597, "learning_rate": 6.103637991514557e-06, "loss": 0.0865, "step": 9981 }, { "epoch": 6.386436340371081, "grad_norm": 0.8474542497263086, "learning_rate": 6.101729685899557e-06, "loss": 0.0835, "step": 9982 }, { "epoch": 6.387076135636597, "grad_norm": 0.866168905566539, "learning_rate": 6.0998215476693525e-06, "loss": 0.0795, "step": 9983 }, { "epoch": 6.387715930902111, "grad_norm": 0.8083017078247423, "learning_rate": 6.097913576905873e-06, "loss": 0.0849, "step": 9984 }, { "epoch": 6.388355726167626, "grad_norm": 0.7554014622661541, "learning_rate": 6.096005773691042e-06, "loss": 0.0726, "step": 9985 }, { "epoch": 6.3889955214331415, "grad_norm": 0.8201816568060003, "learning_rate": 6.094098138106779e-06, "loss": 0.0804, "step": 9986 }, { "epoch": 6.389635316698657, "grad_norm": 0.8334332081789917, "learning_rate": 6.092190670234994e-06, "loss": 0.0846, "step": 9987 }, { "epoch": 6.390275111964171, "grad_norm": 0.8125239639141886, "learning_rate": 6.090283370157588e-06, "loss": 0.0878, "step": 9988 }, { "epoch": 6.3909149072296865, "grad_norm": 0.9321370374385635, "learning_rate": 6.088376237956461e-06, "loss": 0.0853, "step": 9989 }, { "epoch": 6.391554702495202, "grad_norm": 0.892615929766644, "learning_rate": 6.086469273713502e-06, "loss": 0.0847, "step": 9990 }, { "epoch": 6.392194497760716, "grad_norm": 1.1227544890430816, "learning_rate": 6.084562477510589e-06, "loss": 0.0816, "step": 9991 }, { "epoch": 6.392834293026231, "grad_norm": 0.9578370179624465, "learning_rate": 6.082655849429596e-06, "loss": 0.0883, "step": 9992 }, { "epoch": 6.393474088291747, "grad_norm": 0.900803205317271, "learning_rate": 6.080749389552396e-06, "loss": 0.0875, "step": 9993 }, { "epoch": 6.394113883557262, "grad_norm": 0.9772934004143293, "learning_rate": 6.078843097960845e-06, "loss": 0.0923, "step": 9994 }, { "epoch": 6.394753678822776, "grad_norm": 0.8101246694079222, "learning_rate": 6.076936974736793e-06, "loss": 0.0839, "step": 9995 }, { "epoch": 6.395393474088292, "grad_norm": 0.8367969735812415, "learning_rate": 6.0750310199620925e-06, "loss": 0.0845, "step": 9996 }, { "epoch": 6.396033269353807, "grad_norm": 0.7853448460991241, "learning_rate": 6.073125233718578e-06, "loss": 0.0838, "step": 9997 }, { "epoch": 6.396673064619322, "grad_norm": 0.9145569280090607, "learning_rate": 6.071219616088076e-06, "loss": 0.0956, "step": 9998 }, { "epoch": 6.3973128598848366, "grad_norm": 0.8562535148117433, "learning_rate": 6.069314167152418e-06, "loss": 0.0885, "step": 9999 }, { "epoch": 6.397952655150352, "grad_norm": 0.9025304126682671, "learning_rate": 6.067408886993418e-06, "loss": 0.0873, "step": 10000 }, { "epoch": 6.398592450415867, "grad_norm": 0.858696777786411, "learning_rate": 6.065503775692886e-06, "loss": 0.0905, "step": 10001 }, { "epoch": 6.399232245681382, "grad_norm": 0.785908720765894, "learning_rate": 6.06359883333262e-06, "loss": 0.0796, "step": 10002 }, { "epoch": 6.399872040946897, "grad_norm": 0.8196368511367504, "learning_rate": 6.0616940599944206e-06, "loss": 0.087, "step": 10003 }, { "epoch": 6.400511836212412, "grad_norm": 0.8080473392048337, "learning_rate": 6.059789455760071e-06, "loss": 0.0801, "step": 10004 }, { "epoch": 6.401151631477927, "grad_norm": 0.8002401893210019, "learning_rate": 6.0578850207113505e-06, "loss": 0.0827, "step": 10005 }, { "epoch": 6.401791426743442, "grad_norm": 0.8670303194061924, "learning_rate": 6.0559807549300375e-06, "loss": 0.0865, "step": 10006 }, { "epoch": 6.402431222008957, "grad_norm": 0.906322539312927, "learning_rate": 6.054076658497896e-06, "loss": 0.0904, "step": 10007 }, { "epoch": 6.403071017274472, "grad_norm": 0.9014900234619465, "learning_rate": 6.052172731496681e-06, "loss": 0.0796, "step": 10008 }, { "epoch": 6.4037108125399875, "grad_norm": 0.8601302586970077, "learning_rate": 6.050268974008149e-06, "loss": 0.0902, "step": 10009 }, { "epoch": 6.404350607805502, "grad_norm": 0.9782431632969762, "learning_rate": 6.048365386114041e-06, "loss": 0.0842, "step": 10010 }, { "epoch": 6.404990403071017, "grad_norm": 0.8738772781071045, "learning_rate": 6.046461967896093e-06, "loss": 0.0865, "step": 10011 }, { "epoch": 6.4056301983365325, "grad_norm": 0.9160505038813673, "learning_rate": 6.0445587194360335e-06, "loss": 0.0814, "step": 10012 }, { "epoch": 6.406269993602048, "grad_norm": 0.8614148015826602, "learning_rate": 6.042655640815589e-06, "loss": 0.0837, "step": 10013 }, { "epoch": 6.406909788867562, "grad_norm": 0.9736741847602486, "learning_rate": 6.04075273211647e-06, "loss": 0.0859, "step": 10014 }, { "epoch": 6.407549584133077, "grad_norm": 0.7902111051416869, "learning_rate": 6.038849993420386e-06, "loss": 0.0806, "step": 10015 }, { "epoch": 6.408189379398593, "grad_norm": 0.88378861154471, "learning_rate": 6.036947424809039e-06, "loss": 0.0913, "step": 10016 }, { "epoch": 6.408829174664108, "grad_norm": 0.7469994629769842, "learning_rate": 6.0350450263641194e-06, "loss": 0.0867, "step": 10017 }, { "epoch": 6.409468969929622, "grad_norm": 0.9712721967672391, "learning_rate": 6.033142798167309e-06, "loss": 0.0795, "step": 10018 }, { "epoch": 6.410108765195138, "grad_norm": 0.9008401109196604, "learning_rate": 6.0312407403002934e-06, "loss": 0.0867, "step": 10019 }, { "epoch": 6.410748560460653, "grad_norm": 0.8835361625949473, "learning_rate": 6.029338852844742e-06, "loss": 0.0794, "step": 10020 }, { "epoch": 6.411388355726167, "grad_norm": 0.7650651627256186, "learning_rate": 6.027437135882317e-06, "loss": 0.0821, "step": 10021 }, { "epoch": 6.4120281509916826, "grad_norm": 0.926100044277075, "learning_rate": 6.0255355894946685e-06, "loss": 0.0944, "step": 10022 }, { "epoch": 6.412667946257198, "grad_norm": 0.977389748838732, "learning_rate": 6.0236342137634585e-06, "loss": 0.0887, "step": 10023 }, { "epoch": 6.413307741522713, "grad_norm": 0.9175095412967547, "learning_rate": 6.021733008770321e-06, "loss": 0.0825, "step": 10024 }, { "epoch": 6.4139475367882275, "grad_norm": 0.8207734262235179, "learning_rate": 6.019831974596889e-06, "loss": 0.0854, "step": 10025 }, { "epoch": 6.414587332053743, "grad_norm": 0.7751472662281773, "learning_rate": 6.017931111324795e-06, "loss": 0.0839, "step": 10026 }, { "epoch": 6.415227127319258, "grad_norm": 0.9687629688477825, "learning_rate": 6.016030419035654e-06, "loss": 0.0784, "step": 10027 }, { "epoch": 6.415866922584772, "grad_norm": 0.8873821996370429, "learning_rate": 6.0141298978110765e-06, "loss": 0.08, "step": 10028 }, { "epoch": 6.416506717850288, "grad_norm": 0.9311472045317039, "learning_rate": 6.012229547732675e-06, "loss": 0.087, "step": 10029 }, { "epoch": 6.417146513115803, "grad_norm": 0.981302669989736, "learning_rate": 6.010329368882043e-06, "loss": 0.0829, "step": 10030 }, { "epoch": 6.417786308381318, "grad_norm": 0.8910693694280194, "learning_rate": 6.008429361340771e-06, "loss": 0.0846, "step": 10031 }, { "epoch": 6.418426103646833, "grad_norm": 0.9586879100145376, "learning_rate": 6.006529525190439e-06, "loss": 0.0941, "step": 10032 }, { "epoch": 6.419065898912348, "grad_norm": 0.8128797125998692, "learning_rate": 6.004629860512627e-06, "loss": 0.0805, "step": 10033 }, { "epoch": 6.419705694177863, "grad_norm": 0.854140186652562, "learning_rate": 6.002730367388904e-06, "loss": 0.0874, "step": 10034 }, { "epoch": 6.4203454894433785, "grad_norm": 0.9485512559655634, "learning_rate": 6.000831045900822e-06, "loss": 0.0928, "step": 10035 }, { "epoch": 6.420985284708893, "grad_norm": 0.8948356356881365, "learning_rate": 5.998931896129947e-06, "loss": 0.09, "step": 10036 }, { "epoch": 6.421625079974408, "grad_norm": 0.8878410990401394, "learning_rate": 5.997032918157817e-06, "loss": 0.0864, "step": 10037 }, { "epoch": 6.422264875239923, "grad_norm": 0.8733795368513908, "learning_rate": 5.9951341120659725e-06, "loss": 0.0879, "step": 10038 }, { "epoch": 6.422904670505439, "grad_norm": 0.8575982998288492, "learning_rate": 5.9932354779359455e-06, "loss": 0.0829, "step": 10039 }, { "epoch": 6.423544465770953, "grad_norm": 0.8750263444719455, "learning_rate": 5.9913370158492605e-06, "loss": 0.0914, "step": 10040 }, { "epoch": 6.424184261036468, "grad_norm": 0.8473275193608224, "learning_rate": 5.989438725887433e-06, "loss": 0.0878, "step": 10041 }, { "epoch": 6.424824056301984, "grad_norm": 0.8847307030117034, "learning_rate": 5.9875406081319675e-06, "loss": 0.0943, "step": 10042 }, { "epoch": 6.425463851567498, "grad_norm": 0.9449803971536179, "learning_rate": 5.985642662664377e-06, "loss": 0.089, "step": 10043 }, { "epoch": 6.426103646833013, "grad_norm": 0.8062867175982737, "learning_rate": 5.983744889566148e-06, "loss": 0.0816, "step": 10044 }, { "epoch": 6.4267434420985285, "grad_norm": 0.9095446871160878, "learning_rate": 5.981847288918767e-06, "loss": 0.0909, "step": 10045 }, { "epoch": 6.427383237364044, "grad_norm": 0.8225895189386824, "learning_rate": 5.9799498608037175e-06, "loss": 0.0851, "step": 10046 }, { "epoch": 6.428023032629558, "grad_norm": 0.9920896434290359, "learning_rate": 5.97805260530247e-06, "loss": 0.0888, "step": 10047 }, { "epoch": 6.4286628278950735, "grad_norm": 0.9391018531272527, "learning_rate": 5.976155522496485e-06, "loss": 0.0825, "step": 10048 }, { "epoch": 6.429302623160589, "grad_norm": 0.9190915934366939, "learning_rate": 5.97425861246723e-06, "loss": 0.0849, "step": 10049 }, { "epoch": 6.429942418426104, "grad_norm": 0.8677105153391945, "learning_rate": 5.972361875296147e-06, "loss": 0.0812, "step": 10050 }, { "epoch": 6.430582213691618, "grad_norm": 1.0674870055409493, "learning_rate": 5.97046531106468e-06, "loss": 0.1068, "step": 10051 }, { "epoch": 6.431222008957134, "grad_norm": 0.8227010495022126, "learning_rate": 5.968568919854262e-06, "loss": 0.0802, "step": 10052 }, { "epoch": 6.431861804222649, "grad_norm": 0.7185300371895251, "learning_rate": 5.966672701746326e-06, "loss": 0.0797, "step": 10053 }, { "epoch": 6.432501599488164, "grad_norm": 0.8602775687221389, "learning_rate": 5.964776656822291e-06, "loss": 0.0838, "step": 10054 }, { "epoch": 6.433141394753679, "grad_norm": 0.7920222626536044, "learning_rate": 5.962880785163565e-06, "loss": 0.0839, "step": 10055 }, { "epoch": 6.433781190019194, "grad_norm": 0.8127992360755665, "learning_rate": 5.960985086851558e-06, "loss": 0.0852, "step": 10056 }, { "epoch": 6.434420985284709, "grad_norm": 0.9073242750601443, "learning_rate": 5.959089561967668e-06, "loss": 0.0898, "step": 10057 }, { "epoch": 6.435060780550224, "grad_norm": 0.7549362182133116, "learning_rate": 5.957194210593283e-06, "loss": 0.0819, "step": 10058 }, { "epoch": 6.435700575815739, "grad_norm": 0.8403580634198293, "learning_rate": 5.955299032809784e-06, "loss": 0.0836, "step": 10059 }, { "epoch": 6.436340371081254, "grad_norm": 0.9316965550030217, "learning_rate": 5.9534040286985526e-06, "loss": 0.0916, "step": 10060 }, { "epoch": 6.436980166346769, "grad_norm": 0.9433857070824815, "learning_rate": 5.951509198340955e-06, "loss": 0.0877, "step": 10061 }, { "epoch": 6.437619961612284, "grad_norm": 0.8783141431909679, "learning_rate": 5.949614541818347e-06, "loss": 0.0865, "step": 10062 }, { "epoch": 6.438259756877799, "grad_norm": 0.8056963056443185, "learning_rate": 5.9477200592120884e-06, "loss": 0.088, "step": 10063 }, { "epoch": 6.438899552143314, "grad_norm": 0.9246704333889894, "learning_rate": 5.945825750603522e-06, "loss": 0.0849, "step": 10064 }, { "epoch": 6.439539347408829, "grad_norm": 0.8589319033561932, "learning_rate": 5.943931616073983e-06, "loss": 0.0777, "step": 10065 }, { "epoch": 6.440179142674344, "grad_norm": 0.8728450717222219, "learning_rate": 5.942037655704808e-06, "loss": 0.0899, "step": 10066 }, { "epoch": 6.440818937939859, "grad_norm": 0.9448917565614015, "learning_rate": 5.940143869577317e-06, "loss": 0.0873, "step": 10067 }, { "epoch": 6.4414587332053745, "grad_norm": 0.919504733239026, "learning_rate": 5.9382502577728285e-06, "loss": 0.0914, "step": 10068 }, { "epoch": 6.442098528470889, "grad_norm": 0.8647361032683694, "learning_rate": 5.936356820372644e-06, "loss": 0.0841, "step": 10069 }, { "epoch": 6.442738323736404, "grad_norm": 0.9360837926772057, "learning_rate": 5.934463557458072e-06, "loss": 0.0903, "step": 10070 }, { "epoch": 6.4433781190019195, "grad_norm": 0.8147471445325394, "learning_rate": 5.932570469110403e-06, "loss": 0.0757, "step": 10071 }, { "epoch": 6.444017914267435, "grad_norm": 0.8800817326657612, "learning_rate": 5.93067755541092e-06, "loss": 0.0826, "step": 10072 }, { "epoch": 6.444657709532949, "grad_norm": 0.8042036436916437, "learning_rate": 5.928784816440907e-06, "loss": 0.0829, "step": 10073 }, { "epoch": 6.445297504798464, "grad_norm": 0.9431992238415866, "learning_rate": 5.926892252281631e-06, "loss": 0.0919, "step": 10074 }, { "epoch": 6.44593730006398, "grad_norm": 0.8362967082526761, "learning_rate": 5.9249998630143556e-06, "loss": 0.0808, "step": 10075 }, { "epoch": 6.446577095329495, "grad_norm": 0.9390467633166071, "learning_rate": 5.923107648720338e-06, "loss": 0.0834, "step": 10076 }, { "epoch": 6.447216890595009, "grad_norm": 0.8591359672128218, "learning_rate": 5.921215609480826e-06, "loss": 0.084, "step": 10077 }, { "epoch": 6.447856685860525, "grad_norm": 0.9197852442749627, "learning_rate": 5.91932374537706e-06, "loss": 0.0819, "step": 10078 }, { "epoch": 6.44849648112604, "grad_norm": 0.8738265382109782, "learning_rate": 5.91743205649027e-06, "loss": 0.0841, "step": 10079 }, { "epoch": 6.449136276391554, "grad_norm": 0.722589404433596, "learning_rate": 5.915540542901689e-06, "loss": 0.0775, "step": 10080 }, { "epoch": 6.44977607165707, "grad_norm": 0.8761902162149309, "learning_rate": 5.9136492046925305e-06, "loss": 0.0808, "step": 10081 }, { "epoch": 6.450415866922585, "grad_norm": 0.803302861293541, "learning_rate": 5.9117580419440034e-06, "loss": 0.0852, "step": 10082 }, { "epoch": 6.4510556621881, "grad_norm": 0.8116274598045804, "learning_rate": 5.909867054737317e-06, "loss": 0.0851, "step": 10083 }, { "epoch": 6.4516954574536145, "grad_norm": 0.8793661190556042, "learning_rate": 5.907976243153664e-06, "loss": 0.0837, "step": 10084 }, { "epoch": 6.45233525271913, "grad_norm": 0.9762222827698409, "learning_rate": 5.906085607274229e-06, "loss": 0.0889, "step": 10085 }, { "epoch": 6.452975047984645, "grad_norm": 0.9074148752904749, "learning_rate": 5.904195147180198e-06, "loss": 0.0904, "step": 10086 }, { "epoch": 6.45361484325016, "grad_norm": 0.8771598105397354, "learning_rate": 5.902304862952741e-06, "loss": 0.0896, "step": 10087 }, { "epoch": 6.454254638515675, "grad_norm": 0.821530659152466, "learning_rate": 5.900414754673025e-06, "loss": 0.0853, "step": 10088 }, { "epoch": 6.45489443378119, "grad_norm": 0.8964191165878269, "learning_rate": 5.898524822422203e-06, "loss": 0.0888, "step": 10089 }, { "epoch": 6.455534229046705, "grad_norm": 0.7439608009506494, "learning_rate": 5.896635066281432e-06, "loss": 0.0836, "step": 10090 }, { "epoch": 6.4561740243122205, "grad_norm": 0.9329186431107171, "learning_rate": 5.894745486331852e-06, "loss": 0.088, "step": 10091 }, { "epoch": 6.456813819577735, "grad_norm": 0.814453498538395, "learning_rate": 5.892856082654597e-06, "loss": 0.0848, "step": 10092 }, { "epoch": 6.45745361484325, "grad_norm": 0.8682202694917422, "learning_rate": 5.8909668553307974e-06, "loss": 0.0869, "step": 10093 }, { "epoch": 6.4580934101087655, "grad_norm": 0.8899241978273444, "learning_rate": 5.889077804441572e-06, "loss": 0.088, "step": 10094 }, { "epoch": 6.45873320537428, "grad_norm": 0.8818500594305403, "learning_rate": 5.887188930068027e-06, "loss": 0.0933, "step": 10095 }, { "epoch": 6.459373000639795, "grad_norm": 0.8609940553849548, "learning_rate": 5.88530023229128e-06, "loss": 0.0832, "step": 10096 }, { "epoch": 6.46001279590531, "grad_norm": 0.9184659138555907, "learning_rate": 5.88341171119242e-06, "loss": 0.0853, "step": 10097 }, { "epoch": 6.460652591170826, "grad_norm": 0.8467342349810107, "learning_rate": 5.881523366852538e-06, "loss": 0.0835, "step": 10098 }, { "epoch": 6.46129238643634, "grad_norm": 0.8995687949671883, "learning_rate": 5.879635199352715e-06, "loss": 0.0881, "step": 10099 }, { "epoch": 6.461932181701855, "grad_norm": 0.8395876888666651, "learning_rate": 5.8777472087740295e-06, "loss": 0.0831, "step": 10100 }, { "epoch": 6.462571976967371, "grad_norm": 0.9490663041979615, "learning_rate": 5.875859395197545e-06, "loss": 0.0854, "step": 10101 }, { "epoch": 6.463211772232885, "grad_norm": 0.897670220289674, "learning_rate": 5.8739717587043175e-06, "loss": 0.0824, "step": 10102 }, { "epoch": 6.4638515674984, "grad_norm": 0.8683288927063614, "learning_rate": 5.8720842993754075e-06, "loss": 0.0856, "step": 10103 }, { "epoch": 6.464491362763916, "grad_norm": 0.926078811860979, "learning_rate": 5.870197017291855e-06, "loss": 0.0875, "step": 10104 }, { "epoch": 6.465131158029431, "grad_norm": 0.8237175761040658, "learning_rate": 5.868309912534694e-06, "loss": 0.0789, "step": 10105 }, { "epoch": 6.465770953294945, "grad_norm": 0.9131705249763878, "learning_rate": 5.866422985184958e-06, "loss": 0.0856, "step": 10106 }, { "epoch": 6.4664107485604605, "grad_norm": 0.8988129315552987, "learning_rate": 5.8645362353236654e-06, "loss": 0.085, "step": 10107 }, { "epoch": 6.467050543825976, "grad_norm": 0.8640842706997421, "learning_rate": 5.862649663031832e-06, "loss": 0.0855, "step": 10108 }, { "epoch": 6.467690339091491, "grad_norm": 0.8452484928284825, "learning_rate": 5.860763268390457e-06, "loss": 0.0818, "step": 10109 }, { "epoch": 6.468330134357005, "grad_norm": 0.8808338941420576, "learning_rate": 5.858877051480547e-06, "loss": 0.0835, "step": 10110 }, { "epoch": 6.468969929622521, "grad_norm": 0.8690762937808596, "learning_rate": 5.856991012383091e-06, "loss": 0.0901, "step": 10111 }, { "epoch": 6.469609724888036, "grad_norm": 0.9318076846543466, "learning_rate": 5.855105151179068e-06, "loss": 0.0885, "step": 10112 }, { "epoch": 6.470249520153551, "grad_norm": 0.9741543743574161, "learning_rate": 5.853219467949459e-06, "loss": 0.0905, "step": 10113 }, { "epoch": 6.470889315419066, "grad_norm": 0.9196216680753307, "learning_rate": 5.851333962775231e-06, "loss": 0.0847, "step": 10114 }, { "epoch": 6.471529110684581, "grad_norm": 0.898442393223155, "learning_rate": 5.84944863573734e-06, "loss": 0.0813, "step": 10115 }, { "epoch": 6.472168905950096, "grad_norm": 0.9272448373529387, "learning_rate": 5.847563486916743e-06, "loss": 0.0904, "step": 10116 }, { "epoch": 6.472808701215611, "grad_norm": 0.9190016012126974, "learning_rate": 5.845678516394384e-06, "loss": 0.0878, "step": 10117 }, { "epoch": 6.473448496481126, "grad_norm": 0.8689766653922557, "learning_rate": 5.843793724251199e-06, "loss": 0.0795, "step": 10118 }, { "epoch": 6.474088291746641, "grad_norm": 0.8242333278763002, "learning_rate": 5.841909110568114e-06, "loss": 0.0857, "step": 10119 }, { "epoch": 6.474728087012156, "grad_norm": 0.7527382667347018, "learning_rate": 5.840024675426059e-06, "loss": 0.0795, "step": 10120 }, { "epoch": 6.475367882277671, "grad_norm": 0.8444826952991416, "learning_rate": 5.8381404189059466e-06, "loss": 0.0796, "step": 10121 }, { "epoch": 6.476007677543186, "grad_norm": 0.8287503153463399, "learning_rate": 5.836256341088675e-06, "loss": 0.0838, "step": 10122 }, { "epoch": 6.476647472808701, "grad_norm": 0.8529513705865536, "learning_rate": 5.834372442055155e-06, "loss": 0.0885, "step": 10123 }, { "epoch": 6.477287268074217, "grad_norm": 0.883356181564266, "learning_rate": 5.8324887218862714e-06, "loss": 0.09, "step": 10124 }, { "epoch": 6.477927063339731, "grad_norm": 0.8276325932777512, "learning_rate": 5.8306051806629035e-06, "loss": 0.0792, "step": 10125 }, { "epoch": 6.478566858605246, "grad_norm": 0.9059039259316829, "learning_rate": 5.8287218184659375e-06, "loss": 0.0811, "step": 10126 }, { "epoch": 6.479206653870762, "grad_norm": 0.8455356965872428, "learning_rate": 5.826838635376236e-06, "loss": 0.0852, "step": 10127 }, { "epoch": 6.479846449136277, "grad_norm": 0.9145004647472038, "learning_rate": 5.82495563147466e-06, "loss": 0.0966, "step": 10128 }, { "epoch": 6.480486244401791, "grad_norm": 0.8379189507727816, "learning_rate": 5.823072806842059e-06, "loss": 0.0803, "step": 10129 }, { "epoch": 6.4811260396673065, "grad_norm": 0.9482718842298163, "learning_rate": 5.8211901615592845e-06, "loss": 0.0822, "step": 10130 }, { "epoch": 6.481765834932822, "grad_norm": 0.9273050124028193, "learning_rate": 5.8193076957071725e-06, "loss": 0.0759, "step": 10131 }, { "epoch": 6.482405630198336, "grad_norm": 0.917262009014404, "learning_rate": 5.817425409366551e-06, "loss": 0.0816, "step": 10132 }, { "epoch": 6.483045425463851, "grad_norm": 0.8602113422194588, "learning_rate": 5.81554330261824e-06, "loss": 0.0792, "step": 10133 }, { "epoch": 6.483685220729367, "grad_norm": 0.8766737835271464, "learning_rate": 5.813661375543058e-06, "loss": 0.0903, "step": 10134 }, { "epoch": 6.484325015994882, "grad_norm": 0.9240877471985307, "learning_rate": 5.8117796282218085e-06, "loss": 0.0895, "step": 10135 }, { "epoch": 6.484964811260396, "grad_norm": 1.0063619622438829, "learning_rate": 5.809898060735288e-06, "loss": 0.0889, "step": 10136 }, { "epoch": 6.485604606525912, "grad_norm": 0.8205665497443939, "learning_rate": 5.808016673164296e-06, "loss": 0.0801, "step": 10137 }, { "epoch": 6.486244401791427, "grad_norm": 0.831081307677967, "learning_rate": 5.80613546558961e-06, "loss": 0.0891, "step": 10138 }, { "epoch": 6.486884197056942, "grad_norm": 0.8888601584496681, "learning_rate": 5.804254438092003e-06, "loss": 0.0834, "step": 10139 }, { "epoch": 6.487523992322457, "grad_norm": 0.8180827158208948, "learning_rate": 5.802373590752251e-06, "loss": 0.0784, "step": 10140 }, { "epoch": 6.488163787587972, "grad_norm": 0.8692679883913085, "learning_rate": 5.80049292365111e-06, "loss": 0.0854, "step": 10141 }, { "epoch": 6.488803582853487, "grad_norm": 0.840001929477132, "learning_rate": 5.798612436869329e-06, "loss": 0.0873, "step": 10142 }, { "epoch": 6.4894433781190015, "grad_norm": 0.8693679030531255, "learning_rate": 5.796732130487659e-06, "loss": 0.0802, "step": 10143 }, { "epoch": 6.490083173384517, "grad_norm": 0.9574270263591982, "learning_rate": 5.794852004586835e-06, "loss": 0.0894, "step": 10144 }, { "epoch": 6.490722968650032, "grad_norm": 1.1005845377023928, "learning_rate": 5.792972059247586e-06, "loss": 0.0877, "step": 10145 }, { "epoch": 6.491362763915547, "grad_norm": 0.9128604887247337, "learning_rate": 5.791092294550628e-06, "loss": 0.0887, "step": 10146 }, { "epoch": 6.492002559181062, "grad_norm": 0.8894814754432837, "learning_rate": 5.789212710576684e-06, "loss": 0.0912, "step": 10147 }, { "epoch": 6.492642354446577, "grad_norm": 0.7518931952658111, "learning_rate": 5.787333307406457e-06, "loss": 0.0782, "step": 10148 }, { "epoch": 6.493282149712092, "grad_norm": 0.8968632428912161, "learning_rate": 5.785454085120643e-06, "loss": 0.081, "step": 10149 }, { "epoch": 6.4939219449776076, "grad_norm": 0.8827087813068338, "learning_rate": 5.783575043799934e-06, "loss": 0.09, "step": 10150 }, { "epoch": 6.494561740243122, "grad_norm": 0.9785931865568414, "learning_rate": 5.781696183525011e-06, "loss": 0.0891, "step": 10151 }, { "epoch": 6.495201535508637, "grad_norm": 0.8620589318199064, "learning_rate": 5.779817504376547e-06, "loss": 0.0832, "step": 10152 }, { "epoch": 6.4958413307741525, "grad_norm": 0.9773888138029277, "learning_rate": 5.777939006435216e-06, "loss": 0.0954, "step": 10153 }, { "epoch": 6.496481126039667, "grad_norm": 0.8202648698376749, "learning_rate": 5.776060689781674e-06, "loss": 0.0765, "step": 10154 }, { "epoch": 6.497120921305182, "grad_norm": 0.9010972281223005, "learning_rate": 5.774182554496573e-06, "loss": 0.0863, "step": 10155 }, { "epoch": 6.497760716570697, "grad_norm": 0.9227904263742333, "learning_rate": 5.77230460066055e-06, "loss": 0.0862, "step": 10156 }, { "epoch": 6.498400511836213, "grad_norm": 0.8965139253229109, "learning_rate": 5.770426828354252e-06, "loss": 0.0777, "step": 10157 }, { "epoch": 6.499040307101727, "grad_norm": 0.8501935771269282, "learning_rate": 5.768549237658301e-06, "loss": 0.0905, "step": 10158 }, { "epoch": 6.499680102367242, "grad_norm": 0.906623879127372, "learning_rate": 5.766671828653317e-06, "loss": 0.0799, "step": 10159 }, { "epoch": 6.500319897632758, "grad_norm": 0.8603270224566248, "learning_rate": 5.764794601419916e-06, "loss": 0.0845, "step": 10160 }, { "epoch": 6.500959692898273, "grad_norm": 0.7589278871271355, "learning_rate": 5.762917556038702e-06, "loss": 0.0825, "step": 10161 }, { "epoch": 6.501599488163787, "grad_norm": 0.9029246804668164, "learning_rate": 5.76104069259027e-06, "loss": 0.0907, "step": 10162 }, { "epoch": 6.502239283429303, "grad_norm": 0.8336979721287738, "learning_rate": 5.759164011155212e-06, "loss": 0.0889, "step": 10163 }, { "epoch": 6.502879078694818, "grad_norm": 0.7866234727740685, "learning_rate": 5.7572875118141054e-06, "loss": 0.0816, "step": 10164 }, { "epoch": 6.503518873960333, "grad_norm": 0.9415711270406674, "learning_rate": 5.755411194647527e-06, "loss": 0.0813, "step": 10165 }, { "epoch": 6.5041586692258475, "grad_norm": 0.8805344856440308, "learning_rate": 5.753535059736037e-06, "loss": 0.086, "step": 10166 }, { "epoch": 6.504798464491363, "grad_norm": 0.7965383737227797, "learning_rate": 5.751659107160201e-06, "loss": 0.0872, "step": 10167 }, { "epoch": 6.505438259756878, "grad_norm": 0.9170455082693161, "learning_rate": 5.749783337000567e-06, "loss": 0.0833, "step": 10168 }, { "epoch": 6.5060780550223924, "grad_norm": 0.8094954305373547, "learning_rate": 5.747907749337672e-06, "loss": 0.078, "step": 10169 }, { "epoch": 6.506717850287908, "grad_norm": 0.9089090442284208, "learning_rate": 5.746032344252057e-06, "loss": 0.0873, "step": 10170 }, { "epoch": 6.507357645553423, "grad_norm": 0.9067139040315632, "learning_rate": 5.7441571218242475e-06, "loss": 0.0886, "step": 10171 }, { "epoch": 6.507997440818938, "grad_norm": 0.8977105504488686, "learning_rate": 5.742282082134755e-06, "loss": 0.0868, "step": 10172 }, { "epoch": 6.508637236084453, "grad_norm": 0.8260245631149726, "learning_rate": 5.740407225264102e-06, "loss": 0.0832, "step": 10173 }, { "epoch": 6.509277031349968, "grad_norm": 0.8121537603782598, "learning_rate": 5.738532551292784e-06, "loss": 0.0857, "step": 10174 }, { "epoch": 6.509916826615483, "grad_norm": 0.8268757552849718, "learning_rate": 5.7366580603012985e-06, "loss": 0.084, "step": 10175 }, { "epoch": 6.510556621880998, "grad_norm": 0.7144009409720258, "learning_rate": 5.734783752370128e-06, "loss": 0.0809, "step": 10176 }, { "epoch": 6.511196417146513, "grad_norm": 0.8906791010617471, "learning_rate": 5.732909627579759e-06, "loss": 0.0886, "step": 10177 }, { "epoch": 6.511836212412028, "grad_norm": 0.9108911854260651, "learning_rate": 5.731035686010661e-06, "loss": 0.0918, "step": 10178 }, { "epoch": 6.512476007677543, "grad_norm": 0.8990106307444665, "learning_rate": 5.729161927743297e-06, "loss": 0.0825, "step": 10179 }, { "epoch": 6.513115802943059, "grad_norm": 0.9182038255624686, "learning_rate": 5.72728835285812e-06, "loss": 0.083, "step": 10180 }, { "epoch": 6.513755598208573, "grad_norm": 0.8308495012470569, "learning_rate": 5.725414961435584e-06, "loss": 0.0811, "step": 10181 }, { "epoch": 6.514395393474088, "grad_norm": 0.9265205424489938, "learning_rate": 5.723541753556119e-06, "loss": 0.0818, "step": 10182 }, { "epoch": 6.515035188739604, "grad_norm": 0.9262598304464297, "learning_rate": 5.721668729300169e-06, "loss": 0.0839, "step": 10183 }, { "epoch": 6.515674984005118, "grad_norm": 0.89498786367927, "learning_rate": 5.719795888748154e-06, "loss": 0.0837, "step": 10184 }, { "epoch": 6.516314779270633, "grad_norm": 0.9636464046471032, "learning_rate": 5.717923231980489e-06, "loss": 0.0891, "step": 10185 }, { "epoch": 6.516954574536149, "grad_norm": 0.8560376361004767, "learning_rate": 5.716050759077578e-06, "loss": 0.0839, "step": 10186 }, { "epoch": 6.517594369801664, "grad_norm": 0.9048548065697092, "learning_rate": 5.7141784701198335e-06, "loss": 0.0874, "step": 10187 }, { "epoch": 6.518234165067178, "grad_norm": 0.8587267952416789, "learning_rate": 5.712306365187639e-06, "loss": 0.0798, "step": 10188 }, { "epoch": 6.5188739603326935, "grad_norm": 0.9440597508541979, "learning_rate": 5.710434444361379e-06, "loss": 0.0841, "step": 10189 }, { "epoch": 6.519513755598209, "grad_norm": 0.8505248532526294, "learning_rate": 5.708562707721437e-06, "loss": 0.0803, "step": 10190 }, { "epoch": 6.520153550863723, "grad_norm": 0.9142791060990852, "learning_rate": 5.706691155348178e-06, "loss": 0.089, "step": 10191 }, { "epoch": 6.520793346129238, "grad_norm": 0.9047845724271757, "learning_rate": 5.7048197873219656e-06, "loss": 0.082, "step": 10192 }, { "epoch": 6.521433141394754, "grad_norm": 1.0182351837737136, "learning_rate": 5.7029486037231504e-06, "loss": 0.0959, "step": 10193 }, { "epoch": 6.522072936660269, "grad_norm": 0.8170074441327356, "learning_rate": 5.701077604632077e-06, "loss": 0.0798, "step": 10194 }, { "epoch": 6.522712731925783, "grad_norm": 0.9587435613245169, "learning_rate": 5.699206790129085e-06, "loss": 0.0845, "step": 10195 }, { "epoch": 6.523352527191299, "grad_norm": 0.8474226425294821, "learning_rate": 5.6973361602944985e-06, "loss": 0.0849, "step": 10196 }, { "epoch": 6.523992322456814, "grad_norm": 0.811715455161615, "learning_rate": 5.695465715208648e-06, "loss": 0.0777, "step": 10197 }, { "epoch": 6.524632117722329, "grad_norm": 0.8257758857715001, "learning_rate": 5.693595454951841e-06, "loss": 0.0788, "step": 10198 }, { "epoch": 6.525271912987844, "grad_norm": 0.9378711914932025, "learning_rate": 5.6917253796043825e-06, "loss": 0.079, "step": 10199 }, { "epoch": 6.525911708253359, "grad_norm": 0.883778713022197, "learning_rate": 5.6898554892465765e-06, "loss": 0.0867, "step": 10200 }, { "epoch": 6.526551503518874, "grad_norm": 0.8811256429142852, "learning_rate": 5.687985783958707e-06, "loss": 0.0789, "step": 10201 }, { "epoch": 6.527191298784389, "grad_norm": 0.7843865005282527, "learning_rate": 5.686116263821055e-06, "loss": 0.0812, "step": 10202 }, { "epoch": 6.527831094049904, "grad_norm": 0.8848702749708961, "learning_rate": 5.684246928913899e-06, "loss": 0.0856, "step": 10203 }, { "epoch": 6.528470889315419, "grad_norm": 0.8268411364840073, "learning_rate": 5.682377779317505e-06, "loss": 0.0847, "step": 10204 }, { "epoch": 6.529110684580934, "grad_norm": 0.8680341724792762, "learning_rate": 5.680508815112127e-06, "loss": 0.0865, "step": 10205 }, { "epoch": 6.529750479846449, "grad_norm": 0.8793715501433942, "learning_rate": 5.678640036378014e-06, "loss": 0.0857, "step": 10206 }, { "epoch": 6.530390275111964, "grad_norm": 0.826663381897093, "learning_rate": 5.676771443195413e-06, "loss": 0.0804, "step": 10207 }, { "epoch": 6.531030070377479, "grad_norm": 0.853666533879283, "learning_rate": 5.674903035644557e-06, "loss": 0.0802, "step": 10208 }, { "epoch": 6.531669865642995, "grad_norm": 0.9252105199294091, "learning_rate": 5.673034813805671e-06, "loss": 0.0919, "step": 10209 }, { "epoch": 6.532309660908509, "grad_norm": 0.9082419466444783, "learning_rate": 5.671166777758972e-06, "loss": 0.0838, "step": 10210 }, { "epoch": 6.532949456174024, "grad_norm": 0.8663529583915476, "learning_rate": 5.669298927584672e-06, "loss": 0.0795, "step": 10211 }, { "epoch": 6.5335892514395395, "grad_norm": 0.9197863500407749, "learning_rate": 5.667431263362973e-06, "loss": 0.091, "step": 10212 }, { "epoch": 6.534229046705054, "grad_norm": 0.8823960023242992, "learning_rate": 5.665563785174064e-06, "loss": 0.0849, "step": 10213 }, { "epoch": 6.534868841970569, "grad_norm": 0.8644806677839706, "learning_rate": 5.6636964930981385e-06, "loss": 0.0936, "step": 10214 }, { "epoch": 6.535508637236084, "grad_norm": 0.8966006490097996, "learning_rate": 5.661829387215373e-06, "loss": 0.0889, "step": 10215 }, { "epoch": 6.5361484325016, "grad_norm": 0.8515214221821406, "learning_rate": 5.659962467605933e-06, "loss": 0.0841, "step": 10216 }, { "epoch": 6.536788227767115, "grad_norm": 0.8601049762893004, "learning_rate": 5.658095734349987e-06, "loss": 0.0847, "step": 10217 }, { "epoch": 6.537428023032629, "grad_norm": 0.8279784764760755, "learning_rate": 5.656229187527687e-06, "loss": 0.0865, "step": 10218 }, { "epoch": 6.538067818298145, "grad_norm": 0.8408232600539334, "learning_rate": 5.654362827219175e-06, "loss": 0.0813, "step": 10219 }, { "epoch": 6.53870761356366, "grad_norm": 0.8819654957769469, "learning_rate": 5.652496653504598e-06, "loss": 0.0871, "step": 10220 }, { "epoch": 6.539347408829174, "grad_norm": 0.8561306859671926, "learning_rate": 5.650630666464079e-06, "loss": 0.0877, "step": 10221 }, { "epoch": 6.53998720409469, "grad_norm": 1.0113167278881137, "learning_rate": 5.648764866177744e-06, "loss": 0.0901, "step": 10222 }, { "epoch": 6.540626999360205, "grad_norm": 0.8144610070443044, "learning_rate": 5.6468992527257034e-06, "loss": 0.0762, "step": 10223 }, { "epoch": 6.54126679462572, "grad_norm": 0.8880529795113223, "learning_rate": 5.645033826188067e-06, "loss": 0.0852, "step": 10224 }, { "epoch": 6.5419065898912345, "grad_norm": 0.8854441419072726, "learning_rate": 5.6431685866449315e-06, "loss": 0.0901, "step": 10225 }, { "epoch": 6.54254638515675, "grad_norm": 0.8929924955911446, "learning_rate": 5.641303534176381e-06, "loss": 0.0864, "step": 10226 }, { "epoch": 6.543186180422265, "grad_norm": 0.8259564718137677, "learning_rate": 5.639438668862509e-06, "loss": 0.0794, "step": 10227 }, { "epoch": 6.5438259756877795, "grad_norm": 0.8597997637831721, "learning_rate": 5.6375739907833826e-06, "loss": 0.0852, "step": 10228 }, { "epoch": 6.544465770953295, "grad_norm": 0.7374031967760563, "learning_rate": 5.635709500019065e-06, "loss": 0.0764, "step": 10229 }, { "epoch": 6.54510556621881, "grad_norm": 0.8294361851893576, "learning_rate": 5.633845196649624e-06, "loss": 0.0804, "step": 10230 }, { "epoch": 6.545745361484325, "grad_norm": 0.7488204438191112, "learning_rate": 5.6319810807551e-06, "loss": 0.0814, "step": 10231 }, { "epoch": 6.54638515674984, "grad_norm": 0.8852651128539363, "learning_rate": 5.630117152415539e-06, "loss": 0.0851, "step": 10232 }, { "epoch": 6.547024952015355, "grad_norm": 0.8450565787229669, "learning_rate": 5.628253411710972e-06, "loss": 0.0819, "step": 10233 }, { "epoch": 6.54766474728087, "grad_norm": 0.8385976507326715, "learning_rate": 5.62638985872143e-06, "loss": 0.0768, "step": 10234 }, { "epoch": 6.5483045425463855, "grad_norm": 0.7664476220629843, "learning_rate": 5.624526493526927e-06, "loss": 0.0831, "step": 10235 }, { "epoch": 6.5489443378119, "grad_norm": 1.0114020418082046, "learning_rate": 5.62266331620747e-06, "loss": 0.092, "step": 10236 }, { "epoch": 6.549584133077415, "grad_norm": 0.8962691743847141, "learning_rate": 5.620800326843066e-06, "loss": 0.0878, "step": 10237 }, { "epoch": 6.55022392834293, "grad_norm": 0.8553146001918787, "learning_rate": 5.618937525513707e-06, "loss": 0.081, "step": 10238 }, { "epoch": 6.550863723608446, "grad_norm": 0.9034054439887648, "learning_rate": 5.617074912299376e-06, "loss": 0.0871, "step": 10239 }, { "epoch": 6.55150351887396, "grad_norm": 0.8671159516329496, "learning_rate": 5.615212487280053e-06, "loss": 0.0828, "step": 10240 }, { "epoch": 6.552143314139475, "grad_norm": 0.8600990801293068, "learning_rate": 5.613350250535706e-06, "loss": 0.084, "step": 10241 }, { "epoch": 6.552783109404991, "grad_norm": 0.8453104991252645, "learning_rate": 5.611488202146296e-06, "loss": 0.0805, "step": 10242 }, { "epoch": 6.553422904670505, "grad_norm": 0.8972010713599463, "learning_rate": 5.609626342191772e-06, "loss": 0.0772, "step": 10243 }, { "epoch": 6.55406269993602, "grad_norm": 0.8599733313104174, "learning_rate": 5.607764670752085e-06, "loss": 0.0824, "step": 10244 }, { "epoch": 6.554702495201536, "grad_norm": 0.8490171812573397, "learning_rate": 5.605903187907172e-06, "loss": 0.0873, "step": 10245 }, { "epoch": 6.555342290467051, "grad_norm": 0.8791819309829717, "learning_rate": 5.604041893736955e-06, "loss": 0.0866, "step": 10246 }, { "epoch": 6.555982085732565, "grad_norm": 0.9226383389901806, "learning_rate": 5.602180788321363e-06, "loss": 0.0855, "step": 10247 }, { "epoch": 6.5566218809980805, "grad_norm": 0.8049707535765392, "learning_rate": 5.600319871740305e-06, "loss": 0.0778, "step": 10248 }, { "epoch": 6.557261676263596, "grad_norm": 0.8542399968882025, "learning_rate": 5.598459144073682e-06, "loss": 0.0801, "step": 10249 }, { "epoch": 6.557901471529111, "grad_norm": 0.9212861510163399, "learning_rate": 5.596598605401398e-06, "loss": 0.0813, "step": 10250 }, { "epoch": 6.5585412667946255, "grad_norm": 0.8733017130458556, "learning_rate": 5.594738255803336e-06, "loss": 0.0879, "step": 10251 }, { "epoch": 6.559181062060141, "grad_norm": 0.8382053559073079, "learning_rate": 5.592878095359375e-06, "loss": 0.081, "step": 10252 }, { "epoch": 6.559820857325656, "grad_norm": 0.8355667197409705, "learning_rate": 5.5910181241493925e-06, "loss": 0.0817, "step": 10253 }, { "epoch": 6.560460652591171, "grad_norm": 0.8462797735300261, "learning_rate": 5.589158342253247e-06, "loss": 0.078, "step": 10254 }, { "epoch": 6.561100447856686, "grad_norm": 0.8272143467558841, "learning_rate": 5.587298749750796e-06, "loss": 0.0826, "step": 10255 }, { "epoch": 6.561740243122201, "grad_norm": 0.7441269293297252, "learning_rate": 5.585439346721882e-06, "loss": 0.0829, "step": 10256 }, { "epoch": 6.562380038387716, "grad_norm": 0.8696844087059447, "learning_rate": 5.583580133246355e-06, "loss": 0.0786, "step": 10257 }, { "epoch": 6.563019833653231, "grad_norm": 0.8788006404404671, "learning_rate": 5.581721109404041e-06, "loss": 0.08, "step": 10258 }, { "epoch": 6.563659628918746, "grad_norm": 0.8635823247251336, "learning_rate": 5.579862275274759e-06, "loss": 0.086, "step": 10259 }, { "epoch": 6.564299424184261, "grad_norm": 0.829622675051802, "learning_rate": 5.578003630938331e-06, "loss": 0.0886, "step": 10260 }, { "epoch": 6.564939219449776, "grad_norm": 0.9085566589815939, "learning_rate": 5.576145176474562e-06, "loss": 0.0881, "step": 10261 }, { "epoch": 6.565579014715291, "grad_norm": 0.7900892416262407, "learning_rate": 5.574286911963249e-06, "loss": 0.0768, "step": 10262 }, { "epoch": 6.566218809980806, "grad_norm": 0.9002227630993229, "learning_rate": 5.572428837484181e-06, "loss": 0.0844, "step": 10263 }, { "epoch": 6.566858605246321, "grad_norm": 0.8506245163294898, "learning_rate": 5.570570953117145e-06, "loss": 0.0815, "step": 10264 }, { "epoch": 6.567498400511836, "grad_norm": 0.8657852382687717, "learning_rate": 5.5687132589419135e-06, "loss": 0.0859, "step": 10265 }, { "epoch": 6.568138195777351, "grad_norm": 0.9273124563287498, "learning_rate": 5.5668557550382475e-06, "loss": 0.0897, "step": 10266 }, { "epoch": 6.568777991042866, "grad_norm": 0.7779154885401072, "learning_rate": 5.564998441485914e-06, "loss": 0.086, "step": 10267 }, { "epoch": 6.569417786308382, "grad_norm": 0.9750401309720699, "learning_rate": 5.563141318364657e-06, "loss": 0.0911, "step": 10268 }, { "epoch": 6.570057581573896, "grad_norm": 0.8986262542538541, "learning_rate": 5.56128438575422e-06, "loss": 0.0843, "step": 10269 }, { "epoch": 6.570697376839411, "grad_norm": 0.8903778680239555, "learning_rate": 5.559427643734336e-06, "loss": 0.0827, "step": 10270 }, { "epoch": 6.5713371721049265, "grad_norm": 0.8430360286033437, "learning_rate": 5.5575710923847305e-06, "loss": 0.0822, "step": 10271 }, { "epoch": 6.571976967370442, "grad_norm": 0.9415629672772904, "learning_rate": 5.555714731785118e-06, "loss": 0.09, "step": 10272 }, { "epoch": 6.572616762635956, "grad_norm": 0.8598949387636361, "learning_rate": 5.5538585620152065e-06, "loss": 0.0917, "step": 10273 }, { "epoch": 6.5732565579014715, "grad_norm": 0.8644147395186983, "learning_rate": 5.552002583154703e-06, "loss": 0.0798, "step": 10274 }, { "epoch": 6.573896353166987, "grad_norm": 0.7860049142886014, "learning_rate": 5.550146795283295e-06, "loss": 0.0821, "step": 10275 }, { "epoch": 6.574536148432502, "grad_norm": 0.7656816966373929, "learning_rate": 5.548291198480665e-06, "loss": 0.0763, "step": 10276 }, { "epoch": 6.575175943698016, "grad_norm": 0.7742659478743679, "learning_rate": 5.546435792826497e-06, "loss": 0.0716, "step": 10277 }, { "epoch": 6.575815738963532, "grad_norm": 0.808561546946295, "learning_rate": 5.544580578400452e-06, "loss": 0.0876, "step": 10278 }, { "epoch": 6.576455534229047, "grad_norm": 0.7947511099805247, "learning_rate": 5.542725555282188e-06, "loss": 0.0739, "step": 10279 }, { "epoch": 6.577095329494561, "grad_norm": 0.819703427861728, "learning_rate": 5.540870723551364e-06, "loss": 0.0791, "step": 10280 }, { "epoch": 6.577735124760077, "grad_norm": 0.8225812035471989, "learning_rate": 5.539016083287618e-06, "loss": 0.0868, "step": 10281 }, { "epoch": 6.578374920025592, "grad_norm": 0.89462180552004, "learning_rate": 5.537161634570587e-06, "loss": 0.0895, "step": 10282 }, { "epoch": 6.579014715291107, "grad_norm": 0.8937298300027516, "learning_rate": 5.535307377479897e-06, "loss": 0.0876, "step": 10283 }, { "epoch": 6.5796545105566215, "grad_norm": 0.8335295275184671, "learning_rate": 5.533453312095165e-06, "loss": 0.084, "step": 10284 }, { "epoch": 6.580294305822137, "grad_norm": 0.8630642546201965, "learning_rate": 5.5315994384960026e-06, "loss": 0.0848, "step": 10285 }, { "epoch": 6.580934101087652, "grad_norm": 0.8972274038924651, "learning_rate": 5.529745756762008e-06, "loss": 0.0816, "step": 10286 }, { "epoch": 6.581573896353167, "grad_norm": 0.850415267756831, "learning_rate": 5.527892266972783e-06, "loss": 0.0748, "step": 10287 }, { "epoch": 6.582213691618682, "grad_norm": 0.9562290086998334, "learning_rate": 5.526038969207907e-06, "loss": 0.0875, "step": 10288 }, { "epoch": 6.582853486884197, "grad_norm": 0.7961062081880178, "learning_rate": 5.524185863546963e-06, "loss": 0.0847, "step": 10289 }, { "epoch": 6.583493282149712, "grad_norm": 0.8416100581411602, "learning_rate": 5.522332950069509e-06, "loss": 0.0824, "step": 10290 }, { "epoch": 6.584133077415228, "grad_norm": 0.8417195252442902, "learning_rate": 5.520480228855116e-06, "loss": 0.0789, "step": 10291 }, { "epoch": 6.584772872680742, "grad_norm": 0.7745161996326106, "learning_rate": 5.518627699983336e-06, "loss": 0.0761, "step": 10292 }, { "epoch": 6.585412667946257, "grad_norm": 0.8799310393309135, "learning_rate": 5.516775363533706e-06, "loss": 0.081, "step": 10293 }, { "epoch": 6.5860524632117725, "grad_norm": 0.892940295191696, "learning_rate": 5.514923219585771e-06, "loss": 0.0868, "step": 10294 }, { "epoch": 6.586692258477287, "grad_norm": 0.9820113296318812, "learning_rate": 5.513071268219055e-06, "loss": 0.0922, "step": 10295 }, { "epoch": 6.587332053742802, "grad_norm": 0.9197725411532177, "learning_rate": 5.511219509513072e-06, "loss": 0.0899, "step": 10296 }, { "epoch": 6.5879718490083174, "grad_norm": 0.905741713388965, "learning_rate": 5.509367943547344e-06, "loss": 0.084, "step": 10297 }, { "epoch": 6.588611644273833, "grad_norm": 0.9283931038625735, "learning_rate": 5.507516570401366e-06, "loss": 0.0913, "step": 10298 }, { "epoch": 6.589251439539347, "grad_norm": 0.9742140187235886, "learning_rate": 5.505665390154636e-06, "loss": 0.0871, "step": 10299 }, { "epoch": 6.589891234804862, "grad_norm": 0.8687124850698058, "learning_rate": 5.50381440288664e-06, "loss": 0.0828, "step": 10300 }, { "epoch": 6.590531030070378, "grad_norm": 0.8250333219446717, "learning_rate": 5.501963608676855e-06, "loss": 0.0769, "step": 10301 }, { "epoch": 6.591170825335892, "grad_norm": 1.0655154612181768, "learning_rate": 5.5001130076047505e-06, "loss": 0.0885, "step": 10302 }, { "epoch": 6.591810620601407, "grad_norm": 0.9192080241723561, "learning_rate": 5.498262599749787e-06, "loss": 0.0825, "step": 10303 }, { "epoch": 6.592450415866923, "grad_norm": 0.8049281764016851, "learning_rate": 5.496412385191421e-06, "loss": 0.0803, "step": 10304 }, { "epoch": 6.593090211132438, "grad_norm": 0.8863618945712961, "learning_rate": 5.4945623640090964e-06, "loss": 0.084, "step": 10305 }, { "epoch": 6.593730006397953, "grad_norm": 0.7788946906365194, "learning_rate": 5.4927125362822455e-06, "loss": 0.0746, "step": 10306 }, { "epoch": 6.5943698016634675, "grad_norm": 0.889091529822738, "learning_rate": 5.4908629020903044e-06, "loss": 0.0819, "step": 10307 }, { "epoch": 6.595009596928983, "grad_norm": 0.7733727284570872, "learning_rate": 5.48901346151269e-06, "loss": 0.0822, "step": 10308 }, { "epoch": 6.595649392194498, "grad_norm": 0.7369234381822085, "learning_rate": 5.487164214628813e-06, "loss": 0.085, "step": 10309 }, { "epoch": 6.5962891874600125, "grad_norm": 0.9007574918379911, "learning_rate": 5.4853151615180725e-06, "loss": 0.0828, "step": 10310 }, { "epoch": 6.596928982725528, "grad_norm": 0.8226843661350023, "learning_rate": 5.483466302259872e-06, "loss": 0.0858, "step": 10311 }, { "epoch": 6.597568777991043, "grad_norm": 0.8384379328503482, "learning_rate": 5.481617636933594e-06, "loss": 0.0797, "step": 10312 }, { "epoch": 6.598208573256558, "grad_norm": 0.8877270664089995, "learning_rate": 5.479769165618618e-06, "loss": 0.0874, "step": 10313 }, { "epoch": 6.598848368522073, "grad_norm": 0.8501415499365063, "learning_rate": 5.4779208883943125e-06, "loss": 0.0817, "step": 10314 }, { "epoch": 6.599488163787588, "grad_norm": 0.7678892825130307, "learning_rate": 5.476072805340039e-06, "loss": 0.0748, "step": 10315 }, { "epoch": 6.600127959053103, "grad_norm": 0.8165409015863939, "learning_rate": 5.474224916535149e-06, "loss": 0.0849, "step": 10316 }, { "epoch": 6.600767754318618, "grad_norm": 0.8562922510213693, "learning_rate": 5.472377222058993e-06, "loss": 0.0816, "step": 10317 }, { "epoch": 6.601407549584133, "grad_norm": 1.0917570355060184, "learning_rate": 5.470529721990907e-06, "loss": 0.0947, "step": 10318 }, { "epoch": 6.602047344849648, "grad_norm": 0.8867210534236415, "learning_rate": 5.468682416410216e-06, "loss": 0.0792, "step": 10319 }, { "epoch": 6.6026871401151634, "grad_norm": 0.8577618223550008, "learning_rate": 5.466835305396239e-06, "loss": 0.0787, "step": 10320 }, { "epoch": 6.603326935380678, "grad_norm": 0.9524026366370242, "learning_rate": 5.464988389028292e-06, "loss": 0.0908, "step": 10321 }, { "epoch": 6.603966730646193, "grad_norm": 0.9445483224953573, "learning_rate": 5.4631416673856786e-06, "loss": 0.0801, "step": 10322 }, { "epoch": 6.604606525911708, "grad_norm": 0.9464226920142064, "learning_rate": 5.461295140547687e-06, "loss": 0.0904, "step": 10323 }, { "epoch": 6.605246321177224, "grad_norm": 0.8935116390391209, "learning_rate": 5.459448808593611e-06, "loss": 0.0755, "step": 10324 }, { "epoch": 6.605886116442738, "grad_norm": 0.9331983158702646, "learning_rate": 5.457602671602727e-06, "loss": 0.0887, "step": 10325 }, { "epoch": 6.606525911708253, "grad_norm": 0.93410106374688, "learning_rate": 5.4557567296543e-06, "loss": 0.0883, "step": 10326 }, { "epoch": 6.607165706973769, "grad_norm": 0.851292660504022, "learning_rate": 5.453910982827599e-06, "loss": 0.0847, "step": 10327 }, { "epoch": 6.607805502239284, "grad_norm": 0.8333501388533164, "learning_rate": 5.452065431201874e-06, "loss": 0.0776, "step": 10328 }, { "epoch": 6.608445297504798, "grad_norm": 0.8576686952753149, "learning_rate": 5.4502200748563695e-06, "loss": 0.0915, "step": 10329 }, { "epoch": 6.6090850927703135, "grad_norm": 0.8851809286860834, "learning_rate": 5.44837491387032e-06, "loss": 0.0864, "step": 10330 }, { "epoch": 6.609724888035829, "grad_norm": 0.878676918356743, "learning_rate": 5.446529948322956e-06, "loss": 0.0856, "step": 10331 }, { "epoch": 6.610364683301343, "grad_norm": 0.8413649840706209, "learning_rate": 5.444685178293496e-06, "loss": 0.0792, "step": 10332 }, { "epoch": 6.6110044785668585, "grad_norm": 0.8322906589531957, "learning_rate": 5.442840603861148e-06, "loss": 0.0769, "step": 10333 }, { "epoch": 6.611644273832374, "grad_norm": 0.9415580867362943, "learning_rate": 5.440996225105122e-06, "loss": 0.0868, "step": 10334 }, { "epoch": 6.612284069097889, "grad_norm": 0.8057039204984885, "learning_rate": 5.4391520421046076e-06, "loss": 0.0798, "step": 10335 }, { "epoch": 6.612923864363403, "grad_norm": 0.8184741962185584, "learning_rate": 5.437308054938787e-06, "loss": 0.0822, "step": 10336 }, { "epoch": 6.613563659628919, "grad_norm": 0.8713667723284128, "learning_rate": 5.435464263686847e-06, "loss": 0.0867, "step": 10337 }, { "epoch": 6.614203454894434, "grad_norm": 0.9539793490534038, "learning_rate": 5.43362066842795e-06, "loss": 0.0927, "step": 10338 }, { "epoch": 6.614843250159948, "grad_norm": 0.8402128914348026, "learning_rate": 5.431777269241261e-06, "loss": 0.0882, "step": 10339 }, { "epoch": 6.615483045425464, "grad_norm": 0.9087865818158214, "learning_rate": 5.429934066205924e-06, "loss": 0.0862, "step": 10340 }, { "epoch": 6.616122840690979, "grad_norm": 0.9386078991276973, "learning_rate": 5.428091059401093e-06, "loss": 0.0825, "step": 10341 }, { "epoch": 6.616762635956494, "grad_norm": 0.8856363483917117, "learning_rate": 5.4262482489058995e-06, "loss": 0.0859, "step": 10342 }, { "epoch": 6.617402431222009, "grad_norm": 0.8828866699282302, "learning_rate": 5.42440563479947e-06, "loss": 0.083, "step": 10343 }, { "epoch": 6.618042226487524, "grad_norm": 0.9322697425383326, "learning_rate": 5.422563217160923e-06, "loss": 0.0863, "step": 10344 }, { "epoch": 6.618682021753039, "grad_norm": 0.8549372328329419, "learning_rate": 5.4207209960693694e-06, "loss": 0.0825, "step": 10345 }, { "epoch": 6.619321817018554, "grad_norm": 0.8327628221880226, "learning_rate": 5.418878971603907e-06, "loss": 0.0801, "step": 10346 }, { "epoch": 6.619961612284069, "grad_norm": 0.8447099595888534, "learning_rate": 5.417037143843635e-06, "loss": 0.0827, "step": 10347 }, { "epoch": 6.620601407549584, "grad_norm": 0.8720014206237934, "learning_rate": 5.4151955128676346e-06, "loss": 0.0827, "step": 10348 }, { "epoch": 6.621241202815099, "grad_norm": 0.7886751850382963, "learning_rate": 5.413354078754985e-06, "loss": 0.0837, "step": 10349 }, { "epoch": 6.621880998080615, "grad_norm": 0.9815139484101232, "learning_rate": 5.4115128415847484e-06, "loss": 0.081, "step": 10350 }, { "epoch": 6.622520793346129, "grad_norm": 0.9003717798273341, "learning_rate": 5.409671801435991e-06, "loss": 0.0846, "step": 10351 }, { "epoch": 6.623160588611644, "grad_norm": 0.7802345973705636, "learning_rate": 5.407830958387761e-06, "loss": 0.0797, "step": 10352 }, { "epoch": 6.6238003838771595, "grad_norm": 0.8714691245411141, "learning_rate": 5.405990312519098e-06, "loss": 0.0832, "step": 10353 }, { "epoch": 6.624440179142674, "grad_norm": 0.7833224785829863, "learning_rate": 5.4041498639090415e-06, "loss": 0.0788, "step": 10354 }, { "epoch": 6.625079974408189, "grad_norm": 0.7357330234103538, "learning_rate": 5.402309612636616e-06, "loss": 0.0806, "step": 10355 }, { "epoch": 6.6257197696737045, "grad_norm": 0.8792129061882819, "learning_rate": 5.400469558780832e-06, "loss": 0.0866, "step": 10356 }, { "epoch": 6.62635956493922, "grad_norm": 0.7151848060290937, "learning_rate": 5.398629702420707e-06, "loss": 0.0803, "step": 10357 }, { "epoch": 6.626999360204734, "grad_norm": 0.8422344377046478, "learning_rate": 5.396790043635238e-06, "loss": 0.0885, "step": 10358 }, { "epoch": 6.627639155470249, "grad_norm": 1.0535982089655676, "learning_rate": 5.394950582503415e-06, "loss": 0.0833, "step": 10359 }, { "epoch": 6.628278950735765, "grad_norm": 0.8565845149906517, "learning_rate": 5.3931113191042225e-06, "loss": 0.0851, "step": 10360 }, { "epoch": 6.62891874600128, "grad_norm": 0.8958118086290819, "learning_rate": 5.3912722535166355e-06, "loss": 0.096, "step": 10361 }, { "epoch": 6.629558541266794, "grad_norm": 0.9595320687459354, "learning_rate": 5.38943338581962e-06, "loss": 0.0868, "step": 10362 }, { "epoch": 6.63019833653231, "grad_norm": 0.8622015434195065, "learning_rate": 5.387594716092128e-06, "loss": 0.0829, "step": 10363 }, { "epoch": 6.630838131797825, "grad_norm": 0.7902904829581894, "learning_rate": 5.385756244413119e-06, "loss": 0.0826, "step": 10364 }, { "epoch": 6.63147792706334, "grad_norm": 0.9044167423105913, "learning_rate": 5.38391797086153e-06, "loss": 0.0854, "step": 10365 }, { "epoch": 6.6321177223288545, "grad_norm": 0.8613597649391129, "learning_rate": 5.382079895516285e-06, "loss": 0.0873, "step": 10366 }, { "epoch": 6.63275751759437, "grad_norm": 0.9036794999216853, "learning_rate": 5.380242018456321e-06, "loss": 0.0921, "step": 10367 }, { "epoch": 6.633397312859885, "grad_norm": 0.8785617179293838, "learning_rate": 5.378404339760546e-06, "loss": 0.0849, "step": 10368 }, { "epoch": 6.6340371081253995, "grad_norm": 0.8123511679140889, "learning_rate": 5.376566859507868e-06, "loss": 0.0818, "step": 10369 }, { "epoch": 6.634676903390915, "grad_norm": 1.0241396950779593, "learning_rate": 5.37472957777718e-06, "loss": 0.0943, "step": 10370 }, { "epoch": 6.63531669865643, "grad_norm": 0.8657087189222006, "learning_rate": 5.3728924946473804e-06, "loss": 0.0839, "step": 10371 }, { "epoch": 6.635956493921945, "grad_norm": 0.8011339619027293, "learning_rate": 5.371055610197347e-06, "loss": 0.0828, "step": 10372 }, { "epoch": 6.63659628918746, "grad_norm": 0.9370349222203915, "learning_rate": 5.36921892450595e-06, "loss": 0.0814, "step": 10373 }, { "epoch": 6.637236084452975, "grad_norm": 0.8481433452437059, "learning_rate": 5.3673824376520555e-06, "loss": 0.0809, "step": 10374 }, { "epoch": 6.63787587971849, "grad_norm": 0.8453018497674366, "learning_rate": 5.365546149714518e-06, "loss": 0.0873, "step": 10375 }, { "epoch": 6.638515674984005, "grad_norm": 0.8977841945001415, "learning_rate": 5.3637100607721845e-06, "loss": 0.0747, "step": 10376 }, { "epoch": 6.63915547024952, "grad_norm": 0.8114752154737332, "learning_rate": 5.3618741709038904e-06, "loss": 0.0782, "step": 10377 }, { "epoch": 6.639795265515035, "grad_norm": 0.9372469390285088, "learning_rate": 5.360038480188472e-06, "loss": 0.0913, "step": 10378 }, { "epoch": 6.6404350607805505, "grad_norm": 0.8369118812806253, "learning_rate": 5.358202988704748e-06, "loss": 0.0917, "step": 10379 }, { "epoch": 6.641074856046066, "grad_norm": 0.7312791943696969, "learning_rate": 5.356367696531527e-06, "loss": 0.0667, "step": 10380 }, { "epoch": 6.64171465131158, "grad_norm": 0.8338139867902863, "learning_rate": 5.35453260374762e-06, "loss": 0.0833, "step": 10381 }, { "epoch": 6.642354446577095, "grad_norm": 0.8170950780083329, "learning_rate": 5.35269771043182e-06, "loss": 0.0799, "step": 10382 }, { "epoch": 6.642994241842611, "grad_norm": 0.9065987702089499, "learning_rate": 5.3508630166629096e-06, "loss": 0.0855, "step": 10383 }, { "epoch": 6.643634037108125, "grad_norm": 0.852486451448028, "learning_rate": 5.349028522519675e-06, "loss": 0.0827, "step": 10384 }, { "epoch": 6.64427383237364, "grad_norm": 0.8728711062914228, "learning_rate": 5.347194228080881e-06, "loss": 0.082, "step": 10385 }, { "epoch": 6.644913627639156, "grad_norm": 0.8053580051392277, "learning_rate": 5.34536013342529e-06, "loss": 0.0847, "step": 10386 }, { "epoch": 6.645553422904671, "grad_norm": 0.9006886893714865, "learning_rate": 5.343526238631653e-06, "loss": 0.0845, "step": 10387 }, { "epoch": 6.646193218170185, "grad_norm": 0.7872699177014303, "learning_rate": 5.341692543778718e-06, "loss": 0.083, "step": 10388 }, { "epoch": 6.6468330134357005, "grad_norm": 0.9795022310263644, "learning_rate": 5.339859048945219e-06, "loss": 0.0903, "step": 10389 }, { "epoch": 6.647472808701216, "grad_norm": 0.8318790144495787, "learning_rate": 5.338025754209883e-06, "loss": 0.0805, "step": 10390 }, { "epoch": 6.64811260396673, "grad_norm": 0.8132994933011871, "learning_rate": 5.336192659651429e-06, "loss": 0.0865, "step": 10391 }, { "epoch": 6.6487523992322455, "grad_norm": 0.94315490927088, "learning_rate": 5.334359765348566e-06, "loss": 0.0852, "step": 10392 }, { "epoch": 6.649392194497761, "grad_norm": 0.9382009466125554, "learning_rate": 5.332527071379991e-06, "loss": 0.0918, "step": 10393 }, { "epoch": 6.650031989763276, "grad_norm": 0.7822976007682242, "learning_rate": 5.330694577824403e-06, "loss": 0.0815, "step": 10394 }, { "epoch": 6.65067178502879, "grad_norm": 0.81505740023742, "learning_rate": 5.3288622847604856e-06, "loss": 0.0752, "step": 10395 }, { "epoch": 6.651311580294306, "grad_norm": 0.8552434953939119, "learning_rate": 5.327030192266911e-06, "loss": 0.0878, "step": 10396 }, { "epoch": 6.651951375559821, "grad_norm": 0.8847238343127488, "learning_rate": 5.3251983004223455e-06, "loss": 0.0833, "step": 10397 }, { "epoch": 6.652591170825336, "grad_norm": 0.8702824462433735, "learning_rate": 5.323366609305451e-06, "loss": 0.0846, "step": 10398 }, { "epoch": 6.653230966090851, "grad_norm": 0.8101229061085554, "learning_rate": 5.321535118994876e-06, "loss": 0.0825, "step": 10399 }, { "epoch": 6.653870761356366, "grad_norm": 0.8514156749028396, "learning_rate": 5.319703829569256e-06, "loss": 0.0839, "step": 10400 }, { "epoch": 6.654510556621881, "grad_norm": 0.9109399622361011, "learning_rate": 5.317872741107233e-06, "loss": 0.0795, "step": 10401 }, { "epoch": 6.6551503518873965, "grad_norm": 0.8504304778878246, "learning_rate": 5.316041853687426e-06, "loss": 0.0881, "step": 10402 }, { "epoch": 6.655790147152911, "grad_norm": 0.7544337873004637, "learning_rate": 5.314211167388448e-06, "loss": 0.0817, "step": 10403 }, { "epoch": 6.656429942418426, "grad_norm": 0.8397517403842788, "learning_rate": 5.312380682288909e-06, "loss": 0.0815, "step": 10404 }, { "epoch": 6.657069737683941, "grad_norm": 0.8431071557086565, "learning_rate": 5.310550398467403e-06, "loss": 0.0814, "step": 10405 }, { "epoch": 6.657709532949456, "grad_norm": 0.9105233644659227, "learning_rate": 5.3087203160025226e-06, "loss": 0.0905, "step": 10406 }, { "epoch": 6.658349328214971, "grad_norm": 0.9822027183281042, "learning_rate": 5.3068904349728425e-06, "loss": 0.0877, "step": 10407 }, { "epoch": 6.658989123480486, "grad_norm": 0.9046239307852355, "learning_rate": 5.305060755456941e-06, "loss": 0.0878, "step": 10408 }, { "epoch": 6.659628918746002, "grad_norm": 0.7939996611131757, "learning_rate": 5.30323127753338e-06, "loss": 0.0885, "step": 10409 }, { "epoch": 6.660268714011516, "grad_norm": 0.879796866402821, "learning_rate": 5.30140200128071e-06, "loss": 0.0889, "step": 10410 }, { "epoch": 6.660908509277031, "grad_norm": 0.9057370374507636, "learning_rate": 5.299572926777483e-06, "loss": 0.0805, "step": 10411 }, { "epoch": 6.6615483045425465, "grad_norm": 0.8511323050908571, "learning_rate": 5.297744054102234e-06, "loss": 0.085, "step": 10412 }, { "epoch": 6.662188099808061, "grad_norm": 0.7807928266843714, "learning_rate": 5.295915383333485e-06, "loss": 0.0816, "step": 10413 }, { "epoch": 6.662827895073576, "grad_norm": 0.949668943736872, "learning_rate": 5.294086914549767e-06, "loss": 0.0889, "step": 10414 }, { "epoch": 6.6634676903390915, "grad_norm": 0.8607209979400932, "learning_rate": 5.292258647829585e-06, "loss": 0.0791, "step": 10415 }, { "epoch": 6.664107485604607, "grad_norm": 0.8620146393979211, "learning_rate": 5.2904305832514424e-06, "loss": 0.0804, "step": 10416 }, { "epoch": 6.664747280870122, "grad_norm": 0.8969032759104083, "learning_rate": 5.288602720893829e-06, "loss": 0.0897, "step": 10417 }, { "epoch": 6.665387076135636, "grad_norm": 0.8432379265128045, "learning_rate": 5.286775060835238e-06, "loss": 0.0874, "step": 10418 }, { "epoch": 6.666026871401152, "grad_norm": 0.8723044430624186, "learning_rate": 5.284947603154141e-06, "loss": 0.0799, "step": 10419 }, { "epoch": 6.666666666666667, "grad_norm": 0.8212048321680391, "learning_rate": 5.283120347929007e-06, "loss": 0.086, "step": 10420 }, { "epoch": 6.667306461932181, "grad_norm": 0.9799031091038859, "learning_rate": 5.281293295238295e-06, "loss": 0.095, "step": 10421 }, { "epoch": 6.667946257197697, "grad_norm": 0.8779619717063982, "learning_rate": 5.279466445160454e-06, "loss": 0.0875, "step": 10422 }, { "epoch": 6.668586052463212, "grad_norm": 0.7678306836966297, "learning_rate": 5.277639797773923e-06, "loss": 0.0833, "step": 10423 }, { "epoch": 6.669225847728727, "grad_norm": 0.8561165249562479, "learning_rate": 5.275813353157143e-06, "loss": 0.0854, "step": 10424 }, { "epoch": 6.669865642994242, "grad_norm": 0.7648702884254556, "learning_rate": 5.273987111388534e-06, "loss": 0.0832, "step": 10425 }, { "epoch": 6.670505438259757, "grad_norm": 0.8728000274130892, "learning_rate": 5.272161072546511e-06, "loss": 0.0811, "step": 10426 }, { "epoch": 6.671145233525272, "grad_norm": 0.8870234206711299, "learning_rate": 5.270335236709479e-06, "loss": 0.0803, "step": 10427 }, { "epoch": 6.6717850287907865, "grad_norm": 0.8762657799373413, "learning_rate": 5.26850960395584e-06, "loss": 0.0814, "step": 10428 }, { "epoch": 6.672424824056302, "grad_norm": 0.9370254435654911, "learning_rate": 5.266684174363983e-06, "loss": 0.083, "step": 10429 }, { "epoch": 6.673064619321817, "grad_norm": 0.9184518690826196, "learning_rate": 5.264858948012284e-06, "loss": 0.0915, "step": 10430 }, { "epoch": 6.673704414587332, "grad_norm": 0.855398185023739, "learning_rate": 5.263033924979122e-06, "loss": 0.0861, "step": 10431 }, { "epoch": 6.674344209852847, "grad_norm": 0.9951142615479741, "learning_rate": 5.261209105342857e-06, "loss": 0.0893, "step": 10432 }, { "epoch": 6.674984005118362, "grad_norm": 0.8185595507541489, "learning_rate": 5.259384489181844e-06, "loss": 0.0816, "step": 10433 }, { "epoch": 6.675623800383877, "grad_norm": 0.7837723745812353, "learning_rate": 5.257560076574428e-06, "loss": 0.0832, "step": 10434 }, { "epoch": 6.6762635956493925, "grad_norm": 0.9861949617976957, "learning_rate": 5.255735867598946e-06, "loss": 0.0826, "step": 10435 }, { "epoch": 6.676903390914907, "grad_norm": 0.8521091310820701, "learning_rate": 5.253911862333726e-06, "loss": 0.0806, "step": 10436 }, { "epoch": 6.677543186180422, "grad_norm": 0.9138124709503876, "learning_rate": 5.252088060857083e-06, "loss": 0.0796, "step": 10437 }, { "epoch": 6.6781829814459375, "grad_norm": 0.9328209862120701, "learning_rate": 5.250264463247339e-06, "loss": 0.0898, "step": 10438 }, { "epoch": 6.678822776711453, "grad_norm": 0.8337637688278925, "learning_rate": 5.248441069582789e-06, "loss": 0.0901, "step": 10439 }, { "epoch": 6.679462571976967, "grad_norm": 0.7959892024313951, "learning_rate": 5.246617879941724e-06, "loss": 0.078, "step": 10440 }, { "epoch": 6.680102367242482, "grad_norm": 0.8167894812936596, "learning_rate": 5.244794894402434e-06, "loss": 0.0799, "step": 10441 }, { "epoch": 6.680742162507998, "grad_norm": 0.7939928163307023, "learning_rate": 5.242972113043193e-06, "loss": 0.0791, "step": 10442 }, { "epoch": 6.681381957773512, "grad_norm": 0.8737135096139221, "learning_rate": 5.2411495359422635e-06, "loss": 0.0771, "step": 10443 }, { "epoch": 6.682021753039027, "grad_norm": 0.8566273993421493, "learning_rate": 5.239327163177912e-06, "loss": 0.0839, "step": 10444 }, { "epoch": 6.682661548304543, "grad_norm": 0.94679576683554, "learning_rate": 5.237504994828382e-06, "loss": 0.094, "step": 10445 }, { "epoch": 6.683301343570058, "grad_norm": 0.7977242034925386, "learning_rate": 5.235683030971916e-06, "loss": 0.0805, "step": 10446 }, { "epoch": 6.683941138835572, "grad_norm": 0.8150663788336174, "learning_rate": 5.2338612716867445e-06, "loss": 0.0824, "step": 10447 }, { "epoch": 6.684580934101088, "grad_norm": 0.905158836379685, "learning_rate": 5.232039717051093e-06, "loss": 0.0765, "step": 10448 }, { "epoch": 6.685220729366603, "grad_norm": 0.8118831741132441, "learning_rate": 5.230218367143174e-06, "loss": 0.081, "step": 10449 }, { "epoch": 6.685860524632118, "grad_norm": 0.781766524673589, "learning_rate": 5.228397222041195e-06, "loss": 0.0808, "step": 10450 }, { "epoch": 6.6865003198976325, "grad_norm": 0.8964946481711836, "learning_rate": 5.22657628182335e-06, "loss": 0.0893, "step": 10451 }, { "epoch": 6.687140115163148, "grad_norm": 0.9046212339031765, "learning_rate": 5.22475554656783e-06, "loss": 0.0863, "step": 10452 }, { "epoch": 6.687779910428663, "grad_norm": 0.7815127835229754, "learning_rate": 5.222935016352812e-06, "loss": 0.0764, "step": 10453 }, { "epoch": 6.688419705694178, "grad_norm": 0.9094472478972482, "learning_rate": 5.221114691256461e-06, "loss": 0.0837, "step": 10454 }, { "epoch": 6.689059500959693, "grad_norm": 0.8575809394680777, "learning_rate": 5.21929457135695e-06, "loss": 0.0874, "step": 10455 }, { "epoch": 6.689699296225208, "grad_norm": 0.7926833955026472, "learning_rate": 5.217474656732424e-06, "loss": 0.077, "step": 10456 }, { "epoch": 6.690339091490723, "grad_norm": 0.869381272475933, "learning_rate": 5.2156549474610265e-06, "loss": 0.0891, "step": 10457 }, { "epoch": 6.690978886756238, "grad_norm": 0.9623133730279422, "learning_rate": 5.213835443620898e-06, "loss": 0.0918, "step": 10458 }, { "epoch": 6.691618682021753, "grad_norm": 0.7722417156656932, "learning_rate": 5.212016145290162e-06, "loss": 0.0796, "step": 10459 }, { "epoch": 6.692258477287268, "grad_norm": 0.8424820917115995, "learning_rate": 5.210197052546931e-06, "loss": 0.0814, "step": 10460 }, { "epoch": 6.6928982725527835, "grad_norm": 0.8870571375425416, "learning_rate": 5.2083781654693234e-06, "loss": 0.0868, "step": 10461 }, { "epoch": 6.693538067818298, "grad_norm": 0.9151930279185594, "learning_rate": 5.2065594841354315e-06, "loss": 0.0897, "step": 10462 }, { "epoch": 6.694177863083813, "grad_norm": 0.8612827156410763, "learning_rate": 5.204741008623351e-06, "loss": 0.0881, "step": 10463 }, { "epoch": 6.694817658349328, "grad_norm": 0.9707546868836665, "learning_rate": 5.202922739011161e-06, "loss": 0.0835, "step": 10464 }, { "epoch": 6.695457453614843, "grad_norm": 0.8469263446496861, "learning_rate": 5.201104675376933e-06, "loss": 0.0775, "step": 10465 }, { "epoch": 6.696097248880358, "grad_norm": 0.8831453204021674, "learning_rate": 5.199286817798736e-06, "loss": 0.0849, "step": 10466 }, { "epoch": 6.696737044145873, "grad_norm": 0.9113289558227661, "learning_rate": 5.197469166354619e-06, "loss": 0.0915, "step": 10467 }, { "epoch": 6.697376839411389, "grad_norm": 0.8005953348887224, "learning_rate": 5.1956517211226366e-06, "loss": 0.0833, "step": 10468 }, { "epoch": 6.698016634676903, "grad_norm": 0.8337995312521619, "learning_rate": 5.193834482180822e-06, "loss": 0.0857, "step": 10469 }, { "epoch": 6.698656429942418, "grad_norm": 0.9034743226506291, "learning_rate": 5.192017449607203e-06, "loss": 0.0876, "step": 10470 }, { "epoch": 6.6992962252079336, "grad_norm": 0.8406961398243663, "learning_rate": 5.1902006234798045e-06, "loss": 0.0833, "step": 10471 }, { "epoch": 6.699936020473449, "grad_norm": 0.86229926260027, "learning_rate": 5.188384003876637e-06, "loss": 0.0773, "step": 10472 }, { "epoch": 6.700575815738963, "grad_norm": 0.7995779594304232, "learning_rate": 5.1865675908757e-06, "loss": 0.0872, "step": 10473 }, { "epoch": 6.7012156110044785, "grad_norm": 0.8147157999557619, "learning_rate": 5.184751384554983e-06, "loss": 0.0784, "step": 10474 }, { "epoch": 6.701855406269994, "grad_norm": 0.8646253011023906, "learning_rate": 5.182935384992482e-06, "loss": 0.0856, "step": 10475 }, { "epoch": 6.702495201535509, "grad_norm": 0.8700579733139264, "learning_rate": 5.181119592266165e-06, "loss": 0.0865, "step": 10476 }, { "epoch": 6.703134996801023, "grad_norm": 0.8142578225169775, "learning_rate": 5.179304006453998e-06, "loss": 0.0845, "step": 10477 }, { "epoch": 6.703774792066539, "grad_norm": 0.9203676839839199, "learning_rate": 5.177488627633944e-06, "loss": 0.0811, "step": 10478 }, { "epoch": 6.704414587332054, "grad_norm": 0.9817778970455104, "learning_rate": 5.175673455883952e-06, "loss": 0.089, "step": 10479 }, { "epoch": 6.705054382597568, "grad_norm": 0.8310074986263887, "learning_rate": 5.173858491281958e-06, "loss": 0.0827, "step": 10480 }, { "epoch": 6.705694177863084, "grad_norm": 0.8445353341254332, "learning_rate": 5.172043733905896e-06, "loss": 0.0823, "step": 10481 }, { "epoch": 6.706333973128599, "grad_norm": 0.8415496470887402, "learning_rate": 5.1702291838336885e-06, "loss": 0.0831, "step": 10482 }, { "epoch": 6.706973768394114, "grad_norm": 0.8467371240185803, "learning_rate": 5.168414841143248e-06, "loss": 0.0801, "step": 10483 }, { "epoch": 6.707613563659629, "grad_norm": 0.7558743492950563, "learning_rate": 5.166600705912475e-06, "loss": 0.0798, "step": 10484 }, { "epoch": 6.708253358925144, "grad_norm": 0.9525399109497337, "learning_rate": 5.1647867782192725e-06, "loss": 0.089, "step": 10485 }, { "epoch": 6.708893154190659, "grad_norm": 0.7959264492785548, "learning_rate": 5.162973058141527e-06, "loss": 0.0838, "step": 10486 }, { "epoch": 6.709532949456174, "grad_norm": 0.8304425788688865, "learning_rate": 5.161159545757109e-06, "loss": 0.0896, "step": 10487 }, { "epoch": 6.710172744721689, "grad_norm": 0.7500071655296291, "learning_rate": 5.159346241143897e-06, "loss": 0.0793, "step": 10488 }, { "epoch": 6.710812539987204, "grad_norm": 0.9175128377850944, "learning_rate": 5.157533144379746e-06, "loss": 0.0869, "step": 10489 }, { "epoch": 6.711452335252719, "grad_norm": 0.9931181432807753, "learning_rate": 5.155720255542503e-06, "loss": 0.0901, "step": 10490 }, { "epoch": 6.712092130518235, "grad_norm": 0.8526849726272954, "learning_rate": 5.15390757471002e-06, "loss": 0.0813, "step": 10491 }, { "epoch": 6.712731925783749, "grad_norm": 0.8515642246928377, "learning_rate": 5.1520951019601255e-06, "loss": 0.0951, "step": 10492 }, { "epoch": 6.713371721049264, "grad_norm": 0.8841394336339117, "learning_rate": 5.150282837370643e-06, "loss": 0.0886, "step": 10493 }, { "epoch": 6.7140115163147795, "grad_norm": 0.850228651338396, "learning_rate": 5.148470781019388e-06, "loss": 0.0842, "step": 10494 }, { "epoch": 6.714651311580294, "grad_norm": 0.7987057592488501, "learning_rate": 5.1466589329841675e-06, "loss": 0.0768, "step": 10495 }, { "epoch": 6.715291106845809, "grad_norm": 0.9762595389330482, "learning_rate": 5.14484729334278e-06, "loss": 0.0917, "step": 10496 }, { "epoch": 6.7159309021113245, "grad_norm": 0.9494772066072091, "learning_rate": 5.143035862173008e-06, "loss": 0.0844, "step": 10497 }, { "epoch": 6.71657069737684, "grad_norm": 0.9911077736286769, "learning_rate": 5.141224639552641e-06, "loss": 0.0835, "step": 10498 }, { "epoch": 6.717210492642354, "grad_norm": 0.8393117118552521, "learning_rate": 5.139413625559444e-06, "loss": 0.085, "step": 10499 }, { "epoch": 6.717850287907869, "grad_norm": 0.8904338166602781, "learning_rate": 5.137602820271177e-06, "loss": 0.0827, "step": 10500 }, { "epoch": 6.718490083173385, "grad_norm": 0.7509476469766994, "learning_rate": 5.135792223765599e-06, "loss": 0.0798, "step": 10501 }, { "epoch": 6.719129878438899, "grad_norm": 0.8884364731504821, "learning_rate": 5.13398183612045e-06, "loss": 0.0813, "step": 10502 }, { "epoch": 6.719769673704414, "grad_norm": 0.9691246143734575, "learning_rate": 5.132171657413464e-06, "loss": 0.083, "step": 10503 }, { "epoch": 6.72040946896993, "grad_norm": 0.9360682626980643, "learning_rate": 5.130361687722365e-06, "loss": 0.0796, "step": 10504 }, { "epoch": 6.721049264235445, "grad_norm": 0.813235500066831, "learning_rate": 5.1285519271248765e-06, "loss": 0.083, "step": 10505 }, { "epoch": 6.72168905950096, "grad_norm": 0.8772266319612397, "learning_rate": 5.126742375698701e-06, "loss": 0.0829, "step": 10506 }, { "epoch": 6.722328854766475, "grad_norm": 0.829595947501363, "learning_rate": 5.124933033521537e-06, "loss": 0.0795, "step": 10507 }, { "epoch": 6.72296865003199, "grad_norm": 0.818879870141714, "learning_rate": 5.12312390067108e-06, "loss": 0.0838, "step": 10508 }, { "epoch": 6.723608445297505, "grad_norm": 0.8763326622233223, "learning_rate": 5.121314977225007e-06, "loss": 0.0779, "step": 10509 }, { "epoch": 6.7242482405630195, "grad_norm": 0.9599804586214484, "learning_rate": 5.1195062632609906e-06, "loss": 0.0906, "step": 10510 }, { "epoch": 6.724888035828535, "grad_norm": 0.9116790221373802, "learning_rate": 5.117697758856693e-06, "loss": 0.0917, "step": 10511 }, { "epoch": 6.72552783109405, "grad_norm": 0.8415514187910652, "learning_rate": 5.11588946408977e-06, "loss": 0.0823, "step": 10512 }, { "epoch": 6.726167626359565, "grad_norm": 0.8053033636920437, "learning_rate": 5.114081379037864e-06, "loss": 0.0794, "step": 10513 }, { "epoch": 6.72680742162508, "grad_norm": 0.8984119311925074, "learning_rate": 5.112273503778609e-06, "loss": 0.0844, "step": 10514 }, { "epoch": 6.727447216890595, "grad_norm": 0.7998463781587086, "learning_rate": 5.110465838389639e-06, "loss": 0.0823, "step": 10515 }, { "epoch": 6.72808701215611, "grad_norm": 0.8571609161495642, "learning_rate": 5.108658382948568e-06, "loss": 0.082, "step": 10516 }, { "epoch": 6.728726807421625, "grad_norm": 0.8029025189555486, "learning_rate": 5.106851137533001e-06, "loss": 0.0814, "step": 10517 }, { "epoch": 6.72936660268714, "grad_norm": 0.8211734015169179, "learning_rate": 5.105044102220546e-06, "loss": 0.0822, "step": 10518 }, { "epoch": 6.730006397952655, "grad_norm": 0.8591655734394492, "learning_rate": 5.1032372770887895e-06, "loss": 0.081, "step": 10519 }, { "epoch": 6.7306461932181705, "grad_norm": 0.8171114720606647, "learning_rate": 5.1014306622153095e-06, "loss": 0.0801, "step": 10520 }, { "epoch": 6.731285988483685, "grad_norm": 0.9298275928930166, "learning_rate": 5.099624257677688e-06, "loss": 0.0866, "step": 10521 }, { "epoch": 6.7319257837492, "grad_norm": 0.9367301032304646, "learning_rate": 5.097818063553482e-06, "loss": 0.0838, "step": 10522 }, { "epoch": 6.732565579014715, "grad_norm": 0.8636964886547227, "learning_rate": 5.096012079920249e-06, "loss": 0.0834, "step": 10523 }, { "epoch": 6.733205374280231, "grad_norm": 0.7299158516840569, "learning_rate": 5.094206306855535e-06, "loss": 0.0797, "step": 10524 }, { "epoch": 6.733845169545745, "grad_norm": 0.9303232790407713, "learning_rate": 5.092400744436874e-06, "loss": 0.088, "step": 10525 }, { "epoch": 6.73448496481126, "grad_norm": 0.7610798521236152, "learning_rate": 5.090595392741795e-06, "loss": 0.0772, "step": 10526 }, { "epoch": 6.735124760076776, "grad_norm": 0.824382735874173, "learning_rate": 5.088790251847814e-06, "loss": 0.0849, "step": 10527 }, { "epoch": 6.735764555342291, "grad_norm": 0.8501144645923507, "learning_rate": 5.086985321832446e-06, "loss": 0.083, "step": 10528 }, { "epoch": 6.736404350607805, "grad_norm": 0.7746101699022813, "learning_rate": 5.08518060277319e-06, "loss": 0.0859, "step": 10529 }, { "epoch": 6.737044145873321, "grad_norm": 0.9329083388685961, "learning_rate": 5.083376094747534e-06, "loss": 0.0882, "step": 10530 }, { "epoch": 6.737683941138836, "grad_norm": 0.91264138860147, "learning_rate": 5.08157179783296e-06, "loss": 0.0876, "step": 10531 }, { "epoch": 6.73832373640435, "grad_norm": 0.8520517355571932, "learning_rate": 5.079767712106948e-06, "loss": 0.0904, "step": 10532 }, { "epoch": 6.7389635316698655, "grad_norm": 0.8629781957832279, "learning_rate": 5.077963837646957e-06, "loss": 0.0878, "step": 10533 }, { "epoch": 6.739603326935381, "grad_norm": 0.8472047695742577, "learning_rate": 5.07616017453044e-06, "loss": 0.0829, "step": 10534 }, { "epoch": 6.740243122200896, "grad_norm": 0.8768678432448518, "learning_rate": 5.07435672283485e-06, "loss": 0.0818, "step": 10535 }, { "epoch": 6.74088291746641, "grad_norm": 0.9598418827885401, "learning_rate": 5.072553482637621e-06, "loss": 0.09, "step": 10536 }, { "epoch": 6.741522712731926, "grad_norm": 0.9038964532242845, "learning_rate": 5.070750454016176e-06, "loss": 0.0785, "step": 10537 }, { "epoch": 6.742162507997441, "grad_norm": 0.790426527796293, "learning_rate": 5.068947637047941e-06, "loss": 0.0751, "step": 10538 }, { "epoch": 6.742802303262955, "grad_norm": 0.9099540232326588, "learning_rate": 5.067145031810323e-06, "loss": 0.09, "step": 10539 }, { "epoch": 6.743442098528471, "grad_norm": 0.7939285657717685, "learning_rate": 5.065342638380722e-06, "loss": 0.082, "step": 10540 }, { "epoch": 6.744081893793986, "grad_norm": 0.9203037861479048, "learning_rate": 5.063540456836531e-06, "loss": 0.0858, "step": 10541 }, { "epoch": 6.744721689059501, "grad_norm": 0.8796609194747455, "learning_rate": 5.061738487255132e-06, "loss": 0.0819, "step": 10542 }, { "epoch": 6.7453614843250165, "grad_norm": 0.8310673711789752, "learning_rate": 5.0599367297138965e-06, "loss": 0.0893, "step": 10543 }, { "epoch": 6.746001279590531, "grad_norm": 0.88721601141171, "learning_rate": 5.058135184290187e-06, "loss": 0.0844, "step": 10544 }, { "epoch": 6.746641074856046, "grad_norm": 0.8711634020605826, "learning_rate": 5.056333851061366e-06, "loss": 0.0865, "step": 10545 }, { "epoch": 6.747280870121561, "grad_norm": 0.8285531879130448, "learning_rate": 5.054532730104775e-06, "loss": 0.0825, "step": 10546 }, { "epoch": 6.747920665387076, "grad_norm": 0.9280592502765853, "learning_rate": 5.052731821497747e-06, "loss": 0.0868, "step": 10547 }, { "epoch": 6.748560460652591, "grad_norm": 0.842224049107245, "learning_rate": 5.050931125317618e-06, "loss": 0.0839, "step": 10548 }, { "epoch": 6.749200255918106, "grad_norm": 0.83022009270344, "learning_rate": 5.049130641641704e-06, "loss": 0.0849, "step": 10549 }, { "epoch": 6.749840051183622, "grad_norm": 0.8450352742873192, "learning_rate": 5.047330370547312e-06, "loss": 0.0823, "step": 10550 }, { "epoch": 6.750479846449136, "grad_norm": 0.9130464706177214, "learning_rate": 5.045530312111741e-06, "loss": 0.0875, "step": 10551 }, { "epoch": 6.751119641714651, "grad_norm": 0.8295958201596025, "learning_rate": 5.043730466412288e-06, "loss": 0.079, "step": 10552 }, { "epoch": 6.751759436980167, "grad_norm": 0.8156086155903557, "learning_rate": 5.041930833526231e-06, "loss": 0.0835, "step": 10553 }, { "epoch": 6.752399232245681, "grad_norm": 0.8403299474111631, "learning_rate": 5.040131413530846e-06, "loss": 0.0862, "step": 10554 }, { "epoch": 6.753039027511196, "grad_norm": 0.8218814845651251, "learning_rate": 5.038332206503396e-06, "loss": 0.0788, "step": 10555 }, { "epoch": 6.7536788227767115, "grad_norm": 0.8795906449976747, "learning_rate": 5.036533212521134e-06, "loss": 0.0771, "step": 10556 }, { "epoch": 6.754318618042227, "grad_norm": 0.7386440534045645, "learning_rate": 5.0347344316613035e-06, "loss": 0.0795, "step": 10557 }, { "epoch": 6.754958413307741, "grad_norm": 0.8280346476558537, "learning_rate": 5.032935864001146e-06, "loss": 0.0807, "step": 10558 }, { "epoch": 6.755598208573256, "grad_norm": 0.7752625598430433, "learning_rate": 5.031137509617887e-06, "loss": 0.0825, "step": 10559 }, { "epoch": 6.756238003838772, "grad_norm": 0.9068282143598383, "learning_rate": 5.029339368588746e-06, "loss": 0.0896, "step": 10560 }, { "epoch": 6.756877799104287, "grad_norm": 0.8809979558850035, "learning_rate": 5.027541440990926e-06, "loss": 0.0843, "step": 10561 }, { "epoch": 6.757517594369801, "grad_norm": 0.8547756546193551, "learning_rate": 5.025743726901634e-06, "loss": 0.0815, "step": 10562 }, { "epoch": 6.758157389635317, "grad_norm": 0.8441855044986025, "learning_rate": 5.023946226398059e-06, "loss": 0.0801, "step": 10563 }, { "epoch": 6.758797184900832, "grad_norm": 0.8144187768385511, "learning_rate": 5.022148939557376e-06, "loss": 0.0792, "step": 10564 }, { "epoch": 6.759436980166347, "grad_norm": 0.8560754503718394, "learning_rate": 5.020351866456768e-06, "loss": 0.075, "step": 10565 }, { "epoch": 6.760076775431862, "grad_norm": 0.9019029277290563, "learning_rate": 5.018555007173392e-06, "loss": 0.0819, "step": 10566 }, { "epoch": 6.760716570697377, "grad_norm": 0.8272480861786969, "learning_rate": 5.0167583617844e-06, "loss": 0.0746, "step": 10567 }, { "epoch": 6.761356365962892, "grad_norm": 0.9382367759251098, "learning_rate": 5.0149619303669415e-06, "loss": 0.0809, "step": 10568 }, { "epoch": 6.7619961612284065, "grad_norm": 0.9377355426462911, "learning_rate": 5.013165712998151e-06, "loss": 0.0859, "step": 10569 }, { "epoch": 6.762635956493922, "grad_norm": 0.9063139707647954, "learning_rate": 5.011369709755154e-06, "loss": 0.0844, "step": 10570 }, { "epoch": 6.763275751759437, "grad_norm": 0.8274784035817581, "learning_rate": 5.0095739207150675e-06, "loss": 0.0796, "step": 10571 }, { "epoch": 6.763915547024952, "grad_norm": 0.7678053899812628, "learning_rate": 5.0077783459549986e-06, "loss": 0.08, "step": 10572 }, { "epoch": 6.764555342290467, "grad_norm": 0.8083310501644105, "learning_rate": 5.005982985552049e-06, "loss": 0.0819, "step": 10573 }, { "epoch": 6.765195137555982, "grad_norm": 0.8229447184157307, "learning_rate": 5.004187839583301e-06, "loss": 0.0833, "step": 10574 }, { "epoch": 6.765834932821497, "grad_norm": 0.7621126786496926, "learning_rate": 5.002392908125845e-06, "loss": 0.0798, "step": 10575 }, { "epoch": 6.766474728087012, "grad_norm": 0.8275587053382399, "learning_rate": 5.000598191256747e-06, "loss": 0.084, "step": 10576 }, { "epoch": 6.767114523352527, "grad_norm": 0.9037923109445961, "learning_rate": 4.998803689053065e-06, "loss": 0.0871, "step": 10577 }, { "epoch": 6.767754318618042, "grad_norm": 0.7457128650181607, "learning_rate": 4.997009401591861e-06, "loss": 0.0762, "step": 10578 }, { "epoch": 6.7683941138835575, "grad_norm": 0.8583269134504813, "learning_rate": 4.995215328950173e-06, "loss": 0.0836, "step": 10579 }, { "epoch": 6.769033909149073, "grad_norm": 0.9027123583016814, "learning_rate": 4.993421471205037e-06, "loss": 0.0852, "step": 10580 }, { "epoch": 6.769673704414587, "grad_norm": 0.9186619523151524, "learning_rate": 4.991627828433472e-06, "loss": 0.0893, "step": 10581 }, { "epoch": 6.770313499680102, "grad_norm": 0.8604703421763642, "learning_rate": 4.989834400712503e-06, "loss": 0.0831, "step": 10582 }, { "epoch": 6.770953294945618, "grad_norm": 0.9828004277608351, "learning_rate": 4.988041188119133e-06, "loss": 0.0845, "step": 10583 }, { "epoch": 6.771593090211132, "grad_norm": 0.9475017185405313, "learning_rate": 4.986248190730357e-06, "loss": 0.0837, "step": 10584 }, { "epoch": 6.772232885476647, "grad_norm": 0.8196563275621783, "learning_rate": 4.984455408623167e-06, "loss": 0.0802, "step": 10585 }, { "epoch": 6.772872680742163, "grad_norm": 0.8608680245913396, "learning_rate": 4.9826628418745394e-06, "loss": 0.0832, "step": 10586 }, { "epoch": 6.773512476007678, "grad_norm": 0.7351083426062077, "learning_rate": 4.980870490561439e-06, "loss": 0.0746, "step": 10587 }, { "epoch": 6.774152271273192, "grad_norm": 0.9329195957280958, "learning_rate": 4.979078354760837e-06, "loss": 0.0866, "step": 10588 }, { "epoch": 6.774792066538708, "grad_norm": 0.9159648218764767, "learning_rate": 4.977286434549678e-06, "loss": 0.0825, "step": 10589 }, { "epoch": 6.775431861804223, "grad_norm": 0.846599680386207, "learning_rate": 4.975494730004905e-06, "loss": 0.0888, "step": 10590 }, { "epoch": 6.776071657069737, "grad_norm": 1.1158478295207812, "learning_rate": 4.973703241203447e-06, "loss": 0.0817, "step": 10591 }, { "epoch": 6.7767114523352525, "grad_norm": 0.9197783547737918, "learning_rate": 4.9719119682222335e-06, "loss": 0.0841, "step": 10592 }, { "epoch": 6.777351247600768, "grad_norm": 0.8607413415623167, "learning_rate": 4.970120911138177e-06, "loss": 0.0834, "step": 10593 }, { "epoch": 6.777991042866283, "grad_norm": 0.814928337788704, "learning_rate": 4.968330070028175e-06, "loss": 0.0784, "step": 10594 }, { "epoch": 6.7786308381317975, "grad_norm": 0.7986616909671397, "learning_rate": 4.966539444969136e-06, "loss": 0.0861, "step": 10595 }, { "epoch": 6.779270633397313, "grad_norm": 0.9130239093987117, "learning_rate": 4.964749036037938e-06, "loss": 0.0816, "step": 10596 }, { "epoch": 6.779910428662828, "grad_norm": 0.8422940889013936, "learning_rate": 4.962958843311455e-06, "loss": 0.0799, "step": 10597 }, { "epoch": 6.780550223928343, "grad_norm": 0.932974589655048, "learning_rate": 4.961168866866563e-06, "loss": 0.0845, "step": 10598 }, { "epoch": 6.781190019193858, "grad_norm": 0.8964754058828196, "learning_rate": 4.959379106780118e-06, "loss": 0.0851, "step": 10599 }, { "epoch": 6.781829814459373, "grad_norm": 0.7780922420060438, "learning_rate": 4.957589563128966e-06, "loss": 0.0758, "step": 10600 }, { "epoch": 6.782469609724888, "grad_norm": 0.8388771156877074, "learning_rate": 4.95580023598995e-06, "loss": 0.0824, "step": 10601 }, { "epoch": 6.7831094049904035, "grad_norm": 0.793726004356133, "learning_rate": 4.954011125439897e-06, "loss": 0.0828, "step": 10602 }, { "epoch": 6.783749200255918, "grad_norm": 0.8879022224769781, "learning_rate": 4.952222231555631e-06, "loss": 0.0822, "step": 10603 }, { "epoch": 6.784388995521433, "grad_norm": 0.8179293329713365, "learning_rate": 4.950433554413959e-06, "loss": 0.0791, "step": 10604 }, { "epoch": 6.785028790786948, "grad_norm": 1.0362908245183862, "learning_rate": 4.948645094091692e-06, "loss": 0.0754, "step": 10605 }, { "epoch": 6.785668586052463, "grad_norm": 0.804238983708767, "learning_rate": 4.946856850665619e-06, "loss": 0.0805, "step": 10606 }, { "epoch": 6.786308381317978, "grad_norm": 0.8783467677470095, "learning_rate": 4.945068824212523e-06, "loss": 0.085, "step": 10607 }, { "epoch": 6.786948176583493, "grad_norm": 0.80562254973468, "learning_rate": 4.943281014809177e-06, "loss": 0.0835, "step": 10608 }, { "epoch": 6.787587971849009, "grad_norm": 0.8759592809022758, "learning_rate": 4.941493422532353e-06, "loss": 0.0837, "step": 10609 }, { "epoch": 6.788227767114523, "grad_norm": 0.8234312909263966, "learning_rate": 4.939706047458802e-06, "loss": 0.09, "step": 10610 }, { "epoch": 6.788867562380038, "grad_norm": 0.8626817828897593, "learning_rate": 4.937918889665268e-06, "loss": 0.0851, "step": 10611 }, { "epoch": 6.789507357645554, "grad_norm": 0.7958677556855285, "learning_rate": 4.936131949228498e-06, "loss": 0.0747, "step": 10612 }, { "epoch": 6.790147152911068, "grad_norm": 0.8858798686685565, "learning_rate": 4.934345226225212e-06, "loss": 0.0793, "step": 10613 }, { "epoch": 6.790786948176583, "grad_norm": 0.8297064044542998, "learning_rate": 4.932558720732131e-06, "loss": 0.0862, "step": 10614 }, { "epoch": 6.7914267434420985, "grad_norm": 0.7875621163160011, "learning_rate": 4.930772432825966e-06, "loss": 0.0811, "step": 10615 }, { "epoch": 6.792066538707614, "grad_norm": 0.8004778374285584, "learning_rate": 4.9289863625834134e-06, "loss": 0.0788, "step": 10616 }, { "epoch": 6.792706333973129, "grad_norm": 0.7567373666297972, "learning_rate": 4.927200510081166e-06, "loss": 0.0754, "step": 10617 }, { "epoch": 6.7933461292386434, "grad_norm": 0.785321578350478, "learning_rate": 4.925414875395901e-06, "loss": 0.0769, "step": 10618 }, { "epoch": 6.793985924504159, "grad_norm": 0.8690664916026181, "learning_rate": 4.923629458604298e-06, "loss": 0.0807, "step": 10619 }, { "epoch": 6.794625719769674, "grad_norm": 0.7744347891737189, "learning_rate": 4.921844259783017e-06, "loss": 0.0742, "step": 10620 }, { "epoch": 6.795265515035188, "grad_norm": 0.8559422625097977, "learning_rate": 4.920059279008706e-06, "loss": 0.0912, "step": 10621 }, { "epoch": 6.795905310300704, "grad_norm": 0.9295942922927786, "learning_rate": 4.918274516358017e-06, "loss": 0.0838, "step": 10622 }, { "epoch": 6.796545105566219, "grad_norm": 0.8150480130452256, "learning_rate": 4.91648997190758e-06, "loss": 0.0766, "step": 10623 }, { "epoch": 6.797184900831734, "grad_norm": 0.8907602943746611, "learning_rate": 4.914705645734017e-06, "loss": 0.0895, "step": 10624 }, { "epoch": 6.797824696097249, "grad_norm": 0.8054918713432037, "learning_rate": 4.912921537913952e-06, "loss": 0.0779, "step": 10625 }, { "epoch": 6.798464491362764, "grad_norm": 0.9049981975191397, "learning_rate": 4.911137648523986e-06, "loss": 0.0813, "step": 10626 }, { "epoch": 6.799104286628279, "grad_norm": 0.8239472448925972, "learning_rate": 4.909353977640717e-06, "loss": 0.0756, "step": 10627 }, { "epoch": 6.7997440818937935, "grad_norm": 0.8143519639064901, "learning_rate": 4.907570525340729e-06, "loss": 0.0837, "step": 10628 }, { "epoch": 6.800383877159309, "grad_norm": 0.8560956337235898, "learning_rate": 4.9057872917006075e-06, "loss": 0.0826, "step": 10629 }, { "epoch": 6.801023672424824, "grad_norm": 0.7934116220778432, "learning_rate": 4.904004276796919e-06, "loss": 0.0804, "step": 10630 }, { "epoch": 6.801663467690339, "grad_norm": 0.8179302623738757, "learning_rate": 4.902221480706221e-06, "loss": 0.0788, "step": 10631 }, { "epoch": 6.802303262955854, "grad_norm": 1.0340088414608357, "learning_rate": 4.900438903505064e-06, "loss": 0.0865, "step": 10632 }, { "epoch": 6.802943058221369, "grad_norm": 0.8536879977967192, "learning_rate": 4.89865654526999e-06, "loss": 0.0839, "step": 10633 }, { "epoch": 6.803582853486884, "grad_norm": 0.9487206088062811, "learning_rate": 4.896874406077524e-06, "loss": 0.0869, "step": 10634 }, { "epoch": 6.8042226487524, "grad_norm": 0.9514115503861317, "learning_rate": 4.895092486004198e-06, "loss": 0.0852, "step": 10635 }, { "epoch": 6.804862444017914, "grad_norm": 0.9889447076797332, "learning_rate": 4.89331078512652e-06, "loss": 0.0866, "step": 10636 }, { "epoch": 6.805502239283429, "grad_norm": 0.8099921771371339, "learning_rate": 4.891529303520993e-06, "loss": 0.082, "step": 10637 }, { "epoch": 6.8061420345489445, "grad_norm": 0.8669922164479492, "learning_rate": 4.889748041264107e-06, "loss": 0.0795, "step": 10638 }, { "epoch": 6.80678182981446, "grad_norm": 0.9176996617094316, "learning_rate": 4.887966998432352e-06, "loss": 0.0832, "step": 10639 }, { "epoch": 6.807421625079974, "grad_norm": 0.8019858977246602, "learning_rate": 4.886186175102201e-06, "loss": 0.0801, "step": 10640 }, { "epoch": 6.8080614203454894, "grad_norm": 0.8089072333005647, "learning_rate": 4.8844055713501145e-06, "loss": 0.0826, "step": 10641 }, { "epoch": 6.808701215611005, "grad_norm": 0.7656013935535891, "learning_rate": 4.882625187252559e-06, "loss": 0.076, "step": 10642 }, { "epoch": 6.809341010876519, "grad_norm": 0.7721777423685439, "learning_rate": 4.880845022885973e-06, "loss": 0.0842, "step": 10643 }, { "epoch": 6.809980806142034, "grad_norm": 0.8487939592927908, "learning_rate": 4.879065078326796e-06, "loss": 0.0823, "step": 10644 }, { "epoch": 6.81062060140755, "grad_norm": 0.8686022502517479, "learning_rate": 4.877285353651454e-06, "loss": 0.0925, "step": 10645 }, { "epoch": 6.811260396673065, "grad_norm": 0.8813657799609235, "learning_rate": 4.875505848936367e-06, "loss": 0.0781, "step": 10646 }, { "epoch": 6.811900191938579, "grad_norm": 0.8238448169457135, "learning_rate": 4.873726564257943e-06, "loss": 0.0823, "step": 10647 }, { "epoch": 6.812539987204095, "grad_norm": 0.9104220491256025, "learning_rate": 4.871947499692578e-06, "loss": 0.085, "step": 10648 }, { "epoch": 6.81317978246961, "grad_norm": 0.8505966920433188, "learning_rate": 4.870168655316668e-06, "loss": 0.0792, "step": 10649 }, { "epoch": 6.813819577735125, "grad_norm": 0.8350359082500847, "learning_rate": 4.868390031206592e-06, "loss": 0.0796, "step": 10650 }, { "epoch": 6.8144593730006395, "grad_norm": 0.9883264855942796, "learning_rate": 4.866611627438716e-06, "loss": 0.094, "step": 10651 }, { "epoch": 6.815099168266155, "grad_norm": 0.8029657162585322, "learning_rate": 4.864833444089407e-06, "loss": 0.0846, "step": 10652 }, { "epoch": 6.81573896353167, "grad_norm": 0.854735801484443, "learning_rate": 4.863055481235018e-06, "loss": 0.0893, "step": 10653 }, { "epoch": 6.816378758797185, "grad_norm": 0.8282235895263536, "learning_rate": 4.861277738951883e-06, "loss": 0.0802, "step": 10654 }, { "epoch": 6.8170185540627, "grad_norm": 0.9426991693582879, "learning_rate": 4.859500217316344e-06, "loss": 0.0807, "step": 10655 }, { "epoch": 6.817658349328215, "grad_norm": 0.8277036640801656, "learning_rate": 4.857722916404723e-06, "loss": 0.0823, "step": 10656 }, { "epoch": 6.81829814459373, "grad_norm": 0.8850629787775681, "learning_rate": 4.855945836293332e-06, "loss": 0.0869, "step": 10657 }, { "epoch": 6.818937939859245, "grad_norm": 0.9137595963456411, "learning_rate": 4.854168977058472e-06, "loss": 0.0842, "step": 10658 }, { "epoch": 6.81957773512476, "grad_norm": 0.8252240331563998, "learning_rate": 4.8523923387764474e-06, "loss": 0.081, "step": 10659 }, { "epoch": 6.820217530390275, "grad_norm": 0.7921565866973564, "learning_rate": 4.850615921523537e-06, "loss": 0.0802, "step": 10660 }, { "epoch": 6.8208573256557905, "grad_norm": 0.7966302100873366, "learning_rate": 4.848839725376021e-06, "loss": 0.075, "step": 10661 }, { "epoch": 6.821497120921305, "grad_norm": 0.8427462247388695, "learning_rate": 4.847063750410163e-06, "loss": 0.0878, "step": 10662 }, { "epoch": 6.82213691618682, "grad_norm": 0.8900730295325054, "learning_rate": 4.84528799670222e-06, "loss": 0.0829, "step": 10663 }, { "epoch": 6.822776711452335, "grad_norm": 0.8174179771049808, "learning_rate": 4.843512464328439e-06, "loss": 0.0757, "step": 10664 }, { "epoch": 6.82341650671785, "grad_norm": 0.9071393892683167, "learning_rate": 4.841737153365063e-06, "loss": 0.0836, "step": 10665 }, { "epoch": 6.824056301983365, "grad_norm": 0.8161338651544282, "learning_rate": 4.8399620638883186e-06, "loss": 0.0826, "step": 10666 }, { "epoch": 6.82469609724888, "grad_norm": 0.7636461784047183, "learning_rate": 4.838187195974423e-06, "loss": 0.0799, "step": 10667 }, { "epoch": 6.825335892514396, "grad_norm": 0.8458508580029375, "learning_rate": 4.836412549699584e-06, "loss": 0.0767, "step": 10668 }, { "epoch": 6.825975687779911, "grad_norm": 0.8048704577106296, "learning_rate": 4.834638125140007e-06, "loss": 0.0834, "step": 10669 }, { "epoch": 6.826615483045425, "grad_norm": 0.8978144433596549, "learning_rate": 4.832863922371883e-06, "loss": 0.0812, "step": 10670 }, { "epoch": 6.827255278310941, "grad_norm": 0.9434594143826598, "learning_rate": 4.831089941471385e-06, "loss": 0.0864, "step": 10671 }, { "epoch": 6.827895073576456, "grad_norm": 0.9856945493444863, "learning_rate": 4.829316182514695e-06, "loss": 0.0898, "step": 10672 }, { "epoch": 6.82853486884197, "grad_norm": 0.8759969482938256, "learning_rate": 4.827542645577969e-06, "loss": 0.0841, "step": 10673 }, { "epoch": 6.8291746641074855, "grad_norm": 0.9283440347335384, "learning_rate": 4.8257693307373625e-06, "loss": 0.0826, "step": 10674 }, { "epoch": 6.829814459373001, "grad_norm": 0.8915663210013806, "learning_rate": 4.823996238069015e-06, "loss": 0.0769, "step": 10675 }, { "epoch": 6.830454254638516, "grad_norm": 0.9082348365777564, "learning_rate": 4.822223367649063e-06, "loss": 0.0835, "step": 10676 }, { "epoch": 6.8310940499040305, "grad_norm": 0.8715802863175, "learning_rate": 4.820450719553628e-06, "loss": 0.0851, "step": 10677 }, { "epoch": 6.831733845169546, "grad_norm": 0.8204224197050956, "learning_rate": 4.818678293858822e-06, "loss": 0.0853, "step": 10678 }, { "epoch": 6.832373640435061, "grad_norm": 0.894228590477956, "learning_rate": 4.816906090640756e-06, "loss": 0.0888, "step": 10679 }, { "epoch": 6.833013435700575, "grad_norm": 0.8774635361512141, "learning_rate": 4.815134109975524e-06, "loss": 0.0801, "step": 10680 }, { "epoch": 6.833653230966091, "grad_norm": 0.8775484662293217, "learning_rate": 4.813362351939206e-06, "loss": 0.0827, "step": 10681 }, { "epoch": 6.834293026231606, "grad_norm": 0.8728027606430574, "learning_rate": 4.8115908166078854e-06, "loss": 0.0834, "step": 10682 }, { "epoch": 6.834932821497121, "grad_norm": 0.9871786348579689, "learning_rate": 4.809819504057626e-06, "loss": 0.0878, "step": 10683 }, { "epoch": 6.835572616762636, "grad_norm": 0.8713096191449415, "learning_rate": 4.808048414364484e-06, "loss": 0.0896, "step": 10684 }, { "epoch": 6.836212412028151, "grad_norm": 0.9226936473163575, "learning_rate": 4.8062775476045046e-06, "loss": 0.0817, "step": 10685 }, { "epoch": 6.836852207293666, "grad_norm": 0.7748772492495789, "learning_rate": 4.804506903853732e-06, "loss": 0.0778, "step": 10686 }, { "epoch": 6.837492002559181, "grad_norm": 0.8142997203768322, "learning_rate": 4.8027364831881905e-06, "loss": 0.0772, "step": 10687 }, { "epoch": 6.838131797824696, "grad_norm": 0.7627795669254696, "learning_rate": 4.800966285683896e-06, "loss": 0.0835, "step": 10688 }, { "epoch": 6.838771593090211, "grad_norm": 0.7552895082747144, "learning_rate": 4.7991963114168635e-06, "loss": 0.0801, "step": 10689 }, { "epoch": 6.839411388355726, "grad_norm": 0.826458280673026, "learning_rate": 4.79742656046309e-06, "loss": 0.0892, "step": 10690 }, { "epoch": 6.840051183621242, "grad_norm": 0.9084901451484669, "learning_rate": 4.795657032898566e-06, "loss": 0.0876, "step": 10691 }, { "epoch": 6.840690978886756, "grad_norm": 0.7561877605150019, "learning_rate": 4.793887728799272e-06, "loss": 0.0744, "step": 10692 }, { "epoch": 6.841330774152271, "grad_norm": 0.8173114469802698, "learning_rate": 4.792118648241177e-06, "loss": 0.0821, "step": 10693 }, { "epoch": 6.841970569417787, "grad_norm": 0.8609940580943688, "learning_rate": 4.790349791300241e-06, "loss": 0.0788, "step": 10694 }, { "epoch": 6.842610364683301, "grad_norm": 0.9005117839434709, "learning_rate": 4.788581158052418e-06, "loss": 0.0966, "step": 10695 }, { "epoch": 6.843250159948816, "grad_norm": 0.8548596551799572, "learning_rate": 4.78681274857365e-06, "loss": 0.0803, "step": 10696 }, { "epoch": 6.8438899552143315, "grad_norm": 0.8307711644457163, "learning_rate": 4.7850445629398716e-06, "loss": 0.0805, "step": 10697 }, { "epoch": 6.844529750479847, "grad_norm": 0.8949733272979108, "learning_rate": 4.783276601226999e-06, "loss": 0.0882, "step": 10698 }, { "epoch": 6.845169545745361, "grad_norm": 0.8413588417513799, "learning_rate": 4.781508863510951e-06, "loss": 0.0817, "step": 10699 }, { "epoch": 6.8458093410108765, "grad_norm": 0.7663696941613816, "learning_rate": 4.77974134986763e-06, "loss": 0.0767, "step": 10700 }, { "epoch": 6.846449136276392, "grad_norm": 0.9780604970897964, "learning_rate": 4.777974060372926e-06, "loss": 0.086, "step": 10701 }, { "epoch": 6.847088931541906, "grad_norm": 0.8686297837533576, "learning_rate": 4.776206995102731e-06, "loss": 0.0796, "step": 10702 }, { "epoch": 6.847728726807421, "grad_norm": 0.7427448865424401, "learning_rate": 4.774440154132916e-06, "loss": 0.0798, "step": 10703 }, { "epoch": 6.848368522072937, "grad_norm": 0.8862766377781915, "learning_rate": 4.772673537539343e-06, "loss": 0.0833, "step": 10704 }, { "epoch": 6.849008317338452, "grad_norm": 0.8328875999905938, "learning_rate": 4.77090714539787e-06, "loss": 0.0814, "step": 10705 }, { "epoch": 6.849648112603967, "grad_norm": 0.9347956587553242, "learning_rate": 4.769140977784344e-06, "loss": 0.0831, "step": 10706 }, { "epoch": 6.850287907869482, "grad_norm": 0.9748226854342543, "learning_rate": 4.767375034774598e-06, "loss": 0.089, "step": 10707 }, { "epoch": 6.850927703134997, "grad_norm": 0.9210515736700001, "learning_rate": 4.765609316444456e-06, "loss": 0.0926, "step": 10708 }, { "epoch": 6.851567498400512, "grad_norm": 0.8792556221113259, "learning_rate": 4.7638438228697436e-06, "loss": 0.0786, "step": 10709 }, { "epoch": 6.8522072936660265, "grad_norm": 0.7805932850247591, "learning_rate": 4.7620785541262626e-06, "loss": 0.0813, "step": 10710 }, { "epoch": 6.852847088931542, "grad_norm": 0.9905126561244938, "learning_rate": 4.7603135102898065e-06, "loss": 0.0934, "step": 10711 }, { "epoch": 6.853486884197057, "grad_norm": 0.9764498499206844, "learning_rate": 4.758548691436172e-06, "loss": 0.0873, "step": 10712 }, { "epoch": 6.854126679462572, "grad_norm": 0.8691099211013349, "learning_rate": 4.756784097641133e-06, "loss": 0.0789, "step": 10713 }, { "epoch": 6.854766474728087, "grad_norm": 0.8991993298170193, "learning_rate": 4.755019728980458e-06, "loss": 0.0852, "step": 10714 }, { "epoch": 6.855406269993602, "grad_norm": 0.8286827145599027, "learning_rate": 4.753255585529901e-06, "loss": 0.0768, "step": 10715 }, { "epoch": 6.856046065259117, "grad_norm": 0.7217689085465506, "learning_rate": 4.75149166736522e-06, "loss": 0.0838, "step": 10716 }, { "epoch": 6.856685860524632, "grad_norm": 0.9121454414144945, "learning_rate": 4.749727974562151e-06, "loss": 0.0849, "step": 10717 }, { "epoch": 6.857325655790147, "grad_norm": 0.8972994984139752, "learning_rate": 4.747964507196419e-06, "loss": 0.0816, "step": 10718 }, { "epoch": 6.857965451055662, "grad_norm": 0.852442634124988, "learning_rate": 4.7462012653437525e-06, "loss": 0.087, "step": 10719 }, { "epoch": 6.8586052463211775, "grad_norm": 0.8007166894221465, "learning_rate": 4.744438249079858e-06, "loss": 0.0867, "step": 10720 }, { "epoch": 6.859245041586692, "grad_norm": 0.842736014814521, "learning_rate": 4.7426754584804354e-06, "loss": 0.0783, "step": 10721 }, { "epoch": 6.859884836852207, "grad_norm": 0.8440035463990412, "learning_rate": 4.740912893621178e-06, "loss": 0.0871, "step": 10722 }, { "epoch": 6.8605246321177225, "grad_norm": 0.9814512385189932, "learning_rate": 4.739150554577764e-06, "loss": 0.0797, "step": 10723 }, { "epoch": 6.861164427383238, "grad_norm": 0.8599634592212864, "learning_rate": 4.7373884414258675e-06, "loss": 0.0783, "step": 10724 }, { "epoch": 6.861804222648752, "grad_norm": 0.8288858070951156, "learning_rate": 4.735626554241147e-06, "loss": 0.0789, "step": 10725 }, { "epoch": 6.862444017914267, "grad_norm": 0.8619318192421408, "learning_rate": 4.7338648930992605e-06, "loss": 0.0833, "step": 10726 }, { "epoch": 6.863083813179783, "grad_norm": 0.8384072900493987, "learning_rate": 4.732103458075849e-06, "loss": 0.0795, "step": 10727 }, { "epoch": 6.863723608445298, "grad_norm": 0.9114192663628847, "learning_rate": 4.730342249246538e-06, "loss": 0.0936, "step": 10728 }, { "epoch": 6.864363403710812, "grad_norm": 0.9370374128375047, "learning_rate": 4.728581266686963e-06, "loss": 0.0808, "step": 10729 }, { "epoch": 6.865003198976328, "grad_norm": 0.8352435455831096, "learning_rate": 4.726820510472729e-06, "loss": 0.0796, "step": 10730 }, { "epoch": 6.865642994241843, "grad_norm": 0.8699960568189579, "learning_rate": 4.72505998067944e-06, "loss": 0.0875, "step": 10731 }, { "epoch": 6.866282789507357, "grad_norm": 0.8254937601831126, "learning_rate": 4.723299677382696e-06, "loss": 0.0859, "step": 10732 }, { "epoch": 6.8669225847728725, "grad_norm": 0.9106162947315936, "learning_rate": 4.721539600658076e-06, "loss": 0.0826, "step": 10733 }, { "epoch": 6.867562380038388, "grad_norm": 0.8810915767291391, "learning_rate": 4.719779750581156e-06, "loss": 0.0855, "step": 10734 }, { "epoch": 6.868202175303903, "grad_norm": 0.913217984923482, "learning_rate": 4.7180201272275014e-06, "loss": 0.0875, "step": 10735 }, { "epoch": 6.8688419705694175, "grad_norm": 0.751436307352104, "learning_rate": 4.716260730672667e-06, "loss": 0.0814, "step": 10736 }, { "epoch": 6.869481765834933, "grad_norm": 0.909662697685939, "learning_rate": 4.7145015609921975e-06, "loss": 0.0856, "step": 10737 }, { "epoch": 6.870121561100448, "grad_norm": 0.9488938797985031, "learning_rate": 4.712742618261624e-06, "loss": 0.0822, "step": 10738 }, { "epoch": 6.870761356365962, "grad_norm": 0.9252817928144816, "learning_rate": 4.710983902556482e-06, "loss": 0.0899, "step": 10739 }, { "epoch": 6.871401151631478, "grad_norm": 0.8063607889407141, "learning_rate": 4.709225413952283e-06, "loss": 0.0839, "step": 10740 }, { "epoch": 6.872040946896993, "grad_norm": 0.9093980560253345, "learning_rate": 4.707467152524528e-06, "loss": 0.0832, "step": 10741 }, { "epoch": 6.872680742162508, "grad_norm": 0.8209652962992374, "learning_rate": 4.705709118348725e-06, "loss": 0.0864, "step": 10742 }, { "epoch": 6.8733205374280235, "grad_norm": 0.9457969136753118, "learning_rate": 4.703951311500352e-06, "loss": 0.0872, "step": 10743 }, { "epoch": 6.873960332693538, "grad_norm": 0.8650449855880761, "learning_rate": 4.7021937320548905e-06, "loss": 0.0889, "step": 10744 }, { "epoch": 6.874600127959053, "grad_norm": 0.8295044336877254, "learning_rate": 4.700436380087802e-06, "loss": 0.082, "step": 10745 }, { "epoch": 6.8752399232245685, "grad_norm": 0.8468127710065293, "learning_rate": 4.698679255674551e-06, "loss": 0.0849, "step": 10746 }, { "epoch": 6.875879718490083, "grad_norm": 0.8212391155137309, "learning_rate": 4.696922358890583e-06, "loss": 0.0835, "step": 10747 }, { "epoch": 6.876519513755598, "grad_norm": 0.7927896590996083, "learning_rate": 4.695165689811332e-06, "loss": 0.0826, "step": 10748 }, { "epoch": 6.877159309021113, "grad_norm": 0.7613339194688051, "learning_rate": 4.6934092485122325e-06, "loss": 0.0803, "step": 10749 }, { "epoch": 6.877799104286629, "grad_norm": 0.9891390584922691, "learning_rate": 4.691653035068705e-06, "loss": 0.0887, "step": 10750 }, { "epoch": 6.878438899552143, "grad_norm": 0.8246706091570599, "learning_rate": 4.689897049556146e-06, "loss": 0.0815, "step": 10751 }, { "epoch": 6.879078694817658, "grad_norm": 0.9809719778644924, "learning_rate": 4.688141292049965e-06, "loss": 0.0791, "step": 10752 }, { "epoch": 6.879718490083174, "grad_norm": 0.8292969126082316, "learning_rate": 4.686385762625547e-06, "loss": 0.0795, "step": 10753 }, { "epoch": 6.880358285348688, "grad_norm": 0.9054896449152018, "learning_rate": 4.684630461358274e-06, "loss": 0.0856, "step": 10754 }, { "epoch": 6.880998080614203, "grad_norm": 0.8772376691687715, "learning_rate": 4.682875388323509e-06, "loss": 0.0788, "step": 10755 }, { "epoch": 6.8816378758797185, "grad_norm": 0.8414885466091094, "learning_rate": 4.68112054359662e-06, "loss": 0.0822, "step": 10756 }, { "epoch": 6.882277671145234, "grad_norm": 0.879594144996653, "learning_rate": 4.679365927252953e-06, "loss": 0.0868, "step": 10757 }, { "epoch": 6.882917466410748, "grad_norm": 0.8786181976724451, "learning_rate": 4.677611539367846e-06, "loss": 0.082, "step": 10758 }, { "epoch": 6.8835572616762635, "grad_norm": 0.8436277885169815, "learning_rate": 4.6758573800166344e-06, "loss": 0.0879, "step": 10759 }, { "epoch": 6.884197056941779, "grad_norm": 0.8960806898615699, "learning_rate": 4.674103449274635e-06, "loss": 0.0877, "step": 10760 }, { "epoch": 6.884836852207294, "grad_norm": 0.8690546311627928, "learning_rate": 4.672349747217161e-06, "loss": 0.079, "step": 10761 }, { "epoch": 6.885476647472808, "grad_norm": 0.7874138240369507, "learning_rate": 4.670596273919508e-06, "loss": 0.0792, "step": 10762 }, { "epoch": 6.886116442738324, "grad_norm": 0.8393734835811942, "learning_rate": 4.668843029456973e-06, "loss": 0.0878, "step": 10763 }, { "epoch": 6.886756238003839, "grad_norm": 0.9676770328379689, "learning_rate": 4.667090013904836e-06, "loss": 0.0854, "step": 10764 }, { "epoch": 6.887396033269354, "grad_norm": 0.9116880546057566, "learning_rate": 4.665337227338368e-06, "loss": 0.0848, "step": 10765 }, { "epoch": 6.888035828534869, "grad_norm": 0.7749443162983487, "learning_rate": 4.663584669832828e-06, "loss": 0.0747, "step": 10766 }, { "epoch": 6.888675623800384, "grad_norm": 0.746621992378353, "learning_rate": 4.661832341463471e-06, "loss": 0.0766, "step": 10767 }, { "epoch": 6.889315419065899, "grad_norm": 0.8414340347775794, "learning_rate": 4.660080242305534e-06, "loss": 0.0857, "step": 10768 }, { "epoch": 6.889955214331414, "grad_norm": 0.8508528737736483, "learning_rate": 4.658328372434254e-06, "loss": 0.0819, "step": 10769 }, { "epoch": 6.890595009596929, "grad_norm": 0.9124621225494896, "learning_rate": 4.656576731924853e-06, "loss": 0.0824, "step": 10770 }, { "epoch": 6.891234804862444, "grad_norm": 0.8448192173287263, "learning_rate": 4.654825320852543e-06, "loss": 0.0857, "step": 10771 }, { "epoch": 6.891874600127959, "grad_norm": 0.8693438380908035, "learning_rate": 4.65307413929252e-06, "loss": 0.0842, "step": 10772 }, { "epoch": 6.892514395393474, "grad_norm": 0.9090229304183624, "learning_rate": 4.651323187319986e-06, "loss": 0.0877, "step": 10773 }, { "epoch": 6.893154190658989, "grad_norm": 0.8028174684067928, "learning_rate": 4.64957246501012e-06, "loss": 0.0785, "step": 10774 }, { "epoch": 6.893793985924504, "grad_norm": 0.8553380534903509, "learning_rate": 4.647821972438092e-06, "loss": 0.0799, "step": 10775 }, { "epoch": 6.894433781190019, "grad_norm": 1.0054904120088743, "learning_rate": 4.64607170967907e-06, "loss": 0.0923, "step": 10776 }, { "epoch": 6.895073576455534, "grad_norm": 0.8184996149214141, "learning_rate": 4.644321676808206e-06, "loss": 0.0848, "step": 10777 }, { "epoch": 6.895713371721049, "grad_norm": 0.8513620649781926, "learning_rate": 4.642571873900637e-06, "loss": 0.09, "step": 10778 }, { "epoch": 6.8963531669865645, "grad_norm": 0.8520579233904613, "learning_rate": 4.6408223010315105e-06, "loss": 0.0797, "step": 10779 }, { "epoch": 6.89699296225208, "grad_norm": 0.8578945672345426, "learning_rate": 4.639072958275937e-06, "loss": 0.0828, "step": 10780 }, { "epoch": 6.897632757517594, "grad_norm": 0.8285633956672256, "learning_rate": 4.637323845709033e-06, "loss": 0.0829, "step": 10781 }, { "epoch": 6.8982725527831095, "grad_norm": 0.8740283032184519, "learning_rate": 4.635574963405901e-06, "loss": 0.0872, "step": 10782 }, { "epoch": 6.898912348048625, "grad_norm": 0.7969684463243041, "learning_rate": 4.633826311441641e-06, "loss": 0.0791, "step": 10783 }, { "epoch": 6.899552143314139, "grad_norm": 0.8710967574158013, "learning_rate": 4.632077889891334e-06, "loss": 0.0808, "step": 10784 }, { "epoch": 6.900191938579654, "grad_norm": 0.8298110138655405, "learning_rate": 4.630329698830049e-06, "loss": 0.0803, "step": 10785 }, { "epoch": 6.90083173384517, "grad_norm": 0.8486351553636302, "learning_rate": 4.62858173833286e-06, "loss": 0.0742, "step": 10786 }, { "epoch": 6.901471529110685, "grad_norm": 0.7988501549476783, "learning_rate": 4.626834008474815e-06, "loss": 0.0858, "step": 10787 }, { "epoch": 6.902111324376199, "grad_norm": 0.9167326057281117, "learning_rate": 4.625086509330956e-06, "loss": 0.0849, "step": 10788 }, { "epoch": 6.902751119641715, "grad_norm": 1.0714840937603363, "learning_rate": 4.623339240976325e-06, "loss": 0.0937, "step": 10789 }, { "epoch": 6.90339091490723, "grad_norm": 0.7561421210333928, "learning_rate": 4.6215922034859425e-06, "loss": 0.0756, "step": 10790 }, { "epoch": 6.904030710172744, "grad_norm": 0.9241934367930388, "learning_rate": 4.619845396934823e-06, "loss": 0.0803, "step": 10791 }, { "epoch": 6.9046705054382596, "grad_norm": 0.8327133828875746, "learning_rate": 4.61809882139797e-06, "loss": 0.0829, "step": 10792 }, { "epoch": 6.905310300703775, "grad_norm": 0.7502023441435811, "learning_rate": 4.616352476950381e-06, "loss": 0.0775, "step": 10793 }, { "epoch": 6.90595009596929, "grad_norm": 0.7727689722061843, "learning_rate": 4.614606363667041e-06, "loss": 0.0807, "step": 10794 }, { "epoch": 6.9065898912348045, "grad_norm": 0.8703646971348606, "learning_rate": 4.612860481622924e-06, "loss": 0.0822, "step": 10795 }, { "epoch": 6.90722968650032, "grad_norm": 0.7445443716314357, "learning_rate": 4.611114830892994e-06, "loss": 0.085, "step": 10796 }, { "epoch": 6.907869481765835, "grad_norm": 0.858173894471432, "learning_rate": 4.609369411552208e-06, "loss": 0.089, "step": 10797 }, { "epoch": 6.90850927703135, "grad_norm": 0.7926603794346367, "learning_rate": 4.607624223675506e-06, "loss": 0.0812, "step": 10798 }, { "epoch": 6.909149072296865, "grad_norm": 0.7620636970472531, "learning_rate": 4.605879267337832e-06, "loss": 0.0764, "step": 10799 }, { "epoch": 6.90978886756238, "grad_norm": 0.8631368305300179, "learning_rate": 4.604134542614106e-06, "loss": 0.0762, "step": 10800 }, { "epoch": 6.910428662827895, "grad_norm": 0.7746894176739849, "learning_rate": 4.602390049579245e-06, "loss": 0.0819, "step": 10801 }, { "epoch": 6.9110684580934105, "grad_norm": 0.8200826471009864, "learning_rate": 4.60064578830815e-06, "loss": 0.0815, "step": 10802 }, { "epoch": 6.911708253358925, "grad_norm": 0.9430655968709482, "learning_rate": 4.598901758875724e-06, "loss": 0.083, "step": 10803 }, { "epoch": 6.91234804862444, "grad_norm": 0.9452621595494916, "learning_rate": 4.5971579613568474e-06, "loss": 0.073, "step": 10804 }, { "epoch": 6.9129878438899555, "grad_norm": 0.8899463948504563, "learning_rate": 4.595414395826394e-06, "loss": 0.0837, "step": 10805 }, { "epoch": 6.91362763915547, "grad_norm": 0.8507261303156269, "learning_rate": 4.593671062359236e-06, "loss": 0.0844, "step": 10806 }, { "epoch": 6.914267434420985, "grad_norm": 0.8797329396305068, "learning_rate": 4.591927961030226e-06, "loss": 0.0785, "step": 10807 }, { "epoch": 6.9149072296865, "grad_norm": 0.88975457742653, "learning_rate": 4.5901850919142085e-06, "loss": 0.0794, "step": 10808 }, { "epoch": 6.915547024952016, "grad_norm": 0.8337798792905609, "learning_rate": 4.588442455086021e-06, "loss": 0.0771, "step": 10809 }, { "epoch": 6.91618682021753, "grad_norm": 0.9404136237468862, "learning_rate": 4.586700050620486e-06, "loss": 0.0812, "step": 10810 }, { "epoch": 6.916826615483045, "grad_norm": 0.8194592423714374, "learning_rate": 4.584957878592423e-06, "loss": 0.0807, "step": 10811 }, { "epoch": 6.917466410748561, "grad_norm": 0.8849098424133414, "learning_rate": 4.583215939076632e-06, "loss": 0.0772, "step": 10812 }, { "epoch": 6.918106206014075, "grad_norm": 0.8318822471788166, "learning_rate": 4.581474232147916e-06, "loss": 0.0845, "step": 10813 }, { "epoch": 6.91874600127959, "grad_norm": 0.8491106533547554, "learning_rate": 4.579732757881059e-06, "loss": 0.0792, "step": 10814 }, { "epoch": 6.9193857965451055, "grad_norm": 0.8555329588116112, "learning_rate": 4.57799151635083e-06, "loss": 0.0895, "step": 10815 }, { "epoch": 6.920025591810621, "grad_norm": 0.8281755325001332, "learning_rate": 4.576250507632005e-06, "loss": 0.0843, "step": 10816 }, { "epoch": 6.920665387076136, "grad_norm": 0.7907892967515769, "learning_rate": 4.574509731799335e-06, "loss": 0.0831, "step": 10817 }, { "epoch": 6.9213051823416505, "grad_norm": 0.812175491431704, "learning_rate": 4.572769188927563e-06, "loss": 0.079, "step": 10818 }, { "epoch": 6.921944977607166, "grad_norm": 0.8980606730444904, "learning_rate": 4.571028879091431e-06, "loss": 0.0844, "step": 10819 }, { "epoch": 6.922584772872681, "grad_norm": 0.7706065933887223, "learning_rate": 4.56928880236566e-06, "loss": 0.0759, "step": 10820 }, { "epoch": 6.923224568138195, "grad_norm": 0.820077106833053, "learning_rate": 4.567548958824969e-06, "loss": 0.0857, "step": 10821 }, { "epoch": 6.923864363403711, "grad_norm": 0.8119799731645895, "learning_rate": 4.565809348544058e-06, "loss": 0.0798, "step": 10822 }, { "epoch": 6.924504158669226, "grad_norm": 0.8356497543522953, "learning_rate": 4.56406997159763e-06, "loss": 0.0874, "step": 10823 }, { "epoch": 6.925143953934741, "grad_norm": 0.8223944493671485, "learning_rate": 4.562330828060369e-06, "loss": 0.0852, "step": 10824 }, { "epoch": 6.925783749200256, "grad_norm": 0.8641143547371766, "learning_rate": 4.560591918006948e-06, "loss": 0.0804, "step": 10825 }, { "epoch": 6.926423544465771, "grad_norm": 0.8652757148263719, "learning_rate": 4.558853241512035e-06, "loss": 0.0842, "step": 10826 }, { "epoch": 6.927063339731286, "grad_norm": 0.783636832483074, "learning_rate": 4.557114798650284e-06, "loss": 0.0775, "step": 10827 }, { "epoch": 6.927703134996801, "grad_norm": 0.8991393649065275, "learning_rate": 4.5553765894963386e-06, "loss": 0.0793, "step": 10828 }, { "epoch": 6.928342930262316, "grad_norm": 0.8278034560803843, "learning_rate": 4.55363861412484e-06, "loss": 0.0808, "step": 10829 }, { "epoch": 6.928982725527831, "grad_norm": 0.8259469096910909, "learning_rate": 4.551900872610412e-06, "loss": 0.0867, "step": 10830 }, { "epoch": 6.929622520793346, "grad_norm": 0.8855671432675764, "learning_rate": 4.5501633650276675e-06, "loss": 0.088, "step": 10831 }, { "epoch": 6.930262316058861, "grad_norm": 0.7708312621112724, "learning_rate": 4.548426091451211e-06, "loss": 0.0742, "step": 10832 }, { "epoch": 6.930902111324376, "grad_norm": 0.8178563792994059, "learning_rate": 4.546689051955645e-06, "loss": 0.0797, "step": 10833 }, { "epoch": 6.931541906589891, "grad_norm": 0.852792335071319, "learning_rate": 4.544952246615549e-06, "loss": 0.0857, "step": 10834 }, { "epoch": 6.932181701855407, "grad_norm": 0.8687420026586814, "learning_rate": 4.543215675505497e-06, "loss": 0.0795, "step": 10835 }, { "epoch": 6.932821497120921, "grad_norm": 0.8851140189906146, "learning_rate": 4.54147933870006e-06, "loss": 0.0821, "step": 10836 }, { "epoch": 6.933461292386436, "grad_norm": 0.9838394906584822, "learning_rate": 4.53974323627379e-06, "loss": 0.0843, "step": 10837 }, { "epoch": 6.9341010876519515, "grad_norm": 0.9549657972771084, "learning_rate": 4.538007368301234e-06, "loss": 0.0879, "step": 10838 }, { "epoch": 6.934740882917467, "grad_norm": 0.8775447222728846, "learning_rate": 4.536271734856924e-06, "loss": 0.0801, "step": 10839 }, { "epoch": 6.935380678182981, "grad_norm": 0.9111629429984927, "learning_rate": 4.534536336015387e-06, "loss": 0.0827, "step": 10840 }, { "epoch": 6.9360204734484965, "grad_norm": 0.8974090430648524, "learning_rate": 4.532801171851138e-06, "loss": 0.0818, "step": 10841 }, { "epoch": 6.936660268714012, "grad_norm": 0.8468116947531642, "learning_rate": 4.5310662424386765e-06, "loss": 0.0819, "step": 10842 }, { "epoch": 6.937300063979526, "grad_norm": 0.9004108467207346, "learning_rate": 4.529331547852507e-06, "loss": 0.0819, "step": 10843 }, { "epoch": 6.937939859245041, "grad_norm": 0.8868949923908798, "learning_rate": 4.527597088167109e-06, "loss": 0.0834, "step": 10844 }, { "epoch": 6.938579654510557, "grad_norm": 0.8346675620643361, "learning_rate": 4.525862863456952e-06, "loss": 0.0838, "step": 10845 }, { "epoch": 6.939219449776072, "grad_norm": 0.8668655657521883, "learning_rate": 4.5241288737965104e-06, "loss": 0.08, "step": 10846 }, { "epoch": 6.939859245041586, "grad_norm": 0.8141244861146004, "learning_rate": 4.5223951192602355e-06, "loss": 0.0807, "step": 10847 }, { "epoch": 6.940499040307102, "grad_norm": 0.9144851593139969, "learning_rate": 4.520661599922569e-06, "loss": 0.0762, "step": 10848 }, { "epoch": 6.941138835572617, "grad_norm": 0.9371693989481522, "learning_rate": 4.518928315857943e-06, "loss": 0.0898, "step": 10849 }, { "epoch": 6.941778630838132, "grad_norm": 0.7971695176083684, "learning_rate": 4.517195267140788e-06, "loss": 0.0849, "step": 10850 }, { "epoch": 6.942418426103647, "grad_norm": 0.9371254011730101, "learning_rate": 4.515462453845516e-06, "loss": 0.087, "step": 10851 }, { "epoch": 6.943058221369162, "grad_norm": 0.8019796335251461, "learning_rate": 4.513729876046527e-06, "loss": 0.0818, "step": 10852 }, { "epoch": 6.943698016634677, "grad_norm": 0.8354142091815291, "learning_rate": 4.511997533818221e-06, "loss": 0.0765, "step": 10853 }, { "epoch": 6.944337811900192, "grad_norm": 0.8360836633852794, "learning_rate": 4.510265427234978e-06, "loss": 0.0769, "step": 10854 }, { "epoch": 6.944977607165707, "grad_norm": 0.8452816912566037, "learning_rate": 4.508533556371172e-06, "loss": 0.0809, "step": 10855 }, { "epoch": 6.945617402431222, "grad_norm": 0.8376483976505349, "learning_rate": 4.506801921301166e-06, "loss": 0.0798, "step": 10856 }, { "epoch": 6.946257197696737, "grad_norm": 0.8974447263744108, "learning_rate": 4.505070522099315e-06, "loss": 0.0831, "step": 10857 }, { "epoch": 6.946896992962252, "grad_norm": 0.8211856871755956, "learning_rate": 4.503339358839961e-06, "loss": 0.0741, "step": 10858 }, { "epoch": 6.947536788227767, "grad_norm": 0.8678361512383008, "learning_rate": 4.501608431597433e-06, "loss": 0.0849, "step": 10859 }, { "epoch": 6.948176583493282, "grad_norm": 0.7860765741042206, "learning_rate": 4.499877740446061e-06, "loss": 0.0725, "step": 10860 }, { "epoch": 6.9488163787587975, "grad_norm": 0.8595951894327294, "learning_rate": 4.498147285460157e-06, "loss": 0.0841, "step": 10861 }, { "epoch": 6.949456174024312, "grad_norm": 0.7834658708308196, "learning_rate": 4.496417066714016e-06, "loss": 0.075, "step": 10862 }, { "epoch": 6.950095969289827, "grad_norm": 0.9266322647922168, "learning_rate": 4.4946870842819405e-06, "loss": 0.0863, "step": 10863 }, { "epoch": 6.9507357645553425, "grad_norm": 0.8765907138962565, "learning_rate": 4.492957338238209e-06, "loss": 0.088, "step": 10864 }, { "epoch": 6.951375559820857, "grad_norm": 0.7922200645245216, "learning_rate": 4.4912278286570875e-06, "loss": 0.0826, "step": 10865 }, { "epoch": 6.952015355086372, "grad_norm": 0.8966566036087831, "learning_rate": 4.4894985556128486e-06, "loss": 0.0913, "step": 10866 }, { "epoch": 6.952655150351887, "grad_norm": 0.8469348349398732, "learning_rate": 4.48776951917974e-06, "loss": 0.0775, "step": 10867 }, { "epoch": 6.953294945617403, "grad_norm": 0.9189224122910179, "learning_rate": 4.486040719432001e-06, "loss": 0.0821, "step": 10868 }, { "epoch": 6.953934740882918, "grad_norm": 0.8882804350539529, "learning_rate": 4.4843121564438665e-06, "loss": 0.084, "step": 10869 }, { "epoch": 6.954574536148432, "grad_norm": 0.8451236906987106, "learning_rate": 4.482583830289555e-06, "loss": 0.0789, "step": 10870 }, { "epoch": 6.955214331413948, "grad_norm": 0.8933416950810029, "learning_rate": 4.480855741043281e-06, "loss": 0.0872, "step": 10871 }, { "epoch": 6.955854126679463, "grad_norm": 0.8573362858028125, "learning_rate": 4.4791278887792385e-06, "loss": 0.085, "step": 10872 }, { "epoch": 6.956493921944977, "grad_norm": 0.8240929895827903, "learning_rate": 4.477400273571627e-06, "loss": 0.0788, "step": 10873 }, { "epoch": 6.957133717210493, "grad_norm": 0.8760670158162617, "learning_rate": 4.475672895494625e-06, "loss": 0.089, "step": 10874 }, { "epoch": 6.957773512476008, "grad_norm": 0.8587460252576621, "learning_rate": 4.4739457546223975e-06, "loss": 0.0804, "step": 10875 }, { "epoch": 6.958413307741523, "grad_norm": 0.964262910029167, "learning_rate": 4.472218851029113e-06, "loss": 0.0816, "step": 10876 }, { "epoch": 6.9590531030070375, "grad_norm": 0.9300720864125146, "learning_rate": 4.470492184788918e-06, "loss": 0.0869, "step": 10877 }, { "epoch": 6.959692898272553, "grad_norm": 0.8999594250547744, "learning_rate": 4.468765755975952e-06, "loss": 0.0861, "step": 10878 }, { "epoch": 6.960332693538068, "grad_norm": 0.7496193147048172, "learning_rate": 4.467039564664342e-06, "loss": 0.0824, "step": 10879 }, { "epoch": 6.960972488803582, "grad_norm": 0.7782918850510165, "learning_rate": 4.4653136109282146e-06, "loss": 0.0803, "step": 10880 }, { "epoch": 6.961612284069098, "grad_norm": 0.8179946534410915, "learning_rate": 4.463587894841674e-06, "loss": 0.0831, "step": 10881 }, { "epoch": 6.962252079334613, "grad_norm": 0.8786574278200322, "learning_rate": 4.461862416478818e-06, "loss": 0.0828, "step": 10882 }, { "epoch": 6.962891874600128, "grad_norm": 0.9108289939520957, "learning_rate": 4.4601371759137424e-06, "loss": 0.0832, "step": 10883 }, { "epoch": 6.963531669865643, "grad_norm": 0.8911541334320318, "learning_rate": 4.458412173220521e-06, "loss": 0.0869, "step": 10884 }, { "epoch": 6.964171465131158, "grad_norm": 0.9270628121760605, "learning_rate": 4.456687408473222e-06, "loss": 0.0867, "step": 10885 }, { "epoch": 6.964811260396673, "grad_norm": 0.8154305393954183, "learning_rate": 4.4549628817459066e-06, "loss": 0.0864, "step": 10886 }, { "epoch": 6.9654510556621885, "grad_norm": 0.7901829622091955, "learning_rate": 4.453238593112621e-06, "loss": 0.075, "step": 10887 }, { "epoch": 6.966090850927703, "grad_norm": 0.8403917681418874, "learning_rate": 4.4515145426474035e-06, "loss": 0.0804, "step": 10888 }, { "epoch": 6.966730646193218, "grad_norm": 0.8485379469320308, "learning_rate": 4.449790730424277e-06, "loss": 0.082, "step": 10889 }, { "epoch": 6.967370441458733, "grad_norm": 0.9619929185846613, "learning_rate": 4.448067156517268e-06, "loss": 0.0776, "step": 10890 }, { "epoch": 6.968010236724249, "grad_norm": 0.8460354368679173, "learning_rate": 4.446343821000378e-06, "loss": 0.0762, "step": 10891 }, { "epoch": 6.968650031989763, "grad_norm": 0.7883053025589651, "learning_rate": 4.444620723947603e-06, "loss": 0.08, "step": 10892 }, { "epoch": 6.969289827255278, "grad_norm": 0.8560449126367142, "learning_rate": 4.442897865432934e-06, "loss": 0.0825, "step": 10893 }, { "epoch": 6.969929622520794, "grad_norm": 0.838716870385782, "learning_rate": 4.441175245530347e-06, "loss": 0.09, "step": 10894 }, { "epoch": 6.970569417786308, "grad_norm": 0.7796012987596505, "learning_rate": 4.4394528643138025e-06, "loss": 0.079, "step": 10895 }, { "epoch": 6.971209213051823, "grad_norm": 0.8475279291099955, "learning_rate": 4.437730721857263e-06, "loss": 0.0794, "step": 10896 }, { "epoch": 6.971849008317339, "grad_norm": 0.8161503745894538, "learning_rate": 4.436008818234673e-06, "loss": 0.082, "step": 10897 }, { "epoch": 6.972488803582854, "grad_norm": 0.7852372269665142, "learning_rate": 4.434287153519966e-06, "loss": 0.0842, "step": 10898 }, { "epoch": 6.973128598848368, "grad_norm": 0.8847937528806098, "learning_rate": 4.432565727787069e-06, "loss": 0.0769, "step": 10899 }, { "epoch": 6.9737683941138835, "grad_norm": 0.8969230519486826, "learning_rate": 4.4308445411098954e-06, "loss": 0.0865, "step": 10900 }, { "epoch": 6.974408189379399, "grad_norm": 0.936804995617835, "learning_rate": 4.42912359356235e-06, "loss": 0.0858, "step": 10901 }, { "epoch": 6.975047984644913, "grad_norm": 0.895430058621789, "learning_rate": 4.427402885218325e-06, "loss": 0.0809, "step": 10902 }, { "epoch": 6.975687779910428, "grad_norm": 0.8199683679804519, "learning_rate": 4.42568241615171e-06, "loss": 0.0856, "step": 10903 }, { "epoch": 6.976327575175944, "grad_norm": 0.8725036534257634, "learning_rate": 4.4239621864363755e-06, "loss": 0.0853, "step": 10904 }, { "epoch": 6.976967370441459, "grad_norm": 0.8478789073081245, "learning_rate": 4.4222421961461834e-06, "loss": 0.0896, "step": 10905 }, { "epoch": 6.977607165706974, "grad_norm": 0.7991057912035903, "learning_rate": 4.420522445354992e-06, "loss": 0.0812, "step": 10906 }, { "epoch": 6.978246960972489, "grad_norm": 0.8753400067308337, "learning_rate": 4.418802934136642e-06, "loss": 0.091, "step": 10907 }, { "epoch": 6.978886756238004, "grad_norm": 0.8507021682023866, "learning_rate": 4.417083662564966e-06, "loss": 0.0864, "step": 10908 }, { "epoch": 6.979526551503519, "grad_norm": 0.7693805820285284, "learning_rate": 4.4153646307137836e-06, "loss": 0.0787, "step": 10909 }, { "epoch": 6.980166346769034, "grad_norm": 0.8243588957088528, "learning_rate": 4.413645838656911e-06, "loss": 0.0813, "step": 10910 }, { "epoch": 6.980806142034549, "grad_norm": 0.8890507716417114, "learning_rate": 4.411927286468151e-06, "loss": 0.0797, "step": 10911 }, { "epoch": 6.981445937300064, "grad_norm": 0.8378391753132544, "learning_rate": 4.4102089742212885e-06, "loss": 0.0735, "step": 10912 }, { "epoch": 6.982085732565579, "grad_norm": 0.9348411514134299, "learning_rate": 4.408490901990112e-06, "loss": 0.0852, "step": 10913 }, { "epoch": 6.982725527831094, "grad_norm": 0.9294246352384001, "learning_rate": 4.406773069848391e-06, "loss": 0.0854, "step": 10914 }, { "epoch": 6.983365323096609, "grad_norm": 0.9480431210874034, "learning_rate": 4.405055477869884e-06, "loss": 0.0859, "step": 10915 }, { "epoch": 6.984005118362124, "grad_norm": 0.7394189980438096, "learning_rate": 4.403338126128343e-06, "loss": 0.074, "step": 10916 }, { "epoch": 6.984644913627639, "grad_norm": 0.8735218669206415, "learning_rate": 4.401621014697506e-06, "loss": 0.0817, "step": 10917 }, { "epoch": 6.985284708893154, "grad_norm": 0.8904097617963093, "learning_rate": 4.3999041436511035e-06, "loss": 0.0861, "step": 10918 }, { "epoch": 6.985924504158669, "grad_norm": 0.8951583220097624, "learning_rate": 4.398187513062853e-06, "loss": 0.0833, "step": 10919 }, { "epoch": 6.9865642994241846, "grad_norm": 0.799082154539611, "learning_rate": 4.396471123006469e-06, "loss": 0.0816, "step": 10920 }, { "epoch": 6.987204094689699, "grad_norm": 0.8651534446840738, "learning_rate": 4.3947549735556476e-06, "loss": 0.0783, "step": 10921 }, { "epoch": 6.987843889955214, "grad_norm": 0.9062003103865782, "learning_rate": 4.393039064784072e-06, "loss": 0.0837, "step": 10922 }, { "epoch": 6.9884836852207295, "grad_norm": 0.9481703368118458, "learning_rate": 4.391323396765429e-06, "loss": 0.0879, "step": 10923 }, { "epoch": 6.989123480486245, "grad_norm": 0.8827449514169794, "learning_rate": 4.389607969573382e-06, "loss": 0.0816, "step": 10924 }, { "epoch": 6.989763275751759, "grad_norm": 0.8650625935193677, "learning_rate": 4.387892783281589e-06, "loss": 0.0848, "step": 10925 }, { "epoch": 6.990403071017274, "grad_norm": 0.8291489928015342, "learning_rate": 4.3861778379636935e-06, "loss": 0.0809, "step": 10926 }, { "epoch": 6.99104286628279, "grad_norm": 0.8539493618044827, "learning_rate": 4.384463133693339e-06, "loss": 0.0844, "step": 10927 }, { "epoch": 6.991682661548305, "grad_norm": 0.834928660672041, "learning_rate": 4.382748670544147e-06, "loss": 0.0814, "step": 10928 }, { "epoch": 6.992322456813819, "grad_norm": 0.8640623818706793, "learning_rate": 4.381034448589736e-06, "loss": 0.0826, "step": 10929 }, { "epoch": 6.992962252079335, "grad_norm": 0.8738552649711252, "learning_rate": 4.379320467903711e-06, "loss": 0.0789, "step": 10930 }, { "epoch": 6.99360204734485, "grad_norm": 0.8481635345698625, "learning_rate": 4.377606728559664e-06, "loss": 0.0779, "step": 10931 }, { "epoch": 6.994241842610364, "grad_norm": 0.7938733261069719, "learning_rate": 4.375893230631182e-06, "loss": 0.0817, "step": 10932 }, { "epoch": 6.99488163787588, "grad_norm": 0.8655930884429494, "learning_rate": 4.374179974191842e-06, "loss": 0.0816, "step": 10933 }, { "epoch": 6.995521433141395, "grad_norm": 0.8821027986927161, "learning_rate": 4.3724669593152065e-06, "loss": 0.0932, "step": 10934 }, { "epoch": 6.99616122840691, "grad_norm": 0.9348265521269367, "learning_rate": 4.3707541860748305e-06, "loss": 0.0819, "step": 10935 }, { "epoch": 6.9968010236724245, "grad_norm": 0.8991103909731736, "learning_rate": 4.369041654544251e-06, "loss": 0.0883, "step": 10936 }, { "epoch": 6.99744081893794, "grad_norm": 0.9810238927239161, "learning_rate": 4.3673293647970105e-06, "loss": 0.0843, "step": 10937 }, { "epoch": 6.998080614203455, "grad_norm": 0.7848605751170699, "learning_rate": 4.3656173169066275e-06, "loss": 0.0804, "step": 10938 }, { "epoch": 6.9987204094689695, "grad_norm": 0.8281947690315653, "learning_rate": 4.36390551094661e-06, "loss": 0.0872, "step": 10939 }, { "epoch": 6.999360204734485, "grad_norm": 0.853706254042209, "learning_rate": 4.362193946990467e-06, "loss": 0.0879, "step": 10940 }, { "epoch": 7.0, "grad_norm": 0.7912683720860817, "learning_rate": 4.360482625111688e-06, "loss": 0.0771, "step": 10941 }, { "epoch": 7.000639795265515, "grad_norm": 0.6705321769429151, "learning_rate": 4.35877154538375e-06, "loss": 0.0629, "step": 10942 }, { "epoch": 7.00127959053103, "grad_norm": 0.5915472879943466, "learning_rate": 4.35706070788013e-06, "loss": 0.0523, "step": 10943 }, { "epoch": 7.001919385796545, "grad_norm": 0.6464349171847746, "learning_rate": 4.355350112674285e-06, "loss": 0.0626, "step": 10944 }, { "epoch": 7.00255918106206, "grad_norm": 0.5965906888722265, "learning_rate": 4.3536397598396655e-06, "loss": 0.0574, "step": 10945 }, { "epoch": 7.0031989763275755, "grad_norm": 0.600928678473886, "learning_rate": 4.35192964944971e-06, "loss": 0.0587, "step": 10946 }, { "epoch": 7.00383877159309, "grad_norm": 0.5995691198486968, "learning_rate": 4.350219781577848e-06, "loss": 0.0558, "step": 10947 }, { "epoch": 7.004478566858605, "grad_norm": 0.6335921102213913, "learning_rate": 4.3485101562975e-06, "loss": 0.0551, "step": 10948 }, { "epoch": 7.00511836212412, "grad_norm": 0.6365776516915936, "learning_rate": 4.346800773682067e-06, "loss": 0.056, "step": 10949 }, { "epoch": 7.005758157389636, "grad_norm": 0.7940483994072415, "learning_rate": 4.345091633804957e-06, "loss": 0.0627, "step": 10950 }, { "epoch": 7.00639795265515, "grad_norm": 0.7107088147948915, "learning_rate": 4.343382736739553e-06, "loss": 0.0527, "step": 10951 }, { "epoch": 7.007037747920665, "grad_norm": 0.7624252320897981, "learning_rate": 4.341674082559228e-06, "loss": 0.0517, "step": 10952 }, { "epoch": 7.007677543186181, "grad_norm": 0.8716835877995578, "learning_rate": 4.339965671337358e-06, "loss": 0.0639, "step": 10953 }, { "epoch": 7.008317338451695, "grad_norm": 0.7626179889978135, "learning_rate": 4.338257503147291e-06, "loss": 0.0506, "step": 10954 }, { "epoch": 7.00895713371721, "grad_norm": 0.8448211852262075, "learning_rate": 4.336549578062377e-06, "loss": 0.0534, "step": 10955 }, { "epoch": 7.009596928982726, "grad_norm": 0.7542606506854787, "learning_rate": 4.334841896155945e-06, "loss": 0.0499, "step": 10956 }, { "epoch": 7.010236724248241, "grad_norm": 0.7204442527626778, "learning_rate": 4.333134457501329e-06, "loss": 0.0558, "step": 10957 }, { "epoch": 7.010876519513755, "grad_norm": 0.8239745106281221, "learning_rate": 4.331427262171838e-06, "loss": 0.0535, "step": 10958 }, { "epoch": 7.0115163147792705, "grad_norm": 0.813903116487517, "learning_rate": 4.329720310240777e-06, "loss": 0.057, "step": 10959 }, { "epoch": 7.012156110044786, "grad_norm": 0.8062743840164645, "learning_rate": 4.32801360178144e-06, "loss": 0.0527, "step": 10960 }, { "epoch": 7.012795905310301, "grad_norm": 0.7822916376210552, "learning_rate": 4.3263071368671085e-06, "loss": 0.0543, "step": 10961 }, { "epoch": 7.0134357005758154, "grad_norm": 0.7711005813480329, "learning_rate": 4.324600915571052e-06, "loss": 0.0541, "step": 10962 }, { "epoch": 7.014075495841331, "grad_norm": 0.812555122015391, "learning_rate": 4.322894937966541e-06, "loss": 0.0562, "step": 10963 }, { "epoch": 7.014715291106846, "grad_norm": 0.7602061577279404, "learning_rate": 4.321189204126821e-06, "loss": 0.0527, "step": 10964 }, { "epoch": 7.015355086372361, "grad_norm": 0.7645685825907532, "learning_rate": 4.319483714125136e-06, "loss": 0.0593, "step": 10965 }, { "epoch": 7.015994881637876, "grad_norm": 0.7115474009474263, "learning_rate": 4.317778468034711e-06, "loss": 0.0473, "step": 10966 }, { "epoch": 7.016634676903391, "grad_norm": 0.6856517989275497, "learning_rate": 4.3160734659287765e-06, "loss": 0.0569, "step": 10967 }, { "epoch": 7.017274472168906, "grad_norm": 0.8091791511843168, "learning_rate": 4.314368707880534e-06, "loss": 0.0564, "step": 10968 }, { "epoch": 7.017914267434421, "grad_norm": 0.754039327992148, "learning_rate": 4.312664193963182e-06, "loss": 0.0528, "step": 10969 }, { "epoch": 7.018554062699936, "grad_norm": 0.7713162051381149, "learning_rate": 4.310959924249917e-06, "loss": 0.0555, "step": 10970 }, { "epoch": 7.019193857965451, "grad_norm": 0.8290401194113307, "learning_rate": 4.3092558988139126e-06, "loss": 0.053, "step": 10971 }, { "epoch": 7.019833653230966, "grad_norm": 0.7887900077899146, "learning_rate": 4.307552117728333e-06, "loss": 0.0566, "step": 10972 }, { "epoch": 7.020473448496481, "grad_norm": 0.7340973270823523, "learning_rate": 4.305848581066343e-06, "loss": 0.0496, "step": 10973 }, { "epoch": 7.021113243761996, "grad_norm": 0.8069247904069273, "learning_rate": 4.304145288901086e-06, "loss": 0.0536, "step": 10974 }, { "epoch": 7.021753039027511, "grad_norm": 0.8758449415241063, "learning_rate": 4.3024422413056975e-06, "loss": 0.0535, "step": 10975 }, { "epoch": 7.022392834293027, "grad_norm": 0.9023289861464521, "learning_rate": 4.300739438353305e-06, "loss": 0.053, "step": 10976 }, { "epoch": 7.023032629558541, "grad_norm": 0.7449673224732385, "learning_rate": 4.2990368801170225e-06, "loss": 0.0573, "step": 10977 }, { "epoch": 7.023672424824056, "grad_norm": 0.7620304259597758, "learning_rate": 4.297334566669955e-06, "loss": 0.05, "step": 10978 }, { "epoch": 7.024312220089572, "grad_norm": 0.8425636688276987, "learning_rate": 4.295632498085193e-06, "loss": 0.0526, "step": 10979 }, { "epoch": 7.024952015355086, "grad_norm": 0.7232070332262232, "learning_rate": 4.293930674435828e-06, "loss": 0.051, "step": 10980 }, { "epoch": 7.025591810620601, "grad_norm": 0.7321662037391294, "learning_rate": 4.29222909579493e-06, "loss": 0.053, "step": 10981 }, { "epoch": 7.0262316058861165, "grad_norm": 0.8584824374647136, "learning_rate": 4.290527762235558e-06, "loss": 0.0634, "step": 10982 }, { "epoch": 7.026871401151632, "grad_norm": 0.7511688299682787, "learning_rate": 4.288826673830771e-06, "loss": 0.0528, "step": 10983 }, { "epoch": 7.027511196417146, "grad_norm": 0.8989796863232168, "learning_rate": 4.287125830653605e-06, "loss": 0.0586, "step": 10984 }, { "epoch": 7.028150991682661, "grad_norm": 0.6952886996597455, "learning_rate": 4.2854252327770965e-06, "loss": 0.0476, "step": 10985 }, { "epoch": 7.028790786948177, "grad_norm": 0.7659853413580314, "learning_rate": 4.283724880274257e-06, "loss": 0.0565, "step": 10986 }, { "epoch": 7.029430582213692, "grad_norm": 0.8022231229272837, "learning_rate": 4.282024773218107e-06, "loss": 0.0532, "step": 10987 }, { "epoch": 7.030070377479206, "grad_norm": 0.7726704001588496, "learning_rate": 4.280324911681644e-06, "loss": 0.0519, "step": 10988 }, { "epoch": 7.030710172744722, "grad_norm": 0.7607853145659157, "learning_rate": 4.278625295737853e-06, "loss": 0.0509, "step": 10989 }, { "epoch": 7.031349968010237, "grad_norm": 0.7539359128140832, "learning_rate": 4.276925925459713e-06, "loss": 0.0507, "step": 10990 }, { "epoch": 7.031989763275751, "grad_norm": 0.8546204888348677, "learning_rate": 4.2752268009201945e-06, "loss": 0.0549, "step": 10991 }, { "epoch": 7.032629558541267, "grad_norm": 0.8755009952683527, "learning_rate": 4.2735279221922495e-06, "loss": 0.0529, "step": 10992 }, { "epoch": 7.033269353806782, "grad_norm": 0.8498855251692217, "learning_rate": 4.271829289348834e-06, "loss": 0.0535, "step": 10993 }, { "epoch": 7.033909149072297, "grad_norm": 0.7827376212505083, "learning_rate": 4.270130902462878e-06, "loss": 0.0508, "step": 10994 }, { "epoch": 7.0345489443378115, "grad_norm": 0.8035113100868588, "learning_rate": 4.2684327616073085e-06, "loss": 0.0561, "step": 10995 }, { "epoch": 7.035188739603327, "grad_norm": 0.8779993740448568, "learning_rate": 4.266734866855037e-06, "loss": 0.0594, "step": 10996 }, { "epoch": 7.035828534868842, "grad_norm": 0.8478850514747295, "learning_rate": 4.2650372182789745e-06, "loss": 0.0588, "step": 10997 }, { "epoch": 7.036468330134357, "grad_norm": 0.9064317946631922, "learning_rate": 4.263339815952014e-06, "loss": 0.053, "step": 10998 }, { "epoch": 7.037108125399872, "grad_norm": 0.824733795988668, "learning_rate": 4.261642659947032e-06, "loss": 0.0489, "step": 10999 }, { "epoch": 7.037747920665387, "grad_norm": 0.7201752917244775, "learning_rate": 4.2599457503369105e-06, "loss": 0.0513, "step": 11000 }, { "epoch": 7.038387715930902, "grad_norm": 0.7535422512658717, "learning_rate": 4.258249087194507e-06, "loss": 0.0501, "step": 11001 }, { "epoch": 7.039027511196418, "grad_norm": 0.729550900591934, "learning_rate": 4.256552670592674e-06, "loss": 0.0524, "step": 11002 }, { "epoch": 7.039667306461932, "grad_norm": 0.7995180237068843, "learning_rate": 4.2548565006042495e-06, "loss": 0.0601, "step": 11003 }, { "epoch": 7.040307101727447, "grad_norm": 0.8468681851251706, "learning_rate": 4.253160577302071e-06, "loss": 0.0518, "step": 11004 }, { "epoch": 7.0409468969929625, "grad_norm": 0.7845036789419262, "learning_rate": 4.251464900758953e-06, "loss": 0.0513, "step": 11005 }, { "epoch": 7.041586692258477, "grad_norm": 0.7867760666162938, "learning_rate": 4.249769471047708e-06, "loss": 0.059, "step": 11006 }, { "epoch": 7.042226487523992, "grad_norm": 0.8099890798337791, "learning_rate": 4.248074288241133e-06, "loss": 0.0598, "step": 11007 }, { "epoch": 7.042866282789507, "grad_norm": 0.6602745539516469, "learning_rate": 4.246379352412016e-06, "loss": 0.0502, "step": 11008 }, { "epoch": 7.043506078055023, "grad_norm": 0.7464059846839635, "learning_rate": 4.244684663633129e-06, "loss": 0.0517, "step": 11009 }, { "epoch": 7.044145873320537, "grad_norm": 0.7094675286091348, "learning_rate": 4.242990221977251e-06, "loss": 0.0527, "step": 11010 }, { "epoch": 7.044785668586052, "grad_norm": 0.8033808557564659, "learning_rate": 4.241296027517131e-06, "loss": 0.0585, "step": 11011 }, { "epoch": 7.045425463851568, "grad_norm": 0.7688078711978106, "learning_rate": 4.2396020803255156e-06, "loss": 0.0488, "step": 11012 }, { "epoch": 7.046065259117083, "grad_norm": 0.7839625666418701, "learning_rate": 4.2379083804751355e-06, "loss": 0.0507, "step": 11013 }, { "epoch": 7.046705054382597, "grad_norm": 0.8392508164802474, "learning_rate": 4.236214928038724e-06, "loss": 0.0584, "step": 11014 }, { "epoch": 7.047344849648113, "grad_norm": 0.7864701413873163, "learning_rate": 4.23452172308899e-06, "loss": 0.0519, "step": 11015 }, { "epoch": 7.047984644913628, "grad_norm": 0.8896130309224257, "learning_rate": 4.232828765698634e-06, "loss": 0.0596, "step": 11016 }, { "epoch": 7.048624440179142, "grad_norm": 0.7773162484221827, "learning_rate": 4.231136055940356e-06, "loss": 0.0565, "step": 11017 }, { "epoch": 7.0492642354446575, "grad_norm": 0.7776633017930309, "learning_rate": 4.229443593886833e-06, "loss": 0.0509, "step": 11018 }, { "epoch": 7.049904030710173, "grad_norm": 0.8420110109257611, "learning_rate": 4.227751379610737e-06, "loss": 0.0551, "step": 11019 }, { "epoch": 7.050543825975688, "grad_norm": 0.8708768567401494, "learning_rate": 4.226059413184729e-06, "loss": 0.0565, "step": 11020 }, { "epoch": 7.0511836212412025, "grad_norm": 0.828349833166212, "learning_rate": 4.22436769468146e-06, "loss": 0.0593, "step": 11021 }, { "epoch": 7.051823416506718, "grad_norm": 0.7027672063619772, "learning_rate": 4.222676224173565e-06, "loss": 0.0495, "step": 11022 }, { "epoch": 7.052463211772233, "grad_norm": 0.7555376723153125, "learning_rate": 4.220985001733675e-06, "loss": 0.0514, "step": 11023 }, { "epoch": 7.053103007037748, "grad_norm": 0.8356767076916891, "learning_rate": 4.219294027434412e-06, "loss": 0.0511, "step": 11024 }, { "epoch": 7.053742802303263, "grad_norm": 0.8529728249084373, "learning_rate": 4.21760330134838e-06, "loss": 0.0545, "step": 11025 }, { "epoch": 7.054382597568778, "grad_norm": 0.7891962650558125, "learning_rate": 4.215912823548172e-06, "loss": 0.0513, "step": 11026 }, { "epoch": 7.055022392834293, "grad_norm": 0.745885098592556, "learning_rate": 4.214222594106383e-06, "loss": 0.0529, "step": 11027 }, { "epoch": 7.0556621880998085, "grad_norm": 0.7379806625358378, "learning_rate": 4.212532613095582e-06, "loss": 0.0504, "step": 11028 }, { "epoch": 7.056301983365323, "grad_norm": 0.8391000887091123, "learning_rate": 4.210842880588334e-06, "loss": 0.0544, "step": 11029 }, { "epoch": 7.056941778630838, "grad_norm": 0.8928175736747063, "learning_rate": 4.209153396657197e-06, "loss": 0.0591, "step": 11030 }, { "epoch": 7.057581573896353, "grad_norm": 0.9123404645319142, "learning_rate": 4.207464161374713e-06, "loss": 0.0548, "step": 11031 }, { "epoch": 7.058221369161868, "grad_norm": 0.7728876558410732, "learning_rate": 4.205775174813412e-06, "loss": 0.055, "step": 11032 }, { "epoch": 7.058861164427383, "grad_norm": 0.7972244706492865, "learning_rate": 4.204086437045816e-06, "loss": 0.0553, "step": 11033 }, { "epoch": 7.059500959692898, "grad_norm": 0.8462065667245136, "learning_rate": 4.202397948144441e-06, "loss": 0.0546, "step": 11034 }, { "epoch": 7.060140754958414, "grad_norm": 0.8601596501005336, "learning_rate": 4.200709708181785e-06, "loss": 0.0551, "step": 11035 }, { "epoch": 7.060780550223928, "grad_norm": 0.8012977064156558, "learning_rate": 4.199021717230339e-06, "loss": 0.0512, "step": 11036 }, { "epoch": 7.061420345489443, "grad_norm": 0.8752144436260808, "learning_rate": 4.19733397536258e-06, "loss": 0.0562, "step": 11037 }, { "epoch": 7.062060140754959, "grad_norm": 0.6883693364106839, "learning_rate": 4.195646482650978e-06, "loss": 0.0468, "step": 11038 }, { "epoch": 7.062699936020474, "grad_norm": 0.7393659676249292, "learning_rate": 4.193959239167989e-06, "loss": 0.0491, "step": 11039 }, { "epoch": 7.063339731285988, "grad_norm": 0.8403606141552001, "learning_rate": 4.1922722449860634e-06, "loss": 0.0546, "step": 11040 }, { "epoch": 7.0639795265515035, "grad_norm": 0.7587266482764203, "learning_rate": 4.190585500177638e-06, "loss": 0.0509, "step": 11041 }, { "epoch": 7.064619321817019, "grad_norm": 0.8240836299944764, "learning_rate": 4.188899004815135e-06, "loss": 0.0545, "step": 11042 }, { "epoch": 7.065259117082533, "grad_norm": 0.8413395706958042, "learning_rate": 4.187212758970971e-06, "loss": 0.052, "step": 11043 }, { "epoch": 7.0658989123480485, "grad_norm": 0.8129875751656555, "learning_rate": 4.185526762717553e-06, "loss": 0.0526, "step": 11044 }, { "epoch": 7.066538707613564, "grad_norm": 0.866868242982445, "learning_rate": 4.183841016127273e-06, "loss": 0.0604, "step": 11045 }, { "epoch": 7.067178502879079, "grad_norm": 0.7458759694049049, "learning_rate": 4.18215551927251e-06, "loss": 0.0523, "step": 11046 }, { "epoch": 7.067818298144593, "grad_norm": 0.8549236641634688, "learning_rate": 4.1804702722256426e-06, "loss": 0.0547, "step": 11047 }, { "epoch": 7.068458093410109, "grad_norm": 0.7948133105403284, "learning_rate": 4.178785275059031e-06, "loss": 0.0467, "step": 11048 }, { "epoch": 7.069097888675624, "grad_norm": 0.7945358882809256, "learning_rate": 4.177100527845024e-06, "loss": 0.0501, "step": 11049 }, { "epoch": 7.069737683941139, "grad_norm": 0.7412481880065128, "learning_rate": 4.175416030655961e-06, "loss": 0.0535, "step": 11050 }, { "epoch": 7.070377479206654, "grad_norm": 0.7216039494581882, "learning_rate": 4.173731783564174e-06, "loss": 0.0508, "step": 11051 }, { "epoch": 7.071017274472169, "grad_norm": 0.8418706757833964, "learning_rate": 4.172047786641979e-06, "loss": 0.0505, "step": 11052 }, { "epoch": 7.071657069737684, "grad_norm": 0.7837016132008636, "learning_rate": 4.170364039961683e-06, "loss": 0.0542, "step": 11053 }, { "epoch": 7.0722968650031985, "grad_norm": 0.7583110873187537, "learning_rate": 4.168680543595587e-06, "loss": 0.0537, "step": 11054 }, { "epoch": 7.072936660268714, "grad_norm": 0.7714558710686238, "learning_rate": 4.166997297615976e-06, "loss": 0.0518, "step": 11055 }, { "epoch": 7.073576455534229, "grad_norm": 0.8221847544438097, "learning_rate": 4.165314302095121e-06, "loss": 0.0504, "step": 11056 }, { "epoch": 7.074216250799744, "grad_norm": 0.728170408367861, "learning_rate": 4.163631557105294e-06, "loss": 0.0487, "step": 11057 }, { "epoch": 7.074856046065259, "grad_norm": 0.7850529414799565, "learning_rate": 4.1619490627187475e-06, "loss": 0.052, "step": 11058 }, { "epoch": 7.075495841330774, "grad_norm": 0.7917859274503407, "learning_rate": 4.160266819007718e-06, "loss": 0.0525, "step": 11059 }, { "epoch": 7.076135636596289, "grad_norm": 0.7361569072207, "learning_rate": 4.158584826044447e-06, "loss": 0.0552, "step": 11060 }, { "epoch": 7.076775431861805, "grad_norm": 0.7692712955418954, "learning_rate": 4.1569030839011535e-06, "loss": 0.054, "step": 11061 }, { "epoch": 7.077415227127319, "grad_norm": 0.778269691408577, "learning_rate": 4.155221592650047e-06, "loss": 0.0507, "step": 11062 }, { "epoch": 7.078055022392834, "grad_norm": 0.8616327738376968, "learning_rate": 4.153540352363325e-06, "loss": 0.0588, "step": 11063 }, { "epoch": 7.0786948176583495, "grad_norm": 0.8379733264497936, "learning_rate": 4.151859363113182e-06, "loss": 0.0529, "step": 11064 }, { "epoch": 7.079334612923865, "grad_norm": 0.81205329852571, "learning_rate": 4.150178624971797e-06, "loss": 0.0583, "step": 11065 }, { "epoch": 7.079974408189379, "grad_norm": 0.8226507464274116, "learning_rate": 4.148498138011334e-06, "loss": 0.0501, "step": 11066 }, { "epoch": 7.0806142034548945, "grad_norm": 0.7707893140789637, "learning_rate": 4.146817902303953e-06, "loss": 0.0511, "step": 11067 }, { "epoch": 7.08125399872041, "grad_norm": 0.8272158378612752, "learning_rate": 4.1451379179218e-06, "loss": 0.0515, "step": 11068 }, { "epoch": 7.081893793985924, "grad_norm": 0.800763455573167, "learning_rate": 4.143458184937006e-06, "loss": 0.0542, "step": 11069 }, { "epoch": 7.082533589251439, "grad_norm": 0.7302876082537357, "learning_rate": 4.1417787034217015e-06, "loss": 0.05, "step": 11070 }, { "epoch": 7.083173384516955, "grad_norm": 0.9068950435807933, "learning_rate": 4.1400994734480005e-06, "loss": 0.0588, "step": 11071 }, { "epoch": 7.08381317978247, "grad_norm": 0.7682002129638553, "learning_rate": 4.138420495088004e-06, "loss": 0.0528, "step": 11072 }, { "epoch": 7.084452975047984, "grad_norm": 0.8751351944050053, "learning_rate": 4.136741768413799e-06, "loss": 0.0548, "step": 11073 }, { "epoch": 7.0850927703135, "grad_norm": 0.8623203077812264, "learning_rate": 4.135063293497478e-06, "loss": 0.0507, "step": 11074 }, { "epoch": 7.085732565579015, "grad_norm": 0.811556429036948, "learning_rate": 4.133385070411105e-06, "loss": 0.0569, "step": 11075 }, { "epoch": 7.08637236084453, "grad_norm": 0.7926403472351571, "learning_rate": 4.131707099226738e-06, "loss": 0.0612, "step": 11076 }, { "epoch": 7.0870121561100445, "grad_norm": 0.8793417069560658, "learning_rate": 4.130029380016434e-06, "loss": 0.0629, "step": 11077 }, { "epoch": 7.08765195137556, "grad_norm": 0.8102456145104874, "learning_rate": 4.1283519128522255e-06, "loss": 0.0589, "step": 11078 }, { "epoch": 7.088291746641075, "grad_norm": 0.911125546201388, "learning_rate": 4.12667469780614e-06, "loss": 0.0586, "step": 11079 }, { "epoch": 7.0889315419065895, "grad_norm": 0.8494627641569247, "learning_rate": 4.124997734950197e-06, "loss": 0.0592, "step": 11080 }, { "epoch": 7.089571337172105, "grad_norm": 0.7630163315364726, "learning_rate": 4.123321024356399e-06, "loss": 0.0506, "step": 11081 }, { "epoch": 7.09021113243762, "grad_norm": 0.8903516013421865, "learning_rate": 4.1216445660967444e-06, "loss": 0.0577, "step": 11082 }, { "epoch": 7.090850927703135, "grad_norm": 0.7738163269580701, "learning_rate": 4.119968360243209e-06, "loss": 0.055, "step": 11083 }, { "epoch": 7.09149072296865, "grad_norm": 0.7678057351158875, "learning_rate": 4.1182924068677774e-06, "loss": 0.048, "step": 11084 }, { "epoch": 7.092130518234165, "grad_norm": 0.9495822755206603, "learning_rate": 4.1166167060424075e-06, "loss": 0.0614, "step": 11085 }, { "epoch": 7.09277031349968, "grad_norm": 0.8609918805513426, "learning_rate": 4.1149412578390465e-06, "loss": 0.0623, "step": 11086 }, { "epoch": 7.0934101087651955, "grad_norm": 0.7911116522000469, "learning_rate": 4.113266062329644e-06, "loss": 0.0472, "step": 11087 }, { "epoch": 7.09404990403071, "grad_norm": 0.9083245287294659, "learning_rate": 4.111591119586123e-06, "loss": 0.0597, "step": 11088 }, { "epoch": 7.094689699296225, "grad_norm": 0.7894348593129313, "learning_rate": 4.109916429680406e-06, "loss": 0.0514, "step": 11089 }, { "epoch": 7.0953294945617404, "grad_norm": 0.7898367756486262, "learning_rate": 4.1082419926843965e-06, "loss": 0.0528, "step": 11090 }, { "epoch": 7.095969289827256, "grad_norm": 0.9174855410362378, "learning_rate": 4.106567808669998e-06, "loss": 0.0575, "step": 11091 }, { "epoch": 7.09660908509277, "grad_norm": 0.8749775367192458, "learning_rate": 4.1048938777090945e-06, "loss": 0.0525, "step": 11092 }, { "epoch": 7.097248880358285, "grad_norm": 0.9032503016064856, "learning_rate": 4.103220199873558e-06, "loss": 0.0524, "step": 11093 }, { "epoch": 7.097888675623801, "grad_norm": 0.8045321923154085, "learning_rate": 4.101546775235259e-06, "loss": 0.0516, "step": 11094 }, { "epoch": 7.098528470889315, "grad_norm": 0.7872460319527738, "learning_rate": 4.09987360386605e-06, "loss": 0.0519, "step": 11095 }, { "epoch": 7.09916826615483, "grad_norm": 0.8501232698540904, "learning_rate": 4.098200685837772e-06, "loss": 0.0534, "step": 11096 }, { "epoch": 7.099808061420346, "grad_norm": 0.8234944404741521, "learning_rate": 4.096528021222259e-06, "loss": 0.0533, "step": 11097 }, { "epoch": 7.100447856685861, "grad_norm": 0.8861224498989059, "learning_rate": 4.094855610091331e-06, "loss": 0.0592, "step": 11098 }, { "epoch": 7.101087651951375, "grad_norm": 0.8025563923594013, "learning_rate": 4.093183452516799e-06, "loss": 0.059, "step": 11099 }, { "epoch": 7.1017274472168905, "grad_norm": 0.7944549846989511, "learning_rate": 4.0915115485704586e-06, "loss": 0.0491, "step": 11100 }, { "epoch": 7.102367242482406, "grad_norm": 0.8312154219510072, "learning_rate": 4.089839898324105e-06, "loss": 0.0589, "step": 11101 }, { "epoch": 7.103007037747921, "grad_norm": 0.8244763041443182, "learning_rate": 4.088168501849513e-06, "loss": 0.0573, "step": 11102 }, { "epoch": 7.1036468330134355, "grad_norm": 0.9076330191079234, "learning_rate": 4.086497359218446e-06, "loss": 0.0585, "step": 11103 }, { "epoch": 7.104286628278951, "grad_norm": 0.8015762448577337, "learning_rate": 4.084826470502665e-06, "loss": 0.0539, "step": 11104 }, { "epoch": 7.104926423544466, "grad_norm": 0.8481776891403523, "learning_rate": 4.0831558357739155e-06, "loss": 0.0572, "step": 11105 }, { "epoch": 7.10556621880998, "grad_norm": 0.844000446389353, "learning_rate": 4.081485455103924e-06, "loss": 0.0547, "step": 11106 }, { "epoch": 7.106206014075496, "grad_norm": 0.8187126565386789, "learning_rate": 4.079815328564423e-06, "loss": 0.0592, "step": 11107 }, { "epoch": 7.106845809341011, "grad_norm": 0.8417801987454316, "learning_rate": 4.0781454562271206e-06, "loss": 0.0583, "step": 11108 }, { "epoch": 7.107485604606526, "grad_norm": 0.8190811676957653, "learning_rate": 4.076475838163717e-06, "loss": 0.0562, "step": 11109 }, { "epoch": 7.108125399872041, "grad_norm": 0.7774980224607737, "learning_rate": 4.074806474445905e-06, "loss": 0.053, "step": 11110 }, { "epoch": 7.108765195137556, "grad_norm": 0.7886954237274427, "learning_rate": 4.0731373651453635e-06, "loss": 0.0473, "step": 11111 }, { "epoch": 7.109404990403071, "grad_norm": 0.8450362664341886, "learning_rate": 4.071468510333759e-06, "loss": 0.0521, "step": 11112 }, { "epoch": 7.110044785668586, "grad_norm": 0.925597540310652, "learning_rate": 4.069799910082748e-06, "loss": 0.0531, "step": 11113 }, { "epoch": 7.110684580934101, "grad_norm": 0.8006542935016536, "learning_rate": 4.068131564463983e-06, "loss": 0.0554, "step": 11114 }, { "epoch": 7.111324376199616, "grad_norm": 0.8068872523533704, "learning_rate": 4.066463473549096e-06, "loss": 0.0535, "step": 11115 }, { "epoch": 7.111964171465131, "grad_norm": 0.9500913833587031, "learning_rate": 4.06479563740971e-06, "loss": 0.0593, "step": 11116 }, { "epoch": 7.112603966730646, "grad_norm": 0.8607624518093844, "learning_rate": 4.063128056117445e-06, "loss": 0.0557, "step": 11117 }, { "epoch": 7.113243761996161, "grad_norm": 0.8092664287123125, "learning_rate": 4.0614607297439e-06, "loss": 0.0514, "step": 11118 }, { "epoch": 7.113883557261676, "grad_norm": 0.8361695639412914, "learning_rate": 4.059793658360667e-06, "loss": 0.0603, "step": 11119 }, { "epoch": 7.114523352527192, "grad_norm": 0.8228934556294786, "learning_rate": 4.058126842039325e-06, "loss": 0.0535, "step": 11120 }, { "epoch": 7.115163147792706, "grad_norm": 0.7789228912641314, "learning_rate": 4.056460280851449e-06, "loss": 0.0551, "step": 11121 }, { "epoch": 7.115802943058221, "grad_norm": 0.8382147407100302, "learning_rate": 4.054793974868597e-06, "loss": 0.0534, "step": 11122 }, { "epoch": 7.1164427383237365, "grad_norm": 0.8307585890816906, "learning_rate": 4.053127924162311e-06, "loss": 0.0519, "step": 11123 }, { "epoch": 7.117082533589252, "grad_norm": 0.8314565081484567, "learning_rate": 4.0514621288041374e-06, "loss": 0.0598, "step": 11124 }, { "epoch": 7.117722328854766, "grad_norm": 0.7681075324039086, "learning_rate": 4.0497965888655996e-06, "loss": 0.0543, "step": 11125 }, { "epoch": 7.1183621241202815, "grad_norm": 0.852200541981109, "learning_rate": 4.04813130441821e-06, "loss": 0.0587, "step": 11126 }, { "epoch": 7.119001919385797, "grad_norm": 0.7442552641371926, "learning_rate": 4.046466275533476e-06, "loss": 0.0643, "step": 11127 }, { "epoch": 7.119641714651312, "grad_norm": 0.7668815164181941, "learning_rate": 4.044801502282889e-06, "loss": 0.0508, "step": 11128 }, { "epoch": 7.120281509916826, "grad_norm": 0.7237659021708341, "learning_rate": 4.043136984737932e-06, "loss": 0.0537, "step": 11129 }, { "epoch": 7.120921305182342, "grad_norm": 0.7063530092419794, "learning_rate": 4.041472722970072e-06, "loss": 0.0537, "step": 11130 }, { "epoch": 7.121561100447857, "grad_norm": 0.8341816910342135, "learning_rate": 4.03980871705078e-06, "loss": 0.0559, "step": 11131 }, { "epoch": 7.122200895713371, "grad_norm": 0.8597645518333016, "learning_rate": 4.038144967051498e-06, "loss": 0.0588, "step": 11132 }, { "epoch": 7.122840690978887, "grad_norm": 0.8716909167387626, "learning_rate": 4.0364814730436606e-06, "loss": 0.0552, "step": 11133 }, { "epoch": 7.123480486244402, "grad_norm": 0.7764249076061636, "learning_rate": 4.034818235098706e-06, "loss": 0.0541, "step": 11134 }, { "epoch": 7.124120281509917, "grad_norm": 0.8669242635409093, "learning_rate": 4.033155253288044e-06, "loss": 0.0576, "step": 11135 }, { "epoch": 7.1247600767754315, "grad_norm": 0.8291935883293158, "learning_rate": 4.031492527683079e-06, "loss": 0.0508, "step": 11136 }, { "epoch": 7.125399872040947, "grad_norm": 0.8410367217329803, "learning_rate": 4.0298300583552095e-06, "loss": 0.0522, "step": 11137 }, { "epoch": 7.126039667306462, "grad_norm": 0.8743591495968369, "learning_rate": 4.0281678453758175e-06, "loss": 0.0574, "step": 11138 }, { "epoch": 7.126679462571977, "grad_norm": 0.8945770477567101, "learning_rate": 4.026505888816277e-06, "loss": 0.056, "step": 11139 }, { "epoch": 7.127319257837492, "grad_norm": 0.7871784522381495, "learning_rate": 4.024844188747946e-06, "loss": 0.0493, "step": 11140 }, { "epoch": 7.127959053103007, "grad_norm": 0.9261132563924046, "learning_rate": 4.023182745242177e-06, "loss": 0.0564, "step": 11141 }, { "epoch": 7.128598848368522, "grad_norm": 0.8975566336500489, "learning_rate": 4.021521558370308e-06, "loss": 0.0598, "step": 11142 }, { "epoch": 7.129238643634037, "grad_norm": 0.8977019043557025, "learning_rate": 4.019860628203666e-06, "loss": 0.0533, "step": 11143 }, { "epoch": 7.129878438899552, "grad_norm": 0.8826285005836816, "learning_rate": 4.018199954813574e-06, "loss": 0.0576, "step": 11144 }, { "epoch": 7.130518234165067, "grad_norm": 0.8061564061031359, "learning_rate": 4.016539538271336e-06, "loss": 0.0512, "step": 11145 }, { "epoch": 7.1311580294305825, "grad_norm": 0.8599312589050148, "learning_rate": 4.014879378648242e-06, "loss": 0.0616, "step": 11146 }, { "epoch": 7.131797824696097, "grad_norm": 0.7011432899898431, "learning_rate": 4.0132194760155844e-06, "loss": 0.0469, "step": 11147 }, { "epoch": 7.132437619961612, "grad_norm": 0.8887071519628491, "learning_rate": 4.011559830444632e-06, "loss": 0.0577, "step": 11148 }, { "epoch": 7.1330774152271275, "grad_norm": 0.7651399197921676, "learning_rate": 4.009900442006649e-06, "loss": 0.0511, "step": 11149 }, { "epoch": 7.133717210492643, "grad_norm": 0.7859185968116426, "learning_rate": 4.008241310772882e-06, "loss": 0.0516, "step": 11150 }, { "epoch": 7.134357005758157, "grad_norm": 0.7764547671178683, "learning_rate": 4.006582436814578e-06, "loss": 0.0485, "step": 11151 }, { "epoch": 7.134996801023672, "grad_norm": 0.8510689755836047, "learning_rate": 4.004923820202963e-06, "loss": 0.0536, "step": 11152 }, { "epoch": 7.135636596289188, "grad_norm": 0.9402689687711862, "learning_rate": 4.003265461009251e-06, "loss": 0.0594, "step": 11153 }, { "epoch": 7.136276391554702, "grad_norm": 0.8464856680164766, "learning_rate": 4.001607359304656e-06, "loss": 0.0551, "step": 11154 }, { "epoch": 7.136916186820217, "grad_norm": 0.9048047938682707, "learning_rate": 3.999949515160369e-06, "loss": 0.0536, "step": 11155 }, { "epoch": 7.137555982085733, "grad_norm": 0.8064173658804816, "learning_rate": 3.998291928647579e-06, "loss": 0.0574, "step": 11156 }, { "epoch": 7.138195777351248, "grad_norm": 0.8642357497649078, "learning_rate": 3.996634599837457e-06, "loss": 0.0577, "step": 11157 }, { "epoch": 7.138835572616762, "grad_norm": 0.7999184938940368, "learning_rate": 3.994977528801166e-06, "loss": 0.0537, "step": 11158 }, { "epoch": 7.1394753678822775, "grad_norm": 0.823645683774799, "learning_rate": 3.993320715609857e-06, "loss": 0.0534, "step": 11159 }, { "epoch": 7.140115163147793, "grad_norm": 0.904554132606454, "learning_rate": 3.9916641603346685e-06, "loss": 0.0549, "step": 11160 }, { "epoch": 7.140754958413308, "grad_norm": 0.8521192231185811, "learning_rate": 3.990007863046737e-06, "loss": 0.0485, "step": 11161 }, { "epoch": 7.1413947536788225, "grad_norm": 0.7146817580165944, "learning_rate": 3.988351823817176e-06, "loss": 0.0506, "step": 11162 }, { "epoch": 7.142034548944338, "grad_norm": 0.8463199307850594, "learning_rate": 3.986696042717091e-06, "loss": 0.0532, "step": 11163 }, { "epoch": 7.142674344209853, "grad_norm": 0.8163073211555355, "learning_rate": 3.985040519817584e-06, "loss": 0.0574, "step": 11164 }, { "epoch": 7.143314139475368, "grad_norm": 0.7628249751442233, "learning_rate": 3.983385255189738e-06, "loss": 0.055, "step": 11165 }, { "epoch": 7.143953934740883, "grad_norm": 0.8214193845270782, "learning_rate": 3.981730248904625e-06, "loss": 0.0591, "step": 11166 }, { "epoch": 7.144593730006398, "grad_norm": 0.8185936676579599, "learning_rate": 3.980075501033307e-06, "loss": 0.0533, "step": 11167 }, { "epoch": 7.145233525271913, "grad_norm": 0.7059007073218246, "learning_rate": 3.978421011646843e-06, "loss": 0.0485, "step": 11168 }, { "epoch": 7.145873320537428, "grad_norm": 0.9244624372375632, "learning_rate": 3.976766780816267e-06, "loss": 0.0603, "step": 11169 }, { "epoch": 7.146513115802943, "grad_norm": 0.7405773157787516, "learning_rate": 3.9751128086126125e-06, "loss": 0.0507, "step": 11170 }, { "epoch": 7.147152911068458, "grad_norm": 0.8739668719486765, "learning_rate": 3.973459095106895e-06, "loss": 0.0621, "step": 11171 }, { "epoch": 7.1477927063339735, "grad_norm": 0.8402599790224916, "learning_rate": 3.971805640370125e-06, "loss": 0.0561, "step": 11172 }, { "epoch": 7.148432501599488, "grad_norm": 0.7716109129716416, "learning_rate": 3.970152444473293e-06, "loss": 0.0515, "step": 11173 }, { "epoch": 7.149072296865003, "grad_norm": 0.8216522930426875, "learning_rate": 3.968499507487393e-06, "loss": 0.0585, "step": 11174 }, { "epoch": 7.149712092130518, "grad_norm": 0.8455727328880968, "learning_rate": 3.966846829483395e-06, "loss": 0.0482, "step": 11175 }, { "epoch": 7.150351887396034, "grad_norm": 0.8095520154100612, "learning_rate": 3.9651944105322615e-06, "loss": 0.0498, "step": 11176 }, { "epoch": 7.150991682661548, "grad_norm": 0.7780871899660639, "learning_rate": 3.9635422507049415e-06, "loss": 0.0521, "step": 11177 }, { "epoch": 7.151631477927063, "grad_norm": 0.7680531684033929, "learning_rate": 3.961890350072382e-06, "loss": 0.0563, "step": 11178 }, { "epoch": 7.152271273192579, "grad_norm": 0.8931071054895678, "learning_rate": 3.9602387087055116e-06, "loss": 0.0593, "step": 11179 }, { "epoch": 7.152911068458093, "grad_norm": 0.7826397930953188, "learning_rate": 3.9585873266752425e-06, "loss": 0.0534, "step": 11180 }, { "epoch": 7.153550863723608, "grad_norm": 0.7942250411077436, "learning_rate": 3.956936204052491e-06, "loss": 0.0523, "step": 11181 }, { "epoch": 7.1541906589891235, "grad_norm": 0.6999181501540824, "learning_rate": 3.95528534090815e-06, "loss": 0.0466, "step": 11182 }, { "epoch": 7.154830454254639, "grad_norm": 0.8821682814011949, "learning_rate": 3.953634737313098e-06, "loss": 0.0557, "step": 11183 }, { "epoch": 7.155470249520153, "grad_norm": 0.8641016076161884, "learning_rate": 3.951984393338219e-06, "loss": 0.0534, "step": 11184 }, { "epoch": 7.1561100447856685, "grad_norm": 0.7890377055164737, "learning_rate": 3.950334309054372e-06, "loss": 0.0577, "step": 11185 }, { "epoch": 7.156749840051184, "grad_norm": 0.7845712004842202, "learning_rate": 3.948684484532408e-06, "loss": 0.0546, "step": 11186 }, { "epoch": 7.157389635316699, "grad_norm": 0.8997404772635792, "learning_rate": 3.947034919843168e-06, "loss": 0.0515, "step": 11187 }, { "epoch": 7.158029430582213, "grad_norm": 0.8932263336013058, "learning_rate": 3.94538561505748e-06, "loss": 0.0507, "step": 11188 }, { "epoch": 7.158669225847729, "grad_norm": 0.8677847729395218, "learning_rate": 3.943736570246165e-06, "loss": 0.0597, "step": 11189 }, { "epoch": 7.159309021113244, "grad_norm": 0.8712030946161844, "learning_rate": 3.9420877854800246e-06, "loss": 0.0586, "step": 11190 }, { "epoch": 7.159948816378758, "grad_norm": 0.7533800461888894, "learning_rate": 3.94043926082986e-06, "loss": 0.0488, "step": 11191 }, { "epoch": 7.160588611644274, "grad_norm": 0.8716867603725411, "learning_rate": 3.938790996366457e-06, "loss": 0.0521, "step": 11192 }, { "epoch": 7.161228406909789, "grad_norm": 0.9526977919492616, "learning_rate": 3.937142992160582e-06, "loss": 0.066, "step": 11193 }, { "epoch": 7.161868202175304, "grad_norm": 0.7675641966979081, "learning_rate": 3.935495248283006e-06, "loss": 0.0531, "step": 11194 }, { "epoch": 7.162507997440819, "grad_norm": 0.8369507150577803, "learning_rate": 3.933847764804475e-06, "loss": 0.053, "step": 11195 }, { "epoch": 7.163147792706334, "grad_norm": 0.8003577447898984, "learning_rate": 3.932200541795731e-06, "loss": 0.0479, "step": 11196 }, { "epoch": 7.163787587971849, "grad_norm": 0.8138051256725058, "learning_rate": 3.930553579327499e-06, "loss": 0.0503, "step": 11197 }, { "epoch": 7.164427383237364, "grad_norm": 0.8030379910742043, "learning_rate": 3.928906877470501e-06, "loss": 0.0508, "step": 11198 }, { "epoch": 7.165067178502879, "grad_norm": 0.8747183205408504, "learning_rate": 3.927260436295444e-06, "loss": 0.053, "step": 11199 }, { "epoch": 7.165706973768394, "grad_norm": 0.8268653995955196, "learning_rate": 3.925614255873021e-06, "loss": 0.0568, "step": 11200 }, { "epoch": 7.166346769033909, "grad_norm": 0.7713881028487384, "learning_rate": 3.923968336273916e-06, "loss": 0.0542, "step": 11201 }, { "epoch": 7.166986564299425, "grad_norm": 0.9126998528215646, "learning_rate": 3.922322677568802e-06, "loss": 0.0585, "step": 11202 }, { "epoch": 7.167626359564939, "grad_norm": 0.8150545843826082, "learning_rate": 3.920677279828336e-06, "loss": 0.0571, "step": 11203 }, { "epoch": 7.168266154830454, "grad_norm": 0.7856096866955556, "learning_rate": 3.919032143123179e-06, "loss": 0.0545, "step": 11204 }, { "epoch": 7.1689059500959695, "grad_norm": 1.0373390902334696, "learning_rate": 3.917387267523964e-06, "loss": 0.0637, "step": 11205 }, { "epoch": 7.169545745361484, "grad_norm": 0.8648360047984492, "learning_rate": 3.91574265310132e-06, "loss": 0.055, "step": 11206 }, { "epoch": 7.170185540626999, "grad_norm": 0.8509138967681198, "learning_rate": 3.914098299925859e-06, "loss": 0.0508, "step": 11207 }, { "epoch": 7.1708253358925145, "grad_norm": 0.8529911322403144, "learning_rate": 3.912454208068196e-06, "loss": 0.0569, "step": 11208 }, { "epoch": 7.17146513115803, "grad_norm": 0.8640841063480825, "learning_rate": 3.910810377598919e-06, "loss": 0.0544, "step": 11209 }, { "epoch": 7.172104926423544, "grad_norm": 0.8434843060146419, "learning_rate": 3.90916680858861e-06, "loss": 0.0516, "step": 11210 }, { "epoch": 7.172744721689059, "grad_norm": 0.8169113271540204, "learning_rate": 3.907523501107846e-06, "loss": 0.0562, "step": 11211 }, { "epoch": 7.173384516954575, "grad_norm": 0.7631220504825384, "learning_rate": 3.905880455227185e-06, "loss": 0.0521, "step": 11212 }, { "epoch": 7.17402431222009, "grad_norm": 0.837430850342094, "learning_rate": 3.904237671017175e-06, "loss": 0.0562, "step": 11213 }, { "epoch": 7.174664107485604, "grad_norm": 0.8040875433232965, "learning_rate": 3.9025951485483584e-06, "loss": 0.0475, "step": 11214 }, { "epoch": 7.17530390275112, "grad_norm": 0.8741389583899799, "learning_rate": 3.900952887891259e-06, "loss": 0.0552, "step": 11215 }, { "epoch": 7.175943698016635, "grad_norm": 0.929032643187807, "learning_rate": 3.8993108891163934e-06, "loss": 0.0595, "step": 11216 }, { "epoch": 7.176583493282149, "grad_norm": 0.9189111088647173, "learning_rate": 3.897669152294267e-06, "loss": 0.0505, "step": 11217 }, { "epoch": 7.177223288547665, "grad_norm": 0.7513568992670715, "learning_rate": 3.896027677495371e-06, "loss": 0.0527, "step": 11218 }, { "epoch": 7.17786308381318, "grad_norm": 0.9023737265337118, "learning_rate": 3.894386464790189e-06, "loss": 0.0557, "step": 11219 }, { "epoch": 7.178502879078695, "grad_norm": 0.7821315311939705, "learning_rate": 3.892745514249187e-06, "loss": 0.0495, "step": 11220 }, { "epoch": 7.1791426743442095, "grad_norm": 0.8976122642986764, "learning_rate": 3.891104825942832e-06, "loss": 0.0561, "step": 11221 }, { "epoch": 7.179782469609725, "grad_norm": 0.7875722673044458, "learning_rate": 3.889464399941569e-06, "loss": 0.0491, "step": 11222 }, { "epoch": 7.18042226487524, "grad_norm": 0.840269993383709, "learning_rate": 3.887824236315833e-06, "loss": 0.0534, "step": 11223 }, { "epoch": 7.181062060140755, "grad_norm": 0.8576859992343542, "learning_rate": 3.886184335136053e-06, "loss": 0.0604, "step": 11224 }, { "epoch": 7.18170185540627, "grad_norm": 0.8550126231734428, "learning_rate": 3.884544696472643e-06, "loss": 0.0513, "step": 11225 }, { "epoch": 7.182341650671785, "grad_norm": 0.7756747344780324, "learning_rate": 3.882905320396005e-06, "loss": 0.0528, "step": 11226 }, { "epoch": 7.1829814459373, "grad_norm": 0.8207393405088971, "learning_rate": 3.881266206976528e-06, "loss": 0.0523, "step": 11227 }, { "epoch": 7.183621241202815, "grad_norm": 0.8014773884428925, "learning_rate": 3.879627356284599e-06, "loss": 0.0521, "step": 11228 }, { "epoch": 7.18426103646833, "grad_norm": 0.8154483133600583, "learning_rate": 3.877988768390584e-06, "loss": 0.0524, "step": 11229 }, { "epoch": 7.184900831733845, "grad_norm": 0.9133676165428003, "learning_rate": 3.8763504433648406e-06, "loss": 0.0531, "step": 11230 }, { "epoch": 7.1855406269993605, "grad_norm": 0.8797880358431424, "learning_rate": 3.874712381277716e-06, "loss": 0.0584, "step": 11231 }, { "epoch": 7.186180422264875, "grad_norm": 0.7835589073211945, "learning_rate": 3.873074582199546e-06, "loss": 0.0528, "step": 11232 }, { "epoch": 7.18682021753039, "grad_norm": 0.6588602914812605, "learning_rate": 3.871437046200654e-06, "loss": 0.0467, "step": 11233 }, { "epoch": 7.187460012795905, "grad_norm": 0.819110488929994, "learning_rate": 3.869799773351351e-06, "loss": 0.0547, "step": 11234 }, { "epoch": 7.188099808061421, "grad_norm": 0.8789067540350488, "learning_rate": 3.868162763721943e-06, "loss": 0.0576, "step": 11235 }, { "epoch": 7.188739603326935, "grad_norm": 0.8352521871553876, "learning_rate": 3.86652601738272e-06, "loss": 0.0639, "step": 11236 }, { "epoch": 7.18937939859245, "grad_norm": 0.8892845612831745, "learning_rate": 3.864889534403955e-06, "loss": 0.061, "step": 11237 }, { "epoch": 7.190019193857966, "grad_norm": 0.8698694767880015, "learning_rate": 3.863253314855922e-06, "loss": 0.0545, "step": 11238 }, { "epoch": 7.190658989123481, "grad_norm": 0.7711542493890915, "learning_rate": 3.861617358808877e-06, "loss": 0.0499, "step": 11239 }, { "epoch": 7.191298784388995, "grad_norm": 0.7433372010326372, "learning_rate": 3.859981666333059e-06, "loss": 0.0511, "step": 11240 }, { "epoch": 7.1919385796545106, "grad_norm": 0.8256242416408368, "learning_rate": 3.85834623749871e-06, "loss": 0.0525, "step": 11241 }, { "epoch": 7.192578374920026, "grad_norm": 0.857854984329621, "learning_rate": 3.856711072376048e-06, "loss": 0.0593, "step": 11242 }, { "epoch": 7.19321817018554, "grad_norm": 0.8727941590603715, "learning_rate": 3.855076171035285e-06, "loss": 0.0547, "step": 11243 }, { "epoch": 7.1938579654510555, "grad_norm": 0.8398971580125054, "learning_rate": 3.853441533546616e-06, "loss": 0.0521, "step": 11244 }, { "epoch": 7.194497760716571, "grad_norm": 0.8534967079772676, "learning_rate": 3.851807159980238e-06, "loss": 0.0604, "step": 11245 }, { "epoch": 7.195137555982086, "grad_norm": 0.7970966241989403, "learning_rate": 3.850173050406323e-06, "loss": 0.0521, "step": 11246 }, { "epoch": 7.1957773512476, "grad_norm": 0.8243919069947173, "learning_rate": 3.848539204895038e-06, "loss": 0.0476, "step": 11247 }, { "epoch": 7.196417146513116, "grad_norm": 0.8917253740747715, "learning_rate": 3.846905623516536e-06, "loss": 0.0558, "step": 11248 }, { "epoch": 7.197056941778631, "grad_norm": 0.8276004915302068, "learning_rate": 3.845272306340962e-06, "loss": 0.0576, "step": 11249 }, { "epoch": 7.197696737044146, "grad_norm": 0.8370572114660584, "learning_rate": 3.843639253438443e-06, "loss": 0.0549, "step": 11250 }, { "epoch": 7.198336532309661, "grad_norm": 0.8661022142894504, "learning_rate": 3.842006464879107e-06, "loss": 0.0536, "step": 11251 }, { "epoch": 7.198976327575176, "grad_norm": 0.8277736035055694, "learning_rate": 3.8403739407330575e-06, "loss": 0.0538, "step": 11252 }, { "epoch": 7.199616122840691, "grad_norm": 0.877078617674199, "learning_rate": 3.838741681070396e-06, "loss": 0.0559, "step": 11253 }, { "epoch": 7.2002559181062065, "grad_norm": 0.8525067805907638, "learning_rate": 3.837109685961202e-06, "loss": 0.0535, "step": 11254 }, { "epoch": 7.200895713371721, "grad_norm": 0.872816001224857, "learning_rate": 3.83547795547556e-06, "loss": 0.0598, "step": 11255 }, { "epoch": 7.201535508637236, "grad_norm": 0.805331767261021, "learning_rate": 3.833846489683527e-06, "loss": 0.0536, "step": 11256 }, { "epoch": 7.202175303902751, "grad_norm": 0.9313077562149777, "learning_rate": 3.8322152886551545e-06, "loss": 0.0615, "step": 11257 }, { "epoch": 7.202815099168266, "grad_norm": 0.8938180832944347, "learning_rate": 3.830584352460489e-06, "loss": 0.0547, "step": 11258 }, { "epoch": 7.203454894433781, "grad_norm": 0.7941266022431687, "learning_rate": 3.8289536811695575e-06, "loss": 0.0493, "step": 11259 }, { "epoch": 7.204094689699296, "grad_norm": 0.9425662915703782, "learning_rate": 3.827323274852376e-06, "loss": 0.057, "step": 11260 }, { "epoch": 7.204734484964812, "grad_norm": 0.7315388638323534, "learning_rate": 3.825693133578954e-06, "loss": 0.0568, "step": 11261 }, { "epoch": 7.205374280230326, "grad_norm": 0.839536504990991, "learning_rate": 3.824063257419285e-06, "loss": 0.0555, "step": 11262 }, { "epoch": 7.206014075495841, "grad_norm": 0.842482758761126, "learning_rate": 3.822433646443354e-06, "loss": 0.0567, "step": 11263 }, { "epoch": 7.2066538707613566, "grad_norm": 0.9006166540721093, "learning_rate": 3.820804300721129e-06, "loss": 0.0554, "step": 11264 }, { "epoch": 7.207293666026872, "grad_norm": 0.7735288564167732, "learning_rate": 3.81917522032258e-06, "loss": 0.0521, "step": 11265 }, { "epoch": 7.207933461292386, "grad_norm": 0.8544116538650146, "learning_rate": 3.817546405317652e-06, "loss": 0.0548, "step": 11266 }, { "epoch": 7.2085732565579015, "grad_norm": 0.8194335296477957, "learning_rate": 3.81591785577628e-06, "loss": 0.0559, "step": 11267 }, { "epoch": 7.209213051823417, "grad_norm": 0.8749882466181207, "learning_rate": 3.814289571768397e-06, "loss": 0.0562, "step": 11268 }, { "epoch": 7.209852847088931, "grad_norm": 1.0321102732995688, "learning_rate": 3.8126615533639177e-06, "loss": 0.0576, "step": 11269 }, { "epoch": 7.210492642354446, "grad_norm": 0.8857225521535561, "learning_rate": 3.8110338006327406e-06, "loss": 0.0568, "step": 11270 }, { "epoch": 7.211132437619962, "grad_norm": 0.7597977012234526, "learning_rate": 3.809406313644767e-06, "loss": 0.048, "step": 11271 }, { "epoch": 7.211772232885477, "grad_norm": 0.805973821499779, "learning_rate": 3.807779092469873e-06, "loss": 0.056, "step": 11272 }, { "epoch": 7.212412028150991, "grad_norm": 0.7485889590373126, "learning_rate": 3.8061521371779307e-06, "loss": 0.0551, "step": 11273 }, { "epoch": 7.213051823416507, "grad_norm": 0.8285717053491255, "learning_rate": 3.8045254478387937e-06, "loss": 0.0506, "step": 11274 }, { "epoch": 7.213691618682022, "grad_norm": 0.8874451330189088, "learning_rate": 3.8028990245223164e-06, "loss": 0.0583, "step": 11275 }, { "epoch": 7.214331413947537, "grad_norm": 0.8219170328700706, "learning_rate": 3.8012728672983312e-06, "loss": 0.0603, "step": 11276 }, { "epoch": 7.214971209213052, "grad_norm": 0.9732370098445643, "learning_rate": 3.799646976236663e-06, "loss": 0.0598, "step": 11277 }, { "epoch": 7.215611004478567, "grad_norm": 0.8769971777893053, "learning_rate": 3.798021351407124e-06, "loss": 0.055, "step": 11278 }, { "epoch": 7.216250799744082, "grad_norm": 0.7337724441321656, "learning_rate": 3.796395992879516e-06, "loss": 0.0523, "step": 11279 }, { "epoch": 7.2168905950095965, "grad_norm": 0.9025469036540661, "learning_rate": 3.794770900723623e-06, "loss": 0.0543, "step": 11280 }, { "epoch": 7.217530390275112, "grad_norm": 0.8139041205394448, "learning_rate": 3.7931460750092353e-06, "loss": 0.0511, "step": 11281 }, { "epoch": 7.218170185540627, "grad_norm": 0.8719140330943284, "learning_rate": 3.7915215158061126e-06, "loss": 0.0571, "step": 11282 }, { "epoch": 7.218809980806142, "grad_norm": 0.7744064925145578, "learning_rate": 3.7898972231840125e-06, "loss": 0.0507, "step": 11283 }, { "epoch": 7.219449776071657, "grad_norm": 0.7886235599475598, "learning_rate": 3.788273197212675e-06, "loss": 0.0478, "step": 11284 }, { "epoch": 7.220089571337172, "grad_norm": 0.9459108231824739, "learning_rate": 3.7866494379618414e-06, "loss": 0.0628, "step": 11285 }, { "epoch": 7.220729366602687, "grad_norm": 0.8072606495472543, "learning_rate": 3.785025945501226e-06, "loss": 0.0593, "step": 11286 }, { "epoch": 7.2213691618682025, "grad_norm": 0.8442741691505268, "learning_rate": 3.7834027199005395e-06, "loss": 0.0642, "step": 11287 }, { "epoch": 7.222008957133717, "grad_norm": 0.7659062810911494, "learning_rate": 3.7817797612294838e-06, "loss": 0.0504, "step": 11288 }, { "epoch": 7.222648752399232, "grad_norm": 0.770522410106806, "learning_rate": 3.780157069557744e-06, "loss": 0.0488, "step": 11289 }, { "epoch": 7.2232885476647475, "grad_norm": 0.7239418312320902, "learning_rate": 3.7785346449549964e-06, "loss": 0.0476, "step": 11290 }, { "epoch": 7.223928342930263, "grad_norm": 0.8937679142165087, "learning_rate": 3.7769124874909025e-06, "loss": 0.0572, "step": 11291 }, { "epoch": 7.224568138195777, "grad_norm": 0.8012058550644767, "learning_rate": 3.7752905972351175e-06, "loss": 0.057, "step": 11292 }, { "epoch": 7.225207933461292, "grad_norm": 0.8429456705688861, "learning_rate": 3.7736689742572806e-06, "loss": 0.057, "step": 11293 }, { "epoch": 7.225847728726808, "grad_norm": 0.8652799642720023, "learning_rate": 3.7720476186270193e-06, "loss": 0.0567, "step": 11294 }, { "epoch": 7.226487523992322, "grad_norm": 0.8594555069833681, "learning_rate": 3.770426530413959e-06, "loss": 0.0492, "step": 11295 }, { "epoch": 7.227127319257837, "grad_norm": 0.9291502595195155, "learning_rate": 3.7688057096876994e-06, "loss": 0.0615, "step": 11296 }, { "epoch": 7.227767114523353, "grad_norm": 0.7827206201370405, "learning_rate": 3.767185156517836e-06, "loss": 0.0523, "step": 11297 }, { "epoch": 7.228406909788868, "grad_norm": 0.8182102652841741, "learning_rate": 3.765564870973959e-06, "loss": 0.0544, "step": 11298 }, { "epoch": 7.229046705054382, "grad_norm": 0.8345175344670086, "learning_rate": 3.763944853125635e-06, "loss": 0.0595, "step": 11299 }, { "epoch": 7.229686500319898, "grad_norm": 0.8714875910704177, "learning_rate": 3.7623251030424233e-06, "loss": 0.0553, "step": 11300 }, { "epoch": 7.230326295585413, "grad_norm": 0.8229542220192358, "learning_rate": 3.7607056207938786e-06, "loss": 0.0559, "step": 11301 }, { "epoch": 7.230966090850928, "grad_norm": 0.8788610503380915, "learning_rate": 3.759086406449536e-06, "loss": 0.057, "step": 11302 }, { "epoch": 7.2316058861164425, "grad_norm": 0.8550628586182932, "learning_rate": 3.757467460078922e-06, "loss": 0.0577, "step": 11303 }, { "epoch": 7.232245681381958, "grad_norm": 0.9018900810430015, "learning_rate": 3.7558487817515464e-06, "loss": 0.0527, "step": 11304 }, { "epoch": 7.232885476647473, "grad_norm": 0.8480642721614817, "learning_rate": 3.75423037153692e-06, "loss": 0.0542, "step": 11305 }, { "epoch": 7.233525271912987, "grad_norm": 0.8221084412775334, "learning_rate": 3.752612229504531e-06, "loss": 0.0545, "step": 11306 }, { "epoch": 7.234165067178503, "grad_norm": 0.8178285559083279, "learning_rate": 3.7509943557238605e-06, "loss": 0.0548, "step": 11307 }, { "epoch": 7.234804862444018, "grad_norm": 0.8393779409316655, "learning_rate": 3.7493767502643753e-06, "loss": 0.0542, "step": 11308 }, { "epoch": 7.235444657709533, "grad_norm": 0.8805706378841753, "learning_rate": 3.747759413195534e-06, "loss": 0.0628, "step": 11309 }, { "epoch": 7.236084452975048, "grad_norm": 0.8858366555211861, "learning_rate": 3.7461423445867807e-06, "loss": 0.0631, "step": 11310 }, { "epoch": 7.236724248240563, "grad_norm": 0.834377617634654, "learning_rate": 3.744525544507548e-06, "loss": 0.0516, "step": 11311 }, { "epoch": 7.237364043506078, "grad_norm": 0.7763378473523869, "learning_rate": 3.742909013027264e-06, "loss": 0.0524, "step": 11312 }, { "epoch": 7.2380038387715935, "grad_norm": 0.8458139430981345, "learning_rate": 3.7412927502153364e-06, "loss": 0.056, "step": 11313 }, { "epoch": 7.238643634037108, "grad_norm": 0.7992110192699076, "learning_rate": 3.7396767561411605e-06, "loss": 0.0567, "step": 11314 }, { "epoch": 7.239283429302623, "grad_norm": 0.8680331827311016, "learning_rate": 3.7380610308741326e-06, "loss": 0.0553, "step": 11315 }, { "epoch": 7.239923224568138, "grad_norm": 0.8192796643476061, "learning_rate": 3.736445574483626e-06, "loss": 0.0566, "step": 11316 }, { "epoch": 7.240563019833653, "grad_norm": 0.8365392045108008, "learning_rate": 3.734830387038999e-06, "loss": 0.0597, "step": 11317 }, { "epoch": 7.241202815099168, "grad_norm": 0.8098945579905861, "learning_rate": 3.7332154686096154e-06, "loss": 0.0505, "step": 11318 }, { "epoch": 7.241842610364683, "grad_norm": 0.8183644972853401, "learning_rate": 3.7316008192648114e-06, "loss": 0.0542, "step": 11319 }, { "epoch": 7.242482405630199, "grad_norm": 0.8405647930485157, "learning_rate": 3.729986439073917e-06, "loss": 0.0534, "step": 11320 }, { "epoch": 7.243122200895713, "grad_norm": 0.8483839288988979, "learning_rate": 3.728372328106251e-06, "loss": 0.053, "step": 11321 }, { "epoch": 7.243761996161228, "grad_norm": 0.8669935494190341, "learning_rate": 3.7267584864311224e-06, "loss": 0.0548, "step": 11322 }, { "epoch": 7.244401791426744, "grad_norm": 0.7595435702553025, "learning_rate": 3.725144914117824e-06, "loss": 0.0472, "step": 11323 }, { "epoch": 7.245041586692259, "grad_norm": 0.8863286800691815, "learning_rate": 3.723531611235638e-06, "loss": 0.0578, "step": 11324 }, { "epoch": 7.245681381957773, "grad_norm": 0.8058568934148743, "learning_rate": 3.7219185778538425e-06, "loss": 0.0482, "step": 11325 }, { "epoch": 7.2463211772232885, "grad_norm": 0.8236612787136834, "learning_rate": 3.720305814041696e-06, "loss": 0.0566, "step": 11326 }, { "epoch": 7.246960972488804, "grad_norm": 0.8519891847113417, "learning_rate": 3.718693319868444e-06, "loss": 0.0583, "step": 11327 }, { "epoch": 7.247600767754319, "grad_norm": 0.8557107592045863, "learning_rate": 3.7170810954033297e-06, "loss": 0.0558, "step": 11328 }, { "epoch": 7.248240563019833, "grad_norm": 0.890852254361617, "learning_rate": 3.7154691407155774e-06, "loss": 0.0602, "step": 11329 }, { "epoch": 7.248880358285349, "grad_norm": 0.8458947782525218, "learning_rate": 3.713857455874401e-06, "loss": 0.0545, "step": 11330 }, { "epoch": 7.249520153550864, "grad_norm": 0.9483393154349657, "learning_rate": 3.712246040948999e-06, "loss": 0.0565, "step": 11331 }, { "epoch": 7.250159948816378, "grad_norm": 0.7701331367764555, "learning_rate": 3.7106348960085716e-06, "loss": 0.0488, "step": 11332 }, { "epoch": 7.250799744081894, "grad_norm": 0.8462098106388758, "learning_rate": 3.7090240211222927e-06, "loss": 0.0526, "step": 11333 }, { "epoch": 7.251439539347409, "grad_norm": 0.8853326660630808, "learning_rate": 3.707413416359329e-06, "loss": 0.0548, "step": 11334 }, { "epoch": 7.252079334612924, "grad_norm": 0.9289946684541286, "learning_rate": 3.7058030817888424e-06, "loss": 0.0521, "step": 11335 }, { "epoch": 7.252719129878439, "grad_norm": 0.8704172568627659, "learning_rate": 3.7041930174799755e-06, "loss": 0.0564, "step": 11336 }, { "epoch": 7.253358925143954, "grad_norm": 0.8473795207549589, "learning_rate": 3.7025832235018612e-06, "loss": 0.0468, "step": 11337 }, { "epoch": 7.253998720409469, "grad_norm": 0.9981895213947966, "learning_rate": 3.70097369992362e-06, "loss": 0.06, "step": 11338 }, { "epoch": 7.254638515674984, "grad_norm": 0.8747226460269449, "learning_rate": 3.699364446814363e-06, "loss": 0.0545, "step": 11339 }, { "epoch": 7.255278310940499, "grad_norm": 0.8512435671341436, "learning_rate": 3.69775546424319e-06, "loss": 0.0471, "step": 11340 }, { "epoch": 7.255918106206014, "grad_norm": 0.8099433712779456, "learning_rate": 3.6961467522791827e-06, "loss": 0.0566, "step": 11341 }, { "epoch": 7.256557901471529, "grad_norm": 0.8882210442812246, "learning_rate": 3.694538310991422e-06, "loss": 0.0543, "step": 11342 }, { "epoch": 7.257197696737044, "grad_norm": 0.8260909369830114, "learning_rate": 3.6929301404489714e-06, "loss": 0.056, "step": 11343 }, { "epoch": 7.257837492002559, "grad_norm": 0.9166746725058478, "learning_rate": 3.6913222407208783e-06, "loss": 0.0595, "step": 11344 }, { "epoch": 7.258477287268074, "grad_norm": 0.9312974017065485, "learning_rate": 3.689714611876187e-06, "loss": 0.0605, "step": 11345 }, { "epoch": 7.25911708253359, "grad_norm": 0.8211482587144009, "learning_rate": 3.688107253983927e-06, "loss": 0.0489, "step": 11346 }, { "epoch": 7.259756877799104, "grad_norm": 0.8159455915337578, "learning_rate": 3.6865001671131095e-06, "loss": 0.0544, "step": 11347 }, { "epoch": 7.260396673064619, "grad_norm": 0.8684145327843052, "learning_rate": 3.6848933513327478e-06, "loss": 0.0544, "step": 11348 }, { "epoch": 7.2610364683301345, "grad_norm": 0.8348745838075337, "learning_rate": 3.6832868067118323e-06, "loss": 0.052, "step": 11349 }, { "epoch": 7.26167626359565, "grad_norm": 1.0447343767650459, "learning_rate": 3.681680533319344e-06, "loss": 0.0597, "step": 11350 }, { "epoch": 7.262316058861164, "grad_norm": 0.7325915530155185, "learning_rate": 3.6800745312242534e-06, "loss": 0.0495, "step": 11351 }, { "epoch": 7.262955854126679, "grad_norm": 0.8005347104165431, "learning_rate": 3.6784688004955214e-06, "loss": 0.05, "step": 11352 }, { "epoch": 7.263595649392195, "grad_norm": 0.8414390332528405, "learning_rate": 3.676863341202094e-06, "loss": 0.0542, "step": 11353 }, { "epoch": 7.264235444657709, "grad_norm": 0.7203889471866955, "learning_rate": 3.6752581534129026e-06, "loss": 0.0443, "step": 11354 }, { "epoch": 7.264875239923224, "grad_norm": 0.7936874118452015, "learning_rate": 3.673653237196879e-06, "loss": 0.0506, "step": 11355 }, { "epoch": 7.26551503518874, "grad_norm": 0.8966715132534558, "learning_rate": 3.672048592622931e-06, "loss": 0.0509, "step": 11356 }, { "epoch": 7.266154830454255, "grad_norm": 0.876025419178839, "learning_rate": 3.6704442197599576e-06, "loss": 0.057, "step": 11357 }, { "epoch": 7.266794625719769, "grad_norm": 0.8587261363775166, "learning_rate": 3.668840118676853e-06, "loss": 0.0518, "step": 11358 }, { "epoch": 7.267434420985285, "grad_norm": 0.8288339370670327, "learning_rate": 3.6672362894424918e-06, "loss": 0.0502, "step": 11359 }, { "epoch": 7.2680742162508, "grad_norm": 0.9113303019332968, "learning_rate": 3.665632732125738e-06, "loss": 0.0485, "step": 11360 }, { "epoch": 7.268714011516315, "grad_norm": 0.8691357969917398, "learning_rate": 3.664029446795444e-06, "loss": 0.0589, "step": 11361 }, { "epoch": 7.2693538067818295, "grad_norm": 0.7604613161778342, "learning_rate": 3.6624264335204585e-06, "loss": 0.0492, "step": 11362 }, { "epoch": 7.269993602047345, "grad_norm": 0.7794441550773368, "learning_rate": 3.6608236923696085e-06, "loss": 0.0513, "step": 11363 }, { "epoch": 7.27063339731286, "grad_norm": 0.7994244813378358, "learning_rate": 3.6592212234117085e-06, "loss": 0.0542, "step": 11364 }, { "epoch": 7.271273192578375, "grad_norm": 0.7870675617837429, "learning_rate": 3.657619026715573e-06, "loss": 0.0497, "step": 11365 }, { "epoch": 7.27191298784389, "grad_norm": 0.8866612724987669, "learning_rate": 3.6560171023499947e-06, "loss": 0.0554, "step": 11366 }, { "epoch": 7.272552783109405, "grad_norm": 0.8349417031590032, "learning_rate": 3.6544154503837572e-06, "loss": 0.0569, "step": 11367 }, { "epoch": 7.27319257837492, "grad_norm": 0.7860798975340342, "learning_rate": 3.652814070885632e-06, "loss": 0.0501, "step": 11368 }, { "epoch": 7.273832373640435, "grad_norm": 0.8076411219555796, "learning_rate": 3.6512129639243798e-06, "loss": 0.052, "step": 11369 }, { "epoch": 7.27447216890595, "grad_norm": 0.8432642398146695, "learning_rate": 3.6496121295687505e-06, "loss": 0.0544, "step": 11370 }, { "epoch": 7.275111964171465, "grad_norm": 0.8668912278071343, "learning_rate": 3.648011567887475e-06, "loss": 0.0559, "step": 11371 }, { "epoch": 7.2757517594369805, "grad_norm": 0.837686178975033, "learning_rate": 3.6464112789492867e-06, "loss": 0.0545, "step": 11372 }, { "epoch": 7.276391554702495, "grad_norm": 0.8630347275269797, "learning_rate": 3.644811262822897e-06, "loss": 0.058, "step": 11373 }, { "epoch": 7.27703134996801, "grad_norm": 0.8113956331180943, "learning_rate": 3.6432115195770023e-06, "loss": 0.0561, "step": 11374 }, { "epoch": 7.277671145233525, "grad_norm": 0.8990434128952293, "learning_rate": 3.641612049280302e-06, "loss": 0.0551, "step": 11375 }, { "epoch": 7.278310940499041, "grad_norm": 0.8975461849273594, "learning_rate": 3.640012852001468e-06, "loss": 0.0522, "step": 11376 }, { "epoch": 7.278950735764555, "grad_norm": 0.7905327029638737, "learning_rate": 3.638413927809166e-06, "loss": 0.0574, "step": 11377 }, { "epoch": 7.27959053103007, "grad_norm": 0.9084594745936553, "learning_rate": 3.6368152767720565e-06, "loss": 0.0548, "step": 11378 }, { "epoch": 7.280230326295586, "grad_norm": 0.855613145178828, "learning_rate": 3.63521689895878e-06, "loss": 0.0517, "step": 11379 }, { "epoch": 7.280870121561101, "grad_norm": 0.8565695683984416, "learning_rate": 3.6336187944379686e-06, "loss": 0.051, "step": 11380 }, { "epoch": 7.281509916826615, "grad_norm": 0.8516569702015335, "learning_rate": 3.6320209632782398e-06, "loss": 0.0551, "step": 11381 }, { "epoch": 7.282149712092131, "grad_norm": 0.9153656641832827, "learning_rate": 3.6304234055482036e-06, "loss": 0.0539, "step": 11382 }, { "epoch": 7.282789507357646, "grad_norm": 0.7635398461784246, "learning_rate": 3.6288261213164567e-06, "loss": 0.0477, "step": 11383 }, { "epoch": 7.28342930262316, "grad_norm": 0.7506855515539005, "learning_rate": 3.6272291106515777e-06, "loss": 0.0531, "step": 11384 }, { "epoch": 7.2840690978886755, "grad_norm": 0.8417816160109716, "learning_rate": 3.6256323736221487e-06, "loss": 0.0562, "step": 11385 }, { "epoch": 7.284708893154191, "grad_norm": 0.7959779080295057, "learning_rate": 3.624035910296727e-06, "loss": 0.0548, "step": 11386 }, { "epoch": 7.285348688419706, "grad_norm": 0.7494733201594773, "learning_rate": 3.622439720743858e-06, "loss": 0.0506, "step": 11387 }, { "epoch": 7.2859884836852205, "grad_norm": 0.9532099391065064, "learning_rate": 3.620843805032085e-06, "loss": 0.0608, "step": 11388 }, { "epoch": 7.286628278950736, "grad_norm": 0.7814643776786289, "learning_rate": 3.6192481632299312e-06, "loss": 0.0519, "step": 11389 }, { "epoch": 7.287268074216251, "grad_norm": 0.8613730066116189, "learning_rate": 3.6176527954059114e-06, "loss": 0.0588, "step": 11390 }, { "epoch": 7.287907869481765, "grad_norm": 0.8706769579822887, "learning_rate": 3.6160577016285236e-06, "loss": 0.0538, "step": 11391 }, { "epoch": 7.288547664747281, "grad_norm": 0.9863645008460837, "learning_rate": 3.614462881966266e-06, "loss": 0.0687, "step": 11392 }, { "epoch": 7.289187460012796, "grad_norm": 0.8738235702273291, "learning_rate": 3.612868336487613e-06, "loss": 0.0523, "step": 11393 }, { "epoch": 7.289827255278311, "grad_norm": 0.754871712007604, "learning_rate": 3.611274065261028e-06, "loss": 0.0479, "step": 11394 }, { "epoch": 7.290467050543826, "grad_norm": 0.8618758294489756, "learning_rate": 3.6096800683549738e-06, "loss": 0.0556, "step": 11395 }, { "epoch": 7.291106845809341, "grad_norm": 0.7970291483213383, "learning_rate": 3.608086345837889e-06, "loss": 0.0535, "step": 11396 }, { "epoch": 7.291746641074856, "grad_norm": 0.8191220168316358, "learning_rate": 3.6064928977782078e-06, "loss": 0.0561, "step": 11397 }, { "epoch": 7.292386436340371, "grad_norm": 0.8325495673696439, "learning_rate": 3.6048997242443472e-06, "loss": 0.0558, "step": 11398 }, { "epoch": 7.293026231605886, "grad_norm": 0.8259415858277748, "learning_rate": 3.6033068253047167e-06, "loss": 0.0471, "step": 11399 }, { "epoch": 7.293666026871401, "grad_norm": 0.8021031195909498, "learning_rate": 3.601714201027712e-06, "loss": 0.0513, "step": 11400 }, { "epoch": 7.294305822136916, "grad_norm": 0.8811262334222096, "learning_rate": 3.600121851481715e-06, "loss": 0.0528, "step": 11401 }, { "epoch": 7.294945617402432, "grad_norm": 0.8735216860308794, "learning_rate": 3.5985297767351047e-06, "loss": 0.0606, "step": 11402 }, { "epoch": 7.295585412667946, "grad_norm": 0.918226735644631, "learning_rate": 3.596937976856237e-06, "loss": 0.0524, "step": 11403 }, { "epoch": 7.296225207933461, "grad_norm": 0.8926583987936717, "learning_rate": 3.595346451913461e-06, "loss": 0.0552, "step": 11404 }, { "epoch": 7.296865003198977, "grad_norm": 0.9239433987384595, "learning_rate": 3.5937552019751177e-06, "loss": 0.0563, "step": 11405 }, { "epoch": 7.297504798464491, "grad_norm": 0.9428498859197258, "learning_rate": 3.5921642271095303e-06, "loss": 0.0553, "step": 11406 }, { "epoch": 7.298144593730006, "grad_norm": 0.9275153285861588, "learning_rate": 3.5905735273850127e-06, "loss": 0.0538, "step": 11407 }, { "epoch": 7.2987843889955215, "grad_norm": 0.8916580537850566, "learning_rate": 3.5889831028698628e-06, "loss": 0.0524, "step": 11408 }, { "epoch": 7.299424184261037, "grad_norm": 0.956716378815456, "learning_rate": 3.587392953632377e-06, "loss": 0.0535, "step": 11409 }, { "epoch": 7.300063979526551, "grad_norm": 0.8537428651538674, "learning_rate": 3.585803079740832e-06, "loss": 0.06, "step": 11410 }, { "epoch": 7.3007037747920664, "grad_norm": 0.7799533758160271, "learning_rate": 3.584213481263491e-06, "loss": 0.0484, "step": 11411 }, { "epoch": 7.301343570057582, "grad_norm": 0.8296796142596077, "learning_rate": 3.5826241582686115e-06, "loss": 0.0571, "step": 11412 }, { "epoch": 7.301983365323097, "grad_norm": 0.8191966844678843, "learning_rate": 3.5810351108244347e-06, "loss": 0.0517, "step": 11413 }, { "epoch": 7.302623160588611, "grad_norm": 0.8091339474926222, "learning_rate": 3.5794463389991886e-06, "loss": 0.0552, "step": 11414 }, { "epoch": 7.303262955854127, "grad_norm": 0.9124403479673474, "learning_rate": 3.5778578428610978e-06, "loss": 0.0601, "step": 11415 }, { "epoch": 7.303902751119642, "grad_norm": 0.81474341449964, "learning_rate": 3.5762696224783687e-06, "loss": 0.0554, "step": 11416 }, { "epoch": 7.304542546385157, "grad_norm": 0.776592373894537, "learning_rate": 3.574681677919194e-06, "loss": 0.0566, "step": 11417 }, { "epoch": 7.305182341650672, "grad_norm": 0.87354968024761, "learning_rate": 3.573094009251755e-06, "loss": 0.0579, "step": 11418 }, { "epoch": 7.305822136916187, "grad_norm": 0.7853197159152925, "learning_rate": 3.5715066165442304e-06, "loss": 0.0497, "step": 11419 }, { "epoch": 7.306461932181702, "grad_norm": 0.7826347900960957, "learning_rate": 3.5699194998647758e-06, "loss": 0.0539, "step": 11420 }, { "epoch": 7.3071017274472165, "grad_norm": 0.8120621057995483, "learning_rate": 3.568332659281537e-06, "loss": 0.0562, "step": 11421 }, { "epoch": 7.307741522712732, "grad_norm": 0.7952271234929651, "learning_rate": 3.5667460948626554e-06, "loss": 0.0495, "step": 11422 }, { "epoch": 7.308381317978247, "grad_norm": 0.7527793383231746, "learning_rate": 3.5651598066762528e-06, "loss": 0.0515, "step": 11423 }, { "epoch": 7.309021113243762, "grad_norm": 0.740466212575809, "learning_rate": 3.563573794790438e-06, "loss": 0.0489, "step": 11424 }, { "epoch": 7.309660908509277, "grad_norm": 0.8967554113962816, "learning_rate": 3.5619880592733193e-06, "loss": 0.0592, "step": 11425 }, { "epoch": 7.310300703774792, "grad_norm": 0.8757384204361663, "learning_rate": 3.5604026001929803e-06, "loss": 0.0576, "step": 11426 }, { "epoch": 7.310940499040307, "grad_norm": 0.8445731644562889, "learning_rate": 3.558817417617498e-06, "loss": 0.0542, "step": 11427 }, { "epoch": 7.311580294305822, "grad_norm": 0.8205595945328275, "learning_rate": 3.5572325116149386e-06, "loss": 0.0524, "step": 11428 }, { "epoch": 7.312220089571337, "grad_norm": 0.8627487172212697, "learning_rate": 3.5556478822533536e-06, "loss": 0.0533, "step": 11429 }, { "epoch": 7.312859884836852, "grad_norm": 0.8091128139600179, "learning_rate": 3.5540635296007854e-06, "loss": 0.0537, "step": 11430 }, { "epoch": 7.3134996801023675, "grad_norm": 0.7904202495277041, "learning_rate": 3.5524794537252594e-06, "loss": 0.0554, "step": 11431 }, { "epoch": 7.314139475367882, "grad_norm": 0.8473483239890472, "learning_rate": 3.5508956546947994e-06, "loss": 0.0527, "step": 11432 }, { "epoch": 7.314779270633397, "grad_norm": 1.5862725088639678, "learning_rate": 3.5493121325774084e-06, "loss": 0.059, "step": 11433 }, { "epoch": 7.315419065898912, "grad_norm": 0.8274454913710115, "learning_rate": 3.5477288874410765e-06, "loss": 0.0573, "step": 11434 }, { "epoch": 7.316058861164428, "grad_norm": 0.7287275518405012, "learning_rate": 3.5461459193537905e-06, "loss": 0.0546, "step": 11435 }, { "epoch": 7.316698656429942, "grad_norm": 0.8960706003723679, "learning_rate": 3.5445632283835195e-06, "loss": 0.0602, "step": 11436 }, { "epoch": 7.317338451695457, "grad_norm": 0.8298641619866761, "learning_rate": 3.542980814598219e-06, "loss": 0.0576, "step": 11437 }, { "epoch": 7.317978246960973, "grad_norm": 0.7847844656804508, "learning_rate": 3.5413986780658337e-06, "loss": 0.0512, "step": 11438 }, { "epoch": 7.318618042226488, "grad_norm": 0.8585283110037084, "learning_rate": 3.5398168188543037e-06, "loss": 0.0523, "step": 11439 }, { "epoch": 7.319257837492002, "grad_norm": 0.8816940619258633, "learning_rate": 3.5382352370315474e-06, "loss": 0.0562, "step": 11440 }, { "epoch": 7.319897632757518, "grad_norm": 0.7667055033449264, "learning_rate": 3.536653932665476e-06, "loss": 0.0472, "step": 11441 }, { "epoch": 7.320537428023033, "grad_norm": 0.9425898767015073, "learning_rate": 3.535072905823986e-06, "loss": 0.063, "step": 11442 }, { "epoch": 7.321177223288547, "grad_norm": 0.7928869909804047, "learning_rate": 3.5334921565749656e-06, "loss": 0.0564, "step": 11443 }, { "epoch": 7.3218170185540625, "grad_norm": 0.9007846019325569, "learning_rate": 3.5319116849862865e-06, "loss": 0.0542, "step": 11444 }, { "epoch": 7.322456813819578, "grad_norm": 0.7834607211975452, "learning_rate": 3.530331491125817e-06, "loss": 0.0508, "step": 11445 }, { "epoch": 7.323096609085093, "grad_norm": 0.8596576352846949, "learning_rate": 3.528751575061404e-06, "loss": 0.0499, "step": 11446 }, { "epoch": 7.3237364043506075, "grad_norm": 0.7990698706006455, "learning_rate": 3.527171936860887e-06, "loss": 0.0506, "step": 11447 }, { "epoch": 7.324376199616123, "grad_norm": 0.8823235735434439, "learning_rate": 3.5255925765920895e-06, "loss": 0.0502, "step": 11448 }, { "epoch": 7.325015994881638, "grad_norm": 0.7650508131978122, "learning_rate": 3.524013494322833e-06, "loss": 0.0492, "step": 11449 }, { "epoch": 7.325655790147153, "grad_norm": 0.7762428838759818, "learning_rate": 3.522434690120917e-06, "loss": 0.0504, "step": 11450 }, { "epoch": 7.326295585412668, "grad_norm": 0.9075375196798073, "learning_rate": 3.5208561640541308e-06, "loss": 0.057, "step": 11451 }, { "epoch": 7.326935380678183, "grad_norm": 0.806560725882619, "learning_rate": 3.519277916190257e-06, "loss": 0.0579, "step": 11452 }, { "epoch": 7.327575175943698, "grad_norm": 0.7681311560103034, "learning_rate": 3.517699946597062e-06, "loss": 0.0494, "step": 11453 }, { "epoch": 7.3282149712092135, "grad_norm": 0.852591551764322, "learning_rate": 3.516122255342297e-06, "loss": 0.052, "step": 11454 }, { "epoch": 7.328854766474728, "grad_norm": 0.91321349750152, "learning_rate": 3.514544842493711e-06, "loss": 0.0572, "step": 11455 }, { "epoch": 7.329494561740243, "grad_norm": 0.8458717318157849, "learning_rate": 3.512967708119033e-06, "loss": 0.0576, "step": 11456 }, { "epoch": 7.330134357005758, "grad_norm": 0.7124105605414477, "learning_rate": 3.511390852285982e-06, "loss": 0.0479, "step": 11457 }, { "epoch": 7.330774152271273, "grad_norm": 0.823596574602486, "learning_rate": 3.5098142750622654e-06, "loss": 0.0506, "step": 11458 }, { "epoch": 7.331413947536788, "grad_norm": 0.7696941766606283, "learning_rate": 3.5082379765155795e-06, "loss": 0.052, "step": 11459 }, { "epoch": 7.332053742802303, "grad_norm": 0.8006835565838715, "learning_rate": 3.506661956713606e-06, "loss": 0.0543, "step": 11460 }, { "epoch": 7.332693538067819, "grad_norm": 0.8025466998605265, "learning_rate": 3.505086215724015e-06, "loss": 0.0518, "step": 11461 }, { "epoch": 7.333333333333333, "grad_norm": 0.8023621611337212, "learning_rate": 3.5035107536144707e-06, "loss": 0.0526, "step": 11462 }, { "epoch": 7.333973128598848, "grad_norm": 0.8452564958829287, "learning_rate": 3.501935570452619e-06, "loss": 0.0547, "step": 11463 }, { "epoch": 7.334612923864364, "grad_norm": 0.8762136729372869, "learning_rate": 3.500360666306091e-06, "loss": 0.0552, "step": 11464 }, { "epoch": 7.335252719129878, "grad_norm": 0.7845893740892202, "learning_rate": 3.498786041242517e-06, "loss": 0.0499, "step": 11465 }, { "epoch": 7.335892514395393, "grad_norm": 0.81136233841122, "learning_rate": 3.4972116953295067e-06, "loss": 0.0511, "step": 11466 }, { "epoch": 7.3365323096609085, "grad_norm": 0.946448273997158, "learning_rate": 3.495637628634657e-06, "loss": 0.0553, "step": 11467 }, { "epoch": 7.337172104926424, "grad_norm": 0.9580905453778952, "learning_rate": 3.4940638412255557e-06, "loss": 0.0617, "step": 11468 }, { "epoch": 7.337811900191938, "grad_norm": 0.839239532757513, "learning_rate": 3.4924903331697822e-06, "loss": 0.0478, "step": 11469 }, { "epoch": 7.3384516954574535, "grad_norm": 0.9336412529031142, "learning_rate": 3.4909171045348976e-06, "loss": 0.0629, "step": 11470 }, { "epoch": 7.339091490722969, "grad_norm": 0.8004368799292048, "learning_rate": 3.4893441553884544e-06, "loss": 0.0531, "step": 11471 }, { "epoch": 7.339731285988484, "grad_norm": 0.8531776662139934, "learning_rate": 3.4877714857979915e-06, "loss": 0.052, "step": 11472 }, { "epoch": 7.340371081253998, "grad_norm": 0.8039375080110257, "learning_rate": 3.4861990958310367e-06, "loss": 0.0527, "step": 11473 }, { "epoch": 7.341010876519514, "grad_norm": 0.9200545502303968, "learning_rate": 3.484626985555105e-06, "loss": 0.0561, "step": 11474 }, { "epoch": 7.341650671785029, "grad_norm": 0.8894389661023934, "learning_rate": 3.483055155037698e-06, "loss": 0.0613, "step": 11475 }, { "epoch": 7.342290467050544, "grad_norm": 0.8369135788821592, "learning_rate": 3.481483604346313e-06, "loss": 0.056, "step": 11476 }, { "epoch": 7.342930262316059, "grad_norm": 0.8912880328231064, "learning_rate": 3.479912333548425e-06, "loss": 0.056, "step": 11477 }, { "epoch": 7.343570057581574, "grad_norm": 0.8474105655146379, "learning_rate": 3.4783413427115022e-06, "loss": 0.0525, "step": 11478 }, { "epoch": 7.344209852847089, "grad_norm": 0.8747003849346732, "learning_rate": 3.4767706319030025e-06, "loss": 0.0601, "step": 11479 }, { "epoch": 7.3448496481126035, "grad_norm": 0.8492477405026266, "learning_rate": 3.4752002011903685e-06, "loss": 0.0554, "step": 11480 }, { "epoch": 7.345489443378119, "grad_norm": 0.776552449161875, "learning_rate": 3.4736300506410282e-06, "loss": 0.0498, "step": 11481 }, { "epoch": 7.346129238643634, "grad_norm": 0.8411266148708009, "learning_rate": 3.4720601803224076e-06, "loss": 0.0537, "step": 11482 }, { "epoch": 7.346769033909149, "grad_norm": 0.7703412370093425, "learning_rate": 3.4704905903019092e-06, "loss": 0.052, "step": 11483 }, { "epoch": 7.347408829174664, "grad_norm": 0.8055502535486405, "learning_rate": 3.4689212806469308e-06, "loss": 0.0511, "step": 11484 }, { "epoch": 7.348048624440179, "grad_norm": 0.8183255152606694, "learning_rate": 3.4673522514248516e-06, "loss": 0.0527, "step": 11485 }, { "epoch": 7.348688419705694, "grad_norm": 0.7875393900231246, "learning_rate": 3.4657835027030496e-06, "loss": 0.053, "step": 11486 }, { "epoch": 7.34932821497121, "grad_norm": 0.8763586892986426, "learning_rate": 3.464215034548879e-06, "loss": 0.0517, "step": 11487 }, { "epoch": 7.349968010236724, "grad_norm": 0.8287045061309868, "learning_rate": 3.4626468470296892e-06, "loss": 0.0588, "step": 11488 }, { "epoch": 7.350607805502239, "grad_norm": 0.8449859595154786, "learning_rate": 3.4610789402128154e-06, "loss": 0.0582, "step": 11489 }, { "epoch": 7.3512476007677545, "grad_norm": 0.8115421193815641, "learning_rate": 3.45951131416558e-06, "loss": 0.0514, "step": 11490 }, { "epoch": 7.35188739603327, "grad_norm": 0.7652912403372233, "learning_rate": 3.4579439689552907e-06, "loss": 0.0572, "step": 11491 }, { "epoch": 7.352527191298784, "grad_norm": 1.0380640537591428, "learning_rate": 3.456376904649252e-06, "loss": 0.0543, "step": 11492 }, { "epoch": 7.3531669865642995, "grad_norm": 0.8473043824211679, "learning_rate": 3.4548101213147512e-06, "loss": 0.052, "step": 11493 }, { "epoch": 7.353806781829815, "grad_norm": 0.7915897473790946, "learning_rate": 3.453243619019059e-06, "loss": 0.0534, "step": 11494 }, { "epoch": 7.354446577095329, "grad_norm": 0.7665592626444572, "learning_rate": 3.451677397829437e-06, "loss": 0.0461, "step": 11495 }, { "epoch": 7.355086372360844, "grad_norm": 0.8085573304047052, "learning_rate": 3.4501114578131425e-06, "loss": 0.0471, "step": 11496 }, { "epoch": 7.35572616762636, "grad_norm": 0.790140602446712, "learning_rate": 3.4485457990374104e-06, "loss": 0.0562, "step": 11497 }, { "epoch": 7.356365962891875, "grad_norm": 0.8288219696970349, "learning_rate": 3.4469804215694647e-06, "loss": 0.0599, "step": 11498 }, { "epoch": 7.357005758157389, "grad_norm": 0.8260247221497363, "learning_rate": 3.4454153254765256e-06, "loss": 0.0539, "step": 11499 }, { "epoch": 7.357645553422905, "grad_norm": 0.8399026272584008, "learning_rate": 3.4438505108257936e-06, "loss": 0.0545, "step": 11500 }, { "epoch": 7.35828534868842, "grad_norm": 0.7716764742899386, "learning_rate": 3.4422859776844576e-06, "loss": 0.0457, "step": 11501 }, { "epoch": 7.358925143953935, "grad_norm": 0.8341227560666559, "learning_rate": 3.4407217261196957e-06, "loss": 0.0519, "step": 11502 }, { "epoch": 7.3595649392194495, "grad_norm": 0.8236125531804137, "learning_rate": 3.4391577561986755e-06, "loss": 0.0576, "step": 11503 }, { "epoch": 7.360204734484965, "grad_norm": 0.8448225976145488, "learning_rate": 3.437594067988551e-06, "loss": 0.0523, "step": 11504 }, { "epoch": 7.36084452975048, "grad_norm": 0.8824925941070563, "learning_rate": 3.43603066155646e-06, "loss": 0.0596, "step": 11505 }, { "epoch": 7.3614843250159945, "grad_norm": 0.8157472836523646, "learning_rate": 3.434467536969539e-06, "loss": 0.0504, "step": 11506 }, { "epoch": 7.36212412028151, "grad_norm": 0.7819049074453331, "learning_rate": 3.432904694294903e-06, "loss": 0.0514, "step": 11507 }, { "epoch": 7.362763915547025, "grad_norm": 0.7686691525155173, "learning_rate": 3.431342133599653e-06, "loss": 0.05, "step": 11508 }, { "epoch": 7.36340371081254, "grad_norm": 0.7857633689883758, "learning_rate": 3.429779854950892e-06, "loss": 0.0483, "step": 11509 }, { "epoch": 7.364043506078055, "grad_norm": 0.8240748122766521, "learning_rate": 3.4282178584156956e-06, "loss": 0.058, "step": 11510 }, { "epoch": 7.36468330134357, "grad_norm": 0.9435554250586221, "learning_rate": 3.426656144061131e-06, "loss": 0.0523, "step": 11511 }, { "epoch": 7.365323096609085, "grad_norm": 0.867671912921086, "learning_rate": 3.4250947119542623e-06, "loss": 0.057, "step": 11512 }, { "epoch": 7.3659628918746005, "grad_norm": 0.8548574250264578, "learning_rate": 3.42353356216213e-06, "loss": 0.0564, "step": 11513 }, { "epoch": 7.366602687140115, "grad_norm": 0.8181563521456591, "learning_rate": 3.4219726947517683e-06, "loss": 0.0565, "step": 11514 }, { "epoch": 7.36724248240563, "grad_norm": 0.8842712591700436, "learning_rate": 3.420412109790194e-06, "loss": 0.0572, "step": 11515 }, { "epoch": 7.3678822776711455, "grad_norm": 0.8161575140218358, "learning_rate": 3.4188518073444245e-06, "loss": 0.0516, "step": 11516 }, { "epoch": 7.36852207293666, "grad_norm": 0.813007089078441, "learning_rate": 3.417291787481449e-06, "loss": 0.0552, "step": 11517 }, { "epoch": 7.369161868202175, "grad_norm": 0.7450256597159519, "learning_rate": 3.4157320502682555e-06, "loss": 0.049, "step": 11518 }, { "epoch": 7.36980166346769, "grad_norm": 0.8000623466272904, "learning_rate": 3.4141725957718165e-06, "loss": 0.0561, "step": 11519 }, { "epoch": 7.370441458733206, "grad_norm": 0.8920360428806626, "learning_rate": 3.412613424059089e-06, "loss": 0.0556, "step": 11520 }, { "epoch": 7.37108125399872, "grad_norm": 0.8663878091827777, "learning_rate": 3.4110545351970205e-06, "loss": 0.0552, "step": 11521 }, { "epoch": 7.371721049264235, "grad_norm": 0.8725605657430493, "learning_rate": 3.409495929252553e-06, "loss": 0.056, "step": 11522 }, { "epoch": 7.372360844529751, "grad_norm": 0.7997357444629078, "learning_rate": 3.407937606292606e-06, "loss": 0.0515, "step": 11523 }, { "epoch": 7.373000639795266, "grad_norm": 0.8180792236358436, "learning_rate": 3.4063795663840925e-06, "loss": 0.0525, "step": 11524 }, { "epoch": 7.37364043506078, "grad_norm": 0.9140465625542312, "learning_rate": 3.404821809593908e-06, "loss": 0.0612, "step": 11525 }, { "epoch": 7.3742802303262955, "grad_norm": 0.9211247890225664, "learning_rate": 3.403264335988946e-06, "loss": 0.0618, "step": 11526 }, { "epoch": 7.374920025591811, "grad_norm": 0.7579897795399163, "learning_rate": 3.4017071456360785e-06, "loss": 0.0486, "step": 11527 }, { "epoch": 7.375559820857326, "grad_norm": 0.7718258060566493, "learning_rate": 3.4001502386021657e-06, "loss": 0.0495, "step": 11528 }, { "epoch": 7.3761996161228405, "grad_norm": 0.8455216776539652, "learning_rate": 3.3985936149540656e-06, "loss": 0.0564, "step": 11529 }, { "epoch": 7.376839411388356, "grad_norm": 0.8263335931098893, "learning_rate": 3.397037274758611e-06, "loss": 0.0509, "step": 11530 }, { "epoch": 7.377479206653871, "grad_norm": 0.816800674667063, "learning_rate": 3.3954812180826312e-06, "loss": 0.0522, "step": 11531 }, { "epoch": 7.378119001919385, "grad_norm": 0.7799404011022701, "learning_rate": 3.3939254449929392e-06, "loss": 0.0533, "step": 11532 }, { "epoch": 7.378758797184901, "grad_norm": 0.7666865582575063, "learning_rate": 3.392369955556337e-06, "loss": 0.0489, "step": 11533 }, { "epoch": 7.379398592450416, "grad_norm": 0.764954830883003, "learning_rate": 3.390814749839615e-06, "loss": 0.0465, "step": 11534 }, { "epoch": 7.380038387715931, "grad_norm": 0.8264084069546587, "learning_rate": 3.3892598279095477e-06, "loss": 0.0494, "step": 11535 }, { "epoch": 7.380678182981446, "grad_norm": 0.8671267318202602, "learning_rate": 3.3877051898329073e-06, "loss": 0.0558, "step": 11536 }, { "epoch": 7.381317978246961, "grad_norm": 0.8159966781715291, "learning_rate": 3.3861508356764426e-06, "loss": 0.0514, "step": 11537 }, { "epoch": 7.381957773512476, "grad_norm": 0.8387453141823835, "learning_rate": 3.3845967655068935e-06, "loss": 0.0512, "step": 11538 }, { "epoch": 7.3825975687779914, "grad_norm": 0.9284890086999585, "learning_rate": 3.383042979390995e-06, "loss": 0.0608, "step": 11539 }, { "epoch": 7.383237364043506, "grad_norm": 0.7977803108801291, "learning_rate": 3.3814894773954598e-06, "loss": 0.0538, "step": 11540 }, { "epoch": 7.383877159309021, "grad_norm": 0.7839167673132483, "learning_rate": 3.37993625958699e-06, "loss": 0.0524, "step": 11541 }, { "epoch": 7.384516954574536, "grad_norm": 0.7761632834709085, "learning_rate": 3.3783833260322842e-06, "loss": 0.0529, "step": 11542 }, { "epoch": 7.385156749840051, "grad_norm": 0.8095412665041818, "learning_rate": 3.3768306767980184e-06, "loss": 0.0524, "step": 11543 }, { "epoch": 7.385796545105566, "grad_norm": 0.7383289357213949, "learning_rate": 3.3752783119508626e-06, "loss": 0.0491, "step": 11544 }, { "epoch": 7.386436340371081, "grad_norm": 0.8431249590987644, "learning_rate": 3.3737262315574683e-06, "loss": 0.0567, "step": 11545 }, { "epoch": 7.387076135636597, "grad_norm": 0.9002235499389687, "learning_rate": 3.3721744356844854e-06, "loss": 0.0584, "step": 11546 }, { "epoch": 7.387715930902111, "grad_norm": 0.8399633634350618, "learning_rate": 3.3706229243985435e-06, "loss": 0.053, "step": 11547 }, { "epoch": 7.388355726167626, "grad_norm": 0.8092163512438665, "learning_rate": 3.3690716977662584e-06, "loss": 0.0595, "step": 11548 }, { "epoch": 7.3889955214331415, "grad_norm": 0.7214487767743981, "learning_rate": 3.367520755854241e-06, "loss": 0.0465, "step": 11549 }, { "epoch": 7.389635316698657, "grad_norm": 0.8695603964848745, "learning_rate": 3.3659700987290834e-06, "loss": 0.0542, "step": 11550 }, { "epoch": 7.390275111964171, "grad_norm": 0.825085341561006, "learning_rate": 3.3644197264573687e-06, "loss": 0.0524, "step": 11551 }, { "epoch": 7.3909149072296865, "grad_norm": 0.9419488653071312, "learning_rate": 3.3628696391056637e-06, "loss": 0.0583, "step": 11552 }, { "epoch": 7.391554702495202, "grad_norm": 0.8207564333615177, "learning_rate": 3.3613198367405332e-06, "loss": 0.0527, "step": 11553 }, { "epoch": 7.392194497760716, "grad_norm": 0.7439107113765433, "learning_rate": 3.359770319428519e-06, "loss": 0.0477, "step": 11554 }, { "epoch": 7.392834293026231, "grad_norm": 0.8189554591757872, "learning_rate": 3.358221087236153e-06, "loss": 0.054, "step": 11555 }, { "epoch": 7.393474088291747, "grad_norm": 0.8794261991023506, "learning_rate": 3.356672140229962e-06, "loss": 0.0543, "step": 11556 }, { "epoch": 7.394113883557262, "grad_norm": 0.9629783865788731, "learning_rate": 3.3551234784764506e-06, "loss": 0.0664, "step": 11557 }, { "epoch": 7.394753678822776, "grad_norm": 0.7869749367511307, "learning_rate": 3.353575102042115e-06, "loss": 0.0533, "step": 11558 }, { "epoch": 7.395393474088292, "grad_norm": 0.7824613487869632, "learning_rate": 3.352027010993443e-06, "loss": 0.0469, "step": 11559 }, { "epoch": 7.396033269353807, "grad_norm": 0.8231367495254057, "learning_rate": 3.350479205396906e-06, "loss": 0.0506, "step": 11560 }, { "epoch": 7.396673064619322, "grad_norm": 0.8202243781330719, "learning_rate": 3.3489316853189637e-06, "loss": 0.052, "step": 11561 }, { "epoch": 7.3973128598848366, "grad_norm": 0.7360311563581782, "learning_rate": 3.3473844508260634e-06, "loss": 0.0508, "step": 11562 }, { "epoch": 7.397952655150352, "grad_norm": 0.8153533203081906, "learning_rate": 3.3458375019846413e-06, "loss": 0.058, "step": 11563 }, { "epoch": 7.398592450415867, "grad_norm": 0.7621258884866047, "learning_rate": 3.3442908388611186e-06, "loss": 0.0523, "step": 11564 }, { "epoch": 7.399232245681382, "grad_norm": 0.9012763966279481, "learning_rate": 3.3427444615219053e-06, "loss": 0.0574, "step": 11565 }, { "epoch": 7.399872040946897, "grad_norm": 0.9390606908913804, "learning_rate": 3.3411983700334062e-06, "loss": 0.0519, "step": 11566 }, { "epoch": 7.400511836212412, "grad_norm": 0.9038978934063584, "learning_rate": 3.3396525644620025e-06, "loss": 0.0548, "step": 11567 }, { "epoch": 7.401151631477927, "grad_norm": 0.8972219661094971, "learning_rate": 3.338107044874068e-06, "loss": 0.0573, "step": 11568 }, { "epoch": 7.401791426743442, "grad_norm": 0.9416064546482541, "learning_rate": 3.336561811335969e-06, "loss": 0.0604, "step": 11569 }, { "epoch": 7.402431222008957, "grad_norm": 0.8542959751882735, "learning_rate": 3.3350168639140524e-06, "loss": 0.0569, "step": 11570 }, { "epoch": 7.403071017274472, "grad_norm": 0.9329644611947897, "learning_rate": 3.333472202674656e-06, "loss": 0.0577, "step": 11571 }, { "epoch": 7.4037108125399875, "grad_norm": 0.756060655967374, "learning_rate": 3.3319278276841016e-06, "loss": 0.0518, "step": 11572 }, { "epoch": 7.404350607805502, "grad_norm": 0.7489966308468021, "learning_rate": 3.330383739008707e-06, "loss": 0.0492, "step": 11573 }, { "epoch": 7.404990403071017, "grad_norm": 0.9191936525521024, "learning_rate": 3.3288399367147707e-06, "loss": 0.0589, "step": 11574 }, { "epoch": 7.4056301983365325, "grad_norm": 0.8287691990815174, "learning_rate": 3.3272964208685764e-06, "loss": 0.05, "step": 11575 }, { "epoch": 7.406269993602048, "grad_norm": 0.862691195975057, "learning_rate": 3.3257531915364083e-06, "loss": 0.0548, "step": 11576 }, { "epoch": 7.406909788867562, "grad_norm": 0.893399375727836, "learning_rate": 3.324210248784525e-06, "loss": 0.0555, "step": 11577 }, { "epoch": 7.407549584133077, "grad_norm": 0.9329852859694551, "learning_rate": 3.3226675926791795e-06, "loss": 0.0578, "step": 11578 }, { "epoch": 7.408189379398593, "grad_norm": 0.875124344118544, "learning_rate": 3.3211252232866075e-06, "loss": 0.0586, "step": 11579 }, { "epoch": 7.408829174664108, "grad_norm": 0.8594054426188266, "learning_rate": 3.3195831406730394e-06, "loss": 0.0487, "step": 11580 }, { "epoch": 7.409468969929622, "grad_norm": 0.8768529555339224, "learning_rate": 3.3180413449046865e-06, "loss": 0.0516, "step": 11581 }, { "epoch": 7.410108765195138, "grad_norm": 0.862309798390574, "learning_rate": 3.3164998360477495e-06, "loss": 0.0568, "step": 11582 }, { "epoch": 7.410748560460653, "grad_norm": 0.7331964871876558, "learning_rate": 3.3149586141684232e-06, "loss": 0.0499, "step": 11583 }, { "epoch": 7.411388355726167, "grad_norm": 0.7773548982500735, "learning_rate": 3.3134176793328833e-06, "loss": 0.0511, "step": 11584 }, { "epoch": 7.4120281509916826, "grad_norm": 0.8221354886950405, "learning_rate": 3.311877031607289e-06, "loss": 0.0538, "step": 11585 }, { "epoch": 7.412667946257198, "grad_norm": 0.7306826032571565, "learning_rate": 3.3103366710578024e-06, "loss": 0.0529, "step": 11586 }, { "epoch": 7.413307741522713, "grad_norm": 0.7908155882263194, "learning_rate": 3.308796597750558e-06, "loss": 0.0548, "step": 11587 }, { "epoch": 7.4139475367882275, "grad_norm": 0.8322439394612124, "learning_rate": 3.3072568117516822e-06, "loss": 0.049, "step": 11588 }, { "epoch": 7.414587332053743, "grad_norm": 0.9132569440159751, "learning_rate": 3.3057173131272967e-06, "loss": 0.0601, "step": 11589 }, { "epoch": 7.415227127319258, "grad_norm": 0.8724405127521335, "learning_rate": 3.3041781019435017e-06, "loss": 0.0501, "step": 11590 }, { "epoch": 7.415866922584772, "grad_norm": 0.869202449966892, "learning_rate": 3.3026391782663867e-06, "loss": 0.0564, "step": 11591 }, { "epoch": 7.416506717850288, "grad_norm": 0.9218660698580691, "learning_rate": 3.3011005421620333e-06, "loss": 0.057, "step": 11592 }, { "epoch": 7.417146513115803, "grad_norm": 1.0799024035209852, "learning_rate": 3.299562193696506e-06, "loss": 0.0569, "step": 11593 }, { "epoch": 7.417786308381318, "grad_norm": 0.9485296965668029, "learning_rate": 3.2980241329358586e-06, "loss": 0.0556, "step": 11594 }, { "epoch": 7.418426103646833, "grad_norm": 0.8997891114403102, "learning_rate": 3.2964863599461306e-06, "loss": 0.055, "step": 11595 }, { "epoch": 7.419065898912348, "grad_norm": 0.9276817503579516, "learning_rate": 3.2949488747933557e-06, "loss": 0.0527, "step": 11596 }, { "epoch": 7.419705694177863, "grad_norm": 0.8292692066275722, "learning_rate": 3.2934116775435487e-06, "loss": 0.0537, "step": 11597 }, { "epoch": 7.4203454894433785, "grad_norm": 0.7835121721222189, "learning_rate": 3.2918747682627116e-06, "loss": 0.0543, "step": 11598 }, { "epoch": 7.420985284708893, "grad_norm": 0.8633962359764342, "learning_rate": 3.290338147016843e-06, "loss": 0.0555, "step": 11599 }, { "epoch": 7.421625079974408, "grad_norm": 0.8609124079382584, "learning_rate": 3.288801813871918e-06, "loss": 0.0594, "step": 11600 }, { "epoch": 7.422264875239923, "grad_norm": 0.8244399078040292, "learning_rate": 3.2872657688939045e-06, "loss": 0.0504, "step": 11601 }, { "epoch": 7.422904670505439, "grad_norm": 0.879296822301209, "learning_rate": 3.2857300121487557e-06, "loss": 0.0544, "step": 11602 }, { "epoch": 7.423544465770953, "grad_norm": 0.8017380268620886, "learning_rate": 3.284194543702418e-06, "loss": 0.0498, "step": 11603 }, { "epoch": 7.424184261036468, "grad_norm": 0.7891024452213843, "learning_rate": 3.2826593636208214e-06, "loss": 0.05, "step": 11604 }, { "epoch": 7.424824056301984, "grad_norm": 0.7479061471216935, "learning_rate": 3.2811244719698787e-06, "loss": 0.0452, "step": 11605 }, { "epoch": 7.425463851567498, "grad_norm": 0.7586522928577748, "learning_rate": 3.2795898688155026e-06, "loss": 0.0518, "step": 11606 }, { "epoch": 7.426103646833013, "grad_norm": 0.8784023354061304, "learning_rate": 3.2780555542235836e-06, "loss": 0.0553, "step": 11607 }, { "epoch": 7.4267434420985285, "grad_norm": 0.8352539912545416, "learning_rate": 3.2765215282600006e-06, "loss": 0.0484, "step": 11608 }, { "epoch": 7.427383237364044, "grad_norm": 0.9279328518147624, "learning_rate": 3.2749877909906236e-06, "loss": 0.0612, "step": 11609 }, { "epoch": 7.428023032629558, "grad_norm": 0.8413377618426744, "learning_rate": 3.273454342481308e-06, "loss": 0.0579, "step": 11610 }, { "epoch": 7.4286628278950735, "grad_norm": 0.8903575878050342, "learning_rate": 3.271921182797897e-06, "loss": 0.0524, "step": 11611 }, { "epoch": 7.429302623160589, "grad_norm": 0.8927274656156997, "learning_rate": 3.2703883120062198e-06, "loss": 0.0544, "step": 11612 }, { "epoch": 7.429942418426104, "grad_norm": 0.8415375286989069, "learning_rate": 3.268855730172099e-06, "loss": 0.0564, "step": 11613 }, { "epoch": 7.430582213691618, "grad_norm": 0.8195614767818196, "learning_rate": 3.2673234373613407e-06, "loss": 0.0525, "step": 11614 }, { "epoch": 7.431222008957134, "grad_norm": 0.7823532408979041, "learning_rate": 3.2657914336397334e-06, "loss": 0.0515, "step": 11615 }, { "epoch": 7.431861804222649, "grad_norm": 0.8651882330214411, "learning_rate": 3.2642597190730662e-06, "loss": 0.0518, "step": 11616 }, { "epoch": 7.432501599488164, "grad_norm": 0.8862243977727678, "learning_rate": 3.2627282937271055e-06, "loss": 0.0553, "step": 11617 }, { "epoch": 7.433141394753679, "grad_norm": 0.8113540719855302, "learning_rate": 3.261197157667604e-06, "loss": 0.0554, "step": 11618 }, { "epoch": 7.433781190019194, "grad_norm": 0.8249473291957664, "learning_rate": 3.259666310960311e-06, "loss": 0.0516, "step": 11619 }, { "epoch": 7.434420985284709, "grad_norm": 0.8486667285079725, "learning_rate": 3.258135753670957e-06, "loss": 0.0566, "step": 11620 }, { "epoch": 7.435060780550224, "grad_norm": 0.8810979053320911, "learning_rate": 3.2566054858652617e-06, "loss": 0.0545, "step": 11621 }, { "epoch": 7.435700575815739, "grad_norm": 0.8424017058519214, "learning_rate": 3.25507550760893e-06, "loss": 0.0561, "step": 11622 }, { "epoch": 7.436340371081254, "grad_norm": 0.922022094922077, "learning_rate": 3.2535458189676583e-06, "loss": 0.0577, "step": 11623 }, { "epoch": 7.436980166346769, "grad_norm": 0.7543675742188952, "learning_rate": 3.2520164200071293e-06, "loss": 0.0483, "step": 11624 }, { "epoch": 7.437619961612284, "grad_norm": 0.7503132340652316, "learning_rate": 3.250487310793007e-06, "loss": 0.0512, "step": 11625 }, { "epoch": 7.438259756877799, "grad_norm": 0.8341450583757836, "learning_rate": 3.2489584913909566e-06, "loss": 0.0529, "step": 11626 }, { "epoch": 7.438899552143314, "grad_norm": 0.9053071307910066, "learning_rate": 3.2474299618666195e-06, "loss": 0.0572, "step": 11627 }, { "epoch": 7.439539347408829, "grad_norm": 0.7936711359952678, "learning_rate": 3.245901722285627e-06, "loss": 0.0481, "step": 11628 }, { "epoch": 7.440179142674344, "grad_norm": 0.9055185514037024, "learning_rate": 3.244373772713598e-06, "loss": 0.0519, "step": 11629 }, { "epoch": 7.440818937939859, "grad_norm": 0.8662357153961759, "learning_rate": 3.2428461132161437e-06, "loss": 0.0554, "step": 11630 }, { "epoch": 7.4414587332053745, "grad_norm": 0.9504309405094905, "learning_rate": 3.241318743858859e-06, "loss": 0.0571, "step": 11631 }, { "epoch": 7.442098528470889, "grad_norm": 0.9332094170291699, "learning_rate": 3.23979166470732e-06, "loss": 0.0593, "step": 11632 }, { "epoch": 7.442738323736404, "grad_norm": 0.8248093893315, "learning_rate": 3.238264875827105e-06, "loss": 0.051, "step": 11633 }, { "epoch": 7.4433781190019195, "grad_norm": 0.9146041496352496, "learning_rate": 3.2367383772837692e-06, "loss": 0.0566, "step": 11634 }, { "epoch": 7.444017914267435, "grad_norm": 0.8499831310242353, "learning_rate": 3.2352121691428516e-06, "loss": 0.0549, "step": 11635 }, { "epoch": 7.444657709532949, "grad_norm": 0.8235994915364526, "learning_rate": 3.2336862514698943e-06, "loss": 0.0547, "step": 11636 }, { "epoch": 7.445297504798464, "grad_norm": 0.8935704935703316, "learning_rate": 3.2321606243304128e-06, "loss": 0.056, "step": 11637 }, { "epoch": 7.44593730006398, "grad_norm": 0.7477236494271899, "learning_rate": 3.230635287789915e-06, "loss": 0.044, "step": 11638 }, { "epoch": 7.446577095329495, "grad_norm": 0.9139863355333993, "learning_rate": 3.2291102419138964e-06, "loss": 0.0524, "step": 11639 }, { "epoch": 7.447216890595009, "grad_norm": 0.8864214048052034, "learning_rate": 3.227585486767839e-06, "loss": 0.0527, "step": 11640 }, { "epoch": 7.447856685860525, "grad_norm": 0.794238711274903, "learning_rate": 3.226061022417214e-06, "loss": 0.055, "step": 11641 }, { "epoch": 7.44849648112604, "grad_norm": 0.9368440129442936, "learning_rate": 3.2245368489274753e-06, "loss": 0.0605, "step": 11642 }, { "epoch": 7.449136276391554, "grad_norm": 0.8678481507155762, "learning_rate": 3.223012966364074e-06, "loss": 0.0541, "step": 11643 }, { "epoch": 7.44977607165707, "grad_norm": 0.8190116016713643, "learning_rate": 3.2214893747924423e-06, "loss": 0.0509, "step": 11644 }, { "epoch": 7.450415866922585, "grad_norm": 0.8009770617996821, "learning_rate": 3.219966074277995e-06, "loss": 0.0517, "step": 11645 }, { "epoch": 7.4510556621881, "grad_norm": 0.847198287649013, "learning_rate": 3.2184430648861463e-06, "loss": 0.0486, "step": 11646 }, { "epoch": 7.4516954574536145, "grad_norm": 0.7773581871729852, "learning_rate": 3.2169203466822897e-06, "loss": 0.0519, "step": 11647 }, { "epoch": 7.45233525271913, "grad_norm": 0.8157591226121577, "learning_rate": 3.2153979197318076e-06, "loss": 0.051, "step": 11648 }, { "epoch": 7.452975047984645, "grad_norm": 0.946411156726647, "learning_rate": 3.2138757841000656e-06, "loss": 0.0559, "step": 11649 }, { "epoch": 7.45361484325016, "grad_norm": 0.7576771709170034, "learning_rate": 3.2123539398524296e-06, "loss": 0.0457, "step": 11650 }, { "epoch": 7.454254638515675, "grad_norm": 0.7845152883433888, "learning_rate": 3.210832387054241e-06, "loss": 0.0473, "step": 11651 }, { "epoch": 7.45489443378119, "grad_norm": 0.8340258931087927, "learning_rate": 3.209311125770833e-06, "loss": 0.0522, "step": 11652 }, { "epoch": 7.455534229046705, "grad_norm": 0.744332089134898, "learning_rate": 3.207790156067525e-06, "loss": 0.0467, "step": 11653 }, { "epoch": 7.4561740243122205, "grad_norm": 0.8739376349725183, "learning_rate": 3.206269478009626e-06, "loss": 0.0512, "step": 11654 }, { "epoch": 7.456813819577735, "grad_norm": 0.9236401203423159, "learning_rate": 3.204749091662427e-06, "loss": 0.0519, "step": 11655 }, { "epoch": 7.45745361484325, "grad_norm": 0.9272645973127313, "learning_rate": 3.203228997091218e-06, "loss": 0.0593, "step": 11656 }, { "epoch": 7.4580934101087655, "grad_norm": 0.8733683377049756, "learning_rate": 3.2017091943612645e-06, "loss": 0.0541, "step": 11657 }, { "epoch": 7.45873320537428, "grad_norm": 0.8568458957397217, "learning_rate": 3.2001896835378255e-06, "loss": 0.0568, "step": 11658 }, { "epoch": 7.459373000639795, "grad_norm": 0.8941443452286391, "learning_rate": 3.198670464686141e-06, "loss": 0.0514, "step": 11659 }, { "epoch": 7.46001279590531, "grad_norm": 0.8117916443402261, "learning_rate": 3.1971515378714537e-06, "loss": 0.0491, "step": 11660 }, { "epoch": 7.460652591170826, "grad_norm": 0.8522216830566403, "learning_rate": 3.195632903158976e-06, "loss": 0.0568, "step": 11661 }, { "epoch": 7.46129238643634, "grad_norm": 0.8015313919335801, "learning_rate": 3.1941145606139146e-06, "loss": 0.0523, "step": 11662 }, { "epoch": 7.461932181701855, "grad_norm": 0.8782472070135916, "learning_rate": 3.1925965103014713e-06, "loss": 0.0494, "step": 11663 }, { "epoch": 7.462571976967371, "grad_norm": 0.8812323080663296, "learning_rate": 3.191078752286824e-06, "loss": 0.0595, "step": 11664 }, { "epoch": 7.463211772232885, "grad_norm": 0.849640654285512, "learning_rate": 3.189561286635139e-06, "loss": 0.0528, "step": 11665 }, { "epoch": 7.4638515674984, "grad_norm": 0.8105136502774235, "learning_rate": 3.1880441134115814e-06, "loss": 0.0546, "step": 11666 }, { "epoch": 7.464491362763916, "grad_norm": 0.718767327570397, "learning_rate": 3.1865272326812923e-06, "loss": 0.0476, "step": 11667 }, { "epoch": 7.465131158029431, "grad_norm": 0.8267019367738704, "learning_rate": 3.185010644509403e-06, "loss": 0.0571, "step": 11668 }, { "epoch": 7.465770953294945, "grad_norm": 0.7847900697342759, "learning_rate": 3.1834943489610346e-06, "loss": 0.0536, "step": 11669 }, { "epoch": 7.4664107485604605, "grad_norm": 0.8361537996922778, "learning_rate": 3.1819783461012922e-06, "loss": 0.0567, "step": 11670 }, { "epoch": 7.467050543825976, "grad_norm": 0.7472266948212019, "learning_rate": 3.1804626359952717e-06, "loss": 0.0536, "step": 11671 }, { "epoch": 7.467690339091491, "grad_norm": 0.8500125447443848, "learning_rate": 3.1789472187080515e-06, "loss": 0.0521, "step": 11672 }, { "epoch": 7.468330134357005, "grad_norm": 0.8848686032353607, "learning_rate": 3.177432094304708e-06, "loss": 0.0572, "step": 11673 }, { "epoch": 7.468969929622521, "grad_norm": 0.7865214873384037, "learning_rate": 3.175917262850292e-06, "loss": 0.0548, "step": 11674 }, { "epoch": 7.469609724888036, "grad_norm": 0.8623632704615459, "learning_rate": 3.1744027244098485e-06, "loss": 0.0535, "step": 11675 }, { "epoch": 7.470249520153551, "grad_norm": 0.7974036020866904, "learning_rate": 3.1728884790484117e-06, "loss": 0.0498, "step": 11676 }, { "epoch": 7.470889315419066, "grad_norm": 0.9204952113507793, "learning_rate": 3.1713745268310003e-06, "loss": 0.061, "step": 11677 }, { "epoch": 7.471529110684581, "grad_norm": 0.8921956922525479, "learning_rate": 3.1698608678226193e-06, "loss": 0.0539, "step": 11678 }, { "epoch": 7.472168905950096, "grad_norm": 0.8537844532250518, "learning_rate": 3.168347502088259e-06, "loss": 0.0544, "step": 11679 }, { "epoch": 7.472808701215611, "grad_norm": 0.7964337882843494, "learning_rate": 3.166834429692908e-06, "loss": 0.062, "step": 11680 }, { "epoch": 7.473448496481126, "grad_norm": 0.8741928594784096, "learning_rate": 3.1653216507015317e-06, "loss": 0.0547, "step": 11681 }, { "epoch": 7.474088291746641, "grad_norm": 0.8197243649362362, "learning_rate": 3.1638091651790846e-06, "loss": 0.0506, "step": 11682 }, { "epoch": 7.474728087012156, "grad_norm": 0.8795391805051894, "learning_rate": 3.162296973190513e-06, "loss": 0.0563, "step": 11683 }, { "epoch": 7.475367882277671, "grad_norm": 0.8120584671990554, "learning_rate": 3.1607850748007453e-06, "loss": 0.0488, "step": 11684 }, { "epoch": 7.476007677543186, "grad_norm": 0.9444384777651386, "learning_rate": 3.159273470074696e-06, "loss": 0.0573, "step": 11685 }, { "epoch": 7.476647472808701, "grad_norm": 0.7884889106900793, "learning_rate": 3.15776215907728e-06, "loss": 0.055, "step": 11686 }, { "epoch": 7.477287268074217, "grad_norm": 0.8729288824922562, "learning_rate": 3.1562511418733853e-06, "loss": 0.056, "step": 11687 }, { "epoch": 7.477927063339731, "grad_norm": 0.7951915050181918, "learning_rate": 3.1547404185278916e-06, "loss": 0.0513, "step": 11688 }, { "epoch": 7.478566858605246, "grad_norm": 0.8913456594318303, "learning_rate": 3.1532299891056652e-06, "loss": 0.0588, "step": 11689 }, { "epoch": 7.479206653870762, "grad_norm": 0.8774607627859422, "learning_rate": 3.1517198536715664e-06, "loss": 0.0559, "step": 11690 }, { "epoch": 7.479846449136277, "grad_norm": 0.939168344841745, "learning_rate": 3.1502100122904356e-06, "loss": 0.058, "step": 11691 }, { "epoch": 7.480486244401791, "grad_norm": 0.8108848733860191, "learning_rate": 3.148700465027098e-06, "loss": 0.0529, "step": 11692 }, { "epoch": 7.4811260396673065, "grad_norm": 0.7858978555909266, "learning_rate": 3.1471912119463777e-06, "loss": 0.0502, "step": 11693 }, { "epoch": 7.481765834932822, "grad_norm": 0.816363838115637, "learning_rate": 3.145682253113078e-06, "loss": 0.0539, "step": 11694 }, { "epoch": 7.482405630198336, "grad_norm": 0.7956882206116881, "learning_rate": 3.1441735885919856e-06, "loss": 0.0465, "step": 11695 }, { "epoch": 7.483045425463851, "grad_norm": 0.7885793444182422, "learning_rate": 3.1426652184478867e-06, "loss": 0.0472, "step": 11696 }, { "epoch": 7.483685220729367, "grad_norm": 0.7238051801435494, "learning_rate": 3.1411571427455455e-06, "loss": 0.0439, "step": 11697 }, { "epoch": 7.484325015994882, "grad_norm": 0.78032034042168, "learning_rate": 3.139649361549715e-06, "loss": 0.0498, "step": 11698 }, { "epoch": 7.484964811260396, "grad_norm": 0.8751595151727435, "learning_rate": 3.138141874925138e-06, "loss": 0.0619, "step": 11699 }, { "epoch": 7.485604606525912, "grad_norm": 0.8525246632750658, "learning_rate": 3.1366346829365423e-06, "loss": 0.0575, "step": 11700 }, { "epoch": 7.486244401791427, "grad_norm": 0.8162219485640695, "learning_rate": 3.135127785648644e-06, "loss": 0.0496, "step": 11701 }, { "epoch": 7.486884197056942, "grad_norm": 0.8534425148652256, "learning_rate": 3.133621183126144e-06, "loss": 0.0552, "step": 11702 }, { "epoch": 7.487523992322457, "grad_norm": 0.958876220626428, "learning_rate": 3.1321148754337392e-06, "loss": 0.0559, "step": 11703 }, { "epoch": 7.488163787587972, "grad_norm": 0.802069520314775, "learning_rate": 3.130608862636105e-06, "loss": 0.0516, "step": 11704 }, { "epoch": 7.488803582853487, "grad_norm": 0.7728250749654145, "learning_rate": 3.129103144797906e-06, "loss": 0.0543, "step": 11705 }, { "epoch": 7.4894433781190015, "grad_norm": 0.8140934163245982, "learning_rate": 3.127597721983793e-06, "loss": 0.0495, "step": 11706 }, { "epoch": 7.490083173384517, "grad_norm": 0.8722015590557839, "learning_rate": 3.126092594258412e-06, "loss": 0.054, "step": 11707 }, { "epoch": 7.490722968650032, "grad_norm": 0.7309382073768957, "learning_rate": 3.124587761686386e-06, "loss": 0.0465, "step": 11708 }, { "epoch": 7.491362763915547, "grad_norm": 0.8075699124687769, "learning_rate": 3.1230832243323295e-06, "loss": 0.0496, "step": 11709 }, { "epoch": 7.492002559181062, "grad_norm": 0.846207485929223, "learning_rate": 3.121578982260849e-06, "loss": 0.0562, "step": 11710 }, { "epoch": 7.492642354446577, "grad_norm": 0.9367219189236872, "learning_rate": 3.1200750355365307e-06, "loss": 0.057, "step": 11711 }, { "epoch": 7.493282149712092, "grad_norm": 0.8309773716517985, "learning_rate": 3.1185713842239517e-06, "loss": 0.0522, "step": 11712 }, { "epoch": 7.4939219449776076, "grad_norm": 0.8439486748293898, "learning_rate": 3.1170680283876775e-06, "loss": 0.0518, "step": 11713 }, { "epoch": 7.494561740243122, "grad_norm": 0.8394883609461568, "learning_rate": 3.1155649680922574e-06, "loss": 0.0558, "step": 11714 }, { "epoch": 7.495201535508637, "grad_norm": 0.9360941041850128, "learning_rate": 3.1140622034022316e-06, "loss": 0.0507, "step": 11715 }, { "epoch": 7.4958413307741525, "grad_norm": 0.9259366396950011, "learning_rate": 3.1125597343821223e-06, "loss": 0.0577, "step": 11716 }, { "epoch": 7.496481126039667, "grad_norm": 0.9552896474253799, "learning_rate": 3.111057561096449e-06, "loss": 0.0611, "step": 11717 }, { "epoch": 7.497120921305182, "grad_norm": 0.8006421621725677, "learning_rate": 3.1095556836097105e-06, "loss": 0.0554, "step": 11718 }, { "epoch": 7.497760716570697, "grad_norm": 0.9252431494362937, "learning_rate": 3.1080541019863895e-06, "loss": 0.063, "step": 11719 }, { "epoch": 7.498400511836213, "grad_norm": 0.916347303500444, "learning_rate": 3.1065528162909698e-06, "loss": 0.053, "step": 11720 }, { "epoch": 7.499040307101727, "grad_norm": 0.8516439134380549, "learning_rate": 3.105051826587909e-06, "loss": 0.0494, "step": 11721 }, { "epoch": 7.499680102367242, "grad_norm": 0.8018076453976684, "learning_rate": 3.1035511329416546e-06, "loss": 0.0539, "step": 11722 }, { "epoch": 7.500319897632758, "grad_norm": 0.8479560133131548, "learning_rate": 3.1020507354166496e-06, "loss": 0.0501, "step": 11723 }, { "epoch": 7.500959692898273, "grad_norm": 0.8616567797854272, "learning_rate": 3.100550634077316e-06, "loss": 0.053, "step": 11724 }, { "epoch": 7.501599488163787, "grad_norm": 0.8218165615978248, "learning_rate": 3.099050828988065e-06, "loss": 0.0511, "step": 11725 }, { "epoch": 7.502239283429303, "grad_norm": 0.8589800460890288, "learning_rate": 3.0975513202132924e-06, "loss": 0.0515, "step": 11726 }, { "epoch": 7.502879078694818, "grad_norm": 0.8420833247735906, "learning_rate": 3.09605210781739e-06, "loss": 0.0522, "step": 11727 }, { "epoch": 7.503518873960333, "grad_norm": 0.9077773778507763, "learning_rate": 3.0945531918647297e-06, "loss": 0.0543, "step": 11728 }, { "epoch": 7.5041586692258475, "grad_norm": 0.8390544702836183, "learning_rate": 3.0930545724196714e-06, "loss": 0.0531, "step": 11729 }, { "epoch": 7.504798464491363, "grad_norm": 1.079498158308667, "learning_rate": 3.0915562495465634e-06, "loss": 0.0627, "step": 11730 }, { "epoch": 7.505438259756878, "grad_norm": 0.8740005024733686, "learning_rate": 3.090058223309741e-06, "loss": 0.0591, "step": 11731 }, { "epoch": 7.5060780550223924, "grad_norm": 0.8963724788765871, "learning_rate": 3.0885604937735225e-06, "loss": 0.054, "step": 11732 }, { "epoch": 7.506717850287908, "grad_norm": 0.889020998660982, "learning_rate": 3.0870630610022258e-06, "loss": 0.0496, "step": 11733 }, { "epoch": 7.507357645553423, "grad_norm": 0.9104184109909041, "learning_rate": 3.0855659250601445e-06, "loss": 0.0512, "step": 11734 }, { "epoch": 7.507997440818938, "grad_norm": 0.8019921973353455, "learning_rate": 3.0840690860115617e-06, "loss": 0.0523, "step": 11735 }, { "epoch": 7.508637236084453, "grad_norm": 0.8541101944950652, "learning_rate": 3.0825725439207464e-06, "loss": 0.0484, "step": 11736 }, { "epoch": 7.509277031349968, "grad_norm": 0.748501276527209, "learning_rate": 3.081076298851965e-06, "loss": 0.0468, "step": 11737 }, { "epoch": 7.509916826615483, "grad_norm": 0.9493043690132502, "learning_rate": 3.07958035086946e-06, "loss": 0.0548, "step": 11738 }, { "epoch": 7.510556621880998, "grad_norm": 0.7941854271572498, "learning_rate": 3.0780847000374602e-06, "loss": 0.0544, "step": 11739 }, { "epoch": 7.511196417146513, "grad_norm": 0.8332929364483348, "learning_rate": 3.0765893464201936e-06, "loss": 0.0569, "step": 11740 }, { "epoch": 7.511836212412028, "grad_norm": 0.8381715424177645, "learning_rate": 3.0750942900818647e-06, "loss": 0.0554, "step": 11741 }, { "epoch": 7.512476007677543, "grad_norm": 0.7417621846026089, "learning_rate": 3.0735995310866686e-06, "loss": 0.0475, "step": 11742 }, { "epoch": 7.513115802943059, "grad_norm": 0.7630230090310385, "learning_rate": 3.072105069498786e-06, "loss": 0.05, "step": 11743 }, { "epoch": 7.513755598208573, "grad_norm": 0.9518431016747351, "learning_rate": 3.070610905382391e-06, "loss": 0.0611, "step": 11744 }, { "epoch": 7.514395393474088, "grad_norm": 0.9217590047646043, "learning_rate": 3.0691170388016347e-06, "loss": 0.0521, "step": 11745 }, { "epoch": 7.515035188739604, "grad_norm": 0.8549562191440097, "learning_rate": 3.067623469820662e-06, "loss": 0.0515, "step": 11746 }, { "epoch": 7.515674984005118, "grad_norm": 0.8743450140826247, "learning_rate": 3.0661301985036083e-06, "loss": 0.0547, "step": 11747 }, { "epoch": 7.516314779270633, "grad_norm": 0.7880497953090436, "learning_rate": 3.0646372249145895e-06, "loss": 0.0513, "step": 11748 }, { "epoch": 7.516954574536149, "grad_norm": 0.8727973029713443, "learning_rate": 3.063144549117708e-06, "loss": 0.0522, "step": 11749 }, { "epoch": 7.517594369801664, "grad_norm": 0.8926786779833994, "learning_rate": 3.061652171177063e-06, "loss": 0.0633, "step": 11750 }, { "epoch": 7.518234165067178, "grad_norm": 0.8107053926729397, "learning_rate": 3.060160091156732e-06, "loss": 0.0519, "step": 11751 }, { "epoch": 7.5188739603326935, "grad_norm": 0.8105002064270007, "learning_rate": 3.058668309120777e-06, "loss": 0.0514, "step": 11752 }, { "epoch": 7.519513755598209, "grad_norm": 0.7694848580710537, "learning_rate": 3.0571768251332622e-06, "loss": 0.046, "step": 11753 }, { "epoch": 7.520153550863723, "grad_norm": 0.8490567634062886, "learning_rate": 3.0556856392582234e-06, "loss": 0.0531, "step": 11754 }, { "epoch": 7.520793346129238, "grad_norm": 0.8175248476299394, "learning_rate": 3.0541947515596893e-06, "loss": 0.0537, "step": 11755 }, { "epoch": 7.521433141394754, "grad_norm": 0.8405917924284065, "learning_rate": 3.052704162101675e-06, "loss": 0.0529, "step": 11756 }, { "epoch": 7.522072936660269, "grad_norm": 0.8747364101604241, "learning_rate": 3.051213870948189e-06, "loss": 0.052, "step": 11757 }, { "epoch": 7.522712731925783, "grad_norm": 0.8666821486496189, "learning_rate": 3.0497238781632175e-06, "loss": 0.0495, "step": 11758 }, { "epoch": 7.523352527191299, "grad_norm": 0.7473119482731074, "learning_rate": 3.04823418381074e-06, "loss": 0.0489, "step": 11759 }, { "epoch": 7.523992322456814, "grad_norm": 0.9913588591388791, "learning_rate": 3.0467447879547196e-06, "loss": 0.0605, "step": 11760 }, { "epoch": 7.524632117722329, "grad_norm": 0.8954544863942109, "learning_rate": 3.0452556906591092e-06, "loss": 0.057, "step": 11761 }, { "epoch": 7.525271912987844, "grad_norm": 0.9038437665738431, "learning_rate": 3.043766891987845e-06, "loss": 0.0548, "step": 11762 }, { "epoch": 7.525911708253359, "grad_norm": 0.9194278341964562, "learning_rate": 3.0422783920048603e-06, "loss": 0.0542, "step": 11763 }, { "epoch": 7.526551503518874, "grad_norm": 0.7831018971408713, "learning_rate": 3.0407901907740636e-06, "loss": 0.0452, "step": 11764 }, { "epoch": 7.527191298784389, "grad_norm": 0.7859938294751717, "learning_rate": 3.0393022883593572e-06, "loss": 0.0502, "step": 11765 }, { "epoch": 7.527831094049904, "grad_norm": 0.8816257849076156, "learning_rate": 3.037814684824625e-06, "loss": 0.0541, "step": 11766 }, { "epoch": 7.528470889315419, "grad_norm": 0.8922531961409268, "learning_rate": 3.036327380233748e-06, "loss": 0.0559, "step": 11767 }, { "epoch": 7.529110684580934, "grad_norm": 0.8103175631484711, "learning_rate": 3.034840374650586e-06, "loss": 0.0503, "step": 11768 }, { "epoch": 7.529750479846449, "grad_norm": 0.7615426082483444, "learning_rate": 3.0333536681389864e-06, "loss": 0.0474, "step": 11769 }, { "epoch": 7.530390275111964, "grad_norm": 0.7646212827971658, "learning_rate": 3.0318672607627896e-06, "loss": 0.0465, "step": 11770 }, { "epoch": 7.531030070377479, "grad_norm": 0.8027134259642972, "learning_rate": 3.0303811525858174e-06, "loss": 0.0456, "step": 11771 }, { "epoch": 7.531669865642995, "grad_norm": 0.9300697946131796, "learning_rate": 3.0288953436718806e-06, "loss": 0.0555, "step": 11772 }, { "epoch": 7.532309660908509, "grad_norm": 0.8146895060686044, "learning_rate": 3.027409834084777e-06, "loss": 0.0519, "step": 11773 }, { "epoch": 7.532949456174024, "grad_norm": 0.7999406894482826, "learning_rate": 3.025924623888292e-06, "loss": 0.05, "step": 11774 }, { "epoch": 7.5335892514395395, "grad_norm": 1.0263788056995118, "learning_rate": 3.0244397131461976e-06, "loss": 0.0539, "step": 11775 }, { "epoch": 7.534229046705054, "grad_norm": 0.8761862287440364, "learning_rate": 3.02295510192225e-06, "loss": 0.0511, "step": 11776 }, { "epoch": 7.534868841970569, "grad_norm": 0.8029220225635096, "learning_rate": 3.021470790280203e-06, "loss": 0.0451, "step": 11777 }, { "epoch": 7.535508637236084, "grad_norm": 0.7599823050481016, "learning_rate": 3.0199867782837865e-06, "loss": 0.0507, "step": 11778 }, { "epoch": 7.5361484325016, "grad_norm": 0.8949823026250098, "learning_rate": 3.018503065996717e-06, "loss": 0.0513, "step": 11779 }, { "epoch": 7.536788227767115, "grad_norm": 0.9673399237993511, "learning_rate": 3.017019653482711e-06, "loss": 0.0622, "step": 11780 }, { "epoch": 7.537428023032629, "grad_norm": 0.8878079823192964, "learning_rate": 3.0155365408054595e-06, "loss": 0.0548, "step": 11781 }, { "epoch": 7.538067818298145, "grad_norm": 0.7824055150999253, "learning_rate": 3.014053728028643e-06, "loss": 0.0492, "step": 11782 }, { "epoch": 7.53870761356366, "grad_norm": 0.8834813970185283, "learning_rate": 3.0125712152159303e-06, "loss": 0.054, "step": 11783 }, { "epoch": 7.539347408829174, "grad_norm": 0.8968489894997996, "learning_rate": 3.0110890024309835e-06, "loss": 0.0559, "step": 11784 }, { "epoch": 7.53998720409469, "grad_norm": 0.8585729330298658, "learning_rate": 3.009607089737442e-06, "loss": 0.0593, "step": 11785 }, { "epoch": 7.540626999360205, "grad_norm": 0.8380048915344311, "learning_rate": 3.0081254771989333e-06, "loss": 0.0575, "step": 11786 }, { "epoch": 7.54126679462572, "grad_norm": 0.8807287838384961, "learning_rate": 3.0066441648790832e-06, "loss": 0.0563, "step": 11787 }, { "epoch": 7.5419065898912345, "grad_norm": 0.8108582924149621, "learning_rate": 3.0051631528414927e-06, "loss": 0.049, "step": 11788 }, { "epoch": 7.54254638515675, "grad_norm": 0.8225930237840487, "learning_rate": 3.0036824411497523e-06, "loss": 0.0503, "step": 11789 }, { "epoch": 7.543186180422265, "grad_norm": 0.8257012698295179, "learning_rate": 3.0022020298674438e-06, "loss": 0.0513, "step": 11790 }, { "epoch": 7.5438259756877795, "grad_norm": 0.7767961593067566, "learning_rate": 3.0007219190581305e-06, "loss": 0.0585, "step": 11791 }, { "epoch": 7.544465770953295, "grad_norm": 0.8781553928661291, "learning_rate": 2.9992421087853684e-06, "loss": 0.0567, "step": 11792 }, { "epoch": 7.54510556621881, "grad_norm": 0.809222398197561, "learning_rate": 2.9977625991126934e-06, "loss": 0.0521, "step": 11793 }, { "epoch": 7.545745361484325, "grad_norm": 0.8454015945274603, "learning_rate": 2.9962833901036392e-06, "loss": 0.0536, "step": 11794 }, { "epoch": 7.54638515674984, "grad_norm": 0.8310537162030976, "learning_rate": 2.9948044818217183e-06, "loss": 0.0471, "step": 11795 }, { "epoch": 7.547024952015355, "grad_norm": 0.8839450689847179, "learning_rate": 2.9933258743304282e-06, "loss": 0.0577, "step": 11796 }, { "epoch": 7.54766474728087, "grad_norm": 0.8107403682860505, "learning_rate": 2.9918475676932636e-06, "loss": 0.047, "step": 11797 }, { "epoch": 7.5483045425463855, "grad_norm": 0.8297593676290786, "learning_rate": 2.9903695619736985e-06, "loss": 0.0528, "step": 11798 }, { "epoch": 7.5489443378119, "grad_norm": 0.8644359465593212, "learning_rate": 2.988891857235192e-06, "loss": 0.0523, "step": 11799 }, { "epoch": 7.549584133077415, "grad_norm": 0.8992855166604109, "learning_rate": 2.987414453541201e-06, "loss": 0.0593, "step": 11800 }, { "epoch": 7.55022392834293, "grad_norm": 0.7400014603473778, "learning_rate": 2.9859373509551582e-06, "loss": 0.0507, "step": 11801 }, { "epoch": 7.550863723608446, "grad_norm": 0.929918413086519, "learning_rate": 2.9844605495404887e-06, "loss": 0.0524, "step": 11802 }, { "epoch": 7.55150351887396, "grad_norm": 0.8454559369899294, "learning_rate": 2.9829840493606032e-06, "loss": 0.058, "step": 11803 }, { "epoch": 7.552143314139475, "grad_norm": 0.8311388022602554, "learning_rate": 2.9815078504789006e-06, "loss": 0.0536, "step": 11804 }, { "epoch": 7.552783109404991, "grad_norm": 0.8196636290701719, "learning_rate": 2.9800319529587673e-06, "loss": 0.053, "step": 11805 }, { "epoch": 7.553422904670505, "grad_norm": 0.8206698746536205, "learning_rate": 2.9785563568635702e-06, "loss": 0.0472, "step": 11806 }, { "epoch": 7.55406269993602, "grad_norm": 0.8612256419489974, "learning_rate": 2.9770810622566747e-06, "loss": 0.0554, "step": 11807 }, { "epoch": 7.554702495201536, "grad_norm": 0.8294159632283997, "learning_rate": 2.9756060692014276e-06, "loss": 0.0518, "step": 11808 }, { "epoch": 7.555342290467051, "grad_norm": 0.7970346618316566, "learning_rate": 2.9741313777611568e-06, "loss": 0.0528, "step": 11809 }, { "epoch": 7.555982085732565, "grad_norm": 0.8349598180351572, "learning_rate": 2.9726569879991896e-06, "loss": 0.0537, "step": 11810 }, { "epoch": 7.5566218809980805, "grad_norm": 0.8773860994968786, "learning_rate": 2.9711828999788295e-06, "loss": 0.0633, "step": 11811 }, { "epoch": 7.557261676263596, "grad_norm": 0.8410099171279478, "learning_rate": 2.9697091137633737e-06, "loss": 0.053, "step": 11812 }, { "epoch": 7.557901471529111, "grad_norm": 0.8388491944141474, "learning_rate": 2.9682356294160996e-06, "loss": 0.0554, "step": 11813 }, { "epoch": 7.5585412667946255, "grad_norm": 0.8510260001513655, "learning_rate": 2.966762447000282e-06, "loss": 0.0544, "step": 11814 }, { "epoch": 7.559181062060141, "grad_norm": 0.8323964312029748, "learning_rate": 2.9652895665791735e-06, "loss": 0.0479, "step": 11815 }, { "epoch": 7.559820857325656, "grad_norm": 0.9094794921145829, "learning_rate": 2.963816988216014e-06, "loss": 0.0589, "step": 11816 }, { "epoch": 7.560460652591171, "grad_norm": 0.7988138673124062, "learning_rate": 2.96234471197404e-06, "loss": 0.0466, "step": 11817 }, { "epoch": 7.561100447856686, "grad_norm": 0.860125865359586, "learning_rate": 2.9608727379164647e-06, "loss": 0.0526, "step": 11818 }, { "epoch": 7.561740243122201, "grad_norm": 0.8324401446938481, "learning_rate": 2.959401066106493e-06, "loss": 0.0508, "step": 11819 }, { "epoch": 7.562380038387716, "grad_norm": 0.74788123797334, "learning_rate": 2.9579296966073144e-06, "loss": 0.0519, "step": 11820 }, { "epoch": 7.563019833653231, "grad_norm": 0.9044062544099156, "learning_rate": 2.9564586294821084e-06, "loss": 0.0588, "step": 11821 }, { "epoch": 7.563659628918746, "grad_norm": 0.9891893343803155, "learning_rate": 2.954987864794039e-06, "loss": 0.0652, "step": 11822 }, { "epoch": 7.564299424184261, "grad_norm": 0.7648897797384203, "learning_rate": 2.953517402606256e-06, "loss": 0.0501, "step": 11823 }, { "epoch": 7.564939219449776, "grad_norm": 0.8638627285035693, "learning_rate": 2.952047242981904e-06, "loss": 0.0499, "step": 11824 }, { "epoch": 7.565579014715291, "grad_norm": 0.905542172053769, "learning_rate": 2.950577385984106e-06, "loss": 0.0624, "step": 11825 }, { "epoch": 7.566218809980806, "grad_norm": 0.8812934029081986, "learning_rate": 2.9491078316759712e-06, "loss": 0.0507, "step": 11826 }, { "epoch": 7.566858605246321, "grad_norm": 0.869069449506353, "learning_rate": 2.9476385801206075e-06, "loss": 0.0554, "step": 11827 }, { "epoch": 7.567498400511836, "grad_norm": 0.865238143513887, "learning_rate": 2.9461696313810982e-06, "loss": 0.0465, "step": 11828 }, { "epoch": 7.568138195777351, "grad_norm": 0.7984663495371832, "learning_rate": 2.944700985520513e-06, "loss": 0.0487, "step": 11829 }, { "epoch": 7.568777991042866, "grad_norm": 0.8229515011783411, "learning_rate": 2.9432326426019208e-06, "loss": 0.0589, "step": 11830 }, { "epoch": 7.569417786308382, "grad_norm": 0.8901018709072791, "learning_rate": 2.941764602688366e-06, "loss": 0.0544, "step": 11831 }, { "epoch": 7.570057581573896, "grad_norm": 0.826603930874784, "learning_rate": 2.940296865842884e-06, "loss": 0.0507, "step": 11832 }, { "epoch": 7.570697376839411, "grad_norm": 0.9691757134778956, "learning_rate": 2.938829432128496e-06, "loss": 0.051, "step": 11833 }, { "epoch": 7.5713371721049265, "grad_norm": 0.8605020108789558, "learning_rate": 2.9373623016082107e-06, "loss": 0.0507, "step": 11834 }, { "epoch": 7.571976967370442, "grad_norm": 0.7522954386987992, "learning_rate": 2.9358954743450253e-06, "loss": 0.0475, "step": 11835 }, { "epoch": 7.572616762635956, "grad_norm": 0.9013045638907012, "learning_rate": 2.9344289504019184e-06, "loss": 0.0468, "step": 11836 }, { "epoch": 7.5732565579014715, "grad_norm": 0.7969969128652918, "learning_rate": 2.9329627298418685e-06, "loss": 0.0515, "step": 11837 }, { "epoch": 7.573896353166987, "grad_norm": 0.8779704123184048, "learning_rate": 2.9314968127278263e-06, "loss": 0.0497, "step": 11838 }, { "epoch": 7.574536148432502, "grad_norm": 0.836609593378008, "learning_rate": 2.930031199122734e-06, "loss": 0.0496, "step": 11839 }, { "epoch": 7.575175943698016, "grad_norm": 0.7835600031296209, "learning_rate": 2.9285658890895285e-06, "loss": 0.0527, "step": 11840 }, { "epoch": 7.575815738963532, "grad_norm": 0.8599855806652985, "learning_rate": 2.927100882691125e-06, "loss": 0.0551, "step": 11841 }, { "epoch": 7.576455534229047, "grad_norm": 0.8641718317531624, "learning_rate": 2.9256361799904276e-06, "loss": 0.0515, "step": 11842 }, { "epoch": 7.577095329494561, "grad_norm": 0.854957583719205, "learning_rate": 2.9241717810503255e-06, "loss": 0.0531, "step": 11843 }, { "epoch": 7.577735124760077, "grad_norm": 0.8224047567206793, "learning_rate": 2.9227076859337033e-06, "loss": 0.0526, "step": 11844 }, { "epoch": 7.578374920025592, "grad_norm": 0.8007499535153401, "learning_rate": 2.921243894703424e-06, "loss": 0.0476, "step": 11845 }, { "epoch": 7.579014715291107, "grad_norm": 0.7905735017716441, "learning_rate": 2.919780407422336e-06, "loss": 0.0531, "step": 11846 }, { "epoch": 7.5796545105566215, "grad_norm": 0.8865797684364131, "learning_rate": 2.9183172241532887e-06, "loss": 0.0527, "step": 11847 }, { "epoch": 7.580294305822137, "grad_norm": 0.8002082743690443, "learning_rate": 2.916854344959099e-06, "loss": 0.0495, "step": 11848 }, { "epoch": 7.580934101087652, "grad_norm": 0.988544361941462, "learning_rate": 2.915391769902581e-06, "loss": 0.0598, "step": 11849 }, { "epoch": 7.581573896353167, "grad_norm": 0.8061207543564455, "learning_rate": 2.9139294990465416e-06, "loss": 0.0485, "step": 11850 }, { "epoch": 7.582213691618682, "grad_norm": 0.7898350540557995, "learning_rate": 2.9124675324537645e-06, "loss": 0.0502, "step": 11851 }, { "epoch": 7.582853486884197, "grad_norm": 0.8372431228432934, "learning_rate": 2.9110058701870224e-06, "loss": 0.0529, "step": 11852 }, { "epoch": 7.583493282149712, "grad_norm": 0.8656499445328132, "learning_rate": 2.909544512309076e-06, "loss": 0.0538, "step": 11853 }, { "epoch": 7.584133077415228, "grad_norm": 0.8739507304709204, "learning_rate": 2.908083458882678e-06, "loss": 0.0495, "step": 11854 }, { "epoch": 7.584772872680742, "grad_norm": 0.8513891232836711, "learning_rate": 2.906622709970561e-06, "loss": 0.0527, "step": 11855 }, { "epoch": 7.585412667946257, "grad_norm": 0.7843979776322594, "learning_rate": 2.905162265635443e-06, "loss": 0.0525, "step": 11856 }, { "epoch": 7.5860524632117725, "grad_norm": 0.8645775653732858, "learning_rate": 2.90370212594004e-06, "loss": 0.0563, "step": 11857 }, { "epoch": 7.586692258477287, "grad_norm": 1.3054706428550507, "learning_rate": 2.902242290947045e-06, "loss": 0.048, "step": 11858 }, { "epoch": 7.587332053742802, "grad_norm": 0.9015058451565394, "learning_rate": 2.9007827607191364e-06, "loss": 0.0544, "step": 11859 }, { "epoch": 7.5879718490083174, "grad_norm": 0.902805585552059, "learning_rate": 2.8993235353189908e-06, "loss": 0.0566, "step": 11860 }, { "epoch": 7.588611644273833, "grad_norm": 0.8581679245096823, "learning_rate": 2.8978646148092616e-06, "loss": 0.0514, "step": 11861 }, { "epoch": 7.589251439539347, "grad_norm": 0.8781800259745555, "learning_rate": 2.8964059992525918e-06, "loss": 0.0511, "step": 11862 }, { "epoch": 7.589891234804862, "grad_norm": 0.874134539012819, "learning_rate": 2.8949476887116112e-06, "loss": 0.0516, "step": 11863 }, { "epoch": 7.590531030070378, "grad_norm": 0.8740684595487968, "learning_rate": 2.893489683248939e-06, "loss": 0.0552, "step": 11864 }, { "epoch": 7.591170825335892, "grad_norm": 0.922113884355433, "learning_rate": 2.892031982927178e-06, "loss": 0.063, "step": 11865 }, { "epoch": 7.591810620601407, "grad_norm": 0.8203574869669804, "learning_rate": 2.8905745878089166e-06, "loss": 0.0509, "step": 11866 }, { "epoch": 7.592450415866923, "grad_norm": 0.9773022273551577, "learning_rate": 2.8891174979567392e-06, "loss": 0.058, "step": 11867 }, { "epoch": 7.593090211132438, "grad_norm": 0.8450387414776255, "learning_rate": 2.887660713433207e-06, "loss": 0.0564, "step": 11868 }, { "epoch": 7.593730006397953, "grad_norm": 0.802514746510367, "learning_rate": 2.886204234300871e-06, "loss": 0.0463, "step": 11869 }, { "epoch": 7.5943698016634675, "grad_norm": 0.9164207964549533, "learning_rate": 2.884748060622269e-06, "loss": 0.0531, "step": 11870 }, { "epoch": 7.595009596928983, "grad_norm": 0.8325769687190886, "learning_rate": 2.88329219245993e-06, "loss": 0.0517, "step": 11871 }, { "epoch": 7.595649392194498, "grad_norm": 0.8026757602184623, "learning_rate": 2.8818366298763656e-06, "loss": 0.0462, "step": 11872 }, { "epoch": 7.5962891874600125, "grad_norm": 0.7791473222170763, "learning_rate": 2.88038137293407e-06, "loss": 0.0536, "step": 11873 }, { "epoch": 7.596928982725528, "grad_norm": 0.8160336811520846, "learning_rate": 2.8789264216955382e-06, "loss": 0.0505, "step": 11874 }, { "epoch": 7.597568777991043, "grad_norm": 0.8974311147400952, "learning_rate": 2.8774717762232373e-06, "loss": 0.0532, "step": 11875 }, { "epoch": 7.598208573256558, "grad_norm": 0.7580785569167625, "learning_rate": 2.876017436579628e-06, "loss": 0.0453, "step": 11876 }, { "epoch": 7.598848368522073, "grad_norm": 0.8541756596081431, "learning_rate": 2.8745634028271586e-06, "loss": 0.0564, "step": 11877 }, { "epoch": 7.599488163787588, "grad_norm": 0.8121779721457169, "learning_rate": 2.8731096750282605e-06, "loss": 0.0452, "step": 11878 }, { "epoch": 7.600127959053103, "grad_norm": 0.8781762774117832, "learning_rate": 2.871656253245355e-06, "loss": 0.058, "step": 11879 }, { "epoch": 7.600767754318618, "grad_norm": 0.872323620832935, "learning_rate": 2.870203137540847e-06, "loss": 0.0535, "step": 11880 }, { "epoch": 7.601407549584133, "grad_norm": 0.870610832139763, "learning_rate": 2.8687503279771367e-06, "loss": 0.0575, "step": 11881 }, { "epoch": 7.602047344849648, "grad_norm": 0.8676839856901257, "learning_rate": 2.8672978246166017e-06, "loss": 0.0532, "step": 11882 }, { "epoch": 7.6026871401151634, "grad_norm": 0.8585499955825947, "learning_rate": 2.865845627521606e-06, "loss": 0.0552, "step": 11883 }, { "epoch": 7.603326935380678, "grad_norm": 0.9270056382448733, "learning_rate": 2.8643937367545126e-06, "loss": 0.0573, "step": 11884 }, { "epoch": 7.603966730646193, "grad_norm": 0.9049683991704666, "learning_rate": 2.862942152377659e-06, "loss": 0.052, "step": 11885 }, { "epoch": 7.604606525911708, "grad_norm": 0.8462362117601308, "learning_rate": 2.86149087445337e-06, "loss": 0.0543, "step": 11886 }, { "epoch": 7.605246321177224, "grad_norm": 0.8081200759037427, "learning_rate": 2.8600399030439675e-06, "loss": 0.0511, "step": 11887 }, { "epoch": 7.605886116442738, "grad_norm": 0.8339039707903666, "learning_rate": 2.8585892382117495e-06, "loss": 0.0488, "step": 11888 }, { "epoch": 7.606525911708253, "grad_norm": 0.7841772370329457, "learning_rate": 2.857138880019008e-06, "loss": 0.0466, "step": 11889 }, { "epoch": 7.607165706973769, "grad_norm": 0.8499276445191978, "learning_rate": 2.8556888285280126e-06, "loss": 0.052, "step": 11890 }, { "epoch": 7.607805502239284, "grad_norm": 0.779042952011807, "learning_rate": 2.8542390838010337e-06, "loss": 0.0465, "step": 11891 }, { "epoch": 7.608445297504798, "grad_norm": 0.7686698373705152, "learning_rate": 2.852789645900317e-06, "loss": 0.0503, "step": 11892 }, { "epoch": 7.6090850927703135, "grad_norm": 0.8238971270406489, "learning_rate": 2.851340514888098e-06, "loss": 0.0505, "step": 11893 }, { "epoch": 7.609724888035829, "grad_norm": 0.8569450763635544, "learning_rate": 2.8498916908266027e-06, "loss": 0.0578, "step": 11894 }, { "epoch": 7.610364683301343, "grad_norm": 0.7931415773800639, "learning_rate": 2.8484431737780382e-06, "loss": 0.0516, "step": 11895 }, { "epoch": 7.6110044785668585, "grad_norm": 0.8024555287097326, "learning_rate": 2.846994963804599e-06, "loss": 0.0504, "step": 11896 }, { "epoch": 7.611644273832374, "grad_norm": 0.8021125703128306, "learning_rate": 2.8455470609684743e-06, "loss": 0.0499, "step": 11897 }, { "epoch": 7.612284069097889, "grad_norm": 0.8035691420033871, "learning_rate": 2.8440994653318333e-06, "loss": 0.0492, "step": 11898 }, { "epoch": 7.612923864363403, "grad_norm": 0.7511855440774193, "learning_rate": 2.8426521769568318e-06, "loss": 0.0482, "step": 11899 }, { "epoch": 7.613563659628919, "grad_norm": 0.873947749131232, "learning_rate": 2.84120519590561e-06, "loss": 0.061, "step": 11900 }, { "epoch": 7.614203454894434, "grad_norm": 0.8074628043470973, "learning_rate": 2.8397585222403058e-06, "loss": 0.0513, "step": 11901 }, { "epoch": 7.614843250159948, "grad_norm": 0.8593845959352246, "learning_rate": 2.8383121560230343e-06, "loss": 0.0557, "step": 11902 }, { "epoch": 7.615483045425464, "grad_norm": 0.8243305716731413, "learning_rate": 2.8368660973158955e-06, "loss": 0.0561, "step": 11903 }, { "epoch": 7.616122840690979, "grad_norm": 0.8269890290523059, "learning_rate": 2.835420346180987e-06, "loss": 0.0509, "step": 11904 }, { "epoch": 7.616762635956494, "grad_norm": 0.8425346562956074, "learning_rate": 2.8339749026803844e-06, "loss": 0.0522, "step": 11905 }, { "epoch": 7.617402431222009, "grad_norm": 0.8114450939931405, "learning_rate": 2.832529766876152e-06, "loss": 0.0491, "step": 11906 }, { "epoch": 7.618042226487524, "grad_norm": 0.8826849670773046, "learning_rate": 2.831084938830342e-06, "loss": 0.055, "step": 11907 }, { "epoch": 7.618682021753039, "grad_norm": 0.8975622910109912, "learning_rate": 2.829640418604992e-06, "loss": 0.0526, "step": 11908 }, { "epoch": 7.619321817018554, "grad_norm": 1.1018681116016076, "learning_rate": 2.828196206262127e-06, "loss": 0.0599, "step": 11909 }, { "epoch": 7.619961612284069, "grad_norm": 0.910027018499788, "learning_rate": 2.8267523018637566e-06, "loss": 0.0537, "step": 11910 }, { "epoch": 7.620601407549584, "grad_norm": 0.8623984108519989, "learning_rate": 2.8253087054718854e-06, "loss": 0.0519, "step": 11911 }, { "epoch": 7.621241202815099, "grad_norm": 1.5937219560236615, "learning_rate": 2.8238654171484946e-06, "loss": 0.056, "step": 11912 }, { "epoch": 7.621880998080615, "grad_norm": 0.8581971757992394, "learning_rate": 2.822422436955555e-06, "loss": 0.0518, "step": 11913 }, { "epoch": 7.622520793346129, "grad_norm": 0.8581014758544304, "learning_rate": 2.820979764955031e-06, "loss": 0.0535, "step": 11914 }, { "epoch": 7.623160588611644, "grad_norm": 0.9724384236831122, "learning_rate": 2.819537401208866e-06, "loss": 0.0557, "step": 11915 }, { "epoch": 7.6238003838771595, "grad_norm": 0.8467085982683691, "learning_rate": 2.8180953457789884e-06, "loss": 0.0474, "step": 11916 }, { "epoch": 7.624440179142674, "grad_norm": 0.9307152076436607, "learning_rate": 2.8166535987273236e-06, "loss": 0.059, "step": 11917 }, { "epoch": 7.625079974408189, "grad_norm": 0.826053572381208, "learning_rate": 2.8152121601157757e-06, "loss": 0.055, "step": 11918 }, { "epoch": 7.6257197696737045, "grad_norm": 0.8267101793671924, "learning_rate": 2.813771030006237e-06, "loss": 0.0486, "step": 11919 }, { "epoch": 7.62635956493922, "grad_norm": 0.9095789078242486, "learning_rate": 2.812330208460583e-06, "loss": 0.056, "step": 11920 }, { "epoch": 7.626999360204734, "grad_norm": 0.8748879801661421, "learning_rate": 2.8108896955406883e-06, "loss": 0.0569, "step": 11921 }, { "epoch": 7.627639155470249, "grad_norm": 0.9382661994002837, "learning_rate": 2.8094494913084003e-06, "loss": 0.0492, "step": 11922 }, { "epoch": 7.628278950735765, "grad_norm": 0.847853448487078, "learning_rate": 2.8080095958255617e-06, "loss": 0.0577, "step": 11923 }, { "epoch": 7.62891874600128, "grad_norm": 0.9155196372572741, "learning_rate": 2.806570009153996e-06, "loss": 0.0551, "step": 11924 }, { "epoch": 7.629558541266794, "grad_norm": 0.8898679138309402, "learning_rate": 2.8051307313555186e-06, "loss": 0.0542, "step": 11925 }, { "epoch": 7.63019833653231, "grad_norm": 0.7432779819602126, "learning_rate": 2.8036917624919257e-06, "loss": 0.0483, "step": 11926 }, { "epoch": 7.630838131797825, "grad_norm": 0.8719495148483221, "learning_rate": 2.802253102625011e-06, "loss": 0.057, "step": 11927 }, { "epoch": 7.63147792706334, "grad_norm": 0.8687823225144335, "learning_rate": 2.800814751816544e-06, "loss": 0.054, "step": 11928 }, { "epoch": 7.6321177223288545, "grad_norm": 0.8501517771553057, "learning_rate": 2.7993767101282855e-06, "loss": 0.0545, "step": 11929 }, { "epoch": 7.63275751759437, "grad_norm": 0.836844253961396, "learning_rate": 2.7979389776219777e-06, "loss": 0.0561, "step": 11930 }, { "epoch": 7.633397312859885, "grad_norm": 0.9042110691650598, "learning_rate": 2.796501554359363e-06, "loss": 0.0568, "step": 11931 }, { "epoch": 7.6340371081253995, "grad_norm": 0.8508236258530368, "learning_rate": 2.7950644404021576e-06, "loss": 0.0486, "step": 11932 }, { "epoch": 7.634676903390915, "grad_norm": 0.8324715077827317, "learning_rate": 2.793627635812065e-06, "loss": 0.0544, "step": 11933 }, { "epoch": 7.63531669865643, "grad_norm": 0.768203205418242, "learning_rate": 2.792191140650785e-06, "loss": 0.0495, "step": 11934 }, { "epoch": 7.635956493921945, "grad_norm": 0.8604447736385316, "learning_rate": 2.790754954979997e-06, "loss": 0.0565, "step": 11935 }, { "epoch": 7.63659628918746, "grad_norm": 0.9673145843632965, "learning_rate": 2.7893190788613665e-06, "loss": 0.0564, "step": 11936 }, { "epoch": 7.637236084452975, "grad_norm": 0.8046683854180652, "learning_rate": 2.7878835123565474e-06, "loss": 0.0527, "step": 11937 }, { "epoch": 7.63787587971849, "grad_norm": 0.8420850664361483, "learning_rate": 2.7864482555271822e-06, "loss": 0.0554, "step": 11938 }, { "epoch": 7.638515674984005, "grad_norm": 0.8532921472182111, "learning_rate": 2.7850133084348964e-06, "loss": 0.0457, "step": 11939 }, { "epoch": 7.63915547024952, "grad_norm": 0.7756169201206609, "learning_rate": 2.7835786711413005e-06, "loss": 0.0415, "step": 11940 }, { "epoch": 7.639795265515035, "grad_norm": 0.826096956311221, "learning_rate": 2.782144343708003e-06, "loss": 0.0491, "step": 11941 }, { "epoch": 7.6404350607805505, "grad_norm": 0.808271106744474, "learning_rate": 2.7807103261965886e-06, "loss": 0.0538, "step": 11942 }, { "epoch": 7.641074856046066, "grad_norm": 0.9130195882361526, "learning_rate": 2.779276618668627e-06, "loss": 0.0592, "step": 11943 }, { "epoch": 7.64171465131158, "grad_norm": 0.9149030065825661, "learning_rate": 2.7778432211856864e-06, "loss": 0.0546, "step": 11944 }, { "epoch": 7.642354446577095, "grad_norm": 0.830675141214822, "learning_rate": 2.7764101338093108e-06, "loss": 0.0525, "step": 11945 }, { "epoch": 7.642994241842611, "grad_norm": 0.9754230338020823, "learning_rate": 2.774977356601033e-06, "loss": 0.0599, "step": 11946 }, { "epoch": 7.643634037108125, "grad_norm": 0.916912246681488, "learning_rate": 2.773544889622374e-06, "loss": 0.055, "step": 11947 }, { "epoch": 7.64427383237364, "grad_norm": 0.8307566520316965, "learning_rate": 2.772112732934844e-06, "loss": 0.0501, "step": 11948 }, { "epoch": 7.644913627639156, "grad_norm": 0.9253537505489574, "learning_rate": 2.7706808865999356e-06, "loss": 0.0558, "step": 11949 }, { "epoch": 7.645553422904671, "grad_norm": 0.836342797703393, "learning_rate": 2.7692493506791283e-06, "loss": 0.0525, "step": 11950 }, { "epoch": 7.646193218170185, "grad_norm": 0.8699529133353505, "learning_rate": 2.767818125233894e-06, "loss": 0.0544, "step": 11951 }, { "epoch": 7.6468330134357005, "grad_norm": 0.8063591176258332, "learning_rate": 2.766387210325684e-06, "loss": 0.0486, "step": 11952 }, { "epoch": 7.647472808701216, "grad_norm": 0.8571290509202335, "learning_rate": 2.7649566060159396e-06, "loss": 0.0536, "step": 11953 }, { "epoch": 7.64811260396673, "grad_norm": 1.0911843317864194, "learning_rate": 2.763526312366088e-06, "loss": 0.0521, "step": 11954 }, { "epoch": 7.6487523992322455, "grad_norm": 0.7419279254194799, "learning_rate": 2.7620963294375445e-06, "loss": 0.0445, "step": 11955 }, { "epoch": 7.649392194497761, "grad_norm": 0.7847704324059548, "learning_rate": 2.7606666572917094e-06, "loss": 0.052, "step": 11956 }, { "epoch": 7.650031989763276, "grad_norm": 0.86679617839225, "learning_rate": 2.7592372959899673e-06, "loss": 0.0572, "step": 11957 }, { "epoch": 7.65067178502879, "grad_norm": 0.8273504261722747, "learning_rate": 2.7578082455936983e-06, "loss": 0.0544, "step": 11958 }, { "epoch": 7.651311580294306, "grad_norm": 0.7830940482968513, "learning_rate": 2.756379506164261e-06, "loss": 0.0503, "step": 11959 }, { "epoch": 7.651951375559821, "grad_norm": 0.7460647945223472, "learning_rate": 2.7549510777629986e-06, "loss": 0.052, "step": 11960 }, { "epoch": 7.652591170825336, "grad_norm": 0.8240444567571517, "learning_rate": 2.7535229604512516e-06, "loss": 0.0559, "step": 11961 }, { "epoch": 7.653230966090851, "grad_norm": 0.8759192206975452, "learning_rate": 2.7520951542903386e-06, "loss": 0.0548, "step": 11962 }, { "epoch": 7.653870761356366, "grad_norm": 0.8190913696558529, "learning_rate": 2.750667659341564e-06, "loss": 0.0487, "step": 11963 }, { "epoch": 7.654510556621881, "grad_norm": 0.8414361964403219, "learning_rate": 2.7492404756662272e-06, "loss": 0.0532, "step": 11964 }, { "epoch": 7.6551503518873965, "grad_norm": 0.7854798797365967, "learning_rate": 2.7478136033256055e-06, "loss": 0.0531, "step": 11965 }, { "epoch": 7.655790147152911, "grad_norm": 0.8870018046118195, "learning_rate": 2.746387042380968e-06, "loss": 0.0563, "step": 11966 }, { "epoch": 7.656429942418426, "grad_norm": 0.7885606495478805, "learning_rate": 2.7449607928935673e-06, "loss": 0.0533, "step": 11967 }, { "epoch": 7.657069737683941, "grad_norm": 0.8260400977124145, "learning_rate": 2.7435348549246454e-06, "loss": 0.0487, "step": 11968 }, { "epoch": 7.657709532949456, "grad_norm": 0.9334213046370838, "learning_rate": 2.7421092285354277e-06, "loss": 0.0568, "step": 11969 }, { "epoch": 7.658349328214971, "grad_norm": 0.7089451786696243, "learning_rate": 2.7406839137871255e-06, "loss": 0.0444, "step": 11970 }, { "epoch": 7.658989123480486, "grad_norm": 0.7946318404984328, "learning_rate": 2.7392589107409463e-06, "loss": 0.0555, "step": 11971 }, { "epoch": 7.659628918746002, "grad_norm": 0.8972259825222573, "learning_rate": 2.737834219458074e-06, "loss": 0.055, "step": 11972 }, { "epoch": 7.660268714011516, "grad_norm": 0.8658321092351801, "learning_rate": 2.7364098399996784e-06, "loss": 0.0519, "step": 11973 }, { "epoch": 7.660908509277031, "grad_norm": 0.8438450429049299, "learning_rate": 2.7349857724269268e-06, "loss": 0.0496, "step": 11974 }, { "epoch": 7.6615483045425465, "grad_norm": 0.8703970630267476, "learning_rate": 2.7335620168009626e-06, "loss": 0.0575, "step": 11975 }, { "epoch": 7.662188099808061, "grad_norm": 0.8232148451485757, "learning_rate": 2.7321385731829186e-06, "loss": 0.0522, "step": 11976 }, { "epoch": 7.662827895073576, "grad_norm": 0.8425290216872706, "learning_rate": 2.730715441633913e-06, "loss": 0.0536, "step": 11977 }, { "epoch": 7.6634676903390915, "grad_norm": 0.8913069873909961, "learning_rate": 2.7292926222150583e-06, "loss": 0.0546, "step": 11978 }, { "epoch": 7.664107485604607, "grad_norm": 0.8798354379889635, "learning_rate": 2.7278701149874456e-06, "loss": 0.0503, "step": 11979 }, { "epoch": 7.664747280870122, "grad_norm": 0.8625862149999106, "learning_rate": 2.7264479200121507e-06, "loss": 0.0587, "step": 11980 }, { "epoch": 7.665387076135636, "grad_norm": 0.7886119095067292, "learning_rate": 2.7250260373502456e-06, "loss": 0.0555, "step": 11981 }, { "epoch": 7.666026871401152, "grad_norm": 0.8165334699016813, "learning_rate": 2.7236044670627828e-06, "loss": 0.05, "step": 11982 }, { "epoch": 7.666666666666667, "grad_norm": 0.8855440081130113, "learning_rate": 2.7221832092107993e-06, "loss": 0.0546, "step": 11983 }, { "epoch": 7.667306461932181, "grad_norm": 0.8760564151066683, "learning_rate": 2.720762263855323e-06, "loss": 0.054, "step": 11984 }, { "epoch": 7.667946257197697, "grad_norm": 0.8252987752582578, "learning_rate": 2.719341631057366e-06, "loss": 0.0464, "step": 11985 }, { "epoch": 7.668586052463212, "grad_norm": 0.8245989048107989, "learning_rate": 2.717921310877929e-06, "loss": 0.0486, "step": 11986 }, { "epoch": 7.669225847728727, "grad_norm": 0.8285303898773184, "learning_rate": 2.716501303377993e-06, "loss": 0.0552, "step": 11987 }, { "epoch": 7.669865642994242, "grad_norm": 0.8402712018931779, "learning_rate": 2.7150816086185384e-06, "loss": 0.0485, "step": 11988 }, { "epoch": 7.670505438259757, "grad_norm": 0.8258569460423335, "learning_rate": 2.7136622266605205e-06, "loss": 0.0533, "step": 11989 }, { "epoch": 7.671145233525272, "grad_norm": 0.7918798809021446, "learning_rate": 2.7122431575648823e-06, "loss": 0.0509, "step": 11990 }, { "epoch": 7.6717850287907865, "grad_norm": 0.9031112436342426, "learning_rate": 2.710824401392561e-06, "loss": 0.0548, "step": 11991 }, { "epoch": 7.672424824056302, "grad_norm": 0.7635164371547823, "learning_rate": 2.709405958204475e-06, "loss": 0.0491, "step": 11992 }, { "epoch": 7.673064619321817, "grad_norm": 0.8731187362536119, "learning_rate": 2.7079878280615247e-06, "loss": 0.0554, "step": 11993 }, { "epoch": 7.673704414587332, "grad_norm": 0.8248322039821313, "learning_rate": 2.7065700110246074e-06, "loss": 0.0538, "step": 11994 }, { "epoch": 7.674344209852847, "grad_norm": 0.9006337760877449, "learning_rate": 2.7051525071546014e-06, "loss": 0.0541, "step": 11995 }, { "epoch": 7.674984005118362, "grad_norm": 0.8128609738685412, "learning_rate": 2.7037353165123692e-06, "loss": 0.0496, "step": 11996 }, { "epoch": 7.675623800383877, "grad_norm": 0.8123840874276401, "learning_rate": 2.702318439158764e-06, "loss": 0.049, "step": 11997 }, { "epoch": 7.6762635956493925, "grad_norm": 0.7479439916084794, "learning_rate": 2.7009018751546233e-06, "loss": 0.047, "step": 11998 }, { "epoch": 7.676903390914907, "grad_norm": 0.83970298680527, "learning_rate": 2.6994856245607728e-06, "loss": 0.0528, "step": 11999 }, { "epoch": 7.677543186180422, "grad_norm": 0.8475770190072106, "learning_rate": 2.6980696874380184e-06, "loss": 0.0459, "step": 12000 }, { "epoch": 7.6781829814459375, "grad_norm": 0.8610913065688204, "learning_rate": 2.6966540638471662e-06, "loss": 0.0502, "step": 12001 }, { "epoch": 7.678822776711453, "grad_norm": 0.8257459024710395, "learning_rate": 2.6952387538489967e-06, "loss": 0.0482, "step": 12002 }, { "epoch": 7.679462571976967, "grad_norm": 0.981243217100891, "learning_rate": 2.693823757504278e-06, "loss": 0.0579, "step": 12003 }, { "epoch": 7.680102367242482, "grad_norm": 0.8502963534543286, "learning_rate": 2.692409074873774e-06, "loss": 0.0512, "step": 12004 }, { "epoch": 7.680742162507998, "grad_norm": 0.8670793646816092, "learning_rate": 2.690994706018225e-06, "loss": 0.0537, "step": 12005 }, { "epoch": 7.681381957773512, "grad_norm": 0.9369919692462675, "learning_rate": 2.6895806509983613e-06, "loss": 0.056, "step": 12006 }, { "epoch": 7.682021753039027, "grad_norm": 0.8272529160129178, "learning_rate": 2.6881669098748974e-06, "loss": 0.0496, "step": 12007 }, { "epoch": 7.682661548304543, "grad_norm": 0.8572755548920817, "learning_rate": 2.686753482708543e-06, "loss": 0.0527, "step": 12008 }, { "epoch": 7.683301343570058, "grad_norm": 0.8164546592530153, "learning_rate": 2.6853403695599834e-06, "loss": 0.0524, "step": 12009 }, { "epoch": 7.683941138835572, "grad_norm": 0.8271340289071513, "learning_rate": 2.6839275704898947e-06, "loss": 0.0507, "step": 12010 }, { "epoch": 7.684580934101088, "grad_norm": 0.7767497616456412, "learning_rate": 2.6825150855589445e-06, "loss": 0.0486, "step": 12011 }, { "epoch": 7.685220729366603, "grad_norm": 0.9523347246252336, "learning_rate": 2.68110291482778e-06, "loss": 0.0563, "step": 12012 }, { "epoch": 7.685860524632118, "grad_norm": 0.8231071549260762, "learning_rate": 2.679691058357037e-06, "loss": 0.0527, "step": 12013 }, { "epoch": 7.6865003198976325, "grad_norm": 0.8451818505957526, "learning_rate": 2.6782795162073385e-06, "loss": 0.0559, "step": 12014 }, { "epoch": 7.687140115163148, "grad_norm": 0.7915931421618996, "learning_rate": 2.6768682884392928e-06, "loss": 0.0532, "step": 12015 }, { "epoch": 7.687779910428663, "grad_norm": 0.8542492515091729, "learning_rate": 2.675457375113495e-06, "loss": 0.053, "step": 12016 }, { "epoch": 7.688419705694178, "grad_norm": 0.7270767594553817, "learning_rate": 2.6740467762905264e-06, "loss": 0.0429, "step": 12017 }, { "epoch": 7.689059500959693, "grad_norm": 0.8882014810213485, "learning_rate": 2.67263649203096e-06, "loss": 0.0573, "step": 12018 }, { "epoch": 7.689699296225208, "grad_norm": 0.8925340085350724, "learning_rate": 2.671226522395348e-06, "loss": 0.0539, "step": 12019 }, { "epoch": 7.690339091490723, "grad_norm": 0.8427818976781866, "learning_rate": 2.6698168674442304e-06, "loss": 0.0564, "step": 12020 }, { "epoch": 7.690978886756238, "grad_norm": 0.85543938707326, "learning_rate": 2.6684075272381395e-06, "loss": 0.0498, "step": 12021 }, { "epoch": 7.691618682021753, "grad_norm": 0.9009606959567439, "learning_rate": 2.6669985018375864e-06, "loss": 0.0519, "step": 12022 }, { "epoch": 7.692258477287268, "grad_norm": 0.8273852256417223, "learning_rate": 2.665589791303075e-06, "loss": 0.0549, "step": 12023 }, { "epoch": 7.6928982725527835, "grad_norm": 0.8787833599805027, "learning_rate": 2.664181395695087e-06, "loss": 0.0509, "step": 12024 }, { "epoch": 7.693538067818298, "grad_norm": 0.8252044529277407, "learning_rate": 2.662773315074103e-06, "loss": 0.0494, "step": 12025 }, { "epoch": 7.694177863083813, "grad_norm": 0.8524985174341017, "learning_rate": 2.661365549500582e-06, "loss": 0.0481, "step": 12026 }, { "epoch": 7.694817658349328, "grad_norm": 0.8281956039012732, "learning_rate": 2.6599580990349694e-06, "loss": 0.0526, "step": 12027 }, { "epoch": 7.695457453614843, "grad_norm": 0.8552404680695326, "learning_rate": 2.6585509637376996e-06, "loss": 0.0499, "step": 12028 }, { "epoch": 7.696097248880358, "grad_norm": 0.9156678878794141, "learning_rate": 2.6571441436691913e-06, "loss": 0.0547, "step": 12029 }, { "epoch": 7.696737044145873, "grad_norm": 0.8764988209251172, "learning_rate": 2.655737638889848e-06, "loss": 0.0556, "step": 12030 }, { "epoch": 7.697376839411389, "grad_norm": 0.8880882796124758, "learning_rate": 2.6543314494600704e-06, "loss": 0.0567, "step": 12031 }, { "epoch": 7.698016634676903, "grad_norm": 0.8608751821008782, "learning_rate": 2.652925575440233e-06, "loss": 0.0551, "step": 12032 }, { "epoch": 7.698656429942418, "grad_norm": 0.7993079271553212, "learning_rate": 2.651520016890702e-06, "loss": 0.0522, "step": 12033 }, { "epoch": 7.6992962252079336, "grad_norm": 0.9175107240294149, "learning_rate": 2.650114773871826e-06, "loss": 0.0464, "step": 12034 }, { "epoch": 7.699936020473449, "grad_norm": 0.8446848951046843, "learning_rate": 2.648709846443951e-06, "loss": 0.0526, "step": 12035 }, { "epoch": 7.700575815738963, "grad_norm": 0.854156916468435, "learning_rate": 2.6473052346673978e-06, "loss": 0.0543, "step": 12036 }, { "epoch": 7.7012156110044785, "grad_norm": 0.8671752747333287, "learning_rate": 2.645900938602475e-06, "loss": 0.0569, "step": 12037 }, { "epoch": 7.701855406269994, "grad_norm": 0.7825433041886928, "learning_rate": 2.6444969583094872e-06, "loss": 0.0517, "step": 12038 }, { "epoch": 7.702495201535509, "grad_norm": 0.7977851426827571, "learning_rate": 2.6430932938487163e-06, "loss": 0.0474, "step": 12039 }, { "epoch": 7.703134996801023, "grad_norm": 0.9005779435152456, "learning_rate": 2.6416899452804288e-06, "loss": 0.0526, "step": 12040 }, { "epoch": 7.703774792066539, "grad_norm": 0.86390865980829, "learning_rate": 2.640286912664889e-06, "loss": 0.0532, "step": 12041 }, { "epoch": 7.704414587332054, "grad_norm": 0.7433570387336634, "learning_rate": 2.6388841960623367e-06, "loss": 0.0456, "step": 12042 }, { "epoch": 7.705054382597568, "grad_norm": 0.8271452595215236, "learning_rate": 2.6374817955330034e-06, "loss": 0.0537, "step": 12043 }, { "epoch": 7.705694177863084, "grad_norm": 0.8580166214325601, "learning_rate": 2.636079711137104e-06, "loss": 0.0512, "step": 12044 }, { "epoch": 7.706333973128599, "grad_norm": 0.9237512757381805, "learning_rate": 2.6346779429348425e-06, "loss": 0.0599, "step": 12045 }, { "epoch": 7.706973768394114, "grad_norm": 0.7487144468579381, "learning_rate": 2.6332764909864083e-06, "loss": 0.0478, "step": 12046 }, { "epoch": 7.707613563659629, "grad_norm": 0.9844212209686086, "learning_rate": 2.631875355351975e-06, "loss": 0.0588, "step": 12047 }, { "epoch": 7.708253358925144, "grad_norm": 0.8560103191813492, "learning_rate": 2.6304745360917095e-06, "loss": 0.0549, "step": 12048 }, { "epoch": 7.708893154190659, "grad_norm": 0.7900629555182405, "learning_rate": 2.6290740332657584e-06, "loss": 0.053, "step": 12049 }, { "epoch": 7.709532949456174, "grad_norm": 0.882293355102911, "learning_rate": 2.6276738469342533e-06, "loss": 0.0554, "step": 12050 }, { "epoch": 7.710172744721689, "grad_norm": 0.9163918477611164, "learning_rate": 2.626273977157321e-06, "loss": 0.0504, "step": 12051 }, { "epoch": 7.710812539987204, "grad_norm": 0.7607797905979579, "learning_rate": 2.6248744239950686e-06, "loss": 0.0439, "step": 12052 }, { "epoch": 7.711452335252719, "grad_norm": 0.9589711874669948, "learning_rate": 2.623475187507588e-06, "loss": 0.0594, "step": 12053 }, { "epoch": 7.712092130518235, "grad_norm": 0.8101793018552425, "learning_rate": 2.622076267754958e-06, "loss": 0.0567, "step": 12054 }, { "epoch": 7.712731925783749, "grad_norm": 0.8020475635877476, "learning_rate": 2.6206776647972533e-06, "loss": 0.0523, "step": 12055 }, { "epoch": 7.713371721049264, "grad_norm": 0.9187422417817929, "learning_rate": 2.6192793786945213e-06, "loss": 0.0571, "step": 12056 }, { "epoch": 7.7140115163147795, "grad_norm": 0.8437710930054303, "learning_rate": 2.617881409506804e-06, "loss": 0.0578, "step": 12057 }, { "epoch": 7.714651311580294, "grad_norm": 0.8064296865655098, "learning_rate": 2.6164837572941272e-06, "loss": 0.0535, "step": 12058 }, { "epoch": 7.715291106845809, "grad_norm": 0.9167860212303169, "learning_rate": 2.615086422116504e-06, "loss": 0.0581, "step": 12059 }, { "epoch": 7.7159309021113245, "grad_norm": 0.8059873732463829, "learning_rate": 2.613689404033929e-06, "loss": 0.0475, "step": 12060 }, { "epoch": 7.71657069737684, "grad_norm": 0.8578312798305663, "learning_rate": 2.612292703106396e-06, "loss": 0.0529, "step": 12061 }, { "epoch": 7.717210492642354, "grad_norm": 0.9380799541566346, "learning_rate": 2.6108963193938718e-06, "loss": 0.056, "step": 12062 }, { "epoch": 7.717850287907869, "grad_norm": 0.8094938833326807, "learning_rate": 2.609500252956316e-06, "loss": 0.0477, "step": 12063 }, { "epoch": 7.718490083173385, "grad_norm": 0.8337166060644767, "learning_rate": 2.608104503853669e-06, "loss": 0.0523, "step": 12064 }, { "epoch": 7.719129878438899, "grad_norm": 0.836424411757761, "learning_rate": 2.6067090721458687e-06, "loss": 0.0505, "step": 12065 }, { "epoch": 7.719769673704414, "grad_norm": 0.8359210242231009, "learning_rate": 2.60531395789283e-06, "loss": 0.0519, "step": 12066 }, { "epoch": 7.72040946896993, "grad_norm": 0.929563082707588, "learning_rate": 2.6039191611544523e-06, "loss": 0.0511, "step": 12067 }, { "epoch": 7.721049264235445, "grad_norm": 0.9036537872669262, "learning_rate": 2.6025246819906314e-06, "loss": 0.0543, "step": 12068 }, { "epoch": 7.72168905950096, "grad_norm": 0.873846860431665, "learning_rate": 2.6011305204612426e-06, "loss": 0.0542, "step": 12069 }, { "epoch": 7.722328854766475, "grad_norm": 0.8487309332858661, "learning_rate": 2.5997366766261445e-06, "loss": 0.0513, "step": 12070 }, { "epoch": 7.72296865003199, "grad_norm": 0.8637681882231384, "learning_rate": 2.5983431505451917e-06, "loss": 0.0536, "step": 12071 }, { "epoch": 7.723608445297505, "grad_norm": 0.8111680011747661, "learning_rate": 2.5969499422782183e-06, "loss": 0.0468, "step": 12072 }, { "epoch": 7.7242482405630195, "grad_norm": 0.8444286388027671, "learning_rate": 2.595557051885046e-06, "loss": 0.0472, "step": 12073 }, { "epoch": 7.724888035828535, "grad_norm": 0.8009700725231895, "learning_rate": 2.5941644794254804e-06, "loss": 0.0606, "step": 12074 }, { "epoch": 7.72552783109405, "grad_norm": 0.8505213622092986, "learning_rate": 2.5927722249593203e-06, "loss": 0.0556, "step": 12075 }, { "epoch": 7.726167626359565, "grad_norm": 0.8619272794017142, "learning_rate": 2.5913802885463425e-06, "loss": 0.0543, "step": 12076 }, { "epoch": 7.72680742162508, "grad_norm": 0.8184339711485225, "learning_rate": 2.5899886702463138e-06, "loss": 0.051, "step": 12077 }, { "epoch": 7.727447216890595, "grad_norm": 0.8228216622126291, "learning_rate": 2.588597370118994e-06, "loss": 0.0525, "step": 12078 }, { "epoch": 7.72808701215611, "grad_norm": 0.831918909435824, "learning_rate": 2.5872063882241193e-06, "loss": 0.0496, "step": 12079 }, { "epoch": 7.728726807421625, "grad_norm": 0.78824964232914, "learning_rate": 2.5858157246214123e-06, "loss": 0.0526, "step": 12080 }, { "epoch": 7.72936660268714, "grad_norm": 0.8311803592743473, "learning_rate": 2.584425379370592e-06, "loss": 0.0509, "step": 12081 }, { "epoch": 7.730006397952655, "grad_norm": 0.7799158441643427, "learning_rate": 2.5830353525313556e-06, "loss": 0.0492, "step": 12082 }, { "epoch": 7.7306461932181705, "grad_norm": 0.8463702481515728, "learning_rate": 2.5816456441633863e-06, "loss": 0.05, "step": 12083 }, { "epoch": 7.731285988483685, "grad_norm": 0.9437077236754244, "learning_rate": 2.580256254326354e-06, "loss": 0.0564, "step": 12084 }, { "epoch": 7.7319257837492, "grad_norm": 0.8979147901495392, "learning_rate": 2.5788671830799227e-06, "loss": 0.0532, "step": 12085 }, { "epoch": 7.732565579014715, "grad_norm": 0.8583608292225827, "learning_rate": 2.577478430483734e-06, "loss": 0.0545, "step": 12086 }, { "epoch": 7.733205374280231, "grad_norm": 0.9260349334089135, "learning_rate": 2.5760899965974173e-06, "loss": 0.0514, "step": 12087 }, { "epoch": 7.733845169545745, "grad_norm": 0.7930866444702289, "learning_rate": 2.57470188148059e-06, "loss": 0.0501, "step": 12088 }, { "epoch": 7.73448496481126, "grad_norm": 0.8208519992246325, "learning_rate": 2.573314085192855e-06, "loss": 0.0516, "step": 12089 }, { "epoch": 7.735124760076776, "grad_norm": 0.8497787951134927, "learning_rate": 2.5719266077938e-06, "loss": 0.0566, "step": 12090 }, { "epoch": 7.735764555342291, "grad_norm": 0.9482046142015708, "learning_rate": 2.5705394493430056e-06, "loss": 0.054, "step": 12091 }, { "epoch": 7.736404350607805, "grad_norm": 0.8628094049981205, "learning_rate": 2.5691526099000308e-06, "loss": 0.0515, "step": 12092 }, { "epoch": 7.737044145873321, "grad_norm": 0.7788078436651285, "learning_rate": 2.567766089524425e-06, "loss": 0.0467, "step": 12093 }, { "epoch": 7.737683941138836, "grad_norm": 0.8211062157742757, "learning_rate": 2.5663798882757185e-06, "loss": 0.049, "step": 12094 }, { "epoch": 7.73832373640435, "grad_norm": 0.8079871492007407, "learning_rate": 2.5649940062134405e-06, "loss": 0.0545, "step": 12095 }, { "epoch": 7.7389635316698655, "grad_norm": 0.8428306734149491, "learning_rate": 2.563608443397093e-06, "loss": 0.0472, "step": 12096 }, { "epoch": 7.739603326935381, "grad_norm": 0.8841647202222132, "learning_rate": 2.5622231998861678e-06, "loss": 0.0511, "step": 12097 }, { "epoch": 7.740243122200896, "grad_norm": 0.9578035313851541, "learning_rate": 2.560838275740152e-06, "loss": 0.0521, "step": 12098 }, { "epoch": 7.74088291746641, "grad_norm": 0.8553558945211523, "learning_rate": 2.559453671018506e-06, "loss": 0.0544, "step": 12099 }, { "epoch": 7.741522712731926, "grad_norm": 0.8882986114883324, "learning_rate": 2.558069385780684e-06, "loss": 0.0573, "step": 12100 }, { "epoch": 7.742162507997441, "grad_norm": 0.8355044316749408, "learning_rate": 2.5566854200861213e-06, "loss": 0.0521, "step": 12101 }, { "epoch": 7.742802303262955, "grad_norm": 0.8530841839459762, "learning_rate": 2.555301773994249e-06, "loss": 0.0549, "step": 12102 }, { "epoch": 7.743442098528471, "grad_norm": 0.7788139577275568, "learning_rate": 2.553918447564475e-06, "loss": 0.0483, "step": 12103 }, { "epoch": 7.744081893793986, "grad_norm": 0.7944002894458744, "learning_rate": 2.5525354408561974e-06, "loss": 0.0476, "step": 12104 }, { "epoch": 7.744721689059501, "grad_norm": 0.8600329889560314, "learning_rate": 2.5511527539288007e-06, "loss": 0.0519, "step": 12105 }, { "epoch": 7.7453614843250165, "grad_norm": 0.8198244457387982, "learning_rate": 2.5497703868416544e-06, "loss": 0.0468, "step": 12106 }, { "epoch": 7.746001279590531, "grad_norm": 0.8077514112769288, "learning_rate": 2.548388339654111e-06, "loss": 0.0559, "step": 12107 }, { "epoch": 7.746641074856046, "grad_norm": 0.7958094257804827, "learning_rate": 2.54700661242552e-06, "loss": 0.0535, "step": 12108 }, { "epoch": 7.747280870121561, "grad_norm": 0.9412996135198384, "learning_rate": 2.5456252052152074e-06, "loss": 0.0568, "step": 12109 }, { "epoch": 7.747920665387076, "grad_norm": 0.8682494272153681, "learning_rate": 2.5442441180824886e-06, "loss": 0.0498, "step": 12110 }, { "epoch": 7.748560460652591, "grad_norm": 0.8974658049812935, "learning_rate": 2.5428633510866608e-06, "loss": 0.0568, "step": 12111 }, { "epoch": 7.749200255918106, "grad_norm": 0.8817926978902489, "learning_rate": 2.541482904287019e-06, "loss": 0.0532, "step": 12112 }, { "epoch": 7.749840051183622, "grad_norm": 0.8626769731283357, "learning_rate": 2.5401027777428333e-06, "loss": 0.055, "step": 12113 }, { "epoch": 7.750479846449136, "grad_norm": 0.8373820292824433, "learning_rate": 2.538722971513362e-06, "loss": 0.0534, "step": 12114 }, { "epoch": 7.751119641714651, "grad_norm": 0.916209455114244, "learning_rate": 2.537343485657856e-06, "loss": 0.0585, "step": 12115 }, { "epoch": 7.751759436980167, "grad_norm": 0.8333131757394266, "learning_rate": 2.5359643202355456e-06, "loss": 0.0519, "step": 12116 }, { "epoch": 7.752399232245681, "grad_norm": 0.8411492691155955, "learning_rate": 2.53458547530565e-06, "loss": 0.0548, "step": 12117 }, { "epoch": 7.753039027511196, "grad_norm": 0.8221713344316143, "learning_rate": 2.533206950927375e-06, "loss": 0.053, "step": 12118 }, { "epoch": 7.7536788227767115, "grad_norm": 0.8646679536221324, "learning_rate": 2.53182874715991e-06, "loss": 0.053, "step": 12119 }, { "epoch": 7.754318618042227, "grad_norm": 0.8263573383551341, "learning_rate": 2.5304508640624338e-06, "loss": 0.0532, "step": 12120 }, { "epoch": 7.754958413307741, "grad_norm": 0.9005799950698232, "learning_rate": 2.529073301694107e-06, "loss": 0.0504, "step": 12121 }, { "epoch": 7.755598208573256, "grad_norm": 0.9109726542870398, "learning_rate": 2.5276960601140865e-06, "loss": 0.0537, "step": 12122 }, { "epoch": 7.756238003838772, "grad_norm": 0.9422768610078054, "learning_rate": 2.5263191393815057e-06, "loss": 0.0621, "step": 12123 }, { "epoch": 7.756877799104287, "grad_norm": 0.841800688837468, "learning_rate": 2.5249425395554817e-06, "loss": 0.0523, "step": 12124 }, { "epoch": 7.757517594369801, "grad_norm": 0.7436334713143139, "learning_rate": 2.523566260695132e-06, "loss": 0.0473, "step": 12125 }, { "epoch": 7.758157389635317, "grad_norm": 0.7504463954361259, "learning_rate": 2.5221903028595474e-06, "loss": 0.0497, "step": 12126 }, { "epoch": 7.758797184900832, "grad_norm": 0.9127672502271308, "learning_rate": 2.520814666107806e-06, "loss": 0.0494, "step": 12127 }, { "epoch": 7.759436980166347, "grad_norm": 0.9316075398908713, "learning_rate": 2.519439350498981e-06, "loss": 0.0519, "step": 12128 }, { "epoch": 7.760076775431862, "grad_norm": 0.9019360300389043, "learning_rate": 2.5180643560921245e-06, "loss": 0.0526, "step": 12129 }, { "epoch": 7.760716570697377, "grad_norm": 0.8075570178337775, "learning_rate": 2.516689682946275e-06, "loss": 0.0464, "step": 12130 }, { "epoch": 7.761356365962892, "grad_norm": 0.8652644314411918, "learning_rate": 2.515315331120455e-06, "loss": 0.0555, "step": 12131 }, { "epoch": 7.7619961612284065, "grad_norm": 0.8315811255568731, "learning_rate": 2.5139413006736835e-06, "loss": 0.0509, "step": 12132 }, { "epoch": 7.762635956493922, "grad_norm": 0.8430841190942243, "learning_rate": 2.512567591664956e-06, "loss": 0.05, "step": 12133 }, { "epoch": 7.763275751759437, "grad_norm": 0.900972608695804, "learning_rate": 2.511194204153258e-06, "loss": 0.0544, "step": 12134 }, { "epoch": 7.763915547024952, "grad_norm": 0.8724433348494021, "learning_rate": 2.509821138197558e-06, "loss": 0.0518, "step": 12135 }, { "epoch": 7.764555342290467, "grad_norm": 0.8892129855986943, "learning_rate": 2.508448393856815e-06, "loss": 0.0518, "step": 12136 }, { "epoch": 7.765195137555982, "grad_norm": 0.8407977195875415, "learning_rate": 2.507075971189968e-06, "loss": 0.0528, "step": 12137 }, { "epoch": 7.765834932821497, "grad_norm": 0.7489024817785406, "learning_rate": 2.5057038702559533e-06, "loss": 0.049, "step": 12138 }, { "epoch": 7.766474728087012, "grad_norm": 0.8007389184676131, "learning_rate": 2.5043320911136836e-06, "loss": 0.0504, "step": 12139 }, { "epoch": 7.767114523352527, "grad_norm": 0.7757230687333794, "learning_rate": 2.5029606338220593e-06, "loss": 0.0506, "step": 12140 }, { "epoch": 7.767754318618042, "grad_norm": 0.8554337519549252, "learning_rate": 2.501589498439967e-06, "loss": 0.05, "step": 12141 }, { "epoch": 7.7683941138835575, "grad_norm": 0.7984751410222819, "learning_rate": 2.5002186850262857e-06, "loss": 0.048, "step": 12142 }, { "epoch": 7.769033909149073, "grad_norm": 0.850104944383768, "learning_rate": 2.4988481936398723e-06, "loss": 0.0522, "step": 12143 }, { "epoch": 7.769673704414587, "grad_norm": 0.81827819731967, "learning_rate": 2.4974780243395714e-06, "loss": 0.0527, "step": 12144 }, { "epoch": 7.770313499680102, "grad_norm": 0.9012091166879997, "learning_rate": 2.496108177184222e-06, "loss": 0.0547, "step": 12145 }, { "epoch": 7.770953294945618, "grad_norm": 0.8536889510903287, "learning_rate": 2.4947386522326376e-06, "loss": 0.0518, "step": 12146 }, { "epoch": 7.771593090211132, "grad_norm": 0.9185626701353186, "learning_rate": 2.4933694495436255e-06, "loss": 0.0582, "step": 12147 }, { "epoch": 7.772232885476647, "grad_norm": 0.8342399496563871, "learning_rate": 2.4920005691759765e-06, "loss": 0.0487, "step": 12148 }, { "epoch": 7.772872680742163, "grad_norm": 0.8956155982469056, "learning_rate": 2.490632011188466e-06, "loss": 0.0572, "step": 12149 }, { "epoch": 7.773512476007678, "grad_norm": 0.7205685689571156, "learning_rate": 2.4892637756398597e-06, "loss": 0.0429, "step": 12150 }, { "epoch": 7.774152271273192, "grad_norm": 0.8485416173017309, "learning_rate": 2.487895862588904e-06, "loss": 0.0493, "step": 12151 }, { "epoch": 7.774792066538708, "grad_norm": 0.8304735297238283, "learning_rate": 2.4865282720943386e-06, "loss": 0.0499, "step": 12152 }, { "epoch": 7.775431861804223, "grad_norm": 0.8920540453845108, "learning_rate": 2.4851610042148846e-06, "loss": 0.0516, "step": 12153 }, { "epoch": 7.776071657069737, "grad_norm": 0.9115691233151224, "learning_rate": 2.4837940590092458e-06, "loss": 0.0558, "step": 12154 }, { "epoch": 7.7767114523352525, "grad_norm": 0.9203107846028534, "learning_rate": 2.4824274365361225e-06, "loss": 0.0515, "step": 12155 }, { "epoch": 7.777351247600768, "grad_norm": 0.8065571070955236, "learning_rate": 2.4810611368541924e-06, "loss": 0.05, "step": 12156 }, { "epoch": 7.777991042866283, "grad_norm": 0.8330632250887714, "learning_rate": 2.47969516002212e-06, "loss": 0.0533, "step": 12157 }, { "epoch": 7.7786308381317975, "grad_norm": 0.8090080118406091, "learning_rate": 2.478329506098562e-06, "loss": 0.0541, "step": 12158 }, { "epoch": 7.779270633397313, "grad_norm": 0.8854764606336659, "learning_rate": 2.4769641751421545e-06, "loss": 0.0498, "step": 12159 }, { "epoch": 7.779910428662828, "grad_norm": 0.7902281984605248, "learning_rate": 2.475599167211524e-06, "loss": 0.0457, "step": 12160 }, { "epoch": 7.780550223928343, "grad_norm": 0.9063315187382461, "learning_rate": 2.4742344823652773e-06, "loss": 0.0552, "step": 12161 }, { "epoch": 7.781190019193858, "grad_norm": 0.8242361184214274, "learning_rate": 2.4728701206620165e-06, "loss": 0.048, "step": 12162 }, { "epoch": 7.781829814459373, "grad_norm": 0.8149338812089408, "learning_rate": 2.471506082160324e-06, "loss": 0.0484, "step": 12163 }, { "epoch": 7.782469609724888, "grad_norm": 0.8322262801805422, "learning_rate": 2.4701423669187686e-06, "loss": 0.0552, "step": 12164 }, { "epoch": 7.7831094049904035, "grad_norm": 0.9508696711149035, "learning_rate": 2.4687789749959046e-06, "loss": 0.0553, "step": 12165 }, { "epoch": 7.783749200255918, "grad_norm": 0.8167177807671165, "learning_rate": 2.4674159064502744e-06, "loss": 0.0456, "step": 12166 }, { "epoch": 7.784388995521433, "grad_norm": 0.9338522288642126, "learning_rate": 2.4660531613404024e-06, "loss": 0.0531, "step": 12167 }, { "epoch": 7.785028790786948, "grad_norm": 0.7547686541360238, "learning_rate": 2.4646907397248098e-06, "loss": 0.0483, "step": 12168 }, { "epoch": 7.785668586052463, "grad_norm": 0.7559379109221553, "learning_rate": 2.4633286416619916e-06, "loss": 0.0449, "step": 12169 }, { "epoch": 7.786308381317978, "grad_norm": 0.8773574888375065, "learning_rate": 2.461966867210435e-06, "loss": 0.0554, "step": 12170 }, { "epoch": 7.786948176583493, "grad_norm": 0.8665558461013129, "learning_rate": 2.4606054164286096e-06, "loss": 0.0548, "step": 12171 }, { "epoch": 7.787587971849009, "grad_norm": 0.8920539691452533, "learning_rate": 2.459244289374979e-06, "loss": 0.054, "step": 12172 }, { "epoch": 7.788227767114523, "grad_norm": 0.7917936211595993, "learning_rate": 2.457883486107984e-06, "loss": 0.048, "step": 12173 }, { "epoch": 7.788867562380038, "grad_norm": 0.737700147857322, "learning_rate": 2.4565230066860534e-06, "loss": 0.0463, "step": 12174 }, { "epoch": 7.789507357645554, "grad_norm": 0.8664184674043718, "learning_rate": 2.4551628511676086e-06, "loss": 0.0487, "step": 12175 }, { "epoch": 7.790147152911068, "grad_norm": 0.7553402811728052, "learning_rate": 2.4538030196110497e-06, "loss": 0.0469, "step": 12176 }, { "epoch": 7.790786948176583, "grad_norm": 0.9244436867140597, "learning_rate": 2.452443512074766e-06, "loss": 0.0577, "step": 12177 }, { "epoch": 7.7914267434420985, "grad_norm": 0.8996279593324823, "learning_rate": 2.4510843286171317e-06, "loss": 0.0571, "step": 12178 }, { "epoch": 7.792066538707614, "grad_norm": 0.8536937455998793, "learning_rate": 2.4497254692965077e-06, "loss": 0.055, "step": 12179 }, { "epoch": 7.792706333973129, "grad_norm": 0.7742831289727936, "learning_rate": 2.4483669341712403e-06, "loss": 0.0445, "step": 12180 }, { "epoch": 7.7933461292386434, "grad_norm": 0.9442153201662974, "learning_rate": 2.447008723299662e-06, "loss": 0.0634, "step": 12181 }, { "epoch": 7.793985924504159, "grad_norm": 0.8851209452044941, "learning_rate": 2.4456508367400956e-06, "loss": 0.0551, "step": 12182 }, { "epoch": 7.794625719769674, "grad_norm": 0.9010105048727433, "learning_rate": 2.444293274550844e-06, "loss": 0.0499, "step": 12183 }, { "epoch": 7.795265515035188, "grad_norm": 0.9156495852093534, "learning_rate": 2.4429360367901956e-06, "loss": 0.0544, "step": 12184 }, { "epoch": 7.795905310300704, "grad_norm": 0.7710542415147951, "learning_rate": 2.4415791235164345e-06, "loss": 0.0484, "step": 12185 }, { "epoch": 7.796545105566219, "grad_norm": 0.7405574853124185, "learning_rate": 2.4402225347878196e-06, "loss": 0.0471, "step": 12186 }, { "epoch": 7.797184900831734, "grad_norm": 0.814584726218814, "learning_rate": 2.438866270662602e-06, "loss": 0.057, "step": 12187 }, { "epoch": 7.797824696097249, "grad_norm": 0.8714939270925957, "learning_rate": 2.437510331199013e-06, "loss": 0.0529, "step": 12188 }, { "epoch": 7.798464491362764, "grad_norm": 0.9651618869830072, "learning_rate": 2.436154716455281e-06, "loss": 0.0543, "step": 12189 }, { "epoch": 7.799104286628279, "grad_norm": 0.7883153654844139, "learning_rate": 2.4347994264896102e-06, "loss": 0.0458, "step": 12190 }, { "epoch": 7.7997440818937935, "grad_norm": 0.8707572605793851, "learning_rate": 2.4334444613601915e-06, "loss": 0.0509, "step": 12191 }, { "epoch": 7.800383877159309, "grad_norm": 0.8727749219547567, "learning_rate": 2.432089821125212e-06, "loss": 0.0538, "step": 12192 }, { "epoch": 7.801023672424824, "grad_norm": 0.7758882103593241, "learning_rate": 2.430735505842832e-06, "loss": 0.0509, "step": 12193 }, { "epoch": 7.801663467690339, "grad_norm": 0.8474339486812754, "learning_rate": 2.429381515571205e-06, "loss": 0.0559, "step": 12194 }, { "epoch": 7.802303262955854, "grad_norm": 0.8307557498650971, "learning_rate": 2.4280278503684685e-06, "loss": 0.0521, "step": 12195 }, { "epoch": 7.802943058221369, "grad_norm": 0.8114680142918618, "learning_rate": 2.426674510292747e-06, "loss": 0.0541, "step": 12196 }, { "epoch": 7.803582853486884, "grad_norm": 0.8611044808290667, "learning_rate": 2.4253214954021497e-06, "loss": 0.0558, "step": 12197 }, { "epoch": 7.8042226487524, "grad_norm": 0.9256349011618945, "learning_rate": 2.4239688057547705e-06, "loss": 0.0558, "step": 12198 }, { "epoch": 7.804862444017914, "grad_norm": 0.8331291894497492, "learning_rate": 2.422616441408696e-06, "loss": 0.053, "step": 12199 }, { "epoch": 7.805502239283429, "grad_norm": 0.9373933480075164, "learning_rate": 2.4212644024219933e-06, "loss": 0.0531, "step": 12200 }, { "epoch": 7.8061420345489445, "grad_norm": 0.8157616233648974, "learning_rate": 2.4199126888527125e-06, "loss": 0.0509, "step": 12201 }, { "epoch": 7.80678182981446, "grad_norm": 0.8933786535428258, "learning_rate": 2.4185613007589003e-06, "loss": 0.0553, "step": 12202 }, { "epoch": 7.807421625079974, "grad_norm": 0.8780247778368704, "learning_rate": 2.4172102381985797e-06, "loss": 0.0571, "step": 12203 }, { "epoch": 7.8080614203454894, "grad_norm": 0.828397917899598, "learning_rate": 2.4158595012297594e-06, "loss": 0.0494, "step": 12204 }, { "epoch": 7.808701215611005, "grad_norm": 0.8091703130847302, "learning_rate": 2.414509089910443e-06, "loss": 0.0496, "step": 12205 }, { "epoch": 7.809341010876519, "grad_norm": 0.8728506689633575, "learning_rate": 2.4131590042986142e-06, "loss": 0.0565, "step": 12206 }, { "epoch": 7.809980806142034, "grad_norm": 0.9766908091264295, "learning_rate": 2.411809244452241e-06, "loss": 0.0589, "step": 12207 }, { "epoch": 7.81062060140755, "grad_norm": 0.8590104996024788, "learning_rate": 2.4104598104292807e-06, "loss": 0.0503, "step": 12208 }, { "epoch": 7.811260396673065, "grad_norm": 0.8498641025771753, "learning_rate": 2.409110702287675e-06, "loss": 0.0553, "step": 12209 }, { "epoch": 7.811900191938579, "grad_norm": 0.8064250539090267, "learning_rate": 2.407761920085353e-06, "loss": 0.0539, "step": 12210 }, { "epoch": 7.812539987204095, "grad_norm": 0.8366587691044326, "learning_rate": 2.4064134638802262e-06, "loss": 0.052, "step": 12211 }, { "epoch": 7.81317978246961, "grad_norm": 0.6809591219729711, "learning_rate": 2.4050653337301987e-06, "loss": 0.0438, "step": 12212 }, { "epoch": 7.813819577735125, "grad_norm": 0.842470099364067, "learning_rate": 2.403717529693157e-06, "loss": 0.0549, "step": 12213 }, { "epoch": 7.8144593730006395, "grad_norm": 0.799180436570866, "learning_rate": 2.4023700518269675e-06, "loss": 0.0464, "step": 12214 }, { "epoch": 7.815099168266155, "grad_norm": 0.7508551747178731, "learning_rate": 2.401022900189497e-06, "loss": 0.0503, "step": 12215 }, { "epoch": 7.81573896353167, "grad_norm": 0.7964476288132543, "learning_rate": 2.399676074838585e-06, "loss": 0.0533, "step": 12216 }, { "epoch": 7.816378758797185, "grad_norm": 0.9259396835755691, "learning_rate": 2.3983295758320623e-06, "loss": 0.059, "step": 12217 }, { "epoch": 7.8170185540627, "grad_norm": 0.8715138730453882, "learning_rate": 2.3969834032277427e-06, "loss": 0.051, "step": 12218 }, { "epoch": 7.817658349328215, "grad_norm": 0.9504177158181539, "learning_rate": 2.3956375570834346e-06, "loss": 0.0575, "step": 12219 }, { "epoch": 7.81829814459373, "grad_norm": 0.9827161931288356, "learning_rate": 2.394292037456922e-06, "loss": 0.0512, "step": 12220 }, { "epoch": 7.818937939859245, "grad_norm": 0.870578403515796, "learning_rate": 2.3929468444059766e-06, "loss": 0.0511, "step": 12221 }, { "epoch": 7.81957773512476, "grad_norm": 1.054386147201424, "learning_rate": 2.391601977988366e-06, "loss": 0.0515, "step": 12222 }, { "epoch": 7.820217530390275, "grad_norm": 0.8618514268860963, "learning_rate": 2.3902574382618304e-06, "loss": 0.0525, "step": 12223 }, { "epoch": 7.8208573256557905, "grad_norm": 0.8287044806482251, "learning_rate": 2.388913225284105e-06, "loss": 0.0597, "step": 12224 }, { "epoch": 7.821497120921305, "grad_norm": 0.8372589838501605, "learning_rate": 2.387569339112906e-06, "loss": 0.047, "step": 12225 }, { "epoch": 7.82213691618682, "grad_norm": 0.8409705306829631, "learning_rate": 2.3862257798059384e-06, "loss": 0.0528, "step": 12226 }, { "epoch": 7.822776711452335, "grad_norm": 0.8408531028859584, "learning_rate": 2.3848825474208912e-06, "loss": 0.0528, "step": 12227 }, { "epoch": 7.82341650671785, "grad_norm": 0.8125557522456682, "learning_rate": 2.383539642015439e-06, "loss": 0.0589, "step": 12228 }, { "epoch": 7.824056301983365, "grad_norm": 0.8218960558224769, "learning_rate": 2.382197063647249e-06, "loss": 0.0535, "step": 12229 }, { "epoch": 7.82469609724888, "grad_norm": 0.8019480803970731, "learning_rate": 2.3808548123739662e-06, "loss": 0.0488, "step": 12230 }, { "epoch": 7.825335892514396, "grad_norm": 0.9611553322076453, "learning_rate": 2.3795128882532205e-06, "loss": 0.0547, "step": 12231 }, { "epoch": 7.825975687779911, "grad_norm": 0.8729684737068093, "learning_rate": 2.3781712913426393e-06, "loss": 0.054, "step": 12232 }, { "epoch": 7.826615483045425, "grad_norm": 0.8204771964472762, "learning_rate": 2.376830021699824e-06, "loss": 0.049, "step": 12233 }, { "epoch": 7.827255278310941, "grad_norm": 0.8410392130716969, "learning_rate": 2.375489079382365e-06, "loss": 0.0537, "step": 12234 }, { "epoch": 7.827895073576456, "grad_norm": 0.8487624113679043, "learning_rate": 2.374148464447844e-06, "loss": 0.055, "step": 12235 }, { "epoch": 7.82853486884197, "grad_norm": 0.8839393437175319, "learning_rate": 2.3728081769538235e-06, "loss": 0.0581, "step": 12236 }, { "epoch": 7.8291746641074855, "grad_norm": 0.872751185633825, "learning_rate": 2.371468216957852e-06, "loss": 0.0546, "step": 12237 }, { "epoch": 7.829814459373001, "grad_norm": 0.8529155193295278, "learning_rate": 2.370128584517465e-06, "loss": 0.0494, "step": 12238 }, { "epoch": 7.830454254638516, "grad_norm": 0.790572994788134, "learning_rate": 2.3687892796901835e-06, "loss": 0.0464, "step": 12239 }, { "epoch": 7.8310940499040305, "grad_norm": 0.7751375956092432, "learning_rate": 2.367450302533516e-06, "loss": 0.0447, "step": 12240 }, { "epoch": 7.831733845169546, "grad_norm": 0.8678541426275842, "learning_rate": 2.3661116531049524e-06, "loss": 0.049, "step": 12241 }, { "epoch": 7.832373640435061, "grad_norm": 0.8366949344188009, "learning_rate": 2.364773331461977e-06, "loss": 0.0548, "step": 12242 }, { "epoch": 7.833013435700575, "grad_norm": 0.7692699843721164, "learning_rate": 2.3634353376620535e-06, "loss": 0.0501, "step": 12243 }, { "epoch": 7.833653230966091, "grad_norm": 0.8952010367413926, "learning_rate": 2.3620976717626286e-06, "loss": 0.0565, "step": 12244 }, { "epoch": 7.834293026231606, "grad_norm": 0.7758758154741987, "learning_rate": 2.360760333821146e-06, "loss": 0.0492, "step": 12245 }, { "epoch": 7.834932821497121, "grad_norm": 0.7905109744489203, "learning_rate": 2.359423323895026e-06, "loss": 0.0466, "step": 12246 }, { "epoch": 7.835572616762636, "grad_norm": 0.9383172500044971, "learning_rate": 2.358086642041677e-06, "loss": 0.0547, "step": 12247 }, { "epoch": 7.836212412028151, "grad_norm": 0.8442073661115358, "learning_rate": 2.35675028831849e-06, "loss": 0.0568, "step": 12248 }, { "epoch": 7.836852207293666, "grad_norm": 0.827717124845587, "learning_rate": 2.3554142627828535e-06, "loss": 0.0557, "step": 12249 }, { "epoch": 7.837492002559181, "grad_norm": 0.8181636673640541, "learning_rate": 2.3540785654921284e-06, "loss": 0.0453, "step": 12250 }, { "epoch": 7.838131797824696, "grad_norm": 0.8705475366334071, "learning_rate": 2.3527431965036673e-06, "loss": 0.059, "step": 12251 }, { "epoch": 7.838771593090211, "grad_norm": 0.9339055067431523, "learning_rate": 2.3514081558748115e-06, "loss": 0.0557, "step": 12252 }, { "epoch": 7.839411388355726, "grad_norm": 0.8968455905651384, "learning_rate": 2.3500734436628847e-06, "loss": 0.0537, "step": 12253 }, { "epoch": 7.840051183621242, "grad_norm": 0.893225858834723, "learning_rate": 2.3487390599251946e-06, "loss": 0.0539, "step": 12254 }, { "epoch": 7.840690978886756, "grad_norm": 0.9002039128559872, "learning_rate": 2.34740500471904e-06, "loss": 0.0542, "step": 12255 }, { "epoch": 7.841330774152271, "grad_norm": 0.8426373302395849, "learning_rate": 2.3460712781017016e-06, "loss": 0.0507, "step": 12256 }, { "epoch": 7.841970569417787, "grad_norm": 0.9287787771982234, "learning_rate": 2.3447378801304464e-06, "loss": 0.0574, "step": 12257 }, { "epoch": 7.842610364683301, "grad_norm": 0.8607879448367977, "learning_rate": 2.3434048108625275e-06, "loss": 0.0507, "step": 12258 }, { "epoch": 7.843250159948816, "grad_norm": 0.8106364012019757, "learning_rate": 2.342072070355187e-06, "loss": 0.0505, "step": 12259 }, { "epoch": 7.8438899552143315, "grad_norm": 0.9485885586700006, "learning_rate": 2.3407396586656505e-06, "loss": 0.0609, "step": 12260 }, { "epoch": 7.844529750479847, "grad_norm": 0.7810836503561307, "learning_rate": 2.339407575851126e-06, "loss": 0.0504, "step": 12261 }, { "epoch": 7.845169545745361, "grad_norm": 0.9213091420040184, "learning_rate": 2.338075821968815e-06, "loss": 0.0491, "step": 12262 }, { "epoch": 7.8458093410108765, "grad_norm": 0.7363250282154767, "learning_rate": 2.3367443970758984e-06, "loss": 0.0463, "step": 12263 }, { "epoch": 7.846449136276392, "grad_norm": 0.8357985167667323, "learning_rate": 2.3354133012295468e-06, "loss": 0.0517, "step": 12264 }, { "epoch": 7.847088931541906, "grad_norm": 0.8159699110865591, "learning_rate": 2.33408253448691e-06, "loss": 0.0462, "step": 12265 }, { "epoch": 7.847728726807421, "grad_norm": 0.854190948414737, "learning_rate": 2.3327520969051354e-06, "loss": 0.0555, "step": 12266 }, { "epoch": 7.848368522072937, "grad_norm": 0.8064186573142117, "learning_rate": 2.3314219885413468e-06, "loss": 0.0525, "step": 12267 }, { "epoch": 7.849008317338452, "grad_norm": 0.8115017624948347, "learning_rate": 2.330092209452657e-06, "loss": 0.0481, "step": 12268 }, { "epoch": 7.849648112603967, "grad_norm": 0.8777928899250019, "learning_rate": 2.328762759696164e-06, "loss": 0.0559, "step": 12269 }, { "epoch": 7.850287907869482, "grad_norm": 0.7789968645091592, "learning_rate": 2.327433639328952e-06, "loss": 0.0465, "step": 12270 }, { "epoch": 7.850927703134997, "grad_norm": 0.8744636751043683, "learning_rate": 2.3261048484080885e-06, "loss": 0.0529, "step": 12271 }, { "epoch": 7.851567498400512, "grad_norm": 0.869115245963292, "learning_rate": 2.3247763869906337e-06, "loss": 0.0496, "step": 12272 }, { "epoch": 7.8522072936660265, "grad_norm": 0.8652697950869759, "learning_rate": 2.323448255133629e-06, "loss": 0.0543, "step": 12273 }, { "epoch": 7.852847088931542, "grad_norm": 0.8495684818201291, "learning_rate": 2.3221204528941e-06, "loss": 0.0493, "step": 12274 }, { "epoch": 7.853486884197057, "grad_norm": 0.8252716461782047, "learning_rate": 2.3207929803290573e-06, "loss": 0.0532, "step": 12275 }, { "epoch": 7.854126679462572, "grad_norm": 0.7856172035018368, "learning_rate": 2.319465837495508e-06, "loss": 0.0532, "step": 12276 }, { "epoch": 7.854766474728087, "grad_norm": 0.8557860735212967, "learning_rate": 2.318139024450433e-06, "loss": 0.049, "step": 12277 }, { "epoch": 7.855406269993602, "grad_norm": 0.8804234668880981, "learning_rate": 2.3168125412507993e-06, "loss": 0.0541, "step": 12278 }, { "epoch": 7.856046065259117, "grad_norm": 0.8273305924645094, "learning_rate": 2.3154863879535714e-06, "loss": 0.0522, "step": 12279 }, { "epoch": 7.856685860524632, "grad_norm": 0.8396254512716959, "learning_rate": 2.3141605646156882e-06, "loss": 0.0498, "step": 12280 }, { "epoch": 7.857325655790147, "grad_norm": 0.7905126355182046, "learning_rate": 2.3128350712940763e-06, "loss": 0.0467, "step": 12281 }, { "epoch": 7.857965451055662, "grad_norm": 0.8400142231306049, "learning_rate": 2.311509908045654e-06, "loss": 0.0496, "step": 12282 }, { "epoch": 7.8586052463211775, "grad_norm": 1.0105180009613333, "learning_rate": 2.310185074927319e-06, "loss": 0.0579, "step": 12283 }, { "epoch": 7.859245041586692, "grad_norm": 0.8582471929156106, "learning_rate": 2.3088605719959577e-06, "loss": 0.0531, "step": 12284 }, { "epoch": 7.859884836852207, "grad_norm": 0.9505959493663062, "learning_rate": 2.307536399308443e-06, "loss": 0.0477, "step": 12285 }, { "epoch": 7.8605246321177225, "grad_norm": 1.0142011329244585, "learning_rate": 2.306212556921631e-06, "loss": 0.0619, "step": 12286 }, { "epoch": 7.861164427383238, "grad_norm": 0.7419541271011587, "learning_rate": 2.3048890448923655e-06, "loss": 0.0424, "step": 12287 }, { "epoch": 7.861804222648752, "grad_norm": 0.9535000193502436, "learning_rate": 2.303565863277474e-06, "loss": 0.0501, "step": 12288 }, { "epoch": 7.862444017914267, "grad_norm": 0.8599058036156217, "learning_rate": 2.302243012133775e-06, "loss": 0.0499, "step": 12289 }, { "epoch": 7.863083813179783, "grad_norm": 0.8869797099340468, "learning_rate": 2.3009204915180684e-06, "loss": 0.0588, "step": 12290 }, { "epoch": 7.863723608445298, "grad_norm": 0.8413561218461929, "learning_rate": 2.299598301487137e-06, "loss": 0.0503, "step": 12291 }, { "epoch": 7.864363403710812, "grad_norm": 0.8860437983732616, "learning_rate": 2.29827644209776e-06, "loss": 0.0511, "step": 12292 }, { "epoch": 7.865003198976328, "grad_norm": 0.7320261308381457, "learning_rate": 2.2969549134066927e-06, "loss": 0.0421, "step": 12293 }, { "epoch": 7.865642994241843, "grad_norm": 0.8424942730994889, "learning_rate": 2.2956337154706777e-06, "loss": 0.0481, "step": 12294 }, { "epoch": 7.866282789507357, "grad_norm": 0.9257247903662217, "learning_rate": 2.2943128483464426e-06, "loss": 0.055, "step": 12295 }, { "epoch": 7.8669225847728725, "grad_norm": 0.8843203870559782, "learning_rate": 2.2929923120907104e-06, "loss": 0.0547, "step": 12296 }, { "epoch": 7.867562380038388, "grad_norm": 0.9081843661819905, "learning_rate": 2.291672106760179e-06, "loss": 0.051, "step": 12297 }, { "epoch": 7.868202175303903, "grad_norm": 0.8895885651292379, "learning_rate": 2.2903522324115344e-06, "loss": 0.0528, "step": 12298 }, { "epoch": 7.8688419705694175, "grad_norm": 0.7686375259106837, "learning_rate": 2.2890326891014516e-06, "loss": 0.0476, "step": 12299 }, { "epoch": 7.869481765834933, "grad_norm": 0.7991678622123759, "learning_rate": 2.287713476886587e-06, "loss": 0.0491, "step": 12300 }, { "epoch": 7.870121561100448, "grad_norm": 0.8361437747140276, "learning_rate": 2.2863945958235843e-06, "loss": 0.0495, "step": 12301 }, { "epoch": 7.870761356365962, "grad_norm": 0.861769797602796, "learning_rate": 2.285076045969079e-06, "loss": 0.055, "step": 12302 }, { "epoch": 7.871401151631478, "grad_norm": 0.7737271966007898, "learning_rate": 2.2837578273796847e-06, "loss": 0.0457, "step": 12303 }, { "epoch": 7.872040946896993, "grad_norm": 0.8389552167070712, "learning_rate": 2.2824399401120036e-06, "loss": 0.0483, "step": 12304 }, { "epoch": 7.872680742162508, "grad_norm": 0.9304437230219748, "learning_rate": 2.2811223842226193e-06, "loss": 0.0594, "step": 12305 }, { "epoch": 7.8733205374280235, "grad_norm": 0.8159843643179368, "learning_rate": 2.2798051597681127e-06, "loss": 0.0504, "step": 12306 }, { "epoch": 7.873960332693538, "grad_norm": 0.9164724640951253, "learning_rate": 2.2784882668050378e-06, "loss": 0.0546, "step": 12307 }, { "epoch": 7.874600127959053, "grad_norm": 0.8311440921788635, "learning_rate": 2.2771717053899402e-06, "loss": 0.0489, "step": 12308 }, { "epoch": 7.8752399232245685, "grad_norm": 0.9331804152292538, "learning_rate": 2.2758554755793536e-06, "loss": 0.0551, "step": 12309 }, { "epoch": 7.875879718490083, "grad_norm": 0.8189930985505718, "learning_rate": 2.274539577429793e-06, "loss": 0.0554, "step": 12310 }, { "epoch": 7.876519513755598, "grad_norm": 0.8225176711887103, "learning_rate": 2.2732240109977575e-06, "loss": 0.0505, "step": 12311 }, { "epoch": 7.877159309021113, "grad_norm": 0.7623795449647616, "learning_rate": 2.2719087763397405e-06, "loss": 0.0452, "step": 12312 }, { "epoch": 7.877799104286629, "grad_norm": 0.8210529341055963, "learning_rate": 2.2705938735122144e-06, "loss": 0.0482, "step": 12313 }, { "epoch": 7.878438899552143, "grad_norm": 0.9894255434011718, "learning_rate": 2.2692793025716377e-06, "loss": 0.0528, "step": 12314 }, { "epoch": 7.879078694817658, "grad_norm": 0.8151879079464456, "learning_rate": 2.2679650635744553e-06, "loss": 0.0485, "step": 12315 }, { "epoch": 7.879718490083174, "grad_norm": 0.8656079360725978, "learning_rate": 2.266651156577099e-06, "loss": 0.0536, "step": 12316 }, { "epoch": 7.880358285348688, "grad_norm": 0.8273203284743875, "learning_rate": 2.2653375816359845e-06, "loss": 0.0493, "step": 12317 }, { "epoch": 7.880998080614203, "grad_norm": 0.8158701772569592, "learning_rate": 2.2640243388075144e-06, "loss": 0.0519, "step": 12318 }, { "epoch": 7.8816378758797185, "grad_norm": 0.7733093403737032, "learning_rate": 2.26271142814808e-06, "loss": 0.0489, "step": 12319 }, { "epoch": 7.882277671145234, "grad_norm": 0.7808459847268898, "learning_rate": 2.2613988497140526e-06, "loss": 0.05, "step": 12320 }, { "epoch": 7.882917466410748, "grad_norm": 0.8294897479975399, "learning_rate": 2.2600866035617908e-06, "loss": 0.0491, "step": 12321 }, { "epoch": 7.8835572616762635, "grad_norm": 0.8410491917915968, "learning_rate": 2.258774689747645e-06, "loss": 0.0512, "step": 12322 }, { "epoch": 7.884197056941779, "grad_norm": 0.7793357849614234, "learning_rate": 2.257463108327942e-06, "loss": 0.0468, "step": 12323 }, { "epoch": 7.884836852207294, "grad_norm": 0.7423198983378915, "learning_rate": 2.2561518593590014e-06, "loss": 0.0489, "step": 12324 }, { "epoch": 7.885476647472808, "grad_norm": 0.8202820038952144, "learning_rate": 2.254840942897122e-06, "loss": 0.0528, "step": 12325 }, { "epoch": 7.886116442738324, "grad_norm": 0.8613734397980406, "learning_rate": 2.2535303589985967e-06, "loss": 0.0478, "step": 12326 }, { "epoch": 7.886756238003839, "grad_norm": 0.8282629410728285, "learning_rate": 2.2522201077196993e-06, "loss": 0.0497, "step": 12327 }, { "epoch": 7.887396033269354, "grad_norm": 0.8090521444420681, "learning_rate": 2.250910189116687e-06, "loss": 0.0477, "step": 12328 }, { "epoch": 7.888035828534869, "grad_norm": 0.881058894664801, "learning_rate": 2.2496006032458075e-06, "loss": 0.0505, "step": 12329 }, { "epoch": 7.888675623800384, "grad_norm": 0.8170556843731964, "learning_rate": 2.248291350163291e-06, "loss": 0.0542, "step": 12330 }, { "epoch": 7.889315419065899, "grad_norm": 0.7954900838023398, "learning_rate": 2.2469824299253538e-06, "loss": 0.0478, "step": 12331 }, { "epoch": 7.889955214331414, "grad_norm": 0.8550188667218536, "learning_rate": 2.2456738425881976e-06, "loss": 0.0551, "step": 12332 }, { "epoch": 7.890595009596929, "grad_norm": 0.8583516231427948, "learning_rate": 2.2443655882080162e-06, "loss": 0.0564, "step": 12333 }, { "epoch": 7.891234804862444, "grad_norm": 0.7705315260022805, "learning_rate": 2.2430576668409785e-06, "loss": 0.0489, "step": 12334 }, { "epoch": 7.891874600127959, "grad_norm": 0.9109609699805726, "learning_rate": 2.2417500785432444e-06, "loss": 0.0549, "step": 12335 }, { "epoch": 7.892514395393474, "grad_norm": 0.7542739080951948, "learning_rate": 2.240442823370964e-06, "loss": 0.0503, "step": 12336 }, { "epoch": 7.893154190658989, "grad_norm": 0.8377252316302759, "learning_rate": 2.239135901380265e-06, "loss": 0.0521, "step": 12337 }, { "epoch": 7.893793985924504, "grad_norm": 0.7987974707043701, "learning_rate": 2.2378293126272622e-06, "loss": 0.044, "step": 12338 }, { "epoch": 7.894433781190019, "grad_norm": 0.824461813327989, "learning_rate": 2.2365230571680642e-06, "loss": 0.0526, "step": 12339 }, { "epoch": 7.895073576455534, "grad_norm": 0.816707143404491, "learning_rate": 2.235217135058756e-06, "loss": 0.0486, "step": 12340 }, { "epoch": 7.895713371721049, "grad_norm": 0.8980584544210658, "learning_rate": 2.2339115463554104e-06, "loss": 0.0507, "step": 12341 }, { "epoch": 7.8963531669865645, "grad_norm": 0.9431093427070146, "learning_rate": 2.2326062911140865e-06, "loss": 0.0562, "step": 12342 }, { "epoch": 7.89699296225208, "grad_norm": 0.8130950362164289, "learning_rate": 2.2313013693908336e-06, "loss": 0.0483, "step": 12343 }, { "epoch": 7.897632757517594, "grad_norm": 0.7719828823664726, "learning_rate": 2.2299967812416813e-06, "loss": 0.0472, "step": 12344 }, { "epoch": 7.8982725527831095, "grad_norm": 0.8050394283976979, "learning_rate": 2.2286925267226446e-06, "loss": 0.0504, "step": 12345 }, { "epoch": 7.898912348048625, "grad_norm": 0.9360945448419815, "learning_rate": 2.2273886058897265e-06, "loss": 0.0573, "step": 12346 }, { "epoch": 7.899552143314139, "grad_norm": 0.8640616712120833, "learning_rate": 2.226085018798916e-06, "loss": 0.0527, "step": 12347 }, { "epoch": 7.900191938579654, "grad_norm": 0.838234435128685, "learning_rate": 2.224781765506182e-06, "loss": 0.057, "step": 12348 }, { "epoch": 7.90083173384517, "grad_norm": 0.8522852711413633, "learning_rate": 2.223478846067493e-06, "loss": 0.0501, "step": 12349 }, { "epoch": 7.901471529110685, "grad_norm": 0.8594992672667013, "learning_rate": 2.2221762605387866e-06, "loss": 0.0488, "step": 12350 }, { "epoch": 7.902111324376199, "grad_norm": 0.8609518190747893, "learning_rate": 2.2208740089759975e-06, "loss": 0.0454, "step": 12351 }, { "epoch": 7.902751119641715, "grad_norm": 0.9136856600222417, "learning_rate": 2.219572091435037e-06, "loss": 0.0521, "step": 12352 }, { "epoch": 7.90339091490723, "grad_norm": 0.9202648892700813, "learning_rate": 2.2182705079718135e-06, "loss": 0.0449, "step": 12353 }, { "epoch": 7.904030710172744, "grad_norm": 0.8884837517289231, "learning_rate": 2.216969258642212e-06, "loss": 0.0576, "step": 12354 }, { "epoch": 7.9046705054382596, "grad_norm": 0.8684867945537654, "learning_rate": 2.215668343502102e-06, "loss": 0.0516, "step": 12355 }, { "epoch": 7.905310300703775, "grad_norm": 0.848812125430395, "learning_rate": 2.2143677626073502e-06, "loss": 0.0526, "step": 12356 }, { "epoch": 7.90595009596929, "grad_norm": 0.9458651377596597, "learning_rate": 2.213067516013796e-06, "loss": 0.057, "step": 12357 }, { "epoch": 7.9065898912348045, "grad_norm": 0.8928491895065171, "learning_rate": 2.2117676037772705e-06, "loss": 0.0512, "step": 12358 }, { "epoch": 7.90722968650032, "grad_norm": 0.8361771548542447, "learning_rate": 2.2104680259535916e-06, "loss": 0.0454, "step": 12359 }, { "epoch": 7.907869481765835, "grad_norm": 0.819820042375757, "learning_rate": 2.2091687825985575e-06, "loss": 0.0514, "step": 12360 }, { "epoch": 7.90850927703135, "grad_norm": 0.7840451442404587, "learning_rate": 2.207869873767958e-06, "loss": 0.0435, "step": 12361 }, { "epoch": 7.909149072296865, "grad_norm": 0.8925442858649025, "learning_rate": 2.2065712995175616e-06, "loss": 0.055, "step": 12362 }, { "epoch": 7.90978886756238, "grad_norm": 0.7937869432495636, "learning_rate": 2.2052730599031337e-06, "loss": 0.0496, "step": 12363 }, { "epoch": 7.910428662827895, "grad_norm": 0.8095297332029644, "learning_rate": 2.2039751549804147e-06, "loss": 0.0486, "step": 12364 }, { "epoch": 7.9110684580934105, "grad_norm": 0.7808909794973321, "learning_rate": 2.2026775848051308e-06, "loss": 0.0467, "step": 12365 }, { "epoch": 7.911708253358925, "grad_norm": 0.8497315171210246, "learning_rate": 2.2013803494330045e-06, "loss": 0.0549, "step": 12366 }, { "epoch": 7.91234804862444, "grad_norm": 0.7870291632328332, "learning_rate": 2.200083448919733e-06, "loss": 0.0483, "step": 12367 }, { "epoch": 7.9129878438899555, "grad_norm": 0.8316739494454427, "learning_rate": 2.1987868833210013e-06, "loss": 0.0532, "step": 12368 }, { "epoch": 7.91362763915547, "grad_norm": 0.81249530307216, "learning_rate": 2.1974906526924855e-06, "loss": 0.051, "step": 12369 }, { "epoch": 7.914267434420985, "grad_norm": 0.8823210362618019, "learning_rate": 2.196194757089841e-06, "loss": 0.0491, "step": 12370 }, { "epoch": 7.9149072296865, "grad_norm": 0.8427177976277058, "learning_rate": 2.1948991965687127e-06, "loss": 0.0491, "step": 12371 }, { "epoch": 7.915547024952016, "grad_norm": 0.8469429938470977, "learning_rate": 2.1936039711847258e-06, "loss": 0.0472, "step": 12372 }, { "epoch": 7.91618682021753, "grad_norm": 0.8980421367246545, "learning_rate": 2.192309080993501e-06, "loss": 0.0578, "step": 12373 }, { "epoch": 7.916826615483045, "grad_norm": 0.8972693470591335, "learning_rate": 2.191014526050634e-06, "loss": 0.0504, "step": 12374 }, { "epoch": 7.917466410748561, "grad_norm": 0.802994241284513, "learning_rate": 2.1897203064117132e-06, "loss": 0.0462, "step": 12375 }, { "epoch": 7.918106206014075, "grad_norm": 0.7728031884288499, "learning_rate": 2.188426422132309e-06, "loss": 0.0455, "step": 12376 }, { "epoch": 7.91874600127959, "grad_norm": 0.8880092419876682, "learning_rate": 2.1871328732679787e-06, "loss": 0.0506, "step": 12377 }, { "epoch": 7.9193857965451055, "grad_norm": 0.7523567083227919, "learning_rate": 2.185839659874263e-06, "loss": 0.0459, "step": 12378 }, { "epoch": 7.920025591810621, "grad_norm": 0.8610570838925968, "learning_rate": 2.1845467820066925e-06, "loss": 0.0521, "step": 12379 }, { "epoch": 7.920665387076136, "grad_norm": 0.8702151886378513, "learning_rate": 2.1832542397207823e-06, "loss": 0.0513, "step": 12380 }, { "epoch": 7.9213051823416505, "grad_norm": 0.770954444623053, "learning_rate": 2.1819620330720294e-06, "loss": 0.0487, "step": 12381 }, { "epoch": 7.921944977607166, "grad_norm": 0.7933114845701061, "learning_rate": 2.1806701621159166e-06, "loss": 0.0468, "step": 12382 }, { "epoch": 7.922584772872681, "grad_norm": 0.8660322699365192, "learning_rate": 2.179378626907921e-06, "loss": 0.0522, "step": 12383 }, { "epoch": 7.923224568138195, "grad_norm": 1.0116607248387972, "learning_rate": 2.1780874275034948e-06, "loss": 0.0504, "step": 12384 }, { "epoch": 7.923864363403711, "grad_norm": 0.8454308028234775, "learning_rate": 2.176796563958078e-06, "loss": 0.0489, "step": 12385 }, { "epoch": 7.924504158669226, "grad_norm": 0.8410046862089198, "learning_rate": 2.175506036327103e-06, "loss": 0.0549, "step": 12386 }, { "epoch": 7.925143953934741, "grad_norm": 0.8657146419096118, "learning_rate": 2.17421584466598e-06, "loss": 0.0533, "step": 12387 }, { "epoch": 7.925783749200256, "grad_norm": 0.9364618505224562, "learning_rate": 2.172925989030107e-06, "loss": 0.0523, "step": 12388 }, { "epoch": 7.926423544465771, "grad_norm": 0.7463487139417514, "learning_rate": 2.1716364694748693e-06, "loss": 0.044, "step": 12389 }, { "epoch": 7.927063339731286, "grad_norm": 0.8666945320880239, "learning_rate": 2.170347286055636e-06, "loss": 0.0543, "step": 12390 }, { "epoch": 7.927703134996801, "grad_norm": 0.9090060686933283, "learning_rate": 2.1690584388277624e-06, "loss": 0.0561, "step": 12391 }, { "epoch": 7.928342930262316, "grad_norm": 0.8729784611304993, "learning_rate": 2.167769927846587e-06, "loss": 0.0465, "step": 12392 }, { "epoch": 7.928982725527831, "grad_norm": 0.8717901359545369, "learning_rate": 2.1664817531674407e-06, "loss": 0.0529, "step": 12393 }, { "epoch": 7.929622520793346, "grad_norm": 0.8706842319132051, "learning_rate": 2.1651939148456334e-06, "loss": 0.05, "step": 12394 }, { "epoch": 7.930262316058861, "grad_norm": 0.8230324641039666, "learning_rate": 2.1639064129364594e-06, "loss": 0.0591, "step": 12395 }, { "epoch": 7.930902111324376, "grad_norm": 0.8236123493829524, "learning_rate": 2.162619247495208e-06, "loss": 0.0531, "step": 12396 }, { "epoch": 7.931541906589891, "grad_norm": 0.9335606664236454, "learning_rate": 2.1613324185771433e-06, "loss": 0.0511, "step": 12397 }, { "epoch": 7.932181701855407, "grad_norm": 0.9059354035313892, "learning_rate": 2.1600459262375182e-06, "loss": 0.0535, "step": 12398 }, { "epoch": 7.932821497120921, "grad_norm": 0.8244902043293059, "learning_rate": 2.1587597705315788e-06, "loss": 0.0459, "step": 12399 }, { "epoch": 7.933461292386436, "grad_norm": 0.859896425510639, "learning_rate": 2.1574739515145447e-06, "loss": 0.0504, "step": 12400 }, { "epoch": 7.9341010876519515, "grad_norm": 0.8902269230573945, "learning_rate": 2.156188469241629e-06, "loss": 0.0514, "step": 12401 }, { "epoch": 7.934740882917467, "grad_norm": 0.8423154650528871, "learning_rate": 2.154903323768025e-06, "loss": 0.0528, "step": 12402 }, { "epoch": 7.935380678182981, "grad_norm": 0.9003463827200262, "learning_rate": 2.1536185151489198e-06, "loss": 0.0483, "step": 12403 }, { "epoch": 7.9360204734484965, "grad_norm": 0.8407139790887894, "learning_rate": 2.152334043439478e-06, "loss": 0.0509, "step": 12404 }, { "epoch": 7.936660268714012, "grad_norm": 0.7876703015999054, "learning_rate": 2.151049908694852e-06, "loss": 0.043, "step": 12405 }, { "epoch": 7.937300063979526, "grad_norm": 0.904910963990053, "learning_rate": 2.1497661109701808e-06, "loss": 0.0529, "step": 12406 }, { "epoch": 7.937939859245041, "grad_norm": 0.9515008816442571, "learning_rate": 2.148482650320588e-06, "loss": 0.0546, "step": 12407 }, { "epoch": 7.938579654510557, "grad_norm": 0.7357336577685664, "learning_rate": 2.1471995268011815e-06, "loss": 0.0476, "step": 12408 }, { "epoch": 7.939219449776072, "grad_norm": 0.8673501146971484, "learning_rate": 2.14591674046706e-06, "loss": 0.0533, "step": 12409 }, { "epoch": 7.939859245041586, "grad_norm": 0.8797794198080303, "learning_rate": 2.1446342913733033e-06, "loss": 0.0507, "step": 12410 }, { "epoch": 7.940499040307102, "grad_norm": 0.7886208828986556, "learning_rate": 2.1433521795749757e-06, "loss": 0.0525, "step": 12411 }, { "epoch": 7.941138835572617, "grad_norm": 0.7748212568336955, "learning_rate": 2.142070405127127e-06, "loss": 0.0463, "step": 12412 }, { "epoch": 7.941778630838132, "grad_norm": 0.8134041705556708, "learning_rate": 2.1407889680848004e-06, "loss": 0.0454, "step": 12413 }, { "epoch": 7.942418426103647, "grad_norm": 0.8462205760495558, "learning_rate": 2.1395078685030145e-06, "loss": 0.0572, "step": 12414 }, { "epoch": 7.943058221369162, "grad_norm": 0.7927527453724237, "learning_rate": 2.1382271064367755e-06, "loss": 0.0495, "step": 12415 }, { "epoch": 7.943698016634677, "grad_norm": 0.8678226071804576, "learning_rate": 2.136946681941081e-06, "loss": 0.0541, "step": 12416 }, { "epoch": 7.944337811900192, "grad_norm": 0.8727454413084873, "learning_rate": 2.13566659507091e-06, "loss": 0.0535, "step": 12417 }, { "epoch": 7.944977607165707, "grad_norm": 0.970166169186746, "learning_rate": 2.1343868458812254e-06, "loss": 0.0524, "step": 12418 }, { "epoch": 7.945617402431222, "grad_norm": 0.8537706403235907, "learning_rate": 2.1331074344269788e-06, "loss": 0.0511, "step": 12419 }, { "epoch": 7.946257197696737, "grad_norm": 0.8504210992651621, "learning_rate": 2.131828360763104e-06, "loss": 0.0511, "step": 12420 }, { "epoch": 7.946896992962252, "grad_norm": 0.8127951762575087, "learning_rate": 2.1305496249445233e-06, "loss": 0.0485, "step": 12421 }, { "epoch": 7.947536788227767, "grad_norm": 0.8132474147613888, "learning_rate": 2.1292712270261394e-06, "loss": 0.0514, "step": 12422 }, { "epoch": 7.948176583493282, "grad_norm": 0.8642462604764746, "learning_rate": 2.1279931670628516e-06, "loss": 0.0497, "step": 12423 }, { "epoch": 7.9488163787587975, "grad_norm": 0.8600360572626455, "learning_rate": 2.126715445109534e-06, "loss": 0.0507, "step": 12424 }, { "epoch": 7.949456174024312, "grad_norm": 0.8617356131643877, "learning_rate": 2.1254380612210467e-06, "loss": 0.0522, "step": 12425 }, { "epoch": 7.950095969289827, "grad_norm": 0.8495028952858866, "learning_rate": 2.124161015452244e-06, "loss": 0.0527, "step": 12426 }, { "epoch": 7.9507357645553425, "grad_norm": 0.8392595499663585, "learning_rate": 2.1228843078579575e-06, "loss": 0.0502, "step": 12427 }, { "epoch": 7.951375559820857, "grad_norm": 0.9530229055892008, "learning_rate": 2.1216079384930066e-06, "loss": 0.0458, "step": 12428 }, { "epoch": 7.952015355086372, "grad_norm": 0.9229183409317774, "learning_rate": 2.120331907412193e-06, "loss": 0.0557, "step": 12429 }, { "epoch": 7.952655150351887, "grad_norm": 0.8356283614935809, "learning_rate": 2.1190562146703143e-06, "loss": 0.0542, "step": 12430 }, { "epoch": 7.953294945617403, "grad_norm": 0.8224039134242891, "learning_rate": 2.1177808603221418e-06, "loss": 0.0514, "step": 12431 }, { "epoch": 7.953934740882918, "grad_norm": 0.8522679179287628, "learning_rate": 2.1165058444224352e-06, "loss": 0.0576, "step": 12432 }, { "epoch": 7.954574536148432, "grad_norm": 0.830763471941395, "learning_rate": 2.115231167025946e-06, "loss": 0.0495, "step": 12433 }, { "epoch": 7.955214331413948, "grad_norm": 0.7936527175788588, "learning_rate": 2.1139568281874057e-06, "loss": 0.0517, "step": 12434 }, { "epoch": 7.955854126679463, "grad_norm": 0.8508180571732239, "learning_rate": 2.112682827961531e-06, "loss": 0.0495, "step": 12435 }, { "epoch": 7.956493921944977, "grad_norm": 0.9138696740243939, "learning_rate": 2.1114091664030244e-06, "loss": 0.0532, "step": 12436 }, { "epoch": 7.957133717210493, "grad_norm": 0.8164020955503349, "learning_rate": 2.1101358435665753e-06, "loss": 0.0443, "step": 12437 }, { "epoch": 7.957773512476008, "grad_norm": 0.8951403828319265, "learning_rate": 2.1088628595068593e-06, "loss": 0.0543, "step": 12438 }, { "epoch": 7.958413307741523, "grad_norm": 0.8782413729030525, "learning_rate": 2.107590214278531e-06, "loss": 0.0516, "step": 12439 }, { "epoch": 7.9590531030070375, "grad_norm": 0.8998269515942389, "learning_rate": 2.1063179079362428e-06, "loss": 0.0529, "step": 12440 }, { "epoch": 7.959692898272553, "grad_norm": 0.8566479598933998, "learning_rate": 2.1050459405346213e-06, "loss": 0.0494, "step": 12441 }, { "epoch": 7.960332693538068, "grad_norm": 0.8029257834306132, "learning_rate": 2.1037743121282815e-06, "loss": 0.0541, "step": 12442 }, { "epoch": 7.960972488803582, "grad_norm": 0.8220120704774957, "learning_rate": 2.102503022771828e-06, "loss": 0.0547, "step": 12443 }, { "epoch": 7.961612284069098, "grad_norm": 0.7981736586270574, "learning_rate": 2.1012320725198463e-06, "loss": 0.05, "step": 12444 }, { "epoch": 7.962252079334613, "grad_norm": 0.9075339232767142, "learning_rate": 2.099961461426906e-06, "loss": 0.0498, "step": 12445 }, { "epoch": 7.962891874600128, "grad_norm": 0.764273792124336, "learning_rate": 2.0986911895475704e-06, "loss": 0.0505, "step": 12446 }, { "epoch": 7.963531669865643, "grad_norm": 0.9206701010575692, "learning_rate": 2.0974212569363793e-06, "loss": 0.0535, "step": 12447 }, { "epoch": 7.964171465131158, "grad_norm": 0.8544452031102696, "learning_rate": 2.096151663647863e-06, "loss": 0.0505, "step": 12448 }, { "epoch": 7.964811260396673, "grad_norm": 0.755381008416033, "learning_rate": 2.0948824097365338e-06, "loss": 0.0452, "step": 12449 }, { "epoch": 7.9654510556621885, "grad_norm": 0.6957957328322699, "learning_rate": 2.093613495256891e-06, "loss": 0.0453, "step": 12450 }, { "epoch": 7.966090850927703, "grad_norm": 0.8180807231570086, "learning_rate": 2.0923449202634215e-06, "loss": 0.046, "step": 12451 }, { "epoch": 7.966730646193218, "grad_norm": 0.8257356373107466, "learning_rate": 2.0910766848105914e-06, "loss": 0.0506, "step": 12452 }, { "epoch": 7.967370441458733, "grad_norm": 0.9033540550691744, "learning_rate": 2.0898087889528617e-06, "loss": 0.0591, "step": 12453 }, { "epoch": 7.968010236724249, "grad_norm": 0.9489747922003128, "learning_rate": 2.088541232744672e-06, "loss": 0.0482, "step": 12454 }, { "epoch": 7.968650031989763, "grad_norm": 0.7617197747110169, "learning_rate": 2.087274016240445e-06, "loss": 0.0427, "step": 12455 }, { "epoch": 7.969289827255278, "grad_norm": 0.8143985311267944, "learning_rate": 2.0860071394945992e-06, "loss": 0.0528, "step": 12456 }, { "epoch": 7.969929622520794, "grad_norm": 0.9009445071533213, "learning_rate": 2.0847406025615293e-06, "loss": 0.0532, "step": 12457 }, { "epoch": 7.970569417786308, "grad_norm": 0.8444298867938984, "learning_rate": 2.083474405495618e-06, "loss": 0.0478, "step": 12458 }, { "epoch": 7.971209213051823, "grad_norm": 0.7953861177790841, "learning_rate": 2.08220854835123e-06, "loss": 0.0503, "step": 12459 }, { "epoch": 7.971849008317339, "grad_norm": 0.8558906364961161, "learning_rate": 2.080943031182725e-06, "loss": 0.0544, "step": 12460 }, { "epoch": 7.972488803582854, "grad_norm": 0.723803896721993, "learning_rate": 2.0796778540444397e-06, "loss": 0.0432, "step": 12461 }, { "epoch": 7.973128598848368, "grad_norm": 0.8054663126527173, "learning_rate": 2.0784130169906945e-06, "loss": 0.0452, "step": 12462 }, { "epoch": 7.9737683941138835, "grad_norm": 0.7749620499621254, "learning_rate": 2.0771485200758067e-06, "loss": 0.0493, "step": 12463 }, { "epoch": 7.974408189379399, "grad_norm": 0.7759195913838833, "learning_rate": 2.0758843633540683e-06, "loss": 0.0503, "step": 12464 }, { "epoch": 7.975047984644913, "grad_norm": 0.8340818080527993, "learning_rate": 2.0746205468797574e-06, "loss": 0.0461, "step": 12465 }, { "epoch": 7.975687779910428, "grad_norm": 0.8648087777880592, "learning_rate": 2.0733570707071436e-06, "loss": 0.049, "step": 12466 }, { "epoch": 7.976327575175944, "grad_norm": 0.7960288214485679, "learning_rate": 2.0720939348904767e-06, "loss": 0.0527, "step": 12467 }, { "epoch": 7.976967370441459, "grad_norm": 0.8619348112346674, "learning_rate": 2.070831139483993e-06, "loss": 0.0547, "step": 12468 }, { "epoch": 7.977607165706974, "grad_norm": 0.8247472119863009, "learning_rate": 2.069568684541914e-06, "loss": 0.0513, "step": 12469 }, { "epoch": 7.978246960972489, "grad_norm": 0.913953076553354, "learning_rate": 2.068306570118449e-06, "loss": 0.0592, "step": 12470 }, { "epoch": 7.978886756238004, "grad_norm": 0.9374411950896859, "learning_rate": 2.067044796267792e-06, "loss": 0.0533, "step": 12471 }, { "epoch": 7.979526551503519, "grad_norm": 0.766769824724907, "learning_rate": 2.065783363044117e-06, "loss": 0.0442, "step": 12472 }, { "epoch": 7.980166346769034, "grad_norm": 0.8373953964044778, "learning_rate": 2.0645222705015934e-06, "loss": 0.0521, "step": 12473 }, { "epoch": 7.980806142034549, "grad_norm": 0.8452163355835276, "learning_rate": 2.063261518694366e-06, "loss": 0.0487, "step": 12474 }, { "epoch": 7.981445937300064, "grad_norm": 0.8039352545642168, "learning_rate": 2.062001107676569e-06, "loss": 0.0492, "step": 12475 }, { "epoch": 7.982085732565579, "grad_norm": 0.9227806501753333, "learning_rate": 2.0607410375023275e-06, "loss": 0.0497, "step": 12476 }, { "epoch": 7.982725527831094, "grad_norm": 0.8588806919052409, "learning_rate": 2.0594813082257414e-06, "loss": 0.0503, "step": 12477 }, { "epoch": 7.983365323096609, "grad_norm": 0.8872527029173863, "learning_rate": 2.0582219199009034e-06, "loss": 0.0475, "step": 12478 }, { "epoch": 7.984005118362124, "grad_norm": 0.9229010388429221, "learning_rate": 2.056962872581889e-06, "loss": 0.0584, "step": 12479 }, { "epoch": 7.984644913627639, "grad_norm": 0.9010622455176409, "learning_rate": 2.0557041663227582e-06, "loss": 0.0512, "step": 12480 }, { "epoch": 7.985284708893154, "grad_norm": 0.7910827699828807, "learning_rate": 2.0544458011775603e-06, "loss": 0.052, "step": 12481 }, { "epoch": 7.985924504158669, "grad_norm": 0.8612756234586187, "learning_rate": 2.0531877772003217e-06, "loss": 0.0446, "step": 12482 }, { "epoch": 7.9865642994241846, "grad_norm": 0.8612669849931623, "learning_rate": 2.051930094445067e-06, "loss": 0.0502, "step": 12483 }, { "epoch": 7.987204094689699, "grad_norm": 0.80389594503288, "learning_rate": 2.0506727529657954e-06, "loss": 0.0498, "step": 12484 }, { "epoch": 7.987843889955214, "grad_norm": 0.8315977168642844, "learning_rate": 2.0494157528164925e-06, "loss": 0.0438, "step": 12485 }, { "epoch": 7.9884836852207295, "grad_norm": 0.8816193429194387, "learning_rate": 2.0481590940511364e-06, "loss": 0.0488, "step": 12486 }, { "epoch": 7.989123480486245, "grad_norm": 0.8351199134361251, "learning_rate": 2.0469027767236848e-06, "loss": 0.051, "step": 12487 }, { "epoch": 7.989763275751759, "grad_norm": 0.8883236813695888, "learning_rate": 2.045646800888079e-06, "loss": 0.0533, "step": 12488 }, { "epoch": 7.990403071017274, "grad_norm": 1.0350861638257356, "learning_rate": 2.0443911665982473e-06, "loss": 0.0543, "step": 12489 }, { "epoch": 7.99104286628279, "grad_norm": 0.9239849905177728, "learning_rate": 2.0431358739081108e-06, "loss": 0.0516, "step": 12490 }, { "epoch": 7.991682661548305, "grad_norm": 0.8557374258427266, "learning_rate": 2.041880922871565e-06, "loss": 0.0532, "step": 12491 }, { "epoch": 7.992322456813819, "grad_norm": 0.8959917419374706, "learning_rate": 2.040626313542493e-06, "loss": 0.0551, "step": 12492 }, { "epoch": 7.992962252079335, "grad_norm": 0.913650521700941, "learning_rate": 2.0393720459747723e-06, "loss": 0.0576, "step": 12493 }, { "epoch": 7.99360204734485, "grad_norm": 0.7882252615060414, "learning_rate": 2.0381181202222543e-06, "loss": 0.0458, "step": 12494 }, { "epoch": 7.994241842610364, "grad_norm": 0.8130567290429402, "learning_rate": 2.0368645363387806e-06, "loss": 0.0516, "step": 12495 }, { "epoch": 7.99488163787588, "grad_norm": 0.8950854296530181, "learning_rate": 2.035611294378179e-06, "loss": 0.0495, "step": 12496 }, { "epoch": 7.995521433141395, "grad_norm": 0.8940792927796922, "learning_rate": 2.0343583943942612e-06, "loss": 0.0547, "step": 12497 }, { "epoch": 7.99616122840691, "grad_norm": 0.9324258576876239, "learning_rate": 2.033105836440823e-06, "loss": 0.0498, "step": 12498 }, { "epoch": 7.9968010236724245, "grad_norm": 0.8924777305653654, "learning_rate": 2.0318536205716467e-06, "loss": 0.0527, "step": 12499 }, { "epoch": 7.99744081893794, "grad_norm": 0.7796052784466355, "learning_rate": 2.030601746840504e-06, "loss": 0.0458, "step": 12500 }, { "epoch": 7.998080614203455, "grad_norm": 0.7969399197909448, "learning_rate": 2.0293502153011456e-06, "loss": 0.0497, "step": 12501 }, { "epoch": 7.9987204094689695, "grad_norm": 0.8520774125639451, "learning_rate": 2.0280990260073075e-06, "loss": 0.0491, "step": 12502 }, { "epoch": 7.999360204734485, "grad_norm": 0.7906635376121236, "learning_rate": 2.0268481790127192e-06, "loss": 0.0479, "step": 12503 }, { "epoch": 8.0, "grad_norm": 0.7437188644438393, "learning_rate": 2.025597674371087e-06, "loss": 0.0474, "step": 12504 }, { "epoch": 8.000639795265515, "grad_norm": 0.45260231360197356, "learning_rate": 2.0243475121361057e-06, "loss": 0.0239, "step": 12505 }, { "epoch": 8.00127959053103, "grad_norm": 0.5502591472015036, "learning_rate": 2.023097692361451e-06, "loss": 0.0273, "step": 12506 }, { "epoch": 8.001919385796546, "grad_norm": 0.7873233043884211, "learning_rate": 2.0218482151007956e-06, "loss": 0.0317, "step": 12507 }, { "epoch": 8.00255918106206, "grad_norm": 0.5803233032583976, "learning_rate": 2.0205990804077848e-06, "loss": 0.0287, "step": 12508 }, { "epoch": 8.003198976327575, "grad_norm": 0.5123587154029355, "learning_rate": 2.019350288336055e-06, "loss": 0.0258, "step": 12509 }, { "epoch": 8.00383877159309, "grad_norm": 0.6060320731095573, "learning_rate": 2.018101838939228e-06, "loss": 0.0328, "step": 12510 }, { "epoch": 8.004478566858605, "grad_norm": 0.5395474431407392, "learning_rate": 2.016853732270909e-06, "loss": 0.0304, "step": 12511 }, { "epoch": 8.00511836212412, "grad_norm": 0.6134076479740509, "learning_rate": 2.0156059683846875e-06, "loss": 0.0316, "step": 12512 }, { "epoch": 8.005758157389636, "grad_norm": 0.5245044366869195, "learning_rate": 2.0143585473341454e-06, "loss": 0.0265, "step": 12513 }, { "epoch": 8.006397952655151, "grad_norm": 0.593155149304833, "learning_rate": 2.0131114691728416e-06, "loss": 0.0283, "step": 12514 }, { "epoch": 8.007037747920666, "grad_norm": 0.6731860237735423, "learning_rate": 2.011864733954324e-06, "loss": 0.0324, "step": 12515 }, { "epoch": 8.00767754318618, "grad_norm": 0.6272276626665924, "learning_rate": 2.0106183417321213e-06, "loss": 0.0294, "step": 12516 }, { "epoch": 8.008317338451695, "grad_norm": 0.7069162027586225, "learning_rate": 2.009372292559758e-06, "loss": 0.0251, "step": 12517 }, { "epoch": 8.00895713371721, "grad_norm": 0.7045490428812362, "learning_rate": 2.0081265864907353e-06, "loss": 0.0268, "step": 12518 }, { "epoch": 8.009596928982726, "grad_norm": 0.6596786770092672, "learning_rate": 2.0068812235785375e-06, "loss": 0.0258, "step": 12519 }, { "epoch": 8.01023672424824, "grad_norm": 0.841615717695772, "learning_rate": 2.005636203876643e-06, "loss": 0.033, "step": 12520 }, { "epoch": 8.010876519513756, "grad_norm": 0.7806167207797337, "learning_rate": 2.0043915274385095e-06, "loss": 0.0277, "step": 12521 }, { "epoch": 8.011516314779271, "grad_norm": 0.7700220695929516, "learning_rate": 2.0031471943175783e-06, "loss": 0.0294, "step": 12522 }, { "epoch": 8.012156110044785, "grad_norm": 0.918860596868818, "learning_rate": 2.001903204567284e-06, "loss": 0.0324, "step": 12523 }, { "epoch": 8.0127959053103, "grad_norm": 0.7567961632252552, "learning_rate": 2.0006595582410383e-06, "loss": 0.0268, "step": 12524 }, { "epoch": 8.013435700575815, "grad_norm": 0.6827851608499365, "learning_rate": 1.9994162553922415e-06, "loss": 0.0278, "step": 12525 }, { "epoch": 8.01407549584133, "grad_norm": 0.6448602562783736, "learning_rate": 1.9981732960742796e-06, "loss": 0.023, "step": 12526 }, { "epoch": 8.014715291106846, "grad_norm": 0.7035699583476476, "learning_rate": 1.996930680340521e-06, "loss": 0.0275, "step": 12527 }, { "epoch": 8.015355086372361, "grad_norm": 0.7207832188034919, "learning_rate": 1.995688408244323e-06, "loss": 0.0279, "step": 12528 }, { "epoch": 8.015994881637877, "grad_norm": 0.6752060395896043, "learning_rate": 1.994446479839024e-06, "loss": 0.0227, "step": 12529 }, { "epoch": 8.01663467690339, "grad_norm": 0.7088700516459848, "learning_rate": 1.9932048951779547e-06, "loss": 0.0273, "step": 12530 }, { "epoch": 8.017274472168905, "grad_norm": 0.7009186610818121, "learning_rate": 1.991963654314424e-06, "loss": 0.0256, "step": 12531 }, { "epoch": 8.01791426743442, "grad_norm": 0.569261246310615, "learning_rate": 1.9907227573017264e-06, "loss": 0.0247, "step": 12532 }, { "epoch": 8.018554062699936, "grad_norm": 0.754692512594901, "learning_rate": 1.9894822041931485e-06, "loss": 0.0309, "step": 12533 }, { "epoch": 8.019193857965451, "grad_norm": 0.6702147170207756, "learning_rate": 1.988241995041955e-06, "loss": 0.0261, "step": 12534 }, { "epoch": 8.019833653230966, "grad_norm": 0.7439094039718087, "learning_rate": 1.9870021299013995e-06, "loss": 0.026, "step": 12535 }, { "epoch": 8.020473448496482, "grad_norm": 0.6782403041813161, "learning_rate": 1.985762608824715e-06, "loss": 0.0284, "step": 12536 }, { "epoch": 8.021113243761997, "grad_norm": 0.628168567261968, "learning_rate": 1.984523431865131e-06, "loss": 0.0263, "step": 12537 }, { "epoch": 8.02175303902751, "grad_norm": 0.6259416613300959, "learning_rate": 1.983284599075852e-06, "loss": 0.0274, "step": 12538 }, { "epoch": 8.022392834293026, "grad_norm": 0.7004142511920464, "learning_rate": 1.982046110510072e-06, "loss": 0.0259, "step": 12539 }, { "epoch": 8.023032629558541, "grad_norm": 0.6270723691089817, "learning_rate": 1.980807966220969e-06, "loss": 0.0272, "step": 12540 }, { "epoch": 8.023672424824056, "grad_norm": 0.620890893119882, "learning_rate": 1.9795701662617074e-06, "loss": 0.0236, "step": 12541 }, { "epoch": 8.024312220089572, "grad_norm": 0.7124419489894818, "learning_rate": 1.9783327106854322e-06, "loss": 0.0246, "step": 12542 }, { "epoch": 8.024952015355087, "grad_norm": 0.6922407694882728, "learning_rate": 1.9770955995452833e-06, "loss": 0.0278, "step": 12543 }, { "epoch": 8.025591810620602, "grad_norm": 0.6780593744276044, "learning_rate": 1.9758588328943784e-06, "loss": 0.0294, "step": 12544 }, { "epoch": 8.026231605886116, "grad_norm": 0.7099065542698343, "learning_rate": 1.9746224107858216e-06, "loss": 0.0285, "step": 12545 }, { "epoch": 8.026871401151631, "grad_norm": 0.7390590843362894, "learning_rate": 1.9733863332726988e-06, "loss": 0.0264, "step": 12546 }, { "epoch": 8.027511196417146, "grad_norm": 0.6696687976800577, "learning_rate": 1.9721506004080927e-06, "loss": 0.0246, "step": 12547 }, { "epoch": 8.028150991682661, "grad_norm": 0.7489118729015604, "learning_rate": 1.9709152122450582e-06, "loss": 0.0273, "step": 12548 }, { "epoch": 8.028790786948177, "grad_norm": 0.6711353129967309, "learning_rate": 1.9696801688366394e-06, "loss": 0.0252, "step": 12549 }, { "epoch": 8.029430582213692, "grad_norm": 0.7704149054394646, "learning_rate": 1.9684454702358725e-06, "loss": 0.0278, "step": 12550 }, { "epoch": 8.030070377479207, "grad_norm": 0.7406391449899108, "learning_rate": 1.9672111164957697e-06, "loss": 0.0267, "step": 12551 }, { "epoch": 8.030710172744723, "grad_norm": 0.8553127114359705, "learning_rate": 1.96597710766933e-06, "loss": 0.029, "step": 12552 }, { "epoch": 8.031349968010236, "grad_norm": 0.6141297635884835, "learning_rate": 1.9647434438095434e-06, "loss": 0.0233, "step": 12553 }, { "epoch": 8.031989763275751, "grad_norm": 0.7058684105471935, "learning_rate": 1.963510124969381e-06, "loss": 0.0266, "step": 12554 }, { "epoch": 8.032629558541267, "grad_norm": 0.7314185410130086, "learning_rate": 1.962277151201798e-06, "loss": 0.028, "step": 12555 }, { "epoch": 8.033269353806782, "grad_norm": 0.6865952055042517, "learning_rate": 1.9610445225597353e-06, "loss": 0.0229, "step": 12556 }, { "epoch": 8.033909149072297, "grad_norm": 0.808536429340547, "learning_rate": 1.9598122390961216e-06, "loss": 0.028, "step": 12557 }, { "epoch": 8.034548944337812, "grad_norm": 0.7730399925171453, "learning_rate": 1.958580300863867e-06, "loss": 0.0273, "step": 12558 }, { "epoch": 8.035188739603328, "grad_norm": 0.7344408605909916, "learning_rate": 1.957348707915868e-06, "loss": 0.029, "step": 12559 }, { "epoch": 8.035828534868841, "grad_norm": 0.7121381320007325, "learning_rate": 1.9561174603050102e-06, "loss": 0.0238, "step": 12560 }, { "epoch": 8.036468330134356, "grad_norm": 0.7040271945099715, "learning_rate": 1.95488655808416e-06, "loss": 0.0233, "step": 12561 }, { "epoch": 8.037108125399872, "grad_norm": 0.8188663425130867, "learning_rate": 1.9536560013061677e-06, "loss": 0.0296, "step": 12562 }, { "epoch": 8.037747920665387, "grad_norm": 0.7856789497933657, "learning_rate": 1.9524257900238765e-06, "loss": 0.0261, "step": 12563 }, { "epoch": 8.038387715930902, "grad_norm": 0.691853998232024, "learning_rate": 1.951195924290106e-06, "loss": 0.0212, "step": 12564 }, { "epoch": 8.039027511196418, "grad_norm": 0.6818123819729739, "learning_rate": 1.949966404157664e-06, "loss": 0.0242, "step": 12565 }, { "epoch": 8.039667306461933, "grad_norm": 0.7640897929491296, "learning_rate": 1.948737229679343e-06, "loss": 0.0303, "step": 12566 }, { "epoch": 8.040307101727446, "grad_norm": 0.6949337166142737, "learning_rate": 1.947508400907925e-06, "loss": 0.0244, "step": 12567 }, { "epoch": 8.040946896992962, "grad_norm": 0.7638870802633703, "learning_rate": 1.946279917896172e-06, "loss": 0.0291, "step": 12568 }, { "epoch": 8.041586692258477, "grad_norm": 0.7699791111876684, "learning_rate": 1.945051780696834e-06, "loss": 0.0262, "step": 12569 }, { "epoch": 8.042226487523992, "grad_norm": 0.7646775951559667, "learning_rate": 1.943823989362643e-06, "loss": 0.0282, "step": 12570 }, { "epoch": 8.042866282789507, "grad_norm": 0.7035508745915613, "learning_rate": 1.942596543946319e-06, "loss": 0.0242, "step": 12571 }, { "epoch": 8.043506078055023, "grad_norm": 0.7586304768863963, "learning_rate": 1.941369444500567e-06, "loss": 0.026, "step": 12572 }, { "epoch": 8.044145873320538, "grad_norm": 0.6900902399718376, "learning_rate": 1.940142691078073e-06, "loss": 0.0238, "step": 12573 }, { "epoch": 8.044785668586053, "grad_norm": 0.7314479981612187, "learning_rate": 1.9389162837315168e-06, "loss": 0.0246, "step": 12574 }, { "epoch": 8.045425463851567, "grad_norm": 0.6325944257695165, "learning_rate": 1.9376902225135565e-06, "loss": 0.0239, "step": 12575 }, { "epoch": 8.046065259117082, "grad_norm": 0.742010028786259, "learning_rate": 1.936464507476833e-06, "loss": 0.0253, "step": 12576 }, { "epoch": 8.046705054382597, "grad_norm": 0.7429515897705101, "learning_rate": 1.935239138673982e-06, "loss": 0.0222, "step": 12577 }, { "epoch": 8.047344849648113, "grad_norm": 0.6605792272569712, "learning_rate": 1.9340141161576176e-06, "loss": 0.0237, "step": 12578 }, { "epoch": 8.047984644913628, "grad_norm": 0.7233675726931261, "learning_rate": 1.932789439980335e-06, "loss": 0.0291, "step": 12579 }, { "epoch": 8.048624440179143, "grad_norm": 0.7833174482273225, "learning_rate": 1.9315651101947262e-06, "loss": 0.0279, "step": 12580 }, { "epoch": 8.049264235444658, "grad_norm": 0.7209248987861835, "learning_rate": 1.930341126853359e-06, "loss": 0.0234, "step": 12581 }, { "epoch": 8.049904030710172, "grad_norm": 0.661079967380268, "learning_rate": 1.9291174900087892e-06, "loss": 0.0284, "step": 12582 }, { "epoch": 8.050543825975687, "grad_norm": 0.7183006889309815, "learning_rate": 1.9278941997135546e-06, "loss": 0.0264, "step": 12583 }, { "epoch": 8.051183621241202, "grad_norm": 0.6893177002856598, "learning_rate": 1.9266712560201874e-06, "loss": 0.0215, "step": 12584 }, { "epoch": 8.051823416506718, "grad_norm": 0.7115132386165979, "learning_rate": 1.925448658981194e-06, "loss": 0.0256, "step": 12585 }, { "epoch": 8.052463211772233, "grad_norm": 0.6227622585835626, "learning_rate": 1.9242264086490714e-06, "loss": 0.0212, "step": 12586 }, { "epoch": 8.053103007037748, "grad_norm": 0.8646104337887369, "learning_rate": 1.9230045050763026e-06, "loss": 0.0329, "step": 12587 }, { "epoch": 8.053742802303264, "grad_norm": 0.7375877943361905, "learning_rate": 1.9217829483153516e-06, "loss": 0.0282, "step": 12588 }, { "epoch": 8.054382597568779, "grad_norm": 0.7458222170528753, "learning_rate": 1.9205617384186683e-06, "loss": 0.027, "step": 12589 }, { "epoch": 8.055022392834292, "grad_norm": 0.68116360088738, "learning_rate": 1.9193408754386933e-06, "loss": 0.0233, "step": 12590 }, { "epoch": 8.055662188099808, "grad_norm": 0.712033388743366, "learning_rate": 1.9181203594278475e-06, "loss": 0.0254, "step": 12591 }, { "epoch": 8.056301983365323, "grad_norm": 0.6688658118768462, "learning_rate": 1.9169001904385376e-06, "loss": 0.0255, "step": 12592 }, { "epoch": 8.056941778630838, "grad_norm": 0.6874250667061849, "learning_rate": 1.9156803685231518e-06, "loss": 0.0239, "step": 12593 }, { "epoch": 8.057581573896353, "grad_norm": 0.7172251283869132, "learning_rate": 1.9144608937340726e-06, "loss": 0.026, "step": 12594 }, { "epoch": 8.058221369161869, "grad_norm": 0.616200690239714, "learning_rate": 1.913241766123659e-06, "loss": 0.0216, "step": 12595 }, { "epoch": 8.058861164427384, "grad_norm": 0.7137204069372236, "learning_rate": 1.9120229857442573e-06, "loss": 0.0258, "step": 12596 }, { "epoch": 8.059500959692897, "grad_norm": 0.7445202371595429, "learning_rate": 1.910804552648202e-06, "loss": 0.028, "step": 12597 }, { "epoch": 8.060140754958413, "grad_norm": 0.7497025769106366, "learning_rate": 1.909586466887812e-06, "loss": 0.0276, "step": 12598 }, { "epoch": 8.060780550223928, "grad_norm": 0.7727587988507295, "learning_rate": 1.908368728515385e-06, "loss": 0.0264, "step": 12599 }, { "epoch": 8.061420345489443, "grad_norm": 0.7155348422155106, "learning_rate": 1.907151337583213e-06, "loss": 0.0244, "step": 12600 }, { "epoch": 8.062060140754959, "grad_norm": 0.7643172400384656, "learning_rate": 1.9059342941435654e-06, "loss": 0.0219, "step": 12601 }, { "epoch": 8.062699936020474, "grad_norm": 0.6766463027147608, "learning_rate": 1.904717598248701e-06, "loss": 0.021, "step": 12602 }, { "epoch": 8.06333973128599, "grad_norm": 0.7099972467284966, "learning_rate": 1.9035012499508598e-06, "loss": 0.0262, "step": 12603 }, { "epoch": 8.063979526551503, "grad_norm": 0.6825714144754638, "learning_rate": 1.902285249302276e-06, "loss": 0.0238, "step": 12604 }, { "epoch": 8.064619321817018, "grad_norm": 0.7413845389811573, "learning_rate": 1.9010695963551572e-06, "loss": 0.0303, "step": 12605 }, { "epoch": 8.065259117082533, "grad_norm": 0.6891620193017308, "learning_rate": 1.8998542911617023e-06, "loss": 0.0242, "step": 12606 }, { "epoch": 8.065898912348048, "grad_norm": 0.6694403320202752, "learning_rate": 1.898639333774096e-06, "loss": 0.0247, "step": 12607 }, { "epoch": 8.066538707613564, "grad_norm": 0.7311510872282844, "learning_rate": 1.8974247242445066e-06, "loss": 0.0254, "step": 12608 }, { "epoch": 8.067178502879079, "grad_norm": 0.804238038567824, "learning_rate": 1.8962104626250832e-06, "loss": 0.0287, "step": 12609 }, { "epoch": 8.067818298144594, "grad_norm": 0.8383293572969137, "learning_rate": 1.89499654896797e-06, "loss": 0.0307, "step": 12610 }, { "epoch": 8.06845809341011, "grad_norm": 0.861916016149757, "learning_rate": 1.8937829833252875e-06, "loss": 0.0281, "step": 12611 }, { "epoch": 8.069097888675623, "grad_norm": 0.7835650151319774, "learning_rate": 1.8925697657491426e-06, "loss": 0.029, "step": 12612 }, { "epoch": 8.069737683941138, "grad_norm": 0.7950063580186838, "learning_rate": 1.8913568962916285e-06, "loss": 0.0286, "step": 12613 }, { "epoch": 8.070377479206654, "grad_norm": 0.8080728148213234, "learning_rate": 1.8901443750048277e-06, "loss": 0.0313, "step": 12614 }, { "epoch": 8.071017274472169, "grad_norm": 0.743118677217791, "learning_rate": 1.8889322019408008e-06, "loss": 0.0251, "step": 12615 }, { "epoch": 8.071657069737684, "grad_norm": 0.7871125033550529, "learning_rate": 1.887720377151596e-06, "loss": 0.028, "step": 12616 }, { "epoch": 8.0722968650032, "grad_norm": 0.6387904972406088, "learning_rate": 1.8865089006892478e-06, "loss": 0.0197, "step": 12617 }, { "epoch": 8.072936660268715, "grad_norm": 0.7767801967749456, "learning_rate": 1.8852977726057752e-06, "loss": 0.0264, "step": 12618 }, { "epoch": 8.073576455534228, "grad_norm": 0.7422608851529945, "learning_rate": 1.884086992953179e-06, "loss": 0.0263, "step": 12619 }, { "epoch": 8.074216250799743, "grad_norm": 0.7484813612876857, "learning_rate": 1.8828765617834521e-06, "loss": 0.0244, "step": 12620 }, { "epoch": 8.074856046065259, "grad_norm": 0.7982016080137772, "learning_rate": 1.8816664791485672e-06, "loss": 0.0259, "step": 12621 }, { "epoch": 8.075495841330774, "grad_norm": 0.7807295701462085, "learning_rate": 1.8804567451004818e-06, "loss": 0.0289, "step": 12622 }, { "epoch": 8.07613563659629, "grad_norm": 0.7478582178335462, "learning_rate": 1.8792473596911375e-06, "loss": 0.0256, "step": 12623 }, { "epoch": 8.076775431861805, "grad_norm": 0.6479372655099751, "learning_rate": 1.8780383229724685e-06, "loss": 0.0193, "step": 12624 }, { "epoch": 8.07741522712732, "grad_norm": 0.6481140477232765, "learning_rate": 1.8768296349963855e-06, "loss": 0.0229, "step": 12625 }, { "epoch": 8.078055022392835, "grad_norm": 0.760791557611676, "learning_rate": 1.8756212958147868e-06, "loss": 0.0268, "step": 12626 }, { "epoch": 8.078694817658349, "grad_norm": 0.8760488231297846, "learning_rate": 1.8744133054795588e-06, "loss": 0.0266, "step": 12627 }, { "epoch": 8.079334612923864, "grad_norm": 0.8558857767297965, "learning_rate": 1.8732056640425699e-06, "loss": 0.0283, "step": 12628 }, { "epoch": 8.07997440818938, "grad_norm": 0.7795549041502253, "learning_rate": 1.8719983715556733e-06, "loss": 0.0242, "step": 12629 }, { "epoch": 8.080614203454894, "grad_norm": 0.7458325749354411, "learning_rate": 1.8707914280707073e-06, "loss": 0.026, "step": 12630 }, { "epoch": 8.08125399872041, "grad_norm": 0.7488475449536361, "learning_rate": 1.8695848336394974e-06, "loss": 0.025, "step": 12631 }, { "epoch": 8.081893793985925, "grad_norm": 0.709455431060157, "learning_rate": 1.8683785883138516e-06, "loss": 0.026, "step": 12632 }, { "epoch": 8.08253358925144, "grad_norm": 0.7700002185508472, "learning_rate": 1.8671726921455613e-06, "loss": 0.0246, "step": 12633 }, { "epoch": 8.083173384516954, "grad_norm": 0.7389748959133262, "learning_rate": 1.8659671451864114e-06, "loss": 0.0224, "step": 12634 }, { "epoch": 8.083813179782469, "grad_norm": 0.87585208909161, "learning_rate": 1.864761947488163e-06, "loss": 0.027, "step": 12635 }, { "epoch": 8.084452975047984, "grad_norm": 0.6711721451136619, "learning_rate": 1.8635570991025631e-06, "loss": 0.0213, "step": 12636 }, { "epoch": 8.0850927703135, "grad_norm": 0.7052286800227863, "learning_rate": 1.8623526000813496e-06, "loss": 0.0207, "step": 12637 }, { "epoch": 8.085732565579015, "grad_norm": 0.7529830916378623, "learning_rate": 1.861148450476239e-06, "loss": 0.029, "step": 12638 }, { "epoch": 8.08637236084453, "grad_norm": 0.7069049485640112, "learning_rate": 1.859944650338935e-06, "loss": 0.0236, "step": 12639 }, { "epoch": 8.087012156110045, "grad_norm": 0.8164899124304892, "learning_rate": 1.858741199721129e-06, "loss": 0.0268, "step": 12640 }, { "epoch": 8.08765195137556, "grad_norm": 0.8146969251982212, "learning_rate": 1.857538098674494e-06, "loss": 0.0222, "step": 12641 }, { "epoch": 8.088291746641074, "grad_norm": 0.8050604318622139, "learning_rate": 1.8563353472506895e-06, "loss": 0.0271, "step": 12642 }, { "epoch": 8.08893154190659, "grad_norm": 0.7904283968695334, "learning_rate": 1.8551329455013556e-06, "loss": 0.0288, "step": 12643 }, { "epoch": 8.089571337172105, "grad_norm": 0.719768280568245, "learning_rate": 1.8539308934781265e-06, "loss": 0.0247, "step": 12644 }, { "epoch": 8.09021113243762, "grad_norm": 0.728351437207976, "learning_rate": 1.8527291912326151e-06, "loss": 0.0217, "step": 12645 }, { "epoch": 8.090850927703135, "grad_norm": 0.7086774018172011, "learning_rate": 1.8515278388164182e-06, "loss": 0.021, "step": 12646 }, { "epoch": 8.09149072296865, "grad_norm": 0.7816035345957608, "learning_rate": 1.8503268362811222e-06, "loss": 0.0267, "step": 12647 }, { "epoch": 8.092130518234166, "grad_norm": 0.7903256401027714, "learning_rate": 1.8491261836782936e-06, "loss": 0.0268, "step": 12648 }, { "epoch": 8.09277031349968, "grad_norm": 0.8579417679401645, "learning_rate": 1.847925881059488e-06, "loss": 0.0254, "step": 12649 }, { "epoch": 8.093410108765195, "grad_norm": 0.7340539972788771, "learning_rate": 1.8467259284762407e-06, "loss": 0.0231, "step": 12650 }, { "epoch": 8.09404990403071, "grad_norm": 0.8415157307868619, "learning_rate": 1.8455263259800815e-06, "loss": 0.0245, "step": 12651 }, { "epoch": 8.094689699296225, "grad_norm": 0.6938159747010081, "learning_rate": 1.844327073622516e-06, "loss": 0.0226, "step": 12652 }, { "epoch": 8.09532949456174, "grad_norm": 0.7851617705353657, "learning_rate": 1.8431281714550354e-06, "loss": 0.0258, "step": 12653 }, { "epoch": 8.095969289827256, "grad_norm": 0.7359650906163164, "learning_rate": 1.8419296195291236e-06, "loss": 0.0267, "step": 12654 }, { "epoch": 8.096609085092771, "grad_norm": 0.7875235301606953, "learning_rate": 1.8407314178962421e-06, "loss": 0.0251, "step": 12655 }, { "epoch": 8.097248880358284, "grad_norm": 0.685294201103147, "learning_rate": 1.8395335666078373e-06, "loss": 0.0263, "step": 12656 }, { "epoch": 8.0978886756238, "grad_norm": 0.7927781947803991, "learning_rate": 1.838336065715347e-06, "loss": 0.0277, "step": 12657 }, { "epoch": 8.098528470889315, "grad_norm": 0.6859978695009312, "learning_rate": 1.8371389152701868e-06, "loss": 0.0226, "step": 12658 }, { "epoch": 8.09916826615483, "grad_norm": 0.8031010519813976, "learning_rate": 1.8359421153237623e-06, "loss": 0.028, "step": 12659 }, { "epoch": 8.099808061420346, "grad_norm": 0.7572631682468929, "learning_rate": 1.8347456659274587e-06, "loss": 0.0305, "step": 12660 }, { "epoch": 8.10044785668586, "grad_norm": 0.808003072674828, "learning_rate": 1.833549567132653e-06, "loss": 0.029, "step": 12661 }, { "epoch": 8.101087651951376, "grad_norm": 0.781662199066011, "learning_rate": 1.832353818990702e-06, "loss": 0.0244, "step": 12662 }, { "epoch": 8.101727447216891, "grad_norm": 0.7851828189264283, "learning_rate": 1.8311584215529454e-06, "loss": 0.0264, "step": 12663 }, { "epoch": 8.102367242482405, "grad_norm": 0.8660262842004823, "learning_rate": 1.8299633748707179e-06, "loss": 0.0301, "step": 12664 }, { "epoch": 8.10300703774792, "grad_norm": 0.7312658215046617, "learning_rate": 1.8287686789953296e-06, "loss": 0.0244, "step": 12665 }, { "epoch": 8.103646833013435, "grad_norm": 0.6988891852848479, "learning_rate": 1.827574333978076e-06, "loss": 0.0237, "step": 12666 }, { "epoch": 8.10428662827895, "grad_norm": 0.6969379775265596, "learning_rate": 1.8263803398702462e-06, "loss": 0.0244, "step": 12667 }, { "epoch": 8.104926423544466, "grad_norm": 0.8131822188496461, "learning_rate": 1.8251866967231035e-06, "loss": 0.0278, "step": 12668 }, { "epoch": 8.105566218809981, "grad_norm": 0.7091974509908855, "learning_rate": 1.8239934045879038e-06, "loss": 0.024, "step": 12669 }, { "epoch": 8.106206014075497, "grad_norm": 0.7100277474156044, "learning_rate": 1.8228004635158802e-06, "loss": 0.0236, "step": 12670 }, { "epoch": 8.10684580934101, "grad_norm": 0.7899708307341479, "learning_rate": 1.821607873558261e-06, "loss": 0.0272, "step": 12671 }, { "epoch": 8.107485604606525, "grad_norm": 0.8440961283114691, "learning_rate": 1.8204156347662505e-06, "loss": 0.0261, "step": 12672 }, { "epoch": 8.10812539987204, "grad_norm": 0.683093961351882, "learning_rate": 1.8192237471910412e-06, "loss": 0.0218, "step": 12673 }, { "epoch": 8.108765195137556, "grad_norm": 0.726975918089581, "learning_rate": 1.8180322108838134e-06, "loss": 0.0232, "step": 12674 }, { "epoch": 8.109404990403071, "grad_norm": 0.7556039553646091, "learning_rate": 1.816841025895728e-06, "loss": 0.0212, "step": 12675 }, { "epoch": 8.110044785668586, "grad_norm": 0.914664786039177, "learning_rate": 1.8156501922779335e-06, "loss": 0.0267, "step": 12676 }, { "epoch": 8.110684580934102, "grad_norm": 0.7991015098262334, "learning_rate": 1.8144597100815597e-06, "loss": 0.0263, "step": 12677 }, { "epoch": 8.111324376199617, "grad_norm": 0.7698397471218547, "learning_rate": 1.8132695793577248e-06, "loss": 0.0299, "step": 12678 }, { "epoch": 8.11196417146513, "grad_norm": 0.9451031858669936, "learning_rate": 1.8120798001575312e-06, "loss": 0.0304, "step": 12679 }, { "epoch": 8.112603966730646, "grad_norm": 0.8092320637899547, "learning_rate": 1.8108903725320637e-06, "loss": 0.0264, "step": 12680 }, { "epoch": 8.113243761996161, "grad_norm": 0.8135817359849158, "learning_rate": 1.809701296532399e-06, "loss": 0.0299, "step": 12681 }, { "epoch": 8.113883557261676, "grad_norm": 0.693340126801632, "learning_rate": 1.8085125722095909e-06, "loss": 0.0243, "step": 12682 }, { "epoch": 8.114523352527192, "grad_norm": 0.8245761197754888, "learning_rate": 1.8073241996146785e-06, "loss": 0.0274, "step": 12683 }, { "epoch": 8.115163147792707, "grad_norm": 0.8215090658579552, "learning_rate": 1.8061361787986942e-06, "loss": 0.0285, "step": 12684 }, { "epoch": 8.115802943058222, "grad_norm": 0.7837193991130595, "learning_rate": 1.8049485098126474e-06, "loss": 0.0241, "step": 12685 }, { "epoch": 8.116442738323736, "grad_norm": 0.7546993218802657, "learning_rate": 1.8037611927075294e-06, "loss": 0.0282, "step": 12686 }, { "epoch": 8.117082533589251, "grad_norm": 0.7379347126369217, "learning_rate": 1.8025742275343295e-06, "loss": 0.0234, "step": 12687 }, { "epoch": 8.117722328854766, "grad_norm": 0.7172679194009366, "learning_rate": 1.8013876143440101e-06, "loss": 0.0252, "step": 12688 }, { "epoch": 8.118362124120281, "grad_norm": 0.7543190449259333, "learning_rate": 1.8002013531875228e-06, "loss": 0.0285, "step": 12689 }, { "epoch": 8.119001919385797, "grad_norm": 0.7588768168906138, "learning_rate": 1.7990154441158025e-06, "loss": 0.0267, "step": 12690 }, { "epoch": 8.119641714651312, "grad_norm": 0.7352775633483064, "learning_rate": 1.7978298871797705e-06, "loss": 0.0292, "step": 12691 }, { "epoch": 8.120281509916827, "grad_norm": 0.7618097480839342, "learning_rate": 1.796644682430333e-06, "loss": 0.0266, "step": 12692 }, { "epoch": 8.12092130518234, "grad_norm": 0.8026463837176052, "learning_rate": 1.7954598299183778e-06, "loss": 0.0256, "step": 12693 }, { "epoch": 8.121561100447856, "grad_norm": 0.8227656577250798, "learning_rate": 1.7942753296947846e-06, "loss": 0.0248, "step": 12694 }, { "epoch": 8.122200895713371, "grad_norm": 0.7723172974451404, "learning_rate": 1.7930911818104125e-06, "loss": 0.0256, "step": 12695 }, { "epoch": 8.122840690978887, "grad_norm": 0.6893926637950857, "learning_rate": 1.791907386316103e-06, "loss": 0.0243, "step": 12696 }, { "epoch": 8.123480486244402, "grad_norm": 0.7387403465241958, "learning_rate": 1.7907239432626922e-06, "loss": 0.0268, "step": 12697 }, { "epoch": 8.124120281509917, "grad_norm": 0.7720163468626202, "learning_rate": 1.7895408527009917e-06, "loss": 0.0304, "step": 12698 }, { "epoch": 8.124760076775432, "grad_norm": 0.6213168497569449, "learning_rate": 1.788358114681803e-06, "loss": 0.0224, "step": 12699 }, { "epoch": 8.125399872040948, "grad_norm": 0.7910026096490195, "learning_rate": 1.7871757292559056e-06, "loss": 0.0304, "step": 12700 }, { "epoch": 8.126039667306461, "grad_norm": 0.7620895483333716, "learning_rate": 1.7859936964740754e-06, "loss": 0.0262, "step": 12701 }, { "epoch": 8.126679462571976, "grad_norm": 0.7669639374685882, "learning_rate": 1.7848120163870653e-06, "loss": 0.0262, "step": 12702 }, { "epoch": 8.127319257837492, "grad_norm": 0.7665987029914885, "learning_rate": 1.7836306890456102e-06, "loss": 0.0271, "step": 12703 }, { "epoch": 8.127959053103007, "grad_norm": 0.7332956336939652, "learning_rate": 1.7824497145004404e-06, "loss": 0.0253, "step": 12704 }, { "epoch": 8.128598848368522, "grad_norm": 0.7059076753779426, "learning_rate": 1.7812690928022614e-06, "loss": 0.0242, "step": 12705 }, { "epoch": 8.129238643634038, "grad_norm": 0.6873389348390712, "learning_rate": 1.7800888240017688e-06, "loss": 0.0267, "step": 12706 }, { "epoch": 8.129878438899553, "grad_norm": 0.7591032690749505, "learning_rate": 1.7789089081496391e-06, "loss": 0.024, "step": 12707 }, { "epoch": 8.130518234165066, "grad_norm": 0.7879044255010781, "learning_rate": 1.777729345296536e-06, "loss": 0.0262, "step": 12708 }, { "epoch": 8.131158029430582, "grad_norm": 0.7403669304641902, "learning_rate": 1.7765501354931103e-06, "loss": 0.0221, "step": 12709 }, { "epoch": 8.131797824696097, "grad_norm": 0.9692059551121762, "learning_rate": 1.775371278789989e-06, "loss": 0.0225, "step": 12710 }, { "epoch": 8.132437619961612, "grad_norm": 0.8234342971091972, "learning_rate": 1.7741927752377986e-06, "loss": 0.0273, "step": 12711 }, { "epoch": 8.133077415227127, "grad_norm": 0.7673770549004191, "learning_rate": 1.7730146248871371e-06, "loss": 0.0227, "step": 12712 }, { "epoch": 8.133717210492643, "grad_norm": 0.6898809441753422, "learning_rate": 1.77183682778859e-06, "loss": 0.021, "step": 12713 }, { "epoch": 8.134357005758158, "grad_norm": 0.839317066533908, "learning_rate": 1.7706593839927366e-06, "loss": 0.0285, "step": 12714 }, { "epoch": 8.134996801023673, "grad_norm": 0.8234123891789112, "learning_rate": 1.7694822935501288e-06, "loss": 0.0275, "step": 12715 }, { "epoch": 8.135636596289187, "grad_norm": 0.8063557484962477, "learning_rate": 1.7683055565113094e-06, "loss": 0.0262, "step": 12716 }, { "epoch": 8.136276391554702, "grad_norm": 0.7982054913549509, "learning_rate": 1.7671291729268081e-06, "loss": 0.0262, "step": 12717 }, { "epoch": 8.136916186820217, "grad_norm": 0.6962953938864417, "learning_rate": 1.7659531428471343e-06, "loss": 0.0212, "step": 12718 }, { "epoch": 8.137555982085733, "grad_norm": 0.899914176450972, "learning_rate": 1.7647774663227867e-06, "loss": 0.0312, "step": 12719 }, { "epoch": 8.138195777351248, "grad_norm": 0.8063185521927293, "learning_rate": 1.763602143404244e-06, "loss": 0.0269, "step": 12720 }, { "epoch": 8.138835572616763, "grad_norm": 0.8359050993543002, "learning_rate": 1.7624271741419752e-06, "loss": 0.0276, "step": 12721 }, { "epoch": 8.139475367882278, "grad_norm": 0.6597958899482853, "learning_rate": 1.7612525585864293e-06, "loss": 0.0217, "step": 12722 }, { "epoch": 8.140115163147792, "grad_norm": 0.6898862647243004, "learning_rate": 1.7600782967880404e-06, "loss": 0.022, "step": 12723 }, { "epoch": 8.140754958413307, "grad_norm": 0.7568231024887389, "learning_rate": 1.7589043887972347e-06, "loss": 0.0243, "step": 12724 }, { "epoch": 8.141394753678822, "grad_norm": 0.7290368586882978, "learning_rate": 1.7577308346644139e-06, "loss": 0.0244, "step": 12725 }, { "epoch": 8.142034548944338, "grad_norm": 0.7268867299733265, "learning_rate": 1.7565576344399693e-06, "loss": 0.0247, "step": 12726 }, { "epoch": 8.142674344209853, "grad_norm": 0.7501898130943112, "learning_rate": 1.7553847881742737e-06, "loss": 0.0258, "step": 12727 }, { "epoch": 8.143314139475368, "grad_norm": 0.7818139177858242, "learning_rate": 1.7542122959176921e-06, "loss": 0.025, "step": 12728 }, { "epoch": 8.143953934740884, "grad_norm": 0.8535745650503305, "learning_rate": 1.753040157720567e-06, "loss": 0.028, "step": 12729 }, { "epoch": 8.144593730006397, "grad_norm": 0.7766063802956775, "learning_rate": 1.7518683736332233e-06, "loss": 0.0257, "step": 12730 }, { "epoch": 8.145233525271912, "grad_norm": 0.8727674070892996, "learning_rate": 1.750696943705983e-06, "loss": 0.027, "step": 12731 }, { "epoch": 8.145873320537428, "grad_norm": 0.7991738390047625, "learning_rate": 1.7495258679891414e-06, "loss": 0.0259, "step": 12732 }, { "epoch": 8.146513115802943, "grad_norm": 0.7553770136538042, "learning_rate": 1.7483551465329796e-06, "loss": 0.0253, "step": 12733 }, { "epoch": 8.147152911068458, "grad_norm": 0.7505484608354424, "learning_rate": 1.7471847793877716e-06, "loss": 0.0266, "step": 12734 }, { "epoch": 8.147792706333973, "grad_norm": 0.8638936205815175, "learning_rate": 1.7460147666037686e-06, "loss": 0.0271, "step": 12735 }, { "epoch": 8.148432501599489, "grad_norm": 0.8601313950015655, "learning_rate": 1.744845108231209e-06, "loss": 0.0293, "step": 12736 }, { "epoch": 8.149072296865004, "grad_norm": 0.8483262748748406, "learning_rate": 1.7436758043203161e-06, "loss": 0.0264, "step": 12737 }, { "epoch": 8.149712092130518, "grad_norm": 0.7262055317644206, "learning_rate": 1.7425068549212975e-06, "loss": 0.0234, "step": 12738 }, { "epoch": 8.150351887396033, "grad_norm": 0.7777225978131938, "learning_rate": 1.7413382600843455e-06, "loss": 0.0289, "step": 12739 }, { "epoch": 8.150991682661548, "grad_norm": 0.7972752977796921, "learning_rate": 1.740170019859634e-06, "loss": 0.0281, "step": 12740 }, { "epoch": 8.151631477927063, "grad_norm": 0.7695585308618275, "learning_rate": 1.7390021342973318e-06, "loss": 0.0274, "step": 12741 }, { "epoch": 8.152271273192579, "grad_norm": 0.804290598198045, "learning_rate": 1.7378346034475834e-06, "loss": 0.0248, "step": 12742 }, { "epoch": 8.152911068458094, "grad_norm": 0.9883029931361974, "learning_rate": 1.7366674273605166e-06, "loss": 0.028, "step": 12743 }, { "epoch": 8.15355086372361, "grad_norm": 0.8341672904224436, "learning_rate": 1.7355006060862544e-06, "loss": 0.024, "step": 12744 }, { "epoch": 8.154190658989123, "grad_norm": 0.8210580955303589, "learning_rate": 1.7343341396748935e-06, "loss": 0.0271, "step": 12745 }, { "epoch": 8.154830454254638, "grad_norm": 0.7630412284683837, "learning_rate": 1.7331680281765228e-06, "loss": 0.0262, "step": 12746 }, { "epoch": 8.155470249520153, "grad_norm": 0.7375416543089466, "learning_rate": 1.7320022716412077e-06, "loss": 0.0231, "step": 12747 }, { "epoch": 8.156110044785668, "grad_norm": 0.8772192029784597, "learning_rate": 1.7308368701190104e-06, "loss": 0.0292, "step": 12748 }, { "epoch": 8.156749840051184, "grad_norm": 0.8221359087102366, "learning_rate": 1.7296718236599686e-06, "loss": 0.0258, "step": 12749 }, { "epoch": 8.157389635316699, "grad_norm": 0.7819497386437427, "learning_rate": 1.7285071323141068e-06, "loss": 0.0243, "step": 12750 }, { "epoch": 8.158029430582214, "grad_norm": 0.7746815491176514, "learning_rate": 1.7273427961314337e-06, "loss": 0.0201, "step": 12751 }, { "epoch": 8.15866922584773, "grad_norm": 0.9260920884002707, "learning_rate": 1.7261788151619464e-06, "loss": 0.0274, "step": 12752 }, { "epoch": 8.159309021113243, "grad_norm": 1.0009873128195397, "learning_rate": 1.72501518945562e-06, "loss": 0.029, "step": 12753 }, { "epoch": 8.159948816378758, "grad_norm": 0.8143386620802973, "learning_rate": 1.7238519190624226e-06, "loss": 0.0277, "step": 12754 }, { "epoch": 8.160588611644274, "grad_norm": 0.8566533739465105, "learning_rate": 1.722689004032303e-06, "loss": 0.0302, "step": 12755 }, { "epoch": 8.161228406909789, "grad_norm": 0.750267551363349, "learning_rate": 1.7215264444151935e-06, "loss": 0.0288, "step": 12756 }, { "epoch": 8.161868202175304, "grad_norm": 0.7907019157574212, "learning_rate": 1.7203642402610087e-06, "loss": 0.0272, "step": 12757 }, { "epoch": 8.16250799744082, "grad_norm": 0.974876135863408, "learning_rate": 1.7192023916196587e-06, "loss": 0.0328, "step": 12758 }, { "epoch": 8.163147792706335, "grad_norm": 0.8096474971564628, "learning_rate": 1.7180408985410269e-06, "loss": 0.023, "step": 12759 }, { "epoch": 8.163787587971848, "grad_norm": 0.7058205799066313, "learning_rate": 1.7168797610749844e-06, "loss": 0.0219, "step": 12760 }, { "epoch": 8.164427383237364, "grad_norm": 0.9088071340502585, "learning_rate": 1.7157189792713924e-06, "loss": 0.0312, "step": 12761 }, { "epoch": 8.165067178502879, "grad_norm": 0.7195316356219764, "learning_rate": 1.7145585531800912e-06, "loss": 0.0203, "step": 12762 }, { "epoch": 8.165706973768394, "grad_norm": 0.750967013199358, "learning_rate": 1.7133984828509043e-06, "loss": 0.0248, "step": 12763 }, { "epoch": 8.16634676903391, "grad_norm": 0.6883284441440928, "learning_rate": 1.7122387683336483e-06, "loss": 0.0236, "step": 12764 }, { "epoch": 8.166986564299425, "grad_norm": 0.7893464366103783, "learning_rate": 1.711079409678117e-06, "loss": 0.0297, "step": 12765 }, { "epoch": 8.16762635956494, "grad_norm": 0.737368603885995, "learning_rate": 1.7099204069340902e-06, "loss": 0.0225, "step": 12766 }, { "epoch": 8.168266154830455, "grad_norm": 0.7093819934564409, "learning_rate": 1.7087617601513352e-06, "loss": 0.024, "step": 12767 }, { "epoch": 8.168905950095969, "grad_norm": 0.7151978174198412, "learning_rate": 1.7076034693796018e-06, "loss": 0.0274, "step": 12768 }, { "epoch": 8.169545745361484, "grad_norm": 0.6470642217305357, "learning_rate": 1.7064455346686227e-06, "loss": 0.0209, "step": 12769 }, { "epoch": 8.170185540627, "grad_norm": 0.7247345846541986, "learning_rate": 1.705287956068118e-06, "loss": 0.0254, "step": 12770 }, { "epoch": 8.170825335892514, "grad_norm": 0.6824032568206078, "learning_rate": 1.7041307336277957e-06, "loss": 0.0221, "step": 12771 }, { "epoch": 8.17146513115803, "grad_norm": 0.6911059822486996, "learning_rate": 1.7029738673973417e-06, "loss": 0.0245, "step": 12772 }, { "epoch": 8.172104926423545, "grad_norm": 0.7118107773698439, "learning_rate": 1.7018173574264295e-06, "loss": 0.0234, "step": 12773 }, { "epoch": 8.17274472168906, "grad_norm": 0.7843651159468352, "learning_rate": 1.7006612037647196e-06, "loss": 0.0262, "step": 12774 }, { "epoch": 8.173384516954574, "grad_norm": 0.7517471991695182, "learning_rate": 1.6995054064618554e-06, "loss": 0.0257, "step": 12775 }, { "epoch": 8.174024312220089, "grad_norm": 0.8197347330420841, "learning_rate": 1.698349965567463e-06, "loss": 0.0269, "step": 12776 }, { "epoch": 8.174664107485604, "grad_norm": 0.8208502863000625, "learning_rate": 1.6971948811311533e-06, "loss": 0.0258, "step": 12777 }, { "epoch": 8.17530390275112, "grad_norm": 0.7615856925700273, "learning_rate": 1.6960401532025283e-06, "loss": 0.0259, "step": 12778 }, { "epoch": 8.175943698016635, "grad_norm": 0.761520992761032, "learning_rate": 1.694885781831167e-06, "loss": 0.0277, "step": 12779 }, { "epoch": 8.17658349328215, "grad_norm": 0.8001853005065422, "learning_rate": 1.6937317670666375e-06, "loss": 0.0253, "step": 12780 }, { "epoch": 8.177223288547665, "grad_norm": 0.788117828816608, "learning_rate": 1.6925781089584903e-06, "loss": 0.0274, "step": 12781 }, { "epoch": 8.177863083813179, "grad_norm": 0.7444917926062146, "learning_rate": 1.69142480755626e-06, "loss": 0.0245, "step": 12782 }, { "epoch": 8.178502879078694, "grad_norm": 0.8531051811583857, "learning_rate": 1.6902718629094673e-06, "loss": 0.0309, "step": 12783 }, { "epoch": 8.17914267434421, "grad_norm": 0.7315240578351593, "learning_rate": 1.6891192750676212e-06, "loss": 0.026, "step": 12784 }, { "epoch": 8.179782469609725, "grad_norm": 0.7028950284180275, "learning_rate": 1.6879670440802099e-06, "loss": 0.0214, "step": 12785 }, { "epoch": 8.18042226487524, "grad_norm": 0.7925280060873084, "learning_rate": 1.6868151699967072e-06, "loss": 0.0236, "step": 12786 }, { "epoch": 8.181062060140755, "grad_norm": 0.7136725047190609, "learning_rate": 1.6856636528665705e-06, "loss": 0.0247, "step": 12787 }, { "epoch": 8.18170185540627, "grad_norm": 0.8923463038711721, "learning_rate": 1.6845124927392498e-06, "loss": 0.0282, "step": 12788 }, { "epoch": 8.182341650671786, "grad_norm": 0.7649074462128833, "learning_rate": 1.6833616896641692e-06, "loss": 0.0246, "step": 12789 }, { "epoch": 8.1829814459373, "grad_norm": 0.7538833785394486, "learning_rate": 1.6822112436907422e-06, "loss": 0.0244, "step": 12790 }, { "epoch": 8.183621241202815, "grad_norm": 0.6792880293329059, "learning_rate": 1.6810611548683697e-06, "loss": 0.02, "step": 12791 }, { "epoch": 8.18426103646833, "grad_norm": 0.7264745440458583, "learning_rate": 1.6799114232464341e-06, "loss": 0.0246, "step": 12792 }, { "epoch": 8.184900831733845, "grad_norm": 0.7237556196661417, "learning_rate": 1.6787620488742984e-06, "loss": 0.022, "step": 12793 }, { "epoch": 8.18554062699936, "grad_norm": 0.8864395414028033, "learning_rate": 1.6776130318013207e-06, "loss": 0.0285, "step": 12794 }, { "epoch": 8.186180422264876, "grad_norm": 0.871850717072204, "learning_rate": 1.6764643720768347e-06, "loss": 0.0285, "step": 12795 }, { "epoch": 8.186820217530391, "grad_norm": 0.9256035203697957, "learning_rate": 1.6753160697501625e-06, "loss": 0.0305, "step": 12796 }, { "epoch": 8.187460012795905, "grad_norm": 0.7203266822588248, "learning_rate": 1.6741681248706099e-06, "loss": 0.0219, "step": 12797 }, { "epoch": 8.18809980806142, "grad_norm": 0.8267784465621936, "learning_rate": 1.6730205374874674e-06, "loss": 0.0279, "step": 12798 }, { "epoch": 8.188739603326935, "grad_norm": 0.6106940018494238, "learning_rate": 1.6718733076500104e-06, "loss": 0.0169, "step": 12799 }, { "epoch": 8.18937939859245, "grad_norm": 0.9308827578918624, "learning_rate": 1.670726435407497e-06, "loss": 0.0268, "step": 12800 }, { "epoch": 8.190019193857966, "grad_norm": 0.7221180974054122, "learning_rate": 1.6695799208091757e-06, "loss": 0.0246, "step": 12801 }, { "epoch": 8.190658989123481, "grad_norm": 0.7753279005518362, "learning_rate": 1.6684337639042748e-06, "loss": 0.0261, "step": 12802 }, { "epoch": 8.191298784388996, "grad_norm": 0.7506597263451458, "learning_rate": 1.667287964742006e-06, "loss": 0.0257, "step": 12803 }, { "epoch": 8.191938579654511, "grad_norm": 0.7975926165958436, "learning_rate": 1.6661425233715677e-06, "loss": 0.026, "step": 12804 }, { "epoch": 8.192578374920025, "grad_norm": 0.7180944909143875, "learning_rate": 1.6649974398421475e-06, "loss": 0.0229, "step": 12805 }, { "epoch": 8.19321817018554, "grad_norm": 0.7427863035683363, "learning_rate": 1.6638527142029093e-06, "loss": 0.0233, "step": 12806 }, { "epoch": 8.193857965451055, "grad_norm": 0.7917901534392174, "learning_rate": 1.6627083465030048e-06, "loss": 0.0247, "step": 12807 }, { "epoch": 8.19449776071657, "grad_norm": 0.752710353611641, "learning_rate": 1.6615643367915757e-06, "loss": 0.0234, "step": 12808 }, { "epoch": 8.195137555982086, "grad_norm": 0.6951732596026832, "learning_rate": 1.6604206851177407e-06, "loss": 0.0247, "step": 12809 }, { "epoch": 8.195777351247601, "grad_norm": 0.7387882097348767, "learning_rate": 1.659277391530606e-06, "loss": 0.0245, "step": 12810 }, { "epoch": 8.196417146513117, "grad_norm": 0.8380492868190154, "learning_rate": 1.6581344560792644e-06, "loss": 0.0261, "step": 12811 }, { "epoch": 8.19705694177863, "grad_norm": 0.7135486903740687, "learning_rate": 1.6569918788127893e-06, "loss": 0.0256, "step": 12812 }, { "epoch": 8.197696737044145, "grad_norm": 0.755780468081266, "learning_rate": 1.655849659780242e-06, "loss": 0.0268, "step": 12813 }, { "epoch": 8.19833653230966, "grad_norm": 0.7657122037553103, "learning_rate": 1.6547077990306647e-06, "loss": 0.0247, "step": 12814 }, { "epoch": 8.198976327575176, "grad_norm": 0.7709528573738231, "learning_rate": 1.6535662966130927e-06, "loss": 0.0252, "step": 12815 }, { "epoch": 8.199616122840691, "grad_norm": 0.7640784727614272, "learning_rate": 1.6524251525765355e-06, "loss": 0.0266, "step": 12816 }, { "epoch": 8.200255918106206, "grad_norm": 0.7289511117274079, "learning_rate": 1.6512843669699919e-06, "loss": 0.0253, "step": 12817 }, { "epoch": 8.200895713371722, "grad_norm": 0.7987973735104579, "learning_rate": 1.6501439398424479e-06, "loss": 0.0231, "step": 12818 }, { "epoch": 8.201535508637235, "grad_norm": 0.7773066900296628, "learning_rate": 1.6490038712428701e-06, "loss": 0.0292, "step": 12819 }, { "epoch": 8.20217530390275, "grad_norm": 0.7837932154692436, "learning_rate": 1.6478641612202085e-06, "loss": 0.0244, "step": 12820 }, { "epoch": 8.202815099168266, "grad_norm": 0.7220126198377778, "learning_rate": 1.6467248098234045e-06, "loss": 0.0251, "step": 12821 }, { "epoch": 8.203454894433781, "grad_norm": 0.7282843345968192, "learning_rate": 1.6455858171013783e-06, "loss": 0.0244, "step": 12822 }, { "epoch": 8.204094689699296, "grad_norm": 0.7065276285096032, "learning_rate": 1.644447183103035e-06, "loss": 0.0241, "step": 12823 }, { "epoch": 8.204734484964812, "grad_norm": 0.7055931019623033, "learning_rate": 1.6433089078772646e-06, "loss": 0.0219, "step": 12824 }, { "epoch": 8.205374280230327, "grad_norm": 0.8205494050289448, "learning_rate": 1.6421709914729466e-06, "loss": 0.0268, "step": 12825 }, { "epoch": 8.206014075495842, "grad_norm": 0.8302303634619017, "learning_rate": 1.6410334339389389e-06, "loss": 0.0233, "step": 12826 }, { "epoch": 8.206653870761356, "grad_norm": 0.7151945399203495, "learning_rate": 1.6398962353240856e-06, "loss": 0.0227, "step": 12827 }, { "epoch": 8.207293666026871, "grad_norm": 0.7679951728836015, "learning_rate": 1.6387593956772175e-06, "loss": 0.0252, "step": 12828 }, { "epoch": 8.207933461292386, "grad_norm": 0.7525974036690495, "learning_rate": 1.6376229150471467e-06, "loss": 0.0231, "step": 12829 }, { "epoch": 8.208573256557901, "grad_norm": 0.6734060323411788, "learning_rate": 1.63648679348267e-06, "loss": 0.0219, "step": 12830 }, { "epoch": 8.209213051823417, "grad_norm": 0.8040881280975718, "learning_rate": 1.635351031032576e-06, "loss": 0.0273, "step": 12831 }, { "epoch": 8.209852847088932, "grad_norm": 0.6841519384315184, "learning_rate": 1.6342156277456279e-06, "loss": 0.0238, "step": 12832 }, { "epoch": 8.210492642354447, "grad_norm": 0.9926952383594996, "learning_rate": 1.63308058367058e-06, "loss": 0.0304, "step": 12833 }, { "epoch": 8.21113243761996, "grad_norm": 0.8454464156064182, "learning_rate": 1.6319458988561655e-06, "loss": 0.0268, "step": 12834 }, { "epoch": 8.211772232885476, "grad_norm": 0.6905894909015126, "learning_rate": 1.630811573351111e-06, "loss": 0.0228, "step": 12835 }, { "epoch": 8.212412028150991, "grad_norm": 0.8627796264154647, "learning_rate": 1.62967760720412e-06, "loss": 0.0265, "step": 12836 }, { "epoch": 8.213051823416507, "grad_norm": 0.7227997581751743, "learning_rate": 1.6285440004638809e-06, "loss": 0.0213, "step": 12837 }, { "epoch": 8.213691618682022, "grad_norm": 0.7382956483473185, "learning_rate": 1.6274107531790727e-06, "loss": 0.0248, "step": 12838 }, { "epoch": 8.214331413947537, "grad_norm": 0.8442031119624372, "learning_rate": 1.6262778653983525e-06, "loss": 0.0273, "step": 12839 }, { "epoch": 8.214971209213052, "grad_norm": 0.8023609115636028, "learning_rate": 1.6251453371703662e-06, "loss": 0.0234, "step": 12840 }, { "epoch": 8.215611004478568, "grad_norm": 0.6714326052215055, "learning_rate": 1.624013168543741e-06, "loss": 0.0212, "step": 12841 }, { "epoch": 8.216250799744081, "grad_norm": 0.7342936769029806, "learning_rate": 1.62288135956709e-06, "loss": 0.024, "step": 12842 }, { "epoch": 8.216890595009597, "grad_norm": 0.7307515761190577, "learning_rate": 1.6217499102890122e-06, "loss": 0.0234, "step": 12843 }, { "epoch": 8.217530390275112, "grad_norm": 0.7389884227710114, "learning_rate": 1.6206188207580876e-06, "loss": 0.024, "step": 12844 }, { "epoch": 8.218170185540627, "grad_norm": 0.8059848704775914, "learning_rate": 1.619488091022886e-06, "loss": 0.0249, "step": 12845 }, { "epoch": 8.218809980806142, "grad_norm": 0.7777448708010128, "learning_rate": 1.6183577211319602e-06, "loss": 0.025, "step": 12846 }, { "epoch": 8.219449776071658, "grad_norm": 0.8436061045897937, "learning_rate": 1.6172277111338397e-06, "loss": 0.0233, "step": 12847 }, { "epoch": 8.220089571337173, "grad_norm": 0.8129356806367066, "learning_rate": 1.6160980610770526e-06, "loss": 0.0247, "step": 12848 }, { "epoch": 8.220729366602686, "grad_norm": 0.6555314803333742, "learning_rate": 1.6149687710101014e-06, "loss": 0.0216, "step": 12849 }, { "epoch": 8.221369161868202, "grad_norm": 0.8105861040912685, "learning_rate": 1.6138398409814725e-06, "loss": 0.0271, "step": 12850 }, { "epoch": 8.222008957133717, "grad_norm": 0.7508285976427471, "learning_rate": 1.6127112710396453e-06, "loss": 0.0218, "step": 12851 }, { "epoch": 8.222648752399232, "grad_norm": 0.8027838242540452, "learning_rate": 1.6115830612330764e-06, "loss": 0.0261, "step": 12852 }, { "epoch": 8.223288547664747, "grad_norm": 0.7072545957126887, "learning_rate": 1.610455211610209e-06, "loss": 0.0214, "step": 12853 }, { "epoch": 8.223928342930263, "grad_norm": 0.8153327197263103, "learning_rate": 1.6093277222194681e-06, "loss": 0.0267, "step": 12854 }, { "epoch": 8.224568138195778, "grad_norm": 0.7550201956256944, "learning_rate": 1.6082005931092726e-06, "loss": 0.0239, "step": 12855 }, { "epoch": 8.225207933461292, "grad_norm": 0.7525412040575754, "learning_rate": 1.607073824328015e-06, "loss": 0.0248, "step": 12856 }, { "epoch": 8.225847728726807, "grad_norm": 0.7379499344962902, "learning_rate": 1.6059474159240773e-06, "loss": 0.0214, "step": 12857 }, { "epoch": 8.226487523992322, "grad_norm": 0.8072487764417481, "learning_rate": 1.6048213679458258e-06, "loss": 0.0269, "step": 12858 }, { "epoch": 8.227127319257837, "grad_norm": 0.9277112999058936, "learning_rate": 1.6036956804416114e-06, "loss": 0.0265, "step": 12859 }, { "epoch": 8.227767114523353, "grad_norm": 0.7936230525683274, "learning_rate": 1.602570353459766e-06, "loss": 0.0246, "step": 12860 }, { "epoch": 8.228406909788868, "grad_norm": 0.7499733084848496, "learning_rate": 1.6014453870486136e-06, "loss": 0.0257, "step": 12861 }, { "epoch": 8.229046705054383, "grad_norm": 0.818749296209217, "learning_rate": 1.6003207812564569e-06, "loss": 0.0263, "step": 12862 }, { "epoch": 8.229686500319898, "grad_norm": 0.7075058237746682, "learning_rate": 1.5991965361315842e-06, "loss": 0.022, "step": 12863 }, { "epoch": 8.230326295585412, "grad_norm": 0.7859994088878645, "learning_rate": 1.5980726517222656e-06, "loss": 0.0253, "step": 12864 }, { "epoch": 8.230966090850927, "grad_norm": 0.8099957280994582, "learning_rate": 1.5969491280767634e-06, "loss": 0.0268, "step": 12865 }, { "epoch": 8.231605886116443, "grad_norm": 0.9219338159905438, "learning_rate": 1.5958259652433184e-06, "loss": 0.0318, "step": 12866 }, { "epoch": 8.232245681381958, "grad_norm": 0.733194706710943, "learning_rate": 1.594703163270155e-06, "loss": 0.0234, "step": 12867 }, { "epoch": 8.232885476647473, "grad_norm": 0.759224344586337, "learning_rate": 1.5935807222054866e-06, "loss": 0.0277, "step": 12868 }, { "epoch": 8.233525271912988, "grad_norm": 0.8132204761209446, "learning_rate": 1.5924586420975097e-06, "loss": 0.0291, "step": 12869 }, { "epoch": 8.234165067178504, "grad_norm": 0.7816326750193046, "learning_rate": 1.591336922994402e-06, "loss": 0.0254, "step": 12870 }, { "epoch": 8.234804862444017, "grad_norm": 0.7378813455361388, "learning_rate": 1.5902155649443284e-06, "loss": 0.0236, "step": 12871 }, { "epoch": 8.235444657709532, "grad_norm": 0.8300149091178602, "learning_rate": 1.5890945679954395e-06, "loss": 0.029, "step": 12872 }, { "epoch": 8.236084452975048, "grad_norm": 0.8083469480683595, "learning_rate": 1.5879739321958677e-06, "loss": 0.0262, "step": 12873 }, { "epoch": 8.236724248240563, "grad_norm": 0.8037307022396761, "learning_rate": 1.5868536575937287e-06, "loss": 0.0259, "step": 12874 }, { "epoch": 8.237364043506078, "grad_norm": 0.6961567981567583, "learning_rate": 1.5857337442371312e-06, "loss": 0.0233, "step": 12875 }, { "epoch": 8.238003838771593, "grad_norm": 0.836602148733646, "learning_rate": 1.5846141921741587e-06, "loss": 0.0275, "step": 12876 }, { "epoch": 8.238643634037109, "grad_norm": 0.696777271014527, "learning_rate": 1.5834950014528794e-06, "loss": 0.0237, "step": 12877 }, { "epoch": 8.239283429302624, "grad_norm": 0.7811554994654815, "learning_rate": 1.582376172121357e-06, "loss": 0.0245, "step": 12878 }, { "epoch": 8.239923224568138, "grad_norm": 0.944602713070919, "learning_rate": 1.5812577042276267e-06, "loss": 0.0283, "step": 12879 }, { "epoch": 8.240563019833653, "grad_norm": 0.8351588984782646, "learning_rate": 1.5801395978197133e-06, "loss": 0.0265, "step": 12880 }, { "epoch": 8.241202815099168, "grad_norm": 0.6973947041757587, "learning_rate": 1.5790218529456292e-06, "loss": 0.0234, "step": 12881 }, { "epoch": 8.241842610364683, "grad_norm": 0.6843416650132079, "learning_rate": 1.5779044696533673e-06, "loss": 0.0216, "step": 12882 }, { "epoch": 8.242482405630199, "grad_norm": 0.7357172673105963, "learning_rate": 1.5767874479909061e-06, "loss": 0.0236, "step": 12883 }, { "epoch": 8.243122200895714, "grad_norm": 0.7667438219530158, "learning_rate": 1.5756707880062061e-06, "loss": 0.0282, "step": 12884 }, { "epoch": 8.24376199616123, "grad_norm": 0.6418136447072397, "learning_rate": 1.5745544897472176e-06, "loss": 0.0238, "step": 12885 }, { "epoch": 8.244401791426743, "grad_norm": 0.7977859462938915, "learning_rate": 1.5734385532618734e-06, "loss": 0.0238, "step": 12886 }, { "epoch": 8.245041586692258, "grad_norm": 0.7748752815829629, "learning_rate": 1.5723229785980865e-06, "loss": 0.0267, "step": 12887 }, { "epoch": 8.245681381957773, "grad_norm": 0.8225505699910234, "learning_rate": 1.5712077658037606e-06, "loss": 0.0293, "step": 12888 }, { "epoch": 8.246321177223289, "grad_norm": 0.6772433658741293, "learning_rate": 1.5700929149267785e-06, "loss": 0.0254, "step": 12889 }, { "epoch": 8.246960972488804, "grad_norm": 0.7100856086303519, "learning_rate": 1.568978426015012e-06, "loss": 0.0219, "step": 12890 }, { "epoch": 8.247600767754319, "grad_norm": 0.8921863547794542, "learning_rate": 1.5678642991163106e-06, "loss": 0.0269, "step": 12891 }, { "epoch": 8.248240563019834, "grad_norm": 0.6781013610128439, "learning_rate": 1.5667505342785206e-06, "loss": 0.0256, "step": 12892 }, { "epoch": 8.248880358285348, "grad_norm": 0.9668569067219309, "learning_rate": 1.5656371315494601e-06, "loss": 0.0227, "step": 12893 }, { "epoch": 8.249520153550863, "grad_norm": 0.8197544375531918, "learning_rate": 1.5645240909769343e-06, "loss": 0.0293, "step": 12894 }, { "epoch": 8.250159948816378, "grad_norm": 0.7779673119346115, "learning_rate": 1.563411412608743e-06, "loss": 0.0297, "step": 12895 }, { "epoch": 8.250799744081894, "grad_norm": 0.8723094334525363, "learning_rate": 1.5622990964926566e-06, "loss": 0.0305, "step": 12896 }, { "epoch": 8.251439539347409, "grad_norm": 0.7727238181017397, "learning_rate": 1.5611871426764358e-06, "loss": 0.0246, "step": 12897 }, { "epoch": 8.252079334612924, "grad_norm": 0.7375466325710154, "learning_rate": 1.560075551207829e-06, "loss": 0.0242, "step": 12898 }, { "epoch": 8.25271912987844, "grad_norm": 0.8094335808399966, "learning_rate": 1.5589643221345651e-06, "loss": 0.029, "step": 12899 }, { "epoch": 8.253358925143955, "grad_norm": 0.7534364604917154, "learning_rate": 1.557853455504358e-06, "loss": 0.028, "step": 12900 }, { "epoch": 8.253998720409468, "grad_norm": 0.9136188807812458, "learning_rate": 1.5567429513649057e-06, "loss": 0.0301, "step": 12901 }, { "epoch": 8.254638515674984, "grad_norm": 0.801533644774346, "learning_rate": 1.555632809763893e-06, "loss": 0.0242, "step": 12902 }, { "epoch": 8.255278310940499, "grad_norm": 0.7841634147648013, "learning_rate": 1.5545230307489844e-06, "loss": 0.0245, "step": 12903 }, { "epoch": 8.255918106206014, "grad_norm": 0.7379847972162552, "learning_rate": 1.5534136143678324e-06, "loss": 0.0251, "step": 12904 }, { "epoch": 8.25655790147153, "grad_norm": 0.7453984381775349, "learning_rate": 1.5523045606680765e-06, "loss": 0.0216, "step": 12905 }, { "epoch": 8.257197696737045, "grad_norm": 0.8138759445304666, "learning_rate": 1.5511958696973351e-06, "loss": 0.0247, "step": 12906 }, { "epoch": 8.25783749200256, "grad_norm": 0.7892076293933422, "learning_rate": 1.5500875415032113e-06, "loss": 0.0276, "step": 12907 }, { "epoch": 8.258477287268073, "grad_norm": 0.7483206499772654, "learning_rate": 1.5489795761332992e-06, "loss": 0.0246, "step": 12908 }, { "epoch": 8.259117082533589, "grad_norm": 0.7405038089115885, "learning_rate": 1.5478719736351721e-06, "loss": 0.0227, "step": 12909 }, { "epoch": 8.259756877799104, "grad_norm": 0.7219004132041495, "learning_rate": 1.5467647340563873e-06, "loss": 0.0222, "step": 12910 }, { "epoch": 8.26039667306462, "grad_norm": 0.7626008415082852, "learning_rate": 1.5456578574444848e-06, "loss": 0.025, "step": 12911 }, { "epoch": 8.261036468330134, "grad_norm": 0.8221293572396469, "learning_rate": 1.5445513438469972e-06, "loss": 0.0272, "step": 12912 }, { "epoch": 8.26167626359565, "grad_norm": 0.7992426601208289, "learning_rate": 1.5434451933114335e-06, "loss": 0.03, "step": 12913 }, { "epoch": 8.262316058861165, "grad_norm": 0.7396498631858374, "learning_rate": 1.542339405885288e-06, "loss": 0.0238, "step": 12914 }, { "epoch": 8.26295585412668, "grad_norm": 0.884600546359299, "learning_rate": 1.5412339816160459e-06, "loss": 0.0267, "step": 12915 }, { "epoch": 8.263595649392194, "grad_norm": 0.7081126477052258, "learning_rate": 1.5401289205511715e-06, "loss": 0.024, "step": 12916 }, { "epoch": 8.264235444657709, "grad_norm": 0.7670129151809643, "learning_rate": 1.539024222738108e-06, "loss": 0.0243, "step": 12917 }, { "epoch": 8.264875239923224, "grad_norm": 0.7276764065576756, "learning_rate": 1.5379198882242963e-06, "loss": 0.0252, "step": 12918 }, { "epoch": 8.26551503518874, "grad_norm": 0.7287219580329093, "learning_rate": 1.5368159170571517e-06, "loss": 0.0245, "step": 12919 }, { "epoch": 8.266154830454255, "grad_norm": 0.7971913133457064, "learning_rate": 1.5357123092840754e-06, "loss": 0.0253, "step": 12920 }, { "epoch": 8.26679462571977, "grad_norm": 0.813960274198768, "learning_rate": 1.5346090649524548e-06, "loss": 0.0258, "step": 12921 }, { "epoch": 8.267434420985285, "grad_norm": 0.7670496585992003, "learning_rate": 1.5335061841096643e-06, "loss": 0.0237, "step": 12922 }, { "epoch": 8.268074216250799, "grad_norm": 0.7414075015382897, "learning_rate": 1.5324036668030572e-06, "loss": 0.0234, "step": 12923 }, { "epoch": 8.268714011516314, "grad_norm": 0.949551859262686, "learning_rate": 1.5313015130799713e-06, "loss": 0.0252, "step": 12924 }, { "epoch": 8.26935380678183, "grad_norm": 0.7809783188208994, "learning_rate": 1.5301997229877374e-06, "loss": 0.0274, "step": 12925 }, { "epoch": 8.269993602047345, "grad_norm": 0.7135042970844826, "learning_rate": 1.5290982965736601e-06, "loss": 0.0211, "step": 12926 }, { "epoch": 8.27063339731286, "grad_norm": 0.7234379873096287, "learning_rate": 1.5279972338850303e-06, "loss": 0.0218, "step": 12927 }, { "epoch": 8.271273192578375, "grad_norm": 0.7469661023981319, "learning_rate": 1.5268965349691322e-06, "loss": 0.0211, "step": 12928 }, { "epoch": 8.27191298784389, "grad_norm": 0.8237038705969629, "learning_rate": 1.525796199873223e-06, "loss": 0.0284, "step": 12929 }, { "epoch": 8.272552783109404, "grad_norm": 0.8049236750345126, "learning_rate": 1.5246962286445521e-06, "loss": 0.0257, "step": 12930 }, { "epoch": 8.27319257837492, "grad_norm": 0.8974360586084204, "learning_rate": 1.5235966213303478e-06, "loss": 0.0283, "step": 12931 }, { "epoch": 8.273832373640435, "grad_norm": 0.9448589942583414, "learning_rate": 1.5224973779778262e-06, "loss": 0.0277, "step": 12932 }, { "epoch": 8.27447216890595, "grad_norm": 0.8325301506460038, "learning_rate": 1.521398498634188e-06, "loss": 0.0285, "step": 12933 }, { "epoch": 8.275111964171465, "grad_norm": 0.8643180991496228, "learning_rate": 1.5202999833466126e-06, "loss": 0.028, "step": 12934 }, { "epoch": 8.27575175943698, "grad_norm": 0.8291923315399987, "learning_rate": 1.5192018321622736e-06, "loss": 0.0278, "step": 12935 }, { "epoch": 8.276391554702496, "grad_norm": 0.8690621128756166, "learning_rate": 1.5181040451283235e-06, "loss": 0.026, "step": 12936 }, { "epoch": 8.277031349968011, "grad_norm": 0.7726016116374962, "learning_rate": 1.5170066222918934e-06, "loss": 0.0216, "step": 12937 }, { "epoch": 8.277671145233525, "grad_norm": 0.8092629523817036, "learning_rate": 1.515909563700112e-06, "loss": 0.0275, "step": 12938 }, { "epoch": 8.27831094049904, "grad_norm": 0.6818995517505846, "learning_rate": 1.514812869400082e-06, "loss": 0.0245, "step": 12939 }, { "epoch": 8.278950735764555, "grad_norm": 0.738592463870055, "learning_rate": 1.513716539438893e-06, "loss": 0.0232, "step": 12940 }, { "epoch": 8.27959053103007, "grad_norm": 0.8074241797113964, "learning_rate": 1.512620573863618e-06, "loss": 0.0259, "step": 12941 }, { "epoch": 8.280230326295586, "grad_norm": 0.8547614403699598, "learning_rate": 1.51152497272132e-06, "loss": 0.0238, "step": 12942 }, { "epoch": 8.280870121561101, "grad_norm": 0.7446488876864796, "learning_rate": 1.5104297360590404e-06, "loss": 0.0234, "step": 12943 }, { "epoch": 8.281509916826616, "grad_norm": 0.7707452095474258, "learning_rate": 1.5093348639238025e-06, "loss": 0.0256, "step": 12944 }, { "epoch": 8.28214971209213, "grad_norm": 0.7280761141656444, "learning_rate": 1.508240356362627e-06, "loss": 0.0213, "step": 12945 }, { "epoch": 8.282789507357645, "grad_norm": 0.7034661271089054, "learning_rate": 1.5071462134225034e-06, "loss": 0.0202, "step": 12946 }, { "epoch": 8.28342930262316, "grad_norm": 0.7498226783199099, "learning_rate": 1.5060524351504102e-06, "loss": 0.0244, "step": 12947 }, { "epoch": 8.284069097888676, "grad_norm": 0.8260942752826488, "learning_rate": 1.5049590215933185e-06, "loss": 0.027, "step": 12948 }, { "epoch": 8.28470889315419, "grad_norm": 0.9772702853159428, "learning_rate": 1.5038659727981752e-06, "loss": 0.0278, "step": 12949 }, { "epoch": 8.285348688419706, "grad_norm": 0.9003726264285637, "learning_rate": 1.502773288811914e-06, "loss": 0.0281, "step": 12950 }, { "epoch": 8.285988483685221, "grad_norm": 0.8103015023340239, "learning_rate": 1.5016809696814493e-06, "loss": 0.0228, "step": 12951 }, { "epoch": 8.286628278950737, "grad_norm": 0.7632981605300158, "learning_rate": 1.5005890154536894e-06, "loss": 0.0223, "step": 12952 }, { "epoch": 8.28726807421625, "grad_norm": 0.8033412311434549, "learning_rate": 1.499497426175518e-06, "loss": 0.0268, "step": 12953 }, { "epoch": 8.287907869481765, "grad_norm": 0.7107096661922451, "learning_rate": 1.4984062018938028e-06, "loss": 0.022, "step": 12954 }, { "epoch": 8.28854766474728, "grad_norm": 0.6884743628561939, "learning_rate": 1.4973153426554042e-06, "loss": 0.0211, "step": 12955 }, { "epoch": 8.289187460012796, "grad_norm": 0.8060318567939101, "learning_rate": 1.4962248485071607e-06, "loss": 0.027, "step": 12956 }, { "epoch": 8.289827255278311, "grad_norm": 0.8284531858302594, "learning_rate": 1.4951347194958933e-06, "loss": 0.0232, "step": 12957 }, { "epoch": 8.290467050543826, "grad_norm": 0.7101088398781988, "learning_rate": 1.4940449556684133e-06, "loss": 0.0225, "step": 12958 }, { "epoch": 8.291106845809342, "grad_norm": 0.7213075795069244, "learning_rate": 1.4929555570715116e-06, "loss": 0.0221, "step": 12959 }, { "epoch": 8.291746641074855, "grad_norm": 0.6684877064609305, "learning_rate": 1.4918665237519659e-06, "loss": 0.0213, "step": 12960 }, { "epoch": 8.29238643634037, "grad_norm": 0.7622794508962231, "learning_rate": 1.4907778557565379e-06, "loss": 0.0218, "step": 12961 }, { "epoch": 8.293026231605886, "grad_norm": 0.7101739282033429, "learning_rate": 1.4896895531319699e-06, "loss": 0.0232, "step": 12962 }, { "epoch": 8.293666026871401, "grad_norm": 0.8311373671404407, "learning_rate": 1.488601615924996e-06, "loss": 0.0274, "step": 12963 }, { "epoch": 8.294305822136916, "grad_norm": 0.8284418639753195, "learning_rate": 1.4875140441823243e-06, "loss": 0.0262, "step": 12964 }, { "epoch": 8.294945617402432, "grad_norm": 0.8358801488319156, "learning_rate": 1.4864268379506597e-06, "loss": 0.0282, "step": 12965 }, { "epoch": 8.295585412667947, "grad_norm": 0.7703623713134731, "learning_rate": 1.4853399972766824e-06, "loss": 0.0252, "step": 12966 }, { "epoch": 8.29622520793346, "grad_norm": 0.945848128148205, "learning_rate": 1.4842535222070598e-06, "loss": 0.0308, "step": 12967 }, { "epoch": 8.296865003198976, "grad_norm": 0.7415587251857167, "learning_rate": 1.4831674127884399e-06, "loss": 0.0289, "step": 12968 }, { "epoch": 8.297504798464491, "grad_norm": 0.8097083304526487, "learning_rate": 1.4820816690674622e-06, "loss": 0.0245, "step": 12969 }, { "epoch": 8.298144593730006, "grad_norm": 0.7265606484885059, "learning_rate": 1.4809962910907472e-06, "loss": 0.0254, "step": 12970 }, { "epoch": 8.298784388995522, "grad_norm": 0.8813192225399203, "learning_rate": 1.4799112789048942e-06, "loss": 0.028, "step": 12971 }, { "epoch": 8.299424184261037, "grad_norm": 0.7936926075976187, "learning_rate": 1.478826632556497e-06, "loss": 0.0259, "step": 12972 }, { "epoch": 8.300063979526552, "grad_norm": 0.8367695353388185, "learning_rate": 1.4777423520921264e-06, "loss": 0.0227, "step": 12973 }, { "epoch": 8.300703774792067, "grad_norm": 0.7959396793521798, "learning_rate": 1.4766584375583382e-06, "loss": 0.0266, "step": 12974 }, { "epoch": 8.30134357005758, "grad_norm": 0.8437504471717477, "learning_rate": 1.4755748890016762e-06, "loss": 0.0266, "step": 12975 }, { "epoch": 8.301983365323096, "grad_norm": 0.7816288765290363, "learning_rate": 1.4744917064686648e-06, "loss": 0.0251, "step": 12976 }, { "epoch": 8.302623160588611, "grad_norm": 0.7405068521613651, "learning_rate": 1.473408890005813e-06, "loss": 0.0234, "step": 12977 }, { "epoch": 8.303262955854127, "grad_norm": 0.7595550437249688, "learning_rate": 1.4723264396596137e-06, "loss": 0.0216, "step": 12978 }, { "epoch": 8.303902751119642, "grad_norm": 0.7451020321079147, "learning_rate": 1.47124435547655e-06, "loss": 0.0246, "step": 12979 }, { "epoch": 8.304542546385157, "grad_norm": 0.8084670399579138, "learning_rate": 1.4701626375030819e-06, "loss": 0.0266, "step": 12980 }, { "epoch": 8.305182341650672, "grad_norm": 0.7482296824439224, "learning_rate": 1.469081285785655e-06, "loss": 0.0228, "step": 12981 }, { "epoch": 8.305822136916186, "grad_norm": 0.840137109356208, "learning_rate": 1.4680003003707044e-06, "loss": 0.0272, "step": 12982 }, { "epoch": 8.306461932181701, "grad_norm": 0.8051701860843392, "learning_rate": 1.4669196813046439e-06, "loss": 0.0227, "step": 12983 }, { "epoch": 8.307101727447217, "grad_norm": 0.6833015049788027, "learning_rate": 1.4658394286338706e-06, "loss": 0.0222, "step": 12984 }, { "epoch": 8.307741522712732, "grad_norm": 0.8203856526450704, "learning_rate": 1.4647595424047735e-06, "loss": 0.0304, "step": 12985 }, { "epoch": 8.308381317978247, "grad_norm": 0.8756990017754123, "learning_rate": 1.4636800226637183e-06, "loss": 0.029, "step": 12986 }, { "epoch": 8.309021113243762, "grad_norm": 0.7446961909585078, "learning_rate": 1.4626008694570582e-06, "loss": 0.0225, "step": 12987 }, { "epoch": 8.309660908509278, "grad_norm": 0.8638697000982685, "learning_rate": 1.4615220828311284e-06, "loss": 0.0264, "step": 12988 }, { "epoch": 8.310300703774793, "grad_norm": 0.9060828313925405, "learning_rate": 1.4604436628322527e-06, "loss": 0.0269, "step": 12989 }, { "epoch": 8.310940499040306, "grad_norm": 0.8108047078858209, "learning_rate": 1.4593656095067354e-06, "loss": 0.0224, "step": 12990 }, { "epoch": 8.311580294305822, "grad_norm": 0.8387759765566237, "learning_rate": 1.4582879229008673e-06, "loss": 0.0248, "step": 12991 }, { "epoch": 8.312220089571337, "grad_norm": 0.7771711152816811, "learning_rate": 1.4572106030609201e-06, "loss": 0.0229, "step": 12992 }, { "epoch": 8.312859884836852, "grad_norm": 0.9095611502609382, "learning_rate": 1.4561336500331535e-06, "loss": 0.026, "step": 12993 }, { "epoch": 8.313499680102368, "grad_norm": 0.7815798842774332, "learning_rate": 1.4550570638638072e-06, "loss": 0.0266, "step": 12994 }, { "epoch": 8.314139475367883, "grad_norm": 0.8008371691464912, "learning_rate": 1.4539808445991133e-06, "loss": 0.0229, "step": 12995 }, { "epoch": 8.314779270633398, "grad_norm": 0.8244620675173526, "learning_rate": 1.45290499228528e-06, "loss": 0.0247, "step": 12996 }, { "epoch": 8.315419065898912, "grad_norm": 0.7813722396023758, "learning_rate": 1.4518295069685008e-06, "loss": 0.0219, "step": 12997 }, { "epoch": 8.316058861164427, "grad_norm": 0.7998308584490119, "learning_rate": 1.4507543886949548e-06, "loss": 0.0253, "step": 12998 }, { "epoch": 8.316698656429942, "grad_norm": 0.6997962589935756, "learning_rate": 1.4496796375108102e-06, "loss": 0.0208, "step": 12999 }, { "epoch": 8.317338451695457, "grad_norm": 0.8342486472938591, "learning_rate": 1.448605253462212e-06, "loss": 0.027, "step": 13000 }, { "epoch": 8.317978246960973, "grad_norm": 0.7924310396158626, "learning_rate": 1.4475312365952898e-06, "loss": 0.0236, "step": 13001 }, { "epoch": 8.318618042226488, "grad_norm": 0.7605851236497205, "learning_rate": 1.4464575869561659e-06, "loss": 0.0225, "step": 13002 }, { "epoch": 8.319257837492003, "grad_norm": 0.7089109907536021, "learning_rate": 1.4453843045909365e-06, "loss": 0.0176, "step": 13003 }, { "epoch": 8.319897632757517, "grad_norm": 0.7578481681060529, "learning_rate": 1.4443113895456885e-06, "loss": 0.0259, "step": 13004 }, { "epoch": 8.320537428023032, "grad_norm": 0.7744511442338498, "learning_rate": 1.4432388418664888e-06, "loss": 0.0224, "step": 13005 }, { "epoch": 8.321177223288547, "grad_norm": 0.8660472798202611, "learning_rate": 1.4421666615993933e-06, "loss": 0.0249, "step": 13006 }, { "epoch": 8.321817018554063, "grad_norm": 0.7948378838143699, "learning_rate": 1.4410948487904374e-06, "loss": 0.0258, "step": 13007 }, { "epoch": 8.322456813819578, "grad_norm": 0.8069527994525817, "learning_rate": 1.440023403485642e-06, "loss": 0.0255, "step": 13008 }, { "epoch": 8.323096609085093, "grad_norm": 0.867452490599027, "learning_rate": 1.4389523257310156e-06, "loss": 0.025, "step": 13009 }, { "epoch": 8.323736404350608, "grad_norm": 0.8153066997046752, "learning_rate": 1.4378816155725483e-06, "loss": 0.0261, "step": 13010 }, { "epoch": 8.324376199616124, "grad_norm": 0.9006663640058246, "learning_rate": 1.436811273056211e-06, "loss": 0.0271, "step": 13011 }, { "epoch": 8.325015994881637, "grad_norm": 0.8555926077813714, "learning_rate": 1.4357412982279672e-06, "loss": 0.0286, "step": 13012 }, { "epoch": 8.325655790147152, "grad_norm": 0.7900046782066523, "learning_rate": 1.4346716911337576e-06, "loss": 0.0208, "step": 13013 }, { "epoch": 8.326295585412668, "grad_norm": 0.7940721511113856, "learning_rate": 1.433602451819507e-06, "loss": 0.0272, "step": 13014 }, { "epoch": 8.326935380678183, "grad_norm": 0.8004123672067135, "learning_rate": 1.4325335803311313e-06, "loss": 0.0213, "step": 13015 }, { "epoch": 8.327575175943698, "grad_norm": 0.7359973818513967, "learning_rate": 1.4314650767145222e-06, "loss": 0.0243, "step": 13016 }, { "epoch": 8.328214971209214, "grad_norm": 0.7439787026665173, "learning_rate": 1.4303969410155616e-06, "loss": 0.0217, "step": 13017 }, { "epoch": 8.328854766474729, "grad_norm": 0.81988870982047, "learning_rate": 1.4293291732801084e-06, "loss": 0.0261, "step": 13018 }, { "epoch": 8.329494561740242, "grad_norm": 0.8452946381681025, "learning_rate": 1.4282617735540183e-06, "loss": 0.0254, "step": 13019 }, { "epoch": 8.330134357005758, "grad_norm": 0.8045940527588143, "learning_rate": 1.4271947418831188e-06, "loss": 0.0214, "step": 13020 }, { "epoch": 8.330774152271273, "grad_norm": 0.705832302793147, "learning_rate": 1.4261280783132281e-06, "loss": 0.0217, "step": 13021 }, { "epoch": 8.331413947536788, "grad_norm": 0.7526481760963598, "learning_rate": 1.4250617828901448e-06, "loss": 0.025, "step": 13022 }, { "epoch": 8.332053742802303, "grad_norm": 0.751650045561705, "learning_rate": 1.4239958556596556e-06, "loss": 0.0234, "step": 13023 }, { "epoch": 8.332693538067819, "grad_norm": 0.7837835409390526, "learning_rate": 1.4229302966675263e-06, "loss": 0.0236, "step": 13024 }, { "epoch": 8.333333333333334, "grad_norm": 0.818999274066871, "learning_rate": 1.4218651059595156e-06, "loss": 0.0252, "step": 13025 }, { "epoch": 8.33397312859885, "grad_norm": 0.8220241005124931, "learning_rate": 1.4208002835813573e-06, "loss": 0.0203, "step": 13026 }, { "epoch": 8.334612923864363, "grad_norm": 0.7204131547087621, "learning_rate": 1.4197358295787734e-06, "loss": 0.0216, "step": 13027 }, { "epoch": 8.335252719129878, "grad_norm": 0.9605281488372325, "learning_rate": 1.4186717439974685e-06, "loss": 0.031, "step": 13028 }, { "epoch": 8.335892514395393, "grad_norm": 0.8036625857920173, "learning_rate": 1.4176080268831349e-06, "loss": 0.0259, "step": 13029 }, { "epoch": 8.336532309660909, "grad_norm": 0.6744321444003196, "learning_rate": 1.4165446782814462e-06, "loss": 0.0221, "step": 13030 }, { "epoch": 8.337172104926424, "grad_norm": 0.8585397656715563, "learning_rate": 1.4154816982380581e-06, "loss": 0.0293, "step": 13031 }, { "epoch": 8.337811900191939, "grad_norm": 0.8046062221594978, "learning_rate": 1.4144190867986184e-06, "loss": 0.0277, "step": 13032 }, { "epoch": 8.338451695457454, "grad_norm": 0.7295961613850067, "learning_rate": 1.4133568440087497e-06, "loss": 0.0224, "step": 13033 }, { "epoch": 8.339091490722968, "grad_norm": 0.7407324638063533, "learning_rate": 1.4122949699140632e-06, "loss": 0.0288, "step": 13034 }, { "epoch": 8.339731285988483, "grad_norm": 0.76948561064728, "learning_rate": 1.4112334645601556e-06, "loss": 0.0255, "step": 13035 }, { "epoch": 8.340371081253998, "grad_norm": 0.7740654771112542, "learning_rate": 1.4101723279926038e-06, "loss": 0.0238, "step": 13036 }, { "epoch": 8.341010876519514, "grad_norm": 0.6477047274924569, "learning_rate": 1.4091115602569716e-06, "loss": 0.0206, "step": 13037 }, { "epoch": 8.341650671785029, "grad_norm": 0.938195916599871, "learning_rate": 1.4080511613988057e-06, "loss": 0.0298, "step": 13038 }, { "epoch": 8.342290467050544, "grad_norm": 0.8126624297851378, "learning_rate": 1.4069911314636408e-06, "loss": 0.0262, "step": 13039 }, { "epoch": 8.34293026231606, "grad_norm": 0.7353737573649943, "learning_rate": 1.4059314704969917e-06, "loss": 0.0233, "step": 13040 }, { "epoch": 8.343570057581573, "grad_norm": 0.7926218735469904, "learning_rate": 1.4048721785443552e-06, "loss": 0.0262, "step": 13041 }, { "epoch": 8.344209852847088, "grad_norm": 0.8688238449169435, "learning_rate": 1.4038132556512196e-06, "loss": 0.025, "step": 13042 }, { "epoch": 8.344849648112604, "grad_norm": 0.7772921583018347, "learning_rate": 1.4027547018630517e-06, "loss": 0.0214, "step": 13043 }, { "epoch": 8.345489443378119, "grad_norm": 0.654463054471373, "learning_rate": 1.4016965172253027e-06, "loss": 0.0205, "step": 13044 }, { "epoch": 8.346129238643634, "grad_norm": 0.8750995953900266, "learning_rate": 1.4006387017834088e-06, "loss": 0.0279, "step": 13045 }, { "epoch": 8.34676903390915, "grad_norm": 0.7302440080132201, "learning_rate": 1.3995812555827936e-06, "loss": 0.0213, "step": 13046 }, { "epoch": 8.347408829174665, "grad_norm": 0.7950047429976485, "learning_rate": 1.398524178668861e-06, "loss": 0.0247, "step": 13047 }, { "epoch": 8.34804862444018, "grad_norm": 0.7635675376326884, "learning_rate": 1.3974674710869962e-06, "loss": 0.0216, "step": 13048 }, { "epoch": 8.348688419705693, "grad_norm": 0.7952851787123578, "learning_rate": 1.3964111328825791e-06, "loss": 0.0239, "step": 13049 }, { "epoch": 8.349328214971209, "grad_norm": 0.8541900934828315, "learning_rate": 1.395355164100962e-06, "loss": 0.022, "step": 13050 }, { "epoch": 8.349968010236724, "grad_norm": 0.7699469089339877, "learning_rate": 1.394299564787488e-06, "loss": 0.0267, "step": 13051 }, { "epoch": 8.35060780550224, "grad_norm": 0.8286660951743876, "learning_rate": 1.3932443349874824e-06, "loss": 0.0268, "step": 13052 }, { "epoch": 8.351247600767755, "grad_norm": 0.7546709220364254, "learning_rate": 1.392189474746255e-06, "loss": 0.0259, "step": 13053 }, { "epoch": 8.35188739603327, "grad_norm": 0.6563383906908987, "learning_rate": 1.3911349841090993e-06, "loss": 0.0201, "step": 13054 }, { "epoch": 8.352527191298785, "grad_norm": 0.7753005731708482, "learning_rate": 1.3900808631212915e-06, "loss": 0.0227, "step": 13055 }, { "epoch": 8.353166986564299, "grad_norm": 0.7070047235582628, "learning_rate": 1.3890271118280973e-06, "loss": 0.0212, "step": 13056 }, { "epoch": 8.353806781829814, "grad_norm": 0.7827480447099965, "learning_rate": 1.387973730274762e-06, "loss": 0.0252, "step": 13057 }, { "epoch": 8.35444657709533, "grad_norm": 0.7892810527879773, "learning_rate": 1.3869207185065126e-06, "loss": 0.0227, "step": 13058 }, { "epoch": 8.355086372360844, "grad_norm": 0.8555601642038032, "learning_rate": 1.3858680765685672e-06, "loss": 0.0249, "step": 13059 }, { "epoch": 8.35572616762636, "grad_norm": 0.7822867068853311, "learning_rate": 1.3848158045061244e-06, "loss": 0.021, "step": 13060 }, { "epoch": 8.356365962891875, "grad_norm": 0.8529786425127713, "learning_rate": 1.3837639023643634e-06, "loss": 0.0255, "step": 13061 }, { "epoch": 8.35700575815739, "grad_norm": 0.8779651222824297, "learning_rate": 1.3827123701884548e-06, "loss": 0.0269, "step": 13062 }, { "epoch": 8.357645553422905, "grad_norm": 0.7860852717471666, "learning_rate": 1.3816612080235491e-06, "loss": 0.0198, "step": 13063 }, { "epoch": 8.358285348688419, "grad_norm": 0.7878070559890583, "learning_rate": 1.3806104159147792e-06, "loss": 0.0248, "step": 13064 }, { "epoch": 8.358925143953934, "grad_norm": 0.8632680080514957, "learning_rate": 1.3795599939072669e-06, "loss": 0.0281, "step": 13065 }, { "epoch": 8.35956493921945, "grad_norm": 0.6323781676510812, "learning_rate": 1.3785099420461123e-06, "loss": 0.019, "step": 13066 }, { "epoch": 8.360204734484965, "grad_norm": 0.7774762699295024, "learning_rate": 1.3774602603764053e-06, "loss": 0.0261, "step": 13067 }, { "epoch": 8.36084452975048, "grad_norm": 0.6839067218952137, "learning_rate": 1.376410948943213e-06, "loss": 0.0189, "step": 13068 }, { "epoch": 8.361484325015995, "grad_norm": 0.8552672460351953, "learning_rate": 1.3753620077915975e-06, "loss": 0.0236, "step": 13069 }, { "epoch": 8.36212412028151, "grad_norm": 0.7565239742495105, "learning_rate": 1.3743134369665956e-06, "loss": 0.0252, "step": 13070 }, { "epoch": 8.362763915547024, "grad_norm": 0.7814059318723161, "learning_rate": 1.3732652365132282e-06, "loss": 0.0245, "step": 13071 }, { "epoch": 8.36340371081254, "grad_norm": 0.7719412085604006, "learning_rate": 1.3722174064765081e-06, "loss": 0.0264, "step": 13072 }, { "epoch": 8.364043506078055, "grad_norm": 0.7425173765159858, "learning_rate": 1.371169946901425e-06, "loss": 0.0223, "step": 13073 }, { "epoch": 8.36468330134357, "grad_norm": 0.9010926323416667, "learning_rate": 1.370122857832955e-06, "loss": 0.0285, "step": 13074 }, { "epoch": 8.365323096609085, "grad_norm": 0.800639027423464, "learning_rate": 1.3690761393160556e-06, "loss": 0.0263, "step": 13075 }, { "epoch": 8.3659628918746, "grad_norm": 0.7455614867522804, "learning_rate": 1.3680297913956752e-06, "loss": 0.0242, "step": 13076 }, { "epoch": 8.366602687140116, "grad_norm": 0.9156450735111902, "learning_rate": 1.3669838141167412e-06, "loss": 0.0287, "step": 13077 }, { "epoch": 8.36724248240563, "grad_norm": 0.8347292142920745, "learning_rate": 1.3659382075241634e-06, "loss": 0.0244, "step": 13078 }, { "epoch": 8.367882277671145, "grad_norm": 0.8256810864079419, "learning_rate": 1.3648929716628423e-06, "loss": 0.0276, "step": 13079 }, { "epoch": 8.36852207293666, "grad_norm": 0.7484040817475653, "learning_rate": 1.3638481065776555e-06, "loss": 0.0231, "step": 13080 }, { "epoch": 8.369161868202175, "grad_norm": 0.7850840338046697, "learning_rate": 1.3628036123134692e-06, "loss": 0.0226, "step": 13081 }, { "epoch": 8.36980166346769, "grad_norm": 0.7122811244925117, "learning_rate": 1.3617594889151308e-06, "loss": 0.023, "step": 13082 }, { "epoch": 8.370441458733206, "grad_norm": 0.7197435690747687, "learning_rate": 1.3607157364274747e-06, "loss": 0.0231, "step": 13083 }, { "epoch": 8.371081253998721, "grad_norm": 0.6663779997040259, "learning_rate": 1.3596723548953161e-06, "loss": 0.022, "step": 13084 }, { "epoch": 8.371721049264236, "grad_norm": 0.6534628071125004, "learning_rate": 1.3586293443634546e-06, "loss": 0.0192, "step": 13085 }, { "epoch": 8.37236084452975, "grad_norm": 0.7995284665109985, "learning_rate": 1.3575867048766789e-06, "loss": 0.0265, "step": 13086 }, { "epoch": 8.373000639795265, "grad_norm": 0.8196996534869183, "learning_rate": 1.3565444364797574e-06, "loss": 0.0257, "step": 13087 }, { "epoch": 8.37364043506078, "grad_norm": 0.8555585691713831, "learning_rate": 1.3555025392174393e-06, "loss": 0.0286, "step": 13088 }, { "epoch": 8.374280230326296, "grad_norm": 0.6390865993141371, "learning_rate": 1.3544610131344672e-06, "loss": 0.0216, "step": 13089 }, { "epoch": 8.37492002559181, "grad_norm": 0.7722472917497101, "learning_rate": 1.3534198582755598e-06, "loss": 0.0253, "step": 13090 }, { "epoch": 8.375559820857326, "grad_norm": 0.7259624300843066, "learning_rate": 1.3523790746854215e-06, "loss": 0.0249, "step": 13091 }, { "epoch": 8.376199616122841, "grad_norm": 0.7213909263744326, "learning_rate": 1.3513386624087433e-06, "loss": 0.0236, "step": 13092 }, { "epoch": 8.376839411388357, "grad_norm": 0.7741832949775139, "learning_rate": 1.3502986214901991e-06, "loss": 0.0234, "step": 13093 }, { "epoch": 8.37747920665387, "grad_norm": 0.7493306726720015, "learning_rate": 1.349258951974446e-06, "loss": 0.0257, "step": 13094 }, { "epoch": 8.378119001919385, "grad_norm": 0.8008540299879299, "learning_rate": 1.3482196539061244e-06, "loss": 0.0269, "step": 13095 }, { "epoch": 8.3787587971849, "grad_norm": 0.7265720550710804, "learning_rate": 1.34718072732986e-06, "loss": 0.0203, "step": 13096 }, { "epoch": 8.379398592450416, "grad_norm": 0.7856840192364681, "learning_rate": 1.3461421722902634e-06, "loss": 0.0254, "step": 13097 }, { "epoch": 8.380038387715931, "grad_norm": 0.8335891757024502, "learning_rate": 1.3451039888319262e-06, "loss": 0.0284, "step": 13098 }, { "epoch": 8.380678182981447, "grad_norm": 0.6897934702905861, "learning_rate": 1.3440661769994302e-06, "loss": 0.0203, "step": 13099 }, { "epoch": 8.381317978246962, "grad_norm": 0.8279131745833953, "learning_rate": 1.343028736837334e-06, "loss": 0.0298, "step": 13100 }, { "epoch": 8.381957773512475, "grad_norm": 0.7551993879771755, "learning_rate": 1.3419916683901824e-06, "loss": 0.0203, "step": 13101 }, { "epoch": 8.38259756877799, "grad_norm": 0.769683298349727, "learning_rate": 1.3409549717025094e-06, "loss": 0.0247, "step": 13102 }, { "epoch": 8.383237364043506, "grad_norm": 0.6996936338784949, "learning_rate": 1.3399186468188263e-06, "loss": 0.0219, "step": 13103 }, { "epoch": 8.383877159309021, "grad_norm": 0.9659163586215359, "learning_rate": 1.338882693783632e-06, "loss": 0.0299, "step": 13104 }, { "epoch": 8.384516954574536, "grad_norm": 0.8620150017989641, "learning_rate": 1.337847112641406e-06, "loss": 0.0255, "step": 13105 }, { "epoch": 8.385156749840052, "grad_norm": 0.9150483990825826, "learning_rate": 1.3368119034366179e-06, "loss": 0.03, "step": 13106 }, { "epoch": 8.385796545105567, "grad_norm": 0.9427332827780095, "learning_rate": 1.335777066213716e-06, "loss": 0.0325, "step": 13107 }, { "epoch": 8.38643634037108, "grad_norm": 0.8041132270661709, "learning_rate": 1.3347426010171316e-06, "loss": 0.0271, "step": 13108 }, { "epoch": 8.387076135636596, "grad_norm": 0.8303744496541728, "learning_rate": 1.3337085078912892e-06, "loss": 0.0259, "step": 13109 }, { "epoch": 8.387715930902111, "grad_norm": 0.8535657602399416, "learning_rate": 1.3326747868805867e-06, "loss": 0.0275, "step": 13110 }, { "epoch": 8.388355726167626, "grad_norm": 0.7532352966795617, "learning_rate": 1.3316414380294107e-06, "loss": 0.0247, "step": 13111 }, { "epoch": 8.388995521433142, "grad_norm": 0.8530890195754358, "learning_rate": 1.3306084613821312e-06, "loss": 0.0266, "step": 13112 }, { "epoch": 8.389635316698657, "grad_norm": 0.8068833065967799, "learning_rate": 1.3295758569831029e-06, "loss": 0.0272, "step": 13113 }, { "epoch": 8.390275111964172, "grad_norm": 0.7517011625569932, "learning_rate": 1.3285436248766648e-06, "loss": 0.0223, "step": 13114 }, { "epoch": 8.390914907229687, "grad_norm": 0.8773716833320477, "learning_rate": 1.3275117651071346e-06, "loss": 0.0287, "step": 13115 }, { "epoch": 8.3915547024952, "grad_norm": 0.7517903589088929, "learning_rate": 1.3264802777188246e-06, "loss": 0.0243, "step": 13116 }, { "epoch": 8.392194497760716, "grad_norm": 0.8982142590586899, "learning_rate": 1.325449162756023e-06, "loss": 0.0291, "step": 13117 }, { "epoch": 8.392834293026231, "grad_norm": 0.8990721972540774, "learning_rate": 1.3244184202630017e-06, "loss": 0.0281, "step": 13118 }, { "epoch": 8.393474088291747, "grad_norm": 0.8695957058544894, "learning_rate": 1.3233880502840213e-06, "loss": 0.0264, "step": 13119 }, { "epoch": 8.394113883557262, "grad_norm": 0.9330987241084654, "learning_rate": 1.3223580528633251e-06, "loss": 0.0275, "step": 13120 }, { "epoch": 8.394753678822777, "grad_norm": 0.8520401087655317, "learning_rate": 1.3213284280451366e-06, "loss": 0.027, "step": 13121 }, { "epoch": 8.395393474088293, "grad_norm": 0.6655331454790631, "learning_rate": 1.3202991758736662e-06, "loss": 0.0198, "step": 13122 }, { "epoch": 8.396033269353806, "grad_norm": 0.7515598434192163, "learning_rate": 1.3192702963931104e-06, "loss": 0.0212, "step": 13123 }, { "epoch": 8.396673064619321, "grad_norm": 0.8251706018717314, "learning_rate": 1.3182417896476473e-06, "loss": 0.0289, "step": 13124 }, { "epoch": 8.397312859884837, "grad_norm": 0.7776302553553769, "learning_rate": 1.3172136556814375e-06, "loss": 0.0255, "step": 13125 }, { "epoch": 8.397952655150352, "grad_norm": 0.771077265656909, "learning_rate": 1.3161858945386275e-06, "loss": 0.0229, "step": 13126 }, { "epoch": 8.398592450415867, "grad_norm": 0.8259942060911635, "learning_rate": 1.3151585062633486e-06, "loss": 0.0247, "step": 13127 }, { "epoch": 8.399232245681382, "grad_norm": 0.7705824958169577, "learning_rate": 1.3141314908997115e-06, "loss": 0.0201, "step": 13128 }, { "epoch": 8.399872040946898, "grad_norm": 0.7719187890724724, "learning_rate": 1.3131048484918197e-06, "loss": 0.0285, "step": 13129 }, { "epoch": 8.400511836212413, "grad_norm": 0.8255354050054423, "learning_rate": 1.3120785790837531e-06, "loss": 0.0229, "step": 13130 }, { "epoch": 8.401151631477926, "grad_norm": 0.8468118477822614, "learning_rate": 1.3110526827195768e-06, "loss": 0.0287, "step": 13131 }, { "epoch": 8.401791426743442, "grad_norm": 0.8696779129809622, "learning_rate": 1.3100271594433412e-06, "loss": 0.0256, "step": 13132 }, { "epoch": 8.402431222008957, "grad_norm": 0.7529940786100827, "learning_rate": 1.309002009299082e-06, "loss": 0.0202, "step": 13133 }, { "epoch": 8.403071017274472, "grad_norm": 0.7819372954657675, "learning_rate": 1.307977232330817e-06, "loss": 0.0258, "step": 13134 }, { "epoch": 8.403710812539988, "grad_norm": 0.7764861220990837, "learning_rate": 1.306952828582545e-06, "loss": 0.026, "step": 13135 }, { "epoch": 8.404350607805503, "grad_norm": 0.730610058697258, "learning_rate": 1.3059287980982583e-06, "loss": 0.0234, "step": 13136 }, { "epoch": 8.404990403071018, "grad_norm": 0.7284747160617663, "learning_rate": 1.3049051409219227e-06, "loss": 0.0233, "step": 13137 }, { "epoch": 8.405630198336532, "grad_norm": 0.8477520970345037, "learning_rate": 1.3038818570974897e-06, "loss": 0.0271, "step": 13138 }, { "epoch": 8.406269993602047, "grad_norm": 0.7038857543825484, "learning_rate": 1.3028589466689045e-06, "loss": 0.0216, "step": 13139 }, { "epoch": 8.406909788867562, "grad_norm": 0.8229522717630204, "learning_rate": 1.301836409680084e-06, "loss": 0.0263, "step": 13140 }, { "epoch": 8.407549584133077, "grad_norm": 0.7789463664286224, "learning_rate": 1.300814246174935e-06, "loss": 0.0238, "step": 13141 }, { "epoch": 8.408189379398593, "grad_norm": 0.6657025096086554, "learning_rate": 1.2997924561973485e-06, "loss": 0.0207, "step": 13142 }, { "epoch": 8.408829174664108, "grad_norm": 0.8201206923725066, "learning_rate": 1.2987710397911978e-06, "loss": 0.0236, "step": 13143 }, { "epoch": 8.409468969929623, "grad_norm": 0.7669669463066767, "learning_rate": 1.2977499970003393e-06, "loss": 0.0231, "step": 13144 }, { "epoch": 8.410108765195137, "grad_norm": 0.7616254228842401, "learning_rate": 1.2967293278686145e-06, "loss": 0.0237, "step": 13145 }, { "epoch": 8.410748560460652, "grad_norm": 0.7740463375338656, "learning_rate": 1.2957090324398514e-06, "loss": 0.0248, "step": 13146 }, { "epoch": 8.411388355726167, "grad_norm": 0.8034812251668152, "learning_rate": 1.2946891107578596e-06, "loss": 0.0241, "step": 13147 }, { "epoch": 8.412028150991683, "grad_norm": 0.8423323153149848, "learning_rate": 1.2936695628664297e-06, "loss": 0.0227, "step": 13148 }, { "epoch": 8.412667946257198, "grad_norm": 0.8632861056198926, "learning_rate": 1.2926503888093433e-06, "loss": 0.0235, "step": 13149 }, { "epoch": 8.413307741522713, "grad_norm": 0.9321573217967165, "learning_rate": 1.2916315886303599e-06, "loss": 0.0269, "step": 13150 }, { "epoch": 8.413947536788228, "grad_norm": 0.8549643592520829, "learning_rate": 1.2906131623732255e-06, "loss": 0.0245, "step": 13151 }, { "epoch": 8.414587332053744, "grad_norm": 0.8392459367895119, "learning_rate": 1.2895951100816673e-06, "loss": 0.0232, "step": 13152 }, { "epoch": 8.415227127319257, "grad_norm": 0.7009288760482849, "learning_rate": 1.2885774317994015e-06, "loss": 0.0183, "step": 13153 }, { "epoch": 8.415866922584772, "grad_norm": 0.8156515835824818, "learning_rate": 1.2875601275701254e-06, "loss": 0.0281, "step": 13154 }, { "epoch": 8.416506717850288, "grad_norm": 0.919567163102117, "learning_rate": 1.2865431974375188e-06, "loss": 0.027, "step": 13155 }, { "epoch": 8.417146513115803, "grad_norm": 0.7724654963115123, "learning_rate": 1.285526641445247e-06, "loss": 0.0258, "step": 13156 }, { "epoch": 8.417786308381318, "grad_norm": 0.8315347030989939, "learning_rate": 1.2845104596369606e-06, "loss": 0.0245, "step": 13157 }, { "epoch": 8.418426103646834, "grad_norm": 0.8063091560640547, "learning_rate": 1.2834946520562874e-06, "loss": 0.0258, "step": 13158 }, { "epoch": 8.419065898912349, "grad_norm": 0.9095564880740888, "learning_rate": 1.2824792187468516e-06, "loss": 0.0242, "step": 13159 }, { "epoch": 8.419705694177862, "grad_norm": 0.9624586455347357, "learning_rate": 1.2814641597522515e-06, "loss": 0.0284, "step": 13160 }, { "epoch": 8.420345489443378, "grad_norm": 0.8217289686059221, "learning_rate": 1.28044947511607e-06, "loss": 0.0245, "step": 13161 }, { "epoch": 8.420985284708893, "grad_norm": 0.7720369041868853, "learning_rate": 1.279435164881876e-06, "loss": 0.0241, "step": 13162 }, { "epoch": 8.421625079974408, "grad_norm": 0.7320414732624435, "learning_rate": 1.2784212290932242e-06, "loss": 0.0204, "step": 13163 }, { "epoch": 8.422264875239923, "grad_norm": 0.7742947054025103, "learning_rate": 1.2774076677936508e-06, "loss": 0.022, "step": 13164 }, { "epoch": 8.422904670505439, "grad_norm": 0.8861802306217266, "learning_rate": 1.276394481026675e-06, "loss": 0.0263, "step": 13165 }, { "epoch": 8.423544465770954, "grad_norm": 0.9965894865924873, "learning_rate": 1.2753816688358033e-06, "loss": 0.0324, "step": 13166 }, { "epoch": 8.42418426103647, "grad_norm": 0.8674208294038809, "learning_rate": 1.2743692312645228e-06, "loss": 0.0281, "step": 13167 }, { "epoch": 8.424824056301983, "grad_norm": 0.8012200523615991, "learning_rate": 1.273357168356304e-06, "loss": 0.0228, "step": 13168 }, { "epoch": 8.425463851567498, "grad_norm": 0.8416554010888511, "learning_rate": 1.2723454801546065e-06, "loss": 0.0267, "step": 13169 }, { "epoch": 8.426103646833013, "grad_norm": 0.7707844555551275, "learning_rate": 1.2713341667028689e-06, "loss": 0.0245, "step": 13170 }, { "epoch": 8.426743442098529, "grad_norm": 0.8325902330559857, "learning_rate": 1.2703232280445155e-06, "loss": 0.0255, "step": 13171 }, { "epoch": 8.427383237364044, "grad_norm": 0.7508742780139835, "learning_rate": 1.2693126642229535e-06, "loss": 0.0212, "step": 13172 }, { "epoch": 8.428023032629559, "grad_norm": 0.7169582216458946, "learning_rate": 1.2683024752815743e-06, "loss": 0.0244, "step": 13173 }, { "epoch": 8.428662827895074, "grad_norm": 0.7527410075464342, "learning_rate": 1.267292661263756e-06, "loss": 0.025, "step": 13174 }, { "epoch": 8.429302623160588, "grad_norm": 0.6920587982649267, "learning_rate": 1.266283222212853e-06, "loss": 0.02, "step": 13175 }, { "epoch": 8.429942418426103, "grad_norm": 0.7240871046748556, "learning_rate": 1.265274158172215e-06, "loss": 0.0231, "step": 13176 }, { "epoch": 8.430582213691618, "grad_norm": 0.7752481150770644, "learning_rate": 1.2642654691851664e-06, "loss": 0.0228, "step": 13177 }, { "epoch": 8.431222008957134, "grad_norm": 0.7284444325840049, "learning_rate": 1.263257155295018e-06, "loss": 0.0214, "step": 13178 }, { "epoch": 8.431861804222649, "grad_norm": 0.7822318841156459, "learning_rate": 1.262249216545066e-06, "loss": 0.024, "step": 13179 }, { "epoch": 8.432501599488164, "grad_norm": 0.7790241993831877, "learning_rate": 1.261241652978591e-06, "loss": 0.0234, "step": 13180 }, { "epoch": 8.43314139475368, "grad_norm": 0.7719988128318579, "learning_rate": 1.2602344646388542e-06, "loss": 0.0232, "step": 13181 }, { "epoch": 8.433781190019193, "grad_norm": 0.7933831009843847, "learning_rate": 1.2592276515691005e-06, "loss": 0.027, "step": 13182 }, { "epoch": 8.434420985284708, "grad_norm": 0.8647436613975967, "learning_rate": 1.2582212138125638e-06, "loss": 0.0251, "step": 13183 }, { "epoch": 8.435060780550224, "grad_norm": 0.8845204092458614, "learning_rate": 1.2572151514124587e-06, "loss": 0.0264, "step": 13184 }, { "epoch": 8.435700575815739, "grad_norm": 0.8594774297449714, "learning_rate": 1.2562094644119826e-06, "loss": 0.0269, "step": 13185 }, { "epoch": 8.436340371081254, "grad_norm": 0.753986168560006, "learning_rate": 1.255204152854319e-06, "loss": 0.0194, "step": 13186 }, { "epoch": 8.43698016634677, "grad_norm": 0.8186845001638677, "learning_rate": 1.2541992167826323e-06, "loss": 0.0255, "step": 13187 }, { "epoch": 8.437619961612285, "grad_norm": 0.8702865996121162, "learning_rate": 1.253194656240072e-06, "loss": 0.0236, "step": 13188 }, { "epoch": 8.4382597568778, "grad_norm": 0.754392886241205, "learning_rate": 1.2521904712697752e-06, "loss": 0.0236, "step": 13189 }, { "epoch": 8.438899552143313, "grad_norm": 0.9151334991997855, "learning_rate": 1.2511866619148593e-06, "loss": 0.0254, "step": 13190 }, { "epoch": 8.439539347408829, "grad_norm": 0.7734791059764524, "learning_rate": 1.2501832282184244e-06, "loss": 0.0223, "step": 13191 }, { "epoch": 8.440179142674344, "grad_norm": 0.8493673833694955, "learning_rate": 1.2491801702235552e-06, "loss": 0.0262, "step": 13192 }, { "epoch": 8.44081893793986, "grad_norm": 0.7849706531476202, "learning_rate": 1.2481774879733255e-06, "loss": 0.0262, "step": 13193 }, { "epoch": 8.441458733205375, "grad_norm": 0.7607457943990967, "learning_rate": 1.2471751815107858e-06, "loss": 0.0226, "step": 13194 }, { "epoch": 8.44209852847089, "grad_norm": 0.7142121303337052, "learning_rate": 1.2461732508789714e-06, "loss": 0.0203, "step": 13195 }, { "epoch": 8.442738323736405, "grad_norm": 0.8346705896539917, "learning_rate": 1.2451716961209082e-06, "loss": 0.0266, "step": 13196 }, { "epoch": 8.443378119001919, "grad_norm": 0.9172244310427664, "learning_rate": 1.2441705172795981e-06, "loss": 0.0255, "step": 13197 }, { "epoch": 8.444017914267434, "grad_norm": 0.8696861333305278, "learning_rate": 1.2431697143980314e-06, "loss": 0.029, "step": 13198 }, { "epoch": 8.44465770953295, "grad_norm": 0.9794260599896846, "learning_rate": 1.2421692875191783e-06, "loss": 0.033, "step": 13199 }, { "epoch": 8.445297504798464, "grad_norm": 0.7982661270639493, "learning_rate": 1.2411692366859972e-06, "loss": 0.0218, "step": 13200 }, { "epoch": 8.44593730006398, "grad_norm": 0.71766226671838, "learning_rate": 1.24016956194143e-06, "loss": 0.0225, "step": 13201 }, { "epoch": 8.446577095329495, "grad_norm": 0.7681268790674455, "learning_rate": 1.2391702633283986e-06, "loss": 0.0231, "step": 13202 }, { "epoch": 8.44721689059501, "grad_norm": 0.7632620152616612, "learning_rate": 1.2381713408898122e-06, "loss": 0.0226, "step": 13203 }, { "epoch": 8.447856685860526, "grad_norm": 0.792274790656705, "learning_rate": 1.2371727946685618e-06, "loss": 0.0239, "step": 13204 }, { "epoch": 8.448496481126039, "grad_norm": 0.7938278231420886, "learning_rate": 1.2361746247075224e-06, "loss": 0.0242, "step": 13205 }, { "epoch": 8.449136276391554, "grad_norm": 0.6987943039821736, "learning_rate": 1.2351768310495571e-06, "loss": 0.024, "step": 13206 }, { "epoch": 8.44977607165707, "grad_norm": 0.8036262031583485, "learning_rate": 1.2341794137375062e-06, "loss": 0.0223, "step": 13207 }, { "epoch": 8.450415866922585, "grad_norm": 0.7653435472794925, "learning_rate": 1.2331823728141989e-06, "loss": 0.0217, "step": 13208 }, { "epoch": 8.4510556621881, "grad_norm": 0.8572698755300008, "learning_rate": 1.232185708322443e-06, "loss": 0.027, "step": 13209 }, { "epoch": 8.451695457453615, "grad_norm": 0.89548328617587, "learning_rate": 1.231189420305039e-06, "loss": 0.0248, "step": 13210 }, { "epoch": 8.45233525271913, "grad_norm": 0.7857544966147155, "learning_rate": 1.2301935088047622e-06, "loss": 0.0221, "step": 13211 }, { "epoch": 8.452975047984644, "grad_norm": 0.8167612265396199, "learning_rate": 1.2291979738643734e-06, "loss": 0.0214, "step": 13212 }, { "epoch": 8.45361484325016, "grad_norm": 0.8364087919126653, "learning_rate": 1.2282028155266246e-06, "loss": 0.0283, "step": 13213 }, { "epoch": 8.454254638515675, "grad_norm": 0.6515006944811544, "learning_rate": 1.2272080338342418e-06, "loss": 0.0205, "step": 13214 }, { "epoch": 8.45489443378119, "grad_norm": 0.7075694863587401, "learning_rate": 1.226213628829942e-06, "loss": 0.0211, "step": 13215 }, { "epoch": 8.455534229046705, "grad_norm": 0.7618119202270603, "learning_rate": 1.2252196005564198e-06, "loss": 0.0248, "step": 13216 }, { "epoch": 8.45617402431222, "grad_norm": 0.8221665406189843, "learning_rate": 1.2242259490563602e-06, "loss": 0.0249, "step": 13217 }, { "epoch": 8.456813819577736, "grad_norm": 0.7806179410754562, "learning_rate": 1.2232326743724276e-06, "loss": 0.0264, "step": 13218 }, { "epoch": 8.45745361484325, "grad_norm": 0.7092335414849984, "learning_rate": 1.2222397765472682e-06, "loss": 0.0201, "step": 13219 }, { "epoch": 8.458093410108765, "grad_norm": 0.8845948079118114, "learning_rate": 1.2212472556235211e-06, "loss": 0.0259, "step": 13220 }, { "epoch": 8.45873320537428, "grad_norm": 0.8984064950942394, "learning_rate": 1.2202551116438e-06, "loss": 0.0256, "step": 13221 }, { "epoch": 8.459373000639795, "grad_norm": 0.7438786324752823, "learning_rate": 1.2192633446507052e-06, "loss": 0.0225, "step": 13222 }, { "epoch": 8.46001279590531, "grad_norm": 0.811814503604319, "learning_rate": 1.2182719546868249e-06, "loss": 0.0242, "step": 13223 }, { "epoch": 8.460652591170826, "grad_norm": 0.8266773257404723, "learning_rate": 1.217280941794724e-06, "loss": 0.0221, "step": 13224 }, { "epoch": 8.461292386436341, "grad_norm": 0.8391003484387779, "learning_rate": 1.2162903060169552e-06, "loss": 0.0257, "step": 13225 }, { "epoch": 8.461932181701856, "grad_norm": 0.9351054968993354, "learning_rate": 1.2153000473960575e-06, "loss": 0.0304, "step": 13226 }, { "epoch": 8.46257197696737, "grad_norm": 0.9186122750904735, "learning_rate": 1.2143101659745494e-06, "loss": 0.025, "step": 13227 }, { "epoch": 8.463211772232885, "grad_norm": 0.7731600661188093, "learning_rate": 1.2133206617949334e-06, "loss": 0.0185, "step": 13228 }, { "epoch": 8.4638515674984, "grad_norm": 0.8520459683889059, "learning_rate": 1.2123315348996966e-06, "loss": 0.0242, "step": 13229 }, { "epoch": 8.464491362763916, "grad_norm": 0.8305360240269868, "learning_rate": 1.2113427853313132e-06, "loss": 0.0267, "step": 13230 }, { "epoch": 8.46513115802943, "grad_norm": 0.8541640069929746, "learning_rate": 1.2103544131322375e-06, "loss": 0.0235, "step": 13231 }, { "epoch": 8.465770953294946, "grad_norm": 0.8186255931888544, "learning_rate": 1.2093664183449071e-06, "loss": 0.0211, "step": 13232 }, { "epoch": 8.466410748560461, "grad_norm": 0.8082499454719133, "learning_rate": 1.2083788010117458e-06, "loss": 0.025, "step": 13233 }, { "epoch": 8.467050543825975, "grad_norm": 0.8699948570983108, "learning_rate": 1.2073915611751586e-06, "loss": 0.0246, "step": 13234 }, { "epoch": 8.46769033909149, "grad_norm": 0.8303360681909269, "learning_rate": 1.2064046988775358e-06, "loss": 0.0238, "step": 13235 }, { "epoch": 8.468330134357005, "grad_norm": 0.9188593889020571, "learning_rate": 1.2054182141612547e-06, "loss": 0.0268, "step": 13236 }, { "epoch": 8.46896992962252, "grad_norm": 0.754011456948085, "learning_rate": 1.2044321070686715e-06, "loss": 0.0234, "step": 13237 }, { "epoch": 8.469609724888036, "grad_norm": 0.8381780213469996, "learning_rate": 1.2034463776421268e-06, "loss": 0.0261, "step": 13238 }, { "epoch": 8.470249520153551, "grad_norm": 0.7674329092167795, "learning_rate": 1.2024610259239445e-06, "loss": 0.0274, "step": 13239 }, { "epoch": 8.470889315419067, "grad_norm": 0.929241222700526, "learning_rate": 1.2014760519564394e-06, "loss": 0.028, "step": 13240 }, { "epoch": 8.471529110684582, "grad_norm": 0.9137119887383335, "learning_rate": 1.2004914557819004e-06, "loss": 0.0291, "step": 13241 }, { "epoch": 8.472168905950095, "grad_norm": 0.7510888994927868, "learning_rate": 1.199507237442603e-06, "loss": 0.0251, "step": 13242 }, { "epoch": 8.47280870121561, "grad_norm": 0.7323833730233842, "learning_rate": 1.1985233969808118e-06, "loss": 0.0235, "step": 13243 }, { "epoch": 8.473448496481126, "grad_norm": 0.7315510795766557, "learning_rate": 1.1975399344387707e-06, "loss": 0.0223, "step": 13244 }, { "epoch": 8.474088291746641, "grad_norm": 0.7436796202323791, "learning_rate": 1.1965568498587055e-06, "loss": 0.0228, "step": 13245 }, { "epoch": 8.474728087012156, "grad_norm": 0.714812619507915, "learning_rate": 1.1955741432828283e-06, "loss": 0.0235, "step": 13246 }, { "epoch": 8.475367882277672, "grad_norm": 0.7671520083416365, "learning_rate": 1.1945918147533364e-06, "loss": 0.0261, "step": 13247 }, { "epoch": 8.476007677543187, "grad_norm": 0.8031503248334334, "learning_rate": 1.1936098643124083e-06, "loss": 0.0256, "step": 13248 }, { "epoch": 8.4766474728087, "grad_norm": 0.7509783505588165, "learning_rate": 1.1926282920022058e-06, "loss": 0.0233, "step": 13249 }, { "epoch": 8.477287268074216, "grad_norm": 0.7404763220441378, "learning_rate": 1.1916470978648787e-06, "loss": 0.0251, "step": 13250 }, { "epoch": 8.477927063339731, "grad_norm": 0.7533985596464234, "learning_rate": 1.1906662819425563e-06, "loss": 0.0263, "step": 13251 }, { "epoch": 8.478566858605246, "grad_norm": 0.8711108013285058, "learning_rate": 1.1896858442773517e-06, "loss": 0.0235, "step": 13252 }, { "epoch": 8.479206653870762, "grad_norm": 0.7571000450184391, "learning_rate": 1.188705784911366e-06, "loss": 0.0263, "step": 13253 }, { "epoch": 8.479846449136277, "grad_norm": 0.8056631501722357, "learning_rate": 1.1877261038866794e-06, "loss": 0.0269, "step": 13254 }, { "epoch": 8.480486244401792, "grad_norm": 0.7043021550220112, "learning_rate": 1.1867468012453575e-06, "loss": 0.0236, "step": 13255 }, { "epoch": 8.481126039667306, "grad_norm": 0.735289837059371, "learning_rate": 1.1857678770294511e-06, "loss": 0.0242, "step": 13256 }, { "epoch": 8.48176583493282, "grad_norm": 0.7449124479226846, "learning_rate": 1.1847893312809932e-06, "loss": 0.023, "step": 13257 }, { "epoch": 8.482405630198336, "grad_norm": 0.8236171030503754, "learning_rate": 1.1838111640420014e-06, "loss": 0.0258, "step": 13258 }, { "epoch": 8.483045425463851, "grad_norm": 0.78772484477235, "learning_rate": 1.1828333753544718e-06, "loss": 0.0258, "step": 13259 }, { "epoch": 8.483685220729367, "grad_norm": 0.8367108139065098, "learning_rate": 1.1818559652603956e-06, "loss": 0.028, "step": 13260 }, { "epoch": 8.484325015994882, "grad_norm": 0.7987643826019296, "learning_rate": 1.180878933801738e-06, "loss": 0.0249, "step": 13261 }, { "epoch": 8.484964811260397, "grad_norm": 0.8114747748991202, "learning_rate": 1.1799022810204497e-06, "loss": 0.0206, "step": 13262 }, { "epoch": 8.485604606525913, "grad_norm": 0.7864869389882201, "learning_rate": 1.1789260069584696e-06, "loss": 0.0228, "step": 13263 }, { "epoch": 8.486244401791426, "grad_norm": 0.8112708193260368, "learning_rate": 1.177950111657714e-06, "loss": 0.0278, "step": 13264 }, { "epoch": 8.486884197056941, "grad_norm": 0.8212676565031163, "learning_rate": 1.1769745951600852e-06, "loss": 0.027, "step": 13265 }, { "epoch": 8.487523992322457, "grad_norm": 0.7674270612956424, "learning_rate": 1.175999457507475e-06, "loss": 0.0233, "step": 13266 }, { "epoch": 8.488163787587972, "grad_norm": 0.8384267108902262, "learning_rate": 1.1750246987417512e-06, "loss": 0.0251, "step": 13267 }, { "epoch": 8.488803582853487, "grad_norm": 0.8830129814976465, "learning_rate": 1.1740503189047691e-06, "loss": 0.0278, "step": 13268 }, { "epoch": 8.489443378119002, "grad_norm": 0.8259434020799787, "learning_rate": 1.173076318038363e-06, "loss": 0.0249, "step": 13269 }, { "epoch": 8.490083173384518, "grad_norm": 0.7898870730022377, "learning_rate": 1.1721026961843608e-06, "loss": 0.0248, "step": 13270 }, { "epoch": 8.490722968650031, "grad_norm": 0.7984238273507829, "learning_rate": 1.1711294533845651e-06, "loss": 0.0255, "step": 13271 }, { "epoch": 8.491362763915546, "grad_norm": 0.9466933011911416, "learning_rate": 1.1701565896807631e-06, "loss": 0.0291, "step": 13272 }, { "epoch": 8.492002559181062, "grad_norm": 0.8615486539328767, "learning_rate": 1.1691841051147313e-06, "loss": 0.0276, "step": 13273 }, { "epoch": 8.492642354446577, "grad_norm": 0.7926903328056555, "learning_rate": 1.1682119997282259e-06, "loss": 0.0266, "step": 13274 }, { "epoch": 8.493282149712092, "grad_norm": 0.7906237884858621, "learning_rate": 1.1672402735629873e-06, "loss": 0.0244, "step": 13275 }, { "epoch": 8.493921944977608, "grad_norm": 0.7970192922032208, "learning_rate": 1.1662689266607385e-06, "loss": 0.0247, "step": 13276 }, { "epoch": 8.494561740243123, "grad_norm": 0.7106152020883392, "learning_rate": 1.1652979590631884e-06, "loss": 0.0223, "step": 13277 }, { "epoch": 8.495201535508638, "grad_norm": 0.7364813105952339, "learning_rate": 1.1643273708120274e-06, "loss": 0.0225, "step": 13278 }, { "epoch": 8.495841330774152, "grad_norm": 0.7154365598642439, "learning_rate": 1.1633571619489303e-06, "loss": 0.0239, "step": 13279 }, { "epoch": 8.496481126039667, "grad_norm": 0.7210394898822915, "learning_rate": 1.1623873325155598e-06, "loss": 0.0226, "step": 13280 }, { "epoch": 8.497120921305182, "grad_norm": 0.7902475153541123, "learning_rate": 1.161417882553556e-06, "loss": 0.0255, "step": 13281 }, { "epoch": 8.497760716570697, "grad_norm": 0.8149004645857837, "learning_rate": 1.1604488121045431e-06, "loss": 0.0235, "step": 13282 }, { "epoch": 8.498400511836213, "grad_norm": 0.8808682746636111, "learning_rate": 1.1594801212101369e-06, "loss": 0.0253, "step": 13283 }, { "epoch": 8.499040307101728, "grad_norm": 0.834148020881927, "learning_rate": 1.1585118099119275e-06, "loss": 0.0233, "step": 13284 }, { "epoch": 8.499680102367243, "grad_norm": 0.7840639304300625, "learning_rate": 1.1575438782514937e-06, "loss": 0.0218, "step": 13285 }, { "epoch": 8.500319897632757, "grad_norm": 0.7555221946295476, "learning_rate": 1.1565763262703932e-06, "loss": 0.0235, "step": 13286 }, { "epoch": 8.500959692898272, "grad_norm": 0.7802916285064602, "learning_rate": 1.1556091540101777e-06, "loss": 0.025, "step": 13287 }, { "epoch": 8.501599488163787, "grad_norm": 0.7172359821648134, "learning_rate": 1.1546423615123715e-06, "loss": 0.0219, "step": 13288 }, { "epoch": 8.502239283429303, "grad_norm": 0.8684273347638922, "learning_rate": 1.1536759488184846e-06, "loss": 0.021, "step": 13289 }, { "epoch": 8.502879078694818, "grad_norm": 0.8463013815565136, "learning_rate": 1.1527099159700194e-06, "loss": 0.0243, "step": 13290 }, { "epoch": 8.503518873960333, "grad_norm": 0.778797205178478, "learning_rate": 1.151744263008452e-06, "loss": 0.0233, "step": 13291 }, { "epoch": 8.504158669225848, "grad_norm": 0.8237897908067059, "learning_rate": 1.1507789899752453e-06, "loss": 0.0271, "step": 13292 }, { "epoch": 8.504798464491362, "grad_norm": 0.8637547234167013, "learning_rate": 1.1498140969118476e-06, "loss": 0.026, "step": 13293 }, { "epoch": 8.505438259756877, "grad_norm": 0.8968893649457428, "learning_rate": 1.1488495838596903e-06, "loss": 0.0237, "step": 13294 }, { "epoch": 8.506078055022392, "grad_norm": 0.818713919691081, "learning_rate": 1.1478854508601855e-06, "loss": 0.0225, "step": 13295 }, { "epoch": 8.506717850287908, "grad_norm": 0.6475153580557094, "learning_rate": 1.1469216979547315e-06, "loss": 0.018, "step": 13296 }, { "epoch": 8.507357645553423, "grad_norm": 0.8263568389910146, "learning_rate": 1.145958325184713e-06, "loss": 0.0215, "step": 13297 }, { "epoch": 8.507997440818938, "grad_norm": 0.8086953276700302, "learning_rate": 1.1449953325914942e-06, "loss": 0.0214, "step": 13298 }, { "epoch": 8.508637236084454, "grad_norm": 0.8116249407763672, "learning_rate": 1.1440327202164225e-06, "loss": 0.022, "step": 13299 }, { "epoch": 8.509277031349969, "grad_norm": 0.7810961270720826, "learning_rate": 1.1430704881008336e-06, "loss": 0.0256, "step": 13300 }, { "epoch": 8.509916826615482, "grad_norm": 0.848363415047422, "learning_rate": 1.1421086362860422e-06, "loss": 0.0246, "step": 13301 }, { "epoch": 8.510556621880998, "grad_norm": 0.7130855567738387, "learning_rate": 1.1411471648133477e-06, "loss": 0.0228, "step": 13302 }, { "epoch": 8.511196417146513, "grad_norm": 0.8782864831704138, "learning_rate": 1.1401860737240377e-06, "loss": 0.0254, "step": 13303 }, { "epoch": 8.511836212412028, "grad_norm": 0.8679664388471866, "learning_rate": 1.1392253630593764e-06, "loss": 0.0263, "step": 13304 }, { "epoch": 8.512476007677543, "grad_norm": 0.7816591430245772, "learning_rate": 1.138265032860616e-06, "loss": 0.0242, "step": 13305 }, { "epoch": 8.513115802943059, "grad_norm": 0.8606678173878174, "learning_rate": 1.1373050831689913e-06, "loss": 0.0282, "step": 13306 }, { "epoch": 8.513755598208574, "grad_norm": 0.8172172365268546, "learning_rate": 1.1363455140257218e-06, "loss": 0.0244, "step": 13307 }, { "epoch": 8.514395393474087, "grad_norm": 0.9416613193199641, "learning_rate": 1.1353863254720076e-06, "loss": 0.0244, "step": 13308 }, { "epoch": 8.515035188739603, "grad_norm": 0.8049057000019937, "learning_rate": 1.1344275175490337e-06, "loss": 0.0243, "step": 13309 }, { "epoch": 8.515674984005118, "grad_norm": 0.938361766740681, "learning_rate": 1.133469090297975e-06, "loss": 0.0279, "step": 13310 }, { "epoch": 8.516314779270633, "grad_norm": 0.7940558189625866, "learning_rate": 1.1325110437599797e-06, "loss": 0.0233, "step": 13311 }, { "epoch": 8.516954574536149, "grad_norm": 0.8380123774104313, "learning_rate": 1.1315533779761846e-06, "loss": 0.026, "step": 13312 }, { "epoch": 8.517594369801664, "grad_norm": 0.8194849829109373, "learning_rate": 1.1305960929877135e-06, "loss": 0.0224, "step": 13313 }, { "epoch": 8.51823416506718, "grad_norm": 0.8366781428332102, "learning_rate": 1.1296391888356694e-06, "loss": 0.0239, "step": 13314 }, { "epoch": 8.518873960332694, "grad_norm": 0.8330497554338949, "learning_rate": 1.1286826655611384e-06, "loss": 0.025, "step": 13315 }, { "epoch": 8.519513755598208, "grad_norm": 0.883408235115065, "learning_rate": 1.1277265232051914e-06, "loss": 0.0261, "step": 13316 }, { "epoch": 8.520153550863723, "grad_norm": 0.7806462606275317, "learning_rate": 1.1267707618088875e-06, "loss": 0.0269, "step": 13317 }, { "epoch": 8.520793346129238, "grad_norm": 0.8012789429010901, "learning_rate": 1.1258153814132622e-06, "loss": 0.0277, "step": 13318 }, { "epoch": 8.521433141394754, "grad_norm": 0.7703363625688247, "learning_rate": 1.1248603820593362e-06, "loss": 0.0229, "step": 13319 }, { "epoch": 8.522072936660269, "grad_norm": 0.7157930887121423, "learning_rate": 1.12390576378812e-06, "loss": 0.0216, "step": 13320 }, { "epoch": 8.522712731925784, "grad_norm": 0.8012290432329314, "learning_rate": 1.122951526640601e-06, "loss": 0.0265, "step": 13321 }, { "epoch": 8.5233525271913, "grad_norm": 0.7641755076432573, "learning_rate": 1.1219976706577529e-06, "loss": 0.0217, "step": 13322 }, { "epoch": 8.523992322456813, "grad_norm": 0.8049723029430567, "learning_rate": 1.1210441958805306e-06, "loss": 0.0253, "step": 13323 }, { "epoch": 8.524632117722328, "grad_norm": 0.8063914578698794, "learning_rate": 1.1200911023498773e-06, "loss": 0.0247, "step": 13324 }, { "epoch": 8.525271912987844, "grad_norm": 0.7834137701181907, "learning_rate": 1.1191383901067166e-06, "loss": 0.025, "step": 13325 }, { "epoch": 8.525911708253359, "grad_norm": 0.884771448778323, "learning_rate": 1.1181860591919524e-06, "loss": 0.0248, "step": 13326 }, { "epoch": 8.526551503518874, "grad_norm": 0.8241895088105345, "learning_rate": 1.1172341096464812e-06, "loss": 0.0263, "step": 13327 }, { "epoch": 8.52719129878439, "grad_norm": 0.8284039924814324, "learning_rate": 1.1162825415111766e-06, "loss": 0.0225, "step": 13328 }, { "epoch": 8.527831094049905, "grad_norm": 0.7715169728596095, "learning_rate": 1.1153313548268951e-06, "loss": 0.0267, "step": 13329 }, { "epoch": 8.528470889315418, "grad_norm": 0.7756141492089567, "learning_rate": 1.1143805496344818e-06, "loss": 0.0232, "step": 13330 }, { "epoch": 8.529110684580933, "grad_norm": 0.7955770834520086, "learning_rate": 1.1134301259747627e-06, "loss": 0.0238, "step": 13331 }, { "epoch": 8.529750479846449, "grad_norm": 0.842161380305858, "learning_rate": 1.112480083888543e-06, "loss": 0.0243, "step": 13332 }, { "epoch": 8.530390275111964, "grad_norm": 0.9193973622156276, "learning_rate": 1.1115304234166201e-06, "loss": 0.0307, "step": 13333 }, { "epoch": 8.53103007037748, "grad_norm": 0.7506843770180337, "learning_rate": 1.1105811445997705e-06, "loss": 0.0209, "step": 13334 }, { "epoch": 8.531669865642995, "grad_norm": 0.7828902139343451, "learning_rate": 1.1096322474787535e-06, "loss": 0.0233, "step": 13335 }, { "epoch": 8.53230966090851, "grad_norm": 0.8164859742905272, "learning_rate": 1.1086837320943134e-06, "loss": 0.0257, "step": 13336 }, { "epoch": 8.532949456174025, "grad_norm": 0.8685425396040424, "learning_rate": 1.1077355984871763e-06, "loss": 0.025, "step": 13337 }, { "epoch": 8.533589251439539, "grad_norm": 0.8344662889829625, "learning_rate": 1.1067878466980552e-06, "loss": 0.025, "step": 13338 }, { "epoch": 8.534229046705054, "grad_norm": 0.7642073257291029, "learning_rate": 1.1058404767676422e-06, "loss": 0.0234, "step": 13339 }, { "epoch": 8.53486884197057, "grad_norm": 0.670682796987222, "learning_rate": 1.1048934887366191e-06, "loss": 0.0189, "step": 13340 }, { "epoch": 8.535508637236084, "grad_norm": 0.8092229064232989, "learning_rate": 1.1039468826456468e-06, "loss": 0.0229, "step": 13341 }, { "epoch": 8.5361484325016, "grad_norm": 0.731881422190977, "learning_rate": 1.1030006585353692e-06, "loss": 0.0204, "step": 13342 }, { "epoch": 8.536788227767115, "grad_norm": 0.8716744636153546, "learning_rate": 1.1020548164464183e-06, "loss": 0.0302, "step": 13343 }, { "epoch": 8.53742802303263, "grad_norm": 1.3473581383080822, "learning_rate": 1.1011093564194054e-06, "loss": 0.0247, "step": 13344 }, { "epoch": 8.538067818298144, "grad_norm": 0.8028614532809589, "learning_rate": 1.1001642784949263e-06, "loss": 0.0251, "step": 13345 }, { "epoch": 8.538707613563659, "grad_norm": 0.8505291997755303, "learning_rate": 1.09921958271356e-06, "loss": 0.0251, "step": 13346 }, { "epoch": 8.539347408829174, "grad_norm": 0.790761019474027, "learning_rate": 1.0982752691158727e-06, "loss": 0.0257, "step": 13347 }, { "epoch": 8.53998720409469, "grad_norm": 0.861691596896587, "learning_rate": 1.0973313377424112e-06, "loss": 0.0245, "step": 13348 }, { "epoch": 8.540626999360205, "grad_norm": 0.7836458169811302, "learning_rate": 1.0963877886337038e-06, "loss": 0.0241, "step": 13349 }, { "epoch": 8.54126679462572, "grad_norm": 0.8787803463084144, "learning_rate": 1.0954446218302684e-06, "loss": 0.0264, "step": 13350 }, { "epoch": 8.541906589891235, "grad_norm": 0.7801183477655385, "learning_rate": 1.0945018373726001e-06, "loss": 0.0238, "step": 13351 }, { "epoch": 8.54254638515675, "grad_norm": 0.78724880459303, "learning_rate": 1.0935594353011813e-06, "loss": 0.0236, "step": 13352 }, { "epoch": 8.543186180422264, "grad_norm": 0.8794443908755939, "learning_rate": 1.0926174156564773e-06, "loss": 0.029, "step": 13353 }, { "epoch": 8.54382597568778, "grad_norm": 0.7565094395223335, "learning_rate": 1.0916757784789367e-06, "loss": 0.0239, "step": 13354 }, { "epoch": 8.544465770953295, "grad_norm": 0.9592482654633673, "learning_rate": 1.0907345238089916e-06, "loss": 0.0293, "step": 13355 }, { "epoch": 8.54510556621881, "grad_norm": 0.7914264550803074, "learning_rate": 1.0897936516870556e-06, "loss": 0.0252, "step": 13356 }, { "epoch": 8.545745361484325, "grad_norm": 0.7702295892455348, "learning_rate": 1.0888531621535326e-06, "loss": 0.0211, "step": 13357 }, { "epoch": 8.54638515674984, "grad_norm": 0.7856923617394501, "learning_rate": 1.0879130552488016e-06, "loss": 0.0223, "step": 13358 }, { "epoch": 8.547024952015356, "grad_norm": 0.805117732790405, "learning_rate": 1.0869733310132303e-06, "loss": 0.0257, "step": 13359 }, { "epoch": 8.54766474728087, "grad_norm": 0.8763170519887074, "learning_rate": 1.0860339894871697e-06, "loss": 0.0265, "step": 13360 }, { "epoch": 8.548304542546385, "grad_norm": 0.8473180375275714, "learning_rate": 1.0850950307109543e-06, "loss": 0.0235, "step": 13361 }, { "epoch": 8.5489443378119, "grad_norm": 0.8338042468490601, "learning_rate": 1.0841564547248994e-06, "loss": 0.0255, "step": 13362 }, { "epoch": 8.549584133077415, "grad_norm": 0.7711940423851282, "learning_rate": 1.0832182615693033e-06, "loss": 0.0232, "step": 13363 }, { "epoch": 8.55022392834293, "grad_norm": 0.8432675979264402, "learning_rate": 1.0822804512844564e-06, "loss": 0.0263, "step": 13364 }, { "epoch": 8.550863723608446, "grad_norm": 0.7544534501879462, "learning_rate": 1.0813430239106238e-06, "loss": 0.022, "step": 13365 }, { "epoch": 8.551503518873961, "grad_norm": 0.7285848824469762, "learning_rate": 1.0804059794880562e-06, "loss": 0.0211, "step": 13366 }, { "epoch": 8.552143314139474, "grad_norm": 0.9305065048233586, "learning_rate": 1.0794693180569882e-06, "loss": 0.0265, "step": 13367 }, { "epoch": 8.55278310940499, "grad_norm": 0.8387896959876924, "learning_rate": 1.0785330396576398e-06, "loss": 0.0236, "step": 13368 }, { "epoch": 8.553422904670505, "grad_norm": 0.8420450660385304, "learning_rate": 1.0775971443302114e-06, "loss": 0.0234, "step": 13369 }, { "epoch": 8.55406269993602, "grad_norm": 0.8851000333528197, "learning_rate": 1.0766616321148915e-06, "loss": 0.0271, "step": 13370 }, { "epoch": 8.554702495201536, "grad_norm": 0.8403534083312535, "learning_rate": 1.0757265030518482e-06, "loss": 0.0236, "step": 13371 }, { "epoch": 8.55534229046705, "grad_norm": 0.7745889825389582, "learning_rate": 1.0747917571812328e-06, "loss": 0.0206, "step": 13372 }, { "epoch": 8.555982085732566, "grad_norm": 0.8631729612620511, "learning_rate": 1.0738573945431807e-06, "loss": 0.0201, "step": 13373 }, { "epoch": 8.556621880998081, "grad_norm": 0.8876034873604666, "learning_rate": 1.0729234151778168e-06, "loss": 0.0262, "step": 13374 }, { "epoch": 8.557261676263595, "grad_norm": 0.8725105082881263, "learning_rate": 1.071989819125241e-06, "loss": 0.0246, "step": 13375 }, { "epoch": 8.55790147152911, "grad_norm": 0.8328832334150559, "learning_rate": 1.0710566064255378e-06, "loss": 0.0244, "step": 13376 }, { "epoch": 8.558541266794625, "grad_norm": 0.8557580357025107, "learning_rate": 1.0701237771187833e-06, "loss": 0.0263, "step": 13377 }, { "epoch": 8.55918106206014, "grad_norm": 0.8681711647717125, "learning_rate": 1.0691913312450287e-06, "loss": 0.0264, "step": 13378 }, { "epoch": 8.559820857325656, "grad_norm": 0.7710459742851191, "learning_rate": 1.0682592688443094e-06, "loss": 0.0221, "step": 13379 }, { "epoch": 8.560460652591171, "grad_norm": 0.783176046346793, "learning_rate": 1.0673275899566515e-06, "loss": 0.0246, "step": 13380 }, { "epoch": 8.561100447856687, "grad_norm": 0.8555833833361604, "learning_rate": 1.0663962946220563e-06, "loss": 0.0242, "step": 13381 }, { "epoch": 8.561740243122202, "grad_norm": 0.6826816258056921, "learning_rate": 1.0654653828805128e-06, "loss": 0.0189, "step": 13382 }, { "epoch": 8.562380038387715, "grad_norm": 0.903262197786476, "learning_rate": 1.0645348547719925e-06, "loss": 0.0274, "step": 13383 }, { "epoch": 8.56301983365323, "grad_norm": 0.8341898851526963, "learning_rate": 1.06360471033645e-06, "loss": 0.0255, "step": 13384 }, { "epoch": 8.563659628918746, "grad_norm": 0.8420224414427511, "learning_rate": 1.0626749496138245e-06, "loss": 0.0228, "step": 13385 }, { "epoch": 8.564299424184261, "grad_norm": 0.817999443251572, "learning_rate": 1.0617455726440373e-06, "loss": 0.0278, "step": 13386 }, { "epoch": 8.564939219449776, "grad_norm": 0.9475738180293106, "learning_rate": 1.0608165794669968e-06, "loss": 0.0299, "step": 13387 }, { "epoch": 8.565579014715292, "grad_norm": 0.8150926039802843, "learning_rate": 1.0598879701225906e-06, "loss": 0.0242, "step": 13388 }, { "epoch": 8.566218809980807, "grad_norm": 0.8405122276449479, "learning_rate": 1.0589597446506905e-06, "loss": 0.0243, "step": 13389 }, { "epoch": 8.56685860524632, "grad_norm": 0.9221906110666546, "learning_rate": 1.0580319030911556e-06, "loss": 0.027, "step": 13390 }, { "epoch": 8.567498400511836, "grad_norm": 0.9417306693644952, "learning_rate": 1.057104445483824e-06, "loss": 0.028, "step": 13391 }, { "epoch": 8.568138195777351, "grad_norm": 0.7843632421420469, "learning_rate": 1.0561773718685198e-06, "loss": 0.0249, "step": 13392 }, { "epoch": 8.568777991042866, "grad_norm": 0.776118429626554, "learning_rate": 1.0552506822850472e-06, "loss": 0.0229, "step": 13393 }, { "epoch": 8.569417786308382, "grad_norm": 0.8147080566120269, "learning_rate": 1.0543243767732004e-06, "loss": 0.0245, "step": 13394 }, { "epoch": 8.570057581573897, "grad_norm": 0.9275539477718184, "learning_rate": 1.0533984553727527e-06, "loss": 0.0272, "step": 13395 }, { "epoch": 8.570697376839412, "grad_norm": 0.860844288645518, "learning_rate": 1.0524729181234594e-06, "loss": 0.0227, "step": 13396 }, { "epoch": 8.571337172104926, "grad_norm": 0.8176610547179536, "learning_rate": 1.0515477650650642e-06, "loss": 0.024, "step": 13397 }, { "epoch": 8.571976967370441, "grad_norm": 0.8550428013648419, "learning_rate": 1.0506229962372882e-06, "loss": 0.0262, "step": 13398 }, { "epoch": 8.572616762635956, "grad_norm": 0.7657341212620766, "learning_rate": 1.0496986116798403e-06, "loss": 0.0238, "step": 13399 }, { "epoch": 8.573256557901471, "grad_norm": 0.663564022708387, "learning_rate": 1.0487746114324139e-06, "loss": 0.0195, "step": 13400 }, { "epoch": 8.573896353166987, "grad_norm": 0.8356274447310885, "learning_rate": 1.0478509955346827e-06, "loss": 0.0251, "step": 13401 }, { "epoch": 8.574536148432502, "grad_norm": 0.7302758432324375, "learning_rate": 1.0469277640263055e-06, "loss": 0.0257, "step": 13402 }, { "epoch": 8.575175943698017, "grad_norm": 0.8051342550142713, "learning_rate": 1.046004916946921e-06, "loss": 0.0264, "step": 13403 }, { "epoch": 8.57581573896353, "grad_norm": 0.7917768475755684, "learning_rate": 1.0450824543361603e-06, "loss": 0.0224, "step": 13404 }, { "epoch": 8.576455534229046, "grad_norm": 0.824376841079337, "learning_rate": 1.0441603762336294e-06, "loss": 0.0222, "step": 13405 }, { "epoch": 8.577095329494561, "grad_norm": 0.8642350878971264, "learning_rate": 1.0432386826789187e-06, "loss": 0.026, "step": 13406 }, { "epoch": 8.577735124760077, "grad_norm": 0.7595436464220992, "learning_rate": 1.0423173737116076e-06, "loss": 0.0236, "step": 13407 }, { "epoch": 8.578374920025592, "grad_norm": 0.8737868718590306, "learning_rate": 1.0413964493712548e-06, "loss": 0.0295, "step": 13408 }, { "epoch": 8.579014715291107, "grad_norm": 0.9028192951801868, "learning_rate": 1.0404759096974003e-06, "loss": 0.028, "step": 13409 }, { "epoch": 8.579654510556622, "grad_norm": 0.7424509934069773, "learning_rate": 1.0395557547295742e-06, "loss": 0.0234, "step": 13410 }, { "epoch": 8.580294305822138, "grad_norm": 0.7455370374741779, "learning_rate": 1.0386359845072846e-06, "loss": 0.024, "step": 13411 }, { "epoch": 8.580934101087651, "grad_norm": 0.795391017548527, "learning_rate": 1.0377165990700244e-06, "loss": 0.0236, "step": 13412 }, { "epoch": 8.581573896353166, "grad_norm": 0.7544763696116054, "learning_rate": 1.0367975984572709e-06, "loss": 0.0237, "step": 13413 }, { "epoch": 8.582213691618682, "grad_norm": 0.8229954962462718, "learning_rate": 1.0358789827084847e-06, "loss": 0.0236, "step": 13414 }, { "epoch": 8.582853486884197, "grad_norm": 0.634925072500718, "learning_rate": 1.0349607518631089e-06, "loss": 0.02, "step": 13415 }, { "epoch": 8.583493282149712, "grad_norm": 0.8634945359495911, "learning_rate": 1.0340429059605683e-06, "loss": 0.0273, "step": 13416 }, { "epoch": 8.584133077415228, "grad_norm": 0.9322012167477479, "learning_rate": 1.0331254450402783e-06, "loss": 0.0265, "step": 13417 }, { "epoch": 8.584772872680743, "grad_norm": 0.8982222941138108, "learning_rate": 1.0322083691416318e-06, "loss": 0.0291, "step": 13418 }, { "epoch": 8.585412667946258, "grad_norm": 0.787909291763036, "learning_rate": 1.0312916783040027e-06, "loss": 0.0246, "step": 13419 }, { "epoch": 8.586052463211772, "grad_norm": 0.7969885924443717, "learning_rate": 1.0303753725667565e-06, "loss": 0.0229, "step": 13420 }, { "epoch": 8.586692258477287, "grad_norm": 0.894652087624297, "learning_rate": 1.029459451969237e-06, "loss": 0.0262, "step": 13421 }, { "epoch": 8.587332053742802, "grad_norm": 0.8087203743597655, "learning_rate": 1.0285439165507705e-06, "loss": 0.0195, "step": 13422 }, { "epoch": 8.587971849008317, "grad_norm": 0.7525190761993975, "learning_rate": 1.027628766350668e-06, "loss": 0.022, "step": 13423 }, { "epoch": 8.588611644273833, "grad_norm": 0.745403358713505, "learning_rate": 1.0267140014082266e-06, "loss": 0.0208, "step": 13424 }, { "epoch": 8.589251439539348, "grad_norm": 0.9653394005976725, "learning_rate": 1.0257996217627241e-06, "loss": 0.0279, "step": 13425 }, { "epoch": 8.589891234804863, "grad_norm": 0.9100872502913725, "learning_rate": 1.0248856274534224e-06, "loss": 0.0284, "step": 13426 }, { "epoch": 8.590531030070377, "grad_norm": 0.8785256361473344, "learning_rate": 1.0239720185195668e-06, "loss": 0.0232, "step": 13427 }, { "epoch": 8.591170825335892, "grad_norm": 0.7301806545797891, "learning_rate": 1.0230587950003845e-06, "loss": 0.0206, "step": 13428 }, { "epoch": 8.591810620601407, "grad_norm": 0.777304242925706, "learning_rate": 1.0221459569350867e-06, "loss": 0.022, "step": 13429 }, { "epoch": 8.592450415866923, "grad_norm": 0.7975585039577023, "learning_rate": 1.021233504362874e-06, "loss": 0.0239, "step": 13430 }, { "epoch": 8.593090211132438, "grad_norm": 0.7691514084358072, "learning_rate": 1.020321437322922e-06, "loss": 0.0249, "step": 13431 }, { "epoch": 8.593730006397953, "grad_norm": 0.8012830936526233, "learning_rate": 1.0194097558543948e-06, "loss": 0.0215, "step": 13432 }, { "epoch": 8.594369801663468, "grad_norm": 0.7486376966473842, "learning_rate": 1.0184984599964342e-06, "loss": 0.0196, "step": 13433 }, { "epoch": 8.595009596928982, "grad_norm": 0.8019272878066338, "learning_rate": 1.017587549788176e-06, "loss": 0.0245, "step": 13434 }, { "epoch": 8.595649392194497, "grad_norm": 0.9038587942582088, "learning_rate": 1.0166770252687296e-06, "loss": 0.0273, "step": 13435 }, { "epoch": 8.596289187460012, "grad_norm": 0.9044427383812009, "learning_rate": 1.0157668864771896e-06, "loss": 0.0228, "step": 13436 }, { "epoch": 8.596928982725528, "grad_norm": 0.8410745832457689, "learning_rate": 1.0148571334526392e-06, "loss": 0.0239, "step": 13437 }, { "epoch": 8.597568777991043, "grad_norm": 0.7265158999509008, "learning_rate": 1.0139477662341413e-06, "loss": 0.0249, "step": 13438 }, { "epoch": 8.598208573256558, "grad_norm": 0.7754399676708438, "learning_rate": 1.0130387848607404e-06, "loss": 0.0216, "step": 13439 }, { "epoch": 8.598848368522074, "grad_norm": 3.988232625379426, "learning_rate": 1.0121301893714662e-06, "loss": 0.0291, "step": 13440 }, { "epoch": 8.599488163787587, "grad_norm": 0.7275793505506142, "learning_rate": 1.0112219798053347e-06, "loss": 0.0205, "step": 13441 }, { "epoch": 8.600127959053102, "grad_norm": 0.7814064780378112, "learning_rate": 1.0103141562013418e-06, "loss": 0.0249, "step": 13442 }, { "epoch": 8.600767754318618, "grad_norm": 0.8813339224232744, "learning_rate": 1.009406718598468e-06, "loss": 0.0258, "step": 13443 }, { "epoch": 8.601407549584133, "grad_norm": 0.8260373445489919, "learning_rate": 1.0084996670356772e-06, "loss": 0.0276, "step": 13444 }, { "epoch": 8.602047344849648, "grad_norm": 0.848169794539754, "learning_rate": 1.0075930015519153e-06, "loss": 0.0275, "step": 13445 }, { "epoch": 8.602687140115163, "grad_norm": 0.7184134311660367, "learning_rate": 1.0066867221861109e-06, "loss": 0.0227, "step": 13446 }, { "epoch": 8.603326935380679, "grad_norm": 0.8657802353518579, "learning_rate": 1.0057808289771843e-06, "loss": 0.0265, "step": 13447 }, { "epoch": 8.603966730646194, "grad_norm": 0.8016900975850862, "learning_rate": 1.0048753219640284e-06, "loss": 0.0242, "step": 13448 }, { "epoch": 8.604606525911707, "grad_norm": 0.7397734633146754, "learning_rate": 1.0039702011855257e-06, "loss": 0.0233, "step": 13449 }, { "epoch": 8.605246321177223, "grad_norm": 0.7299226045731485, "learning_rate": 1.0030654666805373e-06, "loss": 0.0205, "step": 13450 }, { "epoch": 8.605886116442738, "grad_norm": 0.8827856460094823, "learning_rate": 1.0021611184879165e-06, "loss": 0.0279, "step": 13451 }, { "epoch": 8.606525911708253, "grad_norm": 0.8702949391125857, "learning_rate": 1.0012571566464901e-06, "loss": 0.0258, "step": 13452 }, { "epoch": 8.607165706973769, "grad_norm": 0.8580965095072015, "learning_rate": 1.0003535811950726e-06, "loss": 0.026, "step": 13453 }, { "epoch": 8.607805502239284, "grad_norm": 0.7564446652368723, "learning_rate": 9.994503921724652e-07, "loss": 0.0214, "step": 13454 }, { "epoch": 8.6084452975048, "grad_norm": 0.7945151649820212, "learning_rate": 9.985475896174468e-07, "loss": 0.0256, "step": 13455 }, { "epoch": 8.609085092770314, "grad_norm": 0.6999587393180828, "learning_rate": 9.97645173568783e-07, "loss": 0.0212, "step": 13456 }, { "epoch": 8.609724888035828, "grad_norm": 0.7934751425270875, "learning_rate": 9.967431440652208e-07, "loss": 0.026, "step": 13457 }, { "epoch": 8.610364683301343, "grad_norm": 0.7486814385746681, "learning_rate": 9.958415011454935e-07, "loss": 0.0216, "step": 13458 }, { "epoch": 8.611004478566858, "grad_norm": 0.8316443619825692, "learning_rate": 9.949402448483137e-07, "loss": 0.0245, "step": 13459 }, { "epoch": 8.611644273832374, "grad_norm": 0.9182038469982233, "learning_rate": 9.9403937521238e-07, "loss": 0.027, "step": 13460 }, { "epoch": 8.612284069097889, "grad_norm": 0.8970645123941624, "learning_rate": 9.931388922763774e-07, "loss": 0.0282, "step": 13461 }, { "epoch": 8.612923864363404, "grad_norm": 0.8124363367941292, "learning_rate": 9.922387960789682e-07, "loss": 0.0263, "step": 13462 }, { "epoch": 8.61356365962892, "grad_norm": 0.8927426814124537, "learning_rate": 9.91339086658799e-07, "loss": 0.0256, "step": 13463 }, { "epoch": 8.614203454894433, "grad_norm": 0.7832003579026429, "learning_rate": 9.904397640545071e-07, "loss": 0.0239, "step": 13464 }, { "epoch": 8.614843250159948, "grad_norm": 0.7590460234374925, "learning_rate": 9.895408283047047e-07, "loss": 0.0236, "step": 13465 }, { "epoch": 8.615483045425464, "grad_norm": 0.804787822115558, "learning_rate": 9.886422794479888e-07, "loss": 0.023, "step": 13466 }, { "epoch": 8.616122840690979, "grad_norm": 0.675860070734852, "learning_rate": 9.877441175229462e-07, "loss": 0.0214, "step": 13467 }, { "epoch": 8.616762635956494, "grad_norm": 0.886461426365789, "learning_rate": 9.868463425681384e-07, "loss": 0.0265, "step": 13468 }, { "epoch": 8.61740243122201, "grad_norm": 0.8379482833999093, "learning_rate": 9.859489546221157e-07, "loss": 0.0259, "step": 13469 }, { "epoch": 8.618042226487525, "grad_norm": 0.7785620028464517, "learning_rate": 9.850519537234093e-07, "loss": 0.0212, "step": 13470 }, { "epoch": 8.618682021753038, "grad_norm": 0.8776448859074508, "learning_rate": 9.841553399105364e-07, "loss": 0.0274, "step": 13471 }, { "epoch": 8.619321817018553, "grad_norm": 0.7848855504635468, "learning_rate": 9.832591132219959e-07, "loss": 0.0248, "step": 13472 }, { "epoch": 8.619961612284069, "grad_norm": 0.8634218242424726, "learning_rate": 9.823632736962695e-07, "loss": 0.0249, "step": 13473 }, { "epoch": 8.620601407549584, "grad_norm": 0.7559193368549755, "learning_rate": 9.814678213718222e-07, "loss": 0.0217, "step": 13474 }, { "epoch": 8.6212412028151, "grad_norm": 0.8146778663167766, "learning_rate": 9.805727562871048e-07, "loss": 0.0245, "step": 13475 }, { "epoch": 8.621880998080615, "grad_norm": 0.7895120402131384, "learning_rate": 9.79678078480547e-07, "loss": 0.0225, "step": 13476 }, { "epoch": 8.62252079334613, "grad_norm": 0.7457245004384164, "learning_rate": 9.787837879905682e-07, "loss": 0.0187, "step": 13477 }, { "epoch": 8.623160588611643, "grad_norm": 0.7054868232301267, "learning_rate": 9.778898848555663e-07, "loss": 0.0195, "step": 13478 }, { "epoch": 8.623800383877159, "grad_norm": 1.0534176816169527, "learning_rate": 9.769963691139227e-07, "loss": 0.0268, "step": 13479 }, { "epoch": 8.624440179142674, "grad_norm": 0.761916926442536, "learning_rate": 9.761032408040038e-07, "loss": 0.0228, "step": 13480 }, { "epoch": 8.62507997440819, "grad_norm": 0.7795989312389721, "learning_rate": 9.75210499964161e-07, "loss": 0.0251, "step": 13481 }, { "epoch": 8.625719769673704, "grad_norm": 0.7127259422820049, "learning_rate": 9.743181466327244e-07, "loss": 0.0219, "step": 13482 }, { "epoch": 8.62635956493922, "grad_norm": 0.7612408173020648, "learning_rate": 9.734261808480106e-07, "loss": 0.0229, "step": 13483 }, { "epoch": 8.626999360204735, "grad_norm": 0.8342771248395627, "learning_rate": 9.725346026483207e-07, "loss": 0.0243, "step": 13484 }, { "epoch": 8.62763915547025, "grad_norm": 0.8558692622032168, "learning_rate": 9.716434120719365e-07, "loss": 0.0266, "step": 13485 }, { "epoch": 8.628278950735764, "grad_norm": 0.7644173290475561, "learning_rate": 9.70752609157124e-07, "loss": 0.0245, "step": 13486 }, { "epoch": 8.628918746001279, "grad_norm": 0.752645809356769, "learning_rate": 9.698621939421327e-07, "loss": 0.0223, "step": 13487 }, { "epoch": 8.629558541266794, "grad_norm": 0.8042162816761931, "learning_rate": 9.689721664651951e-07, "loss": 0.0247, "step": 13488 }, { "epoch": 8.63019833653231, "grad_norm": 0.7774175076728175, "learning_rate": 9.680825267645288e-07, "loss": 0.0242, "step": 13489 }, { "epoch": 8.630838131797825, "grad_norm": 0.8435658114265369, "learning_rate": 9.671932748783297e-07, "loss": 0.0248, "step": 13490 }, { "epoch": 8.63147792706334, "grad_norm": 0.9154346484883122, "learning_rate": 9.66304410844785e-07, "loss": 0.0271, "step": 13491 }, { "epoch": 8.632117722328855, "grad_norm": 0.7695839382100975, "learning_rate": 9.654159347020598e-07, "loss": 0.0225, "step": 13492 }, { "epoch": 8.63275751759437, "grad_norm": 0.8082568528664392, "learning_rate": 9.645278464883001e-07, "loss": 0.024, "step": 13493 }, { "epoch": 8.633397312859884, "grad_norm": 0.8863465942369942, "learning_rate": 9.636401462416444e-07, "loss": 0.025, "step": 13494 }, { "epoch": 8.6340371081254, "grad_norm": 0.9650786853480751, "learning_rate": 9.627528340002057e-07, "loss": 0.0274, "step": 13495 }, { "epoch": 8.634676903390915, "grad_norm": 0.8723370705661081, "learning_rate": 9.618659098020832e-07, "loss": 0.0271, "step": 13496 }, { "epoch": 8.63531669865643, "grad_norm": 0.8231561645714595, "learning_rate": 9.609793736853634e-07, "loss": 0.0232, "step": 13497 }, { "epoch": 8.635956493921945, "grad_norm": 0.7688780797591456, "learning_rate": 9.60093225688109e-07, "loss": 0.0215, "step": 13498 }, { "epoch": 8.63659628918746, "grad_norm": 0.8848470293892948, "learning_rate": 9.592074658483708e-07, "loss": 0.0244, "step": 13499 }, { "epoch": 8.637236084452976, "grad_norm": 0.8662700606711625, "learning_rate": 9.583220942041804e-07, "loss": 0.0225, "step": 13500 }, { "epoch": 8.63787587971849, "grad_norm": 0.7655640814659886, "learning_rate": 9.574371107935564e-07, "loss": 0.0248, "step": 13501 }, { "epoch": 8.638515674984005, "grad_norm": 0.7069096892962763, "learning_rate": 9.565525156544986e-07, "loss": 0.0215, "step": 13502 }, { "epoch": 8.63915547024952, "grad_norm": 0.7982067987726161, "learning_rate": 9.556683088249875e-07, "loss": 0.0227, "step": 13503 }, { "epoch": 8.639795265515035, "grad_norm": 0.8250326441930815, "learning_rate": 9.547844903429914e-07, "loss": 0.0233, "step": 13504 }, { "epoch": 8.64043506078055, "grad_norm": 0.9323493798858117, "learning_rate": 9.539010602464594e-07, "loss": 0.0251, "step": 13505 }, { "epoch": 8.641074856046066, "grad_norm": 0.7436343282057208, "learning_rate": 9.530180185733217e-07, "loss": 0.0225, "step": 13506 }, { "epoch": 8.641714651311581, "grad_norm": 0.7700109284726122, "learning_rate": 9.521353653614984e-07, "loss": 0.021, "step": 13507 }, { "epoch": 8.642354446577095, "grad_norm": 0.7473360193534392, "learning_rate": 9.512531006488878e-07, "loss": 0.0228, "step": 13508 }, { "epoch": 8.64299424184261, "grad_norm": 0.7847540791285326, "learning_rate": 9.503712244733732e-07, "loss": 0.025, "step": 13509 }, { "epoch": 8.643634037108125, "grad_norm": 0.8767550957755408, "learning_rate": 9.494897368728184e-07, "loss": 0.0251, "step": 13510 }, { "epoch": 8.64427383237364, "grad_norm": 0.7684189753880717, "learning_rate": 9.486086378850756e-07, "loss": 0.0207, "step": 13511 }, { "epoch": 8.644913627639156, "grad_norm": 0.7528777746551952, "learning_rate": 9.477279275479767e-07, "loss": 0.0199, "step": 13512 }, { "epoch": 8.64555342290467, "grad_norm": 0.8566494590740917, "learning_rate": 9.46847605899337e-07, "loss": 0.0245, "step": 13513 }, { "epoch": 8.646193218170186, "grad_norm": 0.8393317600879848, "learning_rate": 9.459676729769584e-07, "loss": 0.0244, "step": 13514 }, { "epoch": 8.6468330134357, "grad_norm": 0.9326258551842761, "learning_rate": 9.45088128818622e-07, "loss": 0.0253, "step": 13515 }, { "epoch": 8.647472808701215, "grad_norm": 0.70123581575295, "learning_rate": 9.442089734620929e-07, "loss": 0.0206, "step": 13516 }, { "epoch": 8.64811260396673, "grad_norm": 0.7783200314435517, "learning_rate": 9.433302069451222e-07, "loss": 0.024, "step": 13517 }, { "epoch": 8.648752399232245, "grad_norm": 0.7818833226786636, "learning_rate": 9.42451829305443e-07, "loss": 0.0218, "step": 13518 }, { "epoch": 8.64939219449776, "grad_norm": 0.7835339225156184, "learning_rate": 9.415738405807684e-07, "loss": 0.0247, "step": 13519 }, { "epoch": 8.650031989763276, "grad_norm": 0.7469566963820724, "learning_rate": 9.406962408087983e-07, "loss": 0.0211, "step": 13520 }, { "epoch": 8.650671785028791, "grad_norm": 0.7621886331424849, "learning_rate": 9.398190300272181e-07, "loss": 0.0236, "step": 13521 }, { "epoch": 8.651311580294307, "grad_norm": 0.8837447305494018, "learning_rate": 9.38942208273692e-07, "loss": 0.0297, "step": 13522 }, { "epoch": 8.65195137555982, "grad_norm": 0.7910133090209351, "learning_rate": 9.380657755858669e-07, "loss": 0.0233, "step": 13523 }, { "epoch": 8.652591170825335, "grad_norm": 0.8201617271841113, "learning_rate": 9.3718973200138e-07, "loss": 0.0259, "step": 13524 }, { "epoch": 8.65323096609085, "grad_norm": 0.7846578485441491, "learning_rate": 9.363140775578439e-07, "loss": 0.0208, "step": 13525 }, { "epoch": 8.653870761356366, "grad_norm": 0.7028637981961271, "learning_rate": 9.354388122928582e-07, "loss": 0.0232, "step": 13526 }, { "epoch": 8.654510556621881, "grad_norm": 0.7434458937258568, "learning_rate": 9.34563936244004e-07, "loss": 0.0206, "step": 13527 }, { "epoch": 8.655150351887396, "grad_norm": 0.714743298137938, "learning_rate": 9.336894494488502e-07, "loss": 0.0197, "step": 13528 }, { "epoch": 8.655790147152912, "grad_norm": 0.7714530038981702, "learning_rate": 9.328153519449435e-07, "loss": 0.0202, "step": 13529 }, { "epoch": 8.656429942418427, "grad_norm": 0.7463989572672808, "learning_rate": 9.319416437698137e-07, "loss": 0.0234, "step": 13530 }, { "epoch": 8.65706973768394, "grad_norm": 0.8871230481384375, "learning_rate": 9.31068324960982e-07, "loss": 0.0259, "step": 13531 }, { "epoch": 8.657709532949456, "grad_norm": 0.8201770356501047, "learning_rate": 9.301953955559428e-07, "loss": 0.0258, "step": 13532 }, { "epoch": 8.658349328214971, "grad_norm": 0.8304517563408421, "learning_rate": 9.293228555921807e-07, "loss": 0.028, "step": 13533 }, { "epoch": 8.658989123480486, "grad_norm": 0.9272877548083499, "learning_rate": 9.28450705107159e-07, "loss": 0.027, "step": 13534 }, { "epoch": 8.659628918746002, "grad_norm": 0.8155557917668274, "learning_rate": 9.275789441383265e-07, "loss": 0.0228, "step": 13535 }, { "epoch": 8.660268714011517, "grad_norm": 0.7947173797555445, "learning_rate": 9.267075727231156e-07, "loss": 0.0252, "step": 13536 }, { "epoch": 8.660908509277032, "grad_norm": 0.8190050144753134, "learning_rate": 9.258365908989409e-07, "loss": 0.0245, "step": 13537 }, { "epoch": 8.661548304542546, "grad_norm": 0.7067306996837918, "learning_rate": 9.249659987032022e-07, "loss": 0.0205, "step": 13538 }, { "epoch": 8.662188099808061, "grad_norm": 0.8797197589642041, "learning_rate": 9.240957961732799e-07, "loss": 0.0259, "step": 13539 }, { "epoch": 8.662827895073576, "grad_norm": 0.7196846651729659, "learning_rate": 9.232259833465385e-07, "loss": 0.0216, "step": 13540 }, { "epoch": 8.663467690339091, "grad_norm": 0.7455319026762743, "learning_rate": 9.22356560260328e-07, "loss": 0.0212, "step": 13541 }, { "epoch": 8.664107485604607, "grad_norm": 0.8762568126318556, "learning_rate": 9.214875269519785e-07, "loss": 0.0259, "step": 13542 }, { "epoch": 8.664747280870122, "grad_norm": 0.7903373711608175, "learning_rate": 9.206188834588048e-07, "loss": 0.023, "step": 13543 }, { "epoch": 8.665387076135637, "grad_norm": 0.797443688488099, "learning_rate": 9.197506298181058e-07, "loss": 0.0278, "step": 13544 }, { "epoch": 8.66602687140115, "grad_norm": 0.7571693045912813, "learning_rate": 9.188827660671629e-07, "loss": 0.0213, "step": 13545 }, { "epoch": 8.666666666666666, "grad_norm": 0.7884751403313423, "learning_rate": 9.180152922432395e-07, "loss": 0.0246, "step": 13546 }, { "epoch": 8.667306461932181, "grad_norm": 0.7288635000260556, "learning_rate": 9.171482083835847e-07, "loss": 0.021, "step": 13547 }, { "epoch": 8.667946257197697, "grad_norm": 0.8553447823640136, "learning_rate": 9.162815145254278e-07, "loss": 0.0214, "step": 13548 }, { "epoch": 8.668586052463212, "grad_norm": 0.8269427686119771, "learning_rate": 9.154152107059843e-07, "loss": 0.025, "step": 13549 }, { "epoch": 8.669225847728727, "grad_norm": 0.7235089703540917, "learning_rate": 9.145492969624503e-07, "loss": 0.0186, "step": 13550 }, { "epoch": 8.669865642994242, "grad_norm": 0.8229264695311432, "learning_rate": 9.136837733320092e-07, "loss": 0.0229, "step": 13551 }, { "epoch": 8.670505438259756, "grad_norm": 0.843776863790174, "learning_rate": 9.128186398518235e-07, "loss": 0.0242, "step": 13552 }, { "epoch": 8.671145233525271, "grad_norm": 0.8830943295527942, "learning_rate": 9.11953896559038e-07, "loss": 0.0265, "step": 13553 }, { "epoch": 8.671785028790786, "grad_norm": 0.815799804064278, "learning_rate": 9.110895434907885e-07, "loss": 0.0244, "step": 13554 }, { "epoch": 8.672424824056302, "grad_norm": 0.7409462798532142, "learning_rate": 9.102255806841853e-07, "loss": 0.0234, "step": 13555 }, { "epoch": 8.673064619321817, "grad_norm": 0.8118794661119454, "learning_rate": 9.093620081763266e-07, "loss": 0.0216, "step": 13556 }, { "epoch": 8.673704414587332, "grad_norm": 0.8245494803235645, "learning_rate": 9.084988260042904e-07, "loss": 0.0236, "step": 13557 }, { "epoch": 8.674344209852848, "grad_norm": 0.8022547400658817, "learning_rate": 9.076360342051449e-07, "loss": 0.0238, "step": 13558 }, { "epoch": 8.674984005118363, "grad_norm": 0.8227493032039124, "learning_rate": 9.067736328159327e-07, "loss": 0.0221, "step": 13559 }, { "epoch": 8.675623800383876, "grad_norm": 0.7265790358758843, "learning_rate": 9.059116218736841e-07, "loss": 0.0217, "step": 13560 }, { "epoch": 8.676263595649392, "grad_norm": 0.7907233831975834, "learning_rate": 9.05050001415415e-07, "loss": 0.0228, "step": 13561 }, { "epoch": 8.676903390914907, "grad_norm": 0.7562838845293354, "learning_rate": 9.041887714781195e-07, "loss": 0.0208, "step": 13562 }, { "epoch": 8.677543186180422, "grad_norm": 0.7259093905367484, "learning_rate": 9.033279320987786e-07, "loss": 0.022, "step": 13563 }, { "epoch": 8.678182981445937, "grad_norm": 0.8944030830229543, "learning_rate": 9.024674833143543e-07, "loss": 0.0269, "step": 13564 }, { "epoch": 8.678822776711453, "grad_norm": 0.6899856807473734, "learning_rate": 9.016074251617935e-07, "loss": 0.02, "step": 13565 }, { "epoch": 8.679462571976968, "grad_norm": 0.8749243719450542, "learning_rate": 9.007477576780244e-07, "loss": 0.0253, "step": 13566 }, { "epoch": 8.680102367242483, "grad_norm": 0.7411339086978646, "learning_rate": 8.998884808999586e-07, "loss": 0.0213, "step": 13567 }, { "epoch": 8.680742162507997, "grad_norm": 0.8394066950188525, "learning_rate": 8.990295948644956e-07, "loss": 0.0279, "step": 13568 }, { "epoch": 8.681381957773512, "grad_norm": 0.8333388450409978, "learning_rate": 8.981710996085124e-07, "loss": 0.0231, "step": 13569 }, { "epoch": 8.682021753039027, "grad_norm": 0.8114018003432854, "learning_rate": 8.973129951688697e-07, "loss": 0.0245, "step": 13570 }, { "epoch": 8.682661548304543, "grad_norm": 0.8259715637188789, "learning_rate": 8.964552815824168e-07, "loss": 0.0203, "step": 13571 }, { "epoch": 8.683301343570058, "grad_norm": 0.8503147126575941, "learning_rate": 8.955979588859786e-07, "loss": 0.0254, "step": 13572 }, { "epoch": 8.683941138835573, "grad_norm": 0.8491079198494521, "learning_rate": 8.947410271163681e-07, "loss": 0.0283, "step": 13573 }, { "epoch": 8.684580934101088, "grad_norm": 0.7651547653869636, "learning_rate": 8.938844863103824e-07, "loss": 0.0246, "step": 13574 }, { "epoch": 8.685220729366602, "grad_norm": 0.8826754787905664, "learning_rate": 8.930283365047976e-07, "loss": 0.0266, "step": 13575 }, { "epoch": 8.685860524632117, "grad_norm": 0.8142249919054687, "learning_rate": 8.921725777363765e-07, "loss": 0.0246, "step": 13576 }, { "epoch": 8.686500319897632, "grad_norm": 0.7789050912842662, "learning_rate": 8.913172100418643e-07, "loss": 0.0242, "step": 13577 }, { "epoch": 8.687140115163148, "grad_norm": 0.8877683192679271, "learning_rate": 8.904622334579871e-07, "loss": 0.0294, "step": 13578 }, { "epoch": 8.687779910428663, "grad_norm": 0.8034618277743436, "learning_rate": 8.896076480214566e-07, "loss": 0.0239, "step": 13579 }, { "epoch": 8.688419705694178, "grad_norm": 0.7666501990395937, "learning_rate": 8.887534537689668e-07, "loss": 0.022, "step": 13580 }, { "epoch": 8.689059500959694, "grad_norm": 0.7717741310572529, "learning_rate": 8.878996507371973e-07, "loss": 0.0235, "step": 13581 }, { "epoch": 8.689699296225207, "grad_norm": 0.7824482030103597, "learning_rate": 8.870462389628076e-07, "loss": 0.0207, "step": 13582 }, { "epoch": 8.690339091490722, "grad_norm": 0.8006158408191797, "learning_rate": 8.861932184824395e-07, "loss": 0.0269, "step": 13583 }, { "epoch": 8.690978886756238, "grad_norm": 0.8992097412723188, "learning_rate": 8.853405893327249e-07, "loss": 0.0258, "step": 13584 }, { "epoch": 8.691618682021753, "grad_norm": 0.7715769816636873, "learning_rate": 8.844883515502711e-07, "loss": 0.0227, "step": 13585 }, { "epoch": 8.692258477287268, "grad_norm": 0.8925652651070324, "learning_rate": 8.836365051716723e-07, "loss": 0.0265, "step": 13586 }, { "epoch": 8.692898272552783, "grad_norm": 0.850311050614299, "learning_rate": 8.827850502335034e-07, "loss": 0.0254, "step": 13587 }, { "epoch": 8.693538067818299, "grad_norm": 0.7748974982811249, "learning_rate": 8.819339867723287e-07, "loss": 0.0224, "step": 13588 }, { "epoch": 8.694177863083812, "grad_norm": 0.8279717406999241, "learning_rate": 8.810833148246877e-07, "loss": 0.0246, "step": 13589 }, { "epoch": 8.694817658349328, "grad_norm": 0.8050207418088909, "learning_rate": 8.802330344271071e-07, "loss": 0.0238, "step": 13590 }, { "epoch": 8.695457453614843, "grad_norm": 0.8639669479855853, "learning_rate": 8.793831456160984e-07, "loss": 0.0244, "step": 13591 }, { "epoch": 8.696097248880358, "grad_norm": 0.7821812371977107, "learning_rate": 8.785336484281537e-07, "loss": 0.0222, "step": 13592 }, { "epoch": 8.696737044145873, "grad_norm": 0.863241803667818, "learning_rate": 8.776845428997471e-07, "loss": 0.0236, "step": 13593 }, { "epoch": 8.697376839411389, "grad_norm": 1.0003000243663773, "learning_rate": 8.768358290673407e-07, "loss": 0.023, "step": 13594 }, { "epoch": 8.698016634676904, "grad_norm": 0.8594943473441188, "learning_rate": 8.759875069673729e-07, "loss": 0.0277, "step": 13595 }, { "epoch": 8.69865642994242, "grad_norm": 0.7221631438302774, "learning_rate": 8.751395766362725e-07, "loss": 0.0233, "step": 13596 }, { "epoch": 8.699296225207933, "grad_norm": 0.9227919364389853, "learning_rate": 8.742920381104447e-07, "loss": 0.0289, "step": 13597 }, { "epoch": 8.699936020473448, "grad_norm": 0.786262508768772, "learning_rate": 8.73444891426285e-07, "loss": 0.0232, "step": 13598 }, { "epoch": 8.700575815738963, "grad_norm": 0.6651232067628172, "learning_rate": 8.725981366201675e-07, "loss": 0.0182, "step": 13599 }, { "epoch": 8.701215611004478, "grad_norm": 0.8408682992605405, "learning_rate": 8.717517737284475e-07, "loss": 0.0227, "step": 13600 }, { "epoch": 8.701855406269994, "grad_norm": 0.8891711211208515, "learning_rate": 8.709058027874695e-07, "loss": 0.0261, "step": 13601 }, { "epoch": 8.702495201535509, "grad_norm": 0.7771507345974341, "learning_rate": 8.700602238335576e-07, "loss": 0.0229, "step": 13602 }, { "epoch": 8.703134996801024, "grad_norm": 0.8501097896468016, "learning_rate": 8.692150369030195e-07, "loss": 0.026, "step": 13603 }, { "epoch": 8.70377479206654, "grad_norm": 0.9807870400563206, "learning_rate": 8.683702420321427e-07, "loss": 0.0309, "step": 13604 }, { "epoch": 8.704414587332053, "grad_norm": 0.8753625839158783, "learning_rate": 8.67525839257205e-07, "loss": 0.0267, "step": 13605 }, { "epoch": 8.705054382597568, "grad_norm": 0.7850857733890155, "learning_rate": 8.666818286144629e-07, "loss": 0.0241, "step": 13606 }, { "epoch": 8.705694177863084, "grad_norm": 0.8818590586830813, "learning_rate": 8.658382101401564e-07, "loss": 0.0286, "step": 13607 }, { "epoch": 8.706333973128599, "grad_norm": 0.822903024889195, "learning_rate": 8.649949838705086e-07, "loss": 0.0223, "step": 13608 }, { "epoch": 8.706973768394114, "grad_norm": 0.7515937892817007, "learning_rate": 8.641521498417261e-07, "loss": 0.0234, "step": 13609 }, { "epoch": 8.70761356365963, "grad_norm": 0.8086931674206482, "learning_rate": 8.633097080899977e-07, "loss": 0.0229, "step": 13610 }, { "epoch": 8.708253358925145, "grad_norm": 0.9208235704868407, "learning_rate": 8.624676586514991e-07, "loss": 0.0257, "step": 13611 }, { "epoch": 8.708893154190658, "grad_norm": 0.7375958493565041, "learning_rate": 8.616260015623845e-07, "loss": 0.02, "step": 13612 }, { "epoch": 8.709532949456174, "grad_norm": 0.8004646479184055, "learning_rate": 8.607847368587929e-07, "loss": 0.0242, "step": 13613 }, { "epoch": 8.710172744721689, "grad_norm": 0.7718349102182397, "learning_rate": 8.599438645768465e-07, "loss": 0.0213, "step": 13614 }, { "epoch": 8.710812539987204, "grad_norm": 0.8536971551328989, "learning_rate": 8.591033847526531e-07, "loss": 0.0222, "step": 13615 }, { "epoch": 8.71145233525272, "grad_norm": 0.749898899055753, "learning_rate": 8.582632974222993e-07, "loss": 0.0229, "step": 13616 }, { "epoch": 8.712092130518235, "grad_norm": 0.7964022196322269, "learning_rate": 8.574236026218574e-07, "loss": 0.0212, "step": 13617 }, { "epoch": 8.71273192578375, "grad_norm": 0.7819421806869075, "learning_rate": 8.565843003873831e-07, "loss": 0.0251, "step": 13618 }, { "epoch": 8.713371721049263, "grad_norm": 0.7435319569454668, "learning_rate": 8.55745390754914e-07, "loss": 0.0182, "step": 13619 }, { "epoch": 8.714011516314779, "grad_norm": 0.9015198346862328, "learning_rate": 8.549068737604704e-07, "loss": 0.0242, "step": 13620 }, { "epoch": 8.714651311580294, "grad_norm": 0.7179265868430437, "learning_rate": 8.540687494400591e-07, "loss": 0.0203, "step": 13621 }, { "epoch": 8.71529110684581, "grad_norm": 0.888732244388613, "learning_rate": 8.532310178296666e-07, "loss": 0.027, "step": 13622 }, { "epoch": 8.715930902111324, "grad_norm": 0.8525948310633958, "learning_rate": 8.523936789652643e-07, "loss": 0.0255, "step": 13623 }, { "epoch": 8.71657069737684, "grad_norm": 0.86104219873979, "learning_rate": 8.515567328828033e-07, "loss": 0.0232, "step": 13624 }, { "epoch": 8.717210492642355, "grad_norm": 0.7569213563383497, "learning_rate": 8.50720179618223e-07, "loss": 0.0199, "step": 13625 }, { "epoch": 8.71785028790787, "grad_norm": 0.9604859422758069, "learning_rate": 8.498840192074431e-07, "loss": 0.0308, "step": 13626 }, { "epoch": 8.718490083173384, "grad_norm": 0.8274506423116121, "learning_rate": 8.490482516863651e-07, "loss": 0.0252, "step": 13627 }, { "epoch": 8.719129878438899, "grad_norm": 0.899841767930134, "learning_rate": 8.482128770908782e-07, "loss": 0.0225, "step": 13628 }, { "epoch": 8.719769673704414, "grad_norm": 0.816555203465925, "learning_rate": 8.473778954568512e-07, "loss": 0.0233, "step": 13629 }, { "epoch": 8.72040946896993, "grad_norm": 0.7665035536128738, "learning_rate": 8.465433068201334e-07, "loss": 0.0218, "step": 13630 }, { "epoch": 8.721049264235445, "grad_norm": 0.8177195758273905, "learning_rate": 8.45709111216566e-07, "loss": 0.024, "step": 13631 }, { "epoch": 8.72168905950096, "grad_norm": 0.8740569463601248, "learning_rate": 8.44875308681965e-07, "loss": 0.0248, "step": 13632 }, { "epoch": 8.722328854766475, "grad_norm": 0.9032389200457503, "learning_rate": 8.440418992521316e-07, "loss": 0.0269, "step": 13633 }, { "epoch": 8.722968650031989, "grad_norm": 0.811365881101227, "learning_rate": 8.432088829628515e-07, "loss": 0.0242, "step": 13634 }, { "epoch": 8.723608445297504, "grad_norm": 0.8124983659803028, "learning_rate": 8.423762598498942e-07, "loss": 0.0218, "step": 13635 }, { "epoch": 8.72424824056302, "grad_norm": 0.7856504668837188, "learning_rate": 8.415440299490096e-07, "loss": 0.0232, "step": 13636 }, { "epoch": 8.724888035828535, "grad_norm": 0.8664421405822745, "learning_rate": 8.407121932959339e-07, "loss": 0.0255, "step": 13637 }, { "epoch": 8.72552783109405, "grad_norm": 0.8998340786011183, "learning_rate": 8.398807499263828e-07, "loss": 0.0241, "step": 13638 }, { "epoch": 8.726167626359565, "grad_norm": 0.8341060260050862, "learning_rate": 8.390496998760578e-07, "loss": 0.0286, "step": 13639 }, { "epoch": 8.72680742162508, "grad_norm": 0.8183802411946343, "learning_rate": 8.382190431806414e-07, "loss": 0.025, "step": 13640 }, { "epoch": 8.727447216890596, "grad_norm": 0.9013328847903943, "learning_rate": 8.373887798758029e-07, "loss": 0.0257, "step": 13641 }, { "epoch": 8.72808701215611, "grad_norm": 0.8517542366793757, "learning_rate": 8.365589099971915e-07, "loss": 0.0234, "step": 13642 }, { "epoch": 8.728726807421625, "grad_norm": 0.7888559510229844, "learning_rate": 8.3572943358044e-07, "loss": 0.0216, "step": 13643 }, { "epoch": 8.72936660268714, "grad_norm": 0.7435567249849512, "learning_rate": 8.349003506611619e-07, "loss": 0.0211, "step": 13644 }, { "epoch": 8.730006397952655, "grad_norm": 0.8161654572182365, "learning_rate": 8.340716612749622e-07, "loss": 0.022, "step": 13645 }, { "epoch": 8.73064619321817, "grad_norm": 0.8823122266949799, "learning_rate": 8.332433654574201e-07, "loss": 0.0278, "step": 13646 }, { "epoch": 8.731285988483686, "grad_norm": 0.7685700481712563, "learning_rate": 8.324154632440995e-07, "loss": 0.0189, "step": 13647 }, { "epoch": 8.731925783749201, "grad_norm": 0.8947945292392627, "learning_rate": 8.315879546705529e-07, "loss": 0.025, "step": 13648 }, { "epoch": 8.732565579014715, "grad_norm": 0.7836033571243839, "learning_rate": 8.307608397723111e-07, "loss": 0.0249, "step": 13649 }, { "epoch": 8.73320537428023, "grad_norm": 0.7929720883936703, "learning_rate": 8.299341185848852e-07, "loss": 0.0207, "step": 13650 }, { "epoch": 8.733845169545745, "grad_norm": 0.8310590251426001, "learning_rate": 8.291077911437784e-07, "loss": 0.0236, "step": 13651 }, { "epoch": 8.73448496481126, "grad_norm": 0.8159792965662173, "learning_rate": 8.282818574844687e-07, "loss": 0.0225, "step": 13652 }, { "epoch": 8.735124760076776, "grad_norm": 0.7067727471489968, "learning_rate": 8.274563176424211e-07, "loss": 0.0203, "step": 13653 }, { "epoch": 8.735764555342291, "grad_norm": 0.874632134085581, "learning_rate": 8.266311716530828e-07, "loss": 0.025, "step": 13654 }, { "epoch": 8.736404350607806, "grad_norm": 0.8383633452448581, "learning_rate": 8.258064195518834e-07, "loss": 0.025, "step": 13655 }, { "epoch": 8.73704414587332, "grad_norm": 0.8670127735090443, "learning_rate": 8.249820613742365e-07, "loss": 0.0263, "step": 13656 }, { "epoch": 8.737683941138835, "grad_norm": 0.8309144119280677, "learning_rate": 8.241580971555374e-07, "loss": 0.0248, "step": 13657 }, { "epoch": 8.73832373640435, "grad_norm": 0.7946426972606322, "learning_rate": 8.233345269311687e-07, "loss": 0.0246, "step": 13658 }, { "epoch": 8.738963531669866, "grad_norm": 0.8323510312194251, "learning_rate": 8.225113507364913e-07, "loss": 0.0273, "step": 13659 }, { "epoch": 8.73960332693538, "grad_norm": 0.8577669740300924, "learning_rate": 8.216885686068499e-07, "loss": 0.0234, "step": 13660 }, { "epoch": 8.740243122200896, "grad_norm": 0.8922743607516376, "learning_rate": 8.208661805775752e-07, "loss": 0.0295, "step": 13661 }, { "epoch": 8.740882917466411, "grad_norm": 0.807022955791013, "learning_rate": 8.200441866839792e-07, "loss": 0.0199, "step": 13662 }, { "epoch": 8.741522712731927, "grad_norm": 0.7619592791387219, "learning_rate": 8.192225869613557e-07, "loss": 0.0224, "step": 13663 }, { "epoch": 8.74216250799744, "grad_norm": 0.7730120916345173, "learning_rate": 8.184013814449809e-07, "loss": 0.0207, "step": 13664 }, { "epoch": 8.742802303262955, "grad_norm": 0.7250352071690834, "learning_rate": 8.1758057017012e-07, "loss": 0.0188, "step": 13665 }, { "epoch": 8.74344209852847, "grad_norm": 0.7418854668548264, "learning_rate": 8.167601531720149e-07, "loss": 0.0199, "step": 13666 }, { "epoch": 8.744081893793986, "grad_norm": 0.7734356621635051, "learning_rate": 8.159401304858939e-07, "loss": 0.0204, "step": 13667 }, { "epoch": 8.744721689059501, "grad_norm": 1.0502979577382312, "learning_rate": 8.151205021469655e-07, "loss": 0.0274, "step": 13668 }, { "epoch": 8.745361484325016, "grad_norm": 0.8456702025711748, "learning_rate": 8.143012681904238e-07, "loss": 0.0253, "step": 13669 }, { "epoch": 8.746001279590532, "grad_norm": 0.8154617399661882, "learning_rate": 8.134824286514464e-07, "loss": 0.0252, "step": 13670 }, { "epoch": 8.746641074856045, "grad_norm": 0.7768744106986714, "learning_rate": 8.126639835651906e-07, "loss": 0.0247, "step": 13671 }, { "epoch": 8.74728087012156, "grad_norm": 0.8028038153233211, "learning_rate": 8.118459329668016e-07, "loss": 0.0229, "step": 13672 }, { "epoch": 8.747920665387076, "grad_norm": 0.877497618579669, "learning_rate": 8.110282768914035e-07, "loss": 0.0242, "step": 13673 }, { "epoch": 8.748560460652591, "grad_norm": 0.9072565858336991, "learning_rate": 8.102110153741039e-07, "loss": 0.0212, "step": 13674 }, { "epoch": 8.749200255918106, "grad_norm": 0.7089347996828592, "learning_rate": 8.093941484499979e-07, "loss": 0.0212, "step": 13675 }, { "epoch": 8.749840051183622, "grad_norm": 0.8243040418781689, "learning_rate": 8.085776761541574e-07, "loss": 0.0261, "step": 13676 }, { "epoch": 8.750479846449137, "grad_norm": 0.7566010470601142, "learning_rate": 8.077615985216403e-07, "loss": 0.0256, "step": 13677 }, { "epoch": 8.751119641714652, "grad_norm": 0.7474678360007246, "learning_rate": 8.069459155874904e-07, "loss": 0.0224, "step": 13678 }, { "epoch": 8.751759436980166, "grad_norm": 0.8532277073485532, "learning_rate": 8.061306273867287e-07, "loss": 0.0259, "step": 13679 }, { "epoch": 8.752399232245681, "grad_norm": 0.6852577804933098, "learning_rate": 8.053157339543638e-07, "loss": 0.0218, "step": 13680 }, { "epoch": 8.753039027511196, "grad_norm": 0.8364488425373964, "learning_rate": 8.045012353253834e-07, "loss": 0.0277, "step": 13681 }, { "epoch": 8.753678822776711, "grad_norm": 0.8655461149811999, "learning_rate": 8.036871315347638e-07, "loss": 0.025, "step": 13682 }, { "epoch": 8.754318618042227, "grad_norm": 0.7842265060800262, "learning_rate": 8.028734226174606e-07, "loss": 0.0221, "step": 13683 }, { "epoch": 8.754958413307742, "grad_norm": 0.7915098684377787, "learning_rate": 8.020601086084124e-07, "loss": 0.0251, "step": 13684 }, { "epoch": 8.755598208573257, "grad_norm": 0.8316649789549397, "learning_rate": 8.0124718954254e-07, "loss": 0.0267, "step": 13685 }, { "epoch": 8.75623800383877, "grad_norm": 0.8020696275823511, "learning_rate": 8.004346654547512e-07, "loss": 0.0252, "step": 13686 }, { "epoch": 8.756877799104286, "grad_norm": 0.788475194145921, "learning_rate": 7.996225363799303e-07, "loss": 0.0216, "step": 13687 }, { "epoch": 8.757517594369801, "grad_norm": 0.7914543526216626, "learning_rate": 7.98810802352954e-07, "loss": 0.0231, "step": 13688 }, { "epoch": 8.758157389635317, "grad_norm": 0.8460448086982777, "learning_rate": 7.979994634086741e-07, "loss": 0.0218, "step": 13689 }, { "epoch": 8.758797184900832, "grad_norm": 0.7288746960182203, "learning_rate": 7.971885195819284e-07, "loss": 0.0217, "step": 13690 }, { "epoch": 8.759436980166347, "grad_norm": 0.6650727101205122, "learning_rate": 7.963779709075347e-07, "loss": 0.0204, "step": 13691 }, { "epoch": 8.760076775431862, "grad_norm": 0.9436724623807352, "learning_rate": 7.955678174203008e-07, "loss": 0.0267, "step": 13692 }, { "epoch": 8.760716570697376, "grad_norm": 0.8060701653364758, "learning_rate": 7.947580591550119e-07, "loss": 0.022, "step": 13693 }, { "epoch": 8.761356365962891, "grad_norm": 0.7940743181071627, "learning_rate": 7.939486961464349e-07, "loss": 0.0219, "step": 13694 }, { "epoch": 8.761996161228407, "grad_norm": 0.8486469506062562, "learning_rate": 7.931397284293263e-07, "loss": 0.0239, "step": 13695 }, { "epoch": 8.762635956493922, "grad_norm": 0.7277815732764124, "learning_rate": 7.923311560384206e-07, "loss": 0.0212, "step": 13696 }, { "epoch": 8.763275751759437, "grad_norm": 0.8690719998345955, "learning_rate": 7.915229790084356e-07, "loss": 0.0261, "step": 13697 }, { "epoch": 8.763915547024952, "grad_norm": 0.8422132931738742, "learning_rate": 7.907151973740734e-07, "loss": 0.0238, "step": 13698 }, { "epoch": 8.764555342290468, "grad_norm": 0.7842754614322454, "learning_rate": 7.899078111700176e-07, "loss": 0.0235, "step": 13699 }, { "epoch": 8.765195137555983, "grad_norm": 0.8136968527631837, "learning_rate": 7.891008204309381e-07, "loss": 0.0237, "step": 13700 }, { "epoch": 8.765834932821496, "grad_norm": 0.6988968755355415, "learning_rate": 7.882942251914827e-07, "loss": 0.0198, "step": 13701 }, { "epoch": 8.766474728087012, "grad_norm": 0.86810292128307, "learning_rate": 7.874880254862882e-07, "loss": 0.028, "step": 13702 }, { "epoch": 8.767114523352527, "grad_norm": 0.8955230880825303, "learning_rate": 7.866822213499703e-07, "loss": 0.0324, "step": 13703 }, { "epoch": 8.767754318618042, "grad_norm": 0.6612113440767836, "learning_rate": 7.858768128171268e-07, "loss": 0.0191, "step": 13704 }, { "epoch": 8.768394113883557, "grad_norm": 0.8473963822309017, "learning_rate": 7.850717999223434e-07, "loss": 0.0235, "step": 13705 }, { "epoch": 8.769033909149073, "grad_norm": 0.8760372465439239, "learning_rate": 7.842671827001857e-07, "loss": 0.0272, "step": 13706 }, { "epoch": 8.769673704414588, "grad_norm": 0.8413464367582165, "learning_rate": 7.834629611851996e-07, "loss": 0.0223, "step": 13707 }, { "epoch": 8.770313499680102, "grad_norm": 0.945598967529011, "learning_rate": 7.826591354119206e-07, "loss": 0.03, "step": 13708 }, { "epoch": 8.770953294945617, "grad_norm": 0.9289640611635043, "learning_rate": 7.818557054148624e-07, "loss": 0.0278, "step": 13709 }, { "epoch": 8.771593090211132, "grad_norm": 0.8277021390995463, "learning_rate": 7.810526712285216e-07, "loss": 0.0247, "step": 13710 }, { "epoch": 8.772232885476647, "grad_norm": 0.8690693536708435, "learning_rate": 7.802500328873786e-07, "loss": 0.0264, "step": 13711 }, { "epoch": 8.772872680742163, "grad_norm": 0.9784051302536284, "learning_rate": 7.794477904259002e-07, "loss": 0.0286, "step": 13712 }, { "epoch": 8.773512476007678, "grad_norm": 0.805539259289097, "learning_rate": 7.786459438785321e-07, "loss": 0.0257, "step": 13713 }, { "epoch": 8.774152271273193, "grad_norm": 0.9049767802366192, "learning_rate": 7.778444932797025e-07, "loss": 0.0291, "step": 13714 }, { "epoch": 8.774792066538708, "grad_norm": 0.8556122427378323, "learning_rate": 7.770434386638259e-07, "loss": 0.0265, "step": 13715 }, { "epoch": 8.775431861804222, "grad_norm": 0.680378533470568, "learning_rate": 7.762427800652983e-07, "loss": 0.0206, "step": 13716 }, { "epoch": 8.776071657069737, "grad_norm": 5.627539324082963, "learning_rate": 7.754425175184954e-07, "loss": 0.0289, "step": 13717 }, { "epoch": 8.776711452335253, "grad_norm": 0.8379825305524949, "learning_rate": 7.746426510577843e-07, "loss": 0.0264, "step": 13718 }, { "epoch": 8.777351247600768, "grad_norm": 0.8098170266725835, "learning_rate": 7.738431807175062e-07, "loss": 0.0231, "step": 13719 }, { "epoch": 8.777991042866283, "grad_norm": 0.8374272772547647, "learning_rate": 7.730441065319904e-07, "loss": 0.0244, "step": 13720 }, { "epoch": 8.778630838131798, "grad_norm": 0.7586929381719252, "learning_rate": 7.722454285355452e-07, "loss": 0.0223, "step": 13721 }, { "epoch": 8.779270633397314, "grad_norm": 0.8364856116084699, "learning_rate": 7.714471467624684e-07, "loss": 0.0236, "step": 13722 }, { "epoch": 8.779910428662827, "grad_norm": 0.7800482029728868, "learning_rate": 7.70649261247034e-07, "loss": 0.0221, "step": 13723 }, { "epoch": 8.780550223928342, "grad_norm": 0.7227651287039668, "learning_rate": 7.698517720235021e-07, "loss": 0.0196, "step": 13724 }, { "epoch": 8.781190019193858, "grad_norm": 0.7543789866798889, "learning_rate": 7.690546791261166e-07, "loss": 0.0208, "step": 13725 }, { "epoch": 8.781829814459373, "grad_norm": 0.8156491269026384, "learning_rate": 7.682579825891034e-07, "loss": 0.0255, "step": 13726 }, { "epoch": 8.782469609724888, "grad_norm": 1.0308273708746123, "learning_rate": 7.674616824466697e-07, "loss": 0.0238, "step": 13727 }, { "epoch": 8.783109404990403, "grad_norm": 0.7877147795523209, "learning_rate": 7.666657787330089e-07, "loss": 0.0215, "step": 13728 }, { "epoch": 8.783749200255919, "grad_norm": 0.801766301302125, "learning_rate": 7.658702714822941e-07, "loss": 0.025, "step": 13729 }, { "epoch": 8.784388995521432, "grad_norm": 0.7965977709152298, "learning_rate": 7.650751607286833e-07, "loss": 0.0245, "step": 13730 }, { "epoch": 8.785028790786948, "grad_norm": 0.7810770165237759, "learning_rate": 7.642804465063158e-07, "loss": 0.0233, "step": 13731 }, { "epoch": 8.785668586052463, "grad_norm": 0.7907170417633398, "learning_rate": 7.634861288493189e-07, "loss": 0.0264, "step": 13732 }, { "epoch": 8.786308381317978, "grad_norm": 0.8323639079035102, "learning_rate": 7.626922077917975e-07, "loss": 0.0257, "step": 13733 }, { "epoch": 8.786948176583493, "grad_norm": 0.7993528779630955, "learning_rate": 7.618986833678377e-07, "loss": 0.0257, "step": 13734 }, { "epoch": 8.787587971849009, "grad_norm": 0.7669154561553244, "learning_rate": 7.611055556115188e-07, "loss": 0.0237, "step": 13735 }, { "epoch": 8.788227767114524, "grad_norm": 0.9133281422611658, "learning_rate": 7.603128245568913e-07, "loss": 0.0253, "step": 13736 }, { "epoch": 8.78886756238004, "grad_norm": 0.8354016096748722, "learning_rate": 7.595204902379938e-07, "loss": 0.0223, "step": 13737 }, { "epoch": 8.789507357645553, "grad_norm": 0.89171561046605, "learning_rate": 7.587285526888499e-07, "loss": 0.023, "step": 13738 }, { "epoch": 8.790147152911068, "grad_norm": 0.6977045697467245, "learning_rate": 7.579370119434637e-07, "loss": 0.0216, "step": 13739 }, { "epoch": 8.790786948176583, "grad_norm": 0.7573156400180098, "learning_rate": 7.571458680358212e-07, "loss": 0.0196, "step": 13740 }, { "epoch": 8.791426743442099, "grad_norm": 0.8585594792476761, "learning_rate": 7.56355120999892e-07, "loss": 0.022, "step": 13741 }, { "epoch": 8.792066538707614, "grad_norm": 0.7522211182140721, "learning_rate": 7.555647708696334e-07, "loss": 0.0231, "step": 13742 }, { "epoch": 8.792706333973129, "grad_norm": 0.6791144809719023, "learning_rate": 7.547748176789782e-07, "loss": 0.0208, "step": 13743 }, { "epoch": 8.793346129238644, "grad_norm": 0.7244952392461973, "learning_rate": 7.539852614618459e-07, "loss": 0.0213, "step": 13744 }, { "epoch": 8.79398592450416, "grad_norm": 0.72241450556967, "learning_rate": 7.531961022521395e-07, "loss": 0.0245, "step": 13745 }, { "epoch": 8.794625719769673, "grad_norm": 0.6966913427743127, "learning_rate": 7.52407340083744e-07, "loss": 0.0216, "step": 13746 }, { "epoch": 8.795265515035188, "grad_norm": 0.8733250902983192, "learning_rate": 7.516189749905267e-07, "loss": 0.0269, "step": 13747 }, { "epoch": 8.795905310300704, "grad_norm": 0.7928208201907585, "learning_rate": 7.508310070063374e-07, "loss": 0.022, "step": 13748 }, { "epoch": 8.796545105566219, "grad_norm": 0.7344614992952617, "learning_rate": 7.500434361650133e-07, "loss": 0.0216, "step": 13749 }, { "epoch": 8.797184900831734, "grad_norm": 0.7643809168591865, "learning_rate": 7.492562625003697e-07, "loss": 0.0225, "step": 13750 }, { "epoch": 8.79782469609725, "grad_norm": 0.772536636245896, "learning_rate": 7.484694860462038e-07, "loss": 0.0222, "step": 13751 }, { "epoch": 8.798464491362765, "grad_norm": 0.8151606314384601, "learning_rate": 7.476831068363044e-07, "loss": 0.0222, "step": 13752 }, { "epoch": 8.799104286628278, "grad_norm": 0.770523673160279, "learning_rate": 7.468971249044321e-07, "loss": 0.0245, "step": 13753 }, { "epoch": 8.799744081893794, "grad_norm": 0.8725496312484379, "learning_rate": 7.461115402843367e-07, "loss": 0.0266, "step": 13754 }, { "epoch": 8.800383877159309, "grad_norm": 0.9693322011207698, "learning_rate": 7.453263530097521e-07, "loss": 0.0254, "step": 13755 }, { "epoch": 8.801023672424824, "grad_norm": 0.8005497814713417, "learning_rate": 7.445415631143905e-07, "loss": 0.0214, "step": 13756 }, { "epoch": 8.80166346769034, "grad_norm": 0.752982887770054, "learning_rate": 7.437571706319501e-07, "loss": 0.0207, "step": 13757 }, { "epoch": 8.802303262955855, "grad_norm": 0.8347438468924351, "learning_rate": 7.429731755961123e-07, "loss": 0.0288, "step": 13758 }, { "epoch": 8.80294305822137, "grad_norm": 0.8960663177542487, "learning_rate": 7.421895780405386e-07, "loss": 0.0265, "step": 13759 }, { "epoch": 8.803582853486883, "grad_norm": 0.7735278476829042, "learning_rate": 7.414063779988756e-07, "loss": 0.0215, "step": 13760 }, { "epoch": 8.804222648752399, "grad_norm": 0.788549477837996, "learning_rate": 7.406235755047519e-07, "loss": 0.0206, "step": 13761 }, { "epoch": 8.804862444017914, "grad_norm": 0.8592332807885259, "learning_rate": 7.398411705917818e-07, "loss": 0.0239, "step": 13762 }, { "epoch": 8.80550223928343, "grad_norm": 0.8125958229552679, "learning_rate": 7.390591632935595e-07, "loss": 0.0221, "step": 13763 }, { "epoch": 8.806142034548945, "grad_norm": 0.928822550708196, "learning_rate": 7.382775536436615e-07, "loss": 0.0292, "step": 13764 }, { "epoch": 8.80678182981446, "grad_norm": 0.7824740318972319, "learning_rate": 7.374963416756509e-07, "loss": 0.0235, "step": 13765 }, { "epoch": 8.807421625079975, "grad_norm": 0.9073865153600674, "learning_rate": 7.36715527423072e-07, "loss": 0.0264, "step": 13766 }, { "epoch": 8.808061420345489, "grad_norm": 0.7779661121651275, "learning_rate": 7.35935110919449e-07, "loss": 0.0243, "step": 13767 }, { "epoch": 8.808701215611004, "grad_norm": 0.8208538562877564, "learning_rate": 7.351550921982919e-07, "loss": 0.0218, "step": 13768 }, { "epoch": 8.809341010876519, "grad_norm": 0.8935746617819744, "learning_rate": 7.343754712930951e-07, "loss": 0.0252, "step": 13769 }, { "epoch": 8.809980806142034, "grad_norm": 0.8594372676564916, "learning_rate": 7.335962482373338e-07, "loss": 0.024, "step": 13770 }, { "epoch": 8.81062060140755, "grad_norm": 0.7716343652724006, "learning_rate": 7.328174230644647e-07, "loss": 0.023, "step": 13771 }, { "epoch": 8.811260396673065, "grad_norm": 0.8099742670500384, "learning_rate": 7.32038995807931e-07, "loss": 0.0224, "step": 13772 }, { "epoch": 8.81190019193858, "grad_norm": 0.816310959533559, "learning_rate": 7.312609665011572e-07, "loss": 0.0226, "step": 13773 }, { "epoch": 8.812539987204095, "grad_norm": 0.7062297582511856, "learning_rate": 7.304833351775498e-07, "loss": 0.0206, "step": 13774 }, { "epoch": 8.813179782469609, "grad_norm": 0.7073610259828816, "learning_rate": 7.297061018704987e-07, "loss": 0.0217, "step": 13775 }, { "epoch": 8.813819577735124, "grad_norm": 0.8281855945991241, "learning_rate": 7.289292666133774e-07, "loss": 0.0238, "step": 13776 }, { "epoch": 8.81445937300064, "grad_norm": 0.6947551284074057, "learning_rate": 7.281528294395402e-07, "loss": 0.0195, "step": 13777 }, { "epoch": 8.815099168266155, "grad_norm": 0.8194465840082504, "learning_rate": 7.27376790382327e-07, "loss": 0.0208, "step": 13778 }, { "epoch": 8.81573896353167, "grad_norm": 0.6742737689496366, "learning_rate": 7.266011494750602e-07, "loss": 0.0174, "step": 13779 }, { "epoch": 8.816378758797185, "grad_norm": 0.8329042899232281, "learning_rate": 7.258259067510443e-07, "loss": 0.0249, "step": 13780 }, { "epoch": 8.8170185540627, "grad_norm": 0.7763490125337131, "learning_rate": 7.250510622435647e-07, "loss": 0.0209, "step": 13781 }, { "epoch": 8.817658349328216, "grad_norm": 0.8332210911256644, "learning_rate": 7.24276615985896e-07, "loss": 0.0238, "step": 13782 }, { "epoch": 8.81829814459373, "grad_norm": 0.8264521889960202, "learning_rate": 7.235025680112883e-07, "loss": 0.0264, "step": 13783 }, { "epoch": 8.818937939859245, "grad_norm": 0.8576274591084119, "learning_rate": 7.227289183529773e-07, "loss": 0.0237, "step": 13784 }, { "epoch": 8.81957773512476, "grad_norm": 0.920139367789204, "learning_rate": 7.219556670441852e-07, "loss": 0.0245, "step": 13785 }, { "epoch": 8.820217530390275, "grad_norm": 0.7352956019772383, "learning_rate": 7.211828141181132e-07, "loss": 0.0232, "step": 13786 }, { "epoch": 8.82085732565579, "grad_norm": 0.7644814980397909, "learning_rate": 7.204103596079437e-07, "loss": 0.0232, "step": 13787 }, { "epoch": 8.821497120921306, "grad_norm": 0.7755863655319808, "learning_rate": 7.19638303546848e-07, "loss": 0.0225, "step": 13788 }, { "epoch": 8.822136916186821, "grad_norm": 0.819013411608884, "learning_rate": 7.188666459679738e-07, "loss": 0.0265, "step": 13789 }, { "epoch": 8.822776711452335, "grad_norm": 0.7859151282986354, "learning_rate": 7.180953869044572e-07, "loss": 0.0243, "step": 13790 }, { "epoch": 8.82341650671785, "grad_norm": 0.847784663287854, "learning_rate": 7.173245263894114e-07, "loss": 0.0303, "step": 13791 }, { "epoch": 8.824056301983365, "grad_norm": 0.8265833812579547, "learning_rate": 7.165540644559387e-07, "loss": 0.0258, "step": 13792 }, { "epoch": 8.82469609724888, "grad_norm": 0.8091076795998431, "learning_rate": 7.157840011371208e-07, "loss": 0.0254, "step": 13793 }, { "epoch": 8.825335892514396, "grad_norm": 0.8481725949598306, "learning_rate": 7.150143364660212e-07, "loss": 0.0236, "step": 13794 }, { "epoch": 8.825975687779911, "grad_norm": 0.8056965898214546, "learning_rate": 7.142450704756909e-07, "loss": 0.0238, "step": 13795 }, { "epoch": 8.826615483045426, "grad_norm": 0.7606757589911192, "learning_rate": 7.134762031991594e-07, "loss": 0.0224, "step": 13796 }, { "epoch": 8.82725527831094, "grad_norm": 0.8931821634085104, "learning_rate": 7.1270773466944e-07, "loss": 0.0255, "step": 13797 }, { "epoch": 8.827895073576455, "grad_norm": 0.8689198618146676, "learning_rate": 7.119396649195276e-07, "loss": 0.026, "step": 13798 }, { "epoch": 8.82853486884197, "grad_norm": 0.8033078250462735, "learning_rate": 7.111719939824058e-07, "loss": 0.0234, "step": 13799 }, { "epoch": 8.829174664107486, "grad_norm": 0.7154378735255802, "learning_rate": 7.10404721891036e-07, "loss": 0.0198, "step": 13800 }, { "epoch": 8.829814459373, "grad_norm": 0.7560857614692631, "learning_rate": 7.096378486783595e-07, "loss": 0.0225, "step": 13801 }, { "epoch": 8.830454254638516, "grad_norm": 0.8616468535833981, "learning_rate": 7.088713743773101e-07, "loss": 0.0224, "step": 13802 }, { "epoch": 8.831094049904031, "grad_norm": 0.7796219580769269, "learning_rate": 7.081052990207959e-07, "loss": 0.0218, "step": 13803 }, { "epoch": 8.831733845169545, "grad_norm": 0.7969925722484812, "learning_rate": 7.073396226417106e-07, "loss": 0.0242, "step": 13804 }, { "epoch": 8.83237364043506, "grad_norm": 0.7691228417374015, "learning_rate": 7.065743452729324e-07, "loss": 0.0242, "step": 13805 }, { "epoch": 8.833013435700575, "grad_norm": 0.8061249352218877, "learning_rate": 7.058094669473192e-07, "loss": 0.0226, "step": 13806 }, { "epoch": 8.83365323096609, "grad_norm": 0.7493738874643947, "learning_rate": 7.05044987697715e-07, "loss": 0.0204, "step": 13807 }, { "epoch": 8.834293026231606, "grad_norm": 0.7440588467745298, "learning_rate": 7.042809075569435e-07, "loss": 0.0233, "step": 13808 }, { "epoch": 8.834932821497121, "grad_norm": 0.8324101679248199, "learning_rate": 7.03517226557815e-07, "loss": 0.0224, "step": 13809 }, { "epoch": 8.835572616762636, "grad_norm": 0.7913972012649064, "learning_rate": 7.0275394473312e-07, "loss": 0.0223, "step": 13810 }, { "epoch": 8.836212412028152, "grad_norm": 0.8614683870138197, "learning_rate": 7.0199106211563e-07, "loss": 0.0242, "step": 13811 }, { "epoch": 8.836852207293665, "grad_norm": 0.8747205400382005, "learning_rate": 7.012285787381056e-07, "loss": 0.0228, "step": 13812 }, { "epoch": 8.83749200255918, "grad_norm": 0.8556324833595838, "learning_rate": 7.00466494633285e-07, "loss": 0.0251, "step": 13813 }, { "epoch": 8.838131797824696, "grad_norm": 0.8783541732177538, "learning_rate": 6.997048098338888e-07, "loss": 0.0256, "step": 13814 }, { "epoch": 8.838771593090211, "grad_norm": 0.815085313039524, "learning_rate": 6.98943524372625e-07, "loss": 0.0244, "step": 13815 }, { "epoch": 8.839411388355726, "grad_norm": 0.8077878242441475, "learning_rate": 6.981826382821821e-07, "loss": 0.0256, "step": 13816 }, { "epoch": 8.840051183621242, "grad_norm": 0.8863686763817362, "learning_rate": 6.974221515952295e-07, "loss": 0.0205, "step": 13817 }, { "epoch": 8.840690978886757, "grad_norm": 0.7625655235299486, "learning_rate": 6.966620643444222e-07, "loss": 0.0238, "step": 13818 }, { "epoch": 8.841330774152272, "grad_norm": 0.7762837982888072, "learning_rate": 6.959023765623962e-07, "loss": 0.023, "step": 13819 }, { "epoch": 8.841970569417786, "grad_norm": 0.8561801553494234, "learning_rate": 6.951430882817712e-07, "loss": 0.0238, "step": 13820 }, { "epoch": 8.842610364683301, "grad_norm": 0.9781401561698062, "learning_rate": 6.943841995351486e-07, "loss": 0.03, "step": 13821 }, { "epoch": 8.843250159948816, "grad_norm": 0.8364128817292849, "learning_rate": 6.93625710355117e-07, "loss": 0.0248, "step": 13822 }, { "epoch": 8.843889955214332, "grad_norm": 0.8503761343424536, "learning_rate": 6.928676207742425e-07, "loss": 0.0218, "step": 13823 }, { "epoch": 8.844529750479847, "grad_norm": 0.7830497611596242, "learning_rate": 6.921099308250756e-07, "loss": 0.0219, "step": 13824 }, { "epoch": 8.845169545745362, "grad_norm": 0.7654331496510317, "learning_rate": 6.913526405401494e-07, "loss": 0.0209, "step": 13825 }, { "epoch": 8.845809341010877, "grad_norm": 0.7092054513300143, "learning_rate": 6.905957499519844e-07, "loss": 0.0197, "step": 13826 }, { "epoch": 8.84644913627639, "grad_norm": 0.7949924295720078, "learning_rate": 6.898392590930769e-07, "loss": 0.0243, "step": 13827 }, { "epoch": 8.847088931541906, "grad_norm": 0.8222562630317862, "learning_rate": 6.890831679959086e-07, "loss": 0.0214, "step": 13828 }, { "epoch": 8.847728726807421, "grad_norm": 0.7484374300674972, "learning_rate": 6.883274766929482e-07, "loss": 0.0223, "step": 13829 }, { "epoch": 8.848368522072937, "grad_norm": 0.8027041308878294, "learning_rate": 6.875721852166417e-07, "loss": 0.0228, "step": 13830 }, { "epoch": 8.849008317338452, "grad_norm": 0.8214164960535562, "learning_rate": 6.868172935994188e-07, "loss": 0.0235, "step": 13831 }, { "epoch": 8.849648112603967, "grad_norm": 0.7054481681068469, "learning_rate": 6.860628018736948e-07, "loss": 0.0185, "step": 13832 }, { "epoch": 8.850287907869482, "grad_norm": 0.8313933424389215, "learning_rate": 6.853087100718669e-07, "loss": 0.0271, "step": 13833 }, { "epoch": 8.850927703134996, "grad_norm": 0.9337939399278156, "learning_rate": 6.845550182263138e-07, "loss": 0.0259, "step": 13834 }, { "epoch": 8.851567498400511, "grad_norm": 0.9359968513319779, "learning_rate": 6.838017263693964e-07, "loss": 0.0261, "step": 13835 }, { "epoch": 8.852207293666027, "grad_norm": 0.7323232061736339, "learning_rate": 6.830488345334607e-07, "loss": 0.0226, "step": 13836 }, { "epoch": 8.852847088931542, "grad_norm": 0.9002748033070206, "learning_rate": 6.822963427508356e-07, "loss": 0.0296, "step": 13837 }, { "epoch": 8.853486884197057, "grad_norm": 0.858414508484253, "learning_rate": 6.815442510538273e-07, "loss": 0.0276, "step": 13838 }, { "epoch": 8.854126679462572, "grad_norm": 0.9079013780360639, "learning_rate": 6.807925594747355e-07, "loss": 0.0257, "step": 13839 }, { "epoch": 8.854766474728088, "grad_norm": 0.9350784715796291, "learning_rate": 6.800412680458335e-07, "loss": 0.026, "step": 13840 }, { "epoch": 8.855406269993601, "grad_norm": 0.7625391159163503, "learning_rate": 6.792903767993786e-07, "loss": 0.0233, "step": 13841 }, { "epoch": 8.856046065259116, "grad_norm": 0.7703652085420292, "learning_rate": 6.785398857676162e-07, "loss": 0.023, "step": 13842 }, { "epoch": 8.856685860524632, "grad_norm": 0.760489950831594, "learning_rate": 6.777897949827694e-07, "loss": 0.0228, "step": 13843 }, { "epoch": 8.857325655790147, "grad_norm": 0.7812756107772957, "learning_rate": 6.77040104477047e-07, "loss": 0.0239, "step": 13844 }, { "epoch": 8.857965451055662, "grad_norm": 0.8005188205741489, "learning_rate": 6.762908142826352e-07, "loss": 0.022, "step": 13845 }, { "epoch": 8.858605246321178, "grad_norm": 0.7920348853341409, "learning_rate": 6.75541924431713e-07, "loss": 0.0212, "step": 13846 }, { "epoch": 8.859245041586693, "grad_norm": 0.8875360805844724, "learning_rate": 6.747934349564322e-07, "loss": 0.0269, "step": 13847 }, { "epoch": 8.859884836852208, "grad_norm": 0.7367889981608736, "learning_rate": 6.74045345888934e-07, "loss": 0.0212, "step": 13848 }, { "epoch": 8.860524632117722, "grad_norm": 0.8655209562698447, "learning_rate": 6.732976572613393e-07, "loss": 0.0213, "step": 13849 }, { "epoch": 8.861164427383237, "grad_norm": 0.745769918509005, "learning_rate": 6.725503691057523e-07, "loss": 0.0209, "step": 13850 }, { "epoch": 8.861804222648752, "grad_norm": 0.89037748202015, "learning_rate": 6.718034814542585e-07, "loss": 0.0285, "step": 13851 }, { "epoch": 8.862444017914267, "grad_norm": 0.7535479510362362, "learning_rate": 6.7105699433893e-07, "loss": 0.0217, "step": 13852 }, { "epoch": 8.863083813179783, "grad_norm": 0.8404613419483411, "learning_rate": 6.703109077918202e-07, "loss": 0.0209, "step": 13853 }, { "epoch": 8.863723608445298, "grad_norm": 0.6916678036480965, "learning_rate": 6.695652218449644e-07, "loss": 0.0203, "step": 13854 }, { "epoch": 8.864363403710813, "grad_norm": 0.7570616361395365, "learning_rate": 6.688199365303782e-07, "loss": 0.0225, "step": 13855 }, { "epoch": 8.865003198976328, "grad_norm": 0.7269785289189387, "learning_rate": 6.680750518800672e-07, "loss": 0.021, "step": 13856 }, { "epoch": 8.865642994241842, "grad_norm": 0.8139835439469627, "learning_rate": 6.673305679260123e-07, "loss": 0.0234, "step": 13857 }, { "epoch": 8.866282789507357, "grad_norm": 0.8229270605159141, "learning_rate": 6.665864847001802e-07, "loss": 0.024, "step": 13858 }, { "epoch": 8.866922584772873, "grad_norm": 0.8123887560437872, "learning_rate": 6.658428022345232e-07, "loss": 0.0231, "step": 13859 }, { "epoch": 8.867562380038388, "grad_norm": 0.7013623434368957, "learning_rate": 6.650995205609723e-07, "loss": 0.0204, "step": 13860 }, { "epoch": 8.868202175303903, "grad_norm": 0.7911037269460474, "learning_rate": 6.643566397114398e-07, "loss": 0.0242, "step": 13861 }, { "epoch": 8.868841970569418, "grad_norm": 0.9211169372976322, "learning_rate": 6.636141597178292e-07, "loss": 0.0234, "step": 13862 }, { "epoch": 8.869481765834934, "grad_norm": 0.7993074140651022, "learning_rate": 6.628720806120181e-07, "loss": 0.0247, "step": 13863 }, { "epoch": 8.870121561100447, "grad_norm": 0.7323151375633006, "learning_rate": 6.621304024258701e-07, "loss": 0.0211, "step": 13864 }, { "epoch": 8.870761356365962, "grad_norm": 0.7367802630581984, "learning_rate": 6.613891251912318e-07, "loss": 0.0204, "step": 13865 }, { "epoch": 8.871401151631478, "grad_norm": 0.9807440991289276, "learning_rate": 6.606482489399335e-07, "loss": 0.0252, "step": 13866 }, { "epoch": 8.872040946896993, "grad_norm": 0.8322494021652486, "learning_rate": 6.599077737037851e-07, "loss": 0.0253, "step": 13867 }, { "epoch": 8.872680742162508, "grad_norm": 0.6596165996764407, "learning_rate": 6.591676995145801e-07, "loss": 0.0203, "step": 13868 }, { "epoch": 8.873320537428024, "grad_norm": 0.8193311444543018, "learning_rate": 6.584280264040999e-07, "loss": 0.0243, "step": 13869 }, { "epoch": 8.873960332693539, "grad_norm": 0.8018945916798096, "learning_rate": 6.576887544041033e-07, "loss": 0.024, "step": 13870 }, { "epoch": 8.874600127959052, "grad_norm": 0.7899905773301144, "learning_rate": 6.569498835463317e-07, "loss": 0.0238, "step": 13871 }, { "epoch": 8.875239923224568, "grad_norm": 0.9012040729133483, "learning_rate": 6.562114138625131e-07, "loss": 0.026, "step": 13872 }, { "epoch": 8.875879718490083, "grad_norm": 0.7898452202155387, "learning_rate": 6.554733453843542e-07, "loss": 0.0253, "step": 13873 }, { "epoch": 8.876519513755598, "grad_norm": 0.9025430834466487, "learning_rate": 6.547356781435488e-07, "loss": 0.0263, "step": 13874 }, { "epoch": 8.877159309021113, "grad_norm": 0.8363294939853154, "learning_rate": 6.539984121717668e-07, "loss": 0.0224, "step": 13875 }, { "epoch": 8.877799104286629, "grad_norm": 0.7658241906558194, "learning_rate": 6.532615475006699e-07, "loss": 0.0216, "step": 13876 }, { "epoch": 8.878438899552144, "grad_norm": 0.742128587457347, "learning_rate": 6.525250841618946e-07, "loss": 0.0207, "step": 13877 }, { "epoch": 8.879078694817657, "grad_norm": 0.8479759389583819, "learning_rate": 6.51789022187066e-07, "loss": 0.0242, "step": 13878 }, { "epoch": 8.879718490083173, "grad_norm": 0.8286672557494943, "learning_rate": 6.510533616077863e-07, "loss": 0.0259, "step": 13879 }, { "epoch": 8.880358285348688, "grad_norm": 0.7279433542738645, "learning_rate": 6.503181024556448e-07, "loss": 0.0208, "step": 13880 }, { "epoch": 8.880998080614203, "grad_norm": 0.7773184414299431, "learning_rate": 6.495832447622108e-07, "loss": 0.0248, "step": 13881 }, { "epoch": 8.881637875879719, "grad_norm": 0.7917332102537884, "learning_rate": 6.488487885590411e-07, "loss": 0.0215, "step": 13882 }, { "epoch": 8.882277671145234, "grad_norm": 0.8226843423040188, "learning_rate": 6.481147338776705e-07, "loss": 0.0249, "step": 13883 }, { "epoch": 8.882917466410749, "grad_norm": 0.8474851865933664, "learning_rate": 6.473810807496161e-07, "loss": 0.0247, "step": 13884 }, { "epoch": 8.883557261676264, "grad_norm": 0.9015341707904112, "learning_rate": 6.466478292063805e-07, "loss": 0.0284, "step": 13885 }, { "epoch": 8.884197056941778, "grad_norm": 0.8257017091066068, "learning_rate": 6.459149792794506e-07, "loss": 0.0234, "step": 13886 }, { "epoch": 8.884836852207293, "grad_norm": 0.8095863562117339, "learning_rate": 6.451825310002913e-07, "loss": 0.0244, "step": 13887 }, { "epoch": 8.885476647472808, "grad_norm": 0.7646066840036898, "learning_rate": 6.444504844003518e-07, "loss": 0.0204, "step": 13888 }, { "epoch": 8.886116442738324, "grad_norm": 0.8662094978687593, "learning_rate": 6.437188395110672e-07, "loss": 0.0269, "step": 13889 }, { "epoch": 8.886756238003839, "grad_norm": 0.8595307519518216, "learning_rate": 6.429875963638532e-07, "loss": 0.0279, "step": 13890 }, { "epoch": 8.887396033269354, "grad_norm": 0.8181968833303848, "learning_rate": 6.422567549901037e-07, "loss": 0.0259, "step": 13891 }, { "epoch": 8.88803582853487, "grad_norm": 0.7117498716435038, "learning_rate": 6.41526315421206e-07, "loss": 0.0214, "step": 13892 }, { "epoch": 8.888675623800385, "grad_norm": 0.9269328961485195, "learning_rate": 6.407962776885212e-07, "loss": 0.0264, "step": 13893 }, { "epoch": 8.889315419065898, "grad_norm": 0.741421724865454, "learning_rate": 6.400666418233947e-07, "loss": 0.0217, "step": 13894 }, { "epoch": 8.889955214331414, "grad_norm": 0.7680805226728415, "learning_rate": 6.393374078571568e-07, "loss": 0.0217, "step": 13895 }, { "epoch": 8.890595009596929, "grad_norm": 0.7539399170545699, "learning_rate": 6.386085758211191e-07, "loss": 0.0229, "step": 13896 }, { "epoch": 8.891234804862444, "grad_norm": 0.8064760201275378, "learning_rate": 6.378801457465767e-07, "loss": 0.0215, "step": 13897 }, { "epoch": 8.89187460012796, "grad_norm": 0.7356734753953226, "learning_rate": 6.371521176648054e-07, "loss": 0.0212, "step": 13898 }, { "epoch": 8.892514395393475, "grad_norm": 0.8609134018771485, "learning_rate": 6.364244916070684e-07, "loss": 0.0267, "step": 13899 }, { "epoch": 8.89315419065899, "grad_norm": 0.7928397799372957, "learning_rate": 6.35697267604607e-07, "loss": 0.0256, "step": 13900 }, { "epoch": 8.893793985924503, "grad_norm": 0.8608282900256474, "learning_rate": 6.349704456886463e-07, "loss": 0.0237, "step": 13901 }, { "epoch": 8.894433781190019, "grad_norm": 0.7177350160330942, "learning_rate": 6.342440258903959e-07, "loss": 0.0204, "step": 13902 }, { "epoch": 8.895073576455534, "grad_norm": 0.8143299289487723, "learning_rate": 6.335180082410475e-07, "loss": 0.0248, "step": 13903 }, { "epoch": 8.89571337172105, "grad_norm": 0.6872056366051476, "learning_rate": 6.32792392771775e-07, "loss": 0.0195, "step": 13904 }, { "epoch": 8.896353166986565, "grad_norm": 0.8826269056758482, "learning_rate": 6.320671795137312e-07, "loss": 0.0245, "step": 13905 }, { "epoch": 8.89699296225208, "grad_norm": 0.7489099503301112, "learning_rate": 6.313423684980602e-07, "loss": 0.0222, "step": 13906 }, { "epoch": 8.897632757517595, "grad_norm": 0.8819300883005404, "learning_rate": 6.306179597558825e-07, "loss": 0.0263, "step": 13907 }, { "epoch": 8.898272552783109, "grad_norm": 0.7196005261337646, "learning_rate": 6.298939533183035e-07, "loss": 0.0212, "step": 13908 }, { "epoch": 8.898912348048624, "grad_norm": 0.7821936353016945, "learning_rate": 6.291703492164091e-07, "loss": 0.0226, "step": 13909 }, { "epoch": 8.89955214331414, "grad_norm": 0.7021523303954841, "learning_rate": 6.284471474812704e-07, "loss": 0.0188, "step": 13910 }, { "epoch": 8.900191938579654, "grad_norm": 0.8520982391308697, "learning_rate": 6.27724348143941e-07, "loss": 0.024, "step": 13911 }, { "epoch": 8.90083173384517, "grad_norm": 0.838678818020373, "learning_rate": 6.270019512354552e-07, "loss": 0.0239, "step": 13912 }, { "epoch": 8.901471529110685, "grad_norm": 0.8778774067937438, "learning_rate": 6.262799567868327e-07, "loss": 0.0236, "step": 13913 }, { "epoch": 8.9021113243762, "grad_norm": 0.7313548040881112, "learning_rate": 6.255583648290753e-07, "loss": 0.0207, "step": 13914 }, { "epoch": 8.902751119641714, "grad_norm": 0.9710955833248767, "learning_rate": 6.248371753931637e-07, "loss": 0.0252, "step": 13915 }, { "epoch": 8.903390914907229, "grad_norm": 0.8728924989246927, "learning_rate": 6.241163885100676e-07, "loss": 0.0255, "step": 13916 }, { "epoch": 8.904030710172744, "grad_norm": 0.8536622352738933, "learning_rate": 6.233960042107368e-07, "loss": 0.0233, "step": 13917 }, { "epoch": 8.90467050543826, "grad_norm": 0.7649039026363966, "learning_rate": 6.226760225260986e-07, "loss": 0.0185, "step": 13918 }, { "epoch": 8.905310300703775, "grad_norm": 0.8174787952522493, "learning_rate": 6.219564434870739e-07, "loss": 0.0235, "step": 13919 }, { "epoch": 8.90595009596929, "grad_norm": 0.8484630076956894, "learning_rate": 6.21237267124557e-07, "loss": 0.0248, "step": 13920 }, { "epoch": 8.906589891234805, "grad_norm": 0.7755545325514255, "learning_rate": 6.205184934694286e-07, "loss": 0.0232, "step": 13921 }, { "epoch": 8.90722968650032, "grad_norm": 0.7566759727346092, "learning_rate": 6.198001225525497e-07, "loss": 0.0253, "step": 13922 }, { "epoch": 8.907869481765834, "grad_norm": 0.8336001770916007, "learning_rate": 6.190821544047687e-07, "loss": 0.0262, "step": 13923 }, { "epoch": 8.90850927703135, "grad_norm": 0.8517559535814596, "learning_rate": 6.183645890569123e-07, "loss": 0.0244, "step": 13924 }, { "epoch": 8.909149072296865, "grad_norm": 0.7376585960000187, "learning_rate": 6.176474265397925e-07, "loss": 0.0212, "step": 13925 }, { "epoch": 8.90978886756238, "grad_norm": 0.798787721120064, "learning_rate": 6.169306668842012e-07, "loss": 0.0259, "step": 13926 }, { "epoch": 8.910428662827895, "grad_norm": 0.7136778695095218, "learning_rate": 6.162143101209173e-07, "loss": 0.0215, "step": 13927 }, { "epoch": 8.91106845809341, "grad_norm": 0.8330808776514728, "learning_rate": 6.15498356280696e-07, "loss": 0.0256, "step": 13928 }, { "epoch": 8.911708253358926, "grad_norm": 0.8504344039006829, "learning_rate": 6.147828053942828e-07, "loss": 0.0258, "step": 13929 }, { "epoch": 8.912348048624441, "grad_norm": 0.7311549184409103, "learning_rate": 6.14067657492402e-07, "loss": 0.0205, "step": 13930 }, { "epoch": 8.912987843889955, "grad_norm": 0.7704170974568055, "learning_rate": 6.133529126057591e-07, "loss": 0.0237, "step": 13931 }, { "epoch": 8.91362763915547, "grad_norm": 0.7508967487579595, "learning_rate": 6.126385707650428e-07, "loss": 0.0208, "step": 13932 }, { "epoch": 8.914267434420985, "grad_norm": 0.7270890147043304, "learning_rate": 6.119246320009297e-07, "loss": 0.0202, "step": 13933 }, { "epoch": 8.9149072296865, "grad_norm": 0.937861981710403, "learning_rate": 6.112110963440721e-07, "loss": 0.0277, "step": 13934 }, { "epoch": 8.915547024952016, "grad_norm": 0.7265642630990831, "learning_rate": 6.104979638251074e-07, "loss": 0.0202, "step": 13935 }, { "epoch": 8.916186820217531, "grad_norm": 0.8704758707002784, "learning_rate": 6.097852344746591e-07, "loss": 0.0269, "step": 13936 }, { "epoch": 8.916826615483046, "grad_norm": 0.8116511427611139, "learning_rate": 6.090729083233293e-07, "loss": 0.0241, "step": 13937 }, { "epoch": 8.91746641074856, "grad_norm": 0.8314335324262739, "learning_rate": 6.083609854017036e-07, "loss": 0.0249, "step": 13938 }, { "epoch": 8.918106206014075, "grad_norm": 0.827382765800771, "learning_rate": 6.076494657403498e-07, "loss": 0.0259, "step": 13939 }, { "epoch": 8.91874600127959, "grad_norm": 0.850874234243329, "learning_rate": 6.069383493698211e-07, "loss": 0.024, "step": 13940 }, { "epoch": 8.919385796545106, "grad_norm": 0.7453671935739745, "learning_rate": 6.06227636320651e-07, "loss": 0.0208, "step": 13941 }, { "epoch": 8.92002559181062, "grad_norm": 0.8355103041946187, "learning_rate": 6.055173266233538e-07, "loss": 0.0212, "step": 13942 }, { "epoch": 8.920665387076136, "grad_norm": 0.7512232063019239, "learning_rate": 6.048074203084331e-07, "loss": 0.022, "step": 13943 }, { "epoch": 8.921305182341651, "grad_norm": 0.8910089196343509, "learning_rate": 6.040979174063677e-07, "loss": 0.0245, "step": 13944 }, { "epoch": 8.921944977607165, "grad_norm": 0.7655926296413635, "learning_rate": 6.033888179476233e-07, "loss": 0.0226, "step": 13945 }, { "epoch": 8.92258477287268, "grad_norm": 0.8846827381608157, "learning_rate": 6.026801219626499e-07, "loss": 0.0248, "step": 13946 }, { "epoch": 8.923224568138195, "grad_norm": 0.8596257103739425, "learning_rate": 6.019718294818744e-07, "loss": 0.0237, "step": 13947 }, { "epoch": 8.92386436340371, "grad_norm": 0.7853603071956194, "learning_rate": 6.012639405357101e-07, "loss": 0.0197, "step": 13948 }, { "epoch": 8.924504158669226, "grad_norm": 0.8615039411661021, "learning_rate": 6.00556455154554e-07, "loss": 0.0248, "step": 13949 }, { "epoch": 8.925143953934741, "grad_norm": 0.827663110231718, "learning_rate": 5.998493733687838e-07, "loss": 0.0226, "step": 13950 }, { "epoch": 8.925783749200257, "grad_norm": 0.9238441545248753, "learning_rate": 5.991426952087598e-07, "loss": 0.0265, "step": 13951 }, { "epoch": 8.92642354446577, "grad_norm": 0.812361425540939, "learning_rate": 5.984364207048232e-07, "loss": 0.0243, "step": 13952 }, { "epoch": 8.927063339731285, "grad_norm": 0.9175129934143654, "learning_rate": 5.977305498873054e-07, "loss": 0.0322, "step": 13953 }, { "epoch": 8.9277031349968, "grad_norm": 0.8250995984654677, "learning_rate": 5.97025082786512e-07, "loss": 0.0254, "step": 13954 }, { "epoch": 8.928342930262316, "grad_norm": 0.7967945315895311, "learning_rate": 5.963200194327357e-07, "loss": 0.0234, "step": 13955 }, { "epoch": 8.928982725527831, "grad_norm": 0.7696182067961213, "learning_rate": 5.956153598562486e-07, "loss": 0.023, "step": 13956 }, { "epoch": 8.929622520793346, "grad_norm": 0.8298999874766475, "learning_rate": 5.949111040873101e-07, "loss": 0.0265, "step": 13957 }, { "epoch": 8.930262316058862, "grad_norm": 0.8406034547335256, "learning_rate": 5.942072521561559e-07, "loss": 0.0242, "step": 13958 }, { "epoch": 8.930902111324377, "grad_norm": 0.8835919518325938, "learning_rate": 5.935038040930119e-07, "loss": 0.0262, "step": 13959 }, { "epoch": 8.93154190658989, "grad_norm": 0.8304641479708026, "learning_rate": 5.928007599280827e-07, "loss": 0.025, "step": 13960 }, { "epoch": 8.932181701855406, "grad_norm": 0.8768453768677718, "learning_rate": 5.920981196915543e-07, "loss": 0.0271, "step": 13961 }, { "epoch": 8.932821497120921, "grad_norm": 0.7557010489708756, "learning_rate": 5.913958834135958e-07, "loss": 0.0188, "step": 13962 }, { "epoch": 8.933461292386436, "grad_norm": 0.6837226143858903, "learning_rate": 5.906940511243631e-07, "loss": 0.0198, "step": 13963 }, { "epoch": 8.934101087651952, "grad_norm": 0.7816858995323807, "learning_rate": 5.899926228539898e-07, "loss": 0.0247, "step": 13964 }, { "epoch": 8.934740882917467, "grad_norm": 0.7973278941405808, "learning_rate": 5.892915986325931e-07, "loss": 0.022, "step": 13965 }, { "epoch": 8.935380678182982, "grad_norm": 0.7990660974778303, "learning_rate": 5.885909784902766e-07, "loss": 0.022, "step": 13966 }, { "epoch": 8.936020473448497, "grad_norm": 0.8387764605151954, "learning_rate": 5.878907624571207e-07, "loss": 0.0233, "step": 13967 }, { "epoch": 8.93666026871401, "grad_norm": 0.7465011627761821, "learning_rate": 5.871909505631934e-07, "loss": 0.02, "step": 13968 }, { "epoch": 8.937300063979526, "grad_norm": 0.908767375750418, "learning_rate": 5.864915428385432e-07, "loss": 0.0288, "step": 13969 }, { "epoch": 8.937939859245041, "grad_norm": 0.8108084325534715, "learning_rate": 5.857925393131991e-07, "loss": 0.0227, "step": 13970 }, { "epoch": 8.938579654510557, "grad_norm": 0.7850596095521603, "learning_rate": 5.850939400171784e-07, "loss": 0.0213, "step": 13971 }, { "epoch": 8.939219449776072, "grad_norm": 0.7809265795373124, "learning_rate": 5.843957449804738e-07, "loss": 0.0215, "step": 13972 }, { "epoch": 8.939859245041587, "grad_norm": 0.8478275100900301, "learning_rate": 5.836979542330679e-07, "loss": 0.023, "step": 13973 }, { "epoch": 8.940499040307103, "grad_norm": 0.7522306128691653, "learning_rate": 5.830005678049222e-07, "loss": 0.0181, "step": 13974 }, { "epoch": 8.941138835572616, "grad_norm": 0.8514575674195645, "learning_rate": 5.823035857259796e-07, "loss": 0.0274, "step": 13975 }, { "epoch": 8.941778630838131, "grad_norm": 0.9096599782722913, "learning_rate": 5.816070080261693e-07, "loss": 0.0237, "step": 13976 }, { "epoch": 8.942418426103647, "grad_norm": 0.9456277101482696, "learning_rate": 5.809108347353998e-07, "loss": 0.0249, "step": 13977 }, { "epoch": 8.943058221369162, "grad_norm": 0.6934802490648374, "learning_rate": 5.802150658835626e-07, "loss": 0.0194, "step": 13978 }, { "epoch": 8.943698016634677, "grad_norm": 0.7686678254244063, "learning_rate": 5.79519701500535e-07, "loss": 0.0246, "step": 13979 }, { "epoch": 8.944337811900192, "grad_norm": 0.743656254225919, "learning_rate": 5.78824741616174e-07, "loss": 0.019, "step": 13980 }, { "epoch": 8.944977607165708, "grad_norm": 0.8161428113819857, "learning_rate": 5.781301862603194e-07, "loss": 0.0262, "step": 13981 }, { "epoch": 8.945617402431221, "grad_norm": 0.8832256813691634, "learning_rate": 5.774360354627928e-07, "loss": 0.0235, "step": 13982 }, { "epoch": 8.946257197696736, "grad_norm": 0.7571016931647695, "learning_rate": 5.767422892534036e-07, "loss": 0.0224, "step": 13983 }, { "epoch": 8.946896992962252, "grad_norm": 0.8931730927496321, "learning_rate": 5.760489476619391e-07, "loss": 0.025, "step": 13984 }, { "epoch": 8.947536788227767, "grad_norm": 0.8500225540895688, "learning_rate": 5.753560107181655e-07, "loss": 0.0243, "step": 13985 }, { "epoch": 8.948176583493282, "grad_norm": 0.812931416344132, "learning_rate": 5.746634784518412e-07, "loss": 0.0227, "step": 13986 }, { "epoch": 8.948816378758798, "grad_norm": 0.7819737003567961, "learning_rate": 5.739713508927003e-07, "loss": 0.0225, "step": 13987 }, { "epoch": 8.949456174024313, "grad_norm": 0.8092903272553533, "learning_rate": 5.732796280704634e-07, "loss": 0.0232, "step": 13988 }, { "epoch": 8.950095969289826, "grad_norm": 0.7890441143108095, "learning_rate": 5.725883100148277e-07, "loss": 0.023, "step": 13989 }, { "epoch": 8.950735764555342, "grad_norm": 0.6896973696002578, "learning_rate": 5.71897396755482e-07, "loss": 0.0205, "step": 13990 }, { "epoch": 8.951375559820857, "grad_norm": 0.8224487205699017, "learning_rate": 5.7120688832209e-07, "loss": 0.0228, "step": 13991 }, { "epoch": 8.952015355086372, "grad_norm": 0.8126499665320663, "learning_rate": 5.705167847443016e-07, "loss": 0.0235, "step": 13992 }, { "epoch": 8.952655150351887, "grad_norm": 0.8151683523556837, "learning_rate": 5.698270860517486e-07, "loss": 0.0206, "step": 13993 }, { "epoch": 8.953294945617403, "grad_norm": 0.8056379751000807, "learning_rate": 5.691377922740471e-07, "loss": 0.0242, "step": 13994 }, { "epoch": 8.953934740882918, "grad_norm": 0.7714648566638458, "learning_rate": 5.684489034407903e-07, "loss": 0.0211, "step": 13995 }, { "epoch": 8.954574536148433, "grad_norm": 0.898999350367432, "learning_rate": 5.677604195815612e-07, "loss": 0.0271, "step": 13996 }, { "epoch": 8.955214331413947, "grad_norm": 0.8559392866206468, "learning_rate": 5.670723407259215e-07, "loss": 0.0206, "step": 13997 }, { "epoch": 8.955854126679462, "grad_norm": 0.9585957258678455, "learning_rate": 5.663846669034156e-07, "loss": 0.025, "step": 13998 }, { "epoch": 8.956493921944977, "grad_norm": 0.8413654082863736, "learning_rate": 5.65697398143571e-07, "loss": 0.0248, "step": 13999 }, { "epoch": 8.957133717210493, "grad_norm": 0.68731675979548, "learning_rate": 5.650105344758982e-07, "loss": 0.0201, "step": 14000 }, { "epoch": 8.957773512476008, "grad_norm": 0.8576588408842951, "learning_rate": 5.643240759298896e-07, "loss": 0.0285, "step": 14001 }, { "epoch": 8.958413307741523, "grad_norm": 1.046103046420607, "learning_rate": 5.636380225350179e-07, "loss": 0.027, "step": 14002 }, { "epoch": 8.959053103007038, "grad_norm": 0.9508676801280171, "learning_rate": 5.629523743207454e-07, "loss": 0.028, "step": 14003 }, { "epoch": 8.959692898272554, "grad_norm": 0.7567688508784112, "learning_rate": 5.622671313165117e-07, "loss": 0.0225, "step": 14004 }, { "epoch": 8.960332693538067, "grad_norm": 0.698240427319645, "learning_rate": 5.615822935517368e-07, "loss": 0.0197, "step": 14005 }, { "epoch": 8.960972488803582, "grad_norm": 0.7929415957786884, "learning_rate": 5.608978610558291e-07, "loss": 0.0215, "step": 14006 }, { "epoch": 8.961612284069098, "grad_norm": 0.793162372837343, "learning_rate": 5.602138338581775e-07, "loss": 0.0224, "step": 14007 }, { "epoch": 8.962252079334613, "grad_norm": 0.8437318665533503, "learning_rate": 5.595302119881518e-07, "loss": 0.0214, "step": 14008 }, { "epoch": 8.962891874600128, "grad_norm": 0.8278994699094548, "learning_rate": 5.588469954751041e-07, "loss": 0.0221, "step": 14009 }, { "epoch": 8.963531669865644, "grad_norm": 0.8719506820512504, "learning_rate": 5.581641843483731e-07, "loss": 0.0241, "step": 14010 }, { "epoch": 8.964171465131159, "grad_norm": 0.8722037187235553, "learning_rate": 5.574817786372754e-07, "loss": 0.0242, "step": 14011 }, { "epoch": 8.964811260396672, "grad_norm": 0.719329293795289, "learning_rate": 5.567997783711132e-07, "loss": 0.0193, "step": 14012 }, { "epoch": 8.965451055662188, "grad_norm": 0.91213900584151, "learning_rate": 5.561181835791729e-07, "loss": 0.0263, "step": 14013 }, { "epoch": 8.966090850927703, "grad_norm": 0.7966392438137884, "learning_rate": 5.554369942907167e-07, "loss": 0.0217, "step": 14014 }, { "epoch": 8.966730646193218, "grad_norm": 0.8201196474005169, "learning_rate": 5.547562105349935e-07, "loss": 0.0251, "step": 14015 }, { "epoch": 8.967370441458733, "grad_norm": 0.8344590272169763, "learning_rate": 5.540758323412388e-07, "loss": 0.0233, "step": 14016 }, { "epoch": 8.968010236724249, "grad_norm": 0.8126332049688917, "learning_rate": 5.533958597386646e-07, "loss": 0.025, "step": 14017 }, { "epoch": 8.968650031989764, "grad_norm": 0.8861663760804581, "learning_rate": 5.527162927564689e-07, "loss": 0.0268, "step": 14018 }, { "epoch": 8.969289827255277, "grad_norm": 0.773333922053044, "learning_rate": 5.520371314238282e-07, "loss": 0.0214, "step": 14019 }, { "epoch": 8.969929622520793, "grad_norm": 0.7939639129360552, "learning_rate": 5.51358375769907e-07, "loss": 0.0225, "step": 14020 }, { "epoch": 8.970569417786308, "grad_norm": 0.7150761891641269, "learning_rate": 5.50680025823851e-07, "loss": 0.0194, "step": 14021 }, { "epoch": 8.971209213051823, "grad_norm": 0.7851288499655349, "learning_rate": 5.500020816147833e-07, "loss": 0.0247, "step": 14022 }, { "epoch": 8.971849008317339, "grad_norm": 0.768639948314962, "learning_rate": 5.493245431718164e-07, "loss": 0.0207, "step": 14023 }, { "epoch": 8.972488803582854, "grad_norm": 0.6953648366474017, "learning_rate": 5.486474105240436e-07, "loss": 0.0187, "step": 14024 }, { "epoch": 8.973128598848369, "grad_norm": 0.8984491293495437, "learning_rate": 5.479706837005361e-07, "loss": 0.0257, "step": 14025 }, { "epoch": 8.973768394113884, "grad_norm": 0.7809436165020798, "learning_rate": 5.472943627303562e-07, "loss": 0.0215, "step": 14026 }, { "epoch": 8.974408189379398, "grad_norm": 0.802097522385394, "learning_rate": 5.466184476425396e-07, "loss": 0.0232, "step": 14027 }, { "epoch": 8.975047984644913, "grad_norm": 0.8214911226366393, "learning_rate": 5.459429384661119e-07, "loss": 0.0224, "step": 14028 }, { "epoch": 8.975687779910428, "grad_norm": 0.9177122312856928, "learning_rate": 5.452678352300767e-07, "loss": 0.0234, "step": 14029 }, { "epoch": 8.976327575175944, "grad_norm": 0.7802134632870431, "learning_rate": 5.445931379634218e-07, "loss": 0.0263, "step": 14030 }, { "epoch": 8.976967370441459, "grad_norm": 0.8623351764237956, "learning_rate": 5.439188466951173e-07, "loss": 0.0225, "step": 14031 }, { "epoch": 8.977607165706974, "grad_norm": 0.7140156595245214, "learning_rate": 5.432449614541157e-07, "loss": 0.0178, "step": 14032 }, { "epoch": 8.97824696097249, "grad_norm": 0.8041187323897115, "learning_rate": 5.425714822693539e-07, "loss": 0.0189, "step": 14033 }, { "epoch": 8.978886756238003, "grad_norm": 0.792292918914853, "learning_rate": 5.418984091697499e-07, "loss": 0.022, "step": 14034 }, { "epoch": 8.979526551503518, "grad_norm": 0.7381733512832159, "learning_rate": 5.412257421842027e-07, "loss": 0.022, "step": 14035 }, { "epoch": 8.980166346769034, "grad_norm": 0.8155550234923429, "learning_rate": 5.405534813415969e-07, "loss": 0.0233, "step": 14036 }, { "epoch": 8.980806142034549, "grad_norm": 0.852366236497196, "learning_rate": 5.398816266707973e-07, "loss": 0.0262, "step": 14037 }, { "epoch": 8.981445937300064, "grad_norm": 0.7020650817674875, "learning_rate": 5.39210178200652e-07, "loss": 0.0219, "step": 14038 }, { "epoch": 8.98208573256558, "grad_norm": 0.821741428503128, "learning_rate": 5.385391359599923e-07, "loss": 0.0223, "step": 14039 }, { "epoch": 8.982725527831095, "grad_norm": 0.8143881353676243, "learning_rate": 5.378684999776318e-07, "loss": 0.0226, "step": 14040 }, { "epoch": 8.98336532309661, "grad_norm": 0.7552138030742411, "learning_rate": 5.371982702823664e-07, "loss": 0.0228, "step": 14041 }, { "epoch": 8.984005118362123, "grad_norm": 0.8049484109417622, "learning_rate": 5.365284469029741e-07, "loss": 0.0241, "step": 14042 }, { "epoch": 8.984644913627639, "grad_norm": 0.8970078047730422, "learning_rate": 5.358590298682165e-07, "loss": 0.0208, "step": 14043 }, { "epoch": 8.985284708893154, "grad_norm": 0.7288126265326076, "learning_rate": 5.35190019206836e-07, "loss": 0.0203, "step": 14044 }, { "epoch": 8.98592450415867, "grad_norm": 0.771836064927535, "learning_rate": 5.345214149475586e-07, "loss": 0.023, "step": 14045 }, { "epoch": 8.986564299424185, "grad_norm": 0.7562565056696433, "learning_rate": 5.338532171190946e-07, "loss": 0.0211, "step": 14046 }, { "epoch": 8.9872040946897, "grad_norm": 0.798166717491361, "learning_rate": 5.331854257501357e-07, "loss": 0.0218, "step": 14047 }, { "epoch": 8.987843889955215, "grad_norm": 0.7965540181593029, "learning_rate": 5.325180408693531e-07, "loss": 0.0253, "step": 14048 }, { "epoch": 8.988483685220729, "grad_norm": 0.7921654640782413, "learning_rate": 5.318510625054041e-07, "loss": 0.0228, "step": 14049 }, { "epoch": 8.989123480486244, "grad_norm": 0.8552682024857197, "learning_rate": 5.311844906869291e-07, "loss": 0.0234, "step": 14050 }, { "epoch": 8.98976327575176, "grad_norm": 0.76650059527306, "learning_rate": 5.305183254425484e-07, "loss": 0.0245, "step": 14051 }, { "epoch": 8.990403071017274, "grad_norm": 0.7478305787753142, "learning_rate": 5.298525668008648e-07, "loss": 0.0211, "step": 14052 }, { "epoch": 8.99104286628279, "grad_norm": 0.8433447845514773, "learning_rate": 5.291872147904664e-07, "loss": 0.0222, "step": 14053 }, { "epoch": 8.991682661548305, "grad_norm": 0.7527513595052687, "learning_rate": 5.285222694399228e-07, "loss": 0.0211, "step": 14054 }, { "epoch": 8.99232245681382, "grad_norm": 0.8396306194222826, "learning_rate": 5.278577307777821e-07, "loss": 0.0259, "step": 14055 }, { "epoch": 8.992962252079334, "grad_norm": 0.9642218403021884, "learning_rate": 5.271935988325828e-07, "loss": 0.0244, "step": 14056 }, { "epoch": 8.993602047344849, "grad_norm": 0.8428877760614856, "learning_rate": 5.265298736328395e-07, "loss": 0.0247, "step": 14057 }, { "epoch": 8.994241842610364, "grad_norm": 0.80302286415808, "learning_rate": 5.258665552070508e-07, "loss": 0.0248, "step": 14058 }, { "epoch": 8.99488163787588, "grad_norm": 0.8958323434727808, "learning_rate": 5.252036435836993e-07, "loss": 0.0263, "step": 14059 }, { "epoch": 8.995521433141395, "grad_norm": 0.8606369099854938, "learning_rate": 5.245411387912491e-07, "loss": 0.0228, "step": 14060 }, { "epoch": 8.99616122840691, "grad_norm": 0.7802979970183502, "learning_rate": 5.238790408581474e-07, "loss": 0.0206, "step": 14061 }, { "epoch": 8.996801023672425, "grad_norm": 0.7691788874311141, "learning_rate": 5.232173498128212e-07, "loss": 0.0212, "step": 14062 }, { "epoch": 8.99744081893794, "grad_norm": 0.8642561322213158, "learning_rate": 5.225560656836859e-07, "loss": 0.0248, "step": 14063 }, { "epoch": 8.998080614203454, "grad_norm": 0.8806357614120888, "learning_rate": 5.218951884991331e-07, "loss": 0.0223, "step": 14064 }, { "epoch": 8.99872040946897, "grad_norm": 0.7512930440896225, "learning_rate": 5.21234718287541e-07, "loss": 0.0181, "step": 14065 }, { "epoch": 8.999360204734485, "grad_norm": 0.7854751317427363, "learning_rate": 5.205746550772684e-07, "loss": 0.0226, "step": 14066 }, { "epoch": 9.0, "grad_norm": 0.6897309368748862, "learning_rate": 5.199149988966578e-07, "loss": 0.0168, "step": 14067 }, { "epoch": 9.000639795265515, "grad_norm": 0.36166621545055566, "learning_rate": 5.192557497740336e-07, "loss": 0.0098, "step": 14068 }, { "epoch": 9.00127959053103, "grad_norm": 0.4383800020513596, "learning_rate": 5.185969077377007e-07, "loss": 0.0127, "step": 14069 }, { "epoch": 9.001919385796546, "grad_norm": 0.41263361052017106, "learning_rate": 5.179384728159531e-07, "loss": 0.0107, "step": 14070 }, { "epoch": 9.00255918106206, "grad_norm": 0.4062126824229651, "learning_rate": 5.172804450370594e-07, "loss": 0.0111, "step": 14071 }, { "epoch": 9.003198976327575, "grad_norm": 0.38644307792467164, "learning_rate": 5.166228244292748e-07, "loss": 0.009, "step": 14072 }, { "epoch": 9.00383877159309, "grad_norm": 0.45154894021178066, "learning_rate": 5.159656110208367e-07, "loss": 0.0122, "step": 14073 }, { "epoch": 9.004478566858605, "grad_norm": 0.4088974587626801, "learning_rate": 5.153088048399646e-07, "loss": 0.0098, "step": 14074 }, { "epoch": 9.00511836212412, "grad_norm": 0.4700290409488267, "learning_rate": 5.146524059148605e-07, "loss": 0.012, "step": 14075 }, { "epoch": 9.005758157389636, "grad_norm": 0.4780155866381838, "learning_rate": 5.139964142737064e-07, "loss": 0.0136, "step": 14076 }, { "epoch": 9.006397952655151, "grad_norm": 0.3641120082555291, "learning_rate": 5.133408299446752e-07, "loss": 0.0091, "step": 14077 }, { "epoch": 9.007037747920666, "grad_norm": 0.4216425326989467, "learning_rate": 5.126856529559121e-07, "loss": 0.0115, "step": 14078 }, { "epoch": 9.00767754318618, "grad_norm": 0.44654124119885064, "learning_rate": 5.120308833355492e-07, "loss": 0.0128, "step": 14079 }, { "epoch": 9.008317338451695, "grad_norm": 0.4366652920291208, "learning_rate": 5.113765211117039e-07, "loss": 0.0119, "step": 14080 }, { "epoch": 9.00895713371721, "grad_norm": 0.5494492236095135, "learning_rate": 5.107225663124726e-07, "loss": 0.0128, "step": 14081 }, { "epoch": 9.009596928982726, "grad_norm": 0.3429371734859249, "learning_rate": 5.100690189659319e-07, "loss": 0.0085, "step": 14082 }, { "epoch": 9.01023672424824, "grad_norm": 0.4669520267250999, "learning_rate": 5.09415879100148e-07, "loss": 0.0113, "step": 14083 }, { "epoch": 9.010876519513756, "grad_norm": 0.4213645619157107, "learning_rate": 5.08763146743163e-07, "loss": 0.0082, "step": 14084 }, { "epoch": 9.011516314779271, "grad_norm": 0.5105641416236665, "learning_rate": 5.081108219230058e-07, "loss": 0.0138, "step": 14085 }, { "epoch": 9.012156110044785, "grad_norm": 0.45329533881622824, "learning_rate": 5.074589046676837e-07, "loss": 0.0096, "step": 14086 }, { "epoch": 9.0127959053103, "grad_norm": 0.4829903219310644, "learning_rate": 5.06807395005191e-07, "loss": 0.0109, "step": 14087 }, { "epoch": 9.013435700575815, "grad_norm": 0.4972691990801382, "learning_rate": 5.061562929635022e-07, "loss": 0.0119, "step": 14088 }, { "epoch": 9.01407549584133, "grad_norm": 0.48169125973439675, "learning_rate": 5.055055985705748e-07, "loss": 0.0113, "step": 14089 }, { "epoch": 9.014715291106846, "grad_norm": 0.5079010113103211, "learning_rate": 5.048553118543465e-07, "loss": 0.0094, "step": 14090 }, { "epoch": 9.015355086372361, "grad_norm": 0.5255710845490633, "learning_rate": 5.042054328427404e-07, "loss": 0.0128, "step": 14091 }, { "epoch": 9.015994881637877, "grad_norm": 0.4811278144637688, "learning_rate": 5.035559615636609e-07, "loss": 0.0086, "step": 14092 }, { "epoch": 9.01663467690339, "grad_norm": 0.42054017137793603, "learning_rate": 5.029068980449958e-07, "loss": 0.0089, "step": 14093 }, { "epoch": 9.017274472168905, "grad_norm": 0.445427006841383, "learning_rate": 5.022582423146149e-07, "loss": 0.0101, "step": 14094 }, { "epoch": 9.01791426743442, "grad_norm": 0.4605250090977225, "learning_rate": 5.016099944003705e-07, "loss": 0.008, "step": 14095 }, { "epoch": 9.018554062699936, "grad_norm": 0.5119034632386474, "learning_rate": 5.009621543300947e-07, "loss": 0.0103, "step": 14096 }, { "epoch": 9.019193857965451, "grad_norm": 0.43393751008644493, "learning_rate": 5.003147221316085e-07, "loss": 0.0086, "step": 14097 }, { "epoch": 9.019833653230966, "grad_norm": 0.5230852636294949, "learning_rate": 4.996676978327087e-07, "loss": 0.0098, "step": 14098 }, { "epoch": 9.020473448496482, "grad_norm": 0.5032456806897011, "learning_rate": 4.990210814611773e-07, "loss": 0.0098, "step": 14099 }, { "epoch": 9.021113243761997, "grad_norm": 0.38076618358732134, "learning_rate": 4.983748730447802e-07, "loss": 0.0062, "step": 14100 }, { "epoch": 9.02175303902751, "grad_norm": 0.39853671471031515, "learning_rate": 4.977290726112649e-07, "loss": 0.0077, "step": 14101 }, { "epoch": 9.022392834293026, "grad_norm": 0.4220872828479691, "learning_rate": 4.970836801883594e-07, "loss": 0.0085, "step": 14102 }, { "epoch": 9.023032629558541, "grad_norm": 0.492535888230902, "learning_rate": 4.96438695803777e-07, "loss": 0.0098, "step": 14103 }, { "epoch": 9.023672424824056, "grad_norm": 0.5566152595967028, "learning_rate": 4.957941194852112e-07, "loss": 0.0114, "step": 14104 }, { "epoch": 9.024312220089572, "grad_norm": 0.4937951955850685, "learning_rate": 4.951499512603386e-07, "loss": 0.0099, "step": 14105 }, { "epoch": 9.024952015355087, "grad_norm": 0.5115837594217929, "learning_rate": 4.945061911568183e-07, "loss": 0.0097, "step": 14106 }, { "epoch": 9.025591810620602, "grad_norm": 0.4506372448599822, "learning_rate": 4.938628392022948e-07, "loss": 0.0097, "step": 14107 }, { "epoch": 9.026231605886116, "grad_norm": 0.46017660274620154, "learning_rate": 4.932198954243916e-07, "loss": 0.0088, "step": 14108 }, { "epoch": 9.026871401151631, "grad_norm": 0.4941637548632964, "learning_rate": 4.925773598507122e-07, "loss": 0.0093, "step": 14109 }, { "epoch": 9.027511196417146, "grad_norm": 0.5130486859539837, "learning_rate": 4.9193523250885e-07, "loss": 0.0092, "step": 14110 }, { "epoch": 9.028150991682661, "grad_norm": 0.5417798502764294, "learning_rate": 4.912935134263763e-07, "loss": 0.0104, "step": 14111 }, { "epoch": 9.028790786948177, "grad_norm": 0.49614835679569774, "learning_rate": 4.906522026308425e-07, "loss": 0.009, "step": 14112 }, { "epoch": 9.029430582213692, "grad_norm": 0.5458884958365908, "learning_rate": 4.900113001497885e-07, "loss": 0.0116, "step": 14113 }, { "epoch": 9.030070377479207, "grad_norm": 0.4987269345911806, "learning_rate": 4.893708060107328e-07, "loss": 0.0095, "step": 14114 }, { "epoch": 9.030710172744723, "grad_norm": 0.5533797036614381, "learning_rate": 4.887307202411761e-07, "loss": 0.011, "step": 14115 }, { "epoch": 9.031349968010236, "grad_norm": 0.48512277629221073, "learning_rate": 4.880910428686026e-07, "loss": 0.0089, "step": 14116 }, { "epoch": 9.031989763275751, "grad_norm": 0.49480961595558226, "learning_rate": 4.874517739204798e-07, "loss": 0.0097, "step": 14117 }, { "epoch": 9.032629558541267, "grad_norm": 0.4865221451764412, "learning_rate": 4.86812913424256e-07, "loss": 0.0093, "step": 14118 }, { "epoch": 9.033269353806782, "grad_norm": 0.5536413555990529, "learning_rate": 4.861744614073638e-07, "loss": 0.0105, "step": 14119 }, { "epoch": 9.033909149072297, "grad_norm": 0.5921929834905375, "learning_rate": 4.855364178972166e-07, "loss": 0.0103, "step": 14120 }, { "epoch": 9.034548944337812, "grad_norm": 0.47003661198884755, "learning_rate": 4.848987829212104e-07, "loss": 0.0091, "step": 14121 }, { "epoch": 9.035188739603328, "grad_norm": 0.4509633538334119, "learning_rate": 4.842615565067222e-07, "loss": 0.0083, "step": 14122 }, { "epoch": 9.035828534868841, "grad_norm": 0.4281038953621306, "learning_rate": 4.836247386811177e-07, "loss": 0.0083, "step": 14123 }, { "epoch": 9.036468330134356, "grad_norm": 0.5491323621641067, "learning_rate": 4.829883294717396e-07, "loss": 0.0118, "step": 14124 }, { "epoch": 9.037108125399872, "grad_norm": 0.46721025680533257, "learning_rate": 4.823523289059118e-07, "loss": 0.0087, "step": 14125 }, { "epoch": 9.037747920665387, "grad_norm": 0.557202826886429, "learning_rate": 4.817167370109433e-07, "loss": 0.0112, "step": 14126 }, { "epoch": 9.038387715930902, "grad_norm": 0.5059501116915273, "learning_rate": 4.810815538141278e-07, "loss": 0.0109, "step": 14127 }, { "epoch": 9.039027511196418, "grad_norm": 0.5381955439818087, "learning_rate": 4.804467793427381e-07, "loss": 0.0108, "step": 14128 }, { "epoch": 9.039667306461933, "grad_norm": 0.558195683481968, "learning_rate": 4.798124136240279e-07, "loss": 0.0104, "step": 14129 }, { "epoch": 9.040307101727446, "grad_norm": 0.561229733025883, "learning_rate": 4.791784566852386e-07, "loss": 0.0094, "step": 14130 }, { "epoch": 9.040946896992962, "grad_norm": 0.49358168989901674, "learning_rate": 4.785449085535909e-07, "loss": 0.0097, "step": 14131 }, { "epoch": 9.041586692258477, "grad_norm": 0.48092849337735366, "learning_rate": 4.779117692562873e-07, "loss": 0.008, "step": 14132 }, { "epoch": 9.042226487523992, "grad_norm": 0.5767469788638823, "learning_rate": 4.772790388205128e-07, "loss": 0.0127, "step": 14133 }, { "epoch": 9.042866282789507, "grad_norm": 0.4888019309862991, "learning_rate": 4.7664671727343793e-07, "loss": 0.0099, "step": 14134 }, { "epoch": 9.043506078055023, "grad_norm": 0.5220055727766727, "learning_rate": 4.7601480464221196e-07, "loss": 0.0093, "step": 14135 }, { "epoch": 9.044145873320538, "grad_norm": 0.5081704702046427, "learning_rate": 4.753833009539666e-07, "loss": 0.0101, "step": 14136 }, { "epoch": 9.044785668586053, "grad_norm": 0.5340163299344514, "learning_rate": 4.7475220623582117e-07, "loss": 0.0103, "step": 14137 }, { "epoch": 9.045425463851567, "grad_norm": 0.5517381451708196, "learning_rate": 4.741215205148719e-07, "loss": 0.0121, "step": 14138 }, { "epoch": 9.046065259117082, "grad_norm": 0.5130015286315873, "learning_rate": 4.734912438181982e-07, "loss": 0.0083, "step": 14139 }, { "epoch": 9.046705054382597, "grad_norm": 0.5101791675461009, "learning_rate": 4.728613761728651e-07, "loss": 0.01, "step": 14140 }, { "epoch": 9.047344849648113, "grad_norm": 0.45883843338003116, "learning_rate": 4.7223191760591756e-07, "loss": 0.0082, "step": 14141 }, { "epoch": 9.047984644913628, "grad_norm": 0.48591297030593616, "learning_rate": 4.7160286814438184e-07, "loss": 0.0097, "step": 14142 }, { "epoch": 9.048624440179143, "grad_norm": 0.5120158706956098, "learning_rate": 4.709742278152685e-07, "loss": 0.0106, "step": 14143 }, { "epoch": 9.049264235444658, "grad_norm": 0.4505288970861879, "learning_rate": 4.703459966455726e-07, "loss": 0.0085, "step": 14144 }, { "epoch": 9.049904030710172, "grad_norm": 0.5210857211648388, "learning_rate": 4.697181746622681e-07, "loss": 0.0109, "step": 14145 }, { "epoch": 9.050543825975687, "grad_norm": 0.5939271586720617, "learning_rate": 4.690907618923102e-07, "loss": 0.0115, "step": 14146 }, { "epoch": 9.051183621241202, "grad_norm": 0.48129131546561627, "learning_rate": 4.6846375836264167e-07, "loss": 0.0097, "step": 14147 }, { "epoch": 9.051823416506718, "grad_norm": 0.47682445086795794, "learning_rate": 4.6783716410018443e-07, "loss": 0.0092, "step": 14148 }, { "epoch": 9.052463211772233, "grad_norm": 0.48527592797929253, "learning_rate": 4.6721097913184244e-07, "loss": 0.0079, "step": 14149 }, { "epoch": 9.053103007037748, "grad_norm": 0.5239408024327508, "learning_rate": 4.665852034845042e-07, "loss": 0.0099, "step": 14150 }, { "epoch": 9.053742802303264, "grad_norm": 0.4940806854937515, "learning_rate": 4.6595983718503934e-07, "loss": 0.0091, "step": 14151 }, { "epoch": 9.054382597568779, "grad_norm": 0.5201780505349992, "learning_rate": 4.653348802602986e-07, "loss": 0.0103, "step": 14152 }, { "epoch": 9.055022392834292, "grad_norm": 0.5796008025977742, "learning_rate": 4.64710332737115e-07, "loss": 0.0104, "step": 14153 }, { "epoch": 9.055662188099808, "grad_norm": 0.6128058521296723, "learning_rate": 4.640861946423103e-07, "loss": 0.0092, "step": 14154 }, { "epoch": 9.056301983365323, "grad_norm": 0.5805330129513271, "learning_rate": 4.6346246600268096e-07, "loss": 0.0123, "step": 14155 }, { "epoch": 9.056941778630838, "grad_norm": 0.4710546657861823, "learning_rate": 4.6283914684500775e-07, "loss": 0.0095, "step": 14156 }, { "epoch": 9.057581573896353, "grad_norm": 0.4918197451059168, "learning_rate": 4.6221623719605814e-07, "loss": 0.0079, "step": 14157 }, { "epoch": 9.058221369161869, "grad_norm": 0.5257926474667406, "learning_rate": 4.615937370825763e-07, "loss": 0.0092, "step": 14158 }, { "epoch": 9.058861164427384, "grad_norm": 0.546234564297101, "learning_rate": 4.609716465312908e-07, "loss": 0.0089, "step": 14159 }, { "epoch": 9.059500959692897, "grad_norm": 0.5926304885124595, "learning_rate": 4.6034996556891476e-07, "loss": 0.0103, "step": 14160 }, { "epoch": 9.060140754958413, "grad_norm": 0.5459227619633762, "learning_rate": 4.597286942221424e-07, "loss": 0.0105, "step": 14161 }, { "epoch": 9.060780550223928, "grad_norm": 0.5048758670777742, "learning_rate": 4.59107832517649e-07, "loss": 0.0096, "step": 14162 }, { "epoch": 9.061420345489443, "grad_norm": 0.4337022815322127, "learning_rate": 4.584873804820933e-07, "loss": 0.0077, "step": 14163 }, { "epoch": 9.062060140754959, "grad_norm": 0.47724477173302987, "learning_rate": 4.5786733814211613e-07, "loss": 0.0093, "step": 14164 }, { "epoch": 9.062699936020474, "grad_norm": 0.48719055926012267, "learning_rate": 4.5724770552434186e-07, "loss": 0.0075, "step": 14165 }, { "epoch": 9.06333973128599, "grad_norm": 0.5408613744460613, "learning_rate": 4.566284826553735e-07, "loss": 0.0082, "step": 14166 }, { "epoch": 9.063979526551503, "grad_norm": 0.5031039833344181, "learning_rate": 4.560096695618044e-07, "loss": 0.0094, "step": 14167 }, { "epoch": 9.064619321817018, "grad_norm": 0.41652551323511816, "learning_rate": 4.55391266270202e-07, "loss": 0.0072, "step": 14168 }, { "epoch": 9.065259117082533, "grad_norm": 0.47219340695110784, "learning_rate": 4.547732728071186e-07, "loss": 0.0092, "step": 14169 }, { "epoch": 9.065898912348048, "grad_norm": 0.5394115254847536, "learning_rate": 4.5415568919909283e-07, "loss": 0.009, "step": 14170 }, { "epoch": 9.066538707613564, "grad_norm": 0.47717432043323565, "learning_rate": 4.535385154726413e-07, "loss": 0.0078, "step": 14171 }, { "epoch": 9.067178502879079, "grad_norm": 0.44381064112688257, "learning_rate": 4.5292175165426276e-07, "loss": 0.0087, "step": 14172 }, { "epoch": 9.067818298144594, "grad_norm": 0.46646650158557695, "learning_rate": 4.5230539777044167e-07, "loss": 0.0086, "step": 14173 }, { "epoch": 9.06845809341011, "grad_norm": 0.577674390204781, "learning_rate": 4.516894538476424e-07, "loss": 0.0104, "step": 14174 }, { "epoch": 9.069097888675623, "grad_norm": 0.524161640078905, "learning_rate": 4.5107391991231373e-07, "loss": 0.0094, "step": 14175 }, { "epoch": 9.069737683941138, "grad_norm": 0.5667625420647, "learning_rate": 4.504587959908835e-07, "loss": 0.0093, "step": 14176 }, { "epoch": 9.070377479206654, "grad_norm": 0.5105411256137549, "learning_rate": 4.4984408210976715e-07, "loss": 0.0079, "step": 14177 }, { "epoch": 9.071017274472169, "grad_norm": 0.5433205950545873, "learning_rate": 4.4922977829535694e-07, "loss": 0.0096, "step": 14178 }, { "epoch": 9.071657069737684, "grad_norm": 0.5677851960247987, "learning_rate": 4.4861588457403074e-07, "loss": 0.01, "step": 14179 }, { "epoch": 9.0722968650032, "grad_norm": 0.544839594289418, "learning_rate": 4.480024009721473e-07, "loss": 0.0092, "step": 14180 }, { "epoch": 9.072936660268715, "grad_norm": 0.5492673062678696, "learning_rate": 4.4738932751604903e-07, "loss": 0.0083, "step": 14181 }, { "epoch": 9.073576455534228, "grad_norm": 0.6090284725653587, "learning_rate": 4.4677666423206035e-07, "loss": 0.0113, "step": 14182 }, { "epoch": 9.074216250799743, "grad_norm": 0.569703689702374, "learning_rate": 4.461644111464869e-07, "loss": 0.0084, "step": 14183 }, { "epoch": 9.074856046065259, "grad_norm": 0.4689169987046614, "learning_rate": 4.455525682856199e-07, "loss": 0.0095, "step": 14184 }, { "epoch": 9.075495841330774, "grad_norm": 0.5075153232456991, "learning_rate": 4.449411356757294e-07, "loss": 0.0079, "step": 14185 }, { "epoch": 9.07613563659629, "grad_norm": 0.5585580066447522, "learning_rate": 4.443301133430689e-07, "loss": 0.0106, "step": 14186 }, { "epoch": 9.076775431861805, "grad_norm": 0.5672894303493594, "learning_rate": 4.437195013138751e-07, "loss": 0.0099, "step": 14187 }, { "epoch": 9.07741522712732, "grad_norm": 0.49144085664239356, "learning_rate": 4.4310929961436713e-07, "loss": 0.0083, "step": 14188 }, { "epoch": 9.078055022392835, "grad_norm": 0.5077622377019279, "learning_rate": 4.4249950827074504e-07, "loss": 0.0102, "step": 14189 }, { "epoch": 9.078694817658349, "grad_norm": 0.6777761344694172, "learning_rate": 4.4189012730919247e-07, "loss": 0.0115, "step": 14190 }, { "epoch": 9.079334612923864, "grad_norm": 0.4502349851823397, "learning_rate": 4.412811567558761e-07, "loss": 0.0078, "step": 14191 }, { "epoch": 9.07997440818938, "grad_norm": 0.5452078373189394, "learning_rate": 4.40672596636943e-07, "loss": 0.0092, "step": 14192 }, { "epoch": 9.080614203454894, "grad_norm": 0.5623772238197287, "learning_rate": 4.4006444697852424e-07, "loss": 0.0098, "step": 14193 }, { "epoch": 9.08125399872041, "grad_norm": 0.49530650216745287, "learning_rate": 4.3945670780673245e-07, "loss": 0.0083, "step": 14194 }, { "epoch": 9.081893793985925, "grad_norm": 0.6811055387481543, "learning_rate": 4.3884937914766226e-07, "loss": 0.0113, "step": 14195 }, { "epoch": 9.08253358925144, "grad_norm": 0.74387030422806, "learning_rate": 4.382424610273905e-07, "loss": 0.0127, "step": 14196 }, { "epoch": 9.083173384516954, "grad_norm": 0.474514260036943, "learning_rate": 4.3763595347197875e-07, "loss": 0.0073, "step": 14197 }, { "epoch": 9.083813179782469, "grad_norm": 0.47729016210330766, "learning_rate": 4.370298565074704e-07, "loss": 0.0084, "step": 14198 }, { "epoch": 9.084452975047984, "grad_norm": 0.6162729636896874, "learning_rate": 4.36424170159887e-07, "loss": 0.0093, "step": 14199 }, { "epoch": 9.0850927703135, "grad_norm": 0.47381359999622996, "learning_rate": 4.358188944552388e-07, "loss": 0.0084, "step": 14200 }, { "epoch": 9.085732565579015, "grad_norm": 0.5572426229912273, "learning_rate": 4.3521402941951397e-07, "loss": 0.0108, "step": 14201 }, { "epoch": 9.08637236084453, "grad_norm": 0.5823876614065722, "learning_rate": 4.3460957507868496e-07, "loss": 0.0119, "step": 14202 }, { "epoch": 9.087012156110045, "grad_norm": 0.5877700210882042, "learning_rate": 4.340055314587033e-07, "loss": 0.0103, "step": 14203 }, { "epoch": 9.08765195137556, "grad_norm": 0.6316236882918558, "learning_rate": 4.334018985855082e-07, "loss": 0.008, "step": 14204 }, { "epoch": 9.088291746641074, "grad_norm": 0.4639865900803782, "learning_rate": 4.327986764850189e-07, "loss": 0.0079, "step": 14205 }, { "epoch": 9.08893154190659, "grad_norm": 0.5845092513091712, "learning_rate": 4.321958651831337e-07, "loss": 0.0095, "step": 14206 }, { "epoch": 9.089571337172105, "grad_norm": 0.5071489900604547, "learning_rate": 4.315934647057396e-07, "loss": 0.008, "step": 14207 }, { "epoch": 9.09021113243762, "grad_norm": 0.6147018174097848, "learning_rate": 4.309914750787025e-07, "loss": 0.0097, "step": 14208 }, { "epoch": 9.090850927703135, "grad_norm": 0.5798302961800618, "learning_rate": 4.3038989632786853e-07, "loss": 0.0108, "step": 14209 }, { "epoch": 9.09149072296865, "grad_norm": 0.5964835443967756, "learning_rate": 4.2978872847906916e-07, "loss": 0.0097, "step": 14210 }, { "epoch": 9.092130518234166, "grad_norm": 0.5883385587597347, "learning_rate": 4.291879715581182e-07, "loss": 0.0098, "step": 14211 }, { "epoch": 9.09277031349968, "grad_norm": 0.6595286764822376, "learning_rate": 4.285876255908106e-07, "loss": 0.0104, "step": 14212 }, { "epoch": 9.093410108765195, "grad_norm": 0.5867320471915791, "learning_rate": 4.2798769060292233e-07, "loss": 0.0121, "step": 14213 }, { "epoch": 9.09404990403071, "grad_norm": 0.5129136772366524, "learning_rate": 4.273881666202162e-07, "loss": 0.01, "step": 14214 }, { "epoch": 9.094689699296225, "grad_norm": 0.41688913687083173, "learning_rate": 4.267890536684349e-07, "loss": 0.0066, "step": 14215 }, { "epoch": 9.09532949456174, "grad_norm": 0.4565442733674171, "learning_rate": 4.2619035177330014e-07, "loss": 0.0077, "step": 14216 }, { "epoch": 9.095969289827256, "grad_norm": 0.6354970874952987, "learning_rate": 4.255920609605235e-07, "loss": 0.0094, "step": 14217 }, { "epoch": 9.096609085092771, "grad_norm": 0.5036973810048944, "learning_rate": 4.2499418125579116e-07, "loss": 0.0092, "step": 14218 }, { "epoch": 9.097248880358284, "grad_norm": 0.5553197843595411, "learning_rate": 4.24396712684777e-07, "loss": 0.0096, "step": 14219 }, { "epoch": 9.0978886756238, "grad_norm": 0.5730307947172647, "learning_rate": 4.237996552731327e-07, "loss": 0.0113, "step": 14220 }, { "epoch": 9.098528470889315, "grad_norm": 0.4951717021511951, "learning_rate": 4.232030090464978e-07, "loss": 0.0084, "step": 14221 }, { "epoch": 9.09916826615483, "grad_norm": 0.46575596474999603, "learning_rate": 4.2260677403048953e-07, "loss": 0.0068, "step": 14222 }, { "epoch": 9.099808061420346, "grad_norm": 0.5249740295446281, "learning_rate": 4.220109502507097e-07, "loss": 0.0098, "step": 14223 }, { "epoch": 9.10044785668586, "grad_norm": 0.5749114743577364, "learning_rate": 4.2141553773274113e-07, "loss": 0.0092, "step": 14224 }, { "epoch": 9.101087651951376, "grad_norm": 0.49211584425643057, "learning_rate": 4.2082053650215117e-07, "loss": 0.0082, "step": 14225 }, { "epoch": 9.101727447216891, "grad_norm": 0.5358049438910054, "learning_rate": 4.20225946584486e-07, "loss": 0.0082, "step": 14226 }, { "epoch": 9.102367242482405, "grad_norm": 0.5464480333273749, "learning_rate": 4.196317680052786e-07, "loss": 0.0093, "step": 14227 }, { "epoch": 9.10300703774792, "grad_norm": 0.5161500317329804, "learning_rate": 4.1903800079004073e-07, "loss": 0.011, "step": 14228 }, { "epoch": 9.103646833013435, "grad_norm": 0.5059501860963469, "learning_rate": 4.1844464496426876e-07, "loss": 0.0075, "step": 14229 }, { "epoch": 9.10428662827895, "grad_norm": 0.562843136918876, "learning_rate": 4.178517005534377e-07, "loss": 0.0099, "step": 14230 }, { "epoch": 9.104926423544466, "grad_norm": 0.6242795866949855, "learning_rate": 4.172591675830107e-07, "loss": 0.0102, "step": 14231 }, { "epoch": 9.105566218809981, "grad_norm": 0.4645985650187408, "learning_rate": 4.1666704607842833e-07, "loss": 0.0077, "step": 14232 }, { "epoch": 9.106206014075497, "grad_norm": 0.6228198918409547, "learning_rate": 4.1607533606511484e-07, "loss": 0.0103, "step": 14233 }, { "epoch": 9.10684580934101, "grad_norm": 0.6635466894239176, "learning_rate": 4.154840375684788e-07, "loss": 0.0103, "step": 14234 }, { "epoch": 9.107485604606525, "grad_norm": 0.6084564895781048, "learning_rate": 4.1489315061390864e-07, "loss": 0.012, "step": 14235 }, { "epoch": 9.10812539987204, "grad_norm": 0.4559553654610724, "learning_rate": 4.1430267522677424e-07, "loss": 0.0084, "step": 14236 }, { "epoch": 9.108765195137556, "grad_norm": 0.5399838067935859, "learning_rate": 4.1371261143243304e-07, "loss": 0.0098, "step": 14237 }, { "epoch": 9.109404990403071, "grad_norm": 0.58215580508069, "learning_rate": 4.1312295925622024e-07, "loss": 0.0092, "step": 14238 }, { "epoch": 9.110044785668586, "grad_norm": 0.6501252018595177, "learning_rate": 4.125337187234535e-07, "loss": 0.0086, "step": 14239 }, { "epoch": 9.110684580934102, "grad_norm": 0.46023438473053707, "learning_rate": 4.119448898594336e-07, "loss": 0.0077, "step": 14240 }, { "epoch": 9.111324376199617, "grad_norm": 0.5584515501381752, "learning_rate": 4.113564726894448e-07, "loss": 0.0101, "step": 14241 }, { "epoch": 9.11196417146513, "grad_norm": 0.5884864407600883, "learning_rate": 4.107684672387524e-07, "loss": 0.0103, "step": 14242 }, { "epoch": 9.112603966730646, "grad_norm": 0.48416277467199476, "learning_rate": 4.1018087353260185e-07, "loss": 0.0096, "step": 14243 }, { "epoch": 9.113243761996161, "grad_norm": 0.6130477112251024, "learning_rate": 4.095936915962284e-07, "loss": 0.01, "step": 14244 }, { "epoch": 9.113883557261676, "grad_norm": 0.6447036462026124, "learning_rate": 4.090069214548409e-07, "loss": 0.0132, "step": 14245 }, { "epoch": 9.114523352527192, "grad_norm": 0.5304121739217638, "learning_rate": 4.0842056313363356e-07, "loss": 0.0098, "step": 14246 }, { "epoch": 9.115163147792707, "grad_norm": 0.43240553895892236, "learning_rate": 4.0783461665778623e-07, "loss": 0.007, "step": 14247 }, { "epoch": 9.115802943058222, "grad_norm": 0.7482268907072952, "learning_rate": 4.072490820524588e-07, "loss": 0.0133, "step": 14248 }, { "epoch": 9.116442738323736, "grad_norm": 0.5845957279878821, "learning_rate": 4.0666395934279014e-07, "loss": 0.0093, "step": 14249 }, { "epoch": 9.117082533589251, "grad_norm": 0.5266214212006038, "learning_rate": 4.0607924855390557e-07, "loss": 0.0082, "step": 14250 }, { "epoch": 9.117722328854766, "grad_norm": 0.5074815608806003, "learning_rate": 4.0549494971091175e-07, "loss": 0.0088, "step": 14251 }, { "epoch": 9.118362124120281, "grad_norm": 0.5308133899712518, "learning_rate": 4.049110628388986e-07, "loss": 0.0092, "step": 14252 }, { "epoch": 9.119001919385797, "grad_norm": 0.6120012346024939, "learning_rate": 4.04327587962936e-07, "loss": 0.0101, "step": 14253 }, { "epoch": 9.119641714651312, "grad_norm": 0.5738606772258477, "learning_rate": 4.0374452510807737e-07, "loss": 0.0094, "step": 14254 }, { "epoch": 9.120281509916827, "grad_norm": 0.46198643792369165, "learning_rate": 4.0316187429935816e-07, "loss": 0.0084, "step": 14255 }, { "epoch": 9.12092130518234, "grad_norm": 0.6305283324827949, "learning_rate": 4.02579635561795e-07, "loss": 0.0105, "step": 14256 }, { "epoch": 9.121561100447856, "grad_norm": 0.5060597939179945, "learning_rate": 4.019978089203924e-07, "loss": 0.0085, "step": 14257 }, { "epoch": 9.122200895713371, "grad_norm": 0.505408611043416, "learning_rate": 4.0141639440012926e-07, "loss": 0.0086, "step": 14258 }, { "epoch": 9.122840690978887, "grad_norm": 0.8335594908082845, "learning_rate": 4.008353920259711e-07, "loss": 0.0088, "step": 14259 }, { "epoch": 9.123480486244402, "grad_norm": 0.5370157403906416, "learning_rate": 4.002548018228658e-07, "loss": 0.0068, "step": 14260 }, { "epoch": 9.124120281509917, "grad_norm": 0.5301606985279558, "learning_rate": 3.996746238157434e-07, "loss": 0.0075, "step": 14261 }, { "epoch": 9.124760076775432, "grad_norm": 0.5701523470454174, "learning_rate": 3.9909485802951397e-07, "loss": 0.0087, "step": 14262 }, { "epoch": 9.125399872040948, "grad_norm": 0.5328407265121945, "learning_rate": 3.9851550448907316e-07, "loss": 0.0091, "step": 14263 }, { "epoch": 9.126039667306461, "grad_norm": 0.43252665089407694, "learning_rate": 3.979365632192966e-07, "loss": 0.0063, "step": 14264 }, { "epoch": 9.126679462571976, "grad_norm": 0.48124000790031146, "learning_rate": 3.973580342450445e-07, "loss": 0.0075, "step": 14265 }, { "epoch": 9.127319257837492, "grad_norm": 0.5605929400276467, "learning_rate": 3.967799175911546e-07, "loss": 0.01, "step": 14266 }, { "epoch": 9.127959053103007, "grad_norm": 0.6344197914708818, "learning_rate": 3.9620221328245277e-07, "loss": 0.0118, "step": 14267 }, { "epoch": 9.128598848368522, "grad_norm": 0.49054241775219865, "learning_rate": 3.9562492134374577e-07, "loss": 0.0085, "step": 14268 }, { "epoch": 9.129238643634038, "grad_norm": 0.5067625042605916, "learning_rate": 3.9504804179981813e-07, "loss": 0.0089, "step": 14269 }, { "epoch": 9.129878438899553, "grad_norm": 0.6300925085194877, "learning_rate": 3.944715746754435e-07, "loss": 0.0098, "step": 14270 }, { "epoch": 9.130518234165066, "grad_norm": 0.48666531258612195, "learning_rate": 3.938955199953709e-07, "loss": 0.0075, "step": 14271 }, { "epoch": 9.131158029430582, "grad_norm": 0.5636598646976388, "learning_rate": 3.933198777843372e-07, "loss": 0.0102, "step": 14272 }, { "epoch": 9.131797824696097, "grad_norm": 0.5703976718676367, "learning_rate": 3.9274464806705826e-07, "loss": 0.0092, "step": 14273 }, { "epoch": 9.132437619961612, "grad_norm": 0.44666827950794036, "learning_rate": 3.9216983086823536e-07, "loss": 0.007, "step": 14274 }, { "epoch": 9.133077415227127, "grad_norm": 0.6220532635862784, "learning_rate": 3.9159542621254875e-07, "loss": 0.0119, "step": 14275 }, { "epoch": 9.133717210492643, "grad_norm": 0.6292231202347845, "learning_rate": 3.91021434124661e-07, "loss": 0.0093, "step": 14276 }, { "epoch": 9.134357005758158, "grad_norm": 0.5707373987092008, "learning_rate": 3.904478546292212e-07, "loss": 0.0097, "step": 14277 }, { "epoch": 9.134996801023673, "grad_norm": 0.49449064286158806, "learning_rate": 3.8987468775085637e-07, "loss": 0.0083, "step": 14278 }, { "epoch": 9.135636596289187, "grad_norm": 0.5931770199194991, "learning_rate": 3.893019335141779e-07, "loss": 0.0096, "step": 14279 }, { "epoch": 9.136276391554702, "grad_norm": 0.45648447498923467, "learning_rate": 3.8872959194377614e-07, "loss": 0.0075, "step": 14280 }, { "epoch": 9.136916186820217, "grad_norm": 0.5554202791547163, "learning_rate": 3.881576630642303e-07, "loss": 0.0091, "step": 14281 }, { "epoch": 9.137555982085733, "grad_norm": 0.5101828710405328, "learning_rate": 3.875861469000963e-07, "loss": 0.0083, "step": 14282 }, { "epoch": 9.138195777351248, "grad_norm": 0.5467422416861505, "learning_rate": 3.8701504347591457e-07, "loss": 0.0095, "step": 14283 }, { "epoch": 9.138835572616763, "grad_norm": 0.5519052408453928, "learning_rate": 3.864443528162054e-07, "loss": 0.0087, "step": 14284 }, { "epoch": 9.139475367882278, "grad_norm": 0.6136536194477604, "learning_rate": 3.858740749454759e-07, "loss": 0.0118, "step": 14285 }, { "epoch": 9.140115163147792, "grad_norm": 0.43155056734286834, "learning_rate": 3.853042098882087e-07, "loss": 0.0066, "step": 14286 }, { "epoch": 9.140754958413307, "grad_norm": 0.5106766131474065, "learning_rate": 3.847347576688776e-07, "loss": 0.0089, "step": 14287 }, { "epoch": 9.141394753678822, "grad_norm": 0.4746522768478201, "learning_rate": 3.8416571831193186e-07, "loss": 0.0077, "step": 14288 }, { "epoch": 9.142034548944338, "grad_norm": 0.571334182549356, "learning_rate": 3.835970918418053e-07, "loss": 0.0088, "step": 14289 }, { "epoch": 9.142674344209853, "grad_norm": 0.6281113471998968, "learning_rate": 3.8302887828291167e-07, "loss": 0.0098, "step": 14290 }, { "epoch": 9.143314139475368, "grad_norm": 0.5738931599496954, "learning_rate": 3.8246107765965157e-07, "loss": 0.0099, "step": 14291 }, { "epoch": 9.143953934740884, "grad_norm": 0.5340857317674202, "learning_rate": 3.8189368999640544e-07, "loss": 0.0102, "step": 14292 }, { "epoch": 9.144593730006397, "grad_norm": 0.509087205505668, "learning_rate": 3.813267153175326e-07, "loss": 0.0081, "step": 14293 }, { "epoch": 9.145233525271912, "grad_norm": 0.6208418649623921, "learning_rate": 3.807601536473826e-07, "loss": 0.0117, "step": 14294 }, { "epoch": 9.145873320537428, "grad_norm": 0.5020998078336094, "learning_rate": 3.8019400501027926e-07, "loss": 0.0085, "step": 14295 }, { "epoch": 9.146513115802943, "grad_norm": 0.6251569197064442, "learning_rate": 3.7962826943053423e-07, "loss": 0.0107, "step": 14296 }, { "epoch": 9.147152911068458, "grad_norm": 0.5476853342252593, "learning_rate": 3.7906294693243584e-07, "loss": 0.0083, "step": 14297 }, { "epoch": 9.147792706333973, "grad_norm": 0.5428919383755985, "learning_rate": 3.7849803754026136e-07, "loss": 0.0113, "step": 14298 }, { "epoch": 9.148432501599489, "grad_norm": 0.6476755706529016, "learning_rate": 3.77933541278267e-07, "loss": 0.0079, "step": 14299 }, { "epoch": 9.149072296865004, "grad_norm": 0.597318338551763, "learning_rate": 3.7736945817068883e-07, "loss": 0.0094, "step": 14300 }, { "epoch": 9.149712092130518, "grad_norm": 0.5349277512702254, "learning_rate": 3.7680578824174973e-07, "loss": 0.009, "step": 14301 }, { "epoch": 9.150351887396033, "grad_norm": 0.4975898284266324, "learning_rate": 3.762425315156504e-07, "loss": 0.0081, "step": 14302 }, { "epoch": 9.150991682661548, "grad_norm": 0.5563846033788795, "learning_rate": 3.7567968801657693e-07, "loss": 0.0097, "step": 14303 }, { "epoch": 9.151631477927063, "grad_norm": 0.6213606220718918, "learning_rate": 3.75117257768699e-07, "loss": 0.0094, "step": 14304 }, { "epoch": 9.152271273192579, "grad_norm": 0.4820676217570666, "learning_rate": 3.745552407961639e-07, "loss": 0.0077, "step": 14305 }, { "epoch": 9.152911068458094, "grad_norm": 0.5197020482539365, "learning_rate": 3.7399363712310564e-07, "loss": 0.0088, "step": 14306 }, { "epoch": 9.15355086372361, "grad_norm": 0.5179026425354644, "learning_rate": 3.73432446773635e-07, "loss": 0.0078, "step": 14307 }, { "epoch": 9.154190658989123, "grad_norm": 0.44884524876192466, "learning_rate": 3.7287166977185263e-07, "loss": 0.0072, "step": 14308 }, { "epoch": 9.154830454254638, "grad_norm": 0.5301407239067644, "learning_rate": 3.723113061418349e-07, "loss": 0.0078, "step": 14309 }, { "epoch": 9.155470249520153, "grad_norm": 0.5565052845719141, "learning_rate": 3.7175135590764246e-07, "loss": 0.0103, "step": 14310 }, { "epoch": 9.156110044785668, "grad_norm": 0.47802872864204904, "learning_rate": 3.711918190933206e-07, "loss": 0.0073, "step": 14311 }, { "epoch": 9.156749840051184, "grad_norm": 0.5565612773030053, "learning_rate": 3.706326957228934e-07, "loss": 0.009, "step": 14312 }, { "epoch": 9.157389635316699, "grad_norm": 0.5787434822880397, "learning_rate": 3.7007398582036833e-07, "loss": 0.0094, "step": 14313 }, { "epoch": 9.158029430582214, "grad_norm": 0.5910666889588949, "learning_rate": 3.695156894097374e-07, "loss": 0.008, "step": 14314 }, { "epoch": 9.15866922584773, "grad_norm": 0.620560483264095, "learning_rate": 3.689578065149713e-07, "loss": 0.0098, "step": 14315 }, { "epoch": 9.159309021113243, "grad_norm": 0.6179204155006022, "learning_rate": 3.684003371600242e-07, "loss": 0.0107, "step": 14316 }, { "epoch": 9.159948816378758, "grad_norm": 0.5567666048424614, "learning_rate": 3.6784328136883153e-07, "loss": 0.0097, "step": 14317 }, { "epoch": 9.160588611644274, "grad_norm": 0.651645721522286, "learning_rate": 3.6728663916531625e-07, "loss": 0.0106, "step": 14318 }, { "epoch": 9.161228406909789, "grad_norm": 0.5978234087996382, "learning_rate": 3.667304105733771e-07, "loss": 0.0096, "step": 14319 }, { "epoch": 9.161868202175304, "grad_norm": 0.6275386684583097, "learning_rate": 3.6617459561689605e-07, "loss": 0.0092, "step": 14320 }, { "epoch": 9.16250799744082, "grad_norm": 0.4978615798913754, "learning_rate": 3.656191943197429e-07, "loss": 0.0084, "step": 14321 }, { "epoch": 9.163147792706335, "grad_norm": 0.5060463652943316, "learning_rate": 3.65064206705763e-07, "loss": 0.0085, "step": 14322 }, { "epoch": 9.163787587971848, "grad_norm": 0.4540507504549765, "learning_rate": 3.6450963279878517e-07, "loss": 0.0067, "step": 14323 }, { "epoch": 9.164427383237364, "grad_norm": 0.535834590287414, "learning_rate": 3.639554726226258e-07, "loss": 0.0087, "step": 14324 }, { "epoch": 9.165067178502879, "grad_norm": 0.5825317055790319, "learning_rate": 3.6340172620107583e-07, "loss": 0.0095, "step": 14325 }, { "epoch": 9.165706973768394, "grad_norm": 0.5446069529302253, "learning_rate": 3.628483935579141e-07, "loss": 0.0092, "step": 14326 }, { "epoch": 9.16634676903391, "grad_norm": 0.521921087470497, "learning_rate": 3.622954747168983e-07, "loss": 0.0082, "step": 14327 }, { "epoch": 9.166986564299425, "grad_norm": 0.6578677016673706, "learning_rate": 3.617429697017716e-07, "loss": 0.0106, "step": 14328 }, { "epoch": 9.16762635956494, "grad_norm": 0.5158949520826143, "learning_rate": 3.611908785362572e-07, "loss": 0.0083, "step": 14329 }, { "epoch": 9.168266154830455, "grad_norm": 0.5559672933990398, "learning_rate": 3.606392012440596e-07, "loss": 0.0087, "step": 14330 }, { "epoch": 9.168905950095969, "grad_norm": 0.5385847426483508, "learning_rate": 3.6008793784886754e-07, "loss": 0.0088, "step": 14331 }, { "epoch": 9.169545745361484, "grad_norm": 0.5249224075607875, "learning_rate": 3.5953708837435206e-07, "loss": 0.0086, "step": 14332 }, { "epoch": 9.170185540627, "grad_norm": 0.6052821028640332, "learning_rate": 3.5898665284416324e-07, "loss": 0.0099, "step": 14333 }, { "epoch": 9.170825335892514, "grad_norm": 0.46124186097108355, "learning_rate": 3.584366312819387e-07, "loss": 0.0084, "step": 14334 }, { "epoch": 9.17146513115803, "grad_norm": 0.48161935050118054, "learning_rate": 3.5788702371129416e-07, "loss": 0.0086, "step": 14335 }, { "epoch": 9.172104926423545, "grad_norm": 0.5840542893909633, "learning_rate": 3.5733783015582946e-07, "loss": 0.0099, "step": 14336 }, { "epoch": 9.17274472168906, "grad_norm": 0.6680093913702345, "learning_rate": 3.5678905063912363e-07, "loss": 0.0111, "step": 14337 }, { "epoch": 9.173384516954574, "grad_norm": 0.6902920558517333, "learning_rate": 3.562406851847433e-07, "loss": 0.0088, "step": 14338 }, { "epoch": 9.174024312220089, "grad_norm": 0.5619610550416889, "learning_rate": 3.55692733816233e-07, "loss": 0.0082, "step": 14339 }, { "epoch": 9.174664107485604, "grad_norm": 0.511610984092232, "learning_rate": 3.5514519655711955e-07, "loss": 0.0093, "step": 14340 }, { "epoch": 9.17530390275112, "grad_norm": 0.5200483864076137, "learning_rate": 3.545980734309162e-07, "loss": 0.009, "step": 14341 }, { "epoch": 9.175943698016635, "grad_norm": 0.5647619371105072, "learning_rate": 3.540513644611132e-07, "loss": 0.008, "step": 14342 }, { "epoch": 9.17658349328215, "grad_norm": 0.464874324899159, "learning_rate": 3.5350506967118615e-07, "loss": 0.0072, "step": 14343 }, { "epoch": 9.177223288547665, "grad_norm": 0.5670138620850398, "learning_rate": 3.5295918908459183e-07, "loss": 0.009, "step": 14344 }, { "epoch": 9.177863083813179, "grad_norm": 0.48363961729135524, "learning_rate": 3.5241372272476815e-07, "loss": 0.0079, "step": 14345 }, { "epoch": 9.178502879078694, "grad_norm": 0.607373610111977, "learning_rate": 3.5186867061513865e-07, "loss": 0.0102, "step": 14346 }, { "epoch": 9.17914267434421, "grad_norm": 0.49329938165018555, "learning_rate": 3.513240327791034e-07, "loss": 0.0075, "step": 14347 }, { "epoch": 9.179782469609725, "grad_norm": 0.5422834089289224, "learning_rate": 3.507798092400527e-07, "loss": 0.0094, "step": 14348 }, { "epoch": 9.18042226487524, "grad_norm": 0.5943920833827502, "learning_rate": 3.5023600002135226e-07, "loss": 0.0104, "step": 14349 }, { "epoch": 9.181062060140755, "grad_norm": 0.5617771204598508, "learning_rate": 3.4969260514635006e-07, "loss": 0.0099, "step": 14350 }, { "epoch": 9.18170185540627, "grad_norm": 0.593632008717209, "learning_rate": 3.49149624638383e-07, "loss": 0.0119, "step": 14351 }, { "epoch": 9.182341650671786, "grad_norm": 0.5664645729771501, "learning_rate": 3.486070585207635e-07, "loss": 0.0095, "step": 14352 }, { "epoch": 9.1829814459373, "grad_norm": 0.5528148155388982, "learning_rate": 3.480649068167874e-07, "loss": 0.0077, "step": 14353 }, { "epoch": 9.183621241202815, "grad_norm": 0.600235388750073, "learning_rate": 3.4752316954973607e-07, "loss": 0.0097, "step": 14354 }, { "epoch": 9.18426103646833, "grad_norm": 0.5715362419843724, "learning_rate": 3.4698184674286873e-07, "loss": 0.0091, "step": 14355 }, { "epoch": 9.184900831733845, "grad_norm": 0.6050287802091988, "learning_rate": 3.4644093841943005e-07, "loss": 0.0112, "step": 14356 }, { "epoch": 9.18554062699936, "grad_norm": 0.5436357209896316, "learning_rate": 3.4590044460264374e-07, "loss": 0.0082, "step": 14357 }, { "epoch": 9.186180422264876, "grad_norm": 0.5692154244022161, "learning_rate": 3.4536036531572006e-07, "loss": 0.0098, "step": 14358 }, { "epoch": 9.186820217530391, "grad_norm": 0.49253850757663736, "learning_rate": 3.4482070058184825e-07, "loss": 0.0076, "step": 14359 }, { "epoch": 9.187460012795905, "grad_norm": 0.549578333586989, "learning_rate": 3.442814504241998e-07, "loss": 0.0087, "step": 14360 }, { "epoch": 9.18809980806142, "grad_norm": 0.49545614714898895, "learning_rate": 3.4374261486592955e-07, "loss": 0.0087, "step": 14361 }, { "epoch": 9.188739603326935, "grad_norm": 0.5835775309923399, "learning_rate": 3.4320419393017555e-07, "loss": 0.0094, "step": 14362 }, { "epoch": 9.18937939859245, "grad_norm": 0.5251539793744471, "learning_rate": 3.426661876400528e-07, "loss": 0.0076, "step": 14363 }, { "epoch": 9.190019193857966, "grad_norm": 0.44538518845026054, "learning_rate": 3.421285960186671e-07, "loss": 0.0074, "step": 14364 }, { "epoch": 9.190658989123481, "grad_norm": 0.5335338602022873, "learning_rate": 3.4159141908909896e-07, "loss": 0.0089, "step": 14365 }, { "epoch": 9.191298784388996, "grad_norm": 0.5350798802010823, "learning_rate": 3.410546568744144e-07, "loss": 0.0098, "step": 14366 }, { "epoch": 9.191938579654511, "grad_norm": 0.6222444376768431, "learning_rate": 3.405183093976594e-07, "loss": 0.0093, "step": 14367 }, { "epoch": 9.192578374920025, "grad_norm": 0.5307127375738305, "learning_rate": 3.399823766818666e-07, "loss": 0.0079, "step": 14368 }, { "epoch": 9.19321817018554, "grad_norm": 0.6378253921028574, "learning_rate": 3.394468587500477e-07, "loss": 0.0089, "step": 14369 }, { "epoch": 9.193857965451055, "grad_norm": 0.6244858072868185, "learning_rate": 3.389117556251942e-07, "loss": 0.0095, "step": 14370 }, { "epoch": 9.19449776071657, "grad_norm": 0.5397262192841736, "learning_rate": 3.383770673302844e-07, "loss": 0.0092, "step": 14371 }, { "epoch": 9.195137555982086, "grad_norm": 0.6690530045769678, "learning_rate": 3.3784279388827777e-07, "loss": 0.0099, "step": 14372 }, { "epoch": 9.195777351247601, "grad_norm": 0.5713624142780911, "learning_rate": 3.373089353221137e-07, "loss": 0.01, "step": 14373 }, { "epoch": 9.196417146513117, "grad_norm": 0.504102469352542, "learning_rate": 3.36775491654715e-07, "loss": 0.0076, "step": 14374 }, { "epoch": 9.19705694177863, "grad_norm": 0.552202450345911, "learning_rate": 3.3624246290898667e-07, "loss": 0.0073, "step": 14375 }, { "epoch": 9.197696737044145, "grad_norm": 0.4807383578743824, "learning_rate": 3.35709849107817e-07, "loss": 0.007, "step": 14376 }, { "epoch": 9.19833653230966, "grad_norm": 0.4653463184136512, "learning_rate": 3.3517765027407446e-07, "loss": 0.0076, "step": 14377 }, { "epoch": 9.198976327575176, "grad_norm": 0.634856191688881, "learning_rate": 3.346458664306118e-07, "loss": 0.0076, "step": 14378 }, { "epoch": 9.199616122840691, "grad_norm": 0.6659687550990212, "learning_rate": 3.3411449760026194e-07, "loss": 0.0122, "step": 14379 }, { "epoch": 9.200255918106206, "grad_norm": 0.44236794683775715, "learning_rate": 3.3358354380584103e-07, "loss": 0.0071, "step": 14380 }, { "epoch": 9.200895713371722, "grad_norm": 0.5720247329149357, "learning_rate": 3.330530050701486e-07, "loss": 0.009, "step": 14381 }, { "epoch": 9.201535508637235, "grad_norm": 0.49769528912266764, "learning_rate": 3.325228814159631e-07, "loss": 0.0079, "step": 14382 }, { "epoch": 9.20217530390275, "grad_norm": 0.5442322841537617, "learning_rate": 3.319931728660486e-07, "loss": 0.0094, "step": 14383 }, { "epoch": 9.202815099168266, "grad_norm": 0.6364684517032415, "learning_rate": 3.3146387944314796e-07, "loss": 0.0109, "step": 14384 }, { "epoch": 9.203454894433781, "grad_norm": 0.5223776109259248, "learning_rate": 3.3093500116998966e-07, "loss": 0.0099, "step": 14385 }, { "epoch": 9.204094689699296, "grad_norm": 0.6034083981001225, "learning_rate": 3.3040653806928337e-07, "loss": 0.0111, "step": 14386 }, { "epoch": 9.204734484964812, "grad_norm": 0.48056664764081775, "learning_rate": 3.298784901637186e-07, "loss": 0.0073, "step": 14387 }, { "epoch": 9.205374280230327, "grad_norm": 0.5309920872784477, "learning_rate": 3.2935085747597074e-07, "loss": 0.0087, "step": 14388 }, { "epoch": 9.206014075495842, "grad_norm": 0.62001610571225, "learning_rate": 3.288236400286937e-07, "loss": 0.0097, "step": 14389 }, { "epoch": 9.206653870761356, "grad_norm": 0.49700953483212584, "learning_rate": 3.2829683784452616e-07, "loss": 0.0076, "step": 14390 }, { "epoch": 9.207293666026871, "grad_norm": 0.5763256783225394, "learning_rate": 3.277704509460877e-07, "loss": 0.009, "step": 14391 }, { "epoch": 9.207933461292386, "grad_norm": 0.526818198175504, "learning_rate": 3.2724447935598037e-07, "loss": 0.0091, "step": 14392 }, { "epoch": 9.208573256557901, "grad_norm": 0.5697528228264426, "learning_rate": 3.267189230967893e-07, "loss": 0.0094, "step": 14393 }, { "epoch": 9.209213051823417, "grad_norm": 0.6044765650743579, "learning_rate": 3.2619378219107766e-07, "loss": 0.0095, "step": 14394 }, { "epoch": 9.209852847088932, "grad_norm": 0.5062960623969975, "learning_rate": 3.256690566613996e-07, "loss": 0.0091, "step": 14395 }, { "epoch": 9.210492642354447, "grad_norm": 0.5881375358321982, "learning_rate": 3.251447465302815e-07, "loss": 0.0098, "step": 14396 }, { "epoch": 9.21113243761996, "grad_norm": 0.48608131463243387, "learning_rate": 3.246208518202376e-07, "loss": 0.007, "step": 14397 }, { "epoch": 9.211772232885476, "grad_norm": 0.5383348902500226, "learning_rate": 3.240973725537633e-07, "loss": 0.0082, "step": 14398 }, { "epoch": 9.212412028150991, "grad_norm": 0.4886391011070902, "learning_rate": 3.2357430875333606e-07, "loss": 0.0083, "step": 14399 }, { "epoch": 9.213051823416507, "grad_norm": 0.47814470415501237, "learning_rate": 3.2305166044141466e-07, "loss": 0.0069, "step": 14400 }, { "epoch": 9.213691618682022, "grad_norm": 0.583808357814957, "learning_rate": 3.2252942764044114e-07, "loss": 0.009, "step": 14401 }, { "epoch": 9.214331413947537, "grad_norm": 0.6215719976677642, "learning_rate": 3.220076103728398e-07, "loss": 0.0105, "step": 14402 }, { "epoch": 9.214971209213052, "grad_norm": 0.5911327358878241, "learning_rate": 3.21486208661016e-07, "loss": 0.0101, "step": 14403 }, { "epoch": 9.215611004478568, "grad_norm": 0.5509051432452466, "learning_rate": 3.209652225273563e-07, "loss": 0.0088, "step": 14404 }, { "epoch": 9.216250799744081, "grad_norm": 0.5730325328336634, "learning_rate": 3.204446519942339e-07, "loss": 0.0089, "step": 14405 }, { "epoch": 9.216890595009597, "grad_norm": 0.5197698485598478, "learning_rate": 3.1992449708399875e-07, "loss": 0.0092, "step": 14406 }, { "epoch": 9.217530390275112, "grad_norm": 0.5425344457216792, "learning_rate": 3.1940475781898516e-07, "loss": 0.0079, "step": 14407 }, { "epoch": 9.218170185540627, "grad_norm": 0.5771951245243895, "learning_rate": 3.1888543422151195e-07, "loss": 0.0094, "step": 14408 }, { "epoch": 9.218809980806142, "grad_norm": 0.48075729885171364, "learning_rate": 3.1836652631387686e-07, "loss": 0.0077, "step": 14409 }, { "epoch": 9.219449776071658, "grad_norm": 0.5387722970326073, "learning_rate": 3.1784803411835985e-07, "loss": 0.0075, "step": 14410 }, { "epoch": 9.220089571337173, "grad_norm": 0.5410157244962842, "learning_rate": 3.173299576572253e-07, "loss": 0.0092, "step": 14411 }, { "epoch": 9.220729366602686, "grad_norm": 0.6067013897018338, "learning_rate": 3.1681229695271985e-07, "loss": 0.01, "step": 14412 }, { "epoch": 9.221369161868202, "grad_norm": 0.5325769922594957, "learning_rate": 3.16295052027068e-07, "loss": 0.0082, "step": 14413 }, { "epoch": 9.222008957133717, "grad_norm": 0.5462518657176756, "learning_rate": 3.157782229024797e-07, "loss": 0.0082, "step": 14414 }, { "epoch": 9.222648752399232, "grad_norm": 0.5682320238671728, "learning_rate": 3.152618096011495e-07, "loss": 0.0084, "step": 14415 }, { "epoch": 9.223288547664747, "grad_norm": 0.5046720887788344, "learning_rate": 3.1474581214524845e-07, "loss": 0.0079, "step": 14416 }, { "epoch": 9.223928342930263, "grad_norm": 0.603353731584828, "learning_rate": 3.1423023055693334e-07, "loss": 0.0107, "step": 14417 }, { "epoch": 9.224568138195778, "grad_norm": 0.7259652119840788, "learning_rate": 3.13715064858342e-07, "loss": 0.0103, "step": 14418 }, { "epoch": 9.225207933461292, "grad_norm": 0.4669716967799561, "learning_rate": 3.132003150715968e-07, "loss": 0.007, "step": 14419 }, { "epoch": 9.225847728726807, "grad_norm": 0.6269503433723175, "learning_rate": 3.1268598121879767e-07, "loss": 0.0109, "step": 14420 }, { "epoch": 9.226487523992322, "grad_norm": 0.5773707539146289, "learning_rate": 3.1217206332203043e-07, "loss": 0.0092, "step": 14421 }, { "epoch": 9.227127319257837, "grad_norm": 0.523576860977294, "learning_rate": 3.116585614033618e-07, "loss": 0.0071, "step": 14422 }, { "epoch": 9.227767114523353, "grad_norm": 0.48338708368702293, "learning_rate": 3.111454754848398e-07, "loss": 0.0081, "step": 14423 }, { "epoch": 9.228406909788868, "grad_norm": 0.5887148322188795, "learning_rate": 3.106328055884955e-07, "loss": 0.0093, "step": 14424 }, { "epoch": 9.229046705054383, "grad_norm": 0.638875469056594, "learning_rate": 3.101205517363426e-07, "loss": 0.0097, "step": 14425 }, { "epoch": 9.229686500319898, "grad_norm": 0.53659377668046, "learning_rate": 3.096087139503778e-07, "loss": 0.0094, "step": 14426 }, { "epoch": 9.230326295585412, "grad_norm": 0.5516224598506951, "learning_rate": 3.090972922525748e-07, "loss": 0.0082, "step": 14427 }, { "epoch": 9.230966090850927, "grad_norm": 0.5387935842640204, "learning_rate": 3.085862866648959e-07, "loss": 0.0091, "step": 14428 }, { "epoch": 9.231605886116443, "grad_norm": 0.5416079546979848, "learning_rate": 3.080756972092824e-07, "loss": 0.0087, "step": 14429 }, { "epoch": 9.232245681381958, "grad_norm": 0.564198472131026, "learning_rate": 3.07565523907658e-07, "loss": 0.009, "step": 14430 }, { "epoch": 9.232885476647473, "grad_norm": 0.48479806000163916, "learning_rate": 3.070557667819285e-07, "loss": 0.0074, "step": 14431 }, { "epoch": 9.233525271912988, "grad_norm": 0.565900866603907, "learning_rate": 3.0654642585398185e-07, "loss": 0.0081, "step": 14432 }, { "epoch": 9.234165067178504, "grad_norm": 0.5844721576101326, "learning_rate": 3.0603750114568844e-07, "loss": 0.0104, "step": 14433 }, { "epoch": 9.234804862444017, "grad_norm": 0.5724748059090694, "learning_rate": 3.0552899267890067e-07, "loss": 0.0084, "step": 14434 }, { "epoch": 9.235444657709532, "grad_norm": 0.4945554508167761, "learning_rate": 3.050209004754534e-07, "loss": 0.0071, "step": 14435 }, { "epoch": 9.236084452975048, "grad_norm": 0.5846351717309382, "learning_rate": 3.0451322455716246e-07, "loss": 0.0091, "step": 14436 }, { "epoch": 9.236724248240563, "grad_norm": 0.5064069030944431, "learning_rate": 3.040059649458249e-07, "loss": 0.0074, "step": 14437 }, { "epoch": 9.237364043506078, "grad_norm": 0.518243102503188, "learning_rate": 3.034991216632244e-07, "loss": 0.0068, "step": 14438 }, { "epoch": 9.238003838771593, "grad_norm": 0.5349090499185518, "learning_rate": 3.029926947311235e-07, "loss": 0.0075, "step": 14439 }, { "epoch": 9.238643634037109, "grad_norm": 0.5361679319790499, "learning_rate": 3.024866841712659e-07, "loss": 0.0096, "step": 14440 }, { "epoch": 9.239283429302624, "grad_norm": 0.5967938275597298, "learning_rate": 3.019810900053799e-07, "loss": 0.0117, "step": 14441 }, { "epoch": 9.239923224568138, "grad_norm": 0.5995321530943344, "learning_rate": 3.014759122551747e-07, "loss": 0.0104, "step": 14442 }, { "epoch": 9.240563019833653, "grad_norm": 0.5590191934065362, "learning_rate": 3.0097115094234185e-07, "loss": 0.0099, "step": 14443 }, { "epoch": 9.241202815099168, "grad_norm": 0.5562320068772376, "learning_rate": 3.0046680608855295e-07, "loss": 0.0096, "step": 14444 }, { "epoch": 9.241842610364683, "grad_norm": 0.5095159996128322, "learning_rate": 2.999628777154662e-07, "loss": 0.0089, "step": 14445 }, { "epoch": 9.242482405630199, "grad_norm": 0.5035374400745638, "learning_rate": 2.994593658447187e-07, "loss": 0.0072, "step": 14446 }, { "epoch": 9.243122200895714, "grad_norm": 0.6093310933600324, "learning_rate": 2.9895627049792874e-07, "loss": 0.0093, "step": 14447 }, { "epoch": 9.24376199616123, "grad_norm": 0.5686123623284965, "learning_rate": 2.984535916967013e-07, "loss": 0.0072, "step": 14448 }, { "epoch": 9.244401791426743, "grad_norm": 0.5527471412147835, "learning_rate": 2.9795132946261906e-07, "loss": 0.0084, "step": 14449 }, { "epoch": 9.245041586692258, "grad_norm": 0.5103440729081338, "learning_rate": 2.9744948381724813e-07, "loss": 0.0093, "step": 14450 }, { "epoch": 9.245681381957773, "grad_norm": 0.5426537463355998, "learning_rate": 2.969480547821357e-07, "loss": 0.0079, "step": 14451 }, { "epoch": 9.246321177223289, "grad_norm": 0.5576054143589579, "learning_rate": 2.9644704237881463e-07, "loss": 0.0081, "step": 14452 }, { "epoch": 9.246960972488804, "grad_norm": 0.5191560173948948, "learning_rate": 2.9594644662879536e-07, "loss": 0.0095, "step": 14453 }, { "epoch": 9.247600767754319, "grad_norm": 0.5307263592386434, "learning_rate": 2.9544626755357296e-07, "loss": 0.0079, "step": 14454 }, { "epoch": 9.248240563019834, "grad_norm": 0.5636651052297935, "learning_rate": 2.9494650517462585e-07, "loss": 0.0104, "step": 14455 }, { "epoch": 9.248880358285348, "grad_norm": 0.5878037646600915, "learning_rate": 2.944471595134124e-07, "loss": 0.0088, "step": 14456 }, { "epoch": 9.249520153550863, "grad_norm": 0.5489459695102161, "learning_rate": 2.9394823059137103e-07, "loss": 0.0086, "step": 14457 }, { "epoch": 9.250159948816378, "grad_norm": 0.6599556479935859, "learning_rate": 2.93449718429929e-07, "loss": 0.0102, "step": 14458 }, { "epoch": 9.250799744081894, "grad_norm": 0.5735855349165173, "learning_rate": 2.929516230504892e-07, "loss": 0.0096, "step": 14459 }, { "epoch": 9.251439539347409, "grad_norm": 0.5119131456055233, "learning_rate": 2.924539444744379e-07, "loss": 0.0078, "step": 14460 }, { "epoch": 9.252079334612924, "grad_norm": 0.5834505831154907, "learning_rate": 2.919566827231457e-07, "loss": 0.0089, "step": 14461 }, { "epoch": 9.25271912987844, "grad_norm": 0.537244938187694, "learning_rate": 2.914598378179656e-07, "loss": 0.0093, "step": 14462 }, { "epoch": 9.253358925143955, "grad_norm": 0.5129651073696723, "learning_rate": 2.909634097802294e-07, "loss": 0.0077, "step": 14463 }, { "epoch": 9.253998720409468, "grad_norm": 0.6236580123435803, "learning_rate": 2.904673986312545e-07, "loss": 0.0079, "step": 14464 }, { "epoch": 9.254638515674984, "grad_norm": 0.6652505716208892, "learning_rate": 2.8997180439233605e-07, "loss": 0.0099, "step": 14465 }, { "epoch": 9.255278310940499, "grad_norm": 0.5409786441534267, "learning_rate": 2.89476627084756e-07, "loss": 0.0077, "step": 14466 }, { "epoch": 9.255918106206014, "grad_norm": 0.5372335553299844, "learning_rate": 2.8898186672977503e-07, "loss": 0.0078, "step": 14467 }, { "epoch": 9.25655790147153, "grad_norm": 0.48948247235274794, "learning_rate": 2.884875233486395e-07, "loss": 0.0076, "step": 14468 }, { "epoch": 9.257197696737045, "grad_norm": 0.6145332556861176, "learning_rate": 2.879935969625747e-07, "loss": 0.0089, "step": 14469 }, { "epoch": 9.25783749200256, "grad_norm": 0.6714560206456937, "learning_rate": 2.875000875927869e-07, "loss": 0.0113, "step": 14470 }, { "epoch": 9.258477287268073, "grad_norm": 0.41383653083001337, "learning_rate": 2.8700699526046817e-07, "loss": 0.0067, "step": 14471 }, { "epoch": 9.259117082533589, "grad_norm": 0.5305333860934854, "learning_rate": 2.865143199867926e-07, "loss": 0.0101, "step": 14472 }, { "epoch": 9.259756877799104, "grad_norm": 0.625017425036021, "learning_rate": 2.860220617929121e-07, "loss": 0.0083, "step": 14473 }, { "epoch": 9.26039667306462, "grad_norm": 0.5012349499990213, "learning_rate": 2.855302206999633e-07, "loss": 0.0073, "step": 14474 }, { "epoch": 9.261036468330134, "grad_norm": 0.5011570374461436, "learning_rate": 2.85038796729068e-07, "loss": 0.0085, "step": 14475 }, { "epoch": 9.26167626359565, "grad_norm": 0.5149357047475667, "learning_rate": 2.845477899013249e-07, "loss": 0.0092, "step": 14476 }, { "epoch": 9.262316058861165, "grad_norm": 0.4994679529810495, "learning_rate": 2.840572002378161e-07, "loss": 0.0086, "step": 14477 }, { "epoch": 9.26295585412668, "grad_norm": 0.5167535645250805, "learning_rate": 2.835670277596081e-07, "loss": 0.0082, "step": 14478 }, { "epoch": 9.263595649392194, "grad_norm": 0.6690503101237423, "learning_rate": 2.8307727248774954e-07, "loss": 0.0091, "step": 14479 }, { "epoch": 9.264235444657709, "grad_norm": 0.6629862199826027, "learning_rate": 2.825879344432658e-07, "loss": 0.0102, "step": 14480 }, { "epoch": 9.264875239923224, "grad_norm": 0.5692528230018551, "learning_rate": 2.8209901364717127e-07, "loss": 0.0093, "step": 14481 }, { "epoch": 9.26551503518874, "grad_norm": 0.5635078529986488, "learning_rate": 2.8161051012045803e-07, "loss": 0.0083, "step": 14482 }, { "epoch": 9.266154830454255, "grad_norm": 0.6333006599623441, "learning_rate": 2.811224238841026e-07, "loss": 0.0101, "step": 14483 }, { "epoch": 9.26679462571977, "grad_norm": 0.48530786141412463, "learning_rate": 2.8063475495906045e-07, "loss": 0.0072, "step": 14484 }, { "epoch": 9.267434420985285, "grad_norm": 0.5947933047951365, "learning_rate": 2.8014750336627264e-07, "loss": 0.0089, "step": 14485 }, { "epoch": 9.268074216250799, "grad_norm": 0.6353740693813933, "learning_rate": 2.796606691266612e-07, "loss": 0.0115, "step": 14486 }, { "epoch": 9.268714011516314, "grad_norm": 0.5738022998403978, "learning_rate": 2.791742522611285e-07, "loss": 0.008, "step": 14487 }, { "epoch": 9.26935380678183, "grad_norm": 0.5442629637590245, "learning_rate": 2.786882527905632e-07, "loss": 0.0085, "step": 14488 }, { "epoch": 9.269993602047345, "grad_norm": 0.6189189600040524, "learning_rate": 2.7820267073583096e-07, "loss": 0.0089, "step": 14489 }, { "epoch": 9.27063339731286, "grad_norm": 0.5441681244311943, "learning_rate": 2.777175061177817e-07, "loss": 0.0083, "step": 14490 }, { "epoch": 9.271273192578375, "grad_norm": 0.6494201292408313, "learning_rate": 2.7723275895724766e-07, "loss": 0.0108, "step": 14491 }, { "epoch": 9.27191298784389, "grad_norm": 0.5130391454079047, "learning_rate": 2.767484292750444e-07, "loss": 0.0071, "step": 14492 }, { "epoch": 9.272552783109404, "grad_norm": 0.5786411390717213, "learning_rate": 2.7626451709196643e-07, "loss": 0.0078, "step": 14493 }, { "epoch": 9.27319257837492, "grad_norm": 0.5676356293637805, "learning_rate": 2.757810224287938e-07, "loss": 0.0093, "step": 14494 }, { "epoch": 9.273832373640435, "grad_norm": 0.6022343483004925, "learning_rate": 2.752979453062865e-07, "loss": 0.0089, "step": 14495 }, { "epoch": 9.27447216890595, "grad_norm": 0.500734637024035, "learning_rate": 2.748152857451858e-07, "loss": 0.0079, "step": 14496 }, { "epoch": 9.275111964171465, "grad_norm": 0.49947556909966134, "learning_rate": 2.743330437662162e-07, "loss": 0.0082, "step": 14497 }, { "epoch": 9.27575175943698, "grad_norm": 0.5985531744292039, "learning_rate": 2.738512193900855e-07, "loss": 0.0094, "step": 14498 }, { "epoch": 9.276391554702496, "grad_norm": 0.5116640369669547, "learning_rate": 2.7336981263748284e-07, "loss": 0.0069, "step": 14499 }, { "epoch": 9.277031349968011, "grad_norm": 0.4788170157901708, "learning_rate": 2.728888235290783e-07, "loss": 0.0066, "step": 14500 }, { "epoch": 9.277671145233525, "grad_norm": 0.754128102638925, "learning_rate": 2.72408252085522e-07, "loss": 0.0086, "step": 14501 }, { "epoch": 9.27831094049904, "grad_norm": 0.5937992605021015, "learning_rate": 2.7192809832745305e-07, "loss": 0.01, "step": 14502 }, { "epoch": 9.278950735764555, "grad_norm": 0.6115819473420071, "learning_rate": 2.714483622754871e-07, "loss": 0.0087, "step": 14503 }, { "epoch": 9.27959053103007, "grad_norm": 0.6400174944936005, "learning_rate": 2.7096904395022217e-07, "loss": 0.0087, "step": 14504 }, { "epoch": 9.280230326295586, "grad_norm": 0.6652120748061434, "learning_rate": 2.7049014337223956e-07, "loss": 0.0108, "step": 14505 }, { "epoch": 9.280870121561101, "grad_norm": 0.6833797035265652, "learning_rate": 2.7001166056210393e-07, "loss": 0.0094, "step": 14506 }, { "epoch": 9.281509916826616, "grad_norm": 0.5020873744616264, "learning_rate": 2.6953359554035776e-07, "loss": 0.0072, "step": 14507 }, { "epoch": 9.28214971209213, "grad_norm": 0.507293337166435, "learning_rate": 2.6905594832753014e-07, "loss": 0.0079, "step": 14508 }, { "epoch": 9.282789507357645, "grad_norm": 0.7035079926279518, "learning_rate": 2.6857871894413135e-07, "loss": 0.01, "step": 14509 }, { "epoch": 9.28342930262316, "grad_norm": 0.4578901343393944, "learning_rate": 2.681019074106506e-07, "loss": 0.0082, "step": 14510 }, { "epoch": 9.284069097888676, "grad_norm": 0.5292560939886657, "learning_rate": 2.6762551374756253e-07, "loss": 0.0085, "step": 14511 }, { "epoch": 9.28470889315419, "grad_norm": 0.5389059550162221, "learning_rate": 2.6714953797532193e-07, "loss": 0.0082, "step": 14512 }, { "epoch": 9.285348688419706, "grad_norm": 0.6537307397612373, "learning_rate": 2.6667398011436695e-07, "loss": 0.0116, "step": 14513 }, { "epoch": 9.285988483685221, "grad_norm": 0.5099691514027219, "learning_rate": 2.661988401851168e-07, "loss": 0.0076, "step": 14514 }, { "epoch": 9.286628278950737, "grad_norm": 0.4260602139067891, "learning_rate": 2.6572411820797394e-07, "loss": 0.0066, "step": 14515 }, { "epoch": 9.28726807421625, "grad_norm": 0.5789894810869959, "learning_rate": 2.652498142033211e-07, "loss": 0.01, "step": 14516 }, { "epoch": 9.287907869481765, "grad_norm": 0.44483658768596224, "learning_rate": 2.6477592819152407e-07, "loss": 0.0064, "step": 14517 }, { "epoch": 9.28854766474728, "grad_norm": 0.5541529897048408, "learning_rate": 2.6430246019293116e-07, "loss": 0.009, "step": 14518 }, { "epoch": 9.289187460012796, "grad_norm": 0.6117978269356885, "learning_rate": 2.6382941022787376e-07, "loss": 0.0093, "step": 14519 }, { "epoch": 9.289827255278311, "grad_norm": 0.5609758069057753, "learning_rate": 2.6335677831666015e-07, "loss": 0.0092, "step": 14520 }, { "epoch": 9.290467050543826, "grad_norm": 0.5057255525964466, "learning_rate": 2.628845644795863e-07, "loss": 0.0081, "step": 14521 }, { "epoch": 9.291106845809342, "grad_norm": 0.771492161182624, "learning_rate": 2.624127687369293e-07, "loss": 0.0103, "step": 14522 }, { "epoch": 9.291746641074855, "grad_norm": 0.5565873577615291, "learning_rate": 2.619413911089463e-07, "loss": 0.0084, "step": 14523 }, { "epoch": 9.29238643634037, "grad_norm": 0.5895792613653892, "learning_rate": 2.614704316158778e-07, "loss": 0.0091, "step": 14524 }, { "epoch": 9.293026231605886, "grad_norm": 0.5220115286248339, "learning_rate": 2.6099989027794427e-07, "loss": 0.008, "step": 14525 }, { "epoch": 9.293666026871401, "grad_norm": 0.5139152887271943, "learning_rate": 2.605297671153517e-07, "loss": 0.0081, "step": 14526 }, { "epoch": 9.294305822136916, "grad_norm": 0.69715484282345, "learning_rate": 2.6006006214828516e-07, "loss": 0.0103, "step": 14527 }, { "epoch": 9.294945617402432, "grad_norm": 0.6152781820177653, "learning_rate": 2.5959077539691404e-07, "loss": 0.0102, "step": 14528 }, { "epoch": 9.295585412667947, "grad_norm": 0.5923632112575781, "learning_rate": 2.5912190688138885e-07, "loss": 0.0095, "step": 14529 }, { "epoch": 9.29622520793346, "grad_norm": 0.528560047149479, "learning_rate": 2.586534566218402e-07, "loss": 0.0075, "step": 14530 }, { "epoch": 9.296865003198976, "grad_norm": 0.6141682881107222, "learning_rate": 2.5818542463838304e-07, "loss": 0.0095, "step": 14531 }, { "epoch": 9.297504798464491, "grad_norm": 0.582354223486163, "learning_rate": 2.5771781095111583e-07, "loss": 0.0097, "step": 14532 }, { "epoch": 9.298144593730006, "grad_norm": 0.7210042888368151, "learning_rate": 2.572506155801158e-07, "loss": 0.0118, "step": 14533 }, { "epoch": 9.298784388995522, "grad_norm": 0.5589211929669015, "learning_rate": 2.567838385454413e-07, "loss": 0.0092, "step": 14534 }, { "epoch": 9.299424184261037, "grad_norm": 0.6053636561247574, "learning_rate": 2.563174798671386e-07, "loss": 0.0096, "step": 14535 }, { "epoch": 9.300063979526552, "grad_norm": 0.5489716818164878, "learning_rate": 2.5585153956523057e-07, "loss": 0.0085, "step": 14536 }, { "epoch": 9.300703774792067, "grad_norm": 0.534613439687133, "learning_rate": 2.553860176597245e-07, "loss": 0.0086, "step": 14537 }, { "epoch": 9.30134357005758, "grad_norm": 0.6420249179619179, "learning_rate": 2.549209141706066e-07, "loss": 0.0091, "step": 14538 }, { "epoch": 9.301983365323096, "grad_norm": 0.6074149571419035, "learning_rate": 2.5445622911785097e-07, "loss": 0.0097, "step": 14539 }, { "epoch": 9.302623160588611, "grad_norm": 0.5374610717435844, "learning_rate": 2.539919625214082e-07, "loss": 0.0084, "step": 14540 }, { "epoch": 9.303262955854127, "grad_norm": 0.5959771816008569, "learning_rate": 2.5352811440121474e-07, "loss": 0.0104, "step": 14541 }, { "epoch": 9.303902751119642, "grad_norm": 0.5647364282540582, "learning_rate": 2.530646847771856e-07, "loss": 0.0093, "step": 14542 }, { "epoch": 9.304542546385157, "grad_norm": 0.5845772789881157, "learning_rate": 2.526016736692205e-07, "loss": 0.0089, "step": 14543 }, { "epoch": 9.305182341650672, "grad_norm": 0.6066803255685246, "learning_rate": 2.521390810972002e-07, "loss": 0.0108, "step": 14544 }, { "epoch": 9.305822136916186, "grad_norm": 0.6892054749555571, "learning_rate": 2.5167690708098767e-07, "loss": 0.0104, "step": 14545 }, { "epoch": 9.306461932181701, "grad_norm": 0.6288620165390242, "learning_rate": 2.512151516404282e-07, "loss": 0.008, "step": 14546 }, { "epoch": 9.307101727447217, "grad_norm": 0.5188590202466073, "learning_rate": 2.507538147953481e-07, "loss": 0.0083, "step": 14547 }, { "epoch": 9.307741522712732, "grad_norm": 0.5456825607922079, "learning_rate": 2.50292896565556e-07, "loss": 0.0085, "step": 14548 }, { "epoch": 9.308381317978247, "grad_norm": 0.6477624626081816, "learning_rate": 2.4983239697084493e-07, "loss": 0.0098, "step": 14549 }, { "epoch": 9.309021113243762, "grad_norm": 0.5002320426401029, "learning_rate": 2.493723160309858e-07, "loss": 0.0076, "step": 14550 }, { "epoch": 9.309660908509278, "grad_norm": 0.5717464611479275, "learning_rate": 2.4891265376573384e-07, "loss": 0.008, "step": 14551 }, { "epoch": 9.310300703774793, "grad_norm": 0.4901508515688024, "learning_rate": 2.484534101948277e-07, "loss": 0.0073, "step": 14552 }, { "epoch": 9.310940499040306, "grad_norm": 0.4472130801842309, "learning_rate": 2.4799458533798504e-07, "loss": 0.0062, "step": 14553 }, { "epoch": 9.311580294305822, "grad_norm": 0.5727378789605976, "learning_rate": 2.475361792149078e-07, "loss": 0.009, "step": 14554 }, { "epoch": 9.312220089571337, "grad_norm": 0.4883813406363223, "learning_rate": 2.47078191845278e-07, "loss": 0.0079, "step": 14555 }, { "epoch": 9.312859884836852, "grad_norm": 0.538117839139531, "learning_rate": 2.4662062324876225e-07, "loss": 0.0074, "step": 14556 }, { "epoch": 9.313499680102368, "grad_norm": 0.5780507408868872, "learning_rate": 2.461634734450069e-07, "loss": 0.0066, "step": 14557 }, { "epoch": 9.314139475367883, "grad_norm": 0.6045450189410563, "learning_rate": 2.457067424536397e-07, "loss": 0.01, "step": 14558 }, { "epoch": 9.314779270633398, "grad_norm": 0.5145037989418065, "learning_rate": 2.452504302942749e-07, "loss": 0.0073, "step": 14559 }, { "epoch": 9.315419065898912, "grad_norm": 0.5122103591743594, "learning_rate": 2.4479453698650455e-07, "loss": 0.0088, "step": 14560 }, { "epoch": 9.316058861164427, "grad_norm": 0.5140197809993969, "learning_rate": 2.443390625499031e-07, "loss": 0.0088, "step": 14561 }, { "epoch": 9.316698656429942, "grad_norm": 0.5285393559059921, "learning_rate": 2.4388400700402824e-07, "loss": 0.008, "step": 14562 }, { "epoch": 9.317338451695457, "grad_norm": 0.6114247448171459, "learning_rate": 2.4342937036841983e-07, "loss": 0.0097, "step": 14563 }, { "epoch": 9.317978246960973, "grad_norm": 0.5920756912038206, "learning_rate": 2.4297515266259784e-07, "loss": 0.0092, "step": 14564 }, { "epoch": 9.318618042226488, "grad_norm": 0.5344966855481325, "learning_rate": 2.4252135390606666e-07, "loss": 0.0082, "step": 14565 }, { "epoch": 9.319257837492003, "grad_norm": 0.650250488158985, "learning_rate": 2.420679741183118e-07, "loss": 0.0098, "step": 14566 }, { "epoch": 9.319897632757517, "grad_norm": 0.5184255081842369, "learning_rate": 2.416150133188e-07, "loss": 0.0068, "step": 14567 }, { "epoch": 9.320537428023032, "grad_norm": 0.6056572530840248, "learning_rate": 2.4116247152698e-07, "loss": 0.0087, "step": 14568 }, { "epoch": 9.321177223288547, "grad_norm": 0.5680921391750424, "learning_rate": 2.4071034876228527e-07, "loss": 0.0085, "step": 14569 }, { "epoch": 9.321817018554063, "grad_norm": 0.6282797915859116, "learning_rate": 2.402586450441269e-07, "loss": 0.01, "step": 14570 }, { "epoch": 9.322456813819578, "grad_norm": 0.6167258312111725, "learning_rate": 2.398073603919004e-07, "loss": 0.0092, "step": 14571 }, { "epoch": 9.323096609085093, "grad_norm": 0.5312856034518747, "learning_rate": 2.393564948249849e-07, "loss": 0.0087, "step": 14572 }, { "epoch": 9.323736404350608, "grad_norm": 0.6214700733591088, "learning_rate": 2.3890604836273924e-07, "loss": 0.0094, "step": 14573 }, { "epoch": 9.324376199616124, "grad_norm": 0.6674017818221859, "learning_rate": 2.384560210245024e-07, "loss": 0.0089, "step": 14574 }, { "epoch": 9.325015994881637, "grad_norm": 0.5100488760140842, "learning_rate": 2.380064128296e-07, "loss": 0.0073, "step": 14575 }, { "epoch": 9.325655790147152, "grad_norm": 0.5432242623394108, "learning_rate": 2.3755722379733782e-07, "loss": 0.0107, "step": 14576 }, { "epoch": 9.326295585412668, "grad_norm": 0.589525777658414, "learning_rate": 2.3710845394700254e-07, "loss": 0.0094, "step": 14577 }, { "epoch": 9.326935380678183, "grad_norm": 0.6451099524209061, "learning_rate": 2.3666010329786216e-07, "loss": 0.0113, "step": 14578 }, { "epoch": 9.327575175943698, "grad_norm": 0.5174580712797615, "learning_rate": 2.362121718691701e-07, "loss": 0.0075, "step": 14579 }, { "epoch": 9.328214971209214, "grad_norm": 0.6823453870284736, "learning_rate": 2.3576465968015883e-07, "loss": 0.0072, "step": 14580 }, { "epoch": 9.328854766474729, "grad_norm": 0.6175505743045885, "learning_rate": 2.3531756675004403e-07, "loss": 0.0097, "step": 14581 }, { "epoch": 9.329494561740242, "grad_norm": 0.5152233139061511, "learning_rate": 2.348708930980226e-07, "loss": 0.0082, "step": 14582 }, { "epoch": 9.330134357005758, "grad_norm": 0.5830550746398178, "learning_rate": 2.3442463874327471e-07, "loss": 0.0097, "step": 14583 }, { "epoch": 9.330774152271273, "grad_norm": 0.5885583147118574, "learning_rate": 2.3397880370496062e-07, "loss": 0.0095, "step": 14584 }, { "epoch": 9.331413947536788, "grad_norm": 0.43076261590599235, "learning_rate": 2.3353338800222503e-07, "loss": 0.0065, "step": 14585 }, { "epoch": 9.332053742802303, "grad_norm": 0.579488449963249, "learning_rate": 2.3308839165419263e-07, "loss": 0.01, "step": 14586 }, { "epoch": 9.332693538067819, "grad_norm": 0.5367482880128742, "learning_rate": 2.3264381467997033e-07, "loss": 0.0093, "step": 14587 }, { "epoch": 9.333333333333334, "grad_norm": 0.6409058597569025, "learning_rate": 2.3219965709864623e-07, "loss": 0.0097, "step": 14588 }, { "epoch": 9.33397312859885, "grad_norm": 0.4872234665947539, "learning_rate": 2.3175591892929506e-07, "loss": 0.0082, "step": 14589 }, { "epoch": 9.334612923864363, "grad_norm": 0.5657127311341266, "learning_rate": 2.3131260019096824e-07, "loss": 0.0082, "step": 14590 }, { "epoch": 9.335252719129878, "grad_norm": 0.590753024612228, "learning_rate": 2.3086970090270056e-07, "loss": 0.0104, "step": 14591 }, { "epoch": 9.335892514395393, "grad_norm": 0.4427820015922315, "learning_rate": 2.3042722108351013e-07, "loss": 0.0069, "step": 14592 }, { "epoch": 9.336532309660909, "grad_norm": 0.46410074132394724, "learning_rate": 2.299851607523973e-07, "loss": 0.0067, "step": 14593 }, { "epoch": 9.337172104926424, "grad_norm": 0.5220375590937736, "learning_rate": 2.2954351992834023e-07, "loss": 0.0081, "step": 14594 }, { "epoch": 9.337811900191939, "grad_norm": 0.6132774360970669, "learning_rate": 2.2910229863030597e-07, "loss": 0.0091, "step": 14595 }, { "epoch": 9.338451695457454, "grad_norm": 0.5835037647288214, "learning_rate": 2.2866149687723715e-07, "loss": 0.0101, "step": 14596 }, { "epoch": 9.339091490722968, "grad_norm": 0.5358346389856157, "learning_rate": 2.2822111468806197e-07, "loss": 0.008, "step": 14597 }, { "epoch": 9.339731285988483, "grad_norm": 0.5341655962818902, "learning_rate": 2.2778115208168861e-07, "loss": 0.0078, "step": 14598 }, { "epoch": 9.340371081253998, "grad_norm": 0.560729709726384, "learning_rate": 2.2734160907700976e-07, "loss": 0.0103, "step": 14599 }, { "epoch": 9.341010876519514, "grad_norm": 0.6804435059218095, "learning_rate": 2.2690248569289809e-07, "loss": 0.0094, "step": 14600 }, { "epoch": 9.341650671785029, "grad_norm": 0.6348707810876667, "learning_rate": 2.264637819482096e-07, "loss": 0.009, "step": 14601 }, { "epoch": 9.342290467050544, "grad_norm": 0.4736021487645934, "learning_rate": 2.260254978617793e-07, "loss": 0.0085, "step": 14602 }, { "epoch": 9.34293026231606, "grad_norm": 0.6287468757251771, "learning_rate": 2.2558763345242874e-07, "loss": 0.0096, "step": 14603 }, { "epoch": 9.343570057581573, "grad_norm": 0.7027096848290417, "learning_rate": 2.2515018873895622e-07, "loss": 0.0103, "step": 14604 }, { "epoch": 9.344209852847088, "grad_norm": 0.5855874369828306, "learning_rate": 2.2471316374014672e-07, "loss": 0.0092, "step": 14605 }, { "epoch": 9.344849648112604, "grad_norm": 0.5469271707546861, "learning_rate": 2.2427655847476636e-07, "loss": 0.0093, "step": 14606 }, { "epoch": 9.345489443378119, "grad_norm": 0.4946102836841458, "learning_rate": 2.2384037296156014e-07, "loss": 0.0077, "step": 14607 }, { "epoch": 9.346129238643634, "grad_norm": 0.6461050498816244, "learning_rate": 2.234046072192575e-07, "loss": 0.0111, "step": 14608 }, { "epoch": 9.34676903390915, "grad_norm": 0.5719111555018593, "learning_rate": 2.2296926126657015e-07, "loss": 0.0095, "step": 14609 }, { "epoch": 9.347408829174665, "grad_norm": 0.5311731027641124, "learning_rate": 2.225343351221909e-07, "loss": 0.0066, "step": 14610 }, { "epoch": 9.34804862444018, "grad_norm": 0.5015106026008455, "learning_rate": 2.2209982880479374e-07, "loss": 0.0071, "step": 14611 }, { "epoch": 9.348688419705693, "grad_norm": 0.6088345979452734, "learning_rate": 2.2166574233303706e-07, "loss": 0.0095, "step": 14612 }, { "epoch": 9.349328214971209, "grad_norm": 0.6143081188547176, "learning_rate": 2.212320757255604e-07, "loss": 0.0082, "step": 14613 }, { "epoch": 9.349968010236724, "grad_norm": 0.6069797224265133, "learning_rate": 2.2079882900098214e-07, "loss": 0.0112, "step": 14614 }, { "epoch": 9.35060780550224, "grad_norm": 0.5733005627701512, "learning_rate": 2.203660021779064e-07, "loss": 0.0087, "step": 14615 }, { "epoch": 9.351247600767755, "grad_norm": 0.5065002470875767, "learning_rate": 2.199335952749171e-07, "loss": 0.0088, "step": 14616 }, { "epoch": 9.35188739603327, "grad_norm": 0.663667046724343, "learning_rate": 2.1950160831058277e-07, "loss": 0.0107, "step": 14617 }, { "epoch": 9.352527191298785, "grad_norm": 0.5512807926173262, "learning_rate": 2.1907004130345078e-07, "loss": 0.0087, "step": 14618 }, { "epoch": 9.353166986564299, "grad_norm": 0.6041190692929763, "learning_rate": 2.1863889427205297e-07, "loss": 0.0086, "step": 14619 }, { "epoch": 9.353806781829814, "grad_norm": 0.48000312139273593, "learning_rate": 2.1820816723490012e-07, "loss": 0.0072, "step": 14620 }, { "epoch": 9.35444657709533, "grad_norm": 0.5740335550374529, "learning_rate": 2.1777786021048854e-07, "loss": 0.0091, "step": 14621 }, { "epoch": 9.355086372360844, "grad_norm": 0.6022804637457313, "learning_rate": 2.173479732172945e-07, "loss": 0.0104, "step": 14622 }, { "epoch": 9.35572616762636, "grad_norm": 0.5846180610819217, "learning_rate": 2.1691850627377774e-07, "loss": 0.0089, "step": 14623 }, { "epoch": 9.356365962891875, "grad_norm": 0.5467830321678268, "learning_rate": 2.1648945939837685e-07, "loss": 0.0096, "step": 14624 }, { "epoch": 9.35700575815739, "grad_norm": 0.4766197296760667, "learning_rate": 2.160608326095137e-07, "loss": 0.0079, "step": 14625 }, { "epoch": 9.357645553422905, "grad_norm": 0.551269940019024, "learning_rate": 2.1563262592559475e-07, "loss": 0.0078, "step": 14626 }, { "epoch": 9.358285348688419, "grad_norm": 0.6041926768210935, "learning_rate": 2.1520483936500746e-07, "loss": 0.0086, "step": 14627 }, { "epoch": 9.358925143953934, "grad_norm": 0.5495990524894913, "learning_rate": 2.1477747294611607e-07, "loss": 0.009, "step": 14628 }, { "epoch": 9.35956493921945, "grad_norm": 0.5616373415628233, "learning_rate": 2.1435052668727474e-07, "loss": 0.0085, "step": 14629 }, { "epoch": 9.360204734484965, "grad_norm": 0.4876874988850256, "learning_rate": 2.139240006068144e-07, "loss": 0.0082, "step": 14630 }, { "epoch": 9.36084452975048, "grad_norm": 0.6366501956459205, "learning_rate": 2.134978947230504e-07, "loss": 0.0089, "step": 14631 }, { "epoch": 9.361484325015995, "grad_norm": 0.697488908735509, "learning_rate": 2.1307220905427695e-07, "loss": 0.0098, "step": 14632 }, { "epoch": 9.36212412028151, "grad_norm": 0.5942884844219989, "learning_rate": 2.126469436187728e-07, "loss": 0.0079, "step": 14633 }, { "epoch": 9.362763915547024, "grad_norm": 0.6393720582036304, "learning_rate": 2.1222209843480003e-07, "loss": 0.0107, "step": 14634 }, { "epoch": 9.36340371081254, "grad_norm": 0.5063594388452026, "learning_rate": 2.1179767352059734e-07, "loss": 0.0076, "step": 14635 }, { "epoch": 9.364043506078055, "grad_norm": 0.5533139170035882, "learning_rate": 2.1137366889439126e-07, "loss": 0.0081, "step": 14636 }, { "epoch": 9.36468330134357, "grad_norm": 0.5840282956261497, "learning_rate": 2.1095008457438837e-07, "loss": 0.0097, "step": 14637 }, { "epoch": 9.365323096609085, "grad_norm": 0.6198819890187423, "learning_rate": 2.105269205787741e-07, "loss": 0.0105, "step": 14638 }, { "epoch": 9.3659628918746, "grad_norm": 0.5576030202154405, "learning_rate": 2.101041769257206e-07, "loss": 0.0088, "step": 14639 }, { "epoch": 9.366602687140116, "grad_norm": 0.5652178037569372, "learning_rate": 2.096818536333789e-07, "loss": 0.0097, "step": 14640 }, { "epoch": 9.36724248240563, "grad_norm": 0.7128229044387488, "learning_rate": 2.0925995071988114e-07, "loss": 0.0097, "step": 14641 }, { "epoch": 9.367882277671145, "grad_norm": 0.5435701788455216, "learning_rate": 2.0883846820334618e-07, "loss": 0.0092, "step": 14642 }, { "epoch": 9.36852207293666, "grad_norm": 0.6230276242071583, "learning_rate": 2.0841740610187066e-07, "loss": 0.0103, "step": 14643 }, { "epoch": 9.369161868202175, "grad_norm": 0.5183673432601066, "learning_rate": 2.0799676443353346e-07, "loss": 0.0062, "step": 14644 }, { "epoch": 9.36980166346769, "grad_norm": 0.48874255556277524, "learning_rate": 2.0757654321639676e-07, "loss": 0.0071, "step": 14645 }, { "epoch": 9.370441458733206, "grad_norm": 0.5178575989977491, "learning_rate": 2.0715674246850393e-07, "loss": 0.0087, "step": 14646 }, { "epoch": 9.371081253998721, "grad_norm": 0.5307728704184776, "learning_rate": 2.0673736220787943e-07, "loss": 0.009, "step": 14647 }, { "epoch": 9.371721049264236, "grad_norm": 0.5273157791624751, "learning_rate": 2.063184024525322e-07, "loss": 0.0081, "step": 14648 }, { "epoch": 9.37236084452975, "grad_norm": 0.5823076086123009, "learning_rate": 2.058998632204512e-07, "loss": 0.009, "step": 14649 }, { "epoch": 9.373000639795265, "grad_norm": 0.560949498293906, "learning_rate": 2.0548174452960868e-07, "loss": 0.0093, "step": 14650 }, { "epoch": 9.37364043506078, "grad_norm": 0.6166367002349321, "learning_rate": 2.0506404639795474e-07, "loss": 0.0099, "step": 14651 }, { "epoch": 9.374280230326296, "grad_norm": 0.5289480302714218, "learning_rate": 2.0464676884342837e-07, "loss": 0.0076, "step": 14652 }, { "epoch": 9.37492002559181, "grad_norm": 0.6885396646224111, "learning_rate": 2.0422991188394526e-07, "loss": 0.009, "step": 14653 }, { "epoch": 9.375559820857326, "grad_norm": 0.6606268308980323, "learning_rate": 2.0381347553740437e-07, "loss": 0.0104, "step": 14654 }, { "epoch": 9.376199616122841, "grad_norm": 0.6404940096131182, "learning_rate": 2.033974598216859e-07, "loss": 0.0099, "step": 14655 }, { "epoch": 9.376839411388357, "grad_norm": 0.5283189422646105, "learning_rate": 2.0298186475465553e-07, "loss": 0.0082, "step": 14656 }, { "epoch": 9.37747920665387, "grad_norm": 0.6147587884653731, "learning_rate": 2.0256669035415454e-07, "loss": 0.0092, "step": 14657 }, { "epoch": 9.378119001919385, "grad_norm": 0.5687420916040922, "learning_rate": 2.0215193663801203e-07, "loss": 0.009, "step": 14658 }, { "epoch": 9.3787587971849, "grad_norm": 0.5212588959565503, "learning_rate": 2.0173760362403704e-07, "loss": 0.0088, "step": 14659 }, { "epoch": 9.379398592450416, "grad_norm": 0.5332457886891303, "learning_rate": 2.0132369133001873e-07, "loss": 0.0077, "step": 14660 }, { "epoch": 9.380038387715931, "grad_norm": 0.5074407420263357, "learning_rate": 2.0091019977373172e-07, "loss": 0.0081, "step": 14661 }, { "epoch": 9.380678182981447, "grad_norm": 0.5687693167354032, "learning_rate": 2.004971289729285e-07, "loss": 0.008, "step": 14662 }, { "epoch": 9.381317978246962, "grad_norm": 0.6154808375595046, "learning_rate": 2.000844789453471e-07, "loss": 0.0089, "step": 14663 }, { "epoch": 9.381957773512475, "grad_norm": 0.5455424800281602, "learning_rate": 1.9967224970870558e-07, "loss": 0.0082, "step": 14664 }, { "epoch": 9.38259756877799, "grad_norm": 0.5534124285971093, "learning_rate": 1.992604412807031e-07, "loss": 0.0093, "step": 14665 }, { "epoch": 9.383237364043506, "grad_norm": 0.6207252165508857, "learning_rate": 1.988490536790244e-07, "loss": 0.011, "step": 14666 }, { "epoch": 9.383877159309021, "grad_norm": 0.47308202849599984, "learning_rate": 1.9843808692133204e-07, "loss": 0.0062, "step": 14667 }, { "epoch": 9.384516954574536, "grad_norm": 0.5122918144867133, "learning_rate": 1.980275410252719e-07, "loss": 0.0082, "step": 14668 }, { "epoch": 9.385156749840052, "grad_norm": 0.54727491691109, "learning_rate": 1.9761741600847318e-07, "loss": 0.0095, "step": 14669 }, { "epoch": 9.385796545105567, "grad_norm": 0.4141282353932272, "learning_rate": 1.9720771188854626e-07, "loss": 0.0058, "step": 14670 }, { "epoch": 9.38643634037108, "grad_norm": 0.6032408357872193, "learning_rate": 1.9679842868308152e-07, "loss": 0.0092, "step": 14671 }, { "epoch": 9.387076135636596, "grad_norm": 0.5184131005574925, "learning_rate": 1.963895664096538e-07, "loss": 0.0074, "step": 14672 }, { "epoch": 9.387715930902111, "grad_norm": 0.48612235556534494, "learning_rate": 1.95981125085819e-07, "loss": 0.0073, "step": 14673 }, { "epoch": 9.388355726167626, "grad_norm": 0.6029674562188143, "learning_rate": 1.9557310472911538e-07, "loss": 0.0091, "step": 14674 }, { "epoch": 9.388995521433142, "grad_norm": 0.5751610696337016, "learning_rate": 1.9516550535706115e-07, "loss": 0.0087, "step": 14675 }, { "epoch": 9.389635316698657, "grad_norm": 0.6121054484168957, "learning_rate": 1.9475832698716e-07, "loss": 0.011, "step": 14676 }, { "epoch": 9.390275111964172, "grad_norm": 0.5235550251406035, "learning_rate": 1.943515696368925e-07, "loss": 0.0084, "step": 14677 }, { "epoch": 9.390914907229687, "grad_norm": 0.5164212317806425, "learning_rate": 1.9394523332372573e-07, "loss": 0.0083, "step": 14678 }, { "epoch": 9.3915547024952, "grad_norm": 0.6388093206190459, "learning_rate": 1.9353931806510794e-07, "loss": 0.0094, "step": 14679 }, { "epoch": 9.392194497760716, "grad_norm": 0.5437048078887143, "learning_rate": 1.9313382387846635e-07, "loss": 0.0085, "step": 14680 }, { "epoch": 9.392834293026231, "grad_norm": 0.5966716769278029, "learning_rate": 1.9272875078121367e-07, "loss": 0.0082, "step": 14681 }, { "epoch": 9.393474088291747, "grad_norm": 0.8259634759271102, "learning_rate": 1.9232409879074375e-07, "loss": 0.0108, "step": 14682 }, { "epoch": 9.394113883557262, "grad_norm": 0.520173206210993, "learning_rate": 1.919198679244294e-07, "loss": 0.0078, "step": 14683 }, { "epoch": 9.394753678822777, "grad_norm": 0.7128170537696756, "learning_rate": 1.9151605819962894e-07, "loss": 0.0138, "step": 14684 }, { "epoch": 9.395393474088293, "grad_norm": 0.5968987527295981, "learning_rate": 1.911126696336807e-07, "loss": 0.0089, "step": 14685 }, { "epoch": 9.396033269353806, "grad_norm": 0.565388641908235, "learning_rate": 1.907097022439064e-07, "loss": 0.0093, "step": 14686 }, { "epoch": 9.396673064619321, "grad_norm": 0.42588548246248015, "learning_rate": 1.9030715604760775e-07, "loss": 0.0064, "step": 14687 }, { "epoch": 9.397312859884837, "grad_norm": 0.6063745575787618, "learning_rate": 1.8990503106206982e-07, "loss": 0.0103, "step": 14688 }, { "epoch": 9.397952655150352, "grad_norm": 0.5061801176109181, "learning_rate": 1.895033273045588e-07, "loss": 0.0068, "step": 14689 }, { "epoch": 9.398592450415867, "grad_norm": 0.591269887703796, "learning_rate": 1.891020447923242e-07, "loss": 0.0085, "step": 14690 }, { "epoch": 9.399232245681382, "grad_norm": 0.59119639134259, "learning_rate": 1.8870118354259448e-07, "loss": 0.0096, "step": 14691 }, { "epoch": 9.399872040946898, "grad_norm": 0.5181044274143373, "learning_rate": 1.8830074357258367e-07, "loss": 0.0079, "step": 14692 }, { "epoch": 9.400511836212413, "grad_norm": 0.5725097843898355, "learning_rate": 1.8790072489948574e-07, "loss": 0.0093, "step": 14693 }, { "epoch": 9.401151631477926, "grad_norm": 0.6272538916690417, "learning_rate": 1.8750112754047588e-07, "loss": 0.0099, "step": 14694 }, { "epoch": 9.401791426743442, "grad_norm": 0.533667021930139, "learning_rate": 1.871019515127115e-07, "loss": 0.0074, "step": 14695 }, { "epoch": 9.402431222008957, "grad_norm": 0.5354993963498382, "learning_rate": 1.8670319683333438e-07, "loss": 0.0081, "step": 14696 }, { "epoch": 9.403071017274472, "grad_norm": 0.557530010374494, "learning_rate": 1.863048635194664e-07, "loss": 0.0091, "step": 14697 }, { "epoch": 9.403710812539988, "grad_norm": 0.6223040141595267, "learning_rate": 1.8590695158820838e-07, "loss": 0.0081, "step": 14698 }, { "epoch": 9.404350607805503, "grad_norm": 0.5466699966839519, "learning_rate": 1.855094610566499e-07, "loss": 0.0073, "step": 14699 }, { "epoch": 9.404990403071018, "grad_norm": 0.520885479695522, "learning_rate": 1.8511239194185515e-07, "loss": 0.0076, "step": 14700 }, { "epoch": 9.405630198336532, "grad_norm": 0.6399246793453776, "learning_rate": 1.84715744260876e-07, "loss": 0.0108, "step": 14701 }, { "epoch": 9.406269993602047, "grad_norm": 0.5787456878784545, "learning_rate": 1.843195180307411e-07, "loss": 0.01, "step": 14702 }, { "epoch": 9.406909788867562, "grad_norm": 0.680122052913876, "learning_rate": 1.8392371326846682e-07, "loss": 0.0108, "step": 14703 }, { "epoch": 9.407549584133077, "grad_norm": 0.5872012172314337, "learning_rate": 1.835283299910473e-07, "loss": 0.0096, "step": 14704 }, { "epoch": 9.408189379398593, "grad_norm": 0.7266026688393663, "learning_rate": 1.831333682154579e-07, "loss": 0.0116, "step": 14705 }, { "epoch": 9.408829174664108, "grad_norm": 0.5775730356689306, "learning_rate": 1.827388279586595e-07, "loss": 0.0086, "step": 14706 }, { "epoch": 9.409468969929623, "grad_norm": 0.5925565979512577, "learning_rate": 1.8234470923759185e-07, "loss": 0.0109, "step": 14707 }, { "epoch": 9.410108765195137, "grad_norm": 0.5179876571733912, "learning_rate": 1.8195101206917808e-07, "loss": 0.0074, "step": 14708 }, { "epoch": 9.410748560460652, "grad_norm": 0.4375543448545441, "learning_rate": 1.815577364703236e-07, "loss": 0.0066, "step": 14709 }, { "epoch": 9.411388355726167, "grad_norm": 0.5513660739333365, "learning_rate": 1.8116488245791376e-07, "loss": 0.0074, "step": 14710 }, { "epoch": 9.412028150991683, "grad_norm": 0.5547214223935998, "learning_rate": 1.8077245004881728e-07, "loss": 0.0085, "step": 14711 }, { "epoch": 9.412667946257198, "grad_norm": 0.5784473481465751, "learning_rate": 1.8038043925988512e-07, "loss": 0.0071, "step": 14712 }, { "epoch": 9.413307741522713, "grad_norm": 0.5768895072807387, "learning_rate": 1.7998885010794942e-07, "loss": 0.009, "step": 14713 }, { "epoch": 9.413947536788228, "grad_norm": 0.5570542311004401, "learning_rate": 1.7959768260982336e-07, "loss": 0.0088, "step": 14714 }, { "epoch": 9.414587332053744, "grad_norm": 0.5834446565533707, "learning_rate": 1.792069367823035e-07, "loss": 0.0098, "step": 14715 }, { "epoch": 9.415227127319257, "grad_norm": 0.5607592670273953, "learning_rate": 1.7881661264216866e-07, "loss": 0.0074, "step": 14716 }, { "epoch": 9.415866922584772, "grad_norm": 0.5578345404725291, "learning_rate": 1.7842671020617763e-07, "loss": 0.0092, "step": 14717 }, { "epoch": 9.416506717850288, "grad_norm": 0.657366723287526, "learning_rate": 1.7803722949107149e-07, "loss": 0.0093, "step": 14718 }, { "epoch": 9.417146513115803, "grad_norm": 0.559655513258331, "learning_rate": 1.7764817051357687e-07, "loss": 0.0084, "step": 14719 }, { "epoch": 9.417786308381318, "grad_norm": 0.6593870641911252, "learning_rate": 1.7725953329039592e-07, "loss": 0.0079, "step": 14720 }, { "epoch": 9.418426103646834, "grad_norm": 0.5105279636503761, "learning_rate": 1.7687131783821754e-07, "loss": 0.0087, "step": 14721 }, { "epoch": 9.419065898912349, "grad_norm": 0.5517697209094055, "learning_rate": 1.7648352417371174e-07, "loss": 0.0086, "step": 14722 }, { "epoch": 9.419705694177862, "grad_norm": 0.544289673210287, "learning_rate": 1.7609615231352738e-07, "loss": 0.0094, "step": 14723 }, { "epoch": 9.420345489443378, "grad_norm": 0.6180119237047116, "learning_rate": 1.757092022743001e-07, "loss": 0.0094, "step": 14724 }, { "epoch": 9.420985284708893, "grad_norm": 0.5720132338271522, "learning_rate": 1.7532267407264215e-07, "loss": 0.0092, "step": 14725 }, { "epoch": 9.421625079974408, "grad_norm": 0.6343780687460954, "learning_rate": 1.7493656772515244e-07, "loss": 0.0083, "step": 14726 }, { "epoch": 9.422264875239923, "grad_norm": 0.4301538380259259, "learning_rate": 1.7455088324840886e-07, "loss": 0.0061, "step": 14727 }, { "epoch": 9.422904670505439, "grad_norm": 0.46965631923725093, "learning_rate": 1.7416562065897146e-07, "loss": 0.0068, "step": 14728 }, { "epoch": 9.423544465770954, "grad_norm": 0.6007792835706579, "learning_rate": 1.7378077997338483e-07, "loss": 0.0104, "step": 14729 }, { "epoch": 9.42418426103647, "grad_norm": 0.7390697842315712, "learning_rate": 1.733963612081724e-07, "loss": 0.0109, "step": 14730 }, { "epoch": 9.424824056301983, "grad_norm": 0.6200730432839956, "learning_rate": 1.730123643798398e-07, "loss": 0.0092, "step": 14731 }, { "epoch": 9.425463851567498, "grad_norm": 0.6672376515723192, "learning_rate": 1.7262878950487394e-07, "loss": 0.0105, "step": 14732 }, { "epoch": 9.426103646833013, "grad_norm": 0.5999914670442597, "learning_rate": 1.7224563659974713e-07, "loss": 0.0078, "step": 14733 }, { "epoch": 9.426743442098529, "grad_norm": 0.5421043383596602, "learning_rate": 1.718629056809118e-07, "loss": 0.0085, "step": 14734 }, { "epoch": 9.427383237364044, "grad_norm": 0.49512050226627363, "learning_rate": 1.7148059676479922e-07, "loss": 0.0074, "step": 14735 }, { "epoch": 9.428023032629559, "grad_norm": 0.5948858123139252, "learning_rate": 1.7109870986782627e-07, "loss": 0.0098, "step": 14736 }, { "epoch": 9.428662827895074, "grad_norm": 0.6475008064870237, "learning_rate": 1.7071724500639096e-07, "loss": 0.0108, "step": 14737 }, { "epoch": 9.429302623160588, "grad_norm": 0.6421169923610472, "learning_rate": 1.703362021968713e-07, "loss": 0.0079, "step": 14738 }, { "epoch": 9.429942418426103, "grad_norm": 0.5680811821715723, "learning_rate": 1.6995558145562975e-07, "loss": 0.0086, "step": 14739 }, { "epoch": 9.430582213691618, "grad_norm": 0.5804737538968539, "learning_rate": 1.6957538279900988e-07, "loss": 0.0085, "step": 14740 }, { "epoch": 9.431222008957134, "grad_norm": 0.49804146187599063, "learning_rate": 1.691956062433353e-07, "loss": 0.0079, "step": 14741 }, { "epoch": 9.431861804222649, "grad_norm": 0.5154083000317174, "learning_rate": 1.6881625180491413e-07, "loss": 0.0079, "step": 14742 }, { "epoch": 9.432501599488164, "grad_norm": 0.5995448637518098, "learning_rate": 1.684373195000344e-07, "loss": 0.0084, "step": 14743 }, { "epoch": 9.43314139475368, "grad_norm": 0.5326079097218616, "learning_rate": 1.6805880934496754e-07, "loss": 0.0079, "step": 14744 }, { "epoch": 9.433781190019193, "grad_norm": 0.5367655819855228, "learning_rate": 1.6768072135596504e-07, "loss": 0.0072, "step": 14745 }, { "epoch": 9.434420985284708, "grad_norm": 0.5308923503435068, "learning_rate": 1.6730305554926385e-07, "loss": 0.0069, "step": 14746 }, { "epoch": 9.435060780550224, "grad_norm": 0.5632279481891044, "learning_rate": 1.669258119410766e-07, "loss": 0.0077, "step": 14747 }, { "epoch": 9.435700575815739, "grad_norm": 0.47068825673185355, "learning_rate": 1.6654899054760366e-07, "loss": 0.0081, "step": 14748 }, { "epoch": 9.436340371081254, "grad_norm": 0.5017285935492096, "learning_rate": 1.661725913850254e-07, "loss": 0.0076, "step": 14749 }, { "epoch": 9.43698016634677, "grad_norm": 0.5917034904855623, "learning_rate": 1.6579661446950334e-07, "loss": 0.0076, "step": 14750 }, { "epoch": 9.437619961612285, "grad_norm": 0.8156538265558291, "learning_rate": 1.6542105981718016e-07, "loss": 0.0098, "step": 14751 }, { "epoch": 9.4382597568778, "grad_norm": 0.6058779504942353, "learning_rate": 1.6504592744418292e-07, "loss": 0.0097, "step": 14752 }, { "epoch": 9.438899552143313, "grad_norm": 0.5873819465377749, "learning_rate": 1.6467121736661762e-07, "loss": 0.0099, "step": 14753 }, { "epoch": 9.439539347408829, "grad_norm": 0.5156773362863254, "learning_rate": 1.6429692960057586e-07, "loss": 0.0081, "step": 14754 }, { "epoch": 9.440179142674344, "grad_norm": 0.48582109857748, "learning_rate": 1.639230641621259e-07, "loss": 0.0075, "step": 14755 }, { "epoch": 9.44081893793986, "grad_norm": 0.53534436279329, "learning_rate": 1.6354962106732375e-07, "loss": 0.009, "step": 14756 }, { "epoch": 9.441458733205375, "grad_norm": 0.6113936046606042, "learning_rate": 1.631766003322033e-07, "loss": 0.0086, "step": 14757 }, { "epoch": 9.44209852847089, "grad_norm": 0.5635260952290672, "learning_rate": 1.6280400197277945e-07, "loss": 0.01, "step": 14758 }, { "epoch": 9.442738323736405, "grad_norm": 0.5939775442512971, "learning_rate": 1.6243182600505503e-07, "loss": 0.0075, "step": 14759 }, { "epoch": 9.443378119001919, "grad_norm": 0.5880954472025123, "learning_rate": 1.6206007244500721e-07, "loss": 0.0095, "step": 14760 }, { "epoch": 9.444017914267434, "grad_norm": 0.6281867263357681, "learning_rate": 1.616887413085999e-07, "loss": 0.0088, "step": 14761 }, { "epoch": 9.44465770953295, "grad_norm": 0.5311906932037268, "learning_rate": 1.6131783261177591e-07, "loss": 0.008, "step": 14762 }, { "epoch": 9.445297504798464, "grad_norm": 0.5437854658603479, "learning_rate": 1.609473463704647e-07, "loss": 0.0072, "step": 14763 }, { "epoch": 9.44593730006398, "grad_norm": 0.6377694207704662, "learning_rate": 1.6057728260057025e-07, "loss": 0.0104, "step": 14764 }, { "epoch": 9.446577095329495, "grad_norm": 0.5607844446256942, "learning_rate": 1.6020764131798538e-07, "loss": 0.0092, "step": 14765 }, { "epoch": 9.44721689059501, "grad_norm": 0.6087012367471174, "learning_rate": 1.5983842253858074e-07, "loss": 0.0089, "step": 14766 }, { "epoch": 9.447856685860526, "grad_norm": 0.6120440553270877, "learning_rate": 1.5946962627820917e-07, "loss": 0.011, "step": 14767 }, { "epoch": 9.448496481126039, "grad_norm": 0.5350400702927489, "learning_rate": 1.5910125255270803e-07, "loss": 0.0084, "step": 14768 }, { "epoch": 9.449136276391554, "grad_norm": 0.557678598080043, "learning_rate": 1.5873330137789245e-07, "loss": 0.0071, "step": 14769 }, { "epoch": 9.44977607165707, "grad_norm": 0.5233827696024939, "learning_rate": 1.583657727695631e-07, "loss": 0.0075, "step": 14770 }, { "epoch": 9.450415866922585, "grad_norm": 0.5862740607654622, "learning_rate": 1.5799866674350074e-07, "loss": 0.0091, "step": 14771 }, { "epoch": 9.4510556621881, "grad_norm": 0.665110308126606, "learning_rate": 1.5763198331546713e-07, "loss": 0.0103, "step": 14772 }, { "epoch": 9.451695457453615, "grad_norm": 0.587760328829766, "learning_rate": 1.5726572250120864e-07, "loss": 0.0088, "step": 14773 }, { "epoch": 9.45233525271913, "grad_norm": 0.46449377393835967, "learning_rate": 1.5689988431645155e-07, "loss": 0.0069, "step": 14774 }, { "epoch": 9.452975047984644, "grad_norm": 0.5521272752866188, "learning_rate": 1.5653446877690216e-07, "loss": 0.0073, "step": 14775 }, { "epoch": 9.45361484325016, "grad_norm": 0.4863146743380249, "learning_rate": 1.561694758982546e-07, "loss": 0.0064, "step": 14776 }, { "epoch": 9.454254638515675, "grad_norm": 0.5789544059751747, "learning_rate": 1.5580490569617745e-07, "loss": 0.0078, "step": 14777 }, { "epoch": 9.45489443378119, "grad_norm": 0.5470036097672274, "learning_rate": 1.554407581863271e-07, "loss": 0.0085, "step": 14778 }, { "epoch": 9.455534229046705, "grad_norm": 0.532792786506715, "learning_rate": 1.5507703338433767e-07, "loss": 0.0076, "step": 14779 }, { "epoch": 9.45617402431222, "grad_norm": 0.5404817178035433, "learning_rate": 1.547137313058289e-07, "loss": 0.0086, "step": 14780 }, { "epoch": 9.456813819577736, "grad_norm": 0.5517813190381049, "learning_rate": 1.5435085196639833e-07, "loss": 0.01, "step": 14781 }, { "epoch": 9.45745361484325, "grad_norm": 0.5194534659844707, "learning_rate": 1.5398839538162902e-07, "loss": 0.0072, "step": 14782 }, { "epoch": 9.458093410108765, "grad_norm": 0.5317427316862802, "learning_rate": 1.5362636156708188e-07, "loss": 0.0076, "step": 14783 }, { "epoch": 9.45873320537428, "grad_norm": 0.5257211692550733, "learning_rate": 1.5326475053830448e-07, "loss": 0.0075, "step": 14784 }, { "epoch": 9.459373000639795, "grad_norm": 0.5996515704034306, "learning_rate": 1.5290356231082216e-07, "loss": 0.0098, "step": 14785 }, { "epoch": 9.46001279590531, "grad_norm": 0.5857854833107192, "learning_rate": 1.5254279690014474e-07, "loss": 0.009, "step": 14786 }, { "epoch": 9.460652591170826, "grad_norm": 0.538805404175628, "learning_rate": 1.5218245432176205e-07, "loss": 0.0079, "step": 14787 }, { "epoch": 9.461292386436341, "grad_norm": 0.4645276059918946, "learning_rate": 1.5182253459114726e-07, "loss": 0.0056, "step": 14788 }, { "epoch": 9.461932181701856, "grad_norm": 0.6008341109972863, "learning_rate": 1.5146303772375248e-07, "loss": 0.0085, "step": 14789 }, { "epoch": 9.46257197696737, "grad_norm": 0.6229931423071539, "learning_rate": 1.5110396373501758e-07, "loss": 0.01, "step": 14790 }, { "epoch": 9.463211772232885, "grad_norm": 0.5922729390634356, "learning_rate": 1.5074531264035798e-07, "loss": 0.01, "step": 14791 }, { "epoch": 9.4638515674984, "grad_norm": 0.5648425780543148, "learning_rate": 1.503870844551747e-07, "loss": 0.0088, "step": 14792 }, { "epoch": 9.464491362763916, "grad_norm": 0.6440370906317042, "learning_rate": 1.5002927919484765e-07, "loss": 0.0088, "step": 14793 }, { "epoch": 9.46513115802943, "grad_norm": 0.55844195505613, "learning_rate": 1.4967189687474348e-07, "loss": 0.0087, "step": 14794 }, { "epoch": 9.465770953294946, "grad_norm": 0.5408629342412269, "learning_rate": 1.493149375102043e-07, "loss": 0.0081, "step": 14795 }, { "epoch": 9.466410748560461, "grad_norm": 0.5463740542683232, "learning_rate": 1.4895840111656013e-07, "loss": 0.0084, "step": 14796 }, { "epoch": 9.467050543825975, "grad_norm": 0.4735292750368726, "learning_rate": 1.4860228770911757e-07, "loss": 0.0064, "step": 14797 }, { "epoch": 9.46769033909149, "grad_norm": 0.5224978074263399, "learning_rate": 1.4824659730316883e-07, "loss": 0.0071, "step": 14798 }, { "epoch": 9.468330134357005, "grad_norm": 0.5707226184298577, "learning_rate": 1.4789132991398504e-07, "loss": 0.0078, "step": 14799 }, { "epoch": 9.46896992962252, "grad_norm": 0.5553953283683025, "learning_rate": 1.4753648555682288e-07, "loss": 0.0084, "step": 14800 }, { "epoch": 9.469609724888036, "grad_norm": 0.6488712262048173, "learning_rate": 1.4718206424691793e-07, "loss": 0.0111, "step": 14801 }, { "epoch": 9.470249520153551, "grad_norm": 0.5636111860321141, "learning_rate": 1.4682806599948807e-07, "loss": 0.0097, "step": 14802 }, { "epoch": 9.470889315419067, "grad_norm": 0.5764743014272157, "learning_rate": 1.464744908297333e-07, "loss": 0.0067, "step": 14803 }, { "epoch": 9.471529110684582, "grad_norm": 0.5751483068457484, "learning_rate": 1.4612133875283708e-07, "loss": 0.0092, "step": 14804 }, { "epoch": 9.472168905950095, "grad_norm": 0.55999671415428, "learning_rate": 1.4576860978396057e-07, "loss": 0.008, "step": 14805 }, { "epoch": 9.47280870121561, "grad_norm": 0.6156244054176456, "learning_rate": 1.4541630393825167e-07, "loss": 0.0083, "step": 14806 }, { "epoch": 9.473448496481126, "grad_norm": 0.5690240143559788, "learning_rate": 1.450644212308361e-07, "loss": 0.0085, "step": 14807 }, { "epoch": 9.474088291746641, "grad_norm": 0.5568620079004917, "learning_rate": 1.4471296167682392e-07, "loss": 0.0089, "step": 14808 }, { "epoch": 9.474728087012156, "grad_norm": 0.5021130762501121, "learning_rate": 1.4436192529130532e-07, "loss": 0.0087, "step": 14809 }, { "epoch": 9.475367882277672, "grad_norm": 0.6300928642366388, "learning_rate": 1.4401131208935493e-07, "loss": 0.0092, "step": 14810 }, { "epoch": 9.476007677543187, "grad_norm": 0.5416112725440001, "learning_rate": 1.4366112208602623e-07, "loss": 0.0074, "step": 14811 }, { "epoch": 9.4766474728087, "grad_norm": 0.517536327707889, "learning_rate": 1.433113552963561e-07, "loss": 0.0077, "step": 14812 }, { "epoch": 9.477287268074216, "grad_norm": 0.6354964127992866, "learning_rate": 1.429620117353614e-07, "loss": 0.0102, "step": 14813 }, { "epoch": 9.477927063339731, "grad_norm": 0.6057274853082785, "learning_rate": 1.4261309141804458e-07, "loss": 0.0103, "step": 14814 }, { "epoch": 9.478566858605246, "grad_norm": 0.6212869682975535, "learning_rate": 1.4226459435938588e-07, "loss": 0.0101, "step": 14815 }, { "epoch": 9.479206653870762, "grad_norm": 0.5611040985213627, "learning_rate": 1.4191652057435112e-07, "loss": 0.0102, "step": 14816 }, { "epoch": 9.479846449136277, "grad_norm": 0.6735630457368078, "learning_rate": 1.4156887007788388e-07, "loss": 0.0095, "step": 14817 }, { "epoch": 9.480486244401792, "grad_norm": 0.6337082685816611, "learning_rate": 1.4122164288491336e-07, "loss": 0.0088, "step": 14818 }, { "epoch": 9.481126039667306, "grad_norm": 0.5888629508071062, "learning_rate": 1.408748390103476e-07, "loss": 0.0097, "step": 14819 }, { "epoch": 9.48176583493282, "grad_norm": 0.5022522510523989, "learning_rate": 1.4052845846907803e-07, "loss": 0.0075, "step": 14820 }, { "epoch": 9.482405630198336, "grad_norm": 0.4807723766312278, "learning_rate": 1.401825012759783e-07, "loss": 0.0071, "step": 14821 }, { "epoch": 9.483045425463851, "grad_norm": 0.6558639538447277, "learning_rate": 1.3983696744590215e-07, "loss": 0.0081, "step": 14822 }, { "epoch": 9.483685220729367, "grad_norm": 0.632443276444602, "learning_rate": 1.394918569936865e-07, "loss": 0.0095, "step": 14823 }, { "epoch": 9.484325015994882, "grad_norm": 0.7091037578319938, "learning_rate": 1.3914716993415068e-07, "loss": 0.011, "step": 14824 }, { "epoch": 9.484964811260397, "grad_norm": 0.5465305936244054, "learning_rate": 1.3880290628209504e-07, "loss": 0.0092, "step": 14825 }, { "epoch": 9.485604606525913, "grad_norm": 0.6363975988669668, "learning_rate": 1.3845906605229998e-07, "loss": 0.0094, "step": 14826 }, { "epoch": 9.486244401791426, "grad_norm": 0.5888498447008246, "learning_rate": 1.3811564925953035e-07, "loss": 0.01, "step": 14827 }, { "epoch": 9.486884197056941, "grad_norm": 0.644272690153673, "learning_rate": 1.3777265591853218e-07, "loss": 0.0101, "step": 14828 }, { "epoch": 9.487523992322457, "grad_norm": 0.5717033381065465, "learning_rate": 1.3743008604403142e-07, "loss": 0.0091, "step": 14829 }, { "epoch": 9.488163787587972, "grad_norm": 0.6213461651173824, "learning_rate": 1.3708793965073963e-07, "loss": 0.0096, "step": 14830 }, { "epoch": 9.488803582853487, "grad_norm": 0.5912831518604766, "learning_rate": 1.367462167533473e-07, "loss": 0.0103, "step": 14831 }, { "epoch": 9.489443378119002, "grad_norm": 0.5221738180301121, "learning_rate": 1.3640491736652606e-07, "loss": 0.0079, "step": 14832 }, { "epoch": 9.490083173384518, "grad_norm": 0.5512946080535729, "learning_rate": 1.3606404150493191e-07, "loss": 0.008, "step": 14833 }, { "epoch": 9.490722968650031, "grad_norm": 0.61166840000078, "learning_rate": 1.3572358918320205e-07, "loss": 0.0101, "step": 14834 }, { "epoch": 9.491362763915546, "grad_norm": 0.754705108294979, "learning_rate": 1.3538356041595258e-07, "loss": 0.0087, "step": 14835 }, { "epoch": 9.492002559181062, "grad_norm": 0.5749492009153443, "learning_rate": 1.3504395521778625e-07, "loss": 0.0093, "step": 14836 }, { "epoch": 9.492642354446577, "grad_norm": 0.46069098838647243, "learning_rate": 1.3470477360328472e-07, "loss": 0.0064, "step": 14837 }, { "epoch": 9.493282149712092, "grad_norm": 0.6268288881940416, "learning_rate": 1.343660155870108e-07, "loss": 0.0095, "step": 14838 }, { "epoch": 9.493921944977608, "grad_norm": 0.5460286216881324, "learning_rate": 1.340276811835095e-07, "loss": 0.0081, "step": 14839 }, { "epoch": 9.494561740243123, "grad_norm": 0.5379687841227869, "learning_rate": 1.3368977040731034e-07, "loss": 0.0085, "step": 14840 }, { "epoch": 9.495201535508638, "grad_norm": 0.5139636546430996, "learning_rate": 1.333522832729217e-07, "loss": 0.0078, "step": 14841 }, { "epoch": 9.495841330774152, "grad_norm": 0.4883635551455335, "learning_rate": 1.3301521979483423e-07, "loss": 0.0063, "step": 14842 }, { "epoch": 9.496481126039667, "grad_norm": 0.6482212367767162, "learning_rate": 1.3267857998752186e-07, "loss": 0.0102, "step": 14843 }, { "epoch": 9.497120921305182, "grad_norm": 0.5932419952098265, "learning_rate": 1.3234236386543752e-07, "loss": 0.0101, "step": 14844 }, { "epoch": 9.497760716570697, "grad_norm": 0.8481136853623307, "learning_rate": 1.3200657144301965e-07, "loss": 0.0098, "step": 14845 }, { "epoch": 9.498400511836213, "grad_norm": 0.5710246569505962, "learning_rate": 1.316712027346856e-07, "loss": 0.0086, "step": 14846 }, { "epoch": 9.499040307101728, "grad_norm": 0.5657841307276552, "learning_rate": 1.3133625775483604e-07, "loss": 0.0079, "step": 14847 }, { "epoch": 9.499680102367243, "grad_norm": 0.4802732527591258, "learning_rate": 1.3100173651785174e-07, "loss": 0.0074, "step": 14848 }, { "epoch": 9.500319897632757, "grad_norm": 0.5212812616910847, "learning_rate": 1.3066763903809788e-07, "loss": 0.0076, "step": 14849 }, { "epoch": 9.500959692898272, "grad_norm": 0.7347663573143418, "learning_rate": 1.303339653299196e-07, "loss": 0.0103, "step": 14850 }, { "epoch": 9.501599488163787, "grad_norm": 0.5516128238985856, "learning_rate": 1.3000071540764326e-07, "loss": 0.0085, "step": 14851 }, { "epoch": 9.502239283429303, "grad_norm": 0.5793610146285201, "learning_rate": 1.296678892855796e-07, "loss": 0.01, "step": 14852 }, { "epoch": 9.502879078694818, "grad_norm": 0.600716095914811, "learning_rate": 1.2933548697801834e-07, "loss": 0.0101, "step": 14853 }, { "epoch": 9.503518873960333, "grad_norm": 0.568810219763166, "learning_rate": 1.290035084992336e-07, "loss": 0.0093, "step": 14854 }, { "epoch": 9.504158669225848, "grad_norm": 0.6450890953782153, "learning_rate": 1.2867195386347843e-07, "loss": 0.0105, "step": 14855 }, { "epoch": 9.504798464491362, "grad_norm": 0.649534605120271, "learning_rate": 1.283408230849903e-07, "loss": 0.0109, "step": 14856 }, { "epoch": 9.505438259756877, "grad_norm": 0.6683996928076813, "learning_rate": 1.2801011617798676e-07, "loss": 0.0107, "step": 14857 }, { "epoch": 9.506078055022392, "grad_norm": 0.5740243496980052, "learning_rate": 1.2767983315666755e-07, "loss": 0.0095, "step": 14858 }, { "epoch": 9.506717850287908, "grad_norm": 0.4893608716389077, "learning_rate": 1.2734997403521465e-07, "loss": 0.0079, "step": 14859 }, { "epoch": 9.507357645553423, "grad_norm": 0.5631612492780318, "learning_rate": 1.270205388277923e-07, "loss": 0.0089, "step": 14860 }, { "epoch": 9.507997440818938, "grad_norm": 0.7047333881317644, "learning_rate": 1.2669152754854474e-07, "loss": 0.009, "step": 14861 }, { "epoch": 9.508637236084454, "grad_norm": 0.5875134995651572, "learning_rate": 1.2636294021159955e-07, "loss": 0.0087, "step": 14862 }, { "epoch": 9.509277031349969, "grad_norm": 0.5713859849050809, "learning_rate": 1.2603477683106657e-07, "loss": 0.0083, "step": 14863 }, { "epoch": 9.509916826615482, "grad_norm": 0.5096928626024069, "learning_rate": 1.2570703742103562e-07, "loss": 0.0077, "step": 14864 }, { "epoch": 9.510556621880998, "grad_norm": 0.6535850607575996, "learning_rate": 1.2537972199557992e-07, "loss": 0.0095, "step": 14865 }, { "epoch": 9.511196417146513, "grad_norm": 0.6579868048088942, "learning_rate": 1.2505283056875263e-07, "loss": 0.0088, "step": 14866 }, { "epoch": 9.511836212412028, "grad_norm": 0.7851938817903625, "learning_rate": 1.2472636315459142e-07, "loss": 0.0104, "step": 14867 }, { "epoch": 9.512476007677543, "grad_norm": 0.5159263528927497, "learning_rate": 1.2440031976711287e-07, "loss": 0.0071, "step": 14868 }, { "epoch": 9.513115802943059, "grad_norm": 0.5123427925345111, "learning_rate": 1.240747004203169e-07, "loss": 0.008, "step": 14869 }, { "epoch": 9.513755598208574, "grad_norm": 0.5662615783802595, "learning_rate": 1.237495051281856e-07, "loss": 0.0076, "step": 14870 }, { "epoch": 9.514395393474087, "grad_norm": 0.5266716383182625, "learning_rate": 1.2342473390468235e-07, "loss": 0.0079, "step": 14871 }, { "epoch": 9.515035188739603, "grad_norm": 0.588395568922333, "learning_rate": 1.2310038676375257e-07, "loss": 0.0091, "step": 14872 }, { "epoch": 9.515674984005118, "grad_norm": 0.6108108177200866, "learning_rate": 1.2277646371932184e-07, "loss": 0.0097, "step": 14873 }, { "epoch": 9.516314779270633, "grad_norm": 0.5728540388121964, "learning_rate": 1.2245296478530012e-07, "loss": 0.0085, "step": 14874 }, { "epoch": 9.516954574536149, "grad_norm": 0.4840129229852942, "learning_rate": 1.221298899755774e-07, "loss": 0.007, "step": 14875 }, { "epoch": 9.517594369801664, "grad_norm": 0.5832830036873918, "learning_rate": 1.2180723930402373e-07, "loss": 0.0085, "step": 14876 }, { "epoch": 9.51823416506718, "grad_norm": 0.46467689253350875, "learning_rate": 1.2148501278449687e-07, "loss": 0.0059, "step": 14877 }, { "epoch": 9.518873960332694, "grad_norm": 0.5574894130171774, "learning_rate": 1.2116321043083135e-07, "loss": 0.0083, "step": 14878 }, { "epoch": 9.519513755598208, "grad_norm": 0.6898531995100239, "learning_rate": 1.2084183225684386e-07, "loss": 0.0098, "step": 14879 }, { "epoch": 9.520153550863723, "grad_norm": 0.6507572476231889, "learning_rate": 1.2052087827633453e-07, "loss": 0.0097, "step": 14880 }, { "epoch": 9.520793346129238, "grad_norm": 0.6643963595302179, "learning_rate": 1.202003485030845e-07, "loss": 0.01, "step": 14881 }, { "epoch": 9.521433141394754, "grad_norm": 0.5658435182715746, "learning_rate": 1.198802429508572e-07, "loss": 0.0074, "step": 14882 }, { "epoch": 9.522072936660269, "grad_norm": 0.5743710877398585, "learning_rate": 1.1956056163339614e-07, "loss": 0.0087, "step": 14883 }, { "epoch": 9.522712731925784, "grad_norm": 0.4952744664914847, "learning_rate": 1.1924130456442916e-07, "loss": 0.0081, "step": 14884 }, { "epoch": 9.5233525271913, "grad_norm": 0.7093336283130031, "learning_rate": 1.1892247175766425e-07, "loss": 0.0117, "step": 14885 }, { "epoch": 9.523992322456813, "grad_norm": 0.518340877150621, "learning_rate": 1.1860406322679041e-07, "loss": 0.0075, "step": 14886 }, { "epoch": 9.524632117722328, "grad_norm": 0.5110931392637693, "learning_rate": 1.1828607898548117e-07, "loss": 0.0062, "step": 14887 }, { "epoch": 9.525271912987844, "grad_norm": 0.6175821037039941, "learning_rate": 1.1796851904739004e-07, "loss": 0.0087, "step": 14888 }, { "epoch": 9.525911708253359, "grad_norm": 0.6669122076843498, "learning_rate": 1.1765138342615057e-07, "loss": 0.0095, "step": 14889 }, { "epoch": 9.526551503518874, "grad_norm": 0.5416661136426053, "learning_rate": 1.1733467213538186e-07, "loss": 0.0077, "step": 14890 }, { "epoch": 9.52719129878439, "grad_norm": 0.5428975495873186, "learning_rate": 1.1701838518868302e-07, "loss": 0.0089, "step": 14891 }, { "epoch": 9.527831094049905, "grad_norm": 0.5799444077315371, "learning_rate": 1.1670252259963321e-07, "loss": 0.0097, "step": 14892 }, { "epoch": 9.528470889315418, "grad_norm": 0.5220890336106451, "learning_rate": 1.1638708438179713e-07, "loss": 0.0086, "step": 14893 }, { "epoch": 9.529110684580933, "grad_norm": 0.5961069977445843, "learning_rate": 1.1607207054871727e-07, "loss": 0.0087, "step": 14894 }, { "epoch": 9.529750479846449, "grad_norm": 0.551787520442045, "learning_rate": 1.1575748111392172e-07, "loss": 0.0077, "step": 14895 }, { "epoch": 9.530390275111964, "grad_norm": 0.5239797559183115, "learning_rate": 1.1544331609091519e-07, "loss": 0.0075, "step": 14896 }, { "epoch": 9.53103007037748, "grad_norm": 0.5281414603301402, "learning_rate": 1.1512957549319137e-07, "loss": 0.0093, "step": 14897 }, { "epoch": 9.531669865642995, "grad_norm": 0.4575556863397211, "learning_rate": 1.1481625933421836e-07, "loss": 0.0065, "step": 14898 }, { "epoch": 9.53230966090851, "grad_norm": 0.5509626365621874, "learning_rate": 1.1450336762745096e-07, "loss": 0.0082, "step": 14899 }, { "epoch": 9.532949456174025, "grad_norm": 0.5826039106746849, "learning_rate": 1.1419090038632507e-07, "loss": 0.0081, "step": 14900 }, { "epoch": 9.533589251439539, "grad_norm": 0.5594720690365003, "learning_rate": 1.1387885762425555e-07, "loss": 0.0092, "step": 14901 }, { "epoch": 9.534229046705054, "grad_norm": 0.5175866340834587, "learning_rate": 1.1356723935464275e-07, "loss": 0.0087, "step": 14902 }, { "epoch": 9.53486884197057, "grad_norm": 0.6690007126750106, "learning_rate": 1.1325604559086489e-07, "loss": 0.0091, "step": 14903 }, { "epoch": 9.535508637236084, "grad_norm": 0.6101558029939211, "learning_rate": 1.1294527634628682e-07, "loss": 0.0074, "step": 14904 }, { "epoch": 9.5361484325016, "grad_norm": 0.6081827627358644, "learning_rate": 1.126349316342501e-07, "loss": 0.0076, "step": 14905 }, { "epoch": 9.536788227767115, "grad_norm": 0.5749342258848853, "learning_rate": 1.1232501146807961e-07, "loss": 0.0093, "step": 14906 }, { "epoch": 9.53742802303263, "grad_norm": 0.5734864424150863, "learning_rate": 1.1201551586108584e-07, "loss": 0.0091, "step": 14907 }, { "epoch": 9.538067818298144, "grad_norm": 0.5385249428618674, "learning_rate": 1.1170644482655701e-07, "loss": 0.0078, "step": 14908 }, { "epoch": 9.538707613563659, "grad_norm": 0.5416874981550176, "learning_rate": 1.1139779837776254e-07, "loss": 0.0084, "step": 14909 }, { "epoch": 9.539347408829174, "grad_norm": 0.6987575737751442, "learning_rate": 1.1108957652795627e-07, "loss": 0.0116, "step": 14910 }, { "epoch": 9.53998720409469, "grad_norm": 0.6918691997507534, "learning_rate": 1.1078177929037315e-07, "loss": 0.0087, "step": 14911 }, { "epoch": 9.540626999360205, "grad_norm": 0.5758380856910719, "learning_rate": 1.1047440667822818e-07, "loss": 0.0091, "step": 14912 }, { "epoch": 9.54126679462572, "grad_norm": 0.607722977405491, "learning_rate": 1.1016745870471968e-07, "loss": 0.0084, "step": 14913 }, { "epoch": 9.541906589891235, "grad_norm": 0.6190459643883581, "learning_rate": 1.0986093538302933e-07, "loss": 0.0094, "step": 14914 }, { "epoch": 9.54254638515675, "grad_norm": 0.6124834247094592, "learning_rate": 1.0955483672631662e-07, "loss": 0.009, "step": 14915 }, { "epoch": 9.543186180422264, "grad_norm": 0.5477962885818978, "learning_rate": 1.0924916274772546e-07, "loss": 0.0074, "step": 14916 }, { "epoch": 9.54382597568778, "grad_norm": 0.5365337356219136, "learning_rate": 1.0894391346038091e-07, "loss": 0.0072, "step": 14917 }, { "epoch": 9.544465770953295, "grad_norm": 0.6066898456368977, "learning_rate": 1.0863908887738916e-07, "loss": 0.0092, "step": 14918 }, { "epoch": 9.54510556621881, "grad_norm": 0.550934981446949, "learning_rate": 1.0833468901183974e-07, "loss": 0.0083, "step": 14919 }, { "epoch": 9.545745361484325, "grad_norm": 0.6538774387290848, "learning_rate": 1.0803071387680442e-07, "loss": 0.0108, "step": 14920 }, { "epoch": 9.54638515674984, "grad_norm": 0.6125180401626165, "learning_rate": 1.0772716348533274e-07, "loss": 0.0082, "step": 14921 }, { "epoch": 9.547024952015356, "grad_norm": 0.48176328049723954, "learning_rate": 1.0742403785045874e-07, "loss": 0.0077, "step": 14922 }, { "epoch": 9.54766474728087, "grad_norm": 0.5795150628386079, "learning_rate": 1.0712133698519978e-07, "loss": 0.0074, "step": 14923 }, { "epoch": 9.548304542546385, "grad_norm": 0.4581722226631079, "learning_rate": 1.0681906090255323e-07, "loss": 0.0067, "step": 14924 }, { "epoch": 9.5489443378119, "grad_norm": 0.5816821318022455, "learning_rate": 1.0651720961549761e-07, "loss": 0.0092, "step": 14925 }, { "epoch": 9.549584133077415, "grad_norm": 0.5178886367371628, "learning_rate": 1.0621578313699254e-07, "loss": 0.0069, "step": 14926 }, { "epoch": 9.55022392834293, "grad_norm": 0.5521617179439655, "learning_rate": 1.0591478147998324e-07, "loss": 0.008, "step": 14927 }, { "epoch": 9.550863723608446, "grad_norm": 0.5679362032399925, "learning_rate": 1.056142046573927e-07, "loss": 0.008, "step": 14928 }, { "epoch": 9.551503518873961, "grad_norm": 0.6023339129422068, "learning_rate": 1.0531405268212613e-07, "loss": 0.01, "step": 14929 }, { "epoch": 9.552143314139474, "grad_norm": 0.550253806536982, "learning_rate": 1.0501432556707436e-07, "loss": 0.0081, "step": 14930 }, { "epoch": 9.55278310940499, "grad_norm": 0.6646876114169857, "learning_rate": 1.0471502332510597e-07, "loss": 0.0092, "step": 14931 }, { "epoch": 9.553422904670505, "grad_norm": 0.6683206150689248, "learning_rate": 1.044161459690718e-07, "loss": 0.011, "step": 14932 }, { "epoch": 9.55406269993602, "grad_norm": 0.6629797833603387, "learning_rate": 1.0411769351180489e-07, "loss": 0.0095, "step": 14933 }, { "epoch": 9.554702495201536, "grad_norm": 0.5934815166423546, "learning_rate": 1.0381966596612169e-07, "loss": 0.0073, "step": 14934 }, { "epoch": 9.55534229046705, "grad_norm": 0.5666026819133557, "learning_rate": 1.0352206334481751e-07, "loss": 0.0088, "step": 14935 }, { "epoch": 9.555982085732566, "grad_norm": 0.5274161330741768, "learning_rate": 1.0322488566067101e-07, "loss": 0.0087, "step": 14936 }, { "epoch": 9.556621880998081, "grad_norm": 0.49067540895011363, "learning_rate": 1.029281329264431e-07, "loss": 0.0067, "step": 14937 }, { "epoch": 9.557261676263595, "grad_norm": 0.6871658224467133, "learning_rate": 1.026318051548758e-07, "loss": 0.0102, "step": 14938 }, { "epoch": 9.55790147152911, "grad_norm": 0.5751416205297618, "learning_rate": 1.0233590235869229e-07, "loss": 0.0087, "step": 14939 }, { "epoch": 9.558541266794625, "grad_norm": 0.6598995402689742, "learning_rate": 1.0204042455060014e-07, "loss": 0.0095, "step": 14940 }, { "epoch": 9.55918106206014, "grad_norm": 0.6669244677091377, "learning_rate": 1.0174537174328369e-07, "loss": 0.0097, "step": 14941 }, { "epoch": 9.559820857325656, "grad_norm": 0.606639970836292, "learning_rate": 1.0145074394941389e-07, "loss": 0.0082, "step": 14942 }, { "epoch": 9.560460652591171, "grad_norm": 0.589820101664745, "learning_rate": 1.0115654118164064e-07, "loss": 0.01, "step": 14943 }, { "epoch": 9.561100447856687, "grad_norm": 0.5467145136401858, "learning_rate": 1.0086276345259716e-07, "loss": 0.0081, "step": 14944 }, { "epoch": 9.561740243122202, "grad_norm": 0.6477365622209349, "learning_rate": 1.0056941077489668e-07, "loss": 0.0107, "step": 14945 }, { "epoch": 9.562380038387715, "grad_norm": 0.6170773959807362, "learning_rate": 1.0027648316113692e-07, "loss": 0.0104, "step": 14946 }, { "epoch": 9.56301983365323, "grad_norm": 0.5968907278653623, "learning_rate": 9.998398062389448e-08, "loss": 0.0081, "step": 14947 }, { "epoch": 9.563659628918746, "grad_norm": 0.6696438041491506, "learning_rate": 9.96919031757293e-08, "loss": 0.0119, "step": 14948 }, { "epoch": 9.564299424184261, "grad_norm": 0.4727321653778017, "learning_rate": 9.940025082918137e-08, "loss": 0.0069, "step": 14949 }, { "epoch": 9.564939219449776, "grad_norm": 0.6548692669745968, "learning_rate": 9.910902359677621e-08, "loss": 0.0109, "step": 14950 }, { "epoch": 9.565579014715292, "grad_norm": 0.9057836957013456, "learning_rate": 9.881822149101716e-08, "loss": 0.0121, "step": 14951 }, { "epoch": 9.566218809980807, "grad_norm": 0.5688340427653302, "learning_rate": 9.85278445243898e-08, "loss": 0.0084, "step": 14952 }, { "epoch": 9.56685860524632, "grad_norm": 0.5997381272988268, "learning_rate": 9.823789270936413e-08, "loss": 0.0082, "step": 14953 }, { "epoch": 9.567498400511836, "grad_norm": 0.668057129332405, "learning_rate": 9.79483660583891e-08, "loss": 0.0094, "step": 14954 }, { "epoch": 9.568138195777351, "grad_norm": 0.507056061385786, "learning_rate": 9.765926458389808e-08, "loss": 0.0067, "step": 14955 }, { "epoch": 9.568777991042866, "grad_norm": 0.6111521595630242, "learning_rate": 9.737058829830226e-08, "loss": 0.0086, "step": 14956 }, { "epoch": 9.569417786308382, "grad_norm": 0.4970310827964699, "learning_rate": 9.708233721399729e-08, "loss": 0.0061, "step": 14957 }, { "epoch": 9.570057581573897, "grad_norm": 0.5394589306446517, "learning_rate": 9.679451134336216e-08, "loss": 0.0077, "step": 14958 }, { "epoch": 9.570697376839412, "grad_norm": 0.5379170022669072, "learning_rate": 9.650711069875363e-08, "loss": 0.0082, "step": 14959 }, { "epoch": 9.571337172104926, "grad_norm": 0.5569871150105467, "learning_rate": 9.622013529251295e-08, "loss": 0.0087, "step": 14960 }, { "epoch": 9.571976967370441, "grad_norm": 0.5754660230191618, "learning_rate": 9.59335851369625e-08, "loss": 0.0085, "step": 14961 }, { "epoch": 9.572616762635956, "grad_norm": 0.6243972442779081, "learning_rate": 9.564746024440574e-08, "loss": 0.0089, "step": 14962 }, { "epoch": 9.573256557901471, "grad_norm": 0.6610413681607488, "learning_rate": 9.536176062712843e-08, "loss": 0.0093, "step": 14963 }, { "epoch": 9.573896353166987, "grad_norm": 0.6030810399468901, "learning_rate": 9.507648629739852e-08, "loss": 0.0086, "step": 14964 }, { "epoch": 9.574536148432502, "grad_norm": 0.6224814581223196, "learning_rate": 9.479163726746509e-08, "loss": 0.0097, "step": 14965 }, { "epoch": 9.575175943698017, "grad_norm": 0.6878162479851259, "learning_rate": 9.450721354955727e-08, "loss": 0.0104, "step": 14966 }, { "epoch": 9.57581573896353, "grad_norm": 0.5328701404454691, "learning_rate": 9.422321515589084e-08, "loss": 0.0083, "step": 14967 }, { "epoch": 9.576455534229046, "grad_norm": 0.6093031444975665, "learning_rate": 9.393964209865713e-08, "loss": 0.01, "step": 14968 }, { "epoch": 9.577095329494561, "grad_norm": 0.6141355464031841, "learning_rate": 9.365649439003421e-08, "loss": 0.0111, "step": 14969 }, { "epoch": 9.577735124760077, "grad_norm": 0.6822879594549416, "learning_rate": 9.337377204218012e-08, "loss": 0.0088, "step": 14970 }, { "epoch": 9.578374920025592, "grad_norm": 0.5396508761812505, "learning_rate": 9.309147506723293e-08, "loss": 0.0086, "step": 14971 }, { "epoch": 9.579014715291107, "grad_norm": 0.6620412817179403, "learning_rate": 9.280960347731516e-08, "loss": 0.0094, "step": 14972 }, { "epoch": 9.579654510556622, "grad_norm": 0.5078732231164264, "learning_rate": 9.252815728452935e-08, "loss": 0.0078, "step": 14973 }, { "epoch": 9.580294305822138, "grad_norm": 0.4971387112057411, "learning_rate": 9.22471365009603e-08, "loss": 0.007, "step": 14974 }, { "epoch": 9.580934101087651, "grad_norm": 0.6723054256448853, "learning_rate": 9.196654113867498e-08, "loss": 0.0092, "step": 14975 }, { "epoch": 9.581573896353166, "grad_norm": 0.49585204092816526, "learning_rate": 9.168637120972157e-08, "loss": 0.0073, "step": 14976 }, { "epoch": 9.582213691618682, "grad_norm": 0.5431859006245375, "learning_rate": 9.140662672612932e-08, "loss": 0.0076, "step": 14977 }, { "epoch": 9.582853486884197, "grad_norm": 0.6267698190169908, "learning_rate": 9.112730769991085e-08, "loss": 0.0084, "step": 14978 }, { "epoch": 9.583493282149712, "grad_norm": 0.6226822007028912, "learning_rate": 9.084841414305878e-08, "loss": 0.008, "step": 14979 }, { "epoch": 9.584133077415228, "grad_norm": 0.7752869667831449, "learning_rate": 9.05699460675491e-08, "loss": 0.0078, "step": 14980 }, { "epoch": 9.584772872680743, "grad_norm": 0.6846644460228447, "learning_rate": 9.029190348533779e-08, "loss": 0.0103, "step": 14981 }, { "epoch": 9.585412667946258, "grad_norm": 0.5775247355664546, "learning_rate": 9.00142864083653e-08, "loss": 0.0091, "step": 14982 }, { "epoch": 9.586052463211772, "grad_norm": 0.5012834036317184, "learning_rate": 8.973709484854987e-08, "loss": 0.0077, "step": 14983 }, { "epoch": 9.586692258477287, "grad_norm": 0.5390470382649196, "learning_rate": 8.946032881779532e-08, "loss": 0.0066, "step": 14984 }, { "epoch": 9.587332053742802, "grad_norm": 0.5696465233724658, "learning_rate": 8.918398832798436e-08, "loss": 0.0081, "step": 14985 }, { "epoch": 9.587971849008317, "grad_norm": 0.6983328881711666, "learning_rate": 8.890807339098307e-08, "loss": 0.0102, "step": 14986 }, { "epoch": 9.588611644273833, "grad_norm": 0.548768609709792, "learning_rate": 8.863258401863861e-08, "loss": 0.0089, "step": 14987 }, { "epoch": 9.589251439539348, "grad_norm": 0.5145036545291434, "learning_rate": 8.835752022278043e-08, "loss": 0.0072, "step": 14988 }, { "epoch": 9.589891234804863, "grad_norm": 0.6294213654507335, "learning_rate": 8.808288201521797e-08, "loss": 0.0098, "step": 14989 }, { "epoch": 9.590531030070377, "grad_norm": 0.5477317166543477, "learning_rate": 8.780866940774512e-08, "loss": 0.0088, "step": 14990 }, { "epoch": 9.591170825335892, "grad_norm": 0.6137244955726527, "learning_rate": 8.753488241213581e-08, "loss": 0.0099, "step": 14991 }, { "epoch": 9.591810620601407, "grad_norm": 0.6292717168465052, "learning_rate": 8.726152104014619e-08, "loss": 0.0091, "step": 14992 }, { "epoch": 9.592450415866923, "grad_norm": 0.5389010265901844, "learning_rate": 8.69885853035124e-08, "loss": 0.0068, "step": 14993 }, { "epoch": 9.593090211132438, "grad_norm": 0.6325646914336024, "learning_rate": 8.67160752139562e-08, "loss": 0.0107, "step": 14994 }, { "epoch": 9.593730006397953, "grad_norm": 0.5564278779804672, "learning_rate": 8.6443990783176e-08, "loss": 0.0099, "step": 14995 }, { "epoch": 9.594369801663468, "grad_norm": 0.6386167950071614, "learning_rate": 8.61723320228569e-08, "loss": 0.0085, "step": 14996 }, { "epoch": 9.595009596928982, "grad_norm": 0.6291799701218738, "learning_rate": 8.590109894466181e-08, "loss": 0.0086, "step": 14997 }, { "epoch": 9.595649392194497, "grad_norm": 0.5070085564958041, "learning_rate": 8.563029156023805e-08, "loss": 0.0078, "step": 14998 }, { "epoch": 9.596289187460012, "grad_norm": 0.6010611581557204, "learning_rate": 8.535990988121301e-08, "loss": 0.0099, "step": 14999 }, { "epoch": 9.596928982725528, "grad_norm": 0.5643865105666188, "learning_rate": 8.508995391919739e-08, "loss": 0.008, "step": 15000 }, { "epoch": 9.597568777991043, "grad_norm": 0.5250487600756806, "learning_rate": 8.482042368578081e-08, "loss": 0.0088, "step": 15001 }, { "epoch": 9.598208573256558, "grad_norm": 0.5728137482659486, "learning_rate": 8.455131919253846e-08, "loss": 0.0084, "step": 15002 }, { "epoch": 9.598848368522074, "grad_norm": 0.712029024606083, "learning_rate": 8.42826404510233e-08, "loss": 0.0112, "step": 15003 }, { "epoch": 9.599488163787587, "grad_norm": 0.5909118333802043, "learning_rate": 8.40143874727739e-08, "loss": 0.0088, "step": 15004 }, { "epoch": 9.600127959053102, "grad_norm": 0.6764793458770048, "learning_rate": 8.374656026930771e-08, "loss": 0.0113, "step": 15005 }, { "epoch": 9.600767754318618, "grad_norm": 0.5948700342485895, "learning_rate": 8.347915885212332e-08, "loss": 0.0077, "step": 15006 }, { "epoch": 9.601407549584133, "grad_norm": 0.5776333065419267, "learning_rate": 8.321218323270485e-08, "loss": 0.0094, "step": 15007 }, { "epoch": 9.602047344849648, "grad_norm": 0.5921184695929596, "learning_rate": 8.294563342251427e-08, "loss": 0.0088, "step": 15008 }, { "epoch": 9.602687140115163, "grad_norm": 0.4841214871683627, "learning_rate": 8.267950943299685e-08, "loss": 0.0067, "step": 15009 }, { "epoch": 9.603326935380679, "grad_norm": 0.5275913859944735, "learning_rate": 8.241381127558012e-08, "loss": 0.0088, "step": 15010 }, { "epoch": 9.603966730646194, "grad_norm": 0.5775452811731039, "learning_rate": 8.214853896167163e-08, "loss": 0.0086, "step": 15011 }, { "epoch": 9.604606525911707, "grad_norm": 0.6198473536861758, "learning_rate": 8.188369250266337e-08, "loss": 0.0082, "step": 15012 }, { "epoch": 9.605246321177223, "grad_norm": 0.603306365897721, "learning_rate": 8.161927190992513e-08, "loss": 0.0081, "step": 15013 }, { "epoch": 9.605886116442738, "grad_norm": 0.5143683512949103, "learning_rate": 8.13552771948134e-08, "loss": 0.0074, "step": 15014 }, { "epoch": 9.606525911708253, "grad_norm": 0.6413649309915593, "learning_rate": 8.109170836866132e-08, "loss": 0.0101, "step": 15015 }, { "epoch": 9.607165706973769, "grad_norm": 0.6065775318245565, "learning_rate": 8.08285654427865e-08, "loss": 0.0098, "step": 15016 }, { "epoch": 9.607805502239284, "grad_norm": 0.6079664573803878, "learning_rate": 8.056584842848881e-08, "loss": 0.0083, "step": 15017 }, { "epoch": 9.6084452975048, "grad_norm": 0.6121737322366269, "learning_rate": 8.03035573370492e-08, "loss": 0.0088, "step": 15018 }, { "epoch": 9.609085092770314, "grad_norm": 0.607052436099729, "learning_rate": 8.004169217972757e-08, "loss": 0.0096, "step": 15019 }, { "epoch": 9.609724888035828, "grad_norm": 0.644529591035085, "learning_rate": 7.978025296776937e-08, "loss": 0.0106, "step": 15020 }, { "epoch": 9.610364683301343, "grad_norm": 0.6664977816090245, "learning_rate": 7.951923971240117e-08, "loss": 0.0107, "step": 15021 }, { "epoch": 9.611004478566858, "grad_norm": 0.639874031438814, "learning_rate": 7.925865242482955e-08, "loss": 0.01, "step": 15022 }, { "epoch": 9.611644273832374, "grad_norm": 0.5792519875769899, "learning_rate": 7.899849111624447e-08, "loss": 0.0087, "step": 15023 }, { "epoch": 9.612284069097889, "grad_norm": 0.5397207868703104, "learning_rate": 7.873875579781587e-08, "loss": 0.0081, "step": 15024 }, { "epoch": 9.612923864363404, "grad_norm": 0.5534764977011835, "learning_rate": 7.847944648069594e-08, "loss": 0.0075, "step": 15025 }, { "epoch": 9.61356365962892, "grad_norm": 0.4941922198366157, "learning_rate": 7.822056317601911e-08, "loss": 0.0065, "step": 15026 }, { "epoch": 9.614203454894433, "grad_norm": 0.572526797887154, "learning_rate": 7.796210589490205e-08, "loss": 0.0093, "step": 15027 }, { "epoch": 9.614843250159948, "grad_norm": 0.5475571656870529, "learning_rate": 7.770407464844254e-08, "loss": 0.0082, "step": 15028 }, { "epoch": 9.615483045425464, "grad_norm": 0.5252377124923032, "learning_rate": 7.744646944772061e-08, "loss": 0.0072, "step": 15029 }, { "epoch": 9.616122840690979, "grad_norm": 0.576525907709845, "learning_rate": 7.718929030379518e-08, "loss": 0.0097, "step": 15030 }, { "epoch": 9.616762635956494, "grad_norm": 0.6018001448936917, "learning_rate": 7.693253722770966e-08, "loss": 0.01, "step": 15031 }, { "epoch": 9.61740243122201, "grad_norm": 0.4975619339828976, "learning_rate": 7.667621023049077e-08, "loss": 0.0074, "step": 15032 }, { "epoch": 9.618042226487525, "grad_norm": 0.6941459598026541, "learning_rate": 7.642030932314193e-08, "loss": 0.0099, "step": 15033 }, { "epoch": 9.618682021753038, "grad_norm": 0.6435689113145141, "learning_rate": 7.616483451665324e-08, "loss": 0.0107, "step": 15034 }, { "epoch": 9.619321817018553, "grad_norm": 0.5899265101009287, "learning_rate": 7.590978582199371e-08, "loss": 0.011, "step": 15035 }, { "epoch": 9.619961612284069, "grad_norm": 0.5873629835337781, "learning_rate": 7.565516325011346e-08, "loss": 0.0097, "step": 15036 }, { "epoch": 9.620601407549584, "grad_norm": 0.5285089026184874, "learning_rate": 7.540096681194708e-08, "loss": 0.0082, "step": 15037 }, { "epoch": 9.6212412028151, "grad_norm": 0.647003553968383, "learning_rate": 7.514719651840918e-08, "loss": 0.009, "step": 15038 }, { "epoch": 9.621880998080615, "grad_norm": 0.5985814672069033, "learning_rate": 7.489385238039437e-08, "loss": 0.0093, "step": 15039 }, { "epoch": 9.62252079334613, "grad_norm": 0.6206026590225583, "learning_rate": 7.464093440878284e-08, "loss": 0.0099, "step": 15040 }, { "epoch": 9.623160588611643, "grad_norm": 0.6263393569449448, "learning_rate": 7.438844261443367e-08, "loss": 0.0093, "step": 15041 }, { "epoch": 9.623800383877159, "grad_norm": 0.6556755797102964, "learning_rate": 7.413637700818822e-08, "loss": 0.0089, "step": 15042 }, { "epoch": 9.624440179142674, "grad_norm": 0.6750554012488531, "learning_rate": 7.388473760087001e-08, "loss": 0.0097, "step": 15043 }, { "epoch": 9.62507997440819, "grad_norm": 0.6914422151114828, "learning_rate": 7.363352440328486e-08, "loss": 0.0104, "step": 15044 }, { "epoch": 9.625719769673704, "grad_norm": 0.5250399354068718, "learning_rate": 7.338273742621749e-08, "loss": 0.0078, "step": 15045 }, { "epoch": 9.62635956493922, "grad_norm": 0.5512264734829453, "learning_rate": 7.313237668043816e-08, "loss": 0.0074, "step": 15046 }, { "epoch": 9.626999360204735, "grad_norm": 0.478713110839079, "learning_rate": 7.288244217669493e-08, "loss": 0.0063, "step": 15047 }, { "epoch": 9.62763915547025, "grad_norm": 0.4554002601731514, "learning_rate": 7.263293392572146e-08, "loss": 0.0061, "step": 15048 }, { "epoch": 9.628278950735764, "grad_norm": 0.6550109896774899, "learning_rate": 7.238385193823028e-08, "loss": 0.0087, "step": 15049 }, { "epoch": 9.628918746001279, "grad_norm": 0.6038598514447929, "learning_rate": 7.213519622491726e-08, "loss": 0.0084, "step": 15050 }, { "epoch": 9.629558541266794, "grad_norm": 0.8506131887187769, "learning_rate": 7.188696679645724e-08, "loss": 0.0083, "step": 15051 }, { "epoch": 9.63019833653231, "grad_norm": 0.561597316521691, "learning_rate": 7.163916366351165e-08, "loss": 0.0092, "step": 15052 }, { "epoch": 9.630838131797825, "grad_norm": 0.5647175775495203, "learning_rate": 7.139178683671865e-08, "loss": 0.009, "step": 15053 }, { "epoch": 9.63147792706334, "grad_norm": 0.6341569551785214, "learning_rate": 7.114483632669977e-08, "loss": 0.009, "step": 15054 }, { "epoch": 9.632117722328855, "grad_norm": 0.6420054956196719, "learning_rate": 7.089831214406095e-08, "loss": 0.0109, "step": 15055 }, { "epoch": 9.63275751759437, "grad_norm": 0.6294196790959232, "learning_rate": 7.065221429938485e-08, "loss": 0.0102, "step": 15056 }, { "epoch": 9.633397312859884, "grad_norm": 0.6872061745144922, "learning_rate": 7.040654280324077e-08, "loss": 0.0092, "step": 15057 }, { "epoch": 9.6340371081254, "grad_norm": 0.626903724581972, "learning_rate": 7.016129766617586e-08, "loss": 0.0098, "step": 15058 }, { "epoch": 9.634676903390915, "grad_norm": 0.5362498383487247, "learning_rate": 6.991647889872055e-08, "loss": 0.0092, "step": 15059 }, { "epoch": 9.63531669865643, "grad_norm": 0.5943364889305434, "learning_rate": 6.967208651138758e-08, "loss": 0.0097, "step": 15060 }, { "epoch": 9.635956493921945, "grad_norm": 0.6244914622685064, "learning_rate": 6.942812051467073e-08, "loss": 0.01, "step": 15061 }, { "epoch": 9.63659628918746, "grad_norm": 0.5694961340240031, "learning_rate": 6.918458091904612e-08, "loss": 0.0086, "step": 15062 }, { "epoch": 9.637236084452976, "grad_norm": 0.5814910292964883, "learning_rate": 6.894146773496868e-08, "loss": 0.0093, "step": 15063 }, { "epoch": 9.63787587971849, "grad_norm": 0.5171063986641667, "learning_rate": 6.869878097287897e-08, "loss": 0.0078, "step": 15064 }, { "epoch": 9.638515674984005, "grad_norm": 0.5462683415235993, "learning_rate": 6.845652064319752e-08, "loss": 0.0085, "step": 15065 }, { "epoch": 9.63915547024952, "grad_norm": 0.5622762167277203, "learning_rate": 6.821468675632715e-08, "loss": 0.0081, "step": 15066 }, { "epoch": 9.639795265515035, "grad_norm": 0.6804025935865189, "learning_rate": 6.797327932264952e-08, "loss": 0.0096, "step": 15067 }, { "epoch": 9.64043506078055, "grad_norm": 0.5962524232343983, "learning_rate": 6.773229835253192e-08, "loss": 0.0113, "step": 15068 }, { "epoch": 9.641074856046066, "grad_norm": 0.6545166708786316, "learning_rate": 6.749174385632162e-08, "loss": 0.0109, "step": 15069 }, { "epoch": 9.641714651311581, "grad_norm": 0.5363461672360438, "learning_rate": 6.725161584434591e-08, "loss": 0.0069, "step": 15070 }, { "epoch": 9.642354446577095, "grad_norm": 0.5485304949863254, "learning_rate": 6.701191432691878e-08, "loss": 0.0078, "step": 15071 }, { "epoch": 9.64299424184261, "grad_norm": 0.5558013057741559, "learning_rate": 6.677263931432976e-08, "loss": 0.0098, "step": 15072 }, { "epoch": 9.643634037108125, "grad_norm": 0.5596887696964843, "learning_rate": 6.653379081685284e-08, "loss": 0.0082, "step": 15073 }, { "epoch": 9.64427383237364, "grad_norm": 0.6279604246116567, "learning_rate": 6.629536884474652e-08, "loss": 0.0088, "step": 15074 }, { "epoch": 9.644913627639156, "grad_norm": 1.0961130957757865, "learning_rate": 6.605737340824481e-08, "loss": 0.0076, "step": 15075 }, { "epoch": 9.64555342290467, "grad_norm": 0.566516874424501, "learning_rate": 6.581980451756842e-08, "loss": 0.0085, "step": 15076 }, { "epoch": 9.646193218170186, "grad_norm": 0.566224113951118, "learning_rate": 6.55826621829192e-08, "loss": 0.0085, "step": 15077 }, { "epoch": 9.6468330134357, "grad_norm": 0.542563722643535, "learning_rate": 6.534594641447789e-08, "loss": 0.0077, "step": 15078 }, { "epoch": 9.647472808701215, "grad_norm": 0.5070522285438541, "learning_rate": 6.51096572224097e-08, "loss": 0.0072, "step": 15079 }, { "epoch": 9.64811260396673, "grad_norm": 0.5842836095335358, "learning_rate": 6.487379461685871e-08, "loss": 0.0103, "step": 15080 }, { "epoch": 9.648752399232245, "grad_norm": 0.5531789134951688, "learning_rate": 6.463835860795575e-08, "loss": 0.0081, "step": 15081 }, { "epoch": 9.64939219449776, "grad_norm": 0.7108413894083331, "learning_rate": 6.440334920580715e-08, "loss": 0.011, "step": 15082 }, { "epoch": 9.650031989763276, "grad_norm": 0.5588326766148847, "learning_rate": 6.416876642050373e-08, "loss": 0.0073, "step": 15083 }, { "epoch": 9.650671785028791, "grad_norm": 0.6337887354440909, "learning_rate": 6.393461026212078e-08, "loss": 0.0088, "step": 15084 }, { "epoch": 9.651311580294307, "grad_norm": 0.46928487479518854, "learning_rate": 6.370088074070912e-08, "loss": 0.0063, "step": 15085 }, { "epoch": 9.65195137555982, "grad_norm": 0.6185234213802038, "learning_rate": 6.346757786630742e-08, "loss": 0.0083, "step": 15086 }, { "epoch": 9.652591170825335, "grad_norm": 0.5454870398424833, "learning_rate": 6.323470164893208e-08, "loss": 0.0083, "step": 15087 }, { "epoch": 9.65323096609085, "grad_norm": 0.5001251462494021, "learning_rate": 6.30022520985818e-08, "loss": 0.0061, "step": 15088 }, { "epoch": 9.653870761356366, "grad_norm": 0.5419527673514142, "learning_rate": 6.277022922523856e-08, "loss": 0.0074, "step": 15089 }, { "epoch": 9.654510556621881, "grad_norm": 0.67015002980824, "learning_rate": 6.253863303886443e-08, "loss": 0.0099, "step": 15090 }, { "epoch": 9.655150351887396, "grad_norm": 0.4914774454604649, "learning_rate": 6.230746354940476e-08, "loss": 0.007, "step": 15091 }, { "epoch": 9.655790147152912, "grad_norm": 0.557633369389441, "learning_rate": 6.207672076678495e-08, "loss": 0.0093, "step": 15092 }, { "epoch": 9.656429942418427, "grad_norm": 0.5295655618554851, "learning_rate": 6.184640470091152e-08, "loss": 0.0075, "step": 15093 }, { "epoch": 9.65706973768394, "grad_norm": 0.6176860106861092, "learning_rate": 6.161651536167546e-08, "loss": 0.0088, "step": 15094 }, { "epoch": 9.657709532949456, "grad_norm": 0.5578656922738763, "learning_rate": 6.138705275894774e-08, "loss": 0.0099, "step": 15095 }, { "epoch": 9.658349328214971, "grad_norm": 0.5127749295750553, "learning_rate": 6.11580169025805e-08, "loss": 0.0072, "step": 15096 }, { "epoch": 9.658989123480486, "grad_norm": 0.6163869251657702, "learning_rate": 6.092940780240697e-08, "loss": 0.0091, "step": 15097 }, { "epoch": 9.659628918746002, "grad_norm": 0.6167282103535365, "learning_rate": 6.070122546824597e-08, "loss": 0.0089, "step": 15098 }, { "epoch": 9.660268714011517, "grad_norm": 0.6171066985553278, "learning_rate": 6.0473469909893e-08, "loss": 0.0091, "step": 15099 }, { "epoch": 9.660908509277032, "grad_norm": 0.6705422768021916, "learning_rate": 6.02461411371269e-08, "loss": 0.0107, "step": 15100 }, { "epoch": 9.661548304542546, "grad_norm": 0.6699512461428037, "learning_rate": 6.001923915971208e-08, "loss": 0.0114, "step": 15101 }, { "epoch": 9.662188099808061, "grad_norm": 0.49294064187610365, "learning_rate": 5.979276398738854e-08, "loss": 0.0066, "step": 15102 }, { "epoch": 9.662827895073576, "grad_norm": 0.6131181530890122, "learning_rate": 5.95667156298807e-08, "loss": 0.01, "step": 15103 }, { "epoch": 9.663467690339091, "grad_norm": 0.663621780507459, "learning_rate": 5.9341094096896366e-08, "loss": 0.0106, "step": 15104 }, { "epoch": 9.664107485604607, "grad_norm": 0.7538133625815607, "learning_rate": 5.9115899398122236e-08, "loss": 0.0121, "step": 15105 }, { "epoch": 9.664747280870122, "grad_norm": 0.6413286267912278, "learning_rate": 5.889113154322834e-08, "loss": 0.0103, "step": 15106 }, { "epoch": 9.665387076135637, "grad_norm": 0.6244086479559325, "learning_rate": 5.866679054186364e-08, "loss": 0.0092, "step": 15107 }, { "epoch": 9.66602687140115, "grad_norm": 0.5424221752030758, "learning_rate": 5.844287640366486e-08, "loss": 0.0064, "step": 15108 }, { "epoch": 9.666666666666666, "grad_norm": 0.5837146984877147, "learning_rate": 5.821938913824321e-08, "loss": 0.0077, "step": 15109 }, { "epoch": 9.667306461932181, "grad_norm": 0.6060060607991271, "learning_rate": 5.7996328755195455e-08, "loss": 0.0091, "step": 15110 }, { "epoch": 9.667946257197697, "grad_norm": 0.6151155018056662, "learning_rate": 5.77736952641017e-08, "loss": 0.0089, "step": 15111 }, { "epoch": 9.668586052463212, "grad_norm": 0.5970868317394805, "learning_rate": 5.755148867451876e-08, "loss": 0.0109, "step": 15112 }, { "epoch": 9.669225847728727, "grad_norm": 0.6059554909342381, "learning_rate": 5.7329708995987884e-08, "loss": 0.0094, "step": 15113 }, { "epoch": 9.669865642994242, "grad_norm": 0.539248286693972, "learning_rate": 5.710835623803257e-08, "loss": 0.008, "step": 15114 }, { "epoch": 9.670505438259756, "grad_norm": 0.6043172495735871, "learning_rate": 5.688743041015854e-08, "loss": 0.0093, "step": 15115 }, { "epoch": 9.671145233525271, "grad_norm": 0.6426739157503966, "learning_rate": 5.666693152185043e-08, "loss": 0.0091, "step": 15116 }, { "epoch": 9.671785028790786, "grad_norm": 0.6280704612704054, "learning_rate": 5.644685958257623e-08, "loss": 0.0091, "step": 15117 }, { "epoch": 9.672424824056302, "grad_norm": 0.6308767922395494, "learning_rate": 5.622721460178504e-08, "loss": 0.0085, "step": 15118 }, { "epoch": 9.673064619321817, "grad_norm": 0.7079454429787699, "learning_rate": 5.600799658890932e-08, "loss": 0.0106, "step": 15119 }, { "epoch": 9.673704414587332, "grad_norm": 0.5488304825269197, "learning_rate": 5.5789205553361536e-08, "loss": 0.0076, "step": 15120 }, { "epoch": 9.674344209852848, "grad_norm": 0.6627649814464942, "learning_rate": 5.557084150453529e-08, "loss": 0.0108, "step": 15121 }, { "epoch": 9.674984005118363, "grad_norm": 0.7064875482218842, "learning_rate": 5.535290445180752e-08, "loss": 0.0101, "step": 15122 }, { "epoch": 9.675623800383876, "grad_norm": 0.5665834118735761, "learning_rate": 5.51353944045363e-08, "loss": 0.0092, "step": 15123 }, { "epoch": 9.676263595649392, "grad_norm": 0.5166320268581117, "learning_rate": 5.491831137206083e-08, "loss": 0.0081, "step": 15124 }, { "epoch": 9.676903390914907, "grad_norm": 0.6536889275307219, "learning_rate": 5.4701655363701424e-08, "loss": 0.0093, "step": 15125 }, { "epoch": 9.677543186180422, "grad_norm": 0.6094604343987277, "learning_rate": 5.448542638876175e-08, "loss": 0.0101, "step": 15126 }, { "epoch": 9.678182981445937, "grad_norm": 0.5222706646847676, "learning_rate": 5.426962445652661e-08, "loss": 0.0082, "step": 15127 }, { "epoch": 9.678822776711453, "grad_norm": 0.576377195860403, "learning_rate": 5.405424957626193e-08, "loss": 0.0082, "step": 15128 }, { "epoch": 9.679462571976968, "grad_norm": 0.6149595916368715, "learning_rate": 5.383930175721475e-08, "loss": 0.0104, "step": 15129 }, { "epoch": 9.680102367242483, "grad_norm": 0.5637261026995855, "learning_rate": 5.3624781008614343e-08, "loss": 0.0084, "step": 15130 }, { "epoch": 9.680742162507997, "grad_norm": 0.6102383484369237, "learning_rate": 5.341068733967447e-08, "loss": 0.0093, "step": 15131 }, { "epoch": 9.681381957773512, "grad_norm": 0.5539751615954484, "learning_rate": 5.319702075958444e-08, "loss": 0.0085, "step": 15132 }, { "epoch": 9.682021753039027, "grad_norm": 0.5768655384751403, "learning_rate": 5.298378127752135e-08, "loss": 0.0082, "step": 15133 }, { "epoch": 9.682661548304543, "grad_norm": 0.5213438247701211, "learning_rate": 5.277096890264011e-08, "loss": 0.0078, "step": 15134 }, { "epoch": 9.683301343570058, "grad_norm": 0.599591693526526, "learning_rate": 5.255858364407784e-08, "loss": 0.0091, "step": 15135 }, { "epoch": 9.683941138835573, "grad_norm": 0.543457921423943, "learning_rate": 5.234662551095615e-08, "loss": 0.0073, "step": 15136 }, { "epoch": 9.684580934101088, "grad_norm": 0.5616493358663743, "learning_rate": 5.213509451237331e-08, "loss": 0.0081, "step": 15137 }, { "epoch": 9.685220729366602, "grad_norm": 0.5052183679502498, "learning_rate": 5.1923990657415377e-08, "loss": 0.0084, "step": 15138 }, { "epoch": 9.685860524632117, "grad_norm": 0.5865349526279567, "learning_rate": 5.171331395514401e-08, "loss": 0.0093, "step": 15139 }, { "epoch": 9.686500319897632, "grad_norm": 0.5404794943960985, "learning_rate": 5.15030644146064e-08, "loss": 0.0084, "step": 15140 }, { "epoch": 9.687140115163148, "grad_norm": 0.5927521778959034, "learning_rate": 5.129324204483088e-08, "loss": 0.0081, "step": 15141 }, { "epoch": 9.687779910428663, "grad_norm": 0.4844034479086296, "learning_rate": 5.1083846854825816e-08, "loss": 0.0068, "step": 15142 }, { "epoch": 9.688419705694178, "grad_norm": 0.584940028555559, "learning_rate": 5.0874878853581774e-08, "loss": 0.0095, "step": 15143 }, { "epoch": 9.689059500959694, "grad_norm": 0.5261590893024101, "learning_rate": 5.0666338050073816e-08, "loss": 0.0077, "step": 15144 }, { "epoch": 9.689699296225207, "grad_norm": 0.6579104239808158, "learning_rate": 5.0458224453253656e-08, "loss": 0.0096, "step": 15145 }, { "epoch": 9.690339091490722, "grad_norm": 0.6314259476207016, "learning_rate": 5.02505380720586e-08, "loss": 0.0097, "step": 15146 }, { "epoch": 9.690978886756238, "grad_norm": 0.6197556993622435, "learning_rate": 5.004327891540706e-08, "loss": 0.0092, "step": 15147 }, { "epoch": 9.691618682021753, "grad_norm": 0.663795179499559, "learning_rate": 4.983644699219636e-08, "loss": 0.0104, "step": 15148 }, { "epoch": 9.692258477287268, "grad_norm": 0.5004707354617687, "learning_rate": 4.963004231130941e-08, "loss": 0.0071, "step": 15149 }, { "epoch": 9.692898272552783, "grad_norm": 0.5693922185023946, "learning_rate": 4.9424064881608005e-08, "loss": 0.0086, "step": 15150 }, { "epoch": 9.693538067818299, "grad_norm": 0.507032824675886, "learning_rate": 4.921851471193617e-08, "loss": 0.0072, "step": 15151 }, { "epoch": 9.694177863083812, "grad_norm": 0.5494653194285559, "learning_rate": 4.901339181112019e-08, "loss": 0.0087, "step": 15152 }, { "epoch": 9.694817658349328, "grad_norm": 0.6463146737598475, "learning_rate": 4.880869618796746e-08, "loss": 0.0096, "step": 15153 }, { "epoch": 9.695457453614843, "grad_norm": 0.61979856378002, "learning_rate": 4.860442785126873e-08, "loss": 0.0098, "step": 15154 }, { "epoch": 9.696097248880358, "grad_norm": 0.6749471616442746, "learning_rate": 4.840058680979365e-08, "loss": 0.0111, "step": 15155 }, { "epoch": 9.696737044145873, "grad_norm": 0.5631349488799386, "learning_rate": 4.81971730722941e-08, "loss": 0.0092, "step": 15156 }, { "epoch": 9.697376839411389, "grad_norm": 0.5149258528945951, "learning_rate": 4.799418664750532e-08, "loss": 0.0074, "step": 15157 }, { "epoch": 9.698016634676904, "grad_norm": 0.5914467509627177, "learning_rate": 4.7791627544142574e-08, "loss": 0.0091, "step": 15158 }, { "epoch": 9.69865642994242, "grad_norm": 0.6782126363307918, "learning_rate": 4.758949577090444e-08, "loss": 0.01, "step": 15159 }, { "epoch": 9.699296225207933, "grad_norm": 0.5230676751840376, "learning_rate": 4.738779133646842e-08, "loss": 0.0076, "step": 15160 }, { "epoch": 9.699936020473448, "grad_norm": 0.5506726972744106, "learning_rate": 4.7186514249496494e-08, "loss": 0.0089, "step": 15161 }, { "epoch": 9.700575815738963, "grad_norm": 0.6591735473324714, "learning_rate": 4.6985664518631736e-08, "loss": 0.0086, "step": 15162 }, { "epoch": 9.701215611004478, "grad_norm": 0.5877947945269846, "learning_rate": 4.678524215249725e-08, "loss": 0.0105, "step": 15163 }, { "epoch": 9.701855406269994, "grad_norm": 0.5665283734701991, "learning_rate": 4.658524715969837e-08, "loss": 0.0089, "step": 15164 }, { "epoch": 9.702495201535509, "grad_norm": 0.5714540275490125, "learning_rate": 4.638567954882378e-08, "loss": 0.0099, "step": 15165 }, { "epoch": 9.703134996801024, "grad_norm": 0.5972945009960277, "learning_rate": 4.618653932844219e-08, "loss": 0.0073, "step": 15166 }, { "epoch": 9.70377479206654, "grad_norm": 0.5041928967750097, "learning_rate": 4.598782650710343e-08, "loss": 0.0079, "step": 15167 }, { "epoch": 9.704414587332053, "grad_norm": 0.5484938521874668, "learning_rate": 4.578954109334177e-08, "loss": 0.0084, "step": 15168 }, { "epoch": 9.705054382597568, "grad_norm": 0.5441595381877027, "learning_rate": 4.5591683095669305e-08, "loss": 0.0084, "step": 15169 }, { "epoch": 9.705694177863084, "grad_norm": 0.5497084335121708, "learning_rate": 4.539425252258256e-08, "loss": 0.0081, "step": 15170 }, { "epoch": 9.706333973128599, "grad_norm": 0.592800395127063, "learning_rate": 4.51972493825592e-08, "loss": 0.0094, "step": 15171 }, { "epoch": 9.706973768394114, "grad_norm": 0.4846408163945342, "learning_rate": 4.500067368405692e-08, "loss": 0.007, "step": 15172 }, { "epoch": 9.70761356365963, "grad_norm": 0.624769946011516, "learning_rate": 4.480452543551783e-08, "loss": 0.0093, "step": 15173 }, { "epoch": 9.708253358925145, "grad_norm": 0.5391499012936155, "learning_rate": 4.460880464536299e-08, "loss": 0.0104, "step": 15174 }, { "epoch": 9.708893154190658, "grad_norm": 0.4767706861748245, "learning_rate": 4.4413511321996785e-08, "loss": 0.0068, "step": 15175 }, { "epoch": 9.709532949456174, "grad_norm": 0.554406772489375, "learning_rate": 4.4218645473804724e-08, "loss": 0.0088, "step": 15176 }, { "epoch": 9.710172744721689, "grad_norm": 0.5993899689307267, "learning_rate": 4.402420710915456e-08, "loss": 0.0085, "step": 15177 }, { "epoch": 9.710812539987204, "grad_norm": 0.45704621109003596, "learning_rate": 4.3830196236394063e-08, "loss": 0.0071, "step": 15178 }, { "epoch": 9.71145233525272, "grad_norm": 0.9159193875458793, "learning_rate": 4.3636612863854344e-08, "loss": 0.0076, "step": 15179 }, { "epoch": 9.712092130518235, "grad_norm": 0.5254696302117287, "learning_rate": 4.3443456999847644e-08, "loss": 0.0077, "step": 15180 }, { "epoch": 9.71273192578375, "grad_norm": 0.5280842126232965, "learning_rate": 4.325072865266844e-08, "loss": 0.0074, "step": 15181 }, { "epoch": 9.713371721049263, "grad_norm": 0.523528805393712, "learning_rate": 4.3058427830590113e-08, "loss": 0.0074, "step": 15182 }, { "epoch": 9.714011516314779, "grad_norm": 0.6972366945488963, "learning_rate": 4.2866554541871606e-08, "loss": 0.0097, "step": 15183 }, { "epoch": 9.714651311580294, "grad_norm": 0.5988045240422818, "learning_rate": 4.267510879474968e-08, "loss": 0.0082, "step": 15184 }, { "epoch": 9.71529110684581, "grad_norm": 0.6043066107604173, "learning_rate": 4.2484090597446623e-08, "loss": 0.0098, "step": 15185 }, { "epoch": 9.715930902111324, "grad_norm": 0.5859941415471555, "learning_rate": 4.2293499958163674e-08, "loss": 0.009, "step": 15186 }, { "epoch": 9.71657069737684, "grad_norm": 0.5827854248799738, "learning_rate": 4.210333688508539e-08, "loss": 0.0095, "step": 15187 }, { "epoch": 9.717210492642355, "grad_norm": 0.621521010571764, "learning_rate": 4.191360138637523e-08, "loss": 0.0086, "step": 15188 }, { "epoch": 9.71785028790787, "grad_norm": 0.6682417762555403, "learning_rate": 4.172429347018114e-08, "loss": 0.0093, "step": 15189 }, { "epoch": 9.718490083173384, "grad_norm": 0.5629964854766342, "learning_rate": 4.153541314463105e-08, "loss": 0.007, "step": 15190 }, { "epoch": 9.719129878438899, "grad_norm": 0.5306751148961643, "learning_rate": 4.1346960417835146e-08, "loss": 0.0084, "step": 15191 }, { "epoch": 9.719769673704414, "grad_norm": 0.5353604164773971, "learning_rate": 4.115893529788695e-08, "loss": 0.0068, "step": 15192 }, { "epoch": 9.72040946896993, "grad_norm": 0.636730619200986, "learning_rate": 4.097133779285778e-08, "loss": 0.0101, "step": 15193 }, { "epoch": 9.721049264235445, "grad_norm": 0.5492675487248556, "learning_rate": 4.078416791080342e-08, "loss": 0.0088, "step": 15194 }, { "epoch": 9.72168905950096, "grad_norm": 0.6076757072943562, "learning_rate": 4.059742565976188e-08, "loss": 0.0098, "step": 15195 }, { "epoch": 9.722328854766475, "grad_norm": 0.4681739039231141, "learning_rate": 4.041111104775008e-08, "loss": 0.006, "step": 15196 }, { "epoch": 9.722968650031989, "grad_norm": 0.7301805654343223, "learning_rate": 4.02252240827683e-08, "loss": 0.01, "step": 15197 }, { "epoch": 9.723608445297504, "grad_norm": 0.6000019673166589, "learning_rate": 4.003976477279792e-08, "loss": 0.0078, "step": 15198 }, { "epoch": 9.72424824056302, "grad_norm": 0.510848347475871, "learning_rate": 3.985473312580368e-08, "loss": 0.007, "step": 15199 }, { "epoch": 9.724888035828535, "grad_norm": 0.5790187381075333, "learning_rate": 3.9670129149728123e-08, "loss": 0.0089, "step": 15200 }, { "epoch": 9.72552783109405, "grad_norm": 0.5908908881726506, "learning_rate": 3.948595285250045e-08, "loss": 0.0072, "step": 15201 }, { "epoch": 9.726167626359565, "grad_norm": 0.5194140816571375, "learning_rate": 3.930220424202658e-08, "loss": 0.0074, "step": 15202 }, { "epoch": 9.72680742162508, "grad_norm": 0.5714796368117282, "learning_rate": 3.9118883326197953e-08, "loss": 0.0087, "step": 15203 }, { "epoch": 9.727447216890596, "grad_norm": 0.5445446841640119, "learning_rate": 3.8935990112884957e-08, "loss": 0.0088, "step": 15204 }, { "epoch": 9.72808701215611, "grad_norm": 0.615622122029694, "learning_rate": 3.8753524609942415e-08, "loss": 0.0086, "step": 15205 }, { "epoch": 9.728726807421625, "grad_norm": 0.56556345822353, "learning_rate": 3.8571486825202954e-08, "loss": 0.0095, "step": 15206 }, { "epoch": 9.72936660268714, "grad_norm": 0.5892292150260413, "learning_rate": 3.838987676648365e-08, "loss": 0.0085, "step": 15207 }, { "epoch": 9.730006397952655, "grad_norm": 0.5707050206325958, "learning_rate": 3.8208694441582704e-08, "loss": 0.0086, "step": 15208 }, { "epoch": 9.73064619321817, "grad_norm": 0.672543949892058, "learning_rate": 3.802793985828057e-08, "loss": 0.0093, "step": 15209 }, { "epoch": 9.731285988483686, "grad_norm": 0.702594510449332, "learning_rate": 3.7847613024336595e-08, "loss": 0.0083, "step": 15210 }, { "epoch": 9.731925783749201, "grad_norm": 0.5912241427991515, "learning_rate": 3.7667713947494577e-08, "loss": 0.0099, "step": 15211 }, { "epoch": 9.732565579014715, "grad_norm": 0.7038397317062838, "learning_rate": 3.748824263548057e-08, "loss": 0.0096, "step": 15212 }, { "epoch": 9.73320537428023, "grad_norm": 0.5894009744805101, "learning_rate": 3.7309199095999505e-08, "loss": 0.0075, "step": 15213 }, { "epoch": 9.733845169545745, "grad_norm": 0.6972422202165267, "learning_rate": 3.7130583336738576e-08, "loss": 0.0101, "step": 15214 }, { "epoch": 9.73448496481126, "grad_norm": 0.5776491439421316, "learning_rate": 3.695239536536721e-08, "loss": 0.0081, "step": 15215 }, { "epoch": 9.735124760076776, "grad_norm": 0.5931049603575048, "learning_rate": 3.6774635189538163e-08, "loss": 0.0091, "step": 15216 }, { "epoch": 9.735764555342291, "grad_norm": 0.6243421575036439, "learning_rate": 3.659730281688312e-08, "loss": 0.0099, "step": 15217 }, { "epoch": 9.736404350607806, "grad_norm": 0.6168608216178926, "learning_rate": 3.642039825501598e-08, "loss": 0.0092, "step": 15218 }, { "epoch": 9.73704414587332, "grad_norm": 0.5353286298700012, "learning_rate": 3.6243921511532886e-08, "loss": 0.0074, "step": 15219 }, { "epoch": 9.737683941138835, "grad_norm": 0.6270128515845761, "learning_rate": 3.606787259401223e-08, "loss": 0.0095, "step": 15220 }, { "epoch": 9.73832373640435, "grad_norm": 0.5455816190423194, "learning_rate": 3.589225151001241e-08, "loss": 0.0082, "step": 15221 }, { "epoch": 9.738963531669866, "grad_norm": 0.5127496268751803, "learning_rate": 3.5717058267074054e-08, "loss": 0.0074, "step": 15222 }, { "epoch": 9.73960332693538, "grad_norm": 0.5475008023163345, "learning_rate": 3.5542292872720044e-08, "loss": 0.0077, "step": 15223 }, { "epoch": 9.740243122200896, "grad_norm": 0.6675562768017796, "learning_rate": 3.536795533445547e-08, "loss": 0.0086, "step": 15224 }, { "epoch": 9.740882917466411, "grad_norm": 0.6562448167048145, "learning_rate": 3.519404565976437e-08, "loss": 0.01, "step": 15225 }, { "epoch": 9.741522712731927, "grad_norm": 0.6480052254628658, "learning_rate": 3.502056385611519e-08, "loss": 0.0098, "step": 15226 }, { "epoch": 9.74216250799744, "grad_norm": 0.5140545948543507, "learning_rate": 3.4847509930957535e-08, "loss": 0.0075, "step": 15227 }, { "epoch": 9.742802303262955, "grad_norm": 0.5795292304673102, "learning_rate": 3.4674883891719914e-08, "loss": 0.0071, "step": 15228 }, { "epoch": 9.74344209852847, "grad_norm": 0.5738373668545581, "learning_rate": 3.450268574581639e-08, "loss": 0.0078, "step": 15229 }, { "epoch": 9.744081893793986, "grad_norm": 0.5493637132047671, "learning_rate": 3.433091550064105e-08, "loss": 0.008, "step": 15230 }, { "epoch": 9.744721689059501, "grad_norm": 0.6067029981925224, "learning_rate": 3.4159573163567994e-08, "loss": 0.0102, "step": 15231 }, { "epoch": 9.745361484325016, "grad_norm": 1.0241149550658157, "learning_rate": 3.398865874195467e-08, "loss": 0.0095, "step": 15232 }, { "epoch": 9.746001279590532, "grad_norm": 0.6418865112513944, "learning_rate": 3.381817224314077e-08, "loss": 0.01, "step": 15233 }, { "epoch": 9.746641074856045, "grad_norm": 0.7429399279001565, "learning_rate": 3.364811367444487e-08, "loss": 0.0103, "step": 15234 }, { "epoch": 9.74728087012156, "grad_norm": 0.5024841841880995, "learning_rate": 3.347848304317114e-08, "loss": 0.0083, "step": 15235 }, { "epoch": 9.747920665387076, "grad_norm": 0.5681618397523823, "learning_rate": 3.330928035660264e-08, "loss": 0.0082, "step": 15236 }, { "epoch": 9.748560460652591, "grad_norm": 0.6214279092057182, "learning_rate": 3.314050562200355e-08, "loss": 0.0086, "step": 15237 }, { "epoch": 9.749200255918106, "grad_norm": 0.5712124023868973, "learning_rate": 3.297215884662142e-08, "loss": 0.0079, "step": 15238 }, { "epoch": 9.749840051183622, "grad_norm": 0.5888811796332267, "learning_rate": 3.280424003768601e-08, "loss": 0.0094, "step": 15239 }, { "epoch": 9.750479846449137, "grad_norm": 0.5279334710870363, "learning_rate": 3.2636749202404894e-08, "loss": 0.0077, "step": 15240 }, { "epoch": 9.751119641714652, "grad_norm": 0.5780255214889902, "learning_rate": 3.2469686347971205e-08, "loss": 0.008, "step": 15241 }, { "epoch": 9.751759436980166, "grad_norm": 0.5826244036934553, "learning_rate": 3.230305148155921e-08, "loss": 0.0087, "step": 15242 }, { "epoch": 9.752399232245681, "grad_norm": 0.5888254763756516, "learning_rate": 3.213684461032318e-08, "loss": 0.0082, "step": 15243 }, { "epoch": 9.753039027511196, "grad_norm": 0.5806061122339985, "learning_rate": 3.197106574139852e-08, "loss": 0.009, "step": 15244 }, { "epoch": 9.753678822776711, "grad_norm": 0.5426646929579277, "learning_rate": 3.180571488190509e-08, "loss": 0.008, "step": 15245 }, { "epoch": 9.754318618042227, "grad_norm": 0.5311296335902034, "learning_rate": 3.164079203894277e-08, "loss": 0.0086, "step": 15246 }, { "epoch": 9.754958413307742, "grad_norm": 0.4953314055020481, "learning_rate": 3.147629721959145e-08, "loss": 0.0066, "step": 15247 }, { "epoch": 9.755598208573257, "grad_norm": 0.6621541541033443, "learning_rate": 3.131223043091658e-08, "loss": 0.01, "step": 15248 }, { "epoch": 9.75623800383877, "grad_norm": 0.6362586687254683, "learning_rate": 3.114859167996032e-08, "loss": 0.0098, "step": 15249 }, { "epoch": 9.756877799104286, "grad_norm": 0.7241615728603095, "learning_rate": 3.0985380973751476e-08, "loss": 0.0114, "step": 15250 }, { "epoch": 9.757517594369801, "grad_norm": 0.635478645576443, "learning_rate": 3.082259831929557e-08, "loss": 0.0074, "step": 15251 }, { "epoch": 9.758157389635317, "grad_norm": 0.5656666730666419, "learning_rate": 3.066024372358478e-08, "loss": 0.0091, "step": 15252 }, { "epoch": 9.758797184900832, "grad_norm": 0.6063135027560749, "learning_rate": 3.0498317193587975e-08, "loss": 0.009, "step": 15253 }, { "epoch": 9.759436980166347, "grad_norm": 0.5718761557808976, "learning_rate": 3.03368187362596e-08, "loss": 0.0092, "step": 15254 }, { "epoch": 9.760076775431862, "grad_norm": 0.5571459014292613, "learning_rate": 3.0175748358534095e-08, "loss": 0.0082, "step": 15255 }, { "epoch": 9.760716570697376, "grad_norm": 0.6169963157075018, "learning_rate": 3.001510606732594e-08, "loss": 0.0095, "step": 15256 }, { "epoch": 9.761356365962891, "grad_norm": 0.43393327560589673, "learning_rate": 2.985489186953405e-08, "loss": 0.0075, "step": 15257 }, { "epoch": 9.761996161228407, "grad_norm": 0.5263377551029524, "learning_rate": 2.969510577203849e-08, "loss": 0.0072, "step": 15258 }, { "epoch": 9.762635956493922, "grad_norm": 0.707588765129747, "learning_rate": 2.9535747781698208e-08, "loss": 0.0087, "step": 15259 }, { "epoch": 9.763275751759437, "grad_norm": 0.5654289365893124, "learning_rate": 2.9376817905357734e-08, "loss": 0.0081, "step": 15260 }, { "epoch": 9.763915547024952, "grad_norm": 0.5344776827904651, "learning_rate": 2.9218316149840498e-08, "loss": 0.0065, "step": 15261 }, { "epoch": 9.764555342290468, "grad_norm": 0.5581054380296993, "learning_rate": 2.9060242521951054e-08, "loss": 0.0088, "step": 15262 }, { "epoch": 9.765195137555983, "grad_norm": 0.6544408736266245, "learning_rate": 2.8902597028478417e-08, "loss": 0.0116, "step": 15263 }, { "epoch": 9.765834932821496, "grad_norm": 0.73396927811305, "learning_rate": 2.8745379676191622e-08, "loss": 0.0098, "step": 15264 }, { "epoch": 9.766474728087012, "grad_norm": 0.5942015891296155, "learning_rate": 2.8588590471839707e-08, "loss": 0.009, "step": 15265 }, { "epoch": 9.767114523352527, "grad_norm": 0.5394737352440592, "learning_rate": 2.8432229422156178e-08, "loss": 0.0083, "step": 15266 }, { "epoch": 9.767754318618042, "grad_norm": 0.549955387417845, "learning_rate": 2.8276296533854552e-08, "loss": 0.0069, "step": 15267 }, { "epoch": 9.768394113883557, "grad_norm": 0.5518404369547754, "learning_rate": 2.81207918136317e-08, "loss": 0.0081, "step": 15268 }, { "epoch": 9.769033909149073, "grad_norm": 0.5798465876558797, "learning_rate": 2.7965715268162275e-08, "loss": 0.0095, "step": 15269 }, { "epoch": 9.769673704414588, "grad_norm": 0.573753517156998, "learning_rate": 2.7811066904106508e-08, "loss": 0.0076, "step": 15270 }, { "epoch": 9.770313499680102, "grad_norm": 0.6441041890940681, "learning_rate": 2.7656846728103537e-08, "loss": 0.0091, "step": 15271 }, { "epoch": 9.770953294945617, "grad_norm": 0.628238140575647, "learning_rate": 2.7503054746778057e-08, "loss": 0.0109, "step": 15272 }, { "epoch": 9.771593090211132, "grad_norm": 0.5520357937304957, "learning_rate": 2.7349690966730347e-08, "loss": 0.0084, "step": 15273 }, { "epoch": 9.772232885476647, "grad_norm": 0.6131603988044219, "learning_rate": 2.719675539454625e-08, "loss": 0.0085, "step": 15274 }, { "epoch": 9.772872680742163, "grad_norm": 0.6160522096201808, "learning_rate": 2.7044248036794953e-08, "loss": 0.008, "step": 15275 }, { "epoch": 9.773512476007678, "grad_norm": 0.5590848920830521, "learning_rate": 2.689216890002233e-08, "loss": 0.0083, "step": 15276 }, { "epoch": 9.774152271273193, "grad_norm": 0.6261764146481855, "learning_rate": 2.6740517990758718e-08, "loss": 0.0089, "step": 15277 }, { "epoch": 9.774792066538708, "grad_norm": 0.5917559593620272, "learning_rate": 2.6589295315516685e-08, "loss": 0.0096, "step": 15278 }, { "epoch": 9.775431861804222, "grad_norm": 0.49601409214676273, "learning_rate": 2.643850088078992e-08, "loss": 0.0064, "step": 15279 }, { "epoch": 9.776071657069737, "grad_norm": 0.5543913229517553, "learning_rate": 2.6288134693051026e-08, "loss": 0.0084, "step": 15280 }, { "epoch": 9.776711452335253, "grad_norm": 0.5140199151950268, "learning_rate": 2.6138196758758173e-08, "loss": 0.0078, "step": 15281 }, { "epoch": 9.777351247600768, "grad_norm": 0.5934571287080125, "learning_rate": 2.598868708434843e-08, "loss": 0.008, "step": 15282 }, { "epoch": 9.777991042866283, "grad_norm": 0.5680154156552664, "learning_rate": 2.583960567624333e-08, "loss": 0.0089, "step": 15283 }, { "epoch": 9.778630838131798, "grad_norm": 0.5470864831560305, "learning_rate": 2.5690952540842196e-08, "loss": 0.009, "step": 15284 }, { "epoch": 9.779270633397314, "grad_norm": 0.622535643914173, "learning_rate": 2.554272768452881e-08, "loss": 0.0095, "step": 15285 }, { "epoch": 9.779910428662827, "grad_norm": 0.6033234451197428, "learning_rate": 2.5394931113666975e-08, "loss": 0.0082, "step": 15286 }, { "epoch": 9.780550223928342, "grad_norm": 0.6136638957959305, "learning_rate": 2.5247562834603832e-08, "loss": 0.0098, "step": 15287 }, { "epoch": 9.781190019193858, "grad_norm": 0.6267770932699767, "learning_rate": 2.5100622853666545e-08, "loss": 0.0114, "step": 15288 }, { "epoch": 9.781829814459373, "grad_norm": 0.6077050040533579, "learning_rate": 2.495411117716451e-08, "loss": 0.0079, "step": 15289 }, { "epoch": 9.782469609724888, "grad_norm": 0.5688882347729276, "learning_rate": 2.4808027811388247e-08, "loss": 0.0084, "step": 15290 }, { "epoch": 9.783109404990403, "grad_norm": 0.42539852732470074, "learning_rate": 2.466237276261163e-08, "loss": 0.0054, "step": 15291 }, { "epoch": 9.783749200255919, "grad_norm": 0.586325997276871, "learning_rate": 2.4517146037087438e-08, "loss": 0.0086, "step": 15292 }, { "epoch": 9.784388995521432, "grad_norm": 0.6512115885774796, "learning_rate": 2.4372347641052897e-08, "loss": 0.0075, "step": 15293 }, { "epoch": 9.785028790786948, "grad_norm": 0.5740747207544455, "learning_rate": 2.422797758072304e-08, "loss": 0.0086, "step": 15294 }, { "epoch": 9.785668586052463, "grad_norm": 0.568756108174232, "learning_rate": 2.4084035862299572e-08, "loss": 0.01, "step": 15295 }, { "epoch": 9.786308381317978, "grad_norm": 0.5963771021058838, "learning_rate": 2.3940522491960884e-08, "loss": 0.0091, "step": 15296 }, { "epoch": 9.786948176583493, "grad_norm": 0.6174314192116689, "learning_rate": 2.3797437475870933e-08, "loss": 0.0094, "step": 15297 }, { "epoch": 9.787587971849009, "grad_norm": 0.4968618158711749, "learning_rate": 2.3654780820171476e-08, "loss": 0.0071, "step": 15298 }, { "epoch": 9.788227767114524, "grad_norm": 0.680415689860044, "learning_rate": 2.351255253099094e-08, "loss": 0.0123, "step": 15299 }, { "epoch": 9.78886756238004, "grad_norm": 0.6377001464775314, "learning_rate": 2.3370752614433334e-08, "loss": 0.0076, "step": 15300 }, { "epoch": 9.789507357645553, "grad_norm": 0.6216201406868749, "learning_rate": 2.322938107658823e-08, "loss": 0.0092, "step": 15301 }, { "epoch": 9.790147152911068, "grad_norm": 0.5347929384880876, "learning_rate": 2.308843792352744e-08, "loss": 0.0074, "step": 15302 }, { "epoch": 9.790786948176583, "grad_norm": 0.5421014442588089, "learning_rate": 2.294792316130168e-08, "loss": 0.0082, "step": 15303 }, { "epoch": 9.791426743442099, "grad_norm": 0.5284545114426678, "learning_rate": 2.2807836795942784e-08, "loss": 0.0092, "step": 15304 }, { "epoch": 9.792066538707614, "grad_norm": 0.5703612280218631, "learning_rate": 2.266817883346817e-08, "loss": 0.0081, "step": 15305 }, { "epoch": 9.792706333973129, "grad_norm": 0.5878882923784489, "learning_rate": 2.2528949279874145e-08, "loss": 0.0085, "step": 15306 }, { "epoch": 9.793346129238644, "grad_norm": 0.676400737111999, "learning_rate": 2.2390148141138157e-08, "loss": 0.0093, "step": 15307 }, { "epoch": 9.79398592450416, "grad_norm": 0.5561034446831816, "learning_rate": 2.225177542321988e-08, "loss": 0.0076, "step": 15308 }, { "epoch": 9.794625719769673, "grad_norm": 0.519158472596525, "learning_rate": 2.2113831132062334e-08, "loss": 0.0082, "step": 15309 }, { "epoch": 9.795265515035188, "grad_norm": 0.5457557455969005, "learning_rate": 2.1976315273586347e-08, "loss": 0.0065, "step": 15310 }, { "epoch": 9.795905310300704, "grad_norm": 0.5241274635937817, "learning_rate": 2.1839227853698297e-08, "loss": 0.0087, "step": 15311 }, { "epoch": 9.796545105566219, "grad_norm": 0.6533773214797455, "learning_rate": 2.1702568878284592e-08, "loss": 0.0096, "step": 15312 }, { "epoch": 9.797184900831734, "grad_norm": 0.505292059338366, "learning_rate": 2.1566338353212757e-08, "loss": 0.0082, "step": 15313 }, { "epoch": 9.79782469609725, "grad_norm": 0.602424077041206, "learning_rate": 2.143053628433145e-08, "loss": 0.009, "step": 15314 }, { "epoch": 9.798464491362765, "grad_norm": 0.6410981803462494, "learning_rate": 2.129516267747267e-08, "loss": 0.0093, "step": 15315 }, { "epoch": 9.799104286628278, "grad_norm": 0.5596951548807682, "learning_rate": 2.1160217538448434e-08, "loss": 0.0085, "step": 15316 }, { "epoch": 9.799744081893794, "grad_norm": 0.5512865903134917, "learning_rate": 2.102570087305411e-08, "loss": 0.0074, "step": 15317 }, { "epoch": 9.800383877159309, "grad_norm": 0.5745166279197023, "learning_rate": 2.0891612687063967e-08, "loss": 0.0076, "step": 15318 }, { "epoch": 9.801023672424824, "grad_norm": 0.585214689127226, "learning_rate": 2.075795298623784e-08, "loss": 0.009, "step": 15319 }, { "epoch": 9.80166346769034, "grad_norm": 0.6302152727226271, "learning_rate": 2.0624721776312252e-08, "loss": 0.0082, "step": 15320 }, { "epoch": 9.802303262955855, "grad_norm": 0.6716001366933733, "learning_rate": 2.04919190630104e-08, "loss": 0.0097, "step": 15321 }, { "epoch": 9.80294305822137, "grad_norm": 0.6103953667423954, "learning_rate": 2.0359544852032177e-08, "loss": 0.0098, "step": 15322 }, { "epoch": 9.803582853486883, "grad_norm": 0.5673008902451689, "learning_rate": 2.0227599149064136e-08, "loss": 0.0087, "step": 15323 }, { "epoch": 9.804222648752399, "grad_norm": 0.5932881528234616, "learning_rate": 2.009608195976842e-08, "loss": 0.0096, "step": 15324 }, { "epoch": 9.804862444017914, "grad_norm": 0.5838643673139969, "learning_rate": 1.9964993289796065e-08, "loss": 0.0085, "step": 15325 }, { "epoch": 9.80550223928343, "grad_norm": 0.4471243860642375, "learning_rate": 1.9834333144772566e-08, "loss": 0.0059, "step": 15326 }, { "epoch": 9.806142034548945, "grad_norm": 0.5369921220181669, "learning_rate": 1.9704101530310104e-08, "loss": 0.009, "step": 15327 }, { "epoch": 9.80678182981446, "grad_norm": 0.5798614919090477, "learning_rate": 1.957429845199865e-08, "loss": 0.0083, "step": 15328 }, { "epoch": 9.807421625079975, "grad_norm": 0.6153126231626287, "learning_rate": 1.944492391541486e-08, "loss": 0.0102, "step": 15329 }, { "epoch": 9.808061420345489, "grad_norm": 0.5591784911048778, "learning_rate": 1.9315977926110952e-08, "loss": 0.0075, "step": 15330 }, { "epoch": 9.808701215611004, "grad_norm": 0.6354980285004962, "learning_rate": 1.918746048962472e-08, "loss": 0.01, "step": 15331 }, { "epoch": 9.809341010876519, "grad_norm": 0.443713514650932, "learning_rate": 1.9059371611475084e-08, "loss": 0.0066, "step": 15332 }, { "epoch": 9.809980806142034, "grad_norm": 0.6144924895147357, "learning_rate": 1.8931711297160983e-08, "loss": 0.0098, "step": 15333 }, { "epoch": 9.81062060140755, "grad_norm": 0.5618542322618839, "learning_rate": 1.8804479552163578e-08, "loss": 0.0092, "step": 15334 }, { "epoch": 9.811260396673065, "grad_norm": 0.6075073904981819, "learning_rate": 1.8677676381947395e-08, "loss": 0.0101, "step": 15335 }, { "epoch": 9.81190019193858, "grad_norm": 0.625140325796653, "learning_rate": 1.8551301791955856e-08, "loss": 0.0107, "step": 15336 }, { "epoch": 9.812539987204095, "grad_norm": 0.6267350415439534, "learning_rate": 1.842535578761573e-08, "loss": 0.0105, "step": 15337 }, { "epoch": 9.813179782469609, "grad_norm": 0.5564726407490191, "learning_rate": 1.829983837433602e-08, "loss": 0.0101, "step": 15338 }, { "epoch": 9.813819577735124, "grad_norm": 0.5901048767812074, "learning_rate": 1.8174749557504643e-08, "loss": 0.008, "step": 15339 }, { "epoch": 9.81445937300064, "grad_norm": 0.6035818868743866, "learning_rate": 1.805008934249286e-08, "loss": 0.009, "step": 15340 }, { "epoch": 9.815099168266155, "grad_norm": 0.5856686184789703, "learning_rate": 1.7925857734654163e-08, "loss": 0.0083, "step": 15341 }, { "epoch": 9.81573896353167, "grad_norm": 0.4805812364916293, "learning_rate": 1.780205473932206e-08, "loss": 0.0062, "step": 15342 }, { "epoch": 9.816378758797185, "grad_norm": 0.645074293033214, "learning_rate": 1.7678680361813415e-08, "loss": 0.0103, "step": 15343 }, { "epoch": 9.8170185540627, "grad_norm": 0.6426699235495742, "learning_rate": 1.7555734607423992e-08, "loss": 0.0096, "step": 15344 }, { "epoch": 9.817658349328216, "grad_norm": 0.5592557569410367, "learning_rate": 1.7433217481435115e-08, "loss": 0.0096, "step": 15345 }, { "epoch": 9.81829814459373, "grad_norm": 0.5695323193504791, "learning_rate": 1.731112898910592e-08, "loss": 0.0058, "step": 15346 }, { "epoch": 9.818937939859245, "grad_norm": 0.484356952160262, "learning_rate": 1.718946913567776e-08, "loss": 0.0072, "step": 15347 }, { "epoch": 9.81957773512476, "grad_norm": 0.5120435323583188, "learning_rate": 1.7068237926377575e-08, "loss": 0.0075, "step": 15348 }, { "epoch": 9.820217530390275, "grad_norm": 0.6090406902061339, "learning_rate": 1.694743536640786e-08, "loss": 0.0091, "step": 15349 }, { "epoch": 9.82085732565579, "grad_norm": 0.615427856540154, "learning_rate": 1.6827061460956695e-08, "loss": 0.0102, "step": 15350 }, { "epoch": 9.821497120921306, "grad_norm": 0.5670511412164242, "learning_rate": 1.6707116215194384e-08, "loss": 0.0076, "step": 15351 }, { "epoch": 9.822136916186821, "grad_norm": 0.5598111007228057, "learning_rate": 1.6587599634267926e-08, "loss": 0.0091, "step": 15352 }, { "epoch": 9.822776711452335, "grad_norm": 0.5544341466376271, "learning_rate": 1.6468511723310987e-08, "loss": 0.0068, "step": 15353 }, { "epoch": 9.82341650671785, "grad_norm": 0.6490828052812575, "learning_rate": 1.6349852487437256e-08, "loss": 0.0084, "step": 15354 }, { "epoch": 9.824056301983365, "grad_norm": 0.5560975810706325, "learning_rate": 1.6231621931741547e-08, "loss": 0.0093, "step": 15355 }, { "epoch": 9.82469609724888, "grad_norm": 0.5654444210904791, "learning_rate": 1.61138200612998e-08, "loss": 0.0095, "step": 15356 }, { "epoch": 9.825335892514396, "grad_norm": 0.6020856180159727, "learning_rate": 1.5996446881171303e-08, "loss": 0.0094, "step": 15357 }, { "epoch": 9.825975687779911, "grad_norm": 0.6090311568512415, "learning_rate": 1.587950239639424e-08, "loss": 0.0104, "step": 15358 }, { "epoch": 9.826615483045426, "grad_norm": 0.6216627685661951, "learning_rate": 1.576298661199238e-08, "loss": 0.0095, "step": 15359 }, { "epoch": 9.82725527831094, "grad_norm": 0.5940768547009692, "learning_rate": 1.5646899532966165e-08, "loss": 0.0084, "step": 15360 }, { "epoch": 9.827895073576455, "grad_norm": 0.6171628738719853, "learning_rate": 1.5531241164301604e-08, "loss": 0.0095, "step": 15361 }, { "epoch": 9.82853486884197, "grad_norm": 0.6318800267907632, "learning_rate": 1.5416011510964725e-08, "loss": 0.0096, "step": 15362 }, { "epoch": 9.829174664107486, "grad_norm": 0.5619115986545519, "learning_rate": 1.530121057790268e-08, "loss": 0.0088, "step": 15363 }, { "epoch": 9.829814459373, "grad_norm": 0.5435565367953823, "learning_rate": 1.5186838370044865e-08, "loss": 0.0079, "step": 15364 }, { "epoch": 9.830454254638516, "grad_norm": 0.5906961891430803, "learning_rate": 1.5072894892304014e-08, "loss": 0.0082, "step": 15365 }, { "epoch": 9.831094049904031, "grad_norm": 0.559171114669977, "learning_rate": 1.4959380149569548e-08, "loss": 0.0078, "step": 15366 }, { "epoch": 9.831733845169545, "grad_norm": 0.5820976635800073, "learning_rate": 1.4846294146718677e-08, "loss": 0.0106, "step": 15367 }, { "epoch": 9.83237364043506, "grad_norm": 0.5295086024703396, "learning_rate": 1.4733636888605296e-08, "loss": 0.009, "step": 15368 }, { "epoch": 9.833013435700575, "grad_norm": 0.544726839386562, "learning_rate": 1.4621408380066648e-08, "loss": 0.008, "step": 15369 }, { "epoch": 9.83365323096609, "grad_norm": 0.6205873011994064, "learning_rate": 1.450960862592332e-08, "loss": 0.0093, "step": 15370 }, { "epoch": 9.834293026231606, "grad_norm": 0.5802732068639109, "learning_rate": 1.43982376309737e-08, "loss": 0.0094, "step": 15371 }, { "epoch": 9.834932821497121, "grad_norm": 0.6009607111025822, "learning_rate": 1.4287295400000623e-08, "loss": 0.0097, "step": 15372 }, { "epoch": 9.835572616762636, "grad_norm": 0.5160755744178394, "learning_rate": 1.4176781937768058e-08, "loss": 0.0076, "step": 15373 }, { "epoch": 9.836212412028152, "grad_norm": 0.5337470196787703, "learning_rate": 1.40666972490211e-08, "loss": 0.0078, "step": 15374 }, { "epoch": 9.836852207293665, "grad_norm": 0.6582560356359123, "learning_rate": 1.3957041338487077e-08, "loss": 0.0094, "step": 15375 }, { "epoch": 9.83749200255918, "grad_norm": 0.584380691847012, "learning_rate": 1.3847814210873333e-08, "loss": 0.0086, "step": 15376 }, { "epoch": 9.838131797824696, "grad_norm": 0.5624330186864264, "learning_rate": 1.3739015870870565e-08, "loss": 0.0085, "step": 15377 }, { "epoch": 9.838771593090211, "grad_norm": 0.6064585019120836, "learning_rate": 1.3630646323149476e-08, "loss": 0.0091, "step": 15378 }, { "epoch": 9.839411388355726, "grad_norm": 0.5197407027255369, "learning_rate": 1.3522705572365236e-08, "loss": 0.0078, "step": 15379 }, { "epoch": 9.840051183621242, "grad_norm": 0.5604017174894974, "learning_rate": 1.3415193623150802e-08, "loss": 0.009, "step": 15380 }, { "epoch": 9.840690978886757, "grad_norm": 0.454923396997757, "learning_rate": 1.3308110480123592e-08, "loss": 0.0064, "step": 15381 }, { "epoch": 9.841330774152272, "grad_norm": 0.5522319298359353, "learning_rate": 1.3201456147881042e-08, "loss": 0.0089, "step": 15382 }, { "epoch": 9.841970569417786, "grad_norm": 0.5230648590468051, "learning_rate": 1.309523063100282e-08, "loss": 0.0074, "step": 15383 }, { "epoch": 9.842610364683301, "grad_norm": 0.5842579691488061, "learning_rate": 1.2989433934049723e-08, "loss": 0.0077, "step": 15384 }, { "epoch": 9.843250159948816, "grad_norm": 0.5410513427462089, "learning_rate": 1.2884066061565893e-08, "loss": 0.007, "step": 15385 }, { "epoch": 9.843889955214332, "grad_norm": 0.6542793201464268, "learning_rate": 1.2779127018073267e-08, "loss": 0.01, "step": 15386 }, { "epoch": 9.844529750479847, "grad_norm": 0.5562694203518711, "learning_rate": 1.2674616808080464e-08, "loss": 0.0081, "step": 15387 }, { "epoch": 9.845169545745362, "grad_norm": 0.5801174907371994, "learning_rate": 1.2570535436071674e-08, "loss": 0.0094, "step": 15388 }, { "epoch": 9.845809341010877, "grad_norm": 0.6003102634384263, "learning_rate": 1.2466882906518874e-08, "loss": 0.0087, "step": 15389 }, { "epoch": 9.84644913627639, "grad_norm": 0.5552701687147257, "learning_rate": 1.2363659223871838e-08, "loss": 0.0087, "step": 15390 }, { "epoch": 9.847088931541906, "grad_norm": 0.6412269904335515, "learning_rate": 1.2260864392562577e-08, "loss": 0.0089, "step": 15391 }, { "epoch": 9.847728726807421, "grad_norm": 0.596175256332718, "learning_rate": 1.2158498417005338e-08, "loss": 0.0076, "step": 15392 }, { "epoch": 9.848368522072937, "grad_norm": 0.4785580569784116, "learning_rate": 1.2056561301594383e-08, "loss": 0.0063, "step": 15393 }, { "epoch": 9.849008317338452, "grad_norm": 0.6207756818653848, "learning_rate": 1.1955053050707322e-08, "loss": 0.009, "step": 15394 }, { "epoch": 9.849648112603967, "grad_norm": 0.4958363962713152, "learning_rate": 1.1853973668704e-08, "loss": 0.0068, "step": 15395 }, { "epoch": 9.850287907869482, "grad_norm": 0.5843083008895317, "learning_rate": 1.1753323159923169e-08, "loss": 0.0091, "step": 15396 }, { "epoch": 9.850927703134996, "grad_norm": 0.5974087567500938, "learning_rate": 1.1653101528686928e-08, "loss": 0.0089, "step": 15397 }, { "epoch": 9.851567498400511, "grad_norm": 0.500323005095343, "learning_rate": 1.15533087792985e-08, "loss": 0.0076, "step": 15398 }, { "epoch": 9.852207293666027, "grad_norm": 0.555264122203345, "learning_rate": 1.1453944916042237e-08, "loss": 0.0089, "step": 15399 }, { "epoch": 9.852847088931542, "grad_norm": 0.7322501397530876, "learning_rate": 1.1355009943185835e-08, "loss": 0.0076, "step": 15400 }, { "epoch": 9.853486884197057, "grad_norm": 0.59310344110935, "learning_rate": 1.125650386497701e-08, "loss": 0.0088, "step": 15401 }, { "epoch": 9.854126679462572, "grad_norm": 0.5531205786049821, "learning_rate": 1.1158426685645707e-08, "loss": 0.0076, "step": 15402 }, { "epoch": 9.854766474728088, "grad_norm": 0.6672452796016765, "learning_rate": 1.1060778409401895e-08, "loss": 0.0108, "step": 15403 }, { "epoch": 9.855406269993601, "grad_norm": 0.5652490460601645, "learning_rate": 1.0963559040439996e-08, "loss": 0.0081, "step": 15404 }, { "epoch": 9.856046065259116, "grad_norm": 0.5852448667019742, "learning_rate": 1.086676858293334e-08, "loss": 0.0079, "step": 15405 }, { "epoch": 9.856685860524632, "grad_norm": 0.6523396731875686, "learning_rate": 1.07704070410386e-08, "loss": 0.0088, "step": 15406 }, { "epoch": 9.857325655790147, "grad_norm": 0.6098982738266747, "learning_rate": 1.0674474418893577e-08, "loss": 0.0096, "step": 15407 }, { "epoch": 9.857965451055662, "grad_norm": 0.6402026655180789, "learning_rate": 1.057897072061609e-08, "loss": 0.0099, "step": 15408 }, { "epoch": 9.858605246321178, "grad_norm": 0.5025177434208625, "learning_rate": 1.048389595030841e-08, "loss": 0.007, "step": 15409 }, { "epoch": 9.859245041586693, "grad_norm": 0.6715355900395502, "learning_rate": 1.0389250112051719e-08, "loss": 0.0083, "step": 15410 }, { "epoch": 9.859884836852208, "grad_norm": 0.5884076615288364, "learning_rate": 1.0295033209911653e-08, "loss": 0.0085, "step": 15411 }, { "epoch": 9.860524632117722, "grad_norm": 0.8040695307914865, "learning_rate": 1.0201245247931646e-08, "loss": 0.0096, "step": 15412 }, { "epoch": 9.861164427383237, "grad_norm": 0.5315353815454373, "learning_rate": 1.0107886230139585e-08, "loss": 0.007, "step": 15413 }, { "epoch": 9.861804222648752, "grad_norm": 0.674445451866328, "learning_rate": 1.0014956160544487e-08, "loss": 0.0114, "step": 15414 }, { "epoch": 9.862444017914267, "grad_norm": 0.6756981383357837, "learning_rate": 9.922455043136492e-09, "loss": 0.0094, "step": 15415 }, { "epoch": 9.863083813179783, "grad_norm": 0.6304329884692603, "learning_rate": 9.830382881886868e-09, "loss": 0.0099, "step": 15416 }, { "epoch": 9.863723608445298, "grad_norm": 0.5322428341386493, "learning_rate": 9.738739680750232e-09, "loss": 0.0078, "step": 15417 }, { "epoch": 9.864363403710813, "grad_norm": 0.600228450142362, "learning_rate": 9.647525443660099e-09, "loss": 0.0083, "step": 15418 }, { "epoch": 9.865003198976328, "grad_norm": 0.5624968339640865, "learning_rate": 9.556740174533341e-09, "loss": 0.0096, "step": 15419 }, { "epoch": 9.865642994241842, "grad_norm": 0.555665331060938, "learning_rate": 9.466383877270169e-09, "loss": 0.0087, "step": 15420 }, { "epoch": 9.866282789507357, "grad_norm": 0.6214636097794659, "learning_rate": 9.376456555747482e-09, "loss": 0.0077, "step": 15421 }, { "epoch": 9.866922584772873, "grad_norm": 0.6573480872234978, "learning_rate": 9.286958213827746e-09, "loss": 0.0089, "step": 15422 }, { "epoch": 9.867562380038388, "grad_norm": 0.6334715842730468, "learning_rate": 9.197888855354553e-09, "loss": 0.01, "step": 15423 }, { "epoch": 9.868202175303903, "grad_norm": 0.6551421579328016, "learning_rate": 9.109248484151511e-09, "loss": 0.0097, "step": 15424 }, { "epoch": 9.868841970569418, "grad_norm": 0.5498385796865956, "learning_rate": 9.021037104024466e-09, "loss": 0.0093, "step": 15425 }, { "epoch": 9.869481765834934, "grad_norm": 0.5715178897445192, "learning_rate": 8.933254718761497e-09, "loss": 0.0075, "step": 15426 }, { "epoch": 9.870121561100447, "grad_norm": 0.5628980975480693, "learning_rate": 8.845901332132923e-09, "loss": 0.0094, "step": 15427 }, { "epoch": 9.870761356365962, "grad_norm": 0.6875848752782866, "learning_rate": 8.758976947887966e-09, "loss": 0.0103, "step": 15428 }, { "epoch": 9.871401151631478, "grad_norm": 0.4796485427885259, "learning_rate": 8.672481569759194e-09, "loss": 0.0064, "step": 15429 }, { "epoch": 9.872040946896993, "grad_norm": 0.635748289801569, "learning_rate": 8.586415201460308e-09, "loss": 0.0103, "step": 15430 }, { "epoch": 9.872680742162508, "grad_norm": 0.6419051794824254, "learning_rate": 8.500777846688347e-09, "loss": 0.0096, "step": 15431 }, { "epoch": 9.873320537428024, "grad_norm": 0.5633839859619568, "learning_rate": 8.415569509119258e-09, "loss": 0.0093, "step": 15432 }, { "epoch": 9.873960332693539, "grad_norm": 0.5951008665133054, "learning_rate": 8.330790192412341e-09, "loss": 0.009, "step": 15433 }, { "epoch": 9.874600127959052, "grad_norm": 0.7310437024236746, "learning_rate": 8.246439900206904e-09, "loss": 0.0081, "step": 15434 }, { "epoch": 9.875239923224568, "grad_norm": 0.539015776708812, "learning_rate": 8.162518636125605e-09, "loss": 0.0082, "step": 15435 }, { "epoch": 9.875879718490083, "grad_norm": 0.506800303529163, "learning_rate": 8.079026403771117e-09, "loss": 0.0067, "step": 15436 }, { "epoch": 9.876519513755598, "grad_norm": 0.6042021503297386, "learning_rate": 7.995963206729462e-09, "loss": 0.0088, "step": 15437 }, { "epoch": 9.877159309021113, "grad_norm": 0.714268051780581, "learning_rate": 7.913329048566676e-09, "loss": 0.0099, "step": 15438 }, { "epoch": 9.877799104286629, "grad_norm": 0.6863776817552997, "learning_rate": 7.83112393283103e-09, "loss": 0.0103, "step": 15439 }, { "epoch": 9.878438899552144, "grad_norm": 0.6329812195993315, "learning_rate": 7.749347863053035e-09, "loss": 0.0108, "step": 15440 }, { "epoch": 9.879078694817657, "grad_norm": 0.5913185911448791, "learning_rate": 7.668000842742108e-09, "loss": 0.0095, "step": 15441 }, { "epoch": 9.879718490083173, "grad_norm": 0.5770031357686328, "learning_rate": 7.587082875393225e-09, "loss": 0.0087, "step": 15442 }, { "epoch": 9.880358285348688, "grad_norm": 0.6458248959965803, "learning_rate": 7.506593964479169e-09, "loss": 0.0099, "step": 15443 }, { "epoch": 9.880998080614203, "grad_norm": 0.4745901161885508, "learning_rate": 7.4265341134571736e-09, "loss": 0.0061, "step": 15444 }, { "epoch": 9.881637875879719, "grad_norm": 0.5512029093605966, "learning_rate": 7.3469033257633774e-09, "loss": 0.0084, "step": 15445 }, { "epoch": 9.882277671145234, "grad_norm": 0.5052858841375725, "learning_rate": 7.267701604818378e-09, "loss": 0.0072, "step": 15446 }, { "epoch": 9.882917466410749, "grad_norm": 0.6305375539374457, "learning_rate": 7.188928954022789e-09, "loss": 0.0096, "step": 15447 }, { "epoch": 9.883557261676264, "grad_norm": 0.5698745090811397, "learning_rate": 7.11058537675946e-09, "loss": 0.0085, "step": 15448 }, { "epoch": 9.884197056941778, "grad_norm": 0.5720560059093864, "learning_rate": 7.032670876390146e-09, "loss": 0.0094, "step": 15449 }, { "epoch": 9.884836852207293, "grad_norm": 0.7361383296995763, "learning_rate": 6.9551854562632805e-09, "loss": 0.01, "step": 15450 }, { "epoch": 9.885476647472808, "grad_norm": 0.5762121208916474, "learning_rate": 6.87812911970398e-09, "loss": 0.0077, "step": 15451 }, { "epoch": 9.886116442738324, "grad_norm": 0.5799536215056685, "learning_rate": 6.80150187002071e-09, "loss": 0.0087, "step": 15452 }, { "epoch": 9.886756238003839, "grad_norm": 0.5374640079484849, "learning_rate": 6.725303710505282e-09, "loss": 0.0072, "step": 15453 }, { "epoch": 9.887396033269354, "grad_norm": 0.605756615362287, "learning_rate": 6.649534644428412e-09, "loss": 0.0105, "step": 15454 }, { "epoch": 9.88803582853487, "grad_norm": 0.5842120466102565, "learning_rate": 6.574194675044165e-09, "loss": 0.0094, "step": 15455 }, { "epoch": 9.888675623800385, "grad_norm": 0.45432335971393417, "learning_rate": 6.49928380558662e-09, "loss": 0.006, "step": 15456 }, { "epoch": 9.889315419065898, "grad_norm": 0.6135234837712439, "learning_rate": 6.424802039272093e-09, "loss": 0.011, "step": 15457 }, { "epoch": 9.889955214331414, "grad_norm": 0.6210147711256585, "learning_rate": 6.350749379300247e-09, "loss": 0.0078, "step": 15458 }, { "epoch": 9.890595009596929, "grad_norm": 0.5263069367637947, "learning_rate": 6.277125828850761e-09, "loss": 0.0073, "step": 15459 }, { "epoch": 9.891234804862444, "grad_norm": 0.6094948511605663, "learning_rate": 6.203931391082219e-09, "loss": 0.0092, "step": 15460 }, { "epoch": 9.89187460012796, "grad_norm": 0.5418800942381746, "learning_rate": 6.131166069139882e-09, "loss": 0.0081, "step": 15461 }, { "epoch": 9.892514395393475, "grad_norm": 0.5787600999186523, "learning_rate": 6.05882986614792e-09, "loss": 0.0081, "step": 15462 }, { "epoch": 9.89315419065899, "grad_norm": 0.5929277983749207, "learning_rate": 5.986922785211624e-09, "loss": 0.009, "step": 15463 }, { "epoch": 9.893793985924503, "grad_norm": 0.5511760469633614, "learning_rate": 5.915444829419636e-09, "loss": 0.0076, "step": 15464 }, { "epoch": 9.894433781190019, "grad_norm": 0.5303268936031188, "learning_rate": 5.844396001839503e-09, "loss": 0.0076, "step": 15465 }, { "epoch": 9.895073576455534, "grad_norm": 0.6115933995056403, "learning_rate": 5.773776305523226e-09, "loss": 0.0087, "step": 15466 }, { "epoch": 9.89571337172105, "grad_norm": 0.5793614757926606, "learning_rate": 5.703585743502826e-09, "loss": 0.01, "step": 15467 }, { "epoch": 9.896353166986565, "grad_norm": 0.5758962861111062, "learning_rate": 5.633824318791448e-09, "loss": 0.0089, "step": 15468 }, { "epoch": 9.89699296225208, "grad_norm": 0.5793002801537831, "learning_rate": 5.564492034385582e-09, "loss": 0.0082, "step": 15469 }, { "epoch": 9.897632757517595, "grad_norm": 0.5275842291336476, "learning_rate": 5.495588893260628e-09, "loss": 0.0073, "step": 15470 }, { "epoch": 9.898272552783109, "grad_norm": 0.6322947210050177, "learning_rate": 5.42711489837755e-09, "loss": 0.0104, "step": 15471 }, { "epoch": 9.898912348048624, "grad_norm": 0.5861229203741358, "learning_rate": 5.359070052673998e-09, "loss": 0.0087, "step": 15472 }, { "epoch": 9.89955214331414, "grad_norm": 0.5924880804167568, "learning_rate": 5.2914543590743e-09, "loss": 0.0091, "step": 15473 }, { "epoch": 9.900191938579654, "grad_norm": 0.6340447579165479, "learning_rate": 5.224267820479467e-09, "loss": 0.0106, "step": 15474 }, { "epoch": 9.90083173384517, "grad_norm": 0.5562276050967113, "learning_rate": 5.157510439774971e-09, "loss": 0.0076, "step": 15475 }, { "epoch": 9.901471529110685, "grad_norm": 0.5484687678405079, "learning_rate": 5.091182219827406e-09, "loss": 0.0076, "step": 15476 }, { "epoch": 9.9021113243762, "grad_norm": 0.5531477528641955, "learning_rate": 5.025283163485606e-09, "loss": 0.009, "step": 15477 }, { "epoch": 9.902751119641714, "grad_norm": 0.5409111795232151, "learning_rate": 4.959813273578418e-09, "loss": 0.0075, "step": 15478 }, { "epoch": 9.903390914907229, "grad_norm": 0.572800037094327, "learning_rate": 4.894772552916926e-09, "loss": 0.0066, "step": 15479 }, { "epoch": 9.904030710172744, "grad_norm": 0.5550917606909246, "learning_rate": 4.830161004294454e-09, "loss": 0.0078, "step": 15480 }, { "epoch": 9.90467050543826, "grad_norm": 0.6744763762128071, "learning_rate": 4.765978630484336e-09, "loss": 0.009, "step": 15481 }, { "epoch": 9.905310300703775, "grad_norm": 0.5421253818154532, "learning_rate": 4.702225434242147e-09, "loss": 0.0087, "step": 15482 }, { "epoch": 9.90595009596929, "grad_norm": 0.6675661139961242, "learning_rate": 4.6389014183079176e-09, "loss": 0.01, "step": 15483 }, { "epoch": 9.906589891234805, "grad_norm": 0.6001821835294916, "learning_rate": 4.5760065853972524e-09, "loss": 0.0097, "step": 15484 }, { "epoch": 9.90722968650032, "grad_norm": 0.4682678512856688, "learning_rate": 4.513540938213546e-09, "loss": 0.0064, "step": 15485 }, { "epoch": 9.907869481765834, "grad_norm": 0.569487397110264, "learning_rate": 4.451504479436875e-09, "loss": 0.0084, "step": 15486 }, { "epoch": 9.90850927703135, "grad_norm": 0.593970099566877, "learning_rate": 4.3898972117328856e-09, "loss": 0.0085, "step": 15487 }, { "epoch": 9.909149072296865, "grad_norm": 0.580502154153087, "learning_rate": 4.328719137745019e-09, "loss": 0.0086, "step": 15488 }, { "epoch": 9.90978886756238, "grad_norm": 0.6126868913304542, "learning_rate": 4.267970260101173e-09, "loss": 0.0088, "step": 15489 }, { "epoch": 9.910428662827895, "grad_norm": 0.6276303889078675, "learning_rate": 4.207650581410372e-09, "loss": 0.0104, "step": 15490 }, { "epoch": 9.91106845809341, "grad_norm": 0.6264594055393878, "learning_rate": 4.147760104261656e-09, "loss": 0.01, "step": 15491 }, { "epoch": 9.911708253358926, "grad_norm": 0.6247932427346512, "learning_rate": 4.088298831226301e-09, "loss": 0.0088, "step": 15492 }, { "epoch": 9.912348048624441, "grad_norm": 0.5148598573338364, "learning_rate": 4.02926676485782e-09, "loss": 0.0081, "step": 15493 }, { "epoch": 9.912987843889955, "grad_norm": 0.5261036808562666, "learning_rate": 3.9706639076919625e-09, "loss": 0.0078, "step": 15494 }, { "epoch": 9.91362763915547, "grad_norm": 0.5603038681780755, "learning_rate": 3.9124902622444945e-09, "loss": 0.0077, "step": 15495 }, { "epoch": 9.914267434420985, "grad_norm": 0.5624117832401555, "learning_rate": 3.854745831012308e-09, "loss": 0.0083, "step": 15496 }, { "epoch": 9.9149072296865, "grad_norm": 0.5316469233051695, "learning_rate": 3.7974306164756395e-09, "loss": 0.0079, "step": 15497 }, { "epoch": 9.915547024952016, "grad_norm": 0.5220061085610389, "learning_rate": 3.740544621095854e-09, "loss": 0.0073, "step": 15498 }, { "epoch": 9.916186820217531, "grad_norm": 0.5522753588641336, "learning_rate": 3.6840878473143327e-09, "loss": 0.0081, "step": 15499 }, { "epoch": 9.916826615483046, "grad_norm": 0.5688919719898629, "learning_rate": 3.6280602975569122e-09, "loss": 0.0077, "step": 15500 }, { "epoch": 9.91746641074856, "grad_norm": 0.6112910657379886, "learning_rate": 3.572461974228336e-09, "loss": 0.0093, "step": 15501 }, { "epoch": 9.918106206014075, "grad_norm": 0.5534604909262311, "learning_rate": 3.517292879715584e-09, "loss": 0.008, "step": 15502 }, { "epoch": 9.91874600127959, "grad_norm": 0.46872512989722004, "learning_rate": 3.4625530163878706e-09, "loss": 0.0069, "step": 15503 }, { "epoch": 9.919385796545106, "grad_norm": 0.5887852132020394, "learning_rate": 3.4082423865955396e-09, "loss": 0.0088, "step": 15504 }, { "epoch": 9.92002559181062, "grad_norm": 0.5974094523204084, "learning_rate": 3.3543609926711685e-09, "loss": 0.0083, "step": 15505 }, { "epoch": 9.920665387076136, "grad_norm": 0.6010342273442747, "learning_rate": 3.300908836927352e-09, "loss": 0.0089, "step": 15506 }, { "epoch": 9.921305182341651, "grad_norm": 0.6048660604507815, "learning_rate": 3.2478859216611425e-09, "loss": 0.0088, "step": 15507 }, { "epoch": 9.921944977607165, "grad_norm": 0.5527021052060996, "learning_rate": 3.1952922491462755e-09, "loss": 0.0079, "step": 15508 }, { "epoch": 9.92258477287268, "grad_norm": 0.5016997792525612, "learning_rate": 3.143127821643166e-09, "loss": 0.0066, "step": 15509 }, { "epoch": 9.923224568138195, "grad_norm": 0.5296895174815418, "learning_rate": 3.091392641391133e-09, "loss": 0.0079, "step": 15510 }, { "epoch": 9.92386436340371, "grad_norm": 0.5313772797809121, "learning_rate": 3.0400867106117336e-09, "loss": 0.0079, "step": 15511 }, { "epoch": 9.924504158669226, "grad_norm": 0.5548725777365039, "learning_rate": 2.98921003150765e-09, "loss": 0.0079, "step": 15512 }, { "epoch": 9.925143953934741, "grad_norm": 0.4960593151966375, "learning_rate": 2.9387626062638007e-09, "loss": 0.0068, "step": 15513 }, { "epoch": 9.925783749200257, "grad_norm": 0.5327446619572542, "learning_rate": 2.8887444370462315e-09, "loss": 0.0068, "step": 15514 }, { "epoch": 9.92642354446577, "grad_norm": 0.5602525261365131, "learning_rate": 2.839155526001003e-09, "loss": 0.0074, "step": 15515 }, { "epoch": 9.927063339731285, "grad_norm": 0.6460476302978597, "learning_rate": 2.7899958752608535e-09, "loss": 0.0078, "step": 15516 }, { "epoch": 9.9277031349968, "grad_norm": 0.5627686968634378, "learning_rate": 2.7412654869329867e-09, "loss": 0.0082, "step": 15517 }, { "epoch": 9.928342930262316, "grad_norm": 0.6244544531441836, "learning_rate": 2.6929643631123938e-09, "loss": 0.0095, "step": 15518 }, { "epoch": 9.928982725527831, "grad_norm": 0.5646953938403232, "learning_rate": 2.6450925058718603e-09, "loss": 0.0087, "step": 15519 }, { "epoch": 9.929622520793346, "grad_norm": 0.5125365347435291, "learning_rate": 2.5976499172675198e-09, "loss": 0.0078, "step": 15520 }, { "epoch": 9.930262316058862, "grad_norm": 0.524673404206367, "learning_rate": 2.550636599335521e-09, "loss": 0.0073, "step": 15521 }, { "epoch": 9.930902111324377, "grad_norm": 0.6579455475920258, "learning_rate": 2.50405255409536e-09, "loss": 0.0098, "step": 15522 }, { "epoch": 9.93154190658989, "grad_norm": 0.6061864053819026, "learning_rate": 2.457897783547658e-09, "loss": 0.0097, "step": 15523 }, { "epoch": 9.932181701855406, "grad_norm": 0.6472933670050031, "learning_rate": 2.4121722896730536e-09, "loss": 0.0106, "step": 15524 }, { "epoch": 9.932821497120921, "grad_norm": 0.4646374653922105, "learning_rate": 2.366876074435531e-09, "loss": 0.0065, "step": 15525 }, { "epoch": 9.933461292386436, "grad_norm": 0.7113044628193949, "learning_rate": 2.3220091397790913e-09, "loss": 0.0083, "step": 15526 }, { "epoch": 9.934101087651952, "grad_norm": 0.5965500167817275, "learning_rate": 2.2775714876321908e-09, "loss": 0.0087, "step": 15527 }, { "epoch": 9.934740882917467, "grad_norm": 0.6497664350302655, "learning_rate": 2.2335631199010834e-09, "loss": 0.0091, "step": 15528 }, { "epoch": 9.935380678182982, "grad_norm": 0.4901998531143478, "learning_rate": 2.189984038476478e-09, "loss": 0.0068, "step": 15529 }, { "epoch": 9.936020473448497, "grad_norm": 0.5477949033134695, "learning_rate": 2.1468342452291012e-09, "loss": 0.0068, "step": 15530 }, { "epoch": 9.93666026871401, "grad_norm": 0.5998467381517525, "learning_rate": 2.1041137420119152e-09, "loss": 0.0083, "step": 15531 }, { "epoch": 9.937300063979526, "grad_norm": 0.6082729517357071, "learning_rate": 2.061822530659008e-09, "loss": 0.0081, "step": 15532 }, { "epoch": 9.937939859245041, "grad_norm": 0.6581141172679477, "learning_rate": 2.0199606129867045e-09, "loss": 0.0102, "step": 15533 }, { "epoch": 9.938579654510557, "grad_norm": 0.537933998827196, "learning_rate": 1.978527990792456e-09, "loss": 0.0074, "step": 15534 }, { "epoch": 9.939219449776072, "grad_norm": 0.5235267494495034, "learning_rate": 1.9375246658548396e-09, "loss": 0.0071, "step": 15535 }, { "epoch": 9.939859245041587, "grad_norm": 0.5275355730140052, "learning_rate": 1.8969506399346694e-09, "loss": 0.0076, "step": 15536 }, { "epoch": 9.940499040307103, "grad_norm": 0.5685678438825946, "learning_rate": 1.8568059147738848e-09, "loss": 0.0085, "step": 15537 }, { "epoch": 9.941138835572616, "grad_norm": 0.5326666829863634, "learning_rate": 1.8170904920977727e-09, "loss": 0.0087, "step": 15538 }, { "epoch": 9.941778630838131, "grad_norm": 0.45837700925902103, "learning_rate": 1.7778043736094152e-09, "loss": 0.0068, "step": 15539 }, { "epoch": 9.942418426103647, "grad_norm": 0.6548839970900482, "learning_rate": 1.7389475609963512e-09, "loss": 0.01, "step": 15540 }, { "epoch": 9.943058221369162, "grad_norm": 0.46960400481069253, "learning_rate": 1.7005200559283564e-09, "loss": 0.0074, "step": 15541 }, { "epoch": 9.943698016634677, "grad_norm": 0.5575347881820678, "learning_rate": 1.6625218600541114e-09, "loss": 0.0078, "step": 15542 }, { "epoch": 9.944337811900192, "grad_norm": 0.53765396732424, "learning_rate": 1.6249529750056447e-09, "loss": 0.0079, "step": 15543 }, { "epoch": 9.944977607165708, "grad_norm": 0.5580616833900532, "learning_rate": 1.5878134023949999e-09, "loss": 0.01, "step": 15544 }, { "epoch": 9.945617402431221, "grad_norm": 0.6746626866658983, "learning_rate": 1.5511031438186775e-09, "loss": 0.0102, "step": 15545 }, { "epoch": 9.946257197696736, "grad_norm": 0.6385267635325879, "learning_rate": 1.5148222008520842e-09, "loss": 0.0102, "step": 15546 }, { "epoch": 9.946896992962252, "grad_norm": 0.6595820083301099, "learning_rate": 1.4789705750528627e-09, "loss": 0.011, "step": 15547 }, { "epoch": 9.947536788227767, "grad_norm": 0.5257333093491212, "learning_rate": 1.4435482679608925e-09, "loss": 0.0078, "step": 15548 }, { "epoch": 9.948176583493282, "grad_norm": 0.48284028437592097, "learning_rate": 1.4085552810971792e-09, "loss": 0.0069, "step": 15549 }, { "epoch": 9.948816378758798, "grad_norm": 0.5649465639988944, "learning_rate": 1.3739916159627441e-09, "loss": 0.0098, "step": 15550 }, { "epoch": 9.949456174024313, "grad_norm": 0.6297410193233138, "learning_rate": 1.3398572740441762e-09, "loss": 0.0095, "step": 15551 }, { "epoch": 9.950095969289826, "grad_norm": 0.568414742549358, "learning_rate": 1.3061522568058592e-09, "loss": 0.0081, "step": 15552 }, { "epoch": 9.950735764555342, "grad_norm": 0.4655266331070608, "learning_rate": 1.2728765656944141e-09, "loss": 0.0058, "step": 15553 }, { "epoch": 9.951375559820857, "grad_norm": 0.5550708840676555, "learning_rate": 1.240030202139808e-09, "loss": 0.0077, "step": 15554 }, { "epoch": 9.952015355086372, "grad_norm": 0.6142251113895677, "learning_rate": 1.2076131675520241e-09, "loss": 0.0095, "step": 15555 }, { "epoch": 9.952655150351887, "grad_norm": 0.5271041295992673, "learning_rate": 1.175625463322172e-09, "loss": 0.0082, "step": 15556 }, { "epoch": 9.953294945617403, "grad_norm": 0.669359080893446, "learning_rate": 1.1440670908247076e-09, "loss": 0.0081, "step": 15557 }, { "epoch": 9.953934740882918, "grad_norm": 0.5996283436747168, "learning_rate": 1.112938051414103e-09, "loss": 0.0085, "step": 15558 }, { "epoch": 9.954574536148433, "grad_norm": 0.5822153509242343, "learning_rate": 1.0822383464281771e-09, "loss": 0.0081, "step": 15559 }, { "epoch": 9.955214331413947, "grad_norm": 0.6162791881276422, "learning_rate": 1.0519679771836545e-09, "loss": 0.0093, "step": 15560 }, { "epoch": 9.955854126679462, "grad_norm": 0.5329403725242452, "learning_rate": 1.022126944980606e-09, "loss": 0.0065, "step": 15561 }, { "epoch": 9.956493921944977, "grad_norm": 0.5459883807344086, "learning_rate": 9.927152511013393e-10, "loss": 0.0078, "step": 15562 }, { "epoch": 9.957133717210493, "grad_norm": 0.5444445218030178, "learning_rate": 9.637328968081783e-10, "loss": 0.0086, "step": 15563 }, { "epoch": 9.957773512476008, "grad_norm": 0.5889948188778839, "learning_rate": 9.351798833445724e-10, "loss": 0.0085, "step": 15564 }, { "epoch": 9.958413307741523, "grad_norm": 0.5723395297952663, "learning_rate": 9.070562119373182e-10, "loss": 0.008, "step": 15565 }, { "epoch": 9.959053103007038, "grad_norm": 0.6156125462830415, "learning_rate": 8.793618837954487e-10, "loss": 0.0084, "step": 15566 }, { "epoch": 9.959692898272554, "grad_norm": 0.5605091197234707, "learning_rate": 8.520969001057922e-10, "loss": 0.008, "step": 15567 }, { "epoch": 9.960332693538067, "grad_norm": 0.5404698500294672, "learning_rate": 8.252612620396339e-10, "loss": 0.0074, "step": 15568 }, { "epoch": 9.960972488803582, "grad_norm": 0.587582123286707, "learning_rate": 7.988549707493853e-10, "loss": 0.0068, "step": 15569 }, { "epoch": 9.961612284069098, "grad_norm": 0.6418059516623089, "learning_rate": 7.728780273696945e-10, "loss": 0.0098, "step": 15570 }, { "epoch": 9.962252079334613, "grad_norm": 0.6125608276491145, "learning_rate": 7.473304330152253e-10, "loss": 0.011, "step": 15571 }, { "epoch": 9.962891874600128, "grad_norm": 0.600008385134608, "learning_rate": 7.222121887839883e-10, "loss": 0.0098, "step": 15572 }, { "epoch": 9.963531669865644, "grad_norm": 0.638127326966034, "learning_rate": 6.975232957528999e-10, "loss": 0.0114, "step": 15573 }, { "epoch": 9.964171465131159, "grad_norm": 0.559274992227474, "learning_rate": 6.732637549822229e-10, "loss": 0.0089, "step": 15574 }, { "epoch": 9.964811260396672, "grad_norm": 0.6381747769569175, "learning_rate": 6.49433567514457e-10, "loss": 0.0085, "step": 15575 }, { "epoch": 9.965451055662188, "grad_norm": 0.6150597390573417, "learning_rate": 6.260327343732275e-10, "loss": 0.0081, "step": 15576 }, { "epoch": 9.966090850927703, "grad_norm": 0.5651646989347442, "learning_rate": 6.030612565621763e-10, "loss": 0.0084, "step": 15577 }, { "epoch": 9.966730646193218, "grad_norm": 0.5772021111912864, "learning_rate": 5.805191350682915e-10, "loss": 0.0083, "step": 15578 }, { "epoch": 9.967370441458733, "grad_norm": 0.8382929295809602, "learning_rate": 5.584063708585774e-10, "loss": 0.0125, "step": 15579 }, { "epoch": 9.968010236724249, "grad_norm": 0.6213855938956548, "learning_rate": 5.367229648844951e-10, "loss": 0.0094, "step": 15580 }, { "epoch": 9.968650031989764, "grad_norm": 0.5715660620214184, "learning_rate": 5.154689180753014e-10, "loss": 0.0087, "step": 15581 }, { "epoch": 9.969289827255277, "grad_norm": 0.50256691435645, "learning_rate": 4.946442313435995e-10, "loss": 0.0075, "step": 15582 }, { "epoch": 9.969929622520793, "grad_norm": 0.5576383214717878, "learning_rate": 4.742489055853394e-10, "loss": 0.0069, "step": 15583 }, { "epoch": 9.970569417786308, "grad_norm": 0.5243982733856715, "learning_rate": 4.542829416742667e-10, "loss": 0.0071, "step": 15584 }, { "epoch": 9.971209213051823, "grad_norm": 0.5558023070343436, "learning_rate": 4.3474634046858366e-10, "loss": 0.0096, "step": 15585 }, { "epoch": 9.971849008317339, "grad_norm": 0.6152295345120773, "learning_rate": 4.156391028065088e-10, "loss": 0.0096, "step": 15586 }, { "epoch": 9.972488803582854, "grad_norm": 0.6432647970463403, "learning_rate": 3.9696122950960705e-10, "loss": 0.0113, "step": 15587 }, { "epoch": 9.973128598848369, "grad_norm": 0.50766450187642, "learning_rate": 3.787127213794595e-10, "loss": 0.0074, "step": 15588 }, { "epoch": 9.973768394113884, "grad_norm": 0.41922497125348546, "learning_rate": 3.608935791987733e-10, "loss": 0.0059, "step": 15589 }, { "epoch": 9.974408189379398, "grad_norm": 0.5687384944008653, "learning_rate": 3.435038037336025e-10, "loss": 0.0093, "step": 15590 }, { "epoch": 9.975047984644913, "grad_norm": 0.6431124108964293, "learning_rate": 3.26543395731127e-10, "loss": 0.0101, "step": 15591 }, { "epoch": 9.975687779910428, "grad_norm": 2.4370969211494447, "learning_rate": 3.1001235591854307e-10, "loss": 0.0097, "step": 15592 }, { "epoch": 9.976327575175944, "grad_norm": 0.5533477141618093, "learning_rate": 2.9391068500528307e-10, "loss": 0.0065, "step": 15593 }, { "epoch": 9.976967370441459, "grad_norm": 0.5909385759158237, "learning_rate": 2.7823838368412626e-10, "loss": 0.0092, "step": 15594 }, { "epoch": 9.977607165706974, "grad_norm": 0.5667975620051993, "learning_rate": 2.629954526267575e-10, "loss": 0.0079, "step": 15595 }, { "epoch": 9.97824696097249, "grad_norm": 0.48133767688857665, "learning_rate": 2.4818189248820846e-10, "loss": 0.0064, "step": 15596 }, { "epoch": 9.978886756238003, "grad_norm": 0.6058014555951932, "learning_rate": 2.3379770390463686e-10, "loss": 0.0081, "step": 15597 }, { "epoch": 9.979526551503518, "grad_norm": 0.6646397887333915, "learning_rate": 2.1984288749332672e-10, "loss": 0.0065, "step": 15598 }, { "epoch": 9.980166346769034, "grad_norm": 0.6720137353786122, "learning_rate": 2.0631744385379848e-10, "loss": 0.0096, "step": 15599 }, { "epoch": 9.980806142034549, "grad_norm": 0.697716691441665, "learning_rate": 1.9322137356780902e-10, "loss": 0.0095, "step": 15600 }, { "epoch": 9.981445937300064, "grad_norm": 0.6200647403492834, "learning_rate": 1.8055467719602094e-10, "loss": 0.0096, "step": 15601 }, { "epoch": 9.98208573256558, "grad_norm": 0.6633244925090603, "learning_rate": 1.6831735528244353e-10, "loss": 0.0097, "step": 15602 }, { "epoch": 9.982725527831095, "grad_norm": 0.5334120381898402, "learning_rate": 1.565094083533225e-10, "loss": 0.0078, "step": 15603 }, { "epoch": 9.98336532309661, "grad_norm": 0.5117332115054273, "learning_rate": 1.451308369149196e-10, "loss": 0.0074, "step": 15604 }, { "epoch": 9.984005118362123, "grad_norm": 0.6506722208754839, "learning_rate": 1.341816414568431e-10, "loss": 0.0104, "step": 15605 }, { "epoch": 9.984644913627639, "grad_norm": 0.47667076051261853, "learning_rate": 1.2366182244871738e-10, "loss": 0.0075, "step": 15606 }, { "epoch": 9.985284708893154, "grad_norm": 0.616869152574214, "learning_rate": 1.13571380341293e-10, "loss": 0.0099, "step": 15607 }, { "epoch": 9.98592450415867, "grad_norm": 0.5115388371512788, "learning_rate": 1.0391031556977737e-10, "loss": 0.0077, "step": 15608 }, { "epoch": 9.986564299424185, "grad_norm": 0.6506503308607133, "learning_rate": 9.467862854717347e-11, "loss": 0.0112, "step": 15609 }, { "epoch": 9.9872040946897, "grad_norm": 0.6460175807803619, "learning_rate": 8.587631967094112e-11, "loss": 0.0081, "step": 15610 }, { "epoch": 9.987843889955215, "grad_norm": 0.4817331173096211, "learning_rate": 7.750338931855616e-11, "loss": 0.006, "step": 15611 }, { "epoch": 9.988483685220729, "grad_norm": 0.5841851563728601, "learning_rate": 6.955983784973085e-11, "loss": 0.0086, "step": 15612 }, { "epoch": 9.989123480486244, "grad_norm": 0.5914948168473266, "learning_rate": 6.204566560530368e-11, "loss": 0.009, "step": 15613 }, { "epoch": 9.98976327575176, "grad_norm": 0.5907624804267704, "learning_rate": 5.49608729083495e-11, "loss": 0.0086, "step": 15614 }, { "epoch": 9.990403071017274, "grad_norm": 0.9532438500215769, "learning_rate": 4.830546006306947e-11, "loss": 0.011, "step": 15615 }, { "epoch": 9.99104286628279, "grad_norm": 0.5278564501913833, "learning_rate": 4.207942735479087e-11, "loss": 0.0084, "step": 15616 }, { "epoch": 9.991682661548305, "grad_norm": 0.5652603035440582, "learning_rate": 3.628277505107747e-11, "loss": 0.0083, "step": 15617 }, { "epoch": 9.99232245681382, "grad_norm": 0.5702346516318845, "learning_rate": 3.091550340061922e-11, "loss": 0.0093, "step": 15618 }, { "epoch": 9.992962252079334, "grad_norm": 0.5658945568277542, "learning_rate": 2.597761263434251e-11, "loss": 0.0076, "step": 15619 }, { "epoch": 9.993602047344849, "grad_norm": 0.5843517332138803, "learning_rate": 2.146910296429994e-11, "loss": 0.0096, "step": 15620 }, { "epoch": 9.994241842610364, "grad_norm": 0.5573340451691949, "learning_rate": 1.7389974582560088e-11, "loss": 0.0081, "step": 15621 }, { "epoch": 9.99488163787588, "grad_norm": 0.565914352296651, "learning_rate": 1.3740227666758643e-11, "loss": 0.0085, "step": 15622 }, { "epoch": 9.995521433141395, "grad_norm": 0.5175180620627459, "learning_rate": 1.0519862372326827e-11, "loss": 0.0078, "step": 15623 }, { "epoch": 9.99616122840691, "grad_norm": 0.522381281334086, "learning_rate": 7.728878836932297e-12, "loss": 0.0077, "step": 15624 }, { "epoch": 9.996801023672425, "grad_norm": 0.5940281249520306, "learning_rate": 5.367277181589359e-12, "loss": 0.0085, "step": 15625 }, { "epoch": 9.99744081893794, "grad_norm": 0.6011964908500335, "learning_rate": 3.435057506218087e-12, "loss": 0.0101, "step": 15626 }, { "epoch": 9.998080614203454, "grad_norm": 0.5527220860948949, "learning_rate": 1.932219896305654e-12, "loss": 0.0074, "step": 15627 }, { "epoch": 9.99872040946897, "grad_norm": 0.6012080981976905, "learning_rate": 8.587644140245488e-13, "loss": 0.0097, "step": 15628 }, { "epoch": 9.999360204734485, "grad_norm": 0.5245784624821092, "learning_rate": 2.1469110600413902e-13, "loss": 0.008, "step": 15629 }, { "epoch": 10.0, "grad_norm": 0.4256047471940505, "learning_rate": 0.0, "loss": 0.0057, "step": 15630 }, { "epoch": 10.0, "step": 15630, "total_flos": 1118754649055232.0, "train_loss": 0.003334951470948646, "train_runtime": 11361.004, "train_samples_per_second": 176.041, "train_steps_per_second": 1.376 } ], "logging_steps": 1.0, "max_steps": 15630, "num_input_tokens_seen": 0, "num_train_epochs": 10, "save_steps": 50000.0, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 1118754649055232.0, "train_batch_size": 16, "trial_name": null, "trial_params": null }